summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2010-04-12silence warning about undefined preprocessor tokenBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-12libc_regex: __libc_lock primitives are actually available on uClibcCarmelo Amoroso
__libc_lock primitives are actually available on uClibc when threading support is enable, so in this case they can be used. It also fixes the following compiler warnings: In file included from libc/misc/regex/regex.c:55: libc/misc/regex/regex_internal.h:49:1: warning: "__libc_lock_define" redefined [SNIP] libc/misc/regex/regex_internal.h:50:1: warning: "__libc_lock_init" redefined [SNIP] libc/misc/regex/regex_internal.h:51:1: warning: "__libc_lock_lock" redefined [SNIP] libc/misc/regex/regex_internal.h:52:1: warning: "__libc_lock_unlock" redefined Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com> (cherry picked from commit 65f9ccdafd008abd9892dfc46fb9737ec4d964c5) Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-04-11ldso_sh: Provide only one definition for elf_machine_type_classCarmelo Amoroso
It doesn't need to provide different definition of elf_machine_type_class macro depending on TLS support. It also fixes build for sh architecture when thread support is off: CC ldso/ldso/ldso.oS In file included from ./ldso/ldso/sh/dl-sysdep.h:9, from ./ldso/include/dl-string.h:11, from ./ldso/include/ldso.h:40, from ldso/ldso/ldso.c:33: ./include/tls.h:6:22: error: tls.h: No such file or directory make: *** [ldso/ldso/ldso.oS] Error 1 Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-04-09resolv: tentatively fix usage of uninitialized DNS parametersDenys Vlasenko
See "Possible regression from timeout commit for resolv.conf" thread. Also remove superfluous NULL check. Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-08.gitignore getconf binariesBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-08confstr: properly stringify version partsBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-06getconf: print arbitrary GNU_LIBPTHREAD_VERSION stringBernhard Reutner-Fischer
the NPTL should probably print the version-string of the glibc it's compatible with and not the uClibc version (check what mysql expects). Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-06getconf: move to utils/Bernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-06Fix use-after-free bug in __dns_lookupGabor Juhos
If the type of the first answer does not match with the requested type, then the dotted name was freed. If there are no further answers in the DNS reply, this pointer was used later on in the same function. Additionally it is passed to the caller, and caused strange behaviour. Signed-off-by: Gabor Juhos <juhosg@openwrt.org> Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-06i386/bits/syscalls.h: more compact, but stack-hungry syscall code. disabled ↵Denys Vlasenko
so far Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-06i386/bits/syscalls.h: explain _BITS_SYSCALLS_ASMDenys Vlasenko
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-06i386/bits/syscalls.h: generate better code using more restrictive asm ↵Denys Vlasenko
constraints Example: time() doesn't push/pop two registers now, and is smaller: <__GI_time>: -57 push %edi -8b 7c 24 08 mov 0x8(%esp),%edi -53 push %ebx -89 fb mov %edi,%ebx +8b 4c 24 04 mov 0x4(%esp),%ecx +87 cb xchg %ecx,%ebx b8 0d 00 00 00 mov $0xd,%eax cd 80 int $0x80 -5b pop %ebx -5f pop %edi +87 cb xchg %ecx,%ebx c3 ret Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-06i386/bits/syscalls.h: explain how it works. no code changesDenys Vlasenko
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-05libutil/login: was totally broken. fixedDenys Vlasenko
Also made login() and logout() usable with utmpname(non_std_utmp_file) text data bss dec hex filename 111 0 0 111 6f libutil/login.o 217 0 0 217 d9 libutil/login.o text data bss dec hex filename 164 0 0 164 a4 libutil/logout.o 134 0 0 134 86 libutil/logout.o Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-05utent: do not create extra static functions if !THREADSDenys Vlasenko
text data bss dec hex filename 547 8 384 939 3ab libc/misc/utmp/utent.o 519 8 384 911 38f libc/misc/utmp/utent.o Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-05regex_old: convert static flag variable to smallintDenys Vlasenko
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-05getutid is not used internally, removing hidden_protoDenys Vlasenko
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2010-04-02bump version to 0.9.32-gitBernhard Reutner-Fischer
NPTL merge candidate, finally! Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-04-02Release 0.9.31Bernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-30resolv: DEBUG-print nameserver we talk toBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-30utils: hide comment if !VBernhard Reutner-Fischer
2010-03-25pass CFLAGS-dir to CC-mBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-25prettify make cleanBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-18nios2: Define INTERNAL_SYSCALL_NCS, use common sycall definesTobias Klauser
Define the nios2 version of INTERNAL_SYSCALL_NCS and remove all sycall defines already defined by the common syscall defines. This will make the master branch of uClibc compile again for nios2. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Austin Foxley <austinf@cetoncorp.com>
2010-03-18nios2: Update fcntl.h from m68kTobias Klauser
Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Austin Foxley <austinf@cetoncorp.com>
2010-03-17lift printf field width limitMichael Deutschmann
