diff options
Diffstat (limited to 'target/linux/patches/4.1.13')
| -rw-r--r-- | target/linux/patches/4.1.13/aufs.patch | 35215 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/cleankernel.patch | 11 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/cris-header.patch | 12 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/initramfs-nosizelimit.patch | 57 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/realtime.patch | 27435 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/regmap-default-on.patch | 17 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/remove-warn.patch | 11 | ||||
| -rw-r--r-- | target/linux/patches/4.1.13/startup.patch | 37 |
8 files changed, 62795 insertions, 0 deletions
diff --git a/target/linux/patches/4.1.13/aufs.patch b/target/linux/patches/4.1.13/aufs.patch new file mode 100644 index 000000000..749c90989 --- /dev/null +++ b/target/linux/patches/4.1.13/aufs.patch @@ -0,0 +1,35215 @@ +diff -Nur linux-4.1.10.orig/Documentation/ABI/testing/debugfs-aufs linux-4.1.10/Documentation/ABI/testing/debugfs-aufs +--- linux-4.1.10.orig/Documentation/ABI/testing/debugfs-aufs 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/ABI/testing/debugfs-aufs 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,50 @@ ++What: /debug/aufs/si_<id>/ ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ Under /debug/aufs, a directory named si_<id> is created ++ per aufs mount, where <id> is a unique id generated ++ internally. ++ ++What: /debug/aufs/si_<id>/plink ++Date: Apr 2013 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It has three lines and shows the information about the ++ pseudo-link. The first line is a single number ++ representing a number of buckets. The second line is a ++ number of pseudo-links per buckets (separated by a ++ blank). The last line is a single number representing a ++ total number of psedo-links. ++ When the aufs mount option 'noplink' is specified, it ++ will show "1\n0\n0\n". ++ ++What: /debug/aufs/si_<id>/xib ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the consumed blocks by xib (External Inode Number ++ Bitmap), its block size and file size. ++ When the aufs mount option 'noxino' is specified, it ++ will be empty. About XINO files, see the aufs manual. ++ ++What: /debug/aufs/si_<id>/xino0, xino1 ... xinoN ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the consumed blocks by xino (External Inode Number ++ Translation Table), its link count, block size and file ++ size. ++ When the aufs mount option 'noxino' is specified, it ++ will be empty. About XINO files, see the aufs manual. ++ ++What: /debug/aufs/si_<id>/xigen ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the consumed blocks by xigen (External Inode ++ Generation Table), its block size and file size. ++ If CONFIG_AUFS_EXPORT is disabled, this entry will not ++ be created. ++ When the aufs mount option 'noxino' is specified, it ++ will be empty. About XINO files, see the aufs manual. +diff -Nur linux-4.1.10.orig/Documentation/ABI/testing/sysfs-aufs linux-4.1.10/Documentation/ABI/testing/sysfs-aufs +--- linux-4.1.10.orig/Documentation/ABI/testing/sysfs-aufs 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/ABI/testing/sysfs-aufs 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,31 @@ ++What: /sys/fs/aufs/si_<id>/ ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ Under /sys/fs/aufs, a directory named si_<id> is created ++ per aufs mount, where <id> is a unique id generated ++ internally. ++ ++What: /sys/fs/aufs/si_<id>/br0, br1 ... brN ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the abolute path of a member directory (which ++ is called branch) in aufs, and its permission. ++ ++What: /sys/fs/aufs/si_<id>/brid0, brid1 ... bridN ++Date: July 2013 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the id of a member directory (which is called ++ branch) in aufs. ++ ++What: /sys/fs/aufs/si_<id>/xi_path ++Date: March 2009 ++Contact: J. R. Okajima <hooanon05g@gmail.com> ++Description: ++ It shows the abolute path of XINO (External Inode Number ++ Bitmap, Translation Table and Generation Table) file ++ even if it is the default path. ++ When the aufs mount option 'noxino' is specified, it ++ will be empty. About XINO files, see the aufs manual. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/01intro.txt linux-4.1.10/Documentation/filesystems/aufs/design/01intro.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/01intro.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/01intro.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,170 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Introduction ++---------------------------------------- ++ ++aufs [ei ju: ef es] | [a u f s] ++1. abbrev. for "advanced multi-layered unification filesystem". ++2. abbrev. for "another unionfs". ++3. abbrev. for "auf das" in German which means "on the" in English. ++ Ex. "Butter aufs Brot"(G) means "butter onto bread"(E). ++ But "Filesystem aufs Filesystem" is hard to understand. ++ ++AUFS is a filesystem with features: ++- multi layered stackable unification filesystem, the member directory ++ is called as a branch. ++- branch permission and attribute, 'readonly', 'real-readonly', ++ 'readwrite', 'whiteout-able', 'link-able whiteout', etc. and their ++ combination. ++- internal "file copy-on-write". ++- logical deletion, whiteout. ++- dynamic branch manipulation, adding, deleting and changing permission. ++- allow bypassing aufs, user's direct branch access. ++- external inode number translation table and bitmap which maintains the ++ persistent aufs inode number. ++- seekable directory, including NFS readdir. ++- file mapping, mmap and sharing pages. ++- pseudo-link, hardlink over branches. ++- loopback mounted filesystem as a branch. ++- several policies to select one among multiple writable branches. ++- revert a single systemcall when an error occurs in aufs. ++- and more... ++ ++ ++Multi Layered Stackable Unification Filesystem ++---------------------------------------------------------------------- ++Most people already knows what it is. ++It is a filesystem which unifies several directories and provides a ++merged single directory. When users access a file, the access will be ++passed/re-directed/converted (sorry, I am not sure which English word is ++correct) to the real file on the member filesystem. The member ++filesystem is called 'lower filesystem' or 'branch' and has a mode ++'readonly' and 'readwrite.' And the deletion for a file on the lower ++readonly branch is handled by creating 'whiteout' on the upper writable ++branch. ++ ++On LKML, there have been discussions about UnionMount (Jan Blunck, ++Bharata B Rao and Valerie Aurora) and Unionfs (Erez Zadok). They took ++different approaches to implement the merged-view. ++The former tries putting it into VFS, and the latter implements as a ++separate filesystem. ++(If I misunderstand about these implementations, please let me know and ++I shall correct it. Because it is a long time ago when I read their ++source files last time). ++ ++UnionMount's approach will be able to small, but may be hard to share ++branches between several UnionMount since the whiteout in it is ++implemented in the inode on branch filesystem and always ++shared. According to Bharata's post, readdir does not seems to be ++finished yet. ++There are several missing features known in this implementations such as ++- for users, the inode number may change silently. eg. copy-up. ++- link(2) may break by copy-up. ++- read(2) may get an obsoleted filedata (fstat(2) too). ++- fcntl(F_SETLK) may be broken by copy-up. ++- unnecessary copy-up may happen, for example mmap(MAP_PRIVATE) after ++ open(O_RDWR). ++ ++In linux-3.18, "overlay" filesystem (formerly known as "overlayfs") was ++merged into mainline. This is another implementation of UnionMount as a ++separated filesystem. All the limitations and known problems which ++UnionMount are equally inherited to "overlay" filesystem. ++ ++Unionfs has a longer history. When I started implementing a stackable ++filesystem (Aug 2005), it already existed. It has virtual super_block, ++inode, dentry and file objects and they have an array pointing lower ++same kind objects. After contributing many patches for Unionfs, I ++re-started my project AUFS (Jun 2006). ++ ++In AUFS, the structure of filesystem resembles to Unionfs, but I ++implemented my own ideas, approaches and enhancements and it became ++totally different one. ++ ++Comparing DM snapshot and fs based implementation ++- the number of bytes to be copied between devices is much smaller. ++- the type of filesystem must be one and only. ++- the fs must be writable, no readonly fs, even for the lower original ++ device. so the compression fs will not be usable. but if we use ++ loopback mount, we may address this issue. ++ for instance, ++ mount /cdrom/squashfs.img /sq ++ losetup /sq/ext2.img ++ losetup /somewhere/cow ++ dmsetup "snapshot /dev/loop0 /dev/loop1 ..." ++- it will be difficult (or needs more operations) to extract the ++ difference between the original device and COW. ++- DM snapshot-merge may help a lot when users try merging. in the ++ fs-layer union, users will use rsync(1). ++ ++You may want to read my old paper "Filesystems in LiveCD" ++(http://aufs.sourceforge.net/aufs2/report/sq/sq.pdf). ++ ++ ++Several characters/aspects/persona of aufs ++---------------------------------------------------------------------- ++ ++Aufs has several characters, aspects or persona. ++1. a filesystem, callee of VFS helper ++2. sub-VFS, caller of VFS helper for branches ++3. a virtual filesystem which maintains persistent inode number ++4. reader/writer of files on branches such like an application ++ ++1. Callee of VFS Helper ++As an ordinary linux filesystem, aufs is a callee of VFS. For instance, ++unlink(2) from an application reaches sys_unlink() kernel function and ++then vfs_unlink() is called. vfs_unlink() is one of VFS helper and it ++calls filesystem specific unlink operation. Actually aufs implements the ++unlink operation but it behaves like a redirector. ++ ++2. Caller of VFS Helper for Branches ++aufs_unlink() passes the unlink request to the branch filesystem as if ++it were called from VFS. So the called unlink operation of the branch ++filesystem acts as usual. As a caller of VFS helper, aufs should handle ++every necessary pre/post operation for the branch filesystem. ++- acquire the lock for the parent dir on a branch ++- lookup in a branch ++- revalidate dentry on a branch ++- mnt_want_write() for a branch ++- vfs_unlink() for a branch ++- mnt_drop_write() for a branch ++- release the lock on a branch ++ ++3. Persistent Inode Number ++One of the most important issue for a filesystem is to maintain inode ++numbers. This is particularly important to support exporting a ++filesystem via NFS. Aufs is a virtual filesystem which doesn't have a ++backend block device for its own. But some storage is necessary to ++keep and maintain the inode numbers. It may be a large space and may not ++suit to keep in memory. Aufs rents some space from its first writable ++branch filesystem (by default) and creates file(s) on it. These files ++are created by aufs internally and removed soon (currently) keeping ++opened. ++Note: Because these files are removed, they are totally gone after ++ unmounting aufs. It means the inode numbers are not persistent ++ across unmount or reboot. I have a plan to make them really ++ persistent which will be important for aufs on NFS server. ++ ++4. Read/Write Files Internally (copy-on-write) ++Because a branch can be readonly, when you write a file on it, aufs will ++"copy-up" it to the upper writable branch internally. And then write the ++originally requested thing to the file. Generally kernel doesn't ++open/read/write file actively. In aufs, even a single write may cause a ++internal "file copy". This behaviour is very similar to cp(1) command. ++ ++Some people may think it is better to pass such work to user space ++helper, instead of doing in kernel space. Actually I am still thinking ++about it. But currently I have implemented it in kernel space. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/02struct.txt linux-4.1.10/Documentation/filesystems/aufs/design/02struct.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/02struct.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/02struct.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,258 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Basic Aufs Internal Structure ++ ++Superblock/Inode/Dentry/File Objects ++---------------------------------------------------------------------- ++As like an ordinary filesystem, aufs has its own ++superblock/inode/dentry/file objects. All these objects have a ++dynamically allocated array and store the same kind of pointers to the ++lower filesystem, branch. ++For example, when you build a union with one readwrite branch and one ++readonly, mounted /au, /rw and /ro respectively. ++- /au = /rw + /ro ++- /ro/fileA exists but /rw/fileA ++ ++Aufs lookup operation finds /ro/fileA and gets dentry for that. These ++pointers are stored in a aufs dentry. The array in aufs dentry will be, ++- [0] = NULL (because /rw/fileA doesn't exist) ++- [1] = /ro/fileA ++ ++This style of an array is essentially same to the aufs ++superblock/inode/dentry/file objects. ++ ++Because aufs supports manipulating branches, ie. add/delete/change ++branches dynamically, these objects has its own generation. When ++branches are changed, the generation in aufs superblock is ++incremented. And a generation in other object are compared when it is ++accessed. When a generation in other objects are obsoleted, aufs ++refreshes the internal array. ++ ++ ++Superblock ++---------------------------------------------------------------------- ++Additionally aufs superblock has some data for policies to select one ++among multiple writable branches, XIB files, pseudo-links and kobject. ++See below in detail. ++About the policies which supports copy-down a directory, see ++wbr_policy.txt too. ++ ++ ++Branch and XINO(External Inode Number Translation Table) ++---------------------------------------------------------------------- ++Every branch has its own xino (external inode number translation table) ++file. The xino file is created and unlinked by aufs internally. When two ++members of a union exist on the same filesystem, they share the single ++xino file. ++The struct of a xino file is simple, just a sequence of aufs inode ++numbers which is indexed by the lower inode number. ++In the above sample, assume the inode number of /ro/fileA is i111 and ++aufs assigns the inode number i999 for fileA. Then aufs writes 999 as ++4(8) bytes at 111 * 4(8) bytes offset in the xino file. ++ ++When the inode numbers are not contiguous, the xino file will be sparse ++which has a hole in it and doesn't consume as much disk space as it ++might appear. If your branch filesystem consumes disk space for such ++holes, then you should specify 'xino=' option at mounting aufs. ++ ++Aufs has a mount option to free the disk blocks for such holes in XINO ++files on tmpfs or ramdisk. But it is not so effective actually. If you ++meet a problem of disk shortage due to XINO files, then you should try ++"tmpfs-ino.patch" (and "vfs-ino.patch" too) in aufs4-standalone.git. ++The patch localizes the assignment inumbers per tmpfs-mount and avoid ++the holes in XINO files. ++ ++Also a writable branch has three kinds of "whiteout bases". All these ++are existed when the branch is joined to aufs, and their names are ++whiteout-ed doubly, so that users will never see their names in aufs ++hierarchy. ++1. a regular file which will be hardlinked to all whiteouts. ++2. a directory to store a pseudo-link. ++3. a directory to store an "orphan"-ed file temporary. ++ ++1. Whiteout Base ++ When you remove a file on a readonly branch, aufs handles it as a ++ logical deletion and creates a whiteout on the upper writable branch ++ as a hardlink of this file in order not to consume inode on the ++ writable branch. ++2. Pseudo-link Dir ++ See below, Pseudo-link. ++3. Step-Parent Dir ++ When "fileC" exists on the lower readonly branch only and it is ++ opened and removed with its parent dir, and then user writes ++ something into it, then aufs copies-up fileC to this ++ directory. Because there is no other dir to store fileC. After ++ creating a file under this dir, the file is unlinked. ++ ++Because aufs supports manipulating branches, ie. add/delete/change ++dynamically, a branch has its own id. When the branch order changes, ++aufs finds the new index by searching the branch id. ++ ++ ++Pseudo-link ++---------------------------------------------------------------------- ++Assume "fileA" exists on the lower readonly branch only and it is ++hardlinked to "fileB" on the branch. When you write something to fileA, ++aufs copies-up it to the upper writable branch. Additionally aufs ++creates a hardlink under the Pseudo-link Directory of the writable ++branch. The inode of a pseudo-link is kept in aufs super_block as a ++simple list. If fileB is read after unlinking fileA, aufs returns ++filedata from the pseudo-link instead of the lower readonly ++branch. Because the pseudo-link is based upon the inode, to keep the ++inode number by xino (see above) is essentially necessary. ++ ++All the hardlinks under the Pseudo-link Directory of the writable branch ++should be restored in a proper location later. Aufs provides a utility ++to do this. The userspace helpers executed at remounting and unmounting ++aufs by default. ++During this utility is running, it puts aufs into the pseudo-link ++maintenance mode. In this mode, only the process which began the ++maintenance mode (and its child processes) is allowed to operate in ++aufs. Some other processes which are not related to the pseudo-link will ++be allowed to run too, but the rest have to return an error or wait ++until the maintenance mode ends. If a process already acquires an inode ++mutex (in VFS), it has to return an error. ++ ++ ++XIB(external inode number bitmap) ++---------------------------------------------------------------------- ++Addition to the xino file per a branch, aufs has an external inode number ++bitmap in a superblock object. It is also an internal file such like a ++xino file. ++It is a simple bitmap to mark whether the aufs inode number is in-use or ++not. ++To reduce the file I/O, aufs prepares a single memory page to cache xib. ++ ++As well as XINO files, aufs has a feature to truncate/refresh XIB to ++reduce the number of consumed disk blocks for these files. ++ ++ ++Virtual or Vertical Dir, and Readdir in Userspace ++---------------------------------------------------------------------- ++In order to support multiple layers (branches), aufs readdir operation ++constructs a virtual dir block on memory. For readdir, aufs calls ++vfs_readdir() internally for each dir on branches, merges their entries ++with eliminating the whiteout-ed ones, and sets it to file (dir) ++object. So the file object has its entry list until it is closed. The ++entry list will be updated when the file position is zero and becomes ++obsoleted. This decision is made in aufs automatically. ++ ++The dynamically allocated memory block for the name of entries has a ++unit of 512 bytes (by default) and stores the names contiguously (no ++padding). Another block for each entry is handled by kmem_cache too. ++During building dir blocks, aufs creates hash list and judging whether ++the entry is whiteouted by its upper branch or already listed. ++The merged result is cached in the corresponding inode object and ++maintained by a customizable life-time option. ++ ++Some people may call it can be a security hole or invite DoS attack ++since the opened and once readdir-ed dir (file object) holds its entry ++list and becomes a pressure for system memory. But I'd say it is similar ++to files under /proc or /sys. The virtual files in them also holds a ++memory page (generally) while they are opened. When an idea to reduce ++memory for them is introduced, it will be applied to aufs too. ++For those who really hate this situation, I've developed readdir(3) ++library which operates this merging in userspace. You just need to set ++LD_PRELOAD environment variable, and aufs will not consume no memory in ++kernel space for readdir(3). ++ ++ ++Workqueue ++---------------------------------------------------------------------- ++Aufs sometimes requires privilege access to a branch. For instance, ++in copy-up/down operation. When a user process is going to make changes ++to a file which exists in the lower readonly branch only, and the mode ++of one of ancestor directories may not be writable by a user ++process. Here aufs copy-up the file with its ancestors and they may ++require privilege to set its owner/group/mode/etc. ++This is a typical case of a application character of aufs (see ++Introduction). ++ ++Aufs uses workqueue synchronously for this case. It creates its own ++workqueue. The workqueue is a kernel thread and has privilege. Aufs ++passes the request to call mkdir or write (for example), and wait for ++its completion. This approach solves a problem of a signal handler ++simply. ++If aufs didn't adopt the workqueue and changed the privilege of the ++process, then the process may receive the unexpected SIGXFSZ or other ++signals. ++ ++Also aufs uses the system global workqueue ("events" kernel thread) too ++for asynchronous tasks, such like handling inotify/fsnotify, re-creating a ++whiteout base and etc. This is unrelated to a privilege. ++Most of aufs operation tries acquiring a rw_semaphore for aufs ++superblock at the beginning, at the same time waits for the completion ++of all queued asynchronous tasks. ++ ++ ++Whiteout ++---------------------------------------------------------------------- ++The whiteout in aufs is very similar to Unionfs's. That is represented ++by its filename. UnionMount takes an approach of a file mode, but I am ++afraid several utilities (find(1) or something) will have to support it. ++ ++Basically the whiteout represents "logical deletion" which stops aufs to ++lookup further, but also it represents "dir is opaque" which also stop ++further lookup. ++ ++In aufs, rmdir(2) and rename(2) for dir uses whiteout alternatively. ++In order to make several functions in a single systemcall to be ++revertible, aufs adopts an approach to rename a directory to a temporary ++unique whiteouted name. ++For example, in rename(2) dir where the target dir already existed, aufs ++renames the target dir to a temporary unique whiteouted name before the ++actual rename on a branch, and then handles other actions (make it opaque, ++update the attributes, etc). If an error happens in these actions, aufs ++simply renames the whiteouted name back and returns an error. If all are ++succeeded, aufs registers a function to remove the whiteouted unique ++temporary name completely and asynchronously to the system global ++workqueue. ++ ++ ++Copy-up ++---------------------------------------------------------------------- ++It is a well-known feature or concept. ++When user modifies a file on a readonly branch, aufs operate "copy-up" ++internally and makes change to the new file on the upper writable branch. ++When the trigger systemcall does not update the timestamps of the parent ++dir, aufs reverts it after copy-up. ++ ++ ++Move-down (aufs3.9 and later) ++---------------------------------------------------------------------- ++"Copy-up" is one of the essential feature in aufs. It copies a file from ++the lower readonly branch to the upper writable branch when a user ++changes something about the file. ++"Move-down" is an opposite action of copy-up. Basically this action is ++ran manually instead of automatically and internally. ++For desgin and implementation, aufs has to consider these issues. ++- whiteout for the file may exist on the lower branch. ++- ancestor directories may not exist on the lower branch. ++- diropq for the ancestor directories may exist on the upper branch. ++- free space on the lower branch will reduce. ++- another access to the file may happen during moving-down, including ++ UDBA (see "Revalidate Dentry and UDBA"). ++- the file should not be hard-linked nor pseudo-linked. they should be ++ handled by auplink utility later. ++ ++Sometimes users want to move-down a file from the upper writable branch ++to the lower readonly or writable branch. For instance, ++- the free space of the upper writable branch is going to run out. ++- create a new intermediate branch between the upper and lower branch. ++- etc. ++ ++For this purpose, use "aumvdown" command in aufs-util.git. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/03atomic_open.txt linux-4.1.10/Documentation/filesystems/aufs/design/03atomic_open.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/03atomic_open.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/03atomic_open.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,85 @@ ++ ++# Copyright (C) 2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Support for a branch who has its ->atomic_open() ++---------------------------------------------------------------------- ++The filesystems who implement its ->atomic_open() are not majority. For ++example NFSv4 does, and aufs should call NFSv4 ->atomic_open, ++particularly for open(O_CREAT|O_EXCL, 0400) case. Other than ++->atomic_open(), NFSv4 returns an error for this open(2). While I am not ++sure whether all filesystems who have ->atomic_open() behave like this, ++but NFSv4 surely returns the error. ++ ++In order to support ->atomic_open() for aufs, there are a few ++approaches. ++ ++A. Introduce aufs_atomic_open() ++ - calls one of VFS:do_last(), lookup_open() or atomic_open() for ++ branch fs. ++B. Introduce aufs_atomic_open() calling create, open and chmod. this is ++ an aufs user Pip Cet's approach ++ - calls aufs_create(), VFS finish_open() and notify_change(). ++ - pass fake-mode to finish_open(), and then correct the mode by ++ notify_change(). ++C. Extend aufs_open() to call branch fs's ->atomic_open() ++ - no aufs_atomic_open(). ++ - aufs_lookup() registers the TID to an aufs internal object. ++ - aufs_create() does nothing when the matching TID is registered, but ++ registers the mode. ++ - aufs_open() calls branch fs's ->atomic_open() when the matching ++ TID is registered. ++D. Extend aufs_open() to re-try branch fs's ->open() with superuser's ++ credential ++ - no aufs_atomic_open(). ++ - aufs_create() registers the TID to an internal object. this info ++ represents "this process created this file just now." ++ - when aufs gets EACCES from branch fs's ->open(), then confirm the ++ registered TID and re-try open() with superuser's credential. ++ ++Pros and cons for each approach. ++ ++A. ++ - straightforward but highly depends upon VFS internal. ++ - the atomic behavaiour is kept. ++ - some of parameters such as nameidata are hard to reproduce for ++ branch fs. ++ - large overhead. ++B. ++ - easy to implement. ++ - the atomic behavaiour is lost. ++C. ++ - the atomic behavaiour is kept. ++ - dirty and tricky. ++ - VFS checks whether the file is created correctly after calling ++ ->create(), which means this approach doesn't work. ++D. ++ - easy to implement. ++ - the atomic behavaiour is lost. ++ - to open a file with superuser's credential and give it to a user ++ process is a bad idea, since the file object keeps the credential ++ in it. It may affect LSM or something. This approach doesn't work ++ either. ++ ++The approach A is ideal, but it hard to implement. So here is a ++variation of A, which is to be implemented. ++ ++A-1. Introduce aufs_atomic_open() ++ - calls branch fs ->atomic_open() if exists. otherwise calls ++ vfs_create() and finish_open(). ++ - the demerit is that the several checks after branch fs ++ ->atomic_open() are lost. in the ordinary case, the checks are ++ done by VFS:do_last(), lookup_open() and atomic_open(). some can ++ be implemented in aufs, but not all I am afraid. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/03lookup.txt linux-4.1.10/Documentation/filesystems/aufs/design/03lookup.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/03lookup.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/03lookup.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,113 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Lookup in a Branch ++---------------------------------------------------------------------- ++Since aufs has a character of sub-VFS (see Introduction), it operates ++lookup for branches as VFS does. It may be a heavy work. But almost all ++lookup operation in aufs is the simplest case, ie. lookup only an entry ++directly connected to its parent. Digging down the directory hierarchy ++is unnecessary. VFS has a function lookup_one_len() for that use, and ++aufs calls it. ++ ++When a branch is a remote filesystem, aufs basically relies upon its ++->d_revalidate(), also aufs forces the hardest revalidate tests for ++them. ++For d_revalidate, aufs implements three levels of revalidate tests. See ++"Revalidate Dentry and UDBA" in detail. ++ ++ ++Test Only the Highest One for the Directory Permission (dirperm1 option) ++---------------------------------------------------------------------- ++Let's try case study. ++- aufs has two branches, upper readwrite and lower readonly. ++ /au = /rw + /ro ++- "dirA" exists under /ro, but /rw. and its mode is 0700. ++- user invoked "chmod a+rx /au/dirA" ++- the internal copy-up is activated and "/rw/dirA" is created and its ++ permission bits are set to world readable. ++- then "/au/dirA" becomes world readable? ++ ++In this case, /ro/dirA is still 0700 since it exists in readonly branch, ++or it may be a natively readonly filesystem. If aufs respects the lower ++branch, it should not respond readdir request from other users. But user ++allowed it by chmod. Should really aufs rejects showing the entries ++under /ro/dirA? ++ ++To be honest, I don't have a good solution for this case. So aufs ++implements 'dirperm1' and 'nodirperm1' mount options, and leave it to ++users. ++When dirperm1 is specified, aufs checks only the highest one for the ++directory permission, and shows the entries. Otherwise, as usual, checks ++every dir existing on all branches and rejects the request. ++ ++As a side effect, dirperm1 option improves the performance of aufs ++because the number of permission check is reduced when the number of ++branch is many. ++ ++ ++Revalidate Dentry and UDBA (User's Direct Branch Access) ++---------------------------------------------------------------------- ++Generally VFS helpers re-validate a dentry as a part of lookup. ++0. digging down the directory hierarchy. ++1. lock the parent dir by its i_mutex. ++2. lookup the final (child) entry. ++3. revalidate it. ++4. call the actual operation (create, unlink, etc.) ++5. unlock the parent dir ++ ++If the filesystem implements its ->d_revalidate() (step 3), then it is ++called. Actually aufs implements it and checks the dentry on a branch is ++still valid. ++But it is not enough. Because aufs has to release the lock for the ++parent dir on a branch at the end of ->lookup() (step 2) and ++->d_revalidate() (step 3) while the i_mutex of the aufs dir is still ++held by VFS. ++If the file on a branch is changed directly, eg. bypassing aufs, after ++aufs released the lock, then the subsequent operation may cause ++something unpleasant result. ++ ++This situation is a result of VFS architecture, ->lookup() and ++->d_revalidate() is separated. But I never say it is wrong. It is a good ++design from VFS's point of view. It is just not suitable for sub-VFS ++character in aufs. ++ ++Aufs supports such case by three level of revalidation which is ++selectable by user. ++1. Simple Revalidate ++ Addition to the native flow in VFS's, confirm the child-parent ++ relationship on the branch just after locking the parent dir on the ++ branch in the "actual operation" (step 4). When this validation ++ fails, aufs returns EBUSY. ->d_revalidate() (step 3) in aufs still ++ checks the validation of the dentry on branches. ++2. Monitor Changes Internally by Inotify/Fsnotify ++ Addition to above, in the "actual operation" (step 4) aufs re-lookup ++ the dentry on the branch, and returns EBUSY if it finds different ++ dentry. ++ Additionally, aufs sets the inotify/fsnotify watch for every dir on branches ++ during it is in cache. When the event is notified, aufs registers a ++ function to kernel 'events' thread by schedule_work(). And the ++ function sets some special status to the cached aufs dentry and inode ++ private data. If they are not cached, then aufs has nothing to ++ do. When the same file is accessed through aufs (step 0-3) later, ++ aufs will detect the status and refresh all necessary data. ++ In this mode, aufs has to ignore the event which is fired by aufs ++ itself. ++3. No Extra Validation ++ This is the simplest test and doesn't add any additional revalidation ++ test, and skip the revalidation in step 4. It is useful and improves ++ aufs performance when system surely hide the aufs branches from user, ++ by over-mounting something (or another method). +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/04branch.txt linux-4.1.10/Documentation/filesystems/aufs/design/04branch.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/04branch.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/04branch.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,74 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Branch Manipulation ++ ++Since aufs supports dynamic branch manipulation, ie. add/remove a branch ++and changing its permission/attribute, there are a lot of works to do. ++ ++ ++Add a Branch ++---------------------------------------------------------------------- ++o Confirm the adding dir exists outside of aufs, including loopback ++ mount, and its various attributes. ++o Initialize the xino file and whiteout bases if necessary. ++ See struct.txt. ++ ++o Check the owner/group/mode of the directory ++ When the owner/group/mode of the adding directory differs from the ++ existing branch, aufs issues a warning because it may impose a ++ security risk. ++ For example, when a upper writable branch has a world writable empty ++ top directory, a malicious user can create any files on the writable ++ branch directly, like copy-up and modify manually. If something like ++ /etc/{passwd,shadow} exists on the lower readonly branch but the upper ++ writable branch, and the writable branch is world-writable, then a ++ malicious guy may create /etc/passwd on the writable branch directly ++ and the infected file will be valid in aufs. ++ I am afraid it can be a security issue, but aufs can do nothing except ++ producing a warning. ++ ++ ++Delete a Branch ++---------------------------------------------------------------------- ++o Confirm the deleting branch is not busy ++ To be general, there is one merit to adopt "remount" interface to ++ manipulate branches. It is to discard caches. At deleting a branch, ++ aufs checks the still cached (and connected) dentries and inodes. If ++ there are any, then they are all in-use. An inode without its ++ corresponding dentry can be alive alone (for example, inotify/fsnotify case). ++ ++ For the cached one, aufs checks whether the same named entry exists on ++ other branches. ++ If the cached one is a directory, because aufs provides a merged view ++ to users, as long as one dir is left on any branch aufs can show the ++ dir to users. In this case, the branch can be removed from aufs. ++ Otherwise aufs rejects deleting the branch. ++ ++ If any file on the deleting branch is opened by aufs, then aufs ++ rejects deleting. ++ ++ ++Modify the Permission of a Branch ++---------------------------------------------------------------------- ++o Re-initialize or remove the xino file and whiteout bases if necessary. ++ See struct.txt. ++ ++o rw --> ro: Confirm the modifying branch is not busy ++ Aufs rejects the request if any of these conditions are true. ++ - a file on the branch is mmap-ed. ++ - a regular file on the branch is opened for write and there is no ++ same named entry on the upper branch. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/05wbr_policy.txt linux-4.1.10/Documentation/filesystems/aufs/design/05wbr_policy.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/05wbr_policy.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/05wbr_policy.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,64 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Policies to Select One among Multiple Writable Branches ++---------------------------------------------------------------------- ++When the number of writable branch is more than one, aufs has to decide ++the target branch for file creation or copy-up. By default, the highest ++writable branch which has the parent (or ancestor) dir of the target ++file is chosen (top-down-parent policy). ++By user's request, aufs implements some other policies to select the ++writable branch, for file creation several policies, round-robin, ++most-free-space, and other policies. For copy-up, top-down-parent, ++bottom-up-parent, bottom-up and others. ++ ++As expected, the round-robin policy selects the branch in circular. When ++you have two writable branches and creates 10 new files, 5 files will be ++created for each branch. mkdir(2) systemcall is an exception. When you ++create 10 new directories, all will be created on the same branch. ++And the most-free-space policy selects the one which has most free ++space among the writable branches. The amount of free space will be ++checked by aufs internally, and users can specify its time interval. ++ ++The policies for copy-up is more simple, ++top-down-parent is equivalent to the same named on in create policy, ++bottom-up-parent selects the writable branch where the parent dir ++exists and the nearest upper one from the copyup-source, ++bottom-up selects the nearest upper writable branch from the ++copyup-source, regardless the existence of the parent dir. ++ ++There are some rules or exceptions to apply these policies. ++- If there is a readonly branch above the policy-selected branch and ++ the parent dir is marked as opaque (a variation of whiteout), or the ++ target (creating) file is whiteout-ed on the upper readonly branch, ++ then the result of the policy is ignored and the target file will be ++ created on the nearest upper writable branch than the readonly branch. ++- If there is a writable branch above the policy-selected branch and ++ the parent dir is marked as opaque or the target file is whiteouted ++ on the branch, then the result of the policy is ignored and the target ++ file will be created on the highest one among the upper writable ++ branches who has diropq or whiteout. In case of whiteout, aufs removes ++ it as usual. ++- link(2) and rename(2) systemcalls are exceptions in every policy. ++ They try selecting the branch where the source exists as possible ++ since copyup a large file will take long time. If it can't be, ++ ie. the branch where the source exists is readonly, then they will ++ follow the copyup policy. ++- There is an exception for rename(2) when the target exists. ++ If the rename target exists, aufs compares the index of the branches ++ where the source and the target exists and selects the higher ++ one. If the selected branch is readonly, then aufs follows the ++ copyup policy. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/06fhsm.txt linux-4.1.10/Documentation/filesystems/aufs/design/06fhsm.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/06fhsm.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/06fhsm.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,120 @@ ++ ++# Copyright (C) 2011-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program; if not, write to the Free Software ++# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA ++ ++ ++File-based Hierarchical Storage Management (FHSM) ++---------------------------------------------------------------------- ++Hierarchical Storage Management (or HSM) is a well-known feature in the ++storage world. Aufs provides this feature as file-based with multiple ++writable branches, based upon the principle of "Colder, the Lower". ++Here the word "colder" means that the less used files, and "lower" means ++that the position in the order of the stacked branches vertically. ++These multiple writable branches are prioritized, ie. the topmost one ++should be the fastest drive and be used heavily. ++ ++o Characters in aufs FHSM story ++- aufs itself and a new branch attribute. ++- a new ioctl interface to move-down and to establish a connection with ++ the daemon ("move-down" is a converse of "copy-up"). ++- userspace tool and daemon. ++ ++The userspace daemon establishes a connection with aufs and waits for ++the notification. The notified information is very similar to struct ++statfs containing the number of consumed blocks and inodes. ++When the consumed blocks/inodes of a branch exceeds the user-specified ++upper watermark, the daemon activates its move-down process until the ++consumed blocks/inodes reaches the user-specified lower watermark. ++ ++The actual move-down is done by aufs based upon the request from ++user-space since we need to maintain the inode number and the internal ++pointer arrays in aufs. ++ ++Currently aufs FHSM handles the regular files only. Additionally they ++must not be hard-linked nor pseudo-linked. ++ ++ ++o Cowork of aufs and the user-space daemon ++ During the userspace daemon established the connection, aufs sends a ++ small notification to it whenever aufs writes something into the ++ writable branch. But it may cost high since aufs issues statfs(2) ++ internally. So user can specify a new option to cache the ++ info. Actually the notification is controlled by these factors. ++ + the specified cache time. ++ + classified as "force" by aufs internally. ++ Until the specified time expires, aufs doesn't send the info ++ except the forced cases. When aufs decide forcing, the info is always ++ notified to userspace. ++ For example, the number of free inodes is generally large enough and ++ the shortage of it happens rarely. So aufs doesn't force the ++ notification when creating a new file, directory and others. This is ++ the typical case which aufs doesn't force. ++ When aufs writes the actual filedata and the files consumes any of new ++ blocks, the aufs forces notifying. ++ ++ ++o Interfaces in aufs ++- New branch attribute. ++ + fhsm ++ Specifies that the branch is managed by FHSM feature. In other word, ++ participant in the FHSM. ++ When nofhsm is set to the branch, it will not be the source/target ++ branch of the move-down operation. This attribute is set ++ independently from coo and moo attributes, and if you want full ++ FHSM, you should specify them as well. ++- New mount option. ++ + fhsm_sec ++ Specifies a second to suppress many less important info to be ++ notified. ++- New ioctl. ++ + AUFS_CTL_FHSM_FD ++ create a new file descriptor which userspace can read the notification ++ (a subset of struct statfs) from aufs. ++- Module parameter 'brs' ++ It has to be set to 1. Otherwise the new mount option 'fhsm' will not ++ be set. ++- mount helpers /sbin/mount.aufs and /sbin/umount.aufs ++ When there are two or more branches with fhsm attributes, ++ /sbin/mount.aufs invokes the user-space daemon and /sbin/umount.aufs ++ terminates it. As a result of remounting and branch-manipulation, the ++ number of branches with fhsm attribute can be one. In this case, ++ /sbin/mount.aufs will terminate the user-space daemon. ++ ++ ++Finally the operation is done as these steps in kernel-space. ++- make sure that, ++ + no one else is using the file. ++ + the file is not hard-linked. ++ + the file is not pseudo-linked. ++ + the file is a regular file. ++ + the parent dir is not opaqued. ++- find the target writable branch. ++- make sure the file is not whiteout-ed by the upper (than the target) ++ branch. ++- make the parent dir on the target branch. ++- mutex lock the inode on the branch. ++- unlink the whiteout on the target branch (if exists). ++- lookup and create the whiteout-ed temporary name on the target branch. ++- copy the file as the whiteout-ed temporary name on the target branch. ++- rename the whiteout-ed temporary name to the original name. ++- unlink the file on the source branch. ++- maintain the internal pointer array and the external inode number ++ table (XINO). ++- maintain the timestamps and other attributes of the parent dir and the ++ file. ++ ++And of course, in every step, an error may happen. So the operation ++should restore the original file state after an error happens. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/06mmap.txt linux-4.1.10/Documentation/filesystems/aufs/design/06mmap.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/06mmap.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/06mmap.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,72 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++mmap(2) -- File Memory Mapping ++---------------------------------------------------------------------- ++In aufs, the file-mapped pages are handled by a branch fs directly, no ++interaction with aufs. It means aufs_mmap() calls the branch fs's ++->mmap(). ++This approach is simple and good, but there is one problem. ++Under /proc, several entries show the mmapped files by its path (with ++device and inode number), and the printed path will be the path on the ++branch fs's instead of virtual aufs's. ++This is not a problem in most cases, but some utilities lsof(1) (and its ++user) may expect the path on aufs. ++ ++To address this issue, aufs adds a new member called vm_prfile in struct ++vm_area_struct (and struct vm_region). The original vm_file points to ++the file on the branch fs in order to handle everything correctly as ++usual. The new vm_prfile points to a virtual file in aufs, and the ++show-functions in procfs refers to vm_prfile if it is set. ++Also we need to maintain several other places where touching vm_file ++such like ++- fork()/clone() copies vma and the reference count of vm_file is ++ incremented. ++- merging vma maintains the ref count too. ++ ++This is not a good approach. It just fakes the printed path. But it ++leaves all behaviour around f_mapping unchanged. This is surely an ++advantage. ++Actually aufs had adopted another complicated approach which calls ++generic_file_mmap() and handles struct vm_operations_struct. In this ++approach, aufs met a hard problem and I could not solve it without ++switching the approach. ++ ++There may be one more another approach which is ++- bind-mount the branch-root onto the aufs-root internally ++- grab the new vfsmount (ie. struct mount) ++- lazy-umount the branch-root internally ++- in open(2) the aufs-file, open the branch-file with the hidden ++ vfsmount (instead of the original branch's vfsmount) ++- ideally this "bind-mount and lazy-umount" should be done atomically, ++ but it may be possible from userspace by the mount helper. ++ ++Adding the internal hidden vfsmount and using it in opening a file, the ++file path under /proc will be printed correctly. This approach looks ++smarter, but is not possible I am afraid. ++- aufs-root may be bind-mount later. when it happens, another hidden ++ vfsmount will be required. ++- it is hard to get the chance to bind-mount and lazy-umount ++ + in kernel-space, FS can have vfsmount in open(2) via ++ file->f_path, and aufs can know its vfsmount. But several locks are ++ already acquired, and if aufs tries to bind-mount and lazy-umount ++ here, then it may cause a deadlock. ++ + in user-space, bind-mount doesn't invoke the mount helper. ++- since /proc shows dev and ino, aufs has to give vma these info. it ++ means a new member vm_prinode will be necessary. this is essentially ++ equivalent to vm_prfile described above. ++ ++I have to give up this "looks-smater" approach. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/06xattr.txt linux-4.1.10/Documentation/filesystems/aufs/design/06xattr.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/06xattr.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/06xattr.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,96 @@ ++ ++# Copyright (C) 2014-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program; if not, write to the Free Software ++# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA ++ ++ ++Listing XATTR/EA and getting the value ++---------------------------------------------------------------------- ++For the inode standard attributes (owner, group, timestamps, etc.), aufs ++shows the values from the topmost existing file. This behaviour is good ++for the non-dir entries since the bahaviour exactly matches the shown ++information. But for the directories, aufs considers all the same named ++entries on the lower branches. Which means, if one of the lower entry ++rejects readdir call, then aufs returns an error even if the topmost ++entry allows it. This behaviour is necessary to respect the branch fs's ++security, but can make users confused since the user-visible standard ++attributes don't match the behaviour. ++To address this issue, aufs has a mount option called dirperm1 which ++checks the permission for the topmost entry only, and ignores the lower ++entry's permission. ++ ++A similar issue can happen around XATTR. ++getxattr(2) and listxattr(2) families behave as if dirperm1 option is ++always set. Otherwise these very unpleasant situation would happen. ++- listxattr(2) may return the duplicated entries. ++- users may not be able to remove or reset the XATTR forever, ++ ++ ++XATTR/EA support in the internal (copy,move)-(up,down) ++---------------------------------------------------------------------- ++Generally the extended attributes of inode are categorized as these. ++- "security" for LSM and capability. ++- "system" for posix ACL, 'acl' mount option is required for the branch ++ fs generally. ++- "trusted" for userspace, CAP_SYS_ADMIN is required. ++- "user" for userspace, 'user_xattr' mount option is required for the ++ branch fs generally. ++ ++Moreover there are some other categories. Aufs handles these rather ++unpopular categories as the ordinary ones, ie. there is no special ++condition nor exception. ++ ++In copy-up, the support for XATTR on the dst branch may differ from the ++src branch. In this case, the copy-up operation will get an error and ++the original user operation which triggered the copy-up will fail. It ++can happen that even all copy-up will fail. ++When both of src and dst branches support XATTR and if an error occurs ++during copying XATTR, then the copy-up should fail obviously. That is a ++good reason and aufs should return an error to userspace. But when only ++the src branch support that XATTR, aufs should not return an error. ++For example, the src branch supports ACL but the dst branch doesn't ++because the dst branch may natively un-support it or temporary ++un-support it due to "noacl" mount option. Of course, the dst branch fs ++may NOT return an error even if the XATTR is not supported. It is ++totally up to the branch fs. ++ ++Anyway when the aufs internal copy-up gets an error from the dst branch ++fs, then aufs tries removing the just copied entry and returns the error ++to the userspace. The worst case of this situation will be all copy-up ++will fail. ++ ++For the copy-up operation, there two basic approaches. ++- copy the specified XATTR only (by category above), and return the ++ error unconditionally if it happens. ++- copy all XATTR, and ignore the error on the specified category only. ++ ++In order to support XATTR and to implement the correct behaviour, aufs ++chooses the latter approach and introduces some new branch attributes, ++"icexsec", "icexsys", "icextr", "icexusr", and "icexoth". ++They correspond to the XATTR namespaces (see above). Additionally, to be ++convenient, "icex" is also provided which means all "icex*" attributes ++are set (here the word "icex" stands for "ignore copy-error on XATTR"). ++ ++The meaning of these attributes is to ignore the error from setting ++XATTR on that branch. ++Note that aufs tries copying all XATTR unconditionally, and ignores the ++error from the dst branch according to the specified attributes. ++ ++Some XATTR may have its default value. The default value may come from ++the parent dir or the environment. If the default value is set at the ++file creating-time, it will be overwritten by copy-up. ++Some contradiction may happen I am afraid. ++Do we need another attribute to stop copying XATTR? I am unsure. For ++now, aufs implements the branch attributes to ignore the error. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/07export.txt linux-4.1.10/Documentation/filesystems/aufs/design/07export.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/07export.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/07export.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,58 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Export Aufs via NFS ++---------------------------------------------------------------------- ++Here is an approach. ++- like xino/xib, add a new file 'xigen' which stores aufs inode ++ generation. ++- iget_locked(): initialize aufs inode generation for a new inode, and ++ store it in xigen file. ++- destroy_inode(): increment aufs inode generation and store it in xigen ++ file. it is necessary even if it is not unlinked, because any data of ++ inode may be changed by UDBA. ++- encode_fh(): for a root dir, simply return FILEID_ROOT. otherwise ++ build file handle by ++ + branch id (4 bytes) ++ + superblock generation (4 bytes) ++ + inode number (4 or 8 bytes) ++ + parent dir inode number (4 or 8 bytes) ++ + inode generation (4 bytes)) ++ + return value of exportfs_encode_fh() for the parent on a branch (4 ++ bytes) ++ + file handle for a branch (by exportfs_encode_fh()) ++- fh_to_dentry(): ++ + find the index of a branch from its id in handle, and check it is ++ still exist in aufs. ++ + 1st level: get the inode number from handle and search it in cache. ++ + 2nd level: if not found in cache, get the parent inode number from ++ the handle and search it in cache. and then open the found parent ++ dir, find the matching inode number by vfs_readdir() and get its ++ name, and call lookup_one_len() for the target dentry. ++ + 3rd level: if the parent dir is not cached, call ++ exportfs_decode_fh() for a branch and get the parent on a branch, ++ build a pathname of it, convert it a pathname in aufs, call ++ path_lookup(). now aufs gets a parent dir dentry, then handle it as ++ the 2nd level. ++ + to open the dir, aufs needs struct vfsmount. aufs keeps vfsmount ++ for every branch, but not itself. to get this, (currently) aufs ++ searches in current->nsproxy->mnt_ns list. it may not be a good ++ idea, but I didn't get other approach. ++ + test the generation of the gotten inode. ++- every inode operation: they may get EBUSY due to UDBA. in this case, ++ convert it into ESTALE for NFSD. ++- readdir(): call lockdep_on/off() because filldir in NFSD calls ++ lookup_one_len(), vfs_getattr(), encode_fh() and others. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/08shwh.txt linux-4.1.10/Documentation/filesystems/aufs/design/08shwh.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/08shwh.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/08shwh.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,52 @@ ++ ++# Copyright (C) 2005-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Show Whiteout Mode (shwh) ++---------------------------------------------------------------------- ++Generally aufs hides the name of whiteouts. But in some cases, to show ++them is very useful for users. For instance, creating a new middle layer ++(branch) by merging existing layers. ++ ++(borrowing aufs1 HOW-TO from a user, Michael Towers) ++When you have three branches, ++- Bottom: 'system', squashfs (underlying base system), read-only ++- Middle: 'mods', squashfs, read-only ++- Top: 'overlay', ram (tmpfs), read-write ++ ++The top layer is loaded at boot time and saved at shutdown, to preserve ++the changes made to the system during the session. ++When larger changes have been made, or smaller changes have accumulated, ++the size of the saved top layer data grows. At this point, it would be ++nice to be able to merge the two overlay branches ('mods' and 'overlay') ++and rewrite the 'mods' squashfs, clearing the top layer and thus ++restoring save and load speed. ++ ++This merging is simplified by the use of another aufs mount, of just the ++two overlay branches using the 'shwh' option. ++# mount -t aufs -o ro,shwh,br:/livesys/overlay=ro+wh:/livesys/mods=rr+wh \ ++ aufs /livesys/merge_union ++ ++A merged view of these two branches is then available at ++/livesys/merge_union, and the new feature is that the whiteouts are ++visible! ++Note that in 'shwh' mode the aufs mount must be 'ro', which will disable ++writing to all branches. Also the default mode for all branches is 'ro'. ++It is now possible to save the combined contents of the two overlay ++branches to a new squashfs, e.g.: ++# mksquashfs /livesys/merge_union /path/to/newmods.squash ++ ++This new squashfs archive can be stored on the boot device and the ++initramfs will use it to replace the old one at the next boot. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/design/10dynop.txt linux-4.1.10/Documentation/filesystems/aufs/design/10dynop.txt +--- linux-4.1.10.orig/Documentation/filesystems/aufs/design/10dynop.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/design/10dynop.txt 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,47 @@ ++ ++# Copyright (C) 2010-2015 Junjiro R. Okajima ++# ++# This program is free software; you can redistribute it and/or modify ++# it under the terms of the GNU General Public License as published by ++# the Free Software Foundation; either version 2 of the License, or ++# (at your option) any later version. ++# ++# This program is distributed in the hope that it will be useful, ++# but WITHOUT ANY WARRANTY; without even the implied warranty of ++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++# GNU General Public License for more details. ++# ++# You should have received a copy of the GNU General Public License ++# along with this program. If not, see <http://www.gnu.org/licenses/>. ++ ++Dynamically customizable FS operations ++---------------------------------------------------------------------- ++Generally FS operations (struct inode_operations, struct ++address_space_operations, struct file_operations, etc.) are defined as ++"static const", but it never means that FS have only one set of ++operation. Some FS have multiple sets of them. For instance, ext2 has ++three sets, one for XIP, for NOBH, and for normal. ++Since aufs overrides and redirects these operations, sometimes aufs has ++to change its behaviour according to the branch FS type. More importantly ++VFS acts differently if a function (member in the struct) is set or ++not. It means aufs should have several sets of operations and select one ++among them according to the branch FS definition. ++ ++In order to solve this problem and not to affect the behaviour of VFS, ++aufs defines these operations dynamically. For instance, aufs defines ++dummy direct_IO function for struct address_space_operations, but it may ++not be set to the address_space_operations actually. When the branch FS ++doesn't have it, aufs doesn't set it to its address_space_operations ++while the function definition itself is still alive. So the behaviour ++itself will not change, and it will return an error when direct_IO is ++not set. ++ ++The lifetime of these dynamically generated operation object is ++maintained by aufs branch object. When the branch is removed from aufs, ++the reference counter of the object is decremented. When it reaches ++zero, the dynamically generated operation object will be freed. ++ ++This approach is designed to support AIO (io_submit), Direct I/O and ++XIP (DAX) mainly. ++Currently this approach is applied to address_space_operations for ++regular files only. +diff -Nur linux-4.1.10.orig/Documentation/filesystems/aufs/README linux-4.1.10/Documentation/filesystems/aufs/README +--- linux-4.1.10.orig/Documentation/filesystems/aufs/README 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/Documentation/filesystems/aufs/README 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,383 @@ ++ ++Aufs4 -- advanced multi layered unification filesystem version 4.x ++http://aufs.sf.net ++Junjiro R. Okajima ++ ++ ++0. Introduction ++---------------------------------------- ++In the early days, aufs was entirely re-designed and re-implemented ++Unionfs Version 1.x series. Adding many original ideas, approaches, ++improvements and implementations, it becomes totally different from ++Unionfs while keeping the basic features. ++Recently, Unionfs Version 2.x series begin taking some of the same ++approaches to aufs1's. ++Unionfs is being developed by Professor Erez Zadok at Stony Brook ++University and his team. ++ ++Aufs4 supports linux-4.0 and later, and for linux-3.x series try aufs3. ++If you want older kernel version support, try aufs2-2.6.git or ++aufs2-standalone.git repository, aufs1 from CVS on SourceForge. ++ ++Note: it becomes clear that "Aufs was rejected. Let's give it up." ++ According to Christoph Hellwig, linux rejects all union-type ++ filesystems but UnionMount. ++<http://marc.info/?l=linux-kernel&m=123938533724484&w=2> ++ ++PS. Al Viro seems have a plan to merge aufs as well as overlayfs and ++ UnionMount, and he pointed out an issue around a directory mutex ++ lock and aufs addressed it. But it is still unsure whether aufs will ++ be merged (or any other union solution). ++<http://marc.info/?l=linux-kernel&m=136312705029295&w=1> ++ ++ ++1. Features ++---------------------------------------- ++- unite several directories into a single virtual filesystem. The member ++ directory is called as a branch. ++- you can specify the permission flags to the branch, which are 'readonly', ++ 'readwrite' and 'whiteout-able.' ++- by upper writable branch, internal copyup and whiteout, files/dirs on ++ readonly branch are modifiable logically. ++- dynamic branch manipulation, add, del. ++- etc... ++ ++Also there are many enhancements in aufs, such as: ++- test only the highest one for the directory permission (dirperm1) ++- copyup on open (coo=) ++- 'move' policy for copy-up between two writable branches, after ++ checking free space. ++- xattr, acl ++- readdir(3) in userspace. ++- keep inode number by external inode number table ++- keep the timestamps of file/dir in internal copyup operation ++- seekable directory, supporting NFS readdir. ++- whiteout is hardlinked in order to reduce the consumption of inodes ++ on branch ++- do not copyup, nor create a whiteout when it is unnecessary ++- revert a single systemcall when an error occurs in aufs ++- remount interface instead of ioctl ++- maintain /etc/mtab by an external command, /sbin/mount.aufs. ++- loopback mounted filesystem as a branch ++- kernel thread for removing the dir who has a plenty of whiteouts ++- support copyup sparse file (a file which has a 'hole' in it) ++- default permission flags for branches ++- selectable permission flags for ro branch, whether whiteout can ++ exist or not ++- export via NFS. ++- support <sysfs>/fs/aufs and <debugfs>/aufs. ++- support multiple writable branches, some policies to select one ++ among multiple writable branches. ++- a new semantics for link(2) and rename(2) to support multiple ++ writable branches. ++- no glibc changes are required. ++- pseudo hardlink (hardlink over branches) ++- allow a direct access manually to a file on branch, e.g. bypassing aufs. ++ including NFS or remote filesystem branch. ++- userspace wrapper for pathconf(3)/fpathconf(3) with _PC_LINK_MAX. ++- and more... ++ ++Currently these features are dropped temporary from aufs4. ++See design/08plan.txt in detail. ++- nested mount, i.e. aufs as readonly no-whiteout branch of another aufs ++ (robr) ++- statistics of aufs thread (/sys/fs/aufs/stat) ++ ++Features or just an idea in the future (see also design/*.txt), ++- reorder the branch index without del/re-add. ++- permanent xino files for NFSD ++- an option for refreshing the opened files after add/del branches ++- light version, without branch manipulation. (unnecessary?) ++- copyup in userspace ++- inotify in userspace ++- readv/writev ++ ++ ++2. Download ++---------------------------------------- ++There are three GIT trees for aufs4, aufs4-linux.git, ++aufs4-standalone.git, and aufs-util.git. Note that there is no "4" in ++"aufs-util.git." ++While the aufs-util is always necessary, you need either of aufs4-linux ++or aufs4-standalone. ++ ++The aufs4-linux tree includes the whole linux mainline GIT tree, ++git://git.kernel.org/.../torvalds/linux.git. ++And you cannot select CONFIG_AUFS_FS=m for this version, eg. you cannot ++build aufs4 as an external kernel module. ++Several extra patches are not included in this tree. Only ++aufs4-standalone tree contains them. They are describe in the later ++section "Configuration and Compilation." ++ ++On the other hand, the aufs4-standalone tree has only aufs source files ++and necessary patches, and you can select CONFIG_AUFS_FS=m. ++But you need to apply all aufs patches manually. ++ ++You will find GIT branches whose name is in form of "aufs4.x" where "x" ++represents the linux kernel version, "linux-4.x". For instance, ++"aufs4.0" is for linux-4.0. For latest "linux-4.x-rcN", use ++"aufs4.x-rcN" branch. ++ ++o aufs4-linux tree ++$ git clone --reference /your/linux/git/tree \ ++ git://github.com/sfjro/aufs4-linux.git aufs4-linux.git ++- if you don't have linux GIT tree, then remove "--reference ..." ++$ cd aufs4-linux.git ++$ git checkout origin/aufs4.0 ++ ++Or You may want to directly git-pull aufs into your linux GIT tree, and ++leave the patch-work to GIT. ++$ cd /your/linux/git/tree ++$ git remote add aufs4 git://github.com/sfjro/aufs4-linux.git ++$ git fetch aufs4 ++$ git checkout -b my4.0 v4.0 ++$ (add your local change...) ++$ git pull aufs4 aufs4.0 ++- now you have v4.0 + your_changes + aufs4.0 in you my4.0 branch. ++- you may need to solve some conflicts between your_changes and ++ aufs4.0. in this case, git-rerere is recommended so that you can ++ solve the similar conflicts automatically when you upgrade to 4.1 or ++ later in the future. ++ ++o aufs4-standalone tree ++$ git clone git://github.com/sfjro/aufs4-standalone.git aufs4-standalone.git ++$ cd aufs4-standalone.git ++$ git checkout origin/aufs4.0 ++ ++o aufs-util tree ++$ git clone git://git.code.sf.net/p/aufs/aufs-util aufs-util.git ++- note that the public aufs-util.git is on SourceForge instead of ++ GitHUB. ++$ cd aufs-util.git ++$ git checkout origin/aufs4.0 ++ ++Note: The 4.x-rcN branch is to be used with `rc' kernel versions ONLY. ++The minor version number, 'x' in '4.x', of aufs may not always ++follow the minor version number of the kernel. ++Because changes in the kernel that cause the use of a new ++minor version number do not always require changes to aufs-util. ++ ++Since aufs-util has its own minor version number, you may not be ++able to find a GIT branch in aufs-util for your kernel's ++exact minor version number. ++In this case, you should git-checkout the branch for the ++nearest lower number. ++ ++For (an unreleased) example: ++If you are using "linux-4.10" and the "aufs4.10" branch ++does not exist in aufs-util repository, then "aufs4.9", "aufs4.8" ++or something numerically smaller is the branch for your kernel. ++ ++Also you can view all branches by ++ $ git branch -a ++ ++ ++3. Configuration and Compilation ++---------------------------------------- ++Make sure you have git-checkout'ed the correct branch. ++ ++For aufs4-linux tree, ++- enable CONFIG_AUFS_FS. ++- set other aufs configurations if necessary. ++ ++For aufs4-standalone tree, ++There are several ways to build. ++ ++1. ++- apply ./aufs4-kbuild.patch to your kernel source files. ++- apply ./aufs4-base.patch too. ++- apply ./aufs4-mmap.patch too. ++- apply ./aufs4-standalone.patch too, if you have a plan to set ++ CONFIG_AUFS_FS=m. otherwise you don't need ./aufs4-standalone.patch. ++- copy ./{Documentation,fs,include/uapi/linux/aufs_type.h} files to your ++ kernel source tree. Never copy $PWD/include/uapi/linux/Kbuild. ++- enable CONFIG_AUFS_FS, you can select either ++ =m or =y. ++- and build your kernel as usual. ++- install the built kernel. ++ Note: Since linux-3.9, every filesystem module requires an alias ++ "fs-<fsname>". You should make sure that "fs-aufs" is listed in your ++ modules.aliases file if you set CONFIG_AUFS_FS=m. ++- install the header files too by "make headers_install" to the ++ directory where you specify. By default, it is $PWD/usr. ++ "make help" shows a brief note for headers_install. ++- and reboot your system. ++ ++2. ++- module only (CONFIG_AUFS_FS=m). ++- apply ./aufs4-base.patch to your kernel source files. ++- apply ./aufs4-mmap.patch too. ++- apply ./aufs4-standalone.patch too. ++- build your kernel, don't forget "make headers_install", and reboot. ++- edit ./config.mk and set other aufs configurations if necessary. ++ Note: You should read $PWD/fs/aufs/Kconfig carefully which describes ++ every aufs configurations. ++- build the module by simple "make". ++ Note: Since linux-3.9, every filesystem module requires an alias ++ "fs-<fsname>". You should make sure that "fs-aufs" is listed in your ++ modules.aliases file. ++- you can specify ${KDIR} make variable which points to your kernel ++ source tree. ++- install the files ++ + run "make install" to install the aufs module, or copy the built ++ $PWD/aufs.ko to /lib/modules/... and run depmod -a (or reboot simply). ++ + run "make install_headers" (instead of headers_install) to install ++ the modified aufs header file (you can specify DESTDIR which is ++ available in aufs standalone version's Makefile only), or copy ++ $PWD/usr/include/linux/aufs_type.h to /usr/include/linux or wherever ++ you like manually. By default, the target directory is $PWD/usr. ++- no need to apply aufs4-kbuild.patch, nor copying source files to your ++ kernel source tree. ++ ++Note: The header file aufs_type.h is necessary to build aufs-util ++ as well as "make headers_install" in the kernel source tree. ++ headers_install is subject to be forgotten, but it is essentially ++ necessary, not only for building aufs-util. ++ You may not meet problems without headers_install in some older ++ version though. ++ ++And then, ++- read README in aufs-util, build and install it ++- note that your distribution may contain an obsoleted version of ++ aufs_type.h in /usr/include/linux or something. When you build aufs ++ utilities, make sure that your compiler refers the correct aufs header ++ file which is built by "make headers_install." ++- if you want to use readdir(3) in userspace or pathconf(3) wrapper, ++ then run "make install_ulib" too. And refer to the aufs manual in ++ detail. ++ ++There several other patches in aufs4-standalone.git. They are all ++optional. When you meet some problems, they will help you. ++- aufs4-loopback.patch ++ Supports a nested loopback mount in a branch-fs. This patch is ++ unnecessary until aufs produces a message like "you may want to try ++ another patch for loopback file". ++- vfs-ino.patch ++ Modifies a system global kernel internal function get_next_ino() in ++ order to stop assigning 0 for an inode-number. Not directly related to ++ aufs, but recommended generally. ++- tmpfs-idr.patch ++ Keeps the tmpfs inode number as the lowest value. Effective to reduce ++ the size of aufs XINO files for tmpfs branch. Also it prevents the ++ duplication of inode number, which is important for backup tools and ++ other utilities. When you find aufs XINO files for tmpfs branch ++ growing too much, try this patch. ++ ++ ++4. Usage ++---------------------------------------- ++At first, make sure aufs-util are installed, and please read the aufs ++manual, aufs.5 in aufs-util.git tree. ++$ man -l aufs.5 ++ ++And then, ++$ mkdir /tmp/rw /tmp/aufs ++# mount -t aufs -o br=/tmp/rw:${HOME} none /tmp/aufs ++ ++Here is another example. The result is equivalent. ++# mount -t aufs -o br=/tmp/rw=rw:${HOME}=ro none /tmp/aufs ++ Or ++# mount -t aufs -o br:/tmp/rw none /tmp/aufs ++# mount -o remount,append:${HOME} /tmp/aufs ++ ++Then, you can see whole tree of your home dir through /tmp/aufs. If ++you modify a file under /tmp/aufs, the one on your home directory is ++not affected, instead the same named file will be newly created under ++/tmp/rw. And all of your modification to a file will be applied to ++the one under /tmp/rw. This is called the file based Copy on Write ++(COW) method. ++Aufs mount options are described in aufs.5. ++If you run chroot or something and make your aufs as a root directory, ++then you need to customize the shutdown script. See the aufs manual in ++detail. ++ ++Additionally, there are some sample usages of aufs which are a ++diskless system with network booting, and LiveCD over NFS. ++See sample dir in CVS tree on SourceForge. ++ ++ ++5. Contact ++---------------------------------------- ++When you have any problems or strange behaviour in aufs, please let me ++know with: ++- /proc/mounts (instead of the output of mount(8)) ++- /sys/module/aufs/* ++- /sys/fs/aufs/* (if you have them) ++- /debug/aufs/* (if you have them) ++- linux kernel version ++ if your kernel is not plain, for example modified by distributor, ++ the url where i can download its source is necessary too. ++- aufs version which was printed at loading the module or booting the ++ system, instead of the date you downloaded. ++- configuration (define/undefine CONFIG_AUFS_xxx) ++- kernel configuration or /proc/config.gz (if you have it) ++- behaviour which you think to be incorrect ++- actual operation, reproducible one is better ++- mailto: aufs-users at lists.sourceforge.net ++ ++Usually, I don't watch the Public Areas(Bugs, Support Requests, Patches, ++and Feature Requests) on SourceForge. Please join and write to ++aufs-users ML. ++ ++ ++6. Acknowledgements ++---------------------------------------- ++Thanks to everyone who have tried and are using aufs, whoever ++have reported a bug or any feedback. ++ ++Especially donators: ++Tomas Matejicek(slax.org) made a donation (much more than once). ++ Since Apr 2010, Tomas M (the author of Slax and Linux Live ++ scripts) is making "doubling" donations. ++ Unfortunately I cannot list all of the donators, but I really ++ appreciate. ++ It ends Aug 2010, but the ordinary donation URL is still available. ++ <http://sourceforge.net/donate/index.php?group_id=167503> ++Dai Itasaka made a donation (2007/8). ++Chuck Smith made a donation (2008/4, 10 and 12). ++Henk Schoneveld made a donation (2008/9). ++Chih-Wei Huang, ASUS, CTC donated Eee PC 4G (2008/10). ++Francois Dupoux made a donation (2008/11). ++Bruno Cesar Ribas and Luis Carlos Erpen de Bona, C3SL serves public ++ aufs2 GIT tree (2009/2). ++William Grant made a donation (2009/3). ++Patrick Lane made a donation (2009/4). ++The Mail Archive (mail-archive.com) made donations (2009/5). ++Nippy Networks (Ed Wildgoose) made a donation (2009/7). ++New Dream Network, LLC (www.dreamhost.com) made a donation (2009/11). ++Pavel Pronskiy made a donation (2011/2). ++Iridium and Inmarsat satellite phone retailer (www.mailasail.com), Nippy ++ Networks (Ed Wildgoose) made a donation for hardware (2011/3). ++Max Lekomcev (DOM-TV project) made a donation (2011/7, 12, 2012/3, 6 and ++11). ++Sam Liddicott made a donation (2011/9). ++Era Scarecrow made a donation (2013/4). ++Bor Ratajc made a donation (2013/4). ++Alessandro Gorreta made a donation (2013/4). ++POIRETTE Marc made a donation (2013/4). ++Alessandro Gorreta made a donation (2013/4). ++lauri kasvandik made a donation (2013/5). ++"pemasu from Finland" made a donation (2013/7). ++The Parted Magic Project made a donation (2013/9 and 11). ++Pavel Barta made a donation (2013/10). ++Nikolay Pertsev made a donation (2014/5). ++James B made a donation (2014/7 and 2015/7). ++Stefano Di Biase made a donation (2014/8). ++Daniel Epellei made a donation (2015/1). ++ ++Thank you very much. ++Donations are always, including future donations, very important and ++helpful for me to keep on developing aufs. ++ ++ ++7. ++---------------------------------------- ++If you are an experienced user, no explanation is needed. Aufs is ++just a linux filesystem. ++ ++ ++Enjoy! ++ ++# Local variables: ; ++# mode: text; ++# End: ; +diff -Nur linux-4.1.10.orig/drivers/block/loop.c linux-4.1.10/drivers/block/loop.c +--- linux-4.1.10.orig/drivers/block/loop.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/drivers/block/loop.c 2015-10-22 21:35:53.000000000 +0200 +@@ -538,6 +538,24 @@ + return i && S_ISBLK(i->i_mode) && MAJOR(i->i_rdev) == LOOP_MAJOR; + } + ++/* ++ * for AUFS ++ * no get/put for file. ++ */ ++struct file *loop_backing_file(struct super_block *sb) ++{ ++ struct file *ret; ++ struct loop_device *l; ++ ++ ret = NULL; ++ if (MAJOR(sb->s_dev) == LOOP_MAJOR) { ++ l = sb->s_bdev->bd_disk->private_data; ++ ret = l->lo_backing_file; ++ } ++ return ret; ++} ++EXPORT_SYMBOL(loop_backing_file); ++ + /* loop sysfs attributes */ + + static ssize_t loop_attr_show(struct device *dev, char *page, +diff -Nur linux-4.1.10.orig/fs/aufs/aufs.h linux-4.1.10/fs/aufs/aufs.h +--- linux-4.1.10.orig/fs/aufs/aufs.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/aufs.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,59 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * all header files ++ */ ++ ++#ifndef __AUFS_H__ ++#define __AUFS_H__ ++ ++#ifdef __KERNEL__ ++ ++#define AuStub(type, name, body, ...) \ ++ static inline type name(__VA_ARGS__) { body; } ++ ++#define AuStubVoid(name, ...) \ ++ AuStub(void, name, , __VA_ARGS__) ++#define AuStubInt0(name, ...) \ ++ AuStub(int, name, return 0, __VA_ARGS__) ++ ++#include "debug.h" ++ ++#include "branch.h" ++#include "cpup.h" ++#include "dcsub.h" ++#include "dbgaufs.h" ++#include "dentry.h" ++#include "dir.h" ++#include "dynop.h" ++#include "file.h" ++#include "fstype.h" ++#include "inode.h" ++#include "loop.h" ++#include "module.h" ++#include "opts.h" ++#include "rwsem.h" ++#include "spl.h" ++#include "super.h" ++#include "sysaufs.h" ++#include "vfsub.h" ++#include "whout.h" ++#include "wkq.h" ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/branch.c linux-4.1.10/fs/aufs/branch.c +--- linux-4.1.10.orig/fs/aufs/branch.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/branch.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1414 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * branch management ++ */ ++ ++#include <linux/compat.h> ++#include <linux/statfs.h> ++#include "aufs.h" ++ ++/* ++ * free a single branch ++ */ ++static void au_br_do_free(struct au_branch *br) ++{ ++ int i; ++ struct au_wbr *wbr; ++ struct au_dykey **key; ++ ++ au_hnotify_fin_br(br); ++ ++ if (br->br_xino.xi_file) ++ fput(br->br_xino.xi_file); ++ mutex_destroy(&br->br_xino.xi_nondir_mtx); ++ ++ AuDebugOn(atomic_read(&br->br_count)); ++ ++ wbr = br->br_wbr; ++ if (wbr) { ++ for (i = 0; i < AuBrWh_Last; i++) ++ dput(wbr->wbr_wh[i]); ++ AuDebugOn(atomic_read(&wbr->wbr_wh_running)); ++ AuRwDestroy(&wbr->wbr_wh_rwsem); ++ } ++ ++ if (br->br_fhsm) { ++ au_br_fhsm_fin(br->br_fhsm); ++ kfree(br->br_fhsm); ++ } ++ ++ key = br->br_dykey; ++ for (i = 0; i < AuBrDynOp; i++, key++) ++ if (*key) ++ au_dy_put(*key); ++ else ++ break; ++ ++ /* recursive lock, s_umount of branch's */ ++ lockdep_off(); ++ path_put(&br->br_path); ++ lockdep_on(); ++ kfree(wbr); ++ kfree(br); ++} ++ ++/* ++ * frees all branches ++ */ ++void au_br_free(struct au_sbinfo *sbinfo) ++{ ++ aufs_bindex_t bmax; ++ struct au_branch **br; ++ ++ AuRwMustWriteLock(&sbinfo->si_rwsem); ++ ++ bmax = sbinfo->si_bend + 1; ++ br = sbinfo->si_branch; ++ while (bmax--) ++ au_br_do_free(*br++); ++} ++ ++/* ++ * find the index of a branch which is specified by @br_id. ++ */ ++int au_br_index(struct super_block *sb, aufs_bindex_t br_id) ++{ ++ aufs_bindex_t bindex, bend; ++ ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) ++ if (au_sbr_id(sb, bindex) == br_id) ++ return bindex; ++ return -1; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * add a branch ++ */ ++ ++static int test_overlap(struct super_block *sb, struct dentry *h_adding, ++ struct dentry *h_root) ++{ ++ if (unlikely(h_adding == h_root ++ || au_test_loopback_overlap(sb, h_adding))) ++ return 1; ++ if (h_adding->d_sb != h_root->d_sb) ++ return 0; ++ return au_test_subdir(h_adding, h_root) ++ || au_test_subdir(h_root, h_adding); ++} ++ ++/* ++ * returns a newly allocated branch. @new_nbranch is a number of branches ++ * after adding a branch. ++ */ ++static struct au_branch *au_br_alloc(struct super_block *sb, int new_nbranch, ++ int perm) ++{ ++ struct au_branch *add_branch; ++ struct dentry *root; ++ struct inode *inode; ++ int err; ++ ++ err = -ENOMEM; ++ root = sb->s_root; ++ add_branch = kmalloc(sizeof(*add_branch), GFP_NOFS); ++ if (unlikely(!add_branch)) ++ goto out; ++ ++ err = au_hnotify_init_br(add_branch, perm); ++ if (unlikely(err)) ++ goto out_br; ++ ++ add_branch->br_wbr = NULL; ++ if (au_br_writable(perm)) { ++ /* may be freed separately at changing the branch permission */ ++ add_branch->br_wbr = kmalloc(sizeof(*add_branch->br_wbr), ++ GFP_NOFS); ++ if (unlikely(!add_branch->br_wbr)) ++ goto out_hnotify; ++ } ++ ++ add_branch->br_fhsm = NULL; ++ if (au_br_fhsm(perm)) { ++ err = au_fhsm_br_alloc(add_branch); ++ if (unlikely(err)) ++ goto out_wbr; ++ } ++ ++ err = au_sbr_realloc(au_sbi(sb), new_nbranch); ++ if (!err) ++ err = au_di_realloc(au_di(root), new_nbranch); ++ if (!err) { ++ inode = d_inode(root); ++ err = au_ii_realloc(au_ii(inode), new_nbranch); ++ } ++ if (!err) ++ return add_branch; /* success */ ++ ++out_wbr: ++ kfree(add_branch->br_wbr); ++out_hnotify: ++ au_hnotify_fin_br(add_branch); ++out_br: ++ kfree(add_branch); ++out: ++ return ERR_PTR(err); ++} ++ ++/* ++ * test if the branch permission is legal or not. ++ */ ++static int test_br(struct inode *inode, int brperm, char *path) ++{ ++ int err; ++ ++ err = (au_br_writable(brperm) && IS_RDONLY(inode)); ++ if (!err) ++ goto out; ++ ++ err = -EINVAL; ++ pr_err("write permission for readonly mount or inode, %s\n", path); ++ ++out: ++ return err; ++} ++ ++/* ++ * returns: ++ * 0: success, the caller will add it ++ * plus: success, it is already unified, the caller should ignore it ++ * minus: error ++ */ ++static int test_add(struct super_block *sb, struct au_opt_add *add, int remount) ++{ ++ int err; ++ aufs_bindex_t bend, bindex; ++ struct dentry *root, *h_dentry; ++ struct inode *inode, *h_inode; ++ ++ root = sb->s_root; ++ bend = au_sbend(sb); ++ if (unlikely(bend >= 0 ++ && au_find_dbindex(root, add->path.dentry) >= 0)) { ++ err = 1; ++ if (!remount) { ++ err = -EINVAL; ++ pr_err("%s duplicated\n", add->pathname); ++ } ++ goto out; ++ } ++ ++ err = -ENOSPC; /* -E2BIG; */ ++ if (unlikely(AUFS_BRANCH_MAX <= add->bindex ++ || AUFS_BRANCH_MAX - 1 <= bend)) { ++ pr_err("number of branches exceeded %s\n", add->pathname); ++ goto out; ++ } ++ ++ err = -EDOM; ++ if (unlikely(add->bindex < 0 || bend + 1 < add->bindex)) { ++ pr_err("bad index %d\n", add->bindex); ++ goto out; ++ } ++ ++ inode = d_inode(add->path.dentry); ++ err = -ENOENT; ++ if (unlikely(!inode->i_nlink)) { ++ pr_err("no existence %s\n", add->pathname); ++ goto out; ++ } ++ ++ err = -EINVAL; ++ if (unlikely(inode->i_sb == sb)) { ++ pr_err("%s must be outside\n", add->pathname); ++ goto out; ++ } ++ ++ if (unlikely(au_test_fs_unsuppoted(inode->i_sb))) { ++ pr_err("unsupported filesystem, %s (%s)\n", ++ add->pathname, au_sbtype(inode->i_sb)); ++ goto out; ++ } ++ ++ if (unlikely(inode->i_sb->s_stack_depth)) { ++ pr_err("already stacked, %s (%s)\n", ++ add->pathname, au_sbtype(inode->i_sb)); ++ goto out; ++ } ++ ++ err = test_br(d_inode(add->path.dentry), add->perm, add->pathname); ++ if (unlikely(err)) ++ goto out; ++ ++ if (bend < 0) ++ return 0; /* success */ ++ ++ err = -EINVAL; ++ for (bindex = 0; bindex <= bend; bindex++) ++ if (unlikely(test_overlap(sb, add->path.dentry, ++ au_h_dptr(root, bindex)))) { ++ pr_err("%s is overlapped\n", add->pathname); ++ goto out; ++ } ++ ++ err = 0; ++ if (au_opt_test(au_mntflags(sb), WARN_PERM)) { ++ h_dentry = au_h_dptr(root, 0); ++ h_inode = d_inode(h_dentry); ++ if ((h_inode->i_mode & S_IALLUGO) != (inode->i_mode & S_IALLUGO) ++ || !uid_eq(h_inode->i_uid, inode->i_uid) ++ || !gid_eq(h_inode->i_gid, inode->i_gid)) ++ pr_warn("uid/gid/perm %s %u/%u/0%o, %u/%u/0%o\n", ++ add->pathname, ++ i_uid_read(inode), i_gid_read(inode), ++ (inode->i_mode & S_IALLUGO), ++ i_uid_read(h_inode), i_gid_read(h_inode), ++ (h_inode->i_mode & S_IALLUGO)); ++ } ++ ++out: ++ return err; ++} ++ ++/* ++ * initialize or clean the whiteouts for an adding branch ++ */ ++static int au_br_init_wh(struct super_block *sb, struct au_branch *br, ++ int new_perm) ++{ ++ int err, old_perm; ++ aufs_bindex_t bindex; ++ struct mutex *h_mtx; ++ struct au_wbr *wbr; ++ struct au_hinode *hdir; ++ struct dentry *h_dentry; ++ ++ err = vfsub_mnt_want_write(au_br_mnt(br)); ++ if (unlikely(err)) ++ goto out; ++ ++ wbr = br->br_wbr; ++ old_perm = br->br_perm; ++ br->br_perm = new_perm; ++ hdir = NULL; ++ h_mtx = NULL; ++ bindex = au_br_index(sb, br->br_id); ++ if (0 <= bindex) { ++ hdir = au_hi(d_inode(sb->s_root), bindex); ++ au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); ++ } else { ++ h_dentry = au_br_dentry(br); ++ h_mtx = &d_inode(h_dentry)->i_mutex; ++ mutex_lock_nested(h_mtx, AuLsc_I_PARENT); ++ } ++ if (!wbr) ++ err = au_wh_init(br, sb); ++ else { ++ wbr_wh_write_lock(wbr); ++ err = au_wh_init(br, sb); ++ wbr_wh_write_unlock(wbr); ++ } ++ if (hdir) ++ au_hn_imtx_unlock(hdir); ++ else ++ mutex_unlock(h_mtx); ++ vfsub_mnt_drop_write(au_br_mnt(br)); ++ br->br_perm = old_perm; ++ ++ if (!err && wbr && !au_br_writable(new_perm)) { ++ kfree(wbr); ++ br->br_wbr = NULL; ++ } ++ ++out: ++ return err; ++} ++ ++static int au_wbr_init(struct au_branch *br, struct super_block *sb, ++ int perm) ++{ ++ int err; ++ struct kstatfs kst; ++ struct au_wbr *wbr; ++ ++ wbr = br->br_wbr; ++ au_rw_init(&wbr->wbr_wh_rwsem); ++ memset(wbr->wbr_wh, 0, sizeof(wbr->wbr_wh)); ++ atomic_set(&wbr->wbr_wh_running, 0); ++ wbr->wbr_bytes = 0; ++ ++ /* ++ * a limit for rmdir/rename a dir ++ * cf. AUFS_MAX_NAMELEN in include/uapi/linux/aufs_type.h ++ */ ++ err = vfs_statfs(&br->br_path, &kst); ++ if (unlikely(err)) ++ goto out; ++ err = -EINVAL; ++ if (kst.f_namelen >= NAME_MAX) ++ err = au_br_init_wh(sb, br, perm); ++ else ++ pr_err("%pd(%s), unsupported namelen %ld\n", ++ au_br_dentry(br), ++ au_sbtype(au_br_dentry(br)->d_sb), kst.f_namelen); ++ ++out: ++ return err; ++} ++ ++/* initialize a new branch */ ++static int au_br_init(struct au_branch *br, struct super_block *sb, ++ struct au_opt_add *add) ++{ ++ int err; ++ struct inode *h_inode; ++ ++ err = 0; ++ memset(&br->br_xino, 0, sizeof(br->br_xino)); ++ mutex_init(&br->br_xino.xi_nondir_mtx); ++ br->br_perm = add->perm; ++ br->br_path = add->path; /* set first, path_get() later */ ++ spin_lock_init(&br->br_dykey_lock); ++ memset(br->br_dykey, 0, sizeof(br->br_dykey)); ++ atomic_set(&br->br_count, 0); ++ atomic_set(&br->br_xino_running, 0); ++ br->br_id = au_new_br_id(sb); ++ AuDebugOn(br->br_id < 0); ++ ++ if (au_br_writable(add->perm)) { ++ err = au_wbr_init(br, sb, add->perm); ++ if (unlikely(err)) ++ goto out_err; ++ } ++ ++ if (au_opt_test(au_mntflags(sb), XINO)) { ++ h_inode = d_inode(add->path.dentry); ++ err = au_xino_br(sb, br, h_inode->i_ino, ++ au_sbr(sb, 0)->br_xino.xi_file, /*do_test*/1); ++ if (unlikely(err)) { ++ AuDebugOn(br->br_xino.xi_file); ++ goto out_err; ++ } ++ } ++ ++ sysaufs_br_init(br); ++ path_get(&br->br_path); ++ goto out; /* success */ ++ ++out_err: ++ memset(&br->br_path, 0, sizeof(br->br_path)); ++out: ++ return err; ++} ++ ++static void au_br_do_add_brp(struct au_sbinfo *sbinfo, aufs_bindex_t bindex, ++ struct au_branch *br, aufs_bindex_t bend, ++ aufs_bindex_t amount) ++{ ++ struct au_branch **brp; ++ ++ AuRwMustWriteLock(&sbinfo->si_rwsem); ++ ++ brp = sbinfo->si_branch + bindex; ++ memmove(brp + 1, brp, sizeof(*brp) * amount); ++ *brp = br; ++ sbinfo->si_bend++; ++ if (unlikely(bend < 0)) ++ sbinfo->si_bend = 0; ++} ++ ++static void au_br_do_add_hdp(struct au_dinfo *dinfo, aufs_bindex_t bindex, ++ aufs_bindex_t bend, aufs_bindex_t amount) ++{ ++ struct au_hdentry *hdp; ++ ++ AuRwMustWriteLock(&dinfo->di_rwsem); ++ ++ hdp = dinfo->di_hdentry + bindex; ++ memmove(hdp + 1, hdp, sizeof(*hdp) * amount); ++ au_h_dentry_init(hdp); ++ dinfo->di_bend++; ++ if (unlikely(bend < 0)) ++ dinfo->di_bstart = 0; ++} ++ ++static void au_br_do_add_hip(struct au_iinfo *iinfo, aufs_bindex_t bindex, ++ aufs_bindex_t bend, aufs_bindex_t amount) ++{ ++ struct au_hinode *hip; ++ ++ AuRwMustWriteLock(&iinfo->ii_rwsem); ++ ++ hip = iinfo->ii_hinode + bindex; ++ memmove(hip + 1, hip, sizeof(*hip) * amount); ++ hip->hi_inode = NULL; ++ au_hn_init(hip); ++ iinfo->ii_bend++; ++ if (unlikely(bend < 0)) ++ iinfo->ii_bstart = 0; ++} ++ ++static void au_br_do_add(struct super_block *sb, struct au_branch *br, ++ aufs_bindex_t bindex) ++{ ++ struct dentry *root, *h_dentry; ++ struct inode *root_inode, *h_inode; ++ aufs_bindex_t bend, amount; ++ ++ root = sb->s_root; ++ root_inode = d_inode(root); ++ bend = au_sbend(sb); ++ amount = bend + 1 - bindex; ++ h_dentry = au_br_dentry(br); ++ au_sbilist_lock(); ++ au_br_do_add_brp(au_sbi(sb), bindex, br, bend, amount); ++ au_br_do_add_hdp(au_di(root), bindex, bend, amount); ++ au_br_do_add_hip(au_ii(root_inode), bindex, bend, amount); ++ au_set_h_dptr(root, bindex, dget(h_dentry)); ++ h_inode = d_inode(h_dentry); ++ au_set_h_iptr(root_inode, bindex, au_igrab(h_inode), /*flags*/0); ++ au_sbilist_unlock(); ++} ++ ++int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount) ++{ ++ int err; ++ aufs_bindex_t bend, add_bindex; ++ struct dentry *root, *h_dentry; ++ struct inode *root_inode; ++ struct au_branch *add_branch; ++ ++ root = sb->s_root; ++ root_inode = d_inode(root); ++ IMustLock(root_inode); ++ err = test_add(sb, add, remount); ++ if (unlikely(err < 0)) ++ goto out; ++ if (err) { ++ err = 0; ++ goto out; /* success */ ++ } ++ ++ bend = au_sbend(sb); ++ add_branch = au_br_alloc(sb, bend + 2, add->perm); ++ err = PTR_ERR(add_branch); ++ if (IS_ERR(add_branch)) ++ goto out; ++ ++ err = au_br_init(add_branch, sb, add); ++ if (unlikely(err)) { ++ au_br_do_free(add_branch); ++ goto out; ++ } ++ ++ add_bindex = add->bindex; ++ if (!remount) ++ au_br_do_add(sb, add_branch, add_bindex); ++ else { ++ sysaufs_brs_del(sb, add_bindex); ++ au_br_do_add(sb, add_branch, add_bindex); ++ sysaufs_brs_add(sb, add_bindex); ++ } ++ ++ h_dentry = add->path.dentry; ++ if (!add_bindex) { ++ au_cpup_attr_all(root_inode, /*force*/1); ++ sb->s_maxbytes = h_dentry->d_sb->s_maxbytes; ++ } else ++ au_add_nlink(root_inode, d_inode(h_dentry)); ++ ++ /* ++ * this test/set prevents aufs from handling unnecesary notify events ++ * of xino files, in case of re-adding a writable branch which was ++ * once detached from aufs. ++ */ ++ if (au_xino_brid(sb) < 0 ++ && au_br_writable(add_branch->br_perm) ++ && !au_test_fs_bad_xino(h_dentry->d_sb) ++ && add_branch->br_xino.xi_file ++ && add_branch->br_xino.xi_file->f_path.dentry->d_parent == h_dentry) ++ au_xino_brid_set(sb, add_branch->br_id); ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static unsigned long long au_farray_cb(void *a, ++ unsigned long long max __maybe_unused, ++ void *arg) ++{ ++ unsigned long long n; ++ struct file **p, *f; ++ struct au_sphlhead *files; ++ struct au_finfo *finfo; ++ struct super_block *sb = arg; ++ ++ n = 0; ++ p = a; ++ files = &au_sbi(sb)->si_files; ++ spin_lock(&files->spin); ++ hlist_for_each_entry(finfo, &files->head, fi_hlist) { ++ f = finfo->fi_file; ++ if (file_count(f) ++ && !special_file(file_inode(f)->i_mode)) { ++ get_file(f); ++ *p++ = f; ++ n++; ++ AuDebugOn(n > max); ++ } ++ } ++ spin_unlock(&files->spin); ++ ++ return n; ++} ++ ++static struct file **au_farray_alloc(struct super_block *sb, ++ unsigned long long *max) ++{ ++ *max = atomic_long_read(&au_sbi(sb)->si_nfiles); ++ return au_array_alloc(max, au_farray_cb, sb); ++} ++ ++static void au_farray_free(struct file **a, unsigned long long max) ++{ ++ unsigned long long ull; ++ ++ for (ull = 0; ull < max; ull++) ++ if (a[ull]) ++ fput(a[ull]); ++ au_array_free(a); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * delete a branch ++ */ ++ ++/* to show the line number, do not make it inlined function */ ++#define AuVerbose(do_info, fmt, ...) do { \ ++ if (do_info) \ ++ pr_info(fmt, ##__VA_ARGS__); \ ++} while (0) ++ ++static int au_test_ibusy(struct inode *inode, aufs_bindex_t bstart, ++ aufs_bindex_t bend) ++{ ++ return (inode && !S_ISDIR(inode->i_mode)) || bstart == bend; ++} ++ ++static int au_test_dbusy(struct dentry *dentry, aufs_bindex_t bstart, ++ aufs_bindex_t bend) ++{ ++ return au_test_ibusy(d_inode(dentry), bstart, bend); ++} ++ ++/* ++ * test if the branch is deletable or not. ++ */ ++static int test_dentry_busy(struct dentry *root, aufs_bindex_t bindex, ++ unsigned int sigen, const unsigned int verbose) ++{ ++ int err, i, j, ndentry; ++ aufs_bindex_t bstart, bend; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry *d; ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_dcsub_pages(&dpages, root, NULL, NULL); ++ if (unlikely(err)) ++ goto out_dpages; ++ ++ for (i = 0; !err && i < dpages.ndpage; i++) { ++ dpage = dpages.dpages + i; ++ ndentry = dpage->ndentry; ++ for (j = 0; !err && j < ndentry; j++) { ++ d = dpage->dentries[j]; ++ AuDebugOn(au_dcount(d) <= 0); ++ if (!au_digen_test(d, sigen)) { ++ di_read_lock_child(d, AuLock_IR); ++ if (unlikely(au_dbrange_test(d))) { ++ di_read_unlock(d, AuLock_IR); ++ continue; ++ } ++ } else { ++ di_write_lock_child(d); ++ if (unlikely(au_dbrange_test(d))) { ++ di_write_unlock(d); ++ continue; ++ } ++ err = au_reval_dpath(d, sigen); ++ if (!err) ++ di_downgrade_lock(d, AuLock_IR); ++ else { ++ di_write_unlock(d); ++ break; ++ } ++ } ++ ++ /* AuDbgDentry(d); */ ++ bstart = au_dbstart(d); ++ bend = au_dbend(d); ++ if (bstart <= bindex ++ && bindex <= bend ++ && au_h_dptr(d, bindex) ++ && au_test_dbusy(d, bstart, bend)) { ++ err = -EBUSY; ++ AuVerbose(verbose, "busy %pd\n", d); ++ AuDbgDentry(d); ++ } ++ di_read_unlock(d, AuLock_IR); ++ } ++ } ++ ++out_dpages: ++ au_dpages_free(&dpages); ++out: ++ return err; ++} ++ ++static int test_inode_busy(struct super_block *sb, aufs_bindex_t bindex, ++ unsigned int sigen, const unsigned int verbose) ++{ ++ int err; ++ unsigned long long max, ull; ++ struct inode *i, **array; ++ aufs_bindex_t bstart, bend; ++ ++ array = au_iarray_alloc(sb, &max); ++ err = PTR_ERR(array); ++ if (IS_ERR(array)) ++ goto out; ++ ++ err = 0; ++ AuDbg("b%d\n", bindex); ++ for (ull = 0; !err && ull < max; ull++) { ++ i = array[ull]; ++ if (unlikely(!i)) ++ break; ++ if (i->i_ino == AUFS_ROOT_INO) ++ continue; ++ ++ /* AuDbgInode(i); */ ++ if (au_iigen(i, NULL) == sigen) ++ ii_read_lock_child(i); ++ else { ++ ii_write_lock_child(i); ++ err = au_refresh_hinode_self(i); ++ au_iigen_dec(i); ++ if (!err) ++ ii_downgrade_lock(i); ++ else { ++ ii_write_unlock(i); ++ break; ++ } ++ } ++ ++ bstart = au_ibstart(i); ++ bend = au_ibend(i); ++ if (bstart <= bindex ++ && bindex <= bend ++ && au_h_iptr(i, bindex) ++ && au_test_ibusy(i, bstart, bend)) { ++ err = -EBUSY; ++ AuVerbose(verbose, "busy i%lu\n", i->i_ino); ++ AuDbgInode(i); ++ } ++ ii_read_unlock(i); ++ } ++ au_iarray_free(array, max); ++ ++out: ++ return err; ++} ++ ++static int test_children_busy(struct dentry *root, aufs_bindex_t bindex, ++ const unsigned int verbose) ++{ ++ int err; ++ unsigned int sigen; ++ ++ sigen = au_sigen(root->d_sb); ++ DiMustNoWaiters(root); ++ IiMustNoWaiters(d_inode(root)); ++ di_write_unlock(root); ++ err = test_dentry_busy(root, bindex, sigen, verbose); ++ if (!err) ++ err = test_inode_busy(root->d_sb, bindex, sigen, verbose); ++ di_write_lock_child(root); /* aufs_write_lock() calls ..._child() */ ++ ++ return err; ++} ++ ++static int test_dir_busy(struct file *file, aufs_bindex_t br_id, ++ struct file **to_free, int *idx) ++{ ++ int err; ++ unsigned char matched, root; ++ aufs_bindex_t bindex, bend; ++ struct au_fidir *fidir; ++ struct au_hfile *hfile; ++ ++ err = 0; ++ root = IS_ROOT(file->f_path.dentry); ++ if (root) { ++ get_file(file); ++ to_free[*idx] = file; ++ (*idx)++; ++ goto out; ++ } ++ ++ matched = 0; ++ fidir = au_fi(file)->fi_hdir; ++ AuDebugOn(!fidir); ++ bend = au_fbend_dir(file); ++ for (bindex = au_fbstart(file); bindex <= bend; bindex++) { ++ hfile = fidir->fd_hfile + bindex; ++ if (!hfile->hf_file) ++ continue; ++ ++ if (hfile->hf_br->br_id == br_id) { ++ matched = 1; ++ break; ++ } ++ } ++ if (matched) ++ err = -EBUSY; ++ ++out: ++ return err; ++} ++ ++static int test_file_busy(struct super_block *sb, aufs_bindex_t br_id, ++ struct file **to_free, int opened) ++{ ++ int err, idx; ++ unsigned long long ull, max; ++ aufs_bindex_t bstart; ++ struct file *file, **array; ++ struct dentry *root; ++ struct au_hfile *hfile; ++ ++ array = au_farray_alloc(sb, &max); ++ err = PTR_ERR(array); ++ if (IS_ERR(array)) ++ goto out; ++ ++ err = 0; ++ idx = 0; ++ root = sb->s_root; ++ di_write_unlock(root); ++ for (ull = 0; ull < max; ull++) { ++ file = array[ull]; ++ if (unlikely(!file)) ++ break; ++ ++ /* AuDbg("%pD\n", file); */ ++ fi_read_lock(file); ++ bstart = au_fbstart(file); ++ if (!d_is_dir(file->f_path.dentry)) { ++ hfile = &au_fi(file)->fi_htop; ++ if (hfile->hf_br->br_id == br_id) ++ err = -EBUSY; ++ } else ++ err = test_dir_busy(file, br_id, to_free, &idx); ++ fi_read_unlock(file); ++ if (unlikely(err)) ++ break; ++ } ++ di_write_lock_child(root); ++ au_farray_free(array, max); ++ AuDebugOn(idx > opened); ++ ++out: ++ return err; ++} ++ ++static void br_del_file(struct file **to_free, unsigned long long opened, ++ aufs_bindex_t br_id) ++{ ++ unsigned long long ull; ++ aufs_bindex_t bindex, bstart, bend, bfound; ++ struct file *file; ++ struct au_fidir *fidir; ++ struct au_hfile *hfile; ++ ++ for (ull = 0; ull < opened; ull++) { ++ file = to_free[ull]; ++ if (unlikely(!file)) ++ break; ++ ++ /* AuDbg("%pD\n", file); */ ++ AuDebugOn(!d_is_dir(file->f_path.dentry)); ++ bfound = -1; ++ fidir = au_fi(file)->fi_hdir; ++ AuDebugOn(!fidir); ++ fi_write_lock(file); ++ bstart = au_fbstart(file); ++ bend = au_fbend_dir(file); ++ for (bindex = bstart; bindex <= bend; bindex++) { ++ hfile = fidir->fd_hfile + bindex; ++ if (!hfile->hf_file) ++ continue; ++ ++ if (hfile->hf_br->br_id == br_id) { ++ bfound = bindex; ++ break; ++ } ++ } ++ AuDebugOn(bfound < 0); ++ au_set_h_fptr(file, bfound, NULL); ++ if (bfound == bstart) { ++ for (bstart++; bstart <= bend; bstart++) ++ if (au_hf_dir(file, bstart)) { ++ au_set_fbstart(file, bstart); ++ break; ++ } ++ } ++ fi_write_unlock(file); ++ } ++} ++ ++static void au_br_do_del_brp(struct au_sbinfo *sbinfo, ++ const aufs_bindex_t bindex, ++ const aufs_bindex_t bend) ++{ ++ struct au_branch **brp, **p; ++ ++ AuRwMustWriteLock(&sbinfo->si_rwsem); ++ ++ brp = sbinfo->si_branch + bindex; ++ if (bindex < bend) ++ memmove(brp, brp + 1, sizeof(*brp) * (bend - bindex)); ++ sbinfo->si_branch[0 + bend] = NULL; ++ sbinfo->si_bend--; ++ ++ p = krealloc(sbinfo->si_branch, sizeof(*p) * bend, AuGFP_SBILIST); ++ if (p) ++ sbinfo->si_branch = p; ++ /* harmless error */ ++} ++ ++static void au_br_do_del_hdp(struct au_dinfo *dinfo, const aufs_bindex_t bindex, ++ const aufs_bindex_t bend) ++{ ++ struct au_hdentry *hdp, *p; ++ ++ AuRwMustWriteLock(&dinfo->di_rwsem); ++ ++ hdp = dinfo->di_hdentry; ++ if (bindex < bend) ++ memmove(hdp + bindex, hdp + bindex + 1, ++ sizeof(*hdp) * (bend - bindex)); ++ hdp[0 + bend].hd_dentry = NULL; ++ dinfo->di_bend--; ++ ++ p = krealloc(hdp, sizeof(*p) * bend, AuGFP_SBILIST); ++ if (p) ++ dinfo->di_hdentry = p; ++ /* harmless error */ ++} ++ ++static void au_br_do_del_hip(struct au_iinfo *iinfo, const aufs_bindex_t bindex, ++ const aufs_bindex_t bend) ++{ ++ struct au_hinode *hip, *p; ++ ++ AuRwMustWriteLock(&iinfo->ii_rwsem); ++ ++ hip = iinfo->ii_hinode + bindex; ++ if (bindex < bend) ++ memmove(hip, hip + 1, sizeof(*hip) * (bend - bindex)); ++ iinfo->ii_hinode[0 + bend].hi_inode = NULL; ++ au_hn_init(iinfo->ii_hinode + bend); ++ iinfo->ii_bend--; ++ ++ p = krealloc(iinfo->ii_hinode, sizeof(*p) * bend, AuGFP_SBILIST); ++ if (p) ++ iinfo->ii_hinode = p; ++ /* harmless error */ ++} ++ ++static void au_br_do_del(struct super_block *sb, aufs_bindex_t bindex, ++ struct au_branch *br) ++{ ++ aufs_bindex_t bend; ++ struct au_sbinfo *sbinfo; ++ struct dentry *root, *h_root; ++ struct inode *inode, *h_inode; ++ struct au_hinode *hinode; ++ ++ SiMustWriteLock(sb); ++ ++ root = sb->s_root; ++ inode = d_inode(root); ++ sbinfo = au_sbi(sb); ++ bend = sbinfo->si_bend; ++ ++ h_root = au_h_dptr(root, bindex); ++ hinode = au_hi(inode, bindex); ++ h_inode = au_igrab(hinode->hi_inode); ++ au_hiput(hinode); ++ ++ au_sbilist_lock(); ++ au_br_do_del_brp(sbinfo, bindex, bend); ++ au_br_do_del_hdp(au_di(root), bindex, bend); ++ au_br_do_del_hip(au_ii(inode), bindex, bend); ++ au_sbilist_unlock(); ++ ++ dput(h_root); ++ iput(h_inode); ++ au_br_do_free(br); ++} ++ ++static unsigned long long empty_cb(void *array, unsigned long long max, ++ void *arg) ++{ ++ return max; ++} ++ ++int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount) ++{ ++ int err, rerr, i; ++ unsigned long long opened; ++ unsigned int mnt_flags; ++ aufs_bindex_t bindex, bend, br_id; ++ unsigned char do_wh, verbose; ++ struct au_branch *br; ++ struct au_wbr *wbr; ++ struct dentry *root; ++ struct file **to_free; ++ ++ err = 0; ++ opened = 0; ++ to_free = NULL; ++ root = sb->s_root; ++ bindex = au_find_dbindex(root, del->h_path.dentry); ++ if (bindex < 0) { ++ if (remount) ++ goto out; /* success */ ++ err = -ENOENT; ++ pr_err("%s no such branch\n", del->pathname); ++ goto out; ++ } ++ AuDbg("bindex b%d\n", bindex); ++ ++ err = -EBUSY; ++ mnt_flags = au_mntflags(sb); ++ verbose = !!au_opt_test(mnt_flags, VERBOSE); ++ bend = au_sbend(sb); ++ if (unlikely(!bend)) { ++ AuVerbose(verbose, "no more branches left\n"); ++ goto out; ++ } ++ br = au_sbr(sb, bindex); ++ AuDebugOn(!path_equal(&br->br_path, &del->h_path)); ++ ++ br_id = br->br_id; ++ opened = atomic_read(&br->br_count); ++ if (unlikely(opened)) { ++ to_free = au_array_alloc(&opened, empty_cb, NULL); ++ err = PTR_ERR(to_free); ++ if (IS_ERR(to_free)) ++ goto out; ++ ++ err = test_file_busy(sb, br_id, to_free, opened); ++ if (unlikely(err)) { ++ AuVerbose(verbose, "%llu file(s) opened\n", opened); ++ goto out; ++ } ++ } ++ ++ wbr = br->br_wbr; ++ do_wh = wbr && (wbr->wbr_whbase || wbr->wbr_plink || wbr->wbr_orph); ++ if (do_wh) { ++ /* instead of WbrWhMustWriteLock(wbr) */ ++ SiMustWriteLock(sb); ++ for (i = 0; i < AuBrWh_Last; i++) { ++ dput(wbr->wbr_wh[i]); ++ wbr->wbr_wh[i] = NULL; ++ } ++ } ++ ++ err = test_children_busy(root, bindex, verbose); ++ if (unlikely(err)) { ++ if (do_wh) ++ goto out_wh; ++ goto out; ++ } ++ ++ err = 0; ++ if (to_free) { ++ /* ++ * now we confirmed the branch is deletable. ++ * let's free the remaining opened dirs on the branch. ++ */ ++ di_write_unlock(root); ++ br_del_file(to_free, opened, br_id); ++ di_write_lock_child(root); ++ } ++ ++ if (!remount) ++ au_br_do_del(sb, bindex, br); ++ else { ++ sysaufs_brs_del(sb, bindex); ++ au_br_do_del(sb, bindex, br); ++ sysaufs_brs_add(sb, bindex); ++ } ++ ++ if (!bindex) { ++ au_cpup_attr_all(d_inode(root), /*force*/1); ++ sb->s_maxbytes = au_sbr_sb(sb, 0)->s_maxbytes; ++ } else ++ au_sub_nlink(d_inode(root), d_inode(del->h_path.dentry)); ++ if (au_opt_test(mnt_flags, PLINK)) ++ au_plink_half_refresh(sb, br_id); ++ ++ if (au_xino_brid(sb) == br_id) ++ au_xino_brid_set(sb, -1); ++ goto out; /* success */ ++ ++out_wh: ++ /* revert */ ++ rerr = au_br_init_wh(sb, br, br->br_perm); ++ if (rerr) ++ pr_warn("failed re-creating base whiteout, %s. (%d)\n", ++ del->pathname, rerr); ++out: ++ if (to_free) ++ au_farray_free(to_free, opened); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_ibusy(struct super_block *sb, struct aufs_ibusy __user *arg) ++{ ++ int err; ++ aufs_bindex_t bstart, bend; ++ struct aufs_ibusy ibusy; ++ struct inode *inode, *h_inode; ++ ++ err = -EPERM; ++ if (unlikely(!capable(CAP_SYS_ADMIN))) ++ goto out; ++ ++ err = copy_from_user(&ibusy, arg, sizeof(ibusy)); ++ if (!err) ++ err = !access_ok(VERIFY_WRITE, &arg->h_ino, sizeof(arg->h_ino)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ goto out; ++ } ++ ++ err = -EINVAL; ++ si_read_lock(sb, AuLock_FLUSH); ++ if (unlikely(ibusy.bindex < 0 || ibusy.bindex > au_sbend(sb))) ++ goto out_unlock; ++ ++ err = 0; ++ ibusy.h_ino = 0; /* invalid */ ++ inode = ilookup(sb, ibusy.ino); ++ if (!inode ++ || inode->i_ino == AUFS_ROOT_INO ++ || is_bad_inode(inode)) ++ goto out_unlock; ++ ++ ii_read_lock_child(inode); ++ bstart = au_ibstart(inode); ++ bend = au_ibend(inode); ++ if (bstart <= ibusy.bindex && ibusy.bindex <= bend) { ++ h_inode = au_h_iptr(inode, ibusy.bindex); ++ if (h_inode && au_test_ibusy(inode, bstart, bend)) ++ ibusy.h_ino = h_inode->i_ino; ++ } ++ ii_read_unlock(inode); ++ iput(inode); ++ ++out_unlock: ++ si_read_unlock(sb); ++ if (!err) { ++ err = __put_user(ibusy.h_ino, &arg->h_ino); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ } ++ } ++out: ++ return err; ++} ++ ++long au_ibusy_ioctl(struct file *file, unsigned long arg) ++{ ++ return au_ibusy(file->f_path.dentry->d_sb, (void __user *)arg); ++} ++ ++#ifdef CONFIG_COMPAT ++long au_ibusy_compat_ioctl(struct file *file, unsigned long arg) ++{ ++ return au_ibusy(file->f_path.dentry->d_sb, compat_ptr(arg)); ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * change a branch permission ++ */ ++ ++static void au_warn_ima(void) ++{ ++#ifdef CONFIG_IMA ++ /* since it doesn't support mark_files_ro() */ ++ AuWarn1("RW -> RO makes IMA to produce wrong message\n"); ++#endif ++} ++ ++static int do_need_sigen_inc(int a, int b) ++{ ++ return au_br_whable(a) && !au_br_whable(b); ++} ++ ++static int need_sigen_inc(int old, int new) ++{ ++ return do_need_sigen_inc(old, new) ++ || do_need_sigen_inc(new, old); ++} ++ ++static int au_br_mod_files_ro(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ int err, do_warn; ++ unsigned int mnt_flags; ++ unsigned long long ull, max; ++ aufs_bindex_t br_id; ++ unsigned char verbose, writer; ++ struct file *file, *hf, **array; ++ struct au_hfile *hfile; ++ ++ mnt_flags = au_mntflags(sb); ++ verbose = !!au_opt_test(mnt_flags, VERBOSE); ++ ++ array = au_farray_alloc(sb, &max); ++ err = PTR_ERR(array); ++ if (IS_ERR(array)) ++ goto out; ++ ++ do_warn = 0; ++ br_id = au_sbr_id(sb, bindex); ++ for (ull = 0; ull < max; ull++) { ++ file = array[ull]; ++ if (unlikely(!file)) ++ break; ++ ++ /* AuDbg("%pD\n", file); */ ++ fi_read_lock(file); ++ if (unlikely(au_test_mmapped(file))) { ++ err = -EBUSY; ++ AuVerbose(verbose, "mmapped %pD\n", file); ++ AuDbgFile(file); ++ FiMustNoWaiters(file); ++ fi_read_unlock(file); ++ goto out_array; ++ } ++ ++ hfile = &au_fi(file)->fi_htop; ++ hf = hfile->hf_file; ++ if (!d_is_reg(file->f_path.dentry) ++ || !(file->f_mode & FMODE_WRITE) ++ || hfile->hf_br->br_id != br_id ++ || !(hf->f_mode & FMODE_WRITE)) ++ array[ull] = NULL; ++ else { ++ do_warn = 1; ++ get_file(file); ++ } ++ ++ FiMustNoWaiters(file); ++ fi_read_unlock(file); ++ fput(file); ++ } ++ ++ err = 0; ++ if (do_warn) ++ au_warn_ima(); ++ ++ for (ull = 0; ull < max; ull++) { ++ file = array[ull]; ++ if (!file) ++ continue; ++ ++ /* todo: already flushed? */ ++ /* ++ * fs/super.c:mark_files_ro() is gone, but aufs keeps its ++ * approach which resets f_mode and calls mnt_drop_write() and ++ * file_release_write() for each file, because the branch ++ * attribute in aufs world is totally different from the native ++ * fs rw/ro mode. ++ */ ++ /* fi_read_lock(file); */ ++ hfile = &au_fi(file)->fi_htop; ++ hf = hfile->hf_file; ++ /* fi_read_unlock(file); */ ++ spin_lock(&hf->f_lock); ++ writer = !!(hf->f_mode & FMODE_WRITER); ++ hf->f_mode &= ~(FMODE_WRITE | FMODE_WRITER); ++ spin_unlock(&hf->f_lock); ++ if (writer) { ++ put_write_access(file_inode(hf)); ++ __mnt_drop_write(hf->f_path.mnt); ++ } ++ } ++ ++out_array: ++ au_farray_free(array, max); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount, ++ int *do_refresh) ++{ ++ int err, rerr; ++ aufs_bindex_t bindex; ++ struct dentry *root; ++ struct au_branch *br; ++ struct au_br_fhsm *bf; ++ ++ root = sb->s_root; ++ bindex = au_find_dbindex(root, mod->h_root); ++ if (bindex < 0) { ++ if (remount) ++ return 0; /* success */ ++ err = -ENOENT; ++ pr_err("%s no such branch\n", mod->path); ++ goto out; ++ } ++ AuDbg("bindex b%d\n", bindex); ++ ++ err = test_br(d_inode(mod->h_root), mod->perm, mod->path); ++ if (unlikely(err)) ++ goto out; ++ ++ br = au_sbr(sb, bindex); ++ AuDebugOn(mod->h_root != au_br_dentry(br)); ++ if (br->br_perm == mod->perm) ++ return 0; /* success */ ++ ++ /* pre-allocate for non-fhsm --> fhsm */ ++ bf = NULL; ++ if (!au_br_fhsm(br->br_perm) && au_br_fhsm(mod->perm)) { ++ err = au_fhsm_br_alloc(br); ++ if (unlikely(err)) ++ goto out; ++ bf = br->br_fhsm; ++ br->br_fhsm = NULL; ++ } ++ ++ if (au_br_writable(br->br_perm)) { ++ /* remove whiteout base */ ++ err = au_br_init_wh(sb, br, mod->perm); ++ if (unlikely(err)) ++ goto out_bf; ++ ++ if (!au_br_writable(mod->perm)) { ++ /* rw --> ro, file might be mmapped */ ++ DiMustNoWaiters(root); ++ IiMustNoWaiters(d_inode(root)); ++ di_write_unlock(root); ++ err = au_br_mod_files_ro(sb, bindex); ++ /* aufs_write_lock() calls ..._child() */ ++ di_write_lock_child(root); ++ ++ if (unlikely(err)) { ++ rerr = -ENOMEM; ++ br->br_wbr = kmalloc(sizeof(*br->br_wbr), ++ GFP_NOFS); ++ if (br->br_wbr) ++ rerr = au_wbr_init(br, sb, br->br_perm); ++ if (unlikely(rerr)) { ++ AuIOErr("nested error %d (%d)\n", ++ rerr, err); ++ br->br_perm = mod->perm; ++ } ++ } ++ } ++ } else if (au_br_writable(mod->perm)) { ++ /* ro --> rw */ ++ err = -ENOMEM; ++ br->br_wbr = kmalloc(sizeof(*br->br_wbr), GFP_NOFS); ++ if (br->br_wbr) { ++ err = au_wbr_init(br, sb, mod->perm); ++ if (unlikely(err)) { ++ kfree(br->br_wbr); ++ br->br_wbr = NULL; ++ } ++ } ++ } ++ if (unlikely(err)) ++ goto out_bf; ++ ++ if (au_br_fhsm(br->br_perm)) { ++ if (!au_br_fhsm(mod->perm)) { ++ /* fhsm --> non-fhsm */ ++ au_br_fhsm_fin(br->br_fhsm); ++ kfree(br->br_fhsm); ++ br->br_fhsm = NULL; ++ } ++ } else if (au_br_fhsm(mod->perm)) ++ /* non-fhsm --> fhsm */ ++ br->br_fhsm = bf; ++ ++ *do_refresh |= need_sigen_inc(br->br_perm, mod->perm); ++ br->br_perm = mod->perm; ++ goto out; /* success */ ++ ++out_bf: ++ kfree(bf); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_br_stfs(struct au_branch *br, struct aufs_stfs *stfs) ++{ ++ int err; ++ struct kstatfs kstfs; ++ ++ err = vfs_statfs(&br->br_path, &kstfs); ++ if (!err) { ++ stfs->f_blocks = kstfs.f_blocks; ++ stfs->f_bavail = kstfs.f_bavail; ++ stfs->f_files = kstfs.f_files; ++ stfs->f_ffree = kstfs.f_ffree; ++ } ++ ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/branch.h linux-4.1.10/fs/aufs/branch.h +--- linux-4.1.10.orig/fs/aufs/branch.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/branch.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,279 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * branch filesystems and xino for them ++ */ ++ ++#ifndef __AUFS_BRANCH_H__ ++#define __AUFS_BRANCH_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/mount.h> ++#include "dynop.h" ++#include "rwsem.h" ++#include "super.h" ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* a xino file */ ++struct au_xino_file { ++ struct file *xi_file; ++ struct mutex xi_nondir_mtx; ++ ++ /* todo: make xino files an array to support huge inode number */ ++ ++#ifdef CONFIG_DEBUG_FS ++ struct dentry *xi_dbgaufs; ++#endif ++}; ++ ++/* File-based Hierarchical Storage Management */ ++struct au_br_fhsm { ++#ifdef CONFIG_AUFS_FHSM ++ struct mutex bf_lock; ++ unsigned long bf_jiffy; ++ struct aufs_stfs bf_stfs; ++ int bf_readable; ++#endif ++}; ++ ++/* members for writable branch only */ ++enum {AuBrWh_BASE, AuBrWh_PLINK, AuBrWh_ORPH, AuBrWh_Last}; ++struct au_wbr { ++ struct au_rwsem wbr_wh_rwsem; ++ struct dentry *wbr_wh[AuBrWh_Last]; ++ atomic_t wbr_wh_running; ++#define wbr_whbase wbr_wh[AuBrWh_BASE] /* whiteout base */ ++#define wbr_plink wbr_wh[AuBrWh_PLINK] /* pseudo-link dir */ ++#define wbr_orph wbr_wh[AuBrWh_ORPH] /* dir for orphans */ ++ ++ /* mfs mode */ ++ unsigned long long wbr_bytes; ++}; ++ ++/* ext2 has 3 types of operations at least, ext3 has 4 */ ++#define AuBrDynOp (AuDyLast * 4) ++ ++#ifdef CONFIG_AUFS_HFSNOTIFY ++/* support for asynchronous destruction */ ++struct au_br_hfsnotify { ++ struct fsnotify_group *hfsn_group; ++}; ++#endif ++ ++/* sysfs entries */ ++struct au_brsysfs { ++ char name[16]; ++ struct attribute attr; ++}; ++ ++enum { ++ AuBrSysfs_BR, ++ AuBrSysfs_BRID, ++ AuBrSysfs_Last ++}; ++ ++/* protected by superblock rwsem */ ++struct au_branch { ++ struct au_xino_file br_xino; ++ ++ aufs_bindex_t br_id; ++ ++ int br_perm; ++ struct path br_path; ++ spinlock_t br_dykey_lock; ++ struct au_dykey *br_dykey[AuBrDynOp]; ++ atomic_t br_count; ++ ++ struct au_wbr *br_wbr; ++ struct au_br_fhsm *br_fhsm; ++ ++ /* xino truncation */ ++ atomic_t br_xino_running; ++ ++#ifdef CONFIG_AUFS_HFSNOTIFY ++ struct au_br_hfsnotify *br_hfsn; ++#endif ++ ++#ifdef CONFIG_SYSFS ++ /* entries under sysfs per mount-point */ ++ struct au_brsysfs br_sysfs[AuBrSysfs_Last]; ++#endif ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct vfsmount *au_br_mnt(struct au_branch *br) ++{ ++ return br->br_path.mnt; ++} ++ ++static inline struct dentry *au_br_dentry(struct au_branch *br) ++{ ++ return br->br_path.dentry; ++} ++ ++static inline struct super_block *au_br_sb(struct au_branch *br) ++{ ++ return au_br_mnt(br)->mnt_sb; ++} ++ ++static inline int au_br_rdonly(struct au_branch *br) ++{ ++ return ((au_br_sb(br)->s_flags & MS_RDONLY) ++ || !au_br_writable(br->br_perm)) ++ ? -EROFS : 0; ++} ++ ++static inline int au_br_hnotifyable(int brperm __maybe_unused) ++{ ++#ifdef CONFIG_AUFS_HNOTIFY ++ return !(brperm & AuBrPerm_RR); ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_br_test_oflag(int oflag, struct au_branch *br) ++{ ++ int err, exec_flag; ++ ++ err = 0; ++ exec_flag = oflag & __FMODE_EXEC; ++ if (unlikely(exec_flag && (au_br_mnt(br)->mnt_flags & MNT_NOEXEC))) ++ err = -EACCES; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* branch.c */ ++struct au_sbinfo; ++void au_br_free(struct au_sbinfo *sinfo); ++int au_br_index(struct super_block *sb, aufs_bindex_t br_id); ++struct au_opt_add; ++int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount); ++struct au_opt_del; ++int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount); ++long au_ibusy_ioctl(struct file *file, unsigned long arg); ++#ifdef CONFIG_COMPAT ++long au_ibusy_compat_ioctl(struct file *file, unsigned long arg); ++#endif ++struct au_opt_mod; ++int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount, ++ int *do_refresh); ++struct aufs_stfs; ++int au_br_stfs(struct au_branch *br, struct aufs_stfs *stfs); ++ ++/* xino.c */ ++static const loff_t au_loff_max = LLONG_MAX; ++ ++int au_xib_trunc(struct super_block *sb); ++ssize_t xino_fread(vfs_readf_t func, struct file *file, void *buf, size_t size, ++ loff_t *pos); ++ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos); ++struct file *au_xino_create2(struct file *base_file, struct file *copy_src); ++struct file *au_xino_create(struct super_block *sb, char *fname, int silent); ++ino_t au_xino_new_ino(struct super_block *sb); ++void au_xino_delete_inode(struct inode *inode, const int unlinked); ++int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t ino); ++int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t *ino); ++int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t hino, ++ struct file *base_file, int do_test); ++int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex); ++ ++struct au_opt_xino; ++int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount); ++void au_xino_clr(struct super_block *sb); ++struct file *au_xino_def(struct super_block *sb); ++int au_xino_path(struct seq_file *seq, struct file *file); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* Superblock to branch */ ++static inline ++aufs_bindex_t au_sbr_id(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ return au_sbr(sb, bindex)->br_id; ++} ++ ++static inline ++struct vfsmount *au_sbr_mnt(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ return au_br_mnt(au_sbr(sb, bindex)); ++} ++ ++static inline ++struct super_block *au_sbr_sb(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ return au_br_sb(au_sbr(sb, bindex)); ++} ++ ++static inline void au_sbr_put(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ atomic_dec(&au_sbr(sb, bindex)->br_count); ++} ++ ++static inline int au_sbr_perm(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ return au_sbr(sb, bindex)->br_perm; ++} ++ ++static inline int au_sbr_whable(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ return au_br_whable(au_sbr_perm(sb, bindex)); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * wbr_wh_read_lock, wbr_wh_write_lock ++ * wbr_wh_read_unlock, wbr_wh_write_unlock, wbr_wh_downgrade_lock ++ */ ++AuSimpleRwsemFuncs(wbr_wh, struct au_wbr *wbr, &wbr->wbr_wh_rwsem); ++ ++#define WbrWhMustNoWaiters(wbr) AuRwMustNoWaiters(&wbr->wbr_wh_rwsem) ++#define WbrWhMustAnyLock(wbr) AuRwMustAnyLock(&wbr->wbr_wh_rwsem) ++#define WbrWhMustWriteLock(wbr) AuRwMustWriteLock(&wbr->wbr_wh_rwsem) ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_FHSM ++static inline void au_br_fhsm_init(struct au_br_fhsm *brfhsm) ++{ ++ mutex_init(&brfhsm->bf_lock); ++ brfhsm->bf_jiffy = 0; ++ brfhsm->bf_readable = 0; ++} ++ ++static inline void au_br_fhsm_fin(struct au_br_fhsm *brfhsm) ++{ ++ mutex_destroy(&brfhsm->bf_lock); ++} ++#else ++AuStubVoid(au_br_fhsm_init, struct au_br_fhsm *brfhsm) ++AuStubVoid(au_br_fhsm_fin, struct au_br_fhsm *brfhsm) ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_BRANCH_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/conf.mk linux-4.1.10/fs/aufs/conf.mk +--- linux-4.1.10.orig/fs/aufs/conf.mk 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/conf.mk 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,38 @@ ++ ++AuConfStr = CONFIG_AUFS_FS=${CONFIG_AUFS_FS} ++ ++define AuConf ++ifdef ${1} ++AuConfStr += ${1}=${${1}} ++endif ++endef ++ ++AuConfAll = BRANCH_MAX_127 BRANCH_MAX_511 BRANCH_MAX_1023 BRANCH_MAX_32767 \ ++ SBILIST \ ++ HNOTIFY HFSNOTIFY \ ++ EXPORT INO_T_64 \ ++ XATTR \ ++ FHSM \ ++ RDU \ ++ SHWH \ ++ BR_RAMFS \ ++ BR_FUSE POLL \ ++ BR_HFSPLUS \ ++ BDEV_LOOP \ ++ DEBUG MAGIC_SYSRQ ++$(foreach i, ${AuConfAll}, \ ++ $(eval $(call AuConf,CONFIG_AUFS_${i}))) ++ ++AuConfName = ${obj}/conf.str ++${AuConfName}.tmp: FORCE ++ @echo ${AuConfStr} | tr ' ' '\n' | sed -e 's/^/"/' -e 's/$$/\\n"/' > $@ ++${AuConfName}: ${AuConfName}.tmp ++ @diff -q $< $@ > /dev/null 2>&1 || { \ ++ echo ' GEN ' $@; \ ++ cp -p $< $@; \ ++ } ++FORCE: ++clean-files += ${AuConfName} ${AuConfName}.tmp ++${obj}/sysfs.o: ${AuConfName} ++ ++-include ${srctree}/${src}/conf_priv.mk +diff -Nur linux-4.1.10.orig/fs/aufs/cpup.c linux-4.1.10/fs/aufs/cpup.c +--- linux-4.1.10.orig/fs/aufs/cpup.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/cpup.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1319 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * copy-up functions, see wbr_policy.c for copy-down ++ */ ++ ++#include <linux/fs_stack.h> ++#include <linux/mm.h> ++#include "aufs.h" ++ ++void au_cpup_attr_flags(struct inode *dst, unsigned int iflags) ++{ ++ const unsigned int mask = S_DEAD | S_SWAPFILE | S_PRIVATE ++ | S_NOATIME | S_NOCMTIME | S_AUTOMOUNT; ++ ++ BUILD_BUG_ON(sizeof(iflags) != sizeof(dst->i_flags)); ++ ++ dst->i_flags |= iflags & ~mask; ++ if (au_test_fs_notime(dst->i_sb)) ++ dst->i_flags |= S_NOATIME | S_NOCMTIME; ++} ++ ++void au_cpup_attr_timesizes(struct inode *inode) ++{ ++ struct inode *h_inode; ++ ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ fsstack_copy_attr_times(inode, h_inode); ++ fsstack_copy_inode_size(inode, h_inode); ++} ++ ++void au_cpup_attr_nlink(struct inode *inode, int force) ++{ ++ struct inode *h_inode; ++ struct super_block *sb; ++ aufs_bindex_t bindex, bend; ++ ++ sb = inode->i_sb; ++ bindex = au_ibstart(inode); ++ h_inode = au_h_iptr(inode, bindex); ++ if (!force ++ && !S_ISDIR(h_inode->i_mode) ++ && au_opt_test(au_mntflags(sb), PLINK) ++ && au_plink_test(inode)) ++ return; ++ ++ /* ++ * 0 can happen in revalidating. ++ * h_inode->i_mutex may not be held here, but it is harmless since once ++ * i_nlink reaches 0, it will never become positive except O_TMPFILE ++ * case. ++ * todo: O_TMPFILE+linkat(AT_SYMLINK_FOLLOW) bypassing aufs may cause ++ * the incorrect link count. ++ */ ++ set_nlink(inode, h_inode->i_nlink); ++ ++ /* ++ * fewer nlink makes find(1) noisy, but larger nlink doesn't. ++ * it may includes whplink directory. ++ */ ++ if (S_ISDIR(h_inode->i_mode)) { ++ bend = au_ibend(inode); ++ for (bindex++; bindex <= bend; bindex++) { ++ h_inode = au_h_iptr(inode, bindex); ++ if (h_inode) ++ au_add_nlink(inode, h_inode); ++ } ++ } ++} ++ ++void au_cpup_attr_changeable(struct inode *inode) ++{ ++ struct inode *h_inode; ++ ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ inode->i_mode = h_inode->i_mode; ++ inode->i_uid = h_inode->i_uid; ++ inode->i_gid = h_inode->i_gid; ++ au_cpup_attr_timesizes(inode); ++ au_cpup_attr_flags(inode, h_inode->i_flags); ++} ++ ++void au_cpup_igen(struct inode *inode, struct inode *h_inode) ++{ ++ struct au_iinfo *iinfo = au_ii(inode); ++ ++ IiMustWriteLock(inode); ++ ++ iinfo->ii_higen = h_inode->i_generation; ++ iinfo->ii_hsb1 = h_inode->i_sb; ++} ++ ++void au_cpup_attr_all(struct inode *inode, int force) ++{ ++ struct inode *h_inode; ++ ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ au_cpup_attr_changeable(inode); ++ if (inode->i_nlink > 0) ++ au_cpup_attr_nlink(inode, force); ++ inode->i_rdev = h_inode->i_rdev; ++ inode->i_blkbits = h_inode->i_blkbits; ++ au_cpup_igen(inode, h_inode); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* Note: dt_dentry and dt_h_dentry are not dget/dput-ed */ ++ ++/* keep the timestamps of the parent dir when cpup */ ++void au_dtime_store(struct au_dtime *dt, struct dentry *dentry, ++ struct path *h_path) ++{ ++ struct inode *h_inode; ++ ++ dt->dt_dentry = dentry; ++ dt->dt_h_path = *h_path; ++ h_inode = d_inode(h_path->dentry); ++ dt->dt_atime = h_inode->i_atime; ++ dt->dt_mtime = h_inode->i_mtime; ++ /* smp_mb(); */ ++} ++ ++void au_dtime_revert(struct au_dtime *dt) ++{ ++ struct iattr attr; ++ int err; ++ ++ attr.ia_atime = dt->dt_atime; ++ attr.ia_mtime = dt->dt_mtime; ++ attr.ia_valid = ATTR_FORCE | ATTR_MTIME | ATTR_MTIME_SET ++ | ATTR_ATIME | ATTR_ATIME_SET; ++ ++ /* no delegation since this is a directory */ ++ err = vfsub_notify_change(&dt->dt_h_path, &attr, /*delegated*/NULL); ++ if (unlikely(err)) ++ pr_warn("restoring timestamps failed(%d). ignored\n", err); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* internal use only */ ++struct au_cpup_reg_attr { ++ int valid; ++ struct kstat st; ++ unsigned int iflags; /* inode->i_flags */ ++}; ++ ++static noinline_for_stack ++int cpup_iattr(struct dentry *dst, aufs_bindex_t bindex, struct dentry *h_src, ++ struct au_cpup_reg_attr *h_src_attr) ++{ ++ int err, sbits, icex; ++ unsigned int mnt_flags; ++ unsigned char verbose; ++ struct iattr ia; ++ struct path h_path; ++ struct inode *h_isrc, *h_idst; ++ struct kstat *h_st; ++ struct au_branch *br; ++ ++ h_path.dentry = au_h_dptr(dst, bindex); ++ h_idst = d_inode(h_path.dentry); ++ br = au_sbr(dst->d_sb, bindex); ++ h_path.mnt = au_br_mnt(br); ++ h_isrc = d_inode(h_src); ++ ia.ia_valid = ATTR_FORCE | ATTR_UID | ATTR_GID ++ | ATTR_ATIME | ATTR_MTIME ++ | ATTR_ATIME_SET | ATTR_MTIME_SET; ++ if (h_src_attr && h_src_attr->valid) { ++ h_st = &h_src_attr->st; ++ ia.ia_uid = h_st->uid; ++ ia.ia_gid = h_st->gid; ++ ia.ia_atime = h_st->atime; ++ ia.ia_mtime = h_st->mtime; ++ if (h_idst->i_mode != h_st->mode ++ && !S_ISLNK(h_idst->i_mode)) { ++ ia.ia_valid |= ATTR_MODE; ++ ia.ia_mode = h_st->mode; ++ } ++ sbits = !!(h_st->mode & (S_ISUID | S_ISGID)); ++ au_cpup_attr_flags(h_idst, h_src_attr->iflags); ++ } else { ++ ia.ia_uid = h_isrc->i_uid; ++ ia.ia_gid = h_isrc->i_gid; ++ ia.ia_atime = h_isrc->i_atime; ++ ia.ia_mtime = h_isrc->i_mtime; ++ if (h_idst->i_mode != h_isrc->i_mode ++ && !S_ISLNK(h_idst->i_mode)) { ++ ia.ia_valid |= ATTR_MODE; ++ ia.ia_mode = h_isrc->i_mode; ++ } ++ sbits = !!(h_isrc->i_mode & (S_ISUID | S_ISGID)); ++ au_cpup_attr_flags(h_idst, h_isrc->i_flags); ++ } ++ /* no delegation since it is just created */ ++ err = vfsub_notify_change(&h_path, &ia, /*delegated*/NULL); ++ ++ /* is this nfs only? */ ++ if (!err && sbits && au_test_nfs(h_path.dentry->d_sb)) { ++ ia.ia_valid = ATTR_FORCE | ATTR_MODE; ++ ia.ia_mode = h_isrc->i_mode; ++ err = vfsub_notify_change(&h_path, &ia, /*delegated*/NULL); ++ } ++ ++ icex = br->br_perm & AuBrAttr_ICEX; ++ if (!err) { ++ mnt_flags = au_mntflags(dst->d_sb); ++ verbose = !!au_opt_test(mnt_flags, VERBOSE); ++ err = au_cpup_xattr(h_path.dentry, h_src, icex, verbose); ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_do_copy_file(struct file *dst, struct file *src, loff_t len, ++ char *buf, unsigned long blksize) ++{ ++ int err; ++ size_t sz, rbytes, wbytes; ++ unsigned char all_zero; ++ char *p, *zp; ++ struct mutex *h_mtx; ++ /* reduce stack usage */ ++ struct iattr *ia; ++ ++ zp = page_address(ZERO_PAGE(0)); ++ if (unlikely(!zp)) ++ return -ENOMEM; /* possible? */ ++ ++ err = 0; ++ all_zero = 0; ++ while (len) { ++ AuDbg("len %lld\n", len); ++ sz = blksize; ++ if (len < blksize) ++ sz = len; ++ ++ rbytes = 0; ++ /* todo: signal_pending? */ ++ while (!rbytes || err == -EAGAIN || err == -EINTR) { ++ rbytes = vfsub_read_k(src, buf, sz, &src->f_pos); ++ err = rbytes; ++ } ++ if (unlikely(err < 0)) ++ break; ++ ++ all_zero = 0; ++ if (len >= rbytes && rbytes == blksize) ++ all_zero = !memcmp(buf, zp, rbytes); ++ if (!all_zero) { ++ wbytes = rbytes; ++ p = buf; ++ while (wbytes) { ++ size_t b; ++ ++ b = vfsub_write_k(dst, p, wbytes, &dst->f_pos); ++ err = b; ++ /* todo: signal_pending? */ ++ if (unlikely(err == -EAGAIN || err == -EINTR)) ++ continue; ++ if (unlikely(err < 0)) ++ break; ++ wbytes -= b; ++ p += b; ++ } ++ if (unlikely(err < 0)) ++ break; ++ } else { ++ loff_t res; ++ ++ AuLabel(hole); ++ res = vfsub_llseek(dst, rbytes, SEEK_CUR); ++ err = res; ++ if (unlikely(res < 0)) ++ break; ++ } ++ len -= rbytes; ++ err = 0; ++ } ++ ++ /* the last block may be a hole */ ++ if (!err && all_zero) { ++ AuLabel(last hole); ++ ++ err = 1; ++ if (au_test_nfs(dst->f_path.dentry->d_sb)) { ++ /* nfs requires this step to make last hole */ ++ /* is this only nfs? */ ++ do { ++ /* todo: signal_pending? */ ++ err = vfsub_write_k(dst, "\0", 1, &dst->f_pos); ++ } while (err == -EAGAIN || err == -EINTR); ++ if (err == 1) ++ dst->f_pos--; ++ } ++ ++ if (err == 1) { ++ ia = (void *)buf; ++ ia->ia_size = dst->f_pos; ++ ia->ia_valid = ATTR_SIZE | ATTR_FILE; ++ ia->ia_file = dst; ++ h_mtx = &file_inode(dst)->i_mutex; ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD2); ++ /* no delegation since it is just created */ ++ err = vfsub_notify_change(&dst->f_path, ia, ++ /*delegated*/NULL); ++ mutex_unlock(h_mtx); ++ } ++ } ++ ++ return err; ++} ++ ++int au_copy_file(struct file *dst, struct file *src, loff_t len) ++{ ++ int err; ++ unsigned long blksize; ++ unsigned char do_kfree; ++ char *buf; ++ ++ err = -ENOMEM; ++ blksize = dst->f_path.dentry->d_sb->s_blocksize; ++ if (!blksize || PAGE_SIZE < blksize) ++ blksize = PAGE_SIZE; ++ AuDbg("blksize %lu\n", blksize); ++ do_kfree = (blksize != PAGE_SIZE && blksize >= sizeof(struct iattr *)); ++ if (do_kfree) ++ buf = kmalloc(blksize, GFP_NOFS); ++ else ++ buf = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!buf)) ++ goto out; ++ ++ if (len > (1 << 22)) ++ AuDbg("copying a large file %lld\n", (long long)len); ++ ++ src->f_pos = 0; ++ dst->f_pos = 0; ++ err = au_do_copy_file(dst, src, len, buf, blksize); ++ if (do_kfree) ++ kfree(buf); ++ else ++ free_page((unsigned long)buf); ++ ++out: ++ return err; ++} ++ ++/* ++ * to support a sparse file which is opened with O_APPEND, ++ * we need to close the file. ++ */ ++static int au_cp_regular(struct au_cp_generic *cpg) ++{ ++ int err, i; ++ enum { SRC, DST }; ++ struct { ++ aufs_bindex_t bindex; ++ unsigned int flags; ++ struct dentry *dentry; ++ int force_wr; ++ struct file *file; ++ void *label; ++ } *f, file[] = { ++ { ++ .bindex = cpg->bsrc, ++ .flags = O_RDONLY | O_NOATIME | O_LARGEFILE, ++ .label = &&out ++ }, ++ { ++ .bindex = cpg->bdst, ++ .flags = O_WRONLY | O_NOATIME | O_LARGEFILE, ++ .force_wr = !!au_ftest_cpup(cpg->flags, RWDST), ++ .label = &&out_src ++ } ++ }; ++ struct super_block *sb; ++ ++ /* bsrc branch can be ro/rw. */ ++ sb = cpg->dentry->d_sb; ++ f = file; ++ for (i = 0; i < 2; i++, f++) { ++ f->dentry = au_h_dptr(cpg->dentry, f->bindex); ++ f->file = au_h_open(cpg->dentry, f->bindex, f->flags, ++ /*file*/NULL, f->force_wr); ++ err = PTR_ERR(f->file); ++ if (IS_ERR(f->file)) ++ goto *f->label; ++ } ++ ++ /* try stopping to update while we copyup */ ++ IMustLock(d_inode(file[SRC].dentry)); ++ err = au_copy_file(file[DST].file, file[SRC].file, cpg->len); ++ ++ fput(file[DST].file); ++ au_sbr_put(sb, file[DST].bindex); ++ ++out_src: ++ fput(file[SRC].file); ++ au_sbr_put(sb, file[SRC].bindex); ++out: ++ return err; ++} ++ ++static int au_do_cpup_regular(struct au_cp_generic *cpg, ++ struct au_cpup_reg_attr *h_src_attr) ++{ ++ int err, rerr; ++ loff_t l; ++ struct path h_path; ++ struct inode *h_src_inode, *h_dst_inode; ++ ++ err = 0; ++ h_src_inode = au_h_iptr(d_inode(cpg->dentry), cpg->bsrc); ++ l = i_size_read(h_src_inode); ++ if (cpg->len == -1 || l < cpg->len) ++ cpg->len = l; ++ if (cpg->len) { ++ /* try stopping to update while we are referencing */ ++ mutex_lock_nested(&h_src_inode->i_mutex, AuLsc_I_CHILD); ++ au_pin_hdir_unlock(cpg->pin); ++ ++ h_path.dentry = au_h_dptr(cpg->dentry, cpg->bsrc); ++ h_path.mnt = au_sbr_mnt(cpg->dentry->d_sb, cpg->bsrc); ++ h_src_attr->iflags = h_src_inode->i_flags; ++ if (!au_test_nfs(h_src_inode->i_sb)) ++ err = vfs_getattr(&h_path, &h_src_attr->st); ++ else { ++ mutex_unlock(&h_src_inode->i_mutex); ++ err = vfs_getattr(&h_path, &h_src_attr->st); ++ mutex_lock_nested(&h_src_inode->i_mutex, AuLsc_I_CHILD); ++ } ++ if (unlikely(err)) { ++ mutex_unlock(&h_src_inode->i_mutex); ++ goto out; ++ } ++ h_src_attr->valid = 1; ++ err = au_cp_regular(cpg); ++ mutex_unlock(&h_src_inode->i_mutex); ++ rerr = au_pin_hdir_relock(cpg->pin); ++ if (!err && rerr) ++ err = rerr; ++ } ++ if (!err && (h_src_inode->i_state & I_LINKABLE)) { ++ h_path.dentry = au_h_dptr(cpg->dentry, cpg->bdst); ++ h_dst_inode = d_inode(h_path.dentry); ++ spin_lock(&h_dst_inode->i_lock); ++ h_dst_inode->i_state |= I_LINKABLE; ++ spin_unlock(&h_dst_inode->i_lock); ++ } ++ ++out: ++ return err; ++} ++ ++static int au_do_cpup_symlink(struct path *h_path, struct dentry *h_src, ++ struct inode *h_dir) ++{ ++ int err, symlen; ++ mm_segment_t old_fs; ++ union { ++ char *k; ++ char __user *u; ++ } sym; ++ struct inode *h_inode = d_inode(h_src); ++ const struct inode_operations *h_iop = h_inode->i_op; ++ ++ err = -ENOSYS; ++ if (unlikely(!h_iop->readlink)) ++ goto out; ++ ++ err = -ENOMEM; ++ sym.k = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!sym.k)) ++ goto out; ++ ++ /* unnecessary to support mmap_sem since symlink is not mmap-able */ ++ old_fs = get_fs(); ++ set_fs(KERNEL_DS); ++ symlen = h_iop->readlink(h_src, sym.u, PATH_MAX); ++ err = symlen; ++ set_fs(old_fs); ++ ++ if (symlen > 0) { ++ sym.k[symlen] = 0; ++ err = vfsub_symlink(h_dir, h_path, sym.k); ++ } ++ free_page((unsigned long)sym.k); ++ ++out: ++ return err; ++} ++ ++static noinline_for_stack ++int cpup_entry(struct au_cp_generic *cpg, struct dentry *dst_parent, ++ struct au_cpup_reg_attr *h_src_attr) ++{ ++ int err; ++ umode_t mode; ++ unsigned int mnt_flags; ++ unsigned char isdir, isreg, force; ++ const unsigned char do_dt = !!au_ftest_cpup(cpg->flags, DTIME); ++ struct au_dtime dt; ++ struct path h_path; ++ struct dentry *h_src, *h_dst, *h_parent; ++ struct inode *h_inode, *h_dir, *dir, *inode; ++ struct super_block *sb; ++ ++ /* bsrc branch can be ro/rw. */ ++ h_src = au_h_dptr(cpg->dentry, cpg->bsrc); ++ h_inode = d_inode(h_src); ++ AuDebugOn(h_inode != au_h_iptr(d_inode(cpg->dentry), cpg->bsrc)); ++ ++ /* try stopping to be referenced while we are creating */ ++ h_dst = au_h_dptr(cpg->dentry, cpg->bdst); ++ if (au_ftest_cpup(cpg->flags, RENAME)) ++ AuDebugOn(strncmp(h_dst->d_name.name, AUFS_WH_PFX, ++ AUFS_WH_PFX_LEN)); ++ h_parent = h_dst->d_parent; /* dir inode is locked */ ++ h_dir = d_inode(h_parent); ++ IMustLock(h_dir); ++ AuDebugOn(h_parent != h_dst->d_parent); ++ ++ sb = cpg->dentry->d_sb; ++ h_path.mnt = au_sbr_mnt(sb, cpg->bdst); ++ if (do_dt) { ++ h_path.dentry = h_parent; ++ au_dtime_store(&dt, dst_parent, &h_path); ++ } ++ h_path.dentry = h_dst; ++ ++ isreg = 0; ++ isdir = 0; ++ mode = h_inode->i_mode; ++ switch (mode & S_IFMT) { ++ case S_IFREG: ++ isreg = 1; ++ err = vfsub_create(h_dir, &h_path, mode | S_IWUSR, ++ /*want_excl*/true); ++ if (!err) ++ err = au_do_cpup_regular(cpg, h_src_attr); ++ break; ++ case S_IFDIR: ++ isdir = 1; ++ err = vfsub_mkdir(h_dir, &h_path, mode); ++ if (!err) { ++ /* ++ * strange behaviour from the users view, ++ * particularry setattr case ++ */ ++ dir = d_inode(dst_parent); ++ if (au_ibstart(dir) == cpg->bdst) ++ au_cpup_attr_nlink(dir, /*force*/1); ++ inode = d_inode(cpg->dentry); ++ au_cpup_attr_nlink(inode, /*force*/1); ++ } ++ break; ++ case S_IFLNK: ++ err = au_do_cpup_symlink(&h_path, h_src, h_dir); ++ break; ++ case S_IFCHR: ++ case S_IFBLK: ++ AuDebugOn(!capable(CAP_MKNOD)); ++ /*FALLTHROUGH*/ ++ case S_IFIFO: ++ case S_IFSOCK: ++ err = vfsub_mknod(h_dir, &h_path, mode, h_inode->i_rdev); ++ break; ++ default: ++ AuIOErr("Unknown inode type 0%o\n", mode); ++ err = -EIO; ++ } ++ ++ mnt_flags = au_mntflags(sb); ++ if (!au_opt_test(mnt_flags, UDBA_NONE) ++ && !isdir ++ && au_opt_test(mnt_flags, XINO) ++ && (h_inode->i_nlink == 1 ++ || (h_inode->i_state & I_LINKABLE)) ++ /* todo: unnecessary? */ ++ /* && d_inode(cpg->dentry)->i_nlink == 1 */ ++ && cpg->bdst < cpg->bsrc ++ && !au_ftest_cpup(cpg->flags, KEEPLINO)) ++ au_xino_write(sb, cpg->bsrc, h_inode->i_ino, /*ino*/0); ++ /* ignore this error */ ++ ++ if (!err) { ++ force = 0; ++ if (isreg) { ++ force = !!cpg->len; ++ if (cpg->len == -1) ++ force = !!i_size_read(h_inode); ++ } ++ au_fhsm_wrote(sb, cpg->bdst, force); ++ } ++ ++ if (do_dt) ++ au_dtime_revert(&dt); ++ return err; ++} ++ ++static int au_do_ren_after_cpup(struct au_cp_generic *cpg, struct path *h_path) ++{ ++ int err; ++ struct dentry *dentry, *h_dentry, *h_parent, *parent; ++ struct inode *h_dir; ++ aufs_bindex_t bdst; ++ ++ dentry = cpg->dentry; ++ bdst = cpg->bdst; ++ h_dentry = au_h_dptr(dentry, bdst); ++ if (!au_ftest_cpup(cpg->flags, OVERWRITE)) { ++ dget(h_dentry); ++ au_set_h_dptr(dentry, bdst, NULL); ++ err = au_lkup_neg(dentry, bdst, /*wh*/0); ++ if (!err) ++ h_path->dentry = dget(au_h_dptr(dentry, bdst)); ++ au_set_h_dptr(dentry, bdst, h_dentry); ++ } else { ++ err = 0; ++ parent = dget_parent(dentry); ++ h_parent = au_h_dptr(parent, bdst); ++ dput(parent); ++ h_path->dentry = vfsub_lkup_one(&dentry->d_name, h_parent); ++ if (IS_ERR(h_path->dentry)) ++ err = PTR_ERR(h_path->dentry); ++ } ++ if (unlikely(err)) ++ goto out; ++ ++ h_parent = h_dentry->d_parent; /* dir inode is locked */ ++ h_dir = d_inode(h_parent); ++ IMustLock(h_dir); ++ AuDbg("%pd %pd\n", h_dentry, h_path->dentry); ++ /* no delegation since it is just created */ ++ err = vfsub_rename(h_dir, h_dentry, h_dir, h_path, /*delegated*/NULL); ++ dput(h_path->dentry); ++ ++out: ++ return err; ++} ++ ++/* ++ * copyup the @dentry from @bsrc to @bdst. ++ * the caller must set the both of lower dentries. ++ * @len is for truncating when it is -1 copyup the entire file. ++ * in link/rename cases, @dst_parent may be different from the real one. ++ * basic->bsrc can be larger than basic->bdst. ++ */ ++static int au_cpup_single(struct au_cp_generic *cpg, struct dentry *dst_parent) ++{ ++ int err, rerr; ++ aufs_bindex_t old_ibstart; ++ unsigned char isdir, plink; ++ struct dentry *h_src, *h_dst, *h_parent; ++ struct inode *dst_inode, *h_dir, *inode, *delegated, *src_inode; ++ struct super_block *sb; ++ struct au_branch *br; ++ /* to reuduce stack size */ ++ struct { ++ struct au_dtime dt; ++ struct path h_path; ++ struct au_cpup_reg_attr h_src_attr; ++ } *a; ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ a->h_src_attr.valid = 0; ++ ++ sb = cpg->dentry->d_sb; ++ br = au_sbr(sb, cpg->bdst); ++ a->h_path.mnt = au_br_mnt(br); ++ h_dst = au_h_dptr(cpg->dentry, cpg->bdst); ++ h_parent = h_dst->d_parent; /* dir inode is locked */ ++ h_dir = d_inode(h_parent); ++ IMustLock(h_dir); ++ ++ h_src = au_h_dptr(cpg->dentry, cpg->bsrc); ++ inode = d_inode(cpg->dentry); ++ ++ if (!dst_parent) ++ dst_parent = dget_parent(cpg->dentry); ++ else ++ dget(dst_parent); ++ ++ plink = !!au_opt_test(au_mntflags(sb), PLINK); ++ dst_inode = au_h_iptr(inode, cpg->bdst); ++ if (dst_inode) { ++ if (unlikely(!plink)) { ++ err = -EIO; ++ AuIOErr("hi%lu(i%lu) exists on b%d " ++ "but plink is disabled\n", ++ dst_inode->i_ino, inode->i_ino, cpg->bdst); ++ goto out_parent; ++ } ++ ++ if (dst_inode->i_nlink) { ++ const int do_dt = au_ftest_cpup(cpg->flags, DTIME); ++ ++ h_src = au_plink_lkup(inode, cpg->bdst); ++ err = PTR_ERR(h_src); ++ if (IS_ERR(h_src)) ++ goto out_parent; ++ if (unlikely(d_is_negative(h_src))) { ++ err = -EIO; ++ AuIOErr("i%lu exists on a upper branch " ++ "but not pseudo-linked\n", ++ inode->i_ino); ++ dput(h_src); ++ goto out_parent; ++ } ++ ++ if (do_dt) { ++ a->h_path.dentry = h_parent; ++ au_dtime_store(&a->dt, dst_parent, &a->h_path); ++ } ++ ++ a->h_path.dentry = h_dst; ++ delegated = NULL; ++ err = vfsub_link(h_src, h_dir, &a->h_path, &delegated); ++ if (!err && au_ftest_cpup(cpg->flags, RENAME)) ++ err = au_do_ren_after_cpup(cpg, &a->h_path); ++ if (do_dt) ++ au_dtime_revert(&a->dt); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ dput(h_src); ++ goto out_parent; ++ } else ++ /* todo: cpup_wh_file? */ ++ /* udba work */ ++ au_update_ibrange(inode, /*do_put_zero*/1); ++ } ++ ++ isdir = S_ISDIR(inode->i_mode); ++ old_ibstart = au_ibstart(inode); ++ err = cpup_entry(cpg, dst_parent, &a->h_src_attr); ++ if (unlikely(err)) ++ goto out_rev; ++ dst_inode = d_inode(h_dst); ++ mutex_lock_nested(&dst_inode->i_mutex, AuLsc_I_CHILD2); ++ /* todo: necessary? */ ++ /* au_pin_hdir_unlock(cpg->pin); */ ++ ++ err = cpup_iattr(cpg->dentry, cpg->bdst, h_src, &a->h_src_attr); ++ if (unlikely(err)) { ++ /* todo: necessary? */ ++ /* au_pin_hdir_relock(cpg->pin); */ /* ignore an error */ ++ mutex_unlock(&dst_inode->i_mutex); ++ goto out_rev; ++ } ++ ++ if (cpg->bdst < old_ibstart) { ++ if (S_ISREG(inode->i_mode)) { ++ err = au_dy_iaop(inode, cpg->bdst, dst_inode); ++ if (unlikely(err)) { ++ /* ignore an error */ ++ /* au_pin_hdir_relock(cpg->pin); */ ++ mutex_unlock(&dst_inode->i_mutex); ++ goto out_rev; ++ } ++ } ++ au_set_ibstart(inode, cpg->bdst); ++ } else ++ au_set_ibend(inode, cpg->bdst); ++ au_set_h_iptr(inode, cpg->bdst, au_igrab(dst_inode), ++ au_hi_flags(inode, isdir)); ++ ++ /* todo: necessary? */ ++ /* err = au_pin_hdir_relock(cpg->pin); */ ++ mutex_unlock(&dst_inode->i_mutex); ++ if (unlikely(err)) ++ goto out_rev; ++ ++ src_inode = d_inode(h_src); ++ if (!isdir ++ && (src_inode->i_nlink > 1 ++ || src_inode->i_state & I_LINKABLE) ++ && plink) ++ au_plink_append(inode, cpg->bdst, h_dst); ++ ++ if (au_ftest_cpup(cpg->flags, RENAME)) { ++ a->h_path.dentry = h_dst; ++ err = au_do_ren_after_cpup(cpg, &a->h_path); ++ } ++ if (!err) ++ goto out_parent; /* success */ ++ ++ /* revert */ ++out_rev: ++ a->h_path.dentry = h_parent; ++ au_dtime_store(&a->dt, dst_parent, &a->h_path); ++ a->h_path.dentry = h_dst; ++ rerr = 0; ++ if (d_is_positive(h_dst)) { ++ if (!isdir) { ++ /* no delegation since it is just created */ ++ rerr = vfsub_unlink(h_dir, &a->h_path, ++ /*delegated*/NULL, /*force*/0); ++ } else ++ rerr = vfsub_rmdir(h_dir, &a->h_path); ++ } ++ au_dtime_revert(&a->dt); ++ if (rerr) { ++ AuIOErr("failed removing broken entry(%d, %d)\n", err, rerr); ++ err = -EIO; ++ } ++out_parent: ++ dput(dst_parent); ++ kfree(a); ++out: ++ return err; ++} ++ ++#if 0 /* reserved */ ++struct au_cpup_single_args { ++ int *errp; ++ struct au_cp_generic *cpg; ++ struct dentry *dst_parent; ++}; ++ ++static void au_call_cpup_single(void *args) ++{ ++ struct au_cpup_single_args *a = args; ++ ++ au_pin_hdir_acquire_nest(a->cpg->pin); ++ *a->errp = au_cpup_single(a->cpg, a->dst_parent); ++ au_pin_hdir_release(a->cpg->pin); ++} ++#endif ++ ++/* ++ * prevent SIGXFSZ in copy-up. ++ * testing CAP_MKNOD is for generic fs, ++ * but CAP_FSETID is for xfs only, currently. ++ */ ++static int au_cpup_sio_test(struct au_pin *pin, umode_t mode) ++{ ++ int do_sio; ++ struct super_block *sb; ++ struct inode *h_dir; ++ ++ do_sio = 0; ++ sb = au_pinned_parent(pin)->d_sb; ++ if (!au_wkq_test() ++ && (!au_sbi(sb)->si_plink_maint_pid ++ || au_plink_maint(sb, AuLock_NOPLM))) { ++ switch (mode & S_IFMT) { ++ case S_IFREG: ++ /* no condition about RLIMIT_FSIZE and the file size */ ++ do_sio = 1; ++ break; ++ case S_IFCHR: ++ case S_IFBLK: ++ do_sio = !capable(CAP_MKNOD); ++ break; ++ } ++ if (!do_sio) ++ do_sio = ((mode & (S_ISUID | S_ISGID)) ++ && !capable(CAP_FSETID)); ++ /* this workaround may be removed in the future */ ++ if (!do_sio) { ++ h_dir = au_pinned_h_dir(pin); ++ do_sio = h_dir->i_mode & S_ISVTX; ++ } ++ } ++ ++ return do_sio; ++} ++ ++#if 0 /* reserved */ ++int au_sio_cpup_single(struct au_cp_generic *cpg, struct dentry *dst_parent) ++{ ++ int err, wkq_err; ++ struct dentry *h_dentry; ++ ++ h_dentry = au_h_dptr(cpg->dentry, cpg->bsrc); ++ if (!au_cpup_sio_test(pin, d_inode(h_dentry)->i_mode)) ++ err = au_cpup_single(cpg, dst_parent); ++ else { ++ struct au_cpup_single_args args = { ++ .errp = &err, ++ .cpg = cpg, ++ .dst_parent = dst_parent ++ }; ++ wkq_err = au_wkq_wait(au_call_cpup_single, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ return err; ++} ++#endif ++ ++/* ++ * copyup the @dentry from the first active lower branch to @bdst, ++ * using au_cpup_single(). ++ */ ++static int au_cpup_simple(struct au_cp_generic *cpg) ++{ ++ int err; ++ unsigned int flags_orig; ++ struct dentry *dentry; ++ ++ AuDebugOn(cpg->bsrc < 0); ++ ++ dentry = cpg->dentry; ++ DiMustWriteLock(dentry); ++ ++ err = au_lkup_neg(dentry, cpg->bdst, /*wh*/1); ++ if (!err) { ++ flags_orig = cpg->flags; ++ au_fset_cpup(cpg->flags, RENAME); ++ err = au_cpup_single(cpg, NULL); ++ cpg->flags = flags_orig; ++ if (!err) ++ return 0; /* success */ ++ ++ /* revert */ ++ au_set_h_dptr(dentry, cpg->bdst, NULL); ++ au_set_dbstart(dentry, cpg->bsrc); ++ } ++ ++ return err; ++} ++ ++struct au_cpup_simple_args { ++ int *errp; ++ struct au_cp_generic *cpg; ++}; ++ ++static void au_call_cpup_simple(void *args) ++{ ++ struct au_cpup_simple_args *a = args; ++ ++ au_pin_hdir_acquire_nest(a->cpg->pin); ++ *a->errp = au_cpup_simple(a->cpg); ++ au_pin_hdir_release(a->cpg->pin); ++} ++ ++static int au_do_sio_cpup_simple(struct au_cp_generic *cpg) ++{ ++ int err, wkq_err; ++ struct dentry *dentry, *parent; ++ struct file *h_file; ++ struct inode *h_dir; ++ ++ dentry = cpg->dentry; ++ h_file = NULL; ++ if (au_ftest_cpup(cpg->flags, HOPEN)) { ++ AuDebugOn(cpg->bsrc < 0); ++ h_file = au_h_open_pre(dentry, cpg->bsrc, /*force_wr*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ } ++ ++ parent = dget_parent(dentry); ++ h_dir = au_h_iptr(d_inode(parent), cpg->bdst); ++ if (!au_test_h_perm_sio(h_dir, MAY_EXEC | MAY_WRITE) ++ && !au_cpup_sio_test(cpg->pin, d_inode(dentry)->i_mode)) ++ err = au_cpup_simple(cpg); ++ else { ++ struct au_cpup_simple_args args = { ++ .errp = &err, ++ .cpg = cpg ++ }; ++ wkq_err = au_wkq_wait(au_call_cpup_simple, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ dput(parent); ++ if (h_file) ++ au_h_open_post(dentry, cpg->bsrc, h_file); ++ ++out: ++ return err; ++} ++ ++int au_sio_cpup_simple(struct au_cp_generic *cpg) ++{ ++ aufs_bindex_t bsrc, bend; ++ struct dentry *dentry, *h_dentry; ++ ++ if (cpg->bsrc < 0) { ++ dentry = cpg->dentry; ++ bend = au_dbend(dentry); ++ for (bsrc = cpg->bdst + 1; bsrc <= bend; bsrc++) { ++ h_dentry = au_h_dptr(dentry, bsrc); ++ if (h_dentry) { ++ AuDebugOn(d_is_negative(h_dentry)); ++ break; ++ } ++ } ++ AuDebugOn(bsrc > bend); ++ cpg->bsrc = bsrc; ++ } ++ AuDebugOn(cpg->bsrc <= cpg->bdst); ++ return au_do_sio_cpup_simple(cpg); ++} ++ ++int au_sio_cpdown_simple(struct au_cp_generic *cpg) ++{ ++ AuDebugOn(cpg->bdst <= cpg->bsrc); ++ return au_do_sio_cpup_simple(cpg); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * copyup the deleted file for writing. ++ */ ++static int au_do_cpup_wh(struct au_cp_generic *cpg, struct dentry *wh_dentry, ++ struct file *file) ++{ ++ int err; ++ unsigned int flags_orig; ++ aufs_bindex_t bsrc_orig; ++ struct dentry *h_d_dst, *h_d_start; ++ struct au_dinfo *dinfo; ++ struct au_hdentry *hdp; ++ ++ dinfo = au_di(cpg->dentry); ++ AuRwMustWriteLock(&dinfo->di_rwsem); ++ ++ bsrc_orig = cpg->bsrc; ++ cpg->bsrc = dinfo->di_bstart; ++ hdp = dinfo->di_hdentry; ++ h_d_dst = hdp[0 + cpg->bdst].hd_dentry; ++ dinfo->di_bstart = cpg->bdst; ++ hdp[0 + cpg->bdst].hd_dentry = wh_dentry; ++ h_d_start = NULL; ++ if (file) { ++ h_d_start = hdp[0 + cpg->bsrc].hd_dentry; ++ hdp[0 + cpg->bsrc].hd_dentry = au_hf_top(file)->f_path.dentry; ++ } ++ flags_orig = cpg->flags; ++ cpg->flags = !AuCpup_DTIME; ++ err = au_cpup_single(cpg, /*h_parent*/NULL); ++ cpg->flags = flags_orig; ++ if (file) { ++ if (!err) ++ err = au_reopen_nondir(file); ++ hdp[0 + cpg->bsrc].hd_dentry = h_d_start; ++ } ++ hdp[0 + cpg->bdst].hd_dentry = h_d_dst; ++ dinfo->di_bstart = cpg->bsrc; ++ cpg->bsrc = bsrc_orig; ++ ++ return err; ++} ++ ++static int au_cpup_wh(struct au_cp_generic *cpg, struct file *file) ++{ ++ int err; ++ aufs_bindex_t bdst; ++ struct au_dtime dt; ++ struct dentry *dentry, *parent, *h_parent, *wh_dentry; ++ struct au_branch *br; ++ struct path h_path; ++ ++ dentry = cpg->dentry; ++ bdst = cpg->bdst; ++ br = au_sbr(dentry->d_sb, bdst); ++ parent = dget_parent(dentry); ++ h_parent = au_h_dptr(parent, bdst); ++ wh_dentry = au_whtmp_lkup(h_parent, br, &dentry->d_name); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out; ++ ++ h_path.dentry = h_parent; ++ h_path.mnt = au_br_mnt(br); ++ au_dtime_store(&dt, parent, &h_path); ++ err = au_do_cpup_wh(cpg, wh_dentry, file); ++ if (unlikely(err)) ++ goto out_wh; ++ ++ dget(wh_dentry); ++ h_path.dentry = wh_dentry; ++ if (!d_is_dir(wh_dentry)) { ++ /* no delegation since it is just created */ ++ err = vfsub_unlink(d_inode(h_parent), &h_path, ++ /*delegated*/NULL, /*force*/0); ++ } else ++ err = vfsub_rmdir(d_inode(h_parent), &h_path); ++ if (unlikely(err)) { ++ AuIOErr("failed remove copied-up tmp file %pd(%d)\n", ++ wh_dentry, err); ++ err = -EIO; ++ } ++ au_dtime_revert(&dt); ++ au_set_hi_wh(d_inode(dentry), bdst, wh_dentry); ++ ++out_wh: ++ dput(wh_dentry); ++out: ++ dput(parent); ++ return err; ++} ++ ++struct au_cpup_wh_args { ++ int *errp; ++ struct au_cp_generic *cpg; ++ struct file *file; ++}; ++ ++static void au_call_cpup_wh(void *args) ++{ ++ struct au_cpup_wh_args *a = args; ++ ++ au_pin_hdir_acquire_nest(a->cpg->pin); ++ *a->errp = au_cpup_wh(a->cpg, a->file); ++ au_pin_hdir_release(a->cpg->pin); ++} ++ ++int au_sio_cpup_wh(struct au_cp_generic *cpg, struct file *file) ++{ ++ int err, wkq_err; ++ aufs_bindex_t bdst; ++ struct dentry *dentry, *parent, *h_orph, *h_parent; ++ struct inode *dir, *h_dir, *h_tmpdir; ++ struct au_wbr *wbr; ++ struct au_pin wh_pin, *pin_orig; ++ ++ dentry = cpg->dentry; ++ bdst = cpg->bdst; ++ parent = dget_parent(dentry); ++ dir = d_inode(parent); ++ h_orph = NULL; ++ h_parent = NULL; ++ h_dir = au_igrab(au_h_iptr(dir, bdst)); ++ h_tmpdir = h_dir; ++ pin_orig = NULL; ++ if (!h_dir->i_nlink) { ++ wbr = au_sbr(dentry->d_sb, bdst)->br_wbr; ++ h_orph = wbr->wbr_orph; ++ ++ h_parent = dget(au_h_dptr(parent, bdst)); ++ au_set_h_dptr(parent, bdst, dget(h_orph)); ++ h_tmpdir = d_inode(h_orph); ++ au_set_h_iptr(dir, bdst, au_igrab(h_tmpdir), /*flags*/0); ++ ++ mutex_lock_nested(&h_tmpdir->i_mutex, AuLsc_I_PARENT3); ++ /* todo: au_h_open_pre()? */ ++ ++ pin_orig = cpg->pin; ++ au_pin_init(&wh_pin, dentry, bdst, AuLsc_DI_PARENT, ++ AuLsc_I_PARENT3, cpg->pin->udba, AuPin_DI_LOCKED); ++ cpg->pin = &wh_pin; ++ } ++ ++ if (!au_test_h_perm_sio(h_tmpdir, MAY_EXEC | MAY_WRITE) ++ && !au_cpup_sio_test(cpg->pin, d_inode(dentry)->i_mode)) ++ err = au_cpup_wh(cpg, file); ++ else { ++ struct au_cpup_wh_args args = { ++ .errp = &err, ++ .cpg = cpg, ++ .file = file ++ }; ++ wkq_err = au_wkq_wait(au_call_cpup_wh, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ if (h_orph) { ++ mutex_unlock(&h_tmpdir->i_mutex); ++ /* todo: au_h_open_post()? */ ++ au_set_h_iptr(dir, bdst, au_igrab(h_dir), /*flags*/0); ++ au_set_h_dptr(parent, bdst, h_parent); ++ AuDebugOn(!pin_orig); ++ cpg->pin = pin_orig; ++ } ++ iput(h_dir); ++ dput(parent); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * generic routine for both of copy-up and copy-down. ++ */ ++/* cf. revalidate function in file.c */ ++int au_cp_dirs(struct dentry *dentry, aufs_bindex_t bdst, ++ int (*cp)(struct dentry *dentry, aufs_bindex_t bdst, ++ struct au_pin *pin, ++ struct dentry *h_parent, void *arg), ++ void *arg) ++{ ++ int err; ++ struct au_pin pin; ++ struct dentry *d, *parent, *h_parent, *real_parent, *h_dentry; ++ ++ err = 0; ++ parent = dget_parent(dentry); ++ if (IS_ROOT(parent)) ++ goto out; ++ ++ au_pin_init(&pin, dentry, bdst, AuLsc_DI_PARENT2, AuLsc_I_PARENT2, ++ au_opt_udba(dentry->d_sb), AuPin_MNT_WRITE); ++ ++ /* do not use au_dpage */ ++ real_parent = parent; ++ while (1) { ++ dput(parent); ++ parent = dget_parent(dentry); ++ h_parent = au_h_dptr(parent, bdst); ++ if (h_parent) ++ goto out; /* success */ ++ ++ /* find top dir which is necessary to cpup */ ++ do { ++ d = parent; ++ dput(parent); ++ parent = dget_parent(d); ++ di_read_lock_parent3(parent, !AuLock_IR); ++ h_parent = au_h_dptr(parent, bdst); ++ di_read_unlock(parent, !AuLock_IR); ++ } while (!h_parent); ++ ++ if (d != real_parent) ++ di_write_lock_child3(d); ++ ++ /* somebody else might create while we were sleeping */ ++ h_dentry = au_h_dptr(d, bdst); ++ if (!h_dentry || d_is_negative(h_dentry)) { ++ if (h_dentry) ++ au_update_dbstart(d); ++ ++ au_pin_set_dentry(&pin, d); ++ err = au_do_pin(&pin); ++ if (!err) { ++ err = cp(d, bdst, &pin, h_parent, arg); ++ au_unpin(&pin); ++ } ++ } ++ ++ if (d != real_parent) ++ di_write_unlock(d); ++ if (unlikely(err)) ++ break; ++ } ++ ++out: ++ dput(parent); ++ return err; ++} ++ ++static int au_cpup_dir(struct dentry *dentry, aufs_bindex_t bdst, ++ struct au_pin *pin, ++ struct dentry *h_parent __maybe_unused, ++ void *arg __maybe_unused) ++{ ++ struct au_cp_generic cpg = { ++ .dentry = dentry, ++ .bdst = bdst, ++ .bsrc = -1, ++ .len = 0, ++ .pin = pin, ++ .flags = AuCpup_DTIME ++ }; ++ return au_sio_cpup_simple(&cpg); ++} ++ ++int au_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst) ++{ ++ return au_cp_dirs(dentry, bdst, au_cpup_dir, NULL); ++} ++ ++int au_test_and_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst) ++{ ++ int err; ++ struct dentry *parent; ++ struct inode *dir; ++ ++ parent = dget_parent(dentry); ++ dir = d_inode(parent); ++ err = 0; ++ if (au_h_iptr(dir, bdst)) ++ goto out; ++ ++ di_read_unlock(parent, AuLock_IR); ++ di_write_lock_parent(parent); ++ /* someone else might change our inode while we were sleeping */ ++ if (!au_h_iptr(dir, bdst)) ++ err = au_cpup_dirs(dentry, bdst); ++ di_downgrade_lock(parent, AuLock_IR); ++ ++out: ++ dput(parent); ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/cpup.h linux-4.1.10/fs/aufs/cpup.h +--- linux-4.1.10.orig/fs/aufs/cpup.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/cpup.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,94 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * copy-up/down functions ++ */ ++ ++#ifndef __AUFS_CPUP_H__ ++#define __AUFS_CPUP_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/path.h> ++ ++struct inode; ++struct file; ++struct au_pin; ++ ++void au_cpup_attr_flags(struct inode *dst, unsigned int iflags); ++void au_cpup_attr_timesizes(struct inode *inode); ++void au_cpup_attr_nlink(struct inode *inode, int force); ++void au_cpup_attr_changeable(struct inode *inode); ++void au_cpup_igen(struct inode *inode, struct inode *h_inode); ++void au_cpup_attr_all(struct inode *inode, int force); ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_cp_generic { ++ struct dentry *dentry; ++ aufs_bindex_t bdst, bsrc; ++ loff_t len; ++ struct au_pin *pin; ++ unsigned int flags; ++}; ++ ++/* cpup flags */ ++#define AuCpup_DTIME 1 /* do dtime_store/revert */ ++#define AuCpup_KEEPLINO (1 << 1) /* do not clear the lower xino, ++ for link(2) */ ++#define AuCpup_RENAME (1 << 2) /* rename after cpup */ ++#define AuCpup_HOPEN (1 << 3) /* call h_open_pre/post() in ++ cpup */ ++#define AuCpup_OVERWRITE (1 << 4) /* allow overwriting the ++ existing entry */ ++#define AuCpup_RWDST (1 << 5) /* force write target even if ++ the branch is marked as RO */ ++ ++#define au_ftest_cpup(flags, name) ((flags) & AuCpup_##name) ++#define au_fset_cpup(flags, name) \ ++ do { (flags) |= AuCpup_##name; } while (0) ++#define au_fclr_cpup(flags, name) \ ++ do { (flags) &= ~AuCpup_##name; } while (0) ++ ++int au_copy_file(struct file *dst, struct file *src, loff_t len); ++int au_sio_cpup_simple(struct au_cp_generic *cpg); ++int au_sio_cpdown_simple(struct au_cp_generic *cpg); ++int au_sio_cpup_wh(struct au_cp_generic *cpg, struct file *file); ++ ++int au_cp_dirs(struct dentry *dentry, aufs_bindex_t bdst, ++ int (*cp)(struct dentry *dentry, aufs_bindex_t bdst, ++ struct au_pin *pin, ++ struct dentry *h_parent, void *arg), ++ void *arg); ++int au_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst); ++int au_test_and_cpup_dirs(struct dentry *dentry, aufs_bindex_t bdst); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* keep timestamps when copyup */ ++struct au_dtime { ++ struct dentry *dt_dentry; ++ struct path dt_h_path; ++ struct timespec dt_atime, dt_mtime; ++}; ++void au_dtime_store(struct au_dtime *dt, struct dentry *dentry, ++ struct path *h_path); ++void au_dtime_revert(struct au_dtime *dt); ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_CPUP_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/dbgaufs.c linux-4.1.10/fs/aufs/dbgaufs.c +--- linux-4.1.10.orig/fs/aufs/dbgaufs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dbgaufs.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,432 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * debugfs interface ++ */ ++ ++#include <linux/debugfs.h> ++#include "aufs.h" ++ ++#ifndef CONFIG_SYSFS ++#error DEBUG_FS depends upon SYSFS ++#endif ++ ++static struct dentry *dbgaufs; ++static const mode_t dbgaufs_mode = S_IRUSR | S_IRGRP | S_IROTH; ++ ++/* 20 is max digits length of ulong 64 */ ++struct dbgaufs_arg { ++ int n; ++ char a[20 * 4]; ++}; ++ ++/* ++ * common function for all XINO files ++ */ ++static int dbgaufs_xi_release(struct inode *inode __maybe_unused, ++ struct file *file) ++{ ++ kfree(file->private_data); ++ return 0; ++} ++ ++static int dbgaufs_xi_open(struct file *xf, struct file *file, int do_fcnt) ++{ ++ int err; ++ struct kstat st; ++ struct dbgaufs_arg *p; ++ ++ err = -ENOMEM; ++ p = kmalloc(sizeof(*p), GFP_NOFS); ++ if (unlikely(!p)) ++ goto out; ++ ++ err = 0; ++ p->n = 0; ++ file->private_data = p; ++ if (!xf) ++ goto out; ++ ++ err = vfs_getattr(&xf->f_path, &st); ++ if (!err) { ++ if (do_fcnt) ++ p->n = snprintf ++ (p->a, sizeof(p->a), "%ld, %llux%lu %lld\n", ++ (long)file_count(xf), st.blocks, st.blksize, ++ (long long)st.size); ++ else ++ p->n = snprintf(p->a, sizeof(p->a), "%llux%lu %lld\n", ++ st.blocks, st.blksize, ++ (long long)st.size); ++ AuDebugOn(p->n >= sizeof(p->a)); ++ } else { ++ p->n = snprintf(p->a, sizeof(p->a), "err %d\n", err); ++ err = 0; ++ } ++ ++out: ++ return err; ++ ++} ++ ++static ssize_t dbgaufs_xi_read(struct file *file, char __user *buf, ++ size_t count, loff_t *ppos) ++{ ++ struct dbgaufs_arg *p; ++ ++ p = file->private_data; ++ return simple_read_from_buffer(buf, count, ppos, p->a, p->n); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct dbgaufs_plink_arg { ++ int n; ++ char a[]; ++}; ++ ++static int dbgaufs_plink_release(struct inode *inode __maybe_unused, ++ struct file *file) ++{ ++ free_page((unsigned long)file->private_data); ++ return 0; ++} ++ ++static int dbgaufs_plink_open(struct inode *inode, struct file *file) ++{ ++ int err, i, limit; ++ unsigned long n, sum; ++ struct dbgaufs_plink_arg *p; ++ struct au_sbinfo *sbinfo; ++ struct super_block *sb; ++ struct au_sphlhead *sphl; ++ ++ err = -ENOMEM; ++ p = (void *)get_zeroed_page(GFP_NOFS); ++ if (unlikely(!p)) ++ goto out; ++ ++ err = -EFBIG; ++ sbinfo = inode->i_private; ++ sb = sbinfo->si_sb; ++ si_noflush_read_lock(sb); ++ if (au_opt_test(au_mntflags(sb), PLINK)) { ++ limit = PAGE_SIZE - sizeof(p->n); ++ ++ /* the number of buckets */ ++ n = snprintf(p->a + p->n, limit, "%d\n", AuPlink_NHASH); ++ p->n += n; ++ limit -= n; ++ ++ sum = 0; ++ for (i = 0, sphl = sbinfo->si_plink; ++ i < AuPlink_NHASH; ++ i++, sphl++) { ++ n = au_sphl_count(sphl); ++ sum += n; ++ ++ n = snprintf(p->a + p->n, limit, "%lu ", n); ++ p->n += n; ++ limit -= n; ++ if (unlikely(limit <= 0)) ++ goto out_free; ++ } ++ p->a[p->n - 1] = '\n'; ++ ++ /* the sum of plinks */ ++ n = snprintf(p->a + p->n, limit, "%lu\n", sum); ++ p->n += n; ++ limit -= n; ++ if (unlikely(limit <= 0)) ++ goto out_free; ++ } else { ++#define str "1\n0\n0\n" ++ p->n = sizeof(str) - 1; ++ strcpy(p->a, str); ++#undef str ++ } ++ si_read_unlock(sb); ++ ++ err = 0; ++ file->private_data = p; ++ goto out; /* success */ ++ ++out_free: ++ free_page((unsigned long)p); ++out: ++ return err; ++} ++ ++static ssize_t dbgaufs_plink_read(struct file *file, char __user *buf, ++ size_t count, loff_t *ppos) ++{ ++ struct dbgaufs_plink_arg *p; ++ ++ p = file->private_data; ++ return simple_read_from_buffer(buf, count, ppos, p->a, p->n); ++} ++ ++static const struct file_operations dbgaufs_plink_fop = { ++ .owner = THIS_MODULE, ++ .open = dbgaufs_plink_open, ++ .release = dbgaufs_plink_release, ++ .read = dbgaufs_plink_read ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int dbgaufs_xib_open(struct inode *inode, struct file *file) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ struct super_block *sb; ++ ++ sbinfo = inode->i_private; ++ sb = sbinfo->si_sb; ++ si_noflush_read_lock(sb); ++ err = dbgaufs_xi_open(sbinfo->si_xib, file, /*do_fcnt*/0); ++ si_read_unlock(sb); ++ return err; ++} ++ ++static const struct file_operations dbgaufs_xib_fop = { ++ .owner = THIS_MODULE, ++ .open = dbgaufs_xib_open, ++ .release = dbgaufs_xi_release, ++ .read = dbgaufs_xi_read ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define DbgaufsXi_PREFIX "xi" ++ ++static int dbgaufs_xino_open(struct inode *inode, struct file *file) ++{ ++ int err; ++ long l; ++ struct au_sbinfo *sbinfo; ++ struct super_block *sb; ++ struct file *xf; ++ struct qstr *name; ++ ++ err = -ENOENT; ++ xf = NULL; ++ name = &file->f_path.dentry->d_name; ++ if (unlikely(name->len < sizeof(DbgaufsXi_PREFIX) ++ || memcmp(name->name, DbgaufsXi_PREFIX, ++ sizeof(DbgaufsXi_PREFIX) - 1))) ++ goto out; ++ err = kstrtol(name->name + sizeof(DbgaufsXi_PREFIX) - 1, 10, &l); ++ if (unlikely(err)) ++ goto out; ++ ++ sbinfo = inode->i_private; ++ sb = sbinfo->si_sb; ++ si_noflush_read_lock(sb); ++ if (l <= au_sbend(sb)) { ++ xf = au_sbr(sb, (aufs_bindex_t)l)->br_xino.xi_file; ++ err = dbgaufs_xi_open(xf, file, /*do_fcnt*/1); ++ } else ++ err = -ENOENT; ++ si_read_unlock(sb); ++ ++out: ++ return err; ++} ++ ++static const struct file_operations dbgaufs_xino_fop = { ++ .owner = THIS_MODULE, ++ .open = dbgaufs_xino_open, ++ .release = dbgaufs_xi_release, ++ .read = dbgaufs_xi_read ++}; ++ ++void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ aufs_bindex_t bend; ++ struct au_branch *br; ++ struct au_xino_file *xi; ++ ++ if (!au_sbi(sb)->si_dbgaufs) ++ return; ++ ++ bend = au_sbend(sb); ++ for (; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ xi = &br->br_xino; ++ debugfs_remove(xi->xi_dbgaufs); ++ xi->xi_dbgaufs = NULL; ++ } ++} ++ ++void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ struct au_sbinfo *sbinfo; ++ struct dentry *parent; ++ struct au_branch *br; ++ struct au_xino_file *xi; ++ aufs_bindex_t bend; ++ char name[sizeof(DbgaufsXi_PREFIX) + 5]; /* "xi" bindex NULL */ ++ ++ sbinfo = au_sbi(sb); ++ parent = sbinfo->si_dbgaufs; ++ if (!parent) ++ return; ++ ++ bend = au_sbend(sb); ++ for (; bindex <= bend; bindex++) { ++ snprintf(name, sizeof(name), DbgaufsXi_PREFIX "%d", bindex); ++ br = au_sbr(sb, bindex); ++ xi = &br->br_xino; ++ AuDebugOn(xi->xi_dbgaufs); ++ xi->xi_dbgaufs = debugfs_create_file(name, dbgaufs_mode, parent, ++ sbinfo, &dbgaufs_xino_fop); ++ /* ignore an error */ ++ if (unlikely(!xi->xi_dbgaufs)) ++ AuWarn1("failed %s under debugfs\n", name); ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_EXPORT ++static int dbgaufs_xigen_open(struct inode *inode, struct file *file) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ struct super_block *sb; ++ ++ sbinfo = inode->i_private; ++ sb = sbinfo->si_sb; ++ si_noflush_read_lock(sb); ++ err = dbgaufs_xi_open(sbinfo->si_xigen, file, /*do_fcnt*/0); ++ si_read_unlock(sb); ++ return err; ++} ++ ++static const struct file_operations dbgaufs_xigen_fop = { ++ .owner = THIS_MODULE, ++ .open = dbgaufs_xigen_open, ++ .release = dbgaufs_xi_release, ++ .read = dbgaufs_xi_read ++}; ++ ++static int dbgaufs_xigen_init(struct au_sbinfo *sbinfo) ++{ ++ int err; ++ ++ /* ++ * This function is a dynamic '__init' function actually, ++ * so the tiny check for si_rwsem is unnecessary. ++ */ ++ /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ ++ ++ err = -EIO; ++ sbinfo->si_dbgaufs_xigen = debugfs_create_file ++ ("xigen", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, ++ &dbgaufs_xigen_fop); ++ if (sbinfo->si_dbgaufs_xigen) ++ err = 0; ++ ++ return err; ++} ++#else ++static int dbgaufs_xigen_init(struct au_sbinfo *sbinfo) ++{ ++ return 0; ++} ++#endif /* CONFIG_AUFS_EXPORT */ ++ ++/* ---------------------------------------------------------------------- */ ++ ++void dbgaufs_si_fin(struct au_sbinfo *sbinfo) ++{ ++ /* ++ * This function is a dynamic '__fin' function actually, ++ * so the tiny check for si_rwsem is unnecessary. ++ */ ++ /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ ++ ++ debugfs_remove_recursive(sbinfo->si_dbgaufs); ++ sbinfo->si_dbgaufs = NULL; ++ kobject_put(&sbinfo->si_kobj); ++} ++ ++int dbgaufs_si_init(struct au_sbinfo *sbinfo) ++{ ++ int err; ++ char name[SysaufsSiNameLen]; ++ ++ /* ++ * This function is a dynamic '__init' function actually, ++ * so the tiny check for si_rwsem is unnecessary. ++ */ ++ /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ ++ ++ err = -ENOENT; ++ if (!dbgaufs) { ++ AuErr1("/debug/aufs is uninitialized\n"); ++ goto out; ++ } ++ ++ err = -EIO; ++ sysaufs_name(sbinfo, name); ++ sbinfo->si_dbgaufs = debugfs_create_dir(name, dbgaufs); ++ if (unlikely(!sbinfo->si_dbgaufs)) ++ goto out; ++ kobject_get(&sbinfo->si_kobj); ++ ++ sbinfo->si_dbgaufs_xib = debugfs_create_file ++ ("xib", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, ++ &dbgaufs_xib_fop); ++ if (unlikely(!sbinfo->si_dbgaufs_xib)) ++ goto out_dir; ++ ++ sbinfo->si_dbgaufs_plink = debugfs_create_file ++ ("plink", dbgaufs_mode, sbinfo->si_dbgaufs, sbinfo, ++ &dbgaufs_plink_fop); ++ if (unlikely(!sbinfo->si_dbgaufs_plink)) ++ goto out_dir; ++ ++ err = dbgaufs_xigen_init(sbinfo); ++ if (!err) ++ goto out; /* success */ ++ ++out_dir: ++ dbgaufs_si_fin(sbinfo); ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void dbgaufs_fin(void) ++{ ++ debugfs_remove(dbgaufs); ++} ++ ++int __init dbgaufs_init(void) ++{ ++ int err; ++ ++ err = -EIO; ++ dbgaufs = debugfs_create_dir(AUFS_NAME, NULL); ++ if (dbgaufs) ++ err = 0; ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/dbgaufs.h linux-4.1.10/fs/aufs/dbgaufs.h +--- linux-4.1.10.orig/fs/aufs/dbgaufs.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dbgaufs.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,48 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * debugfs interface ++ */ ++ ++#ifndef __DBGAUFS_H__ ++#define __DBGAUFS_H__ ++ ++#ifdef __KERNEL__ ++ ++struct super_block; ++struct au_sbinfo; ++ ++#ifdef CONFIG_DEBUG_FS ++/* dbgaufs.c */ ++void dbgaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex); ++void dbgaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex); ++void dbgaufs_si_fin(struct au_sbinfo *sbinfo); ++int dbgaufs_si_init(struct au_sbinfo *sbinfo); ++void dbgaufs_fin(void); ++int __init dbgaufs_init(void); ++#else ++AuStubVoid(dbgaufs_brs_del, struct super_block *sb, aufs_bindex_t bindex) ++AuStubVoid(dbgaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex) ++AuStubVoid(dbgaufs_si_fin, struct au_sbinfo *sbinfo) ++AuStubInt0(dbgaufs_si_init, struct au_sbinfo *sbinfo) ++AuStubVoid(dbgaufs_fin, void) ++AuStubInt0(__init dbgaufs_init, void) ++#endif /* CONFIG_DEBUG_FS */ ++ ++#endif /* __KERNEL__ */ ++#endif /* __DBGAUFS_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/dcsub.c linux-4.1.10/fs/aufs/dcsub.c +--- linux-4.1.10.orig/fs/aufs/dcsub.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dcsub.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,224 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sub-routines for dentry cache ++ */ ++ ++#include "aufs.h" ++ ++static void au_dpage_free(struct au_dpage *dpage) ++{ ++ int i; ++ struct dentry **p; ++ ++ p = dpage->dentries; ++ for (i = 0; i < dpage->ndentry; i++) ++ dput(*p++); ++ free_page((unsigned long)dpage->dentries); ++} ++ ++int au_dpages_init(struct au_dcsub_pages *dpages, gfp_t gfp) ++{ ++ int err; ++ void *p; ++ ++ err = -ENOMEM; ++ dpages->dpages = kmalloc(sizeof(*dpages->dpages), gfp); ++ if (unlikely(!dpages->dpages)) ++ goto out; ++ ++ p = (void *)__get_free_page(gfp); ++ if (unlikely(!p)) ++ goto out_dpages; ++ ++ dpages->dpages[0].ndentry = 0; ++ dpages->dpages[0].dentries = p; ++ dpages->ndpage = 1; ++ return 0; /* success */ ++ ++out_dpages: ++ kfree(dpages->dpages); ++out: ++ return err; ++} ++ ++void au_dpages_free(struct au_dcsub_pages *dpages) ++{ ++ int i; ++ struct au_dpage *p; ++ ++ p = dpages->dpages; ++ for (i = 0; i < dpages->ndpage; i++) ++ au_dpage_free(p++); ++ kfree(dpages->dpages); ++} ++ ++static int au_dpages_append(struct au_dcsub_pages *dpages, ++ struct dentry *dentry, gfp_t gfp) ++{ ++ int err, sz; ++ struct au_dpage *dpage; ++ void *p; ++ ++ dpage = dpages->dpages + dpages->ndpage - 1; ++ sz = PAGE_SIZE / sizeof(dentry); ++ if (unlikely(dpage->ndentry >= sz)) { ++ AuLabel(new dpage); ++ err = -ENOMEM; ++ sz = dpages->ndpage * sizeof(*dpages->dpages); ++ p = au_kzrealloc(dpages->dpages, sz, ++ sz + sizeof(*dpages->dpages), gfp); ++ if (unlikely(!p)) ++ goto out; ++ ++ dpages->dpages = p; ++ dpage = dpages->dpages + dpages->ndpage; ++ p = (void *)__get_free_page(gfp); ++ if (unlikely(!p)) ++ goto out; ++ ++ dpage->ndentry = 0; ++ dpage->dentries = p; ++ dpages->ndpage++; ++ } ++ ++ AuDebugOn(au_dcount(dentry) <= 0); ++ dpage->dentries[dpage->ndentry++] = dget_dlock(dentry); ++ return 0; /* success */ ++ ++out: ++ return err; ++} ++ ++/* todo: BAD approach */ ++/* copied from linux/fs/dcache.c */ ++enum d_walk_ret { ++ D_WALK_CONTINUE, ++ D_WALK_QUIT, ++ D_WALK_NORETRY, ++ D_WALK_SKIP, ++}; ++ ++extern void d_walk(struct dentry *parent, void *data, ++ enum d_walk_ret (*enter)(void *, struct dentry *), ++ void (*finish)(void *)); ++ ++struct ac_dpages_arg { ++ int err; ++ struct au_dcsub_pages *dpages; ++ struct super_block *sb; ++ au_dpages_test test; ++ void *arg; ++}; ++ ++static enum d_walk_ret au_call_dpages_append(void *_arg, struct dentry *dentry) ++{ ++ enum d_walk_ret ret; ++ struct ac_dpages_arg *arg = _arg; ++ ++ ret = D_WALK_CONTINUE; ++ if (dentry->d_sb == arg->sb ++ && !IS_ROOT(dentry) ++ && au_dcount(dentry) > 0 ++ && au_di(dentry) ++ && (!arg->test || arg->test(dentry, arg->arg))) { ++ arg->err = au_dpages_append(arg->dpages, dentry, GFP_ATOMIC); ++ if (unlikely(arg->err)) ++ ret = D_WALK_QUIT; ++ } ++ ++ return ret; ++} ++ ++int au_dcsub_pages(struct au_dcsub_pages *dpages, struct dentry *root, ++ au_dpages_test test, void *arg) ++{ ++ struct ac_dpages_arg args = { ++ .err = 0, ++ .dpages = dpages, ++ .sb = root->d_sb, ++ .test = test, ++ .arg = arg ++ }; ++ ++ d_walk(root, &args, au_call_dpages_append, NULL); ++ ++ return args.err; ++} ++ ++int au_dcsub_pages_rev(struct au_dcsub_pages *dpages, struct dentry *dentry, ++ int do_include, au_dpages_test test, void *arg) ++{ ++ int err; ++ ++ err = 0; ++ write_seqlock(&rename_lock); ++ spin_lock(&dentry->d_lock); ++ if (do_include ++ && au_dcount(dentry) > 0 ++ && (!test || test(dentry, arg))) ++ err = au_dpages_append(dpages, dentry, GFP_ATOMIC); ++ spin_unlock(&dentry->d_lock); ++ if (unlikely(err)) ++ goto out; ++ ++ /* ++ * RCU for vfsmount is unnecessary since this is a traverse in a single ++ * mount ++ */ ++ while (!IS_ROOT(dentry)) { ++ dentry = dentry->d_parent; /* rename_lock is locked */ ++ spin_lock(&dentry->d_lock); ++ if (au_dcount(dentry) > 0 ++ && (!test || test(dentry, arg))) ++ err = au_dpages_append(dpages, dentry, GFP_ATOMIC); ++ spin_unlock(&dentry->d_lock); ++ if (unlikely(err)) ++ break; ++ } ++ ++out: ++ write_sequnlock(&rename_lock); ++ return err; ++} ++ ++static inline int au_dcsub_dpages_aufs(struct dentry *dentry, void *arg) ++{ ++ return au_di(dentry) && dentry->d_sb == arg; ++} ++ ++int au_dcsub_pages_rev_aufs(struct au_dcsub_pages *dpages, ++ struct dentry *dentry, int do_include) ++{ ++ return au_dcsub_pages_rev(dpages, dentry, do_include, ++ au_dcsub_dpages_aufs, dentry->d_sb); ++} ++ ++int au_test_subdir(struct dentry *d1, struct dentry *d2) ++{ ++ struct path path[2] = { ++ { ++ .dentry = d1 ++ }, ++ { ++ .dentry = d2 ++ } ++ }; ++ ++ return path_is_under(path + 0, path + 1); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/dcsub.h linux-4.1.10/fs/aufs/dcsub.h +--- linux-4.1.10.orig/fs/aufs/dcsub.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dcsub.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,136 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sub-routines for dentry cache ++ */ ++ ++#ifndef __AUFS_DCSUB_H__ ++#define __AUFS_DCSUB_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/dcache.h> ++#include <linux/fs.h> ++ ++struct au_dpage { ++ int ndentry; ++ struct dentry **dentries; ++}; ++ ++struct au_dcsub_pages { ++ int ndpage; ++ struct au_dpage *dpages; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* dcsub.c */ ++int au_dpages_init(struct au_dcsub_pages *dpages, gfp_t gfp); ++void au_dpages_free(struct au_dcsub_pages *dpages); ++typedef int (*au_dpages_test)(struct dentry *dentry, void *arg); ++int au_dcsub_pages(struct au_dcsub_pages *dpages, struct dentry *root, ++ au_dpages_test test, void *arg); ++int au_dcsub_pages_rev(struct au_dcsub_pages *dpages, struct dentry *dentry, ++ int do_include, au_dpages_test test, void *arg); ++int au_dcsub_pages_rev_aufs(struct au_dcsub_pages *dpages, ++ struct dentry *dentry, int do_include); ++int au_test_subdir(struct dentry *d1, struct dentry *d2); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * todo: in linux-3.13, several similar (but faster) helpers are added to ++ * include/linux/dcache.h. Try them (in the future). ++ */ ++ ++static inline int au_d_hashed_positive(struct dentry *d) ++{ ++ int err; ++ struct inode *inode = d_inode(d); ++ ++ err = 0; ++ if (unlikely(d_unhashed(d) ++ || d_is_negative(d) ++ || !inode->i_nlink)) ++ err = -ENOENT; ++ return err; ++} ++ ++static inline int au_d_linkable(struct dentry *d) ++{ ++ int err; ++ struct inode *inode = d_inode(d); ++ ++ err = au_d_hashed_positive(d); ++ if (err ++ && d_is_positive(d) ++ && (inode->i_state & I_LINKABLE)) ++ err = 0; ++ return err; ++} ++ ++static inline int au_d_alive(struct dentry *d) ++{ ++ int err; ++ struct inode *inode; ++ ++ err = 0; ++ if (!IS_ROOT(d)) ++ err = au_d_hashed_positive(d); ++ else { ++ inode = d_inode(d); ++ if (unlikely(d_unlinked(d) ++ || d_is_negative(d) ++ || !inode->i_nlink)) ++ err = -ENOENT; ++ } ++ return err; ++} ++ ++static inline int au_alive_dir(struct dentry *d) ++{ ++ int err; ++ ++ err = au_d_alive(d); ++ if (unlikely(err || IS_DEADDIR(d_inode(d)))) ++ err = -ENOENT; ++ return err; ++} ++ ++static inline int au_qstreq(struct qstr *a, struct qstr *b) ++{ ++ return a->len == b->len ++ && !memcmp(a->name, b->name, a->len); ++} ++ ++/* ++ * by the commit ++ * 360f547 2015-01-25 dcache: let the dentry count go down to zero without ++ * taking d_lock ++ * the type of d_lockref.count became int, but the inlined function d_count() ++ * still returns unsigned int. ++ * I don't know why. Maybe it is for every d_count() users? ++ * Anyway au_dcount() lives on. ++ */ ++static inline int au_dcount(struct dentry *d) ++{ ++ return (int)d_count(d); ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DCSUB_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/debug.c linux-4.1.10/fs/aufs/debug.c +--- linux-4.1.10.orig/fs/aufs/debug.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/debug.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,440 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * debug print functions ++ */ ++ ++#include "aufs.h" ++ ++/* Returns 0, or -errno. arg is in kp->arg. */ ++static int param_atomic_t_set(const char *val, const struct kernel_param *kp) ++{ ++ int err, n; ++ ++ err = kstrtoint(val, 0, &n); ++ if (!err) { ++ if (n > 0) ++ au_debug_on(); ++ else ++ au_debug_off(); ++ } ++ return err; ++} ++ ++/* Returns length written or -errno. Buffer is 4k (ie. be short!) */ ++static int param_atomic_t_get(char *buffer, const struct kernel_param *kp) ++{ ++ atomic_t *a; ++ ++ a = kp->arg; ++ return sprintf(buffer, "%d", atomic_read(a)); ++} ++ ++static struct kernel_param_ops param_ops_atomic_t = { ++ .set = param_atomic_t_set, ++ .get = param_atomic_t_get ++ /* void (*free)(void *arg) */ ++}; ++ ++atomic_t aufs_debug = ATOMIC_INIT(0); ++MODULE_PARM_DESC(debug, "debug print"); ++module_param_named(debug, aufs_debug, atomic_t, S_IRUGO | S_IWUSR | S_IWGRP); ++ ++DEFINE_MUTEX(au_dbg_mtx); /* just to serialize the dbg msgs */ ++char *au_plevel = KERN_DEBUG; ++#define dpri(fmt, ...) do { \ ++ if ((au_plevel \ ++ && strcmp(au_plevel, KERN_DEBUG)) \ ++ || au_debug_test()) \ ++ printk("%s" fmt, au_plevel, ##__VA_ARGS__); \ ++} while (0) ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_dpri_whlist(struct au_nhash *whlist) ++{ ++ unsigned long ul, n; ++ struct hlist_head *head; ++ struct au_vdir_wh *pos; ++ ++ n = whlist->nh_num; ++ head = whlist->nh_head; ++ for (ul = 0; ul < n; ul++) { ++ hlist_for_each_entry(pos, head, wh_hash) ++ dpri("b%d, %.*s, %d\n", ++ pos->wh_bindex, ++ pos->wh_str.len, pos->wh_str.name, ++ pos->wh_str.len); ++ head++; ++ } ++} ++ ++void au_dpri_vdir(struct au_vdir *vdir) ++{ ++ unsigned long ul; ++ union au_vdir_deblk_p p; ++ unsigned char *o; ++ ++ if (!vdir || IS_ERR(vdir)) { ++ dpri("err %ld\n", PTR_ERR(vdir)); ++ return; ++ } ++ ++ dpri("deblk %u, nblk %lu, deblk %p, last{%lu, %p}, ver %lu\n", ++ vdir->vd_deblk_sz, vdir->vd_nblk, vdir->vd_deblk, ++ vdir->vd_last.ul, vdir->vd_last.p.deblk, vdir->vd_version); ++ for (ul = 0; ul < vdir->vd_nblk; ul++) { ++ p.deblk = vdir->vd_deblk[ul]; ++ o = p.deblk; ++ dpri("[%lu]: %p\n", ul, o); ++ } ++} ++ ++static int do_pri_inode(aufs_bindex_t bindex, struct inode *inode, int hn, ++ struct dentry *wh) ++{ ++ char *n = NULL; ++ int l = 0; ++ ++ if (!inode || IS_ERR(inode)) { ++ dpri("i%d: err %ld\n", bindex, PTR_ERR(inode)); ++ return -1; ++ } ++ ++ /* the type of i_blocks depends upon CONFIG_LBDAF */ ++ BUILD_BUG_ON(sizeof(inode->i_blocks) != sizeof(unsigned long) ++ && sizeof(inode->i_blocks) != sizeof(u64)); ++ if (wh) { ++ n = (void *)wh->d_name.name; ++ l = wh->d_name.len; ++ } ++ ++ dpri("i%d: %p, i%lu, %s, cnt %d, nl %u, 0%o, sz %llu, blk %llu," ++ " hn %d, ct %lld, np %lu, st 0x%lx, f 0x%x, v %llu, g %x%s%.*s\n", ++ bindex, inode, ++ inode->i_ino, inode->i_sb ? au_sbtype(inode->i_sb) : "??", ++ atomic_read(&inode->i_count), inode->i_nlink, inode->i_mode, ++ i_size_read(inode), (unsigned long long)inode->i_blocks, ++ hn, (long long)timespec_to_ns(&inode->i_ctime) & 0x0ffff, ++ inode->i_mapping ? inode->i_mapping->nrpages : 0, ++ inode->i_state, inode->i_flags, inode->i_version, ++ inode->i_generation, ++ l ? ", wh " : "", l, n); ++ return 0; ++} ++ ++void au_dpri_inode(struct inode *inode) ++{ ++ struct au_iinfo *iinfo; ++ aufs_bindex_t bindex; ++ int err, hn; ++ ++ err = do_pri_inode(-1, inode, -1, NULL); ++ if (err || !au_test_aufs(inode->i_sb)) ++ return; ++ ++ iinfo = au_ii(inode); ++ if (!iinfo) ++ return; ++ dpri("i-1: bstart %d, bend %d, gen %d\n", ++ iinfo->ii_bstart, iinfo->ii_bend, au_iigen(inode, NULL)); ++ if (iinfo->ii_bstart < 0) ++ return; ++ hn = 0; ++ for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; bindex++) { ++ hn = !!au_hn(iinfo->ii_hinode + bindex); ++ do_pri_inode(bindex, iinfo->ii_hinode[0 + bindex].hi_inode, hn, ++ iinfo->ii_hinode[0 + bindex].hi_whdentry); ++ } ++} ++ ++void au_dpri_dalias(struct inode *inode) ++{ ++ struct dentry *d; ++ ++ spin_lock(&inode->i_lock); ++ hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) ++ au_dpri_dentry(d); ++ spin_unlock(&inode->i_lock); ++} ++ ++static int do_pri_dentry(aufs_bindex_t bindex, struct dentry *dentry) ++{ ++ struct dentry *wh = NULL; ++ int hn; ++ struct au_iinfo *iinfo; ++ ++ if (!dentry || IS_ERR(dentry)) { ++ dpri("d%d: err %ld\n", bindex, PTR_ERR(dentry)); ++ return -1; ++ } ++ /* do not call dget_parent() here */ ++ /* note: access d_xxx without d_lock */ ++ dpri("d%d: %p, %pd2?, %s, cnt %d, flags 0x%x, %shashed\n", ++ bindex, dentry, dentry, ++ dentry->d_sb ? au_sbtype(dentry->d_sb) : "??", ++ au_dcount(dentry), dentry->d_flags, ++ d_unhashed(dentry) ? "un" : ""); ++ hn = -1; ++ if (bindex >= 0 ++ && d_is_positive(dentry) ++ && au_test_aufs(dentry->d_sb)) { ++ iinfo = au_ii(d_inode(dentry)); ++ if (iinfo) { ++ hn = !!au_hn(iinfo->ii_hinode + bindex); ++ wh = iinfo->ii_hinode[0 + bindex].hi_whdentry; ++ } ++ } ++ do_pri_inode(bindex, d_inode(dentry), hn, wh); ++ return 0; ++} ++ ++void au_dpri_dentry(struct dentry *dentry) ++{ ++ struct au_dinfo *dinfo; ++ aufs_bindex_t bindex; ++ int err; ++ struct au_hdentry *hdp; ++ ++ err = do_pri_dentry(-1, dentry); ++ if (err || !au_test_aufs(dentry->d_sb)) ++ return; ++ ++ dinfo = au_di(dentry); ++ if (!dinfo) ++ return; ++ dpri("d-1: bstart %d, bend %d, bwh %d, bdiropq %d, gen %d, tmp %d\n", ++ dinfo->di_bstart, dinfo->di_bend, ++ dinfo->di_bwh, dinfo->di_bdiropq, au_digen(dentry), ++ dinfo->di_tmpfile); ++ if (dinfo->di_bstart < 0) ++ return; ++ hdp = dinfo->di_hdentry; ++ for (bindex = dinfo->di_bstart; bindex <= dinfo->di_bend; bindex++) ++ do_pri_dentry(bindex, hdp[0 + bindex].hd_dentry); ++} ++ ++static int do_pri_file(aufs_bindex_t bindex, struct file *file) ++{ ++ char a[32]; ++ ++ if (!file || IS_ERR(file)) { ++ dpri("f%d: err %ld\n", bindex, PTR_ERR(file)); ++ return -1; ++ } ++ a[0] = 0; ++ if (bindex < 0 ++ && !IS_ERR_OR_NULL(file->f_path.dentry) ++ && au_test_aufs(file->f_path.dentry->d_sb) ++ && au_fi(file)) ++ snprintf(a, sizeof(a), ", gen %d, mmapped %d", ++ au_figen(file), atomic_read(&au_fi(file)->fi_mmapped)); ++ dpri("f%d: mode 0x%x, flags 0%o, cnt %ld, v %llu, pos %llu%s\n", ++ bindex, file->f_mode, file->f_flags, (long)file_count(file), ++ file->f_version, file->f_pos, a); ++ if (!IS_ERR_OR_NULL(file->f_path.dentry)) ++ do_pri_dentry(bindex, file->f_path.dentry); ++ return 0; ++} ++ ++void au_dpri_file(struct file *file) ++{ ++ struct au_finfo *finfo; ++ struct au_fidir *fidir; ++ struct au_hfile *hfile; ++ aufs_bindex_t bindex; ++ int err; ++ ++ err = do_pri_file(-1, file); ++ if (err ++ || IS_ERR_OR_NULL(file->f_path.dentry) ++ || !au_test_aufs(file->f_path.dentry->d_sb)) ++ return; ++ ++ finfo = au_fi(file); ++ if (!finfo) ++ return; ++ if (finfo->fi_btop < 0) ++ return; ++ fidir = finfo->fi_hdir; ++ if (!fidir) ++ do_pri_file(finfo->fi_btop, finfo->fi_htop.hf_file); ++ else ++ for (bindex = finfo->fi_btop; ++ bindex >= 0 && bindex <= fidir->fd_bbot; ++ bindex++) { ++ hfile = fidir->fd_hfile + bindex; ++ do_pri_file(bindex, hfile ? hfile->hf_file : NULL); ++ } ++} ++ ++static int do_pri_br(aufs_bindex_t bindex, struct au_branch *br) ++{ ++ struct vfsmount *mnt; ++ struct super_block *sb; ++ ++ if (!br || IS_ERR(br)) ++ goto out; ++ mnt = au_br_mnt(br); ++ if (!mnt || IS_ERR(mnt)) ++ goto out; ++ sb = mnt->mnt_sb; ++ if (!sb || IS_ERR(sb)) ++ goto out; ++ ++ dpri("s%d: {perm 0x%x, id %d, cnt %d, wbr %p}, " ++ "%s, dev 0x%02x%02x, flags 0x%lx, cnt %d, active %d, " ++ "xino %d\n", ++ bindex, br->br_perm, br->br_id, atomic_read(&br->br_count), ++ br->br_wbr, au_sbtype(sb), MAJOR(sb->s_dev), MINOR(sb->s_dev), ++ sb->s_flags, sb->s_count, ++ atomic_read(&sb->s_active), !!br->br_xino.xi_file); ++ return 0; ++ ++out: ++ dpri("s%d: err %ld\n", bindex, PTR_ERR(br)); ++ return -1; ++} ++ ++void au_dpri_sb(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ aufs_bindex_t bindex; ++ int err; ++ /* to reuduce stack size */ ++ struct { ++ struct vfsmount mnt; ++ struct au_branch fake; ++ } *a; ++ ++ /* this function can be called from magic sysrq */ ++ a = kzalloc(sizeof(*a), GFP_ATOMIC); ++ if (unlikely(!a)) { ++ dpri("no memory\n"); ++ return; ++ } ++ ++ a->mnt.mnt_sb = sb; ++ a->fake.br_perm = 0; ++ a->fake.br_path.mnt = &a->mnt; ++ a->fake.br_xino.xi_file = NULL; ++ atomic_set(&a->fake.br_count, 0); ++ smp_mb(); /* atomic_set */ ++ err = do_pri_br(-1, &a->fake); ++ kfree(a); ++ dpri("dev 0x%x\n", sb->s_dev); ++ if (err || !au_test_aufs(sb)) ++ return; ++ ++ sbinfo = au_sbi(sb); ++ if (!sbinfo) ++ return; ++ dpri("nw %d, gen %u, kobj %d\n", ++ atomic_read(&sbinfo->si_nowait.nw_len), sbinfo->si_generation, ++ atomic_read(&sbinfo->si_kobj.kref.refcount)); ++ for (bindex = 0; bindex <= sbinfo->si_bend; bindex++) ++ do_pri_br(bindex, sbinfo->si_branch[0 + bindex]); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void __au_dbg_verify_dinode(struct dentry *dentry, const char *func, int line) ++{ ++ struct inode *h_inode, *inode = d_inode(dentry); ++ struct dentry *h_dentry; ++ aufs_bindex_t bindex, bend, bi; ++ ++ if (!inode /* || au_di(dentry)->di_lsc == AuLsc_DI_TMP */) ++ return; ++ ++ bend = au_dbend(dentry); ++ bi = au_ibend(inode); ++ if (bi < bend) ++ bend = bi; ++ bindex = au_dbstart(dentry); ++ bi = au_ibstart(inode); ++ if (bi > bindex) ++ bindex = bi; ++ ++ for (; bindex <= bend; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ h_inode = au_h_iptr(inode, bindex); ++ if (unlikely(h_inode != d_inode(h_dentry))) { ++ au_debug_on(); ++ AuDbg("b%d, %s:%d\n", bindex, func, line); ++ AuDbgDentry(dentry); ++ AuDbgInode(inode); ++ au_debug_off(); ++ BUG(); ++ } ++ } ++} ++ ++void au_dbg_verify_gen(struct dentry *parent, unsigned int sigen) ++{ ++ int err, i, j; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry **dentries; ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ AuDebugOn(err); ++ err = au_dcsub_pages_rev_aufs(&dpages, parent, /*do_include*/1); ++ AuDebugOn(err); ++ for (i = dpages.ndpage - 1; !err && i >= 0; i--) { ++ dpage = dpages.dpages + i; ++ dentries = dpage->dentries; ++ for (j = dpage->ndentry - 1; !err && j >= 0; j--) ++ AuDebugOn(au_digen_test(dentries[j], sigen)); ++ } ++ au_dpages_free(&dpages); ++} ++ ++void au_dbg_verify_kthread(void) ++{ ++ if (au_wkq_test()) { ++ au_dbg_blocked(); ++ /* ++ * It may be recursive, but udba=notify between two aufs mounts, ++ * where a single ro branch is shared, is not a problem. ++ */ ++ /* WARN_ON(1); */ ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int __init au_debug_init(void) ++{ ++ aufs_bindex_t bindex; ++ struct au_vdir_destr destr; ++ ++ bindex = -1; ++ AuDebugOn(bindex >= 0); ++ ++ destr.len = -1; ++ AuDebugOn(destr.len < NAME_MAX); ++ ++#ifdef CONFIG_4KSTACKS ++ pr_warn("CONFIG_4KSTACKS is defined.\n"); ++#endif ++ ++ return 0; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/debug.h linux-4.1.10/fs/aufs/debug.h +--- linux-4.1.10.orig/fs/aufs/debug.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/debug.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,225 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * debug print functions ++ */ ++ ++#ifndef __AUFS_DEBUG_H__ ++#define __AUFS_DEBUG_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/atomic.h> ++#include <linux/module.h> ++#include <linux/kallsyms.h> ++#include <linux/sysrq.h> ++ ++#ifdef CONFIG_AUFS_DEBUG ++#define AuDebugOn(a) BUG_ON(a) ++ ++/* module parameter */ ++extern atomic_t aufs_debug; ++static inline void au_debug_on(void) ++{ ++ atomic_inc(&aufs_debug); ++} ++static inline void au_debug_off(void) ++{ ++ atomic_dec_if_positive(&aufs_debug); ++} ++ ++static inline int au_debug_test(void) ++{ ++ return atomic_read(&aufs_debug) > 0; ++} ++#else ++#define AuDebugOn(a) do {} while (0) ++AuStubVoid(au_debug_on, void) ++AuStubVoid(au_debug_off, void) ++AuStubInt0(au_debug_test, void) ++#endif /* CONFIG_AUFS_DEBUG */ ++ ++#define param_check_atomic_t(name, p) __param_check(name, p, atomic_t) ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* debug print */ ++ ++#define AuDbg(fmt, ...) do { \ ++ if (au_debug_test()) \ ++ pr_debug("DEBUG: " fmt, ##__VA_ARGS__); \ ++} while (0) ++#define AuLabel(l) AuDbg(#l "\n") ++#define AuIOErr(fmt, ...) pr_err("I/O Error, " fmt, ##__VA_ARGS__) ++#define AuWarn1(fmt, ...) do { \ ++ static unsigned char _c; \ ++ if (!_c++) \ ++ pr_warn(fmt, ##__VA_ARGS__); \ ++} while (0) ++ ++#define AuErr1(fmt, ...) do { \ ++ static unsigned char _c; \ ++ if (!_c++) \ ++ pr_err(fmt, ##__VA_ARGS__); \ ++} while (0) ++ ++#define AuIOErr1(fmt, ...) do { \ ++ static unsigned char _c; \ ++ if (!_c++) \ ++ AuIOErr(fmt, ##__VA_ARGS__); \ ++} while (0) ++ ++#define AuUnsupportMsg "This operation is not supported." \ ++ " Please report this application to aufs-users ML." ++#define AuUnsupport(fmt, ...) do { \ ++ pr_err(AuUnsupportMsg "\n" fmt, ##__VA_ARGS__); \ ++ dump_stack(); \ ++} while (0) ++ ++#define AuTraceErr(e) do { \ ++ if (unlikely((e) < 0)) \ ++ AuDbg("err %d\n", (int)(e)); \ ++} while (0) ++ ++#define AuTraceErrPtr(p) do { \ ++ if (IS_ERR(p)) \ ++ AuDbg("err %ld\n", PTR_ERR(p)); \ ++} while (0) ++ ++/* dirty macros for debug print, use with "%.*s" and caution */ ++#define AuLNPair(qstr) (qstr)->len, (qstr)->name ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct dentry; ++#ifdef CONFIG_AUFS_DEBUG ++extern struct mutex au_dbg_mtx; ++extern char *au_plevel; ++struct au_nhash; ++void au_dpri_whlist(struct au_nhash *whlist); ++struct au_vdir; ++void au_dpri_vdir(struct au_vdir *vdir); ++struct inode; ++void au_dpri_inode(struct inode *inode); ++void au_dpri_dalias(struct inode *inode); ++void au_dpri_dentry(struct dentry *dentry); ++struct file; ++void au_dpri_file(struct file *filp); ++struct super_block; ++void au_dpri_sb(struct super_block *sb); ++ ++#define au_dbg_verify_dinode(d) __au_dbg_verify_dinode(d, __func__, __LINE__) ++void __au_dbg_verify_dinode(struct dentry *dentry, const char *func, int line); ++void au_dbg_verify_gen(struct dentry *parent, unsigned int sigen); ++void au_dbg_verify_kthread(void); ++ ++int __init au_debug_init(void); ++ ++#define AuDbgWhlist(w) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#w "\n"); \ ++ au_dpri_whlist(w); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgVdir(v) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#v "\n"); \ ++ au_dpri_vdir(v); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgInode(i) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#i "\n"); \ ++ au_dpri_inode(i); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgDAlias(i) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#i "\n"); \ ++ au_dpri_dalias(i); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgDentry(d) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#d "\n"); \ ++ au_dpri_dentry(d); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgFile(f) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#f "\n"); \ ++ au_dpri_file(f); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgSb(sb) do { \ ++ mutex_lock(&au_dbg_mtx); \ ++ AuDbg(#sb "\n"); \ ++ au_dpri_sb(sb); \ ++ mutex_unlock(&au_dbg_mtx); \ ++} while (0) ++ ++#define AuDbgSym(addr) do { \ ++ char sym[KSYM_SYMBOL_LEN]; \ ++ sprint_symbol(sym, (unsigned long)addr); \ ++ AuDbg("%s\n", sym); \ ++} while (0) ++#else ++AuStubVoid(au_dbg_verify_dinode, struct dentry *dentry) ++AuStubVoid(au_dbg_verify_gen, struct dentry *parent, unsigned int sigen) ++AuStubVoid(au_dbg_verify_kthread, void) ++AuStubInt0(__init au_debug_init, void) ++ ++#define AuDbgWhlist(w) do {} while (0) ++#define AuDbgVdir(v) do {} while (0) ++#define AuDbgInode(i) do {} while (0) ++#define AuDbgDAlias(i) do {} while (0) ++#define AuDbgDentry(d) do {} while (0) ++#define AuDbgFile(f) do {} while (0) ++#define AuDbgSb(sb) do {} while (0) ++#define AuDbgSym(addr) do {} while (0) ++#endif /* CONFIG_AUFS_DEBUG */ ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_MAGIC_SYSRQ ++int __init au_sysrq_init(void); ++void au_sysrq_fin(void); ++ ++#ifdef CONFIG_HW_CONSOLE ++#define au_dbg_blocked() do { \ ++ WARN_ON(1); \ ++ handle_sysrq('w'); \ ++} while (0) ++#else ++AuStubVoid(au_dbg_blocked, void) ++#endif ++ ++#else ++AuStubInt0(__init au_sysrq_init, void) ++AuStubVoid(au_sysrq_fin, void) ++AuStubVoid(au_dbg_blocked, void) ++#endif /* CONFIG_AUFS_MAGIC_SYSRQ */ ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DEBUG_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/dentry.c linux-4.1.10/fs/aufs/dentry.c +--- linux-4.1.10.orig/fs/aufs/dentry.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dentry.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1105 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * lookup and dentry operations ++ */ ++ ++#include <linux/namei.h> ++#include "aufs.h" ++ ++#define AuLkup_ALLOW_NEG 1 ++#define AuLkup_IGNORE_PERM (1 << 1) ++#define au_ftest_lkup(flags, name) ((flags) & AuLkup_##name) ++#define au_fset_lkup(flags, name) \ ++ do { (flags) |= AuLkup_##name; } while (0) ++#define au_fclr_lkup(flags, name) \ ++ do { (flags) &= ~AuLkup_##name; } while (0) ++ ++struct au_do_lookup_args { ++ unsigned int flags; ++ mode_t type; ++}; ++ ++/* ++ * returns positive/negative dentry, NULL or an error. ++ * NULL means whiteout-ed or not-found. ++ */ ++static struct dentry* ++au_do_lookup(struct dentry *h_parent, struct dentry *dentry, ++ aufs_bindex_t bindex, struct qstr *wh_name, ++ struct au_do_lookup_args *args) ++{ ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ struct au_branch *br; ++ int wh_found, opq; ++ unsigned char wh_able; ++ const unsigned char allow_neg = !!au_ftest_lkup(args->flags, ALLOW_NEG); ++ const unsigned char ignore_perm = !!au_ftest_lkup(args->flags, ++ IGNORE_PERM); ++ ++ wh_found = 0; ++ br = au_sbr(dentry->d_sb, bindex); ++ wh_able = !!au_br_whable(br->br_perm); ++ if (wh_able) ++ wh_found = au_wh_test(h_parent, wh_name, /*try_sio*/0); ++ h_dentry = ERR_PTR(wh_found); ++ if (!wh_found) ++ goto real_lookup; ++ if (unlikely(wh_found < 0)) ++ goto out; ++ ++ /* We found a whiteout */ ++ /* au_set_dbend(dentry, bindex); */ ++ au_set_dbwh(dentry, bindex); ++ if (!allow_neg) ++ return NULL; /* success */ ++ ++real_lookup: ++ if (!ignore_perm) ++ h_dentry = vfsub_lkup_one(&dentry->d_name, h_parent); ++ else ++ h_dentry = au_sio_lkup_one(&dentry->d_name, h_parent); ++ if (IS_ERR(h_dentry)) { ++ if (PTR_ERR(h_dentry) == -ENAMETOOLONG ++ && !allow_neg) ++ h_dentry = NULL; ++ goto out; ++ } ++ ++ h_inode = d_inode(h_dentry); ++ if (d_is_negative(h_dentry)) { ++ if (!allow_neg) ++ goto out_neg; ++ } else if (wh_found ++ || (args->type && args->type != (h_inode->i_mode & S_IFMT))) ++ goto out_neg; ++ ++ if (au_dbend(dentry) <= bindex) ++ au_set_dbend(dentry, bindex); ++ if (au_dbstart(dentry) < 0 || bindex < au_dbstart(dentry)) ++ au_set_dbstart(dentry, bindex); ++ au_set_h_dptr(dentry, bindex, h_dentry); ++ ++ if (!d_is_dir(h_dentry) ++ || !wh_able ++ || (d_really_is_positive(dentry) && !d_is_dir(dentry))) ++ goto out; /* success */ ++ ++ mutex_lock_nested(&h_inode->i_mutex, AuLsc_I_CHILD); ++ opq = au_diropq_test(h_dentry); ++ mutex_unlock(&h_inode->i_mutex); ++ if (opq > 0) ++ au_set_dbdiropq(dentry, bindex); ++ else if (unlikely(opq < 0)) { ++ au_set_h_dptr(dentry, bindex, NULL); ++ h_dentry = ERR_PTR(opq); ++ } ++ goto out; ++ ++out_neg: ++ dput(h_dentry); ++ h_dentry = NULL; ++out: ++ return h_dentry; ++} ++ ++static int au_test_shwh(struct super_block *sb, const struct qstr *name) ++{ ++ if (unlikely(!au_opt_test(au_mntflags(sb), SHWH) ++ && !strncmp(name->name, AUFS_WH_PFX, AUFS_WH_PFX_LEN))) ++ return -EPERM; ++ return 0; ++} ++ ++/* ++ * returns the number of lower positive dentries, ++ * otherwise an error. ++ * can be called at unlinking with @type is zero. ++ */ ++int au_lkup_dentry(struct dentry *dentry, aufs_bindex_t bstart, mode_t type) ++{ ++ int npositive, err; ++ aufs_bindex_t bindex, btail, bdiropq; ++ unsigned char isdir, dirperm1; ++ struct qstr whname; ++ struct au_do_lookup_args args = { ++ .flags = 0, ++ .type = type ++ }; ++ const struct qstr *name = &dentry->d_name; ++ struct dentry *parent; ++ struct super_block *sb; ++ ++ sb = dentry->d_sb; ++ err = au_test_shwh(sb, name); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_wh_name_alloc(&whname, name); ++ if (unlikely(err)) ++ goto out; ++ ++ isdir = !!d_is_dir(dentry); ++ if (!type) ++ au_fset_lkup(args.flags, ALLOW_NEG); ++ dirperm1 = !!au_opt_test(au_mntflags(sb), DIRPERM1); ++ ++ npositive = 0; ++ parent = dget_parent(dentry); ++ btail = au_dbtaildir(parent); ++ for (bindex = bstart; bindex <= btail; bindex++) { ++ struct dentry *h_parent, *h_dentry; ++ struct inode *h_inode, *h_dir; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry) { ++ if (d_is_positive(h_dentry)) ++ npositive++; ++ if (type != S_IFDIR) ++ break; ++ continue; ++ } ++ h_parent = au_h_dptr(parent, bindex); ++ if (!h_parent || !d_is_dir(h_parent)) ++ continue; ++ ++ h_dir = d_inode(h_parent); ++ mutex_lock_nested(&h_dir->i_mutex, AuLsc_I_PARENT); ++ h_dentry = au_do_lookup(h_parent, dentry, bindex, &whname, ++ &args); ++ mutex_unlock(&h_dir->i_mutex); ++ err = PTR_ERR(h_dentry); ++ if (IS_ERR(h_dentry)) ++ goto out_parent; ++ if (h_dentry) ++ au_fclr_lkup(args.flags, ALLOW_NEG); ++ if (dirperm1) ++ au_fset_lkup(args.flags, IGNORE_PERM); ++ ++ if (au_dbwh(dentry) == bindex) ++ break; ++ if (!h_dentry) ++ continue; ++ if (d_is_negative(h_dentry)) ++ continue; ++ h_inode = d_inode(h_dentry); ++ npositive++; ++ if (!args.type) ++ args.type = h_inode->i_mode & S_IFMT; ++ if (args.type != S_IFDIR) ++ break; ++ else if (isdir) { ++ /* the type of lower may be different */ ++ bdiropq = au_dbdiropq(dentry); ++ if (bdiropq >= 0 && bdiropq <= bindex) ++ break; ++ } ++ } ++ ++ if (npositive) { ++ AuLabel(positive); ++ au_update_dbstart(dentry); ++ } ++ err = npositive; ++ if (unlikely(!au_opt_test(au_mntflags(sb), UDBA_NONE) ++ && au_dbstart(dentry) < 0)) { ++ err = -EIO; ++ AuIOErr("both of real entry and whiteout found, %pd, err %d\n", ++ dentry, err); ++ } ++ ++out_parent: ++ dput(parent); ++ kfree(whname.name); ++out: ++ return err; ++} ++ ++struct dentry *au_sio_lkup_one(struct qstr *name, struct dentry *parent) ++{ ++ struct dentry *dentry; ++ int wkq_err; ++ ++ if (!au_test_h_perm_sio(d_inode(parent), MAY_EXEC)) ++ dentry = vfsub_lkup_one(name, parent); ++ else { ++ struct vfsub_lkup_one_args args = { ++ .errp = &dentry, ++ .name = name, ++ .parent = parent ++ }; ++ ++ wkq_err = au_wkq_wait(vfsub_call_lkup_one, &args); ++ if (unlikely(wkq_err)) ++ dentry = ERR_PTR(wkq_err); ++ } ++ ++ return dentry; ++} ++ ++/* ++ * lookup @dentry on @bindex which should be negative. ++ */ ++int au_lkup_neg(struct dentry *dentry, aufs_bindex_t bindex, int wh) ++{ ++ int err; ++ struct dentry *parent, *h_parent, *h_dentry; ++ struct au_branch *br; ++ ++ parent = dget_parent(dentry); ++ h_parent = au_h_dptr(parent, bindex); ++ br = au_sbr(dentry->d_sb, bindex); ++ if (wh) ++ h_dentry = au_whtmp_lkup(h_parent, br, &dentry->d_name); ++ else ++ h_dentry = au_sio_lkup_one(&dentry->d_name, h_parent); ++ err = PTR_ERR(h_dentry); ++ if (IS_ERR(h_dentry)) ++ goto out; ++ if (unlikely(d_is_positive(h_dentry))) { ++ err = -EIO; ++ AuIOErr("%pd should be negative on b%d.\n", h_dentry, bindex); ++ dput(h_dentry); ++ goto out; ++ } ++ ++ err = 0; ++ if (bindex < au_dbstart(dentry)) ++ au_set_dbstart(dentry, bindex); ++ if (au_dbend(dentry) < bindex) ++ au_set_dbend(dentry, bindex); ++ au_set_h_dptr(dentry, bindex, h_dentry); ++ ++out: ++ dput(parent); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* subset of struct inode */ ++struct au_iattr { ++ unsigned long i_ino; ++ /* unsigned int i_nlink; */ ++ kuid_t i_uid; ++ kgid_t i_gid; ++ u64 i_version; ++/* ++ loff_t i_size; ++ blkcnt_t i_blocks; ++*/ ++ umode_t i_mode; ++}; ++ ++static void au_iattr_save(struct au_iattr *ia, struct inode *h_inode) ++{ ++ ia->i_ino = h_inode->i_ino; ++ /* ia->i_nlink = h_inode->i_nlink; */ ++ ia->i_uid = h_inode->i_uid; ++ ia->i_gid = h_inode->i_gid; ++ ia->i_version = h_inode->i_version; ++/* ++ ia->i_size = h_inode->i_size; ++ ia->i_blocks = h_inode->i_blocks; ++*/ ++ ia->i_mode = (h_inode->i_mode & S_IFMT); ++} ++ ++static int au_iattr_test(struct au_iattr *ia, struct inode *h_inode) ++{ ++ return ia->i_ino != h_inode->i_ino ++ /* || ia->i_nlink != h_inode->i_nlink */ ++ || !uid_eq(ia->i_uid, h_inode->i_uid) ++ || !gid_eq(ia->i_gid, h_inode->i_gid) ++ || ia->i_version != h_inode->i_version ++/* ++ || ia->i_size != h_inode->i_size ++ || ia->i_blocks != h_inode->i_blocks ++*/ ++ || ia->i_mode != (h_inode->i_mode & S_IFMT); ++} ++ ++static int au_h_verify_dentry(struct dentry *h_dentry, struct dentry *h_parent, ++ struct au_branch *br) ++{ ++ int err; ++ struct au_iattr ia; ++ struct inode *h_inode; ++ struct dentry *h_d; ++ struct super_block *h_sb; ++ ++ err = 0; ++ memset(&ia, -1, sizeof(ia)); ++ h_sb = h_dentry->d_sb; ++ h_inode = NULL; ++ if (d_is_positive(h_dentry)) { ++ h_inode = d_inode(h_dentry); ++ au_iattr_save(&ia, h_inode); ++ } else if (au_test_nfs(h_sb) || au_test_fuse(h_sb)) ++ /* nfs d_revalidate may return 0 for negative dentry */ ++ /* fuse d_revalidate always return 0 for negative dentry */ ++ goto out; ++ ++ /* main purpose is namei.c:cached_lookup() and d_revalidate */ ++ h_d = vfsub_lkup_one(&h_dentry->d_name, h_parent); ++ err = PTR_ERR(h_d); ++ if (IS_ERR(h_d)) ++ goto out; ++ ++ err = 0; ++ if (unlikely(h_d != h_dentry ++ || d_inode(h_d) != h_inode ++ || (h_inode && au_iattr_test(&ia, h_inode)))) ++ err = au_busy_or_stale(); ++ dput(h_d); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_h_verify(struct dentry *h_dentry, unsigned int udba, struct inode *h_dir, ++ struct dentry *h_parent, struct au_branch *br) ++{ ++ int err; ++ ++ err = 0; ++ if (udba == AuOpt_UDBA_REVAL ++ && !au_test_fs_remote(h_dentry->d_sb)) { ++ IMustLock(h_dir); ++ err = (d_inode(h_dentry->d_parent) != h_dir); ++ } else if (udba != AuOpt_UDBA_NONE) ++ err = au_h_verify_dentry(h_dentry, h_parent, br); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_do_refresh_hdentry(struct dentry *dentry, struct dentry *parent) ++{ ++ int err; ++ aufs_bindex_t new_bindex, bindex, bend, bwh, bdiropq; ++ struct au_hdentry tmp, *p, *q; ++ struct au_dinfo *dinfo; ++ struct super_block *sb; ++ ++ DiMustWriteLock(dentry); ++ ++ sb = dentry->d_sb; ++ dinfo = au_di(dentry); ++ bend = dinfo->di_bend; ++ bwh = dinfo->di_bwh; ++ bdiropq = dinfo->di_bdiropq; ++ p = dinfo->di_hdentry + dinfo->di_bstart; ++ for (bindex = dinfo->di_bstart; bindex <= bend; bindex++, p++) { ++ if (!p->hd_dentry) ++ continue; ++ ++ new_bindex = au_br_index(sb, p->hd_id); ++ if (new_bindex == bindex) ++ continue; ++ ++ if (dinfo->di_bwh == bindex) ++ bwh = new_bindex; ++ if (dinfo->di_bdiropq == bindex) ++ bdiropq = new_bindex; ++ if (new_bindex < 0) { ++ au_hdput(p); ++ p->hd_dentry = NULL; ++ continue; ++ } ++ ++ /* swap two lower dentries, and loop again */ ++ q = dinfo->di_hdentry + new_bindex; ++ tmp = *q; ++ *q = *p; ++ *p = tmp; ++ if (tmp.hd_dentry) { ++ bindex--; ++ p--; ++ } ++ } ++ ++ dinfo->di_bwh = -1; ++ if (bwh >= 0 && bwh <= au_sbend(sb) && au_sbr_whable(sb, bwh)) ++ dinfo->di_bwh = bwh; ++ ++ dinfo->di_bdiropq = -1; ++ if (bdiropq >= 0 ++ && bdiropq <= au_sbend(sb) ++ && au_sbr_whable(sb, bdiropq)) ++ dinfo->di_bdiropq = bdiropq; ++ ++ err = -EIO; ++ dinfo->di_bstart = -1; ++ dinfo->di_bend = -1; ++ bend = au_dbend(parent); ++ p = dinfo->di_hdentry; ++ for (bindex = 0; bindex <= bend; bindex++, p++) ++ if (p->hd_dentry) { ++ dinfo->di_bstart = bindex; ++ break; ++ } ++ ++ if (dinfo->di_bstart >= 0) { ++ p = dinfo->di_hdentry + bend; ++ for (bindex = bend; bindex >= 0; bindex--, p--) ++ if (p->hd_dentry) { ++ dinfo->di_bend = bindex; ++ err = 0; ++ break; ++ } ++ } ++ ++ return err; ++} ++ ++static void au_do_hide(struct dentry *dentry) ++{ ++ struct inode *inode; ++ ++ if (d_really_is_positive(dentry)) { ++ inode = d_inode(dentry); ++ if (!d_is_dir(dentry)) { ++ if (inode->i_nlink && !d_unhashed(dentry)) ++ drop_nlink(inode); ++ } else { ++ clear_nlink(inode); ++ /* stop next lookup */ ++ inode->i_flags |= S_DEAD; ++ } ++ smp_mb(); /* necessary? */ ++ } ++ d_drop(dentry); ++} ++ ++static int au_hide_children(struct dentry *parent) ++{ ++ int err, i, j, ndentry; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry *dentry; ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_dcsub_pages(&dpages, parent, NULL, NULL); ++ if (unlikely(err)) ++ goto out_dpages; ++ ++ /* in reverse order */ ++ for (i = dpages.ndpage - 1; i >= 0; i--) { ++ dpage = dpages.dpages + i; ++ ndentry = dpage->ndentry; ++ for (j = ndentry - 1; j >= 0; j--) { ++ dentry = dpage->dentries[j]; ++ if (dentry != parent) ++ au_do_hide(dentry); ++ } ++ } ++ ++out_dpages: ++ au_dpages_free(&dpages); ++out: ++ return err; ++} ++ ++static void au_hide(struct dentry *dentry) ++{ ++ int err; ++ ++ AuDbgDentry(dentry); ++ if (d_is_dir(dentry)) { ++ /* shrink_dcache_parent(dentry); */ ++ err = au_hide_children(dentry); ++ if (unlikely(err)) ++ AuIOErr("%pd, failed hiding children, ignored %d\n", ++ dentry, err); ++ } ++ au_do_hide(dentry); ++} ++ ++/* ++ * By adding a dirty branch, a cached dentry may be affected in various ways. ++ * ++ * a dirty branch is added ++ * - on the top of layers ++ * - in the middle of layers ++ * - to the bottom of layers ++ * ++ * on the added branch there exists ++ * - a whiteout ++ * - a diropq ++ * - a same named entry ++ * + exist ++ * * negative --> positive ++ * * positive --> positive ++ * - type is unchanged ++ * - type is changed ++ * + doesn't exist ++ * * negative --> negative ++ * * positive --> negative (rejected by au_br_del() for non-dir case) ++ * - none ++ */ ++static int au_refresh_by_dinfo(struct dentry *dentry, struct au_dinfo *dinfo, ++ struct au_dinfo *tmp) ++{ ++ int err; ++ aufs_bindex_t bindex, bend; ++ struct { ++ struct dentry *dentry; ++ struct inode *inode; ++ mode_t mode; ++ } orig_h, tmp_h; ++ struct au_hdentry *hd; ++ struct inode *inode, *h_inode; ++ struct dentry *h_dentry; ++ ++ err = 0; ++ AuDebugOn(dinfo->di_bstart < 0); ++ orig_h.mode = 0; ++ orig_h.dentry = dinfo->di_hdentry[dinfo->di_bstart].hd_dentry; ++ orig_h.inode = NULL; ++ if (d_is_positive(orig_h.dentry)) { ++ orig_h.inode = d_inode(orig_h.dentry); ++ orig_h.mode = orig_h.inode->i_mode & S_IFMT; ++ } ++ memset(&tmp_h, 0, sizeof(tmp_h)); ++ if (tmp->di_bstart >= 0) { ++ tmp_h.dentry = tmp->di_hdentry[tmp->di_bstart].hd_dentry; ++ tmp_h.inode = NULL; ++ if (d_is_positive(tmp_h.dentry)) { ++ tmp_h.inode = d_inode(tmp_h.dentry); ++ tmp_h.mode = tmp_h.inode->i_mode & S_IFMT; ++ } ++ } ++ ++ inode = NULL; ++ if (d_really_is_positive(dentry)) ++ inode = d_inode(dentry); ++ if (!orig_h.inode) { ++ AuDbg("nagative originally\n"); ++ if (inode) { ++ au_hide(dentry); ++ goto out; ++ } ++ AuDebugOn(inode); ++ AuDebugOn(dinfo->di_bstart != dinfo->di_bend); ++ AuDebugOn(dinfo->di_bdiropq != -1); ++ ++ if (!tmp_h.inode) { ++ AuDbg("negative --> negative\n"); ++ /* should have only one negative lower */ ++ if (tmp->di_bstart >= 0 ++ && tmp->di_bstart < dinfo->di_bstart) { ++ AuDebugOn(tmp->di_bstart != tmp->di_bend); ++ AuDebugOn(dinfo->di_bstart != dinfo->di_bend); ++ au_set_h_dptr(dentry, dinfo->di_bstart, NULL); ++ au_di_cp(dinfo, tmp); ++ hd = tmp->di_hdentry + tmp->di_bstart; ++ au_set_h_dptr(dentry, tmp->di_bstart, ++ dget(hd->hd_dentry)); ++ } ++ au_dbg_verify_dinode(dentry); ++ } else { ++ AuDbg("negative --> positive\n"); ++ /* ++ * similar to the behaviour of creating with bypassing ++ * aufs. ++ * unhash it in order to force an error in the ++ * succeeding create operation. ++ * we should not set S_DEAD here. ++ */ ++ d_drop(dentry); ++ /* au_di_swap(tmp, dinfo); */ ++ au_dbg_verify_dinode(dentry); ++ } ++ } else { ++ AuDbg("positive originally\n"); ++ /* inode may be NULL */ ++ AuDebugOn(inode && (inode->i_mode & S_IFMT) != orig_h.mode); ++ if (!tmp_h.inode) { ++ AuDbg("positive --> negative\n"); ++ /* or bypassing aufs */ ++ au_hide(dentry); ++ if (tmp->di_bwh >= 0 && tmp->di_bwh <= dinfo->di_bstart) ++ dinfo->di_bwh = tmp->di_bwh; ++ if (inode) ++ err = au_refresh_hinode_self(inode); ++ au_dbg_verify_dinode(dentry); ++ } else if (orig_h.mode == tmp_h.mode) { ++ AuDbg("positive --> positive, same type\n"); ++ if (!S_ISDIR(orig_h.mode) ++ && dinfo->di_bstart > tmp->di_bstart) { ++ /* ++ * similar to the behaviour of removing and ++ * creating. ++ */ ++ au_hide(dentry); ++ if (inode) ++ err = au_refresh_hinode_self(inode); ++ au_dbg_verify_dinode(dentry); ++ } else { ++ /* fill empty slots */ ++ if (dinfo->di_bstart > tmp->di_bstart) ++ dinfo->di_bstart = tmp->di_bstart; ++ if (dinfo->di_bend < tmp->di_bend) ++ dinfo->di_bend = tmp->di_bend; ++ dinfo->di_bwh = tmp->di_bwh; ++ dinfo->di_bdiropq = tmp->di_bdiropq; ++ hd = tmp->di_hdentry; ++ bend = dinfo->di_bend; ++ for (bindex = tmp->di_bstart; bindex <= bend; ++ bindex++) { ++ if (au_h_dptr(dentry, bindex)) ++ continue; ++ h_dentry = hd[bindex].hd_dentry; ++ if (!h_dentry) ++ continue; ++ AuDebugOn(d_is_negative(h_dentry)); ++ h_inode = d_inode(h_dentry); ++ AuDebugOn(orig_h.mode ++ != (h_inode->i_mode ++ & S_IFMT)); ++ au_set_h_dptr(dentry, bindex, ++ dget(h_dentry)); ++ } ++ err = au_refresh_hinode(inode, dentry); ++ au_dbg_verify_dinode(dentry); ++ } ++ } else { ++ AuDbg("positive --> positive, different type\n"); ++ /* similar to the behaviour of removing and creating */ ++ au_hide(dentry); ++ if (inode) ++ err = au_refresh_hinode_self(inode); ++ au_dbg_verify_dinode(dentry); ++ } ++ } ++ ++out: ++ return err; ++} ++ ++int au_refresh_dentry(struct dentry *dentry, struct dentry *parent) ++{ ++ int err, ebrange; ++ unsigned int sigen; ++ struct au_dinfo *dinfo, *tmp; ++ struct super_block *sb; ++ struct inode *inode; ++ ++ DiMustWriteLock(dentry); ++ AuDebugOn(IS_ROOT(dentry)); ++ AuDebugOn(d_really_is_negative(parent)); ++ ++ sb = dentry->d_sb; ++ sigen = au_sigen(sb); ++ err = au_digen_test(parent, sigen); ++ if (unlikely(err)) ++ goto out; ++ ++ dinfo = au_di(dentry); ++ err = au_di_realloc(dinfo, au_sbend(sb) + 1); ++ if (unlikely(err)) ++ goto out; ++ ebrange = au_dbrange_test(dentry); ++ if (!ebrange) ++ ebrange = au_do_refresh_hdentry(dentry, parent); ++ ++ if (d_unhashed(dentry) || ebrange /* || dinfo->di_tmpfile */) { ++ AuDebugOn(au_dbstart(dentry) < 0 && au_dbend(dentry) >= 0); ++ if (d_really_is_positive(dentry)) { ++ inode = d_inode(dentry); ++ err = au_refresh_hinode_self(inode); ++ } ++ au_dbg_verify_dinode(dentry); ++ if (!err) ++ goto out_dgen; /* success */ ++ goto out; ++ } ++ ++ /* temporary dinfo */ ++ AuDbgDentry(dentry); ++ err = -ENOMEM; ++ tmp = au_di_alloc(sb, AuLsc_DI_TMP); ++ if (unlikely(!tmp)) ++ goto out; ++ au_di_swap(tmp, dinfo); ++ /* returns the number of positive dentries */ ++ /* ++ * if current working dir is removed, it returns an error. ++ * but the dentry is legal. ++ */ ++ err = au_lkup_dentry(dentry, /*bstart*/0, /*type*/0); ++ AuDbgDentry(dentry); ++ au_di_swap(tmp, dinfo); ++ if (err == -ENOENT) ++ err = 0; ++ if (err >= 0) { ++ /* compare/refresh by dinfo */ ++ AuDbgDentry(dentry); ++ err = au_refresh_by_dinfo(dentry, dinfo, tmp); ++ au_dbg_verify_dinode(dentry); ++ AuTraceErr(err); ++ } ++ au_rw_write_unlock(&tmp->di_rwsem); ++ au_di_free(tmp); ++ if (unlikely(err)) ++ goto out; ++ ++out_dgen: ++ au_update_digen(dentry); ++out: ++ if (unlikely(err && !