diff options
Diffstat (limited to 'target/linux/patches/3.14.40')
26 files changed, 41678 insertions, 0 deletions
diff --git a/target/linux/patches/3.14.40/bsd-compatibility.patch b/target/linux/patches/3.14.40/bsd-compatibility.patch new file mode 100644 index 000000000..b954b658f --- /dev/null +++ b/target/linux/patches/3.14.40/bsd-compatibility.patch @@ -0,0 +1,2538 @@ +diff -Nur linux-3.11.5.orig/scripts/Makefile.lib linux-3.11.5/scripts/Makefile.lib +--- linux-3.11.5.orig/scripts/Makefile.lib 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/scripts/Makefile.lib 2013-10-16 18:09:31.000000000 +0200 +@@ -281,7 +281,12 @@ + size_append = printf $(shell \ + dec_size=0; \ + for F in $1; do \ +- fsize=$$(stat -c "%s" $$F); \ ++ if stat -qs .>/dev/null 2>&1; then \ ++ statcmd='stat -f %z'; \ ++ else \ ++ statcmd='stat -c %s'; \ ++ fi; \ ++ fsize=$$($$statcmd $$F); \ + dec_size=$$(expr $$dec_size + $$fsize); \ + done; \ + printf "%08x\n" $$dec_size | \ +diff -Nur linux-3.11.5.orig/scripts/mod/mk_elfconfig.c linux-3.11.5/scripts/mod/mk_elfconfig.c +--- linux-3.11.5.orig/scripts/mod/mk_elfconfig.c 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/scripts/mod/mk_elfconfig.c 2013-10-16 18:09:31.000000000 +0200 +@@ -1,7 +1,18 @@ + #include <stdio.h> + #include <stdlib.h> + #include <string.h> +-#include <elf.h> ++ ++#define EI_NIDENT (16) ++#define ELFMAG "\177ELF" ++ ++#define SELFMAG 4 ++#define EI_CLASS 4 ++#define ELFCLASS32 1 /* 32-bit objects */ ++#define ELFCLASS64 2 /* 64-bit objects */ ++ ++#define EI_DATA 5 /* Data encoding byte index */ ++#define ELFDATA2LSB 1 /* 2's complement, little endian */ ++#define ELFDATA2MSB 2 /* 2's complement, big endian */ + + int + main(int argc, char **argv) +diff -Nur linux-3.11.5.orig/scripts/mod/modpost.h linux-3.11.5/scripts/mod/modpost.h +--- linux-3.11.5.orig/scripts/mod/modpost.h 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/scripts/mod/modpost.h 2013-10-16 18:09:31.000000000 +0200 +@@ -7,7 +7,2453 @@ + #include <sys/mman.h> + #include <fcntl.h> + #include <unistd.h> +-#include <elf.h> ++ ++ ++/* This file defines standard ELF types, structures, and macros. ++ Copyright (C) 1995-1999,2000,2001,2002,2003 Free Software Foundation, Inc. ++ This file is part of the GNU C Library. ++ ++ The GNU C Library is free software; you can redistribute it and/or ++ modify it under the terms of the GNU Lesser General Public ++ License as published by the Free Software Foundation; either ++ version 2.1 of the License, or (at your option) any later version. ++ ++ The GNU C Library 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 ++ Lesser General Public License for more details. ++ ++ You should have received a copy of the GNU Lesser General Public ++ License along with the GNU C Library; if not, write to the Free ++ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA ++ 02111-1307 USA. */ ++ ++#ifndef _ELF_H ++#define _ELF_H 1 ++ ++__BEGIN_DECLS ++ ++/* Standard ELF types. */ ++ ++#include <stdint.h> ++ ++/* Type for a 16-bit quantity. */ ++typedef uint16_t Elf32_Half; ++typedef uint16_t Elf64_Half; ++ ++/* Types for signed and unsigned 32-bit quantities. */ ++typedef uint32_t Elf32_Word; ++typedef int32_t Elf32_Sword; ++typedef uint32_t Elf64_Word; ++typedef int32_t Elf64_Sword; ++ ++/* Types for signed and unsigned 64-bit quantities. */ ++typedef uint64_t Elf32_Xword; ++typedef int64_t Elf32_Sxword; ++typedef uint64_t Elf64_Xword; ++typedef int64_t Elf64_Sxword; ++ ++/* Type of addresses. */ ++typedef uint32_t Elf32_Addr; ++typedef uint64_t Elf64_Addr; ++ ++/* Type of file offsets. */ ++typedef uint32_t Elf32_Off; ++typedef uint64_t Elf64_Off; ++ ++/* Type for section indices, which are 16-bit quantities. */ ++typedef uint16_t Elf32_Section; ++typedef uint16_t Elf64_Section; ++ ++/* Type for version symbol information. */ ++typedef Elf32_Half Elf32_Versym; ++typedef Elf64_Half Elf64_Versym; ++ ++ ++/* The ELF file header. This appears at the start of every ELF file. */ ++ ++#define EI_NIDENT (16) ++ ++typedef struct ++{ ++ unsigned char e_ident[EI_NIDENT]; /* Magic number and other info */ ++ Elf32_Half e_type; /* Object file type */ ++ Elf32_Half e_machine; /* Architecture */ ++ Elf32_Word e_version; /* Object file version */ ++ Elf32_Addr e_entry; /* Entry point virtual address */ ++ Elf32_Off e_phoff; /* Program header table file offset */ ++ Elf32_Off e_shoff; /* Section header table file offset */ ++ Elf32_Word e_flags; /* Processor-specific flags */ ++ Elf32_Half e_ehsize; /* ELF header size in bytes */ ++ Elf32_Half e_phentsize; /* Program header table entry size */ ++ Elf32_Half e_phnum; /* Program header table entry count */ ++ Elf32_Half e_shentsize; /* Section header table entry size */ ++ Elf32_Half e_shnum; /* Section header table entry count */ ++ Elf32_Half e_shstrndx; /* Section header string table index */ ++} Elf32_Ehdr; ++ ++typedef struct ++{ ++ unsigned char e_ident[EI_NIDENT]; /* Magic number and other info */ ++ Elf64_Half e_type; /* Object file type */ ++ Elf64_Half e_machine; /* Architecture */ ++ Elf64_Word e_version; /* Object file version */ ++ Elf64_Addr e_entry; /* Entry point virtual address */ ++ Elf64_Off e_phoff; /* Program header table file offset */ ++ Elf64_Off e_shoff; /* Section header table file offset */ ++ Elf64_Word e_flags; /* Processor-specific flags */ ++ Elf64_Half e_ehsize; /* ELF header size in bytes */ ++ Elf64_Half e_phentsize; /* Program header table entry size */ ++ Elf64_Half e_phnum; /* Program header table entry count */ ++ Elf64_Half e_shentsize; /* Section header table entry size */ ++ Elf64_Half e_shnum; /* Section header table entry count */ ++ Elf64_Half e_shstrndx; /* Section header string table index */ ++} Elf64_Ehdr; ++ ++/* Fields in the e_ident array. The EI_* macros are indices into the ++ array. The macros under each EI_* macro are the values the byte ++ may have. */ ++ ++#define EI_MAG0 0 /* File identification byte 0 index */ ++#define ELFMAG0 0x7f /* Magic number byte 0 */ ++ ++#define EI_MAG1 1 /* File identification byte 1 index */ ++#define ELFMAG1 'E' /* Magic number byte 1 */ ++ ++#define EI_MAG2 2 /* File identification byte 2 index */ ++#define ELFMAG2 'L' /* Magic number byte 2 */ ++ ++#define EI_MAG3 3 /* File identification byte 3 index */ ++#define ELFMAG3 'F' /* Magic number byte 3 */ ++ ++/* Conglomeration of the identification bytes, for easy testing as a word. */ ++#define ELFMAG "\177ELF" ++#define SELFMAG 4 ++ ++#define EI_CLASS 4 /* File class byte index */ ++#define ELFCLASSNONE 0 /* Invalid class */ ++#define ELFCLASS32 1 /* 32-bit objects */ ++#define ELFCLASS64 2 /* 64-bit objects */ ++#define ELFCLASSNUM 3 ++ ++#define EI_DATA 5 /* Data encoding byte index */ ++#define ELFDATANONE 0 /* Invalid data encoding */ ++#define ELFDATA2LSB 1 /* 2's complement, little endian */ ++#define ELFDATA2MSB 2 /* 2's complement, big endian */ ++#define ELFDATANUM 3 ++ ++#define EI_VERSION 6 /* File version byte index */ ++ /* Value must be EV_CURRENT */ ++ ++#define EI_OSABI 7 /* OS ABI identification */ ++#define ELFOSABI_NONE 0 /* UNIX System V ABI */ ++#define ELFOSABI_SYSV 0 /* Alias. */ ++#define ELFOSABI_HPUX 1 /* HP-UX */ ++#define ELFOSABI_NETBSD 2 /* NetBSD. */ ++#define ELFOSABI_LINUX 3 /* Linux. */ ++#define ELFOSABI_SOLARIS 6 /* Sun Solaris. */ ++#define ELFOSABI_AIX 7 /* IBM AIX. */ ++#define ELFOSABI_IRIX 8 /* SGI Irix. */ ++#define ELFOSABI_FREEBSD 9 /* FreeBSD. */ ++#define ELFOSABI_TRU64 10 /* Compaq TRU64 UNIX. */ ++#define ELFOSABI_MODESTO 11 /* Novell Modesto. */ ++#define ELFOSABI_OPENBSD 12 /* OpenBSD. */ ++#define ELFOSABI_ARM 97 /* ARM */ ++#define ELFOSABI_STANDALONE 255 /* Standalone (embedded) application */ ++ ++#define EI_ABIVERSION 8 /* ABI version */ ++ ++#define EI_PAD 9 /* Byte index of padding bytes */ ++ ++/* Legal values for e_type (object file type). */ ++ ++#define ET_NONE 0 /* No file type */ ++#define ET_REL 1 /* Relocatable file */ ++#define ET_EXEC 2 /* Executable file */ ++#define ET_DYN 3 /* Shared object file */ ++#define ET_CORE 4 /* Core file */ ++#define ET_NUM 5 /* Number of defined types */ ++#define ET_LOOS 0xfe00 /* OS-specific range start */ ++#define ET_HIOS 0xfeff /* OS-specific range end */ ++#define ET_LOPROC 0xff00 /* Processor-specific range start */ ++#define ET_HIPROC 0xffff /* Processor-specific range end */ ++ ++/* Legal values for e_machine (architecture). */ ++ ++#define EM_NONE 0 /* No machine */ ++#define EM_M32 1 /* AT&T WE 32100 */ ++#define EM_SPARC 2 /* SUN SPARC */ ++#define EM_386 3 /* Intel 80386 */ ++#define EM_68K 4 /* Motorola m68k family */ ++#define EM_88K 5 /* Motorola m88k family */ ++#define EM_860 7 /* Intel 80860 */ ++#define EM_MIPS 8 /* MIPS R3000 big-endian */ ++#define EM_S370 9 /* IBM System/370 */ ++#define EM_MIPS_RS3_LE 10 /* MIPS R3000 little-endian */ ++ ++#define EM_PARISC 15 /* HPPA */ ++#define EM_VPP500 17 /* Fujitsu VPP500 */ ++#define EM_SPARC32PLUS 18 /* Sun's "v8plus" */ ++#define EM_960 19 /* Intel 80960 */ ++#define EM_PPC 20 /* PowerPC */ ++#define EM_PPC64 21 /* PowerPC 64-bit */ ++#define EM_S390 22 /* IBM S390 */ ++ ++#define EM_V800 36 /* NEC V800 series */ ++#define EM_FR20 37 /* Fujitsu FR20 */ ++#define EM_RH32 38 /* TRW RH-32 */ ++#define EM_RCE 39 /* Motorola RCE */ ++#define EM_ARM 40 /* ARM */ ++#define EM_FAKE_ALPHA 41 /* Digital Alpha */ ++#define EM_SH 42 /* Hitachi SH */ ++#define EM_SPARCV9 43 /* SPARC v9 64-bit */ ++#define EM_TRICORE 44 /* Siemens Tricore */ ++#define EM_ARC 45 /* Argonaut RISC Core */ ++#define EM_H8_300 46 /* Hitachi H8/300 */ ++#define EM_H8_300H 47 /* Hitachi H8/300H */ ++#define EM_H8S 48 /* Hitachi H8S */ ++#define EM_H8_500 49 /* Hitachi H8/500 */ ++#define EM_IA_64 50 /* Intel Merced */ ++#define EM_MIPS_X 51 /* Stanford MIPS-X */ ++#define EM_COLDFIRE 52 /* Motorola Coldfire */ ++#define EM_68HC12 53 /* Motorola M68HC12 */ ++#define EM_MMA 54 /* Fujitsu MMA Multimedia Accelerator*/ ++#define EM_PCP 55 /* Siemens PCP */ ++#define EM_NCPU 56 /* Sony nCPU embeeded RISC */ ++#define EM_NDR1 57 /* Denso NDR1 microprocessor */ ++#define EM_STARCORE 58 /* Motorola Start*Core processor */ ++#define EM_ME16 59 /* Toyota ME16 processor */ ++#define EM_ST100 60 /* STMicroelectronic ST100 processor */ ++#define EM_TINYJ 61 /* Advanced Logic Corp. Tinyj emb.fam*/ ++#define EM_X86_64 62 /* AMD x86-64 architecture */ ++#define EM_PDSP 63 /* Sony DSP Processor */ ++ ++#define EM_FX66 66 /* Siemens FX66 microcontroller */ ++#define EM_ST9PLUS 67 /* STMicroelectronics ST9+ 8/16 mc */ ++#define EM_ST7 68 /* STmicroelectronics ST7 8 bit mc */ ++#define EM_68HC16 69 /* Motorola MC68HC16 microcontroller */ ++#define EM_68HC11 70 /* Motorola MC68HC11 microcontroller */ ++#define EM_68HC08 71 /* Motorola MC68HC08 microcontroller */ ++#define EM_68HC05 72 /* Motorola MC68HC05 microcontroller */ ++#define EM_SVX 73 /* Silicon Graphics SVx */ ++#define EM_ST19 74 /* STMicroelectronics ST19 8 bit mc */ ++#define EM_VAX 75 /* Digital VAX */ ++#define EM_CRIS 76 /* Axis Communications 32-bit embedded processor */ ++#define EM_JAVELIN 77 /* Infineon Technologies 32-bit embedded processor */ ++#define EM_FIREPATH 78 /* Element 14 64-bit DSP Processor */ ++#define EM_ZSP 79 /* LSI Logic 16-bit DSP Processor */ ++#define EM_MMIX 80 /* Donald Knuth's educational 64-bit processor */ ++#define EM_HUANY 81 /* Harvard University machine-independent object files */ ++#define EM_PRISM 82 /* SiTera Prism */ ++#define EM_AVR 83 /* Atmel AVR 8-bit microcontroller */ ++#define EM_FR30 84 /* Fujitsu FR30 */ ++#define EM_D10V 85 /* Mitsubishi D10V */ ++#define EM_D30V 86 /* Mitsubishi D30V */ ++#define EM_V850 87 /* NEC v850 */ ++#define EM_M32R 88 /* Mitsubishi M32R */ ++#define EM_MN10300 89 /* Matsushita MN10300 */ ++#define EM_MN10200 90 /* Matsushita MN10200 */ ++#define EM_PJ 91 /* picoJava */ ++#define EM_OPENRISC 92 /* OpenRISC 32-bit embedded processor */ ++#define EM_ARC_A5 93 /* ARC Cores Tangent-A5 */ ++#define EM_XTENSA 94 /* Tensilica Xtensa Architecture */ ++#define EM_NUM 95 ++ ++/* If it is necessary to assign new unofficial EM_* values, please ++ pick large random numbers (0x8523, 0xa7f2, etc.) to minimize the ++ chances of collision with official or non-GNU unofficial values. */ ++ ++#define EM_ALPHA 0x9026 ++ ++/* Legal values for e_version (version). */ ++ ++#define EV_NONE 0 /* Invalid ELF version */ ++#define EV_CURRENT 1 /* Current version */ ++#define EV_NUM 2 ++ ++/* Section header. */ ++ ++typedef struct ++{ ++ Elf32_Word sh_name; /* Section name (string tbl index) */ ++ Elf32_Word sh_type; /* Section type */ ++ Elf32_Word sh_flags; /* Section flags */ ++ Elf32_Addr sh_addr; /* Section virtual addr at execution */ ++ Elf32_Off sh_offset; /* Section file offset */ ++ Elf32_Word sh_size; /* Section size in bytes */ ++ Elf32_Word sh_link; /* Link to another section */ ++ Elf32_Word sh_info; /* Additional section information */ ++ Elf32_Word sh_addralign; /* Section alignment */ ++ Elf32_Word sh_entsize; /* Entry size if section holds table */ ++} Elf32_Shdr; ++ ++typedef struct ++{ ++ Elf64_Word sh_name; /* Section name (string tbl index) */ ++ Elf64_Word sh_type; /* Section type */ ++ Elf64_Xword sh_flags; /* Section flags */ ++ Elf64_Addr sh_addr; /* Section virtual addr at execution */ ++ Elf64_Off sh_offset; /* Section file offset */ ++ Elf64_Xword sh_size; /* Section size in bytes */ ++ Elf64_Word sh_link; /* Link to another section */ ++ Elf64_Word sh_info; /* Additional section information */ ++ Elf64_Xword sh_addralign; /* Section alignment */ ++ Elf64_Xword sh_entsize; /* Entry size if section holds table */ ++} Elf64_Shdr; ++ ++/* Special section indices. */ ++ ++#define SHN_UNDEF 0 /* Undefined section */ ++#define SHN_LORESERVE 0xff00 /* Start of reserved indices */ ++#define SHN_LOPROC 0xff00 /* Start of processor-specific */ ++#define SHN_HIPROC 0xff1f /* End of processor-specific */ ++#define SHN_LOOS 0xff20 /* Start of OS-specific */ ++#define SHN_HIOS 0xff3f /* End of OS-specific */ ++#define SHN_ABS 0xfff1 /* Associated symbol is absolute */ ++#define SHN_COMMON 0xfff2 /* Associated symbol is common */ ++#define SHN_XINDEX 0xffff /* Index is in extra table. */ ++#define SHN_HIRESERVE 0xffff /* End of reserved indices */ ++ ++/* Legal values for sh_type (section type). */ ++ ++#define SHT_NULL 0 /* Section header table entry unused */ ++#define SHT_PROGBITS 1 /* Program data */ ++#define SHT_SYMTAB 2 /* Symbol table */ ++#define SHT_STRTAB 3 /* String table */ ++#define SHT_RELA 4 /* Relocation entries with addends */ ++#define SHT_HASH 5 /* Symbol hash table */ ++#define SHT_DYNAMIC 6 /* Dynamic linking information */ ++#define SHT_NOTE 7 /* Notes */ ++#define SHT_NOBITS 8 /* Program space with no data (bss) */ ++#define SHT_REL 9 /* Relocation entries, no addends */ ++#define SHT_SHLIB 10 /* Reserved */ ++#define SHT_DYNSYM 11 /* Dynamic linker symbol table */ ++#define SHT_INIT_ARRAY 14 /* Array of constructors */ ++#define SHT_FINI_ARRAY 15 /* Array of destructors */ ++#define SHT_PREINIT_ARRAY 16 /* Array of pre-constructors */ ++#define SHT_GROUP 17 /* Section group */ ++#define SHT_SYMTAB_SHNDX 18 /* Extended section indeces */ ++#define SHT_NUM 19 /* Number of defined types. */ ++#define SHT_LOOS 0x60000000 /* Start OS-specific */ ++#define SHT_GNU_LIBLIST 0x6ffffff7 /* Prelink library list */ ++#define SHT_CHECKSUM 0x6ffffff8 /* Checksum for DSO content. */ ++#define SHT_LOSUNW 0x6ffffffa /* Sun-specific low bound. */ ++#define SHT_SUNW_move 0x6ffffffa ++#define SHT_SUNW_COMDAT 0x6ffffffb ++#define SHT_SUNW_syminfo 0x6ffffffc ++#define SHT_GNU_verdef 0x6ffffffd /* Version definition section. */ ++#define SHT_GNU_verneed 0x6ffffffe /* Version needs section. */ ++#define SHT_GNU_versym 0x6fffffff /* Version symbol table. */ ++#define SHT_HISUNW 0x6fffffff /* Sun-specific high bound. */ ++#define SHT_HIOS 0x6fffffff /* End OS-specific type */ ++#define SHT_LOPROC 0x70000000 /* Start of processor-specific */ ++#define SHT_HIPROC 0x7fffffff /* End of processor-specific */ ++#define SHT_LOUSER 0x80000000 /* Start of application-specific */ ++#define SHT_HIUSER 0x8fffffff /* End of application-specific */ ++ ++/* Legal values for sh_flags (section flags). */ ++ ++#define SHF_WRITE (1 << 0) /* Writable */ ++#define SHF_ALLOC (1 << 1) /* Occupies memory during execution */ ++#define SHF_EXECINSTR (1 << 2) /* Executable */ ++#define SHF_MERGE (1 << 4) /* Might be merged */ ++#define SHF_STRINGS (1 << 5) /* Contains nul-terminated strings */ ++#define SHF_INFO_LINK (1 << 6) /* `sh_info' contains SHT index */ ++#define SHF_LINK_ORDER (1 << 7) /* Preserve order after combining */ ++#define SHF_OS_NONCONFORMING (1 << 8) /* Non-standard OS specific handling ++ required */ ++#define SHF_GROUP (1 << 9) /* Section is member of a group. */ ++#define SHF_TLS (1 << 10) /* Section hold thread-local data. */ ++#define SHF_MASKOS 0x0ff00000 /* OS-specific. */ ++#define SHF_MASKPROC 0xf0000000 /* Processor-specific */ ++ ++/* Section group handling. */ ++#define GRP_COMDAT 0x1 /* Mark group as COMDAT. */ ++ ++/* Symbol table entry. */ ++ ++typedef struct ++{ ++ Elf32_Word st_name; /* Symbol name (string tbl index) */ ++ Elf32_Addr st_value; /* Symbol value */ ++ Elf32_Word st_size; /* Symbol size */ ++ unsigned char st_info; /* Symbol type and binding */ ++ unsigned char st_other; /* Symbol visibility */ ++ Elf32_Section st_shndx; /* Section index */ ++} Elf32_Sym; ++ ++typedef struct ++{ ++ Elf64_Word st_name; /* Symbol name (string tbl index) */ ++ unsigned char st_info; /* Symbol type and binding */ ++ unsigned char st_other; /* Symbol visibility */ ++ Elf64_Section st_shndx; /* Section index */ ++ Elf64_Addr st_value; /* Symbol value */ ++ Elf64_Xword st_size; /* Symbol size */ ++} Elf64_Sym; ++ ++/* The syminfo section if available contains additional information about ++ every dynamic symbol. */ ++ ++typedef struct ++{ ++ Elf32_Half si_boundto; /* Direct bindings, symbol bound to */ ++ Elf32_Half si_flags; /* Per symbol flags */ ++} Elf32_Syminfo; ++ ++typedef struct ++{ ++ Elf64_Half si_boundto; /* Direct bindings, symbol bound to */ ++ Elf64_Half si_flags; /* Per symbol flags */ ++} Elf64_Syminfo; ++ ++/* Possible values for si_boundto. */ ++#define SYMINFO_BT_SELF 0xffff /* Symbol bound to self */ ++#define SYMINFO_BT_PARENT 0xfffe /* Symbol bound to parent */ ++#define SYMINFO_BT_LOWRESERVE 0xff00 /* Beginning of reserved entries */ ++ ++/* Possible bitmasks for si_flags. */ ++#define SYMINFO_FLG_DIRECT 0x0001 /* Direct bound symbol */ ++#define SYMINFO_FLG_PASSTHRU 0x0002 /* Pass-thru symbol for translator */ ++#define SYMINFO_FLG_COPY 0x0004 /* Symbol is a copy-reloc */ ++#define SYMINFO_FLG_LAZYLOAD 0x0008 /* Symbol bound to object to be lazy ++ loaded */ ++/* Syminfo version values. */ ++#define SYMINFO_NONE 0 ++#define SYMINFO_CURRENT 1 ++#define SYMINFO_NUM 2 ++ ++ ++/* How to extract and insert information held in the st_info field. */ ++ ++#define ELF32_ST_BIND(val) (((unsigned char) (val)) >> 4) ++#define ELF32_ST_TYPE(val) ((val) & 0xf) ++#define ELF32_ST_INFO(bind, type) (((bind) << 4) + ((type) & 0xf)) ++ ++/* Both Elf32_Sym and Elf64_Sym use the same one-byte st_info field. */ ++#define ELF64_ST_BIND(val) ELF32_ST_BIND (val) ++#define ELF64_ST_TYPE(val) ELF32_ST_TYPE (val) ++#define ELF64_ST_INFO(bind, type) ELF32_ST_INFO ((bind), (type)) ++ ++/* Legal values for ST_BIND subfield of st_info (symbol binding). */ ++ ++#define STB_LOCAL 0 /* Local symbol */ ++#define STB_GLOBAL 1 /* Global symbol */ ++#define STB_WEAK 2 /* Weak symbol */ ++#define STB_NUM 3 /* Number of defined types. */ ++#define STB_LOOS 10 /* Start of OS-specific */ ++#define STB_HIOS 12 /* End of OS-specific */ ++#define STB_LOPROC 13 /* Start of processor-specific */ ++#define STB_HIPROC 15 /* End of processor-specific */ ++ ++/* Legal values for ST_TYPE subfield of st_info (symbol type). */ ++ ++#define STT_NOTYPE 0 /* Symbol type is unspecified */ ++#define STT_OBJECT 1 /* Symbol is a data object */ ++#define STT_FUNC 2 /* Symbol is a code object */ ++#define STT_SECTION 3 /* Symbol associated with a section */ ++#define STT_FILE 4 /* Symbol's name is file name */ ++#define STT_COMMON 5 /* Symbol is a common data object */ ++#define STT_TLS 6 /* Symbol is thread-local data object*/ ++#define STT_NUM 7 /* Number of defined types. */ ++#define STT_LOOS 10 /* Start of OS-specific */ ++#define STT_HIOS 12 /* End of OS-specific */ ++#define STT_LOPROC 13 /* Start of processor-specific */ ++#define STT_HIPROC 15 /* End of processor-specific */ ++ ++ ++/* Symbol table indices are found in the hash buckets and chain table ++ of a symbol hash table section. This special index value indicates ++ the end of a chain, meaning no further symbols are found in that bucket. */ ++ ++#define STN_UNDEF 0 /* End of a chain. */ ++ ++ ++/* How to extract and insert information held in the st_other field. */ ++ ++#define ELF32_ST_VISIBILITY(o) ((o) & 0x03) ++ ++/* For ELF64 the definitions are the same. */ ++#define ELF64_ST_VISIBILITY(o) ELF32_ST_VISIBILITY (o) ++ ++/* Symbol visibility specification encoded in the st_other field. */ ++#define STV_DEFAULT 0 /* Default symbol visibility rules */ ++#define STV_INTERNAL 1 /* Processor specific hidden class */ ++#define STV_HIDDEN 2 /* Sym unavailable in other modules */ ++#define STV_PROTECTED 3 /* Not preemptible, not exported */ ++ ++ ++/* Relocation table entry without addend (in section of type SHT_REL). */ ++ ++typedef struct ++{ ++ Elf32_Addr r_offset; /* Address */ ++ Elf32_Word r_info; /* Relocation type and symbol index */ ++} Elf32_Rel; ++ ++/* I have seen two different definitions of the Elf64_Rel and ++ Elf64_Rela structures, so we'll leave them out until Novell (or ++ whoever) gets their act together. */ ++/* The following, at least, is used on Sparc v9, MIPS, and Alpha. */ ++ ++typedef struct ++{ ++ Elf64_Addr r_offset; /* Address */ ++ Elf64_Xword r_info; /* Relocation type and symbol index */ ++} Elf64_Rel; ++ ++/* Relocation table entry with addend (in section of type SHT_RELA). */ ++ ++typedef struct ++{ ++ Elf32_Addr r_offset; /* Address */ ++ Elf32_Word r_info; /* Relocation type and symbol index */ ++ Elf32_Sword r_addend; /* Addend */ ++} Elf32_Rela; ++ ++typedef struct ++{ ++ Elf64_Addr r_offset; /* Address */ ++ Elf64_Xword r_info; /* Relocation type and symbol index */ ++ Elf64_Sxword r_addend; /* Addend */ ++} Elf64_Rela; ++ ++/* How to extract and insert information held in the r_info field. */ ++ ++#define ELF32_R_SYM(val) ((val) >> 8) ++#define ELF32_R_TYPE(val) ((val) & 0xff) ++#define ELF32_R_INFO(sym, type) (((sym) << 8) + ((type) & 0xff)) ++ ++#define ELF64_R_SYM(i) ((i) >> 32) ++#define ELF64_R_TYPE(i) ((i) & 0xffffffff) ++#define ELF64_R_INFO(sym,type) ((((Elf64_Xword) (sym)) << 32) + (type)) ++ ++/* Program segment header. */ ++ ++typedef struct ++{ ++ Elf32_Word p_type; /* Segment type */ ++ Elf32_Off p_offset; /* Segment file offset */ ++ Elf32_Addr p_vaddr; /* Segment virtual address */ ++ Elf32_Addr p_paddr; /* Segment physical address */ ++ Elf32_Word p_filesz; /* Segment size in file */ ++ Elf32_Word p_memsz; /* Segment size in memory */ ++ Elf32_Word p_flags; /* Segment flags */ ++ Elf32_Word p_align; /* Segment alignment */ ++} Elf32_Phdr; ++ ++typedef struct ++{ ++ Elf64_Word p_type; /* Segment type */ ++ Elf64_Word p_flags; /* Segment flags */ ++ Elf64_Off p_offset; /* Segment file offset */ ++ Elf64_Addr p_vaddr; /* Segment virtual address */ ++ Elf64_Addr p_paddr; /* Segment physical address */ ++ Elf64_Xword p_filesz; /* Segment size in file */ ++ Elf64_Xword p_memsz; /* Segment size in memory */ ++ Elf64_Xword p_align; /* Segment alignment */ ++} Elf64_Phdr; ++ ++/* Legal values for p_type (segment type). */ ++ ++#define PT_NULL 0 /* Program header table entry unused */ ++#define PT_LOAD 1 /* Loadable program segment */ ++#define PT_DYNAMIC 2 /* Dynamic linking information */ ++#define PT_INTERP 3 /* Program interpreter */ ++#define PT_NOTE 4 /* Auxiliary information */ ++#define PT_SHLIB 5 /* Reserved */ ++#define PT_PHDR 6 /* Entry for header table itself */ ++#define PT_TLS 7 /* Thread-local storage segment */ ++#define PT_NUM 8 /* Number of defined types */ ++#define PT_LOOS 0x60000000 /* Start of OS-specific */ ++#define PT_GNU_EH_FRAME 0x6474e550 /* GCC .eh_frame_hdr segment */ ++#define PT_GNU_STACK 0x6474e551 /* Indicates stack executability */ ++#define PT_LOSUNW 0x6ffffffa ++#define PT_SUNWBSS 0x6ffffffa /* Sun Specific segment */ ++#define PT_SUNWSTACK 0x6ffffffb /* Stack segment */ ++#define PT_HISUNW 0x6fffffff ++#define PT_HIOS 0x6fffffff /* End of OS-specific */ ++#define PT_LOPROC 0x70000000 /* Start of processor-specific */ ++#define PT_HIPROC 0x7fffffff /* End of processor-specific */ ++ ++/* Legal values for p_flags (segment flags). */ ++ ++#define PF_X (1 << 0) /* Segment is executable */ ++#define PF_W (1 << 1) /* Segment is writable */ ++#define PF_R (1 << 2) /* Segment is readable */ ++#define PF_MASKOS 0x0ff00000 /* OS-specific */ ++#define PF_MASKPROC 0xf0000000 /* Processor-specific */ ++ ++/* Legal values for note segment descriptor types for core files. */ ++ ++#define NT_PRSTATUS 1 /* Contains copy of prstatus struct */ ++#define NT_FPREGSET 2 /* Contains copy of fpregset struct */ ++#define NT_PRPSINFO 3 /* Contains copy of prpsinfo struct */ ++#define NT_PRXREG 4 /* Contains copy of prxregset struct */ ++#define NT_TASKSTRUCT 4 /* Contains copy of task structure */ ++#define NT_PLATFORM 5 /* String from sysinfo(SI_PLATFORM) */ ++#define NT_AUXV 6 /* Contains copy of auxv array */ ++#define NT_GWINDOWS 7 /* Contains copy of gwindows struct */ ++#define NT_ASRS 8 /* Contains copy of asrset struct */ ++#define NT_PSTATUS 10 /* Contains copy of pstatus struct */ ++#define NT_PSINFO 13 /* Contains copy of psinfo struct */ ++#define NT_PRCRED 14 /* Contains copy of prcred struct */ ++#define NT_UTSNAME 15 /* Contains copy of utsname struct */ ++#define NT_LWPSTATUS 16 /* Contains copy of lwpstatus struct */ ++#define NT_LWPSINFO 17 /* Contains copy of lwpinfo struct */ ++#define NT_PRFPXREG 20 /* Contains copy of fprxregset struct*/ ++ ++/* Legal values for the note segment descriptor types for object files. */ ++ ++#define NT_VERSION 1 /* Contains a version string. */ ++ ++ ++/* Dynamic section entry. */ ++ ++typedef struct ++{ ++ Elf32_Sword d_tag; /* Dynamic entry type */ ++ union ++ { ++ Elf32_Word d_val; /* Integer value */ ++ Elf32_Addr d_ptr; /* Address value */ ++ } d_un; ++} Elf32_Dyn; ++ ++typedef struct ++{ ++ Elf64_Sxword d_tag; /* Dynamic entry type */ ++ union ++ { ++ Elf64_Xword d_val; /* Integer value */ ++ Elf64_Addr d_ptr; /* Address value */ ++ } d_un; ++} Elf64_Dyn; ++ ++/* Legal values for d_tag (dynamic entry type). */ ++ ++#define DT_NULL 0 /* Marks end of dynamic section */ ++#define DT_NEEDED 1 /* Name of needed library */ ++#define DT_PLTRELSZ 2 /* Size in bytes of PLT relocs */ ++#define DT_PLTGOT 3 /* Processor defined value */ ++#define DT_HASH 4 /* Address of symbol hash table */ ++#define DT_STRTAB 5 /* Address of string table */ ++#define DT_SYMTAB 6 /* Address of symbol table */ ++#define DT_RELA 7 /* Address of Rela relocs */ ++#define DT_RELASZ 8 /* Total size of Rela relocs */ ++#define DT_RELAENT 9 /* Size of one Rela reloc */ ++#define DT_STRSZ 10 /* Size of string table */ ++#define DT_SYMENT 11 /* Size of one symbol table entry */ ++#define DT_INIT 12 /* Address of init function */ ++#define DT_FINI 13 /* Address of termination function */ ++#define DT_SONAME 14 /* Name of shared object */ ++#define DT_RPATH 15 /* Library search path (deprecated) */ ++#define DT_SYMBOLIC 16 /* Start symbol search here */ ++#define DT_REL 17 /* Address of Rel relocs */ ++#define DT_RELSZ 18 /* Total size of Rel relocs */ ++#define DT_RELENT 19 /* Size of one Rel reloc */ ++#define DT_PLTREL 20 /* Type of reloc in PLT */ ++#define DT_DEBUG 21 /* For debugging; unspecified */ ++#define DT_TEXTREL 22 /* Reloc might modify .text */ ++#define DT_JMPREL 23 /* Address of PLT relocs */ ++#define DT_BIND_NOW 24 /* Process relocations of object */ ++#define DT_INIT_ARRAY 25 /* Array with addresses of init fct */ ++#define DT_FINI_ARRAY 26 /* Array with addresses of fini fct */ ++#define DT_INIT_ARRAYSZ 27 /* Size in bytes of DT_INIT_ARRAY */ ++#define DT_FINI_ARRAYSZ 28 /* Size in bytes of DT_FINI_ARRAY */ ++#define DT_RUNPATH 29 /* Library search path */ ++#define DT_FLAGS 30 /* Flags for the object being loaded */ ++#define DT_ENCODING 32 /* Start of encoded range */ ++#define DT_PREINIT_ARRAY 32 /* Array with addresses of preinit fct*/ ++#define DT_PREINIT_ARRAYSZ 33 /* size in bytes of DT_PREINIT_ARRAY */ ++#define DT_NUM 34 /* Number used */ ++#define DT_LOOS 0x6000000d /* Start of OS-specific */ ++#define DT_HIOS 0x6ffff000 /* End of OS-specific */ ++#define DT_LOPROC 0x70000000 /* Start of processor-specific */ ++#define DT_HIPROC 0x7fffffff /* End of processor-specific */ ++#define DT_PROCNUM DT_MIPS_NUM /* Most used by any processor */ ++ ++/* DT_* entries which fall between DT_VALRNGHI & DT_VALRNGLO use the ++ Dyn.d_un.d_val field of the Elf*_Dyn structure. This follows Sun's ++ approach. */ ++#define DT_VALRNGLO 0x6ffffd00 ++#define DT_GNU_PRELINKED 0x6ffffdf5 /* Prelinking timestamp */ ++#define DT_GNU_CONFLICTSZ 0x6ffffdf6 /* Size of conflict section */ ++#define DT_GNU_LIBLISTSZ 0x6ffffdf7 /* Size of library list */ ++#define DT_CHECKSUM 0x6ffffdf8 ++#define DT_PLTPADSZ 0x6ffffdf9 ++#define DT_MOVEENT 0x6ffffdfa ++#define DT_MOVESZ 0x6ffffdfb ++#define DT_FEATURE_1 0x6ffffdfc /* Feature selection (DTF_*). */ ++#define DT_POSFLAG_1 0x6ffffdfd /* Flags for DT_* entries, effecting ++ the following DT_* entry. */ ++#define DT_SYMINSZ 0x6ffffdfe /* Size of syminfo table (in bytes) */ ++#define DT_SYMINENT 0x6ffffdff /* Entry size of syminfo */ ++#define DT_VALRNGHI 0x6ffffdff ++#define DT_VALTAGIDX(tag) (DT_VALRNGHI - (tag)) /* Reverse order! */ ++#define DT_VALNUM 12 ++ ++/* DT_* entries which fall between DT_ADDRRNGHI & DT_ADDRRNGLO use the ++ Dyn.d_un.d_ptr field of the Elf*_Dyn structure. ++ ++ If any adjustment is made to the ELF object after it has been ++ built these entries will need to be adjusted. */ ++#define DT_ADDRRNGLO 0x6ffffe00 ++#define DT_GNU_CONFLICT 0x6ffffef8 /* Start of conflict section */ ++#define DT_GNU_LIBLIST 0x6ffffef9 /* Library list */ ++#define DT_CONFIG 0x6ffffefa /* Configuration information. */ ++#define DT_DEPAUDIT 0x6ffffefb /* Dependency auditing. */ ++#define DT_AUDIT 0x6ffffefc /* Object auditing. */ ++#define DT_PLTPAD 0x6ffffefd /* PLT padding. */ ++#define DT_MOVETAB 0x6ffffefe /* Move table. */ ++#define DT_SYMINFO 0x6ffffeff /* Syminfo table. */ ++#define DT_ADDRRNGHI 0x6ffffeff ++#define DT_ADDRTAGIDX(tag) (DT_ADDRRNGHI - (tag)) /* Reverse order! */ ++#define DT_ADDRNUM 10 ++ ++/* The versioning entry types. The next are defined as part of the ++ GNU extension. */ ++#define DT_VERSYM 0x6ffffff0 ++ ++#define DT_RELACOUNT 0x6ffffff9 ++#define DT_RELCOUNT 0x6ffffffa ++ ++/* These were chosen by Sun. */ ++#define DT_FLAGS_1 0x6ffffffb /* State flags, see DF_1_* below. */ ++#define DT_VERDEF 0x6ffffffc /* Address of version definition ++ table */ ++#define DT_VERDEFNUM 0x6ffffffd /* Number of version definitions */ ++#define DT_VERNEED 0x6ffffffe /* Address of table with needed ++ versions */ ++#define DT_VERNEEDNUM 0x6fffffff /* Number of needed versions */ ++#define DT_VERSIONTAGIDX(tag) (DT_VERNEEDNUM - (tag)) /* Reverse order! */ ++#define DT_VERSIONTAGNUM 16 ++ ++/* Sun added these machine-independent extensions in the "processor-specific" ++ range. Be compatible. */ ++#define DT_AUXILIARY 0x7ffffffd /* Shared object to load before self */ ++#define DT_FILTER 0x7fffffff /* Shared object to get values from */ ++#define DT_EXTRATAGIDX(tag) ((Elf32_Word)-((Elf32_Sword) (tag) <<1>>1)-1) ++#define DT_EXTRANUM 3 ++ ++/* Values of `d_un.d_val' in the DT_FLAGS entry. */ ++#define DF_ORIGIN 0x00000001 /* Object may use DF_ORIGIN */ ++#define DF_SYMBOLIC 0x00000002 /* Symbol resolutions starts here */ ++#define DF_TEXTREL 0x00000004 /* Object contains text relocations */ ++#define DF_BIND_NOW 0x00000008 /* No lazy binding for this object */ ++#define DF_STATIC_TLS 0x00000010 /* Module uses the static TLS model */ ++ ++/* State flags selectable in the `d_un.d_val' element of the DT_FLAGS_1 ++ entry in the dynamic section. */ ++#define DF_1_NOW 0x00000001 /* Set RTLD_NOW for this object. */ ++#define DF_1_GLOBAL 0x00000002 /* Set RTLD_GLOBAL for this object. */ ++#define DF_1_GROUP 0x00000004 /* Set RTLD_GROUP for this object. */ ++#define DF_1_NODELETE 0x00000008 /* Set RTLD_NODELETE for this object.*/ ++#define DF_1_LOADFLTR 0x00000010 /* Trigger filtee loading at runtime.*/ ++#define DF_1_INITFIRST 0x00000020 /* Set RTLD_INITFIRST for this object*/ ++#define DF_1_NOOPEN 0x00000040 /* Set RTLD_NOOPEN for this object. */ ++#define DF_1_ORIGIN 0x00000080 /* $ORIGIN must be handled. */ ++#define DF_1_DIRECT 0x00000100 /* Direct binding enabled. */ ++#define DF_1_TRANS 0x00000200 ++#define DF_1_INTERPOSE 0x00000400 /* Object is used to interpose. */ ++#define DF_1_NODEFLIB 0x00000800 /* Ignore default lib search path. */ ++#define DF_1_NODUMP 0x00001000 /* Object can't be dldump'ed. */ ++#define DF_1_CONFALT 0x00002000 /* Configuration alternative created.*/ ++#define DF_1_ENDFILTEE 0x00004000 /* Filtee terminates filters search. */ ++#define DF_1_DISPRELDNE 0x00008000 /* Disp reloc applied at build time. */ ++#define DF_1_DISPRELPND 0x00010000 /* Disp reloc applied at run-time. */ ++ ++/* Flags for the feature selection in DT_FEATURE_1. */ ++#define DTF_1_PARINIT 0x00000001 ++#define DTF_1_CONFEXP 0x00000002 ++ ++/* Flags in the DT_POSFLAG_1 entry effecting only the next DT_* entry. */ ++#define DF_P1_LAZYLOAD 0x00000001 /* Lazyload following object. */ ++#define DF_P1_GROUPPERM 0x00000002 /* Symbols from next object are not ++ generally available. */ ++ ++/* Version definition sections. */ ++ ++typedef struct ++{ ++ Elf32_Half vd_version; /* Version revision */ ++ Elf32_Half vd_flags; /* Version information */ ++ Elf32_Half vd_ndx; /* Version Index */ ++ Elf32_Half vd_cnt; /* Number of associated aux entries */ ++ Elf32_Word vd_hash; /* Version name hash value */ ++ Elf32_Word vd_aux; /* Offset in bytes to verdaux array */ ++ Elf32_Word vd_next; /* Offset in bytes to next verdef ++ entry */ ++} Elf32_Verdef; ++ ++typedef struct ++{ ++ Elf64_Half vd_version; /* Version revision */ ++ Elf64_Half vd_flags; /* Version information */ ++ Elf64_Half vd_ndx; /* Version Index */ ++ Elf64_Half vd_cnt; /* Number of associated aux entries */ ++ Elf64_Word vd_hash; /* Version name hash value */ ++ Elf64_Word vd_aux; /* Offset in bytes to verdaux array */ ++ Elf64_Word vd_next; /* Offset in bytes to next verdef ++ entry */ ++} Elf64_Verdef; ++ ++ ++/* Legal values for vd_version (version revision). */ ++#define VER_DEF_NONE 0 /* No version */ ++#define VER_DEF_CURRENT 1 /* Current version */ ++#define VER_DEF_NUM 2 /* Given version number */ ++ ++/* Legal values for vd_flags (version information flags). */ ++#define VER_FLG_BASE 0x1 /* Version definition of file itself */ ++#define VER_FLG_WEAK 0x2 /* Weak version identifier */ ++ ++/* Versym symbol index values. */ ++#define VER_NDX_LOCAL 0 /* Symbol is local. */ ++#define VER_NDX_GLOBAL 1 /* Symbol is global. */ ++#define VER_NDX_LORESERVE 0xff00 /* Beginning of reserved entries. */ ++#define VER_NDX_ELIMINATE 0xff01 /* Symbol is to be eliminated. */ ++ ++/* Auxialiary version information. */ ++ ++typedef struct ++{ ++ Elf32_Word vda_name; /* Version or dependency names */ ++ Elf32_Word vda_next; /* Offset in bytes to next verdaux ++ entry */ ++} Elf32_Verdaux; ++ ++typedef struct ++{ ++ Elf64_Word vda_name; /* Version or dependency names */ ++ Elf64_Word vda_next; /* Offset in bytes to next verdaux ++ entry */ ++} Elf64_Verdaux; ++ ++ ++/* Version dependency section. */ ++ ++typedef struct ++{ ++ Elf32_Half vn_version; /* Version of structure */ ++ Elf32_Half vn_cnt; /* Number of associated aux entries */ ++ Elf32_Word vn_file; /* Offset of filename for this ++ dependency */ ++ Elf32_Word vn_aux; /* Offset in bytes to vernaux array */ ++ Elf32_Word vn_next; /* Offset in bytes to next verneed ++ entry */ ++} Elf32_Verneed; ++ ++typedef struct ++{ ++ Elf64_Half vn_version; /* Version of structure */ ++ Elf64_Half vn_cnt; /* Number of associated aux entries */ ++ Elf64_Word vn_file; /* Offset of filename for this ++ dependency */ ++ Elf64_Word vn_aux; /* Offset in bytes to vernaux array */ ++ Elf64_Word vn_next; /* Offset in bytes to next verneed ++ entry */ ++} Elf64_Verneed; ++ ++ ++/* Legal values for vn_version (version revision). */ ++#define VER_NEED_NONE 0 /* No version */ ++#define VER_NEED_CURRENT 1 /* Current version */ ++#define VER_NEED_NUM 2 /* Given version number */ ++ ++/* Auxiliary needed version information. */ ++ ++typedef struct ++{ ++ Elf32_Word vna_hash; /* Hash value of dependency name */ ++ Elf32_Half vna_flags; /* Dependency specific information */ ++ Elf32_Half vna_other; /* Unused */ ++ Elf32_Word vna_name; /* Dependency name string offset */ ++ Elf32_Word vna_next; /* Offset in bytes to next vernaux ++ entry */ ++} Elf32_Vernaux; ++ ++typedef struct ++{ ++ Elf64_Word vna_hash; /* Hash value of dependency name */ ++ Elf64_Half vna_flags; /* Dependency specific information */ ++ Elf64_Half vna_other; /* Unused */ ++ Elf64_Word vna_name; /* Dependency name string offset */ ++ Elf64_Word vna_next; /* Offset in bytes to next vernaux ++ entry */ ++} Elf64_Vernaux; ++ ++ ++/* Legal values for vna_flags. */ ++#define VER_FLG_WEAK 0x2 /* Weak version identifier */ ++ ++ ++/* Auxiliary vector. */ ++ ++/* This vector is normally only used by the program interpreter. The ++ usual definition in an ABI supplement uses the name auxv_t. The ++ vector is not usually defined in a standard <elf.h> file, but it ++ can't hurt. We rename it to avoid conflicts. The sizes of these ++ types are an arrangement between the exec server and the program ++ interpreter, so we don't fully specify them here. */ ++ ++typedef struct ++{ ++ int a_type; /* Entry type */ ++ union ++ { ++ long int a_val; /* Integer value */ ++ void *a_ptr; /* Pointer value */ ++ void (*a_fcn) (void); /* Function pointer value */ ++ } a_un; ++} Elf32_auxv_t; ++ ++typedef struct ++{ ++ long int a_type; /* Entry type */ ++ union ++ { ++ long int a_val; /* Integer value */ ++ void *a_ptr; /* Pointer value */ ++ void (*a_fcn) (void); /* Function pointer value */ ++ } a_un; ++} Elf64_auxv_t; ++ ++/* Legal values for a_type (entry type). */ ++ ++#define AT_NULL 0 /* End of vector */ ++#define AT_IGNORE 1 /* Entry should be ignored */ ++#define AT_EXECFD 2 /* File descriptor of program */ ++#define AT_PHDR 3 /* Program headers for program */ ++#define AT_PHENT 4 /* Size of program header entry */ ++#define AT_PHNUM 5 /* Number of program headers */ ++#define AT_PAGESZ 6 /* System page size */ ++#define AT_BASE 7 /* Base address of interpreter */ ++#define AT_FLAGS 8 /* Flags */ ++#define AT_ENTRY 9 /* Entry point of program */ ++#define AT_NOTELF 10 /* Program is not ELF */ ++#define AT_UID 11 /* Real uid */ ++#define AT_EUID 12 /* Effective uid */ ++#define AT_GID 13 /* Real gid */ ++#define AT_EGID 14 /* Effective gid */ ++#define AT_CLKTCK 17 /* Frequency of times() */ ++ ++/* Some more special a_type values describing the hardware. */ ++#define AT_PLATFORM 15 /* String identifying platform. */ ++#define AT_HWCAP 16 /* Machine dependent hints about ++ processor capabilities. */ ++ ++/* This entry gives some information about the FPU initialization ++ performed by the kernel. */ ++#define AT_FPUCW 18 /* Used FPU control word. */ ++ ++/* Cache block sizes. */ ++#define AT_DCACHEBSIZE 19 /* Data cache block size. */ ++#define AT_ICACHEBSIZE 20 /* Instruction cache block size. */ ++#define AT_UCACHEBSIZE 21 /* Unified cache block size. */ ++ ++/* A special ignored value for PPC, used by the kernel to control the ++ interpretation of the AUXV. Must be > 16. */ ++#define AT_IGNOREPPC 22 /* Entry should be ignored. */ ++ ++#define AT_SECURE 23 /* Boolean, was exec setuid-like? */ ++ ++/* Pointer to the global system page used for system calls and other ++ nice things. */ ++#define AT_SYSINFO 32 ++#define AT_SYSINFO_EHDR 33 ++ ++ ++/* Note section contents. Each entry in the note section begins with ++ a header of a fixed form. */ ++ ++typedef struct ++{ ++ Elf32_Word n_namesz; /* Length of the note's name. */ ++ Elf32_Word n_descsz; /* Length of the note's descriptor. */ ++ Elf32_Word n_type; /* Type of the note. */ ++} Elf32_Nhdr; ++ ++typedef struct ++{ ++ Elf64_Word n_namesz; /* Length of the note's name. */ ++ Elf64_Word n_descsz; /* Length of the note's descriptor. */ ++ Elf64_Word n_type; /* Type of the note. */ ++} Elf64_Nhdr; ++ ++/* Known names of notes. */ ++ ++/* Solaris entries in the note section have this name. */ ++#define ELF_NOTE_SOLARIS "SUNW Solaris" ++ ++/* Note entries for GNU systems have this name. */ ++#define ELF_NOTE_GNU "GNU" ++ ++ ++/* Defined types of notes for Solaris. */ ++ ++/* Value of descriptor (one word) is desired pagesize for the binary. */ ++#define ELF_NOTE_PAGESIZE_HINT 1 ++ ++ ++/* Defined note types for GNU systems. */ ++ ++/* ABI information. The descriptor consists of words: ++ word 0: OS descriptor ++ word 1: major version of the ABI ++ word 2: minor version of the ABI ++ word 3: subminor version of the ABI ++*/ ++#define ELF_NOTE_ABI 1 ++ ++/* Known OSes. These value can appear in word 0 of an ELF_NOTE_ABI ++ note section entry. */ ++#define ELF_NOTE_OS_LINUX 0 ++#define ELF_NOTE_OS_GNU 1 ++#define ELF_NOTE_OS_SOLARIS2 2 ++#define ELF_NOTE_OS_FREEBSD 3 ++ ++ ++/* Move records. */ ++typedef struct ++{ ++ Elf32_Xword m_value; /* Symbol value. */ ++ Elf32_Word m_info; /* Size and index. */ ++ Elf32_Word m_poffset; /* Symbol offset. */ ++ Elf32_Half m_repeat; /* Repeat count. */ ++ Elf32_Half m_stride; /* Stride info. */ ++} Elf32_Move; ++ ++typedef struct ++{ ++ Elf64_Xword m_value; /* Symbol value. */ ++ Elf64_Xword m_info; /* Size and index. */ ++ Elf64_Xword m_poffset; /* Symbol offset. */ ++ Elf64_Half m_repeat; /* Repeat count. */ ++ Elf64_Half m_stride; /* Stride info. */ ++} Elf64_Move; ++ ++/* Macro to construct move records. */ ++#define ELF32_M_SYM(info) ((info) >> 8) ++#define ELF32_M_SIZE(info) ((unsigned char) (info)) ++#define ELF32_M_INFO(sym, size) (((sym) << 8) + (unsigned char) (size)) ++ ++#define ELF64_M_SYM(info) ELF32_M_SYM (info) ++#define ELF64_M_SIZE(info) ELF32_M_SIZE (info) ++#define ELF64_M_INFO(sym, size) ELF32_M_INFO (sym, size) ++ ++ ++/* Motorola 68k specific definitions. */ ++ ++/* Values for Elf32_Ehdr.e_flags. */ ++#define EF_CPU32 0x00810000 ++ ++/* m68k relocs. */ ++ ++#define R_68K_NONE 0 /* No reloc */ ++#define R_68K_32 1 /* Direct 32 bit */ ++#define R_68K_16 2 /* Direct 16 bit */ ++#define R_68K_8 3 /* Direct 8 bit */ ++#define R_68K_PC32 4 /* PC relative 32 bit */ ++#define R_68K_PC16 5 /* PC relative 16 bit */ ++#define R_68K_PC8 6 /* PC relative 8 bit */ ++#define R_68K_GOT32 7 /* 32 bit PC relative GOT entry */ ++#define R_68K_GOT16 8 /* 16 bit PC relative GOT entry */ ++#define R_68K_GOT8 9 /* 8 bit PC relative GOT entry */ ++#define R_68K_GOT32O 10 /* 32 bit GOT offset */ ++#define R_68K_GOT16O 11 /* 16 bit GOT offset */ ++#define R_68K_GOT8O 12 /* 8 bit GOT offset */ ++#define R_68K_PLT32 13 /* 32 bit PC relative PLT address */ ++#define R_68K_PLT16 14 /* 16 bit PC relative PLT address */ ++#define R_68K_PLT8 15 /* 8 bit PC relative PLT address */ ++#define R_68K_PLT32O 16 /* 32 bit PLT offset */ ++#define R_68K_PLT16O 17 /* 16 bit PLT offset */ ++#define R_68K_PLT8O 18 /* 8 bit PLT offset */ ++#define R_68K_COPY 19 /* Copy symbol at runtime */ ++#define R_68K_GLOB_DAT 20 /* Create GOT entry */ ++#define R_68K_JMP_SLOT 21 /* Create PLT entry */ ++#define R_68K_RELATIVE 22 /* Adjust by program base */ ++/* Keep this the last entry. */ ++#define R_68K_NUM 23 ++ ++/* Intel 80386 specific definitions. */ ++ ++/* i386 relocs. */ ++ ++#define R_386_NONE 0 /* No reloc */ ++#define R_386_32 1 /* Direct 32 bit */ ++#define R_386_PC32 2 /* PC relative 32 bit */ ++#define R_386_GOT32 3 /* 32 bit GOT entry */ ++#define R_386_PLT32 4 /* 32 bit PLT address */ ++#define R_386_COPY 5 /* Copy symbol at runtime */ ++#define R_386_GLOB_DAT 6 /* Create GOT entry */ ++#define R_386_JMP_SLOT 7 /* Create PLT entry */ ++#define R_386_RELATIVE 8 /* Adjust by program base */ ++#define R_386_GOTOFF 9 /* 32 bit offset to GOT */ ++#define R_386_GOTPC 10 /* 32 bit PC relative offset to GOT */ ++#define R_386_32PLT 11 ++#define R_386_TLS_TPOFF 14 /* Offset in static TLS block */ ++#define R_386_TLS_IE 15 /* Address of GOT entry for static TLS ++ block offset */ ++#define R_386_TLS_GOTIE 16 /* GOT entry for static TLS block ++ offset */ ++#define R_386_TLS_LE 17 /* Offset relative to static TLS ++ block */ ++#define R_386_TLS_GD 18 /* Direct 32 bit for GNU version of ++ general dynamic thread local data */ ++#define R_386_TLS_LDM 19 /* Direct 32 bit for GNU version of ++ local dynamic thread local data ++ in LE code */ ++#define R_386_16 20 ++#define R_386_PC16 21 ++#define R_386_8 22 ++#define R_386_PC8 23 ++#define R_386_TLS_GD_32 24 /* Direct 32 bit for general dynamic ++ thread local data */ ++#define R_386_TLS_GD_PUSH 25 /* Tag for pushl in GD TLS code */ ++#define R_386_TLS_GD_CALL 26 /* Relocation for call to ++ __tls_get_addr() */ ++#define R_386_TLS_GD_POP 27 /* Tag for popl in GD TLS code */ ++#define R_386_TLS_LDM_32 28 /* Direct 32 bit for local dynamic ++ thread local data in LE code */ ++#define R_386_TLS_LDM_PUSH 29 /* Tag for pushl in LDM TLS code */ ++#define R_386_TLS_LDM_CALL 30 /* Relocation for call to ++ __tls_get_addr() in LDM code */ ++#define R_386_TLS_LDM_POP 31 /* Tag for popl in LDM TLS code */ ++#define R_386_TLS_LDO_32 32 /* Offset relative to TLS block */ ++#define R_386_TLS_IE_32 33 /* GOT entry for negated static TLS ++ block offset */ ++#define R_386_TLS_LE_32 34 /* Negated offset relative to static ++ TLS block */ ++#define R_386_TLS_DTPMOD32 35 /* ID of module containing symbol */ ++#define R_386_TLS_DTPOFF32 36 /* Offset in TLS block */ ++#define R_386_TLS_TPOFF32 37 /* Negated offset in static TLS block */ ++/* Keep this the last entry. */ ++#define R_386_NUM 38 ++ ++/* SUN SPARC specific definitions. */ ++ ++/* Legal values for ST_TYPE subfield of st_info (symbol type). */ ++ ++#define STT_REGISTER 13 /* Global register reserved to app. */ ++ ++/* Values for Elf64_Ehdr.e_flags. */ ++ ++#define EF_SPARCV9_MM 3 ++#define EF_SPARCV9_TSO 0 ++#define EF_SPARCV9_PSO 1 ++#define EF_SPARCV9_RMO 2 ++#define EF_SPARC_LEDATA 0x800000 /* little endian data */ ++#define EF_SPARC_EXT_MASK 0xFFFF00 ++#define EF_SPARC_32PLUS 0x000100 /* generic V8+ features */ ++#define EF_SPARC_SUN_US1 0x000200 /* Sun UltraSPARC1 extensions */ ++#define EF_SPARC_HAL_R1 0x000400 /* HAL R1 extensions */ ++#define EF_SPARC_SUN_US3 0x000800 /* Sun UltraSPARCIII extensions */ ++ ++/* SPARC relocs. */ ++ ++#define R_SPARC_NONE 0 /* No reloc */ ++#define R_SPARC_8 1 /* Direct 8 bit */ ++#define R_SPARC_16 2 /* Direct 16 bit */ ++#define R_SPARC_32 3 /* Direct 32 bit */ ++#define R_SPARC_DISP8 4 /* PC relative 8 bit */ ++#define R_SPARC_DISP16 5 /* PC relative 16 bit */ ++#define R_SPARC_DISP32 6 /* PC relative 32 bit */ ++#define R_SPARC_WDISP30 7 /* PC relative 30 bit shifted */ ++#define R_SPARC_WDISP22 8 /* PC relative 22 bit shifted */ ++#define R_SPARC_HI22 9 /* High 22 bit */ ++#define R_SPARC_22 10 /* Direct 22 bit */ ++#define R_SPARC_13 11 /* Direct 13 bit */ ++#define R_SPARC_LO10 12 /* Truncated 10 bit */ ++#define R_SPARC_GOT10 13 /* Truncated 10 bit GOT entry */ ++#define R_SPARC_GOT13 14 /* 13 bit GOT entry */ ++#define R_SPARC_GOT22 15 /* 22 bit GOT entry shifted */ ++#define R_SPARC_PC10 16 /* PC relative 10 bit truncated */ ++#define R_SPARC_PC22 17 /* PC relative 22 bit shifted */ ++#define R_SPARC_WPLT30 18 /* 30 bit PC relative PLT address */ ++#define R_SPARC_COPY 19 /* Copy symbol at runtime */ ++#define R_SPARC_GLOB_DAT 20 /* Create GOT entry */ ++#define R_SPARC_JMP_SLOT 21 /* Create PLT entry */ ++#define R_SPARC_RELATIVE 22 /* Adjust by program base */ ++#define R_SPARC_UA32 23 /* Direct 32 bit unaligned */ ++ ++/* Additional Sparc64 relocs. */ ++ ++#define R_SPARC_PLT32 24 /* Direct 32 bit ref to PLT entry */ ++#define R_SPARC_HIPLT22 25 /* High 22 bit PLT entry */ ++#define R_SPARC_LOPLT10 26 /* Truncated 10 bit PLT entry */ ++#define R_SPARC_PCPLT32 27 /* PC rel 32 bit ref to PLT entry */ ++#define R_SPARC_PCPLT22 28 /* PC rel high 22 bit PLT entry */ ++#define R_SPARC_PCPLT10 29 /* PC rel trunc 10 bit PLT entry */ ++#define R_SPARC_10 30 /* Direct 10 bit */ ++#define R_SPARC_11 31 /* Direct 11 bit */ ++#define R_SPARC_64 32 /* Direct 64 bit */ ++#define R_SPARC_OLO10 33 /* 10bit with secondary 13bit addend */ ++#define R_SPARC_HH22 34 /* Top 22 bits of direct 64 bit */ ++#define R_SPARC_HM10 35 /* High middle 10 bits of ... */ ++#define R_SPARC_LM22 36 /* Low middle 22 bits of ... */ ++#define R_SPARC_PC_HH22 37 /* Top 22 bits of pc rel 64 bit */ ++#define R_SPARC_PC_HM10 38 /* High middle 10 bit of ... */ ++#define R_SPARC_PC_LM22 39 /* Low miggle 22 bits of ... */ ++#define R_SPARC_WDISP16 40 /* PC relative 16 bit shifted */ ++#define R_SPARC_WDISP19 41 /* PC relative 19 bit shifted */ ++#define R_SPARC_7 43 /* Direct 7 bit */ ++#define R_SPARC_5 44 /* Direct 5 bit */ ++#define R_SPARC_6 45 /* Direct 6 bit */ ++#define R_SPARC_DISP64 46 /* PC relative 64 bit */ ++#define R_SPARC_PLT64 47 /* Direct 64 bit ref to PLT entry */ ++#define R_SPARC_HIX22 48 /* High 22 bit complemented */ ++#define R_SPARC_LOX10 49 /* Truncated 11 bit complemented */ ++#define R_SPARC_H44 50 /* Direct high 12 of 44 bit */ ++#define R_SPARC_M44 51 /* Direct mid 22 of 44 bit */ ++#define R_SPARC_L44 52 /* Direct low 10 of 44 bit */ ++#define R_SPARC_REGISTER 53 /* Global register usage */ ++#define R_SPARC_UA64 54 /* Direct 64 bit unaligned */ ++#define R_SPARC_UA16 55 /* Direct 16 bit unaligned */ ++#define R_SPARC_TLS_GD_HI22 56 ++#define R_SPARC_TLS_GD_LO10 57 ++#define R_SPARC_TLS_GD_ADD 58 ++#define R_SPARC_TLS_GD_CALL 59 ++#define R_SPARC_TLS_LDM_HI22 60 ++#define R_SPARC_TLS_LDM_LO10 61 ++#define R_SPARC_TLS_LDM_ADD 62 ++#define R_SPARC_TLS_LDM_CALL 63 ++#define R_SPARC_TLS_LDO_HIX22 64 ++#define R_SPARC_TLS_LDO_LOX10 65 ++#define R_SPARC_TLS_LDO_ADD 66 ++#define R_SPARC_TLS_IE_HI22 67 ++#define R_SPARC_TLS_IE_LO10 68 ++#define R_SPARC_TLS_IE_LD 69 ++#define R_SPARC_TLS_IE_LDX 70 ++#define R_SPARC_TLS_IE_ADD 71 ++#define R_SPARC_TLS_LE_HIX22 72 ++#define R_SPARC_TLS_LE_LOX10 73 ++#define R_SPARC_TLS_DTPMOD32 74 ++#define R_SPARC_TLS_DTPMOD64 75 ++#define R_SPARC_TLS_DTPOFF32 76 ++#define R_SPARC_TLS_DTPOFF64 77 ++#define R_SPARC_TLS_TPOFF32 78 ++#define R_SPARC_TLS_TPOFF64 79 ++/* Keep this the last entry. */ ++#define R_SPARC_NUM 80 ++ ++/* For Sparc64, legal values for d_tag of Elf64_Dyn. */ ++ ++#define DT_SPARC_REGISTER 0x70000001 ++#define DT_SPARC_NUM 2 ++ ++/* Bits present in AT_HWCAP, primarily for Sparc32. */ ++ ++#define HWCAP_SPARC_FLUSH 1 /* The cpu supports flush insn. */ ++#define HWCAP_SPARC_STBAR 2 ++#define HWCAP_SPARC_SWAP 4 ++#define HWCAP_SPARC_MULDIV 8 ++#define HWCAP_SPARC_V9 16 /* The cpu is v9, so v8plus is ok. */ ++#define HWCAP_SPARC_ULTRA3 32 ++ ++/* MIPS R3000 specific definitions. */ ++ ++/* Legal values for e_flags field of Elf32_Ehdr. */ ++ ++#define EF_MIPS_NOREORDER 1 /* A .noreorder directive was used */ ++#define EF_MIPS_PIC 2 /* Contains PIC code */ ++#define EF_MIPS_CPIC 4 /* Uses PIC calling sequence */ ++#define EF_MIPS_XGOT 8 ++#define EF_MIPS_64BIT_WHIRL 16 ++#define EF_MIPS_ABI2 32 ++#define EF_MIPS_ABI_ON32 64 ++#define EF_MIPS_ARCH 0xf0000000 /* MIPS architecture level */ ++ ++/* Legal values for MIPS architecture level. */ ++ ++#define EF_MIPS_ARCH_1 0x00000000 /* -mips1 code. */ ++#define EF_MIPS_ARCH_2 0x10000000 /* -mips2 code. */ ++#define EF_MIPS_ARCH_3 0x20000000 /* -mips3 code. */ ++#define EF_MIPS_ARCH_4 0x30000000 /* -mips4 code. */ ++#define EF_MIPS_ARCH_5 0x40000000 /* -mips5 code. */ ++#define EF_MIPS_ARCH_32 0x60000000 /* MIPS32 code. */ ++#define EF_MIPS_ARCH_64 0x70000000 /* MIPS64 code. */ ++ ++/* The following are non-official names and should not be used. */ ++ ++#define E_MIPS_ARCH_1 0x00000000 /* -mips1 code. */ ++#define E_MIPS_ARCH_2 0x10000000 /* -mips2 code. */ ++#define E_MIPS_ARCH_3 0x20000000 /* -mips3 code. */ ++#define E_MIPS_ARCH_4 0x30000000 /* -mips4 code. */ ++#define E_MIPS_ARCH_5 0x40000000 /* -mips5 code. */ ++#define E_MIPS_ARCH_32 0x60000000 /* MIPS32 code. */ ++#define E_MIPS_ARCH_64 0x70000000 /* MIPS64 code. */ ++ ++/* Special section indices. */ ++ ++#define SHN_MIPS_ACOMMON 0xff00 /* Allocated common symbols */ ++#define SHN_MIPS_TEXT 0xff01 /* Allocated test symbols. */ ++#define SHN_MIPS_DATA 0xff02 /* Allocated data symbols. */ ++#define SHN_MIPS_SCOMMON 0xff03 /* Small common symbols */ ++#define SHN_MIPS_SUNDEFINED 0xff04 /* Small undefined symbols */ ++ ++/* Legal values for sh_type field of Elf32_Shdr. */ ++ ++#define SHT_MIPS_LIBLIST 0x70000000 /* Shared objects used in link */ ++#define SHT_MIPS_MSYM 0x70000001 ++#define SHT_MIPS_CONFLICT 0x70000002 /* Conflicting symbols */ ++#define SHT_MIPS_GPTAB 0x70000003 /* Global data area sizes */ ++#define SHT_MIPS_UCODE 0x70000004 /* Reserved for SGI/MIPS compilers */ ++#define SHT_MIPS_DEBUG 0x70000005 /* MIPS ECOFF debugging information*/ ++#define SHT_MIPS_REGINFO 0x70000006 /* Register usage information */ ++#define SHT_MIPS_PACKAGE 0x70000007 ++#define SHT_MIPS_PACKSYM 0x70000008 ++#define SHT_MIPS_RELD 0x70000009 ++#define SHT_MIPS_IFACE 0x7000000b ++#define SHT_MIPS_CONTENT 0x7000000c ++#define SHT_MIPS_OPTIONS 0x7000000d /* Miscellaneous options. */ ++#define SHT_MIPS_SHDR 0x70000010 ++#define SHT_MIPS_FDESC 0x70000011 ++#define SHT_MIPS_EXTSYM 0x70000012 ++#define SHT_MIPS_DENSE 0x70000013 ++#define SHT_MIPS_PDESC 0x70000014 ++#define SHT_MIPS_LOCSYM 0x70000015 ++#define SHT_MIPS_AUXSYM 0x70000016 ++#define SHT_MIPS_OPTSYM 0x70000017 ++#define SHT_MIPS_LOCSTR 0x70000018 ++#define SHT_MIPS_LINE 0x70000019 ++#define SHT_MIPS_RFDESC 0x7000001a ++#define SHT_MIPS_DELTASYM 0x7000001b ++#define SHT_MIPS_DELTAINST 0x7000001c ++#define SHT_MIPS_DELTACLASS 0x7000001d ++#define SHT_MIPS_DWARF 0x7000001e /* DWARF debugging information. */ ++#define SHT_MIPS_DELTADECL 0x7000001f ++#define SHT_MIPS_SYMBOL_LIB 0x70000020 ++#define SHT_MIPS_EVENTS 0x70000021 /* Event section. */ ++#define SHT_MIPS_TRANSLATE 0x70000022 ++#define SHT_MIPS_PIXIE 0x70000023 ++#define SHT_MIPS_XLATE 0x70000024 ++#define SHT_MIPS_XLATE_DEBUG 0x70000025 ++#define SHT_MIPS_WHIRL 0x70000026 ++#define SHT_MIPS_EH_REGION 0x70000027 ++#define SHT_MIPS_XLATE_OLD 0x70000028 ++#define SHT_MIPS_PDR_EXCEPTION 0x70000029 ++ ++/* Legal values for sh_flags field of Elf32_Shdr. */ ++ ++#define SHF_MIPS_GPREL 0x10000000 /* Must be part of global data area */ ++#define SHF_MIPS_MERGE 0x20000000 ++#define SHF_MIPS_ADDR 0x40000000 ++#define SHF_MIPS_STRINGS 0x80000000 ++#define SHF_MIPS_NOSTRIP 0x08000000 ++#define SHF_MIPS_LOCAL 0x04000000 ++#define SHF_MIPS_NAMES 0x02000000 ++#define SHF_MIPS_NODUPE 0x01000000 ++ ++ ++/* Symbol tables. */ ++ ++/* MIPS specific values for `st_other'. */ ++#define STO_MIPS_DEFAULT 0x0 ++#define STO_MIPS_INTERNAL 0x1 ++#define STO_MIPS_HIDDEN 0x2 ++#define STO_MIPS_PROTECTED 0x3 ++#define STO_MIPS_SC_ALIGN_UNUSED 0xff ++ ++/* MIPS specific values for `st_info'. */ ++#define STB_MIPS_SPLIT_COMMON 13 ++ ++/* Entries found in sections of type SHT_MIPS_GPTAB. */ ++ ++typedef union ++{ ++ struct ++ { ++ Elf32_Word gt_current_g_value; /* -G value used for compilation */ ++ Elf32_Word gt_unused; /* Not used */ ++ } gt_header; /* First entry in section */ ++ struct ++ { ++ Elf32_Word gt_g_value; /* If this value were used for -G */ ++ Elf32_Word gt_bytes; /* This many bytes would be used */ ++ } gt_entry; /* Subsequent entries in section */ ++} Elf32_gptab; ++ ++/* Entry found in sections of type SHT_MIPS_REGINFO. */ ++ ++typedef struct ++{ ++ Elf32_Word ri_gprmask; /* General registers used */ ++ Elf32_Word ri_cprmask[4]; /* Coprocessor registers used */ ++ Elf32_Sword ri_gp_value; /* $gp register value */ ++} Elf32_RegInfo; ++ ++/* Entries found in sections of type SHT_MIPS_OPTIONS. */ ++ ++typedef struct ++{ ++ unsigned char kind; /* Determines interpretation of the ++ variable part of descriptor. */ ++ unsigned char size; /* Size of descriptor, including header. */ ++ Elf32_Section section; /* Section header index of section affected, ++ 0 for global options. */ ++ Elf32_Word info; /* Kind-specific information. */ ++} Elf_Options; ++ ++/* Values for `kind' field in Elf_Options. */ ++ ++#define ODK_NULL 0 /* Undefined. */ ++#define ODK_REGINFO 1 /* Register usage information. */ ++#define ODK_EXCEPTIONS 2 /* Exception processing options. */ ++#define ODK_PAD 3 /* Section padding options. */ ++#define ODK_HWPATCH 4 /* Hardware workarounds performed */ ++#define ODK_FILL 5 /* record the fill value used by the linker. */ ++#define ODK_TAGS 6 /* reserve space for desktop tools to write. */ ++#define ODK_HWAND 7 /* HW workarounds. 'AND' bits when merging. */ ++#define ODK_HWOR 8 /* HW workarounds. 'OR' bits when merging. */ ++ ++/* Values for `info' in Elf_Options for ODK_EXCEPTIONS entries. */ ++ ++#define OEX_FPU_MIN 0x1f /* FPE's which MUST be enabled. */ ++#define OEX_FPU_MAX 0x1f00 /* FPE's which MAY be enabled. */ ++#define OEX_PAGE0 0x10000 /* page zero must be mapped. */ ++#define OEX_SMM 0x20000 /* Force sequential memory mode? */ ++#define OEX_FPDBUG 0x40000 /* Force floating point debug mode? */ ++#define OEX_PRECISEFP OEX_FPDBUG ++#define OEX_DISMISS 0x80000 /* Dismiss invalid address faults? */ ++ ++#define OEX_FPU_INVAL 0x10 ++#define OEX_FPU_DIV0 0x08 ++#define OEX_FPU_OFLO 0x04 ++#define OEX_FPU_UFLO 0x02 ++#define OEX_FPU_INEX 0x01 ++ ++/* Masks for `info' in Elf_Options for an ODK_HWPATCH entry. */ ++ ++#define OHW_R4KEOP 0x1 /* R4000 end-of-page patch. */ ++#define OHW_R8KPFETCH 0x2 /* may need R8000 prefetch patch. */ ++#define OHW_R5KEOP 0x4 /* R5000 end-of-page patch. */ ++#define OHW_R5KCVTL 0x8 /* R5000 cvt.[ds].l bug. clean=1. */ ++ ++#define OPAD_PREFIX 0x1 ++#define OPAD_POSTFIX 0x2 ++#define OPAD_SYMBOL 0x4 ++ ++/* Entry found in `.options' section. */ ++ ++typedef struct ++{ ++ Elf32_Word hwp_flags1; /* Extra flags. */ ++ Elf32_Word hwp_flags2; /* Extra flags. */ ++} Elf_Options_Hw; ++ ++/* Masks for `info' in ElfOptions for ODK_HWAND and ODK_HWOR entries. */ ++ ++#define OHWA0_R4KEOP_CHECKED 0x00000001 ++#define OHWA1_R4KEOP_CLEAN 0x00000002 ++ ++/* MIPS relocs. */ ++ ++#define R_MIPS_NONE 0 /* No reloc */ ++#define R_MIPS_16 1 /* Direct 16 bit */ ++#define R_MIPS_32 2 /* Direct 32 bit */ ++#define R_MIPS_REL32 3 /* PC relative 32 bit */ ++#define R_MIPS_26 4 /* Direct 26 bit shifted */ ++#define R_MIPS_HI16 5 /* High 16 bit */ ++#define R_MIPS_LO16 6 /* Low 16 bit */ ++#define R_MIPS_GPREL16 7 /* GP relative 16 bit */ ++#define R_MIPS_LITERAL 8 /* 16 bit literal entry */ ++#define R_MIPS_GOT16 9 /* 16 bit GOT entry */ ++#define R_MIPS_PC16 10 /* PC relative 16 bit */ ++#define R_MIPS_CALL16 11 /* 16 bit GOT entry for function */ ++#define R_MIPS_GPREL32 12 /* GP relative 32 bit */ ++ ++#define R_MIPS_SHIFT5 16 ++#define R_MIPS_SHIFT6 17 ++#define R_MIPS_64 18 ++#define R_MIPS_GOT_DISP 19 ++#define R_MIPS_GOT_PAGE 20 ++#define R_MIPS_GOT_OFST 21 ++#define R_MIPS_GOT_HI16 22 ++#define R_MIPS_GOT_LO16 23 ++#define R_MIPS_SUB 24 ++#define R_MIPS_INSERT_A 25 ++#define R_MIPS_INSERT_B 26 ++#define R_MIPS_DELETE 27 ++#define R_MIPS_HIGHER 28 ++#define R_MIPS_HIGHEST 29 ++#define R_MIPS_CALL_HI16 30 ++#define R_MIPS_CALL_LO16 31 ++#define R_MIPS_SCN_DISP 32 ++#define R_MIPS_REL16 33 ++#define R_MIPS_ADD_IMMEDIATE 34 ++#define R_MIPS_PJUMP 35 ++#define R_MIPS_RELGOT 36 ++#define R_MIPS_JALR 37 ++/* Keep this the last entry. */ ++#define R_MIPS_NUM 38 ++ ++/* Legal values for p_type field of Elf32_Phdr. */ ++ ++#define PT_MIPS_REGINFO 0x70000000 /* Register usage information */ ++#define PT_MIPS_RTPROC 0x70000001 /* Runtime procedure table. */ ++#define PT_MIPS_OPTIONS 0x70000002 ++ ++/* Special program header types. */ ++ ++#define PF_MIPS_LOCAL 0x10000000 ++ ++/* Legal values for d_tag field of Elf32_Dyn. */ ++ ++#define DT_MIPS_RLD_VERSION 0x70000001 /* Runtime linker interface version */ ++#define DT_MIPS_TIME_STAMP 0x70000002 /* Timestamp */ ++#define DT_MIPS_ICHECKSUM 0x70000003 /* Checksum */ ++#define DT_MIPS_IVERSION 0x70000004 /* Version string (string tbl index) */ ++#define DT_MIPS_FLAGS 0x70000005 /* Flags */ ++#define DT_MIPS_BASE_ADDRESS 0x70000006 /* Base address */ ++#define DT_MIPS_MSYM 0x70000007 ++#define DT_MIPS_CONFLICT 0x70000008 /* Address of CONFLICT section */ ++#define DT_MIPS_LIBLIST 0x70000009 /* Address of LIBLIST section */ ++#define DT_MIPS_LOCAL_GOTNO 0x7000000a /* Number of local GOT entries */ ++#define DT_MIPS_CONFLICTNO 0x7000000b /* Number of CONFLICT entries */ ++#define DT_MIPS_LIBLISTNO 0x70000010 /* Number of LIBLIST entries */ ++#define DT_MIPS_SYMTABNO 0x70000011 /* Number of DYNSYM entries */ ++#define DT_MIPS_UNREFEXTNO 0x70000012 /* First external DYNSYM */ ++#define DT_MIPS_GOTSYM 0x70000013 /* First GOT entry in DYNSYM */ ++#define DT_MIPS_HIPAGENO 0x70000014 /* Number of GOT page table entries */ ++#define DT_MIPS_RLD_MAP 0x70000016 /* Address of run time loader map. */ ++#define DT_MIPS_DELTA_CLASS 0x70000017 /* Delta C++ class definition. */ ++#define DT_MIPS_DELTA_CLASS_NO 0x70000018 /* Number of entries in ++ DT_MIPS_DELTA_CLASS. */ ++#define DT_MIPS_DELTA_INSTANCE 0x70000019 /* Delta C++ class instances. */ ++#define DT_MIPS_DELTA_INSTANCE_NO 0x7000001a /* Number of entries in ++ DT_MIPS_DELTA_INSTANCE. */ ++#define DT_MIPS_DELTA_RELOC 0x7000001b /* Delta relocations. */ ++#define DT_MIPS_DELTA_RELOC_NO 0x7000001c /* Number of entries in ++ DT_MIPS_DELTA_RELOC. */ ++#define DT_MIPS_DELTA_SYM 0x7000001d /* Delta symbols that Delta ++ relocations refer to. */ ++#define DT_MIPS_DELTA_SYM_NO 0x7000001e /* Number of entries in ++ DT_MIPS_DELTA_SYM. */ ++#define DT_MIPS_DELTA_CLASSSYM 0x70000020 /* Delta symbols that hold the ++ class declaration. */ ++#define DT_MIPS_DELTA_CLASSSYM_NO 0x70000021 /* Number of entries in ++ DT_MIPS_DELTA_CLASSSYM. */ ++#define DT_MIPS_CXX_FLAGS 0x70000022 /* Flags indicating for C++ flavor. */ ++#define DT_MIPS_PIXIE_INIT 0x70000023 ++#define DT_MIPS_SYMBOL_LIB 0x70000024 ++#define DT_MIPS_LOCALPAGE_GOTIDX 0x70000025 ++#define DT_MIPS_LOCAL_GOTIDX 0x70000026 ++#define DT_MIPS_HIDDEN_GOTIDX 0x70000027 ++#define DT_MIPS_PROTECTED_GOTIDX 0x70000028 ++#define DT_MIPS_OPTIONS 0x70000029 /* Address of .options. */ ++#define DT_MIPS_INTERFACE 0x7000002a /* Address of .interface. */ ++#define DT_MIPS_DYNSTR_ALIGN 0x7000002b ++#define DT_MIPS_INTERFACE_SIZE 0x7000002c /* Size of the .interface section. */ ++#define DT_MIPS_RLD_TEXT_RESOLVE_ADDR 0x7000002d /* Address of rld_text_rsolve ++ function stored in GOT. */ ++#define DT_MIPS_PERF_SUFFIX 0x7000002e /* Default suffix of dso to be added ++ by rld on dlopen() calls. */ ++#define DT_MIPS_COMPACT_SIZE 0x7000002f /* (O32)Size of compact rel section. */ ++#define DT_MIPS_GP_VALUE 0x70000030 /* GP value for aux GOTs. */ ++#define DT_MIPS_AUX_DYNAMIC 0x70000031 /* Address of aux .dynamic. */ ++#define DT_MIPS_NUM 0x32 ++ ++/* Legal values for DT_MIPS_FLAGS Elf32_Dyn entry. */ ++ ++#define RHF_NONE 0 /* No flags */ ++#define RHF_QUICKSTART (1 << 0) /* Use quickstart */ ++#define RHF_NOTPOT (1 << 1) /* Hash size not power of 2 */ ++#define RHF_NO_LIBRARY_REPLACEMENT (1 << 2) /* Ignore LD_LIBRARY_PATH */ ++#define RHF_NO_MOVE (1 << 3) ++#define RHF_SGI_ONLY (1 << 4) ++#define RHF_GUARANTEE_INIT (1 << 5) ++#define RHF_DELTA_C_PLUS_PLUS (1 << 6) ++#define RHF_GUARANTEE_START_INIT (1 << 7) ++#define RHF_PIXIE (1 << 8) ++#define RHF_DEFAULT_DELAY_LOAD (1 << 9) ++#define RHF_REQUICKSTART (1 << 10) ++#define RHF_REQUICKSTARTED (1 << 11) ++#define RHF_CORD (1 << 12) ++#define RHF_NO_UNRES_UNDEF (1 << 13) ++#define RHF_RLD_ORDER_SAFE (1 << 14) ++ ++/* Entries found in sections of type SHT_MIPS_LIBLIST. */ ++ ++typedef struct ++{ ++ Elf32_Word l_name; /* Name (string table index) */ ++ Elf32_Word l_time_stamp; /* Timestamp */ ++ Elf32_Word l_checksum; /* Checksum */ ++ Elf32_Word l_version; /* Interface version */ ++ Elf32_Word l_flags; /* Flags */ ++} Elf32_Lib; ++ ++typedef struct ++{ ++ Elf64_Word l_name; /* Name (string table index) */ ++ Elf64_Word l_time_stamp; /* Timestamp */ ++ Elf64_Word l_checksum; /* Checksum */ ++ Elf64_Word l_version; /* Interface version */ ++ Elf64_Word l_flags; /* Flags */ ++} Elf64_Lib; ++ ++ ++/* Legal values for l_flags. */ ++ ++#define LL_NONE 0 ++#define LL_EXACT_MATCH (1 << 0) /* Require exact match */ ++#define LL_IGNORE_INT_VER (1 << 1) /* Ignore interface version */ ++#define LL_REQUIRE_MINOR (1 << 2) ++#define LL_EXPORTS (1 << 3) ++#define LL_DELAY_LOAD (1 << 4) ++#define LL_DELTA (1 << 5) ++ ++/* Entries found in sections of type SHT_MIPS_CONFLICT. */ ++ ++typedef Elf32_Addr Elf32_Conflict; ++ ++ ++/* HPPA specific definitions. */ ++ ++/* Legal values for e_flags field of Elf32_Ehdr. */ ++ ++#define EF_PARISC_TRAPNIL 0x00010000 /* Trap nil pointer dereference. */ ++#define EF_PARISC_EXT 0x00020000 /* Program uses arch. extensions. */ ++#define EF_PARISC_LSB 0x00040000 /* Program expects little endian. */ ++#define EF_PARISC_WIDE 0x00080000 /* Program expects wide mode. */ ++#define EF_PARISC_NO_KABP 0x00100000 /* No kernel assisted branch ++ prediction. */ ++#define EF_PARISC_LAZYSWAP 0x00400000 /* Allow lazy swapping. */ ++#define EF_PARISC_ARCH 0x0000ffff /* Architecture version. */ ++ ++/* Defined values for `e_flags & EF_PARISC_ARCH' are: */ ++ ++#define EFA_PARISC_1_0 0x020b /* PA-RISC 1.0 big-endian. */ ++#define EFA_PARISC_1_1 0x0210 /* PA-RISC 1.1 big-endian. */ ++#define EFA_PARISC_2_0 0x0214 /* PA-RISC 2.0 big-endian. */ ++ ++/* Additional section indeces. */ ++ ++#define SHN_PARISC_ANSI_COMMON 0xff00 /* Section for tenatively declared ++ symbols in ANSI C. */ ++#define SHN_PARISC_HUGE_COMMON 0xff01 /* Common blocks in huge model. */ ++ ++/* Legal values for sh_type field of Elf32_Shdr. */ ++ ++#define SHT_PARISC_EXT 0x70000000 /* Contains product specific ext. */ ++#define SHT_PARISC_UNWIND 0x70000001 /* Unwind information. */ ++#define SHT_PARISC_DOC 0x70000002 /* Debug info for optimized code. */ ++ ++/* Legal values for sh_flags field of Elf32_Shdr. */ ++ ++#define SHF_PARISC_SHORT 0x20000000 /* Section with short addressing. */ ++#define SHF_PARISC_HUGE 0x40000000 /* Section far from gp. */ ++#define SHF_PARISC_SBP 0x80000000 /* Static branch prediction code. */ ++ ++/* Legal values for ST_TYPE subfield of st_info (symbol type). */ ++ ++#define STT_PARISC_MILLICODE 13 /* Millicode function entry point. */ ++ ++#define STT_HP_OPAQUE (STT_LOOS + 0x1) ++#define STT_HP_STUB (STT_LOOS + 0x2) ++ ++/* HPPA relocs. */ ++ ++#define R_PARISC_NONE 0 /* No reloc. */ ++#define R_PARISC_DIR32 1 /* Direct 32-bit reference. */ ++#define R_PARISC_DIR21L 2 /* Left 21 bits of eff. address. */ ++#define R_PARISC_DIR17R 3 /* Right 17 bits of eff. address. */ ++#define R_PARISC_DIR17F 4 /* 17 bits of eff. address. */ ++#define R_PARISC_DIR14R 6 /* Right 14 bits of eff. address. */ ++#define R_PARISC_PCREL32 9 /* 32-bit rel. address. */ ++#define R_PARISC_PCREL21L 10 /* Left 21 bits of rel. address. */ ++#define R_PARISC_PCREL17R 11 /* Right 17 bits of rel. address. */ ++#define R_PARISC_PCREL17F 12 /* 17 bits of rel. address. */ ++#define R_PARISC_PCREL14R 14 /* Right 14 bits of rel. address. */ ++#define R_PARISC_DPREL21L 18 /* Left 21 bits of rel. address. */ ++#define R_PARISC_DPREL14R 22 /* Right 14 bits of rel. address. */ ++#define R_PARISC_GPREL21L 26 /* GP-relative, left 21 bits. */ ++#define R_PARISC_GPREL14R 30 /* GP-relative, right 14 bits. */ ++#define R_PARISC_LTOFF21L 34 /* LT-relative, left 21 bits. */ ++#define R_PARISC_LTOFF14R 38 /* LT-relative, right 14 bits. */ ++#define R_PARISC_SECREL32 41 /* 32 bits section rel. address. */ ++#define R_PARISC_SEGBASE 48 /* No relocation, set segment base. */ ++#define R_PARISC_SEGREL32 49 /* 32 bits segment rel. address. */ ++#define R_PARISC_PLTOFF21L 50 /* PLT rel. address, left 21 bits. */ ++#define R_PARISC_PLTOFF14R 54 /* PLT rel. address, right 14 bits. */ ++#define R_PARISC_LTOFF_FPTR32 57 /* 32 bits LT-rel. function pointer. */ ++#define R_PARISC_LTOFF_FPTR21L 58 /* LT-rel. fct ptr, left 21 bits. */ ++#define R_PARISC_LTOFF_FPTR14R 62 /* LT-rel. fct ptr, right 14 bits. */ ++#define R_PARISC_FPTR64 64 /* 64 bits function address. */ ++#define R_PARISC_PLABEL32 65 /* 32 bits function address. */ ++#define R_PARISC_PCREL64 72 /* 64 bits PC-rel. address. */ ++#define R_PARISC_PCREL22F 74 /* 22 bits PC-rel. address. */ ++#define R_PARISC_PCREL14WR 75 /* PC-rel. address, right 14 bits. */ ++#define R_PARISC_PCREL14DR 76 /* PC rel. address, right 14 bits. */ ++#define R_PARISC_PCREL16F 77 /* 16 bits PC-rel. address. */ ++#define R_PARISC_PCREL16WF 78 /* 16 bits PC-rel. address. */ ++#define R_PARISC_PCREL16DF 79 /* 16 bits PC-rel. address. */ ++#define R_PARISC_DIR64 80 /* 64 bits of eff. address. */ ++#define R_PARISC_DIR14WR 83 /* 14 bits of eff. address. */ ++#define R_PARISC_DIR14DR 84 /* 14 bits of eff. address. */ ++#define R_PARISC_DIR16F 85 /* 16 bits of eff. address. */ ++#define R_PARISC_DIR16WF 86 /* 16 bits of eff. address. */ ++#define R_PARISC_DIR16DF 87 /* 16 bits of eff. address. */ ++#define R_PARISC_GPREL64 88 /* 64 bits of GP-rel. address. */ ++#define R_PARISC_GPREL14WR 91 /* GP-rel. address, right 14 bits. */ ++#define R_PARISC_GPREL14DR 92 /* GP-rel. address, right 14 bits. */ ++#define R_PARISC_GPREL16F 93 /* 16 bits GP-rel. address. */ ++#define R_PARISC_GPREL16WF 94 /* 16 bits GP-rel. address. */ ++#define R_PARISC_GPREL16DF 95 /* 16 bits GP-rel. address. */ ++#define R_PARISC_LTOFF64 96 /* 64 bits LT-rel. address. */ ++#define R_PARISC_LTOFF14WR 99 /* LT-rel. address, right 14 bits. */ ++#define R_PARISC_LTOFF14DR 100 /* LT-rel. address, right 14 bits. */ ++#define R_PARISC_LTOFF16F 101 /* 16 bits LT-rel. address. */ ++#define R_PARISC_LTOFF16WF 102 /* 16 bits LT-rel. address. */ ++#define R_PARISC_LTOFF16DF 103 /* 16 bits LT-rel. address. */ ++#define R_PARISC_SECREL64 104 /* 64 bits section rel. address. */ ++#define R_PARISC_SEGREL64 112 /* 64 bits segment rel. address. */ ++#define R_PARISC_PLTOFF14WR 115 /* PLT-rel. address, right 14 bits. */ ++#define R_PARISC_PLTOFF14DR 116 /* PLT-rel. address, right 14 bits. */ ++#define R_PARISC_PLTOFF16F 117 /* 16 bits LT-rel. address. */ ++#define R_PARISC_PLTOFF16WF 118 /* 16 bits PLT-rel. address. */ ++#define R_PARISC_PLTOFF16DF 119 /* 16 bits PLT-rel. address. */ ++#define R_PARISC_LTOFF_FPTR64 120 /* 64 bits LT-rel. function ptr. */ ++#define R_PARISC_LTOFF_FPTR14WR 123 /* LT-rel. fct. ptr., right 14 bits. */ ++#define R_PARISC_LTOFF_FPTR14DR 124 /* LT-rel. fct. ptr., right 14 bits. */ ++#define R_PARISC_LTOFF_FPTR16F 125 /* 16 bits LT-rel. function ptr. */ ++#define R_PARISC_LTOFF_FPTR16WF 126 /* 16 bits LT-rel. function ptr. */ ++#define R_PARISC_LTOFF_FPTR16DF 127 /* 16 bits LT-rel. function ptr. */ ++#define R_PARISC_LORESERVE 128 ++#define R_PARISC_COPY 128 /* Copy relocation. */ ++#define R_PARISC_IPLT 129 /* Dynamic reloc, imported PLT */ ++#define R_PARISC_EPLT 130 /* Dynamic reloc, exported PLT */ ++#define R_PARISC_TPREL32 153 /* 32 bits TP-rel. address. */ ++#define R_PARISC_TPREL21L 154 /* TP-rel. address, left 21 bits. */ ++#define R_PARISC_TPREL14R 158 /* TP-rel. address, right 14 bits. */ ++#define R_PARISC_LTOFF_TP21L 162 /* LT-TP-rel. address, left 21 bits. */ ++#define R_PARISC_LTOFF_TP14R 166 /* LT-TP-rel. address, right 14 bits.