uClibc mishandles printf field width limits larger than 40959, as a result of misguided overflow-protection code. This causes spurious test failures with GNU coreutils, which depends on "%65536s" and "%20000000f" working according to spec. Signed-off-by: Michael Deutschmann <michael@talamasca.ocis.net> Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-16nios2: Add sys/user.hTobias Klauser
This is a fixed version containing copyright information. The patch should now apply using 'git am'. linux/user.h is no longer exported during kernel make headers_install Signed-off-by: Atle Nissestad <atle@nissestad.no> Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-13config: tweak textBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12buildsys: touchup 'make dist' a tiny bitBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12bump versionBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12bump version to 0.9.31-rc1Bernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12buildsys: Fix O= PREFIX= caseBernhard Reutner-Fischer
make sure that something like: $ rm -rf /dev/shm/o/* /dev/shm/o/.conf* /dev/shm/DELME $ time (make defconfig O=/dev/shm/o 2>&1 >& /dev/null && make \ install_dev -j O=/dev/shm/o PREFIX=/dev/shm/DELME 2>&1 >& /dev/null) works Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12buildsys: fix dir order-only prereq of unifdefBernhard Reutner-Fischer
|$(@D) -> zilch, so spell it out Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12add MULTILIB_DIR: Path component for libdirsBernhard Reutner-Fischer
forgot to checkin this file Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12config: support make defconfig O=/f/o/oBernhard Reutner-Fischer
previously wanted to create the temporary cfg file in the srcdir (which failed on a RO srcdir) Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12add MULTILIB_DIR: Path component for libdirsBernhard Reutner-Fischer
defaults to "lib". Other prominent values include "lib32" or "lib64" Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12install: fix O= PREFIX= installBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12reduce number of mkdir callsBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-12remove config knobs that belong to the nptl branchBernhard Reutner-Fischer
Fixes bug #1243 Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-05pull kernel-features.h from NPTLBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-03-04poll: unavailable on linux < 2.2.0Bernhard Reutner-Fischer
fixes bug #253 Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-02-10libc_sh: Update memcpy to use the ENTRY macroCarmelo Amoroso
Use the ENTRY macro now available through the sysdep.h header Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-02-10libc_sh: Add a sysdep header for shCarmelo Amoroso
A new sysdep header for sh (ported from NPTL branch) to provide some useful macros in asm code: ENTRY: an entry point visible from C PSEUDO: a wrapper for syscall with proper errno checking SYSCALL_ERROR_HANDLER: errno check (suitable for TLS) This will help in NPTL integration. Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-02-10libc: Add a common sysdep headerCarmelo Amoroso
Add a common header file to provide macros useful in asm code: C_LABEL to construct the asm name for a C symbol cfi_xxx to generate eh_frame unwind information. Ported from NPTL branch. Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-02-08libc: Fix typo in include/rpcCarmelo Amoroso
s/GNU_SOUCE/GNU_SOURCE/ in include/rcp/ Signed-off-by: Jason Woodward <jason.woodward@timesys.com> Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>
2010-02-05getdents: Fix mips64 buildAtsushi Nemoto
On Wed, 27 Jan 2010 07:14:08 +0100, Carmelo AMOROSO <carmelo.amoroso@st.com> wrote: > I would re-write your patch in a simpler way. > > We already have the following > > 136 #if defined __UCLIBC_HAS_LFS__ && ! defined __NR_getdents64 > 137 attribute_hidden strong_alias(__getdents,__getdents64) > 138 #endif > > I think that it's simpler to move in the proper place this statement. Thanks, indeed. If we came into "#elif WORDSIZE == 32" block, above condition never be true. So we can just move this statement out of "#if...#elif...#elif...#endif" block. Here is a revised patch. ------------------------------------------------------ From: Atsushi Nemoto <anemo@mba.ocn.ne.jp> Subject: [PATCH] getdents: Fix mips64 build Some archs (such as mips64) do not have getdents64 syscall but have getdents syscall. Define alias for it. This fixes regression from 0.9.30.1. Backgrounds: This is once done by commit e8b1c674. But after the commit 33bcf733 ("Use getdents syscall if kernel provide supports for this instead of relying upon getdents64."), if __ASSUME_GETDENTS32_D_TYPE was defined the alias for getdents64 is not defined. The macro __ASSUME_GETDENTS32_D_TYPE had been effectively ignored until 0.9.30.1 but the commit 0f0f20ab ("Move kernel-features.h header from the linuxthread directory to a common one...") really enables it. Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp> Signed-off-by: Khem Raj <raj.khem@gmail.com>
2010-02-05prctl: silence shadow warningsBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-02-05ldso: add missing prototypesBernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-02-05Revert "ftw.h installation: fix inverted logic"Bernhard Reutner-Fischer
The logic was fine previously. Thanks to Peter Kjellerstedt for pointing out my confusion. This reverts commit d75f980f379dfbbe70ae8fb0956693a95af2fca2. Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>
2010-02-04libm: explain why some funcs are not implemented (yet)Bernhard Reutner-Fischer
Signed-off-by: Bernhard Reutner-Fischer <rep.dot.nop@gmail.com>