(dentry->d_flags & DCACHE_NFSFS_RENAMED))) { ++ AuIOErr("failed refreshing %pd, %d\n", dentry, err); ++ AuDbgDentry(dentry); ++ } ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_do_h_d_reval(struct dentry *h_dentry, unsigned int flags, ++ struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ int err, valid; ++ ++ err = 0; ++ if (!(h_dentry->d_flags & DCACHE_OP_REVALIDATE)) ++ goto out; ++ ++ AuDbg("b%d\n", bindex); ++ /* ++ * gave up supporting LOOKUP_CREATE/OPEN for lower fs, ++ * due to whiteout and branch permission. ++ */ ++ flags &= ~(/*LOOKUP_PARENT |*/ LOOKUP_OPEN | LOOKUP_CREATE ++ | LOOKUP_FOLLOW | LOOKUP_EXCL); ++ /* it may return tri-state */ ++ valid = h_dentry->d_op->d_revalidate(h_dentry, flags); ++ ++ if (unlikely(valid < 0)) ++ err = valid; ++ else if (!valid) ++ err = -EINVAL; ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* todo: remove this */ ++static int h_d_revalidate(struct dentry *dentry, struct inode *inode, ++ unsigned int flags, int do_udba) ++{ ++ int err; ++ umode_t mode, h_mode; ++ aufs_bindex_t bindex, btail, bstart, ibs, ibe; ++ unsigned char plus, unhashed, is_root, h_plus, h_nfs, tmpfile; ++ struct inode *h_inode, *h_cached_inode; ++ struct dentry *h_dentry; ++ struct qstr *name, *h_name; ++ ++ err = 0; ++ plus = 0; ++ mode = 0; ++ ibs = -1; ++ ibe = -1; ++ unhashed = !!d_unhashed(dentry); ++ is_root = !!IS_ROOT(dentry); ++ name = &dentry->d_name; ++ tmpfile = au_di(dentry)->di_tmpfile; ++ ++ /* ++ * Theoretically, REVAL test should be unnecessary in case of ++ * {FS,I}NOTIFY. ++ * But {fs,i}notify doesn't fire some necessary events, ++ * IN_ATTRIB for atime/nlink/pageio ++ * Let's do REVAL test too. ++ */ ++ if (do_udba && inode) { ++ mode = (inode->i_mode & S_IFMT); ++ plus = (inode->i_nlink > 0); ++ ibs = au_ibstart(inode); ++ ibe = au_ibend(inode); ++ } ++ ++ bstart = au_dbstart(dentry); ++ btail = bstart; ++ if (inode && S_ISDIR(inode->i_mode)) ++ btail = au_dbtaildir(dentry); ++ for (bindex = bstart; bindex <= btail; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ ++ AuDbg("b%d, %pd\n", bindex, h_dentry); ++ h_nfs = !!au_test_nfs(h_dentry->d_sb); ++ spin_lock(&h_dentry->d_lock); ++ h_name = &h_dentry->d_name; ++ if (unlikely(do_udba ++ && !is_root ++ && ((!h_nfs ++ && (unhashed != !!d_unhashed(h_dentry) ++ || (!tmpfile ++ && !au_qstreq(name, h_name)) ++ )) ++ || (h_nfs ++ && !(flags & LOOKUP_OPEN) ++ && (h_dentry->d_flags ++ & DCACHE_NFSFS_RENAMED))) ++ )) { ++ int h_unhashed; ++ ++ h_unhashed = d_unhashed(h_dentry); ++ spin_unlock(&h_dentry->d_lock); ++ AuDbg("unhash 0x%x 0x%x, %pd %pd\n", ++ unhashed, h_unhashed, dentry, h_dentry); ++ goto err; ++ } ++ spin_unlock(&h_dentry->d_lock); ++ ++ err = au_do_h_d_reval(h_dentry, flags, dentry, bindex); ++ if (unlikely(err)) ++ /* do not goto err, to keep the errno */ ++ break; ++ ++ /* todo: plink too? */ ++ if (!do_udba) ++ continue; ++ ++ /* UDBA tests */ ++ if (unlikely(!!inode != d_is_positive(h_dentry))) ++ goto err; ++ ++ h_inode = NULL; ++ if (d_is_positive(h_dentry)) ++ h_inode = d_inode(h_dentry); ++ h_plus = plus; ++ h_mode = mode; ++ h_cached_inode = h_inode; ++ if (h_inode) { ++ h_mode = (h_inode->i_mode & S_IFMT); ++ h_plus = (h_inode->i_nlink > 0); ++ } ++ if (inode && ibs <= bindex && bindex <= ibe) ++ h_cached_inode = au_h_iptr(inode, bindex); ++ ++ if (!h_nfs) { ++ if (unlikely(plus != h_plus && !tmpfile)) ++ goto err; ++ } else { ++ if (unlikely(!(h_dentry->d_flags & DCACHE_NFSFS_RENAMED) ++ && !is_root ++ && !IS_ROOT(h_dentry) ++ && unhashed != d_unhashed(h_dentry))) ++ goto err; ++ } ++ if (unlikely(mode != h_mode ++ || h_cached_inode != h_inode)) ++ goto err; ++ continue; ++ ++err: ++ err = -EINVAL; ++ break; ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++/* todo: consolidate with do_refresh() and au_reval_for_attr() */ ++static int simple_reval_dpath(struct dentry *dentry, unsigned int sigen) ++{ ++ int err; ++ struct dentry *parent; ++ ++ if (!au_digen_test(dentry, sigen)) ++ return 0; ++ ++ parent = dget_parent(dentry); ++ di_read_lock_parent(parent, AuLock_IR); ++ AuDebugOn(au_digen_test(parent, sigen)); ++ au_dbg_verify_gen(parent, sigen); ++ err = au_refresh_dentry(dentry, parent); ++ di_read_unlock(parent, AuLock_IR); ++ dput(parent); ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_reval_dpath(struct dentry *dentry, unsigned int sigen) ++{ ++ int err; ++ struct dentry *d, *parent; ++ ++ if (!au_ftest_si(au_sbi(dentry->d_sb), FAILED_REFRESH_DIR)) ++ return simple_reval_dpath(dentry, sigen); ++ ++ /* slow loop, keep it simple and stupid */ ++ /* cf: au_cpup_dirs() */ ++ err = 0; ++ parent = NULL; ++ while (au_digen_test(dentry, sigen)) { ++ d = dentry; ++ while (1) { ++ dput(parent); ++ parent = dget_parent(d); ++ if (!au_digen_test(parent, sigen)) ++ break; ++ d = parent; ++ } ++ ++ if (d != dentry) ++ di_write_lock_child2(d); ++ ++ /* someone might update our dentry while we were sleeping */ ++ if (au_digen_test(d, sigen)) { ++ /* ++ * todo: consolidate with simple_reval_dpath(), ++ * do_refresh() and au_reval_for_attr(). ++ */ ++ di_read_lock_parent(parent, AuLock_IR); ++ err = au_refresh_dentry(d, parent); ++ di_read_unlock(parent, AuLock_IR); ++ } ++ ++ if (d != dentry) ++ di_write_unlock(d); ++ dput(parent); ++ if (unlikely(err)) ++ break; ++ } ++ ++ return err; ++} ++ ++/* ++ * if valid returns 1, otherwise 0. ++ */ ++static int aufs_d_revalidate(struct dentry *dentry, unsigned int flags) ++{ ++ int valid, err; ++ unsigned int sigen; ++ unsigned char do_udba; ++ struct super_block *sb; ++ struct inode *inode; ++ ++ /* todo: support rcu-walk? */ ++ if (flags & LOOKUP_RCU) ++ return -ECHILD; ++ ++ valid = 0; ++ if (unlikely(!au_di(dentry))) ++ goto out; ++ ++ valid = 1; ++ sb = dentry->d_sb; ++ /* ++ * todo: very ugly ++ * i_mutex of parent dir may be held, ++ * but we should not return 'invalid' due to busy. ++ */ ++ err = aufs_read_lock(dentry, AuLock_FLUSH | AuLock_DW | AuLock_NOPLM); ++ if (unlikely(err)) { ++ valid = err; ++ AuTraceErr(err); ++ goto out; ++ } ++ inode = NULL; ++ if (d_really_is_positive(dentry)) ++ inode = d_inode(dentry); ++ if (unlikely(inode && is_bad_inode(inode))) { ++ err = -EINVAL; ++ AuTraceErr(err); ++ goto out_dgrade; ++ } ++ if (unlikely(au_dbrange_test(dentry))) { ++ err = -EINVAL; ++ AuTraceErr(err); ++ goto out_dgrade; ++ } ++ ++ sigen = au_sigen(sb); ++ if (au_digen_test(dentry, sigen)) { ++ AuDebugOn(IS_ROOT(dentry)); ++ err = au_reval_dpath(dentry, sigen); ++ if (unlikely(err)) { ++ AuTraceErr(err); ++ goto out_dgrade; ++ } ++ } ++ di_downgrade_lock(dentry, AuLock_IR); ++ ++ err = -EINVAL; ++ if (!(flags & (LOOKUP_OPEN | LOOKUP_EMPTY)) ++ && inode ++ && !(inode->i_state && I_LINKABLE) ++ && (IS_DEADDIR(inode) || !inode->i_nlink)) ++ goto out_inval; ++ ++ do_udba = !au_opt_test(au_mntflags(sb), UDBA_NONE); ++ if (do_udba && inode) { ++ aufs_bindex_t bstart = au_ibstart(inode); ++ struct inode *h_inode; ++ ++ if (bstart >= 0) { ++ h_inode = au_h_iptr(inode, bstart); ++ if (h_inode && au_test_higen(inode, h_inode)) ++ goto out_inval; ++ } ++ } ++ ++ err = h_d_revalidate(dentry, inode, flags, do_udba); ++ if (unlikely(!err && do_udba && au_dbstart(dentry) < 0)) { ++ err = -EIO; ++ AuDbg("both of real entry and whiteout found, %p, err %d\n", ++ dentry, err); ++ } ++ goto out_inval; ++ ++out_dgrade: ++ di_downgrade_lock(dentry, AuLock_IR); ++out_inval: ++ aufs_read_unlock(dentry, AuLock_IR); ++ AuTraceErr(err); ++ valid = !err; ++out: ++ if (!valid) { ++ AuDbg("%pd invalid, %d\n", dentry, valid); ++ d_drop(dentry); ++ } ++ return valid; ++} ++ ++static void aufs_d_release(struct dentry *dentry) ++{ ++ if (au_di(dentry)) { ++ au_di_fin(dentry); ++ au_hn_di_reinit(dentry); ++ } ++} ++ ++const struct dentry_operations aufs_dop = { ++ .d_revalidate = aufs_d_revalidate, ++ .d_weak_revalidate = aufs_d_revalidate, ++ .d_release = aufs_d_release ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/dentry.h linux-4.1.10/fs/aufs/dentry.h +--- linux-4.1.10.orig/fs/aufs/dentry.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dentry.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,233 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * lookup and dentry operations ++ */ ++ ++#ifndef __AUFS_DENTRY_H__ ++#define __AUFS_DENTRY_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/dcache.h> ++#include "rwsem.h" ++ ++struct au_hdentry { ++ struct dentry *hd_dentry; ++ aufs_bindex_t hd_id; ++}; ++ ++struct au_dinfo { ++ atomic_t di_generation; ++ ++ struct au_rwsem di_rwsem; ++ aufs_bindex_t di_bstart, di_bend, di_bwh, di_bdiropq; ++ unsigned char di_tmpfile; /* to allow the different name */ ++ struct au_hdentry *di_hdentry; ++} ____cacheline_aligned_in_smp; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* dentry.c */ ++extern const struct dentry_operations aufs_dop; ++struct au_branch; ++struct dentry *au_sio_lkup_one(struct qstr *name, struct dentry *parent); ++int au_h_verify(struct dentry *h_dentry, unsigned int udba, struct inode *h_dir, ++ struct dentry *h_parent, struct au_branch *br); ++ ++int au_lkup_dentry(struct dentry *dentry, aufs_bindex_t bstart, mode_t type); ++int au_lkup_neg(struct dentry *dentry, aufs_bindex_t bindex, int wh); ++int au_refresh_dentry(struct dentry *dentry, struct dentry *parent); ++int au_reval_dpath(struct dentry *dentry, unsigned int sigen); ++ ++/* dinfo.c */ ++void au_di_init_once(void *_di); ++struct au_dinfo *au_di_alloc(struct super_block *sb, unsigned int lsc); ++void au_di_free(struct au_dinfo *dinfo); ++void au_di_swap(struct au_dinfo *a, struct au_dinfo *b); ++void au_di_cp(struct au_dinfo *dst, struct au_dinfo *src); ++int au_di_init(struct dentry *dentry); ++void au_di_fin(struct dentry *dentry); ++int au_di_realloc(struct au_dinfo *dinfo, int nbr); ++ ++void di_read_lock(struct dentry *d, int flags, unsigned int lsc); ++void di_read_unlock(struct dentry *d, int flags); ++void di_downgrade_lock(struct dentry *d, int flags); ++void di_write_lock(struct dentry *d, unsigned int lsc); ++void di_write_unlock(struct dentry *d); ++void di_write_lock2_child(struct dentry *d1, struct dentry *d2, int isdir); ++void di_write_lock2_parent(struct dentry *d1, struct dentry *d2, int isdir); ++void di_write_unlock2(struct dentry *d1, struct dentry *d2); ++ ++struct dentry *au_h_dptr(struct dentry *dentry, aufs_bindex_t bindex); ++struct dentry *au_h_d_alias(struct dentry *dentry, aufs_bindex_t bindex); ++aufs_bindex_t au_dbtail(struct dentry *dentry); ++aufs_bindex_t au_dbtaildir(struct dentry *dentry); ++ ++void au_set_h_dptr(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_dentry); ++int au_digen_test(struct dentry *dentry, unsigned int sigen); ++int au_dbrange_test(struct dentry *dentry); ++void au_update_digen(struct dentry *dentry); ++void au_update_dbrange(struct dentry *dentry, int do_put_zero); ++void au_update_dbstart(struct dentry *dentry); ++void au_update_dbend(struct dentry *dentry); ++int au_find_dbindex(struct dentry *dentry, struct dentry *h_dentry); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct au_dinfo *au_di(struct dentry *dentry) ++{ ++ return dentry->d_fsdata; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* lock subclass for dinfo */ ++enum { ++ AuLsc_DI_CHILD, /* child first */ ++ AuLsc_DI_CHILD2, /* rename(2), link(2), and cpup at hnotify */ ++ AuLsc_DI_CHILD3, /* copyup dirs */ ++ AuLsc_DI_PARENT, ++ AuLsc_DI_PARENT2, ++ AuLsc_DI_PARENT3, ++ AuLsc_DI_TMP /* temp for replacing dinfo */ ++}; ++ ++/* ++ * di_read_lock_child, di_write_lock_child, ++ * di_read_lock_child2, di_write_lock_child2, ++ * di_read_lock_child3, di_write_lock_child3, ++ * di_read_lock_parent, di_write_lock_parent, ++ * di_read_lock_parent2, di_write_lock_parent2, ++ * di_read_lock_parent3, di_write_lock_parent3, ++ */ ++#define AuReadLockFunc(name, lsc) \ ++static inline void di_read_lock_##name(struct dentry *d, int flags) \ ++{ di_read_lock(d, flags, AuLsc_DI_##lsc); } ++ ++#define AuWriteLockFunc(name, lsc) \ ++static inline void di_write_lock_##name(struct dentry *d) \ ++{ di_write_lock(d, AuLsc_DI_##lsc); } ++ ++#define AuRWLockFuncs(name, lsc) \ ++ AuReadLockFunc(name, lsc) \ ++ AuWriteLockFunc(name, lsc) ++ ++AuRWLockFuncs(child, CHILD); ++AuRWLockFuncs(child2, CHILD2); ++AuRWLockFuncs(child3, CHILD3); ++AuRWLockFuncs(parent, PARENT); ++AuRWLockFuncs(parent2, PARENT2); ++AuRWLockFuncs(parent3, PARENT3); ++ ++#undef AuReadLockFunc ++#undef AuWriteLockFunc ++#undef AuRWLockFuncs ++ ++#define DiMustNoWaiters(d) AuRwMustNoWaiters(&au_di(d)->di_rwsem) ++#define DiMustAnyLock(d) AuRwMustAnyLock(&au_di(d)->di_rwsem) ++#define DiMustWriteLock(d) AuRwMustWriteLock(&au_di(d)->di_rwsem) ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* todo: memory barrier? */ ++static inline unsigned int au_digen(struct dentry *d) ++{ ++ return atomic_read(&au_di(d)->di_generation); ++} ++ ++static inline void au_h_dentry_init(struct au_hdentry *hdentry) ++{ ++ hdentry->hd_dentry = NULL; ++} ++ ++static inline void au_hdput(struct au_hdentry *hd) ++{ ++ if (hd) ++ dput(hd->hd_dentry); ++} ++ ++static inline aufs_bindex_t au_dbstart(struct dentry *dentry) ++{ ++ DiMustAnyLock(dentry); ++ return au_di(dentry)->di_bstart; ++} ++ ++static inline aufs_bindex_t au_dbend(struct dentry *dentry) ++{ ++ DiMustAnyLock(dentry); ++ return au_di(dentry)->di_bend; ++} ++ ++static inline aufs_bindex_t au_dbwh(struct dentry *dentry) ++{ ++ DiMustAnyLock(dentry); ++ return au_di(dentry)->di_bwh; ++} ++ ++static inline aufs_bindex_t au_dbdiropq(struct dentry *dentry) ++{ ++ DiMustAnyLock(dentry); ++ return au_di(dentry)->di_bdiropq; ++} ++ ++/* todo: hard/soft set? */ ++static inline void au_set_dbstart(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ DiMustWriteLock(dentry); ++ au_di(dentry)->di_bstart = bindex; ++} ++ ++static inline void au_set_dbend(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ DiMustWriteLock(dentry); ++ au_di(dentry)->di_bend = bindex; ++} ++ ++static inline void au_set_dbwh(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ DiMustWriteLock(dentry); ++ /* dbwh can be outside of bstart - bend range */ ++ au_di(dentry)->di_bwh = bindex; ++} ++ ++static inline void au_set_dbdiropq(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ DiMustWriteLock(dentry); ++ au_di(dentry)->di_bdiropq = bindex; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_HNOTIFY ++static inline void au_digen_dec(struct dentry *d) ++{ ++ atomic_dec(&au_di(d)->di_generation); ++} ++ ++static inline void au_hn_di_reinit(struct dentry *dentry) ++{ ++ dentry->d_fsdata = NULL; ++} ++#else ++AuStubVoid(au_hn_di_reinit, struct dentry *dentry __maybe_unused) ++#endif /* CONFIG_AUFS_HNOTIFY */ ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DENTRY_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/dinfo.c linux-4.1.10/fs/aufs/dinfo.c +--- linux-4.1.10.orig/fs/aufs/dinfo.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dinfo.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,550 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * dentry private data ++ */ ++ ++#include "aufs.h" ++ ++void au_di_init_once(void *_dinfo) ++{ ++ struct au_dinfo *dinfo = _dinfo; ++ static struct lock_class_key aufs_di; ++ ++ au_rw_init(&dinfo->di_rwsem); ++ au_rw_class(&dinfo->di_rwsem, &aufs_di); ++} ++ ++struct au_dinfo *au_di_alloc(struct super_block *sb, unsigned int lsc) ++{ ++ struct au_dinfo *dinfo; ++ int nbr, i; ++ ++ dinfo = au_cache_alloc_dinfo(); ++ if (unlikely(!dinfo)) ++ goto out; ++ ++ nbr = au_sbend(sb) + 1; ++ if (nbr <= 0) ++ nbr = 1; ++ dinfo->di_hdentry = kcalloc(nbr, sizeof(*dinfo->di_hdentry), GFP_NOFS); ++ if (dinfo->di_hdentry) { ++ au_rw_write_lock_nested(&dinfo->di_rwsem, lsc); ++ dinfo->di_bstart = -1; ++ dinfo->di_bend = -1; ++ dinfo->di_bwh = -1; ++ dinfo->di_bdiropq = -1; ++ dinfo->di_tmpfile = 0; ++ for (i = 0; i < nbr; i++) ++ dinfo->di_hdentry[i].hd_id = -1; ++ goto out; ++ } ++ ++ au_cache_free_dinfo(dinfo); ++ dinfo = NULL; ++ ++out: ++ return dinfo; ++} ++ ++void au_di_free(struct au_dinfo *dinfo) ++{ ++ struct au_hdentry *p; ++ aufs_bindex_t bend, bindex; ++ ++ /* dentry may not be revalidated */ ++ bindex = dinfo->di_bstart; ++ if (bindex >= 0) { ++ bend = dinfo->di_bend; ++ p = dinfo->di_hdentry + bindex; ++ while (bindex++ <= bend) ++ au_hdput(p++); ++ } ++ kfree(dinfo->di_hdentry); ++ au_cache_free_dinfo(dinfo); ++} ++ ++void au_di_swap(struct au_dinfo *a, struct au_dinfo *b) ++{ ++ struct au_hdentry *p; ++ aufs_bindex_t bi; ++ ++ AuRwMustWriteLock(&a->di_rwsem); ++ AuRwMustWriteLock(&b->di_rwsem); ++ ++#define DiSwap(v, name) \ ++ do { \ ++ v = a->di_##name; \ ++ a->di_##name = b->di_##name; \ ++ b->di_##name = v; \ ++ } while (0) ++ ++ DiSwap(p, hdentry); ++ DiSwap(bi, bstart); ++ DiSwap(bi, bend); ++ DiSwap(bi, bwh); ++ DiSwap(bi, bdiropq); ++ /* smp_mb(); */ ++ ++#undef DiSwap ++} ++ ++void au_di_cp(struct au_dinfo *dst, struct au_dinfo *src) ++{ ++ AuRwMustWriteLock(&dst->di_rwsem); ++ AuRwMustWriteLock(&src->di_rwsem); ++ ++ dst->di_bstart = src->di_bstart; ++ dst->di_bend = src->di_bend; ++ dst->di_bwh = src->di_bwh; ++ dst->di_bdiropq = src->di_bdiropq; ++ /* smp_mb(); */ ++} ++ ++int au_di_init(struct dentry *dentry) ++{ ++ int err; ++ struct super_block *sb; ++ struct au_dinfo *dinfo; ++ ++ err = 0; ++ sb = dentry->d_sb; ++ dinfo = au_di_alloc(sb, AuLsc_DI_CHILD); ++ if (dinfo) { ++ atomic_set(&dinfo->di_generation, au_sigen(sb)); ++ /* smp_mb(); */ /* atomic_set */ ++ dentry->d_fsdata = dinfo; ++ } else ++ err = -ENOMEM; ++ ++ return err; ++} ++ ++void au_di_fin(struct dentry *dentry) ++{ ++ struct au_dinfo *dinfo; ++ ++ dinfo = au_di(dentry); ++ AuRwDestroy(&dinfo->di_rwsem); ++ au_di_free(dinfo); ++} ++ ++int au_di_realloc(struct au_dinfo *dinfo, int nbr) ++{ ++ int err, sz; ++ struct au_hdentry *hdp; ++ ++ AuRwMustWriteLock(&dinfo->di_rwsem); ++ ++ err = -ENOMEM; ++ sz = sizeof(*hdp) * (dinfo->di_bend + 1); ++ if (!sz) ++ sz = sizeof(*hdp); ++ hdp = au_kzrealloc(dinfo->di_hdentry, sz, sizeof(*hdp) * nbr, GFP_NOFS); ++ if (hdp) { ++ dinfo->di_hdentry = hdp; ++ err = 0; ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void do_ii_write_lock(struct inode *inode, unsigned int lsc) ++{ ++ switch (lsc) { ++ case AuLsc_DI_CHILD: ++ ii_write_lock_child(inode); ++ break; ++ case AuLsc_DI_CHILD2: ++ ii_write_lock_child2(inode); ++ break; ++ case AuLsc_DI_CHILD3: ++ ii_write_lock_child3(inode); ++ break; ++ case AuLsc_DI_PARENT: ++ ii_write_lock_parent(inode); ++ break; ++ case AuLsc_DI_PARENT2: ++ ii_write_lock_parent2(inode); ++ break; ++ case AuLsc_DI_PARENT3: ++ ii_write_lock_parent3(inode); ++ break; ++ default: ++ BUG(); ++ } ++} ++ ++static void do_ii_read_lock(struct inode *inode, unsigned int lsc) ++{ ++ switch (lsc) { ++ case AuLsc_DI_CHILD: ++ ii_read_lock_child(inode); ++ break; ++ case AuLsc_DI_CHILD2: ++ ii_read_lock_child2(inode); ++ break; ++ case AuLsc_DI_CHILD3: ++ ii_read_lock_child3(inode); ++ break; ++ case AuLsc_DI_PARENT: ++ ii_read_lock_parent(inode); ++ break; ++ case AuLsc_DI_PARENT2: ++ ii_read_lock_parent2(inode); ++ break; ++ case AuLsc_DI_PARENT3: ++ ii_read_lock_parent3(inode); ++ break; ++ default: ++ BUG(); ++ } ++} ++ ++void di_read_lock(struct dentry *d, int flags, unsigned int lsc) ++{ ++ struct inode *inode; ++ ++ au_rw_read_lock_nested(&au_di(d)->di_rwsem, lsc); ++ if (d_really_is_positive(d)) { ++ inode = d_inode(d); ++ if (au_ftest_lock(flags, IW)) ++ do_ii_write_lock(inode, lsc); ++ else if (au_ftest_lock(flags, IR)) ++ do_ii_read_lock(inode, lsc); ++ } ++} ++ ++void di_read_unlock(struct dentry *d, int flags) ++{ ++ struct inode *inode; ++ ++ if (d_really_is_positive(d)) { ++ inode = d_inode(d); ++ if (au_ftest_lock(flags, IW)) { ++ au_dbg_verify_dinode(d); ++ ii_write_unlock(inode); ++ } else if (au_ftest_lock(flags, IR)) { ++ au_dbg_verify_dinode(d); ++ ii_read_unlock(inode); ++ } ++ } ++ au_rw_read_unlock(&au_di(d)->di_rwsem); ++} ++ ++void di_downgrade_lock(struct dentry *d, int flags) ++{ ++ if (d_really_is_positive(d) && au_ftest_lock(flags, IR)) ++ ii_downgrade_lock(d_inode(d)); ++ au_rw_dgrade_lock(&au_di(d)->di_rwsem); ++} ++ ++void di_write_lock(struct dentry *d, unsigned int lsc) ++{ ++ au_rw_write_lock_nested(&au_di(d)->di_rwsem, lsc); ++ if (d_really_is_positive(d)) ++ do_ii_write_lock(d_inode(d), lsc); ++} ++ ++void di_write_unlock(struct dentry *d) ++{ ++ au_dbg_verify_dinode(d); ++ if (d_really_is_positive(d)) ++ ii_write_unlock(d_inode(d)); ++ au_rw_write_unlock(&au_di(d)->di_rwsem); ++} ++ ++void di_write_lock2_child(struct dentry *d1, struct dentry *d2, int isdir) ++{ ++ AuDebugOn(d1 == d2 ++ || d_inode(d1) == d_inode(d2) ++ || d1->d_sb != d2->d_sb); ++ ++ if (isdir && au_test_subdir(d1, d2)) { ++ di_write_lock_child(d1); ++ di_write_lock_child2(d2); ++ } else { ++ /* there should be no races */ ++ di_write_lock_child(d2); ++ di_write_lock_child2(d1); ++ } ++} ++ ++void di_write_lock2_parent(struct dentry *d1, struct dentry *d2, int isdir) ++{ ++ AuDebugOn(d1 == d2 ++ || d_inode(d1) == d_inode(d2) ++ || d1->d_sb != d2->d_sb); ++ ++ if (isdir && au_test_subdir(d1, d2)) { ++ di_write_lock_parent(d1); ++ di_write_lock_parent2(d2); ++ } else { ++ /* there should be no races */ ++ di_write_lock_parent(d2); ++ di_write_lock_parent2(d1); ++ } ++} ++ ++void di_write_unlock2(struct dentry *d1, struct dentry *d2) ++{ ++ di_write_unlock(d1); ++ if (d_inode(d1) == d_inode(d2)) ++ au_rw_write_unlock(&au_di(d2)->di_rwsem); ++ else ++ di_write_unlock(d2); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct dentry *au_h_dptr(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ struct dentry *d; ++ ++ DiMustAnyLock(dentry); ++ ++ if (au_dbstart(dentry) < 0 || bindex < au_dbstart(dentry)) ++ return NULL; ++ AuDebugOn(bindex < 0); ++ d = au_di(dentry)->di_hdentry[0 + bindex].hd_dentry; ++ AuDebugOn(d && au_dcount(d) <= 0); ++ return d; ++} ++ ++/* ++ * extended version of au_h_dptr(). ++ * returns a hashed and positive (or linkable) h_dentry in bindex, NULL, or ++ * error. ++ */ ++struct dentry *au_h_d_alias(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ struct dentry *h_dentry; ++ struct inode *inode, *h_inode; ++ ++ AuDebugOn(d_really_is_negative(dentry)); ++ ++ h_dentry = NULL; ++ if (au_dbstart(dentry) <= bindex ++ && bindex <= au_dbend(dentry)) ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry && !au_d_linkable(h_dentry)) { ++ dget(h_dentry); ++ goto out; /* success */ ++ } ++ ++ inode = d_inode(dentry); ++ AuDebugOn(bindex < au_ibstart(inode)); ++ AuDebugOn(au_ibend(inode) < bindex); ++ h_inode = au_h_iptr(inode, bindex); ++ h_dentry = d_find_alias(h_inode); ++ if (h_dentry) { ++ if (!IS_ERR(h_dentry)) { ++ if (!au_d_linkable(h_dentry)) ++ goto out; /* success */ ++ dput(h_dentry); ++ } else ++ goto out; ++ } ++ ++ if (au_opt_test(au_mntflags(dentry->d_sb), PLINK)) { ++ h_dentry = au_plink_lkup(inode, bindex); ++ AuDebugOn(!h_dentry); ++ if (!IS_ERR(h_dentry)) { ++ if (!au_d_hashed_positive(h_dentry)) ++ goto out; /* success */ ++ dput(h_dentry); ++ h_dentry = NULL; ++ } ++ } ++ ++out: ++ AuDbgDentry(h_dentry); ++ return h_dentry; ++} ++ ++aufs_bindex_t au_dbtail(struct dentry *dentry) ++{ ++ aufs_bindex_t bend, bwh; ++ ++ bend = au_dbend(dentry); ++ if (0 <= bend) { ++ bwh = au_dbwh(dentry); ++ if (!bwh) ++ return bwh; ++ if (0 < bwh && bwh < bend) ++ return bwh - 1; ++ } ++ return bend; ++} ++ ++aufs_bindex_t au_dbtaildir(struct dentry *dentry) ++{ ++ aufs_bindex_t bend, bopq; ++ ++ bend = au_dbtail(dentry); ++ if (0 <= bend) { ++ bopq = au_dbdiropq(dentry); ++ if (0 <= bopq && bopq < bend) ++ bend = bopq; ++ } ++ return bend; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_set_h_dptr(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_dentry) ++{ ++ struct au_hdentry *hd = au_di(dentry)->di_hdentry + bindex; ++ struct au_branch *br; ++ ++ DiMustWriteLock(dentry); ++ ++ au_hdput(hd); ++ hd->hd_dentry = h_dentry; ++ if (h_dentry) { ++ br = au_sbr(dentry->d_sb, bindex); ++ hd->hd_id = br->br_id; ++ } ++} ++ ++int au_dbrange_test(struct dentry *dentry) ++{ ++ int err; ++ aufs_bindex_t bstart, bend; ++ ++ err = 0; ++ bstart = au_dbstart(dentry); ++ bend = au_dbend(dentry); ++ if (bstart >= 0) ++ AuDebugOn(bend < 0 && bstart > bend); ++ else { ++ err = -EIO; ++ AuDebugOn(bend >= 0); ++ } ++ ++ return err; ++} ++ ++int au_digen_test(struct dentry *dentry, unsigned int sigen) ++{ ++ int err; ++ ++ err = 0; ++ if (unlikely(au_digen(dentry) != sigen ++ || au_iigen_test(d_inode(dentry), sigen))) ++ err = -EIO; ++ ++ return err; ++} ++ ++void au_update_digen(struct dentry *dentry) ++{ ++ atomic_set(&au_di(dentry)->di_generation, au_sigen(dentry->d_sb)); ++ /* smp_mb(); */ /* atomic_set */ ++} ++ ++void au_update_dbrange(struct dentry *dentry, int do_put_zero) ++{ ++ struct au_dinfo *dinfo; ++ struct dentry *h_d; ++ struct au_hdentry *hdp; ++ ++ DiMustWriteLock(dentry); ++ ++ dinfo = au_di(dentry); ++ if (!dinfo || dinfo->di_bstart < 0) ++ return; ++ ++ hdp = dinfo->di_hdentry; ++ if (do_put_zero) { ++ aufs_bindex_t bindex, bend; ++ ++ bend = dinfo->di_bend; ++ for (bindex = dinfo->di_bstart; bindex <= bend; bindex++) { ++ h_d = hdp[0 + bindex].hd_dentry; ++ if (h_d && d_is_negative(h_d)) ++ au_set_h_dptr(dentry, bindex, NULL); ++ } ++ } ++ ++ dinfo->di_bstart = -1; ++ while (++dinfo->di_bstart <= dinfo->di_bend) ++ if (hdp[0 + dinfo->di_bstart].hd_dentry) ++ break; ++ if (dinfo->di_bstart > dinfo->di_bend) { ++ dinfo->di_bstart = -1; ++ dinfo->di_bend = -1; ++ return; ++ } ++ ++ dinfo->di_bend++; ++ while (0 <= --dinfo->di_bend) ++ if (hdp[0 + dinfo->di_bend].hd_dentry) ++ break; ++ AuDebugOn(dinfo->di_bstart > dinfo->di_bend || dinfo->di_bend < 0); ++} ++ ++void au_update_dbstart(struct dentry *dentry) ++{ ++ aufs_bindex_t bindex, bend; ++ struct dentry *h_dentry; ++ ++ bend = au_dbend(dentry); ++ for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ if (d_is_positive(h_dentry)) { ++ au_set_dbstart(dentry, bindex); ++ return; ++ } ++ au_set_h_dptr(dentry, bindex, NULL); ++ } ++} ++ ++void au_update_dbend(struct dentry *dentry) ++{ ++ aufs_bindex_t bindex, bstart; ++ struct dentry *h_dentry; ++ ++ bstart = au_dbstart(dentry); ++ for (bindex = au_dbend(dentry); bindex >= bstart; bindex--) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ if (d_is_positive(h_dentry)) { ++ au_set_dbend(dentry, bindex); ++ return; ++ } ++ au_set_h_dptr(dentry, bindex, NULL); ++ } ++} ++ ++int au_find_dbindex(struct dentry *dentry, struct dentry *h_dentry) ++{ ++ aufs_bindex_t bindex, bend; ++ ++ bend = au_dbend(dentry); ++ for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) ++ if (au_h_dptr(dentry, bindex) == h_dentry) ++ return bindex; ++ return -1; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/dir.c linux-4.1.10/fs/aufs/dir.c +--- linux-4.1.10.orig/fs/aufs/dir.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dir.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,753 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * directory operations ++ */ ++ ++#include <linux/fs_stack.h> ++#include "aufs.h" ++ ++void au_add_nlink(struct inode *dir, struct inode *h_dir) ++{ ++ unsigned int nlink; ++ ++ AuDebugOn(!S_ISDIR(dir->i_mode) || !S_ISDIR(h_dir->i_mode)); ++ ++ nlink = dir->i_nlink; ++ nlink += h_dir->i_nlink - 2; ++ if (h_dir->i_nlink < 2) ++ nlink += 2; ++ smp_mb(); /* for i_nlink */ ++ /* 0 can happen in revaliding */ ++ set_nlink(dir, nlink); ++} ++ ++void au_sub_nlink(struct inode *dir, struct inode *h_dir) ++{ ++ unsigned int nlink; ++ ++ AuDebugOn(!S_ISDIR(dir->i_mode) || !S_ISDIR(h_dir->i_mode)); ++ ++ nlink = dir->i_nlink; ++ nlink -= h_dir->i_nlink - 2; ++ if (h_dir->i_nlink < 2) ++ nlink -= 2; ++ smp_mb(); /* for i_nlink */ ++ /* nlink == 0 means the branch-fs is broken */ ++ set_nlink(dir, nlink); ++} ++ ++loff_t au_dir_size(struct file *file, struct dentry *dentry) ++{ ++ loff_t sz; ++ aufs_bindex_t bindex, bend; ++ struct file *h_file; ++ struct dentry *h_dentry; ++ ++ sz = 0; ++ if (file) { ++ AuDebugOn(!d_is_dir(file->f_path.dentry)); ++ ++ bend = au_fbend_dir(file); ++ for (bindex = au_fbstart(file); ++ bindex <= bend && sz < KMALLOC_MAX_SIZE; ++ bindex++) { ++ h_file = au_hf_dir(file, bindex); ++ if (h_file && file_inode(h_file)) ++ sz += vfsub_f_size_read(h_file); ++ } ++ } else { ++ AuDebugOn(!dentry); ++ AuDebugOn(!d_is_dir(dentry)); ++ ++ bend = au_dbtaildir(dentry); ++ for (bindex = au_dbstart(dentry); ++ bindex <= bend && sz < KMALLOC_MAX_SIZE; ++ bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry && d_is_positive(h_dentry)) ++ sz += i_size_read(d_inode(h_dentry)); ++ } ++ } ++ if (sz < KMALLOC_MAX_SIZE) ++ sz = roundup_pow_of_two(sz); ++ if (sz > KMALLOC_MAX_SIZE) ++ sz = KMALLOC_MAX_SIZE; ++ else if (sz < NAME_MAX) { ++ BUILD_BUG_ON(AUFS_RDBLK_DEF < NAME_MAX); ++ sz = AUFS_RDBLK_DEF; ++ } ++ return sz; ++} ++ ++struct au_dir_ts_arg { ++ struct dentry *dentry; ++ aufs_bindex_t brid; ++}; ++ ++static void au_do_dir_ts(void *arg) ++{ ++ struct au_dir_ts_arg *a = arg; ++ struct au_dtime dt; ++ struct path h_path; ++ struct inode *dir, *h_dir; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_hinode *hdir; ++ int err; ++ aufs_bindex_t bstart, bindex; ++ ++ sb = a->dentry->d_sb; ++ if (d_really_is_negative(a->dentry)) ++ goto out; ++ aufs_read_lock(a->dentry, AuLock_DW | AuLock_DIR); /* noflush */ ++ ++ /* no dir->i_mutex lock */ ++ dir = d_inode(a->dentry); ++ bstart = au_ibstart(dir); ++ bindex = au_br_index(sb, a->brid); ++ if (bindex < bstart) ++ goto out_unlock; ++ ++ br = au_sbr(sb, bindex); ++ h_path.dentry = au_h_dptr(a->dentry, bindex); ++ if (!h_path.dentry) ++ goto out_unlock; ++ h_path.mnt = au_br_mnt(br); ++ au_dtime_store(&dt, a->dentry, &h_path); ++ ++ br = au_sbr(sb, bstart); ++ if (!au_br_writable(br->br_perm)) ++ goto out_unlock; ++ h_path.dentry = au_h_dptr(a->dentry, bstart); ++ h_path.mnt = au_br_mnt(br); ++ err = vfsub_mnt_want_write(h_path.mnt); ++ if (err) ++ goto out_unlock; ++ hdir = au_hi(dir, bstart); ++ au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); ++ h_dir = au_h_iptr(dir, bstart); ++ if (h_dir->i_nlink ++ && timespec_compare(&h_dir->i_mtime, &dt.dt_mtime) < 0) { ++ dt.dt_h_path = h_path; ++ au_dtime_revert(&dt); ++ } ++ au_hn_imtx_unlock(hdir); ++ vfsub_mnt_drop_write(h_path.mnt); ++ au_cpup_attr_timesizes(dir); ++ ++out_unlock: ++ aufs_read_unlock(a->dentry, AuLock_DW); ++out: ++ dput(a->dentry); ++ au_nwt_done(&au_sbi(sb)->si_nowait); ++ kfree(arg); ++} ++ ++void au_dir_ts(struct inode *dir, aufs_bindex_t bindex) ++{ ++ int perm, wkq_err; ++ aufs_bindex_t bstart; ++ struct au_dir_ts_arg *arg; ++ struct dentry *dentry; ++ struct super_block *sb; ++ ++ IMustLock(dir); ++ ++ dentry = d_find_any_alias(dir); ++ AuDebugOn(!dentry); ++ sb = dentry->d_sb; ++ bstart = au_ibstart(dir); ++ if (bstart == bindex) { ++ au_cpup_attr_timesizes(dir); ++ goto out; ++ } ++ ++ perm = au_sbr_perm(sb, bstart); ++ if (!au_br_writable(perm)) ++ goto out; ++ ++ arg = kmalloc(sizeof(*arg), GFP_NOFS); ++ if (!arg) ++ goto out; ++ ++ arg->dentry = dget(dentry); /* will be dput-ted by au_do_dir_ts() */ ++ arg->brid = au_sbr_id(sb, bindex); ++ wkq_err = au_wkq_nowait(au_do_dir_ts, arg, sb, /*flags*/0); ++ if (unlikely(wkq_err)) { ++ pr_err("wkq %d\n", wkq_err); ++ dput(dentry); ++ kfree(arg); ++ } ++ ++out: ++ dput(dentry); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int reopen_dir(struct file *file) ++{ ++ int err; ++ unsigned int flags; ++ aufs_bindex_t bindex, btail, bstart; ++ struct dentry *dentry, *h_dentry; ++ struct file *h_file; ++ ++ /* open all lower dirs */ ++ dentry = file->f_path.dentry; ++ bstart = au_dbstart(dentry); ++ for (bindex = au_fbstart(file); bindex < bstart; bindex++) ++ au_set_h_fptr(file, bindex, NULL); ++ au_set_fbstart(file, bstart); ++ ++ btail = au_dbtaildir(dentry); ++ for (bindex = au_fbend_dir(file); btail < bindex; bindex--) ++ au_set_h_fptr(file, bindex, NULL); ++ au_set_fbend_dir(file, btail); ++ ++ flags = vfsub_file_flags(file); ++ for (bindex = bstart; bindex <= btail; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ h_file = au_hf_dir(file, bindex); ++ if (h_file) ++ continue; ++ ++ h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; /* close all? */ ++ au_set_h_fptr(file, bindex, h_file); ++ } ++ au_update_figen(file); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ err = 0; ++ ++out: ++ return err; ++} ++ ++static int do_open_dir(struct file *file, int flags, struct file *h_file) ++{ ++ int err; ++ aufs_bindex_t bindex, btail; ++ struct dentry *dentry, *h_dentry; ++ ++ FiMustWriteLock(file); ++ AuDebugOn(h_file); ++ ++ err = 0; ++ dentry = file->f_path.dentry; ++ file->f_version = d_inode(dentry)->i_version; ++ bindex = au_dbstart(dentry); ++ au_set_fbstart(file, bindex); ++ btail = au_dbtaildir(dentry); ++ au_set_fbend_dir(file, btail); ++ for (; !err && bindex <= btail; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!h_dentry) ++ continue; ++ ++ h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); ++ if (IS_ERR(h_file)) { ++ err = PTR_ERR(h_file); ++ break; ++ } ++ au_set_h_fptr(file, bindex, h_file); ++ } ++ au_update_figen(file); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ if (!err) ++ return 0; /* success */ ++ ++ /* close all */ ++ for (bindex = au_fbstart(file); bindex <= btail; bindex++) ++ au_set_h_fptr(file, bindex, NULL); ++ au_set_fbstart(file, -1); ++ au_set_fbend_dir(file, -1); ++ ++ return err; ++} ++ ++static int aufs_open_dir(struct inode *inode __maybe_unused, ++ struct file *file) ++{ ++ int err; ++ struct super_block *sb; ++ struct au_fidir *fidir; ++ ++ err = -ENOMEM; ++ sb = file->f_path.dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ fidir = au_fidir_alloc(sb); ++ if (fidir) { ++ struct au_do_open_args args = { ++ .open = do_open_dir, ++ .fidir = fidir ++ }; ++ err = au_do_open(file, &args); ++ if (unlikely(err)) ++ kfree(fidir); ++ } ++ si_read_unlock(sb); ++ return err; ++} ++ ++static int aufs_release_dir(struct inode *inode __maybe_unused, ++ struct file *file) ++{ ++ struct au_vdir *vdir_cache; ++ struct au_finfo *finfo; ++ struct au_fidir *fidir; ++ aufs_bindex_t bindex, bend; ++ ++ finfo = au_fi(file); ++ fidir = finfo->fi_hdir; ++ if (fidir) { ++ au_sphl_del(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ vdir_cache = fidir->fd_vdir_cache; /* lock-free */ ++ if (vdir_cache) ++ au_vdir_free(vdir_cache); ++ ++ bindex = finfo->fi_btop; ++ if (bindex >= 0) { ++ /* ++ * calls fput() instead of filp_close(), ++ * since no dnotify or lock for the lower file. ++ */ ++ bend = fidir->fd_bbot; ++ for (; bindex <= bend; bindex++) ++ au_set_h_fptr(file, bindex, NULL); ++ } ++ kfree(fidir); ++ finfo->fi_hdir = NULL; ++ } ++ au_finfo_fin(file); ++ return 0; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_do_flush_dir(struct file *file, fl_owner_t id) ++{ ++ int err; ++ aufs_bindex_t bindex, bend; ++ struct file *h_file; ++ ++ err = 0; ++ bend = au_fbend_dir(file); ++ for (bindex = au_fbstart(file); !err && bindex <= bend; bindex++) { ++ h_file = au_hf_dir(file, bindex); ++ if (h_file) ++ err = vfsub_flush(h_file, id); ++ } ++ return err; ++} ++ ++static int aufs_flush_dir(struct file *file, fl_owner_t id) ++{ ++ return au_do_flush(file, id, au_do_flush_dir); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_do_fsync_dir_no_file(struct dentry *dentry, int datasync) ++{ ++ int err; ++ aufs_bindex_t bend, bindex; ++ struct inode *inode; ++ struct super_block *sb; ++ ++ err = 0; ++ sb = dentry->d_sb; ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ bend = au_dbend(dentry); ++ for (bindex = au_dbstart(dentry); !err && bindex <= bend; bindex++) { ++ struct path h_path; ++ ++ if (au_test_ro(sb, bindex, inode)) ++ continue; ++ h_path.dentry = au_h_dptr(dentry, bindex); ++ if (!h_path.dentry) ++ continue; ++ ++ h_path.mnt = au_sbr_mnt(sb, bindex); ++ err = vfsub_fsync(NULL, &h_path, datasync); ++ } ++ ++ return err; ++} ++ ++static int au_do_fsync_dir(struct file *file, int datasync) ++{ ++ int err; ++ aufs_bindex_t bend, bindex; ++ struct file *h_file; ++ struct super_block *sb; ++ struct inode *inode; ++ ++ err = au_reval_and_lock_fdi(file, reopen_dir, /*wlock*/1); ++ if (unlikely(err)) ++ goto out; ++ ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ bend = au_fbend_dir(file); ++ for (bindex = au_fbstart(file); !err && bindex <= bend; bindex++) { ++ h_file = au_hf_dir(file, bindex); ++ if (!h_file || au_test_ro(sb, bindex, inode)) ++ continue; ++ ++ err = vfsub_fsync(h_file, &h_file->f_path, datasync); ++ } ++ ++out: ++ return err; ++} ++ ++/* ++ * @file may be NULL ++ */ ++static int aufs_fsync_dir(struct file *file, loff_t start, loff_t end, ++ int datasync) ++{ ++ int err; ++ struct dentry *dentry; ++ struct inode *inode; ++ struct super_block *sb; ++ struct mutex *mtx; ++ ++ err = 0; ++ dentry = file->f_path.dentry; ++ inode = d_inode(dentry); ++ mtx = &inode->i_mutex; ++ mutex_lock(mtx); ++ sb = dentry->d_sb; ++ si_noflush_read_lock(sb); ++ if (file) ++ err = au_do_fsync_dir(file, datasync); ++ else { ++ di_write_lock_child(dentry); ++ err = au_do_fsync_dir_no_file(dentry, datasync); ++ } ++ au_cpup_attr_timesizes(inode); ++ di_write_unlock(dentry); ++ if (file) ++ fi_write_unlock(file); ++ ++ si_read_unlock(sb); ++ mutex_unlock(mtx); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_iterate(struct file *file, struct dir_context *ctx) ++{ ++ int err; ++ struct dentry *dentry; ++ struct inode *inode, *h_inode; ++ struct super_block *sb; ++ ++ AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); ++ ++ dentry = file->f_path.dentry; ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ ++ sb = dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ err = au_reval_and_lock_fdi(file, reopen_dir, /*wlock*/1); ++ if (unlikely(err)) ++ goto out; ++ err = au_alive_dir(dentry); ++ if (!err) ++ err = au_vdir_init(file); ++ di_downgrade_lock(dentry, AuLock_IR); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ if (!au_test_nfsd()) { ++ err = au_vdir_fill_de(file, ctx); ++ fsstack_copy_attr_atime(inode, h_inode); ++ } else { ++ /* ++ * nfsd filldir may call lookup_one_len(), vfs_getattr(), ++ * encode_fh() and others. ++ */ ++ atomic_inc(&h_inode->i_count); ++ di_read_unlock(dentry, AuLock_IR); ++ si_read_unlock(sb); ++ err = au_vdir_fill_de(file, ctx); ++ fsstack_copy_attr_atime(inode, h_inode); ++ fi_write_unlock(file); ++ iput(h_inode); ++ ++ AuTraceErr(err); ++ return err; ++ } ++ ++out_unlock: ++ di_read_unlock(dentry, AuLock_IR); ++ fi_write_unlock(file); ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define AuTestEmpty_WHONLY 1 ++#define AuTestEmpty_CALLED (1 << 1) ++#define AuTestEmpty_SHWH (1 << 2) ++#define au_ftest_testempty(flags, name) ((flags) & AuTestEmpty_##name) ++#define au_fset_testempty(flags, name) \ ++ do { (flags) |= AuTestEmpty_##name; } while (0) ++#define au_fclr_testempty(flags, name) \ ++ do { (flags) &= ~AuTestEmpty_##name; } while (0) ++ ++#ifndef CONFIG_AUFS_SHWH ++#undef AuTestEmpty_SHWH ++#define AuTestEmpty_SHWH 0 ++#endif ++ ++struct test_empty_arg { ++ struct dir_context ctx; ++ struct au_nhash *whlist; ++ unsigned int flags; ++ int err; ++ aufs_bindex_t bindex; ++}; ++ ++static int test_empty_cb(struct dir_context *ctx, const char *__name, ++ int namelen, loff_t offset __maybe_unused, u64 ino, ++ unsigned int d_type) ++{ ++ struct test_empty_arg *arg = container_of(ctx, struct test_empty_arg, ++ ctx); ++ char *name = (void *)__name; ++ ++ arg->err = 0; ++ au_fset_testempty(arg->flags, CALLED); ++ /* smp_mb(); */ ++ if (name[0] == '.' ++ && (namelen == 1 || (name[1] == '.' && namelen == 2))) ++ goto out; /* success */ ++ ++ if (namelen <= AUFS_WH_PFX_LEN ++ || memcmp(name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { ++ if (au_ftest_testempty(arg->flags, WHONLY) ++ && !au_nhash_test_known_wh(arg->whlist, name, namelen)) ++ arg->err = -ENOTEMPTY; ++ goto out; ++ } ++ ++ name += AUFS_WH_PFX_LEN; ++ namelen -= AUFS_WH_PFX_LEN; ++ if (!au_nhash_test_known_wh(arg->whlist, name, namelen)) ++ arg->err = au_nhash_append_wh ++ (arg->whlist, name, namelen, ino, d_type, arg->bindex, ++ au_ftest_testempty(arg->flags, SHWH)); ++ ++out: ++ /* smp_mb(); */ ++ AuTraceErr(arg->err); ++ return arg->err; ++} ++ ++static int do_test_empty(struct dentry *dentry, struct test_empty_arg *arg) ++{ ++ int err; ++ struct file *h_file; ++ ++ h_file = au_h_open(dentry, arg->bindex, ++ O_RDONLY | O_NONBLOCK | O_DIRECTORY | O_LARGEFILE, ++ /*file*/NULL, /*force_wr*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = 0; ++ if (!au_opt_test(au_mntflags(dentry->d_sb), UDBA_NONE) ++ && !file_inode(h_file)->i_nlink) ++ goto out_put; ++ ++ do { ++ arg->err = 0; ++ au_fclr_testempty(arg->flags, CALLED); ++ /* smp_mb(); */ ++ err = vfsub_iterate_dir(h_file, &arg->ctx); ++ if (err >= 0) ++ err = arg->err; ++ } while (!err && au_ftest_testempty(arg->flags, CALLED)); ++ ++out_put: ++ fput(h_file); ++ au_sbr_put(dentry->d_sb, arg->bindex); ++out: ++ return err; ++} ++ ++struct do_test_empty_args { ++ int *errp; ++ struct dentry *dentry; ++ struct test_empty_arg *arg; ++}; ++ ++static void call_do_test_empty(void *args) ++{ ++ struct do_test_empty_args *a = args; ++ *a->errp = do_test_empty(a->dentry, a->arg); ++} ++ ++static int sio_test_empty(struct dentry *dentry, struct test_empty_arg *arg) ++{ ++ int err, wkq_err; ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ ++ h_dentry = au_h_dptr(dentry, arg->bindex); ++ h_inode = d_inode(h_dentry); ++ /* todo: i_mode changes anytime? */ ++ mutex_lock_nested(&h_inode->i_mutex, AuLsc_I_CHILD); ++ err = au_test_h_perm_sio(h_inode, MAY_EXEC | MAY_READ); ++ mutex_unlock(&h_inode->i_mutex); ++ if (!err) ++ err = do_test_empty(dentry, arg); ++ else { ++ struct do_test_empty_args args = { ++ .errp = &err, ++ .dentry = dentry, ++ .arg = arg ++ }; ++ unsigned int flags = arg->flags; ++ ++ wkq_err = au_wkq_wait(call_do_test_empty, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ arg->flags = flags; ++ } ++ ++ return err; ++} ++ ++int au_test_empty_lower(struct dentry *dentry) ++{ ++ int err; ++ unsigned int rdhash; ++ aufs_bindex_t bindex, bstart, btail; ++ struct au_nhash whlist; ++ struct test_empty_arg arg = { ++ .ctx = { ++ .actor = test_empty_cb ++ } ++ }; ++ int (*test_empty)(struct dentry *dentry, struct test_empty_arg *arg); ++ ++ SiMustAnyLock(dentry->d_sb); ++ ++ rdhash = au_sbi(dentry->d_sb)->si_rdhash; ++ if (!rdhash) ++ rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, dentry)); ++ err = au_nhash_alloc(&whlist, rdhash, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ ++ arg.flags = 0; ++ arg.whlist = &whlist; ++ bstart = au_dbstart(dentry); ++ if (au_opt_test(au_mntflags(dentry->d_sb), SHWH)) ++ au_fset_testempty(arg.flags, SHWH); ++ test_empty = do_test_empty; ++ if (au_opt_test(au_mntflags(dentry->d_sb), DIRPERM1)) ++ test_empty = sio_test_empty; ++ arg.bindex = bstart; ++ err = test_empty(dentry, &arg); ++ if (unlikely(err)) ++ goto out_whlist; ++ ++ au_fset_testempty(arg.flags, WHONLY); ++ btail = au_dbtaildir(dentry); ++ for (bindex = bstart + 1; !err && bindex <= btail; bindex++) { ++ struct dentry *h_dentry; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry && d_is_positive(h_dentry)) { ++ arg.bindex = bindex; ++ err = test_empty(dentry, &arg); ++ } ++ } ++ ++out_whlist: ++ au_nhash_wh_free(&whlist); ++out: ++ return err; ++} ++ ++int au_test_empty(struct dentry *dentry, struct au_nhash *whlist) ++{ ++ int err; ++ struct test_empty_arg arg = { ++ .ctx = { ++ .actor = test_empty_cb ++ } ++ }; ++ aufs_bindex_t bindex, btail; ++ ++ err = 0; ++ arg.whlist = whlist; ++ arg.flags = AuTestEmpty_WHONLY; ++ if (au_opt_test(au_mntflags(dentry->d_sb), SHWH)) ++ au_fset_testempty(arg.flags, SHWH); ++ btail = au_dbtaildir(dentry); ++ for (bindex = au_dbstart(dentry); !err && bindex <= btail; bindex++) { ++ struct dentry *h_dentry; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry && d_is_positive(h_dentry)) { ++ arg.bindex = bindex; ++ err = sio_test_empty(dentry, &arg); ++ } ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++const struct file_operations aufs_dir_fop = { ++ .owner = THIS_MODULE, ++ .llseek = default_llseek, ++ .read = generic_read_dir, ++ .iterate = aufs_iterate, ++ .unlocked_ioctl = aufs_ioctl_dir, ++#ifdef CONFIG_COMPAT ++ .compat_ioctl = aufs_compat_ioctl_dir, ++#endif ++ .open = aufs_open_dir, ++ .release = aufs_release_dir, ++ .flush = aufs_flush_dir, ++ .fsync = aufs_fsync_dir ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/dir.h linux-4.1.10/fs/aufs/dir.h +--- linux-4.1.10.orig/fs/aufs/dir.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dir.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,131 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * directory operations ++ */ ++ ++#ifndef __AUFS_DIR_H__ ++#define __AUFS_DIR_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fs.h> ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* need to be faster and smaller */ ++ ++struct au_nhash { ++ unsigned int nh_num; ++ struct hlist_head *nh_head; ++}; ++ ++struct au_vdir_destr { ++ unsigned char len; ++ unsigned char name[0]; ++} __packed; ++ ++struct au_vdir_dehstr { ++ struct hlist_node hash; ++ struct au_vdir_destr *str; ++} ____cacheline_aligned_in_smp; ++ ++struct au_vdir_de { ++ ino_t de_ino; ++ unsigned char de_type; ++ /* caution: packed */ ++ struct au_vdir_destr de_str; ++} __packed; ++ ++struct au_vdir_wh { ++ struct hlist_node wh_hash; ++#ifdef CONFIG_AUFS_SHWH ++ ino_t wh_ino; ++ aufs_bindex_t wh_bindex; ++ unsigned char wh_type; ++#else ++ aufs_bindex_t wh_bindex; ++#endif ++ /* caution: packed */ ++ struct au_vdir_destr wh_str; ++} __packed; ++ ++union au_vdir_deblk_p { ++ unsigned char *deblk; ++ struct au_vdir_de *de; ++}; ++ ++struct au_vdir { ++ unsigned char **vd_deblk; ++ unsigned long vd_nblk; ++ struct { ++ unsigned long ul; ++ union au_vdir_deblk_p p; ++ } vd_last; ++ ++ unsigned long vd_version; ++ unsigned int vd_deblk_sz; ++ unsigned long vd_jiffy; ++} ____cacheline_aligned_in_smp; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* dir.c */ ++extern const struct file_operations aufs_dir_fop; ++void au_add_nlink(struct inode *dir, struct inode *h_dir); ++void au_sub_nlink(struct inode *dir, struct inode *h_dir); ++loff_t au_dir_size(struct file *file, struct dentry *dentry); ++void au_dir_ts(struct inode *dir, aufs_bindex_t bsrc); ++int au_test_empty_lower(struct dentry *dentry); ++int au_test_empty(struct dentry *dentry, struct au_nhash *whlist); ++ ++/* vdir.c */ ++unsigned int au_rdhash_est(loff_t sz); ++int au_nhash_alloc(struct au_nhash *nhash, unsigned int num_hash, gfp_t gfp); ++void au_nhash_wh_free(struct au_nhash *whlist); ++int au_nhash_test_longer_wh(struct au_nhash *whlist, aufs_bindex_t btgt, ++ int limit); ++int au_nhash_test_known_wh(struct au_nhash *whlist, char *name, int nlen); ++int au_nhash_append_wh(struct au_nhash *whlist, char *name, int nlen, ino_t ino, ++ unsigned int d_type, aufs_bindex_t bindex, ++ unsigned char shwh); ++void au_vdir_free(struct au_vdir *vdir); ++int au_vdir_init(struct file *file); ++int au_vdir_fill_de(struct file *file, struct dir_context *ctx); ++ ++/* ioctl.c */ ++long aufs_ioctl_dir(struct file *file, unsigned int cmd, unsigned long arg); ++ ++#ifdef CONFIG_AUFS_RDU ++/* rdu.c */ ++long au_rdu_ioctl(struct file *file, unsigned int cmd, unsigned long arg); ++#ifdef CONFIG_COMPAT ++long au_rdu_compat_ioctl(struct file *file, unsigned int cmd, ++ unsigned long arg); ++#endif ++#else ++AuStub(long, au_rdu_ioctl, return -EINVAL, struct file *file, ++ unsigned int cmd, unsigned long arg) ++#ifdef CONFIG_COMPAT ++AuStub(long, au_rdu_compat_ioctl, return -EINVAL, struct file *file, ++ unsigned int cmd, unsigned long arg) ++#endif ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DIR_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/dynop.c linux-4.1.10/fs/aufs/dynop.c +--- linux-4.1.10.orig/fs/aufs/dynop.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dynop.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,369 @@ ++/* ++ * Copyright (C) 2010-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * dynamically customizable operations for regular files ++ */ ++ ++#include "aufs.h" ++ ++#define DyPrSym(key) AuDbgSym(key->dk_op.dy_hop) ++ ++/* ++ * How large will these lists be? ++ * Usually just a few elements, 20-30 at most for each, I guess. ++ */ ++static struct au_splhead dynop[AuDyLast]; ++ ++static struct au_dykey *dy_gfind_get(struct au_splhead *spl, const void *h_op) ++{ ++ struct au_dykey *key, *tmp; ++ struct list_head *head; ++ ++ key = NULL; ++ head = &spl->head; ++ rcu_read_lock(); ++ list_for_each_entry_rcu(tmp, head, dk_list) ++ if (tmp->dk_op.dy_hop == h_op) { ++ key = tmp; ++ kref_get(&key->dk_kref); ++ break; ++ } ++ rcu_read_unlock(); ++ ++ return key; ++} ++ ++static struct au_dykey *dy_bradd(struct au_branch *br, struct au_dykey *key) ++{ ++ struct au_dykey **k, *found; ++ const void *h_op = key->dk_op.dy_hop; ++ int i; ++ ++ found = NULL; ++ k = br->br_dykey; ++ for (i = 0; i < AuBrDynOp; i++) ++ if (k[i]) { ++ if (k[i]->dk_op.dy_hop == h_op) { ++ found = k[i]; ++ break; ++ } ++ } else ++ break; ++ if (!found) { ++ spin_lock(&br->br_dykey_lock); ++ for (; i < AuBrDynOp; i++) ++ if (k[i]) { ++ if (k[i]->dk_op.dy_hop == h_op) { ++ found = k[i]; ++ break; ++ } ++ } else { ++ k[i] = key; ++ break; ++ } ++ spin_unlock(&br->br_dykey_lock); ++ BUG_ON(i == AuBrDynOp); /* expand the array */ ++ } ++ ++ return found; ++} ++ ++/* kref_get() if @key is already added */ ++static struct au_dykey *dy_gadd(struct au_splhead *spl, struct au_dykey *key) ++{ ++ struct au_dykey *tmp, *found; ++ struct list_head *head; ++ const void *h_op = key->dk_op.dy_hop; ++ ++ found = NULL; ++ head = &spl->head; ++ spin_lock(&spl->spin); ++ list_for_each_entry(tmp, head, dk_list) ++ if (tmp->dk_op.dy_hop == h_op) { ++ kref_get(&tmp->dk_kref); ++ found = tmp; ++ break; ++ } ++ if (!found) ++ list_add_rcu(&key->dk_list, head); ++ spin_unlock(&spl->spin); ++ ++ if (!found) ++ DyPrSym(key); ++ return found; ++} ++ ++static void dy_free_rcu(struct rcu_head *rcu) ++{ ++ struct au_dykey *key; ++ ++ key = container_of(rcu, struct au_dykey, dk_rcu); ++ DyPrSym(key); ++ kfree(key); ++} ++ ++static void dy_free(struct kref *kref) ++{ ++ struct au_dykey *key; ++ struct au_splhead *spl; ++ ++ key = container_of(kref, struct au_dykey, dk_kref); ++ spl = dynop + key->dk_op.dy_type; ++ au_spl_del_rcu(&key->dk_list, spl); ++ call_rcu(&key->dk_rcu, dy_free_rcu); ++} ++ ++void au_dy_put(struct au_dykey *key) ++{ ++ kref_put(&key->dk_kref, dy_free); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define DyDbgSize(cnt, op) AuDebugOn(cnt != sizeof(op)/sizeof(void *)) ++ ++#ifdef CONFIG_AUFS_DEBUG ++#define DyDbgDeclare(cnt) unsigned int cnt = 0 ++#define DyDbgInc(cnt) do { cnt++; } while (0) ++#else ++#define DyDbgDeclare(cnt) do {} while (0) ++#define DyDbgInc(cnt) do {} while (0) ++#endif ++ ++#define DySet(func, dst, src, h_op, h_sb) do { \ ++ DyDbgInc(cnt); \ ++ if (h_op->func) { \ ++ if (src.func) \ ++ dst.func = src.func; \ ++ else \ ++ AuDbg("%s %s\n", au_sbtype(h_sb), #func); \ ++ } \ ++} while (0) ++ ++#define DySetForce(func, dst, src) do { \ ++ AuDebugOn(!src.func); \ ++ DyDbgInc(cnt); \ ++ dst.func = src.func; \ ++} while (0) ++ ++#define DySetAop(func) \ ++ DySet(func, dyaop->da_op, aufs_aop, h_aop, h_sb) ++#define DySetAopForce(func) \ ++ DySetForce(func, dyaop->da_op, aufs_aop) ++ ++static void dy_aop(struct au_dykey *key, const void *h_op, ++ struct super_block *h_sb __maybe_unused) ++{ ++ struct au_dyaop *dyaop = (void *)key; ++ const struct address_space_operations *h_aop = h_op; ++ DyDbgDeclare(cnt); ++ ++ AuDbg("%s\n", au_sbtype(h_sb)); ++ ++ DySetAop(writepage); ++ DySetAopForce(readpage); /* force */ ++ DySetAop(writepages); ++ DySetAop(set_page_dirty); ++ DySetAop(readpages); ++ DySetAop(write_begin); ++ DySetAop(write_end); ++ DySetAop(bmap); ++ DySetAop(invalidatepage); ++ DySetAop(releasepage); ++ DySetAop(freepage); ++ /* this one will be changed according to an aufs mount option */ ++ DySetAop(direct_IO); ++ DySetAop(migratepage); ++ DySetAop(launder_page); ++ DySetAop(is_partially_uptodate); ++ DySetAop(is_dirty_writeback); ++ DySetAop(error_remove_page); ++ DySetAop(swap_activate); ++ DySetAop(swap_deactivate); ++ ++ DyDbgSize(cnt, *h_aop); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void dy_bug(struct kref *kref) ++{ ++ BUG(); ++} ++ ++static struct au_dykey *dy_get(struct au_dynop *op, struct au_branch *br) ++{ ++ struct au_dykey *key, *old; ++ struct au_splhead *spl; ++ struct op { ++ unsigned int sz; ++ void (*set)(struct au_dykey *key, const void *h_op, ++ struct super_block *h_sb __maybe_unused); ++ }; ++ static const struct op a[] = { ++ [AuDy_AOP] = { ++ .sz = sizeof(struct au_dyaop), ++ .set = dy_aop ++ } ++ }; ++ const struct op *p; ++ ++ spl = dynop + op->dy_type; ++ key = dy_gfind_get(spl, op->dy_hop); ++ if (key) ++ goto out_add; /* success */ ++ ++ p = a + op->dy_type; ++ key = kzalloc(p->sz, GFP_NOFS); ++ if (unlikely(!key)) { ++ key = ERR_PTR(-ENOMEM); ++ goto out; ++ } ++ ++ key->dk_op.dy_hop = op->dy_hop; ++ kref_init(&key->dk_kref); ++ p->set(key, op->dy_hop, au_br_sb(br)); ++ old = dy_gadd(spl, key); ++ if (old) { ++ kfree(key); ++ key = old; ++ } ++ ++out_add: ++ old = dy_bradd(br, key); ++ if (old) ++ /* its ref-count should never be zero here */ ++ kref_put(&key->dk_kref, dy_bug); ++out: ++ return key; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * Aufs prohibits O_DIRECT by defaut even if the branch supports it. ++ * This behaviour is necessary to return an error from open(O_DIRECT) instead ++ * of the succeeding I/O. The dio mount option enables O_DIRECT and makes ++ * open(O_DIRECT) always succeed, but the succeeding I/O may return an error. ++ * See the aufs manual in detail. ++ */ ++static void dy_adx(struct au_dyaop *dyaop, int do_dx) ++{ ++ if (!do_dx) ++ dyaop->da_op.direct_IO = NULL; ++ else ++ dyaop->da_op.direct_IO = aufs_aop.direct_IO; ++} ++ ++static struct au_dyaop *dy_aget(struct au_branch *br, ++ const struct address_space_operations *h_aop, ++ int do_dx) ++{ ++ struct au_dyaop *dyaop; ++ struct au_dynop op; ++ ++ op.dy_type = AuDy_AOP; ++ op.dy_haop = h_aop; ++ dyaop = (void *)dy_get(&op, br); ++ if (IS_ERR(dyaop)) ++ goto out; ++ dy_adx(dyaop, do_dx); ++ ++out: ++ return dyaop; ++} ++ ++int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, ++ struct inode *h_inode) ++{ ++ int err, do_dx; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_dyaop *dyaop; ++ ++ AuDebugOn(!S_ISREG(h_inode->i_mode)); ++ IiMustWriteLock(inode); ++ ++ sb = inode->i_sb; ++ br = au_sbr(sb, bindex); ++ do_dx = !!au_opt_test(au_mntflags(sb), DIO); ++ dyaop = dy_aget(br, h_inode->i_mapping->a_ops, do_dx); ++ err = PTR_ERR(dyaop); ++ if (IS_ERR(dyaop)) ++ /* unnecessary to call dy_fput() */ ++ goto out; ++ ++ err = 0; ++ inode->i_mapping->a_ops = &dyaop->da_op; ++ ++out: ++ return err; ++} ++ ++/* ++ * Is it safe to replace a_ops during the inode/file is in operation? ++ * Yes, I hope so. ++ */ ++int au_dy_irefresh(struct inode *inode) ++{ ++ int err; ++ aufs_bindex_t bstart; ++ struct inode *h_inode; ++ ++ err = 0; ++ if (S_ISREG(inode->i_mode)) { ++ bstart = au_ibstart(inode); ++ h_inode = au_h_iptr(inode, bstart); ++ err = au_dy_iaop(inode, bstart, h_inode); ++ } ++ return err; ++} ++ ++void au_dy_arefresh(int do_dx) ++{ ++ struct au_splhead *spl; ++ struct list_head *head; ++ struct au_dykey *key; ++ ++ spl = dynop + AuDy_AOP; ++ head = &spl->head; ++ spin_lock(&spl->spin); ++ list_for_each_entry(key, head, dk_list) ++ dy_adx((void *)key, do_dx); ++ spin_unlock(&spl->spin); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void __init au_dy_init(void) ++{ ++ int i; ++ ++ /* make sure that 'struct au_dykey *' can be any type */ ++ BUILD_BUG_ON(offsetof(struct au_dyaop, da_key)); ++ ++ for (i = 0; i < AuDyLast; i++) ++ au_spl_init(dynop + i); ++} ++ ++void au_dy_fin(void) ++{ ++ int i; ++ ++ for (i = 0; i < AuDyLast; i++) ++ WARN_ON(!list_empty(&dynop[i].head)); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/dynop.h linux-4.1.10/fs/aufs/dynop.h +--- linux-4.1.10.orig/fs/aufs/dynop.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/dynop.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,74 @@ ++/* ++ * Copyright (C) 2010-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * dynamically customizable operations (for regular files only) ++ */ ++ ++#ifndef __AUFS_DYNOP_H__ ++#define __AUFS_DYNOP_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fs.h> ++#include <linux/kref.h> ++ ++enum {AuDy_AOP, AuDyLast}; ++ ++struct au_dynop { ++ int dy_type; ++ union { ++ const void *dy_hop; ++ const struct address_space_operations *dy_haop; ++ }; ++}; ++ ++struct au_dykey { ++ union { ++ struct list_head dk_list; ++ struct rcu_head dk_rcu; ++ }; ++ struct au_dynop dk_op; ++ ++ /* ++ * during I am in the branch local array, kref is gotten. when the ++ * branch is removed, kref is put. ++ */ ++ struct kref dk_kref; ++}; ++ ++/* stop unioning since their sizes are very different from each other */ ++struct au_dyaop { ++ struct au_dykey da_key; ++ struct address_space_operations da_op; /* not const */ ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* dynop.c */ ++struct au_branch; ++void au_dy_put(struct au_dykey *key); ++int au_dy_iaop(struct inode *inode, aufs_bindex_t bindex, ++ struct inode *h_inode); ++int au_dy_irefresh(struct inode *inode); ++void au_dy_arefresh(int do_dio); ++ ++void __init au_dy_init(void); ++void au_dy_fin(void); ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_DYNOP_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/export.c linux-4.1.10/fs/aufs/export.c +--- linux-4.1.10.orig/fs/aufs/export.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/export.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,832 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * export via nfs ++ */ ++ ++#include <linux/exportfs.h> ++#include <linux/fs_struct.h> ++#include <linux/namei.h> ++#include <linux/nsproxy.h> ++#include <linux/random.h> ++#include <linux/writeback.h> ++#include "../fs/mount.h" ++#include "aufs.h" ++ ++union conv { ++#ifdef CONFIG_AUFS_INO_T_64 ++ __u32 a[2]; ++#else ++ __u32 a[1]; ++#endif ++ ino_t ino; ++}; ++ ++static ino_t decode_ino(__u32 *a) ++{ ++ union conv u; ++ ++ BUILD_BUG_ON(sizeof(u.ino) != sizeof(u.a)); ++ u.a[0] = a[0]; ++#ifdef CONFIG_AUFS_INO_T_64 ++ u.a[1] = a[1]; ++#endif ++ return u.ino; ++} ++ ++static void encode_ino(__u32 *a, ino_t ino) ++{ ++ union conv u; ++ ++ u.ino = ino; ++ a[0] = u.a[0]; ++#ifdef CONFIG_AUFS_INO_T_64 ++ a[1] = u.a[1]; ++#endif ++} ++ ++/* NFS file handle */ ++enum { ++ Fh_br_id, ++ Fh_sigen, ++#ifdef CONFIG_AUFS_INO_T_64 ++ /* support 64bit inode number */ ++ Fh_ino1, ++ Fh_ino2, ++ Fh_dir_ino1, ++ Fh_dir_ino2, ++#else ++ Fh_ino1, ++ Fh_dir_ino1, ++#endif ++ Fh_igen, ++ Fh_h_type, ++ Fh_tail, ++ ++ Fh_ino = Fh_ino1, ++ Fh_dir_ino = Fh_dir_ino1 ++}; ++ ++static int au_test_anon(struct dentry *dentry) ++{ ++ /* note: read d_flags without d_lock */ ++ return !!(dentry->d_flags & DCACHE_DISCONNECTED); ++} ++ ++int au_test_nfsd(void) ++{ ++ int ret; ++ struct task_struct *tsk = current; ++ char comm[sizeof(tsk->comm)]; ++ ++ ret = 0; ++ if (tsk->flags & PF_KTHREAD) { ++ get_task_comm(comm, tsk); ++ ret = !strcmp(comm, "nfsd"); ++ } ++ ++ return ret; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* inode generation external table */ ++ ++void au_xigen_inc(struct inode *inode) ++{ ++ loff_t pos; ++ ssize_t sz; ++ __u32 igen; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ ++ sb = inode->i_sb; ++ AuDebugOn(!au_opt_test(au_mntflags(sb), XINO)); ++ ++ sbinfo = au_sbi(sb); ++ pos = inode->i_ino; ++ pos *= sizeof(igen); ++ igen = inode->i_generation + 1; ++ sz = xino_fwrite(sbinfo->si_xwrite, sbinfo->si_xigen, &igen, ++ sizeof(igen), &pos); ++ if (sz == sizeof(igen)) ++ return; /* success */ ++ ++ if (unlikely(sz >= 0)) ++ AuIOErr("xigen error (%zd)\n", sz); ++} ++ ++int au_xigen_new(struct inode *inode) ++{ ++ int err; ++ loff_t pos; ++ ssize_t sz; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ ++ err = 0; ++ /* todo: dirty, at mount time */ ++ if (inode->i_ino == AUFS_ROOT_INO) ++ goto out; ++ sb = inode->i_sb; ++ SiMustAnyLock(sb); ++ if (unlikely(!au_opt_test(au_mntflags(sb), XINO))) ++ goto out; ++ ++ err = -EFBIG; ++ pos = inode->i_ino; ++ if (unlikely(au_loff_max / sizeof(inode->i_generation) - 1 < pos)) { ++ AuIOErr1("too large i%lld\n", pos); ++ goto out; ++ } ++ pos *= sizeof(inode->i_generation); ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ file = sbinfo->si_xigen; ++ BUG_ON(!file); ++ ++ if (vfsub_f_size_read(file) ++ < pos + sizeof(inode->i_generation)) { ++ inode->i_generation = atomic_inc_return(&sbinfo->si_xigen_next); ++ sz = xino_fwrite(sbinfo->si_xwrite, file, &inode->i_generation, ++ sizeof(inode->i_generation), &pos); ++ } else ++ sz = xino_fread(sbinfo->si_xread, file, &inode->i_generation, ++ sizeof(inode->i_generation), &pos); ++ if (sz == sizeof(inode->i_generation)) ++ goto out; /* success */ ++ ++ err = sz; ++ if (unlikely(sz >= 0)) { ++ err = -EIO; ++ AuIOErr("xigen error (%zd)\n", sz); ++ } ++ ++out: ++ return err; ++} ++ ++int au_xigen_set(struct super_block *sb, struct file *base) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ file = au_xino_create2(base, sbinfo->si_xigen); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ err = 0; ++ if (sbinfo->si_xigen) ++ fput(sbinfo->si_xigen); ++ sbinfo->si_xigen = file; ++ ++out: ++ return err; ++} ++ ++void au_xigen_clr(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ if (sbinfo->si_xigen) { ++ fput(sbinfo->si_xigen); ++ sbinfo->si_xigen = NULL; ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct dentry *decode_by_ino(struct super_block *sb, ino_t ino, ++ ino_t dir_ino) ++{ ++ struct dentry *dentry, *d; ++ struct inode *inode; ++ unsigned int sigen; ++ ++ dentry = NULL; ++ inode = ilookup(sb, ino); ++ if (!inode) ++ goto out; ++ ++ dentry = ERR_PTR(-ESTALE); ++ sigen = au_sigen(sb); ++ if (unlikely(is_bad_inode(inode) ++ || IS_DEADDIR(inode) ++ || sigen != au_iigen(inode, NULL))) ++ goto out_iput; ++ ++ dentry = NULL; ++ if (!dir_ino || S_ISDIR(inode->i_mode)) ++ dentry = d_find_alias(inode); ++ else { ++ spin_lock(&inode->i_lock); ++ hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) { ++ spin_lock(&d->d_lock); ++ if (!au_test_anon(d) ++ && d_inode(d->d_parent)->i_ino == dir_ino) { ++ dentry = dget_dlock(d); ++ spin_unlock(&d->d_lock); ++ break; ++ } ++ spin_unlock(&d->d_lock); ++ } ++ spin_unlock(&inode->i_lock); ++ } ++ if (unlikely(dentry && au_digen_test(dentry, sigen))) { ++ /* need to refresh */ ++ dput(dentry); ++ dentry = NULL; ++ } ++ ++out_iput: ++ iput(inode); ++out: ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* todo: dirty? */ ++/* if exportfs_decode_fh() passed vfsmount*, we could be happy */ ++ ++struct au_compare_mnt_args { ++ /* input */ ++ struct super_block *sb; ++ ++ /* output */ ++ struct vfsmount *mnt; ++}; ++ ++static int au_compare_mnt(struct vfsmount *mnt, void *arg) ++{ ++ struct au_compare_mnt_args *a = arg; ++ ++ if (mnt->mnt_sb != a->sb) ++ return 0; ++ a->mnt = mntget(mnt); ++ return 1; ++} ++ ++static struct vfsmount *au_mnt_get(struct super_block *sb) ++{ ++ int err; ++ struct path root; ++ struct au_compare_mnt_args args = { ++ .sb = sb ++ }; ++ ++ get_fs_root(current->fs, &root); ++ rcu_read_lock(); ++ err = iterate_mounts(au_compare_mnt, &args, root.mnt); ++ rcu_read_unlock(); ++ path_put(&root); ++ AuDebugOn(!err); ++ AuDebugOn(!args.mnt); ++ return args.mnt; ++} ++ ++struct au_nfsd_si_lock { ++ unsigned int sigen; ++ aufs_bindex_t bindex, br_id; ++ unsigned char force_lock; ++}; ++ ++static int si_nfsd_read_lock(struct super_block *sb, ++ struct au_nfsd_si_lock *nsi_lock) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ ++ si_read_lock(sb, AuLock_FLUSH); ++ ++ /* branch id may be wrapped around */ ++ err = 0; ++ bindex = au_br_index(sb, nsi_lock->br_id); ++ if (bindex >= 0 && nsi_lock->sigen + AUFS_BRANCH_MAX > au_sigen(sb)) ++ goto out; /* success */ ++ ++ err = -ESTALE; ++ bindex = -1; ++ if (!nsi_lock->force_lock) ++ si_read_unlock(sb); ++ ++out: ++ nsi_lock->bindex = bindex; ++ return err; ++} ++ ++struct find_name_by_ino { ++ struct dir_context ctx; ++ int called, found; ++ ino_t ino; ++ char *name; ++ int namelen; ++}; ++ ++static int ++find_name_by_ino(struct dir_context *ctx, const char *name, int namelen, ++ loff_t offset, u64 ino, unsigned int d_type) ++{ ++ struct find_name_by_ino *a = container_of(ctx, struct find_name_by_ino, ++ ctx); ++ ++ a->called++; ++ if (a->ino != ino) ++ return 0; ++ ++ memcpy(a->name, name, namelen); ++ a->namelen = namelen; ++ a->found = 1; ++ return 1; ++} ++ ++static struct dentry *au_lkup_by_ino(struct path *path, ino_t ino, ++ struct au_nfsd_si_lock *nsi_lock) ++{ ++ struct dentry *dentry, *parent; ++ struct file *file; ++ struct inode *dir; ++ struct find_name_by_ino arg = { ++ .ctx = { ++ .actor = find_name_by_ino ++ } ++ }; ++ int err; ++ ++ parent = path->dentry; ++ if (nsi_lock) ++ si_read_unlock(parent->d_sb); ++ file = vfsub_dentry_open(path, au_dir_roflags); ++ dentry = (void *)file; ++ if (IS_ERR(file)) ++ goto out; ++ ++ dentry = ERR_PTR(-ENOMEM); ++ arg.name = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!arg.name)) ++ goto out_file; ++ arg.ino = ino; ++ arg.found = 0; ++ do { ++ arg.called = 0; ++ /* smp_mb(); */ ++ err = vfsub_iterate_dir(file, &arg.ctx); ++ } while (!err && !arg.found && arg.called); ++ dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out_name; ++ /* instead of ENOENT */ ++ dentry = ERR_PTR(-ESTALE); ++ if (!arg.found) ++ goto out_name; ++ ++ /* do not call vfsub_lkup_one() */ ++ dir = d_inode(parent); ++ mutex_lock(&dir->i_mutex); ++ dentry = vfsub_lookup_one_len(arg.name, parent, arg.namelen); ++ mutex_unlock(&dir->i_mutex); ++ AuTraceErrPtr(dentry); ++ if (IS_ERR(dentry)) ++ goto out_name; ++ AuDebugOn(au_test_anon(dentry)); ++ if (unlikely(d_really_is_negative(dentry))) { ++ dput(dentry); ++ dentry = ERR_PTR(-ENOENT); ++ } ++ ++out_name: ++ free_page((unsigned long)arg.name); ++out_file: ++ fput(file); ++out: ++ if (unlikely(nsi_lock ++ && si_nfsd_read_lock(parent->d_sb, nsi_lock) < 0)) ++ if (!IS_ERR(dentry)) { ++ dput(dentry); ++ dentry = ERR_PTR(-ESTALE); ++ } ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++static struct dentry *decode_by_dir_ino(struct super_block *sb, ino_t ino, ++ ino_t dir_ino, ++ struct au_nfsd_si_lock *nsi_lock) ++{ ++ struct dentry *dentry; ++ struct path path; ++ ++ if (dir_ino != AUFS_ROOT_INO) { ++ path.dentry = decode_by_ino(sb, dir_ino, 0); ++ dentry = path.dentry; ++ if (!path.dentry || IS_ERR(path.dentry)) ++ goto out; ++ AuDebugOn(au_test_anon(path.dentry)); ++ } else ++ path.dentry = dget(sb->s_root); ++ ++ path.mnt = au_mnt_get(sb); ++ dentry = au_lkup_by_ino(&path, ino, nsi_lock); ++ path_put(&path); ++ ++out: ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int h_acceptable(void *expv, struct dentry *dentry) ++{ ++ return 1; ++} ++ ++static char *au_build_path(struct dentry *h_parent, struct path *h_rootpath, ++ char *buf, int len, struct super_block *sb) ++{ ++ char *p; ++ int n; ++ struct path path; ++ ++ p = d_path(h_rootpath, buf, len); ++ if (IS_ERR(p)) ++ goto out; ++ n = strlen(p); ++ ++ path.mnt = h_rootpath->mnt; ++ path.dentry = h_parent; ++ p = d_path(&path, buf, len); ++ if (IS_ERR(p)) ++ goto out; ++ if (n != 1) ++ p += n; ++ ++ path.mnt = au_mnt_get(sb); ++ path.dentry = sb->s_root; ++ p = d_path(&path, buf, len - strlen(p)); ++ mntput(path.mnt); ++ if (IS_ERR(p)) ++ goto out; ++ if (n != 1) ++ p[strlen(p)] = '/'; ++ ++out: ++ AuTraceErrPtr(p); ++ return p; ++} ++ ++static ++struct dentry *decode_by_path(struct super_block *sb, ino_t ino, __u32 *fh, ++ int fh_len, struct au_nfsd_si_lock *nsi_lock) ++{ ++ struct dentry *dentry, *h_parent, *root; ++ struct super_block *h_sb; ++ char *pathname, *p; ++ struct vfsmount *h_mnt; ++ struct au_branch *br; ++ int err; ++ struct path path; ++ ++ br = au_sbr(sb, nsi_lock->bindex); ++ h_mnt = au_br_mnt(br); ++ h_sb = h_mnt->mnt_sb; ++ /* todo: call lower fh_to_dentry()? fh_to_parent()? */ ++ h_parent = exportfs_decode_fh(h_mnt, (void *)(fh + Fh_tail), ++ fh_len - Fh_tail, fh[Fh_h_type], ++ h_acceptable, /*context*/NULL); ++ dentry = h_parent; ++ if (unlikely(!h_parent || IS_ERR(h_parent))) { ++ AuWarn1("%s decode_fh failed, %ld\n", ++ au_sbtype(h_sb), PTR_ERR(h_parent)); ++ goto out; ++ } ++ dentry = NULL; ++ if (unlikely(au_test_anon(h_parent))) { ++ AuWarn1("%s decode_fh returned a disconnected dentry\n", ++ au_sbtype(h_sb)); ++ goto out_h_parent; ++ } ++ ++ dentry = ERR_PTR(-ENOMEM); ++ pathname = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!pathname)) ++ goto out_h_parent; ++ ++ root = sb->s_root; ++ path.mnt = h_mnt; ++ di_read_lock_parent(root, !AuLock_IR); ++ path.dentry = au_h_dptr(root, nsi_lock->bindex); ++ di_read_unlock(root, !AuLock_IR); ++ p = au_build_path(h_parent, &path, pathname, PAGE_SIZE, sb); ++ dentry = (void *)p; ++ if (IS_ERR(p)) ++ goto out_pathname; ++ ++ si_read_unlock(sb); ++ err = vfsub_kern_path(p, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path); ++ dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out_relock; ++ ++ dentry = ERR_PTR(-ENOENT); ++ AuDebugOn(au_test_anon(path.dentry)); ++ if (unlikely(d_really_is_negative(path.dentry))) ++ goto out_path; ++ ++ if (ino != d_inode(path.dentry)->i_ino) ++ dentry = au_lkup_by_ino(&path, ino, /*nsi_lock*/NULL); ++ else ++ dentry = dget(path.dentry); ++ ++out_path: ++ path_put(&path); ++out_relock: ++ if (unlikely(si_nfsd_read_lock(sb, nsi_lock) < 0)) ++ if (!IS_ERR(dentry)) { ++ dput(dentry); ++ dentry = ERR_PTR(-ESTALE); ++ } ++out_pathname: ++ free_page((unsigned long)pathname); ++out_h_parent: ++ dput(h_parent); ++out: ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct dentry * ++aufs_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, ++ int fh_type) ++{ ++ struct dentry *dentry; ++ __u32 *fh = fid->raw; ++ struct au_branch *br; ++ ino_t ino, dir_ino; ++ struct au_nfsd_si_lock nsi_lock = { ++ .force_lock = 0 ++ }; ++ ++ dentry = ERR_PTR(-ESTALE); ++ /* it should never happen, but the file handle is unreliable */ ++ if (unlikely(fh_len < Fh_tail)) ++ goto out; ++ nsi_lock.sigen = fh[Fh_sigen]; ++ nsi_lock.br_id = fh[Fh_br_id]; ++ ++ /* branch id may be wrapped around */ ++ br = NULL; ++ if (unlikely(si_nfsd_read_lock(sb, &nsi_lock))) ++ goto out; ++ nsi_lock.force_lock = 1; ++ ++ /* is this inode still cached? */ ++ ino = decode_ino(fh + Fh_ino); ++ /* it should never happen */ ++ if (unlikely(ino == AUFS_ROOT_INO)) ++ goto out; ++ ++ dir_ino = decode_ino(fh + Fh_dir_ino); ++ dentry = decode_by_ino(sb, ino, dir_ino); ++ if (IS_ERR(dentry)) ++ goto out_unlock; ++ if (dentry) ++ goto accept; ++ ++ /* is the parent dir cached? */ ++ br = au_sbr(sb, nsi_lock.bindex); ++ atomic_inc(&br->br_count); ++ dentry = decode_by_dir_ino(sb, ino, dir_ino, &nsi_lock); ++ if (IS_ERR(dentry)) ++ goto out_unlock; ++ if (dentry) ++ goto accept; ++ ++ /* lookup path */ ++ dentry = decode_by_path(sb, ino, fh, fh_len, &nsi_lock); ++ if (IS_ERR(dentry)) ++ goto out_unlock; ++ if (unlikely(!dentry)) ++ /* todo?: make it ESTALE */ ++ goto out_unlock; ++ ++accept: ++ if (!au_digen_test(dentry, au_sigen(sb)) ++ && d_inode(dentry)->i_generation == fh[Fh_igen]) ++ goto out_unlock; /* success */ ++ ++ dput(dentry); ++ dentry = ERR_PTR(-ESTALE); ++out_unlock: ++ if (br) ++ atomic_dec(&br->br_count); ++ si_read_unlock(sb); ++out: ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++#if 0 /* reserved for future use */ ++/* support subtreecheck option */ ++static struct dentry *aufs_fh_to_parent(struct super_block *sb, struct fid *fid, ++ int fh_len, int fh_type) ++{ ++ struct dentry *parent; ++ __u32 *fh = fid->raw; ++ ino_t dir_ino; ++ ++ dir_ino = decode_ino(fh + Fh_dir_ino); ++ parent = decode_by_ino(sb, dir_ino, 0); ++ if (IS_ERR(parent)) ++ goto out; ++ if (!parent) ++ parent = decode_by_path(sb, au_br_index(sb, fh[Fh_br_id]), ++ dir_ino, fh, fh_len); ++ ++out: ++ AuTraceErrPtr(parent); ++ return parent; ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_encode_fh(struct inode *inode, __u32 *fh, int *max_len, ++ struct inode *dir) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct super_block *sb, *h_sb; ++ struct dentry *dentry, *parent, *h_parent; ++ struct inode *h_dir; ++ struct au_branch *br; ++ ++ err = -ENOSPC; ++ if (unlikely(*max_len <= Fh_tail)) { ++ AuWarn1("NFSv2 client (max_len %d)?\n", *max_len); ++ goto out; ++ } ++ ++ err = FILEID_ROOT; ++ if (inode->i_ino == AUFS_ROOT_INO) { ++ AuDebugOn(inode->i_ino != AUFS_ROOT_INO); ++ goto out; ++ } ++ ++ h_parent = NULL; ++ sb = inode->i_sb; ++ err = si_read_lock(sb, AuLock_FLUSH); ++ if (unlikely(err)) ++ goto out; ++ ++#ifdef CONFIG_AUFS_DEBUG ++ if (unlikely(!au_opt_test(au_mntflags(sb), XINO))) ++ AuWarn1("NFS-exporting requires xino\n"); ++#endif ++ err = -EIO; ++ parent = NULL; ++ ii_read_lock_child(inode); ++ bindex = au_ibstart(inode); ++ if (!dir) { ++ dentry = d_find_any_alias(inode); ++ if (unlikely(!dentry)) ++ goto out_unlock; ++ AuDebugOn(au_test_anon(dentry)); ++ parent = dget_parent(dentry); ++ dput(dentry); ++ if (unlikely(!parent)) ++ goto out_unlock; ++ if (d_really_is_positive(parent)) ++ dir = d_inode(parent); ++ } ++ ++ ii_read_lock_parent(dir); ++ h_dir = au_h_iptr(dir, bindex); ++ ii_read_unlock(dir); ++ if (unlikely(!h_dir)) ++ goto out_parent; ++ h_parent = d_find_any_alias(h_dir); ++ if (unlikely(!h_parent)) ++ goto out_hparent; ++ ++ err = -EPERM; ++ br = au_sbr(sb, bindex); ++ h_sb = au_br_sb(br); ++ if (unlikely(!h_sb->s_export_op)) { ++ AuErr1("%s branch is not exportable\n", au_sbtype(h_sb)); ++ goto out_hparent; ++ } ++ ++ fh[Fh_br_id] = br->br_id; ++ fh[Fh_sigen] = au_sigen(sb); ++ encode_ino(fh + Fh_ino, inode->i_ino); ++ encode_ino(fh + Fh_dir_ino, dir->i_ino); ++ fh[Fh_igen] = inode->i_generation; ++ ++ *max_len -= Fh_tail; ++ fh[Fh_h_type] = exportfs_encode_fh(h_parent, (void *)(fh + Fh_tail), ++ max_len, ++ /*connectable or subtreecheck*/0); ++ err = fh[Fh_h_type]; ++ *max_len += Fh_tail; ++ /* todo: macros? */ ++ if (err != FILEID_INVALID) ++ err = 99; ++ else ++ AuWarn1("%s encode_fh failed\n", au_sbtype(h_sb)); ++ ++out_hparent: ++ dput(h_parent); ++out_parent: ++ dput(parent); ++out_unlock: ++ ii_read_unlock(inode); ++ si_read_unlock(sb); ++out: ++ if (unlikely(err < 0)) ++ err = FILEID_INVALID; ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_commit_metadata(struct inode *inode) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct super_block *sb; ++ struct inode *h_inode; ++ int (*f)(struct inode *inode); ++ ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ii_write_lock_child(inode); ++ bindex = au_ibstart(inode); ++ AuDebugOn(bindex < 0); ++ h_inode = au_h_iptr(inode, bindex); ++ ++ f = h_inode->i_sb->s_export_op->commit_metadata; ++ if (f) ++ err = f(h_inode); ++ else { ++ struct writeback_control wbc = { ++ .sync_mode = WB_SYNC_ALL, ++ .nr_to_write = 0 /* metadata only */ ++ }; ++ ++ err = sync_inode(h_inode, &wbc); ++ } ++ ++ au_cpup_attr_timesizes(inode); ++ ii_write_unlock(inode); ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct export_operations aufs_export_op = { ++ .fh_to_dentry = aufs_fh_to_dentry, ++ /* .fh_to_parent = aufs_fh_to_parent, */ ++ .encode_fh = aufs_encode_fh, ++ .commit_metadata = aufs_commit_metadata ++}; ++ ++void au_export_init(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ __u32 u; ++ ++ sb->s_export_op = &aufs_export_op; ++ sbinfo = au_sbi(sb); ++ sbinfo->si_xigen = NULL; ++ get_random_bytes(&u, sizeof(u)); ++ BUILD_BUG_ON(sizeof(u) != sizeof(int)); ++ atomic_set(&sbinfo->si_xigen_next, u); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/fhsm.c linux-4.1.10/fs/aufs/fhsm.c +--- linux-4.1.10.orig/fs/aufs/fhsm.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/fhsm.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,426 @@ ++/* ++ * Copyright (C) 2011-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program; if not, write to the Free Software ++ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA ++ */ ++ ++/* ++ * File-based Hierarchy Storage Management ++ */ ++ ++#include <linux/anon_inodes.h> ++#include <linux/poll.h> ++#include <linux/seq_file.h> ++#include <linux/statfs.h> ++#include "aufs.h" ++ ++static aufs_bindex_t au_fhsm_bottom(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ SiMustAnyLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ AuDebugOn(!fhsm); ++ return fhsm->fhsm_bottom; ++} ++ ++void au_fhsm_set_bottom(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ AuDebugOn(!fhsm); ++ fhsm->fhsm_bottom = bindex; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_fhsm_test_jiffy(struct au_sbinfo *sbinfo, struct au_branch *br) ++{ ++ struct au_br_fhsm *bf; ++ ++ bf = br->br_fhsm; ++ MtxMustLock(&bf->bf_lock); ++ ++ return !bf->bf_readable ++ || time_after(jiffies, ++ bf->bf_jiffy + sbinfo->si_fhsm.fhsm_expire); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_fhsm_notify(struct super_block *sb, int val) ++{ ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ SiMustAnyLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ if (au_fhsm_pid(fhsm) ++ && atomic_read(&fhsm->fhsm_readable) != -1) { ++ atomic_set(&fhsm->fhsm_readable, val); ++ if (val) ++ wake_up(&fhsm->fhsm_wqh); ++ } ++} ++ ++static int au_fhsm_stfs(struct super_block *sb, aufs_bindex_t bindex, ++ struct aufs_stfs *rstfs, int do_lock, int do_notify) ++{ ++ int err; ++ struct au_branch *br; ++ struct au_br_fhsm *bf; ++ ++ br = au_sbr(sb, bindex); ++ AuDebugOn(au_br_rdonly(br)); ++ bf = br->br_fhsm; ++ AuDebugOn(!bf); ++ ++ if (do_lock) ++ mutex_lock(&bf->bf_lock); ++ else ++ MtxMustLock(&bf->bf_lock); ++ ++ /* sb->s_root for NFS is unreliable */ ++ err = au_br_stfs(br, &bf->bf_stfs); ++ if (unlikely(err)) { ++ AuErr1("FHSM failed (%d), b%d, ignored.\n", bindex, err); ++ goto out; ++ } ++ ++ bf->bf_jiffy = jiffies; ++ bf->bf_readable = 1; ++ if (do_notify) ++ au_fhsm_notify(sb, /*val*/1); ++ if (rstfs) ++ *rstfs = bf->bf_stfs; ++ ++out: ++ if (do_lock) ++ mutex_unlock(&bf->bf_lock); ++ au_fhsm_notify(sb, /*val*/1); ++ ++ return err; ++} ++ ++void au_fhsm_wrote(struct super_block *sb, aufs_bindex_t bindex, int force) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ struct au_branch *br; ++ struct au_br_fhsm *bf; ++ ++ AuDbg("b%d, force %d\n", bindex, force); ++ SiMustAnyLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ if (!au_ftest_si(sbinfo, FHSM) ++ || fhsm->fhsm_bottom == bindex) ++ return; ++ ++ br = au_sbr(sb, bindex); ++ bf = br->br_fhsm; ++ AuDebugOn(!bf); ++ mutex_lock(&bf->bf_lock); ++ if (force ++ || au_fhsm_pid(fhsm) ++ || au_fhsm_test_jiffy(sbinfo, br)) ++ err = au_fhsm_stfs(sb, bindex, /*rstfs*/NULL, /*do_lock*/0, ++ /*do_notify*/1); ++ mutex_unlock(&bf->bf_lock); ++} ++ ++void au_fhsm_wrote_all(struct super_block *sb, int force) ++{ ++ aufs_bindex_t bindex, bend; ++ struct au_branch *br; ++ ++ /* exclude the bottom */ ++ bend = au_fhsm_bottom(sb); ++ for (bindex = 0; bindex < bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_fhsm(br->br_perm)) ++ au_fhsm_wrote(sb, bindex, force); ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static unsigned int au_fhsm_poll(struct file *file, ++ struct poll_table_struct *wait) ++{ ++ unsigned int mask; ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ mask = 0; ++ sbinfo = file->private_data; ++ fhsm = &sbinfo->si_fhsm; ++ poll_wait(file, &fhsm->fhsm_wqh, wait); ++ if (atomic_read(&fhsm->fhsm_readable)) ++ mask = POLLIN /* | POLLRDNORM */; ++ ++ AuTraceErr((int)mask); ++ return mask; ++} ++ ++static int au_fhsm_do_read_one(struct aufs_stbr __user *stbr, ++ struct aufs_stfs *stfs, __s16 brid) ++{ ++ int err; ++ ++ err = copy_to_user(&stbr->stfs, stfs, sizeof(*stfs)); ++ if (!err) ++ err = __put_user(brid, &stbr->brid); ++ if (unlikely(err)) ++ err = -EFAULT; ++ ++ return err; ++} ++ ++static ssize_t au_fhsm_do_read(struct super_block *sb, ++ struct aufs_stbr __user *stbr, size_t count) ++{ ++ ssize_t err; ++ int nstbr; ++ aufs_bindex_t bindex, bend; ++ struct au_branch *br; ++ struct au_br_fhsm *bf; ++ ++ /* except the bottom branch */ ++ err = 0; ++ nstbr = 0; ++ bend = au_fhsm_bottom(sb); ++ for (bindex = 0; !err && bindex < bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (!au_br_fhsm(br->br_perm)) ++ continue; ++ ++ bf = br->br_fhsm; ++ mutex_lock(&bf->bf_lock); ++ if (bf->bf_readable) { ++ err = -EFAULT; ++ if (count >= sizeof(*stbr)) ++ err = au_fhsm_do_read_one(stbr++, &bf->bf_stfs, ++ br->br_id); ++ if (!err) { ++ bf->bf_readable = 0; ++ count -= sizeof(*stbr); ++ nstbr++; ++ } ++ } ++ mutex_unlock(&bf->bf_lock); ++ } ++ if (!err) ++ err = sizeof(*stbr) * nstbr; ++ ++ return err; ++} ++ ++static ssize_t au_fhsm_read(struct file *file, char __user *buf, size_t count, ++ loff_t *pos) ++{ ++ ssize_t err; ++ int readable; ++ aufs_bindex_t nfhsm, bindex, bend; ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ struct au_branch *br; ++ struct super_block *sb; ++ ++ err = 0; ++ sbinfo = file->private_data; ++ fhsm = &sbinfo->si_fhsm; ++need_data: ++ spin_lock_irq(&fhsm->fhsm_wqh.lock); ++ if (!atomic_read(&fhsm->fhsm_readable)) { ++ if (vfsub_file_flags(file) & O_NONBLOCK) ++ err = -EAGAIN; ++ else ++ err = wait_event_interruptible_locked_irq ++ (fhsm->fhsm_wqh, ++ atomic_read(&fhsm->fhsm_readable)); ++ } ++ spin_unlock_irq(&fhsm->fhsm_wqh.lock); ++ if (unlikely(err)) ++ goto out; ++ ++ /* sb may already be dead */ ++ au_rw_read_lock(&sbinfo->si_rwsem); ++ readable = atomic_read(&fhsm->fhsm_readable); ++ if (readable > 0) { ++ sb = sbinfo->si_sb; ++ AuDebugOn(!sb); ++ /* exclude the bottom branch */ ++ nfhsm = 0; ++ bend = au_fhsm_bottom(sb); ++ for (bindex = 0; bindex < bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_fhsm(br->br_perm)) ++ nfhsm++; ++ } ++ err = -EMSGSIZE; ++ if (nfhsm * sizeof(struct aufs_stbr) <= count) { ++ atomic_set(&fhsm->fhsm_readable, 0); ++ err = au_fhsm_do_read(sbinfo->si_sb, (void __user *)buf, ++ count); ++ } ++ } ++ au_rw_read_unlock(&sbinfo->si_rwsem); ++ if (!readable) ++ goto need_data; ++ ++out: ++ return err; ++} ++ ++static int au_fhsm_release(struct inode *inode, struct file *file) ++{ ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ /* sb may already be dead */ ++ sbinfo = file->private_data; ++ fhsm = &sbinfo->si_fhsm; ++ spin_lock(&fhsm->fhsm_spin); ++ fhsm->fhsm_pid = 0; ++ spin_unlock(&fhsm->fhsm_spin); ++ kobject_put(&sbinfo->si_kobj); ++ ++ return 0; ++} ++ ++static const struct file_operations au_fhsm_fops = { ++ .owner = THIS_MODULE, ++ .llseek = noop_llseek, ++ .read = au_fhsm_read, ++ .poll = au_fhsm_poll, ++ .release = au_fhsm_release ++}; ++ ++int au_fhsm_fd(struct super_block *sb, int oflags) ++{ ++ int err, fd; ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ ++ err = -EPERM; ++ if (unlikely(!capable(CAP_SYS_ADMIN))) ++ goto out; ++ ++ err = -EINVAL; ++ if (unlikely(oflags & ~(O_CLOEXEC | O_NONBLOCK))) ++ goto out; ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ spin_lock(&fhsm->fhsm_spin); ++ if (!fhsm->fhsm_pid) ++ fhsm->fhsm_pid = current->pid; ++ else ++ err = -EBUSY; ++ spin_unlock(&fhsm->fhsm_spin); ++ if (unlikely(err)) ++ goto out; ++ ++ oflags |= O_RDONLY; ++ /* oflags |= FMODE_NONOTIFY; */ ++ fd = anon_inode_getfd("[aufs_fhsm]", &au_fhsm_fops, sbinfo, oflags); ++ err = fd; ++ if (unlikely(fd < 0)) ++ goto out_pid; ++ ++ /* succeed reglardless 'fhsm' status */ ++ kobject_get(&sbinfo->si_kobj); ++ si_noflush_read_lock(sb); ++ if (au_ftest_si(sbinfo, FHSM)) ++ au_fhsm_wrote_all(sb, /*force*/0); ++ si_read_unlock(sb); ++ goto out; /* success */ ++ ++out_pid: ++ spin_lock(&fhsm->fhsm_spin); ++ fhsm->fhsm_pid = 0; ++ spin_unlock(&fhsm->fhsm_spin); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_fhsm_br_alloc(struct au_branch *br) ++{ ++ int err; ++ ++ err = 0; ++ br->br_fhsm = kmalloc(sizeof(*br->br_fhsm), GFP_NOFS); ++ if (br->br_fhsm) ++ au_br_fhsm_init(br->br_fhsm); ++ else ++ err = -ENOMEM; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_fhsm_fin(struct super_block *sb) ++{ ++ au_fhsm_notify(sb, /*val*/-1); ++} ++ ++void au_fhsm_init(struct au_sbinfo *sbinfo) ++{ ++ struct au_fhsm *fhsm; ++ ++ fhsm = &sbinfo->si_fhsm; ++ spin_lock_init(&fhsm->fhsm_spin); ++ init_waitqueue_head(&fhsm->fhsm_wqh); ++ atomic_set(&fhsm->fhsm_readable, 0); ++ fhsm->fhsm_expire ++ = msecs_to_jiffies(AUFS_FHSM_CACHE_DEF_SEC * MSEC_PER_SEC); ++ fhsm->fhsm_bottom = -1; ++} ++ ++void au_fhsm_set(struct au_sbinfo *sbinfo, unsigned int sec) ++{ ++ sbinfo->si_fhsm.fhsm_expire ++ = msecs_to_jiffies(sec * MSEC_PER_SEC); ++} ++ ++void au_fhsm_show(struct seq_file *seq, struct au_sbinfo *sbinfo) ++{ ++ unsigned int u; ++ ++ if (!au_ftest_si(sbinfo, FHSM)) ++ return; ++ ++ u = jiffies_to_msecs(sbinfo->si_fhsm.fhsm_expire) / MSEC_PER_SEC; ++ if (u != AUFS_FHSM_CACHE_DEF_SEC) ++ seq_printf(seq, ",fhsm_sec=%u", u); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/file.c linux-4.1.10/fs/aufs/file.c +--- linux-4.1.10.orig/fs/aufs/file.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/file.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,841 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * handling file/dir, and address_space operation ++ */ ++ ++#ifdef CONFIG_AUFS_DEBUG ++#include <linux/migrate.h> ++#endif ++#include <linux/pagemap.h> ++#include "aufs.h" ++ ++/* drop flags for writing */ ++unsigned int au_file_roflags(unsigned int flags) ++{ ++ flags &= ~(O_WRONLY | O_RDWR | O_APPEND | O_CREAT | O_TRUNC); ++ flags |= O_RDONLY | O_NOATIME; ++ return flags; ++} ++ ++/* common functions to regular file and dir */ ++struct file *au_h_open(struct dentry *dentry, aufs_bindex_t bindex, int flags, ++ struct file *file, int force_wr) ++{ ++ struct file *h_file; ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct path h_path; ++ int err; ++ ++ /* a race condition can happen between open and unlink/rmdir */ ++ h_file = ERR_PTR(-ENOENT); ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (au_test_nfsd() && (!h_dentry || d_is_negative(h_dentry))) ++ goto out; ++ h_inode = d_inode(h_dentry); ++ spin_lock(&h_dentry->d_lock); ++ err = (!d_unhashed(dentry) && d_unlinked(h_dentry)) ++ /* || !d_inode(dentry)->i_nlink */ ++ ; ++ spin_unlock(&h_dentry->d_lock); ++ if (unlikely(err)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ br = au_sbr(sb, bindex); ++ err = au_br_test_oflag(flags, br); ++ h_file = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ /* drop flags for writing */ ++ if (au_test_ro(sb, bindex, d_inode(dentry))) { ++ if (force_wr && !(flags & O_WRONLY)) ++ force_wr = 0; ++ flags = au_file_roflags(flags); ++ if (force_wr) { ++ h_file = ERR_PTR(-EROFS); ++ flags = au_file_roflags(flags); ++ if (unlikely(vfsub_native_ro(h_inode) ++ || IS_APPEND(h_inode))) ++ goto out; ++ flags &= ~O_ACCMODE; ++ flags |= O_WRONLY; ++ } ++ } ++ flags &= ~O_CREAT; ++ atomic_inc(&br->br_count); ++ h_path.dentry = h_dentry; ++ h_path.mnt = au_br_mnt(br); ++ h_file = vfsub_dentry_open(&h_path, flags); ++ if (IS_ERR(h_file)) ++ goto out_br; ++ ++ if (flags & __FMODE_EXEC) { ++ err = deny_write_access(h_file); ++ if (unlikely(err)) { ++ fput(h_file); ++ h_file = ERR_PTR(err); ++ goto out_br; ++ } ++ } ++ fsnotify_open(h_file); ++ goto out; /* success */ ++ ++out_br: ++ atomic_dec(&br->br_count); ++out: ++ return h_file; ++} ++ ++static int au_cmoo(struct dentry *dentry) ++{ ++ int err, cmoo; ++ unsigned int udba; ++ struct path h_path; ++ struct au_pin pin; ++ struct au_cp_generic cpg = { ++ .dentry = dentry, ++ .bdst = -1, ++ .bsrc = -1, ++ .len = -1, ++ .pin = &pin, ++ .flags = AuCpup_DTIME | AuCpup_HOPEN ++ }; ++ struct inode *delegated; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ struct au_fhsm *fhsm; ++ pid_t pid; ++ struct au_branch *br; ++ struct dentry *parent; ++ struct au_hinode *hdir; ++ ++ DiMustWriteLock(dentry); ++ IiMustWriteLock(d_inode(dentry)); ++ ++ err = 0; ++ if (IS_ROOT(dentry)) ++ goto out; ++ cpg.bsrc = au_dbstart(dentry); ++ if (!cpg.bsrc) ++ goto out; ++ ++ sb = dentry->d_sb; ++ sbinfo = au_sbi(sb); ++ fhsm = &sbinfo->si_fhsm; ++ pid = au_fhsm_pid(fhsm); ++ if (pid ++ && (current->pid == pid ++ || current->real_parent->pid == pid)) ++ goto out; ++ ++ br = au_sbr(sb, cpg.bsrc); ++ cmoo = au_br_cmoo(br->br_perm); ++ if (!cmoo) ++ goto out; ++ if (!d_is_reg(dentry)) ++ cmoo &= AuBrAttr_COO_ALL; ++ if (!cmoo) ++ goto out; ++ ++ parent = dget_parent(dentry); ++ di_write_lock_parent(parent); ++ err = au_wbr_do_copyup_bu(dentry, cpg.bsrc - 1); ++ cpg.bdst = err; ++ if (unlikely(err < 0)) { ++ err = 0; /* there is no upper writable branch */ ++ goto out_dgrade; ++ } ++ AuDbg("bsrc %d, bdst %d\n", cpg.bsrc, cpg.bdst); ++ ++ /* do not respect the coo attrib for the target branch */ ++ err = au_cpup_dirs(dentry, cpg.bdst); ++ if (unlikely(err)) ++ goto out_dgrade; ++ ++ di_downgrade_lock(parent, AuLock_IR); ++ udba = au_opt_udba(sb); ++ err = au_pin(&pin, dentry, cpg.bdst, udba, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ err = au_sio_cpup_simple(&cpg); ++ au_unpin(&pin); ++ if (unlikely(err)) ++ goto out_parent; ++ if (!(cmoo & AuBrWAttr_MOO)) ++ goto out_parent; /* success */ ++ ++ err = au_pin(&pin, dentry, cpg.bsrc, udba, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ h_path.mnt = au_br_mnt(br); ++ h_path.dentry = au_h_dptr(dentry, cpg.bsrc); ++ hdir = au_hi(d_inode(parent), cpg.bsrc); ++ delegated = NULL; ++ err = vfsub_unlink(hdir->hi_inode, &h_path, &delegated, /*force*/1); ++ au_unpin(&pin); ++ /* todo: keep h_dentry or not? */ ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ if (unlikely(err)) { ++ pr_err("unlink %pd after coo failed (%d), ignored\n", ++ dentry, err); ++ err = 0; ++ } ++ goto out_parent; /* success */ ++ ++out_dgrade: ++ di_downgrade_lock(parent, AuLock_IR); ++out_parent: ++ di_read_unlock(parent, AuLock_IR); ++ dput(parent); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_do_open(struct file *file, struct au_do_open_args *args) ++{ ++ int err, no_lock = args->no_lock; ++ struct dentry *dentry; ++ struct au_finfo *finfo; ++ ++ if (!no_lock) ++ err = au_finfo_init(file, args->fidir); ++ else { ++ lockdep_off(); ++ err = au_finfo_init(file, args->fidir); ++ lockdep_on(); ++ } ++ if (unlikely(err)) ++ goto out; ++ ++ dentry = file->f_path.dentry; ++ AuDebugOn(IS_ERR_OR_NULL(dentry)); ++ if (!no_lock) { ++ di_write_lock_child(dentry); ++ err = au_cmoo(dentry); ++ di_downgrade_lock(dentry, AuLock_IR); ++ if (!err) ++ err = args->open(file, vfsub_file_flags(file), NULL); ++ di_read_unlock(dentry, AuLock_IR); ++ } else { ++ err = au_cmoo(dentry); ++ if (!err) ++ err = args->open(file, vfsub_file_flags(file), ++ args->h_file); ++ if (!err && au_fbstart(file) != au_dbstart(dentry)) ++ /* ++ * cmoo happens after h_file was opened. ++ * need to refresh file later. ++ */ ++ atomic_dec(&au_fi(file)->fi_generation); ++ } ++ ++ finfo = au_fi(file); ++ if (!err) { ++ finfo->fi_file = file; ++ au_sphl_add(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ } ++ if (!no_lock) ++ fi_write_unlock(file); ++ else { ++ lockdep_off(); ++ fi_write_unlock(file); ++ lockdep_on(); ++ } ++ if (unlikely(err)) { ++ finfo->fi_hdir = NULL; ++ au_finfo_fin(file); ++ } ++ ++out: ++ return err; ++} ++ ++int au_reopen_nondir(struct file *file) ++{ ++ int err; ++ aufs_bindex_t bstart; ++ struct dentry *dentry; ++ struct file *h_file, *h_file_tmp; ++ ++ dentry = file->f_path.dentry; ++ bstart = au_dbstart(dentry); ++ h_file_tmp = NULL; ++ if (au_fbstart(file) == bstart) { ++ h_file = au_hf_top(file); ++ if (file->f_mode == h_file->f_mode) ++ return 0; /* success */ ++ h_file_tmp = h_file; ++ get_file(h_file_tmp); ++ au_set_h_fptr(file, bstart, NULL); ++ } ++ AuDebugOn(au_fi(file)->fi_hdir); ++ /* ++ * it can happen ++ * file exists on both of rw and ro ++ * open --> dbstart and fbstart are both 0 ++ * prepend a branch as rw, "rw" become ro ++ * remove rw/file ++ * delete the top branch, "rw" becomes rw again ++ * --> dbstart is 1, fbstart is still 0 ++ * write --> fbstart is 0 but dbstart is 1 ++ */ ++ /* AuDebugOn(au_fbstart(file) < bstart); */ ++ ++ h_file = au_h_open(dentry, bstart, vfsub_file_flags(file) & ~O_TRUNC, ++ file, /*force_wr*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) { ++ if (h_file_tmp) { ++ atomic_inc(&au_sbr(dentry->d_sb, bstart)->br_count); ++ au_set_h_fptr(file, bstart, h_file_tmp); ++ h_file_tmp = NULL; ++ } ++ goto out; /* todo: close all? */ ++ } ++ ++ err = 0; ++ au_set_fbstart(file, bstart); ++ au_set_h_fptr(file, bstart, h_file); ++ au_update_figen(file); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ ++out: ++ if (h_file_tmp) ++ fput(h_file_tmp); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_reopen_wh(struct file *file, aufs_bindex_t btgt, ++ struct dentry *hi_wh) ++{ ++ int err; ++ aufs_bindex_t bstart; ++ struct au_dinfo *dinfo; ++ struct dentry *h_dentry; ++ struct au_hdentry *hdp; ++ ++ dinfo = au_di(file->f_path.dentry); ++ AuRwMustWriteLock(&dinfo->di_rwsem); ++ ++ bstart = dinfo->di_bstart; ++ dinfo->di_bstart = btgt; ++ hdp = dinfo->di_hdentry; ++ h_dentry = hdp[0 + btgt].hd_dentry; ++ hdp[0 + btgt].hd_dentry = hi_wh; ++ err = au_reopen_nondir(file); ++ hdp[0 + btgt].hd_dentry = h_dentry; ++ dinfo->di_bstart = bstart; ++ ++ return err; ++} ++ ++static int au_ready_to_write_wh(struct file *file, loff_t len, ++ aufs_bindex_t bcpup, struct au_pin *pin) ++{ ++ int err; ++ struct inode *inode, *h_inode; ++ struct dentry *h_dentry, *hi_wh; ++ struct au_cp_generic cpg = { ++ .dentry = file->f_path.dentry, ++ .bdst = bcpup, ++ .bsrc = -1, ++ .len = len, ++ .pin = pin ++ }; ++ ++ au_update_dbstart(cpg.dentry); ++ inode = d_inode(cpg.dentry); ++ h_inode = NULL; ++ if (au_dbstart(cpg.dentry) <= bcpup ++ && au_dbend(cpg.dentry) >= bcpup) { ++ h_dentry = au_h_dptr(cpg.dentry, bcpup); ++ if (h_dentry && d_is_positive(h_dentry)) ++ h_inode = d_inode(h_dentry); ++ } ++ hi_wh = au_hi_wh(inode, bcpup); ++ if (!hi_wh && !h_inode) ++ err = au_sio_cpup_wh(&cpg, file); ++ else ++ /* already copied-up after unlink */ ++ err = au_reopen_wh(file, bcpup, hi_wh); ++ ++ if (!err ++ && (inode->i_nlink > 1 ++ || (inode->i_state & I_LINKABLE)) ++ && au_opt_test(au_mntflags(cpg.dentry->d_sb), PLINK)) ++ au_plink_append(inode, bcpup, au_h_dptr(cpg.dentry, bcpup)); ++ ++ return err; ++} ++ ++/* ++ * prepare the @file for writing. ++ */ ++int au_ready_to_write(struct file *file, loff_t len, struct au_pin *pin) ++{ ++ int err; ++ aufs_bindex_t dbstart; ++ struct dentry *parent; ++ struct inode *inode; ++ struct super_block *sb; ++ struct file *h_file; ++ struct au_cp_generic cpg = { ++ .dentry = file->f_path.dentry, ++ .bdst = -1, ++ .bsrc = -1, ++ .len = len, ++ .pin = pin, ++ .flags = AuCpup_DTIME ++ }; ++ ++ sb = cpg.dentry->d_sb; ++ inode = d_inode(cpg.dentry); ++ cpg.bsrc = au_fbstart(file); ++ err = au_test_ro(sb, cpg.bsrc, inode); ++ if (!err && (au_hf_top(file)->f_mode & FMODE_WRITE)) { ++ err = au_pin(pin, cpg.dentry, cpg.bsrc, AuOpt_UDBA_NONE, ++ /*flags*/0); ++ goto out; ++ } ++ ++ /* need to cpup or reopen */ ++ parent = dget_parent(cpg.dentry); ++ di_write_lock_parent(parent); ++ err = AuWbrCopyup(au_sbi(sb), cpg.dentry); ++ cpg.bdst = err; ++ if (unlikely(err < 0)) ++ goto out_dgrade; ++ err = 0; ++ ++ if (!d_unhashed(cpg.dentry) && !au_h_dptr(parent, cpg.bdst)) { ++ err = au_cpup_dirs(cpg.dentry, cpg.bdst); ++ if (unlikely(err)) ++ goto out_dgrade; ++ } ++ ++ err = au_pin(pin, cpg.dentry, cpg.bdst, AuOpt_UDBA_NONE, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (unlikely(err)) ++ goto out_dgrade; ++ ++ dbstart = au_dbstart(cpg.dentry); ++ if (dbstart <= cpg.bdst) ++ cpg.bsrc = cpg.bdst; ++ ++ if (dbstart <= cpg.bdst /* just reopen */ ++ || !d_unhashed(cpg.dentry) /* copyup and reopen */ ++ ) { ++ h_file = au_h_open_pre(cpg.dentry, cpg.bsrc, /*force_wr*/0); ++ if (IS_ERR(h_file)) ++ err = PTR_ERR(h_file); ++ else { ++ di_downgrade_lock(parent, AuLock_IR); ++ if (dbstart > cpg.bdst) ++ err = au_sio_cpup_simple(&cpg); ++ if (!err) ++ err = au_reopen_nondir(file); ++ au_h_open_post(cpg.dentry, cpg.bsrc, h_file); ++ } ++ } else { /* copyup as wh and reopen */ ++ /* ++ * since writable hfsplus branch is not supported, ++ * h_open_pre/post() are unnecessary. ++ */ ++ err = au_ready_to_write_wh(file, len, cpg.bdst, pin); ++ di_downgrade_lock(parent, AuLock_IR); ++ } ++ ++ if (!err) { ++ au_pin_set_parent_lflag(pin, /*lflag*/0); ++ goto out_dput; /* success */ ++ } ++ au_unpin(pin); ++ goto out_unlock; ++ ++out_dgrade: ++ di_downgrade_lock(parent, AuLock_IR); ++out_unlock: ++ di_read_unlock(parent, AuLock_IR); ++out_dput: ++ dput(parent); ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_do_flush(struct file *file, fl_owner_t id, ++ int (*flush)(struct file *file, fl_owner_t id)) ++{ ++ int err; ++ struct super_block *sb; ++ struct inode *inode; ++ ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ si_noflush_read_lock(sb); ++ fi_read_lock(file); ++ ii_read_lock_child(inode); ++ ++ err = flush(file, id); ++ au_cpup_attr_timesizes(inode); ++ ++ ii_read_unlock(inode); ++ fi_read_unlock(file); ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_file_refresh_by_inode(struct file *file, int *need_reopen) ++{ ++ int err; ++ struct au_pin pin; ++ struct au_finfo *finfo; ++ struct dentry *parent, *hi_wh; ++ struct inode *inode; ++ struct super_block *sb; ++ struct au_cp_generic cpg = { ++ .dentry = file->f_path.dentry, ++ .bdst = -1, ++ .bsrc = -1, ++ .len = -1, ++ .pin = &pin, ++ .flags = AuCpup_DTIME ++ }; ++ ++ FiMustWriteLock(file); ++ ++ err = 0; ++ finfo = au_fi(file); ++ sb = cpg.dentry->d_sb; ++ inode = d_inode(cpg.dentry); ++ cpg.bdst = au_ibstart(inode); ++ if (cpg.bdst == finfo->fi_btop || IS_ROOT(cpg.dentry)) ++ goto out; ++ ++ parent = dget_parent(cpg.dentry); ++ if (au_test_ro(sb, cpg.bdst, inode)) { ++ di_read_lock_parent(parent, !AuLock_IR); ++ err = AuWbrCopyup(au_sbi(sb), cpg.dentry); ++ cpg.bdst = err; ++ di_read_unlock(parent, !AuLock_IR); ++ if (unlikely(err < 0)) ++ goto out_parent; ++ err = 0; ++ } ++ ++ di_read_lock_parent(parent, AuLock_IR); ++ hi_wh = au_hi_wh(inode, cpg.bdst); ++ if (!S_ISDIR(inode->i_mode) ++ && au_opt_test(au_mntflags(sb), PLINK) ++ && au_plink_test(inode) ++ && !d_unhashed(cpg.dentry) ++ && cpg.bdst < au_dbstart(cpg.dentry)) { ++ err = au_test_and_cpup_dirs(cpg.dentry, cpg.bdst); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ /* always superio. */ ++ err = au_pin(&pin, cpg.dentry, cpg.bdst, AuOpt_UDBA_NONE, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (!err) { ++ err = au_sio_cpup_simple(&cpg); ++ au_unpin(&pin); ++ } ++ } else if (hi_wh) { ++ /* already copied-up after unlink */ ++ err = au_reopen_wh(file, cpg.bdst, hi_wh); ++ *need_reopen = 0; ++ } ++ ++out_unlock: ++ di_read_unlock(parent, AuLock_IR); ++out_parent: ++ dput(parent); ++out: ++ return err; ++} ++ ++static void au_do_refresh_dir(struct file *file) ++{ ++ aufs_bindex_t bindex, bend, new_bindex, brid; ++ struct au_hfile *p, tmp, *q; ++ struct au_finfo *finfo; ++ struct super_block *sb; ++ struct au_fidir *fidir; ++ ++ FiMustWriteLock(file); ++ ++ sb = file->f_path.dentry->d_sb; ++ finfo = au_fi(file); ++ fidir = finfo->fi_hdir; ++ AuDebugOn(!fidir); ++ p = fidir->fd_hfile + finfo->fi_btop; ++ brid = p->hf_br->br_id; ++ bend = fidir->fd_bbot; ++ for (bindex = finfo->fi_btop; bindex <= bend; bindex++, p++) { ++ if (!p->hf_file) ++ continue; ++ ++ new_bindex = au_br_index(sb, p->hf_br->br_id); ++ if (new_bindex == bindex) ++ continue; ++ if (new_bindex < 0) { ++ au_set_h_fptr(file, bindex, NULL); ++ continue; ++ } ++ ++ /* swap two lower inode, and loop again */ ++ q = fidir->fd_hfile + new_bindex; ++ tmp = *q; ++ *q = *p; ++ *p = tmp; ++ if (tmp.hf_file) { ++ bindex--; ++ p--; ++ } ++ } ++ ++ p = fidir->fd_hfile; ++ if (!au_test_mmapped(file) && !d_unlinked(file->f_path.dentry)) { ++ bend = au_sbend(sb); ++ for (finfo->fi_btop = 0; finfo->fi_btop <= bend; ++ finfo->fi_btop++, p++) ++ if (p->hf_file) { ++ if (file_inode(p->hf_file)) ++ break; ++ au_hfput(p, file); ++ } ++ } else { ++ bend = au_br_index(sb, brid); ++ for (finfo->fi_btop = 0; finfo->fi_btop < bend; ++ finfo->fi_btop++, p++) ++ if (p->hf_file) ++ au_hfput(p, file); ++ bend = au_sbend(sb); ++ } ++ ++ p = fidir->fd_hfile + bend; ++ for (fidir->fd_bbot = bend; fidir->fd_bbot >= finfo->fi_btop; ++ fidir->fd_bbot--, p--) ++ if (p->hf_file) { ++ if (file_inode(p->hf_file)) ++ break; ++ au_hfput(p, file); ++ } ++ AuDebugOn(fidir->fd_bbot < finfo->fi_btop); ++} ++ ++/* ++ * after branch manipulating, refresh the file. ++ */ ++static int refresh_file(struct file *file, int (*reopen)(struct file *file)) ++{ ++ int err, need_reopen; ++ aufs_bindex_t bend, bindex; ++ struct dentry *dentry; ++ struct au_finfo *finfo; ++ struct au_hfile *hfile; ++ ++ dentry = file->f_path.dentry; ++ finfo = au_fi(file); ++ if (!finfo->fi_hdir) { ++ hfile = &finfo->fi_htop; ++ AuDebugOn(!hfile->hf_file); ++ bindex = au_br_index(dentry->d_sb, hfile->hf_br->br_id); ++ AuDebugOn(bindex < 0); ++ if (bindex != finfo->fi_btop) ++ au_set_fbstart(file, bindex); ++ } else { ++ err = au_fidir_realloc(finfo, au_sbend(dentry->d_sb) + 1); ++ if (unlikely(err)) ++ goto out; ++ au_do_refresh_dir(file); ++ } ++ ++ err = 0; ++ need_reopen = 1; ++ if (!au_test_mmapped(file)) ++ err = au_file_refresh_by_inode(file, &need_reopen); ++ if (!err && need_reopen && !d_unlinked(dentry)) ++ err = reopen(file); ++ if (!err) { ++ au_update_figen(file); ++ goto out; /* success */ ++ } ++ ++ /* error, close all lower files */ ++ if (finfo->fi_hdir) { ++ bend = au_fbend_dir(file); ++ for (bindex = au_fbstart(file); bindex <= bend; bindex++) ++ au_set_h_fptr(file, bindex, NULL); ++ } ++ ++out: ++ return err; ++} ++ ++/* common function to regular file and dir */ ++int au_reval_and_lock_fdi(struct file *file, int (*reopen)(struct file *file), ++ int wlock) ++{ ++ int err; ++ unsigned int sigen, figen; ++ aufs_bindex_t bstart; ++ unsigned char pseudo_link; ++ struct dentry *dentry; ++ struct inode *inode; ++ ++ err = 0; ++ dentry = file->f_path.dentry; ++ inode = d_inode(dentry); ++ sigen = au_sigen(dentry->d_sb); ++ fi_write_lock(file); ++ figen = au_figen(file); ++ di_write_lock_child(dentry); ++ bstart = au_dbstart(dentry); ++ pseudo_link = (bstart != au_ibstart(inode)); ++ if (sigen == figen && !pseudo_link && au_fbstart(file) == bstart) { ++ if (!wlock) { ++ di_downgrade_lock(dentry, AuLock_IR); ++ fi_downgrade_lock(file); ++ } ++ goto out; /* success */ ++ } ++ ++ AuDbg("sigen %d, figen %d\n", sigen, figen); ++ if (au_digen_test(dentry, sigen)) { ++ err = au_reval_dpath(dentry, sigen); ++ AuDebugOn(!err && au_digen_test(dentry, sigen)); ++ } ++ ++ if (!err) ++ err = refresh_file(file, reopen); ++ if (!err) { ++ if (!wlock) { ++ di_downgrade_lock(dentry, AuLock_IR); ++ fi_downgrade_lock(file); ++ } ++ } else { ++ di_write_unlock(dentry); ++ fi_write_unlock(file); ++ } ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* cf. aufs_nopage() */ ++/* for madvise(2) */ ++static int aufs_readpage(struct file *file __maybe_unused, struct page *page) ++{ ++ unlock_page(page); ++ return 0; ++} ++ ++/* it will never be called, but necessary to support O_DIRECT */ ++static ssize_t aufs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, ++ loff_t offset) ++{ BUG(); return 0; } ++ ++/* they will never be called. */ ++#ifdef CONFIG_AUFS_DEBUG ++static int aufs_write_begin(struct file *file, struct address_space *mapping, ++ loff_t pos, unsigned len, unsigned flags, ++ struct page **pagep, void **fsdata) ++{ AuUnsupport(); return 0; } ++static int aufs_write_end(struct file *file, struct address_space *mapping, ++ loff_t pos, unsigned len, unsigned copied, ++ struct page *page, void *fsdata) ++{ AuUnsupport(); return 0; } ++static int aufs_writepage(struct page *page, struct writeback_control *wbc) ++{ AuUnsupport(); return 0; } ++ ++static int aufs_set_page_dirty(struct page *page) ++{ AuUnsupport(); return 0; } ++static void aufs_invalidatepage(struct page *page, unsigned int offset, ++ unsigned int length) ++{ AuUnsupport(); } ++static int aufs_releasepage(struct page *page, gfp_t gfp) ++{ AuUnsupport(); return 0; } ++static int aufs_migratepage(struct address_space *mapping, struct page *newpage, ++ struct page *page, enum migrate_mode mode) ++{ AuUnsupport(); return 0; } ++static int aufs_launder_page(struct page *page) ++{ AuUnsupport(); return 0; } ++static int aufs_is_partially_uptodate(struct page *page, ++ unsigned long from, ++ unsigned long count) ++{ AuUnsupport(); return 0; } ++static void aufs_is_dirty_writeback(struct page *page, bool *dirty, ++ bool *writeback) ++{ AuUnsupport(); } ++static int aufs_error_remove_page(struct address_space *mapping, ++ struct page *page) ++{ AuUnsupport(); return 0; } ++static int aufs_swap_activate(struct swap_info_struct *sis, struct file *file, ++ sector_t *span) ++{ AuUnsupport(); return 0; } ++static void aufs_swap_deactivate(struct file *file) ++{ AuUnsupport(); } ++#endif /* CONFIG_AUFS_DEBUG */ ++ ++const struct address_space_operations aufs_aop = { ++ .readpage = aufs_readpage, ++ .direct_IO = aufs_direct_IO, ++#ifdef CONFIG_AUFS_DEBUG ++ .writepage = aufs_writepage, ++ /* no writepages, because of writepage */ ++ .set_page_dirty = aufs_set_page_dirty, ++ /* no readpages, because of readpage */ ++ .write_begin = aufs_write_begin, ++ .write_end = aufs_write_end, ++ /* no bmap, no block device */ ++ .invalidatepage = aufs_invalidatepage, ++ .releasepage = aufs_releasepage, ++ .migratepage = aufs_migratepage, ++ .launder_page = aufs_launder_page, ++ .is_partially_uptodate = aufs_is_partially_uptodate, ++ .is_dirty_writeback = aufs_is_dirty_writeback, ++ .error_remove_page = aufs_error_remove_page, ++ .swap_activate = aufs_swap_activate, ++ .swap_deactivate = aufs_swap_deactivate ++#endif /* CONFIG_AUFS_DEBUG */ ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/file.h linux-4.1.10/fs/aufs/file.h +--- linux-4.1.10.orig/fs/aufs/file.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/file.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,291 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * file operations ++ */ ++ ++#ifndef __AUFS_FILE_H__ ++#define __AUFS_FILE_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/file.h> ++#include <linux/fs.h> ++#include <linux/poll.h> ++#include "rwsem.h" ++ ++struct au_branch; ++struct au_hfile { ++ struct file *hf_file; ++ struct au_branch *hf_br; ++}; ++ ++struct au_vdir; ++struct au_fidir { ++ aufs_bindex_t fd_bbot; ++ aufs_bindex_t fd_nent; ++ struct au_vdir *fd_vdir_cache; ++ struct au_hfile fd_hfile[]; ++}; ++ ++static inline int au_fidir_sz(int nent) ++{ ++ AuDebugOn(nent < 0); ++ return sizeof(struct au_fidir) + sizeof(struct au_hfile) * nent; ++} ++ ++struct au_finfo { ++ atomic_t fi_generation; ++ ++ struct au_rwsem fi_rwsem; ++ aufs_bindex_t fi_btop; ++ ++ /* do not union them */ ++ struct { /* for non-dir */ ++ struct au_hfile fi_htop; ++ atomic_t fi_mmapped; ++ }; ++ struct au_fidir *fi_hdir; /* for dir only */ ++ ++ struct hlist_node fi_hlist; ++ struct file *fi_file; /* very ugly */ ++} ____cacheline_aligned_in_smp; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* file.c */ ++extern const struct address_space_operations aufs_aop; ++unsigned int au_file_roflags(unsigned int flags); ++struct file *au_h_open(struct dentry *dentry, aufs_bindex_t bindex, int flags, ++ struct file *file, int force_wr); ++struct au_do_open_args { ++ int no_lock; ++ int (*open)(struct file *file, int flags, ++ struct file *h_file); ++ struct au_fidir *fidir; ++ struct file *h_file; ++}; ++int au_do_open(struct file *file, struct au_do_open_args *args); ++int au_reopen_nondir(struct file *file); ++struct au_pin; ++int au_ready_to_write(struct file *file, loff_t len, struct au_pin *pin); ++int au_reval_and_lock_fdi(struct file *file, int (*reopen)(struct file *file), ++ int wlock); ++int au_do_flush(struct file *file, fl_owner_t id, ++ int (*flush)(struct file *file, fl_owner_t id)); ++ ++/* poll.c */ ++#ifdef CONFIG_AUFS_POLL ++unsigned int aufs_poll(struct file *file, poll_table *wait); ++#endif ++ ++#ifdef CONFIG_AUFS_BR_HFSPLUS ++/* hfsplus.c */ ++struct file *au_h_open_pre(struct dentry *dentry, aufs_bindex_t bindex, ++ int force_wr); ++void au_h_open_post(struct dentry *dentry, aufs_bindex_t bindex, ++ struct file *h_file); ++#else ++AuStub(struct file *, au_h_open_pre, return NULL, struct dentry *dentry, ++ aufs_bindex_t bindex, int force_wr) ++AuStubVoid(au_h_open_post, struct dentry *dentry, aufs_bindex_t bindex, ++ struct file *h_file); ++#endif ++ ++/* f_op.c */ ++extern const struct file_operations aufs_file_fop; ++int au_do_open_nondir(struct file *file, int flags, struct file *h_file); ++int aufs_release_nondir(struct inode *inode __maybe_unused, struct file *file); ++struct file *au_read_pre(struct file *file, int keep_fi); ++ ++/* finfo.c */ ++void au_hfput(struct au_hfile *hf, struct file *file); ++void au_set_h_fptr(struct file *file, aufs_bindex_t bindex, ++ struct file *h_file); ++ ++void au_update_figen(struct file *file); ++struct au_fidir *au_fidir_alloc(struct super_block *sb); ++int au_fidir_realloc(struct au_finfo *finfo, int nbr); ++ ++void au_fi_init_once(void *_fi); ++void au_finfo_fin(struct file *file); ++int au_finfo_init(struct file *file, struct au_fidir *fidir); ++ ++/* ioctl.c */ ++long aufs_ioctl_nondir(struct file *file, unsigned int cmd, unsigned long arg); ++#ifdef CONFIG_COMPAT ++long aufs_compat_ioctl_dir(struct file *file, unsigned int cmd, ++ unsigned long arg); ++long aufs_compat_ioctl_nondir(struct file *file, unsigned int cmd, ++ unsigned long arg); ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct au_finfo *au_fi(struct file *file) ++{ ++ return file->private_data; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * fi_read_lock, fi_write_lock, ++ * fi_read_unlock, fi_write_unlock, fi_downgrade_lock ++ */ ++AuSimpleRwsemFuncs(fi, struct file *f, &au_fi(f)->fi_rwsem); ++ ++#define FiMustNoWaiters(f) AuRwMustNoWaiters(&au_fi(f)->fi_rwsem) ++#define FiMustAnyLock(f) AuRwMustAnyLock(&au_fi(f)->fi_rwsem) ++#define FiMustWriteLock(f) AuRwMustWriteLock(&au_fi(f)->fi_rwsem) ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* todo: hard/soft set? */ ++static inline aufs_bindex_t au_fbstart(struct file *file) ++{ ++ FiMustAnyLock(file); ++ return au_fi(file)->fi_btop; ++} ++ ++static inline aufs_bindex_t au_fbend_dir(struct file *file) ++{ ++ FiMustAnyLock(file); ++ AuDebugOn(!au_fi(file)->fi_hdir); ++ return au_fi(file)->fi_hdir->fd_bbot; ++} ++ ++static inline struct au_vdir *au_fvdir_cache(struct file *file) ++{ ++ FiMustAnyLock(file); ++ AuDebugOn(!au_fi(file)->fi_hdir); ++ return au_fi(file)->fi_hdir->fd_vdir_cache; ++} ++ ++static inline void au_set_fbstart(struct file *file, aufs_bindex_t bindex) ++{ ++ FiMustWriteLock(file); ++ au_fi(file)->fi_btop = bindex; ++} ++ ++static inline void au_set_fbend_dir(struct file *file, aufs_bindex_t bindex) ++{ ++ FiMustWriteLock(file); ++ AuDebugOn(!au_fi(file)->fi_hdir); ++ au_fi(file)->fi_hdir->fd_bbot = bindex; ++} ++ ++static inline void au_set_fvdir_cache(struct file *file, ++ struct au_vdir *vdir_cache) ++{ ++ FiMustWriteLock(file); ++ AuDebugOn(!au_fi(file)->fi_hdir); ++ au_fi(file)->fi_hdir->fd_vdir_cache = vdir_cache; ++} ++ ++static inline struct file *au_hf_top(struct file *file) ++{ ++ FiMustAnyLock(file); ++ AuDebugOn(au_fi(file)->fi_hdir); ++ return au_fi(file)->fi_htop.hf_file; ++} ++ ++static inline struct file *au_hf_dir(struct file *file, aufs_bindex_t bindex) ++{ ++ FiMustAnyLock(file); ++ AuDebugOn(!au_fi(file)->fi_hdir); ++ return au_fi(file)->fi_hdir->fd_hfile[0 + bindex].hf_file; ++} ++ ++/* todo: memory barrier? */ ++static inline unsigned int au_figen(struct file *f) ++{ ++ return atomic_read(&au_fi(f)->fi_generation); ++} ++ ++static inline void au_set_mmapped(struct file *f) ++{ ++ if (atomic_inc_return(&au_fi(f)->fi_mmapped)) ++ return; ++ pr_warn("fi_mmapped wrapped around\n"); ++ while (!atomic_inc_return(&au_fi(f)->fi_mmapped)) ++ ; ++} ++ ++static inline void au_unset_mmapped(struct file *f) ++{ ++ atomic_dec(&au_fi(f)->fi_mmapped); ++} ++ ++static inline int au_test_mmapped(struct file *f) ++{ ++ return atomic_read(&au_fi(f)->fi_mmapped); ++} ++ ++/* customize vma->vm_file */ ++ ++static inline void au_do_vm_file_reset(struct vm_area_struct *vma, ++ struct file *file) ++{ ++ struct file *f; ++ ++ f = vma->vm_file; ++ get_file(file); ++ vma->vm_file = file; ++ fput(f); ++} ++ ++#ifdef CONFIG_MMU ++#define AuDbgVmRegion(file, vma) do {} while (0) ++ ++static inline void au_vm_file_reset(struct vm_area_struct *vma, ++ struct file *file) ++{ ++ au_do_vm_file_reset(vma, file); ++} ++#else ++#define AuDbgVmRegion(file, vma) \ ++ AuDebugOn((vma)->vm_region && (vma)->vm_region->vm_file != (file)) ++ ++static inline void au_vm_file_reset(struct vm_area_struct *vma, ++ struct file *file) ++{ ++ struct file *f; ++ ++ au_do_vm_file_reset(vma, file); ++ f = vma->vm_region->vm_file; ++ get_file(file); ++ vma->vm_region->vm_file = file; ++ fput(f); ++} ++#endif /* CONFIG_MMU */ ++ ++/* handle vma->vm_prfile */ ++static inline void au_vm_prfile_set(struct vm_area_struct *vma, ++ struct file *file) ++{ ++ get_file(file); ++ vma->vm_prfile = file; ++#ifndef CONFIG_MMU ++ get_file(file); ++ vma->vm_region->vm_prfile = file; ++#endif ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_FILE_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/finfo.c linux-4.1.10/fs/aufs/finfo.c +--- linux-4.1.10.orig/fs/aufs/finfo.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/finfo.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,157 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * file private data ++ */ ++ ++#include "aufs.h" ++ ++void au_hfput(struct au_hfile *hf, struct file *file) ++{ ++ /* todo: direct access f_flags */ ++ if (vfsub_file_flags(file) & __FMODE_EXEC) ++ allow_write_access(hf->hf_file); ++ fput(hf->hf_file); ++ hf->hf_file = NULL; ++ atomic_dec(&hf->hf_br->br_count); ++ hf->hf_br = NULL; ++} ++ ++void au_set_h_fptr(struct file *file, aufs_bindex_t bindex, struct file *val) ++{ ++ struct au_finfo *finfo = au_fi(file); ++ struct au_hfile *hf; ++ struct au_fidir *fidir; ++ ++ fidir = finfo->fi_hdir; ++ if (!fidir) { ++ AuDebugOn(finfo->fi_btop != bindex); ++ hf = &finfo->fi_htop; ++ } else ++ hf = fidir->fd_hfile + bindex; ++ ++ if (hf && hf->hf_file) ++ au_hfput(hf, file); ++ if (val) { ++ FiMustWriteLock(file); ++ AuDebugOn(IS_ERR_OR_NULL(file->f_path.dentry)); ++ hf->hf_file = val; ++ hf->hf_br = au_sbr(file->f_path.dentry->d_sb, bindex); ++ } ++} ++ ++void au_update_figen(struct file *file) ++{ ++ atomic_set(&au_fi(file)->fi_generation, au_digen(file->f_path.dentry)); ++ /* smp_mb(); */ /* atomic_set */ ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_fidir *au_fidir_alloc(struct super_block *sb) ++{ ++ struct au_fidir *fidir; ++ int nbr; ++ ++ nbr = au_sbend(sb) + 1; ++ if (nbr < 2) ++ nbr = 2; /* initial allocate for 2 branches */ ++ fidir = kzalloc(au_fidir_sz(nbr), GFP_NOFS); ++ if (fidir) { ++ fidir->fd_bbot = -1; ++ fidir->fd_nent = nbr; ++ fidir->fd_vdir_cache = NULL; ++ } ++ ++ return fidir; ++} ++ ++int au_fidir_realloc(struct au_finfo *finfo, int nbr) ++{ ++ int err; ++ struct au_fidir *fidir, *p; ++ ++ AuRwMustWriteLock(&finfo->fi_rwsem); ++ fidir = finfo->fi_hdir; ++ AuDebugOn(!fidir); ++ ++ err = -ENOMEM; ++ p = au_kzrealloc(fidir, au_fidir_sz(fidir->fd_nent), au_fidir_sz(nbr), ++ GFP_NOFS); ++ if (p) { ++ p->fd_nent = nbr; ++ finfo->fi_hdir = p; ++ err = 0; ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_finfo_fin(struct file *file) ++{ ++ struct au_finfo *finfo; ++ ++ au_nfiles_dec(file->f_path.dentry->d_sb); ++ ++ finfo = au_fi(file); ++ AuDebugOn(finfo->fi_hdir); ++ AuRwDestroy(&finfo->fi_rwsem); ++ au_cache_free_finfo(finfo); ++} ++ ++void au_fi_init_once(void *_finfo) ++{ ++ struct au_finfo *finfo = _finfo; ++ static struct lock_class_key aufs_fi; ++ ++ au_rw_init(&finfo->fi_rwsem); ++ au_rw_class(&finfo->fi_rwsem, &aufs_fi); ++} ++ ++int au_finfo_init(struct file *file, struct au_fidir *fidir) ++{ ++ int err; ++ struct au_finfo *finfo; ++ struct dentry *dentry; ++ ++ err = -ENOMEM; ++ dentry = file->f_path.dentry; ++ finfo = au_cache_alloc_finfo(); ++ if (unlikely(!finfo)) ++ goto out; ++ ++ err = 0; ++ au_nfiles_inc(dentry->d_sb); ++ /* verbose coding for lock class name */ ++ if (!fidir) ++ au_rw_class(&finfo->fi_rwsem, au_lc_key + AuLcNonDir_FIINFO); ++ else ++ au_rw_class(&finfo->fi_rwsem, au_lc_key + AuLcDir_FIINFO); ++ au_rw_write_lock(&finfo->fi_rwsem); ++ finfo->fi_btop = -1; ++ finfo->fi_hdir = fidir; ++ atomic_set(&finfo->fi_generation, au_digen(dentry)); ++ /* smp_mb(); */ /* atomic_set */ ++ ++ file->private_data = finfo; ++ ++out: ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/f_op.c linux-4.1.10/fs/aufs/f_op.c +--- linux-4.1.10.orig/fs/aufs/f_op.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/f_op.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,738 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * file and vm operations ++ */ ++ ++#include <linux/aio.h> ++#include <linux/fs_stack.h> ++#include <linux/mman.h> ++#include <linux/security.h> ++#include "aufs.h" ++ ++int au_do_open_nondir(struct file *file, int flags, struct file *h_file) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct dentry *dentry; ++ struct au_finfo *finfo; ++ struct inode *h_inode; ++ ++ FiMustWriteLock(file); ++ ++ err = 0; ++ dentry = file->f_path.dentry; ++ AuDebugOn(IS_ERR_OR_NULL(dentry)); ++ finfo = au_fi(file); ++ memset(&finfo->fi_htop, 0, sizeof(finfo->fi_htop)); ++ atomic_set(&finfo->fi_mmapped, 0); ++ bindex = au_dbstart(dentry); ++ if (!h_file) ++ h_file = au_h_open(dentry, bindex, flags, file, /*force_wr*/0); ++ else ++ get_file(h_file); ++ if (IS_ERR(h_file)) ++ err = PTR_ERR(h_file); ++ else { ++ if ((flags & __O_TMPFILE) ++ && !(flags & O_EXCL)) { ++ h_inode = file_inode(h_file); ++ spin_lock(&h_inode->i_lock); ++ h_inode->i_state |= I_LINKABLE; ++ spin_unlock(&h_inode->i_lock); ++ } ++ au_set_fbstart(file, bindex); ++ au_set_h_fptr(file, bindex, h_file); ++ au_update_figen(file); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ } ++ ++ return err; ++} ++ ++static int aufs_open_nondir(struct inode *inode __maybe_unused, ++ struct file *file) ++{ ++ int err; ++ struct super_block *sb; ++ struct au_do_open_args args = { ++ .open = au_do_open_nondir ++ }; ++ ++ AuDbg("%pD, f_flags 0x%x, f_mode 0x%x\n", ++ file, vfsub_file_flags(file), file->f_mode); ++ ++ sb = file->f_path.dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ err = au_do_open(file, &args); ++ si_read_unlock(sb); ++ return err; ++} ++ ++int aufs_release_nondir(struct inode *inode __maybe_unused, struct file *file) ++{ ++ struct au_finfo *finfo; ++ aufs_bindex_t bindex; ++ ++ finfo = au_fi(file); ++ au_sphl_del(&finfo->fi_hlist, ++ &au_sbi(file->f_path.dentry->d_sb)->si_files); ++ bindex = finfo->fi_btop; ++ if (bindex >= 0) ++ au_set_h_fptr(file, bindex, NULL); ++ ++ au_finfo_fin(file); ++ return 0; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_do_flush_nondir(struct file *file, fl_owner_t id) ++{ ++ int err; ++ struct file *h_file; ++ ++ err = 0; ++ h_file = au_hf_top(file); ++ if (h_file) ++ err = vfsub_flush(h_file, id); ++ return err; ++} ++ ++static int aufs_flush_nondir(struct file *file, fl_owner_t id) ++{ ++ return au_do_flush(file, id, au_do_flush_nondir); ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * read and write functions acquire [fdi]_rwsem once, but release before ++ * mmap_sem. This is because to stop a race condition between mmap(2). ++ * Releasing these aufs-rwsem should be safe, no branch-mamagement (by keeping ++ * si_rwsem), no harmful copy-up should happen. Actually copy-up may happen in ++ * read functions after [fdi]_rwsem are released, but it should be harmless. ++ */ ++ ++/* Callers should call au_read_post() or fput() in the end */ ++struct file *au_read_pre(struct file *file, int keep_fi) ++{ ++ struct file *h_file; ++ int err; ++ ++ err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/0); ++ if (!err) { ++ di_read_unlock(file->f_path.dentry, AuLock_IR); ++ h_file = au_hf_top(file); ++ get_file(h_file); ++ if (!keep_fi) ++ fi_read_unlock(file); ++ } else ++ h_file = ERR_PTR(err); ++ ++ return h_file; ++} ++ ++static void au_read_post(struct inode *inode, struct file *h_file) ++{ ++ /* update without lock, I don't think it a problem */ ++ fsstack_copy_attr_atime(inode, file_inode(h_file)); ++ fput(h_file); ++} ++ ++struct au_write_pre { ++ blkcnt_t blks; ++ aufs_bindex_t bstart; ++}; ++ ++/* ++ * return with iinfo is write-locked ++ * callers should call au_write_post() or iinfo_write_unlock() + fput() in the ++ * end ++ */ ++static struct file *au_write_pre(struct file *file, int do_ready, ++ struct au_write_pre *wpre) ++{ ++ struct file *h_file; ++ struct dentry *dentry; ++ int err; ++ struct au_pin pin; ++ ++ err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/1); ++ h_file = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ dentry = file->f_path.dentry; ++ if (do_ready) { ++ err = au_ready_to_write(file, -1, &pin); ++ if (unlikely(err)) { ++ h_file = ERR_PTR(err); ++ di_write_unlock(dentry); ++ goto out_fi; ++ } ++ } ++ ++ di_downgrade_lock(dentry, /*flags*/0); ++ if (wpre) ++ wpre->bstart = au_fbstart(file); ++ h_file = au_hf_top(file); ++ get_file(h_file); ++ if (wpre) ++ wpre->blks = file_inode(h_file)->i_blocks; ++ if (do_ready) ++ au_unpin(&pin); ++ di_read_unlock(dentry, /*flags*/0); ++ ++out_fi: ++ fi_write_unlock(file); ++out: ++ return h_file; ++} ++ ++static void au_write_post(struct inode *inode, struct file *h_file, ++ struct au_write_pre *wpre, ssize_t written) ++{ ++ struct inode *h_inode; ++ ++ au_cpup_attr_timesizes(inode); ++ AuDebugOn(au_ibstart(inode) != wpre->bstart); ++ h_inode = file_inode(h_file); ++ inode->i_mode = h_inode->i_mode; ++ ii_write_unlock(inode); ++ fput(h_file); ++ ++ /* AuDbg("blks %llu, %llu\n", (u64)blks, (u64)h_inode->i_blocks); */ ++ if (written > 0) ++ au_fhsm_wrote(inode->i_sb, wpre->bstart, ++ /*force*/h_inode->i_blocks > wpre->blks); ++} ++ ++static ssize_t aufs_read(struct file *file, char __user *buf, size_t count, ++ loff_t *ppos) ++{ ++ ssize_t err; ++ struct inode *inode; ++ struct file *h_file; ++ struct super_block *sb; ++ ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ++ h_file = au_read_pre(file, /*keep_fi*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ /* filedata may be obsoleted by concurrent copyup, but no problem */ ++ err = vfsub_read_u(h_file, buf, count, ppos); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ au_read_post(inode, h_file); ++ ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ++ * todo: very ugly ++ * it locks both of i_mutex and si_rwsem for read in safe. ++ * if the plink maintenance mode continues forever (that is the problem), ++ * may loop forever. ++ */ ++static void au_mtx_and_read_lock(struct inode *inode) ++{ ++ int err; ++ struct super_block *sb = inode->i_sb; ++ ++ while (1) { ++ mutex_lock(&inode->i_mutex); ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (!err) ++ break; ++ mutex_unlock(&inode->i_mutex); ++ si_read_lock(sb, AuLock_NOPLMW); ++ si_read_unlock(sb); ++ } ++} ++ ++static ssize_t aufs_write(struct file *file, const char __user *ubuf, ++ size_t count, loff_t *ppos) ++{ ++ ssize_t err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *h_file; ++ char __user *buf = (char __user *)ubuf; ++ ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = vfsub_write_u(h_file, buf, count, ppos); ++ au_write_post(inode, h_file, &wpre, err); ++ ++out: ++ si_read_unlock(inode->i_sb); ++ mutex_unlock(&inode->i_mutex); ++ return err; ++} ++ ++static ssize_t au_do_iter(struct file *h_file, int rw, struct kiocb *kio, ++ struct iov_iter *iov_iter) ++{ ++ ssize_t err; ++ struct file *file; ++ ssize_t (*iter)(struct kiocb *, struct iov_iter *); ++ ++ err = security_file_permission(h_file, rw); ++ if (unlikely(err)) ++ goto out; ++ ++ err = -ENOSYS; ++ iter = NULL; ++ if (rw == MAY_READ) ++ iter = h_file->f_op->read_iter; ++ else if (rw == MAY_WRITE) ++ iter = h_file->f_op->write_iter; ++ ++ file = kio->ki_filp; ++ kio->ki_filp = h_file; ++ if (iter) { ++ lockdep_off(); ++ err = iter(kio, iov_iter); ++ lockdep_on(); ++ } else ++ /* currently there is no such fs */ ++ WARN_ON_ONCE(1); ++ kio->ki_filp = file; ++ ++out: ++ return err; ++} ++ ++static ssize_t aufs_read_iter(struct kiocb *kio, struct iov_iter *iov_iter) ++{ ++ ssize_t err; ++ struct file *file, *h_file; ++ struct inode *inode; ++ struct super_block *sb; ++ ++ file = kio->ki_filp; ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ++ h_file = au_read_pre(file, /*keep_fi*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = au_do_iter(h_file, MAY_READ, kio, iov_iter); ++ /* todo: necessary? */ ++ /* file->f_ra = h_file->f_ra; */ ++ au_read_post(inode, h_file); ++ ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++static ssize_t aufs_write_iter(struct kiocb *kio, struct iov_iter *iov_iter) ++{ ++ ssize_t err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *file, *h_file; ++ ++ file = kio->ki_filp; ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = au_do_iter(h_file, MAY_WRITE, kio, iov_iter); ++ au_write_post(inode, h_file, &wpre, err); ++ ++out: ++ si_read_unlock(inode->i_sb); ++ mutex_unlock(&inode->i_mutex); ++ return err; ++} ++ ++static ssize_t aufs_splice_read(struct file *file, loff_t *ppos, ++ struct pipe_inode_info *pipe, size_t len, ++ unsigned int flags) ++{ ++ ssize_t err; ++ struct file *h_file; ++ struct inode *inode; ++ struct super_block *sb; ++ ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ++ h_file = au_read_pre(file, /*keep_fi*/1); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ if (au_test_loopback_kthread()) { ++ au_warn_loopback(h_file->f_path.dentry->d_sb); ++ if (file->f_mapping != h_file->f_mapping) { ++ file->f_mapping = h_file->f_mapping; ++ smp_mb(); /* unnecessary? */ ++ } ++ } ++ fi_read_unlock(file); ++ ++ err = vfsub_splice_to(h_file, ppos, pipe, len, flags); ++ /* todo: necessasry? */ ++ /* file->f_ra = h_file->f_ra; */ ++ au_read_post(inode, h_file); ++ ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++static ssize_t ++aufs_splice_write(struct pipe_inode_info *pipe, struct file *file, loff_t *ppos, ++ size_t len, unsigned int flags) ++{ ++ ssize_t err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *h_file; ++ ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = vfsub_splice_from(pipe, h_file, ppos, len, flags); ++ au_write_post(inode, h_file, &wpre, err); ++ ++out: ++ si_read_unlock(inode->i_sb); ++ mutex_unlock(&inode->i_mutex); ++ return err; ++} ++ ++static long aufs_fallocate(struct file *file, int mode, loff_t offset, ++ loff_t len) ++{ ++ long err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *h_file; ++ ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_fallocate(h_file, mode, offset, len); ++ lockdep_on(); ++ au_write_post(inode, h_file, &wpre, /*written*/1); ++ ++out: ++ si_read_unlock(inode->i_sb); ++ mutex_unlock(&inode->i_mutex); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * The locking order around current->mmap_sem. ++ * - in most and regular cases ++ * file I/O syscall -- aufs_read() or something ++ * -- si_rwsem for read -- mmap_sem ++ * (Note that [fdi]i_rwsem are released before mmap_sem). ++ * - in mmap case ++ * mmap(2) -- mmap_sem -- aufs_mmap() -- si_rwsem for read -- [fdi]i_rwsem ++ * This AB-BA order is definitly bad, but is not a problem since "si_rwsem for ++ * read" allows muliple processes to acquire it and [fdi]i_rwsem are not held in ++ * file I/O. Aufs needs to stop lockdep in aufs_mmap() though. ++ * It means that when aufs acquires si_rwsem for write, the process should never ++ * acquire mmap_sem. ++ * ++ * Actually aufs_iterate() holds [fdi]i_rwsem before mmap_sem, but this is not a ++ * problem either since any directory is not able to be mmap-ed. ++ * The similar scenario is applied to aufs_readlink() too. ++ */ ++ ++#if 0 /* stop calling security_file_mmap() */ ++/* cf. linux/include/linux/mman.h: calc_vm_prot_bits() */ ++#define AuConv_VM_PROT(f, b) _calc_vm_trans(f, VM_##b, PROT_##b) ++ ++static unsigned long au_arch_prot_conv(unsigned long flags) ++{ ++ /* currently ppc64 only */ ++#ifdef CONFIG_PPC64 ++ /* cf. linux/arch/powerpc/include/asm/mman.h */ ++ AuDebugOn(arch_calc_vm_prot_bits(-1) != VM_SAO); ++ return AuConv_VM_PROT(flags, SAO); ++#else ++ AuDebugOn(arch_calc_vm_prot_bits(-1)); ++ return 0; ++#endif ++} ++ ++static unsigned long au_prot_conv(unsigned long flags) ++{ ++ return AuConv_VM_PROT(flags, READ) ++ | AuConv_VM_PROT(flags, WRITE) ++ | AuConv_VM_PROT(flags, EXEC) ++ | au_arch_prot_conv(flags); ++} ++ ++/* cf. linux/include/linux/mman.h: calc_vm_flag_bits() */ ++#define AuConv_VM_MAP(f, b) _calc_vm_trans(f, VM_##b, MAP_##b) ++ ++static unsigned long au_flag_conv(unsigned long flags) ++{ ++ return AuConv_VM_MAP(flags, GROWSDOWN) ++ | AuConv_VM_MAP(flags, DENYWRITE) ++ | AuConv_VM_MAP(flags, LOCKED); ++} ++#endif ++ ++static int aufs_mmap(struct file *file, struct vm_area_struct *vma) ++{ ++ int err; ++ const unsigned char wlock ++ = (file->f_mode & FMODE_WRITE) && (vma->vm_flags & VM_SHARED); ++ struct super_block *sb; ++ struct file *h_file; ++ struct inode *inode; ++ ++ AuDbgVmRegion(file, vma); ++ ++ inode = file_inode(file); ++ sb = inode->i_sb; ++ lockdep_off(); ++ si_read_lock(sb, AuLock_NOPLMW); ++ ++ h_file = au_write_pre(file, wlock, /*wpre*/NULL); ++ lockdep_on(); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ err = 0; ++ au_set_mmapped(file); ++ au_vm_file_reset(vma, h_file); ++ /* ++ * we cannot call security_mmap_file() here since it may acquire ++ * mmap_sem or i_mutex. ++ * ++ * err = security_mmap_file(h_file, au_prot_conv(vma->vm_flags), ++ * au_flag_conv(vma->vm_flags)); ++ */ ++ if (!err) ++ err = h_file->f_op->mmap(h_file, vma); ++ if (!err) { ++ au_vm_prfile_set(vma, file); ++ fsstack_copy_attr_atime(inode, file_inode(h_file)); ++ goto out_fput; /* success */ ++ } ++ au_unset_mmapped(file); ++ au_vm_file_reset(vma, file); ++ ++out_fput: ++ lockdep_off(); ++ ii_write_unlock(inode); ++ lockdep_on(); ++ fput(h_file); ++out: ++ lockdep_off(); ++ si_read_unlock(sb); ++ lockdep_on(); ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_fsync_nondir(struct file *file, loff_t start, loff_t end, ++ int datasync) ++{ ++ int err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *h_file; ++ ++ err = 0; /* -EBADF; */ /* posix? */ ++ if (unlikely(!(file->f_mode & FMODE_WRITE))) ++ goto out; ++ ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out_unlock; ++ ++ err = vfsub_fsync(h_file, &h_file->f_path, datasync); ++ au_write_post(inode, h_file, &wpre, /*written*/0); ++ ++out_unlock: ++ si_read_unlock(inode->i_sb); ++ mutex_unlock(&inode->i_mutex); ++out: ++ return err; ++} ++ ++/* no one supports this operation, currently */ ++#if 0 ++static int aufs_aio_fsync_nondir(struct kiocb *kio, int datasync) ++{ ++ int err; ++ struct au_write_pre wpre; ++ struct inode *inode; ++ struct file *file, *h_file; ++ ++ err = 0; /* -EBADF; */ /* posix? */ ++ if (unlikely(!(file->f_mode & FMODE_WRITE))) ++ goto out; ++ ++ file = kio->ki_filp; ++ inode = file_inode(file); ++ au_mtx_and_read_lock(inode); ++ ++ h_file = au_write_pre(file, /*do_ready*/1, &wpre); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out_unlock; ++ ++ err = -ENOSYS; ++ h_file = au_hf_top(file); ++ if (h_file->f_op->aio_fsync) { ++ struct mutex *h_mtx; ++ ++ h_mtx = &file_inode(h_file)->i_mutex; ++ if (!is_sync_kiocb(kio)) { ++ get_file(h_file); ++ fput(file); ++ } ++ kio->ki_filp = h_file; ++ err = h_file->f_op->aio_fsync(kio, datasync); ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD); ++ if (!err) ++ vfsub_update_h_iattr(&h_file->f_path, /*did*/NULL); ++ /*ignore*/ ++ mutex_unlock(h_mtx); ++ } ++ au_write_post(inode, h_file, &wpre, /*written*/0); ++ ++out_unlock: ++ si_read_unlock(inode->sb); ++ mutex_unlock(&inode->i_mutex); ++out: ++ return err; ++} ++#endif ++ ++static int aufs_fasync(int fd, struct file *file, int flag) ++{ ++ int err; ++ struct file *h_file; ++ struct super_block *sb; ++ ++ sb = file->f_path.dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ++ h_file = au_read_pre(file, /*keep_fi*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ if (h_file->f_op->fasync) ++ err = h_file->f_op->fasync(fd, h_file, flag); ++ fput(h_file); /* instead of au_read_post() */ ++ ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* no one supports this operation, currently */ ++#if 0 ++static ssize_t aufs_sendpage(struct file *file, struct page *page, int offset, ++ size_t len, loff_t *pos, int more) ++{ ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++const struct file_operations aufs_file_fop = { ++ .owner = THIS_MODULE, ++ ++ .llseek = default_llseek, ++ ++ .read = aufs_read, ++ .write = aufs_write, ++ .read_iter = aufs_read_iter, ++ .write_iter = aufs_write_iter, ++ ++#ifdef CONFIG_AUFS_POLL ++ .poll = aufs_poll, ++#endif ++ .unlocked_ioctl = aufs_ioctl_nondir, ++#ifdef CONFIG_COMPAT ++ .compat_ioctl = aufs_compat_ioctl_nondir, ++#endif ++ .mmap = aufs_mmap, ++ .open = aufs_open_nondir, ++ .flush = aufs_flush_nondir, ++ .release = aufs_release_nondir, ++ .fsync = aufs_fsync_nondir, ++ /* .aio_fsync = aufs_aio_fsync_nondir, */ ++ .fasync = aufs_fasync, ++ /* .sendpage = aufs_sendpage, */ ++ .splice_write = aufs_splice_write, ++ .splice_read = aufs_splice_read, ++#if 0 ++ .aio_splice_write = aufs_aio_splice_write, ++ .aio_splice_read = aufs_aio_splice_read, ++#endif ++ .fallocate = aufs_fallocate ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/fstype.h linux-4.1.10/fs/aufs/fstype.h +--- linux-4.1.10.orig/fs/aufs/fstype.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/fstype.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,400 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * judging filesystem type ++ */ ++ ++#ifndef __AUFS_FSTYPE_H__ ++#define __AUFS_FSTYPE_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fs.h> ++#include <linux/magic.h> ++#include <linux/romfs_fs.h> ++#include <linux/nfs_fs.h> ++ ++static inline int au_test_aufs(struct super_block *sb) ++{ ++ return sb->s_magic == AUFS_SUPER_MAGIC; ++} ++ ++static inline const char *au_sbtype(struct super_block *sb) ++{ ++ return sb->s_type->name; ++} ++ ++static inline int au_test_iso9660(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_ISO9660_FS) || defined(CONFIG_ISO9660_FS_MODULE) ++ return sb->s_magic == ISOFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_romfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_ROMFS_FS) || defined(CONFIG_ROMFS_FS_MODULE) ++ return sb->s_magic == ROMFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_cramfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_CRAMFS) || defined(CONFIG_CRAMFS_MODULE) ++ return sb->s_magic == CRAMFS_MAGIC; ++#endif ++ return 0; ++} ++ ++static inline int au_test_nfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_NFS_FS) || defined(CONFIG_NFS_FS_MODULE) ++ return sb->s_magic == NFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_fuse(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_FUSE_FS) || defined(CONFIG_FUSE_FS_MODULE) ++ return sb->s_magic == FUSE_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_xfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_XFS_FS) || defined(CONFIG_XFS_FS_MODULE) ++ return sb->s_magic == XFS_SB_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_tmpfs(struct super_block *sb __maybe_unused) ++{ ++#ifdef CONFIG_TMPFS ++ return sb->s_magic == TMPFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_ecryptfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_ECRYPT_FS) || defined(CONFIG_ECRYPT_FS_MODULE) ++ return !strcmp(au_sbtype(sb), "ecryptfs"); ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_ramfs(struct super_block *sb) ++{ ++ return sb->s_magic == RAMFS_MAGIC; ++} ++ ++static inline int au_test_ubifs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_UBIFS_FS) || defined(CONFIG_UBIFS_FS_MODULE) ++ return sb->s_magic == UBIFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_procfs(struct super_block *sb __maybe_unused) ++{ ++#ifdef CONFIG_PROC_FS ++ return sb->s_magic == PROC_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_sysfs(struct super_block *sb __maybe_unused) ++{ ++#ifdef CONFIG_SYSFS ++ return sb->s_magic == SYSFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_configfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_CONFIGFS_FS) || defined(CONFIG_CONFIGFS_FS_MODULE) ++ return sb->s_magic == CONFIGFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_minix(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_MINIX_FS) || defined(CONFIG_MINIX_FS_MODULE) ++ return sb->s_magic == MINIX3_SUPER_MAGIC ++ || sb->s_magic == MINIX2_SUPER_MAGIC ++ || sb->s_magic == MINIX2_SUPER_MAGIC2 ++ || sb->s_magic == MINIX_SUPER_MAGIC ++ || sb->s_magic == MINIX_SUPER_MAGIC2; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_fat(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_FAT_FS) || defined(CONFIG_FAT_FS_MODULE) ++ return sb->s_magic == MSDOS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_msdos(struct super_block *sb) ++{ ++ return au_test_fat(sb); ++} ++ ++static inline int au_test_vfat(struct super_block *sb) ++{ ++ return au_test_fat(sb); ++} ++ ++static inline int au_test_securityfs(struct super_block *sb __maybe_unused) ++{ ++#ifdef CONFIG_SECURITYFS ++ return sb->s_magic == SECURITYFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_squashfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_SQUASHFS) || defined(CONFIG_SQUASHFS_MODULE) ++ return sb->s_magic == SQUASHFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_btrfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_BTRFS_FS) || defined(CONFIG_BTRFS_FS_MODULE) ++ return sb->s_magic == BTRFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_xenfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_XENFS) || defined(CONFIG_XENFS_MODULE) ++ return sb->s_magic == XENFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_debugfs(struct super_block *sb __maybe_unused) ++{ ++#ifdef CONFIG_DEBUG_FS ++ return sb->s_magic == DEBUGFS_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_nilfs(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_NILFS) || defined(CONFIG_NILFS_MODULE) ++ return sb->s_magic == NILFS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++static inline int au_test_hfsplus(struct super_block *sb __maybe_unused) ++{ ++#if defined(CONFIG_HFSPLUS_FS) || defined(CONFIG_HFSPLUS_FS_MODULE) ++ return sb->s_magic == HFSPLUS_SUPER_MAGIC; ++#else ++ return 0; ++#endif ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * they can't be an aufs branch. ++ */ ++static inline int au_test_fs_unsuppoted(struct super_block *sb) ++{ ++ return ++#ifndef CONFIG_AUFS_BR_RAMFS ++ au_test_ramfs(sb) || ++#endif ++ au_test_procfs(sb) ++ || au_test_sysfs(sb) ++ || au_test_configfs(sb) ++ || au_test_debugfs(sb) ++ || au_test_securityfs(sb) ++ || au_test_xenfs(sb) ++ || au_test_ecryptfs(sb) ++ /* || !strcmp(au_sbtype(sb), "unionfs") */ ++ || au_test_aufs(sb); /* will be supported in next version */ ++} ++ ++static inline int au_test_fs_remote(struct super_block *sb) ++{ ++ return !au_test_tmpfs(sb) ++#ifdef CONFIG_AUFS_BR_RAMFS ++ && !au_test_ramfs(sb) ++#endif ++ && !(sb->s_type->fs_flags & FS_REQUIRES_DEV); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * Note: these functions (below) are created after reading ->getattr() in all ++ * filesystems under linux/fs. it means we have to do so in every update... ++ */ ++ ++/* ++ * some filesystems require getattr to refresh the inode attributes before ++ * referencing. ++ * in most cases, we can rely on the inode attribute in NFS (or every remote fs) ++ * and leave the work for d_revalidate() ++ */ ++static inline int au_test_fs_refresh_iattr(struct super_block *sb) ++{ ++ return au_test_nfs(sb) ++ || au_test_fuse(sb) ++ /* || au_test_btrfs(sb) */ /* untested */ ++ ; ++} ++ ++/* ++ * filesystems which don't maintain i_size or i_blocks. ++ */ ++static inline int au_test_fs_bad_iattr_size(struct super_block *sb) ++{ ++ return au_test_xfs(sb) ++ || au_test_btrfs(sb) ++ || au_test_ubifs(sb) ++ || au_test_hfsplus(sb) /* maintained, but incorrect */ ++ /* || au_test_minix(sb) */ /* untested */ ++ ; ++} ++ ++/* ++ * filesystems which don't store the correct value in some of their inode ++ * attributes. ++ */ ++static inline int au_test_fs_bad_iattr(struct super_block *sb) ++{ ++ return au_test_fs_bad_iattr_size(sb) ++ || au_test_fat(sb) ++ || au_test_msdos(sb) ++ || au_test_vfat(sb); ++} ++ ++/* they don't check i_nlink in link(2) */ ++static inline int au_test_fs_no_limit_nlink(struct super_block *sb) ++{ ++ return au_test_tmpfs(sb) ++#ifdef CONFIG_AUFS_BR_RAMFS ++ || au_test_ramfs(sb) ++#endif ++ || au_test_ubifs(sb) ++ || au_test_hfsplus(sb); ++} ++ ++/* ++ * filesystems which sets S_NOATIME and S_NOCMTIME. ++ */ ++static inline int au_test_fs_notime(struct super_block *sb) ++{ ++ return au_test_nfs(sb) ++ || au_test_fuse(sb) ++ || au_test_ubifs(sb) ++ ; ++} ++ ++/* temporary support for i#1 in cramfs */ ++static inline int au_test_fs_unique_ino(struct inode *inode) ++{ ++ if (au_test_cramfs(inode->i_sb)) ++ return inode->i_ino != 1; ++ return 1; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * the filesystem where the xino files placed must support i/o after unlink and ++ * maintain i_size and i_blocks. ++ */ ++static inline int au_test_fs_bad_xino(struct super_block *sb) ++{ ++ return au_test_fs_remote(sb) ++ || au_test_fs_bad_iattr_size(sb) ++ /* don't want unnecessary work for xino */ ++ || au_test_aufs(sb) ++ || au_test_ecryptfs(sb) ++ || au_test_nilfs(sb); ++} ++ ++static inline int au_test_fs_trunc_xino(struct super_block *sb) ++{ ++ return au_test_tmpfs(sb) ++ || au_test_ramfs(sb); ++} ++ ++/* ++ * test if the @sb is real-readonly. ++ */ ++static inline int au_test_fs_rr(struct super_block *sb) ++{ ++ return au_test_squashfs(sb) ++ || au_test_iso9660(sb) ++ || au_test_cramfs(sb) ++ || au_test_romfs(sb); ++} ++ ++/* ++ * test if the @inode is nfs with 'noacl' option ++ * NFS always sets MS_POSIXACL regardless its mount option 'noacl.' ++ */ ++static inline int au_test_nfs_noacl(struct inode *inode) ++{ ++ return au_test_nfs(inode->i_sb) ++ /* && IS_POSIXACL(inode) */ ++ && !nfs_server_capable(inode, NFS_CAP_ACLS); ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_FSTYPE_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/hfsnotify.c linux-4.1.10/fs/aufs/hfsnotify.c +--- linux-4.1.10.orig/fs/aufs/hfsnotify.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/hfsnotify.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,288 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * fsnotify for the lower directories ++ */ ++ ++#include "aufs.h" ++ ++/* FS_IN_IGNORED is unnecessary */ ++static const __u32 AuHfsnMask = (FS_MOVED_TO | FS_MOVED_FROM | FS_DELETE ++ | FS_CREATE | FS_EVENT_ON_CHILD); ++static DECLARE_WAIT_QUEUE_HEAD(au_hfsn_wq); ++static __cacheline_aligned_in_smp atomic64_t au_hfsn_ifree = ATOMIC64_INIT(0); ++ ++static void au_hfsn_free_mark(struct fsnotify_mark *mark) ++{ ++ struct au_hnotify *hn = container_of(mark, struct au_hnotify, ++ hn_mark); ++ AuDbg("here\n"); ++ au_cache_free_hnotify(hn); ++ smp_mb__before_atomic(); ++ if (atomic64_dec_and_test(&au_hfsn_ifree)) ++ wake_up(&au_hfsn_wq); ++} ++ ++static int au_hfsn_alloc(struct au_hinode *hinode) ++{ ++ int err; ++ struct au_hnotify *hn; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct fsnotify_mark *mark; ++ aufs_bindex_t bindex; ++ ++ hn = hinode->hi_notify; ++ sb = hn->hn_aufs_inode->i_sb; ++ bindex = au_br_index(sb, hinode->hi_id); ++ br = au_sbr(sb, bindex); ++ AuDebugOn(!br->br_hfsn); ++ ++ mark = &hn->hn_mark; ++ fsnotify_init_mark(mark, au_hfsn_free_mark); ++ mark->mask = AuHfsnMask; ++ /* ++ * by udba rename or rmdir, aufs assign a new inode to the known ++ * h_inode, so specify 1 to allow dups. ++ */ ++ lockdep_off(); ++ err = fsnotify_add_mark(mark, br->br_hfsn->hfsn_group, hinode->hi_inode, ++ /*mnt*/NULL, /*allow_dups*/1); ++ /* even if err */ ++ fsnotify_put_mark(mark); ++ lockdep_on(); ++ ++ return err; ++} ++ ++static int au_hfsn_free(struct au_hinode *hinode, struct au_hnotify *hn) ++{ ++ struct fsnotify_mark *mark; ++ unsigned long long ull; ++ struct fsnotify_group *group; ++ ++ ull = atomic64_inc_return(&au_hfsn_ifree); ++ BUG_ON(!ull); ++ ++ mark = &hn->hn_mark; ++ spin_lock(&mark->lock); ++ group = mark->group; ++ fsnotify_get_group(group); ++ spin_unlock(&mark->lock); ++ lockdep_off(); ++ fsnotify_destroy_mark(mark, group); ++ fsnotify_put_group(group); ++ lockdep_on(); ++ ++ /* free hn by myself */ ++ return 0; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_hfsn_ctl(struct au_hinode *hinode, int do_set) ++{ ++ struct fsnotify_mark *mark; ++ ++ mark = &hinode->hi_notify->hn_mark; ++ spin_lock(&mark->lock); ++ if (do_set) { ++ AuDebugOn(mark->mask & AuHfsnMask); ++ mark->mask |= AuHfsnMask; ++ } else { ++ AuDebugOn(!(mark->mask & AuHfsnMask)); ++ mark->mask &= ~AuHfsnMask; ++ } ++ spin_unlock(&mark->lock); ++ /* fsnotify_recalc_inode_mask(hinode->hi_inode); */ ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* #define AuDbgHnotify */ ++#ifdef AuDbgHnotify ++static char *au_hfsn_name(u32 mask) ++{ ++#ifdef CONFIG_AUFS_DEBUG ++#define test_ret(flag) \ ++ do { \ ++ if (mask & flag) \ ++ return #flag; \ ++ } while (0) ++ test_ret(FS_ACCESS); ++ test_ret(FS_MODIFY); ++ test_ret(FS_ATTRIB); ++ test_ret(FS_CLOSE_WRITE); ++ test_ret(FS_CLOSE_NOWRITE); ++ test_ret(FS_OPEN); ++ test_ret(FS_MOVED_FROM); ++ test_ret(FS_MOVED_TO); ++ test_ret(FS_CREATE); ++ test_ret(FS_DELETE); ++ test_ret(FS_DELETE_SELF); ++ test_ret(FS_MOVE_SELF); ++ test_ret(FS_UNMOUNT); ++ test_ret(FS_Q_OVERFLOW); ++ test_ret(FS_IN_IGNORED); ++ test_ret(FS_ISDIR); ++ test_ret(FS_IN_ONESHOT); ++ test_ret(FS_EVENT_ON_CHILD); ++ return ""; ++#undef test_ret ++#else ++ return "??"; ++#endif ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_hfsn_free_group(struct fsnotify_group *group) ++{ ++ struct au_br_hfsnotify *hfsn = group->private; ++ ++ AuDbg("here\n"); ++ kfree(hfsn); ++} ++ ++static int au_hfsn_handle_event(struct fsnotify_group *group, ++ struct inode *inode, ++ struct fsnotify_mark *inode_mark, ++ struct fsnotify_mark *vfsmount_mark, ++ u32 mask, void *data, int data_type, ++ const unsigned char *file_name, u32 cookie) ++{ ++ int err; ++ struct au_hnotify *hnotify; ++ struct inode *h_dir, *h_inode; ++ struct qstr h_child_qstr = QSTR_INIT(file_name, strlen(file_name)); ++ ++ AuDebugOn(data_type != FSNOTIFY_EVENT_INODE); ++ ++ err = 0; ++ /* if FS_UNMOUNT happens, there must be another bug */ ++ AuDebugOn(mask & FS_UNMOUNT); ++ if (mask & (FS_IN_IGNORED | FS_UNMOUNT)) ++ goto out; ++ ++ h_dir = inode; ++ h_inode = NULL; ++#ifdef AuDbgHnotify ++ au_debug_on(); ++ if (1 || h_child_qstr.len != sizeof(AUFS_XINO_FNAME) - 1 ++ || strncmp(h_child_qstr.name, AUFS_XINO_FNAME, h_child_qstr.len)) { ++ AuDbg("i%lu, mask 0x%x %s, hcname %.*s, hi%lu\n", ++ h_dir->i_ino, mask, au_hfsn_name(mask), ++ AuLNPair(&h_child_qstr), h_inode ? h_inode->i_ino : 0); ++ /* WARN_ON(1); */ ++ } ++ au_debug_off(); ++#endif ++ ++ AuDebugOn(!inode_mark); ++ hnotify = container_of(inode_mark, struct au_hnotify, hn_mark); ++ err = au_hnotify(h_dir, hnotify, mask, &h_child_qstr, h_inode); ++ ++out: ++ return err; ++} ++ ++static struct fsnotify_ops au_hfsn_ops = { ++ .handle_event = au_hfsn_handle_event, ++ .free_group_priv = au_hfsn_free_group ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_hfsn_fin_br(struct au_branch *br) ++{ ++ struct au_br_hfsnotify *hfsn; ++ ++ hfsn = br->br_hfsn; ++ if (hfsn) { ++ lockdep_off(); ++ fsnotify_put_group(hfsn->hfsn_group); ++ lockdep_on(); ++ } ++} ++ ++static int au_hfsn_init_br(struct au_branch *br, int perm) ++{ ++ int err; ++ struct fsnotify_group *group; ++ struct au_br_hfsnotify *hfsn; ++ ++ err = 0; ++ br->br_hfsn = NULL; ++ if (!au_br_hnotifyable(perm)) ++ goto out; ++ ++ err = -ENOMEM; ++ hfsn = kmalloc(sizeof(*hfsn), GFP_NOFS); ++ if (unlikely(!hfsn)) ++ goto out; ++ ++ err = 0; ++ group = fsnotify_alloc_group(&au_hfsn_ops); ++ if (IS_ERR(group)) { ++ err = PTR_ERR(group); ++ pr_err("fsnotify_alloc_group() failed, %d\n", err); ++ goto out_hfsn; ++ } ++ ++ group->private = hfsn; ++ hfsn->hfsn_group = group; ++ br->br_hfsn = hfsn; ++ goto out; /* success */ ++ ++out_hfsn: ++ kfree(hfsn); ++out: ++ return err; ++} ++ ++static int au_hfsn_reset_br(unsigned int udba, struct au_branch *br, int perm) ++{ ++ int err; ++ ++ err = 0; ++ if (!br->br_hfsn) ++ err = au_hfsn_init_br(br, perm); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_hfsn_fin(void) ++{ ++ AuDbg("au_hfsn_ifree %lld\n", (long long)atomic64_read(&au_hfsn_ifree)); ++ wait_event(au_hfsn_wq, !atomic64_read(&au_hfsn_ifree)); ++} ++ ++const struct au_hnotify_op au_hnotify_op = { ++ .ctl = au_hfsn_ctl, ++ .alloc = au_hfsn_alloc, ++ .free = au_hfsn_free, ++ ++ .fin = au_hfsn_fin, ++ ++ .reset_br = au_hfsn_reset_br, ++ .fin_br = au_hfsn_fin_br, ++ .init_br = au_hfsn_init_br ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/hfsplus.c linux-4.1.10/fs/aufs/hfsplus.c +--- linux-4.1.10.orig/fs/aufs/hfsplus.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/hfsplus.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,56 @@ ++/* ++ * Copyright (C) 2010-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * special support for filesystems which aqucires an inode mutex ++ * at final closing a file, eg, hfsplus. ++ * ++ * This trick is very simple and stupid, just to open the file before really ++ * neceeary open to tell hfsplus that this is not the final closing. ++ * The caller should call au_h_open_pre() after acquiring the inode mutex, ++ * and au_h_open_post() after releasing it. ++ */ ++ ++#include "aufs.h" ++ ++struct file *au_h_open_pre(struct dentry *dentry, aufs_bindex_t bindex, ++ int force_wr) ++{ ++ struct file *h_file; ++ struct dentry *h_dentry; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ AuDebugOn(!h_dentry); ++ AuDebugOn(d_is_negative(h_dentry)); ++ ++ h_file = NULL; ++ if (au_test_hfsplus(h_dentry->d_sb) ++ && d_is_reg(h_dentry)) ++ h_file = au_h_open(dentry, bindex, ++ O_RDONLY | O_NOATIME | O_LARGEFILE, ++ /*file*/NULL, force_wr); ++ return h_file; ++} ++ ++void au_h_open_post(struct dentry *dentry, aufs_bindex_t bindex, ++ struct file *h_file) ++{ ++ if (h_file) { ++ fput(h_file); ++ au_sbr_put(dentry->d_sb, bindex); ++ } ++} +diff -Nur linux-4.1.10.orig/fs/aufs/hnotify.c linux-4.1.10/fs/aufs/hnotify.c +--- linux-4.1.10.orig/fs/aufs/hnotify.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/hnotify.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,710 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * abstraction to notify the direct changes on lower directories ++ */ ++ ++#include "aufs.h" ++ ++int au_hn_alloc(struct au_hinode *hinode, struct inode *inode) ++{ ++ int err; ++ struct au_hnotify *hn; ++ ++ err = -ENOMEM; ++ hn = au_cache_alloc_hnotify(); ++ if (hn) { ++ hn->hn_aufs_inode = inode; ++ hinode->hi_notify = hn; ++ err = au_hnotify_op.alloc(hinode); ++ AuTraceErr(err); ++ if (unlikely(err)) { ++ hinode->hi_notify = NULL; ++ au_cache_free_hnotify(hn); ++ /* ++ * The upper dir was removed by udba, but the same named ++ * dir left. In this case, aufs assignes a new inode ++ * number and set the monitor again. ++ * For the lower dir, the old monitnor is still left. ++ */ ++ if (err == -EEXIST) ++ err = 0; ++ } ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++void au_hn_free(struct au_hinode *hinode) ++{ ++ struct au_hnotify *hn; ++ ++ hn = hinode->hi_notify; ++ if (hn) { ++ hinode->hi_notify = NULL; ++ if (au_hnotify_op.free(hinode, hn)) ++ au_cache_free_hnotify(hn); ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_hn_ctl(struct au_hinode *hinode, int do_set) ++{ ++ if (hinode->hi_notify) ++ au_hnotify_op.ctl(hinode, do_set); ++} ++ ++void au_hn_reset(struct inode *inode, unsigned int flags) ++{ ++ aufs_bindex_t bindex, bend; ++ struct inode *hi; ++ struct dentry *iwhdentry; ++ ++ bend = au_ibend(inode); ++ for (bindex = au_ibstart(inode); bindex <= bend; bindex++) { ++ hi = au_h_iptr(inode, bindex); ++ if (!hi) ++ continue; ++ ++ /* mutex_lock_nested(&hi->i_mutex, AuLsc_I_CHILD); */ ++ iwhdentry = au_hi_wh(inode, bindex); ++ if (iwhdentry) ++ dget(iwhdentry); ++ au_igrab(hi); ++ au_set_h_iptr(inode, bindex, NULL, 0); ++ au_set_h_iptr(inode, bindex, au_igrab(hi), ++ flags & ~AuHi_XINO); ++ iput(hi); ++ dput(iwhdentry); ++ /* mutex_unlock(&hi->i_mutex); */ ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int hn_xino(struct inode *inode, struct inode *h_inode) ++{ ++ int err; ++ aufs_bindex_t bindex, bend, bfound, bstart; ++ struct inode *h_i; ++ ++ err = 0; ++ if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { ++ pr_warn("branch root dir was changed\n"); ++ goto out; ++ } ++ ++ bfound = -1; ++ bend = au_ibend(inode); ++ bstart = au_ibstart(inode); ++#if 0 /* reserved for future use */ ++ if (bindex == bend) { ++ /* keep this ino in rename case */ ++ goto out; ++ } ++#endif ++ for (bindex = bstart; bindex <= bend; bindex++) ++ if (au_h_iptr(inode, bindex) == h_inode) { ++ bfound = bindex; ++ break; ++ } ++ if (bfound < 0) ++ goto out; ++ ++ for (bindex = bstart; bindex <= bend; bindex++) { ++ h_i = au_h_iptr(inode, bindex); ++ if (!h_i) ++ continue; ++ ++ err = au_xino_write(inode->i_sb, bindex, h_i->i_ino, /*ino*/0); ++ /* ignore this error */ ++ /* bad action? */ ++ } ++ ++ /* children inode number will be broken */ ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int hn_gen_tree(struct dentry *dentry) ++{ ++ int err, i, j, ndentry; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry **dentries; ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_dcsub_pages(&dpages, dentry, NULL, NULL); ++ if (unlikely(err)) ++ goto out_dpages; ++ ++ for (i = 0; i < dpages.ndpage; i++) { ++ dpage = dpages.dpages + i; ++ dentries = dpage->dentries; ++ ndentry = dpage->ndentry; ++ for (j = 0; j < ndentry; j++) { ++ struct dentry *d; ++ ++ d = dentries[j]; ++ if (IS_ROOT(d)) ++ continue; ++ ++ au_digen_dec(d); ++ if (d_really_is_positive(d)) ++ /* todo: reset children xino? ++ cached children only? */ ++ au_iigen_dec(d_inode(d)); ++ } ++ } ++ ++out_dpages: ++ au_dpages_free(&dpages); ++ ++#if 0 ++ /* discard children */ ++ dentry_unhash(dentry); ++ dput(dentry); ++#endif ++out: ++ return err; ++} ++ ++/* ++ * return 0 if processed. ++ */ ++static int hn_gen_by_inode(char *name, unsigned int nlen, struct inode *inode, ++ const unsigned int isdir) ++{ ++ int err; ++ struct dentry *d; ++ struct qstr *dname; ++ ++ err = 1; ++ if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { ++ pr_warn("branch root dir was changed\n"); ++ err = 0; ++ goto out; ++ } ++ ++ if (!isdir) { ++ AuDebugOn(!name); ++ au_iigen_dec(inode); ++ spin_lock(&inode->i_lock); ++ hlist_for_each_entry(d, &inode->i_dentry, d_u.d_alias) { ++ spin_lock(&d->d_lock); ++ dname = &d->d_name; ++ if (dname->len != nlen ++ && memcmp(dname->name, name, nlen)) { ++ spin_unlock(&d->d_lock); ++ continue; ++ } ++ err = 0; ++ au_digen_dec(d); ++ spin_unlock(&d->d_lock); ++ break; ++ } ++ spin_unlock(&inode->i_lock); ++ } else { ++ au_fset_si(au_sbi(inode->i_sb), FAILED_REFRESH_DIR); ++ d = d_find_any_alias(inode); ++ if (!d) { ++ au_iigen_dec(inode); ++ goto out; ++ } ++ ++ spin_lock(&d->d_lock); ++ dname = &d->d_name; ++ if (dname->len == nlen && !memcmp(dname->name, name, nlen)) { ++ spin_unlock(&d->d_lock); ++ err = hn_gen_tree(d); ++ spin_lock(&d->d_lock); ++ } ++ spin_unlock(&d->d_lock); ++ dput(d); ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int hn_gen_by_name(struct dentry *dentry, const unsigned int isdir) ++{ ++ int err; ++ ++ if (IS_ROOT(dentry)) { ++ pr_warn("branch root dir was changed\n"); ++ return 0; ++ } ++ ++ err = 0; ++ if (!isdir) { ++ au_digen_dec(dentry); ++ if (d_really_is_positive(dentry)) ++ au_iigen_dec(d_inode(dentry)); ++ } else { ++ au_fset_si(au_sbi(dentry->d_sb), FAILED_REFRESH_DIR); ++ if (d_really_is_positive(dentry)) ++ err = hn_gen_tree(dentry); ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* hnotify job flags */ ++#define AuHnJob_XINO0 1 ++#define AuHnJob_GEN (1 << 1) ++#define AuHnJob_DIRENT (1 << 2) ++#define AuHnJob_ISDIR (1 << 3) ++#define AuHnJob_TRYXINO0 (1 << 4) ++#define AuHnJob_MNTPNT (1 << 5) ++#define au_ftest_hnjob(flags, name) ((flags) & AuHnJob_##name) ++#define au_fset_hnjob(flags, name) \ ++ do { (flags) |= AuHnJob_##name; } while (0) ++#define au_fclr_hnjob(flags, name) \ ++ do { (flags) &= ~AuHnJob_##name; } while (0) ++ ++enum { ++ AuHn_CHILD, ++ AuHn_PARENT, ++ AuHnLast ++}; ++ ++struct au_hnotify_args { ++ struct inode *h_dir, *dir, *h_child_inode; ++ u32 mask; ++ unsigned int flags[AuHnLast]; ++ unsigned int h_child_nlen; ++ char h_child_name[]; ++}; ++ ++struct hn_job_args { ++ unsigned int flags; ++ struct inode *inode, *h_inode, *dir, *h_dir; ++ struct dentry *dentry; ++ char *h_name; ++ int h_nlen; ++}; ++ ++static int hn_job(struct hn_job_args *a) ++{ ++ const unsigned int isdir = au_ftest_hnjob(a->flags, ISDIR); ++ int e; ++ ++ /* reset xino */ ++ if (au_ftest_hnjob(a->flags, XINO0) && a->inode) ++ hn_xino(a->inode, a->h_inode); /* ignore this error */ ++ ++ if (au_ftest_hnjob(a->flags, TRYXINO0) ++ && a->inode ++ && a->h_inode) { ++ mutex_lock_nested(&a->h_inode->i_mutex, AuLsc_I_CHILD); ++ if (!a->h_inode->i_nlink ++ && !(a->h_inode->i_state & I_LINKABLE)) ++ hn_xino(a->inode, a->h_inode); /* ignore this error */ ++ mutex_unlock(&a->h_inode->i_mutex); ++ } ++ ++ /* make the generation obsolete */ ++ if (au_ftest_hnjob(a->flags, GEN)) { ++ e = -1; ++ if (a->inode) ++ e = hn_gen_by_inode(a->h_name, a->h_nlen, a->inode, ++ isdir); ++ if (e && a->dentry) ++ hn_gen_by_name(a->dentry, isdir); ++ /* ignore this error */ ++ } ++ ++ /* make dir entries obsolete */ ++ if (au_ftest_hnjob(a->flags, DIRENT) && a->inode) { ++ struct au_vdir *vdir; ++ ++ vdir = au_ivdir(a->inode); ++ if (vdir) ++ vdir->vd_jiffy = 0; ++ /* IMustLock(a->inode); */ ++ /* a->inode->i_version++; */ ++ } ++ ++ /* can do nothing but warn */ ++ if (au_ftest_hnjob(a->flags, MNTPNT) ++ && a->dentry ++ && d_mountpoint(a->dentry)) ++ pr_warn("mount-point %pd is removed or renamed\n", a->dentry); ++ ++ return 0; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct dentry *lookup_wlock_by_name(char *name, unsigned int nlen, ++ struct inode *dir) ++{ ++ struct dentry *dentry, *d, *parent; ++ struct qstr *dname; ++ ++ parent = d_find_any_alias(dir); ++ if (!parent) ++ return NULL; ++ ++ dentry = NULL; ++ spin_lock(&parent->d_lock); ++ list_for_each_entry(d, &parent->d_subdirs, d_child) { ++ /* AuDbg("%pd\n", d); */ ++ spin_lock_nested(&d->d_lock, DENTRY_D_LOCK_NESTED); ++ dname = &d->d_name; ++ if (dname->len != nlen || memcmp(dname->name, name, nlen)) ++ goto cont_unlock; ++ if (au_di(d)) ++ au_digen_dec(d); ++ else ++ goto cont_unlock; ++ if (au_dcount(d) > 0) { ++ dentry = dget_dlock(d); ++ spin_unlock(&d->d_lock); ++ break; ++ } ++ ++cont_unlock: ++ spin_unlock(&d->d_lock); ++ } ++ spin_unlock(&parent->d_lock); ++ dput(parent); ++ ++ if (dentry) ++ di_write_lock_child(dentry); ++ ++ return dentry; ++} ++ ++static struct inode *lookup_wlock_by_ino(struct super_block *sb, ++ aufs_bindex_t bindex, ino_t h_ino) ++{ ++ struct inode *inode; ++ ino_t ino; ++ int err; ++ ++ inode = NULL; ++ err = au_xino_read(sb, bindex, h_ino, &ino); ++ if (!err && ino) ++ inode = ilookup(sb, ino); ++ if (!inode) ++ goto out; ++ ++ if (unlikely(inode->i_ino == AUFS_ROOT_INO)) { ++ pr_warn("wrong root branch\n"); ++ iput(inode); ++ inode = NULL; ++ goto out; ++ } ++ ++ ii_write_lock_child(inode); ++ ++out: ++ return inode; ++} ++ ++static void au_hn_bh(void *_args) ++{ ++ struct au_hnotify_args *a = _args; ++ struct super_block *sb; ++ aufs_bindex_t bindex, bend, bfound; ++ unsigned char xino, try_iput; ++ int err; ++ struct inode *inode; ++ ino_t h_ino; ++ struct hn_job_args args; ++ struct dentry *dentry; ++ struct au_sbinfo *sbinfo; ++ ++ AuDebugOn(!_args); ++ AuDebugOn(!a->h_dir); ++ AuDebugOn(!a->dir); ++ AuDebugOn(!a->mask); ++ AuDbg("mask 0x%x, i%lu, hi%lu, hci%lu\n", ++ a->mask, a->dir->i_ino, a->h_dir->i_ino, ++ a->h_child_inode ? a->h_child_inode->i_ino : 0); ++ ++ inode = NULL; ++ dentry = NULL; ++ /* ++ * do not lock a->dir->i_mutex here ++ * because of d_revalidate() may cause a deadlock. ++ */ ++ sb = a->dir->i_sb; ++ AuDebugOn(!sb); ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!sbinfo); ++ si_write_lock(sb, AuLock_NOPLMW); ++ ++ ii_read_lock_parent(a->dir); ++ bfound = -1; ++ bend = au_ibend(a->dir); ++ for (bindex = au_ibstart(a->dir); bindex <= bend; bindex++) ++ if (au_h_iptr(a->dir, bindex) == a->h_dir) { ++ bfound = bindex; ++ break; ++ } ++ ii_read_unlock(a->dir); ++ if (unlikely(bfound < 0)) ++ goto out; ++ ++ xino = !!au_opt_test(au_mntflags(sb), XINO); ++ h_ino = 0; ++ if (a->h_child_inode) ++ h_ino = a->h_child_inode->i_ino; ++ ++ if (a->h_child_nlen ++ && (au_ftest_hnjob(a->flags[AuHn_CHILD], GEN) ++ || au_ftest_hnjob(a->flags[AuHn_CHILD], MNTPNT))) ++ dentry = lookup_wlock_by_name(a->h_child_name, a->h_child_nlen, ++ a->dir); ++ try_iput = 0; ++ if (dentry && d_really_is_positive(dentry)) ++ inode = d_inode(dentry); ++ if (xino && !inode && h_ino ++ && (au_ftest_hnjob(a->flags[AuHn_CHILD], XINO0) ++ || au_ftest_hnjob(a->flags[AuHn_CHILD], TRYXINO0) ++ || au_ftest_hnjob(a->flags[AuHn_CHILD], GEN))) { ++ inode = lookup_wlock_by_ino(sb, bfound, h_ino); ++ try_iput = 1; ++ } ++ ++ args.flags = a->flags[AuHn_CHILD]; ++ args.dentry = dentry; ++ args.inode = inode; ++ args.h_inode = a->h_child_inode; ++ args.dir = a->dir; ++ args.h_dir = a->h_dir; ++ args.h_name = a->h_child_name; ++ args.h_nlen = a->h_child_nlen; ++ err = hn_job(&args); ++ if (dentry) { ++ if (au_di(dentry)) ++ di_write_unlock(dentry); ++ dput(dentry); ++ } ++ if (inode && try_iput) { ++ ii_write_unlock(inode); ++ iput(inode); ++ } ++ ++ ii_write_lock_parent(a->dir); ++ args.flags = a->flags[AuHn_PARENT]; ++ args.dentry = NULL; ++ args.inode = a->dir; ++ args.h_inode = a->h_dir; ++ args.dir = NULL; ++ args.h_dir = NULL; ++ args.h_name = NULL; ++ args.h_nlen = 0; ++ err = hn_job(&args); ++ ii_write_unlock(a->dir); ++ ++out: ++ iput(a->h_child_inode); ++ iput(a->h_dir); ++ iput(a->dir); ++ si_write_unlock(sb); ++ au_nwt_done(&sbinfo->si_nowait); ++ kfree(a); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, ++ struct qstr *h_child_qstr, struct inode *h_child_inode) ++{ ++ int err, len; ++ unsigned int flags[AuHnLast], f; ++ unsigned char isdir, isroot, wh; ++ struct inode *dir; ++ struct au_hnotify_args *args; ++ char *p, *h_child_name; ++ ++ err = 0; ++ AuDebugOn(!hnotify || !hnotify->hn_aufs_inode); ++ dir = igrab(hnotify->hn_aufs_inode); ++ if (!dir) ++ goto out; ++ ++ isroot = (dir->i_ino == AUFS_ROOT_INO); ++ wh = 0; ++ h_child_name = (void *)h_child_qstr->name; ++ len = h_child_qstr->len; ++ if (h_child_name) { ++ if (len > AUFS_WH_PFX_LEN ++ && !memcmp(h_child_name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { ++ h_child_name += AUFS_WH_PFX_LEN; ++ len -= AUFS_WH_PFX_LEN; ++ wh = 1; ++ } ++ } ++ ++ isdir = 0; ++ if (h_child_inode) ++ isdir = !!S_ISDIR(h_child_inode->i_mode); ++ flags[AuHn_PARENT] = AuHnJob_ISDIR; ++ flags[AuHn_CHILD] = 0; ++ if (isdir) ++ flags[AuHn_CHILD] = AuHnJob_ISDIR; ++ au_fset_hnjob(flags[AuHn_PARENT], DIRENT); ++ au_fset_hnjob(flags[AuHn_CHILD], GEN); ++ switch (mask & FS_EVENTS_POSS_ON_CHILD) { ++ case FS_MOVED_FROM: ++ case FS_MOVED_TO: ++ au_fset_hnjob(flags[AuHn_CHILD], XINO0); ++ au_fset_hnjob(flags[AuHn_CHILD], MNTPNT); ++ /*FALLTHROUGH*/ ++ case FS_CREATE: ++ AuDebugOn(!h_child_name); ++ break; ++ ++ case FS_DELETE: ++ /* ++ * aufs never be able to get this child inode. ++ * revalidation should be in d_revalidate() ++ * by checking i_nlink, i_generation or d_unhashed(). ++ */ ++ AuDebugOn(!h_child_name); ++ au_fset_hnjob(flags[AuHn_CHILD], TRYXINO0); ++ au_fset_hnjob(flags[AuHn_CHILD], MNTPNT); ++ break; ++ ++ default: ++ AuDebugOn(1); ++ } ++ ++ if (wh) ++ h_child_inode = NULL; ++ ++ err = -ENOMEM; ++ /* iput() and kfree() will be called in au_hnotify() */ ++ args = kmalloc(sizeof(*args) + len + 1, GFP_NOFS); ++ if (unlikely(!args)) { ++ AuErr1("no memory\n"); ++ iput(dir); ++ goto out; ++ } ++ args->flags[AuHn_PARENT] = flags[AuHn_PARENT]; ++ args->flags[AuHn_CHILD] = flags[AuHn_CHILD]; ++ args->mask = mask; ++ args->dir = dir; ++ args->h_dir = igrab(h_dir); ++ if (h_child_inode) ++ h_child_inode = igrab(h_child_inode); /* can be NULL */ ++ args->h_child_inode = h_child_inode; ++ args->h_child_nlen = len; ++ if (len) { ++ p = (void *)args; ++ p += sizeof(*args); ++ memcpy(p, h_child_name, len); ++ p[len] = 0; ++ } ++ ++ /* NFS fires the event for silly-renamed one from kworker */ ++ f = 0; ++ if (!dir->i_nlink ++ || (au_test_nfs(h_dir->i_sb) && (mask & FS_DELETE))) ++ f = AuWkq_NEST; ++ err = au_wkq_nowait(au_hn_bh, args, dir->i_sb, f); ++ if (unlikely(err)) { ++ pr_err("wkq %d\n", err); ++ iput(args->h_child_inode); ++ iput(args->h_dir); ++ iput(args->dir); ++ kfree(args); ++ } ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_hnotify_reset_br(unsigned int udba, struct au_branch *br, int perm) ++{ ++ int err; ++ ++ AuDebugOn(!(udba & AuOptMask_UDBA)); ++ ++ err = 0; ++ if (au_hnotify_op.reset_br) ++ err = au_hnotify_op.reset_br(udba, br, perm); ++ ++ return err; ++} ++ ++int au_hnotify_init_br(struct au_branch *br, int perm) ++{ ++ int err; ++ ++ err = 0; ++ if (au_hnotify_op.init_br) ++ err = au_hnotify_op.init_br(br, perm); ++ ++ return err; ++} ++ ++void au_hnotify_fin_br(struct au_branch *br) ++{ ++ if (au_hnotify_op.fin_br) ++ au_hnotify_op.fin_br(br); ++} ++ ++static void au_hn_destroy_cache(void) ++{ ++ kmem_cache_destroy(au_cachep[AuCache_HNOTIFY]); ++ au_cachep[AuCache_HNOTIFY] = NULL; ++} ++ ++int __init au_hnotify_init(void) ++{ ++ int err; ++ ++ err = -ENOMEM; ++ au_cachep[AuCache_HNOTIFY] = AuCache(au_hnotify); ++ if (au_cachep[AuCache_HNOTIFY]) { ++ err = 0; ++ if (au_hnotify_op.init) ++ err = au_hnotify_op.init(); ++ if (unlikely(err)) ++ au_hn_destroy_cache(); ++ } ++ AuTraceErr(err); ++ return err; ++} ++ ++void au_hnotify_fin(void) ++{ ++ if (au_hnotify_op.fin) ++ au_hnotify_op.fin(); ++ /* cf. au_cache_fin() */ ++ if (au_cachep[AuCache_HNOTIFY]) ++ au_hn_destroy_cache(); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/iinfo.c linux-4.1.10/fs/aufs/iinfo.c +--- linux-4.1.10.orig/fs/aufs/iinfo.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/iinfo.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,277 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode private data ++ */ ++ ++#include "aufs.h" ++ ++struct inode *au_h_iptr(struct inode *inode, aufs_bindex_t bindex) ++{ ++ struct inode *h_inode; ++ ++ IiMustAnyLock(inode); ++ ++ h_inode = au_ii(inode)->ii_hinode[0 + bindex].hi_inode; ++ AuDebugOn(h_inode && atomic_read(&h_inode->i_count) <= 0); ++ return h_inode; ++} ++ ++/* todo: hard/soft set? */ ++void au_hiput(struct au_hinode *hinode) ++{ ++ au_hn_free(hinode); ++ dput(hinode->hi_whdentry); ++ iput(hinode->hi_inode); ++} ++ ++unsigned int au_hi_flags(struct inode *inode, int isdir) ++{ ++ unsigned int flags; ++ const unsigned int mnt_flags = au_mntflags(inode->i_sb); ++ ++ flags = 0; ++ if (au_opt_test(mnt_flags, XINO)) ++ au_fset_hi(flags, XINO); ++ if (isdir && au_opt_test(mnt_flags, UDBA_HNOTIFY)) ++ au_fset_hi(flags, HNOTIFY); ++ return flags; ++} ++ ++void au_set_h_iptr(struct inode *inode, aufs_bindex_t bindex, ++ struct inode *h_inode, unsigned int flags) ++{ ++ struct au_hinode *hinode; ++ struct inode *hi; ++ struct au_iinfo *iinfo = au_ii(inode); ++ ++ IiMustWriteLock(inode); ++ ++ hinode = iinfo->ii_hinode + bindex; ++ hi = hinode->hi_inode; ++ AuDebugOn(h_inode && atomic_read(&h_inode->i_count) <= 0); ++ ++ if (hi) ++ au_hiput(hinode); ++ hinode->hi_inode = h_inode; ++ if (h_inode) { ++ int err; ++ struct super_block *sb = inode->i_sb; ++ struct au_branch *br; ++ ++ AuDebugOn(inode->i_mode ++ && (h_inode->i_mode & S_IFMT) ++ != (inode->i_mode & S_IFMT)); ++ if (bindex == iinfo->ii_bstart) ++ au_cpup_igen(inode, h_inode); ++ br = au_sbr(sb, bindex); ++ hinode->hi_id = br->br_id; ++ if (au_ftest_hi(flags, XINO)) { ++ err = au_xino_write(sb, bindex, h_inode->i_ino, ++ inode->i_ino); ++ if (unlikely(err)) ++ AuIOErr1("failed au_xino_write() %d\n", err); ++ } ++ ++ if (au_ftest_hi(flags, HNOTIFY) ++ && au_br_hnotifyable(br->br_perm)) { ++ err = au_hn_alloc(hinode, inode); ++ if (unlikely(err)) ++ AuIOErr1("au_hn_alloc() %d\n", err); ++ } ++ } ++} ++ ++void au_set_hi_wh(struct inode *inode, aufs_bindex_t bindex, ++ struct dentry *h_wh) ++{ ++ struct au_hinode *hinode; ++ ++ IiMustWriteLock(inode); ++ ++ hinode = au_ii(inode)->ii_hinode + bindex; ++ AuDebugOn(hinode->hi_whdentry); ++ hinode->hi_whdentry = h_wh; ++} ++ ++void au_update_iigen(struct inode *inode, int half) ++{ ++ struct au_iinfo *iinfo; ++ struct au_iigen *iigen; ++ unsigned int sigen; ++ ++ sigen = au_sigen(inode->i_sb); ++ iinfo = au_ii(inode); ++ iigen = &iinfo->ii_generation; ++ spin_lock(&iinfo->ii_genspin); ++ iigen->ig_generation = sigen; ++ if (half) ++ au_ig_fset(iigen->ig_flags, HALF_REFRESHED); ++ else ++ au_ig_fclr(iigen->ig_flags, HALF_REFRESHED); ++ spin_unlock(&iinfo->ii_genspin); ++} ++ ++/* it may be called at remount time, too */ ++void au_update_ibrange(struct inode *inode, int do_put_zero) ++{ ++ struct au_iinfo *iinfo; ++ aufs_bindex_t bindex, bend; ++ ++ iinfo = au_ii(inode); ++ if (!iinfo) ++ return; ++ ++ IiMustWriteLock(inode); ++ ++ if (do_put_zero && iinfo->ii_bstart >= 0) { ++ for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; ++ bindex++) { ++ struct inode *h_i; ++ ++ h_i = iinfo->ii_hinode[0 + bindex].hi_inode; ++ if (h_i ++ && !h_i->i_nlink ++ && !(h_i->i_state & I_LINKABLE)) ++ au_set_h_iptr(inode, bindex, NULL, 0); ++ } ++ } ++ ++ iinfo->ii_bstart = -1; ++ iinfo->ii_bend = -1; ++ bend = au_sbend(inode->i_sb); ++ for (bindex = 0; bindex <= bend; bindex++) ++ if (iinfo->ii_hinode[0 + bindex].hi_inode) { ++ iinfo->ii_bstart = bindex; ++ break; ++ } ++ if (iinfo->ii_bstart >= 0) ++ for (bindex = bend; bindex >= iinfo->ii_bstart; bindex--) ++ if (iinfo->ii_hinode[0 + bindex].hi_inode) { ++ iinfo->ii_bend = bindex; ++ break; ++ } ++ AuDebugOn(iinfo->ii_bstart > iinfo->ii_bend); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_icntnr_init_once(void *_c) ++{ ++ struct au_icntnr *c = _c; ++ struct au_iinfo *iinfo = &c->iinfo; ++ static struct lock_class_key aufs_ii; ++ ++ spin_lock_init(&iinfo->ii_genspin); ++ au_rw_init(&iinfo->ii_rwsem); ++ au_rw_class(&iinfo->ii_rwsem, &aufs_ii); ++ inode_init_once(&c->vfs_inode); ++} ++ ++int au_iinfo_init(struct inode *inode) ++{ ++ struct au_iinfo *iinfo; ++ struct super_block *sb; ++ int nbr, i; ++ ++ sb = inode->i_sb; ++ iinfo = &(container_of(inode, struct au_icntnr, vfs_inode)->iinfo); ++ nbr = au_sbend(sb) + 1; ++ if (unlikely(nbr <= 0)) ++ nbr = 1; ++ iinfo->ii_hinode = kcalloc(nbr, sizeof(*iinfo->ii_hinode), GFP_NOFS); ++ if (iinfo->ii_hinode) { ++ au_ninodes_inc(sb); ++ for (i = 0; i < nbr; i++) ++ iinfo->ii_hinode[i].hi_id = -1; ++ ++ iinfo->ii_generation.ig_generation = au_sigen(sb); ++ iinfo->ii_bstart = -1; ++ iinfo->ii_bend = -1; ++ iinfo->ii_vdir = NULL; ++ return 0; ++ } ++ return -ENOMEM; ++} ++ ++int au_ii_realloc(struct au_iinfo *iinfo, int nbr) ++{ ++ int err, sz; ++ struct au_hinode *hip; ++ ++ AuRwMustWriteLock(&iinfo->ii_rwsem); ++ ++ err = -ENOMEM; ++ sz = sizeof(*hip) * (iinfo->ii_bend + 1); ++ if (!sz) ++ sz = sizeof(*hip); ++ hip = au_kzrealloc(iinfo->ii_hinode, sz, sizeof(*hip) * nbr, GFP_NOFS); ++ if (hip) { ++ iinfo->ii_hinode = hip; ++ err = 0; ++ } ++ ++ return err; ++} ++ ++void au_iinfo_fin(struct inode *inode) ++{ ++ struct au_iinfo *iinfo; ++ struct au_hinode *hi; ++ struct super_block *sb; ++ aufs_bindex_t bindex, bend; ++ const unsigned char unlinked = !inode->i_nlink; ++ ++ iinfo = au_ii(inode); ++ /* bad_inode case */ ++ if (!iinfo) ++ return; ++ ++ sb = inode->i_sb; ++ au_ninodes_dec(sb); ++ if (si_pid_test(sb)) ++ au_xino_delete_inode(inode, unlinked); ++ else { ++ /* ++ * it is safe to hide the dependency between sbinfo and ++ * sb->s_umount. ++ */ ++ lockdep_off(); ++ si_noflush_read_lock(sb); ++ au_xino_delete_inode(inode, unlinked); ++ si_read_unlock(sb); ++ lockdep_on(); ++ } ++ ++ if (iinfo->ii_vdir) ++ au_vdir_free(iinfo->ii_vdir); ++ ++ bindex = iinfo->ii_bstart; ++ if (bindex >= 0) { ++ hi = iinfo->ii_hinode + bindex; ++ bend = iinfo->ii_bend; ++ while (bindex++ <= bend) { ++ if (hi->hi_inode) ++ au_hiput(hi); ++ hi++; ++ } ++ } ++ kfree(iinfo->ii_hinode); ++ iinfo->ii_hinode = NULL; ++ AuRwDestroy(&iinfo->ii_rwsem); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/inode.c linux-4.1.10/fs/aufs/inode.c +--- linux-4.1.10.orig/fs/aufs/inode.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/inode.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,500 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode functions ++ */ ++ ++#include "aufs.h" ++ ++struct inode *au_igrab(struct inode *inode) ++{ ++ if (inode) { ++ AuDebugOn(!atomic_read(&inode->i_count)); ++ ihold(inode); ++ } ++ return inode; ++} ++ ++static void au_refresh_hinode_attr(struct inode *inode, int do_version) ++{ ++ au_cpup_attr_all(inode, /*force*/0); ++ au_update_iigen(inode, /*half*/1); ++ if (do_version) ++ inode->i_version++; ++} ++ ++static int au_ii_refresh(struct inode *inode, int *update) ++{ ++ int err, e; ++ umode_t type; ++ aufs_bindex_t bindex, new_bindex; ++ struct super_block *sb; ++ struct au_iinfo *iinfo; ++ struct au_hinode *p, *q, tmp; ++ ++ IiMustWriteLock(inode); ++ ++ *update = 0; ++ sb = inode->i_sb; ++ type = inode->i_mode & S_IFMT; ++ iinfo = au_ii(inode); ++ err = au_ii_realloc(iinfo, au_sbend(sb) + 1); ++ if (unlikely(err)) ++ goto out; ++ ++ AuDebugOn(iinfo->ii_bstart < 0); ++ p = iinfo->ii_hinode + iinfo->ii_bstart; ++ for (bindex = iinfo->ii_bstart; bindex <= iinfo->ii_bend; ++ bindex++, p++) { ++ if (!p->hi_inode) ++ continue; ++ ++ AuDebugOn(type != (p->hi_inode->i_mode & S_IFMT)); ++ new_bindex = au_br_index(sb, p->hi_id); ++ if (new_bindex == bindex) ++ continue; ++ ++ if (new_bindex < 0) { ++ *update = 1; ++ au_hiput(p); ++ p->hi_inode = NULL; ++ continue; ++ } ++ ++ if (new_bindex < iinfo->ii_bstart) ++ iinfo->ii_bstart = new_bindex; ++ if (iinfo->ii_bend < new_bindex) ++ iinfo->ii_bend = new_bindex; ++ /* swap two lower inode, and loop again */ ++ q = iinfo->ii_hinode + new_bindex; ++ tmp = *q; ++ *q = *p; ++ *p = tmp; ++ if (tmp.hi_inode) { ++ bindex--; ++ p--; ++ } ++ } ++ au_update_ibrange(inode, /*do_put_zero*/0); ++ e = au_dy_irefresh(inode); ++ if (unlikely(e && !err)) ++ err = e; ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_refresh_hinode_self(struct inode *inode) ++{ ++ int err, update; ++ ++ err = au_ii_refresh(inode, &update); ++ if (!err) ++ au_refresh_hinode_attr(inode, update && S_ISDIR(inode->i_mode)); ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_refresh_hinode(struct inode *inode, struct dentry *dentry) ++{ ++ int err, e, update; ++ unsigned int flags; ++ umode_t mode; ++ aufs_bindex_t bindex, bend; ++ unsigned char isdir; ++ struct au_hinode *p; ++ struct au_iinfo *iinfo; ++ ++ err = au_ii_refresh(inode, &update); ++ if (unlikely(err)) ++ goto out; ++ ++ update = 0; ++ iinfo = au_ii(inode); ++ p = iinfo->ii_hinode + iinfo->ii_bstart; ++ mode = (inode->i_mode & S_IFMT); ++ isdir = S_ISDIR(mode); ++ flags = au_hi_flags(inode, isdir); ++ bend = au_dbend(dentry); ++ for (bindex = au_dbstart(dentry); bindex <= bend; bindex++) { ++ struct inode *h_i, *h_inode; ++ struct dentry *h_d; ++ ++ h_d = au_h_dptr(dentry, bindex); ++ if (!h_d || d_is_negative(h_d)) ++ continue; ++ ++ h_inode = d_inode(h_d); ++ AuDebugOn(mode != (h_inode->i_mode & S_IFMT)); ++ if (iinfo->ii_bstart <= bindex && bindex <= iinfo->ii_bend) { ++ h_i = au_h_iptr(inode, bindex); ++ if (h_i) { ++ if (h_i == h_inode) ++ continue; ++ err = -EIO; ++ break; ++ } ++ } ++ if (bindex < iinfo->ii_bstart) ++ iinfo->ii_bstart = bindex; ++ if (iinfo->ii_bend < bindex) ++ iinfo->ii_bend = bindex; ++ au_set_h_iptr(inode, bindex, au_igrab(h_inode), flags); ++ update = 1; ++ } ++ au_update_ibrange(inode, /*do_put_zero*/0); ++ e = au_dy_irefresh(inode); ++ if (unlikely(e && !err)) ++ err = e; ++ if (!err) ++ au_refresh_hinode_attr(inode, update && isdir); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int set_inode(struct inode *inode, struct dentry *dentry) ++{ ++ int err; ++ unsigned int flags; ++ umode_t mode; ++ aufs_bindex_t bindex, bstart, btail; ++ unsigned char isdir; ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ struct au_iinfo *iinfo; ++ ++ IiMustWriteLock(inode); ++ ++ err = 0; ++ isdir = 0; ++ bstart = au_dbstart(dentry); ++ h_dentry = au_h_dptr(dentry, bstart); ++ h_inode = d_inode(h_dentry); ++ mode = h_inode->i_mode; ++ switch (mode & S_IFMT) { ++ case S_IFREG: ++ btail = au_dbtail(dentry); ++ inode->i_op = &aufs_iop; ++ inode->i_fop = &aufs_file_fop; ++ err = au_dy_iaop(inode, bstart, h_inode); ++ if (unlikely(err)) ++ goto out; ++ break; ++ case S_IFDIR: ++ isdir = 1; ++ btail = au_dbtaildir(dentry); ++ inode->i_op = &aufs_dir_iop; ++ inode->i_fop = &aufs_dir_fop; ++ break; ++ case S_IFLNK: ++ btail = au_dbtail(dentry); ++ inode->i_op = &aufs_symlink_iop; ++ break; ++ case S_IFBLK: ++ case S_IFCHR: ++ case S_IFIFO: ++ case S_IFSOCK: ++ btail = au_dbtail(dentry); ++ inode->i_op = &aufs_iop; ++ init_special_inode(inode, mode, h_inode->i_rdev); ++ break; ++ default: ++ AuIOErr("Unknown file type 0%o\n", mode); ++ err = -EIO; ++ goto out; ++ } ++ ++ /* do not set hnotify for whiteouted dirs (SHWH mode) */ ++ flags = au_hi_flags(inode, isdir); ++ if (au_opt_test(au_mntflags(dentry->d_sb), SHWH) ++ && au_ftest_hi(flags, HNOTIFY) ++ && dentry->d_name.len > AUFS_WH_PFX_LEN ++ && !memcmp(dentry->d_name.name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) ++ au_fclr_hi(flags, HNOTIFY); ++ iinfo = au_ii(inode); ++ iinfo->ii_bstart = bstart; ++ iinfo->ii_bend = btail; ++ for (bindex = bstart; bindex <= btail; bindex++) { ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (h_dentry) ++ au_set_h_iptr(inode, bindex, ++ au_igrab(d_inode(h_dentry)), flags); ++ } ++ au_cpup_attr_all(inode, /*force*/1); ++ /* ++ * to force calling aufs_get_acl() every time, ++ * do not call cache_no_acl() for aufs inode. ++ */ ++ ++out: ++ return err; ++} ++ ++/* ++ * successful returns with iinfo write_locked ++ * minus: errno ++ * zero: success, matched ++ * plus: no error, but unmatched ++ */ ++static int reval_inode(struct inode *inode, struct dentry *dentry) ++{ ++ int err; ++ unsigned int gen; ++ struct au_iigen iigen; ++ aufs_bindex_t bindex, bend; ++ struct inode *h_inode, *h_dinode; ++ struct dentry *h_dentry; ++ ++ /* ++ * before this function, if aufs got any iinfo lock, it must be only ++ * one, the parent dir. ++ * it can happen by UDBA and the obsoleted inode number. ++ */ ++ err = -EIO; ++ if (unlikely(inode->i_ino == parent_ino(dentry))) ++ goto out; ++ ++ err = 1; ++ ii_write_lock_new_child(inode); ++ h_dentry = au_h_dptr(dentry, au_dbstart(dentry)); ++ h_dinode = d_inode(h_dentry); ++ bend = au_ibend(inode); ++ for (bindex = au_ibstart(inode); bindex <= bend; bindex++) { ++ h_inode = au_h_iptr(inode, bindex); ++ if (!h_inode || h_inode != h_dinode) ++ continue; ++ ++ err = 0; ++ gen = au_iigen(inode, &iigen); ++ if (gen == au_digen(dentry) ++ && !au_ig_ftest(iigen.ig_flags, HALF_REFRESHED)) ++ break; ++ ++ /* fully refresh inode using dentry */ ++ err = au_refresh_hinode(inode, dentry); ++ if (!err) ++ au_update_iigen(inode, /*half*/0); ++ break; ++ } ++ ++ if (unlikely(err)) ++ ii_write_unlock(inode); ++out: ++ return err; ++} ++ ++int au_ino(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ unsigned int d_type, ino_t *ino) ++{ ++ int err; ++ struct mutex *mtx; ++ ++ /* prevent hardlinked inode number from race condition */ ++ mtx = NULL; ++ if (d_type != DT_DIR) { ++ mtx = &au_sbr(sb, bindex)->br_xino.xi_nondir_mtx; ++ mutex_lock(mtx); ++ } ++ err = au_xino_read(sb, bindex, h_ino, ino); ++ if (unlikely(err)) ++ goto out; ++ ++ if (!*ino) { ++ err = -EIO; ++ *ino = au_xino_new_ino(sb); ++ if (unlikely(!*ino)) ++ goto out; ++ err = au_xino_write(sb, bindex, h_ino, *ino); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++out: ++ if (mtx) ++ mutex_unlock(mtx); ++ return err; ++} ++ ++/* successful returns with iinfo write_locked */ ++/* todo: return with unlocked? */ ++struct inode *au_new_inode(struct dentry *dentry, int must_new) ++{ ++ struct inode *inode, *h_inode; ++ struct dentry *h_dentry; ++ struct super_block *sb; ++ struct mutex *mtx; ++ ino_t h_ino, ino; ++ int err; ++ aufs_bindex_t bstart; ++ ++ sb = dentry->d_sb; ++ bstart = au_dbstart(dentry); ++ h_dentry = au_h_dptr(dentry, bstart); ++ h_inode = d_inode(h_dentry); ++ h_ino = h_inode->i_ino; ++ ++ /* ++ * stop 'race'-ing between hardlinks under different ++ * parents. ++ */ ++ mtx = NULL; ++ if (!d_is_dir(h_dentry)) ++ mtx = &au_sbr(sb, bstart)->br_xino.xi_nondir_mtx; ++ ++new_ino: ++ if (mtx) ++ mutex_lock(mtx); ++ err = au_xino_read(sb, bstart, h_ino, &ino); ++ inode = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ if (!ino) { ++ ino = au_xino_new_ino(sb); ++ if (unlikely(!ino)) { ++ inode = ERR_PTR(-EIO); ++ goto out; ++ } ++ } ++ ++ AuDbg("i%lu\n", (unsigned long)ino); ++ inode = au_iget_locked(sb, ino); ++ err = PTR_ERR(inode); ++ if (IS_ERR(inode)) ++ goto out; ++ ++ AuDbg("%lx, new %d\n", inode->i_state, !!(inode->i_state & I_NEW)); ++ if (inode->i_state & I_NEW) { ++ /* verbose coding for lock class name */ ++ if (unlikely(d_is_symlink(h_dentry))) ++ au_rw_class(&au_ii(inode)->ii_rwsem, ++ au_lc_key + AuLcSymlink_IIINFO); ++ else if (unlikely(d_is_dir(h_dentry))) ++ au_rw_class(&au_ii(inode)->ii_rwsem, ++ au_lc_key + AuLcDir_IIINFO); ++ else /* likely */ ++ au_rw_class(&au_ii(inode)->ii_rwsem, ++ au_lc_key + AuLcNonDir_IIINFO); ++ ++ ii_write_lock_new_child(inode); ++ err = set_inode(inode, dentry); ++ if (!err) { ++ unlock_new_inode(inode); ++ goto out; /* success */ ++ } ++ ++ /* ++ * iget_failed() calls iput(), but we need to call ++ * ii_write_unlock() after iget_failed(). so dirty hack for ++ * i_count. ++ */ ++ atomic_inc(&inode->i_count); ++ iget_failed(inode); ++ ii_write_unlock(inode); ++ au_xino_write(sb, bstart, h_ino, /*ino*/0); ++ /* ignore this error */ ++ goto out_iput; ++ } else if (!must_new && !IS_DEADDIR(inode) && inode->i_nlink) { ++ /* ++ * horrible race condition between lookup, readdir and copyup ++ * (or something). ++ */ ++ if (mtx) ++ mutex_unlock(mtx); ++ err = reval_inode(inode, dentry); ++ if (unlikely(err < 0)) { ++ mtx = NULL; ++ goto out_iput; ++ } ++ ++ if (!err) { ++ mtx = NULL; ++ goto out; /* success */ ++ } else if (mtx) ++ mutex_lock(mtx); ++ } ++ ++ if (unlikely(au_test_fs_unique_ino(h_inode))) ++ AuWarn1("Warning: Un-notified UDBA or repeatedly renamed dir," ++ " b%d, %s, %pd, hi%lu, i%lu.\n", ++ bstart, au_sbtype(h_dentry->d_sb), dentry, ++ (unsigned long)h_ino, (unsigned long)ino); ++ ino = 0; ++ err = au_xino_write(sb, bstart, h_ino, /*ino*/0); ++ if (!err) { ++ iput(inode); ++ if (mtx) ++ mutex_unlock(mtx); ++ goto new_ino; ++ } ++ ++out_iput: ++ iput(inode); ++ inode = ERR_PTR(err); ++out: ++ if (mtx) ++ mutex_unlock(mtx); ++ return inode; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_test_ro(struct super_block *sb, aufs_bindex_t bindex, ++ struct inode *inode) ++{ ++ int err; ++ struct inode *hi; ++ ++ err = au_br_rdonly(au_sbr(sb, bindex)); ++ ++ /* pseudo-link after flushed may happen out of bounds */ ++ if (!err ++ && inode ++ && au_ibstart(inode) <= bindex ++ && bindex <= au_ibend(inode)) { ++ /* ++ * permission check is unnecessary since vfsub routine ++ * will be called later ++ */ ++ hi = au_h_iptr(inode, bindex); ++ if (hi) ++ err = IS_IMMUTABLE(hi) ? -EROFS : 0; ++ } ++ ++ return err; ++} ++ ++int au_test_h_perm(struct inode *h_inode, int mask) ++{ ++ if (uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) ++ return 0; ++ return inode_permission(h_inode, mask); ++} ++ ++int au_test_h_perm_sio(struct inode *h_inode, int mask) ++{ ++ if (au_test_nfs(h_inode->i_sb) ++ && (mask & MAY_WRITE) ++ && S_ISDIR(h_inode->i_mode)) ++ mask |= MAY_READ; /* force permission check */ ++ return au_test_h_perm(h_inode, mask); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/inode.h linux-4.1.10/fs/aufs/inode.h +--- linux-4.1.10.orig/fs/aufs/inode.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/inode.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,673 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode operations ++ */ ++ ++#ifndef __AUFS_INODE_H__ ++#define __AUFS_INODE_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fsnotify.h> ++#include "rwsem.h" ++ ++struct vfsmount; ++ ++struct au_hnotify { ++#ifdef CONFIG_AUFS_HNOTIFY ++#ifdef CONFIG_AUFS_HFSNOTIFY ++ /* never use fsnotify_add_vfsmount_mark() */ ++ struct fsnotify_mark hn_mark; ++#endif ++ struct inode *hn_aufs_inode; /* no get/put */ ++#endif ++} ____cacheline_aligned_in_smp; ++ ++struct au_hinode { ++ struct inode *hi_inode; ++ aufs_bindex_t hi_id; ++#ifdef CONFIG_AUFS_HNOTIFY ++ struct au_hnotify *hi_notify; ++#endif ++ ++ /* reference to the copied-up whiteout with get/put */ ++ struct dentry *hi_whdentry; ++}; ++ ++/* ig_flags */ ++#define AuIG_HALF_REFRESHED 1 ++#define au_ig_ftest(flags, name) ((flags) & AuIG_##name) ++#define au_ig_fset(flags, name) \ ++ do { (flags) |= AuIG_##name; } while (0) ++#define au_ig_fclr(flags, name) \ ++ do { (flags) &= ~AuIG_##name; } while (0) ++ ++struct au_iigen { ++ __u32 ig_generation, ig_flags; ++}; ++ ++struct au_vdir; ++struct au_iinfo { ++ spinlock_t ii_genspin; ++ struct au_iigen ii_generation; ++ struct super_block *ii_hsb1; /* no get/put */ ++ ++ struct au_rwsem ii_rwsem; ++ aufs_bindex_t ii_bstart, ii_bend; ++ __u32 ii_higen; ++ struct au_hinode *ii_hinode; ++ struct au_vdir *ii_vdir; ++}; ++ ++struct au_icntnr { ++ struct au_iinfo iinfo; ++ struct inode vfs_inode; ++} ____cacheline_aligned_in_smp; ++ ++/* au_pin flags */ ++#define AuPin_DI_LOCKED 1 ++#define AuPin_MNT_WRITE (1 << 1) ++#define au_ftest_pin(flags, name) ((flags) & AuPin_##name) ++#define au_fset_pin(flags, name) \ ++ do { (flags) |= AuPin_##name; } while (0) ++#define au_fclr_pin(flags, name) \ ++ do { (flags) &= ~AuPin_##name; } while (0) ++ ++struct au_pin { ++ /* input */ ++ struct dentry *dentry; ++ unsigned int udba; ++ unsigned char lsc_di, lsc_hi, flags; ++ aufs_bindex_t bindex; ++ ++ /* output */ ++ struct dentry *parent; ++ struct au_hinode *hdir; ++ struct vfsmount *h_mnt; ++ ++ /* temporary unlock/relock for copyup */ ++ struct dentry *h_dentry, *h_parent; ++ struct au_branch *br; ++ struct task_struct *task; ++}; ++ ++void au_pin_hdir_unlock(struct au_pin *p); ++int au_pin_hdir_lock(struct au_pin *p); ++int au_pin_hdir_relock(struct au_pin *p); ++void au_pin_hdir_set_owner(struct au_pin *p, struct task_struct *task); ++void au_pin_hdir_acquire_nest(struct au_pin *p); ++void au_pin_hdir_release(struct au_pin *p); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct au_iinfo *au_ii(struct inode *inode) ++{ ++ struct au_iinfo *iinfo; ++ ++ iinfo = &(container_of(inode, struct au_icntnr, vfs_inode)->iinfo); ++ if (iinfo->ii_hinode) ++ return iinfo; ++ return NULL; /* debugging bad_inode case */ ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* inode.c */ ++struct inode *au_igrab(struct inode *inode); ++int au_refresh_hinode_self(struct inode *inode); ++int au_refresh_hinode(struct inode *inode, struct dentry *dentry); ++int au_ino(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ unsigned int d_type, ino_t *ino); ++struct inode *au_new_inode(struct dentry *dentry, int must_new); ++int au_test_ro(struct super_block *sb, aufs_bindex_t bindex, ++ struct inode *inode); ++int au_test_h_perm(struct inode *h_inode, int mask); ++int au_test_h_perm_sio(struct inode *h_inode, int mask); ++ ++static inline int au_wh_ino(struct super_block *sb, aufs_bindex_t bindex, ++ ino_t h_ino, unsigned int d_type, ino_t *ino) ++{ ++#ifdef CONFIG_AUFS_SHWH ++ return au_ino(sb, bindex, h_ino, d_type, ino); ++#else ++ return 0; ++#endif ++} ++ ++/* i_op.c */ ++extern struct inode_operations aufs_iop, aufs_symlink_iop, aufs_dir_iop; ++ ++/* au_wr_dir flags */ ++#define AuWrDir_ADD_ENTRY 1 ++#define AuWrDir_ISDIR (1 << 1) ++#define AuWrDir_TMPFILE (1 << 2) ++#define au_ftest_wrdir(flags, name) ((flags) & AuWrDir_##name) ++#define au_fset_wrdir(flags, name) \ ++ do { (flags) |= AuWrDir_##name; } while (0) ++#define au_fclr_wrdir(flags, name) \ ++ do { (flags) &= ~AuWrDir_##name; } while (0) ++ ++struct au_wr_dir_args { ++ aufs_bindex_t force_btgt; ++ unsigned char flags; ++}; ++int au_wr_dir(struct dentry *dentry, struct dentry *src_dentry, ++ struct au_wr_dir_args *args); ++ ++struct dentry *au_pinned_h_parent(struct au_pin *pin); ++void au_pin_init(struct au_pin *pin, struct dentry *dentry, ++ aufs_bindex_t bindex, int lsc_di, int lsc_hi, ++ unsigned int udba, unsigned char flags); ++int au_pin(struct au_pin *pin, struct dentry *dentry, aufs_bindex_t bindex, ++ unsigned int udba, unsigned char flags) __must_check; ++int au_do_pin(struct au_pin *pin) __must_check; ++void au_unpin(struct au_pin *pin); ++int au_reval_for_attr(struct dentry *dentry, unsigned int sigen); ++ ++#define AuIcpup_DID_CPUP 1 ++#define au_ftest_icpup(flags, name) ((flags) & AuIcpup_##name) ++#define au_fset_icpup(flags, name) \ ++ do { (flags) |= AuIcpup_##name; } while (0) ++#define au_fclr_icpup(flags, name) \ ++ do { (flags) &= ~AuIcpup_##name; } while (0) ++ ++struct au_icpup_args { ++ unsigned char flags; ++ unsigned char pin_flags; ++ aufs_bindex_t btgt; ++ unsigned int udba; ++ struct au_pin pin; ++ struct path h_path; ++ struct inode *h_inode; ++}; ++ ++int au_pin_and_icpup(struct dentry *dentry, struct iattr *ia, ++ struct au_icpup_args *a); ++ ++int au_h_path_getattr(struct dentry *dentry, int force, struct path *h_path); ++ ++/* i_op_add.c */ ++int au_may_add(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent, int isdir); ++int aufs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, ++ dev_t dev); ++int aufs_symlink(struct inode *dir, struct dentry *dentry, const char *symname); ++int aufs_create(struct inode *dir, struct dentry *dentry, umode_t mode, ++ bool want_excl); ++struct vfsub_aopen_args; ++int au_aopen_or_create(struct inode *dir, struct dentry *dentry, ++ struct vfsub_aopen_args *args); ++int aufs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode); ++int aufs_link(struct dentry *src_dentry, struct inode *dir, ++ struct dentry *dentry); ++int aufs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode); ++ ++/* i_op_del.c */ ++int au_wr_dir_need_wh(struct dentry *dentry, int isdir, aufs_bindex_t *bcpup); ++int au_may_del(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent, int isdir); ++int aufs_unlink(struct inode *dir, struct dentry *dentry); ++int aufs_rmdir(struct inode *dir, struct dentry *dentry); ++ ++/* i_op_ren.c */ ++int au_wbr(struct dentry *dentry, aufs_bindex_t btgt); ++int aufs_rename(struct inode *src_dir, struct dentry *src_dentry, ++ struct inode *dir, struct dentry *dentry); ++ ++/* iinfo.c */ ++struct inode *au_h_iptr(struct inode *inode, aufs_bindex_t bindex); ++void au_hiput(struct au_hinode *hinode); ++void au_set_hi_wh(struct inode *inode, aufs_bindex_t bindex, ++ struct dentry *h_wh); ++unsigned int au_hi_flags(struct inode *inode, int isdir); ++ ++/* hinode flags */ ++#define AuHi_XINO 1 ++#define AuHi_HNOTIFY (1 << 1) ++#define au_ftest_hi(flags, name) ((flags) & AuHi_##name) ++#define au_fset_hi(flags, name) \ ++ do { (flags) |= AuHi_##name; } while (0) ++#define au_fclr_hi(flags, name) \ ++ do { (flags) &= ~AuHi_##name; } while (0) ++ ++#ifndef CONFIG_AUFS_HNOTIFY ++#undef AuHi_HNOTIFY ++#define AuHi_HNOTIFY 0 ++#endif ++ ++void au_set_h_iptr(struct inode *inode, aufs_bindex_t bindex, ++ struct inode *h_inode, unsigned int flags); ++ ++void au_update_iigen(struct inode *inode, int half); ++void au_update_ibrange(struct inode *inode, int do_put_zero); ++ ++void au_icntnr_init_once(void *_c); ++int au_iinfo_init(struct inode *inode); ++void au_iinfo_fin(struct inode *inode); ++int au_ii_realloc(struct au_iinfo *iinfo, int nbr); ++ ++#ifdef CONFIG_PROC_FS ++/* plink.c */ ++int au_plink_maint(struct super_block *sb, int flags); ++struct au_sbinfo; ++void au_plink_maint_leave(struct au_sbinfo *sbinfo); ++int au_plink_maint_enter(struct super_block *sb); ++#ifdef CONFIG_AUFS_DEBUG ++void au_plink_list(struct super_block *sb); ++#else ++AuStubVoid(au_plink_list, struct super_block *sb) ++#endif ++int au_plink_test(struct inode *inode); ++struct dentry *au_plink_lkup(struct inode *inode, aufs_bindex_t bindex); ++void au_plink_append(struct inode *inode, aufs_bindex_t bindex, ++ struct dentry *h_dentry); ++void au_plink_put(struct super_block *sb, int verbose); ++void au_plink_clean(struct super_block *sb, int verbose); ++void au_plink_half_refresh(struct super_block *sb, aufs_bindex_t br_id); ++#else ++AuStubInt0(au_plink_maint, struct super_block *sb, int flags); ++AuStubVoid(au_plink_maint_leave, struct au_sbinfo *sbinfo); ++AuStubInt0(au_plink_maint_enter, struct super_block *sb); ++AuStubVoid(au_plink_list, struct super_block *sb); ++AuStubInt0(au_plink_test, struct inode *inode); ++AuStub(struct dentry *, au_plink_lkup, return NULL, ++ struct inode *inode, aufs_bindex_t bindex); ++AuStubVoid(au_plink_append, struct inode *inode, aufs_bindex_t bindex, ++ struct dentry *h_dentry); ++AuStubVoid(au_plink_put, struct super_block *sb, int verbose); ++AuStubVoid(au_plink_clean, struct super_block *sb, int verbose); ++AuStubVoid(au_plink_half_refresh, struct super_block *sb, aufs_bindex_t br_id); ++#endif /* CONFIG_PROC_FS */ ++ ++#ifdef CONFIG_AUFS_XATTR ++/* xattr.c */ ++int au_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, int ignore_flags, ++ unsigned int verbose); ++ssize_t aufs_listxattr(struct dentry *dentry, char *list, size_t size); ++ssize_t aufs_getxattr(struct dentry *dentry, const char *name, void *value, ++ size_t size); ++int aufs_setxattr(struct dentry *dentry, const char *name, const void *value, ++ size_t size, int flags); ++int aufs_removexattr(struct dentry *dentry, const char *name); ++ ++/* void au_xattr_init(struct super_block *sb); */ ++#else ++AuStubInt0(au_cpup_xattr, struct dentry *h_dst, struct dentry *h_src, ++ int ignore_flags, unsigned int verbose); ++/* AuStubVoid(au_xattr_init, struct super_block *sb); */ ++#endif ++ ++#ifdef CONFIG_FS_POSIX_ACL ++struct posix_acl *aufs_get_acl(struct inode *inode, int type); ++int aufs_set_acl(struct inode *inode, struct posix_acl *acl, int type); ++#endif ++ ++#if IS_ENABLED(CONFIG_AUFS_XATTR) || IS_ENABLED(CONFIG_FS_POSIX_ACL) ++enum { ++ AU_XATTR_SET, ++ AU_XATTR_REMOVE, ++ AU_ACL_SET ++}; ++ ++struct au_srxattr { ++ int type; ++ union { ++ struct { ++ const char *name; ++ const void *value; ++ size_t size; ++ int flags; ++ } set; ++ struct { ++ const char *name; ++ } remove; ++ struct { ++ struct posix_acl *acl; ++ int type; ++ } acl_set; ++ } u; ++}; ++ssize_t au_srxattr(struct dentry *dentry, struct au_srxattr *arg); ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* lock subclass for iinfo */ ++enum { ++ AuLsc_II_CHILD, /* child first */ ++ AuLsc_II_CHILD2, /* rename(2), link(2), and cpup at hnotify */ ++ AuLsc_II_CHILD3, /* copyup dirs */ ++ AuLsc_II_PARENT, /* see AuLsc_I_PARENT in vfsub.h */ ++ AuLsc_II_PARENT2, ++ AuLsc_II_PARENT3, /* copyup dirs */ ++ AuLsc_II_NEW_CHILD ++}; ++ ++/* ++ * ii_read_lock_child, ii_write_lock_child, ++ * ii_read_lock_child2, ii_write_lock_child2, ++ * ii_read_lock_child3, ii_write_lock_child3, ++ * ii_read_lock_parent, ii_write_lock_parent, ++ * ii_read_lock_parent2, ii_write_lock_parent2, ++ * ii_read_lock_parent3, ii_write_lock_parent3, ++ * ii_read_lock_new_child, ii_write_lock_new_child, ++ */ ++#define AuReadLockFunc(name, lsc) \ ++static inline void ii_read_lock_##name(struct inode *i) \ ++{ \ ++ au_rw_read_lock_nested(&au_ii(i)->ii_rwsem, AuLsc_II_##lsc); \ ++} ++ ++#define AuWriteLockFunc(name, lsc) \ ++static inline void ii_write_lock_##name(struct inode *i) \ ++{ \ ++ au_rw_write_lock_nested(&au_ii(i)->ii_rwsem, AuLsc_II_##lsc); \ ++} ++ ++#define AuRWLockFuncs(name, lsc) \ ++ AuReadLockFunc(name, lsc) \ ++ AuWriteLockFunc(name, lsc) ++ ++AuRWLockFuncs(child, CHILD); ++AuRWLockFuncs(child2, CHILD2); ++AuRWLockFuncs(child3, CHILD3); ++AuRWLockFuncs(parent, PARENT); ++AuRWLockFuncs(parent2, PARENT2); ++AuRWLockFuncs(parent3, PARENT3); ++AuRWLockFuncs(new_child, NEW_CHILD); ++ ++#undef AuReadLockFunc ++#undef AuWriteLockFunc ++#undef AuRWLockFuncs ++ ++/* ++ * ii_read_unlock, ii_write_unlock, ii_downgrade_lock ++ */ ++AuSimpleUnlockRwsemFuncs(ii, struct inode *i, &au_ii(i)->ii_rwsem); ++ ++#define IiMustNoWaiters(i) AuRwMustNoWaiters(&au_ii(i)->ii_rwsem) ++#define IiMustAnyLock(i) AuRwMustAnyLock(&au_ii(i)->ii_rwsem) ++#define IiMustWriteLock(i) AuRwMustWriteLock(&au_ii(i)->ii_rwsem) ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline void au_icntnr_init(struct au_icntnr *c) ++{ ++#ifdef CONFIG_AUFS_DEBUG ++ c->vfs_inode.i_mode = 0; ++#endif ++} ++ ++static inline unsigned int au_iigen(struct inode *inode, struct au_iigen *iigen) ++{ ++ unsigned int gen; ++ struct au_iinfo *iinfo; ++ ++ iinfo = au_ii(inode); ++ spin_lock(&iinfo->ii_genspin); ++ if (iigen) ++ *iigen = iinfo->ii_generation; ++ gen = iinfo->ii_generation.ig_generation; ++ spin_unlock(&iinfo->ii_genspin); ++ ++ return gen; ++} ++ ++/* tiny test for inode number */ ++/* tmpfs generation is too rough */ ++static inline int au_test_higen(struct inode *inode, struct inode *h_inode) ++{ ++ struct au_iinfo *iinfo; ++ ++ iinfo = au_ii(inode); ++ AuRwMustAnyLock(&iinfo->ii_rwsem); ++ return !(iinfo->ii_hsb1 == h_inode->i_sb ++ && iinfo->ii_higen == h_inode->i_generation); ++} ++ ++static inline void au_iigen_dec(struct inode *inode) ++{ ++ struct au_iinfo *iinfo; ++ ++ iinfo = au_ii(inode); ++ spin_lock(&iinfo->ii_genspin); ++ iinfo->ii_generation.ig_generation--; ++ spin_unlock(&iinfo->ii_genspin); ++} ++ ++static inline int au_iigen_test(struct inode *inode, unsigned int sigen) ++{ ++ int err; ++ ++ err = 0; ++ if (unlikely(inode && au_iigen(inode, NULL) != sigen)) ++ err = -EIO; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline aufs_bindex_t au_ii_br_id(struct inode *inode, ++ aufs_bindex_t bindex) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_hinode[0 + bindex].hi_id; ++} ++ ++static inline aufs_bindex_t au_ibstart(struct inode *inode) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_bstart; ++} ++ ++static inline aufs_bindex_t au_ibend(struct inode *inode) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_bend; ++} ++ ++static inline struct au_vdir *au_ivdir(struct inode *inode) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_vdir; ++} ++ ++static inline struct dentry *au_hi_wh(struct inode *inode, aufs_bindex_t bindex) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_hinode[0 + bindex].hi_whdentry; ++} ++ ++static inline void au_set_ibstart(struct inode *inode, aufs_bindex_t bindex) ++{ ++ IiMustWriteLock(inode); ++ au_ii(inode)->ii_bstart = bindex; ++} ++ ++static inline void au_set_ibend(struct inode *inode, aufs_bindex_t bindex) ++{ ++ IiMustWriteLock(inode); ++ au_ii(inode)->ii_bend = bindex; ++} ++ ++static inline void au_set_ivdir(struct inode *inode, struct au_vdir *vdir) ++{ ++ IiMustWriteLock(inode); ++ au_ii(inode)->ii_vdir = vdir; ++} ++ ++static inline struct au_hinode *au_hi(struct inode *inode, aufs_bindex_t bindex) ++{ ++ IiMustAnyLock(inode); ++ return au_ii(inode)->ii_hinode + bindex; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct dentry *au_pinned_parent(struct au_pin *pin) ++{ ++ if (pin) ++ return pin->parent; ++ return NULL; ++} ++ ++static inline struct inode *au_pinned_h_dir(struct au_pin *pin) ++{ ++ if (pin && pin->hdir) ++ return pin->hdir->hi_inode; ++ return NULL; ++} ++ ++static inline struct au_hinode *au_pinned_hdir(struct au_pin *pin) ++{ ++ if (pin) ++ return pin->hdir; ++ return NULL; ++} ++ ++static inline void au_pin_set_dentry(struct au_pin *pin, struct dentry *dentry) ++{ ++ if (pin) ++ pin->dentry = dentry; ++} ++ ++static inline void au_pin_set_parent_lflag(struct au_pin *pin, ++ unsigned char lflag) ++{ ++ if (pin) { ++ if (lflag) ++ au_fset_pin(pin->flags, DI_LOCKED); ++ else ++ au_fclr_pin(pin->flags, DI_LOCKED); ++ } ++} ++ ++#if 0 /* reserved */ ++static inline void au_pin_set_parent(struct au_pin *pin, struct dentry *parent) ++{ ++ if (pin) { ++ dput(pin->parent); ++ pin->parent = dget(parent); ++ } ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_branch; ++#ifdef CONFIG_AUFS_HNOTIFY ++struct au_hnotify_op { ++ void (*ctl)(struct au_hinode *hinode, int do_set); ++ int (*alloc)(struct au_hinode *hinode); ++ ++ /* ++ * if it returns true, the the caller should free hinode->hi_notify, ++ * otherwise ->free() frees it. ++ */ ++ int (*free)(struct au_hinode *hinode, ++ struct au_hnotify *hn) __must_check; ++ ++ void (*fin)(void); ++ int (*init)(void); ++ ++ int (*reset_br)(unsigned int udba, struct au_branch *br, int perm); ++ void (*fin_br)(struct au_branch *br); ++ int (*init_br)(struct au_branch *br, int perm); ++}; ++ ++/* hnotify.c */ ++int au_hn_alloc(struct au_hinode *hinode, struct inode *inode); ++void au_hn_free(struct au_hinode *hinode); ++void au_hn_ctl(struct au_hinode *hinode, int do_set); ++void au_hn_reset(struct inode *inode, unsigned int flags); ++int au_hnotify(struct inode *h_dir, struct au_hnotify *hnotify, u32 mask, ++ struct qstr *h_child_qstr, struct inode *h_child_inode); ++int au_hnotify_reset_br(unsigned int udba, struct au_branch *br, int perm); ++int au_hnotify_init_br(struct au_branch *br, int perm); ++void au_hnotify_fin_br(struct au_branch *br); ++int __init au_hnotify_init(void); ++void au_hnotify_fin(void); ++ ++/* hfsnotify.c */ ++extern const struct au_hnotify_op au_hnotify_op; ++ ++static inline ++void au_hn_init(struct au_hinode *hinode) ++{ ++ hinode->hi_notify = NULL; ++} ++ ++static inline struct au_hnotify *au_hn(struct au_hinode *hinode) ++{ ++ return hinode->hi_notify; ++} ++ ++#else ++AuStub(int, au_hn_alloc, return -EOPNOTSUPP, ++ struct au_hinode *hinode __maybe_unused, ++ struct inode *inode __maybe_unused) ++AuStub(struct au_hnotify *, au_hn, return NULL, struct au_hinode *hinode) ++AuStubVoid(au_hn_free, struct au_hinode *hinode __maybe_unused) ++AuStubVoid(au_hn_ctl, struct au_hinode *hinode __maybe_unused, ++ int do_set __maybe_unused) ++AuStubVoid(au_hn_reset, struct inode *inode __maybe_unused, ++ unsigned int flags __maybe_unused) ++AuStubInt0(au_hnotify_reset_br, unsigned int udba __maybe_unused, ++ struct au_branch *br __maybe_unused, ++ int perm __maybe_unused) ++AuStubInt0(au_hnotify_init_br, struct au_branch *br __maybe_unused, ++ int perm __maybe_unused) ++AuStubVoid(au_hnotify_fin_br, struct au_branch *br __maybe_unused) ++AuStubInt0(__init au_hnotify_init, void) ++AuStubVoid(au_hnotify_fin, void) ++AuStubVoid(au_hn_init, struct au_hinode *hinode __maybe_unused) ++#endif /* CONFIG_AUFS_HNOTIFY */ ++ ++static inline void au_hn_suspend(struct au_hinode *hdir) ++{ ++ au_hn_ctl(hdir, /*do_set*/0); ++} ++ ++static inline void au_hn_resume(struct au_hinode *hdir) ++{ ++ au_hn_ctl(hdir, /*do_set*/1); ++} ++ ++static inline void au_hn_imtx_lock(struct au_hinode *hdir) ++{ ++ mutex_lock(&hdir->hi_inode->i_mutex); ++ au_hn_suspend(hdir); ++} ++ ++static inline void au_hn_imtx_lock_nested(struct au_hinode *hdir, ++ unsigned int sc __maybe_unused) ++{ ++ mutex_lock_nested(&hdir->hi_inode->i_mutex, sc); ++ au_hn_suspend(hdir); ++} ++ ++static inline void au_hn_imtx_unlock(struct au_hinode *hdir) ++{ ++ au_hn_resume(hdir); ++ mutex_unlock(&hdir->hi_inode->i_mutex); ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_INODE_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/ioctl.c linux-4.1.10/fs/aufs/ioctl.c +--- linux-4.1.10.orig/fs/aufs/ioctl.