*/ ++#define R_PARISC_LTOFF_TP14F 167 /* 14 bits LT-TP-rel. address. */ ++#define R_PARISC_TPREL64 216 /* 64 bits TP-rel. address. */ ++#define R_PARISC_TPREL14WR 219 /* TP-rel. address, right 14 bits. */ ++#define R_PARISC_TPREL14DR 220 /* TP-rel. address, right 14 bits. */ ++#define R_PARISC_TPREL16F 221 /* 16 bits TP-rel. address. */ ++#define R_PARISC_TPREL16WF 222 /* 16 bits TP-rel. address. */ ++#define R_PARISC_TPREL16DF 223 /* 16 bits TP-rel. address. */ ++#define R_PARISC_LTOFF_TP64 224 /* 64 bits LT-TP-rel. address. */ ++#define R_PARISC_LTOFF_TP14WR 227 /* LT-TP-rel. address, right 14 bits.*/ ++#define R_PARISC_LTOFF_TP14DR 228 /* LT-TP-rel. address, right 14 bits.*/ ++#define R_PARISC_LTOFF_TP16F 229 /* 16 bits LT-TP-rel. address. */ ++#define R_PARISC_LTOFF_TP16WF 230 /* 16 bits LT-TP-rel. address. */ ++#define R_PARISC_LTOFF_TP16DF 231 /* 16 bits LT-TP-rel. address. */ ++#define R_PARISC_HIRESERVE 255 ++ ++/* Legal values for p_type field of Elf32_Phdr/Elf64_Phdr. */ ++ ++#define PT_HP_TLS (PT_LOOS + 0x0) ++#define PT_HP_CORE_NONE (PT_LOOS + 0x1) ++#define PT_HP_CORE_VERSION (PT_LOOS + 0x2) ++#define PT_HP_CORE_KERNEL (PT_LOOS + 0x3) ++#define PT_HP_CORE_COMM (PT_LOOS + 0x4) ++#define PT_HP_CORE_PROC (PT_LOOS + 0x5) ++#define PT_HP_CORE_LOADABLE (PT_LOOS + 0x6) ++#define PT_HP_CORE_STACK (PT_LOOS + 0x7) ++#define PT_HP_CORE_SHM (PT_LOOS + 0x8) ++#define PT_HP_CORE_MMF (PT_LOOS + 0x9) ++#define PT_HP_PARALLEL (PT_LOOS + 0x10) ++#define PT_HP_FASTBIND (PT_LOOS + 0x11) ++#define PT_HP_OPT_ANNOT (PT_LOOS + 0x12) ++#define PT_HP_HSL_ANNOT (PT_LOOS + 0x13) ++#define PT_HP_STACK (PT_LOOS + 0x14) ++ ++#define PT_PARISC_ARCHEXT 0x70000000 ++#define PT_PARISC_UNWIND 0x70000001 ++ ++/* Legal values for p_flags field of Elf32_Phdr/Elf64_Phdr. */ ++ ++#define PF_PARISC_SBP 0x08000000 ++ ++#define PF_HP_PAGE_SIZE 0x00100000 ++#define PF_HP_FAR_SHARED 0x00200000 ++#define PF_HP_NEAR_SHARED 0x00400000 ++#define PF_HP_CODE 0x01000000 ++#define PF_HP_MODIFY 0x02000000 ++#define PF_HP_LAZYSWAP 0x04000000 ++#define PF_HP_SBP 0x08000000 ++ ++ ++/* Alpha specific definitions. */ ++ ++/* Legal values for e_flags field of Elf64_Ehdr. */ ++ ++#define EF_ALPHA_32BIT 1 /* All addresses must be < 2GB. */ ++#define EF_ALPHA_CANRELAX 2 /* Relocations for relaxing exist. */ ++ ++/* Legal values for sh_type field of Elf64_Shdr. */ ++ ++/* These two are primerily concerned with ECOFF debugging info. */ ++#define SHT_ALPHA_DEBUG 0x70000001 ++#define SHT_ALPHA_REGINFO 0x70000002 ++ ++/* Legal values for sh_flags field of Elf64_Shdr. */ ++ ++#define SHF_ALPHA_GPREL 0x10000000 ++ ++/* Legal values for st_other field of Elf64_Sym. */ ++#define STO_ALPHA_NOPV 0x80 /* No PV required. */ ++#define STO_ALPHA_STD_GPLOAD 0x88 /* PV only used for initial ldgp. */ ++ ++/* Alpha relocs. */ ++ ++#define R_ALPHA_NONE 0 /* No reloc */ ++#define R_ALPHA_REFLONG 1 /* Direct 32 bit */ ++#define R_ALPHA_REFQUAD 2 /* Direct 64 bit */ ++#define R_ALPHA_GPREL32 3 /* GP relative 32 bit */ ++#define R_ALPHA_LITERAL 4 /* GP relative 16 bit w/optimization */ ++#define R_ALPHA_LITUSE 5 /* Optimization hint for LITERAL */ ++#define R_ALPHA_GPDISP 6 /* Add displacement to GP */ ++#define R_ALPHA_BRADDR 7 /* PC+4 relative 23 bit shifted */ ++#define R_ALPHA_HINT 8 /* PC+4 relative 16 bit shifted */ ++#define R_ALPHA_SREL16 9 /* PC relative 16 bit */ ++#define R_ALPHA_SREL32 10 /* PC relative 32 bit */ ++#define R_ALPHA_SREL64 11 /* PC relative 64 bit */ ++#define R_ALPHA_GPRELHIGH 17 /* GP relative 32 bit, high 16 bits */ ++#define R_ALPHA_GPRELLOW 18 /* GP relative 32 bit, low 16 bits */ ++#define R_ALPHA_GPREL16 19 /* GP relative 16 bit */ ++#define R_ALPHA_COPY 24 /* Copy symbol at runtime */ ++#define R_ALPHA_GLOB_DAT 25 /* Create GOT entry */ ++#define R_ALPHA_JMP_SLOT 26 /* Create PLT entry */ ++#define R_ALPHA_RELATIVE 27 /* Adjust by program base */ ++#define R_ALPHA_TLS_GD_HI 28 ++#define R_ALPHA_TLSGD 29 ++#define R_ALPHA_TLS_LDM 30 ++#define R_ALPHA_DTPMOD64 31 ++#define R_ALPHA_GOTDTPREL 32 ++#define R_ALPHA_DTPREL64 33 ++#define R_ALPHA_DTPRELHI 34 ++#define R_ALPHA_DTPRELLO 35 ++#define R_ALPHA_DTPREL16 36 ++#define R_ALPHA_GOTTPREL 37 ++#define R_ALPHA_TPREL64 38 ++#define R_ALPHA_TPRELHI 39 ++#define R_ALPHA_TPRELLO 40 ++#define R_ALPHA_TPREL16 41 ++/* Keep this the last entry. */ ++#define R_ALPHA_NUM 46 ++ ++/* Magic values of the LITUSE relocation addend. */ ++#define LITUSE_ALPHA_ADDR 0 ++#define LITUSE_ALPHA_BASE 1 ++#define LITUSE_ALPHA_BYTOFF 2 ++#define LITUSE_ALPHA_JSR 3 ++#define LITUSE_ALPHA_TLS_GD 4 ++#define LITUSE_ALPHA_TLS_LDM 5 ++ ++ ++/* PowerPC specific declarations */ ++ ++/* Values for Elf32/64_Ehdr.e_flags. */ ++#define EF_PPC_EMB 0x80000000 /* PowerPC embedded flag */ ++ ++/* Cygnus local bits below */ ++#define EF_PPC_RELOCATABLE 0x00010000 /* PowerPC -mrelocatable flag*/ ++#define EF_PPC_RELOCATABLE_LIB 0x00008000 /* PowerPC -mrelocatable-lib ++ flag */ ++ ++/* PowerPC relocations defined by the ABIs */ ++#define R_PPC_NONE 0 ++#define R_PPC_ADDR32 1 /* 32bit absolute address */ ++#define R_PPC_ADDR24 2 /* 26bit address, 2 bits ignored. */ ++#define R_PPC_ADDR16 3 /* 16bit absolute address */ ++#define R_PPC_ADDR16_LO 4 /* lower 16bit of absolute address */ ++#define R_PPC_ADDR16_HI 5 /* high 16bit of absolute address */ ++#define R_PPC_ADDR16_HA 6 /* adjusted high 16bit */ ++#define R_PPC_ADDR14 7 /* 16bit address, 2 bits ignored */ ++#define R_PPC_ADDR14_BRTAKEN 8 ++#define R_PPC_ADDR14_BRNTAKEN 9 ++#define R_PPC_REL24 10 /* PC relative 26 bit */ ++#define R_PPC_REL14 11 /* PC relative 16 bit */ ++#define R_PPC_REL14_BRTAKEN 12 ++#define R_PPC_REL14_BRNTAKEN 13 ++#define R_PPC_GOT16 14 ++#define R_PPC_GOT16_LO 15 ++#define R_PPC_GOT16_HI 16 ++#define R_PPC_GOT16_HA 17 ++#define R_PPC_PLTREL24 18 ++#define R_PPC_COPY 19 ++#define R_PPC_GLOB_DAT 20 ++#define R_PPC_JMP_SLOT 21 ++#define R_PPC_RELATIVE 22 ++#define R_PPC_LOCAL24PC 23 ++#define R_PPC_UADDR32 24 ++#define R_PPC_UADDR16 25 ++#define R_PPC_REL32 26 ++#define R_PPC_PLT32 27 ++#define R_PPC_PLTREL32 28 ++#define R_PPC_PLT16_LO 29 ++#define R_PPC_PLT16_HI 30 ++#define R_PPC_PLT16_HA 31 ++#define R_PPC_SDAREL16 32 ++#define R_PPC_SECTOFF 33 ++#define R_PPC_SECTOFF_LO 34 ++#define R_PPC_SECTOFF_HI 35 ++#define R_PPC_SECTOFF_HA 36 ++ ++/* PowerPC relocations defined for the TLS access ABI. */ ++#define R_PPC_TLS 67 /* none (sym+add)@tls */ ++#define R_PPC_DTPMOD32 68 /* word32 (sym+add)@dtpmod */ ++#define R_PPC_TPREL16 69 /* half16* (sym+add)@tprel */ ++#define R_PPC_TPREL16_LO 70 /* half16 (sym+add)@tprel@l */ ++#define R_PPC_TPREL16_HI 71 /* half16 (sym+add)@tprel@h */ ++#define R_PPC_TPREL16_HA 72 /* half16 (sym+add)@tprel@ha */ ++#define R_PPC_TPREL32 73 /* word32 (sym+add)@tprel */ ++#define R_PPC_DTPREL16 74 /* half16* (sym+add)@dtprel */ ++#define R_PPC_DTPREL16_LO 75 /* half16 (sym+add)@dtprel@l */ ++#define R_PPC_DTPREL16_HI 76 /* half16 (sym+add)@dtprel@h */ ++#define R_PPC_DTPREL16_HA 77 /* half16 (sym+add)@dtprel@ha */ ++#define R_PPC_DTPREL32 78 /* word32 (sym+add)@dtprel */ ++#define R_PPC_GOT_TLSGD16 79 /* half16* (sym+add)@got@tlsgd */ ++#define R_PPC_GOT_TLSGD16_LO 80 /* half16 (sym+add)@got@tlsgd@l */ ++#define R_PPC_GOT_TLSGD16_HI 81 /* half16 (sym+add)@got@tlsgd@h */ ++#define R_PPC_GOT_TLSGD16_HA 82 /* half16 (sym+add)@got@tlsgd@ha */ ++#define R_PPC_GOT_TLSLD16 83 /* half16* (sym+add)@got@tlsld */ ++#define R_PPC_GOT_TLSLD16_LO 84 /* half16 (sym+add)@got@tlsld@l */ ++#define R_PPC_GOT_TLSLD16_HI 85 /* half16 (sym+add)@got@tlsld@h */ ++#define R_PPC_GOT_TLSLD16_HA 86 /* half16 (sym+add)@got@tlsld@ha */ ++#define R_PPC_GOT_TPREL16 87 /* half16* (sym+add)@got@tprel */ ++#define R_PPC_GOT_TPREL16_LO 88 /* half16 (sym+add)@got@tprel@l */ ++#define R_PPC_GOT_TPREL16_HI 89 /* half16 (sym+add)@got@tprel@h */ ++#define R_PPC_GOT_TPREL16_HA 90 /* half16 (sym+add)@got@tprel@ha */ ++#define R_PPC_GOT_DTPREL16 91 /* half16* (sym+add)@got@dtprel */ ++#define R_PPC_GOT_DTPREL16_LO 92 /* half16* (sym+add)@got@dtprel@l */ ++#define R_PPC_GOT_DTPREL16_HI 93 /* half16* (sym+add)@got@dtprel@h */ ++#define R_PPC_GOT_DTPREL16_HA 94 /* half16* (sym+add)@got@dtprel@ha */ ++ ++/* Keep this the last entry. */ ++#define R_PPC_NUM 95 ++ ++/* The remaining relocs are from the Embedded ELF ABI, and are not ++ in the SVR4 ELF ABI. */ ++#define R_PPC_EMB_NADDR32 101 ++#define R_PPC_EMB_NADDR16 102 ++#define R_PPC_EMB_NADDR16_LO 103 ++#define R_PPC_EMB_NADDR16_HI 104 ++#define R_PPC_EMB_NADDR16_HA 105 ++#define R_PPC_EMB_SDAI16 106 ++#define R_PPC_EMB_SDA2I16 107 ++#define R_PPC_EMB_SDA2REL 108 ++#define R_PPC_EMB_SDA21 109 /* 16 bit offset in SDA */ ++#define R_PPC_EMB_MRKREF 110 ++#define R_PPC_EMB_RELSEC16 111 ++#define R_PPC_EMB_RELST_LO 112 ++#define R_PPC_EMB_RELST_HI 113 ++#define R_PPC_EMB_RELST_HA 114 ++#define R_PPC_EMB_BIT_FLD 115 ++#define R_PPC_EMB_RELSDA 116 /* 16 bit relative offset in SDA */ ++ ++/* Diab tool relocations. */ ++#define R_PPC_DIAB_SDA21_LO 180 /* like EMB_SDA21, but lower 16 bit */ ++#define R_PPC_DIAB_SDA21_HI 181 /* like EMB_SDA21, but high 16 bit */ ++#define R_PPC_DIAB_SDA21_HA 182 /* like EMB_SDA21, adjusted high 16 */ ++#define R_PPC_DIAB_RELSDA_LO 183 /* like EMB_RELSDA, but lower 16 bit */ ++#define R_PPC_DIAB_RELSDA_HI 184 /* like EMB_RELSDA, but high 16 bit */ ++#define R_PPC_DIAB_RELSDA_HA 185 /* like EMB_RELSDA, adjusted high 16 */ ++ ++/* This is a phony reloc to handle any old fashioned TOC16 references ++ that may still be in object files. */ ++#define R_PPC_TOC16 255 ++ ++ ++/* PowerPC64 relocations defined by the ABIs */ ++#define R_PPC64_NONE R_PPC_NONE ++#define R_PPC64_ADDR32 R_PPC_ADDR32 /* 32bit absolute address */ ++#define R_PPC64_ADDR24 R_PPC_ADDR24 /* 26bit address, word aligned */ ++#define R_PPC64_ADDR16 R_PPC_ADDR16 /* 16bit absolute address */ ++#define R_PPC64_ADDR16_LO R_PPC_ADDR16_LO /* lower 16bits of address */ ++#define R_PPC64_ADDR16_HI R_PPC_ADDR16_HI /* high 16bits of address. */ ++#define R_PPC64_ADDR16_HA R_PPC_ADDR16_HA /* adjusted high 16bits. */ ++#define R_PPC64_ADDR14 R_PPC_ADDR14 /* 16bit address, word aligned */ ++#define R_PPC64_ADDR14_BRTAKEN R_PPC_ADDR14_BRTAKEN ++#define R_PPC64_ADDR14_BRNTAKEN R_PPC_ADDR14_BRNTAKEN ++#define R_PPC64_REL24 R_PPC_REL24 /* PC-rel. 26 bit, word aligned */ ++#define R_PPC64_REL14 R_PPC_REL14 /* PC relative 16 bit */ ++#define R_PPC64_REL14_BRTAKEN R_PPC_REL14_BRTAKEN ++#define R_PPC64_REL14_BRNTAKEN R_PPC_REL14_BRNTAKEN ++#define R_PPC64_GOT16 R_PPC_GOT16 ++#define R_PPC64_GOT16_LO R_PPC_GOT16_LO ++#define R_PPC64_GOT16_HI R_PPC_GOT16_HI ++#define R_PPC64_GOT16_HA R_PPC_GOT16_HA ++ ++#define R_PPC64_COPY R_PPC_COPY ++#define R_PPC64_GLOB_DAT R_PPC_GLOB_DAT ++#define R_PPC64_JMP_SLOT R_PPC_JMP_SLOT ++#define R_PPC64_RELATIVE R_PPC_RELATIVE ++ ++#define R_PPC64_UADDR32 R_PPC_UADDR32 ++#define R_PPC64_UADDR16 R_PPC_UADDR16 ++#define R_PPC64_REL32 R_PPC_REL32 ++#define R_PPC64_PLT32 R_PPC_PLT32 ++#define R_PPC64_PLTREL32 R_PPC_PLTREL32 ++#define R_PPC64_PLT16_LO R_PPC_PLT16_LO ++#define R_PPC64_PLT16_HI R_PPC_PLT16_HI ++#define R_PPC64_PLT16_HA R_PPC_PLT16_HA ++ ++#define R_PPC64_SECTOFF R_PPC_SECTOFF ++#define R_PPC64_SECTOFF_LO R_PPC_SECTOFF_LO ++#define R_PPC64_SECTOFF_HI R_PPC_SECTOFF_HI ++#define R_PPC64_SECTOFF_HA R_PPC_SECTOFF_HA ++#define R_PPC64_ADDR30 37 /* word30 (S + A - P) >> 2 */ ++#define R_PPC64_ADDR64 38 /* doubleword64 S + A */ ++#define R_PPC64_ADDR16_HIGHER 39 /* half16 #higher(S + A) */ ++#define R_PPC64_ADDR16_HIGHERA 40 /* half16 #highera(S + A) */ ++#define R_PPC64_ADDR16_HIGHEST 41 /* half16 #highest(S + A) */ ++#define R_PPC64_ADDR16_HIGHESTA 42 /* half16 #highesta(S + A) */ ++#define R_PPC64_UADDR64 43 /* doubleword64 S + A */ ++#define R_PPC64_REL64 44 /* doubleword64 S + A - P */ ++#define R_PPC64_PLT64 45 /* doubleword64 L + A */ ++#define R_PPC64_PLTREL64 46 /* doubleword64 L + A - P */ ++#define R_PPC64_TOC16 47 /* half16* S + A - .TOC */ ++#define R_PPC64_TOC16_LO 48 /* half16 #lo(S + A - .TOC.) */ ++#define R_PPC64_TOC16_HI 49 /* half16 #hi(S + A - .TOC.) */ ++#define R_PPC64_TOC16_HA 50 /* half16 #ha(S + A - .TOC.) */ ++#define R_PPC64_TOC 51 /* doubleword64 .TOC */ ++#define R_PPC64_PLTGOT16 52 /* half16* M + A */ ++#define R_PPC64_PLTGOT16_LO 53 /* half16 #lo(M + A) */ ++#define R_PPC64_PLTGOT16_HI 54 /* half16 #hi(M + A) */ ++#define R_PPC64_PLTGOT16_HA 55 /* half16 #ha(M + A) */ ++ ++#define R_PPC64_ADDR16_DS 56 /* half16ds* (S + A) >> 2 */ ++#define R_PPC64_ADDR16_LO_DS 57 /* half16ds #lo(S + A) >> 2 */ ++#define R_PPC64_GOT16_DS 58 /* half16ds* (G + A) >> 2 */ ++#define R_PPC64_GOT16_LO_DS 59 /* half16ds #lo(G + A) >> 2 */ ++#define R_PPC64_PLT16_LO_DS 60 /* half16ds #lo(L + A) >> 2 */ ++#define R_PPC64_SECTOFF_DS 61 /* half16ds* (R + A) >> 2 */ ++#define R_PPC64_SECTOFF_LO_DS 62 /* half16ds #lo(R + A) >> 2 */ ++#define R_PPC64_TOC16_DS 63 /* half16ds* (S + A - .TOC.) >> 2 */ ++#define R_PPC64_TOC16_LO_DS 64 /* half16ds #lo(S + A - .TOC.) >> 2 */ ++#define R_PPC64_PLTGOT16_DS 65 /* half16ds* (M + A) >> 2 */ ++#define R_PPC64_PLTGOT16_LO_DS 66 /* half16ds #lo(M + A) >> 2 */ ++ ++/* PowerPC64 relocations defined for the TLS access ABI. */ ++#define R_PPC64_TLS 67 /* none (sym+add)@tls */ ++#define R_PPC64_DTPMOD64 68 /* doubleword64 (sym+add)@dtpmod */ ++#define R_PPC64_TPREL16 69 /* half16* (sym+add)@tprel */ ++#define R_PPC64_TPREL16_LO 70 /* half16 (sym+add)@tprel@l */ ++#define R_PPC64_TPREL16_HI 71 /* half16 (sym+add)@tprel@h */ ++#define R_PPC64_TPREL16_HA 72 /* half16 (sym+add)@tprel@ha */ ++#define R_PPC64_TPREL64 73 /* doubleword64 (sym+add)@tprel */ ++#define R_PPC64_DTPREL16 74 /* half16* (sym+add)@dtprel */ ++#define R_PPC64_DTPREL16_LO 75 /* half16 (sym+add)@dtprel@l */ ++#define R_PPC64_DTPREL16_HI 76 /* half16 (sym+add)@dtprel@h */ ++#define R_PPC64_DTPREL16_HA 77 /* half16 (sym+add)@dtprel@ha */ ++#define R_PPC64_DTPREL64 78 /* doubleword64 (sym+add)@dtprel */ ++#define R_PPC64_GOT_TLSGD16 79 /* half16* (sym+add)@got@tlsgd */ ++#define R_PPC64_GOT_TLSGD16_LO 80 /* half16 (sym+add)@got@tlsgd@l */ ++#define R_PPC64_GOT_TLSGD16_HI 81 /* half16 (sym+add)@got@tlsgd@h */ ++#define R_PPC64_GOT_TLSGD16_HA 82 /* half16 (sym+add)@got@tlsgd@ha */ ++#define R_PPC64_GOT_TLSLD16 83 /* half16* (sym+add)@got@tlsld */ ++#define R_PPC64_GOT_TLSLD16_LO 84 /* half16 (sym+add)@got@tlsld@l */ ++#define R_PPC64_GOT_TLSLD16_HI 85 /* half16 (sym+add)@got@tlsld@h */ ++#define R_PPC64_GOT_TLSLD16_HA 86 /* half16 (sym+add)@got@tlsld@ha */ ++#define R_PPC64_GOT_TPREL16_DS 87 /* half16ds* (sym+add)@got@tprel */ ++#define R_PPC64_GOT_TPREL16_LO_DS 88 /* half16ds (sym+add)@got@tprel@l */ ++#define R_PPC64_GOT_TPREL16_HI 89 /* half16 (sym+add)@got@tprel@h */ ++#define R_PPC64_GOT_TPREL16_HA 90 /* half16 (sym+add)@got@tprel@ha */ ++#define R_PPC64_GOT_DTPREL16_DS 91 /* half16ds* (sym+add)@got@dtprel */ ++#define R_PPC64_GOT_DTPREL16_LO_DS 92 /* half16ds (sym+add)@got@dtprel@l */ ++#define R_PPC64_GOT_DTPREL16_HI 93 /* half16 (sym+add)@got@dtprel@h */ ++#define R_PPC64_GOT_DTPREL16_HA 94 /* half16 (sym+add)@got@dtprel@ha */ ++#define R_PPC64_TPREL16_DS 95 /* half16ds* (sym+add)@tprel */ ++#define R_PPC64_TPREL16_LO_DS 96 /* half16ds (sym+add)@tprel@l */ ++#define R_PPC64_TPREL16_HIGHER 97 /* half16 (sym+add)@tprel@higher */ ++#define R_PPC64_TPREL16_HIGHERA 98 /* half16 (sym+add)@tprel@highera */ ++#define R_PPC64_TPREL16_HIGHEST 99 /* half16 (sym+add)@tprel@highest */ ++#define R_PPC64_TPREL16_HIGHESTA 100 /* half16 (sym+add)@tprel@highesta */ ++#define R_PPC64_DTPREL16_DS 101 /* half16ds* (sym+add)@dtprel */ ++#define R_PPC64_DTPREL16_LO_DS 102 /* half16ds (sym+add)@dtprel@l */ ++#define R_PPC64_DTPREL16_HIGHER 103 /* half16 (sym+add)@dtprel@higher */ ++#define R_PPC64_DTPREL16_HIGHERA 104 /* half16 (sym+add)@dtprel@highera */ ++#define R_PPC64_DTPREL16_HIGHEST 105 /* half16 (sym+add)@dtprel@highest */ ++#define R_PPC64_DTPREL16_HIGHESTA 106 /* half16 (sym+add)@dtprel@highesta */ ++ ++/* Keep this the last entry. */ ++#define R_PPC64_NUM 107 ++ ++/* PowerPC64 specific values for the Dyn d_tag field. */ ++#define DT_PPC64_GLINK (DT_LOPROC + 0) ++#define DT_PPC64_NUM 1 ++ ++ ++/* ARM specific declarations */ ++ ++/* Processor specific flags for the ELF header e_flags field. */ ++#define EF_ARM_RELEXEC 0x01 ++#define EF_ARM_HASENTRY 0x02 ++#define EF_ARM_INTERWORK 0x04 ++#define EF_ARM_APCS_26 0x08 ++#define EF_ARM_APCS_FLOAT 0x10 ++#define EF_ARM_PIC 0x20 ++#define EF_ARM_ALIGN8 0x40 /* 8-bit structure alignment is in use */ ++#define EF_ARM_NEW_ABI 0x80 ++#define EF_ARM_OLD_ABI 0x100 ++ ++/* Other constants defined in the ARM ELF spec. version B-01. */ ++/* NB. These conflict with values defined above. */ ++#define EF_ARM_SYMSARESORTED 0x04 ++#define EF_ARM_DYNSYMSUSESEGIDX 0x08 ++#define EF_ARM_MAPSYMSFIRST 0x10 ++#define EF_ARM_EABIMASK 0XFF000000 ++ ++#define EF_ARM_EABI_VERSION(flags) ((flags) & EF_ARM_EABIMASK) ++#define EF_ARM_EABI_UNKNOWN 0x00000000 ++#define EF_ARM_EABI_VER1 0x01000000 ++#define EF_ARM_EABI_VER2 0x02000000 ++ ++/* Additional symbol types for Thumb */ ++#define STT_ARM_TFUNC 0xd ++ ++/* ARM-specific values for sh_flags */ ++#define SHF_ARM_ENTRYSECT 0x10000000 /* Section contains an entry point */ ++#define SHF_ARM_COMDEF 0x80000000 /* Section may be multiply defined ++ in the input to a link step */ ++ ++/* ARM-specific program header flags */ ++#define PF_ARM_SB 0x10000000 /* Segment contains the location ++ addressed by the static base */ ++ ++/* ARM relocs. */ ++#define R_ARM_NONE 0 /* No reloc */ ++#define R_ARM_PC24 1 /* PC relative 26 bit branch */ ++#define R_ARM_ABS32 2 /* Direct 32 bit */ ++#define R_ARM_REL32 3 /* PC relative 32 bit */ ++#define R_ARM_PC13 4 ++#define R_ARM_ABS16 5 /* Direct 16 bit */ ++#define R_ARM_ABS12 6 /* Direct 12 bit */ ++#define R_ARM_THM_ABS5 7 ++#define R_ARM_ABS8 8 /* Direct 8 bit */ ++#define R_ARM_SBREL32 9 ++#define R_ARM_THM_PC22 10 ++#define R_ARM_THM_PC8 11 ++#define R_ARM_AMP_VCALL9 12 ++#define R_ARM_SWI24 13 ++#define R_ARM_THM_SWI8 14 ++#define R_ARM_XPC25 15 ++#define R_ARM_THM_XPC22 16 ++#define R_ARM_COPY 20 /* Copy symbol at runtime */ ++#define R_ARM_GLOB_DAT 21 /* Create GOT entry */ ++#define R_ARM_JUMP_SLOT 22 /* Create PLT entry */ ++#define R_ARM_RELATIVE 23 /* Adjust by program base */ ++#define R_ARM_GOTOFF 24 /* 32 bit offset to GOT */ ++#define R_ARM_GOTPC 25 /* 32 bit PC relative offset to GOT */ ++#define R_ARM_GOT32 26 /* 32 bit GOT entry */ ++#define R_ARM_PLT32 27 /* 32 bit PLT address */ ++#define R_ARM_ALU_PCREL_7_0 32 ++#define R_ARM_ALU_PCREL_15_8 33 ++#define R_ARM_ALU_PCREL_23_15 34 ++#define R_ARM_LDR_SBREL_11_0 35 ++#define R_ARM_ALU_SBREL_19_12 36 ++#define R_ARM_ALU_SBREL_27_20 37 ++#define R_ARM_GNU_VTENTRY 100 ++#define R_ARM_GNU_VTINHERIT 101 ++#define R_ARM_THM_PC11 102 /* thumb unconditional branch */ ++#define R_ARM_THM_PC9 103 /* thumb conditional branch */ ++#define R_ARM_RXPC25 249 ++#define R_ARM_RSBREL32 250 ++#define R_ARM_THM_RPC22 251 ++#define R_ARM_RREL32 252 ++#define R_ARM_RABS22 253 ++#define R_ARM_RPC24 254 ++#define R_ARM_RBASE 255 ++/* Keep this the last entry. */ ++#define R_ARM_NUM 256 ++ ++/* IA-64 specific declarations. */ ++ ++/* Processor specific flags for the Ehdr e_flags field. */ ++#define EF_IA_64_MASKOS 0x0000000f /* os-specific flags */ ++#define EF_IA_64_ABI64 0x00000010 /* 64-bit ABI */ ++#define EF_IA_64_ARCH 0xff000000 /* arch. version mask */ ++ ++/* Processor specific values for the Phdr p_type field. */ ++#define PT_IA_64_ARCHEXT (PT_LOPROC + 0) /* arch extension bits */ ++#define PT_IA_64_UNWIND (PT_LOPROC + 1) /* ia64 unwind bits */ ++ ++/* Processor specific flags for the Phdr p_flags field. */ ++#define PF_IA_64_NORECOV 0x80000000 /* spec insns w/o recovery */ ++ ++/* Processor specific values for the Shdr sh_type field. */ ++#define SHT_IA_64_EXT (SHT_LOPROC + 0) /* extension bits */ ++#define SHT_IA_64_UNWIND (SHT_LOPROC + 1) /* unwind bits */ ++ ++/* Processor specific flags for the Shdr sh_flags field. */ ++#define SHF_IA_64_SHORT 0x10000000 /* section near gp */ ++#define SHF_IA_64_NORECOV 0x20000000 /* spec insns w/o recovery */ ++ ++/* Processor specific values for the Dyn d_tag field. */ ++#define DT_IA_64_PLT_RESERVE (DT_LOPROC + 0) ++#define DT_IA_64_NUM 1 ++ ++/* IA-64 relocations. */ ++#define R_IA64_NONE 0x00 /* none */ ++#define R_IA64_IMM14 0x21 /* symbol + addend, add imm14 */ ++#define R_IA64_IMM22 0x22 /* symbol + addend, add imm22 */ ++#define R_IA64_IMM64 0x23 /* symbol + addend, mov imm64 */ ++#define R_IA64_DIR32MSB 0x24 /* symbol + addend, data4 MSB */ ++#define R_IA64_DIR32LSB 0x25 /* symbol + addend, data4 LSB */ ++#define R_IA64_DIR64MSB 0x26 /* symbol + addend, data8 MSB */ ++#define R_IA64_DIR64LSB 0x27 /* symbol + addend, data8 LSB */ ++#define R_IA64_GPREL22 0x2a /* @gprel(sym + add), add imm22 */ ++#define R_IA64_GPREL64I 0x2b /* @gprel(sym + add), mov imm64 */ ++#define R_IA64_GPREL32MSB 0x2c /* @gprel(sym + add), data4 MSB */ ++#define R_IA64_GPREL32LSB 0x2d /* @gprel(sym + add), data4 LSB */ ++#define R_IA64_GPREL64MSB 0x2e /* @gprel(sym + add), data8 MSB */ ++#define R_IA64_GPREL64LSB 0x2f /* @gprel(sym + add), data8 LSB */ ++#define R_IA64_LTOFF22 0x32 /* @ltoff(sym + add), add imm22 */ ++#define R_IA64_LTOFF64I 0x33 /* @ltoff(sym + add), mov imm64 */ ++#define R_IA64_PLTOFF22 0x3a /* @pltoff(sym + add), add imm22 */ ++#define R_IA64_PLTOFF64I 0x3b /* @pltoff(sym + add), mov imm64 */ ++#define R_IA64_PLTOFF64MSB 0x3e /* @pltoff(sym + add), data8 MSB */ ++#define R_IA64_PLTOFF64LSB 0x3f /* @pltoff(sym + add), data8 LSB */ ++#define R_IA64_FPTR64I 0x43 /* @fptr(sym + add), mov imm64 */ ++#define R_IA64_FPTR32MSB 0x44 /* @fptr(sym + add), data4 MSB */ ++#define R_IA64_FPTR32LSB 0x45 /* @fptr(sym + add), data4 LSB */ ++#define R_IA64_FPTR64MSB 0x46 /* @fptr(sym + add), data8 MSB */ ++#define R_IA64_FPTR64LSB 0x47 /* @fptr(sym + add), data8 LSB */ ++#define R_IA64_PCREL60B 0x48 /* @pcrel(sym + add), brl */ ++#define R_IA64_PCREL21B 0x49 /* @pcrel(sym + add), ptb, call */ ++#define R_IA64_PCREL21M 0x4a /* @pcrel(sym + add), chk.s */ ++#define R_IA64_PCREL21F 0x4b /* @pcrel(sym + add), fchkf */ ++#define R_IA64_PCREL32MSB 0x4c /* @pcrel(sym + add), data4 MSB */ ++#define R_IA64_PCREL32LSB 0x4d /* @pcrel(sym + add), data4 LSB */ ++#define R_IA64_PCREL64MSB 0x4e /* @pcrel(sym + add), data8 MSB */ ++#define R_IA64_PCREL64LSB 0x4f /* @pcrel(sym + add), data8 LSB */ ++#define R_IA64_LTOFF_FPTR22 0x52 /* @ltoff(@fptr(s+a)), imm22 */ ++#define R_IA64_LTOFF_FPTR64I 0x53 /* @ltoff(@fptr(s+a)), imm64 */ ++#define R_IA64_LTOFF_FPTR32MSB 0x54 /* @ltoff(@fptr(s+a)), data4 MSB */ ++#define R_IA64_LTOFF_FPTR32LSB 0x55 /* @ltoff(@fptr(s+a)), data4 LSB */ ++#define R_IA64_LTOFF_FPTR64MSB 0x56 /* @ltoff(@fptr(s+a)), data8 MSB */ ++#define R_IA64_LTOFF_FPTR64LSB 0x57 /* @ltoff(@fptr(s+a)), data8 LSB */ ++#define R_IA64_SEGREL32MSB 0x5c /* @segrel(sym + add), data4 MSB */ ++#define R_IA64_SEGREL32LSB 0x5d /* @segrel(sym + add), data4 LSB */ ++#define R_IA64_SEGREL64MSB 0x5e /* @segrel(sym + add), data8 MSB */ ++#define R_IA64_SEGREL64LSB 0x5f /* @segrel(sym + add), data8 LSB */ ++#define R_IA64_SECREL32MSB 0x64 /* @secrel(sym + add), data4 MSB */ ++#define R_IA64_SECREL32LSB 0x65 /* @secrel(sym + add), data4 LSB */ ++#define R_IA64_SECREL64MSB 0x66 /* @secrel(sym + add), data8 MSB */ ++#define R_IA64_SECREL64LSB 0x67 /* @secrel(sym + add), data8 LSB */ ++#define R_IA64_REL32MSB 0x6c /* data 4 + REL */ ++#define R_IA64_REL32LSB 0x6d /* data 4 + REL */ ++#define R_IA64_REL64MSB 0x6e /* data 8 + REL */ ++#define R_IA64_REL64LSB 0x6f /* data 8 + REL */ ++#define R_IA64_LTV32MSB 0x74 /* symbol + addend, data4 MSB */ ++#define R_IA64_LTV32LSB 0x75 /* symbol + addend, data4 LSB */ ++#define R_IA64_LTV64MSB 0x76 /* symbol + addend, data8 MSB */ ++#define R_IA64_LTV64LSB 0x77 /* symbol + addend, data8 LSB */ ++#define R_IA64_PCREL21BI 0x79 /* @pcrel(sym + add), 21bit inst */ ++#define R_IA64_PCREL22 0x7a /* @pcrel(sym + add), 22bit inst */ ++#define R_IA64_PCREL64I 0x7b /* @pcrel(sym + add), 64bit inst */ ++#define R_IA64_IPLTMSB 0x80 /* dynamic reloc, imported PLT, MSB */ ++#define R_IA64_IPLTLSB 0x81 /* dynamic reloc, imported PLT, LSB */ ++#define R_IA64_COPY 0x84 /* copy relocation */ ++#define R_IA64_SUB 0x85 /* Addend and symbol difference */ ++#define R_IA64_LTOFF22X 0x86 /* LTOFF22, relaxable. */ ++#define R_IA64_LDXMOV 0x87 /* Use of LTOFF22X. */ ++#define R_IA64_TPREL14 0x91 /* @tprel(sym + add), imm14 */ ++#define R_IA64_TPREL22 0x92 /* @tprel(sym + add), imm22 */ ++#define R_IA64_TPREL64I 0x93 /* @tprel(sym + add), imm64 */ ++#define R_IA64_TPREL64MSB 0x96 /* @tprel(sym + add), data8 MSB */ ++#define R_IA64_TPREL64LSB 0x97 /* @tprel(sym + add), data8 LSB */ ++#define R_IA64_LTOFF_TPREL22 0x9a /* @ltoff(@tprel(s+a)), imm2 */ ++#define R_IA64_DTPMOD64MSB 0xa6 /* @dtpmod(sym + add), data8 MSB */ ++#define R_IA64_DTPMOD64LSB 0xa7 /* @dtpmod(sym + add), data8 LSB */ ++#define R_IA64_LTOFF_DTPMOD22 0xaa /* @ltoff(@dtpmod(sym + add)), imm22 */ ++#define R_IA64_DTPREL14 0xb1 /* @dtprel(sym + add), imm14 */ ++#define R_IA64_DTPREL22 0xb2 /* @dtprel(sym + add), imm22 */ ++#define R_IA64_DTPREL64I 0xb3 /* @dtprel(sym + add), imm64 */ ++#define R_IA64_DTPREL32MSB 0xb4 /* @dtprel(sym + add), data4 MSB */ ++#define R_IA64_DTPREL32LSB 0xb5 /* @dtprel(sym + add), data4 LSB */ ++#define R_IA64_DTPREL64MSB 0xb6 /* @dtprel(sym + add), data8 MSB */ ++#define R_IA64_DTPREL64LSB 0xb7 /* @dtprel(sym + add), data8 LSB */ ++#define R_IA64_LTOFF_DTPREL22 0xba /* @ltoff(@dtprel(s+a)), imm22 */ ++ ++/* SH specific declarations */ ++ ++/* SH relocs. */ ++#define R_SH_NONE 0 ++#define R_SH_DIR32 1 ++#define R_SH_REL32 2 ++#define R_SH_DIR8WPN 3 ++#define R_SH_IND12W 4 ++#define R_SH_DIR8WPL 5 ++#define R_SH_DIR8WPZ 6 ++#define R_SH_DIR8BP 7 ++#define R_SH_DIR8W 8 ++#define R_SH_DIR8L 9 ++#define R_SH_SWITCH16 25 ++#define R_SH_SWITCH32 26 ++#define R_SH_USES 27 ++#define R_SH_COUNT 28 ++#define R_SH_ALIGN 29 ++#define R_SH_CODE 30 ++#define R_SH_DATA 31 ++#define R_SH_LABEL 32 ++#define R_SH_SWITCH8 33 ++#define R_SH_GNU_VTINHERIT 34 ++#define R_SH_GNU_VTENTRY 35 ++#define R_SH_TLS_GD_32 144 ++#define R_SH_TLS_LD_32 145 ++#define R_SH_TLS_LDO_32 146 ++#define R_SH_TLS_IE_32 147 ++#define R_SH_TLS_LE_32 148 ++#define R_SH_TLS_DTPMOD32 149 ++#define R_SH_TLS_DTPOFF32 150 ++#define R_SH_TLS_TPOFF32 151 ++#define R_SH_GOT32 160 ++#define R_SH_PLT32 161 ++#define R_SH_COPY 162 ++#define R_SH_GLOB_DAT 163 ++#define R_SH_JMP_SLOT 164 ++#define R_SH_RELATIVE 165 ++#define R_SH_GOTOFF 166 ++#define R_SH_GOTPC 167 ++/* Keep this the last entry. */ ++#define R_SH_NUM 256 ++ ++/* Additional s390 relocs */ ++ ++#define R_390_NONE 0 /* No reloc. */ ++#define R_390_8 1 /* Direct 8 bit. */ ++#define R_390_12 2 /* Direct 12 bit. */ ++#define R_390_16 3 /* Direct 16 bit. */ ++#define R_390_32 4 /* Direct 32 bit. */ ++#define R_390_PC32 5 /* PC relative 32 bit. */ ++#define R_390_GOT12 6 /* 12 bit GOT offset. */ ++#define R_390_GOT32 7 /* 32 bit GOT offset. */ ++#define R_390_PLT32 8 /* 32 bit PC relative PLT address. */ ++#define R_390_COPY 9 /* Copy symbol at runtime. */ ++#define R_390_GLOB_DAT 10 /* Create GOT entry. */ ++#define R_390_JMP_SLOT 11 /* Create PLT entry. */ ++#define R_390_RELATIVE 12 /* Adjust by program base. */ ++#define R_390_GOTOFF32 13 /* 32 bit offset to GOT. */ ++#define R_390_GOTPC 14 /* 32 bit PC relative offset to GOT. */ ++#define R_390_GOT16 15 /* 16 bit GOT offset. */ ++#define R_390_PC16 16 /* PC relative 16 bit. */ ++#define R_390_PC16DBL 17 /* PC relative 16 bit shifted by 1. */ ++#define R_390_PLT16DBL 18 /* 16 bit PC rel. PLT shifted by 1. */ ++#define R_390_PC32DBL 19 /* PC relative 32 bit shifted by 1. */ ++#define R_390_PLT32DBL 20 /* 32 bit PC rel. PLT shifted by 1. */ ++#define R_390_GOTPCDBL 21 /* 32 bit PC rel. GOT shifted by 1. */ ++#define R_390_64 22 /* Direct 64 bit. */ ++#define R_390_PC64 23 /* PC relative 64 bit. */ ++#define R_390_GOT64 24 /* 64 bit GOT offset. */ ++#define R_390_PLT64 25 /* 64 bit PC relative PLT address. */ ++#define R_390_GOTENT 26 /* 32 bit PC rel. to GOT entry >> 1. */ ++#define R_390_GOTOFF16 27 /* 16 bit offset to GOT. */ ++#define R_390_GOTOFF64 28 /* 64 bit offset to GOT. */ ++#define R_390_GOTPLT12 29 /* 12 bit offset to jump slot. */ ++#define R_390_GOTPLT16 30 /* 16 bit offset to jump slot. */ ++#define R_390_GOTPLT32 31 /* 32 bit offset to jump slot. */ ++#define R_390_GOTPLT64 32 /* 64 bit offset to jump slot. */ ++#define R_390_GOTPLTENT 33 /* 32 bit rel. offset to jump slot. */ ++#define R_390_PLTOFF16 34 /* 16 bit offset from GOT to PLT. */ ++#define R_390_PLTOFF32 35 /* 32 bit offset from GOT to PLT. */ ++#define R_390_PLTOFF64 36 /* 16 bit offset from GOT to PLT. */ ++#define R_390_TLS_LOAD 37 /* Tag for load insn in TLS code. */ ++#define R_390_TLS_GDCALL 38 /* Tag for function call in general ++ dynamic TLS code. */ ++#define R_390_TLS_LDCALL 39 /* Tag for function call in local ++ dynamic TLS code. */ ++#define R_390_TLS_GD32 40 /* Direct 32 bit for general dynamic ++ thread local data. */ ++#define R_390_TLS_GD64 41 /* Direct 64 bit for general dynamic ++ thread local data. */ ++#define R_390_TLS_GOTIE12 42 /* 12 bit GOT offset for static TLS ++ block offset. */ ++#define R_390_TLS_GOTIE32 43 /* 32 bit GOT offset for static TLS ++ block offset. */ ++#define R_390_TLS_GOTIE64 44 /* 64 bit GOT offset for static TLS ++ block offset. */ ++#define R_390_TLS_LDM32 45 /* Direct 32 bit for local dynamic ++ thread local data in LE code. */ ++#define R_390_TLS_LDM64 46 /* Direct 64 bit for local dynamic ++ thread local data in LE code. */ ++#define R_390_TLS_IE32 47 /* 32 bit address of GOT entry for ++ negated static TLS block offset. */ ++#define R_390_TLS_IE64 48 /* 64 bit address of GOT entry for ++ negated static TLS block offset. */ ++#define R_390_TLS_IEENT 49 /* 32 bit rel. offset to GOT entry for ++ negated static TLS block offset. */ ++#define R_390_TLS_LE32 50 /* 32 bit negated offset relative to ++ static TLS block. */ ++#define R_390_TLS_LE64 51 /* 64 bit negated offset relative to ++ static TLS block. */ ++#define R_390_TLS_LDO32 52 /* 32 bit offset relative to TLS ++ block. */ ++#define R_390_TLS_LDO64 53 /* 64 bit offset relative to TLS ++ block. */ ++#define R_390_TLS_DTPMOD 54 /* ID of module containing symbol. */ ++#define R_390_TLS_DTPOFF 55 /* Offset in TLS block. */ ++#define R_390_TLS_TPOFF 56 /* Negated offset in static TLS ++ block. */ ++ ++/* Keep this the last entry. */ ++#define R_390_NUM 57 ++ ++/* CRIS relocations. */ ++#define R_CRIS_NONE 0 ++#define R_CRIS_8 1 ++#define R_CRIS_16 2 ++#define R_CRIS_32 3 ++#define R_CRIS_8_PCREL 4 ++#define R_CRIS_16_PCREL 5 ++#define R_CRIS_32_PCREL 6 ++#define R_CRIS_GNU_VTINHERIT 7 ++#define R_CRIS_GNU_VTENTRY 8 ++#define R_CRIS_COPY 9 ++#define R_CRIS_GLOB_DAT 10 ++#define R_CRIS_JUMP_SLOT 11 ++#define R_CRIS_RELATIVE 12 ++#define R_CRIS_16_GOT 13 ++#define R_CRIS_32_GOT 14 ++#define R_CRIS_16_GOTPLT 15 ++#define R_CRIS_32_GOTPLT 16 ++#define R_CRIS_32_GOTREL 17 ++#define R_CRIS_32_PLT_GOTREL 18 ++#define R_CRIS_32_PLT_PCREL 19 ++ ++#define R_CRIS_NUM 20 ++ ++/* AMD x86-64 relocations. */ ++#define R_X86_64_NONE 0 /* No reloc */ ++#define R_X86_64_64 1 /* Direct 64 bit */ ++#define R_X86_64_PC32 2 /* PC relative 32 bit signed */ ++#define R_X86_64_GOT32 3 /* 32 bit GOT entry */ ++#define R_X86_64_PLT32 4 /* 32 bit PLT address */ ++#define R_X86_64_COPY 5 /* Copy symbol at runtime */ ++#define R_X86_64_GLOB_DAT 6 /* Create GOT entry */ ++#define R_X86_64_JUMP_SLOT 7 /* Create PLT entry */ ++#define R_X86_64_RELATIVE 8 /* Adjust by program base */ ++#define R_X86_64_GOTPCREL 9 /* 32 bit signed PC relative ++ offset to GOT */ ++#define R_X86_64_32 10 /* Direct 32 bit zero extended */ ++#define R_X86_64_32S 11 /* Direct 32 bit sign extended */ ++#define R_X86_64_16 12 /* Direct 16 bit zero extended */ ++#define R_X86_64_PC16 13 /* 16 bit sign extended pc relative */ ++#define R_X86_64_8 14 /* Direct 8 bit sign extended */ ++#define R_X86_64_PC8 15 /* 8 bit sign extended pc relative */ ++#define R_X86_64_DTPMOD64 16 /* ID of module containing symbol */ ++#define R_X86_64_DTPOFF64 17 /* Offset in module's TLS block */ ++#define R_X86_64_TPOFF64 18 /* Offset in initial TLS block */ ++#define R_X86_64_TLSGD 19 /* 32 bit signed PC relative offset ++ to two GOT entries for GD symbol */ ++#define R_X86_64_TLSLD 20 /* 32 bit signed PC relative offset ++ to two GOT entries for LD symbol */ ++#define R_X86_64_DTPOFF32 21 /* Offset in TLS block */ ++#define R_X86_64_GOTTPOFF 22 /* 32 bit signed PC relative offset ++ to GOT entry for IE symbol */ ++#define R_X86_64_TPOFF32 23 /* Offset in initial TLS block */ ++ ++#define R_X86_64_NUM 24 ++ ++__END_DECLS ++ ++#endif /* elf.h */ + + #include "elfconfig.h" + +@@ -185,3 +2631,4 @@ + void fatal(const char *fmt, ...); + void warn(const char *fmt, ...); + void merror(const char *fmt, ...); ++ +diff -Nur linux-3.11.5.orig/scripts/mod/sumversion.c linux-3.11.5/scripts/mod/sumversion.c +--- linux-3.11.5.orig/scripts/mod/sumversion.c 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/scripts/mod/sumversion.c 2013-10-16 18:09:31.000000000 +0200 +@@ -1,4 +1,4 @@ +-#include <netinet/in.h> ++/* #include <netinet/in.h> */ + #ifdef __sun__ + #include <inttypes.h> + #else +diff -Nur linux-3.11.5.orig/tools/include/tools/linux_types.h linux-3.11.5/tools/include/tools/linux_types.h +--- linux-3.11.5.orig/tools/include/tools/linux_types.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.11.5/tools/include/tools/linux_types.h 2013-10-16 18:09:31.000000000 +0200 +@@ -0,0 +1,22 @@ ++#ifndef __LINUX_TYPES_H ++#define __LINUX_TYPES_H ++ ++#include <stdint.h> ++ ++typedef uint8_t __u8; ++typedef uint8_t __be8; ++typedef uint8_t __le8; ++ ++typedef uint16_t __u16; ++typedef uint16_t __be16; ++typedef uint16_t __le16; ++ ++typedef uint32_t __u32; ++typedef uint32_t __be32; ++typedef uint32_t __le32; ++ ++typedef uint64_t __u64; ++typedef uint64_t __be64; ++typedef uint64_t __le64; ++ ++#endif diff --git a/target/linux/patches/3.14.40/cleankernel.patch b/target/linux/patches/3.14.40/cleankernel.patch new file mode 100644 index 000000000..d8c055dc3 --- /dev/null +++ b/target/linux/patches/3.14.40/cleankernel.patch @@ -0,0 +1,11 @@ +diff -Nur linux-3.11.5.orig/scripts/Makefile.headersinst linux-3.11.5/scripts/Makefile.headersinst +--- linux-3.11.5.orig/scripts/Makefile.headersinst 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/scripts/Makefile.headersinst 2013-10-15 16:33:10.000000000 +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/3.14.40/defaults.patch b/target/linux/patches/3.14.40/defaults.patch new file mode 100644 index 000000000..f071fd1dd --- /dev/null +++ b/target/linux/patches/3.14.40/defaults.patch @@ -0,0 +1,34 @@ +diff -Nur linux-3.0.4.orig/fs/Kconfig linux-3.0.4/fs/Kconfig +--- linux-3.0.4.orig/fs/Kconfig 2011-08-29 22:56:30.000000000 +0200 ++++ linux-3.0.4/fs/Kconfig 2011-10-15 22:08:44.000000000 +0200 +@@ -47,7 +47,7 @@ + def_bool n + + config EXPORTFS +- tristate ++ def_bool y + + config FILE_LOCKING + bool "Enable POSIX file locking API" if EXPERT +diff -Nur linux-3.0.4.orig/fs/notify/Kconfig linux-3.0.4/fs/notify/Kconfig +--- linux-3.0.4.orig/fs/notify/Kconfig 2011-08-29 22:56:30.000000000 +0200 ++++ linux-3.0.4/fs/notify/Kconfig 2011-10-15 22:02:00.000000000 +0200 +@@ -1,5 +1,5 @@ + config FSNOTIFY +- def_bool n ++ def_bool y + + source "fs/notify/dnotify/Kconfig" + source "fs/notify/inotify/Kconfig" +diff -Nur linux-3.11.10.orig/usr/Kconfig linux-3.11.10/usr/Kconfig +--- linux-3.11.10.orig/usr/Kconfig 2013-11-29 19:42:37.000000000 +0100 ++++ linux-3.11.10/usr/Kconfig 2013-12-27 19:15:16.000000000 +0100 +@@ -47,7 +47,7 @@ + + config RD_GZIP + bool "Support initial ramdisks compressed using gzip" if EXPERT +- default y ++ default n + depends on BLK_DEV_INITRD + select DECOMPRESS_GZIP + help diff --git a/target/linux/patches/3.14.40/disable-netfilter.patch b/target/linux/patches/3.14.40/disable-netfilter.patch new file mode 100644 index 000000000..7b1ca013a --- /dev/null +++ b/target/linux/patches/3.14.40/disable-netfilter.patch @@ -0,0 +1,160 @@ +diff -Nur linux-3.7.3.orig/net/Kconfig linux-3.7.3/net/Kconfig +--- linux-3.7.3.orig/net/Kconfig 2013-01-17 17:47:40.000000000 +0100 ++++ linux-3.7.3/net/Kconfig 2013-01-19 18:19:55.000000000 +0100 +@@ -163,7 +163,7 @@ + config NETFILTER_ADVANCED + bool "Advanced netfilter configuration" + depends on NETFILTER +- default y ++ default n + help + If you say Y here you can select between all the netfilter modules. + If you say N the more unusual ones will not be shown and the +@@ -175,7 +175,7 @@ + bool "Bridged IP/ARP packets filtering" + depends on BRIDGE && NETFILTER && INET + depends on NETFILTER_ADVANCED +- default y ++ default n + ---help--- + Enabling this option will let arptables resp. iptables see bridged + ARP resp. IP traffic. If you want a bridging firewall, you probably +diff -Nur linux-3.7.3.orig/net/netfilter/Kconfig linux-3.7.3/net/netfilter/Kconfig +--- linux-3.7.3.orig/net/netfilter/Kconfig 2013-01-17 17:47:40.000000000 +0100 ++++ linux-3.7.3/net/netfilter/Kconfig 2013-01-19 18:21:41.000000000 +0100 +@@ -22,7 +22,6 @@ + + config NETFILTER_NETLINK_LOG + tristate "Netfilter LOG over NFNETLINK interface" +- default m if NETFILTER_ADVANCED=n + select NETFILTER_NETLINK + help + If this option is enabled, the kernel will include support +@@ -34,7 +33,6 @@ + + config NF_CONNTRACK + tristate "Netfilter connection tracking support" +- default m if NETFILTER_ADVANCED=n + help + Connection tracking keeps a record of what packets have passed + through your machine, in order to figure out how they are related +@@ -60,7 +58,6 @@ + config NF_CONNTRACK_SECMARK + bool 'Connection tracking security mark support' + depends on NETWORK_SECMARK +- default m if NETFILTER_ADVANCED=n + help + This option enables security markings to be applied to + connections. Typically they are copied to connections from +@@ -177,7 +174,6 @@ + + config NF_CONNTRACK_FTP + tristate "FTP protocol support" +- default m if NETFILTER_ADVANCED=n + help + Tracking FTP connections is problematic: special helpers are + required for tracking them, and doing masquerading and other forms +@@ -211,7 +207,6 @@ + + config NF_CONNTRACK_IRC + tristate "IRC protocol support" +- default m if NETFILTER_ADVANCED=n + help + There is a commonly-used extension to IRC called + Direct Client-to-Client Protocol (DCC). This enables users to send +@@ -296,7 +291,6 @@ + + config NF_CONNTRACK_SIP + tristate "SIP protocol support" +- default m if NETFILTER_ADVANCED=n + help + SIP is an application-layer control protocol that can establish, + modify, and terminate multimedia sessions (conferences) such as +@@ -320,7 +314,6 @@ + config NF_CT_NETLINK + tristate 'Connection tracking netlink interface' + select NETFILTER_NETLINK +- default m if NETFILTER_ADVANCED=n + help + This option enables support for a netlink-based userspace interface + +@@ -424,7 +417,6 @@ + + config NETFILTER_XTABLES + tristate "Netfilter Xtables support (required for ip_tables)" +- default m if NETFILTER_ADVANCED=n + help + This is required if you intend to use any of ip_tables, + ip6_tables or arp_tables. +@@ -435,7 +427,6 @@ + + config NETFILTER_XT_MARK + tristate 'nfmark target and match support' +- default m if NETFILTER_ADVANCED=n + ---help--- + This option adds the "MARK" target and "mark" match. + +@@ -527,7 +518,6 @@ + config NETFILTER_XT_TARGET_CONNSECMARK + tristate '"CONNSECMARK" target support' + depends on NF_CONNTRACK && NF_CONNTRACK_SECMARK +- default m if NETFILTER_ADVANCED=n + help + The CONNSECMARK target copies security markings from packets + to connections, and restores security markings from connections +@@ -632,7 +622,6 @@ + + config NETFILTER_XT_TARGET_LOG + tristate "LOG target support" +- default m if NETFILTER_ADVANCED=n + help + This option adds a `LOG' target, which allows you to create rules in + any iptables table which records the packet header to the syslog. +@@ -660,7 +649,6 @@ + + config NETFILTER_XT_TARGET_NFLOG + tristate '"NFLOG" target support' +- default m if NETFILTER_ADVANCED=n + select NETFILTER_NETLINK_LOG + help + This option enables the NFLOG target, which allows to LOG +@@ -741,7 +729,6 @@ + config NETFILTER_XT_TARGET_SECMARK + tristate '"SECMARK" target support' + depends on NETWORK_SECMARK +- default m if NETFILTER_ADVANCED=n + help + The SECMARK target allows security marking of network + packets, for use with security subsystems. +@@ -751,7 +738,6 @@ + config NETFILTER_XT_TARGET_TCPMSS + tristate '"TCPMSS" target support' + depends on (IPV6 || IPV6=n) +- default m if NETFILTER_ADVANCED=n + ---help--- + This option adds a `TCPMSS' target, which allows you to alter the + MSS value of TCP SYN packets, to control the maximum size for that +@@ -856,7 +842,6 @@ + config NETFILTER_XT_MATCH_CONNTRACK + tristate '"conntrack" connection tracking match support' + depends on NF_CONNTRACK +- default m if NETFILTER_ADVANCED=n + help + This is a general conntrack match module, a superset of the state match. + +@@ -1063,7 +1048,6 @@ + config NETFILTER_XT_MATCH_POLICY + tristate 'IPsec "policy" match support' + depends on XFRM +- default m if NETFILTER_ADVANCED=n + help + Policy matching allows you to match packets based on the + IPsec policy that was used during decapsulation/will +@@ -1170,7 +1154,6 @@ + config NETFILTER_XT_MATCH_STATE + tristate '"state" match support' + depends on NF_CONNTRACK +- default m if NETFILTER_ADVANCED=n + help + Connection state matching allows you to match packets based on their + relationship to a tracked connection (ie. previous packets). This diff --git a/target/linux/patches/3.14.40/export-symbol-for-exmap.patch b/target/linux/patches/3.14.40/export-symbol-for-exmap.patch new file mode 100644 index 000000000..4f0fc8449 --- /dev/null +++ b/target/linux/patches/3.14.40/export-symbol-for-exmap.patch @@ -0,0 +1,11 @@ +diff -Nur linux-3.11.5.orig/kernel/pid.c linux-3.11.5/kernel/pid.c +--- linux-3.11.5.orig/kernel/pid.c 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/kernel/pid.c 2013-10-29 15:37:02.000000000 +0100 +@@ -450,6 +450,7 @@ + { + return find_task_by_pid_ns(vnr, task_active_pid_ns(current)); + } ++EXPORT_SYMBOL(find_task_by_vpid); + + struct pid *get_task_pid(struct task_struct *task, enum pid_type type) + { diff --git a/target/linux/patches/3.14.40/fblogo.patch b/target/linux/patches/3.14.40/fblogo.patch new file mode 100644 index 000000000..cbbb4216f --- /dev/null +++ b/target/linux/patches/3.14.40/fblogo.patch @@ -0,0 +1,2097 @@ +diff -Nur linux-3.13.7.orig/Documentation/fb/00-INDEX linux-3.13.7/Documentation/fb/00-INDEX +--- linux-3.13.7.orig/Documentation/fb/00-INDEX 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/Documentation/fb/00-INDEX 2014-03-29 16:03:17.000000000 +0100 +@@ -21,6 +21,8 @@ + - info on the driver for EP93xx LCD controller. + fbcon.txt + - intro to and usage guide for the framebuffer console (fbcon). ++fbcondecor.txt ++ - info on the Framebuffer Console Decoration + framebuffer.txt + - introduction to frame buffer devices. + gxfb.txt +diff -Nur linux-3.13.7.orig/Documentation/fb/fbcondecor.txt linux-3.13.7/Documentation/fb/fbcondecor.txt +--- linux-3.13.7.orig/Documentation/fb/fbcondecor.txt 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.7/Documentation/fb/fbcondecor.txt 2014-03-29 16:03:17.000000000 +0100 +@@ -0,0 +1,207 @@ ++What is it? ++----------- ++ ++The framebuffer decorations are a kernel feature which allows displaying a ++background picture on selected consoles. ++ ++What do I need to get it to work? ++--------------------------------- ++ ++To get fbcondecor up-and-running you will have to: ++ 1) get a copy of splashutils [1] or a similar program ++ 2) get some fbcondecor themes ++ 3) build the kernel helper program ++ 4) build your kernel with the FB_CON_DECOR option enabled. ++ ++To get fbcondecor operational right after fbcon initialization is finished, you ++will have to include a theme and the kernel helper into your initramfs image. ++Please refer to splashutils documentation for instructions on how to do that. ++ ++[1] The splashutils package can be downloaded from: ++ http://github.com/alanhaggai/fbsplash ++ ++The userspace helper ++-------------------- ++ ++The userspace fbcondecor helper (by default: /sbin/fbcondecor_helper) is called by the ++kernel whenever an important event occurs and the kernel needs some kind of ++job to be carried out. Important events include console switches and video ++mode switches (the kernel requests background images and configuration ++parameters for the current console). The fbcondecor helper must be accessible at ++all times. If it's not, fbcondecor will be switched off automatically. ++ ++It's possible to set path to the fbcondecor helper by writing it to ++/proc/sys/kernel/fbcondecor. ++ ++***************************************************************************** ++ ++The information below is mostly technical stuff. There's probably no need to ++read it unless you plan to develop a userspace helper. ++ ++The fbcondecor protocol ++----------------------- ++ ++The fbcondecor protocol defines a communication interface between the kernel and ++the userspace fbcondecor helper. ++ ++The kernel side is responsible for: ++ ++ * rendering console text, using an image as a background (instead of a ++ standard solid color fbcon uses), ++ * accepting commands from the user via ioctls on the fbcondecor device, ++ * calling the userspace helper to set things up as soon as the fb subsystem ++ is initialized. ++ ++The userspace helper is responsible for everything else, including parsing ++configuration files, decompressing the image files whenever the kernel needs ++it, and communicating with the kernel if necessary. ++ ++The fbcondecor protocol specifies how communication is done in both ways: ++kernel->userspace and userspace->helper. ++ ++Kernel -> Userspace ++------------------- ++ ++The kernel communicates with the userspace helper by calling it and specifying ++the task to be done in a series of arguments. ++ ++The arguments follow the pattern: ++<fbcondecor protocol version> <command> <parameters> ++ ++All commands defined in fbcondecor protocol v2 have the following parameters: ++ virtual console ++ framebuffer number ++ theme ++ ++Fbcondecor protocol v1 specified an additional 'fbcondecor mode' after the ++framebuffer number. Fbcondecor protocol v1 is deprecated and should not be used. ++ ++Fbcondecor protocol v2 specifies the following commands: ++ ++getpic ++------ ++ The kernel issues this command to request image data. It's up to the ++ userspace helper to find a background image appropriate for the specified ++ theme and the current resolution. The userspace helper should respond by ++ issuing the FBIOCONDECOR_SETPIC ioctl. ++ ++init ++---- ++ The kernel issues this command after the fbcondecor device is created and ++ the fbcondecor interface is initialized. Upon receiving 'init', the userspace ++ helper should parse the kernel command line (/proc/cmdline) or otherwise ++ decide whether fbcondecor is to be activated. ++ ++ To activate fbcondecor on the first console the helper should issue the ++ FBIOCONDECOR_SETCFG, FBIOCONDECOR_SETPIC and FBIOCONDECOR_SETSTATE commands, ++ in the above-mentioned order. ++ ++ When the userspace helper is called in an early phase of the boot process ++ (right after the initialization of fbcon), no filesystems will be mounted. ++ The helper program should mount sysfs and then create the appropriate ++ framebuffer, fbcondecor and tty0 devices (if they don't already exist) to get ++ current display settings and to be able to communicate with the kernel side. ++ It should probably also mount the procfs to be able to parse the kernel ++ command line parameters. ++ ++ Note that the console sem is not held when the kernel calls fbcondecor_helper ++ with the 'init' command. The fbcondecor helper should perform all ioctls with ++ origin set to FBCON_DECOR_IO_ORIG_USER. ++ ++modechange ++---------- ++ The kernel issues this command on a mode change. The helper's response should ++ be similar to the response to the 'init' command. Note that this time the ++ console sem is held and all ioctls must be performed with origin set to ++ FBCON_DECOR_IO_ORIG_KERNEL. ++ ++ ++Userspace -> Kernel ++------------------- ++ ++Userspace programs can communicate with fbcondecor via ioctls on the ++fbcondecor device. These ioctls are to be used by both the userspace helper ++(called only by the kernel) and userspace configuration tools (run by the users). ++ ++The fbcondecor helper should set the origin field to FBCON_DECOR_IO_ORIG_KERNEL ++when doing the appropriate ioctls. All userspace configuration tools should ++use FBCON_DECOR_IO_ORIG_USER. Failure to set the appropriate value in the origin ++field when performing ioctls from the kernel helper will most likely result ++in a console deadlock. ++ ++FBCON_DECOR_IO_ORIG_KERNEL instructs fbcondecor not to try to acquire the console ++semaphore. Not surprisingly, FBCON_DECOR_IO_ORIG_USER instructs it to acquire ++the console sem. ++ ++The framebuffer console decoration provides the following ioctls (all defined in ++linux/fb.h): ++ ++FBIOCONDECOR_SETPIC ++description: loads a background picture for a virtual console ++argument: struct fbcon_decor_iowrapper*; data: struct fb_image* ++notes: ++If called for consoles other than the current foreground one, the picture data ++will be ignored. ++ ++If the current virtual console is running in a 8-bpp mode, the cmap substruct ++of fb_image has to be filled appropriately: start should be set to 16 (first ++16 colors are reserved for fbcon), len to a value <= 240 and red, green and ++blue should point to valid cmap data. The transp field is ingored. The fields ++dx, dy, bg_color, fg_color in fb_image are ignored as well. ++ ++FBIOCONDECOR_SETCFG ++description: sets the fbcondecor config for a virtual console ++argument: struct fbcon_decor_iowrapper*; data: struct vc_decor* ++notes: The structure has to be filled with valid data. ++ ++FBIOCONDECOR_GETCFG ++description: gets the fbcondecor config for a virtual console ++argument: struct fbcon_decor_iowrapper*; data: struct vc_decor* ++ ++FBIOCONDECOR_SETSTATE ++description: sets the fbcondecor state for a virtual console ++argument: struct fbcon_decor_iowrapper*; data: unsigned int* ++ values: 0 = disabled, 1 = enabled. ++ ++FBIOCONDECOR_GETSTATE ++description: gets the fbcondecor state for a virtual console ++argument: struct fbcon_decor_iowrapper*; data: unsigned int* ++ values: as in FBIOCONDECOR_SETSTATE ++ ++Info on used structures: ++ ++Definition of struct vc_decor can be found in linux/console_decor.h. It's ++heavily commented. Note that the 'theme' field should point to a string ++no longer than FBCON_DECOR_THEME_LEN. When FBIOCONDECOR_GETCFG call is ++performed, the theme field should point to a char buffer of length ++FBCON_DECOR_THEME_LEN. ++ ++Definition of struct fbcon_decor_iowrapper can be found in linux/fb.h. ++The fields in this struct have the following meaning: ++ ++vc: ++Virtual console number. ++ ++origin: ++Specifies if the ioctl is performed as a response to a kernel request. The ++fbcondecor helper should set this field to FBCON_DECOR_IO_ORIG_KERNEL, userspace ++programs should set it to FBCON_DECOR_IO_ORIG_USER. This field is necessary to ++avoid console semaphore deadlocks. ++ ++data: ++Pointer to a data structure appropriate for the performed ioctl. Type of ++the data struct is specified in the ioctls description. ++ ++***************************************************************************** ++ ++Credit ++------ ++ ++Original 'bootsplash' project & implementation by: ++ Volker Poplawski <volker@poplawski.de>, Stefan Reinauer <stepan@suse.de>, ++ Steffen Winterfeldt <snwint@suse.de>, Michael Schroeder <mls@suse.de>, ++ Ken Wimer <wimer@suse.de>. ++ ++Fbcondecor, fbcondecor protocol design, current implementation & docs by: ++ Michal Januszewski <michalj+fbcondecor@gmail.com> ++ +diff -Nur linux-3.13.7.orig/drivers/Makefile linux-3.13.7/drivers/Makefile +--- linux-3.13.7.orig/drivers/Makefile 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/Makefile 2014-03-29 16:03:17.000000000 +0100 +@@ -17,6 +17,10 @@ + obj-$(CONFIG_PCI) += pci/ + obj-$(CONFIG_PARISC) += parisc/ + obj-$(CONFIG_RAPIDIO) += rapidio/ ++# tty/ comes before char/ so that the VT console is the boot-time ++# default. ++obj-y += tty/ ++obj-y += char/ + obj-y += video/ + obj-y += idle/ + +@@ -42,11 +46,6 @@ + # reset controllers early, since gpu drivers might rely on them to initialize + obj-$(CONFIG_RESET_CONTROLLER) += reset/ + +-# tty/ comes before char/ so that the VT console is the boot-time +-# default. +-obj-y += tty/ +-obj-y += char/ +- + # gpu/ comes after char for AGP vs DRM startup + obj-y += gpu/ + +diff -Nur linux-3.13.7.orig/drivers/video/Kconfig linux-3.13.7/drivers/video/Kconfig +--- linux-3.13.7.orig/drivers/video/Kconfig 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/Kconfig 2014-03-29 16:03:17.000000000 +0100 +@@ -1231,7 +1231,6 @@ + select FB_CFB_FILLRECT + select FB_CFB_COPYAREA + select FB_CFB_IMAGEBLIT +- select FB_TILEBLITTING + select FB_MACMODES if PPC_PMAC + ---help--- + Say Y here if you have a Matrox Millennium, Matrox Millennium II, +diff -Nur linux-3.13.7.orig/drivers/video/console/Kconfig linux-3.13.7/drivers/video/console/Kconfig +--- linux-3.13.7.orig/drivers/video/console/Kconfig 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/Kconfig 2014-03-29 16:03:17.000000000 +0100 +@@ -125,6 +125,19 @@ + such that other users of the framebuffer will remain normally + oriented. + ++config FB_CON_DECOR ++ bool "Support for the Framebuffer Console Decorations" ++ depends on FRAMEBUFFER_CONSOLE=y && !FB_TILEBLITTING ++ default n ++ ---help--- ++ This option enables support for framebuffer console decorations which ++ makes it possible to display images in the background of the system ++ consoles. Note that userspace utilities are necessary in order to take ++ advantage of these features. Refer to Documentation/fb/fbcondecor.txt ++ for more information. ++ ++ If unsure, say N. ++ + config STI_CONSOLE + bool "STI text console" + depends on PARISC +diff -Nur linux-3.13.7.orig/drivers/video/console/Makefile linux-3.13.7/drivers/video/console/Makefile +--- linux-3.13.7.orig/drivers/video/console/Makefile 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/Makefile 2014-03-29 16:03:17.000000000 +0100 +@@ -16,4 +16,5 @@ + fbcon_ccw.o + endif + ++obj-$(CONFIG_FB_CON_DECOR) += fbcondecor.o cfbcondecor.o + obj-$(CONFIG_FB_STI) += sticore.o +diff -Nur linux-3.13.7.orig/drivers/video/console/bitblit.c linux-3.13.7/drivers/video/console/bitblit.c +--- linux-3.13.7.orig/drivers/video/console/bitblit.c 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/bitblit.c 2014-03-29 16:03:17.000000000 +0100 +@@ -18,6 +18,7 @@ + #include <linux/console.h> + #include <asm/types.h> + #include "fbcon.h" ++#include "fbcondecor.h" + + /* + * Accelerated handlers. +@@ -55,6 +56,13 @@ + area.height = height * vc->vc_font.height; + area.width = width * vc->vc_font.width; + ++ if (fbcon_decor_active(info, vc)) { ++ area.sx += vc->vc_decor.tx; ++ area.sy += vc->vc_decor.ty; ++ area.dx += vc->vc_decor.tx; ++ area.dy += vc->vc_decor.ty; ++ } ++ + info->fbops->fb_copyarea(info, &area); + } + +@@ -380,11 +388,15 @@ + cursor.image.depth = 1; + cursor.rop = ROP_XOR; + +- if (info->fbops->fb_cursor) +- err = info->fbops->fb_cursor(info, &cursor); ++ if (fbcon_decor_active(info, vc)) { ++ fbcon_decor_cursor(info, &cursor); ++ } else { ++ if (info->fbops->fb_cursor) ++ err = info->fbops->fb_cursor(info, &cursor); + +- if (err) +- soft_cursor(info, &cursor); ++ if (err) ++ soft_cursor(info, &cursor); ++ } + + ops->cursor_reset = 0; + } +diff -Nur linux-3.13.7.orig/drivers/video/console/cfbcondecor.c linux-3.13.7/drivers/video/console/cfbcondecor.c +--- linux-3.13.7.orig/drivers/video/console/cfbcondecor.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/cfbcondecor.c 2014-03-29 16:03:17.000000000 +0100 +@@ -0,0 +1,471 @@ ++/* ++ * linux/drivers/video/cfbcon_decor.c -- Framebuffer decor render functions ++ * ++ * Copyright (C) 2004 Michal Januszewski <michalj+fbcondecor@gmail.com> ++ * ++ * Code based upon "Bootdecor" (C) 2001-2003 ++ * Volker Poplawski <volker@poplawski.de>, ++ * Stefan Reinauer <stepan@suse.de>, ++ * Steffen Winterfeldt <snwint@suse.de>, ++ * Michael Schroeder <mls@suse.de>, ++ * Ken Wimer <wimer@suse.de>. ++ * ++ * This file is subject to the terms and conditions of the GNU General Public ++ * License. See the file COPYING in the main directory of this archive for ++ * more details. ++ */ ++#include <linux/module.h> ++#include <linux/types.h> ++#include <linux/fb.h> ++#include <linux/selection.h> ++#include <linux/slab.h> ++#include <linux/vt_kern.h> ++#include <asm/irq.h> ++ ++#include "fbcon.h" ++#include "fbcondecor.h" ++ ++#define parse_pixel(shift,bpp,type) \ ++ do { \ ++ if (d & (0x80 >> (shift))) \ ++ dd2[(shift)] = fgx; \ ++ else \ ++ dd2[(shift)] = transparent ? *(type *)decor_src : bgx; \ ++ decor_src += (bpp); \ ++ } while (0) \ ++ ++extern int get_color(struct vc_data *vc, struct fb_info *info, ++ u16 c, int is_fg); ++ ++void fbcon_decor_fix_pseudo_pal(struct fb_info *info, struct vc_data *vc) ++{ ++ int i, j, k; ++ int minlen = min(min(info->var.red.length, info->var.green.length), ++ info->var.blue.length); ++ u32 col; ++ ++ for (j = i = 0; i < 16; i++) { ++ k = color_table[i]; ++ ++ col = ((vc->vc_palette[j++] >> (8-minlen)) ++ << info->var.red.offset); ++ col |= ((vc->vc_palette[j++] >> (8-minlen)) ++ << info->var.green.offset); ++ col |= ((vc->vc_palette[j++] >> (8-minlen)) ++ << info->var.blue.offset); ++ ((u32 *)info->pseudo_palette)[k] = col; ++ } ++} ++ ++void fbcon_decor_renderc(struct fb_info *info, int ypos, int xpos, int height, ++ int width, u8* src, u32 fgx, u32 bgx, u8 transparent) ++{ ++ unsigned int x, y; ++ u32 dd; ++ int bytespp = ((info->var.bits_per_pixel + 7) >> 3); ++ unsigned int d = ypos * info->fix.line_length + xpos * bytespp; ++ unsigned int ds = (ypos * info->var.xres + xpos) * bytespp; ++ u16 dd2[4]; ++ ++ u8* decor_src = (u8 *)(info->bgdecor.data + ds); ++ u8* dst = (u8 *)(info->screen_base + d); ++ ++ if ((ypos + height) > info->var.yres || (xpos + width) > info->var.xres) ++ return; ++ ++ for (y = 0; y < height; y++) { ++ switch (info->var.bits_per_pixel) { ++ ++ case 32: ++ for (x = 0; x < width; x++) { ++ ++ if ((x & 7) == 0) ++ d = *src++; ++ if (d & 0x80) ++ dd = fgx; ++ else ++ dd = transparent ? ++ *(u32 *)decor_src : bgx; ++ ++ d <<= 1; ++ decor_src += 4; ++ fb_writel(dd, dst); ++ dst += 4; ++ } ++ break; ++ case 24: ++ for (x = 0; x < width; x++) { ++ ++ if ((x & 7) == 0) ++ d = *src++; ++ if (d & 0x80) ++ dd = fgx; ++ else ++ dd = transparent ? ++ (*(u32 *)decor_src & 0xffffff) : bgx; ++ ++ d <<= 1; ++ decor_src += 3; ++#ifdef __LITTLE_ENDIAN ++ fb_writew(dd & 0xffff, dst); ++ dst += 2; ++ fb_writeb((dd >> 16), dst); ++#else ++ fb_writew(dd >> 8, dst); ++ dst += 2; ++ fb_writeb(dd & 0xff, dst); ++#endif ++ dst++; ++ } ++ break; ++ case 16: ++ for (x = 0; x < width; x += 2) { ++ if ((x & 7) == 0) ++ d = *src++; ++ ++ parse_pixel(0, 2, u16); ++ parse_pixel(1, 2, u16); ++#ifdef __LITTLE_ENDIAN ++ dd = dd2[0] | (dd2[1] << 16); ++#else ++ dd = dd2[1] | (dd2[0] << 16); ++#endif ++ d <<= 2; ++ fb_writel(dd, dst); ++ dst += 4; ++ } ++ break; ++ ++ case 8: ++ for (x = 0; x < width; x += 4) { ++ if ((x & 7) == 0) ++ d = *src++; ++ ++ parse_pixel(0, 1, u8); ++ parse_pixel(1, 1, u8); ++ parse_pixel(2, 1, u8); ++ parse_pixel(3, 1, u8); ++ ++#ifdef __LITTLE_ENDIAN ++ dd = dd2[0] | (dd2[1] << 8) | (dd2[2] << 16) | (dd2[3] << 24); ++#else ++ dd = dd2[3] | (dd2[2] << 8) | (dd2[1] << 16) | (dd2[0] << 24); ++#endif ++ d <<= 4; ++ fb_writel(dd, dst); ++ dst += 4; ++ } ++ } ++ ++ dst += info->fix.line_length - width * bytespp; ++ decor_src += (info->var.xres - width) * bytespp; ++ } ++} ++ ++#define cc2cx(a) \ ++ ((info->fix.visual == FB_VISUAL_TRUECOLOR || \ ++ info->fix.visual == FB_VISUAL_DIRECTCOLOR) ? \ ++ ((u32*)info->pseudo_palette)[a] : a) ++ ++void fbcon_decor_putcs(struct vc_data *vc, struct fb_info *info, ++ const unsigned short *s, int count, int yy, int xx) ++{ ++ unsigned short charmask = vc->vc_hi_font_mask ? 0x1ff : 0xff; ++ struct fbcon_ops *ops = info->fbcon_par; ++ int fg_color, bg_color, transparent; ++ u8 *src; ++ u32 bgx, fgx; ++ u16 c = scr_readw(s); ++ ++ fg_color = get_color(vc, info, c, 1); ++ bg_color = get_color(vc, info, c, 0); ++ ++ /* Don't paint the background image if console is blanked */ ++ transparent = ops->blank_state ? 0 : ++ (vc->vc_decor.bg_color == bg_color); ++ ++ xx = xx * vc->vc_font.width + vc->vc_decor.tx; ++ yy = yy * vc->vc_font.height + vc->vc_decor.ty; ++ ++ fgx = cc2cx(fg_color); ++ bgx = cc2cx(bg_color); ++ ++ while (count--) { ++ c = scr_readw(s++); ++ src = vc->vc_font.data + (c & charmask) * vc->vc_font.height * ++ ((vc->vc_font.width + 7) >> 3); ++ ++ fbcon_decor_renderc(info, yy, xx, vc->vc_font.height, ++ vc->vc_font.width, src, fgx, bgx, transparent); ++ xx += vc->vc_font.width; ++ } ++} ++ ++void fbcon_decor_cursor(struct fb_info *info, struct fb_cursor *cursor) ++{ ++ int i; ++ unsigned int dsize, s_pitch; ++ struct fbcon_ops *ops = info->fbcon_par; ++ struct vc_data* vc; ++ u8 *src; ++ ++ /* we really don't need any cursors while the console is blanked */ ++ if (info->state != FBINFO_STATE_RUNNING || ops->blank_state) ++ return; ++ ++ vc = vc_cons[ops->currcon].d; ++ ++ src = kmalloc(64 + sizeof(struct fb_image), GFP_ATOMIC); ++ if (!src) ++ return; ++ ++ s_pitch = (cursor->image.width + 7) >> 3; ++ dsize = s_pitch * cursor->image.height; ++ if (cursor->enable) { ++ switch (cursor->rop) { ++ case ROP_XOR: ++ for (i = 0; i < dsize; i++) ++ src[i] = cursor->image.data[i] ^ cursor->mask[i]; ++ break; ++ case ROP_COPY: ++ default: ++ for (i = 0; i < dsize; i++) ++ src[i] = cursor->image.data[i] & cursor->mask[i]; ++ break; ++ } ++ } else ++ memcpy(src, cursor->image.data, dsize); ++ ++ fbcon_decor_renderc(info, ++ cursor->image.dy + vc->vc_decor.ty, ++ cursor->image.dx + vc->vc_decor.tx, ++ cursor->image.height, ++ cursor->image.width, ++ (u8*)src, ++ cc2cx(cursor->image.fg_color), ++ cc2cx(cursor->image.bg_color), ++ cursor->image.bg_color == vc->vc_decor.bg_color); ++ ++ kfree(src); ++} ++ ++static void decorset(u8 *dst, int height, int width, int dstbytes, ++ u32 bgx, int bpp) ++{ ++ int i; ++ ++ if (bpp == 8) ++ bgx |= bgx << 8; ++ if (bpp == 16 || bpp == 8) ++ bgx |= bgx << 16; ++ ++ while (height-- > 0) { ++ u8 *p = dst; ++ ++ switch (bpp) { ++ ++ case 32: ++ for (i=0; i < width; i++) { ++ fb_writel(bgx, p); p += 4; ++ } ++ break; ++ case 24: ++ for (i=0; i < width; i++) { ++#ifdef __LITTLE_ENDIAN ++ fb_writew((bgx & 0xffff),(u16*)p); p += 2; ++ fb_writeb((bgx >> 16),p++); ++#else ++ fb_writew((bgx >> 8),(u16*)p); p += 2; ++ fb_writeb((bgx & 0xff),p++); ++#endif ++ } ++ case 16: ++ for (i=0; i < width/4; i++) { ++ fb_writel(bgx,p); p += 4; ++ fb_writel(bgx,p); p += 4; ++ } ++ if (width & 2) { ++ fb_writel(bgx,p); p += 4; ++ } ++ if (width & 1) ++ fb_writew(bgx,(u16*)p); ++ break; ++ case 8: ++ for (i=0; i < width/4; i++) { ++ fb_writel(bgx,p); p += 4; ++ } ++ ++ if (width & 2) { ++ fb_writew(bgx,p); p += 2; ++ } ++ if (width & 1) ++ fb_writeb(bgx,(u8*)p); ++ break; ++ ++ } ++ dst += dstbytes; ++ } ++} ++ ++void fbcon_decor_copy(u8 *dst, u8 *src, int height, int width, int linebytes, ++ int srclinebytes, int bpp) ++{ ++ int i; ++ ++ while (height-- > 0) { ++ u32 *p = (u32 *)dst; ++ u32 *q = (u32 *)src; ++ ++ switch (bpp) { ++ ++ case 32: ++ for (i=0; i < width; i++) ++ fb_writel(*q++, p++); ++ break; ++ case 24: ++ for (i=0; i < (width*3/4); i++) ++ fb_writel(*q++, p++); ++ if ((width*3) % 4) { ++ if (width & 2) { ++ fb_writeb(*(u8*)q, (u8*)p); ++ } else if (width & 1) { ++ fb_writew(*(u16*)q, (u16*)p); ++ fb_writeb(*(u8*)((u16*)q+1),(u8*)((u16*)p+2)); ++ } ++ } ++ break; ++ case 16: ++ for (i=0; i < width/4; i++) { ++ fb_writel(*q++, p++); ++ fb_writel(*q++, p++); ++ } ++ if (width & 2) ++ fb_writel(*q++, p++); ++ if (width & 1) ++ fb_writew(*(u16*)q, (u16*)p); ++ break; ++ case 8: ++ for (i=0; i < width/4; i++) ++ fb_writel(*q++, p++); ++ ++ if (width & 2) { ++ fb_writew(*(u16*)q, (u16*)p); ++ q = (u32*) ((u16*)q + 1); ++ p = (u32*) ((u16*)p + 1); ++ } ++ if (width & 1) ++ fb_writeb(*(u8*)q, (u8*)p); ++ break; ++ } ++ ++ dst += linebytes; ++ src += srclinebytes; ++ } ++} ++ ++static void decorfill(struct fb_info *info, int sy, int sx, int height, ++ int width) ++{ ++ int bytespp = ((info->var.bits_per_pixel + 7) >> 3); ++ int d = sy * info->fix.line_length + sx * bytespp; ++ int ds = (sy * info->var.xres + sx) * bytespp; ++ ++ fbcon_decor_copy((u8 *)(info->screen_base + d), (u8 *)(info->bgdecor.data + ds), ++ height, width, info->fix.line_length, info->var.xres * bytespp, ++ info->var.bits_per_pixel); ++} ++ ++void fbcon_decor_clear(struct vc_data *vc, struct fb_info *info, int sy, int sx, ++ int height, int width) ++{ ++ int bgshift = (vc->vc_hi_font_mask) ? 13 : 12; ++ struct fbcon_ops *ops = info->fbcon_par; ++ u8 *dst; ++ int transparent, bg_color = attr_bgcol_ec(bgshift, vc, info); ++ ++ transparent = (vc->vc_decor.bg_color == bg_color); ++ sy = sy * vc->vc_font.height + vc->vc_decor.ty; ++ sx = sx * vc->vc_font.width + vc->vc_decor.tx; ++ height *= vc->vc_font.height; ++ width *= vc->vc_font.width; ++ ++ /* Don't paint the background image if console is blanked */ ++ if (transparent && !ops->blank_state) { ++ decorfill(info, sy, sx, height, width); ++ } else { ++ dst = (u8 *)(info->screen_base + sy * info->fix.line_length + ++ sx * ((info->var.bits_per_pixel + 7) >> 3)); ++ decorset(dst, height, width, info->fix.line_length, cc2cx(bg_color), ++ info->var.bits_per_pixel); ++ } ++} ++ ++void fbcon_decor_clear_margins(struct vc_data *vc, struct fb_info *info, ++ int bottom_only) ++{ ++ unsigned int tw = vc->vc_cols*vc->vc_font.width; ++ unsigned int th = vc->vc_rows*vc->vc_font.height; ++ ++ if (!bottom_only) { ++ /* top margin */ ++ decorfill(info, 0, 0, vc->vc_decor.ty, info->var.xres); ++ /* left margin */ ++ decorfill(info, vc->vc_decor.ty, 0, th, vc->vc_decor.tx); ++ /* right margin */ ++ decorfill(info, vc->vc_decor.ty, vc->vc_decor.tx + tw, th, ++ info->var.xres - vc->vc_decor.tx - tw); ++ } ++ decorfill(info, vc->vc_decor.ty + th, 0, ++ info->var.yres - vc->vc_decor.ty - th, info->var.xres); ++} ++ ++void fbcon_decor_bmove_redraw(struct vc_data *vc, struct fb_info *info, int y, ++ int sx, int dx, int width) ++{ ++ u16 *d = (u16 *) (vc->vc_origin + vc->vc_size_row * y + dx * 2); ++ u16 *s = d + (dx - sx); ++ u16 *start = d; ++ u16 *ls = d; ++ u16 *le = d + width; ++ u16 c; ++ int x = dx; ++ u16 attr = 1; ++ ++ do { ++ c = scr_readw(d); ++ if (attr != (c & 0xff00)) { ++ attr = c & 0xff00; ++ if (d > start) { ++ fbcon_decor_putcs(vc, info, start, d - start, y, x); ++ x += d - start; ++ start = d; ++ } ++ } ++ if (s >= ls && s < le && c == scr_readw(s)) { ++ if (d > start) { ++ fbcon_decor_putcs(vc, info, start, d - start, y, x); ++ x += d - start + 1; ++ start = d + 1; ++ } else { ++ x++; ++ start++; ++ } ++ } ++ s++; ++ d++; ++ } while (d < le); ++ if (d > start) ++ fbcon_decor_putcs(vc, info, start, d - start, y, x); ++} ++ ++void fbcon_decor_blank(struct vc_data *vc, struct fb_info *info, int blank) ++{ ++ if (blank) { ++ decorset((u8 *)info->screen_base, info->var.yres, info->var.xres, ++ info->fix.line_length, 0, info->var.bits_per_pixel); ++ } else { ++ update_screen(vc); ++ fbcon_decor_clear_margins(vc, info, 0); ++ } ++} ++ +diff -Nur linux-3.13.7.orig/drivers/video/console/fbcon.c linux-3.13.7/drivers/video/console/fbcon.c +--- linux-3.13.7.orig/drivers/video/console/fbcon.c 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/fbcon.c 2014-03-29 16:03:17.000000000 +0100 +@@ -79,6 +79,7 @@ + #include <asm/irq.h> + + #include "fbcon.h" ++#include "fbcondecor.h" + + #ifdef FBCONDEBUG + # define DPRINTK(fmt, args...) printk(KERN_DEBUG "%s: " fmt, __func__ , ## args) +@@ -94,7 +95,7 @@ + + static struct display fb_display[MAX_NR_CONSOLES]; + +-static signed char con2fb_map[MAX_NR_CONSOLES]; ++signed char con2fb_map[MAX_NR_CONSOLES]; + static signed char con2fb_map_boot[MAX_NR_CONSOLES]; + + static int logo_lines; +@@ -286,7 +287,7 @@ + !vt_force_oops_output(vc); + } + +-static int get_color(struct vc_data *vc, struct fb_info *info, ++int get_color(struct vc_data *vc, struct fb_info *info, + u16 c, int is_fg) + { + int depth = fb_get_color_depth(&info->var, &info->fix); +@@ -551,6 +552,9 @@ + info_idx = -1; + } else { + fbcon_has_console_bind = 1; ++#ifdef CONFIG_FB_CON_DECOR ++ fbcon_decor_init(); ++#endif + } + + return err; +@@ -1007,6 +1011,12 @@ + rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres); + cols /= vc->vc_font.width; + rows /= vc->vc_font.height; ++ ++ if (fbcon_decor_active(info, vc)) { ++ cols = vc->vc_decor.twidth / vc->vc_font.width; ++ rows = vc->vc_decor.theight / vc->vc_font.height; ++ } ++ + vc_resize(vc, cols, rows); + + DPRINTK("mode: %s\n", info->fix.id); +@@ -1036,7 +1046,7 @@ + cap = info->flags; + + if (vc != svc || logo_shown == FBCON_LOGO_DONTSHOW || +- (info->fix.type == FB_TYPE_TEXT)) ++ (info->fix.type == FB_TYPE_TEXT) || fbcon_decor_active(info, vc)) + logo = 0; + + if (var_to_display(p, &info->var, info)) +@@ -1260,6 +1270,11 @@ + fbcon_clear_margins(vc, 0); + } + ++ if (fbcon_decor_active(info, vc)) { ++ fbcon_decor_clear(vc, info, sy, sx, height, width); ++ return; ++ } ++ + /* Split blits that cross physical y_wrap boundary */ + + y_break = p->vrows - p->yscroll; +@@ -1279,10 +1294,15 @@ + struct display *p = &fb_display[vc->vc_num]; + struct fbcon_ops *ops = info->fbcon_par; + +- if (!fbcon_is_inactive(vc, info)) +- ops->putcs(vc, info, s, count, real_y(p, ypos), xpos, +- get_color(vc, info, scr_readw(s), 1), +- get_color(vc, info, scr_readw(s), 0)); ++ if (!fbcon_is_inactive(vc, info)) { ++ ++ if (fbcon_decor_active(info, vc)) ++ fbcon_decor_putcs(vc, info, s, count, ypos, xpos); ++ else ++ ops->putcs(vc, info, s, count, real_y(p, ypos), xpos, ++ get_color(vc, info, scr_readw(s), 1), ++ get_color(vc, info, scr_readw(s), 0)); ++ } + } + + static void fbcon_putc(struct vc_data *vc, int c, int ypos, int xpos) +@@ -1298,8 +1318,13 @@ + struct fb_info *info = registered_fb[con2fb_map[vc->vc_num]]; + struct fbcon_ops *ops = info->fbcon_par; + +- if (!fbcon_is_inactive(vc, info)) +- ops->clear_margins(vc, info, bottom_only); ++ if (!fbcon_is_inactive(vc, info)) { ++ if (fbcon_decor_active(info, vc)) { ++ fbcon_decor_clear_margins(vc, info, bottom_only); ++ } else { ++ ops->clear_margins(vc, info, bottom_only); ++ } ++ } + } + + static void fbcon_cursor(struct vc_data *vc, int mode) +@@ -1819,7 +1844,7 @@ + count = vc->vc_rows; + if (softback_top) + fbcon_softback_note(vc, t, count); +- if (logo_shown >= 0) ++ if (logo_shown >= 0 || fbcon_decor_active(info, vc)) + goto redraw_up; + switch (p->scrollmode) { + case SCROLL_MOVE: +@@ -1912,6 +1937,8 @@ + count = vc->vc_rows; + if (logo_shown >= 0) + goto redraw_down; ++ if (fbcon_decor_active(info, vc)) ++ goto redraw_down; + switch (p->scrollmode) { + case SCROLL_MOVE: + fbcon_redraw_blit(vc, info, p, b - 1, b - t - count, +@@ -2060,6 +2087,13 @@ + } + return; + } ++ ++ if (fbcon_decor_active(info, vc) && sy == dy && height == 1) { ++ /* must use slower redraw bmove to keep background pic intact */ ++ fbcon_decor_bmove_redraw(vc, info, sy, sx, dx, width); ++ return; ++ } ++ + ops->bmove(vc, info, real_y(p, sy), sx, real_y(p, dy), dx, + height, width); + } +@@ -2130,8 +2164,8 @@ + var.yres = virt_h * virt_fh; + x_diff = info->var.xres - var.xres; + y_diff = info->var.yres - var.yres; +- if (x_diff < 0 || x_diff > virt_fw || +- y_diff < 0 || y_diff > virt_fh) { ++ if ((x_diff < 0 || x_diff > virt_fw || ++ y_diff < 0 || y_diff > virt_fh) && !vc->vc_decor.state) { + const struct fb_videomode *mode; + + DPRINTK("attempting resize %ix%i\n", var.xres, var.yres); +@@ -2167,6 +2201,21 @@ + + info = registered_fb[con2fb_map[vc->vc_num]]; + ops = info->fbcon_par; ++ prev_console = ops->currcon; ++ if (prev_console != -1) ++ old_info = registered_fb[con2fb_map[prev_console]]; ++ ++#ifdef CONFIG_FB_CON_DECOR ++ if (!fbcon_decor_active_vc(vc) && info->fix.visual == FB_VISUAL_DIRECTCOLOR) { ++ struct vc_data *vc_curr = vc_cons[prev_console].d; ++ if (vc_curr && fbcon_decor_active_vc(vc_curr)) { ++ /* Clear the screen to avoid displaying funky colors during ++ * palette updates. */ ++ memset((u8*)info->screen_base + info->fix.line_length * info->var.yoffset, ++ 0, info->var.yres * info->fix.line_length); ++ } ++ } ++#endif + + if (softback_top) { + if (softback_lines) +@@ -2185,9 +2234,6 @@ + logo_shown = FBCON_LOGO_CANSHOW; + } + +- prev_console = ops->currcon; +- if (prev_console != -1) +- old_info = registered_fb[con2fb_map[prev_console]]; + /* + * FIXME: If we have multiple fbdev's loaded, we need to + * update all info->currcon. Perhaps, we can place this +@@ -2231,6 +2277,18 @@ + fbcon_del_cursor_timer(old_info); + } + ++ if (fbcon_decor_active_vc(vc)) { ++ struct vc_data *vc_curr = vc_cons[prev_console].d; ++ ++ if (!vc_curr->vc_decor.theme || ++ strcmp(vc->vc_decor.theme, vc_curr->vc_decor.theme) || ++ (fbcon_decor_active_nores(info, vc_curr) && ++ !fbcon_decor_active(info, vc_curr))) { ++ fbcon_decor_disable(vc, 0); ++ fbcon_decor_call_helper("modechange", vc->vc_num); ++ } ++ } ++ + if (fbcon_is_inactive(vc, info) || + ops->blank_state != FB_BLANK_UNBLANK) + fbcon_del_cursor_timer(info); +@@ -2339,15 +2397,20 @@ + } + } + +- if (!fbcon_is_inactive(vc, info)) { ++ if (!fbcon_is_inactive(vc, info)) { + if (ops->blank_state != blank) { + ops->blank_state = blank; + fbcon_cursor(vc, blank ? CM_ERASE : CM_DRAW); + ops->cursor_flash = (!blank); + +- if (!