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/ioctl.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,219 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * ioctl ++ * plink-management and readdir in userspace. ++ * assist the pathconf(3) wrapper library. ++ * move-down ++ * File-based Hierarchical Storage Management. ++ */ ++ ++#include <linux/compat.h> ++#include <linux/file.h> ++#include "aufs.h" ++ ++static int au_wbr_fd(struct path *path, struct aufs_wbr_fd __user *arg) ++{ ++ int err, fd; ++ aufs_bindex_t wbi, bindex, bend; ++ struct file *h_file; ++ struct super_block *sb; ++ struct dentry *root; ++ struct au_branch *br; ++ struct aufs_wbr_fd wbrfd = { ++ .oflags = au_dir_roflags, ++ .brid = -1 ++ }; ++ const int valid = O_RDONLY | O_NONBLOCK | O_LARGEFILE | O_DIRECTORY ++ | O_NOATIME | O_CLOEXEC; ++ ++ AuDebugOn(wbrfd.oflags & ~valid); ++ ++ if (arg) { ++ err = copy_from_user(&wbrfd, arg, sizeof(wbrfd)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ goto out; ++ } ++ ++ err = -EINVAL; ++ AuDbg("wbrfd{0%o, %d}\n", wbrfd.oflags, wbrfd.brid); ++ wbrfd.oflags |= au_dir_roflags; ++ AuDbg("0%o\n", wbrfd.oflags); ++ if (unlikely(wbrfd.oflags & ~valid)) ++ goto out; ++ } ++ ++ fd = get_unused_fd_flags(0); ++ err = fd; ++ if (unlikely(fd < 0)) ++ goto out; ++ ++ h_file = ERR_PTR(-EINVAL); ++ wbi = 0; ++ br = NULL; ++ sb = path->dentry->d_sb; ++ root = sb->s_root; ++ aufs_read_lock(root, AuLock_IR); ++ bend = au_sbend(sb); ++ if (wbrfd.brid >= 0) { ++ wbi = au_br_index(sb, wbrfd.brid); ++ if (unlikely(wbi < 0 || wbi > bend)) ++ goto out_unlock; ++ } ++ ++ h_file = ERR_PTR(-ENOENT); ++ br = au_sbr(sb, wbi); ++ if (!au_br_writable(br->br_perm)) { ++ if (arg) ++ goto out_unlock; ++ ++ bindex = wbi + 1; ++ wbi = -1; ++ for (; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_writable(br->br_perm)) { ++ wbi = bindex; ++ br = au_sbr(sb, wbi); ++ break; ++ } ++ } ++ } ++ AuDbg("wbi %d\n", wbi); ++ if (wbi >= 0) ++ h_file = au_h_open(root, wbi, wbrfd.oflags, NULL, ++ /*force_wr*/0); ++ ++out_unlock: ++ aufs_read_unlock(root, AuLock_IR); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out_fd; ++ ++ atomic_dec(&br->br_count); /* cf. au_h_open() */ ++ fd_install(fd, h_file); ++ err = fd; ++ goto out; /* success */ ++ ++out_fd: ++ put_unused_fd(fd); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++long aufs_ioctl_dir(struct file *file, unsigned int cmd, unsigned long arg) ++{ ++ long err; ++ struct dentry *dentry; ++ ++ switch (cmd) { ++ case AUFS_CTL_RDU: ++ case AUFS_CTL_RDU_INO: ++ err = au_rdu_ioctl(file, cmd, arg); ++ break; ++ ++ case AUFS_CTL_WBR_FD: ++ err = au_wbr_fd(&file->f_path, (void __user *)arg); ++ break; ++ ++ case AUFS_CTL_IBUSY: ++ err = au_ibusy_ioctl(file, arg); ++ break; ++ ++ case AUFS_CTL_BRINFO: ++ err = au_brinfo_ioctl(file, arg); ++ break; ++ ++ case AUFS_CTL_FHSM_FD: ++ dentry = file->f_path.dentry; ++ if (IS_ROOT(dentry)) ++ err = au_fhsm_fd(dentry->d_sb, arg); ++ else ++ err = -ENOTTY; ++ break; ++ ++ default: ++ /* do not call the lower */ ++ AuDbg("0x%x\n", cmd); ++ err = -ENOTTY; ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++long aufs_ioctl_nondir(struct file *file, unsigned int cmd, unsigned long arg) ++{ ++ long err; ++ ++ switch (cmd) { ++ case AUFS_CTL_MVDOWN: ++ err = au_mvdown(file->f_path.dentry, (void __user *)arg); ++ break; ++ ++ case AUFS_CTL_WBR_FD: ++ err = au_wbr_fd(&file->f_path, (void __user *)arg); ++ break; ++ ++ default: ++ /* do not call the lower */ ++ AuDbg("0x%x\n", cmd); ++ err = -ENOTTY; ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++#ifdef CONFIG_COMPAT ++long aufs_compat_ioctl_dir(struct file *file, unsigned int cmd, ++ unsigned long arg) ++{ ++ long err; ++ ++ switch (cmd) { ++ case AUFS_CTL_RDU: ++ case AUFS_CTL_RDU_INO: ++ err = au_rdu_compat_ioctl(file, cmd, arg); ++ break; ++ ++ case AUFS_CTL_IBUSY: ++ err = au_ibusy_compat_ioctl(file, arg); ++ break; ++ ++ case AUFS_CTL_BRINFO: ++ err = au_brinfo_compat_ioctl(file, arg); ++ break; ++ ++ default: ++ err = aufs_ioctl_dir(file, cmd, arg); ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++long aufs_compat_ioctl_nondir(struct file *file, unsigned int cmd, ++ unsigned long arg) ++{ ++ return aufs_ioctl_nondir(file, cmd, (unsigned long)compat_ptr(arg)); ++} ++#endif +diff -Nur linux-4.1.10.orig/fs/aufs/i_op_add.c linux-4.1.10/fs/aufs/i_op_add.c +--- linux-4.1.10.orig/fs/aufs/i_op_add.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/i_op_add.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,932 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode operations (add entry) ++ */ ++ ++#include "aufs.h" ++ ++/* ++ * final procedure of adding a new entry, except link(2). ++ * remove whiteout, instantiate, copyup the parent dir's times and size ++ * and update version. ++ * if it failed, re-create the removed whiteout. ++ */ ++static int epilog(struct inode *dir, aufs_bindex_t bindex, ++ struct dentry *wh_dentry, struct dentry *dentry) ++{ ++ int err, rerr; ++ aufs_bindex_t bwh; ++ struct path h_path; ++ struct super_block *sb; ++ struct inode *inode, *h_dir; ++ struct dentry *wh; ++ ++ bwh = -1; ++ sb = dir->i_sb; ++ if (wh_dentry) { ++ h_dir = d_inode(wh_dentry->d_parent); /* dir inode is locked */ ++ IMustLock(h_dir); ++ AuDebugOn(au_h_iptr(dir, bindex) != h_dir); ++ bwh = au_dbwh(dentry); ++ h_path.dentry = wh_dentry; ++ h_path.mnt = au_sbr_mnt(sb, bindex); ++ err = au_wh_unlink_dentry(au_h_iptr(dir, bindex), &h_path, ++ dentry); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++ inode = au_new_inode(dentry, /*must_new*/1); ++ if (!IS_ERR(inode)) { ++ d_instantiate(dentry, inode); ++ dir = d_inode(dentry->d_parent); /* dir inode is locked */ ++ IMustLock(dir); ++ au_dir_ts(dir, bindex); ++ dir->i_version++; ++ au_fhsm_wrote(sb, bindex, /*force*/0); ++ return 0; /* success */ ++ } ++ ++ err = PTR_ERR(inode); ++ if (!wh_dentry) ++ goto out; ++ ++ /* revert */ ++ /* dir inode is locked */ ++ wh = au_wh_create(dentry, bwh, wh_dentry->d_parent); ++ rerr = PTR_ERR(wh); ++ if (IS_ERR(wh)) { ++ AuIOErr("%pd reverting whiteout failed(%d, %d)\n", ++ dentry, err, rerr); ++ err = -EIO; ++ } else ++ dput(wh); ++ ++out: ++ return err; ++} ++ ++static int au_d_may_add(struct dentry *dentry) ++{ ++ int err; ++ ++ err = 0; ++ if (unlikely(d_unhashed(dentry))) ++ err = -ENOENT; ++ if (unlikely(d_really_is_positive(dentry))) ++ err = -EEXIST; ++ return err; ++} ++ ++/* ++ * simple tests for the adding inode operations. ++ * following the checks in vfs, plus the parent-child relationship. ++ */ ++int au_may_add(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent, int isdir) ++{ ++ int err; ++ umode_t h_mode; ++ struct dentry *h_dentry; ++ struct inode *h_inode; ++ ++ err = -ENAMETOOLONG; ++ if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) ++ goto out; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (d_really_is_negative(dentry)) { ++ err = -EEXIST; ++ if (unlikely(d_is_positive(h_dentry))) ++ goto out; ++ } else { ++ /* rename(2) case */ ++ err = -EIO; ++ if (unlikely(d_is_negative(h_dentry))) ++ goto out; ++ h_inode = d_inode(h_dentry); ++ if (unlikely(!h_inode->i_nlink)) ++ goto out; ++ ++ h_mode = h_inode->i_mode; ++ if (!isdir) { ++ err = -EISDIR; ++ if (unlikely(S_ISDIR(h_mode))) ++ goto out; ++ } else if (unlikely(!S_ISDIR(h_mode))) { ++ err = -ENOTDIR; ++ goto out; ++ } ++ } ++ ++ err = 0; ++ /* expected parent dir is locked */ ++ if (unlikely(h_parent != h_dentry->d_parent)) ++ err = -EIO; ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ++ * initial procedure of adding a new entry. ++ * prepare writable branch and the parent dir, lock it, ++ * and lookup whiteout for the new entry. ++ */ ++static struct dentry* ++lock_hdir_lkup_wh(struct dentry *dentry, struct au_dtime *dt, ++ struct dentry *src_dentry, struct au_pin *pin, ++ struct au_wr_dir_args *wr_dir_args) ++{ ++ struct dentry *wh_dentry, *h_parent; ++ struct super_block *sb; ++ struct au_branch *br; ++ int err; ++ unsigned int udba; ++ aufs_bindex_t bcpup; ++ ++ AuDbg("%pd\n", dentry); ++ ++ err = au_wr_dir(dentry, src_dentry, wr_dir_args); ++ bcpup = err; ++ wh_dentry = ERR_PTR(err); ++ if (unlikely(err < 0)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ udba = au_opt_udba(sb); ++ err = au_pin(pin, dentry, bcpup, udba, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ wh_dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ h_parent = au_pinned_h_parent(pin); ++ if (udba != AuOpt_UDBA_NONE ++ && au_dbstart(dentry) == bcpup) ++ err = au_may_add(dentry, bcpup, h_parent, ++ au_ftest_wrdir(wr_dir_args->flags, ISDIR)); ++ else if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) ++ err = -ENAMETOOLONG; ++ wh_dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out_unpin; ++ ++ br = au_sbr(sb, bcpup); ++ if (dt) { ++ struct path tmp = { ++ .dentry = h_parent, ++ .mnt = au_br_mnt(br) ++ }; ++ au_dtime_store(dt, au_pinned_parent(pin), &tmp); ++ } ++ ++ wh_dentry = NULL; ++ if (bcpup != au_dbwh(dentry)) ++ goto out; /* success */ ++ ++ /* ++ * ENAMETOOLONG here means that if we allowed create such name, then it ++ * would not be able to removed in the future. So we don't allow such ++ * name here and we don't handle ENAMETOOLONG differently here. ++ */ ++ wh_dentry = au_wh_lkup(h_parent, &dentry->d_name, br); ++ ++out_unpin: ++ if (IS_ERR(wh_dentry)) ++ au_unpin(pin); ++out: ++ return wh_dentry; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++enum { Mknod, Symlink, Creat }; ++struct simple_arg { ++ int type; ++ union { ++ struct { ++ umode_t mode; ++ bool want_excl; ++ bool try_aopen; ++ struct vfsub_aopen_args *aopen; ++ } c; ++ struct { ++ const char *symname; ++ } s; ++ struct { ++ umode_t mode; ++ dev_t dev; ++ } m; ++ } u; ++}; ++ ++static int add_simple(struct inode *dir, struct dentry *dentry, ++ struct simple_arg *arg) ++{ ++ int err, rerr; ++ aufs_bindex_t bstart; ++ unsigned char created; ++ const unsigned char try_aopen ++ = (arg->type == Creat && arg->u.c.try_aopen); ++ struct dentry *wh_dentry, *parent; ++ struct inode *h_dir; ++ struct super_block *sb; ++ struct au_branch *br; ++ /* to reuduce stack size */ ++ struct { ++ struct au_dtime dt; ++ struct au_pin pin; ++ struct path h_path; ++ struct au_wr_dir_args wr_dir_args; ++ } *a; ++ ++ AuDbg("%pd\n", dentry); ++ IMustLock(dir); ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ a->wr_dir_args.force_btgt = -1; ++ a->wr_dir_args.flags = AuWrDir_ADD_ENTRY; ++ ++ parent = dentry->d_parent; /* dir inode is locked */ ++ if (!try_aopen) { ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_free; ++ } ++ err = au_d_may_add(dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ if (!try_aopen) ++ di_write_lock_parent(parent); ++ wh_dentry = lock_hdir_lkup_wh(dentry, &a->dt, /*src_dentry*/NULL, ++ &a->pin, &a->wr_dir_args); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_parent; ++ ++ bstart = au_dbstart(dentry); ++ sb = dentry->d_sb; ++ br = au_sbr(sb, bstart); ++ a->h_path.dentry = au_h_dptr(dentry, bstart); ++ a->h_path.mnt = au_br_mnt(br); ++ h_dir = au_pinned_h_dir(&a->pin); ++ switch (arg->type) { ++ case Creat: ++ err = 0; ++ if (!try_aopen || !h_dir->i_op->atomic_open) ++ err = vfsub_create(h_dir, &a->h_path, arg->u.c.mode, ++ arg->u.c.want_excl); ++ else ++ err = vfsub_atomic_open(h_dir, a->h_path.dentry, ++ arg->u.c.aopen, br); ++ break; ++ case Symlink: ++ err = vfsub_symlink(h_dir, &a->h_path, arg->u.s.symname); ++ break; ++ case Mknod: ++ err = vfsub_mknod(h_dir, &a->h_path, arg->u.m.mode, ++ arg->u.m.dev); ++ break; ++ default: ++ BUG(); ++ } ++ created = !err; ++ if (!err) ++ err = epilog(dir, bstart, wh_dentry, dentry); ++ ++ /* revert */ ++ if (unlikely(created && err && d_is_positive(a->h_path.dentry))) { ++ /* no delegation since it is just created */ ++ rerr = vfsub_unlink(h_dir, &a->h_path, /*delegated*/NULL, ++ /*force*/0); ++ if (rerr) { ++ AuIOErr("%pd revert failure(%d, %d)\n", ++ dentry, err, rerr); ++ err = -EIO; ++ } ++ au_dtime_revert(&a->dt); ++ } ++ ++ if (!err && try_aopen && !h_dir->i_op->atomic_open) ++ *arg->u.c.aopen->opened |= FILE_CREATED; ++ ++ au_unpin(&a->pin); ++ dput(wh_dentry); ++ ++out_parent: ++ if (!try_aopen) ++ di_write_unlock(parent); ++out_unlock: ++ if (unlikely(err)) { ++ au_update_dbstart(dentry); ++ d_drop(dentry); ++ } ++ if (!try_aopen) ++ aufs_read_unlock(dentry, AuLock_DW); ++out_free: ++ kfree(a); ++out: ++ return err; ++} ++ ++int aufs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, ++ dev_t dev) ++{ ++ struct simple_arg arg = { ++ .type = Mknod, ++ .u.m = { ++ .mode = mode, ++ .dev = dev ++ } ++ }; ++ return add_simple(dir, dentry, &arg); ++} ++ ++int aufs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) ++{ ++ struct simple_arg arg = { ++ .type = Symlink, ++ .u.s.symname = symname ++ }; ++ return add_simple(dir, dentry, &arg); ++} ++ ++int aufs_create(struct inode *dir, struct dentry *dentry, umode_t mode, ++ bool want_excl) ++{ ++ struct simple_arg arg = { ++ .type = Creat, ++ .u.c = { ++ .mode = mode, ++ .want_excl = want_excl ++ } ++ }; ++ return add_simple(dir, dentry, &arg); ++} ++ ++int au_aopen_or_create(struct inode *dir, struct dentry *dentry, ++ struct vfsub_aopen_args *aopen_args) ++{ ++ struct simple_arg arg = { ++ .type = Creat, ++ .u.c = { ++ .mode = aopen_args->create_mode, ++ .want_excl = aopen_args->open_flag & O_EXCL, ++ .try_aopen = true, ++ .aopen = aopen_args ++ } ++ }; ++ return add_simple(dir, dentry, &arg); ++} ++ ++int aufs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct super_block *sb; ++ struct dentry *parent, *h_parent, *h_dentry; ++ struct inode *h_dir, *inode; ++ struct vfsmount *h_mnt; ++ struct au_wr_dir_args wr_dir_args = { ++ .force_btgt = -1, ++ .flags = AuWrDir_TMPFILE ++ }; ++ ++ /* copy-up may happen */ ++ mutex_lock(&dir->i_mutex); ++ ++ sb = dir->i_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_di_init(dentry); ++ if (unlikely(err)) ++ goto out_si; ++ ++ err = -EBUSY; ++ parent = d_find_any_alias(dir); ++ AuDebugOn(!parent); ++ di_write_lock_parent(parent); ++ if (unlikely(d_inode(parent) != dir)) ++ goto out_parent; ++ ++ err = au_digen_test(parent, au_sigen(sb)); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ bindex = au_dbstart(parent); ++ au_set_dbstart(dentry, bindex); ++ au_set_dbend(dentry, bindex); ++ err = au_wr_dir(dentry, /*src_dentry*/NULL, &wr_dir_args); ++ bindex = err; ++ if (unlikely(err < 0)) ++ goto out_parent; ++ ++ err = -EOPNOTSUPP; ++ h_dir = au_h_iptr(dir, bindex); ++ if (unlikely(!h_dir->i_op->tmpfile)) ++ goto out_parent; ++ ++ h_mnt = au_sbr_mnt(sb, bindex); ++ err = vfsub_mnt_want_write(h_mnt); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ h_parent = au_h_dptr(parent, bindex); ++ err = inode_permission(d_inode(h_parent), MAY_WRITE | MAY_EXEC); ++ if (unlikely(err)) ++ goto out_mnt; ++ ++ err = -ENOMEM; ++ h_dentry = d_alloc(h_parent, &dentry->d_name); ++ if (unlikely(!h_dentry)) ++ goto out_mnt; ++ ++ err = h_dir->i_op->tmpfile(h_dir, h_dentry, mode); ++ if (unlikely(err)) ++ goto out_dentry; ++ ++ au_set_dbstart(dentry, bindex); ++ au_set_dbend(dentry, bindex); ++ au_set_h_dptr(dentry, bindex, dget(h_dentry)); ++ inode = au_new_inode(dentry, /*must_new*/1); ++ if (IS_ERR(inode)) { ++ err = PTR_ERR(inode); ++ au_set_h_dptr(dentry, bindex, NULL); ++ au_set_dbstart(dentry, -1); ++ au_set_dbend(dentry, -1); ++ } else { ++ if (!inode->i_nlink) ++ set_nlink(inode, 1); ++ d_tmpfile(dentry, inode); ++ au_di(dentry)->di_tmpfile = 1; ++ ++ /* update without i_mutex */ ++ if (au_ibstart(dir) == au_dbstart(dentry)) ++ au_cpup_attr_timesizes(dir); ++ } ++ ++out_dentry: ++ dput(h_dentry); ++out_mnt: ++ vfsub_mnt_drop_write(h_mnt); ++out_parent: ++ di_write_unlock(parent); ++ dput(parent); ++ di_write_unlock(dentry); ++ if (!err) ++#if 0 ++ /* verbose coding for lock class name */ ++ au_rw_class(&au_di(dentry)->di_rwsem, ++ au_lc_key + AuLcNonDir_DIINFO); ++#else ++ ; ++#endif ++ else { ++ au_di_fin(dentry); ++ dentry->d_fsdata = NULL; ++ } ++out_si: ++ si_read_unlock(sb); ++out: ++ mutex_unlock(&dir->i_mutex); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_link_args { ++ aufs_bindex_t bdst, bsrc; ++ struct au_pin pin; ++ struct path h_path; ++ struct dentry *src_parent, *parent; ++}; ++ ++static int au_cpup_before_link(struct dentry *src_dentry, ++ struct au_link_args *a) ++{ ++ int err; ++ struct dentry *h_src_dentry; ++ struct au_cp_generic cpg = { ++ .dentry = src_dentry, ++ .bdst = a->bdst, ++ .bsrc = a->bsrc, ++ .len = -1, ++ .pin = &a->pin, ++ .flags = AuCpup_DTIME | AuCpup_HOPEN /* | AuCpup_KEEPLINO */ ++ }; ++ ++ di_read_lock_parent(a->src_parent, AuLock_IR); ++ err = au_test_and_cpup_dirs(src_dentry, a->bdst); ++ if (unlikely(err)) ++ goto out; ++ ++ h_src_dentry = au_h_dptr(src_dentry, a->bsrc); ++ err = au_pin(&a->pin, src_dentry, a->bdst, ++ au_opt_udba(src_dentry->d_sb), ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_sio_cpup_simple(&cpg); ++ au_unpin(&a->pin); ++ ++out: ++ di_read_unlock(a->src_parent, AuLock_IR); ++ return err; ++} ++ ++static int au_cpup_or_link(struct dentry *src_dentry, struct dentry *dentry, ++ struct au_link_args *a) ++{ ++ int err; ++ unsigned char plink; ++ aufs_bindex_t bend; ++ struct dentry *h_src_dentry; ++ struct inode *h_inode, *inode, *delegated; ++ struct super_block *sb; ++ struct file *h_file; ++ ++ plink = 0; ++ h_inode = NULL; ++ sb = src_dentry->d_sb; ++ inode = d_inode(src_dentry); ++ if (au_ibstart(inode) <= a->bdst) ++ h_inode = au_h_iptr(inode, a->bdst); ++ if (!h_inode || !h_inode->i_nlink) { ++ /* copyup src_dentry as the name of dentry. */ ++ bend = au_dbend(dentry); ++ if (bend < a->bsrc) ++ au_set_dbend(dentry, a->bsrc); ++ au_set_h_dptr(dentry, a->bsrc, ++ dget(au_h_dptr(src_dentry, a->bsrc))); ++ dget(a->h_path.dentry); ++ au_set_h_dptr(dentry, a->bdst, NULL); ++ AuDbg("temporary d_inode...\n"); ++ spin_lock(&dentry->d_lock); ++ dentry->d_inode = d_inode(src_dentry); /* tmp */ ++ spin_unlock(&dentry->d_lock); ++ h_file = au_h_open_pre(dentry, a->bsrc, /*force_wr*/0); ++ if (IS_ERR(h_file)) ++ err = PTR_ERR(h_file); ++ else { ++ struct au_cp_generic cpg = { ++ .dentry = dentry, ++ .bdst = a->bdst, ++ .bsrc = -1, ++ .len = -1, ++ .pin = &a->pin, ++ .flags = AuCpup_KEEPLINO ++ }; ++ err = au_sio_cpup_simple(&cpg); ++ au_h_open_post(dentry, a->bsrc, h_file); ++ if (!err) { ++ dput(a->h_path.dentry); ++ a->h_path.dentry = au_h_dptr(dentry, a->bdst); ++ } else ++ au_set_h_dptr(dentry, a->bdst, ++ a->h_path.dentry); ++ } ++ spin_lock(&dentry->d_lock); ++ dentry->d_inode = NULL; /* restore */ ++ spin_unlock(&dentry->d_lock); ++ AuDbg("temporary d_inode...done\n"); ++ au_set_h_dptr(dentry, a->bsrc, NULL); ++ au_set_dbend(dentry, bend); ++ } else { ++ /* the inode of src_dentry already exists on a.bdst branch */ ++ h_src_dentry = d_find_alias(h_inode); ++ if (!h_src_dentry && au_plink_test(inode)) { ++ plink = 1; ++ h_src_dentry = au_plink_lkup(inode, a->bdst); ++ err = PTR_ERR(h_src_dentry); ++ if (IS_ERR(h_src_dentry)) ++ goto out; ++ ++ if (unlikely(d_is_negative(h_src_dentry))) { ++ dput(h_src_dentry); ++ h_src_dentry = NULL; ++ } ++ ++ } ++ if (h_src_dentry) { ++ delegated = NULL; ++ err = vfsub_link(h_src_dentry, au_pinned_h_dir(&a->pin), ++ &a->h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ dput(h_src_dentry); ++ } else { ++ AuIOErr("no dentry found for hi%lu on b%d\n", ++ h_inode->i_ino, a->bdst); ++ err = -EIO; ++ } ++ } ++ ++ if (!err && !plink) ++ au_plink_append(inode, a->bdst, a->h_path.dentry); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int aufs_link(struct dentry *src_dentry, struct inode *dir, ++ struct dentry *dentry) ++{ ++ int err, rerr; ++ struct au_dtime dt; ++ struct au_link_args *a; ++ struct dentry *wh_dentry, *h_src_dentry; ++ struct inode *inode, *delegated; ++ struct super_block *sb; ++ struct au_wr_dir_args wr_dir_args = { ++ /* .force_btgt = -1, */ ++ .flags = AuWrDir_ADD_ENTRY ++ }; ++ ++ IMustLock(dir); ++ inode = d_inode(src_dentry); ++ IMustLock(inode); ++ ++ err = -ENOMEM; ++ a = kzalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ a->parent = dentry->d_parent; /* dir inode is locked */ ++ err = aufs_read_and_write_lock2(dentry, src_dentry, ++ AuLock_NOPLM | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_kfree; ++ err = au_d_linkable(src_dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ err = au_d_may_add(dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ a->src_parent = dget_parent(src_dentry); ++ wr_dir_args.force_btgt = au_ibstart(inode); ++ ++ di_write_lock_parent(a->parent); ++ wr_dir_args.force_btgt = au_wbr(dentry, wr_dir_args.force_btgt); ++ wh_dentry = lock_hdir_lkup_wh(dentry, &dt, src_dentry, &a->pin, ++ &wr_dir_args); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_parent; ++ ++ err = 0; ++ sb = dentry->d_sb; ++ a->bdst = au_dbstart(dentry); ++ a->h_path.dentry = au_h_dptr(dentry, a->bdst); ++ a->h_path.mnt = au_sbr_mnt(sb, a->bdst); ++ a->bsrc = au_ibstart(inode); ++ h_src_dentry = au_h_d_alias(src_dentry, a->bsrc); ++ if (!h_src_dentry && au_di(src_dentry)->di_tmpfile) ++ h_src_dentry = dget(au_hi_wh(inode, a->bsrc)); ++ if (!h_src_dentry) { ++ a->bsrc = au_dbstart(src_dentry); ++ h_src_dentry = au_h_d_alias(src_dentry, a->bsrc); ++ AuDebugOn(!h_src_dentry); ++ } else if (IS_ERR(h_src_dentry)) { ++ err = PTR_ERR(h_src_dentry); ++ goto out_parent; ++ } ++ ++ if (au_opt_test(au_mntflags(sb), PLINK)) { ++ if (a->bdst < a->bsrc ++ /* && h_src_dentry->d_sb != a->h_path.dentry->d_sb */) ++ err = au_cpup_or_link(src_dentry, dentry, a); ++ else { ++ delegated = NULL; ++ err = vfsub_link(h_src_dentry, au_pinned_h_dir(&a->pin), ++ &a->h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ } ++ dput(h_src_dentry); ++ } else { ++ /* ++ * copyup src_dentry to the branch we process, ++ * and then link(2) to it. ++ */ ++ dput(h_src_dentry); ++ if (a->bdst < a->bsrc ++ /* && h_src_dentry->d_sb != a->h_path.dentry->d_sb */) { ++ au_unpin(&a->pin); ++ di_write_unlock(a->parent); ++ err = au_cpup_before_link(src_dentry, a); ++ di_write_lock_parent(a->parent); ++ if (!err) ++ err = au_pin(&a->pin, dentry, a->bdst, ++ au_opt_udba(sb), ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (unlikely(err)) ++ goto out_wh; ++ } ++ if (!err) { ++ h_src_dentry = au_h_dptr(src_dentry, a->bdst); ++ err = -ENOENT; ++ if (h_src_dentry && d_is_positive(h_src_dentry)) { ++ delegated = NULL; ++ err = vfsub_link(h_src_dentry, ++ au_pinned_h_dir(&a->pin), ++ &a->h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry" ++ " for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ } ++ } ++ } ++ if (unlikely(err)) ++ goto out_unpin; ++ ++ if (wh_dentry) { ++ a->h_path.dentry = wh_dentry; ++ err = au_wh_unlink_dentry(au_pinned_h_dir(&a->pin), &a->h_path, ++ dentry); ++ if (unlikely(err)) ++ goto out_revert; ++ } ++ ++ au_dir_ts(dir, a->bdst); ++ dir->i_version++; ++ inc_nlink(inode); ++ inode->i_ctime = dir->i_ctime; ++ d_instantiate(dentry, au_igrab(inode)); ++ if (d_unhashed(a->h_path.dentry)) ++ /* some filesystem calls d_drop() */ ++ d_drop(dentry); ++ /* some filesystems consume an inode even hardlink */ ++ au_fhsm_wrote(sb, a->bdst, /*force*/0); ++ goto out_unpin; /* success */ ++ ++out_revert: ++ /* no delegation since it is just created */ ++ rerr = vfsub_unlink(au_pinned_h_dir(&a->pin), &a->h_path, ++ /*delegated*/NULL, /*force*/0); ++ if (unlikely(rerr)) { ++ AuIOErr("%pd reverting failed(%d, %d)\n", dentry, err, rerr); ++ err = -EIO; ++ } ++ au_dtime_revert(&dt); ++out_unpin: ++ au_unpin(&a->pin); ++out_wh: ++ dput(wh_dentry); ++out_parent: ++ di_write_unlock(a->parent); ++ dput(a->src_parent); ++out_unlock: ++ if (unlikely(err)) { ++ au_update_dbstart(dentry); ++ d_drop(dentry); ++ } ++ aufs_read_and_write_unlock2(dentry, src_dentry); ++out_kfree: ++ kfree(a); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int aufs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) ++{ ++ int err, rerr; ++ aufs_bindex_t bindex; ++ unsigned char diropq; ++ struct path h_path; ++ struct dentry *wh_dentry, *parent, *opq_dentry; ++ struct mutex *h_mtx; ++ struct super_block *sb; ++ struct { ++ struct au_pin pin; ++ struct au_dtime dt; ++ } *a; /* reduce the stack usage */ ++ struct au_wr_dir_args wr_dir_args = { ++ .force_btgt = -1, ++ .flags = AuWrDir_ADD_ENTRY | AuWrDir_ISDIR ++ }; ++ ++ IMustLock(dir); ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_free; ++ err = au_d_may_add(dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ parent = dentry->d_parent; /* dir inode is locked */ ++ di_write_lock_parent(parent); ++ wh_dentry = lock_hdir_lkup_wh(dentry, &a->dt, /*src_dentry*/NULL, ++ &a->pin, &wr_dir_args); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_parent; ++ ++ sb = dentry->d_sb; ++ bindex = au_dbstart(dentry); ++ h_path.dentry = au_h_dptr(dentry, bindex); ++ h_path.mnt = au_sbr_mnt(sb, bindex); ++ err = vfsub_mkdir(au_pinned_h_dir(&a->pin), &h_path, mode); ++ if (unlikely(err)) ++ goto out_unpin; ++ ++ /* make the dir opaque */ ++ diropq = 0; ++ h_mtx = &d_inode(h_path.dentry)->i_mutex; ++ if (wh_dentry ++ || au_opt_test(au_mntflags(sb), ALWAYS_DIROPQ)) { ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD); ++ opq_dentry = au_diropq_create(dentry, bindex); ++ mutex_unlock(h_mtx); ++ err = PTR_ERR(opq_dentry); ++ if (IS_ERR(opq_dentry)) ++ goto out_dir; ++ dput(opq_dentry); ++ diropq = 1; ++ } ++ ++ err = epilog(dir, bindex, wh_dentry, dentry); ++ if (!err) { ++ inc_nlink(dir); ++ goto out_unpin; /* success */ ++ } ++ ++ /* revert */ ++ if (diropq) { ++ AuLabel(revert opq); ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD); ++ rerr = au_diropq_remove(dentry, bindex); ++ mutex_unlock(h_mtx); ++ if (rerr) { ++ AuIOErr("%pd reverting diropq failed(%d, %d)\n", ++ dentry, err, rerr); ++ err = -EIO; ++ } ++ } ++ ++out_dir: ++ AuLabel(revert dir); ++ rerr = vfsub_rmdir(au_pinned_h_dir(&a->pin), &h_path); ++ if (rerr) { ++ AuIOErr("%pd reverting dir failed(%d, %d)\n", ++ dentry, err, rerr); ++ err = -EIO; ++ } ++ au_dtime_revert(&a->dt); ++out_unpin: ++ au_unpin(&a->pin); ++ dput(wh_dentry); ++out_parent: ++ di_write_unlock(parent); ++out_unlock: ++ if (unlikely(err)) { ++ au_update_dbstart(dentry); ++ d_drop(dentry); ++ } ++ aufs_read_unlock(dentry, AuLock_DW); ++out_free: ++ kfree(a); ++out: ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/i_op.c linux-4.1.10/fs/aufs/i_op.c +--- linux-4.1.10.orig/fs/aufs/i_op.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/i_op.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1447 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode operations (except add/del/rename) ++ */ ++ ++#include <linux/device_cgroup.h> ++#include <linux/fs_stack.h> ++#include <linux/mm.h> ++#include <linux/namei.h> ++#include <linux/security.h> ++#include "aufs.h" ++ ++static int h_permission(struct inode *h_inode, int mask, ++ struct vfsmount *h_mnt, int brperm) ++{ ++ int err; ++ const unsigned char write_mask = !!(mask & (MAY_WRITE | MAY_APPEND)); ++ ++ err = -EACCES; ++ if ((write_mask && IS_IMMUTABLE(h_inode)) ++ || ((mask & MAY_EXEC) ++ && S_ISREG(h_inode->i_mode) ++ && ((h_mnt->mnt_flags & MNT_NOEXEC) ++ || !(h_inode->i_mode & S_IXUGO)))) ++ goto out; ++ ++ /* ++ * - skip the lower fs test in the case of write to ro branch. ++ * - nfs dir permission write check is optimized, but a policy for ++ * link/rename requires a real check. ++ * - nfs always sets MS_POSIXACL regardless its mount option 'noacl.' ++ * in this case, generic_permission() returns -EOPNOTSUPP. ++ */ ++ if ((write_mask && !au_br_writable(brperm)) ++ || (au_test_nfs(h_inode->i_sb) && S_ISDIR(h_inode->i_mode) ++ && write_mask && !(mask & MAY_READ)) ++ || !h_inode->i_op->permission) { ++ /* AuLabel(generic_permission); */ ++ /* AuDbg("get_acl %pf\n", h_inode->i_op->get_acl); */ ++ err = generic_permission(h_inode, mask); ++ if (err == -EOPNOTSUPP && au_test_nfs_noacl(h_inode)) ++ err = h_inode->i_op->permission(h_inode, mask); ++ AuTraceErr(err); ++ } else { ++ /* AuLabel(h_inode->permission); */ ++ err = h_inode->i_op->permission(h_inode, mask); ++ AuTraceErr(err); ++ } ++ ++ if (!err) ++ err = devcgroup_inode_permission(h_inode, mask); ++ if (!err) ++ err = security_inode_permission(h_inode, mask); ++ ++#if 0 ++ if (!err) { ++ /* todo: do we need to call ima_path_check()? */ ++ struct path h_path = { ++ .dentry = ++ .mnt = h_mnt ++ }; ++ err = ima_path_check(&h_path, ++ mask & (MAY_READ | MAY_WRITE | MAY_EXEC), ++ IMA_COUNT_LEAVE); ++ } ++#endif ++ ++out: ++ return err; ++} ++ ++static int aufs_permission(struct inode *inode, int mask) ++{ ++ int err; ++ aufs_bindex_t bindex, bend; ++ const unsigned char isdir = !!S_ISDIR(inode->i_mode), ++ write_mask = !!(mask & (MAY_WRITE | MAY_APPEND)); ++ struct inode *h_inode; ++ struct super_block *sb; ++ struct au_branch *br; ++ ++ /* todo: support rcu-walk? */ ++ if (mask & MAY_NOT_BLOCK) ++ return -ECHILD; ++ ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ ii_read_lock_child(inode); ++#if 0 ++ err = au_iigen_test(inode, au_sigen(sb)); ++ if (unlikely(err)) ++ goto out; ++#endif ++ ++ if (!isdir ++ || write_mask ++ || au_opt_test(au_mntflags(sb), DIRPERM1)) { ++ err = au_busy_or_stale(); ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ if (unlikely(!h_inode ++ || (h_inode->i_mode & S_IFMT) ++ != (inode->i_mode & S_IFMT))) ++ goto out; ++ ++ err = 0; ++ bindex = au_ibstart(inode); ++ br = au_sbr(sb, bindex); ++ err = h_permission(h_inode, mask, au_br_mnt(br), br->br_perm); ++ if (write_mask ++ && !err ++ && !special_file(h_inode->i_mode)) { ++ /* test whether the upper writable branch exists */ ++ err = -EROFS; ++ for (; bindex >= 0; bindex--) ++ if (!au_br_rdonly(au_sbr(sb, bindex))) { ++ err = 0; ++ break; ++ } ++ } ++ goto out; ++ } ++ ++ /* non-write to dir */ ++ err = 0; ++ bend = au_ibend(inode); ++ for (bindex = au_ibstart(inode); !err && bindex <= bend; bindex++) { ++ h_inode = au_h_iptr(inode, bindex); ++ if (h_inode) { ++ err = au_busy_or_stale(); ++ if (unlikely(!S_ISDIR(h_inode->i_mode))) ++ break; ++ ++ br = au_sbr(sb, bindex); ++ err = h_permission(h_inode, mask, au_br_mnt(br), ++ br->br_perm); ++ } ++ } ++ ++out: ++ ii_read_unlock(inode); ++ si_read_unlock(sb); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct dentry *aufs_lookup(struct inode *dir, struct dentry *dentry, ++ unsigned int flags) ++{ ++ struct dentry *ret, *parent; ++ struct inode *inode; ++ struct super_block *sb; ++ int err, npositive; ++ ++ IMustLock(dir); ++ ++ /* todo: support rcu-walk? */ ++ ret = ERR_PTR(-ECHILD); ++ if (flags & LOOKUP_RCU) ++ goto out; ++ ++ ret = ERR_PTR(-ENAMETOOLONG); ++ if (unlikely(dentry->d_name.len > AUFS_MAX_NAMELEN)) ++ goto out; ++ ++ sb = dir->i_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ ret = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_di_init(dentry); ++ ret = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out_si; ++ ++ inode = NULL; ++ npositive = 0; /* suppress a warning */ ++ parent = dentry->d_parent; /* dir inode is locked */ ++ di_read_lock_parent(parent, AuLock_IR); ++ err = au_alive_dir(parent); ++ if (!err) ++ err = au_digen_test(parent, au_sigen(sb)); ++ if (!err) { ++ npositive = au_lkup_dentry(dentry, au_dbstart(parent), ++ /*type*/0); ++ err = npositive; ++ } ++ di_read_unlock(parent, AuLock_IR); ++ ret = ERR_PTR(err); ++ if (unlikely(err < 0)) ++ goto out_unlock; ++ ++ if (npositive) { ++ inode = au_new_inode(dentry, /*must_new*/0); ++ if (IS_ERR(inode)) { ++ ret = (void *)inode; ++ inode = NULL; ++ goto out_unlock; ++ } ++ } ++ ++ if (inode) ++ atomic_inc(&inode->i_count); ++ ret = d_splice_alias(inode, dentry); ++#if 0 ++ if (unlikely(d_need_lookup(dentry))) { ++ spin_lock(&dentry->d_lock); ++ dentry->d_flags &= ~DCACHE_NEED_LOOKUP; ++ spin_unlock(&dentry->d_lock); ++ } else ++#endif ++ if (inode) { ++ if (!IS_ERR(ret)) { ++ iput(inode); ++ if (ret && ret != dentry) ++ ii_write_unlock(inode); ++ } else { ++ ii_write_unlock(inode); ++ iput(inode); ++ inode = NULL; ++ } ++ } ++ ++out_unlock: ++ di_write_unlock(dentry); ++ if (inode) { ++ /* verbose coding for lock class name */ ++ if (unlikely(S_ISLNK(inode->i_mode))) ++ au_rw_class(&au_di(dentry)->di_rwsem, ++ au_lc_key + AuLcSymlink_DIINFO); ++ else if (unlikely(S_ISDIR(inode->i_mode))) ++ au_rw_class(&au_di(dentry)->di_rwsem, ++ au_lc_key + AuLcDir_DIINFO); ++ else /* likely */ ++ au_rw_class(&au_di(dentry)->di_rwsem, ++ au_lc_key + AuLcNonDir_DIINFO); ++ } ++out_si: ++ si_read_unlock(sb); ++out: ++ return ret; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct aopen_node { ++ struct hlist_node hlist; ++ struct file *file, *h_file; ++}; ++ ++static int au_do_aopen(struct inode *inode, struct file *file) ++{ ++ struct au_sphlhead *aopen; ++ struct aopen_node *node; ++ struct au_do_open_args args = { ++ .no_lock = 1, ++ .open = au_do_open_nondir ++ }; ++ ++ aopen = &au_sbi(inode->i_sb)->si_aopen; ++ spin_lock(&aopen->spin); ++ hlist_for_each_entry(node, &aopen->head, hlist) ++ if (node->file == file) { ++ args.h_file = node->h_file; ++ break; ++ } ++ spin_unlock(&aopen->spin); ++ /* AuDebugOn(!args.h_file); */ ++ ++ return au_do_open(file, &args); ++} ++ ++static int aufs_atomic_open(struct inode *dir, struct dentry *dentry, ++ struct file *file, unsigned int open_flag, ++ umode_t create_mode, int *opened) ++{ ++ int err, h_opened = *opened; ++ struct dentry *parent; ++ struct dentry *d; ++ struct au_sphlhead *aopen; ++ struct vfsub_aopen_args args = { ++ .open_flag = open_flag, ++ .create_mode = create_mode, ++ .opened = &h_opened ++ }; ++ struct aopen_node aopen_node = { ++ .file = file ++ }; ++ ++ IMustLock(dir); ++ AuDbg("open_flag 0x%x\n", open_flag); ++ AuDbgDentry(dentry); ++ ++ err = 0; ++ if (!au_di(dentry)) { ++ d = aufs_lookup(dir, dentry, /*flags*/0); ++ if (IS_ERR(d)) { ++ err = PTR_ERR(d); ++ goto out; ++ } else if (d) { ++ /* ++ * obsoleted dentry found. ++ * another error will be returned later. ++ */ ++ d_drop(d); ++ dput(d); ++ AuDbgDentry(d); ++ } ++ AuDbgDentry(dentry); ++ } ++ ++ if (d_is_positive(dentry) ++ || d_unhashed(dentry) ++ || d_unlinked(dentry) ++ || !(open_flag & O_CREAT)) ++ goto out_no_open; ++ ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_GEN); ++ if (unlikely(err)) ++ goto out; ++ ++ parent = dentry->d_parent; /* dir is locked */ ++ di_write_lock_parent(parent); ++ err = au_lkup_dentry(dentry, /*bstart*/0, /*type*/0); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ AuDbgDentry(dentry); ++ if (d_is_positive(dentry)) ++ goto out_unlock; ++ ++ args.file = get_empty_filp(); ++ err = PTR_ERR(args.file); ++ if (IS_ERR(args.file)) ++ goto out_unlock; ++ ++ args.file->f_flags = file->f_flags; ++ err = au_aopen_or_create(dir, dentry, &args); ++ AuTraceErr(err); ++ AuDbgFile(args.file); ++ if (unlikely(err < 0)) { ++ if (h_opened & FILE_OPENED) ++ fput(args.file); ++ else ++ put_filp(args.file); ++ goto out_unlock; ++ } ++ ++ /* some filesystems don't set FILE_CREATED while succeeded? */ ++ *opened |= FILE_CREATED; ++ if (h_opened & FILE_OPENED) ++ aopen_node.h_file = args.file; ++ else { ++ put_filp(args.file); ++ args.file = NULL; ++ } ++ aopen = &au_sbi(dir->i_sb)->si_aopen; ++ au_sphl_add(&aopen_node.hlist, aopen); ++ err = finish_open(file, dentry, au_do_aopen, opened); ++ au_sphl_del(&aopen_node.hlist, aopen); ++ AuTraceErr(err); ++ AuDbgFile(file); ++ if (aopen_node.h_file) ++ fput(aopen_node.h_file); ++ ++out_unlock: ++ di_write_unlock(parent); ++ aufs_read_unlock(dentry, AuLock_DW); ++ AuDbgDentry(dentry); ++ if (unlikely(err)) ++ goto out; ++out_no_open: ++ if (!err && !(*opened & FILE_CREATED)) { ++ AuLabel(out_no_open); ++ dget(dentry); ++ err = finish_no_open(file, dentry); ++ } ++out: ++ AuDbg("%pd%s%s\n", dentry, ++ (*opened & FILE_CREATED) ? " created" : "", ++ (*opened & FILE_OPENED) ? " opened" : ""); ++ AuTraceErr(err); ++ return err; ++} ++ ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_wr_dir_cpup(struct dentry *dentry, struct dentry *parent, ++ const unsigned char add_entry, aufs_bindex_t bcpup, ++ aufs_bindex_t bstart) ++{ ++ int err; ++ struct dentry *h_parent; ++ struct inode *h_dir; ++ ++ if (add_entry) ++ IMustLock(d_inode(parent)); ++ else ++ di_write_lock_parent(parent); ++ ++ err = 0; ++ if (!au_h_dptr(parent, bcpup)) { ++ if (bstart > bcpup) ++ err = au_cpup_dirs(dentry, bcpup); ++ else if (bstart < bcpup) ++ err = au_cpdown_dirs(dentry, bcpup); ++ else ++ BUG(); ++ } ++ if (!err && add_entry && !au_ftest_wrdir(add_entry, TMPFILE)) { ++ h_parent = au_h_dptr(parent, bcpup); ++ h_dir = d_inode(h_parent); ++ mutex_lock_nested(&h_dir->i_mutex, AuLsc_I_PARENT); ++ err = au_lkup_neg(dentry, bcpup, /*wh*/0); ++ /* todo: no unlock here */ ++ mutex_unlock(&h_dir->i_mutex); ++ ++ AuDbg("bcpup %d\n", bcpup); ++ if (!err) { ++ if (d_really_is_negative(dentry)) ++ au_set_h_dptr(dentry, bstart, NULL); ++ au_update_dbrange(dentry, /*do_put_zero*/0); ++ } ++ } ++ ++ if (!add_entry) ++ di_write_unlock(parent); ++ if (!err) ++ err = bcpup; /* success */ ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ++ * decide the branch and the parent dir where we will create a new entry. ++ * returns new bindex or an error. ++ * copyup the parent dir if needed. ++ */ ++int au_wr_dir(struct dentry *dentry, struct dentry *src_dentry, ++ struct au_wr_dir_args *args) ++{ ++ int err; ++ unsigned int flags; ++ aufs_bindex_t bcpup, bstart, src_bstart; ++ const unsigned char add_entry ++ = au_ftest_wrdir(args->flags, ADD_ENTRY) ++ | au_ftest_wrdir(args->flags, TMPFILE); ++ struct super_block *sb; ++ struct dentry *parent; ++ struct au_sbinfo *sbinfo; ++ ++ sb = dentry->d_sb; ++ sbinfo = au_sbi(sb); ++ parent = dget_parent(dentry); ++ bstart = au_dbstart(dentry); ++ bcpup = bstart; ++ if (args->force_btgt < 0) { ++ if (src_dentry) { ++ src_bstart = au_dbstart(src_dentry); ++ if (src_bstart < bstart) ++ bcpup = src_bstart; ++ } else if (add_entry) { ++ flags = 0; ++ if (au_ftest_wrdir(args->flags, ISDIR)) ++ au_fset_wbr(flags, DIR); ++ err = AuWbrCreate(sbinfo, dentry, flags); ++ bcpup = err; ++ } ++ ++ if (bcpup < 0 || au_test_ro(sb, bcpup, d_inode(dentry))) { ++ if (add_entry) ++ err = AuWbrCopyup(sbinfo, dentry); ++ else { ++ if (!IS_ROOT(dentry)) { ++ di_read_lock_parent(parent, !AuLock_IR); ++ err = AuWbrCopyup(sbinfo, dentry); ++ di_read_unlock(parent, !AuLock_IR); ++ } else ++ err = AuWbrCopyup(sbinfo, dentry); ++ } ++ bcpup = err; ++ if (unlikely(err < 0)) ++ goto out; ++ } ++ } else { ++ bcpup = args->force_btgt; ++ AuDebugOn(au_test_ro(sb, bcpup, d_inode(dentry))); ++ } ++ ++ AuDbg("bstart %d, bcpup %d\n", bstart, bcpup); ++ err = bcpup; ++ if (bcpup == bstart) ++ goto out; /* success */ ++ ++ /* copyup the new parent into the branch we process */ ++ err = au_wr_dir_cpup(dentry, parent, add_entry, bcpup, bstart); ++ if (err >= 0) { ++ if (d_really_is_negative(dentry)) { ++ au_set_h_dptr(dentry, bstart, NULL); ++ au_set_dbstart(dentry, bcpup); ++ au_set_dbend(dentry, bcpup); ++ } ++ AuDebugOn(add_entry ++ && !au_ftest_wrdir(args->flags, TMPFILE) ++ && !au_h_dptr(dentry, bcpup)); ++ } ++ ++out: ++ dput(parent); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_pin_hdir_unlock(struct au_pin *p) ++{ ++ if (p->hdir) ++ au_hn_imtx_unlock(p->hdir); ++} ++ ++int au_pin_hdir_lock(struct au_pin *p) ++{ ++ int err; ++ ++ err = 0; ++ if (!p->hdir) ++ goto out; ++ ++ /* even if an error happens later, keep this lock */ ++ au_hn_imtx_lock_nested(p->hdir, p->lsc_hi); ++ ++ err = -EBUSY; ++ if (unlikely(p->hdir->hi_inode != d_inode(p->h_parent))) ++ goto out; ++ ++ err = 0; ++ if (p->h_dentry) ++ err = au_h_verify(p->h_dentry, p->udba, p->hdir->hi_inode, ++ p->h_parent, p->br); ++ ++out: ++ return err; ++} ++ ++int au_pin_hdir_relock(struct au_pin *p) ++{ ++ int err, i; ++ struct inode *h_i; ++ struct dentry *h_d[] = { ++ p->h_dentry, ++ p->h_parent ++ }; ++ ++ err = au_pin_hdir_lock(p); ++ if (unlikely(err)) ++ goto out; ++ ++ for (i = 0; !err && i < sizeof(h_d)/sizeof(*h_d); i++) { ++ if (!h_d[i]) ++ continue; ++ if (d_is_positive(h_d[i])) { ++ h_i = d_inode(h_d[i]); ++ err = !h_i->i_nlink; ++ } ++ } ++ ++out: ++ return err; ++} ++ ++void au_pin_hdir_set_owner(struct au_pin *p, struct task_struct *task) ++{ ++#if defined(CONFIG_DEBUG_MUTEXES) || defined(CONFIG_SMP) ++ p->hdir->hi_inode->i_mutex.owner = task; ++#endif ++} ++ ++void au_pin_hdir_acquire_nest(struct au_pin *p) ++{ ++ if (p->hdir) { ++ mutex_acquire_nest(&p->hdir->hi_inode->i_mutex.dep_map, ++ p->lsc_hi, 0, NULL, _RET_IP_); ++ au_pin_hdir_set_owner(p, current); ++ } ++} ++ ++void au_pin_hdir_release(struct au_pin *p) ++{ ++ if (p->hdir) { ++ au_pin_hdir_set_owner(p, p->task); ++ mutex_release(&p->hdir->hi_inode->i_mutex.dep_map, 1, _RET_IP_); ++ } ++} ++ ++struct dentry *au_pinned_h_parent(struct au_pin *pin) ++{ ++ if (pin && pin->parent) ++ return au_h_dptr(pin->parent, pin->bindex); ++ return NULL; ++} ++ ++void au_unpin(struct au_pin *p) ++{ ++ if (p->hdir) ++ au_pin_hdir_unlock(p); ++ if (p->h_mnt && au_ftest_pin(p->flags, MNT_WRITE)) ++ vfsub_mnt_drop_write(p->h_mnt); ++ if (!p->hdir) ++ return; ++ ++ if (!au_ftest_pin(p->flags, DI_LOCKED)) ++ di_read_unlock(p->parent, AuLock_IR); ++ iput(p->hdir->hi_inode); ++ dput(p->parent); ++ p->parent = NULL; ++ p->hdir = NULL; ++ p->h_mnt = NULL; ++ /* do not clear p->task */ ++} ++ ++int au_do_pin(struct au_pin *p) ++{ ++ int err; ++ struct super_block *sb; ++ struct inode *h_dir; ++ ++ err = 0; ++ sb = p->dentry->d_sb; ++ p->br = au_sbr(sb, p->bindex); ++ if (IS_ROOT(p->dentry)) { ++ if (au_ftest_pin(p->flags, MNT_WRITE)) { ++ p->h_mnt = au_br_mnt(p->br); ++ err = vfsub_mnt_want_write(p->h_mnt); ++ if (unlikely(err)) { ++ au_fclr_pin(p->flags, MNT_WRITE); ++ goto out_err; ++ } ++ } ++ goto out; ++ } ++ ++ p->h_dentry = NULL; ++ if (p->bindex <= au_dbend(p->dentry)) ++ p->h_dentry = au_h_dptr(p->dentry, p->bindex); ++ ++ p->parent = dget_parent(p->dentry); ++ if (!au_ftest_pin(p->flags, DI_LOCKED)) ++ di_read_lock(p->parent, AuLock_IR, p->lsc_di); ++ ++ h_dir = NULL; ++ p->h_parent = au_h_dptr(p->parent, p->bindex); ++ p->hdir = au_hi(d_inode(p->parent), p->bindex); ++ if (p->hdir) ++ h_dir = p->hdir->hi_inode; ++ ++ /* ++ * udba case, or ++ * if DI_LOCKED is not set, then p->parent may be different ++ * and h_parent can be NULL. ++ */ ++ if (unlikely(!p->hdir || !h_dir || !p->h_parent)) { ++ err = -EBUSY; ++ if (!au_ftest_pin(p->flags, DI_LOCKED)) ++ di_read_unlock(p->parent, AuLock_IR); ++ dput(p->parent); ++ p->parent = NULL; ++ goto out_err; ++ } ++ ++ if (au_ftest_pin(p->flags, MNT_WRITE)) { ++ p->h_mnt = au_br_mnt(p->br); ++ err = vfsub_mnt_want_write(p->h_mnt); ++ if (unlikely(err)) { ++ au_fclr_pin(p->flags, MNT_WRITE); ++ if (!au_ftest_pin(p->flags, DI_LOCKED)) ++ di_read_unlock(p->parent, AuLock_IR); ++ dput(p->parent); ++ p->parent = NULL; ++ goto out_err; ++ } ++ } ++ ++ au_igrab(h_dir); ++ err = au_pin_hdir_lock(p); ++ if (!err) ++ goto out; /* success */ ++ ++ au_unpin(p); ++ ++out_err: ++ pr_err("err %d\n", err); ++ err = au_busy_or_stale(); ++out: ++ return err; ++} ++ ++void au_pin_init(struct au_pin *p, struct dentry *dentry, ++ aufs_bindex_t bindex, int lsc_di, int lsc_hi, ++ unsigned int udba, unsigned char flags) ++{ ++ p->dentry = dentry; ++ p->udba = udba; ++ p->lsc_di = lsc_di; ++ p->lsc_hi = lsc_hi; ++ p->flags = flags; ++ p->bindex = bindex; ++ ++ p->parent = NULL; ++ p->hdir = NULL; ++ p->h_mnt = NULL; ++ ++ p->h_dentry = NULL; ++ p->h_parent = NULL; ++ p->br = NULL; ++ p->task = current; ++} ++ ++int au_pin(struct au_pin *pin, struct dentry *dentry, aufs_bindex_t bindex, ++ unsigned int udba, unsigned char flags) ++{ ++ au_pin_init(pin, dentry, bindex, AuLsc_DI_PARENT, AuLsc_I_PARENT2, ++ udba, flags); ++ return au_do_pin(pin); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * ->setattr() and ->getattr() are called in various cases. ++ * chmod, stat: dentry is revalidated. ++ * fchmod, fstat: file and dentry are not revalidated, additionally they may be ++ * unhashed. ++ * for ->setattr(), ia->ia_file is passed from ftruncate only. ++ */ ++/* todo: consolidate with do_refresh() and simple_reval_dpath() */ ++int au_reval_for_attr(struct dentry *dentry, unsigned int sigen) ++{ ++ int err; ++ struct dentry *parent; ++ ++ err = 0; ++ if (au_digen_test(dentry, sigen)) { ++ parent = dget_parent(dentry); ++ di_read_lock_parent(parent, AuLock_IR); ++ err = au_refresh_dentry(dentry, parent); ++ di_read_unlock(parent, AuLock_IR); ++ dput(parent); ++ } ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_pin_and_icpup(struct dentry *dentry, struct iattr *ia, ++ struct au_icpup_args *a) ++{ ++ int err; ++ loff_t sz; ++ aufs_bindex_t bstart, ibstart; ++ struct dentry *hi_wh, *parent; ++ struct inode *inode; ++ struct au_wr_dir_args wr_dir_args = { ++ .force_btgt = -1, ++ .flags = 0 ++ }; ++ ++ if (d_is_dir(dentry)) ++ au_fset_wrdir(wr_dir_args.flags, ISDIR); ++ /* plink or hi_wh() case */ ++ bstart = au_dbstart(dentry); ++ inode = d_inode(dentry); ++ ibstart = au_ibstart(inode); ++ if (bstart != ibstart && !au_test_ro(inode->i_sb, ibstart, inode)) ++ wr_dir_args.force_btgt = ibstart; ++ err = au_wr_dir(dentry, /*src_dentry*/NULL, &wr_dir_args); ++ if (unlikely(err < 0)) ++ goto out; ++ a->btgt = err; ++ if (err != bstart) ++ au_fset_icpup(a->flags, DID_CPUP); ++ ++ err = 0; ++ a->pin_flags = AuPin_MNT_WRITE; ++ parent = NULL; ++ if (!IS_ROOT(dentry)) { ++ au_fset_pin(a->pin_flags, DI_LOCKED); ++ parent = dget_parent(dentry); ++ di_write_lock_parent(parent); ++ } ++ ++ err = au_pin(&a->pin, dentry, a->btgt, a->udba, a->pin_flags); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ a->h_path.dentry = au_h_dptr(dentry, bstart); ++ sz = -1; ++ a->h_inode = d_inode(a->h_path.dentry); ++ if (ia && (ia->ia_valid & ATTR_SIZE)) { ++ mutex_lock_nested(&a->h_inode->i_mutex, AuLsc_I_CHILD); ++ if (ia->ia_size < i_size_read(a->h_inode)) ++ sz = ia->ia_size; ++ mutex_unlock(&a->h_inode->i_mutex); ++ } ++ ++ hi_wh = NULL; ++ if (au_ftest_icpup(a->flags, DID_CPUP) && d_unlinked(dentry)) { ++ hi_wh = au_hi_wh(inode, a->btgt); ++ if (!hi_wh) { ++ struct au_cp_generic cpg = { ++ .dentry = dentry, ++ .bdst = a->btgt, ++ .bsrc = -1, ++ .len = sz, ++ .pin = &a->pin ++ }; ++ err = au_sio_cpup_wh(&cpg, /*file*/NULL); ++ if (unlikely(err)) ++ goto out_unlock; ++ hi_wh = au_hi_wh(inode, a->btgt); ++ /* todo: revalidate hi_wh? */ ++ } ++ } ++ ++ if (parent) { ++ au_pin_set_parent_lflag(&a->pin, /*lflag*/0); ++ di_downgrade_lock(parent, AuLock_IR); ++ dput(parent); ++ parent = NULL; ++ } ++ if (!au_ftest_icpup(a->flags, DID_CPUP)) ++ goto out; /* success */ ++ ++ if (!d_unhashed(dentry)) { ++ struct au_cp_generic cpg = { ++ .dentry = dentry, ++ .bdst = a->btgt, ++ .bsrc = bstart, ++ .len = sz, ++ .pin = &a->pin, ++ .flags = AuCpup_DTIME | AuCpup_HOPEN ++ }; ++ err = au_sio_cpup_simple(&cpg); ++ if (!err) ++ a->h_path.dentry = au_h_dptr(dentry, a->btgt); ++ } else if (!hi_wh) ++ a->h_path.dentry = au_h_dptr(dentry, a->btgt); ++ else ++ a->h_path.dentry = hi_wh; /* do not dget here */ ++ ++out_unlock: ++ a->h_inode = d_inode(a->h_path.dentry); ++ if (!err) ++ goto out; /* success */ ++ au_unpin(&a->pin); ++out_parent: ++ if (parent) { ++ di_write_unlock(parent); ++ dput(parent); ++ } ++out: ++ if (!err) ++ mutex_lock_nested(&a->h_inode->i_mutex, AuLsc_I_CHILD); ++ return err; ++} ++ ++static int aufs_setattr(struct dentry *dentry, struct iattr *ia) ++{ ++ int err; ++ struct inode *inode, *delegated; ++ struct super_block *sb; ++ struct file *file; ++ struct au_icpup_args *a; ++ ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ ++ err = -ENOMEM; ++ a = kzalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ if (ia->ia_valid & (ATTR_KILL_SUID | ATTR_KILL_SGID)) ++ ia->ia_valid &= ~ATTR_MODE; ++ ++ file = NULL; ++ sb = dentry->d_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out_kfree; ++ ++ if (ia->ia_valid & ATTR_FILE) { ++ /* currently ftruncate(2) only */ ++ AuDebugOn(!d_is_reg(dentry)); ++ file = ia->ia_file; ++ err = au_reval_and_lock_fdi(file, au_reopen_nondir, /*wlock*/1); ++ if (unlikely(err)) ++ goto out_si; ++ ia->ia_file = au_hf_top(file); ++ a->udba = AuOpt_UDBA_NONE; ++ } else { ++ /* fchmod() doesn't pass ia_file */ ++ a->udba = au_opt_udba(sb); ++ di_write_lock_child(dentry); ++ /* no d_unlinked(), to set UDBA_NONE for root */ ++ if (d_unhashed(dentry)) ++ a->udba = AuOpt_UDBA_NONE; ++ if (a->udba != AuOpt_UDBA_NONE) { ++ AuDebugOn(IS_ROOT(dentry)); ++ err = au_reval_for_attr(dentry, au_sigen(sb)); ++ if (unlikely(err)) ++ goto out_dentry; ++ } ++ } ++ ++ err = au_pin_and_icpup(dentry, ia, a); ++ if (unlikely(err < 0)) ++ goto out_dentry; ++ if (au_ftest_icpup(a->flags, DID_CPUP)) { ++ ia->ia_file = NULL; ++ ia->ia_valid &= ~ATTR_FILE; ++ } ++ ++ a->h_path.mnt = au_sbr_mnt(sb, a->btgt); ++ if ((ia->ia_valid & (ATTR_MODE | ATTR_CTIME)) ++ == (ATTR_MODE | ATTR_CTIME)) { ++ err = security_path_chmod(&a->h_path, ia->ia_mode); ++ if (unlikely(err)) ++ goto out_unlock; ++ } else if ((ia->ia_valid & (ATTR_UID | ATTR_GID)) ++ && (ia->ia_valid & ATTR_CTIME)) { ++ err = security_path_chown(&a->h_path, ia->ia_uid, ia->ia_gid); ++ if (unlikely(err)) ++ goto out_unlock; ++ } ++ ++ if (ia->ia_valid & ATTR_SIZE) { ++ struct file *f; ++ ++ if (ia->ia_size < i_size_read(inode)) ++ /* unmap only */ ++ truncate_setsize(inode, ia->ia_size); ++ ++ f = NULL; ++ if (ia->ia_valid & ATTR_FILE) ++ f = ia->ia_file; ++ mutex_unlock(&a->h_inode->i_mutex); ++ err = vfsub_trunc(&a->h_path, ia->ia_size, ia->ia_valid, f); ++ mutex_lock_nested(&a->h_inode->i_mutex, AuLsc_I_CHILD); ++ } else { ++ delegated = NULL; ++ while (1) { ++ err = vfsub_notify_change(&a->h_path, ia, &delegated); ++ if (delegated) { ++ err = break_deleg_wait(&delegated); ++ if (!err) ++ continue; ++ } ++ break; ++ } ++ } ++ if (!err) ++ au_cpup_attr_changeable(inode); ++ ++out_unlock: ++ mutex_unlock(&a->h_inode->i_mutex); ++ au_unpin(&a->pin); ++ if (unlikely(err)) ++ au_update_dbstart(dentry); ++out_dentry: ++ di_write_unlock(dentry); ++ if (file) { ++ fi_write_unlock(file); ++ ia->ia_file = file; ++ ia->ia_valid |= ATTR_FILE; ++ } ++out_si: ++ si_read_unlock(sb); ++out_kfree: ++ kfree(a); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++#if IS_ENABLED(CONFIG_AUFS_XATTR) || IS_ENABLED(CONFIG_FS_POSIX_ACL) ++static int au_h_path_to_set_attr(struct dentry *dentry, ++ struct au_icpup_args *a, struct path *h_path) ++{ ++ int err; ++ struct super_block *sb; ++ ++ sb = dentry->d_sb; ++ a->udba = au_opt_udba(sb); ++ /* no d_unlinked(), to set UDBA_NONE for root */ ++ if (d_unhashed(dentry)) ++ a->udba = AuOpt_UDBA_NONE; ++ if (a->udba != AuOpt_UDBA_NONE) { ++ AuDebugOn(IS_ROOT(dentry)); ++ err = au_reval_for_attr(dentry, au_sigen(sb)); ++ if (unlikely(err)) ++ goto out; ++ } ++ err = au_pin_and_icpup(dentry, /*ia*/NULL, a); ++ if (unlikely(err < 0)) ++ goto out; ++ ++ h_path->dentry = a->h_path.dentry; ++ h_path->mnt = au_sbr_mnt(sb, a->btgt); ++ ++out: ++ return err; ++} ++ ++ssize_t au_srxattr(struct dentry *dentry, struct au_srxattr *arg) ++{ ++ int err; ++ struct path h_path; ++ struct super_block *sb; ++ struct au_icpup_args *a; ++ struct inode *inode, *h_inode; ++ ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ ++ err = -ENOMEM; ++ a = kzalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out_kfree; ++ ++ h_path.dentry = NULL; /* silence gcc */ ++ di_write_lock_child(dentry); ++ err = au_h_path_to_set_attr(dentry, a, &h_path); ++ if (unlikely(err)) ++ goto out_di; ++ ++ mutex_unlock(&a->h_inode->i_mutex); ++ switch (arg->type) { ++ case AU_XATTR_SET: ++ err = vfsub_setxattr(h_path.dentry, ++ arg->u.set.name, arg->u.set.value, ++ arg->u.set.size, arg->u.set.flags); ++ break; ++ case AU_XATTR_REMOVE: ++ err = vfsub_removexattr(h_path.dentry, arg->u.remove.name); ++ break; ++ case AU_ACL_SET: ++ err = -EOPNOTSUPP; ++ h_inode = d_inode(h_path.dentry); ++ if (h_inode->i_op->set_acl) ++ err = h_inode->i_op->set_acl(h_inode, ++ arg->u.acl_set.acl, ++ arg->u.acl_set.type); ++ break; ++ } ++ if (!err) ++ au_cpup_attr_timesizes(inode); ++ ++ au_unpin(&a->pin); ++ if (unlikely(err)) ++ au_update_dbstart(dentry); ++ ++out_di: ++ di_write_unlock(dentry); ++ si_read_unlock(sb); ++out_kfree: ++ kfree(a); ++out: ++ AuTraceErr(err); ++ return err; ++} ++#endif ++ ++static void au_refresh_iattr(struct inode *inode, struct kstat *st, ++ unsigned int nlink) ++{ ++ unsigned int n; ++ ++ inode->i_mode = st->mode; ++ /* don't i_[ug]id_write() here */ ++ inode->i_uid = st->uid; ++ inode->i_gid = st->gid; ++ inode->i_atime = st->atime; ++ inode->i_mtime = st->mtime; ++ inode->i_ctime = st->ctime; ++ ++ au_cpup_attr_nlink(inode, /*force*/0); ++ if (S_ISDIR(inode->i_mode)) { ++ n = inode->i_nlink; ++ n -= nlink; ++ n += st->nlink; ++ smp_mb(); /* for i_nlink */ ++ /* 0 can happen */ ++ set_nlink(inode, n); ++ } ++ ++ spin_lock(&inode->i_lock); ++ inode->i_blocks = st->blocks; ++ i_size_write(inode, st->size); ++ spin_unlock(&inode->i_lock); ++} ++ ++/* ++ * common routine for aufs_getattr() and aufs_getxattr(). ++ * returns zero or negative (an error). ++ * @dentry will be read-locked in success. ++ */ ++int au_h_path_getattr(struct dentry *dentry, int force, struct path *h_path) ++{ ++ int err; ++ unsigned int mnt_flags, sigen; ++ unsigned char udba_none; ++ aufs_bindex_t bindex; ++ struct super_block *sb, *h_sb; ++ struct inode *inode; ++ ++ h_path->mnt = NULL; ++ h_path->dentry = NULL; ++ ++ err = 0; ++ sb = dentry->d_sb; ++ mnt_flags = au_mntflags(sb); ++ udba_none = !!au_opt_test(mnt_flags, UDBA_NONE); ++ ++ /* support fstat(2) */ ++ if (!d_unlinked(dentry) && !udba_none) { ++ sigen = au_sigen(sb); ++ err = au_digen_test(dentry, sigen); ++ if (!err) { ++ di_read_lock_child(dentry, AuLock_IR); ++ err = au_dbrange_test(dentry); ++ if (unlikely(err)) { ++ di_read_unlock(dentry, AuLock_IR); ++ goto out; ++ } ++ } else { ++ AuDebugOn(IS_ROOT(dentry)); ++ di_write_lock_child(dentry); ++ err = au_dbrange_test(dentry); ++ if (!err) ++ err = au_reval_for_attr(dentry, sigen); ++ if (!err) ++ di_downgrade_lock(dentry, AuLock_IR); ++ else { ++ di_write_unlock(dentry); ++ goto out; ++ } ++ } ++ } else ++ di_read_lock_child(dentry, AuLock_IR); ++ ++ inode = d_inode(dentry); ++ bindex = au_ibstart(inode); ++ h_path->mnt = au_sbr_mnt(sb, bindex); ++ h_sb = h_path->mnt->mnt_sb; ++ if (!force ++ && !au_test_fs_bad_iattr(h_sb) ++ && udba_none) ++ goto out; /* success */ ++ ++ if (au_dbstart(dentry) == bindex) ++ h_path->dentry = au_h_dptr(dentry, bindex); ++ else if (au_opt_test(mnt_flags, PLINK) && au_plink_test(inode)) { ++ h_path->dentry = au_plink_lkup(inode, bindex); ++ if (IS_ERR(h_path->dentry)) ++ /* pretending success */ ++ h_path->dentry = NULL; ++ else ++ dput(h_path->dentry); ++ } ++ ++out: ++ return err; ++} ++ ++static int aufs_getattr(struct vfsmount *mnt __maybe_unused, ++ struct dentry *dentry, struct kstat *st) ++{ ++ int err; ++ unsigned char positive; ++ struct path h_path; ++ struct inode *inode; ++ struct super_block *sb; ++ ++ inode = d_inode(dentry); ++ sb = dentry->d_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out; ++ err = au_h_path_getattr(dentry, /*force*/0, &h_path); ++ if (unlikely(err)) ++ goto out_si; ++ if (unlikely(!h_path.dentry)) ++ /* illegally overlapped or something */ ++ goto out_fill; /* pretending success */ ++ ++ positive = d_is_positive(h_path.dentry); ++ if (positive) ++ err = vfs_getattr(&h_path, st); ++ if (!err) { ++ if (positive) ++ au_refresh_iattr(inode, st, ++ d_inode(h_path.dentry)->i_nlink); ++ goto out_fill; /* success */ ++ } ++ AuTraceErr(err); ++ goto out_di; ++ ++out_fill: ++ generic_fillattr(inode, st); ++out_di: ++ di_read_unlock(dentry, AuLock_IR); ++out_si: ++ si_read_unlock(sb); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int h_readlink(struct dentry *dentry, int bindex, char __user *buf, ++ int bufsiz) ++{ ++ int err; ++ struct super_block *sb; ++ struct dentry *h_dentry; ++ struct inode *inode, *h_inode; ++ ++ err = -EINVAL; ++ h_dentry = au_h_dptr(dentry, bindex); ++ h_inode = d_inode(h_dentry); ++ if (unlikely(!h_inode->i_op->readlink)) ++ goto out; ++ ++ err = security_inode_readlink(h_dentry); ++ if (unlikely(err)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ inode = d_inode(dentry); ++ if (!au_test_ro(sb, bindex, inode)) { ++ vfsub_touch_atime(au_sbr_mnt(sb, bindex), h_dentry); ++ fsstack_copy_attr_atime(inode, h_inode); ++ } ++ err = h_inode->i_op->readlink(h_dentry, buf, bufsiz); ++ ++out: ++ return err; ++} ++ ++static int aufs_readlink(struct dentry *dentry, char __user *buf, int bufsiz) ++{ ++ int err; ++ ++ err = aufs_read_lock(dentry, AuLock_IR | AuLock_GEN); ++ if (unlikely(err)) ++ goto out; ++ err = au_d_hashed_positive(dentry); ++ if (!err) ++ err = h_readlink(dentry, au_dbstart(dentry), buf, bufsiz); ++ aufs_read_unlock(dentry, AuLock_IR); ++ ++out: ++ return err; ++} ++ ++static void *aufs_follow_link(struct dentry *dentry, struct nameidata *nd) ++{ ++ int err; ++ mm_segment_t old_fs; ++ union { ++ char *k; ++ char __user *u; ++ } buf; ++ ++ err = -ENOMEM; ++ buf.k = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!buf.k)) ++ goto out; ++ ++ err = aufs_read_lock(dentry, AuLock_IR | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_name; ++ ++ err = au_d_hashed_positive(dentry); ++ if (!err) { ++ old_fs = get_fs(); ++ set_fs(KERNEL_DS); ++ err = h_readlink(dentry, au_dbstart(dentry), buf.u, PATH_MAX); ++ set_fs(old_fs); ++ } ++ aufs_read_unlock(dentry, AuLock_IR); ++ ++ if (err >= 0) { ++ buf.k[err] = 0; ++ /* will be freed by put_link */ ++ nd_set_link(nd, buf.k); ++ return NULL; /* success */ ++ } ++ ++out_name: ++ free_page((unsigned long)buf.k); ++out: ++ AuTraceErr(err); ++ return ERR_PTR(err); ++} ++ ++static void aufs_put_link(struct dentry *dentry __maybe_unused, ++ struct nameidata *nd, void *cookie __maybe_unused) ++{ ++ char *p; ++ ++ p = nd_get_link(nd); ++ if (!IS_ERR_OR_NULL(p)) ++ free_page((unsigned long)p); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_update_time(struct inode *inode, struct timespec *ts, int flags) ++{ ++ int err; ++ struct super_block *sb; ++ struct inode *h_inode; ++ ++ sb = inode->i_sb; ++ /* mmap_sem might be acquired already, cf. aufs_mmap() */ ++ lockdep_off(); ++ si_read_lock(sb, AuLock_FLUSH); ++ ii_write_lock_child(inode); ++ lockdep_on(); ++ h_inode = au_h_iptr(inode, au_ibstart(inode)); ++ err = vfsub_update_time(h_inode, ts, flags); ++ lockdep_off(); ++ if (!err) ++ au_cpup_attr_timesizes(inode); ++ ii_write_unlock(inode); ++ si_read_unlock(sb); ++ lockdep_on(); ++ ++ if (!err && (flags & S_VERSION)) ++ inode_inc_iversion(inode); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct inode_operations aufs_symlink_iop = { ++ .permission = aufs_permission, ++#ifdef CONFIG_FS_POSIX_ACL ++ .get_acl = aufs_get_acl, ++ .set_acl = aufs_set_acl, /* unsupport for symlink? */ ++#endif ++ ++ .setattr = aufs_setattr, ++ .getattr = aufs_getattr, ++ ++#ifdef CONFIG_AUFS_XATTR ++ .setxattr = aufs_setxattr, ++ .getxattr = aufs_getxattr, ++ .listxattr = aufs_listxattr, ++ .removexattr = aufs_removexattr, ++#endif ++ ++ .readlink = aufs_readlink, ++ .follow_link = aufs_follow_link, ++ .put_link = aufs_put_link, ++ ++ /* .update_time = aufs_update_time */ ++}; ++ ++struct inode_operations aufs_dir_iop = { ++ .create = aufs_create, ++ .lookup = aufs_lookup, ++ .link = aufs_link, ++ .unlink = aufs_unlink, ++ .symlink = aufs_symlink, ++ .mkdir = aufs_mkdir, ++ .rmdir = aufs_rmdir, ++ .mknod = aufs_mknod, ++ .rename = aufs_rename, ++ ++ .permission = aufs_permission, ++#ifdef CONFIG_FS_POSIX_ACL ++ .get_acl = aufs_get_acl, ++ .set_acl = aufs_set_acl, ++#endif ++ ++ .setattr = aufs_setattr, ++ .getattr = aufs_getattr, ++ ++#ifdef CONFIG_AUFS_XATTR ++ .setxattr = aufs_setxattr, ++ .getxattr = aufs_getxattr, ++ .listxattr = aufs_listxattr, ++ .removexattr = aufs_removexattr, ++#endif ++ ++ .update_time = aufs_update_time, ++ .atomic_open = aufs_atomic_open, ++ .tmpfile = aufs_tmpfile ++}; ++ ++struct inode_operations aufs_iop = { ++ .permission = aufs_permission, ++#ifdef CONFIG_FS_POSIX_ACL ++ .get_acl = aufs_get_acl, ++ .set_acl = aufs_set_acl, ++#endif ++ ++ .setattr = aufs_setattr, ++ .getattr = aufs_getattr, ++ ++#ifdef CONFIG_AUFS_XATTR ++ .setxattr = aufs_setxattr, ++ .getxattr = aufs_getxattr, ++ .listxattr = aufs_listxattr, ++ .removexattr = aufs_removexattr, ++#endif ++ ++ .update_time = aufs_update_time ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/i_op_del.c linux-4.1.10/fs/aufs/i_op_del.c +--- linux-4.1.10.orig/fs/aufs/i_op_del.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/i_op_del.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,510 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode operations (del entry) ++ */ ++ ++#include "aufs.h" ++ ++/* ++ * decide if a new whiteout for @dentry is necessary or not. ++ * when it is necessary, prepare the parent dir for the upper branch whose ++ * branch index is @bcpup for creation. the actual creation of the whiteout will ++ * be done by caller. ++ * return value: ++ * 0: wh is unnecessary ++ * plus: wh is necessary ++ * minus: error ++ */ ++int au_wr_dir_need_wh(struct dentry *dentry, int isdir, aufs_bindex_t *bcpup) ++{ ++ int need_wh, err; ++ aufs_bindex_t bstart; ++ struct super_block *sb; ++ ++ sb = dentry->d_sb; ++ bstart = au_dbstart(dentry); ++ if (*bcpup < 0) { ++ *bcpup = bstart; ++ if (au_test_ro(sb, bstart, d_inode(dentry))) { ++ err = AuWbrCopyup(au_sbi(sb), dentry); ++ *bcpup = err; ++ if (unlikely(err < 0)) ++ goto out; ++ } ++ } else ++ AuDebugOn(bstart < *bcpup ++ || au_test_ro(sb, *bcpup, d_inode(dentry))); ++ AuDbg("bcpup %d, bstart %d\n", *bcpup, bstart); ++ ++ if (*bcpup != bstart) { ++ err = au_cpup_dirs(dentry, *bcpup); ++ if (unlikely(err)) ++ goto out; ++ need_wh = 1; ++ } else { ++ struct au_dinfo *dinfo, *tmp; ++ ++ need_wh = -ENOMEM; ++ dinfo = au_di(dentry); ++ tmp = au_di_alloc(sb, AuLsc_DI_TMP); ++ if (tmp) { ++ au_di_cp(tmp, dinfo); ++ au_di_swap(tmp, dinfo); ++ /* returns the number of positive dentries */ ++ need_wh = au_lkup_dentry(dentry, bstart + 1, /*type*/0); ++ au_di_swap(tmp, dinfo); ++ au_rw_write_unlock(&tmp->di_rwsem); ++ au_di_free(tmp); ++ } ++ } ++ AuDbg("need_wh %d\n", need_wh); ++ err = need_wh; ++ ++out: ++ return err; ++} ++ ++/* ++ * simple tests for the del-entry operations. ++ * following the checks in vfs, plus the parent-child relationship. ++ */ ++int au_may_del(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent, int isdir) ++{ ++ int err; ++ umode_t h_mode; ++ struct dentry *h_dentry, *h_latest; ++ struct inode *h_inode; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (d_really_is_positive(dentry)) { ++ err = -ENOENT; ++ if (unlikely(d_is_negative(h_dentry))) ++ goto out; ++ h_inode = d_inode(h_dentry); ++ if (unlikely(!h_inode->i_nlink)) ++ goto out; ++ ++ h_mode = h_inode->i_mode; ++ if (!isdir) { ++ err = -EISDIR; ++ if (unlikely(S_ISDIR(h_mode))) ++ goto out; ++ } else if (unlikely(!S_ISDIR(h_mode))) { ++ err = -ENOTDIR; ++ goto out; ++ } ++ } else { ++ /* rename(2) case */ ++ err = -EIO; ++ if (unlikely(d_is_positive(h_dentry))) ++ goto out; ++ } ++ ++ err = -ENOENT; ++ /* expected parent dir is locked */ ++ if (unlikely(h_parent != h_dentry->d_parent)) ++ goto out; ++ err = 0; ++ ++ /* ++ * rmdir a dir may break the consistency on some filesystem. ++ * let's try heavy test. ++ */ ++ err = -EACCES; ++ if (unlikely(!au_opt_test(au_mntflags(dentry->d_sb), DIRPERM1) ++ && au_test_h_perm(d_inode(h_parent), ++ MAY_EXEC | MAY_WRITE))) ++ goto out; ++ ++ h_latest = au_sio_lkup_one(&dentry->d_name, h_parent); ++ err = -EIO; ++ if (IS_ERR(h_latest)) ++ goto out; ++ if (h_latest == h_dentry) ++ err = 0; ++ dput(h_latest); ++ ++out: ++ return err; ++} ++ ++/* ++ * decide the branch where we operate for @dentry. the branch index will be set ++ * @rbcpup. after diciding it, 'pin' it and store the timestamps of the parent ++ * dir for reverting. ++ * when a new whiteout is necessary, create it. ++ */ ++static struct dentry* ++lock_hdir_create_wh(struct dentry *dentry, int isdir, aufs_bindex_t *rbcpup, ++ struct au_dtime *dt, struct au_pin *pin) ++{ ++ struct dentry *wh_dentry; ++ struct super_block *sb; ++ struct path h_path; ++ int err, need_wh; ++ unsigned int udba; ++ aufs_bindex_t bcpup; ++ ++ need_wh = au_wr_dir_need_wh(dentry, isdir, rbcpup); ++ wh_dentry = ERR_PTR(need_wh); ++ if (unlikely(need_wh < 0)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ udba = au_opt_udba(sb); ++ bcpup = *rbcpup; ++ err = au_pin(pin, dentry, bcpup, udba, ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ wh_dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out; ++ ++ h_path.dentry = au_pinned_h_parent(pin); ++ if (udba != AuOpt_UDBA_NONE ++ && au_dbstart(dentry) == bcpup) { ++ err = au_may_del(dentry, bcpup, h_path.dentry, isdir); ++ wh_dentry = ERR_PTR(err); ++ if (unlikely(err)) ++ goto out_unpin; ++ } ++ ++ h_path.mnt = au_sbr_mnt(sb, bcpup); ++ au_dtime_store(dt, au_pinned_parent(pin), &h_path); ++ wh_dentry = NULL; ++ if (!need_wh) ++ goto out; /* success, no need to create whiteout */ ++ ++ wh_dentry = au_wh_create(dentry, bcpup, h_path.dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_unpin; ++ ++ /* returns with the parent is locked and wh_dentry is dget-ed */ ++ goto out; /* success */ ++ ++out_unpin: ++ au_unpin(pin); ++out: ++ return wh_dentry; ++} ++ ++/* ++ * when removing a dir, rename it to a unique temporary whiteout-ed name first ++ * in order to be revertible and save time for removing many child whiteouts ++ * under the dir. ++ * returns 1 when there are too many child whiteout and caller should remove ++ * them asynchronously. returns 0 when the number of children is enough small to ++ * remove now or the branch fs is a remote fs. ++ * otherwise return an error. ++ */ ++static int renwh_and_rmdir(struct dentry *dentry, aufs_bindex_t bindex, ++ struct au_nhash *whlist, struct inode *dir) ++{ ++ int rmdir_later, err, dirwh; ++ struct dentry *h_dentry; ++ struct super_block *sb; ++ struct inode *inode; ++ ++ sb = dentry->d_sb; ++ SiMustAnyLock(sb); ++ h_dentry = au_h_dptr(dentry, bindex); ++ err = au_whtmp_ren(h_dentry, au_sbr(sb, bindex)); ++ if (unlikely(err)) ++ goto out; ++ ++ /* stop monitoring */ ++ inode = d_inode(dentry); ++ au_hn_free(au_hi(inode, bindex)); ++ ++ if (!au_test_fs_remote(h_dentry->d_sb)) { ++ dirwh = au_sbi(sb)->si_dirwh; ++ rmdir_later = (dirwh <= 1); ++ if (!rmdir_later) ++ rmdir_later = au_nhash_test_longer_wh(whlist, bindex, ++ dirwh); ++ if (rmdir_later) ++ return rmdir_later; ++ } ++ ++ err = au_whtmp_rmdir(dir, bindex, h_dentry, whlist); ++ if (unlikely(err)) { ++ AuIOErr("rmdir %pd, b%d failed, %d. ignored\n", ++ h_dentry, bindex, err); ++ err = 0; ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ++ * final procedure for deleting a entry. ++ * maintain dentry and iattr. ++ */ ++static void epilog(struct inode *dir, struct dentry *dentry, ++ aufs_bindex_t bindex) ++{ ++ struct inode *inode; ++ ++ inode = d_inode(dentry); ++ d_drop(dentry); ++ inode->i_ctime = dir->i_ctime; ++ ++ au_dir_ts(dir, bindex); ++ dir->i_version++; ++} ++ ++/* ++ * when an error happened, remove the created whiteout and revert everything. ++ */ ++static int do_revert(int err, struct inode *dir, aufs_bindex_t bindex, ++ aufs_bindex_t bwh, struct dentry *wh_dentry, ++ struct dentry *dentry, struct au_dtime *dt) ++{ ++ int rerr; ++ struct path h_path = { ++ .dentry = wh_dentry, ++ .mnt = au_sbr_mnt(dir->i_sb, bindex) ++ }; ++ ++ rerr = au_wh_unlink_dentry(au_h_iptr(dir, bindex), &h_path, dentry); ++ if (!rerr) { ++ au_set_dbwh(dentry, bwh); ++ au_dtime_revert(dt); ++ return 0; ++ } ++ ++ AuIOErr("%pd reverting whiteout failed(%d, %d)\n", dentry, err, rerr); ++ return -EIO; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int aufs_unlink(struct inode *dir, struct dentry *dentry) ++{ ++ int err; ++ aufs_bindex_t bwh, bindex, bstart; ++ struct inode *inode, *h_dir, *delegated; ++ struct dentry *parent, *wh_dentry; ++ /* to reuduce stack size */ ++ struct { ++ struct au_dtime dt; ++ struct au_pin pin; ++ struct path h_path; ++ } *a; ++ ++ IMustLock(dir); ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_free; ++ err = au_d_hashed_positive(dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ err = -EISDIR; ++ if (unlikely(d_is_dir(dentry))) ++ goto out_unlock; /* possible? */ ++ ++ bstart = au_dbstart(dentry); ++ bwh = au_dbwh(dentry); ++ bindex = -1; ++ parent = dentry->d_parent; /* dir inode is locked */ ++ di_write_lock_parent(parent); ++ wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/0, &bindex, &a->dt, ++ &a->pin); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_parent; ++ ++ a->h_path.mnt = au_sbr_mnt(dentry->d_sb, bstart); ++ a->h_path.dentry = au_h_dptr(dentry, bstart); ++ dget(a->h_path.dentry); ++ if (bindex == bstart) { ++ h_dir = au_pinned_h_dir(&a->pin); ++ delegated = NULL; ++ err = vfsub_unlink(h_dir, &a->h_path, &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ } else { ++ /* dir inode is locked */ ++ h_dir = d_inode(wh_dentry->d_parent); ++ IMustLock(h_dir); ++ err = 0; ++ } ++ ++ if (!err) { ++ vfsub_drop_nlink(inode); ++ epilog(dir, dentry, bindex); ++ ++ /* update target timestamps */ ++ if (bindex == bstart) { ++ vfsub_update_h_iattr(&a->h_path, /*did*/NULL); ++ /*ignore*/ ++ inode->i_ctime = d_inode(a->h_path.dentry)->i_ctime; ++ } else ++ /* todo: this timestamp may be reverted later */ ++ inode->i_ctime = h_dir->i_ctime; ++ goto out_unpin; /* success */ ++ } ++ ++ /* revert */ ++ if (wh_dentry) { ++ int rerr; ++ ++ rerr = do_revert(err, dir, bindex, bwh, wh_dentry, dentry, ++ &a->dt); ++ if (rerr) ++ err = rerr; ++ } ++ ++out_unpin: ++ au_unpin(&a->pin); ++ dput(wh_dentry); ++ dput(a->h_path.dentry); ++out_parent: ++ di_write_unlock(parent); ++out_unlock: ++ aufs_read_unlock(dentry, AuLock_DW); ++out_free: ++ kfree(a); ++out: ++ return err; ++} ++ ++int aufs_rmdir(struct inode *dir, struct dentry *dentry) ++{ ++ int err, rmdir_later; ++ aufs_bindex_t bwh, bindex, bstart; ++ struct inode *inode; ++ struct dentry *parent, *wh_dentry, *h_dentry; ++ struct au_whtmp_rmdir *args; ++ /* to reuduce stack size */ ++ struct { ++ struct au_dtime dt; ++ struct au_pin pin; ++ } *a; ++ ++ IMustLock(dir); ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH | AuLock_GEN); ++ if (unlikely(err)) ++ goto out_free; ++ err = au_alive_dir(dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ inode = d_inode(dentry); ++ IMustLock(inode); ++ err = -ENOTDIR; ++ if (unlikely(!d_is_dir(dentry))) ++ goto out_unlock; /* possible? */ ++ ++ err = -ENOMEM; ++ args = au_whtmp_rmdir_alloc(dir->i_sb, GFP_NOFS); ++ if (unlikely(!args)) ++ goto out_unlock; ++ ++ parent = dentry->d_parent; /* dir inode is locked */ ++ di_write_lock_parent(parent); ++ err = au_test_empty(dentry, &args->whlist); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ bstart = au_dbstart(dentry); ++ bwh = au_dbwh(dentry); ++ bindex = -1; ++ wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/1, &bindex, &a->dt, ++ &a->pin); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) ++ goto out_parent; ++ ++ h_dentry = au_h_dptr(dentry, bstart); ++ dget(h_dentry); ++ rmdir_later = 0; ++ if (bindex == bstart) { ++ err = renwh_and_rmdir(dentry, bstart, &args->whlist, dir); ++ if (err > 0) { ++ rmdir_later = err; ++ err = 0; ++ } ++ } else { ++ /* stop monitoring */ ++ au_hn_free(au_hi(inode, bstart)); ++ ++ /* dir inode is locked */ ++ IMustLock(d_inode(wh_dentry->d_parent)); ++ err = 0; ++ } ++ ++ if (!err) { ++ vfsub_dead_dir(inode); ++ au_set_dbdiropq(dentry, -1); ++ epilog(dir, dentry, bindex); ++ ++ if (rmdir_later) { ++ au_whtmp_kick_rmdir(dir, bstart, h_dentry, args); ++ args = NULL; ++ } ++ ++ goto out_unpin; /* success */ ++ } ++ ++ /* revert */ ++ AuLabel(revert); ++ if (wh_dentry) { ++ int rerr; ++ ++ rerr = do_revert(err, dir, bindex, bwh, wh_dentry, dentry, ++ &a->dt); ++ if (rerr) ++ err = rerr; ++ } ++ ++out_unpin: ++ au_unpin(&a->pin); ++ dput(wh_dentry); ++ dput(h_dentry); ++out_parent: ++ di_write_unlock(parent); ++ if (args) ++ au_whtmp_rmdir_free(args); ++out_unlock: ++ aufs_read_unlock(dentry, AuLock_DW); ++out_free: ++ kfree(a); ++out: ++ AuTraceErr(err); ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/i_op_ren.c linux-4.1.10/fs/aufs/i_op_ren.c +--- linux-4.1.10.orig/fs/aufs/i_op_ren.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/i_op_ren.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1017 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * inode operation (rename entry) ++ * todo: this is crazy monster ++ */ ++ ++#include "aufs.h" ++ ++enum { AuSRC, AuDST, AuSrcDst }; ++enum { AuPARENT, AuCHILD, AuParentChild }; ++ ++#define AuRen_ISDIR 1 ++#define AuRen_ISSAMEDIR (1 << 1) ++#define AuRen_WHSRC (1 << 2) ++#define AuRen_WHDST (1 << 3) ++#define AuRen_MNT_WRITE (1 << 4) ++#define AuRen_DT_DSTDIR (1 << 5) ++#define AuRen_DIROPQ (1 << 6) ++#define au_ftest_ren(flags, name) ((flags) & AuRen_##name) ++#define au_fset_ren(flags, name) \ ++ do { (flags) |= AuRen_##name; } while (0) ++#define au_fclr_ren(flags, name) \ ++ do { (flags) &= ~AuRen_##name; } while (0) ++ ++struct au_ren_args { ++ struct { ++ struct dentry *dentry, *h_dentry, *parent, *h_parent, ++ *wh_dentry; ++ struct inode *dir, *inode; ++ struct au_hinode *hdir; ++ struct au_dtime dt[AuParentChild]; ++ aufs_bindex_t bstart; ++ } sd[AuSrcDst]; ++ ++#define src_dentry sd[AuSRC].dentry ++#define src_dir sd[AuSRC].dir ++#define src_inode sd[AuSRC].inode ++#define src_h_dentry sd[AuSRC].h_dentry ++#define src_parent sd[AuSRC].parent ++#define src_h_parent sd[AuSRC].h_parent ++#define src_wh_dentry sd[AuSRC].wh_dentry ++#define src_hdir sd[AuSRC].hdir ++#define src_h_dir sd[AuSRC].hdir->hi_inode ++#define src_dt sd[AuSRC].dt ++#define src_bstart sd[AuSRC].bstart ++ ++#define dst_dentry sd[AuDST].dentry ++#define dst_dir sd[AuDST].dir ++#define dst_inode sd[AuDST].inode ++#define dst_h_dentry sd[AuDST].h_dentry ++#define dst_parent sd[AuDST].parent ++#define dst_h_parent sd[AuDST].h_parent ++#define dst_wh_dentry sd[AuDST].wh_dentry ++#define dst_hdir sd[AuDST].hdir ++#define dst_h_dir sd[AuDST].hdir->hi_inode ++#define dst_dt sd[AuDST].dt ++#define dst_bstart sd[AuDST].bstart ++ ++ struct dentry *h_trap; ++ struct au_branch *br; ++ struct au_hinode *src_hinode; ++ struct path h_path; ++ struct au_nhash whlist; ++ aufs_bindex_t btgt, src_bwh, src_bdiropq; ++ ++ unsigned int flags; ++ ++ struct au_whtmp_rmdir *thargs; ++ struct dentry *h_dst; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * functions for reverting. ++ * when an error happened in a single rename systemcall, we should revert ++ * everything as if nothing happened. ++ * we don't need to revert the copied-up/down the parent dir since they are ++ * harmless. ++ */ ++ ++#define RevertFailure(fmt, ...) do { \ ++ AuIOErr("revert failure: " fmt " (%d, %d)\n", \ ++ ##__VA_ARGS__, err, rerr); \ ++ err = -EIO; \ ++} while (0) ++ ++static void au_ren_rev_diropq(int err, struct au_ren_args *a) ++{ ++ int rerr; ++ ++ au_hn_imtx_lock_nested(a->src_hinode, AuLsc_I_CHILD); ++ rerr = au_diropq_remove(a->src_dentry, a->btgt); ++ au_hn_imtx_unlock(a->src_hinode); ++ au_set_dbdiropq(a->src_dentry, a->src_bdiropq); ++ if (rerr) ++ RevertFailure("remove diropq %pd", a->src_dentry); ++} ++ ++static void au_ren_rev_rename(int err, struct au_ren_args *a) ++{ ++ int rerr; ++ struct inode *delegated; ++ ++ a->h_path.dentry = vfsub_lkup_one(&a->src_dentry->d_name, ++ a->src_h_parent); ++ rerr = PTR_ERR(a->h_path.dentry); ++ if (IS_ERR(a->h_path.dentry)) { ++ RevertFailure("lkup one %pd", a->src_dentry); ++ return; ++ } ++ ++ delegated = NULL; ++ rerr = vfsub_rename(a->dst_h_dir, ++ au_h_dptr(a->src_dentry, a->btgt), ++ a->src_h_dir, &a->h_path, &delegated); ++ if (unlikely(rerr == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal rename\n"); ++ iput(delegated); ++ } ++ d_drop(a->h_path.dentry); ++ dput(a->h_path.dentry); ++ /* au_set_h_dptr(a->src_dentry, a->btgt, NULL); */ ++ if (rerr) ++ RevertFailure("rename %pd", a->src_dentry); ++} ++ ++static void au_ren_rev_whtmp(int err, struct au_ren_args *a) ++{ ++ int rerr; ++ struct inode *delegated; ++ ++ a->h_path.dentry = vfsub_lkup_one(&a->dst_dentry->d_name, ++ a->dst_h_parent); ++ rerr = PTR_ERR(a->h_path.dentry); ++ if (IS_ERR(a->h_path.dentry)) { ++ RevertFailure("lkup one %pd", a->dst_dentry); ++ return; ++ } ++ if (d_is_positive(a->h_path.dentry)) { ++ d_drop(a->h_path.dentry); ++ dput(a->h_path.dentry); ++ return; ++ } ++ ++ delegated = NULL; ++ rerr = vfsub_rename(a->dst_h_dir, a->h_dst, a->dst_h_dir, &a->h_path, ++ &delegated); ++ if (unlikely(rerr == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal rename\n"); ++ iput(delegated); ++ } ++ d_drop(a->h_path.dentry); ++ dput(a->h_path.dentry); ++ if (!rerr) ++ au_set_h_dptr(a->dst_dentry, a->btgt, dget(a->h_dst)); ++ else ++ RevertFailure("rename %pd", a->h_dst); ++} ++ ++static void au_ren_rev_whsrc(int err, struct au_ren_args *a) ++{ ++ int rerr; ++ ++ a->h_path.dentry = a->src_wh_dentry; ++ rerr = au_wh_unlink_dentry(a->src_h_dir, &a->h_path, a->src_dentry); ++ au_set_dbwh(a->src_dentry, a->src_bwh); ++ if (rerr) ++ RevertFailure("unlink %pd", a->src_wh_dentry); ++} ++#undef RevertFailure ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * when we have to copyup the renaming entry, do it with the rename-target name ++ * in order to minimize the cost (the later actual rename is unnecessary). ++ * otherwise rename it on the target branch. ++ */ ++static int au_ren_or_cpup(struct au_ren_args *a) ++{ ++ int err; ++ struct dentry *d; ++ struct inode *delegated; ++ ++ d = a->src_dentry; ++ if (au_dbstart(d) == a->btgt) { ++ a->h_path.dentry = a->dst_h_dentry; ++ if (au_ftest_ren(a->flags, DIROPQ) ++ && au_dbdiropq(d) == a->btgt) ++ au_fclr_ren(a->flags, DIROPQ); ++ AuDebugOn(au_dbstart(d) != a->btgt); ++ delegated = NULL; ++ err = vfsub_rename(a->src_h_dir, au_h_dptr(d, a->btgt), ++ a->dst_h_dir, &a->h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal rename\n"); ++ iput(delegated); ++ } ++ } else ++ BUG(); ++ ++ if (!err && a->h_dst) ++ /* it will be set to dinfo later */ ++ dget(a->h_dst); ++ ++ return err; ++} ++ ++/* cf. aufs_rmdir() */ ++static int au_ren_del_whtmp(struct au_ren_args *a) ++{ ++ int err; ++ struct inode *dir; ++ ++ dir = a->dst_dir; ++ SiMustAnyLock(dir->i_sb); ++ if (!au_nhash_test_longer_wh(&a->whlist, a->btgt, ++ au_sbi(dir->i_sb)->si_dirwh) ++ || au_test_fs_remote(a->h_dst->d_sb)) { ++ err = au_whtmp_rmdir(dir, a->btgt, a->h_dst, &a->whlist); ++ if (unlikely(err)) ++ pr_warn("failed removing whtmp dir %pd (%d), " ++ "ignored.\n", a->h_dst, err); ++ } else { ++ au_nhash_wh_free(&a->thargs->whlist); ++ a->thargs->whlist = a->whlist; ++ a->whlist.nh_num = 0; ++ au_whtmp_kick_rmdir(dir, a->btgt, a->h_dst, a->thargs); ++ dput(a->h_dst); ++ a->thargs = NULL; ++ } ++ ++ return 0; ++} ++ ++/* make it 'opaque' dir. */ ++static int au_ren_diropq(struct au_ren_args *a) ++{ ++ int err; ++ struct dentry *diropq; ++ ++ err = 0; ++ a->src_bdiropq = au_dbdiropq(a->src_dentry); ++ a->src_hinode = au_hi(a->src_inode, a->btgt); ++ au_hn_imtx_lock_nested(a->src_hinode, AuLsc_I_CHILD); ++ diropq = au_diropq_create(a->src_dentry, a->btgt); ++ au_hn_imtx_unlock(a->src_hinode); ++ if (IS_ERR(diropq)) ++ err = PTR_ERR(diropq); ++ else ++ dput(diropq); ++ ++ return err; ++} ++ ++static int do_rename(struct au_ren_args *a) ++{ ++ int err; ++ struct dentry *d, *h_d; ++ ++ /* prepare workqueue args for asynchronous rmdir */ ++ h_d = a->dst_h_dentry; ++ if (au_ftest_ren(a->flags, ISDIR) && d_is_positive(h_d)) { ++ err = -ENOMEM; ++ a->thargs = au_whtmp_rmdir_alloc(a->src_dentry->d_sb, GFP_NOFS); ++ if (unlikely(!a->thargs)) ++ goto out; ++ a->h_dst = dget(h_d); ++ } ++ ++ /* create whiteout for src_dentry */ ++ if (au_ftest_ren(a->flags, WHSRC)) { ++ a->src_bwh = au_dbwh(a->src_dentry); ++ AuDebugOn(a->src_bwh >= 0); ++ a->src_wh_dentry ++ = au_wh_create(a->src_dentry, a->btgt, a->src_h_parent); ++ err = PTR_ERR(a->src_wh_dentry); ++ if (IS_ERR(a->src_wh_dentry)) ++ goto out_thargs; ++ } ++ ++ /* lookup whiteout for dentry */ ++ if (au_ftest_ren(a->flags, WHDST)) { ++ h_d = au_wh_lkup(a->dst_h_parent, &a->dst_dentry->d_name, ++ a->br); ++ err = PTR_ERR(h_d); ++ if (IS_ERR(h_d)) ++ goto out_whsrc; ++ if (d_is_negative(h_d)) ++ dput(h_d); ++ else ++ a->dst_wh_dentry = h_d; ++ } ++ ++ /* rename dentry to tmpwh */ ++ if (a->thargs) { ++ err = au_whtmp_ren(a->dst_h_dentry, a->br); ++ if (unlikely(err)) ++ goto out_whdst; ++ ++ d = a->dst_dentry; ++ au_set_h_dptr(d, a->btgt, NULL); ++ err = au_lkup_neg(d, a->btgt, /*wh*/0); ++ if (unlikely(err)) ++ goto out_whtmp; ++ a->dst_h_dentry = au_h_dptr(d, a->btgt); ++ } ++ ++ BUG_ON(d_is_positive(a->dst_h_dentry) && a->src_bstart != a->btgt); ++ ++ /* rename by vfs_rename or cpup */ ++ d = a->dst_dentry; ++ if (au_ftest_ren(a->flags, ISDIR) ++ && (a->dst_wh_dentry ++ || au_dbdiropq(d) == a->btgt ++ /* hide the lower to keep xino */ ++ || a->btgt < au_dbend(d) ++ || au_opt_test(au_mntflags(d->d_sb), ALWAYS_DIROPQ))) ++ au_fset_ren(a->flags, DIROPQ); ++ err = au_ren_or_cpup(a); ++ if (unlikely(err)) ++ /* leave the copied-up one */ ++ goto out_whtmp; ++ ++ /* make dir opaque */ ++ if (au_ftest_ren(a->flags, DIROPQ)) { ++ err = au_ren_diropq(a); ++ if (unlikely(err)) ++ goto out_rename; ++ } ++ ++ /* update target timestamps */ ++ AuDebugOn(au_dbstart(a->src_dentry) != a->btgt); ++ a->h_path.dentry = au_h_dptr(a->src_dentry, a->btgt); ++ vfsub_update_h_iattr(&a->h_path, /*did*/NULL); /*ignore*/ ++ a->src_inode->i_ctime = d_inode(a->h_path.dentry)->i_ctime; ++ ++ /* remove whiteout for dentry */ ++ if (a->dst_wh_dentry) { ++ a->h_path.dentry = a->dst_wh_dentry; ++ err = au_wh_unlink_dentry(a->dst_h_dir, &a->h_path, ++ a->dst_dentry); ++ if (unlikely(err)) ++ goto out_diropq; ++ } ++ ++ /* remove whtmp */ ++ if (a->thargs) ++ au_ren_del_whtmp(a); /* ignore this error */ ++ ++ au_fhsm_wrote(a->src_dentry->d_sb, a->btgt, /*force*/0); ++ err = 0; ++ goto out_success; ++ ++out_diropq: ++ if (au_ftest_ren(a->flags, DIROPQ)) ++ au_ren_rev_diropq(err, a); ++out_rename: ++ au_ren_rev_rename(err, a); ++ dput(a->h_dst); ++out_whtmp: ++ if (a->thargs) ++ au_ren_rev_whtmp(err, a); ++out_whdst: ++ dput(a->dst_wh_dentry); ++ a->dst_wh_dentry = NULL; ++out_whsrc: ++ if (a->src_wh_dentry) ++ au_ren_rev_whsrc(err, a); ++out_success: ++ dput(a->src_wh_dentry); ++ dput(a->dst_wh_dentry); ++out_thargs: ++ if (a->thargs) { ++ dput(a->h_dst); ++ au_whtmp_rmdir_free(a->thargs); ++ a->thargs = NULL; ++ } ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * test if @dentry dir can be rename destination or not. ++ * success means, it is a logically empty dir. ++ */ ++static int may_rename_dstdir(struct dentry *dentry, struct au_nhash *whlist) ++{ ++ return au_test_empty(dentry, whlist); ++} ++ ++/* ++ * test if @dentry dir can be rename source or not. ++ * if it can, return 0 and @children is filled. ++ * success means, ++ * - it is a logically empty dir. ++ * - or, it exists on writable branch and has no children including whiteouts ++ * on the lower branch. ++ */ ++static int may_rename_srcdir(struct dentry *dentry, aufs_bindex_t btgt) ++{ ++ int err; ++ unsigned int rdhash; ++ aufs_bindex_t bstart; ++ ++ bstart = au_dbstart(dentry); ++ if (bstart != btgt) { ++ struct au_nhash whlist; ++ ++ SiMustAnyLock(dentry->d_sb); ++ rdhash = au_sbi(dentry->d_sb)->si_rdhash; ++ if (!rdhash) ++ rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, ++ dentry)); ++ err = au_nhash_alloc(&whlist, rdhash, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_test_empty(dentry, &whlist); ++ au_nhash_wh_free(&whlist); ++ goto out; ++ } ++ ++ if (bstart == au_dbtaildir(dentry)) ++ return 0; /* success */ ++ ++ err = au_test_empty_lower(dentry); ++ ++out: ++ if (err == -ENOTEMPTY) { ++ AuWarn1("renaming dir who has child(ren) on multiple branches," ++ " is not supported\n"); ++ err = -EXDEV; ++ } ++ return err; ++} ++ ++/* side effect: sets whlist and h_dentry */ ++static int au_ren_may_dir(struct au_ren_args *a) ++{ ++ int err; ++ unsigned int rdhash; ++ struct dentry *d; ++ ++ d = a->dst_dentry; ++ SiMustAnyLock(d->d_sb); ++ ++ err = 0; ++ if (au_ftest_ren(a->flags, ISDIR) && a->dst_inode) { ++ rdhash = au_sbi(d->d_sb)->si_rdhash; ++ if (!rdhash) ++ rdhash = au_rdhash_est(au_dir_size(/*file*/NULL, d)); ++ err = au_nhash_alloc(&a->whlist, rdhash, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ ++ au_set_dbstart(d, a->dst_bstart); ++ err = may_rename_dstdir(d, &a->whlist); ++ au_set_dbstart(d, a->btgt); ++ } ++ a->dst_h_dentry = au_h_dptr(d, au_dbstart(d)); ++ if (unlikely(err)) ++ goto out; ++ ++ d = a->src_dentry; ++ a->src_h_dentry = au_h_dptr(d, au_dbstart(d)); ++ if (au_ftest_ren(a->flags, ISDIR)) { ++ err = may_rename_srcdir(d, a->btgt); ++ if (unlikely(err)) { ++ au_nhash_wh_free(&a->whlist); ++ a->whlist.nh_num = 0; ++ } ++ } ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * simple tests for rename. ++ * following the checks in vfs, plus the parent-child relationship. ++ */ ++static int au_may_ren(struct au_ren_args *a) ++{ ++ int err, isdir; ++ struct inode *h_inode; ++ ++ if (a->src_bstart == a->btgt) { ++ err = au_may_del(a->src_dentry, a->btgt, a->src_h_parent, ++ au_ftest_ren(a->flags, ISDIR)); ++ if (unlikely(err)) ++ goto out; ++ err = -EINVAL; ++ if (unlikely(a->src_h_dentry == a->h_trap)) ++ goto out; ++ } ++ ++ err = 0; ++ if (a->dst_bstart != a->btgt) ++ goto out; ++ ++ err = -ENOTEMPTY; ++ if (unlikely(a->dst_h_dentry == a->h_trap)) ++ goto out; ++ ++ err = -EIO; ++ isdir = !!au_ftest_ren(a->flags, ISDIR); ++ if (d_really_is_negative(a->dst_dentry)) { ++ if (d_is_negative(a->dst_h_dentry)) ++ err = au_may_add(a->dst_dentry, a->btgt, ++ a->dst_h_parent, isdir); ++ } else { ++ if (unlikely(d_is_negative(a->dst_h_dentry))) ++ goto out; ++ h_inode = d_inode(a->dst_h_dentry); ++ if (h_inode->i_nlink) ++ err = au_may_del(a->dst_dentry, a->btgt, ++ a->dst_h_parent, isdir); ++ } ++ ++out: ++ if (unlikely(err == -ENOENT || err == -EEXIST)) ++ err = -EIO; ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * locking order ++ * (VFS) ++ * - src_dir and dir by lock_rename() ++ * - inode if exitsts ++ * (aufs) ++ * - lock all ++ * + src_dentry and dentry by aufs_read_and_write_lock2() which calls, ++ * + si_read_lock ++ * + di_write_lock2_child() ++ * + di_write_lock_child() ++ * + ii_write_lock_child() ++ * + di_write_lock_child2() ++ * + ii_write_lock_child2() ++ * + src_parent and parent ++ * + di_write_lock_parent() ++ * + ii_write_lock_parent() ++ * + di_write_lock_parent2() ++ * + ii_write_lock_parent2() ++ * + lower src_dir and dir by vfsub_lock_rename() ++ * + verify the every relationships between child and parent. if any ++ * of them failed, unlock all and return -EBUSY. ++ */ ++static void au_ren_unlock(struct au_ren_args *a) ++{ ++ vfsub_unlock_rename(a->src_h_parent, a->src_hdir, ++ a->dst_h_parent, a->dst_hdir); ++ if (au_ftest_ren(a->flags, MNT_WRITE)) ++ vfsub_mnt_drop_write(au_br_mnt(a->br)); ++} ++ ++static int au_ren_lock(struct au_ren_args *a) ++{ ++ int err; ++ unsigned int udba; ++ ++ err = 0; ++ a->src_h_parent = au_h_dptr(a->src_parent, a->btgt); ++ a->src_hdir = au_hi(a->src_dir, a->btgt); ++ a->dst_h_parent = au_h_dptr(a->dst_parent, a->btgt); ++ a->dst_hdir = au_hi(a->dst_dir, a->btgt); ++ ++ err = vfsub_mnt_want_write(au_br_mnt(a->br)); ++ if (unlikely(err)) ++ goto out; ++ au_fset_ren(a->flags, MNT_WRITE); ++ a->h_trap = vfsub_lock_rename(a->src_h_parent, a->src_hdir, ++ a->dst_h_parent, a->dst_hdir); ++ udba = au_opt_udba(a->src_dentry->d_sb); ++ if (unlikely(a->src_hdir->hi_inode != d_inode(a->src_h_parent) ++ || a->dst_hdir->hi_inode != d_inode(a->dst_h_parent))) ++ err = au_busy_or_stale(); ++ if (!err && au_dbstart(a->src_dentry) == a->btgt) ++ err = au_h_verify(a->src_h_dentry, udba, ++ d_inode(a->src_h_parent), a->src_h_parent, ++ a->br); ++ if (!err && au_dbstart(a->dst_dentry) == a->btgt) ++ err = au_h_verify(a->dst_h_dentry, udba, ++ d_inode(a->dst_h_parent), a->dst_h_parent, ++ a->br); ++ if (!err) ++ goto out; /* success */ ++ ++ err = au_busy_or_stale(); ++ au_ren_unlock(a); ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_ren_refresh_dir(struct au_ren_args *a) ++{ ++ struct inode *dir; ++ ++ dir = a->dst_dir; ++ dir->i_version++; ++ if (au_ftest_ren(a->flags, ISDIR)) { ++ /* is this updating defined in POSIX? */ ++ au_cpup_attr_timesizes(a->src_inode); ++ au_cpup_attr_nlink(dir, /*force*/1); ++ } ++ ++ au_dir_ts(dir, a->btgt); ++ ++ if (au_ftest_ren(a->flags, ISSAMEDIR)) ++ return; ++ ++ dir = a->src_dir; ++ dir->i_version++; ++ if (au_ftest_ren(a->flags, ISDIR)) ++ au_cpup_attr_nlink(dir, /*force*/1); ++ au_dir_ts(dir, a->btgt); ++} ++ ++static void au_ren_refresh(struct au_ren_args *a) ++{ ++ aufs_bindex_t bend, bindex; ++ struct dentry *d, *h_d; ++ struct inode *i, *h_i; ++ struct super_block *sb; ++ ++ d = a->dst_dentry; ++ d_drop(d); ++ if (a->h_dst) ++ /* already dget-ed by au_ren_or_cpup() */ ++ au_set_h_dptr(d, a->btgt, a->h_dst); ++ ++ i = a->dst_inode; ++ if (i) { ++ if (!au_ftest_ren(a->flags, ISDIR)) ++ vfsub_drop_nlink(i); ++ else { ++ vfsub_dead_dir(i); ++ au_cpup_attr_timesizes(i); ++ } ++ au_update_dbrange(d, /*do_put_zero*/1); ++ } else { ++ bend = a->btgt; ++ for (bindex = au_dbstart(d); bindex < bend; bindex++) ++ au_set_h_dptr(d, bindex, NULL); ++ bend = au_dbend(d); ++ for (bindex = a->btgt + 1; bindex <= bend; bindex++) ++ au_set_h_dptr(d, bindex, NULL); ++ au_update_dbrange(d, /*do_put_zero*/0); ++ } ++ ++ d = a->src_dentry; ++ au_set_dbwh(d, -1); ++ bend = au_dbend(d); ++ for (bindex = a->btgt + 1; bindex <= bend; bindex++) { ++ h_d = au_h_dptr(d, bindex); ++ if (h_d) ++ au_set_h_dptr(d, bindex, NULL); ++ } ++ au_set_dbend(d, a->btgt); ++ ++ sb = d->d_sb; ++ i = a->src_inode; ++ if (au_opt_test(au_mntflags(sb), PLINK) && au_plink_test(i)) ++ return; /* success */ ++ ++ bend = au_ibend(i); ++ for (bindex = a->btgt + 1; bindex <= bend; bindex++) { ++ h_i = au_h_iptr(i, bindex); ++ if (h_i) { ++ au_xino_write(sb, bindex, h_i->i_ino, /*ino*/0); ++ /* ignore this error */ ++ au_set_h_iptr(i, bindex, NULL, 0); ++ } ++ } ++ au_set_ibend(i, a->btgt); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* mainly for link(2) and rename(2) */ ++int au_wbr(struct dentry *dentry, aufs_bindex_t btgt) ++{ ++ aufs_bindex_t bdiropq, bwh; ++ struct dentry *parent; ++ struct au_branch *br; ++ ++ parent = dentry->d_parent; ++ IMustLock(d_inode(parent)); /* dir is locked */ ++ ++ bdiropq = au_dbdiropq(parent); ++ bwh = au_dbwh(dentry); ++ br = au_sbr(dentry->d_sb, btgt); ++ if (au_br_rdonly(br) ++ || (0 <= bdiropq && bdiropq < btgt) ++ || (0 <= bwh && bwh < btgt)) ++ btgt = -1; ++ ++ AuDbg("btgt %d\n", btgt); ++ return btgt; ++} ++ ++/* sets src_bstart, dst_bstart and btgt */ ++static int au_ren_wbr(struct au_ren_args *a) ++{ ++ int err; ++ struct au_wr_dir_args wr_dir_args = { ++ /* .force_btgt = -1, */ ++ .flags = AuWrDir_ADD_ENTRY ++ }; ++ ++ a->src_bstart = au_dbstart(a->src_dentry); ++ a->dst_bstart = au_dbstart(a->dst_dentry); ++ if (au_ftest_ren(a->flags, ISDIR)) ++ au_fset_wrdir(wr_dir_args.flags, ISDIR); ++ wr_dir_args.force_btgt = a->src_bstart; ++ if (a->dst_inode && a->dst_bstart < a->src_bstart) ++ wr_dir_args.force_btgt = a->dst_bstart; ++ wr_dir_args.force_btgt = au_wbr(a->dst_dentry, wr_dir_args.force_btgt); ++ err = au_wr_dir(a->dst_dentry, a->src_dentry, &wr_dir_args); ++ a->btgt = err; ++ ++ return err; ++} ++ ++static void au_ren_dt(struct au_ren_args *a) ++{ ++ a->h_path.dentry = a->src_h_parent; ++ au_dtime_store(a->src_dt + AuPARENT, a->src_parent, &a->h_path); ++ if (!au_ftest_ren(a->flags, ISSAMEDIR)) { ++ a->h_path.dentry = a->dst_h_parent; ++ au_dtime_store(a->dst_dt + AuPARENT, a->dst_parent, &a->h_path); ++ } ++ ++ au_fclr_ren(a->flags, DT_DSTDIR); ++ if (!au_ftest_ren(a->flags, ISDIR)) ++ return; ++ ++ a->h_path.dentry = a->src_h_dentry; ++ au_dtime_store(a->src_dt + AuCHILD, a->src_dentry, &a->h_path); ++ if (d_is_positive(a->dst_h_dentry)) { ++ au_fset_ren(a->flags, DT_DSTDIR); ++ a->h_path.dentry = a->dst_h_dentry; ++ au_dtime_store(a->dst_dt + AuCHILD, a->dst_dentry, &a->h_path); ++ } ++} ++ ++static void au_ren_rev_dt(int err, struct au_ren_args *a) ++{ ++ struct dentry *h_d; ++ struct mutex *h_mtx; ++ ++ au_dtime_revert(a->src_dt + AuPARENT); ++ if (!au_ftest_ren(a->flags, ISSAMEDIR)) ++ au_dtime_revert(a->dst_dt + AuPARENT); ++ ++ if (au_ftest_ren(a->flags, ISDIR) && err != -EIO) { ++ h_d = a->src_dt[AuCHILD].dt_h_path.dentry; ++ h_mtx = &d_inode(h_d)->i_mutex; ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD); ++ au_dtime_revert(a->src_dt + AuCHILD); ++ mutex_unlock(h_mtx); ++ ++ if (au_ftest_ren(a->flags, DT_DSTDIR)) { ++ h_d = a->dst_dt[AuCHILD].dt_h_path.dentry; ++ h_mtx = &d_inode(h_d)->i_mutex; ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD); ++ au_dtime_revert(a->dst_dt + AuCHILD); ++ mutex_unlock(h_mtx); ++ } ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int aufs_rename(struct inode *_src_dir, struct dentry *_src_dentry, ++ struct inode *_dst_dir, struct dentry *_dst_dentry) ++{ ++ int err, flags; ++ /* reduce stack space */ ++ struct au_ren_args *a; ++ ++ AuDbg("%pd, %pd\n", _src_dentry, _dst_dentry); ++ IMustLock(_src_dir); ++ IMustLock(_dst_dir); ++ ++ err = -ENOMEM; ++ BUILD_BUG_ON(sizeof(*a) > PAGE_SIZE); ++ a = kzalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ a->src_dir = _src_dir; ++ a->src_dentry = _src_dentry; ++ a->src_inode = NULL; ++ if (d_really_is_positive(a->src_dentry)) ++ a->src_inode = d_inode(a->src_dentry); ++ a->src_parent = a->src_dentry->d_parent; /* dir inode is locked */ ++ a->dst_dir = _dst_dir; ++ a->dst_dentry = _dst_dentry; ++ a->dst_inode = NULL; ++ if (d_really_is_positive(a->dst_dentry)) ++ a->dst_inode = d_inode(a->dst_dentry); ++ a->dst_parent = a->dst_dentry->d_parent; /* dir inode is locked */ ++ if (a->dst_inode) { ++ IMustLock(a->dst_inode); ++ au_igrab(a->dst_inode); ++ } ++ ++ err = -ENOTDIR; ++ flags = AuLock_FLUSH | AuLock_NOPLM | AuLock_GEN; ++ if (d_is_dir(a->src_dentry)) { ++ au_fset_ren(a->flags, ISDIR); ++ if (unlikely(d_really_is_positive(a->dst_dentry) ++ && !d_is_dir(a->dst_dentry))) ++ goto out_free; ++ err = aufs_read_and_write_lock2(a->dst_dentry, a->src_dentry, ++ AuLock_DIR | flags); ++ } else ++ err = aufs_read_and_write_lock2(a->dst_dentry, a->src_dentry, ++ flags); ++ if (unlikely(err)) ++ goto out_free; ++ ++ err = au_d_hashed_positive(a->src_dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ err = -ENOENT; ++ if (a->dst_inode) { ++ /* ++ * If it is a dir, VFS unhash dst_dentry before this ++ * function. It means we cannot rely upon d_unhashed(). ++ */ ++ if (unlikely(!a->dst_inode->i_nlink)) ++ goto out_unlock; ++ if (!S_ISDIR(a->dst_inode->i_mode)) { ++ err = au_d_hashed_positive(a->dst_dentry); ++ if (unlikely(err)) ++ goto out_unlock; ++ } else if (unlikely(IS_DEADDIR(a->dst_inode))) ++ goto out_unlock; ++ } else if (unlikely(d_unhashed(a->dst_dentry))) ++ goto out_unlock; ++ ++ /* ++ * is it possible? ++ * yes, it happened (in linux-3.3-rcN) but I don't know why. ++ * there may exist a problem somewhere else. ++ */ ++ err = -EINVAL; ++ if (unlikely(d_inode(a->dst_parent) == d_inode(a->src_dentry))) ++ goto out_unlock; ++ ++ au_fset_ren(a->flags, ISSAMEDIR); /* temporary */ ++ di_write_lock_parent(a->dst_parent); ++ ++ /* which branch we process */ ++ err = au_ren_wbr(a); ++ if (unlikely(err < 0)) ++ goto out_parent; ++ a->br = au_sbr(a->dst_dentry->d_sb, a->btgt); ++ a->h_path.mnt = au_br_mnt(a->br); ++ ++ /* are they available to be renamed */ ++ err = au_ren_may_dir(a); ++ if (unlikely(err)) ++ goto out_children; ++ ++ /* prepare the writable parent dir on the same branch */ ++ if (a->dst_bstart == a->btgt) { ++ au_fset_ren(a->flags, WHDST); ++ } else { ++ err = au_cpup_dirs(a->dst_dentry, a->btgt); ++ if (unlikely(err)) ++ goto out_children; ++ } ++ ++ if (a->src_dir != a->dst_dir) { ++ /* ++ * this temporary unlock is safe, ++ * because both dir->i_mutex are locked. ++ */ ++ di_write_unlock(a->dst_parent); ++ di_write_lock_parent(a->src_parent); ++ err = au_wr_dir_need_wh(a->src_dentry, ++ au_ftest_ren(a->flags, ISDIR), ++ &a->btgt); ++ di_write_unlock(a->src_parent); ++ di_write_lock2_parent(a->src_parent, a->dst_parent, /*isdir*/1); ++ au_fclr_ren(a->flags, ISSAMEDIR); ++ } else ++ err = au_wr_dir_need_wh(a->src_dentry, ++ au_ftest_ren(a->flags, ISDIR), ++ &a->btgt); ++ if (unlikely(err < 0)) ++ goto out_children; ++ if (err) ++ au_fset_ren(a->flags, WHSRC); ++ ++ /* cpup src */ ++ if (a->src_bstart != a->btgt) { ++ struct au_pin pin; ++ ++ err = au_pin(&pin, a->src_dentry, a->btgt, ++ au_opt_udba(a->src_dentry->d_sb), ++ AuPin_DI_LOCKED | AuPin_MNT_WRITE); ++ if (!err) { ++ struct au_cp_generic cpg = { ++ .dentry = a->src_dentry, ++ .bdst = a->btgt, ++ .bsrc = a->src_bstart, ++ .len = -1, ++ .pin = &pin, ++ .flags = AuCpup_DTIME | AuCpup_HOPEN ++ }; ++ AuDebugOn(au_dbstart(a->src_dentry) != a->src_bstart); ++ err = au_sio_cpup_simple(&cpg); ++ au_unpin(&pin); ++ } ++ if (unlikely(err)) ++ goto out_children; ++ a->src_bstart = a->btgt; ++ a->src_h_dentry = au_h_dptr(a->src_dentry, a->btgt); ++ au_fset_ren(a->flags, WHSRC); ++ } ++ ++ /* lock them all */ ++ err = au_ren_lock(a); ++ if (unlikely(err)) ++ /* leave the copied-up one */ ++ goto out_children; ++ ++ if (!au_opt_test(au_mntflags(a->dst_dir->i_sb), UDBA_NONE)) ++ err = au_may_ren(a); ++ else if (unlikely(a->dst_dentry->d_name.len > AUFS_MAX_NAMELEN)) ++ err = -ENAMETOOLONG; ++ if (unlikely(err)) ++ goto out_hdir; ++ ++ /* store timestamps to be revertible */ ++ au_ren_dt(a); ++ ++ /* here we go */ ++ err = do_rename(a); ++ if (unlikely(err)) ++ goto out_dt; ++ ++ /* update dir attributes */ ++ au_ren_refresh_dir(a); ++ ++ /* dput/iput all lower dentries */ ++ au_ren_refresh(a); ++ ++ goto out_hdir; /* success */ ++ ++out_dt: ++ au_ren_rev_dt(err, a); ++out_hdir: ++ au_ren_unlock(a); ++out_children: ++ au_nhash_wh_free(&a->whlist); ++ if (err && a->dst_inode && a->dst_bstart != a->btgt) { ++ AuDbg("bstart %d, btgt %d\n", a->dst_bstart, a->btgt); ++ au_set_h_dptr(a->dst_dentry, a->btgt, NULL); ++ au_set_dbstart(a->dst_dentry, a->dst_bstart); ++ } ++out_parent: ++ if (!err) ++ d_move(a->src_dentry, a->dst_dentry); ++ else { ++ au_update_dbstart(a->dst_dentry); ++ if (!a->dst_inode) ++ d_drop(a->dst_dentry); ++ } ++ if (au_ftest_ren(a->flags, ISSAMEDIR)) ++ di_write_unlock(a->dst_parent); ++ else ++ di_write_unlock2(a->src_parent, a->dst_parent); ++out_unlock: ++ aufs_read_and_write_unlock2(a->dst_dentry, a->src_dentry); ++out_free: ++ iput(a->dst_inode); ++ if (a->thargs) ++ au_whtmp_rmdir_free(a->thargs); ++ kfree(a); ++out: ++ AuTraceErr(err); ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/Kconfig linux-4.1.10/fs/aufs/Kconfig +--- linux-4.1.10.orig/fs/aufs/Kconfig 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/Kconfig 2015-10-22 21:36:44.000000000 +0200 +@@ -0,0 +1,185 @@ ++config AUFS_FS ++ tristate "Aufs (Advanced multi layered unification filesystem) support" ++ help ++ Aufs is a stackable unification filesystem such as Unionfs, ++ which unifies several directories and provides a merged single ++ directory. ++ In the early days, aufs was entirely re-designed and ++ re-implemented Unionfs Version 1.x series. Introducing many ++ original ideas, approaches and improvements, it becomes totally ++ different from Unionfs while keeping the basic features. ++ ++if AUFS_FS ++choice ++ prompt "Maximum number of branches" ++ default AUFS_BRANCH_MAX_127 ++ help ++ Specifies the maximum number of branches (or member directories) ++ in a single aufs. The larger value consumes more system ++ resources and has a minor impact to performance. ++config AUFS_BRANCH_MAX_127 ++ bool "127" ++ help ++ Specifies the maximum number of branches (or member directories) ++ in a single aufs. The larger value consumes more system ++ resources and has a minor impact to performance. ++config AUFS_BRANCH_MAX_511 ++ bool "511" ++ help ++ Specifies the maximum number of branches (or member directories) ++ in a single aufs. The larger value consumes more system ++ resources and has a minor impact to performance. ++config AUFS_BRANCH_MAX_1023 ++ bool "1023" ++ help ++ Specifies the maximum number of branches (or member directories) ++ in a single aufs. The larger value consumes more system ++ resources and has a minor impact to performance. ++config AUFS_BRANCH_MAX_32767 ++ bool "32767" ++ help ++ Specifies the maximum number of branches (or member directories) ++ in a single aufs. The larger value consumes more system ++ resources and has a minor impact to performance. ++endchoice ++ ++config AUFS_SBILIST ++ bool ++ depends on AUFS_MAGIC_SYSRQ || PROC_FS ++ default y ++ help ++ Automatic configuration for internal use. ++ When aufs supports Magic SysRq or /proc, enabled automatically. ++ ++config AUFS_HNOTIFY ++ bool "Detect direct branch access (bypassing aufs)" ++ help ++ If you want to modify files on branches directly, eg. bypassing aufs, ++ and want aufs to detect the changes of them fully, then enable this ++ option and use 'udba=notify' mount option. ++ Currently there is only one available configuration, "fsnotify". ++ It will have a negative impact to the performance. ++ See detail in aufs.5. ++ ++choice ++ prompt "method" if AUFS_HNOTIFY ++ default AUFS_HFSNOTIFY ++config AUFS_HFSNOTIFY ++ bool "fsnotify" ++ select FSNOTIFY ++endchoice ++ ++config AUFS_EXPORT ++ bool "NFS-exportable aufs" ++ select EXPORTFS ++ help ++ If you want to export your mounted aufs via NFS, then enable this ++ option. There are several requirements for this configuration. ++ See detail in aufs.5. ++ ++config AUFS_INO_T_64 ++ bool ++ depends on AUFS_EXPORT ++ depends on 64BIT && !(ALPHA || S390) ++ default y ++ help ++ Automatic configuration for internal use. ++ /* typedef unsigned long/int __kernel_ino_t */ ++ /* alpha and s390x are int */ ++ ++config AUFS_XATTR ++ bool "support for XATTR/EA (including Security Labels)" ++ help ++ If your branch fs supports XATTR/EA and you want to make them ++ available in aufs too, then enable this opsion and specify the ++ branch attributes for EA. ++ See detail in aufs.5. ++ ++config AUFS_FHSM ++ bool "File-based Hierarchical Storage Management" ++ help ++ Hierarchical Storage Management (or HSM) is a well-known feature ++ in the storage world. Aufs provides this feature as file-based. ++ with multiple branches. ++ These multiple branches are prioritized, ie. the topmost one ++ should be the fastest drive and be used heavily. ++ ++config AUFS_RDU ++ bool "Readdir in userspace" ++ help ++ Aufs has two methods to provide a merged view for a directory, ++ by a user-space library and by kernel-space natively. The latter ++ is always enabled but sometimes large and slow. ++ If you enable this option, install the library in aufs2-util ++ package, and set some environment variables for your readdir(3), ++ then the work will be handled in user-space which generally ++ shows better performance in most cases. ++ See detail in aufs.5. ++ ++config AUFS_SHWH ++ bool "Show whiteouts" ++ help ++ If you want to make the whiteouts in aufs visible, then enable ++ this option and specify 'shwh' mount option. Although it may ++ sounds like philosophy or something, but in technically it ++ simply shows the name of whiteout with keeping its behaviour. ++ ++config AUFS_BR_RAMFS ++ bool "Ramfs (initramfs/rootfs) as an aufs branch" ++ help ++ If you want to use ramfs as an aufs branch fs, then enable this ++ option. Generally tmpfs is recommended. ++ Aufs prohibited them to be a branch fs by default, because ++ initramfs becomes unusable after switch_root or something ++ generally. If you sets initramfs as an aufs branch and boot your ++ system by switch_root, you will meet a problem easily since the ++ files in initramfs may be inaccessible. ++ Unless you are going to use ramfs as an aufs branch fs without ++ switch_root or something, leave it N. ++ ++config AUFS_BR_FUSE ++ bool "Fuse fs as an aufs branch" ++ depends on FUSE_FS ++ select AUFS_POLL ++ help ++ If you want to use fuse-based userspace filesystem as an aufs ++ branch fs, then enable this option. ++ It implements the internal poll(2) operation which is ++ implemented by fuse only (curretnly). ++ ++config AUFS_POLL ++ bool ++ help ++ Automatic configuration for internal use. ++ ++config AUFS_BR_HFSPLUS ++ bool "Hfsplus as an aufs branch" ++ depends on HFSPLUS_FS ++ default y ++ help ++ If you want to use hfsplus fs as an aufs branch fs, then enable ++ this option. This option introduces a small overhead at ++ copying-up a file on hfsplus. ++ ++config AUFS_BDEV_LOOP ++ bool ++ depends on BLK_DEV_LOOP ++ default y ++ help ++ Automatic configuration for internal use. ++ Convert =[ym] into =y. ++ ++config AUFS_DEBUG ++ bool "Debug aufs" ++ help ++ Enable this to compile aufs internal debug code. ++ It will have a negative impact to the performance. ++ ++config AUFS_MAGIC_SYSRQ ++ bool ++ depends on AUFS_DEBUG && MAGIC_SYSRQ ++ default y ++ help ++ Automatic configuration for internal use. ++ When aufs supports Magic SysRq, enabled automatically. ++endif +diff -Nur linux-4.1.10.orig/fs/aufs/loop.c linux-4.1.10/fs/aufs/loop.c +--- linux-4.1.10.orig/fs/aufs/loop.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/loop.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,145 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * support for loopback block device as a branch ++ */ ++ ++#include "aufs.h" ++ ++/* added into drivers/block/loop.c */ ++static struct file *(*backing_file_func)(struct super_block *sb); ++ ++/* ++ * test if two lower dentries have overlapping branches. ++ */ ++int au_test_loopback_overlap(struct super_block *sb, struct dentry *h_adding) ++{ ++ struct super_block *h_sb; ++ struct file *backing_file; ++ ++ if (unlikely(!backing_file_func)) { ++ /* don't load "loop" module here */ ++ backing_file_func = symbol_get(loop_backing_file); ++ if (unlikely(!backing_file_func)) ++ /* "loop" module is not loaded */ ++ return 0; ++ } ++ ++ h_sb = h_adding->d_sb; ++ backing_file = backing_file_func(h_sb); ++ if (!backing_file) ++ return 0; ++ ++ h_adding = backing_file->f_path.dentry; ++ /* ++ * h_adding can be local NFS. ++ * in this case aufs cannot detect the loop. ++ */ ++ if (unlikely(h_adding->d_sb == sb)) ++ return 1; ++ return !!au_test_subdir(h_adding, sb->s_root); ++} ++ ++/* true if a kernel thread named 'loop[0-9].*' accesses a file */ ++int au_test_loopback_kthread(void) ++{ ++ int ret; ++ struct task_struct *tsk = current; ++ char c, comm[sizeof(tsk->comm)]; ++ ++ ret = 0; ++ if (tsk->flags & PF_KTHREAD) { ++ get_task_comm(comm, tsk); ++ c = comm[4]; ++ ret = ('0' <= c && c <= '9' ++ && !strncmp(comm, "loop", 4)); ++ } ++ ++ return ret; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define au_warn_loopback_step 16 ++static int au_warn_loopback_nelem = au_warn_loopback_step; ++static unsigned long *au_warn_loopback_array; ++ ++void au_warn_loopback(struct super_block *h_sb) ++{ ++ int i, new_nelem; ++ unsigned long *a, magic; ++ static DEFINE_SPINLOCK(spin); ++ ++ magic = h_sb->s_magic; ++ spin_lock(&spin); ++ a = au_warn_loopback_array; ++ for (i = 0; i < au_warn_loopback_nelem && *a; i++) ++ if (a[i] == magic) { ++ spin_unlock(&spin); ++ return; ++ } ++ ++ /* h_sb is new to us, print it */ ++ if (i < au_warn_loopback_nelem) { ++ a[i] = magic; ++ goto pr; ++ } ++ ++ /* expand the array */ ++ new_nelem = au_warn_loopback_nelem + au_warn_loopback_step; ++ a = au_kzrealloc(au_warn_loopback_array, ++ au_warn_loopback_nelem * sizeof(unsigned long), ++ new_nelem * sizeof(unsigned long), GFP_ATOMIC); ++ if (a) { ++ au_warn_loopback_nelem = new_nelem; ++ au_warn_loopback_array = a; ++ a[i] = magic; ++ goto pr; ++ } ++ ++ spin_unlock(&spin); ++ AuWarn1("realloc failed, ignored\n"); ++ return; ++ ++pr: ++ spin_unlock(&spin); ++ pr_warn("you may want to try another patch for loopback file " ++ "on %s(0x%lx) branch\n", au_sbtype(h_sb), magic); ++} ++ ++int au_loopback_init(void) ++{ ++ int err; ++ struct super_block *sb __maybe_unused; ++ ++ AuDebugOn(sizeof(sb->s_magic) != sizeof(unsigned long)); ++ ++ err = 0; ++ au_warn_loopback_array = kcalloc(au_warn_loopback_step, ++ sizeof(unsigned long), GFP_NOFS); ++ if (unlikely(!au_warn_loopback_array)) ++ err = -ENOMEM; ++ ++ return err; ++} ++ ++void au_loopback_fin(void) ++{ ++ symbol_put(loop_backing_file); ++ kfree(au_warn_loopback_array); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/loop.h linux-4.1.10/fs/aufs/loop.h +--- linux-4.1.10.orig/fs/aufs/loop.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/loop.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,52 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * support for loopback mount as a branch ++ */ ++ ++#ifndef __AUFS_LOOP_H__ ++#define __AUFS_LOOP_H__ ++ ++#ifdef __KERNEL__ ++ ++struct dentry; ++struct super_block; ++ ++#ifdef CONFIG_AUFS_BDEV_LOOP ++/* drivers/block/loop.c */ ++struct file *loop_backing_file(struct super_block *sb); ++ ++/* loop.c */ ++int au_test_loopback_overlap(struct super_block *sb, struct dentry *h_adding); ++int au_test_loopback_kthread(void); ++void au_warn_loopback(struct super_block *h_sb); ++ ++int au_loopback_init(void); ++void au_loopback_fin(void); ++#else ++AuStubInt0(au_test_loopback_overlap, struct super_block *sb, ++ struct dentry *h_adding) ++AuStubInt0(au_test_loopback_kthread, void) ++AuStubVoid(au_warn_loopback, struct super_block *h_sb) ++ ++AuStubInt0(au_loopback_init, void) ++AuStubVoid(au_loopback_fin, void) ++#endif /* BLK_DEV_LOOP */ ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_LOOP_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/magic.mk linux-4.1.10/fs/aufs/magic.mk +--- linux-4.1.10.orig/fs/aufs/magic.mk 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/magic.mk 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,30 @@ ++ ++# defined in ${srctree}/fs/fuse/inode.c ++# tristate ++ifdef CONFIG_FUSE_FS ++ccflags-y += -DFUSE_SUPER_MAGIC=0x65735546 ++endif ++ ++# defined in ${srctree}/fs/xfs/xfs_sb.h ++# tristate ++ifdef CONFIG_XFS_FS ++ccflags-y += -DXFS_SB_MAGIC=0x58465342 ++endif ++ ++# defined in ${srctree}/fs/configfs/mount.c ++# tristate ++ifdef CONFIG_CONFIGFS_FS ++ccflags-y += -DCONFIGFS_MAGIC=0x62656570 ++endif ++ ++# defined in ${srctree}/fs/ubifs/ubifs.h ++# tristate ++ifdef CONFIG_UBIFS_FS ++ccflags-y += -DUBIFS_SUPER_MAGIC=0x24051905 ++endif ++ ++# defined in ${srctree}/fs/hfsplus/hfsplus_raw.h ++# tristate ++ifdef CONFIG_HFSPLUS_FS ++ccflags-y += -DHFSPLUS_SUPER_MAGIC=0x482b ++endif +diff -Nur linux-4.1.10.orig/fs/aufs/Makefile linux-4.1.10/fs/aufs/Makefile +--- linux-4.1.10.orig/fs/aufs/Makefile 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/Makefile 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,44 @@ ++ ++include ${src}/magic.mk ++ifeq (${CONFIG_AUFS_FS},m) ++include ${src}/conf.mk ++endif ++-include ${src}/priv_def.mk ++ ++# cf. include/linux/kernel.h ++# enable pr_debug ++ccflags-y += -DDEBUG ++# sparse requires the full pathname ++ifdef M ++ccflags-y += -include ${M}/../../include/uapi/linux/aufs_type.h ++else ++ccflags-y += -include ${srctree}/include/uapi/linux/aufs_type.h ++endif ++ ++obj-$(CONFIG_AUFS_FS) += aufs.o ++aufs-y := module.o sbinfo.o super.o branch.o xino.o sysaufs.o opts.o \ ++ wkq.o vfsub.o dcsub.o \ ++ cpup.o whout.o wbr_policy.o \ ++ dinfo.o dentry.o \ ++ dynop.o \ ++ finfo.o file.o f_op.o \ ++ dir.o vdir.o \ ++ iinfo.o inode.o i_op.o i_op_add.o i_op_del.o i_op_ren.o \ ++ mvdown.o ioctl.o ++ ++# all are boolean ++aufs-$(CONFIG_PROC_FS) += procfs.o plink.o ++aufs-$(CONFIG_SYSFS) += sysfs.o ++aufs-$(CONFIG_DEBUG_FS) += dbgaufs.o ++aufs-$(CONFIG_AUFS_BDEV_LOOP) += loop.o ++aufs-$(CONFIG_AUFS_HNOTIFY) += hnotify.o ++aufs-$(CONFIG_AUFS_HFSNOTIFY) += hfsnotify.o ++aufs-$(CONFIG_AUFS_EXPORT) += export.o ++aufs-$(CONFIG_AUFS_XATTR) += xattr.o ++aufs-$(CONFIG_FS_POSIX_ACL) += posix_acl.o ++aufs-$(CONFIG_AUFS_FHSM) += fhsm.o ++aufs-$(CONFIG_AUFS_POLL) += poll.o ++aufs-$(CONFIG_AUFS_RDU) += rdu.o ++aufs-$(CONFIG_AUFS_BR_HFSPLUS) += hfsplus.o ++aufs-$(CONFIG_AUFS_DEBUG) += debug.o ++aufs-$(CONFIG_AUFS_MAGIC_SYSRQ) += sysrq.o +diff -Nur linux-4.1.10.orig/fs/aufs/module.c linux-4.1.10/fs/aufs/module.c +--- linux-4.1.10.orig/fs/aufs/module.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/module.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,218 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * module global variables and operations ++ */ ++ ++#include <linux/module.h> ++#include <linux/seq_file.h> ++#include "aufs.h" ++ ++void *au_kzrealloc(void *p, unsigned int nused, unsigned int new_sz, gfp_t gfp) ++{ ++ if (new_sz <= nused) ++ return p; ++ ++ p = krealloc(p, new_sz, gfp); ++ if (p) ++ memset(p + nused, 0, new_sz - nused); ++ return p; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * aufs caches ++ */ ++struct kmem_cache *au_cachep[AuCache_Last]; ++static int __init au_cache_init(void) ++{ ++ au_cachep[AuCache_DINFO] = AuCacheCtor(au_dinfo, au_di_init_once); ++ if (au_cachep[AuCache_DINFO]) ++ /* SLAB_DESTROY_BY_RCU */ ++ au_cachep[AuCache_ICNTNR] = AuCacheCtor(au_icntnr, ++ au_icntnr_init_once); ++ if (au_cachep[AuCache_ICNTNR]) ++ au_cachep[AuCache_FINFO] = AuCacheCtor(au_finfo, ++ au_fi_init_once); ++ if (au_cachep[AuCache_FINFO]) ++ au_cachep[AuCache_VDIR] = AuCache(au_vdir); ++ if (au_cachep[AuCache_VDIR]) ++ au_cachep[AuCache_DEHSTR] = AuCache(au_vdir_dehstr); ++ if (au_cachep[AuCache_DEHSTR]) ++ return 0; ++ ++ return -ENOMEM; ++} ++ ++static void au_cache_fin(void) ++{ ++ int i; ++ ++ /* ++ * Make sure all delayed rcu free inodes are flushed before we ++ * destroy cache. ++ */ ++ rcu_barrier(); ++ ++ /* excluding AuCache_HNOTIFY */ ++ BUILD_BUG_ON(AuCache_HNOTIFY + 1 != AuCache_Last); ++ for (i = 0; i < AuCache_HNOTIFY; i++) ++ if (au_cachep[i]) { ++ kmem_cache_destroy(au_cachep[i]); ++ au_cachep[i] = NULL; ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_dir_roflags; ++ ++#ifdef CONFIG_AUFS_SBILIST ++/* ++ * iterate_supers_type() doesn't protect us from ++ * remounting (branch management) ++ */ ++struct au_splhead au_sbilist; ++#endif ++ ++struct lock_class_key au_lc_key[AuLcKey_Last]; ++ ++/* ++ * functions for module interface. ++ */ ++MODULE_LICENSE("GPL"); ++/* MODULE_LICENSE("GPL v2"); */ ++MODULE_AUTHOR("Junjiro R. Okajima <aufs-users@lists.sourceforge.net>"); ++MODULE_DESCRIPTION(AUFS_NAME ++ " -- Advanced multi layered unification filesystem"); ++MODULE_VERSION(AUFS_VERSION); ++MODULE_ALIAS_FS(AUFS_NAME); ++ ++/* this module parameter has no meaning when SYSFS is disabled */ ++int sysaufs_brs = 1; ++MODULE_PARM_DESC(brs, "use <sysfs>/fs/aufs/si_*/brN"); ++module_param_named(brs, sysaufs_brs, int, S_IRUGO); ++ ++/* this module parameter has no meaning when USER_NS is disabled */ ++static bool au_userns; ++MODULE_PARM_DESC(allow_userns, "allow unprivileged to mount under userns"); ++module_param_named(allow_userns, au_userns, bool, S_IRUGO); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static char au_esc_chars[0x20 + 3]; /* 0x01-0x20, backslash, del, and NULL */ ++ ++int au_seq_path(struct seq_file *seq, struct path *path) ++{ ++ int err; ++ ++ err = seq_path(seq, path, au_esc_chars); ++ if (err > 0) ++ err = 0; ++ else if (err < 0) ++ err = -ENOMEM; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int __init aufs_init(void) ++{ ++ int err, i; ++ char *p; ++ ++ p = au_esc_chars; ++ for (i = 1; i <= ' '; i++) ++ *p++ = i; ++ *p++ = '\\'; ++ *p++ = '\x7f'; ++ *p = 0; ++ ++ au_dir_roflags = au_file_roflags(O_DIRECTORY | O_LARGEFILE); ++ ++ au_sbilist_init(); ++ sysaufs_brs_init(); ++ au_debug_init(); ++ au_dy_init(); ++ err = sysaufs_init(); ++ if (unlikely(err)) ++ goto out; ++ err = au_procfs_init(); ++ if (unlikely(err)) ++ goto out_sysaufs; ++ err = au_wkq_init(); ++ if (unlikely(err)) ++ goto out_procfs; ++ err = au_loopback_init(); ++ if (unlikely(err)) ++ goto out_wkq; ++ err = au_hnotify_init(); ++ if (unlikely(err)) ++ goto out_loopback; ++ err = au_sysrq_init(); ++ if (unlikely(err)) ++ goto out_hin; ++ err = au_cache_init(); ++ if (unlikely(err)) ++ goto out_sysrq; ++ ++ aufs_fs_type.fs_flags |= au_userns ? FS_USERNS_MOUNT : 0; ++ err = register_filesystem(&aufs_fs_type); ++ if (unlikely(err)) ++ goto out_cache; ++ ++ /* since we define pr_fmt, call printk directly */ ++ printk(KERN_INFO AUFS_NAME " " AUFS_VERSION "\n"); ++ goto out; /* success */ ++ ++out_cache: ++ au_cache_fin(); ++out_sysrq: ++ au_sysrq_fin(); ++out_hin: ++ au_hnotify_fin(); ++out_loopback: ++ au_loopback_fin(); ++out_wkq: ++ au_wkq_fin(); ++out_procfs: ++ au_procfs_fin(); ++out_sysaufs: ++ sysaufs_fin(); ++ au_dy_fin(); ++out: ++ return err; ++} ++ ++static void __exit aufs_exit(void) ++{ ++ unregister_filesystem(&aufs_fs_type); ++ au_cache_fin(); ++ au_sysrq_fin(); ++ au_hnotify_fin(); ++ au_loopback_fin(); ++ au_wkq_fin(); ++ au_procfs_fin(); ++ sysaufs_fin(); ++ au_dy_fin(); ++} ++ ++module_init(aufs_init); ++module_exit(aufs_exit); +diff -Nur linux-4.1.10.orig/fs/aufs/module.h linux-4.1.10/fs/aufs/module.h +--- linux-4.1.10.orig/fs/aufs/module.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/module.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,104 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * module initialization and module-global ++ */ ++ ++#ifndef __AUFS_MODULE_H__ ++#define __AUFS_MODULE_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/slab.h> ++ ++struct path; ++struct seq_file; ++ ++/* module parameters */ ++extern int sysaufs_brs; ++ ++/* ---------------------------------------------------------------------- */ ++ ++extern int au_dir_roflags; ++ ++enum { ++ AuLcNonDir_FIINFO, ++ AuLcNonDir_DIINFO, ++ AuLcNonDir_IIINFO, ++ ++ AuLcDir_FIINFO, ++ AuLcDir_DIINFO, ++ AuLcDir_IIINFO, ++ ++ AuLcSymlink_DIINFO, ++ AuLcSymlink_IIINFO, ++ ++ AuLcKey_Last ++}; ++extern struct lock_class_key au_lc_key[AuLcKey_Last]; ++ ++void *au_kzrealloc(void *p, unsigned int nused, unsigned int new_sz, gfp_t gfp); ++int au_seq_path(struct seq_file *seq, struct path *path); ++ ++#ifdef CONFIG_PROC_FS ++/* procfs.c */ ++int __init au_procfs_init(void); ++void au_procfs_fin(void); ++#else ++AuStubInt0(au_procfs_init, void); ++AuStubVoid(au_procfs_fin, void); ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* kmem cache */ ++enum { ++ AuCache_DINFO, ++ AuCache_ICNTNR, ++ AuCache_FINFO, ++ AuCache_VDIR, ++ AuCache_DEHSTR, ++ AuCache_HNOTIFY, /* must be last */ ++ AuCache_Last ++}; ++ ++#define AuCacheFlags (SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD) ++#define AuCache(type) KMEM_CACHE(type, AuCacheFlags) ++#define AuCacheCtor(type, ctor) \ ++ kmem_cache_create(#type, sizeof(struct type), \ ++ __alignof__(struct type), AuCacheFlags, ctor) ++ ++extern struct kmem_cache *au_cachep[]; ++ ++#define AuCacheFuncs(name, index) \ ++static inline struct au_##name *au_cache_alloc_##name(void) \ ++{ return kmem_cache_alloc(au_cachep[AuCache_##index], GFP_NOFS); } \ ++static inline void au_cache_free_##name(struct au_##name *p) \ ++{ kmem_cache_free(au_cachep[AuCache_##index], p); } ++ ++AuCacheFuncs(dinfo, DINFO); ++AuCacheFuncs(icntnr, ICNTNR); ++AuCacheFuncs(finfo, FINFO); ++AuCacheFuncs(vdir, VDIR); ++AuCacheFuncs(vdir_dehstr, DEHSTR); ++#ifdef CONFIG_AUFS_HNOTIFY ++AuCacheFuncs(hnotify, HNOTIFY); ++#endif ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_MODULE_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/mvdown.c linux-4.1.10/fs/aufs/mvdown.c +--- linux-4.1.10.orig/fs/aufs/mvdown.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/mvdown.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,694 @@ ++/* ++ * Copyright (C) 2011-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * move-down, opposite of copy-up ++ */ ++ ++#include "aufs.h" ++ ++struct au_mvd_args { ++ struct { ++ struct super_block *h_sb; ++ struct dentry *h_parent; ++ struct au_hinode *hdir; ++ struct inode *h_dir, *h_inode; ++ struct au_pin pin; ++ } info[AUFS_MVDOWN_NARRAY]; ++ ++ struct aufs_mvdown mvdown; ++ struct dentry *dentry, *parent; ++ struct inode *inode, *dir; ++ struct super_block *sb; ++ aufs_bindex_t bopq, bwh, bfound; ++ unsigned char rename_lock; ++}; ++ ++#define mvd_errno mvdown.au_errno ++#define mvd_bsrc mvdown.stbr[AUFS_MVDOWN_UPPER].bindex ++#define mvd_src_brid mvdown.stbr[AUFS_MVDOWN_UPPER].brid ++#define mvd_bdst mvdown.stbr[AUFS_MVDOWN_LOWER].bindex ++#define mvd_dst_brid mvdown.stbr[AUFS_MVDOWN_LOWER].brid ++ ++#define mvd_h_src_sb info[AUFS_MVDOWN_UPPER].h_sb ++#define mvd_h_src_parent info[AUFS_MVDOWN_UPPER].h_parent ++#define mvd_hdir_src info[AUFS_MVDOWN_UPPER].hdir ++#define mvd_h_src_dir info[AUFS_MVDOWN_UPPER].h_dir ++#define mvd_h_src_inode info[AUFS_MVDOWN_UPPER].h_inode ++#define mvd_pin_src info[AUFS_MVDOWN_UPPER].pin ++ ++#define mvd_h_dst_sb info[AUFS_MVDOWN_LOWER].h_sb ++#define mvd_h_dst_parent info[AUFS_MVDOWN_LOWER].h_parent ++#define mvd_hdir_dst info[AUFS_MVDOWN_LOWER].hdir ++#define mvd_h_dst_dir info[AUFS_MVDOWN_LOWER].h_dir ++#define mvd_h_dst_inode info[AUFS_MVDOWN_LOWER].h_inode ++#define mvd_pin_dst info[AUFS_MVDOWN_LOWER].pin ++ ++#define AU_MVD_PR(flag, ...) do { \ ++ if (flag) \ ++ pr_err(__VA_ARGS__); \ ++ } while (0) ++ ++static int find_lower_writable(struct au_mvd_args *a) ++{ ++ struct super_block *sb; ++ aufs_bindex_t bindex, bend; ++ struct au_branch *br; ++ ++ sb = a->sb; ++ bindex = a->mvd_bsrc; ++ bend = au_sbend(sb); ++ if (a->mvdown.flags & AUFS_MVDOWN_FHSM_LOWER) ++ for (bindex++; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_fhsm(br->br_perm) ++ && (!(au_br_sb(br)->s_flags & MS_RDONLY))) ++ return bindex; ++ } ++ else if (!(a->mvdown.flags & AUFS_MVDOWN_ROLOWER)) ++ for (bindex++; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (!au_br_rdonly(br)) ++ return bindex; ++ } ++ else ++ for (bindex++; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (!(au_br_sb(br)->s_flags & MS_RDONLY)) { ++ if (au_br_rdonly(br)) ++ a->mvdown.flags ++ |= AUFS_MVDOWN_ROLOWER_R; ++ return bindex; ++ } ++ } ++ ++ return -1; ++} ++ ++/* make the parent dir on bdst */ ++static int au_do_mkdir(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ ++ err = 0; ++ a->mvd_hdir_src = au_hi(a->dir, a->mvd_bsrc); ++ a->mvd_hdir_dst = au_hi(a->dir, a->mvd_bdst); ++ a->mvd_h_src_parent = au_h_dptr(a->parent, a->mvd_bsrc); ++ a->mvd_h_dst_parent = NULL; ++ if (au_dbend(a->parent) >= a->mvd_bdst) ++ a->mvd_h_dst_parent = au_h_dptr(a->parent, a->mvd_bdst); ++ if (!a->mvd_h_dst_parent) { ++ err = au_cpdown_dirs(a->dentry, a->mvd_bdst); ++ if (unlikely(err)) { ++ AU_MVD_PR(dmsg, "cpdown_dirs failed\n"); ++ goto out; ++ } ++ a->mvd_h_dst_parent = au_h_dptr(a->parent, a->mvd_bdst); ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* lock them all */ ++static int au_do_lock(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct dentry *h_trap; ++ ++ a->mvd_h_src_sb = au_sbr_sb(a->sb, a->mvd_bsrc); ++ a->mvd_h_dst_sb = au_sbr_sb(a->sb, a->mvd_bdst); ++ err = au_pin(&a->mvd_pin_dst, a->dentry, a->mvd_bdst, ++ au_opt_udba(a->sb), ++ AuPin_MNT_WRITE | AuPin_DI_LOCKED); ++ AuTraceErr(err); ++ if (unlikely(err)) { ++ AU_MVD_PR(dmsg, "pin_dst failed\n"); ++ goto out; ++ } ++ ++ if (a->mvd_h_src_sb != a->mvd_h_dst_sb) { ++ a->rename_lock = 0; ++ au_pin_init(&a->mvd_pin_src, a->dentry, a->mvd_bsrc, ++ AuLsc_DI_PARENT, AuLsc_I_PARENT3, ++ au_opt_udba(a->sb), ++ AuPin_MNT_WRITE | AuPin_DI_LOCKED); ++ err = au_do_pin(&a->mvd_pin_src); ++ AuTraceErr(err); ++ a->mvd_h_src_dir = d_inode(a->mvd_h_src_parent); ++ if (unlikely(err)) { ++ AU_MVD_PR(dmsg, "pin_src failed\n"); ++ goto out_dst; ++ } ++ goto out; /* success */ ++ } ++ ++ a->rename_lock = 1; ++ au_pin_hdir_unlock(&a->mvd_pin_dst); ++ err = au_pin(&a->mvd_pin_src, a->dentry, a->mvd_bsrc, ++ au_opt_udba(a->sb), ++ AuPin_MNT_WRITE | AuPin_DI_LOCKED); ++ AuTraceErr(err); ++ a->mvd_h_src_dir = d_inode(a->mvd_h_src_parent); ++ if (unlikely(err)) { ++ AU_MVD_PR(dmsg, "pin_src failed\n"); ++ au_pin_hdir_lock(&a->mvd_pin_dst); ++ goto out_dst; ++ } ++ au_pin_hdir_unlock(&a->mvd_pin_src); ++ h_trap = vfsub_lock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, ++ a->mvd_h_dst_parent, a->mvd_hdir_dst); ++ if (h_trap) { ++ err = (h_trap != a->mvd_h_src_parent); ++ if (err) ++ err = (h_trap != a->mvd_h_dst_parent); ++ } ++ BUG_ON(err); /* it should never happen */ ++ if (unlikely(a->mvd_h_src_dir != au_pinned_h_dir(&a->mvd_pin_src))) { ++ err = -EBUSY; ++ AuTraceErr(err); ++ vfsub_unlock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, ++ a->mvd_h_dst_parent, a->mvd_hdir_dst); ++ au_pin_hdir_lock(&a->mvd_pin_src); ++ au_unpin(&a->mvd_pin_src); ++ au_pin_hdir_lock(&a->mvd_pin_dst); ++ goto out_dst; ++ } ++ goto out; /* success */ ++ ++out_dst: ++ au_unpin(&a->mvd_pin_dst); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static void au_do_unlock(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ if (!a->rename_lock) ++ au_unpin(&a->mvd_pin_src); ++ else { ++ vfsub_unlock_rename(a->mvd_h_src_parent, a->mvd_hdir_src, ++ a->mvd_h_dst_parent, a->mvd_hdir_dst); ++ au_pin_hdir_lock(&a->mvd_pin_src); ++ au_unpin(&a->mvd_pin_src); ++ au_pin_hdir_lock(&a->mvd_pin_dst); ++ } ++ au_unpin(&a->mvd_pin_dst); ++} ++ ++/* copy-down the file */ ++static int au_do_cpdown(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct au_cp_generic cpg = { ++ .dentry = a->dentry, ++ .bdst = a->mvd_bdst, ++ .bsrc = a->mvd_bsrc, ++ .len = -1, ++ .pin = &a->mvd_pin_dst, ++ .flags = AuCpup_DTIME | AuCpup_HOPEN ++ }; ++ ++ AuDbg("b%d, b%d\n", cpg.bsrc, cpg.bdst); ++ if (a->mvdown.flags & AUFS_MVDOWN_OWLOWER) ++ au_fset_cpup(cpg.flags, OVERWRITE); ++ if (a->mvdown.flags & AUFS_MVDOWN_ROLOWER) ++ au_fset_cpup(cpg.flags, RWDST); ++ err = au_sio_cpdown_simple(&cpg); ++ if (unlikely(err)) ++ AU_MVD_PR(dmsg, "cpdown failed\n"); ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ++ * unlink the whiteout on bdst if exist which may be created by UDBA while we ++ * were sleeping ++ */ ++static int au_do_unlink_wh(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct path h_path; ++ struct au_branch *br; ++ struct inode *delegated; ++ ++ br = au_sbr(a->sb, a->mvd_bdst); ++ h_path.dentry = au_wh_lkup(a->mvd_h_dst_parent, &a->dentry->d_name, br); ++ err = PTR_ERR(h_path.dentry); ++ if (IS_ERR(h_path.dentry)) { ++ AU_MVD_PR(dmsg, "wh_lkup failed\n"); ++ goto out; ++ } ++ ++ err = 0; ++ if (d_is_positive(h_path.dentry)) { ++ h_path.mnt = au_br_mnt(br); ++ delegated = NULL; ++ err = vfsub_unlink(d_inode(a->mvd_h_dst_parent), &h_path, ++ &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ if (unlikely(err)) ++ AU_MVD_PR(dmsg, "wh_unlink failed\n"); ++ } ++ dput(h_path.dentry); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ++ * unlink the topmost h_dentry ++ */ ++static int au_do_unlink(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct path h_path; ++ struct inode *delegated; ++ ++ h_path.mnt = au_sbr_mnt(a->sb, a->mvd_bsrc); ++ h_path.dentry = au_h_dptr(a->dentry, a->mvd_bsrc); ++ delegated = NULL; ++ err = vfsub_unlink(a->mvd_h_src_dir, &h_path, &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ if (unlikely(err)) ++ AU_MVD_PR(dmsg, "unlink failed\n"); ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++/* Since mvdown succeeded, we ignore an error of this function */ ++static void au_do_stfs(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct au_branch *br; ++ ++ a->mvdown.flags |= AUFS_MVDOWN_STFS_FAILED; ++ br = au_sbr(a->sb, a->mvd_bsrc); ++ err = au_br_stfs(br, &a->mvdown.stbr[AUFS_MVDOWN_UPPER].stfs); ++ if (!err) { ++ br = au_sbr(a->sb, a->mvd_bdst); ++ a->mvdown.stbr[AUFS_MVDOWN_LOWER].brid = br->br_id; ++ err = au_br_stfs(br, &a->mvdown.stbr[AUFS_MVDOWN_LOWER].stfs); ++ } ++ if (!err) ++ a->mvdown.flags &= ~AUFS_MVDOWN_STFS_FAILED; ++ else ++ AU_MVD_PR(dmsg, "statfs failed (%d), ignored\n", err); ++} ++ ++/* ++ * copy-down the file and unlink the bsrc file. ++ * - unlink the bdst whout if exist ++ * - copy-down the file (with whtmp name and rename) ++ * - unlink the bsrc file ++ */ ++static int au_do_mvdown(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ ++ err = au_do_mkdir(dmsg, a); ++ if (!err) ++ err = au_do_lock(dmsg, a); ++ if (unlikely(err)) ++ goto out; ++ ++ /* ++ * do not revert the activities we made on bdst since they should be ++ * harmless in aufs. ++ */ ++ ++ err = au_do_cpdown(dmsg, a); ++ if (!err) ++ err = au_do_unlink_wh(dmsg, a); ++ if (!err && !(a->mvdown.flags & AUFS_MVDOWN_KUPPER)) ++ err = au_do_unlink(dmsg, a); ++ if (unlikely(err)) ++ goto out_unlock; ++ ++ AuDbg("%pd2, 0x%x, %d --> %d\n", ++ a->dentry, a->mvdown.flags, a->mvd_bsrc, a->mvd_bdst); ++ if (find_lower_writable(a) < 0) ++ a->mvdown.flags |= AUFS_MVDOWN_BOTTOM; ++ ++ if (a->mvdown.flags & AUFS_MVDOWN_STFS) ++ au_do_stfs(dmsg, a); ++ ++ /* maintain internal array */ ++ if (!(a->mvdown.flags & AUFS_MVDOWN_KUPPER)) { ++ au_set_h_dptr(a->dentry, a->mvd_bsrc, NULL); ++ au_set_dbstart(a->dentry, a->mvd_bdst); ++ au_set_h_iptr(a->inode, a->mvd_bsrc, NULL, /*flags*/0); ++ au_set_ibstart(a->inode, a->mvd_bdst); ++ } ++ if (au_dbend(a->dentry) < a->mvd_bdst) ++ au_set_dbend(a->dentry, a->mvd_bdst); ++ if (au_ibend(a->inode) < a->mvd_bdst) ++ au_set_ibend(a->inode, a->mvd_bdst); ++ ++out_unlock: ++ au_do_unlock(dmsg, a); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* make sure the file is idle */ ++static int au_mvd_args_busy(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err, plinked; ++ ++ err = 0; ++ plinked = !!au_opt_test(au_mntflags(a->sb), PLINK); ++ if (au_dbstart(a->dentry) == a->mvd_bsrc ++ && au_dcount(a->dentry) == 1 ++ && atomic_read(&a->inode->i_count) == 1 ++ /* && a->mvd_h_src_inode->i_nlink == 1 */ ++ && (!plinked || !au_plink_test(a->inode)) ++ && a->inode->i_nlink == 1) ++ goto out; ++ ++ err = -EBUSY; ++ AU_MVD_PR(dmsg, ++ "b%d, d{b%d, c%d?}, i{c%d?, l%u}, hi{l%u}, p{%d, %d}\n", ++ a->mvd_bsrc, au_dbstart(a->dentry), au_dcount(a->dentry), ++ atomic_read(&a->inode->i_count), a->inode->i_nlink, ++ a->mvd_h_src_inode->i_nlink, ++ plinked, plinked ? au_plink_test(a->inode) : 0); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* make sure the parent dir is fine */ ++static int au_mvd_args_parent(const unsigned char dmsg, ++ struct au_mvd_args *a) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ ++ err = 0; ++ if (unlikely(au_alive_dir(a->parent))) { ++ err = -ENOENT; ++ AU_MVD_PR(dmsg, "parent dir is dead\n"); ++ goto out; ++ } ++ ++ a->bopq = au_dbdiropq(a->parent); ++ bindex = au_wbr_nonopq(a->dentry, a->mvd_bdst); ++ AuDbg("b%d\n", bindex); ++ if (unlikely((bindex >= 0 && bindex < a->mvd_bdst) ++ || (a->bopq != -1 && a->bopq < a->mvd_bdst))) { ++ err = -EINVAL; ++ a->mvd_errno = EAU_MVDOWN_OPAQUE; ++ AU_MVD_PR(dmsg, "ancestor is opaque b%d, b%d\n", ++ a->bopq, a->mvd_bdst); ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_mvd_args_intermediate(const unsigned char dmsg, ++ struct au_mvd_args *a) ++{ ++ int err; ++ struct au_dinfo *dinfo, *tmp; ++ ++ /* lookup the next lower positive entry */ ++ err = -ENOMEM; ++ tmp = au_di_alloc(a->sb, AuLsc_DI_TMP); ++ if (unlikely(!tmp)) ++ goto out; ++ ++ a->bfound = -1; ++ a->bwh = -1; ++ dinfo = au_di(a->dentry); ++ au_di_cp(tmp, dinfo); ++ au_di_swap(tmp, dinfo); ++ ++ /* returns the number of positive dentries */ ++ err = au_lkup_dentry(a->dentry, a->mvd_bsrc + 1, /*type*/0); ++ if (!err) ++ a->bwh = au_dbwh(a->dentry); ++ else if (err > 0) ++ a->bfound = au_dbstart(a->dentry); ++ ++ au_di_swap(tmp, dinfo); ++ au_rw_write_unlock(&tmp->di_rwsem); ++ au_di_free(tmp); ++ if (unlikely(err < 0)) ++ AU_MVD_PR(dmsg, "failed look-up lower\n"); ++ ++ /* ++ * here, we have these cases. ++ * bfound == -1 ++ * no positive dentry under bsrc. there are more sub-cases. ++ * bwh < 0 ++ * there no whiteout, we can safely move-down. ++ * bwh <= bsrc ++ * impossible ++ * bsrc < bwh && bwh < bdst ++ * there is a whiteout on RO branch. cannot proceed. ++ * bwh == bdst ++ * there is a whiteout on the RW target branch. it should ++ * be removed. ++ * bdst < bwh ++ * there is a whiteout somewhere unrelated branch. ++ * -1 < bfound && bfound <= bsrc ++ * impossible. ++ * bfound < bdst ++ * found, but it is on RO branch between bsrc and bdst. cannot ++ * proceed. ++ * bfound == bdst ++ * found, replace it if AUFS_MVDOWN_FORCE is set. otherwise return ++ * error. ++ * bdst < bfound ++ * found, after we create the file on bdst, it will be hidden. ++ */ ++ ++ AuDebugOn(a->bfound == -1 ++ && a->bwh != -1 ++ && a->bwh <= a->mvd_bsrc); ++ AuDebugOn(-1 < a->bfound ++ && a->bfound <= a->mvd_bsrc); ++ ++ err = -EINVAL; ++ if (a->bfound == -1 ++ && a->mvd_bsrc < a->bwh ++ && a->bwh != -1 ++ && a->bwh < a->mvd_bdst) { ++ a->mvd_errno = EAU_MVDOWN_WHITEOUT; ++ AU_MVD_PR(dmsg, "bsrc %d, bdst %d, bfound %d, bwh %d\n", ++ a->mvd_bsrc, a->mvd_bdst, a->bfound, a->bwh); ++ goto out; ++ } else if (a->bfound != -1 && a->bfound < a->mvd_bdst) { ++ a->mvd_errno = EAU_MVDOWN_UPPER; ++ AU_MVD_PR(dmsg, "bdst %d, bfound %d\n", ++ a->mvd_bdst, a->bfound); ++ goto out; ++ } ++ ++ err = 0; /* success */ ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_mvd_args_exist(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ ++ err = 0; ++ if (!(a->mvdown.flags & AUFS_MVDOWN_OWLOWER) ++ && a->bfound == a->mvd_bdst) ++ err = -EEXIST; ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_mvd_args(const unsigned char dmsg, struct au_mvd_args *a) ++{ ++ int err; ++ struct au_branch *br; ++ ++ err = -EISDIR; ++ if (unlikely(S_ISDIR(a->inode->i_mode))) ++ goto out; ++ ++ err = -EINVAL; ++ if (!(a->mvdown.flags & AUFS_MVDOWN_BRID_UPPER)) ++ a->mvd_bsrc = au_ibstart(a->inode); ++ else { ++ a->mvd_bsrc = au_br_index(a->sb, a->mvd_src_brid); ++ if (unlikely(a->mvd_bsrc < 0 ++ || (a->mvd_bsrc < au_dbstart(a->dentry) ++ || au_dbend(a->dentry) < a->mvd_bsrc ++ || !au_h_dptr(a->dentry, a->mvd_bsrc)) ++ || (a->mvd_bsrc < au_ibstart(a->inode) ++ || au_ibend(a->inode) < a->mvd_bsrc ++ || !au_h_iptr(a->inode, a->mvd_bsrc)))) { ++ a->mvd_errno = EAU_MVDOWN_NOUPPER; ++ AU_MVD_PR(dmsg, "no upper\n"); ++ goto out; ++ } ++ } ++ if (unlikely(a->mvd_bsrc == au_sbend(a->sb))) { ++ a->mvd_errno = EAU_MVDOWN_BOTTOM; ++ AU_MVD_PR(dmsg, "on the bottom\n"); ++ goto out; ++ } ++ a->mvd_h_src_inode = au_h_iptr(a->inode, a->mvd_bsrc); ++ br = au_sbr(a->sb, a->mvd_bsrc); ++ err = au_br_rdonly(br); ++ if (!(a->mvdown.flags & AUFS_MVDOWN_ROUPPER)) { ++ if (unlikely(err)) ++ goto out; ++ } else if (!(vfsub_native_ro(a->mvd_h_src_inode) ++ || IS_APPEND(a->mvd_h_src_inode))) { ++ if (err) ++ a->mvdown.flags |= AUFS_MVDOWN_ROUPPER_R; ++ /* go on */ ++ } else ++ goto out; ++ ++ err = -EINVAL; ++ if (!(a->mvdown.flags & AUFS_MVDOWN_BRID_LOWER)) { ++ a->mvd_bdst = find_lower_writable(a); ++ if (unlikely(a->mvd_bdst < 0)) { ++ a->mvd_errno = EAU_MVDOWN_BOTTOM; ++ AU_MVD_PR(dmsg, "no writable lower branch\n"); ++ goto out; ++ } ++ } else { ++ a->mvd_bdst = au_br_index(a->sb, a->mvd_dst_brid); ++ if (unlikely(a->mvd_bdst < 0 ++ || au_sbend(a->sb) < a->mvd_bdst)) { ++ a->mvd_errno = EAU_MVDOWN_NOLOWERBR; ++ AU_MVD_PR(dmsg, "no lower brid\n"); ++ goto out; ++ } ++ } ++ ++ err = au_mvd_args_busy(dmsg, a); ++ if (!err) ++ err = au_mvd_args_parent(dmsg, a); ++ if (!err) ++ err = au_mvd_args_intermediate(dmsg, a); ++ if (!err) ++ err = au_mvd_args_exist(dmsg, a); ++ if (!err) ++ AuDbg("b%d, b%d\n", a->mvd_bsrc, a->mvd_bdst); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++int au_mvdown(struct dentry *dentry, struct aufs_mvdown __user *uarg) ++{ ++ int err, e; ++ unsigned char dmsg; ++ struct au_mvd_args *args; ++ ++ err = -EPERM; ++ if (unlikely(!capable(CAP_SYS_ADMIN))) ++ goto out; ++ ++ err = -ENOMEM; ++ args = kmalloc(sizeof(*args), GFP_NOFS); ++ if (unlikely(!args)) ++ goto out; ++ ++ err = copy_from_user(&args->mvdown, uarg, sizeof(args->mvdown)); ++ if (!err) ++ err = !access_ok(VERIFY_WRITE, uarg, sizeof(*uarg)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ goto out_free; ++ } ++ AuDbg("flags 0x%x\n", args->mvdown.flags); ++ args->mvdown.flags &= ~(AUFS_MVDOWN_ROLOWER_R | AUFS_MVDOWN_ROUPPER_R); ++ args->mvdown.au_errno = 0; ++ args->dentry = dentry; ++ args->inode = d_inode(dentry); ++ args->sb = dentry->d_sb; ++ ++ err = -ENOENT; ++ dmsg = !!(args->mvdown.flags & AUFS_MVDOWN_DMSG); ++ args->parent = dget_parent(dentry); ++ args->dir = d_inode(args->parent); ++ mutex_lock_nested(&args->dir->i_mutex, I_MUTEX_PARENT); ++ dput(args->parent); ++ if (unlikely(args->parent != dentry->d_parent)) { ++ AU_MVD_PR(dmsg, "parent dir is moved\n"); ++ goto out_dir; ++ } ++ ++ mutex_lock_nested(&args->inode->i_mutex, I_MUTEX_CHILD); ++ err = aufs_read_lock(dentry, AuLock_DW | AuLock_FLUSH); ++ if (unlikely(err)) ++ goto out_inode; ++ ++ di_write_lock_parent(args->parent); ++ err = au_mvd_args(dmsg, args); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ err = au_do_mvdown(dmsg, args); ++ if (unlikely(err)) ++ goto out_parent; ++ ++ au_cpup_attr_timesizes(args->dir); ++ au_cpup_attr_timesizes(args->inode); ++ au_cpup_igen(args->inode, au_h_iptr(args->inode, args->mvd_bdst)); ++ /* au_digen_dec(dentry); */ ++ ++out_parent: ++ di_write_unlock(args->parent); ++ aufs_read_unlock(dentry, AuLock_DW); ++out_inode: ++ mutex_unlock(&args->inode->i_mutex); ++out_dir: ++ mutex_unlock(&args->dir->i_mutex); ++out_free: ++ e = copy_to_user(uarg, &args->mvdown, sizeof(args->mvdown)); ++ if (unlikely(e)) ++ err = -EFAULT; ++ kfree(args); ++out: ++ AuTraceErr(err); ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/opts.c linux-4.1.10/fs/aufs/opts.c +--- linux-4.1.10.orig/fs/aufs/opts.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/opts.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1835 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * mount options/flags ++ */ ++ ++#include <linux/namei.h> ++#include <linux/types.h> /* a distribution requires */ ++#include <linux/parser.h> ++#include "aufs.h" ++ ++/* ---------------------------------------------------------------------- */ ++ ++enum { ++ Opt_br, ++ Opt_add, Opt_del, Opt_mod, Opt_append, Opt_prepend, ++ Opt_idel, Opt_imod, ++ Opt_dirwh, Opt_rdcache, Opt_rdblk, Opt_rdhash, ++ Opt_rdblk_def, Opt_rdhash_def, ++ Opt_xino, Opt_noxino, ++ Opt_trunc_xino, Opt_trunc_xino_v, Opt_notrunc_xino, ++ Opt_trunc_xino_path, Opt_itrunc_xino, ++ Opt_trunc_xib, Opt_notrunc_xib, ++ Opt_shwh, Opt_noshwh, ++ Opt_plink, Opt_noplink, Opt_list_plink, ++ Opt_udba, ++ Opt_dio, Opt_nodio, ++ Opt_diropq_a, Opt_diropq_w, ++ Opt_warn_perm, Opt_nowarn_perm, ++ Opt_wbr_copyup, Opt_wbr_create, ++ Opt_fhsm_sec, ++ Opt_verbose, Opt_noverbose, ++ Opt_sum, Opt_nosum, Opt_wsum, ++ Opt_dirperm1, Opt_nodirperm1, ++ Opt_acl, Opt_noacl, ++ Opt_tail, Opt_ignore, Opt_ignore_silent, Opt_err ++}; ++ ++static match_table_t options = { ++ {Opt_br, "br=%s"}, ++ {Opt_br, "br:%s"}, ++ ++ {Opt_add, "add=%d:%s"}, ++ {Opt_add, "add:%d:%s"}, ++ {Opt_add, "ins=%d:%s"}, ++ {Opt_add, "ins:%d:%s"}, ++ {Opt_append, "append=%s"}, ++ {Opt_append, "append:%s"}, ++ {Opt_prepend, "prepend=%s"}, ++ {Opt_prepend, "prepend:%s"}, ++ ++ {Opt_del, "del=%s"}, ++ {Opt_del, "del:%s"}, ++ /* {Opt_idel, "idel:%d"}, */ ++ {Opt_mod, "mod=%s"}, ++ {Opt_mod, "mod:%s"}, ++ /* {Opt_imod, "imod:%d:%s"}, */ ++ ++ {Opt_dirwh, "dirwh=%d"}, ++ ++ {Opt_xino, "xino=%s"}, ++ {Opt_noxino, "noxino"}, ++ {Opt_trunc_xino, "trunc_xino"}, ++ {Opt_trunc_xino_v, "trunc_xino_v=%d:%d"}, ++ {Opt_notrunc_xino, "notrunc_xino"}, ++ {Opt_trunc_xino_path, "trunc_xino=%s"}, ++ {Opt_itrunc_xino, "itrunc_xino=%d"}, ++ /* {Opt_zxino, "zxino=%s"}, */ ++ {Opt_trunc_xib, "trunc_xib"}, ++ {Opt_notrunc_xib, "notrunc_xib"}, ++ ++#ifdef CONFIG_PROC_FS ++ {Opt_plink, "plink"}, ++#else ++ {Opt_ignore_silent, "plink"}, ++#endif ++ ++ {Opt_noplink, "noplink"}, ++ ++#ifdef CONFIG_AUFS_DEBUG ++ {Opt_list_plink, "list_plink"}, ++#endif ++ ++ {Opt_udba, "udba=%s"}, ++ ++ {Opt_dio, "dio"}, ++ {Opt_nodio, "nodio"}, ++ ++#ifdef CONFIG_AUFS_FHSM ++ {Opt_fhsm_sec, "fhsm_sec=%d"}, ++#else ++ {Opt_ignore_silent, "fhsm_sec=%d"}, ++#endif ++ ++ {Opt_diropq_a, "diropq=always"}, ++ {Opt_diropq_a, "diropq=a"}, ++ {Opt_diropq_w, "diropq=whiteouted"}, ++ {Opt_diropq_w, "diropq=w"}, ++ ++ {Opt_warn_perm, "warn_perm"}, ++ {Opt_nowarn_perm, "nowarn_perm"}, ++ ++ /* keep them temporary */ ++ {Opt_ignore_silent, "nodlgt"}, ++ {Opt_ignore_silent, "clean_plink"}, ++ ++#ifdef CONFIG_AUFS_SHWH ++ {Opt_shwh, "shwh"}, ++#endif ++ {Opt_noshwh, "noshwh"}, ++ ++ {Opt_dirperm1, "dirperm1"}, ++ {Opt_nodirperm1, "nodirperm1"}, ++ ++ {Opt_verbose, "verbose"}, ++ {Opt_verbose, "v"}, ++ {Opt_noverbose, "noverbose"}, ++ {Opt_noverbose, "quiet"}, ++ {Opt_noverbose, "q"}, ++ {Opt_noverbose, "silent"}, ++ ++ {Opt_sum, "sum"}, ++ {Opt_nosum, "nosum"}, ++ {Opt_wsum, "wsum"}, ++ ++ {Opt_rdcache, "rdcache=%d"}, ++ {Opt_rdblk, "rdblk=%d"}, ++ {Opt_rdblk_def, "rdblk=def"}, ++ {Opt_rdhash, "rdhash=%d"}, ++ {Opt_rdhash_def, "rdhash=def"}, ++ ++ {Opt_wbr_create, "create=%s"}, ++ {Opt_wbr_create, "create_policy=%s"}, ++ {Opt_wbr_copyup, "cpup=%s"}, ++ {Opt_wbr_copyup, "copyup=%s"}, ++ {Opt_wbr_copyup, "copyup_policy=%s"}, ++ ++ /* generic VFS flag */ ++#ifdef CONFIG_FS_POSIX_ACL ++ {Opt_acl, "acl"}, ++ {Opt_noacl, "noacl"}, ++#else ++ {Opt_ignore_silent, "acl"}, ++ {Opt_ignore_silent, "noacl"}, ++#endif ++ ++ /* internal use for the scripts */ ++ {Opt_ignore_silent, "si=%s"}, ++ ++ {Opt_br, "dirs=%s"}, ++ {Opt_ignore, "debug=%d"}, ++ {Opt_ignore, "delete=whiteout"}, ++ {Opt_ignore, "delete=all"}, ++ {Opt_ignore, "imap=%s"}, ++ ++ /* temporary workaround, due to old mount(8)? */ ++ {Opt_ignore_silent, "relatime"}, ++ ++ {Opt_err, NULL} ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static const char *au_parser_pattern(int val, match_table_t tbl) ++{ ++ struct match_token *p; ++ ++ p = tbl; ++ while (p->pattern) { ++ if (p->token == val) ++ return p->pattern; ++ p++; ++ } ++ BUG(); ++ return "??"; ++} ++ ++static const char *au_optstr(int *val, match_table_t tbl) ++{ ++ struct match_token *p; ++ int v; ++ ++ v = *val; ++ if (!v) ++ goto out; ++ p = tbl; ++ while (p->pattern) { ++ if (p->token ++ && (v & p->token) == p->token) { ++ *val &= ~p->token; ++ return p->pattern; ++ } ++ p++; ++ } ++ ++out: ++ return NULL; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static match_table_t brperm = { ++ {AuBrPerm_RO, AUFS_BRPERM_RO}, ++ {AuBrPerm_RR, AUFS_BRPERM_RR}, ++ {AuBrPerm_RW, AUFS_BRPERM_RW}, ++ {0, NULL} ++}; ++ ++static match_table_t brattr = { ++ /* general */ ++ {AuBrAttr_COO_REG, AUFS_BRATTR_COO_REG}, ++ {AuBrAttr_COO_ALL, AUFS_BRATTR_COO_ALL}, ++ /* 'unpin' attrib is meaningless since linux-3.18-rc1 */ ++ {AuBrAttr_UNPIN, AUFS_BRATTR_UNPIN}, ++#ifdef CONFIG_AUFS_FHSM ++ {AuBrAttr_FHSM, AUFS_BRATTR_FHSM}, ++#endif ++#ifdef CONFIG_AUFS_XATTR ++ {AuBrAttr_ICEX, AUFS_BRATTR_ICEX}, ++ {AuBrAttr_ICEX_SEC, AUFS_BRATTR_ICEX_SEC}, ++ {AuBrAttr_ICEX_SYS, AUFS_BRATTR_ICEX_SYS}, ++ {AuBrAttr_ICEX_TR, AUFS_BRATTR_ICEX_TR}, ++ {AuBrAttr_ICEX_USR, AUFS_BRATTR_ICEX_USR}, ++ {AuBrAttr_ICEX_OTH, AUFS_BRATTR_ICEX_OTH}, ++#endif ++ ++ /* ro/rr branch */ ++ {AuBrRAttr_WH, AUFS_BRRATTR_WH}, ++ ++ /* rw branch */ ++ {AuBrWAttr_MOO, AUFS_BRWATTR_MOO}, ++ {AuBrWAttr_NoLinkWH, AUFS_BRWATTR_NLWH}, ++ ++ {0, NULL} ++}; ++ ++static int br_attr_val(char *str, match_table_t table, substring_t args[]) ++{ ++ int attr, v; ++ char *p; ++ ++ attr = 0; ++ do { ++ p = strchr(str, '+'); ++ if (p) ++ *p = 0; ++ v = match_token(str, table, args); ++ if (v) { ++ if (v & AuBrAttr_CMOO_Mask) ++ attr &= ~AuBrAttr_CMOO_Mask; ++ attr |= v; ++ } else { ++ if (p) ++ *p = '+'; ++ pr_warn("ignored branch attribute %s\n", str); ++ break; ++ } ++ if (p) ++ str = p + 1; ++ } while (p); ++ ++ return attr; ++} ++ ++static int au_do_optstr_br_attr(au_br_perm_str_t *str, int perm) ++{ ++ int sz; ++ const char *p; ++ char *q; ++ ++ q = str->a; ++ *q = 0; ++ p = au_optstr(&perm, brattr); ++ if (p) { ++ sz = strlen(p); ++ memcpy(q, p, sz + 1); ++ q += sz; ++ } else ++ goto out; ++ ++ do { ++ p = au_optstr(&perm, brattr); ++ if (p) { ++ *q++ = '+'; ++ sz = strlen(p); ++ memcpy(q, p, sz + 1); ++ q += sz; ++ } ++ } while (p); ++ ++out: ++ return q - str->a; ++} ++ ++static int noinline_for_stack br_perm_val(char *perm) ++{ ++ int val, bad, sz; ++ char *p; ++ substring_t args[MAX_OPT_ARGS]; ++ au_br_perm_str_t attr; ++ ++ p = strchr(perm, '+'); ++ if (p) ++ *p = 0; ++ val = match_token(perm, brperm, args); ++ if (!val) { ++ if (p) ++ *p = '+'; ++ pr_warn("ignored branch permission %s\n", perm); ++ val = AuBrPerm_RO; ++ goto out; ++ } ++ if (!p) ++ goto out; ++ ++ val |= br_attr_val(p + 1, brattr, args); ++ ++ bad = 0; ++ switch (val & AuBrPerm_Mask) { ++ case AuBrPerm_RO: ++ case AuBrPerm_RR: ++ bad = val & AuBrWAttr_Mask; ++ val &= ~AuBrWAttr_Mask; ++ break; ++ case AuBrPerm_RW: ++ bad = val & AuBrRAttr_Mask; ++ val &= ~AuBrRAttr_Mask; ++ break; ++ } ++ ++ /* ++ * 'unpin' attrib becomes meaningless since linux-3.18-rc1, but aufs ++ * does not treat it as an error, just warning. ++ * this is a tiny guard for the user operation. ++ */ ++ if (val & AuBrAttr_UNPIN) { ++ bad |= AuBrAttr_UNPIN; ++ val &= ~AuBrAttr_UNPIN; ++ } ++ ++ if (unlikely(bad)) { ++ sz = au_do_optstr_br_attr(&attr, bad); ++ AuDebugOn(!sz); ++ pr_warn("ignored branch attribute %s\n", attr.a); ++ } ++ ++out: ++ return val; ++} ++ ++void au_optstr_br_perm(au_br_perm_str_t *str, int perm) ++{ ++ au_br_perm_str_t attr; ++ const char *p; ++ char *q; ++ int sz; ++ ++ q = str->a; ++ p = au_optstr(&perm, brperm); ++ AuDebugOn(!p || !*p); ++ sz = strlen(p); ++ memcpy(q, p, sz + 1); ++ q += sz; ++ ++ sz = au_do_optstr_br_attr(&attr, perm); ++ if (sz) { ++ *q++ = '+'; ++ memcpy(q, attr.a, sz + 1); ++ } ++ ++ AuDebugOn(strlen(str->a) >= sizeof(str->a)); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static match_table_t udbalevel = { ++ {AuOpt_UDBA_REVAL, "reval"}, ++ {AuOpt_UDBA_NONE, "none"}, ++#ifdef CONFIG_AUFS_HNOTIFY ++ {AuOpt_UDBA_HNOTIFY, "notify"}, /* abstraction */ ++#ifdef CONFIG_AUFS_HFSNOTIFY ++ {AuOpt_UDBA_HNOTIFY, "fsnotify"}, ++#endif ++#endif ++ {-1, NULL} ++}; ++ ++static int noinline_for_stack udba_val(char *str) ++{ ++ substring_t args[MAX_OPT_ARGS]; ++ ++ return match_token(str, udbalevel, args); ++} ++ ++const char *au_optstr_udba(int udba) ++{ ++ return au_parser_pattern(udba, udbalevel); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static match_table_t au_wbr_create_policy = { ++ {AuWbrCreate_TDP, "tdp"}, ++ {AuWbrCreate_TDP, "top-down-parent"}, ++ {AuWbrCreate_RR, "rr"}, ++ {AuWbrCreate_RR, "round-robin"}, ++ {AuWbrCreate_MFS, "mfs"}, ++ {AuWbrCreate_MFS, "most-free-space"}, ++ {AuWbrCreate_MFSV, "mfs:%d"}, ++ {AuWbrCreate_MFSV, "most-free-space:%d"}, ++ ++ {AuWbrCreate_MFSRR, "mfsrr:%d"}, ++ {AuWbrCreate_MFSRRV, "mfsrr:%d:%d"}, ++ {AuWbrCreate_PMFS, "pmfs"}, ++ {AuWbrCreate_PMFSV, "pmfs:%d"}, ++ {AuWbrCreate_PMFSRR, "pmfsrr:%d"}, ++ {AuWbrCreate_PMFSRRV, "pmfsrr:%d:%d"}, ++ ++ {-1, NULL} ++}; ++ ++/* ++ * cf. linux/lib/parser.c and cmdline.c ++ * gave up calling memparse() since it uses simple_strtoull() instead of ++ * kstrto...(). ++ */ ++static int noinline_for_stack ++au_match_ull(substring_t *s, unsigned long long *result) ++{ ++ int err; ++ unsigned int len; ++ char a[32]; ++ ++ err = -ERANGE; ++ len = s->to - s->from; ++ if (len + 1 <= sizeof(a)) { ++ memcpy(a, s->from, len); ++ a[len] = '\0'; ++ err = kstrtoull(a, 0, result); ++ } ++ return err; ++} ++ ++static int au_wbr_mfs_wmark(substring_t *arg, char *str, ++ struct au_opt_wbr_create *create) ++{ ++ int err; ++ unsigned long long ull; ++ ++ err = 0; ++ if (!au_match_ull(arg, &ull)) ++ create->mfsrr_watermark = ull; ++ else { ++ pr_err("bad integer in %s\n", str); ++ err = -EINVAL; ++ } ++ ++ return err; ++} ++ ++static int au_wbr_mfs_sec(substring_t *arg, char *str, ++ struct au_opt_wbr_create *create) ++{ ++ int n, err; ++ ++ err = 0; ++ if (!match_int(arg, &n) && 0 <= n && n <= AUFS_MFS_MAX_SEC) ++ create->mfs_second = n; ++ else { ++ pr_err("bad integer in %s\n", str); ++ err = -EINVAL; ++ } ++ ++ return err; ++} ++ ++static int noinline_for_stack ++au_wbr_create_val(char *str, struct au_opt_wbr_create *create) ++{ ++ int err, e; ++ substring_t args[MAX_OPT_ARGS]; ++ ++ err = match_token(str, au_wbr_create_policy, args); ++ create->wbr_create = err; ++ switch (err) { ++ case AuWbrCreate_MFSRRV: ++ case AuWbrCreate_PMFSRRV: ++ e = au_wbr_mfs_wmark(&args[0], str, create); ++ if (!e) ++ e = au_wbr_mfs_sec(&args[1], str, create); ++ if (unlikely(e)) ++ err = e; ++ break; ++ case AuWbrCreate_MFSRR: ++ case AuWbrCreate_PMFSRR: ++ e = au_wbr_mfs_wmark(&args[0], str, create); ++ if (unlikely(e)) { ++ err = e; ++ break; ++ } ++ /*FALLTHROUGH*/ ++ case AuWbrCreate_MFS: ++ case AuWbrCreate_PMFS: ++ create->mfs_second = AUFS_MFS_DEF_SEC; ++ break; ++ case AuWbrCreate_MFSV: ++ case AuWbrCreate_PMFSV: ++ e = au_wbr_mfs_sec(&args[0], str, create); ++ if (unlikely(e)) ++ err = e; ++ break; ++ } ++ ++ return err; ++} ++ ++const char *au_optstr_wbr_create(int wbr_create) ++{ ++ return au_parser_pattern(wbr_create, au_wbr_create_policy); ++} ++ ++static match_table_t au_wbr_copyup_policy = { ++ {AuWbrCopyup_TDP, "tdp"}, ++ {AuWbrCopyup_TDP, "top-down-parent"}, ++ {AuWbrCopyup_BUP, "bup"}, ++ {AuWbrCopyup_BUP, "bottom-up-parent"}, ++ {AuWbrCopyup_BU, "bu"}, ++ {AuWbrCopyup_BU, "bottom-up"}, ++ {-1, NULL} ++}; ++ ++static int noinline_for_stack au_wbr_copyup_val(char *str) ++{ ++ substring_t args[MAX_OPT_ARGS]; ++ ++ return match_token(str, au_wbr_copyup_policy, args); ++} ++ ++const char *au_optstr_wbr_copyup(int wbr_copyup) ++{ ++ return au_parser_pattern(wbr_copyup, au_wbr_copyup_policy); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static const int lkup_dirflags = LOOKUP_FOLLOW | LOOKUP_DIRECTORY; ++ ++static void dump_opts(struct au_opts *opts) ++{ ++#ifdef CONFIG_AUFS_DEBUG ++ /* reduce stack space */ ++ union { ++ struct au_opt_add *add; ++ struct au_opt_del *del; ++ struct au_opt_mod *mod; ++ struct au_opt_xino *xino; ++ struct au_opt_xino_itrunc *xino_itrunc; ++ struct au_opt_wbr_create *create; ++ } u; ++ struct au_opt *opt; ++ ++ opt = opts->opt; ++ while (opt->type != Opt_tail) { ++ switch (opt->type) { ++ case Opt_add: ++ u.add = &opt->add; ++ AuDbg("add {b%d, %s, 0x%x, %p}\n", ++ u.add->bindex, u.add->pathname, u.add->perm, ++ u.add->path.dentry); ++ break; ++ case Opt_del: ++ case Opt_idel: ++ u.del = &opt->del; ++ AuDbg("del {%s, %p}\n", ++ u.del->pathname, u.del->h_path.dentry); ++ break; ++ case Opt_mod: ++ case Opt_imod: ++ u.mod = &opt->mod; ++ AuDbg("mod {%s, 0x%x, %p}\n", ++ u.mod->path, u.mod->perm, u.mod->h_root); ++ break; ++ case Opt_append: ++ u.add = &opt->add; ++ AuDbg("append {b%d, %s, 0x%x, %p}\n", ++ u.add->bindex, u.add->pathname, u.add->perm, ++ u.add->path.dentry); ++ break; ++ case Opt_prepend: ++ u.add = &opt->add; ++ AuDbg("prepend {b%d, %s, 0x%x, %p}\n", ++ u.add->bindex, u.add->pathname, u.add->perm, ++ u.add->path.dentry); ++ break; ++ case Opt_dirwh: ++ AuDbg("dirwh %d\n", opt->dirwh); ++ break; ++ case Opt_rdcache: ++ AuDbg("rdcache %d\n", opt->rdcache); ++ break; ++ case Opt_rdblk: ++ AuDbg("rdblk %u\n", opt->rdblk); ++ break; ++ case Opt_rdblk_def: ++ AuDbg("rdblk_def\n"); ++ break; ++ case Opt_rdhash: ++ AuDbg("rdhash %u\n", opt->rdhash); ++ break; ++ case Opt_rdhash_def: ++ AuDbg("rdhash_def\n"); ++ break; ++ case Opt_xino: ++ u.xino = &opt->xino; ++ AuDbg("xino {%s %pD}\n", u.xino->path, u.xino->file); ++ break; ++ case Opt_trunc_xino: ++ AuLabel(trunc_xino); ++ break; ++ case Opt_notrunc_xino: ++ AuLabel(notrunc_xino); ++ break; ++ case Opt_trunc_xino_path: ++ case Opt_itrunc_xino: ++ u.xino_itrunc = &opt->xino_itrunc; ++ AuDbg("trunc_xino %d\n", u.xino_itrunc->bindex); ++ break; ++ case Opt_noxino: ++ AuLabel(noxino); ++ break; ++ case Opt_trunc_xib: ++ AuLabel(trunc_xib); ++ break; ++ case Opt_notrunc_xib: ++ AuLabel(notrunc_xib); ++ break; ++ case Opt_shwh: ++ AuLabel(shwh); ++ break; ++ case Opt_noshwh: ++ AuLabel(noshwh); ++ break; ++ case Opt_dirperm1: ++ AuLabel(dirperm1); ++ break; ++ case Opt_nodirperm1: ++ AuLabel(nodirperm1); ++ break; ++ case Opt_plink: ++ AuLabel(plink); ++ break; ++ case Opt_noplink: ++ AuLabel(noplink); ++ break; ++ case Opt_list_plink: ++ AuLabel(list_plink); ++ break; ++ case Opt_udba: ++ AuDbg("udba %d, %s\n", ++ opt->udba, au_optstr_udba(opt->udba)); ++ break; ++ case Opt_dio: ++ AuLabel(dio); ++ break; ++ case Opt_nodio: ++ AuLabel(nodio); ++ break; ++ case Opt_diropq_a: ++ AuLabel(diropq_a); ++ break; ++ case Opt_diropq_w: ++ AuLabel(diropq_w); ++ break; ++ case Opt_warn_perm: ++ AuLabel(warn_perm); ++ break; ++ case Opt_nowarn_perm: ++ AuLabel(nowarn_perm); ++ break; ++ case Opt_verbose: ++ AuLabel(verbose); ++ break; ++ case Opt_noverbose: ++ AuLabel(noverbose); ++ break; ++ case Opt_sum: ++ AuLabel(sum); ++ break; ++ case Opt_nosum: ++ AuLabel(nosum); ++ break; ++ case Opt_wsum: ++ AuLabel(wsum); ++ break; ++ case Opt_wbr_create: ++ u.create = &opt->wbr_create; ++ AuDbg("create %d, %s\n", u.create->wbr_create, ++ au_optstr_wbr_create(u.create->wbr_create)); ++ switch (u.create->wbr_create) { ++ case AuWbrCreate_MFSV: ++ case AuWbrCreate_PMFSV: ++ AuDbg("%d sec\n", u.create->mfs_second); ++ break; ++ case AuWbrCreate_MFSRR: ++ AuDbg("%llu watermark\n", ++ u.create->mfsrr_watermark); ++ break; ++ case AuWbrCreate_MFSRRV: ++ case AuWbrCreate_PMFSRRV: ++ AuDbg("%llu watermark, %d sec\n", ++ u.create->mfsrr_watermark, ++ u.create->mfs_second); ++ break; ++ } ++ break; ++ case Opt_wbr_copyup: ++ AuDbg("copyup %d, %s\n", opt->wbr_copyup, ++ au_optstr_wbr_copyup(opt->wbr_copyup)); ++ break; ++ case Opt_fhsm_sec: ++ AuDbg("fhsm_sec %u\n", opt->fhsm_second); ++ break; ++ case Opt_acl: ++ AuLabel(acl); ++ break; ++ case Opt_noacl: ++ AuLabel(noacl); ++ break; ++ default: ++ BUG(); ++ } ++ opt++; ++ } ++#endif ++} ++ ++void au_opts_free(struct au_opts *opts) ++{ ++ struct au_opt *opt; ++ ++ opt = opts->opt; ++ while (opt->type != Opt_tail) { ++ switch (opt->type) { ++ case Opt_add: ++ case Opt_append: ++ case Opt_prepend: ++ path_put(&opt->add.path); ++ break; ++ case Opt_del: ++ case Opt_idel: ++ path_put(&opt->del.h_path); ++ break; ++ case Opt_mod: ++ case Opt_imod: ++ dput(opt->mod.h_root); ++ break; ++ case Opt_xino: ++ fput(opt->xino.file); ++ break; ++ } ++ opt++; ++ } ++} ++ ++static int opt_add(struct au_opt *opt, char *opt_str, unsigned long sb_flags, ++ aufs_bindex_t bindex) ++{ ++ int err; ++ struct au_opt_add *add = &opt->add; ++ char *p; ++ ++ add->bindex = bindex; ++ add->perm = AuBrPerm_RO; ++ add->pathname = opt_str; ++ p = strchr(opt_str, '='); ++ if (p) { ++ *p++ = 0; ++ if (*p) ++ add->perm = br_perm_val(p); ++ } ++ ++ err = vfsub_kern_path(add->pathname, lkup_dirflags, &add->path); ++ if (!err) { ++ if (!p) { ++ add->perm = AuBrPerm_RO; ++ if (au_test_fs_rr(add->path.dentry->d_sb)) ++ add->perm = AuBrPerm_RR; ++ else if (!bindex && !(sb_flags & MS_RDONLY)) ++ add->perm = AuBrPerm_RW; ++ } ++ opt->type = Opt_add; ++ goto out; ++ } ++ pr_err("lookup failed %s (%d)\n", add->pathname, err); ++ err = -EINVAL; ++ ++out: ++ return err; ++} ++ ++static int au_opts_parse_del(struct au_opt_del *del, substring_t args[]) ++{ ++ int err; ++ ++ del->pathname = args[0].from; ++ AuDbg("del path %s\n", del->pathname); ++ ++ err = vfsub_kern_path(del->pathname, lkup_dirflags, &del->h_path); ++ if (unlikely(err)) ++ pr_err("lookup failed %s (%d)\n", del->pathname, err); ++ ++ return err; ++} ++ ++#if 0 /* reserved for future use */ ++static int au_opts_parse_idel(struct super_block *sb, aufs_bindex_t bindex, ++ struct au_opt_del *del, substring_t args[]) ++{ ++ int err; ++ struct dentry *root; ++ ++ err = -EINVAL; ++ root = sb->s_root; ++ aufs_read_lock(root, AuLock_FLUSH); ++ if (bindex < 0 || au_sbend(sb) < bindex) { ++ pr_err("out of bounds, %d\n", bindex); ++ goto out; ++ } ++ ++ err = 0; ++ del->h_path.dentry = dget(au_h_dptr(root, bindex)); ++ del->h_path.mnt = mntget(au_sbr_mnt(sb, bindex)); ++ ++out: ++ aufs_read_unlock(root, !AuLock_IR); ++ return err; ++} ++#endif ++ ++static int noinline_for_stack ++au_opts_parse_mod(struct au_opt_mod *mod, substring_t args[]) ++{ ++ int err; ++ struct path path; ++ char *p; ++ ++ err = -EINVAL; ++ mod->path = args[0].from; ++ p = strchr(mod->path, '='); ++ if (unlikely(!p)) { ++ pr_err("no permssion %s\n", args[0].from); ++ goto out; ++ } ++ ++ *p++ = 0; ++ err = vfsub_kern_path(mod->path, lkup_dirflags, &path); ++ if (unlikely(err)) { ++ pr_err("lookup failed %s (%d)\n", mod->path, err); ++ goto out; ++ } ++ ++ mod->perm = br_perm_val(p); ++ AuDbg("mod path %s, perm 0x%x, %s\n", mod->path, mod->perm, p); ++ mod->h_root = dget(path.dentry); ++ path_put(&path); ++ ++out: ++ return err; ++} ++ ++#if 0 /* reserved for future use */ ++static int au_opts_parse_imod(struct super_block *sb, aufs_bindex_t bindex, ++ struct au_opt_mod *mod, substring_t args[]) ++{ ++ int err; ++ struct dentry *root; ++ ++ err = -EINVAL; ++ root = sb->s_root; ++ aufs_read_lock(root, AuLock_FLUSH); ++ if (bindex < 0 || au_sbend(sb) < bindex) { ++ pr_err("out of bounds, %d\n", bindex); ++ goto out; ++ } ++ ++ err = 0; ++ mod->perm = br_perm_val(args[1].from); ++ AuDbg("mod path %s, perm 0x%x, %s\n", ++ mod->path, mod->perm, args[1].from); ++ mod->h_root = dget(au_h_dptr(root, bindex)); ++ ++out: ++ aufs_read_unlock(root, !AuLock_IR); ++ return err; ++} ++#endif ++ ++static int au_opts_parse_xino(struct super_block *sb, struct au_opt_xino *xino, ++ substring_t args[]) ++{ ++ int err; ++ struct file *file; ++ ++ file = au_xino_create(sb, args[0].from, /*silent*/0); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ ++ err = -EINVAL; ++ if (unlikely(file->f_path.dentry->d_sb == sb)) { ++ fput(file); ++ pr_err("%s must be outside\n", args[0].from); ++ goto out; ++ } ++ ++ err = 0; ++ xino->file = file; ++ xino->path = args[0].from; ++ ++out: ++ return err; ++} ++ ++static int noinline_for_stack ++au_opts_parse_xino_itrunc_path(struct super_block *sb, ++ struct au_opt_xino_itrunc *xino_itrunc, ++ substring_t args[]) ++{ ++ int err; ++ aufs_bindex_t bend, bindex; ++ struct path path; ++ struct dentry *root; ++ ++ err = vfsub_kern_path(args[0].from, lkup_dirflags, &path); ++ if (unlikely(err)) { ++ pr_err("lookup failed %s (%d)\n", args[0].from, err); ++ goto out; ++ } ++ ++ xino_itrunc->bindex = -1; ++ root = sb->s_root; ++ aufs_read_lock(root, AuLock_FLUSH); ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ if (au_h_dptr(root, bindex) == path.dentry) { ++ xino_itrunc->bindex = bindex; ++ break; ++ } ++ } ++ aufs_read_unlock(root, !AuLock_IR); ++ path_put(&path); ++ ++ if (unlikely(xino_itrunc->bindex < 0)) { ++ pr_err("no such branch %s\n", args[0].from); ++ err = -EINVAL; ++ } ++ ++out: ++ return err; ++} ++ ++/* called without aufs lock */ ++int au_opts_parse(struct super_block *sb, char *str, struct au_opts *opts) ++{ ++ int err, n, token; ++ aufs_bindex_t bindex; ++ unsigned char skipped; ++ struct dentry *root; ++ struct au_opt *opt, *opt_tail; ++ char *opt_str; ++ /* reduce the stack space */ ++ union { ++ struct au_opt_xino_itrunc *xino_itrunc; ++ struct au_opt_wbr_create *create; ++ } u; ++ struct { ++ substring_t args[MAX_OPT_ARGS]; ++ } *a; ++ ++ err = -ENOMEM; ++ a = kmalloc(sizeof(*a), GFP_NOFS); ++ if (unlikely(!a)) ++ goto out; ++ ++ root = sb->s_root; ++ err = 0; ++ bindex = 0; ++ opt = opts->opt; ++ opt_tail = opt + opts->max_opt - 1; ++ opt->type = Opt_tail; ++ while (!err && (opt_str = strsep(&str, ",")) && *opt_str) { ++ err = -EINVAL; ++ skipped = 0; ++ token = match_token(opt_str, options, a->args); ++ switch (token) { ++ case Opt_br: ++ err = 0; ++ while (!err && (opt_str = strsep(&a->args[0].from, ":")) ++ && *opt_str) { ++ err = opt_add(opt, opt_str, opts->sb_flags, ++ bindex++); ++ if (unlikely(!err && ++opt > opt_tail)) { ++ err = -E2BIG; ++ break; ++ } ++ opt->type = Opt_tail; ++ skipped = 1; ++ } ++ break; ++ case Opt_add: ++ if (unlikely(match_int(&a->args[0], &n))) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ bindex = n; ++ err = opt_add(opt, a->args[1].from, opts->sb_flags, ++ bindex); ++ if (!err) ++ opt->type = token; ++ break; ++ case Opt_append: ++ err = opt_add(opt, a->args[0].from, opts->sb_flags, ++ /*dummy bindex*/1); ++ if (!err) ++ opt->type = token; ++ break; ++ case Opt_prepend: ++ err = opt_add(opt, a->args[0].from, opts->sb_flags, ++ /*bindex*/0); ++ if (!err) ++ opt->type = token; ++ break; ++ case Opt_del: ++ err = au_opts_parse_del(&opt->del, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++#if 0 /* reserved for future use */ ++ case Opt_idel: ++ del->pathname = "(indexed)"; ++ if (unlikely(match_int(&args[0], &n))) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ err = au_opts_parse_idel(sb, n, &opt->del, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++#endif ++ case Opt_mod: ++ err = au_opts_parse_mod(&opt->mod, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++#ifdef IMOD /* reserved for future use */ ++ case Opt_imod: ++ u.mod->path = "(indexed)"; ++ if (unlikely(match_int(&a->args[0], &n))) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ err = au_opts_parse_imod(sb, n, &opt->mod, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++#endif ++ case Opt_xino: ++ err = au_opts_parse_xino(sb, &opt->xino, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++ ++ case Opt_trunc_xino_path: ++ err = au_opts_parse_xino_itrunc_path ++ (sb, &opt->xino_itrunc, a->args); ++ if (!err) ++ opt->type = token; ++ break; ++ ++ case Opt_itrunc_xino: ++ u.xino_itrunc = &opt->xino_itrunc; ++ if (unlikely(match_int(&a->args[0], &n))) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ u.xino_itrunc->bindex = n; ++ aufs_read_lock(root, AuLock_FLUSH); ++ if (n < 0 || au_sbend(sb) < n) { ++ pr_err("out of bounds, %d\n", n); ++ aufs_read_unlock(root, !AuLock_IR); ++ break; ++ } ++ aufs_read_unlock(root, !AuLock_IR); ++ err = 0; ++ opt->type = token; ++ break; ++ ++ case Opt_dirwh: ++ if (unlikely(match_int(&a->args[0], &opt->dirwh))) ++ break; ++ err = 0; ++ opt->type = token; ++ break; ++ ++ case Opt_rdcache: ++ if (unlikely(match_int(&a->args[0], &n))) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ if (unlikely(n > AUFS_RDCACHE_MAX)) { ++ pr_err("rdcache must be smaller than %d\n", ++ AUFS_RDCACHE_MAX); ++ break; ++ } ++ opt->rdcache = n; ++ err = 0; ++ opt->type = token; ++ break; ++ case Opt_rdblk: ++ if (unlikely(match_int(&a->args[0], &n) ++ || n < 0 ++ || n > KMALLOC_MAX_SIZE)) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ if (unlikely(n && n < NAME_MAX)) { ++ pr_err("rdblk must be larger than %d\n", ++ NAME_MAX); ++ break; ++ } ++ opt->rdblk = n; ++ err = 0; ++ opt->type = token; ++ break; ++ case Opt_rdhash: ++ if (unlikely(match_int(&a->args[0], &n) ++ || n < 0 ++ || n * sizeof(struct hlist_head) ++ > KMALLOC_MAX_SIZE)) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ opt->rdhash = n; ++ err = 0; ++ opt->type = token; ++ break; ++ ++ case Opt_trunc_xino: ++ case Opt_notrunc_xino: ++ case Opt_noxino: ++ case Opt_trunc_xib: ++ case Opt_notrunc_xib: ++ case Opt_shwh: ++ case Opt_noshwh: ++ case Opt_dirperm1: ++ case Opt_nodirperm1: ++ case Opt_plink: ++ case Opt_noplink: ++ case Opt_list_plink: ++ case Opt_dio: ++ case Opt_nodio: ++ case Opt_diropq_a: ++ case Opt_diropq_w: ++ case Opt_warn_perm: ++ case Opt_nowarn_perm: ++ case Opt_verbose: ++ case Opt_noverbose: ++ case Opt_sum: ++ case Opt_nosum: ++ case Opt_wsum: ++ case Opt_rdblk_def: ++ case Opt_rdhash_def: ++ case Opt_acl: ++ case Opt_noacl: ++ err = 0; ++ opt->type = token; ++ break; ++ ++ case Opt_udba: ++ opt->udba = udba_val(a->args[0].from); ++ if (opt->udba >= 0) { ++ err = 0; ++ opt->type = token; ++ } else ++ pr_err("wrong value, %s\n", opt_str); ++ break; ++ ++ case Opt_wbr_create: ++ u.create = &opt->wbr_create; ++ u.create->wbr_create ++ = au_wbr_create_val(a->args[0].from, u.create); ++ if (u.create->wbr_create >= 0) { ++ err = 0; ++ opt->type = token; ++ } else ++ pr_err("wrong value, %s\n", opt_str); ++ break; ++ case Opt_wbr_copyup: ++ opt->wbr_copyup = au_wbr_copyup_val(a->args[0].from); ++ if (opt->wbr_copyup >= 0) { ++ err = 0; ++ opt->type = token; ++ } else ++ pr_err("wrong value, %s\n", opt_str); ++ break; ++ ++ case Opt_fhsm_sec: ++ if (unlikely(match_int(&a->args[0], &n) ++ || n < 0)) { ++ pr_err("bad integer in %s\n", opt_str); ++ break; ++ } ++ if (sysaufs_brs) { ++ opt->fhsm_second = n; ++ opt->type = token; ++ } else ++ pr_warn("ignored %s\n", opt_str); ++ err = 0; ++ break; ++ ++ case Opt_ignore: ++ pr_warn("ignored %s\n", opt_str); ++ /*FALLTHROUGH*/ ++ case Opt_ignore_silent: ++ skipped = 1; ++ err = 0; ++ break; ++ case Opt_err: ++ pr_err("unknown option %s\n", opt_str); ++ break; ++ } ++ ++ if (!err && !skipped) { ++ if (unlikely(++opt > opt_tail)) { ++ err = -E2BIG; ++ opt--; ++ opt->type = Opt_tail; ++ break; ++ } ++ opt->type = Opt_tail; ++ } ++ } ++ ++ kfree(a); ++ dump_opts(opts); ++ if (unlikely(err)) ++ au_opts_free(opts); ++ ++out: ++ return err; ++} ++ ++static int au_opt_wbr_create(struct super_block *sb, ++ struct au_opt_wbr_create *create) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ err = 1; /* handled */ ++ sbinfo = au_sbi(sb); ++ if (sbinfo->si_wbr_create_ops->fin) { ++ err = sbinfo->si_wbr_create_ops->fin(sb); ++ if (!err) ++ err = 1; ++ } ++ ++ sbinfo->si_wbr_create = create->wbr_create; ++ sbinfo->si_wbr_create_ops = au_wbr_create_ops + create->wbr_create; ++ switch (create->wbr_create) { ++ case AuWbrCreate_MFSRRV: ++ case AuWbrCreate_MFSRR: ++ case AuWbrCreate_PMFSRR: ++ case AuWbrCreate_PMFSRRV: ++ sbinfo->si_wbr_mfs.mfsrr_watermark = create->mfsrr_watermark; ++ /*FALLTHROUGH*/ ++ case AuWbrCreate_MFS: ++ case AuWbrCreate_MFSV: ++ case AuWbrCreate_PMFS: ++ case AuWbrCreate_PMFSV: ++ sbinfo->si_wbr_mfs.mfs_expire ++ = msecs_to_jiffies(create->mfs_second * MSEC_PER_SEC); ++ break; ++ } ++ ++ if (sbinfo->si_wbr_create_ops->init) ++ sbinfo->si_wbr_create_ops->init(sb); /* ignore */ ++ ++ return err; ++} ++ ++/* ++ * returns, ++ * plus: processed without an error ++ * zero: unprocessed ++ */ ++static int au_opt_simple(struct super_block *sb, struct au_opt *opt, ++ struct au_opts *opts) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ err = 1; /* handled */ ++ sbinfo = au_sbi(sb); ++ switch (opt->type) { ++ case Opt_udba: ++ sbinfo->si_mntflags &= ~AuOptMask_UDBA; ++ sbinfo->si_mntflags |= opt->udba; ++ opts->given_udba |= opt->udba; ++ break; ++ ++ case Opt_plink: ++ au_opt_set(sbinfo->si_mntflags, PLINK); ++ break; ++ case Opt_noplink: ++ if (au_opt_test(sbinfo->si_mntflags, PLINK)) ++ au_plink_put(sb, /*verbose*/1); ++ au_opt_clr(sbinfo->si_mntflags, PLINK); ++ break; ++ case Opt_list_plink: ++ if (au_opt_test(sbinfo->si_mntflags, PLINK)) ++ au_plink_list(sb); ++ break; ++ ++ case Opt_dio: ++ au_opt_set(sbinfo->si_mntflags, DIO); ++ au_fset_opts(opts->flags, REFRESH_DYAOP); ++ break; ++ case Opt_nodio: ++ au_opt_clr(sbinfo->si_mntflags, DIO); ++ au_fset_opts(opts->flags, REFRESH_DYAOP); ++ break; ++ ++ case Opt_fhsm_sec: ++ au_fhsm_set(sbinfo, opt->fhsm_second); ++ break; ++ ++ case Opt_diropq_a: ++ au_opt_set(sbinfo->si_mntflags, ALWAYS_DIROPQ); ++ break; ++ case Opt_diropq_w: ++ au_opt_clr(sbinfo->si_mntflags, ALWAYS_DIROPQ); ++ break; ++ ++ case Opt_warn_perm: ++ au_opt_set(sbinfo->si_mntflags, WARN_PERM); ++ break; ++ case Opt_nowarn_perm: ++ au_opt_clr(sbinfo->si_mntflags, WARN_PERM); ++ break; ++ ++ case Opt_verbose: ++ au_opt_set(sbinfo->si_mntflags, VERBOSE); ++ break; ++ case Opt_noverbose: ++ au_opt_clr(sbinfo->si_mntflags, VERBOSE); ++ break; ++ ++ case Opt_sum: ++ au_opt_set(sbinfo->si_mntflags, SUM); ++ break; ++ case Opt_wsum: ++ au_opt_clr(sbinfo->si_mntflags, SUM); ++ au_opt_set(sbinfo->si_mntflags, SUM_W); ++ case Opt_nosum: ++ au_opt_clr(sbinfo->si_mntflags, SUM); ++ au_opt_clr(sbinfo->si_mntflags, SUM_W); ++ break; ++ ++ case Opt_wbr_create: ++ err = au_opt_wbr_create(sb, &opt->wbr_create); ++ break; ++ case Opt_wbr_copyup: ++ sbinfo->si_wbr_copyup = opt->wbr_copyup; ++ sbinfo->si_wbr_copyup_ops = au_wbr_copyup_ops + opt->wbr_copyup; ++ break; ++ ++ case Opt_dirwh: ++ sbinfo->si_dirwh = opt->dirwh; ++ break; ++ ++ case Opt_rdcache: ++ sbinfo->si_rdcache ++ = msecs_to_jiffies(opt->rdcache * MSEC_PER_SEC); ++ break; ++ case Opt_rdblk: ++ sbinfo->si_rdblk = opt->rdblk; ++ break; ++ case Opt_rdblk_def: ++ sbinfo->si_rdblk = AUFS_RDBLK_DEF; ++ break; ++ case Opt_rdhash: ++ sbinfo->si_rdhash = opt->rdhash; ++ break; ++ case Opt_rdhash_def: ++ sbinfo->si_rdhash = AUFS_RDHASH_DEF; ++ break; ++ ++ case Opt_shwh: ++ au_opt_set(sbinfo->si_mntflags, SHWH); ++ break; ++ case Opt_noshwh: ++ au_opt_clr(sbinfo->si_mntflags, SHWH); ++ break; ++ ++ case Opt_dirperm1: ++ au_opt_set(sbinfo->si_mntflags, DIRPERM1); ++ break; ++ case Opt_nodirperm1: ++ au_opt_clr(sbinfo->si_mntflags, DIRPERM1); ++ break; ++ ++ case Opt_trunc_xino: ++ au_opt_set(sbinfo->si_mntflags, TRUNC_XINO); ++ break; ++ case Opt_notrunc_xino: ++ au_opt_clr(sbinfo->si_mntflags, TRUNC_XINO); ++ break; ++ ++ case Opt_trunc_xino_path: ++ case Opt_itrunc_xino: ++ err = au_xino_trunc(sb, opt->xino_itrunc.bindex); ++ if (!err) ++ err = 1; ++ break; ++ ++ case Opt_trunc_xib: ++ au_fset_opts(opts->flags, TRUNC_XIB); ++ break; ++ case Opt_notrunc_xib: ++ au_fclr_opts(opts->flags, TRUNC_XIB); ++ break; ++ ++ case Opt_acl: ++ sb->s_flags |= MS_POSIXACL; ++ break; ++ case Opt_noacl: ++ sb->s_flags &= ~MS_POSIXACL; ++ break; ++ ++ default: ++ err = 0; ++ break; ++ } ++ ++ return err; ++} ++ ++/* ++ * returns tri-state. ++ * plus: processed without an error ++ * zero: unprocessed ++ * minus: error ++ */ ++static int au_opt_br(struct super_block *sb, struct au_opt *opt, ++ struct au_opts *opts) ++{ ++ int err, do_refresh; ++ ++ err = 0; ++ switch (opt->type) { ++ case Opt_append: ++ opt->add.bindex = au_sbend(sb) + 1; ++ if (opt->add.bindex < 0) ++ opt->add.bindex = 0; ++ goto add; ++ case Opt_prepend: ++ opt->add.bindex = 0; ++ add: /* indented label */ ++ case Opt_add: ++ err = au_br_add(sb, &opt->add, ++ au_ftest_opts(opts->flags, REMOUNT)); ++ if (!err) { ++ err = 1; ++ au_fset_opts(opts->flags, REFRESH); ++ } ++ break; ++ ++ case Opt_del: ++ case Opt_idel: ++ err = au_br_del(sb, &opt->del, ++ au_ftest_opts(opts->flags, REMOUNT)); ++ if (!err) { ++ err = 1; ++ au_fset_opts(opts->flags, TRUNC_XIB); ++ au_fset_opts(opts->flags, REFRESH); ++ } ++ break; ++ ++ case Opt_mod: ++ case Opt_imod: ++ err = au_br_mod(sb, &opt->mod, ++ au_ftest_opts(opts->flags, REMOUNT), ++ &do_refresh); ++ if (!err) { ++ err = 1; ++ if (do_refresh) ++ au_fset_opts(opts->flags, REFRESH); ++ } ++ break; ++ } ++ ++ return err; ++} ++ ++static int au_opt_xino(struct super_block *sb, struct au_opt *opt, ++ struct au_opt_xino **opt_xino, ++ struct au_opts *opts) ++{ ++ int err; ++ aufs_bindex_t bend, bindex; ++ struct dentry *root, *parent, *h_root; ++ ++ err = 0; ++ switch (opt->type) { ++ case Opt_xino: ++ err = au_xino_set(sb, &opt->xino, ++ !!au_ftest_opts(opts->flags, REMOUNT)); ++ if (unlikely(err)) ++ break; ++ ++ *opt_xino = &opt->xino; ++ au_xino_brid_set(sb, -1); ++ ++ /* safe d_parent access */ ++ parent = opt->xino.file->f_path.dentry->d_parent; ++ root = sb->s_root; ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ h_root = au_h_dptr(root, bindex); ++ if (h_root == parent) { ++ au_xino_brid_set(sb, au_sbr_id(sb, bindex)); ++ break; ++ } ++ } ++ break; ++ ++ case Opt_noxino: ++ au_xino_clr(sb); ++ au_xino_brid_set(sb, -1); ++ *opt_xino = (void *)-1; ++ break; ++ } ++ ++ return err; ++} ++ ++int au_opts_verify(struct super_block *sb, unsigned long sb_flags, ++ unsigned int pending) ++{ ++ int err, fhsm; ++ aufs_bindex_t bindex, bend; ++ unsigned char do_plink, skip, do_free; ++ struct au_branch *br; ++ struct au_wbr *wbr; ++ struct dentry *root; ++ struct inode *dir, *h_dir; ++ struct au_sbinfo *sbinfo; ++ struct au_hinode *hdir; ++ ++ SiMustAnyLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!(sbinfo->si_mntflags & AuOptMask_UDBA)); ++ ++ if (!(sb_flags & MS_RDONLY)) { ++ if (unlikely(!au_br_writable(au_sbr_perm(sb, 0)))) ++ pr_warn("first branch should be rw\n"); ++ if (unlikely(au_opt_test(sbinfo->si_mntflags, SHWH))) ++ pr_warn("shwh should be used with ro\n"); ++ } ++ ++ if (au_opt_test((sbinfo->si_mntflags | pending), UDBA_HNOTIFY) ++ && !au_opt_test(sbinfo->si_mntflags, XINO)) ++ pr_warn("udba=*notify requires xino\n"); ++ ++ if (au_opt_test(sbinfo->si_mntflags, DIRPERM1)) ++ pr_warn("dirperm1 breaks the protection" ++ " by the permission bits on the lower branch\n"); ++ ++ err = 0; ++ fhsm = 0; ++ root = sb->s_root; ++ dir = d_inode(root); ++ do_plink = !!au_opt_test(sbinfo->si_mntflags, PLINK); ++ bend = au_sbend(sb); ++ for (bindex = 0; !err && bindex <= bend; bindex++) { ++ skip = 0; ++ h_dir = au_h_iptr(dir, bindex); ++ br = au_sbr(sb, bindex); ++ ++ if ((br->br_perm & AuBrAttr_ICEX) ++ && !h_dir->i_op->listxattr) ++ br->br_perm &= ~AuBrAttr_ICEX; ++#if 0 ++ if ((br->br_perm & AuBrAttr_ICEX_SEC) ++ && (au_br_sb(br)->s_flags & MS_NOSEC)) ++ br->br_perm &= ~AuBrAttr_ICEX_SEC; ++#endif ++ ++ do_free = 0; ++ wbr = br->br_wbr; ++ if (wbr) ++ wbr_wh_read_lock(wbr); ++ ++ if (!au_br_writable(br->br_perm)) { ++ do_free = !!wbr; ++ skip = (!wbr ++ || (!wbr->wbr_whbase ++ && !wbr->wbr_plink ++ && !wbr->wbr_orph)); ++ } else if (!au_br_wh_linkable(br->br_perm)) { ++ /* skip = (!br->br_whbase && !br->br_orph); */ ++ skip = (!wbr || !wbr->wbr_whbase); ++ if (skip && wbr) { ++ if (do_plink) ++ skip = !!wbr->wbr_plink; ++ else ++ skip = !wbr->wbr_plink; ++ } ++ } else { ++ /* skip = (br->br_whbase && br->br_ohph); */ ++ skip = (wbr && wbr->wbr_whbase); ++ if (skip) { ++ if (do_plink) ++ skip = !!wbr->wbr_plink; ++ else ++ skip = !wbr->wbr_plink; ++ } ++ } ++ if (wbr) ++ wbr_wh_read_unlock(wbr); ++ ++ if (au_br_fhsm(br->br_perm)) { ++ fhsm++; ++ AuDebugOn(!br->br_fhsm); ++ } ++ ++ if (skip) ++ continue; ++ ++ hdir = au_hi(dir, bindex); ++ au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); ++ if (wbr) ++ wbr_wh_write_lock(wbr); ++ err = au_wh_init(br, sb); ++ if (wbr) ++ wbr_wh_write_unlock(wbr); ++ au_hn_imtx_unlock(hdir); ++ ++ if (!err && do_free) { ++ kfree(wbr); ++ br->br_wbr = NULL; ++ } ++ } ++ ++ if (fhsm >= 2) { ++ au_fset_si(sbinfo, FHSM); ++ for (bindex = bend; bindex >= 0; bindex--) { ++ br = au_sbr(sb, bindex); ++ if (au_br_fhsm(br->br_perm)) { ++ au_fhsm_set_bottom(sb, bindex); ++ break; ++ } ++ } ++ } else { ++ au_fclr_si(sbinfo, FHSM); ++ au_fhsm_set_bottom(sb, -1); ++ } ++ ++ return err; ++} ++ ++int au_opts_mount(struct super_block *sb, struct au_opts *opts) ++{ ++ int err; ++ unsigned int tmp; ++ aufs_bindex_t bindex, bend; ++ struct au_opt *opt; ++ struct au_opt_xino *opt_xino, xino; ++ struct au_sbinfo *sbinfo; ++ struct au_branch *br; ++ struct inode *dir; ++ ++ SiMustWriteLock(sb); ++ ++ err = 0; ++ opt_xino = NULL; ++ opt = opts->opt; ++ while (err >= 0 && opt->type != Opt_tail) ++ err = au_opt_simple(sb, opt++, opts); ++ if (err > 0) ++ err = 0; ++ else if (unlikely(err < 0)) ++ goto out; ++ ++ /* disable xino and udba temporary */ ++ sbinfo = au_sbi(sb); ++ tmp = sbinfo->si_mntflags; ++ au_opt_clr(sbinfo->si_mntflags, XINO); ++ au_opt_set_udba(sbinfo->si_mntflags, UDBA_REVAL); ++ ++ opt = opts->opt; ++ while (err >= 0 && opt->type != Opt_tail) ++ err = au_opt_br(sb, opt++, opts); ++ if (err > 0) ++ err = 0; ++ else if (unlikely(err < 0)) ++ goto out; ++ ++ bend = au_sbend(sb); ++ if (unlikely(bend < 0)) { ++ err = -EINVAL; ++ pr_err("no branches\n"); ++ goto out; ++ } ++ ++ if (au_opt_test(tmp, XINO)) ++ au_opt_set(sbinfo->si_mntflags, XINO); ++ opt = opts->opt; ++ while (!err && opt->type != Opt_tail) ++ err = au_opt_xino(sb, opt++, &opt_xino, opts); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_opts_verify(sb, sb->s_flags, tmp); ++ if (unlikely(err)) ++ goto out; ++ ++ /* restore xino */ ++ if (au_opt_test(tmp, XINO) && !opt_xino) { ++ xino.file = au_xino_def(sb); ++ err = PTR_ERR(xino.file); ++ if (IS_ERR(xino.file)) ++ goto out; ++ ++ err = au_xino_set(sb, &xino, /*remount*/0); ++ fput(xino.file); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++ /* restore udba */ ++ tmp &= AuOptMask_UDBA; ++ sbinfo->si_mntflags &= ~AuOptMask_UDBA; ++ sbinfo->si_mntflags |= tmp; ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ err = au_hnotify_reset_br(tmp, br, br->br_perm); ++ if (unlikely(err)) ++ AuIOErr("hnotify failed on br %d, %d, ignored\n", ++ bindex, err); ++ /* go on even if err */ ++ } ++ if (au_opt_test(tmp, UDBA_HNOTIFY)) { ++ dir = d_inode(sb->s_root); ++ au_hn_reset(dir, au_hi_flags(dir, /*isdir*/1) & ~AuHi_XINO); ++ } ++ ++out: ++ return err; ++} ++ ++int au_opts_remount(struct super_block *sb, struct au_opts *opts) ++{ ++ int err, rerr; ++ struct inode *dir; ++ struct au_opt_xino *opt_xino; ++ struct au_opt *opt; ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ dir = d_inode(sb->s_root); ++ sbinfo = au_sbi(sb); ++ err = 0; ++ opt_xino = NULL; ++ opt = opts->opt; ++ while (err >= 0 && opt->type != Opt_tail) { ++ err = au_opt_simple(sb, opt, opts); ++ if (!err) ++ err = au_opt_br(sb, opt, opts); ++ if (!err) ++ err = au_opt_xino(sb, opt, &opt_xino, opts); ++ opt++; ++ } ++ if (err > 0) ++ err = 0; ++ AuTraceErr(err); ++ /* go on even err */ ++ ++ rerr = au_opts_verify(sb, opts->sb_flags, /*pending*/0); ++ if (unlikely(rerr && !err)) ++ err = rerr; ++ ++ if (au_ftest_opts(opts->flags, TRUNC_XIB)) { ++ rerr = au_xib_trunc(sb); ++ if (unlikely(rerr && !err)) ++ err = rerr; ++ } ++ ++ /* will be handled by the caller */ ++ if (!au_ftest_opts(opts->flags, REFRESH) ++ && (opts->given_udba || au_opt_test(sbinfo->si_mntflags, XINO))) ++ au_fset_opts(opts->flags, REFRESH); ++ ++ AuDbg("status 0x%x\n", opts->flags); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++unsigned int au_opt_udba(struct super_block *sb) ++{ ++ return au_mntflags(sb) & AuOptMask_UDBA; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/opts.h linux-4.1.10/fs/aufs/opts.h +--- linux-4.1.10.orig/fs/aufs/opts.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/opts.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,210 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * mount options/flags ++ */ ++ ++#ifndef __AUFS_OPTS_H__ ++#define __AUFS_OPTS_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/path.h> ++ ++struct file; ++struct super_block; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* mount flags */ ++#define AuOpt_XINO 1 /* external inode number bitmap ++ and translation table */ ++#define AuOpt_TRUNC_XINO (1 << 1) /* truncate xino files */ ++#define AuOpt_UDBA_NONE (1 << 2) /* users direct branch access */ ++#define AuOpt_UDBA_REVAL (1 << 3) ++#define AuOpt_UDBA_HNOTIFY (1 << 4) ++#define AuOpt_SHWH (1 << 5) /* show whiteout */ ++#define AuOpt_PLINK (1 << 6) /* pseudo-link */ ++#define AuOpt_DIRPERM1 (1 << 7) /* ignore the lower dir's perm ++ bits */ ++#define AuOpt_ALWAYS_DIROPQ (1 << 9) /* policy to creating diropq */ ++#define AuOpt_SUM (1 << 10) /* summation for statfs(2) */ ++#define AuOpt_SUM_W (1 << 11) /* unimplemented */ ++#define AuOpt_WARN_PERM (1 << 12) /* warn when add-branch */ ++#define AuOpt_VERBOSE (1 << 13) /* busy inode when del-branch */ ++#define AuOpt_DIO (1 << 14) /* direct io */ ++ ++#ifndef CONFIG_AUFS_HNOTIFY ++#undef AuOpt_UDBA_HNOTIFY ++#define AuOpt_UDBA_HNOTIFY 0 ++#endif ++#ifndef CONFIG_AUFS_SHWH ++#undef AuOpt_SHWH ++#define AuOpt_SHWH 0 ++#endif ++ ++#define AuOpt_Def (AuOpt_XINO \ ++ | AuOpt_UDBA_REVAL \ ++ | AuOpt_PLINK \ ++ /* | AuOpt_DIRPERM1 */ \ ++ | AuOpt_WARN_PERM) ++#define AuOptMask_UDBA (AuOpt_UDBA_NONE \ ++ | AuOpt_UDBA_REVAL \ ++ | AuOpt_UDBA_HNOTIFY) ++ ++#define au_opt_test(flags, name) (flags & AuOpt_##name) ++#define au_opt_set(flags, name) do { \ ++ BUILD_BUG_ON(AuOpt_##name & AuOptMask_UDBA); \ ++ ((flags) |= AuOpt_##name); \ ++} while (0) ++#define au_opt_set_udba(flags, name) do { \ ++ (flags) &= ~AuOptMask_UDBA; \ ++ ((flags) |= AuOpt_##name); \ ++} while (0) ++#define au_opt_clr(flags, name) do { \ ++ ((flags) &= ~AuOpt_##name); \ ++} while (0) ++ ++static inline unsigned int au_opts_plink(unsigned int mntflags) ++{ ++#ifdef CONFIG_PROC_FS ++ return mntflags; ++#else ++ return mntflags & ~AuOpt_PLINK; ++#endif ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* policies to select one among multiple writable branches */ ++enum { ++ AuWbrCreate_TDP, /* top down parent */ ++ AuWbrCreate_RR, /* round robin */ ++ AuWbrCreate_MFS, /* most free space */ ++ AuWbrCreate_MFSV, /* mfs with seconds */ ++ AuWbrCreate_MFSRR, /* mfs then rr */ ++ AuWbrCreate_MFSRRV, /* mfs then rr with seconds */ ++ AuWbrCreate_PMFS, /* parent and mfs */ ++ AuWbrCreate_PMFSV, /* parent and mfs with seconds */ ++ AuWbrCreate_PMFSRR, /* parent, mfs and round-robin */ ++ AuWbrCreate_PMFSRRV, /* plus seconds */ ++ ++ AuWbrCreate_Def = AuWbrCreate_TDP ++}; ++ ++enum { ++ AuWbrCopyup_TDP, /* top down parent */ ++ AuWbrCopyup_BUP, /* bottom up parent */ ++ AuWbrCopyup_BU, /* bottom up */ ++ ++ AuWbrCopyup_Def = AuWbrCopyup_TDP ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_opt_add { ++ aufs_bindex_t bindex; ++ char *pathname; ++ int perm; ++ struct path path; ++}; ++ ++struct au_opt_del { ++ char *pathname; ++ struct path h_path; ++}; ++ ++struct au_opt_mod { ++ char *path; ++ int perm; ++ struct dentry *h_root; ++}; ++ ++struct au_opt_xino { ++ char *path; ++ struct file *file; ++}; ++ ++struct au_opt_xino_itrunc { ++ aufs_bindex_t bindex; ++}; ++ ++struct au_opt_wbr_create { ++ int wbr_create; ++ int mfs_second; ++ unsigned long long mfsrr_watermark; ++}; ++ ++struct au_opt { ++ int type; ++ union { ++ struct au_opt_xino xino; ++ struct au_opt_xino_itrunc xino_itrunc; ++ struct au_opt_add add; ++ struct au_opt_del del; ++ struct au_opt_mod mod; ++ int dirwh; ++ int rdcache; ++ unsigned int rdblk; ++ unsigned int rdhash; ++ int udba; ++ struct au_opt_wbr_create wbr_create; ++ int wbr_copyup; ++ unsigned int fhsm_second; ++ }; ++}; ++ ++/* opts flags */ ++#define AuOpts_REMOUNT 1 ++#define AuOpts_REFRESH (1 << 1) ++#define AuOpts_TRUNC_XIB (1 << 2) ++#define AuOpts_REFRESH_DYAOP (1 << 3) ++#define au_ftest_opts(flags, name) ((flags) & AuOpts_##name) ++#define au_fset_opts(flags, name) \ ++ do { (flags) |= AuOpts_##name; } while (0) ++#define au_fclr_opts(flags, name) \ ++ do { (flags) &= ~AuOpts_##name; } while (0) ++ ++struct au_opts { ++ struct au_opt *opt; ++ int max_opt; ++ ++ unsigned int given_udba; ++ unsigned int flags; ++ unsigned long sb_flags; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* opts.c */ ++void au_optstr_br_perm(au_br_perm_str_t *str, int perm); ++const char *au_optstr_udba(int udba); ++const char *au_optstr_wbr_copyup(int wbr_copyup); ++const char *au_optstr_wbr_create(int wbr_create); ++ ++void au_opts_free(struct au_opts *opts); ++int au_opts_parse(struct super_block *sb, char *str, struct au_opts *opts); ++int au_opts_verify(struct super_block *sb, unsigned long sb_flags, ++ unsigned int pending); ++int au_opts_mount(struct super_block *sb, struct au_opts *opts); ++int au_opts_remount(struct super_block *sb, struct au_opts *opts); ++ ++unsigned int au_opt_udba(struct super_block *sb); ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_OPTS_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/plink.c linux-4.1.10/fs/aufs/plink.c +--- linux-4.1.10.orig/fs/aufs/plink.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/plink.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,528 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * pseudo-link ++ */ ++ ++#include "aufs.h" ++ ++/* ++ * the pseudo-link maintenance mode. ++ * during a user process maintains the pseudo-links, ++ * prohibit adding a new plink and branch manipulation. ++ * ++ * Flags ++ * NOPLM: ++ * For entry functions which will handle plink, and i_mutex is already held ++ * in VFS. ++ * They cannot wait and should return an error at once. ++ * Callers has to check the error. ++ * NOPLMW: ++ * For entry functions which will handle plink, but i_mutex is not held ++ * in VFS. ++ * They can wait the plink maintenance mode to finish. ++ * ++ * They behave like F_SETLK and F_SETLKW. ++ * If the caller never handle plink, then both flags are unnecessary. ++ */ ++ ++int au_plink_maint(struct super_block *sb, int flags) ++{ ++ int err; ++ pid_t pid, ppid; ++ struct au_sbinfo *sbi; ++ ++ SiMustAnyLock(sb); ++ ++ err = 0; ++ if (!au_opt_test(au_mntflags(sb), PLINK)) ++ goto out; ++ ++ sbi = au_sbi(sb); ++ pid = sbi->si_plink_maint_pid; ++ if (!pid || pid == current->pid) ++ goto out; ++ ++ /* todo: it highly depends upon /sbin/mount.aufs */ ++ rcu_read_lock(); ++ ppid = task_pid_vnr(rcu_dereference(current->real_parent)); ++ rcu_read_unlock(); ++ if (pid == ppid) ++ goto out; ++ ++ if (au_ftest_lock(flags, NOPLMW)) { ++ /* if there is no i_mutex lock in VFS, we don't need to wait */ ++ /* AuDebugOn(!lockdep_depth(current)); */ ++ while (sbi->si_plink_maint_pid) { ++ si_read_unlock(sb); ++ /* gave up wake_up_bit() */ ++ wait_event(sbi->si_plink_wq, !sbi->si_plink_maint_pid); ++ ++ if (au_ftest_lock(flags, FLUSH)) ++ au_nwt_flush(&sbi->si_nowait); ++ si_noflush_read_lock(sb); ++ } ++ } else if (au_ftest_lock(flags, NOPLM)) { ++ AuDbg("ppid %d, pid %d\n", ppid, pid); ++ err = -EAGAIN; ++ } ++ ++out: ++ return err; ++} ++ ++void au_plink_maint_leave(struct au_sbinfo *sbinfo) ++{ ++ spin_lock(&sbinfo->si_plink_maint_lock); ++ sbinfo->si_plink_maint_pid = 0; ++ spin_unlock(&sbinfo->si_plink_maint_lock); ++ wake_up_all(&sbinfo->si_plink_wq); ++} ++ ++int au_plink_maint_enter(struct super_block *sb) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ /* make sure i am the only one in this fs */ ++ si_write_lock(sb, AuLock_FLUSH); ++ if (au_opt_test(au_mntflags(sb), PLINK)) { ++ spin_lock(&sbinfo->si_plink_maint_lock); ++ if (!sbinfo->si_plink_maint_pid) ++ sbinfo->si_plink_maint_pid = current->pid; ++ else ++ err = -EBUSY; ++ spin_unlock(&sbinfo->si_plink_maint_lock); ++ } ++ si_write_unlock(sb); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_DEBUG ++void au_plink_list(struct super_block *sb) ++{ ++ int i; ++ struct au_sbinfo *sbinfo; ++ struct hlist_head *plink_hlist; ++ struct pseudo_link *plink; ++ ++ SiMustAnyLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); ++ AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); ++ ++ for (i = 0; i < AuPlink_NHASH; i++) { ++ plink_hlist = &sbinfo->si_plink[i].head; ++ rcu_read_lock(); ++ hlist_for_each_entry_rcu(plink, plink_hlist, hlist) ++ AuDbg("%lu\n", plink->inode->i_ino); ++ rcu_read_unlock(); ++ } ++} ++#endif ++ ++/* is the inode pseudo-linked? */ ++int au_plink_test(struct inode *inode) ++{ ++ int found, i; ++ struct au_sbinfo *sbinfo; ++ struct hlist_head *plink_hlist; ++ struct pseudo_link *plink; ++ ++ sbinfo = au_sbi(inode->i_sb); ++ AuRwMustAnyLock(&sbinfo->si_rwsem); ++ AuDebugOn(!au_opt_test(au_mntflags(inode->i_sb), PLINK)); ++ AuDebugOn(au_plink_maint(inode->i_sb, AuLock_NOPLM)); ++ ++ found = 0; ++ i = au_plink_hash(inode->i_ino); ++ plink_hlist = &sbinfo->si_plink[i].head; ++ rcu_read_lock(); ++ hlist_for_each_entry_rcu(plink, plink_hlist, hlist) ++ if (plink->inode == inode) { ++ found = 1; ++ break; ++ } ++ rcu_read_unlock(); ++ return found; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * generate a name for plink. ++ * the file will be stored under AUFS_WH_PLINKDIR. ++ */ ++/* 20 is max digits length of ulong 64 */ ++#define PLINK_NAME_LEN ((20 + 1) * 2) ++ ++static int plink_name(char *name, int len, struct inode *inode, ++ aufs_bindex_t bindex) ++{ ++ int rlen; ++ struct inode *h_inode; ++ ++ h_inode = au_h_iptr(inode, bindex); ++ rlen = snprintf(name, len, "%lu.%lu", inode->i_ino, h_inode->i_ino); ++ return rlen; ++} ++ ++struct au_do_plink_lkup_args { ++ struct dentry **errp; ++ struct qstr *tgtname; ++ struct dentry *h_parent; ++ struct au_branch *br; ++}; ++ ++static struct dentry *au_do_plink_lkup(struct qstr *tgtname, ++ struct dentry *h_parent, ++ struct au_branch *br) ++{ ++ struct dentry *h_dentry; ++ struct mutex *h_mtx; ++ ++ h_mtx = &d_inode(h_parent)->i_mutex; ++ mutex_lock_nested(h_mtx, AuLsc_I_CHILD2); ++ h_dentry = vfsub_lkup_one(tgtname, h_parent); ++ mutex_unlock(h_mtx); ++ return h_dentry; ++} ++ ++static void au_call_do_plink_lkup(void *args) ++{ ++ struct au_do_plink_lkup_args *a = args; ++ *a->errp = au_do_plink_lkup(a->tgtname, a->h_parent, a->br); ++} ++ ++/* lookup the plink-ed @inode under the branch at @bindex */ ++struct dentry *au_plink_lkup(struct inode *inode, aufs_bindex_t bindex) ++{ ++ struct dentry *h_dentry, *h_parent; ++ struct au_branch *br; ++ int wkq_err; ++ char a[PLINK_NAME_LEN]; ++ struct qstr tgtname = QSTR_INIT(a, 0); ++ ++ AuDebugOn(au_plink_maint(inode->i_sb, AuLock_NOPLM)); ++ ++ br = au_sbr(inode->i_sb, bindex); ++ h_parent = br->br_wbr->wbr_plink; ++ tgtname.len = plink_name(a, sizeof(a), inode, bindex); ++ ++ if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) { ++ struct au_do_plink_lkup_args args = { ++ .errp = &h_dentry, ++ .tgtname = &tgtname, ++ .h_parent = h_parent, ++ .br = br ++ }; ++ ++ wkq_err = au_wkq_wait(au_call_do_plink_lkup, &args); ++ if (unlikely(wkq_err)) ++ h_dentry = ERR_PTR(wkq_err); ++ } else ++ h_dentry = au_do_plink_lkup(&tgtname, h_parent, br); ++ ++ return h_dentry; ++} ++ ++/* create a pseudo-link */ ++static int do_whplink(struct qstr *tgt, struct dentry *h_parent, ++ struct dentry *h_dentry, struct au_branch *br) ++{ ++ int err; ++ struct path h_path = { ++ .mnt = au_br_mnt(br) ++ }; ++ struct inode *h_dir, *delegated; ++ ++ h_dir = d_inode(h_parent); ++ mutex_lock_nested(&h_dir->i_mutex, AuLsc_I_CHILD2); ++again: ++ h_path.dentry = vfsub_lkup_one(tgt, h_parent); ++ err = PTR_ERR(h_path.dentry); ++ if (IS_ERR(h_path.dentry)) ++ goto out; ++ ++ err = 0; ++ /* wh.plink dir is not monitored */ ++ /* todo: is it really safe? */ ++ if (d_is_positive(h_path.dentry) ++ && d_inode(h_path.dentry) != d_inode(h_dentry)) { ++ delegated = NULL; ++ err = vfsub_unlink(h_dir, &h_path, &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ dput(h_path.dentry); ++ h_path.dentry = NULL; ++ if (!err) ++ goto again; ++ } ++ if (!err && d_is_negative(h_path.dentry)) { ++ delegated = NULL; ++ err = vfsub_link(h_dentry, h_dir, &h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ } ++ dput(h_path.dentry); ++ ++out: ++ mutex_unlock(&h_dir->i_mutex); ++ return err; ++} ++ ++struct do_whplink_args { ++ int *errp; ++ struct qstr *tgt; ++ struct dentry *h_parent; ++ struct dentry *h_dentry; ++ struct au_branch *br; ++}; ++ ++static void call_do_whplink(void *args) ++{ ++ struct do_whplink_args *a = args; ++ *a->errp = do_whplink(a->tgt, a->h_parent, a->h_dentry, a->br); ++} ++ ++static int whplink(struct dentry *h_dentry, struct inode *inode, ++ aufs_bindex_t bindex, struct au_branch *br) ++{ ++ int err, wkq_err; ++ struct au_wbr *wbr; ++ struct dentry *h_parent; ++ char a[PLINK_NAME_LEN]; ++ struct qstr tgtname = QSTR_INIT(a, 0); ++ ++ wbr = au_sbr(inode->i_sb, bindex)->br_wbr; ++ h_parent = wbr->wbr_plink; ++ tgtname.len = plink_name(a, sizeof(a), inode, bindex); ++ ++ /* always superio. */ ++ if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) { ++ struct do_whplink_args args = { ++ .errp = &err, ++ .tgt = &tgtname, ++ .h_parent = h_parent, ++ .h_dentry = h_dentry, ++ .br = br ++ }; ++ wkq_err = au_wkq_wait(call_do_whplink, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } else ++ err = do_whplink(&tgtname, h_parent, h_dentry, br); ++ ++ return err; ++} ++ ++/* free a single plink */ ++static void do_put_plink(struct pseudo_link *plink, int do_del) ++{ ++ if (do_del) ++ hlist_del(&plink->hlist); ++ iput(plink->inode); ++ kfree(plink); ++} ++ ++static void do_put_plink_rcu(struct rcu_head *rcu) ++{ ++ struct pseudo_link *plink; ++ ++ plink = container_of(rcu, struct pseudo_link, rcu); ++ iput(plink->inode); ++ kfree(plink); ++} ++ ++/* ++ * create a new pseudo-link for @h_dentry on @bindex. ++ * the linked inode is held in aufs @inode. ++ */ ++void au_plink_append(struct inode *inode, aufs_bindex_t bindex, ++ struct dentry *h_dentry) ++{ ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ struct hlist_head *plink_hlist; ++ struct pseudo_link *plink, *tmp; ++ struct au_sphlhead *sphl; ++ int found, err, cnt, i; ++ ++ sb = inode->i_sb; ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); ++ AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); ++ ++ found = au_plink_test(inode); ++ if (found) ++ return; ++ ++ i = au_plink_hash(inode->i_ino); ++ sphl = sbinfo->si_plink + i; ++ plink_hlist = &sphl->head; ++ tmp = kmalloc(sizeof(*plink), GFP_NOFS); ++ if (tmp) ++ tmp->inode = au_igrab(inode); ++ else { ++ err = -ENOMEM; ++ goto out; ++ } ++ ++ spin_lock(&sphl->spin); ++ hlist_for_each_entry(plink, plink_hlist, hlist) { ++ if (plink->inode == inode) { ++ found = 1; ++ break; ++ } ++ } ++ if (!found) ++ hlist_add_head_rcu(&tmp->hlist, plink_hlist); ++ spin_unlock(&sphl->spin); ++ if (!found) { ++ cnt = au_sphl_count(sphl); ++#define msg "unexpectedly unblanced or too many pseudo-links" ++ if (cnt > AUFS_PLINK_WARN) ++ AuWarn1(msg ", %d\n", cnt); ++#undef msg ++ err = whplink(h_dentry, inode, bindex, au_sbr(sb, bindex)); ++ } else { ++ do_put_plink(tmp, 0); ++ return; ++ } ++ ++out: ++ if (unlikely(err)) { ++ pr_warn("err %d, damaged pseudo link.\n", err); ++ if (tmp) { ++ au_sphl_del_rcu(&tmp->hlist, sphl); ++ call_rcu(&tmp->rcu, do_put_plink_rcu); ++ } ++ } ++} ++ ++/* free all plinks */ ++void au_plink_put(struct super_block *sb, int verbose) ++{ ++ int i, warned; ++ struct au_sbinfo *sbinfo; ++ struct hlist_head *plink_hlist; ++ struct hlist_node *tmp; ++ struct pseudo_link *plink; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); ++ AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); ++ ++ /* no spin_lock since sbinfo is write-locked */ ++ warned = 0; ++ for (i = 0; i < AuPlink_NHASH; i++) { ++ plink_hlist = &sbinfo->si_plink[i].head; ++ if (!warned && verbose && !hlist_empty(plink_hlist)) { ++ pr_warn("pseudo-link is not flushed"); ++ warned = 1; ++ } ++ hlist_for_each_entry_safe(plink, tmp, plink_hlist, hlist) ++ do_put_plink(plink, 0); ++ INIT_HLIST_HEAD(plink_hlist); ++ } ++} ++ ++void au_plink_clean(struct super_block *sb, int verbose) ++{ ++ struct dentry *root; ++ ++ root = sb->s_root; ++ aufs_write_lock(root); ++ if (au_opt_test(au_mntflags(sb), PLINK)) ++ au_plink_put(sb, verbose); ++ aufs_write_unlock(root); ++} ++ ++static int au_plink_do_half_refresh(struct inode *inode, aufs_bindex_t br_id) ++{ ++ int do_put; ++ aufs_bindex_t bstart, bend, bindex; ++ ++ do_put = 0; ++ bstart = au_ibstart(inode); ++ bend = au_ibend(inode); ++ if (bstart >= 0) { ++ for (bindex = bstart; bindex <= bend; bindex++) { ++ if (!au_h_iptr(inode, bindex) ++ || au_ii_br_id(inode, bindex) != br_id) ++ continue; ++ au_set_h_iptr(inode, bindex, NULL, 0); ++ do_put = 1; ++ break; ++ } ++ if (do_put) ++ for (bindex = bstart; bindex <= bend; bindex++) ++ if (au_h_iptr(inode, bindex)) { ++ do_put = 0; ++ break; ++ } ++ } else ++ do_put = 1; ++ ++ return do_put; ++} ++ ++/* free the plinks on a branch specified by @br_id */ ++void au_plink_half_refresh(struct super_block *sb, aufs_bindex_t br_id) ++{ ++ struct au_sbinfo *sbinfo; ++ struct hlist_head *plink_hlist; ++ struct hlist_node *tmp; ++ struct pseudo_link *plink; ++ struct inode *inode; ++ int i, do_put; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ AuDebugOn(!au_opt_test(au_mntflags(sb), PLINK)); ++ AuDebugOn(au_plink_maint(sb, AuLock_NOPLM)); ++ ++ /* no spin_lock since sbinfo is write-locked */ ++ for (i = 0; i < AuPlink_NHASH; i++) { ++ plink_hlist = &sbinfo->si_plink[i].head; ++ hlist_for_each_entry_safe(plink, tmp, plink_hlist, hlist) { ++ inode = au_igrab(plink->inode); ++ ii_write_lock_child(inode); ++ do_put = au_plink_do_half_refresh(inode, br_id); ++ if (do_put) ++ do_put_plink(plink, 1); ++ ii_write_unlock(inode); ++ iput(inode); ++ } ++ } ++} +diff -Nur linux-4.1.10.orig/fs/aufs/poll.c linux-4.1.10/fs/aufs/poll.c +--- linux-4.1.10.orig/fs/aufs/poll.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/poll.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,52 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * poll operation ++ * There is only one filesystem which implements ->poll operation, currently. ++ */ ++ ++#include "aufs.h" ++ ++unsigned int aufs_poll(struct file *file, poll_table *wait) ++{ ++ unsigned int mask; ++ int err; ++ struct file *h_file; ++ struct super_block *sb; ++ ++ /* We should pretend an error happened. */ ++ mask = POLLERR /* | POLLIN | POLLOUT */; ++ sb = file->f_path.dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLMW); ++ ++ h_file = au_read_pre(file, /*keep_fi*/0); ++ err = PTR_ERR(h_file); ++ if (IS_ERR(h_file)) ++ goto out; ++ ++ /* it is not an error if h_file has no operation */ ++ mask = DEFAULT_POLLMASK; ++ if (h_file->f_op->poll) ++ mask = h_file->f_op->poll(h_file, wait); ++ fput(h_file); /* instead of au_read_post() */ ++ ++out: ++ si_read_unlock(sb); ++ AuTraceErr((int)mask); ++ return mask; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/posix_acl.c linux-4.1.10/fs/aufs/posix_acl.c +--- linux-4.1.10.orig/fs/aufs/posix_acl.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/posix_acl.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,99 @@ ++/* ++ * Copyright (C) 2014-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * posix acl operations ++ */ ++ ++#include <linux/fs.h> ++#include <linux/posix_acl.h> ++#include "aufs.h" ++ ++struct posix_acl *aufs_get_acl(struct inode *inode, int type) ++{ ++ struct posix_acl *acl; ++ int err; ++ aufs_bindex_t bindex; ++ struct inode *h_inode; ++ struct super_block *sb; ++ ++ acl = NULL; ++ sb = inode->i_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ ii_read_lock_child(inode); ++ if (!(sb->s_flags & MS_POSIXACL)) ++ goto out; ++ ++ bindex = au_ibstart(inode); ++ h_inode = au_h_iptr(inode, bindex); ++ if (unlikely(!h_inode ++ || ((h_inode->i_mode & S_IFMT) ++ != (inode->i_mode & S_IFMT)))) { ++ err = au_busy_or_stale(); ++ acl = ERR_PTR(err); ++ goto out; ++ } ++ ++ /* always topmost only */ ++ acl = get_acl(h_inode, type); ++ ++out: ++ ii_read_unlock(inode); ++ si_read_unlock(sb); ++ ++ AuTraceErrPtr(acl); ++ return acl; ++} ++ ++int aufs_set_acl(struct inode *inode, struct posix_acl *acl, int type) ++{ ++ int err; ++ ssize_t ssz; ++ struct dentry *dentry; ++ struct au_srxattr arg = { ++ .type = AU_ACL_SET, ++ .u.acl_set = { ++ .acl = acl, ++ .type = type ++ }, ++ }; ++ ++ mutex_lock(&inode->i_mutex); ++ if (inode->i_ino == AUFS_ROOT_INO) ++ dentry = dget(inode->i_sb->s_root); ++ else { ++ dentry = d_find_alias(inode); ++ if (!dentry) ++ dentry = d_find_any_alias(inode); ++ if (!dentry) { ++ pr_warn("cannot handle this inode, " ++ "please report to aufs-users ML\n"); ++ err = -ENOENT; ++ goto out; ++ } ++ } ++ ++ ssz = au_srxattr(dentry, &arg); ++ dput(dentry); ++ err = ssz; ++ if (ssz >= 0) ++ err = 0; ++ ++out: ++ mutex_unlock(&inode->i_mutex); ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/procfs.c linux-4.1.10/fs/aufs/procfs.c +--- linux-4.1.10.orig/fs/aufs/procfs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/procfs.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,169 @@ ++/* ++ * Copyright (C) 2010-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * procfs interfaces ++ */ ++ ++#include <linux/proc_fs.h> ++#include "aufs.h" ++ ++static int au_procfs_plm_release(struct inode *inode, struct file *file) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ sbinfo = file->private_data; ++ if (sbinfo) { ++ au_plink_maint_leave(sbinfo); ++ kobject_put(&sbinfo->si_kobj); ++ } ++ ++ return 0; ++} ++ ++static void au_procfs_plm_write_clean(struct file *file) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ sbinfo = file->private_data; ++ if (sbinfo) ++ au_plink_clean(sbinfo->si_sb, /*verbose*/0); ++} ++ ++static int au_procfs_plm_write_si(struct file *file, unsigned long id) ++{ ++ int err; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ ++ err = -EBUSY; ++ if (unlikely(file->private_data)) ++ goto out; ++ ++ sb = NULL; ++ /* don't use au_sbilist_lock() here */ ++ spin_lock(&au_sbilist.spin); ++ list_for_each_entry(sbinfo, &au_sbilist.head, si_list) ++ if (id == sysaufs_si_id(sbinfo)) { ++ kobject_get(&sbinfo->si_kobj); ++ sb = sbinfo->si_sb; ++ break; ++ } ++ spin_unlock(&au_sbilist.spin); ++ ++ err = -EINVAL; ++ if (unlikely(!sb)) ++ goto out; ++ ++ err = au_plink_maint_enter(sb); ++ if (!err) ++ /* keep kobject_get() */ ++ file->private_data = sbinfo; ++ else ++ kobject_put(&sbinfo->si_kobj); ++out: ++ return err; ++} ++ ++/* ++ * Accept a valid "si=xxxx" only. ++ * Once it is accepted successfully, accept "clean" too. ++ */ ++static ssize_t au_procfs_plm_write(struct file *file, const char __user *ubuf, ++ size_t count, loff_t *ppos) ++{ ++ ssize_t err; ++ unsigned long id; ++ /* last newline is allowed */ ++ char buf[3 + sizeof(unsigned long) * 2 + 1]; ++ ++ err = -EACCES; ++ if (unlikely(!capable(CAP_SYS_ADMIN))) ++ goto out; ++ ++ err = -EINVAL; ++ if (unlikely(count > sizeof(buf))) ++ goto out; ++ ++ err = copy_from_user(buf, ubuf, count); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ goto out; ++ } ++ buf[count] = 0; ++ ++ err = -EINVAL; ++ if (!strcmp("clean", buf)) { ++ au_procfs_plm_write_clean(file); ++ goto out_success; ++ } else if (unlikely(strncmp("si=", buf, 3))) ++ goto out; ++ ++ err = kstrtoul(buf + 3, 16, &id); ++ if (unlikely(err)) ++ goto out; ++ ++ err = au_procfs_plm_write_si(file, id); ++ if (unlikely(err)) ++ goto out; ++ ++out_success: ++ err = count; /* success */ ++out: ++ return err; ++} ++ ++static const struct file_operations au_procfs_plm_fop = { ++ .write = au_procfs_plm_write, ++ .release = au_procfs_plm_release, ++ .owner = THIS_MODULE ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct proc_dir_entry *au_procfs_dir; ++ ++void au_procfs_fin(void) ++{ ++ remove_proc_entry(AUFS_PLINK_MAINT_NAME, au_procfs_dir); ++ remove_proc_entry(AUFS_PLINK_MAINT_DIR, NULL); ++} ++ ++int __init au_procfs_init(void) ++{ ++ int err; ++ struct proc_dir_entry *entry; ++ ++ err = -ENOMEM; ++ au_procfs_dir = proc_mkdir(AUFS_PLINK_MAINT_DIR, NULL); ++ if (unlikely(!au_procfs_dir)) ++ goto out; ++ ++ entry = proc_create(AUFS_PLINK_MAINT_NAME, S_IFREG | S_IWUSR, ++ au_procfs_dir, &au_procfs_plm_fop); ++ if (unlikely(!entry)) ++ goto out_dir; ++ ++ err = 0; ++ goto out; /* success */ ++ ++ ++out_dir: ++ remove_proc_entry(AUFS_PLINK_MAINT_DIR, NULL); ++out: ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/rdu.c linux-4.1.10/fs/aufs/rdu.c +--- linux-4.1.10.orig/fs/aufs/rdu.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/rdu.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,388 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * readdir in userspace. ++ */ ++ ++#include <linux/compat.h> ++#include <linux/fs_stack.h> ++#include <linux/security.h> ++#include "aufs.h" ++ ++/* bits for struct aufs_rdu.flags */ ++#define AuRdu_CALLED 1 ++#define AuRdu_CONT (1 << 1) ++#define AuRdu_FULL (1 << 2) ++#define au_ftest_rdu(flags, name) ((flags) & AuRdu_##name) ++#define au_fset_rdu(flags, name) \ ++ do { (flags) |= AuRdu_##name; } while (0) ++#define au_fclr_rdu(flags, name) \ ++ do { (flags) &= ~AuRdu_##name; } while (0) ++ ++struct au_rdu_arg { ++ struct dir_context ctx; ++ struct aufs_rdu *rdu; ++ union au_rdu_ent_ul ent; ++ unsigned long end; ++ ++ struct super_block *sb; ++ int err; ++}; ++ ++static int au_rdu_fill(struct dir_context *ctx, const char *name, int nlen, ++ loff_t offset, u64 h_ino, unsigned int d_type) ++{ ++ int err, len; ++ struct au_rdu_arg *arg = container_of(ctx, struct au_rdu_arg, ctx); ++ struct aufs_rdu *rdu = arg->rdu; ++ struct au_rdu_ent ent; ++ ++ err = 0; ++ arg->err = 0; ++ au_fset_rdu(rdu->cookie.flags, CALLED); ++ len = au_rdu_len(nlen); ++ if (arg->ent.ul + len < arg->end) { ++ ent.ino = h_ino; ++ ent.bindex = rdu->cookie.bindex; ++ ent.type = d_type; ++ ent.nlen = nlen; ++ if (unlikely(nlen > AUFS_MAX_NAMELEN)) ++ ent.type = DT_UNKNOWN; ++ ++ /* unnecessary to support mmap_sem since this is a dir */ ++ err = -EFAULT; ++ if (copy_to_user(arg->ent.e, &ent, sizeof(ent))) ++ goto out; ++ if (copy_to_user(arg->ent.e->name, name, nlen)) ++ goto out; ++ /* the terminating NULL */ ++ if (__put_user(0, arg->ent.e->name + nlen)) ++ goto out; ++ err = 0; ++ /* AuDbg("%p, %.*s\n", arg->ent.p, nlen, name); */ ++ arg->ent.ul += len; ++ rdu->rent++; ++ } else { ++ err = -EFAULT; ++ au_fset_rdu(rdu->cookie.flags, FULL); ++ rdu->full = 1; ++ rdu->tail = arg->ent; ++ } ++ ++out: ++ /* AuTraceErr(err); */ ++ return err; ++} ++ ++static int au_rdu_do(struct file *h_file, struct au_rdu_arg *arg) ++{ ++ int err; ++ loff_t offset; ++ struct au_rdu_cookie *cookie = &arg->rdu->cookie; ++ ++ /* we don't have to care (FMODE_32BITHASH | FMODE_64BITHASH) for ext4 */ ++ offset = vfsub_llseek(h_file, cookie->h_pos, SEEK_SET); ++ err = offset; ++ if (unlikely(offset != cookie->h_pos)) ++ goto out; ++ ++ err = 0; ++ do { ++ arg->err = 0; ++ au_fclr_rdu(cookie->flags, CALLED); ++ /* smp_mb(); */ ++ err = vfsub_iterate_dir(h_file, &arg->ctx); ++ if (err >= 0) ++ err = arg->err; ++ } while (!err ++ && au_ftest_rdu(cookie->flags, CALLED) ++ && !au_ftest_rdu(cookie->flags, FULL)); ++ cookie->h_pos = h_file->f_pos; ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_rdu(struct file *file, struct aufs_rdu *rdu) ++{ ++ int err; ++ aufs_bindex_t bend; ++ struct au_rdu_arg arg = { ++ .ctx = { ++ .actor = au_rdu_fill ++ } ++ }; ++ struct dentry *dentry; ++ struct inode *inode; ++ struct file *h_file; ++ struct au_rdu_cookie *cookie = &rdu->cookie; ++ ++ err = !access_ok(VERIFY_WRITE, rdu->ent.e, rdu->sz); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ goto out; ++ } ++ rdu->rent = 0; ++ rdu->tail = rdu->ent; ++ rdu->full = 0; ++ arg.rdu = rdu; ++ arg.ent = rdu->ent; ++ arg.end = arg.ent.ul; ++ arg.end += rdu->sz; ++ ++ err = -ENOTDIR; ++ if (unlikely(!file->f_op->iterate)) ++ goto out; ++ ++ err = security_file_permission(file, MAY_READ); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out; ++ ++ dentry = file->f_path.dentry; ++ inode = d_inode(dentry); ++#if 1 ++ mutex_lock(&inode->i_mutex); ++#else ++ err = mutex_lock_killable(&inode->i_mutex); ++ AuTraceErr(err); ++ if (unlikely(err)) ++ goto out; ++#endif ++ ++ arg.sb = inode->i_sb; ++ err = si_read_lock(arg.sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out_mtx; ++ err = au_alive_dir(dentry); ++ if (unlikely(err)) ++ goto out_si; ++ /* todo: reval? */ ++ fi_read_lock(file); ++ ++ err = -EAGAIN; ++ if (unlikely(au_ftest_rdu(cookie->flags, CONT) ++ && cookie->generation != au_figen(file))) ++ goto out_unlock; ++ ++ err = 0; ++ if (!rdu->blk) { ++ rdu->blk = au_sbi(arg.sb)->si_rdblk; ++ if (!rdu->blk) ++ rdu->blk = au_dir_size(file, /*dentry*/NULL); ++ } ++ bend = au_fbstart(file); ++ if (cookie->bindex < bend) ++ cookie->bindex = bend; ++ bend = au_fbend_dir(file); ++ /* AuDbg("b%d, b%d\n", cookie->bindex, bend); */ ++ for (; !err && cookie->bindex <= bend; ++ cookie->bindex++, cookie->h_pos = 0) { ++ h_file = au_hf_dir(file, cookie->bindex); ++ if (!h_file) ++ continue; ++ ++ au_fclr_rdu(cookie->flags, FULL); ++ err = au_rdu_do(h_file, &arg); ++ AuTraceErr(err); ++ if (unlikely(au_ftest_rdu(cookie->flags, FULL) || err)) ++ break; ++ } ++ AuDbg("rent %llu\n", rdu->rent); ++ ++ if (!err && !au_ftest_rdu(cookie->flags, CONT)) { ++ rdu->shwh = !!au_opt_test(au_sbi(arg.sb)->si_mntflags, SHWH); ++ au_fset_rdu(cookie->flags, CONT); ++ cookie->generation = au_figen(file); ++ } ++ ++ ii_read_lock_child(inode); ++ fsstack_copy_attr_atime(inode, au_h_iptr(inode, au_ibstart(inode))); ++ ii_read_unlock(inode); ++ ++out_unlock: ++ fi_read_unlock(file); ++out_si: ++ si_read_unlock(arg.sb); ++out_mtx: ++ mutex_unlock(&inode->i_mutex); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_rdu_ino(struct file *file, struct aufs_rdu *rdu) ++{ ++ int err; ++ ino_t ino; ++ unsigned long long nent; ++ union au_rdu_ent_ul *u; ++ struct au_rdu_ent ent; ++ struct super_block *sb; ++ ++ err = 0; ++ nent = rdu->nent; ++ u = &rdu->ent; ++ sb = file->f_path.dentry->d_sb; ++ si_read_lock(sb, AuLock_FLUSH); ++ while (nent-- > 0) { ++ /* unnecessary to support mmap_sem since this is a dir */ ++ err = copy_from_user(&ent, u->e, sizeof(ent)); ++ if (!err) ++ err = !access_ok(VERIFY_WRITE, &u->e->ino, sizeof(ino)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ break; ++ } ++ ++ /* AuDbg("b%d, i%llu\n", ent.bindex, ent.ino); */ ++ if (!ent.wh) ++ err = au_ino(sb, ent.bindex, ent.ino, ent.type, &ino); ++ else ++ err = au_wh_ino(sb, ent.bindex, ent.ino, ent.type, ++ &ino); ++ if (unlikely(err)) { ++ AuTraceErr(err); ++ break; ++ } ++ ++ err = __put_user(ino, &u->e->ino); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ break; ++ } ++ u->ul += au_rdu_len(ent.nlen); ++ } ++ si_read_unlock(sb); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_rdu_verify(struct aufs_rdu *rdu) ++{ ++ AuDbg("rdu{%llu, %p, %u | %u | %llu, %u, %u | " ++ "%llu, b%d, 0x%x, g%u}\n", ++ rdu->sz, rdu->ent.e, rdu->verify[AufsCtlRduV_SZ], ++ rdu->blk, ++ rdu->rent, rdu->shwh, rdu->full, ++ rdu->cookie.h_pos, rdu->cookie.bindex, rdu->cookie.flags, ++ rdu->cookie.generation); ++ ++ if (rdu->verify[AufsCtlRduV_SZ] == sizeof(*rdu)) ++ return 0; ++ ++ AuDbg("%u:%u\n", ++ rdu->verify[AufsCtlRduV_SZ], (unsigned int)sizeof(*rdu)); ++ return -EINVAL; ++} ++ ++long au_rdu_ioctl(struct file *file, unsigned int cmd, unsigned long arg) ++{ ++ long err, e; ++ struct aufs_rdu rdu; ++ void __user *p = (void __user *)arg; ++ ++ err = copy_from_user(&rdu, p, sizeof(rdu)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ goto out; ++ } ++ err = au_rdu_verify(&rdu); ++ if (unlikely(err)) ++ goto out; ++ ++ switch (cmd) { ++ case AUFS_CTL_RDU: ++ err = au_rdu(file, &rdu); ++ if (unlikely(err)) ++ break; ++ ++ e = copy_to_user(p, &rdu, sizeof(rdu)); ++ if (unlikely(e)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ } ++ break; ++ case AUFS_CTL_RDU_INO: ++ err = au_rdu_ino(file, &rdu); ++ break; ++ ++ default: ++ /* err = -ENOTTY; */ ++ err = -EINVAL; ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++#ifdef CONFIG_COMPAT ++long au_rdu_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) ++{ ++ long err, e; ++ struct aufs_rdu rdu; ++ void __user *p = compat_ptr(arg); ++ ++ /* todo: get_user()? */ ++ err = copy_from_user(&rdu, p, sizeof(rdu)); ++ if (unlikely(err)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ goto out; ++ } ++ rdu.ent.e = compat_ptr(rdu.ent.ul); ++ err = au_rdu_verify(&rdu); ++ if (unlikely(err)) ++ goto out; ++ ++ switch (cmd) { ++ case AUFS_CTL_RDU: ++ err = au_rdu(file, &rdu); ++ if (unlikely(err)) ++ break; ++ ++ rdu.ent.ul = ptr_to_compat(rdu.ent.e); ++ rdu.tail.ul = ptr_to_compat(rdu.tail.e); ++ e = copy_to_user(p, &rdu, sizeof(rdu)); ++ if (unlikely(e)) { ++ err = -EFAULT; ++ AuTraceErr(err); ++ } ++ break; ++ case AUFS_CTL_RDU_INO: ++ err = au_rdu_ino(file, &rdu); ++ break; ++ ++ default: ++ /* err = -ENOTTY; */ ++ err = -EINVAL; ++ } ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++#endif +diff -Nur linux-4.1.10.orig/fs/aufs/rwsem.h linux-4.1.10/fs/aufs/rwsem.h +--- linux-4.1.10.orig/fs/aufs/rwsem.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/rwsem.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,191 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * simple read-write semaphore wrappers ++ */ ++ ++#ifndef __AUFS_RWSEM_H__ ++#define __AUFS_RWSEM_H__ ++ ++#ifdef __KERNEL__ ++ ++#include "debug.h" ++ ++struct au_rwsem { ++ struct rw_semaphore rwsem; ++#ifdef CONFIG_AUFS_DEBUG ++ /* just for debugging, not almighty counter */ ++ atomic_t rcnt, wcnt; ++#endif ++}; ++ ++#ifdef CONFIG_AUFS_DEBUG ++#define AuDbgCntInit(rw) do { \ ++ atomic_set(&(rw)->rcnt, 0); \ ++ atomic_set(&(rw)->wcnt, 0); \ ++ smp_mb(); /* atomic set */ \ ++} while (0) ++ ++#define AuDbgRcntInc(rw) atomic_inc(&(rw)->rcnt) ++#define AuDbgRcntDec(rw) WARN_ON(atomic_dec_return(&(rw)->rcnt) < 0) ++#define AuDbgWcntInc(rw) atomic_inc(&(rw)->wcnt) ++#define AuDbgWcntDec(rw) WARN_ON(atomic_dec_return(&(rw)->wcnt) < 0) ++#else ++#define AuDbgCntInit(rw) do {} while (0) ++#define AuDbgRcntInc(rw) do {} while (0) ++#define AuDbgRcntDec(rw) do {} while (0) ++#define AuDbgWcntInc(rw) do {} while (0) ++#define AuDbgWcntDec(rw) do {} while (0) ++#endif /* CONFIG_AUFS_DEBUG */ ++ ++/* to debug easier, do not make them inlined functions */ ++#define AuRwMustNoWaiters(rw) AuDebugOn(!list_empty(&(rw)->rwsem.wait_list)) ++/* rwsem_is_locked() is unusable */ ++#define AuRwMustReadLock(rw) AuDebugOn(atomic_read(&(rw)->rcnt) <= 0) ++#define AuRwMustWriteLock(rw) AuDebugOn(atomic_read(&(rw)->wcnt) <= 0) ++#define AuRwMustAnyLock(rw) AuDebugOn(atomic_read(&(rw)->rcnt) <= 0 \ ++ && atomic_read(&(rw)->wcnt) <= 0) ++#define AuRwDestroy(rw) AuDebugOn(atomic_read(&(rw)->rcnt) \ ++ || atomic_read(&(rw)->wcnt)) ++ ++#define au_rw_class(rw, key) lockdep_set_class(&(rw)->rwsem, key) ++ ++static inline void au_rw_init(struct au_rwsem *rw) ++{ ++ AuDbgCntInit(rw); ++ init_rwsem(&rw->rwsem); ++} ++ ++static inline void au_rw_init_wlock(struct au_rwsem *rw) ++{ ++ au_rw_init(rw); ++ down_write(&rw->rwsem); ++ AuDbgWcntInc(rw); ++} ++ ++static inline void au_rw_init_wlock_nested(struct au_rwsem *rw, ++ unsigned int lsc) ++{ ++ au_rw_init(rw); ++ down_write_nested(&rw->rwsem, lsc); ++ AuDbgWcntInc(rw); ++} ++ ++static inline void au_rw_read_lock(struct au_rwsem *rw) ++{ ++ down_read(&rw->rwsem); ++ AuDbgRcntInc(rw); ++} ++ ++static inline void au_rw_read_lock_nested(struct au_rwsem *rw, unsigned int lsc) ++{ ++ down_read_nested(&rw->rwsem, lsc); ++ AuDbgRcntInc(rw); ++} ++ ++static inline void au_rw_read_unlock(struct au_rwsem *rw) ++{ ++ AuRwMustReadLock(rw); ++ AuDbgRcntDec(rw); ++ up_read(&rw->rwsem); ++} ++ ++static inline void au_rw_dgrade_lock(struct au_rwsem *rw) ++{ ++ AuRwMustWriteLock(rw); ++ AuDbgRcntInc(rw); ++ AuDbgWcntDec(rw); ++ downgrade_write(&rw->rwsem); ++} ++ ++static inline void au_rw_write_lock(struct au_rwsem *rw) ++{ ++ down_write(&rw->rwsem); ++ AuDbgWcntInc(rw); ++} ++ ++static inline void au_rw_write_lock_nested(struct au_rwsem *rw, ++ unsigned int lsc) ++{ ++ down_write_nested(&rw->rwsem, lsc); ++ AuDbgWcntInc(rw); ++} ++ ++static inline void au_rw_write_unlock(struct au_rwsem *rw) ++{ ++ AuRwMustWriteLock(rw); ++ AuDbgWcntDec(rw); ++ up_write(&rw->rwsem); ++} ++ ++/* why is not _nested version defined */ ++static inline int au_rw_read_trylock(struct au_rwsem *rw) ++{ ++ int ret; ++ ++ ret = down_read_trylock(&rw->rwsem); ++ if (ret) ++ AuDbgRcntInc(rw); ++ return ret; ++} ++ ++static inline int au_rw_write_trylock(struct au_rwsem *rw) ++{ ++ int ret; ++ ++ ret = down_write_trylock(&rw->rwsem); ++ if (ret) ++ AuDbgWcntInc(rw); ++ return ret; ++} ++ ++#undef AuDbgCntInit ++#undef AuDbgRcntInc ++#undef AuDbgRcntDec ++#undef AuDbgWcntInc ++#undef AuDbgWcntDec ++ ++#define AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ ++static inline void prefix##_read_lock(param) \ ++{ au_rw_read_lock(rwsem); } \ ++static inline void prefix##_write_lock(param) \ ++{ au_rw_write_lock(rwsem); } \ ++static inline int prefix##_read_trylock(param) \ ++{ return au_rw_read_trylock(rwsem); } \ ++static inline int prefix##_write_trylock(param) \ ++{ return au_rw_write_trylock(rwsem); } ++/* why is not _nested version defined */ ++/* static inline void prefix##_read_trylock_nested(param, lsc) ++{ au_rw_read_trylock_nested(rwsem, lsc)); } ++static inline void prefix##_write_trylock_nestd(param, lsc) ++{ au_rw_write_trylock_nested(rwsem, lsc); } */ ++ ++#define AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) \ ++static inline void prefix##_read_unlock(param) \ ++{ au_rw_read_unlock(rwsem); } \ ++static inline void prefix##_write_unlock(param) \ ++{ au_rw_write_unlock(rwsem); } \ ++static inline void prefix##_downgrade_lock(param) \ ++{ au_rw_dgrade_lock(rwsem); } ++ ++#define AuSimpleRwsemFuncs(prefix, param, rwsem) \ ++ AuSimpleLockRwsemFuncs(prefix, param, rwsem) \ ++ AuSimpleUnlockRwsemFuncs(prefix, param, rwsem) ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_RWSEM_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/sbinfo.c linux-4.1.10/fs/aufs/sbinfo.c +--- linux-4.1.10.orig/fs/aufs/sbinfo.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/sbinfo.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,356 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * superblock private data ++ */ ++ ++#include "aufs.h" ++ ++/* ++ * they are necessary regardless sysfs is disabled. ++ */ ++void au_si_free(struct kobject *kobj) ++{ ++ int i; ++ struct au_sbinfo *sbinfo; ++ char *locked __maybe_unused; /* debug only */ ++ ++ sbinfo = container_of(kobj, struct au_sbinfo, si_kobj); ++ for (i = 0; i < AuPlink_NHASH; i++) ++ AuDebugOn(!hlist_empty(&sbinfo->si_plink[i].head)); ++ AuDebugOn(atomic_read(&sbinfo->si_nowait.nw_len)); ++ ++ au_rw_write_lock(&sbinfo->si_rwsem); ++ au_br_free(sbinfo); ++ au_rw_write_unlock(&sbinfo->si_rwsem); ++ ++ AuDebugOn(radix_tree_gang_lookup ++ (&sbinfo->au_si_pid.tree, (void **)&locked, ++ /*first_index*/PID_MAX_DEFAULT - 1, ++ /*max_items*/sizeof(locked)/sizeof(*locked))); ++ ++ kfree(sbinfo->si_branch); ++ kfree(sbinfo->au_si_pid.bitmap); ++ mutex_destroy(&sbinfo->si_xib_mtx); ++ AuRwDestroy(&sbinfo->si_rwsem); ++ ++ kfree(sbinfo); ++} ++ ++int au_si_alloc(struct super_block *sb) ++{ ++ int err, i; ++ struct au_sbinfo *sbinfo; ++ static struct lock_class_key aufs_si; ++ ++ err = -ENOMEM; ++ sbinfo = kzalloc(sizeof(*sbinfo), GFP_NOFS); ++ if (unlikely(!sbinfo)) ++ goto out; ++ ++ BUILD_BUG_ON(sizeof(unsigned long) != ++ sizeof(*sbinfo->au_si_pid.bitmap)); ++ sbinfo->au_si_pid.bitmap = kcalloc(BITS_TO_LONGS(PID_MAX_DEFAULT), ++ sizeof(*sbinfo->au_si_pid.bitmap), ++ GFP_NOFS); ++ if (unlikely(!sbinfo->au_si_pid.bitmap)) ++ goto out_sbinfo; ++ ++ /* will be reallocated separately */ ++ sbinfo->si_branch = kzalloc(sizeof(*sbinfo->si_branch), GFP_NOFS); ++ if (unlikely(!sbinfo->si_branch)) ++ goto out_pidmap; ++ ++ err = sysaufs_si_init(sbinfo); ++ if (unlikely(err)) ++ goto out_br; ++ ++ au_nwt_init(&sbinfo->si_nowait); ++ au_rw_init_wlock(&sbinfo->si_rwsem); ++ au_rw_class(&sbinfo->si_rwsem, &aufs_si); ++ spin_lock_init(&sbinfo->au_si_pid.tree_lock); ++ INIT_RADIX_TREE(&sbinfo->au_si_pid.tree, GFP_ATOMIC | __GFP_NOFAIL); ++ ++ atomic_long_set(&sbinfo->si_ninodes, 0); ++ atomic_long_set(&sbinfo->si_nfiles, 0); ++ ++ sbinfo->si_bend = -1; ++ sbinfo->si_last_br_id = AUFS_BRANCH_MAX / 2; ++ ++ sbinfo->si_wbr_copyup = AuWbrCopyup_Def; ++ sbinfo->si_wbr_create = AuWbrCreate_Def; ++ sbinfo->si_wbr_copyup_ops = au_wbr_copyup_ops + sbinfo->si_wbr_copyup; ++ sbinfo->si_wbr_create_ops = au_wbr_create_ops + sbinfo->si_wbr_create; ++ ++ au_fhsm_init(sbinfo); ++ ++ sbinfo->si_mntflags = au_opts_plink(AuOpt_Def); ++ ++ sbinfo->si_xino_jiffy = jiffies; ++ sbinfo->si_xino_expire ++ = msecs_to_jiffies(AUFS_XINO_DEF_SEC * MSEC_PER_SEC); ++ mutex_init(&sbinfo->si_xib_mtx); ++ sbinfo->si_xino_brid = -1; ++ /* leave si_xib_last_pindex and si_xib_next_bit */ ++ ++ au_sphl_init(&sbinfo->si_aopen); ++ ++ sbinfo->si_rdcache = msecs_to_jiffies(AUFS_RDCACHE_DEF * MSEC_PER_SEC); ++ sbinfo->si_rdblk = AUFS_RDBLK_DEF; ++ sbinfo->si_rdhash = AUFS_RDHASH_DEF; ++ sbinfo->si_dirwh = AUFS_DIRWH_DEF; ++ ++ for (i = 0; i < AuPlink_NHASH; i++) ++ au_sphl_init(sbinfo->si_plink + i); ++ init_waitqueue_head(&sbinfo->si_plink_wq); ++ spin_lock_init(&sbinfo->si_plink_maint_lock); ++ ++ au_sphl_init(&sbinfo->si_files); ++ ++ /* leave other members for sysaufs and si_mnt. */ ++ sbinfo->si_sb = sb; ++ sb->s_fs_info = sbinfo; ++ si_pid_set(sb); ++ return 0; /* success */ ++ ++out_br: ++ kfree(sbinfo->si_branch); ++out_pidmap: ++ kfree(sbinfo->au_si_pid.bitmap); ++out_sbinfo: ++ kfree(sbinfo); ++out: ++ return err; ++} ++ ++int au_sbr_realloc(struct au_sbinfo *sbinfo, int nbr) ++{ ++ int err, sz; ++ struct au_branch **brp; ++ ++ AuRwMustWriteLock(&sbinfo->si_rwsem); ++ ++ err = -ENOMEM; ++ sz = sizeof(*brp) * (sbinfo->si_bend + 1); ++ if (unlikely(!sz)) ++ sz = sizeof(*brp); ++ brp = au_kzrealloc(sbinfo->si_branch, sz, sizeof(*brp) * nbr, GFP_NOFS); ++ if (brp) { ++ sbinfo->si_branch = brp; ++ err = 0; ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++unsigned int au_sigen_inc(struct super_block *sb) ++{ ++ unsigned int gen; ++ struct inode *inode; ++ ++ SiMustWriteLock(sb); ++ ++ gen = ++au_sbi(sb)->si_generation; ++ au_update_digen(sb->s_root); ++ inode = d_inode(sb->s_root); ++ au_update_iigen(inode, /*half*/0); ++ inode->i_version++; ++ return gen; ++} ++ ++aufs_bindex_t au_new_br_id(struct super_block *sb) ++{ ++ aufs_bindex_t br_id; ++ int i; ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ for (i = 0; i <= AUFS_BRANCH_MAX; i++) { ++ br_id = ++sbinfo->si_last_br_id; ++ AuDebugOn(br_id < 0); ++ if (br_id && au_br_index(sb, br_id) < 0) ++ return br_id; ++ } ++ ++ return -1; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* it is ok that new 'nwt' tasks are appended while we are sleeping */ ++int si_read_lock(struct super_block *sb, int flags) ++{ ++ int err; ++ ++ err = 0; ++ if (au_ftest_lock(flags, FLUSH)) ++ au_nwt_flush(&au_sbi(sb)->si_nowait); ++ ++ si_noflush_read_lock(sb); ++ err = au_plink_maint(sb, flags); ++ if (unlikely(err)) ++ si_read_unlock(sb); ++ ++ return err; ++} ++ ++int si_write_lock(struct super_block *sb, int flags) ++{ ++ int err; ++ ++ if (au_ftest_lock(flags, FLUSH)) ++ au_nwt_flush(&au_sbi(sb)->si_nowait); ++ ++ si_noflush_write_lock(sb); ++ err = au_plink_maint(sb, flags); ++ if (unlikely(err)) ++ si_write_unlock(sb); ++ ++ return err; ++} ++ ++/* dentry and super_block lock. call at entry point */ ++int aufs_read_lock(struct dentry *dentry, int flags) ++{ ++ int err; ++ struct super_block *sb; ++ ++ sb = dentry->d_sb; ++ err = si_read_lock(sb, flags); ++ if (unlikely(err)) ++ goto out; ++ ++ if (au_ftest_lock(flags, DW)) ++ di_write_lock_child(dentry); ++ else ++ di_read_lock_child(dentry, flags); ++ ++ if (au_ftest_lock(flags, GEN)) { ++ err = au_digen_test(dentry, au_sigen(sb)); ++ AuDebugOn(!err && au_dbrange_test(dentry)); ++ if (unlikely(err)) ++ aufs_read_unlock(dentry, flags); ++ } ++ ++out: ++ return err; ++} ++ ++void aufs_read_unlock(struct dentry *dentry, int flags) ++{ ++ if (au_ftest_lock(flags, DW)) ++ di_write_unlock(dentry); ++ else ++ di_read_unlock(dentry, flags); ++ si_read_unlock(dentry->d_sb); ++} ++ ++void aufs_write_lock(struct dentry *dentry) ++{ ++ si_write_lock(dentry->d_sb, AuLock_FLUSH | AuLock_NOPLMW); ++ di_write_lock_child(dentry); ++} ++ ++void aufs_write_unlock(struct dentry *dentry) ++{ ++ di_write_unlock(dentry); ++ si_write_unlock(dentry->d_sb); ++} ++ ++int aufs_read_and_write_lock2(struct dentry *d1, struct dentry *d2, int flags) ++{ ++ int err; ++ unsigned int sigen; ++ struct super_block *sb; ++ ++ sb = d1->d_sb; ++ err = si_read_lock(sb, flags); ++ if (unlikely(err)) ++ goto out; ++ ++ di_write_lock2_child(d1, d2, au_ftest_lock(flags, DIR)); ++ ++ if (au_ftest_lock(flags, GEN)) { ++ sigen = au_sigen(sb); ++ err = au_digen_test(d1, sigen); ++ AuDebugOn(!err && au_dbrange_test(d1)); ++ if (!err) { ++ err = au_digen_test(d2, sigen); ++ AuDebugOn(!err && au_dbrange_test(d2)); ++ } ++ if (unlikely(err)) ++ aufs_read_and_write_unlock2(d1, d2); ++ } ++ ++out: ++ return err; ++} ++ ++void aufs_read_and_write_unlock2(struct dentry *d1, struct dentry *d2) ++{ ++ di_write_unlock2(d1, d2); ++ si_read_unlock(d1->d_sb); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int si_pid_test_slow(struct super_block *sb) ++{ ++ void *p; ++ ++ rcu_read_lock(); ++ p = radix_tree_lookup(&au_sbi(sb)->au_si_pid.tree, current->pid); ++ rcu_read_unlock(); ++ ++ return (long)!!p; ++} ++ ++void si_pid_set_slow(struct super_block *sb) ++{ ++ int err; ++ struct au_sbinfo *sbinfo; ++ ++ AuDebugOn(si_pid_test_slow(sb)); ++ ++ sbinfo = au_sbi(sb); ++ err = radix_tree_preload(GFP_NOFS | __GFP_NOFAIL); ++ AuDebugOn(err); ++ spin_lock(&sbinfo->au_si_pid.tree_lock); ++ err = radix_tree_insert(&sbinfo->au_si_pid.tree, current->pid, ++ /*any valid ptr*/sb); ++ spin_unlock(&sbinfo->au_si_pid.tree_lock); ++ AuDebugOn(err); ++ radix_tree_preload_end(); ++} ++ ++void si_pid_clr_slow(struct super_block *sb) ++{ ++ void *p; ++ struct au_sbinfo *sbinfo; ++ ++ AuDebugOn(!si_pid_test_slow(sb)); ++ ++ sbinfo = au_sbi(sb); ++ spin_lock(&sbinfo->au_si_pid.tree_lock); ++ p = radix_tree_delete(&sbinfo->au_si_pid.tree, current->pid); ++ spin_unlock(&sbinfo->au_si_pid.tree_lock); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/spl.h linux-4.1.10/fs/aufs/spl.h +--- linux-4.1.10.orig/fs/aufs/spl.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/spl.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,111 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * simple list protected by a spinlock ++ */ ++ ++#ifndef __AUFS_SPL_H__ ++#define __AUFS_SPL_H__ ++ ++#ifdef __KERNEL__ ++ ++struct au_splhead { ++ spinlock_t spin; ++ struct list_head head; ++}; ++ ++static inline void au_spl_init(struct au_splhead *spl) ++{ ++ spin_lock_init(&spl->spin); ++ INIT_LIST_HEAD(&spl->head); ++} ++ ++static inline void au_spl_add(struct list_head *list, struct au_splhead *spl) ++{ ++ spin_lock(&spl->spin); ++ list_add(list, &spl->head); ++ spin_unlock(&spl->spin); ++} ++ ++static inline void au_spl_del(struct list_head *list, struct au_splhead *spl) ++{ ++ spin_lock(&spl->spin); ++ list_del(list); ++ spin_unlock(&spl->spin); ++} ++ ++static inline void au_spl_del_rcu(struct list_head *list, ++ struct au_splhead *spl) ++{ ++ spin_lock(&spl->spin); ++ list_del_rcu(list); ++ spin_unlock(&spl->spin); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_sphlhead { ++ spinlock_t spin; ++ struct hlist_head head; ++}; ++ ++static inline void au_sphl_init(struct au_sphlhead *sphl) ++{ ++ spin_lock_init(&sphl->spin); ++ INIT_HLIST_HEAD(&sphl->head); ++} ++ ++static inline void au_sphl_add(struct hlist_node *hlist, ++ struct au_sphlhead *sphl) ++{ ++ spin_lock(&sphl->spin); ++ hlist_add_head(hlist, &sphl->head); ++ spin_unlock(&sphl->spin); ++} ++ ++static inline void au_sphl_del(struct hlist_node *hlist, ++ struct au_sphlhead *sphl) ++{ ++ spin_lock(&sphl->spin); ++ hlist_del(hlist); ++ spin_unlock(&sphl->spin); ++} ++ ++static inline void au_sphl_del_rcu(struct hlist_node *hlist, ++ struct au_sphlhead *sphl) ++{ ++ spin_lock(&sphl->spin); ++ hlist_del_rcu(hlist); ++ spin_unlock(&sphl->spin); ++} ++ ++static inline unsigned long au_sphl_count(struct au_sphlhead *sphl) ++{ ++ unsigned long cnt; ++ struct hlist_node *pos; ++ ++ cnt = 0; ++ spin_lock(&sphl->spin); ++ hlist_for_each(pos, &sphl->head) ++ cnt++; ++ spin_unlock(&sphl->spin); ++ return cnt; ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_SPL_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/super.c linux-4.1.10/fs/aufs/super.c +--- linux-4.1.10.orig/fs/aufs/super.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/super.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1004 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * mount and super_block operations ++ */ ++ ++#include <linux/mm.h> ++#include <linux/seq_file.h> ++#include <linux/statfs.h> ++#include <linux/vmalloc.h> ++#include "aufs.h" ++ ++/* ++ * super_operations ++ */ ++static struct inode *aufs_alloc_inode(struct super_block *sb __maybe_unused) ++{ ++ struct au_icntnr *c; ++ ++ c = au_cache_alloc_icntnr(); ++ if (c) { ++ au_icntnr_init(c); ++ c->vfs_inode.i_version = 1; /* sigen(sb); */ ++ c->iinfo.ii_hinode = NULL; ++ return &c->vfs_inode; ++ } ++ return NULL; ++} ++ ++static void aufs_destroy_inode_cb(struct rcu_head *head) ++{ ++ struct inode *inode = container_of(head, struct inode, i_rcu); ++ ++ INIT_HLIST_HEAD(&inode->i_dentry); ++ au_cache_free_icntnr(container_of(inode, struct au_icntnr, vfs_inode)); ++} ++ ++static void aufs_destroy_inode(struct inode *inode) ++{ ++ au_iinfo_fin(inode); ++ call_rcu(&inode->i_rcu, aufs_destroy_inode_cb); ++} ++ ++struct inode *au_iget_locked(struct super_block *sb, ino_t ino) ++{ ++ struct inode *inode; ++ int err; ++ ++ inode = iget_locked(sb, ino); ++ if (unlikely(!inode)) { ++ inode = ERR_PTR(-ENOMEM); ++ goto out; ++ } ++ if (!(inode->i_state & I_NEW)) ++ goto out; ++ ++ err = au_xigen_new(inode); ++ if (!err) ++ err = au_iinfo_init(inode); ++ if (!err) ++ inode->i_version++; ++ else { ++ iget_failed(inode); ++ inode = ERR_PTR(err); ++ } ++ ++out: ++ /* never return NULL */ ++ AuDebugOn(!inode); ++ AuTraceErrPtr(inode); ++ return inode; ++} ++ ++/* lock free root dinfo */ ++static int au_show_brs(struct seq_file *seq, struct super_block *sb) ++{ ++ int err; ++ aufs_bindex_t bindex, bend; ++ struct path path; ++ struct au_hdentry *hdp; ++ struct au_branch *br; ++ au_br_perm_str_t perm; ++ ++ err = 0; ++ bend = au_sbend(sb); ++ hdp = au_di(sb->s_root)->di_hdentry; ++ for (bindex = 0; !err && bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ path.mnt = au_br_mnt(br); ++ path.dentry = hdp[bindex].hd_dentry; ++ err = au_seq_path(seq, &path); ++ if (!err) { ++ au_optstr_br_perm(&perm, br->br_perm); ++ err = seq_printf(seq, "=%s", perm.a); ++ if (err == -1) ++ err = -E2BIG; ++ } ++ if (!err && bindex != bend) ++ err = seq_putc(seq, ':'); ++ } ++ ++ return err; ++} ++ ++static void au_show_wbr_create(struct seq_file *m, int v, ++ struct au_sbinfo *sbinfo) ++{ ++ const char *pat; ++ ++ AuRwMustAnyLock(&sbinfo->si_rwsem); ++ ++ seq_puts(m, ",create="); ++ pat = au_optstr_wbr_create(v); ++ switch (v) { ++ case AuWbrCreate_TDP: ++ case AuWbrCreate_RR: ++ case AuWbrCreate_MFS: ++ case AuWbrCreate_PMFS: ++ seq_puts(m, pat); ++ break; ++ case AuWbrCreate_MFSV: ++ seq_printf(m, /*pat*/"mfs:%lu", ++ jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) ++ / MSEC_PER_SEC); ++ break; ++ case AuWbrCreate_PMFSV: ++ seq_printf(m, /*pat*/"pmfs:%lu", ++ jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) ++ / MSEC_PER_SEC); ++ break; ++ case AuWbrCreate_MFSRR: ++ seq_printf(m, /*pat*/"mfsrr:%llu", ++ sbinfo->si_wbr_mfs.mfsrr_watermark); ++ break; ++ case AuWbrCreate_MFSRRV: ++ seq_printf(m, /*pat*/"mfsrr:%llu:%lu", ++ sbinfo->si_wbr_mfs.mfsrr_watermark, ++ jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) ++ / MSEC_PER_SEC); ++ break; ++ case AuWbrCreate_PMFSRR: ++ seq_printf(m, /*pat*/"pmfsrr:%llu", ++ sbinfo->si_wbr_mfs.mfsrr_watermark); ++ break; ++ case AuWbrCreate_PMFSRRV: ++ seq_printf(m, /*pat*/"pmfsrr:%llu:%lu", ++ sbinfo->si_wbr_mfs.mfsrr_watermark, ++ jiffies_to_msecs(sbinfo->si_wbr_mfs.mfs_expire) ++ / MSEC_PER_SEC); ++ break; ++ } ++} ++ ++static int au_show_xino(struct seq_file *seq, struct super_block *sb) ++{ ++#ifdef CONFIG_SYSFS ++ return 0; ++#else ++ int err; ++ const int len = sizeof(AUFS_XINO_FNAME) - 1; ++ aufs_bindex_t bindex, brid; ++ struct qstr *name; ++ struct file *f; ++ struct dentry *d, *h_root; ++ struct au_hdentry *hdp; ++ ++ AuRwMustAnyLock(&sbinfo->si_rwsem); ++ ++ err = 0; ++ f = au_sbi(sb)->si_xib; ++ if (!f) ++ goto out; ++ ++ /* stop printing the default xino path on the first writable branch */ ++ h_root = NULL; ++ brid = au_xino_brid(sb); ++ if (brid >= 0) { ++ bindex = au_br_index(sb, brid); ++ hdp = au_di(sb->s_root)->di_hdentry; ++ h_root = hdp[0 + bindex].hd_dentry; ++ } ++ d = f->f_path.dentry; ++ name = &d->d_name; ++ /* safe ->d_parent because the file is unlinked */ ++ if (d->d_parent == h_root ++ && name->len == len ++ && !memcmp(name->name, AUFS_XINO_FNAME, len)) ++ goto out; ++ ++ seq_puts(seq, ",xino="); ++ err = au_xino_path(seq, f); ++ ++out: ++ return err; ++#endif ++} ++ ++/* seq_file will re-call me in case of too long string */ ++static int aufs_show_options(struct seq_file *m, struct dentry *dentry) ++{ ++ int err; ++ unsigned int mnt_flags, v; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ ++#define AuBool(name, str) do { \ ++ v = au_opt_test(mnt_flags, name); \ ++ if (v != au_opt_test(AuOpt_Def, name)) \ ++ seq_printf(m, ",%s" #str, v ? "" : "no"); \ ++} while (0) ++ ++#define AuStr(name, str) do { \ ++ v = mnt_flags & AuOptMask_##name; \ ++ if (v != (AuOpt_Def & AuOptMask_##name)) \ ++ seq_printf(m, "," #str "=%s", au_optstr_##str(v)); \ ++} while (0) ++ ++#define AuUInt(name, str, val) do { \ ++ if (val != AUFS_##name##_DEF) \ ++ seq_printf(m, "," #str "=%u", val); \ ++} while (0) ++ ++ sb = dentry->d_sb; ++ if (sb->s_flags & MS_POSIXACL) ++ seq_puts(m, ",acl"); ++ ++ /* lock free root dinfo */ ++ si_noflush_read_lock(sb); ++ sbinfo = au_sbi(sb); ++ seq_printf(m, ",si=%lx", sysaufs_si_id(sbinfo)); ++ ++ mnt_flags = au_mntflags(sb); ++ if (au_opt_test(mnt_flags, XINO)) { ++ err = au_show_xino(m, sb); ++ if (unlikely(err)) ++ goto out; ++ } else ++ seq_puts(m, ",noxino"); ++ ++ AuBool(TRUNC_XINO, trunc_xino); ++ AuStr(UDBA, udba); ++ AuBool(SHWH, shwh); ++ AuBool(PLINK, plink); ++ AuBool(DIO, dio); ++ AuBool(DIRPERM1, dirperm1); ++ ++ v = sbinfo->si_wbr_create; ++ if (v != AuWbrCreate_Def) ++ au_show_wbr_create(m, v, sbinfo); ++ ++ v = sbinfo->si_wbr_copyup; ++ if (v != AuWbrCopyup_Def) ++ seq_printf(m, ",cpup=%s", au_optstr_wbr_copyup(v)); ++ ++ v = au_opt_test(mnt_flags, ALWAYS_DIROPQ); ++ if (v != au_opt_test(AuOpt_Def, ALWAYS_DIROPQ)) ++ seq_printf(m, ",diropq=%c", v ? 'a' : 'w'); ++ ++ AuUInt(DIRWH, dirwh, sbinfo->si_dirwh); ++ ++ v = jiffies_to_msecs(sbinfo->si_rdcache) / MSEC_PER_SEC; ++ AuUInt(RDCACHE, rdcache, v); ++ ++ AuUInt(RDBLK, rdblk, sbinfo->si_rdblk); ++ AuUInt(RDHASH, rdhash, sbinfo->si_rdhash); ++ ++ au_fhsm_show(m, sbinfo); ++ ++ AuBool(SUM, sum); ++ /* AuBool(SUM_W, wsum); */ ++ AuBool(WARN_PERM, warn_perm); ++ AuBool(VERBOSE, verbose); ++ ++out: ++ /* be sure to print "br:" last */ ++ if (!sysaufs_brs) { ++ seq_puts(m, ",br:"); ++ au_show_brs(m, sb); ++ } ++ si_read_unlock(sb); ++ return 0; ++ ++#undef AuBool ++#undef AuStr ++#undef AuUInt ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* sum mode which returns the summation for statfs(2) */ ++ ++static u64 au_add_till_max(u64 a, u64 b) ++{ ++ u64 old; ++ ++ old = a; ++ a += b; ++ if (old <= a) ++ return a; ++ return ULLONG_MAX; ++} ++ ++static u64 au_mul_till_max(u64 a, long mul) ++{ ++ u64 old; ++ ++ old = a; ++ a *= mul; ++ if (old <= a) ++ return a; ++ return ULLONG_MAX; ++} ++ ++static int au_statfs_sum(struct super_block *sb, struct kstatfs *buf) ++{ ++ int err; ++ long bsize, factor; ++ u64 blocks, bfree, bavail, files, ffree; ++ aufs_bindex_t bend, bindex, i; ++ unsigned char shared; ++ struct path h_path; ++ struct super_block *h_sb; ++ ++ err = 0; ++ bsize = LONG_MAX; ++ files = 0; ++ ffree = 0; ++ blocks = 0; ++ bfree = 0; ++ bavail = 0; ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ h_path.mnt = au_sbr_mnt(sb, bindex); ++ h_sb = h_path.mnt->mnt_sb; ++ shared = 0; ++ for (i = 0; !shared && i < bindex; i++) ++ shared = (au_sbr_sb(sb, i) == h_sb); ++ if (shared) ++ continue; ++ ++ /* sb->s_root for NFS is unreliable */ ++ h_path.dentry = h_path.mnt->mnt_root; ++ err = vfs_statfs(&h_path, buf); ++ if (unlikely(err)) ++ goto out; ++ ++ if (bsize > buf->f_bsize) { ++ /* ++ * we will reduce bsize, so we have to expand blocks ++ * etc. to match them again ++ */ ++ factor = (bsize / buf->f_bsize); ++ blocks = au_mul_till_max(blocks, factor); ++ bfree = au_mul_till_max(bfree, factor); ++ bavail = au_mul_till_max(bavail, factor); ++ bsize = buf->f_bsize; ++ } ++ ++ factor = (buf->f_bsize / bsize); ++ blocks = au_add_till_max(blocks, ++ au_mul_till_max(buf->f_blocks, factor)); ++ bfree = au_add_till_max(bfree, ++ au_mul_till_max(buf->f_bfree, factor)); ++ bavail = au_add_till_max(bavail, ++ au_mul_till_max(buf->f_bavail, factor)); ++ files = au_add_till_max(files, buf->f_files); ++ ffree = au_add_till_max(ffree, buf->f_ffree); ++ } ++ ++ buf->f_bsize = bsize; ++ buf->f_blocks = blocks; ++ buf->f_bfree = bfree; ++ buf->f_bavail = bavail; ++ buf->f_files = files; ++ buf->f_ffree = ffree; ++ buf->f_frsize = 0; ++ ++out: ++ return err; ++} ++ ++static int aufs_statfs(struct dentry *dentry, struct kstatfs *buf) ++{ ++ int err; ++ struct path h_path; ++ struct super_block *sb; ++ ++ /* lock free root dinfo */ ++ sb = dentry->d_sb; ++ si_noflush_read_lock(sb); ++ if (!au_opt_test(au_mntflags(sb), SUM)) { ++ /* sb->s_root for NFS is unreliable */ ++ h_path.mnt = au_sbr_mnt(sb, 0); ++ h_path.dentry = h_path.mnt->mnt_root; ++ err = vfs_statfs(&h_path, buf); ++ } else ++ err = au_statfs_sum(sb, buf); ++ si_read_unlock(sb); ++ ++ if (!err) { ++ buf->f_type = AUFS_SUPER_MAGIC; ++ buf->f_namelen = AUFS_MAX_NAMELEN; ++ memset(&buf->f_fsid, 0, sizeof(buf->f_fsid)); ++ } ++ /* buf->f_bsize = buf->f_blocks = buf->f_bfree = buf->f_bavail = -1; */ ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int aufs_sync_fs(struct super_block *sb, int wait) ++{ ++ int err, e; ++ aufs_bindex_t bend, bindex; ++ struct au_branch *br; ++ struct super_block *h_sb; ++ ++ err = 0; ++ si_noflush_read_lock(sb); ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (!au_br_writable(br->br_perm)) ++ continue; ++ ++ h_sb = au_sbr_sb(sb, bindex); ++ if (h_sb->s_op->sync_fs) { ++ e = h_sb->s_op->sync_fs(h_sb, wait); ++ if (unlikely(e && !err)) ++ err = e; ++ /* go on even if an error happens */ ++ } ++ } ++ si_read_unlock(sb); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* final actions when unmounting a file system */ ++static void aufs_put_super(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ sbinfo = au_sbi(sb); ++ if (!sbinfo) ++ return; ++ ++ dbgaufs_si_fin(sbinfo); ++ kobject_put(&sbinfo->si_kobj); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_array_free(void *array) ++{ ++ if (array) { ++ if (!is_vmalloc_addr(array)) ++ kfree(array); ++ else ++ vfree(array); ++ } ++} ++ ++void *au_array_alloc(unsigned long long *hint, au_arraycb_t cb, void *arg) ++{ ++ void *array; ++ unsigned long long n, sz; ++ ++ array = NULL; ++ n = 0; ++ if (!*hint) ++ goto out; ++ ++ if (*hint > ULLONG_MAX / sizeof(array)) { ++ array = ERR_PTR(-EMFILE); ++ pr_err("hint %llu\n", *hint); ++ goto out; ++ } ++ ++ sz = sizeof(array) * *hint; ++ array = kzalloc(sz, GFP_NOFS); ++ if (unlikely(!array)) ++ array = vzalloc(sz); ++ if (unlikely(!array)) { ++ array = ERR_PTR(-ENOMEM); ++ goto out; ++ } ++ ++ n = cb(array, *hint, arg); ++ AuDebugOn(n > *hint); ++ ++out: ++ *hint = n; ++ return array; ++} ++ ++static unsigned long long au_iarray_cb(void *a, ++ unsigned long long max __maybe_unused, ++ void *arg) ++{ ++ unsigned long long n; ++ struct inode **p, *inode; ++ struct list_head *head; ++ ++ n = 0; ++ p = a; ++ head = arg; ++ spin_lock(&inode_sb_list_lock); ++ list_for_each_entry(inode, head, i_sb_list) { ++ if (!is_bad_inode(inode) ++ && au_ii(inode)->ii_bstart >= 0) { ++ spin_lock(&inode->i_lock); ++ if (atomic_read(&inode->i_count)) { ++ au_igrab(inode); ++ *p++ = inode; ++ n++; ++ AuDebugOn(n > max); ++ } ++ spin_unlock(&inode->i_lock); ++ } ++ } ++ spin_unlock(&inode_sb_list_lock); ++ ++ return n; ++} ++ ++struct inode **au_iarray_alloc(struct super_block *sb, unsigned long long *max) ++{ ++ *max = atomic_long_read(&au_sbi(sb)->si_ninodes); ++ return au_array_alloc(max, au_iarray_cb, &sb->s_inodes); ++} ++ ++void au_iarray_free(struct inode **a, unsigned long long max) ++{ ++ unsigned long long ull; ++ ++ for (ull = 0; ull < max; ull++) ++ iput(a[ull]); ++ au_array_free(a); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * refresh dentry and inode at remount time. ++ */ ++/* todo: consolidate with simple_reval_dpath() and au_reval_for_attr() */ ++static int au_do_refresh(struct dentry *dentry, unsigned int dir_flags, ++ struct dentry *parent) ++{ ++ int err; ++ ++ di_write_lock_child(dentry); ++ di_read_lock_parent(parent, AuLock_IR); ++ err = au_refresh_dentry(dentry, parent); ++ if (!err && dir_flags) ++ au_hn_reset(d_inode(dentry), dir_flags); ++ di_read_unlock(parent, AuLock_IR); ++ di_write_unlock(dentry); ++ ++ return err; ++} ++ ++static int au_do_refresh_d(struct dentry *dentry, unsigned int sigen, ++ struct au_sbinfo *sbinfo, ++ const unsigned int dir_flags) ++{ ++ int err; ++ struct dentry *parent; ++ ++ err = 0; ++ parent = dget_parent(dentry); ++ if (!au_digen_test(parent, sigen) && au_digen_test(dentry, sigen)) { ++ if (d_really_is_positive(dentry)) { ++ if (!d_is_dir(dentry)) ++ err = au_do_refresh(dentry, /*dir_flags*/0, ++ parent); ++ else { ++ err = au_do_refresh(dentry, dir_flags, parent); ++ if (unlikely(err)) ++ au_fset_si(sbinfo, FAILED_REFRESH_DIR); ++ } ++ } else ++ err = au_do_refresh(dentry, /*dir_flags*/0, parent); ++ AuDbgDentry(dentry); ++ } ++ dput(parent); ++ ++ AuTraceErr(err); ++ return err; ++} ++ ++static int au_refresh_d(struct super_block *sb) ++{ ++ int err, i, j, ndentry, e; ++ unsigned int sigen; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry **dentries, *d; ++ struct au_sbinfo *sbinfo; ++ struct dentry *root = sb->s_root; ++ const unsigned int dir_flags = au_hi_flags(d_inode(root), /*isdir*/1); ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_dcsub_pages(&dpages, root, NULL, NULL); ++ if (unlikely(err)) ++ goto out_dpages; ++ ++ sigen = au_sigen(sb); ++ sbinfo = au_sbi(sb); ++ for (i = 0; i < dpages.ndpage; i++) { ++ dpage = dpages.dpages + i; ++ dentries = dpage->dentries; ++ ndentry = dpage->ndentry; ++ for (j = 0; j < ndentry; j++) { ++ d = dentries[j]; ++ e = au_do_refresh_d(d, sigen, sbinfo, dir_flags); ++ if (unlikely(e && !err)) ++ err = e; ++ /* go on even err */ ++ } ++ } ++ ++out_dpages: ++ au_dpages_free(&dpages); ++out: ++ return err; ++} ++ ++static int au_refresh_i(struct super_block *sb) ++{ ++ int err, e; ++ unsigned int sigen; ++ unsigned long long max, ull; ++ struct inode *inode, **array; ++ ++ array = au_iarray_alloc(sb, &max); ++ err = PTR_ERR(array); ++ if (IS_ERR(array)) ++ goto out; ++ ++ err = 0; ++ sigen = au_sigen(sb); ++ for (ull = 0; ull < max; ull++) { ++ inode = array[ull]; ++ if (unlikely(!inode)) ++ break; ++ if (au_iigen(inode, NULL) != sigen) { ++ ii_write_lock_child(inode); ++ e = au_refresh_hinode_self(inode); ++ ii_write_unlock(inode); ++ if (unlikely(e)) { ++ pr_err("error %d, i%lu\n", e, inode->i_ino); ++ if (!err) ++ err = e; ++ /* go on even if err */ ++ } ++ } ++ } ++ ++ au_iarray_free(array, max); ++ ++out: ++ return err; ++} ++ ++static void au_remount_refresh(struct super_block *sb) ++{ ++ int err, e; ++ unsigned int udba; ++ aufs_bindex_t bindex, bend; ++ struct dentry *root; ++ struct inode *inode; ++ struct au_branch *br; ++ ++ au_sigen_inc(sb); ++ au_fclr_si(au_sbi(sb), FAILED_REFRESH_DIR); ++ ++ root = sb->s_root; ++ DiMustNoWaiters(root); ++ inode = d_inode(root); ++ IiMustNoWaiters(inode); ++ ++ udba = au_opt_udba(sb); ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ err = au_hnotify_reset_br(udba, br, br->br_perm); ++ if (unlikely(err)) ++ AuIOErr("hnotify failed on br %d, %d, ignored\n", ++ bindex, err); ++ /* go on even if err */ ++ } ++ au_hn_reset(inode, au_hi_flags(inode, /*isdir*/1)); ++ ++ di_write_unlock(root); ++ err = au_refresh_d(sb); ++ e = au_refresh_i(sb); ++ if (unlikely(e && !err)) ++ err = e; ++ /* aufs_write_lock() calls ..._child() */ ++ di_write_lock_child(root); ++ ++ au_cpup_attr_all(inode, /*force*/1); ++ ++ if (unlikely(err)) ++ AuIOErr("refresh failed, ignored, %d\n", err); ++} ++ ++/* stop extra interpretation of errno in mount(8), and strange error messages */ ++static int cvt_err(int err) ++{ ++ AuTraceErr(err); ++ ++ switch (err) { ++ case -ENOENT: ++ case -ENOTDIR: ++ case -EEXIST: ++ case -EIO: ++ err = -EINVAL; ++ } ++ return err; ++} ++ ++static int aufs_remount_fs(struct super_block *sb, int *flags, char *data) ++{ ++ int err, do_dx; ++ unsigned int mntflags; ++ struct au_opts opts; ++ struct dentry *root; ++ struct inode *inode; ++ struct au_sbinfo *sbinfo; ++ ++ err = 0; ++ root = sb->s_root; ++ if (!data || !*data) { ++ err = si_write_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (!err) { ++ di_write_lock_child(root); ++ err = au_opts_verify(sb, *flags, /*pending*/0); ++ aufs_write_unlock(root); ++ } ++ goto out; ++ } ++ ++ err = -ENOMEM; ++ memset(&opts, 0, sizeof(opts)); ++ opts.opt = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!opts.opt)) ++ goto out; ++ opts.max_opt = PAGE_SIZE / sizeof(*opts.opt); ++ opts.flags = AuOpts_REMOUNT; ++ opts.sb_flags = *flags; ++ ++ /* parse it before aufs lock */ ++ err = au_opts_parse(sb, data, &opts); ++ if (unlikely(err)) ++ goto out_opts; ++ ++ sbinfo = au_sbi(sb); ++ inode = d_inode(root); ++ mutex_lock(&inode->i_mutex); ++ err = si_write_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out_mtx; ++ di_write_lock_child(root); ++ ++ /* au_opts_remount() may return an error */ ++ err = au_opts_remount(sb, &opts); ++ au_opts_free(&opts); ++ ++ if (au_ftest_opts(opts.flags, REFRESH)) ++ au_remount_refresh(sb); ++ ++ if (au_ftest_opts(opts.flags, REFRESH_DYAOP)) { ++ mntflags = au_mntflags(sb); ++ do_dx = !!au_opt_test(mntflags, DIO); ++ au_dy_arefresh(do_dx); ++ } ++ ++ au_fhsm_wrote_all(sb, /*force*/1); /* ?? */ ++ aufs_write_unlock(root); ++ ++out_mtx: ++ mutex_unlock(&inode->i_mutex); ++out_opts: ++ free_page((unsigned long)opts.opt); ++out: ++ err = cvt_err(err); ++ AuTraceErr(err); ++ return err; ++} ++ ++static const struct super_operations aufs_sop = { ++ .alloc_inode = aufs_alloc_inode, ++ .destroy_inode = aufs_destroy_inode, ++ /* always deleting, no clearing */ ++ .drop_inode = generic_delete_inode, ++ .show_options = aufs_show_options, ++ .statfs = aufs_statfs, ++ .put_super = aufs_put_super, ++ .sync_fs = aufs_sync_fs, ++ .remount_fs = aufs_remount_fs ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int alloc_root(struct super_block *sb) ++{ ++ int err; ++ struct inode *inode; ++ struct dentry *root; ++ ++ err = -ENOMEM; ++ inode = au_iget_locked(sb, AUFS_ROOT_INO); ++ err = PTR_ERR(inode); ++ if (IS_ERR(inode)) ++ goto out; ++ ++ inode->i_op = &aufs_dir_iop; ++ inode->i_fop = &aufs_dir_fop; ++ inode->i_mode = S_IFDIR; ++ set_nlink(inode, 2); ++ unlock_new_inode(inode); ++ ++ root = d_make_root(inode); ++ if (unlikely(!root)) ++ goto out; ++ err = PTR_ERR(root); ++ if (IS_ERR(root)) ++ goto out; ++ ++ err = au_di_init(root); ++ if (!err) { ++ sb->s_root = root; ++ return 0; /* success */ ++ } ++ dput(root); ++ ++out: ++ return err; ++} ++ ++static int aufs_fill_super(struct super_block *sb, void *raw_data, ++ int silent __maybe_unused) ++{ ++ int err; ++ struct au_opts opts; ++ struct dentry *root; ++ struct inode *inode; ++ char *arg = raw_data; ++ ++ if (unlikely(!arg || !*arg)) { ++ err = -EINVAL; ++ pr_err("no arg\n"); ++ goto out; ++ } ++ ++ err = -ENOMEM; ++ memset(&opts, 0, sizeof(opts)); ++ opts.opt = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!opts.opt)) ++ goto out; ++ opts.max_opt = PAGE_SIZE / sizeof(*opts.opt); ++ opts.sb_flags = sb->s_flags; ++ ++ err = au_si_alloc(sb); ++ if (unlikely(err)) ++ goto out_opts; ++ ++ /* all timestamps always follow the ones on the branch */ ++ sb->s_flags |= MS_NOATIME | MS_NODIRATIME; ++ sb->s_op = &aufs_sop; ++ sb->s_d_op = &aufs_dop; ++ sb->s_magic = AUFS_SUPER_MAGIC; ++ sb->s_maxbytes = 0; ++ sb->s_stack_depth = 1; ++ au_export_init(sb); ++ /* au_xattr_init(sb); */ ++ ++ err = alloc_root(sb); ++ if (unlikely(err)) { ++ si_write_unlock(sb); ++ goto out_info; ++ } ++ root = sb->s_root; ++ inode = d_inode(root); ++ ++ /* ++ * actually we can parse options regardless aufs lock here. ++ * but at remount time, parsing must be done before aufs lock. ++ * so we follow the same rule. ++ */ ++ ii_write_lock_parent(inode); ++ aufs_write_unlock(root); ++ err = au_opts_parse(sb, arg, &opts); ++ if (unlikely(err)) ++ goto out_root; ++ ++ /* lock vfs_inode first, then aufs. */ ++ mutex_lock(&inode->i_mutex); ++ aufs_write_lock(root); ++ err = au_opts_mount(sb, &opts); ++ au_opts_free(&opts); ++ aufs_write_unlock(root); ++ mutex_unlock(&inode->i_mutex); ++ if (!err) ++ goto out_opts; /* success */ ++ ++out_root: ++ dput(root); ++ sb->s_root = NULL; ++out_info: ++ dbgaufs_si_fin(au_sbi(sb)); ++ kobject_put(&au_sbi(sb)->si_kobj); ++ sb->s_fs_info = NULL; ++out_opts: ++ free_page((unsigned long)opts.opt); ++out: ++ AuTraceErr(err); ++ err = cvt_err(err); ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct dentry *aufs_mount(struct file_system_type *fs_type, int flags, ++ const char *dev_name __maybe_unused, ++ void *raw_data) ++{ ++ struct dentry *root; ++ struct super_block *sb; ++ ++ /* all timestamps always follow the ones on the branch */ ++ /* mnt->mnt_flags |= MNT_NOATIME | MNT_NODIRATIME; */ ++ root = mount_nodev(fs_type, flags, raw_data, aufs_fill_super); ++ if (IS_ERR(root)) ++ goto out; ++ ++ sb = root->d_sb; ++ si_write_lock(sb, !AuLock_FLUSH); ++ sysaufs_brs_add(sb, 0); ++ si_write_unlock(sb); ++ au_sbilist_add(sb); ++ ++out: ++ return root; ++} ++ ++static void aufs_kill_sb(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ sbinfo = au_sbi(sb); ++ if (sbinfo) { ++ au_sbilist_del(sb); ++ aufs_write_lock(sb->s_root); ++ au_fhsm_fin(sb); ++ if (sbinfo->si_wbr_create_ops->fin) ++ sbinfo->si_wbr_create_ops->fin(sb); ++ if (au_opt_test(sbinfo->si_mntflags, UDBA_HNOTIFY)) { ++ au_opt_set_udba(sbinfo->si_mntflags, UDBA_NONE); ++ au_remount_refresh(sb); ++ } ++ if (au_opt_test(sbinfo->si_mntflags, PLINK)) ++ au_plink_put(sb, /*verbose*/1); ++ au_xino_clr(sb); ++ sbinfo->si_sb = NULL; ++ aufs_write_unlock(sb->s_root); ++ au_nwt_flush(&sbinfo->si_nowait); ++ } ++ kill_anon_super(sb); ++} ++ ++struct file_system_type aufs_fs_type = { ++ .name = AUFS_FSTYPE, ++ /* a race between rename and others */ ++ .fs_flags = FS_RENAME_DOES_D_MOVE, ++ .mount = aufs_mount, ++ .kill_sb = aufs_kill_sb, ++ /* no need to __module_get() and module_put(). */ ++ .owner = THIS_MODULE, ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/super.h linux-4.1.10/fs/aufs/super.h +--- linux-4.1.10.orig/fs/aufs/super.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/super.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,635 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * super_block operations ++ */ ++ ++#ifndef __AUFS_SUPER_H__ ++#define __AUFS_SUPER_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fs.h> ++#include <linux/kobject.h> ++#include "rwsem.h" ++#include "spl.h" ++#include "wkq.h" ++ ++/* policies to select one among multiple writable branches */ ++struct au_wbr_copyup_operations { ++ int (*copyup)(struct dentry *dentry); ++}; ++ ++#define AuWbr_DIR 1 /* target is a dir */ ++#define AuWbr_PARENT (1 << 1) /* always require a parent */ ++ ++#define au_ftest_wbr(flags, name) ((flags) & AuWbr_##name) ++#define au_fset_wbr(flags, name) { (flags) |= AuWbr_##name; } ++#define au_fclr_wbr(flags, name) { (flags) &= ~AuWbr_##name; } ++ ++struct au_wbr_create_operations { ++ int (*create)(struct dentry *dentry, unsigned int flags); ++ int (*init)(struct super_block *sb); ++ int (*fin)(struct super_block *sb); ++}; ++ ++struct au_wbr_mfs { ++ struct mutex mfs_lock; /* protect this structure */ ++ unsigned long mfs_jiffy; ++ unsigned long mfs_expire; ++ aufs_bindex_t mfs_bindex; ++ ++ unsigned long long mfsrr_bytes; ++ unsigned long long mfsrr_watermark; ++}; ++ ++struct pseudo_link { ++ union { ++ struct hlist_node hlist; ++ struct rcu_head rcu; ++ }; ++ struct inode *inode; ++}; ++ ++#define AuPlink_NHASH 100 ++static inline int au_plink_hash(ino_t ino) ++{ ++ return ino % AuPlink_NHASH; ++} ++ ++/* File-based Hierarchical Storage Management */ ++struct au_fhsm { ++#ifdef CONFIG_AUFS_FHSM ++ /* allow only one process who can receive the notification */ ++ spinlock_t fhsm_spin; ++ pid_t fhsm_pid; ++ wait_queue_head_t fhsm_wqh; ++ atomic_t fhsm_readable; ++ ++ /* these are protected by si_rwsem */ ++ unsigned long fhsm_expire; ++ aufs_bindex_t fhsm_bottom; ++#endif ++}; ++ ++struct au_branch; ++struct au_sbinfo { ++ /* nowait tasks in the system-wide workqueue */ ++ struct au_nowait_tasks si_nowait; ++ ++ /* ++ * tried sb->s_umount, but failed due to the dependecy between i_mutex. ++ * rwsem for au_sbinfo is necessary. ++ */ ++ struct au_rwsem si_rwsem; ++ ++ /* prevent recursive locking in deleting inode */ ++ struct { ++ unsigned long *bitmap; ++ spinlock_t tree_lock; ++ struct radix_tree_root tree; ++ } au_si_pid; ++ ++ /* ++ * dirty approach to protect sb->sb_inodes and ->s_files (gone) from ++ * remount. ++ */ ++ atomic_long_t si_ninodes, si_nfiles; ++ ++ /* branch management */ ++ unsigned int si_generation; ++ ++ /* see AuSi_ flags */ ++ unsigned char au_si_status; ++ ++ aufs_bindex_t si_bend; ++ ++ /* dirty trick to keep br_id plus */ ++ unsigned int si_last_br_id : ++ sizeof(aufs_bindex_t) * BITS_PER_BYTE - 1; ++ struct au_branch **si_branch; ++ ++ /* policy to select a writable branch */ ++ unsigned char si_wbr_copyup; ++ unsigned char si_wbr_create; ++ struct au_wbr_copyup_operations *si_wbr_copyup_ops; ++ struct au_wbr_create_operations *si_wbr_create_ops; ++ ++ /* round robin */ ++ atomic_t si_wbr_rr_next; ++ ++ /* most free space */ ++ struct au_wbr_mfs si_wbr_mfs; ++ ++ /* File-based Hierarchical Storage Management */ ++ struct au_fhsm si_fhsm; ++ ++ /* mount flags */ ++ /* include/asm-ia64/siginfo.h defines a macro named si_flags */ ++ unsigned int si_mntflags; ++ ++ /* external inode number (bitmap and translation table) */ ++ vfs_readf_t si_xread; ++ vfs_writef_t si_xwrite; ++ struct file *si_xib; ++ struct mutex si_xib_mtx; /* protect xib members */ ++ unsigned long *si_xib_buf; ++ unsigned long si_xib_last_pindex; ++ int si_xib_next_bit; ++ aufs_bindex_t si_xino_brid; ++ unsigned long si_xino_jiffy; ++ unsigned long si_xino_expire; ++ /* reserved for future use */ ++ /* unsigned long long si_xib_limit; */ /* Max xib file size */ ++ ++#ifdef CONFIG_AUFS_EXPORT ++ /* i_generation */ ++ struct file *si_xigen; ++ atomic_t si_xigen_next; ++#endif ++ ++ /* dirty trick to suppoer atomic_open */ ++ struct au_sphlhead si_aopen; ++ ++ /* vdir parameters */ ++ unsigned long si_rdcache; /* max cache time in jiffies */ ++ unsigned int si_rdblk; /* deblk size */ ++ unsigned int si_rdhash; /* hash size */ ++ ++ /* ++ * If the number of whiteouts are larger than si_dirwh, leave all of ++ * them after au_whtmp_ren to reduce the cost of rmdir(2). ++ * future fsck.aufs or kernel thread will remove them later. ++ * Otherwise, remove all whiteouts and the dir in rmdir(2). ++ */ ++ unsigned int si_dirwh; ++ ++ /* pseudo_link list */ ++ struct au_sphlhead si_plink[AuPlink_NHASH]; ++ wait_queue_head_t si_plink_wq; ++ spinlock_t si_plink_maint_lock; ++ pid_t si_plink_maint_pid; ++ ++ /* file list */ ++ struct au_sphlhead si_files; ++ ++ /* ++ * sysfs and lifetime management. ++ * this is not a small structure and it may be a waste of memory in case ++ * of sysfs is disabled, particulary when many aufs-es are mounted. ++ * but using sysfs is majority. ++ */ ++ struct kobject si_kobj; ++#ifdef CONFIG_DEBUG_FS ++ struct dentry *si_dbgaufs; ++ struct dentry *si_dbgaufs_plink; ++ struct dentry *si_dbgaufs_xib; ++#ifdef CONFIG_AUFS_EXPORT ++ struct dentry *si_dbgaufs_xigen; ++#endif ++#endif ++ ++#ifdef CONFIG_AUFS_SBILIST ++ struct list_head si_list; ++#endif ++ ++ /* dirty, necessary for unmounting, sysfs and sysrq */ ++ struct super_block *si_sb; ++}; ++ ++/* sbinfo status flags */ ++/* ++ * set true when refresh_dirs() failed at remount time. ++ * then try refreshing dirs at access time again. ++ * if it is false, refreshing dirs at access time is unnecesary ++ */ ++#define AuSi_FAILED_REFRESH_DIR 1 ++ ++#define AuSi_FHSM (1 << 1) /* fhsm is active now */ ++ ++#ifndef CONFIG_AUFS_FHSM ++#undef AuSi_FHSM ++#define AuSi_FHSM 0 ++#endif ++ ++static inline unsigned char au_do_ftest_si(struct au_sbinfo *sbi, ++ unsigned int flag) ++{ ++ AuRwMustAnyLock(&sbi->si_rwsem); ++ return sbi->au_si_status & flag; ++} ++#define au_ftest_si(sbinfo, name) au_do_ftest_si(sbinfo, AuSi_##name) ++#define au_fset_si(sbinfo, name) do { \ ++ AuRwMustWriteLock(&(sbinfo)->si_rwsem); \ ++ (sbinfo)->au_si_status |= AuSi_##name; \ ++} while (0) ++#define au_fclr_si(sbinfo, name) do { \ ++ AuRwMustWriteLock(&(sbinfo)->si_rwsem); \ ++ (sbinfo)->au_si_status &= ~AuSi_##name; \ ++} while (0) ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* policy to select one among writable branches */ ++#define AuWbrCopyup(sbinfo, ...) \ ++ ((sbinfo)->si_wbr_copyup_ops->copyup(__VA_ARGS__)) ++#define AuWbrCreate(sbinfo, ...) \ ++ ((sbinfo)->si_wbr_create_ops->create(__VA_ARGS__)) ++ ++/* flags for si_read_lock()/aufs_read_lock()/di_read_lock() */ ++#define AuLock_DW 1 /* write-lock dentry */ ++#define AuLock_IR (1 << 1) /* read-lock inode */ ++#define AuLock_IW (1 << 2) /* write-lock inode */ ++#define AuLock_FLUSH (1 << 3) /* wait for 'nowait' tasks */ ++#define AuLock_DIR (1 << 4) /* target is a dir */ ++#define AuLock_NOPLM (1 << 5) /* return err in plm mode */ ++#define AuLock_NOPLMW (1 << 6) /* wait for plm mode ends */ ++#define AuLock_GEN (1 << 7) /* test digen/iigen */ ++#define au_ftest_lock(flags, name) ((flags) & AuLock_##name) ++#define au_fset_lock(flags, name) \ ++ do { (flags) |= AuLock_##name; } while (0) ++#define au_fclr_lock(flags, name) \ ++ do { (flags) &= ~AuLock_##name; } while (0) ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* super.c */ ++extern struct file_system_type aufs_fs_type; ++struct inode *au_iget_locked(struct super_block *sb, ino_t ino); ++typedef unsigned long long (*au_arraycb_t)(void *array, unsigned long long max, ++ void *arg); ++void au_array_free(void *array); ++void *au_array_alloc(unsigned long long *hint, au_arraycb_t cb, void *arg); ++struct inode **au_iarray_alloc(struct super_block *sb, unsigned long long *max); ++void au_iarray_free(struct inode **a, unsigned long long max); ++ ++/* sbinfo.c */ ++void au_si_free(struct kobject *kobj); ++int au_si_alloc(struct super_block *sb); ++int au_sbr_realloc(struct au_sbinfo *sbinfo, int nbr); ++ ++unsigned int au_sigen_inc(struct super_block *sb); ++aufs_bindex_t au_new_br_id(struct super_block *sb); ++ ++int si_read_lock(struct super_block *sb, int flags); ++int si_write_lock(struct super_block *sb, int flags); ++int aufs_read_lock(struct dentry *dentry, int flags); ++void aufs_read_unlock(struct dentry *dentry, int flags); ++void aufs_write_lock(struct dentry *dentry); ++void aufs_write_unlock(struct dentry *dentry); ++int aufs_read_and_write_lock2(struct dentry *d1, struct dentry *d2, int flags); ++void aufs_read_and_write_unlock2(struct dentry *d1, struct dentry *d2); ++ ++int si_pid_test_slow(struct super_block *sb); ++void si_pid_set_slow(struct super_block *sb); ++void si_pid_clr_slow(struct super_block *sb); ++ ++/* wbr_policy.c */ ++extern struct au_wbr_copyup_operations au_wbr_copyup_ops[]; ++extern struct au_wbr_create_operations au_wbr_create_ops[]; ++int au_cpdown_dirs(struct dentry *dentry, aufs_bindex_t bdst); ++int au_wbr_nonopq(struct dentry *dentry, aufs_bindex_t bindex); ++int au_wbr_do_copyup_bu(struct dentry *dentry, aufs_bindex_t bstart); ++ ++/* mvdown.c */ ++int au_mvdown(struct dentry *dentry, struct aufs_mvdown __user *arg); ++ ++#ifdef CONFIG_AUFS_FHSM ++/* fhsm.c */ ++ ++static inline pid_t au_fhsm_pid(struct au_fhsm *fhsm) ++{ ++ pid_t pid; ++ ++ spin_lock(&fhsm->fhsm_spin); ++ pid = fhsm->fhsm_pid; ++ spin_unlock(&fhsm->fhsm_spin); ++ ++ return pid; ++} ++ ++void au_fhsm_wrote(struct super_block *sb, aufs_bindex_t bindex, int force); ++void au_fhsm_wrote_all(struct super_block *sb, int force); ++int au_fhsm_fd(struct super_block *sb, int oflags); ++int au_fhsm_br_alloc(struct au_branch *br); ++void au_fhsm_set_bottom(struct super_block *sb, aufs_bindex_t bindex); ++void au_fhsm_fin(struct super_block *sb); ++void au_fhsm_init(struct au_sbinfo *sbinfo); ++void au_fhsm_set(struct au_sbinfo *sbinfo, unsigned int sec); ++void au_fhsm_show(struct seq_file *seq, struct au_sbinfo *sbinfo); ++#else ++AuStubVoid(au_fhsm_wrote, struct super_block *sb, aufs_bindex_t bindex, ++ int force) ++AuStubVoid(au_fhsm_wrote_all, struct super_block *sb, int force) ++AuStub(int, au_fhsm_fd, return -EOPNOTSUPP, struct super_block *sb, int oflags) ++AuStub(pid_t, au_fhsm_pid, return 0, struct au_fhsm *fhsm) ++AuStubInt0(au_fhsm_br_alloc, struct au_branch *br) ++AuStubVoid(au_fhsm_set_bottom, struct super_block *sb, aufs_bindex_t bindex) ++AuStubVoid(au_fhsm_fin, struct super_block *sb) ++AuStubVoid(au_fhsm_init, struct au_sbinfo *sbinfo) ++AuStubVoid(au_fhsm_set, struct au_sbinfo *sbinfo, unsigned int sec) ++AuStubVoid(au_fhsm_show, struct seq_file *seq, struct au_sbinfo *sbinfo) ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct au_sbinfo *au_sbi(struct super_block *sb) ++{ ++ return sb->s_fs_info; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_EXPORT ++int au_test_nfsd(void); ++void au_export_init(struct super_block *sb); ++void au_xigen_inc(struct inode *inode); ++int au_xigen_new(struct inode *inode); ++int au_xigen_set(struct super_block *sb, struct file *base); ++void au_xigen_clr(struct super_block *sb); ++ ++static inline int au_busy_or_stale(void) ++{ ++ if (!au_test_nfsd()) ++ return -EBUSY; ++ return -ESTALE; ++} ++#else ++AuStubInt0(au_test_nfsd, void) ++AuStubVoid(au_export_init, struct super_block *sb) ++AuStubVoid(au_xigen_inc, struct inode *inode) ++AuStubInt0(au_xigen_new, struct inode *inode) ++AuStubInt0(au_xigen_set, struct super_block *sb, struct file *base) ++AuStubVoid(au_xigen_clr, struct super_block *sb) ++AuStub(int, au_busy_or_stale, return -EBUSY, void) ++#endif /* CONFIG_AUFS_EXPORT */ ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_SBILIST ++/* module.c */ ++extern struct au_splhead au_sbilist; ++ ++static inline void au_sbilist_init(void) ++{ ++ au_spl_init(&au_sbilist); ++} ++ ++static inline void au_sbilist_add(struct super_block *sb) ++{ ++ au_spl_add(&au_sbi(sb)->si_list, &au_sbilist); ++} ++ ++static inline void au_sbilist_del(struct super_block *sb) ++{ ++ au_spl_del(&au_sbi(sb)->si_list, &au_sbilist); ++} ++ ++#ifdef CONFIG_AUFS_MAGIC_SYSRQ ++static inline void au_sbilist_lock(void) ++{ ++ spin_lock(&au_sbilist.spin); ++} ++ ++static inline void au_sbilist_unlock(void) ++{ ++ spin_unlock(&au_sbilist.spin); ++} ++#define AuGFP_SBILIST GFP_ATOMIC ++#else ++AuStubVoid(au_sbilist_lock, void) ++AuStubVoid(au_sbilist_unlock, void) ++#define AuGFP_SBILIST GFP_NOFS ++#endif /* CONFIG_AUFS_MAGIC_SYSRQ */ ++#else ++AuStubVoid(au_sbilist_init, void) ++AuStubVoid(au_sbilist_add, struct super_block *sb) ++AuStubVoid(au_sbilist_del, struct super_block *sb) ++AuStubVoid(au_sbilist_lock, void) ++AuStubVoid(au_sbilist_unlock, void) ++#define AuGFP_SBILIST GFP_NOFS ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline void dbgaufs_si_null(struct au_sbinfo *sbinfo) ++{ ++ /* ++ * This function is a dynamic '__init' function actually, ++ * so the tiny check for si_rwsem is unnecessary. ++ */ ++ /* AuRwMustWriteLock(&sbinfo->si_rwsem); */ ++#ifdef CONFIG_DEBUG_FS ++ sbinfo->si_dbgaufs = NULL; ++ sbinfo->si_dbgaufs_plink = NULL; ++ sbinfo->si_dbgaufs_xib = NULL; ++#ifdef CONFIG_AUFS_EXPORT ++ sbinfo->si_dbgaufs_xigen = NULL; ++#endif ++#endif ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline pid_t si_pid_bit(void) ++{ ++ /* the origin of pid is 1, but the bitmap's is 0 */ ++ return current->pid - 1; ++} ++ ++static inline int si_pid_test(struct super_block *sb) ++{ ++ pid_t bit; ++ ++ bit = si_pid_bit(); ++ if (bit < PID_MAX_DEFAULT) ++ return test_bit(bit, au_sbi(sb)->au_si_pid.bitmap); ++ return si_pid_test_slow(sb); ++} ++ ++static inline void si_pid_set(struct super_block *sb) ++{ ++ pid_t bit; ++ ++ bit = si_pid_bit(); ++ if (bit < PID_MAX_DEFAULT) { ++ AuDebugOn(test_bit(bit, au_sbi(sb)->au_si_pid.bitmap)); ++ set_bit(bit, au_sbi(sb)->au_si_pid.bitmap); ++ /* smp_mb(); */ ++ } else ++ si_pid_set_slow(sb); ++} ++ ++static inline void si_pid_clr(struct super_block *sb) ++{ ++ pid_t bit; ++ ++ bit = si_pid_bit(); ++ if (bit < PID_MAX_DEFAULT) { ++ AuDebugOn(!test_bit(bit, au_sbi(sb)->au_si_pid.bitmap)); ++ clear_bit(bit, au_sbi(sb)->au_si_pid.bitmap); ++ /* smp_mb(); */ ++ } else ++ si_pid_clr_slow(sb); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* lock superblock. mainly for entry point functions */ ++/* ++ * __si_read_lock, __si_write_lock, ++ * __si_read_unlock, __si_write_unlock, __si_downgrade_lock ++ */ ++AuSimpleRwsemFuncs(__si, struct super_block *sb, &au_sbi(sb)->si_rwsem); ++ ++#define SiMustNoWaiters(sb) AuRwMustNoWaiters(&au_sbi(sb)->si_rwsem) ++#define SiMustAnyLock(sb) AuRwMustAnyLock(&au_sbi(sb)->si_rwsem) ++#define SiMustWriteLock(sb) AuRwMustWriteLock(&au_sbi(sb)->si_rwsem) ++ ++static inline void si_noflush_read_lock(struct super_block *sb) ++{ ++ __si_read_lock(sb); ++ si_pid_set(sb); ++} ++ ++static inline int si_noflush_read_trylock(struct super_block *sb) ++{ ++ int locked; ++ ++ locked = __si_read_trylock(sb); ++ if (locked) ++ si_pid_set(sb); ++ return locked; ++} ++ ++static inline void si_noflush_write_lock(struct super_block *sb) ++{ ++ __si_write_lock(sb); ++ si_pid_set(sb); ++} ++ ++static inline int si_noflush_write_trylock(struct super_block *sb) ++{ ++ int locked; ++ ++ locked = __si_write_trylock(sb); ++ if (locked) ++ si_pid_set(sb); ++ return locked; ++} ++ ++#if 0 /* reserved */ ++static inline int si_read_trylock(struct super_block *sb, int flags) ++{ ++ if (au_ftest_lock(flags, FLUSH)) ++ au_nwt_flush(&au_sbi(sb)->si_nowait); ++ return si_noflush_read_trylock(sb); ++} ++#endif ++ ++static inline void si_read_unlock(struct super_block *sb) ++{ ++ si_pid_clr(sb); ++ __si_read_unlock(sb); ++} ++ ++#if 0 /* reserved */ ++static inline int si_write_trylock(struct super_block *sb, int flags) ++{ ++ if (au_ftest_lock(flags, FLUSH)) ++ au_nwt_flush(&au_sbi(sb)->si_nowait); ++ return si_noflush_write_trylock(sb); ++} ++#endif ++ ++static inline void si_write_unlock(struct super_block *sb) ++{ ++ si_pid_clr(sb); ++ __si_write_unlock(sb); ++} ++ ++#if 0 /* reserved */ ++static inline void si_downgrade_lock(struct super_block *sb) ++{ ++ __si_downgrade_lock(sb); ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline aufs_bindex_t au_sbend(struct super_block *sb) ++{ ++ SiMustAnyLock(sb); ++ return au_sbi(sb)->si_bend; ++} ++ ++static inline unsigned int au_mntflags(struct super_block *sb) ++{ ++ SiMustAnyLock(sb); ++ return au_sbi(sb)->si_mntflags; ++} ++ ++static inline unsigned int au_sigen(struct super_block *sb) ++{ ++ SiMustAnyLock(sb); ++ return au_sbi(sb)->si_generation; ++} ++ ++static inline void au_ninodes_inc(struct super_block *sb) ++{ ++ atomic_long_inc(&au_sbi(sb)->si_ninodes); ++} ++ ++static inline void au_ninodes_dec(struct super_block *sb) ++{ ++ AuDebugOn(!atomic_long_read(&au_sbi(sb)->si_ninodes)); ++ atomic_long_dec(&au_sbi(sb)->si_ninodes); ++} ++ ++static inline void au_nfiles_inc(struct super_block *sb) ++{ ++ atomic_long_inc(&au_sbi(sb)->si_nfiles); ++} ++ ++static inline void au_nfiles_dec(struct super_block *sb) ++{ ++ AuDebugOn(!atomic_long_read(&au_sbi(sb)->si_nfiles)); ++ atomic_long_dec(&au_sbi(sb)->si_nfiles); ++} ++ ++static inline struct au_branch *au_sbr(struct super_block *sb, ++ aufs_bindex_t bindex) ++{ ++ SiMustAnyLock(sb); ++ return au_sbi(sb)->si_branch[0 + bindex]; ++} ++ ++static inline void au_xino_brid_set(struct super_block *sb, aufs_bindex_t brid) ++{ ++ SiMustWriteLock(sb); ++ au_sbi(sb)->si_xino_brid = brid; ++} ++ ++static inline aufs_bindex_t au_xino_brid(struct super_block *sb) ++{ ++ SiMustAnyLock(sb); ++ return au_sbi(sb)->si_xino_brid; ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_SUPER_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/sysaufs.c linux-4.1.10/fs/aufs/sysaufs.c +--- linux-4.1.10.orig/fs/aufs/sysaufs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/sysaufs.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,104 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sysfs interface and lifetime management ++ * they are necessary regardless sysfs is disabled. ++ */ ++ ++#include <linux/random.h> ++#include "aufs.h" ++ ++unsigned long sysaufs_si_mask; ++struct kset *sysaufs_kset; ++ ++#define AuSiAttr(_name) { \ ++ .attr = { .name = __stringify(_name), .mode = 0444 }, \ ++ .show = sysaufs_si_##_name, \ ++} ++ ++static struct sysaufs_si_attr sysaufs_si_attr_xi_path = AuSiAttr(xi_path); ++struct attribute *sysaufs_si_attrs[] = { ++ &sysaufs_si_attr_xi_path.attr, ++ NULL, ++}; ++ ++static const struct sysfs_ops au_sbi_ops = { ++ .show = sysaufs_si_show ++}; ++ ++static struct kobj_type au_sbi_ktype = { ++ .release = au_si_free, ++ .sysfs_ops = &au_sbi_ops, ++ .default_attrs = sysaufs_si_attrs ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++int sysaufs_si_init(struct au_sbinfo *sbinfo) ++{ ++ int err; ++ ++ sbinfo->si_kobj.kset = sysaufs_kset; ++ /* cf. sysaufs_name() */ ++ err = kobject_init_and_add ++ (&sbinfo->si_kobj, &au_sbi_ktype, /*&sysaufs_kset->kobj*/NULL, ++ SysaufsSiNamePrefix "%lx", sysaufs_si_id(sbinfo)); ++ ++ dbgaufs_si_null(sbinfo); ++ if (!err) { ++ err = dbgaufs_si_init(sbinfo); ++ if (unlikely(err)) ++ kobject_put(&sbinfo->si_kobj); ++ } ++ return err; ++} ++ ++void sysaufs_fin(void) ++{ ++ dbgaufs_fin(); ++ sysfs_remove_group(&sysaufs_kset->kobj, sysaufs_attr_group); ++ kset_unregister(sysaufs_kset); ++} ++ ++int __init sysaufs_init(void) ++{ ++ int err; ++ ++ do { ++ get_random_bytes(&sysaufs_si_mask, sizeof(sysaufs_si_mask)); ++ } while (!sysaufs_si_mask); ++ ++ err = -EINVAL; ++ sysaufs_kset = kset_create_and_add(AUFS_NAME, NULL, fs_kobj); ++ if (unlikely(!sysaufs_kset)) ++ goto out; ++ err = PTR_ERR(sysaufs_kset); ++ if (IS_ERR(sysaufs_kset)) ++ goto out; ++ err = sysfs_create_group(&sysaufs_kset->kobj, sysaufs_attr_group); ++ if (unlikely(err)) { ++ kset_unregister(sysaufs_kset); ++ goto out; ++ } ++ ++ err = dbgaufs_init(); ++ if (unlikely(err)) ++ sysaufs_fin(); ++out: ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/sysaufs.h linux-4.1.10/fs/aufs/sysaufs.h +--- linux-4.1.10.orig/fs/aufs/sysaufs.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/sysaufs.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,101 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sysfs interface and mount lifetime management ++ */ ++ ++#ifndef __SYSAUFS_H__ ++#define __SYSAUFS_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/sysfs.h> ++#include "module.h" ++ ++struct super_block; ++struct au_sbinfo; ++ ++struct sysaufs_si_attr { ++ struct attribute attr; ++ int (*show)(struct seq_file *seq, struct super_block *sb); ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* sysaufs.c */ ++extern unsigned long sysaufs_si_mask; ++extern struct kset *sysaufs_kset; ++extern struct attribute *sysaufs_si_attrs[]; ++int sysaufs_si_init(struct au_sbinfo *sbinfo); ++int __init sysaufs_init(void); ++void sysaufs_fin(void); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* some people doesn't like to show a pointer in kernel */ ++static inline unsigned long sysaufs_si_id(struct au_sbinfo *sbinfo) ++{ ++ return sysaufs_si_mask ^ (unsigned long)sbinfo; ++} ++ ++#define SysaufsSiNamePrefix "si_" ++#define SysaufsSiNameLen (sizeof(SysaufsSiNamePrefix) + 16) ++static inline void sysaufs_name(struct au_sbinfo *sbinfo, char *name) ++{ ++ snprintf(name, SysaufsSiNameLen, SysaufsSiNamePrefix "%lx", ++ sysaufs_si_id(sbinfo)); ++} ++ ++struct au_branch; ++#ifdef CONFIG_SYSFS ++/* sysfs.c */ ++extern struct attribute_group *sysaufs_attr_group; ++ ++int sysaufs_si_xi_path(struct seq_file *seq, struct super_block *sb); ++ssize_t sysaufs_si_show(struct kobject *kobj, struct attribute *attr, ++ char *buf); ++long au_brinfo_ioctl(struct file *file, unsigned long arg); ++#ifdef CONFIG_COMPAT ++long au_brinfo_compat_ioctl(struct file *file, unsigned long arg); ++#endif ++ ++void sysaufs_br_init(struct au_branch *br); ++void sysaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex); ++void sysaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex); ++ ++#define sysaufs_brs_init() do {} while (0) ++ ++#else ++#define sysaufs_attr_group NULL ++ ++AuStubInt0(sysaufs_si_xi_path, struct seq_file *seq, struct super_block *sb) ++AuStub(ssize_t, sysaufs_si_show, return 0, struct kobject *kobj, ++ struct attribute *attr, char *buf) ++AuStubVoid(sysaufs_br_init, struct au_branch *br) ++AuStubVoid(sysaufs_brs_add, struct super_block *sb, aufs_bindex_t bindex) ++AuStubVoid(sysaufs_brs_del, struct super_block *sb, aufs_bindex_t bindex) ++ ++static inline void sysaufs_brs_init(void) ++{ ++ sysaufs_brs = 0; ++} ++ ++#endif /* CONFIG_SYSFS */ ++ ++#endif /* __KERNEL__ */ ++#endif /* __SYSAUFS_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/sysfs.c linux-4.1.10/fs/aufs/sysfs.c +--- linux-4.1.10.orig/fs/aufs/sysfs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/sysfs.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,376 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sysfs interface ++ */ ++ ++#include <linux/compat.h> ++#include <linux/seq_file.h> ++#include "aufs.h" ++ ++#ifdef CONFIG_AUFS_FS_MODULE ++/* this entry violates the "one line per file" policy of sysfs */ ++static ssize_t config_show(struct kobject *kobj, struct kobj_attribute *attr, ++ char *buf) ++{ ++ ssize_t err; ++ static char *conf = ++/* this file is generated at compiling */ ++#include "conf.str" ++ ; ++ ++ err = snprintf(buf, PAGE_SIZE, conf); ++ if (unlikely(err >= PAGE_SIZE)) ++ err = -EFBIG; ++ return err; ++} ++ ++static struct kobj_attribute au_config_attr = __ATTR_RO(config); ++#endif ++ ++static struct attribute *au_attr[] = { ++#ifdef CONFIG_AUFS_FS_MODULE ++ &au_config_attr.attr, ++#endif ++ NULL, /* need to NULL terminate the list of attributes */ ++}; ++ ++static struct attribute_group sysaufs_attr_group_body = { ++ .attrs = au_attr ++}; ++ ++struct attribute_group *sysaufs_attr_group = &sysaufs_attr_group_body; ++ ++/* ---------------------------------------------------------------------- */ ++ ++int sysaufs_si_xi_path(struct seq_file *seq, struct super_block *sb) ++{ ++ int err; ++ ++ SiMustAnyLock(sb); ++ ++ err = 0; ++ if (au_opt_test(au_mntflags(sb), XINO)) { ++ err = au_xino_path(seq, au_sbi(sb)->si_xib); ++ seq_putc(seq, '\n'); ++ } ++ return err; ++} ++ ++/* ++ * the lifetime of branch is independent from the entry under sysfs. ++ * sysfs handles the lifetime of the entry, and never call ->show() after it is ++ * unlinked. ++ */ ++static int sysaufs_si_br(struct seq_file *seq, struct super_block *sb, ++ aufs_bindex_t bindex, int idx) ++{ ++ int err; ++ struct path path; ++ struct dentry *root; ++ struct au_branch *br; ++ au_br_perm_str_t perm; ++ ++ AuDbg("b%d\n", bindex); ++ ++ err = 0; ++ root = sb->s_root; ++ di_read_lock_parent(root, !AuLock_IR); ++ br = au_sbr(sb, bindex); ++ ++ switch (idx) { ++ case AuBrSysfs_BR: ++ path.mnt = au_br_mnt(br); ++ path.dentry = au_h_dptr(root, bindex); ++ err = au_seq_path(seq, &path); ++ if (!err) { ++ au_optstr_br_perm(&perm, br->br_perm); ++ err = seq_printf(seq, "=%s\n", perm.a); ++ } ++ break; ++ case AuBrSysfs_BRID: ++ err = seq_printf(seq, "%d\n", br->br_id); ++ break; ++ } ++ di_read_unlock(root, !AuLock_IR); ++ if (err == -1) ++ err = -E2BIG; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static struct seq_file *au_seq(char *p, ssize_t len) ++{ ++ struct seq_file *seq; ++ ++ seq = kzalloc(sizeof(*seq), GFP_NOFS); ++ if (seq) { ++ /* mutex_init(&seq.lock); */ ++ seq->buf = p; ++ seq->size = len; ++ return seq; /* success */ ++ } ++ ++ seq = ERR_PTR(-ENOMEM); ++ return seq; ++} ++ ++#define SysaufsBr_PREFIX "br" ++#define SysaufsBrid_PREFIX "brid" ++ ++/* todo: file size may exceed PAGE_SIZE */ ++ssize_t sysaufs_si_show(struct kobject *kobj, struct attribute *attr, ++ char *buf) ++{ ++ ssize_t err; ++ int idx; ++ long l; ++ aufs_bindex_t bend; ++ struct au_sbinfo *sbinfo; ++ struct super_block *sb; ++ struct seq_file *seq; ++ char *name; ++ struct attribute **cattr; ++ ++ sbinfo = container_of(kobj, struct au_sbinfo, si_kobj); ++ sb = sbinfo->si_sb; ++ ++ /* ++ * prevent a race condition between sysfs and aufs. ++ * for instance, sysfs_file_read() calls sysfs_get_active_two() which ++ * prohibits maintaining the sysfs entries. ++ * hew we acquire read lock after sysfs_get_active_two(). ++ * on the other hand, the remount process may maintain the sysfs/aufs ++ * entries after acquiring write lock. ++ * it can cause a deadlock. ++ * simply we gave up processing read here. ++ */ ++ err = -EBUSY; ++ if (unlikely(!si_noflush_read_trylock(sb))) ++ goto out; ++ ++ seq = au_seq(buf, PAGE_SIZE); ++ err = PTR_ERR(seq); ++ if (IS_ERR(seq)) ++ goto out_unlock; ++ ++ name = (void *)attr->name; ++ cattr = sysaufs_si_attrs; ++ while (*cattr) { ++ if (!strcmp(name, (*cattr)->name)) { ++ err = container_of(*cattr, struct sysaufs_si_attr, attr) ++ ->show(seq, sb); ++ goto out_seq; ++ } ++ cattr++; ++ } ++ ++ if (!strncmp(name, SysaufsBrid_PREFIX, ++ sizeof(SysaufsBrid_PREFIX) - 1)) { ++ idx = AuBrSysfs_BRID; ++ name += sizeof(SysaufsBrid_PREFIX) - 1; ++ } else if (!strncmp(name, SysaufsBr_PREFIX, ++ sizeof(SysaufsBr_PREFIX) - 1)) { ++ idx = AuBrSysfs_BR; ++ name += sizeof(SysaufsBr_PREFIX) - 1; ++ } else ++ BUG(); ++ ++ err = kstrtol(name, 10, &l); ++ if (!err) { ++ bend = au_sbend(sb); ++ if (l <= bend) ++ err = sysaufs_si_br(seq, sb, (aufs_bindex_t)l, idx); ++ else ++ err = -ENOENT; ++ } ++ ++out_seq: ++ if (!err) { ++ err = seq->count; ++ /* sysfs limit */ ++ if (unlikely(err == PAGE_SIZE)) ++ err = -EFBIG; ++ } ++ kfree(seq); ++out_unlock: ++ si_read_unlock(sb); ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_brinfo(struct super_block *sb, union aufs_brinfo __user *arg) ++{ ++ int err; ++ int16_t brid; ++ aufs_bindex_t bindex, bend; ++ size_t sz; ++ char *buf; ++ struct seq_file *seq; ++ struct au_branch *br; ++ ++ si_read_lock(sb, AuLock_FLUSH); ++ bend = au_sbend(sb); ++ err = bend + 1; ++ if (!arg) ++ goto out; ++ ++ err = -ENOMEM; ++ buf = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!buf)) ++ goto out; ++ ++ seq = au_seq(buf, PAGE_SIZE); ++ err = PTR_ERR(seq); ++ if (IS_ERR(seq)) ++ goto out_buf; ++ ++ sz = sizeof(*arg) - offsetof(union aufs_brinfo, path); ++ for (bindex = 0; bindex <= bend; bindex++, arg++) { ++ err = !access_ok(VERIFY_WRITE, arg, sizeof(*arg)); ++ if (unlikely(err)) ++ break; ++ ++ br = au_sbr(sb, bindex); ++ brid = br->br_id; ++ BUILD_BUG_ON(sizeof(brid) != sizeof(arg->id)); ++ err = __put_user(brid, &arg->id); ++ if (unlikely(err)) ++ break; ++ ++ BUILD_BUG_ON(sizeof(br->br_perm) != sizeof(arg->perm)); ++ err = __put_user(br->br_perm, &arg->perm); ++ if (unlikely(err)) ++ break; ++ ++ err = au_seq_path(seq, &br->br_path); ++ if (unlikely(err)) ++ break; ++ err = seq_putc(seq, '\0'); ++ if (!err && seq->count <= sz) { ++ err = copy_to_user(arg->path, seq->buf, seq->count); ++ seq->count = 0; ++ if (unlikely(err)) ++ break; ++ } else { ++ err = -E2BIG; ++ goto out_seq; ++ } ++ } ++ if (unlikely(err)) ++ err = -EFAULT; ++ ++out_seq: ++ kfree(seq); ++out_buf: ++ free_page((unsigned long)buf); ++out: ++ si_read_unlock(sb); ++ return err; ++} ++ ++long au_brinfo_ioctl(struct file *file, unsigned long arg) ++{ ++ return au_brinfo(file->f_path.dentry->d_sb, (void __user *)arg); ++} ++ ++#ifdef CONFIG_COMPAT ++long au_brinfo_compat_ioctl(struct file *file, unsigned long arg) ++{ ++ return au_brinfo(file->f_path.dentry->d_sb, compat_ptr(arg)); ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++void sysaufs_br_init(struct au_branch *br) ++{ ++ int i; ++ struct au_brsysfs *br_sysfs; ++ struct attribute *attr; ++ ++ br_sysfs = br->br_sysfs; ++ for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { ++ attr = &br_sysfs->attr; ++ sysfs_attr_init(attr); ++ attr->name = br_sysfs->name; ++ attr->mode = S_IRUGO; ++ br_sysfs++; ++ } ++} ++ ++void sysaufs_brs_del(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ struct au_branch *br; ++ struct kobject *kobj; ++ struct au_brsysfs *br_sysfs; ++ int i; ++ aufs_bindex_t bend; ++ ++ dbgaufs_brs_del(sb, bindex); ++ ++ if (!sysaufs_brs) ++ return; ++ ++ kobj = &au_sbi(sb)->si_kobj; ++ bend = au_sbend(sb); ++ for (; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ br_sysfs = br->br_sysfs; ++ for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { ++ sysfs_remove_file(kobj, &br_sysfs->attr); ++ br_sysfs++; ++ } ++ } ++} ++ ++void sysaufs_brs_add(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ int err, i; ++ aufs_bindex_t bend; ++ struct kobject *kobj; ++ struct au_branch *br; ++ struct au_brsysfs *br_sysfs; ++ ++ dbgaufs_brs_add(sb, bindex); ++ ++ if (!sysaufs_brs) ++ return; ++ ++ kobj = &au_sbi(sb)->si_kobj; ++ bend = au_sbend(sb); ++ for (; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ br_sysfs = br->br_sysfs; ++ snprintf(br_sysfs[AuBrSysfs_BR].name, sizeof(br_sysfs->name), ++ SysaufsBr_PREFIX "%d", bindex); ++ snprintf(br_sysfs[AuBrSysfs_BRID].name, sizeof(br_sysfs->name), ++ SysaufsBrid_PREFIX "%d", bindex); ++ for (i = 0; i < ARRAY_SIZE(br->br_sysfs); i++) { ++ err = sysfs_create_file(kobj, &br_sysfs->attr); ++ if (unlikely(err)) ++ pr_warn("failed %s under sysfs(%d)\n", ++ br_sysfs->name, err); ++ br_sysfs++; ++ } ++ } ++} +diff -Nur linux-4.1.10.orig/fs/aufs/sysrq.c linux-4.1.10/fs/aufs/sysrq.c +--- linux-4.1.10.orig/fs/aufs/sysrq.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/sysrq.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,157 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * magic sysrq hanlder ++ */ ++ ++/* #include <linux/sysrq.h> */ ++#include <linux/writeback.h> ++#include "aufs.h" ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void sysrq_sb(struct super_block *sb) ++{ ++ char *plevel; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ struct au_sphlhead *files; ++ struct au_finfo *finfo; ++ ++ plevel = au_plevel; ++ au_plevel = KERN_WARNING; ++ ++ /* since we define pr_fmt, call printk directly */ ++#define pr(str) printk(KERN_WARNING AUFS_NAME ": " str) ++ ++ sbinfo = au_sbi(sb); ++ printk(KERN_WARNING "si=%lx\n", sysaufs_si_id(sbinfo)); ++ pr("superblock\n"); ++ au_dpri_sb(sb); ++ ++#if 0 ++ pr("root dentry\n"); ++ au_dpri_dentry(sb->s_root); ++ pr("root inode\n"); ++ au_dpri_inode(d_inode(sb->s_root)); ++#endif ++ ++#if 0 ++ do { ++ int err, i, j, ndentry; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ ++ err = au_dpages_init(&dpages, GFP_ATOMIC); ++ if (unlikely(err)) ++ break; ++ err = au_dcsub_pages(&dpages, sb->s_root, NULL, NULL); ++ if (!err) ++ for (i = 0; i < dpages.ndpage; i++) { ++ dpage = dpages.dpages + i; ++ ndentry = dpage->ndentry; ++ for (j = 0; j < ndentry; j++) ++ au_dpri_dentry(dpage->dentries[j]); ++ } ++ au_dpages_free(&dpages); ++ } while (0); ++#endif ++ ++#if 1 ++ { ++ struct inode *i; ++ ++ pr("isolated inode\n"); ++ spin_lock(&inode_sb_list_lock); ++ list_for_each_entry(i, &sb->s_inodes, i_sb_list) { ++ spin_lock(&i->i_lock); ++ if (1 || hlist_empty(&i->i_dentry)) ++ au_dpri_inode(i); ++ spin_unlock(&i->i_lock); ++ } ++ spin_unlock(&inode_sb_list_lock); ++ } ++#endif ++ pr("files\n"); ++ files = &au_sbi(sb)->si_files; ++ spin_lock(&files->spin); ++ hlist_for_each_entry(finfo, &files->head, fi_hlist) { ++ umode_t mode; ++ ++ file = finfo->fi_file; ++ mode = file_inode(file)->i_mode; ++ if (!special_file(mode)) ++ au_dpri_file(file); ++ } ++ spin_unlock(&files->spin); ++ pr("done\n"); ++ ++#undef pr ++ au_plevel = plevel; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* module parameter */ ++static char *aufs_sysrq_key = "a"; ++module_param_named(sysrq, aufs_sysrq_key, charp, S_IRUGO); ++MODULE_PARM_DESC(sysrq, "MagicSysRq key for " AUFS_NAME); ++ ++static void au_sysrq(int key __maybe_unused) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ lockdep_off(); ++ au_sbilist_lock(); ++ list_for_each_entry(sbinfo, &au_sbilist.head, si_list) ++ sysrq_sb(sbinfo->si_sb); ++ au_sbilist_unlock(); ++ lockdep_on(); ++} ++ ++static struct sysrq_key_op au_sysrq_op = { ++ .handler = au_sysrq, ++ .help_msg = "Aufs", ++ .action_msg = "Aufs", ++ .enable_mask = SYSRQ_ENABLE_DUMP ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++int __init au_sysrq_init(void) ++{ ++ int err; ++ char key; ++ ++ err = -1; ++ key = *aufs_sysrq_key; ++ if ('a' <= key && key <= 'z') ++ err = register_sysrq_key(key, &au_sysrq_op); ++ if (unlikely(err)) ++ pr_err("err %d, sysrq=%c\n", err, key); ++ return err; ++} ++ ++void au_sysrq_fin(void) ++{ ++ int err; ++ ++ err = unregister_sysrq_key(*aufs_sysrq_key, &au_sysrq_op); ++ if (unlikely(err)) ++ pr_err("err %d (ignored)\n", err); ++} +diff -Nur linux-4.1.10.orig/fs/aufs/vdir.c linux-4.1.10/fs/aufs/vdir.c +--- linux-4.1.10.orig/fs/aufs/vdir.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/vdir.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,888 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * virtual or vertical directory ++ */ ++ ++#include "aufs.h" ++ ++static unsigned int calc_size(int nlen) ++{ ++ return ALIGN(sizeof(struct au_vdir_de) + nlen, sizeof(ino_t)); ++} ++ ++static int set_deblk_end(union au_vdir_deblk_p *p, ++ union au_vdir_deblk_p *deblk_end) ++{ ++ if (calc_size(0) <= deblk_end->deblk - p->deblk) { ++ p->de->de_str.len = 0; ++ /* smp_mb(); */ ++ return 0; ++ } ++ return -1; /* error */ ++} ++ ++/* returns true or false */ ++static int is_deblk_end(union au_vdir_deblk_p *p, ++ union au_vdir_deblk_p *deblk_end) ++{ ++ if (calc_size(0) <= deblk_end->deblk - p->deblk) ++ return !p->de->de_str.len; ++ return 1; ++} ++ ++static unsigned char *last_deblk(struct au_vdir *vdir) ++{ ++ return vdir->vd_deblk[vdir->vd_nblk - 1]; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* estimate the appropriate size for name hash table */ ++unsigned int au_rdhash_est(loff_t sz) ++{ ++ unsigned int n; ++ ++ n = UINT_MAX; ++ sz >>= 10; ++ if (sz < n) ++ n = sz; ++ if (sz < AUFS_RDHASH_DEF) ++ n = AUFS_RDHASH_DEF; ++ /* pr_info("n %u\n", n); */ ++ return n; ++} ++ ++/* ++ * the allocated memory has to be freed by ++ * au_nhash_wh_free() or au_nhash_de_free(). ++ */ ++int au_nhash_alloc(struct au_nhash *nhash, unsigned int num_hash, gfp_t gfp) ++{ ++ struct hlist_head *head; ++ unsigned int u; ++ size_t sz; ++ ++ sz = sizeof(*nhash->nh_head) * num_hash; ++ head = kmalloc(sz, gfp); ++ if (head) { ++ nhash->nh_num = num_hash; ++ nhash->nh_head = head; ++ for (u = 0; u < num_hash; u++) ++ INIT_HLIST_HEAD(head++); ++ return 0; /* success */ ++ } ++ ++ return -ENOMEM; ++} ++ ++static void nhash_count(struct hlist_head *head) ++{ ++#if 0 ++ unsigned long n; ++ struct hlist_node *pos; ++ ++ n = 0; ++ hlist_for_each(pos, head) ++ n++; ++ pr_info("%lu\n", n); ++#endif ++} ++ ++static void au_nhash_wh_do_free(struct hlist_head *head) ++{ ++ struct au_vdir_wh *pos; ++ struct hlist_node *node; ++ ++ hlist_for_each_entry_safe(pos, node, head, wh_hash) ++ kfree(pos); ++} ++ ++static void au_nhash_de_do_free(struct hlist_head *head) ++{ ++ struct au_vdir_dehstr *pos; ++ struct hlist_node *node; ++ ++ hlist_for_each_entry_safe(pos, node, head, hash) ++ au_cache_free_vdir_dehstr(pos); ++} ++ ++static void au_nhash_do_free(struct au_nhash *nhash, ++ void (*free)(struct hlist_head *head)) ++{ ++ unsigned int n; ++ struct hlist_head *head; ++ ++ n = nhash->nh_num; ++ if (!n) ++ return; ++ ++ head = nhash->nh_head; ++ while (n-- > 0) { ++ nhash_count(head); ++ free(head++); ++ } ++ kfree(nhash->nh_head); ++} ++ ++void au_nhash_wh_free(struct au_nhash *whlist) ++{ ++ au_nhash_do_free(whlist, au_nhash_wh_do_free); ++} ++ ++static void au_nhash_de_free(struct au_nhash *delist) ++{ ++ au_nhash_do_free(delist, au_nhash_de_do_free); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_nhash_test_longer_wh(struct au_nhash *whlist, aufs_bindex_t btgt, ++ int limit) ++{ ++ int num; ++ unsigned int u, n; ++ struct hlist_head *head; ++ struct au_vdir_wh *pos; ++ ++ num = 0; ++ n = whlist->nh_num; ++ head = whlist->nh_head; ++ for (u = 0; u < n; u++, head++) ++ hlist_for_each_entry(pos, head, wh_hash) ++ if (pos->wh_bindex == btgt && ++num > limit) ++ return 1; ++ return 0; ++} ++ ++static struct hlist_head *au_name_hash(struct au_nhash *nhash, ++ unsigned char *name, ++ unsigned int len) ++{ ++ unsigned int v; ++ /* const unsigned int magic_bit = 12; */ ++ ++ AuDebugOn(!nhash->nh_num || !nhash->nh_head); ++ ++ v = 0; ++ while (len--) ++ v += *name++; ++ /* v = hash_long(v, magic_bit); */ ++ v %= nhash->nh_num; ++ return nhash->nh_head + v; ++} ++ ++static int au_nhash_test_name(struct au_vdir_destr *str, const char *name, ++ int nlen) ++{ ++ return str->len == nlen && !memcmp(str->name, name, nlen); ++} ++ ++/* returns found or not */ ++int au_nhash_test_known_wh(struct au_nhash *whlist, char *name, int nlen) ++{ ++ struct hlist_head *head; ++ struct au_vdir_wh *pos; ++ struct au_vdir_destr *str; ++ ++ head = au_name_hash(whlist, name, nlen); ++ hlist_for_each_entry(pos, head, wh_hash) { ++ str = &pos->wh_str; ++ AuDbg("%.*s\n", str->len, str->name); ++ if (au_nhash_test_name(str, name, nlen)) ++ return 1; ++ } ++ return 0; ++} ++ ++/* returns found(true) or not */ ++static int test_known(struct au_nhash *delist, char *name, int nlen) ++{ ++ struct hlist_head *head; ++ struct au_vdir_dehstr *pos; ++ struct au_vdir_destr *str; ++ ++ head = au_name_hash(delist, name, nlen); ++ hlist_for_each_entry(pos, head, hash) { ++ str = pos->str; ++ AuDbg("%.*s\n", str->len, str->name); ++ if (au_nhash_test_name(str, name, nlen)) ++ return 1; ++ } ++ return 0; ++} ++ ++static void au_shwh_init_wh(struct au_vdir_wh *wh, ino_t ino, ++ unsigned char d_type) ++{ ++#ifdef CONFIG_AUFS_SHWH ++ wh->wh_ino = ino; ++ wh->wh_type = d_type; ++#endif ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_nhash_append_wh(struct au_nhash *whlist, char *name, int nlen, ino_t ino, ++ unsigned int d_type, aufs_bindex_t bindex, ++ unsigned char shwh) ++{ ++ int err; ++ struct au_vdir_destr *str; ++ struct au_vdir_wh *wh; ++ ++ AuDbg("%.*s\n", nlen, name); ++ AuDebugOn(!whlist->nh_num || !whlist->nh_head); ++ ++ err = -ENOMEM; ++ wh = kmalloc(sizeof(*wh) + nlen, GFP_NOFS); ++ if (unlikely(!wh)) ++ goto out; ++ ++ err = 0; ++ wh->wh_bindex = bindex; ++ if (shwh) ++ au_shwh_init_wh(wh, ino, d_type); ++ str = &wh->wh_str; ++ str->len = nlen; ++ memcpy(str->name, name, nlen); ++ hlist_add_head(&wh->wh_hash, au_name_hash(whlist, name, nlen)); ++ /* smp_mb(); */ ++ ++out: ++ return err; ++} ++ ++static int append_deblk(struct au_vdir *vdir) ++{ ++ int err; ++ unsigned long ul; ++ const unsigned int deblk_sz = vdir->vd_deblk_sz; ++ union au_vdir_deblk_p p, deblk_end; ++ unsigned char **o; ++ ++ err = -ENOMEM; ++ o = krealloc(vdir->vd_deblk, sizeof(*o) * (vdir->vd_nblk + 1), ++ GFP_NOFS); ++ if (unlikely(!o)) ++ goto out; ++ ++ vdir->vd_deblk = o; ++ p.deblk = kmalloc(deblk_sz, GFP_NOFS); ++ if (p.deblk) { ++ ul = vdir->vd_nblk++; ++ vdir->vd_deblk[ul] = p.deblk; ++ vdir->vd_last.ul = ul; ++ vdir->vd_last.p.deblk = p.deblk; ++ deblk_end.deblk = p.deblk + deblk_sz; ++ err = set_deblk_end(&p, &deblk_end); ++ } ++ ++out: ++ return err; ++} ++ ++static int append_de(struct au_vdir *vdir, char *name, int nlen, ino_t ino, ++ unsigned int d_type, struct au_nhash *delist) ++{ ++ int err; ++ unsigned int sz; ++ const unsigned int deblk_sz = vdir->vd_deblk_sz; ++ union au_vdir_deblk_p p, *room, deblk_end; ++ struct au_vdir_dehstr *dehstr; ++ ++ p.deblk = last_deblk(vdir); ++ deblk_end.deblk = p.deblk + deblk_sz; ++ room = &vdir->vd_last.p; ++ AuDebugOn(room->deblk < p.deblk || deblk_end.deblk <= room->deblk ++ || !is_deblk_end(room, &deblk_end)); ++ ++ sz = calc_size(nlen); ++ if (unlikely(sz > deblk_end.deblk - room->deblk)) { ++ err = append_deblk(vdir); ++ if (unlikely(err)) ++ goto out; ++ ++ p.deblk = last_deblk(vdir); ++ deblk_end.deblk = p.deblk + deblk_sz; ++ /* smp_mb(); */ ++ AuDebugOn(room->deblk != p.deblk); ++ } ++ ++ err = -ENOMEM; ++ dehstr = au_cache_alloc_vdir_dehstr(); ++ if (unlikely(!dehstr)) ++ goto out; ++ ++ dehstr->str = &room->de->de_str; ++ hlist_add_head(&dehstr->hash, au_name_hash(delist, name, nlen)); ++ room->de->de_ino = ino; ++ room->de->de_type = d_type; ++ room->de->de_str.len = nlen; ++ memcpy(room->de->de_str.name, name, nlen); ++ ++ err = 0; ++ room->deblk += sz; ++ if (unlikely(set_deblk_end(room, &deblk_end))) ++ err = append_deblk(vdir); ++ /* smp_mb(); */ ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_vdir_free(struct au_vdir *vdir) ++{ ++ unsigned char **deblk; ++ ++ deblk = vdir->vd_deblk; ++ while (vdir->vd_nblk--) ++ kfree(*deblk++); ++ kfree(vdir->vd_deblk); ++ au_cache_free_vdir(vdir); ++} ++ ++static struct au_vdir *alloc_vdir(struct file *file) ++{ ++ struct au_vdir *vdir; ++ struct super_block *sb; ++ int err; ++ ++ sb = file->f_path.dentry->d_sb; ++ SiMustAnyLock(sb); ++ ++ err = -ENOMEM; ++ vdir = au_cache_alloc_vdir(); ++ if (unlikely(!vdir)) ++ goto out; ++ ++ vdir->vd_deblk = kzalloc(sizeof(*vdir->vd_deblk), GFP_NOFS); ++ if (unlikely(!vdir->vd_deblk)) ++ goto out_free; ++ ++ vdir->vd_deblk_sz = au_sbi(sb)->si_rdblk; ++ if (!vdir->vd_deblk_sz) { ++ /* estimate the appropriate size for deblk */ ++ vdir->vd_deblk_sz = au_dir_size(file, /*dentry*/NULL); ++ /* pr_info("vd_deblk_sz %u\n", vdir->vd_deblk_sz); */ ++ } ++ vdir->vd_nblk = 0; ++ vdir->vd_version = 0; ++ vdir->vd_jiffy = 0; ++ err = append_deblk(vdir); ++ if (!err) ++ return vdir; /* success */ ++ ++ kfree(vdir->vd_deblk); ++ ++out_free: ++ au_cache_free_vdir(vdir); ++out: ++ vdir = ERR_PTR(err); ++ return vdir; ++} ++ ++static int reinit_vdir(struct au_vdir *vdir) ++{ ++ int err; ++ union au_vdir_deblk_p p, deblk_end; ++ ++ while (vdir->vd_nblk > 1) { ++ kfree(vdir->vd_deblk[vdir->vd_nblk - 1]); ++ /* vdir->vd_deblk[vdir->vd_nblk - 1] = NULL; */ ++ vdir->vd_nblk--; ++ } ++ p.deblk = vdir->vd_deblk[0]; ++ deblk_end.deblk = p.deblk + vdir->vd_deblk_sz; ++ err = set_deblk_end(&p, &deblk_end); ++ /* keep vd_dblk_sz */ ++ vdir->vd_last.ul = 0; ++ vdir->vd_last.p.deblk = vdir->vd_deblk[0]; ++ vdir->vd_version = 0; ++ vdir->vd_jiffy = 0; ++ /* smp_mb(); */ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define AuFillVdir_CALLED 1 ++#define AuFillVdir_WHABLE (1 << 1) ++#define AuFillVdir_SHWH (1 << 2) ++#define au_ftest_fillvdir(flags, name) ((flags) & AuFillVdir_##name) ++#define au_fset_fillvdir(flags, name) \ ++ do { (flags) |= AuFillVdir_##name; } while (0) ++#define au_fclr_fillvdir(flags, name) \ ++ do { (flags) &= ~AuFillVdir_##name; } while (0) ++ ++#ifndef CONFIG_AUFS_SHWH ++#undef AuFillVdir_SHWH ++#define AuFillVdir_SHWH 0 ++#endif ++ ++struct fillvdir_arg { ++ struct dir_context ctx; ++ struct file *file; ++ struct au_vdir *vdir; ++ struct au_nhash delist; ++ struct au_nhash whlist; ++ aufs_bindex_t bindex; ++ unsigned int flags; ++ int err; ++}; ++ ++static int fillvdir(struct dir_context *ctx, const char *__name, int nlen, ++ loff_t offset __maybe_unused, u64 h_ino, ++ unsigned int d_type) ++{ ++ struct fillvdir_arg *arg = container_of(ctx, struct fillvdir_arg, ctx); ++ char *name = (void *)__name; ++ struct super_block *sb; ++ ino_t ino; ++ const unsigned char shwh = !!au_ftest_fillvdir(arg->flags, SHWH); ++ ++ arg->err = 0; ++ sb = arg->file->f_path.dentry->d_sb; ++ au_fset_fillvdir(arg->flags, CALLED); ++ /* smp_mb(); */ ++ if (nlen <= AUFS_WH_PFX_LEN ++ || memcmp(name, AUFS_WH_PFX, AUFS_WH_PFX_LEN)) { ++ if (test_known(&arg->delist, name, nlen) ++ || au_nhash_test_known_wh(&arg->whlist, name, nlen)) ++ goto out; /* already exists or whiteouted */ ++ ++ arg->err = au_ino(sb, arg->bindex, h_ino, d_type, &ino); ++ if (!arg->err) { ++ if (unlikely(nlen > AUFS_MAX_NAMELEN)) ++ d_type = DT_UNKNOWN; ++ arg->err = append_de(arg->vdir, name, nlen, ino, ++ d_type, &arg->delist); ++ } ++ } else if (au_ftest_fillvdir(arg->flags, WHABLE)) { ++ name += AUFS_WH_PFX_LEN; ++ nlen -= AUFS_WH_PFX_LEN; ++ if (au_nhash_test_known_wh(&arg->whlist, name, nlen)) ++ goto out; /* already whiteouted */ ++ ++ if (shwh) ++ arg->err = au_wh_ino(sb, arg->bindex, h_ino, d_type, ++ &ino); ++ if (!arg->err) { ++ if (nlen <= AUFS_MAX_NAMELEN + AUFS_WH_PFX_LEN) ++ d_type = DT_UNKNOWN; ++ arg->err = au_nhash_append_wh ++ (&arg->whlist, name, nlen, ino, d_type, ++ arg->bindex, shwh); ++ } ++ } ++ ++out: ++ if (!arg->err) ++ arg->vdir->vd_jiffy = jiffies; ++ /* smp_mb(); */ ++ AuTraceErr(arg->err); ++ return arg->err; ++} ++ ++static int au_handle_shwh(struct super_block *sb, struct au_vdir *vdir, ++ struct au_nhash *whlist, struct au_nhash *delist) ++{ ++#ifdef CONFIG_AUFS_SHWH ++ int err; ++ unsigned int nh, u; ++ struct hlist_head *head; ++ struct au_vdir_wh *pos; ++ struct hlist_node *n; ++ char *p, *o; ++ struct au_vdir_destr *destr; ++ ++ AuDebugOn(!au_opt_test(au_mntflags(sb), SHWH)); ++ ++ err = -ENOMEM; ++ o = p = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!p)) ++ goto out; ++ ++ err = 0; ++ nh = whlist->nh_num; ++ memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); ++ p += AUFS_WH_PFX_LEN; ++ for (u = 0; u < nh; u++) { ++ head = whlist->nh_head + u; ++ hlist_for_each_entry_safe(pos, n, head, wh_hash) { ++ destr = &pos->wh_str; ++ memcpy(p, destr->name, destr->len); ++ err = append_de(vdir, o, destr->len + AUFS_WH_PFX_LEN, ++ pos->wh_ino, pos->wh_type, delist); ++ if (unlikely(err)) ++ break; ++ } ++ } ++ ++ free_page((unsigned long)o); ++ ++out: ++ AuTraceErr(err); ++ return err; ++#else ++ return 0; ++#endif ++} ++ ++static int au_do_read_vdir(struct fillvdir_arg *arg) ++{ ++ int err; ++ unsigned int rdhash; ++ loff_t offset; ++ aufs_bindex_t bend, bindex, bstart; ++ unsigned char shwh; ++ struct file *hf, *file; ++ struct super_block *sb; ++ ++ file = arg->file; ++ sb = file->f_path.dentry->d_sb; ++ SiMustAnyLock(sb); ++ ++ rdhash = au_sbi(sb)->si_rdhash; ++ if (!rdhash) ++ rdhash = au_rdhash_est(au_dir_size(file, /*dentry*/NULL)); ++ err = au_nhash_alloc(&arg->delist, rdhash, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ err = au_nhash_alloc(&arg->whlist, rdhash, GFP_NOFS); ++ if (unlikely(err)) ++ goto out_delist; ++ ++ err = 0; ++ arg->flags = 0; ++ shwh = 0; ++ if (au_opt_test(au_mntflags(sb), SHWH)) { ++ shwh = 1; ++ au_fset_fillvdir(arg->flags, SHWH); ++ } ++ bstart = au_fbstart(file); ++ bend = au_fbend_dir(file); ++ for (bindex = bstart; !err && bindex <= bend; bindex++) { ++ hf = au_hf_dir(file, bindex); ++ if (!hf) ++ continue; ++ ++ offset = vfsub_llseek(hf, 0, SEEK_SET); ++ err = offset; ++ if (unlikely(offset)) ++ break; ++ ++ arg->bindex = bindex; ++ au_fclr_fillvdir(arg->flags, WHABLE); ++ if (shwh ++ || (bindex != bend ++ && au_br_whable(au_sbr_perm(sb, bindex)))) ++ au_fset_fillvdir(arg->flags, WHABLE); ++ do { ++ arg->err = 0; ++ au_fclr_fillvdir(arg->flags, CALLED); ++ /* smp_mb(); */ ++ err = vfsub_iterate_dir(hf, &arg->ctx); ++ if (err >= 0) ++ err = arg->err; ++ } while (!err && au_ftest_fillvdir(arg->flags, CALLED)); ++ ++ /* ++ * dir_relax() may be good for concurrency, but aufs should not ++ * use it since it will cause a lockdep problem. ++ */ ++ } ++ ++ if (!err && shwh) ++ err = au_handle_shwh(sb, arg->vdir, &arg->whlist, &arg->delist); ++ ++ au_nhash_wh_free(&arg->whlist); ++ ++out_delist: ++ au_nhash_de_free(&arg->delist); ++out: ++ return err; ++} ++ ++static int read_vdir(struct file *file, int may_read) ++{ ++ int err; ++ unsigned long expire; ++ unsigned char do_read; ++ struct fillvdir_arg arg = { ++ .ctx = { ++ .actor = fillvdir ++ } ++ }; ++ struct inode *inode; ++ struct au_vdir *vdir, *allocated; ++ ++ err = 0; ++ inode = file_inode(file); ++ IMustLock(inode); ++ SiMustAnyLock(inode->i_sb); ++ ++ allocated = NULL; ++ do_read = 0; ++ expire = au_sbi(inode->i_sb)->si_rdcache; ++ vdir = au_ivdir(inode); ++ if (!vdir) { ++ do_read = 1; ++ vdir = alloc_vdir(file); ++ err = PTR_ERR(vdir); ++ if (IS_ERR(vdir)) ++ goto out; ++ err = 0; ++ allocated = vdir; ++ } else if (may_read ++ && (inode->i_version != vdir->vd_version ++ || time_after(jiffies, vdir->vd_jiffy + expire))) { ++ do_read = 1; ++ err = reinit_vdir(vdir); ++ if (unlikely(err)) ++ goto out; ++ } ++ ++ if (!do_read) ++ return 0; /* success */ ++ ++ arg.file = file; ++ arg.vdir = vdir; ++ err = au_do_read_vdir(&arg); ++ if (!err) { ++ /* file->f_pos = 0; */ /* todo: ctx->pos? */ ++ vdir->vd_version = inode->i_version; ++ vdir->vd_last.ul = 0; ++ vdir->vd_last.p.deblk = vdir->vd_deblk[0]; ++ if (allocated) ++ au_set_ivdir(inode, allocated); ++ } else if (allocated) ++ au_vdir_free(allocated); ++ ++out: ++ return err; ++} ++ ++static int copy_vdir(struct au_vdir *tgt, struct au_vdir *src) ++{ ++ int err, rerr; ++ unsigned long ul, n; ++ const unsigned int deblk_sz = src->vd_deblk_sz; ++ ++ AuDebugOn(tgt->vd_nblk != 1); ++ ++ err = -ENOMEM; ++ if (tgt->vd_nblk < src->vd_nblk) { ++ unsigned char **p; ++ ++ p = krealloc(tgt->vd_deblk, sizeof(*p) * src->vd_nblk, ++ GFP_NOFS); ++ if (unlikely(!p)) ++ goto out; ++ tgt->vd_deblk = p; ++ } ++ ++ if (tgt->vd_deblk_sz != deblk_sz) { ++ unsigned char *p; ++ ++ tgt->vd_deblk_sz = deblk_sz; ++ p = krealloc(tgt->vd_deblk[0], deblk_sz, GFP_NOFS); ++ if (unlikely(!p)) ++ goto out; ++ tgt->vd_deblk[0] = p; ++ } ++ memcpy(tgt->vd_deblk[0], src->vd_deblk[0], deblk_sz); ++ tgt->vd_version = src->vd_version; ++ tgt->vd_jiffy = src->vd_jiffy; ++ ++ n = src->vd_nblk; ++ for (ul = 1; ul < n; ul++) { ++ tgt->vd_deblk[ul] = kmemdup(src->vd_deblk[ul], deblk_sz, ++ GFP_NOFS); ++ if (unlikely(!tgt->vd_deblk[ul])) ++ goto out; ++ tgt->vd_nblk++; ++ } ++ tgt->vd_nblk = n; ++ tgt->vd_last.ul = tgt->vd_last.ul; ++ tgt->vd_last.p.deblk = tgt->vd_deblk[tgt->vd_last.ul]; ++ tgt->vd_last.p.deblk += src->vd_last.p.deblk ++ - src->vd_deblk[src->vd_last.ul]; ++ /* smp_mb(); */ ++ return 0; /* success */ ++ ++out: ++ rerr = reinit_vdir(tgt); ++ BUG_ON(rerr); ++ return err; ++} ++ ++int au_vdir_init(struct file *file) ++{ ++ int err; ++ struct inode *inode; ++ struct au_vdir *vdir_cache, *allocated; ++ ++ /* test file->f_pos here instead of ctx->pos */ ++ err = read_vdir(file, !file->f_pos); ++ if (unlikely(err)) ++ goto out; ++ ++ allocated = NULL; ++ vdir_cache = au_fvdir_cache(file); ++ if (!vdir_cache) { ++ vdir_cache = alloc_vdir(file); ++ err = PTR_ERR(vdir_cache); ++ if (IS_ERR(vdir_cache)) ++ goto out; ++ allocated = vdir_cache; ++ } else if (!file->f_pos && vdir_cache->vd_version != file->f_version) { ++ /* test file->f_pos here instead of ctx->pos */ ++ err = reinit_vdir(vdir_cache); ++ if (unlikely(err)) ++ goto out; ++ } else ++ return 0; /* success */ ++ ++ inode = file_inode(file); ++ err = copy_vdir(vdir_cache, au_ivdir(inode)); ++ if (!err) { ++ file->f_version = inode->i_version; ++ if (allocated) ++ au_set_fvdir_cache(file, allocated); ++ } else if (allocated) ++ au_vdir_free(allocated); ++ ++out: ++ return err; ++} ++ ++static loff_t calc_offset(struct au_vdir *vdir) ++{ ++ loff_t offset; ++ union au_vdir_deblk_p p; ++ ++ p.deblk = vdir->vd_deblk[vdir->vd_last.ul]; ++ offset = vdir->vd_last.p.deblk - p.deblk; ++ offset += vdir->vd_deblk_sz * vdir->vd_last.ul; ++ return offset; ++} ++ ++/* returns true or false */ ++static int seek_vdir(struct file *file, struct dir_context *ctx) ++{ ++ int valid; ++ unsigned int deblk_sz; ++ unsigned long ul, n; ++ loff_t offset; ++ union au_vdir_deblk_p p, deblk_end; ++ struct au_vdir *vdir_cache; ++ ++ valid = 1; ++ vdir_cache = au_fvdir_cache(file); ++ offset = calc_offset(vdir_cache); ++ AuDbg("offset %lld\n", offset); ++ if (ctx->pos == offset) ++ goto out; ++ ++ vdir_cache->vd_last.ul = 0; ++ vdir_cache->vd_last.p.deblk = vdir_cache->vd_deblk[0]; ++ if (!ctx->pos) ++ goto out; ++ ++ valid = 0; ++ deblk_sz = vdir_cache->vd_deblk_sz; ++ ul = div64_u64(ctx->pos, deblk_sz); ++ AuDbg("ul %lu\n", ul); ++ if (ul >= vdir_cache->vd_nblk) ++ goto out; ++ ++ n = vdir_cache->vd_nblk; ++ for (; ul < n; ul++) { ++ p.deblk = vdir_cache->vd_deblk[ul]; ++ deblk_end.deblk = p.deblk + deblk_sz; ++ offset = ul; ++ offset *= deblk_sz; ++ while (!is_deblk_end(&p, &deblk_end) && offset < ctx->pos) { ++ unsigned int l; ++ ++ l = calc_size(p.de->de_str.len); ++ offset += l; ++ p.deblk += l; ++ } ++ if (!is_deblk_end(&p, &deblk_end)) { ++ valid = 1; ++ vdir_cache->vd_last.ul = ul; ++ vdir_cache->vd_last.p = p; ++ break; ++ } ++ } ++ ++out: ++ /* smp_mb(); */ ++ AuTraceErr(!valid); ++ return valid; ++} ++ ++int au_vdir_fill_de(struct file *file, struct dir_context *ctx) ++{ ++ unsigned int l, deblk_sz; ++ union au_vdir_deblk_p deblk_end; ++ struct au_vdir *vdir_cache; ++ struct au_vdir_de *de; ++ ++ vdir_cache = au_fvdir_cache(file); ++ if (!seek_vdir(file, ctx)) ++ return 0; ++ ++ deblk_sz = vdir_cache->vd_deblk_sz; ++ while (1) { ++ deblk_end.deblk = vdir_cache->vd_deblk[vdir_cache->vd_last.ul]; ++ deblk_end.deblk += deblk_sz; ++ while (!is_deblk_end(&vdir_cache->vd_last.p, &deblk_end)) { ++ de = vdir_cache->vd_last.p.de; ++ AuDbg("%.*s, off%lld, i%lu, dt%d\n", ++ de->de_str.len, de->de_str.name, ctx->pos, ++ (unsigned long)de->de_ino, de->de_type); ++ if (unlikely(!dir_emit(ctx, de->de_str.name, ++ de->de_str.len, de->de_ino, ++ de->de_type))) { ++ /* todo: ignore the error caused by udba? */ ++ /* return err; */ ++ return 0; ++ } ++ ++ l = calc_size(de->de_str.len); ++ vdir_cache->vd_last.p.deblk += l; ++ ctx->pos += l; ++ } ++ if (vdir_cache->vd_last.ul < vdir_cache->vd_nblk - 1) { ++ vdir_cache->vd_last.ul++; ++ vdir_cache->vd_last.p.deblk ++ = vdir_cache->vd_deblk[vdir_cache->vd_last.ul]; ++ ctx->pos = deblk_sz * vdir_cache->vd_last.ul; ++ continue; ++ } ++ break; ++ } ++ ++ /* smp_mb(); */ ++ return 0; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/vfsub.c linux-4.1.10/fs/aufs/vfsub.c +--- linux-4.1.10.orig/fs/aufs/vfsub.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/vfsub.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,848 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sub-routines for VFS ++ */ ++ ++#include <linux/namei.h> ++#include <linux/security.h> ++#include <linux/splice.h> ++#include "aufs.h" ++ ++int vfsub_update_h_iattr(struct path *h_path, int *did) ++{ ++ int err; ++ struct kstat st; ++ struct super_block *h_sb; ++ ++ /* for remote fs, leave work for its getattr or d_revalidate */ ++ /* for bad i_attr fs, handle them in aufs_getattr() */ ++ /* still some fs may acquire i_mutex. we need to skip them */ ++ err = 0; ++ if (!did) ++ did = &err; ++ h_sb = h_path->dentry->d_sb; ++ *did = (!au_test_fs_remote(h_sb) && au_test_fs_refresh_iattr(h_sb)); ++ if (*did) ++ err = vfs_getattr(h_path, &st); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct file *vfsub_dentry_open(struct path *path, int flags) ++{ ++ struct file *file; ++ ++ file = dentry_open(path, flags /* | __FMODE_NONOTIFY */, ++ current_cred()); ++ if (!IS_ERR_OR_NULL(file) ++ && (file->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) ++ i_readcount_inc(d_inode(path->dentry)); ++ ++ return file; ++} ++ ++struct file *vfsub_filp_open(const char *path, int oflags, int mode) ++{ ++ struct file *file; ++ ++ lockdep_off(); ++ file = filp_open(path, ++ oflags /* | __FMODE_NONOTIFY */, ++ mode); ++ lockdep_on(); ++ if (IS_ERR(file)) ++ goto out; ++ vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ ++ ++out: ++ return file; ++} ++ ++/* ++ * Ideally this function should call VFS:do_last() in order to keep all its ++ * checkings. But it is very hard for aufs to regenerate several VFS internal ++ * structure such as nameidata. This is a second (or third) best approach. ++ * cf. linux/fs/namei.c:do_last(), lookup_open() and atomic_open(). ++ */ ++int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, ++ struct vfsub_aopen_args *args, struct au_branch *br) ++{ ++ int err; ++ struct file *file = args->file; ++ /* copied from linux/fs/namei.c:atomic_open() */ ++ struct dentry *const DENTRY_NOT_SET = (void *)-1UL; ++ ++ IMustLock(dir); ++ AuDebugOn(!dir->i_op->atomic_open); ++ ++ err = au_br_test_oflag(args->open_flag, br); ++ if (unlikely(err)) ++ goto out; ++ ++ args->file->f_path.dentry = DENTRY_NOT_SET; ++ args->file->f_path.mnt = au_br_mnt(br); ++ err = dir->i_op->atomic_open(dir, dentry, file, args->open_flag, ++ args->create_mode, args->opened); ++ if (err >= 0) { ++ /* some filesystems don't set FILE_CREATED while succeeded? */ ++ if (*args->opened & FILE_CREATED) ++ fsnotify_create(dir, dentry); ++ } else ++ goto out; ++ ++ ++ if (!err) { ++ /* todo: call VFS:may_open() here */ ++ err = open_check_o_direct(file); ++ /* todo: ima_file_check() too? */ ++ if (!err && (args->open_flag & __FMODE_EXEC)) ++ err = deny_write_access(file); ++ if (unlikely(err)) ++ /* note that the file is created and still opened */ ++ goto out; ++ } ++ ++ atomic_inc(&br->br_count); ++ fsnotify_open(file); ++ ++out: ++ return err; ++} ++ ++int vfsub_kern_path(const char *name, unsigned int flags, struct path *path) ++{ ++ int err; ++ ++ err = kern_path(name, flags, path); ++ if (!err && d_is_positive(path->dentry)) ++ vfsub_update_h_iattr(path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++struct dentry *vfsub_lookup_one_len(const char *name, struct dentry *parent, ++ int len) ++{ ++ struct path path = { ++ .mnt = NULL ++ }; ++ ++ /* VFS checks it too, but by WARN_ON_ONCE() */ ++ IMustLock(d_inode(parent)); ++ ++ path.dentry = lookup_one_len(name, parent, len); ++ if (IS_ERR(path.dentry)) ++ goto out; ++ if (d_is_positive(path.dentry)) ++ vfsub_update_h_iattr(&path, /*did*/NULL); /*ignore*/ ++ ++out: ++ AuTraceErrPtr(path.dentry); ++ return path.dentry; ++} ++ ++void vfsub_call_lkup_one(void *args) ++{ ++ struct vfsub_lkup_one_args *a = args; ++ *a->errp = vfsub_lkup_one(a->name, a->parent); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct dentry *vfsub_lock_rename(struct dentry *d1, struct au_hinode *hdir1, ++ struct dentry *d2, struct au_hinode *hdir2) ++{ ++ struct dentry *d; ++ ++ lockdep_off(); ++ d = lock_rename(d1, d2); ++ lockdep_on(); ++ au_hn_suspend(hdir1); ++ if (hdir1 != hdir2) ++ au_hn_suspend(hdir2); ++ ++ return d; ++} ++ ++void vfsub_unlock_rename(struct dentry *d1, struct au_hinode *hdir1, ++ struct dentry *d2, struct au_hinode *hdir2) ++{ ++ au_hn_resume(hdir1); ++ if (hdir1 != hdir2) ++ au_hn_resume(hdir2); ++ lockdep_off(); ++ unlock_rename(d1, d2); ++ lockdep_on(); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int vfsub_create(struct inode *dir, struct path *path, int mode, bool want_excl) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_mknod(path, d, mode, 0); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_create(dir, path->dentry, mode, want_excl); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = *path; ++ int did; ++ ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = path->dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++int vfsub_symlink(struct inode *dir, struct path *path, const char *symname) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_symlink(path, d, symname); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_symlink(dir, path->dentry, symname); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = *path; ++ int did; ++ ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = path->dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++int vfsub_mknod(struct inode *dir, struct path *path, int mode, dev_t dev) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_mknod(path, d, mode, new_encode_dev(dev)); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_mknod(dir, path->dentry, mode, dev); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = *path; ++ int did; ++ ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = path->dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++static int au_test_nlink(struct inode *inode) ++{ ++ const unsigned int link_max = UINT_MAX >> 1; /* rough margin */ ++ ++ if (!au_test_fs_no_limit_nlink(inode->i_sb) ++ || inode->i_nlink < link_max) ++ return 0; ++ return -EMLINK; ++} ++ ++int vfsub_link(struct dentry *src_dentry, struct inode *dir, struct path *path, ++ struct inode **delegated_inode) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ err = au_test_nlink(d_inode(src_dentry)); ++ if (unlikely(err)) ++ return err; ++ ++ /* we don't call may_linkat() */ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_link(src_dentry, path, d); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_link(src_dentry, dir, path->dentry, delegated_inode); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = *path; ++ int did; ++ ++ /* fuse has different memory inode for the same inumber */ ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = path->dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ tmp.dentry = src_dentry; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++int vfsub_rename(struct inode *src_dir, struct dentry *src_dentry, ++ struct inode *dir, struct path *path, ++ struct inode **delegated_inode) ++{ ++ int err; ++ struct path tmp = { ++ .mnt = path->mnt ++ }; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ IMustLock(src_dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ tmp.dentry = src_dentry->d_parent; ++ err = security_path_rename(&tmp, src_dentry, path, d, /*flags*/0); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_rename(src_dir, src_dentry, dir, path->dentry, ++ delegated_inode, /*flags*/0); ++ lockdep_on(); ++ if (!err) { ++ int did; ++ ++ tmp.dentry = d->d_parent; ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = src_dentry; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ tmp.dentry = src_dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++int vfsub_mkdir(struct inode *dir, struct path *path, int mode) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_mkdir(path, d, mode); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_mkdir(dir, path->dentry, mode); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = *path; ++ int did; ++ ++ vfsub_update_h_iattr(&tmp, &did); ++ if (did) { ++ tmp.dentry = path->dentry->d_parent; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); ++ } ++ /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++int vfsub_rmdir(struct inode *dir, struct path *path) ++{ ++ int err; ++ struct dentry *d; ++ ++ IMustLock(dir); ++ ++ d = path->dentry; ++ path->dentry = d->d_parent; ++ err = security_path_rmdir(path, d); ++ path->dentry = d; ++ if (unlikely(err)) ++ goto out; ++ ++ lockdep_off(); ++ err = vfs_rmdir(dir, path->dentry); ++ lockdep_on(); ++ if (!err) { ++ struct path tmp = { ++ .dentry = path->dentry->d_parent, ++ .mnt = path->mnt ++ }; ++ ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); /*ignore*/ ++ } ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* todo: support mmap_sem? */ ++ssize_t vfsub_read_u(struct file *file, char __user *ubuf, size_t count, ++ loff_t *ppos) ++{ ++ ssize_t err; ++ ++ lockdep_off(); ++ err = vfs_read(file, ubuf, count, ppos); ++ lockdep_on(); ++ if (err >= 0) ++ vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++/* todo: kernel_read()? */ ++ssize_t vfsub_read_k(struct file *file, void *kbuf, size_t count, ++ loff_t *ppos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ char __user *u; ++ } buf; ++ ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ err = vfsub_read_u(file, buf.u, count, ppos); ++ set_fs(oldfs); ++ return err; ++} ++ ++ssize_t vfsub_write_u(struct file *file, const char __user *ubuf, size_t count, ++ loff_t *ppos) ++{ ++ ssize_t err; ++ ++ lockdep_off(); ++ err = vfs_write(file, ubuf, count, ppos); ++ lockdep_on(); ++ if (err >= 0) ++ vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++ssize_t vfsub_write_k(struct file *file, void *kbuf, size_t count, loff_t *ppos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ const char __user *u; ++ } buf; ++ ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ err = vfsub_write_u(file, buf.u, count, ppos); ++ set_fs(oldfs); ++ return err; ++} ++ ++int vfsub_flush(struct file *file, fl_owner_t id) ++{ ++ int err; ++ ++ err = 0; ++ if (file->f_op->flush) { ++ if (!au_test_nfs(file->f_path.dentry->d_sb)) ++ err = file->f_op->flush(file, id); ++ else { ++ lockdep_off(); ++ err = file->f_op->flush(file, id); ++ lockdep_on(); ++ } ++ if (!err) ++ vfsub_update_h_iattr(&file->f_path, /*did*/NULL); ++ /*ignore*/ ++ } ++ return err; ++} ++ ++int vfsub_iterate_dir(struct file *file, struct dir_context *ctx) ++{ ++ int err; ++ ++ AuDbg("%pD, ctx{%pf, %llu}\n", file, ctx->actor, ctx->pos); ++ ++ lockdep_off(); ++ err = iterate_dir(file, ctx); ++ lockdep_on(); ++ if (err >= 0) ++ vfsub_update_h_iattr(&file->f_path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++long vfsub_splice_to(struct file *in, loff_t *ppos, ++ struct pipe_inode_info *pipe, size_t len, ++ unsigned int flags) ++{ ++ long err; ++ ++ lockdep_off(); ++ err = do_splice_to(in, ppos, pipe, len, flags); ++ lockdep_on(); ++ file_accessed(in); ++ if (err >= 0) ++ vfsub_update_h_iattr(&in->f_path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++long vfsub_splice_from(struct pipe_inode_info *pipe, struct file *out, ++ loff_t *ppos, size_t len, unsigned int flags) ++{ ++ long err; ++ ++ lockdep_off(); ++ err = do_splice_from(pipe, out, ppos, len, flags); ++ lockdep_on(); ++ if (err >= 0) ++ vfsub_update_h_iattr(&out->f_path, /*did*/NULL); /*ignore*/ ++ return err; ++} ++ ++int vfsub_fsync(struct file *file, struct path *path, int datasync) ++{ ++ int err; ++ ++ /* file can be NULL */ ++ lockdep_off(); ++ err = vfs_fsync(file, datasync); ++ lockdep_on(); ++ if (!err) { ++ if (!path) { ++ AuDebugOn(!file); ++ path = &file->f_path; ++ } ++ vfsub_update_h_iattr(path, /*did*/NULL); /*ignore*/ ++ } ++ return err; ++} ++ ++/* cf. open.c:do_sys_truncate() and do_sys_ftruncate() */ ++int vfsub_trunc(struct path *h_path, loff_t length, unsigned int attr, ++ struct file *h_file) ++{ ++ int err; ++ struct inode *h_inode; ++ struct super_block *h_sb; ++ ++ if (!h_file) { ++ err = vfsub_truncate(h_path, length); ++ goto out; ++ } ++ ++ h_inode = d_inode(h_path->dentry); ++ h_sb = h_inode->i_sb; ++ lockdep_off(); ++ sb_start_write(h_sb); ++ lockdep_on(); ++ err = locks_verify_truncate(h_inode, h_file, length); ++ if (!err) ++ err = security_path_truncate(h_path); ++ if (!err) { ++ lockdep_off(); ++ err = do_truncate(h_path->dentry, length, attr, h_file); ++ lockdep_on(); ++ } ++ lockdep_off(); ++ sb_end_write(h_sb); ++ lockdep_on(); ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_vfsub_mkdir_args { ++ int *errp; ++ struct inode *dir; ++ struct path *path; ++ int mode; ++}; ++ ++static void au_call_vfsub_mkdir(void *args) ++{ ++ struct au_vfsub_mkdir_args *a = args; ++ *a->errp = vfsub_mkdir(a->dir, a->path, a->mode); ++} ++ ++int vfsub_sio_mkdir(struct inode *dir, struct path *path, int mode) ++{ ++ int err, do_sio, wkq_err; ++ ++ do_sio = au_test_h_perm_sio(dir, MAY_EXEC | MAY_WRITE); ++ if (!do_sio) { ++ lockdep_off(); ++ err = vfsub_mkdir(dir, path, mode); ++ lockdep_on(); ++ } else { ++ struct au_vfsub_mkdir_args args = { ++ .errp = &err, ++ .dir = dir, ++ .path = path, ++ .mode = mode ++ }; ++ wkq_err = au_wkq_wait(au_call_vfsub_mkdir, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ return err; ++} ++ ++struct au_vfsub_rmdir_args { ++ int *errp; ++ struct inode *dir; ++ struct path *path; ++}; ++ ++static void au_call_vfsub_rmdir(void *args) ++{ ++ struct au_vfsub_rmdir_args *a = args; ++ *a->errp = vfsub_rmdir(a->dir, a->path); ++} ++ ++int vfsub_sio_rmdir(struct inode *dir, struct path *path) ++{ ++ int err, do_sio, wkq_err; ++ ++ do_sio = au_test_h_perm_sio(dir, MAY_EXEC | MAY_WRITE); ++ if (!do_sio) { ++ lockdep_off(); ++ err = vfsub_rmdir(dir, path); ++ lockdep_on(); ++ } else { ++ struct au_vfsub_rmdir_args args = { ++ .errp = &err, ++ .dir = dir, ++ .path = path ++ }; ++ wkq_err = au_wkq_wait(au_call_vfsub_rmdir, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct notify_change_args { ++ int *errp; ++ struct path *path; ++ struct iattr *ia; ++ struct inode **delegated_inode; ++}; ++ ++static void call_notify_change(void *args) ++{ ++ struct notify_change_args *a = args; ++ struct inode *h_inode; ++ ++ h_inode = d_inode(a->path->dentry); ++ IMustLock(h_inode); ++ ++ *a->errp = -EPERM; ++ if (!IS_IMMUTABLE(h_inode) && !IS_APPEND(h_inode)) { ++ lockdep_off(); ++ *a->errp = notify_change(a->path->dentry, a->ia, ++ a->delegated_inode); ++ lockdep_on(); ++ if (!*a->errp) ++ vfsub_update_h_iattr(a->path, /*did*/NULL); /*ignore*/ ++ } ++ AuTraceErr(*a->errp); ++} ++ ++int vfsub_notify_change(struct path *path, struct iattr *ia, ++ struct inode **delegated_inode) ++{ ++ int err; ++ struct notify_change_args args = { ++ .errp = &err, ++ .path = path, ++ .ia = ia, ++ .delegated_inode = delegated_inode ++ }; ++ ++ call_notify_change(&args); ++ ++ return err; ++} ++ ++int vfsub_sio_notify_change(struct path *path, struct iattr *ia, ++ struct inode **delegated_inode) ++{ ++ int err, wkq_err; ++ struct notify_change_args args = { ++ .errp = &err, ++ .path = path, ++ .ia = ia, ++ .delegated_inode = delegated_inode ++ }; ++ ++ wkq_err = au_wkq_wait(call_notify_change, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct unlink_args { ++ int *errp; ++ struct inode *dir; ++ struct path *path; ++ struct inode **delegated_inode; ++}; ++ ++static void call_unlink(void *args) ++{ ++ struct unlink_args *a = args; ++ struct dentry *d = a->path->dentry; ++ struct inode *h_inode; ++ const int stop_sillyrename = (au_test_nfs(d->d_sb) ++ && au_dcount(d) == 1); ++ ++ IMustLock(a->dir); ++ ++ a->path->dentry = d->d_parent; ++ *a->errp = security_path_unlink(a->path, d); ++ a->path->dentry = d; ++ if (unlikely(*a->errp)) ++ return; ++ ++ if (!stop_sillyrename) ++ dget(d); ++ h_inode = NULL; ++ if (d_is_positive(d)) { ++ h_inode = d_inode(d); ++ ihold(h_inode); ++ } ++ ++ lockdep_off(); ++ *a->errp = vfs_unlink(a->dir, d, a->delegated_inode); ++ lockdep_on(); ++ if (!*a->errp) { ++ struct path tmp = { ++ .dentry = d->d_parent, ++ .mnt = a->path->mnt ++ }; ++ vfsub_update_h_iattr(&tmp, /*did*/NULL); /*ignore*/ ++ } ++ ++ if (!stop_sillyrename) ++ dput(d); ++ if (h_inode) ++ iput(h_inode); ++ ++ AuTraceErr(*a->errp); ++} ++ ++/* ++ * @dir: must be locked. ++ * @dentry: target dentry. ++ */ ++int vfsub_unlink(struct inode *dir, struct path *path, ++ struct inode **delegated_inode, int force) ++{ ++ int err; ++ struct unlink_args args = { ++ .errp = &err, ++ .dir = dir, ++ .path = path, ++ .delegated_inode = delegated_inode ++ }; ++ ++ if (!force) ++ call_unlink(&args); ++ else { ++ int wkq_err; ++ ++ wkq_err = au_wkq_wait(call_unlink, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/vfsub.h linux-4.1.10/fs/aufs/vfsub.h +--- linux-4.1.10.orig/fs/aufs/vfsub.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/vfsub.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,286 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * sub-routines for VFS ++ */ ++ ++#ifndef __AUFS_VFSUB_H__ ++#define __AUFS_VFSUB_H__ ++ ++#ifdef __KERNEL__ ++ ++#include <linux/fs.h> ++#include <linux/mount.h> ++#include <linux/xattr.h> ++#include "debug.h" ++ ++/* copied from linux/fs/internal.h */ ++/* todo: BAD approach!! */ ++extern void __mnt_drop_write(struct vfsmount *); ++extern spinlock_t inode_sb_list_lock; ++extern int open_check_o_direct(struct file *f); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* lock subclass for lower inode */ ++/* default MAX_LOCKDEP_SUBCLASSES(8) is not enough */ ++/* reduce? gave up. */ ++enum { ++ AuLsc_I_Begin = I_MUTEX_PARENT2, /* 5 */ ++ AuLsc_I_PARENT, /* lower inode, parent first */ ++ AuLsc_I_PARENT2, /* copyup dirs */ ++ AuLsc_I_PARENT3, /* copyup wh */ ++ AuLsc_I_CHILD, ++ AuLsc_I_CHILD2, ++ AuLsc_I_End ++}; ++ ++/* to debug easier, do not make them inlined functions */ ++#define MtxMustLock(mtx) AuDebugOn(!mutex_is_locked(mtx)) ++#define IMustLock(i) MtxMustLock(&(i)->i_mutex) ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline void vfsub_drop_nlink(struct inode *inode) ++{ ++ AuDebugOn(!inode->i_nlink); ++ drop_nlink(inode); ++} ++ ++static inline void vfsub_dead_dir(struct inode *inode) ++{ ++ AuDebugOn(!S_ISDIR(inode->i_mode)); ++ inode->i_flags |= S_DEAD; ++ clear_nlink(inode); ++} ++ ++static inline int vfsub_native_ro(struct inode *inode) ++{ ++ return (inode->i_sb->s_flags & MS_RDONLY) ++ || IS_RDONLY(inode) ++ /* || IS_APPEND(inode) */ ++ || IS_IMMUTABLE(inode); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int vfsub_update_h_iattr(struct path *h_path, int *did); ++struct file *vfsub_dentry_open(struct path *path, int flags); ++struct file *vfsub_filp_open(const char *path, int oflags, int mode); ++struct vfsub_aopen_args { ++ struct file *file; ++ unsigned int open_flag; ++ umode_t create_mode; ++ int *opened; ++}; ++struct au_branch; ++int vfsub_atomic_open(struct inode *dir, struct dentry *dentry, ++ struct vfsub_aopen_args *args, struct au_branch *br); ++int vfsub_kern_path(const char *name, unsigned int flags, struct path *path); ++ ++struct dentry *vfsub_lookup_one_len(const char *name, struct dentry *parent, ++ int len); ++ ++struct vfsub_lkup_one_args { ++ struct dentry **errp; ++ struct qstr *name; ++ struct dentry *parent; ++}; ++ ++static inline struct dentry *vfsub_lkup_one(struct qstr *name, ++ struct dentry *parent) ++{ ++ return vfsub_lookup_one_len(name->name, parent, name->len); ++} ++ ++void vfsub_call_lkup_one(void *args); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline int vfsub_mnt_want_write(struct vfsmount *mnt) ++{ ++ int err; ++ ++ lockdep_off(); ++ err = mnt_want_write(mnt); ++ lockdep_on(); ++ return err; ++} ++ ++static inline void vfsub_mnt_drop_write(struct vfsmount *mnt) ++{ ++ lockdep_off(); ++ mnt_drop_write(mnt); ++ lockdep_on(); ++} ++ ++#if 0 /* reserved */ ++static inline void vfsub_mnt_drop_write_file(struct file *file) ++{ ++ lockdep_off(); ++ mnt_drop_write_file(file); ++ lockdep_on(); ++} ++#endif ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_hinode; ++struct dentry *vfsub_lock_rename(struct dentry *d1, struct au_hinode *hdir1, ++ struct dentry *d2, struct au_hinode *hdir2); ++void vfsub_unlock_rename(struct dentry *d1, struct au_hinode *hdir1, ++ struct dentry *d2, struct au_hinode *hdir2); ++ ++int vfsub_create(struct inode *dir, struct path *path, int mode, ++ bool want_excl); ++int vfsub_symlink(struct inode *dir, struct path *path, ++ const char *symname); ++int vfsub_mknod(struct inode *dir, struct path *path, int mode, dev_t dev); ++int vfsub_link(struct dentry *src_dentry, struct inode *dir, ++ struct path *path, struct inode **delegated_inode); ++int vfsub_rename(struct inode *src_hdir, struct dentry *src_dentry, ++ struct inode *hdir, struct path *path, ++ struct inode **delegated_inode); ++int vfsub_mkdir(struct inode *dir, struct path *path, int mode); ++int vfsub_rmdir(struct inode *dir, struct path *path); ++ ++/* ---------------------------------------------------------------------- */ ++ ++ssize_t vfsub_read_u(struct file *file, char __user *ubuf, size_t count, ++ loff_t *ppos); ++ssize_t vfsub_read_k(struct file *file, void *kbuf, size_t count, ++ loff_t *ppos); ++ssize_t vfsub_write_u(struct file *file, const char __user *ubuf, size_t count, ++ loff_t *ppos); ++ssize_t vfsub_write_k(struct file *file, void *kbuf, size_t count, ++ loff_t *ppos); ++int vfsub_flush(struct file *file, fl_owner_t id); ++int vfsub_iterate_dir(struct file *file, struct dir_context *ctx); ++ ++static inline loff_t vfsub_f_size_read(struct file *file) ++{ ++ return i_size_read(file_inode(file)); ++} ++ ++static inline unsigned int vfsub_file_flags(struct file *file) ++{ ++ unsigned int flags; ++ ++ spin_lock(&file->f_lock); ++ flags = file->f_flags; ++ spin_unlock(&file->f_lock); ++ ++ return flags; ++} ++ ++#if 0 /* reserved */ ++static inline void vfsub_file_accessed(struct file *h_file) ++{ ++ file_accessed(h_file); ++ vfsub_update_h_iattr(&h_file->f_path, /*did*/NULL); /*ignore*/ ++} ++#endif ++ ++static inline void vfsub_touch_atime(struct vfsmount *h_mnt, ++ struct dentry *h_dentry) ++{ ++ struct path h_path = { ++ .dentry = h_dentry, ++ .mnt = h_mnt ++ }; ++ touch_atime(&h_path); ++ vfsub_update_h_iattr(&h_path, /*did*/NULL); /*ignore*/ ++} ++ ++static inline int vfsub_update_time(struct inode *h_inode, struct timespec *ts, ++ int flags) ++{ ++ return generic_update_time(h_inode, ts, flags); ++ /* no vfsub_update_h_iattr() since we don't have struct path */ ++} ++ ++long vfsub_splice_to(struct file *in, loff_t *ppos, ++ struct pipe_inode_info *pipe, size_t len, ++ unsigned int flags); ++long vfsub_splice_from(struct pipe_inode_info *pipe, struct file *out, ++ loff_t *ppos, size_t len, unsigned int flags); ++ ++static inline long vfsub_truncate(struct path *path, loff_t length) ++{ ++ long err; ++ ++ lockdep_off(); ++ err = vfs_truncate(path, length); ++ lockdep_on(); ++ return err; ++} ++ ++int vfsub_trunc(struct path *h_path, loff_t length, unsigned int attr, ++ struct file *h_file); ++int vfsub_fsync(struct file *file, struct path *path, int datasync); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline loff_t vfsub_llseek(struct file *file, loff_t offset, int origin) ++{ ++ loff_t err; ++ ++ lockdep_off(); ++ err = vfs_llseek(file, offset, origin); ++ lockdep_on(); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int vfsub_sio_mkdir(struct inode *dir, struct path *path, int mode); ++int vfsub_sio_rmdir(struct inode *dir, struct path *path); ++int vfsub_sio_notify_change(struct path *path, struct iattr *ia, ++ struct inode **delegated_inode); ++int vfsub_notify_change(struct path *path, struct iattr *ia, ++ struct inode **delegated_inode); ++int vfsub_unlink(struct inode *dir, struct path *path, ++ struct inode **delegated_inode, int force); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline int vfsub_setxattr(struct dentry *dentry, const char *name, ++ const void *value, size_t size, int flags) ++{ ++ int err; ++ ++ lockdep_off(); ++ err = vfs_setxattr(dentry, name, value, size, flags); ++ lockdep_on(); ++ ++ return err; ++} ++ ++static inline int vfsub_removexattr(struct dentry *dentry, const char *name) ++{ ++ int err; ++ ++ lockdep_off(); ++ err = vfs_removexattr(dentry, name); ++ lockdep_on(); ++ ++ return err; ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_VFSUB_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/wbr_policy.c linux-4.1.10/fs/aufs/wbr_policy.c +--- linux-4.1.10.orig/fs/aufs/wbr_policy.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/wbr_policy.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,765 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * policies for selecting one among multiple writable branches ++ */ ++ ++#include <linux/statfs.h> ++#include "aufs.h" ++ ++/* subset of cpup_attr() */ ++static noinline_for_stack ++int au_cpdown_attr(struct path *h_path, struct dentry *h_src) ++{ ++ int err, sbits; ++ struct iattr ia; ++ struct inode *h_isrc; ++ ++ h_isrc = d_inode(h_src); ++ ia.ia_valid = ATTR_FORCE | ATTR_MODE | ATTR_UID | ATTR_GID; ++ ia.ia_mode = h_isrc->i_mode; ++ ia.ia_uid = h_isrc->i_uid; ++ ia.ia_gid = h_isrc->i_gid; ++ sbits = !!(ia.ia_mode & (S_ISUID | S_ISGID)); ++ au_cpup_attr_flags(d_inode(h_path->dentry), h_isrc->i_flags); ++ /* no delegation since it is just created */ ++ err = vfsub_sio_notify_change(h_path, &ia, /*delegated*/NULL); ++ ++ /* is this nfs only? */ ++ if (!err && sbits && au_test_nfs(h_path->dentry->d_sb)) { ++ ia.ia_valid = ATTR_FORCE | ATTR_MODE; ++ ia.ia_mode = h_isrc->i_mode; ++ err = vfsub_sio_notify_change(h_path, &ia, /*delegated*/NULL); ++ } ++ ++ return err; ++} ++ ++#define AuCpdown_PARENT_OPQ 1 ++#define AuCpdown_WHED (1 << 1) ++#define AuCpdown_MADE_DIR (1 << 2) ++#define AuCpdown_DIROPQ (1 << 3) ++#define au_ftest_cpdown(flags, name) ((flags) & AuCpdown_##name) ++#define au_fset_cpdown(flags, name) \ ++ do { (flags) |= AuCpdown_##name; } while (0) ++#define au_fclr_cpdown(flags, name) \ ++ do { (flags) &= ~AuCpdown_##name; } while (0) ++ ++static int au_cpdown_dir_opq(struct dentry *dentry, aufs_bindex_t bdst, ++ unsigned int *flags) ++{ ++ int err; ++ struct dentry *opq_dentry; ++ ++ opq_dentry = au_diropq_create(dentry, bdst); ++ err = PTR_ERR(opq_dentry); ++ if (IS_ERR(opq_dentry)) ++ goto out; ++ dput(opq_dentry); ++ au_fset_cpdown(*flags, DIROPQ); ++ ++out: ++ return err; ++} ++ ++static int au_cpdown_dir_wh(struct dentry *dentry, struct dentry *h_parent, ++ struct inode *dir, aufs_bindex_t bdst) ++{ ++ int err; ++ struct path h_path; ++ struct au_branch *br; ++ ++ br = au_sbr(dentry->d_sb, bdst); ++ h_path.dentry = au_wh_lkup(h_parent, &dentry->d_name, br); ++ err = PTR_ERR(h_path.dentry); ++ if (IS_ERR(h_path.dentry)) ++ goto out; ++ ++ err = 0; ++ if (d_is_positive(h_path.dentry)) { ++ h_path.mnt = au_br_mnt(br); ++ err = au_wh_unlink_dentry(au_h_iptr(dir, bdst), &h_path, ++ dentry); ++ } ++ dput(h_path.dentry); ++ ++out: ++ return err; ++} ++ ++static int au_cpdown_dir(struct dentry *dentry, aufs_bindex_t bdst, ++ struct au_pin *pin, ++ struct dentry *h_parent, void *arg) ++{ ++ int err, rerr; ++ aufs_bindex_t bopq, bstart; ++ struct path h_path; ++ struct dentry *parent; ++ struct inode *h_dir, *h_inode, *inode, *dir; ++ unsigned int *flags = arg; ++ ++ bstart = au_dbstart(dentry); ++ /* dentry is di-locked */ ++ parent = dget_parent(dentry); ++ dir = d_inode(parent); ++ h_dir = d_inode(h_parent); ++ AuDebugOn(h_dir != au_h_iptr(dir, bdst)); ++ IMustLock(h_dir); ++ ++ err = au_lkup_neg(dentry, bdst, /*wh*/0); ++ if (unlikely(err < 0)) ++ goto out; ++ h_path.dentry = au_h_dptr(dentry, bdst); ++ h_path.mnt = au_sbr_mnt(dentry->d_sb, bdst); ++ err = vfsub_sio_mkdir(au_h_iptr(dir, bdst), &h_path, ++ S_IRWXU | S_IRUGO | S_IXUGO); ++ if (unlikely(err)) ++ goto out_put; ++ au_fset_cpdown(*flags, MADE_DIR); ++ ++ bopq = au_dbdiropq(dentry); ++ au_fclr_cpdown(*flags, WHED); ++ au_fclr_cpdown(*flags, DIROPQ); ++ if (au_dbwh(dentry) == bdst) ++ au_fset_cpdown(*flags, WHED); ++ if (!au_ftest_cpdown(*flags, PARENT_OPQ) && bopq <= bdst) ++ au_fset_cpdown(*flags, PARENT_OPQ); ++ h_inode = d_inode(h_path.dentry); ++ mutex_lock_nested(&h_inode->i_mutex, AuLsc_I_CHILD); ++ if (au_ftest_cpdown(*flags, WHED)) { ++ err = au_cpdown_dir_opq(dentry, bdst, flags); ++ if (unlikely(err)) { ++ mutex_unlock(&h_inode->i_mutex); ++ goto out_dir; ++ } ++ } ++ ++ err = au_cpdown_attr(&h_path, au_h_dptr(dentry, bstart)); ++ mutex_unlock(&h_inode->i_mutex); ++ if (unlikely(err)) ++ goto out_opq; ++ ++ if (au_ftest_cpdown(*flags, WHED)) { ++ err = au_cpdown_dir_wh(dentry, h_parent, dir, bdst); ++ if (unlikely(err)) ++ goto out_opq; ++ } ++ ++ inode = d_inode(dentry); ++ if (au_ibend(inode) < bdst) ++ au_set_ibend(inode, bdst); ++ au_set_h_iptr(inode, bdst, au_igrab(h_inode), ++ au_hi_flags(inode, /*isdir*/1)); ++ au_fhsm_wrote(dentry->d_sb, bdst, /*force*/0); ++ goto out; /* success */ ++ ++ /* revert */ ++out_opq: ++ if (au_ftest_cpdown(*flags, DIROPQ)) { ++ mutex_lock_nested(&h_inode->i_mutex, AuLsc_I_CHILD); ++ rerr = au_diropq_remove(dentry, bdst); ++ mutex_unlock(&h_inode->i_mutex); ++ if (unlikely(rerr)) { ++ AuIOErr("failed removing diropq for %pd b%d (%d)\n", ++ dentry, bdst, rerr); ++ err = -EIO; ++ goto out; ++ } ++ } ++out_dir: ++ if (au_ftest_cpdown(*flags, MADE_DIR)) { ++ rerr = vfsub_sio_rmdir(au_h_iptr(dir, bdst), &h_path); ++ if (unlikely(rerr)) { ++ AuIOErr("failed removing %pd b%d (%d)\n", ++ dentry, bdst, rerr); ++ err = -EIO; ++ } ++ } ++out_put: ++ au_set_h_dptr(dentry, bdst, NULL); ++ if (au_dbend(dentry) == bdst) ++ au_update_dbend(dentry); ++out: ++ dput(parent); ++ return err; ++} ++ ++int au_cpdown_dirs(struct dentry *dentry, aufs_bindex_t bdst) ++{ ++ int err; ++ unsigned int flags; ++ ++ flags = 0; ++ err = au_cp_dirs(dentry, bdst, au_cpdown_dir, &flags); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* policies for create */ ++ ++int au_wbr_nonopq(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ int err, i, j, ndentry; ++ aufs_bindex_t bopq; ++ struct au_dcsub_pages dpages; ++ struct au_dpage *dpage; ++ struct dentry **dentries, *parent, *d; ++ ++ err = au_dpages_init(&dpages, GFP_NOFS); ++ if (unlikely(err)) ++ goto out; ++ parent = dget_parent(dentry); ++ err = au_dcsub_pages_rev_aufs(&dpages, parent, /*do_include*/0); ++ if (unlikely(err)) ++ goto out_free; ++ ++ err = bindex; ++ for (i = 0; i < dpages.ndpage; i++) { ++ dpage = dpages.dpages + i; ++ dentries = dpage->dentries; ++ ndentry = dpage->ndentry; ++ for (j = 0; j < ndentry; j++) { ++ d = dentries[j]; ++ di_read_lock_parent2(d, !AuLock_IR); ++ bopq = au_dbdiropq(d); ++ di_read_unlock(d, !AuLock_IR); ++ if (bopq >= 0 && bopq < err) ++ err = bopq; ++ } ++ } ++ ++out_free: ++ dput(parent); ++ au_dpages_free(&dpages); ++out: ++ return err; ++} ++ ++static int au_wbr_bu(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ for (; bindex >= 0; bindex--) ++ if (!au_br_rdonly(au_sbr(sb, bindex))) ++ return bindex; ++ return -EROFS; ++} ++ ++/* top down parent */ ++static int au_wbr_create_tdp(struct dentry *dentry, ++ unsigned int flags __maybe_unused) ++{ ++ int err; ++ aufs_bindex_t bstart, bindex; ++ struct super_block *sb; ++ struct dentry *parent, *h_parent; ++ ++ sb = dentry->d_sb; ++ bstart = au_dbstart(dentry); ++ err = bstart; ++ if (!au_br_rdonly(au_sbr(sb, bstart))) ++ goto out; ++ ++ err = -EROFS; ++ parent = dget_parent(dentry); ++ for (bindex = au_dbstart(parent); bindex < bstart; bindex++) { ++ h_parent = au_h_dptr(parent, bindex); ++ if (!h_parent || d_is_negative(h_parent)) ++ continue; ++ ++ if (!au_br_rdonly(au_sbr(sb, bindex))) { ++ err = bindex; ++ break; ++ } ++ } ++ dput(parent); ++ ++ /* bottom up here */ ++ if (unlikely(err < 0)) { ++ err = au_wbr_bu(sb, bstart - 1); ++ if (err >= 0) ++ err = au_wbr_nonopq(dentry, err); ++ } ++ ++out: ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* an exception for the policy other than tdp */ ++static int au_wbr_create_exp(struct dentry *dentry) ++{ ++ int err; ++ aufs_bindex_t bwh, bdiropq; ++ struct dentry *parent; ++ ++ err = -1; ++ bwh = au_dbwh(dentry); ++ parent = dget_parent(dentry); ++ bdiropq = au_dbdiropq(parent); ++ if (bwh >= 0) { ++ if (bdiropq >= 0) ++ err = min(bdiropq, bwh); ++ else ++ err = bwh; ++ AuDbg("%d\n", err); ++ } else if (bdiropq >= 0) { ++ err = bdiropq; ++ AuDbg("%d\n", err); ++ } ++ dput(parent); ++ ++ if (err >= 0) ++ err = au_wbr_nonopq(dentry, err); ++ ++ if (err >= 0 && au_br_rdonly(au_sbr(dentry->d_sb, err))) ++ err = -1; ++ ++ AuDbg("%d\n", err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* round robin */ ++static int au_wbr_create_init_rr(struct super_block *sb) ++{ ++ int err; ++ ++ err = au_wbr_bu(sb, au_sbend(sb)); ++ atomic_set(&au_sbi(sb)->si_wbr_rr_next, -err); /* less important */ ++ /* smp_mb(); */ ++ ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++static int au_wbr_create_rr(struct dentry *dentry, unsigned int flags) ++{ ++ int err, nbr; ++ unsigned int u; ++ aufs_bindex_t bindex, bend; ++ struct super_block *sb; ++ atomic_t *next; ++ ++ err = au_wbr_create_exp(dentry); ++ if (err >= 0) ++ goto out; ++ ++ sb = dentry->d_sb; ++ next = &au_sbi(sb)->si_wbr_rr_next; ++ bend = au_sbend(sb); ++ nbr = bend + 1; ++ for (bindex = 0; bindex <= bend; bindex++) { ++ if (!au_ftest_wbr(flags, DIR)) { ++ err = atomic_dec_return(next) + 1; ++ /* modulo for 0 is meaningless */ ++ if (unlikely(!err)) ++ err = atomic_dec_return(next) + 1; ++ } else ++ err = atomic_read(next); ++ AuDbg("%d\n", err); ++ u = err; ++ err = u % nbr; ++ AuDbg("%d\n", err); ++ if (!au_br_rdonly(au_sbr(sb, err))) ++ break; ++ err = -EROFS; ++ } ++ ++ if (err >= 0) ++ err = au_wbr_nonopq(dentry, err); ++ ++out: ++ AuDbg("%d\n", err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* most free space */ ++static void au_mfs(struct dentry *dentry, struct dentry *parent) ++{ ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_wbr_mfs *mfs; ++ struct dentry *h_parent; ++ aufs_bindex_t bindex, bend; ++ int err; ++ unsigned long long b, bavail; ++ struct path h_path; ++ /* reduce the stack usage */ ++ struct kstatfs *st; ++ ++ st = kmalloc(sizeof(*st), GFP_NOFS); ++ if (unlikely(!st)) { ++ AuWarn1("failed updating mfs(%d), ignored\n", -ENOMEM); ++ return; ++ } ++ ++ bavail = 0; ++ sb = dentry->d_sb; ++ mfs = &au_sbi(sb)->si_wbr_mfs; ++ MtxMustLock(&mfs->mfs_lock); ++ mfs->mfs_bindex = -EROFS; ++ mfs->mfsrr_bytes = 0; ++ if (!parent) { ++ bindex = 0; ++ bend = au_sbend(sb); ++ } else { ++ bindex = au_dbstart(parent); ++ bend = au_dbtaildir(parent); ++ } ++ ++ for (; bindex <= bend; bindex++) { ++ if (parent) { ++ h_parent = au_h_dptr(parent, bindex); ++ if (!h_parent || d_is_negative(h_parent)) ++ continue; ++ } ++ br = au_sbr(sb, bindex); ++ if (au_br_rdonly(br)) ++ continue; ++ ++ /* sb->s_root for NFS is unreliable */ ++ h_path.mnt = au_br_mnt(br); ++ h_path.dentry = h_path.mnt->mnt_root; ++ err = vfs_statfs(&h_path, st); ++ if (unlikely(err)) { ++ AuWarn1("failed statfs, b%d, %d\n", bindex, err); ++ continue; ++ } ++ ++ /* when the available size is equal, select the lower one */ ++ BUILD_BUG_ON(sizeof(b) < sizeof(st->f_bavail) ++ || sizeof(b) < sizeof(st->f_bsize)); ++ b = st->f_bavail * st->f_bsize; ++ br->br_wbr->wbr_bytes = b; ++ if (b >= bavail) { ++ bavail = b; ++ mfs->mfs_bindex = bindex; ++ mfs->mfs_jiffy = jiffies; ++ } ++ } ++ ++ mfs->mfsrr_bytes = bavail; ++ AuDbg("b%d\n", mfs->mfs_bindex); ++ kfree(st); ++} ++ ++static int au_wbr_create_mfs(struct dentry *dentry, unsigned int flags) ++{ ++ int err; ++ struct dentry *parent; ++ struct super_block *sb; ++ struct au_wbr_mfs *mfs; ++ ++ err = au_wbr_create_exp(dentry); ++ if (err >= 0) ++ goto out; ++ ++ sb = dentry->d_sb; ++ parent = NULL; ++ if (au_ftest_wbr(flags, PARENT)) ++ parent = dget_parent(dentry); ++ mfs = &au_sbi(sb)->si_wbr_mfs; ++ mutex_lock(&mfs->mfs_lock); ++ if (time_after(jiffies, mfs->mfs_jiffy + mfs->mfs_expire) ++ || mfs->mfs_bindex < 0 ++ || au_br_rdonly(au_sbr(sb, mfs->mfs_bindex))) ++ au_mfs(dentry, parent); ++ mutex_unlock(&mfs->mfs_lock); ++ err = mfs->mfs_bindex; ++ dput(parent); ++ ++ if (err >= 0) ++ err = au_wbr_nonopq(dentry, err); ++ ++out: ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++static int au_wbr_create_init_mfs(struct super_block *sb) ++{ ++ struct au_wbr_mfs *mfs; ++ ++ mfs = &au_sbi(sb)->si_wbr_mfs; ++ mutex_init(&mfs->mfs_lock); ++ mfs->mfs_jiffy = 0; ++ mfs->mfs_bindex = -EROFS; ++ ++ return 0; ++} ++ ++static int au_wbr_create_fin_mfs(struct super_block *sb __maybe_unused) ++{ ++ mutex_destroy(&au_sbi(sb)->si_wbr_mfs.mfs_lock); ++ return 0; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* most free space and then round robin */ ++static int au_wbr_create_mfsrr(struct dentry *dentry, unsigned int flags) ++{ ++ int err; ++ struct au_wbr_mfs *mfs; ++ ++ err = au_wbr_create_mfs(dentry, flags); ++ if (err >= 0) { ++ mfs = &au_sbi(dentry->d_sb)->si_wbr_mfs; ++ mutex_lock(&mfs->mfs_lock); ++ if (mfs->mfsrr_bytes < mfs->mfsrr_watermark) ++ err = au_wbr_create_rr(dentry, flags); ++ mutex_unlock(&mfs->mfs_lock); ++ } ++ ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++static int au_wbr_create_init_mfsrr(struct super_block *sb) ++{ ++ int err; ++ ++ au_wbr_create_init_mfs(sb); /* ignore */ ++ err = au_wbr_create_init_rr(sb); ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* top down parent and most free space */ ++static int au_wbr_create_pmfs(struct dentry *dentry, unsigned int flags) ++{ ++ int err, e2; ++ unsigned long long b; ++ aufs_bindex_t bindex, bstart, bend; ++ struct super_block *sb; ++ struct dentry *parent, *h_parent; ++ struct au_branch *br; ++ ++ err = au_wbr_create_tdp(dentry, flags); ++ if (unlikely(err < 0)) ++ goto out; ++ parent = dget_parent(dentry); ++ bstart = au_dbstart(parent); ++ bend = au_dbtaildir(parent); ++ if (bstart == bend) ++ goto out_parent; /* success */ ++ ++ e2 = au_wbr_create_mfs(dentry, flags); ++ if (e2 < 0) ++ goto out_parent; /* success */ ++ ++ /* when the available size is equal, select upper one */ ++ sb = dentry->d_sb; ++ br = au_sbr(sb, err); ++ b = br->br_wbr->wbr_bytes; ++ AuDbg("b%d, %llu\n", err, b); ++ ++ for (bindex = bstart; bindex <= bend; bindex++) { ++ h_parent = au_h_dptr(parent, bindex); ++ if (!h_parent || d_is_negative(h_parent)) ++ continue; ++ ++ br = au_sbr(sb, bindex); ++ if (!au_br_rdonly(br) && br->br_wbr->wbr_bytes > b) { ++ b = br->br_wbr->wbr_bytes; ++ err = bindex; ++ AuDbg("b%d, %llu\n", err, b); ++ } ++ } ++ ++ if (err >= 0) ++ err = au_wbr_nonopq(dentry, err); ++ ++out_parent: ++ dput(parent); ++out: ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * - top down parent ++ * - most free space with parent ++ * - most free space round-robin regardless parent ++ */ ++static int au_wbr_create_pmfsrr(struct dentry *dentry, unsigned int flags) ++{ ++ int err; ++ unsigned long long watermark; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct au_wbr_mfs *mfs; ++ ++ err = au_wbr_create_pmfs(dentry, flags | AuWbr_PARENT); ++ if (unlikely(err < 0)) ++ goto out; ++ ++ sb = dentry->d_sb; ++ br = au_sbr(sb, err); ++ mfs = &au_sbi(sb)->si_wbr_mfs; ++ mutex_lock(&mfs->mfs_lock); ++ watermark = mfs->mfsrr_watermark; ++ mutex_unlock(&mfs->mfs_lock); ++ if (br->br_wbr->wbr_bytes < watermark) ++ /* regardless the parent dir */ ++ err = au_wbr_create_mfsrr(dentry, flags); ++ ++out: ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* policies for copyup */ ++ ++/* top down parent */ ++static int au_wbr_copyup_tdp(struct dentry *dentry) ++{ ++ return au_wbr_create_tdp(dentry, /*flags, anything is ok*/0); ++} ++ ++/* bottom up parent */ ++static int au_wbr_copyup_bup(struct dentry *dentry) ++{ ++ int err; ++ aufs_bindex_t bindex, bstart; ++ struct dentry *parent, *h_parent; ++ struct super_block *sb; ++ ++ err = -EROFS; ++ sb = dentry->d_sb; ++ parent = dget_parent(dentry); ++ bstart = au_dbstart(parent); ++ for (bindex = au_dbstart(dentry); bindex >= bstart; bindex--) { ++ h_parent = au_h_dptr(parent, bindex); ++ if (!h_parent || d_is_negative(h_parent)) ++ continue; ++ ++ if (!au_br_rdonly(au_sbr(sb, bindex))) { ++ err = bindex; ++ break; ++ } ++ } ++ dput(parent); ++ ++ /* bottom up here */ ++ if (unlikely(err < 0)) ++ err = au_wbr_bu(sb, bstart - 1); ++ ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++/* bottom up */ ++int au_wbr_do_copyup_bu(struct dentry *dentry, aufs_bindex_t bstart) ++{ ++ int err; ++ ++ err = au_wbr_bu(dentry->d_sb, bstart); ++ AuDbg("b%d\n", err); ++ if (err > bstart) ++ err = au_wbr_nonopq(dentry, err); ++ ++ AuDbg("b%d\n", err); ++ return err; ++} ++ ++static int au_wbr_copyup_bu(struct dentry *dentry) ++{ ++ int err; ++ aufs_bindex_t bstart; ++ ++ bstart = au_dbstart(dentry); ++ err = au_wbr_do_copyup_bu(dentry, bstart); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_wbr_copyup_operations au_wbr_copyup_ops[] = { ++ [AuWbrCopyup_TDP] = { ++ .copyup = au_wbr_copyup_tdp ++ }, ++ [AuWbrCopyup_BUP] = { ++ .copyup = au_wbr_copyup_bup ++ }, ++ [AuWbrCopyup_BU] = { ++ .copyup = au_wbr_copyup_bu ++ } ++}; ++ ++struct au_wbr_create_operations au_wbr_create_ops[] = { ++ [AuWbrCreate_TDP] = { ++ .create = au_wbr_create_tdp ++ }, ++ [AuWbrCreate_RR] = { ++ .create = au_wbr_create_rr, ++ .init = au_wbr_create_init_rr ++ }, ++ [AuWbrCreate_MFS] = { ++ .create = au_wbr_create_mfs, ++ .init = au_wbr_create_init_mfs, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_MFSV] = { ++ .create = au_wbr_create_mfs, ++ .init = au_wbr_create_init_mfs, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_MFSRR] = { ++ .create = au_wbr_create_mfsrr, ++ .init = au_wbr_create_init_mfsrr, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_MFSRRV] = { ++ .create = au_wbr_create_mfsrr, ++ .init = au_wbr_create_init_mfsrr, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_PMFS] = { ++ .create = au_wbr_create_pmfs, ++ .init = au_wbr_create_init_mfs, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_PMFSV] = { ++ .create = au_wbr_create_pmfs, ++ .init = au_wbr_create_init_mfs, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_PMFSRR] = { ++ .create = au_wbr_create_pmfsrr, ++ .init = au_wbr_create_init_mfsrr, ++ .fin = au_wbr_create_fin_mfs ++ }, ++ [AuWbrCreate_PMFSRRV] = { ++ .create = au_wbr_create_pmfsrr, ++ .init = au_wbr_create_init_mfsrr, ++ .fin = au_wbr_create_fin_mfs ++ } ++}; +diff -Nur linux-4.1.10.orig/fs/aufs/whout.c linux-4.1.10/fs/aufs/whout.c +--- linux-4.1.10.orig/fs/aufs/whout.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/whout.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1063 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * whiteout for logical deletion and opaque directory ++ */ ++ ++#include "aufs.h" ++ ++#define WH_MASK S_IRUGO ++ ++/* ++ * If a directory contains this file, then it is opaque. We start with the ++ * .wh. flag so that it is blocked by lookup. ++ */ ++static struct qstr diropq_name = QSTR_INIT(AUFS_WH_DIROPQ, ++ sizeof(AUFS_WH_DIROPQ) - 1); ++ ++/* ++ * generate whiteout name, which is NOT terminated by NULL. ++ * @name: original d_name.name ++ * @len: original d_name.len ++ * @wh: whiteout qstr ++ * returns zero when succeeds, otherwise error. ++ * succeeded value as wh->name should be freed by kfree(). ++ */ ++int au_wh_name_alloc(struct qstr *wh, const struct qstr *name) ++{ ++ char *p; ++ ++ if (unlikely(name->len > PATH_MAX - AUFS_WH_PFX_LEN)) ++ return -ENAMETOOLONG; ++ ++ wh->len = name->len + AUFS_WH_PFX_LEN; ++ p = kmalloc(wh->len, GFP_NOFS); ++ wh->name = p; ++ if (p) { ++ memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); ++ memcpy(p + AUFS_WH_PFX_LEN, name->name, name->len); ++ /* smp_mb(); */ ++ return 0; ++ } ++ return -ENOMEM; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * test if the @wh_name exists under @h_parent. ++ * @try_sio specifies the necessary of super-io. ++ */ ++int au_wh_test(struct dentry *h_parent, struct qstr *wh_name, int try_sio) ++{ ++ int err; ++ struct dentry *wh_dentry; ++ ++ if (!try_sio) ++ wh_dentry = vfsub_lkup_one(wh_name, h_parent); ++ else ++ wh_dentry = au_sio_lkup_one(wh_name, h_parent); ++ err = PTR_ERR(wh_dentry); ++ if (IS_ERR(wh_dentry)) { ++ if (err == -ENAMETOOLONG) ++ err = 0; ++ goto out; ++ } ++ ++ err = 0; ++ if (d_is_negative(wh_dentry)) ++ goto out_wh; /* success */ ++ ++ err = 1; ++ if (d_is_reg(wh_dentry)) ++ goto out_wh; /* success */ ++ ++ err = -EIO; ++ AuIOErr("%pd Invalid whiteout entry type 0%o.\n", ++ wh_dentry, d_inode(wh_dentry)->i_mode); ++ ++out_wh: ++ dput(wh_dentry); ++out: ++ return err; ++} ++ ++/* ++ * test if the @h_dentry sets opaque or not. ++ */ ++int au_diropq_test(struct dentry *h_dentry) ++{ ++ int err; ++ struct inode *h_dir; ++ ++ h_dir = d_inode(h_dentry); ++ err = au_wh_test(h_dentry, &diropq_name, ++ au_test_h_perm_sio(h_dir, MAY_EXEC)); ++ return err; ++} ++ ++/* ++ * returns a negative dentry whose name is unique and temporary. ++ */ ++struct dentry *au_whtmp_lkup(struct dentry *h_parent, struct au_branch *br, ++ struct qstr *prefix) ++{ ++ struct dentry *dentry; ++ int i; ++ char defname[NAME_MAX - AUFS_MAX_NAMELEN + DNAME_INLINE_LEN + 1], ++ *name, *p; ++ /* strict atomic_t is unnecessary here */ ++ static unsigned short cnt; ++ struct qstr qs; ++ ++ BUILD_BUG_ON(sizeof(cnt) * 2 > AUFS_WH_TMP_LEN); ++ ++ name = defname; ++ qs.len = sizeof(defname) - DNAME_INLINE_LEN + prefix->len - 1; ++ if (unlikely(prefix->len > DNAME_INLINE_LEN)) { ++ dentry = ERR_PTR(-ENAMETOOLONG); ++ if (unlikely(qs.len > NAME_MAX)) ++ goto out; ++ dentry = ERR_PTR(-ENOMEM); ++ name = kmalloc(qs.len + 1, GFP_NOFS); ++ if (unlikely(!name)) ++ goto out; ++ } ++ ++ /* doubly whiteout-ed */ ++ memcpy(name, AUFS_WH_PFX AUFS_WH_PFX, AUFS_WH_PFX_LEN * 2); ++ p = name + AUFS_WH_PFX_LEN * 2; ++ memcpy(p, prefix->name, prefix->len); ++ p += prefix->len; ++ *p++ = '.'; ++ AuDebugOn(name + qs.len + 1 - p <= AUFS_WH_TMP_LEN); ++ ++ qs.name = name; ++ for (i = 0; i < 3; i++) { ++ sprintf(p, "%.*x", AUFS_WH_TMP_LEN, cnt++); ++ dentry = au_sio_lkup_one(&qs, h_parent); ++ if (IS_ERR(dentry) || d_is_negative(dentry)) ++ goto out_name; ++ dput(dentry); ++ } ++ /* pr_warn("could not get random name\n"); */ ++ dentry = ERR_PTR(-EEXIST); ++ AuDbg("%.*s\n", AuLNPair(&qs)); ++ BUG(); ++ ++out_name: ++ if (name != defname) ++ kfree(name); ++out: ++ AuTraceErrPtr(dentry); ++ return dentry; ++} ++ ++/* ++ * rename the @h_dentry on @br to the whiteouted temporary name. ++ */ ++int au_whtmp_ren(struct dentry *h_dentry, struct au_branch *br) ++{ ++ int err; ++ struct path h_path = { ++ .mnt = au_br_mnt(br) ++ }; ++ struct inode *h_dir, *delegated; ++ struct dentry *h_parent; ++ ++ h_parent = h_dentry->d_parent; /* dir inode is locked */ ++ h_dir = d_inode(h_parent); ++ IMustLock(h_dir); ++ ++ h_path.dentry = au_whtmp_lkup(h_parent, br, &h_dentry->d_name); ++ err = PTR_ERR(h_path.dentry); ++ if (IS_ERR(h_path.dentry)) ++ goto out; ++ ++ /* under the same dir, no need to lock_rename() */ ++ delegated = NULL; ++ err = vfsub_rename(h_dir, h_dentry, h_dir, &h_path, &delegated); ++ AuTraceErr(err); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal rename\n"); ++ iput(delegated); ++ } ++ dput(h_path.dentry); ++ ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * functions for removing a whiteout ++ */ ++ ++static int do_unlink_wh(struct inode *h_dir, struct path *h_path) ++{ ++ int err, force; ++ struct inode *delegated; ++ ++ /* ++ * forces superio when the dir has a sticky bit. ++ * this may be a violation of unix fs semantics. ++ */ ++ force = (h_dir->i_mode & S_ISVTX) ++ && !uid_eq(current_fsuid(), d_inode(h_path->dentry)->i_uid); ++ delegated = NULL; ++ err = vfsub_unlink(h_dir, h_path, &delegated, force); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ return err; ++} ++ ++int au_wh_unlink_dentry(struct inode *h_dir, struct path *h_path, ++ struct dentry *dentry) ++{ ++ int err; ++ ++ err = do_unlink_wh(h_dir, h_path); ++ if (!err && dentry) ++ au_set_dbwh(dentry, -1); ++ ++ return err; ++} ++ ++static int unlink_wh_name(struct dentry *h_parent, struct qstr *wh, ++ struct au_branch *br) ++{ ++ int err; ++ struct path h_path = { ++ .mnt = au_br_mnt(br) ++ }; ++ ++ err = 0; ++ h_path.dentry = vfsub_lkup_one(wh, h_parent); ++ if (IS_ERR(h_path.dentry)) ++ err = PTR_ERR(h_path.dentry); ++ else { ++ if (d_is_reg(h_path.dentry)) ++ err = do_unlink_wh(d_inode(h_parent), &h_path); ++ dput(h_path.dentry); ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * initialize/clean whiteout for a branch ++ */ ++ ++static void au_wh_clean(struct inode *h_dir, struct path *whpath, ++ const int isdir) ++{ ++ int err; ++ struct inode *delegated; ++ ++ if (d_is_negative(whpath->dentry)) ++ return; ++ ++ if (isdir) ++ err = vfsub_rmdir(h_dir, whpath); ++ else { ++ delegated = NULL; ++ err = vfsub_unlink(h_dir, whpath, &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ } ++ if (unlikely(err)) ++ pr_warn("failed removing %pd (%d), ignored.\n", ++ whpath->dentry, err); ++} ++ ++static int test_linkable(struct dentry *h_root) ++{ ++ struct inode *h_dir = d_inode(h_root); ++ ++ if (h_dir->i_op->link) ++ return 0; ++ ++ pr_err("%pd (%s) doesn't support link(2), use noplink and rw+nolwh\n", ++ h_root, au_sbtype(h_root->d_sb)); ++ return -ENOSYS; ++} ++ ++/* todo: should this mkdir be done in /sbin/mount.aufs helper? */ ++static int au_whdir(struct inode *h_dir, struct path *path) ++{ ++ int err; ++ ++ err = -EEXIST; ++ if (d_is_negative(path->dentry)) { ++ int mode = S_IRWXU; ++ ++ if (au_test_nfs(path->dentry->d_sb)) ++ mode |= S_IXUGO; ++ err = vfsub_mkdir(h_dir, path, mode); ++ } else if (d_is_dir(path->dentry)) ++ err = 0; ++ else ++ pr_err("unknown %pd exists\n", path->dentry); ++ ++ return err; ++} ++ ++struct au_wh_base { ++ const struct qstr *name; ++ struct dentry *dentry; ++}; ++ ++static void au_wh_init_ro(struct inode *h_dir, struct au_wh_base base[], ++ struct path *h_path) ++{ ++ h_path->dentry = base[AuBrWh_BASE].dentry; ++ au_wh_clean(h_dir, h_path, /*isdir*/0); ++ h_path->dentry = base[AuBrWh_PLINK].dentry; ++ au_wh_clean(h_dir, h_path, /*isdir*/1); ++ h_path->dentry = base[AuBrWh_ORPH].dentry; ++ au_wh_clean(h_dir, h_path, /*isdir*/1); ++} ++ ++/* ++ * returns tri-state, ++ * minus: error, caller should print the message ++ * zero: succuess ++ * plus: error, caller should NOT print the message ++ */ ++static int au_wh_init_rw_nolink(struct dentry *h_root, struct au_wbr *wbr, ++ int do_plink, struct au_wh_base base[], ++ struct path *h_path) ++{ ++ int err; ++ struct inode *h_dir; ++ ++ h_dir = d_inode(h_root); ++ h_path->dentry = base[AuBrWh_BASE].dentry; ++ au_wh_clean(h_dir, h_path, /*isdir*/0); ++ h_path->dentry = base[AuBrWh_PLINK].dentry; ++ if (do_plink) { ++ err = test_linkable(h_root); ++ if (unlikely(err)) { ++ err = 1; ++ goto out; ++ } ++ ++ err = au_whdir(h_dir, h_path); ++ if (unlikely(err)) ++ goto out; ++ wbr->wbr_plink = dget(base[AuBrWh_PLINK].dentry); ++ } else ++ au_wh_clean(h_dir, h_path, /*isdir*/1); ++ h_path->dentry = base[AuBrWh_ORPH].dentry; ++ err = au_whdir(h_dir, h_path); ++ if (unlikely(err)) ++ goto out; ++ wbr->wbr_orph = dget(base[AuBrWh_ORPH].dentry); ++ ++out: ++ return err; ++} ++ ++/* ++ * for the moment, aufs supports the branch filesystem which does not support ++ * link(2). testing on FAT which does not support i_op->setattr() fully either, ++ * copyup failed. finally, such filesystem will not be used as the writable ++ * branch. ++ * ++ * returns tri-state, see above. ++ */ ++static int au_wh_init_rw(struct dentry *h_root, struct au_wbr *wbr, ++ int do_plink, struct au_wh_base base[], ++ struct path *h_path) ++{ ++ int err; ++ struct inode *h_dir; ++ ++ WbrWhMustWriteLock(wbr); ++ ++ err = test_linkable(h_root); ++ if (unlikely(err)) { ++ err = 1; ++ goto out; ++ } ++ ++ /* ++ * todo: should this create be done in /sbin/mount.aufs helper? ++ */ ++ err = -EEXIST; ++ h_dir = d_inode(h_root); ++ if (d_is_negative(base[AuBrWh_BASE].dentry)) { ++ h_path->dentry = base[AuBrWh_BASE].dentry; ++ err = vfsub_create(h_dir, h_path, WH_MASK, /*want_excl*/true); ++ } else if (d_is_reg(base[AuBrWh_BASE].dentry)) ++ err = 0; ++ else ++ pr_err("unknown %pd2 exists\n", base[AuBrWh_BASE].dentry); ++ if (unlikely(err)) ++ goto out; ++ ++ h_path->dentry = base[AuBrWh_PLINK].dentry; ++ if (do_plink) { ++ err = au_whdir(h_dir, h_path); ++ if (unlikely(err)) ++ goto out; ++ wbr->wbr_plink = dget(base[AuBrWh_PLINK].dentry); ++ } else ++ au_wh_clean(h_dir, h_path, /*isdir*/1); ++ wbr->wbr_whbase = dget(base[AuBrWh_BASE].dentry); ++ ++ h_path->dentry = base[AuBrWh_ORPH].dentry; ++ err = au_whdir(h_dir, h_path); ++ if (unlikely(err)) ++ goto out; ++ wbr->wbr_orph = dget(base[AuBrWh_ORPH].dentry); ++ ++out: ++ return err; ++} ++ ++/* ++ * initialize the whiteout base file/dir for @br. ++ */ ++int au_wh_init(struct au_branch *br, struct super_block *sb) ++{ ++ int err, i; ++ const unsigned char do_plink ++ = !!au_opt_test(au_mntflags(sb), PLINK); ++ struct inode *h_dir; ++ struct path path = br->br_path; ++ struct dentry *h_root = path.dentry; ++ struct au_wbr *wbr = br->br_wbr; ++ static const struct qstr base_name[] = { ++ [AuBrWh_BASE] = QSTR_INIT(AUFS_BASE_NAME, ++ sizeof(AUFS_BASE_NAME) - 1), ++ [AuBrWh_PLINK] = QSTR_INIT(AUFS_PLINKDIR_NAME, ++ sizeof(AUFS_PLINKDIR_NAME) - 1), ++ [AuBrWh_ORPH] = QSTR_INIT(AUFS_ORPHDIR_NAME, ++ sizeof(AUFS_ORPHDIR_NAME) - 1) ++ }; ++ struct au_wh_base base[] = { ++ [AuBrWh_BASE] = { ++ .name = base_name + AuBrWh_BASE, ++ .dentry = NULL ++ }, ++ [AuBrWh_PLINK] = { ++ .name = base_name + AuBrWh_PLINK, ++ .dentry = NULL ++ }, ++ [AuBrWh_ORPH] = { ++ .name = base_name + AuBrWh_ORPH, ++ .dentry = NULL ++ } ++ }; ++ ++ if (wbr) ++ WbrWhMustWriteLock(wbr); ++ ++ for (i = 0; i < AuBrWh_Last; i++) { ++ /* doubly whiteouted */ ++ struct dentry *d; ++ ++ d = au_wh_lkup(h_root, (void *)base[i].name, br); ++ err = PTR_ERR(d); ++ if (IS_ERR(d)) ++ goto out; ++ ++ base[i].dentry = d; ++ AuDebugOn(wbr ++ && wbr->wbr_wh[i] ++ && wbr->wbr_wh[i] != base[i].dentry); ++ } ++ ++ if (wbr) ++ for (i = 0; i < AuBrWh_Last; i++) { ++ dput(wbr->wbr_wh[i]); ++ wbr->wbr_wh[i] = NULL; ++ } ++ ++ err = 0; ++ if (!au_br_writable(br->br_perm)) { ++ h_dir = d_inode(h_root); ++ au_wh_init_ro(h_dir, base, &path); ++ } else if (!au_br_wh_linkable(br->br_perm)) { ++ err = au_wh_init_rw_nolink(h_root, wbr, do_plink, base, &path); ++ if (err > 0) ++ goto out; ++ else if (err) ++ goto out_err; ++ } else { ++ err = au_wh_init_rw(h_root, wbr, do_plink, base, &path); ++ if (err > 0) ++ goto out; ++ else if (err) ++ goto out_err; ++ } ++ goto out; /* success */ ++ ++out_err: ++ pr_err("an error(%d) on the writable branch %pd(%s)\n", ++ err, h_root, au_sbtype(h_root->d_sb)); ++out: ++ for (i = 0; i < AuBrWh_Last; i++) ++ dput(base[i].dentry); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++/* ++ * whiteouts are all hard-linked usually. ++ * when its link count reaches a ceiling, we create a new whiteout base ++ * asynchronously. ++ */ ++ ++struct reinit_br_wh { ++ struct super_block *sb; ++ struct au_branch *br; ++}; ++ ++static void reinit_br_wh(void *arg) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct path h_path; ++ struct reinit_br_wh *a = arg; ++ struct au_wbr *wbr; ++ struct inode *dir, *delegated; ++ struct dentry *h_root; ++ struct au_hinode *hdir; ++ ++ err = 0; ++ wbr = a->br->br_wbr; ++ /* big aufs lock */ ++ si_noflush_write_lock(a->sb); ++ if (!au_br_writable(a->br->br_perm)) ++ goto out; ++ bindex = au_br_index(a->sb, a->br->br_id); ++ if (unlikely(bindex < 0)) ++ goto out; ++ ++ di_read_lock_parent(a->sb->s_root, AuLock_IR); ++ dir = d_inode(a->sb->s_root); ++ hdir = au_hi(dir, bindex); ++ h_root = au_h_dptr(a->sb->s_root, bindex); ++ AuDebugOn(h_root != au_br_dentry(a->br)); ++ ++ au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); ++ wbr_wh_write_lock(wbr); ++ err = au_h_verify(wbr->wbr_whbase, au_opt_udba(a->sb), hdir->hi_inode, ++ h_root, a->br); ++ if (!err) { ++ h_path.dentry = wbr->wbr_whbase; ++ h_path.mnt = au_br_mnt(a->br); ++ delegated = NULL; ++ err = vfsub_unlink(hdir->hi_inode, &h_path, &delegated, ++ /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ } else { ++ pr_warn("%pd is moved, ignored\n", wbr->wbr_whbase); ++ err = 0; ++ } ++ dput(wbr->wbr_whbase); ++ wbr->wbr_whbase = NULL; ++ if (!err) ++ err = au_wh_init(a->br, a->sb); ++ wbr_wh_write_unlock(wbr); ++ au_hn_imtx_unlock(hdir); ++ di_read_unlock(a->sb->s_root, AuLock_IR); ++ if (!err) ++ au_fhsm_wrote(a->sb, bindex, /*force*/0); ++ ++out: ++ if (wbr) ++ atomic_dec(&wbr->wbr_wh_running); ++ atomic_dec(&a->br->br_count); ++ si_write_unlock(a->sb); ++ au_nwt_done(&au_sbi(a->sb)->si_nowait); ++ kfree(arg); ++ if (unlikely(err)) ++ AuIOErr("err %d\n", err); ++} ++ ++static void kick_reinit_br_wh(struct super_block *sb, struct au_branch *br) ++{ ++ int do_dec, wkq_err; ++ struct reinit_br_wh *arg; ++ ++ do_dec = 1; ++ if (atomic_inc_return(&br->br_wbr->wbr_wh_running) != 1) ++ goto out; ++ ++ /* ignore ENOMEM */ ++ arg = kmalloc(sizeof(*arg), GFP_NOFS); ++ if (arg) { ++ /* ++ * dec(wh_running), kfree(arg) and dec(br_count) ++ * in reinit function ++ */ ++ arg->sb = sb; ++ arg->br = br; ++ atomic_inc(&br->br_count); ++ wkq_err = au_wkq_nowait(reinit_br_wh, arg, sb, /*flags*/0); ++ if (unlikely(wkq_err)) { ++ atomic_dec(&br->br_wbr->wbr_wh_running); ++ atomic_dec(&br->br_count); ++ kfree(arg); ++ } ++ do_dec = 0; ++ } ++ ++out: ++ if (do_dec) ++ atomic_dec(&br->br_wbr->wbr_wh_running); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * create the whiteout @wh. ++ */ ++static int link_or_create_wh(struct super_block *sb, aufs_bindex_t bindex, ++ struct dentry *wh) ++{ ++ int err; ++ struct path h_path = { ++ .dentry = wh ++ }; ++ struct au_branch *br; ++ struct au_wbr *wbr; ++ struct dentry *h_parent; ++ struct inode *h_dir, *delegated; ++ ++ h_parent = wh->d_parent; /* dir inode is locked */ ++ h_dir = d_inode(h_parent); ++ IMustLock(h_dir); ++ ++ br = au_sbr(sb, bindex); ++ h_path.mnt = au_br_mnt(br); ++ wbr = br->br_wbr; ++ wbr_wh_read_lock(wbr); ++ if (wbr->wbr_whbase) { ++ delegated = NULL; ++ err = vfsub_link(wbr->wbr_whbase, h_dir, &h_path, &delegated); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal link\n"); ++ iput(delegated); ++ } ++ if (!err || err != -EMLINK) ++ goto out; ++ ++ /* link count full. re-initialize br_whbase. */ ++ kick_reinit_br_wh(sb, br); ++ } ++ ++ /* return this error in this context */ ++ err = vfsub_create(h_dir, &h_path, WH_MASK, /*want_excl*/true); ++ if (!err) ++ au_fhsm_wrote(sb, bindex, /*force*/0); ++ ++out: ++ wbr_wh_read_unlock(wbr); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * create or remove the diropq. ++ */ ++static struct dentry *do_diropq(struct dentry *dentry, aufs_bindex_t bindex, ++ unsigned int flags) ++{ ++ struct dentry *opq_dentry, *h_dentry; ++ struct super_block *sb; ++ struct au_branch *br; ++ int err; ++ ++ sb = dentry->d_sb; ++ br = au_sbr(sb, bindex); ++ h_dentry = au_h_dptr(dentry, bindex); ++ opq_dentry = vfsub_lkup_one(&diropq_name, h_dentry); ++ if (IS_ERR(opq_dentry)) ++ goto out; ++ ++ if (au_ftest_diropq(flags, CREATE)) { ++ err = link_or_create_wh(sb, bindex, opq_dentry); ++ if (!err) { ++ au_set_dbdiropq(dentry, bindex); ++ goto out; /* success */ ++ } ++ } else { ++ struct path tmp = { ++ .dentry = opq_dentry, ++ .mnt = au_br_mnt(br) ++ }; ++ err = do_unlink_wh(au_h_iptr(d_inode(dentry), bindex), &tmp); ++ if (!err) ++ au_set_dbdiropq(dentry, -1); ++ } ++ dput(opq_dentry); ++ opq_dentry = ERR_PTR(err); ++ ++out: ++ return opq_dentry; ++} ++ ++struct do_diropq_args { ++ struct dentry **errp; ++ struct dentry *dentry; ++ aufs_bindex_t bindex; ++ unsigned int flags; ++}; ++ ++static void call_do_diropq(void *args) ++{ ++ struct do_diropq_args *a = args; ++ *a->errp = do_diropq(a->dentry, a->bindex, a->flags); ++} ++ ++struct dentry *au_diropq_sio(struct dentry *dentry, aufs_bindex_t bindex, ++ unsigned int flags) ++{ ++ struct dentry *diropq, *h_dentry; ++ ++ h_dentry = au_h_dptr(dentry, bindex); ++ if (!au_test_h_perm_sio(d_inode(h_dentry), MAY_EXEC | MAY_WRITE)) ++ diropq = do_diropq(dentry, bindex, flags); ++ else { ++ int wkq_err; ++ struct do_diropq_args args = { ++ .errp = &diropq, ++ .dentry = dentry, ++ .bindex = bindex, ++ .flags = flags ++ }; ++ ++ wkq_err = au_wkq_wait(call_do_diropq, &args); ++ if (unlikely(wkq_err)) ++ diropq = ERR_PTR(wkq_err); ++ } ++ ++ return diropq; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * lookup whiteout dentry. ++ * @h_parent: lower parent dentry which must exist and be locked ++ * @base_name: name of dentry which will be whiteouted ++ * returns dentry for whiteout. ++ */ ++struct dentry *au_wh_lkup(struct dentry *h_parent, struct qstr *base_name, ++ struct au_branch *br) ++{ ++ int err; ++ struct qstr wh_name; ++ struct dentry *wh_dentry; ++ ++ err = au_wh_name_alloc(&wh_name, base_name); ++ wh_dentry = ERR_PTR(err); ++ if (!err) { ++ wh_dentry = vfsub_lkup_one(&wh_name, h_parent); ++ kfree(wh_name.name); ++ } ++ return wh_dentry; ++} ++ ++/* ++ * link/create a whiteout for @dentry on @bindex. ++ */ ++struct dentry *au_wh_create(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent) ++{ ++ struct dentry *wh_dentry; ++ struct super_block *sb; ++ int err; ++ ++ sb = dentry->d_sb; ++ wh_dentry = au_wh_lkup(h_parent, &dentry->d_name, au_sbr(sb, bindex)); ++ if (!IS_ERR(wh_dentry) && d_is_negative(wh_dentry)) { ++ err = link_or_create_wh(sb, bindex, wh_dentry); ++ if (!err) { ++ au_set_dbwh(dentry, bindex); ++ au_fhsm_wrote(sb, bindex, /*force*/0); ++ } else { ++ dput(wh_dentry); ++ wh_dentry = ERR_PTR(err); ++ } ++ } ++ ++ return wh_dentry; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* Delete all whiteouts in this directory on branch bindex. */ ++static int del_wh_children(struct dentry *h_dentry, struct au_nhash *whlist, ++ aufs_bindex_t bindex, struct au_branch *br) ++{ ++ int err; ++ unsigned long ul, n; ++ struct qstr wh_name; ++ char *p; ++ struct hlist_head *head; ++ struct au_vdir_wh *pos; ++ struct au_vdir_destr *str; ++ ++ err = -ENOMEM; ++ p = (void *)__get_free_page(GFP_NOFS); ++ wh_name.name = p; ++ if (unlikely(!wh_name.name)) ++ goto out; ++ ++ err = 0; ++ memcpy(p, AUFS_WH_PFX, AUFS_WH_PFX_LEN); ++ p += AUFS_WH_PFX_LEN; ++ n = whlist->nh_num; ++ head = whlist->nh_head; ++ for (ul = 0; !err && ul < n; ul++, head++) { ++ hlist_for_each_entry(pos, head, wh_hash) { ++ if (pos->wh_bindex != bindex) ++ continue; ++ ++ str = &pos->wh_str; ++ if (str->len + AUFS_WH_PFX_LEN <= PATH_MAX) { ++ memcpy(p, str->name, str->len); ++ wh_name.len = AUFS_WH_PFX_LEN + str->len; ++ err = unlink_wh_name(h_dentry, &wh_name, br); ++ if (!err) ++ continue; ++ break; ++ } ++ AuIOErr("whiteout name too long %.*s\n", ++ str->len, str->name); ++ err = -EIO; ++ break; ++ } ++ } ++ free_page((unsigned long)wh_name.name); ++ ++out: ++ return err; ++} ++ ++struct del_wh_children_args { ++ int *errp; ++ struct dentry *h_dentry; ++ struct au_nhash *whlist; ++ aufs_bindex_t bindex; ++ struct au_branch *br; ++}; ++ ++static void call_del_wh_children(void *args) ++{ ++ struct del_wh_children_args *a = args; ++ *a->errp = del_wh_children(a->h_dentry, a->whlist, a->bindex, a->br); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct au_whtmp_rmdir *au_whtmp_rmdir_alloc(struct super_block *sb, gfp_t gfp) ++{ ++ struct au_whtmp_rmdir *whtmp; ++ int err; ++ unsigned int rdhash; ++ ++ SiMustAnyLock(sb); ++ ++ whtmp = kmalloc(sizeof(*whtmp), gfp); ++ if (unlikely(!whtmp)) { ++ whtmp = ERR_PTR(-ENOMEM); ++ goto out; ++ } ++ ++ whtmp->dir = NULL; ++ whtmp->br = NULL; ++ whtmp->wh_dentry = NULL; ++ /* no estimation for dir size */ ++ rdhash = au_sbi(sb)->si_rdhash; ++ if (!rdhash) ++ rdhash = AUFS_RDHASH_DEF; ++ err = au_nhash_alloc(&whtmp->whlist, rdhash, gfp); ++ if (unlikely(err)) { ++ kfree(whtmp); ++ whtmp = ERR_PTR(err); ++ } ++ ++out: ++ return whtmp; ++} ++ ++void au_whtmp_rmdir_free(struct au_whtmp_rmdir *whtmp) ++{ ++ if (whtmp->br) ++ atomic_dec(&whtmp->br->br_count); ++ dput(whtmp->wh_dentry); ++ iput(whtmp->dir); ++ au_nhash_wh_free(&whtmp->whlist); ++ kfree(whtmp); ++} ++ ++/* ++ * rmdir the whiteouted temporary named dir @h_dentry. ++ * @whlist: whiteouted children. ++ */ ++int au_whtmp_rmdir(struct inode *dir, aufs_bindex_t bindex, ++ struct dentry *wh_dentry, struct au_nhash *whlist) ++{ ++ int err; ++ unsigned int h_nlink; ++ struct path h_tmp; ++ struct inode *wh_inode, *h_dir; ++ struct au_branch *br; ++ ++ h_dir = d_inode(wh_dentry->d_parent); /* dir inode is locked */ ++ IMustLock(h_dir); ++ ++ br = au_sbr(dir->i_sb, bindex); ++ wh_inode = d_inode(wh_dentry); ++ mutex_lock_nested(&wh_inode->i_mutex, AuLsc_I_CHILD); ++ ++ /* ++ * someone else might change some whiteouts while we were sleeping. ++ * it means this whlist may have an obsoleted entry. ++ */ ++ if (!au_test_h_perm_sio(wh_inode, MAY_EXEC | MAY_WRITE)) ++ err = del_wh_children(wh_dentry, whlist, bindex, br); ++ else { ++ int wkq_err; ++ struct del_wh_children_args args = { ++ .errp = &err, ++ .h_dentry = wh_dentry, ++ .whlist = whlist, ++ .bindex = bindex, ++ .br = br ++ }; ++ ++ wkq_err = au_wkq_wait(call_del_wh_children, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ mutex_unlock(&wh_inode->i_mutex); ++ ++ if (!err) { ++ h_tmp.dentry = wh_dentry; ++ h_tmp.mnt = au_br_mnt(br); ++ h_nlink = h_dir->i_nlink; ++ err = vfsub_rmdir(h_dir, &h_tmp); ++ /* some fs doesn't change the parent nlink in some cases */ ++ h_nlink -= h_dir->i_nlink; ++ } ++ ++ if (!err) { ++ if (au_ibstart(dir) == bindex) { ++ /* todo: dir->i_mutex is necessary */ ++ au_cpup_attr_timesizes(dir); ++ if (h_nlink) ++ vfsub_drop_nlink(dir); ++ } ++ return 0; /* success */ ++ } ++ ++ pr_warn("failed removing %pd(%d), ignored\n", wh_dentry, err); ++ return err; ++} ++ ++static void call_rmdir_whtmp(void *args) ++{ ++ int err; ++ aufs_bindex_t bindex; ++ struct au_whtmp_rmdir *a = args; ++ struct super_block *sb; ++ struct dentry *h_parent; ++ struct inode *h_dir; ++ struct au_hinode *hdir; ++ ++ /* rmdir by nfsd may cause deadlock with this i_mutex */ ++ /* mutex_lock(&a->dir->i_mutex); */ ++ err = -EROFS; ++ sb = a->dir->i_sb; ++ si_read_lock(sb, !AuLock_FLUSH); ++ if (!au_br_writable(a->br->br_perm)) ++ goto out; ++ bindex = au_br_index(sb, a->br->br_id); ++ if (unlikely(bindex < 0)) ++ goto out; ++ ++ err = -EIO; ++ ii_write_lock_parent(a->dir); ++ h_parent = dget_parent(a->wh_dentry); ++ h_dir = d_inode(h_parent); ++ hdir = au_hi(a->dir, bindex); ++ err = vfsub_mnt_want_write(au_br_mnt(a->br)); ++ if (unlikely(err)) ++ goto out_mnt; ++ au_hn_imtx_lock_nested(hdir, AuLsc_I_PARENT); ++ err = au_h_verify(a->wh_dentry, au_opt_udba(sb), h_dir, h_parent, ++ a->br); ++ if (!err) ++ err = au_whtmp_rmdir(a->dir, bindex, a->wh_dentry, &a->whlist); ++ au_hn_imtx_unlock(hdir); ++ vfsub_mnt_drop_write(au_br_mnt(a->br)); ++ ++out_mnt: ++ dput(h_parent); ++ ii_write_unlock(a->dir); ++out: ++ /* mutex_unlock(&a->dir->i_mutex); */ ++ au_whtmp_rmdir_free(a); ++ si_read_unlock(sb); ++ au_nwt_done(&au_sbi(sb)->si_nowait); ++ if (unlikely(err)) ++ AuIOErr("err %d\n", err); ++} ++ ++void au_whtmp_kick_rmdir(struct inode *dir, aufs_bindex_t bindex, ++ struct dentry *wh_dentry, struct au_whtmp_rmdir *args) ++{ ++ int wkq_err; ++ struct super_block *sb; ++ ++ IMustLock(dir); ++ ++ /* all post-process will be done in do_rmdir_whtmp(). */ ++ sb = dir->i_sb; ++ args->dir = au_igrab(dir); ++ args->br = au_sbr(sb, bindex); ++ atomic_inc(&args->br->br_count); ++ args->wh_dentry = dget(wh_dentry); ++ wkq_err = au_wkq_nowait(call_rmdir_whtmp, args, sb, /*flags*/0); ++ if (unlikely(wkq_err)) { ++ pr_warn("rmdir error %pd (%d), ignored\n", wh_dentry, wkq_err); ++ au_whtmp_rmdir_free(args); ++ } ++} +diff -Nur linux-4.1.10.orig/fs/aufs/whout.h linux-4.1.10/fs/aufs/whout.h +--- linux-4.1.10.orig/fs/aufs/whout.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/whout.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,85 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * whiteout for logical deletion and opaque directory ++ */ ++ ++#ifndef __AUFS_WHOUT_H__ ++#define __AUFS_WHOUT_H__ ++ ++#ifdef __KERNEL__ ++ ++#include "dir.h" ++ ++/* whout.c */ ++int au_wh_name_alloc(struct qstr *wh, const struct qstr *name); ++int au_wh_test(struct dentry *h_parent, struct qstr *wh_name, int try_sio); ++int au_diropq_test(struct dentry *h_dentry); ++struct au_branch; ++struct dentry *au_whtmp_lkup(struct dentry *h_parent, struct au_branch *br, ++ struct qstr *prefix); ++int au_whtmp_ren(struct dentry *h_dentry, struct au_branch *br); ++int au_wh_unlink_dentry(struct inode *h_dir, struct path *h_path, ++ struct dentry *dentry); ++int au_wh_init(struct au_branch *br, struct super_block *sb); ++ ++/* diropq flags */ ++#define AuDiropq_CREATE 1 ++#define au_ftest_diropq(flags, name) ((flags) & AuDiropq_##name) ++#define au_fset_diropq(flags, name) \ ++ do { (flags) |= AuDiropq_##name; } while (0) ++#define au_fclr_diropq(flags, name) \ ++ do { (flags) &= ~AuDiropq_##name; } while (0) ++ ++struct dentry *au_diropq_sio(struct dentry *dentry, aufs_bindex_t bindex, ++ unsigned int flags); ++struct dentry *au_wh_lkup(struct dentry *h_parent, struct qstr *base_name, ++ struct au_branch *br); ++struct dentry *au_wh_create(struct dentry *dentry, aufs_bindex_t bindex, ++ struct dentry *h_parent); ++ ++/* real rmdir for the whiteout-ed dir */ ++struct au_whtmp_rmdir { ++ struct inode *dir; ++ struct au_branch *br; ++ struct dentry *wh_dentry; ++ struct au_nhash whlist; ++}; ++ ++struct au_whtmp_rmdir *au_whtmp_rmdir_alloc(struct super_block *sb, gfp_t gfp); ++void au_whtmp_rmdir_free(struct au_whtmp_rmdir *whtmp); ++int au_whtmp_rmdir(struct inode *dir, aufs_bindex_t bindex, ++ struct dentry *wh_dentry, struct au_nhash *whlist); ++void au_whtmp_kick_rmdir(struct inode *dir, aufs_bindex_t bindex, ++ struct dentry *wh_dentry, struct au_whtmp_rmdir *args); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline struct dentry *au_diropq_create(struct dentry *dentry, ++ aufs_bindex_t bindex) ++{ ++ return au_diropq_sio(dentry, bindex, AuDiropq_CREATE); ++} ++ ++static inline int au_diropq_remove(struct dentry *dentry, aufs_bindex_t bindex) ++{ ++ return PTR_ERR(au_diropq_sio(dentry, bindex, !AuDiropq_CREATE)); ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_WHOUT_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/wkq.c linux-4.1.10/fs/aufs/wkq.c +--- linux-4.1.10.orig/fs/aufs/wkq.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/wkq.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,213 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * workqueue for asynchronous/super-io operations ++ * todo: try new dredential scheme ++ */ ++ ++#include <linux/module.h> ++#include "aufs.h" ++ ++/* internal workqueue named AUFS_WKQ_NAME */ ++ ++static struct workqueue_struct *au_wkq; ++ ++struct au_wkinfo { ++ struct work_struct wk; ++ struct kobject *kobj; ++ ++ unsigned int flags; /* see wkq.h */ ++ ++ au_wkq_func_t func; ++ void *args; ++ ++ struct completion *comp; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void wkq_func(struct work_struct *wk) ++{ ++ struct au_wkinfo *wkinfo = container_of(wk, struct au_wkinfo, wk); ++ ++ AuDebugOn(!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)); ++ AuDebugOn(rlimit(RLIMIT_FSIZE) != RLIM_INFINITY); ++ ++ wkinfo->func(wkinfo->args); ++ if (au_ftest_wkq(wkinfo->flags, WAIT)) ++ complete(wkinfo->comp); ++ else { ++ kobject_put(wkinfo->kobj); ++ module_put(THIS_MODULE); /* todo: ?? */ ++ kfree(wkinfo); ++ } ++} ++ ++/* ++ * Since struct completion is large, try allocating it dynamically. ++ */ ++#if 1 /* defined(CONFIG_4KSTACKS) || defined(AuTest4KSTACKS) */ ++#define AuWkqCompDeclare(name) struct completion *comp = NULL ++ ++static int au_wkq_comp_alloc(struct au_wkinfo *wkinfo, struct completion **comp) ++{ ++ *comp = kmalloc(sizeof(**comp), GFP_NOFS); ++ if (*comp) { ++ init_completion(*comp); ++ wkinfo->comp = *comp; ++ return 0; ++ } ++ return -ENOMEM; ++} ++ ++static void au_wkq_comp_free(struct completion *comp) ++{ ++ kfree(comp); ++} ++ ++#else ++ ++/* no braces */ ++#define AuWkqCompDeclare(name) \ ++ DECLARE_COMPLETION_ONSTACK(_ ## name); \ ++ struct completion *comp = &_ ## name ++ ++static int au_wkq_comp_alloc(struct au_wkinfo *wkinfo, struct completion **comp) ++{ ++ wkinfo->comp = *comp; ++ return 0; ++} ++ ++static void au_wkq_comp_free(struct completion *comp __maybe_unused) ++{ ++ /* empty */ ++} ++#endif /* 4KSTACKS */ ++ ++static void au_wkq_run(struct au_wkinfo *wkinfo) ++{ ++ if (au_ftest_wkq(wkinfo->flags, NEST)) { ++ if (au_wkq_test()) { ++ AuWarn1("wkq from wkq, unless silly-rename on NFS," ++ " due to a dead dir by UDBA?\n"); ++ AuDebugOn(au_ftest_wkq(wkinfo->flags, WAIT)); ++ } ++ } else ++ au_dbg_verify_kthread(); ++ ++ if (au_ftest_wkq(wkinfo->flags, WAIT)) { ++ INIT_WORK_ONSTACK(&wkinfo->wk, wkq_func); ++ queue_work(au_wkq, &wkinfo->wk); ++ } else { ++ INIT_WORK(&wkinfo->wk, wkq_func); ++ schedule_work(&wkinfo->wk); ++ } ++} ++ ++/* ++ * Be careful. It is easy to make deadlock happen. ++ * processA: lock, wkq and wait ++ * processB: wkq and wait, lock in wkq ++ * --> deadlock ++ */ ++int au_wkq_do_wait(unsigned int flags, au_wkq_func_t func, void *args) ++{ ++ int err; ++ AuWkqCompDeclare(comp); ++ struct au_wkinfo wkinfo = { ++ .flags = flags, ++ .func = func, ++ .args = args ++ }; ++ ++ err = au_wkq_comp_alloc(&wkinfo, &comp); ++ if (!err) { ++ au_wkq_run(&wkinfo); ++ /* no timeout, no interrupt */ ++ wait_for_completion(wkinfo.comp); ++ au_wkq_comp_free(comp); ++ destroy_work_on_stack(&wkinfo.wk); ++ } ++ ++ return err; ++ ++} ++ ++/* ++ * Note: dget/dput() in func for aufs dentries are not supported. It will be a ++ * problem in a concurrent umounting. ++ */ ++int au_wkq_nowait(au_wkq_func_t func, void *args, struct super_block *sb, ++ unsigned int flags) ++{ ++ int err; ++ struct au_wkinfo *wkinfo; ++ ++ atomic_inc(&au_sbi(sb)->si_nowait.nw_len); ++ ++ /* ++ * wkq_func() must free this wkinfo. ++ * it highly depends upon the implementation of workqueue. ++ */ ++ err = 0; ++ wkinfo = kmalloc(sizeof(*wkinfo), GFP_NOFS); ++ if (wkinfo) { ++ wkinfo->kobj = &au_sbi(sb)->si_kobj; ++ wkinfo->flags = flags & ~AuWkq_WAIT; ++ wkinfo->func = func; ++ wkinfo->args = args; ++ wkinfo->comp = NULL; ++ kobject_get(wkinfo->kobj); ++ __module_get(THIS_MODULE); /* todo: ?? */ ++ ++ au_wkq_run(wkinfo); ++ } else { ++ err = -ENOMEM; ++ au_nwt_done(&au_sbi(sb)->si_nowait); ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++void au_nwt_init(struct au_nowait_tasks *nwt) ++{ ++ atomic_set(&nwt->nw_len, 0); ++ /* smp_mb(); */ /* atomic_set */ ++ init_waitqueue_head(&nwt->nw_wq); ++} ++ ++void au_wkq_fin(void) ++{ ++ destroy_workqueue(au_wkq); ++} ++ ++int __init au_wkq_init(void) ++{ ++ int err; ++ ++ err = 0; ++ au_wkq = alloc_workqueue(AUFS_WKQ_NAME, 0, WQ_DFL_ACTIVE); ++ if (IS_ERR(au_wkq)) ++ err = PTR_ERR(au_wkq); ++ else if (!au_wkq) ++ err = -ENOMEM; ++ ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/aufs/wkq.h linux-4.1.10/fs/aufs/wkq.h +--- linux-4.1.10.orig/fs/aufs/wkq.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/wkq.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,91 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * workqueue for asynchronous/super-io operations ++ * todo: try new credentials management scheme ++ */ ++ ++#ifndef __AUFS_WKQ_H__ ++#define __AUFS_WKQ_H__ ++ ++#ifdef __KERNEL__ ++ ++struct super_block; ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * in the next operation, wait for the 'nowait' tasks in system-wide workqueue ++ */ ++struct au_nowait_tasks { ++ atomic_t nw_len; ++ wait_queue_head_t nw_wq; ++}; ++ ++/* ---------------------------------------------------------------------- */ ++ ++typedef void (*au_wkq_func_t)(void *args); ++ ++/* wkq flags */ ++#define AuWkq_WAIT 1 ++#define AuWkq_NEST (1 << 1) ++#define au_ftest_wkq(flags, name) ((flags) & AuWkq_##name) ++#define au_fset_wkq(flags, name) \ ++ do { (flags) |= AuWkq_##name; } while (0) ++#define au_fclr_wkq(flags, name) \ ++ do { (flags) &= ~AuWkq_##name; } while (0) ++ ++#ifndef CONFIG_AUFS_HNOTIFY ++#undef AuWkq_NEST ++#define AuWkq_NEST 0 ++#endif ++ ++/* wkq.c */ ++int au_wkq_do_wait(unsigned int flags, au_wkq_func_t func, void *args); ++int au_wkq_nowait(au_wkq_func_t func, void *args, struct super_block *sb, ++ unsigned int flags); ++void au_nwt_init(struct au_nowait_tasks *nwt); ++int __init au_wkq_init(void); ++void au_wkq_fin(void); ++ ++/* ---------------------------------------------------------------------- */ ++ ++static inline int au_wkq_test(void) ++{ ++ return current->flags & PF_WQ_WORKER; ++} ++ ++static inline int au_wkq_wait(au_wkq_func_t func, void *args) ++{ ++ return au_wkq_do_wait(AuWkq_WAIT, func, args); ++} ++ ++static inline void au_nwt_done(struct au_nowait_tasks *nwt) ++{ ++ if (atomic_dec_and_test(&nwt->nw_len)) ++ wake_up_all(&nwt->nw_wq); ++} ++ ++static inline int au_nwt_flush(struct au_nowait_tasks *nwt) ++{ ++ wait_event(nwt->nw_wq, !atomic_read(&nwt->nw_len)); ++ return 0; ++} ++ ++#endif /* __KERNEL__ */ ++#endif /* __AUFS_WKQ_H__ */ +diff -Nur linux-4.1.10.orig/fs/aufs/xattr.c linux-4.1.10/fs/aufs/xattr.c +--- linux-4.1.10.orig/fs/aufs/xattr.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/xattr.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,344 @@ ++/* ++ * Copyright (C) 2014-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * handling xattr functions ++ */ ++ ++#include <linux/xattr.h> ++#include "aufs.h" ++ ++static int au_xattr_ignore(int err, char *name, unsigned int ignore_flags) ++{ ++ if (!ignore_flags) ++ goto out; ++ switch (err) { ++ case -ENOMEM: ++ case -EDQUOT: ++ goto out; ++ } ++ ++ if ((ignore_flags & AuBrAttr_ICEX) == AuBrAttr_ICEX) { ++ err = 0; ++ goto out; ++ } ++ ++#define cmp(brattr, prefix) do { \ ++ if (!strncmp(name, XATTR_##prefix##_PREFIX, \ ++ XATTR_##prefix##_PREFIX_LEN)) { \ ++ if (ignore_flags & AuBrAttr_ICEX_##brattr) \ ++ err = 0; \ ++ goto out; \ ++ } \ ++ } while (0) ++ ++ cmp(SEC, SECURITY); ++ cmp(SYS, SYSTEM); ++ cmp(TR, TRUSTED); ++ cmp(USR, USER); ++#undef cmp ++ ++ if (ignore_flags & AuBrAttr_ICEX_OTH) ++ err = 0; ++ ++out: ++ return err; ++} ++ ++static const int au_xattr_out_of_list = AuBrAttr_ICEX_OTH << 1; ++ ++static int au_do_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, ++ char *name, char **buf, unsigned int ignore_flags, ++ unsigned int verbose) ++{ ++ int err; ++ ssize_t ssz; ++ struct inode *h_idst; ++ ++ ssz = vfs_getxattr_alloc(h_src, name, buf, 0, GFP_NOFS); ++ err = ssz; ++ if (unlikely(err <= 0)) { ++ if (err == -ENODATA ++ || (err == -EOPNOTSUPP ++ && ((ignore_flags & au_xattr_out_of_list) ++ || (au_test_nfs_noacl(d_inode(h_src)) ++ && (!strcmp(name, XATTR_NAME_POSIX_ACL_ACCESS) ++ || !strcmp(name, ++ XATTR_NAME_POSIX_ACL_DEFAULT)))) ++ )) ++ err = 0; ++ if (err && (verbose || au_debug_test())) ++ pr_err("%s, err %d\n", name, err); ++ goto out; ++ } ++ ++ /* unlock it temporary */ ++ h_idst = d_inode(h_dst); ++ mutex_unlock(&h_idst->i_mutex); ++ err = vfsub_setxattr(h_dst, name, *buf, ssz, /*flags*/0); ++ mutex_lock_nested(&h_idst->i_mutex, AuLsc_I_CHILD2); ++ if (unlikely(err)) { ++ if (verbose || au_debug_test()) ++ pr_err("%s, err %d\n", name, err); ++ err = au_xattr_ignore(err, name, ignore_flags); ++ } ++ ++out: ++ return err; ++} ++ ++int au_cpup_xattr(struct dentry *h_dst, struct dentry *h_src, int ignore_flags, ++ unsigned int verbose) ++{ ++ int err, unlocked, acl_access, acl_default; ++ ssize_t ssz; ++ struct inode *h_isrc, *h_idst; ++ char *value, *p, *o, *e; ++ ++ /* try stopping to update the source inode while we are referencing */ ++ /* there should not be the parent-child relationship between them */ ++ h_isrc = d_inode(h_src); ++ h_idst = d_inode(h_dst); ++ mutex_unlock(&h_idst->i_mutex); ++ mutex_lock_nested(&h_isrc->i_mutex, AuLsc_I_CHILD); ++ mutex_lock_nested(&h_idst->i_mutex, AuLsc_I_CHILD2); ++ unlocked = 0; ++ ++ /* some filesystems don't list POSIX ACL, for example tmpfs */ ++ ssz = vfs_listxattr(h_src, NULL, 0); ++ err = ssz; ++ if (unlikely(err < 0)) { ++ AuTraceErr(err); ++ if (err == -ENODATA ++ || err == -EOPNOTSUPP) ++ err = 0; /* ignore */ ++ goto out; ++ } ++ ++ err = 0; ++ p = NULL; ++ o = NULL; ++ if (ssz) { ++ err = -ENOMEM; ++ p = kmalloc(ssz, GFP_NOFS); ++ o = p; ++ if (unlikely(!p)) ++ goto out; ++ err = vfs_listxattr(h_src, p, ssz); ++ } ++ mutex_unlock(&h_isrc->i_mutex); ++ unlocked = 1; ++ AuDbg("err %d, ssz %zd\n", err, ssz); ++ if (unlikely(err < 0)) ++ goto out_free; ++ ++ err = 0; ++ e = p + ssz; ++ value = NULL; ++ acl_access = 0; ++ acl_default = 0; ++ while (!err && p < e) { ++ acl_access |= !strncmp(p, XATTR_NAME_POSIX_ACL_ACCESS, ++ sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1); ++ acl_default |= !strncmp(p, XATTR_NAME_POSIX_ACL_DEFAULT, ++ sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) ++ - 1); ++ err = au_do_cpup_xattr(h_dst, h_src, p, &value, ignore_flags, ++ verbose); ++ p += strlen(p) + 1; ++ } ++ AuTraceErr(err); ++ ignore_flags |= au_xattr_out_of_list; ++ if (!err && !acl_access) { ++ err = au_do_cpup_xattr(h_dst, h_src, ++ XATTR_NAME_POSIX_ACL_ACCESS, &value, ++ ignore_flags, verbose); ++ AuTraceErr(err); ++ } ++ if (!err && !acl_default) { ++ err = au_do_cpup_xattr(h_dst, h_src, ++ XATTR_NAME_POSIX_ACL_DEFAULT, &value, ++ ignore_flags, verbose); ++ AuTraceErr(err); ++ } ++ ++ kfree(value); ++ ++out_free: ++ kfree(o); ++out: ++ if (!unlocked) ++ mutex_unlock(&h_isrc->i_mutex); ++ AuTraceErr(err); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++enum { ++ AU_XATTR_LIST, ++ AU_XATTR_GET ++}; ++ ++struct au_lgxattr { ++ int type; ++ union { ++ struct { ++ char *list; ++ size_t size; ++ } list; ++ struct { ++ const char *name; ++ void *value; ++ size_t size; ++ } get; ++ } u; ++}; ++ ++static ssize_t au_lgxattr(struct dentry *dentry, struct au_lgxattr *arg) ++{ ++ ssize_t err; ++ struct path h_path; ++ struct super_block *sb; ++ ++ sb = dentry->d_sb; ++ err = si_read_lock(sb, AuLock_FLUSH | AuLock_NOPLM); ++ if (unlikely(err)) ++ goto out; ++ err = au_h_path_getattr(dentry, /*force*/1, &h_path); ++ if (unlikely(err)) ++ goto out_si; ++ if (unlikely(!h_path.dentry)) ++ /* illegally overlapped or something */ ++ goto out_di; /* pretending success */ ++ ++ /* always topmost entry only */ ++ switch (arg->type) { ++ case AU_XATTR_LIST: ++ err = vfs_listxattr(h_path.dentry, ++ arg->u.list.list, arg->u.list.size); ++ break; ++ case AU_XATTR_GET: ++ err = vfs_getxattr(h_path.dentry, ++ arg->u.get.name, arg->u.get.value, ++ arg->u.get.size); ++ break; ++ } ++ ++out_di: ++ di_read_unlock(dentry, AuLock_IR); ++out_si: ++ si_read_unlock(sb); ++out: ++ AuTraceErr(err); ++ return err; ++} ++ ++ssize_t aufs_listxattr(struct dentry *dentry, char *list, size_t size) ++{ ++ struct au_lgxattr arg = { ++ .type = AU_XATTR_LIST, ++ .u.list = { ++ .list = list, ++ .size = size ++ }, ++ }; ++ ++ return au_lgxattr(dentry, &arg); ++} ++ ++ssize_t aufs_getxattr(struct dentry *dentry, const char *name, void *value, ++ size_t size) ++{ ++ struct au_lgxattr arg = { ++ .type = AU_XATTR_GET, ++ .u.get = { ++ .name = name, ++ .value = value, ++ .size = size ++ }, ++ }; ++ ++ return au_lgxattr(dentry, &arg); ++} ++ ++int aufs_setxattr(struct dentry *dentry, const char *name, const void *value, ++ size_t size, int flags) ++{ ++ struct au_srxattr arg = { ++ .type = AU_XATTR_SET, ++ .u.set = { ++ .name = name, ++ .value = value, ++ .size = size, ++ .flags = flags ++ }, ++ }; ++ ++ return au_srxattr(dentry, &arg); ++} ++ ++int aufs_removexattr(struct dentry *dentry, const char *name) ++{ ++ struct au_srxattr arg = { ++ .type = AU_XATTR_REMOVE, ++ .u.remove = { ++ .name = name ++ }, ++ }; ++ ++ return au_srxattr(dentry, &arg); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++#if 0 ++static size_t au_xattr_list(struct dentry *dentry, char *list, size_t list_size, ++ const char *name, size_t name_len, int type) ++{ ++ return aufs_listxattr(dentry, list, list_size); ++} ++ ++static int au_xattr_get(struct dentry *dentry, const char *name, void *buffer, ++ size_t size, int type) ++{ ++ return aufs_getxattr(dentry, name, buffer, size); ++} ++ ++static int au_xattr_set(struct dentry *dentry, const char *name, ++ const void *value, size_t size, int flags, int type) ++{ ++ return aufs_setxattr(dentry, name, value, size, flags); ++} ++ ++static const struct xattr_handler au_xattr_handler = { ++ /* no prefix, no flags */ ++ .list = au_xattr_list, ++ .get = au_xattr_get, ++ .set = au_xattr_set ++ /* why no remove? */ ++}; ++ ++static const struct xattr_handler *au_xattr_handlers[] = { ++ &au_xattr_handler ++}; ++ ++void au_xattr_init(struct super_block *sb) ++{ ++ /* sb->s_xattr = au_xattr_handlers; */ ++} ++#endif +diff -Nur linux-4.1.10.orig/fs/aufs/xino.c linux-4.1.10/fs/aufs/xino.c +--- linux-4.1.10.orig/fs/aufs/xino.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/fs/aufs/xino.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,1296 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++/* ++ * external inode number translation table and bitmap ++ */ ++ ++#include <linux/seq_file.h> ++#include <linux/statfs.h> ++#include "aufs.h" ++ ++/* todo: unnecessary to support mmap_sem since kernel-space? */ ++ssize_t xino_fread(vfs_readf_t func, struct file *file, void *kbuf, size_t size, ++ loff_t *pos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ char __user *u; ++ } buf; ++ ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ do { ++ /* todo: signal_pending? */ ++ err = func(file, buf.u, size, pos); ++ } while (err == -EAGAIN || err == -EINTR); ++ set_fs(oldfs); ++ ++#if 0 /* reserved for future use */ ++ if (err > 0) ++ fsnotify_access(file->f_path.dentry); ++#endif ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static ssize_t do_xino_fwrite(vfs_writef_t func, struct file *file, void *kbuf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; ++ mm_segment_t oldfs; ++ union { ++ void *k; ++ const char __user *u; ++ } buf; ++ ++ buf.k = kbuf; ++ oldfs = get_fs(); ++ set_fs(KERNEL_DS); ++ do { ++ /* todo: signal_pending? */ ++ err = func(file, buf.u, size, pos); ++ } while (err == -EAGAIN || err == -EINTR); ++ set_fs(oldfs); ++ ++#if 0 /* reserved for future use */ ++ if (err > 0) ++ fsnotify_modify(file->f_path.dentry); ++#endif ++ ++ return err; ++} ++ ++struct do_xino_fwrite_args { ++ ssize_t *errp; ++ vfs_writef_t func; ++ struct file *file; ++ void *buf; ++ size_t size; ++ loff_t *pos; ++}; ++ ++static void call_do_xino_fwrite(void *args) ++{ ++ struct do_xino_fwrite_args *a = args; ++ *a->errp = do_xino_fwrite(a->func, a->file, a->buf, a->size, a->pos); ++} ++ ++ssize_t xino_fwrite(vfs_writef_t func, struct file *file, void *buf, ++ size_t size, loff_t *pos) ++{ ++ ssize_t err; ++ ++ /* todo: signal block and no wkq? */ ++ if (rlimit(RLIMIT_FSIZE) == RLIM_INFINITY) { ++ lockdep_off(); ++ err = do_xino_fwrite(func, file, buf, size, pos); ++ lockdep_on(); ++ } else { ++ /* ++ * it breaks RLIMIT_FSIZE and normal user's limit, ++ * users should care about quota and real 'filesystem full.' ++ */ ++ int wkq_err; ++ struct do_xino_fwrite_args args = { ++ .errp = &err, ++ .func = func, ++ .file = file, ++ .buf = buf, ++ .size = size, ++ .pos = pos ++ }; ++ ++ wkq_err = au_wkq_wait(call_do_xino_fwrite, &args); ++ if (unlikely(wkq_err)) ++ err = wkq_err; ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * create a new xinofile at the same place/path as @base_file. ++ */ ++struct file *au_xino_create2(struct file *base_file, struct file *copy_src) ++{ ++ struct file *file; ++ struct dentry *base, *parent; ++ struct inode *dir, *delegated; ++ struct qstr *name; ++ struct path path; ++ int err; ++ ++ base = base_file->f_path.dentry; ++ parent = base->d_parent; /* dir inode is locked */ ++ dir = d_inode(parent); ++ IMustLock(dir); ++ ++ file = ERR_PTR(-EINVAL); ++ name = &base->d_name; ++ path.dentry = vfsub_lookup_one_len(name->name, parent, name->len); ++ if (IS_ERR(path.dentry)) { ++ file = (void *)path.dentry; ++ pr_err("%pd lookup err %ld\n", ++ base, PTR_ERR(path.dentry)); ++ goto out; ++ } ++ ++ /* no need to mnt_want_write() since we call dentry_open() later */ ++ err = vfs_create(dir, path.dentry, S_IRUGO | S_IWUGO, NULL); ++ if (unlikely(err)) { ++ file = ERR_PTR(err); ++ pr_err("%pd create err %d\n", base, err); ++ goto out_dput; ++ } ++ ++ path.mnt = base_file->f_path.mnt; ++ file = vfsub_dentry_open(&path, ++ O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE ++ /* | __FMODE_NONOTIFY */); ++ if (IS_ERR(file)) { ++ pr_err("%pd open err %ld\n", base, PTR_ERR(file)); ++ goto out_dput; ++ } ++ ++ delegated = NULL; ++ err = vfsub_unlink(dir, &file->f_path, &delegated, /*force*/0); ++ if (unlikely(err == -EWOULDBLOCK)) { ++ pr_warn("cannot retry for NFSv4 delegation" ++ " for an internal unlink\n"); ++ iput(delegated); ++ } ++ if (unlikely(err)) { ++ pr_err("%pd unlink err %d\n", base, err); ++ goto out_fput; ++ } ++ ++ if (copy_src) { ++ /* no one can touch copy_src xino */ ++ err = au_copy_file(file, copy_src, vfsub_f_size_read(copy_src)); ++ if (unlikely(err)) { ++ pr_err("%pd copy err %d\n", base, err); ++ goto out_fput; ++ } ++ } ++ goto out_dput; /* success */ ++ ++out_fput: ++ fput(file); ++ file = ERR_PTR(err); ++out_dput: ++ dput(path.dentry); ++out: ++ return file; ++} ++ ++struct au_xino_lock_dir { ++ struct au_hinode *hdir; ++ struct dentry *parent; ++ struct mutex *mtx; ++}; ++ ++static void au_xino_lock_dir(struct super_block *sb, struct file *xino, ++ struct au_xino_lock_dir *ldir) ++{ ++ aufs_bindex_t brid, bindex; ++ ++ ldir->hdir = NULL; ++ bindex = -1; ++ brid = au_xino_brid(sb); ++ if (brid >= 0) ++ bindex = au_br_index(sb, brid); ++ if (bindex >= 0) { ++ ldir->hdir = au_hi(d_inode(sb->s_root), bindex); ++ au_hn_imtx_lock_nested(ldir->hdir, AuLsc_I_PARENT); ++ } else { ++ ldir->parent = dget_parent(xino->f_path.dentry); ++ ldir->mtx = &d_inode(ldir->parent)->i_mutex; ++ mutex_lock_nested(ldir->mtx, AuLsc_I_PARENT); ++ } ++} ++ ++static void au_xino_unlock_dir(struct au_xino_lock_dir *ldir) ++{ ++ if (ldir->hdir) ++ au_hn_imtx_unlock(ldir->hdir); ++ else { ++ mutex_unlock(ldir->mtx); ++ dput(ldir->parent); ++ } ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* trucate xino files asynchronously */ ++ ++int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex) ++{ ++ int err; ++ unsigned long jiffy; ++ blkcnt_t blocks; ++ aufs_bindex_t bi, bend; ++ struct kstatfs *st; ++ struct au_branch *br; ++ struct file *new_xino, *file; ++ struct super_block *h_sb; ++ struct au_xino_lock_dir ldir; ++ ++ err = -ENOMEM; ++ st = kzalloc(sizeof(*st), GFP_NOFS); ++ if (unlikely(!st)) ++ goto out; ++ ++ err = -EINVAL; ++ bend = au_sbend(sb); ++ if (unlikely(bindex < 0 || bend < bindex)) ++ goto out_st; ++ br = au_sbr(sb, bindex); ++ file = br->br_xino.xi_file; ++ if (!file) ++ goto out_st; ++ ++ err = vfs_statfs(&file->f_path, st); ++ if (unlikely(err)) ++ AuErr1("statfs err %d, ignored\n", err); ++ jiffy = jiffies; ++ blocks = file_inode(file)->i_blocks; ++ pr_info("begin truncating xino(b%d), ib%llu, %llu/%llu free blks\n", ++ bindex, (u64)blocks, st->f_bfree, st->f_blocks); ++ ++ au_xino_lock_dir(sb, file, &ldir); ++ /* mnt_want_write() is unnecessary here */ ++ new_xino = au_xino_create2(file, file); ++ au_xino_unlock_dir(&ldir); ++ err = PTR_ERR(new_xino); ++ if (IS_ERR(new_xino)) { ++ pr_err("err %d, ignored\n", err); ++ goto out_st; ++ } ++ err = 0; ++ fput(file); ++ br->br_xino.xi_file = new_xino; ++ ++ h_sb = au_br_sb(br); ++ for (bi = 0; bi <= bend; bi++) { ++ if (unlikely(bi == bindex)) ++ continue; ++ br = au_sbr(sb, bi); ++ if (au_br_sb(br) != h_sb) ++ continue; ++ ++ fput(br->br_xino.xi_file); ++ br->br_xino.xi_file = new_xino; ++ get_file(new_xino); ++ } ++ ++ err = vfs_statfs(&new_xino->f_path, st); ++ if (!err) { ++ pr_info("end truncating xino(b%d), ib%llu, %llu/%llu free blks\n", ++ bindex, (u64)file_inode(new_xino)->i_blocks, ++ st->f_bfree, st->f_blocks); ++ if (file_inode(new_xino)->i_blocks < blocks) ++ au_sbi(sb)->si_xino_jiffy = jiffy; ++ } else ++ AuErr1("statfs err %d, ignored\n", err); ++ ++out_st: ++ kfree(st); ++out: ++ return err; ++} ++ ++struct xino_do_trunc_args { ++ struct super_block *sb; ++ struct au_branch *br; ++}; ++ ++static void xino_do_trunc(void *_args) ++{ ++ struct xino_do_trunc_args *args = _args; ++ struct super_block *sb; ++ struct au_branch *br; ++ struct inode *dir; ++ int err; ++ aufs_bindex_t bindex; ++ ++ err = 0; ++ sb = args->sb; ++ dir = d_inode(sb->s_root); ++ br = args->br; ++ ++ si_noflush_write_lock(sb); ++ ii_read_lock_parent(dir); ++ bindex = au_br_index(sb, br->br_id); ++ err = au_xino_trunc(sb, bindex); ++ ii_read_unlock(dir); ++ if (unlikely(err)) ++ pr_warn("err b%d, (%d)\n", bindex, err); ++ atomic_dec(&br->br_xino_running); ++ atomic_dec(&br->br_count); ++ si_write_unlock(sb); ++ au_nwt_done(&au_sbi(sb)->si_nowait); ++ kfree(args); ++} ++ ++static int xino_trunc_test(struct super_block *sb, struct au_branch *br) ++{ ++ int err; ++ struct kstatfs st; ++ struct au_sbinfo *sbinfo; ++ ++ /* todo: si_xino_expire and the ratio should be customizable */ ++ sbinfo = au_sbi(sb); ++ if (time_before(jiffies, ++ sbinfo->si_xino_jiffy + sbinfo->si_xino_expire)) ++ return 0; ++ ++ /* truncation border */ ++ err = vfs_statfs(&br->br_xino.xi_file->f_path, &st); ++ if (unlikely(err)) { ++ AuErr1("statfs err %d, ignored\n", err); ++ return 0; ++ } ++ if (div64_u64(st.f_bfree * 100, st.f_blocks) >= AUFS_XINO_DEF_TRUNC) ++ return 0; ++ ++ return 1; ++} ++ ++static void xino_try_trunc(struct super_block *sb, struct au_branch *br) ++{ ++ struct xino_do_trunc_args *args; ++ int wkq_err; ++ ++ if (!xino_trunc_test(sb, br)) ++ return; ++ ++ if (atomic_inc_return(&br->br_xino_running) > 1) ++ goto out; ++ ++ /* lock and kfree() will be called in trunc_xino() */ ++ args = kmalloc(sizeof(*args), GFP_NOFS); ++ if (unlikely(!args)) { ++ AuErr1("no memory\n"); ++ goto out_args; ++ } ++ ++ atomic_inc(&br->br_count); ++ args->sb = sb; ++ args->br = br; ++ wkq_err = au_wkq_nowait(xino_do_trunc, args, sb, /*flags*/0); ++ if (!wkq_err) ++ return; /* success */ ++ ++ pr_err("wkq %d\n", wkq_err); ++ atomic_dec(&br->br_count); ++ ++out_args: ++ kfree(args); ++out: ++ atomic_dec(&br->br_xino_running); ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static int au_xino_do_write(vfs_writef_t write, struct file *file, ++ ino_t h_ino, ino_t ino) ++{ ++ loff_t pos; ++ ssize_t sz; ++ ++ pos = h_ino; ++ if (unlikely(au_loff_max / sizeof(ino) - 1 < pos)) { ++ AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); ++ return -EFBIG; ++ } ++ pos *= sizeof(ino); ++ sz = xino_fwrite(write, file, &ino, sizeof(ino), &pos); ++ if (sz == sizeof(ino)) ++ return 0; /* success */ ++ ++ AuIOErr("write failed (%zd)\n", sz); ++ return -EIO; ++} ++ ++/* ++ * write @ino to the xinofile for the specified branch{@sb, @bindex} ++ * at the position of @h_ino. ++ * even if @ino is zero, it is written to the xinofile and means no entry. ++ * if the size of the xino file on a specific filesystem exceeds the watermark, ++ * try truncating it. ++ */ ++int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t ino) ++{ ++ int err; ++ unsigned int mnt_flags; ++ struct au_branch *br; ++ ++ BUILD_BUG_ON(sizeof(long long) != sizeof(au_loff_max) ++ || ((loff_t)-1) > 0); ++ SiMustAnyLock(sb); ++ ++ mnt_flags = au_mntflags(sb); ++ if (!au_opt_test(mnt_flags, XINO)) ++ return 0; ++ ++ br = au_sbr(sb, bindex); ++ err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, ++ h_ino, ino); ++ if (!err) { ++ if (au_opt_test(mnt_flags, TRUNC_XINO) ++ && au_test_fs_trunc_xino(au_br_sb(br))) ++ xino_try_trunc(sb, br); ++ return 0; /* success */ ++ } ++ ++ AuIOErr("write failed (%d)\n", err); ++ return -EIO; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* aufs inode number bitmap */ ++ ++static const int page_bits = (int)PAGE_SIZE * BITS_PER_BYTE; ++static ino_t xib_calc_ino(unsigned long pindex, int bit) ++{ ++ ino_t ino; ++ ++ AuDebugOn(bit < 0 || page_bits <= bit); ++ ino = AUFS_FIRST_INO + pindex * page_bits + bit; ++ return ino; ++} ++ ++static void xib_calc_bit(ino_t ino, unsigned long *pindex, int *bit) ++{ ++ AuDebugOn(ino < AUFS_FIRST_INO); ++ ino -= AUFS_FIRST_INO; ++ *pindex = ino / page_bits; ++ *bit = ino % page_bits; ++} ++ ++static int xib_pindex(struct super_block *sb, unsigned long pindex) ++{ ++ int err; ++ loff_t pos; ++ ssize_t sz; ++ struct au_sbinfo *sbinfo; ++ struct file *xib; ++ unsigned long *p; ++ ++ sbinfo = au_sbi(sb); ++ MtxMustLock(&sbinfo->si_xib_mtx); ++ AuDebugOn(pindex > ULONG_MAX / PAGE_SIZE ++ || !au_opt_test(sbinfo->si_mntflags, XINO)); ++ ++ if (pindex == sbinfo->si_xib_last_pindex) ++ return 0; ++ ++ xib = sbinfo->si_xib; ++ p = sbinfo->si_xib_buf; ++ pos = sbinfo->si_xib_last_pindex; ++ pos *= PAGE_SIZE; ++ sz = xino_fwrite(sbinfo->si_xwrite, xib, p, PAGE_SIZE, &pos); ++ if (unlikely(sz != PAGE_SIZE)) ++ goto out; ++ ++ pos = pindex; ++ pos *= PAGE_SIZE; ++ if (vfsub_f_size_read(xib) >= pos + PAGE_SIZE) ++ sz = xino_fread(sbinfo->si_xread, xib, p, PAGE_SIZE, &pos); ++ else { ++ memset(p, 0, PAGE_SIZE); ++ sz = xino_fwrite(sbinfo->si_xwrite, xib, p, PAGE_SIZE, &pos); ++ } ++ if (sz == PAGE_SIZE) { ++ sbinfo->si_xib_last_pindex = pindex; ++ return 0; /* success */ ++ } ++ ++out: ++ AuIOErr1("write failed (%zd)\n", sz); ++ err = sz; ++ if (sz >= 0) ++ err = -EIO; ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++static void au_xib_clear_bit(struct inode *inode) ++{ ++ int err, bit; ++ unsigned long pindex; ++ struct super_block *sb; ++ struct au_sbinfo *sbinfo; ++ ++ AuDebugOn(inode->i_nlink); ++ ++ sb = inode->i_sb; ++ xib_calc_bit(inode->i_ino, &pindex, &bit); ++ AuDebugOn(page_bits <= bit); ++ sbinfo = au_sbi(sb); ++ mutex_lock(&sbinfo->si_xib_mtx); ++ err = xib_pindex(sb, pindex); ++ if (!err) { ++ clear_bit(bit, sbinfo->si_xib_buf); ++ sbinfo->si_xib_next_bit = bit; ++ } ++ mutex_unlock(&sbinfo->si_xib_mtx); ++} ++ ++/* for s_op->delete_inode() */ ++void au_xino_delete_inode(struct inode *inode, const int unlinked) ++{ ++ int err; ++ unsigned int mnt_flags; ++ aufs_bindex_t bindex, bend, bi; ++ unsigned char try_trunc; ++ struct au_iinfo *iinfo; ++ struct super_block *sb; ++ struct au_hinode *hi; ++ struct inode *h_inode; ++ struct au_branch *br; ++ vfs_writef_t xwrite; ++ ++ sb = inode->i_sb; ++ mnt_flags = au_mntflags(sb); ++ if (!au_opt_test(mnt_flags, XINO) ++ || inode->i_ino == AUFS_ROOT_INO) ++ return; ++ ++ if (unlinked) { ++ au_xigen_inc(inode); ++ au_xib_clear_bit(inode); ++ } ++ ++ iinfo = au_ii(inode); ++ if (!iinfo) ++ return; ++ ++ bindex = iinfo->ii_bstart; ++ if (bindex < 0) ++ return; ++ ++ xwrite = au_sbi(sb)->si_xwrite; ++ try_trunc = !!au_opt_test(mnt_flags, TRUNC_XINO); ++ hi = iinfo->ii_hinode + bindex; ++ bend = iinfo->ii_bend; ++ for (; bindex <= bend; bindex++, hi++) { ++ h_inode = hi->hi_inode; ++ if (!h_inode ++ || (!unlinked && h_inode->i_nlink)) ++ continue; ++ ++ /* inode may not be revalidated */ ++ bi = au_br_index(sb, hi->hi_id); ++ if (bi < 0) ++ continue; ++ ++ br = au_sbr(sb, bi); ++ err = au_xino_do_write(xwrite, br->br_xino.xi_file, ++ h_inode->i_ino, /*ino*/0); ++ if (!err && try_trunc ++ && au_test_fs_trunc_xino(au_br_sb(br))) ++ xino_try_trunc(sb, br); ++ } ++} ++ ++/* get an unused inode number from bitmap */ ++ino_t au_xino_new_ino(struct super_block *sb) ++{ ++ ino_t ino; ++ unsigned long *p, pindex, ul, pend; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ int free_bit, err; ++ ++ if (!au_opt_test(au_mntflags(sb), XINO)) ++ return iunique(sb, AUFS_FIRST_INO); ++ ++ sbinfo = au_sbi(sb); ++ mutex_lock(&sbinfo->si_xib_mtx); ++ p = sbinfo->si_xib_buf; ++ free_bit = sbinfo->si_xib_next_bit; ++ if (free_bit < page_bits && !test_bit(free_bit, p)) ++ goto out; /* success */ ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ ++ pindex = sbinfo->si_xib_last_pindex; ++ for (ul = pindex - 1; ul < ULONG_MAX; ul--) { ++ err = xib_pindex(sb, ul); ++ if (unlikely(err)) ++ goto out_err; ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ } ++ ++ file = sbinfo->si_xib; ++ pend = vfsub_f_size_read(file) / PAGE_SIZE; ++ for (ul = pindex + 1; ul <= pend; ul++) { ++ err = xib_pindex(sb, ul); ++ if (unlikely(err)) ++ goto out_err; ++ free_bit = find_first_zero_bit(p, page_bits); ++ if (free_bit < page_bits) ++ goto out; /* success */ ++ } ++ BUG(); ++ ++out: ++ set_bit(free_bit, p); ++ sbinfo->si_xib_next_bit = free_bit + 1; ++ pindex = sbinfo->si_xib_last_pindex; ++ mutex_unlock(&sbinfo->si_xib_mtx); ++ ino = xib_calc_ino(pindex, free_bit); ++ AuDbg("i%lu\n", (unsigned long)ino); ++ return ino; ++out_err: ++ mutex_unlock(&sbinfo->si_xib_mtx); ++ AuDbg("i0\n"); ++ return 0; ++} ++ ++/* ++ * read @ino from xinofile for the specified branch{@sb, @bindex} ++ * at the position of @h_ino. ++ * if @ino does not exist and @do_new is true, get new one. ++ */ ++int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino, ++ ino_t *ino) ++{ ++ int err; ++ ssize_t sz; ++ loff_t pos; ++ struct file *file; ++ struct au_sbinfo *sbinfo; ++ ++ *ino = 0; ++ if (!au_opt_test(au_mntflags(sb), XINO)) ++ return 0; /* no xino */ ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ pos = h_ino; ++ if (unlikely(au_loff_max / sizeof(*ino) - 1 < pos)) { ++ AuIOErr1("too large hi%lu\n", (unsigned long)h_ino); ++ return -EFBIG; ++ } ++ pos *= sizeof(*ino); ++ ++ file = au_sbr(sb, bindex)->br_xino.xi_file; ++ if (vfsub_f_size_read(file) < pos + sizeof(*ino)) ++ return 0; /* no ino */ ++ ++ sz = xino_fread(sbinfo->si_xread, file, ino, sizeof(*ino), &pos); ++ if (sz == sizeof(*ino)) ++ return 0; /* success */ ++ ++ err = sz; ++ if (unlikely(sz >= 0)) { ++ err = -EIO; ++ AuIOErr("xino read error (%zd)\n", sz); ++ } ++ ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* create and set a new xino file */ ++ ++struct file *au_xino_create(struct super_block *sb, char *fname, int silent) ++{ ++ struct file *file; ++ struct dentry *h_parent, *d; ++ struct inode *h_dir, *inode; ++ int err; ++ ++ /* ++ * at mount-time, and the xino file is the default path, ++ * hnotify is disabled so we have no notify events to ignore. ++ * when a user specified the xino, we cannot get au_hdir to be ignored. ++ */ ++ file = vfsub_filp_open(fname, O_RDWR | O_CREAT | O_EXCL | O_LARGEFILE ++ /* | __FMODE_NONOTIFY */, ++ S_IRUGO | S_IWUGO); ++ if (IS_ERR(file)) { ++ if (!silent) ++ pr_err("open %s(%ld)\n", fname, PTR_ERR(file)); ++ return file; ++ } ++ ++ /* keep file count */ ++ err = 0; ++ inode = file_inode(file); ++ h_parent = dget_parent(file->f_path.dentry); ++ h_dir = d_inode(h_parent); ++ mutex_lock_nested(&h_dir->i_mutex, AuLsc_I_PARENT); ++ /* mnt_want_write() is unnecessary here */ ++ /* no delegation since it is just created */ ++ if (inode->i_nlink) ++ err = vfsub_unlink(h_dir, &file->f_path, /*delegated*/NULL, ++ /*force*/0); ++ mutex_unlock(&h_dir->i_mutex); ++ dput(h_parent); ++ if (unlikely(err)) { ++ if (!silent) ++ pr_err("unlink %s(%d)\n", fname, err); ++ goto out; ++ } ++ ++ err = -EINVAL; ++ d = file->f_path.dentry; ++ if (unlikely(sb == d->d_sb)) { ++ if (!silent) ++ pr_err("%s must be outside\n", fname); ++ goto out; ++ } ++ if (unlikely(au_test_fs_bad_xino(d->d_sb))) { ++ if (!silent) ++ pr_err("xino doesn't support %s(%s)\n", ++ fname, au_sbtype(d->d_sb)); ++ goto out; ++ } ++ return file; /* success */ ++ ++out: ++ fput(file); ++ file = ERR_PTR(err); ++ return file; ++} ++ ++/* ++ * find another branch who is on the same filesystem of the specified ++ * branch{@btgt}. search until @bend. ++ */ ++static int is_sb_shared(struct super_block *sb, aufs_bindex_t btgt, ++ aufs_bindex_t bend) ++{ ++ aufs_bindex_t bindex; ++ struct super_block *tgt_sb = au_sbr_sb(sb, btgt); ++ ++ for (bindex = 0; bindex < btgt; bindex++) ++ if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) ++ return bindex; ++ for (bindex++; bindex <= bend; bindex++) ++ if (unlikely(tgt_sb == au_sbr_sb(sb, bindex))) ++ return bindex; ++ return -1; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * initialize the xinofile for the specified branch @br ++ * at the place/path where @base_file indicates. ++ * test whether another branch is on the same filesystem or not, ++ * if @do_test is true. ++ */ ++int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t h_ino, ++ struct file *base_file, int do_test) ++{ ++ int err; ++ ino_t ino; ++ aufs_bindex_t bend, bindex; ++ struct au_branch *shared_br, *b; ++ struct file *file; ++ struct super_block *tgt_sb; ++ ++ shared_br = NULL; ++ bend = au_sbend(sb); ++ if (do_test) { ++ tgt_sb = au_br_sb(br); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ b = au_sbr(sb, bindex); ++ if (tgt_sb == au_br_sb(b)) { ++ shared_br = b; ++ break; ++ } ++ } ++ } ++ ++ if (!shared_br || !shared_br->br_xino.xi_file) { ++ struct au_xino_lock_dir ldir; ++ ++ au_xino_lock_dir(sb, base_file, &ldir); ++ /* mnt_want_write() is unnecessary here */ ++ file = au_xino_create2(base_file, NULL); ++ au_xino_unlock_dir(&ldir); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ br->br_xino.xi_file = file; ++ } else { ++ br->br_xino.xi_file = shared_br->br_xino.xi_file; ++ get_file(br->br_xino.xi_file); ++ } ++ ++ ino = AUFS_ROOT_INO; ++ err = au_xino_do_write(au_sbi(sb)->si_xwrite, br->br_xino.xi_file, ++ h_ino, ino); ++ if (unlikely(err)) { ++ fput(br->br_xino.xi_file); ++ br->br_xino.xi_file = NULL; ++ } ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* trucate a xino bitmap file */ ++ ++/* todo: slow */ ++static int do_xib_restore(struct super_block *sb, struct file *file, void *page) ++{ ++ int err, bit; ++ ssize_t sz; ++ unsigned long pindex; ++ loff_t pos, pend; ++ struct au_sbinfo *sbinfo; ++ vfs_readf_t func; ++ ino_t *ino; ++ unsigned long *p; ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ MtxMustLock(&sbinfo->si_xib_mtx); ++ p = sbinfo->si_xib_buf; ++ func = sbinfo->si_xread; ++ pend = vfsub_f_size_read(file); ++ pos = 0; ++ while (pos < pend) { ++ sz = xino_fread(func, file, page, PAGE_SIZE, &pos); ++ err = sz; ++ if (unlikely(sz <= 0)) ++ goto out; ++ ++ err = 0; ++ for (ino = page; sz > 0; ino++, sz -= sizeof(ino)) { ++ if (unlikely(*ino < AUFS_FIRST_INO)) ++ continue; ++ ++ xib_calc_bit(*ino, &pindex, &bit); ++ AuDebugOn(page_bits <= bit); ++ err = xib_pindex(sb, pindex); ++ if (!err) ++ set_bit(bit, p); ++ else ++ goto out; ++ } ++ } ++ ++out: ++ return err; ++} ++ ++static int xib_restore(struct super_block *sb) ++{ ++ int err; ++ aufs_bindex_t bindex, bend; ++ void *page; ++ ++ err = -ENOMEM; ++ page = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!page)) ++ goto out; ++ ++ err = 0; ++ bend = au_sbend(sb); ++ for (bindex = 0; !err && bindex <= bend; bindex++) ++ if (!bindex || is_sb_shared(sb, bindex, bindex - 1) < 0) ++ err = do_xib_restore ++ (sb, au_sbr(sb, bindex)->br_xino.xi_file, page); ++ else ++ AuDbg("b%d\n", bindex); ++ free_page((unsigned long)page); ++ ++out: ++ return err; ++} ++ ++int au_xib_trunc(struct super_block *sb) ++{ ++ int err; ++ ssize_t sz; ++ loff_t pos; ++ struct au_xino_lock_dir ldir; ++ struct au_sbinfo *sbinfo; ++ unsigned long *p; ++ struct file *file; ++ ++ SiMustWriteLock(sb); ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ if (!au_opt_test(sbinfo->si_mntflags, XINO)) ++ goto out; ++ ++ file = sbinfo->si_xib; ++ if (vfsub_f_size_read(file) <= PAGE_SIZE) ++ goto out; ++ ++ au_xino_lock_dir(sb, file, &ldir); ++ /* mnt_want_write() is unnecessary here */ ++ file = au_xino_create2(sbinfo->si_xib, NULL); ++ au_xino_unlock_dir(&ldir); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ fput(sbinfo->si_xib); ++ sbinfo->si_xib = file; ++ ++ p = sbinfo->si_xib_buf; ++ memset(p, 0, PAGE_SIZE); ++ pos = 0; ++ sz = xino_fwrite(sbinfo->si_xwrite, sbinfo->si_xib, p, PAGE_SIZE, &pos); ++ if (unlikely(sz != PAGE_SIZE)) { ++ err = sz; ++ AuIOErr("err %d\n", err); ++ if (sz >= 0) ++ err = -EIO; ++ goto out; ++ } ++ ++ mutex_lock(&sbinfo->si_xib_mtx); ++ /* mnt_want_write() is unnecessary here */ ++ err = xib_restore(sb); ++ mutex_unlock(&sbinfo->si_xib_mtx); ++ ++out: ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * xino mount option handlers ++ */ ++ ++/* xino bitmap */ ++static void xino_clear_xib(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ sbinfo->si_xread = NULL; ++ sbinfo->si_xwrite = NULL; ++ if (sbinfo->si_xib) ++ fput(sbinfo->si_xib); ++ sbinfo->si_xib = NULL; ++ free_page((unsigned long)sbinfo->si_xib_buf); ++ sbinfo->si_xib_buf = NULL; ++} ++ ++static int au_xino_set_xib(struct super_block *sb, struct file *base) ++{ ++ int err; ++ loff_t pos; ++ struct au_sbinfo *sbinfo; ++ struct file *file; ++ ++ SiMustWriteLock(sb); ++ ++ sbinfo = au_sbi(sb); ++ file = au_xino_create2(base, sbinfo->si_xib); ++ err = PTR_ERR(file); ++ if (IS_ERR(file)) ++ goto out; ++ if (sbinfo->si_xib) ++ fput(sbinfo->si_xib); ++ sbinfo->si_xib = file; ++ sbinfo->si_xread = vfs_readf(file); ++ sbinfo->si_xwrite = vfs_writef(file); ++ ++ err = -ENOMEM; ++ if (!sbinfo->si_xib_buf) ++ sbinfo->si_xib_buf = (void *)get_zeroed_page(GFP_NOFS); ++ if (unlikely(!sbinfo->si_xib_buf)) ++ goto out_unset; ++ ++ sbinfo->si_xib_last_pindex = 0; ++ sbinfo->si_xib_next_bit = 0; ++ if (vfsub_f_size_read(file) < PAGE_SIZE) { ++ pos = 0; ++ err = xino_fwrite(sbinfo->si_xwrite, file, sbinfo->si_xib_buf, ++ PAGE_SIZE, &pos); ++ if (unlikely(err != PAGE_SIZE)) ++ goto out_free; ++ } ++ err = 0; ++ goto out; /* success */ ++ ++out_free: ++ free_page((unsigned long)sbinfo->si_xib_buf); ++ sbinfo->si_xib_buf = NULL; ++ if (err >= 0) ++ err = -EIO; ++out_unset: ++ fput(sbinfo->si_xib); ++ sbinfo->si_xib = NULL; ++ sbinfo->si_xread = NULL; ++ sbinfo->si_xwrite = NULL; ++out: ++ return err; ++} ++ ++/* xino for each branch */ ++static void xino_clear_br(struct super_block *sb) ++{ ++ aufs_bindex_t bindex, bend; ++ struct au_branch *br; ++ ++ bend = au_sbend(sb); ++ for (bindex = 0; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (!br || !br->br_xino.xi_file) ++ continue; ++ ++ fput(br->br_xino.xi_file); ++ br->br_xino.xi_file = NULL; ++ } ++} ++ ++static int au_xino_set_br(struct super_block *sb, struct file *base) ++{ ++ int err; ++ ino_t ino; ++ aufs_bindex_t bindex, bend, bshared; ++ struct { ++ struct file *old, *new; ++ } *fpair, *p; ++ struct au_branch *br; ++ struct inode *inode; ++ vfs_writef_t writef; ++ ++ SiMustWriteLock(sb); ++ ++ err = -ENOMEM; ++ bend = au_sbend(sb); ++ fpair = kcalloc(bend + 1, sizeof(*fpair), GFP_NOFS); ++ if (unlikely(!fpair)) ++ goto out; ++ ++ inode = d_inode(sb->s_root); ++ ino = AUFS_ROOT_INO; ++ writef = au_sbi(sb)->si_xwrite; ++ for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) { ++ br = au_sbr(sb, bindex); ++ bshared = is_sb_shared(sb, bindex, bindex - 1); ++ if (bshared >= 0) { ++ /* shared xino */ ++ *p = fpair[bshared]; ++ get_file(p->new); ++ } ++ ++ if (!p->new) { ++ /* new xino */ ++ p->old = br->br_xino.xi_file; ++ p->new = au_xino_create2(base, br->br_xino.xi_file); ++ err = PTR_ERR(p->new); ++ if (IS_ERR(p->new)) { ++ p->new = NULL; ++ goto out_pair; ++ } ++ } ++ ++ err = au_xino_do_write(writef, p->new, ++ au_h_iptr(inode, bindex)->i_ino, ino); ++ if (unlikely(err)) ++ goto out_pair; ++ } ++ ++ for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) { ++ br = au_sbr(sb, bindex); ++ if (br->br_xino.xi_file) ++ fput(br->br_xino.xi_file); ++ get_file(p->new); ++ br->br_xino.xi_file = p->new; ++ } ++ ++out_pair: ++ for (bindex = 0, p = fpair; bindex <= bend; bindex++, p++) ++ if (p->new) ++ fput(p->new); ++ else ++ break; ++ kfree(fpair); ++out: ++ return err; ++} ++ ++void au_xino_clr(struct super_block *sb) ++{ ++ struct au_sbinfo *sbinfo; ++ ++ au_xigen_clr(sb); ++ xino_clear_xib(sb); ++ xino_clear_br(sb); ++ sbinfo = au_sbi(sb); ++ /* lvalue, do not call au_mntflags() */ ++ au_opt_clr(sbinfo->si_mntflags, XINO); ++} ++ ++int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount) ++{ ++ int err, skip; ++ struct dentry *parent, *cur_parent; ++ struct qstr *dname, *cur_name; ++ struct file *cur_xino; ++ struct inode *dir; ++ struct au_sbinfo *sbinfo; ++ ++ SiMustWriteLock(sb); ++ ++ err = 0; ++ sbinfo = au_sbi(sb); ++ parent = dget_parent(xino->file->f_path.dentry); ++ if (remount) { ++ skip = 0; ++ dname = &xino->file->f_path.dentry->d_name; ++ cur_xino = sbinfo->si_xib; ++ if (cur_xino) { ++ cur_parent = dget_parent(cur_xino->f_path.dentry); ++ cur_name = &cur_xino->f_path.dentry->d_name; ++ skip = (cur_parent == parent ++ && au_qstreq(dname, cur_name)); ++ dput(cur_parent); ++ } ++ if (skip) ++ goto out; ++ } ++ ++ au_opt_set(sbinfo->si_mntflags, XINO); ++ dir = d_inode(parent); ++ mutex_lock_nested(&dir->i_mutex, AuLsc_I_PARENT); ++ /* mnt_want_write() is unnecessary here */ ++ err = au_xino_set_xib(sb, xino->file); ++ if (!err) ++ err = au_xigen_set(sb, xino->file); ++ if (!err) ++ err = au_xino_set_br(sb, xino->file); ++ mutex_unlock(&dir->i_mutex); ++ if (!err) ++ goto out; /* success */ ++ ++ /* reset all */ ++ AuIOErr("failed creating xino(%d).