(info->flags & FBINFO_MISC_USEREVENT)) +- if (fb_blank(info, blank)) +- fbcon_generic_blank(vc, info, blank); ++ if (!(info->flags & FBINFO_MISC_USEREVENT)) { ++ if (fb_blank(info, blank)) { ++ if (fbcon_decor_active(info, vc)) ++ fbcon_decor_blank(vc, info, blank); ++ else ++ fbcon_generic_blank(vc, info, blank); ++ } ++ } + } + + if (!blank) +@@ -2522,13 +2585,22 @@ + } + + if (resize) { ++ /* reset wrap/pan */ + int cols, rows; + + cols = FBCON_SWAP(ops->rotate, info->var.xres, info->var.yres); + rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres); ++ ++ if (fbcon_decor_active(info, vc)) { ++ info->var.xoffset = info->var.yoffset = p->yscroll = 0; ++ cols = vc->vc_decor.twidth; ++ rows = vc->vc_decor.theight; ++ } + cols /= w; + rows /= h; ++ + vc_resize(vc, cols, rows); ++ + if (CON_IS_VISIBLE(vc) && softback_buf) + fbcon_update_softback(vc); + } else if (CON_IS_VISIBLE(vc) +@@ -2657,7 +2729,11 @@ + int i, j, k, depth; + u8 val; + +- if (fbcon_is_inactive(vc, info)) ++ if (fbcon_is_inactive(vc, info) ++#ifdef CONFIG_FB_CON_DECOR ++ || vc->vc_num != fg_console ++#endif ++ ) + return -EINVAL; + + if (!CON_IS_VISIBLE(vc)) +@@ -2683,14 +2759,56 @@ + } else + fb_copy_cmap(fb_default_cmap(1 << depth), &palette_cmap); + +- return fb_set_cmap(&palette_cmap, info); ++ if (fbcon_decor_active(info, vc_cons[fg_console].d) && ++ info->fix.visual == FB_VISUAL_DIRECTCOLOR) { ++ ++ u16 *red, *green, *blue; ++ int minlen = min(min(info->var.red.length, info->var.green.length), ++ info->var.blue.length); ++ int h; ++ ++ struct fb_cmap cmap = { ++ .start = 0, ++ .len = (1 << minlen), ++ .red = NULL, ++ .green = NULL, ++ .blue = NULL, ++ .transp = NULL ++ }; ++ ++ red = kmalloc(256 * sizeof(u16) * 3, GFP_KERNEL); ++ ++ if (!red) ++ goto out; ++ ++ green = red + 256; ++ blue = green + 256; ++ cmap.red = red; ++ cmap.green = green; ++ cmap.blue = blue; ++ ++ for (i = 0; i < cmap.len; i++) { ++ red[i] = green[i] = blue[i] = (0xffff * i)/(cmap.len-1); ++ } ++ ++ h = fb_set_cmap(&cmap, info); ++ fbcon_decor_fix_pseudo_pal(info, vc_cons[fg_console].d); ++ kfree(red); ++ ++ return h; ++ ++ } else if (fbcon_decor_active(info, vc_cons[fg_console].d) && ++ info->var.bits_per_pixel == 8 && info->bgdecor.cmap.red != NULL) ++ fb_set_cmap(&info->bgdecor.cmap, info); ++ ++out: return fb_set_cmap(&palette_cmap, info); + } + + static u16 *fbcon_screen_pos(struct vc_data *vc, int offset) + { + unsigned long p; + int line; +- ++ + if (vc->vc_num != fg_console || !softback_lines) + return (u16 *) (vc->vc_origin + offset); + line = offset / vc->vc_size_row; +@@ -2909,7 +3027,14 @@ + rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres); + cols /= vc->vc_font.width; + rows /= vc->vc_font.height; +- vc_resize(vc, cols, rows); ++ ++ if (!fbcon_decor_active_nores(info, vc)) { ++ vc_resize(vc, cols, rows); ++ } else { ++ fbcon_decor_disable(vc, 0); ++ fbcon_decor_call_helper("modechange", vc->vc_num); ++ } ++ + updatescrollmode(p, info, vc); + scrollback_max = 0; + scrollback_current = 0; +@@ -2954,7 +3079,9 @@ + rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres); + cols /= vc->vc_font.width; + rows /= vc->vc_font.height; +- vc_resize(vc, cols, rows); ++ if (!fbcon_decor_active_nores(info, vc)) { ++ vc_resize(vc, cols, rows); ++ } + } + + if (fg != -1) +@@ -3570,6 +3697,7 @@ + } + } + ++ fbcon_decor_exit(); + fbcon_has_exited = 1; + } + +diff -Nur linux-3.13.7.orig/drivers/video/console/fbcondecor.c linux-3.13.7/drivers/video/console/fbcondecor.c +--- linux-3.13.7.orig/drivers/video/console/fbcondecor.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/fbcondecor.c 2014-03-29 16:03:17.000000000 +0100 +@@ -0,0 +1,555 @@ ++/* ++ * linux/drivers/video/console/fbcondecor.c -- Framebuffer console decorations ++ * ++ * Copyright (C) 2004-2009 Michal Januszewski <michalj+fbcondecor@gmail.com> ++ * ++ * Code based upon "Bootsplash" (C) 2001-2003 ++ * Volker Poplawski <volker@poplawski.de>, ++ * Stefan Reinauer <stepan@suse.de>, ++ * Steffen Winterfeldt <snwint@suse.de>, ++ * Michael Schroeder <mls@suse.de>, ++ * Ken Wimer <wimer@suse.de>. ++ * ++ * Compat ioctl support by Thorsten Klein <TK@Thorsten-Klein.de>. ++ * ++ * This file is subject to the terms and conditions of the GNU General Public ++ * License. See the file COPYING in the main directory of this archive for ++ * more details. ++ * ++ */ ++#include <linux/module.h> ++#include <linux/kernel.h> ++#include <linux/string.h> ++#include <linux/types.h> ++#include <linux/fb.h> ++#include <linux/vt_kern.h> ++#include <linux/vmalloc.h> ++#include <linux/unistd.h> ++#include <linux/syscalls.h> ++#include <linux/init.h> ++#include <linux/proc_fs.h> ++#include <linux/workqueue.h> ++#include <linux/kmod.h> ++#include <linux/miscdevice.h> ++#include <linux/device.h> ++#include <linux/fs.h> ++#include <linux/compat.h> ++#include <linux/console.h> ++ ++#include <asm/uaccess.h> ++#include <asm/irq.h> ++ ++#include "fbcon.h" ++#include "fbcondecor.h" ++ ++extern signed char con2fb_map[]; ++static int fbcon_decor_enable(struct vc_data *vc); ++char fbcon_decor_path[KMOD_PATH_LEN] = "/sbin/fbcondecor_helper"; ++static int initialized = 0; ++ ++int fbcon_decor_call_helper(char* cmd, unsigned short vc) ++{ ++ char *envp[] = { ++ "HOME=/", ++ "PATH=/sbin:/bin", ++ NULL ++ }; ++ ++ char tfb[5]; ++ char tcons[5]; ++ unsigned char fb = (int) con2fb_map[vc]; ++ ++ char *argv[] = { ++ fbcon_decor_path, ++ "2", ++ cmd, ++ tcons, ++ tfb, ++ vc_cons[vc].d->vc_decor.theme, ++ NULL ++ }; ++ ++ snprintf(tfb,5,"%d",fb); ++ snprintf(tcons,5,"%d",vc); ++ ++ return call_usermodehelper(fbcon_decor_path, argv, envp, UMH_WAIT_EXEC); ++} ++ ++/* Disables fbcondecor on a virtual console; called with console sem held. */ ++int fbcon_decor_disable(struct vc_data *vc, unsigned char redraw) ++{ ++ struct fb_info* info; ++ ++ if (!vc->vc_decor.state) ++ return -EINVAL; ++ ++ info = registered_fb[(int) con2fb_map[vc->vc_num]]; ++ ++ if (info == NULL) ++ return -EINVAL; ++ ++ vc->vc_decor.state = 0; ++ vc_resize(vc, info->var.xres / vc->vc_font.width, ++ info->var.yres / vc->vc_font.height); ++ ++ if (fg_console == vc->vc_num && redraw) { ++ redraw_screen(vc, 0); ++ update_region(vc, vc->vc_origin + ++ vc->vc_size_row * vc->vc_top, ++ vc->vc_size_row * (vc->vc_bottom - vc->vc_top) / 2); ++ } ++ ++ printk(KERN_INFO "fbcondecor: switched decor state to 'off' on console %d\n", ++ vc->vc_num); ++ ++ return 0; ++} ++ ++/* Enables fbcondecor on a virtual console; called with console sem held. */ ++static int fbcon_decor_enable(struct vc_data *vc) ++{ ++ struct fb_info* info; ++ ++ info = registered_fb[(int) con2fb_map[vc->vc_num]]; ++ ++ if (vc->vc_decor.twidth == 0 || vc->vc_decor.theight == 0 || ++ info == NULL || vc->vc_decor.state || (!info->bgdecor.data && ++ vc->vc_num == fg_console)) ++ return -EINVAL; ++ ++ vc->vc_decor.state = 1; ++ vc_resize(vc, vc->vc_decor.twidth / vc->vc_font.width, ++ vc->vc_decor.theight / vc->vc_font.height); ++ ++ if (fg_console == vc->vc_num) { ++ redraw_screen(vc, 0); ++ update_region(vc, vc->vc_origin + ++ vc->vc_size_row * vc->vc_top, ++ vc->vc_size_row * (vc->vc_bottom - vc->vc_top) / 2); ++ fbcon_decor_clear_margins(vc, info, 0); ++ } ++ ++ printk(KERN_INFO "fbcondecor: switched decor state to 'on' on console %d\n", ++ vc->vc_num); ++ ++ return 0; ++} ++ ++static inline int fbcon_decor_ioctl_dosetstate(struct vc_data *vc, unsigned int state, unsigned char origin) ++{ ++ int ret; ++ ++// if (origin == FBCON_DECOR_IO_ORIG_USER) ++ console_lock(); ++ if (!state) ++ ret = fbcon_decor_disable(vc, 1); ++ else ++ ret = fbcon_decor_enable(vc); ++// if (origin == FBCON_DECOR_IO_ORIG_USER) ++ console_unlock(); ++ ++ return ret; ++} ++ ++static inline void fbcon_decor_ioctl_dogetstate(struct vc_data *vc, unsigned int *state) ++{ ++ *state = vc->vc_decor.state; ++} ++ ++static int fbcon_decor_ioctl_dosetcfg(struct vc_data *vc, struct vc_decor *cfg, unsigned char origin) ++{ ++ struct fb_info *info; ++ int len; ++ char *tmp; ++ ++ info = registered_fb[(int) con2fb_map[vc->vc_num]]; ++ ++ if (info == NULL || !cfg->twidth || !cfg->theight || ++ cfg->tx + cfg->twidth > info->var.xres || ++ cfg->ty + cfg->theight > info->var.yres) ++ return -EINVAL; ++ ++ len = strlen_user(cfg->theme); ++ if (!len || len > FBCON_DECOR_THEME_LEN) ++ return -EINVAL; ++ tmp = kmalloc(len, GFP_KERNEL); ++ if (!tmp) ++ return -ENOMEM; ++ if (copy_from_user(tmp, (void __user *)cfg->theme, len)) ++ return -EFAULT; ++ cfg->theme = tmp; ++ cfg->state = 0; ++ ++ /* If this ioctl is a response to a request from kernel, the console sem ++ * is already held; we also don't need to disable decor because either the ++ * new config and background picture will be successfully loaded, and the ++ * decor will stay on, or in case of a failure it'll be turned off in fbcon. */ ++// if (origin == FBCON_DECOR_IO_ORIG_USER) { ++ console_lock(); ++ if (vc->vc_decor.state) ++ fbcon_decor_disable(vc, 1); ++// } ++ ++ if (vc->vc_decor.theme) ++ kfree(vc->vc_decor.theme); ++ ++ vc->vc_decor = *cfg; ++ ++// if (origin == FBCON_DECOR_IO_ORIG_USER) ++ console_unlock(); ++ ++ printk(KERN_INFO "fbcondecor: console %d using theme '%s'\n", ++ vc->vc_num, vc->vc_decor.theme); ++ return 0; ++} ++ ++static int fbcon_decor_ioctl_dogetcfg(struct vc_data *vc, struct vc_decor *decor) ++{ ++ char __user *tmp; ++ ++ tmp = decor->theme; ++ *decor = vc->vc_decor; ++ decor->theme = tmp; ++ ++ if (vc->vc_decor.theme) { ++ if (copy_to_user(tmp, vc->vc_decor.theme, strlen(vc->vc_decor.theme) + 1)) ++ return -EFAULT; ++ } else ++ if (put_user(0, tmp)) ++ return -EFAULT; ++ ++ return 0; ++} ++ ++static int fbcon_decor_ioctl_dosetpic(struct vc_data *vc, struct fb_image *img, unsigned char origin) ++{ ++ struct fb_info *info; ++ int len; ++ u8 *tmp; ++ ++ if (vc->vc_num != fg_console) ++ return -EINVAL; ++ ++ info = registered_fb[(int) con2fb_map[vc->vc_num]]; ++ ++ if (info == NULL) ++ return -EINVAL; ++ ++ if (img->width != info->var.xres || img->height != info->var.yres) { ++ printk(KERN_ERR "fbcondecor: picture dimensions mismatch\n"); ++ printk(KERN_ERR "%dx%d vs %dx%d\n", img->width, img->height, info->var.xres, info->var.yres); ++ return -EINVAL; ++ } ++ ++ if (img->depth != info->var.bits_per_pixel) { ++ printk(KERN_ERR "fbcondecor: picture depth mismatch\n"); ++ return -EINVAL; ++ } ++ ++ if (img->depth == 8) { ++ if (!img->cmap.len || !img->cmap.red || !img->cmap.green || ++ !img->cmap.blue) ++ return -EINVAL; ++ ++ tmp = vmalloc(img->cmap.len * 3 * 2); ++ if (!tmp) ++ return -ENOMEM; ++ ++ if (copy_from_user(tmp, ++ (void __user*)img->cmap.red, (img->cmap.len << 1)) || ++ copy_from_user(tmp + (img->cmap.len << 1), ++ (void __user*)img->cmap.green, (img->cmap.len << 1)) || ++ copy_from_user(tmp + (img->cmap.len << 2), ++ (void __user*)img->cmap.blue, (img->cmap.len << 1))) { ++ vfree(tmp); ++ return -EFAULT; ++ } ++ ++ img->cmap.transp = NULL; ++ img->cmap.red = (u16*)tmp; ++ img->cmap.green = img->cmap.red + img->cmap.len; ++ img->cmap.blue = img->cmap.green + img->cmap.len; ++ } else { ++ img->cmap.red = NULL; ++ } ++ ++ len = ((img->depth + 7) >> 3) * img->width * img->height; ++ ++ /* ++ * Allocate an additional byte so that we never go outside of the ++ * buffer boundaries in the rendering functions in a 24 bpp mode. ++ */ ++ tmp = vmalloc(len + 1); ++ ++ if (!tmp) ++ goto out; ++ ++ if (copy_from_user(tmp, (void __user*)img->data, len)) ++ goto out; ++ ++ img->data = tmp; ++ ++ /* If this ioctl is a response to a request from kernel, the console sem ++ * is already held. */ ++// if (origin == FBCON_DECOR_IO_ORIG_USER) ++ console_lock(); ++ ++ if (info->bgdecor.data) ++ vfree((u8*)info->bgdecor.data); ++ if (info->bgdecor.cmap.red) ++ vfree(info->bgdecor.cmap.red); ++ ++ info->bgdecor = *img; ++ ++ if (fbcon_decor_active_vc(vc) && fg_console == vc->vc_num) { ++ redraw_screen(vc, 0); ++ update_region(vc, vc->vc_origin + ++ vc->vc_size_row * vc->vc_top, ++ vc->vc_size_row * (vc->vc_bottom - vc->vc_top) / 2); ++ fbcon_decor_clear_margins(vc, info, 0); ++ } ++ ++// if (origin == FBCON_DECOR_IO_ORIG_USER) ++ console_unlock(); ++ ++ return 0; ++ ++out: if (img->cmap.red) ++ vfree(img->cmap.red); ++ ++ if (tmp) ++ vfree(tmp); ++ return -ENOMEM; ++} ++ ++static long fbcon_decor_ioctl(struct file *filp, u_int cmd, u_long arg) ++{ ++ struct fbcon_decor_iowrapper __user *wrapper = (void __user*) arg; ++ struct vc_data *vc = NULL; ++ unsigned short vc_num = 0; ++ unsigned char origin = 0; ++ void __user *data = NULL; ++ ++ if (!access_ok(VERIFY_READ, wrapper, ++ sizeof(struct fbcon_decor_iowrapper))) ++ return -EFAULT; ++ ++ __get_user(vc_num, &wrapper->vc); ++ __get_user(origin, &wrapper->origin); ++ __get_user(data, &wrapper->data); ++ ++ if (!vc_cons_allocated(vc_num)) ++ return -EINVAL; ++ ++ vc = vc_cons[vc_num].d; ++ ++ switch (cmd) { ++ case FBIOCONDECOR_SETPIC: ++ { ++ struct fb_image img; ++ if (copy_from_user(&img, (struct fb_image __user *)data, sizeof(struct fb_image))) ++ return -EFAULT; ++ ++ return fbcon_decor_ioctl_dosetpic(vc, &img, origin); ++ } ++ case FBIOCONDECOR_SETCFG: ++ { ++ struct vc_decor cfg; ++ if (copy_from_user(&cfg, (struct vc_decor __user *)data, sizeof(struct vc_decor))) ++ return -EFAULT; ++ ++ return fbcon_decor_ioctl_dosetcfg(vc, &cfg, origin); ++ } ++ case FBIOCONDECOR_GETCFG: ++ { ++ int rval; ++ struct vc_decor cfg; ++ ++ if (copy_from_user(&cfg, (struct vc_decor __user *)data, sizeof(struct vc_decor))) ++ return -EFAULT; ++ ++ rval = fbcon_decor_ioctl_dogetcfg(vc, &cfg); ++ ++ if (copy_to_user(data, &cfg, sizeof(struct vc_decor))) ++ return -EFAULT; ++ return rval; ++ } ++ case FBIOCONDECOR_SETSTATE: ++ { ++ unsigned int state = 0; ++ if (get_user(state, (unsigned int __user *)data)) ++ return -EFAULT; ++ return fbcon_decor_ioctl_dosetstate(vc, state, origin); ++ } ++ case FBIOCONDECOR_GETSTATE: ++ { ++ unsigned int state = 0; ++ fbcon_decor_ioctl_dogetstate(vc, &state); ++ return put_user(state, (unsigned int __user *)data); ++ } ++ ++ default: ++ return -ENOIOCTLCMD; ++ } ++} ++ ++#ifdef CONFIG_COMPAT ++ ++static long fbcon_decor_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { ++ ++ struct fbcon_decor_iowrapper32 __user *wrapper = (void __user *)arg; ++ struct vc_data *vc = NULL; ++ unsigned short vc_num = 0; ++ unsigned char origin = 0; ++ compat_uptr_t data_compat = 0; ++ void __user *data = NULL; ++ ++ if (!access_ok(VERIFY_READ, wrapper, ++ sizeof(struct fbcon_decor_iowrapper32))) ++ return -EFAULT; ++ ++ __get_user(vc_num, &wrapper->vc); ++ __get_user(origin, &wrapper->origin); ++ __get_user(data_compat, &wrapper->data); ++ data = compat_ptr(data_compat); ++ ++ if (!vc_cons_allocated(vc_num)) ++ return -EINVAL; ++ ++ vc = vc_cons[vc_num].d; ++ ++ switch (cmd) { ++ case FBIOCONDECOR_SETPIC32: ++ { ++ struct fb_image32 img_compat; ++ struct fb_image img; ++ ++ if (copy_from_user(&img_compat, (struct fb_image32 __user *)data, sizeof(struct fb_image32))) ++ return -EFAULT; ++ ++ fb_image_from_compat(img, img_compat); ++ ++ return fbcon_decor_ioctl_dosetpic(vc, &img, origin); ++ } ++ ++ case FBIOCONDECOR_SETCFG32: ++ { ++ struct vc_decor32 cfg_compat; ++ struct vc_decor cfg; ++ ++ if (copy_from_user(&cfg_compat, (struct vc_decor32 __user *)data, sizeof(struct vc_decor32))) ++ return -EFAULT; ++ ++ vc_decor_from_compat(cfg, cfg_compat); ++ ++ return fbcon_decor_ioctl_dosetcfg(vc, &cfg, origin); ++ } ++ ++ case FBIOCONDECOR_GETCFG32: ++ { ++ int rval; ++ struct vc_decor32 cfg_compat; ++ struct vc_decor cfg; ++ ++ if (copy_from_user(&cfg_compat, (struct vc_decor32 __user *)data, sizeof(struct vc_decor32))) ++ return -EFAULT; ++ cfg.theme = compat_ptr(cfg_compat.theme); ++ ++ rval = fbcon_decor_ioctl_dogetcfg(vc, &cfg); ++ ++ vc_decor_to_compat(cfg_compat, cfg); ++ ++ if (copy_to_user((struct vc_decor32 __user *)data, &cfg_compat, sizeof(struct vc_decor32))) ++ return -EFAULT; ++ return rval; ++ } ++ ++ case FBIOCONDECOR_SETSTATE32: ++ { ++ compat_uint_t state_compat = 0; ++ unsigned int state = 0; ++ ++ if (get_user(state_compat, (compat_uint_t __user *)data)) ++ return -EFAULT; ++ ++ state = (unsigned int)state_compat; ++ ++ return fbcon_decor_ioctl_dosetstate(vc, state, origin); ++ } ++ ++ case FBIOCONDECOR_GETSTATE32: ++ { ++ compat_uint_t state_compat = 0; ++ unsigned int state = 0; ++ ++ fbcon_decor_ioctl_dogetstate(vc, &state); ++ state_compat = (compat_uint_t)state; ++ ++ return put_user(state_compat, (compat_uint_t __user *)data); ++ } ++ ++ default: ++ return -ENOIOCTLCMD; ++ } ++} ++#else ++ #define fbcon_decor_compat_ioctl NULL ++#endif ++ ++static struct file_operations fbcon_decor_ops = { ++ .owner = THIS_MODULE, ++ .unlocked_ioctl = fbcon_decor_ioctl, ++ .compat_ioctl = fbcon_decor_compat_ioctl ++}; ++ ++static struct miscdevice fbcon_decor_dev = { ++ .minor = MISC_DYNAMIC_MINOR, ++ .name = "fbcondecor", ++ .fops = &fbcon_decor_ops ++}; ++ ++void fbcon_decor_reset() ++{ ++ int i; ++ ++ for (i = 0; i < num_registered_fb; i++) { ++ registered_fb[i]->bgdecor.data = NULL; ++ registered_fb[i]->bgdecor.cmap.red = NULL; ++ } ++ ++ for (i = 0; i < MAX_NR_CONSOLES && vc_cons[i].d; i++) { ++ vc_cons[i].d->vc_decor.state = vc_cons[i].d->vc_decor.twidth = ++ vc_cons[i].d->vc_decor.theight = 0; ++ vc_cons[i].d->vc_decor.theme = NULL; ++ } ++ ++ return; ++} ++ ++int fbcon_decor_init() ++{ ++ int i; ++ ++ fbcon_decor_reset(); ++ ++ if (initialized) ++ return 0; ++ ++ i = misc_register(&fbcon_decor_dev); ++ if (i) { ++ printk(KERN_ERR "fbcondecor: failed to register device\n"); ++ return i; ++ } ++ ++ fbcon_decor_call_helper("init", 0); ++ initialized = 1; ++ return 0; ++} ++ ++int fbcon_decor_exit(void) ++{ ++ fbcon_decor_reset(); ++ return 0; ++} ++ ++EXPORT_SYMBOL(fbcon_decor_path); +diff -Nur linux-3.13.7.orig/drivers/video/console/fbcondecor.h linux-3.13.7/drivers/video/console/fbcondecor.h +--- linux-3.13.7.orig/drivers/video/console/fbcondecor.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.7/drivers/video/console/fbcondecor.h 2014-03-29 21:17:27.000000000 +0100 +@@ -0,0 +1,79 @@ ++/* ++ * linux/drivers/video/console/fbcondecor.h -- Framebuffer Console Decoration headers ++ * ++ * Copyright (C) 2004 Michal Januszewski <michalj+fbcondecor@gmail.com> ++ * ++ */ ++ ++#ifndef __FBCON_DECOR_H ++#define __FBCON_DECOR_H ++ ++#ifndef _LINUX_FB_H ++#include <linux/fb.h> ++#endif ++ ++/* This is needed for vc_cons in fbcmap.c */ ++#include <linux/vt_kern.h> ++ ++struct fb_cursor; ++struct fb_info; ++struct vc_data; ++ ++#ifdef CONFIG_FB_CON_DECOR ++/* fbcondecor.c */ ++int fbcon_decor_init(void); ++void fbcon_decor_reset(void); ++int fbcon_decor_exit(void); ++int fbcon_decor_call_helper(char* cmd, unsigned short cons); ++int fbcon_decor_disable(struct vc_data *vc, unsigned char redraw); ++ ++/* cfbcondecor.c */ ++void fbcon_decor_putcs(struct vc_data *vc, struct fb_info *info, const unsigned short *s, int count, int yy, int xx); ++void fbcon_decor_cursor(struct fb_info *info, struct fb_cursor *cursor); ++void fbcon_decor_clear(struct vc_data *vc, struct fb_info *info, int sy, int sx, int height, int width); ++void fbcon_decor_clear_margins(struct vc_data *vc, struct fb_info *info, int bottom_only); ++void fbcon_decor_blank(struct vc_data *vc, struct fb_info *info, int blank); ++void fbcon_decor_bmove_redraw(struct vc_data *vc, struct fb_info *info, int y, int sx, int dx, int width); ++void fbcon_decor_copy(u8 *dst, u8 *src, int height, int width, int linebytes, int srclinesbytes, int bpp); ++void fbcon_decor_fix_pseudo_pal(struct fb_info *info, struct vc_data *vc); ++ ++/* vt.c */ ++void acquire_console_sem(void); ++void release_console_sem(void); ++void do_unblank_screen(int entering_gfx); ++ ++/* struct vc_data *y */ ++#define fbcon_decor_active_vc(y) (y->vc_decor.state && y->vc_decor.theme) ++ ++/* struct fb_info *x, struct vc_data *y */ ++#define fbcon_decor_active_nores(x,y) (x->bgdecor.data && fbcon_decor_active_vc(y)) ++ ++/* struct fb_info *x, struct vc_data *y */ ++#define fbcon_decor_active(x,y) (fbcon_decor_active_nores(x,y) && \ ++ x->bgdecor.width == x->var.xres && \ ++ x->bgdecor.height == x->var.yres && \ ++ x->bgdecor.depth == x->var.bits_per_pixel) ++ ++ ++#else /* CONFIG_FB_CON_DECOR */ ++ ++static inline void fbcon_decor_putcs(struct vc_data *vc, struct fb_info *info, const unsigned short *s, int count, int yy, int xx) {} ++static inline void fbcon_decor_putc(struct vc_data *vc, struct fb_info *info, int c, int ypos, int xpos) {} ++static inline void fbcon_decor_cursor(struct fb_info *info, struct fb_cursor *cursor) {} ++static inline void fbcon_decor_clear(struct vc_data *vc, struct fb_info *info, int sy, int sx, int height, int width) {} ++static inline void fbcon_decor_clear_margins(struct vc_data *vc, struct fb_info *info, int bottom_only) {} ++static inline void fbcon_decor_blank(struct vc_data *vc, struct fb_info *info, int blank) {} ++static inline void fbcon_decor_bmove_redraw(struct vc_data *vc, struct fb_info *info, int y, int sx, int dx, int width) {} ++static inline void fbcon_decor_fix_pseudo_pal(struct fb_info *info, struct vc_data *vc) {} ++static inline int fbcon_decor_call_helper(char* cmd, unsigned short cons) { return 0; } ++static inline int fbcon_decor_init(void) { return 0; } ++static inline int fbcon_decor_exit(void) { return 0; } ++static inline int fbcon_decor_disable(struct vc_data *vc, unsigned char redraw) { return 0; } ++ ++#define fbcon_decor_active_vc(y) (0) ++#define fbcon_decor_active_nores(x,y) (0) ++#define fbcon_decor_active(x,y) (0) ++ ++#endif /* CONFIG_FB_CON_DECOR */ ++ ++#endif /* __FBCON_DECOR_H */ +diff -Nur linux-3.13.7.orig/drivers/video/fbcmap.c linux-3.13.7/drivers/video/fbcmap.c +--- linux-3.13.7.orig/drivers/video/fbcmap.c 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/fbcmap.c 2014-03-29 16:03:17.000000000 +0100 +@@ -17,6 +17,8 @@ + #include <linux/slab.h> + #include <linux/uaccess.h> + ++#include "console/fbcondecor.h" ++ + static u16 red2[] __read_mostly = { + 0x0000, 0xaaaa + }; +@@ -249,14 +251,17 @@ + if (transp) + htransp = *transp++; + if (info->fbops->fb_setcolreg(start++, +- hred, hgreen, hblue, ++ hred, hgreen, hblue, + htransp, info)) + break; + } + } +- if (rc == 0) ++ if (rc == 0) { + fb_copy_cmap(cmap, &info->cmap); +- ++ if (fbcon_decor_active(info, vc_cons[fg_console].d) && ++ info->fix.visual == FB_VISUAL_DIRECTCOLOR) ++ fbcon_decor_fix_pseudo_pal(info, vc_cons[fg_console].d); ++ } + return rc; + } + +diff -Nur linux-3.13.7.orig/drivers/video/fbmem.c linux-3.13.7/drivers/video/fbmem.c +--- linux-3.13.7.orig/drivers/video/fbmem.c 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/drivers/video/fbmem.c 2014-03-29 16:03:17.000000000 +0100 +@@ -1250,15 +1250,6 @@ + u16 reserved[3]; + }; + +-struct fb_cmap32 { +- u32 start; +- u32 len; +- compat_caddr_t red; +- compat_caddr_t green; +- compat_caddr_t blue; +- compat_caddr_t transp; +-}; +- + static int fb_getput_cmap(struct fb_info *info, unsigned int cmd, + unsigned long arg) + { +diff -Nur linux-3.13.7.orig/include/linux/console_decor.h linux-3.13.7/include/linux/console_decor.h +--- linux-3.13.7.orig/include/linux/console_decor.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.7/include/linux/console_decor.h 2014-03-29 16:03:17.000000000 +0100 +@@ -0,0 +1,46 @@ ++#ifndef _LINUX_CONSOLE_DECOR_H_ ++#define _LINUX_CONSOLE_DECOR_H_ 1 ++ ++/* A structure used by the framebuffer console decorations (drivers/video/console/fbcondecor.c) */ ++struct vc_decor { ++ __u8 bg_color; /* The color that is to be treated as transparent */ ++ __u8 state; /* Current decor state: 0 = off, 1 = on */ ++ __u16 tx, ty; /* Top left corner coordinates of the text field */ ++ __u16 twidth, theight; /* Width and height of the text field */ ++ char* theme; ++}; ++ ++#ifdef __KERNEL__ ++#ifdef CONFIG_COMPAT ++#include <linux/compat.h> ++ ++struct vc_decor32 { ++ __u8 bg_color; /* The color that is to be treated as transparent */ ++ __u8 state; /* Current decor state: 0 = off, 1 = on */ ++ __u16 tx, ty; /* Top left corner coordinates of the text field */ ++ __u16 twidth, theight; /* Width and height of the text field */ ++ compat_uptr_t theme; ++}; ++ ++#define vc_decor_from_compat(to, from) \ ++ (to).bg_color = (from).bg_color; \ ++ (to).state = (from).state; \ ++ (to).tx = (from).tx; \ ++ (to).ty = (from).ty; \ ++ (to).twidth = (from).twidth; \ ++ (to).theight = (from).theight; \ ++ (to).theme = compat_ptr((from).theme) ++ ++#define vc_decor_to_compat(to, from) \ ++ (to).bg_color = (from).bg_color; \ ++ (to).state = (from).state; \ ++ (to).tx = (from).tx; \ ++ (to).ty = (from).ty; \ ++ (to).twidth = (from).twidth; \ ++ (to).theight = (from).theight; \ ++ (to).theme = ptr_to_compat((from).theme) ++ ++#endif /* CONFIG_COMPAT */ ++#endif /* __KERNEL__ */ ++ ++#endif +diff -Nur linux-3.13.7.orig/include/linux/console_struct.h linux-3.13.7/include/linux/console_struct.h +--- linux-3.13.7.orig/include/linux/console_struct.h 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/include/linux/console_struct.h 2014-03-29 16:03:17.000000000 +0100 +@@ -19,6 +19,7 @@ + struct vt_struct; + + #define NPAR 16 ++#include <linux/console_decor.h> + + struct vc_data { + struct tty_port port; /* Upper level data */ +@@ -107,6 +108,8 @@ + unsigned long vc_uni_pagedir; + unsigned long *vc_uni_pagedir_loc; /* [!] Location of uni_pagedir variable for this console */ + bool vc_panic_force_write; /* when oops/panic this VC can accept forced output/blanking */ ++ ++ struct vc_decor vc_decor; + /* additional information is in vt_kern.h */ + }; + +diff -Nur linux-3.13.7.orig/include/linux/fb.h linux-3.13.7/include/linux/fb.h +--- linux-3.13.7.orig/include/linux/fb.h 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/include/linux/fb.h 2014-03-29 16:03:17.000000000 +0100 +@@ -219,6 +219,34 @@ + }; + #endif + ++#ifdef __KERNEL__ ++#ifdef CONFIG_COMPAT ++struct fb_image32 { ++ __u32 dx; /* Where to place image */ ++ __u32 dy; ++ __u32 width; /* Size of image */ ++ __u32 height; ++ __u32 fg_color; /* Only used when a mono bitmap */ ++ __u32 bg_color; ++ __u8 depth; /* Depth of the image */ ++ const compat_uptr_t data; /* Pointer to image data */ ++ struct fb_cmap32 cmap; /* color map info */ ++}; ++ ++#define fb_image_from_compat(to, from) \ ++ (to).dx = (from).dx; \ ++ (to).dy = (from).dy; \ ++ (to).width = (from).width; \ ++ (to).height = (from).height; \ ++ (to).fg_color = (from).fg_color; \ ++ (to).bg_color = (from).bg_color; \ ++ (to).depth = (from).depth; \ ++ (to).data = compat_ptr((from).data); \ ++ fb_cmap_from_compat((to).cmap, (from).cmap) ++ ++#endif /* CONFIG_COMPAT */ ++#endif /* __KERNEL__ */ ++ + /* + * Frame buffer operations + * +@@ -489,6 +517,9 @@ + #define FBINFO_STATE_SUSPENDED 1 + u32 state; /* Hardware state i.e suspend */ + void *fbcon_par; /* fbcon use-only private area */ ++ ++ struct fb_image bgdecor; ++ + /* From here on everything is device dependent */ + void *par; + /* we need the PCI or similar aperture base/size not +diff -Nur linux-3.13.7.orig/include/uapi/linux/fb.h linux-3.13.7/include/uapi/linux/fb.h +--- linux-3.13.7.orig/include/uapi/linux/fb.h 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/include/uapi/linux/fb.h 2014-03-29 16:03:17.000000000 +0100 +@@ -8,6 +8,25 @@ + + #define FB_MAX 32 /* sufficient for now */ + ++struct fbcon_decor_iowrapper ++{ ++ unsigned short vc; /* Virtual console */ ++ unsigned char origin; /* Point of origin of the request */ ++ void *data; ++}; ++ ++#ifdef __KERNEL__ ++#ifdef CONFIG_COMPAT ++#include <linux/compat.h> ++struct fbcon_decor_iowrapper32 ++{ ++ unsigned short vc; /* Virtual console */ ++ unsigned char origin; /* Point of origin of the request */ ++ compat_uptr_t data; ++}; ++#endif /* CONFIG_COMPAT */ ++#endif /* __KERNEL__ */ ++ + /* ioctls + 0x46 is 'F' */ + #define FBIOGET_VSCREENINFO 0x4600 +@@ -35,6 +54,25 @@ + #define FBIOGET_DISPINFO 0x4618 + #define FBIO_WAITFORVSYNC _IOW('F', 0x20, __u32) + ++#define FBIOCONDECOR_SETCFG _IOWR('F', 0x19, struct fbcon_decor_iowrapper) ++#define FBIOCONDECOR_GETCFG _IOR('F', 0x1A, struct fbcon_decor_iowrapper) ++#define FBIOCONDECOR_SETSTATE _IOWR('F', 0x1B, struct fbcon_decor_iowrapper) ++#define FBIOCONDECOR_GETSTATE _IOR('F', 0x1C, struct fbcon_decor_iowrapper) ++#define FBIOCONDECOR_SETPIC _IOWR('F', 0x1D, struct fbcon_decor_iowrapper) ++#ifdef __KERNEL__ ++#ifdef CONFIG_COMPAT ++#define FBIOCONDECOR_SETCFG32 _IOWR('F', 0x19, struct fbcon_decor_iowrapper32) ++#define FBIOCONDECOR_GETCFG32 _IOR('F', 0x1A, struct fbcon_decor_iowrapper32) ++#define FBIOCONDECOR_SETSTATE32 _IOWR('F', 0x1B, struct fbcon_decor_iowrapper32) ++#define FBIOCONDECOR_GETSTATE32 _IOR('F', 0x1C, struct fbcon_decor_iowrapper32) ++#define FBIOCONDECOR_SETPIC32 _IOWR('F', 0x1D, struct fbcon_decor_iowrapper32) ++#endif /* CONFIG_COMPAT */ ++#endif /* __KERNEL__ */ ++ ++#define FBCON_DECOR_THEME_LEN 128 /* Maximum lenght of a theme name */ ++#define FBCON_DECOR_IO_ORIG_KERNEL 0 /* Kernel ioctl origin */ ++#define FBCON_DECOR_IO_ORIG_USER 1 /* User ioctl origin */ ++ + #define FB_TYPE_PACKED_PIXELS 0 /* Packed Pixels */ + #define FB_TYPE_PLANES 1 /* Non interleaved planes */ + #define FB_TYPE_INTERLEAVED_PLANES 2 /* Interleaved planes */ +@@ -277,6 +315,29 @@ + __u32 reserved[4]; /* Reserved for future compatibility */ + }; + ++#ifdef __KERNEL__ ++#ifdef CONFIG_COMPAT ++struct fb_cmap32 { ++ __u32 start; ++ __u32 len; /* Number of entries */ ++ compat_uptr_t red; /* Red values */ ++ compat_uptr_t green; ++ compat_uptr_t blue; ++ compat_uptr_t transp; /* transparency, can be NULL */ ++}; ++ ++#define fb_cmap_from_compat(to, from) \ ++ (to).start = (from).start; \ ++ (to).len = (from).len; \ ++ (to).red = compat_ptr((from).red); \ ++ (to).green = compat_ptr((from).green); \ ++ (to).blue = compat_ptr((from).blue); \ ++ (to).transp = compat_ptr((from).transp) ++ ++#endif /* CONFIG_COMPAT */ ++#endif /* __KERNEL__ */ ++ ++ + struct fb_cmap { + __u32 start; /* First entry */ + __u32 len; /* Number of entries */ +diff -Nur linux-3.13.7.orig/kernel/sysctl.c linux-3.13.7/kernel/sysctl.c +--- linux-3.13.7.orig/kernel/sysctl.c 2014-03-24 05:45:42.000000000 +0100 ++++ linux-3.13.7/kernel/sysctl.c 2014-03-29 16:03:17.000000000 +0100 +@@ -143,6 +143,10 @@ + static int ngroups_max = NGROUPS_MAX; + static const int cap_last_cap = CAP_LAST_CAP; + ++#ifdef CONFIG_FB_CON_DECOR ++extern char fbcon_decor_path[]; ++#endif ++ + #ifdef CONFIG_INOTIFY_USER + #include <linux/inotify.h> + #endif +@@ -252,6 +256,15 @@ + .mode = 0555, + .child = dev_table, + }, ++#ifdef CONFIG_FB_CON_DECOR ++ { ++ .procname = "fbcondecor", ++ .data = &fbcon_decor_path, ++ .maxlen = KMOD_PATH_LEN, ++ .mode = 0644, ++ .proc_handler = &proc_dostring, ++ }, ++#endif + { } + }; + diff --git a/target/linux/patches/3.14.40/gemalto.patch b/target/linux/patches/3.14.40/gemalto.patch new file mode 100644 index 000000000..65f7af1d7 --- /dev/null +++ b/target/linux/patches/3.14.40/gemalto.patch @@ -0,0 +1,11 @@ +diff -Nur linux-2.6.36.orig/drivers/tty/serial/8250/serial_cs.c linux-2.6.36/drivers/serial/8250/serial_cs.c +--- linux-2.6.36.orig/drivers/tty/serial/8250/serial_cs.c 2010-10-20 22:30:22.000000000 +0200 ++++ linux-2.6.36/drivers/tty/serial/8250/serial_cs.c 2010-12-13 23:03:40.000000000 +0100 +@@ -794,6 +794,7 @@ + PCMCIA_DEVICE_MANF_CARD(0x0137, 0x0025), + PCMCIA_DEVICE_MANF_CARD(0x0137, 0x0045), + PCMCIA_DEVICE_MANF_CARD(0x0137, 0x0052), ++ PCMCIA_DEVICE_MANF_CARD(0x0157, 0x0100), /* Gemalto SCR */ + PCMCIA_DEVICE_MANF_CARD(0x016c, 0x0006), /* Psion 56K+Fax */ + PCMCIA_DEVICE_MANF_CARD(0x0200, 0x0001), /* MultiMobile */ + PCMCIA_DEVICE_PROD_ID134("ADV", "TECH", "COMpad-32/85", 0x67459937, 0x916d02ba, 0x8fbe92ae), diff --git a/target/linux/patches/3.14.40/initramfs-nosizelimit.patch b/target/linux/patches/3.14.40/initramfs-nosizelimit.patch new file mode 100644 index 000000000..40d2f6bd8 --- /dev/null +++ b/target/linux/patches/3.14.40/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/3.14.40/lemote-rfkill.patch b/target/linux/patches/3.14.40/lemote-rfkill.patch new file mode 100644 index 000000000..a61488434 --- /dev/null +++ b/target/linux/patches/3.14.40/lemote-rfkill.patch @@ -0,0 +1,21 @@ +diff -Nur linux-3.3.orig/drivers/net/wireless/rtl818x/rtl8187/rfkill.c linux-3.3/drivers/net/wireless/rtl818x/rtl8187/rfkill.c +--- linux-3.3.orig/drivers/net/wireless/rtl818x/rtl8187/rfkill.c 2012-03-19 00:15:34.000000000 +0100 ++++ linux-3.3/drivers/net/wireless/rtl818x/rtl8187/rfkill.c 2012-03-27 23:29:46.000000000 +0200 +@@ -22,6 +22,9 @@ + + static bool rtl8187_is_radio_enabled(struct rtl8187_priv *priv) + { ++#ifdef CONFIG_LEMOTE_MACH2F ++ return 1; ++#else + u8 gpio; + + gpio = rtl818x_ioread8(priv, &priv->map->GPIO0); +@@ -29,6 +32,7 @@ + gpio = rtl818x_ioread8(priv, &priv->map->GPIO1); + + return gpio & priv->rfkill_mask; ++#endif + } + + void rtl8187_rfkill_init(struct ieee80211_hw *hw) diff --git a/target/linux/patches/3.14.40/microblaze-axi.patch b/target/linux/patches/3.14.40/microblaze-axi.patch new file mode 100644 index 000000000..1a4b17d8c --- /dev/null +++ b/target/linux/patches/3.14.40/microblaze-axi.patch @@ -0,0 +1,11 @@ +diff -Nur linux-3.13.3.orig/drivers/net/ethernet/xilinx/xilinx_axienet_main.c linux-3.13.3/drivers/net/ethernet/xilinx/xilinx_axienet_main.c +--- linux-3.13.3.orig/drivers/net/ethernet/xilinx/xilinx_axienet_main.c 2014-02-13 23:00:14.000000000 +0100 ++++ linux-3.13.3/drivers/net/ethernet/xilinx/xilinx_axienet_main.c 2014-02-24 08:03:57.000000000 +0100 +@@ -25,6 +25,7 @@ + #include <linux/init.h> + #include <linux/module.h> + #include <linux/netdevice.h> ++#include <linux/of_irq.h> + #include <linux/of_mdio.h> + #include <linux/of_platform.h> + #include <linux/of_address.h> diff --git a/target/linux/patches/3.14.40/microblaze-ethernet.patch b/target/linux/patches/3.14.40/microblaze-ethernet.patch new file mode 100644 index 000000000..742ab477e --- /dev/null +++ b/target/linux/patches/3.14.40/microblaze-ethernet.patch @@ -0,0 +1,11 @@ +diff -Nur linux-3.11.10.orig/drivers/net/ethernet/xilinx/xilinx_emaclite.c linux-3.11.10/drivers/net/ethernet/xilinx/xilinx_emaclite.c +--- linux-3.11.10.orig/drivers/net/ethernet/xilinx/xilinx_emaclite.c 2013-11-29 19:42:37.000000000 +0100 ++++ linux-3.11.10/drivers/net/ethernet/xilinx/xilinx_emaclite.c 2013-12-23 20:01:14.000000000 +0100 +@@ -1282,6 +1282,7 @@ + { .compatible = "xlnx,opb-ethernetlite-1.01.b", }, + { .compatible = "xlnx,xps-ethernetlite-1.00.a", }, + { .compatible = "xlnx,xps-ethernetlite-2.00.a", }, ++ { .compatible = "xlnx,xps-ethernetlite-2.00.b", }, + { .compatible = "xlnx,xps-ethernetlite-2.01.a", }, + { .compatible = "xlnx,xps-ethernetlite-3.00.a", }, + { /* end of list */ }, diff --git a/target/linux/patches/3.14.40/mkpiggy.patch b/target/linux/patches/3.14.40/mkpiggy.patch new file mode 100644 index 000000000..751678b74 --- /dev/null +++ b/target/linux/patches/3.14.40/mkpiggy.patch @@ -0,0 +1,28 @@ +diff -Nur linux-3.13.3.orig/arch/x86/boot/compressed/mkpiggy.c linux-3.13.3/arch/x86/boot/compressed/mkpiggy.c +--- linux-3.13.3.orig/arch/x86/boot/compressed/mkpiggy.c 2014-02-13 23:00:14.000000000 +0100 ++++ linux-3.13.3/arch/x86/boot/compressed/mkpiggy.c 2014-02-17 11:09:06.000000000 +0100 +@@ -29,7 +29,14 @@ + #include <stdio.h> + #include <string.h> + #include <inttypes.h> +-#include <tools/le_byteshift.h> ++ ++static uint32_t getle32(const void *p) ++{ ++ const uint8_t *cp = p; ++ ++ return (uint32_t)cp[0] + ((uint32_t)cp[1] << 8) + ++ ((uint32_t)cp[2] << 16) + ((uint32_t)cp[3] << 24); ++} + + int main(int argc, char *argv[]) + { +@@ -63,7 +70,7 @@ + } + + ilen = ftell(f); +- olen = get_unaligned_le32(&olen); ++ olen = getle32(&olen); + + /* + * Now we have the input (compressed) and output (uncompressed) diff --git a/target/linux/patches/3.14.40/mptcp.patch b/target/linux/patches/3.14.40/mptcp.patch new file mode 100644 index 000000000..9784e0577 --- /dev/null +++ b/target/linux/patches/3.14.40/mptcp.patch @@ -0,0 +1,17203 @@ +diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c +index d286bde..34d56d7 100644 +--- a/drivers/infiniband/hw/cxgb4/cm.c ++++ b/drivers/infiniband/hw/cxgb4/cm.c +@@ -3162,7 +3162,7 @@ static void build_cpl_pass_accept_req(struct sk_buff *skb, int stid , u8 tos) + */ + memset(&tmp_opt, 0, sizeof(tmp_opt)); + tcp_clear_options(&tmp_opt); +- tcp_parse_options(skb, &tmp_opt, 0, NULL); ++ tcp_parse_options(skb, &tmp_opt, NULL, 0, NULL); + + req = (struct cpl_pass_accept_req *)__skb_push(skb, sizeof(*req)); + memset(req, 0, sizeof(*req)); +diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h +index 2faef33..9c12362 100644 +--- a/include/linux/ipv6.h ++++ b/include/linux/ipv6.h +@@ -309,12 +309,6 @@ static inline struct ipv6_pinfo * inet6_sk(const struct sock *__sk) + return NULL; + } + +-static inline struct inet6_request_sock * +- inet6_rsk(const struct request_sock *rsk) +-{ +- return NULL; +-} +- + static inline struct raw6_sock *raw6_sk(const struct sock *sk) + { + return NULL; +diff --git a/include/linux/tcp.h b/include/linux/tcp.h +index 4ad0706..a230dd0 100644 +--- a/include/linux/tcp.h ++++ b/include/linux/tcp.h +@@ -72,6 +72,53 @@ struct tcp_sack_block { + u32 end_seq; + }; + ++struct tcp_out_options { ++ u16 options; /* bit field of OPTION_* */ ++ u8 ws; /* window scale, 0 to disable */ ++ u8 num_sack_blocks;/* number of SACK blocks to include */ ++ u8 hash_size; /* bytes in hash_location */ ++ u16 mss; /* 0 to disable */ ++ __u8 *hash_location; /* temporary pointer, overloaded */ ++ __u32 tsval, tsecr; /* need to include OPTION_TS */ ++ struct tcp_fastopen_cookie *fastopen_cookie; /* Fast open cookie */ ++#ifdef CONFIG_MPTCP ++ u16 mptcp_options; /* bit field of MPTCP related OPTION_* */ ++ u8 dss_csum:1, ++ add_addr_v4:1, ++ add_addr_v6:1; /* dss-checksum required? */ ++ ++ __u32 data_seq; /* data sequence number, for MPTCP */ ++ __u32 data_ack; /* data ack, for MPTCP */ ++ ++ union { ++ struct { ++ __u64 sender_key; /* sender's key for mptcp */ ++ __u64 receiver_key; /* receiver's key for mptcp */ ++ } mp_capable; ++ ++ struct { ++ __u64 sender_truncated_mac; ++ __u32 sender_nonce; ++ /* random number of the sender */ ++ __u32 token; /* token for mptcp */ ++ } mp_join_syns; ++ }; ++ ++ struct { ++ struct in_addr addr; ++ u8 addr_id; ++ } add_addr4; ++ ++ struct { ++ struct in6_addr addr; ++ u8 addr_id; ++ } add_addr6; ++ ++ u16 remove_addrs; /* list of address id */ ++ u8 addr_id; /* address id (mp_join or add_address) */ ++#endif /* CONFIG_MPTCP */ ++}; ++ + /*These are used to set the sack_ok field in struct tcp_options_received */ + #define TCP_SACK_SEEN (1 << 0) /*1 = peer is SACK capable, */ + #define TCP_FACK_ENABLED (1 << 1) /*1 = FACK is enabled locally*/ +@@ -95,6 +142,9 @@ struct tcp_options_received { + u16 mss_clamp; /* Maximal mss, negotiated at connection setup */ + }; + ++struct mptcp_cb; ++struct mptcp_tcp_sock; ++ + static inline void tcp_clear_options(struct tcp_options_received *rx_opt) + { + rx_opt->tstamp_ok = rx_opt->sack_ok = 0; +@@ -123,6 +173,7 @@ struct tcp_request_sock { + * FastOpen it's the seq# + * after data-in-SYN. + */ ++ u8 saw_mpc:1; + }; + + static inline struct tcp_request_sock *tcp_rsk(const struct request_sock *req) +@@ -130,6 +181,8 @@ static inline struct tcp_request_sock *tcp_rsk(const struct request_sock *req) + return (struct tcp_request_sock *)req; + } + ++struct tcp_md5sig_key; ++ + struct tcp_sock { + /* inet_connection_sock has to be the first member of tcp_sock */ + struct inet_connection_sock inet_conn; +@@ -323,6 +376,45 @@ struct tcp_sock { + * socket. Used to retransmit SYNACKs etc. + */ + struct request_sock *fastopen_rsk; ++ ++ ++ struct mptcp_cb *mpcb; ++ struct sock *meta_sk; ++ /* We keep these flags even if CONFIG_MPTCP is not checked, because ++ * it allows checking MPTCP capability just by checking the mpc flag, ++ * rather than adding ifdefs everywhere. ++ */ ++ u16 mpc:1, /* Other end is multipath capable */ ++ inside_tk_table:1, /* Is the tcp_sock inside the token-table? */ ++ send_mp_fclose:1, ++ request_mptcp:1, /* Did we send out an MP_CAPABLE? ++ * (this speeds up mptcp_doit() in tcp_recvmsg) ++ */ ++ mptcp_enabled:1, /* Is MPTCP enabled from the application ? */ ++ pf:1, /* Potentially Failed state: when this flag is set, we ++ * stop using the subflow ++ */ ++ mp_killed:1, /* Killed with a tcp_done in mptcp? */ ++ was_meta_sk:1, /* This was a meta sk (in case of reuse) */ ++ close_it:1, /* Must close socket in mptcp_data_ready? */ ++ closing:1; ++ struct mptcp_tcp_sock *mptcp; ++#ifdef CONFIG_MPTCP ++ struct hlist_nulls_node tk_table; ++ u32 mptcp_loc_token; ++ u64 mptcp_loc_key; ++#endif /* CONFIG_MPTCP */ ++ ++ /* Functions that depend on the value of the mpc flag */ ++ u32 (*__select_window)(struct sock *sk); ++ u16 (*select_window)(struct sock *sk); ++ void (*select_initial_window)(int __space, __u32 mss, __u32 *rcv_wnd, ++ __u32 *window_clamp, int wscale_ok, ++ __u8 *rcv_wscale, __u32 init_rcv_wnd, ++ const struct sock *sk); ++ void (*init_buffer_space)(struct sock *sk); ++ void (*set_rto)(struct sock *sk); ++ bool (*should_expand_sndbuf)(const struct sock *sk); + }; + + enum tsq_flags { +@@ -334,6 +426,8 @@ enum tsq_flags { + TCP_MTU_REDUCED_DEFERRED, /* tcp_v{4|6}_err() could not call + * tcp_v{4|6}_mtu_reduced() + */ ++ MPTCP_PATH_MANAGER, /* MPTCP deferred creation of new subflows */ ++ MPTCP_SUB_DEFERRED, /* A subflow got deferred - process them */ + }; + + static inline struct tcp_sock *tcp_sk(const struct sock *sk) +@@ -352,6 +446,7 @@ struct tcp_timewait_sock { + #ifdef CONFIG_TCP_MD5SIG + struct tcp_md5sig_key *tw_md5_key; + #endif ++ struct mptcp_tw *mptcp_tw; + }; + + static inline struct tcp_timewait_sock *tcp_twsk(const struct sock *sk) +diff --git a/include/net/inet6_connection_sock.h b/include/net/inet6_connection_sock.h +index f981ba7..0144c65 100644 +--- a/include/net/inet6_connection_sock.h ++++ b/include/net/inet6_connection_sock.h +@@ -27,6 +27,8 @@ int inet6_csk_bind_conflict(const struct sock *sk, + + struct dst_entry *inet6_csk_route_req(struct sock *sk, struct flowi6 *fl6, + const struct request_sock *req); ++u32 inet6_synq_hash(const struct in6_addr *raddr, const __be16 rport, ++ const u32 rnd, const u32 synq_hsize); + + struct request_sock *inet6_csk_search_req(const struct sock *sk, + struct request_sock ***prevp, +diff --git a/include/net/inet_common.h b/include/net/inet_common.h +index fe7994c..780f229 100644 +--- a/include/net/inet_common.h ++++ b/include/net/inet_common.h +@@ -1,6 +1,8 @@ + #ifndef _INET_COMMON_H + #define _INET_COMMON_H + ++#include <net/sock.h> ++ + extern const struct proto_ops inet_stream_ops; + extern const struct proto_ops inet_dgram_ops; + +@@ -13,6 +15,8 @@ struct sock; + struct sockaddr; + struct socket; + ++int inet_create(struct net *net, struct socket *sock, int protocol, int kern); ++int inet6_create(struct net *net, struct socket *sock, int protocol, int kern); + int inet_release(struct socket *sock); + int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, + int addr_len, int flags); +diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h +index c55aeed..84d1927 100644 +--- a/include/net/inet_connection_sock.h ++++ b/include/net/inet_connection_sock.h +@@ -243,6 +243,9 @@ static inline void inet_csk_reset_xmit_timer(struct sock *sk, const int what, + + struct sock *inet_csk_accept(struct sock *sk, int flags, int *err); + ++u32 inet_synq_hash(const __be32 raddr, const __be16 rport, const u32 rnd, ++ const u32 synq_hsize); ++ + struct request_sock *inet_csk_search_req(const struct sock *sk, + struct request_sock ***prevp, + const __be16 rport, +diff --git a/include/net/mptcp.h b/include/net/mptcp.h +new file mode 100644 +index 0000000..6454535 +--- /dev/null ++++ b/include/net/mptcp.h +@@ -0,0 +1,1471 @@ ++/* ++ * MPTCP implementation ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#ifndef _MPTCP_H ++#define _MPTCP_H ++ ++#include <linux/inetdevice.h> ++#include <linux/ipv6.h> ++#include <linux/list.h> ++#include <linux/net.h> ++#include <linux/netpoll.h> ++#include <linux/skbuff.h> ++#include <linux/socket.h> ++#include <linux/tcp.h> ++#include <linux/kernel.h> ++ ++#include <asm/byteorder.h> ++#include <asm/unaligned.h> ++#include <crypto/hash.h> ++#include <net/tcp.h> ++ ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ #define ntohll(x) be64_to_cpu(x) ++ #define htonll(x) cpu_to_be64(x) ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ #define ntohll(x) (x) ++ #define htonll(x) (x) ++#endif ++ ++/* Max number of local or remote addresses we can store. ++ * When changing, see the bitfield below in mptcp_loc4/6. */ ++#define MPTCP_MAX_ADDR 8 ++ ++#define MPTCP_SUBFLOW_RETRY_DELAY 1000 ++ ++struct mptcp_loc4 { ++ u8 loc4_id; ++ u8 low_prio:1; ++ struct in_addr addr; ++}; ++ ++struct mptcp_rem4 { ++ u8 rem4_id; ++ u8 bitfield; ++ u8 retry_bitfield; ++ __be16 port; ++ struct in_addr addr; ++}; ++ ++struct mptcp_loc6 { ++ u8 loc6_id; ++ u8 low_prio:1; ++ struct in6_addr addr; ++}; ++ ++struct mptcp_rem6 { ++ u8 rem6_id; ++ u8 bitfield; ++ u8 retry_bitfield; ++ __be16 port; ++ struct in6_addr addr; ++}; ++ ++struct mptcp_request_sock { ++ struct tcp_request_sock req; ++ struct mptcp_cb *mpcb; ++ /* Collision list in the tuple hashtable. We need to find ++ * the req sock when receiving the third msg of the 3-way handshake, ++ * since that one does not contain the token. If this makes ++ * the request sock too long, we can use kmalloc'ed specific entries for ++ * that tuple hashtable. At the moment, though, I extend the ++ * request_sock. ++ */ ++ struct list_head collide_tuple; ++ struct hlist_nulls_node collide_tk; ++ u32 mptcp_rem_nonce; ++ u32 mptcp_loc_token; ++ u64 mptcp_loc_key; ++ u64 mptcp_rem_key; ++ u64 mptcp_hash_tmac; ++ u32 mptcp_loc_nonce; ++ u8 loc_id; ++ u8 rem_id; /* Address-id in the MP_JOIN */ ++ u8 dss_csum:1, ++ low_prio:1; ++}; ++ ++struct mptcp_options_received { ++ u16 saw_mpc:1, ++ dss_csum:1, ++ drop_me:1, ++ ++ is_mp_join:1, ++ join_ack:1, ++ ++ saw_low_prio:2, /* 0x1 - low-prio set for this subflow ++ * 0x2 - low-prio set for another subflow ++ */ ++ low_prio:1, ++ ++ saw_add_addr:2, /* Saw at least one add_addr option: ++ * 0x1: IPv4 - 0x2: IPv6 ++ */ ++ more_add_addr:1, /* Saw one more add-addr. */ ++ ++ saw_rem_addr:1, /* Saw at least one rem_addr option */ ++ more_rem_addr:1, /* Saw one more rem-addr. */ ++ ++ mp_fail:1, ++ mp_fclose:1; ++ u8 rem_id; /* Address-id in the MP_JOIN */ ++ u8 prio_addr_id; /* Address-id in the MP_PRIO */ ++ ++ const unsigned char *add_addr_ptr; /* Pointer to add-address option */ ++ const unsigned char *rem_addr_ptr; /* Pointer to rem-address option */ ++ ++ u32 data_ack; ++ u32 data_seq; ++ u16 data_len; ++ ++ u32 mptcp_rem_token;/* Remote token */ ++ ++ /* Key inside the option (from mp_capable or fast_close) */ ++ u64 mptcp_key; ++ ++ u32 mptcp_recv_nonce; ++ u64 mptcp_recv_tmac; ++ u8 mptcp_recv_mac[20]; ++}; ++ ++struct mptcp_tcp_sock { ++ struct tcp_sock *next; /* Next subflow socket */ ++ struct list_head cb_list; ++ struct mptcp_options_received rx_opt; ++ ++ /* Those three fields record the current mapping */ ++ u64 map_data_seq; ++ u32 map_subseq; ++ u16 map_data_len; ++ u16 slave_sk:1, ++ fully_established:1, ++ establish_increased:1, ++ second_packet:1, ++ attached:1, ++ send_mp_fail:1, ++ include_mpc:1, ++ mapping_present:1, ++ map_data_fin:1, ++ low_prio:1, /* use this socket as backup */ ++ rcv_low_prio:1, /* Peer sent low-prio option to us */ ++ send_mp_prio:1, /* Trigger to send mp_prio on this socket */ ++ pre_established:1; /* State between sending 3rd ACK and ++ * receiving the fourth ack of new subflows. ++ */ ++ ++ /* isn: needed to translate abs to relative subflow seqnums */ ++ u32 snt_isn; ++ u32 rcv_isn; ++ u32 last_data_seq; ++ u8 path_index; ++ u8 loc_id; ++ u8 rem_id; ++ ++ u32 last_rbuf_opti; /* Timestamp of last rbuf optimization */ ++ unsigned int sent_pkts; ++ ++ struct sk_buff *shortcut_ofoqueue; /* Shortcut to the current modified ++ * skb in the ofo-queue. ++ */ ++ ++ int init_rcv_wnd; ++ u32 infinite_cutoff_seq; ++ struct delayed_work work; ++ u32 mptcp_loc_nonce; ++ struct tcp_sock *tp; /* Where is my daddy? */ ++ u32 last_end_data_seq; ++ ++ /* MP_JOIN subflow: timer for retransmitting the 3rd ack */ ++ struct timer_list mptcp_ack_timer; ++ ++ /* HMAC of the third ack */ ++ char sender_mac[20]; ++}; ++ ++struct mptcp_tw { ++ struct list_head list; ++ u64 loc_key; ++ u64 rcv_nxt; ++ struct mptcp_cb __rcu *mpcb; ++ u8 meta_tw:1, ++ in_list:1; ++}; ++ ++#define MPTCP_PM_NAME_MAX 16 ++struct mptcp_pm_ops { ++ struct list_head list; ++ ++ /* Signal the creation of a new MPTCP-session. */ ++ void (*new_session)(struct sock *meta_sk, int index); ++ void (*release_sock)(struct sock *meta_sk); ++ void (*fully_established)(struct sock *meta_sk); ++ void (*new_remote_address)(struct sock *meta_sk); ++ int (*get_local_index)(sa_family_t family, union inet_addr *addr, ++ struct net *net); ++ int (*get_local_id)(sa_family_t family, union inet_addr *addr, ++ struct net *net); ++ void (*addr_signal)(struct sock *sk, unsigned *size, ++ struct tcp_out_options *opts, struct sk_buff *skb); ++ ++ char name[MPTCP_PM_NAME_MAX]; ++ struct module *owner; ++}; ++ ++struct mptcp_cb { ++ struct sock *meta_sk; ++ ++ /* list of sockets in this multipath connection */ ++ struct tcp_sock *connection_list; ++ /* list of sockets that need a call to release_cb */ ++ struct list_head callback_list; ++ ++ spinlock_t tw_lock; ++ struct list_head tw_list; ++ unsigned char mptw_state; ++ ++ atomic_t mpcb_refcnt; ++ ++ /* High-order bits of 64-bit sequence numbers */ ++ u32 snd_high_order[2]; ++ u32 rcv_high_order[2]; ++ ++ u16 send_infinite_mapping:1, ++ in_time_wait:1, ++ list_rcvd:1, /* XXX TO REMOVE */ ++ dss_csum:1, ++ server_side:1, ++ infinite_mapping_rcv:1, ++ infinite_mapping_snd:1, ++ dfin_combined:1, /* Was the DFIN combined with subflow-fin? */ ++ passive_close:1, ++ snd_hiseq_index:1, /* Index in snd_high_order of snd_nxt */ ++ rcv_hiseq_index:1; /* Index in rcv_high_order of rcv_nxt */ ++ ++ /* socket count in this connection */ ++ u8 cnt_subflows; ++ u8 cnt_established; ++ ++ u32 noneligible; /* Path mask of temporarily non ++ * eligible subflows by the scheduler ++ */ ++ ++ struct sk_buff_head reinject_queue; ++ ++ u8 dfin_path_index; ++ ++#define MPTCP_PM_SIZE 320 ++ u8 mptcp_pm[MPTCP_PM_SIZE] __aligned(8); ++ struct mptcp_pm_ops *pm_ops; ++ ++ /* Mutex needed, because otherwise mptcp_close will complain that the ++ * socket is owned by the user. ++ * E.g., mptcp_sub_close_wq is taking the meta-lock. ++ */ ++ struct mutex mpcb_mutex; ++ ++ /* Master socket, also part of the connection_list, this ++ * socket is the one that the application sees. ++ */ ++ struct sock *master_sk; ++ ++ u64 csum_cutoff_seq; ++ ++ __u64 mptcp_loc_key; ++ __u32 mptcp_loc_token; ++ __u64 mptcp_rem_key; ++ __u32 mptcp_rem_token; ++ ++ /* Create a new subflow - necessary because the meta-sk may be IPv4, but ++ * the new subflow can be IPv6 ++ */ ++ struct sock *(*syn_recv_sock)(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst); ++ ++ /* Remote addresses */ ++ struct mptcp_rem4 remaddr4[MPTCP_MAX_ADDR]; ++ u8 rem4_bits; ++ ++ struct mptcp_rem6 remaddr6[MPTCP_MAX_ADDR]; ++ u8 rem6_bits; ++ ++ u32 path_index_bits; ++ /* Next pi to pick up in case a new path becomes available */ ++ u8 next_path_index; ++ ++ /* Original snd/rcvbuf of the initial subflow. ++ * Used for the new subflows on the server-side to allow correct ++ * autotuning ++ */ ++ int orig_sk_rcvbuf; ++ int orig_sk_sndbuf; ++ u32 orig_window_clamp; ++}; ++ ++#define MPTCP_SUB_CAPABLE 0 ++#define MPTCP_SUB_LEN_CAPABLE_SYN 12 ++#define MPTCP_SUB_LEN_CAPABLE_SYN_ALIGN 12 ++#define MPTCP_SUB_LEN_CAPABLE_ACK 20 ++#define MPTCP_SUB_LEN_CAPABLE_ACK_ALIGN 20 ++ ++#define MPTCP_SUB_JOIN 1 ++#define MPTCP_SUB_LEN_JOIN_SYN 12 ++#define MPTCP_SUB_LEN_JOIN_SYN_ALIGN 12 ++#define MPTCP_SUB_LEN_JOIN_SYNACK 16 ++#define MPTCP_SUB_LEN_JOIN_SYNACK_ALIGN 16 ++#define MPTCP_SUB_LEN_JOIN_ACK 24 ++#define MPTCP_SUB_LEN_JOIN_ACK_ALIGN 24 ++ ++#define MPTCP_SUB_DSS 2 ++#define MPTCP_SUB_LEN_DSS 4 ++#define MPTCP_SUB_LEN_DSS_ALIGN 4 ++ ++/* Lengths for seq and ack are the ones without the generic MPTCP-option header, ++ * as they are part of the DSS-option. ++ * To get the total length, just add the different options together. ++ */ ++#define MPTCP_SUB_LEN_SEQ 10 ++#define MPTCP_SUB_LEN_SEQ_CSUM 12 ++#define MPTCP_SUB_LEN_SEQ_ALIGN 12 ++ ++#define MPTCP_SUB_LEN_SEQ_64 14 ++#define MPTCP_SUB_LEN_SEQ_CSUM_64 16 ++#define MPTCP_SUB_LEN_SEQ_64_ALIGN 16 ++ ++#define MPTCP_SUB_LEN_ACK 4 ++#define MPTCP_SUB_LEN_ACK_ALIGN 4 ++ ++#define MPTCP_SUB_LEN_ACK_64 8 ++#define MPTCP_SUB_LEN_ACK_64_ALIGN 8 ++ ++/* This is the "default" option-length we will send out most often. ++ * MPTCP DSS-header ++ * 32-bit data sequence number ++ * 32-bit data ack ++ * ++ * It is necessary to calculate the effective MSS we will be using when ++ * sending data. ++ */ ++#define MPTCP_SUB_LEN_DSM_ALIGN (MPTCP_SUB_LEN_DSS_ALIGN + \ ++ MPTCP_SUB_LEN_SEQ_ALIGN + \ ++ MPTCP_SUB_LEN_ACK_ALIGN) ++ ++#define MPTCP_SUB_ADD_ADDR 3 ++#define MPTCP_SUB_LEN_ADD_ADDR4 8 ++#define MPTCP_SUB_LEN_ADD_ADDR6 20 ++#define MPTCP_SUB_LEN_ADD_ADDR4_ALIGN 8 ++#define MPTCP_SUB_LEN_ADD_ADDR6_ALIGN 20 ++ ++#define MPTCP_SUB_REMOVE_ADDR 4 ++#define MPTCP_SUB_LEN_REMOVE_ADDR 4 ++ ++#define MPTCP_SUB_PRIO 5 ++#define MPTCP_SUB_LEN_PRIO 3 ++#define MPTCP_SUB_LEN_PRIO_ADDR 4 ++#define MPTCP_SUB_LEN_PRIO_ALIGN 4 ++ ++#define MPTCP_SUB_FAIL 6 ++#define MPTCP_SUB_LEN_FAIL 12 ++#define MPTCP_SUB_LEN_FAIL_ALIGN 12 ++ ++#define MPTCP_SUB_FCLOSE 7 ++#define MPTCP_SUB_LEN_FCLOSE 12 ++#define MPTCP_SUB_LEN_FCLOSE_ALIGN 12 ++ ++ ++#define OPTION_MPTCP (1 << 5) ++ ++static inline void reset_mpc(struct tcp_sock *tp) ++{ ++ tp->mpc = 0; ++ ++ tp->__select_window = __tcp_select_window; ++ tp->select_window = tcp_select_window; ++ tp->select_initial_window = tcp_select_initial_window; ++ tp->init_buffer_space = tcp_init_buffer_space; ++ tp->set_rto = tcp_set_rto; ++ tp->should_expand_sndbuf = tcp_should_expand_sndbuf; ++} ++ ++/* Initializes MPTCP flags in tcp_sock (and other tcp_sock members that depend ++ * on those flags). ++ */ ++static inline void mptcp_init_tcp_sock(struct tcp_sock *tp) ++{ ++ reset_mpc(tp); ++} ++ ++#ifdef CONFIG_MPTCP ++ ++/* Used for checking if the mptcp initialization has been successful */ ++extern bool mptcp_init_failed; ++ ++/* MPTCP options */ ++#define OPTION_TYPE_SYN (1 << 0) ++#define OPTION_TYPE_SYNACK (1 << 1) ++#define OPTION_TYPE_ACK (1 << 2) ++#define OPTION_MP_CAPABLE (1 << 3) ++#define OPTION_DATA_ACK (1 << 4) ++#define OPTION_ADD_ADDR (1 << 5) ++#define OPTION_MP_JOIN (1 << 6) ++#define OPTION_MP_FAIL (1 << 7) ++#define OPTION_MP_FCLOSE (1 << 8) ++#define OPTION_REMOVE_ADDR (1 << 9) ++#define OPTION_MP_PRIO (1 << 10) ++ ++/* MPTCP flags */ ++#define MPTCPHDR_ACK 0x01 ++#define MPTCPHDR_SEQ 0x02 ++#define MPTCPHDR_FIN 0x04 ++#define MPTCPHDR_INF 0x08 ++#define MPTCPHDR_SEQ64_SET 0x10 /* Did we received a 64-bit seq number */ ++#define MPTCPHDR_SEQ64_OFO 0x20 /* Is it not in our circular array? */ ++#define MPTCPHDR_SEQ64_INDEX 0x40 /* Index of seq in mpcb->snd_high_order */ ++#define MPTCPHDR_DSS_CSUM 0x80 ++ ++/* It is impossible, that all 8 bits of mptcp_flags are set to 1 with the above ++ * Thus, defining MPTCPHDR_JOIN as 0xFF is safe. ++ */ ++#define MPTCPHDR_JOIN 0xFF ++ ++struct mptcp_option { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 ver:4, ++ sub:4; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ ver:4; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++}; ++ ++struct mp_capable { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 ver:4, ++ sub:4; ++ __u8 h:1, ++ rsv:5, ++ b:1, ++ a:1; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ ver:4; ++ __u8 a:1, ++ b:1, ++ rsv:5, ++ h:1; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __u64 sender_key; ++ __u64 receiver_key; ++} __attribute__((__packed__)); ++ ++struct mp_join { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 b:1, ++ rsv:3, ++ sub:4; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ rsv:3, ++ b:1; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __u8 addr_id; ++ union { ++ struct { ++ u32 token; ++ u32 nonce; ++ } syn; ++ struct { ++ __u64 mac; ++ u32 nonce; ++ } synack; ++ struct { ++ __u8 mac[20]; ++ } ack; ++ } u; ++} __attribute__((__packed__)); ++ ++struct mp_dss { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u16 rsv1:4, ++ sub:4, ++ A:1, ++ a:1, ++ M:1, ++ m:1, ++ F:1, ++ rsv2:3; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u16 sub:4, ++ rsv1:4, ++ rsv2:3, ++ F:1, ++ m:1, ++ M:1, ++ a:1, ++ A:1; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++}; ++ ++struct mp_add_addr { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 ipver:4, ++ sub:4; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ ipver:4; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __u8 addr_id; ++ union { ++ struct { ++ struct in_addr addr; ++ __be16 port; ++ } v4; ++ struct { ++ struct in6_addr addr; ++ __be16 port; ++ } v6; ++ } u; ++} __attribute__((__packed__)); ++ ++struct mp_remove_addr { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 rsv:4, ++ sub:4; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ rsv:4; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ /* list of addr_id */ ++ __u8 addrs_id; ++}; ++ ++struct mp_fail { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u16 rsv1:4, ++ sub:4, ++ rsv2:8; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u16 sub:4, ++ rsv1:4, ++ rsv2:8; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __be64 data_seq; ++} __attribute__((__packed__)); ++ ++struct mp_fclose { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u16 rsv1:4, ++ sub:4, ++ rsv2:8; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u16 sub:4, ++ rsv1:4, ++ rsv2:8; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __u64 key; ++} __attribute__((__packed__)); ++ ++struct mp_prio { ++ __u8 kind; ++ __u8 len; ++#if defined(__LITTLE_ENDIAN_BITFIELD) ++ __u8 b:1, ++ rsv:3, ++ sub:4; ++#elif defined(__BIG_ENDIAN_BITFIELD) ++ __u8 sub:4, ++ rsv:3, ++ b:1; ++#else ++#error "Adjust your <asm/byteorder.h> defines" ++#endif ++ __u8 addr_id; ++} __attribute__((__packed__)); ++ ++static inline int mptcp_sub_len_dss(struct mp_dss *m, int csum) ++{ ++ return 4 + m->A * (4 + m->a * 4) + m->M * (10 + m->m * 4 + csum * 2); ++} ++ ++#define MPTCP_APP 2 ++ ++extern int sysctl_mptcp_enabled; ++extern int sysctl_mptcp_checksum; ++extern int sysctl_mptcp_debug; ++extern int sysctl_mptcp_syn_retries; ++ ++extern struct workqueue_struct *mptcp_wq; ++ ++#define mptcp_debug(fmt, args...) \ ++ do { \ ++ if (unlikely(sysctl_mptcp_debug)) \ ++ pr_err(__FILE__ ": " fmt, ##args); \ ++ } while (0) ++ ++/* Iterates over all subflows */ ++#define mptcp_for_each_tp(mpcb, tp) \ ++ for ((tp) = (mpcb)->connection_list; (tp); (tp) = (tp)->mptcp->next) ++ ++#define mptcp_for_each_sk(mpcb, sk) \ ++ for ((sk) = (struct sock *)(mpcb)->connection_list; \ ++ sk; \ ++ sk = (struct sock *)tcp_sk(sk)->mptcp->next) ++ ++#define mptcp_for_each_sk_safe(__mpcb, __sk, __temp) \ ++ for (__sk = (struct sock *)(__mpcb)->connection_list, \ ++ __temp = __sk ? (struct sock *)tcp_sk(__sk)->mptcp->next : NULL; \ ++ __sk; \ ++ __sk = __temp, \ ++ __temp = __sk ? (struct sock *)tcp_sk(__sk)->mptcp->next : NULL) ++ ++/* Iterates over all bit set to 1 in a bitset */ ++#define mptcp_for_each_bit_set(b, i) \ ++ for (i = ffs(b) - 1; i >= 0; i = ffs(b >> (i + 1) << (i + 1)) - 1) ++ ++#define mptcp_for_each_bit_unset(b, i) \ ++ mptcp_for_each_bit_set(~b, i) ++ ++extern struct lock_class_key meta_key; ++extern struct lock_class_key meta_slock_key; ++extern u32 mptcp_secret[MD5_MESSAGE_BYTES / 4]; ++ ++/* This is needed to ensure that two subsequent key-generation result in ++ * different keys if the IPs and ports are the same. ++ */ ++extern u32 mptcp_key_seed; ++ ++#define MPTCP_HASH_SIZE 1024 ++ ++extern struct hlist_nulls_head tk_hashtable[MPTCP_HASH_SIZE]; ++ ++/* This second hashtable is needed to retrieve request socks ++ * created as a result of a join request. While the SYN contains ++ * the token, the final ack does not, so we need a separate hashtable ++ * to retrieve the mpcb. ++ */ ++extern struct list_head mptcp_reqsk_htb[MPTCP_HASH_SIZE]; ++extern spinlock_t mptcp_reqsk_hlock; /* hashtable protection */ ++ ++/* Lock, protecting the two hash-tables that hold the token. Namely, ++ * mptcp_reqsk_tk_htb and tk_hashtable ++ */ ++extern spinlock_t mptcp_tk_hashlock; /* hashtable protection */ ++ ++void mptcp_data_ready(struct sock *sk, int bytes); ++void mptcp_write_space(struct sock *sk); ++ ++void mptcp_add_meta_ofo_queue(struct sock *meta_sk, struct sk_buff *skb, ++ struct sock *sk); ++void mptcp_ofo_queue(struct sock *meta_sk); ++void mptcp_purge_ofo_queue(struct tcp_sock *meta_tp); ++void mptcp_cleanup_rbuf(struct sock *meta_sk, int copied); ++int mptcp_alloc_mpcb(struct sock *master_sk, __u64 remote_key, u32 window); ++int mptcp_add_sock(struct sock *meta_sk, struct sock *sk, u8 loc_id, u8 rem_id, ++ gfp_t flags); ++void mptcp_del_sock(struct sock *sk); ++void mptcp_update_metasocket(struct sock *sock, struct sock *meta_sk); ++void mptcp_reinject_data(struct sock *orig_sk, int clone_it); ++void mptcp_update_sndbuf(struct mptcp_cb *mpcb); ++struct sk_buff *mptcp_next_segment(struct sock *sk, int *reinject); ++void mptcp_send_fin(struct sock *meta_sk); ++void mptcp_send_active_reset(struct sock *meta_sk, gfp_t priority); ++int mptcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, ++ int push_one, gfp_t gfp); ++void mptcp_parse_options(const uint8_t *ptr, int opsize, ++ struct tcp_options_received *opt_rx, ++ struct mptcp_options_received *mopt, ++ const struct sk_buff *skb); ++void mptcp_syn_options(struct sock *sk, struct tcp_out_options *opts, ++ unsigned *remaining); ++void mptcp_synack_options(struct request_sock *req, ++ struct tcp_out_options *opts, ++ unsigned *remaining); ++void mptcp_established_options(struct sock *sk, struct sk_buff *skb, ++ struct tcp_out_options *opts, unsigned *size); ++void mptcp_options_write(__be32 *ptr, struct tcp_sock *tp, ++ struct tcp_out_options *opts, ++ struct sk_buff *skb); ++void mptcp_close(struct sock *meta_sk, long timeout); ++int mptcp_doit(struct sock *sk); ++int mptcp_create_master_sk(struct sock *meta_sk, __u64 remote_key, u32 window); ++int mptcp_check_req_master(struct sock *sk, struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ struct mptcp_options_received *mopt); ++struct sock *mptcp_check_req_child(struct sock *sk, struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ struct mptcp_options_received *mopt); ++u32 __mptcp_select_window(struct sock *sk); ++void mptcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd, ++ __u32 *window_clamp, int wscale_ok, ++ __u8 *rcv_wscale, __u32 init_rcv_wnd, ++ const struct sock *sk); ++unsigned int mptcp_current_mss(struct sock *meta_sk); ++int mptcp_select_size(const struct sock *meta_sk, bool sg); ++void mptcp_key_sha1(u64 key, u32 *token, u64 *idsn); ++void mptcp_hmac_sha1(u8 *key_1, u8 *key_2, u8 *rand_1, u8 *rand_2, ++ u32 *hash_out); ++void mptcp_clean_rtx_infinite(struct sk_buff *skb, struct sock *sk); ++void mptcp_fin(struct sock *meta_sk); ++void mptcp_retransmit_timer(struct sock *meta_sk); ++int mptcp_write_wakeup(struct sock *meta_sk); ++void mptcp_sub_close_wq(struct work_struct *work); ++void mptcp_sub_close(struct sock *sk, unsigned long delay); ++struct sock *mptcp_select_ack_sock(const struct sock *meta_sk, int copied); ++void mptcp_fallback_meta_sk(struct sock *meta_sk); ++int mptcp_backlog_rcv(struct sock *meta_sk, struct sk_buff *skb); ++struct sock *mptcp_sk_clone(const struct sock *sk, int family, const gfp_t priority); ++void mptcp_ack_handler(unsigned long); ++int mptcp_check_rtt(const struct tcp_sock *tp, int time); ++int mptcp_check_snd_buf(const struct tcp_sock *tp); ++int mptcp_handle_options(struct sock *sk, const struct tcphdr *th, struct sk_buff *skb); ++void __init mptcp_init(void); ++int mptcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len); ++int mptcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, ++ unsigned int mss_now, int reinject); ++int mptso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, ++ unsigned int mss_now, gfp_t gfp, int reinject); ++void mptcp_destroy_sock(struct sock *sk); ++int mptcp_rcv_synsent_state_process(struct sock *sk, struct sock **skptr, ++ struct sk_buff *skb, ++ struct mptcp_options_received *mopt); ++unsigned int mptcp_xmit_size_goal(struct sock *meta_sk, u32 mss_now, ++ int large_allowed); ++int mptcp_time_wait(struct sock *sk, struct tcp_timewait_sock *tw); ++void mptcp_twsk_destructor(struct tcp_timewait_sock *tw); ++void mptcp_update_tw_socks(const struct tcp_sock *tp, int state); ++void mptcp_disconnect(struct sock *sk); ++bool mptcp_should_expand_sndbuf(const struct sock *sk); ++int mptcp_retransmit_skb(struct sock *meta_sk, struct sk_buff *skb); ++void mptcp_tsq_flags(struct sock *sk); ++void mptcp_tsq_sub_deferred(struct sock *meta_sk); ++struct mp_join *mptcp_find_join(struct sk_buff *skb); ++void mptcp_hash_remove_bh(struct tcp_sock *meta_tp); ++void mptcp_hash_remove(struct tcp_sock *meta_tp); ++struct sock *mptcp_hash_find(struct net *net, u32 token); ++int mptcp_lookup_join(struct sk_buff *skb, struct inet_timewait_sock *tw); ++int mptcp_do_join_short(struct sk_buff *skb, struct mptcp_options_received *mopt, ++ struct tcp_options_received *tmp_opt, struct net *net); ++void mptcp_reqsk_destructor(struct request_sock *req); ++void mptcp_reqsk_new_mptcp(struct request_sock *req, ++ const struct tcp_options_received *rx_opt, ++ const struct mptcp_options_received *mopt, ++ const struct sk_buff *skb); ++int mptcp_check_req(struct sk_buff *skb, struct net *net); ++void mptcp_connect_init(struct sock *sk); ++void mptcp_sub_force_close(struct sock *sk); ++int mptcp_sub_len_remove_addr_align(u16 bitfield); ++void mptcp_remove_shortcuts(const struct mptcp_cb *mpcb, ++ const struct sk_buff *skb); ++void mptcp_init_buffer_space(struct sock *sk); ++ ++/* MPTCP-path-manager registration/initialization functions */ ++int mptcp_register_path_manager(struct mptcp_pm_ops *pm); ++void mptcp_unregister_path_manager(struct mptcp_pm_ops *pm); ++void mptcp_init_path_manager(struct mptcp_cb *mpcb); ++void mptcp_cleanup_path_manager(struct mptcp_cb *mpcb); ++void mptcp_fallback_default(struct mptcp_cb *mpcb); ++void mptcp_get_default_path_manager(char *name); ++int mptcp_set_default_path_manager(const char *name); ++extern struct mptcp_pm_ops mptcp_pm_default; ++ ++static inline ++struct mptcp_request_sock *mptcp_rsk(const struct request_sock *req) ++{ ++ return (struct mptcp_request_sock *)req; ++} ++ ++static inline ++struct request_sock *rev_mptcp_rsk(const struct mptcp_request_sock *req) ++{ ++ return (struct request_sock *)req; ++} ++ ++static inline bool mptcp_can_sendpage(struct sock *sk) ++{ ++ struct sock *sk_it; ++ ++ if (tcp_sk(sk)->mpcb->dss_csum) ++ return false; ++ ++ mptcp_for_each_sk(tcp_sk(sk)->mpcb, sk_it) { ++ if (!