\n", err); ++ au_xigen_clr(sb); ++ xino_clear_xib(sb); ++ ++out: ++ dput(parent); ++ return err; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ++ * create a xinofile at the default place/path. ++ */ ++struct file *au_xino_def(struct super_block *sb) ++{ ++ struct file *file; ++ char *page, *p; ++ struct au_branch *br; ++ struct super_block *h_sb; ++ struct path path; ++ aufs_bindex_t bend, bindex, bwr; ++ ++ br = NULL; ++ bend = au_sbend(sb); ++ bwr = -1; ++ for (bindex = 0; bindex <= bend; bindex++) { ++ br = au_sbr(sb, bindex); ++ if (au_br_writable(br->br_perm) ++ && !au_test_fs_bad_xino(au_br_sb(br))) { ++ bwr = bindex; ++ break; ++ } ++ } ++ ++ if (bwr >= 0) { ++ file = ERR_PTR(-ENOMEM); ++ page = (void *)__get_free_page(GFP_NOFS); ++ if (unlikely(!page)) ++ goto out; ++ path.mnt = au_br_mnt(br); ++ path.dentry = au_h_dptr(sb->s_root, bwr); ++ p = d_path(&path, page, PATH_MAX - sizeof(AUFS_XINO_FNAME)); ++ file = (void *)p; ++ if (!IS_ERR(p)) { ++ strcat(p, "/" AUFS_XINO_FNAME); ++ AuDbg("%s\n", p); ++ file = au_xino_create(sb, p, /*silent*/0); ++ if (!IS_ERR(file)) ++ au_xino_brid_set(sb, br->br_id); ++ } ++ free_page((unsigned long)page); ++ } else { ++ file = au_xino_create(sb, AUFS_XINO_DEFPATH, /*silent*/0); ++ if (IS_ERR(file)) ++ goto out; ++ h_sb = file->f_path.dentry->d_sb; ++ if (unlikely(au_test_fs_bad_xino(h_sb))) { ++ pr_err("xino doesn't support %s(%s)\n", ++ AUFS_XINO_DEFPATH, au_sbtype(h_sb)); ++ fput(file); ++ file = ERR_PTR(-EINVAL); ++ } ++ if (!IS_ERR(file)) ++ au_xino_brid_set(sb, -1); ++ } ++ ++out: ++ return file; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++int au_xino_path(struct seq_file *seq, struct file *file) ++{ ++ int err; ++ ++ err = au_seq_path(seq, &file->f_path); ++ if (unlikely(err)) ++ goto out; ++ ++#define Deleted "\\040(deleted)" ++ seq->count -= sizeof(Deleted) - 1; ++ AuDebugOn(memcmp(seq->buf + seq->count, Deleted, ++ sizeof(Deleted) - 1)); ++#undef Deleted ++ ++out: ++ return err; ++} +diff -Nur linux-4.1.10.orig/fs/buffer.c linux-4.1.10/fs/buffer.c +--- linux-4.1.10.orig/fs/buffer.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/buffer.c 2015-10-22 21:35:53.000000000 +0200 +@@ -2450,7 +2450,7 @@ + * Update file times before taking page lock. We may end up failing the + * fault so this update may be superfluous but who really cares... + */ +- file_update_time(vma->vm_file); ++ vma_file_update_time(vma); + + ret = __block_page_mkwrite(vma, vmf, get_block); + sb_end_pagefault(sb); +diff -Nur linux-4.1.10.orig/fs/dcache.c linux-4.1.10/fs/dcache.c +--- linux-4.1.10.orig/fs/dcache.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/dcache.c 2015-10-22 21:35:53.000000000 +0200 +@@ -1167,7 +1167,7 @@ + * + * The @enter() and @finish() callbacks are called with d_lock held. + */ +-static void d_walk(struct dentry *parent, void *data, ++void d_walk(struct dentry *parent, void *data, + enum d_walk_ret (*enter)(void *, struct dentry *), + void (*finish)(void *)) + { +@@ -1272,6 +1272,7 @@ + seq = 1; + goto again; + } ++EXPORT_SYMBOL(d_walk); + + /* + * Search for at least 1 mount point in the dentry's subdirs. +diff -Nur linux-4.1.10.orig/fs/file_table.c linux-4.1.10/fs/file_table.c +--- linux-4.1.10.orig/fs/file_table.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/file_table.c 2015-10-22 21:35:53.000000000 +0200 +@@ -147,6 +147,7 @@ + } + return ERR_PTR(-ENFILE); + } ++EXPORT_SYMBOL(get_empty_filp); + + /** + * alloc_file - allocate and initialize a 'struct file' +@@ -308,6 +309,7 @@ + file_free(file); + } + } ++EXPORT_SYMBOL(put_filp); + + void __init files_init(unsigned long mempages) + { +diff -Nur linux-4.1.10.orig/fs/inode.c linux-4.1.10/fs/inode.c +--- linux-4.1.10.orig/fs/inode.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/inode.c 2015-10-22 21:35:53.000000000 +0200 +@@ -58,6 +58,7 @@ + static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock); + + __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock); ++EXPORT_SYMBOL(inode_sb_list_lock); + + /* + * Empty aops. Can be used for the cases where the user does not +diff -Nur linux-4.1.10.orig/fs/Kconfig linux-4.1.10/fs/Kconfig +--- linux-4.1.10.orig/fs/Kconfig 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/Kconfig 2015-10-22 21:35:53.000000000 +0200 +@@ -218,6 +218,7 @@ + source "fs/sysv/Kconfig" + source "fs/ufs/Kconfig" + source "fs/exofs/Kconfig" ++source "fs/aufs/Kconfig" + + endif # MISC_FILESYSTEMS + +diff -Nur linux-4.1.10.orig/fs/Makefile linux-4.1.10/fs/Makefile +--- linux-4.1.10.orig/fs/Makefile 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/Makefile 2015-10-22 21:35:53.000000000 +0200 +@@ -127,3 +127,4 @@ + obj-$(CONFIG_CEPH_FS) += ceph/ + obj-$(CONFIG_PSTORE) += pstore/ + obj-$(CONFIG_EFIVAR_FS) += efivarfs/ ++obj-$(CONFIG_AUFS_FS) += aufs/ +diff -Nur linux-4.1.10.orig/fs/namespace.c linux-4.1.10/fs/namespace.c +--- linux-4.1.10.orig/fs/namespace.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/namespace.c 2015-10-22 21:35:53.000000000 +0200 +@@ -463,6 +463,7 @@ + mnt_dec_writers(real_mount(mnt)); + preempt_enable(); + } ++EXPORT_SYMBOL_GPL(__mnt_drop_write); + + /** + * mnt_drop_write - give up write access to a mount +@@ -1792,6 +1793,7 @@ + } + return 0; + } ++EXPORT_SYMBOL(iterate_mounts); + + static void cleanup_group_ids(struct mount *mnt, struct mount *end) + { +diff -Nur linux-4.1.10.orig/fs/notify/group.c linux-4.1.10/fs/notify/group.c +--- linux-4.1.10.orig/fs/notify/group.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/notify/group.c 2015-10-22 21:35:53.000000000 +0200 +@@ -22,6 +22,7 @@ + #include <linux/srcu.h> + #include <linux/rculist.h> + #include <linux/wait.h> ++#include <linux/module.h> + + #include <linux/fsnotify_backend.h> + #include "fsnotify.h" +@@ -72,6 +73,7 @@ + { + atomic_inc(&group->refcnt); + } ++EXPORT_SYMBOL(fsnotify_get_group); + + /* + * Drop a reference to a group. Free it if it's through. +@@ -81,6 +83,7 @@ + if (atomic_dec_and_test(&group->refcnt)) + fsnotify_final_destroy_group(group); + } ++EXPORT_SYMBOL(fsnotify_put_group); + + /* + * Create a new fsnotify_group and hold a reference for the group returned. +@@ -109,6 +112,7 @@ + + return group; + } ++EXPORT_SYMBOL(fsnotify_alloc_group); + + int fsnotify_fasync(int fd, struct file *file, int on) + { +diff -Nur linux-4.1.10.orig/fs/notify/mark.c linux-4.1.10/fs/notify/mark.c +--- linux-4.1.10.orig/fs/notify/mark.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/notify/mark.c 2015-10-22 21:35:53.000000000 +0200 +@@ -109,6 +109,7 @@ + mark->free_mark(mark); + } + } ++EXPORT_SYMBOL(fsnotify_put_mark); + + /* Calculate mask of events for a list of marks */ + u32 fsnotify_recalc_mask(struct hlist_head *head) +@@ -202,6 +203,7 @@ + fsnotify_destroy_mark_locked(mark, group); + mutex_unlock(&group->mark_mutex); + } ++EXPORT_SYMBOL(fsnotify_destroy_mark); + + /* + * Destroy all marks in the given list. The marks must be already detached from +@@ -376,6 +378,7 @@ + + return ret; + } ++EXPORT_SYMBOL(fsnotify_add_mark); + + int fsnotify_add_mark(struct fsnotify_mark *mark, struct fsnotify_group *group, + struct inode *inode, struct vfsmount *mnt, int allow_dups) +@@ -475,6 +478,7 @@ + atomic_set(&mark->refcnt, 1); + mark->free_mark = free_mark; + } ++EXPORT_SYMBOL(fsnotify_init_mark); + + static int fsnotify_mark_destroy(void *ignored) + { +diff -Nur linux-4.1.10.orig/fs/open.c linux-4.1.10/fs/open.c +--- linux-4.1.10.orig/fs/open.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/open.c 2015-10-22 21:35:53.000000000 +0200 +@@ -62,6 +62,7 @@ + mutex_unlock(&dentry->d_inode->i_mutex); + return ret; + } ++EXPORT_SYMBOL(do_truncate); + + long vfs_truncate(struct path *path, loff_t length) + { +@@ -676,6 +677,7 @@ + } + return 0; + } ++EXPORT_SYMBOL(open_check_o_direct); + + static int do_dentry_open(struct file *f, + int (*open)(struct inode *, struct file *), +diff -Nur linux-4.1.10.orig/fs/proc/base.c linux-4.1.10/fs/proc/base.c +--- linux-4.1.10.orig/fs/proc/base.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/proc/base.c 2015-10-22 21:35:53.000000000 +0200 +@@ -1744,7 +1744,7 @@ + down_read(&mm->mmap_sem); + vma = find_exact_vma(mm, vm_start, vm_end); + if (vma && vma->vm_file) { +- *path = vma->vm_file->f_path; ++ *path = vma_pr_or_file(vma)->f_path; + path_get(path); + rc = 0; + } +diff -Nur linux-4.1.10.orig/fs/proc/nommu.c linux-4.1.10/fs/proc/nommu.c +--- linux-4.1.10.orig/fs/proc/nommu.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/proc/nommu.c 2015-10-22 21:35:53.000000000 +0200 +@@ -45,7 +45,10 @@ + file = region->vm_file; + + if (file) { +- struct inode *inode = file_inode(region->vm_file); ++ struct inode *inode; ++ ++ file = vmr_pr_or_file(region); ++ inode = file_inode(file); + dev = inode->i_sb->s_dev; + ino = inode->i_ino; + } +diff -Nur linux-4.1.10.orig/fs/proc/task_mmu.c linux-4.1.10/fs/proc/task_mmu.c +--- linux-4.1.10.orig/fs/proc/task_mmu.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/proc/task_mmu.c 2015-10-22 21:35:53.000000000 +0200 +@@ -279,7 +279,10 @@ + const char *name = NULL; + + if (file) { +- struct inode *inode = file_inode(vma->vm_file); ++ struct inode *inode; ++ ++ file = vma_pr_or_file(vma); ++ inode = file_inode(file); + dev = inode->i_sb->s_dev; + ino = inode->i_ino; + pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT; +@@ -1479,7 +1482,7 @@ + struct proc_maps_private *proc_priv = &numa_priv->proc_maps; + struct vm_area_struct *vma = v; + struct numa_maps *md = &numa_priv->md; +- struct file *file = vma->vm_file; ++ struct file *file = vma_pr_or_file(vma); + struct mm_struct *mm = vma->vm_mm; + struct mm_walk walk = { + .hugetlb_entry = gather_hugetlb_stats, +diff -Nur linux-4.1.10.orig/fs/proc/task_nommu.c linux-4.1.10/fs/proc/task_nommu.c +--- linux-4.1.10.orig/fs/proc/task_nommu.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/proc/task_nommu.c 2015-10-22 21:35:53.000000000 +0200 +@@ -160,7 +160,10 @@ + file = vma->vm_file; + + if (file) { +- struct inode *inode = file_inode(vma->vm_file); ++ struct inode *inode; ++ ++ file = vma_pr_or_file(vma); ++ inode = file_inode(file); + dev = inode->i_sb->s_dev; + ino = inode->i_ino; + pgoff = (loff_t)vma->vm_pgoff << PAGE_SHIFT; +diff -Nur linux-4.1.10.orig/fs/read_write.c linux-4.1.10/fs/read_write.c +--- linux-4.1.10.orig/fs/read_write.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/read_write.c 2015-10-22 21:35:53.000000000 +0200 +@@ -494,6 +494,30 @@ + } + EXPORT_SYMBOL(__vfs_write); + ++vfs_readf_t vfs_readf(struct file *file) ++{ ++ const struct file_operations *fop = file->f_op; ++ ++ if (fop->read) ++ return fop->read; ++ if (fop->read_iter) ++ return new_sync_read; ++ return ERR_PTR(-ENOSYS); ++} ++EXPORT_SYMBOL(vfs_readf); ++ ++vfs_writef_t vfs_writef(struct file *file) ++{ ++ const struct file_operations *fop = file->f_op; ++ ++ if (fop->write) ++ return fop->write; ++ if (fop->write_iter) ++ return new_sync_write; ++ return ERR_PTR(-ENOSYS); ++} ++EXPORT_SYMBOL(vfs_writef); ++ + ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t *pos) + { + mm_segment_t old_fs; +diff -Nur linux-4.1.10.orig/fs/splice.c linux-4.1.10/fs/splice.c +--- linux-4.1.10.orig/fs/splice.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/splice.c 2015-10-22 21:35:53.000000000 +0200 +@@ -1101,8 +1101,8 @@ + /* + * Attempt to initiate a splice from pipe to file. + */ +-static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, +- loff_t *ppos, size_t len, unsigned int flags) ++long do_splice_from(struct pipe_inode_info *pipe, struct file *out, ++ loff_t *ppos, size_t len, unsigned int flags) + { + ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, + loff_t *, size_t, unsigned int); +@@ -1114,13 +1114,14 @@ + + return splice_write(pipe, out, ppos, len, flags); + } ++EXPORT_SYMBOL(do_splice_from); + + /* + * Attempt to initiate a splice from a file to a pipe. + */ +-static long do_splice_to(struct file *in, loff_t *ppos, +- struct pipe_inode_info *pipe, size_t len, +- unsigned int flags) ++long do_splice_to(struct file *in, loff_t *ppos, ++ struct pipe_inode_info *pipe, size_t len, ++ unsigned int flags) + { + ssize_t (*splice_read)(struct file *, loff_t *, + struct pipe_inode_info *, size_t, unsigned int); +@@ -1140,6 +1141,7 @@ + + return splice_read(in, ppos, pipe, len, flags); + } ++EXPORT_SYMBOL(do_splice_to); + + /** + * splice_direct_to_actor - splices data directly between two non-pipes +diff -Nur linux-4.1.10.orig/fs/xattr.c linux-4.1.10/fs/xattr.c +--- linux-4.1.10.orig/fs/xattr.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/fs/xattr.c 2015-10-22 21:35:53.000000000 +0200 +@@ -207,6 +207,7 @@ + *xattr_value = value; + return error; + } ++EXPORT_SYMBOL(vfs_getxattr_alloc); + + /* Compare an extended attribute value with the given value */ + int vfs_xattr_cmp(struct dentry *dentry, const char *xattr_name, +diff -Nur linux-4.1.10.orig/include/linux/file.h linux-4.1.10/include/linux/file.h +--- linux-4.1.10.orig/include/linux/file.h 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/linux/file.h 2015-10-22 21:35:53.000000000 +0200 +@@ -19,6 +19,7 @@ + struct path; + extern struct file *alloc_file(struct path *, fmode_t mode, + const struct file_operations *fop); ++extern struct file *get_empty_filp(void); + + static inline void fput_light(struct file *file, int fput_needed) + { +diff -Nur linux-4.1.10.orig/include/linux/fs.h linux-4.1.10/include/linux/fs.h +--- linux-4.1.10.orig/include/linux/fs.h 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/linux/fs.h 2015-10-22 21:35:53.000000000 +0200 +@@ -1649,6 +1649,12 @@ + struct iovec *fast_pointer, + struct iovec **ret_pointer); + ++typedef ssize_t (*vfs_readf_t)(struct file *, char __user *, size_t, loff_t *); ++typedef ssize_t (*vfs_writef_t)(struct file *, const char __user *, size_t, ++ loff_t *); ++vfs_readf_t vfs_readf(struct file *file); ++vfs_writef_t vfs_writef(struct file *file); ++ + extern ssize_t __vfs_read(struct file *, char __user *, size_t, loff_t *); + extern ssize_t __vfs_write(struct file *, const char __user *, size_t, loff_t *); + extern ssize_t vfs_read(struct file *, char __user *, size_t, loff_t *); +diff -Nur linux-4.1.10.orig/include/linux/mm.h linux-4.1.10/include/linux/mm.h +--- linux-4.1.10.orig/include/linux/mm.h 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/linux/mm.h 2015-10-22 21:35:53.000000000 +0200 +@@ -1200,6 +1200,28 @@ + } + #endif + ++extern void vma_do_file_update_time(struct vm_area_struct *, const char[], int); ++extern struct file *vma_do_pr_or_file(struct vm_area_struct *, const char[], ++ int); ++extern void vma_do_get_file(struct vm_area_struct *, const char[], int); ++extern void vma_do_fput(struct vm_area_struct *, const char[], int); ++ ++#define vma_file_update_time(vma) vma_do_file_update_time(vma, __func__, \ ++ __LINE__) ++#define vma_pr_or_file(vma) vma_do_pr_or_file(vma, __func__, \ ++ __LINE__) ++#define vma_get_file(vma) vma_do_get_file(vma, __func__, __LINE__) ++#define vma_fput(vma) vma_do_fput(vma, __func__, __LINE__) ++ ++#ifndef CONFIG_MMU ++extern struct file *vmr_do_pr_or_file(struct vm_region *, const char[], int); ++extern void vmr_do_fput(struct vm_region *, const char[], int); ++ ++#define vmr_pr_or_file(region) vmr_do_pr_or_file(region, __func__, \ ++ __LINE__) ++#define vmr_fput(region) vmr_do_fput(region, __func__, __LINE__) ++#endif /* !CONFIG_MMU */ ++ + extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write); + extern int access_remote_vm(struct mm_struct *mm, unsigned long addr, + void *buf, int len, int write); +diff -Nur linux-4.1.10.orig/include/linux/mm_types.h linux-4.1.10/include/linux/mm_types.h +--- linux-4.1.10.orig/include/linux/mm_types.h 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/linux/mm_types.h 2015-10-22 21:35:53.000000000 +0200 +@@ -232,6 +232,7 @@ + unsigned long vm_top; /* region allocated to here */ + unsigned long vm_pgoff; /* the offset in vm_file corresponding to vm_start */ + struct file *vm_file; /* the backing file or NULL */ ++ struct file *vm_prfile; /* the virtual backing file or NULL */ + + int vm_usage; /* region usage count (access under nommu_region_sem) */ + bool vm_icache_flushed : 1; /* true if the icache has been flushed for +@@ -296,6 +297,7 @@ + unsigned long vm_pgoff; /* Offset (within vm_file) in PAGE_SIZE + units, *not* PAGE_CACHE_SIZE */ + struct file * vm_file; /* File we map to (can be NULL). */ ++ struct file *vm_prfile; /* shadow of vm_file */ + void * vm_private_data; /* was vm_pte (shared mem) */ + + #ifndef CONFIG_MMU +diff -Nur linux-4.1.10.orig/include/linux/splice.h linux-4.1.10/include/linux/splice.h +--- linux-4.1.10.orig/include/linux/splice.h 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/linux/splice.h 2015-10-22 21:35:53.000000000 +0200 +@@ -83,4 +83,10 @@ + extern void spd_release_page(struct splice_pipe_desc *, unsigned int); + + extern const struct pipe_buf_operations page_cache_pipe_buf_ops; ++ ++extern long do_splice_from(struct pipe_inode_info *pipe, struct file *out, ++ loff_t *ppos, size_t len, unsigned int flags); ++extern long do_splice_to(struct file *in, loff_t *ppos, ++ struct pipe_inode_info *pipe, size_t len, ++ unsigned int flags); + #endif +diff -Nur linux-4.1.10.orig/include/uapi/linux/aufs_type.h linux-4.1.10/include/uapi/linux/aufs_type.h +--- linux-4.1.10.orig/include/uapi/linux/aufs_type.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/include/uapi/linux/aufs_type.h 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,419 @@ ++/* ++ * Copyright (C) 2005-2015 Junjiro R. Okajima ++ * ++ * This program, aufs is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License as published by ++ * the Free Software Foundation; either version 2 of the License, or ++ * (at your option) any later version. ++ * ++ * This program is distributed in the hope that it will be useful, ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ++ * GNU General Public License for more details. ++ * ++ * You should have received a copy of the GNU General Public License ++ * along with this program. If not, see <http://www.gnu.org/licenses/>. ++ */ ++ ++#ifndef __AUFS_TYPE_H__ ++#define __AUFS_TYPE_H__ ++ ++#define AUFS_NAME "aufs" ++ ++#ifdef __KERNEL__ ++/* ++ * define it before including all other headers. ++ * sched.h may use pr_* macros before defining "current", so define the ++ * no-current version first, and re-define later. ++ */ ++#define pr_fmt(fmt) AUFS_NAME " %s:%d: " fmt, __func__, __LINE__ ++#include <linux/sched.h> ++#undef pr_fmt ++#define pr_fmt(fmt) \ ++ AUFS_NAME " %s:%d:%.*s[%d]: " fmt, __func__, __LINE__, \ ++ (int)sizeof(current->comm), current->comm, current->pid ++#else ++#include <stdint.h> ++#include <sys/types.h> ++#endif /* __KERNEL__ */ ++ ++#include <linux/limits.h> ++ ++#define AUFS_VERSION "4.1-20151012" ++ ++/* todo? move this to linux-2.6.19/include/magic.h */ ++#define AUFS_SUPER_MAGIC ('a' << 24 | 'u' << 16 | 'f' << 8 | 's') ++ ++/* ---------------------------------------------------------------------- */ ++ ++#ifdef CONFIG_AUFS_BRANCH_MAX_127 ++typedef int8_t aufs_bindex_t; ++#define AUFS_BRANCH_MAX 127 ++#else ++typedef int16_t aufs_bindex_t; ++#ifdef CONFIG_AUFS_BRANCH_MAX_511 ++#define AUFS_BRANCH_MAX 511 ++#elif defined(CONFIG_AUFS_BRANCH_MAX_1023) ++#define AUFS_BRANCH_MAX 1023 ++#elif defined(CONFIG_AUFS_BRANCH_MAX_32767) ++#define AUFS_BRANCH_MAX 32767 ++#endif ++#endif ++ ++#ifdef __KERNEL__ ++#ifndef AUFS_BRANCH_MAX ++#error unknown CONFIG_AUFS_BRANCH_MAX value ++#endif ++#endif /* __KERNEL__ */ ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define AUFS_FSTYPE AUFS_NAME ++ ++#define AUFS_ROOT_INO 2 ++#define AUFS_FIRST_INO 11 ++ ++#define AUFS_WH_PFX ".wh." ++#define AUFS_WH_PFX_LEN ((int)sizeof(AUFS_WH_PFX) - 1) ++#define AUFS_WH_TMP_LEN 4 ++/* a limit for rmdir/rename a dir and copyup */ ++#define AUFS_MAX_NAMELEN (NAME_MAX \ ++ - AUFS_WH_PFX_LEN * 2 /* doubly whiteouted */\ ++ - 1 /* dot */\ ++ - AUFS_WH_TMP_LEN) /* hex */ ++#define AUFS_XINO_FNAME "." AUFS_NAME ".xino" ++#define AUFS_XINO_DEFPATH "/tmp/" AUFS_XINO_FNAME ++#define AUFS_XINO_DEF_SEC 30 /* seconds */ ++#define AUFS_XINO_DEF_TRUNC 45 /* percentage */ ++#define AUFS_DIRWH_DEF 3 ++#define AUFS_RDCACHE_DEF 10 /* seconds */ ++#define AUFS_RDCACHE_MAX 3600 /* seconds */ ++#define AUFS_RDBLK_DEF 512 /* bytes */ ++#define AUFS_RDHASH_DEF 32 ++#define AUFS_WKQ_NAME AUFS_NAME "d" ++#define AUFS_MFS_DEF_SEC 30 /* seconds */ ++#define AUFS_MFS_MAX_SEC 3600 /* seconds */ ++#define AUFS_FHSM_CACHE_DEF_SEC 30 /* seconds */ ++#define AUFS_PLINK_WARN 50 /* number of plinks in a single bucket */ ++ ++/* pseudo-link maintenace under /proc */ ++#define AUFS_PLINK_MAINT_NAME "plink_maint" ++#define AUFS_PLINK_MAINT_DIR "fs/" AUFS_NAME ++#define AUFS_PLINK_MAINT_PATH AUFS_PLINK_MAINT_DIR "/" AUFS_PLINK_MAINT_NAME ++ ++#define AUFS_DIROPQ_NAME AUFS_WH_PFX ".opq" /* whiteouted doubly */ ++#define AUFS_WH_DIROPQ AUFS_WH_PFX AUFS_DIROPQ_NAME ++ ++#define AUFS_BASE_NAME AUFS_WH_PFX AUFS_NAME ++#define AUFS_PLINKDIR_NAME AUFS_WH_PFX "plnk" ++#define AUFS_ORPHDIR_NAME AUFS_WH_PFX "orph" ++ ++/* doubly whiteouted */ ++#define AUFS_WH_BASE AUFS_WH_PFX AUFS_BASE_NAME ++#define AUFS_WH_PLINKDIR AUFS_WH_PFX AUFS_PLINKDIR_NAME ++#define AUFS_WH_ORPHDIR AUFS_WH_PFX AUFS_ORPHDIR_NAME ++ ++/* branch permissions and attributes */ ++#define AUFS_BRPERM_RW "rw" ++#define AUFS_BRPERM_RO "ro" ++#define AUFS_BRPERM_RR "rr" ++#define AUFS_BRATTR_COO_REG "coo_reg" ++#define AUFS_BRATTR_COO_ALL "coo_all" ++#define AUFS_BRATTR_FHSM "fhsm" ++#define AUFS_BRATTR_UNPIN "unpin" ++#define AUFS_BRATTR_ICEX "icex" ++#define AUFS_BRATTR_ICEX_SEC "icexsec" ++#define AUFS_BRATTR_ICEX_SYS "icexsys" ++#define AUFS_BRATTR_ICEX_TR "icextr" ++#define AUFS_BRATTR_ICEX_USR "icexusr" ++#define AUFS_BRATTR_ICEX_OTH "icexoth" ++#define AUFS_BRRATTR_WH "wh" ++#define AUFS_BRWATTR_NLWH "nolwh" ++#define AUFS_BRWATTR_MOO "moo" ++ ++#define AuBrPerm_RW 1 /* writable, hardlinkable wh */ ++#define AuBrPerm_RO (1 << 1) /* readonly */ ++#define AuBrPerm_RR (1 << 2) /* natively readonly */ ++#define AuBrPerm_Mask (AuBrPerm_RW | AuBrPerm_RO | AuBrPerm_RR) ++ ++#define AuBrAttr_COO_REG (1 << 3) /* copy-up on open */ ++#define AuBrAttr_COO_ALL (1 << 4) ++#define AuBrAttr_COO_Mask (AuBrAttr_COO_REG | AuBrAttr_COO_ALL) ++ ++#define AuBrAttr_FHSM (1 << 5) /* file-based hsm */ ++#define AuBrAttr_UNPIN (1 << 6) /* rename-able top dir of ++ branch. meaningless since ++ linux-3.18-rc1 */ ++ ++/* ignore error in copying XATTR */ ++#define AuBrAttr_ICEX_SEC (1 << 7) ++#define AuBrAttr_ICEX_SYS (1 << 8) ++#define AuBrAttr_ICEX_TR (1 << 9) ++#define AuBrAttr_ICEX_USR (1 << 10) ++#define AuBrAttr_ICEX_OTH (1 << 11) ++#define AuBrAttr_ICEX (AuBrAttr_ICEX_SEC \ ++ | AuBrAttr_ICEX_SYS \ ++ | AuBrAttr_ICEX_TR \ ++ | AuBrAttr_ICEX_USR \ ++ | AuBrAttr_ICEX_OTH) ++ ++#define AuBrRAttr_WH (1 << 12) /* whiteout-able */ ++#define AuBrRAttr_Mask AuBrRAttr_WH ++ ++#define AuBrWAttr_NoLinkWH (1 << 13) /* un-hardlinkable whiteouts */ ++#define AuBrWAttr_MOO (1 << 14) /* move-up on open */ ++#define AuBrWAttr_Mask (AuBrWAttr_NoLinkWH | AuBrWAttr_MOO) ++ ++#define AuBrAttr_CMOO_Mask (AuBrAttr_COO_Mask | AuBrWAttr_MOO) ++ ++/* #warning test userspace */ ++#ifdef __KERNEL__ ++#ifndef CONFIG_AUFS_FHSM ++#undef AuBrAttr_FHSM ++#define AuBrAttr_FHSM 0 ++#endif ++#ifndef CONFIG_AUFS_XATTR ++#undef AuBrAttr_ICEX ++#define AuBrAttr_ICEX 0 ++#undef AuBrAttr_ICEX_SEC ++#define AuBrAttr_ICEX_SEC 0 ++#undef AuBrAttr_ICEX_SYS ++#define AuBrAttr_ICEX_SYS 0 ++#undef AuBrAttr_ICEX_TR ++#define AuBrAttr_ICEX_TR 0 ++#undef AuBrAttr_ICEX_USR ++#define AuBrAttr_ICEX_USR 0 ++#undef AuBrAttr_ICEX_OTH ++#define AuBrAttr_ICEX_OTH 0 ++#endif ++#endif ++ ++/* the longest combination */ ++/* AUFS_BRATTR_ICEX and AUFS_BRATTR_ICEX_TR don't affect here */ ++#define AuBrPermStrSz sizeof(AUFS_BRPERM_RW \ ++ "+" AUFS_BRATTR_COO_REG \ ++ "+" AUFS_BRATTR_FHSM \ ++ "+" AUFS_BRATTR_UNPIN \ ++ "+" AUFS_BRATTR_ICEX_SEC \ ++ "+" AUFS_BRATTR_ICEX_SYS \ ++ "+" AUFS_BRATTR_ICEX_USR \ ++ "+" AUFS_BRATTR_ICEX_OTH \ ++ "+" AUFS_BRWATTR_NLWH) ++ ++typedef struct { ++ char a[AuBrPermStrSz]; ++} au_br_perm_str_t; ++ ++static inline int au_br_writable(int brperm) ++{ ++ return brperm & AuBrPerm_RW; ++} ++ ++static inline int au_br_whable(int brperm) ++{ ++ return brperm & (AuBrPerm_RW | AuBrRAttr_WH); ++} ++ ++static inline int au_br_wh_linkable(int brperm) ++{ ++ return !(brperm & AuBrWAttr_NoLinkWH); ++} ++ ++static inline int au_br_cmoo(int brperm) ++{ ++ return brperm & AuBrAttr_CMOO_Mask; ++} ++ ++static inline int au_br_fhsm(int brperm) ++{ ++ return brperm & AuBrAttr_FHSM; ++} ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* ioctl */ ++enum { ++ /* readdir in userspace */ ++ AuCtl_RDU, ++ AuCtl_RDU_INO, ++ ++ AuCtl_WBR_FD, /* pathconf wrapper */ ++ AuCtl_IBUSY, /* busy inode */ ++ AuCtl_MVDOWN, /* move-down */ ++ AuCtl_BR, /* info about branches */ ++ AuCtl_FHSM_FD /* connection for fhsm */ ++}; ++ ++/* borrowed from linux/include/linux/kernel.h */ ++#ifndef ALIGN ++#define ALIGN(x, a) __ALIGN_MASK(x, (typeof(x))(a)-1) ++#define __ALIGN_MASK(x, mask) (((x)+(mask))&~(mask)) ++#endif ++ ++/* borrowed from linux/include/linux/compiler-gcc3.h */ ++#ifndef __aligned ++#define __aligned(x) __attribute__((aligned(x))) ++#endif ++ ++#ifdef __KERNEL__ ++#ifndef __packed ++#define __packed __attribute__((packed)) ++#endif ++#endif ++ ++struct au_rdu_cookie { ++ uint64_t h_pos; ++ int16_t bindex; ++ uint8_t flags; ++ uint8_t pad; ++ uint32_t generation; ++} __aligned(8); ++ ++struct au_rdu_ent { ++ uint64_t ino; ++ int16_t bindex; ++ uint8_t type; ++ uint8_t nlen; ++ uint8_t wh; ++ char name[0]; ++} __aligned(8); ++ ++static inline int au_rdu_len(int nlen) ++{ ++ /* include the terminating NULL */ ++ return ALIGN(sizeof(struct au_rdu_ent) + nlen + 1, ++ sizeof(uint64_t)); ++} ++ ++union au_rdu_ent_ul { ++ struct au_rdu_ent __user *e; ++ uint64_t ul; ++}; ++ ++enum { ++ AufsCtlRduV_SZ, ++ AufsCtlRduV_End ++}; ++ ++struct aufs_rdu { ++ /* input */ ++ union { ++ uint64_t sz; /* AuCtl_RDU */ ++ uint64_t nent; /* AuCtl_RDU_INO */ ++ }; ++ union au_rdu_ent_ul ent; ++ uint16_t verify[AufsCtlRduV_End]; ++ ++ /* input/output */ ++ uint32_t blk; ++ ++ /* output */ ++ union au_rdu_ent_ul tail; ++ /* number of entries which were added in a single call */ ++ uint64_t rent; ++ uint8_t full; ++ uint8_t shwh; ++ ++ struct au_rdu_cookie cookie; ++} __aligned(8); ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct aufs_wbr_fd { ++ uint32_t oflags; ++ int16_t brid; ++} __aligned(8); ++ ++/* ---------------------------------------------------------------------- */ ++ ++struct aufs_ibusy { ++ uint64_t ino, h_ino; ++ int16_t bindex; ++} __aligned(8); ++ ++/* ---------------------------------------------------------------------- */ ++ ++/* error code for move-down */ ++/* the actual message strings are implemented in aufs-util.git */ ++enum { ++ EAU_MVDOWN_OPAQUE = 1, ++ EAU_MVDOWN_WHITEOUT, ++ EAU_MVDOWN_UPPER, ++ EAU_MVDOWN_BOTTOM, ++ EAU_MVDOWN_NOUPPER, ++ EAU_MVDOWN_NOLOWERBR, ++ EAU_Last ++}; ++ ++/* flags for move-down */ ++#define AUFS_MVDOWN_DMSG 1 ++#define AUFS_MVDOWN_OWLOWER (1 << 1) /* overwrite lower */ ++#define AUFS_MVDOWN_KUPPER (1 << 2) /* keep upper */ ++#define AUFS_MVDOWN_ROLOWER (1 << 3) /* do even if lower is RO */ ++#define AUFS_MVDOWN_ROLOWER_R (1 << 4) /* did on lower RO */ ++#define AUFS_MVDOWN_ROUPPER (1 << 5) /* do even if upper is RO */ ++#define AUFS_MVDOWN_ROUPPER_R (1 << 6) /* did on upper RO */ ++#define AUFS_MVDOWN_BRID_UPPER (1 << 7) /* upper brid */ ++#define AUFS_MVDOWN_BRID_LOWER (1 << 8) /* lower brid */ ++#define AUFS_MVDOWN_FHSM_LOWER (1 << 9) /* find fhsm attr for lower */ ++#define AUFS_MVDOWN_STFS (1 << 10) /* req. stfs */ ++#define AUFS_MVDOWN_STFS_FAILED (1 << 11) /* output: stfs is unusable */ ++#define AUFS_MVDOWN_BOTTOM (1 << 12) /* output: no more lowers */ ++ ++/* index for move-down */ ++enum { ++ AUFS_MVDOWN_UPPER, ++ AUFS_MVDOWN_LOWER, ++ AUFS_MVDOWN_NARRAY ++}; ++ ++/* ++ * additional info of move-down ++ * number of free blocks and inodes. ++ * subset of struct kstatfs, but smaller and always 64bit. ++ */ ++struct aufs_stfs { ++ uint64_t f_blocks; ++ uint64_t f_bavail; ++ uint64_t f_files; ++ uint64_t f_ffree; ++}; ++ ++struct aufs_stbr { ++ int16_t brid; /* optional input */ ++ int16_t bindex; /* output */ ++ struct aufs_stfs stfs; /* output when AUFS_MVDOWN_STFS set */ ++} __aligned(8); ++ ++struct aufs_mvdown { ++ uint32_t flags; /* input/output */ ++ struct aufs_stbr stbr[AUFS_MVDOWN_NARRAY]; /* input/output */ ++ int8_t au_errno; /* output */ ++} __aligned(8); ++ ++/* ---------------------------------------------------------------------- */ ++ ++union aufs_brinfo { ++ /* PATH_MAX may differ between kernel-space and user-space */ ++ char _spacer[4096]; ++ struct { ++ int16_t id; ++ int perm; ++ char path[0]; ++ }; ++} __aligned(8); ++ ++/* ---------------------------------------------------------------------- */ ++ ++#define AuCtlType 'A' ++#define AUFS_CTL_RDU _IOWR(AuCtlType, AuCtl_RDU, struct aufs_rdu) ++#define AUFS_CTL_RDU_INO _IOWR(AuCtlType, AuCtl_RDU_INO, struct aufs_rdu) ++#define AUFS_CTL_WBR_FD _IOW(AuCtlType, AuCtl_WBR_FD, \ ++ struct aufs_wbr_fd) ++#define AUFS_CTL_IBUSY _IOWR(AuCtlType, AuCtl_IBUSY, struct aufs_ibusy) ++#define AUFS_CTL_MVDOWN _IOWR(AuCtlType, AuCtl_MVDOWN, \ ++ struct aufs_mvdown) ++#define AUFS_CTL_BRINFO _IOW(AuCtlType, AuCtl_BR, union aufs_brinfo) ++#define AUFS_CTL_FHSM_FD _IOW(AuCtlType, AuCtl_FHSM_FD, int) ++ ++#endif /* __AUFS_TYPE_H__ */ +diff -Nur linux-4.1.10.orig/include/uapi/linux/Kbuild linux-4.1.10/include/uapi/linux/Kbuild +--- linux-4.1.10.orig/include/uapi/linux/Kbuild 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/include/uapi/linux/Kbuild 2015-10-22 21:35:53.000000000 +0200 +@@ -59,6 +59,7 @@ + header-y += atm_tcp.h + header-y += atm_zatm.h + header-y += audit.h ++header-y += aufs_type.h + header-y += auto_fs4.h + header-y += auto_fs.h + header-y += auxvec.h +diff -Nur linux-4.1.10.orig/kernel/fork.c linux-4.1.10/kernel/fork.c +--- linux-4.1.10.orig/kernel/fork.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/kernel/fork.c 2015-10-22 21:35:53.000000000 +0200 +@@ -456,7 +456,7 @@ + struct inode *inode = file_inode(file); + struct address_space *mapping = file->f_mapping; + +- get_file(file); ++ vma_get_file(tmp); + if (tmp->vm_flags & VM_DENYWRITE) + atomic_dec(&inode->i_writecount); + i_mmap_lock_write(mapping); +diff -Nur linux-4.1.10.orig/MAINTAINERS linux-4.1.10/MAINTAINERS +--- linux-4.1.10.orig/MAINTAINERS 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/MAINTAINERS 2015-10-22 21:35:53.000000000 +0200 +@@ -1880,6 +1880,19 @@ + F: include/uapi/linux/audit.h + F: kernel/audit* + ++AUFS (advanced multi layered unification filesystem) FILESYSTEM ++M: "J. R. Okajima" <hooanon05g@gmail.com> ++L: linux-unionfs@vger.kernel.org ++L: aufs-users@lists.sourceforge.net (members only) ++W: http://aufs.sourceforge.net ++T: git://github.com/sfjro/aufs4-linux.git ++S: Supported ++F: Documentation/filesystems/aufs/ ++F: Documentation/ABI/testing/debugfs-aufs ++F: Documentation/ABI/testing/sysfs-aufs ++F: fs/aufs/ ++F: include/uapi/linux/aufs_type.h ++ + AUXILIARY DISPLAY DRIVERS + M: Miguel Ojeda Sandonis <miguel.ojeda.sandonis@gmail.com> + W: http://miguelojeda.es/auxdisplay.htm +diff -Nur linux-4.1.10.orig/mm/filemap.c linux-4.1.10/mm/filemap.c +--- linux-4.1.10.orig/mm/filemap.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/mm/filemap.c 2015-10-22 21:35:53.000000000 +0200 +@@ -2062,7 +2062,7 @@ + int ret = VM_FAULT_LOCKED; + + sb_start_pagefault(inode->i_sb); +- file_update_time(vma->vm_file); ++ vma_file_update_time(vma); + lock_page(page); + if (page->mapping != inode->i_mapping) { + unlock_page(page); +diff -Nur linux-4.1.10.orig/mm/Makefile linux-4.1.10/mm/Makefile +--- linux-4.1.10.orig/mm/Makefile 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/mm/Makefile 2015-10-22 21:35:53.000000000 +0200 +@@ -21,7 +21,7 @@ + mm_init.o mmu_context.o percpu.o slab_common.o \ + compaction.o vmacache.o \ + interval_tree.o list_lru.o workingset.o \ +- debug.o $(mmu-y) ++ prfile.o debug.o $(mmu-y) + + obj-y += init-mm.o + +diff -Nur linux-4.1.10.orig/mm/memory.c linux-4.1.10/mm/memory.c +--- linux-4.1.10.orig/mm/memory.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/mm/memory.c 2015-10-22 21:35:53.000000000 +0200 +@@ -2034,7 +2034,7 @@ + } + + if (!page_mkwrite) +- file_update_time(vma->vm_file); ++ vma_file_update_time(vma); + } + + return VM_FAULT_WRITE; +diff -Nur linux-4.1.10.orig/mm/mmap.c linux-4.1.10/mm/mmap.c +--- linux-4.1.10.orig/mm/mmap.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/mm/mmap.c 2015-10-22 21:35:53.000000000 +0200 +@@ -274,7 +274,7 @@ + if (vma->vm_ops && vma->vm_ops->close) + vma->vm_ops->close(vma); + if (vma->vm_file) +- fput(vma->vm_file); ++ vma_fput(vma); + mpol_put(vma_policy(vma)); + kmem_cache_free(vm_area_cachep, vma); + return next; +@@ -886,7 +886,7 @@ + if (remove_next) { + if (file) { + uprobe_munmap(next, next->vm_start, next->vm_end); +- fput(file); ++ vma_fput(vma); + } + if (next->anon_vma) + anon_vma_merge(vma, next); +@@ -1671,8 +1671,8 @@ + return addr; + + unmap_and_free_vma: ++ vma_fput(vma); + vma->vm_file = NULL; +- fput(file); + + /* Undo any partial mapping done by a device driver. */ + unmap_region(mm, vma, prev, vma->vm_start, vma->vm_end); +@@ -2473,7 +2473,7 @@ + goto out_free_mpol; + + if (new->vm_file) +- get_file(new->vm_file); ++ vma_get_file(new); + + if (new->vm_ops && new->vm_ops->open) + new->vm_ops->open(new); +@@ -2492,7 +2492,7 @@ + if (new->vm_ops && new->vm_ops->close) + new->vm_ops->close(new); + if (new->vm_file) +- fput(new->vm_file); ++ vma_fput(new); + unlink_anon_vmas(new); + out_free_mpol: + mpol_put(vma_policy(new)); +@@ -2635,7 +2635,6 @@ + struct vm_area_struct *vma; + unsigned long populate = 0; + unsigned long ret = -EINVAL; +- struct file *file; + + pr_warn_once("%s (%d) uses deprecated remap_file_pages() syscall. " + "See Documentation/vm/remap_file_pages.txt.\n", +@@ -2679,10 +2678,10 @@ + munlock_vma_pages_range(vma, start, start + size); + } + +- file = get_file(vma->vm_file); ++ vma_get_file(vma); + ret = do_mmap_pgoff(vma->vm_file, start, size, + prot, flags, pgoff, &populate); +- fput(file); ++ vma_fput(vma); + out: + up_write(&mm->mmap_sem); + if (populate) +@@ -2949,7 +2948,7 @@ + if (anon_vma_clone(new_vma, vma)) + goto out_free_mempol; + if (new_vma->vm_file) +- get_file(new_vma->vm_file); ++ vma_get_file(new_vma); + if (new_vma->vm_ops && new_vma->vm_ops->open) + new_vma->vm_ops->open(new_vma); + vma_link(mm, new_vma, prev, rb_link, rb_parent); +diff -Nur linux-4.1.10.orig/mm/nommu.c linux-4.1.10/mm/nommu.c +--- linux-4.1.10.orig/mm/nommu.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/mm/nommu.c 2015-10-22 21:35:53.000000000 +0200 +@@ -693,7 +693,7 @@ + up_write(&nommu_region_sem); + + if (region->vm_file) +- fput(region->vm_file); ++ vmr_fput(region); + + /* IO memory and memory shared directly out of the pagecache + * from ramfs/tmpfs mustn't be released here */ +@@ -858,7 +858,7 @@ + if (vma->vm_ops && vma->vm_ops->close) + vma->vm_ops->close(vma); + if (vma->vm_file) +- fput(vma->vm_file); ++ vma_fput(vma); + put_nommu_region(vma->vm_region); + kmem_cache_free(vm_area_cachep, vma); + } +@@ -1398,7 +1398,7 @@ + goto error_just_free; + } + } +- fput(region->vm_file); ++ vmr_fput(region); + kmem_cache_free(vm_region_jar, region); + region = pregion; + result = start; +@@ -1474,10 +1474,10 @@ + up_write(&nommu_region_sem); + error: + if (region->vm_file) +- fput(region->vm_file); ++ vmr_fput(region); + kmem_cache_free(vm_region_jar, region); + if (vma->vm_file) +- fput(vma->vm_file); ++ vma_fput(vma); + kmem_cache_free(vm_area_cachep, vma); + kleave(" = %d", ret); + return ret; +diff -Nur linux-4.1.10.orig/mm/prfile.c linux-4.1.10/mm/prfile.c +--- linux-4.1.10.orig/mm/prfile.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.10/mm/prfile.c 2015-10-22 21:35:53.000000000 +0200 +@@ -0,0 +1,86 @@ ++/* ++ * Mainly for aufs which mmap(2) diffrent file and wants to print different path ++ * in /proc/PID/maps. ++ * Call these functions via macros defined in linux/mm.h. ++ * ++ * See Documentation/filesystems/aufs/design/06mmap.txt ++ * ++ * Copyright (c) 2014 Junjro R. Okajima ++ * Copyright (c) 2014 Ian Campbell ++ */ ++ ++#include <linux/mm.h> ++#include <linux/file.h> ++#include <linux/fs.h> ++ ++/* #define PRFILE_TRACE */ ++static inline void prfile_trace(struct file *f, struct file *pr, ++ const char func[], int line, const char func2[]) ++{ ++#ifdef PRFILE_TRACE ++ if (pr) ++ pr_info("%s:%d: %s, %s\n", func, line, func2, ++ f ? (char *)f->f_path.dentry->d_name.name : "(null)"); ++#endif ++} ++ ++void vma_do_file_update_time(struct vm_area_struct *vma, const char func[], ++ int line) ++{ ++ struct file *f = vma->vm_file, *pr = vma->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ file_update_time(f); ++ if (f && pr) ++ file_update_time(pr); ++} ++ ++struct file *vma_do_pr_or_file(struct vm_area_struct *vma, const char func[], ++ int line) ++{ ++ struct file *f = vma->vm_file, *pr = vma->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ return (f && pr) ? pr : f; ++} ++ ++void vma_do_get_file(struct vm_area_struct *vma, const char func[], int line) ++{ ++ struct file *f = vma->vm_file, *pr = vma->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ get_file(f); ++ if (f && pr) ++ get_file(pr); ++} ++ ++void vma_do_fput(struct vm_area_struct *vma, const char func[], int line) ++{ ++ struct file *f = vma->vm_file, *pr = vma->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ fput(f); ++ if (f && pr) ++ fput(pr); ++} ++ ++#ifndef CONFIG_MMU ++struct file *vmr_do_pr_or_file(struct vm_region *region, const char func[], ++ int line) ++{ ++ struct file *f = region->vm_file, *pr = region->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ return (f && pr) ? pr : f; ++} ++ ++void vmr_do_fput(struct vm_region *region, const char func[], int line) ++{ ++ struct file *f = region->vm_file, *pr = region->vm_prfile; ++ ++ prfile_trace(f, pr, func, line, __func__); ++ fput(f); ++ if (f && pr) ++ fput(pr); ++} ++#endif /* !CONFIG_MMU */ +diff -Nur linux-4.1.10.orig/security/commoncap.c linux-4.1.10/security/commoncap.c +--- linux-4.1.10.orig/security/commoncap.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/security/commoncap.c 2015-10-22 21:35:53.000000000 +0200 +@@ -975,9 +975,11 @@ + } + return ret; + } ++EXPORT_SYMBOL(cap_mmap_addr); + + int cap_mmap_file(struct file *file, unsigned long reqprot, + unsigned long prot, unsigned long flags) + { + return 0; + } ++EXPORT_SYMBOL(cap_mmap_file); +diff -Nur linux-4.1.10.orig/security/device_cgroup.c linux-4.1.10/security/device_cgroup.c +--- linux-4.1.10.orig/security/device_cgroup.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/security/device_cgroup.c 2015-10-22 21:35:53.000000000 +0200 +@@ -7,6 +7,7 @@ + #include <linux/device_cgroup.h> + #include <linux/cgroup.h> + #include <linux/ctype.h> ++#include <linux/export.h> + #include <linux/list.h> + #include <linux/uaccess.h> + #include <linux/seq_file.h> +@@ -849,6 +850,7 @@ + return __devcgroup_check_permission(type, imajor(inode), iminor(inode), + access); + } ++EXPORT_SYMBOL(__devcgroup_inode_permission); + + int devcgroup_inode_mknod(int mode, dev_t dev) + { +diff -Nur linux-4.1.10.orig/security/security.c linux-4.1.10/security/security.c +--- linux-4.1.10.orig/security/security.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/security/security.c 2015-10-22 21:35:53.000000000 +0200 +@@ -430,6 +430,7 @@ + return 0; + return security_ops->path_rmdir(dir, dentry); + } ++EXPORT_SYMBOL(security_path_rmdir); + + int security_path_unlink(struct path *dir, struct dentry *dentry) + { +@@ -446,6 +447,7 @@ + return 0; + return security_ops->path_symlink(dir, dentry, old_name); + } ++EXPORT_SYMBOL(security_path_symlink); + + int security_path_link(struct dentry *old_dentry, struct path *new_dir, + struct dentry *new_dentry) +@@ -454,6 +456,7 @@ + return 0; + return security_ops->path_link(old_dentry, new_dir, new_dentry); + } ++EXPORT_SYMBOL(security_path_link); + + int security_path_rename(struct path *old_dir, struct dentry *old_dentry, + struct path *new_dir, struct dentry *new_dentry, +@@ -481,6 +484,7 @@ + return 0; + return security_ops->path_truncate(path); + } ++EXPORT_SYMBOL(security_path_truncate); + + int security_path_chmod(struct path *path, umode_t mode) + { +@@ -488,6 +492,7 @@ + return 0; + return security_ops->path_chmod(path, mode); + } ++EXPORT_SYMBOL(security_path_chmod); + + int security_path_chown(struct path *path, kuid_t uid, kgid_t gid) + { +@@ -495,6 +500,7 @@ + return 0; + return security_ops->path_chown(path, uid, gid); + } ++EXPORT_SYMBOL(security_path_chown); + + int security_path_chroot(struct path *path) + { +@@ -580,6 +586,7 @@ + return 0; + return security_ops->inode_readlink(dentry); + } ++EXPORT_SYMBOL(security_inode_readlink); + + int security_inode_follow_link(struct dentry *dentry, struct nameidata *nd) + { +@@ -594,6 +601,7 @@ + return 0; + return security_ops->inode_permission(inode, mask); + } ++EXPORT_SYMBOL(security_inode_permission); + + int security_inode_setattr(struct dentry *dentry, struct iattr *attr) + { +@@ -716,6 +724,7 @@ + + return fsnotify_perm(file, mask); + } ++EXPORT_SYMBOL(security_file_permission); + + int security_file_alloc(struct file *file) + { +@@ -775,6 +784,7 @@ + return ret; + return ima_file_mmap(file, prot); + } ++EXPORT_SYMBOL(security_mmap_file); + + int security_mmap_addr(unsigned long addr) + { diff --git a/target/linux/patches/4.1.13/cleankernel.patch b/target/linux/patches/4.1.13/cleankernel.patch new file mode 100644 index 000000000..59693f426 --- /dev/null +++ b/target/linux/patches/4.1.13/cleankernel.patch @@ -0,0 +1,11 @@ +diff -Nur linux-4.1.10.orig/scripts/Makefile.headersinst linux-4.1.10/scripts/Makefile.headersinst +--- linux-4.1.10.orig/scripts/Makefile.headersinst 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/scripts/Makefile.headersinst 2015-10-15 11:23:35.000000000 +0200 +@@ -107,7 +107,6 @@ + + targets += $(install-file) + $(install-file): scripts/headers_install.sh $(input-files1) $(input-files2) $(input-files3) FORCE +- $(if $(unwanted),$(call cmd,remove),) + $(if $(wildcard $(dir $@)),,$(shell mkdir -p $(dir $@))) + $(call if_changed,install) + diff --git a/target/linux/patches/4.1.13/cris-header.patch b/target/linux/patches/4.1.13/cris-header.patch new file mode 100644 index 000000000..2b5a88461 --- /dev/null +++ b/target/linux/patches/4.1.13/cris-header.patch @@ -0,0 +1,12 @@ +diff -Nur linux-3.16.2.orig/arch/cris/include/arch-v10/arch/Kbuild linux-3.16.2/arch/cris/include/arch-v10/arch/Kbuild +--- linux-3.16.2.orig/arch/cris/include/arch-v10/arch/Kbuild 2014-09-06 01:37:11.000000000 +0200 ++++ linux-3.16.2/arch/cris/include/arch-v10/arch/Kbuild 2014-09-26 19:24:50.000000000 +0200 +@@ -1 +1,2 @@ + # CRISv10 arch ++header-y += ptrace.h +diff -Nur linux-3.16.2.orig/arch/cris/include/arch-v32/arch/Kbuild linux-3.16.2/arch/cris/include/arch-v32/arch/Kbuild +--- linux-3.16.2.orig/arch/cris/include/arch-v32/arch/Kbuild 2014-09-06 01:37:11.000000000 +0200 ++++ linux-3.16.2/arch/cris/include/arch-v32/arch/Kbuild 2014-09-26 19:24:31.000000000 +0200 +@@ -1 +1,2 @@ + # CRISv32 arch ++header-y += ptrace.h diff --git a/target/linux/patches/4.1.13/initramfs-nosizelimit.patch b/target/linux/patches/4.1.13/initramfs-nosizelimit.patch new file mode 100644 index 000000000..40d2f6bd8 --- /dev/null +++ b/target/linux/patches/4.1.13/initramfs-nosizelimit.patch @@ -0,0 +1,57 @@ +From 9a18df7a71bfa620b1278777d64783a359d7eb4e Mon Sep 17 00:00:00 2001 +From: Thorsten Glaser <tg@mirbsd.org> +Date: Sun, 4 May 2014 01:37:54 +0200 +Subject: [PATCH] mount tmpfs-as-rootfs (initramfs) with -o + nr_blocks=0,nr_inodes=0 + +I would have preferred to write this patch to be able to pass +rootflags=nr_blocks=0,nr_inodes=0 on the kernel command line, +and then hand these rootflags over to the initramfs (tmpfs) +mount in the same way the kernel hands them over to the block +device rootfs mount. But at least the Debian/m68k initrd also +parses $rootflags from the environment and adds it to the call +to the user-space mount for the eventual root device, which +would make the kernel command line rootflags option be used in +both places (tmpfs and e.g. ext4) which is guaranteed to error +out in at least one of them. + +This change is intended to aid people in a setup where the +initrd is the final root filesystem, i.e. not mounted over. +This is especially useful in automated tests running on qemu +for boards with constrained memory (e.g. 64 MiB on sh4). + +Considering that the initramfs is normally emptied out then +overmounted, this change is probably safe for setups where +initramfs just hosts early userspace, too, since the tmpfs +backing it is not accessible any more later on, AFAICT. + +Signed-off-by: Thorsten Glaser <tg@mirbsd.org> +--- + init/do_mounts.c | 4 ++++ + 1 file changed, 4 insertions(+) + +diff --git a/init/do_mounts.c b/init/do_mounts.c +index 82f2288..55a4cfe 100644 +--- a/init/do_mounts.c ++++ b/init/do_mounts.c +@@ -594,6 +594,7 @@ out: + } + + static bool is_tmpfs; ++static char tmpfs_rootflags[] = "nr_blocks=0,nr_inodes=0"; + static struct dentry *rootfs_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data) + { +@@ -606,6 +607,9 @@ static struct dentry *rootfs_mount(struct file_system_type *fs_type, + if (IS_ENABLED(CONFIG_TMPFS) && is_tmpfs) + fill = shmem_fill_super; + ++ if (is_tmpfs) ++ data = tmpfs_rootflags; ++ + return mount_nodev(fs_type, flags, data, fill); + } + +-- +2.0.0.rc0 + diff --git a/target/linux/patches/4.1.13/realtime.patch b/target/linux/patches/4.1.13/realtime.patch new file mode 100644 index 000000000..3b65f6148 --- /dev/null +++ b/target/linux/patches/4.1.13/realtime.patch @@ -0,0 +1,27435 @@ +diff -Nur linux-4.1.13.orig/arch/alpha/mm/fault.c linux-4.1.13/arch/alpha/mm/fault.c +--- linux-4.1.13.orig/arch/alpha/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/alpha/mm/fault.c 2015-11-29 09:23:09.477622951 +0100 +@@ -23,8 +23,7 @@ + #include <linux/smp.h> + #include <linux/interrupt.h> + #include <linux/module.h> +- +-#include <asm/uaccess.h> ++#include <linux/uaccess.h> + + extern void die_if_kernel(char *,struct pt_regs *,long, unsigned long *); + +@@ -107,7 +106,7 @@ + + /* If we're in an interrupt context, or have no user context, + we must not take the fault. */ +- if (!mm || in_atomic()) ++ if (!mm || faulthandler_disabled()) + goto no_context; + + #ifdef CONFIG_ALPHA_LARGE_VMALLOC +diff -Nur linux-4.1.13.orig/arch/arc/include/asm/futex.h linux-4.1.13/arch/arc/include/asm/futex.h +--- linux-4.1.13.orig/arch/arc/include/asm/futex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arc/include/asm/futex.h 2015-11-29 09:23:09.477622951 +0100 +@@ -53,7 +53,7 @@ + if (!access_ok(VERIFY_WRITE, uaddr, sizeof(int))) + return -EFAULT; + +- pagefault_disable(); /* implies preempt_disable() */ ++ pagefault_disable(); + + switch (op) { + case FUTEX_OP_SET: +@@ -75,7 +75,7 @@ + ret = -ENOSYS; + } + +- pagefault_enable(); /* subsumes preempt_enable() */ ++ pagefault_enable(); + + if (!ret) { + switch (cmp) { +@@ -104,7 +104,7 @@ + return ret; + } + +-/* Compare-xchg with preemption disabled. ++/* Compare-xchg with pagefaults disabled. + * Notes: + * -Best-Effort: Exchg happens only if compare succeeds. + * If compare fails, returns; leaving retry/looping to upper layers +@@ -121,7 +121,7 @@ + if (!access_ok(VERIFY_WRITE, uaddr, sizeof(int))) + return -EFAULT; + +- pagefault_disable(); /* implies preempt_disable() */ ++ pagefault_disable(); + + /* TBD : can use llock/scond */ + __asm__ __volatile__( +@@ -142,7 +142,7 @@ + : "r"(oldval), "r"(newval), "r"(uaddr), "ir"(-EFAULT) + : "cc", "memory"); + +- pagefault_enable(); /* subsumes preempt_enable() */ ++ pagefault_enable(); + + *uval = val; + return val; +diff -Nur linux-4.1.13.orig/arch/arc/mm/fault.c linux-4.1.13/arch/arc/mm/fault.c +--- linux-4.1.13.orig/arch/arc/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arc/mm/fault.c 2015-11-29 09:23:09.477622951 +0100 +@@ -86,7 +86,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/arm/include/asm/cmpxchg.h linux-4.1.13/arch/arm/include/asm/cmpxchg.h +--- linux-4.1.13.orig/arch/arm/include/asm/cmpxchg.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/include/asm/cmpxchg.h 2015-11-29 09:23:09.477622951 +0100 +@@ -129,6 +129,8 @@ + + #else /* min ARCH >= ARMv6 */ + ++#define __HAVE_ARCH_CMPXCHG 1 ++ + extern void __bad_cmpxchg(volatile void *ptr, int size); + + /* +diff -Nur linux-4.1.13.orig/arch/arm/include/asm/futex.h linux-4.1.13/arch/arm/include/asm/futex.h +--- linux-4.1.13.orig/arch/arm/include/asm/futex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/include/asm/futex.h 2015-11-29 09:23:09.477622951 +0100 +@@ -93,6 +93,7 @@ + if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) + return -EFAULT; + ++ preempt_disable(); + __asm__ __volatile__("@futex_atomic_cmpxchg_inatomic\n" + "1: " TUSER(ldr) " %1, [%4]\n" + " teq %1, %2\n" +@@ -104,6 +105,8 @@ + : "cc", "memory"); + + *uval = val; ++ preempt_enable(); ++ + return ret; + } + +@@ -124,7 +127,10 @@ + if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) + return -EFAULT; + +- pagefault_disable(); /* implies preempt_disable() */ ++#ifndef CONFIG_SMP ++ preempt_disable(); ++#endif ++ pagefault_disable(); + + switch (op) { + case FUTEX_OP_SET: +@@ -146,7 +152,10 @@ + ret = -ENOSYS; + } + +- pagefault_enable(); /* subsumes preempt_enable() */ ++ pagefault_enable(); ++#ifndef CONFIG_SMP ++ preempt_enable(); ++#endif + + if (!ret) { + switch (cmp) { +diff -Nur linux-4.1.13.orig/arch/arm/include/asm/switch_to.h linux-4.1.13/arch/arm/include/asm/switch_to.h +--- linux-4.1.13.orig/arch/arm/include/asm/switch_to.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/include/asm/switch_to.h 2015-11-29 09:23:09.477622951 +0100 +@@ -3,6 +3,13 @@ + + #include <linux/thread_info.h> + ++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p); ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + /* + * For v7 SMP cores running a preemptible kernel we may be pre-empted + * during a TLB maintenance operation, so execute an inner-shareable dsb +@@ -22,6 +29,7 @@ + + #define switch_to(prev,next,last) \ + do { \ ++ switch_kmaps(prev, next); \ + last = __switch_to(prev,task_thread_info(prev), task_thread_info(next)); \ + } while (0) + +diff -Nur linux-4.1.13.orig/arch/arm/include/asm/thread_info.h linux-4.1.13/arch/arm/include/asm/thread_info.h +--- linux-4.1.13.orig/arch/arm/include/asm/thread_info.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/include/asm/thread_info.h 2015-11-29 09:23:09.477622951 +0100 +@@ -50,6 +50,7 @@ + struct thread_info { + unsigned long flags; /* low level flags */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + __u32 cpu; /* cpu */ +@@ -147,6 +148,7 @@ + #define TIF_SIGPENDING 0 + #define TIF_NEED_RESCHED 1 + #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ ++#define TIF_NEED_RESCHED_LAZY 3 + #define TIF_UPROBE 7 + #define TIF_SYSCALL_TRACE 8 + #define TIF_SYSCALL_AUDIT 9 +@@ -160,6 +162,7 @@ + #define _TIF_SIGPENDING (1 << TIF_SIGPENDING) + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +diff -Nur linux-4.1.13.orig/arch/arm/Kconfig linux-4.1.13/arch/arm/Kconfig +--- linux-4.1.13.orig/arch/arm/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/Kconfig 2015-11-29 09:23:09.477622951 +0100 +@@ -31,7 +31,7 @@ + select HARDIRQS_SW_RESEND + select HAVE_ARCH_AUDITSYSCALL if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_BITREVERSE if (CPU_32v7M || CPU_32v7) && !CPU_32v6 +- select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL ++ select HAVE_ARCH_JUMP_LABEL if (!XIP_KERNEL && !PREEMPT_RT_BASE) + select HAVE_ARCH_KGDB + select HAVE_ARCH_SECCOMP_FILTER if (AEABI && !OABI_COMPAT) + select HAVE_ARCH_TRACEHOOK +@@ -66,6 +66,7 @@ + select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP ++ select HAVE_PREEMPT_LAZY + select HAVE_RCU_TABLE_FREE if (SMP && ARM_LPAE) + select HAVE_REGS_AND_STACK_ACCESS_API + select HAVE_SYSCALL_TRACEPOINTS +diff -Nur linux-4.1.13.orig/arch/arm/kernel/asm-offsets.c linux-4.1.13/arch/arm/kernel/asm-offsets.c +--- linux-4.1.13.orig/arch/arm/kernel/asm-offsets.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/asm-offsets.c 2015-11-29 09:23:09.477622951 +0100 +@@ -65,6 +65,7 @@ + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff -Nur linux-4.1.13.orig/arch/arm/kernel/entry-armv.S linux-4.1.13/arch/arm/kernel/entry-armv.S +--- linux-4.1.13.orig/arch/arm/kernel/entry-armv.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/entry-armv.S 2015-11-29 09:23:09.477622951 +0100 +@@ -208,11 +208,18 @@ + #ifdef CONFIG_PREEMPT + get_thread_info tsk + ldr r8, [tsk, #TI_PREEMPT] @ get preempt count +- ldr r0, [tsk, #TI_FLAGS] @ get flags + teq r8, #0 @ if preempt count != 0 ++ bne 1f @ return from exeption ++ ldr r0, [tsk, #TI_FLAGS] @ get flags ++ tst r0, #_TIF_NEED_RESCHED @ if NEED_RESCHED is set ++ blne svc_preempt @ preempt! ++ ++ ldr r8, [tsk, #TI_PREEMPT_LAZY] @ get preempt lazy count ++ teq r8, #0 @ if preempt lazy count != 0 + movne r0, #0 @ force flags to 0 +- tst r0, #_TIF_NEED_RESCHED ++ tst r0, #_TIF_NEED_RESCHED_LAZY + blne svc_preempt ++1: + #endif + + svc_exit r5, irq = 1 @ return from exception +@@ -227,6 +234,8 @@ + 1: bl preempt_schedule_irq @ irq en/disable is done inside + ldr r0, [tsk, #TI_FLAGS] @ get new tasks TI_FLAGS + tst r0, #_TIF_NEED_RESCHED ++ bne 1b ++ tst r0, #_TIF_NEED_RESCHED_LAZY + reteq r8 @ go again + b 1b + #endif +diff -Nur linux-4.1.13.orig/arch/arm/kernel/process.c linux-4.1.13/arch/arm/kernel/process.c +--- linux-4.1.13.orig/arch/arm/kernel/process.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/process.c 2015-11-29 09:23:09.477622951 +0100 +@@ -290,6 +290,30 @@ + } + + #ifdef CONFIG_MMU ++/* ++ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not ++ * initialized by pgtable_page_ctor() then a coredump of the vector page will ++ * fail. ++ */ ++static int __init vectors_user_mapping_init_page(void) ++{ ++ struct page *page; ++ unsigned long addr = 0xffff0000; ++ pgd_t *pgd; ++ pud_t *pud; ++ pmd_t *pmd; ++ ++ pgd = pgd_offset_k(addr); ++ pud = pud_offset(pgd, addr); ++ pmd = pmd_offset(pud, addr); ++ page = pmd_page(*(pmd)); ++ ++ pgtable_page_ctor(page); ++ ++ return 0; ++} ++late_initcall(vectors_user_mapping_init_page); ++ + #ifdef CONFIG_KUSER_HELPERS + /* + * The vectors page is always readable from user space for the +diff -Nur linux-4.1.13.orig/arch/arm/kernel/signal.c linux-4.1.13/arch/arm/kernel/signal.c +--- linux-4.1.13.orig/arch/arm/kernel/signal.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/signal.c 2015-11-29 09:23:09.477622951 +0100 +@@ -568,7 +568,8 @@ + do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall) + { + do { +- if (likely(thread_flags & _TIF_NEED_RESCHED)) { ++ if (likely(thread_flags & (_TIF_NEED_RESCHED | ++ _TIF_NEED_RESCHED_LAZY))) { + schedule(); + } else { + if (unlikely(!user_mode(regs))) +diff -Nur linux-4.1.13.orig/arch/arm/kernel/smp.c linux-4.1.13/arch/arm/kernel/smp.c +--- linux-4.1.13.orig/arch/arm/kernel/smp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/smp.c 2015-11-29 09:23:09.481622687 +0100 +@@ -213,8 +213,6 @@ + flush_cache_louis(); + local_flush_tlb_all(); + +- clear_tasks_mm_cpumask(cpu); +- + return 0; + } + +@@ -230,6 +228,9 @@ + pr_err("CPU%u: cpu didn't die\n", cpu); + return; + } ++ ++ clear_tasks_mm_cpumask(cpu); ++ + pr_notice("CPU%u: shutdown\n", cpu); + + /* +diff -Nur linux-4.1.13.orig/arch/arm/kernel/unwind.c linux-4.1.13/arch/arm/kernel/unwind.c +--- linux-4.1.13.orig/arch/arm/kernel/unwind.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kernel/unwind.c 2015-11-29 09:23:09.481622687 +0100 +@@ -93,7 +93,7 @@ + static const struct unwind_idx *__origin_unwind_idx; + extern const struct unwind_idx __stop_unwind_idx[]; + +-static DEFINE_SPINLOCK(unwind_lock); ++static DEFINE_RAW_SPINLOCK(unwind_lock); + static LIST_HEAD(unwind_tables); + + /* Convert a prel31 symbol to an absolute address */ +@@ -201,7 +201,7 @@ + /* module unwind tables */ + struct unwind_table *table; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_for_each_entry(table, &unwind_tables, list) { + if (addr >= table->begin_addr && + addr < table->end_addr) { +@@ -213,7 +213,7 @@ + break; + } + } +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + } + + pr_debug("%s: idx = %p\n", __func__, idx); +@@ -529,9 +529,9 @@ + tab->begin_addr = text_addr; + tab->end_addr = text_addr + text_size; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_add_tail(&tab->list, &unwind_tables); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + + return tab; + } +@@ -543,9 +543,9 @@ + if (!tab) + return; + +- spin_lock_irqsave(&unwind_lock, flags); ++ raw_spin_lock_irqsave(&unwind_lock, flags); + list_del(&tab->list); +- spin_unlock_irqrestore(&unwind_lock, flags); ++ raw_spin_unlock_irqrestore(&unwind_lock, flags); + + kfree(tab); + } +diff -Nur linux-4.1.13.orig/arch/arm/kvm/arm.c linux-4.1.13/arch/arm/kvm/arm.c +--- linux-4.1.13.orig/arch/arm/kvm/arm.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kvm/arm.c 2015-11-29 09:23:09.481622687 +0100 +@@ -474,9 +474,9 @@ + + static void vcpu_pause(struct kvm_vcpu *vcpu) + { +- wait_queue_head_t *wq = kvm_arch_vcpu_wq(vcpu); ++ struct swait_head *wq = kvm_arch_vcpu_wq(vcpu); + +- wait_event_interruptible(*wq, !vcpu->arch.pause); ++ swait_event_interruptible(*wq, !vcpu->arch.pause); + } + + static int kvm_vcpu_initialized(struct kvm_vcpu *vcpu) +diff -Nur linux-4.1.13.orig/arch/arm/kvm/psci.c linux-4.1.13/arch/arm/kvm/psci.c +--- linux-4.1.13.orig/arch/arm/kvm/psci.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/kvm/psci.c 2015-11-29 09:23:09.501621354 +0100 +@@ -68,7 +68,7 @@ + { + struct kvm *kvm = source_vcpu->kvm; + struct kvm_vcpu *vcpu = NULL; +- wait_queue_head_t *wq; ++ struct swait_head *wq; + unsigned long cpu_id; + unsigned long context_id; + phys_addr_t target_pc; +@@ -117,7 +117,7 @@ + smp_mb(); /* Make sure the above is visible */ + + wq = kvm_arch_vcpu_wq(vcpu); +- wake_up_interruptible(wq); ++ swait_wake_interruptible(wq); + + return PSCI_RET_SUCCESS; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-exynos/platsmp.c linux-4.1.13/arch/arm/mach-exynos/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-exynos/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-exynos/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -231,7 +231,7 @@ + return (void __iomem *)(S5P_VA_SCU); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void exynos_secondary_init(unsigned int cpu) + { +@@ -244,8 +244,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int exynos_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -259,7 +259,7 @@ + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -286,7 +286,7 @@ + + if (timeout == 0) { + printk(KERN_ERR "cpu1 power enable failed"); +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + return -ETIMEDOUT; + } + } +@@ -342,7 +342,7 @@ + * calibrations, then wait for it to finish + */ + fail: +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? ret : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-hisi/platmcpm.c linux-4.1.13/arch/arm/mach-hisi/platmcpm.c +--- linux-4.1.13.orig/arch/arm/mach-hisi/platmcpm.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-hisi/platmcpm.c 2015-11-29 09:23:09.501621354 +0100 +@@ -57,7 +57,7 @@ + + static void __iomem *sysctrl, *fabric; + static int hip04_cpu_table[HIP04_MAX_CLUSTERS][HIP04_MAX_CPUS_PER_CLUSTER]; +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + static u32 fabric_phys_addr; + /* + * [0]: bootwrapper physical address +@@ -104,7 +104,7 @@ + if (cluster >= HIP04_MAX_CLUSTERS || cpu >= HIP04_MAX_CPUS_PER_CLUSTER) + return -EINVAL; + +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + + if (hip04_cpu_table[cluster][cpu]) + goto out; +@@ -133,7 +133,7 @@ + udelay(20); + out: + hip04_cpu_table[cluster][cpu]++; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + + return 0; + } +@@ -149,7 +149,7 @@ + + __mcpm_cpu_going_down(cpu, cluster); + +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + BUG_ON(__mcpm_cluster_state(cluster) != CLUSTER_UP); + hip04_cpu_table[cluster][cpu]--; + if (hip04_cpu_table[cluster][cpu] == 1) { +@@ -162,7 +162,7 @@ + + last_man = hip04_cluster_is_down(cluster); + if (last_man && __mcpm_outbound_enter_critical(cpu, cluster)) { +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + /* Since it's Cortex A15, disable L2 prefetching. */ + asm volatile( + "mcr p15, 1, %0, c15, c0, 3 \n\t" +@@ -173,7 +173,7 @@ + hip04_set_snoop_filter(cluster, 0); + __mcpm_outbound_leave_critical(cluster, CLUSTER_DOWN); + } else { +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + v7_exit_coherency_flush(louis); + } + +@@ -192,7 +192,7 @@ + cpu >= HIP04_MAX_CPUS_PER_CLUSTER); + + count = TIMEOUT_MSEC / POLL_MSEC; +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + for (tries = 0; tries < count; tries++) { + if (hip04_cpu_table[cluster][cpu]) { + ret = -EBUSY; +@@ -202,10 +202,10 @@ + data = readl_relaxed(sysctrl + SC_CPU_RESET_STATUS(cluster)); + if (data & CORE_WFI_STATUS(cpu)) + break; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + /* Wait for clean L2 when the whole cluster is down. */ + msleep(POLL_MSEC); +- spin_lock_irq(&boot_lock); ++ raw_spin_lock_irq(&boot_lock); + } + if (tries >= count) + goto err; +@@ -220,10 +220,10 @@ + } + if (tries >= count) + goto err; +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return 0; + err: +- spin_unlock_irq(&boot_lock); ++ raw_spin_unlock_irq(&boot_lock); + return ret; + } + +@@ -235,10 +235,10 @@ + cpu = MPIDR_AFFINITY_LEVEL(mpidr, 0); + cluster = MPIDR_AFFINITY_LEVEL(mpidr, 1); + +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + if (!hip04_cpu_table[cluster][cpu]) + hip04_cpu_table[cluster][cpu] = 1; +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static void __naked hip04_mcpm_power_up_setup(unsigned int affinity_level) +diff -Nur linux-4.1.13.orig/arch/arm/mach-omap2/omap-smp.c linux-4.1.13/arch/arm/mach-omap2/omap-smp.c +--- linux-4.1.13.orig/arch/arm/mach-omap2/omap-smp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-omap2/omap-smp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -43,7 +43,7 @@ + /* SCU base address */ + static void __iomem *scu_base; + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + void __iomem *omap4_get_scu_base(void) + { +@@ -74,8 +74,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int omap4_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -89,7 +89,7 @@ + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * Update the AuxCoreBoot0 with boot state for secondary core. +@@ -166,7 +166,7 @@ + * Now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-prima2/platsmp.c linux-4.1.13/arch/arm/mach-prima2/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-prima2/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-prima2/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -22,7 +22,7 @@ + + static void __iomem *clk_base; + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void sirfsoc_secondary_init(unsigned int cpu) + { +@@ -36,8 +36,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static const struct of_device_id clk_ids[] = { +@@ -75,7 +75,7 @@ + /* make sure write buffer is drained */ + mb(); + +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -107,7 +107,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-qcom/platsmp.c linux-4.1.13/arch/arm/mach-qcom/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-qcom/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-qcom/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -46,7 +46,7 @@ + + extern void secondary_startup_arm(void); + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + #ifdef CONFIG_HOTPLUG_CPU + static void __ref qcom_cpu_die(unsigned int cpu) +@@ -60,8 +60,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int scss_release_secondary(unsigned int cpu) +@@ -284,7 +284,7 @@ + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * Send the secondary CPU a soft interrupt, thereby causing +@@ -297,7 +297,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return ret; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-spear/platsmp.c linux-4.1.13/arch/arm/mach-spear/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-spear/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-spear/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -32,7 +32,7 @@ + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void __iomem *scu_base = IOMEM(VA_SCU_BASE); + +@@ -47,8 +47,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int spear13xx_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -59,7 +59,7 @@ + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -84,7 +84,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-sti/platsmp.c linux-4.1.13/arch/arm/mach-sti/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-sti/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-sti/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -34,7 +34,7 @@ + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void sti_secondary_init(unsigned int cpu) + { +@@ -49,8 +49,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int sti_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -61,7 +61,7 @@ + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -92,7 +92,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mach-ux500/platsmp.c linux-4.1.13/arch/arm/mach-ux500/platsmp.c +--- linux-4.1.13.orig/arch/arm/mach-ux500/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mach-ux500/platsmp.c 2015-11-29 09:23:09.501621354 +0100 +@@ -51,7 +51,7 @@ + return NULL; + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + static void ux500_secondary_init(unsigned int cpu) + { +@@ -64,8 +64,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + static int ux500_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -76,7 +76,7 @@ + * set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * The secondary processor is waiting to be released from +@@ -97,7 +97,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mm/fault.c linux-4.1.13/arch/arm/mm/fault.c +--- linux-4.1.13.orig/arch/arm/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mm/fault.c 2015-11-29 09:23:09.501621354 +0100 +@@ -276,7 +276,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +@@ -430,6 +430,9 @@ + if (addr < TASK_SIZE) + return do_page_fault(addr, fsr, regs); + ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + if (user_mode(regs)) + goto bad_area; + +@@ -497,6 +500,9 @@ + static int + do_sect_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) + { ++ if (interrupts_enabled(regs)) ++ local_irq_enable(); ++ + do_bad_area(addr, fsr, regs); + return 0; + } +diff -Nur linux-4.1.13.orig/arch/arm/mm/highmem.c linux-4.1.13/arch/arm/mm/highmem.c +--- linux-4.1.13.orig/arch/arm/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/mm/highmem.c 2015-11-29 09:23:09.505621090 +0100 +@@ -54,11 +54,13 @@ + + void *kmap_atomic(struct page *page) + { ++ pte_t pte = mk_pte(page, kmap_prot); + unsigned int idx; + unsigned long vaddr; + void *kmap; + int type; + ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -92,7 +94,10 @@ + * in place, so the contained TLB flush ensures the TLB is updated + * with the new mapping. + */ +- set_fixmap_pte(idx, mk_pte(page, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); + + return (void *)vaddr; + } +@@ -109,27 +114,33 @@ + + if (cache_is_vivt()) + __cpuc_flush_dcache_area((void *)vaddr, PAGE_SIZE); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(vaddr != __fix_to_virt(idx)); +- set_fixmap_pte(idx, __pte(0)); + #else + (void) idx; /* to kill a warning */ + #endif ++ set_fixmap_pte(idx, __pte(0)); + kmap_atomic_idx_pop(); + } else if (vaddr >= PKMAP_ADDR(0) && vaddr < PKMAP_ADDR(LAST_PKMAP)) { + /* this address was obtained through kmap_high_get() */ + kunmap_high(pte_page(pkmap_page_table[PKMAP_NR(vaddr)])); + } + pagefault_enable(); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(__kunmap_atomic); + + void *kmap_atomic_pfn(unsigned long pfn) + { ++ pte_t pte = pfn_pte(pfn, kmap_prot); + unsigned long vaddr; + int idx, type; + struct page *page = pfn_to_page(pfn); + ++ preempt_disable_nort(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -140,7 +151,10 @@ + #ifdef CONFIG_DEBUG_HIGHMEM + BUG_ON(!pte_none(get_fixmap_pte(vaddr))); + #endif +- set_fixmap_pte(idx, pfn_pte(pfn, kmap_prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_fixmap_pte(idx, pte); + + return (void *)vaddr; + } +@@ -154,3 +168,28 @@ + + return pte_page(get_fixmap_pte(vaddr)); + } ++ ++#if defined CONFIG_PREEMPT_RT_FULL ++void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++{ ++ int i; ++ ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ ++ set_fixmap_pte(idx, __pte(0)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_fixmap_pte(idx, next_p->kmap_pte[i]); ++ } ++} ++#endif +diff -Nur linux-4.1.13.orig/arch/arm/plat-versatile/platsmp.c linux-4.1.13/arch/arm/plat-versatile/platsmp.c +--- linux-4.1.13.orig/arch/arm/plat-versatile/platsmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm/plat-versatile/platsmp.c 2015-11-29 09:23:09.505621090 +0100 +@@ -30,7 +30,7 @@ + sync_cache_w(&pen_release); + } + +-static DEFINE_SPINLOCK(boot_lock); ++static DEFINE_RAW_SPINLOCK(boot_lock); + + void versatile_secondary_init(unsigned int cpu) + { +@@ -43,8 +43,8 @@ + /* + * Synchronise with the boot thread. + */ +- spin_lock(&boot_lock); +- spin_unlock(&boot_lock); ++ raw_spin_lock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + } + + int versatile_boot_secondary(unsigned int cpu, struct task_struct *idle) +@@ -55,7 +55,7 @@ + * Set synchronisation state between this boot processor + * and the secondary one + */ +- spin_lock(&boot_lock); ++ raw_spin_lock(&boot_lock); + + /* + * This is really belt and braces; we hold unintended secondary +@@ -85,7 +85,7 @@ + * now the secondary core is starting up let it run its + * calibrations, then wait for it to finish + */ +- spin_unlock(&boot_lock); ++ raw_spin_unlock(&boot_lock); + + return pen_release != -1 ? -ENOSYS : 0; + } +diff -Nur linux-4.1.13.orig/arch/arm64/include/asm/futex.h linux-4.1.13/arch/arm64/include/asm/futex.h +--- linux-4.1.13.orig/arch/arm64/include/asm/futex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/include/asm/futex.h 2015-11-29 09:23:09.505621090 +0100 +@@ -58,7 +58,7 @@ + if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))) + return -EFAULT; + +- pagefault_disable(); /* implies preempt_disable() */ ++ pagefault_disable(); + + switch (op) { + case FUTEX_OP_SET: +@@ -85,7 +85,7 @@ + ret = -ENOSYS; + } + +- pagefault_enable(); /* subsumes preempt_enable() */ ++ pagefault_enable(); + + if (!ret) { + switch (cmp) { +diff -Nur linux-4.1.13.orig/arch/arm64/include/asm/thread_info.h linux-4.1.13/arch/arm64/include/asm/thread_info.h +--- linux-4.1.13.orig/arch/arm64/include/asm/thread_info.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/include/asm/thread_info.h 2015-11-29 09:23:09.505621090 +0100 +@@ -47,6 +47,7 @@ + mm_segment_t addr_limit; /* address limit */ + struct task_struct *task; /* main task structure */ + int preempt_count; /* 0 => preemptable, <0 => bug */ ++ int preempt_lazy_count; /* 0 => preemptable, <0 => bug */ + int cpu; /* cpu */ + }; + +@@ -101,6 +102,7 @@ + #define TIF_NEED_RESCHED 1 + #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ + #define TIF_FOREIGN_FPSTATE 3 /* CPU's FP state is not current's */ ++#define TIF_NEED_RESCHED_LAZY 4 + #define TIF_NOHZ 7 + #define TIF_SYSCALL_TRACE 8 + #define TIF_SYSCALL_AUDIT 9 +@@ -117,6 +119,7 @@ + #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) + #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) + #define _TIF_FOREIGN_FPSTATE (1 << TIF_FOREIGN_FPSTATE) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_NOHZ (1 << TIF_NOHZ) + #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +diff -Nur linux-4.1.13.orig/arch/arm64/Kconfig linux-4.1.13/arch/arm64/Kconfig +--- linux-4.1.13.orig/arch/arm64/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/Kconfig 2015-11-29 09:23:09.505621090 +0100 +@@ -69,8 +69,10 @@ + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP + select HAVE_RCU_TABLE_FREE ++ select HAVE_PREEMPT_LAZY + select HAVE_SYSCALL_TRACEPOINTS + select IRQ_DOMAIN ++ select IRQ_FORCED_THREADING + select MODULES_USE_ELF_RELA + select NO_BOOTMEM + select OF +@@ -599,7 +601,7 @@ + + config XEN + bool "Xen guest support on ARM64" +- depends on ARM64 && OF ++ depends on ARM64 && OF && !PREEMPT_RT_FULL + select SWIOTLB_XEN + help + Say Y if you want to run Linux in a Virtual Machine on Xen on ARM64. +diff -Nur linux-4.1.13.orig/arch/arm64/kernel/asm-offsets.c linux-4.1.13/arch/arm64/kernel/asm-offsets.c +--- linux-4.1.13.orig/arch/arm64/kernel/asm-offsets.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/kernel/asm-offsets.c 2015-11-29 09:23:09.505621090 +0100 +@@ -35,6 +35,7 @@ + BLANK(); + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); +diff -Nur linux-4.1.13.orig/arch/arm64/kernel/debug-monitors.c linux-4.1.13/arch/arm64/kernel/debug-monitors.c +--- linux-4.1.13.orig/arch/arm64/kernel/debug-monitors.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/kernel/debug-monitors.c 2015-11-29 09:23:09.505621090 +0100 +@@ -271,20 +271,21 @@ + * Use reader/writer locks instead of plain spinlock. + */ + static LIST_HEAD(break_hook); +-static DEFINE_RWLOCK(break_hook_lock); ++static DEFINE_SPINLOCK(break_hook_lock); + + void register_break_hook(struct break_hook *hook) + { +- write_lock(&break_hook_lock); +- list_add(&hook->node, &break_hook); +- write_unlock(&break_hook_lock); ++ spin_lock(&break_hook_lock); ++ list_add_rcu(&hook->node, &break_hook); ++ spin_unlock(&break_hook_lock); + } + + void unregister_break_hook(struct break_hook *hook) + { +- write_lock(&break_hook_lock); +- list_del(&hook->node); +- write_unlock(&break_hook_lock); ++ spin_lock(&break_hook_lock); ++ list_del_rcu(&hook->node); ++ spin_unlock(&break_hook_lock); ++ synchronize_rcu(); + } + + static int call_break_hook(struct pt_regs *regs, unsigned int esr) +@@ -292,11 +293,11 @@ + struct break_hook *hook; + int (*fn)(struct pt_regs *regs, unsigned int esr) = NULL; + +- read_lock(&break_hook_lock); +- list_for_each_entry(hook, &break_hook, node) ++ rcu_read_lock(); ++ list_for_each_entry_rcu(hook, &break_hook, node) + if ((esr & hook->esr_mask) == hook->esr_val) + fn = hook->fn; +- read_unlock(&break_hook_lock); ++ rcu_read_unlock(); + + return fn ? fn(regs, esr) : DBG_HOOK_ERROR; + } +diff -Nur linux-4.1.13.orig/arch/arm64/kernel/entry.S linux-4.1.13/arch/arm64/kernel/entry.S +--- linux-4.1.13.orig/arch/arm64/kernel/entry.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/kernel/entry.S 2015-11-29 09:23:09.505621090 +0100 +@@ -367,11 +367,16 @@ + #ifdef CONFIG_PREEMPT + get_thread_info tsk + ldr w24, [tsk, #TI_PREEMPT] // get preempt count +- cbnz w24, 1f // preempt count != 0 ++ cbnz w24, 2f // preempt count != 0 + ldr x0, [tsk, #TI_FLAGS] // get flags +- tbz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? +- bl el1_preempt ++ tbnz x0, #TIF_NEED_RESCHED, 1f // needs rescheduling? ++ ++ ldr w24, [tsk, #TI_PREEMPT_LAZY] // get preempt lazy count ++ cbnz w24, 2f // preempt lazy count != 0 ++ tbz x0, #TIF_NEED_RESCHED_LAZY, 2f // needs rescheduling? + 1: ++ bl el1_preempt ++2: + #endif + #ifdef CONFIG_TRACE_IRQFLAGS + bl trace_hardirqs_on +@@ -385,6 +390,7 @@ + 1: bl preempt_schedule_irq // irq en/disable is done inside + ldr x0, [tsk, #TI_FLAGS] // get new tasks TI_FLAGS + tbnz x0, #TIF_NEED_RESCHED, 1b // needs rescheduling? ++ tbnz x0, #TIF_NEED_RESCHED_LAZY, 1b // needs rescheduling? + ret x24 + #endif + +@@ -622,6 +628,7 @@ + str x0, [sp, #S_X0] // returned x0 + work_pending: + tbnz x1, #TIF_NEED_RESCHED, work_resched ++ tbnz x1, #TIF_NEED_RESCHED_LAZY, work_resched + /* TIF_SIGPENDING, TIF_NOTIFY_RESUME or TIF_FOREIGN_FPSTATE case */ + ldr x2, [sp, #S_PSTATE] + mov x0, sp // 'regs' +diff -Nur linux-4.1.13.orig/arch/arm64/kernel/insn.c linux-4.1.13/arch/arm64/kernel/insn.c +--- linux-4.1.13.orig/arch/arm64/kernel/insn.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/kernel/insn.c 2015-11-29 09:23:09.505621090 +0100 +@@ -77,7 +77,7 @@ + } + } + +-static DEFINE_SPINLOCK(patch_lock); ++static DEFINE_RAW_SPINLOCK(patch_lock); + + static void __kprobes *patch_map(void *addr, int fixmap) + { +@@ -124,13 +124,13 @@ + unsigned long flags = 0; + int ret; + +- spin_lock_irqsave(&patch_lock, flags); ++ raw_spin_lock_irqsave(&patch_lock, flags); + waddr = patch_map(addr, FIX_TEXT_POKE0); + + ret = probe_kernel_write(waddr, &insn, AARCH64_INSN_SIZE); + + patch_unmap(FIX_TEXT_POKE0); +- spin_unlock_irqrestore(&patch_lock, flags); ++ raw_spin_unlock_irqrestore(&patch_lock, flags); + + return ret; + } +diff -Nur linux-4.1.13.orig/arch/arm64/kernel/perf_event.c linux-4.1.13/arch/arm64/kernel/perf_event.c +--- linux-4.1.13.orig/arch/arm64/kernel/perf_event.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/kernel/perf_event.c 2015-11-29 09:23:09.505621090 +0100 +@@ -488,7 +488,7 @@ + } + + err = request_irq(irq, armpmu->handle_irq, +- IRQF_NOBALANCING, ++ IRQF_NOBALANCING | IRQF_NO_THREAD, + "arm-pmu", armpmu); + if (err) { + pr_err("unable to request IRQ%d for ARM PMU counters\n", +diff -Nur linux-4.1.13.orig/arch/arm64/mm/fault.c linux-4.1.13/arch/arm64/mm/fault.c +--- linux-4.1.13.orig/arch/arm64/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/arm64/mm/fault.c 2015-11-29 09:23:09.505621090 +0100 +@@ -211,7 +211,7 @@ + * If we're in an interrupt or have no user context, we must not take + * the fault. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/avr32/include/asm/uaccess.h linux-4.1.13/arch/avr32/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/avr32/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/avr32/include/asm/uaccess.h 2015-11-29 09:23:09.505621090 +0100 +@@ -97,7 +97,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -116,7 +117,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -136,7 +138,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -158,7 +161,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +diff -Nur linux-4.1.13.orig/arch/avr32/mm/fault.c linux-4.1.13/arch/avr32/mm/fault.c +--- linux-4.1.13.orig/arch/avr32/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/avr32/mm/fault.c 2015-11-29 09:23:09.505621090 +0100 +@@ -14,11 +14,11 @@ + #include <linux/pagemap.h> + #include <linux/kdebug.h> + #include <linux/kprobes.h> ++#include <linux/uaccess.h> + + #include <asm/mmu_context.h> + #include <asm/sysreg.h> + #include <asm/tlb.h> +-#include <asm/uaccess.h> + + #ifdef CONFIG_KPROBES + static inline int notify_page_fault(struct pt_regs *regs, int trap) +@@ -81,7 +81,7 @@ + * If we're in an interrupt or have no user context, we must + * not take the fault... + */ +- if (in_atomic() || !mm || regs->sr & SYSREG_BIT(GM)) ++ if (faulthandler_disabled() || !mm || regs->sr & SYSREG_BIT(GM)) + goto no_context; + + local_irq_enable(); +diff -Nur linux-4.1.13.orig/arch/cris/mm/fault.c linux-4.1.13/arch/cris/mm/fault.c +--- linux-4.1.13.orig/arch/cris/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/cris/mm/fault.c 2015-11-29 09:23:09.505621090 +0100 +@@ -8,7 +8,7 @@ + #include <linux/interrupt.h> + #include <linux/module.h> + #include <linux/wait.h> +-#include <asm/uaccess.h> ++#include <linux/uaccess.h> + #include <arch/system.h> + + extern int find_fixup_code(struct pt_regs *); +@@ -109,11 +109,11 @@ + info.si_code = SEGV_MAPERR; + + /* +- * If we're in an interrupt or "atomic" operation or have no ++ * If we're in an interrupt, have pagefaults disabled or have no + * user context, we must not take the fault. + */ + +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/frv/mm/fault.c linux-4.1.13/arch/frv/mm/fault.c +--- linux-4.1.13.orig/arch/frv/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/frv/mm/fault.c 2015-11-29 09:23:09.505621090 +0100 +@@ -19,9 +19,9 @@ + #include <linux/kernel.h> + #include <linux/ptrace.h> + #include <linux/hardirq.h> ++#include <linux/uaccess.h> + + #include <asm/pgtable.h> +-#include <asm/uaccess.h> + #include <asm/gdb-stub.h> + + /*****************************************************************************/ +@@ -78,7 +78,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(__frame)) +diff -Nur linux-4.1.13.orig/arch/frv/mm/highmem.c linux-4.1.13/arch/frv/mm/highmem.c +--- linux-4.1.13.orig/arch/frv/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/frv/mm/highmem.c 2015-11-29 09:23:09.505621090 +0100 +@@ -42,6 +42,7 @@ + unsigned long paddr; + int type; + ++ preempt_disable(); + pagefault_disable(); + type = kmap_atomic_idx_push(); + paddr = page_to_phys(page); +@@ -85,5 +86,6 @@ + } + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); +diff -Nur linux-4.1.13.orig/arch/hexagon/include/asm/uaccess.h linux-4.1.13/arch/hexagon/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/hexagon/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/hexagon/include/asm/uaccess.h 2015-11-29 09:23:09.505621090 +0100 +@@ -36,7 +36,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +diff -Nur linux-4.1.13.orig/arch/ia64/mm/fault.c linux-4.1.13/arch/ia64/mm/fault.c +--- linux-4.1.13.orig/arch/ia64/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/ia64/mm/fault.c 2015-11-29 09:23:09.505621090 +0100 +@@ -11,10 +11,10 @@ + #include <linux/kprobes.h> + #include <linux/kdebug.h> + #include <linux/prefetch.h> ++#include <linux/uaccess.h> + + #include <asm/pgtable.h> + #include <asm/processor.h> +-#include <asm/uaccess.h> + + extern int die(char *, struct pt_regs *, long); + +@@ -96,7 +96,7 @@ + /* + * If we're in an interrupt or have no user context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + #ifdef CONFIG_VIRTUAL_MEM_MAP +diff -Nur linux-4.1.13.orig/arch/Kconfig linux-4.1.13/arch/Kconfig +--- linux-4.1.13.orig/arch/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/Kconfig 2015-11-29 09:23:09.477622951 +0100 +@@ -6,6 +6,7 @@ + tristate "OProfile system profiling" + depends on PROFILING + depends on HAVE_OPROFILE ++ depends on !PREEMPT_RT_FULL + select RING_BUFFER + select RING_BUFFER_ALLOW_SWAP + help +diff -Nur linux-4.1.13.orig/arch/m32r/include/asm/uaccess.h linux-4.1.13/arch/m32r/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/m32r/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/m32r/include/asm/uaccess.h 2015-11-29 09:23:09.509620826 +0100 +@@ -91,7 +91,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +@@ -155,7 +156,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -175,7 +177,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -194,7 +197,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -274,7 +278,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -568,7 +573,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -588,7 +594,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. + * +@@ -606,7 +613,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -626,7 +634,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. + * +@@ -677,7 +686,8 @@ + * strlen_user: - Get the size of a string in user space. + * @str: The string to measure. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +diff -Nur linux-4.1.13.orig/arch/m32r/mm/fault.c linux-4.1.13/arch/m32r/mm/fault.c +--- linux-4.1.13.orig/arch/m32r/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/m32r/mm/fault.c 2015-11-29 09:23:09.509620826 +0100 +@@ -24,9 +24,9 @@ + #include <linux/vt_kern.h> /* For unblank_screen() */ + #include <linux/highmem.h> + #include <linux/module.h> ++#include <linux/uaccess.h> + + #include <asm/m32r.h> +-#include <asm/uaccess.h> + #include <asm/hardirq.h> + #include <asm/mmu_context.h> + #include <asm/tlbflush.h> +@@ -111,10 +111,10 @@ + mm = tsk->mm; + + /* +- * If we're in an interrupt or have no user context or are running in an +- * atomic region then we must not take the fault.. ++ * If we're in an interrupt or have no user context or have pagefaults ++ * disabled then we must not take the fault. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto bad_area_nosemaphore; + + if (error_code & ACE_USERMODE) +diff -Nur linux-4.1.13.orig/arch/m68k/mm/fault.c linux-4.1.13/arch/m68k/mm/fault.c +--- linux-4.1.13.orig/arch/m68k/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/m68k/mm/fault.c 2015-11-29 09:23:09.509620826 +0100 +@@ -10,10 +10,10 @@ + #include <linux/ptrace.h> + #include <linux/interrupt.h> + #include <linux/module.h> ++#include <linux/uaccess.h> + + #include <asm/setup.h> + #include <asm/traps.h> +-#include <asm/uaccess.h> + #include <asm/pgalloc.h> + + extern void die_if_kernel(char *, struct pt_regs *, long); +@@ -81,7 +81,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/metag/mm/fault.c linux-4.1.13/arch/metag/mm/fault.c +--- linux-4.1.13.orig/arch/metag/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/metag/mm/fault.c 2015-11-29 09:23:09.509620826 +0100 +@@ -105,7 +105,7 @@ + + mm = tsk->mm; + +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/metag/mm/highmem.c linux-4.1.13/arch/metag/mm/highmem.c +--- linux-4.1.13.orig/arch/metag/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/metag/mm/highmem.c 2015-11-29 09:23:09.509620826 +0100 +@@ -43,7 +43,7 @@ + unsigned long vaddr; + int type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -82,6 +82,7 @@ + } + + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +@@ -95,6 +96,7 @@ + unsigned long vaddr; + int type; + ++ preempt_disable(); + pagefault_disable(); + + type = kmap_atomic_idx_push(); +diff -Nur linux-4.1.13.orig/arch/microblaze/include/asm/uaccess.h linux-4.1.13/arch/microblaze/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/microblaze/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/microblaze/include/asm/uaccess.h 2015-11-29 09:23:09.509620826 +0100 +@@ -178,7 +178,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -290,7 +291,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +diff -Nur linux-4.1.13.orig/arch/microblaze/mm/fault.c linux-4.1.13/arch/microblaze/mm/fault.c +--- linux-4.1.13.orig/arch/microblaze/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/microblaze/mm/fault.c 2015-11-29 09:23:09.509620826 +0100 +@@ -107,14 +107,14 @@ + if ((error_code & 0x13) == 0x13 || (error_code & 0x11) == 0x11) + is_write = 0; + +- if (unlikely(in_atomic() || !mm)) { ++ if (unlikely(faulthandler_disabled() || !mm)) { + if (kernel_mode(regs)) + goto bad_area_nosemaphore; + +- /* in_atomic() in user mode is really bad, ++ /* faulthandler_disabled() in user mode is really bad, + as is current->mm == NULL. */ +- pr_emerg("Page fault in user mode with in_atomic(), mm = %p\n", +- mm); ++ pr_emerg("Page fault in user mode with faulthandler_disabled(), mm = %p\n", ++ mm); + pr_emerg("r15 = %lx MSR = %lx\n", + regs->r15, regs->msr); + die("Weird page fault", regs, SIGSEGV); +diff -Nur linux-4.1.13.orig/arch/microblaze/mm/highmem.c linux-4.1.13/arch/microblaze/mm/highmem.c +--- linux-4.1.13.orig/arch/microblaze/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/microblaze/mm/highmem.c 2015-11-29 09:23:09.509620826 +0100 +@@ -37,7 +37,7 @@ + unsigned long vaddr; + int idx, type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -63,6 +63,7 @@ + + if (vaddr < __fix_to_virt(FIX_KMAP_END)) { + pagefault_enable(); ++ preempt_enable(); + return; + } + +@@ -84,5 +85,6 @@ + #endif + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); +diff -Nur linux-4.1.13.orig/arch/mips/include/asm/uaccess.h linux-4.1.13/arch/mips/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/mips/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/include/asm/uaccess.h 2015-11-29 09:23:09.513620562 +0100 +@@ -103,7 +103,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +@@ -138,7 +139,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -157,7 +159,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -177,7 +180,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -199,7 +203,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -498,7 +503,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -517,7 +523,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -537,7 +544,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -559,7 +567,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -815,7 +824,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -888,7 +898,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. + * +@@ -1075,7 +1086,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -1107,7 +1119,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. + * +@@ -1329,7 +1342,8 @@ + * strlen_user: - Get the size of a string in user space. + * @str: The string to measure. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +@@ -1398,7 +1412,8 @@ + * strnlen_user: - Get the size of a string in user space. + * @str: The string to measure. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +diff -Nur linux-4.1.13.orig/arch/mips/Kconfig linux-4.1.13/arch/mips/Kconfig +--- linux-4.1.13.orig/arch/mips/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/Kconfig 2015-11-29 09:23:09.513620562 +0100 +@@ -2366,7 +2366,7 @@ + # + config HIGHMEM + bool "High Memory Support" +- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA ++ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !CPU_MIPS32_3_5_EVA && !PREEMPT_RT_FULL + + config CPU_SUPPORTS_HIGHMEM + bool +diff -Nur linux-4.1.13.orig/arch/mips/kernel/signal-common.h linux-4.1.13/arch/mips/kernel/signal-common.h +--- linux-4.1.13.orig/arch/mips/kernel/signal-common.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/kernel/signal-common.h 2015-11-29 09:23:09.513620562 +0100 +@@ -28,12 +28,7 @@ + extern int fpcsr_pending(unsigned int __user *fpcsr); + + /* Make sure we will not lose FPU ownership */ +-#ifdef CONFIG_PREEMPT +-#define lock_fpu_owner() preempt_disable() +-#define unlock_fpu_owner() preempt_enable() +-#else +-#define lock_fpu_owner() pagefault_disable() +-#define unlock_fpu_owner() pagefault_enable() +-#endif ++#define lock_fpu_owner() ({ preempt_disable(); pagefault_disable(); }) ++#define unlock_fpu_owner() ({ pagefault_enable(); preempt_enable(); }) + + #endif /* __SIGNAL_COMMON_H */ +diff -Nur linux-4.1.13.orig/arch/mips/mm/fault.c linux-4.1.13/arch/mips/mm/fault.c +--- linux-4.1.13.orig/arch/mips/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/mm/fault.c 2015-11-29 09:23:09.513620562 +0100 +@@ -21,10 +21,10 @@ + #include <linux/module.h> + #include <linux/kprobes.h> + #include <linux/perf_event.h> ++#include <linux/uaccess.h> + + #include <asm/branch.h> + #include <asm/mmu_context.h> +-#include <asm/uaccess.h> + #include <asm/ptrace.h> + #include <asm/highmem.h> /* For VMALLOC_END */ + #include <linux/kdebug.h> +@@ -94,7 +94,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto bad_area_nosemaphore; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/mips/mm/highmem.c linux-4.1.13/arch/mips/mm/highmem.c +--- linux-4.1.13.orig/arch/mips/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/mm/highmem.c 2015-11-29 09:23:09.513620562 +0100 +@@ -47,7 +47,7 @@ + unsigned long vaddr; + int idx, type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -72,6 +72,7 @@ + + if (vaddr < FIXADDR_START) { // FIXME + pagefault_enable(); ++ preempt_enable(); + return; + } + +@@ -92,6 +93,7 @@ + #endif + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +@@ -104,6 +106,7 @@ + unsigned long vaddr; + int idx, type; + ++ preempt_disable(); + pagefault_disable(); + + type = kmap_atomic_idx_push(); +diff -Nur linux-4.1.13.orig/arch/mips/mm/init.c linux-4.1.13/arch/mips/mm/init.c +--- linux-4.1.13.orig/arch/mips/mm/init.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mips/mm/init.c 2015-11-29 09:23:09.513620562 +0100 +@@ -90,6 +90,7 @@ + + BUG_ON(Page_dcache_dirty(page)); + ++ preempt_disable(); + pagefault_disable(); + idx = (addr >> PAGE_SHIFT) & (FIX_N_COLOURS - 1); + idx += in_interrupt() ? FIX_N_COLOURS : 0; +@@ -152,6 +153,7 @@ + write_c0_entryhi(old_ctx); + local_irq_restore(flags); + pagefault_enable(); ++ preempt_enable(); + } + + void copy_user_highpage(struct page *to, struct page *from, +diff -Nur linux-4.1.13.orig/arch/mn10300/include/asm/highmem.h linux-4.1.13/arch/mn10300/include/asm/highmem.h +--- linux-4.1.13.orig/arch/mn10300/include/asm/highmem.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mn10300/include/asm/highmem.h 2015-11-29 09:23:09.513620562 +0100 +@@ -75,6 +75,7 @@ + unsigned long vaddr; + int idx, type; + ++ preempt_disable(); + pagefault_disable(); + if (page < highmem_start_page) + return page_address(page); +@@ -98,6 +99,7 @@ + + if (vaddr < FIXADDR_START) { /* FIXME */ + pagefault_enable(); ++ preempt_enable(); + return; + } + +@@ -122,6 +124,7 @@ + + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + #endif /* __KERNEL__ */ + +diff -Nur linux-4.1.13.orig/arch/mn10300/mm/fault.c linux-4.1.13/arch/mn10300/mm/fault.c +--- linux-4.1.13.orig/arch/mn10300/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/mn10300/mm/fault.c 2015-11-29 09:23:09.513620562 +0100 +@@ -23,8 +23,8 @@ + #include <linux/interrupt.h> + #include <linux/init.h> + #include <linux/vt_kern.h> /* For unblank_screen() */ ++#include <linux/uaccess.h> + +-#include <asm/uaccess.h> + #include <asm/pgalloc.h> + #include <asm/hardirq.h> + #include <asm/cpu-regs.h> +@@ -168,7 +168,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if ((fault_code & MMUFCR_xFC_ACCESS) == MMUFCR_xFC_ACCESS_USR) +diff -Nur linux-4.1.13.orig/arch/nios2/mm/fault.c linux-4.1.13/arch/nios2/mm/fault.c +--- linux-4.1.13.orig/arch/nios2/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/nios2/mm/fault.c 2015-11-29 09:23:09.513620562 +0100 +@@ -77,7 +77,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto bad_area_nosemaphore; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/parisc/include/asm/cacheflush.h linux-4.1.13/arch/parisc/include/asm/cacheflush.h +--- linux-4.1.13.orig/arch/parisc/include/asm/cacheflush.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/parisc/include/asm/cacheflush.h 2015-11-29 09:23:09.513620562 +0100 +@@ -142,6 +142,7 @@ + + static inline void *kmap_atomic(struct page *page) + { ++ preempt_disable(); + pagefault_disable(); + return page_address(page); + } +@@ -150,6 +151,7 @@ + { + flush_kernel_dcache_page_addr(addr); + pagefault_enable(); ++ preempt_enable(); + } + + #define kmap_atomic_prot(page, prot) kmap_atomic(page) +diff -Nur linux-4.1.13.orig/arch/parisc/kernel/traps.c linux-4.1.13/arch/parisc/kernel/traps.c +--- linux-4.1.13.orig/arch/parisc/kernel/traps.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/parisc/kernel/traps.c 2015-11-29 09:23:09.513620562 +0100 +@@ -26,9 +26,9 @@ + #include <linux/console.h> + #include <linux/bug.h> + #include <linux/ratelimit.h> ++#include <linux/uaccess.h> + + #include <asm/assembly.h> +-#include <asm/uaccess.h> + #include <asm/io.h> + #include <asm/irq.h> + #include <asm/traps.h> +@@ -796,7 +796,7 @@ + * unless pagefault_disable() was called before. + */ + +- if (fault_space == 0 && !in_atomic()) ++ if (fault_space == 0 && !faulthandler_disabled()) + { + pdc_chassis_send_status(PDC_CHASSIS_DIRECT_PANIC); + parisc_terminate("Kernel Fault", regs, code, fault_address); +diff -Nur linux-4.1.13.orig/arch/parisc/mm/fault.c linux-4.1.13/arch/parisc/mm/fault.c +--- linux-4.1.13.orig/arch/parisc/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/parisc/mm/fault.c 2015-11-29 09:23:09.513620562 +0100 +@@ -15,8 +15,8 @@ + #include <linux/sched.h> + #include <linux/interrupt.h> + #include <linux/module.h> ++#include <linux/uaccess.h> + +-#include <asm/uaccess.h> + #include <asm/traps.h> + + /* Various important other fields */ +@@ -207,7 +207,7 @@ + int fault; + unsigned int flags; + +- if (in_atomic()) ++ if (pagefault_disabled()) + goto no_context; + + tsk = current; +diff -Nur linux-4.1.13.orig/arch/powerpc/include/asm/kvm_host.h linux-4.1.13/arch/powerpc/include/asm/kvm_host.h +--- linux-4.1.13.orig/arch/powerpc/include/asm/kvm_host.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/include/asm/kvm_host.h 2015-11-29 09:23:09.517620297 +0100 +@@ -280,7 +280,7 @@ + u8 in_guest; + struct list_head runnable_threads; + spinlock_t lock; +- wait_queue_head_t wq; ++ struct swait_head wq; + spinlock_t stoltb_lock; /* protects stolen_tb and preempt_tb */ + u64 stolen_tb; + u64 preempt_tb; +@@ -613,7 +613,7 @@ + u8 prodded; + u32 last_inst; + +- wait_queue_head_t *wqp; ++ struct swait_head *wqp; + struct kvmppc_vcore *vcore; + int ret; + int trap; +diff -Nur linux-4.1.13.orig/arch/powerpc/include/asm/thread_info.h linux-4.1.13/arch/powerpc/include/asm/thread_info.h +--- linux-4.1.13.orig/arch/powerpc/include/asm/thread_info.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/include/asm/thread_info.h 2015-11-29 09:23:09.517620297 +0100 +@@ -42,6 +42,8 @@ + int cpu; /* cpu we're on */ + int preempt_count; /* 0 => preemptable, + <0 => BUG */ ++ int preempt_lazy_count; /* 0 => preemptable, ++ <0 => BUG */ + unsigned long local_flags; /* private flags for thread */ + + /* low level flags - has atomic operations done on it */ +@@ -82,8 +84,7 @@ + #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ + #define TIF_SIGPENDING 1 /* signal pending */ + #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ +-#define TIF_POLLING_NRFLAG 3 /* true if poll_idle() is polling +- TIF_NEED_RESCHED */ ++#define TIF_NEED_RESCHED_LAZY 3 /* lazy rescheduling necessary */ + #define TIF_32BIT 4 /* 32 bit binary */ + #define TIF_RESTORE_TM 5 /* need to restore TM FP/VEC/VSX */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ +@@ -101,6 +102,8 @@ + #if defined(CONFIG_PPC64) + #define TIF_ELF2ABI 18 /* function descriptors must die! */ + #endif ++#define TIF_POLLING_NRFLAG 19 /* true if poll_idle() is polling ++ TIF_NEED_RESCHED */ + + /* as above, but as bit values */ + #define _TIF_SYSCALL_TRACE (1<<TIF_SYSCALL_TRACE) +@@ -119,14 +122,16 @@ + #define _TIF_SYSCALL_TRACEPOINT (1<<TIF_SYSCALL_TRACEPOINT) + #define _TIF_EMULATE_STACK_STORE (1<<TIF_EMULATE_STACK_STORE) + #define _TIF_NOHZ (1<<TIF_NOHZ) ++#define _TIF_NEED_RESCHED_LAZY (1<<TIF_NEED_RESCHED_LAZY) + #define _TIF_SYSCALL_DOTRACE (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ + _TIF_SECCOMP | _TIF_SYSCALL_TRACEPOINT | \ + _TIF_NOHZ) + + #define _TIF_USER_WORK_MASK (_TIF_SIGPENDING | _TIF_NEED_RESCHED | \ + _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ +- _TIF_RESTORE_TM) ++ _TIF_RESTORE_TM | _TIF_NEED_RESCHED_LAZY) + #define _TIF_PERSYSCALL_MASK (_TIF_RESTOREALL|_TIF_NOERROR) ++#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) + + /* Bits in local_flags */ + /* Don't move TLF_NAPPING without adjusting the code in entry_32.S */ +diff -Nur linux-4.1.13.orig/arch/powerpc/Kconfig linux-4.1.13/arch/powerpc/Kconfig +--- linux-4.1.13.orig/arch/powerpc/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/Kconfig 2015-11-29 09:23:09.513620562 +0100 +@@ -60,10 +60,11 @@ + + config RWSEM_GENERIC_SPINLOCK + bool ++ default y if PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM + bool +- default y ++ default y if !PREEMPT_RT_FULL + + config GENERIC_LOCKBREAK + bool +@@ -138,6 +139,7 @@ + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST + select GENERIC_STRNCPY_FROM_USER + select GENERIC_STRNLEN_USER ++ select HAVE_PREEMPT_LAZY + select HAVE_MOD_ARCH_SPECIFIC + select MODULES_USE_ELF_RELA + select CLONE_BACKWARDS +@@ -312,7 +314,7 @@ + + config HIGHMEM + bool "High memory support" +- depends on PPC32 ++ depends on PPC32 && !PREEMPT_RT_FULL + + source kernel/Kconfig.hz + source kernel/Kconfig.preempt +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/asm-offsets.c linux-4.1.13/arch/powerpc/kernel/asm-offsets.c +--- linux-4.1.13.orig/arch/powerpc/kernel/asm-offsets.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/asm-offsets.c 2015-11-29 09:23:09.517620297 +0100 +@@ -160,6 +160,7 @@ + DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); + DEFINE(TI_LOCAL_FLAGS, offsetof(struct thread_info, local_flags)); + DEFINE(TI_PREEMPT, offsetof(struct thread_info, preempt_count)); ++ DEFINE(TI_PREEMPT_LAZY, offsetof(struct thread_info, preempt_lazy_count)); + DEFINE(TI_TASK, offsetof(struct thread_info, task)); + DEFINE(TI_CPU, offsetof(struct thread_info, cpu)); + +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/entry_32.S linux-4.1.13/arch/powerpc/kernel/entry_32.S +--- linux-4.1.13.orig/arch/powerpc/kernel/entry_32.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/entry_32.S 2015-11-29 09:23:09.517620297 +0100 +@@ -813,7 +813,14 @@ + cmpwi 0,r0,0 /* if non-zero, just restore regs and return */ + bne restore + andi. r8,r8,_TIF_NEED_RESCHED ++ bne+ 1f ++ lwz r0,TI_PREEMPT_LAZY(r9) ++ cmpwi 0,r0,0 /* if non-zero, just restore regs and return */ ++ bne restore ++ lwz r0,TI_FLAGS(r9) ++ andi. r0,r0,_TIF_NEED_RESCHED_LAZY + beq+ restore ++1: + lwz r3,_MSR(r1) + andi. r0,r3,MSR_EE /* interrupts off? */ + beq restore /* don't schedule if so */ +@@ -824,11 +831,11 @@ + */ + bl trace_hardirqs_off + #endif +-1: bl preempt_schedule_irq ++2: bl preempt_schedule_irq + CURRENT_THREAD_INFO(r9, r1) + lwz r3,TI_FLAGS(r9) +- andi. r0,r3,_TIF_NEED_RESCHED +- bne- 1b ++ andi. r0,r3,_TIF_NEED_RESCHED_MASK ++ bne- 2b + #ifdef CONFIG_TRACE_IRQFLAGS + /* And now, to properly rebalance the above, we tell lockdep they + * are being turned back on, which will happen when we return +@@ -1149,7 +1156,7 @@ + #endif /* !(CONFIG_4xx || CONFIG_BOOKE) */ + + do_work: /* r10 contains MSR_KERNEL here */ +- andi. r0,r9,_TIF_NEED_RESCHED ++ andi. r0,r9,_TIF_NEED_RESCHED_MASK + beq do_user_signal + + do_resched: /* r10 contains MSR_KERNEL here */ +@@ -1170,7 +1177,7 @@ + MTMSRD(r10) /* disable interrupts */ + CURRENT_THREAD_INFO(r9, r1) + lwz r9,TI_FLAGS(r9) +- andi. r0,r9,_TIF_NEED_RESCHED ++ andi. r0,r9,_TIF_NEED_RESCHED_MASK + bne- do_resched + andi. r0,r9,_TIF_USER_WORK_MASK + beq restore_user +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/entry_64.S linux-4.1.13/arch/powerpc/kernel/entry_64.S +--- linux-4.1.13.orig/arch/powerpc/kernel/entry_64.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/entry_64.S 2015-11-29 09:23:09.517620297 +0100 +@@ -636,7 +636,7 @@ + #else + beq restore + #endif +-1: andi. r0,r4,_TIF_NEED_RESCHED ++1: andi. r0,r4,_TIF_NEED_RESCHED_MASK + beq 2f + bl restore_interrupts + SCHEDULE_USER +@@ -698,10 +698,18 @@ + + #ifdef CONFIG_PREEMPT + /* Check if we need to preempt */ ++ lwz r8,TI_PREEMPT(r9) ++ cmpwi 0,r8,0 /* if non-zero, just restore regs and return */ ++ bne restore + andi. r0,r4,_TIF_NEED_RESCHED ++ bne+ check_count ++ ++ andi. r0,r4,_TIF_NEED_RESCHED_LAZY + beq+ restore ++ lwz r8,TI_PREEMPT_LAZY(r9) ++ + /* Check that preempt_count() == 0 and interrupts are enabled */ +- lwz r8,TI_PREEMPT(r9) ++check_count: + cmpwi cr1,r8,0 + ld r0,SOFTE(r1) + cmpdi r0,0 +@@ -718,7 +726,7 @@ + /* Re-test flags and eventually loop */ + CURRENT_THREAD_INFO(r9, r1) + ld r4,TI_FLAGS(r9) +- andi. r0,r4,_TIF_NEED_RESCHED ++ andi. r0,r4,_TIF_NEED_RESCHED_MASK + bne 1b + + /* +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/irq.c linux-4.1.13/arch/powerpc/kernel/irq.c +--- linux-4.1.13.orig/arch/powerpc/kernel/irq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/irq.c 2015-11-29 09:23:09.517620297 +0100 +@@ -614,6 +614,7 @@ + } + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct thread_info *curtp, *irqtp; +@@ -631,6 +632,7 @@ + if (irqtp->flags) + set_bits(irqtp->flags, &curtp->flags); + } ++#endif + + irq_hw_number_t virq_to_hw(unsigned int virq) + { +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/misc_32.S linux-4.1.13/arch/powerpc/kernel/misc_32.S +--- linux-4.1.13.orig/arch/powerpc/kernel/misc_32.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/misc_32.S 2015-11-29 09:23:09.517620297 +0100 +@@ -40,6 +40,7 @@ + * We store the saved ksp_limit in the unused part + * of the STACK_FRAME_OVERHEAD + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + stw r0,4(r1) +@@ -56,6 +57,7 @@ + stw r10,THREAD+KSP_LIMIT(r2) + mtlr r0 + blr ++#endif + + /* + * void call_do_irq(struct pt_regs *regs, struct thread_info *irqtp); +diff -Nur linux-4.1.13.orig/arch/powerpc/kernel/misc_64.S linux-4.1.13/arch/powerpc/kernel/misc_64.S +--- linux-4.1.13.orig/arch/powerpc/kernel/misc_64.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kernel/misc_64.S 2015-11-29 09:23:09.517620297 +0100 +@@ -29,6 +29,7 @@ + + .text + ++#ifndef CONFIG_PREEMPT_RT_FULL + _GLOBAL(call_do_softirq) + mflr r0 + std r0,16(r1) +@@ -39,6 +40,7 @@ + ld r0,16(r1) + mtlr r0 + blr ++#endif + + _GLOBAL(call_do_irq) + mflr r0 +diff -Nur linux-4.1.13.orig/arch/powerpc/kvm/book3s_hv.c linux-4.1.13/arch/powerpc/kvm/book3s_hv.c +--- linux-4.1.13.orig/arch/powerpc/kvm/book3s_hv.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kvm/book3s_hv.c 2015-11-29 09:23:09.517620297 +0100 +@@ -115,11 +115,11 @@ + static void kvmppc_fast_vcpu_kick_hv(struct kvm_vcpu *vcpu) + { + int cpu = vcpu->cpu; +- wait_queue_head_t *wqp; ++ struct swait_head *wqp; + + wqp = kvm_arch_vcpu_wq(vcpu); +- if (waitqueue_active(wqp)) { +- wake_up_interruptible(wqp); ++ if (swaitqueue_active(wqp)) { ++ swait_wake_interruptible(wqp); + ++vcpu->stat.halt_wakeup; + } + +@@ -686,8 +686,8 @@ + tvcpu->arch.prodded = 1; + smp_mb(); + if (vcpu->arch.ceded) { +- if (waitqueue_active(&vcpu->wq)) { +- wake_up_interruptible(&vcpu->wq); ++ if (swaitqueue_active(&vcpu->wq)) { ++ swait_wake_interruptible(&vcpu->wq); + vcpu->stat.halt_wakeup++; + } + } +@@ -1426,7 +1426,7 @@ + INIT_LIST_HEAD(&vcore->runnable_threads); + spin_lock_init(&vcore->lock); + spin_lock_init(&vcore->stoltb_lock); +- init_waitqueue_head(&vcore->wq); ++ init_swait_head(&vcore->wq); + vcore->preempt_tb = TB_NIL; + vcore->lpcr = kvm->arch.lpcr; + vcore->first_vcpuid = core * threads_per_subcore; +@@ -2073,10 +2073,9 @@ + { + struct kvm_vcpu *vcpu; + int do_sleep = 1; ++ DEFINE_SWAITER(wait); + +- DEFINE_WAIT(wait); +- +- prepare_to_wait(&vc->wq, &wait, TASK_INTERRUPTIBLE); ++ swait_prepare(&vc->wq, &wait, TASK_INTERRUPTIBLE); + + /* + * Check one last time for pending exceptions and ceded state after +@@ -2090,7 +2089,7 @@ + } + + if (!do_sleep) { +- finish_wait(&vc->wq, &wait); ++ swait_finish(&vc->wq, &wait); + return; + } + +@@ -2098,7 +2097,7 @@ + trace_kvmppc_vcore_blocked(vc, 0); + spin_unlock(&vc->lock); + schedule(); +- finish_wait(&vc->wq, &wait); ++ swait_finish(&vc->wq, &wait); + spin_lock(&vc->lock); + vc->vcore_state = VCORE_INACTIVE; + trace_kvmppc_vcore_blocked(vc, 1); +@@ -2142,7 +2141,7 @@ + kvmppc_start_thread(vcpu); + trace_kvm_guest_enter(vcpu); + } else if (vc->vcore_state == VCORE_SLEEPING) { +- wake_up(&vc->wq); ++ swait_wake(&vc->wq); + } + + } +diff -Nur linux-4.1.13.orig/arch/powerpc/kvm/Kconfig linux-4.1.13/arch/powerpc/kvm/Kconfig +--- linux-4.1.13.orig/arch/powerpc/kvm/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/kvm/Kconfig 2015-11-29 09:23:09.517620297 +0100 +@@ -172,6 +172,7 @@ + config KVM_MPIC + bool "KVM in-kernel MPIC emulation" + depends on KVM && E500 ++ depends on !PREEMPT_RT_FULL + select HAVE_KVM_IRQCHIP + select HAVE_KVM_IRQFD + select HAVE_KVM_IRQ_ROUTING +diff -Nur linux-4.1.13.orig/arch/powerpc/mm/fault.c linux-4.1.13/arch/powerpc/mm/fault.c +--- linux-4.1.13.orig/arch/powerpc/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/mm/fault.c 2015-11-29 09:23:09.517620297 +0100 +@@ -33,13 +33,13 @@ + #include <linux/ratelimit.h> + #include <linux/context_tracking.h> + #include <linux/hugetlb.h> ++#include <linux/uaccess.h> + + #include <asm/firmware.h> + #include <asm/page.h> + #include <asm/pgtable.h> + #include <asm/mmu.h> + #include <asm/mmu_context.h> +-#include <asm/uaccess.h> + #include <asm/tlbflush.h> + #include <asm/siginfo.h> + #include <asm/debug.h> +@@ -272,15 +272,16 @@ + if (!arch_irq_disabled_regs(regs)) + local_irq_enable(); + +- if (in_atomic() || mm == NULL) { ++ if (faulthandler_disabled() || mm == NULL) { + if (!user_mode(regs)) { + rc = SIGSEGV; + goto bail; + } +- /* in_atomic() in user mode is really bad, ++ /* faulthandler_disabled() in user mode is really bad, + as is current->mm == NULL. */ + printk(KERN_EMERG "Page fault in user mode with " +- "in_atomic() = %d mm = %p\n", in_atomic(), mm); ++ "faulthandler_disabled() = %d mm = %p\n", ++ faulthandler_disabled(), mm); + printk(KERN_EMERG "NIP = %lx MSR = %lx\n", + regs->nip, regs->msr); + die("Weird page fault", regs, SIGSEGV); +diff -Nur linux-4.1.13.orig/arch/powerpc/mm/highmem.c linux-4.1.13/arch/powerpc/mm/highmem.c +--- linux-4.1.13.orig/arch/powerpc/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/mm/highmem.c 2015-11-29 09:23:09.517620297 +0100 +@@ -34,7 +34,7 @@ + unsigned long vaddr; + int idx, type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -59,6 +59,7 @@ + + if (vaddr < __fix_to_virt(FIX_KMAP_END)) { + pagefault_enable(); ++ preempt_enable(); + return; + } + +@@ -82,5 +83,6 @@ + + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); +diff -Nur linux-4.1.13.orig/arch/powerpc/platforms/ps3/device-init.c linux-4.1.13/arch/powerpc/platforms/ps3/device-init.c +--- linux-4.1.13.orig/arch/powerpc/platforms/ps3/device-init.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/powerpc/platforms/ps3/device-init.c 2015-11-29 09:23:09.521620031 +0100 +@@ -752,7 +752,7 @@ + } + pr_debug("%s:%u: notification %s issued\n", __func__, __LINE__, op); + +- res = wait_event_interruptible(dev->done.wait, ++ res = swait_event_interruptible(dev->done.wait, + dev->done.done || kthread_should_stop()); + if (kthread_should_stop()) + res = -EINTR; +diff -Nur linux-4.1.13.orig/arch/s390/include/asm/kvm_host.h linux-4.1.13/arch/s390/include/asm/kvm_host.h +--- linux-4.1.13.orig/arch/s390/include/asm/kvm_host.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/s390/include/asm/kvm_host.h 2015-11-29 09:23:09.521620031 +0100 +@@ -419,7 +419,7 @@ + struct kvm_s390_local_interrupt { + spinlock_t lock; + struct kvm_s390_float_interrupt *float_int; +- wait_queue_head_t *wq; ++ struct swait_head *wq; + atomic_t *cpuflags; + DECLARE_BITMAP(sigp_emerg_pending, KVM_MAX_VCPUS); + struct kvm_s390_irq_payload irq; +diff -Nur linux-4.1.13.orig/arch/s390/include/asm/uaccess.h linux-4.1.13/arch/s390/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/s390/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/s390/include/asm/uaccess.h 2015-11-29 09:23:09.521620031 +0100 +@@ -98,7 +98,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -118,7 +119,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -264,7 +266,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. + * +@@ -290,7 +293,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. + * +@@ -348,7 +352,8 @@ + * strlen_user: - Get the size of a string in user space. + * @str: The string to measure. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +diff -Nur linux-4.1.13.orig/arch/s390/kvm/interrupt.c linux-4.1.13/arch/s390/kvm/interrupt.c +--- linux-4.1.13.orig/arch/s390/kvm/interrupt.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/s390/kvm/interrupt.c 2015-11-29 09:23:09.521620031 +0100 +@@ -875,13 +875,13 @@ + + void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu) + { +- if (waitqueue_active(&vcpu->wq)) { ++ if (swaitqueue_active(&vcpu->wq)) { + /* + * The vcpu gave up the cpu voluntarily, mark it as a good + * yield-candidate. + */ + vcpu->preempted = true; +- wake_up_interruptible(&vcpu->wq); ++ swait_wake_interruptible(&vcpu->wq); + vcpu->stat.halt_wakeup++; + } + } +@@ -987,7 +987,7 @@ + spin_lock(&li->lock); + irq.u.pgm.code = code; + __inject_prog(vcpu, &irq); +- BUG_ON(waitqueue_active(li->wq)); ++ BUG_ON(swaitqueue_active(li->wq)); + spin_unlock(&li->lock); + return 0; + } +@@ -1006,7 +1006,7 @@ + spin_lock(&li->lock); + irq.u.pgm = *pgm_info; + rc = __inject_prog(vcpu, &irq); +- BUG_ON(waitqueue_active(li->wq)); ++ BUG_ON(swaitqueue_active(li->wq)); + spin_unlock(&li->lock); + return rc; + } +diff -Nur linux-4.1.13.orig/arch/s390/mm/fault.c linux-4.1.13/arch/s390/mm/fault.c +--- linux-4.1.13.orig/arch/s390/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/s390/mm/fault.c 2015-11-29 09:23:09.521620031 +0100 +@@ -399,7 +399,7 @@ + * user context. + */ + fault = VM_FAULT_BADCONTEXT; +- if (unlikely(!user_space_fault(regs) || in_atomic() || !mm)) ++ if (unlikely(!user_space_fault(regs) || faulthandler_disabled() || !mm)) + goto out; + + address = trans_exc_code & __FAIL_ADDR_MASK; +diff -Nur linux-4.1.13.orig/arch/score/include/asm/uaccess.h linux-4.1.13/arch/score/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/score/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/score/include/asm/uaccess.h 2015-11-29 09:23:09.521620031 +0100 +@@ -36,7 +36,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +@@ -61,7 +62,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -79,7 +81,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -98,7 +101,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -119,7 +123,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +diff -Nur linux-4.1.13.orig/arch/score/mm/fault.c linux-4.1.13/arch/score/mm/fault.c +--- linux-4.1.13.orig/arch/score/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/score/mm/fault.c 2015-11-29 09:23:09.521620031 +0100 +@@ -34,6 +34,7 @@ + #include <linux/string.h> + #include <linux/types.h> + #include <linux/ptrace.h> ++#include <linux/uaccess.h> + + /* + * This routine handles page faults. It determines the address, +@@ -73,7 +74,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (pagefault_disabled() || !mm) + goto bad_area_nosemaphore; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/sh/kernel/irq.c linux-4.1.13/arch/sh/kernel/irq.c +--- linux-4.1.13.orig/arch/sh/kernel/irq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sh/kernel/irq.c 2015-11-29 09:23:09.521620031 +0100 +@@ -147,6 +147,7 @@ + hardirq_ctx[cpu] = NULL; + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct thread_info *curctx; +@@ -174,6 +175,7 @@ + "r5", "r6", "r7", "r8", "r9", "r15", "t", "pr" + ); + } ++#endif + #else + static inline void handle_one_irq(unsigned int irq) + { +diff -Nur linux-4.1.13.orig/arch/sh/mm/fault.c linux-4.1.13/arch/sh/mm/fault.c +--- linux-4.1.13.orig/arch/sh/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sh/mm/fault.c 2015-11-29 09:23:09.521620031 +0100 +@@ -17,6 +17,7 @@ + #include <linux/kprobes.h> + #include <linux/perf_event.h> + #include <linux/kdebug.h> ++#include <linux/uaccess.h> + #include <asm/io_trapped.h> + #include <asm/mmu_context.h> + #include <asm/tlbflush.h> +@@ -438,9 +439,9 @@ + + /* + * If we're in an interrupt, have no user context or are running +- * in an atomic region then we must not take the fault: ++ * with pagefaults disabled then we must not take the fault: + */ +- if (unlikely(in_atomic() || !mm)) { ++ if (unlikely(faulthandler_disabled() || !mm)) { + bad_area_nosemaphore(regs, error_code, address); + return; + } +diff -Nur linux-4.1.13.orig/arch/sparc/Kconfig linux-4.1.13/arch/sparc/Kconfig +--- linux-4.1.13.orig/arch/sparc/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/Kconfig 2015-11-29 09:23:09.521620031 +0100 +@@ -189,12 +189,10 @@ + source kernel/Kconfig.hz + + config RWSEM_GENERIC_SPINLOCK +- bool +- default y if SPARC32 ++ def_bool PREEMPT_RT_FULL + + config RWSEM_XCHGADD_ALGORITHM +- bool +- default y if SPARC64 ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_HWEIGHT + bool +diff -Nur linux-4.1.13.orig/arch/sparc/kernel/irq_64.c linux-4.1.13/arch/sparc/kernel/irq_64.c +--- linux-4.1.13.orig/arch/sparc/kernel/irq_64.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/kernel/irq_64.c 2015-11-29 09:23:09.521620031 +0100 +@@ -849,6 +849,7 @@ + set_irq_regs(old_regs); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + void *orig_sp, *sp = softirq_stack[smp_processor_id()]; +@@ -863,6 +864,7 @@ + __asm__ __volatile__("mov %0, %%sp" + : : "r" (orig_sp)); + } ++#endif + + #ifdef CONFIG_HOTPLUG_CPU + void fixup_irqs(void) +diff -Nur linux-4.1.13.orig/arch/sparc/mm/fault_32.c linux-4.1.13/arch/sparc/mm/fault_32.c +--- linux-4.1.13.orig/arch/sparc/mm/fault_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/mm/fault_32.c 2015-11-29 09:23:09.521620031 +0100 +@@ -21,6 +21,7 @@ + #include <linux/perf_event.h> + #include <linux/interrupt.h> + #include <linux/kdebug.h> ++#include <linux/uaccess.h> + + #include <asm/page.h> + #include <asm/pgtable.h> +@@ -29,7 +30,6 @@ + #include <asm/setup.h> + #include <asm/smp.h> + #include <asm/traps.h> +-#include <asm/uaccess.h> + + #include "mm_32.h" + +@@ -196,7 +196,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (pagefault_disabled() || !mm) + goto no_context; + + perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address); +diff -Nur linux-4.1.13.orig/arch/sparc/mm/fault_64.c linux-4.1.13/arch/sparc/mm/fault_64.c +--- linux-4.1.13.orig/arch/sparc/mm/fault_64.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/mm/fault_64.c 2015-11-29 09:23:09.521620031 +0100 +@@ -22,12 +22,12 @@ + #include <linux/kdebug.h> + #include <linux/percpu.h> + #include <linux/context_tracking.h> ++#include <linux/uaccess.h> + + #include <asm/page.h> + #include <asm/pgtable.h> + #include <asm/openprom.h> + #include <asm/oplib.h> +-#include <asm/uaccess.h> + #include <asm/asi.h> + #include <asm/lsu.h> + #include <asm/sections.h> +@@ -330,7 +330,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto intr_or_no_mm; + + perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address); +diff -Nur linux-4.1.13.orig/arch/sparc/mm/highmem.c linux-4.1.13/arch/sparc/mm/highmem.c +--- linux-4.1.13.orig/arch/sparc/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/mm/highmem.c 2015-11-29 09:23:09.521620031 +0100 +@@ -53,7 +53,7 @@ + unsigned long vaddr; + long idx, type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -91,6 +91,7 @@ + + if (vaddr < FIXADDR_START) { // FIXME + pagefault_enable(); ++ preempt_enable(); + return; + } + +@@ -126,5 +127,6 @@ + + kmap_atomic_idx_pop(); + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); +diff -Nur linux-4.1.13.orig/arch/sparc/mm/init_64.c linux-4.1.13/arch/sparc/mm/init_64.c +--- linux-4.1.13.orig/arch/sparc/mm/init_64.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/sparc/mm/init_64.c 2015-11-29 09:23:09.521620031 +0100 +@@ -2738,7 +2738,7 @@ + struct mm_struct *mm = current->mm; + struct tsb_config *tp; + +- if (in_atomic() || !mm) { ++ if (faulthandler_disabled() || !mm) { + const struct exception_table_entry *entry; + + entry = search_exception_tables(regs->tpc); +diff -Nur linux-4.1.13.orig/arch/tile/include/asm/uaccess.h linux-4.1.13/arch/tile/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/tile/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/tile/include/asm/uaccess.h 2015-11-29 09:23:09.521620031 +0100 +@@ -78,7 +78,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +@@ -192,7 +193,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -274,7 +276,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -330,7 +333,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -366,7 +370,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -437,7 +442,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to user space. Caller must check + * the specified blocks with access_ok() before calling this function. +diff -Nur linux-4.1.13.orig/arch/tile/mm/fault.c linux-4.1.13/arch/tile/mm/fault.c +--- linux-4.1.13.orig/arch/tile/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/tile/mm/fault.c 2015-11-29 09:23:09.521620031 +0100 +@@ -354,9 +354,9 @@ + + /* + * If we're in an interrupt, have no user context or are running in an +- * atomic region then we must not take the fault. ++ * region with pagefaults disabled then we must not take the fault. + */ +- if (in_atomic() || !mm) { ++ if (pagefault_disabled() || !mm) { + vma = NULL; /* happy compiler */ + goto bad_area_nosemaphore; + } +diff -Nur linux-4.1.13.orig/arch/tile/mm/highmem.c linux-4.1.13/arch/tile/mm/highmem.c +--- linux-4.1.13.orig/arch/tile/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/tile/mm/highmem.c 2015-11-29 09:23:09.521620031 +0100 +@@ -201,7 +201,7 @@ + int idx, type; + pte_t *pte; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable(); + pagefault_disable(); + + /* Avoid icache flushes by disallowing atomic executable mappings. */ +@@ -259,6 +259,7 @@ + } + + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +diff -Nur linux-4.1.13.orig/arch/um/kernel/trap.c linux-4.1.13/arch/um/kernel/trap.c +--- linux-4.1.13.orig/arch/um/kernel/trap.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/um/kernel/trap.c 2015-11-29 09:23:09.521620031 +0100 +@@ -35,10 +35,10 @@ + *code_out = SEGV_MAPERR; + + /* +- * If the fault was during atomic operation, don't take the fault, just ++ * If the fault was with pagefaults disabled, don't take the fault, just + * fail. + */ +- if (in_atomic()) ++ if (faulthandler_disabled()) + goto out_nosemaphore; + + if (is_user) +diff -Nur linux-4.1.13.orig/arch/unicore32/mm/fault.c linux-4.1.13/arch/unicore32/mm/fault.c +--- linux-4.1.13.orig/arch/unicore32/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/unicore32/mm/fault.c 2015-11-29 09:23:09.525619763 +0100 +@@ -218,7 +218,7 @@ + * If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) ++ if (faulthandler_disabled() || !mm) + goto no_context; + + if (user_mode(regs)) +diff -Nur linux-4.1.13.orig/arch/x86/crypto/aesni-intel_glue.c linux-4.1.13/arch/x86/crypto/aesni-intel_glue.c +--- linux-4.1.13.orig/arch/x86/crypto/aesni-intel_glue.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/crypto/aesni-intel_glue.c 2015-11-29 09:23:09.525619763 +0100 +@@ -382,14 +382,14 @@ + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, +- nbytes & AES_BLOCK_MASK); ++ nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -406,14 +406,14 @@ + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -430,14 +430,14 @@ + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -454,14 +454,14 @@ + err = blkcipher_walk_virt(desc, &walk); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes)) { ++ kernel_fpu_begin(); + aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- kernel_fpu_end(); + + return err; + } +@@ -513,18 +513,20 @@ + err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + +- kernel_fpu_begin(); + while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { ++ kernel_fpu_begin(); + aesni_ctr_enc_tfm(ctx, walk.dst.virt.addr, walk.src.virt.addr, + nbytes & AES_BLOCK_MASK, walk.iv); ++ kernel_fpu_end(); + nbytes &= AES_BLOCK_SIZE - 1; + err = blkcipher_walk_done(desc, &walk, nbytes); + } + if (walk.nbytes) { ++ kernel_fpu_begin(); + ctr_crypt_final(ctx, &walk); ++ kernel_fpu_end(); + err = blkcipher_walk_done(desc, &walk, 0); + } +- kernel_fpu_end(); + + return err; + } +diff -Nur linux-4.1.13.orig/arch/x86/crypto/cast5_avx_glue.c linux-4.1.13/arch/x86/crypto/cast5_avx_glue.c +--- linux-4.1.13.orig/arch/x86/crypto/cast5_avx_glue.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/crypto/cast5_avx_glue.c 2015-11-29 09:23:09.525619763 +0100 +@@ -60,7 +60,7 @@ + static int ecb_crypt(struct blkcipher_desc *desc, struct blkcipher_walk *walk, + bool enc) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct cast5_ctx *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = CAST5_BLOCK_SIZE; + unsigned int nbytes; +@@ -76,7 +76,7 @@ + u8 *wsrc = walk->src.virt.addr; + u8 *wdst = walk->dst.virt.addr; + +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + + /* Process multi-block batch */ + if (nbytes >= bsize * CAST5_PARALLEL_BLOCKS) { +@@ -104,10 +104,9 @@ + } while (nbytes >= bsize); + + done: ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; + } + +@@ -228,7 +227,7 @@ + static int cbc_decrypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -237,12 +236,11 @@ + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + + while ((nbytes = walk.nbytes)) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __cbc_decrypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } +- +- cast5_fpu_end(fpu_enabled); + return err; + } + +@@ -312,7 +310,7 @@ + static int ctr_crypt(struct blkcipher_desc *desc, struct scatterlist *dst, + struct scatterlist *src, unsigned int nbytes) + { +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -321,13 +319,12 @@ + desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP; + + while ((nbytes = walk.nbytes) >= CAST5_BLOCK_SIZE) { +- fpu_enabled = cast5_fpu_begin(fpu_enabled, nbytes); ++ fpu_enabled = cast5_fpu_begin(false, nbytes); + nbytes = __ctr_crypt(desc, &walk); ++ cast5_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- cast5_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + ctr_crypt_final(desc, &walk); + err = blkcipher_walk_done(desc, &walk, 0); +diff -Nur linux-4.1.13.orig/arch/x86/crypto/glue_helper.c linux-4.1.13/arch/x86/crypto/glue_helper.c +--- linux-4.1.13.orig/arch/x86/crypto/glue_helper.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/crypto/glue_helper.c 2015-11-29 09:23:09.525619763 +0100 +@@ -39,7 +39,7 @@ + void *ctx = crypto_blkcipher_ctx(desc->tfm); + const unsigned int bsize = 128 / 8; + unsigned int nbytes, i, func_bytes; +- bool fpu_enabled = false; ++ bool fpu_enabled; + int err; + + err = blkcipher_walk_virt(desc, walk); +@@ -49,7 +49,7 @@ + u8 *wdst = walk->dst.virt.addr; + + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + + for (i = 0; i < gctx->num_funcs; i++) { + func_bytes = bsize * gctx->funcs[i].num_blocks; +@@ -71,10 +71,10 @@ + } + + done: ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); + return err; + } + +@@ -194,7 +194,7 @@ + struct scatterlist *src, unsigned int nbytes) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -203,12 +203,12 @@ + + while ((nbytes = walk.nbytes)) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_cbc_decrypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); + return err; + } + EXPORT_SYMBOL_GPL(glue_cbc_decrypt_128bit); +@@ -277,7 +277,7 @@ + struct scatterlist *src, unsigned int nbytes) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -286,13 +286,12 @@ + + while ((nbytes = walk.nbytes) >= bsize) { + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, nbytes); ++ desc, false, nbytes); + nbytes = __glue_ctr_crypt_128bit(gctx, desc, &walk); ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + } + +- glue_fpu_end(fpu_enabled); +- + if (walk.nbytes) { + glue_ctr_crypt_final_128bit( + gctx->funcs[gctx->num_funcs - 1].fn_u.ctr, desc, &walk); +@@ -347,7 +346,7 @@ + void *tweak_ctx, void *crypt_ctx) + { + const unsigned int bsize = 128 / 8; +- bool fpu_enabled = false; ++ bool fpu_enabled; + struct blkcipher_walk walk; + int err; + +@@ -360,21 +359,21 @@ + + /* set minimum length to bsize, for tweak_fn */ + fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, +- desc, fpu_enabled, ++ desc, false, + nbytes < bsize ? bsize : nbytes); +- + /* calculate first value of T */ + tweak_fn(tweak_ctx, walk.iv, walk.iv); ++ glue_fpu_end(fpu_enabled); + + while (nbytes) { ++ fpu_enabled = glue_fpu_begin(bsize, gctx->fpu_blocks_limit, ++ desc, false, nbytes); + nbytes = __glue_xts_crypt_128bit(gctx, crypt_ctx, desc, &walk); + ++ glue_fpu_end(fpu_enabled); + err = blkcipher_walk_done(desc, &walk, nbytes); + nbytes = walk.nbytes; + } +- +- glue_fpu_end(fpu_enabled); +- + return err; + } + EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit); +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/preempt.h linux-4.1.13/arch/x86/include/asm/preempt.h +--- linux-4.1.13.orig/arch/x86/include/asm/preempt.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/preempt.h 2015-11-29 09:23:09.525619763 +0100 +@@ -82,17 +82,33 @@ + * a decrement which hits zero means we have no preempt_count and should + * reschedule. + */ +-static __always_inline bool __preempt_count_dec_and_test(void) ++static __always_inline bool ____preempt_count_dec_and_test(void) + { + GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), "e"); + } + ++static __always_inline bool __preempt_count_dec_and_test(void) ++{ ++ if (____preempt_count_dec_and_test()) ++ return true; ++#ifdef CONFIG_PREEMPT_LAZY ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++#else ++ return false; ++#endif ++} ++ + /* + * Returns true when we need to resched and can (barring IRQ state). + */ + static __always_inline bool should_resched(int preempt_offset) + { ++#ifdef CONFIG_PREEMPT_LAZY ++ return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset || ++ test_thread_flag(TIF_NEED_RESCHED_LAZY)); ++#else + return unlikely(raw_cpu_read_4(__preempt_count) == preempt_offset); ++#endif + } + + #ifdef CONFIG_PREEMPT +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/signal.h linux-4.1.13/arch/x86/include/asm/signal.h +--- linux-4.1.13.orig/arch/x86/include/asm/signal.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/signal.h 2015-11-29 09:23:09.525619763 +0100 +@@ -23,6 +23,19 @@ + unsigned long sig[_NSIG_WORDS]; + } sigset_t; + ++/* ++ * Because some traps use the IST stack, we must keep preemption ++ * disabled while calling do_trap(), but do_trap() may call ++ * force_sig_info() which will grab the signal spin_locks for the ++ * task, which in PREEMPT_RT_FULL are mutexes. By defining ++ * ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will set ++ * TIF_NOTIFY_RESUME and set up the signal to be sent on exit of the ++ * trap. ++ */ ++#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_X86_64) ++#define ARCH_RT_DELAYS_SIGNAL_SEND ++#endif ++ + #ifndef CONFIG_COMPAT + typedef sigset_t compat_sigset_t; + #endif +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/stackprotector.h linux-4.1.13/arch/x86/include/asm/stackprotector.h +--- linux-4.1.13.orig/arch/x86/include/asm/stackprotector.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/stackprotector.h 2015-11-29 09:23:09.525619763 +0100 +@@ -57,7 +57,7 @@ + */ + static __always_inline void boot_init_stack_canary(void) + { +- u64 canary; ++ u64 uninitialized_var(canary); + u64 tsc; + + #ifdef CONFIG_X86_64 +@@ -68,8 +68,16 @@ + * of randomness. The TSC only matters for very early init, + * there it already has some randomness on most systems. Later + * on during the bootup the random pool has true entropy too. ++ * ++ * For preempt-rt we need to weaken the randomness a bit, as ++ * we can't call into the random generator from atomic context ++ * due to locking constraints. We just leave canary ++ * uninitialized and use the TSC based randomness on top of ++ * it. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + get_random_bytes(&canary, sizeof(canary)); ++#endif + tsc = __native_read_tsc(); + canary += tsc + (tsc << 32UL); + +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/thread_info.h linux-4.1.13/arch/x86/include/asm/thread_info.h +--- linux-4.1.13.orig/arch/x86/include/asm/thread_info.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/thread_info.h 2015-11-29 09:23:09.525619763 +0100 +@@ -55,6 +55,8 @@ + __u32 status; /* thread synchronous flags */ + __u32 cpu; /* current CPU */ + int saved_preempt_count; ++ int preempt_lazy_count; /* 0 => lazy preemptable ++ <0 => BUG */ + mm_segment_t addr_limit; + void __user *sysenter_return; + unsigned int sig_on_uaccess_error:1; +@@ -95,6 +97,7 @@ + #define TIF_SYSCALL_EMU 6 /* syscall emulation active */ + #define TIF_SYSCALL_AUDIT 7 /* syscall auditing active */ + #define TIF_SECCOMP 8 /* secure computing */ ++#define TIF_NEED_RESCHED_LAZY 9 /* lazy rescheduling necessary */ + #define TIF_USER_RETURN_NOTIFY 11 /* notify kernel of userspace return */ + #define TIF_UPROBE 12 /* breakpointed or singlestepping */ + #define TIF_NOTSC 16 /* TSC is not accessible in userland */ +@@ -119,6 +122,7 @@ + #define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) + #define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) + #define _TIF_SECCOMP (1 << TIF_SECCOMP) ++#define _TIF_NEED_RESCHED_LAZY (1 << TIF_NEED_RESCHED_LAZY) + #define _TIF_USER_RETURN_NOTIFY (1 << TIF_USER_RETURN_NOTIFY) + #define _TIF_UPROBE (1 << TIF_UPROBE) + #define _TIF_NOTSC (1 << TIF_NOTSC) +@@ -168,6 +172,8 @@ + #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) + #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW) + ++#define _TIF_NEED_RESCHED_MASK (_TIF_NEED_RESCHED | _TIF_NEED_RESCHED_LAZY) ++ + #define STACK_WARN (THREAD_SIZE/8) + + /* +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/uaccess_32.h linux-4.1.13/arch/x86/include/asm/uaccess_32.h +--- linux-4.1.13.orig/arch/x86/include/asm/uaccess_32.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/uaccess_32.h 2015-11-29 09:23:09.525619763 +0100 +@@ -70,7 +70,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. Caller must check + * the specified block with access_ok() before calling this function. +@@ -117,7 +118,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. Caller must check + * the specified block with access_ok() before calling this function. +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/uaccess.h linux-4.1.13/arch/x86/include/asm/uaccess.h +--- linux-4.1.13.orig/arch/x86/include/asm/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/uaccess.h 2015-11-29 09:23:09.525619763 +0100 +@@ -74,7 +74,8 @@ + * @addr: User space pointer to start of block to check + * @size: Size of block to check + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Checks if a pointer to a block of memory in user space is valid. + * +@@ -145,7 +146,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -240,7 +242,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +@@ -455,7 +458,8 @@ + * @x: Variable to store result. + * @ptr: Source address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple variable from user space to kernel + * space. It supports simple types like char and int, but not larger +@@ -479,7 +483,8 @@ + * @x: Value to copy to user space. + * @ptr: Destination address, in user space. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * This macro copies a single simple value from kernel space to user + * space. It supports simple types like char and int, but not larger +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/uv/uv_bau.h linux-4.1.13/arch/x86/include/asm/uv/uv_bau.h +--- linux-4.1.13.orig/arch/x86/include/asm/uv/uv_bau.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/uv/uv_bau.h 2015-11-29 09:23:09.525619763 +0100 +@@ -615,9 +615,9 @@ + cycles_t send_message; + cycles_t period_end; + cycles_t period_time; +- spinlock_t uvhub_lock; +- spinlock_t queue_lock; +- spinlock_t disable_lock; ++ raw_spinlock_t uvhub_lock; ++ raw_spinlock_t queue_lock; ++ raw_spinlock_t disable_lock; + /* tunables */ + int max_concurr; + int max_concurr_const; +@@ -776,15 +776,15 @@ + * to be lowered below the current 'v'. atomic_add_unless can only stop + * on equal. + */ +-static inline int atomic_inc_unless_ge(spinlock_t *lock, atomic_t *v, int u) ++static inline int atomic_inc_unless_ge(raw_spinlock_t *lock, atomic_t *v, int u) + { +- spin_lock(lock); ++ raw_spin_lock(lock); + if (atomic_read(v) >= u) { +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 0; + } + atomic_inc(v); +- spin_unlock(lock); ++ raw_spin_unlock(lock); + return 1; + } + +diff -Nur linux-4.1.13.orig/arch/x86/include/asm/uv/uv_hub.h linux-4.1.13/arch/x86/include/asm/uv/uv_hub.h +--- linux-4.1.13.orig/arch/x86/include/asm/uv/uv_hub.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/include/asm/uv/uv_hub.h 2015-11-29 09:23:09.525619763 +0100 +@@ -492,7 +492,7 @@ + unsigned short nr_online_cpus; + unsigned short pnode; + short memory_nid; +- spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ ++ raw_spinlock_t nmi_lock; /* obsolete, see uv_hub_nmi */ + unsigned long nmi_count; /* obsolete, see uv_hub_nmi */ + }; + extern struct uv_blade_info *uv_blade_info; +diff -Nur linux-4.1.13.orig/arch/x86/Kconfig linux-4.1.13/arch/x86/Kconfig +--- linux-4.1.13.orig/arch/x86/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/Kconfig 2015-11-29 09:23:09.525619763 +0100 +@@ -22,6 +22,7 @@ + ### Arch settings + config X86 + def_bool y ++ select HAVE_PREEMPT_LAZY + select ACPI_SYSTEM_POWER_STATES_SUPPORT if ACPI + select ARCH_MIGHT_HAVE_ACPI_PDC if ACPI + select ARCH_HAS_DEBUG_STRICT_USER_COPY_CHECKS +@@ -203,8 +204,11 @@ + def_bool y + depends on ISA_DMA_API + ++config RWSEM_GENERIC_SPINLOCK ++ def_bool PREEMPT_RT_FULL ++ + config RWSEM_XCHGADD_ALGORITHM +- def_bool y ++ def_bool !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL + + config GENERIC_CALIBRATE_DELAY + def_bool y +@@ -838,7 +842,7 @@ + config MAXSMP + bool "Enable Maximum number of SMP Processors and NUMA Nodes" + depends on X86_64 && SMP && DEBUG_KERNEL +- select CPUMASK_OFFSTACK ++ select CPUMASK_OFFSTACK if !PREEMPT_RT_FULL + ---help--- + Enable maximum number of CPUS and NUMA Nodes for this architecture. + If unsure, say N. +diff -Nur linux-4.1.13.orig/arch/x86/kernel/apic/io_apic.c linux-4.1.13/arch/x86/kernel/apic/io_apic.c +--- linux-4.1.13.orig/arch/x86/kernel/apic/io_apic.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/apic/io_apic.c 2015-11-29 09:23:09.525619763 +0100 +@@ -1891,7 +1891,8 @@ + static inline bool ioapic_irqd_mask(struct irq_data *data, struct irq_cfg *cfg) + { + /* If we are moving the irq we need to mask it */ +- if (unlikely(irqd_is_setaffinity_pending(data))) { ++ if (unlikely(irqd_is_setaffinity_pending(data) && ++ !irqd_irq_inprogress(data))) { + mask_ioapic(cfg); + return true; + } +diff -Nur linux-4.1.13.orig/arch/x86/kernel/apic/x2apic_uv_x.c linux-4.1.13/arch/x86/kernel/apic/x2apic_uv_x.c +--- linux-4.1.13.orig/arch/x86/kernel/apic/x2apic_uv_x.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/apic/x2apic_uv_x.c 2015-11-29 09:23:09.525619763 +0100 +@@ -949,7 +949,7 @@ + uv_blade_info[blade].pnode = pnode; + uv_blade_info[blade].nr_possible_cpus = 0; + uv_blade_info[blade].nr_online_cpus = 0; +- spin_lock_init(&uv_blade_info[blade].nmi_lock); ++ raw_spin_lock_init(&uv_blade_info[blade].nmi_lock); + min_pnode = min(pnode, min_pnode); + max_pnode = max(pnode, max_pnode); + blade++; +diff -Nur linux-4.1.13.orig/arch/x86/kernel/asm-offsets.c linux-4.1.13/arch/x86/kernel/asm-offsets.c +--- linux-4.1.13.orig/arch/x86/kernel/asm-offsets.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/asm-offsets.c 2015-11-29 09:23:09.525619763 +0100 +@@ -32,6 +32,7 @@ + OFFSET(TI_flags, thread_info, flags); + OFFSET(TI_status, thread_info, status); + OFFSET(TI_addr_limit, thread_info, addr_limit); ++ OFFSET(TI_preempt_lazy_count, thread_info, preempt_lazy_count); + + BLANK(); + OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); +@@ -71,4 +72,5 @@ + + BLANK(); + DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); ++ DEFINE(_PREEMPT_ENABLED, PREEMPT_ENABLED); + } +diff -Nur linux-4.1.13.orig/arch/x86/kernel/cpu/mcheck/mce.c linux-4.1.13/arch/x86/kernel/cpu/mcheck/mce.c +--- linux-4.1.13.orig/arch/x86/kernel/cpu/mcheck/mce.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/cpu/mcheck/mce.c 2015-11-29 09:23:09.529619496 +0100 +@@ -41,6 +41,8 @@ + #include <linux/debugfs.h> + #include <linux/irq_work.h> + #include <linux/export.h> ++#include <linux/jiffies.h> ++#include <linux/work-simple.h> + + #include <asm/processor.h> + #include <asm/traps.h> +@@ -1267,7 +1269,7 @@ + static unsigned long check_interval = INITIAL_CHECK_INTERVAL; + + static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ +-static DEFINE_PER_CPU(struct timer_list, mce_timer); ++static DEFINE_PER_CPU(struct hrtimer, mce_timer); + + static unsigned long mce_adjust_timer_default(unsigned long interval) + { +@@ -1276,32 +1278,18 @@ + + static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; + +-static void __restart_timer(struct timer_list *t, unsigned long interval) ++static enum hrtimer_restart __restart_timer(struct hrtimer *timer, unsigned long interval) + { +- unsigned long when = jiffies + interval; +- unsigned long flags; +- +- local_irq_save(flags); +- +- if (timer_pending(t)) { +- if (time_before(when, t->expires)) +- mod_timer_pinned(t, when); +- } else { +- t->expires = round_jiffies(when); +- add_timer_on(t, smp_processor_id()); +- } +- +- local_irq_restore(flags); ++ if (!interval) ++ return HRTIMER_NORESTART; ++ hrtimer_forward_now(timer, ns_to_ktime(jiffies_to_nsecs(interval))); ++ return HRTIMER_RESTART; + } + +-static void mce_timer_fn(unsigned long data) ++static enum hrtimer_restart mce_timer_fn(struct hrtimer *timer) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); +- int cpu = smp_processor_id(); + unsigned long iv; + +- WARN_ON(cpu != data); +- + iv = __this_cpu_read(mce_next_interval); + + if (mce_available(this_cpu_ptr(&cpu_info))) { +@@ -1324,7 +1312,7 @@ + + done: + __this_cpu_write(mce_next_interval, iv); +- __restart_timer(t, iv); ++ return __restart_timer(timer, iv); + } + + /* +@@ -1332,7 +1320,7 @@ + */ + void mce_timer_kick(unsigned long interval) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned long iv = __this_cpu_read(mce_next_interval); + + __restart_timer(t, interval); +@@ -1347,7 +1335,7 @@ + int cpu; + + for_each_online_cpu(cpu) +- del_timer_sync(&per_cpu(mce_timer, cpu)); ++ hrtimer_cancel(&per_cpu(mce_timer, cpu)); + } + + static void mce_do_trigger(struct work_struct *work) +@@ -1357,6 +1345,56 @@ + + static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + ++static void __mce_notify_work(struct swork_event *event) ++{ ++ /* Not more than two messages every minute */ ++ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); ++ ++ /* wake processes polling /dev/mcelog */ ++ wake_up_interruptible(&mce_chrdev_wait); ++ ++ /* ++ * There is no risk of missing notifications because ++ * work_pending is always cleared before the function is ++ * executed. ++ */ ++ if (mce_helper[0] && !work_pending(&mce_trigger_work)) ++ schedule_work(&mce_trigger_work); ++ ++ if (__ratelimit(&ratelimit)) ++ pr_info(HW_ERR "Machine check events logged\n"); ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static bool notify_work_ready __read_mostly; ++static struct swork_event notify_work; ++ ++static int mce_notify_work_init(void) ++{ ++ int err; ++ ++ err = swork_get(); ++ if (err) ++ return err; ++ ++ INIT_SWORK(¬ify_work, __mce_notify_work); ++ notify_work_ready = true; ++ return 0; ++} ++ ++static void mce_notify_work(void) ++{ ++ if (notify_work_ready) ++ swork_queue(¬ify_work); ++} ++#else ++static void mce_notify_work(void) ++{ ++ __mce_notify_work(NULL); ++} ++static inline int mce_notify_work_init(void) { return 0; } ++#endif ++ + /* + * Notify the user(s) about new machine check events. + * Can be called from interrupt context, but not from machine check/NMI +@@ -1364,19 +1402,8 @@ + */ + int mce_notify_irq(void) + { +- /* Not more than two messages every minute */ +- static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); +- + if (test_and_clear_bit(0, &mce_need_notify)) { +- /* wake processes polling /dev/mcelog */ +- wake_up_interruptible(&mce_chrdev_wait); +- +- if (mce_helper[0]) +- schedule_work(&mce_trigger_work); +- +- if (__ratelimit(&ratelimit)) +- pr_info(HW_ERR "Machine check events logged\n"); +- ++ mce_notify_work(); + return 1; + } + return 0; +@@ -1649,7 +1676,7 @@ + } + } + +-static void mce_start_timer(unsigned int cpu, struct timer_list *t) ++static void mce_start_timer(unsigned int cpu, struct hrtimer *t) + { + unsigned long iv = check_interval * HZ; + +@@ -1658,16 +1685,17 @@ + + per_cpu(mce_next_interval, cpu) = iv; + +- t->expires = round_jiffies(jiffies + iv); +- add_timer_on(t, cpu); ++ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(iv) * 1000ULL), ++ 0, HRTIMER_MODE_REL_PINNED); + } + + static void __mcheck_cpu_init_timer(void) + { +- struct timer_list *t = this_cpu_ptr(&mce_timer); ++ struct hrtimer *t = this_cpu_ptr(&mce_timer); + unsigned int cpu = smp_processor_id(); + +- setup_timer(t, mce_timer_fn, cpu); ++ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL); ++ t->function = mce_timer_fn; + mce_start_timer(cpu, t); + } + +@@ -2345,6 +2373,8 @@ + if (!mce_available(raw_cpu_ptr(&cpu_info))) + return; + ++ hrtimer_cancel(this_cpu_ptr(&mce_timer)); ++ + if (!(action & CPU_TASKS_FROZEN)) + cmci_clear(); + for (i = 0; i < mca_cfg.banks; i++) { +@@ -2371,6 +2401,7 @@ + if (b->init) + wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl); + } ++ __mcheck_cpu_init_timer(); + } + + /* Get notified when a cpu comes on/off. Be hotplug friendly. */ +@@ -2378,7 +2409,6 @@ + mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) + { + unsigned int cpu = (unsigned long)hcpu; +- struct timer_list *t = &per_cpu(mce_timer, cpu); + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_ONLINE: +@@ -2398,11 +2428,9 @@ + break; + case CPU_DOWN_PREPARE: + smp_call_function_single(cpu, mce_disable_cpu, &action, 1); +- del_timer_sync(t); + break; + case CPU_DOWN_FAILED: + smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); +- mce_start_timer(cpu, t); + break; + } + +@@ -2441,6 +2469,10 @@ + goto err_out; + } + ++ err = mce_notify_work_init(); ++ if (err) ++ goto err_out; ++ + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; +diff -Nur linux-4.1.13.orig/arch/x86/kernel/dumpstack_32.c linux-4.1.13/arch/x86/kernel/dumpstack_32.c +--- linux-4.1.13.orig/arch/x86/kernel/dumpstack_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/dumpstack_32.c 2015-11-29 09:23:09.529619496 +0100 +@@ -42,7 +42,7 @@ + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) + { +- const unsigned cpu = get_cpu(); ++ const unsigned cpu = get_cpu_light(); + int graph = 0; + u32 *prev_esp; + +@@ -86,7 +86,7 @@ + break; + touch_nmi_watchdog(); + } +- put_cpu(); ++ put_cpu_light(); + } + EXPORT_SYMBOL(dump_trace); + +diff -Nur linux-4.1.13.orig/arch/x86/kernel/dumpstack_64.c linux-4.1.13/arch/x86/kernel/dumpstack_64.c +--- linux-4.1.13.orig/arch/x86/kernel/dumpstack_64.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/dumpstack_64.c 2015-11-29 09:23:09.529619496 +0100 +@@ -152,7 +152,7 @@ + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) + { +- const unsigned cpu = get_cpu(); ++ const unsigned cpu = get_cpu_light(); + struct thread_info *tinfo; + unsigned long *irq_stack = (unsigned long *)per_cpu(irq_stack_ptr, cpu); + unsigned long dummy; +@@ -241,7 +241,7 @@ + * This handles the process stack: + */ + bp = ops->walk_stack(tinfo, stack, bp, ops, data, NULL, &graph); +- put_cpu(); ++ put_cpu_light(); + } + EXPORT_SYMBOL(dump_trace); + +@@ -255,7 +255,7 @@ + int cpu; + int i; + +- preempt_disable(); ++ migrate_disable(); + cpu = smp_processor_id(); + + irq_stack_end = (unsigned long *)(per_cpu(irq_stack_ptr, cpu)); +@@ -291,7 +291,7 @@ + pr_cont(" %016lx", *stack++); + touch_nmi_watchdog(); + } +- preempt_enable(); ++ migrate_enable(); + + pr_cont("\n"); + show_trace_log_lvl(task, regs, sp, bp, log_lvl); +diff -Nur linux-4.1.13.orig/arch/x86/kernel/entry_32.S linux-4.1.13/arch/x86/kernel/entry_32.S +--- linux-4.1.13.orig/arch/x86/kernel/entry_32.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/entry_32.S 2015-11-29 09:23:09.529619496 +0100 +@@ -359,8 +359,24 @@ + ENTRY(resume_kernel) + DISABLE_INTERRUPTS(CLBR_ANY) + need_resched: ++ # preempt count == 0 + NEED_RS set? + cmpl $0,PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz restore_all ++#else ++ jz test_int_off ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jne restore_all ++ ++ cmpl $0,TI_preempt_lazy_count(%ebp) # non-zero preempt_lazy_count ? ++ jnz restore_all ++ ++ testl $_TIF_NEED_RESCHED_LAZY, TI_flags(%ebp) ++ jz restore_all ++test_int_off: ++#endif + testl $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off (exception path) ? + jz restore_all + call preempt_schedule_irq +@@ -594,7 +610,7 @@ + ALIGN + RING0_PTREGS_FRAME # can't unwind into user space anyway + work_pending: +- testb $_TIF_NEED_RESCHED, %cl ++ testl $_TIF_NEED_RESCHED_MASK, %ecx + jz work_notifysig + work_resched: + call schedule +@@ -607,7 +623,7 @@ + andl $_TIF_WORK_MASK, %ecx # is there any work to be done other + # than syscall tracing? + jz restore_all +- testb $_TIF_NEED_RESCHED, %cl ++ testl $_TIF_NEED_RESCHED_MASK, %ecx + jnz work_resched + + work_notifysig: # deal with pending signals and +diff -Nur linux-4.1.13.orig/arch/x86/kernel/entry_64.S linux-4.1.13/arch/x86/kernel/entry_64.S +--- linux-4.1.13.orig/arch/x86/kernel/entry_64.S 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/entry_64.S 2015-11-29 09:23:09.529619496 +0100 +@@ -370,8 +370,8 @@ + /* First do a reschedule test. */ + /* edx: work, edi: workmask */ + int_careful: +- bt $TIF_NEED_RESCHED,%edx +- jnc int_very_careful ++ testl $_TIF_NEED_RESCHED_MASK,%edx ++ jz int_very_careful + TRACE_IRQS_ON + ENABLE_INTERRUPTS(CLBR_NONE) + pushq_cfi %rdi +@@ -776,7 +776,23 @@ + bt $9,EFLAGS(%rsp) /* interrupts were off? */ + jnc 1f + 0: cmpl $0,PER_CPU_VAR(__preempt_count) ++#ifndef CONFIG_PREEMPT_LAZY + jnz 1f ++#else ++ jz do_preempt_schedule_irq ++ ++ # atleast preempt count == 0 ? ++ cmpl $_PREEMPT_ENABLED,PER_CPU_VAR(__preempt_count) ++ jnz 1f ++ ++ GET_THREAD_INFO(%rcx) ++ cmpl $0, TI_preempt_lazy_count(%rcx) ++ jnz 1f ++ ++ bt $TIF_NEED_RESCHED_LAZY,TI_flags(%rcx) ++ jnc 1f ++do_preempt_schedule_irq: ++#endif + call preempt_schedule_irq + jmp 0b + 1: +@@ -844,8 +860,8 @@ + /* edi: workmask, edx: work */ + retint_careful: + CFI_RESTORE_STATE +- bt $TIF_NEED_RESCHED,%edx +- jnc retint_signal ++ testl $_TIF_NEED_RESCHED_MASK,%edx ++ jz retint_signal + TRACE_IRQS_ON + ENABLE_INTERRUPTS(CLBR_NONE) + pushq_cfi %rdi +@@ -1118,6 +1134,7 @@ + jmp 2b + .previous + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* Call softirq on interrupt stack. Interrupts are off. */ + ENTRY(do_softirq_own_stack) + CFI_STARTPROC +@@ -1137,6 +1154,7 @@ + ret + CFI_ENDPROC + END(do_softirq_own_stack) ++#endif + + #ifdef CONFIG_XEN + idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 +diff -Nur linux-4.1.13.orig/arch/x86/kernel/irq_32.c linux-4.1.13/arch/x86/kernel/irq_32.c +--- linux-4.1.13.orig/arch/x86/kernel/irq_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/irq_32.c 2015-11-29 09:23:09.529619496 +0100 +@@ -135,6 +135,7 @@ + cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + void do_softirq_own_stack(void) + { + struct thread_info *curstk; +@@ -153,6 +154,7 @@ + + call_on_stack(__do_softirq, isp); + } ++#endif + + bool handle_irq(unsigned irq, struct pt_regs *regs) + { +diff -Nur linux-4.1.13.orig/arch/x86/kernel/process_32.c linux-4.1.13/arch/x86/kernel/process_32.c +--- linux-4.1.13.orig/arch/x86/kernel/process_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/process_32.c 2015-11-29 09:23:09.529619496 +0100 +@@ -35,6 +35,7 @@ + #include <linux/uaccess.h> + #include <linux/io.h> + #include <linux/kdebug.h> ++#include <linux/highmem.h> + + #include <asm/pgtable.h> + #include <asm/ldt.h> +@@ -210,6 +211,35 @@ + } + EXPORT_SYMBOL_GPL(start_thread); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) ++{ ++ int i; ++ ++ /* ++ * Clear @prev's kmap_atomic mappings ++ */ ++ for (i = 0; i < prev_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ pte_t *ptep = kmap_pte - idx; ++ ++ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx)); ++ } ++ /* ++ * Restore @next_p's kmap_atomic mappings ++ */ ++ for (i = 0; i < next_p->kmap_idx; i++) { ++ int idx = i + KM_TYPE_NR * smp_processor_id(); ++ ++ if (!pte_none(next_p->kmap_pte[i])) ++ set_pte(kmap_pte - idx, next_p->kmap_pte[i]); ++ } ++} ++#else ++static inline void ++switch_kmaps(struct task_struct *prev_p, struct task_struct *next_p) { } ++#endif ++ + + /* + * switch_to(x,y) should switch tasks from x to y. +@@ -292,6 +322,8 @@ + task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT)) + __switch_to_xtra(prev_p, next_p, tss); + ++ switch_kmaps(prev_p, next_p); ++ + /* + * Leave lazy mode, flushing any hypercalls made here. + * This must be done before restoring TLS segments so +diff -Nur linux-4.1.13.orig/arch/x86/kernel/signal.c linux-4.1.13/arch/x86/kernel/signal.c +--- linux-4.1.13.orig/arch/x86/kernel/signal.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/signal.c 2015-11-29 09:23:09.529619496 +0100 +@@ -723,6 +723,14 @@ + { + user_exit(); + ++#ifdef ARCH_RT_DELAYS_SIGNAL_SEND ++ if (unlikely(current->forced_info.si_signo)) { ++ struct task_struct *t = current; ++ force_sig_info(t->forced_info.si_signo, &t->forced_info, t); ++ t->forced_info.si_signo = 0; ++ } ++#endif ++ + if (thread_info_flags & _TIF_UPROBE) + uprobe_notify_resume(regs); + +diff -Nur linux-4.1.13.orig/arch/x86/kernel/traps.c linux-4.1.13/arch/x86/kernel/traps.c +--- linux-4.1.13.orig/arch/x86/kernel/traps.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kernel/traps.c 2015-11-29 09:23:09.529619496 +0100 +@@ -88,9 +88,21 @@ + local_irq_enable(); + } + +-static inline void preempt_conditional_sti(struct pt_regs *regs) ++static inline void conditional_sti_ist(struct pt_regs *regs) + { ++#ifdef CONFIG_X86_64 ++ /* ++ * X86_64 uses a per CPU stack on the IST for certain traps ++ * like int3. The task can not be preempted when using one ++ * of these stacks, thus preemption must be disabled, otherwise ++ * the stack can be corrupted if the task is scheduled out, ++ * and another task comes in and uses this stack. ++ * ++ * On x86_32 the task keeps its own stack and it is OK if the ++ * task schedules out. ++ */ + preempt_count_inc(); ++#endif + if (regs->flags & X86_EFLAGS_IF) + local_irq_enable(); + } +@@ -101,11 +113,13 @@ + local_irq_disable(); + } + +-static inline void preempt_conditional_cli(struct pt_regs *regs) ++static inline void conditional_cli_ist(struct pt_regs *regs) + { + if (regs->flags & X86_EFLAGS_IF) + local_irq_disable(); ++#ifdef CONFIG_X86_64 + preempt_count_dec(); ++#endif + } + + enum ctx_state ist_enter(struct pt_regs *regs) +@@ -536,9 +550,9 @@ + * as we may switch to the interrupt stack. + */ + debug_stack_usage_inc(); +- preempt_conditional_sti(regs); ++ conditional_sti_ist(regs); + do_trap(X86_TRAP_BP, SIGTRAP, "int3", regs, error_code, NULL); +- preempt_conditional_cli(regs); ++ conditional_cli_ist(regs); + debug_stack_usage_dec(); + exit: + ist_exit(regs, prev_state); +@@ -668,12 +682,12 @@ + debug_stack_usage_inc(); + + /* It's safe to allow irq's after DR6 has been saved */ +- preempt_conditional_sti(regs); ++ conditional_sti_ist(regs); + + if (v8086_mode(regs)) { + handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, + X86_TRAP_DB); +- preempt_conditional_cli(regs); ++ conditional_cli_ist(regs); + debug_stack_usage_dec(); + goto exit; + } +@@ -693,7 +707,7 @@ + si_code = get_si_code(tsk->thread.debugreg6); + if (tsk->thread.debugreg6 & (DR_STEP | DR_TRAP_BITS) || user_icebp) + send_sigtrap(tsk, regs, error_code, si_code); +- preempt_conditional_cli(regs); ++ conditional_cli_ist(regs); + debug_stack_usage_dec(); + + exit: +diff -Nur linux-4.1.13.orig/arch/x86/kvm/lapic.c linux-4.1.13/arch/x86/kvm/lapic.c +--- linux-4.1.13.orig/arch/x86/kvm/lapic.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kvm/lapic.c 2015-11-29 09:23:09.529619496 +0100 +@@ -1104,7 +1104,7 @@ + static void apic_timer_expired(struct kvm_lapic *apic) + { + struct kvm_vcpu *vcpu = apic->vcpu; +- wait_queue_head_t *q = &vcpu->wq; ++ struct swait_head *q = &vcpu->wq; + struct kvm_timer *ktimer = &apic->lapic_timer; + + if (atomic_read(&apic->lapic_timer.pending)) +@@ -1113,8 +1113,8 @@ + atomic_inc(&apic->lapic_timer.pending); + kvm_set_pending_timer(vcpu); + +- if (waitqueue_active(q)) +- wake_up_interruptible(q); ++ if (swaitqueue_active(q)) ++ swait_wake_interruptible(q); + + if (apic_lvtt_tscdeadline(apic)) + ktimer->expired_tscdeadline = ktimer->tscdeadline; +@@ -1167,8 +1167,36 @@ + __delay(tsc_deadline - guest_tsc); + } + ++static enum hrtimer_restart apic_timer_fn(struct hrtimer *data); ++ ++static void __apic_timer_expired(struct hrtimer *data) ++{ ++ int ret, i = 0; ++ enum hrtimer_restart r; ++ struct kvm_timer *ktimer = container_of(data, struct kvm_timer, timer); ++ ++ r = apic_timer_fn(data); ++ ++ if (r == HRTIMER_RESTART) { ++ do { ++ ret = hrtimer_start_expires(data, HRTIMER_MODE_ABS); ++ if (ret == -ETIME) ++ hrtimer_add_expires_ns(&ktimer->timer, ++ ktimer->period); ++ i++; ++ } while (ret == -ETIME && i < 10); ++ ++ if (ret == -ETIME) { ++ printk_once(KERN_ERR "%s: failed to reprogram timer\n", ++ __func__); ++ WARN_ON_ONCE(1); ++ } ++ } ++} ++ + static void start_apic_timer(struct kvm_lapic *apic) + { ++ int ret; + ktime_t now; + + atomic_set(&apic->lapic_timer.pending, 0); +@@ -1199,9 +1227,11 @@ + } + } + +- hrtimer_start(&apic->lapic_timer.timer, ++ ret = hrtimer_start(&apic->lapic_timer.timer, + ktime_add_ns(now, apic->lapic_timer.period), + HRTIMER_MODE_ABS); ++ if (ret == -ETIME) ++ __apic_timer_expired(&apic->lapic_timer.timer); + + apic_debug("%s: bus cycle is %" PRId64 "ns, now 0x%016" + PRIx64 ", " +@@ -1233,8 +1263,10 @@ + do_div(ns, this_tsc_khz); + expire = ktime_add_ns(now, ns); + expire = ktime_sub_ns(expire, lapic_timer_advance_ns); +- hrtimer_start(&apic->lapic_timer.timer, ++ ret = hrtimer_start(&apic->lapic_timer.timer, + expire, HRTIMER_MODE_ABS); ++ if (ret == -ETIME) ++ __apic_timer_expired(&apic->lapic_timer.timer); + } else + apic_timer_expired(apic); + +@@ -1707,6 +1739,7 @@ + hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC, + HRTIMER_MODE_ABS); + apic->lapic_timer.timer.function = apic_timer_fn; ++ apic->lapic_timer.timer.irqsafe = 1; + + /* + * APIC is created enabled. This will prevent kvm_lapic_set_base from +@@ -1834,7 +1867,8 @@ + + timer = &vcpu->arch.apic->lapic_timer.timer; + if (hrtimer_cancel(timer)) +- hrtimer_start_expires(timer, HRTIMER_MODE_ABS); ++ if (hrtimer_start_expires(timer, HRTIMER_MODE_ABS) == -ETIME) ++ __apic_timer_expired(timer); + } + + /* +diff -Nur linux-4.1.13.orig/arch/x86/kvm/x86.c linux-4.1.13/arch/x86/kvm/x86.c +--- linux-4.1.13.orig/arch/x86/kvm/x86.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/kvm/x86.c 2015-11-29 09:23:09.529619496 +0100 +@@ -5813,6 +5813,13 @@ + goto out; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) { ++ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n"); ++ return -EOPNOTSUPP; ++ } ++#endif ++ + r = kvm_mmu_module_init(); + if (r) + goto out_free_percpu; +diff -Nur linux-4.1.13.orig/arch/x86/lib/usercopy_32.c linux-4.1.13/arch/x86/lib/usercopy_32.c +--- linux-4.1.13.orig/arch/x86/lib/usercopy_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/lib/usercopy_32.c 2015-11-29 09:23:09.533619230 +0100 +@@ -647,7 +647,8 @@ + * @from: Source address, in kernel space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from kernel space to user space. + * +@@ -668,7 +669,8 @@ + * @from: Source address, in user space. + * @n: Number of bytes to copy. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Copy data from user space to kernel space. + * +diff -Nur linux-4.1.13.orig/arch/x86/mm/fault.c linux-4.1.13/arch/x86/mm/fault.c +--- linux-4.1.13.orig/arch/x86/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/mm/fault.c 2015-11-29 09:23:09.533619230 +0100 +@@ -13,6 +13,7 @@ + #include <linux/hugetlb.h> /* hstate_index_to_shift */ + #include <linux/prefetch.h> /* prefetchw */ + #include <linux/context_tracking.h> /* exception_enter(), ... */ ++#include <linux/uaccess.h> /* faulthandler_disabled() */ + + #include <asm/traps.h> /* dotraplinkage, ... */ + #include <asm/pgalloc.h> /* pgd_*(), ... */ +@@ -1126,9 +1127,9 @@ + + /* + * If we're in an interrupt, have no user context or are running +- * in an atomic region then we must not take the fault: ++ * in a region with pagefaults disabled then we must not take the fault + */ +- if (unlikely(in_atomic() || !mm)) { ++ if (unlikely(faulthandler_disabled() || !mm)) { + bad_area_nosemaphore(regs, error_code, address); + return; + } +diff -Nur linux-4.1.13.orig/arch/x86/mm/highmem_32.c linux-4.1.13/arch/x86/mm/highmem_32.c +--- linux-4.1.13.orig/arch/x86/mm/highmem_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/mm/highmem_32.c 2015-11-29 09:23:09.533619230 +0100 +@@ -32,10 +32,11 @@ + */ + void *kmap_atomic_prot(struct page *page, pgprot_t prot) + { ++ pte_t pte = mk_pte(page, prot); + unsigned long vaddr; + int idx, type; + +- /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ ++ preempt_disable_nort(); + pagefault_disable(); + + if (!PageHighMem(page)) +@@ -45,7 +46,10 @@ + idx = type + KM_TYPE_NR*smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); + BUG_ON(!pte_none(*(kmap_pte-idx))); +- set_pte(kmap_pte-idx, mk_pte(page, prot)); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte-idx, pte); + arch_flush_lazy_mmu_mode(); + + return (void *)vaddr; +@@ -88,6 +92,9 @@ + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + arch_flush_lazy_mmu_mode(); +@@ -100,6 +107,7 @@ + #endif + + pagefault_enable(); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +diff -Nur linux-4.1.13.orig/arch/x86/mm/iomap_32.c linux-4.1.13/arch/x86/mm/iomap_32.c +--- linux-4.1.13.orig/arch/x86/mm/iomap_32.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/mm/iomap_32.c 2015-11-29 09:23:09.533619230 +0100 +@@ -56,15 +56,22 @@ + + void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) + { ++ pte_t pte = pfn_pte(pfn, prot); + unsigned long vaddr; + int idx, type; + ++ preempt_disable(); + pagefault_disable(); + + type = kmap_atomic_idx_push(); + idx = type + KM_TYPE_NR * smp_processor_id(); + vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); +- set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); ++ WARN_ON(!pte_none(*(kmap_pte - idx))); ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = pte; ++#endif ++ set_pte(kmap_pte - idx, pte); + arch_flush_lazy_mmu_mode(); + + return (void *)vaddr; +@@ -112,10 +119,14 @@ + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ current->kmap_pte[type] = __pte(0); ++#endif + kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + } + + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL_GPL(iounmap_atomic); +diff -Nur linux-4.1.13.orig/arch/x86/platform/uv/tlb_uv.c linux-4.1.13/arch/x86/platform/uv/tlb_uv.c +--- linux-4.1.13.orig/arch/x86/platform/uv/tlb_uv.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/platform/uv/tlb_uv.c 2015-11-29 09:23:09.533619230 +0100 +@@ -714,9 +714,9 @@ + + quiesce_local_uvhub(hmaster); + +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); + + end_uvhub_quiesce(hmaster); + +@@ -736,9 +736,9 @@ + + quiesce_local_uvhub(hmaster); + +- spin_lock(&hmaster->queue_lock); ++ raw_spin_lock(&hmaster->queue_lock); + reset_with_ipi(&bau_desc->distribution, bcp); +- spin_unlock(&hmaster->queue_lock); ++ raw_spin_unlock(&hmaster->queue_lock); + + end_uvhub_quiesce(hmaster); + +@@ -759,7 +759,7 @@ + cycles_t tm1; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (!bcp->baudisabled) { + stat->s_bau_disabled++; + tm1 = get_cycles(); +@@ -772,7 +772,7 @@ + } + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + } + + static void count_max_concurr(int stat, struct bau_control *bcp, +@@ -835,7 +835,7 @@ + */ + static void uv1_throttle(struct bau_control *hmaster, struct ptc_stats *stat) + { +- spinlock_t *lock = &hmaster->uvhub_lock; ++ raw_spinlock_t *lock = &hmaster->uvhub_lock; + atomic_t *v; + + v = &hmaster->active_descriptor_count; +@@ -968,7 +968,7 @@ + struct bau_control *hmaster; + + hmaster = bcp->uvhub_master; +- spin_lock(&hmaster->disable_lock); ++ raw_spin_lock(&hmaster->disable_lock); + if (bcp->baudisabled && (get_cycles() >= bcp->set_bau_on_time)) { + stat->s_bau_reenabled++; + for_each_present_cpu(tcpu) { +@@ -980,10 +980,10 @@ + tbcp->period_giveups = 0; + } + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return 0; + } +- spin_unlock(&hmaster->disable_lock); ++ raw_spin_unlock(&hmaster->disable_lock); + return -1; + } + +@@ -1901,9 +1901,9 @@ + bcp->cong_reps = congested_reps; + bcp->disabled_period = sec_2_cycles(disabled_period); + bcp->giveup_limit = giveup_limit; +- spin_lock_init(&bcp->queue_lock); +- spin_lock_init(&bcp->uvhub_lock); +- spin_lock_init(&bcp->disable_lock); ++ raw_spin_lock_init(&bcp->queue_lock); ++ raw_spin_lock_init(&bcp->uvhub_lock); ++ raw_spin_lock_init(&bcp->disable_lock); + } + } + +diff -Nur linux-4.1.13.orig/arch/x86/platform/uv/uv_time.c linux-4.1.13/arch/x86/platform/uv/uv_time.c +--- linux-4.1.13.orig/arch/x86/platform/uv/uv_time.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/x86/platform/uv/uv_time.c 2015-11-29 09:23:09.533619230 +0100 +@@ -58,7 +58,7 @@ + + /* There is one of these allocated per node */ + struct uv_rtc_timer_head { +- spinlock_t lock; ++ raw_spinlock_t lock; + /* next cpu waiting for timer, local node relative: */ + int next_cpu; + /* number of cpus on this node: */ +@@ -178,7 +178,7 @@ + uv_rtc_deallocate_timers(); + return -ENOMEM; + } +- spin_lock_init(&head->lock); ++ raw_spin_lock_init(&head->lock); + head->ncpus = uv_blade_nr_possible_cpus(bid); + head->next_cpu = -1; + blade_info[bid] = head; +@@ -232,7 +232,7 @@ + unsigned long flags; + int next_cpu; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + next_cpu = head->next_cpu; + *t = expires; +@@ -244,12 +244,12 @@ + if (uv_setup_intr(cpu, expires)) { + *t = ULLONG_MAX; + uv_rtc_find_next_timer(head, pnode); +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return -ETIME; + } + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + return 0; + } + +@@ -268,7 +268,7 @@ + unsigned long flags; + int rc = 0; + +- spin_lock_irqsave(&head->lock, flags); ++ raw_spin_lock_irqsave(&head->lock, flags); + + if ((head->next_cpu == bcpu && uv_read_rtc(NULL) >= *t) || force) + rc = 1; +@@ -280,7 +280,7 @@ + uv_rtc_find_next_timer(head, pnode); + } + +- spin_unlock_irqrestore(&head->lock, flags); ++ raw_spin_unlock_irqrestore(&head->lock, flags); + + return rc; + } +@@ -300,13 +300,18 @@ + static cycle_t uv_read_rtc(struct clocksource *cs) + { + unsigned long offset; ++ cycle_t cycles; + ++ preempt_disable(); + if (uv_get_min_hub_revision_id() == 1) + offset = 0; + else + offset = (uv_blade_processor_id() * L1_CACHE_BYTES) % PAGE_SIZE; + +- return (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ cycles = (cycle_t)uv_read_local_mmr(UVH_RTC | offset); ++ preempt_enable(); ++ ++ return cycles; + } + + /* +diff -Nur linux-4.1.13.orig/arch/xtensa/mm/fault.c linux-4.1.13/arch/xtensa/mm/fault.c +--- linux-4.1.13.orig/arch/xtensa/mm/fault.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/xtensa/mm/fault.c 2015-11-29 09:23:09.533619230 +0100 +@@ -15,10 +15,10 @@ + #include <linux/mm.h> + #include <linux/module.h> + #include <linux/hardirq.h> ++#include <linux/uaccess.h> + #include <asm/mmu_context.h> + #include <asm/cacheflush.h> + #include <asm/hardirq.h> +-#include <asm/uaccess.h> + #include <asm/pgalloc.h> + + DEFINE_PER_CPU(unsigned long, asid_cache) = ASID_USER_FIRST; +@@ -57,7 +57,7 @@ + /* If we're in an interrupt or have no user + * context, we must not take the fault.. + */ +- if (in_atomic() || !mm) { ++ if (faulthandler_disabled() || !mm) { + bad_page_fault(regs, address, SIGSEGV); + return; + } +diff -Nur linux-4.1.13.orig/arch/xtensa/mm/highmem.c linux-4.1.13/arch/xtensa/mm/highmem.c +--- linux-4.1.13.orig/arch/xtensa/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/arch/xtensa/mm/highmem.c 2015-11-29 09:23:09.533619230 +0100 +@@ -42,6 +42,7 @@ + enum fixed_addresses idx; + unsigned long vaddr; + ++ preempt_disable(); + pagefault_disable(); + if (!PageHighMem(page)) + return page_address(page); +@@ -79,6 +80,7 @@ + } + + pagefault_enable(); ++ preempt_enable(); + } + EXPORT_SYMBOL(__kunmap_atomic); + +diff -Nur linux-4.1.13.orig/block/blk-core.c linux-4.1.13/block/blk-core.c +--- linux-4.1.13.orig/block/blk-core.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-core.c 2015-11-29 09:23:09.533619230 +0100 +@@ -100,6 +100,9 @@ + + INIT_LIST_HEAD(&rq->queuelist); + INIT_LIST_HEAD(&rq->timeout_list); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + rq->cpu = -1; + rq->q = q; + rq->__sector = (sector_t) -1; +@@ -194,7 +197,7 @@ + **/ + void blk_start_queue(struct request_queue *q) + { +- WARN_ON(!irqs_disabled()); ++ WARN_ON_NONRT(!irqs_disabled()); + + queue_flag_clear(QUEUE_FLAG_STOPPED, q); + __blk_run_queue(q); +@@ -661,7 +664,7 @@ + q->bypass_depth = 1; + __set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + +- init_waitqueue_head(&q->mq_freeze_wq); ++ init_swait_head(&q->mq_freeze_wq); + + if (blkcg_init_queue(q)) + goto fail_bdi; +@@ -3077,7 +3080,7 @@ + blk_run_queue_async(q); + else + __blk_run_queue(q); +- spin_unlock(q->queue_lock); ++ spin_unlock_irq(q->queue_lock); + } + + static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) +@@ -3125,7 +3128,6 @@ + void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) + { + struct request_queue *q; +- unsigned long flags; + struct request *rq; + LIST_HEAD(list); + unsigned int depth; +@@ -3145,11 +3147,6 @@ + q = NULL; + depth = 0; + +- /* +- * Save and disable interrupts here, to avoid doing it for every +- * queue lock we have to take. +- */ +- local_irq_save(flags); + while (!list_empty(&list)) { + rq = list_entry_rq(list.next); + list_del_init(&rq->queuelist); +@@ -3162,7 +3159,7 @@ + queue_unplugged(q, depth, from_schedule); + q = rq->q; + depth = 0; +- spin_lock(q->queue_lock); ++ spin_lock_irq(q->queue_lock); + } + + /* +@@ -3189,8 +3186,6 @@ + */ + if (q) + queue_unplugged(q, depth, from_schedule); +- +- local_irq_restore(flags); + } + + void blk_finish_plug(struct blk_plug *plug) +diff -Nur linux-4.1.13.orig/block/blk-ioc.c linux-4.1.13/block/blk-ioc.c +--- linux-4.1.13.orig/block/blk-ioc.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-ioc.c 2015-11-29 09:23:09.533619230 +0100 +@@ -7,6 +7,7 @@ + #include <linux/bio.h> + #include <linux/blkdev.h> + #include <linux/slab.h> ++#include <linux/delay.h> + + #include "blk.h" + +@@ -109,7 +110,7 @@ + spin_unlock(q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + spin_lock_irqsave_nested(&ioc->lock, flags, 1); + } + } +@@ -187,7 +188,7 @@ + spin_unlock(icq->q->queue_lock); + } else { + spin_unlock_irqrestore(&ioc->lock, flags); +- cpu_relax(); ++ cpu_chill(); + goto retry; + } + } +diff -Nur linux-4.1.13.orig/block/blk-iopoll.c linux-4.1.13/block/blk-iopoll.c +--- linux-4.1.13.orig/block/blk-iopoll.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-iopoll.c 2015-11-29 09:23:09.533619230 +0100 +@@ -35,6 +35,7 @@ + list_add_tail(&iop->list, this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(blk_iopoll_sched); + +@@ -132,6 +133,7 @@ + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + /** +@@ -201,6 +203,7 @@ + this_cpu_ptr(&blk_cpu_iopoll)); + __raise_softirq_irqoff(BLOCK_IOPOLL_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + return NOTIFY_OK; +diff -Nur linux-4.1.13.orig/block/blk-mq.c linux-4.1.13/block/blk-mq.c +--- linux-4.1.13.orig/block/blk-mq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-mq.c 2015-11-29 09:23:09.533619230 +0100 +@@ -88,7 +88,7 @@ + if (!(gfp & __GFP_WAIT)) + return -EBUSY; + +- ret = wait_event_interruptible(q->mq_freeze_wq, ++ ret = swait_event_interruptible(q->mq_freeze_wq, + !q->mq_freeze_depth || blk_queue_dying(q)); + if (blk_queue_dying(q)) + return -ENODEV; +@@ -107,7 +107,7 @@ + struct request_queue *q = + container_of(ref, struct request_queue, mq_usage_counter); + +- wake_up_all(&q->mq_freeze_wq); ++ swait_wake_all(&q->mq_freeze_wq); + } + + void blk_mq_freeze_queue_start(struct request_queue *q) +@@ -127,7 +127,7 @@ + + static void blk_mq_freeze_queue_wait(struct request_queue *q) + { +- wait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->mq_usage_counter)); ++ swait_event(q->mq_freeze_wq, percpu_ref_is_zero(&q->mq_usage_counter)); + } + + /* +@@ -151,7 +151,7 @@ + spin_unlock_irq(q->queue_lock); + if (wake) { + percpu_ref_reinit(&q->mq_usage_counter); +- wake_up_all(&q->mq_freeze_wq); ++ swait_wake_all(&q->mq_freeze_wq); + } + } + EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); +@@ -170,7 +170,7 @@ + * dying, we need to ensure that processes currently waiting on + * the queue are notified as well. + */ +- wake_up_all(&q->mq_freeze_wq); ++ swait_wake_all(&q->mq_freeze_wq); + } + + bool blk_mq_can_queue(struct blk_mq_hw_ctx *hctx) +@@ -217,6 +217,9 @@ + rq->resid_len = 0; + rq->sense = NULL; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ INIT_WORK(&rq->work, __blk_mq_complete_request_remote_work); ++#endif + INIT_LIST_HEAD(&rq->timeout_list); + rq->timeout = 0; + +@@ -346,6 +349,17 @@ + } + EXPORT_SYMBOL(blk_mq_end_request); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++void __blk_mq_complete_request_remote_work(struct work_struct *work) ++{ ++ struct request *rq = container_of(work, struct request, work); ++ ++ rq->q->softirq_done_fn(rq); ++} ++ ++#else ++ + static void __blk_mq_complete_request_remote(void *data) + { + struct request *rq = data; +@@ -353,6 +367,8 @@ + rq->q->softirq_done_fn(rq); + } + ++#endif ++ + static void blk_mq_ipi_complete_request(struct request *rq) + { + struct blk_mq_ctx *ctx = rq->mq_ctx; +@@ -364,19 +380,23 @@ + return; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + if (!test_bit(QUEUE_FLAG_SAME_FORCE, &rq->q->queue_flags)) + shared = cpus_share_cache(cpu, ctx->cpu); + + if (cpu != ctx->cpu && !shared && cpu_online(ctx->cpu)) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ schedule_work_on(ctx->cpu, &rq->work); ++#else + rq->csd.func = __blk_mq_complete_request_remote; + rq->csd.info = rq; + rq->csd.flags = 0; + smp_call_function_single_async(ctx->cpu, &rq->csd); ++#endif + } else { + rq->q->softirq_done_fn(rq); + } +- put_cpu(); ++ put_cpu_light(); + } + + void __blk_mq_complete_request(struct request *rq) +@@ -905,14 +925,14 @@ + return; + + if (!async) { +- int cpu = get_cpu(); ++ int cpu = get_cpu_light(); + if (cpumask_test_cpu(cpu, hctx->cpumask)) { + __blk_mq_run_hw_queue(hctx); +- put_cpu(); ++ put_cpu_light(); + return; + } + +- put_cpu(); ++ put_cpu_light(); + } + + kblockd_schedule_delayed_work_on(blk_mq_hctx_next_cpu(hctx), +@@ -1589,7 +1609,7 @@ + { + struct blk_mq_hw_ctx *hctx = data; + +- if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) ++ if (action == CPU_POST_DEAD) + return blk_mq_hctx_cpu_offline(hctx, cpu); + + /* +diff -Nur linux-4.1.13.orig/block/blk-mq-cpu.c linux-4.1.13/block/blk-mq-cpu.c +--- linux-4.1.13.orig/block/blk-mq-cpu.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-mq-cpu.c 2015-11-29 09:23:09.533619230 +0100 +@@ -16,7 +16,7 @@ + #include "blk-mq.h" + + static LIST_HEAD(blk_mq_cpu_notify_list); +-static DEFINE_RAW_SPINLOCK(blk_mq_cpu_notify_lock); ++static DEFINE_SPINLOCK(blk_mq_cpu_notify_lock); + + static int blk_mq_main_cpu_notify(struct notifier_block *self, + unsigned long action, void *hcpu) +@@ -25,7 +25,10 @@ + struct blk_mq_cpu_notifier *notify; + int ret = NOTIFY_OK; + +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ if (action != CPU_POST_DEAD) ++ return NOTIFY_OK; ++ ++ spin_lock(&blk_mq_cpu_notify_lock); + + list_for_each_entry(notify, &blk_mq_cpu_notify_list, list) { + ret = notify->notify(notify->data, action, cpu); +@@ -33,7 +36,7 @@ + break; + } + +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + return ret; + } + +@@ -41,16 +44,16 @@ + { + BUG_ON(!notifier->notify); + +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ spin_lock(&blk_mq_cpu_notify_lock); + list_add_tail(¬ifier->list, &blk_mq_cpu_notify_list); +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + } + + void blk_mq_unregister_cpu_notifier(struct blk_mq_cpu_notifier *notifier) + { +- raw_spin_lock(&blk_mq_cpu_notify_lock); ++ spin_lock(&blk_mq_cpu_notify_lock); + list_del(¬ifier->list); +- raw_spin_unlock(&blk_mq_cpu_notify_lock); ++ spin_unlock(&blk_mq_cpu_notify_lock); + } + + void blk_mq_init_cpu_notifier(struct blk_mq_cpu_notifier *notifier, +diff -Nur linux-4.1.13.orig/block/blk-mq.h linux-4.1.13/block/blk-mq.h +--- linux-4.1.13.orig/block/blk-mq.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-mq.h 2015-11-29 09:23:09.533619230 +0100 +@@ -76,7 +76,10 @@ + static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, + unsigned int cpu) + { +- return per_cpu_ptr(q->queue_ctx, cpu); ++ struct blk_mq_ctx *ctx; ++ ++ ctx = per_cpu_ptr(q->queue_ctx, cpu); ++ return ctx; + } + + /* +@@ -87,12 +90,12 @@ + */ + static inline struct blk_mq_ctx *blk_mq_get_ctx(struct request_queue *q) + { +- return __blk_mq_get_ctx(q, get_cpu()); ++ return __blk_mq_get_ctx(q, get_cpu_light()); + } + + static inline void blk_mq_put_ctx(struct blk_mq_ctx *ctx) + { +- put_cpu(); ++ put_cpu_light(); + } + + struct blk_mq_alloc_data { +diff -Nur linux-4.1.13.orig/block/blk-softirq.c linux-4.1.13/block/blk-softirq.c +--- linux-4.1.13.orig/block/blk-softirq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/blk-softirq.c 2015-11-29 09:23:09.533619230 +0100 +@@ -51,6 +51,7 @@ + raise_softirq_irqoff(BLOCK_SOFTIRQ); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /* +@@ -93,6 +94,7 @@ + this_cpu_ptr(&blk_cpu_done)); + raise_softirq_irqoff(BLOCK_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + return NOTIFY_OK; +@@ -150,6 +152,7 @@ + goto do_local; + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + /** +diff -Nur linux-4.1.13.orig/block/bounce.c linux-4.1.13/block/bounce.c +--- linux-4.1.13.orig/block/bounce.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/block/bounce.c 2015-11-29 09:23:09.537618965 +0100 +@@ -54,11 +54,11 @@ + unsigned long flags; + unsigned char *vto; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + vto = kmap_atomic(to->bv_page); + memcpy(vto + to->bv_offset, vfrom, to->bv_len); + kunmap_atomic(vto); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + #else /* CONFIG_HIGHMEM */ +diff -Nur linux-4.1.13.orig/crypto/algapi.c linux-4.1.13/crypto/algapi.c +--- linux-4.1.13.orig/crypto/algapi.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/crypto/algapi.c 2015-11-29 09:23:09.537618965 +0100 +@@ -695,13 +695,13 @@ + + int crypto_register_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_register(&crypto_chain, nb); ++ return srcu_notifier_chain_register(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_register_notifier); + + int crypto_unregister_notifier(struct notifier_block *nb) + { +- return blocking_notifier_chain_unregister(&crypto_chain, nb); ++ return srcu_notifier_chain_unregister(&crypto_chain, nb); + } + EXPORT_SYMBOL_GPL(crypto_unregister_notifier); + +diff -Nur linux-4.1.13.orig/crypto/api.c linux-4.1.13/crypto/api.c +--- linux-4.1.13.orig/crypto/api.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/crypto/api.c 2015-11-29 09:23:09.537618965 +0100 +@@ -31,7 +31,7 @@ + DECLARE_RWSEM(crypto_alg_sem); + EXPORT_SYMBOL_GPL(crypto_alg_sem); + +-BLOCKING_NOTIFIER_HEAD(crypto_chain); ++SRCU_NOTIFIER_HEAD(crypto_chain); + EXPORT_SYMBOL_GPL(crypto_chain); + + static struct crypto_alg *crypto_larval_wait(struct crypto_alg *alg); +@@ -236,10 +236,10 @@ + { + int ok; + +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + if (ok == NOTIFY_DONE) { + request_module("cryptomgr"); +- ok = blocking_notifier_call_chain(&crypto_chain, val, v); ++ ok = srcu_notifier_call_chain(&crypto_chain, val, v); + } + + return ok; +diff -Nur linux-4.1.13.orig/crypto/internal.h linux-4.1.13/crypto/internal.h +--- linux-4.1.13.orig/crypto/internal.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/crypto/internal.h 2015-11-29 09:23:09.537618965 +0100 +@@ -48,7 +48,7 @@ + + extern struct list_head crypto_alg_list; + extern struct rw_semaphore crypto_alg_sem; +-extern struct blocking_notifier_head crypto_chain; ++extern struct srcu_notifier_head crypto_chain; + + #ifdef CONFIG_PROC_FS + void __init crypto_init_proc(void); +@@ -142,7 +142,7 @@ + + static inline void crypto_notify(unsigned long val, void *v) + { +- blocking_notifier_call_chain(&crypto_chain, val, v); ++ srcu_notifier_call_chain(&crypto_chain, val, v); + } + + #endif /* _CRYPTO_INTERNAL_H */ +diff -Nur linux-4.1.13.orig/Documentation/hwlat_detector.txt linux-4.1.13/Documentation/hwlat_detector.txt +--- linux-4.1.13.orig/Documentation/hwlat_detector.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/Documentation/hwlat_detector.txt 2015-11-29 09:23:09.477622951 +0100 +@@ -0,0 +1,64 @@ ++Introduction: ++------------- ++ ++The module hwlat_detector is a special purpose kernel module that is used to ++detect large system latencies induced by the behavior of certain underlying ++hardware or firmware, independent of Linux itself. The code was developed ++originally to detect SMIs (System Management Interrupts) on x86 systems, ++however there is nothing x86 specific about this patchset. It was ++originally written for use by the "RT" patch since the Real Time ++kernel is highly latency sensitive. ++ ++SMIs are usually not serviced by the Linux kernel, which typically does not ++even know that they are occuring. SMIs are instead are set up by BIOS code ++and are serviced by BIOS code, usually for "critical" events such as ++management of thermal sensors and fans. Sometimes though, SMIs are used for ++other tasks and those tasks can spend an inordinate amount of time in the ++handler (sometimes measured in milliseconds). Obviously this is a problem if ++you are trying to keep event service latencies down in the microsecond range. ++ ++The hardware latency detector works by hogging all of the cpus for configurable ++amounts of time (by calling stop_machine()), polling the CPU Time Stamp Counter ++for some period, then looking for gaps in the TSC data. Any gap indicates a ++time when the polling was interrupted and since the machine is stopped and ++interrupts turned off the only thing that could do that would be an SMI. ++ ++Note that the SMI detector should *NEVER* be used in a production environment. ++It is intended to be run manually to determine if the hardware platform has a ++problem with long system firmware service routines. ++ ++Usage: ++------ ++ ++Loading the module hwlat_detector passing the parameter "enabled=1" (or by ++setting the "enable" entry in "hwlat_detector" debugfs toggled on) is the only ++step required to start the hwlat_detector. It is possible to redefine the ++threshold in microseconds (us) above which latency spikes will be taken ++into account (parameter "threshold="). ++ ++Example: ++ ++ # modprobe hwlat_detector enabled=1 threshold=100 ++ ++After the module is loaded, it creates a directory named "hwlat_detector" under ++the debugfs mountpoint, "/debug/hwlat_detector" for this text. It is necessary ++to have debugfs mounted, which might be on /sys/debug on your system. ++ ++The /debug/hwlat_detector interface contains the following files: ++ ++count - number of latency spikes observed since last reset ++enable - a global enable/disable toggle (0/1), resets count ++max - maximum hardware latency actually observed (usecs) ++sample - a pipe from which to read current raw sample data ++ in the format <timestamp> <latency observed usecs> ++ (can be opened O_NONBLOCK for a single sample) ++threshold - minimum latency value to be considered (usecs) ++width - time period to sample with CPUs held (usecs) ++ must be less than the total window size (enforced) ++window - total period of sampling, width being inside (usecs) ++ ++By default we will set width to 500,000 and window to 1,000,000, meaning that ++we will sample every 1,000,000 usecs (1s) for 500,000 usecs (0.5s). If we ++observe any latencies that exceed the threshold (initially 100 usecs), ++then we write to a global sample ring buffer of 8K samples, which is ++consumed by reading from the "sample" (pipe) debugfs file interface. +diff -Nur linux-4.1.13.orig/Documentation/sysrq.txt linux-4.1.13/Documentation/sysrq.txt +--- linux-4.1.13.orig/Documentation/sysrq.txt 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/Documentation/sysrq.txt 2015-11-29 09:23:09.477622951 +0100 +@@ -59,10 +59,17 @@ + On other - If you know of the key combos for other architectures, please + let me know so I can add them to this section. + +-On all - write a character to /proc/sysrq-trigger. e.g.: +- ++On all - write a character to /proc/sysrq-trigger, e.g.: + echo t > /proc/sysrq-trigger + ++On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g. ++ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq ++ Send an ICMP echo request with this pattern plus the particular ++ SysRq command key. Example: ++ # ping -c1 -s57 -p0102030468 ++ will trigger the SysRq-H (help) command. ++ ++ + * What are the 'command' keys? + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + 'b' - Will immediately reboot the system without syncing or unmounting +diff -Nur linux-4.1.13.orig/Documentation/trace/histograms.txt linux-4.1.13/Documentation/trace/histograms.txt +--- linux-4.1.13.orig/Documentation/trace/histograms.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/Documentation/trace/histograms.txt 2015-11-29 09:23:09.477622951 +0100 +@@ -0,0 +1,186 @@ ++ Using the Linux Kernel Latency Histograms ++ ++ ++This document gives a short explanation how to enable, configure and use ++latency histograms. Latency histograms are primarily relevant in the ++context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT) ++and are used in the quality management of the Linux real-time ++capabilities. ++ ++ ++* Purpose of latency histograms ++ ++A latency histogram continuously accumulates the frequencies of latency ++data. There are two types of histograms ++- potential sources of latencies ++- effective latencies ++ ++ ++* Potential sources of latencies ++ ++Potential sources of latencies are code segments where interrupts, ++preemption or both are disabled (aka critical sections). To create ++histograms of potential sources of latency, the kernel stores the time ++stamp at the start of a critical section, determines the time elapsed ++when the end of the section is reached, and increments the frequency ++counter of that latency value - irrespective of whether any concurrently ++running process is affected by latency or not. ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_INTERRUPT_OFF_LATENCY ++ CONFIG_PREEMPT_OFF_LATENCY ++ ++ ++* Effective latencies ++ ++Effective latencies are actually occuring during wakeup of a process. To ++determine effective latencies, the kernel stores the time stamp when a ++process is scheduled to be woken up, and determines the duration of the ++wakeup time shortly before control is passed over to this process. Note ++that the apparent latency in user space may be somewhat longer, since the ++process may be interrupted after control is passed over to it but before ++the execution in user space takes place. Simply measuring the interval ++between enqueuing and wakeup may also not appropriate in cases when a ++process is scheduled as a result of a timer expiration. The timer may have ++missed its deadline, e.g. due to disabled interrupts, but this latency ++would not be registered. Therefore, the offsets of missed timers are ++recorded in a separate histogram. If both wakeup latency and missed timer ++offsets are configured and enabled, a third histogram may be enabled that ++records the overall latency as a sum of the timer latency, if any, and the ++wakeup latency. This histogram is called "timerandwakeup". ++- Configuration items (in the Kernel hacking/Tracers submenu) ++ CONFIG_WAKEUP_LATENCY ++ CONFIG_MISSED_TIMER_OFSETS ++ ++ ++* Usage ++ ++The interface to the administration of the latency histograms is located ++in the debugfs file system. To mount it, either enter ++ ++mount -t sysfs nodev /sys ++mount -t debugfs nodev /sys/kernel/debug ++ ++from shell command line level, or add ++ ++nodev /sys sysfs defaults 0 0 ++nodev /sys/kernel/debug debugfs defaults 0 0 ++ ++to the file /etc/fstab. All latency histogram related files are then ++available in the directory /sys/kernel/debug/tracing/latency_hist. A ++particular histogram type is enabled by writing non-zero to the related ++variable in the /sys/kernel/debug/tracing/latency_hist/enable directory. ++Select "preemptirqsoff" for the histograms of potential sources of ++latencies and "wakeup" for histograms of effective latencies etc. The ++histogram data - one per CPU - are available in the files ++ ++/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx ++/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx ++/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx ++/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx ++ ++The histograms are reset by writing non-zero to the file "reset" in a ++particular latency directory. To reset all latency data, use ++ ++#!/bin/sh ++ ++TRACINGDIR=/sys/kernel/debug/tracing ++HISTDIR=$TRACINGDIR/latency_hist ++ ++if test -d $HISTDIR ++then ++ cd $HISTDIR ++ for i in `find . | grep /reset$` ++ do ++ echo 1 >$i ++ done ++fi ++ ++ ++* Data format ++ ++Latency data are stored with a resolution of one microsecond. The ++maximum latency is 10,240 microseconds. The data are only valid, if the ++overflow register is empty. Every output line contains the latency in ++microseconds in the first row and the number of samples in the second ++row. To display only lines with a positive latency count, use, for ++example, ++ ++grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0 ++ ++#Minimum latency: 0 microseconds. ++#Average latency: 0 microseconds. ++#Maximum latency: 25 microseconds. ++#Total samples: 3104770694 ++#There are 0 samples greater or equal than 10240 microseconds ++#usecs samples ++ 0 2984486876 ++ 1 49843506 ++ 2 58219047 ++ 3 5348126 ++ 4 2187960 ++ 5 3388262 ++ 6 959289 ++ 7 208294 ++ 8 40420 ++ 9 4485 ++ 10 14918 ++ 11 18340 ++ 12 25052 ++ 13 19455 ++ 14 5602 ++ 15 969 ++ 16 47 ++ 17 18 ++ 18 14 ++ 19 1 ++ 20 3 ++ 21 2 ++ 22 5 ++ 23 2 ++ 25 1 ++ ++ ++* Wakeup latency of a selected process ++ ++To only collect wakeup latency data of a particular process, write the ++PID of the requested process to ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/pid ++ ++PIDs are not considered, if this variable is set to 0. ++ ++ ++* Details of the process with the highest wakeup latency so far ++ ++Selected data of the process that suffered from the highest wakeup ++latency that occurred in a particular CPU are available in the file ++ ++/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx. ++ ++In addition, other relevant system data at the time when the ++latency occurred are given. ++ ++The format of the data is (all in one line): ++<PID> <Priority> <Latency> (<Timeroffset>) <Command> \ ++<- <PID> <Priority> <Command> <Timestamp> ++ ++The value of <Timeroffset> is only relevant in the combined timer ++and wakeup latency recording. In the wakeup recording, it is ++always 0, in the missed_timer_offsets recording, it is the same ++as <Latency>. ++ ++When retrospectively searching for the origin of a latency and ++tracing was not enabled, it may be helpful to know the name and ++some basic data of the task that (finally) was switching to the ++late real-tlme task. In addition to the victim's data, also the ++data of the possible culprit are therefore displayed after the ++"<-" symbol. ++ ++Finally, the timestamp of the time when the latency occurred ++in <seconds>.<microseconds> after the most recent system boot ++is provided. ++ ++These data are also reset when the wakeup histogram is reset. +diff -Nur linux-4.1.13.orig/drivers/acpi/acpica/acglobal.h linux-4.1.13/drivers/acpi/acpica/acglobal.h +--- linux-4.1.13.orig/drivers/acpi/acpica/acglobal.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/acpi/acpica/acglobal.h 2015-11-29 09:23:09.537618965 +0100 +@@ -112,7 +112,7 @@ + * interrupt level + */ + ACPI_GLOBAL(acpi_spinlock, acpi_gbl_gpe_lock); /* For GPE data structs and registers */ +-ACPI_GLOBAL(acpi_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ ++ACPI_GLOBAL(acpi_raw_spinlock, acpi_gbl_hardware_lock); /* For ACPI H/W except GPE registers */ + ACPI_GLOBAL(acpi_spinlock, acpi_gbl_reference_count_lock); + + /* Mutex for _OSI support */ +diff -Nur linux-4.1.13.orig/drivers/acpi/acpica/hwregs.c linux-4.1.13/drivers/acpi/acpica/hwregs.c +--- linux-4.1.13.orig/drivers/acpi/acpica/hwregs.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/acpi/acpica/hwregs.c 2015-11-29 09:23:09.537618965 +0100 +@@ -269,14 +269,14 @@ + ACPI_BITMASK_ALL_FIXED_STATUS, + ACPI_FORMAT_UINT64(acpi_gbl_xpm1a_status.address))); + +- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); ++ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); + + /* Clear the fixed events in PM1 A/B */ + + status = acpi_hw_register_write(ACPI_REGISTER_PM1_STATUS, + ACPI_BITMASK_ALL_FIXED_STATUS); + +- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); ++ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); + + if (ACPI_FAILURE(status)) { + goto exit; +diff -Nur linux-4.1.13.orig/drivers/acpi/acpica/hwxface.c linux-4.1.13/drivers/acpi/acpica/hwxface.c +--- linux-4.1.13.orig/drivers/acpi/acpica/hwxface.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/acpi/acpica/hwxface.c 2015-11-29 09:23:09.537618965 +0100 +@@ -374,7 +374,7 @@ + return_ACPI_STATUS(AE_BAD_PARAMETER); + } + +- lock_flags = acpi_os_acquire_lock(acpi_gbl_hardware_lock); ++ raw_spin_lock_irqsave(acpi_gbl_hardware_lock, lock_flags); + + /* + * At this point, we know that the parent register is one of the +@@ -435,7 +435,7 @@ + + unlock_and_exit: + +- acpi_os_release_lock(acpi_gbl_hardware_lock, lock_flags); ++ raw_spin_unlock_irqrestore(acpi_gbl_hardware_lock, lock_flags); + return_ACPI_STATUS(status); + } + +diff -Nur linux-4.1.13.orig/drivers/acpi/acpica/utmutex.c linux-4.1.13/drivers/acpi/acpica/utmutex.c +--- linux-4.1.13.orig/drivers/acpi/acpica/utmutex.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/acpi/acpica/utmutex.c 2015-11-29 09:23:09.537618965 +0100 +@@ -88,7 +88,7 @@ + return_ACPI_STATUS (status); + } + +- status = acpi_os_create_lock (&acpi_gbl_hardware_lock); ++ status = acpi_os_create_raw_lock (&acpi_gbl_hardware_lock); + if (ACPI_FAILURE (status)) { + return_ACPI_STATUS (status); + } +@@ -141,7 +141,7 @@ + /* Delete the spinlocks */ + + acpi_os_delete_lock(acpi_gbl_gpe_lock); +- acpi_os_delete_lock(acpi_gbl_hardware_lock); ++ acpi_os_delete_raw_lock(acpi_gbl_hardware_lock); + acpi_os_delete_lock(acpi_gbl_reference_count_lock); + + /* Delete the reader/writer lock */ +diff -Nur linux-4.1.13.orig/drivers/ata/libata-sff.c linux-4.1.13/drivers/ata/libata-sff.c +--- linux-4.1.13.orig/drivers/ata/libata-sff.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ata/libata-sff.c 2015-11-29 09:23:09.537618965 +0100 +@@ -678,9 +678,9 @@ + unsigned long flags; + unsigned int consumed; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + consumed = ata_sff_data_xfer32(dev, buf, buflen, rw); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + return consumed; + } +@@ -719,7 +719,7 @@ + unsigned long flags; + + /* FIXME: use a bounce buffer */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + buf = kmap_atomic(page); + + /* do the actual data transfer */ +@@ -727,7 +727,7 @@ + do_write); + + kunmap_atomic(buf); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } else { + buf = page_address(page); + ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size, +@@ -864,7 +864,7 @@ + unsigned long flags; + + /* FIXME: use bounce buffer */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + buf = kmap_atomic(page); + + /* do the actual data transfer */ +@@ -872,7 +872,7 @@ + count, rw); + + kunmap_atomic(buf); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } else { + buf = page_address(page); + consumed = ap->ops->sff_data_xfer(dev, buf + offset, +diff -Nur linux-4.1.13.orig/drivers/char/random.c linux-4.1.13/drivers/char/random.c +--- linux-4.1.13.orig/drivers/char/random.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/char/random.c 2015-11-29 09:23:09.537618965 +0100 +@@ -776,8 +776,6 @@ + } sample; + long delta, delta2, delta3; + +- preempt_disable(); +- + sample.jiffies = jiffies; + sample.cycles = random_get_entropy(); + sample.num = num; +@@ -818,7 +816,6 @@ + */ + credit_entropy_bits(r, min_t(int, fls(delta>>1), 11)); + } +- preempt_enable(); + } + + void add_input_randomness(unsigned int type, unsigned int code, +@@ -871,28 +868,27 @@ + return *(ptr + f->reg_idx++); + } + +-void add_interrupt_randomness(int irq, int irq_flags) ++void add_interrupt_randomness(int irq, int irq_flags, __u64 ip) + { + struct entropy_store *r; + struct fast_pool *fast_pool = this_cpu_ptr(&irq_randomness); +- struct pt_regs *regs = get_irq_regs(); + unsigned long now = jiffies; + cycles_t cycles = random_get_entropy(); + __u32 c_high, j_high; +- __u64 ip; + unsigned long seed; + int credit = 0; + + if (cycles == 0) +- cycles = get_reg(fast_pool, regs); ++ cycles = get_reg(fast_pool, NULL); + c_high = (sizeof(cycles) > 4) ? cycles >> 32 : 0; + j_high = (sizeof(now) > 4) ? now >> 32 : 0; + fast_pool->pool[0] ^= cycles ^ j_high ^ irq; + fast_pool->pool[1] ^= now ^ c_high; +- ip = regs ? instruction_pointer(regs) : _RET_IP_; ++ if (!ip) ++ ip = _RET_IP_; + fast_pool->pool[2] ^= ip; + fast_pool->pool[3] ^= (sizeof(ip) > 4) ? ip >> 32 : +- get_reg(fast_pool, regs); ++ get_reg(fast_pool, NULL); + + fast_mix(fast_pool); + add_interrupt_bench(cycles); +diff -Nur linux-4.1.13.orig/drivers/clocksource/tcb_clksrc.c linux-4.1.13/drivers/clocksource/tcb_clksrc.c +--- linux-4.1.13.orig/drivers/clocksource/tcb_clksrc.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/clocksource/tcb_clksrc.c 2015-11-29 09:23:09.537618965 +0100 +@@ -23,8 +23,7 @@ + * this 32 bit free-running counter. the second channel is not used. + * + * - The third channel may be used to provide a 16-bit clockevent +- * source, used in either periodic or oneshot mode. This runs +- * at 32 KiHZ, and can handle delays of up to two seconds. ++ * source, used in either periodic or oneshot mode. + * + * A boot clocksource and clockevent source are also currently needed, + * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so +@@ -74,6 +73,7 @@ + struct tc_clkevt_device { + struct clock_event_device clkevt; + struct clk *clk; ++ u32 freq; + void __iomem *regs; + }; + +@@ -82,13 +82,6 @@ + return container_of(clkevt, struct tc_clkevt_device, clkevt); + } + +-/* For now, we always use the 32K clock ... this optimizes for NO_HZ, +- * because using one of the divided clocks would usually mean the +- * tick rate can never be less than several dozen Hz (vs 0.5 Hz). +- * +- * A divided clock could be good for high resolution timers, since +- * 30.5 usec resolution can seem "low". +- */ + static u32 timer_clock; + + static void tc_mode(enum clock_event_mode m, struct clock_event_device *d) +@@ -111,11 +104,12 @@ + case CLOCK_EVT_MODE_PERIODIC: + clk_enable(tcd->clk); + +- /* slow clock, count up to RC, then irq and restart */ ++ /* count up to RC, then irq and restart */ + __raw_writel(timer_clock + | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, + regs + ATMEL_TC_REG(2, CMR)); +- __raw_writel((32768 + HZ/2) / HZ, tcaddr + ATMEL_TC_REG(2, RC)); ++ __raw_writel((tcd->freq + HZ / 2) / HZ, ++ tcaddr + ATMEL_TC_REG(2, RC)); + + /* Enable clock and interrupts on RC compare */ + __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER)); +@@ -128,7 +122,7 @@ + case CLOCK_EVT_MODE_ONESHOT: + clk_enable(tcd->clk); + +- /* slow clock, count up to RC, then irq and stop */ ++ /* count up to RC, then irq and stop */ + __raw_writel(timer_clock | ATMEL_TC_CPCSTOP + | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO, + regs + ATMEL_TC_REG(2, CMR)); +@@ -157,8 +151,12 @@ + .name = "tc_clkevt", + .features = CLOCK_EVT_FEAT_PERIODIC + | CLOCK_EVT_FEAT_ONESHOT, ++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK + /* Should be lower than at91rm9200's system timer */ + .rating = 125, ++#else ++ .rating = 200, ++#endif + .set_next_event = tc_next_event, + .set_mode = tc_mode, + }, +@@ -178,8 +176,9 @@ + return IRQ_NONE; + } + +-static int __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx) ++static int __init setup_clkevents(struct atmel_tc *tc, int divisor_idx) + { ++ unsigned divisor = atmel_tc_divisors[divisor_idx]; + int ret; + struct clk *t2_clk = tc->clk[2]; + int irq = tc->irq[2]; +@@ -193,7 +192,11 @@ + clkevt.regs = tc->regs; + clkevt.clk = t2_clk; + +- timer_clock = clk32k_divisor_idx; ++ timer_clock = divisor_idx; ++ if (!divisor) ++ clkevt.freq = 32768; ++ else ++ clkevt.freq = clk_get_rate(t2_clk) / divisor; + + clkevt.clkevt.cpumask = cpumask_of(0); + +@@ -203,7 +206,7 @@ + return ret; + } + +- clockevents_config_and_register(&clkevt.clkevt, 32768, 1, 0xffff); ++ clockevents_config_and_register(&clkevt.clkevt, clkevt.freq, 1, 0xffff); + + return ret; + } +@@ -340,7 +343,11 @@ + goto err_disable_t1; + + /* channel 2: periodic and oneshot timer support */ ++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK + ret = setup_clkevents(tc, clk32k_divisor_idx); ++#else ++ ret = setup_clkevents(tc, best_divisor_idx); ++#endif + if (ret) + goto err_unregister_clksrc; + +diff -Nur linux-4.1.13.orig/drivers/clocksource/timer-atmel-pit.c linux-4.1.13/drivers/clocksource/timer-atmel-pit.c +--- linux-4.1.13.orig/drivers/clocksource/timer-atmel-pit.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/clocksource/timer-atmel-pit.c 2015-11-29 09:23:09.537618965 +0100 +@@ -90,6 +90,7 @@ + return elapsed; + } + ++static struct irqaction at91sam926x_pit_irq; + /* + * Clockevent device: interrupts every 1/HZ (== pit_cycles * MCK/16) + */ +@@ -100,6 +101,8 @@ + + switch (mode) { + case CLOCK_EVT_MODE_PERIODIC: ++ /* Set up irq handler */ ++ setup_irq(at91sam926x_pit_irq.irq, &at91sam926x_pit_irq); + /* update clocksource counter */ + data->cnt += data->cycle * PIT_PICNT(pit_read(data->base, AT91_PIT_PIVR)); + pit_write(data->base, AT91_PIT_MR, +@@ -113,6 +116,7 @@ + /* disable irq, leaving the clocksource active */ + pit_write(data->base, AT91_PIT_MR, + (data->cycle - 1) | AT91_PIT_PITEN); ++ remove_irq(at91sam926x_pit_irq.irq, &at91sam926x_pit_irq); + break; + case CLOCK_EVT_MODE_RESUME: + break; +diff -Nur linux-4.1.13.orig/drivers/clocksource/timer-atmel-st.c linux-4.1.13/drivers/clocksource/timer-atmel-st.c +--- linux-4.1.13.orig/drivers/clocksource/timer-atmel-st.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/clocksource/timer-atmel-st.c 2015-11-29 09:23:09.537618965 +0100 +@@ -131,6 +131,7 @@ + break; + case CLOCK_EVT_MODE_SHUTDOWN: + case CLOCK_EVT_MODE_UNUSED: ++ remove_irq(NR_IRQS_LEGACY + AT91_ID_SYS, &at91rm9200_timer_irq); + case CLOCK_EVT_MODE_RESUME: + irqmask = 0; + break; +diff -Nur linux-4.1.13.orig/drivers/cpufreq/cpufreq.c linux-4.1.13/drivers/cpufreq/cpufreq.c +--- linux-4.1.13.orig/drivers/cpufreq/cpufreq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/cpufreq/cpufreq.c 2015-11-29 09:23:09.545618436 +0100 +@@ -64,12 +64,6 @@ + return cpufreq_driver->target_index || cpufreq_driver->target; + } + +-/* +- * rwsem to guarantee that cpufreq driver module doesn't unload during critical +- * sections +- */ +-static DECLARE_RWSEM(cpufreq_rwsem); +- + /* internal prototypes */ + static int __cpufreq_governor(struct cpufreq_policy *policy, + unsigned int event); +@@ -215,9 +209,6 @@ + if (cpu >= nr_cpu_ids) + return NULL; + +- if (!down_read_trylock(&cpufreq_rwsem)) +- return NULL; +- + /* get the cpufreq driver */ + read_lock_irqsave(&cpufreq_driver_lock, flags); + +@@ -230,9 +221,6 @@ + + read_unlock_irqrestore(&cpufreq_driver_lock, flags); + +- if (!policy) +- up_read(&cpufreq_rwsem); +- + return policy; + } + EXPORT_SYMBOL_GPL(cpufreq_cpu_get); +@@ -240,7 +228,6 @@ + void cpufreq_cpu_put(struct cpufreq_policy *policy) + { + kobject_put(&policy->kobj); +- up_read(&cpufreq_rwsem); + } + EXPORT_SYMBOL_GPL(cpufreq_cpu_put); + +@@ -765,9 +752,6 @@ + struct freq_attr *fattr = to_attr(attr); + ssize_t ret; + +- if (!down_read_trylock(&cpufreq_rwsem)) +- return -EINVAL; +- + down_read(&policy->rwsem); + + if (fattr->show) +@@ -776,7 +760,6 @@ + ret = -EIO; + + up_read(&policy->rwsem); +- up_read(&cpufreq_rwsem); + + return ret; + } +@@ -793,9 +776,6 @@ + if (!cpu_online(policy->cpu)) + goto unlock; + +- if (!down_read_trylock(&cpufreq_rwsem)) +- goto unlock; +- + down_write(&policy->rwsem); + + if (fattr->store) +@@ -804,8 +784,6 @@ + ret = -EIO; + + up_write(&policy->rwsem); +- +- up_read(&cpufreq_rwsem); + unlock: + put_online_cpus(); + +@@ -1117,16 +1095,12 @@ + if (unlikely(policy)) + return 0; + +- if (!down_read_trylock(&cpufreq_rwsem)) +- return 0; +- + /* Check if this cpu was hot-unplugged earlier and has siblings */ + read_lock_irqsave(&cpufreq_driver_lock, flags); + for_each_policy(policy) { + if (cpumask_test_cpu(cpu, policy->related_cpus)) { + read_unlock_irqrestore(&cpufreq_driver_lock, flags); + ret = cpufreq_add_policy_cpu(policy, cpu, dev); +- up_read(&cpufreq_rwsem); + return ret; + } + } +@@ -1269,8 +1243,6 @@ + + kobject_uevent(&policy->kobj, KOBJ_ADD); + +- up_read(&cpufreq_rwsem); +- + /* Callback for handling stuff after policy is ready */ + if (cpufreq_driver->ready) + cpufreq_driver->ready(policy); +@@ -1304,8 +1276,6 @@ + cpufreq_policy_free(policy); + + nomem_out: +- up_read(&cpufreq_rwsem); +- + return ret; + } + +@@ -2499,19 +2469,20 @@ + + pr_debug("unregistering driver %s\n", driver->name); + ++ /* Protect against concurrent cpu hotplug */ ++ get_online_cpus(); + subsys_interface_unregister(&cpufreq_interface); + if (cpufreq_boost_supported()) + cpufreq_sysfs_remove_file(&boost.attr); + + unregister_hotcpu_notifier(&cpufreq_cpu_notifier); + +- down_write(&cpufreq_rwsem); + write_lock_irqsave(&cpufreq_driver_lock, flags); + + cpufreq_driver = NULL; + + write_unlock_irqrestore(&cpufreq_driver_lock, flags); +- up_write(&cpufreq_rwsem); ++ put_online_cpus(); + + return 0; + } +diff -Nur linux-4.1.13.orig/drivers/cpufreq/Kconfig.x86 linux-4.1.13/drivers/cpufreq/Kconfig.x86 +--- linux-4.1.13.orig/drivers/cpufreq/Kconfig.x86 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/cpufreq/Kconfig.x86 2015-11-29 09:23:09.545618436 +0100 +@@ -123,7 +123,7 @@ + + config X86_POWERNOW_K8 + tristate "AMD Opteron/Athlon64 PowerNow!" +- depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ ++ depends on ACPI && ACPI_PROCESSOR && X86_ACPI_CPUFREQ && !PREEMPT_RT_BASE + help + This adds the CPUFreq driver for K8/early Opteron/Athlon64 processors. + Support for K10 and newer processors is now in acpi-cpufreq. +diff -Nur linux-4.1.13.orig/drivers/gpio/gpio-omap.c linux-4.1.13/drivers/gpio/gpio-omap.c +--- linux-4.1.13.orig/drivers/gpio/gpio-omap.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/gpio/gpio-omap.c 2015-11-29 09:23:09.545618436 +0100 +@@ -57,7 +57,7 @@ + u32 saved_datain; + u32 level_mask; + u32 toggle_mask; +- spinlock_t lock; ++ raw_spinlock_t lock; + struct gpio_chip chip; + struct clk *dbck; + u32 mod_usage; +@@ -498,14 +498,14 @@ + (type & (IRQ_TYPE_LEVEL_LOW|IRQ_TYPE_LEVEL_HIGH))) + return -EINVAL; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + retval = omap_set_gpio_triggering(bank, offset, type); + omap_gpio_init_irq(bank, offset); + if (!omap_gpio_is_input(bank, offset)) { +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return -EINVAL; + } +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + if (type & (IRQ_TYPE_LEVEL_LOW | IRQ_TYPE_LEVEL_HIGH)) + __irq_set_handler_locked(d->irq, handle_level_irq); +@@ -626,14 +626,14 @@ + return -EINVAL; + } + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + if (enable) + bank->context.wake_en |= gpio_bit; + else + bank->context.wake_en &= ~gpio_bit; + + writel_relaxed(bank->context.wake_en, bank->base + bank->regs->wkup_en); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -668,7 +668,7 @@ + if (!BANK_USED(bank)) + pm_runtime_get_sync(bank->dev); + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + /* Set trigger to none. You need to enable the desired trigger with + * request_irq() or set_irq_type(). Only do this if the IRQ line has + * not already been requested. +@@ -678,7 +678,7 @@ + omap_enable_gpio_module(bank, offset); + } + bank->mod_usage |= BIT(offset); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -688,11 +688,11 @@ + struct gpio_bank *bank = container_of(chip, struct gpio_bank, chip); + unsigned long flags; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + bank->mod_usage &= ~(BIT(offset)); + omap_disable_gpio_module(bank, offset); + omap_reset_gpio(bank, offset); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + /* + * If this is the last gpio to be freed in the bank, +@@ -794,9 +794,9 @@ + if (!BANK_USED(bank)) + pm_runtime_get_sync(bank->dev); + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + omap_gpio_init_irq(bank, offset); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + omap_gpio_unmask_irq(d); + + return 0; +@@ -808,11 +808,11 @@ + unsigned long flags; + unsigned offset = d->hwirq; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + bank->irq_usage &= ~(BIT(offset)); + omap_disable_gpio_module(bank, offset); + omap_reset_gpio(bank, offset); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + /* + * If this is the last IRQ to be freed in the bank, +@@ -836,10 +836,10 @@ + unsigned offset = d->hwirq; + unsigned long flags; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + omap_set_gpio_irqenable(bank, offset, 0); + omap_set_gpio_triggering(bank, offset, IRQ_TYPE_NONE); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + } + + static void omap_gpio_unmask_irq(struct irq_data *d) +@@ -849,7 +849,7 @@ + u32 trigger = irqd_get_trigger_type(d); + unsigned long flags; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + if (trigger) + omap_set_gpio_triggering(bank, offset, trigger); + +@@ -861,7 +861,7 @@ + } + + omap_set_gpio_irqenable(bank, offset, 1); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + } + + /*---------------------------------------------------------------------*/ +@@ -874,9 +874,9 @@ + OMAP_MPUIO_GPIO_MASKIT / bank->stride; + unsigned long flags; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + writel_relaxed(0xffff & ~bank->context.wake_en, mask_reg); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -889,9 +889,9 @@ + OMAP_MPUIO_GPIO_MASKIT / bank->stride; + unsigned long flags; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + writel_relaxed(bank->context.wake_en, mask_reg); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -937,9 +937,9 @@ + + bank = container_of(chip, struct gpio_bank, chip); + reg = bank->base + bank->regs->direction; +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + dir = !!(readl_relaxed(reg) & BIT(offset)); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return dir; + } + +@@ -949,9 +949,9 @@ + unsigned long flags; + + bank = container_of(chip, struct gpio_bank, chip); +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + omap_set_gpio_direction(bank, offset, 1); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return 0; + } + +@@ -973,10 +973,10 @@ + unsigned long flags; + + bank = container_of(chip, struct gpio_bank, chip); +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + bank->set_dataout(bank, offset, value); + omap_set_gpio_direction(bank, offset, 0); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return 0; + } + +@@ -988,9 +988,9 @@ + + bank = container_of(chip, struct gpio_bank, chip); + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + omap2_set_gpio_debounce(bank, offset, debounce); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -1001,9 +1001,9 @@ + unsigned long flags; + + bank = container_of(chip, struct gpio_bank, chip); +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + bank->set_dataout(bank, offset, value); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + } + + /*---------------------------------------------------------------------*/ +@@ -1199,7 +1199,7 @@ + else + bank->set_dataout = omap_set_gpio_dataout_mask; + +- spin_lock_init(&bank->lock); ++ raw_spin_lock_init(&bank->lock); + + /* Static mapping, never released */ + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); +@@ -1246,7 +1246,7 @@ + unsigned long flags; + u32 wake_low, wake_hi; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + + /* + * Only edges can generate a wakeup event to the PRCM. +@@ -1299,7 +1299,7 @@ + bank->get_context_loss_count(bank->dev); + + omap_gpio_dbck_disable(bank); +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +@@ -1314,7 +1314,7 @@ + unsigned long flags; + int c; + +- spin_lock_irqsave(&bank->lock, flags); ++ raw_spin_lock_irqsave(&bank->lock, flags); + + /* + * On the first resume during the probe, the context has not +@@ -1350,14 +1350,14 @@ + if (c != bank->context_loss_count) { + omap_gpio_restore_context(bank); + } else { +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return 0; + } + } + } + + if (!bank->workaround_enabled) { +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + return 0; + } + +@@ -1412,7 +1412,7 @@ + } + + bank->workaround_enabled = false; +- spin_unlock_irqrestore(&bank->lock, flags); ++ raw_spin_unlock_irqrestore(&bank->lock, flags); + + return 0; + } +diff -Nur linux-4.1.13.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c linux-4.1.13/drivers/gpu/drm/i915/i915_gem_execbuffer.c +--- linux-4.1.13.orig/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/gpu/drm/i915/i915_gem_execbuffer.c 2015-11-29 09:23:09.545618436 +0100 +@@ -32,6 +32,7 @@ + #include "i915_trace.h" + #include "intel_drv.h" + #include <linux/dma_remapping.h> ++#include <linux/uaccess.h> + + #define __EXEC_OBJECT_HAS_PIN (1<<31) + #define __EXEC_OBJECT_HAS_FENCE (1<<30) +@@ -465,7 +466,7 @@ + } + + /* We can't wait for rendering with pagefaults disabled */ +- if (obj->active && in_atomic()) ++ if (obj->active && pagefault_disabled()) + return -EFAULT; + + if (use_cpu_reloc(obj)) +@@ -1338,7 +1339,9 @@ + return ret; + } + ++#ifndef CONFIG_PREEMPT_RT_BASE + trace_i915_gem_ring_dispatch(intel_ring_get_request(ring), dispatch_flags); ++#endif + + i915_gem_execbuffer_move_to_active(vmas, ring); + i915_gem_execbuffer_retire_commands(dev, file, ring, batch_obj); +diff -Nur linux-4.1.13.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c linux-4.1.13/drivers/gpu/drm/i915/i915_gem_shrinker.c +--- linux-4.1.13.orig/drivers/gpu/drm/i915/i915_gem_shrinker.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/gpu/drm/i915/i915_gem_shrinker.c 2015-11-29 09:23:09.545618436 +0100 +@@ -39,7 +39,7 @@ + if (!mutex_is_locked(mutex)) + return false; + +-#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES) ++#if (defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES)) && !defined(CONFIG_PREEMPT_RT_BASE) + return mutex->owner == task; + #else + /* Since UP may be pre-empted, we cannot assume that we own the lock */ +diff -Nur linux-4.1.13.orig/drivers/gpu/drm/i915/intel_display.c linux-4.1.13/drivers/gpu/drm/i915/intel_display.c +--- linux-4.1.13.orig/drivers/gpu/drm/i915/intel_display.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/gpu/drm/i915/intel_display.c 2015-11-29 09:23:09.549618170 +0100 +@@ -10088,7 +10088,7 @@ + struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe]; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + +- WARN_ON(!in_interrupt()); ++ WARN_ON_NONRT(!in_interrupt()); + + if (crtc == NULL) + return; +diff -Nur linux-4.1.13.orig/drivers/i2c/busses/i2c-omap.c linux-4.1.13/drivers/i2c/busses/i2c-omap.c +--- linux-4.1.13.orig/drivers/i2c/busses/i2c-omap.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/i2c/busses/i2c-omap.c 2015-11-29 09:23:09.549618170 +0100 +@@ -996,15 +996,12 @@ + u16 mask; + u16 stat; + +- spin_lock(&dev->lock); +- mask = omap_i2c_read_reg(dev, OMAP_I2C_IE_REG); + stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG); ++ mask = omap_i2c_read_reg(dev, OMAP_I2C_IE_REG); + + if (stat & mask) + ret = IRQ_WAKE_THREAD; + +- spin_unlock(&dev->lock); +- + return ret; + } + +diff -Nur linux-4.1.13.orig/drivers/ide/alim15x3.c linux-4.1.13/drivers/ide/alim15x3.c +--- linux-4.1.13.orig/drivers/ide/alim15x3.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/alim15x3.c 2015-11-29 09:23:09.549618170 +0100 +@@ -234,7 +234,7 @@ + + isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + if (m5229_revision < 0xC2) { + /* +@@ -325,7 +325,7 @@ + } + pci_dev_put(north); + pci_dev_put(isa_dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return 0; + } + +diff -Nur linux-4.1.13.orig/drivers/ide/hpt366.c linux-4.1.13/drivers/ide/hpt366.c +--- linux-4.1.13.orig/drivers/ide/hpt366.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/hpt366.c 2015-11-29 09:23:09.557617634 +0100 +@@ -1241,7 +1241,7 @@ + + dma_old = inb(base + 2); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + dma_new = dma_old; + pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma); +@@ -1252,7 +1252,7 @@ + if (dma_new != dma_old) + outb(dma_new, base + 2); + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n", + hwif->name, base, base + 7); +diff -Nur linux-4.1.13.orig/drivers/ide/ide-io.c linux-4.1.13/drivers/ide/ide-io.c +--- linux-4.1.13.orig/drivers/ide/ide-io.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/ide-io.c 2015-11-29 09:23:09.557617634 +0100 +@@ -659,7 +659,7 @@ + /* disable_irq_nosync ?? */ + disable_irq(hwif->irq); + /* local CPU only, as if we were handling an interrupt */ +- local_irq_disable(); ++ local_irq_disable_nort(); + if (hwif->polling) { + startstop = handler(drive); + } else if (drive_is_ready(drive)) { +diff -Nur linux-4.1.13.orig/drivers/ide/ide-iops.c linux-4.1.13/drivers/ide/ide-iops.c +--- linux-4.1.13.orig/drivers/ide/ide-iops.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/ide-iops.c 2015-11-29 09:23:09.557617634 +0100 +@@ -129,12 +129,12 @@ + if ((stat & ATA_BUSY) == 0) + break; + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + *rstat = stat; + return -EBUSY; + } + } +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + /* + * Allow status to settle, then read it again. +diff -Nur linux-4.1.13.orig/drivers/ide/ide-io-std.c linux-4.1.13/drivers/ide/ide-io-std.c +--- linux-4.1.13.orig/drivers/ide/ide-io-std.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/ide-io-std.c 2015-11-29 09:23:09.557617634 +0100 +@@ -175,7 +175,7 @@ + unsigned long uninitialized_var(flags); + + if ((io_32bit & 2) && !mmio) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + ata_vlb_sync(io_ports->nsect_addr); + } + +@@ -186,7 +186,7 @@ + insl(data_addr, buf, words); + + if ((io_32bit & 2) && !mmio) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + if (((len + 1) & 3) < 2) + return; +@@ -219,7 +219,7 @@ + unsigned long uninitialized_var(flags); + + if ((io_32bit & 2) && !mmio) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + ata_vlb_sync(io_ports->nsect_addr); + } + +@@ -230,7 +230,7 @@ + outsl(data_addr, buf, words); + + if ((io_32bit & 2) && !mmio) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + if (((len + 1) & 3) < 2) + return; +diff -Nur linux-4.1.13.orig/drivers/ide/ide-probe.c linux-4.1.13/drivers/ide/ide-probe.c +--- linux-4.1.13.orig/drivers/ide/ide-probe.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/ide-probe.c 2015-11-29 09:23:09.557617634 +0100 +@@ -196,10 +196,10 @@ + int bswap = 1; + + /* local CPU only; some systems need this */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + /* read 512 bytes of id info */ + hwif->tp_ops->input_data(drive, NULL, id, SECTOR_SIZE); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + drive->dev_flags |= IDE_DFLAG_ID_READ; + #ifdef DEBUG +diff -Nur linux-4.1.13.orig/drivers/ide/ide-taskfile.c linux-4.1.13/drivers/ide/ide-taskfile.c +--- linux-4.1.13.orig/drivers/ide/ide-taskfile.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/ide/ide-taskfile.c 2015-11-29 09:23:09.557617634 +0100 +@@ -250,7 +250,7 @@ + + page_is_high = PageHighMem(page); + if (page_is_high) +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + buf = kmap_atomic(page) + offset; + +@@ -271,7 +271,7 @@ + kunmap_atomic(buf); + + if (page_is_high) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + len -= nr_bytes; + } +@@ -414,7 +414,7 @@ + } + + if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0) +- local_irq_disable(); ++ local_irq_disable_nort(); + + ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE); + +diff -Nur linux-4.1.13.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c linux-4.1.13/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +--- linux-4.1.13.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/infiniband/ulp/ipoib/ipoib_multicast.c 2015-11-29 09:23:09.557617634 +0100 +@@ -821,7 +821,7 @@ + + ipoib_dbg_mcast(priv, "restarting multicast task\n"); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + netif_addr_lock(dev); + spin_lock(&priv->lock); + +@@ -903,7 +903,7 @@ + + spin_unlock(&priv->lock); + netif_addr_unlock(dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + /* + * make sure the in-flight joins have finished before we attempt +diff -Nur linux-4.1.13.orig/drivers/input/gameport/gameport.c linux-4.1.13/drivers/input/gameport/gameport.c +--- linux-4.1.13.orig/drivers/input/gameport/gameport.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/input/gameport/gameport.c 2015-11-29 09:23:09.557617634 +0100 +@@ -124,12 +124,12 @@ + tx = 1 << 30; + + for(i = 0; i < 50; i++) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + GET_TIME(t1); + for (t = 0; t < 50; t++) gameport_read(gameport); + GET_TIME(t2); + GET_TIME(t3); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + udelay(i * 10); + if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t; + } +@@ -148,11 +148,11 @@ + tx = 1 << 30; + + for(i = 0; i < 50; i++) { +- local_irq_save(flags); ++ local_irq_save_nort(flags); + rdtscl(t1); + for (t = 0; t < 50; t++) gameport_read(gameport); + rdtscl(t2); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + udelay(i * 10); + if (t2 - t1 < tx) tx = t2 - t1; + } +diff -Nur linux-4.1.13.orig/drivers/leds/trigger/Kconfig linux-4.1.13/drivers/leds/trigger/Kconfig +--- linux-4.1.13.orig/drivers/leds/trigger/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/leds/trigger/Kconfig 2015-11-29 09:23:09.557617634 +0100 +@@ -61,7 +61,7 @@ + + config LEDS_TRIGGER_CPU + bool "LED CPU Trigger" +- depends on LEDS_TRIGGERS ++ depends on LEDS_TRIGGERS && !PREEMPT_RT_BASE + help + This allows LEDs to be controlled by active CPUs. This shows + the active CPUs across an array of LEDs so you can see which +diff -Nur linux-4.1.13.orig/drivers/md/bcache/Kconfig linux-4.1.13/drivers/md/bcache/Kconfig +--- linux-4.1.13.orig/drivers/md/bcache/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/md/bcache/Kconfig 2015-11-29 09:23:09.557617634 +0100 +@@ -1,6 +1,7 @@ + + config BCACHE + tristate "Block device as cache" ++ depends on !PREEMPT_RT_FULL + ---help--- + Allows a block device to be used as cache for other devices; uses + a btree for indexing and the layout is optimized for SSDs. +diff -Nur linux-4.1.13.orig/drivers/md/dm.c linux-4.1.13/drivers/md/dm.c +--- linux-4.1.13.orig/drivers/md/dm.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/md/dm.c 2015-11-29 09:23:09.557617634 +0100 +@@ -2132,7 +2132,7 @@ + /* Establish tio->ti before queuing work (map_tio_request) */ + tio->ti = ti; + queue_kthread_work(&md->kworker, &tio->work); +- BUG_ON(!irqs_disabled()); ++ BUG_ON_NONRT(!irqs_disabled()); + } + + goto out; +diff -Nur linux-4.1.13.orig/drivers/md/raid5.c linux-4.1.13/drivers/md/raid5.c +--- linux-4.1.13.orig/drivers/md/raid5.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/md/raid5.c 2015-11-29 09:23:09.561617368 +0100 +@@ -1918,8 +1918,9 @@ + struct raid5_percpu *percpu; + unsigned long cpu; + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + percpu = per_cpu_ptr(conf->percpu, cpu); ++ spin_lock(&percpu->lock); + if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) { + ops_run_biofill(sh); + overlap_clear++; +@@ -1975,7 +1976,8 @@ + if (test_and_clear_bit(R5_Overlap, &dev->flags)) + wake_up(&sh->raid_conf->wait_for_overlap); + } +- put_cpu(); ++ spin_unlock(&percpu->lock); ++ put_cpu_light(); + } + + static struct stripe_head *alloc_stripe(struct kmem_cache *sc, gfp_t gfp) +@@ -6363,6 +6365,7 @@ + __func__, cpu); + break; + } ++ spin_lock_init(&per_cpu_ptr(conf->percpu, cpu)->lock); + } + put_online_cpus(); + +diff -Nur linux-4.1.13.orig/drivers/md/raid5.h linux-4.1.13/drivers/md/raid5.h +--- linux-4.1.13.orig/drivers/md/raid5.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/md/raid5.h 2015-11-29 09:23:09.561617368 +0100 +@@ -495,6 +495,7 @@ + int recovery_disabled; + /* per cpu variables */ + struct raid5_percpu { ++ spinlock_t lock; /* Protection for -RT */ + struct page *spare_page; /* Used when checking P/Q in raid6 */ + struct flex_array *scribble; /* space for constructing buffer + * lists and performing address +diff -Nur linux-4.1.13.orig/drivers/misc/hwlat_detector.c linux-4.1.13/drivers/misc/hwlat_detector.c +--- linux-4.1.13.orig/drivers/misc/hwlat_detector.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/drivers/misc/hwlat_detector.c 2015-11-29 09:23:09.561617368 +0100 +@@ -0,0 +1,1240 @@ ++/* ++ * hwlat_detector.c - A simple Hardware Latency detector. ++ * ++ * Use this module to detect large system latencies induced by the behavior of ++ * certain underlying system hardware or firmware, independent of Linux itself. ++ * The code was developed originally to detect the presence of SMIs on Intel ++ * and AMD systems, although there is no dependency upon x86 herein. ++ * ++ * The classical example usage of this module is in detecting the presence of ++ * SMIs or System Management Interrupts on Intel and AMD systems. An SMI is a ++ * somewhat special form of hardware interrupt spawned from earlier CPU debug ++ * modes in which the (BIOS/EFI/etc.) firmware arranges for the South Bridge ++ * LPC (or other device) to generate a special interrupt under certain ++ * circumstances, for example, upon expiration of a special SMI timer device, ++ * due to certain external thermal readings, on certain I/O address accesses, ++ * and other situations. An SMI hits a special CPU pin, triggers a special ++ * SMI mode (complete with special memory map), and the OS is unaware. ++ * ++ * Although certain hardware-inducing latencies are necessary (for example, ++ * a modern system often requires an SMI handler for correct thermal control ++ * and remote management) they can wreak havoc upon any OS-level performance ++ * guarantees toward low-latency, especially when the OS is not even made ++ * aware of the presence of these interrupts. For this reason, we need a ++ * somewhat brute force mechanism to detect these interrupts. In this case, ++ * we do it by hogging all of the CPU(s) for configurable timer intervals, ++ * sampling the built-in CPU timer, looking for discontiguous readings. ++ * ++ * WARNING: This implementation necessarily introduces latencies. Therefore, ++ * you should NEVER use this module in a production environment ++ * requiring any kind of low-latency performance guarantee(s). ++ * ++ * Copyright (C) 2008-2009 Jon Masters, Red Hat, Inc. <jcm@redhat.com> ++ * ++ * Includes useful feedback from Clark Williams <clark@redhat.com> ++ * ++ * This file is licensed under the terms of the GNU General Public ++ * License version 2. This program is licensed "as is" without any ++ * warranty of any kind, whether express or implied. ++ */ ++ ++#include <linux/module.h> ++#include <linux/init.h> ++#include <linux/ring_buffer.h> ++#include <linux/time.h> ++#include <linux/hrtimer.h> ++#include <linux/kthread.h> ++#include <linux/debugfs.h> ++#include <linux/seq_file.h> ++#include <linux/uaccess.h> ++#include <linux/version.h> ++#include <linux/delay.h> ++#include <linux/slab.h> ++#include <linux/trace_clock.h> ++ ++#define BUF_SIZE_DEFAULT 262144UL /* 8K*(sizeof(entry)) */ ++#define BUF_FLAGS (RB_FL_OVERWRITE) /* no block on full */ ++#define U64STR_SIZE 22 /* 20 digits max */ ++ ++#define VERSION "1.0.0" ++#define BANNER "hwlat_detector: " ++#define DRVNAME "hwlat_detector" ++#define DEFAULT_SAMPLE_WINDOW 1000000 /* 1s */ ++#define DEFAULT_SAMPLE_WIDTH 500000 /* 0.5s */ ++#define DEFAULT_LAT_THRESHOLD 10 /* 10us */ ++ ++/* Module metadata */ ++ ++MODULE_LICENSE("GPL"); ++MODULE_AUTHOR("Jon Masters <jcm@redhat.com>"); ++MODULE_DESCRIPTION("A simple hardware latency detector"); ++MODULE_VERSION(VERSION); ++ ++/* Module parameters */ ++ ++static int debug; ++static int enabled; ++static int threshold; ++ ++module_param(debug, int, 0); /* enable debug */ ++module_param(enabled, int, 0); /* enable detector */ ++module_param(threshold, int, 0); /* latency threshold */ ++ ++/* Buffering and sampling */ ++ ++static struct ring_buffer *ring_buffer; /* sample buffer */ ++static DEFINE_MUTEX(ring_buffer_mutex); /* lock changes */ ++static unsigned long buf_size = BUF_SIZE_DEFAULT; ++static struct task_struct *kthread; /* sampling thread */ ++ ++/* DebugFS filesystem entries */ ++ ++static struct dentry *debug_dir; /* debugfs directory */ ++static struct dentry *debug_max; /* maximum TSC delta */ ++static struct dentry *debug_count; /* total detect count */ ++static struct dentry *debug_sample_width; /* sample width us */ ++static struct dentry *debug_sample_window; /* sample window us */ ++static struct dentry *debug_sample; /* raw samples us */ ++static struct dentry *debug_threshold; /* threshold us */ ++static struct dentry *debug_enable; /* enable/disable */ ++ ++/* Individual samples and global state */ ++ ++struct sample; /* latency sample */ ++struct data; /* Global state */ ++ ++/* Sampling functions */ ++static int __buffer_add_sample(struct sample *sample); ++static struct sample *buffer_get_sample(struct sample *sample); ++ ++/* Threading and state */ ++static int kthread_fn(void *unused); ++static int start_kthread(void); ++static int stop_kthread(void); ++static void __reset_stats(void); ++static int init_stats(void); ++ ++/* Debugfs interface */ ++static ssize_t simple_data_read(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos, const u64 *entry); ++static ssize_t simple_data_write(struct file *filp, const char __user *ubuf, ++ size_t cnt, loff_t *ppos, u64 *entry); ++static int debug_sample_fopen(struct inode *inode, struct file *filp); ++static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos); ++static int debug_sample_release(struct inode *inode, struct file *filp); ++static int debug_enable_fopen(struct inode *inode, struct file *filp); ++static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos); ++static ssize_t debug_enable_fwrite(struct file *file, ++ const char __user *user_buffer, ++ size_t user_size, loff_t *offset); ++ ++/* Initialization functions */ ++static int init_debugfs(void); ++static void free_debugfs(void); ++static int detector_init(void); ++static void detector_exit(void); ++ ++/* Individual latency samples are stored here when detected and packed into ++ * the ring_buffer circular buffer, where they are overwritten when ++ * more than buf_size/sizeof(sample) samples are received. */ ++struct sample { ++ u64 seqnum; /* unique sequence */ ++ u64 duration; /* ktime delta */ ++ u64 outer_duration; /* ktime delta (outer loop) */ ++ struct timespec timestamp; /* wall time */ ++ unsigned long lost; ++}; ++ ++/* keep the global state somewhere. */ ++static struct data { ++ ++ struct mutex lock; /* protect changes */ ++ ++ u64 count; /* total since reset */ ++ u64 max_sample; /* max hardware latency */ ++ u64 threshold; /* sample threshold level */ ++ ++ u64 sample_window; /* total sampling window (on+off) */ ++ u64 sample_width; /* active sampling portion of window */ ++ ++ atomic_t sample_open; /* whether the sample file is open */ ++ ++ wait_queue_head_t wq; /* waitqeue for new sample values */ ++ ++} data; ++ ++/** ++ * __buffer_add_sample - add a new latency sample recording to the ring buffer ++ * @sample: The new latency sample value ++ * ++ * This receives a new latency sample and records it in a global ring buffer. ++ * No additional locking is used in this case. ++ */ ++static int __buffer_add_sample(struct sample *sample) ++{ ++ return ring_buffer_write(ring_buffer, ++ sizeof(struct sample), sample); ++} ++ ++/** ++ * buffer_get_sample - remove a hardware latency sample from the ring buffer ++ * @sample: Pre-allocated storage for the sample ++ * ++ * This retrieves a hardware latency sample from the global circular buffer ++ */ ++static struct sample *buffer_get_sample(struct sample *sample) ++{ ++ struct ring_buffer_event *e = NULL; ++ struct sample *s = NULL; ++ unsigned int cpu = 0; ++ ++ if (!sample) ++ return NULL; ++ ++ mutex_lock(&ring_buffer_mutex); ++ for_each_online_cpu(cpu) { ++ e = ring_buffer_consume(ring_buffer, cpu, NULL, &sample->lost); ++ if (e) ++ break; ++ } ++ ++ if (e) { ++ s = ring_buffer_event_data(e); ++ memcpy(sample, s, sizeof(struct sample)); ++ } else ++ sample = NULL; ++ mutex_unlock(&ring_buffer_mutex); ++ ++ return sample; ++} ++ ++#ifndef CONFIG_TRACING ++#define time_type ktime_t ++#define time_get() ktime_get() ++#define time_to_us(x) ktime_to_us(x) ++#define time_sub(a, b) ktime_sub(a, b) ++#define init_time(a, b) (a).tv64 = b ++#define time_u64(a) ((a).tv64) ++#else ++#define time_type u64 ++#define time_get() trace_clock_local() ++#define time_to_us(x) div_u64(x, 1000) ++#define time_sub(a, b) ((a) - (b)) ++#define init_time(a, b) (a = b) ++#define time_u64(a) a ++#endif ++/** ++ * get_sample - sample the CPU TSC and look for likely hardware latencies ++ * ++ * Used to repeatedly capture the CPU TSC (or similar), looking for potential ++ * hardware-induced latency. Called with interrupts disabled and with ++ * data.lock held. ++ */ ++static int get_sample(void) ++{ ++ time_type start, t1, t2, last_t2; ++ s64 diff, total = 0; ++ u64 sample = 0; ++ u64 outer_sample = 0; ++ int ret = -1; ++ ++ init_time(last_t2, 0); ++ start = time_get(); /* start timestamp */ ++ ++ do { ++ ++ t1 = time_get(); /* we'll look for a discontinuity */ ++ t2 = time_get(); ++ ++ if (time_u64(last_t2)) { ++ /* Check the delta from outer loop (t2 to next t1) */ ++ diff = time_to_us(time_sub(t1, last_t2)); ++ /* This shouldn't happen */ ++ if (diff < 0) { ++ pr_err(BANNER "time running backwards\n"); ++ goto out; ++ } ++ if (diff > outer_sample) ++ outer_sample = diff; ++ } ++ last_t2 = t2; ++ ++ total = time_to_us(time_sub(t2, start)); /* sample width */ ++ ++ /* This checks the inner loop (t1 to t2) */ ++ diff = time_to_us(time_sub(t2, t1)); /* current diff */ ++ ++ /* This shouldn't happen */ ++ if (diff < 0) { ++ pr_err(BANNER "time running backwards\n"); ++ goto out; ++ } ++ ++ if (diff > sample) ++ sample = diff; /* only want highest value */ ++ ++ } while (total <= data.sample_width); ++ ++ ret = 0; ++ ++ /* If we exceed the threshold value, we have found a hardware latency */ ++ if (sample > data.threshold || outer_sample > data.threshold) { ++ struct sample s; ++ ++ ret = 1; ++ ++ data.count++; ++ s.seqnum = data.count; ++ s.duration = sample; ++ s.outer_duration = outer_sample; ++ s.timestamp = CURRENT_TIME; ++ __buffer_add_sample(&s); ++ ++ /* Keep a running maximum ever recorded hardware latency */ ++ if (sample > data.max_sample) ++ data.max_sample = sample; ++ } ++ ++out: ++ return ret; ++} ++ ++/* ++ * kthread_fn - The CPU time sampling/hardware latency detection kernel thread ++ * @unused: A required part of the kthread API. ++ * ++ * Used to periodically sample the CPU TSC via a call to get_sample. We ++ * disable interrupts, which does (intentionally) introduce latency since we ++ * need to ensure nothing else might be running (and thus pre-empting). ++ * Obviously this should never be used in production environments. ++ * ++ * Currently this runs on which ever CPU it was scheduled on, but most ++ * real-worald hardware latency situations occur across several CPUs, ++ * but we might later generalize this if we find there are any actualy ++ * systems with alternate SMI delivery or other hardware latencies. ++ */ ++static int kthread_fn(void *unused) ++{ ++ int ret; ++ u64 interval; ++ ++ while (!kthread_should_stop()) { ++ ++ mutex_lock(&data.lock); ++ ++ local_irq_disable(); ++ ret = get_sample(); ++ local_irq_enable(); ++ ++ if (ret > 0) ++ wake_up(&data.wq); /* wake up reader(s) */ ++ ++ interval = data.sample_window - data.sample_width; ++ do_div(interval, USEC_PER_MSEC); /* modifies interval value */ ++ ++ mutex_unlock(&data.lock); ++ ++ if (msleep_interruptible(interval)) ++ break; ++ } ++ ++ return 0; ++} ++ ++/** ++ * start_kthread - Kick off the hardware latency sampling/detector kthread ++ * ++ * This starts a kernel thread that will sit and sample the CPU timestamp ++ * counter (TSC or similar) and look for potential hardware latencies. ++ */ ++static int start_kthread(void) ++{ ++ kthread = kthread_run(kthread_fn, NULL, ++ DRVNAME); ++ if (IS_ERR(kthread)) { ++ pr_err(BANNER "could not start sampling thread\n"); ++ enabled = 0; ++ return -ENOMEM; ++ } ++ ++ return 0; ++} ++ ++/** ++ * stop_kthread - Inform the hardware latency samping/detector kthread to stop ++ * ++ * This kicks the running hardware latency sampling/detector kernel thread and ++ * tells it to stop sampling now. Use this on unload and at system shutdown. ++ */ ++static int stop_kthread(void) ++{ ++ int ret; ++ ++ ret = kthread_stop(kthread); ++ ++ return ret; ++} ++ ++/** ++ * __reset_stats - Reset statistics for the hardware latency detector ++ * ++ * We use data to store various statistics and global state. We call this ++ * function in order to reset those when "enable" is toggled on or off, and ++ * also at initialization. Should be called with data.lock held. ++ */ ++static void __reset_stats(void) ++{ ++ data.count = 0; ++ data.max_sample = 0; ++ ring_buffer_reset(ring_buffer); /* flush out old sample entries */ ++} ++ ++/** ++ * init_stats - Setup global state statistics for the hardware latency detector ++ * ++ * We use data to store various statistics and global state. We also use ++ * a global ring buffer (ring_buffer) to keep raw samples of detected hardware ++ * induced system latencies. This function initializes these structures and ++ * allocates the global ring buffer also. ++ */ ++static int init_stats(void) ++{ ++ int ret = -ENOMEM; ++ ++ mutex_init(&data.lock); ++ init_waitqueue_head(&data.wq); ++ atomic_set(&data.sample_open, 0); ++ ++ ring_buffer = ring_buffer_alloc(buf_size, BUF_FLAGS); ++ ++ if (WARN(!ring_buffer, KERN_ERR BANNER ++ "failed to allocate ring buffer!\n")) ++ goto out; ++ ++ __reset_stats(); ++ data.threshold = threshold ?: DEFAULT_LAT_THRESHOLD; /* threshold us */ ++ data.sample_window = DEFAULT_SAMPLE_WINDOW; /* window us */ ++ data.sample_width = DEFAULT_SAMPLE_WIDTH; /* width us */ ++ ++ ret = 0; ++ ++out: ++ return ret; ++ ++} ++ ++/* ++ * simple_data_read - Wrapper read function for global state debugfs entries ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * @entry: The entry to read from ++ * ++ * This function provides a generic read implementation for the global state ++ * "data" structure debugfs filesystem entries. It would be nice to use ++ * simple_attr_read directly, but we need to make sure that the data.lock ++ * is held during the actual read. ++ */ ++static ssize_t simple_data_read(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos, const u64 *entry) ++{ ++ char buf[U64STR_SIZE]; ++ u64 val = 0; ++ int len = 0; ++ ++ memset(buf, 0, sizeof(buf)); ++ ++ if (!entry) ++ return -EFAULT; ++ ++ mutex_lock(&data.lock); ++ val = *entry; ++ mutex_unlock(&data.lock); ++ ++ len = snprintf(buf, sizeof(buf), "%llu\n", (unsigned long long)val); ++ ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, len); ++ ++} ++ ++/* ++ * simple_data_write - Wrapper write function for global state debugfs entries ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to write value from ++ * @cnt: The maximum number of bytes to write ++ * @ppos: The current "file" position ++ * @entry: The entry to write to ++ * ++ * This function provides a generic write implementation for the global state ++ * "data" structure debugfs filesystem entries. It would be nice to use ++ * simple_attr_write directly, but we need to make sure that the data.lock ++ * is held during the actual write. ++ */ ++static ssize_t simple_data_write(struct file *filp, const char __user *ubuf, ++ size_t cnt, loff_t *ppos, u64 *entry) ++{ ++ char buf[U64STR_SIZE]; ++ int csize = min(cnt, sizeof(buf)); ++ u64 val = 0; ++ int err = 0; ++ ++ memset(buf, '\0', sizeof(buf)); ++ if (copy_from_user(buf, ubuf, csize)) ++ return -EFAULT; ++ ++ buf[U64STR_SIZE-1] = '\0'; /* just in case */ ++ err = kstrtoull(buf, 10, &val); ++ if (err) ++ return -EINVAL; ++ ++ mutex_lock(&data.lock); ++ *entry = val; ++ mutex_unlock(&data.lock); ++ ++ return csize; ++} ++ ++/** ++ * debug_count_fopen - Open function for "count" debugfs entry ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "count" debugfs ++ * interface to the hardware latency detector. ++ */ ++static int debug_count_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_count_fread - Read function for "count" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "count" debugfs ++ * interface to the hardware latency detector. Can be used to read the ++ * number of latency readings exceeding the configured threshold since ++ * the detector was last reset (e.g. by writing a zero into "count"). ++ */ ++static ssize_t debug_count_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ return simple_data_read(filp, ubuf, cnt, ppos, &data.count); ++} ++ ++/** ++ * debug_count_fwrite - Write function for "count" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "count" debugfs ++ * interface to the hardware latency detector. Can be used to write a ++ * desired value, especially to zero the total count. ++ */ ++static ssize_t debug_count_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ return simple_data_write(filp, ubuf, cnt, ppos, &data.count); ++} ++ ++/** ++ * debug_enable_fopen - Dummy open function for "enable" debugfs interface ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "enable" debugfs ++ * interface to the hardware latency detector. ++ */ ++static int debug_enable_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_enable_fread - Read function for "enable" debugfs interface ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "enable" debugfs ++ * interface to the hardware latency detector. Can be used to determine ++ * whether the detector is currently enabled ("0\n" or "1\n" returned). ++ */ ++static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ char buf[4]; ++ ++ if ((cnt < sizeof(buf)) || (*ppos)) ++ return 0; ++ ++ buf[0] = enabled ? '1' : '0'; ++ buf[1] = '\n'; ++ buf[2] = '\0'; ++ if (copy_to_user(ubuf, buf, strlen(buf))) ++ return -EFAULT; ++ return *ppos = strlen(buf); ++} ++ ++/** ++ * debug_enable_fwrite - Write function for "enable" debugfs interface ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "enable" debugfs ++ * interface to the hardware latency detector. Can be used to enable or ++ * disable the detector, which will have the side-effect of possibly ++ * also resetting the global stats and kicking off the measuring ++ * kthread (on an enable) or the converse (upon a disable). ++ */ ++static ssize_t debug_enable_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ char buf[4]; ++ int csize = min(cnt, sizeof(buf)); ++ long val = 0; ++ int err = 0; ++ ++ memset(buf, '\0', sizeof(buf)); ++ if (copy_from_user(buf, ubuf, csize)) ++ return -EFAULT; ++ ++ buf[sizeof(buf)-1] = '\0'; /* just in case */ ++ err = kstrtoul(buf, 10, &val); ++ if (0 != err) ++ return -EINVAL; ++ ++ if (val) { ++ if (enabled) ++ goto unlock; ++ enabled = 1; ++ __reset_stats(); ++ if (start_kthread()) ++ return -EFAULT; ++ } else { ++ if (!enabled) ++ goto unlock; ++ enabled = 0; ++ err = stop_kthread(); ++ if (err) { ++ pr_err(BANNER "cannot stop kthread\n"); ++ return -EFAULT; ++ } ++ wake_up(&data.wq); /* reader(s) should return */ ++ } ++unlock: ++ return csize; ++} ++ ++/** ++ * debug_max_fopen - Open function for "max" debugfs entry ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "max" debugfs ++ * interface to the hardware latency detector. ++ */ ++static int debug_max_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_max_fread - Read function for "max" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "max" debugfs ++ * interface to the hardware latency detector. Can be used to determine ++ * the maximum latency value observed since it was last reset. ++ */ ++static ssize_t debug_max_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ return simple_data_read(filp, ubuf, cnt, ppos, &data.max_sample); ++} ++ ++/** ++ * debug_max_fwrite - Write function for "max" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "max" debugfs ++ * interface to the hardware latency detector. Can be used to reset the ++ * maximum or set it to some other desired value - if, then, subsequent ++ * measurements exceed this value, the maximum will be updated. ++ */ ++static ssize_t debug_max_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ return simple_data_write(filp, ubuf, cnt, ppos, &data.max_sample); ++} ++ ++ ++/** ++ * debug_sample_fopen - An open function for "sample" debugfs interface ++ * @inode: The in-kernel inode representation of this debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function handles opening the "sample" file within the hardware ++ * latency detector debugfs directory interface. This file is used to read ++ * raw samples from the global ring_buffer and allows the user to see a ++ * running latency history. Can be opened blocking or non-blocking, ++ * affecting whether it behaves as a buffer read pipe, or does not. ++ * Implements simple locking to prevent multiple simultaneous use. ++ */ ++static int debug_sample_fopen(struct inode *inode, struct file *filp) ++{ ++ if (!atomic_add_unless(&data.sample_open, 1, 1)) ++ return -EBUSY; ++ else ++ return 0; ++} ++ ++/** ++ * debug_sample_fread - A read function for "sample" debugfs interface ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that will contain the samples read ++ * @cnt: The maximum bytes to read from the debugfs "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function handles reading from the "sample" file within the hardware ++ * latency detector debugfs directory interface. This file is used to read ++ * raw samples from the global ring_buffer and allows the user to see a ++ * running latency history. By default this will block pending a new ++ * value written into the sample buffer, unless there are already a ++ * number of value(s) waiting in the buffer, or the sample file was ++ * previously opened in a non-blocking mode of operation. ++ */ ++static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ int len = 0; ++ char buf[64]; ++ struct sample *sample = NULL; ++ ++ if (!enabled) ++ return 0; ++ ++ sample = kzalloc(sizeof(struct sample), GFP_KERNEL); ++ if (!sample) ++ return -ENOMEM; ++ ++ while (!buffer_get_sample(sample)) { ++ ++ DEFINE_WAIT(wait); ++ ++ if (filp->f_flags & O_NONBLOCK) { ++ len = -EAGAIN; ++ goto out; ++ } ++ ++ prepare_to_wait(&data.wq, &wait, TASK_INTERRUPTIBLE); ++ schedule(); ++ finish_wait(&data.wq, &wait); ++ ++ if (signal_pending(current)) { ++ len = -EINTR; ++ goto out; ++ } ++ ++ if (!enabled) { /* enable was toggled */ ++ len = 0; ++ goto out; ++ } ++ } ++ ++ len = snprintf(buf, sizeof(buf), "%010lu.%010lu\t%llu\t%llu\n", ++ sample->timestamp.tv_sec, ++ sample->timestamp.tv_nsec, ++ sample->duration, ++ sample->outer_duration); ++ ++ ++ /* handling partial reads is more trouble than it's worth */ ++ if (len > cnt) ++ goto out; ++ ++ if (copy_to_user(ubuf, buf, len)) ++ len = -EFAULT; ++ ++out: ++ kfree(sample); ++ return len; ++} ++ ++/** ++ * debug_sample_release - Release function for "sample" debugfs interface ++ * @inode: The in-kernel inode represenation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function completes the close of the debugfs interface "sample" file. ++ * Frees the sample_open "lock" so that other users may open the interface. ++ */ ++static int debug_sample_release(struct inode *inode, struct file *filp) ++{ ++ atomic_dec(&data.sample_open); ++ ++ return 0; ++} ++ ++/** ++ * debug_threshold_fopen - Open function for "threshold" debugfs entry ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "threshold" debugfs ++ * interface to the hardware latency detector. ++ */ ++static int debug_threshold_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_threshold_fread - Read function for "threshold" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "threshold" debugfs ++ * interface to the hardware latency detector. It can be used to determine ++ * the current threshold level at which a latency will be recorded in the ++ * global ring buffer, typically on the order of 10us. ++ */ ++static ssize_t debug_threshold_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ return simple_data_read(filp, ubuf, cnt, ppos, &data.threshold); ++} ++ ++/** ++ * debug_threshold_fwrite - Write function for "threshold" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "threshold" debugfs ++ * interface to the hardware latency detector. It can be used to configure ++ * the threshold level at which any subsequently detected latencies will ++ * be recorded into the global ring buffer. ++ */ ++static ssize_t debug_threshold_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ int ret; ++ ++ ret = simple_data_write(filp, ubuf, cnt, ppos, &data.threshold); ++ ++ if (enabled) ++ wake_up_process(kthread); ++ ++ return ret; ++} ++ ++/** ++ * debug_width_fopen - Open function for "width" debugfs entry ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "width" debugfs ++ * interface to the hardware latency detector. ++ */ ++static int debug_width_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_width_fread - Read function for "width" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "width" debugfs ++ * interface to the hardware latency detector. It can be used to determine ++ * for how many us of the total window us we will actively sample for any ++ * hardware-induced latecy periods. Obviously, it is not possible to ++ * sample constantly and have the system respond to a sample reader, or, ++ * worse, without having the system appear to have gone out to lunch. ++ */ ++static ssize_t debug_width_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_width); ++} ++ ++/** ++ * debug_width_fwrite - Write function for "width" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "width" debugfs ++ * interface to the hardware latency detector. It can be used to configure ++ * for how many us of the total window us we will actively sample for any ++ * hardware-induced latency periods. Obviously, it is not possible to ++ * sample constantly and have the system respond to a sample reader, or, ++ * worse, without having the system appear to have gone out to lunch. It ++ * is enforced that width is less that the total window size. ++ */ ++static ssize_t debug_width_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ char buf[U64STR_SIZE]; ++ int csize = min(cnt, sizeof(buf)); ++ u64 val = 0; ++ int err = 0; ++ ++ memset(buf, '\0', sizeof(buf)); ++ if (copy_from_user(buf, ubuf, csize)) ++ return -EFAULT; ++ ++ buf[U64STR_SIZE-1] = '\0'; /* just in case */ ++ err = kstrtoull(buf, 10, &val); ++ if (0 != err) ++ return -EINVAL; ++ ++ mutex_lock(&data.lock); ++ if (val < data.sample_window) ++ data.sample_width = val; ++ else { ++ mutex_unlock(&data.lock); ++ return -EINVAL; ++ } ++ mutex_unlock(&data.lock); ++ ++ if (enabled) ++ wake_up_process(kthread); ++ ++ return csize; ++} ++ ++/** ++ * debug_window_fopen - Open function for "window" debugfs entry ++ * @inode: The in-kernel inode representation of the debugfs "file" ++ * @filp: The active open file structure for the debugfs "file" ++ * ++ * This function provides an open implementation for the "window" debugfs ++ * interface to the hardware latency detector. The window is the total time ++ * in us that will be considered one sample period. Conceptually, windows ++ * occur back-to-back and contain a sample width period during which ++ * actual sampling occurs. ++ */ ++static int debug_window_fopen(struct inode *inode, struct file *filp) ++{ ++ return 0; ++} ++ ++/** ++ * debug_window_fread - Read function for "window" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The userspace provided buffer to read value into ++ * @cnt: The maximum number of bytes to read ++ * @ppos: The current "file" position ++ * ++ * This function provides a read implementation for the "window" debugfs ++ * interface to the hardware latency detector. The window is the total time ++ * in us that will be considered one sample period. Conceptually, windows ++ * occur back-to-back and contain a sample width period during which ++ * actual sampling occurs. Can be used to read the total window size. ++ */ ++static ssize_t debug_window_fread(struct file *filp, char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_window); ++} ++ ++/** ++ * debug_window_fwrite - Write function for "window" debugfs entry ++ * @filp: The active open file structure for the debugfs "file" ++ * @ubuf: The user buffer that contains the value to write ++ * @cnt: The maximum number of bytes to write to "file" ++ * @ppos: The current position in the debugfs "file" ++ * ++ * This function provides a write implementation for the "window" debufds ++ * interface to the hardware latency detetector. The window is the total time ++ * in us that will be considered one sample period. Conceptually, windows ++ * occur back-to-back and contain a sample width period during which ++ * actual sampling occurs. Can be used to write a new total window size. It ++ * is enfoced that any value written must be greater than the sample width ++ * size, or an error results. ++ */ ++static ssize_t debug_window_fwrite(struct file *filp, ++ const char __user *ubuf, ++ size_t cnt, ++ loff_t *ppos) ++{ ++ char buf[U64STR_SIZE]; ++ int csize = min(cnt, sizeof(buf)); ++ u64 val = 0; ++ int err = 0; ++ ++ memset(buf, '\0', sizeof(buf)); ++ if (copy_from_user(buf, ubuf, csize)) ++ return -EFAULT; ++ ++ buf[U64STR_SIZE-1] = '\0'; /* just in case */ ++ err = kstrtoull(buf, 10, &val); ++ if (0 != err) ++ return -EINVAL; ++ ++ mutex_lock(&data.lock); ++ if (data.sample_width < val) ++ data.sample_window = val; ++ else { ++ mutex_unlock(&data.lock); ++ return -EINVAL; ++ } ++ mutex_unlock(&data.lock); ++ ++ return csize; ++} ++ ++/* ++ * Function pointers for the "count" debugfs file operations ++ */ ++static const struct file_operations count_fops = { ++ .open = debug_count_fopen, ++ .read = debug_count_fread, ++ .write = debug_count_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "enable" debugfs file operations ++ */ ++static const struct file_operations enable_fops = { ++ .open = debug_enable_fopen, ++ .read = debug_enable_fread, ++ .write = debug_enable_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "max" debugfs file operations ++ */ ++static const struct file_operations max_fops = { ++ .open = debug_max_fopen, ++ .read = debug_max_fread, ++ .write = debug_max_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "sample" debugfs file operations ++ */ ++static const struct file_operations sample_fops = { ++ .open = debug_sample_fopen, ++ .read = debug_sample_fread, ++ .release = debug_sample_release, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "threshold" debugfs file operations ++ */ ++static const struct file_operations threshold_fops = { ++ .open = debug_threshold_fopen, ++ .read = debug_threshold_fread, ++ .write = debug_threshold_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "width" debugfs file operations ++ */ ++static const struct file_operations width_fops = { ++ .open = debug_width_fopen, ++ .read = debug_width_fread, ++ .write = debug_width_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/* ++ * Function pointers for the "window" debugfs file operations ++ */ ++static const struct file_operations window_fops = { ++ .open = debug_window_fopen, ++ .read = debug_window_fread, ++ .write = debug_window_fwrite, ++ .owner = THIS_MODULE, ++}; ++ ++/** ++ * init_debugfs - A function to initialize the debugfs interface files ++ * ++ * This function creates entries in debugfs for "hwlat_detector", including ++ * files to read values from the detector, current samples, and the ++ * maximum sample that has been captured since the hardware latency ++ * dectector was started. ++ */ ++static int init_debugfs(void) ++{ ++ int ret = -ENOMEM; ++ ++ debug_dir = debugfs_create_dir(DRVNAME, NULL); ++ if (!debug_dir) ++ goto err_debug_dir; ++ ++ debug_sample = debugfs_create_file("sample", 0444, ++ debug_dir, NULL, ++ &sample_fops); ++ if (!debug_sample) ++ goto err_sample; ++ ++ debug_count = debugfs_create_file("count", 0444, ++ debug_dir, NULL, ++ &count_fops); ++ if (!debug_count) ++ goto err_count; ++ ++ debug_max = debugfs_create_file("max", 0444, ++ debug_dir, NULL, ++ &max_fops); ++ if (!debug_max) ++ goto err_max; ++ ++ debug_sample_window = debugfs_create_file("window", 0644, ++ debug_dir, NULL, ++ &window_fops); ++ if (!debug_sample_window) ++ goto err_window; ++ ++ debug_sample_width = debugfs_create_file("width", 0644, ++ debug_dir, NULL, ++ &width_fops); ++ if (!debug_sample_width) ++ goto err_width; ++ ++ debug_threshold = debugfs_create_file("threshold", 0644, ++ debug_dir, NULL, ++ &threshold_fops); ++ if (!debug_threshold) ++ goto err_threshold; ++ ++ debug_enable = debugfs_create_file("enable", 0644, ++ debug_dir, &enabled, ++ &enable_fops); ++ if (!debug_enable) ++ goto err_enable; ++ ++ else { ++ ret = 0; ++ goto out; ++ } ++ ++err_enable: ++ debugfs_remove(debug_threshold); ++err_threshold: ++ debugfs_remove(debug_sample_width); ++err_width: ++ debugfs_remove(debug_sample_window); ++err_window: ++ debugfs_remove(debug_max); ++err_max: ++ debugfs_remove(debug_count); ++err_count: ++ debugfs_remove(debug_sample); ++err_sample: ++ debugfs_remove(debug_dir); ++err_debug_dir: ++out: ++ return ret; ++} ++ ++/** ++ * free_debugfs - A function to cleanup the debugfs file interface ++ */ ++static void free_debugfs(void) ++{ ++ /* could also use a debugfs_remove_recursive */ ++ debugfs_remove(debug_enable); ++ debugfs_remove(debug_threshold); ++ debugfs_remove(debug_sample_width); ++ debugfs_remove(debug_sample_window); ++ debugfs_remove(debug_max); ++ debugfs_remove(debug_count); ++ debugfs_remove(debug_sample); ++ debugfs_remove(debug_dir); ++} ++ ++/** ++ * detector_init - Standard module initialization code ++ */ ++static int detector_init(void) ++{ ++ int ret = -ENOMEM; ++ ++ pr_info(BANNER "version %s\n", VERSION); ++ ++ ret = init_stats(); ++ if (0 != ret) ++ goto out; ++ ++ ret = init_debugfs(); ++ if (0 != ret) ++ goto err_stats; ++ ++ if (enabled) ++ ret = start_kthread(); ++ ++ goto out; ++ ++err_stats: ++ ring_buffer_free(ring_buffer); ++out: ++ return ret; ++ ++} ++ ++/** ++ * detector_exit - Standard module cleanup code ++ */ ++static void detector_exit(void) ++{ ++ int err; ++ ++ if (enabled) { ++ enabled = 0; ++ err = stop_kthread(); ++ if (err) ++ pr_err(BANNER "cannot stop kthread\n"); ++ } ++ ++ free_debugfs(); ++ ring_buffer_free(ring_buffer); /* free up the ring buffer */ ++ ++} ++ ++module_init(detector_init); ++module_exit(detector_exit); +diff -Nur linux-4.1.13.orig/drivers/misc/Kconfig linux-4.1.13/drivers/misc/Kconfig +--- linux-4.1.13.orig/drivers/misc/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/misc/Kconfig 2015-11-29 09:23:09.561617368 +0100 +@@ -54,6 +54,7 @@ + config ATMEL_TCLIB + bool "Atmel AT32/AT91 Timer/Counter Library" + depends on (AVR32 || ARCH_AT91) ++ default y if PREEMPT_RT_FULL + help + Select this if you want a library to allocate the Timer/Counter + blocks found on many Atmel processors. This facilitates using +@@ -69,8 +70,7 @@ + are combined to make a single 32-bit timer. + + When GENERIC_CLOCKEVENTS is defined, the third timer channel +- may be used as a clock event device supporting oneshot mode +- (delays of up to two seconds) based on the 32 KiHz clock. ++ may be used as a clock event device supporting oneshot mode. + + config ATMEL_TCB_CLKSRC_BLOCK + int +@@ -84,6 +84,15 @@ + TC can be used for other purposes, such as PWM generation and + interval timing. + ++config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK ++ bool "TC Block use 32 KiHz clock" ++ depends on ATMEL_TCB_CLKSRC ++ default y if !PREEMPT_RT_FULL ++ help ++ Select this to use 32 KiHz base clock rate as TC block clock ++ source for clock events. ++ ++ + config DUMMY_IRQ + tristate "Dummy IRQ handler" + default n +@@ -113,6 +122,35 @@ + for information on the specific driver level and support statement + for your IBM server. + ++config HWLAT_DETECTOR ++ tristate "Testing module to detect hardware-induced latencies" ++ depends on DEBUG_FS ++ depends on RING_BUFFER ++ default m ++ ---help--- ++ A simple hardware latency detector. Use this module to detect ++ large latencies introduced by the behavior of the underlying ++ system firmware external to Linux. We do this using periodic ++ use of stop_machine to grab all available CPUs and measure ++ for unexplainable gaps in the CPU timestamp counter(s). By ++ default, the module is not enabled until the "enable" file ++ within the "hwlat_detector" debugfs directory is toggled. ++ ++ This module is often used to detect SMI (System Management ++ Interrupts) on x86 systems, though is not x86 specific. To ++ this end, we default to using a sample window of 1 second, ++ during which we will sample for 0.5 seconds. If an SMI or ++ similar event occurs during that time, it is recorded ++ into an 8K samples global ring buffer until retreived. ++ ++ WARNING: This software should never be enabled (it can be built ++ but should not be turned on after it is loaded) in a production ++ environment where high latencies are a concern since the ++ sampling mechanism actually introduces latencies for ++ regular tasks while the CPU(s) are being held. ++ ++ If unsure, say N ++ + config PHANTOM + tristate "Sensable PHANToM (PCI)" + depends on PCI +diff -Nur linux-4.1.13.orig/drivers/misc/Makefile linux-4.1.13/drivers/misc/Makefile +--- linux-4.1.13.orig/drivers/misc/Makefile 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/misc/Makefile 2015-11-29 09:23:09.561617368 +0100 +@@ -38,6 +38,7 @@ + obj-$(CONFIG_HMC6352) += hmc6352.o + obj-y += eeprom/ + obj-y += cb710/ ++obj-$(CONFIG_HWLAT_DETECTOR) += hwlat_detector.o + obj-$(CONFIG_SPEAR13XX_PCIE_GADGET) += spear13xx_pcie_gadget.o + obj-$(CONFIG_VMWARE_BALLOON) += vmw_balloon.o + obj-$(CONFIG_ARM_CHARLCD) += arm-charlcd.o +diff -Nur linux-4.1.13.orig/drivers/mmc/host/mmci.c linux-4.1.13/drivers/mmc/host/mmci.c +--- linux-4.1.13.orig/drivers/mmc/host/mmci.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/mmc/host/mmci.c 2015-11-29 09:23:09.561617368 +0100 +@@ -1155,15 +1155,12 @@ + struct sg_mapping_iter *sg_miter = &host->sg_miter; + struct variant_data *variant = host->variant; + void __iomem *base = host->base; +- unsigned long flags; + u32 status; + + status = readl(base + MMCISTATUS); + + dev_dbg(mmc_dev(host->mmc), "irq1 (pio) %08x\n", status); + +- local_irq_save(flags); +- + do { + unsigned int remain, len; + char *buffer; +@@ -1203,8 +1200,6 @@ + + sg_miter_stop(sg_miter); + +- local_irq_restore(flags); +- + /* + * If we have less than the fifo 'half-full' threshold to transfer, + * trigger a PIO interrupt as soon as any data is available. +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/3com/3c59x.c linux-4.1.13/drivers/net/ethernet/3com/3c59x.c +--- linux-4.1.13.orig/drivers/net/ethernet/3com/3c59x.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/3com/3c59x.c 2015-11-29 09:23:09.561617368 +0100 +@@ -842,9 +842,9 @@ + { + struct vortex_private *vp = netdev_priv(dev); + unsigned long flags; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + (vp->full_bus_master_rx ? boomerang_interrupt:vortex_interrupt)(dev->irq,dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + #endif + +@@ -1916,12 +1916,12 @@ + * Block interrupts because vortex_interrupt does a bare spin_lock() + */ + unsigned long flags; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (vp->full_bus_master_tx) + boomerang_interrupt(dev->irq, dev); + else + vortex_interrupt(dev->irq, dev); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + } + +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/atheros/atl1c/atl1c_main.c linux-4.1.13/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +--- linux-4.1.13.orig/drivers/net/ethernet/atheros/atl1c/atl1c_main.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/atheros/atl1c/atl1c_main.c 2015-11-29 09:23:09.561617368 +0100 +@@ -2213,11 +2213,7 @@ + } + + tpd_req = atl1c_cal_tpd_req(skb); +- if (!spin_trylock_irqsave(&adapter->tx_lock, flags)) { +- if (netif_msg_pktdata(adapter)) +- dev_info(&adapter->pdev->dev, "tx locked\n"); +- return NETDEV_TX_LOCKED; +- } ++ spin_lock_irqsave(&adapter->tx_lock, flags); + + if (atl1c_tpd_avail(adapter, type) < tpd_req) { + /* no enough descriptor, just stop queue */ +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/atheros/atl1e/atl1e_main.c linux-4.1.13/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +--- linux-4.1.13.orig/drivers/net/ethernet/atheros/atl1e/atl1e_main.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/atheros/atl1e/atl1e_main.c 2015-11-29 09:23:09.561617368 +0100 +@@ -1880,8 +1880,7 @@ + return NETDEV_TX_OK; + } + tpd_req = atl1e_cal_tdp_req(skb); +- if (!spin_trylock_irqsave(&adapter->tx_lock, flags)) +- return NETDEV_TX_LOCKED; ++ spin_lock_irqsave(&adapter->tx_lock, flags); + + if (atl1e_tpd_avail(adapter) < tpd_req) { + /* no enough descriptor, just stop queue */ +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/chelsio/cxgb/sge.c linux-4.1.13/drivers/net/ethernet/chelsio/cxgb/sge.c +--- linux-4.1.13.orig/drivers/net/ethernet/chelsio/cxgb/sge.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/chelsio/cxgb/sge.c 2015-11-29 09:23:09.569616840 +0100 +@@ -1664,8 +1664,7 @@ + struct cmdQ *q = &sge->cmdQ[qid]; + unsigned int credits, pidx, genbit, count, use_sched_skb = 0; + +- if (!spin_trylock(&q->lock)) +- return NETDEV_TX_LOCKED; ++ spin_lock(&q->lock); + + reclaim_completed_tx(sge, q); + +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/freescale/gianfar.c linux-4.1.13/drivers/net/ethernet/freescale/gianfar.c +--- linux-4.1.13.orig/drivers/net/ethernet/freescale/gianfar.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/freescale/gianfar.c 2015-11-29 09:23:09.569616840 +0100 +@@ -1540,7 +1540,7 @@ + + if (netif_running(ndev)) { + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + lock_tx_qs(priv); + + gfar_halt_nodisable(priv); +@@ -1556,7 +1556,7 @@ + gfar_write(®s->maccfg1, tempval); + + unlock_tx_qs(priv); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + disable_napi(priv); + +@@ -1598,7 +1598,7 @@ + /* Disable Magic Packet mode, in case something + * else woke us up. + */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + lock_tx_qs(priv); + + tempval = gfar_read(®s->maccfg2); +@@ -1608,7 +1608,7 @@ + gfar_start(priv); + + unlock_tx_qs(priv); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + netif_device_attach(ndev); + +@@ -3418,14 +3418,14 @@ + dev->stats.tx_dropped++; + atomic64_inc(&priv->extra_stats.tx_underrun); + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + lock_tx_qs(priv); + + /* Reactivate the Tx Queues */ + gfar_write(®s->tstat, gfargrp->tstat); + + unlock_tx_qs(priv); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + netif_dbg(priv, tx_err, dev, "Transmit Error\n"); + } +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/neterion/s2io.c linux-4.1.13/drivers/net/ethernet/neterion/s2io.c +--- linux-4.1.13.orig/drivers/net/ethernet/neterion/s2io.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/neterion/s2io.c 2015-11-29 09:23:09.569616840 +0100 +@@ -4084,12 +4084,7 @@ + [skb->priority & (MAX_TX_FIFOS - 1)]; + fifo = &mac_control->fifos[queue]; + +- if (do_spin_lock) +- spin_lock_irqsave(&fifo->tx_lock, flags); +- else { +- if (unlikely(!spin_trylock_irqsave(&fifo->tx_lock, flags))) +- return NETDEV_TX_LOCKED; +- } ++ spin_lock_irqsave(&fifo->tx_lock, flags); + + if (sp->config.multiq) { + if (__netif_subqueue_stopped(dev, fifo->fifo_no)) { +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c linux-4.1.13/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +--- linux-4.1.13.orig/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c 2015-11-29 09:23:09.569616840 +0100 +@@ -2137,10 +2137,8 @@ + struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring; + unsigned long flags; + +- if (!spin_trylock_irqsave(&tx_ring->tx_lock, flags)) { +- /* Collision - tell upper layer to requeue */ +- return NETDEV_TX_LOCKED; +- } ++ spin_lock_irqsave(&tx_ring->tx_lock, flags); ++ + if (unlikely(!PCH_GBE_DESC_UNUSED(tx_ring))) { + netif_stop_queue(netdev); + spin_unlock_irqrestore(&tx_ring->tx_lock, flags); +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/realtek/8139too.c linux-4.1.13/drivers/net/ethernet/realtek/8139too.c +--- linux-4.1.13.orig/drivers/net/ethernet/realtek/8139too.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/realtek/8139too.c 2015-11-29 09:23:09.569616840 +0100 +@@ -2229,7 +2229,7 @@ + struct rtl8139_private *tp = netdev_priv(dev); + const int irq = tp->pci_dev->irq; + +- disable_irq(irq); ++ disable_irq_nosync(irq); + rtl8139_interrupt(irq, dev); + enable_irq(irq); + } +diff -Nur linux-4.1.13.orig/drivers/net/ethernet/tehuti/tehuti.c linux-4.1.13/drivers/net/ethernet/tehuti/tehuti.c +--- linux-4.1.13.orig/drivers/net/ethernet/tehuti/tehuti.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/ethernet/tehuti/tehuti.c 2015-11-29 09:23:09.569616840 +0100 +@@ -1629,13 +1629,8 @@ + unsigned long flags; + + ENTER; +- local_irq_save(flags); +- if (!spin_trylock(&priv->tx_lock)) { +- local_irq_restore(flags); +- DBG("%s[%s]: TX locked, returning NETDEV_TX_LOCKED\n", +- BDX_DRV_NAME, ndev->name); +- return NETDEV_TX_LOCKED; +- } ++ ++ spin_lock_irqsave(&priv->tx_lock, flags); + + /* build tx descriptor */ + BDX_ASSERT(f->m.wptr >= f->m.memsz); /* started with valid wptr */ +diff -Nur linux-4.1.13.orig/drivers/net/rionet.c linux-4.1.13/drivers/net/rionet.c +--- linux-4.1.13.orig/drivers/net/rionet.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/rionet.c 2015-11-29 09:23:09.581616045 +0100 +@@ -174,11 +174,7 @@ + unsigned long flags; + int add_num = 1; + +- local_irq_save(flags); +- if (!spin_trylock(&rnet->tx_lock)) { +- local_irq_restore(flags); +- return NETDEV_TX_LOCKED; +- } ++ spin_lock_irqsave(&rnet->tx_lock, flags); + + if (is_multicast_ether_addr(eth->h_dest)) + add_num = nets[rnet->mport->id].nact; +diff -Nur linux-4.1.13.orig/drivers/net/wireless/orinoco/orinoco_usb.c linux-4.1.13/drivers/net/wireless/orinoco/orinoco_usb.c +--- linux-4.1.13.orig/drivers/net/wireless/orinoco/orinoco_usb.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/net/wireless/orinoco/orinoco_usb.c 2015-11-29 09:23:09.581616045 +0100 +@@ -697,7 +697,7 @@ + while (!ctx->done.done && msecs--) + udelay(1000); + } else { +- wait_event_interruptible(ctx->done.wait, ++ swait_event_interruptible(ctx->done.wait, + ctx->done.done); + } + break; +diff -Nur linux-4.1.13.orig/drivers/pci/access.c linux-4.1.13/drivers/pci/access.c +--- linux-4.1.13.orig/drivers/pci/access.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/pci/access.c 2015-11-29 09:23:09.581616045 +0100 +@@ -561,7 +561,7 @@ + WARN_ON(!dev->block_cfg_access); + + dev->block_cfg_access = 0; +- wake_up_all(&pci_cfg_wait); ++ wake_up_all_locked(&pci_cfg_wait); + raw_spin_unlock_irqrestore(&pci_lock, flags); + } + EXPORT_SYMBOL_GPL(pci_cfg_access_unlock); +diff -Nur linux-4.1.13.orig/drivers/scsi/fcoe/fcoe.c linux-4.1.13/drivers/scsi/fcoe/fcoe.c +--- linux-4.1.13.orig/drivers/scsi/fcoe/fcoe.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/scsi/fcoe/fcoe.c 2015-11-29 09:23:09.581616045 +0100 +@@ -1287,7 +1287,7 @@ + struct sk_buff *skb; + #ifdef CONFIG_SMP + struct fcoe_percpu_s *p0; +- unsigned targ_cpu = get_cpu(); ++ unsigned targ_cpu = get_cpu_light(); + #endif /* CONFIG_SMP */ + + FCOE_DBG("Destroying receive thread for CPU %d\n", cpu); +@@ -1343,7 +1343,7 @@ + kfree_skb(skb); + spin_unlock_bh(&p->fcoe_rx_list.lock); + } +- put_cpu(); ++ put_cpu_light(); + #else + /* + * This a non-SMP scenario where the singular Rx thread is +@@ -1567,11 +1567,11 @@ + static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) + { + struct fcoe_percpu_s *fps; +- int rc; ++ int rc, cpu = get_cpu_light(); + +- fps = &get_cpu_var(fcoe_percpu); ++ fps = &per_cpu(fcoe_percpu, cpu); + rc = fcoe_get_paged_crc_eof(skb, tlen, fps); +- put_cpu_var(fcoe_percpu); ++ put_cpu_light(); + + return rc; + } +@@ -1767,11 +1767,11 @@ + return 0; + } + +- stats = per_cpu_ptr(lport->stats, get_cpu()); ++ stats = per_cpu_ptr(lport->stats, get_cpu_light()); + stats->InvalidCRCCount++; + if (stats->InvalidCRCCount < 5) + printk(KERN_WARNING "fcoe: dropping frame with CRC error\n"); +- put_cpu(); ++ put_cpu_light(); + return -EINVAL; + } + +@@ -1847,13 +1847,13 @@ + goto drop; + + if (!fcoe_filter_frames(lport, fp)) { +- put_cpu(); ++ put_cpu_light(); + fc_exch_recv(lport, fp); + return; + } + drop: + stats->ErrorFrames++; +- put_cpu(); ++ put_cpu_light(); + kfree_skb(skb); + } + +diff -Nur linux-4.1.13.orig/drivers/scsi/fcoe/fcoe_ctlr.c linux-4.1.13/drivers/scsi/fcoe/fcoe_ctlr.c +--- linux-4.1.13.orig/drivers/scsi/fcoe/fcoe_ctlr.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/scsi/fcoe/fcoe_ctlr.c 2015-11-29 09:23:09.581616045 +0100 +@@ -831,7 +831,7 @@ + + INIT_LIST_HEAD(&del_list); + +- stats = per_cpu_ptr(fip->lp->stats, get_cpu()); ++ stats = per_cpu_ptr(fip->lp->stats, get_cpu_light()); + + list_for_each_entry_safe(fcf, next, &fip->fcfs, list) { + deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2; +@@ -867,7 +867,7 @@ + sel_time = fcf->time; + } + } +- put_cpu(); ++ put_cpu_light(); + + list_for_each_entry_safe(fcf, next, &del_list, list) { + /* Removes fcf from current list */ +diff -Nur linux-4.1.13.orig/drivers/scsi/libfc/fc_exch.c linux-4.1.13/drivers/scsi/libfc/fc_exch.c +--- linux-4.1.13.orig/drivers/scsi/libfc/fc_exch.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/scsi/libfc/fc_exch.c 2015-11-29 09:23:09.581616045 +0100 +@@ -814,10 +814,10 @@ + } + memset(ep, 0, sizeof(*ep)); + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + pool = per_cpu_ptr(mp->pool, cpu); + spin_lock_bh(&pool->lock); +- put_cpu(); ++ put_cpu_light(); + + /* peek cache of free slot */ + if (pool->left != FC_XID_UNKNOWN) { +diff -Nur linux-4.1.13.orig/drivers/scsi/libsas/sas_ata.c linux-4.1.13/drivers/scsi/libsas/sas_ata.c +--- linux-4.1.13.orig/drivers/scsi/libsas/sas_ata.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/scsi/libsas/sas_ata.c 2015-11-29 09:23:09.581616045 +0100 +@@ -190,7 +190,7 @@ + /* TODO: audit callers to ensure they are ready for qc_issue to + * unconditionally re-enable interrupts + */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + spin_unlock(ap->lock); + + /* If the device fell off, no sense in issuing commands */ +@@ -255,7 +255,7 @@ + + out: + spin_lock(ap->lock); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return ret; + } + +diff -Nur linux-4.1.13.orig/drivers/scsi/qla2xxx/qla_inline.h linux-4.1.13/drivers/scsi/qla2xxx/qla_inline.h +--- linux-4.1.13.orig/drivers/scsi/qla2xxx/qla_inline.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/scsi/qla2xxx/qla_inline.h 2015-11-29 09:23:09.581616045 +0100 +@@ -59,12 +59,12 @@ + { + unsigned long flags; + struct qla_hw_data *ha = rsp->hw; +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (IS_P3P_TYPE(ha)) + qla82xx_poll(0, rsp); + else + ha->isp_ops->intr_handler(0, rsp); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + static inline uint8_t * +diff -Nur linux-4.1.13.orig/drivers/thermal/x86_pkg_temp_thermal.c linux-4.1.13/drivers/thermal/x86_pkg_temp_thermal.c +--- linux-4.1.13.orig/drivers/thermal/x86_pkg_temp_thermal.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/thermal/x86_pkg_temp_thermal.c 2015-11-29 09:23:09.581616045 +0100 +@@ -29,6 +29,7 @@ + #include <linux/pm.h> + #include <linux/thermal.h> + #include <linux/debugfs.h> ++#include <linux/work-simple.h> + #include <asm/cpu_device_id.h> + #include <asm/mce.h> + +@@ -352,7 +353,7 @@ + } + } + +-static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++static void platform_thermal_notify_work(struct swork_event *event) + { + unsigned long flags; + int cpu = smp_processor_id(); +@@ -369,7 +370,7 @@ + pkg_work_scheduled[phy_id]) { + disable_pkg_thres_interrupt(); + spin_unlock_irqrestore(&pkg_work_lock, flags); +- return -EINVAL; ++ return; + } + pkg_work_scheduled[phy_id] = 1; + spin_unlock_irqrestore(&pkg_work_lock, flags); +@@ -378,9 +379,48 @@ + schedule_delayed_work_on(cpu, + &per_cpu(pkg_temp_thermal_threshold_work, cpu), + msecs_to_jiffies(notify_delay_ms)); ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static struct swork_event notify_work; ++ ++static int thermal_notify_work_init(void) ++{ ++ int err; ++ ++ err = swork_get(); ++ if (err) ++ return err; ++ ++ INIT_SWORK(¬ify_work, platform_thermal_notify_work); + return 0; + } + ++static void thermal_notify_work_cleanup(void) ++{ ++ swork_put(); ++} ++ ++static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++{ ++ swork_queue(¬ify_work); ++ return 0; ++} ++ ++#else /* !CONFIG_PREEMPT_RT_FULL */ ++ ++static int thermal_notify_work_init(void) { return 0; } ++ ++static void thermal_notify_work_cleanup(void) { } ++ ++static int pkg_temp_thermal_platform_thermal_notify(__u64 msr_val) ++{ ++ platform_thermal_notify_work(NULL); ++ ++ return 0; ++} ++#endif /* CONFIG_PREEMPT_RT_FULL */ ++ + static int find_siblings_cpu(int cpu) + { + int i; +@@ -584,6 +624,9 @@ + if (!x86_match_cpu(pkg_temp_thermal_ids)) + return -ENODEV; + ++ if (!thermal_notify_work_init()) ++ return -ENODEV; ++ + spin_lock_init(&pkg_work_lock); + platform_thermal_package_notify = + pkg_temp_thermal_platform_thermal_notify; +@@ -608,7 +651,7 @@ + kfree(pkg_work_scheduled); + platform_thermal_package_notify = NULL; + platform_thermal_package_rate_control = NULL; +- ++ thermal_notify_work_cleanup(); + return -ENODEV; + } + +@@ -633,6 +676,7 @@ + mutex_unlock(&phy_dev_list_mutex); + platform_thermal_package_notify = NULL; + platform_thermal_package_rate_control = NULL; ++ thermal_notify_work_cleanup(); + for_each_online_cpu(i) + cancel_delayed_work_sync( + &per_cpu(pkg_temp_thermal_threshold_work, i)); +diff -Nur linux-4.1.13.orig/drivers/tty/serial/8250/8250_core.c linux-4.1.13/drivers/tty/serial/8250/8250_core.c +--- linux-4.1.13.orig/drivers/tty/serial/8250/8250_core.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/tty/serial/8250/8250_core.c 2015-11-29 09:23:09.581616045 +0100 +@@ -36,6 +36,7 @@ + #include <linux/nmi.h> + #include <linux/mutex.h> + #include <linux/slab.h> ++#include <linux/kdb.h> + #include <linux/uaccess.h> + #include <linux/pm_runtime.h> + #ifdef CONFIG_SPARC +@@ -80,7 +81,16 @@ + #define DEBUG_INTR(fmt...) do { } while (0) + #endif + +-#define PASS_LIMIT 512 ++/* ++ * On -rt we can have a more delays, and legitimately ++ * so - so don't drop work spuriously and spam the ++ * syslog: ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define PASS_LIMIT 1000000 ++#else ++# define PASS_LIMIT 512 ++#endif + + #define BOTH_EMPTY (UART_LSR_TEMT | UART_LSR_THRE) + +@@ -3372,7 +3382,7 @@ + + if (port->sysrq) + locked = 0; +- else if (oops_in_progress) ++ else if (oops_in_progress || in_kdb_printk()) + locked = spin_trylock_irqsave(&port->lock, flags); + else + spin_lock_irqsave(&port->lock, flags); +diff -Nur linux-4.1.13.orig/drivers/tty/serial/amba-pl011.c linux-4.1.13/drivers/tty/serial/amba-pl011.c +--- linux-4.1.13.orig/drivers/tty/serial/amba-pl011.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/tty/serial/amba-pl011.c 2015-11-29 09:23:09.585615778 +0100 +@@ -2000,13 +2000,19 @@ + + clk_enable(uap->clk); + +- local_irq_save(flags); ++ /* ++ * local_irq_save(flags); ++ * ++ * This local_irq_save() is nonsense. If we come in via sysrq ++ * handling then interrupts are already disabled. Aside of ++ * that the port.sysrq check is racy on SMP regardless. ++ */ + if (uap->port.sysrq) + locked = 0; + else if (oops_in_progress) +- locked = spin_trylock(&uap->port.lock); ++ locked = spin_trylock_irqsave(&uap->port.lock, flags); + else +- spin_lock(&uap->port.lock); ++ spin_lock_irqsave(&uap->port.lock, flags); + + /* + * First save the CR then disable the interrupts +@@ -2028,8 +2034,7 @@ + writew(old_cr, uap->port.membase + UART011_CR); + + if (locked) +- spin_unlock(&uap->port.lock); +- local_irq_restore(flags); ++ spin_unlock_irqrestore(&uap->port.lock, flags); + + clk_disable(uap->clk); + } +diff -Nur linux-4.1.13.orig/drivers/tty/serial/omap-serial.c linux-4.1.13/drivers/tty/serial/omap-serial.c +--- linux-4.1.13.orig/drivers/tty/serial/omap-serial.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/tty/serial/omap-serial.c 2015-11-29 09:23:09.585615778 +0100 +@@ -1282,13 +1282,10 @@ + + pm_runtime_get_sync(up->dev); + +- local_irq_save(flags); +- if (up->port.sysrq) +- locked = 0; +- else if (oops_in_progress) +- locked = spin_trylock(&up->port.lock); ++ if (up->port.sysrq || oops_in_progress) ++ locked = spin_trylock_irqsave(&up->port.lock, flags); + else +- spin_lock(&up->port.lock); ++ spin_lock_irqsave(&up->port.lock, flags); + + /* + * First save the IER then disable the interrupts +@@ -1317,8 +1314,7 @@ + pm_runtime_mark_last_busy(up->dev); + pm_runtime_put_autosuspend(up->dev); + if (locked) +- spin_unlock(&up->port.lock); +- local_irq_restore(flags); ++ spin_unlock_irqrestore(&up->port.lock, flags); + } + + static int __init +diff -Nur linux-4.1.13.orig/drivers/usb/core/hcd.c linux-4.1.13/drivers/usb/core/hcd.c +--- linux-4.1.13.orig/drivers/usb/core/hcd.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/usb/core/hcd.c 2015-11-29 09:23:09.585615778 +0100 +@@ -1684,9 +1684,9 @@ + * and no one may trigger the above deadlock situation when + * running complete() in tasklet. + */ +- local_irq_save(flags); ++ local_irq_save_nort(flags); + urb->complete(urb); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + + usb_anchor_resume_wakeups(anchor); + atomic_dec(&urb->use_count); +diff -Nur linux-4.1.13.orig/drivers/usb/gadget/function/f_fs.c linux-4.1.13/drivers/usb/gadget/function/f_fs.c +--- linux-4.1.13.orig/drivers/usb/gadget/function/f_fs.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/usb/gadget/function/f_fs.c 2015-11-29 09:23:09.585615778 +0100 +@@ -1405,7 +1405,7 @@ + pr_info("%s(): freeing\n", __func__); + ffs_data_clear(ffs); + BUG_ON(waitqueue_active(&ffs->ev.waitq) || +- waitqueue_active(&ffs->ep0req_completion.wait)); ++ swaitqueue_active(&ffs->ep0req_completion.wait)); + kfree(ffs->dev_name); + kfree(ffs); + } +diff -Nur linux-4.1.13.orig/drivers/usb/gadget/legacy/inode.c linux-4.1.13/drivers/usb/gadget/legacy/inode.c +--- linux-4.1.13.orig/drivers/usb/gadget/legacy/inode.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/drivers/usb/gadget/legacy/inode.c 2015-11-29 09:23:09.585615778 +0100 +@@ -345,7 +345,7 @@ + spin_unlock_irq (&epdata->dev->lock); + + if (likely (value == 0)) { +- value = wait_event_interruptible (done.wait, done.done); ++ value = swait_event_interruptible (done.wait, done.done); + if (value != 0) { + spin_lock_irq (&epdata->dev->lock); + if (likely (epdata->ep != NULL)) { +@@ -354,7 +354,7 @@ + usb_ep_dequeue (epdata->ep, epdata->req); + spin_unlock_irq (&epdata->dev->lock); + +- wait_event (done.wait, done.done); ++ swait_event (done.wait, done.done); + if (epdata->status == -ECONNRESET) + epdata->status = -EINTR; + } else { +diff -Nur linux-4.1.13.orig/fs/aio.c linux-4.1.13/fs/aio.c +--- linux-4.1.13.orig/fs/aio.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/aio.c 2015-11-29 09:23:09.585615778 +0100 +@@ -40,6 +40,7 @@ + #include <linux/ramfs.h> + #include <linux/percpu-refcount.h> + #include <linux/mount.h> ++#include <linux/work-simple.h> + + #include <asm/kmap_types.h> + #include <asm/uaccess.h> +@@ -115,7 +116,7 @@ + struct page **ring_pages; + long nr_pages; + +- struct work_struct free_work; ++ struct swork_event free_work; + + /* + * signals when all in-flight requests are done +@@ -253,6 +254,7 @@ + .mount = aio_mount, + .kill_sb = kill_anon_super, + }; ++ BUG_ON(swork_get()); + aio_mnt = kern_mount(&aio_fs); + if (IS_ERR(aio_mnt)) + panic("Failed to create aio fs mount."); +@@ -559,9 +561,9 @@ + return cancel(&kiocb->common); + } + +-static void free_ioctx(struct work_struct *work) ++static void free_ioctx(struct swork_event *sev) + { +- struct kioctx *ctx = container_of(work, struct kioctx, free_work); ++ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); + + pr_debug("freeing %p\n", ctx); + +@@ -580,8 +582,8 @@ + if (ctx->rq_wait && atomic_dec_and_test(&ctx->rq_wait->count)) + complete(&ctx->rq_wait->comp); + +- INIT_WORK(&ctx->free_work, free_ioctx); +- schedule_work(&ctx->free_work); ++ INIT_SWORK(&ctx->free_work, free_ioctx); ++ swork_queue(&ctx->free_work); + } + + /* +@@ -589,9 +591,9 @@ + * and ctx->users has dropped to 0, so we know no more kiocbs can be submitted - + * now it's safe to cancel any that need to be. + */ +-static void free_ioctx_users(struct percpu_ref *ref) ++static void free_ioctx_users_work(struct swork_event *sev) + { +- struct kioctx *ctx = container_of(ref, struct kioctx, users); ++ struct kioctx *ctx = container_of(sev, struct kioctx, free_work); + struct aio_kiocb *req; + + spin_lock_irq(&ctx->ctx_lock); +@@ -610,6 +612,14 @@ + percpu_ref_put(&ctx->reqs); + } + ++static void free_ioctx_users(struct percpu_ref *ref) ++{ ++ struct kioctx *ctx = container_of(ref, struct kioctx, users); ++ ++ INIT_SWORK(&ctx->free_work, free_ioctx_users_work); ++ swork_queue(&ctx->free_work); ++} ++ + static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) + { + unsigned i, new_nr; +diff -Nur linux-4.1.13.orig/fs/autofs4/autofs_i.h linux-4.1.13/fs/autofs4/autofs_i.h +--- linux-4.1.13.orig/fs/autofs4/autofs_i.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/autofs4/autofs_i.h 2015-11-29 09:23:09.585615778 +0100 +@@ -34,6 +34,7 @@ + #include <linux/sched.h> + #include <linux/mount.h> + #include <linux/namei.h> ++#include <linux/delay.h> + #include <asm/current.h> + #include <asm/uaccess.h> + +diff -Nur linux-4.1.13.orig/fs/autofs4/expire.c linux-4.1.13/fs/autofs4/expire.c +--- linux-4.1.13.orig/fs/autofs4/expire.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/autofs4/expire.c 2015-11-29 09:23:09.585615778 +0100 +@@ -150,7 +150,7 @@ + parent = p->d_parent; + if (!spin_trylock(&parent->d_lock)) { + spin_unlock(&p->d_lock); +- cpu_relax(); ++ cpu_chill(); + goto relock; + } + spin_unlock(&p->d_lock); +diff -Nur linux-4.1.13.orig/fs/buffer.c linux-4.1.13/fs/buffer.c +--- linux-4.1.13.orig/fs/buffer.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/buffer.c 2015-11-29 09:23:09.585615778 +0100 +@@ -301,8 +301,7 @@ + * decide that the page is now completely done. + */ + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + clear_buffer_async_read(bh); + unlock_buffer(bh); + tmp = bh; +@@ -315,8 +314,7 @@ + } + tmp = tmp->b_this_page; + } while (tmp != bh); +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + + /* + * If none of the buffers had errors and they are all +@@ -328,9 +326,7 @@ + return; + + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + + /* +@@ -358,8 +354,7 @@ + } + + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + + clear_buffer_async_write(bh); + unlock_buffer(bh); +@@ -371,15 +366,12 @@ + } + tmp = tmp->b_this_page; + } +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + end_page_writeback(page); + return; + + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + EXPORT_SYMBOL(end_buffer_async_write); + +@@ -3325,6 +3317,7 @@ + struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); + if (ret) { + INIT_LIST_HEAD(&ret->b_assoc_buffers); ++ buffer_head_init_locks(ret); + preempt_disable(); + __this_cpu_inc(bh_accounting.nr); + recalc_bh_state(); +diff -Nur linux-4.1.13.orig/fs/dcache.c linux-4.1.13/fs/dcache.c +--- linux-4.1.13.orig/fs/dcache.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/dcache.c 2015-11-29 09:23:09.585615778 +0100 +@@ -19,6 +19,7 @@ + #include <linux/mm.h> + #include <linux/fs.h> + #include <linux/fsnotify.h> ++#include <linux/delay.h> + #include <linux/slab.h> + #include <linux/init.h> + #include <linux/hash.h> +@@ -589,7 +590,7 @@ + + failed: + spin_unlock(&dentry->d_lock); +- cpu_relax(); ++ cpu_chill(); + return dentry; /* try again with same dentry */ + } + +@@ -2398,7 +2399,7 @@ + if (dentry->d_lockref.count == 1) { + if (!spin_trylock(&inode->i_lock)) { + spin_unlock(&dentry->d_lock); +- cpu_relax(); ++ cpu_chill(); + goto again; + } + dentry->d_flags &= ~DCACHE_CANT_MOUNT; +diff -Nur linux-4.1.13.orig/fs/eventpoll.c linux-4.1.13/fs/eventpoll.c +--- linux-4.1.13.orig/fs/eventpoll.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/eventpoll.c 2015-11-29 09:23:09.585615778 +0100 +@@ -505,12 +505,12 @@ + */ + static void ep_poll_safewake(wait_queue_head_t *wq) + { +- int this_cpu = get_cpu(); ++ int this_cpu = get_cpu_light(); + + ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS, + ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu); + +- put_cpu(); ++ put_cpu_light(); + } + + static void ep_remove_wait_queue(struct eppoll_entry *pwq) +diff -Nur linux-4.1.13.orig/fs/exec.c linux-4.1.13/fs/exec.c +--- linux-4.1.13.orig/fs/exec.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/exec.c 2015-11-29 09:23:09.593615242 +0100 +@@ -859,12 +859,14 @@ + } + } + task_lock(tsk); ++ preempt_disable_rt(); + active_mm = tsk->active_mm; + tsk->mm = mm; + tsk->active_mm = mm; + activate_mm(active_mm, mm); + tsk->mm->vmacache_seqnum = 0; + vmacache_flush(tsk); ++ preempt_enable_rt(); + task_unlock(tsk); + if (old_mm) { + up_read(&old_mm->mmap_sem); +diff -Nur linux-4.1.13.orig/fs/jbd/checkpoint.c linux-4.1.13/fs/jbd/checkpoint.c +--- linux-4.1.13.orig/fs/jbd/checkpoint.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/jbd/checkpoint.c 2015-11-29 09:23:09.593615242 +0100 +@@ -129,6 +129,8 @@ + if (journal->j_flags & JFS_ABORT) + return; + spin_unlock(&journal->j_state_lock); ++ if (current->plug) ++ io_schedule(); + mutex_lock(&journal->j_checkpoint_mutex); + + /* +diff -Nur linux-4.1.13.orig/fs/jbd2/checkpoint.c linux-4.1.13/fs/jbd2/checkpoint.c +--- linux-4.1.13.orig/fs/jbd2/checkpoint.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/jbd2/checkpoint.c 2015-11-29 09:23:09.593615242 +0100 +@@ -116,6 +116,8 @@ + nblocks = jbd2_space_needed(journal); + while (jbd2_log_space_left(journal) < nblocks) { + write_unlock(&journal->j_state_lock); ++ if (current->plug) ++ io_schedule(); + mutex_lock(&journal->j_checkpoint_mutex); + + /* +diff -Nur linux-4.1.13.orig/fs/namespace.c linux-4.1.13/fs/namespace.c +--- linux-4.1.13.orig/fs/namespace.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/namespace.c 2015-11-29 09:23:09.593615242 +0100 +@@ -14,6 +14,7 @@ + #include <linux/mnt_namespace.h> + #include <linux/user_namespace.h> + #include <linux/namei.h> ++#include <linux/delay.h> + #include <linux/security.h> + #include <linux/idr.h> + #include <linux/init.h> /* init_rootfs */ +@@ -353,8 +354,11 @@ + * incremented count after it has set MNT_WRITE_HOLD. + */ + smp_mb(); +- while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) +- cpu_relax(); ++ while (ACCESS_ONCE(mnt->mnt.mnt_flags) & MNT_WRITE_HOLD) { ++ preempt_enable(); ++ cpu_chill(); ++ preempt_disable(); ++ } + /* + * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will + * be set to match its requirements. So we must not load that until +diff -Nur linux-4.1.13.orig/fs/ntfs/aops.c linux-4.1.13/fs/ntfs/aops.c +--- linux-4.1.13.orig/fs/ntfs/aops.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/ntfs/aops.c 2015-11-29 09:23:09.593615242 +0100 +@@ -107,8 +107,7 @@ + "0x%llx.", (unsigned long long)bh->b_blocknr); + } + first = page_buffers(page); +- local_irq_save(flags); +- bit_spin_lock(BH_Uptodate_Lock, &first->b_state); ++ flags = bh_uptodate_lock_irqsave(first); + clear_buffer_async_read(bh); + unlock_buffer(bh); + tmp = bh; +@@ -123,8 +122,7 @@ + } + tmp = tmp->b_this_page; + } while (tmp != bh); +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); ++ bh_uptodate_unlock_irqrestore(first, flags); + /* + * If none of the buffers had errors then we can set the page uptodate, + * but we first have to perform the post read mst fixups, if the +@@ -145,13 +143,13 @@ + recs = PAGE_CACHE_SIZE / rec_size; + /* Should have been verified before we got here... */ + BUG_ON(!recs); +- local_irq_save(flags); ++ local_irq_save_nort(flags); + kaddr = kmap_atomic(page); + for (i = 0; i < recs; i++) + post_read_mst_fixup((NTFS_RECORD*)(kaddr + + i * rec_size), rec_size); + kunmap_atomic(kaddr); +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + flush_dcache_page(page); + if (likely(page_uptodate && !PageError(page))) + SetPageUptodate(page); +@@ -159,9 +157,7 @@ + unlock_page(page); + return; + still_busy: +- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); +- local_irq_restore(flags); +- return; ++ bh_uptodate_unlock_irqrestore(first, flags); + } + + /** +diff -Nur linux-4.1.13.orig/fs/timerfd.c linux-4.1.13/fs/timerfd.c +--- linux-4.1.13.orig/fs/timerfd.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/timerfd.c 2015-11-29 09:23:09.593615242 +0100 +@@ -450,7 +450,10 @@ + break; + } + spin_unlock_irq(&ctx->wqh.lock); +- cpu_relax(); ++ if (isalarm(ctx)) ++ hrtimer_wait_for_timer(&ctx->t.alarm.timer); ++ else ++ hrtimer_wait_for_timer(&ctx->t.tmr); + } + + /* +diff -Nur linux-4.1.13.orig/fs/xfs/xfs_inode.c linux-4.1.13/fs/xfs/xfs_inode.c +--- linux-4.1.13.orig/fs/xfs/xfs_inode.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/xfs/xfs_inode.c 2015-11-29 09:23:09.593615242 +0100 +@@ -164,7 +164,7 @@ + (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); +- ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); ++ ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); + + if (lock_flags & XFS_IOLOCK_EXCL) + mrupdate_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags)); +@@ -212,7 +212,7 @@ + (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); +- ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); ++ ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); + + if (lock_flags & XFS_IOLOCK_EXCL) { + if (!mrtryupdate(&ip->i_iolock)) +@@ -281,7 +281,7 @@ + (XFS_MMAPLOCK_SHARED | XFS_MMAPLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); +- ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); ++ ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_SUBCLASS_MASK)) == 0); + ASSERT(lock_flags != 0); + + if (lock_flags & XFS_IOLOCK_EXCL) +@@ -364,30 +364,38 @@ + + /* + * Bump the subclass so xfs_lock_inodes() acquires each lock with a different +- * value. This shouldn't be called for page fault locking, but we also need to +- * ensure we don't overrun the number of lockdep subclasses for the iolock or +- * mmaplock as that is limited to 12 by the mmap lock lockdep annotations. ++ * value. This can be called for any type of inode lock combination, including ++ * parent locking. Care must be taken to ensure we don't overrun the subclass ++ * storage fields in the class mask we build. + */ + static inline int + xfs_lock_inumorder(int lock_mode, int subclass) + { ++ int class = 0; ++ ++ ASSERT(!(lock_mode & (XFS_ILOCK_PARENT | XFS_ILOCK_RTBITMAP | ++ XFS_ILOCK_RTSUM))); ++ + if (lock_mode & (XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL)) { +- ASSERT(subclass + XFS_LOCK_INUMORDER < +- (1 << (XFS_MMAPLOCK_SHIFT - XFS_IOLOCK_SHIFT))); +- lock_mode |= (subclass + XFS_LOCK_INUMORDER) << XFS_IOLOCK_SHIFT; ++ ASSERT(subclass <= XFS_IOLOCK_MAX_SUBCLASS); ++ ASSERT(subclass + XFS_IOLOCK_PARENT_VAL < ++ MAX_LOCKDEP_SUBCLASSES); ++ class += subclass << XFS_IOLOCK_SHIFT; ++ if (lock_mode & XFS_IOLOCK_PARENT) ++ class += XFS_IOLOCK_PARENT_VAL << XFS_IOLOCK_SHIFT; + } + + if (lock_mode & (XFS_MMAPLOCK_SHARED|XFS_MMAPLOCK_EXCL)) { +- ASSERT(subclass + XFS_LOCK_INUMORDER < +- (1 << (XFS_ILOCK_SHIFT - XFS_MMAPLOCK_SHIFT))); +- lock_mode |= (subclass + XFS_LOCK_INUMORDER) << +- XFS_MMAPLOCK_SHIFT; ++ ASSERT(subclass <= XFS_MMAPLOCK_MAX_SUBCLASS); ++ class += subclass << XFS_MMAPLOCK_SHIFT; + } + +- if (lock_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL)) +- lock_mode |= (subclass + XFS_LOCK_INUMORDER) << XFS_ILOCK_SHIFT; ++ if (lock_mode & (XFS_ILOCK_SHARED|XFS_ILOCK_EXCL)) { ++ ASSERT(subclass <= XFS_ILOCK_MAX_SUBCLASS); ++ class += subclass << XFS_ILOCK_SHIFT; ++ } + +- return lock_mode; ++ return (lock_mode & ~XFS_LOCK_SUBCLASS_MASK) | class; + } + + /* +@@ -399,6 +407,11 @@ + * transaction (such as truncate). This can result in deadlock since the long + * running trans might need to wait for the inode we just locked in order to + * push the tail and free space in the log. ++ * ++ * xfs_lock_inodes() can only be used to lock one type of lock at a time - ++ * the iolock, the mmaplock or the ilock, but not more than one at a time. If we ++ * lock more than one at a time, lockdep will report false positives saying we ++ * have violated locking orders. + */ + void + xfs_lock_inodes( +@@ -409,8 +422,29 @@ + int attempts = 0, i, j, try_lock; + xfs_log_item_t *lp; + +- /* currently supports between 2 and 5 inodes */ ++ /* ++ * Currently supports between 2 and 5 inodes with exclusive locking. We ++ * support an arbitrary depth of locking here, but absolute limits on ++ * inodes depend on the the type of locking and the limits placed by ++ * lockdep annotations in xfs_lock_inumorder. These are all checked by ++ * the asserts. ++ */ + ASSERT(ips && inodes >= 2 && inodes <= 5); ++ ASSERT(lock_mode & (XFS_IOLOCK_EXCL | XFS_MMAPLOCK_EXCL | ++ XFS_ILOCK_EXCL)); ++ ASSERT(!(lock_mode & (XFS_IOLOCK_SHARED | XFS_MMAPLOCK_SHARED | ++ XFS_ILOCK_SHARED))); ++ ASSERT(!(lock_mode & XFS_IOLOCK_EXCL) || ++ inodes <= XFS_IOLOCK_MAX_SUBCLASS + 1); ++ ASSERT(!(lock_mode & XFS_MMAPLOCK_EXCL) || ++ inodes <= XFS_MMAPLOCK_MAX_SUBCLASS + 1); ++ ASSERT(!(lock_mode & XFS_ILOCK_EXCL) || ++ inodes <= XFS_ILOCK_MAX_SUBCLASS + 1); ++ ++ if (lock_mode & XFS_IOLOCK_EXCL) { ++ ASSERT(!(lock_mode & (XFS_MMAPLOCK_EXCL | XFS_ILOCK_EXCL))); ++ } else if (lock_mode & XFS_MMAPLOCK_EXCL) ++ ASSERT(!(lock_mode & XFS_ILOCK_EXCL)); + + try_lock = 0; + i = 0; +diff -Nur linux-4.1.13.orig/fs/xfs/xfs_inode.h linux-4.1.13/fs/xfs/xfs_inode.h +--- linux-4.1.13.orig/fs/xfs/xfs_inode.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/fs/xfs/xfs_inode.h 2015-11-29 09:23:09.593615242 +0100 +@@ -284,9 +284,9 @@ + * Flags for lockdep annotations. + * + * XFS_LOCK_PARENT - for directory operations that require locking a +- * parent directory inode and a child entry inode. The parent gets locked +- * with this flag so it gets a lockdep subclass of 1 and the child entry +- * lock will have a lockdep subclass of 0. ++ * parent directory inode and a child entry inode. IOLOCK requires nesting, ++ * MMAPLOCK does not support this class, ILOCK requires a single subclass ++ * to differentiate parent from child. + * + * XFS_LOCK_RTBITMAP/XFS_LOCK_RTSUM - the realtime device bitmap and summary + * inodes do not participate in the normal lock order, and thus have their +@@ -295,30 +295,63 @@ + * XFS_LOCK_INUMORDER - for locking several inodes at the some time + * with xfs_lock_inodes(). This flag is used as the starting subclass + * and each subsequent lock acquired will increment the subclass by one. +- * So the first lock acquired will have a lockdep subclass of 4, the +- * second lock will have a lockdep subclass of 5, and so on. It is +- * the responsibility of the class builder to shift this to the correct +- * portion of the lock_mode lockdep mask. ++ * However, MAX_LOCKDEP_SUBCLASSES == 8, which means we are greatly ++ * limited to the subclasses we can represent via nesting. We need at least ++ * 5 inodes nest depth for the ILOCK through rename, and we also have to support ++ * XFS_ILOCK_PARENT, which gives 6 subclasses. Then we have XFS_ILOCK_RTBITMAP ++ * and XFS_ILOCK_RTSUM, which are another 2 unique subclasses, so that's all ++ * 8 subclasses supported by lockdep. ++ * ++ * This also means we have to number the sub-classes in the lowest bits of ++ * the mask we keep, and we have to ensure we never exceed 3 bits of lockdep ++ * mask and we can't use bit-masking to build the subclasses. What a mess. ++ * ++ * Bit layout: ++ * ++ * Bit Lock Region ++ * 16-19 XFS_IOLOCK_SHIFT dependencies ++ * 20-23 XFS_MMAPLOCK_SHIFT dependencies ++ * 24-31 XFS_ILOCK_SHIFT dependencies ++ * ++ * IOLOCK values ++ * ++ * 0-3 subclass value ++ * 4-7 PARENT subclass values ++ * ++ * MMAPLOCK values ++ * ++ * 0-3 subclass value ++ * 4-7 unused ++ * ++ * ILOCK values ++ * 0-4 subclass values ++ * 5 PARENT subclass (not nestable) ++ * 6 RTBITMAP subclass (not nestable) ++ * 7 RTSUM subclass (not nestable) ++ * + */ +-#define XFS_LOCK_PARENT 1 +-#define XFS_LOCK_RTBITMAP 2 +-#define XFS_LOCK_RTSUM 3 +-#define XFS_LOCK_INUMORDER 4 +- +-#define XFS_IOLOCK_SHIFT 16 +-#define XFS_IOLOCK_PARENT (XFS_LOCK_PARENT << XFS_IOLOCK_SHIFT) ++#define XFS_IOLOCK_SHIFT 16 ++#define XFS_IOLOCK_PARENT_VAL 4 ++#define XFS_IOLOCK_MAX_SUBCLASS (XFS_IOLOCK_PARENT_VAL - 1) ++#define XFS_IOLOCK_DEP_MASK 0x000f0000 ++#define XFS_IOLOCK_PARENT (XFS_IOLOCK_PARENT_VAL << XFS_IOLOCK_SHIFT) + +-#define XFS_MMAPLOCK_SHIFT 20 ++#define XFS_MMAPLOCK_SHIFT 20 ++#define XFS_MMAPLOCK_NUMORDER 0 ++#define XFS_MMAPLOCK_MAX_SUBCLASS 3 ++#define XFS_MMAPLOCK_DEP_MASK 0x00f00000 + +-#define XFS_ILOCK_SHIFT 24 +-#define XFS_ILOCK_PARENT (XFS_LOCK_PARENT << XFS_ILOCK_SHIFT) +-#define XFS_ILOCK_RTBITMAP (XFS_LOCK_RTBITMAP << XFS_ILOCK_SHIFT) +-#define XFS_ILOCK_RTSUM (XFS_LOCK_RTSUM << XFS_ILOCK_SHIFT) ++#define XFS_ILOCK_SHIFT 24 ++#define XFS_ILOCK_PARENT_VAL 5 ++#define XFS_ILOCK_MAX_SUBCLASS (XFS_ILOCK_PARENT_VAL - 1) ++#define XFS_ILOCK_RTBITMAP_VAL 6 ++#define XFS_ILOCK_RTSUM_VAL 7 ++#define XFS_ILOCK_DEP_MASK 0xff000000 ++#define XFS_ILOCK_PARENT (XFS_ILOCK_PARENT_VAL << XFS_ILOCK_SHIFT) ++#define XFS_ILOCK_RTBITMAP (XFS_ILOCK_RTBITMAP_VAL << XFS_ILOCK_SHIFT) ++#define XFS_ILOCK_RTSUM (XFS_ILOCK_RTSUM_VAL << XFS_ILOCK_SHIFT) + +-#define XFS_IOLOCK_DEP_MASK 0x000f0000 +-#define XFS_MMAPLOCK_DEP_MASK 0x00f00000 +-#define XFS_ILOCK_DEP_MASK 0xff000000 +-#define XFS_LOCK_DEP_MASK (XFS_IOLOCK_DEP_MASK | \ ++#define XFS_LOCK_SUBCLASS_MASK (XFS_IOLOCK_DEP_MASK | \ + XFS_MMAPLOCK_DEP_MASK | \ + XFS_ILOCK_DEP_MASK) + +diff -Nur linux-4.1.13.orig/include/acpi/platform/aclinux.h linux-4.1.13/include/acpi/platform/aclinux.h +--- linux-4.1.13.orig/include/acpi/platform/aclinux.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/acpi/platform/aclinux.h 2015-11-29 09:23:09.593615242 +0100 +@@ -123,6 +123,7 @@ + + #define acpi_cache_t struct kmem_cache + #define acpi_spinlock spinlock_t * ++#define acpi_raw_spinlock raw_spinlock_t * + #define acpi_cpu_flags unsigned long + + /* Use native linux version of acpi_os_allocate_zeroed */ +@@ -141,6 +142,20 @@ + #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_get_thread_id + #define ACPI_USE_ALTERNATE_PROTOTYPE_acpi_os_create_lock + ++#define acpi_os_create_raw_lock(__handle) \ ++({ \ ++ raw_spinlock_t *lock = ACPI_ALLOCATE(sizeof(*lock)); \ ++ \ ++ if (lock) { \ ++ *(__handle) = lock; \ ++ raw_spin_lock_init(*(__handle)); \ ++ } \ ++ lock ? AE_OK : AE_NO_MEMORY; \ ++ }) ++ ++#define acpi_os_delete_raw_lock(__handle) kfree(__handle) ++ ++ + /* + * OSL interfaces used by debugger/disassembler + */ +diff -Nur linux-4.1.13.orig/include/asm-generic/bug.h linux-4.1.13/include/asm-generic/bug.h +--- linux-4.1.13.orig/include/asm-generic/bug.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/asm-generic/bug.h 2015-11-29 09:23:09.593615242 +0100 +@@ -206,6 +206,20 @@ + # define WARN_ON_SMP(x) ({0;}) + #endif + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define BUG_ON_RT(c) BUG_ON(c) ++# define BUG_ON_NONRT(c) do { } while (0) ++# define WARN_ON_RT(condition) WARN_ON(condition) ++# define WARN_ON_NONRT(condition) do { } while (0) ++# define WARN_ON_ONCE_NONRT(condition) do { } while (0) ++#else ++# define BUG_ON_RT(c) do { } while (0) ++# define BUG_ON_NONRT(c) BUG_ON(c) ++# define WARN_ON_RT(condition) do { } while (0) ++# define WARN_ON_NONRT(condition) WARN_ON(condition) ++# define WARN_ON_ONCE_NONRT(condition) WARN_ON_ONCE(condition) ++#endif ++ + #endif /* __ASSEMBLY__ */ + + #endif +diff -Nur linux-4.1.13.orig/include/asm-generic/futex.h linux-4.1.13/include/asm-generic/futex.h +--- linux-4.1.13.orig/include/asm-generic/futex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/asm-generic/futex.h 2015-11-29 09:23:09.593615242 +0100 +@@ -8,8 +8,7 @@ + #ifndef CONFIG_SMP + /* + * The following implementation only for uniprocessor machines. +- * For UP, it's relies on the fact that pagefault_disable() also disables +- * preemption to ensure mutual exclusion. ++ * It relies on preempt_disable() ensuring mutual exclusion. + * + */ + +@@ -38,6 +37,7 @@ + if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) + oparg = 1 << oparg; + ++ preempt_disable(); + pagefault_disable(); + + ret = -EFAULT; +@@ -72,6 +72,7 @@ + + out_pagefault_enable: + pagefault_enable(); ++ preempt_enable(); + + if (ret == 0) { + switch (cmp) { +@@ -106,6 +107,7 @@ + { + u32 val; + ++ preempt_disable(); + if (unlikely(get_user(val, uaddr) != 0)) + return -EFAULT; + +@@ -113,6 +115,7 @@ + return -EFAULT; + + *uval = val; ++ preempt_enable(); + + return 0; + } +diff -Nur linux-4.1.13.orig/include/linux/blkdev.h linux-4.1.13/include/linux/blkdev.h +--- linux-4.1.13.orig/include/linux/blkdev.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/blkdev.h 2015-11-29 09:23:09.593615242 +0100 +@@ -101,6 +101,7 @@ + struct list_head queuelist; + union { + struct call_single_data csd; ++ struct work_struct work; + unsigned long fifo_time; + }; + +@@ -482,7 +483,7 @@ + struct throtl_data *td; + #endif + struct rcu_head rcu_head; +- wait_queue_head_t mq_freeze_wq; ++ struct swait_head mq_freeze_wq; + struct percpu_ref mq_usage_counter; + struct list_head all_q_node; + +diff -Nur linux-4.1.13.orig/include/linux/blk-mq.h linux-4.1.13/include/linux/blk-mq.h +--- linux-4.1.13.orig/include/linux/blk-mq.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/blk-mq.h 2015-11-29 09:23:09.593615242 +0100 +@@ -202,6 +202,7 @@ + + struct blk_mq_hw_ctx *blk_mq_map_queue(struct request_queue *, const int ctx_index); + struct blk_mq_hw_ctx *blk_mq_alloc_single_hw_queue(struct blk_mq_tag_set *, unsigned int, int); ++void __blk_mq_complete_request_remote_work(struct work_struct *work); + + int blk_mq_request_started(struct request *rq); + void blk_mq_start_request(struct request *rq); +diff -Nur linux-4.1.13.orig/include/linux/bottom_half.h linux-4.1.13/include/linux/bottom_half.h +--- linux-4.1.13.orig/include/linux/bottom_half.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/bottom_half.h 2015-11-29 09:23:09.593615242 +0100 +@@ -4,6 +4,39 @@ + #include <linux/preempt.h> + #include <linux/preempt_mask.h> + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++extern void __local_bh_disable(void); ++extern void _local_bh_enable(void); ++extern void __local_bh_enable(void); ++ ++static inline void local_bh_disable(void) ++{ ++ __local_bh_disable(); ++} ++ ++static inline void __local_bh_disable_ip(unsigned long ip, unsigned int cnt) ++{ ++ __local_bh_disable(); ++} ++ ++static inline void local_bh_enable(void) ++{ ++ __local_bh_enable(); ++} ++ ++static inline void __local_bh_enable_ip(unsigned long ip, unsigned int cnt) ++{ ++ __local_bh_enable(); ++} ++ ++static inline void local_bh_enable_ip(unsigned long ip) ++{ ++ __local_bh_enable(); ++} ++ ++#else ++ + #ifdef CONFIG_TRACE_IRQFLAGS + extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt); + #else +@@ -31,5 +64,6 @@ + { + __local_bh_enable_ip(_THIS_IP_, SOFTIRQ_DISABLE_OFFSET); + } ++#endif + + #endif /* _LINUX_BH_H */ +diff -Nur linux-4.1.13.orig/include/linux/buffer_head.h linux-4.1.13/include/linux/buffer_head.h +--- linux-4.1.13.orig/include/linux/buffer_head.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/buffer_head.h 2015-11-29 09:23:09.601614714 +0100 +@@ -75,8 +75,52 @@ + struct address_space *b_assoc_map; /* mapping this buffer is + associated with */ + atomic_t b_count; /* users using this buffer_head */ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ spinlock_t b_uptodate_lock; ++#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \ ++ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE) ++ spinlock_t b_state_lock; ++ spinlock_t b_journal_head_lock; ++#endif ++#endif + }; + ++static inline unsigned long bh_uptodate_lock_irqsave(struct buffer_head *bh) ++{ ++ unsigned long flags; ++ ++#ifndef CONFIG_PREEMPT_RT_BASE ++ local_irq_save(flags); ++ bit_spin_lock(BH_Uptodate_Lock, &bh->b_state); ++#else ++ spin_lock_irqsave(&bh->b_uptodate_lock, flags); ++#endif ++ return flags; ++} ++ ++static inline void ++bh_uptodate_unlock_irqrestore(struct buffer_head *bh, unsigned long flags) ++{ ++#ifndef CONFIG_PREEMPT_RT_BASE ++ bit_spin_unlock(BH_Uptodate_Lock, &bh->b_state); ++ local_irq_restore(flags); ++#else ++ spin_unlock_irqrestore(&bh->b_uptodate_lock, flags); ++#endif ++} ++ ++static inline void buffer_head_init_locks(struct buffer_head *bh) ++{ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ spin_lock_init(&bh->b_uptodate_lock); ++#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \ ++ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE) ++ spin_lock_init(&bh->b_state_lock); ++ spin_lock_init(&bh->b_journal_head_lock); ++#endif ++#endif ++} ++ + /* + * macro tricks to expand the set_buffer_foo(), clear_buffer_foo() + * and buffer_foo() functions. +diff -Nur linux-4.1.13.orig/include/linux/cgroup.h linux-4.1.13/include/linux/cgroup.h +--- linux-4.1.13.orig/include/linux/cgroup.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/cgroup.h 2015-11-29 09:23:09.601614714 +0100 +@@ -22,6 +22,7 @@ + #include <linux/seq_file.h> + #include <linux/kernfs.h> + #include <linux/wait.h> ++#include <linux/work-simple.h> + + #ifdef CONFIG_CGROUPS + +@@ -91,6 +92,7 @@ + /* percpu_ref killing and RCU release */ + struct rcu_head rcu_head; + struct work_struct destroy_work; ++ struct swork_event destroy_swork; + }; + + /* bits in struct cgroup_subsys_state flags field */ +diff -Nur linux-4.1.13.orig/include/linux/completion.h linux-4.1.13/include/linux/completion.h +--- linux-4.1.13.orig/include/linux/completion.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/completion.h 2015-11-29 09:23:09.601614714 +0100 +@@ -7,8 +7,7 @@ + * Atomic wait-for-completion handler data structures. + * See kernel/sched/completion.c for details. + */ +- +-#include <linux/wait.h> ++#include <linux/wait-simple.h> + + /* + * struct completion - structure used to maintain state for a "completion" +@@ -24,11 +23,11 @@ + */ + struct completion { + unsigned int done; +- wait_queue_head_t wait; ++ struct swait_head wait; + }; + + #define COMPLETION_INITIALIZER(work) \ +- { 0, __WAIT_QUEUE_HEAD_INITIALIZER((work).wait) } ++ { 0, SWAIT_HEAD_INITIALIZER((work).wait) } + + #define COMPLETION_INITIALIZER_ONSTACK(work) \ + ({ init_completion(&work); work; }) +@@ -73,7 +72,7 @@ + static inline void init_completion(struct completion *x) + { + x->done = 0; +- init_waitqueue_head(&x->wait); ++ init_swait_head(&x->wait); + } + + /** +diff -Nur linux-4.1.13.orig/include/linux/cpu.h linux-4.1.13/include/linux/cpu.h +--- linux-4.1.13.orig/include/linux/cpu.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/cpu.h 2015-11-29 09:23:09.601614714 +0100 +@@ -231,6 +231,8 @@ + extern void put_online_cpus(void); + extern void cpu_hotplug_disable(void); + extern void cpu_hotplug_enable(void); ++extern void pin_current_cpu(void); ++extern void unpin_current_cpu(void); + #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri) + #define __hotcpu_notifier(fn, pri) __cpu_notifier(fn, pri) + #define register_hotcpu_notifier(nb) register_cpu_notifier(nb) +@@ -249,6 +251,8 @@ + #define put_online_cpus() do { } while (0) + #define cpu_hotplug_disable() do { } while (0) + #define cpu_hotplug_enable() do { } while (0) ++static inline void pin_current_cpu(void) { } ++static inline void unpin_current_cpu(void) { } + #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) + #define __hotcpu_notifier(fn, pri) do { (void)(fn); } while (0) + /* These aren't inline functions due to a GCC bug. */ +diff -Nur linux-4.1.13.orig/include/linux/delay.h linux-4.1.13/include/linux/delay.h +--- linux-4.1.13.orig/include/linux/delay.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/delay.h 2015-11-29 09:23:09.601614714 +0100 +@@ -52,4 +52,10 @@ + msleep(seconds * 1000); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++extern void cpu_chill(void); ++#else ++# define cpu_chill() cpu_relax() ++#endif ++ + #endif /* defined(_LINUX_DELAY_H) */ +diff -Nur linux-4.1.13.orig/include/linux/ftrace_event.h linux-4.1.13/include/linux/ftrace_event.h +--- linux-4.1.13.orig/include/linux/ftrace_event.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/ftrace_event.h 2015-11-29 09:23:09.601614714 +0100 +@@ -66,6 +66,9 @@ + unsigned char flags; + unsigned char preempt_count; + int pid; ++ unsigned short migrate_disable; ++ unsigned short padding; ++ unsigned char preempt_lazy_count; + }; + + #define FTRACE_MAX_EVENT \ +diff -Nur linux-4.1.13.orig/include/linux/highmem.h linux-4.1.13/include/linux/highmem.h +--- linux-4.1.13.orig/include/linux/highmem.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/highmem.h 2015-11-29 09:23:09.601614714 +0100 +@@ -7,6 +7,7 @@ + #include <linux/mm.h> + #include <linux/uaccess.h> + #include <linux/hardirq.h> ++#include <linux/sched.h> + + #include <asm/cacheflush.h> + +@@ -65,6 +66,7 @@ + + static inline void *kmap_atomic(struct page *page) + { ++ preempt_disable_nort(); + pagefault_disable(); + return page_address(page); + } +@@ -73,6 +75,7 @@ + static inline void __kunmap_atomic(void *addr) + { + pagefault_enable(); ++ preempt_enable_nort(); + } + + #define kmap_atomic_pfn(pfn) kmap_atomic(pfn_to_page(pfn)) +@@ -85,32 +88,51 @@ + + #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) + ++#ifndef CONFIG_PREEMPT_RT_FULL + DECLARE_PER_CPU(int, __kmap_atomic_idx); ++#endif + + static inline int kmap_atomic_idx_push(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + int idx = __this_cpu_inc_return(__kmap_atomic_idx) - 1; + +-#ifdef CONFIG_DEBUG_HIGHMEM ++# ifdef CONFIG_DEBUG_HIGHMEM + WARN_ON_ONCE(in_irq() && !irqs_disabled()); + BUG_ON(idx >= KM_TYPE_NR); +-#endif ++# endif + return idx; ++#else ++ current->kmap_idx++; ++ BUG_ON(current->kmap_idx > KM_TYPE_NR); ++ return current->kmap_idx - 1; ++#endif + } + + static inline int kmap_atomic_idx(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + return __this_cpu_read(__kmap_atomic_idx) - 1; ++#else ++ return current->kmap_idx - 1; ++#endif + } + + static inline void kmap_atomic_idx_pop(void) + { +-#ifdef CONFIG_DEBUG_HIGHMEM ++#ifndef CONFIG_PREEMPT_RT_FULL ++# ifdef CONFIG_DEBUG_HIGHMEM + int idx = __this_cpu_dec_return(__kmap_atomic_idx); + + BUG_ON(idx < 0); +-#else ++# else + __this_cpu_dec(__kmap_atomic_idx); ++# endif ++#else ++ current->kmap_idx--; ++# ifdef CONFIG_DEBUG_HIGHMEM ++ BUG_ON(current->kmap_idx < 0); ++# endif + #endif + } + +diff -Nur linux-4.1.13.orig/include/linux/hrtimer.h linux-4.1.13/include/linux/hrtimer.h +--- linux-4.1.13.orig/include/linux/hrtimer.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/hrtimer.h 2015-11-29 09:23:09.601614714 +0100 +@@ -111,6 +111,11 @@ + enum hrtimer_restart (*function)(struct hrtimer *); + struct hrtimer_clock_base *base; + unsigned long state; ++ struct list_head cb_entry; ++ int irqsafe; ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ ktime_t praecox; ++#endif + #ifdef CONFIG_TIMER_STATS + int start_pid; + void *start_site; +@@ -147,6 +152,7 @@ + int index; + clockid_t clockid; + struct timerqueue_head active; ++ struct list_head expired; + ktime_t resolution; + ktime_t (*get_time)(void); + ktime_t softirq_time; +@@ -194,6 +200,9 @@ + unsigned long nr_hangs; + ktime_t max_hang_time; + #endif ++#ifdef CONFIG_PREEMPT_RT_BASE ++ wait_queue_head_t wait; ++#endif + struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES]; + }; + +@@ -381,6 +390,13 @@ + return hrtimer_start_expires(timer, HRTIMER_MODE_ABS); + } + ++/* Softirq preemption could deadlock timer removal */ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ extern void hrtimer_wait_for_timer(const struct hrtimer *timer); ++#else ++# define hrtimer_wait_for_timer(timer) do { cpu_relax(); } while (0) ++#endif ++ + /* Query timers: */ + extern ktime_t hrtimer_get_remaining(const struct hrtimer *timer); + extern int hrtimer_get_res(const clockid_t which_clock, struct timespec *tp); +diff -Nur linux-4.1.13.orig/include/linux/idr.h linux-4.1.13/include/linux/idr.h +--- linux-4.1.13.orig/include/linux/idr.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/idr.h 2015-11-29 09:23:09.601614714 +0100 +@@ -95,10 +95,14 @@ + * Each idr_preload() should be matched with an invocation of this + * function. See idr_preload() for details. + */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++void idr_preload_end(void); ++#else + static inline void idr_preload_end(void) + { + preempt_enable(); + } ++#endif + + /** + * idr_find - return pointer for given id +diff -Nur linux-4.1.13.orig/include/linux/init_task.h linux-4.1.13/include/linux/init_task.h +--- linux-4.1.13.orig/include/linux/init_task.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/init_task.h 2015-11-29 09:23:09.601614714 +0100 +@@ -147,9 +147,16 @@ + # define INIT_PERF_EVENTS(tsk) + #endif + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define INIT_TIMER_LIST .posix_timer_list = NULL, ++#else ++# define INIT_TIMER_LIST ++#endif ++ + #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN + # define INIT_VTIME(tsk) \ +- .vtime_seqlock = __SEQLOCK_UNLOCKED(tsk.vtime_seqlock), \ ++ .vtime_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.vtime_lock), \ ++ .vtime_seq = SEQCNT_ZERO(tsk.vtime_seq), \ + .vtime_snap = 0, \ + .vtime_snap_whence = VTIME_SYS, + #else +@@ -238,6 +245,7 @@ + .cpu_timers = INIT_CPU_TIMERS(tsk.cpu_timers), \ + .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \ + .timer_slack_ns = 50000, /* 50 usec default slack */ \ ++ INIT_TIMER_LIST \ + .pids = { \ + [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \ + [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ +diff -Nur linux-4.1.13.orig/include/linux/interrupt.h linux-4.1.13/include/linux/interrupt.h +--- linux-4.1.13.orig/include/linux/interrupt.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/interrupt.h 2015-11-29 09:23:09.601614714 +0100 +@@ -61,6 +61,7 @@ + * interrupt handler after suspending interrupts. For system + * wakeup devices users need to implement wakeup detection in + * their interrupt handlers. ++ * IRQF_NO_SOFTIRQ_CALL - Do not process softirqs in the irq thread context (RT) + */ + #define IRQF_SHARED 0x00000080 + #define IRQF_PROBE_SHARED 0x00000100 +@@ -74,6 +75,7 @@ + #define IRQF_NO_THREAD 0x00010000 + #define IRQF_EARLY_RESUME 0x00020000 + #define IRQF_COND_SUSPEND 0x00040000 ++#define IRQF_NO_SOFTIRQ_CALL 0x00080000 + + #define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD) + +@@ -102,6 +104,7 @@ + * @flags: flags (see IRQF_* above) + * @thread_fn: interrupt handler function for threaded interrupts + * @thread: thread pointer for threaded interrupts ++ * @secondary: pointer to secondary irqaction (force threading) + * @thread_flags: flags related to @thread + * @thread_mask: bitmask for keeping track of @thread activity + * @dir: pointer to the proc/irq/NN/name entry +@@ -113,6 +116,7 @@ + struct irqaction *next; + irq_handler_t thread_fn; + struct task_struct *thread; ++ struct irqaction *secondary; + unsigned int irq; + unsigned int flags; + unsigned long thread_flags; +@@ -184,7 +188,7 @@ + #ifdef CONFIG_LOCKDEP + # define local_irq_enable_in_hardirq() do { } while (0) + #else +-# define local_irq_enable_in_hardirq() local_irq_enable() ++# define local_irq_enable_in_hardirq() local_irq_enable_nort() + #endif + + extern void disable_irq_nosync(unsigned int irq); +@@ -215,6 +219,7 @@ + unsigned int irq; + struct kref kref; + struct work_struct work; ++ struct list_head list; + void (*notify)(struct irq_affinity_notify *, const cpumask_t *mask); + void (*release)(struct kref *ref); + }; +@@ -377,9 +382,13 @@ + bool state); + + #ifdef CONFIG_IRQ_FORCED_THREADING ++# ifndef CONFIG_PREEMPT_RT_BASE + extern bool force_irqthreads; ++# else ++# define force_irqthreads (true) ++# endif + #else +-#define force_irqthreads (0) ++#define force_irqthreads (false) + #endif + + #ifndef __ARCH_SET_SOFTIRQ_PENDING +@@ -435,9 +444,10 @@ + void (*action)(struct softirq_action *); + }; + ++#ifndef CONFIG_PREEMPT_RT_FULL + asmlinkage void do_softirq(void); + asmlinkage void __do_softirq(void); +- ++static inline void thread_do_softirq(void) { do_softirq(); } + #ifdef __ARCH_HAS_DO_SOFTIRQ + void do_softirq_own_stack(void); + #else +@@ -446,6 +456,9 @@ + __do_softirq(); + } + #endif ++#else ++extern void thread_do_softirq(void); ++#endif + + extern void open_softirq(int nr, void (*action)(struct softirq_action *)); + extern void softirq_init(void); +@@ -453,6 +466,7 @@ + + extern void raise_softirq_irqoff(unsigned int nr); + extern void raise_softirq(unsigned int nr); ++extern void softirq_check_pending_idle(void); + + DECLARE_PER_CPU(struct task_struct *, ksoftirqd); + +@@ -474,8 +488,9 @@ + to be executed on some cpu at least once after this. + * If the tasklet is already scheduled, but its execution is still not + started, it will be executed only once. +- * If this tasklet is already running on another CPU (or schedule is called +- from tasklet itself), it is rescheduled for later. ++ * If this tasklet is already running on another CPU, it is rescheduled ++ for later. ++ * Schedule must not be called from the tasklet itself (a lockup occurs) + * Tasklet is strictly serialized wrt itself, but not + wrt another tasklets. If client needs some intertask synchronization, + he makes it with spinlocks. +@@ -500,27 +515,36 @@ + enum + { + TASKLET_STATE_SCHED, /* Tasklet is scheduled for execution */ +- TASKLET_STATE_RUN /* Tasklet is running (SMP only) */ ++ TASKLET_STATE_RUN, /* Tasklet is running (SMP only) */ ++ TASKLET_STATE_PENDING /* Tasklet is pending */ + }; + +-#ifdef CONFIG_SMP ++#define TASKLET_STATEF_SCHED (1 << TASKLET_STATE_SCHED) ++#define TASKLET_STATEF_RUN (1 << TASKLET_STATE_RUN) ++#define TASKLET_STATEF_PENDING (1 << TASKLET_STATE_PENDING) ++ ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + static inline int tasklet_trylock(struct tasklet_struct *t) + { + return !test_and_set_bit(TASKLET_STATE_RUN, &(t)->state); + } + ++static inline int tasklet_tryunlock(struct tasklet_struct *t) ++{ ++ return cmpxchg(&t->state, TASKLET_STATEF_RUN, 0) == TASKLET_STATEF_RUN; ++} ++ + static inline void tasklet_unlock(struct tasklet_struct *t) + { + smp_mb__before_atomic(); + clear_bit(TASKLET_STATE_RUN, &(t)->state); + } + +-static inline void tasklet_unlock_wait(struct tasklet_struct *t) +-{ +- while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { barrier(); } +-} ++extern void tasklet_unlock_wait(struct tasklet_struct *t); ++ + #else + #define tasklet_trylock(t) 1 ++#define tasklet_tryunlock(t) 1 + #define tasklet_unlock_wait(t) do { } while (0) + #define tasklet_unlock(t) do { } while (0) + #endif +@@ -569,12 +593,7 @@ + smp_mb(); + } + +-static inline void tasklet_enable(struct tasklet_struct *t) +-{ +- smp_mb__before_atomic(); +- atomic_dec(&t->count); +-} +- ++extern void tasklet_enable(struct tasklet_struct *t); + extern void tasklet_kill(struct tasklet_struct *t); + extern void tasklet_kill_immediate(struct tasklet_struct *t, unsigned int cpu); + extern void tasklet_init(struct tasklet_struct *t, +@@ -605,6 +624,12 @@ + tasklet_kill(&ttimer->tasklet); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++extern void softirq_early_init(void); ++#else ++static inline void softirq_early_init(void) { } ++#endif ++ + /* + * Autoprobing for irqs: + * +diff -Nur linux-4.1.13.orig/include/linux/io-mapping.h linux-4.1.13/include/linux/io-mapping.h +--- linux-4.1.13.orig/include/linux/io-mapping.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/io-mapping.h 2015-11-29 09:23:09.601614714 +0100 +@@ -141,6 +141,7 @@ + io_mapping_map_atomic_wc(struct io_mapping *mapping, + unsigned long offset) + { ++ preempt_disable(); + pagefault_disable(); + return ((char __force __iomem *) mapping) + offset; + } +@@ -149,6 +150,7 @@ + io_mapping_unmap_atomic(void __iomem *vaddr) + { + pagefault_enable(); ++ preempt_enable(); + } + + /* Non-atomic map/unmap */ +diff -Nur linux-4.1.13.orig/include/linux/irqdesc.h linux-4.1.13/include/linux/irqdesc.h +--- linux-4.1.13.orig/include/linux/irqdesc.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/irqdesc.h 2015-11-29 09:23:09.601614714 +0100 +@@ -63,6 +63,7 @@ + unsigned int irqs_unhandled; + atomic_t threads_handled; + int threads_handled_last; ++ u64 random_ip; + raw_spinlock_t lock; + struct cpumask *percpu_enabled; + #ifdef CONFIG_SMP +diff -Nur linux-4.1.13.orig/include/linux/irqflags.h linux-4.1.13/include/linux/irqflags.h +--- linux-4.1.13.orig/include/linux/irqflags.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/irqflags.h 2015-11-29 09:23:09.601614714 +0100 +@@ -25,8 +25,6 @@ + # define trace_softirqs_enabled(p) ((p)->softirqs_enabled) + # define trace_hardirq_enter() do { current->hardirq_context++; } while (0) + # define trace_hardirq_exit() do { current->hardirq_context--; } while (0) +-# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) +-# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) + # define INIT_TRACE_IRQFLAGS .softirqs_enabled = 1, + #else + # define trace_hardirqs_on() do { } while (0) +@@ -39,9 +37,15 @@ + # define trace_softirqs_enabled(p) 0 + # define trace_hardirq_enter() do { } while (0) + # define trace_hardirq_exit() do { } while (0) ++# define INIT_TRACE_IRQFLAGS ++#endif ++ ++#if defined(CONFIG_TRACE_IRQFLAGS) && !defined(CONFIG_PREEMPT_RT_FULL) ++# define lockdep_softirq_enter() do { current->softirq_context++; } while (0) ++# define lockdep_softirq_exit() do { current->softirq_context--; } while (0) ++#else + # define lockdep_softirq_enter() do { } while (0) + # define lockdep_softirq_exit() do { } while (0) +-# define INIT_TRACE_IRQFLAGS + #endif + + #if defined(CONFIG_IRQSOFF_TRACER) || \ +@@ -148,4 +152,23 @@ + + #define irqs_disabled_flags(flags) raw_irqs_disabled_flags(flags) + ++/* ++ * local_irq* variants depending on RT/!RT ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define local_irq_disable_nort() do { } while (0) ++# define local_irq_enable_nort() do { } while (0) ++# define local_irq_save_nort(flags) local_save_flags(flags) ++# define local_irq_restore_nort(flags) (void)(flags) ++# define local_irq_disable_rt() local_irq_disable() ++# define local_irq_enable_rt() local_irq_enable() ++#else ++# define local_irq_disable_nort() local_irq_disable() ++# define local_irq_enable_nort() local_irq_enable() ++# define local_irq_save_nort(flags) local_irq_save(flags) ++# define local_irq_restore_nort(flags) local_irq_restore(flags) ++# define local_irq_disable_rt() do { } while (0) ++# define local_irq_enable_rt() do { } while (0) ++#endif ++ + #endif +diff -Nur linux-4.1.13.orig/include/linux/irq.h linux-4.1.13/include/linux/irq.h +--- linux-4.1.13.orig/include/linux/irq.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/irq.h 2015-11-29 09:23:09.601614714 +0100 +@@ -72,6 +72,7 @@ + * IRQ_IS_POLLED - Always polled by another interrupt. Exclude + * it from the spurious interrupt detection + * mechanism and from core side polling. ++ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT) + */ + enum { + IRQ_TYPE_NONE = 0x00000000, +@@ -97,13 +98,14 @@ + IRQ_NOTHREAD = (1 << 16), + IRQ_PER_CPU_DEVID = (1 << 17), + IRQ_IS_POLLED = (1 << 18), ++ IRQ_NO_SOFTIRQ_CALL = (1 << 19), + }; + + #define IRQF_MODIFY_MASK \ + (IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \ + IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \ + IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NOTHREAD | IRQ_PER_CPU_DEVID | \ +- IRQ_IS_POLLED) ++ IRQ_IS_POLLED | IRQ_NO_SOFTIRQ_CALL) + + #define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING) + +diff -Nur linux-4.1.13.orig/include/linux/irq_work.h linux-4.1.13/include/linux/irq_work.h +--- linux-4.1.13.orig/include/linux/irq_work.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/irq_work.h 2015-11-29 09:23:09.601614714 +0100 +@@ -16,6 +16,7 @@ + #define IRQ_WORK_BUSY 2UL + #define IRQ_WORK_FLAGS 3UL + #define IRQ_WORK_LAZY 4UL /* Doesn't want IPI, wait for tick */ ++#define IRQ_WORK_HARD_IRQ 8UL /* Run hard IRQ context, even on RT */ + + struct irq_work { + unsigned long flags; +@@ -51,4 +52,10 @@ + static inline void irq_work_run(void) { } + #endif + ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void); ++#else ++static inline void irq_work_tick_soft(void) { } ++#endif ++ + #endif /* _LINUX_IRQ_WORK_H */ +diff -Nur linux-4.1.13.orig/include/linux/jbd_common.h linux-4.1.13/include/linux/jbd_common.h +--- linux-4.1.13.orig/include/linux/jbd_common.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/jbd_common.h 2015-11-29 09:23:09.601614714 +0100 +@@ -15,32 +15,56 @@ + + static inline void jbd_lock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(BH_State, &bh->b_state); ++#else ++ spin_lock(&bh->b_state_lock); ++#endif + } + + static inline int jbd_trylock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + return bit_spin_trylock(BH_State, &bh->b_state); ++#else ++ return spin_trylock(&bh->b_state_lock); ++#endif + } + + static inline int jbd_is_locked_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + return bit_spin_is_locked(BH_State, &bh->b_state); ++#else ++ return spin_is_locked(&bh->b_state_lock); ++#endif + } + + static inline void jbd_unlock_bh_state(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_unlock(BH_State, &bh->b_state); ++#else ++ spin_unlock(&bh->b_state_lock); ++#endif + } + + static inline void jbd_lock_bh_journal_head(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(BH_JournalHead, &bh->b_state); ++#else ++ spin_lock(&bh->b_journal_head_lock); ++#endif + } + + static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_unlock(BH_JournalHead, &bh->b_state); ++#else ++ spin_unlock(&bh->b_journal_head_lock); ++#endif + } + + #endif +diff -Nur linux-4.1.13.orig/include/linux/kdb.h linux-4.1.13/include/linux/kdb.h +--- linux-4.1.13.orig/include/linux/kdb.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/kdb.h 2015-11-29 09:23:09.601614714 +0100 +@@ -167,6 +167,7 @@ + extern __printf(1, 2) int kdb_printf(const char *, ...); + typedef __printf(1, 2) int (*kdb_printf_t)(const char *, ...); + ++#define in_kdb_printk() (kdb_trap_printk) + extern void kdb_init(int level); + + /* Access to kdb specific polling devices */ +@@ -201,6 +202,7 @@ + extern int kdb_unregister(char *); + #else /* ! CONFIG_KGDB_KDB */ + static inline __printf(1, 2) int kdb_printf(const char *fmt, ...) { return 0; } ++#define in_kdb_printk() (0) + static inline void kdb_init(int level) {} + static inline int kdb_register(char *cmd, kdb_func_t func, char *usage, + char *help, short minlen) { return 0; } +diff -Nur linux-4.1.13.orig/include/linux/kernel.h linux-4.1.13/include/linux/kernel.h +--- linux-4.1.13.orig/include/linux/kernel.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/kernel.h 2015-11-29 09:23:09.601614714 +0100 +@@ -188,6 +188,9 @@ + */ + # define might_sleep() \ + do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) ++ ++# define might_sleep_no_state_check() \ ++ do { ___might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) + # define sched_annotate_sleep() (current->task_state_change = 0) + #else + static inline void ___might_sleep(const char *file, int line, +@@ -195,6 +198,7 @@ + static inline void __might_sleep(const char *file, int line, + int preempt_offset) { } + # define might_sleep() do { might_resched(); } while (0) ++# define might_sleep_no_state_check() do { might_resched(); } while (0) + # define sched_annotate_sleep() do { } while (0) + #endif + +@@ -244,7 +248,8 @@ + + #if defined(CONFIG_MMU) && \ + (defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_DEBUG_ATOMIC_SLEEP)) +-void might_fault(void); ++#define might_fault() __might_fault(__FILE__, __LINE__) ++void __might_fault(const char *file, int line); + #else + static inline void might_fault(void) { } + #endif +@@ -466,6 +471,7 @@ + SYSTEM_HALT, + SYSTEM_POWER_OFF, + SYSTEM_RESTART, ++ SYSTEM_SUSPEND, + } system_state; + + #define TAINT_PROPRIETARY_MODULE 0 +diff -Nur linux-4.1.13.orig/include/linux/kvm_host.h linux-4.1.13/include/linux/kvm_host.h +--- linux-4.1.13.orig/include/linux/kvm_host.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/kvm_host.h 2015-11-29 09:23:09.601614714 +0100 +@@ -230,7 +230,7 @@ + + int fpu_active; + int guest_fpu_loaded, guest_xcr0_loaded; +- wait_queue_head_t wq; ++ struct swait_head wq; + struct pid *pid; + int sigset_active; + sigset_t sigset; +@@ -690,7 +690,7 @@ + } + #endif + +-static inline wait_queue_head_t *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu) ++static inline struct swait_head *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu) + { + #ifdef __KVM_HAVE_ARCH_WQP + return vcpu->arch.wqp; +diff -Nur linux-4.1.13.orig/include/linux/lglock.h linux-4.1.13/include/linux/lglock.h +--- linux-4.1.13.orig/include/linux/lglock.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/lglock.h 2015-11-29 09:23:09.601614714 +0100 +@@ -34,22 +34,39 @@ + #endif + + struct lglock { ++#ifndef CONFIG_PREEMPT_RT_FULL + arch_spinlock_t __percpu *lock; ++#else ++ struct rt_mutex __percpu *lock; ++#endif + #ifdef CONFIG_DEBUG_LOCK_ALLOC + struct lock_class_key lock_key; + struct lockdep_map lock_dep_map; + #endif + }; + +-#define DEFINE_LGLOCK(name) \ ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define DEFINE_LGLOCK(name) \ + static DEFINE_PER_CPU(arch_spinlock_t, name ## _lock) \ + = __ARCH_SPIN_LOCK_UNLOCKED; \ + struct lglock name = { .lock = &name ## _lock } + +-#define DEFINE_STATIC_LGLOCK(name) \ ++# define DEFINE_STATIC_LGLOCK(name) \ + static DEFINE_PER_CPU(arch_spinlock_t, name ## _lock) \ + = __ARCH_SPIN_LOCK_UNLOCKED; \ + static struct lglock name = { .lock = &name ## _lock } ++#else ++ ++# define DEFINE_LGLOCK(name) \ ++ static DEFINE_PER_CPU(struct rt_mutex, name ## _lock) \ ++ = __RT_MUTEX_INITIALIZER( name ## _lock); \ ++ struct lglock name = { .lock = &name ## _lock } ++ ++# define DEFINE_STATIC_LGLOCK(name) \ ++ static DEFINE_PER_CPU(struct rt_mutex, name ## _lock) \ ++ = __RT_MUTEX_INITIALIZER( name ## _lock); \ ++ static struct lglock name = { .lock = &name ## _lock } ++#endif + + void lg_lock_init(struct lglock *lg, char *name); + void lg_local_lock(struct lglock *lg); +@@ -59,6 +76,12 @@ + void lg_global_lock(struct lglock *lg); + void lg_global_unlock(struct lglock *lg); + ++#ifndef CONFIG_PREEMPT_RT_FULL ++#define lg_global_trylock_relax(name) lg_global_lock(name) ++#else ++void lg_global_trylock_relax(struct lglock *lg); ++#endif ++ + #else + /* When !CONFIG_SMP, map lglock to spinlock */ + #define lglock spinlock +diff -Nur linux-4.1.13.orig/include/linux/list_bl.h linux-4.1.13/include/linux/list_bl.h +--- linux-4.1.13.orig/include/linux/list_bl.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/list_bl.h 2015-11-29 09:23:09.601614714 +0100 +@@ -2,6 +2,7 @@ + #define _LINUX_LIST_BL_H + + #include <linux/list.h> ++#include <linux/spinlock.h> + #include <linux/bit_spinlock.h> + + /* +@@ -32,13 +33,22 @@ + + struct hlist_bl_head { + struct hlist_bl_node *first; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ raw_spinlock_t lock; ++#endif + }; + + struct hlist_bl_node { + struct hlist_bl_node *next, **pprev; + }; +-#define INIT_HLIST_BL_HEAD(ptr) \ +- ((ptr)->first = NULL) ++ ++static inline void INIT_HLIST_BL_HEAD(struct hlist_bl_head *h) ++{ ++ h->first = NULL; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ raw_spin_lock_init(&h->lock); ++#endif ++} + + static inline void INIT_HLIST_BL_NODE(struct hlist_bl_node *h) + { +@@ -117,12 +127,26 @@ + + static inline void hlist_bl_lock(struct hlist_bl_head *b) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + bit_spin_lock(0, (unsigned long *)b); ++#else ++ raw_spin_lock(&b->lock); ++#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) ++ __set_bit(0, (unsigned long *)b); ++#endif ++#endif + } + + static inline void hlist_bl_unlock(struct hlist_bl_head *b) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + __bit_spin_unlock(0, (unsigned long *)b); ++#else ++#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) ++ __clear_bit(0, (unsigned long *)b); ++#endif ++ raw_spin_unlock(&b->lock); ++#endif + } + + static inline bool hlist_bl_is_locked(struct hlist_bl_head *b) +diff -Nur linux-4.1.13.orig/include/linux/locallock.h linux-4.1.13/include/linux/locallock.h +--- linux-4.1.13.orig/include/linux/locallock.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/locallock.h 2015-11-29 09:23:09.605614450 +0100 +@@ -0,0 +1,270 @@ ++#ifndef _LINUX_LOCALLOCK_H ++#define _LINUX_LOCALLOCK_H ++ ++#include <linux/percpu.h> ++#include <linux/spinlock.h> ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++ ++#ifdef CONFIG_DEBUG_SPINLOCK ++# define LL_WARN(cond) WARN_ON(cond) ++#else ++# define LL_WARN(cond) do { } while (0) ++#endif ++ ++/* ++ * per cpu lock based substitute for local_irq_*() ++ */ ++struct local_irq_lock { ++ spinlock_t lock; ++ struct task_struct *owner; ++ int nestcnt; ++ unsigned long flags; ++}; ++ ++#define DEFINE_LOCAL_IRQ_LOCK(lvar) \ ++ DEFINE_PER_CPU(struct local_irq_lock, lvar) = { \ ++ .lock = __SPIN_LOCK_UNLOCKED((lvar).lock) } ++ ++#define DECLARE_LOCAL_IRQ_LOCK(lvar) \ ++ DECLARE_PER_CPU(struct local_irq_lock, lvar) ++ ++#define local_irq_lock_init(lvar) \ ++ do { \ ++ int __cpu; \ ++ for_each_possible_cpu(__cpu) \ ++ spin_lock_init(&per_cpu(lvar, __cpu).lock); \ ++ } while (0) ++ ++/* ++ * spin_lock|trylock|unlock_local flavour that does not migrate disable ++ * used for __local_lock|trylock|unlock where get_local_var/put_local_var ++ * already takes care of the migrate_disable/enable ++ * for CONFIG_PREEMPT_BASE map to the normal spin_* calls. ++ */ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define spin_lock_local(lock) rt_spin_lock(lock) ++# define spin_trylock_local(lock) rt_spin_trylock(lock) ++# define spin_unlock_local(lock) rt_spin_unlock(lock) ++#else ++# define spin_lock_local(lock) spin_lock(lock) ++# define spin_trylock_local(lock) spin_trylock(lock) ++# define spin_unlock_local(lock) spin_unlock(lock) ++#endif ++ ++static inline void __local_lock(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current) { ++ spin_lock_local(&lv->lock); ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ } ++ lv->nestcnt++; ++} ++ ++#define local_lock(lvar) \ ++ do { __local_lock(&get_local_var(lvar)); } while (0) ++ ++static inline int __local_trylock(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current && spin_trylock_local(&lv->lock)) { ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ lv->nestcnt = 1; ++ return 1; ++ } ++ return 0; ++} ++ ++#define local_trylock(lvar) \ ++ ({ \ ++ int __locked; \ ++ __locked = __local_trylock(&get_local_var(lvar)); \ ++ if (!__locked) \ ++ put_local_var(lvar); \ ++ __locked; \ ++ }) ++ ++static inline void __local_unlock(struct local_irq_lock *lv) ++{ ++ LL_WARN(lv->nestcnt == 0); ++ LL_WARN(lv->owner != current); ++ if (--lv->nestcnt) ++ return; ++ ++ lv->owner = NULL; ++ spin_unlock_local(&lv->lock); ++} ++ ++#define local_unlock(lvar) \ ++ do { \ ++ __local_unlock(this_cpu_ptr(&lvar)); \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++static inline void __local_lock_irq(struct local_irq_lock *lv) ++{ ++ spin_lock_irqsave(&lv->lock, lv->flags); ++ LL_WARN(lv->owner); ++ LL_WARN(lv->nestcnt); ++ lv->owner = current; ++ lv->nestcnt = 1; ++} ++ ++#define local_lock_irq(lvar) \ ++ do { __local_lock_irq(&get_local_var(lvar)); } while (0) ++ ++#define local_lock_irq_on(lvar, cpu) \ ++ do { __local_lock_irq(&per_cpu(lvar, cpu)); } while (0) ++ ++static inline void __local_unlock_irq(struct local_irq_lock *lv) ++{ ++ LL_WARN(!lv->nestcnt); ++ LL_WARN(lv->owner != current); ++ lv->owner = NULL; ++ lv->nestcnt = 0; ++ spin_unlock_irq(&lv->lock); ++} ++ ++#define local_unlock_irq(lvar) \ ++ do { \ ++ __local_unlock_irq(this_cpu_ptr(&lvar)); \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++#define local_unlock_irq_on(lvar, cpu) \ ++ do { \ ++ __local_unlock_irq(&per_cpu(lvar, cpu)); \ ++ } while (0) ++ ++static inline int __local_lock_irqsave(struct local_irq_lock *lv) ++{ ++ if (lv->owner != current) { ++ __local_lock_irq(lv); ++ return 0; ++ } else { ++ lv->nestcnt++; ++ return 1; ++ } ++} ++ ++#define local_lock_irqsave(lvar, _flags) \ ++ do { \ ++ if (__local_lock_irqsave(&get_local_var(lvar))) \ ++ put_local_var(lvar); \ ++ _flags = __this_cpu_read(lvar.flags); \ ++ } while (0) ++ ++#define local_lock_irqsave_on(lvar, _flags, cpu) \ ++ do { \ ++ __local_lock_irqsave(&per_cpu(lvar, cpu)); \ ++ _flags = per_cpu(lvar, cpu).flags; \ ++ } while (0) ++ ++static inline int __local_unlock_irqrestore(struct local_irq_lock *lv, ++ unsigned long flags) ++{ ++ LL_WARN(!lv->nestcnt); ++ LL_WARN(lv->owner != current); ++ if (--lv->nestcnt) ++ return 0; ++ ++ lv->owner = NULL; ++ spin_unlock_irqrestore(&lv->lock, lv->flags); ++ return 1; ++} ++ ++#define local_unlock_irqrestore(lvar, flags) \ ++ do { \ ++ if (__local_unlock_irqrestore(this_cpu_ptr(&lvar), flags)) \ ++ put_local_var(lvar); \ ++ } while (0) ++ ++#define local_unlock_irqrestore_on(lvar, flags, cpu) \ ++ do { \ ++ __local_unlock_irqrestore(&per_cpu(lvar, cpu), flags); \ ++ } while (0) ++ ++#define local_spin_trylock_irq(lvar, lock) \ ++ ({ \ ++ int __locked; \ ++ local_lock_irq(lvar); \ ++ __locked = spin_trylock(lock); \ ++ if (!__locked) \ ++ local_unlock_irq(lvar); \ ++ __locked; \ ++ }) ++ ++#define local_spin_lock_irq(lvar, lock) \ ++ do { \ ++ local_lock_irq(lvar); \ ++ spin_lock(lock); \ ++ } while (0) ++ ++#define local_spin_unlock_irq(lvar, lock) \ ++ do { \ ++ spin_unlock(lock); \ ++ local_unlock_irq(lvar); \ ++ } while (0) ++ ++#define local_spin_lock_irqsave(lvar, lock, flags) \ ++ do { \ ++ local_lock_irqsave(lvar, flags); \ ++ spin_lock(lock); \ ++ } while (0) ++ ++#define local_spin_unlock_irqrestore(lvar, lock, flags) \ ++ do { \ ++ spin_unlock(lock); \ ++ local_unlock_irqrestore(lvar, flags); \ ++ } while (0) ++ ++#define get_locked_var(lvar, var) \ ++ (*({ \ ++ local_lock(lvar); \ ++ this_cpu_ptr(&var); \ ++ })) ++ ++#define put_locked_var(lvar, var) local_unlock(lvar); ++ ++#define local_lock_cpu(lvar) \ ++ ({ \ ++ local_lock(lvar); \ ++ smp_processor_id(); \ ++ }) ++ ++#define local_unlock_cpu(lvar) local_unlock(lvar) ++ ++#else /* PREEMPT_RT_BASE */ ++ ++#define DEFINE_LOCAL_IRQ_LOCK(lvar) __typeof__(const int) lvar ++#define DECLARE_LOCAL_IRQ_LOCK(lvar) extern __typeof__(const int) lvar ++ ++static inline void local_irq_lock_init(int lvar) { } ++ ++#define local_lock(lvar) preempt_disable() ++#define local_unlock(lvar) preempt_enable() ++#define local_lock_irq(lvar) local_irq_disable() ++#define local_unlock_irq(lvar) local_irq_enable() ++#define local_lock_irqsave(lvar, flags) local_irq_save(flags) ++#define local_unlock_irqrestore(lvar, flags) local_irq_restore(flags) ++ ++#define local_spin_trylock_irq(lvar, lock) spin_trylock_irq(lock) ++#define local_spin_lock_irq(lvar, lock) spin_lock_irq(lock) ++#define local_spin_unlock_irq(lvar, lock) spin_unlock_irq(lock) ++#define local_spin_lock_irqsave(lvar, lock, flags) \ ++ spin_lock_irqsave(lock, flags) ++#define local_spin_unlock_irqrestore(lvar, lock, flags) \ ++ spin_unlock_irqrestore(lock, flags) ++ ++#define get_locked_var(lvar, var) get_cpu_var(var) ++#define put_locked_var(lvar, var) put_cpu_var(var) ++ ++#define local_lock_cpu(lvar) get_cpu() ++#define local_unlock_cpu(lvar) put_cpu() ++ ++#endif ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/mm_types.h linux-4.1.13/include/linux/mm_types.h +--- linux-4.1.13.orig/include/linux/mm_types.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/mm_types.h 2015-11-29 09:23:09.605614450 +0100 +@@ -11,6 +11,7 @@ + #include <linux/completion.h> + #include <linux/cpumask.h> + #include <linux/uprobes.h> ++#include <linux/rcupdate.h> + #include <linux/page-flags-layout.h> + #include <asm/page.h> + #include <asm/mmu.h> +@@ -453,6 +454,9 @@ + bool tlb_flush_pending; + #endif + struct uprobes_state uprobes_state; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct rcu_head delayed_drop; ++#endif + #ifdef CONFIG_X86_INTEL_MPX + /* address of the bounds directory */ + void __user *bd_addr; +diff -Nur linux-4.1.13.orig/include/linux/mutex.h linux-4.1.13/include/linux/mutex.h +--- linux-4.1.13.orig/include/linux/mutex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/mutex.h 2015-11-29 09:23:09.609614185 +0100 +@@ -19,6 +19,17 @@ + #include <asm/processor.h> + #include <linux/osq_lock.h> + ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ ++ , .dep_map = { .name = #lockname } ++#else ++# define __DEP_MAP_MUTEX_INITIALIZER(lockname) ++#endif ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include <linux/mutex_rt.h> ++#else ++ + /* + * Simple, straightforward mutexes with strict semantics: + * +@@ -99,13 +110,6 @@ + static inline void mutex_destroy(struct mutex *lock) {} + #endif + +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \ +- , .dep_map = { .name = #lockname } +-#else +-# define __DEP_MAP_MUTEX_INITIALIZER(lockname) +-#endif +- + #define __MUTEX_INITIALIZER(lockname) \ + { .count = ATOMIC_INIT(1) \ + , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \ +@@ -173,6 +177,8 @@ + extern int mutex_trylock(struct mutex *lock); + extern void mutex_unlock(struct mutex *lock); + ++#endif /* !PREEMPT_RT_FULL */ ++ + extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); + + #endif /* __LINUX_MUTEX_H */ +diff -Nur linux-4.1.13.orig/include/linux/mutex_rt.h linux-4.1.13/include/linux/mutex_rt.h +--- linux-4.1.13.orig/include/linux/mutex_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/mutex_rt.h 2015-11-29 09:23:09.609614185 +0100 +@@ -0,0 +1,84 @@ ++#ifndef __LINUX_MUTEX_RT_H ++#define __LINUX_MUTEX_RT_H ++ ++#ifndef __LINUX_MUTEX_H ++#error "Please include mutex.h" ++#endif ++ ++#include <linux/rtmutex.h> ++ ++/* FIXME: Just for __lockfunc */ ++#include <linux/spinlock.h> ++ ++struct mutex { ++ struct rt_mutex lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++}; ++ ++#define __MUTEX_INITIALIZER(mutexname) \ ++ { \ ++ .lock = __RT_MUTEX_INITIALIZER(mutexname.lock) \ ++ __DEP_MAP_MUTEX_INITIALIZER(mutexname) \ ++ } ++ ++#define DEFINE_MUTEX(mutexname) \ ++ struct mutex mutexname = __MUTEX_INITIALIZER(mutexname) ++ ++extern void __mutex_do_init(struct mutex *lock, const char *name, struct lock_class_key *key); ++extern void __lockfunc _mutex_lock(struct mutex *lock); ++extern int __lockfunc _mutex_lock_interruptible(struct mutex *lock); ++extern int __lockfunc _mutex_lock_killable(struct mutex *lock); ++extern void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass); ++extern void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest_lock); ++extern int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass); ++extern int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass); ++extern int __lockfunc _mutex_trylock(struct mutex *lock); ++extern void __lockfunc _mutex_unlock(struct mutex *lock); ++ ++#define mutex_is_locked(l) rt_mutex_is_locked(&(l)->lock) ++#define mutex_lock(l) _mutex_lock(l) ++#define mutex_lock_interruptible(l) _mutex_lock_interruptible(l) ++#define mutex_lock_killable(l) _mutex_lock_killable(l) ++#define mutex_trylock(l) _mutex_trylock(l) ++#define mutex_unlock(l) _mutex_unlock(l) ++#define mutex_destroy(l) rt_mutex_destroy(&(l)->lock) ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define mutex_lock_nested(l, s) _mutex_lock_nested(l, s) ++# define mutex_lock_interruptible_nested(l, s) \ ++ _mutex_lock_interruptible_nested(l, s) ++# define mutex_lock_killable_nested(l, s) \ ++ _mutex_lock_killable_nested(l, s) ++ ++# define mutex_lock_nest_lock(lock, nest_lock) \ ++do { \ ++ typecheck(struct lockdep_map *, &(nest_lock)->dep_map); \ ++ _mutex_lock_nest_lock(lock, &(nest_lock)->dep_map); \ ++} while (0) ++ ++#else ++# define mutex_lock_nested(l, s) _mutex_lock(l) ++# define mutex_lock_interruptible_nested(l, s) \ ++ _mutex_lock_interruptible(l) ++# define mutex_lock_killable_nested(l, s) \ ++ _mutex_lock_killable(l) ++# define mutex_lock_nest_lock(lock, nest_lock) mutex_lock(lock) ++#endif ++ ++# define mutex_init(mutex) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(mutex)->lock); \ ++ __mutex_do_init((mutex), #mutex, &__key); \ ++} while (0) ++ ++# define __mutex_init(mutex, name, key) \ ++do { \ ++ rt_mutex_init(&(mutex)->lock); \ ++ __mutex_do_init((mutex), name, key); \ ++} while (0) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/netdevice.h linux-4.1.13/include/linux/netdevice.h +--- linux-4.1.13.orig/include/linux/netdevice.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/netdevice.h 2015-11-29 09:23:09.609614185 +0100 +@@ -2469,6 +2469,7 @@ + unsigned int dropped; + struct sk_buff_head input_pkt_queue; + struct napi_struct backlog; ++ struct sk_buff_head tofree_queue; + + }; + +diff -Nur linux-4.1.13.orig/include/linux/netfilter/x_tables.h linux-4.1.13/include/linux/netfilter/x_tables.h +--- linux-4.1.13.orig/include/linux/netfilter/x_tables.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/netfilter/x_tables.h 2015-11-29 09:23:09.609614185 +0100 +@@ -3,6 +3,7 @@ + + + #include <linux/netdevice.h> ++#include <linux/locallock.h> + #include <uapi/linux/netfilter/x_tables.h> + + /** +@@ -282,6 +283,8 @@ + */ + DECLARE_PER_CPU(seqcount_t, xt_recseq); + ++DECLARE_LOCAL_IRQ_LOCK(xt_write_lock); ++ + /** + * xt_write_recseq_begin - start of a write section + * +@@ -296,6 +299,9 @@ + { + unsigned int addend; + ++ /* RT protection */ ++ local_lock(xt_write_lock); ++ + /* + * Low order bit of sequence is set if we already + * called xt_write_recseq_begin(). +@@ -326,6 +332,7 @@ + /* this is kind of a write_seqcount_end(), but addend is 0 or 1 */ + smp_wmb(); + __this_cpu_add(xt_recseq.sequence, addend); ++ local_unlock(xt_write_lock); + } + + /* +diff -Nur linux-4.1.13.orig/include/linux/notifier.h linux-4.1.13/include/linux/notifier.h +--- linux-4.1.13.orig/include/linux/notifier.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/notifier.h 2015-11-29 09:23:09.609614185 +0100 +@@ -6,7 +6,7 @@ + * + * Alan Cox <Alan.Cox@linux.org> + */ +- ++ + #ifndef _LINUX_NOTIFIER_H + #define _LINUX_NOTIFIER_H + #include <linux/errno.h> +@@ -42,9 +42,7 @@ + * in srcu_notifier_call_chain(): no cache bounces and no memory barriers. + * As compensation, srcu_notifier_chain_unregister() is rather expensive. + * SRCU notifier chains should be used when the chain will be called very +- * often but notifier_blocks will seldom be removed. Also, SRCU notifier +- * chains are slightly more difficult to use because they require special +- * runtime initialization. ++ * often but notifier_blocks will seldom be removed. + */ + + typedef int (*notifier_fn_t)(struct notifier_block *nb, +@@ -88,7 +86,7 @@ + (name)->head = NULL; \ + } while (0) + +-/* srcu_notifier_heads must be initialized and cleaned up dynamically */ ++/* srcu_notifier_heads must be cleaned up dynamically */ + extern void srcu_init_notifier_head(struct srcu_notifier_head *nh); + #define srcu_cleanup_notifier_head(name) \ + cleanup_srcu_struct(&(name)->srcu); +@@ -101,7 +99,13 @@ + .head = NULL } + #define RAW_NOTIFIER_INIT(name) { \ + .head = NULL } +-/* srcu_notifier_heads cannot be initialized statically */ ++ ++#define SRCU_NOTIFIER_INIT(name, pcpu) \ ++ { \ ++ .mutex = __MUTEX_INITIALIZER(name.mutex), \ ++ .head = NULL, \ ++ .srcu = __SRCU_STRUCT_INIT(name.srcu, pcpu), \ ++ } + + #define ATOMIC_NOTIFIER_HEAD(name) \ + struct atomic_notifier_head name = \ +@@ -113,6 +117,18 @@ + struct raw_notifier_head name = \ + RAW_NOTIFIER_INIT(name) + ++#define _SRCU_NOTIFIER_HEAD(name, mod) \ ++ static DEFINE_PER_CPU(struct srcu_struct_array, \ ++ name##_head_srcu_array); \ ++ mod struct srcu_notifier_head name = \ ++ SRCU_NOTIFIER_INIT(name, name##_head_srcu_array) ++ ++#define SRCU_NOTIFIER_HEAD(name) \ ++ _SRCU_NOTIFIER_HEAD(name, ) ++ ++#define SRCU_NOTIFIER_HEAD_STATIC(name) \ ++ _SRCU_NOTIFIER_HEAD(name, static) ++ + #ifdef __KERNEL__ + + extern int atomic_notifier_chain_register(struct atomic_notifier_head *nh, +@@ -182,12 +198,12 @@ + + /* + * Declared notifiers so far. I can imagine quite a few more chains +- * over time (eg laptop power reset chains, reboot chain (to clean ++ * over time (eg laptop power reset chains, reboot chain (to clean + * device units up), device [un]mount chain, module load/unload chain, +- * low memory chain, screenblank chain (for plug in modular screenblankers) ++ * low memory chain, screenblank chain (for plug in modular screenblankers) + * VC switch chains (for loadable kernel svgalib VC switch helpers) etc... + */ +- ++ + /* CPU notfiers are defined in include/linux/cpu.h. */ + + /* netdevice notifiers are defined in include/linux/netdevice.h */ +diff -Nur linux-4.1.13.orig/include/linux/percpu.h linux-4.1.13/include/linux/percpu.h +--- linux-4.1.13.orig/include/linux/percpu.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/percpu.h 2015-11-29 09:23:09.609614185 +0100 +@@ -24,6 +24,35 @@ + PERCPU_MODULE_RESERVE) + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ ++#define get_local_var(var) (*({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(&var); })) ++ ++#define put_local_var(var) do { \ ++ (void)&(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++# define get_local_ptr(var) ({ \ ++ migrate_disable(); \ ++ this_cpu_ptr(var); }) ++ ++# define put_local_ptr(var) do { \ ++ (void)(var); \ ++ migrate_enable(); \ ++} while (0) ++ ++#else ++ ++#define get_local_var(var) get_cpu_var(var) ++#define put_local_var(var) put_cpu_var(var) ++#define get_local_ptr(var) get_cpu_ptr(var) ++#define put_local_ptr(var) put_cpu_ptr(var) ++ ++#endif ++ + /* minimum unit size, also is the maximum supported allocation size */ + #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10) + +diff -Nur linux-4.1.13.orig/include/linux/pid.h linux-4.1.13/include/linux/pid.h +--- linux-4.1.13.orig/include/linux/pid.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/pid.h 2015-11-29 09:23:09.609614185 +0100 +@@ -2,6 +2,7 @@ + #define _LINUX_PID_H + + #include <linux/rcupdate.h> ++#include <linux/atomic.h> + + enum pid_type + { +diff -Nur linux-4.1.13.orig/include/linux/preempt.h linux-4.1.13/include/linux/preempt.h +--- linux-4.1.13.orig/include/linux/preempt.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/preempt.h 2015-11-29 09:23:09.609614185 +0100 +@@ -34,6 +34,20 @@ + #define preempt_count_inc() preempt_count_add(1) + #define preempt_count_dec() preempt_count_sub(1) + ++#ifdef CONFIG_PREEMPT_LAZY ++#define add_preempt_lazy_count(val) do { preempt_lazy_count() += (val); } while (0) ++#define sub_preempt_lazy_count(val) do { preempt_lazy_count() -= (val); } while (0) ++#define inc_preempt_lazy_count() add_preempt_lazy_count(1) ++#define dec_preempt_lazy_count() sub_preempt_lazy_count(1) ++#define preempt_lazy_count() (current_thread_info()->preempt_lazy_count) ++#else ++#define add_preempt_lazy_count(val) do { } while (0) ++#define sub_preempt_lazy_count(val) do { } while (0) ++#define inc_preempt_lazy_count() do { } while (0) ++#define dec_preempt_lazy_count() do { } while (0) ++#define preempt_lazy_count() (0) ++#endif ++ + #ifdef CONFIG_PREEMPT_COUNT + + #define preempt_disable() \ +@@ -42,13 +56,25 @@ + barrier(); \ + } while (0) + ++#define preempt_lazy_disable() \ ++do { \ ++ inc_preempt_lazy_count(); \ ++ barrier(); \ ++} while (0) ++ + #define sched_preempt_enable_no_resched() \ + do { \ + barrier(); \ + preempt_count_dec(); \ + } while (0) + +-#define preempt_enable_no_resched() sched_preempt_enable_no_resched() ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define preempt_enable_no_resched() sched_preempt_enable_no_resched() ++# define preempt_check_resched_rt() preempt_check_resched() ++#else ++# define preempt_enable_no_resched() preempt_enable() ++# define preempt_check_resched_rt() barrier(); ++#endif + + #ifdef CONFIG_PREEMPT + #define preempt_enable() \ +@@ -64,6 +90,13 @@ + __preempt_schedule(); \ + } while (0) + ++#define preempt_lazy_enable() \ ++do { \ ++ dec_preempt_lazy_count(); \ ++ barrier(); \ ++ preempt_check_resched(); \ ++} while (0) ++ + #else + #define preempt_enable() \ + do { \ +@@ -122,6 +155,7 @@ + #define preempt_disable_notrace() barrier() + #define preempt_enable_no_resched_notrace() barrier() + #define preempt_enable_notrace() barrier() ++#define preempt_check_resched_rt() barrier() + + #endif /* CONFIG_PREEMPT_COUNT */ + +@@ -141,10 +175,31 @@ + } while (0) + #define preempt_fold_need_resched() \ + do { \ +- if (tif_need_resched()) \ ++ if (tif_need_resched_now()) \ + set_preempt_need_resched(); \ + } while (0) + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define preempt_disable_rt() preempt_disable() ++# define preempt_enable_rt() preempt_enable() ++# define preempt_disable_nort() barrier() ++# define preempt_enable_nort() barrier() ++# ifdef CONFIG_SMP ++ extern void migrate_disable(void); ++ extern void migrate_enable(void); ++# else /* CONFIG_SMP */ ++# define migrate_disable() barrier() ++# define migrate_enable() barrier() ++# endif /* CONFIG_SMP */ ++#else ++# define preempt_disable_rt() barrier() ++# define preempt_enable_rt() barrier() ++# define preempt_disable_nort() preempt_disable() ++# define preempt_enable_nort() preempt_enable() ++# define migrate_disable() preempt_disable() ++# define migrate_enable() preempt_enable() ++#endif ++ + #ifdef CONFIG_PREEMPT_NOTIFIERS + + struct preempt_notifier; +diff -Nur linux-4.1.13.orig/include/linux/preempt_mask.h linux-4.1.13/include/linux/preempt_mask.h +--- linux-4.1.13.orig/include/linux/preempt_mask.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/preempt_mask.h 2015-11-29 09:23:09.609614185 +0100 +@@ -44,16 +44,26 @@ + #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT) + #define NMI_OFFSET (1UL << NMI_SHIFT) + +-#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET) ++#else ++# define SOFTIRQ_DISABLE_OFFSET (0) ++#endif + + #define PREEMPT_ACTIVE_BITS 1 + #define PREEMPT_ACTIVE_SHIFT (NMI_SHIFT + NMI_BITS) + #define PREEMPT_ACTIVE (__IRQ_MASK(PREEMPT_ACTIVE_BITS) << PREEMPT_ACTIVE_SHIFT) + + #define hardirq_count() (preempt_count() & HARDIRQ_MASK) +-#define softirq_count() (preempt_count() & SOFTIRQ_MASK) + #define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \ + | NMI_MASK)) ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define softirq_count() (preempt_count() & SOFTIRQ_MASK) ++# define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) ++#else ++# define softirq_count() (0UL) ++extern int in_serving_softirq(void); ++#endif + + /* + * Are we doing bottom half or hardware interrupt processing? +@@ -64,7 +74,6 @@ + #define in_irq() (hardirq_count()) + #define in_softirq() (softirq_count()) + #define in_interrupt() (irq_count()) +-#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET) + + /* + * Are we in NMI context? +@@ -83,7 +92,11 @@ + /* + * The preempt_count offset after spin_lock() + */ ++#if !defined(CONFIG_PREEMPT_RT_FULL) + #define PREEMPT_LOCK_OFFSET PREEMPT_DISABLE_OFFSET ++#else ++#define PREEMPT_LOCK_OFFSET 0 ++#endif + + /* + * The preempt_count offset needed for things like: +diff -Nur linux-4.1.13.orig/include/linux/printk.h linux-4.1.13/include/linux/printk.h +--- linux-4.1.13.orig/include/linux/printk.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/printk.h 2015-11-29 09:23:09.609614185 +0100 +@@ -115,9 +115,11 @@ + #ifdef CONFIG_EARLY_PRINTK + extern asmlinkage __printf(1, 2) + void early_printk(const char *fmt, ...); ++extern void printk_kill(void); + #else + static inline __printf(1, 2) __cold + void early_printk(const char *s, ...) { } ++static inline void printk_kill(void) { } + #endif + + typedef int(*printk_func_t)(const char *fmt, va_list args); +diff -Nur linux-4.1.13.orig/include/linux/radix-tree.h linux-4.1.13/include/linux/radix-tree.h +--- linux-4.1.13.orig/include/linux/radix-tree.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/radix-tree.h 2015-11-29 09:23:09.609614185 +0100 +@@ -277,8 +277,13 @@ + unsigned int radix_tree_gang_lookup_slot(struct radix_tree_root *root, + void ***results, unsigned long *indices, + unsigned long first_index, unsigned int max_items); ++#ifndef CONFIG_PREEMPT_RT_FULL + int radix_tree_preload(gfp_t gfp_mask); + int radix_tree_maybe_preload(gfp_t gfp_mask); ++#else ++static inline int radix_tree_preload(gfp_t gm) { return 0; } ++static inline int radix_tree_maybe_preload(gfp_t gfp_mask) { return 0; } ++#endif + void radix_tree_init(void); + void *radix_tree_tag_set(struct radix_tree_root *root, + unsigned long index, unsigned int tag); +@@ -303,7 +308,7 @@ + + static inline void radix_tree_preload_end(void) + { +- preempt_enable(); ++ preempt_enable_nort(); + } + + /** +diff -Nur linux-4.1.13.orig/include/linux/random.h linux-4.1.13/include/linux/random.h +--- linux-4.1.13.orig/include/linux/random.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/random.h 2015-11-29 09:23:09.609614185 +0100 +@@ -11,7 +11,7 @@ + extern void add_device_randomness(const void *, unsigned int); + extern void add_input_randomness(unsigned int type, unsigned int code, + unsigned int value); +-extern void add_interrupt_randomness(int irq, int irq_flags); ++extern void add_interrupt_randomness(int irq, int irq_flags, __u64 ip); + + extern void get_random_bytes(void *buf, int nbytes); + extern void get_random_bytes_arch(void *buf, int nbytes); +diff -Nur linux-4.1.13.orig/include/linux/rcupdate.h linux-4.1.13/include/linux/rcupdate.h +--- linux-4.1.13.orig/include/linux/rcupdate.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/rcupdate.h 2015-11-29 09:23:09.609614185 +0100 +@@ -167,6 +167,9 @@ + + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + ++#ifdef CONFIG_PREEMPT_RT_FULL ++#define call_rcu_bh call_rcu ++#else + /** + * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period. + * @head: structure to be used for queueing the RCU updates. +@@ -190,6 +193,7 @@ + */ + void call_rcu_bh(struct rcu_head *head, + void (*func)(struct rcu_head *head)); ++#endif + + /** + * call_rcu_sched() - Queue an RCU for invocation after sched grace period. +@@ -260,6 +264,11 @@ + * types of kernel builds, the rcu_read_lock() nesting depth is unknowable. + */ + #define rcu_preempt_depth() (current->rcu_read_lock_nesting) ++#ifndef CONFIG_PREEMPT_RT_FULL ++#define sched_rcu_preempt_depth() rcu_preempt_depth() ++#else ++static inline int sched_rcu_preempt_depth(void) { return 0; } ++#endif + + #else /* #ifdef CONFIG_PREEMPT_RCU */ + +@@ -283,6 +292,8 @@ + return 0; + } + ++#define sched_rcu_preempt_depth() rcu_preempt_depth() ++ + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + + /* Internal to kernel */ +@@ -463,7 +474,14 @@ + int debug_lockdep_rcu_enabled(void); + + int rcu_read_lock_held(void); ++#ifdef CONFIG_PREEMPT_RT_FULL ++static inline int rcu_read_lock_bh_held(void) ++{ ++ return rcu_read_lock_held(); ++} ++#else + int rcu_read_lock_bh_held(void); ++#endif + + /** + * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? +@@ -990,10 +1008,14 @@ + static inline void rcu_read_lock_bh(void) + { + local_bh_disable(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ rcu_read_lock(); ++#else + __acquire(RCU_BH); + rcu_lock_acquire(&rcu_bh_lock_map); + rcu_lockdep_assert(rcu_is_watching(), + "rcu_read_lock_bh() used illegally while idle"); ++#endif + } + + /* +@@ -1003,10 +1025,14 @@ + */ + static inline void rcu_read_unlock_bh(void) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ rcu_read_unlock(); ++#else + rcu_lockdep_assert(rcu_is_watching(), + "rcu_read_unlock_bh() used illegally while idle"); + rcu_lock_release(&rcu_bh_lock_map); + __release(RCU_BH); ++#endif + local_bh_enable(); + } + +diff -Nur linux-4.1.13.orig/include/linux/rcutree.h linux-4.1.13/include/linux/rcutree.h +--- linux-4.1.13.orig/include/linux/rcutree.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/rcutree.h 2015-11-29 09:23:09.609614185 +0100 +@@ -46,7 +46,11 @@ + rcu_note_context_switch(); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define synchronize_rcu_bh synchronize_rcu ++#else + void synchronize_rcu_bh(void); ++#endif + void synchronize_sched_expedited(void); + void synchronize_rcu_expedited(void); + +@@ -74,7 +78,11 @@ + } + + void rcu_barrier(void); ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define rcu_barrier_bh rcu_barrier ++#else + void rcu_barrier_bh(void); ++#endif + void rcu_barrier_sched(void); + unsigned long get_state_synchronize_rcu(void); + void cond_synchronize_rcu(unsigned long oldstate); +@@ -85,12 +93,10 @@ + unsigned long rcu_batches_started_bh(void); + unsigned long rcu_batches_started_sched(void); + unsigned long rcu_batches_completed(void); +-unsigned long rcu_batches_completed_bh(void); + unsigned long rcu_batches_completed_sched(void); + void show_rcu_gp_kthreads(void); + + void rcu_force_quiescent_state(void); +-void rcu_bh_force_quiescent_state(void); + void rcu_sched_force_quiescent_state(void); + + void exit_rcu(void); +@@ -100,6 +106,14 @@ + + bool rcu_is_watching(void); + ++#ifndef CONFIG_PREEMPT_RT_FULL ++void rcu_bh_force_quiescent_state(void); ++unsigned long rcu_batches_completed_bh(void); ++#else ++# define rcu_bh_force_quiescent_state rcu_force_quiescent_state ++# define rcu_batches_completed_bh rcu_batches_completed ++#endif ++ + void rcu_all_qs(void); + + #endif /* __LINUX_RCUTREE_H */ +diff -Nur linux-4.1.13.orig/include/linux/rtmutex.h linux-4.1.13/include/linux/rtmutex.h +--- linux-4.1.13.orig/include/linux/rtmutex.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/rtmutex.h 2015-11-29 09:23:09.609614185 +0100 +@@ -14,10 +14,14 @@ + + #include <linux/linkage.h> + #include <linux/rbtree.h> +-#include <linux/spinlock_types.h> ++#include <linux/spinlock_types_raw.h> + + extern int max_lock_depth; /* for sysctl */ + ++#ifdef CONFIG_DEBUG_MUTEXES ++#include <linux/debug_locks.h> ++#endif ++ + /** + * The rt_mutex structure + * +@@ -31,8 +35,8 @@ + struct rb_root waiters; + struct rb_node *waiters_leftmost; + struct task_struct *owner; +-#ifdef CONFIG_DEBUG_RT_MUTEXES + int save_state; ++#ifdef CONFIG_DEBUG_RT_MUTEXES + const char *name, *file; + int line; + void *magic; +@@ -55,22 +59,33 @@ + # define rt_mutex_debug_check_no_locks_held(task) do { } while (0) + #endif + ++# define rt_mutex_init(mutex) \ ++ do { \ ++ raw_spin_lock_init(&(mutex)->wait_lock); \ ++ __rt_mutex_init(mutex, #mutex); \ ++ } while (0) ++ + #ifdef CONFIG_DEBUG_RT_MUTEXES + # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) \ + , .name = #mutexname, .file = __FILE__, .line = __LINE__ +-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, __func__) + extern void rt_mutex_debug_task_free(struct task_struct *tsk); + #else + # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) +-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, NULL) + # define rt_mutex_debug_task_free(t) do { } while (0) + #endif + +-#define __RT_MUTEX_INITIALIZER(mutexname) \ +- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ ++#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \ + , .waiters = RB_ROOT \ + , .owner = NULL \ +- __DEBUG_RT_MUTEX_INITIALIZER(mutexname)} ++ __DEBUG_RT_MUTEX_INITIALIZER(mutexname) ++ ++#define __RT_MUTEX_INITIALIZER(mutexname) \ ++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) } ++ ++#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \ ++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \ ++ , .save_state = 1 } + + #define DEFINE_RT_MUTEX(mutexname) \ + struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname) +@@ -91,6 +106,7 @@ + + extern void rt_mutex_lock(struct rt_mutex *lock); + extern int rt_mutex_lock_interruptible(struct rt_mutex *lock); ++extern int rt_mutex_lock_killable(struct rt_mutex *lock); + extern int rt_mutex_timed_lock(struct rt_mutex *lock, + struct hrtimer_sleeper *timeout); + +diff -Nur linux-4.1.13.orig/include/linux/rwlock_rt.h linux-4.1.13/include/linux/rwlock_rt.h +--- linux-4.1.13.orig/include/linux/rwlock_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/rwlock_rt.h 2015-11-29 09:23:09.609614185 +0100 +@@ -0,0 +1,99 @@ ++#ifndef __LINUX_RWLOCK_RT_H ++#define __LINUX_RWLOCK_RT_H ++ ++#ifndef __LINUX_SPINLOCK_H ++#error Do not include directly. Use spinlock.h ++#endif ++ ++#define rwlock_init(rwl) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(rwl)->lock); \ ++ __rt_rwlock_init(rwl, #rwl, &__key); \ ++} while (0) ++ ++extern void __lockfunc rt_write_lock(rwlock_t *rwlock); ++extern void __lockfunc rt_read_lock(rwlock_t *rwlock); ++extern int __lockfunc rt_write_trylock(rwlock_t *rwlock); ++extern int __lockfunc rt_write_trylock_irqsave(rwlock_t *trylock, unsigned long *flags); ++extern int __lockfunc rt_read_trylock(rwlock_t *rwlock); ++extern void __lockfunc rt_write_unlock(rwlock_t *rwlock); ++extern void __lockfunc rt_read_unlock(rwlock_t *rwlock); ++extern unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock); ++extern unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock); ++extern void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key); ++ ++#define read_trylock(lock) __cond_lock(lock, rt_read_trylock(lock)) ++#define write_trylock(lock) __cond_lock(lock, rt_write_trylock(lock)) ++ ++#define write_trylock_irqsave(lock, flags) \ ++ __cond_lock(lock, rt_write_trylock_irqsave(lock, &flags)) ++ ++#define read_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = rt_read_lock_irqsave(lock); \ ++ } while (0) ++ ++#define write_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = rt_write_lock_irqsave(lock); \ ++ } while (0) ++ ++#define read_lock(lock) rt_read_lock(lock) ++ ++#define read_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_read_lock(lock); \ ++ } while (0) ++ ++#define read_lock_irq(lock) read_lock(lock) ++ ++#define write_lock(lock) rt_write_lock(lock) ++ ++#define write_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ rt_write_lock(lock); \ ++ } while (0) ++ ++#define write_lock_irq(lock) write_lock(lock) ++ ++#define read_unlock(lock) rt_read_unlock(lock) ++ ++#define read_unlock_bh(lock) \ ++ do { \ ++ rt_read_unlock(lock); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define read_unlock_irq(lock) read_unlock(lock) ++ ++#define write_unlock(lock) rt_write_unlock(lock) ++ ++#define write_unlock_bh(lock) \ ++ do { \ ++ rt_write_unlock(lock); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define write_unlock_irq(lock) write_unlock(lock) ++ ++#define read_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ rt_read_unlock(lock); \ ++ } while (0) ++ ++#define write_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ rt_write_unlock(lock); \ ++ } while (0) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/rwlock_types.h linux-4.1.13/include/linux/rwlock_types.h +--- linux-4.1.13.orig/include/linux/rwlock_types.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/rwlock_types.h 2015-11-29 09:23:09.609614185 +0100 +@@ -1,6 +1,10 @@ + #ifndef __LINUX_RWLOCK_TYPES_H + #define __LINUX_RWLOCK_TYPES_H + ++#if !defined(__LINUX_SPINLOCK_TYPES_H) ++# error "Do not include directly, include spinlock_types.h" ++#endif ++ + /* + * include/linux/rwlock_types.h - generic rwlock type definitions + * and initializers +@@ -43,6 +47,7 @@ + RW_DEP_MAP_INIT(lockname) } + #endif + +-#define DEFINE_RWLOCK(x) rwlock_t x = __RW_LOCK_UNLOCKED(x) ++#define DEFINE_RWLOCK(name) \ ++ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name) + + #endif /* __LINUX_RWLOCK_TYPES_H */ +diff -Nur linux-4.1.13.orig/include/linux/rwlock_types_rt.h linux-4.1.13/include/linux/rwlock_types_rt.h +--- linux-4.1.13.orig/include/linux/rwlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/rwlock_types_rt.h 2015-11-29 09:23:09.609614185 +0100 +@@ -0,0 +1,33 @@ ++#ifndef __LINUX_RWLOCK_TYPES_RT_H ++#define __LINUX_RWLOCK_TYPES_RT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++/* ++ * rwlocks - rtmutex which allows single reader recursion ++ */ ++typedef struct { ++ struct rt_mutex lock; ++ int read_depth; ++ unsigned int break_lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} rwlock_t; ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define RW_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } ++#else ++# define RW_DEP_MAP_INIT(lockname) ++#endif ++ ++#define __RW_LOCK_UNLOCKED(name) \ ++ { .lock = __RT_MUTEX_INITIALIZER_SAVE_STATE(name.lock), \ ++ RW_DEP_MAP_INIT(name) } ++ ++#define DEFINE_RWLOCK(name) \ ++ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/rwsem.h linux-4.1.13/include/linux/rwsem.h +--- linux-4.1.13.orig/include/linux/rwsem.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/rwsem.h 2015-11-29 09:23:09.609614185 +0100 +@@ -18,6 +18,10 @@ + #include <linux/osq_lock.h> + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++#include <linux/rwsem_rt.h> ++#else /* PREEMPT_RT_FULL */ ++ + struct rw_semaphore; + + #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK +@@ -177,4 +181,6 @@ + # define up_read_non_owner(sem) up_read(sem) + #endif + ++#endif /* !PREEMPT_RT_FULL */ ++ + #endif /* _LINUX_RWSEM_H */ +diff -Nur linux-4.1.13.orig/include/linux/rwsem_rt.h linux-4.1.13/include/linux/rwsem_rt.h +--- linux-4.1.13.orig/include/linux/rwsem_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/rwsem_rt.h 2015-11-29 09:23:09.609614185 +0100 +@@ -0,0 +1,140 @@ ++#ifndef _LINUX_RWSEM_RT_H ++#define _LINUX_RWSEM_RT_H ++ ++#ifndef _LINUX_RWSEM_H ++#error "Include rwsem.h" ++#endif ++ ++/* ++ * RW-semaphores are a spinlock plus a reader-depth count. ++ * ++ * Note that the semantics are different from the usual ++ * Linux rw-sems, in PREEMPT_RT mode we do not allow ++ * multiple readers to hold the lock at once, we only allow ++ * a read-lock owner to read-lock recursively. This is ++ * better for latency, makes the implementation inherently ++ * fair and makes it simpler as well. ++ */ ++ ++#include <linux/rtmutex.h> ++ ++struct rw_semaphore { ++ struct rt_mutex lock; ++ int read_depth; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++}; ++ ++#define __RWSEM_INITIALIZER(name) \ ++ { .lock = __RT_MUTEX_INITIALIZER(name.lock), \ ++ RW_DEP_MAP_INIT(name) } ++ ++#define DECLARE_RWSEM(lockname) \ ++ struct rw_semaphore lockname = __RWSEM_INITIALIZER(lockname) ++ ++extern void __rt_rwsem_init(struct rw_semaphore *rwsem, const char *name, ++ struct lock_class_key *key); ++ ++#define __rt_init_rwsem(sem, name, key) \ ++ do { \ ++ rt_mutex_init(&(sem)->lock); \ ++ __rt_rwsem_init((sem), (name), (key));\ ++ } while (0) ++ ++#define __init_rwsem(sem, name, key) __rt_init_rwsem(sem, name, key) ++ ++# define rt_init_rwsem(sem) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ __rt_init_rwsem((sem), #sem, &__key); \ ++} while (0) ++ ++extern void rt_down_write(struct rw_semaphore *rwsem); ++extern void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass); ++extern void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass); ++extern void rt_down_write_nested_lock(struct rw_semaphore *rwsem, ++ struct lockdep_map *nest); ++extern void rt_down_read(struct rw_semaphore *rwsem); ++extern int rt_down_write_trylock(struct rw_semaphore *rwsem); ++extern int rt_down_read_trylock(struct rw_semaphore *rwsem); ++extern void __rt_up_read(struct rw_semaphore *rwsem); ++extern void rt_up_read(struct rw_semaphore *rwsem); ++extern void rt_up_write(struct rw_semaphore *rwsem); ++extern void rt_downgrade_write(struct rw_semaphore *rwsem); ++ ++#define init_rwsem(sem) rt_init_rwsem(sem) ++#define rwsem_is_locked(s) rt_mutex_is_locked(&(s)->lock) ++ ++static inline int rwsem_is_contended(struct rw_semaphore *sem) ++{ ++ /* rt_mutex_has_waiters() */ ++ return !RB_EMPTY_ROOT(&sem->lock.waiters); ++} ++ ++static inline void down_read(struct rw_semaphore *sem) ++{ ++ rt_down_read(sem); ++} ++ ++static inline int down_read_trylock(struct rw_semaphore *sem) ++{ ++ return rt_down_read_trylock(sem); ++} ++ ++static inline void down_write(struct rw_semaphore *sem) ++{ ++ rt_down_write(sem); ++} ++ ++static inline int down_write_trylock(struct rw_semaphore *sem) ++{ ++ return rt_down_write_trylock(sem); ++} ++ ++static inline void __up_read(struct rw_semaphore *sem) ++{ ++ __rt_up_read(sem); ++} ++ ++static inline void up_read(struct rw_semaphore *sem) ++{ ++ rt_up_read(sem); ++} ++ ++static inline void up_write(struct rw_semaphore *sem) ++{ ++ rt_up_write(sem); ++} ++ ++static inline void downgrade_write(struct rw_semaphore *sem) ++{ ++ rt_downgrade_write(sem); ++} ++ ++static inline void down_read_nested(struct rw_semaphore *sem, int subclass) ++{ ++ return rt_down_read_nested(sem, subclass); ++} ++ ++static inline void down_write_nested(struct rw_semaphore *sem, int subclass) ++{ ++ rt_down_write_nested(sem, subclass); ++} ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++static inline void down_write_nest_lock(struct rw_semaphore *sem, ++ struct rw_semaphore *nest_lock) ++{ ++ rt_down_write_nested_lock(sem, &nest_lock->dep_map); ++} ++ ++#else ++ ++static inline void down_write_nest_lock(struct rw_semaphore *sem, ++ struct rw_semaphore *nest_lock) ++{ ++ rt_down_write_nested_lock(sem, NULL); ++} ++#endif ++#endif +diff -Nur linux-4.1.13.orig/include/linux/sched.h linux-4.1.13/include/linux/sched.h +--- linux-4.1.13.orig/include/linux/sched.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/sched.h 2015-11-29 09:23:09.617613651 +0100 +@@ -26,6 +26,7 @@ + #include <linux/nodemask.h> + #include <linux/mm_types.h> + #include <linux/preempt_mask.h> ++#include <asm/kmap_types.h> + + #include <asm/page.h> + #include <asm/ptrace.h> +@@ -234,10 +235,7 @@ + TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ + __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) + +-#define task_is_traced(task) ((task->state & __TASK_TRACED) != 0) + #define task_is_stopped(task) ((task->state & __TASK_STOPPED) != 0) +-#define task_is_stopped_or_traced(task) \ +- ((task->state & (__TASK_STOPPED | __TASK_TRACED)) != 0) + #define task_contributes_to_load(task) \ + ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \ + (task->flags & PF_FROZEN) == 0) +@@ -302,6 +300,11 @@ + + #endif + ++#define __set_current_state_no_track(state_value) \ ++ do { current->state = (state_value); } while (0) ++#define set_current_state_no_track(state_value) \ ++ set_mb(current->state, (state_value)) ++ + /* Task command name length */ + #define TASK_COMM_LEN 16 + +@@ -900,6 +903,50 @@ + #define SCHED_CAPACITY_SCALE (1L << SCHED_CAPACITY_SHIFT) + + /* ++ * Wake-queues are lists of tasks with a pending wakeup, whose ++ * callers have already marked the task as woken internally, ++ * and can thus carry on. A common use case is being able to ++ * do the wakeups once the corresponding user lock as been ++ * released. ++ * ++ * We hold reference to each task in the list across the wakeup, ++ * thus guaranteeing that the memory is still valid by the time ++ * the actual wakeups are performed in wake_up_q(). ++ * ++ * One per task suffices, because there's never a need for a task to be ++ * in two wake queues simultaneously; it is forbidden to abandon a task ++ * in a wake queue (a call to wake_up_q() _must_ follow), so if a task is ++ * already in a wake queue, the wakeup will happen soon and the second ++ * waker can just skip it. ++ * ++ * The WAKE_Q macro declares and initializes the list head. ++ * wake_up_q() does NOT reinitialize the list; it's expected to be ++ * called near the end of a function, where the fact that the queue is ++ * not used again will be easy to see by inspection. ++ * ++ * Note that this can cause spurious wakeups. schedule() callers ++ * must ensure the call is done inside a loop, confirming that the ++ * wakeup condition has in fact occurred. ++ */ ++struct wake_q_node { ++ struct wake_q_node *next; ++}; ++ ++struct wake_q_head { ++ struct wake_q_node *first; ++ struct wake_q_node **lastp; ++}; ++ ++#define WAKE_Q_TAIL ((struct wake_q_node *) 0x01) ++ ++#define WAKE_Q(name) \ ++ struct wake_q_head name = { WAKE_Q_TAIL, &name.first } ++ ++extern void wake_q_add(struct wake_q_head *head, ++ struct task_struct *task); ++extern void wake_up_q(struct wake_q_head *head); ++ ++/* + * sched-domains (multiprocessor balancing) declarations: + */ + #ifdef CONFIG_SMP +@@ -1291,6 +1338,7 @@ + + struct task_struct { + volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */ ++ volatile long saved_state; /* saved state for "spinlock sleepers" */ + void *stack; + atomic_t usage; + unsigned int flags; /* per process flags, defined below */ +@@ -1327,6 +1375,12 @@ + #endif + + unsigned int policy; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ int migrate_disable; ++# ifdef CONFIG_SCHED_DEBUG ++ int migrate_disable_atomic; ++# endif ++#endif + int nr_cpus_allowed; + cpumask_t cpus_allowed; + +@@ -1434,7 +1488,8 @@ + struct cputime prev_cputime; + #endif + #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN +- seqlock_t vtime_seqlock; ++ raw_spinlock_t vtime_lock; ++ seqcount_t vtime_seq; + unsigned long long vtime_snap; + enum { + VTIME_SLEEPING = 0, +@@ -1450,6 +1505,9 @@ + + struct task_cputime cputime_expires; + struct list_head cpu_timers[3]; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct task_struct *posix_timer_list; ++#endif + + /* process credentials */ + const struct cred __rcu *real_cred; /* objective and real subjective task +@@ -1482,10 +1540,15 @@ + /* signal handlers */ + struct signal_struct *signal; + struct sighand_struct *sighand; ++ struct sigqueue *sigqueue_cache; + + sigset_t blocked, real_blocked; + sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */ + struct sigpending pending; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* TODO: move me into ->restart_block ? */ ++ struct siginfo forced_info; ++#endif + + unsigned long sas_ss_sp; + size_t sas_ss_size; +@@ -1511,6 +1574,8 @@ + /* Protection of the PI data structures: */ + raw_spinlock_t pi_lock; + ++ struct wake_q_node wake_q; ++ + #ifdef CONFIG_RT_MUTEXES + /* PI waiters blocked on a rt_mutex held by this task */ + struct rb_root pi_waiters; +@@ -1705,6 +1770,12 @@ + unsigned long trace; + /* bitmask and counter of trace recursion */ + unsigned long trace_recursion; ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ u64 preempt_timestamp_hist; ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ long timer_offset; ++#endif ++#endif + #endif /* CONFIG_TRACING */ + #ifdef CONFIG_MEMCG + struct memcg_oom_info { +@@ -1721,14 +1792,23 @@ + unsigned int sequential_io; + unsigned int sequential_io_avg; + #endif ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct rcu_head put_rcu; ++ int softirq_nestcnt; ++ unsigned int softirqs_raised; ++#endif ++#ifdef CONFIG_PREEMPT_RT_FULL ++# if defined CONFIG_HIGHMEM || defined CONFIG_X86_32 ++ int kmap_idx; ++ pte_t kmap_pte[KM_TYPE_NR]; ++# endif ++#endif + #ifdef CONFIG_DEBUG_ATOMIC_SLEEP + unsigned long task_state_change; + #endif ++ int pagefault_disabled; + }; + +-/* Future-safe accessor for struct task_struct's cpus_allowed. */ +-#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed) +- + #define TNF_MIGRATED 0x01 + #define TNF_NO_GROUP 0x02 + #define TNF_SHARED 0x04 +@@ -1917,6 +1997,15 @@ + extern void free_task(struct task_struct *tsk); + #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0) + ++#ifdef CONFIG_PREEMPT_RT_BASE ++extern void __put_task_struct_cb(struct rcu_head *rhp); ++ ++static inline void put_task_struct(struct task_struct *t) ++{ ++ if (atomic_dec_and_test(&t->usage)) ++ call_rcu(&t->put_rcu, __put_task_struct_cb); ++} ++#else + extern void __put_task_struct(struct task_struct *t); + + static inline void put_task_struct(struct task_struct *t) +@@ -1924,6 +2013,7 @@ + if (atomic_dec_and_test(&t->usage)) + __put_task_struct(t); + } ++#endif + + #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN + extern void task_cputime(struct task_struct *t, +@@ -1962,6 +2052,7 @@ + /* + * Per process flags + */ ++#define PF_IN_SOFTIRQ 0x00000001 /* Task is serving softirq */ + #define PF_EXITING 0x00000004 /* getting shut down */ + #define PF_EXITPIDONE 0x00000008 /* pi exit done on shut down */ + #define PF_VCPU 0x00000010 /* I'm a virtual CPU */ +@@ -2126,6 +2217,10 @@ + + extern int set_cpus_allowed_ptr(struct task_struct *p, + const struct cpumask *new_mask); ++int migrate_me(void); ++void tell_sched_cpu_down_begin(int cpu); ++void tell_sched_cpu_down_done(int cpu); ++ + #else + static inline void do_set_cpus_allowed(struct task_struct *p, + const struct cpumask *new_mask) +@@ -2138,6 +2233,9 @@ + return -EINVAL; + return 0; + } ++static inline int migrate_me(void) { return 0; } ++static inline void tell_sched_cpu_down_begin(int cpu) { } ++static inline void tell_sched_cpu_down_done(int cpu) { } + #endif + + #ifdef CONFIG_NO_HZ_COMMON +@@ -2354,6 +2452,7 @@ + + extern int wake_up_state(struct task_struct *tsk, unsigned int state); + extern int wake_up_process(struct task_struct *tsk); ++extern int wake_up_lock_sleeper(struct task_struct * tsk); + extern void wake_up_new_task(struct task_struct *tsk); + #ifdef CONFIG_SMP + extern void kick_process(struct task_struct *tsk); +@@ -2470,12 +2569,24 @@ + + /* mmdrop drops the mm and the page tables */ + extern void __mmdrop(struct mm_struct *); ++ + static inline void mmdrop(struct mm_struct * mm) + { + if (unlikely(atomic_dec_and_test(&mm->mm_count))) + __mmdrop(mm); + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++extern void __mmdrop_delayed(struct rcu_head *rhp); ++static inline void mmdrop_delayed(struct mm_struct *mm) ++{ ++ if (atomic_dec_and_test(&mm->mm_count)) ++ call_rcu(&mm->delayed_drop, __mmdrop_delayed); ++} ++#else ++# define mmdrop_delayed(mm) mmdrop(mm) ++#endif ++ + /* mmput gets rid of the mappings and all user-space */ + extern void mmput(struct mm_struct *); + /* Grab a reference to a task's mm, if it is not already going away */ +@@ -2787,6 +2898,43 @@ + return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); + } + ++#ifdef CONFIG_PREEMPT_LAZY ++static inline void set_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ set_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline int test_tsk_need_resched_lazy(struct task_struct *tsk) ++{ ++ return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED_LAZY)); ++} ++ ++static inline int need_resched_lazy(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED_LAZY); ++} ++ ++static inline int need_resched_now(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED); ++} ++ ++#else ++static inline void clear_tsk_need_resched_lazy(struct task_struct *tsk) { } ++static inline int need_resched_lazy(void) { return 0; } ++ ++static inline int need_resched_now(void) ++{ ++ return test_thread_flag(TIF_NEED_RESCHED); ++} ++ ++#endif ++ + static inline int restart_syscall(void) + { + set_tsk_thread_flag(current, TIF_SIGPENDING); +@@ -2818,6 +2966,51 @@ + return (state & TASK_INTERRUPTIBLE) || __fatal_signal_pending(p); + } + ++static inline bool __task_is_stopped_or_traced(struct task_struct *task) ++{ ++ if (task->state & (__TASK_STOPPED | __TASK_TRACED)) ++ return true; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (task->saved_state & (__TASK_STOPPED | __TASK_TRACED)) ++ return true; ++#endif ++ return false; ++} ++ ++static inline bool task_is_stopped_or_traced(struct task_struct *task) ++{ ++ bool traced_stopped; ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&task->pi_lock, flags); ++ traced_stopped = __task_is_stopped_or_traced(task); ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); ++#else ++ traced_stopped = __task_is_stopped_or_traced(task); ++#endif ++ return traced_stopped; ++} ++ ++static inline bool task_is_traced(struct task_struct *task) ++{ ++ bool traced = false; ++ ++ if (task->state & __TASK_TRACED) ++ return true; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* in case the task is sleeping on tasklist_lock */ ++ raw_spin_lock_irq(&task->pi_lock); ++ if (task->state & __TASK_TRACED) ++ traced = true; ++ else if (task->saved_state & __TASK_TRACED) ++ traced = true; ++ raw_spin_unlock_irq(&task->pi_lock); ++#endif ++ return traced; ++} ++ + /* + * cond_resched() and cond_resched_lock(): latency reduction via + * explicit rescheduling in places that are safe. The return +@@ -2839,12 +3032,16 @@ + __cond_resched_lock(lock); \ + }) + ++#ifndef CONFIG_PREEMPT_RT_FULL + extern int __cond_resched_softirq(void); + + #define cond_resched_softirq() ({ \ + ___might_sleep(__FILE__, __LINE__, SOFTIRQ_DISABLE_OFFSET); \ + __cond_resched_softirq(); \ + }) ++#else ++# define cond_resched_softirq() cond_resched() ++#endif + + static inline void cond_resched_rcu(void) + { +@@ -3011,6 +3208,26 @@ + + #endif /* CONFIG_SMP */ + ++static inline int __migrate_disabled(struct task_struct *p) ++{ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ return p->migrate_disable; ++#else ++ return 0; ++#endif ++} ++ ++/* Future-safe accessor for struct task_struct's cpus_allowed. */ ++static inline const struct cpumask *tsk_cpus_allowed(struct task_struct *p) ++{ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (p->migrate_disable) ++ return cpumask_of(task_cpu(p)); ++#endif ++ ++ return &p->cpus_allowed; ++} ++ + extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask); + extern long sched_getaffinity(pid_t pid, struct cpumask *mask); + +diff -Nur linux-4.1.13.orig/include/linux/seqlock.h linux-4.1.13/include/linux/seqlock.h +--- linux-4.1.13.orig/include/linux/seqlock.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/seqlock.h 2015-11-29 09:23:09.617613651 +0100 +@@ -219,20 +219,30 @@ + return __read_seqcount_retry(s, start); + } + +- +- +-static inline void raw_write_seqcount_begin(seqcount_t *s) ++static inline void __raw_write_seqcount_begin(seqcount_t *s) + { + s->sequence++; + smp_wmb(); + } + +-static inline void raw_write_seqcount_end(seqcount_t *s) ++static inline void raw_write_seqcount_begin(seqcount_t *s) ++{ ++ preempt_disable_rt(); ++ __raw_write_seqcount_begin(s); ++} ++ ++static inline void __raw_write_seqcount_end(seqcount_t *s) + { + smp_wmb(); + s->sequence++; + } + ++static inline void raw_write_seqcount_end(seqcount_t *s) ++{ ++ __raw_write_seqcount_end(s); ++ preempt_enable_rt(); ++} ++ + /* + * raw_write_seqcount_latch - redirect readers to even/odd copy + * @s: pointer to seqcount_t +@@ -305,10 +315,32 @@ + /* + * Read side functions for starting and finalizing a read side section. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + static inline unsigned read_seqbegin(const seqlock_t *sl) + { + return read_seqcount_begin(&sl->seqcount); + } ++#else ++/* ++ * Starvation safe read side for RT ++ */ ++static inline unsigned read_seqbegin(seqlock_t *sl) ++{ ++ unsigned ret; ++ ++repeat: ++ ret = ACCESS_ONCE(sl->seqcount.sequence); ++ if (unlikely(ret & 1)) { ++ /* ++ * Take the lock and let the writer proceed (i.e. evtl ++ * boost it), otherwise we could loop here forever. ++ */ ++ spin_unlock_wait(&sl->lock); ++ goto repeat; ++ } ++ return ret; ++} ++#endif + + static inline unsigned read_seqretry(const seqlock_t *sl, unsigned start) + { +@@ -323,36 +355,36 @@ + static inline void write_seqlock(seqlock_t *sl) + { + spin_lock(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + } + + static inline void write_sequnlock(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock(&sl->lock); + } + + static inline void write_seqlock_bh(seqlock_t *sl) + { + spin_lock_bh(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + } + + static inline void write_sequnlock_bh(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_bh(&sl->lock); + } + + static inline void write_seqlock_irq(seqlock_t *sl) + { + spin_lock_irq(&sl->lock); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + } + + static inline void write_sequnlock_irq(seqlock_t *sl) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_irq(&sl->lock); + } + +@@ -361,7 +393,7 @@ + unsigned long flags; + + spin_lock_irqsave(&sl->lock, flags); +- write_seqcount_begin(&sl->seqcount); ++ __raw_write_seqcount_begin(&sl->seqcount); + return flags; + } + +@@ -371,7 +403,7 @@ + static inline void + write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags) + { +- write_seqcount_end(&sl->seqcount); ++ __raw_write_seqcount_end(&sl->seqcount); + spin_unlock_irqrestore(&sl->lock, flags); + } + +diff -Nur linux-4.1.13.orig/include/linux/signal.h linux-4.1.13/include/linux/signal.h +--- linux-4.1.13.orig/include/linux/signal.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/signal.h 2015-11-29 09:23:09.617613651 +0100 +@@ -218,6 +218,7 @@ + } + + extern void flush_sigqueue(struct sigpending *queue); ++extern void flush_task_sigqueue(struct task_struct *tsk); + + /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */ + static inline int valid_signal(unsigned long sig) +diff -Nur linux-4.1.13.orig/include/linux/skbuff.h linux-4.1.13/include/linux/skbuff.h +--- linux-4.1.13.orig/include/linux/skbuff.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/skbuff.h 2015-11-29 09:23:09.617613651 +0100 +@@ -187,6 +187,7 @@ + + __u32 qlen; + spinlock_t lock; ++ raw_spinlock_t raw_lock; + }; + + struct sk_buff; +@@ -1336,6 +1337,12 @@ + __skb_queue_head_init(list); + } + ++static inline void skb_queue_head_init_raw(struct sk_buff_head *list) ++{ ++ raw_spin_lock_init(&list->raw_lock); ++ __skb_queue_head_init(list); ++} ++ + static inline void skb_queue_head_init_class(struct sk_buff_head *list, + struct lock_class_key *class) + { +diff -Nur linux-4.1.13.orig/include/linux/smp.h linux-4.1.13/include/linux/smp.h +--- linux-4.1.13.orig/include/linux/smp.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/smp.h 2015-11-29 09:23:09.617613651 +0100 +@@ -185,6 +185,9 @@ + #define get_cpu() ({ preempt_disable(); smp_processor_id(); }) + #define put_cpu() preempt_enable() + ++#define get_cpu_light() ({ migrate_disable(); smp_processor_id(); }) ++#define put_cpu_light() migrate_enable() ++ + /* + * Callback to arch code if there's nosmp or maxcpus=0 on the + * boot command line: +diff -Nur linux-4.1.13.orig/include/linux/spinlock_api_smp.h linux-4.1.13/include/linux/spinlock_api_smp.h +--- linux-4.1.13.orig/include/linux/spinlock_api_smp.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_api_smp.h 2015-11-29 09:23:09.617613651 +0100 +@@ -189,6 +189,8 @@ + return 0; + } + +-#include <linux/rwlock_api_smp.h> ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include <linux/rwlock_api_smp.h> ++#endif + + #endif /* __LINUX_SPINLOCK_API_SMP_H */ +diff -Nur linux-4.1.13.orig/include/linux/spinlock.h linux-4.1.13/include/linux/spinlock.h +--- linux-4.1.13.orig/include/linux/spinlock.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock.h 2015-11-29 09:23:09.617613651 +0100 +@@ -281,7 +281,11 @@ + #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock)) + + /* Include rwlock functions */ +-#include <linux/rwlock.h> ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include <linux/rwlock_rt.h> ++#else ++# include <linux/rwlock.h> ++#endif + + /* + * Pull the _spin_*()/_read_*()/_write_*() functions/declarations: +@@ -292,6 +296,10 @@ + # include <linux/spinlock_api_up.h> + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++# include <linux/spinlock_rt.h> ++#else /* PREEMPT_RT_FULL */ ++ + /* + * Map the spin_lock functions to the raw variants for PREEMPT_RT=n + */ +@@ -426,4 +434,6 @@ + #define atomic_dec_and_lock(atomic, lock) \ + __cond_lock(lock, _atomic_dec_and_lock(atomic, lock)) + ++#endif /* !PREEMPT_RT_FULL */ ++ + #endif /* __LINUX_SPINLOCK_H */ +diff -Nur linux-4.1.13.orig/include/linux/spinlock_rt.h linux-4.1.13/include/linux/spinlock_rt.h +--- linux-4.1.13.orig/include/linux/spinlock_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_rt.h 2015-11-29 09:23:09.617613651 +0100 +@@ -0,0 +1,174 @@ ++#ifndef __LINUX_SPINLOCK_RT_H ++#define __LINUX_SPINLOCK_RT_H ++ ++#ifndef __LINUX_SPINLOCK_H ++#error Do not include directly. Use spinlock.h ++#endif ++ ++#include <linux/bug.h> ++ ++extern void ++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key); ++ ++#define spin_lock_init(slock) \ ++do { \ ++ static struct lock_class_key __key; \ ++ \ ++ rt_mutex_init(&(slock)->lock); \ ++ __rt_spin_lock_init(slock, #slock, &__key); \ ++} while (0) ++ ++extern void __lockfunc rt_spin_lock(spinlock_t *lock); ++extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock); ++extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass); ++extern void __lockfunc rt_spin_unlock(spinlock_t *lock); ++extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock); ++extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags); ++extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock); ++extern int __lockfunc rt_spin_trylock(spinlock_t *lock); ++extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock); ++ ++/* ++ * lockdep-less calls, for derived types like rwlock: ++ * (for trylock they can use rt_mutex_trylock() directly. ++ */ ++extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock); ++extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock); ++extern int __lockfunc __rt_spin_trylock(struct rt_mutex *lock); ++ ++#define spin_lock(lock) \ ++ do { \ ++ migrate_disable(); \ ++ rt_spin_lock(lock); \ ++ } while (0) ++ ++#define spin_lock_bh(lock) \ ++ do { \ ++ local_bh_disable(); \ ++ migrate_disable(); \ ++ rt_spin_lock(lock); \ ++ } while (0) ++ ++#define spin_lock_irq(lock) spin_lock(lock) ++ ++#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock)) ++ ++#define spin_trylock(lock) \ ++({ \ ++ int __locked; \ ++ migrate_disable(); \ ++ __locked = spin_do_trylock(lock); \ ++ if (!__locked) \ ++ migrate_enable(); \ ++ __locked; \ ++}) ++ ++#ifdef CONFIG_LOCKDEP ++# define spin_lock_nested(lock, subclass) \ ++ do { \ ++ migrate_disable(); \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++ ++#define spin_lock_bh_nested(lock, subclass) \ ++ do { \ ++ local_bh_disable(); \ ++ migrate_disable(); \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++ ++# define spin_lock_irqsave_nested(lock, flags, subclass) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ migrate_disable(); \ ++ rt_spin_lock_nested(lock, subclass); \ ++ } while (0) ++#else ++# define spin_lock_nested(lock, subclass) spin_lock(lock) ++# define spin_lock_bh_nested(lock, subclass) spin_lock_bh(lock) ++ ++# define spin_lock_irqsave_nested(lock, flags, subclass) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ spin_lock(lock); \ ++ } while (0) ++#endif ++ ++#define spin_lock_irqsave(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ flags = 0; \ ++ spin_lock(lock); \ ++ } while (0) ++ ++static inline unsigned long spin_lock_trace_flags(spinlock_t *lock) ++{ ++ unsigned long flags = 0; ++#ifdef CONFIG_TRACE_IRQFLAGS ++ flags = rt_spin_lock_trace_flags(lock); ++#else ++ spin_lock(lock); /* lock_local */ ++#endif ++ return flags; ++} ++ ++/* FIXME: we need rt_spin_lock_nest_lock */ ++#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0) ++ ++#define spin_unlock(lock) \ ++ do { \ ++ rt_spin_unlock(lock); \ ++ migrate_enable(); \ ++ } while (0) ++ ++#define spin_unlock_bh(lock) \ ++ do { \ ++ rt_spin_unlock(lock); \ ++ migrate_enable(); \ ++ local_bh_enable(); \ ++ } while (0) ++ ++#define spin_unlock_irq(lock) spin_unlock(lock) ++ ++#define spin_unlock_irqrestore(lock, flags) \ ++ do { \ ++ typecheck(unsigned long, flags); \ ++ (void) flags; \ ++ spin_unlock(lock); \ ++ } while (0) ++ ++#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock)) ++#define spin_trylock_irq(lock) spin_trylock(lock) ++ ++#define spin_trylock_irqsave(lock, flags) \ ++ rt_spin_trylock_irqsave(lock, &(flags)) ++ ++#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock) ++ ++#ifdef CONFIG_GENERIC_LOCKBREAK ++# define spin_is_contended(lock) ((lock)->break_lock) ++#else ++# define spin_is_contended(lock) (((void)(lock), 0)) ++#endif ++ ++static inline int spin_can_lock(spinlock_t *lock) ++{ ++ return !rt_mutex_is_locked(&lock->lock); ++} ++ ++static inline int spin_is_locked(spinlock_t *lock) ++{ ++ return rt_mutex_is_locked(&lock->lock); ++} ++ ++static inline void assert_spin_locked(spinlock_t *lock) ++{ ++ BUG_ON(!spin_is_locked(lock)); ++} ++ ++#define atomic_dec_and_lock(atomic, lock) \ ++ atomic_dec_and_spin_lock(atomic, lock) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/spinlock_types.h linux-4.1.13/include/linux/spinlock_types.h +--- linux-4.1.13.orig/include/linux/spinlock_types.