(sk_it->sk_route_caps & NETIF_F_SG) || ++ !(sk_it->sk_route_caps & NETIF_F_ALL_CSUM)) ++ return false; ++ } ++ ++ return true; ++} ++ ++static inline void mptcp_push_pending_frames(struct sock *meta_sk) ++{ ++ if (mptcp_next_segment(meta_sk, NULL)) { ++ struct tcp_sock *tp = tcp_sk(meta_sk); ++ ++ /* We don't care about the MSS, because it will be set in ++ * mptcp_write_xmit. ++ */ ++ __tcp_push_pending_frames(meta_sk, 0, tp->nonagle); ++ } ++} ++ ++static inline void mptcp_send_reset(struct sock *sk) ++{ ++ tcp_send_active_reset(sk, GFP_ATOMIC); ++ mptcp_sub_force_close(sk); ++} ++ ++static inline int mptcp_is_data_seq(const struct sk_buff *skb) ++{ ++ return TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_SEQ; ++} ++ ++static inline int mptcp_is_data_fin(const struct sk_buff *skb) ++{ ++ return mptcp_is_data_seq(skb) && ++ (TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_FIN); ++} ++ ++/* Is it a data-fin while in infinite mapping mode? ++ * In infinite mode, a subflow-fin is in fact a data-fin. ++ */ ++static inline int mptcp_is_data_fin2(const struct sk_buff *skb, ++ const struct tcp_sock *tp) ++{ ++ return mptcp_is_data_fin(skb) || ++ (tp->mpcb->infinite_mapping_rcv && tcp_hdr(skb)->fin); ++} ++ ++static inline void mptcp_skb_entail_init(const struct tcp_sock *tp, ++ struct sk_buff *skb) ++{ ++ TCP_SKB_CB(skb)->mptcp_flags = MPTCPHDR_SEQ; ++} ++ ++static inline u8 mptcp_get_64_bit(u64 data_seq, struct mptcp_cb *mpcb) ++{ ++ u64 data_seq_high = (u32)(data_seq >> 32); ++ ++ if (mpcb->rcv_high_order[0] == data_seq_high) ++ return 0; ++ else if (mpcb->rcv_high_order[1] == data_seq_high) ++ return MPTCPHDR_SEQ64_INDEX; ++ else ++ return MPTCPHDR_SEQ64_OFO; ++} ++ ++/* Sets the data_seq and returns pointer to the in-skb field of the data_seq. ++ * If the packet has a 64-bit dseq, the pointer points to the last 32 bits. ++ */ ++static inline __u32 *mptcp_skb_set_data_seq(const struct sk_buff *skb, ++ u32 *data_seq, ++ struct mptcp_cb *mpcb) ++{ ++ __u32 *ptr = (__u32 *)(skb_transport_header(skb) + TCP_SKB_CB(skb)->dss_off); ++ ++ if (TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_SEQ64_SET) { ++ u64 data_seq64 = get_unaligned_be64(ptr); ++ ++ if (mpcb) ++ TCP_SKB_CB(skb)->mptcp_flags |= mptcp_get_64_bit(data_seq64, mpcb); ++ ++ *data_seq = (u32)data_seq64 ; ++ ptr++; ++ } else { ++ *data_seq = get_unaligned_be32(ptr); ++ } ++ ++ return ptr; ++} ++ ++static inline struct sock *mptcp_meta_sk(const struct sock *sk) ++{ ++ return tcp_sk(sk)->meta_sk; ++} ++ ++static inline struct tcp_sock *mptcp_meta_tp(const struct tcp_sock *tp) ++{ ++ return tcp_sk(tp->meta_sk); ++} ++ ++static inline int is_meta_tp(const struct tcp_sock *tp) ++{ ++ return tp->mpcb && mptcp_meta_tp(tp) == tp; ++} ++ ++static inline int is_meta_sk(const struct sock *sk) ++{ ++ return sk->sk_type == SOCK_STREAM && sk->sk_protocol == IPPROTO_TCP && ++ tcp_sk(sk)->mpc && mptcp_meta_sk(sk) == sk; ++} ++ ++static inline int is_master_tp(const struct tcp_sock *tp) ++{ ++ return !tp->mpc || (!tp->mptcp->slave_sk && !is_meta_tp(tp)); ++} ++ ++static inline void mptcp_hash_request_remove(struct request_sock *req) ++{ ++ int in_softirq = 0; ++ ++ if (list_empty(&mptcp_rsk(req)->collide_tuple)) ++ return; ++ ++ if (in_softirq()) { ++ spin_lock(&mptcp_reqsk_hlock); ++ in_softirq = 1; ++ } else { ++ spin_lock_bh(&mptcp_reqsk_hlock); ++ } ++ ++ list_del(&mptcp_rsk(req)->collide_tuple); ++ ++ if (in_softirq) ++ spin_unlock(&mptcp_reqsk_hlock); ++ else ++ spin_unlock_bh(&mptcp_reqsk_hlock); ++} ++ ++static inline void mptcp_init_mp_opt(struct mptcp_options_received *mopt) ++{ ++ mopt->saw_mpc = 0; ++ mopt->dss_csum = 0; ++ mopt->drop_me = 0; ++ ++ mopt->is_mp_join = 0; ++ mopt->join_ack = 0; ++ ++ mopt->saw_low_prio = 0; ++ mopt->low_prio = 0; ++ ++ mopt->saw_add_addr = 0; ++ mopt->more_add_addr = 0; ++ ++ mopt->saw_rem_addr = 0; ++ mopt->more_rem_addr = 0; ++ ++ mopt->mp_fail = 0; ++ mopt->mp_fclose = 0; ++} ++ ++static inline void mptcp_reset_mopt(struct tcp_sock *tp) ++{ ++ struct mptcp_options_received *mopt = &tp->mptcp->rx_opt; ++ ++ mopt->saw_low_prio = 0; ++ mopt->saw_add_addr = 0; ++ mopt->more_add_addr = 0; ++ mopt->saw_rem_addr = 0; ++ mopt->more_rem_addr = 0; ++ mopt->join_ack = 0; ++ mopt->mp_fail = 0; ++ mopt->mp_fclose = 0; ++} ++ ++static inline __be32 mptcp_get_highorder_sndbits(const struct sk_buff *skb, ++ const struct mptcp_cb *mpcb) ++{ ++ return htonl(mpcb->snd_high_order[(TCP_SKB_CB(skb)->mptcp_flags & ++ MPTCPHDR_SEQ64_INDEX) ? 1 : 0]); ++} ++ ++static inline u64 mptcp_get_data_seq_64(const struct mptcp_cb *mpcb, int index, ++ u32 data_seq_32) ++{ ++ return ((u64)mpcb->rcv_high_order[index] << 32) | data_seq_32; ++} ++ ++static inline u64 mptcp_get_rcv_nxt_64(const struct tcp_sock *meta_tp) ++{ ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ return mptcp_get_data_seq_64(mpcb, mpcb->rcv_hiseq_index, ++ meta_tp->rcv_nxt); ++} ++ ++static inline void mptcp_check_sndseq_wrap(struct tcp_sock *meta_tp, int inc) ++{ ++ if (unlikely(meta_tp->snd_nxt > meta_tp->snd_nxt + inc)) { ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ mpcb->snd_hiseq_index = mpcb->snd_hiseq_index ? 0 : 1; ++ mpcb->snd_high_order[mpcb->snd_hiseq_index] += 2; ++ } ++} ++ ++static inline void mptcp_check_rcvseq_wrap(struct tcp_sock *meta_tp, ++ u32 old_rcv_nxt) ++{ ++ if (unlikely(old_rcv_nxt > meta_tp->rcv_nxt)) { ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ mpcb->rcv_high_order[mpcb->rcv_hiseq_index] += 2; ++ mpcb->rcv_hiseq_index = mpcb->rcv_hiseq_index ? 0 : 1; ++ } ++} ++ ++static inline int mptcp_sk_can_send(const struct sock *sk) ++{ ++ return (1 << sk->sk_state) & (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT) && ++ !tcp_sk(sk)->mptcp->pre_established; ++} ++ ++static inline int mptcp_sk_can_recv(const struct sock *sk) ++{ ++ return (1 << sk->sk_state) & (TCPF_ESTABLISHED | TCP_FIN_WAIT1 | TCP_FIN_WAIT2); ++} ++ ++static inline int mptcp_sk_can_send_ack(const struct sock *sk) ++{ ++ return !((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV | ++ TCPF_CLOSE | TCPF_LISTEN)) && ++ !tcp_sk(sk)->mptcp->pre_established; ++} ++ ++/* Only support GSO if all subflows supports it */ ++static inline bool mptcp_sk_can_gso(const struct sock *meta_sk) ++{ ++ struct sock *sk; ++ ++ if (tcp_sk(meta_sk)->mpcb->dss_csum) ++ return 0; ++ ++ mptcp_for_each_sk(tcp_sk(meta_sk)->mpcb, sk) { ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ if (!sk_can_gso(sk)) ++ return false; ++ } ++ return true; ++} ++ ++static inline bool mptcp_can_sg(const struct sock *meta_sk) ++{ ++ struct sock *sk; ++ ++ if (tcp_sk(meta_sk)->mpcb->dss_csum) ++ return 0; ++ ++ mptcp_for_each_sk(tcp_sk(meta_sk)->mpcb, sk) { ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ if (!(sk->sk_route_caps & NETIF_F_SG)) ++ return false; ++ } ++ return true; ++} ++ ++static inline void mptcp_set_rto(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *sk_it; ++ struct inet_connection_sock *micsk = inet_csk(mptcp_meta_sk(sk)); ++ __u32 max_rto = 0; ++ ++ /* We are in recovery-phase on the MPTCP-level. Do not update the ++ * RTO, because this would kill exponential backoff. ++ */ ++ if (micsk->icsk_retransmits) ++ return; ++ ++ mptcp_for_each_sk(tp->mpcb, sk_it) { ++ if (mptcp_sk_can_send(sk_it) && ++ inet_csk(sk_it)->icsk_rto > max_rto) ++ max_rto = inet_csk(sk_it)->icsk_rto; ++ } ++ if (max_rto) { ++ micsk->icsk_rto = max_rto << 1; ++ ++ /* A successfull rto-measurement - reset backoff counter */ ++ micsk->icsk_backoff = 0; ++ } ++} ++ ++static inline int mptcp_sysctl_syn_retries(void) ++{ ++ return sysctl_mptcp_syn_retries; ++} ++ ++static inline void mptcp_sub_close_passive(struct sock *sk) ++{ ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = tcp_sk(meta_sk); ++ ++ /* Only close, if the app did a send-shutdown (passive close), and we ++ * received the data-ack of the data-fin. ++ */ ++ if (tp->mpcb->passive_close && meta_tp->snd_una == meta_tp->write_seq) ++ mptcp_sub_close(sk, 0); ++} ++ ++static inline bool mptcp_fallback_infinite(struct sock *sk, int flag) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ /* If data has been acknowleged on the meta-level, fully_established ++ * will have been set before and thus we will not fall back to infinite ++ * mapping. ++ */ ++ if (likely(tp->mptcp->fully_established)) ++ return false; ++ ++ if (!(flag & MPTCP_FLAG_DATA_ACKED)) ++ return false; ++ ++ /* Don't fallback twice ;) */ ++ if (tp->mpcb->infinite_mapping_snd) ++ return false; ++ ++ pr_err("%s %#x will fallback - pi %d, src %pI4 dst %pI4 from %pS\n", ++ __func__, tp->mpcb->mptcp_loc_token, tp->mptcp->path_index, ++ &inet_sk(sk)->inet_saddr, &inet_sk(sk)->inet_daddr, ++ __builtin_return_address(0)); ++ if (!is_master_tp(tp)) ++ return true; ++ ++ tp->mpcb->infinite_mapping_snd = 1; ++ tp->mpcb->infinite_mapping_rcv = 1; ++ tp->mptcp->fully_established = 1; ++ ++ return false; ++} ++ ++/* Find the first free index in the bitfield */ ++static inline int __mptcp_find_free_index(u8 bitfield, int j, u8 base) ++{ ++ int i; ++ mptcp_for_each_bit_unset(bitfield >> base, i) { ++ /* We wrapped at the bitfield - try from 0 on */ ++ if (i + base >= sizeof(bitfield) * 8) { ++ mptcp_for_each_bit_unset(bitfield, i) { ++ if (i >= sizeof(bitfield) * 8) ++ goto exit; ++ ++ if (i != j) ++ return i; ++ } ++ goto exit; ++ } ++ if (i + base >= sizeof(bitfield) * 8) ++ break; ++ ++ if (i + base != j) ++ return i + base; ++ } ++exit: ++ return -1; ++} ++ ++static inline int mptcp_find_free_index(u8 bitfield) ++{ ++ return __mptcp_find_free_index(bitfield, -1, 0); ++} ++ ++/* Find the first index whose bit in the bit-field == 0 */ ++static inline u8 mptcp_set_new_pathindex(struct mptcp_cb *mpcb) ++{ ++ u8 base = mpcb->next_path_index; ++ int i; ++ ++ /* Start at 1, because 0 is reserved for the meta-sk */ ++ mptcp_for_each_bit_unset(mpcb->path_index_bits >> base, i) { ++ if (i + base < 1) ++ continue; ++ if (i + base >= sizeof(mpcb->path_index_bits) * 8) ++ break; ++ i += base; ++ mpcb->path_index_bits |= (1 << i); ++ mpcb->next_path_index = i + 1; ++ return i; ++ } ++ mptcp_for_each_bit_unset(mpcb->path_index_bits, i) { ++ if (i >= sizeof(mpcb->path_index_bits) * 8) ++ break; ++ if (i < 1) ++ continue; ++ mpcb->path_index_bits |= (1 << i); ++ mpcb->next_path_index = i + 1; ++ return i; ++ } ++ ++ return 0; ++} ++ ++static inline int mptcp_v6_is_v4_mapped(struct sock *sk) ++{ ++ return sk->sk_family == AF_INET6 && ++ ipv6_addr_type(&inet6_sk(sk)->saddr) == IPV6_ADDR_MAPPED; ++} ++ ++/* TCP and MPTCP mpc flag-depending functions */ ++u16 mptcp_select_window(struct sock *sk); ++void mptcp_init_buffer_space(struct sock *sk); ++void mptcp_tcp_set_rto(struct sock *sk); ++ ++static inline void set_mpc(struct tcp_sock *tp) ++{ ++ tp->mpc = 1; ++ ++ tp->__select_window = __mptcp_select_window; ++ tp->select_window = mptcp_select_window; ++ tp->select_initial_window = mptcp_select_initial_window; ++ tp->init_buffer_space = mptcp_init_buffer_space; ++ tp->set_rto = mptcp_tcp_set_rto; ++ tp->should_expand_sndbuf = mptcp_should_expand_sndbuf; ++} ++ ++#else /* CONFIG_MPTCP */ ++#define mptcp_debug(fmt, args...) \ ++ do { \ ++ } while (0) ++ ++/* Without MPTCP, we just do one iteration ++ * over the only socket available. This assumes that ++ * the sk/tp arg is the socket in that case. ++ */ ++#define mptcp_for_each_sk(mpcb, sk) ++#define mptcp_for_each_sk_safe(__mpcb, __sk, __temp) ++ ++static inline int mptcp_is_data_fin(const struct sk_buff *skb) ++{ ++ return 0; ++} ++static inline int mptcp_is_data_seq(const struct sk_buff *skb) ++{ ++ return 0; ++} ++static inline struct sock *mptcp_meta_sk(const struct sock *sk) ++{ ++ return NULL; ++} ++static inline struct tcp_sock *mptcp_meta_tp(const struct tcp_sock *tp) ++{ ++ return NULL; ++} ++static inline int is_meta_sk(const struct sock *sk) ++{ ++ return 0; ++} ++static inline int is_master_tp(const struct tcp_sock *tp) ++{ ++ return 0; ++} ++static inline void mptcp_purge_ofo_queue(struct tcp_sock *meta_tp) {} ++static inline void mptcp_cleanup_rbuf(const struct sock *meta_sk, int copied) {} ++static inline void mptcp_del_sock(const struct sock *sk) {} ++static inline void mptcp_reinject_data(struct sock *orig_sk, int clone_it) {} ++static inline void mptcp_update_sndbuf(const struct mptcp_cb *mpcb) {} ++static inline void mptcp_skb_entail_init(const struct tcp_sock *tp, ++ const struct sk_buff *skb) {} ++static inline void mptcp_clean_rtx_infinite(const struct sk_buff *skb, ++ const struct sock *sk) {} ++static inline void mptcp_retransmit_timer(const struct sock *meta_sk) {} ++static inline int mptcp_write_wakeup(struct sock *meta_sk) ++{ ++ return 0; ++} ++static inline void mptcp_sub_close(struct sock *sk, unsigned long delay) {} ++static inline void mptcp_set_rto(const struct sock *sk) {} ++static inline void mptcp_send_fin(const struct sock *meta_sk) {} ++static inline void mptcp_parse_options(const uint8_t *ptr, const int opsize, ++ const struct tcp_options_received *opt_rx, ++ const struct mptcp_options_received *mopt, ++ const struct sk_buff *skb) {} ++static inline void mptcp_syn_options(struct sock *sk, ++ struct tcp_out_options *opts, ++ unsigned *remaining) {} ++static inline void mptcp_synack_options(struct request_sock *req, ++ struct tcp_out_options *opts, ++ unsigned *remaining) {} ++ ++static inline void mptcp_established_options(struct sock *sk, ++ struct sk_buff *skb, ++ struct tcp_out_options *opts, ++ unsigned *size) {} ++static inline void mptcp_options_write(__be32 *ptr, struct tcp_sock *tp, ++ struct tcp_out_options *opts, ++ struct sk_buff *skb) {} ++static inline void mptcp_close(struct sock *meta_sk, long timeout) {} ++static inline int mptcp_doit(struct sock *sk) ++{ ++ return 0; ++} ++static inline int mptcp_check_req_master(const struct sock *sk, ++ const struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ const struct mptcp_options_received *mopt) ++{ ++ return 1; ++} ++static inline struct sock *mptcp_check_req_child(struct sock *sk, ++ struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ struct mptcp_options_received *mopt) ++{ ++ return NULL; ++} ++static inline unsigned int mptcp_current_mss(struct sock *meta_sk) ++{ ++ return 0; ++} ++static inline int mptcp_select_size(const struct sock *meta_sk, bool sg) ++{ ++ return 0; ++} ++static inline void mptcp_sub_close_passive(struct sock *sk) {} ++static inline bool mptcp_fallback_infinite(const struct sock *sk, int flag) ++{ ++ return false; ++} ++static inline void mptcp_init_mp_opt(const struct mptcp_options_received *mopt) {} ++static inline int mptcp_check_rtt(const struct tcp_sock *tp, int time) ++{ ++ return 0; ++} ++static inline int mptcp_check_snd_buf(const struct tcp_sock *tp) ++{ ++ return 0; ++} ++static inline int mptcp_sysctl_syn_retries(void) ++{ ++ return 0; ++} ++static inline void mptcp_send_reset(const struct sock *sk) {} ++static inline void mptcp_send_active_reset(struct sock *meta_sk, ++ gfp_t priority) {} ++static inline int mptcp_write_xmit(struct sock *sk, unsigned int mss_now, ++ int nonagle, int push_one, gfp_t gfp) ++{ ++ return 0; ++} ++static inline struct sock *mptcp_sk_clone(const struct sock *sk, int family, ++ const gfp_t priority) ++{ ++ return NULL; ++} ++static inline int mptcp_handle_options(struct sock *sk, ++ const struct tcphdr *th, ++ struct sk_buff *skb) ++{ ++ return 0; ++} ++static inline void mptcp_reset_mopt(struct tcp_sock *tp) {} ++static inline void __init mptcp_init(void) {} ++static inline int mptcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len) ++{ ++ return 0; ++} ++static inline int mptcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, ++ unsigned int mss_now, int reinject) ++{ ++ return 0; ++} ++static inline int mptso_fragment(struct sock *sk, struct sk_buff *skb, ++ unsigned int len, unsigned int mss_now, ++ gfp_t gfp, int reinject) ++{ ++ return 0; ++} ++static inline bool mptcp_sk_can_gso(const struct sock *sk) ++{ ++ return false; ++} ++static inline bool mptcp_can_sg(const struct sock *meta_sk) ++{ ++ return false; ++} ++static inline unsigned int mptcp_xmit_size_goal(struct sock *meta_sk, ++ u32 mss_now, int large_allowed) ++{ ++ return 0; ++} ++static inline void mptcp_destroy_sock(struct sock *sk) {} ++static inline int mptcp_rcv_synsent_state_process(struct sock *sk, ++ struct sock **skptr, ++ struct sk_buff *skb, ++ struct mptcp_options_received *mopt) ++{ ++ return 0; ++} ++static inline bool mptcp_can_sendpage(struct sock *sk) ++{ ++ return false; ++} ++static inline int mptcp_time_wait(struct sock *sk, struct tcp_timewait_sock *tw) ++{ ++ return 0; ++} ++static inline void mptcp_twsk_destructor(struct tcp_timewait_sock *tw) {} ++static inline void mptcp_update_tw_socks(const struct tcp_sock *tp, int state) {} ++static inline void mptcp_disconnect(struct sock *sk) {} ++static inline void mptcp_tsq_flags(struct sock *sk) {} ++static inline void mptcp_tsq_sub_deferred(struct sock *meta_sk) {} ++static inline void mptcp_hash_remove_bh(struct tcp_sock *meta_tp) {} ++static inline void mptcp_hash_remove(struct tcp_sock *meta_tp) {} ++static inline void mptcp_reqsk_new_mptcp(struct request_sock *req, ++ const struct tcp_options_received *rx_opt, ++ const struct mptcp_options_received *mopt, ++ const struct sk_buff *skb) {} ++static inline void mptcp_remove_shortcuts(const struct mptcp_cb *mpcb, ++ const struct sk_buff *skb) {} ++#endif /* CONFIG_MPTCP */ ++ ++#endif /* _MPTCP_H */ +diff --git a/include/net/mptcp_v4.h b/include/net/mptcp_v4.h +new file mode 100644 +index 0000000..047884c +--- /dev/null ++++ b/include/net/mptcp_v4.h +@@ -0,0 +1,69 @@ ++/* ++ * MPTCP implementation ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#ifndef MPTCP_V4_H_ ++#define MPTCP_V4_H_ ++ ++ ++#include <linux/in.h> ++#include <linux/skbuff.h> ++#include <net/mptcp.h> ++#include <net/request_sock.h> ++#include <net/sock.h> ++ ++extern struct request_sock_ops mptcp_request_sock_ops; ++ ++#ifdef CONFIG_MPTCP ++ ++int mptcp_v4_do_rcv(struct sock *meta_sk, struct sk_buff *skb); ++int mptcp_v4_rem_raddress(struct mptcp_cb *mpcb, u8 id); ++int mptcp_v4_add_raddress(struct mptcp_cb *mpcb, const struct in_addr *addr, ++ __be16 port, u8 id); ++void mptcp_v4_set_init_addr_bit(struct mptcp_cb *mpcb, __be32 daddr, int index); ++struct sock *mptcp_v4_search_req(const __be16 rport, const __be32 raddr, ++ const __be32 laddr, const struct net *net); ++int mptcp_init4_subsockets(struct sock *meta_sk, const struct mptcp_loc4 *loc, ++ struct mptcp_rem4 *rem); ++int mptcp_pm_v4_init(void); ++void mptcp_pm_v4_undo(void); ++u32 mptcp_v4_get_nonce(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport, ++ u32 seq); ++u64 mptcp_v4_get_key(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport); ++ ++#else ++ ++static inline int mptcp_v4_do_rcv(const struct sock *meta_sk, ++ const struct sk_buff *skb) ++{ ++ return 0; ++} ++ ++#endif /* CONFIG_MPTCP */ ++ ++#endif /* MPTCP_V4_H_ */ +diff --git a/include/net/mptcp_v6.h b/include/net/mptcp_v6.h +new file mode 100644 +index 0000000..c303208 +--- /dev/null ++++ b/include/net/mptcp_v6.h +@@ -0,0 +1,72 @@ ++/* ++ * MPTCP implementation ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#ifndef _MPTCP_V6_H ++#define _MPTCP_V6_H ++ ++#include <linux/in6.h> ++#include <net/if_inet6.h> ++ ++#include <net/mptcp.h> ++ ++extern struct request_sock_ops mptcp6_request_sock_ops; ++extern struct proto mptcpv6_prot; ++ ++#ifdef CONFIG_MPTCP ++ ++int mptcp_v6_do_rcv(struct sock *meta_sk, struct sk_buff *skb); ++int mptcp_v6_rem_raddress(struct mptcp_cb *mpcb, u8 id); ++int mptcp_v6_add_raddress(struct mptcp_cb *mpcb, const struct in6_addr *addr, ++ __be16 port, u8 id); ++void mptcp_v6_set_init_addr_bit(struct mptcp_cb *mpcb, ++ const struct in6_addr *daddr, int index); ++struct sock *mptcp_v6_search_req(const __be16 rport, const struct in6_addr *raddr, ++ const struct in6_addr *laddr, const struct net *net); ++int mptcp_init6_subsockets(struct sock *meta_sk, const struct mptcp_loc6 *loc, ++ struct mptcp_rem6 *rem); ++int mptcp_pm_v6_init(void); ++void mptcp_pm_v6_undo(void); ++struct sock *mptcp_v6v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst); ++__u32 mptcp_v6_get_nonce(const __be32 *saddr, const __be32 *daddr, ++ __be16 sport, __be16 dport, u32 seq); ++u64 mptcp_v6_get_key(const __be32 *saddr, const __be32 *daddr, ++ __be16 sport, __be16 dport); ++ ++#else /* CONFIG_MPTCP */ ++ ++static inline int mptcp_v6_do_rcv(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ return 0; ++} ++ ++#endif /* CONFIG_MPTCP */ ++ ++#endif /* _MPTCP_V6_H */ +diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h +index 991dcd9..6297c97 100644 +--- a/include/net/net_namespace.h ++++ b/include/net/net_namespace.h +@@ -15,6 +15,7 @@ + #include <net/netns/packet.h> + #include <net/netns/ipv4.h> + #include <net/netns/ipv6.h> ++#include <net/netns/mptcp.h> + #include <net/netns/sctp.h> + #include <net/netns/dccp.h> + #include <net/netns/netfilter.h> +@@ -90,6 +91,9 @@ struct net { + #if IS_ENABLED(CONFIG_IPV6) + struct netns_ipv6 ipv6; + #endif ++#if IS_ENABLED(CONFIG_MPTCP) ++ struct netns_mptcp mptcp; ++#endif + #if defined(CONFIG_IP_SCTP) || defined(CONFIG_IP_SCTP_MODULE) + struct netns_sctp sctp; + #endif +diff --git a/include/net/netns/mptcp.h b/include/net/netns/mptcp.h +new file mode 100644 +index 0000000..bad418b +--- /dev/null ++++ b/include/net/netns/mptcp.h +@@ -0,0 +1,44 @@ ++/* ++ * MPTCP implementation - MPTCP namespace ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#ifndef __NETNS_MPTCP_H__ ++#define __NETNS_MPTCP_H__ ++ ++#include <linux/compiler.h> ++ ++enum { ++ MPTCP_PM_FULLMESH = 0, ++ MPTCP_PM_MAX ++}; ++ ++struct netns_mptcp { ++ void *path_managers[MPTCP_PM_MAX]; ++}; ++ ++#endif /* __NETNS_MPTCP_H__ */ +diff --git a/include/net/request_sock.h b/include/net/request_sock.h +index 7f830ff..e79e87a 100644 +--- a/include/net/request_sock.h ++++ b/include/net/request_sock.h +@@ -164,7 +164,7 @@ struct request_sock_queue { + }; + + int reqsk_queue_alloc(struct request_sock_queue *queue, +- unsigned int nr_table_entries); ++ unsigned int nr_table_entries, gfp_t flags); + + void __reqsk_queue_destroy(struct request_sock_queue *queue); + void reqsk_queue_destroy(struct request_sock_queue *queue); +diff --git a/include/net/sock.h b/include/net/sock.h +index b9586a1..09a682e 100644 +--- a/include/net/sock.h ++++ b/include/net/sock.h +@@ -899,6 +899,16 @@ void sk_clear_memalloc(struct sock *sk); + + int sk_wait_data(struct sock *sk, long *timeo); + ++/* START - needed for MPTCP */ ++extern void sock_def_error_report(struct sock *sk); ++extern struct sock *sk_prot_alloc(struct proto *prot, gfp_t priority, ++ int family); ++extern void sock_lock_init(struct sock *sk); ++ ++extern struct lock_class_key af_callback_keys[AF_MAX]; ++extern char *const af_family_clock_key_strings[AF_MAX+1]; ++/* END - needed for MPTCP */ ++ + struct request_sock_ops; + struct timewait_sock_ops; + struct inet_hashinfo; +diff --git a/include/net/tcp.h b/include/net/tcp.h +index 743acce..db0cc04 100644 +--- a/include/net/tcp.h ++++ b/include/net/tcp.h +@@ -176,6 +176,7 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); + #define TCPOPT_SACK 5 /* SACK Block */ + #define TCPOPT_TIMESTAMP 8 /* Better RTT estimations/PAWS */ + #define TCPOPT_MD5SIG 19 /* MD5 Signature (RFC2385) */ ++#define TCPOPT_MPTCP 30 + #define TCPOPT_EXP 254 /* Experimental */ + /* Magic number to be after the option value for sharing TCP + * experimental options. See draft-ietf-tcpm-experimental-options-00.txt +@@ -234,6 +235,27 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); + */ + #define TFO_SERVER_ALWAYS 0x1000 + ++/* Flags from tcp_input.c for tcp_ack */ ++#define FLAG_DATA 0x01 /* Incoming frame contained data. */ ++#define FLAG_WIN_UPDATE 0x02 /* Incoming ACK was a window update. */ ++#define FLAG_DATA_ACKED 0x04 /* This ACK acknowledged new data. */ ++#define FLAG_RETRANS_DATA_ACKED 0x08 /* "" "" some of which was retransmitted. */ ++#define FLAG_SYN_ACKED 0x10 /* This ACK acknowledged SYN. */ ++#define FLAG_DATA_SACKED 0x20 /* New SACK. */ ++#define FLAG_ECE 0x40 /* ECE in this ACK */ ++#define FLAG_SLOWPATH 0x100 /* Do not skip RFC checks for window update.*/ ++#define FLAG_ORIG_SACK_ACKED 0x200 /* Never retransmitted data are (s)acked */ ++#define FLAG_SND_UNA_ADVANCED 0x400 /* Snd_una was changed (!= FLAG_DATA_ACKED) */ ++#define FLAG_DSACKING_ACK 0x800 /* SACK blocks contained D-SACK info */ ++#define FLAG_SACK_RENEGING 0x2000 /* snd_una advanced to a sacked seq */ ++#define FLAG_UPDATE_TS_RECENT 0x4000 /* tcp_replace_ts_recent() */ ++#define MPTCP_FLAG_DATA_ACKED 0x8000 ++ ++#define FLAG_ACKED (FLAG_DATA_ACKED|FLAG_SYN_ACKED) ++#define FLAG_NOT_DUP (FLAG_DATA|FLAG_WIN_UPDATE|FLAG_ACKED) ++#define FLAG_CA_ALERT (FLAG_DATA_SACKED|FLAG_ECE) ++#define FLAG_FORWARD_PROGRESS (FLAG_ACKED|FLAG_DATA_SACKED) ++ + extern struct inet_timewait_death_row tcp_death_row; + + /* sysctl variables for tcp */ +@@ -349,6 +371,112 @@ extern struct proto tcp_prot; + #define TCP_ADD_STATS_USER(net, field, val) SNMP_ADD_STATS_USER((net)->mib.tcp_statistics, field, val) + #define TCP_ADD_STATS(net, field, val) SNMP_ADD_STATS((net)->mib.tcp_statistics, field, val) + ++/**** START - Exports needed for MPTCP ****/ ++extern const struct inet_connection_sock_af_ops ipv4_specific; ++extern const struct inet_connection_sock_af_ops ipv6_specific; ++extern const struct inet_connection_sock_af_ops ipv6_mapped; ++extern const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops; ++extern const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops; ++ ++struct mptcp_options_received; ++ ++int tcp_close_state(struct sock *sk); ++void tcp_push(struct sock *sk, int flags, int mss_now, int nonagle, int ++ size_goal); ++void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss_now, ++ const struct sk_buff *skb); ++int tcp_xmit_probe_skb(struct sock *sk, int urgent); ++void tcp_cwnd_validate(struct sock *sk); ++void tcp_event_new_data_sent(struct sock *sk, const struct sk_buff *skb); ++int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, ++ gfp_t gfp_mask); ++unsigned int tcp_mss_split_point(const struct sock *sk, ++ const struct sk_buff *skb, ++ unsigned int mss_now, ++ unsigned int max_segs, ++ int nonagle); ++bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb); ++bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buff *skb, ++ unsigned int cur_mss, int nonagle); ++bool tcp_snd_wnd_test(const struct tcp_sock *tp, const struct sk_buff *skb, ++ unsigned int cur_mss); ++unsigned int tcp_cwnd_test(const struct tcp_sock *tp, const struct sk_buff *skb); ++int tcp_mtu_probe(struct sock *sk); ++int tcp_init_tso_segs(const struct sock *sk, struct sk_buff *skb, ++ unsigned int mss_now); ++void __pskb_trim_head(struct sk_buff *skb, int len); ++void tcp_queue_skb(struct sock *sk, struct sk_buff *skb); ++void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags); ++void tcp_reset(struct sock *sk); ++bool tcp_may_update_window(const struct tcp_sock *tp, const u32 ack, ++ const u32 ack_seq, const u32 nwin); ++bool tcp_urg_mode(const struct tcp_sock *tp); ++void tcp_ack_probe(struct sock *sk); ++void tcp_rearm_rto(struct sock *sk); ++int tcp_write_timeout(struct sock *sk); ++bool retransmits_timed_out(struct sock *sk, unsigned int boundary, ++ unsigned int timeout, bool syn_set); ++void tcp_write_err(struct sock *sk); ++void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int decr); ++void tcp_set_skb_tso_segs(const struct sock *sk, struct sk_buff *skb, ++ unsigned int mss_now); ++ ++int tcp_v4_rtx_synack(struct sock *sk, struct request_sock *req); ++void tcp_v4_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req); ++__u32 tcp_v4_init_sequence(const struct sk_buff *skb); ++int tcp_v4_send_synack(struct sock *sk, struct dst_entry *dst, ++ struct request_sock *req, ++ u16 queue_mapping); ++void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb); ++struct ip_options_rcu *tcp_v4_save_options(struct sk_buff *skb); ++struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb); ++void tcp_v4_reqsk_destructor(struct request_sock *req); ++ ++int tcp_v6_rtx_synack(struct sock *sk, struct request_sock *req); ++void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req); ++__u32 tcp_v6_init_sequence(const struct sk_buff *skb); ++int tcp_v6_send_synack(struct sock *sk, struct dst_entry *dst, ++ struct flowi6 *fl6, struct request_sock *req, ++ u16 queue_mapping); ++void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb); ++int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb); ++int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); ++void tcp_v6_destroy_sock(struct sock *sk); ++void inet6_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb); ++void tcp_v6_hash(struct sock *sk); ++struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb); ++struct sock *tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst); ++void tcp_v6_reqsk_destructor(struct request_sock *req); ++ ++void sock_valbool_flag(struct sock *sk, int bit, int valbool); ++unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now, ++ int large_allowed); ++u32 tcp_tso_acked(struct sock *sk, struct sk_buff *skb); ++ ++void skb_clone_fraglist(struct sk_buff *skb); ++void copy_skb_header(struct sk_buff *new, const struct sk_buff *old); ++ ++void inet_twsk_free(struct inet_timewait_sock *tw); ++/* These states need RST on ABORT according to RFC793 */ ++static inline bool tcp_need_reset(int state) ++{ ++ return (1 << state) & ++ (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT | TCPF_FIN_WAIT1 | ++ TCPF_FIN_WAIT2 | TCPF_SYN_RECV); ++} ++ ++bool tcp_dma_try_early_copy(struct sock *sk, struct sk_buff *skb, ++ int hlen); ++int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb, int hdrlen, ++ bool *fragstolen); ++bool tcp_try_coalesce(struct sock *sk, struct sk_buff *to, ++ struct sk_buff *from, bool *fragstolen); ++/**** END - Exports needed for MPTCP ****/ ++ + void tcp_tasklet_init(void); + + void tcp_v4_err(struct sk_buff *skb, u32); +@@ -445,6 +573,7 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + size_t len, int nonblock, int flags, int *addr_len); + void tcp_parse_options(const struct sk_buff *skb, + struct tcp_options_received *opt_rx, ++ struct mptcp_options_received *mopt_rx, + int estab, struct tcp_fastopen_cookie *foc); + const u8 *tcp_parse_md5sig_option(const struct tcphdr *th); + +@@ -557,11 +686,15 @@ void tcp_send_delayed_ack(struct sock *sk); + void tcp_send_loss_probe(struct sock *sk); + bool tcp_schedule_loss_probe(struct sock *sk); + ++u16 tcp_select_window(struct sock *sk); ++ + /* tcp_input.c */ + void tcp_cwnd_application_limited(struct sock *sk); + void tcp_resume_early_retransmit(struct sock *sk); + void tcp_rearm_rto(struct sock *sk); + void tcp_reset(struct sock *sk); ++void tcp_set_rto(struct sock *sk); ++bool tcp_should_expand_sndbuf(const struct sock *sk); + + /* tcp_timer.c */ + void tcp_init_xmit_timers(struct sock *); +@@ -705,14 +838,24 @@ void tcp_send_window_probe(struct sock *sk); + */ + struct tcp_skb_cb { + union { +- struct inet_skb_parm h4; ++ union { ++ struct inet_skb_parm h4; + #if IS_ENABLED(CONFIG_IPV6) +- struct inet6_skb_parm h6; ++ struct inet6_skb_parm h6; + #endif +- } header; /* For incoming frames */ ++ } header; /* For incoming frames */ ++#ifdef CONFIG_MPTCP ++ __u32 path_mask; /* path indices that tried to send this skb */ ++#endif ++ }; + __u32 seq; /* Starting sequence number */ + __u32 end_seq; /* SEQ + FIN + SYN + datalen */ + __u32 when; /* used to compute rtt's */ ++#ifdef CONFIG_MPTCP ++ __u8 mptcp_flags; /* flags for the MPTCP layer */ ++ __u8 dss_off; /* Number of 4-byte words until ++ * seq-number */ ++#endif + __u8 tcp_flags; /* TCP header flags. (tcp[13]) */ + + __u8 sacked; /* State flags for SACK/FACK. */ +@@ -1058,7 +1201,8 @@ u32 tcp_default_init_rwnd(u32 mss); + /* Determine a window scaling and initial window to offer. */ + void tcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd, + __u32 *window_clamp, int wscale_ok, +- __u8 *rcv_wscale, __u32 init_rcv_wnd); ++ __u8 *rcv_wscale, __u32 init_rcv_wnd, ++ const struct sock *sk); + + static inline int tcp_win_from_space(int space) + { +@@ -1070,12 +1214,18 @@ static inline int tcp_win_from_space(int space) + /* Note: caller must be prepared to deal with negative returns */ + static inline int tcp_space(const struct sock *sk) + { ++ if (tcp_sk(sk)->mpc) ++ sk = tcp_sk(sk)->meta_sk; ++ + return tcp_win_from_space(sk->sk_rcvbuf - + atomic_read(&sk->sk_rmem_alloc)); + } + + static inline int tcp_full_space(const struct sock *sk) + { ++ if (tcp_sk(sk)->mpc) ++ sk = tcp_sk(sk)->meta_sk; ++ + return tcp_win_from_space(sk->sk_rcvbuf); + } + +@@ -1090,6 +1240,7 @@ static inline void tcp_openreq_init(struct request_sock *req, + tcp_rsk(req)->rcv_isn = TCP_SKB_CB(skb)->seq; + tcp_rsk(req)->rcv_nxt = TCP_SKB_CB(skb)->seq + 1; + tcp_rsk(req)->snt_synack = 0; ++ tcp_rsk(req)->saw_mpc = 0; + req->mss = rx_opt->mss_clamp; + req->ts_recent = rx_opt->saw_tstamp ? rx_opt->rcv_tsval : 0; + ireq->tstamp_ok = rx_opt->tstamp_ok; +diff --git a/include/uapi/linux/if.h b/include/uapi/linux/if.h +index d758163..3d81e49 100644 +--- a/include/uapi/linux/if.h ++++ b/include/uapi/linux/if.h +@@ -53,6 +53,9 @@ + + #define IFF_ECHO 0x40000 /* echo sent packets */ + ++#define IFF_NOMULTIPATH 0x80000 /* Disable for MPTCP */ ++#define IFF_MPBACKUP 0x100000 /* Use as backup path for MPTCP */ ++ + #define IFF_VOLATILE (IFF_LOOPBACK|IFF_POINTOPOINT|IFF_BROADCAST|IFF_ECHO|\ + IFF_MASTER|IFF_SLAVE|IFF_RUNNING|IFF_LOWER_UP|IFF_DORMANT) + +diff --git a/include/uapi/linux/tcp.h b/include/uapi/linux/tcp.h +index 377f1e5..2ffcb03 100644 +--- a/include/uapi/linux/tcp.h ++++ b/include/uapi/linux/tcp.h +@@ -112,6 +112,7 @@ enum { + #define TCP_FASTOPEN 23 /* Enable FastOpen on listeners */ + #define TCP_TIMESTAMP 24 + #define TCP_NOTSENT_LOWAT 25 /* limit number of unsent bytes in write queue */ ++#define MPTCP_ENABLED 26 + + struct tcp_repair_opt { + __u32 opt_code; +diff --git a/net/Kconfig b/net/Kconfig +index e411046..3e4b278 100644 +--- a/net/Kconfig ++++ b/net/Kconfig +@@ -79,6 +79,7 @@ if INET + source "net/ipv4/Kconfig" + source "net/ipv6/Kconfig" + source "net/netlabel/Kconfig" ++source "net/mptcp/Kconfig" + + endif # if INET + +diff --git a/net/Makefile b/net/Makefile +index cbbbe6d..244bac1 100644 +--- a/net/Makefile ++++ b/net/Makefile +@@ -20,6 +20,7 @@ obj-$(CONFIG_INET) += ipv4/ + obj-$(CONFIG_XFRM) += xfrm/ + obj-$(CONFIG_UNIX) += unix/ + obj-$(CONFIG_NET) += ipv6/ ++obj-$(CONFIG_MPTCP) += mptcp/ + obj-$(CONFIG_PACKET) += packet/ + obj-$(CONFIG_NET_KEY) += key/ + obj-$(CONFIG_BRIDGE) += bridge/ +diff --git a/net/core/dev.c b/net/core/dev.c +index 45fa2f1..3cfdbc0 100644 +--- a/net/core/dev.c ++++ b/net/core/dev.c +@@ -5271,7 +5271,7 @@ int __dev_change_flags(struct net_device *dev, unsigned int flags) + + dev->flags = (flags & (IFF_DEBUG | IFF_NOTRAILERS | IFF_NOARP | + IFF_DYNAMIC | IFF_MULTICAST | IFF_PORTSEL | +- IFF_AUTOMEDIA)) | ++ IFF_AUTOMEDIA | IFF_NOMULTIPATH | IFF_MPBACKUP)) | + (dev->flags & (IFF_UP | IFF_VOLATILE | IFF_PROMISC | + IFF_ALLMULTI)); + +diff --git a/net/core/request_sock.c b/net/core/request_sock.c +index 4425148..e128f08 100644 +--- a/net/core/request_sock.c ++++ b/net/core/request_sock.c +@@ -38,7 +38,8 @@ int sysctl_max_syn_backlog = 256; + EXPORT_SYMBOL(sysctl_max_syn_backlog); + + int reqsk_queue_alloc(struct request_sock_queue *queue, +- unsigned int nr_table_entries) ++ unsigned int nr_table_entries, ++ gfp_t flags) + { + size_t lopt_size = sizeof(struct listen_sock); + struct listen_sock *lopt; +@@ -48,9 +49,11 @@ int reqsk_queue_alloc(struct request_sock_queue *queue, + nr_table_entries = roundup_pow_of_two(nr_table_entries + 1); + lopt_size += nr_table_entries * sizeof(struct request_sock *); + if (lopt_size > PAGE_SIZE) +- lopt = vzalloc(lopt_size); ++ lopt = __vmalloc(lopt_size, ++ flags | __GFP_HIGHMEM | __GFP_ZERO, ++ PAGE_KERNEL); + else +- lopt = kzalloc(lopt_size, GFP_KERNEL); ++ lopt = kzalloc(lopt_size, flags); + if (lopt == NULL) + return -ENOMEM; + +diff --git a/net/core/skbuff.c b/net/core/skbuff.c +index 90b96a1..2564d89 100644 +--- a/net/core/skbuff.c ++++ b/net/core/skbuff.c +@@ -472,7 +472,7 @@ static inline void skb_drop_fraglist(struct sk_buff *skb) + skb_drop_list(&skb_shinfo(skb)->frag_list); + } + +-static void skb_clone_fraglist(struct sk_buff *skb) ++void skb_clone_fraglist(struct sk_buff *skb) + { + struct sk_buff *list; + +@@ -894,7 +894,7 @@ static void skb_headers_offset_update(struct sk_buff *skb, int off) + skb->inner_mac_header += off; + } + +-static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old) ++void copy_skb_header(struct sk_buff *new, const struct sk_buff *old) + { + __copy_skb_header(new, old); + +diff --git a/net/core/sock.c b/net/core/sock.c +index c0fc6bd..7314971 100644 +--- a/net/core/sock.c ++++ b/net/core/sock.c +@@ -231,7 +231,7 @@ static const char *const af_family_slock_key_strings[AF_MAX+1] = { + "slock-AF_IEEE802154", "slock-AF_CAIF" , "slock-AF_ALG" , + "slock-AF_NFC" , "slock-AF_VSOCK" ,"slock-AF_MAX" + }; +-static const char *const af_family_clock_key_strings[AF_MAX+1] = { ++char *const af_family_clock_key_strings[AF_MAX+1] = { + "clock-AF_UNSPEC", "clock-AF_UNIX" , "clock-AF_INET" , + "clock-AF_AX25" , "clock-AF_IPX" , "clock-AF_APPLETALK", + "clock-AF_NETROM", "clock-AF_BRIDGE" , "clock-AF_ATMPVC" , +@@ -252,7 +252,7 @@ static const char *const af_family_clock_key_strings[AF_MAX+1] = { + * sk_callback_lock locking rules are per-address-family, + * so split the lock classes by using a per-AF key: + */ +-static struct lock_class_key af_callback_keys[AF_MAX]; ++struct lock_class_key af_callback_keys[AF_MAX]; + + /* Take into consideration the size of the struct sk_buff overhead in the + * determination of these values, since that is non-constant across +@@ -602,7 +602,7 @@ out: + return ret; + } + +-static inline void sock_valbool_flag(struct sock *sk, int bit, int valbool) ++void sock_valbool_flag(struct sock *sk, int bit, int valbool) + { + if (valbool) + sock_set_flag(sk, bit); +@@ -1204,7 +1204,7 @@ lenout: + * + * (We also register the sk_lock with the lock validator.) + */ +-static inline void sock_lock_init(struct sock *sk) ++void sock_lock_init(struct sock *sk) + { + sock_lock_init_class_and_name(sk, + af_family_slock_key_strings[sk->sk_family], +@@ -1252,7 +1252,7 @@ void sk_prot_clear_portaddr_nulls(struct sock *sk, int size) + } + EXPORT_SYMBOL(sk_prot_clear_portaddr_nulls); + +-static struct sock *sk_prot_alloc(struct proto *prot, gfp_t priority, ++struct sock *sk_prot_alloc(struct proto *prot, gfp_t priority, + int family) + { + struct sock *sk; +@@ -2184,7 +2184,7 @@ static void sock_def_wakeup(struct sock *sk) + rcu_read_unlock(); + } + +-static void sock_def_error_report(struct sock *sk) ++void sock_def_error_report(struct sock *sk) + { + struct socket_wq *wq; + +diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig +index 05c57f0..630434d 100644 +--- a/net/ipv4/Kconfig ++++ b/net/ipv4/Kconfig +@@ -556,6 +556,30 @@ config TCP_CONG_ILLINOIS + For further details see: + http://www.ews.uiuc.edu/~shaoliu/tcpillinois/index.html + ++config TCP_CONG_COUPLED ++ tristate "MPTCP COUPLED CONGESTION CONTROL" ++ depends on MPTCP ++ default n ++ ---help--- ++ MultiPath TCP Coupled Congestion Control ++ To enable it, just put 'coupled' in tcp_congestion_control ++ ++config TCP_CONG_OLIA ++ tristate "MPTCP Opportunistic Linked Increase" ++ depends on MPTCP ++ default n ++ ---help--- ++ MultiPath TCP Opportunistic Linked Increase Congestion Control ++ To enable it, just put 'olia' in tcp_congestion_control ++ ++config TCP_CONG_WVEGAS ++ tristate "MPTCP WVEGAS CONGESTION CONTROL" ++ depends on MPTCP ++ default n ++ ---help--- ++ wVegas congestion control for MPTCP ++ To enable it, just put 'wvegas' in tcp_congestion_control ++ + choice + prompt "Default TCP congestion control" + default DEFAULT_CUBIC +@@ -584,6 +608,15 @@ choice + config DEFAULT_WESTWOOD + bool "Westwood" if TCP_CONG_WESTWOOD=y + ++ config DEFAULT_COUPLED ++ bool "Coupled" if TCP_CONG_COUPLED=y ++ ++ config DEFAULT_OLIA ++ bool "Olia" if TCP_CONG_OLIA=y ++ ++ config DEFAULT_WVEGAS ++ bool "Wvegas" if TCP_CONG_WVEGAS=y ++ + config DEFAULT_RENO + bool "Reno" + +@@ -605,6 +638,8 @@ config DEFAULT_TCP_CONG + default "vegas" if DEFAULT_VEGAS + default "westwood" if DEFAULT_WESTWOOD + default "veno" if DEFAULT_VENO ++ default "coupled" if DEFAULT_COUPLED ++ default "wvegas" if DEFAULT_WVEGAS + default "reno" if DEFAULT_RENO + default "cubic" + +diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c +index 19ab78a..567918a 100644 +--- a/net/ipv4/af_inet.c ++++ b/net/ipv4/af_inet.c +@@ -104,6 +104,7 @@ + #include <net/ip_fib.h> + #include <net/inet_connection_sock.h> + #include <net/tcp.h> ++#include <net/mptcp.h> + #include <net/udp.h> + #include <net/udplite.h> + #include <net/ping.h> +@@ -246,8 +247,7 @@ EXPORT_SYMBOL(inet_listen); + * Create an inet socket. + */ + +-static int inet_create(struct net *net, struct socket *sock, int protocol, +- int kern) ++int inet_create(struct net *net, struct socket *sock, int protocol, int kern) + { + struct sock *sk; + struct inet_protosw *answer; +@@ -679,6 +679,23 @@ int inet_accept(struct socket *sock, struct socket *newsock, int flags) + lock_sock(sk2); + + sock_rps_record_flow(sk2); ++ ++ if (sk2->sk_protocol == IPPROTO_TCP && tcp_sk(sk2)->mpc) { ++ struct sock *sk_it = sk2; ++ ++ mptcp_for_each_sk(tcp_sk(sk2)->mpcb, sk_it) ++ sock_rps_record_flow(sk_it); ++ ++ if (tcp_sk(sk2)->mpcb->master_sk) { ++ sk_it = tcp_sk(sk2)->mpcb->master_sk; ++ ++ write_lock_bh(&sk_it->sk_callback_lock); ++ sk_it->sk_wq = newsock->wq; ++ sk_it->sk_socket = newsock; ++ write_unlock_bh(&sk_it->sk_callback_lock); ++ } ++ } ++ + WARN_ON(!((1 << sk2->sk_state) & + (TCPF_ESTABLISHED | TCPF_SYN_RECV | + TCPF_CLOSE_WAIT | TCPF_CLOSE))); +@@ -1767,6 +1784,9 @@ static int __init inet_init(void) + + ip_init(); + ++ /* We must initialize MPTCP before TCP. */ ++ mptcp_init(); ++ + tcp_v4_init(); + + /* Setup TCP slab cache for open requests. */ +diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c +index 0d1e2cb..423dfb6 100644 +--- a/net/ipv4/inet_connection_sock.c ++++ b/net/ipv4/inet_connection_sock.c +@@ -23,6 +23,7 @@ + #include <net/route.h> + #include <net/tcp_states.h> + #include <net/xfrm.h> ++#include <net/mptcp.h> + + #ifdef INET_CSK_DEBUG + const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; +@@ -468,8 +469,8 @@ no_route: + } + EXPORT_SYMBOL_GPL(inet_csk_route_child_sock); + +-static inline u32 inet_synq_hash(const __be32 raddr, const __be16 rport, +- const u32 rnd, const u32 synq_hsize) ++u32 inet_synq_hash(const __be32 raddr, const __be16 rport, const u32 rnd, ++ const u32 synq_hsize) + { + return jhash_2words((__force u32)raddr, (__force u32)rport, rnd) & (synq_hsize - 1); + } +@@ -667,7 +668,12 @@ struct sock *inet_csk_clone_lock(const struct sock *sk, + const struct request_sock *req, + const gfp_t priority) + { +- struct sock *newsk = sk_clone_lock(sk, priority); ++ struct sock *newsk; ++ ++ if (sk->sk_protocol == IPPROTO_TCP && tcp_sk(sk)->mpc) ++ newsk = mptcp_sk_clone(sk, req->rsk_ops->family, priority); ++ else ++ newsk = sk_clone_lock(sk, priority); + + if (newsk != NULL) { + struct inet_connection_sock *newicsk = inet_csk(newsk); +@@ -744,7 +750,8 @@ int inet_csk_listen_start(struct sock *sk, const int nr_table_entries) + { + struct inet_sock *inet = inet_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); +- int rc = reqsk_queue_alloc(&icsk->icsk_accept_queue, nr_table_entries); ++ int rc = reqsk_queue_alloc(&icsk->icsk_accept_queue, nr_table_entries, ++ GFP_KERNEL); + + if (rc != 0) + return rc; +@@ -802,9 +809,14 @@ void inet_csk_listen_stop(struct sock *sk) + + while ((req = acc_req) != NULL) { + struct sock *child = req->sk; ++ bool mutex_taken = false; + + acc_req = req->dl_next; + ++ if (is_meta_sk(child)) { ++ mutex_lock(&tcp_sk(child)->mpcb->mpcb_mutex); ++ mutex_taken = true; ++ } + local_bh_disable(); + bh_lock_sock(child); + WARN_ON(sock_owned_by_user(child)); +@@ -833,6 +845,8 @@ void inet_csk_listen_stop(struct sock *sk) + + bh_unlock_sock(child); + local_bh_enable(); ++ if (mutex_taken) ++ mutex_unlock(&tcp_sk(child)->mpcb->mpcb_mutex); + sock_put(child); + + sk_acceptq_removed(sk); +diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c +index f2ed13c..f08addc 100644 +--- a/net/ipv4/syncookies.c ++++ b/net/ipv4/syncookies.c +@@ -284,7 +284,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb, + + /* check for timestamp cookie support */ + memset(&tcp_opt, 0, sizeof(tcp_opt)); +- tcp_parse_options(skb, &tcp_opt, 0, NULL); ++ tcp_parse_options(skb, &tcp_opt, NULL, 0, NULL); + + if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok)) + goto out; +@@ -354,10 +354,10 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb, + /* Try to redo what tcp_v4_send_synack did. */ + req->window_clamp = tp->window_clamp ? :dst_metric(&rt->dst, RTAX_WINDOW); + +- tcp_select_initial_window(tcp_full_space(sk), req->mss, ++ tp->select_initial_window(tcp_full_space(sk), req->mss, + &req->rcv_wnd, &req->window_clamp, + ireq->wscale_ok, &rcv_wscale, +- dst_metric(&rt->dst, RTAX_INITRWND)); ++ dst_metric(&rt->dst, RTAX_INITRWND), sk); + + ireq->rcv_wscale = rcv_wscale; + +diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c +index 97c8f56..be72a40 100644 +--- a/net/ipv4/tcp.c ++++ b/net/ipv4/tcp.c +@@ -271,6 +271,7 @@ + + #include <net/icmp.h> + #include <net/inet_common.h> ++#include <net/mptcp.h> + #include <net/tcp.h> + #include <net/xfrm.h> + #include <net/ip.h> +@@ -419,6 +420,9 @@ void tcp_init_sock(struct sock *sk) + sk->sk_sndbuf = sysctl_tcp_wmem[1]; + sk->sk_rcvbuf = sysctl_tcp_rmem[1]; + ++ /* Set function pointers in tcp_sock to tcp functions. */ ++ mptcp_init_tcp_sock(tp); ++ + local_bh_disable(); + sock_update_memcg(sk); + sk_sockets_allocated_inc(sk); +@@ -607,6 +611,8 @@ static inline void skb_entail(struct sock *sk, struct sk_buff *skb) + tcb->seq = tcb->end_seq = tp->write_seq; + tcb->tcp_flags = TCPHDR_ACK; + tcb->sacked = 0; ++ if (tp->mpc) ++ mptcp_skb_entail_init(tp, skb); + skb_header_release(skb); + tcp_add_write_queue_tail(sk, skb); + sk->sk_wmem_queued += skb->truesize; +@@ -640,8 +646,8 @@ static bool tcp_should_autocork(struct sock *sk, struct sk_buff *skb, + atomic_read(&sk->sk_wmem_alloc) > skb->truesize; + } + +-static void tcp_push(struct sock *sk, int flags, int mss_now, +- int nonagle, int size_goal) ++void tcp_push(struct sock *sk, int flags, int mss_now, int nonagle, ++ int size_goal) + { + struct tcp_sock *tp = tcp_sk(sk); + struct sk_buff *skb; +@@ -726,6 +732,14 @@ ssize_t tcp_splice_read(struct socket *sock, loff_t *ppos, + int ret; + + sock_rps_record_flow(sk); ++ ++#ifdef CONFIG_MPTCP ++ if (tcp_sk(sk)->mpc) { ++ struct sock *sk_it; ++ mptcp_for_each_sk(tcp_sk(sk)->mpcb, sk_it) ++ sock_rps_record_flow(sk_it); ++ } ++#endif + /* + * We can't seek on a socket input + */ +@@ -821,8 +835,7 @@ struct sk_buff *sk_stream_alloc_skb(struct sock *sk, int size, gfp_t gfp) + return NULL; + } + +-static unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now, +- int large_allowed) ++unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now, int large_allowed) + { + struct tcp_sock *tp = tcp_sk(sk); + u32 xmit_size_goal, old_size_goal; +@@ -872,8 +885,13 @@ static int tcp_send_mss(struct sock *sk, int *size_goal, int flags) + { + int mss_now; + +- mss_now = tcp_current_mss(sk); +- *size_goal = tcp_xmit_size_goal(sk, mss_now, !(flags & MSG_OOB)); ++ if (tcp_sk(sk)->mpc) { ++ mss_now = mptcp_current_mss(sk); ++ *size_goal = mptcp_xmit_size_goal(sk, mss_now, !(flags & MSG_OOB)); ++ } else { ++ mss_now = tcp_current_mss(sk); ++ *size_goal = tcp_xmit_size_goal(sk, mss_now, !(flags & MSG_OOB)); ++ } + + return mss_now; + } +@@ -897,6 +915,26 @@ static ssize_t do_tcp_sendpages(struct sock *sk, struct page *page, int offset, + goto out_err; + } + ++ if (tp->mpc) { ++ struct sock *sk_it = sk; ++ ++ /* We must check this with socket-lock hold because we iterate ++ * over the subflows. ++ */ ++ if (!mptcp_can_sendpage(sk)) { ++ ssize_t ret; ++ ++ release_sock(sk); ++ ret = sock_no_sendpage(sk->sk_socket, page, offset, ++ size, flags); ++ lock_sock(sk); ++ return ret; ++ } ++ ++ mptcp_for_each_sk(tp->mpcb, sk_it) ++ sock_rps_record_flow(sk_it); ++ } ++ + clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags); + + mss_now = tcp_send_mss(sk, &size_goal, flags); +@@ -1001,8 +1039,9 @@ int tcp_sendpage(struct sock *sk, struct page *page, int offset, + { + ssize_t res; + +- if (!(sk->sk_route_caps & NETIF_F_SG) || +- !(sk->sk_route_caps & NETIF_F_ALL_CSUM)) ++ /* If MPTCP is enabled, we check it later after establishment */ ++ if (!tcp_sk(sk)->mpc && (!(sk->sk_route_caps & NETIF_F_SG) || ++ !(sk->sk_route_caps & NETIF_F_ALL_CSUM))) + return sock_no_sendpage(sk->sk_socket, page, offset, size, + flags); + +@@ -1018,6 +1057,9 @@ static inline int select_size(const struct sock *sk, bool sg) + const struct tcp_sock *tp = tcp_sk(sk); + int tmp = tp->mss_cache; + ++ if (tp->mpc) ++ return mptcp_select_size(sk, sg); ++ + if (sg) { + if (sk_can_gso(sk)) { + /* Small frames wont use a full page: +@@ -1105,6 +1147,12 @@ int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + goto do_error; + } + ++ if (tp->mpc) { ++ struct sock *sk_it = sk; ++ mptcp_for_each_sk(tp->mpcb, sk_it) ++ sock_rps_record_flow(sk_it); ++ } ++ + if (unlikely(tp->repair)) { + if (tp->repair_queue == TCP_RECV_QUEUE) { + copied = tcp_send_rcvq(sk, msg, size); +@@ -1132,7 +1180,10 @@ int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + if (sk->sk_err || (sk->sk_shutdown & SEND_SHUTDOWN)) + goto out_err; + +- sg = !!(sk->sk_route_caps & NETIF_F_SG); ++ if (tp->mpc) ++ sg = mptcp_can_sg(sk); ++ else ++ sg = !!(sk->sk_route_caps & NETIF_F_SG); + + while (--iovlen >= 0) { + size_t seglen = iov->iov_len; +@@ -1183,8 +1234,15 @@ new_segment: + + /* + * Check whether we can use HW checksum. ++ * ++ * If dss-csum is enabled, we do not do hw-csum. ++ * In case of non-mptcp we check the ++ * device-capabilities. ++ * In case of mptcp, hw-csum's will be handled ++ * later in mptcp_write_xmit. + */ +- if (sk->sk_route_caps & NETIF_F_ALL_CSUM) ++ if (((tp->mpc && !tp->mpcb->dss_csum) || !tp->mpc) && ++ (tp->mpc || sk->sk_route_caps & NETIF_F_ALL_CSUM)) + skb->ip_summed = CHECKSUM_PARTIAL; + + skb_entail(sk, skb); +@@ -1385,6 +1443,11 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied) + + struct sk_buff *skb = skb_peek(&sk->sk_receive_queue); + ++ if (is_meta_sk(sk)) { ++ mptcp_cleanup_rbuf(sk, copied); ++ return; ++ } ++ + WARN(skb && !before(tp->copied_seq, TCP_SKB_CB(skb)->end_seq), + "cleanup rbuf bug: copied %X seq %X rcvnxt %X\n", + tp->copied_seq, TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt); +@@ -1421,7 +1484,7 @@ void tcp_cleanup_rbuf(struct sock *sk, int copied) + + /* Optimize, __tcp_select_window() is not cheap. */ + if (2*rcv_window_now <= tp->window_clamp) { +- __u32 new_window = __tcp_select_window(sk); ++ __u32 new_window = tp->__select_window(sk); + + /* Send ACK now, if this read freed lots of space + * in our buffer. Certainly, new_window is new window. +@@ -1622,6 +1685,14 @@ int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, + + lock_sock(sk); + ++#ifdef CONFIG_MPTCP ++ if (tp->mpc) { ++ struct sock *sk_it; ++ mptcp_for_each_sk(tp->mpcb, sk_it) ++ sock_rps_record_flow(sk_it); ++ } ++#endif ++ + err = -ENOTCONN; + if (sk->sk_state == TCP_LISTEN) + goto out; +@@ -2069,7 +2140,7 @@ static const unsigned char new_state[16] = { + /* TCP_CLOSING */ TCP_CLOSING, + }; + +-static int tcp_close_state(struct sock *sk) ++int tcp_close_state(struct sock *sk) + { + int next = (int)new_state[sk->sk_state]; + int ns = next & TCP_STATE_MASK; +@@ -2098,8 +2169,12 @@ void tcp_shutdown(struct sock *sk, int how) + (TCPF_ESTABLISHED | TCPF_SYN_SENT | + TCPF_SYN_RECV | TCPF_CLOSE_WAIT)) { + /* Clear out any half completed packets. FIN if needed. */ +- if (tcp_close_state(sk)) +- tcp_send_fin(sk); ++ if (tcp_close_state(sk)) { ++ if (!is_meta_sk(sk)) ++ tcp_send_fin(sk); ++ else ++ mptcp_send_fin(sk); ++ } + } + } + EXPORT_SYMBOL(tcp_shutdown); +@@ -2124,6 +2199,11 @@ void tcp_close(struct sock *sk, long timeout) + int data_was_unread = 0; + int state; + ++ if (is_meta_sk(sk)) { ++ mptcp_close(sk, timeout); ++ return; ++ } ++ + lock_sock(sk); + sk->sk_shutdown = SHUTDOWN_MASK; + +@@ -2290,15 +2370,6 @@ out: + } + EXPORT_SYMBOL(tcp_close); + +-/* These states need RST on ABORT according to RFC793 */ +- +-static inline bool tcp_need_reset(int state) +-{ +- return (1 << state) & +- (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT | TCPF_FIN_WAIT1 | +- TCPF_FIN_WAIT2 | TCPF_SYN_RECV); +-} +- + int tcp_disconnect(struct sock *sk, int flags) + { + struct inet_sock *inet = inet_sk(sk); +@@ -2339,6 +2410,13 @@ int tcp_disconnect(struct sock *sk, int flags) + if (!(sk->sk_userlocks & SOCK_BINDADDR_LOCK)) + inet_reset_saddr(sk); + ++ if (is_meta_sk(sk)) { ++ mptcp_disconnect(sk); ++ } else { ++ if (tp->inside_tk_table) ++ mptcp_hash_remove_bh(tp); ++ } ++ + sk->sk_shutdown = 0; + sock_reset_flag(sk, SOCK_DONE); + tp->srtt = 0; +@@ -2698,6 +2776,18 @@ static int do_tcp_setsockopt(struct sock *sk, int level, + tp->notsent_lowat = val; + sk->sk_write_space(sk); + break; ++#ifdef CONFIG_MPTCP ++ case MPTCP_ENABLED: ++ if (sk->sk_state == TCP_CLOSE || sk->sk_state == TCP_LISTEN) { ++ if (val) ++ tp->mptcp_enabled = 1; ++ else ++ tp->mptcp_enabled = 0; ++ } else { ++ err = -EPERM; ++ } ++ break; ++#endif + default: + err = -ENOPROTOOPT; + break; +@@ -2917,6 +3007,11 @@ static int do_tcp_getsockopt(struct sock *sk, int level, + case TCP_NOTSENT_LOWAT: + val = tp->notsent_lowat; + break; ++#ifdef CONFIG_MPTCP ++ case MPTCP_ENABLED: ++ val = tp->mptcp_enabled; ++ break; ++#endif + default: + return -ENOPROTOOPT; + } +@@ -3106,8 +3201,11 @@ void tcp_done(struct sock *sk) + if (sk->sk_state == TCP_SYN_SENT || sk->sk_state == TCP_SYN_RECV) + TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_ATTEMPTFAILS); + ++ WARN_ON(sk->sk_state == TCP_CLOSE); + tcp_set_state(sk, TCP_CLOSE); ++ + tcp_clear_xmit_timers(sk); ++ + if (req != NULL) + reqsk_fastopen_remove(sk, req, false); + +diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c +index eeaac39..cb06531 100644 +--- a/net/ipv4/tcp_input.c ++++ b/net/ipv4/tcp_input.c +@@ -74,6 +74,9 @@ + #include <linux/ipsec.h> + #include <asm/unaligned.h> + #include <net/netdma.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#include <net/mptcp_v6.h> + + int sysctl_tcp_timestamps __read_mostly = 1; + int sysctl_tcp_window_scaling __read_mostly = 1; +@@ -99,25 +102,6 @@ int sysctl_tcp_thin_dupack __read_mostly; + int sysctl_tcp_moderate_rcvbuf __read_mostly = 1; + int sysctl_tcp_early_retrans __read_mostly = 3; + +-#define FLAG_DATA 0x01 /* Incoming frame contained data. */ +-#define FLAG_WIN_UPDATE 0x02 /* Incoming ACK was a window update. */ +-#define FLAG_DATA_ACKED 0x04 /* This ACK acknowledged new data. */ +-#define FLAG_RETRANS_DATA_ACKED 0x08 /* "" "" some of which was retransmitted. */ +-#define FLAG_SYN_ACKED 0x10 /* This ACK acknowledged SYN. */ +-#define FLAG_DATA_SACKED 0x20 /* New SACK. */ +-#define FLAG_ECE 0x40 /* ECE in this ACK */ +-#define FLAG_SLOWPATH 0x100 /* Do not skip RFC checks for window update.*/ +-#define FLAG_ORIG_SACK_ACKED 0x200 /* Never retransmitted data are (s)acked */ +-#define FLAG_SND_UNA_ADVANCED 0x400 /* Snd_una was changed (!= FLAG_DATA_ACKED) */ +-#define FLAG_DSACKING_ACK 0x800 /* SACK blocks contained D-SACK info */ +-#define FLAG_SACK_RENEGING 0x2000 /* snd_una advanced to a sacked seq */ +-#define FLAG_UPDATE_TS_RECENT 0x4000 /* tcp_replace_ts_recent() */ +- +-#define FLAG_ACKED (FLAG_DATA_ACKED|FLAG_SYN_ACKED) +-#define FLAG_NOT_DUP (FLAG_DATA|FLAG_WIN_UPDATE|FLAG_ACKED) +-#define FLAG_CA_ALERT (FLAG_DATA_SACKED|FLAG_ECE) +-#define FLAG_FORWARD_PROGRESS (FLAG_ACKED|FLAG_DATA_SACKED) +- + #define TCP_REMNANT (TCP_FLAG_FIN|TCP_FLAG_URG|TCP_FLAG_SYN|TCP_FLAG_PSH) + #define TCP_HP_BITS (~(TCP_RESERVED_BITS|TCP_FLAG_PSH)) + +@@ -283,8 +267,12 @@ static void tcp_sndbuf_expand(struct sock *sk) + per_mss = roundup_pow_of_two(per_mss) + + SKB_DATA_ALIGN(sizeof(struct sk_buff)); + +- nr_segs = max_t(u32, TCP_INIT_CWND, tp->snd_cwnd); +- nr_segs = max_t(u32, nr_segs, tp->reordering + 1); ++ if (tp->mpc) { ++ nr_segs = mptcp_check_snd_buf(tp); ++ } else { ++ nr_segs = max_t(u32, TCP_INIT_CWND, tp->snd_cwnd); ++ nr_segs = max_t(u32, nr_segs, tp->reordering + 1); ++ } + + /* Fast Recovery (RFC 5681 3.2) : + * Cubic needs 1.7 factor, rounded to 2 to include +@@ -292,8 +280,16 @@ static void tcp_sndbuf_expand(struct sock *sk) + */ + sndmem = 2 * nr_segs * per_mss; + +- if (sk->sk_sndbuf < sndmem) ++ /* MPTCP: after this sndmem is the new contribution of the ++ * current subflow to the aggregated sndbuf */ ++ if (sk->sk_sndbuf < sndmem) { ++ int old_sndbuf = sk->sk_sndbuf; + sk->sk_sndbuf = min(sndmem, sysctl_tcp_wmem[2]); ++ /* MPTCP: ok, the subflow sndbuf has grown, reflect ++ * this in the aggregate buffer.*/ ++ if (tp->mpc && old_sndbuf != sk->sk_sndbuf) ++ mptcp_update_sndbuf(tp->mpcb); ++ } + } + + /* 2. Tuning advertised window (window_clamp, rcv_ssthresh) +@@ -342,10 +338,12 @@ static int __tcp_grow_window(const struct sock *sk, const struct sk_buff *skb) + static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb) + { + struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); + + /* Check #1 */ +- if (tp->rcv_ssthresh < tp->window_clamp && +- (int)tp->rcv_ssthresh < tcp_space(sk) && ++ if (meta_tp->rcv_ssthresh < meta_tp->window_clamp && ++ (int)meta_tp->rcv_ssthresh < tcp_space(sk) && + !sk_under_memory_pressure(sk)) { + int incr; + +@@ -353,14 +351,14 @@ static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb) + * will fit to rcvbuf in future. + */ + if (tcp_win_from_space(skb->truesize) <= skb->len) +- incr = 2 * tp->advmss; ++ incr = 2 * meta_tp->advmss; + else +- incr = __tcp_grow_window(sk, skb); ++ incr = __tcp_grow_window(meta_sk, skb); + + if (incr) { + incr = max_t(int, incr, 2 * skb->len); +- tp->rcv_ssthresh = min(tp->rcv_ssthresh + incr, +- tp->window_clamp); ++ meta_tp->rcv_ssthresh = min(meta_tp->rcv_ssthresh + incr, ++ meta_tp->window_clamp); + inet_csk(sk)->icsk_ack.quick |= 1; + } + } +@@ -543,7 +541,10 @@ void tcp_rcv_space_adjust(struct sock *sk) + int copied; + + time = tcp_time_stamp - tp->rcvq_space.time; +- if (time < (tp->rcv_rtt_est.rtt >> 3) || tp->rcv_rtt_est.rtt == 0) ++ if (tp->mpc) { ++ if (mptcp_check_rtt(tp, time)) ++ return; ++ } else if (time < (tp->rcv_rtt_est.rtt >> 3) || tp->rcv_rtt_est.rtt == 0) + return; + + /* Number of bytes copied to user in last RTT */ +@@ -768,7 +769,7 @@ static void tcp_update_pacing_rate(struct sock *sk) + /* Calculate rto without backoff. This is the second half of Van Jacobson's + * routine referred to above. + */ +-static void tcp_set_rto(struct sock *sk) ++void tcp_set_rto(struct sock *sk) + { + const struct tcp_sock *tp = tcp_sk(sk); + /* Old crap is replaced with new one. 8) +@@ -2914,7 +2915,7 @@ static inline bool tcp_ack_update_rtt(struct sock *sk, const int flag, + return false; + + tcp_rtt_estimator(sk, seq_rtt); +- tcp_set_rto(sk); ++ tp->set_rto(sk); + + /* RFC6298: only reset backoff on valid RTT measurement. */ + inet_csk(sk)->icsk_backoff = 0; +@@ -2998,7 +2999,7 @@ void tcp_resume_early_retransmit(struct sock *sk) + } + + /* If we get here, the whole TSO packet has not been acked. */ +-static u32 tcp_tso_acked(struct sock *sk, struct sk_buff *skb) ++u32 tcp_tso_acked(struct sock *sk, struct sk_buff *skb) + { + struct tcp_sock *tp = tcp_sk(sk); + u32 packets_acked; +@@ -3092,6 +3093,8 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, + */ + if (!(scb->tcp_flags & TCPHDR_SYN)) { + flag |= FLAG_DATA_ACKED; ++ if (tp->mpc && mptcp_is_data_seq(skb)) ++ flag |= MPTCP_FLAG_DATA_ACKED; + } else { + flag |= FLAG_SYN_ACKED; + tp->retrans_stamp = 0; +@@ -3194,7 +3197,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, + return flag; + } + +-static void tcp_ack_probe(struct sock *sk) ++void tcp_ack_probe(struct sock *sk) + { + const struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); +@@ -3241,9 +3244,8 @@ static inline bool tcp_may_raise_cwnd(const struct sock *sk, const int flag) + /* Check that window update is acceptable. + * The function assumes that snd_una<=ack<=snd_next. + */ +-static inline bool tcp_may_update_window(const struct tcp_sock *tp, +- const u32 ack, const u32 ack_seq, +- const u32 nwin) ++bool tcp_may_update_window(const struct tcp_sock *tp, const u32 ack, ++ const u32 ack_seq, const u32 nwin) + { + return after(ack, tp->snd_una) || + after(ack_seq, tp->snd_wl1) || +@@ -3362,7 +3364,7 @@ static void tcp_process_tlp_ack(struct sock *sk, u32 ack, int flag) + } + + /* This routine deals with incoming acks, but not outgoing ones. */ +-static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) ++static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) + { + struct inet_connection_sock *icsk = inet_csk(sk); + struct tcp_sock *tp = tcp_sk(sk); +@@ -3455,6 +3457,16 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) + flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, sack_rtt); + acked -= tp->packets_out; + ++ if (tp->mpc) { ++ if (mptcp_fallback_infinite(sk, flag)) { ++ pr_err("%s resetting flow\n", __func__); ++ mptcp_send_reset(sk); ++ goto invalid_ack; ++ } ++ ++ mptcp_clean_rtx_infinite(skb, sk); ++ } ++ + /* Advance cwnd if state allows */ + if (tcp_may_raise_cwnd(sk, flag)) + tcp_cong_avoid(sk, ack, acked, prior_in_flight); +@@ -3519,8 +3531,9 @@ old_ack: + * the fast version below fails. + */ + void tcp_parse_options(const struct sk_buff *skb, +- struct tcp_options_received *opt_rx, int estab, +- struct tcp_fastopen_cookie *foc) ++ struct tcp_options_received *opt_rx, ++ struct mptcp_options_received *mopt, ++ int estab, struct tcp_fastopen_cookie *foc) + { + const unsigned char *ptr; + const struct tcphdr *th = tcp_hdr(skb); +@@ -3603,6 +3616,10 @@ void tcp_parse_options(const struct sk_buff *skb, + */ + break; + #endif ++ case TCPOPT_MPTCP: ++ mptcp_parse_options(ptr - 2, opsize, opt_rx, ++ mopt, skb); ++ break; + case TCPOPT_EXP: + /* Fast Open option shares code 254 using a + * 16 bits magic number. It's valid only in +@@ -3664,8 +3681,8 @@ static bool tcp_fast_parse_options(const struct sk_buff *skb, + if (tcp_parse_aligned_timestamp(tp, th)) + return true; + } +- +- tcp_parse_options(skb, &tp->rx_opt, 1, NULL); ++ tcp_parse_options(skb, &tp->rx_opt, tp->mpc ? &tp->mptcp->rx_opt : NULL, ++ 1, NULL); + if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr) + tp->rx_opt.rcv_tsecr -= tp->tsoffset; + +@@ -3838,6 +3855,8 @@ static void tcp_fin(struct sock *sk) + dst = __sk_dst_get(sk); + if (!dst || !dst_metric(dst, RTAX_QUICKACK)) + inet_csk(sk)->icsk_ack.pingpong = 1; ++ if (tp->mpc) ++ mptcp_sub_close_passive(sk); + break; + + case TCP_CLOSE_WAIT: +@@ -3859,6 +3878,13 @@ static void tcp_fin(struct sock *sk) + tcp_set_state(sk, TCP_CLOSING); + break; + case TCP_FIN_WAIT2: ++ if (tp->mpc) { ++ /* The socket will get closed by mptcp_data_ready. ++ * We first have to process all data-sequences. ++ */ ++ tp->close_it = 1; ++ break; ++ } + /* Received a FIN -- send ACK and enter TIME_WAIT. */ + tcp_send_ack(sk); + tcp_time_wait(sk, TCP_TIME_WAIT, 0); +@@ -3883,6 +3909,10 @@ static void tcp_fin(struct sock *sk) + if (!sock_flag(sk, SOCK_DEAD)) { + sk->sk_state_change(sk); + ++ /* Don't wake up MPTCP-subflows */ ++ if (tp->mpc) ++ return; ++ + /* Do not send POLL_HUP for half duplex close. */ + if (sk->sk_shutdown == SHUTDOWN_MASK || + sk->sk_state == TCP_CLOSE) +@@ -4080,7 +4110,11 @@ static void tcp_ofo_queue(struct sock *sk) + tcp_dsack_extend(sk, TCP_SKB_CB(skb)->seq, dsack); + } + +- if (!after(TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt)) { ++ /* In case of MPTCP, the segment may be empty if it's a ++ * non-data DATA_FIN. (see beginning of tcp_data_queue) ++ */ ++ if (!after(TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt) && ++ !(tp->mpc && TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq)) { + SOCK_DEBUG(sk, "ofo packet was already received\n"); + __skb_unlink(skb, &tp->out_of_order_queue); + __kfree_skb(skb); +@@ -4104,6 +4138,9 @@ static int tcp_prune_queue(struct sock *sk); + static int tcp_try_rmem_schedule(struct sock *sk, struct sk_buff *skb, + unsigned int size) + { ++ if (tcp_sk(sk)->mpc) ++ sk = mptcp_meta_sk(sk); ++ + if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf || + !sk_rmem_schedule(sk, skb, size)) { + +@@ -4134,15 +4171,16 @@ static int tcp_try_rmem_schedule(struct sock *sk, struct sk_buff *skb, + * Better try to coalesce them right now to avoid future collapses. + * Returns true if caller should free @from instead of queueing it + */ +-static bool tcp_try_coalesce(struct sock *sk, +- struct sk_buff *to, +- struct sk_buff *from, +- bool *fragstolen) ++bool tcp_try_coalesce(struct sock *sk, struct sk_buff *to, struct sk_buff *from, ++ bool *fragstolen) + { + int delta; + + *fragstolen = false; + ++ if (tcp_sk(sk)->mpc && !is_meta_sk(sk)) ++ return false; ++ + if (tcp_hdr(from)->fin) + return false; + +@@ -4232,7 +4270,9 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb) + + /* Do skb overlap to previous one? */ + if (skb1 && before(seq, TCP_SKB_CB(skb1)->end_seq)) { +- if (!after(end_seq, TCP_SKB_CB(skb1)->end_seq)) { ++ /* MPTCP allows non-data data-fin to be in the ofo-queue */ ++ if (!after(end_seq, TCP_SKB_CB(skb1)->end_seq) && ++ !(tp->mpc && end_seq == seq)) { + /* All the bits are present. Drop. */ + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPOFOMERGE); + __kfree_skb(skb); +@@ -4270,6 +4310,9 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb) + end_seq); + break; + } ++ /* MPTCP allows non-data data-fin to be in the ofo-queue */ ++ if (tp->mpc && TCP_SKB_CB(skb1)->seq == TCP_SKB_CB(skb1)->end_seq) ++ continue; + __skb_unlink(skb1, &tp->out_of_order_queue); + tcp_dsack_extend(sk, TCP_SKB_CB(skb1)->seq, + TCP_SKB_CB(skb1)->end_seq); +@@ -4287,8 +4330,8 @@ end: + } + } + +-static int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb, int hdrlen, +- bool *fragstolen) ++int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb, int hdrlen, ++ bool *fragstolen) + { + int eaten; + struct sk_buff *tail = skb_peek_tail(&sk->sk_receive_queue); +@@ -4350,7 +4393,10 @@ static void tcp_data_queue(struct sock *sk, struct sk_buff *skb) + int eaten = -1; + bool fragstolen = false; + +- if (TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq) ++ /* If no data is present, but a data_fin is in the options, we still ++ * have to call mptcp_queue_skb later on. */ ++ if (TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq && ++ !(tp->mpc && mptcp_is_data_fin(skb))) + goto drop; + + skb_dst_drop(skb); +@@ -4396,7 +4442,7 @@ queue_and_out: + eaten = tcp_queue_rcv(sk, skb, 0, &fragstolen); + } + tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; +- if (skb->len) ++ if (skb->len || mptcp_is_data_fin(skb)) + tcp_event_data_recv(sk, skb); + if (th->fin) + tcp_fin(sk); +@@ -4418,7 +4464,11 @@ queue_and_out: + + if (eaten > 0) + kfree_skb_partial(skb, fragstolen); +- if (!sock_flag(sk, SOCK_DEAD)) ++ if (!sock_flag(sk, SOCK_DEAD) || tp->mpc) ++ /* MPTCP: we always have to call data_ready, because ++ * we may be about to receive a data-fin, which still ++ * must get queued. ++ */ + sk->sk_data_ready(sk, 0); + return; + } +@@ -4470,6 +4520,8 @@ static struct sk_buff *tcp_collapse_one(struct sock *sk, struct sk_buff *skb, + next = skb_queue_next(list, skb); + + __skb_unlink(skb, list); ++ if (tcp_sk(sk)->mpc) ++ mptcp_remove_shortcuts(tcp_sk(sk)->mpcb, skb); + __kfree_skb(skb); + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPRCVCOLLAPSED); + +@@ -4642,6 +4694,18 @@ static bool tcp_prune_ofo_queue(struct sock *sk) + struct tcp_sock *tp = tcp_sk(sk); + bool res = false; + ++ if (is_meta_sk(sk)) { ++ if (!skb_queue_empty(&tp->out_of_order_queue)) { ++ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_OFOPRUNED); ++ mptcp_purge_ofo_queue(tp); ++ ++ /* No sack at the mptcp-level */ ++ sk_mem_reclaim(sk); ++ res = true; ++ } ++ return res; ++ } ++ + if (!skb_queue_empty(&tp->out_of_order_queue)) { + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_OFOPRUNED); + __skb_queue_purge(&tp->out_of_order_queue); +@@ -4731,7 +4795,7 @@ void tcp_cwnd_application_limited(struct sock *sk) + tp->snd_cwnd_stamp = tcp_time_stamp; + } + +-static bool tcp_should_expand_sndbuf(const struct sock *sk) ++bool tcp_should_expand_sndbuf(const struct sock *sk) + { + const struct tcp_sock *tp = tcp_sk(sk); + +@@ -4766,7 +4830,7 @@ static void tcp_new_space(struct sock *sk) + { + struct tcp_sock *tp = tcp_sk(sk); + +- if (tcp_should_expand_sndbuf(sk)) { ++ if (tp->should_expand_sndbuf(sk)) { + tcp_sndbuf_expand(sk); + tp->snd_cwnd_stamp = tcp_time_stamp; + } +@@ -4778,8 +4842,9 @@ static void tcp_check_space(struct sock *sk) + { + if (sock_flag(sk, SOCK_QUEUE_SHRUNK)) { + sock_reset_flag(sk, SOCK_QUEUE_SHRUNK); +- if (sk->sk_socket && +- test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) ++ if (tcp_sk(sk)->mpc || ++ (sk->sk_socket && ++ test_bit(SOCK_NOSPACE, &sk->sk_socket->flags))) + tcp_new_space(sk); + } + } +@@ -4802,7 +4867,7 @@ static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible) + /* ... and right edge of window advances far enough. + * (tcp_recvmsg() will send ACK otherwise). Or... + */ +- __tcp_select_window(sk) >= tp->rcv_wnd) || ++ tp->__select_window(sk) >= tp->rcv_wnd) || + /* We ACK each frame or... */ + tcp_in_quickack_mode(sk) || + /* We have out of order data. */ +@@ -4904,6 +4969,10 @@ static void tcp_urg(struct sock *sk, struct sk_buff *skb, const struct tcphdr *t + { + struct tcp_sock *tp = tcp_sk(sk); + ++ /* MPTCP urgent data is not yet supported */ ++ if (tp->mpc) ++ return; ++ + /* Check if we get a new urgent pointer - normally not. */ + if (th->urg) + tcp_check_urg(sk, th); +@@ -4971,8 +5040,7 @@ static inline bool tcp_checksum_complete_user(struct sock *sk, + } + + #ifdef CONFIG_NET_DMA +-static bool tcp_dma_try_early_copy(struct sock *sk, struct sk_buff *skb, +- int hlen) ++bool tcp_dma_try_early_copy(struct sock *sk, struct sk_buff *skb, int hlen) + { + struct tcp_sock *tp = tcp_sk(sk); + int chunk = skb->len - hlen; +@@ -5081,9 +5149,15 @@ syn_challenge: + goto discard; + } + ++ /* If valid: post process the received MPTCP options. */ ++ if (tp->mpc && mptcp_handle_options(sk, th, skb)) ++ goto discard; ++ + return true; + + discard: ++ if (tp->mpc) ++ mptcp_reset_mopt(tp); + __kfree_skb(skb); + return false; + } +@@ -5135,6 +5209,10 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb, + + tp->rx_opt.saw_tstamp = 0; + ++ /* MPTCP: force slowpath. */ ++ if (tp->mpc) ++ goto slow_path; ++ + /* pred_flags is 0xS?10 << 16 + snd_wnd + * if header_prediction is to be made + * 'S' will always be tp->tcp_header_len >> 2 +@@ -5349,7 +5427,7 @@ void tcp_finish_connect(struct sock *sk, struct sk_buff *skb) + */ + tp->lsndtime = tcp_time_stamp; + +- tcp_init_buffer_space(sk); ++ tp->init_buffer_space(sk); + + if (sock_flag(sk, SOCK_KEEPOPEN)) + inet_csk_reset_keepalive_timer(sk, keepalive_time_when(tp)); +@@ -5379,7 +5457,7 @@ static bool tcp_rcv_fastopen_synack(struct sock *sk, struct sk_buff *synack, + /* Get original SYNACK MSS value if user MSS sets mss_clamp */ + tcp_clear_options(&opt); + opt.user_mss = opt.mss_clamp = 0; +- tcp_parse_options(synack, &opt, 0, NULL); ++ tcp_parse_options(synack, &opt, NULL, 0, NULL); + mss = opt.mss_clamp; + } + +@@ -5414,8 +5492,11 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, + struct tcp_sock *tp = tcp_sk(sk); + struct tcp_fastopen_cookie foc = { .len = -1 }; + int saved_clamp = tp->rx_opt.mss_clamp; ++ struct mptcp_options_received mopt; ++ mptcp_init_mp_opt(&mopt); + +- tcp_parse_options(skb, &tp->rx_opt, 0, &foc); ++ tcp_parse_options(skb, &tp->rx_opt, ++ tp->mpc ? &tp->mptcp->rx_opt : &mopt, 0, &foc); + if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr) + tp->rx_opt.rcv_tsecr -= tp->tsoffset; + +@@ -5462,6 +5543,21 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, + if (!th->syn) + goto discard_and_undo; + ++ if (tp->request_mptcp || tp->mpc) { ++ int ret; ++ ret = mptcp_rcv_synsent_state_process(sk, &sk, ++ skb, &mopt); ++ ++ /* May have changed if we support MPTCP */ ++ tp = tcp_sk(sk); ++ icsk = inet_csk(sk); ++ ++ if (ret == 1) ++ goto reset_and_undo; ++ if (ret == 2) ++ goto discard; ++ } ++ + /* rfc793: + * "If the SYN bit is on ... + * are acceptable then ... +@@ -5474,6 +5570,15 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, + tcp_init_wl(tp, TCP_SKB_CB(skb)->seq); + tcp_ack(sk, skb, FLAG_SLOWPATH); + ++ if (tp->mpc && !is_master_tp(tp)) { ++ /* Timer for repeating the ACK until an answer ++ * arrives. Used only when establishing an additional ++ * subflow inside of an MPTCP connection. ++ */ ++ sk_reset_timer(sk, &tp->mptcp->mptcp_ack_timer, ++ jiffies + icsk->icsk_rto); ++ } ++ + /* Ok.. it's good. Set up sequence numbers and + * move to established. + */ +@@ -5500,6 +5605,11 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, + tp->tcp_header_len = sizeof(struct tcphdr); + } + ++ if (tp->mpc) { ++ tp->tcp_header_len += MPTCP_SUB_LEN_DSM_ALIGN; ++ tp->advmss -= MPTCP_SUB_LEN_DSM_ALIGN; ++ } ++ + if (tcp_is_sack(tp) && sysctl_tcp_fack) + tcp_enable_fack(tp); + +@@ -5520,7 +5630,9 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb, + tcp_rcv_fastopen_synack(sk, skb, &foc)) + return -1; + +- if (sk->sk_write_pending || ++ /* With MPTCP we cannot send data on the third ack due to the ++ * lack of option-space */ ++ if ((sk->sk_write_pending && !tp->mpc) || + icsk->icsk_accept_queue.rskq_defer_accept || + icsk->icsk_ack.pingpong) { + /* Save one ACK. Data will be ready after +@@ -5562,6 +5674,7 @@ discard: + tcp_paws_reject(&tp->rx_opt, 0)) + goto discard_and_undo; + ++ /* TODO - check this here for MPTCP */ + if (th->syn) { + /* We see SYN without ACK. It is attempt of + * simultaneous connect with crossed SYNs. +@@ -5578,6 +5691,11 @@ discard: + tp->tcp_header_len = sizeof(struct tcphdr); + } + ++ if (tp->mpc) { ++ tp->tcp_header_len += MPTCP_SUB_LEN_DSM_ALIGN; ++ tp->advmss -= MPTCP_SUB_LEN_DSM_ALIGN; ++ } ++ + tp->rcv_nxt = TCP_SKB_CB(skb)->seq + 1; + tp->rcv_wup = TCP_SKB_CB(skb)->seq + 1; + +@@ -5636,6 +5754,7 @@ reset_and_undo: + + int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + const struct tcphdr *th, unsigned int len) ++ __releases(&sk->sk_lock.slock) + { + struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); +@@ -5687,6 +5806,10 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + + case TCP_SYN_SENT: + queued = tcp_rcv_synsent_state_process(sk, skb, th, len); ++ if (is_meta_sk(sk)) { ++ sk = tcp_sk(sk)->mpcb->master_sk; ++ tp = tcp_sk(sk); ++ } + if (queued >= 0) + return queued; + +@@ -5694,6 +5817,8 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + tcp_urg(sk, skb, th); + __kfree_skb(skb); + tcp_data_snd_check(sk); ++ if (tp->mpc && is_master_tp(tp)) ++ bh_unlock_sock(sk); + return 0; + } + +@@ -5736,7 +5861,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + + tcp_mtup_init(sk); + tp->copied_seq = tp->rcv_nxt; +- tcp_init_buffer_space(sk); ++ tp->init_buffer_space(sk); + } + smp_mb(); + tcp_set_state(sk, TCP_ESTABLISHED); +@@ -5756,6 +5881,8 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + + if (tp->rx_opt.tstamp_ok) + tp->advmss -= TCPOLEN_TSTAMP_ALIGNED; ++ if (tp->mpc) ++ tp->advmss -= MPTCP_SUB_LEN_DSM_ALIGN; + + if (req) { + /* Re-arm the timer because data may have been sent out. +@@ -5777,6 +5904,12 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + + tcp_initialize_rcv_mss(sk); + tcp_fast_path_on(tp); ++ /* Send an ACK when establishing a new ++ * MPTCP subflow, i.e. using an MP_JOIN ++ * subtype. ++ */ ++ if (tp->mpc && !is_master_tp(tp)) ++ tcp_send_ack(sk); + break; + + case TCP_FIN_WAIT1: { +@@ -5828,7 +5961,8 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + tmo = tcp_fin_time(sk); + if (tmo > TCP_TIMEWAIT_LEN) { + inet_csk_reset_keepalive_timer(sk, tmo - TCP_TIMEWAIT_LEN); +- } else if (th->fin || sock_owned_by_user(sk)) { ++ } else if (th->fin || mptcp_is_data_fin(skb) || ++ sock_owned_by_user(sk)) { + /* Bad case. We could lose such FIN otherwise. + * It is not a big problem, but it looks confusing + * and not so rare event. We still can lose it now, +@@ -5857,6 +5991,9 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + goto discard; + } + break; ++ case TCP_CLOSE: ++ if (tp->mp_killed) ++ goto discard; + } + + /* step 6: check the URG bit */ +@@ -5877,7 +6014,11 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, + */ + if (sk->sk_shutdown & RCV_SHUTDOWN) { + if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq && +- after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt)) { ++ after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt) && ++ !tp->mpc) { ++ /* In case of mptcp, the reset is handled by ++ * mptcp_rcv_state_process ++ */ + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPABORTONDATA); + tcp_reset(sk); + return 1; +diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c +index 1e4eac7..5891fa6 100644 +--- a/net/ipv4/tcp_ipv4.c ++++ b/net/ipv4/tcp_ipv4.c +@@ -67,6 +67,8 @@ + #include <net/icmp.h> + #include <net/inet_hashtables.h> + #include <net/tcp.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> + #include <net/transp_v6.h> + #include <net/ipv6.h> + #include <net/inet_common.h> +@@ -99,7 +101,7 @@ static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, + struct inet_hashinfo tcp_hashinfo; + EXPORT_SYMBOL(tcp_hashinfo); + +-static inline __u32 tcp_v4_init_sequence(const struct sk_buff *skb) ++__u32 tcp_v4_init_sequence(const struct sk_buff *skb) + { + return secure_tcp_sequence_number(ip_hdr(skb)->daddr, + ip_hdr(skb)->saddr, +@@ -334,7 +336,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + struct inet_sock *inet; + const int type = icmp_hdr(icmp_skb)->type; + const int code = icmp_hdr(icmp_skb)->code; +- struct sock *sk; ++ struct sock *sk, *meta_sk; + struct sk_buff *skb; + struct request_sock *req; + __u32 seq; +@@ -358,13 +360,19 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + return; + } + +- bh_lock_sock(sk); ++ tp = tcp_sk(sk); ++ if (tp->mpc) ++ meta_sk = mptcp_meta_sk(sk); ++ else ++ meta_sk = sk; ++ ++ bh_lock_sock(meta_sk); + /* If too many ICMPs get dropped on busy + * servers this needs to be solved differently. + * We do take care of PMTU discovery (RFC1191) special case : + * we can receive locally generated ICMP messages while socket is held. + */ +- if (sock_owned_by_user(sk)) { ++ if (sock_owned_by_user(meta_sk)) { + if (!(type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED)) + NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS); + } +@@ -377,7 +385,6 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + } + + icsk = inet_csk(sk); +- tp = tcp_sk(sk); + req = tp->fastopen_rsk; + seq = ntohl(th->seq); + if (sk->sk_state != TCP_LISTEN && +@@ -411,11 +418,13 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + goto out; + + tp->mtu_info = info; +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + tcp_v4_mtu_reduced(sk); + } else { + if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, &tp->tsq_flags)) + sock_hold(sk); ++ if (tp->mpc) ++ mptcp_tsq_flags(sk); + } + goto out; + } +@@ -431,7 +440,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + + /* XXX (TFO) - revisit the following logic for TFO */ + +- if (sock_owned_by_user(sk)) ++ if (sock_owned_by_user(meta_sk)) + break; + + icsk->icsk_backoff--; +@@ -473,7 +482,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + switch (sk->sk_state) { + struct request_sock *req, **prev; + case TCP_LISTEN: +- if (sock_owned_by_user(sk)) ++ if (sock_owned_by_user(meta_sk)) + goto out; + + req = inet_csk_search_req(sk, &prev, th->dest, +@@ -506,7 +515,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + It can f.e. if SYNs crossed, + or Fast Open. + */ +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + sk->sk_err = err; + + sk->sk_error_report(sk); +@@ -535,7 +544,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + */ + + inet = inet_sk(sk); +- if (!sock_owned_by_user(sk) && inet->recverr) { ++ if (!sock_owned_by_user(meta_sk) && inet->recverr) { + sk->sk_err = err; + sk->sk_error_report(sk); + } else { /* Only an error on timeout */ +@@ -543,7 +552,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) + } + + out: +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + sock_put(sk); + } + +@@ -585,7 +594,7 @@ EXPORT_SYMBOL(tcp_v4_send_check); + * Exception: precedence violation. We do not implement it in any case. + */ + +-static void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb) ++void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb) + { + const struct tcphdr *th = tcp_hdr(skb); + struct { +@@ -709,10 +718,10 @@ release_sk1: + outside socket context is ugly, certainly. What can I do? + */ + +-static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack, ++static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 data_ack, + u32 win, u32 tsval, u32 tsecr, int oif, + struct tcp_md5sig_key *key, +- int reply_flags, u8 tos) ++ int reply_flags, u8 tos, int mptcp) + { + const struct tcphdr *th = tcp_hdr(skb); + struct { +@@ -721,6 +730,10 @@ static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack, + #ifdef CONFIG_TCP_MD5SIG + + (TCPOLEN_MD5SIG_ALIGNED >> 2) + #endif ++#ifdef CONFIG_MPTCP ++ + ((MPTCP_SUB_LEN_DSS >> 2) + ++ (MPTCP_SUB_LEN_ACK >> 2)) ++#endif + ]; + } rep; + struct ip_reply_arg arg; +@@ -765,6 +778,21 @@ static void tcp_v4_send_ack(struct sk_buff *skb, u32 seq, u32 ack, + ip_hdr(skb)->daddr, &rep.th); + } + #endif ++#ifdef CONFIG_MPTCP ++ if (mptcp) { ++ int offset = (tsecr) ? 3 : 0; ++ /* Construction of 32-bit data_ack */ ++ rep.opt[offset++] = htonl((TCPOPT_MPTCP << 24) | ++ ((MPTCP_SUB_LEN_DSS + MPTCP_SUB_LEN_ACK) << 16) | ++ (0x20 << 8) | ++ (0x01)); ++ rep.opt[offset] = htonl(data_ack); ++ ++ arg.iov[0].iov_len += MPTCP_SUB_LEN_DSS + MPTCP_SUB_LEN_ACK; ++ rep.th.doff = arg.iov[0].iov_len / 4; ++ } ++#endif /* CONFIG_MPTCP */ ++ + arg.flags = reply_flags; + arg.csum = csum_tcpudp_nofold(ip_hdr(skb)->daddr, + ip_hdr(skb)->saddr, /* XXX */ +@@ -783,36 +811,44 @@ static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb) + { + struct inet_timewait_sock *tw = inet_twsk(sk); + struct tcp_timewait_sock *tcptw = tcp_twsk(sk); ++ u32 data_ack = 0; ++ int mptcp = 0; ++ ++ if (tcptw->mptcp_tw && tcptw->mptcp_tw->meta_tw) { ++ data_ack = (u32)tcptw->mptcp_tw->rcv_nxt; ++ mptcp = 1; ++ } + + tcp_v4_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt, ++ data_ack, + tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale, + tcp_time_stamp + tcptw->tw_ts_offset, + tcptw->tw_ts_recent, + tw->tw_bound_dev_if, + tcp_twsk_md5_key(tcptw), + tw->tw_transparent ? IP_REPLY_ARG_NOSRCCHECK : 0, +- tw->tw_tos ++ tw->tw_tos, mptcp + ); + + inet_twsk_put(tw); + } + +-static void tcp_v4_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, +- struct request_sock *req) ++void tcp_v4_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req) + { + /* sk->sk_state == TCP_LISTEN -> for regular TCP_SYN_RECV + * sk->sk_state == TCP_SYN_RECV -> for Fast Open. + */ + tcp_v4_send_ack(skb, (sk->sk_state == TCP_LISTEN) ? + tcp_rsk(req)->snt_isn + 1 : tcp_sk(sk)->snd_nxt, +- tcp_rsk(req)->rcv_nxt, req->rcv_wnd, ++ tcp_rsk(req)->rcv_nxt, 0, req->rcv_wnd, + tcp_time_stamp, + req->ts_recent, + 0, + tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr, + AF_INET), + inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0, +- ip_hdr(skb)->tos); ++ ip_hdr(skb)->tos, 0); + } + + /* +@@ -820,9 +856,9 @@ static void tcp_v4_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, + * This still operates on a request_sock only, not on a big + * socket. + */ +-static int tcp_v4_send_synack(struct sock *sk, struct dst_entry *dst, +- struct request_sock *req, +- u16 queue_mapping) ++int tcp_v4_send_synack(struct sock *sk, struct dst_entry *dst, ++ struct request_sock *req, ++ u16 queue_mapping) + { + const struct inet_request_sock *ireq = inet_rsk(req); + struct flowi4 fl4; +@@ -850,7 +886,7 @@ static int tcp_v4_send_synack(struct sock *sk, struct dst_entry *dst, + return err; + } + +-static int tcp_v4_rtx_synack(struct sock *sk, struct request_sock *req) ++int tcp_v4_rtx_synack(struct sock *sk, struct request_sock *req) + { + int res = tcp_v4_send_synack(sk, NULL, req, 0); + +@@ -862,7 +898,7 @@ static int tcp_v4_rtx_synack(struct sock *sk, struct request_sock *req) + /* + * IPv4 request_sock destructor. + */ +-static void tcp_v4_reqsk_destructor(struct request_sock *req) ++void tcp_v4_reqsk_destructor(struct request_sock *req) + { + kfree(inet_rsk(req)->opt); + } +@@ -902,7 +938,7 @@ EXPORT_SYMBOL(tcp_syn_flood_action); + /* + * Save and compile IPv4 options into the request_sock if needed. + */ +-static struct ip_options_rcu *tcp_v4_save_options(struct sk_buff *skb) ++struct ip_options_rcu *tcp_v4_save_options(struct sk_buff *skb) + { + const struct ip_options *opt = &(IPCB(skb)->opt); + struct ip_options_rcu *dopt = NULL; +@@ -1254,7 +1290,7 @@ struct request_sock_ops tcp_request_sock_ops __read_mostly = { + }; + + #ifdef CONFIG_TCP_MD5SIG +-static const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = { ++const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = { + .md5_lookup = tcp_v4_reqsk_md5_lookup, + .calc_md5_hash = tcp_v4_md5_hash_skb, + }; +@@ -1412,7 +1448,7 @@ static int tcp_v4_conn_req_fastopen(struct sock *sk, + tcp_init_congestion_control(child); + tcp_mtup_init(child); + tcp_init_metrics(child); +- tcp_init_buffer_space(child); ++ tp->init_buffer_space(child); + + /* Queue the data carried in the SYN packet. We need to first + * bump skb's refcnt because the caller will attempt to free it. +@@ -1444,6 +1480,7 @@ static int tcp_v4_conn_req_fastopen(struct sock *sk, + int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) + { + struct tcp_options_received tmp_opt; ++ struct mptcp_options_received mopt; + struct request_sock *req; + struct inet_request_sock *ireq; + struct tcp_sock *tp = tcp_sk(sk); +@@ -1458,6 +1495,22 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) + struct sk_buff *skb_synack; + int do_fastopen; + ++ tcp_clear_options(&tmp_opt); ++ tmp_opt.mss_clamp = TCP_MSS_DEFAULT; ++ tmp_opt.user_mss = tp->rx_opt.user_mss; ++ mptcp_init_mp_opt(&mopt); ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, want_cookie ? NULL : &foc); ++ ++#ifdef CONFIG_MPTCP ++ /* MPTCP structures not initialized, so clear MPTCP fields */ ++ if (mptcp_init_failed) ++ mptcp_init_mp_opt(&mopt); ++ ++ if (mopt.is_mp_join) ++ return mptcp_do_join_short(skb, &mopt, &tmp_opt, sock_net(sk)); ++ if (mopt.drop_me) ++ goto drop; ++#endif + /* Never answer to SYNs send to broadcast or multicast */ + if (skb_rtable(skb)->rt_flags & (RTCF_BROADCAST | RTCF_MULTICAST)) + goto drop; +@@ -1483,7 +1536,22 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) + goto drop; + } + +- req = inet_reqsk_alloc(&tcp_request_sock_ops); ++#ifdef CONFIG_MPTCP ++ if (sysctl_mptcp_enabled == MPTCP_APP && !tp->mptcp_enabled) ++ mopt.saw_mpc = 0; ++ if (mopt.saw_mpc && !want_cookie) { ++ req = inet_reqsk_alloc(&mptcp_request_sock_ops); ++ ++ if (!req) ++ goto drop; ++ ++ mptcp_rsk(req)->mpcb = NULL; ++ mptcp_rsk(req)->dss_csum = mopt.dss_csum; ++ mptcp_rsk(req)->collide_tk.pprev = NULL; ++ } else ++#endif ++ req = inet_reqsk_alloc(&tcp_request_sock_ops); ++ + if (!req) + goto drop; + +@@ -1491,17 +1559,15 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb) + tcp_rsk(req)->af_specific = &tcp_request_sock_ipv4_ops; + #endif + +- tcp_clear_options(&tmp_opt); +- tmp_opt.mss_clamp = TCP_MSS_DEFAULT; +- tmp_opt.user_mss = tp->rx_opt.user_mss; +- tcp_parse_options(skb, &tmp_opt, 0, want_cookie ? NULL : &foc); +- + if (want_cookie && !tmp_opt.saw_tstamp) + tcp_clear_options(&tmp_opt); + + tmp_opt.tstamp_ok = tmp_opt.saw_tstamp; + tcp_openreq_init(req, &tmp_opt, skb); + ++ if (mopt.saw_mpc && !want_cookie) ++ mptcp_reqsk_new_mptcp(req, &tmp_opt, &mopt, skb); ++ + ireq = inet_rsk(req); + ireq->ir_loc_addr = daddr; + ireq->ir_rmt_addr = saddr; +@@ -1713,7 +1779,7 @@ put_and_exit: + } + EXPORT_SYMBOL(tcp_v4_syn_recv_sock); + +-static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) ++struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) + { + struct tcphdr *th = tcp_hdr(skb); + const struct iphdr *iph = ip_hdr(skb); +@@ -1730,8 +1796,15 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) + + if (nsk) { + if (nsk->sk_state != TCP_TIME_WAIT) { ++ /* Don't lock again the meta-sk. It has been locked ++ * before mptcp_v4_do_rcv. ++ */ ++ if (tcp_sk(nsk)->mpc && !is_meta_sk(sk)) ++ bh_lock_sock(mptcp_meta_sk(nsk)); + bh_lock_sock(nsk); ++ + return nsk; ++ + } + inet_twsk_put(inet_twsk(nsk)); + return NULL; +@@ -1788,6 +1861,9 @@ int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb) + goto discard; + #endif + ++ if (is_meta_sk(sk)) ++ return mptcp_v4_do_rcv(sk, skb); ++ + if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */ + struct dst_entry *dst = sk->sk_rx_dst; + +@@ -1919,7 +1995,7 @@ bool tcp_prequeue(struct sock *sk, struct sk_buff *skb) + } else if (skb_queue_len(&tp->ucopy.prequeue) == 1) { + wake_up_interruptible_sync_poll(sk_sleep(sk), + POLLIN | POLLRDNORM | POLLRDBAND); +- if (!inet_csk_ack_scheduled(sk)) ++ if (!inet_csk_ack_scheduled(sk) && !tp->mpc) + inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, + (3 * tcp_rto_min(sk)) / 4, + TCP_RTO_MAX); +@@ -1936,7 +2012,7 @@ int tcp_v4_rcv(struct sk_buff *skb) + { + const struct iphdr *iph; + const struct tcphdr *th; +- struct sock *sk; ++ struct sock *sk, *meta_sk = NULL; + int ret; + struct net *net = dev_net(skb->dev); + +@@ -1969,18 +2045,42 @@ int tcp_v4_rcv(struct sk_buff *skb) + TCP_SKB_CB(skb)->end_seq = (TCP_SKB_CB(skb)->seq + th->syn + th->fin + + skb->len - th->doff * 4); + TCP_SKB_CB(skb)->ack_seq = ntohl(th->ack_seq); ++#ifdef CONFIG_MPTCP ++ TCP_SKB_CB(skb)->mptcp_flags = 0; ++ TCP_SKB_CB(skb)->dss_off = 0; ++#endif + TCP_SKB_CB(skb)->when = 0; + TCP_SKB_CB(skb)->ip_dsfield = ipv4_get_dsfield(iph); + TCP_SKB_CB(skb)->sacked = 0; + + sk = __inet_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest); +- if (!sk) +- goto no_tcp_socket; + + process: +- if (sk->sk_state == TCP_TIME_WAIT) ++ if (sk && sk->sk_state == TCP_TIME_WAIT) + goto do_time_wait; + ++#ifdef CONFIG_MPTCP ++ if (!sk && th->syn && !th->ack) { ++ int ret = mptcp_lookup_join(skb, NULL); ++ ++ if (ret < 0) { ++ tcp_v4_send_reset(NULL, skb); ++ goto discard_it; ++ } else if (ret > 0) { ++ return 0; ++ } ++ } ++ ++ /* Is there a pending request sock for this segment ? */ ++ if ((!sk || sk->sk_state == TCP_LISTEN) && mptcp_check_req(skb, net)) { ++ if (sk) ++ sock_put(sk); ++ return 0; ++ } ++#endif ++ if (!sk) ++ goto no_tcp_socket; ++ + if (unlikely(iph->ttl < inet_sk(sk)->min_ttl)) { + NET_INC_STATS_BH(net, LINUX_MIB_TCPMINTTLDROP); + goto discard_and_relse; +@@ -1996,11 +2096,21 @@ process: + sk_mark_napi_id(sk, skb); + skb->dev = NULL; + +- bh_lock_sock_nested(sk); ++ if (tcp_sk(sk)->mpc) { ++ meta_sk = mptcp_meta_sk(sk); ++ ++ bh_lock_sock_nested(meta_sk); ++ if (sock_owned_by_user(meta_sk)) ++ skb->sk = sk; ++ } else { ++ meta_sk = sk; ++ bh_lock_sock_nested(sk); ++ } ++ + ret = 0; +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + #ifdef CONFIG_NET_DMA +- struct tcp_sock *tp = tcp_sk(sk); ++ struct tcp_sock *tp = tcp_sk(meta_sk); + if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list) + tp->ucopy.dma_chan = net_dma_find_channel(); + if (tp->ucopy.dma_chan) +@@ -2008,16 +2118,16 @@ process: + else + #endif + { +- if (!tcp_prequeue(sk, skb)) ++ if (!tcp_prequeue(meta_sk, skb)) + ret = tcp_v4_do_rcv(sk, skb); + } +- } else if (unlikely(sk_add_backlog(sk, skb, +- sk->sk_rcvbuf + sk->sk_sndbuf))) { +- bh_unlock_sock(sk); ++ } else if (unlikely(sk_add_backlog(meta_sk, skb, ++ meta_sk->sk_rcvbuf + meta_sk->sk_sndbuf))) { ++ bh_unlock_sock(meta_sk); + NET_INC_STATS_BH(net, LINUX_MIB_TCPBACKLOGDROP); + goto discard_and_relse; + } +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + + sock_put(sk); + +@@ -2072,6 +2182,18 @@ do_time_wait: + sk = sk2; + goto process; + } ++#ifdef CONFIG_MPTCP ++ if (th->syn && !th->ack) { ++ int ret = mptcp_lookup_join(skb, inet_twsk(sk)); ++ ++ if (ret < 0) { ++ tcp_v4_send_reset(NULL, skb); ++ goto discard_it; ++ } else if (ret > 0) { ++ return 0; ++ } ++ } ++#endif + /* Fall through to ACK */ + } + case TCP_TW_ACK: +@@ -2154,6 +2276,11 @@ void tcp_v4_destroy_sock(struct sock *sk) + + tcp_cleanup_congestion_control(sk); + ++ if (tp->mpc) ++ mptcp_destroy_sock(sk); ++ if (tp->inside_tk_table) ++ mptcp_hash_remove(tp); ++ + /* Cleanup up the write buffer. */ + tcp_write_queue_purge(sk); + +diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c +index 7a436c5..72f9b8e 100644 +--- a/net/ipv4/tcp_minisocks.c ++++ b/net/ipv4/tcp_minisocks.c +@@ -18,11 +18,13 @@ + * Jorge Cwik, <jorge@laser.satlink.net> + */ + ++#include <linux/kconfig.h> + #include <linux/mm.h> + #include <linux/module.h> + #include <linux/slab.h> + #include <linux/sysctl.h> + #include <linux/workqueue.h> ++#include <net/mptcp.h> + #include <net/tcp.h> + #include <net/inet_common.h> + #include <net/xfrm.h> +@@ -95,10 +97,13 @@ tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb, + struct tcp_options_received tmp_opt; + struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw); + bool paws_reject = false; ++ struct mptcp_options_received mopt; + + tmp_opt.saw_tstamp = 0; + if (th->doff > (sizeof(*th) >> 2) && tcptw->tw_ts_recent_stamp) { +- tcp_parse_options(skb, &tmp_opt, 0, NULL); ++ mptcp_init_mp_opt(&mopt); ++ ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, NULL); + + if (tmp_opt.saw_tstamp) { + tmp_opt.rcv_tsecr -= tcptw->tw_ts_offset; +@@ -106,6 +111,11 @@ tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb, + tmp_opt.ts_recent_stamp = tcptw->tw_ts_recent_stamp; + paws_reject = tcp_paws_reject(&tmp_opt, th->rst); + } ++ ++ if (unlikely(mopt.mp_fclose) && tcptw->mptcp_tw) { ++ if (mopt.mptcp_key == tcptw->mptcp_tw->loc_key) ++ goto kill_with_rst; ++ } + } + + if (tw->tw_substate == TCP_FIN_WAIT2) { +@@ -128,6 +138,16 @@ tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb, + if (!th->ack || + !after(TCP_SKB_CB(skb)->end_seq, tcptw->tw_rcv_nxt) || + TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) { ++ /* If mptcp_is_data_fin() returns true, we are sure that ++ * mopt has been initialized - otherwise it would not ++ * be a DATA_FIN. ++ */ ++ if (tcptw->mptcp_tw && tcptw->mptcp_tw->meta_tw && ++ mptcp_is_data_fin(skb) && ++ TCP_SKB_CB(skb)->seq == tcptw->tw_rcv_nxt && ++ mopt.data_seq + 1 == (u32)tcptw->mptcp_tw->rcv_nxt) ++ return TCP_TW_ACK; ++ + inet_twsk_put(tw); + return TCP_TW_SUCCESS; + } +@@ -270,6 +290,11 @@ void tcp_time_wait(struct sock *sk, int state, int timeo) + const struct tcp_sock *tp = tcp_sk(sk); + bool recycle_ok = false; + ++ if (is_meta_sk(sk)) { ++ mptcp_update_tw_socks(tp, state); ++ goto tcp_done; ++ } ++ + if (tcp_death_row.sysctl_tw_recycle && tp->rx_opt.ts_recent_stamp) + recycle_ok = tcp_remember_stamp(sk); + +@@ -290,6 +315,15 @@ void tcp_time_wait(struct sock *sk, int state, int timeo) + tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp; + tcptw->tw_ts_offset = tp->tsoffset; + ++ if (tp->mpc) { ++ if (mptcp_time_wait(sk, tcptw)) { ++ inet_twsk_free(tw); ++ goto exit; ++ } ++ } else { ++ tcptw->mptcp_tw = NULL; ++ } ++ + #if IS_ENABLED(CONFIG_IPV6) + if (tw->tw_family == PF_INET6) { + struct ipv6_pinfo *np = inet6_sk(sk); +@@ -347,15 +381,19 @@ void tcp_time_wait(struct sock *sk, int state, int timeo) + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPTIMEWAITOVERFLOW); + } + ++exit: + tcp_update_metrics(sk); ++tcp_done: + tcp_done(sk); + } + + void tcp_twsk_destructor(struct sock *sk) + { +-#ifdef CONFIG_TCP_MD5SIG + struct tcp_timewait_sock *twsk = tcp_twsk(sk); + ++ if (twsk->mptcp_tw) ++ mptcp_twsk_destructor(twsk); ++#ifdef CONFIG_TCP_MD5SIG + if (twsk->tw_md5_key) + kfree_rcu(twsk->tw_md5_key, rcu); + #endif +@@ -392,6 +430,9 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, + + newtp->snd_sml = newtp->snd_una = + newtp->snd_nxt = newtp->snd_up = treq->snt_isn + 1; ++#ifdef CONFIG_MPTCP ++ memset(&newtp->rcvq_space, 0, sizeof(newtp->rcvq_space)); ++#endif + + tcp_prequeue_init(newtp); + INIT_LIST_HEAD(&newtp->tsq_node); +@@ -436,7 +477,11 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, + + newtp->urg_data = 0; + +- if (sock_flag(newsk, SOCK_KEEPOPEN)) ++ /* MPTCP: If we are creating a subflow, KEEPOPEN might have been ++ * set on the meta. But, keepalive is entirely handled at the ++ * meta-socket, so let's keep it there. ++ */ ++ if (sock_flag(newsk, SOCK_KEEPOPEN) && is_meta_sk(sk)) + inet_csk_reset_keepalive_timer(newsk, + keepalive_time_when(newtp)); + +@@ -468,6 +513,8 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, + newtp->rx_opt.ts_recent_stamp = 0; + newtp->tcp_header_len = sizeof(struct tcphdr); + } ++ if (treq->saw_mpc) ++ newtp->tcp_header_len += MPTCP_SUB_LEN_DSM_ALIGN; + newtp->tsoffset = 0; + #ifdef CONFIG_TCP_MD5SIG + newtp->md5sig_info = NULL; /*XXX*/ +@@ -504,16 +551,20 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, + bool fastopen) + { + struct tcp_options_received tmp_opt; ++ struct mptcp_options_received mopt; + struct sock *child; + const struct tcphdr *th = tcp_hdr(skb); + __be32 flg = tcp_flag_word(th) & (TCP_FLAG_RST|TCP_FLAG_SYN|TCP_FLAG_ACK); + bool paws_reject = false; + +- BUG_ON(fastopen == (sk->sk_state == TCP_LISTEN)); ++ BUG_ON(!tcp_sk(sk)->mpc && fastopen == (sk->sk_state == TCP_LISTEN)); + + tmp_opt.saw_tstamp = 0; ++ ++ mptcp_init_mp_opt(&mopt); ++ + if (th->doff > (sizeof(struct tcphdr)>>2)) { +- tcp_parse_options(skb, &tmp_opt, 0, NULL); ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, NULL); + + if (tmp_opt.saw_tstamp) { + tmp_opt.ts_recent = req->ts_recent; +@@ -552,7 +603,14 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, + * + * Reset timer after retransmitting SYNACK, similar to + * the idea of fast retransmit in recovery. ++ * ++ * Fall back to TCP if MP_CAPABLE is not set. + */ ++ ++ if (tcp_rsk(req)->saw_mpc && !mopt.saw_mpc) ++ tcp_rsk(req)->saw_mpc = false; ++ ++ + if (!inet_rtx_syn_ack(sk, req)) + req->expires = min(TCP_TIMEOUT_INIT << req->num_timeout, + TCP_RTO_MAX) + jiffies; +@@ -674,7 +732,20 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, + + /* While TCP_DEFER_ACCEPT is active, drop bare ACK. */ + if (req->num_timeout < inet_csk(sk)->icsk_accept_queue.rskq_defer_accept && +- TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) { ++ TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1 && ++ /* TODO MPTCP: ++ * We do this here, because otherwise options sent in the third ack, ++ * or duplicate fourth ack will get lost. Options like MP_PRIO, ADD_ADDR,... ++ * ++ * We could store them in request_sock, but this would mean that we ++ * have to put tcp_options_received and mptcp_options_received in there, ++ * increasing considerably the size of the request-sock. ++ * ++ * As soon as we have reworked the request-sock MPTCP-fields and ++ * created a mptcp_request_sock structure, we can handle options ++ * correclty there without increasing request_sock. ++ */ ++ !tcp_rsk(req)->saw_mpc) { + inet_rsk(req)->acked = 1; + NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPDEFERACCEPTDROP); + return NULL; +@@ -686,10 +757,29 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, + * ESTABLISHED STATE. If it will be dropped after + * socket is created, wait for troubles. + */ +- child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb, req, NULL); ++#ifdef CONFIG_MPTCP ++ if (tcp_sk(sk)->mpc) ++ /* MPTCP: We call the mptcp-specific syn_recv_sock */ ++ child = tcp_sk(sk)->mpcb->syn_recv_sock(sk, skb, req, NULL); ++ else ++#endif ++ child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb, ++ req, NULL); ++ + if (child == NULL) + goto listen_overflow; + ++ if (!is_meta_sk(sk)) { ++ int ret = mptcp_check_req_master(sk, child, req, prev, &mopt); ++ if (ret < 0) ++ goto listen_overflow; ++ ++ /* MPTCP-supported */ ++ if (!ret) ++ return tcp_sk(child)->mpcb->master_sk; ++ } else { ++ return mptcp_check_req_child(sk, child, req, prev, &mopt); ++ } + inet_csk_reqsk_queue_unlink(sk, req, prev); + inet_csk_reqsk_queue_removed(sk, req); + +@@ -739,8 +829,9 @@ int tcp_child_process(struct sock *parent, struct sock *child, + { + int ret = 0; + int state = child->sk_state; ++ struct sock *meta_sk = tcp_sk(child)->mpc ? mptcp_meta_sk(child) : child; + +- if (!sock_owned_by_user(child)) { ++ if (!sock_owned_by_user(meta_sk)) { + ret = tcp_rcv_state_process(child, skb, tcp_hdr(skb), + skb->len); + /* Wakeup parent, send SIGIO */ +@@ -751,10 +842,14 @@ int tcp_child_process(struct sock *parent, struct sock *child, + * in main socket hash table and lock on listening + * socket does not protect us more. + */ +- __sk_add_backlog(child, skb); ++ if (tcp_sk(child)->mpc) ++ skb->sk = child; ++ __sk_add_backlog(meta_sk, skb); + } + +- bh_unlock_sock(child); ++ if (tcp_sk(child)->mpc) ++ bh_unlock_sock(child); ++ bh_unlock_sock(meta_sk); + sock_put(child); + return ret; + } +diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c +index 17a11e6..6b45057 100644 +--- a/net/ipv4/tcp_output.c ++++ b/net/ipv4/tcp_output.c +@@ -36,6 +36,12 @@ + + #define pr_fmt(fmt) "TCP: " fmt + ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#if IS_ENABLED(CONFIG_IPV6) ++#include <net/mptcp_v6.h> ++#endif ++#include <net/ipv6.h> + #include <net/tcp.h> + + #include <linux/compiler.h> +@@ -72,7 +78,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, + int push_one, gfp_t gfp); + + /* Account for new data that has been sent to the network. */ +-static void tcp_event_new_data_sent(struct sock *sk, const struct sk_buff *skb) ++void tcp_event_new_data_sent(struct sock *sk, const struct sk_buff *skb) + { + struct inet_connection_sock *icsk = inet_csk(sk); + struct tcp_sock *tp = tcp_sk(sk); +@@ -211,7 +217,7 @@ u32 tcp_default_init_rwnd(u32 mss) + void tcp_select_initial_window(int __space, __u32 mss, + __u32 *rcv_wnd, __u32 *window_clamp, + int wscale_ok, __u8 *rcv_wscale, +- __u32 init_rcv_wnd) ++ __u32 init_rcv_wnd, const struct sock *sk) + { + unsigned int space = (__space < 0 ? 0 : __space); + +@@ -266,11 +272,15 @@ EXPORT_SYMBOL(tcp_select_initial_window); + * value can be stuffed directly into th->window for an outgoing + * frame. + */ +-static u16 tcp_select_window(struct sock *sk) ++u16 tcp_select_window(struct sock *sk) + { + struct tcp_sock *tp = tcp_sk(sk); +- u32 cur_win = tcp_receive_window(tp); +- u32 new_win = __tcp_select_window(sk); ++ /* The window must never shrink at the meta-level. At the subflow we ++ * have to allow this. Otherwise we may announce a window too large ++ * for the current meta-level sk_rcvbuf. ++ */ ++ u32 cur_win = tcp_receive_window(tp->mpc ? tcp_sk(mptcp_meta_sk(sk)) : tp); ++ u32 new_win = tp->__select_window(sk); + + /* Never shrink the offered window */ + if (new_win < cur_win) { +@@ -283,6 +293,7 @@ static u16 tcp_select_window(struct sock *sk) + */ + new_win = ALIGN(cur_win, 1 << tp->rx_opt.rcv_wscale); + } ++ + tp->rcv_wnd = new_win; + tp->rcv_wup = tp->rcv_nxt; + +@@ -361,7 +372,7 @@ static inline void TCP_ECN_send(struct sock *sk, struct sk_buff *skb, + /* Constructs common control bits of non-data skb. If SYN/FIN is present, + * auto increment end seqno. + */ +-static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags) ++void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags) + { + struct skb_shared_info *shinfo = skb_shinfo(skb); + +@@ -381,7 +392,7 @@ static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags) + TCP_SKB_CB(skb)->end_seq = seq; + } + +-static inline bool tcp_urg_mode(const struct tcp_sock *tp) ++bool tcp_urg_mode(const struct tcp_sock *tp) + { + return tp->snd_una != tp->snd_up; + } +@@ -391,17 +402,7 @@ static inline bool tcp_urg_mode(const struct tcp_sock *tp) + #define OPTION_MD5 (1 << 2) + #define OPTION_WSCALE (1 << 3) + #define OPTION_FAST_OPEN_COOKIE (1 << 8) +- +-struct tcp_out_options { +- u16 options; /* bit field of OPTION_* */ +- u16 mss; /* 0 to disable */ +- u8 ws; /* window scale, 0 to disable */ +- u8 num_sack_blocks; /* number of SACK blocks to include */ +- u8 hash_size; /* bytes in hash_location */ +- __u8 *hash_location; /* temporary pointer, overloaded */ +- __u32 tsval, tsecr; /* need to include OPTION_TS */ +- struct tcp_fastopen_cookie *fastopen_cookie; /* Fast open cookie */ +-}; ++/* Before adding here - take a look at OPTION_MPTCP in include/net/mptcp.h */ + + /* Write previously computed TCP options to the packet. + * +@@ -417,7 +418,7 @@ struct tcp_out_options { + * (but it may well be that other scenarios fail similarly). + */ + static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, +- struct tcp_out_options *opts) ++ struct tcp_out_options *opts, struct sk_buff *skb) + { + u16 options = opts->options; /* mungable copy */ + +@@ -500,6 +501,9 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, + } + ptr += (foc->len + 3) >> 2; + } ++ ++ if (unlikely(OPTION_MPTCP & opts->options)) ++ mptcp_options_write(ptr, tp, opts, skb); + } + + /* Compute TCP options for SYN packets. This is not the final +@@ -551,6 +555,8 @@ static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb, + if (unlikely(!(OPTION_TS & opts->options))) + remaining -= TCPOLEN_SACKPERM_ALIGNED; + } ++ if (tp->request_mptcp || tp->mpc) ++ mptcp_syn_options(sk, opts, &remaining); + + if (fastopen && fastopen->cookie.len >= 0) { + u32 need = TCPOLEN_EXP_FASTOPEN_BASE + fastopen->cookie.len; +@@ -624,6 +630,9 @@ static unsigned int tcp_synack_options(struct sock *sk, + } + } + ++ if (tcp_rsk(req)->saw_mpc) ++ mptcp_synack_options(req, opts, &remaining); ++ + return MAX_TCP_OPTION_SPACE - remaining; + } + +@@ -657,16 +666,22 @@ static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb + opts->tsecr = tp->rx_opt.ts_recent; + size += TCPOLEN_TSTAMP_ALIGNED; + } ++ if (tp->mpc) ++ mptcp_established_options(sk, skb, opts, &size); + + eff_sacks = tp->rx_opt.num_sacks + tp->rx_opt.dsack; + if (unlikely(eff_sacks)) { +- const unsigned int remaining = MAX_TCP_OPTION_SPACE - size; +- opts->num_sack_blocks = +- min_t(unsigned int, eff_sacks, +- (remaining - TCPOLEN_SACK_BASE_ALIGNED) / +- TCPOLEN_SACK_PERBLOCK); +- size += TCPOLEN_SACK_BASE_ALIGNED + +- opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK; ++ const unsigned remaining = MAX_TCP_OPTION_SPACE - size; ++ if (remaining < TCPOLEN_SACK_BASE_ALIGNED) ++ opts->num_sack_blocks = 0; ++ else ++ opts->num_sack_blocks = ++ min_t(unsigned int, eff_sacks, ++ (remaining - TCPOLEN_SACK_BASE_ALIGNED) / ++ TCPOLEN_SACK_PERBLOCK); ++ if (opts->num_sack_blocks) ++ size += TCPOLEN_SACK_BASE_ALIGNED + ++ opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK; + } + + return size; +@@ -714,7 +729,7 @@ static void tcp_tasklet_func(unsigned long data) + unsigned long flags; + struct list_head *q, *n; + struct tcp_sock *tp; +- struct sock *sk; ++ struct sock *sk, *meta_sk; + + local_irq_save(flags); + list_splice_init(&tsq->head, &list); +@@ -725,15 +740,27 @@ static void tcp_tasklet_func(unsigned long data) + list_del(&tp->tsq_node); + + sk = (struct sock *)tp; +- bh_lock_sock(sk); ++ meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; ++ bh_lock_sock(meta_sk); + +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + tcp_tsq_handler(sk); ++ if (tp->mpc) ++ tcp_tsq_handler(meta_sk); + } else { + /* defer the work to tcp_release_cb() */ + set_bit(TCP_TSQ_DEFERRED, &tp->tsq_flags); ++ ++ /* For MPTCP, we set the tsq-bit on the meta, and the ++ * subflow as we don't know if the limitation happened ++ * while inside mptcp_write_xmit or during tcp_write_xmit. ++ */ ++ if (tp->mpc) { ++ set_bit(TCP_TSQ_DEFERRED, &tcp_sk(meta_sk)->tsq_flags); ++ mptcp_tsq_flags(sk); ++ } + } +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + + clear_bit(TSQ_QUEUED, &tp->tsq_flags); + sk_free(sk); +@@ -743,7 +770,10 @@ static void tcp_tasklet_func(unsigned long data) + #define TCP_DEFERRED_ALL ((1UL << TCP_TSQ_DEFERRED) | \ + (1UL << TCP_WRITE_TIMER_DEFERRED) | \ + (1UL << TCP_DELACK_TIMER_DEFERRED) | \ +- (1UL << TCP_MTU_REDUCED_DEFERRED)) ++ (1UL << TCP_MTU_REDUCED_DEFERRED) | \ ++ (1UL << MPTCP_PATH_MANAGER) | \ ++ (1UL << MPTCP_SUB_DEFERRED)) ++ + /** + * tcp_release_cb - tcp release_sock() callback + * @sk: socket +@@ -790,6 +820,13 @@ void tcp_release_cb(struct sock *sk) + sk->sk_prot->mtu_reduced(sk); + __sock_put(sk); + } ++ if (flags & (1UL << MPTCP_PATH_MANAGER)) { ++ if (tcp_sk(sk)->mpcb->pm_ops->release_sock) ++ tcp_sk(sk)->mpcb->pm_ops->release_sock(sk); ++ __sock_put(sk); ++ } ++ if (flags & (1UL << MPTCP_SUB_DEFERRED)) ++ mptcp_tsq_sub_deferred(sk); + } + EXPORT_SYMBOL(tcp_release_cb); + +@@ -849,8 +886,8 @@ void tcp_wfree(struct sk_buff *skb) + * We are working here with either a clone of the original + * SKB, or a fresh unique copy made by the retransmit engine. + */ +-static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, +- gfp_t gfp_mask) ++int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, ++ gfp_t gfp_mask) + { + const struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_sock *inet; +@@ -878,10 +915,28 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, + NET_INC_STATS(sock_net(sk), + LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES); + +- if (unlikely(skb_cloned(skb))) +- skb = pskb_copy(skb, gfp_mask); +- else ++ if (unlikely(skb_cloned(skb))) { ++ struct sk_buff *newskb; ++ if (mptcp_is_data_seq(skb)) ++ skb_push(skb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ ++ newskb = pskb_copy(skb, gfp_mask); ++ ++ if (mptcp_is_data_seq(skb)) { ++ skb_pull(skb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ if (newskb) ++ skb_pull(newskb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ } ++ skb = newskb; ++ } else { + skb = skb_clone(skb, gfp_mask); ++ } + if (unlikely(!skb)) + return -ENOBUFS; + } +@@ -929,7 +984,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, + */ + th->window = htons(min(tp->rcv_wnd, 65535U)); + } else { +- th->window = htons(tcp_select_window(sk)); ++ th->window = htons(tp->select_window(sk)); + } + th->check = 0; + th->urg_ptr = 0; +@@ -945,7 +1000,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, + } + } + +- tcp_options_write((__be32 *)(th + 1), tp, &opts); ++ tcp_options_write((__be32 *)(th + 1), tp, &opts, skb); + if (likely((tcb->tcp_flags & TCPHDR_SYN) == 0)) + TCP_ECN_send(sk, skb, tcp_header_size); + +@@ -984,7 +1039,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, + * NOTE: probe0 timer is not checked, do not forget tcp_push_pending_frames, + * otherwise socket can stall. + */ +-static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb) ++void tcp_queue_skb(struct sock *sk, struct sk_buff *skb) + { + struct tcp_sock *tp = tcp_sk(sk); + +@@ -997,15 +1052,16 @@ static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb) + } + + /* Initialize TSO segments for a packet. */ +-static void tcp_set_skb_tso_segs(const struct sock *sk, struct sk_buff *skb, +- unsigned int mss_now) ++void tcp_set_skb_tso_segs(const struct sock *sk, struct sk_buff *skb, ++ unsigned int mss_now) + { + struct skb_shared_info *shinfo = skb_shinfo(skb); + + /* Make sure we own this skb before messing gso_size/gso_segs */ + WARN_ON_ONCE(skb_cloned(skb)); + +- if (skb->len <= mss_now || skb->ip_summed == CHECKSUM_NONE) { ++ if (skb->len <= mss_now || (is_meta_sk(sk) && !mptcp_sk_can_gso(sk)) || ++ (!is_meta_sk(sk) && !sk_can_gso(sk)) || skb->ip_summed == CHECKSUM_NONE) { + /* Avoid the costly divide in the normal + * non-TSO case. + */ +@@ -1037,7 +1093,7 @@ static void tcp_adjust_fackets_out(struct sock *sk, const struct sk_buff *skb, + /* Pcount in the middle of the write queue got changed, we need to do various + * tweaks to fix counters + */ +-static void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int decr) ++void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int decr) + { + struct tcp_sock *tp = tcp_sk(sk); + +@@ -1078,6 +1134,9 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, + int nlen; + u8 flags; + ++ if (tcp_sk(sk)->mpc && mptcp_is_data_seq(skb)) ++ mptcp_fragment(sk, skb, len, mss_now, 0); ++ + if (WARN_ON(len > skb->len)) + return -EINVAL; + +@@ -1160,7 +1219,7 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, + * eventually). The difference is that pulled data not copied, but + * immediately discarded. + */ +-static void __pskb_trim_head(struct sk_buff *skb, int len) ++void __pskb_trim_head(struct sk_buff *skb, int len) + { + struct skb_shared_info *shinfo; + int i, k, eat; +@@ -1201,6 +1260,9 @@ static void __pskb_trim_head(struct sk_buff *skb, int len) + /* Remove acked data from a packet in the transmit queue. */ + int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len) + { ++ if (tcp_sk(sk)->mpc && !is_meta_sk(sk) && mptcp_is_data_seq(skb)) ++ return mptcp_trim_head(sk, skb, len); ++ + if (skb_unclone(skb, GFP_ATOMIC)) + return -ENOMEM; + +@@ -1218,6 +1280,15 @@ int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len) + if (tcp_skb_pcount(skb) > 1) + tcp_set_skb_tso_segs(sk, skb, tcp_skb_mss(skb)); + ++#ifdef CONFIG_MPTCP ++ /* Some data got acked - we assume that the seq-number reached the dest. ++ * Anyway, our MPTCP-option has been trimmed above - we lost it here. ++ * Only remove the SEQ if the call does not come from a meta retransmit. ++ */ ++ if (tcp_sk(sk)->mpc && !is_meta_sk(sk)) ++ TCP_SKB_CB(skb)->mptcp_flags &= ~MPTCPHDR_SEQ; ++#endif ++ + return 0; + } + +@@ -1377,7 +1448,7 @@ unsigned int tcp_current_mss(struct sock *sk) + } + + /* Congestion window validation. (RFC2861) */ +-static void tcp_cwnd_validate(struct sock *sk) ++void tcp_cwnd_validate(struct sock *sk) + { + struct tcp_sock *tp = tcp_sk(sk); + +@@ -1411,8 +1482,8 @@ static bool tcp_minshall_check(const struct tcp_sock *tp) + * But we can avoid doing the divide again given we already have + * skb_pcount = skb->len / mss_now + */ +-static void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss_now, +- const struct sk_buff *skb) ++void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss_now, ++ const struct sk_buff *skb) + { + if (skb->len < tcp_skb_pcount(skb) * mss_now) + tp->snd_sml = TCP_SKB_CB(skb)->end_seq; +@@ -1433,19 +1504,28 @@ static bool tcp_nagle_check(bool partial, const struct tcp_sock *tp, + (!nonagle && tp->packets_out && tcp_minshall_check(tp))); + } + /* Returns the portion of skb which can be sent right away */ +-static unsigned int tcp_mss_split_point(const struct sock *sk, +- const struct sk_buff *skb, +- unsigned int mss_now, +- unsigned int max_segs, +- int nonagle) ++unsigned int tcp_mss_split_point(const struct sock *sk, ++ const struct sk_buff *skb, ++ unsigned int mss_now, ++ unsigned int max_segs, ++ int nonagle) + { + const struct tcp_sock *tp = tcp_sk(sk); ++ const struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; + u32 partial, needed, window, max_len; + +- window = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; ++ if (!tp->mpc) ++ window = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; ++ else ++ /* We need to evaluate the available space in the sending window ++ * at the subflow level. However, the subflow seq has not yet ++ * been set. Nevertheless we know that the caller will set it to ++ * write_seq. ++ */ ++ window = tcp_wnd_end(tp) - tp->write_seq; + max_len = mss_now * max_segs; + +- if (likely(max_len <= window && skb != tcp_write_queue_tail(sk))) ++ if (likely(max_len <= window && skb != tcp_write_queue_tail(meta_sk))) + return max_len; + + needed = min(skb->len, window); +@@ -1467,13 +1547,14 @@ static unsigned int tcp_mss_split_point(const struct sock *sk, + /* Can at least one segment of SKB be sent right now, according to the + * congestion window rules? If so, return how many segments are allowed. + */ +-static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp, +- const struct sk_buff *skb) ++unsigned int tcp_cwnd_test(const struct tcp_sock *tp, ++ const struct sk_buff *skb) + { + u32 in_flight, cwnd; + + /* Don't be strict about the congestion window for the final FIN. */ +- if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) && ++ if (skb && ++ ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) || mptcp_is_data_fin(skb)) && + tcp_skb_pcount(skb) == 1) + return 1; + +@@ -1489,8 +1570,8 @@ static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp, + * This must be invoked the first time we consider transmitting + * SKB onto the wire. + */ +-static int tcp_init_tso_segs(const struct sock *sk, struct sk_buff *skb, +- unsigned int mss_now) ++int tcp_init_tso_segs(const struct sock *sk, struct sk_buff *skb, ++ unsigned int mss_now) + { + int tso_segs = tcp_skb_pcount(skb); + +@@ -1505,8 +1586,8 @@ static int tcp_init_tso_segs(const struct sock *sk, struct sk_buff *skb, + /* Return true if the Nagle test allows this packet to be + * sent now. + */ +-static inline bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buff *skb, +- unsigned int cur_mss, int nonagle) ++bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buff *skb, ++ unsigned int cur_mss, int nonagle) + { + /* Nagle rule does not apply to frames, which sit in the middle of the + * write_queue (they have no chances to get new data). +@@ -1518,7 +1599,8 @@ static inline bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buf + return true; + + /* Don't use the nagle rule for urgent data (or for the final FIN). */ +- if (tcp_urg_mode(tp) || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) ++ if (tcp_urg_mode(tp) || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) || ++ mptcp_is_data_fin(skb)) + return true; + + if (!tcp_nagle_check(skb->len < cur_mss, tp, cur_mss, nonagle)) +@@ -1528,9 +1610,8 @@ static inline bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buf + } + + /* Does at least the first segment of SKB fit into the send window? */ +-static bool tcp_snd_wnd_test(const struct tcp_sock *tp, +- const struct sk_buff *skb, +- unsigned int cur_mss) ++bool tcp_snd_wnd_test(const struct tcp_sock *tp, const struct sk_buff *skb, ++ unsigned int cur_mss) + { + u32 end_seq = TCP_SKB_CB(skb)->end_seq; + +@@ -1549,14 +1630,16 @@ static unsigned int tcp_snd_test(const struct sock *sk, struct sk_buff *skb, + { + const struct tcp_sock *tp = tcp_sk(sk); + unsigned int cwnd_quota; ++ const struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; ++ const struct tcp_sock *meta_tp = tcp_sk(meta_sk); + +- tcp_init_tso_segs(sk, skb, cur_mss); ++ tcp_init_tso_segs(meta_sk, skb, cur_mss); + +- if (!tcp_nagle_test(tp, skb, cur_mss, nonagle)) ++ if (!tcp_nagle_test(meta_tp, skb, cur_mss, nonagle)) + return 0; + + cwnd_quota = tcp_cwnd_test(tp, skb); +- if (cwnd_quota && !tcp_snd_wnd_test(tp, skb, cur_mss)) ++ if (cwnd_quota && !tcp_snd_wnd_test(meta_tp, skb, cur_mss)) + cwnd_quota = 0; + + return cwnd_quota; +@@ -1566,12 +1649,16 @@ static unsigned int tcp_snd_test(const struct sock *sk, struct sk_buff *skb, + bool tcp_may_send_now(struct sock *sk) + { + const struct tcp_sock *tp = tcp_sk(sk); +- struct sk_buff *skb = tcp_send_head(sk); ++ struct sk_buff *skb; ++ const struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; ++ const struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ ++ skb = tcp_send_head(meta_sk); + + return skb && + tcp_snd_test(sk, skb, tcp_current_mss(sk), +- (tcp_skb_is_last(sk, skb) ? +- tp->nonagle : TCP_NAGLE_PUSH)); ++ (tcp_skb_is_last(meta_sk, skb) ? ++ meta_tp->nonagle : TCP_NAGLE_PUSH)); + } + + /* Trim TSO SKB to LEN bytes, put the remaining data into a new packet +@@ -1588,6 +1675,9 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, + int nlen = skb->len - len; + u8 flags; + ++ if (tcp_sk(sk)->mpc && mptcp_is_data_seq(skb)) ++ mptso_fragment(sk, skb, len, mss_now, gfp, 0); ++ + /* All of a TSO frame must be composed of paged data. */ + if (skb->len != skb->data_len) + return tcp_fragment(sk, skb, len, mss_now); +@@ -1633,29 +1723,39 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, + * + * This algorithm is from John Heffner. + */ +-static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb) ++bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb) + { + struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); + const struct inet_connection_sock *icsk = inet_csk(sk); + u32 send_win, cong_win, limit, in_flight; + int win_divisor; + +- if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) ++ if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) || mptcp_is_data_fin(skb)) + goto send_now; + + if (icsk->icsk_ca_state != TCP_CA_Open) + goto send_now; + + /* Defer for less than two clock ticks. */ +- if (tp->tso_deferred && +- (((u32)jiffies << 1) >> 1) - (tp->tso_deferred >> 1) > 1) ++ if (meta_tp->tso_deferred && ++ (((u32)jiffies << 1) >> 1) - (meta_tp->tso_deferred >> 1) > 1) + goto send_now; + + in_flight = tcp_packets_in_flight(tp); + + BUG_ON(tcp_skb_pcount(skb) <= 1 || (tp->snd_cwnd <= in_flight)); + +- send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; ++ if (!tp->mpc) ++ send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; ++ else ++ /* We need to evaluate the available space in the sending window ++ * at the subflow level. However, the subflow seq has not yet ++ * been set. Nevertheless we know that the caller will set it to ++ * write_seq. ++ */ ++ send_win = tcp_wnd_end(tp) - tp->write_seq; + + /* From in_flight test above, we know that cwnd > in_flight. */ + cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache; +@@ -1668,7 +1768,7 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb) + goto send_now; + + /* Middle in queue won't get any more data, full sendable already? */ +- if ((skb != tcp_write_queue_tail(sk)) && (limit >= skb->len)) ++ if ((skb != tcp_write_queue_tail(meta_sk)) && (limit >= skb->len)) + goto send_now; + + win_divisor = ACCESS_ONCE(sysctl_tcp_tso_win_divisor); +@@ -1694,13 +1794,13 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb) + /* Ok, it looks like it is advisable to defer. + * Do not rearm the timer if already set to not break TCP ACK clocking. + */ +- if (!tp->tso_deferred) +- tp->tso_deferred = 1 | (jiffies << 1); ++ if (!meta_tp->tso_deferred) ++ meta_tp->tso_deferred = 1 | (jiffies << 1); + + return true; + + send_now: +- tp->tso_deferred = 0; ++ meta_tp->tso_deferred = 0; + return false; + } + +@@ -1713,7 +1813,7 @@ send_now: + * 1 if a probe was sent, + * -1 otherwise + */ +-static int tcp_mtu_probe(struct sock *sk) ++int tcp_mtu_probe(struct sock *sk) + { + struct tcp_sock *tp = tcp_sk(sk); + struct inet_connection_sock *icsk = inet_csk(sk); +@@ -1858,6 +1958,9 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, + int cwnd_quota; + int result; + ++ if (is_meta_sk(sk)) ++ return mptcp_write_xmit(sk, mss_now, nonagle, push_one, gfp); ++ + sent_pkts = 0; + + if (!push_one) { +@@ -2313,6 +2416,10 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *to, + if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + return; + ++ /* Currently not supported for MPTCP - but it should be possible */ ++ if (tp->mpc) ++ return; ++ + tcp_for_write_queue_from_safe(skb, tmp, sk) { + if (!tcp_can_collapse(sk, skb)) + break; +@@ -2410,10 +2517,26 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb) + */ + if (unlikely((NET_IP_ALIGN && ((unsigned long)skb->data & 3)) || + skb_headroom(skb) >= 0xFFFF)) { +- struct sk_buff *nskb = __pskb_copy(skb, MAX_TCP_HEADER, +- GFP_ATOMIC); ++ struct sk_buff *nskb; ++ ++ if (mptcp_is_data_seq(skb)) ++ skb_push(skb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ ++ nskb = __pskb_copy(skb, MAX_TCP_HEADER, GFP_ATOMIC); ++ ++ if (mptcp_is_data_seq(skb)) { ++ skb_pull(skb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ if (nskb) ++ skb_pull(nskb, MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN); ++ } + err = nskb ? tcp_transmit_skb(sk, nskb, 0, GFP_ATOMIC) : +- -ENOBUFS; ++ -ENOBUFS; + } else { + err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC); + } +@@ -2640,6 +2763,11 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority) + { + struct sk_buff *skb; + ++ if (is_meta_sk(sk)) { ++ mptcp_send_active_reset(sk, priority); ++ return; ++ } ++ + /* NOTE: No TCP options attached and we never retransmit this. */ + skb = alloc_skb(MAX_TCP_HEADER, priority); + if (!skb) { +@@ -2742,14 +2870,14 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst, + (req->window_clamp > tcp_full_space(sk) || req->window_clamp == 0)) + req->window_clamp = tcp_full_space(sk); + +- /* tcp_full_space because it is guaranteed to be the first packet */ +- tcp_select_initial_window(tcp_full_space(sk), +- mss - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0), ++ tp->select_initial_window(tcp_full_space(sk), ++ mss - (ireq->tstamp_ok ? TCPOLEN_TSTAMP_ALIGNED : 0) - ++ (tcp_rsk(req)->saw_mpc ? MPTCP_SUB_LEN_DSM_ALIGN : 0), + &req->rcv_wnd, + &req->window_clamp, + ireq->wscale_ok, + &rcv_wscale, +- dst_metric(dst, RTAX_INITRWND)); ++ dst_metric(dst, RTAX_INITRWND), sk); + ireq->rcv_wscale = rcv_wscale; + } + +@@ -2785,7 +2913,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst, + + /* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */ + th->window = htons(min(req->rcv_wnd, 65535U)); +- tcp_options_write((__be32 *)(th + 1), tp, &opts); ++ tcp_options_write((__be32 *)(th + 1), tp, &opts, skb); + th->doff = (tcp_header_size >> 2); + TCP_ADD_STATS(sock_net(sk), TCP_MIB_OUTSEGS, tcp_skb_pcount(skb)); + +@@ -2839,13 +2967,13 @@ static void tcp_connect_init(struct sock *sk) + (tp->window_clamp > tcp_full_space(sk) || tp->window_clamp == 0)) + tp->window_clamp = tcp_full_space(sk); + +- tcp_select_initial_window(tcp_full_space(sk), ++ tp->select_initial_window(tcp_full_space(sk), + tp->advmss - (tp->rx_opt.ts_recent_stamp ? tp->tcp_header_len - sizeof(struct tcphdr) : 0), + &tp->rcv_wnd, + &tp->window_clamp, + sysctl_tcp_window_scaling, + &rcv_wscale, +- dst_metric(dst, RTAX_INITRWND)); ++ dst_metric(dst, RTAX_INITRWND), sk); + + tp->rx_opt.rcv_wscale = rcv_wscale; + tp->rcv_ssthresh = tp->rcv_wnd; +@@ -2869,6 +2997,38 @@ static void tcp_connect_init(struct sock *sk) + inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT; + inet_csk(sk)->icsk_retransmits = 0; + tcp_clear_retrans(tp); ++ ++#ifdef CONFIG_MPTCP ++ if (sysctl_mptcp_enabled && mptcp_doit(sk)) { ++ if (is_master_tp(tp)) { ++ tp->request_mptcp = 1; ++ mptcp_connect_init(sk); ++ } else if (tp->mptcp) { ++ struct inet_sock *inet = inet_sk(sk); ++ ++ tp->mptcp->snt_isn = tp->write_seq; ++ tp->mptcp->init_rcv_wnd = tp->rcv_wnd; ++ ++ /* Set nonce for new subflows */ ++ if (sk->sk_family == AF_INET) ++ tp->mptcp->mptcp_loc_nonce = mptcp_v4_get_nonce( ++ inet->inet_saddr, ++ inet->inet_daddr, ++ inet->inet_sport, ++ inet->inet_dport, ++ tp->write_seq); ++#if IS_ENABLED(CONFIG_IPV6) ++ else ++ tp->mptcp->mptcp_loc_nonce = mptcp_v6_get_nonce( ++ inet6_sk(sk)->saddr.s6_addr32, ++ sk->sk_v6_daddr.s6_addr32, ++ inet->inet_sport, ++ inet->inet_dport, ++ tp->write_seq); ++#endif ++ } ++ } ++#endif + } + + static void tcp_connect_queue_skb(struct sock *sk, struct sk_buff *skb) +@@ -3111,6 +3271,7 @@ void tcp_send_ack(struct sock *sk) + TCP_SKB_CB(buff)->when = tcp_time_stamp; + tcp_transmit_skb(sk, buff, 0, sk_gfp_atomic(sk, GFP_ATOMIC)); + } ++EXPORT_SYMBOL(tcp_send_ack); + + /* This routine sends a packet with an out of date sequence + * number. It assumes the other end will try to ack it. +@@ -3123,7 +3284,7 @@ void tcp_send_ack(struct sock *sk) + * one is with SEG.SEQ=SND.UNA to deliver urgent pointer, another is + * out-of-date with SND.UNA-1 to probe window. + */ +-static int tcp_xmit_probe_skb(struct sock *sk, int urgent) ++int tcp_xmit_probe_skb(struct sock *sk, int urgent) + { + struct tcp_sock *tp = tcp_sk(sk); + struct sk_buff *skb; +@@ -3161,6 +3322,9 @@ int tcp_write_wakeup(struct sock *sk) + if (sk->sk_state == TCP_CLOSE) + return -1; + ++ if (is_meta_sk(sk)) ++ return mptcp_write_wakeup(sk); ++ + if ((skb = tcp_send_head(sk)) != NULL && + before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp))) { + int err; +diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c +index 64f0354..7b55b9a 100644 +--- a/net/ipv4/tcp_timer.c ++++ b/net/ipv4/tcp_timer.c +@@ -20,6 +20,7 @@ + + #include <linux/module.h> + #include <linux/gfp.h> ++#include <net/mptcp.h> + #include <net/tcp.h> + + int sysctl_tcp_syn_retries __read_mostly = TCP_SYN_RETRIES; +@@ -32,7 +33,7 @@ int sysctl_tcp_retries2 __read_mostly = TCP_RETR2; + int sysctl_tcp_orphan_retries __read_mostly; + int sysctl_tcp_thin_linear_timeouts __read_mostly; + +-static void tcp_write_err(struct sock *sk) ++void tcp_write_err(struct sock *sk) + { + sk->sk_err = sk->sk_err_soft ? : ETIMEDOUT; + sk->sk_error_report(sk); +@@ -124,10 +125,8 @@ static void tcp_mtu_probing(struct inet_connection_sock *icsk, struct sock *sk) + * retransmissions with an initial RTO of TCP_RTO_MIN or TCP_TIMEOUT_INIT if + * syn_set flag is set. + */ +-static bool retransmits_timed_out(struct sock *sk, +- unsigned int boundary, +- unsigned int timeout, +- bool syn_set) ++bool retransmits_timed_out(struct sock *sk, unsigned int boundary, ++ unsigned int timeout, bool syn_set) + { + unsigned int linear_backoff_thresh, start_ts; + unsigned int rto_base = syn_set ? TCP_TIMEOUT_INIT : TCP_RTO_MIN; +@@ -153,7 +152,7 @@ static bool retransmits_timed_out(struct sock *sk, + } + + /* A write timeout has occurred. Process the after effects. */ +-static int tcp_write_timeout(struct sock *sk) ++int tcp_write_timeout(struct sock *sk) + { + struct inet_connection_sock *icsk = inet_csk(sk); + struct tcp_sock *tp = tcp_sk(sk); +@@ -168,6 +167,10 @@ static int tcp_write_timeout(struct sock *sk) + } + retry_until = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; + syn_set = true; ++ /* Stop retransmitting MP_CAPABLE options in SYN if timed out. */ ++ if (tcp_sk(sk)->request_mptcp && ++ icsk->icsk_retransmits >= mptcp_sysctl_syn_retries()) ++ tcp_sk(sk)->request_mptcp = 0; + } else { + if (retransmits_timed_out(sk, sysctl_tcp_retries1, 0, 0)) { + /* Black hole detection */ +@@ -248,18 +251,22 @@ out: + static void tcp_delack_timer(unsigned long data) + { + struct sock *sk = (struct sock *)data; ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; + +- bh_lock_sock(sk); +- if (!sock_owned_by_user(sk)) { ++ bh_lock_sock(meta_sk); ++ if (!sock_owned_by_user(meta_sk)) { + tcp_delack_timer_handler(sk); + } else { + inet_csk(sk)->icsk_ack.blocked = 1; +- NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_DELAYEDACKLOCKED); ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_DELAYEDACKLOCKED); + /* deleguate our work to tcp_release_cb() */ + if (!test_and_set_bit(TCP_DELACK_TIMER_DEFERRED, &tcp_sk(sk)->tsq_flags)) + sock_hold(sk); ++ if (tp->mpc) ++ mptcp_tsq_flags(sk); + } +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + sock_put(sk); + } + +@@ -421,6 +428,9 @@ void tcp_retransmit_timer(struct sock *sk) + + tcp_enter_loss(sk, 0); + ++ if (tp->mpc) ++ mptcp_reinject_data(sk, 1); ++ + if (tcp_retransmit_skb(sk, tcp_write_queue_head(sk)) > 0) { + /* Retransmission failed because of local congestion, + * do not backoff. +@@ -471,6 +481,8 @@ out_reset_timer: + /* Use normal (exponential) backoff */ + icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); + } ++ if (tp->mpc) ++ mptcp_set_rto(sk); + inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, TCP_RTO_MAX); + if (retransmits_timed_out(sk, sysctl_tcp_retries1 + 1, 0, 0)) + __sk_dst_reset(sk); +@@ -502,7 +514,10 @@ void tcp_write_timer_handler(struct sock *sk) + break; + case ICSK_TIME_RETRANS: + icsk->icsk_pending = 0; +- tcp_retransmit_timer(sk); ++ if (is_meta_sk(sk)) ++ mptcp_retransmit_timer(sk); ++ else ++ tcp_retransmit_timer(sk); + break; + case ICSK_TIME_PROBE0: + icsk->icsk_pending = 0; +@@ -517,16 +532,19 @@ out: + static void tcp_write_timer(unsigned long data) + { + struct sock *sk = (struct sock *)data; ++ struct sock *meta_sk = tcp_sk(sk)->mpc ? mptcp_meta_sk(sk) : sk; + +- bh_lock_sock(sk); +- if (!sock_owned_by_user(sk)) { ++ bh_lock_sock(meta_sk); ++ if (!sock_owned_by_user(meta_sk)) { + tcp_write_timer_handler(sk); + } else { + /* deleguate our work to tcp_release_cb() */ + if (!test_and_set_bit(TCP_WRITE_TIMER_DEFERRED, &tcp_sk(sk)->tsq_flags)) + sock_hold(sk); ++ if (tcp_sk(sk)->mpc) ++ mptcp_tsq_flags(sk); + } +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + sock_put(sk); + } + +@@ -563,11 +581,12 @@ static void tcp_keepalive_timer (unsigned long data) + struct sock *sk = (struct sock *) data; + struct inet_connection_sock *icsk = inet_csk(sk); + struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = tp->mpc ? mptcp_meta_sk(sk) : sk; + u32 elapsed; + + /* Only process if socket is not in use. */ +- bh_lock_sock(sk); +- if (sock_owned_by_user(sk)) { ++ bh_lock_sock(meta_sk); ++ if (sock_owned_by_user(meta_sk)) { + /* Try again later. */ + inet_csk_reset_keepalive_timer (sk, HZ/20); + goto out; +@@ -578,6 +597,29 @@ static void tcp_keepalive_timer (unsigned long data) + goto out; + } + ++ if (tp->send_mp_fclose) { ++ /* MUST do this before tcp_write_timeout, because retrans_stamp ++ * may have been set to 0 in another part while we are ++ * retransmitting MP_FASTCLOSE. Then, we would crash, because ++ * retransmits_timed_out accesses the meta-write-queue. ++ * ++ * We make sure that the timestamp is != 0. ++ */ ++ if (!tp->retrans_stamp) ++ tp->retrans_stamp = tcp_time_stamp ? : 1; ++ ++ if (tcp_write_timeout(sk)) ++ goto out; ++ ++ tcp_send_ack(sk); ++ icsk->icsk_backoff++; ++ icsk->icsk_retransmits++; ++ ++ icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); ++ elapsed = icsk->icsk_rto; ++ goto resched; ++ } ++ + if (sk->sk_state == TCP_FIN_WAIT2 && sock_flag(sk, SOCK_DEAD)) { + if (tp->linger2 >= 0) { + const int tmo = tcp_fin_time(sk) - TCP_TIMEWAIT_LEN; +@@ -639,7 +681,7 @@ death: + tcp_done(sk); + + out: +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + sock_put(sk); + } + +diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c +index 6c7fa08..733d602 100644 +--- a/net/ipv6/addrconf.c ++++ b/net/ipv6/addrconf.c +@@ -765,6 +765,7 @@ void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp) + + kfree_rcu(ifp, rcu); + } ++EXPORT_SYMBOL(inet6_ifa_finish_destroy); + + static void + ipv6_link_dev_addr(struct inet6_dev *idev, struct inet6_ifaddr *ifp) +diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c +index d935889..9f0fd80 100644 +--- a/net/ipv6/af_inet6.c ++++ b/net/ipv6/af_inet6.c +@@ -97,8 +97,7 @@ static __inline__ struct ipv6_pinfo *inet6_sk_generic(struct sock *sk) + return (struct ipv6_pinfo *)(((u8 *)sk) + offset); + } + +-static int inet6_create(struct net *net, struct socket *sock, int protocol, +- int kern) ++int inet6_create(struct net *net, struct socket *sock, int protocol, int kern) + { + struct inet_sock *inet; + struct ipv6_pinfo *np; +diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c +index c913818..2f5b4c5 100644 +--- a/net/ipv6/inet6_connection_sock.c ++++ b/net/ipv6/inet6_connection_sock.c +@@ -96,8 +96,8 @@ struct dst_entry *inet6_csk_route_req(struct sock *sk, + /* + * request_sock (formerly open request) hash tables. + */ +-static u32 inet6_synq_hash(const struct in6_addr *raddr, const __be16 rport, +- const u32 rnd, const u32 synq_hsize) ++u32 inet6_synq_hash(const struct in6_addr *raddr, const __be16 rport, ++ const u32 rnd, const u32 synq_hsize) + { + u32 c; + +diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c +index bb53a5e7..0d29995 100644 +--- a/net/ipv6/syncookies.c ++++ b/net/ipv6/syncookies.c +@@ -181,7 +181,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) + + /* check for timestamp cookie support */ + memset(&tcp_opt, 0, sizeof(tcp_opt)); +- tcp_parse_options(skb, &tcp_opt, 0, NULL); ++ tcp_parse_options(skb, &tcp_opt, NULL, 0, NULL); + + if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok)) + goto out; +@@ -253,10 +253,10 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) + } + + req->window_clamp = tp->window_clamp ? :dst_metric(dst, RTAX_WINDOW); +- tcp_select_initial_window(tcp_full_space(sk), req->mss, ++ tp->select_initial_window(tcp_full_space(sk), req->mss, + &req->rcv_wnd, &req->window_clamp, + ireq->wscale_ok, &rcv_wscale, +- dst_metric(dst, RTAX_INITRWND)); ++ dst_metric(dst, RTAX_INITRWND), sk); + + ireq->rcv_wscale = rcv_wscale; + +diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c +index 889079b..d7f8b5f 100644 +--- a/net/ipv6/tcp_ipv6.c ++++ b/net/ipv6/tcp_ipv6.c +@@ -63,6 +63,8 @@ + #include <net/inet_common.h> + #include <net/secure_seq.h> + #include <net/tcp_memcontrol.