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_types.h 2015-11-29 09:23:09.617613651 +0100 +@@ -9,80 +9,15 @@ + * Released under the General Public License (GPL). + */ + +-#if defined(CONFIG_SMP) +-# include <asm/spinlock_types.h> +-#else +-# include <linux/spinlock_types_up.h> +-#endif +- +-#include <linux/lockdep.h> +- +-typedef struct raw_spinlock { +- arch_spinlock_t raw_lock; +-#ifdef CONFIG_GENERIC_LOCKBREAK +- unsigned int break_lock; +-#endif +-#ifdef CONFIG_DEBUG_SPINLOCK +- unsigned int magic, owner_cpu; +- void *owner; +-#endif +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +- struct lockdep_map dep_map; +-#endif +-} raw_spinlock_t; +- +-#define SPINLOCK_MAGIC 0xdead4ead +- +-#define SPINLOCK_OWNER_INIT ((void *)-1L) +- +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } +-#else +-# define SPIN_DEP_MAP_INIT(lockname) +-#endif ++#include <linux/spinlock_types_raw.h> + +-#ifdef CONFIG_DEBUG_SPINLOCK +-# define SPIN_DEBUG_INIT(lockname) \ +- .magic = SPINLOCK_MAGIC, \ +- .owner_cpu = -1, \ +- .owner = SPINLOCK_OWNER_INIT, ++#ifndef CONFIG_PREEMPT_RT_FULL ++# include <linux/spinlock_types_nort.h> ++# include <linux/rwlock_types.h> + #else +-# define SPIN_DEBUG_INIT(lockname) ++# include <linux/rtmutex.h> ++# include <linux/spinlock_types_rt.h> ++# include <linux/rwlock_types_rt.h> + #endif + +-#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ +- { \ +- .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ +- SPIN_DEBUG_INIT(lockname) \ +- SPIN_DEP_MAP_INIT(lockname) } +- +-#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ +- (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) +- +-#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) +- +-typedef struct spinlock { +- union { +- struct raw_spinlock rlock; +- +-#ifdef CONFIG_DEBUG_LOCK_ALLOC +-# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) +- struct { +- u8 __padding[LOCK_PADSIZE]; +- struct lockdep_map dep_map; +- }; +-#endif +- }; +-} spinlock_t; +- +-#define __SPIN_LOCK_INITIALIZER(lockname) \ +- { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } +- +-#define __SPIN_LOCK_UNLOCKED(lockname) \ +- (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) +- +-#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) +- +-#include <linux/rwlock_types.h> +- + #endif /* __LINUX_SPINLOCK_TYPES_H */ +diff -Nur linux-4.1.13.orig/include/linux/spinlock_types_nort.h linux-4.1.13/include/linux/spinlock_types_nort.h +--- linux-4.1.13.orig/include/linux/spinlock_types_nort.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_types_nort.h 2015-11-29 09:23:09.617613651 +0100 +@@ -0,0 +1,33 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_NORT_H ++#define __LINUX_SPINLOCK_TYPES_NORT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++/* ++ * The non RT version maps spinlocks to raw_spinlocks ++ */ ++typedef struct spinlock { ++ union { ++ struct raw_spinlock rlock; ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map)) ++ struct { ++ u8 __padding[LOCK_PADSIZE]; ++ struct lockdep_map dep_map; ++ }; ++#endif ++ }; ++} spinlock_t; ++ ++#define __SPIN_LOCK_INITIALIZER(lockname) \ ++ { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } } ++ ++#define __SPIN_LOCK_UNLOCKED(lockname) \ ++ (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname) ++ ++#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/spinlock_types_raw.h linux-4.1.13/include/linux/spinlock_types_raw.h +--- linux-4.1.13.orig/include/linux/spinlock_types_raw.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_types_raw.h 2015-11-29 09:23:09.617613651 +0100 +@@ -0,0 +1,56 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_RAW_H ++#define __LINUX_SPINLOCK_TYPES_RAW_H ++ ++#if defined(CONFIG_SMP) ++# include <asm/spinlock_types.h> ++#else ++# include <linux/spinlock_types_up.h> ++#endif ++ ++#include <linux/lockdep.h> ++ ++typedef struct raw_spinlock { ++ arch_spinlock_t raw_lock; ++#ifdef CONFIG_GENERIC_LOCKBREAK ++ unsigned int break_lock; ++#endif ++#ifdef CONFIG_DEBUG_SPINLOCK ++ unsigned int magic, owner_cpu; ++ void *owner; ++#endif ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} raw_spinlock_t; ++ ++#define SPINLOCK_MAGIC 0xdead4ead ++ ++#define SPINLOCK_OWNER_INIT ((void *)-1L) ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname } ++#else ++# define SPIN_DEP_MAP_INIT(lockname) ++#endif ++ ++#ifdef CONFIG_DEBUG_SPINLOCK ++# define SPIN_DEBUG_INIT(lockname) \ ++ .magic = SPINLOCK_MAGIC, \ ++ .owner_cpu = -1, \ ++ .owner = SPINLOCK_OWNER_INIT, ++#else ++# define SPIN_DEBUG_INIT(lockname) ++#endif ++ ++#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \ ++ { \ ++ .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \ ++ SPIN_DEBUG_INIT(lockname) \ ++ SPIN_DEP_MAP_INIT(lockname) } ++ ++#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \ ++ (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname) ++ ++#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/spinlock_types_rt.h linux-4.1.13/include/linux/spinlock_types_rt.h +--- linux-4.1.13.orig/include/linux/spinlock_types_rt.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/spinlock_types_rt.h 2015-11-29 09:23:09.617613651 +0100 +@@ -0,0 +1,51 @@ ++#ifndef __LINUX_SPINLOCK_TYPES_RT_H ++#define __LINUX_SPINLOCK_TYPES_RT_H ++ ++#ifndef __LINUX_SPINLOCK_TYPES_H ++#error "Do not include directly. Include spinlock_types.h instead" ++#endif ++ ++#include <linux/cache.h> ++ ++/* ++ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field: ++ */ ++typedef struct spinlock { ++ struct rt_mutex lock; ++ unsigned int break_lock; ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ struct lockdep_map dep_map; ++#endif ++} spinlock_t; ++ ++#ifdef CONFIG_DEBUG_RT_MUTEXES ++# define __RT_SPIN_INITIALIZER(name) \ ++ { \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ ++ .save_state = 1, \ ++ .file = __FILE__, \ ++ .line = __LINE__ , \ ++ } ++#else ++# define __RT_SPIN_INITIALIZER(name) \ ++ { \ ++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \ ++ .save_state = 1, \ ++ } ++#endif ++ ++/* ++.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock) ++*/ ++ ++#define __SPIN_LOCK_UNLOCKED(name) \ ++ { .lock = __RT_SPIN_INITIALIZER(name.lock), \ ++ SPIN_DEP_MAP_INIT(name) } ++ ++#define __DEFINE_SPINLOCK(name) \ ++ spinlock_t name = __SPIN_LOCK_UNLOCKED(name) ++ ++#define DEFINE_SPINLOCK(name) \ ++ spinlock_t name __cacheline_aligned_in_smp = __SPIN_LOCK_UNLOCKED(name) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/srcu.h linux-4.1.13/include/linux/srcu.h +--- linux-4.1.13.orig/include/linux/srcu.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/srcu.h 2015-11-29 09:23:09.617613651 +0100 +@@ -84,10 +84,10 @@ + + void process_srcu(struct work_struct *work); + +-#define __SRCU_STRUCT_INIT(name) \ ++#define __SRCU_STRUCT_INIT(name, pcpu_name) \ + { \ + .completed = -300, \ +- .per_cpu_ref = &name##_srcu_array, \ ++ .per_cpu_ref = &pcpu_name, \ + .queue_lock = __SPIN_LOCK_UNLOCKED(name.queue_lock), \ + .running = false, \ + .batch_queue = RCU_BATCH_INIT(name.batch_queue), \ +@@ -104,7 +104,7 @@ + */ + #define __DEFINE_SRCU(name, is_static) \ + static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ +- is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name) ++ is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name, name##_srcu_array) + #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) + #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) + +diff -Nur linux-4.1.13.orig/include/linux/swap.h linux-4.1.13/include/linux/swap.h +--- linux-4.1.13.orig/include/linux/swap.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/swap.h 2015-11-29 09:23:09.617613651 +0100 +@@ -11,6 +11,7 @@ + #include <linux/fs.h> + #include <linux/atomic.h> + #include <linux/page-flags.h> ++#include <linux/locallock.h> + #include <asm/page.h> + + struct notifier_block; +@@ -252,7 +253,8 @@ + void *workingset_eviction(struct address_space *mapping, struct page *page); + bool workingset_refault(void *shadow); + void workingset_activation(struct page *page); +-extern struct list_lru workingset_shadow_nodes; ++extern struct list_lru __workingset_shadow_nodes; ++DECLARE_LOCAL_IRQ_LOCK(workingset_shadow_lock); + + static inline unsigned int workingset_node_pages(struct radix_tree_node *node) + { +@@ -296,6 +298,7 @@ + + + /* linux/mm/swap.c */ ++DECLARE_LOCAL_IRQ_LOCK(swapvec_lock); + extern void lru_cache_add(struct page *); + extern void lru_cache_add_anon(struct page *page); + extern void lru_cache_add_file(struct page *page); +diff -Nur linux-4.1.13.orig/include/linux/thread_info.h linux-4.1.13/include/linux/thread_info.h +--- linux-4.1.13.orig/include/linux/thread_info.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/thread_info.h 2015-11-29 09:23:09.617613651 +0100 +@@ -102,7 +102,17 @@ + #define test_thread_flag(flag) \ + test_ti_thread_flag(current_thread_info(), flag) + +-#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) ++#ifdef CONFIG_PREEMPT_LAZY ++#define tif_need_resched() (test_thread_flag(TIF_NEED_RESCHED) || \ ++ test_thread_flag(TIF_NEED_RESCHED_LAZY)) ++#define tif_need_resched_now() (test_thread_flag(TIF_NEED_RESCHED)) ++#define tif_need_resched_lazy() test_thread_flag(TIF_NEED_RESCHED_LAZY)) ++ ++#else ++#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) ++#define tif_need_resched_now() test_thread_flag(TIF_NEED_RESCHED) ++#define tif_need_resched_lazy() 0 ++#endif + + #if defined TIF_RESTORE_SIGMASK && !defined HAVE_SET_RESTORE_SIGMASK + /* +diff -Nur linux-4.1.13.orig/include/linux/timer.h linux-4.1.13/include/linux/timer.h +--- linux-4.1.13.orig/include/linux/timer.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/timer.h 2015-11-29 09:23:09.617613651 +0100 +@@ -241,7 +241,7 @@ + + extern int try_to_del_timer_sync(struct timer_list *timer); + +-#ifdef CONFIG_SMP ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + extern int del_timer_sync(struct timer_list *timer); + #else + # define del_timer_sync(t) del_timer(t) +diff -Nur linux-4.1.13.orig/include/linux/uaccess.h linux-4.1.13/include/linux/uaccess.h +--- linux-4.1.13.orig/include/linux/uaccess.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/uaccess.h 2015-11-29 09:23:09.617613651 +0100 +@@ -1,21 +1,31 @@ + #ifndef __LINUX_UACCESS_H__ + #define __LINUX_UACCESS_H__ + +-#include <linux/preempt.h> ++#include <linux/sched.h> + #include <asm/uaccess.h> + ++static __always_inline void pagefault_disabled_inc(void) ++{ ++ current->pagefault_disabled++; ++} ++ ++static __always_inline void pagefault_disabled_dec(void) ++{ ++ current->pagefault_disabled--; ++ WARN_ON(current->pagefault_disabled < 0); ++} ++ + /* +- * These routines enable/disable the pagefault handler in that +- * it will not take any locks and go straight to the fixup table. ++ * These routines enable/disable the pagefault handler. If disabled, it will ++ * not take any locks and go straight to the fixup table. + * +- * They have great resemblance to the preempt_disable/enable calls +- * and in fact they are identical; this is because currently there is +- * no other way to make the pagefault handlers do this. So we do +- * disable preemption but we don't necessarily care about that. ++ * User access methods will not sleep when called from a pagefault_disabled() ++ * environment. + */ + static inline void pagefault_disable(void) + { +- preempt_count_inc(); ++ migrate_disable(); ++ pagefault_disabled_inc(); + /* + * make sure to have issued the store before a pagefault + * can hit. +@@ -25,18 +35,32 @@ + + static inline void pagefault_enable(void) + { +-#ifndef CONFIG_PREEMPT + /* + * make sure to issue those last loads/stores before enabling + * the pagefault handler again. + */ + barrier(); +- preempt_count_dec(); +-#else +- preempt_enable(); +-#endif ++ pagefault_disabled_dec(); ++ migrate_enable(); + } + ++/* ++ * Is the pagefault handler disabled? If so, user access methods will not sleep. ++ */ ++#define pagefault_disabled() (current->pagefault_disabled != 0) ++ ++/* ++ * The pagefault handler is in general disabled by pagefault_disable() or ++ * when in irq context (via in_atomic()). ++ * ++ * This function should only be used by the fault handlers. Other users should ++ * stick to pagefault_disabled(). ++ * Please NEVER use preempt_disable() to disable the fault handler. With ++ * !CONFIG_PREEMPT_COUNT, this is like a NOP. So the handler won't be disabled. ++ * in_atomic() will report different values based on !CONFIG_PREEMPT_COUNT. ++ */ ++#define faulthandler_disabled() (pagefault_disabled() || in_atomic()) ++ + #ifndef ARCH_HAS_NOCACHE_UACCESS + + static inline unsigned long __copy_from_user_inatomic_nocache(void *to, +diff -Nur linux-4.1.13.orig/include/linux/uprobes.h linux-4.1.13/include/linux/uprobes.h +--- linux-4.1.13.orig/include/linux/uprobes.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/uprobes.h 2015-11-29 09:23:09.617613651 +0100 +@@ -27,6 +27,7 @@ + #include <linux/errno.h> + #include <linux/rbtree.h> + #include <linux/types.h> ++#include <linux/wait.h> + + struct vm_area_struct; + struct mm_struct; +diff -Nur linux-4.1.13.orig/include/linux/vmstat.h linux-4.1.13/include/linux/vmstat.h +--- linux-4.1.13.orig/include/linux/vmstat.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/vmstat.h 2015-11-29 09:23:09.617613651 +0100 +@@ -33,7 +33,9 @@ + */ + static inline void __count_vm_event(enum vm_event_item item) + { ++ preempt_disable_rt(); + raw_cpu_inc(vm_event_states.event[item]); ++ preempt_enable_rt(); + } + + static inline void count_vm_event(enum vm_event_item item) +@@ -43,7 +45,9 @@ + + static inline void __count_vm_events(enum vm_event_item item, long delta) + { ++ preempt_disable_rt(); + raw_cpu_add(vm_event_states.event[item], delta); ++ preempt_enable_rt(); + } + + static inline void count_vm_events(enum vm_event_item item, long delta) +diff -Nur linux-4.1.13.orig/include/linux/wait.h linux-4.1.13/include/linux/wait.h +--- linux-4.1.13.orig/include/linux/wait.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/linux/wait.h 2015-11-29 09:23:09.621613384 +0100 +@@ -8,6 +8,7 @@ + #include <linux/spinlock.h> + #include <asm/current.h> + #include <uapi/linux/wait.h> ++#include <linux/atomic.h> + + typedef struct __wait_queue wait_queue_t; + typedef int (*wait_queue_func_t)(wait_queue_t *wait, unsigned mode, int flags, void *key); +diff -Nur linux-4.1.13.orig/include/linux/wait-simple.h linux-4.1.13/include/linux/wait-simple.h +--- linux-4.1.13.orig/include/linux/wait-simple.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/wait-simple.h 2015-11-29 09:23:09.617613651 +0100 +@@ -0,0 +1,207 @@ ++#ifndef _LINUX_WAIT_SIMPLE_H ++#define _LINUX_WAIT_SIMPLE_H ++ ++#include <linux/spinlock.h> ++#include <linux/list.h> ++ ++#include <asm/current.h> ++ ++struct swaiter { ++ struct task_struct *task; ++ struct list_head node; ++}; ++ ++#define DEFINE_SWAITER(name) \ ++ struct swaiter name = { \ ++ .task = current, \ ++ .node = LIST_HEAD_INIT((name).node), \ ++ } ++ ++struct swait_head { ++ raw_spinlock_t lock; ++ struct list_head list; ++}; ++ ++#define SWAIT_HEAD_INITIALIZER(name) { \ ++ .lock = __RAW_SPIN_LOCK_UNLOCKED(name.lock), \ ++ .list = LIST_HEAD_INIT((name).list), \ ++ } ++ ++#define DEFINE_SWAIT_HEAD(name) \ ++ struct swait_head name = SWAIT_HEAD_INITIALIZER(name) ++ ++extern void __init_swait_head(struct swait_head *h, struct lock_class_key *key); ++ ++#define init_swait_head(swh) \ ++ do { \ ++ static struct lock_class_key __key; \ ++ \ ++ __init_swait_head((swh), &__key); \ ++ } while (0) ++ ++/* ++ * Waiter functions ++ */ ++extern void swait_prepare_locked(struct swait_head *head, struct swaiter *w); ++extern void swait_prepare(struct swait_head *head, struct swaiter *w, int state); ++extern void swait_finish_locked(struct swait_head *head, struct swaiter *w); ++extern void swait_finish(struct swait_head *head, struct swaiter *w); ++ ++/* Check whether a head has waiters enqueued */ ++static inline bool swaitqueue_active(struct swait_head *h) ++{ ++ /* Make sure the condition is visible before checking list_empty() */ ++ smp_mb(); ++ return !list_empty(&h->list); ++} ++ ++/* ++ * Wakeup functions ++ */ ++extern unsigned int __swait_wake(struct swait_head *head, unsigned int state, unsigned int num); ++extern unsigned int __swait_wake_locked(struct swait_head *head, unsigned int state, unsigned int num); ++ ++#define swait_wake(head) __swait_wake(head, TASK_NORMAL, 1) ++#define swait_wake_interruptible(head) __swait_wake(head, TASK_INTERRUPTIBLE, 1) ++#define swait_wake_all(head) __swait_wake(head, TASK_NORMAL, 0) ++#define swait_wake_all_interruptible(head) __swait_wake(head, TASK_INTERRUPTIBLE, 0) ++ ++/* ++ * Event API ++ */ ++#define __swait_event(wq, condition) \ ++do { \ ++ DEFINE_SWAITER(__wait); \ ++ \ ++ for (;;) { \ ++ swait_prepare(&wq, &__wait, TASK_UNINTERRUPTIBLE); \ ++ if (condition) \ ++ break; \ ++ schedule(); \ ++ } \ ++ swait_finish(&wq, &__wait); \ ++} while (0) ++ ++/** ++ * swait_event - sleep until a condition gets true ++ * @wq: the waitqueue to wait on ++ * @condition: a C expression for the event to wait for ++ * ++ * The process is put to sleep (TASK_UNINTERRUPTIBLE) until the ++ * @condition evaluates to true. The @condition is checked each time ++ * the waitqueue @wq is woken up. ++ * ++ * wake_up() has to be called after changing any variable that could ++ * change the result of the wait condition. ++ */ ++#define swait_event(wq, condition) \ ++do { \ ++ if (condition) \ ++ break; \ ++ __swait_event(wq, condition); \ ++} while (0) ++ ++#define __swait_event_interruptible(wq, condition, ret) \ ++do { \ ++ DEFINE_SWAITER(__wait); \ ++ \ ++ for (;;) { \ ++ swait_prepare(&wq, &__wait, TASK_INTERRUPTIBLE); \ ++ if (condition) \ ++ break; \ ++ if (signal_pending(current)) { \ ++ ret = -ERESTARTSYS; \ ++ break; \ ++ } \ ++ schedule(); \ ++ } \ ++ swait_finish(&wq, &__wait); \ ++} while (0) ++ ++#define __swait_event_interruptible_timeout(wq, condition, ret) \ ++do { \ ++ DEFINE_SWAITER(__wait); \ ++ \ ++ for (;;) { \ ++ swait_prepare(&wq, &__wait, TASK_INTERRUPTIBLE); \ ++ if (condition) \ ++ break; \ ++ if (signal_pending(current)) { \ ++ ret = -ERESTARTSYS; \ ++ break; \ ++ } \ ++ ret = schedule_timeout(ret); \ ++ if (!ret) \ ++ break; \ ++ } \ ++ swait_finish(&wq, &__wait); \ ++} while (0) ++ ++/** ++ * swait_event_interruptible - sleep until a condition gets true ++ * @wq: the waitqueue to wait on ++ * @condition: a C expression for the event to wait for ++ * ++ * The process is put to sleep (TASK_INTERRUPTIBLE) until the ++ * @condition evaluates to true. The @condition is checked each time ++ * the waitqueue @wq is woken up. ++ * ++ * wake_up() has to be called after changing any variable that could ++ * change the result of the wait condition. ++ */ ++#define swait_event_interruptible(wq, condition) \ ++({ \ ++ int __ret = 0; \ ++ if (!(condition)) \ ++ __swait_event_interruptible(wq, condition, __ret); \ ++ __ret; \ ++}) ++ ++#define swait_event_interruptible_timeout(wq, condition, timeout) \ ++({ \ ++ int __ret = timeout; \ ++ if (!(condition)) \ ++ __swait_event_interruptible_timeout(wq, condition, __ret); \ ++ __ret; \ ++}) ++ ++#define __swait_event_timeout(wq, condition, ret) \ ++do { \ ++ DEFINE_SWAITER(__wait); \ ++ \ ++ for (;;) { \ ++ swait_prepare(&wq, &__wait, TASK_UNINTERRUPTIBLE); \ ++ if (condition) \ ++ break; \ ++ ret = schedule_timeout(ret); \ ++ if (!ret) \ ++ break; \ ++ } \ ++ swait_finish(&wq, &__wait); \ ++} while (0) ++ ++/** ++ * swait_event_timeout - sleep until a condition gets true or a timeout elapses ++ * @wq: the waitqueue to wait on ++ * @condition: a C expression for the event to wait for ++ * @timeout: timeout, in jiffies ++ * ++ * The process is put to sleep (TASK_UNINTERRUPTIBLE) until the ++ * @condition evaluates to true. The @condition is checked each time ++ * the waitqueue @wq is woken up. ++ * ++ * wake_up() has to be called after changing any variable that could ++ * change the result of the wait condition. ++ * ++ * The function returns 0 if the @timeout elapsed, and the remaining ++ * jiffies if the condition evaluated to true before the timeout elapsed. ++ */ ++#define swait_event_timeout(wq, condition, timeout) \ ++({ \ ++ long __ret = timeout; \ ++ if (!(condition)) \ ++ __swait_event_timeout(wq, condition, __ret); \ ++ __ret; \ ++}) ++ ++#endif +diff -Nur linux-4.1.13.orig/include/linux/work-simple.h linux-4.1.13/include/linux/work-simple.h +--- linux-4.1.13.orig/include/linux/work-simple.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/linux/work-simple.h 2015-11-29 09:23:09.621613384 +0100 +@@ -0,0 +1,24 @@ ++#ifndef _LINUX_SWORK_H ++#define _LINUX_SWORK_H ++ ++#include <linux/list.h> ++ ++struct swork_event { ++ struct list_head item; ++ unsigned long flags; ++ void (*func)(struct swork_event *); ++}; ++ ++static inline void INIT_SWORK(struct swork_event *event, ++ void (*func)(struct swork_event *)) ++{ ++ event->flags = 0; ++ event->func = func; ++} ++ ++bool swork_queue(struct swork_event *sev); ++ ++int swork_get(void); ++void swork_put(void); ++ ++#endif /* _LINUX_SWORK_H */ +diff -Nur linux-4.1.13.orig/include/net/dst.h linux-4.1.13/include/net/dst.h +--- linux-4.1.13.orig/include/net/dst.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/net/dst.h 2015-11-29 09:23:09.629612854 +0100 +@@ -403,7 +403,7 @@ + static inline int dst_neigh_output(struct dst_entry *dst, struct neighbour *n, + struct sk_buff *skb) + { +- const struct hh_cache *hh; ++ struct hh_cache *hh; + + if (dst->pending_confirm) { + unsigned long now = jiffies; +diff -Nur linux-4.1.13.orig/include/net/neighbour.h linux-4.1.13/include/net/neighbour.h +--- linux-4.1.13.orig/include/net/neighbour.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/net/neighbour.h 2015-11-29 09:23:09.629612854 +0100 +@@ -445,7 +445,7 @@ + } + #endif + +-static inline int neigh_hh_output(const struct hh_cache *hh, struct sk_buff *skb) ++static inline int neigh_hh_output(struct hh_cache *hh, struct sk_buff *skb) + { + unsigned int seq; + int hh_len; +@@ -500,7 +500,7 @@ + + #define NEIGH_CB(skb) ((struct neighbour_cb *)(skb)->cb) + +-static inline void neigh_ha_snapshot(char *dst, const struct neighbour *n, ++static inline void neigh_ha_snapshot(char *dst, struct neighbour *n, + const struct net_device *dev) + { + unsigned int seq; +diff -Nur linux-4.1.13.orig/include/net/netns/ipv4.h linux-4.1.13/include/net/netns/ipv4.h +--- linux-4.1.13.orig/include/net/netns/ipv4.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/net/netns/ipv4.h 2015-11-29 09:23:09.629612854 +0100 +@@ -69,6 +69,7 @@ + + int sysctl_icmp_echo_ignore_all; + int sysctl_icmp_echo_ignore_broadcasts; ++ int sysctl_icmp_echo_sysrq; + int sysctl_icmp_ignore_bogus_error_responses; + int sysctl_icmp_ratelimit; + int sysctl_icmp_ratemask; +diff -Nur linux-4.1.13.orig/include/trace/events/hist.h linux-4.1.13/include/trace/events/hist.h +--- linux-4.1.13.orig/include/trace/events/hist.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/trace/events/hist.h 2015-11-29 09:23:09.629612854 +0100 +@@ -0,0 +1,74 @@ ++#undef TRACE_SYSTEM ++#define TRACE_SYSTEM hist ++ ++#if !defined(_TRACE_HIST_H) || defined(TRACE_HEADER_MULTI_READ) ++#define _TRACE_HIST_H ++ ++#include "latency_hist.h" ++#include <linux/tracepoint.h> ++ ++#if !defined(CONFIG_PREEMPT_OFF_HIST) && !defined(CONFIG_INTERRUPT_OFF_HIST) ++#define trace_preemptirqsoff_hist(a, b) ++#define trace_preemptirqsoff_hist_rcuidle(a, b) ++#else ++TRACE_EVENT(preemptirqsoff_hist, ++ ++ TP_PROTO(int reason, int starthist), ++ ++ TP_ARGS(reason, starthist), ++ ++ TP_STRUCT__entry( ++ __field(int, reason) ++ __field(int, starthist) ++ ), ++ ++ TP_fast_assign( ++ __entry->reason = reason; ++ __entry->starthist = starthist; ++ ), ++ ++ TP_printk("reason=%s starthist=%s", getaction(__entry->reason), ++ __entry->starthist ? "start" : "stop") ++); ++#endif ++ ++#ifndef CONFIG_MISSED_TIMER_OFFSETS_HIST ++#define trace_hrtimer_interrupt(a, b, c, d) ++#define trace_hrtimer_interrupt_rcuidle(a, b, c, d) ++#else ++TRACE_EVENT(hrtimer_interrupt, ++ ++ TP_PROTO(int cpu, long long offset, struct task_struct *curr, ++ struct task_struct *task), ++ ++ TP_ARGS(cpu, offset, curr, task), ++ ++ TP_STRUCT__entry( ++ __field(int, cpu) ++ __field(long long, offset) ++ __array(char, ccomm, TASK_COMM_LEN) ++ __field(int, cprio) ++ __array(char, tcomm, TASK_COMM_LEN) ++ __field(int, tprio) ++ ), ++ ++ TP_fast_assign( ++ __entry->cpu = cpu; ++ __entry->offset = offset; ++ memcpy(__entry->ccomm, curr->comm, TASK_COMM_LEN); ++ __entry->cprio = curr->prio; ++ memcpy(__entry->tcomm, task != NULL ? task->comm : "<none>", ++ task != NULL ? TASK_COMM_LEN : 7); ++ __entry->tprio = task != NULL ? task->prio : -1; ++ ), ++ ++ TP_printk("cpu=%d offset=%lld curr=%s[%d] thread=%s[%d]", ++ __entry->cpu, __entry->offset, __entry->ccomm, ++ __entry->cprio, __entry->tcomm, __entry->tprio) ++); ++#endif ++ ++#endif /* _TRACE_HIST_H */ ++ ++/* This part must be outside protection */ ++#include <trace/define_trace.h> +diff -Nur linux-4.1.13.orig/include/trace/events/latency_hist.h linux-4.1.13/include/trace/events/latency_hist.h +--- linux-4.1.13.orig/include/trace/events/latency_hist.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/include/trace/events/latency_hist.h 2015-11-29 09:23:09.629612854 +0100 +@@ -0,0 +1,29 @@ ++#ifndef _LATENCY_HIST_H ++#define _LATENCY_HIST_H ++ ++enum hist_action { ++ IRQS_ON, ++ PREEMPT_ON, ++ TRACE_STOP, ++ IRQS_OFF, ++ PREEMPT_OFF, ++ TRACE_START, ++}; ++ ++static char *actions[] = { ++ "IRQS_ON", ++ "PREEMPT_ON", ++ "TRACE_STOP", ++ "IRQS_OFF", ++ "PREEMPT_OFF", ++ "TRACE_START", ++}; ++ ++static inline char *getaction(int action) ++{ ++ if (action >= 0 && action <= sizeof(actions)/sizeof(actions[0])) ++ return actions[action]; ++ return "unknown"; ++} ++ ++#endif /* _LATENCY_HIST_H */ +diff -Nur linux-4.1.13.orig/include/trace/events/sched.h linux-4.1.13/include/trace/events/sched.h +--- linux-4.1.13.orig/include/trace/events/sched.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/include/trace/events/sched.h 2015-11-29 09:23:09.629612854 +0100 +@@ -55,9 +55,9 @@ + */ + DECLARE_EVENT_CLASS(sched_wakeup_template, + +- TP_PROTO(struct task_struct *p, int success), ++ TP_PROTO(struct task_struct *p), + +- TP_ARGS(__perf_task(p), success), ++ TP_ARGS(__perf_task(p)), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) +@@ -71,25 +71,37 @@ + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; + __entry->prio = p->prio; +- __entry->success = success; ++ __entry->success = 1; /* rudiment, kill when possible */ + __entry->target_cpu = task_cpu(p); + ), + +- TP_printk("comm=%s pid=%d prio=%d success=%d target_cpu=%03d", ++ TP_printk("comm=%s pid=%d prio=%d target_cpu=%03d", + __entry->comm, __entry->pid, __entry->prio, +- __entry->success, __entry->target_cpu) ++ __entry->target_cpu) + ); + ++/* ++ * Tracepoint called when waking a task; this tracepoint is guaranteed to be ++ * called from the waking context. ++ */ ++DEFINE_EVENT(sched_wakeup_template, sched_waking, ++ TP_PROTO(struct task_struct *p), ++ TP_ARGS(p)); ++ ++/* ++ * Tracepoint called when the task is actually woken; p->state == TASK_RUNNNG. ++ * It it not always called from the waking context. ++ */ + DEFINE_EVENT(sched_wakeup_template, sched_wakeup, +- TP_PROTO(struct task_struct *p, int success), +- TP_ARGS(p, success)); ++ TP_PROTO(struct task_struct *p), ++ TP_ARGS(p)); + + /* + * Tracepoint for waking up a new task: + */ + DEFINE_EVENT(sched_wakeup_template, sched_wakeup_new, +- TP_PROTO(struct task_struct *p, int success), +- TP_ARGS(p, success)); ++ TP_PROTO(struct task_struct *p), ++ TP_ARGS(p)); + + #ifdef CREATE_TRACE_POINTS + static inline long __trace_sched_switch_state(struct task_struct *p) +diff -Nur linux-4.1.13.orig/init/Kconfig linux-4.1.13/init/Kconfig +--- linux-4.1.13.orig/init/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/init/Kconfig 2015-11-29 09:23:09.629612854 +0100 +@@ -637,7 +637,7 @@ + + config RCU_FAST_NO_HZ + bool "Accelerate last non-dyntick-idle CPU's grace periods" +- depends on NO_HZ_COMMON && SMP ++ depends on NO_HZ_COMMON && SMP && !PREEMPT_RT_FULL + default n + help + This option permits CPUs to enter dynticks-idle state even if +@@ -664,7 +664,7 @@ + config RCU_BOOST + bool "Enable RCU priority boosting" + depends on RT_MUTEXES && PREEMPT_RCU +- default n ++ default y if PREEMPT_RT_FULL + help + This option boosts the priority of preempted RCU readers that + block the current preemptible RCU grace period for too long. +@@ -1101,6 +1101,7 @@ + config RT_GROUP_SCHED + bool "Group scheduling for SCHED_RR/FIFO" + depends on CGROUP_SCHED ++ depends on !PREEMPT_RT_FULL + default n + help + This feature lets you explicitly allocate real CPU bandwidth +@@ -1688,6 +1689,7 @@ + + config SLAB + bool "SLAB" ++ depends on !PREEMPT_RT_FULL + help + The regular slab allocator that is established and known to work + well in all environments. It organizes cache hot objects in +@@ -1706,6 +1708,7 @@ + config SLOB + depends on EXPERT + bool "SLOB (Simple Allocator)" ++ depends on !PREEMPT_RT_FULL + help + SLOB replaces the stock allocator with a drastically simpler + allocator. SLOB is generally more space efficient but +@@ -1715,7 +1718,7 @@ + + config SLUB_CPU_PARTIAL + default y +- depends on SLUB && SMP ++ depends on SLUB && SMP && !PREEMPT_RT_FULL + bool "SLUB per cpu partial cache" + help + Per cpu partial caches accellerate objects allocation and freeing +diff -Nur linux-4.1.13.orig/init/main.c linux-4.1.13/init/main.c +--- linux-4.1.13.orig/init/main.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/init/main.c 2015-11-29 09:23:09.629612854 +0100 +@@ -525,6 +525,7 @@ + setup_command_line(command_line); + setup_nr_cpu_ids(); + setup_per_cpu_areas(); ++ softirq_early_init(); + smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */ + + build_all_zonelists(NULL, NULL); +diff -Nur linux-4.1.13.orig/init/Makefile linux-4.1.13/init/Makefile +--- linux-4.1.13.orig/init/Makefile 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/init/Makefile 2015-11-29 09:23:09.629612854 +0100 +@@ -33,4 +33,4 @@ + include/generated/compile.h: FORCE + @$($(quiet)chk_compile.h) + $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \ +- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)" ++ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)" +diff -Nur linux-4.1.13.orig/ipc/mqueue.c linux-4.1.13/ipc/mqueue.c +--- linux-4.1.13.orig/ipc/mqueue.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/ipc/mqueue.c 2015-11-29 09:23:09.629612854 +0100 +@@ -47,8 +47,7 @@ + #define RECV 1 + + #define STATE_NONE 0 +-#define STATE_PENDING 1 +-#define STATE_READY 2 ++#define STATE_READY 1 + + struct posix_msg_tree_node { + struct rb_node rb_node; +@@ -568,15 +567,12 @@ + wq_add(info, sr, ewp); + + for (;;) { +- set_current_state(TASK_INTERRUPTIBLE); ++ __set_current_state(TASK_INTERRUPTIBLE); + + spin_unlock(&info->lock); + time = schedule_hrtimeout_range_clock(timeout, 0, + HRTIMER_MODE_ABS, CLOCK_REALTIME); + +- while (ewp->state == STATE_PENDING) +- cpu_relax(); +- + if (ewp->state == STATE_READY) { + retval = 0; + goto out; +@@ -904,11 +900,15 @@ + * list of waiting receivers. A sender checks that list before adding the new + * message into the message array. If there is a waiting receiver, then it + * bypasses the message array and directly hands the message over to the +- * receiver. +- * The receiver accepts the message and returns without grabbing the queue +- * spinlock. Therefore an intermediate STATE_PENDING state and memory barriers +- * are necessary. The same algorithm is used for sysv semaphores, see +- * ipc/sem.c for more details. ++ * receiver. The receiver accepts the message and returns without grabbing the ++ * queue spinlock: ++ * ++ * - Set pointer to message. ++ * - Queue the receiver task for later wakeup (without the info->lock). ++ * - Update its state to STATE_READY. Now the receiver can continue. ++ * - Wake up the process after the lock is dropped. Should the process wake up ++ * before this wakeup (due to a timeout or a signal) it will either see ++ * STATE_READY and continue or acquire the lock to check the state again. + * + * The same algorithm is used for senders. + */ +@@ -916,21 +916,29 @@ + /* pipelined_send() - send a message directly to the task waiting in + * sys_mq_timedreceive() (without inserting message into a queue). + */ +-static inline void pipelined_send(struct mqueue_inode_info *info, ++static inline void pipelined_send(struct wake_q_head *wake_q, ++ struct mqueue_inode_info *info, + struct msg_msg *message, + struct ext_wait_queue *receiver) + { + receiver->msg = message; + list_del(&receiver->list); +- receiver->state = STATE_PENDING; +- wake_up_process(receiver->task); +- smp_wmb(); ++ wake_q_add(wake_q, receiver->task); ++ /* ++ * Rely on the implicit cmpxchg barrier from wake_q_add such ++ * that we can ensure that updating receiver->state is the last ++ * write operation: As once set, the receiver can continue, ++ * and if we don't have the reference count from the wake_q, ++ * yet, at that point we can later have a use-after-free ++ * condition and bogus wakeup. ++ */ + receiver->state = STATE_READY; + } + + /* pipelined_receive() - if there is task waiting in sys_mq_timedsend() + * gets its message and put to the queue (we have one free place for sure). */ +-static inline void pipelined_receive(struct mqueue_inode_info *info) ++static inline void pipelined_receive(struct wake_q_head *wake_q, ++ struct mqueue_inode_info *info) + { + struct ext_wait_queue *sender = wq_get_first_waiter(info, SEND); + +@@ -941,10 +949,9 @@ + } + if (msg_insert(sender->msg, info)) + return; ++ + list_del(&sender->list); +- sender->state = STATE_PENDING; +- wake_up_process(sender->task); +- smp_wmb(); ++ wake_q_add(wake_q, sender->task); + sender->state = STATE_READY; + } + +@@ -962,6 +969,7 @@ + struct timespec ts; + struct posix_msg_tree_node *new_leaf = NULL; + int ret = 0; ++ WAKE_Q(wake_q); + + if (u_abs_timeout) { + int res = prepare_timeout(u_abs_timeout, &expires, &ts); +@@ -1045,7 +1053,7 @@ + } else { + receiver = wq_get_first_waiter(info, RECV); + if (receiver) { +- pipelined_send(info, msg_ptr, receiver); ++ pipelined_send(&wake_q, info, msg_ptr, receiver); + } else { + /* adds message to the queue */ + ret = msg_insert(msg_ptr, info); +@@ -1058,6 +1066,7 @@ + } + out_unlock: + spin_unlock(&info->lock); ++ wake_up_q(&wake_q); + out_free: + if (ret) + free_msg(msg_ptr); +@@ -1144,14 +1153,17 @@ + msg_ptr = wait.msg; + } + } else { ++ WAKE_Q(wake_q); ++ + msg_ptr = msg_get(info); + + inode->i_atime = inode->i_mtime = inode->i_ctime = + CURRENT_TIME; + + /* There is now free space in queue. */ +- pipelined_receive(info); ++ pipelined_receive(&wake_q, info); + spin_unlock(&info->lock); ++ wake_up_q(&wake_q); + ret = 0; + } + if (ret == 0) { +diff -Nur linux-4.1.13.orig/ipc/msg.c linux-4.1.13/ipc/msg.c +--- linux-4.1.13.orig/ipc/msg.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/ipc/msg.c 2015-11-29 09:23:09.629612854 +0100 +@@ -188,6 +188,12 @@ + struct msg_receiver *msr, *t; + + list_for_each_entry_safe(msr, t, &msq->q_receivers, r_list) { ++ /* ++ * Make sure that the wakeup doesnt preempt ++ * this CPU prematurely. (on PREEMPT_RT) ++ */ ++ preempt_disable_rt(); ++ + msr->r_msg = NULL; /* initialize expunge ordering */ + wake_up_process(msr->r_tsk); + /* +@@ -198,6 +204,8 @@ + */ + smp_mb(); + msr->r_msg = ERR_PTR(res); ++ ++ preempt_enable_rt(); + } + } + +@@ -574,6 +582,11 @@ + if (testmsg(msg, msr->r_msgtype, msr->r_mode) && + !security_msg_queue_msgrcv(msq, msg, msr->r_tsk, + msr->r_msgtype, msr->r_mode)) { ++ /* ++ * Make sure that the wakeup doesnt preempt ++ * this CPU prematurely. (on PREEMPT_RT) ++ */ ++ preempt_disable_rt(); + + list_del(&msr->r_list); + if (msr->r_maxsize < msg->m_ts) { +@@ -595,12 +608,13 @@ + */ + smp_mb(); + msr->r_msg = msg; ++ preempt_enable_rt(); + + return 1; + } ++ preempt_enable_rt(); + } + } +- + return 0; + } + +diff -Nur linux-4.1.13.orig/ipc/sem.c linux-4.1.13/ipc/sem.c +--- linux-4.1.13.orig/ipc/sem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/ipc/sem.c 2015-11-29 09:23:09.629612854 +0100 +@@ -690,6 +690,13 @@ + static void wake_up_sem_queue_prepare(struct list_head *pt, + struct sem_queue *q, int error) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ struct task_struct *p = q->sleeper; ++ get_task_struct(p); ++ q->status = error; ++ wake_up_process(p); ++ put_task_struct(p); ++#else + if (list_empty(pt)) { + /* + * Hold preempt off so that we don't get preempted and have the +@@ -701,6 +708,7 @@ + q->pid = error; + + list_add_tail(&q->list, pt); ++#endif + } + + /** +@@ -714,6 +722,7 @@ + */ + static void wake_up_sem_queue_do(struct list_head *pt) + { ++#ifndef CONFIG_PREEMPT_RT_BASE + struct sem_queue *q, *t; + int did_something; + +@@ -726,6 +735,7 @@ + } + if (did_something) + preempt_enable(); ++#endif + } + + static void unlink_queue(struct sem_array *sma, struct sem_queue *q) +diff -Nur linux-4.1.13.orig/kernel/bpf/hashtab.c linux-4.1.13/kernel/bpf/hashtab.c +--- linux-4.1.13.orig/kernel/bpf/hashtab.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/bpf/hashtab.c 2015-11-29 09:23:09.629612854 +0100 +@@ -17,7 +17,7 @@ + struct bpf_htab { + struct bpf_map map; + struct hlist_head *buckets; +- spinlock_t lock; ++ raw_spinlock_t lock; + u32 count; /* number of elements in this hashtable */ + u32 n_buckets; /* number of hash buckets */ + u32 elem_size; /* size of each element in bytes */ +@@ -82,7 +82,7 @@ + for (i = 0; i < htab->n_buckets; i++) + INIT_HLIST_HEAD(&htab->buckets[i]); + +- spin_lock_init(&htab->lock); ++ raw_spin_lock_init(&htab->lock); + htab->count = 0; + + htab->elem_size = sizeof(struct htab_elem) + +@@ -230,7 +230,7 @@ + l_new->hash = htab_map_hash(l_new->key, key_size); + + /* bpf_map_update_elem() can be called in_irq() */ +- spin_lock_irqsave(&htab->lock, flags); ++ raw_spin_lock_irqsave(&htab->lock, flags); + + head = select_bucket(htab, l_new->hash); + +@@ -266,11 +266,11 @@ + } else { + htab->count++; + } +- spin_unlock_irqrestore(&htab->lock, flags); ++ raw_spin_unlock_irqrestore(&htab->lock, flags); + + return 0; + err: +- spin_unlock_irqrestore(&htab->lock, flags); ++ raw_spin_unlock_irqrestore(&htab->lock, flags); + kfree(l_new); + return ret; + } +@@ -291,7 +291,7 @@ + + hash = htab_map_hash(key, key_size); + +- spin_lock_irqsave(&htab->lock, flags); ++ raw_spin_lock_irqsave(&htab->lock, flags); + + head = select_bucket(htab, hash); + +@@ -304,7 +304,7 @@ + ret = 0; + } + +- spin_unlock_irqrestore(&htab->lock, flags); ++ raw_spin_unlock_irqrestore(&htab->lock, flags); + return ret; + } + +diff -Nur linux-4.1.13.orig/kernel/cgroup.c linux-4.1.13/kernel/cgroup.c +--- linux-4.1.13.orig/kernel/cgroup.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/cgroup.c 2015-11-29 09:23:09.629612854 +0100 +@@ -4422,10 +4422,10 @@ + queue_work(cgroup_destroy_wq, &css->destroy_work); + } + +-static void css_release_work_fn(struct work_struct *work) ++static void css_release_work_fn(struct swork_event *sev) + { + struct cgroup_subsys_state *css = +- container_of(work, struct cgroup_subsys_state, destroy_work); ++ container_of(sev, struct cgroup_subsys_state, destroy_swork); + struct cgroup_subsys *ss = css->ss; + struct cgroup *cgrp = css->cgroup; + +@@ -4464,8 +4464,8 @@ + struct cgroup_subsys_state *css = + container_of(ref, struct cgroup_subsys_state, refcnt); + +- INIT_WORK(&css->destroy_work, css_release_work_fn); +- queue_work(cgroup_destroy_wq, &css->destroy_work); ++ INIT_SWORK(&css->destroy_swork, css_release_work_fn); ++ swork_queue(&css->destroy_swork); + } + + static void init_and_link_css(struct cgroup_subsys_state *css, +@@ -5069,6 +5069,7 @@ + */ + cgroup_destroy_wq = alloc_workqueue("cgroup_destroy", 0, 1); + BUG_ON(!cgroup_destroy_wq); ++ BUG_ON(swork_get()); + + /* + * Used to destroy pidlists and separate to serve as flush domain. +diff -Nur linux-4.1.13.orig/kernel/cpu.c linux-4.1.13/kernel/cpu.c +--- linux-4.1.13.orig/kernel/cpu.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/cpu.c 2015-11-29 09:23:09.633612589 +0100 +@@ -74,8 +74,8 @@ + #endif + } cpu_hotplug = { + .active_writer = NULL, +- .wq = __WAIT_QUEUE_HEAD_INITIALIZER(cpu_hotplug.wq), + .lock = __MUTEX_INITIALIZER(cpu_hotplug.lock), ++ .wq = __WAIT_QUEUE_HEAD_INITIALIZER(cpu_hotplug.wq), + #ifdef CONFIG_DEBUG_LOCK_ALLOC + .dep_map = {.name = "cpu_hotplug.lock" }, + #endif +@@ -88,6 +88,289 @@ + #define cpuhp_lock_acquire() lock_map_acquire(&cpu_hotplug.dep_map) + #define cpuhp_lock_release() lock_map_release(&cpu_hotplug.dep_map) + ++/** ++ * hotplug_pcp - per cpu hotplug descriptor ++ * @unplug: set when pin_current_cpu() needs to sync tasks ++ * @sync_tsk: the task that waits for tasks to finish pinned sections ++ * @refcount: counter of tasks in pinned sections ++ * @grab_lock: set when the tasks entering pinned sections should wait ++ * @synced: notifier for @sync_tsk to tell cpu_down it's finished ++ * @mutex: the mutex to make tasks wait (used when @grab_lock is true) ++ * @mutex_init: zero if the mutex hasn't been initialized yet. ++ * ++ * Although @unplug and @sync_tsk may point to the same task, the @unplug ++ * is used as a flag and still exists after @sync_tsk has exited and ++ * @sync_tsk set to NULL. ++ */ ++struct hotplug_pcp { ++ struct task_struct *unplug; ++ struct task_struct *sync_tsk; ++ int refcount; ++ int grab_lock; ++ struct completion synced; ++ struct completion unplug_wait; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * Note, on PREEMPT_RT, the hotplug lock must save the state of ++ * the task, otherwise the mutex will cause the task to fail ++ * to sleep when required. (Because it's called from migrate_disable()) ++ * ++ * The spinlock_t on PREEMPT_RT is a mutex that saves the task's ++ * state. ++ */ ++ spinlock_t lock; ++#else ++ struct mutex mutex; ++#endif ++ int mutex_init; ++}; ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++# define hotplug_lock(hp) rt_spin_lock(&(hp)->lock) ++# define hotplug_unlock(hp) rt_spin_unlock(&(hp)->lock) ++#else ++# define hotplug_lock(hp) mutex_lock(&(hp)->mutex) ++# define hotplug_unlock(hp) mutex_unlock(&(hp)->mutex) ++#endif ++ ++static DEFINE_PER_CPU(struct hotplug_pcp, hotplug_pcp); ++ ++/** ++ * pin_current_cpu - Prevent the current cpu from being unplugged ++ * ++ * Lightweight version of get_online_cpus() to prevent cpu from being ++ * unplugged when code runs in a migration disabled region. ++ * ++ * Must be called with preemption disabled (preempt_count = 1)! ++ */ ++void pin_current_cpu(void) ++{ ++ struct hotplug_pcp *hp; ++ int force = 0; ++ ++retry: ++ hp = this_cpu_ptr(&hotplug_pcp); ++ ++ if (!hp->unplug || hp->refcount || force || preempt_count() > 1 || ++ hp->unplug == current) { ++ hp->refcount++; ++ return; ++ } ++ if (hp->grab_lock) { ++ preempt_enable(); ++ hotplug_lock(hp); ++ hotplug_unlock(hp); ++ } else { ++ preempt_enable(); ++ /* ++ * Try to push this task off of this CPU. ++ */ ++ if (!migrate_me()) { ++ preempt_disable(); ++ hp = this_cpu_ptr(&hotplug_pcp); ++ if (!hp->grab_lock) { ++ /* ++ * Just let it continue it's already pinned ++ * or about to sleep. ++ */ ++ force = 1; ++ goto retry; ++ } ++ preempt_enable(); ++ } ++ } ++ preempt_disable(); ++ goto retry; ++} ++ ++/** ++ * unpin_current_cpu - Allow unplug of current cpu ++ * ++ * Must be called with preemption or interrupts disabled! ++ */ ++void unpin_current_cpu(void) ++{ ++ struct hotplug_pcp *hp = this_cpu_ptr(&hotplug_pcp); ++ ++ WARN_ON(hp->refcount <= 0); ++ ++ /* This is safe. sync_unplug_thread is pinned to this cpu */ ++ if (!--hp->refcount && hp->unplug && hp->unplug != current) ++ wake_up_process(hp->unplug); ++} ++ ++static void wait_for_pinned_cpus(struct hotplug_pcp *hp) ++{ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ while (hp->refcount) { ++ schedule_preempt_disabled(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++} ++ ++static int sync_unplug_thread(void *data) ++{ ++ struct hotplug_pcp *hp = data; ++ ++ wait_for_completion(&hp->unplug_wait); ++ preempt_disable(); ++ hp->unplug = current; ++ wait_for_pinned_cpus(hp); ++ ++ /* ++ * This thread will synchronize the cpu_down() with threads ++ * that have pinned the CPU. When the pinned CPU count reaches ++ * zero, we inform the cpu_down code to continue to the next step. ++ */ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ preempt_enable(); ++ complete(&hp->synced); ++ ++ /* ++ * If all succeeds, the next step will need tasks to wait till ++ * the CPU is offline before continuing. To do this, the grab_lock ++ * is set and tasks going into pin_current_cpu() will block on the ++ * mutex. But we still need to wait for those that are already in ++ * pinned CPU sections. If the cpu_down() failed, the kthread_should_stop() ++ * will kick this thread out. ++ */ ++ while (!hp->grab_lock && !kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++ ++ /* Make sure grab_lock is seen before we see a stale completion */ ++ smp_mb(); ++ ++ /* ++ * Now just before cpu_down() enters stop machine, we need to make ++ * sure all tasks that are in pinned CPU sections are out, and new ++ * tasks will now grab the lock, keeping them from entering pinned ++ * CPU sections. ++ */ ++ if (!kthread_should_stop()) { ++ preempt_disable(); ++ wait_for_pinned_cpus(hp); ++ preempt_enable(); ++ complete(&hp->synced); ++ } ++ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ while (!kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++ set_current_state(TASK_RUNNING); ++ ++ /* ++ * Force this thread off this CPU as it's going down and ++ * we don't want any more work on this CPU. ++ */ ++ current->flags &= ~PF_NO_SETAFFINITY; ++ set_cpus_allowed_ptr(current, cpu_present_mask); ++ migrate_me(); ++ return 0; ++} ++ ++static void __cpu_unplug_sync(struct hotplug_pcp *hp) ++{ ++ wake_up_process(hp->sync_tsk); ++ wait_for_completion(&hp->synced); ++} ++ ++static void __cpu_unplug_wait(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ complete(&hp->unplug_wait); ++ wait_for_completion(&hp->synced); ++} ++ ++/* ++ * Start the sync_unplug_thread on the target cpu and wait for it to ++ * complete. ++ */ ++static int cpu_unplug_begin(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ int err; ++ ++ /* Protected by cpu_hotplug.lock */ ++ if (!hp->mutex_init) { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ spin_lock_init(&hp->lock); ++#else ++ mutex_init(&hp->mutex); ++#endif ++ hp->mutex_init = 1; ++ } ++ ++ /* Inform the scheduler to migrate tasks off this CPU */ ++ tell_sched_cpu_down_begin(cpu); ++ ++ init_completion(&hp->synced); ++ init_completion(&hp->unplug_wait); ++ ++ hp->sync_tsk = kthread_create(sync_unplug_thread, hp, "sync_unplug/%d", cpu); ++ if (IS_ERR(hp->sync_tsk)) { ++ err = PTR_ERR(hp->sync_tsk); ++ hp->sync_tsk = NULL; ++ return err; ++ } ++ kthread_bind(hp->sync_tsk, cpu); ++ ++ /* ++ * Wait for tasks to get out of the pinned sections, ++ * it's still OK if new tasks enter. Some CPU notifiers will ++ * wait for tasks that are going to enter these sections and ++ * we must not have them block. ++ */ ++ wake_up_process(hp->sync_tsk); ++ return 0; ++} ++ ++static void cpu_unplug_sync(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ init_completion(&hp->synced); ++ /* The completion needs to be initialzied before setting grab_lock */ ++ smp_wmb(); ++ ++ /* Grab the mutex before setting grab_lock */ ++ hotplug_lock(hp); ++ hp->grab_lock = 1; ++ ++ /* ++ * The CPU notifiers have been completed. ++ * Wait for tasks to get out of pinned CPU sections and have new ++ * tasks block until the CPU is completely down. ++ */ ++ __cpu_unplug_sync(hp); ++ ++ /* All done with the sync thread */ ++ kthread_stop(hp->sync_tsk); ++ hp->sync_tsk = NULL; ++} ++ ++static void cpu_unplug_done(unsigned int cpu) ++{ ++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu); ++ ++ hp->unplug = NULL; ++ /* Let all tasks know cpu unplug is finished before cleaning up */ ++ smp_wmb(); ++ ++ if (hp->sync_tsk) ++ kthread_stop(hp->sync_tsk); ++ ++ if (hp->grab_lock) { ++ hotplug_unlock(hp); ++ /* protected by cpu_hotplug.lock */ ++ hp->grab_lock = 0; ++ } ++ tell_sched_cpu_down_done(cpu); ++} + + void get_online_cpus(void) + { +@@ -349,13 +632,15 @@ + /* Requires cpu_add_remove_lock to be held */ + static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) + { +- int err, nr_calls = 0; ++ int mycpu, err, nr_calls = 0; + void *hcpu = (void *)(long)cpu; + unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0; + struct take_cpu_down_param tcd_param = { + .mod = mod, + .hcpu = hcpu, + }; ++ cpumask_var_t cpumask; ++ cpumask_var_t cpumask_org; + + if (num_online_cpus() == 1) + return -EBUSY; +@@ -363,7 +648,34 @@ + if (!cpu_online(cpu)) + return -EINVAL; + ++ /* Move the downtaker off the unplug cpu */ ++ if (!alloc_cpumask_var(&cpumask, GFP_KERNEL)) ++ return -ENOMEM; ++ if (!alloc_cpumask_var(&cpumask_org, GFP_KERNEL)) { ++ free_cpumask_var(cpumask); ++ return -ENOMEM; ++ } ++ ++ cpumask_copy(cpumask_org, tsk_cpus_allowed(current)); ++ cpumask_andnot(cpumask, cpu_online_mask, cpumask_of(cpu)); ++ set_cpus_allowed_ptr(current, cpumask); ++ free_cpumask_var(cpumask); ++ migrate_disable(); ++ mycpu = smp_processor_id(); ++ if (mycpu == cpu) { ++ printk(KERN_ERR "Yuck! Still on unplug CPU\n!"); ++ migrate_enable(); ++ err = -EBUSY; ++ goto restore_cpus; ++ } ++ migrate_enable(); ++ + cpu_hotplug_begin(); ++ err = cpu_unplug_begin(cpu); ++ if (err) { ++ printk("cpu_unplug_begin(%d) failed\n", cpu); ++ goto out_cancel; ++ } + + err = __cpu_notify(CPU_DOWN_PREPARE | mod, hcpu, -1, &nr_calls); + if (err) { +@@ -389,8 +701,12 @@ + #endif + synchronize_rcu(); + ++ __cpu_unplug_wait(cpu); + smpboot_park_threads(cpu); + ++ /* Notifiers are done. Don't let any more tasks pin this CPU. */ ++ cpu_unplug_sync(cpu); ++ + /* + * So now all preempt/rcu users must observe !cpu_active(). + */ +@@ -427,9 +743,14 @@ + check_for_tasks(cpu); + + out_release: ++ cpu_unplug_done(cpu); ++out_cancel: + cpu_hotplug_done(); + if (!err) + cpu_notify_nofail(CPU_POST_DEAD | mod, hcpu); ++restore_cpus: ++ set_cpus_allowed_ptr(current, cpumask_org); ++ free_cpumask_var(cpumask_org); + return err; + } + +diff -Nur linux-4.1.13.orig/kernel/debug/kdb/kdb_io.c linux-4.1.13/kernel/debug/kdb/kdb_io.c +--- linux-4.1.13.orig/kernel/debug/kdb/kdb_io.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/debug/kdb/kdb_io.c 2015-11-29 09:23:09.633612589 +0100 +@@ -554,7 +554,6 @@ + int linecount; + int colcount; + int logging, saved_loglevel = 0; +- int saved_trap_printk; + int got_printf_lock = 0; + int retlen = 0; + int fnd, len; +@@ -565,8 +564,6 @@ + unsigned long uninitialized_var(flags); + + preempt_disable(); +- saved_trap_printk = kdb_trap_printk; +- kdb_trap_printk = 0; + + /* Serialize kdb_printf if multiple cpus try to write at once. + * But if any cpu goes recursive in kdb, just print the output, +@@ -855,7 +852,6 @@ + } else { + __release(kdb_printf_lock); + } +- kdb_trap_printk = saved_trap_printk; + preempt_enable(); + return retlen; + } +@@ -865,9 +861,11 @@ + va_list ap; + int r; + ++ kdb_trap_printk++; + va_start(ap, fmt); + r = vkdb_printf(KDB_MSGSRC_INTERNAL, fmt, ap); + va_end(ap); ++ kdb_trap_printk--; + + return r; + } +diff -Nur linux-4.1.13.orig/kernel/events/core.c linux-4.1.13/kernel/events/core.c +--- linux-4.1.13.orig/kernel/events/core.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/events/core.c 2015-11-29 09:23:09.633612589 +0100 +@@ -6925,6 +6925,7 @@ + + hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hwc->hrtimer.function = perf_swevent_hrtimer; ++ hwc->hrtimer.irqsafe = 1; + + /* + * Since hrtimers have a fixed rate, we can do a static freq->period +diff -Nur linux-4.1.13.orig/kernel/exit.c linux-4.1.13/kernel/exit.c +--- linux-4.1.13.orig/kernel/exit.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/exit.c 2015-11-29 09:23:09.633612589 +0100 +@@ -144,7 +144,7 @@ + * Do this under ->siglock, we can race with another thread + * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals. + */ +- flush_sigqueue(&tsk->pending); ++ flush_task_sigqueue(tsk); + tsk->sighand = NULL; + spin_unlock(&sighand->siglock); + +diff -Nur linux-4.1.13.orig/kernel/fork.c linux-4.1.13/kernel/fork.c +--- linux-4.1.13.orig/kernel/fork.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/fork.c 2015-11-29 09:23:09.633612589 +0100 +@@ -108,7 +108,7 @@ + + DEFINE_PER_CPU(unsigned long, process_counts) = 0; + +-__cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */ ++DEFINE_RWLOCK(tasklist_lock); /* outer */ + + #ifdef CONFIG_PROVE_RCU + int lockdep_tasklist_lock_is_held(void) +@@ -244,7 +244,9 @@ + if (atomic_dec_and_test(&sig->sigcnt)) + free_signal_struct(sig); + } +- ++#ifdef CONFIG_PREEMPT_RT_BASE ++static ++#endif + void __put_task_struct(struct task_struct *tsk) + { + WARN_ON(!tsk->exit_state); +@@ -260,7 +262,18 @@ + if (!profile_handoff_task(tsk)) + free_task(tsk); + } ++#ifndef CONFIG_PREEMPT_RT_BASE + EXPORT_SYMBOL_GPL(__put_task_struct); ++#else ++void __put_task_struct_cb(struct rcu_head *rhp) ++{ ++ struct task_struct *tsk = container_of(rhp, struct task_struct, put_rcu); ++ ++ __put_task_struct(tsk); ++ ++} ++EXPORT_SYMBOL_GPL(__put_task_struct_cb); ++#endif + + void __init __weak arch_task_cache_init(void) { } + +@@ -680,6 +693,19 @@ + } + EXPORT_SYMBOL_GPL(__mmdrop); + ++#ifdef CONFIG_PREEMPT_RT_BASE ++/* ++ * RCU callback for delayed mm drop. Not strictly rcu, but we don't ++ * want another facility to make this work. ++ */ ++void __mmdrop_delayed(struct rcu_head *rhp) ++{ ++ struct mm_struct *mm = container_of(rhp, struct mm_struct, delayed_drop); ++ ++ __mmdrop(mm); ++} ++#endif ++ + /* + * Decrement the use count and release all resources for an mm. + */ +@@ -1214,6 +1240,9 @@ + */ + static void posix_cpu_timers_init(struct task_struct *tsk) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ tsk->posix_timer_list = NULL; ++#endif + tsk->cputime_expires.prof_exp = 0; + tsk->cputime_expires.virt_exp = 0; + tsk->cputime_expires.sched_exp = 0; +@@ -1338,6 +1367,7 @@ + spin_lock_init(&p->alloc_lock); + + init_sigpending(&p->pending); ++ p->sigqueue_cache = NULL; + + p->utime = p->stime = p->gtime = 0; + p->utimescaled = p->stimescaled = 0; +@@ -1345,7 +1375,8 @@ + p->prev_cputime.utime = p->prev_cputime.stime = 0; + #endif + #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN +- seqlock_init(&p->vtime_seqlock); ++ raw_spin_lock_init(&p->vtime_lock); ++ seqcount_init(&p->vtime_seq); + p->vtime_snap = 0; + p->vtime_snap_whence = VTIME_SLEEPING; + #endif +@@ -1396,6 +1427,9 @@ + p->hardirq_context = 0; + p->softirq_context = 0; + #endif ++ ++ p->pagefault_disabled = 0; ++ + #ifdef CONFIG_LOCKDEP + p->lockdep_depth = 0; /* no locks held yet */ + p->curr_chain_key = 0; +diff -Nur linux-4.1.13.orig/kernel/futex.c linux-4.1.13/kernel/futex.c +--- linux-4.1.13.orig/kernel/futex.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/futex.c 2015-11-29 09:23:09.633612589 +0100 +@@ -738,7 +738,9 @@ + * task still owns the PI-state: + */ + if (head->next != next) { ++ raw_spin_unlock_irq(&curr->pi_lock); + spin_unlock(&hb->lock); ++ raw_spin_lock_irq(&curr->pi_lock); + continue; + } + +@@ -1090,9 +1092,11 @@ + + /* + * The hash bucket lock must be held when this is called. +- * Afterwards, the futex_q must not be accessed. ++ * Afterwards, the futex_q must not be accessed. Callers ++ * must ensure to later call wake_up_q() for the actual ++ * wakeups to occur. + */ +-static void wake_futex(struct futex_q *q) ++static void mark_wake_futex(struct wake_q_head *wake_q, struct futex_q *q) + { + struct task_struct *p = q->task; + +@@ -1100,14 +1104,10 @@ + return; + + /* +- * We set q->lock_ptr = NULL _before_ we wake up the task. If +- * a non-futex wake up happens on another CPU then the task +- * might exit and p would dereference a non-existing task +- * struct. Prevent this by holding a reference on p across the +- * wake up. ++ * Queue the task for later wakeup for after we've released ++ * the hb->lock. wake_q_add() grabs reference to p. + */ +- get_task_struct(p); +- ++ wake_q_add(wake_q, p); + __unqueue_futex(q); + /* + * The waiting task can free the futex_q as soon as +@@ -1117,16 +1117,15 @@ + */ + smp_wmb(); + q->lock_ptr = NULL; +- +- wake_up_state(p, TASK_NORMAL); +- put_task_struct(p); + } + +-static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this) ++static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, ++ struct futex_hash_bucket *hb) + { + struct task_struct *new_owner; + struct futex_pi_state *pi_state = this->pi_state; + u32 uninitialized_var(curval), newval; ++ bool deboost; + int ret = 0; + + if (!pi_state) +@@ -1178,7 +1177,17 @@ + raw_spin_unlock_irq(&new_owner->pi_lock); + + raw_spin_unlock(&pi_state->pi_mutex.wait_lock); +- rt_mutex_unlock(&pi_state->pi_mutex); ++ ++ deboost = rt_mutex_futex_unlock(&pi_state->pi_mutex); ++ ++ /* ++ * We deboost after dropping hb->lock. That prevents a double ++ * wakeup on RT. ++ */ ++ spin_unlock(&hb->lock); ++ ++ if (deboost) ++ rt_mutex_adjust_prio(current); + + return 0; + } +@@ -1217,6 +1226,7 @@ + struct futex_q *this, *next; + union futex_key key = FUTEX_KEY_INIT; + int ret; ++ WAKE_Q(wake_q); + + if (!bitset) + return -EINVAL; +@@ -1244,13 +1254,14 @@ + if (!(this->bitset & bitset)) + continue; + +- wake_futex(this); ++ mark_wake_futex(&wake_q, this); + if (++ret >= nr_wake) + break; + } + } + + spin_unlock(&hb->lock); ++ wake_up_q(&wake_q); + out_put_key: + put_futex_key(&key); + out: +@@ -1269,6 +1280,7 @@ + struct futex_hash_bucket *hb1, *hb2; + struct futex_q *this, *next; + int ret, op_ret; ++ WAKE_Q(wake_q); + + retry: + ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1, VERIFY_READ); +@@ -1320,7 +1332,7 @@ + ret = -EINVAL; + goto out_unlock; + } +- wake_futex(this); ++ mark_wake_futex(&wake_q, this); + if (++ret >= nr_wake) + break; + } +@@ -1334,7 +1346,7 @@ + ret = -EINVAL; + goto out_unlock; + } +- wake_futex(this); ++ mark_wake_futex(&wake_q, this); + if (++op_ret >= nr_wake2) + break; + } +@@ -1344,6 +1356,7 @@ + + out_unlock: + double_unlock_hb(hb1, hb2); ++ wake_up_q(&wake_q); + out_put_keys: + put_futex_key(&key2); + out_put_key1: +@@ -1503,6 +1516,7 @@ + struct futex_pi_state *pi_state = NULL; + struct futex_hash_bucket *hb1, *hb2; + struct futex_q *this, *next; ++ WAKE_Q(wake_q); + + if (requeue_pi) { + /* +@@ -1679,7 +1693,7 @@ + * woken by futex_unlock_pi(). + */ + if (++task_count <= nr_wake && !requeue_pi) { +- wake_futex(this); ++ mark_wake_futex(&wake_q, this); + continue; + } + +@@ -1705,6 +1719,16 @@ + requeue_pi_wake_futex(this, &key2, hb2); + drop_count++; + continue; ++ } else if (ret == -EAGAIN) { ++ /* ++ * Waiter was woken by timeout or ++ * signal and has set pi_blocked_on to ++ * PI_WAKEUP_INPROGRESS before we ++ * tried to enqueue it on the rtmutex. ++ */ ++ this->pi_state = NULL; ++ free_pi_state(pi_state); ++ continue; + } else if (ret) { + /* -EDEADLK */ + this->pi_state = NULL; +@@ -1719,6 +1743,7 @@ + out_unlock: + free_pi_state(pi_state); + double_unlock_hb(hb1, hb2); ++ wake_up_q(&wake_q); + hb_waiters_dec(hb2); + + /* +@@ -2412,13 +2437,26 @@ + */ + match = futex_top_waiter(hb, &key); + if (match) { +- ret = wake_futex_pi(uaddr, uval, match); ++ ret = wake_futex_pi(uaddr, uval, match, hb); ++ ++ /* ++ * In case of success wake_futex_pi dropped the hash ++ * bucket lock. ++ */ ++ if (!ret) ++ goto out_putkey; ++ + /* + * The atomic access to the futex value generated a + * pagefault, so retry the user-access and the wakeup: + */ + if (ret == -EFAULT) + goto pi_faulted; ++ ++ /* ++ * wake_futex_pi has detected invalid state. Tell user ++ * space. ++ */ + goto out_unlock; + } + +@@ -2439,6 +2477,7 @@ + + out_unlock: + spin_unlock(&hb->lock); ++out_putkey: + put_futex_key(&key); + return ret; + +@@ -2549,7 +2588,7 @@ + struct hrtimer_sleeper timeout, *to = NULL; + struct rt_mutex_waiter rt_waiter; + struct rt_mutex *pi_mutex = NULL; +- struct futex_hash_bucket *hb; ++ struct futex_hash_bucket *hb, *hb2; + union futex_key key2 = FUTEX_KEY_INIT; + struct futex_q q = futex_q_init; + int res, ret; +@@ -2574,10 +2613,7 @@ + * The waiter is allocated on our stack, manipulated by the requeue + * code while we sleep on uaddr. + */ +- debug_rt_mutex_init_waiter(&rt_waiter); +- RB_CLEAR_NODE(&rt_waiter.pi_tree_entry); +- RB_CLEAR_NODE(&rt_waiter.tree_entry); +- rt_waiter.task = NULL; ++ rt_mutex_init_waiter(&rt_waiter, false); + + ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE); + if (unlikely(ret != 0)) +@@ -2608,20 +2644,55 @@ + /* Queue the futex_q, drop the hb lock, wait for wakeup. */ + futex_wait_queue_me(hb, &q, to); + +- spin_lock(&hb->lock); +- ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); +- spin_unlock(&hb->lock); +- if (ret) +- goto out_put_keys; ++ /* ++ * On RT we must avoid races with requeue and trying to block ++ * on two mutexes (hb->lock and uaddr2's rtmutex) by ++ * serializing access to pi_blocked_on with pi_lock. ++ */ ++ raw_spin_lock_irq(¤t->pi_lock); ++ if (current->pi_blocked_on) { ++ /* ++ * We have been requeued or are in the process of ++ * being requeued. ++ */ ++ raw_spin_unlock_irq(¤t->pi_lock); ++ } else { ++ /* ++ * Setting pi_blocked_on to PI_WAKEUP_INPROGRESS ++ * prevents a concurrent requeue from moving us to the ++ * uaddr2 rtmutex. After that we can safely acquire ++ * (and possibly block on) hb->lock. ++ */ ++ current->pi_blocked_on = PI_WAKEUP_INPROGRESS; ++ raw_spin_unlock_irq(¤t->pi_lock); ++ ++ spin_lock(&hb->lock); ++ ++ /* ++ * Clean up pi_blocked_on. We might leak it otherwise ++ * when we succeeded with the hb->lock in the fast ++ * path. ++ */ ++ raw_spin_lock_irq(¤t->pi_lock); ++ current->pi_blocked_on = NULL; ++ raw_spin_unlock_irq(¤t->pi_lock); ++ ++ ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); ++ spin_unlock(&hb->lock); ++ if (ret) ++ goto out_put_keys; ++ } + + /* +- * In order for us to be here, we know our q.key == key2, and since +- * we took the hb->lock above, we also know that futex_requeue() has +- * completed and we no longer have to concern ourselves with a wakeup +- * race with the atomic proxy lock acquisition by the requeue code. The +- * futex_requeue dropped our key1 reference and incremented our key2 +- * reference count. ++ * In order to be here, we have either been requeued, are in ++ * the process of being requeued, or requeue successfully ++ * acquired uaddr2 on our behalf. If pi_blocked_on was ++ * non-null above, we may be racing with a requeue. Do not ++ * rely on q->lock_ptr to be hb2->lock until after blocking on ++ * hb->lock or hb2->lock. The futex_requeue dropped our key1 ++ * reference and incremented our key2 reference count. + */ ++ hb2 = hash_futex(&key2); + + /* Check if the requeue code acquired the second futex for us. */ + if (!q.rt_waiter) { +@@ -2630,9 +2701,10 @@ + * did a lock-steal - fix up the PI-state in that case. + */ + if (q.pi_state && (q.pi_state->owner != current)) { +- spin_lock(q.lock_ptr); ++ spin_lock(&hb2->lock); ++ BUG_ON(&hb2->lock != q.lock_ptr); + ret = fixup_pi_state_owner(uaddr2, &q, current); +- spin_unlock(q.lock_ptr); ++ spin_unlock(&hb2->lock); + } + } else { + /* +@@ -2645,7 +2717,8 @@ + ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter); + debug_rt_mutex_free_waiter(&rt_waiter); + +- spin_lock(q.lock_ptr); ++ spin_lock(&hb2->lock); ++ BUG_ON(&hb2->lock != q.lock_ptr); + /* + * Fixup the pi_state owner and possibly acquire the lock if we + * haven't already. +diff -Nur linux-4.1.13.orig/kernel/irq/handle.c linux-4.1.13/kernel/irq/handle.c +--- linux-4.1.13.orig/kernel/irq/handle.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/irq/handle.c 2015-11-29 09:23:09.633612589 +0100 +@@ -133,6 +133,8 @@ + irqreturn_t + handle_irq_event_percpu(struct irq_desc *desc, struct irqaction *action) + { ++ struct pt_regs *regs = get_irq_regs(); ++ u64 ip = regs ? instruction_pointer(regs) : 0; + irqreturn_t retval = IRQ_NONE; + unsigned int flags = 0, irq = desc->irq_data.irq; + +@@ -173,7 +175,11 @@ + action = action->next; + } while (action); + +- add_interrupt_randomness(irq, flags); ++#ifndef CONFIG_PREEMPT_RT_FULL ++ add_interrupt_randomness(irq, flags, ip); ++#else ++ desc->random_ip = ip; ++#endif + + if (!noirqdebug) + note_interrupt(irq, desc, retval); +diff -Nur linux-4.1.13.orig/kernel/irq/manage.c linux-4.1.13/kernel/irq/manage.c +--- linux-4.1.13.orig/kernel/irq/manage.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/irq/manage.c 2015-11-29 09:23:09.633612589 +0100 +@@ -22,6 +22,7 @@ + #include "internals.h" + + #ifdef CONFIG_IRQ_FORCED_THREADING ++# ifndef CONFIG_PREEMPT_RT_BASE + __read_mostly bool force_irqthreads; + + static int __init setup_forced_irqthreads(char *arg) +@@ -30,6 +31,7 @@ + return 0; + } + early_param("threadirqs", setup_forced_irqthreads); ++# endif + #endif + + static void __synchronize_hardirq(struct irq_desc *desc) +@@ -179,6 +181,62 @@ + irq_get_pending(struct cpumask *mask, struct irq_desc *desc) { } + #endif + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void _irq_affinity_notify(struct irq_affinity_notify *notify); ++static struct task_struct *set_affinity_helper; ++static LIST_HEAD(affinity_list); ++static DEFINE_RAW_SPINLOCK(affinity_list_lock); ++ ++static int set_affinity_thread(void *unused) ++{ ++ while (1) { ++ struct irq_affinity_notify *notify; ++ int empty; ++ ++ set_current_state(TASK_INTERRUPTIBLE); ++ ++ raw_spin_lock_irq(&affinity_list_lock); ++ empty = list_empty(&affinity_list); ++ raw_spin_unlock_irq(&affinity_list_lock); ++ ++ if (empty) ++ schedule(); ++ if (kthread_should_stop()) ++ break; ++ set_current_state(TASK_RUNNING); ++try_next: ++ notify = NULL; ++ ++ raw_spin_lock_irq(&affinity_list_lock); ++ if (!list_empty(&affinity_list)) { ++ notify = list_first_entry(&affinity_list, ++ struct irq_affinity_notify, list); ++ list_del_init(¬ify->list); ++ } ++ raw_spin_unlock_irq(&affinity_list_lock); ++ ++ if (!notify) ++ continue; ++ _irq_affinity_notify(notify); ++ goto try_next; ++ } ++ return 0; ++} ++ ++static void init_helper_thread(void) ++{ ++ if (set_affinity_helper) ++ return; ++ set_affinity_helper = kthread_run(set_affinity_thread, NULL, ++ "affinity-cb"); ++ WARN_ON(IS_ERR(set_affinity_helper)); ++} ++#else ++ ++static inline void init_helper_thread(void) { } ++ ++#endif ++ + int irq_do_set_affinity(struct irq_data *data, const struct cpumask *mask, + bool force) + { +@@ -218,7 +276,17 @@ + + if (desc->affinity_notify) { + kref_get(&desc->affinity_notify->kref); ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ raw_spin_lock(&affinity_list_lock); ++ if (list_empty(&desc->affinity_notify->list)) ++ list_add_tail(&affinity_list, ++ &desc->affinity_notify->list); ++ raw_spin_unlock(&affinity_list_lock); ++ wake_up_process(set_affinity_helper); ++#else + schedule_work(&desc->affinity_notify->work); ++#endif + } + irqd_set(data, IRQD_AFFINITY_SET); + +@@ -256,10 +324,8 @@ + } + EXPORT_SYMBOL_GPL(irq_set_affinity_hint); + +-static void irq_affinity_notify(struct work_struct *work) ++static void _irq_affinity_notify(struct irq_affinity_notify *notify) + { +- struct irq_affinity_notify *notify = +- container_of(work, struct irq_affinity_notify, work); + struct irq_desc *desc = irq_to_desc(notify->irq); + cpumask_var_t cpumask; + unsigned long flags; +@@ -281,6 +347,13 @@ + kref_put(¬ify->kref, notify->release); + } + ++static void irq_affinity_notify(struct work_struct *work) ++{ ++ struct irq_affinity_notify *notify = ++ container_of(work, struct irq_affinity_notify, work); ++ _irq_affinity_notify(notify); ++} ++ + /** + * irq_set_affinity_notifier - control notification of IRQ affinity changes + * @irq: Interrupt for which to enable/disable notification +@@ -310,6 +383,8 @@ + notify->irq = irq; + kref_init(¬ify->kref); + INIT_WORK(¬ify->work, irq_affinity_notify); ++ INIT_LIST_HEAD(¬ify->list); ++ init_helper_thread(); + } + + raw_spin_lock_irqsave(&desc->lock, flags); +@@ -697,6 +772,12 @@ + return IRQ_NONE; + } + ++static irqreturn_t irq_forced_secondary_handler(int irq, void *dev_id) ++{ ++ WARN(1, "Secondary action handler called for irq %d\n", irq); ++ return IRQ_NONE; ++} ++ + static int irq_wait_for_interrupt(struct irqaction *action) + { + set_current_state(TASK_INTERRUPTIBLE); +@@ -723,7 +804,8 @@ + static void irq_finalize_oneshot(struct irq_desc *desc, + struct irqaction *action) + { +- if (!(desc->istate & IRQS_ONESHOT)) ++ if (!(desc->istate & IRQS_ONESHOT) || ++ action->handler == irq_forced_secondary_handler) + return; + again: + chip_bus_lock(desc); +@@ -825,7 +907,15 @@ + local_bh_disable(); + ret = action->thread_fn(action->irq, action->dev_id); + irq_finalize_oneshot(desc, action); +- local_bh_enable(); ++ /* ++ * Interrupts which have real time requirements can be set up ++ * to avoid softirq processing in the thread handler. This is ++ * safe as these interrupts do not raise soft interrupts. ++ */ ++ if (irq_settings_no_softirq_call(desc)) ++ _local_bh_enable(); ++ else ++ local_bh_enable(); + return ret; + } + +@@ -877,6 +967,18 @@ + irq_finalize_oneshot(desc, action); + } + ++static void irq_wake_secondary(struct irq_desc *desc, struct irqaction *action) ++{ ++ struct irqaction *secondary = action->secondary; ++ ++ if (WARN_ON_ONCE(!secondary)) ++ return; ++ ++ raw_spin_lock_irq(&desc->lock); ++ __irq_wake_thread(desc, secondary); ++ raw_spin_unlock_irq(&desc->lock); ++} ++ + /* + * Interrupt handler thread + */ +@@ -907,7 +1009,15 @@ + action_ret = handler_fn(desc, action); + if (action_ret == IRQ_HANDLED) + atomic_inc(&desc->threads_handled); ++ if (action_ret == IRQ_WAKE_THREAD) ++ irq_wake_secondary(desc, action); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ migrate_disable(); ++ add_interrupt_randomness(action->irq, 0, ++ desc->random_ip ^ (unsigned long) action); ++ migrate_enable(); ++#endif + wake_threads_waitq(desc); + } + +@@ -951,20 +1061,36 @@ + } + EXPORT_SYMBOL_GPL(irq_wake_thread); + +-static void irq_setup_forced_threading(struct irqaction *new) ++static int irq_setup_forced_threading(struct irqaction *new) + { + if (!force_irqthreads) +- return; ++ return 0; + if (new->flags & (IRQF_NO_THREAD | IRQF_PERCPU | IRQF_ONESHOT)) +- return; ++ return 0; + + new->flags |= IRQF_ONESHOT; + +- if (!new->thread_fn) { +- set_bit(IRQTF_FORCED_THREAD, &new->thread_flags); +- new->thread_fn = new->handler; +- new->handler = irq_default_primary_handler; +- } ++ /* ++ * Handle the case where we have a real primary handler and a ++ * thread handler. We force thread them as well by creating a ++ * secondary action. ++ */ ++ if (new->handler != irq_default_primary_handler && new->thread_fn) { ++ /* Allocate the secondary action */ ++ new->secondary = kzalloc(sizeof(struct irqaction), GFP_KERNEL); ++ if (!new->secondary) ++ return -ENOMEM; ++ new->secondary->handler = irq_forced_secondary_handler; ++ new->secondary->thread_fn = new->thread_fn; ++ new->secondary->dev_id = new->dev_id; ++ new->secondary->irq = new->irq; ++ new->secondary->name = new->name; ++ } ++ /* Deal with the primary handler */ ++ set_bit(IRQTF_FORCED_THREAD, &new->thread_flags); ++ new->thread_fn = new->handler; ++ new->handler = irq_default_primary_handler; ++ return 0; + } + + static int irq_request_resources(struct irq_desc *desc) +@@ -984,6 +1110,48 @@ + c->irq_release_resources(d); + } + ++static int ++setup_irq_thread(struct irqaction *new, unsigned int irq, bool secondary) ++{ ++ struct task_struct *t; ++ struct sched_param param = { ++ .sched_priority = MAX_USER_RT_PRIO/2, ++ }; ++ ++ if (!secondary) { ++ t = kthread_create(irq_thread, new, "irq/%d-%s", irq, ++ new->name); ++ } else { ++ t = kthread_create(irq_thread, new, "irq/%d-s-%s", irq, ++ new->name); ++ param.sched_priority += 1; ++ } ++ ++ if (IS_ERR(t)) ++ return PTR_ERR(t); ++ ++ sched_setscheduler_nocheck(t, SCHED_FIFO, ¶m); ++ ++ /* ++ * We keep the reference to the task struct even if ++ * the thread dies to avoid that the interrupt code ++ * references an already freed task_struct. ++ */ ++ get_task_struct(t); ++ new->thread = t; ++ /* ++ * Tell the thread to set its affinity. This is ++ * important for shared interrupt handlers as we do ++ * not invoke setup_affinity() for the secondary ++ * handlers as everything is already set up. Even for ++ * interrupts marked with IRQF_NO_BALANCE this is ++ * correct as we want the thread to move to the cpu(s) ++ * on which the requesting code placed the interrupt. ++ */ ++ set_bit(IRQTF_AFFINITY, &new->thread_flags); ++ return 0; ++} ++ + /* + * Internal function to register an irqaction - typically used to + * allocate special interrupts that are part of the architecture. +@@ -1004,6 +1172,8 @@ + if (!try_module_get(desc->owner)) + return -ENODEV; + ++ new->irq = irq; ++ + /* + * Check whether the interrupt nests into another interrupt + * thread. +@@ -1021,8 +1191,11 @@ + */ + new->handler = irq_nested_primary_handler; + } else { +- if (irq_settings_can_thread(desc)) +- irq_setup_forced_threading(new); ++ if (irq_settings_can_thread(desc)) { ++ ret = irq_setup_forced_threading(new); ++ if (ret) ++ goto out_mput; ++ } + } + + /* +@@ -1031,37 +1204,14 @@ + * thread. + */ + if (new->thread_fn && !nested) { +- struct task_struct *t; +- static const struct sched_param param = { +- .sched_priority = MAX_USER_RT_PRIO/2, +- }; +- +- t = kthread_create(irq_thread, new, "irq/%d-%s", irq, +- new->name); +- if (IS_ERR(t)) { +- ret = PTR_ERR(t); ++ ret = setup_irq_thread(new, irq, false); ++ if (ret) + goto out_mput; ++ if (new->secondary) { ++ ret = setup_irq_thread(new->secondary, irq, true); ++ if (ret) ++ goto out_thread; + } +- +- sched_setscheduler_nocheck(t, SCHED_FIFO, ¶m); +- +- /* +- * We keep the reference to the task struct even if +- * the thread dies to avoid that the interrupt code +- * references an already freed task_struct. +- */ +- get_task_struct(t); +- new->thread = t; +- /* +- * Tell the thread to set its affinity. This is +- * important for shared interrupt handlers as we do +- * not invoke setup_affinity() for the secondary +- * handlers as everything is already set up. Even for +- * interrupts marked with IRQF_NO_BALANCE this is +- * correct as we want the thread to move to the cpu(s) +- * on which the requesting code placed the interrupt. +- */ +- set_bit(IRQTF_AFFINITY, &new->thread_flags); + } + + if (!alloc_cpumask_var(&mask, GFP_KERNEL)) { +@@ -1221,6 +1371,9 @@ + irqd_set(&desc->irq_data, IRQD_NO_BALANCING); + } + ++ if (new->flags & IRQF_NO_SOFTIRQ_CALL) ++ irq_settings_set_no_softirq_call(desc); ++ + /* Set default affinity mask once everything is setup */ + setup_affinity(irq, desc, mask); + +@@ -1234,7 +1387,6 @@ + irq, nmsk, omsk); + } + +- new->irq = irq; + *old_ptr = new; + + irq_pm_install_action(desc, new); +@@ -1260,6 +1412,8 @@ + */ + if (new->thread) + wake_up_process(new->thread); ++ if (new->secondary) ++ wake_up_process(new->secondary->thread); + + register_irq_proc(irq, desc); + new->dir = NULL; +@@ -1290,6 +1444,13 @@ + kthread_stop(t); + put_task_struct(t); + } ++ if (new->secondary && new->secondary->thread) { ++ struct task_struct *t = new->secondary->thread; ++ ++ new->secondary->thread = NULL; ++ kthread_stop(t); ++ put_task_struct(t); ++ } + out_mput: + module_put(desc->owner); + return ret; +@@ -1397,9 +1558,14 @@ + if (action->thread) { + kthread_stop(action->thread); + put_task_struct(action->thread); ++ if (action->secondary && action->secondary->thread) { ++ kthread_stop(action->secondary->thread); ++ put_task_struct(action->secondary->thread); ++ } + } + + module_put(desc->owner); ++ kfree(action->secondary); + return action; + } + +@@ -1543,8 +1709,10 @@ + retval = __setup_irq(irq, desc, action); + chip_bus_sync_unlock(desc); + +- if (retval) ++ if (retval) { ++ kfree(action->secondary); + kfree(action); ++ } + + #ifdef CONFIG_DEBUG_SHIRQ_FIXME + if (!retval && (irqflags & IRQF_SHARED)) { +diff -Nur linux-4.1.13.orig/kernel/irq/settings.h linux-4.1.13/kernel/irq/settings.h +--- linux-4.1.13.orig/kernel/irq/settings.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/irq/settings.h 2015-11-29 09:23:09.637612322 +0100 +@@ -15,6 +15,7 @@ + _IRQ_NESTED_THREAD = IRQ_NESTED_THREAD, + _IRQ_PER_CPU_DEVID = IRQ_PER_CPU_DEVID, + _IRQ_IS_POLLED = IRQ_IS_POLLED, ++ _IRQ_NO_SOFTIRQ_CALL = IRQ_NO_SOFTIRQ_CALL, + _IRQF_MODIFY_MASK = IRQF_MODIFY_MASK, + }; + +@@ -28,6 +29,7 @@ + #define IRQ_NESTED_THREAD GOT_YOU_MORON + #define IRQ_PER_CPU_DEVID GOT_YOU_MORON + #define IRQ_IS_POLLED GOT_YOU_MORON ++#define IRQ_NO_SOFTIRQ_CALL GOT_YOU_MORON + #undef IRQF_MODIFY_MASK + #define IRQF_MODIFY_MASK GOT_YOU_MORON + +@@ -38,6 +40,16 @@ + desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK); + } + ++static inline bool irq_settings_no_softirq_call(struct irq_desc *desc) ++{ ++ return desc->status_use_accessors & _IRQ_NO_SOFTIRQ_CALL; ++} ++ ++static inline void irq_settings_set_no_softirq_call(struct irq_desc *desc) ++{ ++ desc->status_use_accessors |= _IRQ_NO_SOFTIRQ_CALL; ++} ++ + static inline bool irq_settings_is_per_cpu(struct irq_desc *desc) + { + return desc->status_use_accessors & _IRQ_PER_CPU; +diff -Nur linux-4.1.13.orig/kernel/irq/spurious.c linux-4.1.13/kernel/irq/spurious.c +--- linux-4.1.13.orig/kernel/irq/spurious.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/irq/spurious.c 2015-11-29 09:23:09.637612322 +0100 +@@ -444,6 +444,10 @@ + + static int __init irqfixup_setup(char *str) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ pr_warn("irqfixup boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); ++ return 1; ++#endif + irqfixup = 1; + printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n"); + printk(KERN_WARNING "This may impact system performance.\n"); +@@ -456,6 +460,10 @@ + + static int __init irqpoll_setup(char *str) + { ++#ifdef CONFIG_PREEMPT_RT_BASE ++ pr_warn("irqpoll boot option not supported w/ CONFIG_PREEMPT_RT_BASE\n"); ++ return 1; ++#endif + irqfixup = 2; + printk(KERN_WARNING "Misrouted IRQ fixup and polling support " + "enabled\n"); +diff -Nur linux-4.1.13.orig/kernel/irq_work.c linux-4.1.13/kernel/irq_work.c +--- linux-4.1.13.orig/kernel/irq_work.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/irq_work.c 2015-11-29 09:23:09.637612322 +0100 +@@ -17,6 +17,7 @@ + #include <linux/cpu.h> + #include <linux/notifier.h> + #include <linux/smp.h> ++#include <linux/interrupt.h> + #include <asm/processor.h> + + +@@ -65,6 +66,8 @@ + */ + bool irq_work_queue_on(struct irq_work *work, int cpu) + { ++ struct llist_head *list; ++ + /* All work should have been flushed before going offline */ + WARN_ON_ONCE(cpu_is_offline(cpu)); + +@@ -75,7 +78,12 @@ + if (!irq_work_claim(work)) + return false; + +- if (llist_add(&work->llnode, &per_cpu(raised_list, cpu))) ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL) && !(work->flags & IRQ_WORK_HARD_IRQ)) ++ list = &per_cpu(lazy_list, cpu); ++ else ++ list = &per_cpu(raised_list, cpu); ++ ++ if (llist_add(&work->llnode, list)) + arch_send_call_function_single_ipi(cpu); + + return true; +@@ -86,6 +94,9 @@ + /* Enqueue the irq work @work on the current CPU */ + bool irq_work_queue(struct irq_work *work) + { ++ struct llist_head *list; ++ bool lazy_work, realtime = IS_ENABLED(CONFIG_PREEMPT_RT_FULL); ++ + /* Only queue if not already pending */ + if (!irq_work_claim(work)) + return false; +@@ -93,13 +104,15 @@ + /* Queue the entry and raise the IPI if needed. */ + preempt_disable(); + +- /* If the work is "lazy", handle it from next tick if any */ +- if (work->flags & IRQ_WORK_LAZY) { +- if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) && +- tick_nohz_tick_stopped()) +- arch_irq_work_raise(); +- } else { +- if (llist_add(&work->llnode, this_cpu_ptr(&raised_list))) ++ lazy_work = work->flags & IRQ_WORK_LAZY; ++ ++ if (lazy_work || (realtime && !(work->flags & IRQ_WORK_HARD_IRQ))) ++ list = this_cpu_ptr(&lazy_list); ++ else ++ list = this_cpu_ptr(&raised_list); ++ ++ if (llist_add(&work->llnode, list)) { ++ if (!lazy_work || tick_nohz_tick_stopped()) + arch_irq_work_raise(); + } + +@@ -116,9 +129,8 @@ + raised = this_cpu_ptr(&raised_list); + lazy = this_cpu_ptr(&lazy_list); + +- if (llist_empty(raised) || arch_irq_work_has_interrupt()) +- if (llist_empty(lazy)) +- return false; ++ if (llist_empty(raised) && llist_empty(lazy)) ++ return false; + + /* All work should have been flushed before going offline */ + WARN_ON_ONCE(cpu_is_offline(smp_processor_id())); +@@ -132,7 +144,7 @@ + struct irq_work *work; + struct llist_node *llnode; + +- BUG_ON(!irqs_disabled()); ++ BUG_ON_NONRT(!irqs_disabled()); + + if (llist_empty(list)) + return; +@@ -169,7 +181,16 @@ + void irq_work_run(void) + { + irq_work_run_list(this_cpu_ptr(&raised_list)); +- irq_work_run_list(this_cpu_ptr(&lazy_list)); ++ if (IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) { ++ /* ++ * NOTE: we raise softirq via IPI for safety, ++ * and execute in irq_work_tick() to move the ++ * overhead from hard to soft irq context. ++ */ ++ if (!llist_empty(this_cpu_ptr(&lazy_list))) ++ raise_softirq(TIMER_SOFTIRQ); ++ } else ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); + } + EXPORT_SYMBOL_GPL(irq_work_run); + +@@ -179,8 +200,17 @@ + + if (!llist_empty(raised) && !arch_irq_work_has_interrupt()) + irq_work_run_list(raised); ++ ++ if (!IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) ++ irq_work_run_list(this_cpu_ptr(&lazy_list)); ++} ++ ++#if defined(CONFIG_IRQ_WORK) && defined(CONFIG_PREEMPT_RT_FULL) ++void irq_work_tick_soft(void) ++{ + irq_work_run_list(this_cpu_ptr(&lazy_list)); + } ++#endif + + /* + * Synchronize against the irq_work @entry, ensures the entry is not +diff -Nur linux-4.1.13.orig/kernel/Kconfig.locks linux-4.1.13/kernel/Kconfig.locks +--- linux-4.1.13.orig/kernel/Kconfig.locks 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/Kconfig.locks 2015-11-29 09:23:09.629612854 +0100 +@@ -225,11 +225,11 @@ + + config MUTEX_SPIN_ON_OWNER + def_bool y +- depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && !DEBUG_MUTEXES && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config RWSEM_SPIN_ON_OWNER + def_bool y +- depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW ++ depends on SMP && RWSEM_XCHGADD_ALGORITHM && ARCH_SUPPORTS_ATOMIC_RMW && !PREEMPT_RT_FULL + + config LOCK_SPIN_ON_OWNER + def_bool y +diff -Nur linux-4.1.13.orig/kernel/Kconfig.preempt linux-4.1.13/kernel/Kconfig.preempt +--- linux-4.1.13.orig/kernel/Kconfig.preempt 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/Kconfig.preempt 2015-11-29 09:23:09.629612854 +0100 +@@ -1,3 +1,16 @@ ++config PREEMPT ++ bool ++ select PREEMPT_COUNT ++ ++config PREEMPT_RT_BASE ++ bool ++ select PREEMPT ++ ++config HAVE_PREEMPT_LAZY ++ bool ++ ++config PREEMPT_LAZY ++ def_bool y if HAVE_PREEMPT_LAZY && PREEMPT_RT_FULL + + choice + prompt "Preemption Model" +@@ -33,9 +46,9 @@ + + Select this if you are building a kernel for a desktop system. + +-config PREEMPT ++config PREEMPT__LL + bool "Preemptible Kernel (Low-Latency Desktop)" +- select PREEMPT_COUNT ++ select PREEMPT + select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK + help + This option reduces the latency of the kernel by making +@@ -52,6 +65,22 @@ + embedded system with latency requirements in the milliseconds + range. + ++config PREEMPT_RTB ++ bool "Preemptible Kernel (Basic RT)" ++ select PREEMPT_RT_BASE ++ help ++ This option is basically the same as (Low-Latency Desktop) but ++ enables changes which are preliminary for the full preemptible ++ RT kernel. ++ ++config PREEMPT_RT_FULL ++ bool "Fully Preemptible Kernel (RT)" ++ depends on IRQ_FORCED_THREADING ++ select PREEMPT_RT_BASE ++ select PREEMPT_RCU ++ help ++ All and everything ++ + endchoice + + config PREEMPT_COUNT +diff -Nur linux-4.1.13.orig/kernel/ksysfs.c linux-4.1.13/kernel/ksysfs.c +--- linux-4.1.13.orig/kernel/ksysfs.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/ksysfs.c 2015-11-29 09:23:09.637612322 +0100 +@@ -136,6 +136,15 @@ + + #endif /* CONFIG_KEXEC */ + ++#if defined(CONFIG_PREEMPT_RT_FULL) ++static ssize_t realtime_show(struct kobject *kobj, ++ struct kobj_attribute *attr, char *buf) ++{ ++ return sprintf(buf, "%d\n", 1); ++} ++KERNEL_ATTR_RO(realtime); ++#endif ++ + /* whether file capabilities are enabled */ + static ssize_t fscaps_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +@@ -203,6 +212,9 @@ + &vmcoreinfo_attr.attr, + #endif + &rcu_expedited_attr.attr, ++#ifdef CONFIG_PREEMPT_RT_FULL ++ &realtime_attr.attr, ++#endif + NULL + }; + +diff -Nur linux-4.1.13.orig/kernel/locking/lglock.c linux-4.1.13/kernel/locking/lglock.c +--- linux-4.1.13.orig/kernel/locking/lglock.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/lglock.c 2015-11-29 09:23:09.637612322 +0100 +@@ -4,6 +4,15 @@ + #include <linux/cpu.h> + #include <linux/string.h> + ++#ifndef CONFIG_PREEMPT_RT_FULL ++# define lg_lock_ptr arch_spinlock_t ++# define lg_do_lock(l) arch_spin_lock(l) ++# define lg_do_unlock(l) arch_spin_unlock(l) ++#else ++# define lg_lock_ptr struct rt_mutex ++# define lg_do_lock(l) __rt_spin_lock(l) ++# define lg_do_unlock(l) __rt_spin_unlock(l) ++#endif + /* + * Note there is no uninit, so lglocks cannot be defined in + * modules (but it's fine to use them from there) +@@ -12,51 +21,60 @@ + + void lg_lock_init(struct lglock *lg, char *name) + { ++#ifdef CONFIG_PREEMPT_RT_FULL ++ int i; ++ ++ for_each_possible_cpu(i) { ++ struct rt_mutex *lock = per_cpu_ptr(lg->lock, i); ++ ++ rt_mutex_init(lock); ++ } ++#endif + LOCKDEP_INIT_MAP(&lg->lock_dep_map, name, &lg->lock_key, 0); + } + EXPORT_SYMBOL(lg_lock_init); + + void lg_local_lock(struct lglock *lg) + { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + +- preempt_disable(); ++ migrate_disable(); + lock_acquire_shared(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); + lock = this_cpu_ptr(lg->lock); +- arch_spin_lock(lock); ++ lg_do_lock(lock); + } + EXPORT_SYMBOL(lg_local_lock); + + void lg_local_unlock(struct lglock *lg) + { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + + lock_release(&lg->lock_dep_map, 1, _RET_IP_); + lock = this_cpu_ptr(lg->lock); +- arch_spin_unlock(lock); +- preempt_enable(); ++ lg_do_unlock(lock); ++ migrate_enable(); + } + EXPORT_SYMBOL(lg_local_unlock); + + void lg_local_lock_cpu(struct lglock *lg, int cpu) + { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + +- preempt_disable(); ++ preempt_disable_nort(); + lock_acquire_shared(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); + lock = per_cpu_ptr(lg->lock, cpu); +- arch_spin_lock(lock); ++ lg_do_lock(lock); + } + EXPORT_SYMBOL(lg_local_lock_cpu); + + void lg_local_unlock_cpu(struct lglock *lg, int cpu) + { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + + lock_release(&lg->lock_dep_map, 1, _RET_IP_); + lock = per_cpu_ptr(lg->lock, cpu); +- arch_spin_unlock(lock); +- preempt_enable(); ++ lg_do_unlock(lock); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(lg_local_unlock_cpu); + +@@ -64,12 +82,12 @@ + { + int i; + +- preempt_disable(); ++ preempt_disable_nort(); + lock_acquire_exclusive(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); + for_each_possible_cpu(i) { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + lock = per_cpu_ptr(lg->lock, i); +- arch_spin_lock(lock); ++ lg_do_lock(lock); + } + } + EXPORT_SYMBOL(lg_global_lock); +@@ -80,10 +98,35 @@ + + lock_release(&lg->lock_dep_map, 1, _RET_IP_); + for_each_possible_cpu(i) { +- arch_spinlock_t *lock; ++ lg_lock_ptr *lock; + lock = per_cpu_ptr(lg->lock, i); +- arch_spin_unlock(lock); ++ lg_do_unlock(lock); + } +- preempt_enable(); ++ preempt_enable_nort(); + } + EXPORT_SYMBOL(lg_global_unlock); ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * HACK: If you use this, you get to keep the pieces. ++ * Used in queue_stop_cpus_work() when stop machinery ++ * is called from inactive CPU, so we can't schedule. ++ */ ++# define lg_do_trylock_relax(l) \ ++ do { \ ++ while (!__rt_spin_trylock(l)) \ ++ cpu_relax(); \ ++ } while (0) ++ ++void lg_global_trylock_relax(struct lglock *lg) ++{ ++ int i; ++ ++ lock_acquire_exclusive(&lg->lock_dep_map, 0, 0, NULL, _RET_IP_); ++ for_each_possible_cpu(i) { ++ lg_lock_ptr *lock; ++ lock = per_cpu_ptr(lg->lock, i); ++ lg_do_trylock_relax(lock); ++ } ++} ++#endif +diff -Nur linux-4.1.13.orig/kernel/locking/lockdep.c linux-4.1.13/kernel/locking/lockdep.c +--- linux-4.1.13.orig/kernel/locking/lockdep.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/lockdep.c 2015-11-29 09:23:09.637612322 +0100 +@@ -3563,6 +3563,7 @@ + } + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * We dont accurately track softirq state in e.g. + * hardirq contexts (such as on 4KSTACKS), so only +@@ -3577,6 +3578,7 @@ + DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled); + } + } ++#endif + + if (!debug_locks) + print_irqtrace_events(current); +diff -Nur linux-4.1.13.orig/kernel/locking/locktorture.c linux-4.1.13/kernel/locking/locktorture.c +--- linux-4.1.13.orig/kernel/locking/locktorture.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/locktorture.c 2015-11-29 09:23:09.637612322 +0100 +@@ -24,7 +24,6 @@ + #include <linux/module.h> + #include <linux/kthread.h> + #include <linux/spinlock.h> +-#include <linux/rwlock.h> + #include <linux/mutex.h> + #include <linux/rwsem.h> + #include <linux/smp.h> +diff -Nur linux-4.1.13.orig/kernel/locking/Makefile linux-4.1.13/kernel/locking/Makefile +--- linux-4.1.13.orig/kernel/locking/Makefile 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/Makefile 2015-11-29 09:23:09.637612322 +0100 +@@ -1,5 +1,5 @@ + +-obj-y += mutex.o semaphore.o rwsem.o ++obj-y += semaphore.o + + ifdef CONFIG_FUNCTION_TRACER + CFLAGS_REMOVE_lockdep.o = $(CC_FLAGS_FTRACE) +@@ -8,7 +8,11 @@ + CFLAGS_REMOVE_rtmutex-debug.o = $(CC_FLAGS_FTRACE) + endif + ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) ++obj-y += mutex.o + obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o ++obj-y += rwsem.o ++endif + obj-$(CONFIG_LOCKDEP) += lockdep.o + ifeq ($(CONFIG_PROC_FS),y) + obj-$(CONFIG_LOCKDEP) += lockdep_proc.o +@@ -22,8 +26,11 @@ + obj-$(CONFIG_RT_MUTEX_TESTER) += rtmutex-tester.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o + obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o ++ifneq ($(CONFIG_PREEMPT_RT_FULL),y) + obj-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o + obj-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem-xadd.o ++endif + obj-$(CONFIG_PERCPU_RWSEM) += percpu-rwsem.o ++obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o + obj-$(CONFIG_QUEUE_RWLOCK) += qrwlock.o + obj-$(CONFIG_LOCK_TORTURE_TEST) += locktorture.o +diff -Nur linux-4.1.13.orig/kernel/locking/rt.c linux-4.1.13/kernel/locking/rt.c +--- linux-4.1.13.orig/kernel/locking/rt.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/kernel/locking/rt.c 2015-11-29 09:23:09.637612322 +0100 +@@ -0,0 +1,461 @@ ++/* ++ * kernel/rt.c ++ * ++ * Real-Time Preemption Support ++ * ++ * started by Ingo Molnar: ++ * ++ * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com> ++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx@timesys.com> ++ * ++ * historic credit for proving that Linux spinlocks can be implemented via ++ * RT-aware mutexes goes to many people: The Pmutex project (Dirk Grambow ++ * and others) who prototyped it on 2.4 and did lots of comparative ++ * research and analysis; TimeSys, for proving that you can implement a ++ * fully preemptible kernel via the use of IRQ threading and mutexes; ++ * Bill Huey for persuasively arguing on lkml that the mutex model is the ++ * right one; and to MontaVista, who ported pmutexes to 2.6. ++ * ++ * This code is a from-scratch implementation and is not based on pmutexes, ++ * but the idea of converting spinlocks to mutexes is used here too. ++ * ++ * lock debugging, locking tree, deadlock detection: ++ * ++ * Copyright (C) 2004, LynuxWorks, Inc., Igor Manyilov, Bill Huey ++ * Released under the General Public License (GPL). ++ * ++ * Includes portions of the generic R/W semaphore implementation from: ++ * ++ * Copyright (c) 2001 David Howells (dhowells@redhat.com). ++ * - Derived partially from idea by Andrea Arcangeli <andrea@suse.de> ++ * - Derived also from comments by Linus ++ * ++ * Pending ownership of locks and ownership stealing: ++ * ++ * Copyright (C) 2005, Kihon Technologies Inc., Steven Rostedt ++ * ++ * (also by Steven Rostedt) ++ * - Converted single pi_lock to individual task locks. ++ * ++ * By Esben Nielsen: ++ * Doing priority inheritance with help of the scheduler. ++ * ++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx@timesys.com> ++ * - major rework based on Esben Nielsens initial patch ++ * - replaced thread_info references by task_struct refs ++ * - removed task->pending_owner dependency ++ * - BKL drop/reacquire for semaphore style locks to avoid deadlocks ++ * in the scheduler return path as discussed with Steven Rostedt ++ * ++ * Copyright (C) 2006, Kihon Technologies Inc. ++ * Steven Rostedt <rostedt@goodmis.org> ++ * - debugged and patched Thomas Gleixner's rework. ++ * - added back the cmpxchg to the rework. ++ * - turned atomic require back on for SMP. ++ */ ++ ++#include <linux/spinlock.h> ++#include <linux/rtmutex.h> ++#include <linux/sched.h> ++#include <linux/delay.h> ++#include <linux/module.h> ++#include <linux/kallsyms.h> ++#include <linux/syscalls.h> ++#include <linux/interrupt.h> ++#include <linux/plist.h> ++#include <linux/fs.h> ++#include <linux/futex.h> ++#include <linux/hrtimer.h> ++ ++#include "rtmutex_common.h" ++ ++/* ++ * struct mutex functions ++ */ ++void __mutex_do_init(struct mutex *mutex, const char *name, ++ struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)mutex, sizeof(*mutex)); ++ lockdep_init_map(&mutex->dep_map, name, key, 0); ++#endif ++ mutex->lock.save_state = 0; ++} ++EXPORT_SYMBOL(__mutex_do_init); ++ ++void __lockfunc _mutex_lock(struct mutex *lock) ++{ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock); ++ ++int __lockfunc _mutex_lock_interruptible(struct mutex *lock) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ ret = rt_mutex_lock_interruptible(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_interruptible); ++ ++int __lockfunc _mutex_lock_killable(struct mutex *lock) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++ ret = rt_mutex_lock_killable(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_killable); ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass) ++{ ++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock_nested); ++ ++void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest) ++{ ++ mutex_acquire_nest(&lock->dep_map, 0, 0, nest, _RET_IP_); ++ rt_mutex_lock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_lock_nest_lock); ++ ++int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass) ++{ ++ int ret; ++ ++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_); ++ ret = rt_mutex_lock_interruptible(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_interruptible_nested); ++ ++int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass) ++{ ++ int ret; ++ ++ mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_); ++ ret = rt_mutex_lock_killable(&lock->lock); ++ if (ret) ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_lock_killable_nested); ++#endif ++ ++int __lockfunc _mutex_trylock(struct mutex *lock) ++{ ++ int ret = rt_mutex_trylock(&lock->lock); ++ ++ if (ret) ++ mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ ++ return ret; ++} ++EXPORT_SYMBOL(_mutex_trylock); ++ ++void __lockfunc _mutex_unlock(struct mutex *lock) ++{ ++ mutex_release(&lock->dep_map, 1, _RET_IP_); ++ rt_mutex_unlock(&lock->lock); ++} ++EXPORT_SYMBOL(_mutex_unlock); ++ ++/* ++ * rwlock_t functions ++ */ ++int __lockfunc rt_write_trylock(rwlock_t *rwlock) ++{ ++ int ret; ++ ++ migrate_disable(); ++ ret = rt_mutex_trylock(&rwlock->lock); ++ if (ret) ++ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); ++ else ++ migrate_enable(); ++ ++ return ret; ++} ++EXPORT_SYMBOL(rt_write_trylock); ++ ++int __lockfunc rt_write_trylock_irqsave(rwlock_t *rwlock, unsigned long *flags) ++{ ++ int ret; ++ ++ *flags = 0; ++ ret = rt_write_trylock(rwlock); ++ return ret; ++} ++EXPORT_SYMBOL(rt_write_trylock_irqsave); ++ ++int __lockfunc rt_read_trylock(rwlock_t *rwlock) ++{ ++ struct rt_mutex *lock = &rwlock->lock; ++ int ret = 1; ++ ++ /* ++ * recursive read locks succeed when current owns the lock, ++ * but not when read_depth == 0 which means that the lock is ++ * write locked. ++ */ ++ if (rt_mutex_owner(lock) != current) { ++ migrate_disable(); ++ ret = rt_mutex_trylock(lock); ++ if (ret) ++ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_); ++ else ++ migrate_enable(); ++ ++ } else if (!rwlock->read_depth) { ++ ret = 0; ++ } ++ ++ if (ret) ++ rwlock->read_depth++; ++ ++ return ret; ++} ++EXPORT_SYMBOL(rt_read_trylock); ++ ++void __lockfunc rt_write_lock(rwlock_t *rwlock) ++{ ++ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); ++ migrate_disable(); ++ __rt_spin_lock(&rwlock->lock); ++} ++EXPORT_SYMBOL(rt_write_lock); ++ ++void __lockfunc rt_read_lock(rwlock_t *rwlock) ++{ ++ struct rt_mutex *lock = &rwlock->lock; ++ ++ ++ /* ++ * recursive read locks succeed when current owns the lock ++ */ ++ if (rt_mutex_owner(lock) != current) { ++ migrate_disable(); ++ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_); ++ __rt_spin_lock(lock); ++ } ++ rwlock->read_depth++; ++} ++ ++EXPORT_SYMBOL(rt_read_lock); ++ ++void __lockfunc rt_write_unlock(rwlock_t *rwlock) ++{ ++ /* NOTE: we always pass in '1' for nested, for simplicity */ ++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); ++ __rt_spin_unlock(&rwlock->lock); ++ migrate_enable(); ++} ++EXPORT_SYMBOL(rt_write_unlock); ++ ++void __lockfunc rt_read_unlock(rwlock_t *rwlock) ++{ ++ /* Release the lock only when read_depth is down to 0 */ ++ if (--rwlock->read_depth == 0) { ++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_); ++ __rt_spin_unlock(&rwlock->lock); ++ migrate_enable(); ++ } ++} ++EXPORT_SYMBOL(rt_read_unlock); ++ ++unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock) ++{ ++ rt_write_lock(rwlock); ++ ++ return 0; ++} ++EXPORT_SYMBOL(rt_write_lock_irqsave); ++ ++unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock) ++{ ++ rt_read_lock(rwlock); ++ ++ return 0; ++} ++EXPORT_SYMBOL(rt_read_lock_irqsave); ++ ++void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)rwlock, sizeof(*rwlock)); ++ lockdep_init_map(&rwlock->dep_map, name, key, 0); ++#endif ++ rwlock->lock.save_state = 1; ++ rwlock->read_depth = 0; ++} ++EXPORT_SYMBOL(__rt_rwlock_init); ++ ++/* ++ * rw_semaphores ++ */ ++ ++void rt_up_write(struct rw_semaphore *rwsem) ++{ ++ rwsem_release(&rwsem->dep_map, 1, _RET_IP_); ++ rt_mutex_unlock(&rwsem->lock); ++} ++EXPORT_SYMBOL(rt_up_write); ++ ++void __rt_up_read(struct rw_semaphore *rwsem) ++{ ++ if (--rwsem->read_depth == 0) ++ rt_mutex_unlock(&rwsem->lock); ++} ++ ++void rt_up_read(struct rw_semaphore *rwsem) ++{ ++ rwsem_release(&rwsem->dep_map, 1, _RET_IP_); ++ __rt_up_read(rwsem); ++} ++EXPORT_SYMBOL(rt_up_read); ++ ++/* ++ * downgrade a write lock into a read lock ++ * - just wake up any readers at the front of the queue ++ */ ++void rt_downgrade_write(struct rw_semaphore *rwsem) ++{ ++ BUG_ON(rt_mutex_owner(&rwsem->lock) != current); ++ rwsem->read_depth = 1; ++} ++EXPORT_SYMBOL(rt_downgrade_write); ++ ++int rt_down_write_trylock(struct rw_semaphore *rwsem) ++{ ++ int ret = rt_mutex_trylock(&rwsem->lock); ++ ++ if (ret) ++ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(rt_down_write_trylock); ++ ++void rt_down_write(struct rw_semaphore *rwsem) ++{ ++ rwsem_acquire(&rwsem->dep_map, 0, 0, _RET_IP_); ++ rt_mutex_lock(&rwsem->lock); ++} ++EXPORT_SYMBOL(rt_down_write); ++ ++void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass) ++{ ++ rwsem_acquire(&rwsem->dep_map, subclass, 0, _RET_IP_); ++ rt_mutex_lock(&rwsem->lock); ++} ++EXPORT_SYMBOL(rt_down_write_nested); ++ ++void rt_down_write_nested_lock(struct rw_semaphore *rwsem, ++ struct lockdep_map *nest) ++{ ++ rwsem_acquire_nest(&rwsem->dep_map, 0, 0, nest, _RET_IP_); ++ rt_mutex_lock(&rwsem->lock); ++} ++EXPORT_SYMBOL(rt_down_write_nested_lock); ++ ++int rt_down_read_trylock(struct rw_semaphore *rwsem) ++{ ++ struct rt_mutex *lock = &rwsem->lock; ++ int ret = 1; ++ ++ /* ++ * recursive read locks succeed when current owns the rwsem, ++ * but not when read_depth == 0 which means that the rwsem is ++ * write locked. ++ */ ++ if (rt_mutex_owner(lock) != current) ++ ret = rt_mutex_trylock(&rwsem->lock); ++ else if (!rwsem->read_depth) ++ ret = 0; ++ ++ if (ret) { ++ rwsem->read_depth++; ++ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_); ++ } ++ return ret; ++} ++EXPORT_SYMBOL(rt_down_read_trylock); ++ ++static void __rt_down_read(struct rw_semaphore *rwsem, int subclass) ++{ ++ struct rt_mutex *lock = &rwsem->lock; ++ ++ rwsem_acquire_read(&rwsem->dep_map, subclass, 0, _RET_IP_); ++ ++ if (rt_mutex_owner(lock) != current) ++ rt_mutex_lock(&rwsem->lock); ++ rwsem->read_depth++; ++} ++ ++void rt_down_read(struct rw_semaphore *rwsem) ++{ ++ __rt_down_read(rwsem, 0); ++} ++EXPORT_SYMBOL(rt_down_read); ++ ++void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass) ++{ ++ __rt_down_read(rwsem, subclass); ++} ++EXPORT_SYMBOL(rt_down_read_nested); ++ ++void __rt_rwsem_init(struct rw_semaphore *rwsem, const char *name, ++ struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)rwsem, sizeof(*rwsem)); ++ lockdep_init_map(&rwsem->dep_map, name, key, 0); ++#endif ++ rwsem->read_depth = 0; ++ rwsem->lock.save_state = 0; ++} ++EXPORT_SYMBOL(__rt_rwsem_init); ++ ++/** ++ * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0 ++ * @cnt: the atomic which we are to dec ++ * @lock: the mutex to return holding if we dec to 0 ++ * ++ * return true and hold lock if we dec to 0, return false otherwise ++ */ ++int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock) ++{ ++ /* dec if we can't possibly hit 0 */ ++ if (atomic_add_unless(cnt, -1, 1)) ++ return 0; ++ /* we might hit 0, so take the lock */ ++ mutex_lock(lock); ++ if (!atomic_dec_and_test(cnt)) { ++ /* when we actually did the dec, we didn't hit 0 */ ++ mutex_unlock(lock); ++ return 0; ++ } ++ /* we hit 0, and we hold the lock */ ++ return 1; ++} ++EXPORT_SYMBOL(atomic_dec_and_mutex_lock); +diff -Nur linux-4.1.13.orig/kernel/locking/rtmutex.c linux-4.1.13/kernel/locking/rtmutex.c +--- linux-4.1.13.orig/kernel/locking/rtmutex.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/rtmutex.c 2015-11-29 09:23:09.637612322 +0100 +@@ -7,6 +7,11 @@ + * Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner <tglx@timesys.com> + * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt + * Copyright (C) 2006 Esben Nielsen ++ * Adaptive Spinlocks: ++ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich, ++ * and Peter Morreale, ++ * Adaptive Spinlocks simplification: ++ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt <srostedt@redhat.com> + * + * See Documentation/locking/rt-mutex-design.txt for details. + */ +@@ -16,6 +21,7 @@ + #include <linux/sched/rt.h> + #include <linux/sched/deadline.h> + #include <linux/timer.h> ++#include <linux/ww_mutex.h> + + #include "rtmutex_common.h" + +@@ -69,6 +75,12 @@ + clear_rt_mutex_waiters(lock); + } + ++static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter) ++{ ++ return waiter && waiter != PI_WAKEUP_INPROGRESS && ++ waiter != PI_REQUEUE_INPROGRESS; ++} ++ + /* + * We can speed up the acquire/release, if the architecture + * supports cmpxchg and if there's no debugging state to be set up +@@ -300,7 +312,7 @@ + * of task. We do not use the spin_xx_mutex() variants here as we are + * outside of the debug path.) + */ +-static void rt_mutex_adjust_prio(struct task_struct *task) ++void rt_mutex_adjust_prio(struct task_struct *task) + { + unsigned long flags; + +@@ -335,6 +347,14 @@ + return debug_rt_mutex_detect_deadlock(waiter, chwalk); + } + ++static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter) ++{ ++ if (waiter->savestate) ++ wake_up_lock_sleeper(waiter->task); ++ else ++ wake_up_process(waiter->task); ++} ++ + /* + * Max number of times we'll walk the boosting chain: + */ +@@ -342,7 +362,8 @@ + + static inline struct rt_mutex *task_blocked_on_lock(struct task_struct *p) + { +- return p->pi_blocked_on ? p->pi_blocked_on->lock : NULL; ++ return rt_mutex_real_waiter(p->pi_blocked_on) ? ++ p->pi_blocked_on->lock : NULL; + } + + /* +@@ -479,7 +500,7 @@ + * reached or the state of the chain has changed while we + * dropped the locks. + */ +- if (!waiter) ++ if (!rt_mutex_real_waiter(waiter)) + goto out_unlock_pi; + + /* +@@ -641,13 +662,16 @@ + * follow here. This is the end of the chain we are walking. + */ + if (!rt_mutex_owner(lock)) { ++ struct rt_mutex_waiter *lock_top_waiter; ++ + /* + * If the requeue [7] above changed the top waiter, + * then we need to wake the new top waiter up to try + * to get the lock. + */ +- if (prerequeue_top_waiter != rt_mutex_top_waiter(lock)) +- wake_up_process(rt_mutex_top_waiter(lock)->task); ++ lock_top_waiter = rt_mutex_top_waiter(lock); ++ if (prerequeue_top_waiter != lock_top_waiter) ++ rt_mutex_wake_waiter(lock_top_waiter); + raw_spin_unlock(&lock->wait_lock); + return 0; + } +@@ -740,6 +764,25 @@ + return ret; + } + ++ ++#define STEAL_NORMAL 0 ++#define STEAL_LATERAL 1 ++ ++/* ++ * Note that RT tasks are excluded from lateral-steals to prevent the ++ * introduction of an unbounded latency ++ */ ++static inline int lock_is_stealable(struct task_struct *task, ++ struct task_struct *pendowner, int mode) ++{ ++ if (mode == STEAL_NORMAL || rt_task(task)) { ++ if (task->prio >= pendowner->prio) ++ return 0; ++ } else if (task->prio > pendowner->prio) ++ return 0; ++ return 1; ++} ++ + /* + * Try to take an rt-mutex + * +@@ -750,8 +793,9 @@ + * @waiter: The waiter that is queued to the lock's wait list if the + * callsite called task_blocked_on_lock(), otherwise NULL + */ +-static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, +- struct rt_mutex_waiter *waiter) ++static int __try_to_take_rt_mutex(struct rt_mutex *lock, ++ struct task_struct *task, ++ struct rt_mutex_waiter *waiter, int mode) + { + unsigned long flags; + +@@ -790,8 +834,10 @@ + * If waiter is not the highest priority waiter of + * @lock, give up. + */ +- if (waiter != rt_mutex_top_waiter(lock)) ++ if (waiter != rt_mutex_top_waiter(lock)) { ++ /* XXX lock_is_stealable() ? */ + return 0; ++ } + + /* + * We can acquire the lock. Remove the waiter from the +@@ -809,14 +855,10 @@ + * not need to be dequeued. + */ + if (rt_mutex_has_waiters(lock)) { +- /* +- * If @task->prio is greater than or equal to +- * the top waiter priority (kernel view), +- * @task lost. +- */ +- if (task->prio >= rt_mutex_top_waiter(lock)->prio) +- return 0; ++ struct task_struct *pown = rt_mutex_top_waiter(lock)->task; + ++ if (task != pown && !lock_is_stealable(task, pown, mode)) ++ return 0; + /* + * The current top waiter stays enqueued. We + * don't have to change anything in the lock +@@ -865,6 +907,347 @@ + return 1; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * preemptible spin_lock functions: ++ */ ++static inline void rt_spin_lock_fastlock(struct rt_mutex *lock, ++ void (*slowfn)(struct rt_mutex *lock)) ++{ ++ might_sleep_no_state_check(); ++ ++ if (likely(rt_mutex_cmpxchg(lock, NULL, current))) ++ rt_mutex_deadlock_account_lock(lock, current); ++ else ++ slowfn(lock); ++} ++ ++static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock, ++ void (*slowfn)(struct rt_mutex *lock)) ++{ ++ if (likely(rt_mutex_cmpxchg(lock, current, NULL))) ++ rt_mutex_deadlock_account_unlock(current); ++ else ++ slowfn(lock); ++} ++#ifdef CONFIG_SMP ++/* ++ * Note that owner is a speculative pointer and dereferencing relies ++ * on rcu_read_lock() and the check against the lock owner. ++ */ ++static int adaptive_wait(struct rt_mutex *lock, ++ struct task_struct *owner) ++{ ++ int res = 0; ++ ++ rcu_read_lock(); ++ for (;;) { ++ if (owner != rt_mutex_owner(lock)) ++ break; ++ /* ++ * Ensure that owner->on_cpu is dereferenced _after_ ++ * checking the above to be valid. ++ */ ++ barrier(); ++ if (!owner->on_cpu) { ++ res = 1; ++ break; ++ } ++ cpu_relax(); ++ } ++ rcu_read_unlock(); ++ return res; ++} ++#else ++static int adaptive_wait(struct rt_mutex *lock, ++ struct task_struct *orig_owner) ++{ ++ return 1; ++} ++#endif ++ ++# define pi_lock(lock) raw_spin_lock_irq(lock) ++# define pi_unlock(lock) raw_spin_unlock_irq(lock) ++ ++static int task_blocks_on_rt_mutex(struct rt_mutex *lock, ++ struct rt_mutex_waiter *waiter, ++ struct task_struct *task, ++ enum rtmutex_chainwalk chwalk); ++/* ++ * Slow path lock function spin_lock style: this variant is very ++ * careful not to miss any non-lock wakeups. ++ * ++ * We store the current state under p->pi_lock in p->saved_state and ++ * the try_to_wake_up() code handles this accordingly. ++ */ ++static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock) ++{ ++ struct task_struct *lock_owner, *self = current; ++ struct rt_mutex_waiter waiter, *top_waiter; ++ int ret; ++ ++ rt_mutex_init_waiter(&waiter, true); ++ ++ raw_spin_lock(&lock->wait_lock); ++ ++ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) { ++ raw_spin_unlock(&lock->wait_lock); ++ return; ++ } ++ ++ BUG_ON(rt_mutex_owner(lock) == self); ++ ++ /* ++ * We save whatever state the task is in and we'll restore it ++ * after acquiring the lock taking real wakeups into account ++ * as well. We are serialized via pi_lock against wakeups. See ++ * try_to_wake_up(). ++ */ ++ pi_lock(&self->pi_lock); ++ self->saved_state = self->state; ++ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); ++ pi_unlock(&self->pi_lock); ++ ++ ret = task_blocks_on_rt_mutex(lock, &waiter, self, 0); ++ BUG_ON(ret); ++ ++ for (;;) { ++ /* Try to acquire the lock again. */ ++ if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL)) ++ break; ++ ++ top_waiter = rt_mutex_top_waiter(lock); ++ lock_owner = rt_mutex_owner(lock); ++ ++ raw_spin_unlock(&lock->wait_lock); ++ ++ debug_rt_mutex_print_deadlock(&waiter); ++ ++ if (top_waiter != &waiter || adaptive_wait(lock, lock_owner)) ++ schedule_rt_mutex(lock); ++ ++ raw_spin_lock(&lock->wait_lock); ++ ++ pi_lock(&self->pi_lock); ++ __set_current_state_no_track(TASK_UNINTERRUPTIBLE); ++ pi_unlock(&self->pi_lock); ++ } ++ ++ /* ++ * Restore the task state to current->saved_state. We set it ++ * to the original state above and the try_to_wake_up() code ++ * has possibly updated it when a real (non-rtmutex) wakeup ++ * happened while we were blocked. Clear saved_state so ++ * try_to_wakeup() does not get confused. ++ */ ++ pi_lock(&self->pi_lock); ++ __set_current_state_no_track(self->saved_state); ++ self->saved_state = TASK_RUNNING; ++ pi_unlock(&self->pi_lock); ++ ++ /* ++ * try_to_take_rt_mutex() sets the waiter bit ++ * unconditionally. We might have to fix that up: ++ */ ++ fixup_rt_mutex_waiters(lock); ++ ++ BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock)); ++ BUG_ON(!RB_EMPTY_NODE(&waiter.tree_entry)); ++ ++ raw_spin_unlock(&lock->wait_lock); ++ ++ debug_rt_mutex_free_waiter(&waiter); ++} ++ ++static void wakeup_next_waiter(struct rt_mutex *lock); ++/* ++ * Slow path to release a rt_mutex spin_lock style ++ */ ++static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock) ++{ ++ raw_spin_lock(&lock->wait_lock); ++ ++ debug_rt_mutex_unlock(lock); ++ ++ rt_mutex_deadlock_account_unlock(current); ++ ++ if (!rt_mutex_has_waiters(lock)) { ++ lock->owner = NULL; ++ raw_spin_unlock(&lock->wait_lock); ++ return; ++ } ++ ++ wakeup_next_waiter(lock); ++ ++ raw_spin_unlock(&lock->wait_lock); ++ ++ /* Undo pi boosting.when necessary */ ++ rt_mutex_adjust_prio(current); ++} ++ ++void __lockfunc rt_spin_lock(spinlock_t *lock) ++{ ++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock); ++ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_); ++} ++EXPORT_SYMBOL(rt_spin_lock); ++ ++void __lockfunc __rt_spin_lock(struct rt_mutex *lock) ++{ ++ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock); ++} ++EXPORT_SYMBOL(__rt_spin_lock); ++ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass) ++{ ++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock); ++ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_); ++} ++EXPORT_SYMBOL(rt_spin_lock_nested); ++#endif ++ ++void __lockfunc rt_spin_unlock(spinlock_t *lock) ++{ ++ /* NOTE: we always pass in '1' for nested, for simplicity */ ++ spin_release(&lock->dep_map, 1, _RET_IP_); ++ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock); ++} ++EXPORT_SYMBOL(rt_spin_unlock); ++ ++void __lockfunc __rt_spin_unlock(struct rt_mutex *lock) ++{ ++ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock); ++} ++EXPORT_SYMBOL(__rt_spin_unlock); ++ ++/* ++ * Wait for the lock to get unlocked: instead of polling for an unlock ++ * (like raw spinlocks do), we lock and unlock, to force the kernel to ++ * schedule if there's contention: ++ */ ++void __lockfunc rt_spin_unlock_wait(spinlock_t *lock) ++{ ++ spin_lock(lock); ++ spin_unlock(lock); ++} ++EXPORT_SYMBOL(rt_spin_unlock_wait); ++ ++int __lockfunc __rt_spin_trylock(struct rt_mutex *lock) ++{ ++ return rt_mutex_trylock(lock); ++} ++ ++int __lockfunc rt_spin_trylock(spinlock_t *lock) ++{ ++ int ret = rt_mutex_trylock(&lock->lock); ++ ++ if (ret) ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock); ++ ++int __lockfunc rt_spin_trylock_bh(spinlock_t *lock) ++{ ++ int ret; ++ ++ local_bh_disable(); ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) { ++ migrate_disable(); ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ } else ++ local_bh_enable(); ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock_bh); ++ ++int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags) ++{ ++ int ret; ++ ++ *flags = 0; ++ ret = rt_mutex_trylock(&lock->lock); ++ if (ret) { ++ migrate_disable(); ++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_); ++ } ++ return ret; ++} ++EXPORT_SYMBOL(rt_spin_trylock_irqsave); ++ ++int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock) ++{ ++ /* Subtract 1 from counter unless that drops it to 0 (ie. it was 1) */ ++ if (atomic_add_unless(atomic, -1, 1)) ++ return 0; ++ migrate_disable(); ++ rt_spin_lock(lock); ++ if (atomic_dec_and_test(atomic)) ++ return 1; ++ rt_spin_unlock(lock); ++ migrate_enable(); ++ return 0; ++} ++EXPORT_SYMBOL(atomic_dec_and_spin_lock); ++ ++ void ++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key) ++{ ++#ifdef CONFIG_DEBUG_LOCK_ALLOC ++ /* ++ * Make sure we are not reinitializing a held lock: ++ */ ++ debug_check_no_locks_freed((void *)lock, sizeof(*lock)); ++ lockdep_init_map(&lock->dep_map, name, key, 0); ++#endif ++} ++EXPORT_SYMBOL(__rt_spin_lock_init); ++ ++#endif /* PREEMPT_RT_FULL */ ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ static inline int __sched ++__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); ++ struct ww_acquire_ctx *hold_ctx = ACCESS_ONCE(ww->ctx); ++ ++ if (!hold_ctx) ++ return 0; ++ ++ if (unlikely(ctx == hold_ctx)) ++ return -EALREADY; ++ ++ if (ctx->stamp - hold_ctx->stamp <= LONG_MAX && ++ (ctx->stamp != hold_ctx->stamp || ctx > hold_ctx)) { ++#ifdef CONFIG_DEBUG_MUTEXES ++ DEBUG_LOCKS_WARN_ON(ctx->contending_lock); ++ ctx->contending_lock = ww; ++#endif ++ return -EDEADLK; ++ } ++ ++ return 0; ++} ++#else ++ static inline int __sched ++__mutex_lock_check_stamp(struct rt_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++ BUG(); ++ return 0; ++} ++ ++#endif ++ ++static inline int ++try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task, ++ struct rt_mutex_waiter *waiter) ++{ ++ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL); ++} ++ + /* + * Task blocks on lock. + * +@@ -896,6 +1279,23 @@ + return -EDEADLK; + + raw_spin_lock_irqsave(&task->pi_lock, flags); ++ ++ /* ++ * In the case of futex requeue PI, this will be a proxy ++ * lock. The task will wake unaware that it is enqueueed on ++ * this lock. Avoid blocking on two locks and corrupting ++ * pi_blocked_on via the PI_WAKEUP_INPROGRESS ++ * flag. futex_wait_requeue_pi() sets this when it wakes up ++ * before requeue (due to a signal or timeout). Do not enqueue ++ * the task if PI_WAKEUP_INPROGRESS is set. ++ */ ++ if (task != current && task->pi_blocked_on == PI_WAKEUP_INPROGRESS) { ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); ++ return -EAGAIN; ++ } ++ ++ BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)); ++ + __rt_mutex_adjust_prio(task); + waiter->task = task; + waiter->lock = lock; +@@ -919,7 +1319,7 @@ + rt_mutex_enqueue_pi(owner, waiter); + + __rt_mutex_adjust_prio(owner); +- if (owner->pi_blocked_on) ++ if (rt_mutex_real_waiter(owner->pi_blocked_on)) + chain_walk = 1; + } else if (rt_mutex_cond_detect_deadlock(waiter, chwalk)) { + chain_walk = 1; +@@ -957,8 +1357,9 @@ + /* + * Wake up the next waiter on the lock. + * +- * Remove the top waiter from the current tasks pi waiter list and +- * wake it up. ++ * Remove the top waiter from the current tasks pi waiter list, ++ * wake it up and return whether the current task needs to undo ++ * a potential priority boosting. + * + * Called with lock->wait_lock held. + */ +@@ -996,7 +1397,7 @@ + * long as we hold lock->wait_lock. The waiter task needs to + * acquire it in order to dequeue the waiter. + */ +- wake_up_process(waiter->task); ++ rt_mutex_wake_waiter(waiter); + } + + /* +@@ -1010,7 +1411,7 @@ + { + bool is_top_waiter = (waiter == rt_mutex_top_waiter(lock)); + struct task_struct *owner = rt_mutex_owner(lock); +- struct rt_mutex *next_lock; ++ struct rt_mutex *next_lock = NULL; + unsigned long flags; + + raw_spin_lock_irqsave(¤t->pi_lock, flags); +@@ -1035,7 +1436,8 @@ + __rt_mutex_adjust_prio(owner); + + /* Store the lock on which owner is blocked or NULL */ +- next_lock = task_blocked_on_lock(owner); ++ if (rt_mutex_real_waiter(owner->pi_blocked_on)) ++ next_lock = task_blocked_on_lock(owner); + + raw_spin_unlock_irqrestore(&owner->pi_lock, flags); + +@@ -1071,17 +1473,17 @@ + raw_spin_lock_irqsave(&task->pi_lock, flags); + + waiter = task->pi_blocked_on; +- if (!waiter || (waiter->prio == task->prio && ++ if (!rt_mutex_real_waiter(waiter) || (waiter->prio == task->prio && + !dl_prio(task->prio))) { + raw_spin_unlock_irqrestore(&task->pi_lock, flags); + return; + } + next_lock = waiter->lock; +- raw_spin_unlock_irqrestore(&task->pi_lock, flags); + + /* gets dropped in rt_mutex_adjust_prio_chain()! */ + get_task_struct(task); + ++ raw_spin_unlock_irqrestore(&task->pi_lock, flags); + rt_mutex_adjust_prio_chain(task, RT_MUTEX_MIN_CHAINWALK, NULL, + next_lock, NULL, task); + } +@@ -1099,7 +1501,8 @@ + static int __sched + __rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- struct rt_mutex_waiter *waiter) ++ struct rt_mutex_waiter *waiter, ++ struct ww_acquire_ctx *ww_ctx) + { + int ret = 0; + +@@ -1122,6 +1525,12 @@ + break; + } + ++ if (ww_ctx && ww_ctx->acquired > 0) { ++ ret = __mutex_lock_check_stamp(lock, ww_ctx); ++ if (ret) ++ break; ++ } ++ + raw_spin_unlock(&lock->wait_lock); + + debug_rt_mutex_print_deadlock(waiter); +@@ -1156,25 +1565,102 @@ + } + } + ++static __always_inline void ww_mutex_lock_acquired(struct ww_mutex *ww, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++#ifdef CONFIG_DEBUG_MUTEXES ++ /* ++ * If this WARN_ON triggers, you used ww_mutex_lock to acquire, ++ * but released with a normal mutex_unlock in this call. ++ * ++ * This should never happen, always use ww_mutex_unlock. ++ */ ++ DEBUG_LOCKS_WARN_ON(ww->ctx); ++ ++ /* ++ * Not quite done after calling ww_acquire_done() ? ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->done_acquire); ++ ++ if (ww_ctx->contending_lock) { ++ /* ++ * After -EDEADLK you tried to ++ * acquire a different ww_mutex? Bad! ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->contending_lock != ww); ++ ++ /* ++ * You called ww_mutex_lock after receiving -EDEADLK, ++ * but 'forgot' to unlock everything else first? ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->acquired > 0); ++ ww_ctx->contending_lock = NULL; ++ } ++ ++ /* ++ * Naughty, using a different class will lead to undefined behavior! ++ */ ++ DEBUG_LOCKS_WARN_ON(ww_ctx->ww_class != ww->ww_class); ++#endif ++ ww_ctx->acquired++; ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void ww_mutex_account_lock(struct rt_mutex *lock, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++ struct ww_mutex *ww = container_of(lock, struct ww_mutex, base.lock); ++ struct rt_mutex_waiter *waiter, *n; ++ ++ /* ++ * This branch gets optimized out for the common case, ++ * and is only important for ww_mutex_lock. ++ */ ++ ww_mutex_lock_acquired(ww, ww_ctx); ++ ww->ctx = ww_ctx; ++ ++ /* ++ * Give any possible sleeping processes the chance to wake up, ++ * so they can recheck if they have to back off. ++ */ ++ rbtree_postorder_for_each_entry_safe(waiter, n, &lock->waiters, ++ tree_entry) { ++ /* XXX debug rt mutex waiter wakeup */ ++ ++ BUG_ON(waiter->lock != lock); ++ rt_mutex_wake_waiter(waiter); ++ } ++} ++ ++#else ++ ++static void ww_mutex_account_lock(struct rt_mutex *lock, ++ struct ww_acquire_ctx *ww_ctx) ++{ ++ BUG(); ++} ++#endif ++ + /* + * Slow path lock function: + */ + static int __sched + rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx) + { + struct rt_mutex_waiter waiter; + int ret = 0; + +- debug_rt_mutex_init_waiter(&waiter); +- RB_CLEAR_NODE(&waiter.pi_tree_entry); +- RB_CLEAR_NODE(&waiter.tree_entry); ++ rt_mutex_init_waiter(&waiter, false); + + raw_spin_lock(&lock->wait_lock); + + /* Try to acquire the lock again: */ + if (try_to_take_rt_mutex(lock, current, NULL)) { ++ if (ww_ctx) ++ ww_mutex_account_lock(lock, ww_ctx); + raw_spin_unlock(&lock->wait_lock); + return 0; + } +@@ -1192,13 +1678,23 @@ + + if (likely(!ret)) + /* sleep on the mutex */ +- ret = __rt_mutex_slowlock(lock, state, timeout, &waiter); ++ ret = __rt_mutex_slowlock(lock, state, timeout, &waiter, ++ ww_ctx); ++ else if (ww_ctx) { ++ /* ww_mutex received EDEADLK, let it become EALREADY */ ++ ret = __mutex_lock_check_stamp(lock, ww_ctx); ++ BUG_ON(!ret); ++ } + + if (unlikely(ret)) { + __set_current_state(TASK_RUNNING); + if (rt_mutex_has_waiters(lock)) + remove_waiter(lock, &waiter); +- rt_mutex_handle_deadlock(ret, chwalk, &waiter); ++ /* ww_mutex want to report EDEADLK/EALREADY, let them */ ++ if (!ww_ctx) ++ rt_mutex_handle_deadlock(ret, chwalk, &waiter); ++ } else if (ww_ctx) { ++ ww_mutex_account_lock(lock, ww_ctx); + } + + /* +@@ -1255,7 +1751,7 @@ + /* + * Slow path to release a rt-mutex: + */ +-static void __sched ++static bool __sched + rt_mutex_slowunlock(struct rt_mutex *lock) + { + raw_spin_lock(&lock->wait_lock); +@@ -1298,7 +1794,7 @@ + while (!rt_mutex_has_waiters(lock)) { + /* Drops lock->wait_lock ! */ + if (unlock_rt_mutex_safe(lock) == true) +- return; ++ return false; + /* Relock the rtmutex and try again */ + raw_spin_lock(&lock->wait_lock); + } +@@ -1311,8 +1807,7 @@ + + raw_spin_unlock(&lock->wait_lock); + +- /* Undo pi boosting if necessary: */ +- rt_mutex_adjust_prio(current); ++ return true; + } + + /* +@@ -1323,31 +1818,36 @@ + */ + static inline int + rt_mutex_fastlock(struct rt_mutex *lock, int state, ++ struct ww_acquire_ctx *ww_ctx, + int (*slowfn)(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk)) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx)) + { + if (likely(rt_mutex_cmpxchg(lock, NULL, current))) { + rt_mutex_deadlock_account_lock(lock, current); + return 0; + } else +- return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK); ++ return slowfn(lock, state, NULL, RT_MUTEX_MIN_CHAINWALK, ++ ww_ctx); + } + + static inline int + rt_mutex_timed_fastlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, + enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx, + int (*slowfn)(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, +- enum rtmutex_chainwalk chwalk)) ++ enum rtmutex_chainwalk chwalk, ++ struct ww_acquire_ctx *ww_ctx)) + { + if (chwalk == RT_MUTEX_MIN_CHAINWALK && + likely(rt_mutex_cmpxchg(lock, NULL, current))) { + rt_mutex_deadlock_account_lock(lock, current); + return 0; + } else +- return slowfn(lock, state, timeout, chwalk); ++ return slowfn(lock, state, timeout, chwalk, ww_ctx); + } + + static inline int +@@ -1363,12 +1863,14 @@ + + static inline void + rt_mutex_fastunlock(struct rt_mutex *lock, +- void (*slowfn)(struct rt_mutex *lock)) ++ bool (*slowfn)(struct rt_mutex *lock)) + { +- if (likely(rt_mutex_cmpxchg(lock, current, NULL))) ++ if (likely(rt_mutex_cmpxchg(lock, current, NULL))) { + rt_mutex_deadlock_account_unlock(current); +- else +- slowfn(lock); ++ } else if (slowfn(lock)) { ++ /* Undo pi boosting if necessary: */ ++ rt_mutex_adjust_prio(current); ++ } + } + + /** +@@ -1380,7 +1882,7 @@ + { + might_sleep(); + +- rt_mutex_fastlock(lock, TASK_UNINTERRUPTIBLE, rt_mutex_slowlock); ++ rt_mutex_fastlock(lock, TASK_UNINTERRUPTIBLE, NULL, rt_mutex_slowlock); + } + EXPORT_SYMBOL_GPL(rt_mutex_lock); + +@@ -1397,7 +1899,7 @@ + { + might_sleep(); + +- return rt_mutex_fastlock(lock, TASK_INTERRUPTIBLE, rt_mutex_slowlock); ++ return rt_mutex_fastlock(lock, TASK_INTERRUPTIBLE, NULL, rt_mutex_slowlock); + } + EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible); + +@@ -1410,11 +1912,30 @@ + might_sleep(); + + return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, +- RT_MUTEX_FULL_CHAINWALK, ++ RT_MUTEX_FULL_CHAINWALK, NULL, + rt_mutex_slowlock); + } + + /** ++ * rt_mutex_lock_killable - lock a rt_mutex killable ++ * ++ * @lock: the rt_mutex to be locked ++ * @detect_deadlock: deadlock detection on/off ++ * ++ * Returns: ++ * 0 on success ++ * -EINTR when interrupted by a signal ++ * -EDEADLK when the lock would deadlock (when deadlock detection is on) ++ */ ++int __sched rt_mutex_lock_killable(struct rt_mutex *lock) ++{ ++ might_sleep(); ++ ++ return rt_mutex_fastlock(lock, TASK_KILLABLE, NULL, rt_mutex_slowlock); ++} ++EXPORT_SYMBOL_GPL(rt_mutex_lock_killable); ++ ++/** + * rt_mutex_timed_lock - lock a rt_mutex interruptible + * the timeout structure is provided + * by the caller +@@ -1434,6 +1955,7 @@ + + return rt_mutex_timed_fastlock(lock, TASK_INTERRUPTIBLE, timeout, + RT_MUTEX_MIN_CHAINWALK, ++ NULL, + rt_mutex_slowlock); + } + EXPORT_SYMBOL_GPL(rt_mutex_timed_lock); +@@ -1463,6 +1985,22 @@ + EXPORT_SYMBOL_GPL(rt_mutex_unlock); + + /** ++ * rt_mutex_futex_unlock - Futex variant of rt_mutex_unlock ++ * @lock: the rt_mutex to be unlocked ++ * ++ * Returns: true/false indicating whether priority adjustment is ++ * required or not. ++ */ ++bool __sched rt_mutex_futex_unlock(struct rt_mutex *lock) ++{ ++ if (likely(rt_mutex_cmpxchg(lock, current, NULL))) { ++ rt_mutex_deadlock_account_unlock(current); ++ return false; ++ } ++ return rt_mutex_slowunlock(lock); ++} ++ ++/** + * rt_mutex_destroy - mark a mutex unusable + * @lock: the mutex to be destroyed + * +@@ -1492,13 +2030,12 @@ + void __rt_mutex_init(struct rt_mutex *lock, const char *name) + { + lock->owner = NULL; +- raw_spin_lock_init(&lock->wait_lock); + lock->waiters = RB_ROOT; + lock->waiters_leftmost = NULL; + + debug_rt_mutex_init(lock, name); + } +-EXPORT_SYMBOL_GPL(__rt_mutex_init); ++EXPORT_SYMBOL(__rt_mutex_init); + + /** + * rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a +@@ -1513,7 +2050,7 @@ + void rt_mutex_init_proxy_locked(struct rt_mutex *lock, + struct task_struct *proxy_owner) + { +- __rt_mutex_init(lock, NULL); ++ rt_mutex_init(lock); + debug_rt_mutex_proxy_lock(lock, proxy_owner); + rt_mutex_set_owner(lock, proxy_owner); + rt_mutex_deadlock_account_lock(lock, proxy_owner); +@@ -1561,6 +2098,35 @@ + return 1; + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * In PREEMPT_RT there's an added race. ++ * If the task, that we are about to requeue, times out, ++ * it can set the PI_WAKEUP_INPROGRESS. This tells the requeue ++ * to skip this task. But right after the task sets ++ * its pi_blocked_on to PI_WAKEUP_INPROGRESS it can then ++ * block on the spin_lock(&hb->lock), which in RT is an rtmutex. ++ * This will replace the PI_WAKEUP_INPROGRESS with the actual ++ * lock that it blocks on. We *must not* place this task ++ * on this proxy lock in that case. ++ * ++ * To prevent this race, we first take the task's pi_lock ++ * and check if it has updated its pi_blocked_on. If it has, ++ * we assume that it woke up and we return -EAGAIN. ++ * Otherwise, we set the task's pi_blocked_on to ++ * PI_REQUEUE_INPROGRESS, so that if the task is waking up ++ * it will know that we are in the process of requeuing it. ++ */ ++ raw_spin_lock_irq(&task->pi_lock); ++ if (task->pi_blocked_on) { ++ raw_spin_unlock_irq(&task->pi_lock); ++ raw_spin_unlock(&lock->wait_lock); ++ return -EAGAIN; ++ } ++ task->pi_blocked_on = PI_REQUEUE_INPROGRESS; ++ raw_spin_unlock_irq(&task->pi_lock); ++#endif ++ + /* We enforce deadlock detection for futexes */ + ret = task_blocks_on_rt_mutex(lock, waiter, task, + RT_MUTEX_FULL_CHAINWALK); +@@ -1575,7 +2141,7 @@ + ret = 0; + } + +- if (unlikely(ret)) ++ if (ret && rt_mutex_has_waiters(lock)) + remove_waiter(lock, waiter); + + raw_spin_unlock(&lock->wait_lock); +@@ -1631,7 +2197,7 @@ + set_current_state(TASK_INTERRUPTIBLE); + + /* sleep on the mutex */ +- ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter); ++ ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter, NULL); + + if (unlikely(ret)) + remove_waiter(lock, waiter); +@@ -1646,3 +2212,89 @@ + + return ret; + } ++ ++static inline int ++ww_mutex_deadlock_injection(struct ww_mutex *lock, struct ww_acquire_ctx *ctx) ++{ ++#ifdef CONFIG_DEBUG_WW_MUTEX_SLOWPATH ++ unsigned tmp; ++ ++ if (ctx->deadlock_inject_countdown-- == 0) { ++ tmp = ctx->deadlock_inject_interval; ++ if (tmp > UINT_MAX/4) ++ tmp = UINT_MAX; ++ else ++ tmp = tmp*2 + tmp + tmp/2; ++ ++ ctx->deadlock_inject_interval = tmp; ++ ctx->deadlock_inject_countdown = tmp; ++ ctx->contending_lock = lock; ++ ++ ww_mutex_unlock(lock); ++ ++ return -EDEADLK; ++ } ++#endif ++ ++ return 0; ++} ++ ++#ifdef CONFIG_PREEMPT_RT_FULL ++int __sched ++__ww_mutex_lock_interruptible(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) ++{ ++ int ret; ++ ++ might_sleep(); ++ ++ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); ++ ret = rt_mutex_slowlock(&lock->base.lock, TASK_INTERRUPTIBLE, NULL, 0, ww_ctx); ++ if (ret) ++ mutex_release(&lock->base.dep_map, 1, _RET_IP_); ++ else if (!ret && ww_ctx->acquired > 1) ++ return ww_mutex_deadlock_injection(lock, ww_ctx); ++ ++ return ret; ++} ++EXPORT_SYMBOL_GPL(__ww_mutex_lock_interruptible); ++ ++int __sched ++__ww_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx) ++{ ++ int ret; ++ ++ might_sleep(); ++ ++ mutex_acquire_nest(&lock->base.dep_map, 0, 0, &ww_ctx->dep_map, _RET_IP_); ++ ret = rt_mutex_slowlock(&lock->base.lock, TASK_UNINTERRUPTIBLE, NULL, 0, ww_ctx); ++ if (ret) ++ mutex_release(&lock->base.dep_map, 1, _RET_IP_); ++ else if (!ret && ww_ctx->acquired > 1) ++ return ww_mutex_deadlock_injection(lock, ww_ctx); ++ ++ return ret; ++} ++EXPORT_SYMBOL_GPL(__ww_mutex_lock); ++ ++void __sched ww_mutex_unlock(struct ww_mutex *lock) ++{ ++ int nest = !!lock->ctx; ++ ++ /* ++ * The unlocking fastpath is the 0->1 transition from 'locked' ++ * into 'unlocked' state: ++ */ ++ if (nest) { ++#ifdef CONFIG_DEBUG_MUTEXES ++ DEBUG_LOCKS_WARN_ON(!lock->ctx->acquired); ++#endif ++ if (lock->ctx->acquired > 0) ++ lock->ctx->acquired--; ++ lock->ctx = NULL; ++ } ++ ++ mutex_release(&lock->base.dep_map, nest, _RET_IP_); ++ rt_mutex_unlock(&lock->base.lock); ++} ++EXPORT_SYMBOL(ww_mutex_unlock); ++#endif +diff -Nur linux-4.1.13.orig/kernel/locking/rtmutex_common.h linux-4.1.13/kernel/locking/rtmutex_common.h +--- linux-4.1.13.orig/kernel/locking/rtmutex_common.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/rtmutex_common.h 2015-11-29 09:23:09.637612322 +0100 +@@ -49,6 +49,7 @@ + struct rb_node pi_tree_entry; + struct task_struct *task; + struct rt_mutex *lock; ++ bool savestate; + #ifdef CONFIG_DEBUG_RT_MUTEXES + unsigned long ip; + struct pid *deadlock_task_pid; +@@ -119,6 +120,9 @@ + /* + * PI-futex support (proxy locking functions, etc.): + */ ++#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1) ++#define PI_REQUEUE_INPROGRESS ((struct rt_mutex_waiter *) 2) ++ + extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock); + extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock, + struct task_struct *proxy_owner); +@@ -132,10 +136,24 @@ + struct rt_mutex_waiter *waiter); + extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to); + ++extern bool rt_mutex_futex_unlock(struct rt_mutex *lock); ++ ++extern void rt_mutex_adjust_prio(struct task_struct *task); ++ + #ifdef CONFIG_DEBUG_RT_MUTEXES + # include "rtmutex-debug.h" + #else + # include "rtmutex.h" + #endif + ++static inline void ++rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate) ++{ ++ debug_rt_mutex_init_waiter(waiter); ++ waiter->task = NULL; ++ waiter->savestate = savestate; ++ RB_CLEAR_NODE(&waiter->pi_tree_entry); ++ RB_CLEAR_NODE(&waiter->tree_entry); ++} ++ + #endif +diff -Nur linux-4.1.13.orig/kernel/locking/spinlock.c linux-4.1.13/kernel/locking/spinlock.c +--- linux-4.1.13.orig/kernel/locking/spinlock.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/spinlock.c 2015-11-29 09:23:09.637612322 +0100 +@@ -124,8 +124,11 @@ + * __[spin|read|write]_lock_bh() + */ + BUILD_LOCK_OPS(spin, raw_spinlock); ++ ++#ifndef CONFIG_PREEMPT_RT_FULL + BUILD_LOCK_OPS(read, rwlock); + BUILD_LOCK_OPS(write, rwlock); ++#endif + + #endif + +@@ -209,6 +212,8 @@ + EXPORT_SYMBOL(_raw_spin_unlock_bh); + #endif + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #ifndef CONFIG_INLINE_READ_TRYLOCK + int __lockfunc _raw_read_trylock(rwlock_t *lock) + { +@@ -353,6 +358,8 @@ + EXPORT_SYMBOL(_raw_write_unlock_bh); + #endif + ++#endif /* !PREEMPT_RT_FULL */ ++ + #ifdef CONFIG_DEBUG_LOCK_ALLOC + + void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass) +diff -Nur linux-4.1.13.orig/kernel/locking/spinlock_debug.c linux-4.1.13/kernel/locking/spinlock_debug.c +--- linux-4.1.13.orig/kernel/locking/spinlock_debug.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/locking/spinlock_debug.c 2015-11-29 09:23:09.637612322 +0100 +@@ -31,6 +31,7 @@ + + EXPORT_SYMBOL(__raw_spin_lock_init); + ++#ifndef CONFIG_PREEMPT_RT_FULL + void __rwlock_init(rwlock_t *lock, const char *name, + struct lock_class_key *key) + { +@@ -48,6 +49,7 @@ + } + + EXPORT_SYMBOL(__rwlock_init); ++#endif + + static void spin_dump(raw_spinlock_t *lock, const char *msg) + { +@@ -159,6 +161,7 @@ + arch_spin_unlock(&lock->raw_lock); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + static void rwlock_bug(rwlock_t *lock, const char *msg) + { + if (!debug_locks_off()) +@@ -300,3 +303,5 @@ + debug_write_unlock(lock); + arch_write_unlock(&lock->raw_lock); + } ++ ++#endif +diff -Nur linux-4.1.13.orig/kernel/panic.c linux-4.1.13/kernel/panic.c +--- linux-4.1.13.orig/kernel/panic.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/panic.c 2015-11-29 09:23:09.637612322 +0100 +@@ -387,9 +387,11 @@ + + static int init_oops_id(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + if (!oops_id) + get_random_bytes(&oops_id, sizeof(oops_id)); + else ++#endif + oops_id++; + + return 0; +diff -Nur linux-4.1.13.orig/kernel/power/hibernate.c linux-4.1.13/kernel/power/hibernate.c +--- linux-4.1.13.orig/kernel/power/hibernate.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/power/hibernate.c 2015-11-29 09:23:09.637612322 +0100 +@@ -285,6 +285,8 @@ + + local_irq_disable(); + ++ system_state = SYSTEM_SUSPEND; ++ + error = syscore_suspend(); + if (error) { + printk(KERN_ERR "PM: Some system devices failed to power down, " +@@ -314,6 +316,7 @@ + syscore_resume(); + + Enable_irqs: ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + + Enable_cpus: +@@ -437,6 +440,7 @@ + goto Enable_cpus; + + local_irq_disable(); ++ system_state = SYSTEM_SUSPEND; + + error = syscore_suspend(); + if (error) +@@ -470,6 +474,7 @@ + syscore_resume(); + + Enable_irqs: ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + + Enable_cpus: +@@ -555,6 +560,7 @@ + goto Platform_finish; + + local_irq_disable(); ++ system_state = SYSTEM_SUSPEND; + syscore_suspend(); + if (pm_wakeup_pending()) { + error = -EAGAIN; +@@ -567,6 +573,7 @@ + + Power_up: + syscore_resume(); ++ system_state = SYSTEM_RUNNING; + local_irq_enable(); + enable_nonboot_cpus(); + +diff -Nur linux-4.1.13.orig/kernel/power/suspend.c linux-4.1.13/kernel/power/suspend.c +--- linux-4.1.13.orig/kernel/power/suspend.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/power/suspend.c 2015-11-29 09:23:09.637612322 +0100 +@@ -356,6 +356,8 @@ + arch_suspend_disable_irqs(); + BUG_ON(!irqs_disabled()); + ++ system_state = SYSTEM_SUSPEND; ++ + error = syscore_suspend(); + if (!error) { + *wakeup = pm_wakeup_pending(); +@@ -370,6 +372,8 @@ + syscore_resume(); + } + ++ system_state = SYSTEM_RUNNING; ++ + arch_suspend_enable_irqs(); + BUG_ON(irqs_disabled()); + +diff -Nur linux-4.1.13.orig/kernel/printk/printk.c linux-4.1.13/kernel/printk/printk.c +--- linux-4.1.13.orig/kernel/printk/printk.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/printk/printk.c 2015-11-29 09:23:09.641612055 +0100 +@@ -1163,6 +1163,7 @@ + { + char *text; + int len = 0; ++ int attempts = 0; + + text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL); + if (!text) +@@ -1174,7 +1175,14 @@ + u64 seq; + u32 idx; + enum log_flags prev; +- ++ int num_msg; ++try_again: ++ attempts++; ++ if (attempts > 10) { ++ len = -EBUSY; ++ goto out; ++ } ++ num_msg = 0; + if (clear_seq < log_first_seq) { + /* messages are gone, move to first available one */ + clear_seq = log_first_seq; +@@ -1195,6 +1203,14 @@ + prev = msg->flags; + idx = log_next(idx); + seq++; ++ num_msg++; ++ if (num_msg > 5) { ++ num_msg = 0; ++ raw_spin_unlock_irq(&logbuf_lock); ++ raw_spin_lock_irq(&logbuf_lock); ++ if (clear_seq < log_first_seq) ++ goto try_again; ++ } + } + + /* move first record forward until length fits into the buffer */ +@@ -1208,6 +1224,14 @@ + prev = msg->flags; + idx = log_next(idx); + seq++; ++ num_msg++; ++ if (num_msg > 5) { ++ num_msg = 0; ++ raw_spin_unlock_irq(&logbuf_lock); ++ raw_spin_lock_irq(&logbuf_lock); ++ if (clear_seq < log_first_seq) ++ goto try_again; ++ } + } + + /* last message fitting into this dump */ +@@ -1248,6 +1272,7 @@ + clear_seq = log_next_seq; + clear_idx = log_next_idx; + } ++out: + raw_spin_unlock_irq(&logbuf_lock); + + kfree(text); +@@ -1401,6 +1426,7 @@ + if (!console_drivers) + return; + ++ migrate_disable(); + for_each_console(con) { + if (exclusive_console && con != exclusive_console) + continue; +@@ -1413,6 +1439,7 @@ + continue; + con->write(con, text, len); + } ++ migrate_enable(); + } + + /* +@@ -1473,6 +1500,15 @@ + static int console_trylock_for_printk(void) + { + unsigned int cpu = smp_processor_id(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ int lock = !early_boot_irqs_disabled && (preempt_count() == 0) && ++ !irqs_disabled(); ++#else ++ int lock = 1; ++#endif ++ ++ if (!lock) ++ return 0; + + if (!console_trylock()) + return 0; +@@ -1607,6 +1643,62 @@ + return textlen; + } + ++#ifdef CONFIG_EARLY_PRINTK ++struct console *early_console; ++ ++static void early_vprintk(const char *fmt, va_list ap) ++{ ++ if (early_console) { ++ char buf[512]; ++ int n = vscnprintf(buf, sizeof(buf), fmt, ap); ++ ++ early_console->write(early_console, buf, n); ++ } ++} ++ ++asmlinkage void early_printk(const char *fmt, ...) ++{ ++ va_list ap; ++ ++ va_start(ap, fmt); ++ early_vprintk(fmt, ap); ++ va_end(ap); ++} ++ ++/* ++ * This is independent of any log levels - a global ++ * kill switch that turns off all of printk. ++ * ++ * Used by the NMI watchdog if early-printk is enabled. ++ */ ++static bool __read_mostly printk_killswitch; ++ ++static int __init force_early_printk_setup(char *str) ++{ ++ printk_killswitch = true; ++ return 0; ++} ++early_param("force_early_printk", force_early_printk_setup); ++ ++void printk_kill(void) ++{ ++ printk_killswitch = true; ++} ++ ++static int forced_early_printk(const char *fmt, va_list ap) ++{ ++ if (!printk_killswitch) ++ return 0; ++ early_vprintk(fmt, ap); ++ return 1; ++} ++#else ++static inline int forced_early_printk(const char *fmt, va_list ap) ++{ ++ return 0; ++} ++#endif ++ + asmlinkage int vprintk_emit(int facility, int level, + const char *dict, size_t dictlen, + const char *fmt, va_list args) +@@ -1623,6 +1715,13 @@ + /* cpu currently holding logbuf_lock in this function */ + static unsigned int logbuf_cpu = UINT_MAX; + ++ /* ++ * Fall back to early_printk if a debugging subsystem has ++ * killed printk output ++ */ ++ if (unlikely(forced_early_printk(fmt, args))) ++ return 1; ++ + if (level == LOGLEVEL_SCHED) { + level = LOGLEVEL_DEFAULT; + in_sched = true; +@@ -1764,8 +1863,7 @@ + * console_sem which would prevent anyone from printing to + * console + */ +- preempt_disable(); +- ++ migrate_disable(); + /* + * Try to acquire and then immediately release the console + * semaphore. The release will print out buffers and wake up +@@ -1773,7 +1871,7 @@ + */ + if (console_trylock_for_printk()) + console_unlock(); +- preempt_enable(); ++ migrate_enable(); + lockdep_on(); + } + +@@ -1902,26 +2000,6 @@ + + #endif /* CONFIG_PRINTK */ + +-#ifdef CONFIG_EARLY_PRINTK +-struct console *early_console; +- +-asmlinkage __visible void early_printk(const char *fmt, ...) +-{ +- va_list ap; +- char buf[512]; +- int n; +- +- if (!early_console) +- return; +- +- va_start(ap, fmt); +- n = vscnprintf(buf, sizeof(buf), fmt, ap); +- va_end(ap); +- +- early_console->write(early_console, buf, n); +-} +-#endif +- + static int __add_preferred_console(char *name, int idx, char *options, + char *brl_options) + { +@@ -2143,11 +2221,16 @@ + goto out; + + len = cont_print_text(text, size); ++#ifndef CONFIG_PREEMPT_RT_FULL + raw_spin_unlock(&logbuf_lock); + stop_critical_timings(); + call_console_drivers(cont.level, text, len); + start_critical_timings(); + local_irq_restore(flags); ++#else ++ raw_spin_unlock_irqrestore(&logbuf_lock, flags); ++ call_console_drivers(cont.level, text, len); ++#endif + return; + out: + raw_spin_unlock_irqrestore(&logbuf_lock, flags); +@@ -2235,12 +2318,17 @@ + console_idx = log_next(console_idx); + console_seq++; + console_prev = msg->flags; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ raw_spin_unlock_irqrestore(&logbuf_lock, flags); ++ call_console_drivers(level, text, len); ++#else + raw_spin_unlock(&logbuf_lock); + + stop_critical_timings(); /* don't trace print latency */ + call_console_drivers(level, text, len); + start_critical_timings(); + local_irq_restore(flags); ++#endif + } + console_locked = 0; + +diff -Nur linux-4.1.13.orig/kernel/ptrace.c linux-4.1.13/kernel/ptrace.c +--- linux-4.1.13.orig/kernel/ptrace.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/ptrace.c 2015-11-29 09:23:09.641612055 +0100 +@@ -129,7 +129,12 @@ + + spin_lock_irq(&task->sighand->siglock); + if (task_is_traced(task) && !__fatal_signal_pending(task)) { +- task->state = __TASK_TRACED; ++ raw_spin_lock_irq(&task->pi_lock); ++ if (task->state & __TASK_TRACED) ++ task->state = __TASK_TRACED; ++ else ++ task->saved_state = __TASK_TRACED; ++ raw_spin_unlock_irq(&task->pi_lock); + ret = true; + } + spin_unlock_irq(&task->sighand->siglock); +diff -Nur linux-4.1.13.orig/kernel/rcu/tree.c linux-4.1.13/kernel/rcu/tree.c +--- linux-4.1.13.orig/kernel/rcu/tree.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/rcu/tree.c 2015-11-29 09:23:09.641612055 +0100 +@@ -56,6 +56,11 @@ + #include <linux/random.h> + #include <linux/ftrace_event.h> + #include <linux/suspend.h> ++#include <linux/delay.h> ++#include <linux/gfp.h> ++#include <linux/oom.h> ++#include <linux/smpboot.h> ++#include "../time/tick-internal.h" + + #include "tree.h" + #include "rcu.h" +@@ -220,6 +225,19 @@ + } + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static void rcu_preempt_qs(void); ++ ++void rcu_bh_qs(void) ++{ ++ unsigned long flags; ++ ++ /* Callers to this function, rcu_preempt_qs(), must disable irqs. */ ++ local_irq_save(flags); ++ rcu_preempt_qs(); ++ local_irq_restore(flags); ++} ++#else + void rcu_bh_qs(void) + { + if (!__this_cpu_read(rcu_bh_data.passed_quiesce)) { +@@ -229,6 +247,7 @@ + __this_cpu_write(rcu_bh_data.passed_quiesce, 1); + } + } ++#endif + + static DEFINE_PER_CPU(int, rcu_sched_qs_mask); + +@@ -404,6 +423,7 @@ + } + EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Return the number of RCU BH batches completed thus far for debug & stats. + */ +@@ -431,6 +451,13 @@ + } + EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); + ++#else ++void rcu_force_quiescent_state(void) ++{ ++} ++EXPORT_SYMBOL_GPL(rcu_force_quiescent_state); ++#endif ++ + /* + * Force a quiescent state for RCU-sched. + */ +@@ -1545,7 +1572,7 @@ + !ACCESS_ONCE(rsp->gp_flags) || + !rsp->gp_kthread) + return; +- wake_up(&rsp->gp_wq); ++ swait_wake(&rsp->gp_wq); + } + + /* +@@ -1986,7 +2013,7 @@ + ACCESS_ONCE(rsp->gpnum), + TPS("reqwait")); + rsp->gp_state = RCU_GP_WAIT_GPS; +- wait_event_interruptible(rsp->gp_wq, ++ swait_event_interruptible(rsp->gp_wq, + ACCESS_ONCE(rsp->gp_flags) & + RCU_GP_FLAG_INIT); + /* Locking provides needed memory barrier. */ +@@ -2015,7 +2042,7 @@ + ACCESS_ONCE(rsp->gpnum), + TPS("fqswait")); + rsp->gp_state = RCU_GP_WAIT_FQS; +- ret = wait_event_interruptible_timeout(rsp->gp_wq, ++ ret = swait_event_interruptible_timeout(rsp->gp_wq, + ((gf = ACCESS_ONCE(rsp->gp_flags)) & + RCU_GP_FLAG_FQS) || + (!ACCESS_ONCE(rnp->qsmask) && +@@ -2860,18 +2887,17 @@ + /* + * Do RCU core processing for the current CPU. + */ +-static void rcu_process_callbacks(struct softirq_action *unused) ++static void rcu_process_callbacks(void) + { + struct rcu_state *rsp; + + if (cpu_is_offline(smp_processor_id())) + return; +- trace_rcu_utilization(TPS("Start RCU core")); + for_each_rcu_flavor(rsp) + __rcu_process_callbacks(rsp); +- trace_rcu_utilization(TPS("End RCU core")); + } + ++static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); + /* + * Schedule RCU callback invocation. If the specified type of RCU + * does not support RCU priority boosting, just do a direct call, +@@ -2883,18 +2909,105 @@ + { + if (unlikely(!ACCESS_ONCE(rcu_scheduler_fully_active))) + return; +- if (likely(!rsp->boost)) { +- rcu_do_batch(rsp, rdp); ++ rcu_do_batch(rsp, rdp); ++} ++ ++static void rcu_wake_cond(struct task_struct *t, int status) ++{ ++ /* ++ * If the thread is yielding, only wake it when this ++ * is invoked from idle ++ */ ++ if (t && (status != RCU_KTHREAD_YIELDING || is_idle_task(current))) ++ wake_up_process(t); ++} ++ ++/* ++ * Wake up this CPU's rcuc kthread to do RCU core processing. ++ */ ++static void invoke_rcu_core(void) ++{ ++ unsigned long flags; ++ struct task_struct *t; ++ ++ if (!cpu_online(smp_processor_id())) + return; ++ local_irq_save(flags); ++ __this_cpu_write(rcu_cpu_has_work, 1); ++ t = __this_cpu_read(rcu_cpu_kthread_task); ++ if (t != NULL && current != t) ++ rcu_wake_cond(t, __this_cpu_read(rcu_cpu_kthread_status)); ++ local_irq_restore(flags); ++} ++ ++static void rcu_cpu_kthread_park(unsigned int cpu) ++{ ++ per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; ++} ++ ++static int rcu_cpu_kthread_should_run(unsigned int cpu) ++{ ++ return __this_cpu_read(rcu_cpu_has_work); ++} ++ ++/* ++ * Per-CPU kernel thread that invokes RCU callbacks. This replaces the ++ * RCU softirq used in flavors and configurations of RCU that do not ++ * support RCU priority boosting. ++ */ ++static void rcu_cpu_kthread(unsigned int cpu) ++{ ++ unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); ++ char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); ++ int spincnt; ++ ++ for (spincnt = 0; spincnt < 10; spincnt++) { ++ trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); ++ local_bh_disable(); ++ *statusp = RCU_KTHREAD_RUNNING; ++ this_cpu_inc(rcu_cpu_kthread_loops); ++ local_irq_disable(); ++ work = *workp; ++ *workp = 0; ++ local_irq_enable(); ++ if (work) ++ rcu_process_callbacks(); ++ local_bh_enable(); ++ if (*workp == 0) { ++ trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); ++ *statusp = RCU_KTHREAD_WAITING; ++ return; ++ } + } +- invoke_rcu_callbacks_kthread(); ++ *statusp = RCU_KTHREAD_YIELDING; ++ trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); ++ schedule_timeout_interruptible(2); ++ trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); ++ *statusp = RCU_KTHREAD_WAITING; + } + +-static void invoke_rcu_core(void) ++static struct smp_hotplug_thread rcu_cpu_thread_spec = { ++ .store = &rcu_cpu_kthread_task, ++ .thread_should_run = rcu_cpu_kthread_should_run, ++ .thread_fn = rcu_cpu_kthread, ++ .thread_comm = "rcuc/%u", ++ .setup = rcu_cpu_kthread_setup, ++ .park = rcu_cpu_kthread_park, ++}; ++ ++/* ++ * Spawn per-CPU RCU core processing kthreads. ++ */ ++static int __init rcu_spawn_core_kthreads(void) + { +- if (cpu_online(smp_processor_id())) +- raise_softirq(RCU_SOFTIRQ); ++ int cpu; ++ ++ for_each_possible_cpu(cpu) ++ per_cpu(rcu_cpu_has_work, cpu) = 0; ++ BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); ++ return 0; + } ++early_initcall(rcu_spawn_core_kthreads); + + /* + * Handle any core-RCU processing required by a call_rcu() invocation. +@@ -3040,6 +3153,7 @@ + } + EXPORT_SYMBOL_GPL(call_rcu_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Queue an RCU callback for invocation after a quicker grace period. + */ +@@ -3048,6 +3162,7 @@ + __call_rcu(head, func, &rcu_bh_state, -1, 0); + } + EXPORT_SYMBOL_GPL(call_rcu_bh); ++#endif + + /* + * Queue an RCU callback for lazy invocation after a grace period. +@@ -3139,6 +3254,7 @@ + } + EXPORT_SYMBOL_GPL(synchronize_sched); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed. + * +@@ -3165,6 +3281,7 @@ + wait_rcu_gp(call_rcu_bh); + } + EXPORT_SYMBOL_GPL(synchronize_rcu_bh); ++#endif + + /** + * get_state_synchronize_rcu - Snapshot current RCU state +@@ -3677,6 +3794,7 @@ + mutex_unlock(&rsp->barrier_mutex); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete. + */ +@@ -3685,6 +3803,7 @@ + _rcu_barrier(&rcu_bh_state); + } + EXPORT_SYMBOL_GPL(rcu_barrier_bh); ++#endif + + /** + * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks. +@@ -4021,7 +4140,7 @@ + } + } + +- init_waitqueue_head(&rsp->gp_wq); ++ init_swait_head(&rsp->gp_wq); + rnp = rsp->level[rcu_num_lvls - 1]; + for_each_possible_cpu(i) { + while (i > rnp->grphi) +@@ -4120,7 +4239,6 @@ + rcu_init_one(&rcu_bh_state, &rcu_bh_data); + rcu_init_one(&rcu_sched_state, &rcu_sched_data); + __rcu_init_preempt(); +- open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); + + /* + * We don't need protection against CPU-hotplug here because +diff -Nur linux-4.1.13.orig/kernel/rcu/tree.h linux-4.1.13/kernel/rcu/tree.h +--- linux-4.1.13.orig/kernel/rcu/tree.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/rcu/tree.h 2015-11-29 09:23:09.641612055 +0100 +@@ -27,6 +27,7 @@ + #include <linux/threads.h> + #include <linux/cpumask.h> + #include <linux/seqlock.h> ++#include <linux/wait-simple.h> + + /* + * Define shape of hierarchy based on NR_CPUS, CONFIG_RCU_FANOUT, and +@@ -210,7 +211,7 @@ + /* This can happen due to race conditions. */ + #endif /* #ifdef CONFIG_RCU_BOOST */ + #ifdef CONFIG_RCU_NOCB_CPU +- wait_queue_head_t nocb_gp_wq[2]; ++ struct swait_head nocb_gp_wq[2]; + /* Place for rcu_nocb_kthread() to wait GP. */ + #endif /* #ifdef CONFIG_RCU_NOCB_CPU */ + int need_future_gp[2]; +@@ -349,7 +350,7 @@ + atomic_long_t nocb_q_count_lazy; /* invocation (all stages). */ + struct rcu_head *nocb_follower_head; /* CBs ready to invoke. */ + struct rcu_head **nocb_follower_tail; +- wait_queue_head_t nocb_wq; /* For nocb kthreads to sleep on. */ ++ struct swait_head nocb_wq; /* For nocb kthreads to sleep on. */ + struct task_struct *nocb_kthread; + int nocb_defer_wakeup; /* Defer wakeup of nocb_kthread. */ + +@@ -438,7 +439,7 @@ + unsigned long gpnum; /* Current gp number. */ + unsigned long completed; /* # of last completed gp. */ + struct task_struct *gp_kthread; /* Task for grace periods. */ +- wait_queue_head_t gp_wq; /* Where GP task waits. */ ++ struct swait_head gp_wq; /* Where GP task waits. */ + short gp_flags; /* Commands for GP task. */ + short gp_state; /* GP kthread sleep state. */ + +@@ -529,12 +530,10 @@ + DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data); + #endif /* #ifdef CONFIG_PREEMPT_RCU */ + +-#ifdef CONFIG_RCU_BOOST + DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status); + DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu); + DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); + DECLARE_PER_CPU(char, rcu_cpu_has_work); +-#endif /* #ifdef CONFIG_RCU_BOOST */ + + #ifndef RCU_TREE_NONCORE + +@@ -553,10 +552,9 @@ + static void __init __rcu_init_preempt(void); + static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); + static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); +-static void invoke_rcu_callbacks_kthread(void); + static bool rcu_is_callbacks_kthread(void); ++static void rcu_cpu_kthread_setup(unsigned int cpu); + #ifdef CONFIG_RCU_BOOST +-static void rcu_preempt_do_callbacks(void); + static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp, + struct rcu_node *rnp); + #endif /* #ifdef CONFIG_RCU_BOOST */ +diff -Nur linux-4.1.13.orig/kernel/rcu/tree_plugin.h linux-4.1.13/kernel/rcu/tree_plugin.h +--- linux-4.1.13.orig/kernel/rcu/tree_plugin.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/rcu/tree_plugin.h 2015-11-29 09:23:09.641612055 +0100 +@@ -24,27 +24,20 @@ + * Paul E. McKenney <paulmck@linux.vnet.ibm.com> + */ + +-#include <linux/delay.h> +-#include <linux/gfp.h> +-#include <linux/oom.h> +-#include <linux/smpboot.h> +-#include "../time/tick-internal.h" +- + #ifdef CONFIG_RCU_BOOST + + #include "../locking/rtmutex_common.h" + ++#endif /* #ifdef CONFIG_RCU_BOOST */ ++ + /* + * Control variables for per-CPU and per-rcu_node kthreads. These + * handle all flavors of RCU. + */ +-static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task); + DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); + DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); + DEFINE_PER_CPU(char, rcu_cpu_has_work); + +-#endif /* #ifdef CONFIG_RCU_BOOST */ +- + #ifdef CONFIG_RCU_NOCB_CPU + static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */ + static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */ +@@ -291,7 +284,7 @@ + } + + /* Hardware IRQ handlers cannot block, complain if they get here. */ +- if (in_irq() || in_serving_softirq()) { ++ if (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET)) { + lockdep_rcu_suspicious(__FILE__, __LINE__, + "rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n"); + pr_alert("->rcu_read_unlock_special: %#x (b: %d, nq: %d)\n", +@@ -496,15 +489,6 @@ + t->rcu_read_unlock_special.b.need_qs = true; + } + +-#ifdef CONFIG_RCU_BOOST +- +-static void rcu_preempt_do_callbacks(void) +-{ +- rcu_do_batch(&rcu_preempt_state, this_cpu_ptr(&rcu_preempt_data)); +-} +- +-#endif /* #ifdef CONFIG_RCU_BOOST */ +- + /* + * Queue a preemptible-RCU callback for invocation after a grace period. + */ +@@ -939,6 +923,19 @@ + + #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ + ++/* ++ * If boosting, set rcuc kthreads to realtime priority. ++ */ ++static void rcu_cpu_kthread_setup(unsigned int cpu) ++{ ++#ifdef CONFIG_RCU_BOOST ++ struct sched_param sp; ++ ++ sp.sched_priority = kthread_prio; ++ sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); ++#endif /* #ifdef CONFIG_RCU_BOOST */ ++} ++ + #ifdef CONFIG_RCU_BOOST + + #include "../locking/rtmutex_common.h" +@@ -970,16 +967,6 @@ + + #endif /* #else #ifdef CONFIG_RCU_TRACE */ + +-static void rcu_wake_cond(struct task_struct *t, int status) +-{ +- /* +- * If the thread is yielding, only wake it when this +- * is invoked from idle +- */ +- if (status != RCU_KTHREAD_YIELDING || is_idle_task(current)) +- wake_up_process(t); +-} +- + /* + * Carry out RCU priority boosting on the task indicated by ->exp_tasks + * or ->boost_tasks, advancing the pointer to the next task in the +@@ -1125,23 +1112,6 @@ + } + + /* +- * Wake up the per-CPU kthread to invoke RCU callbacks. +- */ +-static void invoke_rcu_callbacks_kthread(void) +-{ +- unsigned long flags; +- +- local_irq_save(flags); +- __this_cpu_write(rcu_cpu_has_work, 1); +- if (__this_cpu_read(rcu_cpu_kthread_task) != NULL && +- current != __this_cpu_read(rcu_cpu_kthread_task)) { +- rcu_wake_cond(__this_cpu_read(rcu_cpu_kthread_task), +- __this_cpu_read(rcu_cpu_kthread_status)); +- } +- local_irq_restore(flags); +-} +- +-/* + * Is the current CPU running the RCU-callbacks kthread? + * Caller must have preemption disabled. + */ +@@ -1196,67 +1166,6 @@ + return 0; + } + +-static void rcu_kthread_do_work(void) +-{ +- rcu_do_batch(&rcu_sched_state, this_cpu_ptr(&rcu_sched_data)); +- rcu_do_batch(&rcu_bh_state, this_cpu_ptr(&rcu_bh_data)); +- rcu_preempt_do_callbacks(); +-} +- +-static void rcu_cpu_kthread_setup(unsigned int cpu) +-{ +- struct sched_param sp; +- +- sp.sched_priority = kthread_prio; +- sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); +-} +- +-static void rcu_cpu_kthread_park(unsigned int cpu) +-{ +- per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; +-} +- +-static int rcu_cpu_kthread_should_run(unsigned int cpu) +-{ +- return __this_cpu_read(rcu_cpu_has_work); +-} +- +-/* +- * Per-CPU kernel thread that invokes RCU callbacks. This replaces the +- * RCU softirq used in flavors and configurations of RCU that do not +- * support RCU priority boosting. +- */ +-static void rcu_cpu_kthread(unsigned int cpu) +-{ +- unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status); +- char work, *workp = this_cpu_ptr(&rcu_cpu_has_work); +- int spincnt; +- +- for (spincnt = 0; spincnt < 10; spincnt++) { +- trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait")); +- local_bh_disable(); +- *statusp = RCU_KTHREAD_RUNNING; +- this_cpu_inc(rcu_cpu_kthread_loops); +- local_irq_disable(); +- work = *workp; +- *workp = 0; +- local_irq_enable(); +- if (work) +- rcu_kthread_do_work(); +- local_bh_enable(); +- if (*workp == 0) { +- trace_rcu_utilization(TPS("End CPU kthread@rcu_wait")); +- *statusp = RCU_KTHREAD_WAITING; +- return; +- } +- } +- *statusp = RCU_KTHREAD_YIELDING; +- trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield")); +- schedule_timeout_interruptible(2); +- trace_rcu_utilization(TPS("End CPU kthread@rcu_yield")); +- *statusp = RCU_KTHREAD_WAITING; +-} +- + /* + * Set the per-rcu_node kthread's affinity to cover all CPUs that are + * served by the rcu_node in question. The CPU hotplug lock is still +@@ -1286,26 +1195,12 @@ + free_cpumask_var(cm); + } + +-static struct smp_hotplug_thread rcu_cpu_thread_spec = { +- .store = &rcu_cpu_kthread_task, +- .thread_should_run = rcu_cpu_kthread_should_run, +- .thread_fn = rcu_cpu_kthread, +- .thread_comm = "rcuc/%u", +- .setup = rcu_cpu_kthread_setup, +- .park = rcu_cpu_kthread_park, +-}; +- + /* + * Spawn boost kthreads -- called as soon as the scheduler is running. + */ + static void __init rcu_spawn_boost_kthreads(void) + { + struct rcu_node *rnp; +- int cpu; +- +- for_each_possible_cpu(cpu) +- per_cpu(rcu_cpu_has_work, cpu) = 0; +- BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec)); + rcu_for_each_leaf_node(rcu_state_p, rnp) + (void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp); + } +@@ -1328,11 +1223,6 @@ + raw_spin_unlock_irqrestore(&rnp->lock, flags); + } + +-static void invoke_rcu_callbacks_kthread(void) +-{ +- WARN_ON_ONCE(1); +-} +- + static bool rcu_is_callbacks_kthread(void) + { + return false; +@@ -1356,7 +1246,7 @@ + + #endif /* #else #ifdef CONFIG_RCU_BOOST */ + +-#if !defined(CONFIG_RCU_FAST_NO_HZ) ++#if !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) + + /* + * Check to see if any future RCU-related work will need to be done +@@ -1374,7 +1264,9 @@ + return rcu_cpu_has_callbacks(NULL); + } + #endif /* #ifndef CONFIG_RCU_NOCB_CPU_ALL */ ++#endif /* !defined(CONFIG_RCU_FAST_NO_HZ) || defined(CONFIG_PREEMPT_RT_FULL) */ + ++#if !defined(CONFIG_RCU_FAST_NO_HZ) + /* + * Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up + * after it. +@@ -1472,6 +1364,8 @@ + return cbs_ready; + } + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * Allow the CPU to enter dyntick-idle mode unless it has callbacks ready + * to invoke. If the CPU has callbacks, try to advance them. Tell the +@@ -1512,7 +1406,7 @@ + return 0; + } + #endif /* #ifndef CONFIG_RCU_NOCB_CPU_ALL */ +- ++#endif /* #ifndef CONFIG_PREEMPT_RT_FULL */ + /* + * Prepare a CPU for idle from an RCU perspective. The first major task + * is to sense whether nohz mode has been enabled or disabled via sysfs. +@@ -1859,7 +1753,7 @@ + */ + static void rcu_nocb_gp_cleanup(struct rcu_state *rsp, struct rcu_node *rnp) + { +- wake_up_all(&rnp->nocb_gp_wq[rnp->completed & 0x1]); ++ swait_wake_all(&rnp->nocb_gp_wq[rnp->completed & 0x1]); + } + + /* +@@ -1877,8 +1771,8 @@ + + static void rcu_init_one_nocb(struct rcu_node *rnp) + { +- init_waitqueue_head(&rnp->nocb_gp_wq[0]); +- init_waitqueue_head(&rnp->nocb_gp_wq[1]); ++ init_swait_head(&rnp->nocb_gp_wq[0]); ++ init_swait_head(&rnp->nocb_gp_wq[1]); + } + + #ifndef CONFIG_RCU_NOCB_CPU_ALL +@@ -1903,7 +1797,7 @@ + if (ACCESS_ONCE(rdp_leader->nocb_leader_sleep) || force) { + /* Prior smp_mb__after_atomic() orders against prior enqueue. */ + ACCESS_ONCE(rdp_leader->nocb_leader_sleep) = false; +- wake_up(&rdp_leader->nocb_wq); ++ swait_wake(&rdp_leader->nocb_wq); + } + } + +@@ -2116,7 +2010,7 @@ + */ + trace_rcu_future_gp(rnp, rdp, c, TPS("StartWait")); + for (;;) { +- wait_event_interruptible( ++ swait_event_interruptible( + rnp->nocb_gp_wq[c & 0x1], + (d = ULONG_CMP_GE(ACCESS_ONCE(rnp->completed), c))); + if (likely(d)) +@@ -2144,7 +2038,7 @@ + /* Wait for callbacks to appear. */ + if (!rcu_nocb_poll) { + trace_rcu_nocb_wake(my_rdp->rsp->name, my_rdp->cpu, "Sleep"); +- wait_event_interruptible(my_rdp->nocb_wq, ++ swait_event_interruptible(my_rdp->nocb_wq, + !ACCESS_ONCE(my_rdp->nocb_leader_sleep)); + /* Memory barrier handled by smp_mb() calls below and repoll. */ + } else if (firsttime) { +@@ -2219,7 +2113,7 @@ + * List was empty, wake up the follower. + * Memory barriers supplied by atomic_long_add(). + */ +- wake_up(&rdp->nocb_wq); ++ swait_wake(&rdp->nocb_wq); + } + } + +@@ -2240,7 +2134,7 @@ + if (!rcu_nocb_poll) { + trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, + "FollowerSleep"); +- wait_event_interruptible(rdp->nocb_wq, ++ swait_event_interruptible(rdp->nocb_wq, + ACCESS_ONCE(rdp->nocb_follower_head)); + } else if (firsttime) { + /* Don't drown trace log with "Poll"! */ +@@ -2399,7 +2293,7 @@ + static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp) + { + rdp->nocb_tail = &rdp->nocb_head; +- init_waitqueue_head(&rdp->nocb_wq); ++ init_swait_head(&rdp->nocb_wq); + rdp->nocb_follower_tail = &rdp->nocb_follower_head; + } + +diff -Nur linux-4.1.13.orig/kernel/rcu/update.c linux-4.1.13/kernel/rcu/update.c +--- linux-4.1.13.orig/kernel/rcu/update.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/rcu/update.c 2015-11-29 09:23:09.641612055 +0100 +@@ -227,6 +227,7 @@ + } + EXPORT_SYMBOL_GPL(rcu_read_lock_held); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /** + * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section? + * +@@ -253,6 +254,7 @@ + return in_softirq() || irqs_disabled(); + } + EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held); ++#endif + + #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */ + +diff -Nur linux-4.1.13.orig/kernel/relay.c linux-4.1.13/kernel/relay.c +--- linux-4.1.13.orig/kernel/relay.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/relay.c 2015-11-29 09:23:09.641612055 +0100 +@@ -339,6 +339,10 @@ + { + struct rchan_buf *buf = (struct rchan_buf *)data; + wake_up_interruptible(&buf->read_wait); ++ /* ++ * Stupid polling for now: ++ */ ++ mod_timer(&buf->timer, jiffies + 1); + } + + /** +@@ -356,6 +360,7 @@ + init_waitqueue_head(&buf->read_wait); + kref_init(&buf->kref); + setup_timer(&buf->timer, wakeup_readers, (unsigned long)buf); ++ mod_timer(&buf->timer, jiffies + 1); + } else + del_timer_sync(&buf->timer); + +@@ -739,15 +744,6 @@ + else + buf->early_bytes += buf->chan->subbuf_size - + buf->padding[old_subbuf]; +- smp_mb(); +- if (waitqueue_active(&buf->read_wait)) +- /* +- * Calling wake_up_interruptible() from here +- * will deadlock if we happen to be logging +- * from the scheduler (trying to re-grab +- * rq->lock), so defer it. +- */ +- mod_timer(&buf->timer, jiffies + 1); + } + + old = buf->data; +diff -Nur linux-4.1.13.orig/kernel/sched/completion.c linux-4.1.13/kernel/sched/completion.c +--- linux-4.1.13.orig/kernel/sched/completion.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/completion.c 2015-11-29 09:23:09.641612055 +0100 +@@ -30,10 +30,10 @@ + { + unsigned long flags; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + x->done++; +- __wake_up_locked(&x->wait, TASK_NORMAL, 1); +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ __swait_wake_locked(&x->wait, TASK_NORMAL, 1); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + } + EXPORT_SYMBOL(complete); + +@@ -50,10 +50,10 @@ + { + unsigned long flags; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + x->done += UINT_MAX/2; +- __wake_up_locked(&x->wait, TASK_NORMAL, 0); +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ __swait_wake_locked(&x->wait, TASK_NORMAL, 0); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + } + EXPORT_SYMBOL(complete_all); + +@@ -62,20 +62,20 @@ + long (*action)(long), long timeout, int state) + { + if (!x->done) { +- DECLARE_WAITQUEUE(wait, current); ++ DEFINE_SWAITER(wait); + +- __add_wait_queue_tail_exclusive(&x->wait, &wait); ++ swait_prepare_locked(&x->wait, &wait); + do { + if (signal_pending_state(state, current)) { + timeout = -ERESTARTSYS; + break; + } + __set_current_state(state); +- spin_unlock_irq(&x->wait.lock); ++ raw_spin_unlock_irq(&x->wait.lock); + timeout = action(timeout); +- spin_lock_irq(&x->wait.lock); ++ raw_spin_lock_irq(&x->wait.lock); + } while (!x->done && timeout); +- __remove_wait_queue(&x->wait, &wait); ++ swait_finish_locked(&x->wait, &wait); + if (!x->done) + return timeout; + } +@@ -89,9 +89,9 @@ + { + might_sleep(); + +- spin_lock_irq(&x->wait.lock); ++ raw_spin_lock_irq(&x->wait.lock); + timeout = do_wait_for_common(x, action, timeout, state); +- spin_unlock_irq(&x->wait.lock); ++ raw_spin_unlock_irq(&x->wait.lock); + return timeout; + } + +@@ -277,12 +277,12 @@ + if (!READ_ONCE(x->done)) + return 0; + +- spin_lock_irqsave(&x->wait.lock, flags); ++ raw_spin_lock_irqsave(&x->wait.lock, flags); + if (!x->done) + ret = 0; + else + x->done--; +- spin_unlock_irqrestore(&x->wait.lock, flags); ++ raw_spin_unlock_irqrestore(&x->wait.lock, flags); + return ret; + } + EXPORT_SYMBOL(try_wait_for_completion); +@@ -311,7 +311,7 @@ + * after it's acquired the lock. + */ + smp_rmb(); +- spin_unlock_wait(&x->wait.lock); ++ raw_spin_unlock_wait(&x->wait.lock); + return true; + } + EXPORT_SYMBOL(completion_done); +diff -Nur linux-4.1.13.orig/kernel/sched/core.c linux-4.1.13/kernel/sched/core.c +--- linux-4.1.13.orig/kernel/sched/core.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/core.c 2015-11-29 09:23:09.645611789 +0100 +@@ -282,7 +282,11 @@ + * Number of tasks to iterate in a single balance run. + * Limited because this is done with IRQs disabled. + */ ++#ifndef CONFIG_PREEMPT_RT_FULL + const_debug unsigned int sysctl_sched_nr_migrate = 32; ++#else ++const_debug unsigned int sysctl_sched_nr_migrate = 8; ++#endif + + /* + * period over which we average the RT time consumption, measured +@@ -461,6 +465,7 @@ + + hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + rq->hrtick_timer.function = hrtick; ++ rq->hrtick_timer.irqsafe = 1; + } + #else /* CONFIG_SCHED_HRTICK */ + static inline void hrtick_clear(struct rq *rq) +@@ -541,6 +546,52 @@ + #endif + #endif + ++void wake_q_add(struct wake_q_head *head, struct task_struct *task) ++{ ++ struct wake_q_node *node = &task->wake_q; ++ ++ /* ++ * Atomically grab the task, if ->wake_q is !nil already it means ++ * its already queued (either by us or someone else) and will get the ++ * wakeup due to that. ++ * ++ * This cmpxchg() implies a full barrier, which pairs with the write ++ * barrier implied by the wakeup in wake_up_list(). ++ */ ++ if (cmpxchg(&node->next, NULL, WAKE_Q_TAIL)) ++ return; ++ ++ get_task_struct(task); ++ ++ /* ++ * The head is context local, there can be no concurrency. ++ */ ++ *head->lastp = node; ++ head->lastp = &node->next; ++} ++ ++void wake_up_q(struct wake_q_head *head) ++{ ++ struct wake_q_node *node = head->first; ++ ++ while (node != WAKE_Q_TAIL) { ++ struct task_struct *task; ++ ++ task = container_of(node, struct task_struct, wake_q); ++ BUG_ON(!task); ++ /* task can safely be re-inserted now */ ++ node = node->next; ++ task->wake_q.next = NULL; ++ ++ /* ++ * wake_up_process() implies a wmb() to pair with the queueing ++ * in wake_q_add() so as not to miss wakeups. ++ */ ++ wake_up_process(task); ++ put_task_struct(task); ++ } ++} ++ + /* + * resched_curr - mark rq's current task 'to be rescheduled now'. + * +@@ -572,6 +623,38 @@ + trace_sched_wake_idle_without_ipi(cpu); + } + ++#ifdef CONFIG_PREEMPT_LAZY ++void resched_curr_lazy(struct rq *rq) ++{ ++ struct task_struct *curr = rq->curr; ++ int cpu; ++ ++ if (!sched_feat(PREEMPT_LAZY)) { ++ resched_curr(rq); ++ return; ++ } ++ ++ lockdep_assert_held(&rq->lock); ++ ++ if (test_tsk_need_resched(curr)) ++ return; ++ ++ if (test_tsk_need_resched_lazy(curr)) ++ return; ++ ++ set_tsk_need_resched_lazy(curr); ++ ++ cpu = cpu_of(rq); ++ if (cpu == smp_processor_id()) ++ return; ++ ++ /* NEED_RESCHED_LAZY must be visible before we test polling */ ++ smp_mb(); ++ if (!tsk_is_polling(curr)) ++ smp_send_reschedule(cpu); ++} ++#endif ++ + void resched_cpu(int cpu) + { + struct rq *rq = cpu_rq(cpu); +@@ -595,12 +678,14 @@ + */ + int get_nohz_timer_target(int pinned) + { +- int cpu = smp_processor_id(); ++ int cpu; + int i; + struct sched_domain *sd; + ++ preempt_disable_rt(); ++ cpu = smp_processor_id(); + if (pinned || !get_sysctl_timer_migration() || !idle_cpu(cpu)) +- return cpu; ++ goto preempt_en_rt; + + rcu_read_lock(); + for_each_domain(cpu, sd) { +@@ -613,6 +698,8 @@ + } + unlock: + rcu_read_unlock(); ++preempt_en_rt: ++ preempt_enable_rt(); + return cpu; + } + /* +@@ -1164,6 +1251,18 @@ + + static int migration_cpu_stop(void *data); + ++static bool check_task_state(struct task_struct *p, long match_state) ++{ ++ bool match = false; ++ ++ raw_spin_lock_irq(&p->pi_lock); ++ if (p->state == match_state || p->saved_state == match_state) ++ match = true; ++ raw_spin_unlock_irq(&p->pi_lock); ++ ++ return match; ++} ++ + /* + * wait_task_inactive - wait for a thread to unschedule. + * +@@ -1208,7 +1307,7 @@ + * is actually now running somewhere else! + */ + while (task_running(rq, p)) { +- if (match_state && unlikely(p->state != match_state)) ++ if (match_state && !check_task_state(p, match_state)) + return 0; + cpu_relax(); + } +@@ -1223,7 +1322,8 @@ + running = task_running(rq, p); + queued = task_on_rq_queued(p); + ncsw = 0; +- if (!match_state || p->state == match_state) ++ if (!match_state || p->state == match_state || ++ p->saved_state == match_state) + ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ + task_rq_unlock(rq, p, &flags); + +@@ -1449,10 +1549,6 @@ + { + activate_task(rq, p, en_flags); + p->on_rq = TASK_ON_RQ_QUEUED; +- +- /* if a worker is waking up, notify workqueue */ +- if (p->flags & PF_WQ_WORKER) +- wq_worker_waking_up(p, cpu_of(rq)); + } + + /* +@@ -1462,9 +1558,9 @@ + ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) + { + check_preempt_curr(rq, p, wake_flags); +- trace_sched_wakeup(p, true); +- + p->state = TASK_RUNNING; ++ trace_sched_wakeup(p); ++ + #ifdef CONFIG_SMP + if (p->sched_class->task_woken) + p->sched_class->task_woken(rq, p); +@@ -1666,8 +1762,29 @@ + */ + smp_mb__before_spinlock(); + raw_spin_lock_irqsave(&p->pi_lock, flags); +- if (!(p->state & state)) ++ if (!(p->state & state)) { ++ /* ++ * The task might be running due to a spinlock sleeper ++ * wakeup. Check the saved state and set it to running ++ * if the wakeup condition is true. ++ */ ++ if (!(wake_flags & WF_LOCK_SLEEPER)) { ++ if (p->saved_state & state) { ++ p->saved_state = TASK_RUNNING; ++ success = 1; ++ } ++ } + goto out; ++ } ++ ++ /* ++ * If this is a regular wakeup, then we can unconditionally ++ * clear the saved state of a "lock sleeper". ++ */ ++ if (!(wake_flags & WF_LOCK_SLEEPER)) ++ p->saved_state = TASK_RUNNING; ++ ++ trace_sched_waking(p); + + success = 1; /* we're going to change ->state */ + cpu = task_cpu(p); +@@ -1710,42 +1827,6 @@ + } + + /** +- * try_to_wake_up_local - try to wake up a local task with rq lock held +- * @p: the thread to be awakened +- * +- * Put @p on the run-queue if it's not already there. The caller must +- * ensure that this_rq() is locked, @p is bound to this_rq() and not +- * the current task. +- */ +-static void try_to_wake_up_local(struct task_struct *p) +-{ +- struct rq *rq = task_rq(p); +- +- if (WARN_ON_ONCE(rq != this_rq()) || +- WARN_ON_ONCE(p == current)) +- return; +- +- lockdep_assert_held(&rq->lock); +- +- if (!raw_spin_trylock(&p->pi_lock)) { +- raw_spin_unlock(&rq->lock); +- raw_spin_lock(&p->pi_lock); +- raw_spin_lock(&rq->lock); +- } +- +- if (!(p->state & TASK_NORMAL)) +- goto out; +- +- if (!task_on_rq_queued(p)) +- ttwu_activate(rq, p, ENQUEUE_WAKEUP); +- +- ttwu_do_wakeup(rq, p, 0); +- ttwu_stat(p, smp_processor_id(), 0); +-out: +- raw_spin_unlock(&p->pi_lock); +-} +- +-/** + * wake_up_process - Wake up a specific process + * @p: The process to be woken up. + * +@@ -1759,11 +1840,23 @@ + */ + int wake_up_process(struct task_struct *p) + { +- WARN_ON(task_is_stopped_or_traced(p)); ++ WARN_ON(__task_is_stopped_or_traced(p)); + return try_to_wake_up(p, TASK_NORMAL, 0); + } + EXPORT_SYMBOL(wake_up_process); + ++/** ++ * wake_up_lock_sleeper - Wake up a specific process blocked on a "sleeping lock" ++ * @p: The process to be woken up. ++ * ++ * Same as wake_up_process() above, but wake_flags=WF_LOCK_SLEEPER to indicate ++ * the nature of the wakeup. ++ */ ++int wake_up_lock_sleeper(struct task_struct *p) ++{ ++ return try_to_wake_up(p, TASK_ALL, WF_LOCK_SLEEPER); ++} ++ + int wake_up_state(struct task_struct *p, unsigned int state) + { + return try_to_wake_up(p, state, 0); +@@ -1959,6 +2052,9 @@ + p->on_cpu = 0; + #endif + init_task_preempt_count(p); ++#ifdef CONFIG_HAVE_PREEMPT_LAZY ++ task_thread_info(p)->preempt_lazy_count = 0; ++#endif + #ifdef CONFIG_SMP + plist_node_init(&p->pushable_tasks, MAX_PRIO); + RB_CLEAR_NODE(&p->pushable_dl_tasks); +@@ -2094,7 +2190,7 @@ + rq = __task_rq_lock(p); + activate_task(rq, p, 0); + p->on_rq = TASK_ON_RQ_QUEUED; +- trace_sched_wakeup_new(p, true); ++ trace_sched_wakeup_new(p); + check_preempt_curr(rq, p, WF_FORK); + #ifdef CONFIG_SMP + if (p->sched_class->task_woken) +@@ -2231,8 +2327,12 @@ + finish_arch_post_lock_switch(); + + fire_sched_in_preempt_notifiers(current); ++ /* ++ * We use mmdrop_delayed() here so we don't have to do the ++ * full __mmdrop() when we are the last user. ++ */ + if (mm) +- mmdrop(mm); ++ mmdrop_delayed(mm); + if (unlikely(prev_state == TASK_DEAD)) { + if (prev->sched_class->task_dead) + prev->sched_class->task_dead(prev); +@@ -2657,6 +2757,133 @@ + schedstat_inc(this_rq(), sched_count); + } + ++#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP) ++#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */ ++#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN) ++#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN) ++ ++static inline void update_migrate_disable(struct task_struct *p) ++{ ++ const struct cpumask *mask; ++ ++ if (likely(!p->migrate_disable)) ++ return; ++ ++ /* Did we already update affinity? */ ++ if (unlikely(migrate_disabled_updated(p))) ++ return; ++ ++ /* ++ * Since this is always current we can get away with only locking ++ * rq->lock, the ->cpus_allowed value can normally only be changed ++ * while holding both p->pi_lock and rq->lock, but seeing that this ++ * is current, we cannot actually be waking up, so all code that ++ * relies on serialization against p->pi_lock is out of scope. ++ * ++ * Having rq->lock serializes us against things like ++ * set_cpus_allowed_ptr() that can still happen concurrently. ++ */ ++ mask = tsk_cpus_allowed(p); ++ ++ if (p->sched_class->set_cpus_allowed) ++ p->sched_class->set_cpus_allowed(p, mask); ++ /* mask==cpumask_of(task_cpu(p)) which has a cpumask_weight==1 */ ++ p->nr_cpus_allowed = 1; ++ ++ /* Let migrate_enable know to fix things back up */ ++ p->migrate_disable |= MIGRATE_DISABLE_SET_AFFIN; ++} ++ ++void migrate_disable(void) ++{ ++ struct task_struct *p = current; ++ ++ if (in_atomic()) { ++#ifdef CONFIG_SCHED_DEBUG ++ p->migrate_disable_atomic++; ++#endif ++ return; ++ } ++ ++#ifdef CONFIG_SCHED_DEBUG ++ if (unlikely(p->migrate_disable_atomic)) { ++ tracing_off(); ++ WARN_ON_ONCE(1); ++ } ++#endif ++ ++ if (p->migrate_disable) { ++ p->migrate_disable++; ++ return; ++ } ++ ++ preempt_disable(); ++ preempt_lazy_disable(); ++ pin_current_cpu(); ++ p->migrate_disable = 1; ++ preempt_enable(); ++} ++EXPORT_SYMBOL(migrate_disable); ++ ++void migrate_enable(void) ++{ ++ struct task_struct *p = current; ++ const struct cpumask *mask; ++ unsigned long flags; ++ struct rq *rq; ++ ++ if (in_atomic()) { ++#ifdef CONFIG_SCHED_DEBUG ++ p->migrate_disable_atomic--; ++#endif ++ return; ++ } ++ ++#ifdef CONFIG_SCHED_DEBUG ++ if (unlikely(p->migrate_disable_atomic)) { ++ tracing_off(); ++ WARN_ON_ONCE(1); ++ } ++#endif ++ WARN_ON_ONCE(p->migrate_disable <= 0); ++ ++ if (migrate_disable_count(p) > 1) { ++ p->migrate_disable--; ++ return; ++ } ++ ++ preempt_disable(); ++ if (unlikely(migrate_disabled_updated(p))) { ++ /* ++ * Undo whatever update_migrate_disable() did, also see there ++ * about locking. ++ */ ++ rq = this_rq(); ++ raw_spin_lock_irqsave(&rq->lock, flags); ++ ++ /* ++ * Clearing migrate_disable causes tsk_cpus_allowed to ++ * show the tasks original cpu affinity. ++ */ ++ p->migrate_disable = 0; ++ mask = tsk_cpus_allowed(p); ++ if (p->sched_class->set_cpus_allowed) ++ p->sched_class->set_cpus_allowed(p, mask); ++ p->nr_cpus_allowed = cpumask_weight(mask); ++ raw_spin_unlock_irqrestore(&rq->lock, flags); ++ } else ++ p->migrate_disable = 0; ++ ++ unpin_current_cpu(); ++ preempt_enable(); ++ preempt_lazy_enable(); ++} ++EXPORT_SYMBOL(migrate_enable); ++#else ++static inline void update_migrate_disable(struct task_struct *p) { } ++#define migrate_disabled_updated(p) 0 ++#endif ++ + /* + * Pick up the highest-prio task: + */ +@@ -2763,6 +2990,8 @@ + smp_mb__before_spinlock(); + raw_spin_lock_irq(&rq->lock); + ++ update_migrate_disable(prev); ++ + rq->clock_skip_update <<= 1; /* promote REQ to ACT */ + + switch_count = &prev->nivcsw; +@@ -2772,19 +3001,6 @@ + } else { + deactivate_task(rq, prev, DEQUEUE_SLEEP); + prev->on_rq = 0; +- +- /* +- * If a worker went to sleep, notify and ask workqueue +- * whether it wants to wake up a task to maintain +- * concurrency. +- */ +- if (prev->flags & PF_WQ_WORKER) { +- struct task_struct *to_wakeup; +- +- to_wakeup = wq_worker_sleeping(prev, cpu); +- if (to_wakeup) +- try_to_wake_up_local(to_wakeup); +- } + } + switch_count = &prev->nvcsw; + } +@@ -2794,6 +3010,7 @@ + + next = pick_next_task(rq, prev); + clear_tsk_need_resched(prev); ++ clear_tsk_need_resched_lazy(prev); + clear_preempt_need_resched(); + rq->clock_skip_update = 0; + +@@ -2814,8 +3031,19 @@ + + static inline void sched_submit_work(struct task_struct *tsk) + { +- if (!tsk->state || tsk_is_pi_blocked(tsk)) ++ if (!tsk->state) ++ return; ++ /* ++ * If a worker went to sleep, notify and ask workqueue whether ++ * it wants to wake up a task to maintain concurrency. ++ */ ++ if (tsk->flags & PF_WQ_WORKER) ++ wq_worker_sleeping(tsk); ++ ++ ++ if (tsk_is_pi_blocked(tsk)) + return; ++ + /* + * If we are going to sleep and we have plugged IO queued, + * make sure to submit it to avoid deadlocks. +@@ -2824,6 +3052,12 @@ + blk_schedule_flush_plug(tsk); + } + ++static void sched_update_worker(struct task_struct *tsk) ++{ ++ if (tsk->flags & PF_WQ_WORKER) ++ wq_worker_running(tsk); ++} ++ + asmlinkage __visible void __sched schedule(void) + { + struct task_struct *tsk = current; +@@ -2832,6 +3066,7 @@ + do { + __schedule(); + } while (need_resched()); ++ sched_update_worker(tsk); + } + EXPORT_SYMBOL(schedule); + +@@ -2923,6 +3158,14 @@ + if (likely(!preemptible())) + return; + ++#ifdef CONFIG_PREEMPT_LAZY ++ /* ++ * Check for lazy preemption ++ */ ++ if (current_thread_info()->preempt_lazy_count && ++ !test_thread_flag(TIF_NEED_RESCHED)) ++ return; ++#endif + do { + __preempt_count_add(PREEMPT_ACTIVE); + /* +@@ -2931,7 +3174,16 @@ + * an infinite recursion. + */ + prev_ctx = exception_enter(); ++ /* ++ * The add/subtract must not be traced by the function ++ * tracer. But we still want to account for the ++ * preempt off latency tracer. Since the _notrace versions ++ * of add/subtract skip the accounting for latency tracer ++ * we must force it manually. ++ */ ++ start_critical_timings(); + __schedule(); ++ stop_critical_timings(); + exception_exit(prev_ctx); + + __preempt_count_sub(PREEMPT_ACTIVE); +@@ -4268,6 +4520,7 @@ + } + EXPORT_SYMBOL(__cond_resched_lock); + ++#ifndef CONFIG_PREEMPT_RT_FULL + int __sched __cond_resched_softirq(void) + { + BUG_ON(!in_softirq()); +@@ -4281,6 +4534,7 @@ + return 0; + } + EXPORT_SYMBOL(__cond_resched_softirq); ++#endif + + /** + * yield - yield the current processor to other threads. +@@ -4635,7 +4889,9 @@ + + /* Set the preempt count _outside_ the spinlocks! */ + init_idle_preempt_count(idle, cpu); +- ++#ifdef CONFIG_HAVE_PREEMPT_LAZY ++ task_thread_info(idle)->preempt_lazy_count = 0; ++#endif + /* + * The idle tasks have their own, simple scheduling class: + */ +@@ -4755,11 +5011,91 @@ + + void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) + { +- if (p->sched_class->set_cpus_allowed) +- p->sched_class->set_cpus_allowed(p, new_mask); ++ if (!migrate_disabled_updated(p)) { ++ if (p->sched_class->set_cpus_allowed) ++ p->sched_class->set_cpus_allowed(p, new_mask); ++ p->nr_cpus_allowed = cpumask_weight(new_mask); ++ } + + cpumask_copy(&p->cpus_allowed, new_mask); +- p->nr_cpus_allowed = cpumask_weight(new_mask); ++} ++ ++static DEFINE_PER_CPU(struct cpumask, sched_cpumasks); ++static DEFINE_MUTEX(sched_down_mutex); ++static cpumask_t sched_down_cpumask; ++ ++void tell_sched_cpu_down_begin(int cpu) ++{ ++ mutex_lock(&sched_down_mutex); ++ cpumask_set_cpu(cpu, &sched_down_cpumask); ++ mutex_unlock(&sched_down_mutex); ++} ++ ++void tell_sched_cpu_down_done(int cpu) ++{ ++ mutex_lock(&sched_down_mutex); ++ cpumask_clear_cpu(cpu, &sched_down_cpumask); ++ mutex_unlock(&sched_down_mutex); ++} ++ ++/** ++ * migrate_me - try to move the current task off this cpu ++ * ++ * Used by the pin_current_cpu() code to try to get tasks ++ * to move off the current CPU as it is going down. ++ * It will only move the task if the task isn't pinned to ++ * the CPU (with migrate_disable, affinity or NO_SETAFFINITY) ++ * and the task has to be in a RUNNING state. Otherwise the ++ * movement of the task will wake it up (change its state ++ * to running) when the task did not expect it. ++ * ++ * Returns 1 if it succeeded in moving the current task ++ * 0 otherwise. ++ */ ++int migrate_me(void) ++{ ++ struct task_struct *p = current; ++ struct migration_arg arg; ++ struct cpumask *cpumask; ++ struct cpumask *mask; ++ unsigned long flags; ++ unsigned int dest_cpu; ++ struct rq *rq; ++ ++ /* ++ * We can not migrate tasks bounded to a CPU or tasks not ++ * running. The movement of the task will wake it up. ++ */ ++ if (p->flags & PF_NO_SETAFFINITY || p->state) ++ return 0; ++ ++ mutex_lock(&sched_down_mutex); ++ rq = task_rq_lock(p, &flags); ++ ++ cpumask = this_cpu_ptr(&sched_cpumasks); ++ mask = &p->cpus_allowed; ++ ++ cpumask_andnot(cpumask, mask, &sched_down_cpumask); ++ ++ if (!cpumask_weight(cpumask)) { ++ /* It's only on this CPU? */ ++ task_rq_unlock(rq, p, &flags); ++ mutex_unlock(&sched_down_mutex); ++ return 0; ++ } ++ ++ dest_cpu = cpumask_any_and(cpu_active_mask, cpumask); ++ ++ arg.task = p; ++ arg.dest_cpu = dest_cpu; ++ ++ task_rq_unlock(rq, p, &flags); ++ ++ stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg); ++ tlb_migrate_finish(p->mm); ++ mutex_unlock(&sched_down_mutex); ++ ++ return 1; + } + + /* +@@ -4805,7 +5141,7 @@ + do_set_cpus_allowed(p, new_mask); + + /* Can the task run on the task's current CPU? If so, we're done */ +- if (cpumask_test_cpu(task_cpu(p), new_mask)) ++ if (cpumask_test_cpu(task_cpu(p), new_mask) || __migrate_disabled(p)) + goto out; + + dest_cpu = cpumask_any_and(cpu_active_mask, new_mask); +@@ -4945,6 +5281,8 @@ + + #ifdef CONFIG_HOTPLUG_CPU + ++static DEFINE_PER_CPU(struct mm_struct *, idle_last_mm); ++ + /* + * Ensures that the idle task is using init_mm right before its cpu goes + * offline. +@@ -4959,7 +5297,11 @@ + switch_mm(mm, &init_mm, current); + finish_arch_post_lock_switch(); + } +- mmdrop(mm); ++ /* ++ * Defer the cleanup to an alive cpu. On RT we can neither ++ * call mmdrop() nor mmdrop_delayed() from here. ++ */ ++ per_cpu(idle_last_mm, smp_processor_id()) = mm; + } + + /* +@@ -5302,6 +5644,10 @@ + + case CPU_DEAD: + calc_load_migrate(rq); ++ if (per_cpu(idle_last_mm, cpu)) { ++ mmdrop(per_cpu(idle_last_mm, cpu)); ++ per_cpu(idle_last_mm, cpu) = NULL; ++ } + break; + #endif + } +@@ -7281,7 +7627,8 @@ + #ifdef CONFIG_DEBUG_ATOMIC_SLEEP + static inline int preempt_count_equals(int preempt_offset) + { +- int nested = (preempt_count() & ~PREEMPT_ACTIVE) + rcu_preempt_depth(); ++ int nested = (preempt_count() & ~PREEMPT_ACTIVE) + ++ sched_rcu_preempt_depth(); + + return (nested == preempt_offset); + } +diff -Nur linux-4.1.13.orig/kernel/sched/cputime.c linux-4.1.13/kernel/sched/cputime.c +--- linux-4.1.13.orig/kernel/sched/cputime.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/cputime.c 2015-11-29 09:23:09.645611789 +0100 +@@ -675,37 +675,45 @@ + + void vtime_account_system(struct task_struct *tsk) + { +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + __vtime_account_system(tsk); +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + + void vtime_gen_account_irq_exit(struct task_struct *tsk) + { +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + __vtime_account_system(tsk); + if (context_tracking_in_user()) + tsk->vtime_snap_whence = VTIME_USER; +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + + void vtime_account_user(struct task_struct *tsk) + { + cputime_t delta_cpu; + +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + delta_cpu = get_vtime_delta(tsk); + tsk->vtime_snap_whence = VTIME_SYS; + account_user_time(tsk, delta_cpu, cputime_to_scaled(delta_cpu)); +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + + void vtime_user_enter(struct task_struct *tsk) + { +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + __vtime_account_system(tsk); + tsk->vtime_snap_whence = VTIME_USER; +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + + void vtime_guest_enter(struct task_struct *tsk) +@@ -717,19 +725,23 @@ + * synchronization against the reader (task_gtime()) + * that can thus safely catch up with a tickless delta. + */ +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + __vtime_account_system(tsk); + current->flags |= PF_VCPU; +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + EXPORT_SYMBOL_GPL(vtime_guest_enter); + + void vtime_guest_exit(struct task_struct *tsk) + { +- write_seqlock(&tsk->vtime_seqlock); ++ raw_spin_lock(&tsk->vtime_lock); ++ write_seqcount_begin(&tsk->vtime_seq); + __vtime_account_system(tsk); + current->flags &= ~PF_VCPU; +- write_sequnlock(&tsk->vtime_seqlock); ++ write_seqcount_end(&tsk->vtime_seq); ++ raw_spin_unlock(&tsk->vtime_lock); + } + EXPORT_SYMBOL_GPL(vtime_guest_exit); + +@@ -742,24 +754,30 @@ + + void arch_vtime_task_switch(struct task_struct *prev) + { +- write_seqlock(&prev->vtime_seqlock); ++ raw_spin_lock(&prev->vtime_lock); ++ write_seqcount_begin(&prev->vtime_seq); + prev->vtime_snap_whence = VTIME_SLEEPING; +- write_sequnlock(&prev->vtime_seqlock); ++ write_seqcount_end(&prev->vtime_seq); ++ raw_spin_unlock(&prev->vtime_lock); + +- write_seqlock(¤t->vtime_seqlock); ++ raw_spin_lock(¤t->vtime_lock); ++ write_seqcount_begin(¤t->vtime_seq); + current->vtime_snap_whence = VTIME_SYS; + current->vtime_snap = sched_clock_cpu(smp_processor_id()); +- write_sequnlock(¤t->vtime_seqlock); ++ write_seqcount_end(¤t->vtime_seq); ++ raw_spin_unlock(¤t->vtime_lock); + } + + void vtime_init_idle(struct task_struct *t, int cpu) + { + unsigned long flags; + +- write_seqlock_irqsave(&t->vtime_seqlock, flags); ++ raw_spin_lock_irqsave(&t->vtime_lock, flags); ++ write_seqcount_begin(&t->vtime_seq); + t->vtime_snap_whence = VTIME_SYS; + t->vtime_snap = sched_clock_cpu(cpu); +- write_sequnlock_irqrestore(&t->vtime_seqlock, flags); ++ write_seqcount_end(&t->vtime_seq); ++ raw_spin_unlock_irqrestore(&t->vtime_lock, flags); + } + + cputime_t task_gtime(struct task_struct *t) +@@ -768,13 +786,13 @@ + cputime_t gtime; + + do { +- seq = read_seqbegin(&t->vtime_seqlock); ++ seq = read_seqcount_begin(&t->vtime_seq); + + gtime = t->gtime; + if (t->flags & PF_VCPU) + gtime += vtime_delta(t); + +- } while (read_seqretry(&t->vtime_seqlock, seq)); ++ } while (read_seqcount_retry(&t->vtime_seq, seq)); + + return gtime; + } +@@ -797,7 +815,7 @@ + *udelta = 0; + *sdelta = 0; + +- seq = read_seqbegin(&t->vtime_seqlock); ++ seq = read_seqcount_begin(&t->vtime_seq); + + if (u_dst) + *u_dst = *u_src; +@@ -821,7 +839,7 @@ + if (t->vtime_snap_whence == VTIME_SYS) + *sdelta = delta; + } +- } while (read_seqretry(&t->vtime_seqlock, seq)); ++ } while (read_seqcount_retry(&t->vtime_seq, seq)); + } + + +diff -Nur linux-4.1.13.orig/kernel/sched/deadline.c linux-4.1.13/kernel/sched/deadline.c +--- linux-4.1.13.orig/kernel/sched/deadline.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/deadline.c 2015-11-29 09:23:09.645611789 +0100 +@@ -637,6 +637,7 @@ + + hrtimer_init(timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + timer->function = dl_task_timer; ++ timer->irqsafe = 1; + } + + static +diff -Nur linux-4.1.13.orig/kernel/sched/debug.c linux-4.1.13/kernel/sched/debug.c +--- linux-4.1.13.orig/kernel/sched/debug.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/debug.c 2015-11-29 09:23:09.645611789 +0100 +@@ -260,6 +260,9 @@ + P(rt_throttled); + PN(rt_time); + PN(rt_runtime); ++#ifdef CONFIG_SMP ++ P(rt_nr_migratory); ++#endif + + #undef PN + #undef P +@@ -648,6 +651,10 @@ + #endif + P(policy); + P(prio); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ P(migrate_disable); ++#endif ++ P(nr_cpus_allowed); + #undef PN + #undef __PN + #undef P +diff -Nur linux-4.1.13.orig/kernel/sched/fair.c linux-4.1.13/kernel/sched/fair.c +--- linux-4.1.13.orig/kernel/sched/fair.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/fair.c 2015-11-29 09:23:09.645611789 +0100 +@@ -3201,7 +3201,7 @@ + ideal_runtime = sched_slice(cfs_rq, curr); + delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; + if (delta_exec > ideal_runtime) { +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + /* + * The current task ran long enough, ensure it doesn't get + * re-elected due to buddy favours. +@@ -3225,7 +3225,7 @@ + return; + + if (delta > ideal_runtime) +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + } + + static void +@@ -3366,7 +3366,7 @@ + * validating it and just reschedule. + */ + if (queued) { +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + return; + } + /* +@@ -3557,7 +3557,7 @@ + * hierarchy can be throttled + */ + if (!assign_cfs_rq_runtime(cfs_rq) && likely(cfs_rq->curr)) +- resched_curr(rq_of(cfs_rq)); ++ resched_curr_lazy(rq_of(cfs_rq)); + } + + static __always_inline +@@ -4180,7 +4180,7 @@ + + if (delta < 0) { + if (rq->curr == p) +- resched_curr(rq); ++ resched_curr_lazy(rq); + return; + } + hrtick_start(rq, delta); +@@ -5076,7 +5076,7 @@ + return; + + preempt: +- resched_curr(rq); ++ resched_curr_lazy(rq); + /* + * Only set the backward buddy when the current task is still + * on the rq. This can happen when a wakeup gets interleaved +@@ -7869,7 +7869,7 @@ + * 'current' within the tree based on its new key value. + */ + swap(curr->vruntime, se->vruntime); +- resched_curr(rq); ++ resched_curr_lazy(rq); + } + + se->vruntime -= cfs_rq->min_vruntime; +@@ -7894,7 +7894,7 @@ + */ + if (rq->curr == p) { + if (p->prio > oldprio) +- resched_curr(rq); ++ resched_curr_lazy(rq); + } else + check_preempt_curr(rq, p, 0); + } +diff -Nur linux-4.1.13.orig/kernel/sched/features.h linux-4.1.13/kernel/sched/features.h +--- linux-4.1.13.orig/kernel/sched/features.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/features.h 2015-11-29 09:23:09.645611789 +0100 +@@ -50,11 +50,19 @@ + */ + SCHED_FEAT(NONTASK_CAPACITY, true) + ++#ifdef CONFIG_PREEMPT_RT_FULL ++SCHED_FEAT(TTWU_QUEUE, false) ++# ifdef CONFIG_PREEMPT_LAZY ++SCHED_FEAT(PREEMPT_LAZY, true) ++# endif ++#else ++ + /* + * Queue remote wakeups on the target CPU and process them + * using the scheduler IPI. Reduces rq->lock contention/bounces. + */ + SCHED_FEAT(TTWU_QUEUE, true) ++#endif + + #ifdef HAVE_RT_PUSH_IPI + /* +diff -Nur linux-4.1.13.orig/kernel/sched/Makefile linux-4.1.13/kernel/sched/Makefile +--- linux-4.1.13.orig/kernel/sched/Makefile 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/Makefile 2015-11-29 09:23:09.641612055 +0100 +@@ -13,7 +13,7 @@ + + obj-y += core.o proc.o clock.o cputime.o + obj-y += idle_task.o fair.o rt.o deadline.o stop_task.o +-obj-y += wait.o completion.o idle.o ++obj-y += wait.o wait-simple.o work-simple.o completion.o idle.o + obj-$(CONFIG_SMP) += cpupri.o cpudeadline.o + obj-$(CONFIG_SCHED_AUTOGROUP) += auto_group.o + obj-$(CONFIG_SCHEDSTATS) += stats.o +diff -Nur linux-4.1.13.orig/kernel/sched/rt.c linux-4.1.13/kernel/sched/rt.c +--- linux-4.1.13.orig/kernel/sched/rt.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/rt.c 2015-11-29 09:23:09.645611789 +0100 +@@ -44,6 +44,7 @@ + + hrtimer_init(&rt_b->rt_period_timer, + CLOCK_MONOTONIC, HRTIMER_MODE_REL); ++ rt_b->rt_period_timer.irqsafe = 1; + rt_b->rt_period_timer.function = sched_rt_period_timer; + } + +@@ -89,6 +90,7 @@ + rt_rq->push_cpu = nr_cpu_ids; + raw_spin_lock_init(&rt_rq->push_lock); + init_irq_work(&rt_rq->push_work, push_irq_work_func); ++ rt_rq->push_work.flags |= IRQ_WORK_HARD_IRQ; + #endif + #endif /* CONFIG_SMP */ + /* We start is dequeued state, because no RT tasks are queued */ +diff -Nur linux-4.1.13.orig/kernel/sched/sched.h linux-4.1.13/kernel/sched/sched.h +--- linux-4.1.13.orig/kernel/sched/sched.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/sched/sched.h 2015-11-29 09:23:09.645611789 +0100 +@@ -1093,6 +1093,7 @@ + #define WF_SYNC 0x01 /* waker goes to sleep after wakeup */ + #define WF_FORK 0x02 /* child wakeup after fork */ + #define WF_MIGRATED 0x4 /* internal use, task got migrated */ ++#define WF_LOCK_SLEEPER 0x08 /* wakeup spinlock "sleeper" */ + + /* + * To aid in avoiding the subversion of "niceness" due to uneven distribution +@@ -1290,6 +1291,15 @@ + extern void resched_curr(struct rq *rq); + extern void resched_cpu(int cpu); + ++#ifdef CONFIG_PREEMPT_LAZY ++extern void resched_curr_lazy(struct rq *rq); ++#else ++static inline void resched_curr_lazy(struct rq *rq) ++{ ++ resched_curr(rq); ++} ++#endif ++ + extern struct rt_bandwidth def_rt_bandwidth; + extern void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime); + +diff -Nur linux-4.1.13.orig/kernel/sched/wait-simple.c linux-4.1.13/kernel/sched/wait-simple.c +--- linux-4.1.13.orig/kernel/sched/wait-simple.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/kernel/sched/wait-simple.c 2015-11-29 09:23:09.645611789 +0100 +@@ -0,0 +1,115 @@ ++/* ++ * Simple waitqueues without fancy flags and callbacks ++ * ++ * (C) 2011 Thomas Gleixner <tglx@linutronix.de> ++ * ++ * Based on kernel/wait.c ++ * ++ * For licencing details see kernel-base/COPYING ++ */ ++#include <linux/init.h> ++#include <linux/export.h> ++#include <linux/sched.h> ++#include <linux/wait-simple.h> ++ ++/* Adds w to head->list. Must be called with head->lock locked. */ ++static inline void __swait_enqueue(struct swait_head *head, struct swaiter *w) ++{ ++ list_add(&w->node, &head->list); ++ /* We can't let the condition leak before the setting of head */ ++ smp_mb(); ++} ++ ++/* Removes w from head->list. Must be called with head->lock locked. */ ++static inline void __swait_dequeue(struct swaiter *w) ++{ ++ list_del_init(&w->node); ++} ++ ++void __init_swait_head(struct swait_head *head, struct lock_class_key *key) ++{ ++ raw_spin_lock_init(&head->lock); ++ lockdep_set_class(&head->lock, key); ++ INIT_LIST_HEAD(&head->list); ++} ++EXPORT_SYMBOL(__init_swait_head); ++ ++void swait_prepare_locked(struct swait_head *head, struct swaiter *w) ++{ ++ w->task = current; ++ if (list_empty(&w->node)) ++ __swait_enqueue(head, w); ++} ++ ++void swait_prepare(struct swait_head *head, struct swaiter *w, int state) ++{ ++ unsigned long flags; ++ ++ raw_spin_lock_irqsave(&head->lock, flags); ++ swait_prepare_locked(head, w); ++ __set_current_state(state); ++ raw_spin_unlock_irqrestore(&head->lock, flags); ++} ++EXPORT_SYMBOL(swait_prepare); ++ ++void swait_finish_locked(struct swait_head *head, struct swaiter *w) ++{ ++ __set_current_state(TASK_RUNNING); ++ if (w->task) ++ __swait_dequeue(w); ++} ++ ++void swait_finish(struct swait_head *head, struct swaiter *w) ++{ ++ unsigned long flags; ++ ++ __set_current_state(TASK_RUNNING); ++ if (w->task) { ++ raw_spin_lock_irqsave(&head->lock, flags); ++ __swait_dequeue(w); ++ raw_spin_unlock_irqrestore(&head->lock, flags); ++ } ++} ++EXPORT_SYMBOL(swait_finish); ++ ++unsigned int ++__swait_wake_locked(struct swait_head *head, unsigned int state, unsigned int num) ++{ ++ struct swaiter *curr, *next; ++ int woken = 0; ++ ++ list_for_each_entry_safe(curr, next, &head->list, node) { ++ if (wake_up_state(curr->task, state)) { ++ __swait_dequeue(curr); ++ /* ++ * The waiting task can free the waiter as ++ * soon as curr->task = NULL is written, ++ * without taking any locks. A memory barrier ++ * is required here to prevent the following ++ * store to curr->task from getting ahead of ++ * the dequeue operation. ++ */ ++ smp_wmb(); ++ curr->task = NULL; ++ if (++woken == num) ++ break; ++ } ++ } ++ return woken; ++} ++ ++unsigned int ++__swait_wake(struct swait_head *head, unsigned int state, unsigned int num) ++{ ++ unsigned long flags; ++ int woken; ++ ++ if (!swaitqueue_active(head)) ++ return 0; ++ ++ raw_spin_lock_irqsave(&head->lock, flags); ++ woken = __swait_wake_locked(head, state, num); ++ raw_spin_unlock_irqrestore(&head->lock, flags); ++ return woken; ++} ++EXPORT_SYMBOL(__swait_wake); +diff -Nur linux-4.1.13.orig/kernel/sched/work-simple.c linux-4.1.13/kernel/sched/work-simple.c +--- linux-4.1.13.orig/kernel/sched/work-simple.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/kernel/sched/work-simple.c 2015-11-29 09:23:09.645611789 +0100 +@@ -0,0 +1,173 @@ ++/* ++ * Copyright (C) 2014 BMW Car IT GmbH, Daniel Wagner daniel.wagner@bmw-carit.de ++ * ++ * Provides a framework for enqueuing callbacks from irq context ++ * PREEMPT_RT_FULL safe. The callbacks are executed in kthread context. ++ */ ++ ++#include <linux/wait-simple.h> ++#include <linux/work-simple.h> ++#include <linux/kthread.h> ++#include <linux/slab.h> ++#include <linux/spinlock.h> ++#include <linux/export.h> ++ ++#define SWORK_EVENT_PENDING (1 << 0) ++ ++static DEFINE_MUTEX(worker_mutex); ++static struct sworker *glob_worker; ++ ++struct sworker { ++ struct list_head events; ++ struct swait_head wq; ++ ++ raw_spinlock_t lock; ++ ++ struct task_struct *task; ++ int refs; ++}; ++ ++static bool swork_readable(struct sworker *worker) ++{ ++ bool r; ++ ++ if (kthread_should_stop()) ++ return true; ++ ++ raw_spin_lock_irq(&worker->lock); ++ r = !list_empty(&worker->events); ++ raw_spin_unlock_irq(&worker->lock); ++ ++ return r; ++} ++ ++static int swork_kthread(void *arg) ++{ ++ struct sworker *worker = arg; ++ ++ for (;;) { ++ swait_event_interruptible(worker->wq, ++ swork_readable(worker)); ++ if (kthread_should_stop()) ++ break; ++ ++ raw_spin_lock_irq(&worker->lock); ++ while (!list_empty(&worker->events)) { ++ struct swork_event *sev; ++ ++ sev = list_first_entry(&worker->events, ++ struct swork_event, item); ++ list_del(&sev->item); ++ raw_spin_unlock_irq(&worker->lock); ++ ++ WARN_ON_ONCE(!test_and_clear_bit(SWORK_EVENT_PENDING, ++ &sev->flags)); ++ sev->func(sev); ++ raw_spin_lock_irq(&worker->lock); ++ } ++ raw_spin_unlock_irq(&worker->lock); ++ } ++ return 0; ++} ++ ++static struct sworker *swork_create(void) ++{ ++ struct sworker *worker; ++ ++ worker = kzalloc(sizeof(*worker), GFP_KERNEL); ++ if (!worker) ++ return ERR_PTR(-ENOMEM); ++ ++ INIT_LIST_HEAD(&worker->events); ++ raw_spin_lock_init(&worker->lock); ++ init_swait_head(&worker->wq); ++ ++ worker->task = kthread_run(swork_kthread, worker, "kswork"); ++ if (IS_ERR(worker->task)) { ++ kfree(worker); ++ return ERR_PTR(-ENOMEM); ++ } ++ ++ return worker; ++} ++ ++static void swork_destroy(struct sworker *worker) ++{ ++ kthread_stop(worker->task); ++ ++ WARN_ON(!list_empty(&worker->events)); ++ kfree(worker); ++} ++ ++/** ++ * swork_queue - queue swork ++ * ++ * Returns %false if @work was already on a queue, %true otherwise. ++ * ++ * The work is queued and processed on a random CPU ++ */ ++bool swork_queue(struct swork_event *sev) ++{ ++ unsigned long flags; ++ ++ if (test_and_set_bit(SWORK_EVENT_PENDING, &sev->flags)) ++ return false; ++ ++ raw_spin_lock_irqsave(&glob_worker->lock, flags); ++ list_add_tail(&sev->item, &glob_worker->events); ++ raw_spin_unlock_irqrestore(&glob_worker->lock, flags); ++ ++ swait_wake(&glob_worker->wq); ++ return true; ++} ++EXPORT_SYMBOL_GPL(swork_queue); ++ ++/** ++ * swork_get - get an instance of the sworker ++ * ++ * Returns an negative error code if the initialization if the worker did not ++ * work, %0 otherwise. ++ * ++ */ ++int swork_get(void) ++{ ++ struct sworker *worker; ++ ++ mutex_lock(&worker_mutex); ++ if (!glob_worker) { ++ worker = swork_create(); ++ if (IS_ERR(worker)) { ++ mutex_unlock(&worker_mutex); ++ return -ENOMEM; ++ } ++ ++ glob_worker = worker; ++ } ++ ++ glob_worker->refs++; ++ mutex_unlock(&worker_mutex); ++ ++ return 0; ++} ++EXPORT_SYMBOL_GPL(swork_get); ++ ++/** ++ * swork_put - puts an instance of the sworker ++ * ++ * Will destroy the sworker thread. This function must not be called until all ++ * queued events have been completed. ++ */ ++void swork_put(void) ++{ ++ mutex_lock(&worker_mutex); ++ ++ glob_worker->refs--; ++ if (glob_worker->refs > 0) ++ goto out; ++ ++ swork_destroy(glob_worker); ++ glob_worker = NULL; ++out: ++ mutex_unlock(&worker_mutex); ++} ++EXPORT_SYMBOL_GPL(swork_put); +diff -Nur linux-4.1.13.orig/kernel/signal.c linux-4.1.13/kernel/signal.c +--- linux-4.1.13.orig/kernel/signal.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/signal.c 2015-11-29 09:23:09.649611524 +0100 +@@ -14,6 +14,7 @@ + #include <linux/export.h> + #include <linux/init.h> + #include <linux/sched.h> ++#include <linux/sched/rt.h> + #include <linux/fs.h> + #include <linux/tty.h> + #include <linux/binfmts.h> +@@ -352,13 +353,45 @@ + return false; + } + ++#ifdef __HAVE_ARCH_CMPXCHG ++static inline struct sigqueue *get_task_cache(struct task_struct *t) ++{ ++ struct sigqueue *q = t->sigqueue_cache; ++ ++ if (cmpxchg(&t->sigqueue_cache, q, NULL) != q) ++ return NULL; ++ return q; ++} ++ ++static inline int put_task_cache(struct task_struct *t, struct sigqueue *q) ++{ ++ if (cmpxchg(&t->sigqueue_cache, NULL, q) == NULL) ++ return 0; ++ return 1; ++} ++ ++#else ++ ++static inline struct sigqueue *get_task_cache(struct task_struct *t) ++{ ++ return NULL; ++} ++ ++static inline int put_task_cache(struct task_struct *t, struct sigqueue *q) ++{ ++ return 1; ++} ++ ++#endif ++ + /* + * allocate a new signal queue record + * - this may be called without locks if and only if t == current, otherwise an + * appropriate lock must be held to stop the target task from exiting + */ + static struct sigqueue * +-__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit) ++__sigqueue_do_alloc(int sig, struct task_struct *t, gfp_t flags, ++ int override_rlimit, int fromslab) + { + struct sigqueue *q = NULL; + struct user_struct *user; +@@ -375,7 +408,10 @@ + if (override_rlimit || + atomic_read(&user->sigpending) <= + task_rlimit(t, RLIMIT_SIGPENDING)) { +- q = kmem_cache_alloc(sigqueue_cachep, flags); ++ if (!fromslab) ++ q = get_task_cache(t); ++ if (!q) ++ q = kmem_cache_alloc(sigqueue_cachep, flags); + } else { + print_dropped_signal(sig); + } +@@ -392,6 +428,13 @@ + return q; + } + ++static struct sigqueue * ++__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, ++ int override_rlimit) ++{ ++ return __sigqueue_do_alloc(sig, t, flags, override_rlimit, 0); ++} ++ + static void __sigqueue_free(struct sigqueue *q) + { + if (q->flags & SIGQUEUE_PREALLOC) +@@ -401,6 +444,21 @@ + kmem_cache_free(sigqueue_cachep, q); + } + ++static void sigqueue_free_current(struct sigqueue *q) ++{ ++ struct user_struct *up; ++ ++ if (q->flags & SIGQUEUE_PREALLOC) ++ return; ++ ++ up = q->user; ++ if (rt_prio(current->normal_prio) && !put_task_cache(current, q)) { ++ atomic_dec(&up->sigpending); ++ free_uid(up); ++ } else ++ __sigqueue_free(q); ++} ++ + void flush_sigqueue(struct sigpending *queue) + { + struct sigqueue *q; +@@ -414,6 +472,21 @@ + } + + /* ++ * Called from __exit_signal. Flush tsk->pending and ++ * tsk->sigqueue_cache ++ */ ++void flush_task_sigqueue(struct task_struct *tsk) ++{ ++ struct sigqueue *q; ++ ++ flush_sigqueue(&tsk->pending); ++ ++ q = get_task_cache(tsk); ++ if (q) ++ kmem_cache_free(sigqueue_cachep, q); ++} ++ ++/* + * Flush all pending signals for a task. + */ + void __flush_signals(struct task_struct *t) +@@ -565,7 +638,7 @@ + still_pending: + list_del_init(&first->list); + copy_siginfo(info, &first->info); +- __sigqueue_free(first); ++ sigqueue_free_current(first); + } else { + /* + * Ok, it wasn't in the queue. This must be +@@ -611,6 +684,8 @@ + { + int signr; + ++ WARN_ON_ONCE(tsk != current); ++ + /* We only dequeue private signals from ourselves, we don't let + * signalfd steal them + */ +@@ -1207,8 +1282,8 @@ + * We don't want to have recursive SIGSEGV's etc, for example, + * that is why we also clear SIGNAL_UNKILLABLE. + */ +-int +-force_sig_info(int sig, struct siginfo *info, struct task_struct *t) ++static int ++do_force_sig_info(int sig, struct siginfo *info, struct task_struct *t) + { + unsigned long int flags; + int ret, blocked, ignored; +@@ -1233,6 +1308,39 @@ + return ret; + } + ++int force_sig_info(int sig, struct siginfo *info, struct task_struct *t) ++{ ++/* ++ * On some archs, PREEMPT_RT has to delay sending a signal from a trap ++ * since it can not enable preemption, and the signal code's spin_locks ++ * turn into mutexes. Instead, it must set TIF_NOTIFY_RESUME which will ++ * send the signal on exit of the trap. ++ */ ++#ifdef ARCH_RT_DELAYS_SIGNAL_SEND ++ if (in_atomic()) { ++ if (WARN_ON_ONCE(t != current)) ++ return 0; ++ if (WARN_ON_ONCE(t->forced_info.si_signo)) ++ return 0; ++ ++ if (is_si_special(info)) { ++ WARN_ON_ONCE(info != SEND_SIG_PRIV); ++ t->forced_info.si_signo = sig; ++ t->forced_info.si_errno = 0; ++ t->forced_info.si_code = SI_KERNEL; ++ t->forced_info.si_pid = 0; ++ t->forced_info.si_uid = 0; ++ } else { ++ t->forced_info = *info; ++ } ++ ++ set_tsk_thread_flag(t, TIF_NOTIFY_RESUME); ++ return 0; ++ } ++#endif ++ return do_force_sig_info(sig, info, t); ++} ++ + /* + * Nuke all other threads in the group. + */ +@@ -1267,12 +1375,12 @@ + * Disable interrupts early to avoid deadlocks. + * See rcu_read_unlock() comment header for details. + */ +- local_irq_save(*flags); ++ local_irq_save_nort(*flags); + rcu_read_lock(); + sighand = rcu_dereference(tsk->sighand); + if (unlikely(sighand == NULL)) { + rcu_read_unlock(); +- local_irq_restore(*flags); ++ local_irq_restore_nort(*flags); + break; + } + /* +@@ -1293,7 +1401,7 @@ + } + spin_unlock(&sighand->siglock); + rcu_read_unlock(); +- local_irq_restore(*flags); ++ local_irq_restore_nort(*flags); + } + + return sighand; +@@ -1536,7 +1644,8 @@ + */ + struct sigqueue *sigqueue_alloc(void) + { +- struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0); ++ /* Preallocated sigqueue objects always from the slabcache ! */ ++ struct sigqueue *q = __sigqueue_do_alloc(-1, current, GFP_KERNEL, 0, 1); + + if (q) + q->flags |= SIGQUEUE_PREALLOC; +@@ -1897,15 +2006,7 @@ + if (gstop_done && ptrace_reparented(current)) + do_notify_parent_cldstop(current, false, why); + +- /* +- * Don't want to allow preemption here, because +- * sys_ptrace() needs this task to be inactive. +- * +- * XXX: implement read_unlock_no_resched(). +- */ +- preempt_disable(); + read_unlock(&tasklist_lock); +- preempt_enable_no_resched(); + freezable_schedule(); + } else { + /* +diff -Nur linux-4.1.13.orig/kernel/softirq.c linux-4.1.13/kernel/softirq.c +--- linux-4.1.13.orig/kernel/softirq.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/softirq.c 2015-11-29 09:23:09.649611524 +0100 +@@ -21,10 +21,12 @@ + #include <linux/freezer.h> + #include <linux/kthread.h> + #include <linux/rcupdate.h> ++#include <linux/delay.h> + #include <linux/ftrace.h> + #include <linux/smp.h> + #include <linux/smpboot.h> + #include <linux/tick.h> ++#include <linux/locallock.h> + #include <linux/irq.h> + + #define CREATE_TRACE_POINTS +@@ -62,6 +64,98 @@ + "TASKLET", "SCHED", "HRTIMER", "RCU" + }; + ++#ifdef CONFIG_NO_HZ_COMMON ++# ifdef CONFIG_PREEMPT_RT_FULL ++ ++struct softirq_runner { ++ struct task_struct *runner[NR_SOFTIRQS]; ++}; ++ ++static DEFINE_PER_CPU(struct softirq_runner, softirq_runners); ++ ++static inline void softirq_set_runner(unsigned int sirq) ++{ ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ ++ sr->runner[sirq] = current; ++} ++ ++static inline void softirq_clr_runner(unsigned int sirq) ++{ ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ ++ sr->runner[sirq] = NULL; ++} ++ ++/* ++ * On preempt-rt a softirq running context might be blocked on a ++ * lock. There might be no other runnable task on this CPU because the ++ * lock owner runs on some other CPU. So we have to go into idle with ++ * the pending bit set. Therefor we need to check this otherwise we ++ * warn about false positives which confuses users and defeats the ++ * whole purpose of this test. ++ * ++ * This code is called with interrupts disabled. ++ */ ++void softirq_check_pending_idle(void) ++{ ++ static int rate_limit; ++ struct softirq_runner *sr = this_cpu_ptr(&softirq_runners); ++ u32 warnpending; ++ int i; ++ ++ if (rate_limit >= 10) ++ return; ++ ++ warnpending = local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK; ++ for (i = 0; i < NR_SOFTIRQS; i++) { ++ struct task_struct *tsk = sr->runner[i]; ++ ++ /* ++ * The wakeup code in rtmutex.c wakes up the task ++ * _before_ it sets pi_blocked_on to NULL under ++ * tsk->pi_lock. So we need to check for both: state ++ * and pi_blocked_on. ++ */ ++ if (tsk) { ++ raw_spin_lock(&tsk->pi_lock); ++ if (tsk->pi_blocked_on || tsk->state == TASK_RUNNING) { ++ /* Clear all bits pending in that task */ ++ warnpending &= ~(tsk->softirqs_raised); ++ warnpending &= ~(1 << i); ++ } ++ raw_spin_unlock(&tsk->pi_lock); ++ } ++ } ++ ++ if (warnpending) { ++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", ++ warnpending); ++ rate_limit++; ++ } ++} ++# else ++/* ++ * On !PREEMPT_RT we just printk rate limited: ++ */ ++void softirq_check_pending_idle(void) ++{ ++ static int rate_limit; ++ ++ if (rate_limit < 10 && ++ (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { ++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n", ++ local_softirq_pending()); ++ rate_limit++; ++ } ++} ++# endif ++ ++#else /* !CONFIG_NO_HZ_COMMON */ ++static inline void softirq_set_runner(unsigned int sirq) { } ++static inline void softirq_clr_runner(unsigned int sirq) { } ++#endif ++ + /* + * we cannot loop indefinitely here to avoid userspace starvation, + * but we also don't want to introduce a worst case 1/HZ latency +@@ -77,6 +171,68 @@ + wake_up_process(tsk); + } + ++static void handle_softirq(unsigned int vec_nr) ++{ ++ struct softirq_action *h = softirq_vec + vec_nr; ++ int prev_count; ++ ++ prev_count = preempt_count(); ++ ++ kstat_incr_softirqs_this_cpu(vec_nr); ++ ++ trace_softirq_entry(vec_nr); ++ h->action(h); ++ trace_softirq_exit(vec_nr); ++ if (unlikely(prev_count != preempt_count())) { ++ pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", ++ vec_nr, softirq_to_name[vec_nr], h->action, ++ prev_count, preempt_count()); ++ preempt_count_set(prev_count); ++ } ++} ++ ++#ifndef CONFIG_PREEMPT_RT_FULL ++static inline int ksoftirqd_softirq_pending(void) ++{ ++ return local_softirq_pending(); ++} ++ ++static void handle_pending_softirqs(u32 pending) ++{ ++ struct softirq_action *h = softirq_vec; ++ int softirq_bit; ++ ++ local_irq_enable(); ++ ++ h = softirq_vec; ++ ++ while ((softirq_bit = ffs(pending))) { ++ unsigned int vec_nr; ++ ++ h += softirq_bit - 1; ++ vec_nr = h - softirq_vec; ++ handle_softirq(vec_nr); ++ ++ h++; ++ pending >>= softirq_bit; ++ } ++ ++ rcu_bh_qs(); ++ local_irq_disable(); ++} ++ ++static void run_ksoftirqd(unsigned int cpu) ++{ ++ local_irq_disable(); ++ if (ksoftirqd_softirq_pending()) { ++ __do_softirq(); ++ local_irq_enable(); ++ cond_resched_rcu_qs(); ++ return; ++ } ++ local_irq_enable(); ++} ++ + /* + * preempt_count and SOFTIRQ_OFFSET usage: + * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving +@@ -232,10 +388,8 @@ + unsigned long end = jiffies + MAX_SOFTIRQ_TIME; + unsigned long old_flags = current->flags; + int max_restart = MAX_SOFTIRQ_RESTART; +- struct softirq_action *h; + bool in_hardirq; + __u32 pending; +- int softirq_bit; + + /* + * Mask out PF_MEMALLOC s current task context is borrowed for the +@@ -254,36 +408,7 @@ + /* Reset the pending bitmask before enabling irqs */ + set_softirq_pending(0); + +- local_irq_enable(); +- +- h = softirq_vec; +- +- while ((softirq_bit = ffs(pending))) { +- unsigned int vec_nr; +- int prev_count; +- +- h += softirq_bit - 1; +- +- vec_nr = h - softirq_vec; +- prev_count = preempt_count(); +- +- kstat_incr_softirqs_this_cpu(vec_nr); +- +- trace_softirq_entry(vec_nr); +- h->action(h); +- trace_softirq_exit(vec_nr); +- if (unlikely(prev_count != preempt_count())) { +- pr_err("huh, entered softirq %u %s %p with preempt_count %08x, exited with %08x?\n", +- vec_nr, softirq_to_name[vec_nr], h->action, +- prev_count, preempt_count()); +- preempt_count_set(prev_count); +- } +- h++; +- pending >>= softirq_bit; +- } +- +- rcu_bh_qs(); +- local_irq_disable(); ++ handle_pending_softirqs(pending); + + pending = local_softirq_pending(); + if (pending) { +@@ -320,6 +445,255 @@ + } + + /* ++ * This function must run with irqs disabled! ++ */ ++void raise_softirq_irqoff(unsigned int nr) ++{ ++ __raise_softirq_irqoff(nr); ++ ++ /* ++ * If we're in an interrupt or softirq, we're done ++ * (this also catches softirq-disabled code). We will ++ * actually run the softirq once we return from ++ * the irq or softirq. ++ * ++ * Otherwise we wake up ksoftirqd to make sure we ++ * schedule the softirq soon. ++ */ ++ if (!in_interrupt()) ++ wakeup_softirqd(); ++} ++ ++void __raise_softirq_irqoff(unsigned int nr) ++{ ++ trace_softirq_raise(nr); ++ or_softirq_pending(1UL << nr); ++} ++ ++static inline void local_bh_disable_nort(void) { local_bh_disable(); } ++static inline void _local_bh_enable_nort(void) { _local_bh_enable(); } ++static void ksoftirqd_set_sched_params(unsigned int cpu) { } ++static void ksoftirqd_clr_sched_params(unsigned int cpu, bool online) { } ++ ++#else /* !PREEMPT_RT_FULL */ ++ ++/* ++ * On RT we serialize softirq execution with a cpu local lock per softirq ++ */ ++static DEFINE_PER_CPU(struct local_irq_lock [NR_SOFTIRQS], local_softirq_locks); ++ ++void __init softirq_early_init(void) ++{ ++ int i; ++ ++ for (i = 0; i < NR_SOFTIRQS; i++) ++ local_irq_lock_init(local_softirq_locks[i]); ++} ++ ++static void lock_softirq(int which) ++{ ++ local_lock(local_softirq_locks[which]); ++} ++ ++static void unlock_softirq(int which) ++{ ++ local_unlock(local_softirq_locks[which]); ++} ++ ++static void do_single_softirq(int which) ++{ ++ unsigned long old_flags = current->flags; ++ ++ current->flags &= ~PF_MEMALLOC; ++ vtime_account_irq_enter(current); ++ current->flags |= PF_IN_SOFTIRQ; ++ lockdep_softirq_enter(); ++ local_irq_enable(); ++ handle_softirq(which); ++ local_irq_disable(); ++ lockdep_softirq_exit(); ++ current->flags &= ~PF_IN_SOFTIRQ; ++ vtime_account_irq_enter(current); ++ tsk_restore_flags(current, old_flags, PF_MEMALLOC); ++} ++ ++/* ++ * Called with interrupts disabled. Process softirqs which were raised ++ * in current context (or on behalf of ksoftirqd). ++ */ ++static void do_current_softirqs(void) ++{ ++ while (current->softirqs_raised) { ++ int i = __ffs(current->softirqs_raised); ++ unsigned int pending, mask = (1U << i); ++ ++ current->softirqs_raised &= ~mask; ++ local_irq_enable(); ++ ++ /* ++ * If the lock is contended, we boost the owner to ++ * process the softirq or leave the critical section ++ * now. ++ */ ++ lock_softirq(i); ++ local_irq_disable(); ++ softirq_set_runner(i); ++ /* ++ * Check with the local_softirq_pending() bits, ++ * whether we need to process this still or if someone ++ * else took care of it. ++ */ ++ pending = local_softirq_pending(); ++ if (pending & mask) { ++ set_softirq_pending(pending & ~mask); ++ do_single_softirq(i); ++ } ++ softirq_clr_runner(i); ++ unlock_softirq(i); ++ WARN_ON(current->softirq_nestcnt != 1); ++ } ++} ++ ++void __local_bh_disable(void) ++{ ++ if (++current->softirq_nestcnt == 1) ++ migrate_disable(); ++} ++EXPORT_SYMBOL(__local_bh_disable); ++ ++void __local_bh_enable(void) ++{ ++ if (WARN_ON(current->softirq_nestcnt == 0)) ++ return; ++ ++ local_irq_disable(); ++ if (current->softirq_nestcnt == 1 && current->softirqs_raised) ++ do_current_softirqs(); ++ local_irq_enable(); ++ ++ if (--current->softirq_nestcnt == 0) ++ migrate_enable(); ++} ++EXPORT_SYMBOL(__local_bh_enable); ++ ++void _local_bh_enable(void) ++{ ++ if (WARN_ON(current->softirq_nestcnt == 0)) ++ return; ++ if (--current->softirq_nestcnt == 0) ++ migrate_enable(); ++} ++EXPORT_SYMBOL(_local_bh_enable); ++ ++int in_serving_softirq(void) ++{ ++ return current->flags & PF_IN_SOFTIRQ; ++} ++EXPORT_SYMBOL(in_serving_softirq); ++ ++/* Called with preemption disabled */ ++static void run_ksoftirqd(unsigned int cpu) ++{ ++ local_irq_disable(); ++ current->softirq_nestcnt++; ++ ++ do_current_softirqs(); ++ current->softirq_nestcnt--; ++ rcu_note_context_switch(); ++ local_irq_enable(); ++} ++ ++/* ++ * Called from netif_rx_ni(). Preemption enabled, but migration ++ * disabled. So the cpu can't go away under us. ++ */ ++void thread_do_softirq(void) ++{ ++ if (!in_serving_softirq() && current->softirqs_raised) { ++ current->softirq_nestcnt++; ++ do_current_softirqs(); ++ current->softirq_nestcnt--; ++ } ++} ++ ++static void do_raise_softirq_irqoff(unsigned int nr) ++{ ++ trace_softirq_raise(nr); ++ or_softirq_pending(1UL << nr); ++ ++ /* ++ * If we are not in a hard interrupt and inside a bh disabled ++ * region, we simply raise the flag on current. local_bh_enable() ++ * will make sure that the softirq is executed. Otherwise we ++ * delegate it to ksoftirqd. ++ */ ++ if (!in_irq() && current->softirq_nestcnt) ++ current->softirqs_raised |= (1U << nr); ++ else if (__this_cpu_read(ksoftirqd)) ++ __this_cpu_read(ksoftirqd)->softirqs_raised |= (1U << nr); ++} ++ ++void __raise_softirq_irqoff(unsigned int nr) ++{ ++ do_raise_softirq_irqoff(nr); ++ if (!in_irq() && !current->softirq_nestcnt) ++ wakeup_softirqd(); ++} ++ ++/* ++ * This function must run with irqs disabled! ++ */ ++void raise_softirq_irqoff(unsigned int nr) ++{ ++ do_raise_softirq_irqoff(nr); ++ ++ /* ++ * If we're in an hard interrupt we let irq return code deal ++ * with the wakeup of ksoftirqd. ++ */ ++ if (in_irq()) ++ return; ++ /* ++ * If we are in thread context but outside of a bh disabled ++ * region, we need to wake ksoftirqd as well. ++ * ++ * CHECKME: Some of the places which do that could be wrapped ++ * into local_bh_disable/enable pairs. Though it's unclear ++ * whether this is worth the effort. To find those places just ++ * raise a WARN() if the condition is met. ++ */ ++ if (!current->softirq_nestcnt) ++ wakeup_softirqd(); ++} ++ ++static inline int ksoftirqd_softirq_pending(void) ++{ ++ return current->softirqs_raised; ++} ++ ++static inline void local_bh_disable_nort(void) { } ++static inline void _local_bh_enable_nort(void) { } ++ ++static inline void ksoftirqd_set_sched_params(unsigned int cpu) ++{ ++ struct sched_param param = { .sched_priority = 1 }; ++ ++ sched_setscheduler(current, SCHED_FIFO, ¶m); ++ /* Take over all pending softirqs when starting */ ++ local_irq_disable(); ++ current->softirqs_raised = local_softirq_pending(); ++ local_irq_enable(); ++} ++ ++static inline void ksoftirqd_clr_sched_params(unsigned int cpu, bool online) ++{ ++ struct sched_param param = { .sched_priority = 0 }; ++ ++ sched_setscheduler(current, SCHED_NORMAL, ¶m); ++} ++ ++#endif /* PREEMPT_RT_FULL */ ++/* + * Enter an interrupt context. + */ + void irq_enter(void) +@@ -330,9 +704,9 @@ + * Prevent raise_softirq from needlessly waking up ksoftirqd + * here, as softirq will be serviced on return from interrupt. + */ +- local_bh_disable(); ++ local_bh_disable_nort(); + tick_irq_enter(); +- _local_bh_enable(); ++ _local_bh_enable_nort(); + } + + __irq_enter(); +@@ -340,6 +714,7 @@ + + static inline void invoke_softirq(void) + { ++#ifndef CONFIG_PREEMPT_RT_FULL + if (!force_irqthreads) { + #ifdef CONFIG_HAVE_IRQ_EXIT_ON_IRQ_STACK + /* +@@ -359,6 +734,15 @@ + } else { + wakeup_softirqd(); + } ++#else /* PREEMPT_RT_FULL */ ++ unsigned long flags; ++ ++ local_irq_save(flags); ++ if (__this_cpu_read(ksoftirqd) && ++ __this_cpu_read(ksoftirqd)->softirqs_raised) ++ wakeup_softirqd(); ++ local_irq_restore(flags); ++#endif + } + + static inline void tick_irq_exit(void) +@@ -395,26 +779,6 @@ + trace_hardirq_exit(); /* must be last! */ + } + +-/* +- * This function must run with irqs disabled! +- */ +-inline void raise_softirq_irqoff(unsigned int nr) +-{ +- __raise_softirq_irqoff(nr); +- +- /* +- * If we're in an interrupt or softirq, we're done +- * (this also catches softirq-disabled code). We will +- * actually run the softirq once we return from +- * the irq or softirq. +- * +- * Otherwise we wake up ksoftirqd to make sure we +- * schedule the softirq soon. +- */ +- if (!in_interrupt()) +- wakeup_softirqd(); +-} +- + void raise_softirq(unsigned int nr) + { + unsigned long flags; +@@ -424,12 +788,6 @@ + local_irq_restore(flags); + } + +-void __raise_softirq_irqoff(unsigned int nr) +-{ +- trace_softirq_raise(nr); +- or_softirq_pending(1UL << nr); +-} +- + void open_softirq(int nr, void (*action)(struct softirq_action *)) + { + softirq_vec[nr].action = action; +@@ -446,15 +804,45 @@ + static DEFINE_PER_CPU(struct tasklet_head, tasklet_vec); + static DEFINE_PER_CPU(struct tasklet_head, tasklet_hi_vec); + ++static void inline ++__tasklet_common_schedule(struct tasklet_struct *t, struct tasklet_head *head, unsigned int nr) ++{ ++ if (tasklet_trylock(t)) { ++again: ++ /* We may have been preempted before tasklet_trylock ++ * and __tasklet_action may have already run. ++ * So double check the sched bit while the takslet ++ * is locked before adding it to the list. ++ */ ++ if (test_bit(TASKLET_STATE_SCHED, &t->state)) { ++ t->next = NULL; ++ *head->tail = t; ++ head->tail = &(t->next); ++ raise_softirq_irqoff(nr); ++ tasklet_unlock(t); ++ } else { ++ /* This is subtle. If we hit the corner case above ++ * It is possible that we get preempted right here, ++ * and another task has successfully called ++ * tasklet_schedule(), then this function, and ++ * failed on the trylock. Thus we must be sure ++ * before releasing the tasklet lock, that the ++ * SCHED_BIT is clear. Otherwise the tasklet ++ * may get its SCHED_BIT set, but not added to the ++ * list ++ */ ++ if (!tasklet_tryunlock(t)) ++ goto again; ++ } ++ } ++} ++ + void __tasklet_schedule(struct tasklet_struct *t) + { + unsigned long flags; + + local_irq_save(flags); +- t->next = NULL; +- *__this_cpu_read(tasklet_vec.tail) = t; +- __this_cpu_write(tasklet_vec.tail, &(t->next)); +- raise_softirq_irqoff(TASKLET_SOFTIRQ); ++ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_vec), TASKLET_SOFTIRQ); + local_irq_restore(flags); + } + EXPORT_SYMBOL(__tasklet_schedule); +@@ -464,10 +852,7 @@ + unsigned long flags; + + local_irq_save(flags); +- t->next = NULL; +- *__this_cpu_read(tasklet_hi_vec.tail) = t; +- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); +- raise_softirq_irqoff(HI_SOFTIRQ); ++ __tasklet_common_schedule(t, this_cpu_ptr(&tasklet_hi_vec), HI_SOFTIRQ); + local_irq_restore(flags); + } + EXPORT_SYMBOL(__tasklet_hi_schedule); +@@ -476,82 +861,122 @@ + { + BUG_ON(!irqs_disabled()); + +- t->next = __this_cpu_read(tasklet_hi_vec.head); +- __this_cpu_write(tasklet_hi_vec.head, t); +- __raise_softirq_irqoff(HI_SOFTIRQ); ++ __tasklet_hi_schedule(t); + } + EXPORT_SYMBOL(__tasklet_hi_schedule_first); + +-static void tasklet_action(struct softirq_action *a) ++void tasklet_enable(struct tasklet_struct *t) + { +- struct tasklet_struct *list; ++ if (!atomic_dec_and_test(&t->count)) ++ return; ++ if (test_and_clear_bit(TASKLET_STATE_PENDING, &t->state)) ++ tasklet_schedule(t); ++} ++EXPORT_SYMBOL(tasklet_enable); + +- local_irq_disable(); +- list = __this_cpu_read(tasklet_vec.head); +- __this_cpu_write(tasklet_vec.head, NULL); +- __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); +- local_irq_enable(); ++static void __tasklet_action(struct softirq_action *a, ++ struct tasklet_struct *list) ++{ ++ int loops = 1000000; + + while (list) { + struct tasklet_struct *t = list; + + list = list->next; + +- if (tasklet_trylock(t)) { +- if (!atomic_read(&t->count)) { +- if (!test_and_clear_bit(TASKLET_STATE_SCHED, +- &t->state)) +- BUG(); +- t->func(t->data); +- tasklet_unlock(t); +- continue; +- } +- tasklet_unlock(t); ++ /* ++ * Should always succeed - after a tasklist got on the ++ * list (after getting the SCHED bit set from 0 to 1), ++ * nothing but the tasklet softirq it got queued to can ++ * lock it: ++ */ ++ if (!tasklet_trylock(t)) { ++ WARN_ON(1); ++ continue; + } + +- local_irq_disable(); + t->next = NULL; +- *__this_cpu_read(tasklet_vec.tail) = t; +- __this_cpu_write(tasklet_vec.tail, &(t->next)); +- __raise_softirq_irqoff(TASKLET_SOFTIRQ); +- local_irq_enable(); ++ ++ /* ++ * If we cannot handle the tasklet because it's disabled, ++ * mark it as pending. tasklet_enable() will later ++ * re-schedule the tasklet. ++ */ ++ if (unlikely(atomic_read(&t->count))) { ++out_disabled: ++ /* implicit unlock: */ ++ wmb(); ++ t->state = TASKLET_STATEF_PENDING; ++ continue; ++ } ++ ++ /* ++ * After this point on the tasklet might be rescheduled ++ * on another CPU, but it can only be added to another ++ * CPU's tasklet list if we unlock the tasklet (which we ++ * dont do yet). ++ */ ++ if (!test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) ++ WARN_ON(1); ++ ++again: ++ t->func(t->data); ++ ++ /* ++ * Try to unlock the tasklet. We must use cmpxchg, because ++ * another CPU might have scheduled or disabled the tasklet. ++ * We only allow the STATE_RUN -> 0 transition here. ++ */ ++ while (!tasklet_tryunlock(t)) { ++ /* ++ * If it got disabled meanwhile, bail out: ++ */ ++ if (atomic_read(&t->count)) ++ goto out_disabled; ++ /* ++ * If it got scheduled meanwhile, re-execute ++ * the tasklet function: ++ */ ++ if (test_and_clear_bit(TASKLET_STATE_SCHED, &t->state)) ++ goto again; ++ if (!--loops) { ++ printk("hm, tasklet state: %08lx\n", t->state); ++ WARN_ON(1); ++ tasklet_unlock(t); ++ break; ++ } ++ } + } + } + ++static void tasklet_action(struct softirq_action *a) ++{ ++ struct tasklet_struct *list; ++ ++ local_irq_disable(); ++ ++ list = __this_cpu_read(tasklet_vec.head); ++ __this_cpu_write(tasklet_vec.head, NULL); ++ __this_cpu_write(tasklet_vec.tail, this_cpu_ptr(&tasklet_vec.head)); ++ ++ local_irq_enable(); ++ ++ __tasklet_action(a, list); ++} ++ + static void tasklet_hi_action(struct softirq_action *a) + { + struct tasklet_struct *list; + + local_irq_disable(); ++ + list = __this_cpu_read(tasklet_hi_vec.head); + __this_cpu_write(tasklet_hi_vec.head, NULL); + __this_cpu_write(tasklet_hi_vec.tail, this_cpu_ptr(&tasklet_hi_vec.head)); +- local_irq_enable(); + +- while (list) { +- struct tasklet_struct *t = list; +- +- list = list->next; +- +- if (tasklet_trylock(t)) { +- if (!atomic_read(&t->count)) { +- if (!test_and_clear_bit(TASKLET_STATE_SCHED, +- &t->state)) +- BUG(); +- t->func(t->data); +- tasklet_unlock(t); +- continue; +- } +- tasklet_unlock(t); +- } ++ local_irq_enable(); + +- local_irq_disable(); +- t->next = NULL; +- *__this_cpu_read(tasklet_hi_vec.tail) = t; +- __this_cpu_write(tasklet_hi_vec.tail, &(t->next)); +- __raise_softirq_irqoff(HI_SOFTIRQ); +- local_irq_enable(); +- } ++ __tasklet_action(a, list); + } + + void tasklet_init(struct tasklet_struct *t, +@@ -572,7 +997,7 @@ + + while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { + do { +- yield(); ++ msleep(1); + } while (test_bit(TASKLET_STATE_SCHED, &t->state)); + } + tasklet_unlock_wait(t); +@@ -646,25 +1071,26 @@ + open_softirq(HI_SOFTIRQ, tasklet_hi_action); + } + +-static int ksoftirqd_should_run(unsigned int cpu) ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) ++void tasklet_unlock_wait(struct tasklet_struct *t) + { +- return local_softirq_pending(); +-} +- +-static void run_ksoftirqd(unsigned int cpu) +-{ +- local_irq_disable(); +- if (local_softirq_pending()) { ++ while (test_bit(TASKLET_STATE_RUN, &(t)->state)) { + /* +- * We can safely run softirq on inline stack, as we are not deep +- * in the task stack here. ++ * Hack for now to avoid this busy-loop: + */ +- __do_softirq(); +- local_irq_enable(); +- cond_resched_rcu_qs(); +- return; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ msleep(1); ++#else ++ barrier(); ++#endif + } +- local_irq_enable(); ++} ++EXPORT_SYMBOL(tasklet_unlock_wait); ++#endif ++ ++static int ksoftirqd_should_run(unsigned int cpu) ++{ ++ return ksoftirqd_softirq_pending(); + } + + #ifdef CONFIG_HOTPLUG_CPU +@@ -746,6 +1172,8 @@ + + static struct smp_hotplug_thread softirq_threads = { + .store = &ksoftirqd, ++ .setup = ksoftirqd_set_sched_params, ++ .cleanup = ksoftirqd_clr_sched_params, + .thread_should_run = ksoftirqd_should_run, + .thread_fn = run_ksoftirqd, + .thread_comm = "ksoftirqd/%u", +diff -Nur linux-4.1.13.orig/kernel/stop_machine.c linux-4.1.13/kernel/stop_machine.c +--- linux-4.1.13.orig/kernel/stop_machine.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/stop_machine.c 2015-11-29 09:23:09.649611524 +0100 +@@ -30,12 +30,12 @@ + atomic_t nr_todo; /* nr left to execute */ + bool executed; /* actually executed? */ + int ret; /* collected return value */ +- struct completion completion; /* fired if nr_todo reaches 0 */ ++ struct task_struct *waiter; /* woken when nr_todo reaches 0 */ + }; + + /* the actual stopper, one per every possible cpu, enabled on online cpus */ + struct cpu_stopper { +- spinlock_t lock; ++ raw_spinlock_t lock; + bool enabled; /* is this stopper enabled? */ + struct list_head works; /* list of pending works */ + }; +@@ -56,7 +56,7 @@ + { + memset(done, 0, sizeof(*done)); + atomic_set(&done->nr_todo, nr_todo); +- init_completion(&done->completion); ++ done->waiter = current; + } + + /* signal completion unless @done is NULL */ +@@ -65,8 +65,10 @@ + if (done) { + if (executed) + done->executed = true; +- if (atomic_dec_and_test(&done->nr_todo)) +- complete(&done->completion); ++ if (atomic_dec_and_test(&done->nr_todo)) { ++ wake_up_process(done->waiter); ++ done->waiter = NULL; ++ } + } + } + +@@ -78,7 +80,7 @@ + + unsigned long flags; + +- spin_lock_irqsave(&stopper->lock, flags); ++ raw_spin_lock_irqsave(&stopper->lock, flags); + + if (stopper->enabled) { + list_add_tail(&work->list, &stopper->works); +@@ -86,7 +88,23 @@ + } else + cpu_stop_signal_done(work->done, false); + +- spin_unlock_irqrestore(&stopper->lock, flags); ++ raw_spin_unlock_irqrestore(&stopper->lock, flags); ++} ++ ++static void wait_for_stop_done(struct cpu_stop_done *done) ++{ ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ while (atomic_read(&done->nr_todo)) { ++ schedule(); ++ set_current_state(TASK_UNINTERRUPTIBLE); ++ } ++ /* ++ * We need to wait until cpu_stop_signal_done() has cleared ++ * done->waiter. ++ */ ++ while (done->waiter) ++ cpu_relax(); ++ set_current_state(TASK_RUNNING); + } + + /** +@@ -120,7 +138,7 @@ + + cpu_stop_init_done(&done, 1); + cpu_stop_queue_work(cpu, &work); +- wait_for_completion(&done.completion); ++ wait_for_stop_done(&done); + return done.executed ? done.ret : -ENOENT; + } + +@@ -248,7 +266,7 @@ + struct irq_cpu_stop_queue_work_info call_args; + struct multi_stop_data msdata; + +- preempt_disable(); ++ preempt_disable_nort(); + msdata = (struct multi_stop_data){ + .fn = fn, + .data = arg, +@@ -281,7 +299,7 @@ + * This relies on the stopper workqueues to be FIFO. + */ + if (!cpu_active(cpu1) || !cpu_active(cpu2)) { +- preempt_enable(); ++ preempt_enable_nort(); + return -ENOENT; + } + +@@ -295,9 +313,9 @@ + &irq_cpu_stop_queue_work, + &call_args, 1); + lg_local_unlock(&stop_cpus_lock); +- preempt_enable(); ++ preempt_enable_nort(); + +- wait_for_completion(&done.completion); ++ wait_for_stop_done(&done); + + return done.executed ? done.ret : -ENOENT; + } +@@ -329,7 +347,7 @@ + + static void queue_stop_cpus_work(const struct cpumask *cpumask, + cpu_stop_fn_t fn, void *arg, +- struct cpu_stop_done *done) ++ struct cpu_stop_done *done, bool inactive) + { + struct cpu_stop_work *work; + unsigned int cpu; +@@ -343,11 +361,13 @@ + } + + /* +- * Disable preemption while queueing to avoid getting +- * preempted by a stopper which might wait for other stoppers +- * to enter @fn which can lead to deadlock. ++ * Make sure that all work is queued on all cpus before ++ * any of the cpus can execute it. + */ +- lg_global_lock(&stop_cpus_lock); ++ if (!inactive) ++ lg_global_lock(&stop_cpus_lock); ++ else ++ lg_global_trylock_relax(&stop_cpus_lock); + for_each_cpu(cpu, cpumask) + cpu_stop_queue_work(cpu, &per_cpu(stop_cpus_work, cpu)); + lg_global_unlock(&stop_cpus_lock); +@@ -359,8 +379,8 @@ + struct cpu_stop_done done; + + cpu_stop_init_done(&done, cpumask_weight(cpumask)); +- queue_stop_cpus_work(cpumask, fn, arg, &done); +- wait_for_completion(&done.completion); ++ queue_stop_cpus_work(cpumask, fn, arg, &done, false); ++ wait_for_stop_done(&done); + return done.executed ? done.ret : -ENOENT; + } + +@@ -439,9 +459,9 @@ + unsigned long flags; + int run; + +- spin_lock_irqsave(&stopper->lock, flags); ++ raw_spin_lock_irqsave(&stopper->lock, flags); + run = !list_empty(&stopper->works); +- spin_unlock_irqrestore(&stopper->lock, flags); ++ raw_spin_unlock_irqrestore(&stopper->lock, flags); + return run; + } + +@@ -453,13 +473,13 @@ + + repeat: + work = NULL; +- spin_lock_irq(&stopper->lock); ++ raw_spin_lock_irq(&stopper->lock); + if (!list_empty(&stopper->works)) { + work = list_first_entry(&stopper->works, + struct cpu_stop_work, list); + list_del_init(&work->list); + } +- spin_unlock_irq(&stopper->lock); ++ raw_spin_unlock_irq(&stopper->lock); + + if (work) { + cpu_stop_fn_t fn = work->fn; +@@ -467,6 +487,16 @@ + struct cpu_stop_done *done = work->done; + char ksym_buf[KSYM_NAME_LEN] __maybe_unused; + ++ /* ++ * Wait until the stopper finished scheduling on all ++ * cpus ++ */ ++ lg_global_lock(&stop_cpus_lock); ++ /* ++ * Let other cpu threads continue as well ++ */ ++ lg_global_unlock(&stop_cpus_lock); ++ + /* cpu stop callbacks are not allowed to sleep */ + preempt_disable(); + +@@ -481,7 +511,13 @@ + kallsyms_lookup((unsigned long)fn, NULL, NULL, NULL, + ksym_buf), arg); + ++ /* ++ * Make sure that the wakeup and setting done->waiter ++ * to NULL is atomic. ++ */ ++ local_irq_disable(); + cpu_stop_signal_done(done, true); ++ local_irq_enable(); + goto repeat; + } + } +@@ -500,20 +536,20 @@ + unsigned long flags; + + /* drain remaining works */ +- spin_lock_irqsave(&stopper->lock, flags); ++ raw_spin_lock_irqsave(&stopper->lock, flags); + list_for_each_entry(work, &stopper->works, list) + cpu_stop_signal_done(work->done, false); + stopper->enabled = false; +- spin_unlock_irqrestore(&stopper->lock, flags); ++ raw_spin_unlock_irqrestore(&stopper->lock, flags); + } + + static void cpu_stop_unpark(unsigned int cpu) + { + struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); + +- spin_lock_irq(&stopper->lock); ++ raw_spin_lock_irq(&stopper->lock); + stopper->enabled = true; +- spin_unlock_irq(&stopper->lock); ++ raw_spin_unlock_irq(&stopper->lock); + } + + static struct smp_hotplug_thread cpu_stop_threads = { +@@ -535,10 +571,12 @@ + for_each_possible_cpu(cpu) { + struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu); + +- spin_lock_init(&stopper->lock); ++ raw_spin_lock_init(&stopper->lock); + INIT_LIST_HEAD(&stopper->works); + } + ++ lg_lock_init(&stop_cpus_lock, "stop_cpus_lock"); ++ + BUG_ON(smpboot_register_percpu_thread(&cpu_stop_threads)); + stop_machine_initialized = true; + return 0; +@@ -634,11 +672,11 @@ + set_state(&msdata, MULTI_STOP_PREPARE); + cpu_stop_init_done(&done, num_active_cpus()); + queue_stop_cpus_work(cpu_active_mask, multi_cpu_stop, &msdata, +- &done); ++ &done, true); + ret = multi_cpu_stop(&msdata); + + /* Busy wait for completion. */ +- while (!completion_done(&done.completion)) ++ while (atomic_read(&done.nr_todo)) + cpu_relax(); + + mutex_unlock(&stop_cpus_mutex); +diff -Nur linux-4.1.13.orig/kernel/time/hrtimer.c linux-4.1.13/kernel/time/hrtimer.c +--- linux-4.1.13.orig/kernel/time/hrtimer.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/hrtimer.c 2015-11-29 09:23:09.649611524 +0100 +@@ -48,11 +48,13 @@ + #include <linux/sched/rt.h> + #include <linux/sched/deadline.h> + #include <linux/timer.h> ++#include <linux/kthread.h> + #include <linux/freezer.h> + + #include <asm/uaccess.h> + + #include <trace/events/timer.h> ++#include <trace/events/hist.h> + + #include "tick-internal.h" + +@@ -576,8 +578,7 @@ + * When the callback is running, we do not reprogram the clock event + * device. The timer callback is either running on a different CPU or + * the callback is executed in the hrtimer_interrupt context. The +- * reprogramming is handled either by the softirq, which called the +- * callback or at the end of the hrtimer_interrupt. ++ * reprogramming is handled at the end of the hrtimer_interrupt. + */ + if (hrtimer_callback_running(timer)) + return 0; +@@ -621,6 +622,9 @@ + return res; + } + ++static void __run_hrtimer(struct hrtimer *timer, ktime_t *now); ++static int hrtimer_rt_defer(struct hrtimer *timer); ++ + /* + * Initialize the high resolution related parts of cpu_base + */ +@@ -630,6 +634,21 @@ + base->hres_active = 0; + } + ++static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer, ++ struct hrtimer_clock_base *base, ++ int wakeup) ++{ ++ if (!hrtimer_reprogram(timer, base)) ++ return 0; ++ if (!wakeup) ++ return -ETIME; ++#ifdef CONFIG_PREEMPT_RT_BASE ++ if (!hrtimer_rt_defer(timer)) ++ return -ETIME; ++#endif ++ return 1; ++} ++ + static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base) + { + ktime_t *offs_real = &base->clock_base[HRTIMER_BASE_REALTIME].offset; +@@ -695,6 +714,44 @@ + + static DECLARE_WORK(hrtimer_work, clock_was_set_work); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * RT can not call schedule_work from real interrupt context. ++ * Need to make a thread to do the real work. ++ */ ++static struct task_struct *clock_set_delay_thread; ++static bool do_clock_set_delay; ++ ++static int run_clock_set_delay(void *ignore) ++{ ++ while (!kthread_should_stop()) { ++ set_current_state(TASK_INTERRUPTIBLE); ++ if (do_clock_set_delay) { ++ do_clock_set_delay = false; ++ schedule_work(&hrtimer_work); ++ } ++ schedule(); ++ } ++ __set_current_state(TASK_RUNNING); ++ return 0; ++} ++ ++void clock_was_set_delayed(void) ++{ ++ do_clock_set_delay = true; ++ /* Make visible before waking up process */ ++ smp_wmb(); ++ wake_up_process(clock_set_delay_thread); ++} ++ ++static __init int create_clock_set_delay_thread(void) ++{ ++ clock_set_delay_thread = kthread_run(run_clock_set_delay, NULL, "kclksetdelayd"); ++ BUG_ON(!clock_set_delay_thread); ++ return 0; ++} ++early_initcall(create_clock_set_delay_thread); ++#else /* PREEMPT_RT_FULL */ + /* + * Called from timekeeping and resume code to reprogramm the hrtimer + * interrupt device on all cpus. +@@ -703,6 +760,7 @@ + { + schedule_work(&hrtimer_work); + } ++#endif + + #else + +@@ -711,6 +769,13 @@ + static inline int hrtimer_switch_to_hres(void) { return 0; } + static inline void + hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { } ++static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer, ++ struct hrtimer_clock_base *base, ++ int wakeup) ++{ ++ return 0; ++} ++ + static inline int hrtimer_reprogram(struct hrtimer *timer, + struct hrtimer_clock_base *base) + { +@@ -718,7 +783,6 @@ + } + static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { } + static inline void retrigger_next_event(void *arg) { } +- + #endif /* CONFIG_HIGH_RES_TIMERS */ + + /* +@@ -836,6 +900,32 @@ + } + EXPORT_SYMBOL_GPL(hrtimer_forward); + ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define wake_up_timer_waiters(b) wake_up(&(b)->wait) ++ ++/** ++ * hrtimer_wait_for_timer - Wait for a running timer ++ * ++ * @timer: timer to wait for ++ * ++ * The function waits in case the timers callback function is ++ * currently executed on the waitqueue of the timer base. The ++ * waitqueue is woken up after the timer callback function has ++ * finished execution. ++ */ ++void hrtimer_wait_for_timer(const struct hrtimer *timer) ++{ ++ struct hrtimer_clock_base *base = timer->base; ++ ++ if (base && base->cpu_base && !timer->irqsafe) ++ wait_event(base->cpu_base->wait, ++ !(timer->state & HRTIMER_STATE_CALLBACK)); ++} ++ ++#else ++# define wake_up_timer_waiters(b) do { } while (0) ++#endif ++ + /* + * enqueue_hrtimer - internal function to (re)start a timer + * +@@ -879,6 +969,11 @@ + if (!(timer->state & HRTIMER_STATE_ENQUEUED)) + goto out; + ++ if (unlikely(!list_empty(&timer->cb_entry))) { ++ list_del_init(&timer->cb_entry); ++ goto out; ++ } ++ + next_timer = timerqueue_getnext(&base->active); + timerqueue_del(&base->active, &timer->node); + if (&timer->node == next_timer) { +@@ -966,7 +1061,16 @@ + new_base = switch_hrtimer_base(timer, base, mode & HRTIMER_MODE_PINNED); + + timer_stats_hrtimer_set_start_info(timer); ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ { ++ ktime_t now = new_base->get_time(); + ++ if (ktime_to_ns(tim) < ktime_to_ns(now)) ++ timer->praecox = now; ++ else ++ timer->praecox = ktime_set(0, 0); ++ } ++#endif + leftmost = enqueue_hrtimer(timer, new_base); + + if (!leftmost) { +@@ -980,15 +1084,26 @@ + * on dynticks target. + */ + wake_up_nohz_cpu(new_base->cpu_base->cpu); +- } else if (new_base->cpu_base == this_cpu_ptr(&hrtimer_bases) && +- hrtimer_reprogram(timer, new_base)) { ++ } else if (new_base->cpu_base == this_cpu_ptr(&hrtimer_bases)) { ++ ++ ret = hrtimer_enqueue_reprogram(timer, new_base, wakeup); ++ if (ret < 0) { ++ /* ++ * In case we failed to reprogram the timer (mostly ++ * because out current timer is already elapsed), ++ * remove it again and report a failure. This avoids ++ * stale base->first entries. ++ */ ++ debug_deactivate(timer); ++ __remove_hrtimer(timer, new_base, ++ timer->state & HRTIMER_STATE_CALLBACK, 0); ++ } else if (ret > 0) { + /* + * Only allow reprogramming if the new base is on this CPU. + * (it might still be on another CPU if the timer was pending) + * + * XXX send_remote_softirq() ? + */ +- if (wakeup) { + /* + * We need to drop cpu_base->lock to avoid a + * lock ordering issue vs. rq->lock. +@@ -996,9 +1111,7 @@ + raw_spin_unlock(&new_base->cpu_base->lock); + raise_softirq_irqoff(HRTIMER_SOFTIRQ); + local_irq_restore(flags); +- return ret; +- } else { +- __raise_softirq_irqoff(HRTIMER_SOFTIRQ); ++ return 0; + } + } + +@@ -1089,7 +1202,7 @@ + + if (ret >= 0) + return ret; +- cpu_relax(); ++ hrtimer_wait_for_timer(timer); + } + } + EXPORT_SYMBOL_GPL(hrtimer_cancel); +@@ -1153,6 +1266,7 @@ + + base = hrtimer_clockid_to_base(clock_id); + timer->base = &cpu_base->clock_base[base]; ++ INIT_LIST_HEAD(&timer->cb_entry); + timerqueue_init(&timer->node); + + #ifdef CONFIG_TIMER_STATS +@@ -1236,6 +1350,126 @@ + timer->state &= ~HRTIMER_STATE_CALLBACK; + } + ++static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer); ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++static void hrtimer_rt_reprogram(int restart, struct hrtimer *timer, ++ struct hrtimer_clock_base *base) ++{ ++ /* ++ * Note, we clear the callback flag before we requeue the ++ * timer otherwise we trigger the callback_running() check ++ * in hrtimer_reprogram(). ++ */ ++ timer->state &= ~HRTIMER_STATE_CALLBACK; ++ ++ if (restart != HRTIMER_NORESTART) { ++ BUG_ON(hrtimer_active(timer)); ++ /* ++ * Enqueue the timer, if it's the leftmost timer then ++ * we need to reprogram it. ++ */ ++ if (!enqueue_hrtimer(timer, base)) ++ return; ++ ++#ifndef CONFIG_HIGH_RES_TIMERS ++ } ++#else ++ if (base->cpu_base->hres_active && ++ hrtimer_reprogram(timer, base)) ++ goto requeue; ++ ++ } else if (hrtimer_active(timer)) { ++ /* ++ * If the timer was rearmed on another CPU, reprogram ++ * the event device. ++ */ ++ if (&timer->node == base->active.next && ++ base->cpu_base->hres_active && ++ hrtimer_reprogram(timer, base)) ++ goto requeue; ++ } ++ return; ++ ++requeue: ++ /* ++ * Timer is expired. Thus move it from tree to pending list ++ * again. ++ */ ++ __remove_hrtimer(timer, base, timer->state, 0); ++ list_add_tail(&timer->cb_entry, &base->expired); ++#endif ++} ++ ++/* ++ * The changes in mainline which removed the callback modes from ++ * hrtimer are not yet working with -rt. The non wakeup_process() ++ * based callbacks which involve sleeping locks need to be treated ++ * seperately. ++ */ ++static void hrtimer_rt_run_pending(void) ++{ ++ enum hrtimer_restart (*fn)(struct hrtimer *); ++ struct hrtimer_cpu_base *cpu_base; ++ struct hrtimer_clock_base *base; ++ struct hrtimer *timer; ++ int index, restart; ++ ++ local_irq_disable(); ++ cpu_base = &per_cpu(hrtimer_bases, smp_processor_id()); ++ ++ raw_spin_lock(&cpu_base->lock); ++ ++ for (index = 0; index < HRTIMER_MAX_CLOCK_BASES; index++) { ++ base = &cpu_base->clock_base[index]; ++ ++ while (!list_empty(&base->expired)) { ++ timer = list_first_entry(&base->expired, ++ struct hrtimer, cb_entry); ++ ++ /* ++ * Same as the above __run_hrtimer function ++ * just we run with interrupts enabled. ++ */ ++ debug_hrtimer_deactivate(timer); ++ __remove_hrtimer(timer, base, HRTIMER_STATE_CALLBACK, 0); ++ timer_stats_account_hrtimer(timer); ++ fn = timer->function; ++ ++ raw_spin_unlock_irq(&cpu_base->lock); ++ restart = fn(timer); ++ raw_spin_lock_irq(&cpu_base->lock); ++ ++ hrtimer_rt_reprogram(restart, timer, base); ++ } ++ } ++ ++ raw_spin_unlock_irq(&cpu_base->lock); ++ ++ wake_up_timer_waiters(cpu_base); ++} ++ ++static int hrtimer_rt_defer(struct hrtimer *timer) ++{ ++ if (timer->irqsafe) ++ return 0; ++ ++ __remove_hrtimer(timer, timer->base, timer->state, 0); ++ list_add_tail(&timer->cb_entry, &timer->base->expired); ++ return 1; ++} ++ ++#else ++ ++static inline void hrtimer_rt_run_pending(void) ++{ ++ hrtimer_peek_ahead_timers(); ++} ++ ++static inline int hrtimer_rt_defer(struct hrtimer *timer) { return 0; } ++ ++#endif ++ + #ifdef CONFIG_HIGH_RES_TIMERS + + /* +@@ -1246,7 +1480,7 @@ + { + struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + ktime_t expires_next, now, entry_time, delta; +- int i, retries = 0; ++ int i, retries = 0, raise = 0; + + BUG_ON(!cpu_base->hres_active); + cpu_base->nr_events++; +@@ -1281,6 +1515,15 @@ + + timer = container_of(node, struct hrtimer, node); + ++ trace_hrtimer_interrupt(raw_smp_processor_id(), ++ ktime_to_ns(ktime_sub(ktime_to_ns(timer->praecox) ? ++ timer->praecox : hrtimer_get_expires(timer), ++ basenow)), ++ current, ++ timer->function == hrtimer_wakeup ? ++ container_of(timer, struct hrtimer_sleeper, ++ timer)->task : NULL); ++ + /* + * The immediate goal for using the softexpires is + * minimizing wakeups, not running timers at the +@@ -1296,7 +1539,10 @@ + if (basenow.tv64 < hrtimer_get_softexpires_tv64(timer)) + break; + +- __run_hrtimer(timer, &basenow); ++ if (!hrtimer_rt_defer(timer)) ++ __run_hrtimer(timer, &basenow); ++ else ++ raise = 1; + } + } + /* Reevaluate the clock bases for the next expiry */ +@@ -1313,7 +1559,7 @@ + if (expires_next.tv64 == KTIME_MAX || + !tick_program_event(expires_next, 0)) { + cpu_base->hang_detected = 0; +- return; ++ goto out; + } + + /* +@@ -1357,6 +1603,9 @@ + tick_program_event(expires_next, 1); + printk_once(KERN_WARNING "hrtimer: interrupt took %llu ns\n", + ktime_to_ns(delta)); ++out: ++ if (raise) ++ raise_softirq_irqoff(HRTIMER_SOFTIRQ); + } + + /* +@@ -1392,18 +1641,18 @@ + __hrtimer_peek_ahead_timers(); + local_irq_restore(flags); + } +- +-static void run_hrtimer_softirq(struct softirq_action *h) +-{ +- hrtimer_peek_ahead_timers(); +-} +- + #else /* CONFIG_HIGH_RES_TIMERS */ + + static inline void __hrtimer_peek_ahead_timers(void) { } + + #endif /* !CONFIG_HIGH_RES_TIMERS */ + ++ ++static void run_hrtimer_softirq(struct softirq_action *h) ++{ ++ hrtimer_rt_run_pending(); ++} ++ + /* + * Called from timer softirq every jiffy, expire hrtimers: + * +@@ -1436,7 +1685,7 @@ + struct timerqueue_node *node; + struct hrtimer_cpu_base *cpu_base = this_cpu_ptr(&hrtimer_bases); + struct hrtimer_clock_base *base; +- int index, gettime = 1; ++ int index, gettime = 1, raise = 0; + + if (hrtimer_hres_active()) + return; +@@ -1461,10 +1710,16 @@ + hrtimer_get_expires_tv64(timer)) + break; + +- __run_hrtimer(timer, &base->softirq_time); ++ if (!hrtimer_rt_defer(timer)) ++ __run_hrtimer(timer, &base->softirq_time); ++ else ++ raise = 1; + } + raw_spin_unlock(&cpu_base->lock); + } ++ ++ if (raise) ++ raise_softirq_irqoff(HRTIMER_SOFTIRQ); + } + + /* +@@ -1486,16 +1741,18 @@ + void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task) + { + sl->timer.function = hrtimer_wakeup; ++ sl->timer.irqsafe = 1; + sl->task = task; + } + EXPORT_SYMBOL_GPL(hrtimer_init_sleeper); + +-static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode) ++static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mode, ++ unsigned long state) + { + hrtimer_init_sleeper(t, current); + + do { +- set_current_state(TASK_INTERRUPTIBLE); ++ set_current_state(state); + hrtimer_start_expires(&t->timer, mode); + if (!hrtimer_active(&t->timer)) + t->task = NULL; +@@ -1539,7 +1796,8 @@ + HRTIMER_MODE_ABS); + hrtimer_set_expires_tv64(&t.timer, restart->nanosleep.expires); + +- if (do_nanosleep(&t, HRTIMER_MODE_ABS)) ++ /* cpu_chill() does not care about restart state. */ ++ if (do_nanosleep(&t, HRTIMER_MODE_ABS, TASK_INTERRUPTIBLE)) + goto out; + + rmtp = restart->nanosleep.rmtp; +@@ -1556,8 +1814,10 @@ + return ret; + } + +-long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, +- const enum hrtimer_mode mode, const clockid_t clockid) ++static long ++__hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, ++ const enum hrtimer_mode mode, const clockid_t clockid, ++ unsigned long state) + { + struct restart_block *restart; + struct hrtimer_sleeper t; +@@ -1570,7 +1830,7 @@ + + hrtimer_init_on_stack(&t.timer, clockid, mode); + hrtimer_set_expires_range_ns(&t.timer, timespec_to_ktime(*rqtp), slack); +- if (do_nanosleep(&t, mode)) ++ if (do_nanosleep(&t, mode, state)) + goto out; + + /* Absolute timers do not update the rmtp value and restart: */ +@@ -1597,6 +1857,12 @@ + return ret; + } + ++long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp, ++ const enum hrtimer_mode mode, const clockid_t clockid) ++{ ++ return __hrtimer_nanosleep(rqtp, rmtp, mode, clockid, TASK_INTERRUPTIBLE); ++} ++ + SYSCALL_DEFINE2(nanosleep, struct timespec __user *, rqtp, + struct timespec __user *, rmtp) + { +@@ -1611,6 +1877,26 @@ + return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * Sleep for 1 ms in hope whoever holds what we want will let it go. ++ */ ++void cpu_chill(void) ++{ ++ struct timespec tu = { ++ .tv_nsec = NSEC_PER_MSEC, ++ }; ++ unsigned int freeze_flag = current->flags & PF_NOFREEZE; ++ ++ current->flags |= PF_NOFREEZE; ++ __hrtimer_nanosleep(&tu, NULL, HRTIMER_MODE_REL, CLOCK_MONOTONIC, ++ TASK_UNINTERRUPTIBLE); ++ if (!freeze_flag) ++ current->flags &= ~PF_NOFREEZE; ++} ++EXPORT_SYMBOL(cpu_chill); ++#endif ++ + /* + * Functions related to boot-time initialization: + */ +@@ -1622,10 +1908,14 @@ + for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { + cpu_base->clock_base[i].cpu_base = cpu_base; + timerqueue_init_head(&cpu_base->clock_base[i].active); ++ INIT_LIST_HEAD(&cpu_base->clock_base[i].expired); + } + + cpu_base->cpu = cpu; + hrtimer_init_hres(cpu_base); ++#ifdef CONFIG_PREEMPT_RT_BASE ++ init_waitqueue_head(&cpu_base->wait); ++#endif + } + + #ifdef CONFIG_HOTPLUG_CPU +@@ -1731,9 +2021,7 @@ + hrtimer_cpu_notify(&hrtimers_nb, (unsigned long)CPU_UP_PREPARE, + (void *)(long)smp_processor_id()); + register_cpu_notifier(&hrtimers_nb); +-#ifdef CONFIG_HIGH_RES_TIMERS + open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq); +-#endif + } + + /** +diff -Nur linux-4.1.13.orig/kernel/time/itimer.c linux-4.1.13/kernel/time/itimer.c +--- linux-4.1.13.orig/kernel/time/itimer.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/itimer.c 2015-11-29 09:23:09.649611524 +0100 +@@ -213,6 +213,7 @@ + /* We are sharing ->siglock with it_real_fn() */ + if (hrtimer_try_to_cancel(timer) < 0) { + spin_unlock_irq(&tsk->sighand->siglock); ++ hrtimer_wait_for_timer(&tsk->signal->real_timer); + goto again; + } + expires = timeval_to_ktime(value->it_value); +diff -Nur linux-4.1.13.orig/kernel/time/jiffies.c linux-4.1.13/kernel/time/jiffies.c +--- linux-4.1.13.orig/kernel/time/jiffies.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/jiffies.c 2015-11-29 09:23:09.649611524 +0100 +@@ -74,7 +74,8 @@ + .max_cycles = 10, + }; + +-__cacheline_aligned_in_smp DEFINE_SEQLOCK(jiffies_lock); ++__cacheline_aligned_in_smp DEFINE_RAW_SPINLOCK(jiffies_lock); ++__cacheline_aligned_in_smp seqcount_t jiffies_seq; + + #if (BITS_PER_LONG < 64) + u64 get_jiffies_64(void) +@@ -83,9 +84,9 @@ + u64 ret; + + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + ret = jiffies_64; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + return ret; + } + EXPORT_SYMBOL(get_jiffies_64); +diff -Nur linux-4.1.13.orig/kernel/time/ntp.c linux-4.1.13/kernel/time/ntp.c +--- linux-4.1.13.orig/kernel/time/ntp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/ntp.c 2015-11-29 09:23:09.649611524 +0100 +@@ -10,6 +10,7 @@ + #include <linux/workqueue.h> + #include <linux/hrtimer.h> + #include <linux/jiffies.h> ++#include <linux/kthread.h> + #include <linux/math64.h> + #include <linux/timex.h> + #include <linux/time.h> +@@ -529,10 +530,52 @@ + &sync_cmos_work, timespec_to_jiffies(&next)); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * RT can not call schedule_delayed_work from real interrupt context. ++ * Need to make a thread to do the real work. ++ */ ++static struct task_struct *cmos_delay_thread; ++static bool do_cmos_delay; ++ ++static int run_cmos_delay(void *ignore) ++{ ++ while (!kthread_should_stop()) { ++ set_current_state(TASK_INTERRUPTIBLE); ++ if (do_cmos_delay) { ++ do_cmos_delay = false; ++ queue_delayed_work(system_power_efficient_wq, ++ &sync_cmos_work, 0); ++ } ++ schedule(); ++ } ++ __set_current_state(TASK_RUNNING); ++ return 0; ++} ++ ++void ntp_notify_cmos_timer(void) ++{ ++ do_cmos_delay = true; ++ /* Make visible before waking up process */ ++ smp_wmb(); ++ wake_up_process(cmos_delay_thread); ++} ++ ++static __init int create_cmos_delay_thread(void) ++{ ++ cmos_delay_thread = kthread_run(run_cmos_delay, NULL, "kcmosdelayd"); ++ BUG_ON(!cmos_delay_thread); ++ return 0; ++} ++early_initcall(create_cmos_delay_thread); ++ ++#else ++ + void ntp_notify_cmos_timer(void) + { + queue_delayed_work(system_power_efficient_wq, &sync_cmos_work, 0); + } ++#endif /* CONFIG_PREEMPT_RT_FULL */ + + #else + void ntp_notify_cmos_timer(void) { } +diff -Nur linux-4.1.13.orig/kernel/time/posix-cpu-timers.c linux-4.1.13/kernel/time/posix-cpu-timers.c +--- linux-4.1.13.orig/kernel/time/posix-cpu-timers.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/posix-cpu-timers.c 2015-11-29 09:23:09.649611524 +0100 +@@ -3,6 +3,7 @@ + */ + + #include <linux/sched.h> ++#include <linux/sched/rt.h> + #include <linux/posix-timers.h> + #include <linux/errno.h> + #include <linux/math64.h> +@@ -626,7 +627,7 @@ + /* + * Disarm any old timer after extracting its expiry time. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + ret = 0; + old_incr = timer->it.cpu.incr; +@@ -1047,7 +1048,7 @@ + /* + * Now re-arm for the new expiry time. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + arm_timer(timer); + unlock_task_sighand(p, &flags); + +@@ -1113,10 +1114,11 @@ + sig = tsk->signal; + if (sig->cputimer.running) { + struct task_cputime group_sample; ++ unsigned long flags; + +- raw_spin_lock(&sig->cputimer.lock); ++ raw_spin_lock_irqsave(&sig->cputimer.lock, flags); + group_sample = sig->cputimer.cputime; +- raw_spin_unlock(&sig->cputimer.lock); ++ raw_spin_unlock_irqrestore(&sig->cputimer.lock, flags); + + if (task_cputime_expired(&group_sample, &sig->cputime_expires)) + return 1; +@@ -1130,13 +1132,13 @@ + * already updated our counts. We need to check if any timers fire now. + * Interrupts are disabled. + */ +-void run_posix_cpu_timers(struct task_struct *tsk) ++static void __run_posix_cpu_timers(struct task_struct *tsk) + { + LIST_HEAD(firing); + struct k_itimer *timer, *next; + unsigned long flags; + +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + /* + * The fast path checks that there are no expired thread or thread +@@ -1194,6 +1196,190 @@ + } + } + ++#ifdef CONFIG_PREEMPT_RT_BASE ++#include <linux/kthread.h> ++#include <linux/cpu.h> ++DEFINE_PER_CPU(struct task_struct *, posix_timer_task); ++DEFINE_PER_CPU(struct task_struct *, posix_timer_tasklist); ++ ++static int posix_cpu_timers_thread(void *data) ++{ ++ int cpu = (long)data; ++ ++ BUG_ON(per_cpu(posix_timer_task,cpu) != current); ++ ++ while (!kthread_should_stop()) { ++ struct task_struct *tsk = NULL; ++ struct task_struct *next = NULL; ++ ++ if (cpu_is_offline(cpu)) ++ goto wait_to_die; ++ ++ /* grab task list */ ++ raw_local_irq_disable(); ++ tsk = per_cpu(posix_timer_tasklist, cpu); ++ per_cpu(posix_timer_tasklist, cpu) = NULL; ++ raw_local_irq_enable(); ++ ++ /* its possible the list is empty, just return */ ++ if (!tsk) { ++ set_current_state(TASK_INTERRUPTIBLE); ++ schedule(); ++ __set_current_state(TASK_RUNNING); ++ continue; ++ } ++ ++ /* Process task list */ ++ while (1) { ++ /* save next */ ++ next = tsk->posix_timer_list; ++ ++ /* run the task timers, clear its ptr and ++ * unreference it ++ */ ++ __run_posix_cpu_timers(tsk); ++ tsk->posix_timer_list = NULL; ++ put_task_struct(tsk); ++ ++ /* check if this is the last on the list */ ++ if (next == tsk) ++ break; ++ tsk = next; ++ } ++ } ++ return 0; ++ ++wait_to_die: ++ /* Wait for kthread_stop */ ++ set_current_state(TASK_INTERRUPTIBLE); ++ while (!kthread_should_stop()) { ++ schedule(); ++ set_current_state(TASK_INTERRUPTIBLE); ++ } ++ __set_current_state(TASK_RUNNING); ++ return 0; ++} ++ ++static inline int __fastpath_timer_check(struct task_struct *tsk) ++{ ++ /* tsk == current, ensure it is safe to use ->signal/sighand */ ++ if (unlikely(tsk->exit_state)) ++ return 0; ++ ++ if (!task_cputime_zero(&tsk->cputime_expires)) ++ return 1; ++ ++ if (!task_cputime_zero(&tsk->signal->cputime_expires)) ++ return 1; ++ ++ return 0; ++} ++ ++void run_posix_cpu_timers(struct task_struct *tsk) ++{ ++ unsigned long cpu = smp_processor_id(); ++ struct task_struct *tasklist; ++ ++ BUG_ON(!irqs_disabled()); ++ if(!per_cpu(posix_timer_task, cpu)) ++ return; ++ /* get per-cpu references */ ++ tasklist = per_cpu(posix_timer_tasklist, cpu); ++ ++ /* check to see if we're already queued */ ++ if (!tsk->posix_timer_list && __fastpath_timer_check(tsk)) { ++ get_task_struct(tsk); ++ if (tasklist) { ++ tsk->posix_timer_list = tasklist; ++ } else { ++ /* ++ * The list is terminated by a self-pointing ++ * task_struct ++ */ ++ tsk->posix_timer_list = tsk; ++ } ++ per_cpu(posix_timer_tasklist, cpu) = tsk; ++ ++ wake_up_process(per_cpu(posix_timer_task, cpu)); ++ } ++} ++ ++/* ++ * posix_cpu_thread_call - callback that gets triggered when a CPU is added. ++ * Here we can start up the necessary migration thread for the new CPU. ++ */ ++static int posix_cpu_thread_call(struct notifier_block *nfb, ++ unsigned long action, void *hcpu) ++{ ++ int cpu = (long)hcpu; ++ struct task_struct *p; ++ struct sched_param param; ++ ++ switch (action) { ++ case CPU_UP_PREPARE: ++ p = kthread_create(posix_cpu_timers_thread, hcpu, ++ "posixcputmr/%d",cpu); ++ if (IS_ERR(p)) ++ return NOTIFY_BAD; ++ p->flags |= PF_NOFREEZE; ++ kthread_bind(p, cpu); ++ /* Must be high prio to avoid getting starved */ ++ param.sched_priority = MAX_RT_PRIO-1; ++ sched_setscheduler(p, SCHED_FIFO, ¶m); ++ per_cpu(posix_timer_task,cpu) = p; ++ break; ++ case CPU_ONLINE: ++ /* Strictly unneccessary, as first user will wake it. */ ++ wake_up_process(per_cpu(posix_timer_task,cpu)); ++ break; ++#ifdef CONFIG_HOTPLUG_CPU ++ case CPU_UP_CANCELED: ++ /* Unbind it from offline cpu so it can run. Fall thru. */ ++ kthread_bind(per_cpu(posix_timer_task, cpu), ++ cpumask_any(cpu_online_mask)); ++ kthread_stop(per_cpu(posix_timer_task,cpu)); ++ per_cpu(posix_timer_task,cpu) = NULL; ++ break; ++ case CPU_DEAD: ++ kthread_stop(per_cpu(posix_timer_task,cpu)); ++ per_cpu(posix_timer_task,cpu) = NULL; ++ break; ++#endif ++ } ++ return NOTIFY_OK; ++} ++ ++/* Register at highest priority so that task migration (migrate_all_tasks) ++ * happens before everything else. ++ */ ++static struct notifier_block posix_cpu_thread_notifier = { ++ .notifier_call = posix_cpu_thread_call, ++ .priority = 10 ++}; ++ ++static int __init posix_cpu_thread_init(void) ++{ ++ void *hcpu = (void *)(long)smp_processor_id(); ++ /* Start one for boot CPU. */ ++ unsigned long cpu; ++ ++ /* init the per-cpu posix_timer_tasklets */ ++ for_each_possible_cpu(cpu) ++ per_cpu(posix_timer_tasklist, cpu) = NULL; ++ ++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_UP_PREPARE, hcpu); ++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_ONLINE, hcpu); ++ register_cpu_notifier(&posix_cpu_thread_notifier); ++ return 0; ++} ++early_initcall(posix_cpu_thread_init); ++#else /* CONFIG_PREEMPT_RT_BASE */ ++void run_posix_cpu_timers(struct task_struct *tsk) ++{ ++ __run_posix_cpu_timers(tsk); ++} ++#endif /* CONFIG_PREEMPT_RT_BASE */ ++ + /* + * Set one of the process-wide special case CPU timers or RLIMIT_CPU. + * The tsk->sighand->siglock must be held by the caller. +diff -Nur linux-4.1.13.orig/kernel/time/posix-timers.c linux-4.1.13/kernel/time/posix-timers.c +--- linux-4.1.13.orig/kernel/time/posix-timers.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/posix-timers.c 2015-11-29 09:23:09.649611524 +0100 +@@ -499,6 +499,7 @@ + static struct pid *good_sigevent(sigevent_t * event) + { + struct task_struct *rtn = current->group_leader; ++ int sig = event->sigev_signo; + + if ((event->sigev_notify & SIGEV_THREAD_ID ) && + (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) || +@@ -507,7 +508,8 @@ + return NULL; + + if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) && +- ((event->sigev_signo <= 0) || (event->sigev_signo > SIGRTMAX))) ++ (sig <= 0 || sig > SIGRTMAX || sig_kernel_only(sig) || ++ sig_kernel_coredump(sig))) + return NULL; + + return task_pid(rtn); +@@ -819,6 +821,20 @@ + return overrun; + } + ++/* ++ * Protected by RCU! ++ */ ++static void timer_wait_for_callback(struct k_clock *kc, struct k_itimer *timr) ++{ ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (kc->timer_set == common_timer_set) ++ hrtimer_wait_for_timer(&timr->it.real.timer); ++ else ++ /* FIXME: Whacky hack for posix-cpu-timers */ ++ schedule_timeout(1); ++#endif ++} ++ + /* Set a POSIX.1b interval timer. */ + /* timr->it_lock is taken. */ + static int +@@ -896,6 +912,7 @@ + if (!timr) + return -EINVAL; + ++ rcu_read_lock(); + kc = clockid_to_kclock(timr->it_clock); + if (WARN_ON_ONCE(!kc || !kc->timer_set)) + error = -EINVAL; +@@ -904,9 +921,12 @@ + + unlock_timer(timr, flag); + if (error == TIMER_RETRY) { ++ timer_wait_for_callback(kc, timr); + rtn = NULL; // We already got the old time... ++ rcu_read_unlock(); + goto retry; + } ++ rcu_read_unlock(); + + if (old_setting && !error && + copy_to_user(old_setting, &old_spec, sizeof (old_spec))) +@@ -944,10 +964,15 @@ + if (!timer) + return -EINVAL; + ++ rcu_read_lock(); + if (timer_delete_hook(timer) == TIMER_RETRY) { + unlock_timer(timer, flags); ++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), ++ timer); ++ rcu_read_unlock(); + goto retry_delete; + } ++ rcu_read_unlock(); + + spin_lock(¤t->sighand->siglock); + list_del(&timer->list); +@@ -973,8 +998,18 @@ + retry_delete: + spin_lock_irqsave(&timer->it_lock, flags); + ++ /* On RT we can race with a deletion */ ++ if (!timer->it_signal) { ++ unlock_timer(timer, flags); ++ return; ++ } ++ + if (timer_delete_hook(timer) == TIMER_RETRY) { ++ rcu_read_lock(); + unlock_timer(timer, flags); ++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock), ++ timer); ++ rcu_read_unlock(); + goto retry_delete; + } + list_del(&timer->list); +diff -Nur linux-4.1.13.orig/kernel/time/tick-common.c linux-4.1.13/kernel/time/tick-common.c +--- linux-4.1.13.orig/kernel/time/tick-common.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/tick-common.c 2015-11-29 09:23:09.649611524 +0100 +@@ -78,13 +78,15 @@ + static void tick_periodic(int cpu) + { + if (tick_do_timer_cpu == cpu) { +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + + /* Keep track of the next tick event */ + tick_next_period = ktime_add(tick_next_period, tick_period); + + do_timer(1); +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } + +@@ -146,9 +148,9 @@ + ktime_t next; + + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + next = tick_next_period; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + + clockevents_set_state(dev, CLOCK_EVT_STATE_ONESHOT); + +diff -Nur linux-4.1.13.orig/kernel/time/tick-sched.c linux-4.1.13/kernel/time/tick-sched.c +--- linux-4.1.13.orig/kernel/time/tick-sched.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/tick-sched.c 2015-11-29 09:23:09.649611524 +0100 +@@ -62,7 +62,8 @@ + return; + + /* Reevalute with jiffies_lock held */ +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + + delta = ktime_sub(now, last_jiffies_update); + if (delta.tv64 >= tick_period.tv64) { +@@ -85,10 +86,12 @@ + /* Keep the tick_next_period variable up to date */ + tick_next_period = ktime_add(last_jiffies_update, tick_period); + } else { +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + return; + } +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } + +@@ -99,12 +102,14 @@ + { + ktime_t period; + +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + /* Did we start the jiffies update yet ? */ + if (last_jiffies_update.tv64 == 0) + last_jiffies_update = tick_next_period; + period = last_jiffies_update; +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + return period; + } + +@@ -176,6 +181,11 @@ + return false; + } + ++ if (!arch_irq_work_has_interrupt()) { ++ trace_tick_stop(0, "missing irq work interrupt\n"); ++ return false; ++ } ++ + /* sched_clock_tick() needs us? */ + #ifdef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK + /* +@@ -222,6 +232,7 @@ + + static DEFINE_PER_CPU(struct irq_work, nohz_full_kick_work) = { + .func = nohz_full_kick_work_func, ++ .flags = IRQ_WORK_HARD_IRQ, + }; + + /* +@@ -578,10 +589,10 @@ + + /* Read jiffies and the time when jiffies were updated last */ + do { +- seq = read_seqbegin(&jiffies_lock); ++ seq = read_seqcount_begin(&jiffies_seq); + last_update = last_jiffies_update; + last_jiffies = jiffies; +- } while (read_seqretry(&jiffies_lock, seq)); ++ } while (read_seqcount_retry(&jiffies_seq, seq)); + + if (rcu_needs_cpu(&rcu_delta_jiffies) || + arch_needs_cpu() || irq_work_needs_cpu()) { +@@ -759,14 +770,7 @@ + return false; + + if (unlikely(local_softirq_pending() && cpu_online(cpu))) { +- static int ratelimit; +- +- if (ratelimit < 10 && +- (local_softirq_pending() & SOFTIRQ_STOP_IDLE_MASK)) { +- pr_warn("NOHZ: local_softirq_pending %02x\n", +- (unsigned int) local_softirq_pending()); +- ratelimit++; +- } ++ softirq_check_pending_idle(); + return false; + } + +@@ -1154,6 +1158,7 @@ + * Emulate tick processing via per-CPU hrtimers: + */ + hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); ++ ts->sched_timer.irqsafe = 1; + ts->sched_timer.function = tick_sched_timer; + + /* Get the next period (per cpu) */ +diff -Nur linux-4.1.13.orig/kernel/time/timekeeping.c linux-4.1.13/kernel/time/timekeeping.c +--- linux-4.1.13.orig/kernel/time/timekeeping.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/timekeeping.c 2015-11-29 09:23:09.649611524 +0100 +@@ -2065,8 +2065,10 @@ + */ + void xtime_update(unsigned long ticks) + { +- write_seqlock(&jiffies_lock); ++ raw_spin_lock(&jiffies_lock); ++ write_seqcount_begin(&jiffies_seq); + do_timer(ticks); +- write_sequnlock(&jiffies_lock); ++ write_seqcount_end(&jiffies_seq); ++ raw_spin_unlock(&jiffies_lock); + update_wall_time(); + } +diff -Nur linux-4.1.13.orig/kernel/time/timekeeping.h linux-4.1.13/kernel/time/timekeeping.h +--- linux-4.1.13.orig/kernel/time/timekeeping.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/timekeeping.h 2015-11-29 09:23:09.653611260 +0100 +@@ -22,7 +22,8 @@ + extern void do_timer(unsigned long ticks); + extern void update_wall_time(void); + +-extern seqlock_t jiffies_lock; ++extern raw_spinlock_t jiffies_lock; ++extern seqcount_t jiffies_seq; + + #define CS_NAME_LEN 32 + +diff -Nur linux-4.1.13.orig/kernel/time/timer.c linux-4.1.13/kernel/time/timer.c +--- linux-4.1.13.orig/kernel/time/timer.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/time/timer.c 2015-11-29 09:23:09.653611260 +0100 +@@ -78,6 +78,9 @@ + struct tvec_base { + spinlock_t lock; + struct timer_list *running_timer; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ wait_queue_head_t wait_for_running_timer; ++#endif + unsigned long timer_jiffies; + unsigned long next_timer; + unsigned long active_timers; +@@ -768,6 +771,36 @@ + } + } + ++#ifndef CONFIG_PREEMPT_RT_FULL ++static inline struct tvec_base *switch_timer_base(struct timer_list *timer, ++ struct tvec_base *old, ++ struct tvec_base *new) ++{ ++ /* See the comment in lock_timer_base() */ ++ timer_set_base(timer, NULL); ++ spin_unlock(&old->lock); ++ spin_lock(&new->lock); ++ timer_set_base(timer, new); ++ return new; ++} ++#else ++static inline struct tvec_base *switch_timer_base(struct timer_list *timer, ++ struct tvec_base *old, ++ struct tvec_base *new) ++{ ++ /* ++ * We cannot do the above because we might be preempted and ++ * then the preempter would see NULL and loop forever. ++ */ ++ if (spin_trylock(&new->lock)) { ++ timer_set_base(timer, new); ++ spin_unlock(&old->lock); ++ return new; ++ } ++ return old; ++} ++#endif ++ + static inline int + __mod_timer(struct timer_list *timer, unsigned long expires, + bool pending_only, int pinned) +@@ -798,14 +831,8 @@ + * handler yet has not finished. This also guarantees that + * the timer is serialized wrt itself. + */ +- if (likely(base->running_timer != timer)) { +- /* See the comment in lock_timer_base() */ +- timer_set_base(timer, NULL); +- spin_unlock(&base->lock); +- base = new_base; +- spin_lock(&base->lock); +- timer_set_base(timer, base); +- } ++ if (likely(base->running_timer != timer)) ++ base = switch_timer_base(timer, base, new_base); + } + + timer->expires = expires; +@@ -979,6 +1006,29 @@ + } + EXPORT_SYMBOL_GPL(add_timer_on); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * Wait for a running timer ++ */ ++static void wait_for_running_timer(struct timer_list *timer) ++{ ++ struct tvec_base *base = timer->base; ++ ++ if (base->running_timer == timer) ++ wait_event(base->wait_for_running_timer, ++ base->running_timer != timer); ++} ++ ++# define wakeup_timer_waiters(b) wake_up(&(b)->wait_for_running_timer) ++#else ++static inline void wait_for_running_timer(struct timer_list *timer) ++{ ++ cpu_relax(); ++} ++ ++# define wakeup_timer_waiters(b) do { } while (0) ++#endif ++ + /** + * del_timer - deactive a timer. + * @timer: the timer to be deactivated +@@ -1036,7 +1086,7 @@ + } + EXPORT_SYMBOL(try_to_del_timer_sync); + +-#ifdef CONFIG_SMP ++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL) + static DEFINE_PER_CPU(struct tvec_base, __tvec_bases); + + /** +@@ -1098,7 +1148,7 @@ + int ret = try_to_del_timer_sync(timer); + if (ret >= 0) + return ret; +- cpu_relax(); ++ wait_for_running_timer(timer); + } + } + EXPORT_SYMBOL(del_timer_sync); +@@ -1219,15 +1269,17 @@ + if (irqsafe) { + spin_unlock(&base->lock); + call_timer_fn(timer, fn, data); ++ base->running_timer = NULL; + spin_lock(&base->lock); + } else { + spin_unlock_irq(&base->lock); + call_timer_fn(timer, fn, data); ++ base->running_timer = NULL; + spin_lock_irq(&base->lock); + } + } + } +- base->running_timer = NULL; ++ wakeup_timer_waiters(base); + spin_unlock_irq(&base->lock); + } + +@@ -1367,6 +1419,14 @@ + if (cpu_is_offline(smp_processor_id())) + return expires; + ++#ifdef CONFIG_PREEMPT_RT_FULL ++ /* ++ * On PREEMPT_RT we cannot sleep here. As a result we can't take ++ * the base lock to check when the next timer is pending and so ++ * we assume the next jiffy. ++ */ ++ return now + 1; ++#endif + spin_lock(&base->lock); + if (base->active_timers) { + if (time_before_eq(base->next_timer, base->timer_jiffies)) +@@ -1392,13 +1452,13 @@ + + /* Note: this timer irq context must be accounted for as well. */ + account_process_tick(p, user_tick); ++ scheduler_tick(); + run_local_timers(); + rcu_check_callbacks(user_tick); +-#ifdef CONFIG_IRQ_WORK ++#if defined(CONFIG_IRQ_WORK) + if (in_irq()) + irq_work_tick(); + #endif +- scheduler_tick(); + run_posix_cpu_timers(p); + } + +@@ -1411,6 +1471,8 @@ + + hrtimer_run_pending(); + ++ irq_work_tick_soft(); ++ + if (time_after_eq(jiffies, base->timer_jiffies)) + __run_timers(base); + } +@@ -1566,7 +1628,7 @@ + + BUG_ON(cpu_online(cpu)); + old_base = per_cpu(tvec_bases, cpu); +- new_base = get_cpu_var(tvec_bases); ++ new_base = get_local_var(tvec_bases); + /* + * The caller is globally serialized and nobody else + * takes two locks at once, deadlock is not possible. +@@ -1590,7 +1652,7 @@ + + spin_unlock(&old_base->lock); + spin_unlock_irq(&new_base->lock); +- put_cpu_var(tvec_bases); ++ put_local_var(tvec_bases); + } + + static int timer_cpu_notify(struct notifier_block *self, +@@ -1625,6 +1687,9 @@ + base->cpu = cpu; + per_cpu(tvec_bases, cpu) = base; + spin_lock_init(&base->lock); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ init_waitqueue_head(&base->wait_for_running_timer); ++#endif + + for (j = 0; j < TVN_SIZE; j++) { + INIT_LIST_HEAD(base->tv5.vec + j); +diff -Nur linux-4.1.13.orig/kernel/trace/Kconfig linux-4.1.13/kernel/trace/Kconfig +--- linux-4.1.13.orig/kernel/trace/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/Kconfig 2015-11-29 09:23:09.653611260 +0100 +@@ -187,6 +187,24 @@ + enabled. This option and the preempt-off timing option can be + used together or separately.) + ++config INTERRUPT_OFF_HIST ++ bool "Interrupts-off Latency Histogram" ++ depends on IRQSOFF_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the duration of time periods with interrupts disabled. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff ++ ++ If PREEMPT_OFF_HIST is also selected, additional histograms (one ++ per cpu) are generated that accumulate the duration of time periods ++ when both interrupts and preemption are disabled. The histogram data ++ will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/irqsoff ++ + config PREEMPT_TRACER + bool "Preemption-off Latency Tracer" + default n +@@ -211,6 +229,24 @@ + enabled. This option and the irqs-off timing option can be + used together or separately.) + ++config PREEMPT_OFF_HIST ++ bool "Preemption-off Latency Histogram" ++ depends on PREEMPT_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the duration of time periods with preemption disabled. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff ++ ++ If INTERRUPT_OFF_HIST is also selected, additional histograms (one ++ per cpu) are generated that accumulate the duration of time periods ++ when both interrupts and preemption are disabled. The histogram data ++ will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/preemptoff ++ + config SCHED_TRACER + bool "Scheduling Latency Tracer" + select GENERIC_TRACER +@@ -221,6 +257,74 @@ + This tracer tracks the latency of the highest priority task + to be scheduled in, starting from the point it has woken up. + ++config WAKEUP_LATENCY_HIST ++ bool "Scheduling Latency Histogram" ++ depends on SCHED_TRACER ++ help ++ This option generates continuously updated histograms (one per cpu) ++ of the scheduling latency of the highest priority task. ++ The histograms are disabled by default. To enable them, write a ++ non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/wakeup ++ ++ Two different algorithms are used, one to determine the latency of ++ processes that exclusively use the highest priority of the system and ++ another one to determine the latency of processes that share the ++ highest system priority with other processes. The former is used to ++ improve hardware and system software, the latter to optimize the ++ priority design of a given system. The histogram data will be ++ located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/wakeup ++ ++ and ++ ++ /sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio ++ ++ If both Scheduling Latency Histogram and Missed Timer Offsets ++ Histogram are selected, additional histogram data will be collected ++ that contain, in addition to the wakeup latency, the timer latency, in ++ case the wakeup was triggered by an expired timer. These histograms ++ are available in the ++ ++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup ++ ++ directory. They reflect the apparent interrupt and scheduling latency ++ and are best suitable to determine the worst-case latency of a given ++ system. To enable these histograms, write a non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup ++ ++config MISSED_TIMER_OFFSETS_HIST ++ depends on HIGH_RES_TIMERS ++ select GENERIC_TRACER ++ bool "Missed Timer Offsets Histogram" ++ help ++ Generate a histogram of missed timer offsets in microseconds. The ++ histograms are disabled by default. To enable them, write a non-zero ++ number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/missed_timer_offsets ++ ++ The histogram data will be located in the debug file system at ++ ++ /sys/kernel/debug/tracing/latency_hist/missed_timer_offsets ++ ++ If both Scheduling Latency Histogram and Missed Timer Offsets ++ Histogram are selected, additional histogram data will be collected ++ that contain, in addition to the wakeup latency, the timer latency, in ++ case the wakeup was triggered by an expired timer. These histograms ++ are available in the ++ ++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup ++ ++ directory. They reflect the apparent interrupt and scheduling latency ++ and are best suitable to determine the worst-case latency of a given ++ system. To enable these histograms, write a non-zero number to ++ ++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup ++ + config ENABLE_DEFAULT_TRACERS + bool "Trace process context switches and events" + depends on !GENERIC_TRACER +diff -Nur linux-4.1.13.orig/kernel/trace/latency_hist.c linux-4.1.13/kernel/trace/latency_hist.c +--- linux-4.1.13.orig/kernel/trace/latency_hist.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-4.1.13/kernel/trace/latency_hist.c 2015-11-29 09:23:09.653611260 +0100 +@@ -0,0 +1,1178 @@ ++/* ++ * kernel/trace/latency_hist.c ++ * ++ * Add support for histograms of preemption-off latency and ++ * interrupt-off latency and wakeup latency, it depends on ++ * Real-Time Preemption Support. ++ * ++ * Copyright (C) 2005 MontaVista Software, Inc. ++ * Yi Yang <yyang@ch.mvista.com> ++ * ++ * Converted to work with the new latency tracer. ++ * Copyright (C) 2008 Red Hat, Inc. ++ * Steven Rostedt <srostedt@redhat.com> ++ * ++ */ ++#include <linux/module.h> ++#include <linux/debugfs.h> ++#include <linux/seq_file.h> ++#include <linux/percpu.h> ++#include <linux/kallsyms.h> ++#include <linux/uaccess.h> ++#include <linux/sched.h> ++#include <linux/sched/rt.h> ++#include <linux/slab.h> ++#include <linux/atomic.h> ++#include <asm/div64.h> ++ ++#include "trace.h" ++#include <trace/events/sched.h> ++ ++#define NSECS_PER_USECS 1000L ++ ++#define CREATE_TRACE_POINTS ++#include <trace/events/hist.h> ++ ++enum { ++ IRQSOFF_LATENCY = 0, ++ PREEMPTOFF_LATENCY, ++ PREEMPTIRQSOFF_LATENCY, ++ WAKEUP_LATENCY, ++ WAKEUP_LATENCY_SHAREDPRIO, ++ MISSED_TIMER_OFFSETS, ++ TIMERANDWAKEUP_LATENCY, ++ MAX_LATENCY_TYPE, ++}; ++ ++#define MAX_ENTRY_NUM 10240 ++ ++struct hist_data { ++ atomic_t hist_mode; /* 0 log, 1 don't log */ ++ long offset; /* set it to MAX_ENTRY_NUM/2 for a bipolar scale */ ++ long min_lat; ++ long max_lat; ++ unsigned long long below_hist_bound_samples; ++ unsigned long long above_hist_bound_samples; ++ long long accumulate_lat; ++ unsigned long long total_samples; ++ unsigned long long hist_array[MAX_ENTRY_NUM]; ++}; ++ ++struct enable_data { ++ int latency_type; ++ int enabled; ++}; ++ ++static char *latency_hist_dir_root = "latency_hist"; ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++static DEFINE_PER_CPU(struct hist_data, irqsoff_hist); ++static char *irqsoff_hist_dir = "irqsoff"; ++static DEFINE_PER_CPU(cycles_t, hist_irqsoff_start); ++static DEFINE_PER_CPU(int, hist_irqsoff_counting); ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++static DEFINE_PER_CPU(struct hist_data, preemptoff_hist); ++static char *preemptoff_hist_dir = "preemptoff"; ++static DEFINE_PER_CPU(cycles_t, hist_preemptoff_start); ++static DEFINE_PER_CPU(int, hist_preemptoff_counting); ++#endif ++ ++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) ++static DEFINE_PER_CPU(struct hist_data, preemptirqsoff_hist); ++static char *preemptirqsoff_hist_dir = "preemptirqsoff"; ++static DEFINE_PER_CPU(cycles_t, hist_preemptirqsoff_start); ++static DEFINE_PER_CPU(int, hist_preemptirqsoff_counting); ++#endif ++ ++#if defined(CONFIG_PREEMPT_OFF_HIST) || defined(CONFIG_INTERRUPT_OFF_HIST) ++static notrace void probe_preemptirqsoff_hist(void *v, int reason, int start); ++static struct enable_data preemptirqsoff_enabled_data = { ++ .latency_type = PREEMPTIRQSOFF_LATENCY, ++ .enabled = 0, ++}; ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++struct maxlatproc_data { ++ char comm[FIELD_SIZEOF(struct task_struct, comm)]; ++ char current_comm[FIELD_SIZEOF(struct task_struct, comm)]; ++ int pid; ++ int current_pid; ++ int prio; ++ int current_prio; ++ long latency; ++ long timeroffset; ++ cycle_t timestamp; ++}; ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist); ++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist_sharedprio); ++static char *wakeup_latency_hist_dir = "wakeup"; ++static char *wakeup_latency_hist_dir_sharedprio = "sharedprio"; ++static notrace void probe_wakeup_latency_hist_start(void *v, ++ struct task_struct *p); ++static notrace void probe_wakeup_latency_hist_stop(void *v, ++ struct task_struct *prev, struct task_struct *next); ++static notrace void probe_sched_migrate_task(void *, ++ struct task_struct *task, int cpu); ++static struct enable_data wakeup_latency_enabled_data = { ++ .latency_type = WAKEUP_LATENCY, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc); ++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc_sharedprio); ++static DEFINE_PER_CPU(struct task_struct *, wakeup_task); ++static DEFINE_PER_CPU(int, wakeup_sharedprio); ++static unsigned long wakeup_pid; ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++static DEFINE_PER_CPU(struct hist_data, missed_timer_offsets); ++static char *missed_timer_offsets_dir = "missed_timer_offsets"; ++static notrace void probe_hrtimer_interrupt(void *v, int cpu, ++ long long offset, struct task_struct *curr, struct task_struct *task); ++static struct enable_data missed_timer_offsets_enabled_data = { ++ .latency_type = MISSED_TIMER_OFFSETS, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, missed_timer_offsets_maxlatproc); ++static unsigned long missed_timer_offsets_pid; ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static DEFINE_PER_CPU(struct hist_data, timerandwakeup_latency_hist); ++static char *timerandwakeup_latency_hist_dir = "timerandwakeup"; ++static struct enable_data timerandwakeup_enabled_data = { ++ .latency_type = TIMERANDWAKEUP_LATENCY, ++ .enabled = 0, ++}; ++static DEFINE_PER_CPU(struct maxlatproc_data, timerandwakeup_maxlatproc); ++#endif ++ ++void notrace latency_hist(int latency_type, int cpu, long latency, ++ long timeroffset, cycle_t stop, ++ struct task_struct *p) ++{ ++ struct hist_data *my_hist; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ struct maxlatproc_data *mp = NULL; ++#endif ++ ++ if (!cpu_possible(cpu) || latency_type < 0 || ++ latency_type >= MAX_LATENCY_TYPE) ++ return; ++ ++ switch (latency_type) { ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ case IRQSOFF_LATENCY: ++ my_hist = &per_cpu(irqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ case PREEMPTOFF_LATENCY: ++ my_hist = &per_cpu(preemptoff_hist, cpu); ++ break; ++#endif ++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ my_hist = &per_cpu(preemptirqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ my_hist = &per_cpu(wakeup_latency_hist, cpu); ++ mp = &per_cpu(wakeup_maxlatproc, cpu); ++ break; ++ case WAKEUP_LATENCY_SHAREDPRIO: ++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ my_hist = &per_cpu(missed_timer_offsets, cpu); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ my_hist = &per_cpu(timerandwakeup_latency_hist, cpu); ++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); ++ break; ++#endif ++ ++ default: ++ return; ++ } ++ ++ latency += my_hist->offset; ++ ++ if (atomic_read(&my_hist->hist_mode) == 0) ++ return; ++ ++ if (latency < 0 || latency >= MAX_ENTRY_NUM) { ++ if (latency < 0) ++ my_hist->below_hist_bound_samples++; ++ else ++ my_hist->above_hist_bound_samples++; ++ } else ++ my_hist->hist_array[latency]++; ++ ++ if (unlikely(latency > my_hist->max_lat || ++ my_hist->min_lat == LONG_MAX)) { ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ if (latency_type == WAKEUP_LATENCY || ++ latency_type == WAKEUP_LATENCY_SHAREDPRIO || ++ latency_type == MISSED_TIMER_OFFSETS || ++ latency_type == TIMERANDWAKEUP_LATENCY) { ++ strncpy(mp->comm, p->comm, sizeof(mp->comm)); ++ strncpy(mp->current_comm, current->comm, ++ sizeof(mp->current_comm)); ++ mp->pid = task_pid_nr(p); ++ mp->current_pid = task_pid_nr(current); ++ mp->prio = p->prio; ++ mp->current_prio = current->prio; ++ mp->latency = latency; ++ mp->timeroffset = timeroffset; ++ mp->timestamp = stop; ++ } ++#endif ++ my_hist->max_lat = latency; ++ } ++ if (unlikely(latency < my_hist->min_lat)) ++ my_hist->min_lat = latency; ++ my_hist->total_samples++; ++ my_hist->accumulate_lat += latency; ++} ++ ++static void *l_start(struct seq_file *m, loff_t *pos) ++{ ++ loff_t *index_ptr = NULL; ++ loff_t index = *pos; ++ struct hist_data *my_hist = m->private; ++ ++ if (index == 0) { ++ char minstr[32], avgstr[32], maxstr[32]; ++ ++ atomic_dec(&my_hist->hist_mode); ++ ++ if (likely(my_hist->total_samples)) { ++ long avg = (long) div64_s64(my_hist->accumulate_lat, ++ my_hist->total_samples); ++ snprintf(minstr, sizeof(minstr), "%ld", ++ my_hist->min_lat - my_hist->offset); ++ snprintf(avgstr, sizeof(avgstr), "%ld", ++ avg - my_hist->offset); ++ snprintf(maxstr, sizeof(maxstr), "%ld", ++ my_hist->max_lat - my_hist->offset); ++ } else { ++ strcpy(minstr, "<undef>"); ++ strcpy(avgstr, minstr); ++ strcpy(maxstr, minstr); ++ } ++ ++ seq_printf(m, "#Minimum latency: %s microseconds\n" ++ "#Average latency: %s microseconds\n" ++ "#Maximum latency: %s microseconds\n" ++ "#Total samples: %llu\n" ++ "#There are %llu samples lower than %ld" ++ " microseconds.\n" ++ "#There are %llu samples greater or equal" ++ " than %ld microseconds.\n" ++ "#usecs\t%16s\n", ++ minstr, avgstr, maxstr, ++ my_hist->total_samples, ++ my_hist->below_hist_bound_samples, ++ -my_hist->offset, ++ my_hist->above_hist_bound_samples, ++ MAX_ENTRY_NUM - my_hist->offset, ++ "samples"); ++ } ++ if (index < MAX_ENTRY_NUM) { ++ index_ptr = kmalloc(sizeof(loff_t), GFP_KERNEL); ++ if (index_ptr) ++ *index_ptr = index; ++ } ++ ++ return index_ptr; ++} ++ ++static void *l_next(struct seq_file *m, void *p, loff_t *pos) ++{ ++ loff_t *index_ptr = p; ++ struct hist_data *my_hist = m->private; ++ ++ if (++*pos >= MAX_ENTRY_NUM) { ++ atomic_inc(&my_hist->hist_mode); ++ return NULL; ++ } ++ *index_ptr = *pos; ++ return index_ptr; ++} ++ ++static void l_stop(struct seq_file *m, void *p) ++{ ++ kfree(p); ++} ++ ++static int l_show(struct seq_file *m, void *p) ++{ ++ int index = *(loff_t *) p; ++ struct hist_data *my_hist = m->private; ++ ++ seq_printf(m, "%6ld\t%16llu\n", index - my_hist->offset, ++ my_hist->hist_array[index]); ++ return 0; ++} ++ ++static const struct seq_operations latency_hist_seq_op = { ++ .start = l_start, ++ .next = l_next, ++ .stop = l_stop, ++ .show = l_show ++}; ++ ++static int latency_hist_open(struct inode *inode, struct file *file) ++{ ++ int ret; ++ ++ ret = seq_open(file, &latency_hist_seq_op); ++ if (!ret) { ++ struct seq_file *seq = file->private_data; ++ seq->private = inode->i_private; ++ } ++ return ret; ++} ++ ++static const struct file_operations latency_hist_fops = { ++ .open = latency_hist_open, ++ .read = seq_read, ++ .llseek = seq_lseek, ++ .release = seq_release, ++}; ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static void clear_maxlatprocdata(struct maxlatproc_data *mp) ++{ ++ mp->comm[0] = mp->current_comm[0] = '\0'; ++ mp->prio = mp->current_prio = mp->pid = mp->current_pid = ++ mp->latency = mp->timeroffset = -1; ++ mp->timestamp = 0; ++} ++#endif ++ ++static void hist_reset(struct hist_data *hist) ++{ ++ atomic_dec(&hist->hist_mode); ++ ++ memset(hist->hist_array, 0, sizeof(hist->hist_array)); ++ hist->below_hist_bound_samples = 0ULL; ++ hist->above_hist_bound_samples = 0ULL; ++ hist->min_lat = LONG_MAX; ++ hist->max_lat = LONG_MIN; ++ hist->total_samples = 0ULL; ++ hist->accumulate_lat = 0LL; ++ ++ atomic_inc(&hist->hist_mode); ++} ++ ++static ssize_t ++latency_hist_reset(struct file *file, const char __user *a, ++ size_t size, loff_t *off) ++{ ++ int cpu; ++ struct hist_data *hist = NULL; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ struct maxlatproc_data *mp = NULL; ++#endif ++ off_t latency_type = (off_t) file->private_data; ++ ++ for_each_online_cpu(cpu) { ++ ++ switch (latency_type) { ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ case PREEMPTOFF_LATENCY: ++ hist = &per_cpu(preemptoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ case IRQSOFF_LATENCY: ++ hist = &per_cpu(irqsoff_hist, cpu); ++ break; ++#endif ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ hist = &per_cpu(preemptirqsoff_hist, cpu); ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ hist = &per_cpu(wakeup_latency_hist, cpu); ++ mp = &per_cpu(wakeup_maxlatproc, cpu); ++ break; ++ case WAKEUP_LATENCY_SHAREDPRIO: ++ hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu); ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu); ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ hist = &per_cpu(missed_timer_offsets, cpu); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu); ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ hist = &per_cpu(timerandwakeup_latency_hist, cpu); ++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu); ++ break; ++#endif ++ } ++ ++ hist_reset(hist); ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ if (latency_type == WAKEUP_LATENCY || ++ latency_type == WAKEUP_LATENCY_SHAREDPRIO || ++ latency_type == MISSED_TIMER_OFFSETS || ++ latency_type == TIMERANDWAKEUP_LATENCY) ++ clear_maxlatprocdata(mp); ++#endif ++ } ++ ++ return size; ++} ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static ssize_t ++show_pid(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ int r; ++ unsigned long *this_pid = file->private_data; ++ ++ r = snprintf(buf, sizeof(buf), "%lu\n", *this_pid); ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++} ++ ++static ssize_t do_pid(struct file *file, const char __user *ubuf, ++ size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ unsigned long pid; ++ unsigned long *this_pid = file->private_data; ++ ++ if (cnt >= sizeof(buf)) ++ return -EINVAL; ++ ++ if (copy_from_user(&buf, ubuf, cnt)) ++ return -EFAULT; ++ ++ buf[cnt] = '\0'; ++ ++ if (kstrtoul(buf, 10, &pid)) ++ return -EINVAL; ++ ++ *this_pid = pid; ++ ++ return cnt; ++} ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static ssize_t ++show_maxlatproc(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ int r; ++ struct maxlatproc_data *mp = file->private_data; ++ int strmaxlen = (TASK_COMM_LEN * 2) + (8 * 8); ++ unsigned long long t; ++ unsigned long usecs, secs; ++ char *buf; ++ ++ if (mp->pid == -1 || mp->current_pid == -1) { ++ buf = "(none)\n"; ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, ++ strlen(buf)); ++ } ++ ++ buf = kmalloc(strmaxlen, GFP_KERNEL); ++ if (buf == NULL) ++ return -ENOMEM; ++ ++ t = ns2usecs(mp->timestamp); ++ usecs = do_div(t, USEC_PER_SEC); ++ secs = (unsigned long) t; ++ r = snprintf(buf, strmaxlen, ++ "%d %d %ld (%ld) %s <- %d %d %s %lu.%06lu\n", mp->pid, ++ MAX_RT_PRIO-1 - mp->prio, mp->latency, mp->timeroffset, mp->comm, ++ mp->current_pid, MAX_RT_PRIO-1 - mp->current_prio, mp->current_comm, ++ secs, usecs); ++ r = simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++ kfree(buf); ++ return r; ++} ++#endif ++ ++static ssize_t ++show_enable(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ struct enable_data *ed = file->private_data; ++ int r; ++ ++ r = snprintf(buf, sizeof(buf), "%d\n", ed->enabled); ++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); ++} ++ ++static ssize_t ++do_enable(struct file *file, const char __user *ubuf, size_t cnt, loff_t *ppos) ++{ ++ char buf[64]; ++ long enable; ++ struct enable_data *ed = file->private_data; ++ ++ if (cnt >= sizeof(buf)) ++ return -EINVAL; ++ ++ if (copy_from_user(&buf, ubuf, cnt)) ++ return -EFAULT; ++ ++ buf[cnt] = 0; ++ ++ if (kstrtoul(buf, 10, &enable)) ++ return -EINVAL; ++ ++ if ((enable && ed->enabled) || (!enable && !ed->enabled)) ++ return cnt; ++ ++ if (enable) { ++ int ret; ++ ++ switch (ed->latency_type) { ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ ret = register_trace_preemptirqsoff_hist( ++ probe_preemptirqsoff_hist, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_preemptirqsoff_hist " ++ "to trace_preemptirqsoff_hist\n"); ++ return ret; ++ } ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ ret = register_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_start " ++ "to trace_sched_wakeup\n"); ++ return ret; ++ } ++ ret = register_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_start " ++ "to trace_sched_wakeup_new\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ return ret; ++ } ++ ret = register_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_wakeup_latency_hist_stop " ++ "to trace_sched_switch\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ return ret; ++ } ++ ret = register_trace_sched_migrate_task( ++ probe_sched_migrate_task, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_sched_migrate_task " ++ "to trace_sched_migrate_task\n"); ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ return ret; ++ } ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ ret = register_trace_hrtimer_interrupt( ++ probe_hrtimer_interrupt, NULL); ++ if (ret) { ++ pr_info("wakeup trace: Couldn't assign " ++ "probe_hrtimer_interrupt " ++ "to trace_hrtimer_interrupt\n"); ++ return ret; ++ } ++ break; ++#endif ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ case TIMERANDWAKEUP_LATENCY: ++ if (!wakeup_latency_enabled_data.enabled || ++ !missed_timer_offsets_enabled_data.enabled) ++ return -EINVAL; ++ break; ++#endif ++ default: ++ break; ++ } ++ } else { ++ switch (ed->latency_type) { ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ case PREEMPTIRQSOFF_LATENCY: ++ { ++ int cpu; ++ ++ unregister_trace_preemptirqsoff_hist( ++ probe_preemptirqsoff_hist, NULL); ++ for_each_online_cpu(cpu) { ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ per_cpu(hist_irqsoff_counting, ++ cpu) = 0; ++#endif ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ per_cpu(hist_preemptoff_counting, ++ cpu) = 0; ++#endif ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ per_cpu(hist_preemptirqsoff_counting, ++ cpu) = 0; ++#endif ++ } ++ } ++ break; ++#endif ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ case WAKEUP_LATENCY: ++ { ++ int cpu; ++ ++ unregister_trace_sched_wakeup( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_wakeup_new( ++ probe_wakeup_latency_hist_start, NULL); ++ unregister_trace_sched_switch( ++ probe_wakeup_latency_hist_stop, NULL); ++ unregister_trace_sched_migrate_task( ++ probe_sched_migrate_task, NULL); ++ ++ for_each_online_cpu(cpu) { ++ per_cpu(wakeup_task, cpu) = NULL; ++ per_cpu(wakeup_sharedprio, cpu) = 0; ++ } ++ } ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ timerandwakeup_enabled_data.enabled = 0; ++#endif ++ break; ++#endif ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ case MISSED_TIMER_OFFSETS: ++ unregister_trace_hrtimer_interrupt( ++ probe_hrtimer_interrupt, NULL); ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ timerandwakeup_enabled_data.enabled = 0; ++#endif ++ break; ++#endif ++ default: ++ break; ++ } ++ } ++ ed->enabled = enable; ++ return cnt; ++} ++ ++static const struct file_operations latency_hist_reset_fops = { ++ .open = tracing_open_generic, ++ .write = latency_hist_reset, ++}; ++ ++static const struct file_operations enable_fops = { ++ .open = tracing_open_generic, ++ .read = show_enable, ++ .write = do_enable, ++}; ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++static const struct file_operations pid_fops = { ++ .open = tracing_open_generic, ++ .read = show_pid, ++ .write = do_pid, ++}; ++ ++static const struct file_operations maxlatproc_fops = { ++ .open = tracing_open_generic, ++ .read = show_maxlatproc, ++}; ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++static notrace void probe_preemptirqsoff_hist(void *v, int reason, ++ int starthist) ++{ ++ int cpu = raw_smp_processor_id(); ++ int time_set = 0; ++ ++ if (starthist) { ++ cycle_t uninitialized_var(start); ++ ++ if (!preempt_count() && !irqs_disabled()) ++ return; ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ if ((reason == IRQS_OFF || reason == TRACE_START) && ++ !per_cpu(hist_irqsoff_counting, cpu)) { ++ per_cpu(hist_irqsoff_counting, cpu) = 1; ++ start = ftrace_now(cpu); ++ time_set++; ++ per_cpu(hist_irqsoff_start, cpu) = start; ++ } ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ if ((reason == PREEMPT_OFF || reason == TRACE_START) && ++ !per_cpu(hist_preemptoff_counting, cpu)) { ++ per_cpu(hist_preemptoff_counting, cpu) = 1; ++ if (!(time_set++)) ++ start = ftrace_now(cpu); ++ per_cpu(hist_preemptoff_start, cpu) = start; ++ } ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ if (per_cpu(hist_irqsoff_counting, cpu) && ++ per_cpu(hist_preemptoff_counting, cpu) && ++ !per_cpu(hist_preemptirqsoff_counting, cpu)) { ++ per_cpu(hist_preemptirqsoff_counting, cpu) = 1; ++ if (!time_set) ++ start = ftrace_now(cpu); ++ per_cpu(hist_preemptirqsoff_start, cpu) = start; ++ } ++#endif ++ } else { ++ cycle_t uninitialized_var(stop); ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ if ((reason == IRQS_ON || reason == TRACE_STOP) && ++ per_cpu(hist_irqsoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_irqsoff_start, cpu); ++ ++ stop = ftrace_now(cpu); ++ time_set++; ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(IRQSOFF_LATENCY, cpu, latency, 0, ++ stop, NULL); ++ } ++ per_cpu(hist_irqsoff_counting, cpu) = 0; ++ } ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ if ((reason == PREEMPT_ON || reason == TRACE_STOP) && ++ per_cpu(hist_preemptoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_preemptoff_start, cpu); ++ ++ if (!(time_set++)) ++ stop = ftrace_now(cpu); ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(PREEMPTOFF_LATENCY, cpu, latency, ++ 0, stop, NULL); ++ } ++ per_cpu(hist_preemptoff_counting, cpu) = 0; ++ } ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ if ((!per_cpu(hist_irqsoff_counting, cpu) || ++ !per_cpu(hist_preemptoff_counting, cpu)) && ++ per_cpu(hist_preemptirqsoff_counting, cpu)) { ++ cycle_t start = per_cpu(hist_preemptirqsoff_start, cpu); ++ ++ if (!time_set) ++ stop = ftrace_now(cpu); ++ if (start) { ++ long latency = ((long) (stop - start)) / ++ NSECS_PER_USECS; ++ ++ latency_hist(PREEMPTIRQSOFF_LATENCY, cpu, ++ latency, 0, stop, NULL); ++ } ++ per_cpu(hist_preemptirqsoff_counting, cpu) = 0; ++ } ++#endif ++ } ++} ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++static DEFINE_RAW_SPINLOCK(wakeup_lock); ++static notrace void probe_sched_migrate_task(void *v, struct task_struct *task, ++ int cpu) ++{ ++ int old_cpu = task_cpu(task); ++ ++ if (cpu != old_cpu) { ++ unsigned long flags; ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, old_cpu); ++ if (task == cpu_wakeup_task) { ++ put_task_struct(cpu_wakeup_task); ++ per_cpu(wakeup_task, old_cpu) = NULL; ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = task; ++ get_task_struct(cpu_wakeup_task); ++ } ++ ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++ } ++} ++ ++static notrace void probe_wakeup_latency_hist_start(void *v, ++ struct task_struct *p) ++{ ++ unsigned long flags; ++ struct task_struct *curr = current; ++ int cpu = task_cpu(p); ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu); ++ ++ if (wakeup_pid) { ++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || ++ p->prio == curr->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ if (likely(wakeup_pid != task_pid_nr(p))) ++ goto out; ++ } else { ++ if (likely(!rt_task(p)) || ++ (cpu_wakeup_task && p->prio > cpu_wakeup_task->prio) || ++ p->prio > curr->prio) ++ goto out; ++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) || ++ p->prio == curr->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ } ++ ++ if (cpu_wakeup_task) ++ put_task_struct(cpu_wakeup_task); ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = p; ++ get_task_struct(cpu_wakeup_task); ++ cpu_wakeup_task->preempt_timestamp_hist = ++ ftrace_now(raw_smp_processor_id()); ++out: ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++} ++ ++static notrace void probe_wakeup_latency_hist_stop(void *v, ++ struct task_struct *prev, struct task_struct *next) ++{ ++ unsigned long flags; ++ int cpu = task_cpu(next); ++ long latency; ++ cycle_t stop; ++ struct task_struct *cpu_wakeup_task; ++ ++ raw_spin_lock_irqsave(&wakeup_lock, flags); ++ ++ cpu_wakeup_task = per_cpu(wakeup_task, cpu); ++ ++ if (cpu_wakeup_task == NULL) ++ goto out; ++ ++ /* Already running? */ ++ if (unlikely(current == cpu_wakeup_task)) ++ goto out_reset; ++ ++ if (next != cpu_wakeup_task) { ++ if (next->prio < cpu_wakeup_task->prio) ++ goto out_reset; ++ ++ if (next->prio == cpu_wakeup_task->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ ++ goto out; ++ } ++ ++ if (current->prio == cpu_wakeup_task->prio) ++ per_cpu(wakeup_sharedprio, cpu) = 1; ++ ++ /* ++ * The task we are waiting for is about to be switched to. ++ * Calculate latency and store it in histogram. ++ */ ++ stop = ftrace_now(raw_smp_processor_id()); ++ ++ latency = ((long) (stop - next->preempt_timestamp_hist)) / ++ NSECS_PER_USECS; ++ ++ if (per_cpu(wakeup_sharedprio, cpu)) { ++ latency_hist(WAKEUP_LATENCY_SHAREDPRIO, cpu, latency, 0, stop, ++ next); ++ per_cpu(wakeup_sharedprio, cpu) = 0; ++ } else { ++ latency_hist(WAKEUP_LATENCY, cpu, latency, 0, stop, next); ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ if (timerandwakeup_enabled_data.enabled) { ++ latency_hist(TIMERANDWAKEUP_LATENCY, cpu, ++ next->timer_offset + latency, next->timer_offset, ++ stop, next); ++ } ++#endif ++ } ++ ++out_reset: ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ next->timer_offset = 0; ++#endif ++ put_task_struct(cpu_wakeup_task); ++ per_cpu(wakeup_task, cpu) = NULL; ++out: ++ raw_spin_unlock_irqrestore(&wakeup_lock, flags); ++} ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++static notrace void probe_hrtimer_interrupt(void *v, int cpu, ++ long long latency_ns, struct task_struct *curr, ++ struct task_struct *task) ++{ ++ if (latency_ns <= 0 && task != NULL && rt_task(task) && ++ (task->prio < curr->prio || ++ (task->prio == curr->prio && ++ !cpumask_test_cpu(cpu, &task->cpus_allowed)))) { ++ long latency; ++ cycle_t now; ++ ++ if (missed_timer_offsets_pid) { ++ if (likely(missed_timer_offsets_pid != ++ task_pid_nr(task))) ++ return; ++ } ++ ++ now = ftrace_now(cpu); ++ latency = (long) div_s64(-latency_ns, NSECS_PER_USECS); ++ latency_hist(MISSED_TIMER_OFFSETS, cpu, latency, latency, now, ++ task); ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ task->timer_offset = latency; ++#endif ++ } ++} ++#endif ++ ++static __init int latency_hist_init(void) ++{ ++ struct dentry *latency_hist_root = NULL; ++ struct dentry *dentry; ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ struct dentry *dentry_sharedprio; ++#endif ++ struct dentry *entry; ++ struct dentry *enable_root; ++ int i = 0; ++ struct hist_data *my_hist; ++ char name[64]; ++ char *cpufmt = "CPU%d"; ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ char *cpufmt_maxlatproc = "max_latency-CPU%d"; ++ struct maxlatproc_data *mp = NULL; ++#endif ++ ++ dentry = tracing_init_dentry(); ++ latency_hist_root = debugfs_create_dir(latency_hist_dir_root, dentry); ++ enable_root = debugfs_create_dir("enable", latency_hist_root); ++ ++#ifdef CONFIG_INTERRUPT_OFF_HIST ++ dentry = debugfs_create_dir(irqsoff_hist_dir, latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(irqsoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(irqsoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)IRQSOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#ifdef CONFIG_PREEMPT_OFF_HIST ++ dentry = debugfs_create_dir(preemptoff_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(preemptoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(preemptoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)PREEMPTOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST) ++ dentry = debugfs_create_dir(preemptirqsoff_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(preemptirqsoff_hist, i), &latency_hist_fops); ++ my_hist = &per_cpu(preemptirqsoff_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)PREEMPTIRQSOFF_LATENCY, &latency_hist_reset_fops); ++#endif ++ ++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST) ++ entry = debugfs_create_file("preemptirqsoff", 0644, ++ enable_root, (void *)&preemptirqsoff_enabled_data, ++ &enable_fops); ++#endif ++ ++#ifdef CONFIG_WAKEUP_LATENCY_HIST ++ dentry = debugfs_create_dir(wakeup_latency_hist_dir, ++ latency_hist_root); ++ dentry_sharedprio = debugfs_create_dir( ++ wakeup_latency_hist_dir_sharedprio, dentry); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(wakeup_latency_hist, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(wakeup_latency_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ entry = debugfs_create_file(name, 0444, dentry_sharedprio, ++ &per_cpu(wakeup_latency_hist_sharedprio, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ ++ mp = &per_cpu(wakeup_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ ++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, i); ++ entry = debugfs_create_file(name, 0444, dentry_sharedprio, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("pid", 0644, dentry, ++ (void *)&wakeup_pid, &pid_fops); ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)WAKEUP_LATENCY, &latency_hist_reset_fops); ++ entry = debugfs_create_file("reset", 0644, dentry_sharedprio, ++ (void *)WAKEUP_LATENCY_SHAREDPRIO, &latency_hist_reset_fops); ++ entry = debugfs_create_file("wakeup", 0644, ++ enable_root, (void *)&wakeup_latency_enabled_data, ++ &enable_fops); ++#endif ++ ++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST ++ dentry = debugfs_create_dir(missed_timer_offsets_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(missed_timer_offsets, i), &latency_hist_fops); ++ my_hist = &per_cpu(missed_timer_offsets, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ mp = &per_cpu(missed_timer_offsets_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("pid", 0644, dentry, ++ (void *)&missed_timer_offsets_pid, &pid_fops); ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)MISSED_TIMER_OFFSETS, &latency_hist_reset_fops); ++ entry = debugfs_create_file("missed_timer_offsets", 0644, ++ enable_root, (void *)&missed_timer_offsets_enabled_data, ++ &enable_fops); ++#endif ++ ++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \ ++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST) ++ dentry = debugfs_create_dir(timerandwakeup_latency_hist_dir, ++ latency_hist_root); ++ for_each_possible_cpu(i) { ++ sprintf(name, cpufmt, i); ++ entry = debugfs_create_file(name, 0444, dentry, ++ &per_cpu(timerandwakeup_latency_hist, i), ++ &latency_hist_fops); ++ my_hist = &per_cpu(timerandwakeup_latency_hist, i); ++ atomic_set(&my_hist->hist_mode, 1); ++ my_hist->min_lat = LONG_MAX; ++ ++ sprintf(name, cpufmt_maxlatproc, i); ++ mp = &per_cpu(timerandwakeup_maxlatproc, i); ++ entry = debugfs_create_file(name, 0444, dentry, mp, ++ &maxlatproc_fops); ++ clear_maxlatprocdata(mp); ++ } ++ entry = debugfs_create_file("reset", 0644, dentry, ++ (void *)TIMERANDWAKEUP_LATENCY, &latency_hist_reset_fops); ++ entry = debugfs_create_file("timerandwakeup", 0644, ++ enable_root, (void *)&timerandwakeup_enabled_data, ++ &enable_fops); ++#endif ++ return 0; ++} ++ ++device_initcall(latency_hist_init); +diff -Nur linux-4.1.13.orig/kernel/trace/Makefile linux-4.1.13/kernel/trace/Makefile +--- linux-4.1.13.orig/kernel/trace/Makefile 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/Makefile 2015-11-29 09:23:09.653611260 +0100 +@@ -36,6 +36,10 @@ + obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o + obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o + obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o ++obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o ++obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o ++obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o + obj-$(CONFIG_NOP_TRACER) += trace_nop.o + obj-$(CONFIG_STACK_TRACER) += trace_stack.o + obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o +diff -Nur linux-4.1.13.orig/kernel/trace/trace.c linux-4.1.13/kernel/trace/trace.c +--- linux-4.1.13.orig/kernel/trace/trace.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace.c 2015-11-29 09:23:09.653611260 +0100 +@@ -1630,6 +1630,7 @@ + struct task_struct *tsk = current; + + entry->preempt_count = pc & 0xff; ++ entry->preempt_lazy_count = preempt_lazy_count(); + entry->pid = (tsk) ? tsk->pid : 0; + entry->flags = + #ifdef CONFIG_TRACE_IRQFLAGS_SUPPORT +@@ -1639,8 +1640,11 @@ + #endif + ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) | + ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | +- (tif_need_resched() ? TRACE_FLAG_NEED_RESCHED : 0) | ++ (tif_need_resched_now() ? TRACE_FLAG_NEED_RESCHED : 0) | ++ (need_resched_lazy() ? TRACE_FLAG_NEED_RESCHED_LAZY : 0) | + (test_preempt_need_resched() ? TRACE_FLAG_PREEMPT_RESCHED : 0); ++ ++ entry->migrate_disable = (tsk) ? __migrate_disabled(tsk) & 0xFF : 0; + } + EXPORT_SYMBOL_GPL(tracing_generic_entry_update); + +@@ -2558,14 +2562,17 @@ + + static void print_lat_help_header(struct seq_file *m) + { +- seq_puts(m, "# _------=> CPU# \n" +- "# / _-----=> irqs-off \n" +- "# | / _----=> need-resched \n" +- "# || / _---=> hardirq/softirq \n" +- "# ||| / _--=> preempt-depth \n" +- "# |||| / delay \n" +- "# cmd pid ||||| time | caller \n" +- "# \\ / ||||| \\ | / \n"); ++ seq_puts(m, "# _--------=> CPU# \n" ++ "# / _-------=> irqs-off \n" ++ "# | / _------=> need-resched \n" ++ "# || / _-----=> need-resched_lazy \n" ++ "# ||| / _----=> hardirq/softirq \n" ++ "# |||| / _---=> preempt-depth \n" ++ "# ||||| / _--=> preempt-lazy-depth\n" ++ "# |||||| / _-=> migrate-disable \n" ++ "# ||||||| / delay \n" ++ "# cmd pid |||||||| time | caller \n" ++ "# \\ / |||||||| \\ | / \n"); + } + + static void print_event_info(struct trace_buffer *buf, struct seq_file *m) +@@ -2591,11 +2598,14 @@ + print_event_info(buf, m); + seq_puts(m, "# _-----=> irqs-off\n" + "# / _----=> need-resched\n" +- "# | / _---=> hardirq/softirq\n" +- "# || / _--=> preempt-depth\n" +- "# ||| / delay\n" +- "# TASK-PID CPU# |||| TIMESTAMP FUNCTION\n" +- "# | | | |||| | |\n"); ++ "# |/ _-----=> need-resched_lazy\n" ++ "# || / _---=> hardirq/softirq\n" ++ "# ||| / _--=> preempt-depth\n" ++ "# |||| /_--=> preempt-lazy-depth\n" ++ "# ||||| _-=> migrate-disable \n" ++ "# ||||| / delay\n" ++ "# TASK-PID CPU# |||||| TIMESTAMP FUNCTION\n" ++ "# | | | |||||| | |\n"); + } + + void +diff -Nur linux-4.1.13.orig/kernel/trace/trace_events.c linux-4.1.13/kernel/trace/trace_events.c +--- linux-4.1.13.orig/kernel/trace/trace_events.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace_events.c 2015-11-29 09:23:09.653611260 +0100 +@@ -162,6 +162,8 @@ + __common_field(unsigned char, flags); + __common_field(unsigned char, preempt_count); + __common_field(int, pid); ++ __common_field(unsigned short, migrate_disable); ++ __common_field(unsigned short, padding); + + return ret; + } +diff -Nur linux-4.1.13.orig/kernel/trace/trace.h linux-4.1.13/kernel/trace/trace.h +--- linux-4.1.13.orig/kernel/trace/trace.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace.h 2015-11-29 09:23:09.653611260 +0100 +@@ -120,6 +120,7 @@ + * NEED_RESCHED - reschedule is requested + * HARDIRQ - inside an interrupt handler + * SOFTIRQ - inside a softirq handler ++ * NEED_RESCHED_LAZY - lazy reschedule is requested + */ + enum trace_flag_type { + TRACE_FLAG_IRQS_OFF = 0x01, +@@ -128,6 +129,7 @@ + TRACE_FLAG_HARDIRQ = 0x08, + TRACE_FLAG_SOFTIRQ = 0x10, + TRACE_FLAG_PREEMPT_RESCHED = 0x20, ++ TRACE_FLAG_NEED_RESCHED_LAZY = 0x40, + }; + + #define TRACE_BUF_SIZE 1024 +diff -Nur linux-4.1.13.orig/kernel/trace/trace_irqsoff.c linux-4.1.13/kernel/trace/trace_irqsoff.c +--- linux-4.1.13.orig/kernel/trace/trace_irqsoff.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace_irqsoff.c 2015-11-29 09:23:09.653611260 +0100 +@@ -13,6 +13,7 @@ + #include <linux/uaccess.h> + #include <linux/module.h> + #include <linux/ftrace.h> ++#include <trace/events/hist.h> + + #include "trace.h" + +@@ -433,11 +434,13 @@ + { + if (preempt_trace() || irq_trace()) + start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); ++ trace_preemptirqsoff_hist(TRACE_START, 1); + } + EXPORT_SYMBOL_GPL(start_critical_timings); + + void stop_critical_timings(void) + { ++ trace_preemptirqsoff_hist(TRACE_STOP, 0); + if (preempt_trace() || irq_trace()) + stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); + } +@@ -447,6 +450,7 @@ + #ifdef CONFIG_PROVE_LOCKING + void time_hardirqs_on(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist_rcuidle(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(a0, a1); + } +@@ -455,6 +459,7 @@ + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(a0, a1); ++ trace_preemptirqsoff_hist_rcuidle(IRQS_OFF, 1); + } + + #else /* !CONFIG_PROVE_LOCKING */ +@@ -480,6 +485,7 @@ + */ + void trace_hardirqs_on(void) + { ++ trace_preemptirqsoff_hist(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1); + } +@@ -489,11 +495,13 @@ + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(CALLER_ADDR0, CALLER_ADDR1); ++ trace_preemptirqsoff_hist(IRQS_OFF, 1); + } + EXPORT_SYMBOL(trace_hardirqs_off); + + __visible void trace_hardirqs_on_caller(unsigned long caller_addr) + { ++ trace_preemptirqsoff_hist(IRQS_ON, 0); + if (!preempt_trace() && irq_trace()) + stop_critical_timing(CALLER_ADDR0, caller_addr); + } +@@ -503,6 +511,7 @@ + { + if (!preempt_trace() && irq_trace()) + start_critical_timing(CALLER_ADDR0, caller_addr); ++ trace_preemptirqsoff_hist(IRQS_OFF, 1); + } + EXPORT_SYMBOL(trace_hardirqs_off_caller); + +@@ -512,12 +521,14 @@ + #ifdef CONFIG_PREEMPT_TRACER + void trace_preempt_on(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist(PREEMPT_ON, 0); + if (preempt_trace() && !irq_trace()) + stop_critical_timing(a0, a1); + } + + void trace_preempt_off(unsigned long a0, unsigned long a1) + { ++ trace_preemptirqsoff_hist(PREEMPT_ON, 1); + if (preempt_trace() && !irq_trace()) + start_critical_timing(a0, a1); + } +diff -Nur linux-4.1.13.orig/kernel/trace/trace_output.c linux-4.1.13/kernel/trace/trace_output.c +--- linux-4.1.13.orig/kernel/trace/trace_output.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace_output.c 2015-11-29 09:23:09.653611260 +0100 +@@ -430,6 +430,7 @@ + { + char hardsoft_irq; + char need_resched; ++ char need_resched_lazy; + char irqs_off; + int hardirq; + int softirq; +@@ -457,6 +458,8 @@ + need_resched = '.'; + break; + } ++ need_resched_lazy = ++ (entry->flags & TRACE_FLAG_NEED_RESCHED_LAZY) ? 'L' : '.'; + + hardsoft_irq = + (hardirq && softirq) ? 'H' : +@@ -464,14 +467,25 @@ + softirq ? 's' : + '.'; + +- trace_seq_printf(s, "%c%c%c", +- irqs_off, need_resched, hardsoft_irq); ++ trace_seq_printf(s, "%c%c%c%c", ++ irqs_off, need_resched, need_resched_lazy, ++ hardsoft_irq); + + if (entry->preempt_count) + trace_seq_printf(s, "%x", entry->preempt_count); + else + trace_seq_putc(s, '.'); + ++ if (entry->preempt_lazy_count) ++ trace_seq_printf(s, "%x", entry->preempt_lazy_count); ++ else ++ trace_seq_putc(s, '.'); ++ ++ if (entry->migrate_disable) ++ trace_seq_printf(s, "%x", entry->migrate_disable); ++ else ++ trace_seq_putc(s, '.'); ++ + return !trace_seq_has_overflowed(s); + } + +diff -Nur linux-4.1.13.orig/kernel/trace/trace_sched_switch.c linux-4.1.13/kernel/trace/trace_sched_switch.c +--- linux-4.1.13.orig/kernel/trace/trace_sched_switch.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace_sched_switch.c 2015-11-29 09:23:09.653611260 +0100 +@@ -26,7 +26,7 @@ + } + + static void +-probe_sched_wakeup(void *ignore, struct task_struct *wakee, int success) ++probe_sched_wakeup(void *ignore, struct task_struct *wakee) + { + if (unlikely(!sched_ref)) + return; +diff -Nur linux-4.1.13.orig/kernel/trace/trace_sched_wakeup.c linux-4.1.13/kernel/trace/trace_sched_wakeup.c +--- linux-4.1.13.orig/kernel/trace/trace_sched_wakeup.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/trace/trace_sched_wakeup.c 2015-11-29 09:23:09.657610995 +0100 +@@ -514,7 +514,7 @@ + } + + static void +-probe_wakeup(void *ignore, struct task_struct *p, int success) ++probe_wakeup(void *ignore, struct task_struct *p) + { + struct trace_array_cpu *data; + int cpu = smp_processor_id(); +diff -Nur linux-4.1.13.orig/kernel/user.c linux-4.1.13/kernel/user.c +--- linux-4.1.13.orig/kernel/user.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/user.c 2015-11-29 09:23:09.657610995 +0100 +@@ -161,11 +161,11 @@ + if (!up) + return; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + if (atomic_dec_and_lock(&up->__count, &uidhash_lock)) + free_user(up, flags); + else +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + + struct user_struct *alloc_uid(kuid_t uid) +diff -Nur linux-4.1.13.orig/kernel/watchdog.c linux-4.1.13/kernel/watchdog.c +--- linux-4.1.13.orig/kernel/watchdog.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/watchdog.c 2015-11-29 09:23:09.657610995 +0100 +@@ -262,6 +262,8 @@ + + #ifdef CONFIG_HARDLOCKUP_DETECTOR + ++static DEFINE_RAW_SPINLOCK(watchdog_output_lock); ++ + static struct perf_event_attr wd_hw_attr = { + .type = PERF_TYPE_HARDWARE, + .config = PERF_COUNT_HW_CPU_CYCLES, +@@ -295,13 +297,21 @@ + /* only print hardlockups once */ + if (__this_cpu_read(hard_watchdog_warn) == true) + return; ++ /* ++ * If early-printk is enabled then make sure we do not ++ * lock up in printk() and kill console logging: ++ */ ++ printk_kill(); + +- if (hardlockup_panic) ++ if (hardlockup_panic) { + panic("Watchdog detected hard LOCKUP on cpu %d", + this_cpu); +- else ++ } else { ++ raw_spin_lock(&watchdog_output_lock); + WARN(1, "Watchdog detected hard LOCKUP on cpu %d", + this_cpu); ++ raw_spin_unlock(&watchdog_output_lock); ++ } + + __this_cpu_write(hard_watchdog_warn, true); + return; +@@ -444,6 +454,7 @@ + /* kick off the timer for the hardlockup detector */ + hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hrtimer->function = watchdog_timer_fn; ++ hrtimer->irqsafe = 1; + + /* Enable the perf event */ + watchdog_nmi_enable(cpu); +diff -Nur linux-4.1.13.orig/kernel/workqueue.c linux-4.1.13/kernel/workqueue.c +--- linux-4.1.13.orig/kernel/workqueue.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/workqueue.c 2015-11-29 09:23:09.657610995 +0100 +@@ -48,6 +48,8 @@ + #include <linux/nodemask.h> + #include <linux/moduleparam.h> + #include <linux/uaccess.h> ++#include <linux/locallock.h> ++#include <linux/delay.h> + + #include "workqueue_internal.h" + +@@ -121,15 +123,20 @@ + * cpu or grabbing pool->lock is enough for read access. If + * POOL_DISASSOCIATED is set, it's identical to L. + * ++ * On RT we need the extra protection via rt_lock_idle_list() for ++ * the list manipulations against read access from ++ * wq_worker_sleeping(). All other places are nicely serialized via ++ * pool->lock. ++ * + * A: pool->attach_mutex protected. + * + * PL: wq_pool_mutex protected. + * +- * PR: wq_pool_mutex protected for writes. Sched-RCU protected for reads. ++ * PR: wq_pool_mutex protected for writes. RCU protected for reads. + * + * WQ: wq->mutex protected. + * +- * WR: wq->mutex protected for writes. Sched-RCU protected for reads. ++ * WR: wq->mutex protected for writes. RCU protected for reads. + * + * MD: wq_mayday_lock protected. + */ +@@ -178,7 +185,7 @@ + atomic_t nr_running ____cacheline_aligned_in_smp; + + /* +- * Destruction of pool is sched-RCU protected to allow dereferences ++ * Destruction of pool is RCU protected to allow dereferences + * from get_work_pool(). + */ + struct rcu_head rcu; +@@ -207,7 +214,7 @@ + /* + * Release of unbound pwq is punted to system_wq. See put_pwq() + * and pwq_unbound_release_workfn() for details. pool_workqueue +- * itself is also sched-RCU protected so that the first pwq can be ++ * itself is also RCU protected so that the first pwq can be + * determined without grabbing wq->mutex. + */ + struct work_struct unbound_release_work; +@@ -329,6 +336,8 @@ + struct workqueue_struct *system_freezable_power_efficient_wq __read_mostly; + EXPORT_SYMBOL_GPL(system_freezable_power_efficient_wq); + ++static DEFINE_LOCAL_IRQ_LOCK(pendingb_lock); ++ + static int worker_thread(void *__worker); + static void copy_workqueue_attrs(struct workqueue_attrs *to, + const struct workqueue_attrs *from); +@@ -338,14 +347,14 @@ + #include <trace/events/workqueue.h> + + #define assert_rcu_or_pool_mutex() \ +- rcu_lockdep_assert(rcu_read_lock_sched_held() || \ ++ rcu_lockdep_assert(rcu_read_lock_held() || \ + lockdep_is_held(&wq_pool_mutex), \ +- "sched RCU or wq_pool_mutex should be held") ++ "RCU or wq_pool_mutex should be held") + + #define assert_rcu_or_wq_mutex(wq) \ +- rcu_lockdep_assert(rcu_read_lock_sched_held() || \ ++ rcu_lockdep_assert(rcu_read_lock_held() || \ + lockdep_is_held(&wq->mutex), \ +- "sched RCU or wq->mutex should be held") ++ "RCU or wq->mutex should be held") + + #define for_each_cpu_worker_pool(pool, cpu) \ + for ((pool) = &per_cpu(cpu_worker_pools, cpu)[0]; \ +@@ -357,7 +366,7 @@ + * @pool: iteration cursor + * @pi: integer used for iteration + * +- * This must be called either with wq_pool_mutex held or sched RCU read ++ * This must be called either with wq_pool_mutex held or RCU read + * locked. If the pool needs to be used beyond the locking in effect, the + * caller is responsible for guaranteeing that the pool stays online. + * +@@ -389,7 +398,7 @@ + * @pwq: iteration cursor + * @wq: the target workqueue + * +- * This must be called either with wq->mutex held or sched RCU read locked. ++ * This must be called either with wq->mutex held or RCU read locked. + * If the pwq needs to be used beyond the locking in effect, the caller is + * responsible for guaranteeing that the pwq stays online. + * +@@ -401,6 +410,31 @@ + if (({ assert_rcu_or_wq_mutex(wq); false; })) { } \ + else + ++#ifdef CONFIG_PREEMPT_RT_BASE ++static inline void rt_lock_idle_list(struct worker_pool *pool) ++{ ++ preempt_disable(); ++} ++static inline void rt_unlock_idle_list(struct worker_pool *pool) ++{ ++ preempt_enable(); ++} ++static inline void sched_lock_idle_list(struct worker_pool *pool) { } ++static inline void sched_unlock_idle_list(struct worker_pool *pool) { } ++#else ++static inline void rt_lock_idle_list(struct worker_pool *pool) { } ++static inline void rt_unlock_idle_list(struct worker_pool *pool) { } ++static inline void sched_lock_idle_list(struct worker_pool *pool) ++{ ++ spin_lock_irq(&pool->lock); ++} ++static inline void sched_unlock_idle_list(struct worker_pool *pool) ++{ ++ spin_unlock_irq(&pool->lock); ++} ++#endif ++ ++ + #ifdef CONFIG_DEBUG_OBJECTS_WORK + + static struct debug_obj_descr work_debug_descr; +@@ -551,7 +585,7 @@ + * @wq: the target workqueue + * @node: the node ID + * +- * This must be called either with pwq_lock held or sched RCU read locked. ++ * This must be called either with pwq_lock held or RCU read locked. + * If the pwq needs to be used beyond the locking in effect, the caller is + * responsible for guaranteeing that the pwq stays online. + * +@@ -655,8 +689,8 @@ + * @work: the work item of interest + * + * Pools are created and destroyed under wq_pool_mutex, and allows read +- * access under sched-RCU read lock. As such, this function should be +- * called under wq_pool_mutex or with preemption disabled. ++ * access under RCU read lock. As such, this function should be ++ * called under wq_pool_mutex or inside of a rcu_read_lock() region. + * + * All fields of the returned pool are accessible as long as the above + * mentioned locking is in effect. If the returned pool needs to be used +@@ -793,51 +827,44 @@ + */ + static void wake_up_worker(struct worker_pool *pool) + { +- struct worker *worker = first_idle_worker(pool); ++ struct worker *worker; ++ ++ rt_lock_idle_list(pool); ++ ++ worker = first_idle_worker(pool); + + if (likely(worker)) + wake_up_process(worker->task); ++ ++ rt_unlock_idle_list(pool); + } + + /** +- * wq_worker_waking_up - a worker is waking up +- * @task: task waking up +- * @cpu: CPU @task is waking up to ++ * wq_worker_running - a worker is running again ++ * @task: task returning from sleep + * +- * This function is called during try_to_wake_up() when a worker is +- * being awoken. +- * +- * CONTEXT: +- * spin_lock_irq(rq->lock) ++ * This function is called when a worker returns from schedule() + */ +-void wq_worker_waking_up(struct task_struct *task, int cpu) ++void wq_worker_running(struct task_struct *task) + { + struct worker *worker = kthread_data(task); + +- if (!(worker->flags & WORKER_NOT_RUNNING)) { +- WARN_ON_ONCE(worker->pool->cpu != cpu); ++ if (!worker->sleeping) ++ return; ++ if (!(worker->flags & WORKER_NOT_RUNNING)) + atomic_inc(&worker->pool->nr_running); +- } ++ worker->sleeping = 0; + } + + /** + * wq_worker_sleeping - a worker is going to sleep + * @task: task going to sleep +- * @cpu: CPU in question, must be the current CPU number +- * +- * This function is called during schedule() when a busy worker is +- * going to sleep. Worker on the same cpu can be woken up by +- * returning pointer to its task. +- * +- * CONTEXT: +- * spin_lock_irq(rq->lock) +- * +- * Return: +- * Worker task on @cpu to wake up, %NULL if none. ++ * This function is called from schedule() when a busy worker is ++ * going to sleep. + */ +-struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu) ++void wq_worker_sleeping(struct task_struct *task) + { +- struct worker *worker = kthread_data(task), *to_wakeup = NULL; ++ struct worker *worker = kthread_data(task); + struct worker_pool *pool; + + /* +@@ -846,29 +873,26 @@ + * checking NOT_RUNNING. + */ + if (worker->flags & WORKER_NOT_RUNNING) +- return NULL; ++ return; + + pool = worker->pool; + +- /* this can only happen on the local cpu */ +- if (WARN_ON_ONCE(cpu != raw_smp_processor_id() || pool->cpu != cpu)) +- return NULL; ++ if (WARN_ON_ONCE(worker->sleeping)) ++ return; ++ ++ worker->sleeping = 1; + + /* + * The counterpart of the following dec_and_test, implied mb, + * worklist not empty test sequence is in insert_work(). + * Please read comment there. +- * +- * NOT_RUNNING is clear. This means that we're bound to and +- * running on the local cpu w/ rq lock held and preemption +- * disabled, which in turn means that none else could be +- * manipulating idle_list, so dereferencing idle_list without pool +- * lock is safe. + */ + if (atomic_dec_and_test(&pool->nr_running) && +- !list_empty(&pool->worklist)) +- to_wakeup = first_idle_worker(pool); +- return to_wakeup ? to_wakeup->task : NULL; ++ !list_empty(&pool->worklist)) { ++ sched_lock_idle_list(pool); ++ wake_up_worker(pool); ++ sched_unlock_idle_list(pool); ++ } + } + + /** +@@ -1062,12 +1086,12 @@ + { + if (pwq) { + /* +- * As both pwqs and pools are sched-RCU protected, the ++ * As both pwqs and pools are RCU protected, the + * following lock operations are safe. + */ +- spin_lock_irq(&pwq->pool->lock); ++ local_spin_lock_irq(pendingb_lock, &pwq->pool->lock); + put_pwq(pwq); +- spin_unlock_irq(&pwq->pool->lock); ++ local_spin_unlock_irq(pendingb_lock, &pwq->pool->lock); + } + } + +@@ -1169,7 +1193,7 @@ + struct worker_pool *pool; + struct pool_workqueue *pwq; + +- local_irq_save(*flags); ++ local_lock_irqsave(pendingb_lock, *flags); + + /* try to steal the timer if it exists */ + if (is_dwork) { +@@ -1188,6 +1212,7 @@ + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) + return 0; + ++ rcu_read_lock(); + /* + * The queueing is in progress, or it is already queued. Try to + * steal it from ->worklist without clearing WORK_STRUCT_PENDING. +@@ -1226,14 +1251,16 @@ + set_work_pool_and_keep_pending(work, pool->id); + + spin_unlock(&pool->lock); ++ rcu_read_unlock(); + return 1; + } + spin_unlock(&pool->lock); + fail: +- local_irq_restore(*flags); ++ rcu_read_unlock(); ++ local_unlock_irqrestore(pendingb_lock, *flags); + if (work_is_canceling(work)) + return -ENOENT; +- cpu_relax(); ++ cpu_chill(); + return -EAGAIN; + } + +@@ -1302,7 +1329,7 @@ + * queued or lose PENDING. Grabbing PENDING and queueing should + * happen with IRQ disabled. + */ +- WARN_ON_ONCE(!irqs_disabled()); ++ WARN_ON_ONCE_NONRT(!irqs_disabled()); + + debug_work_activate(work); + +@@ -1310,6 +1337,8 @@ + if (unlikely(wq->flags & __WQ_DRAINING) && + WARN_ON_ONCE(!is_chained_work(wq))) + return; ++ ++ rcu_read_lock(); + retry: + if (req_cpu == WORK_CPU_UNBOUND) + cpu = raw_smp_processor_id(); +@@ -1366,10 +1395,8 @@ + /* pwq determined, queue */ + trace_workqueue_queue_work(req_cpu, pwq, work); + +- if (WARN_ON(!list_empty(&work->entry))) { +- spin_unlock(&pwq->pool->lock); +- return; +- } ++ if (WARN_ON(!list_empty(&work->entry))) ++ goto out; + + pwq->nr_in_flight[pwq->work_color]++; + work_flags = work_color_to_flags(pwq->work_color); +@@ -1385,7 +1412,9 @@ + + insert_work(pwq, work, worklist, work_flags); + ++out: + spin_unlock(&pwq->pool->lock); ++ rcu_read_unlock(); + } + + /** +@@ -1405,14 +1434,14 @@ + bool ret = false; + unsigned long flags; + +- local_irq_save(flags); ++ local_lock_irqsave(pendingb_lock,flags); + + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { + __queue_work(cpu, wq, work); + ret = true; + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + EXPORT_SYMBOL(queue_work_on); +@@ -1479,14 +1508,14 @@ + unsigned long flags; + + /* read the comment in __queue_work() */ +- local_irq_save(flags); ++ local_lock_irqsave(pendingb_lock, flags); + + if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) { + __queue_delayed_work(cpu, wq, dwork, delay); + ret = true; + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + EXPORT_SYMBOL(queue_delayed_work_on); +@@ -1521,7 +1550,7 @@ + + if (likely(ret >= 0)) { + __queue_delayed_work(cpu, wq, dwork, delay); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + } + + /* -ENOENT from try_to_grab_pending() becomes %true */ +@@ -1554,7 +1583,9 @@ + worker->last_active = jiffies; + + /* idle_list is LIFO */ ++ rt_lock_idle_list(pool); + list_add(&worker->entry, &pool->idle_list); ++ rt_unlock_idle_list(pool); + + if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) + mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT); +@@ -1587,7 +1618,9 @@ + return; + worker_clr_flags(worker, WORKER_IDLE); + pool->nr_idle--; ++ rt_lock_idle_list(pool); + list_del_init(&worker->entry); ++ rt_unlock_idle_list(pool); + } + + static struct worker *alloc_worker(int node) +@@ -1755,7 +1788,9 @@ + pool->nr_workers--; + pool->nr_idle--; + ++ rt_lock_idle_list(pool); + list_del_init(&worker->entry); ++ rt_unlock_idle_list(pool); + worker->flags |= WORKER_DIE; + wake_up_process(worker->task); + } +@@ -2672,14 +2707,14 @@ + + might_sleep(); + +- local_irq_disable(); ++ rcu_read_lock(); + pool = get_work_pool(work); + if (!pool) { +- local_irq_enable(); ++ rcu_read_unlock(); + return false; + } + +- spin_lock(&pool->lock); ++ spin_lock_irq(&pool->lock); + /* see the comment in try_to_grab_pending() with the same code */ + pwq = get_work_pwq(work); + if (pwq) { +@@ -2706,10 +2741,11 @@ + else + lock_map_acquire_read(&pwq->wq->lockdep_map); + lock_map_release(&pwq->wq->lockdep_map); +- ++ rcu_read_unlock(); + return true; + already_gone: + spin_unlock_irq(&pool->lock); ++ rcu_read_unlock(); + return false; + } + +@@ -2796,7 +2832,7 @@ + + /* tell other tasks trying to grab @work to back off */ + mark_work_canceling(work); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + + flush_work(work); + clear_work_data(work); +@@ -2851,10 +2887,10 @@ + */ + bool flush_delayed_work(struct delayed_work *dwork) + { +- local_irq_disable(); ++ local_lock_irq(pendingb_lock); + if (del_timer_sync(&dwork->timer)) + __queue_work(dwork->cpu, dwork->wq, &dwork->work); +- local_irq_enable(); ++ local_unlock_irq(pendingb_lock); + return flush_work(&dwork->work); + } + EXPORT_SYMBOL(flush_delayed_work); +@@ -2889,7 +2925,7 @@ + + set_work_pool_and_clear_pending(&dwork->work, + get_work_pool_id(&dwork->work)); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pendingb_lock, flags); + return ret; + } + EXPORT_SYMBOL(cancel_delayed_work); +@@ -3147,7 +3183,7 @@ + * put_unbound_pool - put a worker_pool + * @pool: worker_pool to put + * +- * Put @pool. If its refcnt reaches zero, it gets destroyed in sched-RCU ++ * Put @pool. If its refcnt reaches zero, it gets destroyed in RCU + * safe manner. get_unbound_pool() calls this function on its failure path + * and this function should be able to release pools which went through, + * successfully or not, init_worker_pool(). +@@ -3201,8 +3237,8 @@ + del_timer_sync(&pool->idle_timer); + del_timer_sync(&pool->mayday_timer); + +- /* sched-RCU protected to allow dereferences from get_work_pool() */ +- call_rcu_sched(&pool->rcu, rcu_free_pool); ++ /* RCU protected to allow dereferences from get_work_pool() */ ++ call_rcu(&pool->rcu, rcu_free_pool); + } + + /** +@@ -3307,14 +3343,14 @@ + put_unbound_pool(pool); + mutex_unlock(&wq_pool_mutex); + +- call_rcu_sched(&pwq->rcu, rcu_free_pwq); ++ call_rcu(&pwq->rcu, rcu_free_pwq); + + /* + * If we're the last pwq going away, @wq is already dead and no one + * is gonna access it anymore. Schedule RCU free. + */ + if (is_last) +- call_rcu_sched(&wq->rcu, rcu_free_wq); ++ call_rcu(&wq->rcu, rcu_free_wq); + } + + /** +@@ -3920,7 +3956,7 @@ + * The base ref is never dropped on per-cpu pwqs. Directly + * schedule RCU free. + */ +- call_rcu_sched(&wq->rcu, rcu_free_wq); ++ call_rcu(&wq->rcu, rcu_free_wq); + } else { + /* + * We're the sole accessor of @wq at this point. Directly +@@ -4013,7 +4049,8 @@ + struct pool_workqueue *pwq; + bool ret; + +- rcu_read_lock_sched(); ++ rcu_read_lock(); ++ preempt_disable(); + + if (cpu == WORK_CPU_UNBOUND) + cpu = smp_processor_id(); +@@ -4024,7 +4061,8 @@ + pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); + + ret = !list_empty(&pwq->delayed_works); +- rcu_read_unlock_sched(); ++ preempt_enable(); ++ rcu_read_unlock(); + + return ret; + } +@@ -4050,15 +4088,15 @@ + if (work_pending(work)) + ret |= WORK_BUSY_PENDING; + +- local_irq_save(flags); ++ rcu_read_lock(); + pool = get_work_pool(work); + if (pool) { +- spin_lock(&pool->lock); ++ spin_lock_irqsave(&pool->lock, flags); + if (find_worker_executing_work(pool, work)) + ret |= WORK_BUSY_RUNNING; +- spin_unlock(&pool->lock); ++ spin_unlock_irqrestore(&pool->lock, flags); + } +- local_irq_restore(flags); ++ rcu_read_unlock(); + + return ret; + } +@@ -4247,7 +4285,7 @@ + unsigned long flags; + int pi; + +- rcu_read_lock_sched(); ++ rcu_read_lock(); + + pr_info("Showing busy workqueues and worker pools:\n"); + +@@ -4298,7 +4336,7 @@ + spin_unlock_irqrestore(&pool->lock, flags); + } + +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + } + + /* +@@ -4648,16 +4686,16 @@ + * nr_active is monotonically decreasing. It's safe + * to peek without lock. + */ +- rcu_read_lock_sched(); ++ rcu_read_lock(); + for_each_pwq(pwq, wq) { + WARN_ON_ONCE(pwq->nr_active < 0); + if (pwq->nr_active) { + busy = true; +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + goto out_unlock; + } + } +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); + } + out_unlock: + mutex_unlock(&wq_pool_mutex); +@@ -4771,7 +4809,8 @@ + const char *delim = ""; + int node, written = 0; + +- rcu_read_lock_sched(); ++ get_online_cpus(); ++ rcu_read_lock(); + for_each_node(node) { + written += scnprintf(buf + written, PAGE_SIZE - written, + "%s%d:%d", delim, node, +@@ -4779,7 +4818,8 @@ + delim = " "; + } + written += scnprintf(buf + written, PAGE_SIZE - written, "\n"); +- rcu_read_unlock_sched(); ++ rcu_read_unlock(); ++ put_online_cpus(); + + return written; + } +diff -Nur linux-4.1.13.orig/kernel/workqueue_internal.h linux-4.1.13/kernel/workqueue_internal.h +--- linux-4.1.13.orig/kernel/workqueue_internal.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/kernel/workqueue_internal.h 2015-11-29 09:23:09.657610995 +0100 +@@ -43,6 +43,7 @@ + unsigned long last_active; /* L: last active timestamp */ + unsigned int flags; /* X: flags */ + int id; /* I: worker id */ ++ int sleeping; /* None */ + + /* + * Opaque string set with work_set_desc(). Printed out with task +@@ -68,7 +69,7 @@ + * Scheduler hooks for concurrency managed workqueue. Only to be used from + * sched/core.c and workqueue.c. + */ +-void wq_worker_waking_up(struct task_struct *task, int cpu); +-struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu); ++void wq_worker_running(struct task_struct *task); ++void wq_worker_sleeping(struct task_struct *task); + + #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */ +diff -Nur linux-4.1.13.orig/lib/debugobjects.c linux-4.1.13/lib/debugobjects.c +--- linux-4.1.13.orig/lib/debugobjects.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/debugobjects.c 2015-11-29 09:23:09.657610995 +0100 +@@ -309,7 +309,10 @@ + struct debug_obj *obj; + unsigned long flags; + +- fill_pool(); ++#ifdef CONFIG_PREEMPT_RT_FULL ++ if (preempt_count() == 0 && !irqs_disabled()) ++#endif ++ fill_pool(); + + db = get_bucket((unsigned long) addr); + +diff -Nur linux-4.1.13.orig/lib/dump_stack.c linux-4.1.13/lib/dump_stack.c +--- linux-4.1.13.orig/lib/dump_stack.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/dump_stack.c 2015-11-29 09:23:09.657610995 +0100 +@@ -33,7 +33,7 @@ + * Permit this cpu to perform nested stack dumps while serialising + * against other CPUs + */ +- preempt_disable(); ++ migrate_disable(); + + retry: + cpu = smp_processor_id(); +@@ -52,7 +52,7 @@ + if (!was_locked) + atomic_set(&dump_lock, -1); + +- preempt_enable(); ++ migrate_enable(); + } + #else + asmlinkage __visible void dump_stack(void) +diff -Nur linux-4.1.13.orig/lib/idr.c linux-4.1.13/lib/idr.c +--- linux-4.1.13.orig/lib/idr.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/idr.c 2015-11-29 09:23:09.657610995 +0100 +@@ -30,6 +30,7 @@ + #include <linux/idr.h> + #include <linux/spinlock.h> + #include <linux/percpu.h> ++#include <linux/locallock.h> + + #define MAX_IDR_SHIFT (sizeof(int) * 8 - 1) + #define MAX_IDR_BIT (1U << MAX_IDR_SHIFT) +@@ -366,6 +367,35 @@ + idr_mark_full(pa, id); + } + ++#ifdef CONFIG_PREEMPT_RT_FULL ++static DEFINE_LOCAL_IRQ_LOCK(idr_lock); ++ ++static inline void idr_preload_lock(void) ++{ ++ local_lock(idr_lock); ++} ++ ++static inline void idr_preload_unlock(void) ++{ ++ local_unlock(idr_lock); ++} ++ ++void idr_preload_end(void) ++{ ++ idr_preload_unlock(); ++} ++EXPORT_SYMBOL(idr_preload_end); ++#else ++static inline void idr_preload_lock(void) ++{ ++ preempt_disable(); ++} ++ ++static inline void idr_preload_unlock(void) ++{ ++ preempt_enable(); ++} ++#endif + + /** + * idr_preload - preload for idr_alloc() +@@ -401,7 +431,7 @@ + WARN_ON_ONCE(in_interrupt()); + might_sleep_if(gfp_mask & __GFP_WAIT); + +- preempt_disable(); ++ idr_preload_lock(); + + /* + * idr_alloc() is likely to succeed w/o full idr_layer buffer and +@@ -413,9 +443,9 @@ + while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) { + struct idr_layer *new; + +- preempt_enable(); ++ idr_preload_unlock(); + new = kmem_cache_zalloc(idr_layer_cache, gfp_mask); +- preempt_disable(); ++ idr_preload_lock(); + if (!new) + break; + +diff -Nur linux-4.1.13.orig/lib/Kconfig linux-4.1.13/lib/Kconfig +--- linux-4.1.13.orig/lib/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/Kconfig 2015-11-29 09:23:09.657610995 +0100 +@@ -391,6 +391,7 @@ + + config CPUMASK_OFFSTACK + bool "Force CPU masks off stack" if DEBUG_PER_CPU_MAPS ++ depends on !PREEMPT_RT_FULL + help + Use dynamic allocation for cpumask_var_t, instead of putting + them on the stack. This is a bit more expensive, but avoids +diff -Nur linux-4.1.13.orig/lib/locking-selftest.c linux-4.1.13/lib/locking-selftest.c +--- linux-4.1.13.orig/lib/locking-selftest.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/locking-selftest.c 2015-11-29 09:23:09.657610995 +0100 +@@ -590,6 +590,8 @@ + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_hard_rlock) + +@@ -605,9 +607,12 @@ + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe1_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Enabling hardirqs with a softirq-safe lock held: + */ +@@ -640,6 +645,8 @@ + #undef E1 + #undef E2 + ++#endif ++ + /* + * Enabling irqs with an irq-safe lock held: + */ +@@ -663,6 +670,8 @@ + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_hard_rlock) + +@@ -678,6 +687,8 @@ + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_2_EVENTS(irqsafe2B_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + +@@ -709,6 +720,8 @@ + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_hard_rlock) + +@@ -724,6 +737,8 @@ + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe3_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + #undef E3 +@@ -757,6 +772,8 @@ + #include "locking-selftest-spin-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_spin) + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + #include "locking-selftest-rlock-hardirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_hard_rlock) + +@@ -772,10 +789,14 @@ + #include "locking-selftest-wlock-softirq.h" + GENERATE_PERMUTATIONS_3_EVENTS(irqsafe4_soft_wlock) + ++#endif ++ + #undef E1 + #undef E2 + #undef E3 + ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * read-lock / write-lock irq inversion. + * +@@ -838,6 +859,10 @@ + #undef E2 + #undef E3 + ++#endif ++ ++#ifndef CONFIG_PREEMPT_RT_FULL ++ + /* + * read-lock / write-lock recursion that is actually safe. + */ +@@ -876,6 +901,8 @@ + #undef E2 + #undef E3 + ++#endif ++ + /* + * read-lock / write-lock recursion that is unsafe. + */ +@@ -1858,6 +1885,7 @@ + + printk(" --------------------------------------------------------------------------\n"); + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * irq-context testcases: + */ +@@ -1870,6 +1898,28 @@ + + DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion); + // DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2); ++#else ++ /* On -rt, we only do hardirq context test for raw spinlock */ ++ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 12); ++ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 21); ++ ++ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 12); ++ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 21); ++ ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 123); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 132); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 213); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 231); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 312); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 321); ++ ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 123); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 132); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 213); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 231); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 312); ++ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 321); ++#endif + + ww_tests(); + +diff -Nur linux-4.1.13.orig/lib/percpu_ida.c linux-4.1.13/lib/percpu_ida.c +--- linux-4.1.13.orig/lib/percpu_ida.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/percpu_ida.c 2015-11-29 09:23:09.657610995 +0100 +@@ -26,6 +26,9 @@ + #include <linux/string.h> + #include <linux/spinlock.h> + #include <linux/percpu_ida.h> ++#include <linux/locallock.h> ++ ++static DEFINE_LOCAL_IRQ_LOCK(irq_off_lock); + + struct percpu_ida_cpu { + /* +@@ -148,13 +151,13 @@ + unsigned long flags; + int tag; + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + + /* Fastpath */ + tag = alloc_local_tag(tags); + if (likely(tag >= 0)) { +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + return tag; + } + +@@ -173,6 +176,7 @@ + + if (!tags->nr_free) + alloc_global_tags(pool, tags); ++ + if (!tags->nr_free) + steal_tags(pool, tags); + +@@ -184,7 +188,7 @@ + } + + spin_unlock(&pool->lock); +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + + if (tag >= 0 || state == TASK_RUNNING) + break; +@@ -196,7 +200,7 @@ + + schedule(); + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + } + if (state != TASK_RUNNING) +@@ -221,7 +225,7 @@ + + BUG_ON(tag >= pool->nr_tags); + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + tags = this_cpu_ptr(pool->tag_cpu); + + spin_lock(&tags->lock); +@@ -253,7 +257,7 @@ + spin_unlock(&pool->lock); + } + +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + } + EXPORT_SYMBOL_GPL(percpu_ida_free); + +@@ -345,7 +349,7 @@ + struct percpu_ida_cpu *remote; + unsigned cpu, i, err = 0; + +- local_irq_save(flags); ++ local_lock_irqsave(irq_off_lock, flags); + for_each_possible_cpu(cpu) { + remote = per_cpu_ptr(pool->tag_cpu, cpu); + spin_lock(&remote->lock); +@@ -367,7 +371,7 @@ + } + spin_unlock(&pool->lock); + out: +- local_irq_restore(flags); ++ local_unlock_irqrestore(irq_off_lock, flags); + return err; + } + EXPORT_SYMBOL_GPL(percpu_ida_for_each_free); +diff -Nur linux-4.1.13.orig/lib/radix-tree.c linux-4.1.13/lib/radix-tree.c +--- linux-4.1.13.orig/lib/radix-tree.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/radix-tree.c 2015-11-29 09:23:09.657610995 +0100 +@@ -195,12 +195,13 @@ + * succeed in getting a node here (and never reach + * kmem_cache_alloc) + */ +- rtp = this_cpu_ptr(&radix_tree_preloads); ++ rtp = &get_cpu_var(radix_tree_preloads); + if (rtp->nr) { + ret = rtp->nodes[rtp->nr - 1]; + rtp->nodes[rtp->nr - 1] = NULL; + rtp->nr--; + } ++ put_cpu_var(radix_tree_preloads); + /* + * Update the allocation stack trace as this is more useful + * for debugging. +@@ -240,6 +241,7 @@ + call_rcu(&node->rcu_head, radix_tree_node_rcu_free); + } + ++#ifndef CONFIG_PREEMPT_RT_FULL + /* + * Load up this CPU's radix_tree_node buffer with sufficient objects to + * ensure that the addition of a single element in the tree cannot fail. On +@@ -305,6 +307,7 @@ + return 0; + } + EXPORT_SYMBOL(radix_tree_maybe_preload); ++#endif + + /* + * Return the maximum key which can be store into a +diff -Nur linux-4.1.13.orig/lib/scatterlist.c linux-4.1.13/lib/scatterlist.c +--- linux-4.1.13.orig/lib/scatterlist.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/scatterlist.c 2015-11-29 09:23:09.657610995 +0100 +@@ -592,7 +592,7 @@ + flush_kernel_dcache_page(miter->page); + + if (miter->__flags & SG_MITER_ATOMIC) { +- WARN_ON_ONCE(preemptible()); ++ WARN_ON_ONCE(!pagefault_disabled()); + kunmap_atomic(miter->addr); + } else + kunmap(miter->page); +@@ -637,7 +637,7 @@ + if (!sg_miter_skip(&miter, skip)) + return false; + +- local_irq_save(flags); ++ local_irq_save_nort(flags); + + while (sg_miter_next(&miter) && offset < buflen) { + unsigned int len; +@@ -654,7 +654,7 @@ + + sg_miter_stop(&miter); + +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + return offset; + } + +diff -Nur linux-4.1.13.orig/lib/smp_processor_id.c linux-4.1.13/lib/smp_processor_id.c +--- linux-4.1.13.orig/lib/smp_processor_id.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/smp_processor_id.c 2015-11-29 09:23:09.657610995 +0100 +@@ -39,8 +39,9 @@ + if (!printk_ratelimit()) + goto out_enable; + +- printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x] code: %s/%d\n", +- what1, what2, preempt_count() - 1, current->comm, current->pid); ++ printk(KERN_ERR "BUG: using %s%s() in preemptible [%08x %08x] code: %s/%d\n", ++ what1, what2, preempt_count() - 1, __migrate_disabled(current), ++ current->comm, current->pid); + + print_symbol("caller is %s\n", (long)__builtin_return_address(0)); + dump_stack(); +diff -Nur linux-4.1.13.orig/lib/strnlen_user.c linux-4.1.13/lib/strnlen_user.c +--- linux-4.1.13.orig/lib/strnlen_user.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/lib/strnlen_user.c 2015-11-29 09:23:09.657610995 +0100 +@@ -85,7 +85,8 @@ + * @str: The string to measure. + * @count: Maximum count (including NUL character) + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +@@ -121,7 +122,8 @@ + * strlen_user: - Get the size of a user string INCLUDING final NUL. + * @str: The string to measure. + * +- * Context: User context only. This function may sleep. ++ * Context: User context only. This function may sleep if pagefaults are ++ * enabled. + * + * Get the size of a NUL-terminated string in user space. + * +diff -Nur linux-4.1.13.orig/mm/compaction.c linux-4.1.13/mm/compaction.c +--- linux-4.1.13.orig/mm/compaction.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/compaction.c 2015-11-29 09:23:09.661610728 +0100 +@@ -1406,10 +1406,12 @@ + cc->migrate_pfn & ~((1UL << cc->order) - 1); + + if (last_migrated_pfn < current_block_start) { +- cpu = get_cpu(); ++ cpu = get_cpu_light(); ++ local_lock_irq(swapvec_lock); + lru_add_drain_cpu(cpu); ++ local_unlock_irq(swapvec_lock); + drain_local_pages(zone); +- put_cpu(); ++ put_cpu_light(); + /* No more flushing until we migrate again */ + last_migrated_pfn = 0; + } +diff -Nur linux-4.1.13.orig/mm/filemap.c linux-4.1.13/mm/filemap.c +--- linux-4.1.13.orig/mm/filemap.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/filemap.c 2015-11-29 09:23:09.661610728 +0100 +@@ -167,7 +167,9 @@ + if (!workingset_node_pages(node) && + list_empty(&node->private_list)) { + node->private_data = mapping; +- list_lru_add(&workingset_shadow_nodes, &node->private_list); ++ local_lock(workingset_shadow_lock); ++ list_lru_add(&__workingset_shadow_nodes, &node->private_list); ++ local_unlock(workingset_shadow_lock); + } + } + +@@ -533,9 +535,12 @@ + * node->private_list is protected by + * mapping->tree_lock. + */ +- if (!list_empty(&node->private_list)) +- list_lru_del(&workingset_shadow_nodes, ++ if (!list_empty(&node->private_list)) { ++ local_lock(workingset_shadow_lock); ++ list_lru_del(&__workingset_shadow_nodes, + &node->private_list); ++ local_unlock(workingset_shadow_lock); ++ } + } + return 0; + } +diff -Nur linux-4.1.13.orig/mm/highmem.c linux-4.1.13/mm/highmem.c +--- linux-4.1.13.orig/mm/highmem.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/highmem.c 2015-11-29 09:23:09.661610728 +0100 +@@ -29,10 +29,11 @@ + #include <linux/kgdb.h> + #include <asm/tlbflush.h> + +- ++#ifndef CONFIG_PREEMPT_RT_FULL + #if defined(CONFIG_HIGHMEM) || defined(CONFIG_X86_32) + DEFINE_PER_CPU(int, __kmap_atomic_idx); + #endif ++#endif + + /* + * Virtual_count is not a pure "count". +@@ -107,8 +108,9 @@ + unsigned long totalhigh_pages __read_mostly; + EXPORT_SYMBOL(totalhigh_pages); + +- ++#ifndef CONFIG_PREEMPT_RT_FULL + EXPORT_PER_CPU_SYMBOL(__kmap_atomic_idx); ++#endif + + unsigned int nr_free_highpages (void) + { +diff -Nur linux-4.1.13.orig/mm/Kconfig linux-4.1.13/mm/Kconfig +--- linux-4.1.13.orig/mm/Kconfig 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/Kconfig 2015-11-29 09:23:09.661610728 +0100 +@@ -409,7 +409,7 @@ + + config TRANSPARENT_HUGEPAGE + bool "Transparent Hugepage Support" +- depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE ++ depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT_FULL + select COMPACTION + help + Transparent Hugepages allows the kernel to use huge pages and +diff -Nur linux-4.1.13.orig/mm/memcontrol.c linux-4.1.13/mm/memcontrol.c +--- linux-4.1.13.orig/mm/memcontrol.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/memcontrol.c 2015-11-29 09:23:09.661610728 +0100 +@@ -66,6 +66,8 @@ + #include <net/sock.h> + #include <net/ip.h> + #include <net/tcp_memcontrol.h> ++#include <linux/locallock.h> ++ + #include "slab.h" + + #include <asm/uaccess.h> +@@ -85,6 +87,7 @@ + #define do_swap_account 0 + #endif + ++static DEFINE_LOCAL_IRQ_LOCK(event_lock); + static const char * const mem_cgroup_stat_names[] = { + "cache", + "rss", +@@ -2124,14 +2127,17 @@ + */ + static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages) + { +- struct memcg_stock_pcp *stock = &get_cpu_var(memcg_stock); ++ struct memcg_stock_pcp *stock; ++ int cpu = get_cpu_light(); ++ ++ stock = &per_cpu(memcg_stock, cpu); + + if (stock->cached != memcg) { /* reset if necessary */ + drain_stock(stock); + stock->cached = memcg; + } + stock->nr_pages += nr_pages; +- put_cpu_var(memcg_stock); ++ put_cpu_light(); + } + + /* +@@ -2147,7 +2153,7 @@ + return; + /* Notify other cpus that system-wide "drain" is running */ + get_online_cpus(); +- curcpu = get_cpu(); ++ curcpu = get_cpu_light(); + for_each_online_cpu(cpu) { + struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); + struct mem_cgroup *memcg; +@@ -2164,7 +2170,7 @@ + schedule_work_on(cpu, &stock->work); + } + } +- put_cpu(); ++ put_cpu_light(); + put_online_cpus(); + mutex_unlock(&percpu_charge_mutex); + } +@@ -4802,12 +4808,12 @@ + + ret = 0; + +- local_irq_disable(); ++ local_lock_irq(event_lock); + mem_cgroup_charge_statistics(to, page, nr_pages); + memcg_check_events(to, page); + mem_cgroup_charge_statistics(from, page, -nr_pages); + memcg_check_events(from, page); +- local_irq_enable(); ++ local_unlock_irq(event_lock); + out_unlock: + unlock_page(page); + out: +@@ -5544,10 +5550,10 @@ + VM_BUG_ON_PAGE(!PageTransHuge(page), page); + } + +- local_irq_disable(); ++ local_lock_irq(event_lock); + mem_cgroup_charge_statistics(memcg, page, nr_pages); + memcg_check_events(memcg, page); +- local_irq_enable(); ++ local_unlock_irq(event_lock); + + if (do_swap_account && PageSwapCache(page)) { + swp_entry_t entry = { .val = page_private(page) }; +@@ -5603,14 +5609,14 @@ + memcg_oom_recover(memcg); + } + +- local_irq_save(flags); ++ local_lock_irqsave(event_lock, flags); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS], nr_anon); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_CACHE], nr_file); + __this_cpu_sub(memcg->stat->count[MEM_CGROUP_STAT_RSS_HUGE], nr_huge); + __this_cpu_add(memcg->stat->events[MEM_CGROUP_EVENTS_PGPGOUT], pgpgout); + __this_cpu_add(memcg->stat->nr_page_events, nr_pages); + memcg_check_events(memcg, dummy_page); +- local_irq_restore(flags); ++ local_unlock_irqrestore(event_lock, flags); + + if (!mem_cgroup_is_root(memcg)) + css_put_many(&memcg->css, nr_pages); +@@ -5814,6 +5820,7 @@ + { + struct mem_cgroup *memcg; + unsigned short oldid; ++ unsigned long flags; + + VM_BUG_ON_PAGE(PageLRU(page), page); + VM_BUG_ON_PAGE(page_count(page), page); +@@ -5836,9 +5843,11 @@ + if (!mem_cgroup_is_root(memcg)) + page_counter_uncharge(&memcg->memory, 1); + ++ local_lock_irqsave(event_lock, flags); + /* Caller disabled preemption with mapping->tree_lock */ + mem_cgroup_charge_statistics(memcg, page, -1); + memcg_check_events(memcg, page); ++ local_unlock_irqrestore(event_lock, flags); + } + + /** +diff -Nur linux-4.1.13.orig/mm/memory.c linux-4.1.13/mm/memory.c +--- linux-4.1.13.orig/mm/memory.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/memory.c 2015-11-29 09:23:09.661610728 +0100 +@@ -3743,7 +3743,7 @@ + } + + #if defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_DEBUG_ATOMIC_SLEEP) +-void might_fault(void) ++void __might_fault(const char *file, int line) + { + /* + * Some code (nfs/sunrpc) uses socket ops on kernel memory while +@@ -3753,21 +3753,15 @@ + */ + if (segment_eq(get_fs(), KERNEL_DS)) + return; +- +- /* +- * it would be nicer only to annotate paths which are not under +- * pagefault_disable, however that requires a larger audit and +- * providing helpers like get_user_atomic. +- */ +- if (in_atomic()) ++ if (pagefault_disabled()) + return; +- +- __might_sleep(__FILE__, __LINE__, 0); +- ++ __might_sleep(file, line, 0); ++#if defined(CONFIG_DEBUG_ATOMIC_SLEEP) + if (current->mm) + might_lock_read(¤t->mm->mmap_sem); ++#endif + } +-EXPORT_SYMBOL(might_fault); ++EXPORT_SYMBOL(__might_fault); + #endif + + #if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_HUGETLBFS) +diff -Nur linux-4.1.13.orig/mm/mmu_context.c linux-4.1.13/mm/mmu_context.c +--- linux-4.1.13.orig/mm/mmu_context.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/mmu_context.c 2015-11-29 09:23:09.661610728 +0100 +@@ -23,6 +23,7 @@ + struct task_struct *tsk = current; + + task_lock(tsk); ++ preempt_disable_rt(); + active_mm = tsk->active_mm; + if (active_mm != mm) { + atomic_inc(&mm->mm_count); +@@ -30,6 +31,7 @@ + } + tsk->mm = mm; + switch_mm(active_mm, mm, tsk); ++ preempt_enable_rt(); + task_unlock(tsk); + #ifdef finish_arch_post_lock_switch + finish_arch_post_lock_switch(); +diff -Nur linux-4.1.13.orig/mm/page_alloc.c linux-4.1.13/mm/page_alloc.c +--- linux-4.1.13.orig/mm/page_alloc.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/page_alloc.c 2015-11-29 09:23:09.661610728 +0100 +@@ -60,6 +60,7 @@ + #include <linux/page_ext.h> + #include <linux/hugetlb.h> + #include <linux/sched/rt.h> ++#include <linux/locallock.h> + #include <linux/page_owner.h> + + #include <asm/sections.h> +@@ -233,6 +234,18 @@ + EXPORT_SYMBOL(nr_online_nodes); + #endif + ++static DEFINE_LOCAL_IRQ_LOCK(pa_lock); ++ ++#ifdef CONFIG_PREEMPT_RT_BASE ++# define cpu_lock_irqsave(cpu, flags) \ ++ local_lock_irqsave_on(pa_lock, flags, cpu) ++# define cpu_unlock_irqrestore(cpu, flags) \ ++ local_unlock_irqrestore_on(pa_lock, flags, cpu) ++#else ++# define cpu_lock_irqsave(cpu, flags) local_irq_save(flags) ++# define cpu_unlock_irqrestore(cpu, flags) local_irq_restore(flags) ++#endif ++ + int page_group_by_mobility_disabled __read_mostly; + + void set_pageblock_migratetype(struct page *page, int migratetype) +@@ -681,7 +694,7 @@ + } + + /* +- * Frees a number of pages from the PCP lists ++ * Frees a number of pages which have been collected from the pcp lists. + * Assumes all pages on list are in same zone, and of same order. + * count is the number of pages to free. + * +@@ -692,18 +705,51 @@ + * pinned" detection logic. + */ + static void free_pcppages_bulk(struct zone *zone, int count, +- struct per_cpu_pages *pcp) ++ struct list_head *list) + { +- int migratetype = 0; +- int batch_free = 0; + int to_free = count; + unsigned long nr_scanned; ++ unsigned long flags; ++ ++ spin_lock_irqsave(&zone->lock, flags); + +- spin_lock(&zone->lock); + nr_scanned = zone_page_state(zone, NR_PAGES_SCANNED); + if (nr_scanned) + __mod_zone_page_state(zone, NR_PAGES_SCANNED, -nr_scanned); + ++ while (!list_empty(list)) { ++ struct page *page = list_first_entry(list, struct page, lru); ++ int mt; /* migratetype of the to-be-freed page */ ++ ++ /* must delete as __free_one_page list manipulates */ ++ list_del(&page->lru); ++ ++ mt = get_freepage_migratetype(page); ++ if (unlikely(has_isolate_pageblock(zone))) ++ mt = get_pageblock_migratetype(page); ++ ++ /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */ ++ __free_one_page(page, page_to_pfn(page), zone, 0, mt); ++ trace_mm_page_pcpu_drain(page, 0, mt); ++ to_free--; ++ } ++ WARN_ON(to_free != 0); ++ spin_unlock_irqrestore(&zone->lock, flags); ++} ++ ++/* ++ * Moves a number of pages from the PCP lists to free list which ++ * is freed outside of the locked region. ++ * ++ * Assumes all pages on list are in same zone, and of same order. ++ * count is the number of pages to free. ++ */ ++static void isolate_pcp_pages(int to_free, struct per_cpu_pages *src, ++ struct list_head *dst) ++{ ++ int migratetype = 0; ++ int batch_free = 0; ++ + while (to_free) { + struct page *page; + struct list_head *list; +@@ -719,7 +765,7 @@ + batch_free++; + if (++migratetype == MIGRATE_PCPTYPES) + migratetype = 0; +- list = &pcp->lists[migratetype]; ++ list = &src->lists[migratetype]; + } while (list_empty(list)); + + /* This is the only non-empty list. Free them all. */ +@@ -727,21 +773,11 @@ + batch_free = to_free; + + do { +- int mt; /* migratetype of the to-be-freed page */ +- +- page = list_entry(list->prev, struct page, lru); +- /* must delete as __free_one_page list manipulates */ ++ page = list_last_entry(list, struct page, lru); + list_del(&page->lru); +- mt = get_freepage_migratetype(page); +- if (unlikely(has_isolate_pageblock(zone))) +- mt = get_pageblock_migratetype(page); +- +- /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */ +- __free_one_page(page, page_to_pfn(page), zone, 0, mt); +- trace_mm_page_pcpu_drain(page, 0, mt); ++ list_add(&page->lru, dst); + } while (--to_free && --batch_free && !list_empty(list)); + } +- spin_unlock(&zone->lock); + } + + static void free_one_page(struct zone *zone, +@@ -750,7 +786,9 @@ + int migratetype) + { + unsigned long nr_scanned; +- spin_lock(&zone->lock); ++ unsigned long flags; ++ ++ spin_lock_irqsave(&zone->lock, flags); + nr_scanned = zone_page_state(zone, NR_PAGES_SCANNED); + if (nr_scanned) + __mod_zone_page_state(zone, NR_PAGES_SCANNED, -nr_scanned); +@@ -760,7 +798,7 @@ + migratetype = get_pfnblock_migratetype(page, pfn); + } + __free_one_page(page, pfn, zone, order, migratetype); +- spin_unlock(&zone->lock); ++ spin_unlock_irqrestore(&zone->lock, flags); + } + + static int free_tail_pages_check(struct page *head_page, struct page *page) +@@ -825,11 +863,11 @@ + return; + + migratetype = get_pfnblock_migratetype(page, pfn); +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + __count_vm_events(PGFREE, 1 << order); + set_freepage_migratetype(page, migratetype); + free_one_page(page_zone(page), page, pfn, order, migratetype); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + void __init __free_pages_bootmem(struct page *page, unsigned int order) +@@ -1371,16 +1409,18 @@ + void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp) + { + unsigned long flags; ++ LIST_HEAD(dst); + int to_drain, batch; + +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + batch = READ_ONCE(pcp->batch); + to_drain = min(pcp->count, batch); + if (to_drain > 0) { +- free_pcppages_bulk(zone, to_drain, pcp); ++ isolate_pcp_pages(to_drain, pcp, &dst); + pcp->count -= to_drain; + } +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); ++ free_pcppages_bulk(zone, to_drain, &dst); + } + #endif + +@@ -1396,16 +1436,21 @@ + unsigned long flags; + struct per_cpu_pageset *pset; + struct per_cpu_pages *pcp; ++ LIST_HEAD(dst); ++ int count; + +- local_irq_save(flags); ++ cpu_lock_irqsave(cpu, flags); + pset = per_cpu_ptr(zone->pageset, cpu); + + pcp = &pset->pcp; +- if (pcp->count) { +- free_pcppages_bulk(zone, pcp->count, pcp); ++ count = pcp->count; ++ if (count) { ++ isolate_pcp_pages(count, pcp, &dst); + pcp->count = 0; + } +- local_irq_restore(flags); ++ cpu_unlock_irqrestore(cpu, flags); ++ if (count) ++ free_pcppages_bulk(zone, count, &dst); + } + + /* +@@ -1491,8 +1536,17 @@ + else + cpumask_clear_cpu(cpu, &cpus_with_pcps); + } ++#ifndef CONFIG_PREEMPT_RT_BASE + on_each_cpu_mask(&cpus_with_pcps, (smp_call_func_t) drain_local_pages, + zone, 1); ++#else ++ for_each_cpu(cpu, &cpus_with_pcps) { ++ if (zone) ++ drain_pages_zone(cpu, zone); ++ else ++ drain_pages(cpu); ++ } ++#endif + } + + #ifdef CONFIG_HIBERNATION +@@ -1548,7 +1602,7 @@ + + migratetype = get_pfnblock_migratetype(page, pfn); + set_freepage_migratetype(page, migratetype); +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + __count_vm_event(PGFREE); + + /* +@@ -1574,12 +1628,17 @@ + pcp->count++; + if (pcp->count >= pcp->high) { + unsigned long batch = READ_ONCE(pcp->batch); +- free_pcppages_bulk(zone, batch, pcp); ++ LIST_HEAD(dst); ++ ++ isolate_pcp_pages(batch, pcp, &dst); + pcp->count -= batch; ++ local_unlock_irqrestore(pa_lock, flags); ++ free_pcppages_bulk(zone, batch, &dst); ++ return; + } + + out: +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + /* +@@ -1710,7 +1769,7 @@ + struct per_cpu_pages *pcp; + struct list_head *list; + +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + pcp = &this_cpu_ptr(zone->pageset)->pcp; + list = &pcp->lists[migratetype]; + if (list_empty(list)) { +@@ -1742,13 +1801,15 @@ + */ + WARN_ON_ONCE(order > 1); + } +- spin_lock_irqsave(&zone->lock, flags); ++ local_spin_lock_irqsave(pa_lock, &zone->lock, flags); + page = __rmqueue(zone, order, migratetype); +- spin_unlock(&zone->lock); +- if (!page) ++ if (!page) { ++ spin_unlock(&zone->lock); + goto failed; ++ } + __mod_zone_freepage_state(zone, -(1 << order), + get_freepage_migratetype(page)); ++ spin_unlock(&zone->lock); + } + + __mod_zone_page_state(zone, NR_ALLOC_BATCH, -(1 << order)); +@@ -1758,13 +1819,13 @@ + + __count_zone_vm_events(PGALLOC, zone, 1 << order); + zone_statistics(preferred_zone, zone, gfp_flags); +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + + VM_BUG_ON_PAGE(bad_range(zone, page), page); + return page; + + failed: +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + return NULL; + } + +@@ -5653,6 +5714,7 @@ + void __init page_alloc_init(void) + { + hotcpu_notifier(page_alloc_cpu_notify, 0); ++ local_irq_lock_init(pa_lock); + } + + /* +@@ -6547,7 +6609,7 @@ + struct per_cpu_pageset *pset; + + /* avoid races with drain_pages() */ +- local_irq_save(flags); ++ local_lock_irqsave(pa_lock, flags); + if (zone->pageset != &boot_pageset) { + for_each_online_cpu(cpu) { + pset = per_cpu_ptr(zone->pageset, cpu); +@@ -6556,7 +6618,7 @@ + free_percpu(zone->pageset); + zone->pageset = &boot_pageset; + } +- local_irq_restore(flags); ++ local_unlock_irqrestore(pa_lock, flags); + } + + #ifdef CONFIG_MEMORY_HOTREMOVE +diff -Nur linux-4.1.13.orig/mm/slab.h linux-4.1.13/mm/slab.h +--- linux-4.1.13.orig/mm/slab.h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/slab.h 2015-11-29 09:23:09.661610728 +0100 +@@ -330,7 +330,11 @@ + * The slab lists for all objects. + */ + struct kmem_cache_node { ++#ifdef CONFIG_SLUB ++ raw_spinlock_t list_lock; ++#else + spinlock_t list_lock; ++#endif + + #ifdef CONFIG_SLAB + struct list_head slabs_partial; /* partial list first, better asm code */ +diff -Nur linux-4.1.13.orig/mm/slub.c linux-4.1.13/mm/slub.c +--- linux-4.1.13.orig/mm/slub.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/slub.c 2015-11-29 09:23:09.665610461 +0100 +@@ -1069,7 +1069,7 @@ + { + struct kmem_cache_node *n = get_node(s, page_to_nid(page)); + +- spin_lock_irqsave(&n->list_lock, *flags); ++ raw_spin_lock_irqsave(&n->list_lock, *flags); + slab_lock(page); + + if (!check_slab(s, page)) +@@ -1116,7 +1116,7 @@ + + fail: + slab_unlock(page); +- spin_unlock_irqrestore(&n->list_lock, *flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, *flags); + slab_fix(s, "Object at 0x%p not freed", object); + return NULL; + } +@@ -1242,6 +1242,12 @@ + + #endif /* CONFIG_SLUB_DEBUG */ + ++struct slub_free_list { ++ raw_spinlock_t lock; ++ struct list_head list; ++}; ++static DEFINE_PER_CPU(struct slub_free_list, slub_free_list); ++ + /* + * Hooks for other subsystems that check memory allocations. In a typical + * production configuration these hooks all should produce no code at all. +@@ -1306,6 +1312,17 @@ + kasan_slab_free(s, x); + } + ++static void setup_object(struct kmem_cache *s, struct page *page, ++ void *object) ++{ ++ setup_object_debug(s, page, object); ++ if (unlikely(s->ctor)) { ++ kasan_unpoison_object_data(s, object); ++ s->ctor(object); ++ kasan_poison_object_data(s, object); ++ } ++} ++ + /* + * Slab allocation and freeing + */ +@@ -1336,10 +1353,17 @@ + struct page *page; + struct kmem_cache_order_objects oo = s->oo; + gfp_t alloc_gfp; ++ void *start, *p; ++ int idx, order; ++ bool enableirqs; + + flags &= gfp_allowed_mask; + +- if (flags & __GFP_WAIT) ++ enableirqs = (flags & __GFP_WAIT) != 0; ++#ifdef CONFIG_PREEMPT_RT_FULL ++ enableirqs |= system_state == SYSTEM_RUNNING; ++#endif ++ if (enableirqs) + local_irq_enable(); + + flags |= s->allocflags; +@@ -1359,13 +1383,13 @@ + * Try a lower order alloc if possible + */ + page = alloc_slab_page(s, alloc_gfp, node, oo); +- +- if (page) +- stat(s, ORDER_FALLBACK); ++ if (unlikely(!page)) ++ goto out; ++ stat(s, ORDER_FALLBACK); + } + +- if (kmemcheck_enabled && page +- && !(s->flags & (SLAB_NOTRACK | DEBUG_DEFAULT_FLAGS))) { ++ if (kmemcheck_enabled && ++ !(s->flags & (SLAB_NOTRACK | DEBUG_DEFAULT_FLAGS))) { + int pages = 1 << oo_order(oo); + + kmemcheck_alloc_shadow(page, oo_order(oo), alloc_gfp, node); +@@ -1380,51 +1404,9 @@ + kmemcheck_mark_unallocated_pages(page, pages); + } + +- if (flags & __GFP_WAIT) +- local_irq_disable(); +- if (!page) +- return NULL; +- + page->objects = oo_objects(oo); +- mod_zone_page_state(page_zone(page), +- (s->flags & SLAB_RECLAIM_ACCOUNT) ? +- NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE, +- 1 << oo_order(oo)); +- +- return page; +-} +- +-static void setup_object(struct kmem_cache *s, struct page *page, +- void *object) +-{ +- setup_object_debug(s, page, object); +- if (unlikely(s->ctor)) { +- kasan_unpoison_object_data(s, object); +- s->ctor(object); +- kasan_poison_object_data(s, object); +- } +-} +- +-static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node) +-{ +- struct page *page; +- void *start; +- void *p; +- int order; +- int idx; +- +- if (unlikely(flags & GFP_SLAB_BUG_MASK)) { +- pr_emerg("gfp: %u\n", flags & GFP_SLAB_BUG_MASK); +- BUG(); +- } +- +- page = allocate_slab(s, +- flags & (GFP_RECLAIM_MASK | GFP_CONSTRAINT_MASK), node); +- if (!page) +- goto out; + + order = compound_order(page); +- inc_slabs_node(s, page_to_nid(page), page->objects); + page->slab_cache = s; + __SetPageSlab(page); + if (page_is_pfmemalloc(page)) +@@ -1448,10 +1430,34 @@ + page->freelist = start; + page->inuse = page->objects; + page->frozen = 1; ++ + out: ++ if (enableirqs) ++ local_irq_disable(); ++ if (!page) ++ return NULL; ++ ++ mod_zone_page_state(page_zone(page), ++ (s->flags & SLAB_RECLAIM_ACCOUNT) ? ++ NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE, ++ 1 << oo_order(oo)); ++ ++ inc_slabs_node(s, page_to_nid(page), page->objects); ++ + return page; + } + ++static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node) ++{ ++ if (unlikely(flags & GFP_SLAB_BUG_MASK)) { ++ pr_emerg("gfp: %u\n", flags & GFP_SLAB_BUG_MASK); ++ BUG(); ++ } ++ ++ return allocate_slab(s, ++ flags & (GFP_RECLAIM_MASK | GFP_CONSTRAINT_MASK), node); ++} ++ + static void __free_slab(struct kmem_cache *s, struct page *page) + { + int order = compound_order(page); +@@ -1483,6 +1489,16 @@ + memcg_uncharge_slab(s, order); + } + ++static void free_delayed(struct list_head *h) ++{ ++ while(!list_empty(h)) { ++ struct page *page = list_first_entry(h, struct page, lru); ++ ++ list_del(&page->lru); ++ __free_slab(page->slab_cache, page); ++ } ++} ++ + #define need_reserve_slab_rcu \ + (sizeof(((struct page *)NULL)->lru) < sizeof(struct rcu_head)) + +@@ -1517,6 +1533,12 @@ + } + + call_rcu(head, rcu_free_slab); ++ } else if (irqs_disabled()) { ++ struct slub_free_list *f = this_cpu_ptr(&slub_free_list); ++ ++ raw_spin_lock(&f->lock); ++ list_add(&page->lru, &f->list); ++ raw_spin_unlock(&f->lock); + } else + __free_slab(s, page); + } +@@ -1630,7 +1652,7 @@ + if (!n || !n->nr_partial) + return NULL; + +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + list_for_each_entry_safe(page, page2, &n->partial, lru) { + void *t; + +@@ -1655,7 +1677,7 @@ + break; + + } +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + return object; + } + +@@ -1901,7 +1923,7 @@ + * that acquire_slab() will see a slab page that + * is frozen + */ +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + } else { + m = M_FULL; +@@ -1912,7 +1934,7 @@ + * slabs from diagnostic functions will not see + * any frozen slabs. + */ +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + } + +@@ -1947,7 +1969,7 @@ + goto redo; + + if (lock) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + if (m == M_FREE) { + stat(s, DEACTIVATE_EMPTY); +@@ -1979,10 +2001,10 @@ + n2 = get_node(s, page_to_nid(page)); + if (n != n2) { + if (n) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + n = n2; +- spin_lock(&n->list_lock); ++ raw_spin_lock(&n->list_lock); + } + + do { +@@ -2011,7 +2033,7 @@ + } + + if (n) +- spin_unlock(&n->list_lock); ++ raw_spin_unlock(&n->list_lock); + + while (discard_page) { + page = discard_page; +@@ -2050,14 +2072,21 @@ + pobjects = oldpage->pobjects; + pages = oldpage->pages; + if (drain && pobjects > s->cpu_partial) { ++ struct slub_free_list *f; + unsigned long flags; ++ LIST_HEAD(tofree); + /* + * partial array is full. Move the existing + * set to the per node partial list. + */ + local_irq_save(flags); + unfreeze_partials(s, this_cpu_ptr(s->cpu_slab)); ++ f = this_cpu_ptr(&slub_free_list); ++ raw_spin_lock(&f->lock); ++ list_splice_init(&f->list, &tofree); ++ raw_spin_unlock(&f->lock); + local_irq_restore(flags); ++ free_delayed(&tofree); + oldpage = NULL; + pobjects = 0; + pages = 0; +@@ -2129,7 +2158,22 @@ + + static void flush_all(struct kmem_cache *s) + { ++ LIST_HEAD(tofree); ++ int cpu; ++ + on_each_cpu_cond(has_cpu_slab, flush_cpu_slab, s, 1, GFP_ATOMIC); ++ for_each_online_cpu(cpu) { ++ struct slub_free_list *f; ++ ++ if (!has_cpu_slab(cpu, s)) ++ continue; ++ ++ f = &per_cpu(slub_free_list, cpu); ++ raw_spin_lock_irq(&f->lock); ++ list_splice_init(&f->list, &tofree); ++ raw_spin_unlock_irq(&f->lock); ++ free_delayed(&tofree); ++ } + } + + /* +@@ -2165,10 +2209,10 @@ + unsigned long x = 0; + struct page *page; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + list_for_each_entry(page, &n->partial, lru) + x += get_count(page); +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return x; + } + #endif /* CONFIG_SLUB_DEBUG || CONFIG_SYSFS */ +@@ -2305,9 +2349,11 @@ + static void *__slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node, + unsigned long addr, struct kmem_cache_cpu *c) + { ++ struct slub_free_list *f; + void *freelist; + struct page *page; + unsigned long flags; ++ LIST_HEAD(tofree); + + local_irq_save(flags); + #ifdef CONFIG_PREEMPT +@@ -2375,7 +2421,13 @@ + VM_BUG_ON(!c->page->frozen); + c->freelist = get_freepointer(s, freelist); + c->tid = next_tid(c->tid); ++out: ++ f = this_cpu_ptr(&slub_free_list); ++ raw_spin_lock(&f->lock); ++ list_splice_init(&f->list, &tofree); ++ raw_spin_unlock(&f->lock); + local_irq_restore(flags); ++ free_delayed(&tofree); + return freelist; + + new_slab: +@@ -2392,8 +2444,7 @@ + + if (unlikely(!freelist)) { + slab_out_of_memory(s, gfpflags, node); +- local_irq_restore(flags); +- return NULL; ++ goto out; + } + + page = c->page; +@@ -2408,8 +2459,7 @@ + deactivate_slab(s, page, get_freepointer(s, freelist)); + c->page = NULL; + c->freelist = NULL; +- local_irq_restore(flags); +- return freelist; ++ goto out; + } + + /* +@@ -2593,7 +2643,7 @@ + + do { + if (unlikely(n)) { +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + n = NULL; + } + prior = page->freelist; +@@ -2625,7 +2675,7 @@ + * Otherwise the list_lock will synchronize with + * other processors updating the list of slabs. + */ +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + } + } +@@ -2667,7 +2717,7 @@ + add_partial(n, page, DEACTIVATE_TO_TAIL); + stat(s, FREE_ADD_PARTIAL); + } +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return; + + slab_empty: +@@ -2682,7 +2732,7 @@ + remove_full(s, n, page); + } + +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + stat(s, FREE_SLAB); + discard_slab(s, page); + } +@@ -2881,7 +2931,7 @@ + init_kmem_cache_node(struct kmem_cache_node *n) + { + n->nr_partial = 0; +- spin_lock_init(&n->list_lock); ++ raw_spin_lock_init(&n->list_lock); + INIT_LIST_HEAD(&n->partial); + #ifdef CONFIG_SLUB_DEBUG + atomic_long_set(&n->nr_slabs, 0); +@@ -3463,7 +3513,7 @@ + for (i = 0; i < SHRINK_PROMOTE_MAX; i++) + INIT_LIST_HEAD(promote + i); + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + /* + * Build lists of slabs to discard or promote. +@@ -3494,7 +3544,7 @@ + for (i = SHRINK_PROMOTE_MAX - 1; i >= 0; i--) + list_splice(promote + i, &n->partial); + +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + + /* Release empty slabs */ + list_for_each_entry_safe(page, t, &discard, lru) +@@ -3670,6 +3720,12 @@ + { + static __initdata struct kmem_cache boot_kmem_cache, + boot_kmem_cache_node; ++ int cpu; ++ ++ for_each_possible_cpu(cpu) { ++ raw_spin_lock_init(&per_cpu(slub_free_list, cpu).lock); ++ INIT_LIST_HEAD(&per_cpu(slub_free_list, cpu).list); ++ } + + if (debug_guardpage_minorder()) + slub_max_order = 0; +@@ -3912,7 +3968,7 @@ + struct page *page; + unsigned long flags; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + + list_for_each_entry(page, &n->partial, lru) { + validate_slab_slab(s, page, map); +@@ -3934,7 +3990,7 @@ + s->name, count, atomic_long_read(&n->nr_slabs)); + + out: +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + return count; + } + +@@ -4122,12 +4178,12 @@ + if (!atomic_long_read(&n->nr_slabs)) + continue; + +- spin_lock_irqsave(&n->list_lock, flags); ++ raw_spin_lock_irqsave(&n->list_lock, flags); + list_for_each_entry(page, &n->partial, lru) + process_slab(&t, s, page, alloc, map); + list_for_each_entry(page, &n->full, lru) + process_slab(&t, s, page, alloc, map); +- spin_unlock_irqrestore(&n->list_lock, flags); ++ raw_spin_unlock_irqrestore(&n->list_lock, flags); + } + + for (i = 0; i < t.count; i++) { +diff -Nur linux-4.1.13.orig/mm/swap.c linux-4.1.13/mm/swap.c +--- linux-4.1.13.orig/mm/swap.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/swap.c 2015-11-29 09:23:09.665610461 +0100 +@@ -32,6 +32,7 @@ + #include <linux/gfp.h> + #include <linux/uio.h> + #include <linux/hugetlb.h> ++#include <linux/locallock.h> + + #include "internal.h" + +@@ -45,6 +46,9 @@ + static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs); + static DEFINE_PER_CPU(struct pagevec, lru_deactivate_file_pvecs); + ++static DEFINE_LOCAL_IRQ_LOCK(rotate_lock); ++DEFINE_LOCAL_IRQ_LOCK(swapvec_lock); ++ + /* + * This path almost never happens for VM activity - pages are normally + * freed via pagevecs. But it gets used by networking. +@@ -481,11 +485,11 @@ + unsigned long flags; + + page_cache_get(page); +- local_irq_save(flags); ++ local_lock_irqsave(rotate_lock, flags); + pvec = this_cpu_ptr(&lru_rotate_pvecs); + if (!pagevec_add(pvec, page)) + pagevec_move_tail(pvec); +- local_irq_restore(flags); ++ local_unlock_irqrestore(rotate_lock, flags); + } + } + +@@ -536,12 +540,13 @@ + void activate_page(struct page *page) + { + if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) { +- struct pagevec *pvec = &get_cpu_var(activate_page_pvecs); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, ++ activate_page_pvecs); + + page_cache_get(page); + if (!pagevec_add(pvec, page)) + pagevec_lru_move_fn(pvec, __activate_page, NULL); +- put_cpu_var(activate_page_pvecs); ++ put_locked_var(swapvec_lock, activate_page_pvecs); + } + } + +@@ -567,7 +572,7 @@ + + static void __lru_cache_activate_page(struct page *page) + { +- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); + int i; + + /* +@@ -589,7 +594,7 @@ + } + } + +- put_cpu_var(lru_add_pvec); ++ put_locked_var(swapvec_lock, lru_add_pvec); + } + + /* +@@ -628,13 +633,13 @@ + + static void __lru_cache_add(struct page *page) + { +- struct pagevec *pvec = &get_cpu_var(lru_add_pvec); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, lru_add_pvec); + + page_cache_get(page); + if (!pagevec_space(pvec)) + __pagevec_lru_add(pvec); + pagevec_add(pvec, page); +- put_cpu_var(lru_add_pvec); ++ put_locked_var(swapvec_lock, lru_add_pvec); + } + + /** +@@ -814,9 +819,9 @@ + unsigned long flags; + + /* No harm done if a racing interrupt already did this */ +- local_irq_save(flags); ++ local_lock_irqsave(rotate_lock, flags); + pagevec_move_tail(pvec); +- local_irq_restore(flags); ++ local_unlock_irqrestore(rotate_lock, flags); + } + + pvec = &per_cpu(lru_deactivate_file_pvecs, cpu); +@@ -844,18 +849,19 @@ + return; + + if (likely(get_page_unless_zero(page))) { +- struct pagevec *pvec = &get_cpu_var(lru_deactivate_file_pvecs); ++ struct pagevec *pvec = &get_locked_var(swapvec_lock, ++ lru_deactivate_file_pvecs); + + if (!pagevec_add(pvec, page)) + pagevec_lru_move_fn(pvec, lru_deactivate_file_fn, NULL); +- put_cpu_var(lru_deactivate_file_pvecs); ++ put_locked_var(swapvec_lock, lru_deactivate_file_pvecs); + } + } + + void lru_add_drain(void) + { +- lru_add_drain_cpu(get_cpu()); +- put_cpu(); ++ lru_add_drain_cpu(local_lock_cpu(swapvec_lock)); ++ local_unlock_cpu(swapvec_lock); + } + + static void lru_add_drain_per_cpu(struct work_struct *dummy) +diff -Nur linux-4.1.13.orig/mm/truncate.c linux-4.1.13/mm/truncate.c +--- linux-4.1.13.orig/mm/truncate.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/truncate.c 2015-11-29 09:23:09.665610461 +0100 +@@ -56,8 +56,11 @@ + * protected by mapping->tree_lock. + */ + if (!workingset_node_shadows(node) && +- !list_empty(&node->private_list)) +- list_lru_del(&workingset_shadow_nodes, &node->private_list); ++ !list_empty(&node->private_list)) { ++ local_lock(workingset_shadow_lock); ++ list_lru_del(&__workingset_shadow_nodes, &node->private_list); ++ local_unlock(workingset_shadow_lock); ++ } + __radix_tree_delete_node(&mapping->page_tree, node); + unlock: + spin_unlock_irq(&mapping->tree_lock); +diff -Nur linux-4.1.13.orig/mm/vmalloc.c linux-4.1.13/mm/vmalloc.c +--- linux-4.1.13.orig/mm/vmalloc.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/vmalloc.c 2015-11-29 09:23:09.665610461 +0100 +@@ -819,7 +819,7 @@ + struct vmap_block *vb; + struct vmap_area *va; + unsigned long vb_idx; +- int node, err; ++ int node, err, cpu; + void *vaddr; + + node = numa_node_id(); +@@ -862,11 +862,12 @@ + BUG_ON(err); + radix_tree_preload_end(); + +- vbq = &get_cpu_var(vmap_block_queue); ++ cpu = get_cpu_light(); ++ vbq = this_cpu_ptr(&vmap_block_queue); + spin_lock(&vbq->lock); + list_add_tail_rcu(&vb->free_list, &vbq->free); + spin_unlock(&vbq->lock); +- put_cpu_var(vmap_block_queue); ++ put_cpu_light(); + + return vaddr; + } +@@ -935,6 +936,7 @@ + struct vmap_block *vb; + void *vaddr = NULL; + unsigned int order; ++ int cpu; + + BUG_ON(size & ~PAGE_MASK); + BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC); +@@ -949,7 +951,8 @@ + order = get_order(size); + + rcu_read_lock(); +- vbq = &get_cpu_var(vmap_block_queue); ++ cpu = get_cpu_light(); ++ vbq = this_cpu_ptr(&vmap_block_queue); + list_for_each_entry_rcu(vb, &vbq->free, free_list) { + unsigned long pages_off; + +@@ -972,7 +975,7 @@ + break; + } + +- put_cpu_var(vmap_block_queue); ++ put_cpu_light(); + rcu_read_unlock(); + + /* Allocate new block if nothing was found */ +diff -Nur linux-4.1.13.orig/mm/vmstat.c linux-4.1.13/mm/vmstat.c +--- linux-4.1.13.orig/mm/vmstat.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/vmstat.c 2015-11-29 09:23:09.665610461 +0100 +@@ -226,6 +226,7 @@ + long x; + long t; + ++ preempt_disable_rt(); + x = delta + __this_cpu_read(*p); + + t = __this_cpu_read(pcp->stat_threshold); +@@ -235,6 +236,7 @@ + x = 0; + } + __this_cpu_write(*p, x); ++ preempt_enable_rt(); + } + EXPORT_SYMBOL(__mod_zone_page_state); + +@@ -267,6 +269,7 @@ + s8 __percpu *p = pcp->vm_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_inc_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v > t)) { +@@ -275,6 +278,7 @@ + zone_page_state_add(v + overstep, zone, item); + __this_cpu_write(*p, -overstep); + } ++ preempt_enable_rt(); + } + + void __inc_zone_page_state(struct page *page, enum zone_stat_item item) +@@ -289,6 +293,7 @@ + s8 __percpu *p = pcp->vm_stat_diff + item; + s8 v, t; + ++ preempt_disable_rt(); + v = __this_cpu_dec_return(*p); + t = __this_cpu_read(pcp->stat_threshold); + if (unlikely(v < - t)) { +@@ -297,6 +302,7 @@ + zone_page_state_add(v - overstep, zone, item); + __this_cpu_write(*p, overstep); + } ++ preempt_enable_rt(); + } + + void __dec_zone_page_state(struct page *page, enum zone_stat_item item) +diff -Nur linux-4.1.13.orig/mm/workingset.c linux-4.1.13/mm/workingset.c +--- linux-4.1.13.orig/mm/workingset.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/mm/workingset.c 2015-11-29 09:23:09.665610461 +0100 +@@ -264,7 +264,8 @@ + * point where they would still be useful. + */ + +-struct list_lru workingset_shadow_nodes; ++struct list_lru __workingset_shadow_nodes; ++DEFINE_LOCAL_IRQ_LOCK(workingset_shadow_lock); + + static unsigned long count_shadow_nodes(struct shrinker *shrinker, + struct shrink_control *sc) +@@ -274,9 +275,9 @@ + unsigned long pages; + + /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ +- local_irq_disable(); +- shadow_nodes = list_lru_shrink_count(&workingset_shadow_nodes, sc); +- local_irq_enable(); ++ local_lock_irq(workingset_shadow_lock); ++ shadow_nodes = list_lru_shrink_count(&__workingset_shadow_nodes, sc); ++ local_unlock_irq(workingset_shadow_lock); + + pages = node_present_pages(sc->nid); + /* +@@ -363,9 +364,9 @@ + spin_unlock(&mapping->tree_lock); + ret = LRU_REMOVED_RETRY; + out: +- local_irq_enable(); ++ local_unlock_irq(workingset_shadow_lock); + cond_resched(); +- local_irq_disable(); ++ local_lock_irq(workingset_shadow_lock); + spin_lock(lru_lock); + return ret; + } +@@ -376,10 +377,10 @@ + unsigned long ret; + + /* list_lru lock nests inside IRQ-safe mapping->tree_lock */ +- local_irq_disable(); +- ret = list_lru_shrink_walk(&workingset_shadow_nodes, sc, ++ local_lock_irq(workingset_shadow_lock); ++ ret = list_lru_shrink_walk(&__workingset_shadow_nodes, sc, + shadow_lru_isolate, NULL); +- local_irq_enable(); ++ local_unlock_irq(workingset_shadow_lock); + return ret; + } + +@@ -400,7 +401,7 @@ + { + int ret; + +- ret = list_lru_init_key(&workingset_shadow_nodes, &shadow_nodes_key); ++ ret = list_lru_init_key(&__workingset_shadow_nodes, &shadow_nodes_key); + if (ret) + goto err; + ret = register_shrinker(&workingset_shadow_shrinker); +@@ -408,7 +409,7 @@ + goto err_list_lru; + return 0; + err_list_lru: +- list_lru_destroy(&workingset_shadow_nodes); ++ list_lru_destroy(&__workingset_shadow_nodes); + err: + return ret; + } +diff -Nur linux-4.1.13.orig/net/core/dev.c linux-4.1.13/net/core/dev.c +--- linux-4.1.13.orig/net/core/dev.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/core/dev.c 2015-11-29 09:23:09.665610461 +0100 +@@ -184,6 +184,7 @@ + static DEFINE_HASHTABLE(napi_hash, 8); + + static seqcount_t devnet_rename_seq; ++static DEFINE_MUTEX(devnet_rename_mutex); + + static inline void dev_base_seq_inc(struct net *net) + { +@@ -205,14 +206,14 @@ + static inline void rps_lock(struct softnet_data *sd) + { + #ifdef CONFIG_RPS +- spin_lock(&sd->input_pkt_queue.lock); ++ raw_spin_lock(&sd->input_pkt_queue.raw_lock); + #endif + } + + static inline void rps_unlock(struct softnet_data *sd) + { + #ifdef CONFIG_RPS +- spin_unlock(&sd->input_pkt_queue.lock); ++ raw_spin_unlock(&sd->input_pkt_queue.raw_lock); + #endif + } + +@@ -852,7 +853,8 @@ + strcpy(name, dev->name); + rcu_read_unlock(); + if (read_seqcount_retry(&devnet_rename_seq, seq)) { +- cond_resched(); ++ mutex_lock(&devnet_rename_mutex); ++ mutex_unlock(&devnet_rename_mutex); + goto retry; + } + +@@ -1121,20 +1123,17 @@ + if (dev->flags & IFF_UP) + return -EBUSY; + +- write_seqcount_begin(&devnet_rename_seq); ++ mutex_lock(&devnet_rename_mutex); ++ __raw_write_seqcount_begin(&devnet_rename_seq); + +- if (strncmp(newname, dev->name, IFNAMSIZ) == 0) { +- write_seqcount_end(&devnet_rename_seq); +- return 0; +- } ++ if (strncmp(newname, dev->name, IFNAMSIZ) == 0) ++ goto outunlock; + + memcpy(oldname, dev->name, IFNAMSIZ); + + err = dev_get_valid_name(net, dev, newname); +- if (err < 0) { +- write_seqcount_end(&devnet_rename_seq); +- return err; +- } ++ if (err < 0) ++ goto outunlock; + + if (oldname[0] && !strchr(oldname, '%')) + netdev_info(dev, "renamed from %s\n", oldname); +@@ -1147,11 +1146,12 @@ + if (ret) { + memcpy(dev->name, oldname, IFNAMSIZ); + dev->name_assign_type = old_assign_type; +- write_seqcount_end(&devnet_rename_seq); +- return ret; ++ err = ret; ++ goto outunlock; + } + +- write_seqcount_end(&devnet_rename_seq); ++ __raw_write_seqcount_end(&devnet_rename_seq); ++ mutex_unlock(&devnet_rename_mutex); + + netdev_adjacent_rename_links(dev, oldname); + +@@ -1172,7 +1172,8 @@ + /* err >= 0 after dev_alloc_name() or stores the first errno */ + if (err >= 0) { + err = ret; +- write_seqcount_begin(&devnet_rename_seq); ++ mutex_lock(&devnet_rename_mutex); ++ __raw_write_seqcount_begin(&devnet_rename_seq); + memcpy(dev->name, oldname, IFNAMSIZ); + memcpy(oldname, newname, IFNAMSIZ); + dev->name_assign_type = old_assign_type; +@@ -1185,6 +1186,11 @@ + } + + return err; ++ ++outunlock: ++ __raw_write_seqcount_end(&devnet_rename_seq); ++ mutex_unlock(&devnet_rename_mutex); ++ return err; + } + + /** +@@ -2214,6 +2220,7 @@ + sd->output_queue_tailp = &q->next_sched; + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + + void __netif_schedule(struct Qdisc *q) +@@ -2295,6 +2302,7 @@ + __this_cpu_write(softnet_data.completion_queue, skb); + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(__dev_kfree_skb_irq); + +@@ -3365,6 +3373,7 @@ + rps_unlock(sd); + + local_irq_restore(flags); ++ preempt_check_resched_rt(); + + atomic_long_inc(&skb->dev->rx_dropped); + kfree_skb(skb); +@@ -3383,7 +3392,7 @@ + struct rps_dev_flow voidflow, *rflow = &voidflow; + int cpu; + +- preempt_disable(); ++ migrate_disable(); + rcu_read_lock(); + + cpu = get_rps_cpu(skb->dev, skb, &rflow); +@@ -3393,13 +3402,13 @@ + ret = enqueue_to_backlog(skb, cpu, &rflow->last_qtail); + + rcu_read_unlock(); +- preempt_enable(); ++ migrate_enable(); + } else + #endif + { + unsigned int qtail; +- ret = enqueue_to_backlog(skb, get_cpu(), &qtail); +- put_cpu(); ++ ret = enqueue_to_backlog(skb, get_cpu_light(), &qtail); ++ put_cpu_light(); + } + return ret; + } +@@ -3433,16 +3442,44 @@ + + trace_netif_rx_ni_entry(skb); + +- preempt_disable(); ++ local_bh_disable(); + err = netif_rx_internal(skb); +- if (local_softirq_pending()) +- do_softirq(); +- preempt_enable(); ++ local_bh_enable(); + + return err; + } + EXPORT_SYMBOL(netif_rx_ni); + ++#ifdef CONFIG_PREEMPT_RT_FULL ++/* ++ * RT runs ksoftirqd as a real time thread and the root_lock is a ++ * "sleeping spinlock". If the trylock fails then we can go into an ++ * infinite loop when ksoftirqd preempted the task which actually ++ * holds the lock, because we requeue q and raise NET_TX softirq ++ * causing ksoftirqd to loop forever. ++ * ++ * It's safe to use spin_lock on RT here as softirqs run in thread ++ * context and cannot deadlock against the thread which is holding ++ * root_lock. ++ * ++ * On !RT the trylock might fail, but there we bail out from the ++ * softirq loop after 10 attempts which we can't do on RT. And the ++ * task holding root_lock cannot be preempted, so the only downside of ++ * that trylock is that we need 10 loops to decide that we should have ++ * given up in the first one :) ++ */ ++static inline int take_root_lock(spinlock_t *lock) ++{ ++ spin_lock(lock); ++ return 1; ++} ++#else ++static inline int take_root_lock(spinlock_t *lock) ++{ ++ return spin_trylock(lock); ++} ++#endif ++ + static void net_tx_action(struct softirq_action *h) + { + struct softnet_data *sd = this_cpu_ptr(&softnet_data); +@@ -3484,7 +3521,7 @@ + head = head->next_sched; + + root_lock = qdisc_lock(q); +- if (spin_trylock(root_lock)) { ++ if (take_root_lock(root_lock)) { + smp_mb__before_atomic(); + clear_bit(__QDISC_STATE_SCHED, + &q->state); +@@ -3881,7 +3918,7 @@ + skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) { + if (skb->dev == dev) { + __skb_unlink(skb, &sd->input_pkt_queue); +- kfree_skb(skb); ++ __skb_queue_tail(&sd->tofree_queue, skb); + input_queue_head_incr(sd); + } + } +@@ -3890,10 +3927,13 @@ + skb_queue_walk_safe(&sd->process_queue, skb, tmp) { + if (skb->dev == dev) { + __skb_unlink(skb, &sd->process_queue); +- kfree_skb(skb); ++ __skb_queue_tail(&sd->tofree_queue, skb); + input_queue_head_incr(sd); + } + } ++ ++ if (!skb_queue_empty(&sd->tofree_queue)) ++ raise_softirq_irqoff(NET_RX_SOFTIRQ); + } + + static int napi_gro_complete(struct sk_buff *skb) +@@ -4344,6 +4384,7 @@ + sd->rps_ipi_list = NULL; + + local_irq_enable(); ++ preempt_check_resched_rt(); + + /* Send pending IPI's to kick RPS processing on remote cpus. */ + while (remsd) { +@@ -4357,6 +4398,7 @@ + } else + #endif + local_irq_enable(); ++ preempt_check_resched_rt(); + } + + static bool sd_has_rps_ipi_waiting(struct softnet_data *sd) +@@ -4438,6 +4480,7 @@ + local_irq_save(flags); + ____napi_schedule(this_cpu_ptr(&softnet_data), n); + local_irq_restore(flags); ++ preempt_check_resched_rt(); + } + EXPORT_SYMBOL(__napi_schedule); + +@@ -6926,7 +6969,7 @@ + void synchronize_net(void) + { + might_sleep(); +- if (rtnl_is_locked()) ++ if (rtnl_is_locked() && !IS_ENABLED(CONFIG_PREEMPT_RT_FULL)) + synchronize_rcu_expedited(); + else + synchronize_rcu(); +@@ -7167,16 +7210,20 @@ + + raise_softirq_irqoff(NET_TX_SOFTIRQ); + local_irq_enable(); ++ preempt_check_resched_rt(); + + /* Process offline CPU's input_pkt_queue */ + while ((skb = __skb_dequeue(&oldsd->process_queue))) { + netif_rx_ni(skb); + input_queue_head_incr(oldsd); + } +- while ((skb = skb_dequeue(&oldsd->input_pkt_queue))) { ++ while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) { + netif_rx_ni(skb); + input_queue_head_incr(oldsd); + } ++ while ((skb = __skb_dequeue(&oldsd->tofree_queue))) { ++ kfree_skb(skb); ++ } + + return NOTIFY_OK; + } +@@ -7478,8 +7525,9 @@ + for_each_possible_cpu(i) { + struct softnet_data *sd = &per_cpu(softnet_data, i); + +- skb_queue_head_init(&sd->input_pkt_queue); +- skb_queue_head_init(&sd->process_queue); ++ skb_queue_head_init_raw(&sd->input_pkt_queue); ++ skb_queue_head_init_raw(&sd->process_queue); ++ skb_queue_head_init_raw(&sd->tofree_queue); + INIT_LIST_HEAD(&sd->poll_list); + sd->output_queue_tailp = &sd->output_queue; + #ifdef CONFIG_RPS +diff -Nur linux-4.1.13.orig/net/core/skbuff.c linux-4.1.13/net/core/skbuff.c +--- linux-4.1.13.orig/net/core/skbuff.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/core/skbuff.c 2015-11-29 09:23:09.669610195 +0100 +@@ -63,6 +63,7 @@ + #include <linux/errqueue.h> + #include <linux/prefetch.h> + #include <linux/if_vlan.h> ++#include <linux/locallock.h> + + #include <net/protocol.h> + #include <net/dst.h> +@@ -356,6 +357,7 @@ + }; + static DEFINE_PER_CPU(struct netdev_alloc_cache, netdev_alloc_cache); + static DEFINE_PER_CPU(struct netdev_alloc_cache, napi_alloc_cache); ++static DEFINE_LOCAL_IRQ_LOCK(netdev_alloc_lock); + + static struct page *__page_frag_refill(struct netdev_alloc_cache *nc, + gfp_t gfp_mask) +@@ -433,9 +435,9 @@ + unsigned long flags; + void *data; + +- local_irq_save(flags); ++ local_lock_irqsave(netdev_alloc_lock, flags); + data = __alloc_page_frag(&netdev_alloc_cache, fragsz, gfp_mask); +- local_irq_restore(flags); ++ local_unlock_irqrestore(netdev_alloc_lock, flags); + return data; + } + +diff -Nur linux-4.1.13.orig/net/core/sock.c linux-4.1.13/net/core/sock.c +--- linux-4.1.13.orig/net/core/sock.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/core/sock.c 2015-11-29 09:23:09.669610195 +0100 +@@ -2370,12 +2370,11 @@ + if (sk->sk_lock.owned) + __lock_sock(sk); + sk->sk_lock.owned = 1; +- spin_unlock(&sk->sk_lock.slock); ++ spin_unlock_bh(&sk->sk_lock.slock); + /* + * The sk_lock has mutex_lock() semantics here: + */ + mutex_acquire(&sk->sk_lock.dep_map, subclass, 0, _RET_IP_); +- local_bh_enable(); + } + EXPORT_SYMBOL(lock_sock_nested); + +diff -Nur linux-4.1.13.orig/net/ipv4/icmp.c linux-4.1.13/net/ipv4/icmp.c +--- linux-4.1.13.orig/net/ipv4/icmp.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/ipv4/icmp.c 2015-11-29 09:23:09.669610195 +0100 +@@ -69,6 +69,7 @@ + #include <linux/jiffies.h> + #include <linux/kernel.h> + #include <linux/fcntl.h> ++#include <linux/sysrq.h> + #include <linux/socket.h> + #include <linux/in.h> + #include <linux/inet.h> +@@ -867,6 +868,30 @@ + } + + /* ++ * 32bit and 64bit have different timestamp length, so we check for ++ * the cookie at offset 20 and verify it is repeated at offset 50 ++ */ ++#define CO_POS0 20 ++#define CO_POS1 50 ++#define CO_SIZE sizeof(int) ++#define ICMP_SYSRQ_SIZE 57 ++ ++/* ++ * We got a ICMP_SYSRQ_SIZE sized ping request. Check for the cookie ++ * pattern and if it matches send the next byte as a trigger to sysrq. ++ */ ++static void icmp_check_sysrq(struct net *net, struct sk_buff *skb) ++{ ++ int cookie = htonl(net->ipv4.sysctl_icmp_echo_sysrq); ++ char *p = skb->data; ++ ++ if (!memcmp(&cookie, p + CO_POS0, CO_SIZE) && ++ !memcmp(&cookie, p + CO_POS1, CO_SIZE) && ++ p[CO_POS0 + CO_SIZE] == p[CO_POS1 + CO_SIZE]) ++ handle_sysrq(p[CO_POS0 + CO_SIZE]); ++} ++ ++/* + * Handle ICMP_ECHO ("ping") requests. + * + * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo +@@ -893,6 +918,11 @@ + icmp_param.data_len = skb->len; + icmp_param.head_len = sizeof(struct icmphdr); + icmp_reply(&icmp_param, skb); ++ ++ if (skb->len == ICMP_SYSRQ_SIZE && ++ net->ipv4.sysctl_icmp_echo_sysrq) { ++ icmp_check_sysrq(net, skb); ++ } + } + /* should there be an ICMP stat for ignored echos? */ + return true; +diff -Nur linux-4.1.13.orig/net/ipv4/sysctl_net_ipv4.c linux-4.1.13/net/ipv4/sysctl_net_ipv4.c +--- linux-4.1.13.orig/net/ipv4/sysctl_net_ipv4.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/ipv4/sysctl_net_ipv4.c 2015-11-29 09:23:09.669610195 +0100 +@@ -779,6 +779,13 @@ + .proc_handler = proc_dointvec + }, + { ++ .procname = "icmp_echo_sysrq", ++ .data = &init_net.ipv4.sysctl_icmp_echo_sysrq, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = proc_dointvec ++ }, ++ { + .procname = "icmp_ignore_bogus_error_responses", + .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses, + .maxlen = sizeof(int), +diff -Nur linux-4.1.13.orig/net/mac80211/rx.c linux-4.1.13/net/mac80211/rx.c +--- linux-4.1.13.orig/net/mac80211/rx.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/mac80211/rx.c 2015-11-29 09:23:09.669610195 +0100 +@@ -3554,7 +3554,7 @@ + struct ieee80211_supported_band *sband; + struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb); + +- WARN_ON_ONCE(softirq_count() == 0); ++ WARN_ON_ONCE_NONRT(softirq_count() == 0); + + if (WARN_ON(status->band >= IEEE80211_NUM_BANDS)) + goto drop; +diff -Nur linux-4.1.13.orig/net/netfilter/core.c linux-4.1.13/net/netfilter/core.c +--- linux-4.1.13.orig/net/netfilter/core.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/netfilter/core.c 2015-11-29 09:23:09.669610195 +0100 +@@ -22,11 +22,17 @@ + #include <linux/proc_fs.h> + #include <linux/mutex.h> + #include <linux/slab.h> ++#include <linux/locallock.h> + #include <net/net_namespace.h> + #include <net/sock.h> + + #include "nf_internals.h" + ++#ifdef CONFIG_PREEMPT_RT_BASE ++DEFINE_LOCAL_IRQ_LOCK(xt_write_lock); ++EXPORT_PER_CPU_SYMBOL(xt_write_lock); ++#endif ++ + static DEFINE_MUTEX(afinfo_mutex); + + const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly; +diff -Nur linux-4.1.13.orig/net/packet/af_packet.c linux-4.1.13/net/packet/af_packet.c +--- linux-4.1.13.orig/net/packet/af_packet.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/packet/af_packet.c 2015-11-29 09:23:09.669610195 +0100 +@@ -63,6 +63,7 @@ + #include <linux/if_packet.h> + #include <linux/wireless.h> + #include <linux/kernel.h> ++#include <linux/delay.h> + #include <linux/kmod.h> + #include <linux/slab.h> + #include <linux/vmalloc.h> +@@ -698,7 +699,7 @@ + if (BLOCK_NUM_PKTS(pbd)) { + while (atomic_read(&pkc->blk_fill_in_prog)) { + /* Waiting for skb_copy_bits to finish... */ +- cpu_relax(); ++ cpu_chill(); + } + } + +@@ -960,7 +961,7 @@ + if (!(status & TP_STATUS_BLK_TMO)) { + while (atomic_read(&pkc->blk_fill_in_prog)) { + /* Waiting for skb_copy_bits to finish... */ +- cpu_relax(); ++ cpu_chill(); + } + } + prb_close_block(pkc, pbd, po, status); +diff -Nur linux-4.1.13.orig/net/rds/ib_rdma.c linux-4.1.13/net/rds/ib_rdma.c +--- linux-4.1.13.orig/net/rds/ib_rdma.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/rds/ib_rdma.c 2015-11-29 09:23:09.669610195 +0100 +@@ -34,6 +34,7 @@ + #include <linux/slab.h> + #include <linux/rculist.h> + #include <linux/llist.h> ++#include <linux/delay.h> + + #include "rds.h" + #include "ib.h" +@@ -286,7 +287,7 @@ + for_each_online_cpu(cpu) { + flag = &per_cpu(clean_list_grace, cpu); + while (test_bit(CLEAN_LIST_BUSY_BIT, flag)) +- cpu_relax(); ++ cpu_chill(); + } + } + +diff -Nur linux-4.1.13.orig/net/sched/sch_generic.c linux-4.1.13/net/sched/sch_generic.c +--- linux-4.1.13.orig/net/sched/sch_generic.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/sched/sch_generic.c 2015-11-29 09:23:09.669610195 +0100 +@@ -894,7 +894,7 @@ + /* Wait for outstanding qdisc_run calls. */ + list_for_each_entry(dev, head, close_list) + while (some_qdisc_is_busy(dev)) +- yield(); ++ msleep(1); + } + + void dev_deactivate(struct net_device *dev) +diff -Nur linux-4.1.13.orig/net/sunrpc/svc_xprt.c linux-4.1.13/net/sunrpc/svc_xprt.c +--- linux-4.1.13.orig/net/sunrpc/svc_xprt.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/net/sunrpc/svc_xprt.c 2015-11-29 09:23:09.669610195 +0100 +@@ -341,7 +341,7 @@ + goto out; + } + +- cpu = get_cpu(); ++ cpu = get_cpu_light(); + pool = svc_pool_for_cpu(xprt->xpt_server, cpu); + + atomic_long_inc(&pool->sp_stats.packets); +@@ -377,7 +377,7 @@ + + atomic_long_inc(&pool->sp_stats.threads_woken); + wake_up_process(rqstp->rq_task); +- put_cpu(); ++ put_cpu_light(); + goto out; + } + rcu_read_unlock(); +@@ -398,7 +398,7 @@ + goto redo_search; + } + rqstp = NULL; +- put_cpu(); ++ put_cpu_light(); + out: + trace_svc_xprt_do_enqueue(xprt, rqstp); + } +diff -Nur linux-4.1.13.orig/scripts/mkcompile_h linux-4.1.13/scripts/mkcompile_h +--- linux-4.1.13.orig/scripts/mkcompile_h 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/scripts/mkcompile_h 2015-11-29 09:23:09.669610195 +0100 +@@ -4,7 +4,8 @@ + ARCH=$2 + SMP=$3 + PREEMPT=$4 +-CC=$5 ++RT=$5 ++CC=$6 + + vecho() { [ "${quiet}" = "silent_" ] || echo "$@" ; } + +@@ -57,6 +58,7 @@ + CONFIG_FLAGS="" + if [ -n "$SMP" ] ; then CONFIG_FLAGS="SMP"; fi + if [ -n "$PREEMPT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS PREEMPT"; fi ++if [ -n "$RT" ] ; then CONFIG_FLAGS="$CONFIG_FLAGS RT"; fi + UTS_VERSION="$UTS_VERSION $CONFIG_FLAGS $TIMESTAMP" + + # Truncate to maximum length +diff -Nur linux-4.1.13.orig/sound/core/pcm_native.c linux-4.1.13/sound/core/pcm_native.c +--- linux-4.1.13.orig/sound/core/pcm_native.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/sound/core/pcm_native.c 2015-11-29 09:23:09.669610195 +0100 +@@ -123,7 +123,7 @@ + void snd_pcm_stream_lock_irq(struct snd_pcm_substream *substream) + { + if (!substream->pcm->nonatomic) +- local_irq_disable(); ++ local_irq_disable_nort(); + snd_pcm_stream_lock(substream); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_lock_irq); +@@ -138,7 +138,7 @@ + { + snd_pcm_stream_unlock(substream); + if (!substream->pcm->nonatomic) +- local_irq_enable(); ++ local_irq_enable_nort(); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irq); + +@@ -146,7 +146,7 @@ + { + unsigned long flags = 0; + if (!substream->pcm->nonatomic) +- local_irq_save(flags); ++ local_irq_save_nort(flags); + snd_pcm_stream_lock(substream); + return flags; + } +@@ -164,7 +164,7 @@ + { + snd_pcm_stream_unlock(substream); + if (!substream->pcm->nonatomic) +- local_irq_restore(flags); ++ local_irq_restore_nort(flags); + } + EXPORT_SYMBOL_GPL(snd_pcm_stream_unlock_irqrestore); + +diff -Nur linux-4.1.13.orig/sound/soc/intel/atom/sst/sst.c linux-4.1.13/sound/soc/intel/atom/sst/sst.c +--- linux-4.1.13.orig/sound/soc/intel/atom/sst/sst.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/sound/soc/intel/atom/sst/sst.c 2015-11-29 09:23:09.673609931 +0100 +@@ -368,8 +368,8 @@ + * initialize by FW or driver when firmware is loaded + */ + spin_lock_irqsave(&ctx->ipc_spin_lock, irq_flags); +- sst_shim_write64(shim, SST_IMRX, shim_regs->imrx), +- sst_shim_write64(shim, SST_CSR, shim_regs->csr), ++ sst_shim_write64(shim, SST_IMRX, shim_regs->imrx); ++ sst_shim_write64(shim, SST_CSR, shim_regs->csr); + spin_unlock_irqrestore(&ctx->ipc_spin_lock, irq_flags); + } + +diff -Nur linux-4.1.13.orig/virt/kvm/async_pf.c linux-4.1.13/virt/kvm/async_pf.c +--- linux-4.1.13.orig/virt/kvm/async_pf.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/virt/kvm/async_pf.c 2015-11-29 09:23:09.673609931 +0100 +@@ -94,8 +94,8 @@ + + trace_kvm_async_pf_completed(addr, gva); + +- if (waitqueue_active(&vcpu->wq)) +- wake_up_interruptible(&vcpu->wq); ++ if (swaitqueue_active(&vcpu->wq)) ++ swait_wake_interruptible(&vcpu->wq); + + mmput(mm); + kvm_put_kvm(vcpu->kvm); +diff -Nur linux-4.1.13.orig/virt/kvm/kvm_main.c linux-4.1.13/virt/kvm/kvm_main.c +--- linux-4.1.13.orig/virt/kvm/kvm_main.c 2015-11-09 23:34:10.000000000 +0100 ++++ linux-4.1.13/virt/kvm/kvm_main.c 2015-11-29 09:23:09.701608072 +0100 +@@ -218,7 +218,7 @@ + vcpu->kvm = kvm; + vcpu->vcpu_id = id; + vcpu->pid = NULL; +- init_waitqueue_head(&vcpu->wq); ++ init_swait_head(&vcpu->wq); + kvm_async_pf_vcpu_init(vcpu); + + page = alloc_page(GFP_KERNEL | __GFP_ZERO); +@@ -1779,7 +1779,7 @@ + void kvm_vcpu_block(struct kvm_vcpu *vcpu) + { + ktime_t start, cur; +- DEFINE_WAIT(wait); ++ DEFINE_SWAITER(wait); + bool waited = false; + + start = cur = ktime_get(); +@@ -1800,7 +1800,7 @@ + } + + for (;;) { +- prepare_to_wait(&vcpu->wq, &wait, TASK_INTERRUPTIBLE); ++ swait_prepare(&vcpu->wq, &wait, TASK_INTERRUPTIBLE); + + if (kvm_vcpu_check_block(vcpu) < 0) + break; +@@ -1809,7 +1809,7 @@ + schedule(); + } + +- finish_wait(&vcpu->wq, &wait); ++ swait_finish(&vcpu->wq, &wait); + cur = ktime_get(); + + out: +@@ -1825,11 +1825,11 @@ + { + int me; + int cpu = vcpu->cpu; +- wait_queue_head_t *wqp; ++ struct swait_head *wqp; + + wqp = kvm_arch_vcpu_wq(vcpu); +- if (waitqueue_active(wqp)) { +- wake_up_interruptible(wqp); ++ if (swaitqueue_active(wqp)) { ++ swait_wake_interruptible(wqp); + ++vcpu->stat.halt_wakeup; + } + +@@ -1930,7 +1930,7 @@ + continue; + if (vcpu == me) + continue; +- if (waitqueue_active(&vcpu->wq) && !kvm_arch_vcpu_runnable(vcpu)) ++ if (swaitqueue_active(&vcpu->wq) && !kvm_arch_vcpu_runnable(vcpu)) + continue; + if (!kvm_vcpu_eligible_for_directed_yield(vcpu)) + continue; diff --git a/target/linux/patches/4.1.13/regmap-default-on.patch b/target/linux/patches/4.1.13/regmap-default-on.patch new file mode 100644 index 000000000..8d72224bf --- /dev/null +++ b/target/linux/patches/4.1.13/regmap-default-on.patch @@ -0,0 +1,17 @@ +diff -Nur linux-4.1.6.orig/drivers/base/regmap/Kconfig linux-4.1.6/drivers/base/regmap/Kconfig +--- linux-4.1.6.orig/drivers/base/regmap/Kconfig 2015-08-17 05:52:51.000000000 +0200 ++++ linux-4.1.6/drivers/base/regmap/Kconfig 2015-08-29 22:18:50.329683337 +0200 +@@ -3,7 +3,7 @@ + # subsystems should select the appropriate symbols. + + config REGMAP +- default y if (REGMAP_I2C || REGMAP_SPI || REGMAP_SPMI || REGMAP_AC97 || REGMAP_MMIO || REGMAP_IRQ) ++ default y + select LZO_COMPRESS + select LZO_DECOMPRESS + select IRQ_DOMAIN if REGMAP_IRQ +@@ -29,3 +29,4 @@ + + config REGMAP_IRQ + bool ++ default y diff --git a/target/linux/patches/4.1.13/remove-warn.patch b/target/linux/patches/4.1.13/remove-warn.patch new file mode 100644 index 000000000..1f89c710d --- /dev/null +++ b/target/linux/patches/4.1.13/remove-warn.patch @@ -0,0 +1,11 @@ +diff -Nur linux-4.1.10.orig/drivers/media/v4l2-core/videobuf2-core.c linux-4.1.10/drivers/media/v4l2-core/videobuf2-core.c +--- linux-4.1.10.orig/drivers/media/v4l2-core/videobuf2-core.c 2015-10-03 13:49:38.000000000 +0200 ++++ linux-4.1.10/drivers/media/v4l2-core/videobuf2-core.c 2015-10-18 18:18:47.000000000 +0200 +@@ -1245,7 +1245,6 @@ + return; + + __check_once = true; +- __WARN(); + + pr_warn_once("use of bytesused == 0 is deprecated and will be removed in the future,\n"); + if (vb->vb2_queue->allow_zero_bytesused) diff --git a/target/linux/patches/4.1.13/startup.patch b/target/linux/patches/4.1.13/startup.patch new file mode 100644 index 000000000..d396b75e4 --- /dev/null +++ b/target/linux/patches/4.1.13/startup.patch @@ -0,0 +1,37 @@ +diff -Nur linux-3.13.3.orig/init/main.c linux-3.13.3/init/main.c +--- linux-3.13.3.orig/init/main.c 2014-02-13 23:00:14.000000000 +0100 ++++ linux-3.13.3/init/main.c 2014-02-17 11:35:14.000000000 +0100 +@@ -916,6 +917,8 @@ + if (sys_open((const char __user *) "/dev/console", O_RDWR, 0) < 0) + pr_err("Warning: unable to open an initial console.\n"); + ++ printk(KERN_WARNING "Starting Linux (built with OpenADK).\n"); ++ + (void) sys_dup(0); + (void) sys_dup(0); + /* +diff -Nur linux-3.13.6.orig/init/initramfs.c linux-3.13.6/init/initramfs.c +--- linux-3.13.6.orig/init/initramfs.c 2014-03-07 07:07:02.000000000 +0100 ++++ linux-3.13.6/init/initramfs.c 2014-03-15 12:11:31.882731916 +0100 +@@ -622,6 +622,9 @@ + */ + load_default_modules(); + } ++#ifdef CONFIG_DEVTMPFS_MOUNT ++ devtmpfs_mount("dev"); ++#endif + return 0; + } + rootfs_initcall(populate_rootfs); +diff -Nur linux-3.13.6.orig/init/main.c linux-3.13.6/init/main.c +--- linux-3.13.6.orig/init/main.c 2014-03-07 07:07:02.000000000 +0100 ++++ linux-3.13.6/init/main.c 2014-03-15 12:13:16.459024452 +0100 +@@ -924,7 +924,7 @@ + */ + + if (!ramdisk_execute_command) +- ramdisk_execute_command = "/init"; ++ ramdisk_execute_command = "/sbin/init"; + + if (sys_access((const char __user *) ramdisk_execute_command, 0) != 0) { + ramdisk_execute_command = NULL; |