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v6.h> + #include <net/busy_poll.h> + + #include <asm/uaccess.h> +@@ -73,14 +75,6 @@ + #include <linux/crypto.h> + #include <linux/scatterlist.h> + +-static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb); +-static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, +- struct request_sock *req); +- +-static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb); +- +-static const struct inet_connection_sock_af_ops ipv6_mapped; +-static const struct inet_connection_sock_af_ops ipv6_specific; + #ifdef CONFIG_TCP_MD5SIG + static const struct tcp_sock_af_ops tcp_sock_ipv6_specific; + static const struct tcp_sock_af_ops tcp_sock_ipv6_mapped_specific; +@@ -92,7 +86,7 @@ static struct tcp_md5sig_key *tcp_v6_md5_do_lookup(struct sock *sk, + } + #endif + +-static void inet6_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb) ++void inet6_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb) + { + struct dst_entry *dst = skb_dst(skb); + const struct rt6_info *rt = (const struct rt6_info *)dst; +@@ -104,7 +98,7 @@ static void inet6_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb) + inet6_sk(sk)->rx_dst_cookie = rt->rt6i_node->fn_sernum; + } + +-static void tcp_v6_hash(struct sock *sk) ++void tcp_v6_hash(struct sock *sk) + { + if (sk->sk_state != TCP_CLOSE) { + if (inet_csk(sk)->icsk_af_ops == &ipv6_mapped) { +@@ -117,7 +111,7 @@ static void tcp_v6_hash(struct sock *sk) + } + } + +-static __u32 tcp_v6_init_sequence(const struct sk_buff *skb) ++__u32 tcp_v6_init_sequence(const struct sk_buff *skb) + { + return secure_tcpv6_sequence_number(ipv6_hdr(skb)->daddr.s6_addr32, + ipv6_hdr(skb)->saddr.s6_addr32, +@@ -125,7 +119,7 @@ static __u32 tcp_v6_init_sequence(const struct sk_buff *skb) + tcp_hdr(skb)->source); + } + +-static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, ++int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, + int addr_len) + { + struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr; +@@ -339,7 +333,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + const struct ipv6hdr *hdr = (const struct ipv6hdr *)skb->data; + const struct tcphdr *th = (struct tcphdr *)(skb->data+offset); + struct ipv6_pinfo *np; +- struct sock *sk; ++ struct sock *sk, *meta_sk; + int err; + struct tcp_sock *tp; + __u32 seq; +@@ -359,8 +353,14 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + return; + } + +- bh_lock_sock(sk); +- if (sock_owned_by_user(sk) && type != ICMPV6_PKT_TOOBIG) ++ tp = tcp_sk(sk); ++ if (tp->mpc) ++ meta_sk = mptcp_meta_sk(sk); ++ else ++ meta_sk = sk; ++ ++ bh_lock_sock(meta_sk); ++ if (sock_owned_by_user(meta_sk) && type != ICMPV6_PKT_TOOBIG) + NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS); + + if (sk->sk_state == TCP_CLOSE) +@@ -371,7 +371,6 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + goto out; + } + +- tp = tcp_sk(sk); + seq = ntohl(th->seq); + if (sk->sk_state != TCP_LISTEN && + !between(seq, tp->snd_una, tp->snd_nxt)) { +@@ -401,11 +400,15 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + goto out; + + tp->mtu_info = ntohl(info); +- if (!sock_owned_by_user(sk)) ++ if (!sock_owned_by_user(meta_sk)) + tcp_v6_mtu_reduced(sk); +- else if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, ++ else { ++ if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, + &tp->tsq_flags)) +- sock_hold(sk); ++ sock_hold(sk); ++ if (tp->mpc) ++ mptcp_tsq_flags(sk); ++ } + goto out; + } + +@@ -415,7 +418,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + switch (sk->sk_state) { + struct request_sock *req, **prev; + case TCP_LISTEN: +- if (sock_owned_by_user(sk)) ++ if (sock_owned_by_user(meta_sk)) + goto out; + + req = inet6_csk_search_req(sk, &prev, th->dest, &hdr->daddr, +@@ -440,7 +443,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + case TCP_SYN_SENT: + case TCP_SYN_RECV: /* Cannot happen. + It can, it SYNs are crossed. --ANK */ +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + sk->sk_err = err; + sk->sk_error_report(sk); /* Wake people up to see the error (see connect in sock.c) */ + +@@ -450,22 +453,22 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + goto out; + } + +- if (!sock_owned_by_user(sk) && np->recverr) { ++ if (!sock_owned_by_user(meta_sk) && np->recverr) { + sk->sk_err = err; + sk->sk_error_report(sk); + } else + sk->sk_err_soft = err; + + out: +- bh_unlock_sock(sk); ++ bh_unlock_sock(meta_sk); + sock_put(sk); + } + + +-static int tcp_v6_send_synack(struct sock *sk, struct dst_entry *dst, +- struct flowi6 *fl6, +- struct request_sock *req, +- u16 queue_mapping) ++int tcp_v6_send_synack(struct sock *sk, struct dst_entry *dst, ++ struct flowi6 *fl6, ++ struct request_sock *req, ++ u16 queue_mapping) + { + struct inet_request_sock *ireq = inet_rsk(req); + struct ipv6_pinfo *np = inet6_sk(sk); +@@ -495,7 +498,7 @@ done: + return err; + } + +-static int tcp_v6_rtx_synack(struct sock *sk, struct request_sock *req) ++int tcp_v6_rtx_synack(struct sock *sk, struct request_sock *req) + { + struct flowi6 fl6; + int res; +@@ -506,7 +509,7 @@ static int tcp_v6_rtx_synack(struct sock *sk, struct request_sock *req) + return res; + } + +-static void tcp_v6_reqsk_destructor(struct request_sock *req) ++void tcp_v6_reqsk_destructor(struct request_sock *req) + { + kfree_skb(inet_rsk(req)->pktopts); + } +@@ -719,16 +722,16 @@ struct request_sock_ops tcp6_request_sock_ops __read_mostly = { + }; + + #ifdef CONFIG_TCP_MD5SIG +-static const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = { ++const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = { + .md5_lookup = tcp_v6_reqsk_md5_lookup, + .calc_md5_hash = tcp_v6_md5_hash_skb, + }; + #endif + +-static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win, +- u32 tsval, u32 tsecr, ++static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, ++ u32 data_ack, u32 win, u32 tsval, u32 tsecr, + struct tcp_md5sig_key *key, int rst, u8 tclass, +- u32 label) ++ u32 label, int mptcp) + { + const struct tcphdr *th = tcp_hdr(skb); + struct tcphdr *t1; +@@ -746,7 +749,10 @@ static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win, + if (key) + tot_len += TCPOLEN_MD5SIG_ALIGNED; + #endif +- ++#ifdef CONFIG_MPTCP ++ if (mptcp) ++ tot_len += MPTCP_SUB_LEN_DSS + MPTCP_SUB_LEN_ACK; ++#endif + buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len, + GFP_ATOMIC); + if (buff == NULL) +@@ -784,6 +790,17 @@ static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win, + tcp_v6_md5_hash_hdr((__u8 *)topt, key, + &ipv6_hdr(skb)->saddr, + &ipv6_hdr(skb)->daddr, t1); ++ topt += 4; ++ } ++#endif ++#ifdef CONFIG_MPTCP ++ if (mptcp) { ++ /* Construction of 32-bit data_ack */ ++ *topt++ = htonl((TCPOPT_MPTCP << 24) | ++ ((MPTCP_SUB_LEN_DSS + MPTCP_SUB_LEN_ACK) << 16) | ++ (0x20 << 8) | ++ (0x01)); ++ *topt++ = htonl(data_ack); + } + #endif + +@@ -821,7 +838,7 @@ static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win, + kfree_skb(buff); + } + +-static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb) ++void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb) + { + const struct tcphdr *th = tcp_hdr(skb); + u32 seq = 0, ack_seq = 0; +@@ -876,7 +893,7 @@ static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb) + ack_seq = ntohl(th->seq) + th->syn + th->fin + skb->len - + (th->doff << 2); + +- tcp_v6_send_response(skb, seq, ack_seq, 0, 0, 0, key, 1, 0, 0); ++ tcp_v6_send_response(skb, seq, ack_seq, 0, 0, 0, 0, key, 1, 0, 0, 0); + + #ifdef CONFIG_TCP_MD5SIG + release_sk1: +@@ -887,40 +904,47 @@ release_sk1: + #endif + } + +-static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, ++static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 data_ack, + u32 win, u32 tsval, u32 tsecr, +- struct tcp_md5sig_key *key, u8 tclass, +- u32 label) ++ struct tcp_md5sig_key *key, u8 tclass, u32 label, ++ int mptcp) + { +- tcp_v6_send_response(skb, seq, ack, win, tsval, tsecr, key, 0, tclass, +- label); ++ tcp_v6_send_response(skb, seq, ack, data_ack, win, tsval, tsecr, key, 0, ++ tclass, label, mptcp); + } + + static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb) + { + struct inet_timewait_sock *tw = inet_twsk(sk); + struct tcp_timewait_sock *tcptw = tcp_twsk(sk); ++ u32 data_ack = 0; ++ int mptcp = 0; + ++ if (tcptw->mptcp_tw && tcptw->mptcp_tw->meta_tw) { ++ data_ack = (u32)tcptw->mptcp_tw->rcv_nxt; ++ mptcp = 1; ++ } + tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt, ++ data_ack, + tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale, + tcp_time_stamp + tcptw->tw_ts_offset, + tcptw->tw_ts_recent, tcp_twsk_md5_key(tcptw), +- tw->tw_tclass, (tw->tw_flowlabel << 12)); ++ tw->tw_tclass, (tw->tw_flowlabel << 12), mptcp); + + inet_twsk_put(tw); + } + +-static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, +- struct request_sock *req) ++void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req) + { + tcp_v6_send_ack(skb, tcp_rsk(req)->snt_isn + 1, tcp_rsk(req)->rcv_isn + 1, +- req->rcv_wnd, tcp_time_stamp, req->ts_recent, ++ 0, req->rcv_wnd, tcp_time_stamp, req->ts_recent, + tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), +- 0, 0); ++ 0, 0, 0); + } + + +-static struct sock *tcp_v6_hnd_req(struct sock *sk, struct sk_buff *skb) ++struct sock *tcp_v6_hnd_req(struct sock *sk, struct sk_buff *skb) + { + struct request_sock *req, **prev; + const struct tcphdr *th = tcp_hdr(skb); +@@ -939,7 +963,13 @@ static struct sock *tcp_v6_hnd_req(struct sock *sk, struct sk_buff *skb) + + if (nsk) { + if (nsk->sk_state != TCP_TIME_WAIT) { ++ /* Don't lock again the meta-sk. It has been locked ++ * before mptcp_v6_do_rcv. ++ */ ++ if (tcp_sk(nsk)->mpc && !is_meta_sk(sk)) ++ bh_lock_sock(mptcp_meta_sk(nsk)); + bh_lock_sock(nsk); ++ + return nsk; + } + inet_twsk_put(inet_twsk(nsk)); +@@ -959,6 +989,7 @@ static struct sock *tcp_v6_hnd_req(struct sock *sk, struct sk_buff *skb) + static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) + { + struct tcp_options_received tmp_opt; ++ struct mptcp_options_received mopt; + struct request_sock *req; + struct inet_request_sock *ireq; + struct ipv6_pinfo *np = inet6_sk(sk); +@@ -971,6 +1002,23 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) + if (skb->protocol == htons(ETH_P_IP)) + return tcp_v4_conn_request(sk, skb); + ++ tcp_clear_options(&tmp_opt); ++ tmp_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr); ++ tmp_opt.user_mss = tp->rx_opt.user_mss; ++ mptcp_init_mp_opt(&mopt); ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, NULL); ++ ++#ifdef CONFIG_MPTCP ++ /*MPTCP structures not initialized, so return error */ ++ if (mptcp_init_failed) ++ mptcp_init_mp_opt(&mopt); ++ ++ if (mopt.is_mp_join) ++ return mptcp_do_join_short(skb, &mopt, &tmp_opt, sock_net(sk)); ++ if (mopt.drop_me) ++ goto drop; ++#endif ++ + if (!ipv6_unicast_destination(skb)) + goto drop; + +@@ -986,7 +1034,22 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) + goto drop; + } + +- req = inet6_reqsk_alloc(&tcp6_request_sock_ops); ++#ifdef CONFIG_MPTCP ++ if (sysctl_mptcp_enabled == MPTCP_APP && !tp->mptcp_enabled) ++ mopt.saw_mpc = 0; ++ if (mopt.saw_mpc && !want_cookie) { ++ req = inet6_reqsk_alloc(&mptcp6_request_sock_ops); ++ ++ if (req == NULL) ++ goto drop; ++ ++ mptcp_rsk(req)->mpcb = NULL; ++ mptcp_rsk(req)->dss_csum = mopt.dss_csum; ++ mptcp_rsk(req)->collide_tk.pprev = NULL; ++ } else ++#endif ++ req = inet6_reqsk_alloc(&tcp6_request_sock_ops); ++ + if (req == NULL) + goto drop; + +@@ -994,17 +1057,15 @@ static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb) + tcp_rsk(req)->af_specific = &tcp_request_sock_ipv6_ops; + #endif + +- tcp_clear_options(&tmp_opt); +- tmp_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr); +- tmp_opt.user_mss = tp->rx_opt.user_mss; +- tcp_parse_options(skb, &tmp_opt, 0, NULL); +- + if (want_cookie && !tmp_opt.saw_tstamp) + tcp_clear_options(&tmp_opt); + + tmp_opt.tstamp_ok = tmp_opt.saw_tstamp; + tcp_openreq_init(req, &tmp_opt, skb); + ++ if (mopt.saw_mpc && !want_cookie) ++ mptcp_reqsk_new_mptcp(req, &tmp_opt, &mopt, skb); ++ + ireq = inet_rsk(req); + ireq->ir_v6_rmt_addr = ipv6_hdr(skb)->saddr; + ireq->ir_v6_loc_addr = ipv6_hdr(skb)->daddr; +@@ -1094,9 +1155,9 @@ drop: + return 0; /* don't send reset */ + } + +-static struct sock *tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, +- struct request_sock *req, +- struct dst_entry *dst) ++struct sock *tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst) + { + struct inet_request_sock *ireq; + struct ipv6_pinfo *newnp, *np = inet6_sk(sk); +@@ -1317,7 +1378,7 @@ static __sum16 tcp_v6_checksum_init(struct sk_buff *skb) + * This is because we cannot sleep with the original spinlock + * held. + */ +-static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) ++int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) + { + struct ipv6_pinfo *np = inet6_sk(sk); + struct tcp_sock *tp; +@@ -1339,6 +1400,9 @@ static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) + goto discard; + #endif + ++ if (is_meta_sk(sk)) ++ return mptcp_v6_do_rcv(sk, skb); ++ + if (sk_filter(sk, skb)) + goto discard; + +@@ -1460,7 +1524,7 @@ static int tcp_v6_rcv(struct sk_buff *skb) + { + const struct tcphdr *th; + const struct ipv6hdr *hdr; +- struct sock *sk; ++ struct sock *sk, *meta_sk = NULL; + int ret; + struct net *net = dev_net(skb->dev); + +@@ -1491,18 +1555,43 @@ static int tcp_v6_rcv(struct sk_buff *skb) + TCP_SKB_CB(skb)->end_seq = (TCP_SKB_CB(skb)->seq + th->syn + th->fin + + skb->len - th->doff*4); + TCP_SKB_CB(skb)->ack_seq = ntohl(th->ack_seq); ++#ifdef CONFIG_MPTCP ++ TCP_SKB_CB(skb)->mptcp_flags = 0; ++ TCP_SKB_CB(skb)->dss_off = 0; ++#endif + TCP_SKB_CB(skb)->when = 0; + TCP_SKB_CB(skb)->ip_dsfield = ipv6_get_dsfield(hdr); + TCP_SKB_CB(skb)->sacked = 0; + + sk = __inet6_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest); +- if (!sk) +- goto no_tcp_socket; + + process: +- if (sk->sk_state == TCP_TIME_WAIT) ++ if (sk && sk->sk_state == TCP_TIME_WAIT) + goto do_time_wait; + ++#ifdef CONFIG_MPTCP ++ if (!sk && th->syn && !th->ack) { ++ int ret = mptcp_lookup_join(skb, NULL); ++ ++ if (ret < 0) { ++ tcp_v6_send_reset(NULL, skb); ++ goto discard_it; ++ } else if (ret > 0) { ++ return 0; ++ } ++ } ++ ++ /* Is there a pending request sock for this segment ? */ ++ if ((!sk || sk->sk_state == TCP_LISTEN) && mptcp_check_req(skb, net)) { ++ if (sk) ++ sock_put(sk); ++ return 0; ++ } ++#endif ++ ++ if (!sk) ++ goto no_tcp_socket; ++ + if (hdr->hop_limit < inet6_sk(sk)->min_hopcount) { + NET_INC_STATS_BH(net, LINUX_MIB_TCPMINTTLDROP); + goto discard_and_relse; +@@ -1517,11 +1606,21 @@ process: + sk_mark_napi_id(sk, skb); + skb->dev = NULL; + +- bh_lock_sock_nested(sk); ++ if (tcp_sk(sk)->mpc) { ++ meta_sk = mptcp_meta_sk(sk); ++ ++ bh_lock_sock_nested(meta_sk); ++ if (sock_owned_by_user(meta_sk)) ++ skb->sk = sk; ++ } else { ++ meta_sk = sk; ++ bh_lock_sock_nested(sk); ++ } ++ + ret = 0; +- if (!sock_owned_by_user(sk)) { ++ if (!sock_owned_by_user(meta_sk)) { + #ifdef CONFIG_NET_DMA +- struct tcp_sock *tp = tcp_sk(sk); ++ struct tcp_sock *tp = tcp_sk(meta_sk); + if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list) + tp->ucopy.dma_chan = net_dma_find_channel(); + if (tp->ucopy.dma_chan) +@@ -1529,16 +1628,17 @@ process: + else + #endif + { +- if (!tcp_prequeue(sk, skb)) ++ if (!tcp_prequeue(meta_sk, skb)) + ret = tcp_v6_do_rcv(sk, skb); + } +- } else if (unlikely(sk_add_backlog(sk, skb, +- sk->sk_rcvbuf + sk->sk_sndbuf))) { +- bh_unlock_sock(sk); ++ } else if (unlikely(sk_add_backlog(meta_sk, skb, ++ meta_sk->sk_rcvbuf + meta_sk->sk_sndbuf))) { ++ bh_unlock_sock(meta_sk); + NET_INC_STATS_BH(net, LINUX_MIB_TCPBACKLOGDROP); + goto discard_and_relse; + } +- bh_unlock_sock(sk); ++ ++ bh_unlock_sock(meta_sk); + + sock_put(sk); + return ret ? -1 : 0; +@@ -1595,6 +1695,18 @@ do_time_wait: + sk = sk2; + goto process; + } ++#ifdef CONFIG_MPTCP ++ if (th->syn && !th->ack) { ++ int ret = mptcp_lookup_join(skb, inet_twsk(sk)); ++ ++ if (ret < 0) { ++ tcp_v6_send_reset(NULL, skb); ++ goto discard_it; ++ } else if (ret > 0) { ++ return 0; ++ } ++ } ++#endif + /* Fall through to ACK */ + } + case TCP_TW_ACK: +@@ -1644,13 +1756,13 @@ static void tcp_v6_early_demux(struct sk_buff *skb) + } + } + +-static struct timewait_sock_ops tcp6_timewait_sock_ops = { ++struct timewait_sock_ops tcp6_timewait_sock_ops = { + .twsk_obj_size = sizeof(struct tcp6_timewait_sock), + .twsk_unique = tcp_twsk_unique, + .twsk_destructor= tcp_twsk_destructor, + }; + +-static const struct inet_connection_sock_af_ops ipv6_specific = { ++const struct inet_connection_sock_af_ops ipv6_specific = { + .queue_xmit = inet6_csk_xmit, + .send_check = tcp_v6_send_check, + .rebuild_header = inet6_sk_rebuild_header, +@@ -1682,7 +1794,7 @@ static const struct tcp_sock_af_ops tcp_sock_ipv6_specific = { + * TCP over IPv4 via INET6 API + */ + +-static const struct inet_connection_sock_af_ops ipv6_mapped = { ++const struct inet_connection_sock_af_ops ipv6_mapped = { + .queue_xmit = ip_queue_xmit, + .send_check = tcp_v4_send_check, + .rebuild_header = inet_sk_rebuild_header, +@@ -1727,7 +1839,7 @@ static int tcp_v6_init_sock(struct sock *sk) + return 0; + } + +-static void tcp_v6_destroy_sock(struct sock *sk) ++void tcp_v6_destroy_sock(struct sock *sk) + { + tcp_v4_destroy_sock(sk); + inet6_destroy_sock(sk); +diff --git a/net/mptcp/Kconfig b/net/mptcp/Kconfig +new file mode 100644 +index 0000000..88a05b1 +--- /dev/null ++++ b/net/mptcp/Kconfig +@@ -0,0 +1,58 @@ ++# ++# MPTCP configuration ++# ++config MPTCP ++ bool "MPTCP protocol" ++ depends on (IPV6=y || IPV6=n) ++ ---help--- ++ This replaces the normal TCP stack with a Multipath TCP stack, ++ able to use several paths at once. ++ ++menuconfig MPTCP_PM_ADVANCED ++ bool "MPTCP: advanced path-manager control" ++ depends on MPTCP=y ++ ---help--- ++ Support for selection of different path-managers. You should choose 'Y' here, ++ because otherwise you will not actively create new MPTCP-subflows. ++ ++if MPTCP_PM_ADVANCED ++ ++config MPTCP_FULLMESH ++ tristate "MPTCP Full-Mesh Path-Manager" ++ depends on MPTCP=y ++ ---help--- ++ This path-management module will create a full-mesh among all IP-addresses. ++ ++config MPTCP_NDIFFPORTS ++ tristate "MPTCP ndiff-ports" ++ depends on MPTCP=y ++ ---help--- ++ This path-management module will create multiple subflows between the same ++ pair of IP-addresses, modifying the source-port. You can set the number ++ of subflows via the mptcp_ndiffports-sysctl. ++ ++choice ++ prompt "Default MPTCP Path-Manager" ++ default DEFAULT ++ help ++ Select the Path-Manager of your choice ++ ++ config DEFAULT_FULLMESH ++ bool "Full mesh" if MPTCP_FULLMESH=y ++ ++ config DEFAULT_NDIFFPORTS ++ bool "ndiff-ports" if MPTCP_NDIFFPORTS=y ++ ++ config DEFAULT_DUMMY ++ bool "Default" ++ ++endchoice ++ ++endif ++ ++config DEFAULT_MPTCP_PM ++ string ++ default "default" if DEFAULT_DUMMY ++ default "fullmesh" if DEFAULT_FULLMESH ++ default "ndiffports" if DEFAULT_NDIFFPORTS ++ default "default" +diff --git a/net/mptcp/Makefile b/net/mptcp/Makefile +new file mode 100644 +index 0000000..e7238f5 +--- /dev/null ++++ b/net/mptcp/Makefile +@@ -0,0 +1,18 @@ ++# ++## Makefile for MultiPath TCP support code. ++# ++# ++ ++obj-$(CONFIG_MPTCP) += mptcp.o ++ ++mptcp-y := mptcp_ctrl.o mptcp_ipv4.o mptcp_ofo_queue.o mptcp_pm.o \ ++ mptcp_output.o mptcp_input.o ++ ++obj-$(CONFIG_TCP_CONG_COUPLED) += mptcp_coupled.o ++obj-$(CONFIG_TCP_CONG_OLIA) += mptcp_olia.o ++obj-$(CONFIG_TCP_CONG_WVEGAS) += mptcp_wvegas.o ++obj-$(CONFIG_MPTCP_FULLMESH) += mptcp_fullmesh.o ++obj-$(CONFIG_MPTCP_NDIFFPORTS) += mptcp_ndiffports.o ++ ++mptcp-$(subst m,y,$(CONFIG_IPV6)) += mptcp_ipv6.o ++ +diff --git a/net/mptcp/mptcp_coupled.c b/net/mptcp/mptcp_coupled.c +new file mode 100644 +index 0000000..d71f96e +--- /dev/null ++++ b/net/mptcp/mptcp_coupled.c +@@ -0,0 +1,273 @@ ++/* ++ * MPTCP implementation - Coupled Congestion Control ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++#include <net/tcp.h> ++#include <net/mptcp.h> ++ ++#include <linux/module.h> ++ ++/* Scaling is done in the numerator with alpha_scale_num and in the denominator ++ * with alpha_scale_den. ++ * ++ * To downscale, we just need to use alpha_scale. ++ * ++ * We have: alpha_scale = alpha_scale_num / (alpha_scale_den ^ 2) ++ */ ++static int alpha_scale_den = 10; ++static int alpha_scale_num = 32; ++static int alpha_scale = 12; ++ ++struct mptcp_ccc { ++ u64 alpha; ++ bool forced_update; ++}; ++ ++static inline int mptcp_ccc_sk_can_send(const struct sock *sk) ++{ ++ return mptcp_sk_can_send(sk) && tcp_sk(sk)->srtt; ++} ++ ++static inline u64 mptcp_get_alpha(struct sock *meta_sk) ++{ ++ struct mptcp_ccc *mptcp_ccc = inet_csk_ca(meta_sk); ++ return mptcp_ccc->alpha; ++} ++ ++static inline void mptcp_set_alpha(struct sock *meta_sk, u64 alpha) ++{ ++ struct mptcp_ccc *mptcp_ccc = inet_csk_ca(meta_sk); ++ mptcp_ccc->alpha = alpha; ++} ++ ++static inline u64 mptcp_ccc_scale(u32 val, int scale) ++{ ++ return (u64) val << scale; ++} ++ ++static inline bool mptcp_get_forced(struct sock *meta_sk) ++{ ++ struct mptcp_ccc *mptcp_ccc = inet_csk_ca(meta_sk); ++ return mptcp_ccc->forced_update; ++} ++ ++static inline void mptcp_set_forced(struct sock *meta_sk, bool force) ++{ ++ struct mptcp_ccc *mptcp_ccc = inet_csk_ca(meta_sk); ++ mptcp_ccc->forced_update = force; ++} ++ ++static void mptcp_ccc_recalc_alpha(struct sock *sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(sk)->mpcb; ++ struct sock *sub_sk; ++ int best_cwnd = 0, best_rtt = 0, can_send = 0; ++ u64 max_numerator = 0, sum_denominator = 0, alpha = 1; ++ ++ if (!mpcb) ++ return; ++ ++ /* Only one subflow left - fall back to normal reno-behavior ++ * (set alpha to 1) */ ++ if (mpcb->cnt_established <= 1) ++ goto exit; ++ ++ /* Do regular alpha-calculation for multiple subflows */ ++ ++ /* Find the max numerator of the alpha-calculation */ ++ mptcp_for_each_sk(mpcb, sub_sk) { ++ struct tcp_sock *sub_tp = tcp_sk(sub_sk); ++ u64 tmp; ++ ++ if (!mptcp_ccc_sk_can_send(sub_sk)) ++ continue; ++ ++ can_send++; ++ ++ /* We need to look for the path, that provides the max-value. ++ * Integer-overflow is not possible here, because ++ * tmp will be in u64. ++ */ ++ tmp = div64_u64(mptcp_ccc_scale(sub_tp->snd_cwnd, ++ alpha_scale_num), (u64)sub_tp->srtt * sub_tp->srtt); ++ ++ if (tmp >= max_numerator) { ++ max_numerator = tmp; ++ best_cwnd = sub_tp->snd_cwnd; ++ best_rtt = sub_tp->srtt; ++ } ++ } ++ ++ /* No subflow is able to send - we don't care anymore */ ++ if (unlikely(!can_send)) ++ goto exit; ++ ++ /* Calculate the denominator */ ++ mptcp_for_each_sk(mpcb, sub_sk) { ++ struct tcp_sock *sub_tp = tcp_sk(sub_sk); ++ ++ if (!mptcp_ccc_sk_can_send(sub_sk)) ++ continue; ++ ++ sum_denominator += div_u64( ++ mptcp_ccc_scale(sub_tp->snd_cwnd, ++ alpha_scale_den) * best_rtt, ++ sub_tp->srtt); ++ } ++ sum_denominator *= sum_denominator; ++ if (unlikely(!sum_denominator)) { ++ pr_err("%s: sum_denominator == 0, cnt_established:%d\n", ++ __func__, mpcb->cnt_established); ++ mptcp_for_each_sk(mpcb, sub_sk) { ++ struct tcp_sock *sub_tp = tcp_sk(sub_sk); ++ pr_err("%s: pi:%d, state:%d\n, rtt:%u, cwnd: %u", ++ __func__, sub_tp->mptcp->path_index, ++ sub_sk->sk_state, sub_tp->srtt, ++ sub_tp->snd_cwnd); ++ } ++ } ++ ++ alpha = div64_u64(mptcp_ccc_scale(best_cwnd, alpha_scale_num), sum_denominator); ++ ++ if (unlikely(!alpha)) ++ alpha = 1; ++ ++exit: ++ mptcp_set_alpha(mptcp_meta_sk(sk), alpha); ++} ++ ++static void mptcp_ccc_init(struct sock *sk) ++{ ++ if (tcp_sk(sk)->mpc) { ++ mptcp_set_forced(mptcp_meta_sk(sk), 0); ++ mptcp_set_alpha(mptcp_meta_sk(sk), 1); ++ } ++ /* If we do not mptcp, behave like reno: return */ ++} ++ ++static void mptcp_ccc_cwnd_event(struct sock *sk, enum tcp_ca_event event) ++{ ++ if (event == CA_EVENT_LOSS) ++ mptcp_ccc_recalc_alpha(sk); ++} ++ ++static void mptcp_ccc_set_state(struct sock *sk, u8 ca_state) ++{ ++ if (!tcp_sk(sk)->mpc) ++ return; ++ ++ mptcp_set_forced(mptcp_meta_sk(sk), 1); ++} ++ ++static void mptcp_ccc_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ int snd_cwnd; ++ ++ if (!tp->mpc) { ++ tcp_reno_cong_avoid(sk, ack, acked, in_flight); ++ return; ++ } ++ ++ if (!tcp_is_cwnd_limited(sk, in_flight)) ++ return; ++ ++ if (tp->snd_cwnd <= tp->snd_ssthresh) { ++ /* In "safe" area, increase. */ ++ tcp_slow_start(tp, acked); ++ mptcp_ccc_recalc_alpha(sk); ++ return; ++ } ++ ++ if (mptcp_get_forced(mptcp_meta_sk(sk))) { ++ mptcp_ccc_recalc_alpha(sk); ++ mptcp_set_forced(mptcp_meta_sk(sk), 0); ++ } ++ ++ if (mpcb->cnt_established > 1) { ++ u64 alpha = mptcp_get_alpha(mptcp_meta_sk(sk)); ++ ++ /* This may happen, if at the initialization, the mpcb ++ * was not yet attached to the sock, and thus ++ * initializing alpha failed. ++ */ ++ if (unlikely(!alpha)) ++ alpha = 1; ++ ++ snd_cwnd = (int) div_u64 ((u64) mptcp_ccc_scale(1, alpha_scale), ++ alpha); ++ ++ /* snd_cwnd_cnt >= max (scale * tot_cwnd / alpha, cwnd) ++ * Thus, we select here the max value. */ ++ if (snd_cwnd < tp->snd_cwnd) ++ snd_cwnd = tp->snd_cwnd; ++ } else { ++ snd_cwnd = tp->snd_cwnd; ++ } ++ ++ if (tp->snd_cwnd_cnt >= snd_cwnd) { ++ if (tp->snd_cwnd < tp->snd_cwnd_clamp) { ++ tp->snd_cwnd++; ++ mptcp_ccc_recalc_alpha(sk); ++ } ++ ++ tp->snd_cwnd_cnt = 0; ++ } else { ++ tp->snd_cwnd_cnt++; ++ } ++} ++ ++static struct tcp_congestion_ops mptcp_ccc = { ++ .init = mptcp_ccc_init, ++ .ssthresh = tcp_reno_ssthresh, ++ .cong_avoid = mptcp_ccc_cong_avoid, ++ .cwnd_event = mptcp_ccc_cwnd_event, ++ .set_state = mptcp_ccc_set_state, ++ .min_cwnd = tcp_reno_min_cwnd, ++ .owner = THIS_MODULE, ++ .name = "coupled", ++}; ++ ++static int __init mptcp_ccc_register(void) ++{ ++ BUILD_BUG_ON(sizeof(struct mptcp_ccc) > ICSK_CA_PRIV_SIZE); ++ return tcp_register_congestion_control(&mptcp_ccc); ++} ++ ++static void __exit mptcp_ccc_unregister(void) ++{ ++ tcp_unregister_congestion_control(&mptcp_ccc); ++} ++ ++module_init(mptcp_ccc_register); ++module_exit(mptcp_ccc_unregister); ++ ++MODULE_AUTHOR("Christoph Paasch, Sébastien Barré"); ++MODULE_LICENSE("GPL"); ++MODULE_DESCRIPTION("MPTCP COUPLED CONGESTION CONTROL"); ++MODULE_VERSION("0.1"); +diff --git a/net/mptcp/mptcp_ctrl.c b/net/mptcp/mptcp_ctrl.c +new file mode 100644 +index 0000000..6a7654d +--- /dev/null ++++ b/net/mptcp/mptcp_ctrl.c +@@ -0,0 +1,2270 @@ ++/* ++ * MPTCP implementation - MPTCP-control ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#include <net/inet_common.h> ++#include <net/inet6_hashtables.h> ++#include <net/ipv6.h> ++#include <net/ip6_checksum.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#if IS_ENABLED(CONFIG_IPV6) ++#include <net/mptcp_v6.h> ++#endif ++#include <net/sock.h> ++#include <net/tcp.h> ++#include <net/tcp_states.h> ++#include <net/transp_v6.h> ++#include <net/xfrm.h> ++ ++#include <linux/cryptohash.h> ++#include <linux/kconfig.h> ++#include <linux/module.h> ++#include <linux/netpoll.h> ++#include <linux/list.h> ++#include <linux/jhash.h> ++#include <linux/tcp.h> ++#include <linux/net.h> ++#include <linux/in.h> ++#include <linux/random.h> ++#include <linux/inetdevice.h> ++#include <linux/workqueue.h> ++#include <linux/atomic.h> ++#include <linux/sysctl.h> ++ ++static struct kmem_cache *mptcp_sock_cache __read_mostly; ++static struct kmem_cache *mptcp_cb_cache __read_mostly; ++static struct kmem_cache *mptcp_tw_cache __read_mostly; ++ ++int sysctl_mptcp_enabled __read_mostly = 1; ++int sysctl_mptcp_checksum __read_mostly = 1; ++int sysctl_mptcp_debug __read_mostly; ++EXPORT_SYMBOL(sysctl_mptcp_debug); ++int sysctl_mptcp_syn_retries __read_mostly = 3; ++ ++bool mptcp_init_failed __read_mostly; ++ ++static int proc_mptcp_path_manager(ctl_table *ctl, int write, ++ void __user *buffer, size_t *lenp, ++ loff_t *ppos) ++{ ++ char val[MPTCP_PM_NAME_MAX]; ++ ctl_table tbl = { ++ .data = val, ++ .maxlen = MPTCP_PM_NAME_MAX, ++ }; ++ int ret; ++ ++ mptcp_get_default_path_manager(val); ++ ++ ret = proc_dostring(&tbl, write, buffer, lenp, ppos); ++ if (write && ret == 0) ++ ret = mptcp_set_default_path_manager(val); ++ return ret; ++} ++ ++static struct ctl_table mptcp_table[] = { ++ { ++ .procname = "mptcp_enabled", ++ .data = &sysctl_mptcp_enabled, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = &proc_dointvec ++ }, ++ { ++ .procname = "mptcp_checksum", ++ .data = &sysctl_mptcp_checksum, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = &proc_dointvec ++ }, ++ { ++ .procname = "mptcp_debug", ++ .data = &sysctl_mptcp_debug, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = &proc_dointvec ++ }, ++ { ++ .procname = "mptcp_syn_retries", ++ .data = &sysctl_mptcp_syn_retries, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = &proc_dointvec ++ }, ++ { ++ .procname = "mptcp_path_manager", ++ .mode = 0644, ++ .maxlen = MPTCP_PM_NAME_MAX, ++ .proc_handler = proc_mptcp_path_manager, ++ }, ++ { } ++}; ++ ++static inline u32 mptcp_hash_tk(u32 token) ++{ ++ return token % MPTCP_HASH_SIZE; ++} ++ ++struct hlist_nulls_head tk_hashtable[MPTCP_HASH_SIZE]; ++EXPORT_SYMBOL(tk_hashtable); ++ ++/* This second hashtable is needed to retrieve request socks ++ * created as a result of a join request. While the SYN contains ++ * the token, the final ack does not, so we need a separate hashtable ++ * to retrieve the mpcb. ++ */ ++struct list_head mptcp_reqsk_htb[MPTCP_HASH_SIZE]; ++spinlock_t mptcp_reqsk_hlock; /* hashtable protection */ ++ ++/* The following hash table is used to avoid collision of token */ ++static struct hlist_nulls_head mptcp_reqsk_tk_htb[MPTCP_HASH_SIZE]; ++spinlock_t mptcp_tk_hashlock; /* hashtable protection */ ++ ++static int mptcp_reqsk_find_tk(u32 token) ++{ ++ u32 hash = mptcp_hash_tk(token); ++ struct mptcp_request_sock *mtreqsk; ++ const struct hlist_nulls_node *node; ++ ++ hlist_nulls_for_each_entry_rcu(mtreqsk, node, ++ &mptcp_reqsk_tk_htb[hash], collide_tk) { ++ if (token == mtreqsk->mptcp_loc_token) ++ return 1; ++ } ++ return 0; ++} ++ ++static void mptcp_reqsk_insert_tk(struct request_sock *reqsk, u32 token) ++{ ++ u32 hash = mptcp_hash_tk(token); ++ ++ hlist_nulls_add_head_rcu(&mptcp_rsk(reqsk)->collide_tk, ++ &mptcp_reqsk_tk_htb[hash]); ++} ++ ++static void mptcp_reqsk_remove_tk(struct request_sock *reqsk) ++{ ++ rcu_read_lock(); ++ spin_lock(&mptcp_tk_hashlock); ++ hlist_nulls_del_init_rcu(&mptcp_rsk(reqsk)->collide_tk); ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock(); ++} ++ ++void mptcp_reqsk_destructor(struct request_sock *req) ++{ ++ if (!mptcp_rsk(req)->mpcb) { ++ if (in_softirq()) { ++ mptcp_reqsk_remove_tk(req); ++ } else { ++ rcu_read_lock_bh(); ++ spin_lock(&mptcp_tk_hashlock); ++ hlist_nulls_del_init_rcu(&mptcp_rsk(req)->collide_tk); ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock_bh(); ++ } ++ } else { ++ mptcp_hash_request_remove(req); ++ } ++} ++ ++static void __mptcp_hash_insert(struct tcp_sock *meta_tp, u32 token) ++{ ++ u32 hash = mptcp_hash_tk(token); ++ hlist_nulls_add_head_rcu(&meta_tp->tk_table, &tk_hashtable[hash]); ++ meta_tp->inside_tk_table = 1; ++} ++ ++static int mptcp_find_token(u32 token) ++{ ++ u32 hash = mptcp_hash_tk(token); ++ struct tcp_sock *meta_tp; ++ const struct hlist_nulls_node *node; ++ ++ hlist_nulls_for_each_entry_rcu(meta_tp, node, &tk_hashtable[hash], tk_table) { ++ if (token == meta_tp->mptcp_loc_token) ++ return 1; ++ } ++ return 0; ++} ++ ++static void mptcp_set_key_reqsk(struct request_sock *req, ++ const struct sk_buff *skb) ++{ ++ struct inet_request_sock *ireq = inet_rsk(req); ++ struct mptcp_request_sock *mtreq = mptcp_rsk(req); ++ ++ if (skb->protocol == htons(ETH_P_IP)) { ++ mtreq->mptcp_loc_key = mptcp_v4_get_key(ip_hdr(skb)->saddr, ++ ip_hdr(skb)->daddr, ++ htons(ireq->ir_num), ++ ireq->ir_rmt_port); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else { ++ mtreq->mptcp_loc_key = mptcp_v6_get_key(ipv6_hdr(skb)->saddr.s6_addr32, ++ ipv6_hdr(skb)->daddr.s6_addr32, ++ htons(ireq->ir_num), ++ ireq->ir_rmt_port); ++#endif ++ } ++ ++ mptcp_key_sha1(mtreq->mptcp_loc_key, &mtreq->mptcp_loc_token, NULL); ++} ++ ++/* New MPTCP-connection request, prepare a new token for the meta-socket that ++ * will be created in mptcp_check_req_master(), and store the received token. ++ */ ++void mptcp_reqsk_new_mptcp(struct request_sock *req, ++ const struct tcp_options_received *rx_opt, ++ const struct mptcp_options_received *mopt, ++ const struct sk_buff *skb) ++{ ++ struct mptcp_request_sock *mtreq = mptcp_rsk(req); ++ ++ tcp_rsk(req)->saw_mpc = 1; ++ ++ rcu_read_lock(); ++ spin_lock(&mptcp_tk_hashlock); ++ do { ++ mptcp_set_key_reqsk(req, skb); ++ } while (mptcp_reqsk_find_tk(mtreq->mptcp_loc_token) || ++ mptcp_find_token(mtreq->mptcp_loc_token)); ++ ++ mptcp_reqsk_insert_tk(req, mtreq->mptcp_loc_token); ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock(); ++ mtreq->mptcp_rem_key = mopt->mptcp_key; ++} ++ ++static void mptcp_set_key_sk(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct inet_sock *isk = inet_sk(sk); ++ ++ if (sk->sk_family == AF_INET) ++ tp->mptcp_loc_key = mptcp_v4_get_key(isk->inet_saddr, ++ isk->inet_daddr, ++ isk->inet_sport, ++ isk->inet_dport); ++#if IS_ENABLED(CONFIG_IPV6) ++ else ++ tp->mptcp_loc_key = mptcp_v6_get_key(inet6_sk(sk)->saddr.s6_addr32, ++ sk->sk_v6_daddr.s6_addr32, ++ isk->inet_sport, ++ isk->inet_dport); ++#endif ++ ++ mptcp_key_sha1(tp->mptcp_loc_key, ++ &tp->mptcp_loc_token, NULL); ++} ++ ++void mptcp_connect_init(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ rcu_read_lock_bh(); ++ spin_lock(&mptcp_tk_hashlock); ++ do { ++ mptcp_set_key_sk(sk); ++ } while (mptcp_reqsk_find_tk(tp->mptcp_loc_token) || ++ mptcp_find_token(tp->mptcp_loc_token)); ++ ++ __mptcp_hash_insert(tp, tp->mptcp_loc_token); ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock_bh(); ++} ++ ++/** ++ * This function increments the refcount of the mpcb struct. ++ * It is the responsibility of the caller to decrement when releasing ++ * the structure. ++ */ ++struct sock *mptcp_hash_find(struct net *net, u32 token) ++{ ++ u32 hash = mptcp_hash_tk(token); ++ struct tcp_sock *meta_tp; ++ struct sock *meta_sk = NULL; ++ struct hlist_nulls_node *node; ++ ++ rcu_read_lock(); ++ hlist_nulls_for_each_entry_rcu(meta_tp, node, &tk_hashtable[hash], ++ tk_table) { ++ meta_sk = (struct sock *)meta_tp; ++ if (token == meta_tp->mptcp_loc_token && ++ net_eq(net, sock_net(meta_sk)) && ++ atomic_inc_not_zero(&meta_sk->sk_refcnt)) ++ break; ++ meta_sk = NULL; ++ } ++ rcu_read_unlock(); ++ return meta_sk; ++} ++ ++void mptcp_hash_remove_bh(struct tcp_sock *meta_tp) ++{ ++ /* remove from the token hashtable */ ++ rcu_read_lock_bh(); ++ spin_lock(&mptcp_tk_hashlock); ++ hlist_nulls_del_init_rcu(&meta_tp->tk_table); ++ meta_tp->inside_tk_table = 0; ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock_bh(); ++} ++ ++void mptcp_hash_remove(struct tcp_sock *meta_tp) ++{ ++ rcu_read_lock(); ++ spin_lock(&mptcp_tk_hashlock); ++ hlist_nulls_del_init_rcu(&meta_tp->tk_table); ++ meta_tp->inside_tk_table = 0; ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock(); ++} ++ ++static struct sock *mptcp_syn_recv_sock(struct sock *sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst) ++{ ++#if IS_ENABLED(CONFIG_IPV6) ++ if (sk->sk_family == AF_INET6) ++ return tcp_v6_syn_recv_sock(sk, skb, req, dst); ++ ++ /* sk->sk_family == AF_INET */ ++ if (req->rsk_ops->family == AF_INET6) ++ return mptcp_v6v4_syn_recv_sock(sk, skb, req, dst); ++#endif ++ ++ /* sk->sk_family == AF_INET && req->rsk_ops->family == AF_INET */ ++ return tcp_v4_syn_recv_sock(sk, skb, req, dst); ++} ++ ++struct sock *mptcp_select_ack_sock(const struct sock *meta_sk, int copied) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sock *sk, *subsk = NULL; ++ u32 max_data_seq = 0; ++ /* max_data_seq initialized to correct compiler-warning. ++ * But the initialization is handled by max_data_seq_set ++ */ ++ short max_data_seq_set = 0; ++ u32 min_time = 0xffffffff; ++ ++ /* How do we select the subflow to send the window-update on? ++ * ++ * 1. He has to be in a state where he can send an ack and is ++ * operational (pf = 0). ++ * 2. He has to be one of those subflow who recently ++ * contributed to the received stream ++ * (this guarantees a working subflow) ++ * a) its latest data_seq received is after the original ++ * copied_seq. ++ * We select the one with the lowest rtt, so that the ++ * window-update reaches our peer the fastest. ++ * b) if no subflow has this kind of data_seq (e.g., very ++ * strange meta-level retransmissions going on), we take ++ * the subflow who last sent the highest data_seq. ++ */ ++ mptcp_for_each_sk(meta_tp->mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ if (!mptcp_sk_can_send_ack(sk) || tp->pf) ++ continue; ++ ++ /* Select among those who contributed to the ++ * current receive-queue. ++ */ ++ if (copied && after(tp->mptcp->last_data_seq, meta_tp->copied_seq - copied)) { ++ if (tp->srtt < min_time) { ++ min_time = tp->srtt; ++ subsk = sk; ++ max_data_seq_set = 0; ++ } ++ continue; ++ } ++ ++ if (!subsk && !max_data_seq_set) { ++ max_data_seq = tp->mptcp->last_data_seq; ++ max_data_seq_set = 1; ++ subsk = sk; ++ } ++ ++ /* Otherwise, take the one with the highest data_seq */ ++ if ((!subsk || max_data_seq_set) && ++ after(tp->mptcp->last_data_seq, max_data_seq)) { ++ max_data_seq = tp->mptcp->last_data_seq; ++ subsk = sk; ++ } ++ } ++ ++ if (!subsk) { ++ mptcp_debug("%s subsk is null, copied %d, cseq %u\n", __func__, ++ copied, meta_tp->copied_seq); ++ mptcp_for_each_sk(meta_tp->mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ mptcp_debug("%s pi %d state %u last_dseq %u\n", ++ __func__, tp->mptcp->path_index, sk->sk_state, ++ tp->mptcp->last_data_seq); ++ } ++ } ++ ++ return subsk; ++} ++EXPORT_SYMBOL(mptcp_select_ack_sock); ++ ++static void mptcp_sock_def_error_report(struct sock *sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(sk)->mpcb; ++ ++ if (!sock_flag(sk, SOCK_DEAD)) ++ mptcp_sub_close(sk, 0); ++ ++ if (mpcb->infinite_mapping_rcv || mpcb->infinite_mapping_snd || ++ mpcb->send_infinite_mapping) { ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ ++ meta_sk->sk_err = sk->sk_err; ++ meta_sk->sk_err_soft = sk->sk_err_soft; ++ ++ if (!sock_flag(meta_sk, SOCK_DEAD)) ++ meta_sk->sk_error_report(meta_sk); ++ ++ tcp_done(meta_sk); ++ } ++ ++ sk->sk_err = 0; ++ return; ++} ++ ++static void mptcp_mpcb_put(struct mptcp_cb *mpcb) ++{ ++ if (atomic_dec_and_test(&mpcb->mpcb_refcnt)) { ++ mptcp_cleanup_path_manager(mpcb); ++ kmem_cache_free(mptcp_cb_cache, mpcb); ++ } ++} ++ ++static void mptcp_sock_destruct(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ inet_sock_destruct(sk); ++ ++ BUG_ON(!list_empty(&tp->mptcp->cb_list)); ++ ++ kmem_cache_free(mptcp_sock_cache, tp->mptcp); ++ tp->mptcp = NULL; ++ ++ if (!is_meta_sk(sk) && !tp->was_meta_sk) { ++ /* Taken when mpcb pointer was set */ ++ sock_put(mptcp_meta_sk(sk)); ++ mptcp_mpcb_put(tp->mpcb); ++ } else { ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct mptcp_tw *mptw; ++ ++ /* The mpcb is disappearing - we can make the final ++ * update to the rcv_nxt of the time-wait-sock and remove ++ * its reference to the mpcb. ++ */ ++ spin_lock_bh(&mpcb->tw_lock); ++ list_for_each_entry_rcu(mptw, &mpcb->tw_list, list) { ++ list_del_rcu(&mptw->list); ++ mptw->in_list = 0; ++ mptcp_mpcb_put(mpcb); ++ rcu_assign_pointer(mptw->mpcb, NULL); ++ } ++ spin_unlock_bh(&mpcb->tw_lock); ++ ++ mptcp_mpcb_put(mpcb); ++ ++ mptcp_debug("%s destroying meta-sk\n", __func__); ++ } ++} ++ ++void mptcp_destroy_sock(struct sock *sk) ++{ ++ if (is_meta_sk(sk)) { ++ struct sock *sk_it, *tmpsk; ++ ++ __skb_queue_purge(&tcp_sk(sk)->mpcb->reinject_queue); ++ mptcp_purge_ofo_queue(tcp_sk(sk)); ++ ++ /* We have to close all remaining subflows. Normally, they ++ * should all be about to get closed. But, if the kernel is ++ * forcing a closure (e.g., tcp_write_err), the subflows might ++ * not have been closed properly (as we are waiting for the ++ * DATA_ACK of the DATA_FIN). ++ */ ++ mptcp_for_each_sk_safe(tcp_sk(sk)->mpcb, sk_it, tmpsk) { ++ /* Already did call tcp_close - waiting for graceful ++ * closure, or if we are retransmitting fast-close on ++ * the subflow. The reset (or timeout) will kill the ++ * subflow.. ++ */ ++ if (tcp_sk(sk_it)->closing || ++ tcp_sk(sk_it)->send_mp_fclose) ++ continue; ++ ++ /* Allow the delayed work first to prevent time-wait state */ ++ if (delayed_work_pending(&tcp_sk(sk_it)->mptcp->work)) ++ continue; ++ ++ mptcp_sub_close(sk_it, 0); ++ } ++ } else { ++ mptcp_del_sock(sk); ++ } ++} ++ ++static void mptcp_set_state(struct sock *sk) ++{ ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ ++ /* Meta is not yet established - wake up the application */ ++ if ((1 << meta_sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV) && ++ sk->sk_state == TCP_ESTABLISHED) { ++ tcp_set_state(meta_sk, TCP_ESTABLISHED); ++ ++ if (!sock_flag(meta_sk, SOCK_DEAD)) { ++ meta_sk->sk_state_change(meta_sk); ++ sk_wake_async(meta_sk, SOCK_WAKE_IO, POLL_OUT); ++ } ++ } ++ ++ if (sk->sk_state == TCP_ESTABLISHED) { ++ tcp_sk(sk)->mptcp->establish_increased = 1; ++ tcp_sk(sk)->mpcb->cnt_established++; ++ } ++} ++ ++u32 mptcp_secret[MD5_MESSAGE_BYTES / 4] ____cacheline_aligned; ++u32 mptcp_key_seed = 0; ++ ++void mptcp_key_sha1(u64 key, u32 *token, u64 *idsn) ++{ ++ u32 workspace[SHA_WORKSPACE_WORDS]; ++ u32 mptcp_hashed_key[SHA_DIGEST_WORDS]; ++ u8 input[64]; ++ int i; ++ ++ memset(workspace, 0, sizeof(workspace)); ++ ++ /* Initialize input with appropriate padding */ ++ memset(&input[9], 0, sizeof(input) - 10); /* -10, because the last byte ++ * is explicitly set too */ ++ memcpy(input, &key, sizeof(key)); /* Copy key to the msg beginning */ ++ input[8] = 0x80; /* Padding: First bit after message = 1 */ ++ input[63] = 0x40; /* Padding: Length of the message = 64 bits */ ++ ++ sha_init(mptcp_hashed_key); ++ sha_transform(mptcp_hashed_key, input, workspace); ++ ++ for (i = 0; i < 5; i++) ++ mptcp_hashed_key[i] = cpu_to_be32(mptcp_hashed_key[i]); ++ ++ if (token) ++ *token = mptcp_hashed_key[0]; ++ if (idsn) ++ *idsn = *((u64 *)&mptcp_hashed_key[3]); ++} ++ ++void mptcp_hmac_sha1(u8 *key_1, u8 *key_2, u8 *rand_1, u8 *rand_2, ++ u32 *hash_out) ++{ ++ u32 workspace[SHA_WORKSPACE_WORDS]; ++ u8 input[128]; /* 2 512-bit blocks */ ++ int i; ++ ++ memset(workspace, 0, sizeof(workspace)); ++ ++ /* Generate key xored with ipad */ ++ memset(input, 0x36, 64); ++ for (i = 0; i < 8; i++) ++ input[i] ^= key_1[i]; ++ for (i = 0; i < 8; i++) ++ input[i + 8] ^= key_2[i]; ++ ++ memcpy(&input[64], rand_1, 4); ++ memcpy(&input[68], rand_2, 4); ++ input[72] = 0x80; /* Padding: First bit after message = 1 */ ++ memset(&input[73], 0, 53); ++ ++ /* Padding: Length of the message = 512 + 64 bits */ ++ input[126] = 0x02; ++ input[127] = 0x40; ++ ++ sha_init(hash_out); ++ sha_transform(hash_out, input, workspace); ++ memset(workspace, 0, sizeof(workspace)); ++ ++ sha_transform(hash_out, &input[64], workspace); ++ memset(workspace, 0, sizeof(workspace)); ++ ++ for (i = 0; i < 5; i++) ++ hash_out[i] = cpu_to_be32(hash_out[i]); ++ ++ /* Prepare second part of hmac */ ++ memset(input, 0x5C, 64); ++ for (i = 0; i < 8; i++) ++ input[i] ^= key_1[i]; ++ for (i = 0; i < 8; i++) ++ input[i + 8] ^= key_2[i]; ++ ++ memcpy(&input[64], hash_out, 20); ++ input[84] = 0x80; ++ memset(&input[85], 0, 41); ++ ++ /* Padding: Length of the message = 512 + 160 bits */ ++ input[126] = 0x02; ++ input[127] = 0xA0; ++ ++ sha_init(hash_out); ++ sha_transform(hash_out, input, workspace); ++ memset(workspace, 0, sizeof(workspace)); ++ ++ sha_transform(hash_out, &input[64], workspace); ++ ++ for (i = 0; i < 5; i++) ++ hash_out[i] = cpu_to_be32(hash_out[i]); ++} ++ ++static void mptcp_mpcb_inherit_sockopts(struct sock *meta_sk, struct sock *master_sk) ++{ ++ /* Socket-options handled by mptcp_inherit_sk while creating the meta-sk. ++ * ====== ++ * SO_SNDBUF, SO_SNDBUFFORCE, SO_RCVBUF, SO_RCVBUFFORCE, SO_RCVLOWAT, ++ * SO_RCVTIMEO, SO_SNDTIMEO, SO_ATTACH_FILTER, SO_DETACH_FILTER, ++ * TCP_NODELAY, TCP_CORK ++ * ++ * Socket-options handled in this function here ++ * ====== ++ * TCP_DEFER_ACCEPT ++ * ++ * Socket-options on the todo-list ++ * ====== ++ * SO_BINDTODEVICE - should probably prevent creation of new subsocks ++ * across other devices. - what about the api-draft? ++ * SO_DEBUG ++ * SO_REUSEADDR - probably we don't care about this ++ * SO_DONTROUTE, SO_BROADCAST ++ * SO_OOBINLINE ++ * SO_LINGER ++ * SO_TIMESTAMP* - I don't think this is of concern for a SOCK_STREAM ++ * SO_PASSSEC - I don't think this is of concern for a SOCK_STREAM ++ * SO_RXQ_OVFL ++ * TCP_COOKIE_TRANSACTIONS ++ * TCP_MAXSEG ++ * TCP_THIN_* - Handled by mptcp_inherit_sk, but we need to support this ++ * in mptcp_retransmit_timer. AND we need to check what is ++ * about the subsockets. ++ * TCP_LINGER2 ++ * TCP_WINDOW_CLAMP ++ * TCP_USER_TIMEOUT ++ * TCP_MD5SIG ++ * ++ * Socket-options of no concern for the meta-socket (but for the subsocket) ++ * ====== ++ * SO_PRIORITY ++ * SO_MARK ++ * TCP_CONGESTION ++ * TCP_SYNCNT ++ * TCP_QUICKACK ++ * SO_KEEPALIVE ++ */ ++ ++ /****** DEFER_ACCEPT-handler ******/ ++ ++ /* DEFER_ACCEPT is not of concern for new subflows - we always accept ++ * them ++ */ ++ inet_csk(meta_sk)->icsk_accept_queue.rskq_defer_accept = 0; ++} ++ ++static void mptcp_sub_inherit_sockopts(struct sock *meta_sk, struct sock *sub_sk) ++{ ++ /* IP_TOS also goes to the subflow. */ ++ if (inet_sk(sub_sk)->tos != inet_sk(meta_sk)->tos) { ++ inet_sk(sub_sk)->tos = inet_sk(meta_sk)->tos; ++ sub_sk->sk_priority = meta_sk->sk_priority; ++ sk_dst_reset(sub_sk); ++ } ++ ++ /* Inherit SO_REUSEADDR */ ++ sub_sk->sk_reuse = meta_sk->sk_reuse; ++ ++ /* Inherit snd/rcv-buffer locks */ ++ sub_sk->sk_userlocks = meta_sk->sk_userlocks & ~SOCK_BINDPORT_LOCK; ++} ++ ++int mptcp_backlog_rcv(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ /* skb-sk may be NULL if we receive a packet immediatly after the ++ * SYN/ACK + MP_CAPABLE. ++ */ ++ struct sock *sk = skb->sk ? skb->sk : meta_sk; ++ int ret = 0; ++ ++ skb->sk = NULL; ++ ++ if (unlikely(!atomic_inc_not_zero(&sk->sk_refcnt))) { ++ kfree_skb(skb); ++ return 0; ++ } ++ ++ if (sk->sk_family == AF_INET) ++ ret = tcp_v4_do_rcv(sk, skb); ++#if IS_ENABLED(CONFIG_IPV6) ++ else ++ ret = tcp_v6_do_rcv(sk, skb); ++#endif ++ ++ sock_put(sk); ++ return ret; ++} ++ ++struct lock_class_key meta_key; ++struct lock_class_key meta_slock_key; ++ ++/* Code heavily inspired from sk_clone() */ ++static int mptcp_inherit_sk(const struct sock *sk, struct sock *newsk, ++ int family, const gfp_t flags) ++{ ++ struct sk_filter *filter; ++ struct proto *prot = newsk->sk_prot; ++ const struct inet_connection_sock_af_ops *af_ops = inet_csk(newsk)->icsk_af_ops; ++#ifdef CONFIG_SECURITY_NETWORK ++ void *sptr = newsk->sk_security; ++#endif ++ ++ if (sk->sk_family == AF_INET) { ++ memcpy(newsk, sk, offsetof(struct sock, sk_dontcopy_begin)); ++ memcpy(&newsk->sk_dontcopy_end, &sk->sk_dontcopy_end, ++ sizeof(struct tcp_sock) - offsetof(struct sock, sk_dontcopy_end)); ++ } else { ++ memcpy(newsk, sk, offsetof(struct sock, sk_dontcopy_begin)); ++ memcpy(&newsk->sk_dontcopy_end, &sk->sk_dontcopy_end, ++ sizeof(struct tcp6_sock) - offsetof(struct sock, sk_dontcopy_end)); ++ } ++ ++#ifdef CONFIG_SECURITY_NETWORK ++ newsk->sk_security = sptr; ++ security_sk_clone(sk, newsk); ++#endif ++ ++ /* Has been changed by sock_copy above - we may need an IPv6-socket */ ++ newsk->sk_family = family; ++ newsk->sk_prot = prot; ++ newsk->sk_prot_creator = prot; ++ inet_csk(newsk)->icsk_af_ops = af_ops; ++ ++ /* We don't yet have the mptcp-point. Thus we still need inet_sock_destruct */ ++ newsk->sk_destruct = inet_sock_destruct; ++ ++ /* SANITY */ ++ get_net(sock_net(newsk)); ++ sk_node_init(&newsk->sk_node); ++ sock_lock_init_class_and_name(newsk, "slock-AF_INET-MPTCP", ++ &meta_slock_key, "sk_lock-AF_INET-MPTCP", ++ &meta_key); ++ ++ /* Unlocks are in: ++ * ++ * 1. If we are creating the master-sk ++ * * on client-side in tcp_rcv_state_process, "case TCP_SYN_SENT" ++ * * on server-side in tcp_child_process ++ * 2. If we are creating another subsock ++ * * Also in tcp_child_process ++ */ ++ bh_lock_sock(newsk); ++ newsk->sk_backlog.head = NULL; ++ newsk->sk_backlog.tail = NULL; ++ newsk->sk_backlog.len = 0; ++ ++ atomic_set(&newsk->sk_rmem_alloc, 0); ++ atomic_set(&newsk->sk_wmem_alloc, 1); ++ atomic_set(&newsk->sk_omem_alloc, 0); ++ ++ skb_queue_head_init(&newsk->sk_receive_queue); ++ skb_queue_head_init(&newsk->sk_write_queue); ++#ifdef CONFIG_NET_DMA ++ skb_queue_head_init(&newsk->sk_async_wait_queue); ++#endif ++ ++ spin_lock_init(&newsk->sk_dst_lock); ++ rwlock_init(&newsk->sk_callback_lock); ++ lockdep_set_class_and_name(&newsk->sk_callback_lock, ++ af_callback_keys + newsk->sk_family, ++ af_family_clock_key_strings[newsk->sk_family]); ++ newsk->sk_dst_cache = NULL; ++ newsk->sk_rx_dst = NULL; ++ newsk->sk_wmem_queued = 0; ++ newsk->sk_forward_alloc = 0; ++ newsk->sk_send_head = NULL; ++ newsk->sk_userlocks = sk->sk_userlocks & ~SOCK_BINDPORT_LOCK; ++ ++ tcp_sk(newsk)->mptcp = NULL; ++ ++ sock_reset_flag(newsk, SOCK_DONE); ++ skb_queue_head_init(&newsk->sk_error_queue); ++ ++ filter = rcu_dereference_protected(newsk->sk_filter, 1); ++ if (filter != NULL) ++ sk_filter_charge(newsk, filter); ++ ++ if (unlikely(xfrm_sk_clone_policy(newsk))) { ++ /* It is still raw copy of parent, so invalidate ++ * destructor and make plain sk_free() ++ */ ++ newsk->sk_destruct = NULL; ++ bh_unlock_sock(newsk); ++ sk_free(newsk); ++ newsk = NULL; ++ return -ENOMEM; ++ } ++ ++ newsk->sk_err = 0; ++ newsk->sk_priority = 0; ++ /* Before updating sk_refcnt, we must commit prior changes to memory ++ * (Documentation/RCU/rculist_nulls.txt for details) ++ */ ++ smp_wmb(); ++ atomic_set(&newsk->sk_refcnt, 2); ++ ++ /* Increment the counter in the same struct proto as the master ++ * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that ++ * is the same as sk->sk_prot->socks, as this field was copied ++ * with memcpy). ++ * ++ * This _changes_ the previous behaviour, where ++ * tcp_create_openreq_child always was incrementing the ++ * equivalent to tcp_prot->socks (inet_sock_nr), so this have ++ * to be taken into account in all callers. -acme ++ */ ++ sk_refcnt_debug_inc(newsk); ++ sk_set_socket(newsk, NULL); ++ newsk->sk_wq = NULL; ++ ++ if (newsk->sk_prot->sockets_allocated) ++ percpu_counter_inc(newsk->sk_prot->sockets_allocated); ++ ++ if (sock_flag(newsk, SOCK_TIMESTAMP) || ++ sock_flag(newsk, SOCK_TIMESTAMPING_RX_SOFTWARE)) ++ net_enable_timestamp(); ++ ++ return 0; ++} ++ ++int mptcp_alloc_mpcb(struct sock *meta_sk, __u64 remote_key, u32 window) ++{ ++ struct mptcp_cb *mpcb; ++ struct sock *master_sk; ++ struct inet_connection_sock *master_icsk, *meta_icsk = inet_csk(meta_sk); ++ struct tcp_sock *master_tp, *meta_tp = tcp_sk(meta_sk); ++ struct sk_buff *skb, *tmp; ++ u64 idsn; ++ ++ master_sk = sk_prot_alloc(meta_sk->sk_prot, GFP_ATOMIC | __GFP_ZERO, ++ meta_sk->sk_family); ++ if (!master_sk) ++ return -ENOBUFS; ++ ++ master_tp = tcp_sk(master_sk); ++ master_icsk = inet_csk(master_sk); ++ ++ /* Need to set this here - it is needed by mptcp_inherit_sk */ ++ master_sk->sk_prot = meta_sk->sk_prot; ++ master_sk->sk_prot_creator = meta_sk->sk_prot; ++ master_icsk->icsk_af_ops = meta_icsk->icsk_af_ops; ++ ++ mpcb = kmem_cache_zalloc(mptcp_cb_cache, GFP_ATOMIC); ++ if (!mpcb) { ++ sk_free(master_sk); ++ return -ENOBUFS; ++ } ++ ++ /* master_sk inherits from meta_sk */ ++ if (mptcp_inherit_sk(meta_sk, master_sk, meta_sk->sk_family, GFP_ATOMIC)) { ++ kmem_cache_free(mptcp_cb_cache, mpcb); ++ return -ENOBUFS; ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ if (meta_icsk->icsk_af_ops == &ipv6_mapped) { ++ struct ipv6_pinfo *newnp, *np = inet6_sk(meta_sk); ++ ++ inet_sk(master_sk)->pinet6 = &((struct tcp6_sock *)master_sk)->inet6; ++ ++ newnp = inet6_sk(master_sk); ++ memcpy(newnp, np, sizeof(struct ipv6_pinfo)); ++ ++ newnp->ipv6_mc_list = NULL; ++ newnp->ipv6_ac_list = NULL; ++ newnp->ipv6_fl_list = NULL; ++ newnp->opt = NULL; ++ newnp->pktoptions = NULL; ++ (void)xchg(&newnp->rxpmtu, NULL); ++ } else if (meta_sk->sk_family == AF_INET6) { ++ struct ipv6_pinfo *newnp, *np = inet6_sk(meta_sk); ++ ++ inet_sk(master_sk)->pinet6 = &((struct tcp6_sock *)master_sk)->inet6; ++ ++ newnp = inet6_sk(master_sk); ++ memcpy(newnp, np, sizeof(struct ipv6_pinfo)); ++ ++ newnp->hop_limit = -1; ++ newnp->mcast_hops = IPV6_DEFAULT_MCASTHOPS; ++ newnp->mc_loop = 1; ++ newnp->pmtudisc = IPV6_PMTUDISC_WANT; ++ newnp->ipv6only = sock_net(master_sk)->ipv6.sysctl.bindv6only; ++ } ++#endif ++ ++ meta_tp->mptcp = kmem_cache_zalloc(mptcp_sock_cache, GFP_ATOMIC); ++ if (!meta_tp->mptcp) { ++ kmem_cache_free(mptcp_cb_cache, mpcb); ++ sk_free(master_sk); ++ return -ENOBUFS; ++ } ++ ++ INIT_LIST_HEAD(&meta_tp->mptcp->cb_list); ++ ++ /* Store the keys and generate the peer's token */ ++ mpcb->mptcp_loc_key = meta_tp->mptcp_loc_key; ++ mpcb->mptcp_loc_token = meta_tp->mptcp_loc_token; ++ ++ /* Generate Initial data-sequence-numbers */ ++ mptcp_key_sha1(mpcb->mptcp_loc_key, NULL, &idsn); ++ idsn = ntohll(idsn) + 1; ++ mpcb->snd_high_order[0] = idsn >> 32; ++ mpcb->snd_high_order[1] = mpcb->snd_high_order[0] - 1; ++ ++ meta_tp->write_seq = (u32)idsn; ++ meta_tp->snd_sml = meta_tp->write_seq; ++ meta_tp->snd_una = meta_tp->write_seq; ++ meta_tp->snd_nxt = meta_tp->write_seq; ++ meta_tp->pushed_seq = meta_tp->write_seq; ++ meta_tp->snd_up = meta_tp->write_seq; ++ ++ mpcb->mptcp_rem_key = remote_key; ++ mptcp_key_sha1(mpcb->mptcp_rem_key, &mpcb->mptcp_rem_token, &idsn); ++ idsn = ntohll(idsn) + 1; ++ mpcb->rcv_high_order[0] = idsn >> 32; ++ mpcb->rcv_high_order[1] = mpcb->rcv_high_order[0] + 1; ++ meta_tp->copied_seq = (u32) idsn; ++ meta_tp->rcv_nxt = (u32) idsn; ++ meta_tp->rcv_wup = (u32) idsn; ++ ++ meta_tp->snd_wl1 = meta_tp->rcv_nxt - 1; ++ meta_tp->snd_wnd = window; ++ meta_tp->retrans_stamp = 0; /* Set in tcp_connect() */ ++ ++ meta_tp->packets_out = 0; ++ meta_tp->mptcp->snt_isn = meta_tp->write_seq; /* Initial data-sequence-number */ ++ meta_icsk->icsk_probes_out = 0; ++ ++ /* Set mptcp-pointers */ ++ master_tp->mpcb = mpcb; ++ master_tp->meta_sk = meta_sk; ++ meta_tp->mpcb = mpcb; ++ meta_tp->meta_sk = meta_sk; ++ mpcb->meta_sk = meta_sk; ++ mpcb->master_sk = master_sk; ++ ++ set_mpc(meta_tp); ++ meta_tp->mptcp->attached = 0; ++ meta_tp->was_meta_sk = 0; ++ ++ /* Initialize the queues */ ++ skb_queue_head_init(&mpcb->reinject_queue); ++ skb_queue_head_init(&master_tp->out_of_order_queue); ++ tcp_prequeue_init(master_tp); ++ INIT_LIST_HEAD(&master_tp->tsq_node); ++ ++ master_tp->tsq_flags = 0; ++ ++ /* Copy the write-queue from the meta down to the master. ++ * This is necessary to get the SYN to the master-write-queue. ++ * No other data can be queued, before tcp_sendmsg waits for the ++ * connection to finish. ++ */ ++ skb_queue_walk_safe(&meta_sk->sk_write_queue, skb, tmp) { ++ skb_unlink(skb, &meta_sk->sk_write_queue); ++ skb_queue_tail(&master_sk->sk_write_queue, skb); ++ ++ master_sk->sk_wmem_queued += skb->truesize; ++ sk_mem_charge(master_sk, skb->truesize); ++ } ++ ++ meta_sk->sk_wmem_queued = 0; ++ meta_sk->sk_forward_alloc = 0; ++ ++ mutex_init(&mpcb->mpcb_mutex); ++ ++ /* Init the accept_queue structure, we support a queue of 32 pending ++ * connections, it does not need to be huge, since we only store here ++ * pending subflow creations. ++ */ ++ if (reqsk_queue_alloc(&meta_icsk->icsk_accept_queue, 32, GFP_ATOMIC)) { ++ inet_put_port(master_sk); ++ kmem_cache_free(mptcp_sock_cache, meta_tp->mptcp); ++ kmem_cache_free(mptcp_cb_cache, mpcb); ++ sk_free(master_sk); ++ reset_mpc(meta_tp); ++ return -ENOMEM; ++ } ++ ++ /* Redefine function-pointers as the meta-sk is now fully ready */ ++ meta_sk->sk_backlog_rcv = mptcp_backlog_rcv; ++ meta_sk->sk_destruct = mptcp_sock_destruct; ++ mpcb->syn_recv_sock = mptcp_syn_recv_sock; ++ ++ /* Meta-level retransmit timer */ ++ meta_icsk->icsk_rto *= 2; /* Double of initial - rto */ ++ ++ tcp_init_xmit_timers(master_sk); ++ /* Has been set for sending out the SYN */ ++ inet_csk_clear_xmit_timer(meta_sk, ICSK_TIME_RETRANS); ++ ++ if (!meta_tp->inside_tk_table) { ++ /* Adding the meta_tp in the token hashtable - coming from server-side */ ++ rcu_read_lock(); ++ spin_lock(&mptcp_tk_hashlock); ++ ++ __mptcp_hash_insert(meta_tp, mpcb->mptcp_loc_token); ++ ++ spin_unlock(&mptcp_tk_hashlock); ++ rcu_read_unlock(); ++ } ++ master_tp->inside_tk_table = 0; ++ ++ /* Init time-wait stuff */ ++ INIT_LIST_HEAD(&mpcb->tw_list); ++ spin_lock_init(&mpcb->tw_lock); ++ ++ INIT_LIST_HEAD(&mpcb->callback_list); ++ ++ mptcp_mpcb_inherit_sockopts(meta_sk, master_sk); ++ ++ mpcb->orig_sk_rcvbuf = meta_sk->sk_rcvbuf; ++ mpcb->orig_sk_sndbuf = meta_sk->sk_sndbuf; ++ mpcb->orig_window_clamp = meta_tp->window_clamp; ++ ++ /* The meta is directly linked - set refcnt to 1 */ ++ atomic_set(&mpcb->mpcb_refcnt, 1); ++ ++ mptcp_init_path_manager(mpcb); ++ ++ mptcp_debug("%s: created mpcb with token %#x\n", ++ __func__, mpcb->mptcp_loc_token); ++ ++ return 0; ++} ++ ++struct sock *mptcp_sk_clone(const struct sock *sk, int family, ++ const gfp_t priority) ++{ ++ struct sock *newsk = NULL; ++ ++ if (family == AF_INET && sk->sk_family == AF_INET) { ++ newsk = sk_prot_alloc(&tcp_prot, priority, family); ++ if (!newsk) ++ return NULL; ++ ++ /* Set these pointers - they are needed by mptcp_inherit_sk */ ++ newsk->sk_prot = &tcp_prot; ++ newsk->sk_prot_creator = &tcp_prot; ++ inet_csk(newsk)->icsk_af_ops = &ipv4_specific; ++ newsk->sk_family = AF_INET; ++ } ++#if IS_ENABLED(CONFIG_IPV6) ++ else { ++ newsk = sk_prot_alloc(&tcpv6_prot, priority, family); ++ if (!newsk) ++ return NULL; ++ ++ newsk->sk_prot = &tcpv6_prot; ++ newsk->sk_prot_creator = &tcpv6_prot; ++ if (family == AF_INET) ++ inet_csk(newsk)->icsk_af_ops = &ipv6_mapped; ++ else ++ inet_csk(newsk)->icsk_af_ops = &ipv6_specific; ++ newsk->sk_family = AF_INET6; ++ } ++#endif ++ ++ if (mptcp_inherit_sk(sk, newsk, family, priority)) ++ return NULL; ++ ++ return newsk; ++} ++ ++void mptcp_fallback_meta_sk(struct sock *meta_sk) ++{ ++ kfree(inet_csk(meta_sk)->icsk_accept_queue.listen_opt); ++ kmem_cache_free(mptcp_sock_cache, tcp_sk(meta_sk)->mptcp); ++ kmem_cache_free(mptcp_cb_cache, tcp_sk(meta_sk)->mpcb); ++} ++ ++int mptcp_add_sock(struct sock *meta_sk, struct sock *sk, u8 loc_id, u8 rem_id, ++ gfp_t flags) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ tp->mptcp = kmem_cache_zalloc(mptcp_sock_cache, flags); ++ if (!tp->mptcp) ++ return -ENOMEM; ++ ++ tp->mptcp->path_index = mptcp_set_new_pathindex(mpcb); ++ /* No more space for more subflows? */ ++ if (!tp->mptcp->path_index) { ++ kmem_cache_free(mptcp_sock_cache, tp->mptcp); ++ return -EPERM; ++ } ++ ++ INIT_LIST_HEAD(&tp->mptcp->cb_list); ++ ++ tp->mptcp->tp = tp; ++ tp->mpcb = mpcb; ++ tp->meta_sk = meta_sk; ++ set_mpc(tp); ++ tp->mptcp->loc_id = loc_id; ++ tp->mptcp->rem_id = rem_id; ++ tp->mptcp->last_rbuf_opti = tcp_time_stamp; ++ ++ /* The corresponding sock_put is in mptcp_sock_destruct(). It cannot be ++ * included in mptcp_del_sock(), because the mpcb must remain alive ++ * until the last subsocket is completely destroyed. ++ */ ++ sock_hold(meta_sk); ++ atomic_inc(&mpcb->mpcb_refcnt); ++ ++ tp->mptcp->next = mpcb->connection_list; ++ mpcb->connection_list = tp; ++ tp->mptcp->attached = 1; ++ ++ mpcb->cnt_subflows++; ++ atomic_add(atomic_read(&((struct sock *)tp)->sk_rmem_alloc), ++ &meta_sk->sk_rmem_alloc); ++ ++ mptcp_sub_inherit_sockopts(meta_sk, sk); ++ INIT_DELAYED_WORK(&tp->mptcp->work, mptcp_sub_close_wq); ++ ++ /* As we successfully allocated the mptcp_tcp_sock, we have to ++ * change the function-pointers here (for sk_destruct to work correctly) ++ */ ++ sk->sk_error_report = mptcp_sock_def_error_report; ++ sk->sk_data_ready = mptcp_data_ready; ++ sk->sk_write_space = mptcp_write_space; ++ sk->sk_state_change = mptcp_set_state; ++ sk->sk_destruct = mptcp_sock_destruct; ++ ++ if (sk->sk_family == AF_INET) ++ mptcp_debug("%s: token %#x pi %d, src_addr:%pI4:%d dst_addr:%pI4:%d, cnt_subflows now %d\n", ++ __func__ , mpcb->mptcp_loc_token, ++ tp->mptcp->path_index, ++ &((struct inet_sock *)tp)->inet_saddr, ++ ntohs(((struct inet_sock *)tp)->inet_sport), ++ &((struct inet_sock *)tp)->inet_daddr, ++ ntohs(((struct inet_sock *)tp)->inet_dport), ++ mpcb->cnt_subflows); ++#if IS_ENABLED(CONFIG_IPV6) ++ else ++ mptcp_debug("%s: token %#x pi %d, src_addr:%pI6:%d dst_addr:%pI6:%d, cnt_subflows now %d\n", ++ __func__ , mpcb->mptcp_loc_token, ++ tp->mptcp->path_index, &inet6_sk(sk)->saddr, ++ ntohs(((struct inet_sock *)tp)->inet_sport), ++ &sk->sk_v6_daddr, ++ ntohs(((struct inet_sock *)tp)->inet_dport), ++ mpcb->cnt_subflows); ++#endif ++ ++ return 0; ++} ++ ++void mptcp_del_sock(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk), *tp_prev; ++ struct mptcp_cb *mpcb; ++ ++ if (!tp->mptcp || !tp->mptcp->attached) ++ return; ++ ++ mpcb = tp->mpcb; ++ tp_prev = mpcb->connection_list; ++ ++ mptcp_debug("%s: Removing subsock tok %#x pi:%d state %d is_meta? %d\n", ++ __func__, mpcb->mptcp_loc_token, tp->mptcp->path_index, ++ sk->sk_state, is_meta_sk(sk)); ++ ++ if (tp_prev == tp) { ++ mpcb->connection_list = tp->mptcp->next; ++ } else { ++ for (; tp_prev && tp_prev->mptcp->next; tp_prev = tp_prev->mptcp->next) { ++ if (tp_prev->mptcp->next == tp) { ++ tp_prev->mptcp->next = tp->mptcp->next; ++ break; ++ } ++ } ++ } ++ mpcb->cnt_subflows--; ++ if (tp->mptcp->establish_increased) ++ mpcb->cnt_established--; ++ ++ tp->mptcp->next = NULL; ++ tp->mptcp->attached = 0; ++ mpcb->path_index_bits &= ~(1 << tp->mptcp->path_index); ++ ++ if (!skb_queue_empty(&sk->sk_write_queue)) ++ mptcp_reinject_data(sk, 0); ++ ++ if (is_master_tp(tp)) ++ mpcb->master_sk = NULL; ++ else if (tp->mptcp->pre_established) ++ sk_stop_timer(sk, &tp->mptcp->mptcp_ack_timer); ++ ++ rcu_assign_pointer(inet_sk(sk)->inet_opt, NULL); ++} ++ ++/* Updates the metasocket ULID/port data, based on the given sock. ++ * The argument sock must be the sock accessible to the application. ++ * In this function, we update the meta socket info, based on the changes ++ * in the application socket (bind, address allocation, ...) ++ */ ++void mptcp_update_metasocket(struct sock *sk, struct sock *meta_sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ union inet_addr addr; ++ int index; ++ ++ /* Get the index of the local address */ ++ if (sk->sk_family == AF_INET || mptcp_v6_is_v4_mapped(sk)) { ++ addr.ip = inet_sk(sk)->inet_saddr; ++ index = mpcb->pm_ops->get_local_index(AF_INET, &addr, sock_net(meta_sk)); ++ } else { ++ addr.in6 = inet6_sk(sk)->saddr; ++ index = mpcb->pm_ops->get_local_index(AF_INET6, &addr, sock_net(meta_sk)); ++ } ++ ++ if (sk->sk_family == AF_INET || mptcp_v6_is_v4_mapped(sk)) { ++ mptcp_v4_add_raddress(mpcb, ++ (struct in_addr *)&inet_sk(sk)->inet_daddr, ++ 0, 0); ++ if (index >= 0) ++ mptcp_v4_set_init_addr_bit(mpcb, inet_sk(sk)->inet_daddr, index); ++ } else { ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_v6_add_raddress(mpcb, &sk->sk_v6_daddr, 0, 0); ++ if (index >= 0) ++ mptcp_v6_set_init_addr_bit(mpcb, &sk->sk_v6_daddr, index); ++#endif ++ } ++ ++ if (mpcb->pm_ops->new_session) ++ mpcb->pm_ops->new_session(meta_sk, index); ++ ++ tcp_sk(sk)->mptcp->send_mp_prio = tcp_sk(sk)->mptcp->low_prio; ++} ++ ++/* Clean up the receive buffer for full frames taken by the user, ++ * then send an ACK if necessary. COPIED is the number of bytes ++ * tcp_recvmsg has given to the user so far, it speeds up the ++ * calculation of whether or not we must ACK for the sake of ++ * a window update. ++ */ ++void mptcp_cleanup_rbuf(struct sock *meta_sk, int copied) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sock *sk; ++ __u32 rcv_window_now = 0; ++ ++ if (copied > 0 && !(meta_sk->sk_shutdown & RCV_SHUTDOWN)) { ++ rcv_window_now = tcp_receive_window(meta_tp); ++ ++ if (2 * rcv_window_now > meta_tp->window_clamp) ++ rcv_window_now = 0; ++ } ++ ++ mptcp_for_each_sk(meta_tp->mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ const struct inet_connection_sock *icsk = inet_csk(sk); ++ ++ if (!mptcp_sk_can_send_ack(sk)) ++ continue; ++ ++ if (!inet_csk_ack_scheduled(sk)) ++ goto second_part; ++ /* Delayed ACKs frequently hit locked sockets during bulk ++ * receive. ++ */ ++ if (icsk->icsk_ack.blocked || ++ /* Once-per-two-segments ACK was not sent by tcp_input.c */ ++ tp->rcv_nxt - tp->rcv_wup > icsk->icsk_ack.rcv_mss || ++ /* If this read emptied read buffer, we send ACK, if ++ * connection is not bidirectional, user drained ++ * receive buffer and there was a small segment ++ * in queue. ++ */ ++ (copied > 0 && ++ ((icsk->icsk_ack.pending & ICSK_ACK_PUSHED2) || ++ ((icsk->icsk_ack.pending & ICSK_ACK_PUSHED) && ++ !icsk->icsk_ack.pingpong)) && ++ !atomic_read(&meta_sk->sk_rmem_alloc))) { ++ tcp_send_ack(sk); ++ continue; ++ } ++ ++second_part: ++ /* This here is the second part of tcp_cleanup_rbuf */ ++ if (rcv_window_now) { ++ __u32 new_window = tp->__select_window(sk); ++ ++ /* Send ACK now, if this read freed lots of space ++ * in our buffer. Certainly, new_window is new window. ++ * We can advertise it now, if it is not less than ++ * current one. ++ * "Lots" means "at least twice" here. ++ */ ++ if (new_window && new_window >= 2 * rcv_window_now) ++ tcp_send_ack(sk); ++ } ++ } ++} ++ ++static int mptcp_sub_send_fin(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sk_buff *skb = tcp_write_queue_tail(sk); ++ int mss_now; ++ ++ /* Optimization, tack on the FIN if we have a queue of ++ * unsent frames. But be careful about outgoing SACKS ++ * and IP options. ++ */ ++ mss_now = tcp_current_mss(sk); ++ ++ if (tcp_send_head(sk) != NULL) { ++ TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_FIN; ++ TCP_SKB_CB(skb)->end_seq++; ++ tp->write_seq++; ++ } else { ++ skb = alloc_skb_fclone(MAX_TCP_HEADER, GFP_ATOMIC); ++ if (!skb) ++ return 1; ++ ++ /* Reserve space for headers and prepare control bits. */ ++ skb_reserve(skb, MAX_TCP_HEADER); ++ /* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */ ++ tcp_init_nondata_skb(skb, tp->write_seq, ++ TCPHDR_ACK | TCPHDR_FIN); ++ tcp_queue_skb(sk, skb); ++ } ++ __tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_OFF); ++ ++ return 0; ++} ++ ++void mptcp_sub_close_wq(struct work_struct *work) ++{ ++ struct mptcp_tcp_sock *mptcp = container_of(work, struct mptcp_tcp_sock, work.work); ++ struct tcp_sock *tp = mptcp->tp; ++ struct sock *sk = (struct sock *)tp; ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ ++ mutex_lock(&tp->mpcb->mpcb_mutex); ++ lock_sock_nested(meta_sk, SINGLE_DEPTH_NESTING); ++ ++ if (sock_flag(sk, SOCK_DEAD)) ++ goto exit; ++ ++ /* We come from tcp_disconnect. We are sure that meta_sk is set */ ++ if (!tp->mpc) { ++ tp->closing = 1; ++ sock_rps_reset_flow(sk); ++ tcp_close(sk, 0); ++ goto exit; ++ } ++ ++ if (meta_sk->sk_shutdown == SHUTDOWN_MASK || sk->sk_state == TCP_CLOSE) { ++ tp->closing = 1; ++ sock_rps_reset_flow(sk); ++ tcp_close(sk, 0); ++ } else if (tcp_close_state(sk)) { ++ sk->sk_shutdown |= SEND_SHUTDOWN; ++ tcp_send_fin(sk); ++ } ++ ++exit: ++ release_sock(meta_sk); ++ mutex_unlock(&tp->mpcb->mpcb_mutex); ++ sock_put(sk); ++} ++ ++void mptcp_sub_close(struct sock *sk, unsigned long delay) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct delayed_work *work = &tcp_sk(sk)->mptcp->work; ++ ++ /* We are already closing - e.g., call from sock_def_error_report upon ++ * tcp_disconnect in tcp_close. ++ */ ++ if (tp->closing) ++ return; ++ ++ /* Work already scheduled ? */ ++ if (work_pending(&work->work)) { ++ /* Work present - who will be first ? */ ++ if (jiffies + delay > work->timer.expires) ++ return; ++ ++ /* Try canceling - if it fails, work will be executed soon */ ++ if (!cancel_delayed_work(work)) ++ return; ++ sock_put(sk); ++ } ++ ++ if (!delay) { ++ unsigned char old_state = sk->sk_state; ++ ++ /* If we are in user-context we can directly do the closing ++ * procedure. No need to schedule a work-queue. ++ */ ++ if (!in_softirq()) { ++ if (sock_flag(sk, SOCK_DEAD)) ++ return; ++ ++ if (!tp->mpc) { ++ tp->closing = 1; ++ sock_rps_reset_flow(sk); ++ tcp_close(sk, 0); ++ return; ++ } ++ ++ if (mptcp_meta_sk(sk)->sk_shutdown == SHUTDOWN_MASK || ++ sk->sk_state == TCP_CLOSE) { ++ tp->closing = 1; ++ sock_rps_reset_flow(sk); ++ tcp_close(sk, 0); ++ } else if (tcp_close_state(sk)) { ++ sk->sk_shutdown |= SEND_SHUTDOWN; ++ tcp_send_fin(sk); ++ } ++ ++ return; ++ } ++ ++ /* We directly send the FIN. Because it may take so a long time, ++ * untile the work-queue will get scheduled... ++ * ++ * If mptcp_sub_send_fin returns 1, it failed and thus we reset ++ * the old state so that tcp_close will finally send the fin ++ * in user-context. ++ */ ++ if (!sk->sk_err && old_state != TCP_CLOSE && ++ tcp_close_state(sk) && mptcp_sub_send_fin(sk)) { ++ if (old_state == TCP_ESTABLISHED) ++ TCP_INC_STATS(sock_net(sk), TCP_MIB_CURRESTAB); ++ sk->sk_state = old_state; ++ } ++ } ++ ++ sock_hold(sk); ++ queue_delayed_work(mptcp_wq, work, delay); ++} ++ ++void mptcp_sub_force_close(struct sock *sk) ++{ ++ /* The below tcp_done may have freed the socket, if he is already dead. ++ * Thus, we are not allowed to access it afterwards. That's why ++ * we have to store the dead-state in this local variable. ++ */ ++ int sock_is_dead = sock_flag(sk, SOCK_DEAD); ++ ++ tcp_sk(sk)->mp_killed = 1; ++ ++ if (sk->sk_state != TCP_CLOSE) ++ tcp_done(sk); ++ ++ if (!sock_is_dead) ++ mptcp_sub_close(sk, 0); ++} ++EXPORT_SYMBOL(mptcp_sub_force_close); ++ ++/* Update the mpcb send window, based on the contributions ++ * of each subflow ++ */ ++void mptcp_update_sndbuf(struct mptcp_cb *mpcb) ++{ ++ struct sock *meta_sk = mpcb->meta_sk, *sk; ++ int new_sndbuf = 0, old_sndbuf = meta_sk->sk_sndbuf; ++ mptcp_for_each_sk(mpcb, sk) { ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ ++ new_sndbuf += sk->sk_sndbuf; ++ ++ if (new_sndbuf > sysctl_tcp_wmem[2] || new_sndbuf < 0) { ++ new_sndbuf = sysctl_tcp_wmem[2]; ++ break; ++ } ++ } ++ meta_sk->sk_sndbuf = max(min(new_sndbuf, sysctl_tcp_wmem[2]), meta_sk->sk_sndbuf); ++ ++ /* The subflow's call to sk_write_space in tcp_new_space ends up in ++ * mptcp_write_space. ++ * It has nothing to do with waking up the application. ++ * So, we do it here. ++ */ ++ if (old_sndbuf != meta_sk->sk_sndbuf) ++ meta_sk->sk_write_space(meta_sk); ++} ++ ++void mptcp_close(struct sock *meta_sk, long timeout) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sock *sk_it, *tmpsk; ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sk_buff *skb; ++ int data_was_unread = 0; ++ int state; ++ ++ mptcp_debug("%s: Close of meta_sk with tok %#x\n", ++ __func__, mpcb->mptcp_loc_token); ++ ++ mutex_lock(&mpcb->mpcb_mutex); ++ lock_sock(meta_sk); ++ ++ if (meta_tp->inside_tk_table) { ++ /* Detach the mpcb from the token hashtable */ ++ mptcp_hash_remove_bh(meta_tp); ++ reqsk_queue_destroy(&inet_csk(meta_sk)->icsk_accept_queue); ++ } ++ ++ meta_sk->sk_shutdown = SHUTDOWN_MASK; ++ /* We need to flush the recv. buffs. We do this only on the ++ * descriptor close, not protocol-sourced closes, because the ++ * reader process may not have drained the data yet! ++ */ ++ while ((skb = __skb_dequeue(&meta_sk->sk_receive_queue)) != NULL) { ++ u32 len = TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq - ++ tcp_hdr(skb)->fin; ++ data_was_unread += len; ++ __kfree_skb(skb); ++ } ++ ++ sk_mem_reclaim(meta_sk); ++ ++ /* If socket has been already reset (e.g. in tcp_reset()) - kill it. */ ++ if (meta_sk->sk_state == TCP_CLOSE) { ++ mptcp_for_each_sk_safe(mpcb, sk_it, tmpsk) { ++ if (tcp_sk(sk_it)->send_mp_fclose) ++ continue; ++ mptcp_sub_close(sk_it, 0); ++ } ++ goto adjudge_to_death; ++ } ++ ++ if (data_was_unread) { ++ /* Unread data was tossed, zap the connection. */ ++ NET_INC_STATS_USER(sock_net(meta_sk), LINUX_MIB_TCPABORTONCLOSE); ++ tcp_set_state(meta_sk, TCP_CLOSE); ++ tcp_send_active_reset(meta_sk, meta_sk->sk_allocation); ++ } else if (sock_flag(meta_sk, SOCK_LINGER) && !meta_sk->sk_lingertime) { ++ /* Check zero linger _after_ checking for unread data. */ ++ meta_sk->sk_prot->disconnect(meta_sk, 0); ++ NET_INC_STATS_USER(sock_net(meta_sk), LINUX_MIB_TCPABORTONDATA); ++ } else if (tcp_close_state(meta_sk)) { ++ mptcp_send_fin(meta_sk); ++ } else if (meta_tp->snd_una == meta_tp->write_seq) { ++ /* The DATA_FIN has been sent and acknowledged ++ * (e.g., by sk_shutdown). Close all the other subflows ++ */ ++ mptcp_for_each_sk_safe(mpcb, sk_it, tmpsk) { ++ unsigned long delay = 0; ++ /* If we are the passive closer, don't trigger ++ * subflow-fin until the subflow has been finned ++ * by the peer. - thus we add a delay ++ */ ++ if (mpcb->passive_close && ++ sk_it->sk_state == TCP_ESTABLISHED) ++ delay = inet_csk(sk_it)->icsk_rto << 3; ++ ++ mptcp_sub_close(sk_it, delay); ++ } ++ } ++ ++ sk_stream_wait_close(meta_sk, timeout); ++ ++adjudge_to_death: ++ state = meta_sk->sk_state; ++ sock_hold(meta_sk); ++ sock_orphan(meta_sk); ++ ++ /* socket will be freed after mptcp_close - we have to prevent ++ * access from the subflows. ++ */ ++ mptcp_for_each_sk(mpcb, sk_it) { ++ /* Similar to sock_orphan, but we don't set it DEAD, because ++ * the callbacks are still set and must be called. ++ */ ++ write_lock_bh(&sk_it->sk_callback_lock); ++ sk_set_socket(sk_it, NULL); ++ sk_it->sk_wq = NULL; ++ write_unlock_bh(&sk_it->sk_callback_lock); ++ } ++ ++ /* It is the last release_sock in its life. It will remove backlog. */ ++ release_sock(meta_sk); ++ ++ /* Now socket is owned by kernel and we acquire BH lock ++ * to finish close. No need to check for user refs. ++ */ ++ local_bh_disable(); ++ bh_lock_sock(meta_sk); ++ WARN_ON(sock_owned_by_user(meta_sk)); ++ ++ percpu_counter_inc(meta_sk->sk_prot->orphan_count); ++ ++ /* Have we already been destroyed by a softirq or backlog? */ ++ if (state != TCP_CLOSE && meta_sk->sk_state == TCP_CLOSE) ++ goto out; ++ ++ /* This is a (useful) BSD violating of the RFC. There is a ++ * problem with TCP as specified in that the other end could ++ * keep a socket open forever with no application left this end. ++ * We use a 3 minute timeout (about the same as BSD) then kill ++ * our end. If they send after that then tough - BUT: long enough ++ * that we won't make the old 4*rto = almost no time - whoops ++ * reset mistake. ++ * ++ * Nope, it was not mistake. It is really desired behaviour ++ * f.e. on http servers, when such sockets are useless, but ++ * consume significant resources. Let's do it with special ++ * linger2 option. --ANK ++ */ ++ ++ if (meta_sk->sk_state == TCP_FIN_WAIT2) { ++ if (meta_tp->linger2 < 0) { ++ tcp_set_state(meta_sk, TCP_CLOSE); ++ tcp_send_active_reset(meta_sk, GFP_ATOMIC); ++ NET_INC_STATS_BH(sock_net(meta_sk), ++ LINUX_MIB_TCPABORTONLINGER); ++ } else { ++ const int tmo = tcp_fin_time(meta_sk); ++ ++ if (tmo > TCP_TIMEWAIT_LEN) { ++ inet_csk_reset_keepalive_timer(meta_sk, ++ tmo - TCP_TIMEWAIT_LEN); ++ } else { ++ tcp_time_wait(meta_sk, TCP_FIN_WAIT2, tmo); ++ goto out; ++ } ++ } ++ } ++ if (meta_sk->sk_state != TCP_CLOSE) { ++ sk_mem_reclaim(meta_sk); ++ if (tcp_too_many_orphans(meta_sk, 0)) { ++ if (net_ratelimit()) ++ pr_info("MPTCP: too many of orphaned sockets\n"); ++ tcp_set_state(meta_sk, TCP_CLOSE); ++ tcp_send_active_reset(meta_sk, GFP_ATOMIC); ++ NET_INC_STATS_BH(sock_net(meta_sk), ++ LINUX_MIB_TCPABORTONMEMORY); ++ } ++ } ++ ++ ++ if (meta_sk->sk_state == TCP_CLOSE) ++ inet_csk_destroy_sock(meta_sk); ++ /* Otherwise, socket is reprieved until protocol close. */ ++ ++out: ++ bh_unlock_sock(meta_sk); ++ local_bh_enable(); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ sock_put(meta_sk); /* Taken by sock_hold */ ++} ++ ++void mptcp_disconnect(struct sock *sk) ++{ ++ struct sock *subsk, *tmpsk; ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ __skb_queue_purge(&tp->mpcb->reinject_queue); ++ ++ if (tp->inside_tk_table) { ++ mptcp_hash_remove_bh(tp); ++ reqsk_queue_destroy(&inet_csk(tp->meta_sk)->icsk_accept_queue); ++ } ++ ++ local_bh_disable(); ++ mptcp_for_each_sk_safe(tp->mpcb, subsk, tmpsk) { ++ /* The socket will get removed from the subsocket-list ++ * and made non-mptcp by setting mpc to 0. ++ * ++ * This is necessary, because tcp_disconnect assumes ++ * that the connection is completly dead afterwards. ++ * Thus we need to do a mptcp_del_sock. Due to this call ++ * we have to make it non-mptcp. ++ * ++ * We have to lock the socket, because we set mpc to 0. ++ * An incoming packet would take the subsocket's lock ++ * and go on into the receive-path. ++ * This would be a race. ++ */ ++ ++ bh_lock_sock(subsk); ++ mptcp_del_sock(subsk); ++ reset_mpc(tcp_sk(subsk)); ++ mptcp_sub_force_close(subsk); ++ bh_unlock_sock(subsk); ++ } ++ local_bh_enable(); ++ ++ tp->was_meta_sk = 1; ++ reset_mpc(tp); ++} ++ ++ ++/* Returns 1 if we should enable MPTCP for that socket. */ ++int mptcp_doit(struct sock *sk) ++{ ++ /* Do not allow MPTCP enabling if the MPTCP initialization failed */ ++ if (mptcp_init_failed) ++ return 0; ++ ++ if (sysctl_mptcp_enabled == MPTCP_APP && !tcp_sk(sk)->mptcp_enabled) ++ return 0; ++ ++ /* Socket may already be established (e.g., called from tcp_recvmsg) */ ++ if (tcp_sk(sk)->mpc || tcp_sk(sk)->request_mptcp) ++ return 1; ++ ++ /* Don't do mptcp over loopback */ ++ if (sk->sk_family == AF_INET && ++ (ipv4_is_loopback(inet_sk(sk)->inet_daddr) || ++ ipv4_is_loopback(inet_sk(sk)->inet_saddr))) ++ return 0; ++#if IS_ENABLED(CONFIG_IPV6) ++ if (sk->sk_family == AF_INET6 && ++ (ipv6_addr_loopback(&sk->sk_v6_daddr) || ++ ipv6_addr_loopback(&inet6_sk(sk)->saddr))) ++ return 0; ++#endif ++ if (mptcp_v6_is_v4_mapped(sk) && ++ ipv4_is_loopback(inet_sk(sk)->inet_saddr)) ++ return 0; ++ ++#ifdef CONFIG_TCP_MD5SIG ++ /* If TCP_MD5SIG is enabled, do not do MPTCP - there is no Option-Space */ ++ if (tcp_sk(sk)->af_specific->md5_lookup(sk, sk)) ++ return 0; ++#endif ++ ++ return 1; ++} ++ ++int mptcp_create_master_sk(struct sock *meta_sk, __u64 remote_key, u32 window) ++{ ++ struct tcp_sock *master_tp; ++ struct sock *master_sk; ++ ++ if (mptcp_alloc_mpcb(meta_sk, remote_key, window)) ++ goto err_alloc_mpcb; ++ ++ master_sk = tcp_sk(meta_sk)->mpcb->master_sk; ++ master_tp = tcp_sk(master_sk); ++ ++ if (mptcp_add_sock(meta_sk, master_sk, 0, 0, GFP_ATOMIC)) ++ goto err_add_sock; ++ ++ if (__inet_inherit_port(meta_sk, master_sk) < 0) ++ goto err_add_sock; ++ ++ meta_sk->sk_prot->unhash(meta_sk); ++ ++ if (master_sk->sk_family == AF_INET || mptcp_v6_is_v4_mapped(master_sk)) ++ __inet_hash_nolisten(master_sk, NULL); ++#if IS_ENABLED(CONFIG_IPV6) ++ else ++ __inet6_hash(master_sk, NULL); ++#endif ++ ++ master_tp->mptcp->init_rcv_wnd = master_tp->rcv_wnd; ++ ++ return 0; ++ ++err_add_sock: ++ mptcp_fallback_meta_sk(meta_sk); ++ ++ inet_csk_prepare_forced_close(master_sk); ++ tcp_done(master_sk); ++ inet_csk_prepare_forced_close(meta_sk); ++ tcp_done(meta_sk); ++ ++err_alloc_mpcb: ++ return -ENOBUFS; ++} ++ ++int mptcp_check_req_master(struct sock *sk, struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ struct mptcp_options_received *mopt) ++{ ++ struct tcp_sock *child_tp = tcp_sk(child); ++ struct sock *meta_sk = child; ++ struct mptcp_cb *mpcb; ++ struct mptcp_request_sock *mtreq; ++ ++ if (!tcp_rsk(req)->saw_mpc) ++ return 1; ++ ++ /* Just set this values to pass them to mptcp_alloc_mpcb */ ++ mtreq = mptcp_rsk(req); ++ child_tp->mptcp_loc_key = mtreq->mptcp_loc_key; ++ child_tp->mptcp_loc_token = mtreq->mptcp_loc_token; ++ ++ if (mptcp_create_master_sk(meta_sk, mtreq->mptcp_rem_key, ++ child_tp->snd_wnd)) ++ return -ENOBUFS; ++ ++ child = tcp_sk(child)->mpcb->master_sk; ++ child_tp = tcp_sk(child); ++ mpcb = child_tp->mpcb; ++ ++ child_tp->mptcp->snt_isn = tcp_rsk(req)->snt_isn; ++ child_tp->mptcp->rcv_isn = tcp_rsk(req)->rcv_isn; ++ ++ mpcb->dss_csum = mtreq->dss_csum; ++ mpcb->server_side = 1; ++ ++ /* Will be moved to ESTABLISHED by tcp_rcv_state_process() */ ++ mptcp_update_metasocket(child, meta_sk); ++ ++ /* Needs to be done here additionally, because when accepting a ++ * new connection we pass by __reqsk_free and not reqsk_free. ++ */ ++ mptcp_reqsk_remove_tk(req); ++ ++ /* Hold when creating the meta-sk in tcp_vX_syn_recv_sock. */ ++ sock_put(meta_sk); ++ ++ inet_csk_reqsk_queue_unlink(sk, req, prev); ++ inet_csk_reqsk_queue_removed(sk, req); ++ inet_csk_reqsk_queue_add(sk, req, meta_sk); ++ ++ return 0; ++} ++ ++struct sock *mptcp_check_req_child(struct sock *meta_sk, struct sock *child, ++ struct request_sock *req, ++ struct request_sock **prev, ++ struct mptcp_options_received *mopt) ++{ ++ struct tcp_sock *child_tp = tcp_sk(child); ++ struct mptcp_request_sock *mtreq = mptcp_rsk(req); ++ struct mptcp_cb *mpcb = mtreq->mpcb; ++ u8 hash_mac_check[20]; ++ ++ child_tp->inside_tk_table = 0; ++ ++ if (!mopt->join_ack) ++ goto teardown; ++ ++ mptcp_hmac_sha1((u8 *)&mpcb->mptcp_rem_key, ++ (u8 *)&mpcb->mptcp_loc_key, ++ (u8 *)&mtreq->mptcp_rem_nonce, ++ (u8 *)&mtreq->mptcp_loc_nonce, ++ (u32 *)hash_mac_check); ++ ++ if (memcmp(hash_mac_check, (char *)&mopt->mptcp_recv_mac, 20)) ++ goto teardown; ++ ++ /* Point it to the same struct socket and wq as the meta_sk */ ++ sk_set_socket(child, meta_sk->sk_socket); ++ child->sk_wq = meta_sk->sk_wq; ++ ++ if (mptcp_add_sock(meta_sk, child, mtreq->loc_id, mtreq->rem_id, GFP_ATOMIC)) { ++ reset_mpc(child_tp); /* Has been inherited, but now ++ * child_tp->mptcp is NULL ++ */ ++ /* TODO when we support acking the third ack for new subflows, ++ * we should silently discard this third ack, by returning NULL. ++ * ++ * Maybe, at the retransmission we will have enough memory to ++ * fully add the socket to the meta-sk. ++ */ ++ goto teardown; ++ } ++ ++ /* The child is a clone of the meta socket, we must now reset ++ * some of the fields ++ */ ++ child_tp->mptcp->rcv_low_prio = mtreq->low_prio; ++ ++ /* We should allow proper increase of the snd/rcv-buffers. Thus, we ++ * use the original values instead of the bloated up ones from the ++ * clone. ++ */ ++ child->sk_sndbuf = mpcb->orig_sk_sndbuf; ++ child->sk_rcvbuf = mpcb->orig_sk_rcvbuf; ++ ++ child_tp->mptcp->slave_sk = 1; ++ child_tp->mptcp->snt_isn = tcp_rsk(req)->snt_isn; ++ child_tp->mptcp->rcv_isn = tcp_rsk(req)->rcv_isn; ++ child_tp->mptcp->init_rcv_wnd = req->rcv_wnd; ++ ++ child_tp->tsq_flags = 0; ++ ++ /* Subflows do not use the accept queue, as they ++ * are attached immediately to the mpcb. ++ */ ++ inet_csk_reqsk_queue_drop(meta_sk, req, prev); ++ return child; ++ ++teardown: ++ /* Drop this request - sock creation failed. */ ++ inet_csk_reqsk_queue_drop(meta_sk, req, prev); ++ inet_csk_prepare_forced_close(child); ++ tcp_done(child); ++ return meta_sk; ++} ++ ++int mptcp_time_wait(struct sock *sk, struct tcp_timewait_sock *tw) ++{ ++ struct mptcp_tw *mptw; ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ ++ /* Alloc MPTCP-tw-sock */ ++ mptw = kmem_cache_alloc(mptcp_tw_cache, GFP_ATOMIC); ++ if (!mptw) ++ return -ENOBUFS; ++ ++ atomic_inc(&mpcb->mpcb_refcnt); ++ ++ tw->mptcp_tw = mptw; ++ mptw->loc_key = mpcb->mptcp_loc_key; ++ mptw->meta_tw = mpcb->in_time_wait; ++ if (mptw->meta_tw) { ++ mptw->rcv_nxt = mptcp_get_rcv_nxt_64(mptcp_meta_tp(tp)); ++ if (mpcb->mptw_state != TCP_TIME_WAIT) ++ mptw->rcv_nxt++; ++ } ++ rcu_assign_pointer(mptw->mpcb, mpcb); ++ ++ spin_lock(&mpcb->tw_lock); ++ list_add_rcu(&mptw->list, &tp->mpcb->tw_list); ++ mptw->in_list = 1; ++ spin_unlock(&mpcb->tw_lock); ++ ++ return 0; ++} ++ ++void mptcp_twsk_destructor(struct tcp_timewait_sock *tw) ++{ ++ struct mptcp_cb *mpcb; ++ ++ rcu_read_lock(); ++ mpcb = rcu_dereference(tw->mptcp_tw->mpcb); ++ ++ /* If we are still holding a ref to the mpcb, we have to remove ourself ++ * from the list and drop the ref properly. ++ */ ++ if (mpcb && atomic_inc_not_zero(&mpcb->mpcb_refcnt)) { ++ spin_lock(&mpcb->tw_lock); ++ if (tw->mptcp_tw->in_list) { ++ list_del_rcu(&tw->mptcp_tw->list); ++ tw->mptcp_tw->in_list = 0; ++ } ++ spin_unlock(&mpcb->tw_lock); ++ ++ /* Twice, because we increased it above */ ++ mptcp_mpcb_put(mpcb); ++ mptcp_mpcb_put(mpcb); ++ } ++ ++ rcu_read_unlock(); ++ ++ kmem_cache_free(mptcp_tw_cache, tw->mptcp_tw); ++} ++ ++/* Updates the rcv_nxt of the time-wait-socks and allows them to ack a ++ * data-fin. ++ */ ++void mptcp_update_tw_socks(const struct tcp_sock *tp, int state) ++{ ++ struct mptcp_tw *mptw; ++ ++ /* Used for sockets that go into tw after the meta ++ * (see mptcp_time_wait()) ++ */ ++ tp->mpcb->in_time_wait = 1; ++ tp->mpcb->mptw_state = state; ++ ++ /* Update the time-wait-sock's information */ ++ rcu_read_lock_bh(); ++ list_for_each_entry_rcu(mptw, &tp->mpcb->tw_list, list) { ++ mptw->meta_tw = 1; ++ mptw->rcv_nxt = mptcp_get_rcv_nxt_64(tp); ++ ++ /* We want to ack a DATA_FIN, but are yet in FIN_WAIT_2 - ++ * pretend as if the DATA_FIN has already reached us, that way ++ * the checks in tcp_timewait_state_process will be good as the ++ * DATA_FIN comes in. ++ */ ++ if (state != TCP_TIME_WAIT) ++ mptw->rcv_nxt++; ++ } ++ rcu_read_unlock_bh(); ++} ++ ++void mptcp_tsq_flags(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ ++ /* It will be handled as a regular deferred-call */ ++ if (is_meta_sk(sk)) ++ return; ++ ++ if (list_empty(&tp->mptcp->cb_list)) { ++ list_add(&tp->mptcp->cb_list, &tp->mpcb->callback_list); ++ /* We need to hold it here, as the sock_hold is not assured ++ * by the release_sock as it is done in regular TCP. ++ * ++ * The subsocket may get inet_csk_destroy'd while it is inside ++ * the callback_list. ++ */ ++ sock_hold(sk); ++ } ++ ++ if (!test_and_set_bit(MPTCP_SUB_DEFERRED, &tcp_sk(meta_sk)->tsq_flags)) ++ sock_hold(meta_sk); ++} ++ ++void mptcp_tsq_sub_deferred(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_tcp_sock *mptcp, *tmp; ++ ++ BUG_ON(!is_meta_sk(meta_sk) && !meta_tp->was_meta_sk); ++ ++ __sock_put(meta_sk); ++ list_for_each_entry_safe(mptcp, tmp, &meta_tp->mpcb->callback_list, cb_list) { ++ struct tcp_sock *tp = mptcp->tp; ++ struct sock *sk = (struct sock *)tp; ++ ++ list_del_init(&mptcp->cb_list); ++ sk->sk_prot->release_cb(sk); ++ /* Final sock_put (cfr. mptcp_tsq_flags */ ++ sock_put(sk); ++ } ++} ++ ++struct workqueue_struct *mptcp_wq; ++EXPORT_SYMBOL(mptcp_wq); ++ ++/* Output /proc/net/mptcp */ ++static int mptcp_pm_seq_show(struct seq_file *seq, void *v) ++{ ++ struct tcp_sock *meta_tp; ++ struct net *net = seq->private; ++ int i, n = 0; ++ ++ seq_printf(seq, " sl loc_tok rem_tok v6 " ++ "local_address " ++ "remote_address " ++ "st ns tx_queue rx_queue inode"); ++ seq_putc(seq, '\n'); ++ ++ for (i = 0; i < MPTCP_HASH_SIZE; i++) { ++ struct hlist_nulls_node *node; ++ rcu_read_lock_bh(); ++ hlist_nulls_for_each_entry_rcu(meta_tp, node, ++ &tk_hashtable[i], tk_table) { ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sock *meta_sk = (struct sock *)meta_tp; ++ struct inet_sock *isk = inet_sk(meta_sk); ++ ++ if (!meta_tp->mpc || !net_eq(net, sock_net(meta_sk))) ++ continue; ++ ++ seq_printf(seq, "%4d: %04X %04X ", n++, ++ mpcb->mptcp_loc_token, ++ mpcb->mptcp_rem_token); ++ if (meta_sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(meta_sk)) { ++ seq_printf(seq, " 0 %08X:%04X %08X:%04X ", ++ isk->inet_saddr, ++ ntohs(isk->inet_sport), ++ isk->inet_daddr, ++ ntohs(isk->inet_dport)); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else if (meta_sk->sk_family == AF_INET6) { ++ struct in6_addr *src = &isk->pinet6->saddr; ++ struct in6_addr *dst = &meta_sk->sk_v6_daddr; ++ seq_printf(seq, " 1 %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X", ++ src->s6_addr32[0], src->s6_addr32[1], ++ src->s6_addr32[2], src->s6_addr32[3], ++ ntohs(isk->inet_sport), ++ dst->s6_addr32[0], dst->s6_addr32[1], ++ dst->s6_addr32[2], dst->s6_addr32[3], ++ ntohs(isk->inet_dport)); ++#endif ++ } ++ seq_printf(seq, " %02X %02X %08X:%08X %lu", ++ meta_sk->sk_state, mpcb->cnt_subflows, ++ meta_tp->write_seq - meta_tp->snd_una, ++ max_t(int, meta_tp->rcv_nxt - ++ meta_tp->copied_seq, 0), ++ sock_i_ino(meta_sk)); ++ seq_putc(seq, '\n'); ++ } ++ rcu_read_unlock_bh(); ++ } ++ ++ return 0; ++} ++ ++static int mptcp_pm_seq_open(struct inode *inode, struct file *file) ++{ ++ return single_open_net(inode, file, mptcp_pm_seq_show); ++} ++ ++static const struct file_operations mptcp_pm_seq_fops = { ++ .owner = THIS_MODULE, ++ .open = mptcp_pm_seq_open, ++ .read = seq_read, ++ .llseek = seq_lseek, ++ .release = single_release_net, ++}; ++ ++static int mptcp_pm_init_net(struct net *net) ++{ ++ if (!proc_create("mptcp", S_IRUGO, net->proc_net, &mptcp_pm_seq_fops)) ++ return -ENOMEM; ++ ++ return 0; ++} ++ ++static void mptcp_pm_exit_net(struct net *net) ++{ ++ remove_proc_entry("mptcp", net->proc_net); ++} ++ ++static struct pernet_operations mptcp_pm_proc_ops = { ++ .init = mptcp_pm_init_net, ++ .exit = mptcp_pm_exit_net, ++}; ++ ++/* General initialization of mptcp */ ++void __init mptcp_init(void) ++{ ++ int i; ++ struct ctl_table_header *mptcp_sysctl; ++ ++ mptcp_sock_cache = kmem_cache_create("mptcp_sock", ++ sizeof(struct mptcp_tcp_sock), ++ 0, SLAB_HWCACHE_ALIGN, ++ NULL); ++ if (!mptcp_sock_cache) ++ goto mptcp_sock_cache_failed; ++ ++ mptcp_cb_cache = kmem_cache_create("mptcp_cb", sizeof(struct mptcp_cb), ++ 0, SLAB_DESTROY_BY_RCU|SLAB_HWCACHE_ALIGN, ++ NULL); ++ if (!mptcp_cb_cache) ++ goto mptcp_cb_cache_failed; ++ ++ mptcp_tw_cache = kmem_cache_create("mptcp_tw", sizeof(struct mptcp_tw), ++ 0, SLAB_DESTROY_BY_RCU|SLAB_HWCACHE_ALIGN, ++ NULL); ++ if (!mptcp_tw_cache) ++ goto mptcp_tw_cache_failed; ++ ++ get_random_bytes(mptcp_secret, sizeof(mptcp_secret)); ++ ++ mptcp_wq = alloc_workqueue("mptcp_wq", WQ_UNBOUND | WQ_MEM_RECLAIM, 8); ++ if (!mptcp_wq) ++ goto alloc_workqueue_failed; ++ ++ for (i = 0; i < MPTCP_HASH_SIZE; i++) { ++ INIT_HLIST_NULLS_HEAD(&tk_hashtable[i], i); ++ INIT_LIST_HEAD(&mptcp_reqsk_htb[i]); ++ INIT_HLIST_NULLS_HEAD(&mptcp_reqsk_tk_htb[i], i); ++ } ++ ++ spin_lock_init(&mptcp_reqsk_hlock); ++ spin_lock_init(&mptcp_tk_hashlock); ++ ++ if (register_pernet_subsys(&mptcp_pm_proc_ops)) ++ goto pernet_failed; ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ if (mptcp_pm_v6_init()) ++ goto mptcp_pm_v6_failed; ++#endif ++ if (mptcp_pm_v4_init()) ++ goto mptcp_pm_v4_failed; ++ ++ mptcp_sysctl = register_net_sysctl(&init_net, "net/mptcp", mptcp_table); ++ if (!mptcp_sysctl) ++ goto register_sysctl_failed; ++ ++ if (mptcp_register_path_manager(&mptcp_pm_default)) ++ goto register_pm_failed; ++ ++ pr_info("MPTCP: Stable release v0.89.0-rc"); ++ ++ mptcp_init_failed = false; ++ ++ return; ++ ++register_pm_failed: ++ unregister_net_sysctl_table(mptcp_sysctl); ++register_sysctl_failed: ++ mptcp_pm_v4_undo(); ++mptcp_pm_v4_failed: ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_pm_v6_undo(); ++mptcp_pm_v6_failed: ++#endif ++ unregister_pernet_subsys(&mptcp_pm_proc_ops); ++pernet_failed: ++ destroy_workqueue(mptcp_wq); ++alloc_workqueue_failed: ++ kmem_cache_destroy(mptcp_tw_cache); ++mptcp_tw_cache_failed: ++ kmem_cache_destroy(mptcp_cb_cache); ++mptcp_cb_cache_failed: ++ kmem_cache_destroy(mptcp_sock_cache); ++mptcp_sock_cache_failed: ++ mptcp_init_failed = true; ++} +diff --git a/net/mptcp/mptcp_fullmesh.c b/net/mptcp/mptcp_fullmesh.c +new file mode 100644 +index 0000000..49bddf3 +--- /dev/null ++++ b/net/mptcp/mptcp_fullmesh.c +@@ -0,0 +1,1313 @@ ++#include <linux/module.h> ++ ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++ ++#if IS_ENABLED(CONFIG_IPV6) ++#include <net/mptcp_v6.h> ++#include <net/addrconf.h> ++#endif ++ ++enum { ++ MPTCP_EVENT_ADD = 1, ++ MPTCP_EVENT_DEL, ++ MPTCP_EVENT_MOD, ++}; ++ ++struct mptcp_loc_addr { ++ struct mptcp_loc4 locaddr4[MPTCP_MAX_ADDR]; ++ u8 loc4_bits; ++ u8 next_v4_index; ++ ++ struct mptcp_loc6 locaddr6[MPTCP_MAX_ADDR]; ++ u8 loc6_bits; ++ u8 next_v6_index; ++}; ++ ++struct mptcp_addr_event { ++ struct list_head list; ++ unsigned short family; ++ u8 code:7, ++ low_prio:1; ++ union inet_addr addr; ++}; ++ ++struct fullmesh_priv { ++ /* Worker struct for subflow establishment */ ++ struct work_struct subflow_work; ++ /* Delayed worker, when the routing-tables are not yet ready. */ ++ struct delayed_work subflow_retry_work; ++ ++ struct mptcp_cb *mpcb; ++ ++ u16 remove_addrs; /* Addresses to remove */ ++ u8 announced_addrs_v4; /* IPv4 Addresses we did announce */ ++ u8 announced_addrs_v6; /* IPv4 Addresses we did announce */ ++ ++ u8 add_addr; /* Are we sending an add_addr? */ ++}; ++ ++struct mptcp_fm_ns { ++ struct mptcp_loc_addr __rcu *local; ++ spinlock_t local_lock; /* Protecting the above pointer */ ++ struct list_head events; ++ struct delayed_work address_worker; ++ ++ struct net *net; ++}; ++ ++static struct mptcp_pm_ops full_mesh __read_mostly; ++ ++static struct mptcp_fm_ns *fm_get_ns(struct net *net) ++{ ++ return (struct mptcp_fm_ns *)net->mptcp.path_managers[MPTCP_PM_FULLMESH]; ++} ++ ++static void full_mesh_create_subflows(struct sock *meta_sk); ++ ++static void retry_subflow_worker(struct work_struct *work) ++{ ++ struct delayed_work *delayed_work = container_of(work, ++ struct delayed_work, ++ work); ++ struct fullmesh_priv *pm_priv = container_of(delayed_work, ++ struct fullmesh_priv, ++ subflow_retry_work); ++ struct mptcp_cb *mpcb = pm_priv->mpcb; ++ struct sock *meta_sk = mpcb->meta_sk; ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(sock_net(meta_sk)); ++ int iter = 0, i; ++ ++ /* We need a local (stable) copy of the address-list. Really, it is not ++ * such a big deal, if the address-list is not 100% up-to-date. ++ */ ++ rcu_read_lock_bh(); ++ mptcp_local = rcu_dereference_bh(fm_ns->local); ++ mptcp_local = kmemdup(mptcp_local, sizeof(*mptcp_local), GFP_ATOMIC); ++ rcu_read_unlock_bh(); ++ ++ if (!mptcp_local) ++ return; ++ ++next_subflow: ++ if (iter) { ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ ++ yield(); ++ } ++ mutex_lock(&mpcb->mpcb_mutex); ++ lock_sock_nested(meta_sk, SINGLE_DEPTH_NESTING); ++ ++ iter++; ++ ++ if (sock_flag(meta_sk, SOCK_DEAD)) ++ goto exit; ++ ++ mptcp_for_each_bit_set(mpcb->rem4_bits, i) { ++ struct mptcp_rem4 *rem = &mpcb->remaddr4[i]; ++ /* Do we need to retry establishing a subflow ? */ ++ if (rem->retry_bitfield) { ++ int i = mptcp_find_free_index(~rem->retry_bitfield); ++ ++ rem->bitfield |= (1 << i); ++ rem->retry_bitfield &= ~(1 << i); ++ ++ mptcp_init4_subsockets(meta_sk, &mptcp_local->locaddr4[i], rem); ++ goto next_subflow; ++ } ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_for_each_bit_set(mpcb->rem6_bits, i) { ++ struct mptcp_rem6 *rem = &mpcb->remaddr6[i]; ++ ++ /* Do we need to retry establishing a subflow ? */ ++ if (rem->retry_bitfield) { ++ int i = mptcp_find_free_index(~rem->retry_bitfield); ++ ++ rem->bitfield |= (1 << i); ++ rem->retry_bitfield &= ~(1 << i); ++ ++ mptcp_init6_subsockets(meta_sk, &mptcp_local->locaddr6[i], rem); ++ goto next_subflow; ++ } ++ } ++#endif ++ ++exit: ++ kfree(mptcp_local); ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ sock_put(meta_sk); ++} ++ ++/** ++ * Create all new subflows, by doing calls to mptcp_initX_subsockets ++ * ++ * This function uses a goto next_subflow, to allow releasing the lock between ++ * new subflows and giving other processes a chance to do some work on the ++ * socket and potentially finishing the communication. ++ **/ ++static void create_subflow_worker(struct work_struct *work) ++{ ++ struct fullmesh_priv *pm_priv = container_of(work, ++ struct fullmesh_priv, ++ subflow_work); ++ struct mptcp_cb *mpcb = pm_priv->mpcb; ++ struct sock *meta_sk = mpcb->meta_sk; ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(sock_net(meta_sk)); ++ int iter = 0, retry = 0; ++ int i; ++ ++ /* We need a local (stable) copy of the address-list. Really, it is not ++ * such a big deal, if the address-list is not 100% up-to-date. ++ */ ++ rcu_read_lock_bh(); ++ mptcp_local = rcu_dereference_bh(fm_ns->local); ++ mptcp_local = kmemdup(mptcp_local, sizeof(*mptcp_local), GFP_ATOMIC); ++ rcu_read_unlock_bh(); ++ ++ if (!mptcp_local) ++ return; ++ ++next_subflow: ++ if (iter) { ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ ++ yield(); ++ } ++ mutex_lock(&mpcb->mpcb_mutex); ++ lock_sock_nested(meta_sk, SINGLE_DEPTH_NESTING); ++ ++ iter++; ++ ++ if (sock_flag(meta_sk, SOCK_DEAD)) ++ goto exit; ++ ++ if (mpcb->master_sk && ++ !tcp_sk(mpcb->master_sk)->mptcp->fully_established) ++ goto exit; ++ ++ mptcp_for_each_bit_set(mpcb->rem4_bits, i) { ++ struct mptcp_rem4 *rem; ++ u8 remaining_bits; ++ ++ rem = &mpcb->remaddr4[i]; ++ remaining_bits = ~(rem->bitfield) & mptcp_local->loc4_bits; ++ ++ /* Are there still combinations to handle? */ ++ if (remaining_bits) { ++ int i = mptcp_find_free_index(~remaining_bits); ++ ++ rem->bitfield |= (1 << i); ++ ++ /* If a route is not yet available then retry once */ ++ if (mptcp_init4_subsockets(meta_sk, &mptcp_local->locaddr4[i], ++ rem) == -ENETUNREACH) ++ retry = rem->retry_bitfield |= (1 << i); ++ goto next_subflow; ++ } ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_for_each_bit_set(mpcb->rem6_bits, i) { ++ struct mptcp_rem6 *rem; ++ u8 remaining_bits; ++ ++ rem = &mpcb->remaddr6[i]; ++ remaining_bits = ~(rem->bitfield) & mptcp_local->loc6_bits; ++ ++ /* Are there still combinations to handle? */ ++ if (remaining_bits) { ++ int i = mptcp_find_free_index(~remaining_bits); ++ ++ rem->bitfield |= (1 << i); ++ ++ /* If a route is not yet available then retry once */ ++ if (mptcp_init6_subsockets(meta_sk, &mptcp_local->locaddr6[i], ++ rem) == -ENETUNREACH) ++ retry = rem->retry_bitfield |= (1 << i); ++ goto next_subflow; ++ } ++ } ++#endif ++ ++ if (retry && !delayed_work_pending(&pm_priv->subflow_retry_work)) { ++ sock_hold(meta_sk); ++ queue_delayed_work(mptcp_wq, &pm_priv->subflow_retry_work, ++ msecs_to_jiffies(MPTCP_SUBFLOW_RETRY_DELAY)); ++ } ++ ++exit: ++ kfree(mptcp_local); ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ sock_put(meta_sk); ++} ++ ++static void update_remove_addrs(u8 addr_id, struct sock *meta_sk, ++ struct mptcp_loc_addr *mptcp_local) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct fullmesh_priv *fmp = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ struct sock *sk; ++ int i; ++ ++ fmp->remove_addrs |= (1 << addr_id); ++ /* v4 goes from 0 to MPTCP_MAX_ADDR, v6 beyond */ ++ if (addr_id < MPTCP_MAX_ADDR) { ++ fmp->announced_addrs_v4 &= ~(1 << addr_id); ++ ++ mptcp_for_each_bit_set(mpcb->rem4_bits, i) { ++ mpcb->remaddr4[i].bitfield &= mptcp_local->loc4_bits; ++ mpcb->remaddr4[i].retry_bitfield &= mptcp_local->loc4_bits; ++ } ++ } else { ++ fmp->announced_addrs_v6 &= ~(1 << (addr_id - MPTCP_MAX_ADDR)); ++ ++ mptcp_for_each_bit_set(mpcb->rem6_bits, i) { ++ mpcb->remaddr6[i].bitfield &= mptcp_local->loc6_bits; ++ mpcb->remaddr6[i].retry_bitfield &= mptcp_local->loc6_bits; ++ } ++ } ++ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ if (sk) ++ tcp_send_ack(sk); ++} ++ ++static int mptcp_find_address(struct mptcp_loc_addr *mptcp_local, ++ sa_family_t family, union inet_addr *addr) ++{ ++ int i; ++ u8 loc_bits; ++ bool found = false; ++ ++ if (family == AF_INET) ++ loc_bits = mptcp_local->loc4_bits; ++ else ++ loc_bits = mptcp_local->loc6_bits; ++ ++ mptcp_for_each_bit_set(loc_bits, i) { ++ if (family == AF_INET && ++ mptcp_local->locaddr4[i].addr.s_addr == addr->in.s_addr) { ++ found = true; ++ break; ++ } ++ if (family == AF_INET6 && ++ ipv6_addr_equal(&mptcp_local->locaddr6[i].addr, ++ &addr->in6)) { ++ found = true; ++ break; ++ } ++ } ++ ++ if (!found) ++ return -1; ++ ++ return i; ++} ++ ++static void mptcp_address_worker(struct work_struct *work) ++{ ++ struct delayed_work *delayed_work = container_of(work, ++ struct delayed_work, ++ work); ++ struct mptcp_fm_ns *fm_ns = container_of(delayed_work, ++ struct mptcp_fm_ns, ++ address_worker); ++ struct net *net = fm_ns->net; ++ struct mptcp_addr_event *event = NULL; ++ struct mptcp_loc_addr *mptcp_local, *old; ++ int i, id = -1; /* id is used in the socket-code on a delete-event */ ++ bool success; /* Used to indicate if we succeeded handling the event */ ++ ++next_event: ++ success = false; ++ kfree(event); ++ ++ /* First, let's dequeue an event from our event-list */ ++ rcu_read_lock_bh(); ++ spin_lock(&fm_ns->local_lock); ++ ++ event = list_first_entry_or_null(&fm_ns->events, ++ struct mptcp_addr_event, list); ++ if (!event) { ++ spin_unlock(&fm_ns->local_lock); ++ rcu_read_unlock_bh(); ++ return; ++ } ++ ++ list_del(&event->list); ++ ++ mptcp_local = rcu_dereference_bh(fm_ns->local); ++ ++ if (event->code == MPTCP_EVENT_DEL) { ++ id = mptcp_find_address(mptcp_local, event->family, &event->addr); ++ ++ /* Not in the list - so we don't care */ ++ if (id < 0) ++ goto duno; ++ ++ old = mptcp_local; ++ mptcp_local = kmemdup(mptcp_local, sizeof(*mptcp_local), ++ GFP_ATOMIC); ++ if (!mptcp_local) ++ goto duno; ++ ++ if (event->family == AF_INET) ++ mptcp_local->loc4_bits &= ~(1 << id); ++ else ++ mptcp_local->loc6_bits &= ~(1 << id); ++ ++ rcu_assign_pointer(fm_ns->local, mptcp_local); ++ kfree(old); ++ } else { ++ int i = mptcp_find_address(mptcp_local, event->family, &event->addr); ++ int j = i; ++ ++ if (j < 0) { ++ /* Not in the list, so we have to find an empty slot */ ++ if (event->family == AF_INET) ++ i = __mptcp_find_free_index(mptcp_local->loc4_bits, -1, ++ mptcp_local->next_v4_index); ++ if (event->family == AF_INET6) ++ i = __mptcp_find_free_index(mptcp_local->loc6_bits, -1, ++ mptcp_local->next_v6_index); ++ ++ if (i < 0) { ++ mptcp_debug("%s no more space\n", __func__); ++ goto duno; ++ } ++ ++ /* It might have been a MOD-event. */ ++ event->code = MPTCP_EVENT_ADD; ++ } else { ++ /* Let's check if anything changes */ ++ if (event->family == AF_INET && ++ event->low_prio == mptcp_local->locaddr4[i].low_prio) ++ goto duno; ++ ++ if (event->family == AF_INET6 && ++ event->low_prio == mptcp_local->locaddr6[i].low_prio) ++ goto duno; ++ } ++ ++ old = mptcp_local; ++ mptcp_local = kmemdup(mptcp_local, sizeof(*mptcp_local), ++ GFP_ATOMIC); ++ if (!mptcp_local) ++ goto duno; ++ ++ if (event->family == AF_INET) { ++ mptcp_local->locaddr4[i].addr.s_addr = event->addr.in.s_addr; ++ mptcp_local->locaddr4[i].loc4_id = i + 1; ++ mptcp_local->locaddr4[i].low_prio = event->low_prio; ++ } else { ++ mptcp_local->locaddr6[i].addr = event->addr.in6; ++ mptcp_local->locaddr6[i].loc6_id = i + MPTCP_MAX_ADDR; ++ mptcp_local->locaddr6[i].low_prio = event->low_prio; ++ } ++ ++ if (j < 0) { ++ if (event->family == AF_INET) { ++ mptcp_local->loc4_bits |= (1 << i); ++ mptcp_local->next_v4_index = i + 1; ++ } else { ++ mptcp_local->loc6_bits |= (1 << i); ++ mptcp_local->next_v6_index = i + 1; ++ } ++ } ++ ++ rcu_assign_pointer(fm_ns->local, mptcp_local); ++ kfree(old); ++ } ++ success = true; ++ ++duno: ++ spin_unlock(&fm_ns->local_lock); ++ rcu_read_unlock_bh(); ++ ++ if (!success) ++ goto next_event; ++ ++ /* Now we iterate over the MPTCP-sockets and apply the event. */ ++ for (i = 0; i < MPTCP_HASH_SIZE; i++) { ++ struct hlist_nulls_node *node; ++ struct tcp_sock *meta_tp; ++ ++ rcu_read_lock_bh(); ++ hlist_nulls_for_each_entry_rcu(meta_tp, node, &tk_hashtable[i], ++ tk_table) { ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sock *meta_sk = (struct sock *)meta_tp, *sk; ++ struct fullmesh_priv *fmp = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ ++ if (sock_net(meta_sk) != net) ++ continue; ++ ++ if (unlikely(!atomic_inc_not_zero(&meta_sk->sk_refcnt))) ++ continue; ++ ++ bh_lock_sock(meta_sk); ++ ++ if (!meta_tp->mpc || !is_meta_sk(meta_sk) || ++ mpcb->infinite_mapping_snd || ++ mpcb->infinite_mapping_rcv || ++ mpcb->send_infinite_mapping) ++ goto next; ++ ++ /* May be that the pm has changed in-between */ ++ if (mpcb->pm_ops != &full_mesh) ++ goto next; ++ ++ if (sock_owned_by_user(meta_sk)) { ++ if (!test_and_set_bit(MPTCP_PATH_MANAGER, ++ &meta_tp->tsq_flags)) ++ sock_hold(meta_sk); ++ ++ goto next; ++ } ++ ++ if (event->code == MPTCP_EVENT_ADD) { ++ if (event->family == AF_INET) ++ fmp->add_addr++; ++#if IS_ENABLED(CONFIG_IPV6) ++ if (event->family == AF_INET6) ++ fmp->add_addr++; ++#endif ++ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ if (sk) ++ tcp_send_ack(sk); ++ ++ full_mesh_create_subflows(meta_sk); ++ } ++ ++ if (event->code == MPTCP_EVENT_DEL) { ++ struct sock *sk, *tmpsk; ++ struct mptcp_loc_addr *mptcp_local; ++ bool found = false; ++ ++ mptcp_local = rcu_dereference_bh(fm_ns->local); ++ ++ /* Look for the socket and remove him */ ++ mptcp_for_each_sk_safe(mpcb, sk, tmpsk) { ++ if ((event->family == AF_INET6 && ++ (sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(sk))) || ++ (event->family == AF_INET && ++ (sk->sk_family == AF_INET6 && ++ !mptcp_v6_is_v4_mapped(sk)))) ++ continue; ++ ++ if (event->family == AF_INET && ++ (sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(sk)) && ++ inet_sk(sk)->inet_saddr != event->addr.in.s_addr) ++ continue; ++ ++ if (event->family == AF_INET6 && ++ sk->sk_family == AF_INET6 && ++ !ipv6_addr_equal(&inet6_sk(sk)->saddr, &event->addr.in6)) ++ continue; ++ ++ /* Reinject, so that pf = 1 and so we ++ * won't select this one as the ++ * ack-sock. ++ */ ++ mptcp_reinject_data(sk, 0); ++ ++ /* A master is special, it has ++ * address-id 0 ++ */ ++ if (!tcp_sk(sk)->mptcp->loc_id) ++ update_remove_addrs(0, meta_sk, mptcp_local); ++ else if (tcp_sk(sk)->mptcp->loc_id != id) ++ update_remove_addrs(tcp_sk(sk)->mptcp->loc_id, meta_sk, mptcp_local); ++ ++ mptcp_sub_force_close(sk); ++ found = true; ++ } ++ ++ if (!found) ++ goto next; ++ ++ /* The id may have been given by the event, ++ * matching on a local address. And it may not ++ * have matched on one of the above sockets, ++ * because the client never created a subflow. ++ * So, we have to finally remove it here. ++ */ ++ if (id > 0) ++ update_remove_addrs(id, meta_sk, mptcp_local); ++ } ++ ++ if (event->code == MPTCP_EVENT_MOD) { ++ struct sock *sk; ++ ++ mptcp_for_each_sk(mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ if (event->family == AF_INET && ++ (sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(sk)) && ++ inet_sk(sk)->inet_saddr == event->addr.in.s_addr) { ++ if (event->low_prio != tp->mptcp->low_prio) { ++ tp->mptcp->send_mp_prio = 1; ++ tp->mptcp->low_prio = event->low_prio; ++ ++ tcp_send_ack(sk); ++ } ++ } ++ ++ if (event->family == AF_INET6 && ++ sk->sk_family == AF_INET6 && ++ !ipv6_addr_equal(&inet6_sk(sk)->saddr, &event->addr.in6)) { ++ if (event->low_prio != tp->mptcp->low_prio) { ++ tp->mptcp->send_mp_prio = 1; ++ tp->mptcp->low_prio = event->low_prio; ++ ++ tcp_send_ack(sk); ++ } ++ } ++ } ++ } ++next: ++ bh_unlock_sock(meta_sk); ++ sock_put(meta_sk); ++ } ++ rcu_read_unlock_bh(); ++ } ++ goto next_event; ++} ++ ++static struct mptcp_addr_event *lookup_similar_event(struct net *net, ++ struct mptcp_addr_event *event) ++{ ++ struct mptcp_addr_event *eventq; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ ++ list_for_each_entry(eventq, &fm_ns->events, list) { ++ if (eventq->family != event->family) ++ continue; ++ if (event->family == AF_INET) { ++ if (eventq->addr.in.s_addr == event->addr.in.s_addr) ++ return eventq; ++ } else { ++ if (ipv6_addr_equal(&eventq->addr.in6, &event->addr.in6)) ++ return eventq; ++ } ++ } ++ return NULL; ++} ++ ++/* We already hold the net-namespace MPTCP-lock */ ++static void add_pm_event(struct net *net, struct mptcp_addr_event *event) ++{ ++ struct mptcp_addr_event *eventq = lookup_similar_event(net, event); ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ ++ if (eventq) { ++ switch (event->code) { ++ case MPTCP_EVENT_DEL: ++ list_del(&eventq->list); ++ kfree(eventq); ++ break; ++ case MPTCP_EVENT_ADD: ++ eventq->low_prio = event->low_prio; ++ eventq->code = MPTCP_EVENT_ADD; ++ return; ++ case MPTCP_EVENT_MOD: ++ eventq->low_prio = event->low_prio; ++ return; ++ } ++ } ++ ++ /* OK, we have to add the new address to the wait queue */ ++ eventq = kmemdup(event, sizeof(struct mptcp_addr_event), GFP_ATOMIC); ++ if (!eventq) ++ return; ++ ++ list_add_tail(&eventq->list, &fm_ns->events); ++ ++ /* Create work-queue */ ++ if (!delayed_work_pending(&fm_ns->address_worker)) ++ queue_delayed_work(mptcp_wq, &fm_ns->address_worker, ++ msecs_to_jiffies(500)); ++} ++ ++static void addr4_event_handler(struct in_ifaddr *ifa, unsigned long event, ++ struct net *net) ++{ ++ struct net_device *netdev = ifa->ifa_dev->dev; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ struct mptcp_addr_event mpevent; ++ ++ if (ifa->ifa_scope > RT_SCOPE_LINK || ++ ipv4_is_loopback(ifa->ifa_local)) ++ return; ++ ++ spin_lock_bh(&fm_ns->local_lock); ++ ++ mpevent.family = AF_INET; ++ mpevent.addr.in.s_addr = ifa->ifa_local; ++ mpevent.low_prio = (netdev->flags & IFF_MPBACKUP) ? 1 : 0; ++ ++ if (event == NETDEV_DOWN || !netif_running(netdev) || ++ (netdev->flags & IFF_NOMULTIPATH)) ++ mpevent.code = MPTCP_EVENT_DEL; ++ else if (event == NETDEV_UP) ++ mpevent.code = MPTCP_EVENT_ADD; ++ else if (event == NETDEV_CHANGE) ++ mpevent.code = MPTCP_EVENT_MOD; ++ ++ add_pm_event(net, &mpevent); ++ ++ spin_unlock_bh(&fm_ns->local_lock); ++ return; ++} ++ ++/* React on IPv4-addr add/rem-events */ ++static int mptcp_pm_inetaddr_event(struct notifier_block *this, ++ unsigned long event, void *ptr) ++{ ++ struct in_ifaddr *ifa = (struct in_ifaddr *)ptr; ++ struct net *net = dev_net(ifa->ifa_dev->dev); ++ ++ addr4_event_handler(ifa, event, net); ++ ++ return NOTIFY_DONE; ++} ++ ++static struct notifier_block mptcp_pm_inetaddr_notifier = { ++ .notifier_call = mptcp_pm_inetaddr_event, ++}; ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ ++/* IPV6-related address/interface watchers */ ++struct mptcp_dad_data { ++ struct timer_list timer; ++ struct inet6_ifaddr *ifa; ++}; ++ ++static void dad_callback(unsigned long arg); ++static int inet6_addr_event(struct notifier_block *this, ++ unsigned long event, void *ptr); ++ ++static int ipv6_is_in_dad_state(struct inet6_ifaddr *ifa) ++{ ++ return ((ifa->flags & IFA_F_TENTATIVE) && ++ ifa->state == INET6_IFADDR_STATE_DAD); ++} ++ ++static void dad_init_timer(struct mptcp_dad_data *data, ++ struct inet6_ifaddr *ifa) ++{ ++ data->ifa = ifa; ++ data->timer.data = (unsigned long)data; ++ data->timer.function = dad_callback; ++ if (ifa->idev->cnf.rtr_solicit_delay) ++ data->timer.expires = jiffies + ifa->idev->cnf.rtr_solicit_delay; ++ else ++ data->timer.expires = jiffies + (HZ/10); ++} ++ ++static void dad_callback(unsigned long arg) ++{ ++ struct mptcp_dad_data *data = (struct mptcp_dad_data *)arg; ++ ++ if (ipv6_is_in_dad_state(data->ifa)) { ++ dad_init_timer(data, data->ifa); ++ add_timer(&data->timer); ++ } else { ++ inet6_addr_event(NULL, NETDEV_UP, data->ifa); ++ in6_ifa_put(data->ifa); ++ kfree(data); ++ } ++} ++ ++static inline void dad_setup_timer(struct inet6_ifaddr *ifa) ++{ ++ struct mptcp_dad_data *data; ++ ++ data = kmalloc(sizeof(*data), GFP_ATOMIC); ++ ++ if (!data) ++ return; ++ ++ init_timer(&data->timer); ++ dad_init_timer(data, ifa); ++ add_timer(&data->timer); ++ in6_ifa_hold(ifa); ++} ++ ++static void addr6_event_handler(struct inet6_ifaddr *ifa, unsigned long event, ++ struct net *net) ++{ ++ struct net_device *netdev = ifa->idev->dev; ++ int addr_type = ipv6_addr_type(&ifa->addr); ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ struct mptcp_addr_event mpevent; ++ ++ if (ifa->scope > RT_SCOPE_LINK || ++ addr_type == IPV6_ADDR_ANY || ++ (addr_type & IPV6_ADDR_LOOPBACK) || ++ (addr_type & IPV6_ADDR_LINKLOCAL)) ++ return; ++ ++ spin_lock_bh(&fm_ns->local_lock); ++ ++ mpevent.family = AF_INET6; ++ mpevent.addr.in6 = ifa->addr; ++ mpevent.low_prio = (netdev->flags & IFF_MPBACKUP) ? 1 : 0; ++ ++ if (event == NETDEV_DOWN ||!netif_running(netdev) || ++ (netdev->flags & IFF_NOMULTIPATH)) ++ mpevent.code = MPTCP_EVENT_DEL; ++ else if (event == NETDEV_UP) ++ mpevent.code = MPTCP_EVENT_ADD; ++ else if (event == NETDEV_CHANGE) ++ mpevent.code = MPTCP_EVENT_MOD; ++ ++ add_pm_event(net, &mpevent); ++ ++ spin_unlock_bh(&fm_ns->local_lock); ++ return; ++} ++ ++/* React on IPv6-addr add/rem-events */ ++static int inet6_addr_event(struct notifier_block *this, unsigned long event, ++ void *ptr) ++{ ++ struct inet6_ifaddr *ifa6 = (struct inet6_ifaddr *)ptr; ++ struct net *net = dev_net(ifa6->idev->dev); ++ ++ if (ipv6_is_in_dad_state(ifa6)) ++ dad_setup_timer(ifa6); ++ else ++ addr6_event_handler(ifa6, event, net); ++ ++ return NOTIFY_DONE; ++} ++ ++static struct notifier_block inet6_addr_notifier = { ++ .notifier_call = inet6_addr_event, ++}; ++ ++#endif ++ ++/* React on ifup/down-events */ ++static int netdev_event(struct notifier_block *this, unsigned long event, ++ void *ptr) ++{ ++ struct net_device *dev = netdev_notifier_info_to_dev(ptr); ++ struct in_device *in_dev; ++#if IS_ENABLED(CONFIG_IPV6) ++ struct inet6_dev *in6_dev; ++#endif ++ ++ if (!(event == NETDEV_UP || event == NETDEV_DOWN || ++ event == NETDEV_CHANGE)) ++ return NOTIFY_DONE; ++ ++ rcu_read_lock(); ++ in_dev = __in_dev_get_rtnl(dev); ++ ++ if (in_dev) { ++ for_ifa(in_dev) { ++ mptcp_pm_inetaddr_event(NULL, event, ifa); ++ } endfor_ifa(in_dev); ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ in6_dev = __in6_dev_get(dev); ++ ++ if (in6_dev) { ++ struct inet6_ifaddr *ifa6; ++ list_for_each_entry(ifa6, &in6_dev->addr_list, if_list) ++ inet6_addr_event(NULL, event, ifa6); ++ } ++#endif ++ ++ rcu_read_unlock(); ++ return NOTIFY_DONE; ++} ++ ++static struct notifier_block mptcp_pm_netdev_notifier = { ++ .notifier_call = netdev_event, ++}; ++ ++static void full_mesh_new_session(struct sock *meta_sk, int index) ++{ ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct fullmesh_priv *fmp = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ struct net *net = sock_net(meta_sk); ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ struct sock *sk; ++ int i; ++ ++ if (index == -1) { ++ mptcp_fallback_default(mpcb); ++ return; ++ } ++ ++ /* Initialize workqueue-struct */ ++ INIT_WORK(&fmp->subflow_work, create_subflow_worker); ++ INIT_DELAYED_WORK(&fmp->subflow_retry_work, retry_subflow_worker); ++ fmp->mpcb = mpcb; ++ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ ++ rcu_read_lock(); ++ mptcp_local = rcu_dereference(fm_ns->local); ++ ++ /* Look for the address among the local addresses */ ++ mptcp_for_each_bit_set(mptcp_local->loc4_bits, i) { ++ __be32 ifa_address = mptcp_local->locaddr4[i].addr.s_addr; ++ ++ /* We do not need to announce the initial subflow's address again */ ++ if ((meta_sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(meta_sk)) && ++ inet_sk(meta_sk)->inet_saddr == ifa_address) ++ continue; ++ ++ fmp->add_addr++; ++ ++ if (sk) ++ tcp_send_ack(sk); ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_for_each_bit_set(mptcp_local->loc6_bits, i) { ++ struct in6_addr *ifa6 = &mptcp_local->locaddr6[i].addr; ++ ++ /* We do not need to announce the initial subflow's address again */ ++ if (meta_sk->sk_family == AF_INET6 && ++ ipv6_addr_equal(&inet6_sk(meta_sk)->saddr, ifa6)) ++ continue; ++ ++ fmp->add_addr++; ++ ++ if (sk) ++ tcp_send_ack(sk); ++ } ++#endif ++ ++ rcu_read_unlock(); ++ ++ if (meta_sk->sk_family == AF_INET || mptcp_v6_is_v4_mapped(meta_sk)) ++ fmp->announced_addrs_v4 |= (1 << index); ++ else ++ fmp->announced_addrs_v6 |= (1 << index); ++} ++ ++static void full_mesh_create_subflows(struct sock *meta_sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct fullmesh_priv *pm_priv = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ ++ if (mpcb->infinite_mapping_snd || mpcb->infinite_mapping_rcv || ++ mpcb->send_infinite_mapping || ++ mpcb->server_side || sock_flag(meta_sk, SOCK_DEAD)) ++ return; ++ ++ /* The master may not yet be fully established (address added through ++ * mptcp_update_metasocket). Then, we should not attempt to create new ++ * subflows. ++ */ ++ if (mpcb->master_sk && ++ !tcp_sk(mpcb->master_sk)->mptcp->fully_established) ++ return; ++ ++ if (!work_pending(&pm_priv->subflow_work)) { ++ sock_hold(meta_sk); ++ queue_work(mptcp_wq, &pm_priv->subflow_work); ++ } ++} ++ ++/* Called upon release_sock, if the socket was owned by the user during ++ * a path-management event. ++ */ ++static void full_mesh_release_sock(struct sock *meta_sk) ++{ ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct fullmesh_priv *fmp = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(sock_net(meta_sk)); ++ struct sock *sk, *tmpsk; ++ int i; ++ ++ rcu_read_lock(); ++ mptcp_local = rcu_dereference(fm_ns->local); ++ ++ /* First, detect modifications or additions */ ++ mptcp_for_each_bit_set(mptcp_local->loc4_bits, i) { ++ struct in_addr ifa = mptcp_local->locaddr4[i].addr; ++ bool found = false; ++ ++ mptcp_for_each_sk(mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ if (sk->sk_family == AF_INET6 && ++ !mptcp_v6_is_v4_mapped(sk)) ++ continue; ++ ++ if (inet_sk(sk)->inet_saddr != ifa.s_addr) ++ continue; ++ ++ found = true; ++ ++ if (mptcp_local->locaddr4[i].low_prio != tp->mptcp->low_prio) { ++ tp->mptcp->send_mp_prio = 1; ++ tp->mptcp->low_prio = mptcp_local->locaddr4[i].low_prio; ++ ++ tcp_send_ack(sk); ++ } ++ } ++ ++ if (!found) { ++ fmp->add_addr++; ++ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ if (sk) ++ tcp_send_ack(sk); ++ full_mesh_create_subflows(meta_sk); ++ } ++ } ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ mptcp_for_each_bit_set(mptcp_local->loc6_bits, i) { ++ struct in6_addr ifa = mptcp_local->locaddr6[i].addr; ++ bool found = false; ++ ++ mptcp_for_each_sk(mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ if (sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(sk)) ++ continue; ++ ++ if (!ipv6_addr_equal(&inet6_sk(sk)->saddr, &ifa)) ++ continue; ++ ++ found = true; ++ ++ if (mptcp_local->locaddr6[i].low_prio != tp->mptcp->low_prio) { ++ tp->mptcp->send_mp_prio = 1; ++ tp->mptcp->low_prio = mptcp_local->locaddr6[i].low_prio; ++ ++ tcp_send_ack(sk); ++ } ++ } ++ ++ if (!found) { ++ fmp->add_addr++; ++ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ if (sk) ++ tcp_send_ack(sk); ++ full_mesh_create_subflows(meta_sk); ++ } ++ } ++#endif ++ ++ /* Now, detect address-removals */ ++ mptcp_for_each_sk_safe(mpcb, sk, tmpsk) { ++ bool shall_remove = true; ++ ++ if (sk->sk_family == AF_INET || mptcp_v6_is_v4_mapped(sk)) { ++ mptcp_for_each_bit_set(mptcp_local->loc4_bits, i) { ++ if (inet_sk(sk)->inet_saddr == mptcp_local->locaddr4[i].addr.s_addr) { ++ shall_remove = false; ++ break; ++ } ++ } ++ } else { ++ mptcp_for_each_bit_set(mptcp_local->loc6_bits, i) { ++ if (ipv6_addr_equal(&inet6_sk(sk)->saddr, &mptcp_local->locaddr6[i].addr)) { ++ shall_remove = false; ++ break; ++ } ++ } ++ } ++ ++ if (shall_remove) { ++ /* Reinject, so that pf = 1 and so we ++ * won't select this one as the ++ * ack-sock. ++ */ ++ mptcp_reinject_data(sk, 0); ++ ++ update_remove_addrs(tcp_sk(sk)->mptcp->loc_id, meta_sk, ++ mptcp_local); ++ ++ if (mpcb->master_sk == sk) ++ update_remove_addrs(0, meta_sk, mptcp_local); ++ ++ mptcp_sub_force_close(sk); ++ } ++ } ++ rcu_read_unlock(); ++} ++ ++static int full_mesh_get_local_index(sa_family_t family, union inet_addr *addr, ++ struct net *net) ++{ ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ int index; ++ ++ /* Handle the backup-flows */ ++ rcu_read_lock(); ++ mptcp_local = rcu_dereference(fm_ns->local); ++ ++ index = mptcp_find_address(mptcp_local, family, addr); ++ ++ rcu_read_unlock(); ++ ++ return index; ++} ++ ++static int full_mesh_get_local_id(sa_family_t family, union inet_addr *addr, ++ struct net *net) ++{ ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(net); ++ int index, id = -1; ++ ++ /* Handle the backup-flows */ ++ rcu_read_lock(); ++ mptcp_local = rcu_dereference(fm_ns->local); ++ ++ index = mptcp_find_address(mptcp_local, family, addr); ++ ++ if (index != -1) { ++ if (family == AF_INET) ++ id = mptcp_local->locaddr4[index].loc4_id; ++ else ++ id = mptcp_local->locaddr6[index].loc6_id; ++ } ++ ++ ++ rcu_read_unlock(); ++ ++ return id; ++} ++ ++static void full_mesh_addr_signal(struct sock *sk, unsigned *size, ++ struct tcp_out_options *opts, ++ struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct fullmesh_priv *fmp = (struct fullmesh_priv *)&mpcb->mptcp_pm[0]; ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns = fm_get_ns(sock_net(sk)); ++ int remove_addr_len; ++ u8 unannouncedv4, unannouncedv6; ++ ++ if (likely(!fmp->add_addr)) ++ goto remove_addr; ++ ++ rcu_read_lock(); ++ mptcp_local = rcu_dereference(fm_ns->local); ++ ++ /* IPv4 */ ++ unannouncedv4 = (~fmp->announced_addrs_v4) & mptcp_local->loc4_bits; ++ if (unannouncedv4 && ++ MAX_TCP_OPTION_SPACE - *size >= MPTCP_SUB_LEN_ADD_ADDR4_ALIGN) { ++ int ind = mptcp_find_free_index(~unannouncedv4); ++ ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_ADD_ADDR; ++ opts->add_addr4.addr_id = mptcp_local->locaddr4[ind].loc4_id; ++ opts->add_addr4.addr = mptcp_local->locaddr4[ind].addr; ++ opts->add_addr_v4 = 1; ++ ++ if (skb) { ++ fmp->announced_addrs_v4 |= (1 << ind); ++ fmp->add_addr--; ++ } ++ *size += MPTCP_SUB_LEN_ADD_ADDR4_ALIGN; ++ } ++ ++ /* IPv6 */ ++ unannouncedv6 = (~fmp->announced_addrs_v6) & mptcp_local->loc6_bits; ++ if (unannouncedv6 && ++ MAX_TCP_OPTION_SPACE - *size >= MPTCP_SUB_LEN_ADD_ADDR6_ALIGN) { ++ int ind = mptcp_find_free_index(~unannouncedv6); ++ ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_ADD_ADDR; ++ opts->add_addr6.addr_id = mptcp_local->locaddr6[ind].loc6_id; ++ opts->add_addr6.addr = mptcp_local->locaddr6[ind].addr; ++ opts->add_addr_v6 = 1; ++ ++ if (skb) { ++ fmp->announced_addrs_v6 |= (1 << ind); ++ fmp->add_addr--; ++ } ++ *size += MPTCP_SUB_LEN_ADD_ADDR6_ALIGN; ++ } ++ ++ rcu_read_unlock(); ++ ++ if (!unannouncedv4 && !unannouncedv6 && skb) { ++ fmp->add_addr--; ++ } ++ ++remove_addr: ++ if (likely(!fmp->remove_addrs)) ++ return; ++ ++ remove_addr_len = mptcp_sub_len_remove_addr_align(fmp->remove_addrs); ++ if (MAX_TCP_OPTION_SPACE - *size < remove_addr_len) ++ return; ++ ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_REMOVE_ADDR; ++ opts->remove_addrs = fmp->remove_addrs; ++ *size += remove_addr_len; ++ if (skb) ++ fmp->remove_addrs = 0; ++} ++ ++static int mptcp_fm_init_net(struct net *net) ++{ ++ struct mptcp_loc_addr *mptcp_local; ++ struct mptcp_fm_ns *fm_ns; ++ ++ fm_ns = kzalloc(sizeof(*fm_ns), GFP_KERNEL); ++ if (!fm_ns) ++ return -ENOBUFS; ++ ++ mptcp_local = kzalloc(sizeof(*mptcp_local), GFP_KERNEL); ++ if (!mptcp_local) { ++ kfree(fm_ns); ++ return -ENOBUFS; ++ } ++ ++ mptcp_local->next_v4_index = 1; ++ ++ rcu_assign_pointer(fm_ns->local, mptcp_local); ++ INIT_DELAYED_WORK(&fm_ns->address_worker, mptcp_address_worker); ++ INIT_LIST_HEAD(&fm_ns->events); ++ spin_lock_init(&fm_ns->local_lock); ++ fm_ns->net = net; ++ net->mptcp.path_managers[MPTCP_PM_FULLMESH] = fm_ns; ++ ++ return 0; ++} ++ ++static void mptcp_fm_exit_net(struct net *net) ++{ ++ struct mptcp_addr_event *eventq, *tmp; ++ struct mptcp_fm_ns *fm_ns; ++ struct mptcp_loc_addr *mptcp_local; ++ ++ fm_ns = fm_get_ns(net); ++ cancel_delayed_work_sync(&fm_ns->address_worker); ++ ++ rcu_read_lock_bh(); ++ ++ mptcp_local = rcu_dereference_bh(fm_ns->local); ++ kfree(mptcp_local); ++ ++ spin_lock(&fm_ns->local_lock); ++ list_for_each_entry_safe(eventq, tmp, &fm_ns->events, list) { ++ list_del(&eventq->list); ++ kfree(eventq); ++ } ++ spin_unlock(&fm_ns->local_lock); ++ ++ rcu_read_unlock_bh(); ++ ++ kfree(fm_ns); ++} ++ ++static struct pernet_operations full_mesh_net_ops = { ++ .init = mptcp_fm_init_net, ++ .exit = mptcp_fm_exit_net, ++}; ++ ++static struct mptcp_pm_ops full_mesh __read_mostly = { ++ .new_session = full_mesh_new_session, ++ .release_sock = full_mesh_release_sock, ++ .fully_established = full_mesh_create_subflows, ++ .new_remote_address = full_mesh_create_subflows, ++ .get_local_index = full_mesh_get_local_index, ++ .get_local_id = full_mesh_get_local_id, ++ .addr_signal = full_mesh_addr_signal, ++ .name = "fullmesh", ++ .owner = THIS_MODULE, ++}; ++ ++/* General initialization of MPTCP_PM */ ++static int __init full_mesh_register(void) ++{ ++ int ret; ++ ++ BUILD_BUG_ON(sizeof(struct fullmesh_priv) > MPTCP_PM_SIZE); ++ ++ ret = register_pernet_subsys(&full_mesh_net_ops); ++ if (ret) ++ goto out; ++ ++ ret = register_inetaddr_notifier(&mptcp_pm_inetaddr_notifier); ++ if (ret) ++ goto err_reg_inetaddr; ++ ret = register_netdevice_notifier(&mptcp_pm_netdev_notifier); ++ if (ret) ++ goto err_reg_netdev; ++ ++#if IS_ENABLED(CONFIG_IPV6) ++ ret = register_inet6addr_notifier(&inet6_addr_notifier); ++ if (ret) ++ goto err_reg_inet6addr; ++#endif ++ ++ ret = mptcp_register_path_manager(&full_mesh); ++ if (ret) ++ goto err_reg_pm; ++ ++out: ++ return ret; ++ ++ ++err_reg_pm: ++#if IS_ENABLED(CONFIG_IPV6) ++ unregister_inet6addr_notifier(&inet6_addr_notifier); ++err_reg_inet6addr: ++#endif ++ unregister_netdevice_notifier(&mptcp_pm_netdev_notifier); ++err_reg_netdev: ++ unregister_inetaddr_notifier(&mptcp_pm_inetaddr_notifier); ++err_reg_inetaddr: ++ unregister_pernet_subsys(&full_mesh_net_ops); ++ goto out; ++} ++ ++static void full_mesh_unregister(void) ++{ ++#if IS_ENABLED(CONFIG_IPV6) ++ unregister_inet6addr_notifier(&inet6_addr_notifier); ++#endif ++ unregister_netdevice_notifier(&mptcp_pm_netdev_notifier); ++ unregister_inetaddr_notifier(&mptcp_pm_inetaddr_notifier); ++ unregister_pernet_subsys(&full_mesh_net_ops); ++ mptcp_unregister_path_manager(&full_mesh); ++} ++ ++module_init(full_mesh_register); ++module_exit(full_mesh_unregister); ++ ++MODULE_AUTHOR("Christoph Paasch"); ++MODULE_LICENSE("GPL"); ++MODULE_DESCRIPTION("Full-Mesh MPTCP"); ++MODULE_VERSION("0.88"); +diff --git a/net/mptcp/mptcp_input.c b/net/mptcp/mptcp_input.c +new file mode 100644 +index 0000000..f3c9057 +--- /dev/null ++++ b/net/mptcp/mptcp_input.c +@@ -0,0 +1,2254 @@ ++/* ++ * MPTCP implementation - Sending side ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#include <asm/unaligned.h> ++ ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#include <net/mptcp_v6.h> ++ ++#include <linux/kconfig.h> ++ ++/* is seq1 < seq2 ? */ ++static inline int before64(const u64 seq1, const u64 seq2) ++{ ++ return (s64)(seq1 - seq2) < 0; ++} ++ ++/* is seq1 > seq2 ? */ ++#define after64(seq1, seq2) before64(seq2, seq1) ++ ++static inline void mptcp_become_fully_estab(struct sock *sk) ++{ ++ tcp_sk(sk)->mptcp->fully_established = 1; ++ ++ if (is_master_tp(tcp_sk(sk)) && ++ tcp_sk(sk)->mpcb->pm_ops->fully_established) ++ tcp_sk(sk)->mpcb->pm_ops->fully_established(mptcp_meta_sk(sk)); ++} ++ ++/* Similar to tcp_tso_acked without any memory accounting */ ++static inline int mptcp_tso_acked_reinject(struct sock *sk, struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ u32 packets_acked, len; ++ ++ BUG_ON(!after(TCP_SKB_CB(skb)->end_seq, tp->snd_una)); ++ ++ packets_acked = tcp_skb_pcount(skb); ++ ++ if (skb_unclone(skb, GFP_ATOMIC)) ++ return 0; ++ ++ len = tp->snd_una - TCP_SKB_CB(skb)->seq; ++ __pskb_trim_head(skb, len); ++ ++ TCP_SKB_CB(skb)->seq += len; ++ skb->ip_summed = CHECKSUM_PARTIAL; ++ skb->truesize -= len; ++ ++ /* Any change of skb->len requires recalculation of tso factor. */ ++ if (tcp_skb_pcount(skb) > 1) ++ tcp_set_skb_tso_segs(sk, skb, tcp_skb_mss(skb)); ++ packets_acked -= tcp_skb_pcount(skb); ++ ++ if (packets_acked) { ++ BUG_ON(tcp_skb_pcount(skb) == 0); ++ BUG_ON(!before(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq)); ++ } ++ ++ return packets_acked; ++} ++ ++/** ++ * Cleans the meta-socket retransmission queue and the reinject-queue. ++ * @sk must be the metasocket. ++ */ ++static void mptcp_clean_rtx_queue(struct sock *meta_sk, u32 prior_snd_una) ++{ ++ struct sk_buff *skb, *tmp; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ bool acked = false; ++ u32 acked_pcount; ++ ++ while ((skb = tcp_write_queue_head(meta_sk)) && ++ skb != tcp_send_head(meta_sk)) { ++ bool fully_acked = true; ++ ++ if (before(meta_tp->snd_una, TCP_SKB_CB(skb)->end_seq)) { ++ if (tcp_skb_pcount(skb) == 1 || ++ !after(meta_tp->snd_una, TCP_SKB_CB(skb)->seq)) ++ break; ++ ++ acked_pcount = tcp_tso_acked(meta_sk, skb); ++ if (!acked_pcount) ++ break; ++ ++ fully_acked = false; ++ } else { ++ acked_pcount = tcp_skb_pcount(skb); ++ } ++ ++ acked = true; ++ meta_tp->packets_out -= acked_pcount; ++ meta_tp->retrans_stamp = 0; ++ ++ if (!fully_acked) ++ break; ++ ++ tcp_unlink_write_queue(skb, meta_sk); ++ ++ if (mptcp_is_data_fin(skb)) { ++ struct sock *sk_it; ++ ++ /* DATA_FIN has been acknowledged - now we can close ++ * the subflows ++ */ ++ mptcp_for_each_sk(mpcb, sk_it) { ++ unsigned long delay = 0; ++ ++ /* If we are the passive closer, don't trigger ++ * subflow-fin until the subflow has been finned ++ * by the peer - thus we add a delay. ++ */ ++ if (mpcb->passive_close && ++ sk_it->sk_state == TCP_ESTABLISHED) ++ delay = inet_csk(sk_it)->icsk_rto << 3; ++ ++ mptcp_sub_close(sk_it, delay); ++ } ++ } ++ sk_wmem_free_skb(meta_sk, skb); ++ } ++ /* Remove acknowledged data from the reinject queue */ ++ skb_queue_walk_safe(&mpcb->reinject_queue, skb, tmp) { ++ if (before(meta_tp->snd_una, TCP_SKB_CB(skb)->end_seq)) { ++ if (tcp_skb_pcount(skb) == 1 || ++ !after(meta_tp->snd_una, TCP_SKB_CB(skb)->seq)) ++ break; ++ ++ mptcp_tso_acked_reinject(meta_sk, skb); ++ break; ++ } ++ ++ __skb_unlink(skb, &mpcb->reinject_queue); ++ __kfree_skb(skb); ++ } ++ ++ if (likely(between(meta_tp->snd_up, prior_snd_una, meta_tp->snd_una))) ++ meta_tp->snd_up = meta_tp->snd_una; ++ ++ if (acked) { ++ tcp_rearm_rto(meta_sk); ++ /* Normally this is done in tcp_try_undo_loss - but MPTCP ++ * does not call this function. ++ */ ++ inet_csk(meta_sk)->icsk_retransmits = 0; ++ } ++} ++ ++/* Inspired by tcp_rcv_state_process */ ++static int mptcp_rcv_state_process(struct sock *meta_sk, struct sock *sk, ++ const struct sk_buff *skb, u32 data_seq, ++ u16 data_len) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk), *tp = tcp_sk(sk); ++ struct tcphdr *th = tcp_hdr(skb); ++ ++ /* State-machine handling if FIN has been enqueued and he has ++ * been acked (snd_una == write_seq) - it's important that this ++ * here is after sk_wmem_free_skb because otherwise ++ * sk_forward_alloc is wrong upon inet_csk_destroy_sock() ++ */ ++ switch (meta_sk->sk_state) { ++ case TCP_FIN_WAIT1: ++ if (meta_tp->snd_una == meta_tp->write_seq) { ++ struct dst_entry *dst = __sk_dst_get(meta_sk); ++ ++ tcp_set_state(meta_sk, TCP_FIN_WAIT2); ++ meta_sk->sk_shutdown |= SEND_SHUTDOWN; ++ ++ dst = __sk_dst_get(sk); ++ if (dst) ++ dst_confirm(dst); ++ ++ if (!sock_flag(meta_sk, SOCK_DEAD)) { ++ /* Wake up lingering close() */ ++ meta_sk->sk_state_change(meta_sk); ++ } else { ++ int tmo; ++ ++ if (meta_tp->linger2 < 0 || ++ (data_len && ++ after(data_seq + data_len - (mptcp_is_data_fin2(skb, tp) ? 1 : 0), ++ meta_tp->rcv_nxt))) { ++ mptcp_send_active_reset(meta_sk, GFP_ATOMIC); ++ tcp_done(meta_sk); ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_TCPABORTONDATA); ++ return 1; ++ } ++ ++ tmo = tcp_fin_time(meta_sk); ++ if (tmo > TCP_TIMEWAIT_LEN) { ++ inet_csk_reset_keepalive_timer(meta_sk, tmo - TCP_TIMEWAIT_LEN); ++ } else if (mptcp_is_data_fin2(skb, tp) || ++ sock_owned_by_user(meta_sk)) { ++ /* Bad case. We could lose such FIN otherwise. ++ * It is not a big problem, but it looks confusing ++ * and not so rare event. We still can lose it now, ++ * if it spins in bh_lock_sock(), but it is really ++ * marginal case. ++ */ ++ inet_csk_reset_keepalive_timer(meta_sk, tmo); ++ } else { ++ tcp_time_wait(meta_sk, TCP_FIN_WAIT2, tmo); ++ } ++ } ++ } ++ break; ++ case TCP_CLOSING: ++ case TCP_LAST_ACK: ++ if (meta_tp->snd_una == meta_tp->write_seq) { ++ tcp_done(meta_sk); ++ return 1; ++ } ++ break; ++ } ++ ++ /* step 7: process the segment text */ ++ switch (meta_sk->sk_state) { ++ case TCP_FIN_WAIT1: ++ case TCP_FIN_WAIT2: ++ /* RFC 793 says to queue data in these states, ++ * RFC 1122 says we MUST send a reset. ++ * BSD 4.4 also does reset. ++ */ ++ if (meta_sk->sk_shutdown & RCV_SHUTDOWN) { ++ if (TCP_SKB_CB(skb)->end_seq != TCP_SKB_CB(skb)->seq && ++ after(TCP_SKB_CB(skb)->end_seq - th->fin, tp->rcv_nxt) && ++ !mptcp_is_data_fin2(skb, tp)) { ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_TCPABORTONDATA); ++ mptcp_send_active_reset(meta_sk, GFP_ATOMIC); ++ tcp_reset(meta_sk); ++ return 1; ++ } ++ } ++ break; ++ } ++ ++ return 0; ++} ++ ++/** ++ * @return: ++ * i) 1: Everything's fine. ++ * ii) -1: A reset has been sent on the subflow - csum-failure ++ * iii) 0: csum-failure but no reset sent, because it's the last subflow. ++ * Last packet should not be destroyed by the caller because it has ++ * been done here. ++ */ ++static int mptcp_verif_dss_csum(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sk_buff *tmp, *tmp1, *last = NULL; ++ __wsum csum_tcp = 0; /* cumulative checksum of pld + mptcp-header */ ++ int ans = 1, overflowed = 0, offset = 0, dss_csum_added = 0; ++ int iter = 0; ++ ++ skb_queue_walk_safe(&sk->sk_receive_queue, tmp, tmp1) { ++ unsigned int csum_len; ++ ++ if (before(tp->mptcp->map_subseq + tp->mptcp->map_data_len, TCP_SKB_CB(tmp)->end_seq)) ++ /* Mapping ends in the middle of the packet - ++ * csum only these bytes ++ */ ++ csum_len = tp->mptcp->map_subseq + tp->mptcp->map_data_len - TCP_SKB_CB(tmp)->seq; ++ else ++ csum_len = tmp->len; ++ ++ offset = 0; ++ if (overflowed) { ++ char first_word[4]; ++ first_word[0] = 0; ++ first_word[1] = 0; ++ first_word[2] = 0; ++ first_word[3] = *(tmp->data); ++ csum_tcp = csum_partial(first_word, 4, csum_tcp); ++ offset = 1; ++ csum_len--; ++ overflowed = 0; ++ } ++ ++ csum_tcp = skb_checksum(tmp, offset, csum_len, csum_tcp); ++ ++ /* Was it on an odd-length? Then we have to merge the next byte ++ * correctly (see above) ++ */ ++ if (csum_len != (csum_len & (~1))) ++ overflowed = 1; ++ ++ if (mptcp_is_data_seq(tmp) && !dss_csum_added) { ++ __be32 data_seq = htonl((u32)(tp->mptcp->map_data_seq >> 32)); ++ ++ /* If a 64-bit dss is present, we increase the offset ++ * by 4 bytes, as the high-order 64-bits will be added ++ * in the final csum_partial-call. ++ */ ++ u32 offset = skb_transport_offset(tmp) + ++ TCP_SKB_CB(tmp)->dss_off; ++ if (TCP_SKB_CB(tmp)->mptcp_flags & MPTCPHDR_SEQ64_SET) ++ offset += 4; ++ ++ csum_tcp = skb_checksum(tmp, offset, ++ MPTCP_SUB_LEN_SEQ_CSUM, ++ csum_tcp); ++ ++ csum_tcp = csum_partial(&data_seq, ++ sizeof(data_seq), csum_tcp); ++ ++ dss_csum_added = 1; /* Just do it once */ ++ } ++ last = tmp; ++ iter++; ++ ++ if (!skb_queue_is_last(&sk->sk_receive_queue, tmp) && ++ !before(TCP_SKB_CB(tmp1)->seq, ++ tp->mptcp->map_subseq + tp->mptcp->map_data_len)) ++ break; ++ } ++ ++ /* Now, checksum must be 0 */ ++ if (unlikely(csum_fold(csum_tcp))) { ++ pr_err("%s csum is wrong: %#x data_seq %u dss_csum_added %d overflowed %d iterations %d\n", ++ __func__, csum_fold(csum_tcp), ++ TCP_SKB_CB(last)->seq, dss_csum_added, overflowed, ++ iter); ++ ++ tp->mptcp->send_mp_fail = 1; ++ ++ /* map_data_seq is the data-seq number of the ++ * mapping we are currently checking ++ */ ++ tp->mpcb->csum_cutoff_seq = tp->mptcp->map_data_seq; ++ ++ if (tp->mpcb->cnt_subflows > 1) { ++ mptcp_send_reset(sk); ++ ans = -1; ++ } else { ++ tp->mpcb->send_infinite_mapping = 1; ++ ++ /* Need to purge the rcv-queue as it's no more valid */ ++ while ((tmp = __skb_dequeue(&sk->sk_receive_queue)) != NULL) { ++ tp->copied_seq = TCP_SKB_CB(tmp)->end_seq; ++ kfree_skb(tmp); ++ } ++ ++ ans = 0; ++ } ++ } ++ ++ return ans; ++} ++ ++static inline void mptcp_prepare_skb(struct sk_buff *skb, struct sk_buff *next, ++ struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); ++ /* Adapt data-seq's to the packet itself. We kinda transform the ++ * dss-mapping to a per-packet granularity. This is necessary to ++ * correctly handle overlapping mappings coming from different ++ * subflows. Otherwise it would be a complete mess. ++ */ ++ tcb->seq = ((u32)tp->mptcp->map_data_seq) + tcb->seq - tp->mptcp->map_subseq; ++ tcb->end_seq = tcb->seq + skb->len; ++ ++ /* If cur is the last one in the rcv-queue (or the last one for this ++ * mapping), and data_fin is enqueued, the end_data_seq is +1. ++ */ ++ if (skb_queue_is_last(&sk->sk_receive_queue, skb) || ++ after(TCP_SKB_CB(next)->end_seq, tp->mptcp->map_subseq + tp->mptcp->map_data_len)) { ++ tcb->end_seq += tp->mptcp->map_data_fin; ++ ++ /* We manually set the fin-flag if it is a data-fin. For easy ++ * processing in tcp_recvmsg. ++ */ ++ if (mptcp_is_data_fin2(skb, tp)) ++ tcp_hdr(skb)->fin = 1; ++ else ++ tcp_hdr(skb)->fin = 0; ++ } else { ++ /* We may have a subflow-fin with data but without data-fin */ ++ tcp_hdr(skb)->fin = 0; ++ } ++} ++ ++/** ++ * @return: 1 if the segment has been eaten and can be suppressed, ++ * otherwise 0. ++ */ ++static inline int mptcp_direct_copy(struct sk_buff *skb, struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ int chunk = min_t(unsigned int, skb->len, meta_tp->ucopy.len); ++ int eaten = 0; ++ ++ __set_current_state(TASK_RUNNING); ++ ++ local_bh_enable(); ++ if (!skb_copy_datagram_iovec(skb, 0, meta_tp->ucopy.iov, chunk)) { ++ meta_tp->ucopy.len -= chunk; ++ meta_tp->copied_seq += chunk; ++ eaten = (chunk == skb->len); ++ tcp_rcv_space_adjust(meta_sk); ++ } ++ local_bh_disable(); ++ return eaten; ++} ++ ++static inline void mptcp_reset_mapping(struct tcp_sock *tp) ++{ ++ tp->mptcp->map_data_len = 0; ++ tp->mptcp->map_data_seq = 0; ++ tp->mptcp->map_subseq = 0; ++ tp->mptcp->map_data_fin = 0; ++ tp->mptcp->mapping_present = 0; ++} ++ ++/* The DSS-mapping received on the sk only covers the second half of the skb ++ * (cut at seq). We trim the head from the skb. ++ * Data will be freed upon kfree(). ++ * ++ * Inspired by tcp_trim_head(). ++ */ ++static void mptcp_skb_trim_head(struct sk_buff *skb, struct sock *sk, u32 seq) ++{ ++ int len = seq - TCP_SKB_CB(skb)->seq; ++ u32 new_seq = TCP_SKB_CB(skb)->seq + len; ++ ++ if (len < skb_headlen(skb)) ++ __skb_pull(skb, len); ++ else ++ __pskb_trim_head(skb, len - skb_headlen(skb)); ++ ++ TCP_SKB_CB(skb)->seq = new_seq; ++ ++ skb->truesize -= len; ++ atomic_sub(len, &sk->sk_rmem_alloc); ++ sk_mem_uncharge(sk, len); ++} ++ ++/* The DSS-mapping received on the sk only covers the first half of the skb ++ * (cut at seq). We create a second skb (@return), and queue it in the rcv-queue ++ * as further packets may resolve the mapping of the second half of data. ++ * ++ * Inspired by tcp_fragment(). ++ */ ++static int mptcp_skb_split_tail(struct sk_buff *skb, struct sock *sk, u32 seq) ++{ ++ struct sk_buff *buff; ++ int nsize; ++ int nlen, len; ++ ++ len = seq - TCP_SKB_CB(skb)->seq; ++ nsize = skb_headlen(skb) - len + tcp_sk(sk)->tcp_header_len; ++ if (nsize < 0) ++ nsize = 0; ++ ++ /* Get a new skb... force flag on. */ ++ buff = alloc_skb(nsize, GFP_ATOMIC); ++ if (buff == NULL) ++ return -ENOMEM; ++ ++ skb_reserve(buff, tcp_sk(sk)->tcp_header_len); ++ skb_reset_transport_header(buff); ++ ++ tcp_hdr(buff)->fin = tcp_hdr(skb)->fin; ++ tcp_hdr(skb)->fin = 0; ++ ++ /* We absolutly need to call skb_set_owner_r before refreshing the ++ * truesize of buff, otherwise the moved data will account twice. ++ */ ++ skb_set_owner_r(buff, sk); ++ nlen = skb->len - len - nsize; ++ buff->truesize += nlen; ++ skb->truesize -= nlen; ++ ++ /* Correct the sequence numbers. */ ++ TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len; ++ TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq; ++ TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq; ++ ++ skb_split(skb, buff, len); ++ ++ __skb_queue_after(&sk->sk_receive_queue, skb, buff); ++ ++ return 0; ++} ++ ++/* @return: 0 everything is fine. Just continue processing ++ * 1 subflow is broken stop everything ++ * -1 this packet was broken - continue with the next one. ++ */ ++static int mptcp_prevalidate_skb(struct sock *sk, struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ /* If we are in infinite mode, the subflow-fin is in fact a data-fin. */ ++ if (!skb->len && tcp_hdr(skb)->fin && !mptcp_is_data_fin(skb) && ++ !tp->mpcb->infinite_mapping_rcv) { ++ /* Remove a pure subflow-fin from the queue and increase ++ * copied_seq. ++ */ ++ tp->copied_seq = TCP_SKB_CB(skb)->end_seq; ++ __skb_unlink(skb, &sk->sk_receive_queue); ++ __kfree_skb(skb); ++ return -1; ++ } ++ ++ /* If we are not yet fully established and do not know the mapping for ++ * this segment, this path has to fallback to infinite or be torn down. ++ */ ++ if (!tp->mptcp->fully_established && !mptcp_is_data_seq(skb) && ++ !tp->mptcp->mapping_present && !tp->mpcb->infinite_mapping_rcv) { ++ pr_err("%s %#x will fallback - pi %d from %pS, seq %u\n", ++ __func__, tp->mpcb->mptcp_loc_token, ++ tp->mptcp->path_index, __builtin_return_address(0), ++ TCP_SKB_CB(skb)->seq); ++ ++ if (!is_master_tp(tp)) { ++ mptcp_send_reset(sk); ++ return 1; ++ } ++ ++ tp->mpcb->infinite_mapping_snd = 1; ++ tp->mpcb->infinite_mapping_rcv = 1; ++ tp->mptcp->fully_established = 1; ++ } ++ ++ /* Receiver-side becomes fully established when a whole rcv-window has ++ * been received without the need to fallback due to the previous ++ * condition. */ ++ if (!tp->mptcp->fully_established) { ++ tp->mptcp->init_rcv_wnd -= skb->len; ++ if (tp->mptcp->init_rcv_wnd < 0) ++ mptcp_become_fully_estab(sk); ++ } ++ ++ return 0; ++} ++ ++/* @return: 0 everything is fine. Just continue processing ++ * 1 subflow is broken stop everything ++ * -1 this packet was broken - continue with the next one. ++ */ ++static int mptcp_detect_mapping(struct sock *sk, struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = mptcp_meta_tp(tp); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); ++ u32 *ptr; ++ u32 data_seq, sub_seq, data_len, tcp_end_seq; ++ ++ /* If we are in infinite-mapping-mode, the subflow is guaranteed to be ++ * in-order at the data-level. Thus data-seq-numbers can be inferred ++ * from what is expected at the data-level. ++ */ ++ if (mpcb->infinite_mapping_rcv) { ++ tp->mptcp->map_data_seq = mptcp_get_rcv_nxt_64(meta_tp); ++ tp->mptcp->map_subseq = tcb->seq; ++ tp->mptcp->map_data_len = skb->len; ++ tp->mptcp->map_data_fin = tcp_hdr(skb)->fin; ++ tp->mptcp->mapping_present = 1; ++ return 0; ++ } ++ ++ /* No mapping here? Exit - it is either already set or still on its way */ ++ if (!mptcp_is_data_seq(skb)) { ++ /* Too many packets without a mapping - this subflow is broken */ ++ if (!tp->mptcp->mapping_present && ++ tp->rcv_nxt - tp->copied_seq > 65536) { ++ mptcp_send_reset(sk); ++ return 1; ++ } ++ ++ return 0; ++ } ++ ++ ptr = mptcp_skb_set_data_seq(skb, &data_seq, mpcb); ++ ptr++; ++ sub_seq = get_unaligned_be32(ptr) + tp->mptcp->rcv_isn; ++ ptr++; ++ data_len = get_unaligned_be16(ptr); ++ ++ /* If it's an empty skb with DATA_FIN, sub_seq must get fixed. ++ * The draft sets it to 0, but we really would like to have the ++ * real value, to have an easy handling afterwards here in this ++ * function. ++ */ ++ if (mptcp_is_data_fin(skb) && skb->len == 0) ++ sub_seq = TCP_SKB_CB(skb)->seq; ++ ++ /* If there is already a mapping - we check if it maps with the current ++ * one. If not - we reset. ++ */ ++ if (tp->mptcp->mapping_present && ++ (data_seq != (u32)tp->mptcp->map_data_seq || ++ sub_seq != tp->mptcp->map_subseq || ++ data_len != tp->mptcp->map_data_len + tp->mptcp->map_data_fin || ++ mptcp_is_data_fin(skb) != tp->mptcp->map_data_fin)) { ++ /* Mapping in packet is different from what we want */ ++ pr_err("%s Mappings do not match!\n", __func__); ++ pr_err("%s dseq %u mdseq %u, sseq %u msseq %u dlen %u mdlen %u dfin %d mdfin %d\n", ++ __func__, data_seq, (u32)tp->mptcp->map_data_seq, ++ sub_seq, tp->mptcp->map_subseq, data_len, ++ tp->mptcp->map_data_len, mptcp_is_data_fin(skb), ++ tp->mptcp->map_data_fin); ++ mptcp_send_reset(sk); ++ return 1; ++ } ++ ++ /* If the previous check was good, the current mapping is valid and we exit. */ ++ if (tp->mptcp->mapping_present) ++ return 0; ++ ++ /* Mapping not yet set on this subflow - we set it here! */ ++ ++ if (!data_len) { ++ mpcb->infinite_mapping_rcv = 1; ++ tp->mptcp->fully_established = 1; ++ /* We need to repeat mp_fail's until the sender felt ++ * back to infinite-mapping - here we stop repeating it. ++ */ ++ tp->mptcp->send_mp_fail = 0; ++ ++ /* We have to fixup data_len - it must be the same as skb->len */ ++ data_len = skb->len + (mptcp_is_data_fin(skb) ? 1 : 0); ++ sub_seq = tcb->seq; ++ ++ /* TODO kill all other subflows than this one */ ++ /* data_seq and so on are set correctly */ ++ ++ /* At this point, the meta-ofo-queue has to be emptied, ++ * as the following data is guaranteed to be in-order at ++ * the data and subflow-level ++ */ ++ mptcp_purge_ofo_queue(meta_tp); ++ } ++ ++ /* We are sending mp-fail's and thus are in fallback mode. ++ * Ignore packets which do not announce the fallback and still ++ * want to provide a mapping. ++ */ ++ if (tp->mptcp->send_mp_fail) { ++ tp->copied_seq = TCP_SKB_CB(skb)->end_seq; ++ __skb_unlink(skb, &sk->sk_receive_queue); ++ __kfree_skb(skb); ++ return -1; ++ } ++ ++ /* FIN increased the mapping-length by 1 */ ++ if (mptcp_is_data_fin(skb)) ++ data_len--; ++ ++ /* Subflow-sequences of packet must be ++ * (at least partially) be part of the DSS-mapping's ++ * subflow-sequence-space. ++ * ++ * Basically the mapping is not valid, if either of the ++ * following conditions is true: ++ * ++ * 1. It's not a data_fin and ++ * MPTCP-sub_seq >= TCP-end_seq ++ * ++ * 2. It's a data_fin and TCP-end_seq > TCP-seq and ++ * MPTCP-sub_seq >= TCP-end_seq ++ * ++ * The previous two can be merged into: ++ * TCP-end_seq > TCP-seq and MPTCP-sub_seq >= TCP-end_seq ++ * Because if it's not a data-fin, TCP-end_seq > TCP-seq ++ * ++ * 3. It's a data_fin and skb->len == 0 and ++ * MPTCP-sub_seq > TCP-end_seq ++ * ++ * 4. It's not a data_fin and TCP-end_seq > TCP-seq and ++ * MPTCP-sub_seq + MPTCP-data_len <= TCP-seq ++ * ++ * 5. MPTCP-sub_seq is prior to what we already copied (copied_seq) ++ */ ++ ++ /* subflow-fin is not part of the mapping - ignore it here ! */ ++ tcp_end_seq = tcb->end_seq - tcp_hdr(skb)->fin; ++ if ((!before(sub_seq, tcb->end_seq) && after(tcp_end_seq, tcb->seq)) || ++ (mptcp_is_data_fin(skb) && skb->len == 0 && after(sub_seq, tcb->end_seq)) || ++ (!after(sub_seq + data_len, tcb->seq) && after(tcp_end_seq, tcb->seq)) || ++ before(sub_seq, tp->copied_seq)) { ++ /* Subflow-sequences of packet is different from what is in the ++ * packet's dss-mapping. The peer is misbehaving - reset ++ */ ++ pr_err("%s Packet's mapping does not map to the DSS sub_seq %u " ++ "end_seq %u, tcp_end_seq %u seq %u dfin %u len %u data_len %u" ++ "copied_seq %u\n", __func__, sub_seq, tcb->end_seq, tcp_end_seq, tcb->seq, mptcp_is_data_fin(skb), ++ skb->len, data_len, tp->copied_seq); ++ mptcp_send_reset(sk); ++ return 1; ++ } ++ ++ /* Does the DSS had 64-bit seqnum's ? */ ++ if (!(tcb->mptcp_flags & MPTCPHDR_SEQ64_SET)) { ++ /* Wrapped around? */ ++ if (unlikely(after(data_seq, meta_tp->rcv_nxt) && data_seq < meta_tp->rcv_nxt)) { ++ tp->mptcp->map_data_seq = mptcp_get_data_seq_64(mpcb, !mpcb->rcv_hiseq_index, data_seq); ++ } else { ++ /* Else, access the default high-order bits */ ++ tp->mptcp->map_data_seq = mptcp_get_data_seq_64(mpcb, mpcb->rcv_hiseq_index, data_seq); ++ } ++ } else { ++ tp->mptcp->map_data_seq = mptcp_get_data_seq_64(mpcb, (tcb->mptcp_flags & MPTCPHDR_SEQ64_INDEX) ? 1 : 0, data_seq); ++ ++ if (unlikely(tcb->mptcp_flags & MPTCPHDR_SEQ64_OFO)) { ++ /* We make sure that the data_seq is invalid. ++ * It will be dropped later. ++ */ ++ tp->mptcp->map_data_seq += 0xFFFFFFFF; ++ tp->mptcp->map_data_seq += 0xFFFFFFFF; ++ } ++ } ++ ++ tp->mptcp->map_data_len = data_len; ++ tp->mptcp->map_subseq = sub_seq; ++ tp->mptcp->map_data_fin = mptcp_is_data_fin(skb) ? 1 : 0; ++ tp->mptcp->mapping_present = 1; ++ ++ return 0; ++} ++ ++/* Similar to tcp_sequence(...) */ ++static inline int mptcp_sequence(const struct tcp_sock *meta_tp, ++ u64 data_seq, u64 end_data_seq) ++{ ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ u64 rcv_wup64; ++ ++ /* Wrap-around? */ ++ if (meta_tp->rcv_wup > meta_tp->rcv_nxt) { ++ rcv_wup64 = ((u64)(mpcb->rcv_high_order[mpcb->rcv_hiseq_index] - 1) << 32) | ++ meta_tp->rcv_wup; ++ } else { ++ rcv_wup64 = mptcp_get_data_seq_64(mpcb, mpcb->rcv_hiseq_index, ++ meta_tp->rcv_wup); ++ } ++ ++ return !before64(end_data_seq, rcv_wup64) && ++ !after64(data_seq, mptcp_get_rcv_nxt_64(meta_tp) + tcp_receive_window(meta_tp)); ++} ++ ++/* @return: 0 everything is fine. Just continue processing ++ * -1 this packet was broken - continue with the next one. ++ */ ++static int mptcp_validate_mapping(struct sock *sk, struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sk_buff *tmp, *tmp1; ++ u32 tcp_end_seq; ++ ++ if (!tp->mptcp->mapping_present) ++ return 0; ++ ++ /* either, the new skb gave us the mapping and the first segment ++ * in the sub-rcv-queue has to be trimmed ... ++ */ ++ tmp = skb_peek(&sk->sk_receive_queue); ++ if (before(TCP_SKB_CB(tmp)->seq, tp->mptcp->map_subseq) && ++ after(TCP_SKB_CB(tmp)->end_seq, tp->mptcp->map_subseq)) ++ mptcp_skb_trim_head(tmp, sk, tp->mptcp->map_subseq); ++ ++ /* ... or the new skb (tail) has to be split at the end. */ ++ tcp_end_seq = TCP_SKB_CB(skb)->end_seq - (tcp_hdr(skb)->fin ? 1 : 0); ++ if (after(tcp_end_seq, tp->mptcp->map_subseq + tp->mptcp->map_data_len)) { ++ u32 seq = tp->mptcp->map_subseq + tp->mptcp->map_data_len; ++ if (mptcp_skb_split_tail(skb, sk, seq)) { /* Allocation failed */ ++ /* TODO : maybe handle this here better. ++ * We now just force meta-retransmission. ++ */ ++ tp->copied_seq = TCP_SKB_CB(skb)->end_seq; ++ __skb_unlink(skb, &sk->sk_receive_queue); ++ __kfree_skb(skb); ++ return -1; ++ } ++ } ++ ++ /* Now, remove old sk_buff's from the receive-queue. ++ * This may happen if the mapping has been lost for these segments and ++ * the next mapping has already been received. ++ */ ++ if (tp->mptcp->mapping_present && ++ before(TCP_SKB_CB(skb_peek(&sk->sk_receive_queue))->seq, tp->mptcp->map_subseq)) { ++ skb_queue_walk_safe(&sk->sk_receive_queue, tmp1, tmp) { ++ if (!before(TCP_SKB_CB(tmp1)->seq, tp->mptcp->map_subseq)) ++ break; ++ ++ tp->copied_seq = TCP_SKB_CB(tmp1)->end_seq; ++ __skb_unlink(tmp1, &sk->sk_receive_queue); ++ ++ /* Impossible that we could free skb here, because his ++ * mapping is known to be valid from previous checks ++ */ ++ __kfree_skb(tmp1); ++ } ++ } ++ ++ return 0; ++} ++ ++/* @return: 0 everything is fine. Just continue processing ++ * 1 subflow is broken stop everything ++ * -1 this mapping has been put in the meta-receive-queue ++ * -2 this mapping has been eaten by the application ++ */ ++static int mptcp_queue_skb(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = mptcp_meta_tp(tp); ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct sk_buff *tmp, *tmp1; ++ u64 rcv_nxt64 = mptcp_get_rcv_nxt_64(meta_tp); ++ bool data_queued = false; ++ ++ /* Have we not yet received the full mapping? */ ++ if (!tp->mptcp->mapping_present || ++ before(tp->rcv_nxt, tp->mptcp->map_subseq + tp->mptcp->map_data_len)) ++ return 0; ++ ++ /* Is this an overlapping mapping? rcv_nxt >= end_data_seq ++ * OR ++ * This mapping is out of window ++ */ ++ if (!before64(rcv_nxt64, tp->mptcp->map_data_seq + tp->mptcp->map_data_len + tp->mptcp->map_data_fin) || ++ !mptcp_sequence(meta_tp, tp->mptcp->map_data_seq, ++ tp->mptcp->map_data_seq + tp->mptcp->map_data_len + tp->mptcp->map_data_fin)) { ++ skb_queue_walk_safe(&sk->sk_receive_queue, tmp1, tmp) { ++ __skb_unlink(tmp1, &sk->sk_receive_queue); ++ tp->copied_seq = TCP_SKB_CB(tmp1)->end_seq; ++ __kfree_skb(tmp1); ++ ++ if (!skb_queue_empty(&sk->sk_receive_queue) && ++ !before(TCP_SKB_CB(tmp)->seq, ++ tp->mptcp->map_subseq + tp->mptcp->map_data_len)) ++ break; ++ } ++ ++ mptcp_reset_mapping(tp); ++ ++ return -1; ++ } ++ ++ /* Record it, because we want to send our data_fin on the same path */ ++ if (tp->mptcp->map_data_fin) { ++ mpcb->dfin_path_index = tp->mptcp->path_index; ++ mpcb->dfin_combined = !!(sk->sk_shutdown & RCV_SHUTDOWN); ++ } ++ ++ /* Verify the checksum */ ++ if (mpcb->dss_csum && !mpcb->infinite_mapping_rcv) { ++ int ret = mptcp_verif_dss_csum(sk); ++ ++ if (ret <= 0) { ++ mptcp_reset_mapping(tp); ++ return 1; ++ } ++ } ++ ++ if (before64(rcv_nxt64, tp->mptcp->map_data_seq)) { ++ /* Seg's have to go to the meta-ofo-queue */ ++ skb_queue_walk_safe(&sk->sk_receive_queue, tmp1, tmp) { ++ tp->copied_seq = TCP_SKB_CB(tmp1)->end_seq; ++ mptcp_prepare_skb(tmp1, tmp, sk); ++ __skb_unlink(tmp1, &sk->sk_receive_queue); ++ /* MUST be done here, because fragstolen may be true later. ++ * Then, kfree_skb_partial will not account the memory. ++ */ ++ skb_orphan(tmp1); ++ ++ if (!mpcb->in_time_wait) /* In time-wait, do not receive data */ ++ mptcp_add_meta_ofo_queue(meta_sk, tmp1, sk); ++ else ++ __kfree_skb(tmp1); ++ ++ if (!skb_queue_empty(&sk->sk_receive_queue) && ++ !before(TCP_SKB_CB(tmp)->seq, ++ tp->mptcp->map_subseq + tp->mptcp->map_data_len)) ++ break; ++ ++ } ++ } else { ++ /* Ready for the meta-rcv-queue */ ++ skb_queue_walk_safe(&sk->sk_receive_queue, tmp1, tmp) { ++ int eaten = 0; ++ int copied_early = 0; ++ bool fragstolen = false; ++ u32 old_rcv_nxt = meta_tp->rcv_nxt; ++ ++ tp->copied_seq = TCP_SKB_CB(tmp1)->end_seq; ++ mptcp_prepare_skb(tmp1, tmp, sk); ++ __skb_unlink(tmp1, &sk->sk_receive_queue); ++ /* MUST be done here, because fragstolen may be true. ++ * Then, kfree_skb_partial will not account the memory. ++ */ ++ skb_orphan(tmp1); ++ ++ /* This segment has already been received */ ++ if (!after(TCP_SKB_CB(tmp1)->end_seq, meta_tp->rcv_nxt)) { ++ __kfree_skb(tmp1); ++ goto next; ++ } ++ ++#ifdef CONFIG_NET_DMA ++ if (TCP_SKB_CB(tmp1)->seq == meta_tp->rcv_nxt && ++ meta_tp->ucopy.task == current && ++ meta_tp->copied_seq == meta_tp->rcv_nxt && ++ tmp1->len <= meta_tp->ucopy.len && ++ sock_owned_by_user(meta_sk) && ++ tcp_dma_try_early_copy(meta_sk, tmp1, 0)) { ++ copied_early = 1; ++ eaten = 1; ++ } ++#endif ++ ++ /* Is direct copy possible ? */ ++ if (TCP_SKB_CB(tmp1)->seq == meta_tp->rcv_nxt && ++ meta_tp->ucopy.task == current && ++ meta_tp->copied_seq == meta_tp->rcv_nxt && ++ meta_tp->ucopy.len && sock_owned_by_user(meta_sk) && ++ !copied_early) ++ eaten = mptcp_direct_copy(tmp1, meta_sk); ++ ++ if (mpcb->in_time_wait) /* In time-wait, do not receive data */ ++ eaten = 1; ++ ++ if (!eaten) ++ eaten = tcp_queue_rcv(meta_sk, tmp1, 0, &fragstolen); ++ ++ meta_tp->rcv_nxt = TCP_SKB_CB(tmp1)->end_seq; ++ mptcp_check_rcvseq_wrap(meta_tp, old_rcv_nxt); ++ ++ if (copied_early) ++ tcp_cleanup_rbuf(meta_sk, tmp1->len); ++ ++ if (tcp_hdr(tmp1)->fin && !mpcb->in_time_wait) ++ mptcp_fin(meta_sk); ++ ++ /* Check if this fills a gap in the ofo queue */ ++ if (!skb_queue_empty(&meta_tp->out_of_order_queue)) ++ mptcp_ofo_queue(meta_sk); ++ ++#ifdef CONFIG_NET_DMA ++ if (copied_early) ++ __skb_queue_tail(&meta_sk->sk_async_wait_queue, ++ tmp1); ++ else ++#endif ++ if (eaten) ++ kfree_skb_partial(tmp1, fragstolen); ++ ++ data_queued = true; ++next: ++ if (!skb_queue_empty(&sk->sk_receive_queue) && ++ !before(TCP_SKB_CB(tmp)->seq, ++ tp->mptcp->map_subseq + tp->mptcp->map_data_len)) ++ break; ++ } ++ } ++ ++ inet_csk(meta_sk)->icsk_ack.lrcvtime = tcp_time_stamp; ++ tp->mptcp->last_data_seq = tp->mptcp->map_data_seq; ++ mptcp_reset_mapping(tp); ++ ++ return data_queued ? -1 : -2; ++} ++ ++void mptcp_data_ready(struct sock *sk, int bytes) ++{ ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct sk_buff *skb, *tmp; ++ int queued = 0; ++ ++ /* If the meta is already closed, there is no point in pushing data */ ++ if (meta_sk->sk_state == TCP_CLOSE && !tcp_sk(sk)->mpcb->in_time_wait) { ++ skb_queue_purge(&sk->sk_receive_queue); ++ tcp_sk(sk)->copied_seq = tcp_sk(sk)->rcv_nxt; ++ goto exit; ++ } ++ ++restart: ++ /* Iterate over all segments, detect their mapping (if we don't have ++ * one yet), validate them and push everything one level higher. ++ */ ++ skb_queue_walk_safe(&sk->sk_receive_queue, skb, tmp) { ++ int ret; ++ /* Pre-validation - e.g., early fallback */ ++ ret = mptcp_prevalidate_skb(sk, skb); ++ if (ret < 0) ++ goto restart; ++ else if (ret > 0) ++ break; ++ ++ /* Set the current mapping */ ++ ret = mptcp_detect_mapping(sk, skb); ++ if (ret < 0) ++ goto restart; ++ else if (ret > 0) ++ break; ++ ++ /* Validation */ ++ if (mptcp_validate_mapping(sk, skb) < 0) ++ goto restart; ++ ++ /* Push a level higher */ ++ ret = mptcp_queue_skb(sk); ++ if (ret < 0) { ++ if (ret == -1) ++ queued = ret; ++ goto restart; ++ } else if (ret == 0) { ++ continue; ++ } else { /* ret == 1 */ ++ break; ++ } ++ } ++ ++exit: ++ if (tcp_sk(sk)->close_it) { ++ tcp_send_ack(sk); ++ tcp_time_wait(sk, TCP_TIME_WAIT, 0); ++ } ++ ++ if (queued == -1 && !sock_flag(meta_sk, SOCK_DEAD)) ++ meta_sk->sk_data_ready(meta_sk, 0); ++} ++ ++ ++int mptcp_check_req(struct sk_buff *skb, struct net *net) ++{ ++ struct tcphdr *th = tcp_hdr(skb); ++ struct sock *meta_sk = NULL; ++ ++ /* MPTCP structures not initialized */ ++ if (mptcp_init_failed) ++ return 0; ++ ++ if (skb->protocol == htons(ETH_P_IP)) ++ meta_sk = mptcp_v4_search_req(th->source, ip_hdr(skb)->saddr, ++ ip_hdr(skb)->daddr, net); ++#if IS_ENABLED(CONFIG_IPV6) ++ else /* IPv6 */ ++ meta_sk = mptcp_v6_search_req(th->source, &ipv6_hdr(skb)->saddr, ++ &ipv6_hdr(skb)->daddr, net); ++#endif /* CONFIG_IPV6 */ ++ ++ if (!meta_sk) ++ return 0; ++ ++ TCP_SKB_CB(skb)->mptcp_flags = MPTCPHDR_JOIN; ++ ++ bh_lock_sock_nested(meta_sk); ++ if (sock_owned_by_user(meta_sk)) { ++ skb->sk = meta_sk; ++ if (unlikely(sk_add_backlog(meta_sk, skb, ++ meta_sk->sk_rcvbuf + meta_sk->sk_sndbuf))) { ++ bh_unlock_sock(meta_sk); ++ NET_INC_STATS_BH(net, LINUX_MIB_TCPBACKLOGDROP); ++ sock_put(meta_sk); /* Taken by mptcp_search_req */ ++ kfree_skb(skb); ++ return 1; ++ } ++ } else if (skb->protocol == htons(ETH_P_IP)) { ++ tcp_v4_do_rcv(meta_sk, skb); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else { /* IPv6 */ ++ tcp_v6_do_rcv(meta_sk, skb); ++#endif /* CONFIG_IPV6 */ ++ } ++ bh_unlock_sock(meta_sk); ++ sock_put(meta_sk); /* Taken by mptcp_vX_search_req */ ++ return 1; ++} ++ ++struct mp_join *mptcp_find_join(struct sk_buff *skb) ++{ ++ struct tcphdr *th = tcp_hdr(skb); ++ unsigned char *ptr; ++ int length = (th->doff * 4) - sizeof(struct tcphdr); ++ ++ /* Jump through the options to check whether JOIN is there */ ++ ptr = (unsigned char *)(th + 1); ++ while (length > 0) { ++ int opcode = *ptr++; ++ int opsize; ++ ++ switch (opcode) { ++ case TCPOPT_EOL: ++ return NULL; ++ case TCPOPT_NOP: /* Ref: RFC 793 section 3.1 */ ++ length--; ++ continue; ++ default: ++ opsize = *ptr++; ++ if (opsize < 2) /* "silly options" */ ++ return NULL; ++ if (opsize > length) ++ return NULL; /* don't parse partial options */ ++ if (opcode == TCPOPT_MPTCP && ++ ((struct mptcp_option *)(ptr - 2))->sub == MPTCP_SUB_JOIN) { ++ return (struct mp_join *)(ptr - 2); ++ } ++ ptr += opsize - 2; ++ length -= opsize; ++ } ++ } ++ return NULL; ++} ++ ++int mptcp_lookup_join(struct sk_buff *skb, struct inet_timewait_sock *tw) ++{ ++ struct mptcp_cb *mpcb; ++ struct sock *meta_sk; ++ u32 token; ++ struct mp_join *join_opt = mptcp_find_join(skb); ++ if (!join_opt) ++ return 0; ++ ++ /* MPTCP structures were not initialized, so return error */ ++ if (mptcp_init_failed) ++ return -1; ++ ++ token = join_opt->u.syn.token; ++ meta_sk = mptcp_hash_find(dev_net(skb_dst(skb)->dev), token); ++ if (!meta_sk) { ++ mptcp_debug("%s:mpcb not found:%x\n", __func__, token); ++ return -1; ++ } ++ ++ mpcb = tcp_sk(meta_sk)->mpcb; ++ if (mpcb->infinite_mapping_rcv || mpcb->send_infinite_mapping) { ++ /* We are in fallback-mode on the reception-side - ++ * no new subflows! ++ */ ++ sock_put(meta_sk); /* Taken by mptcp_hash_find */ ++ return -1; ++ } ++ ++ /* Coming from time-wait-sock processing in tcp_v4_rcv. ++ * We have to deschedule it before continuing, because otherwise ++ * mptcp_v4_do_rcv will hit again on it inside tcp_v4_hnd_req. ++ */ ++ if (tw) { ++ inet_twsk_deschedule(tw, &tcp_death_row); ++ inet_twsk_put(tw); ++ } ++ ++ TCP_SKB_CB(skb)->mptcp_flags = MPTCPHDR_JOIN; ++ /* OK, this is a new syn/join, let's create a new open request and ++ * send syn+ack ++ */ ++ bh_lock_sock_nested(meta_sk); ++ if (sock_owned_by_user(meta_sk)) { ++ skb->sk = meta_sk; ++ if (unlikely(sk_add_backlog(meta_sk, skb, ++ meta_sk->sk_rcvbuf + meta_sk->sk_sndbuf))) { ++ bh_unlock_sock(meta_sk); ++ NET_INC_STATS_BH(sock_net(meta_sk), ++ LINUX_MIB_TCPBACKLOGDROP); ++ sock_put(meta_sk); /* Taken by mptcp_hash_find */ ++ kfree_skb(skb); ++ return 1; ++ } ++ } else if (skb->protocol == htons(ETH_P_IP)) { ++ tcp_v4_do_rcv(meta_sk, skb); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else { ++ tcp_v6_do_rcv(meta_sk, skb); ++#endif /* CONFIG_IPV6 */ ++ } ++ bh_unlock_sock(meta_sk); ++ sock_put(meta_sk); /* Taken by mptcp_hash_find */ ++ return 1; ++} ++ ++int mptcp_do_join_short(struct sk_buff *skb, struct mptcp_options_received *mopt, ++ struct tcp_options_received *tmp_opt, struct net *net) ++{ ++ struct sock *meta_sk; ++ u32 token; ++ ++ token = mopt->mptcp_rem_token; ++ meta_sk = mptcp_hash_find(net, token); ++ if (!meta_sk) { ++ mptcp_debug("%s:mpcb not found:%x\n", __func__, token); ++ return -1; ++ } ++ ++ TCP_SKB_CB(skb)->mptcp_flags = MPTCPHDR_JOIN; ++ ++ /* OK, this is a new syn/join, let's create a new open request and ++ * send syn+ack ++ */ ++ bh_lock_sock(meta_sk); ++ ++ /* This check is also done in mptcp_vX_do_rcv. But, there we cannot ++ * call tcp_vX_send_reset, because we hold already two socket-locks. ++ * (the listener and the meta from above) ++ * ++ * And the send-reset will try to take yet another one (ip_send_reply). ++ * Thus, we propagate the reset up to tcp_rcv_state_process. ++ */ ++ if (tcp_sk(meta_sk)->mpcb->infinite_mapping_rcv || ++ tcp_sk(meta_sk)->mpcb->send_infinite_mapping || ++ meta_sk->sk_state == TCP_CLOSE || !tcp_sk(meta_sk)->inside_tk_table) { ++ bh_unlock_sock(meta_sk); ++ sock_put(meta_sk); /* Taken by mptcp_hash_find */ ++ return -1; ++ } ++ ++ if (sock_owned_by_user(meta_sk)) { ++ skb->sk = meta_sk; ++ if (unlikely(sk_add_backlog(meta_sk, skb, ++ meta_sk->sk_rcvbuf + meta_sk->sk_sndbuf))) ++ NET_INC_STATS_BH(net, LINUX_MIB_TCPBACKLOGDROP); ++ else ++ /* Must make sure that upper layers won't free the ++ * skb if it is added to the backlog-queue. ++ */ ++ skb_get(skb); ++ } else { ++ /* mptcp_v4_do_rcv tries to free the skb - we prevent this, as ++ * the skb will finally be freed by tcp_v4_do_rcv (where we are ++ * coming from) ++ */ ++ skb_get(skb); ++ if (skb->protocol == htons(ETH_P_IP)) { ++ tcp_v4_do_rcv(meta_sk, skb); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else { /* IPv6 */ ++ tcp_v6_do_rcv(meta_sk, skb); ++#endif /* CONFIG_IPV6 */ ++ } ++ } ++ ++ bh_unlock_sock(meta_sk); ++ sock_put(meta_sk); /* Taken by mptcp_hash_find */ ++ return 0; ++} ++ ++/** ++ * Equivalent of tcp_fin() for MPTCP ++ * Can be called only when the FIN is validly part ++ * of the data seqnum space. Not before when we get holes. ++ */ ++void mptcp_fin(struct sock *meta_sk) ++{ ++ struct sock *sk = NULL, *sk_it; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ ++ mptcp_for_each_sk(mpcb, sk_it) { ++ if (tcp_sk(sk_it)->mptcp->path_index == mpcb->dfin_path_index) { ++ sk = sk_it; ++ break; ++ } ++ } ++ ++ if (!sk || sk->sk_state == TCP_CLOSE) ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ ++ inet_csk_schedule_ack(sk); ++ ++ meta_sk->sk_shutdown |= RCV_SHUTDOWN; ++ sock_set_flag(meta_sk, SOCK_DONE); ++ ++ switch (meta_sk->sk_state) { ++ case TCP_SYN_RECV: ++ case TCP_ESTABLISHED: ++ /* Move to CLOSE_WAIT */ ++ tcp_set_state(meta_sk, TCP_CLOSE_WAIT); ++ inet_csk(sk)->icsk_ack.pingpong = 1; ++ break; ++ ++ case TCP_CLOSE_WAIT: ++ case TCP_CLOSING: ++ /* Received a retransmission of the FIN, do ++ * nothing. ++ */ ++ break; ++ case TCP_LAST_ACK: ++ /* RFC793: Remain in the LAST-ACK state. */ ++ break; ++ ++ case TCP_FIN_WAIT1: ++ /* This case occurs when a simultaneous close ++ * happens, we must ack the received FIN and ++ * enter the CLOSING state. ++ */ ++ tcp_send_ack(sk); ++ tcp_set_state(meta_sk, TCP_CLOSING); ++ break; ++ case TCP_FIN_WAIT2: ++ /* Received a FIN -- send ACK and enter TIME_WAIT. */ ++ tcp_send_ack(sk); ++ tcp_time_wait(meta_sk, TCP_TIME_WAIT, 0); ++ break; ++ default: ++ /* Only TCP_LISTEN and TCP_CLOSE are left, in these ++ * cases we should never reach this piece of code. ++ */ ++ pr_err("%s: Impossible, meta_sk->sk_state=%d\n", __func__, ++ meta_sk->sk_state); ++ break; ++ } ++ ++ /* It _is_ possible, that we have something out-of-order _after_ FIN. ++ * Probably, we should reset in this case. For now drop them. ++ */ ++ mptcp_purge_ofo_queue(meta_tp); ++ sk_mem_reclaim(meta_sk); ++ ++ if (!sock_flag(meta_sk, SOCK_DEAD)) { ++ meta_sk->sk_state_change(meta_sk); ++ ++ /* Do not send POLL_HUP for half duplex close. */ ++ if (meta_sk->sk_shutdown == SHUTDOWN_MASK || ++ meta_sk->sk_state == TCP_CLOSE) ++ sk_wake_async(meta_sk, SOCK_WAKE_WAITD, POLL_HUP); ++ else ++ sk_wake_async(meta_sk, SOCK_WAKE_WAITD, POLL_IN); ++ } ++ ++ return; ++} ++ ++static void mptcp_xmit_retransmit_queue(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sk_buff *skb; ++ ++ if (!meta_tp->packets_out) ++ return; ++ ++ tcp_for_write_queue(skb, meta_sk) { ++ if (skb == tcp_send_head(meta_sk)) ++ break; ++ ++ if (mptcp_retransmit_skb(meta_sk, skb)) ++ return; ++ ++ if (skb == tcp_write_queue_head(meta_sk)) ++ inet_csk_reset_xmit_timer(meta_sk, ICSK_TIME_RETRANS, ++ inet_csk(meta_sk)->icsk_rto, ++ TCP_RTO_MAX); ++ } ++} ++ ++/* Handle the DATA_ACK */ ++static void mptcp_data_ack(struct sock *sk, const struct sk_buff *skb) ++{ ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk), *tp = tcp_sk(sk); ++ struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); ++ u32 prior_snd_una = meta_tp->snd_una; ++ int prior_packets; ++ u32 nwin, data_ack, data_seq; ++ u16 data_len = 0; ++ ++ /* A valid packet came in - subflow is operational again */ ++ tp->pf = 0; ++ ++ /* Even if there is no data-ack, we stop retransmitting. ++ * Except if this is a SYN/ACK. Then it is just a retransmission ++ */ ++ if (tp->mptcp->pre_established && !tcp_hdr(skb)->syn) { ++ tp->mptcp->pre_established = 0; ++ sk_stop_timer(sk, &tp->mptcp->mptcp_ack_timer); ++ } ++ ++ /* If we are in infinite mapping mode, rx_opt.data_ack has been ++ * set by mptcp_clean_rtx_infinite. ++ */ ++ if (!(tcb->mptcp_flags & MPTCPHDR_ACK) && !tp->mpcb->infinite_mapping_snd) ++ goto exit; ++ ++ data_ack = tp->mptcp->rx_opt.data_ack; ++ ++ if (unlikely(!tp->mptcp->fully_established) && ++ (data_ack != meta_tp->mptcp->snt_isn || ++ tp->mptcp->snt_isn + 1 != TCP_SKB_CB(skb)->ack_seq)) ++ /* As soon as data has been data-acked, ++ * or a subflow-data-ack (not acking syn - thus snt_isn + 1) ++ * includes a data-ack, we are fully established ++ */ ++ mptcp_become_fully_estab(sk); ++ ++ /* Get the data_seq */ ++ if (mptcp_is_data_seq(skb)) { ++ data_seq = tp->mptcp->rx_opt.data_seq; ++ data_len = tp->mptcp->rx_opt.data_len; ++ } else { ++ data_seq = meta_tp->snd_wl1; ++ } ++ ++ /* If the ack is older than previous acks ++ * then we can probably ignore it. ++ */ ++ if (before(data_ack, prior_snd_una)) ++ goto exit; ++ ++ /* If the ack includes data we haven't sent yet, discard ++ * this segment (RFC793 Section 3.9). ++ */ ++ if (after(data_ack, meta_tp->snd_nxt)) ++ goto exit; ++ ++ /*** Now, update the window - inspired by tcp_ack_update_window ***/ ++ nwin = ntohs(tcp_hdr(skb)->window); ++ ++ if (likely(!tcp_hdr(skb)->syn)) ++ nwin <<= tp->rx_opt.snd_wscale; ++ ++ if (tcp_may_update_window(meta_tp, data_ack, data_seq, nwin)) { ++ tcp_update_wl(meta_tp, data_seq); ++ ++ /* Draft v09, Section 3.3.5: ++ * [...] It should only update its local receive window values ++ * when the largest sequence number allowed (i.e. DATA_ACK + ++ * receive window) increases. [...] ++ */ ++ if (meta_tp->snd_wnd != nwin && ++ !before(data_ack + nwin, tcp_wnd_end(meta_tp))) { ++ meta_tp->snd_wnd = nwin; ++ ++ if (nwin > meta_tp->max_window) ++ meta_tp->max_window = nwin; ++ } ++ } ++ /*** Done, update the window ***/ ++ ++ /* We passed data and got it acked, remove any soft error ++ * log. Something worked... ++ */ ++ sk->sk_err_soft = 0; ++ inet_csk(meta_sk)->icsk_probes_out = 0; ++ meta_tp->rcv_tstamp = tcp_time_stamp; ++ prior_packets = meta_tp->packets_out; ++ if (!prior_packets) ++ goto no_queue; ++ ++ meta_tp->snd_una = data_ack; ++ ++ mptcp_clean_rtx_queue(meta_sk, prior_snd_una); ++ ++ /* We are in loss-state, and something got acked, retransmit the whole ++ * queue now! ++ */ ++ if (inet_csk(meta_sk)->icsk_ca_state == TCP_CA_Loss && ++ after(data_ack, prior_snd_una)) { ++ mptcp_xmit_retransmit_queue(meta_sk); ++ inet_csk(meta_sk)->icsk_ca_state = TCP_CA_Open; ++ } ++ ++ /* Simplified version of tcp_new_space, because the snd-buffer ++ * is handled by all the subflows. ++ */ ++ if (sock_flag(meta_sk, SOCK_QUEUE_SHRUNK)) { ++ sock_reset_flag(meta_sk, SOCK_QUEUE_SHRUNK); ++ if (meta_sk->sk_socket && ++ test_bit(SOCK_NOSPACE, &meta_sk->sk_socket->flags)) ++ meta_sk->sk_write_space(meta_sk); ++ } ++ ++ if (meta_sk->sk_state != TCP_ESTABLISHED && ++ mptcp_rcv_state_process(meta_sk, sk, skb, data_seq, data_len)) ++ return; ++ ++exit: ++ mptcp_push_pending_frames(meta_sk); ++ ++ return; ++ ++no_queue: ++ if (tcp_send_head(meta_sk)) ++ tcp_ack_probe(meta_sk); ++ ++ mptcp_push_pending_frames(meta_sk); ++ ++ return; ++} ++ ++void mptcp_clean_rtx_infinite(struct sk_buff *skb, struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = tcp_sk(mptcp_meta_sk(sk)); ++ ++ if (!tp->mpcb->infinite_mapping_snd) ++ return; ++ ++ /* The difference between both write_seq's represents the offset between ++ * data-sequence and subflow-sequence. As we are infinite, this must ++ * match. ++ * ++ * Thus, from this difference we can infer the meta snd_una. ++ */ ++ tp->mptcp->rx_opt.data_ack = meta_tp->snd_nxt - tp->snd_nxt + ++ tp->snd_una; ++ ++ mptcp_data_ack(sk, skb); ++} ++ ++/**** static functions used by mptcp_parse_options */ ++ ++static inline int mptcp_rem_raddress(struct mptcp_cb *mpcb, u8 rem_id) ++{ ++ if (mptcp_v4_rem_raddress(mpcb, rem_id) < 0) { ++#if IS_ENABLED(CONFIG_IPV6) ++ if (mptcp_v6_rem_raddress(mpcb, rem_id) < 0) ++ return -1; ++#else ++ return -1; ++#endif /* CONFIG_IPV6 */ ++ } ++ return 0; ++} ++ ++static void mptcp_send_reset_rem_id(const struct mptcp_cb *mpcb, u8 rem_id) ++{ ++ struct sock *sk_it, *tmpsk; ++ ++ mptcp_for_each_sk_safe(mpcb, sk_it, tmpsk) { ++ if (tcp_sk(sk_it)->mptcp->rem_id == rem_id) { ++ mptcp_reinject_data(sk_it, 0); ++ sk_it->sk_err = ECONNRESET; ++ if (tcp_need_reset(sk_it->sk_state)) ++ tcp_send_active_reset(sk_it, GFP_ATOMIC); ++ mptcp_sub_force_close(sk_it); ++ } ++ } ++} ++ ++void mptcp_parse_options(const uint8_t *ptr, int opsize, ++ struct tcp_options_received *opt_rx, ++ struct mptcp_options_received *mopt, ++ const struct sk_buff *skb) ++{ ++ struct mptcp_option *mp_opt = (struct mptcp_option *)ptr; ++ ++ /* If the socket is mp-capable we would have a mopt. */ ++ if (!mopt) ++ return; ++ ++ switch (mp_opt->sub) { ++ case MPTCP_SUB_CAPABLE: ++ { ++ struct mp_capable *mpcapable = (struct mp_capable *)ptr; ++ ++ if (opsize != MPTCP_SUB_LEN_CAPABLE_SYN && ++ opsize != MPTCP_SUB_LEN_CAPABLE_ACK) { ++ mptcp_debug("%s: mp_capable: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ if (!sysctl_mptcp_enabled) ++ break; ++ ++ /* We only support MPTCP version 0 */ ++ if (mpcapable->ver != 0) ++ break; ++ ++ /* MPTCP-RFC 6824: ++ * "If receiving a message with the 'B' flag set to 1, and this ++ * is not understood, then this SYN MUST be silently ignored; ++ */ ++ if (mpcapable->b) { ++ mopt->drop_me = 1; ++ break; ++ } ++ ++ /* MPTCP-RFC 6824: ++ * "An implementation that only supports this method MUST set ++ * bit "H" to 1, and bits "C" through "G" to 0." ++ */ ++ if (!mpcapable->h) ++ break; ++ ++ mopt->saw_mpc = 1; ++ mopt->dss_csum = sysctl_mptcp_checksum || mpcapable->a; ++ ++ if (opsize >= MPTCP_SUB_LEN_CAPABLE_SYN) ++ mopt->mptcp_key = mpcapable->sender_key; ++ ++ break; ++ } ++ case MPTCP_SUB_JOIN: ++ { ++ struct mp_join *mpjoin = (struct mp_join *)ptr; ++ ++ if (opsize != MPTCP_SUB_LEN_JOIN_SYN && ++ opsize != MPTCP_SUB_LEN_JOIN_SYNACK && ++ opsize != MPTCP_SUB_LEN_JOIN_ACK) { ++ mptcp_debug("%s: mp_join: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ /* saw_mpc must be set, because in tcp_check_req we assume that ++ * it is set to support falling back to reg. TCP if a rexmitted ++ * SYN has no MP_CAPABLE or MP_JOIN ++ */ ++ switch (opsize) { ++ case MPTCP_SUB_LEN_JOIN_SYN: ++ mopt->is_mp_join = 1; ++ mopt->saw_mpc = 1; ++ mopt->low_prio = mpjoin->b; ++ mopt->rem_id = mpjoin->addr_id; ++ mopt->mptcp_rem_token = mpjoin->u.syn.token; ++ mopt->mptcp_recv_nonce = mpjoin->u.syn.nonce; ++ break; ++ case MPTCP_SUB_LEN_JOIN_SYNACK: ++ mopt->saw_mpc = 1; ++ mopt->low_prio = mpjoin->b; ++ mopt->rem_id = mpjoin->addr_id; ++ mopt->mptcp_recv_tmac = mpjoin->u.synack.mac; ++ mopt->mptcp_recv_nonce = mpjoin->u.synack.nonce; ++ break; ++ case MPTCP_SUB_LEN_JOIN_ACK: ++ mopt->saw_mpc = 1; ++ mopt->join_ack = 1; ++ memcpy(mopt->mptcp_recv_mac, mpjoin->u.ack.mac, 20); ++ break; ++ } ++ break; ++ } ++ case MPTCP_SUB_DSS: ++ { ++ struct mp_dss *mdss = (struct mp_dss *)ptr; ++ struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); ++ ++ /* We check opsize for the csum and non-csum case. We do this, ++ * because the draft says that the csum SHOULD be ignored if ++ * it has not been negotiated in the MP_CAPABLE but still is ++ * present in the data. ++ * ++ * It will get ignored later in mptcp_queue_skb. ++ */ ++ if (opsize != mptcp_sub_len_dss(mdss, 0) && ++ opsize != mptcp_sub_len_dss(mdss, 1)) { ++ mptcp_debug("%s: mp_dss: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ ptr += 4; ++ ++ if (mdss->A) { ++ tcb->mptcp_flags |= MPTCPHDR_ACK; ++ ++ if (mdss->a) { ++ mopt->data_ack = (u32) get_unaligned_be64(ptr); ++ ptr += MPTCP_SUB_LEN_ACK_64; ++ } else { ++ mopt->data_ack = get_unaligned_be32(ptr); ++ ptr += MPTCP_SUB_LEN_ACK; ++ } ++ } ++ ++ tcb->dss_off = (ptr - skb_transport_header(skb)); ++ ++ if (mdss->M) { ++ if (mdss->m) { ++ u64 data_seq64 = get_unaligned_be64(ptr); ++ ++ tcb->mptcp_flags |= MPTCPHDR_SEQ64_SET; ++ mopt->data_seq = (u32) data_seq64; ++ ++ ptr += 12; /* 64-bit dseq + subseq */ ++ } else { ++ mopt->data_seq = get_unaligned_be32(ptr); ++ ptr += 8; /* 32-bit dseq + subseq */ ++ } ++ mopt->data_len = get_unaligned_be16(ptr); ++ ++ tcb->mptcp_flags |= MPTCPHDR_SEQ; ++ ++ /* Is a check-sum present? */ ++ if (opsize == mptcp_sub_len_dss(mdss, 1)) ++ tcb->mptcp_flags |= MPTCPHDR_DSS_CSUM; ++ ++ /* DATA_FIN only possible with DSS-mapping */ ++ if (mdss->F) ++ tcb->mptcp_flags |= MPTCPHDR_FIN; ++ } ++ ++ break; ++ } ++ case MPTCP_SUB_ADD_ADDR: ++ { ++#if IS_ENABLED(CONFIG_IPV6) ++ struct mp_add_addr *mpadd = (struct mp_add_addr *)ptr; ++ ++ if ((mpadd->ipver == 4 && opsize != MPTCP_SUB_LEN_ADD_ADDR4 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR4 + 2) || ++ (mpadd->ipver == 6 && opsize != MPTCP_SUB_LEN_ADD_ADDR6 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR6 + 2)) { ++#else ++ if (opsize != MPTCP_SUB_LEN_ADD_ADDR4 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR4 + 2) { ++#endif /* CONFIG_IPV6 */ ++ mptcp_debug("%s: mp_add_addr: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ /* We have to manually parse the options if we got two of them. */ ++ if (mopt->saw_add_addr) { ++ mopt->more_add_addr = 1; ++ break; ++ } ++ mopt->saw_add_addr = 1; ++ mopt->add_addr_ptr = ptr; ++ break; ++ } ++ case MPTCP_SUB_REMOVE_ADDR: ++ if ((opsize - MPTCP_SUB_LEN_REMOVE_ADDR) < 0) { ++ mptcp_debug("%s: mp_remove_addr: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ if (mopt->saw_rem_addr) { ++ mopt->more_rem_addr = 1; ++ break; ++ } ++ mopt->saw_rem_addr = 1; ++ mopt->rem_addr_ptr = ptr; ++ break; ++ case MPTCP_SUB_PRIO: ++ { ++ struct mp_prio *mpprio = (struct mp_prio *)ptr; ++ ++ if (opsize != MPTCP_SUB_LEN_PRIO && ++ opsize != MPTCP_SUB_LEN_PRIO_ADDR) { ++ mptcp_debug("%s: mp_prio: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ mopt->saw_low_prio = 1; ++ mopt->low_prio = mpprio->b; ++ ++ if (opsize == MPTCP_SUB_LEN_PRIO_ADDR) { ++ mopt->saw_low_prio = 2; ++ mopt->prio_addr_id = mpprio->addr_id; ++ } ++ break; ++ } ++ case MPTCP_SUB_FAIL: ++ if (opsize != MPTCP_SUB_LEN_FAIL) { ++ mptcp_debug("%s: mp_fail: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ mopt->mp_fail = 1; ++ break; ++ case MPTCP_SUB_FCLOSE: ++ if (opsize != MPTCP_SUB_LEN_FCLOSE) { ++ mptcp_debug("%s: mp_fclose: bad option size %d\n", ++ __func__, opsize); ++ break; ++ } ++ ++ mopt->mp_fclose = 1; ++ mopt->mptcp_key = ((struct mp_fclose *)ptr)->key; ++ ++ break; ++ default: ++ mptcp_debug("%s: Received unkown subtype: %d\n", ++ __func__, mp_opt->sub); ++ break; ++ } ++} ++ ++int mptcp_check_rtt(const struct tcp_sock *tp, int time) ++{ ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct sock *sk; ++ u32 rtt_max = 0; ++ ++ /* In MPTCP, we take the max delay across all flows, ++ * in order to take into account meta-reordering buffers. ++ */ ++ mptcp_for_each_sk(mpcb, sk) { ++ if (!mptcp_sk_can_recv(sk)) ++ continue; ++ ++ if (rtt_max < tcp_sk(sk)->rcv_rtt_est.rtt) ++ rtt_max = tcp_sk(sk)->rcv_rtt_est.rtt; ++ } ++ if (time < (rtt_max >> 3) || !rtt_max) ++ return 1; ++ ++ return 0; ++} ++ ++static void mptcp_handle_add_addr(const unsigned char *ptr, struct sock *sk) ++{ ++ struct mp_add_addr *mpadd = (struct mp_add_addr *)ptr; ++ ++ if (mpadd->ipver == 4) { ++ __be16 port = 0; ++ if (mpadd->len == MPTCP_SUB_LEN_ADD_ADDR4 + 2) ++ port = mpadd->u.v4.port; ++ ++ mptcp_v4_add_raddress(tcp_sk(sk)->mpcb, &mpadd->u.v4.addr, port, ++ mpadd->addr_id); ++#if IS_ENABLED(CONFIG_IPV6) ++ } else if (mpadd->ipver == 6) { ++ __be16 port = 0; ++ if (mpadd->len == MPTCP_SUB_LEN_ADD_ADDR6 + 2) ++ port = mpadd->u.v6.port; ++ ++ mptcp_v6_add_raddress(tcp_sk(sk)->mpcb, &mpadd->u.v6.addr, port, ++ mpadd->addr_id); ++#endif /* CONFIG_IPV6 */ ++ } ++} ++ ++static void mptcp_handle_rem_addr(const unsigned char *ptr, struct sock *sk) ++{ ++ struct mp_remove_addr *mprem = (struct mp_remove_addr *)ptr; ++ int i; ++ u8 rem_id; ++ ++ for (i = 0; i <= mprem->len - MPTCP_SUB_LEN_REMOVE_ADDR; i++) { ++ rem_id = (&mprem->addrs_id)[i]; ++ if (!mptcp_rem_raddress(tcp_sk(sk)->mpcb, rem_id)) ++ mptcp_send_reset_rem_id(tcp_sk(sk)->mpcb, rem_id); ++ } ++} ++ ++static void mptcp_parse_addropt(const struct sk_buff *skb, struct sock *sk) ++{ ++ struct tcphdr *th = tcp_hdr(skb); ++ unsigned char *ptr; ++ int length = (th->doff * 4) - sizeof(struct tcphdr); ++ ++ /* Jump through the options to check whether ADD_ADDR is there */ ++ ptr = (unsigned char *)(th + 1); ++ while (length > 0) { ++ int opcode = *ptr++; ++ int opsize; ++ ++ switch (opcode) { ++ case TCPOPT_EOL: ++ return; ++ case TCPOPT_NOP: ++ length--; ++ continue; ++ default: ++ opsize = *ptr++; ++ if (opsize < 2) ++ return; ++ if (opsize > length) ++ return; /* don't parse partial options */ ++ if (opcode == TCPOPT_MPTCP && ++ ((struct mptcp_option *)ptr)->sub == MPTCP_SUB_ADD_ADDR) { ++#if IS_ENABLED(CONFIG_IPV6) ++ struct mp_add_addr *mpadd = (struct mp_add_addr *)ptr; ++ if ((mpadd->ipver == 4 && opsize != MPTCP_SUB_LEN_ADD_ADDR4 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR4 + 2) || ++ (mpadd->ipver == 6 && opsize != MPTCP_SUB_LEN_ADD_ADDR6 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR6 + 2)) ++#else ++ if (opsize != MPTCP_SUB_LEN_ADD_ADDR4 && ++ opsize != MPTCP_SUB_LEN_ADD_ADDR4 + 2) ++#endif /* CONFIG_IPV6 */ ++ goto cont; ++ ++ mptcp_handle_add_addr(ptr, sk); ++ } ++ if (opcode == TCPOPT_MPTCP && ++ ((struct mptcp_option *)ptr)->sub == MPTCP_SUB_REMOVE_ADDR) { ++ if ((opsize - MPTCP_SUB_LEN_REMOVE_ADDR) < 0) ++ goto cont; ++ ++ mptcp_handle_rem_addr(ptr, sk); ++ } ++cont: ++ ptr += opsize - 2; ++ length -= opsize; ++ } ++ } ++ return; ++} ++ ++static inline int mptcp_mp_fail_rcvd(struct sock *sk, const struct tcphdr *th) ++{ ++ struct mptcp_tcp_sock *mptcp = tcp_sk(sk)->mptcp; ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct mptcp_cb *mpcb = tcp_sk(sk)->mpcb; ++ ++ if (unlikely(mptcp->rx_opt.mp_fail)) { ++ mptcp->rx_opt.mp_fail = 0; ++ ++ if (!th->rst && !mpcb->infinite_mapping_snd) { ++ struct sock *sk_it; ++ ++ mpcb->send_infinite_mapping = 1; ++ /* We resend everything that has not been acknowledged */ ++ meta_sk->sk_send_head = tcp_write_queue_head(meta_sk); ++ ++ /* We artificially restart the whole send-queue. Thus, ++ * it is as if no packets are in flight ++ */ ++ tcp_sk(meta_sk)->packets_out = 0; ++ ++ /* If the snd_nxt already wrapped around, we have to ++ * undo the wrapping, as we are restarting from snd_una ++ * on. ++ */ ++ if (tcp_sk(meta_sk)->snd_nxt < tcp_sk(meta_sk)->snd_una) { ++ mpcb->snd_high_order[mpcb->snd_hiseq_index] -= 2; ++ mpcb->snd_hiseq_index = mpcb->snd_hiseq_index ? 0 : 1; ++ } ++ tcp_sk(meta_sk)->snd_nxt = tcp_sk(meta_sk)->snd_una; ++ ++ /* Trigger a sending on the meta. */ ++ mptcp_push_pending_frames(meta_sk); ++ ++ mptcp_for_each_sk(mpcb, sk_it) { ++ if (sk != sk_it) ++ mptcp_sub_force_close(sk_it); ++ } ++ } ++ ++ return 0; ++ } ++ ++ if (unlikely(mptcp->rx_opt.mp_fclose)) { ++ struct sock *sk_it, *tmpsk; ++ ++ mptcp->rx_opt.mp_fclose = 0; ++ if (mptcp->rx_opt.mptcp_key != mpcb->mptcp_loc_key) ++ return 0; ++ ++ if (tcp_need_reset(sk->sk_state)) ++ tcp_send_active_reset(sk, GFP_ATOMIC); ++ ++ mptcp_for_each_sk_safe(mpcb, sk_it, tmpsk) ++ mptcp_sub_force_close(sk_it); ++ ++ tcp_reset(meta_sk); ++ ++ return 1; ++ } ++ ++ return 0; ++} ++ ++static inline void mptcp_path_array_check(struct sock *meta_sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ ++ if (unlikely(mpcb->list_rcvd)) { ++ mpcb->list_rcvd = 0; ++ if (mpcb->pm_ops->new_remote_address) ++ mpcb->pm_ops->new_remote_address(meta_sk); ++ } ++} ++ ++int mptcp_handle_options(struct sock *sk, const struct tcphdr *th, struct sk_buff *skb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_options_received *mopt = &tp->mptcp->rx_opt; ++ ++ if (tp->mpcb->infinite_mapping_rcv || tp->mpcb->infinite_mapping_snd) ++ return 0; ++ ++ if (mptcp_mp_fail_rcvd(sk, th)) ++ return 1; ++ ++ /* RFC 6824, Section 3.3: ++ * If a checksum is not present when its use has been negotiated, the ++ * receiver MUST close the subflow with a RST as it is considered broken. ++ */ ++ if (mptcp_is_data_seq(skb) && tp->mpcb->dss_csum && ++ !(TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_DSS_CSUM)) { ++ if (tcp_need_reset(sk->sk_state)) ++ tcp_send_active_reset(sk, GFP_ATOMIC); ++ ++ mptcp_sub_force_close(sk); ++ return 1; ++ } ++ ++ /* We have to acknowledge retransmissions of the third ++ * ack. ++ */ ++ if (mopt->join_ack) { ++ tcp_send_delayed_ack(sk); ++ mopt->join_ack = 0; ++ } ++ ++ if (mopt->saw_add_addr || mopt->saw_rem_addr) { ++ if (mopt->more_add_addr || mopt->more_rem_addr) { ++ mptcp_parse_addropt(skb, sk); ++ } else { ++ if (mopt->saw_add_addr) ++ mptcp_handle_add_addr(mopt->add_addr_ptr, sk); ++ if (mopt->saw_rem_addr) ++ mptcp_handle_rem_addr(mopt->rem_addr_ptr, sk); ++ } ++ ++ mopt->more_add_addr = 0; ++ mopt->saw_add_addr = 0; ++ mopt->more_rem_addr = 0; ++ mopt->saw_rem_addr = 0; ++ } ++ if (mopt->saw_low_prio) { ++ if (mopt->saw_low_prio == 1) { ++ tp->mptcp->rcv_low_prio = mopt->low_prio; ++ } else { ++ struct sock *sk_it; ++ mptcp_for_each_sk(tp->mpcb, sk_it) { ++ struct mptcp_tcp_sock *mptcp = tcp_sk(sk_it)->mptcp; ++ if (mptcp->rem_id == mopt->prio_addr_id) ++ mptcp->rcv_low_prio = mopt->low_prio; ++ } ++ } ++ mopt->saw_low_prio = 0; ++ } ++ ++ mptcp_data_ack(sk, skb); ++ ++ mptcp_path_array_check(mptcp_meta_sk(sk)); ++ /* Socket may have been mp_killed by a REMOVE_ADDR */ ++ if (tp->mp_killed) ++ return 1; ++ ++ return 0; ++} ++ ++/* The skptr is needed, because if we become MPTCP-capable, we have to switch ++ * from meta-socket to master-socket. ++ * ++ * @return: 1 - we want to reset this connection ++ * 2 - we want to discard the received syn/ack ++ * 0 - everything is fine - continue ++ */ ++int mptcp_rcv_synsent_state_process(struct sock *sk, struct sock **skptr, ++ struct sk_buff *skb, ++ struct mptcp_options_received *mopt) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ if (tp->mpc) { ++ u8 hash_mac_check[20]; ++ struct mptcp_cb *mpcb = tp->mpcb; ++ ++ mptcp_hmac_sha1((u8 *)&mpcb->mptcp_rem_key, ++ (u8 *)&mpcb->mptcp_loc_key, ++ (u8 *)&tp->mptcp->rx_opt.mptcp_recv_nonce, ++ (u8 *)&tp->mptcp->mptcp_loc_nonce, ++ (u32 *)hash_mac_check); ++ if (memcmp(hash_mac_check, ++ (char *)&tp->mptcp->rx_opt.mptcp_recv_tmac, 8)) { ++ mptcp_sub_force_close(sk); ++ return 1; ++ } ++ ++ /* Set this flag in order to postpone data sending ++ * until the 4th ack arrives. ++ */ ++ tp->mptcp->pre_established = 1; ++ tp->mptcp->rcv_low_prio = tp->mptcp->rx_opt.low_prio; ++ ++ mptcp_hmac_sha1((u8 *)&mpcb->mptcp_loc_key, ++ (u8 *)&mpcb->mptcp_rem_key, ++ (u8 *)&tp->mptcp->mptcp_loc_nonce, ++ (u8 *)&tp->mptcp->rx_opt.mptcp_recv_nonce, ++ (u32 *)&tp->mptcp->sender_mac[0]); ++ ++ } else if (mopt->saw_mpc) { ++ if (mptcp_create_master_sk(sk, mopt->mptcp_key, ++ ntohs(tcp_hdr(skb)->window))) ++ return 2; ++ ++ sk = tcp_sk(sk)->mpcb->master_sk; ++ *skptr = sk; ++ tp = tcp_sk(sk); ++ ++ /* snd_nxt - 1, because it has been incremented ++ * by tcp_connect for the SYN ++ */ ++ tp->mptcp->snt_isn = tp->snd_nxt - 1; ++ tp->mpcb->dss_csum = mopt->dss_csum; ++ tp->mptcp->include_mpc = 1; ++ ++ sk_set_socket(sk, mptcp_meta_sk(sk)->sk_socket); ++ sk->sk_wq = mptcp_meta_sk(sk)->sk_wq; ++ ++ mptcp_update_metasocket(sk, mptcp_meta_sk(sk)); ++ ++ /* hold in mptcp_inherit_sk due to initialization to 2 */ ++ sock_put(sk); ++ } else { ++ tp->request_mptcp = 0; ++ ++ if (tp->inside_tk_table) ++ mptcp_hash_remove(tp); ++ } ++ ++ if (tp->mpc) ++ tp->mptcp->rcv_isn = TCP_SKB_CB(skb)->seq; ++ ++ return 0; ++} ++ ++bool mptcp_should_expand_sndbuf(const struct sock *sk) ++{ ++ struct sock *sk_it; ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ int cnt_backups = 0; ++ int backup_available = 0; ++ ++ /* We circumvent this check in tcp_check_space, because we want to ++ * always call sk_write_space. So, we reproduce the check here. ++ */ ++ if (!meta_sk->sk_socket || ++ !test_bit(SOCK_NOSPACE, &meta_sk->sk_socket->flags)) ++ return false; ++ ++ /* If the user specified a specific send buffer setting, do ++ * not modify it. ++ */ ++ if (meta_sk->sk_userlocks & SOCK_SNDBUF_LOCK) ++ return false; ++ ++ /* If we are under global TCP memory pressure, do not expand. */ ++ if (sk_under_memory_pressure(meta_sk)) ++ return false; ++ ++ /* If we are under soft global TCP memory pressure, do not expand. */ ++ if (sk_memory_allocated(meta_sk) >= sk_prot_mem_limits(meta_sk, 0)) ++ return false; ++ ++ ++ /* For MPTCP we look for a subsocket that could send data. ++ * If we found one, then we update the send-buffer. ++ */ ++ mptcp_for_each_sk(meta_tp->mpcb, sk_it) { ++ struct tcp_sock *tp_it = tcp_sk(sk_it); ++ ++ if (!mptcp_sk_can_send(sk_it)) ++ continue; ++ ++ /* Backup-flows have to be counted - if there is no other ++ * subflow we take the backup-flow into account. */ ++ if (tp_it->mptcp->rcv_low_prio || tp_it->mptcp->low_prio) { ++ cnt_backups++; ++ } ++ ++ if (tp_it->packets_out < tp_it->snd_cwnd) { ++ if (tp_it->mptcp->rcv_low_prio || tp_it->mptcp->low_prio) { ++ backup_available = 1; ++ continue; ++ } ++ return true; ++ } ++ } ++ ++ /* Backup-flow is available for sending - update send-buffer */ ++ if (meta_tp->mpcb->cnt_established == cnt_backups && backup_available) ++ return true; ++ return false; ++} ++ ++void mptcp_init_buffer_space(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ int space; ++ ++ tcp_init_buffer_space(sk); ++ ++ if (is_master_tp(tp)) { ++ /* If there is only one subflow, we just use regular TCP ++ * autotuning. User-locks are handled already by ++ * tcp_init_buffer_space ++ */ ++ meta_tp->window_clamp = tp->window_clamp; ++ meta_tp->rcv_ssthresh = tp->rcv_ssthresh; ++ meta_sk->sk_rcvbuf = sk->sk_rcvbuf; ++ meta_sk->sk_sndbuf = sk->sk_sndbuf; ++ ++ return; ++ } ++ ++ if (meta_sk->sk_userlocks & SOCK_RCVBUF_LOCK) ++ goto snd_buf; ++ ++ /* Adding a new subflow to the rcv-buffer space. We make a simple ++ * addition, to give some space to allow traffic on the new subflow. ++ * Autotuning will increase it further later on. ++ */ ++ space = min(meta_sk->sk_rcvbuf + sk->sk_rcvbuf, sysctl_tcp_rmem[2]); ++ if (space > meta_sk->sk_rcvbuf) { ++ meta_tp->window_clamp += tp->window_clamp; ++ meta_tp->rcv_ssthresh += tp->rcv_ssthresh; ++ meta_sk->sk_rcvbuf = space; ++ } ++ ++snd_buf: ++ if (meta_sk->sk_userlocks & SOCK_SNDBUF_LOCK) ++ return; ++ ++ /* Adding a new subflow to the send-buffer space. We make a simple ++ * addition, to give some space to allow traffic on the new subflow. ++ * Autotuning will increase it further later on. ++ */ ++ space = min(meta_sk->sk_sndbuf + sk->sk_sndbuf, sysctl_tcp_wmem[2]); ++ if (space > meta_sk->sk_sndbuf) { ++ meta_sk->sk_sndbuf = space; ++ meta_sk->sk_write_space(meta_sk); ++ } ++} ++ ++void mptcp_tcp_set_rto(struct sock *sk) ++{ ++ tcp_set_rto(sk); ++ mptcp_set_rto(sk); ++} +diff --git a/net/mptcp/mptcp_ipv4.c b/net/mptcp/mptcp_ipv4.c +new file mode 100644 +index 0000000..b6053f1 +--- /dev/null ++++ b/net/mptcp/mptcp_ipv4.c +@@ -0,0 +1,603 @@ ++/* ++ * MPTCP implementation - IPv4-specific functions ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#include <linux/export.h> ++#include <linux/ip.h> ++#include <linux/list.h> ++#include <linux/skbuff.h> ++#include <linux/spinlock.h> ++#include <linux/tcp.h> ++ ++#include <net/inet_common.h> ++#include <net/inet_connection_sock.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#include <net/request_sock.h> ++#include <net/tcp.h> ++ ++u32 mptcp_v4_get_nonce(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport, ++ u32 seq) ++{ ++ u32 hash[MD5_DIGEST_WORDS]; ++ ++ hash[0] = (__force u32)saddr; ++ hash[1] = (__force u32)daddr; ++ hash[2] = ((__force u16)sport << 16) + (__force u16)dport; ++ hash[3] = seq; ++ ++ md5_transform(hash, mptcp_secret); ++ ++ return hash[0]; ++} ++ ++u64 mptcp_v4_get_key(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport) ++{ ++ u32 hash[MD5_DIGEST_WORDS]; ++ ++ hash[0] = (__force u32)saddr; ++ hash[1] = (__force u32)daddr; ++ hash[2] = ((__force u16)sport << 16) + (__force u16)dport; ++ hash[3] = mptcp_key_seed++; ++ ++ md5_transform(hash, mptcp_secret); ++ ++ return *((u64 *)hash); ++} ++ ++ ++static void mptcp_v4_reqsk_destructor(struct request_sock *req) ++{ ++ mptcp_reqsk_destructor(req); ++ ++ tcp_v4_reqsk_destructor(req); ++} ++ ++/* Similar to tcp_request_sock_ops */ ++struct request_sock_ops mptcp_request_sock_ops __read_mostly = { ++ .family = PF_INET, ++ .obj_size = sizeof(struct mptcp_request_sock), ++ .rtx_syn_ack = tcp_v4_rtx_synack, ++ .send_ack = tcp_v4_reqsk_send_ack, ++ .destructor = mptcp_v4_reqsk_destructor, ++ .send_reset = tcp_v4_send_reset, ++ .syn_ack_timeout = tcp_syn_ack_timeout, ++}; ++ ++static void mptcp_v4_reqsk_queue_hash_add(struct sock *meta_sk, ++ struct request_sock *req, ++ unsigned long timeout) ++{ ++ const u32 h1 = inet_synq_hash(inet_rsk(req)->ir_rmt_addr, ++ inet_rsk(req)->ir_rmt_port, ++ 0, MPTCP_HASH_SIZE); ++ /* We cannot call inet_csk_reqsk_queue_hash_add(), because we do not ++ * want to reset the keepalive-timer (responsible for retransmitting ++ * SYN/ACKs). We do not retransmit SYN/ACKs+MP_JOINs, because we cannot ++ * overload the keepalive timer. Also, it's not a big deal, because the ++ * third ACK of the MP_JOIN-handshake is sent in a reliable manner. So, ++ * if the third ACK gets lost, the client will handle the retransmission ++ * anyways. If our SYN/ACK gets lost, the client will retransmit the ++ * SYN. ++ */ ++ struct inet_connection_sock *meta_icsk = inet_csk(meta_sk); ++ struct listen_sock *lopt = meta_icsk->icsk_accept_queue.listen_opt; ++ const u32 h2 = inet_synq_hash(inet_rsk(req)->ir_rmt_addr, ++ inet_rsk(req)->ir_rmt_port, ++ lopt->hash_rnd, lopt->nr_table_entries); ++ ++ reqsk_queue_hash_req(&meta_icsk->icsk_accept_queue, h2, req, timeout); ++ reqsk_queue_added(&meta_icsk->icsk_accept_queue); ++ ++ spin_lock(&mptcp_reqsk_hlock); ++ list_add(&mptcp_rsk(req)->collide_tuple, &mptcp_reqsk_htb[h1]); ++ spin_unlock(&mptcp_reqsk_hlock); ++} ++ ++/* Similar to tcp_v4_conn_request */ ++static void mptcp_v4_join_request(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct tcp_options_received tmp_opt; ++ struct mptcp_options_received mopt; ++ struct request_sock *req; ++ struct inet_request_sock *ireq; ++ struct mptcp_request_sock *mtreq; ++ struct dst_entry *dst = NULL; ++ u8 mptcp_hash_mac[20]; ++ __be32 saddr = ip_hdr(skb)->saddr; ++ __be32 daddr = ip_hdr(skb)->daddr; ++ __u32 isn = TCP_SKB_CB(skb)->when; ++ int want_cookie = 0; ++ union inet_addr addr; ++ ++ tcp_clear_options(&tmp_opt); ++ mptcp_init_mp_opt(&mopt); ++ tmp_opt.mss_clamp = TCP_MSS_DEFAULT; ++ tmp_opt.user_mss = tcp_sk(meta_sk)->rx_opt.user_mss; ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, NULL); ++ ++ req = inet_reqsk_alloc(&mptcp_request_sock_ops); ++ if (!req) ++ return; ++ ++#ifdef CONFIG_TCP_MD5SIG ++ tcp_rsk(req)->af_specific = &tcp_request_sock_ipv4_ops; ++#endif ++ ++ tmp_opt.tstamp_ok = tmp_opt.saw_tstamp; ++ tcp_openreq_init(req, &tmp_opt, skb); ++ ++ ireq = inet_rsk(req); ++ ireq->ir_loc_addr = daddr; ++ ireq->ir_rmt_addr = saddr; ++ ireq->no_srccheck = inet_sk(meta_sk)->transparent; ++ ireq->opt = tcp_v4_save_options(skb); ++ ++ if (security_inet_conn_request(meta_sk, skb, req)) ++ goto drop_and_free; ++ ++ if (!want_cookie || tmp_opt.tstamp_ok) ++ TCP_ECN_create_request(req, skb, sock_net(meta_sk)); ++ ++ if (!isn) { ++ struct flowi4 fl4; ++ ++ /* VJ's idea. We save last timestamp seen ++ * from the destination in peer table, when entering ++ * state TIME-WAIT, and check against it before ++ * accepting new connection request. ++ * ++ * If "isn" is not zero, this request hit alive ++ * timewait bucket, so that all the necessary checks ++ * are made in the function processing timewait state. ++ */ ++ if (tmp_opt.saw_tstamp && ++ tcp_death_row.sysctl_tw_recycle && ++ (dst = inet_csk_route_req(meta_sk, &fl4, req)) != NULL && ++ fl4.daddr == saddr) { ++ if (!tcp_peer_is_proven(req, dst, true)) { ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_PAWSPASSIVEREJECTED); ++ goto drop_and_release; ++ } ++ } ++ /* Kill the following clause, if you dislike this way. */ ++ else if (!sysctl_tcp_syncookies && ++ (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(meta_sk) < ++ (sysctl_max_syn_backlog >> 2)) && ++ !tcp_peer_is_proven(req, dst, false)) { ++ /* Without syncookies last quarter of ++ * backlog is filled with destinations, ++ * proven to be alive. ++ * It means that we continue to communicate ++ * to destinations, already remembered ++ * to the moment of synflood. ++ */ ++ LIMIT_NETDEBUG(KERN_DEBUG pr_fmt("drop open request from %pI4/%u\n"), ++ &saddr, ntohs(tcp_hdr(skb)->source)); ++ goto drop_and_release; ++ } ++ ++ isn = tcp_v4_init_sequence(skb); ++ } ++ tcp_rsk(req)->snt_isn = isn; ++ tcp_rsk(req)->snt_synack = tcp_time_stamp; ++ tcp_rsk(req)->listener = NULL; ++ ++ mtreq = mptcp_rsk(req); ++ mtreq->mpcb = mpcb; ++ INIT_LIST_HEAD(&mtreq->collide_tuple); ++ mtreq->mptcp_rem_nonce = mopt.mptcp_recv_nonce; ++ mtreq->mptcp_rem_key = mpcb->mptcp_rem_key; ++ mtreq->mptcp_loc_key = mpcb->mptcp_loc_key; ++ mtreq->mptcp_loc_nonce = mptcp_v4_get_nonce(saddr, daddr, ++ tcp_hdr(skb)->source, ++ tcp_hdr(skb)->dest, isn); ++ mptcp_hmac_sha1((u8 *)&mtreq->mptcp_loc_key, ++ (u8 *)&mtreq->mptcp_rem_key, ++ (u8 *)&mtreq->mptcp_loc_nonce, ++ (u8 *)&mtreq->mptcp_rem_nonce, (u32 *)mptcp_hash_mac); ++ mtreq->mptcp_hash_tmac = *(u64 *)mptcp_hash_mac; ++ ++ addr.ip = ireq->ir_loc_addr; ++ mtreq->loc_id = mpcb->pm_ops->get_local_id(AF_INET, &addr, sock_net(meta_sk)); ++ if (mtreq->loc_id == -1) /* Address not part of the allowed ones */ ++ goto drop_and_release; ++ mtreq->rem_id = mopt.rem_id; ++ mtreq->low_prio = mopt.low_prio; ++ tcp_rsk(req)->saw_mpc = 1; ++ ++ if (tcp_v4_send_synack(meta_sk, dst, req, skb_get_queue_mapping(skb))) ++ goto drop_and_free; ++ ++ /* Adding to request queue in metasocket */ ++ mptcp_v4_reqsk_queue_hash_add(meta_sk, req, TCP_TIMEOUT_INIT); ++ ++ return; ++ ++drop_and_release: ++ dst_release(dst); ++drop_and_free: ++ reqsk_free(req); ++ return; ++} ++ ++int mptcp_v4_rem_raddress(struct mptcp_cb *mpcb, u8 id) ++{ ++ int i; ++ ++ for (i = 0; i < MPTCP_MAX_ADDR; i++) { ++ if (!((1 << i) & mpcb->rem4_bits)) ++ continue; ++ ++ if (mpcb->remaddr4[i].rem4_id == id) { ++ /* remove address from bitfield */ ++ mpcb->rem4_bits &= ~(1 << i); ++ ++ return 0; ++ } ++ } ++ ++ return -1; ++} ++ ++/* Based on function tcp_v4_conn_request (tcp_ipv4.c) ++ * Returns -1 if there is no space anymore to store an additional ++ * address ++ */ ++int mptcp_v4_add_raddress(struct mptcp_cb *mpcb, const struct in_addr *addr, ++ __be16 port, u8 id) ++{ ++ int i; ++ struct mptcp_rem4 *rem4; ++ ++ mptcp_for_each_bit_set(mpcb->rem4_bits, i) { ++ rem4 = &mpcb->remaddr4[i]; ++ ++ /* Address is already in the list --- continue */ ++ if (rem4->rem4_id == id && ++ rem4->addr.s_addr == addr->s_addr && rem4->port == port) ++ return 0; ++ ++ /* This may be the case, when the peer is behind a NAT. He is ++ * trying to JOIN, thus sending the JOIN with a certain ID. ++ * However the src_addr of the IP-packet has been changed. We ++ * update the addr in the list, because this is the address as ++ * OUR BOX sees it. ++ */ ++ if (rem4->rem4_id == id && rem4->addr.s_addr != addr->s_addr) { ++ /* update the address */ ++ mptcp_debug("%s: updating old addr:%pI4 to addr %pI4 with id:%d\n", ++ __func__, &rem4->addr.s_addr, ++ &addr->s_addr, id); ++ rem4->addr.s_addr = addr->s_addr; ++ rem4->port = port; ++ mpcb->list_rcvd = 1; ++ return 0; ++ } ++ } ++ ++ i = mptcp_find_free_index(mpcb->rem4_bits); ++ /* Do we have already the maximum number of local/remote addresses? */ ++ if (i < 0) { ++ mptcp_debug("%s: At max num of remote addresses: %d --- not adding address: %pI4\n", ++ __func__, MPTCP_MAX_ADDR, &addr->s_addr); ++ return -1; ++ } ++ ++ rem4 = &mpcb->remaddr4[i]; ++ ++ /* Address is not known yet, store it */ ++ rem4->addr.s_addr = addr->s_addr; ++ rem4->port = port; ++ rem4->bitfield = 0; ++ rem4->retry_bitfield = 0; ++ rem4->rem4_id = id; ++ mpcb->list_rcvd = 1; ++ mpcb->rem4_bits |= (1 << i); ++ ++ return 0; ++} ++ ++/* Sets the bitfield of the remote-address field ++ * local address is not set as it will disappear with the global address-list ++ */ ++void mptcp_v4_set_init_addr_bit(struct mptcp_cb *mpcb, __be32 daddr, int index) ++{ ++ int i; ++ ++ mptcp_for_each_bit_set(mpcb->rem4_bits, i) { ++ if (mpcb->remaddr4[i].addr.s_addr == daddr) { ++ mpcb->remaddr4[i].bitfield |= (1 << index); ++ return; ++ } ++ } ++} ++ ++/* We only process join requests here. (either the SYN or the final ACK) */ ++int mptcp_v4_do_rcv(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct sock *child, *rsk = NULL; ++ int ret; ++ ++ if (!(TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_JOIN)) { ++ struct tcphdr *th = tcp_hdr(skb); ++ const struct iphdr *iph = ip_hdr(skb); ++ struct sock *sk; ++ ++ sk = inet_lookup_established(sock_net(meta_sk), &tcp_hashinfo, ++ iph->saddr, th->source, iph->daddr, ++ th->dest, inet_iif(skb)); ++ ++ if (!sk) { ++ kfree_skb(skb); ++ return 0; ++ } ++ if (is_meta_sk(sk)) { ++ WARN("%s Did not find a sub-sk - did found the meta!\n", __func__); ++ kfree_skb(skb); ++ sock_put(sk); ++ return 0; ++ } ++ ++ if (sk->sk_state == TCP_TIME_WAIT) { ++ inet_twsk_put(inet_twsk(sk)); ++ kfree_skb(skb); ++ return 0; ++ } ++ ++ ret = tcp_v4_do_rcv(sk, skb); ++ sock_put(sk); ++ ++ return ret; ++ } ++ TCP_SKB_CB(skb)->mptcp_flags = 0; ++ ++ /* Has been removed from the tk-table. Thus, no new subflows. ++ * ++ * Check for close-state is necessary, because we may have been closed ++ * without passing by mptcp_close(). ++ * ++ * When falling back, no new subflows are allowed either. ++ */ ++ if (meta_sk->sk_state == TCP_CLOSE || !tcp_sk(meta_sk)->inside_tk_table || ++ mpcb->infinite_mapping_rcv || mpcb->send_infinite_mapping) ++ goto reset_and_discard; ++ ++ child = tcp_v4_hnd_req(meta_sk, skb); ++ ++ if (!child) ++ goto discard; ++ ++ if (child != meta_sk) { ++ sock_rps_save_rxhash(child, skb); ++ /* We don't call tcp_child_process here, because we hold ++ * already the meta-sk-lock and are sure that it is not owned ++ * by the user. ++ */ ++ ret = tcp_rcv_state_process(child, skb, tcp_hdr(skb), skb->len); ++ bh_unlock_sock(child); ++ sock_put(child); ++ if (ret) { ++ rsk = child; ++ goto reset_and_discard; ++ } ++ } else { ++ if (tcp_hdr(skb)->syn) { ++ struct mp_join *join_opt = mptcp_find_join(skb); ++ /* Currently we make two calls to mptcp_find_join(). This ++ * can probably be optimized. ++ */ ++ if (mptcp_v4_add_raddress(mpcb, ++ (struct in_addr *)&ip_hdr(skb)->saddr, ++ 0, ++ join_opt->addr_id) < 0) ++ goto reset_and_discard; ++ mpcb->list_rcvd = 0; ++ ++ mptcp_v4_join_request(meta_sk, skb); ++ goto discard; ++ } ++ goto reset_and_discard; ++ } ++ return 0; ++ ++reset_and_discard: ++ tcp_v4_send_reset(rsk, skb); ++discard: ++ kfree_skb(skb); ++ return 0; ++} ++ ++/* After this, the ref count of the meta_sk associated with the request_sock ++ * is incremented. Thus it is the responsibility of the caller ++ * to call sock_put() when the reference is not needed anymore. ++ */ ++struct sock *mptcp_v4_search_req(const __be16 rport, const __be32 raddr, ++ const __be32 laddr, const struct net *net) ++{ ++ struct mptcp_request_sock *mtreq; ++ struct sock *meta_sk = NULL; ++ ++ spin_lock(&mptcp_reqsk_hlock); ++ list_for_each_entry(mtreq, ++ &mptcp_reqsk_htb[inet_synq_hash(raddr, rport, 0, ++ MPTCP_HASH_SIZE)], ++ collide_tuple) { ++ struct inet_request_sock *ireq = inet_rsk(rev_mptcp_rsk(mtreq)); ++ meta_sk = mtreq->mpcb->meta_sk; ++ ++ if (ireq->ir_rmt_port == rport && ++ ireq->ir_rmt_addr == raddr && ++ ireq->ir_loc_addr == laddr && ++ rev_mptcp_rsk(mtreq)->rsk_ops->family == AF_INET && ++ net_eq(net, sock_net(meta_sk))) ++ break; ++ meta_sk = NULL; ++ } ++ ++ if (meta_sk && unlikely(!atomic_inc_not_zero(&meta_sk->sk_refcnt))) ++ meta_sk = NULL; ++ spin_unlock(&mptcp_reqsk_hlock); ++ ++ return meta_sk; ++} ++ ++/* Create a new IPv4 subflow. ++ * ++ * We are in user-context and meta-sock-lock is hold. ++ */ ++int mptcp_init4_subsockets(struct sock *meta_sk, const struct mptcp_loc4 *loc, ++ struct mptcp_rem4 *rem) ++{ ++ struct tcp_sock *tp; ++ struct sock *sk; ++ struct sockaddr_in loc_in, rem_in; ++ struct socket sock; ++ int ulid_size = 0, ret; ++ ++ /** First, create and prepare the new socket */ ++ ++ sock.type = meta_sk->sk_socket->type; ++ sock.state = SS_UNCONNECTED; ++ sock.wq = meta_sk->sk_socket->wq; ++ sock.file = meta_sk->sk_socket->file; ++ sock.ops = NULL; ++ ++ ret = inet_create(sock_net(meta_sk), &sock, IPPROTO_TCP, 1); ++ if (unlikely(ret < 0)) { ++ mptcp_debug("%s inet_create failed ret: %d\n", __func__, ret); ++ return ret; ++ } ++ ++ sk = sock.sk; ++ tp = tcp_sk(sk); ++ ++ /* All subsockets need the MPTCP-lock-class */ ++ lockdep_set_class_and_name(&(sk)->sk_lock.slock, &meta_slock_key, "slock-AF_INET-MPTCP"); ++ lockdep_init_map(&(sk)->sk_lock.dep_map, "sk_lock-AF_INET-MPTCP", &meta_key, 0); ++ ++ if (mptcp_add_sock(meta_sk, sk, loc->loc4_id, rem->rem4_id, GFP_KERNEL)) ++ goto error; ++ ++ tp->mptcp->slave_sk = 1; ++ tp->mptcp->low_prio = loc->low_prio; ++ ++ /* Initializing the timer for an MPTCP subflow */ ++ setup_timer(&tp->mptcp->mptcp_ack_timer, mptcp_ack_handler, (unsigned long)sk); ++ ++ /** Then, connect the socket to the peer */ ++ ++ ulid_size = sizeof(struct sockaddr_in); ++ loc_in.sin_family = AF_INET; ++ rem_in.sin_family = AF_INET; ++ loc_in.sin_port = 0; ++ if (rem->port) ++ rem_in.sin_port = rem->port; ++ else ++ rem_in.sin_port = inet_sk(meta_sk)->inet_dport; ++ loc_in.sin_addr = loc->addr; ++ rem_in.sin_addr = rem->addr; ++ ++ ret = sock.ops->bind(&sock, (struct sockaddr *)&loc_in, ulid_size); ++ if (ret < 0) { ++ mptcp_debug("%s: MPTCP subsocket bind() failed, error %d\n", ++ __func__, ret); ++ goto error; ++ } ++ ++ mptcp_debug("%s: token %#x pi %d src_addr:%pI4:%d dst_addr:%pI4:%d\n", ++ __func__, tcp_sk(meta_sk)->mpcb->mptcp_loc_token, ++ tp->mptcp->path_index, &loc_in.sin_addr, ++ ntohs(loc_in.sin_port), &rem_in.sin_addr, ++ ntohs(rem_in.sin_port)); ++ ++ ret = sock.ops->connect(&sock, (struct sockaddr *)&rem_in, ++ ulid_size, O_NONBLOCK); ++ if (ret < 0 && ret != -EINPROGRESS) { ++ mptcp_debug("%s: MPTCP subsocket connect() failed, error %d\n", ++ __func__, ret); ++ goto error; ++ } ++ ++ sk_set_socket(sk, meta_sk->sk_socket); ++ sk->sk_wq = meta_sk->sk_wq; ++ ++ return 0; ++ ++error: ++ /* May happen if mptcp_add_sock fails first */ ++ if (!tp->mpc) { ++ tcp_close(sk, 0); ++ } else { ++ local_bh_disable(); ++ mptcp_sub_force_close(sk); ++ local_bh_enable(); ++ } ++ return ret; ++} ++EXPORT_SYMBOL(mptcp_init4_subsockets); ++ ++/* General initialization of IPv4 for MPTCP */ ++int mptcp_pm_v4_init(void) ++{ ++ int ret = 0; ++ struct request_sock_ops *ops = &mptcp_request_sock_ops; ++ ++ ops->slab_name = kasprintf(GFP_KERNEL, "request_sock_%s", "MPTCP"); ++ if (ops->slab_name == NULL) { ++ ret = -ENOMEM; ++ goto out; ++ } ++ ++ ops->slab = kmem_cache_create(ops->slab_name, ops->obj_size, 0, ++ SLAB_DESTROY_BY_RCU|SLAB_HWCACHE_ALIGN, ++ NULL); ++ ++ if (ops->slab == NULL) { ++ ret = -ENOMEM; ++ goto err_reqsk_create; ++ } ++ ++out: ++ return ret; ++ ++err_reqsk_create: ++ kfree(ops->slab_name); ++ ops->slab_name = NULL; ++ goto out; ++} ++ ++void mptcp_pm_v4_undo(void) ++{ ++ kmem_cache_destroy(mptcp_request_sock_ops.slab); ++ kfree(mptcp_request_sock_ops.slab_name); ++} ++ ++ +diff --git a/net/mptcp/mptcp_ipv6.c b/net/mptcp/mptcp_ipv6.c +new file mode 100644 +index 0000000..b6b444d +--- /dev/null ++++ b/net/mptcp/mptcp_ipv6.c +@@ -0,0 +1,822 @@ ++/* ++ * MPTCP implementation - IPv6-specific functions ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#include <linux/export.h> ++#include <linux/in6.h> ++#include <linux/kernel.h> ++ ++#include <net/addrconf.h> ++#include <net/flow.h> ++#include <net/inet6_connection_sock.h> ++#include <net/inet6_hashtables.h> ++#include <net/inet_common.h> ++#include <net/ipv6.h> ++#include <net/ip6_checksum.h> ++#include <net/ip6_route.h> ++#include <net/mptcp.h> ++#include <net/mptcp_v6.h> ++#include <net/tcp.h> ++#include <net/transp_v6.h> ++ ++static int mptcp_v6v4_send_synack(struct sock *meta_sk, struct request_sock *req, ++ u16 queue_mapping); ++ ++__u32 mptcp_v6_get_nonce(const __be32 *saddr, const __be32 *daddr, ++ __be16 sport, __be16 dport, u32 seq) ++{ ++ u32 secret[MD5_MESSAGE_BYTES / 4]; ++ u32 hash[MD5_DIGEST_WORDS]; ++ u32 i; ++ ++ memcpy(hash, saddr, 16); ++ for (i = 0; i < 4; i++) ++ secret[i] = mptcp_secret[i] + (__force u32)daddr[i]; ++ secret[4] = mptcp_secret[4] + ++ (((__force u16)sport << 16) + (__force u16)dport); ++ secret[5] = seq; ++ for (i = 6; i < MD5_MESSAGE_BYTES / 4; i++) ++ secret[i] = mptcp_secret[i]; ++ ++ md5_transform(hash, secret); ++ ++ return hash[0]; ++} ++ ++u64 mptcp_v6_get_key(const __be32 *saddr, const __be32 *daddr, ++ __be16 sport, __be16 dport) ++{ ++ u32 secret[MD5_MESSAGE_BYTES / 4]; ++ u32 hash[MD5_DIGEST_WORDS]; ++ u32 i; ++ ++ memcpy(hash, saddr, 16); ++ for (i = 0; i < 4; i++) ++ secret[i] = mptcp_secret[i] + (__force u32)daddr[i]; ++ secret[4] = mptcp_secret[4] + ++ (((__force u16)sport << 16) + (__force u16)dport); ++ secret[5] = mptcp_key_seed++; ++ for (i = 5; i < MD5_MESSAGE_BYTES / 4; i++) ++ secret[i] = mptcp_secret[i]; ++ ++ md5_transform(hash, secret); ++ ++ return *((u64 *)hash); ++} ++ ++static void mptcp_v6_reqsk_destructor(struct request_sock *req) ++{ ++ mptcp_reqsk_destructor(req); ++ ++ tcp_v6_reqsk_destructor(req); ++} ++ ++/* Similar to tcp_v6_rtx_synack */ ++static int mptcp_v6_rtx_synack(struct sock *meta_sk, struct request_sock *req) ++{ ++ if (meta_sk->sk_family == AF_INET6) ++ return tcp_v6_rtx_synack(meta_sk, req); ++ ++ TCP_INC_STATS_BH(sock_net(meta_sk), TCP_MIB_RETRANSSEGS); ++ return mptcp_v6v4_send_synack(meta_sk, req, 0); ++} ++ ++/* Similar to tcp6_request_sock_ops */ ++struct request_sock_ops mptcp6_request_sock_ops __read_mostly = { ++ .family = AF_INET6, ++ .obj_size = sizeof(struct mptcp_request_sock), ++ .rtx_syn_ack = mptcp_v6_rtx_synack, ++ .send_ack = tcp_v6_reqsk_send_ack, ++ .destructor = mptcp_v6_reqsk_destructor, ++ .send_reset = tcp_v6_send_reset, ++ .syn_ack_timeout = tcp_syn_ack_timeout, ++}; ++ ++static void mptcp_v6_reqsk_queue_hash_add(struct sock *meta_sk, ++ struct request_sock *req, ++ unsigned long timeout) ++{ ++ const u32 h1 = inet6_synq_hash(&inet_rsk(req)->ir_v6_rmt_addr, ++ inet_rsk(req)->ir_rmt_port, ++ 0, MPTCP_HASH_SIZE); ++ /* We cannot call inet6_csk_reqsk_queue_hash_add(), because we do not ++ * want to reset the keepalive-timer (responsible for retransmitting ++ * SYN/ACKs). We do not retransmit SYN/ACKs+MP_JOINs, because we cannot ++ * overload the keepalive timer. Also, it's not a big deal, because the ++ * third ACK of the MP_JOIN-handshake is sent in a reliable manner. So, ++ * if the third ACK gets lost, the client will handle the retransmission ++ * anyways. If our SYN/ACK gets lost, the client will retransmit the ++ * SYN. ++ */ ++ struct inet_connection_sock *meta_icsk = inet_csk(meta_sk); ++ struct listen_sock *lopt = meta_icsk->icsk_accept_queue.listen_opt; ++ const u32 h2 = inet6_synq_hash(&inet_rsk(req)->ir_v6_rmt_addr, ++ inet_rsk(req)->ir_rmt_port, ++ lopt->hash_rnd, lopt->nr_table_entries); ++ ++ reqsk_queue_hash_req(&meta_icsk->icsk_accept_queue, h2, req, timeout); ++ reqsk_queue_added(&meta_icsk->icsk_accept_queue); ++ ++ spin_lock(&mptcp_reqsk_hlock); ++ list_add(&mptcp_rsk(req)->collide_tuple, &mptcp_reqsk_htb[h1]); ++ spin_unlock(&mptcp_reqsk_hlock); ++} ++ ++/* Similar to tcp_v6_send_synack ++ * ++ * The meta-socket is IPv4, but a new subsocket is IPv6 ++ */ ++static int mptcp_v6v4_send_synack(struct sock *meta_sk, struct request_sock *req, ++ u16 queue_mapping) ++{ ++ struct inet_request_sock *treq = inet_rsk(req); ++ struct sk_buff *skb; ++ struct flowi6 fl6; ++ struct dst_entry *dst; ++ int err = -ENOMEM; ++ ++ memset(&fl6, 0, sizeof(fl6)); ++ fl6.flowi6_proto = IPPROTO_TCP; ++ fl6.daddr = treq->ir_v6_rmt_addr; ++ fl6.saddr = treq->ir_v6_loc_addr; ++ fl6.flowlabel = 0; ++ fl6.flowi6_oif = treq->ir_iif; ++ fl6.flowi6_mark = meta_sk->sk_mark; ++ fl6.fl6_dport = inet_rsk(req)->ir_rmt_port; ++ fl6.fl6_sport = htons(inet_rsk(req)->ir_num); ++ security_req_classify_flow(req, flowi6_to_flowi(&fl6)); ++ ++ dst = ip6_dst_lookup_flow(meta_sk, &fl6, NULL); ++ if (IS_ERR(dst)) { ++ err = PTR_ERR(dst); ++ return err; ++ } ++ skb = tcp_make_synack(meta_sk, dst, req, NULL); ++ ++ if (skb) { ++ __tcp_v6_send_check(skb, &treq->ir_v6_loc_addr, ++ &treq->ir_v6_rmt_addr); ++ ++ fl6.daddr = treq->ir_v6_rmt_addr; ++ skb_set_queue_mapping(skb, queue_mapping); ++ err = ip6_xmit(meta_sk, skb, &fl6, NULL, 0); ++ err = net_xmit_eval(err); ++ } ++ ++ return err; ++} ++ ++/* Similar to tcp_v6_syn_recv_sock ++ * ++ * The meta-socket is IPv4, but a new subsocket is IPv6 ++ */ ++struct sock *mptcp_v6v4_syn_recv_sock(struct sock *meta_sk, struct sk_buff *skb, ++ struct request_sock *req, ++ struct dst_entry *dst) ++{ ++ struct inet_request_sock *treq; ++ struct ipv6_pinfo *newnp; ++ struct tcp6_sock *newtcp6sk; ++ struct inet_sock *newinet; ++ struct tcp_sock *newtp; ++ struct sock *newsk; ++ ++ treq = inet_rsk(req); ++ ++ if (sk_acceptq_is_full(meta_sk)) ++ goto out_overflow; ++ ++ if (!dst) { ++ /* This code is similar to inet6_csk_route_req, but as we ++ * don't have a np-pointer in the meta, we have to do it ++ * manually. ++ */ ++ struct flowi6 fl6; ++ ++ memset(&fl6, 0, sizeof(fl6)); ++ fl6.flowi6_proto = IPPROTO_TCP; ++ fl6.daddr = treq->ir_v6_rmt_addr; ++ fl6.saddr = treq->ir_v6_loc_addr; ++ fl6.flowi6_oif = treq->ir_iif; ++ fl6.flowi6_mark = meta_sk->sk_mark; ++ fl6.fl6_dport = inet_rsk(req)->ir_rmt_port; ++ fl6.fl6_sport = htons(inet_rsk(req)->ir_num); ++ security_req_classify_flow(req, flowi6_to_flowi(&fl6)); ++ ++ dst = ip6_dst_lookup_flow(meta_sk, &fl6, NULL); ++ if (IS_ERR(dst)) ++ goto out; ++ } ++ ++ newsk = tcp_create_openreq_child(meta_sk, req, skb); ++ if (newsk == NULL) ++ goto out_nonewsk; ++ ++ /* Diff to tcp_v6_syn_recv_sock: Must do this prior to __ip6_dst_store, ++ * as it tries to access the pinet6-pointer. ++ */ ++ newtcp6sk = (struct tcp6_sock *)newsk; ++ inet_sk(newsk)->pinet6 = &newtcp6sk->inet6; ++ ++ /* ++ * No need to charge this sock to the relevant IPv6 refcnt debug socks ++ * count here, tcp_create_openreq_child now does this for us, see the ++ * comment in that function for the gory details. -acme ++ */ ++ ++ newsk->sk_gso_type = SKB_GSO_TCPV6; ++ __ip6_dst_store(newsk, dst, NULL, NULL); ++ inet6_sk_rx_dst_set(newsk, skb); ++ ++ newtp = tcp_sk(newsk); ++ newinet = inet_sk(newsk); ++ newnp = inet6_sk(newsk); ++ ++ newsk->sk_v6_daddr = treq->ir_v6_rmt_addr; ++ newnp->saddr = treq->ir_v6_loc_addr; ++ newsk->sk_v6_rcv_saddr = treq->ir_v6_loc_addr; ++ newsk->sk_bound_dev_if = treq->ir_iif; ++ ++ /* Now IPv6 options... ++ ++ First: no IPv4 options. ++ */ ++ newinet->inet_opt = NULL; ++ newnp->ipv6_ac_list = NULL; ++ newnp->ipv6_fl_list = NULL; ++ newnp->rxopt.all = 0; ++ ++ /* Clone pktoptions received with SYN */ ++ newnp->pktoptions = NULL; ++ if (treq->pktopts != NULL) { ++ newnp->pktoptions = skb_clone(treq->pktopts, ++ sk_gfp_atomic(meta_sk, GFP_ATOMIC)); ++ consume_skb(treq->pktopts); ++ treq->pktopts = NULL; ++ if (newnp->pktoptions) ++ skb_set_owner_r(newnp->pktoptions, newsk); ++ } ++ newnp->opt = NULL; ++ newnp->mcast_oif = inet6_iif(skb); ++ newnp->mcast_hops = ipv6_hdr(skb)->hop_limit; ++ newnp->rcv_flowinfo = ip6_flowinfo(ipv6_hdr(skb)); ++ ++ /* Initialization copied from inet6_create - normally this should have ++ * been handled by the memcpy as in tcp_v6_syn_recv_sock ++ */ ++ newnp->hop_limit = -1; ++ newnp->mc_loop = 1; ++ newnp->pmtudisc = IPV6_PMTUDISC_WANT; ++ (void)xchg(&newnp->rxpmtu, NULL); ++ ++ inet_csk(newsk)->icsk_ext_hdr_len = 0; ++ ++ tcp_mtup_init(newsk); ++ tcp_sync_mss(newsk, dst_mtu(dst)); ++ newtp->advmss = dst_metric_advmss(dst); ++ if (tcp_sk(meta_sk)->rx_opt.user_mss && ++ tcp_sk(meta_sk)->rx_opt.user_mss < newtp->advmss) ++ newtp->advmss = tcp_sk(meta_sk)->rx_opt.user_mss; ++ ++ tcp_initialize_rcv_mss(newsk); ++ ++ newinet->inet_daddr = LOOPBACK4_IPV6; ++ newinet->inet_saddr = LOOPBACK4_IPV6; ++ newinet->inet_rcv_saddr = LOOPBACK4_IPV6; ++ ++ if (__inet_inherit_port(meta_sk, newsk) < 0) { ++ inet_csk_prepare_forced_close(newsk); ++ tcp_done(newsk); ++ goto out; ++ } ++ __inet6_hash(newsk, NULL); ++ ++ return newsk; ++ ++out_overflow: ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_LISTENOVERFLOWS); ++out_nonewsk: ++ dst_release(dst); ++out: ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_LISTENDROPS); ++ return NULL; ++} ++ ++/* Similar to tcp_v6_conn_request */ ++static void mptcp_v6_join_request(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct tcp_options_received tmp_opt; ++ struct mptcp_options_received mopt; ++ struct ipv6_pinfo *np = inet6_sk(meta_sk); ++ struct request_sock *req; ++ struct inet_request_sock *treq; ++ struct mptcp_request_sock *mtreq; ++ u8 mptcp_hash_mac[20]; ++ __u32 isn = TCP_SKB_CB(skb)->when; ++ struct dst_entry *dst = NULL; ++ struct flowi6 fl6; ++ int want_cookie = 0; ++ union inet_addr addr; ++ ++ tcp_clear_options(&tmp_opt); ++ mptcp_init_mp_opt(&mopt); ++ tmp_opt.mss_clamp = TCP_MSS_DEFAULT; ++ tmp_opt.user_mss = tcp_sk(meta_sk)->rx_opt.user_mss; ++ tcp_parse_options(skb, &tmp_opt, &mopt, 0, NULL); ++ ++ req = inet6_reqsk_alloc(&mptcp6_request_sock_ops); ++ if (!req) ++ return; ++ ++#ifdef CONFIG_TCP_MD5SIG ++ tcp_rsk(req)->af_specific = &tcp_request_sock_ipv6_ops; ++#endif ++ ++ tmp_opt.tstamp_ok = tmp_opt.saw_tstamp; ++ tcp_openreq_init(req, &tmp_opt, skb); ++ ++ treq = inet_rsk(req); ++ treq->ir_v6_rmt_addr = ipv6_hdr(skb)->saddr; ++ treq->ir_v6_loc_addr = ipv6_hdr(skb)->daddr; ++ ++ if (!want_cookie || tmp_opt.tstamp_ok) ++ TCP_ECN_create_request(req, skb, sock_net(meta_sk)); ++ ++ treq->ir_iif = meta_sk->sk_bound_dev_if; ++ ++ /* So that link locals have meaning */ ++ if (!meta_sk->sk_bound_dev_if && ++ ipv6_addr_type(&treq->ir_v6_rmt_addr) & IPV6_ADDR_LINKLOCAL) ++ treq->ir_iif = inet6_iif(skb); ++ ++ if (!isn) { ++ if (meta_sk->sk_family == AF_INET6 && ++ (ipv6_opt_accepted(meta_sk, skb) || ++ np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo || ++ np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim)) { ++ atomic_inc(&skb->users); ++ treq->pktopts = skb; ++ } ++ ++ /* VJ's idea. We save last timestamp seen ++ * from the destination in peer table, when entering ++ * state TIME-WAIT, and check against it before ++ * accepting new connection request. ++ * ++ * If "isn" is not zero, this request hit alive ++ * timewait bucket, so that all the necessary checks ++ * are made in the function processing timewait state. ++ */ ++ if (tmp_opt.saw_tstamp && ++ tcp_death_row.sysctl_tw_recycle && ++ (dst = inet6_csk_route_req(meta_sk, &fl6, req)) != NULL) { ++ if (!tcp_peer_is_proven(req, dst, true)) { ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_PAWSPASSIVEREJECTED); ++ goto drop_and_release; ++ } ++ } ++ /* Kill the following clause, if you dislike this way. */ ++ else if (!sysctl_tcp_syncookies && ++ (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(meta_sk) < ++ (sysctl_max_syn_backlog >> 2)) && ++ !tcp_peer_is_proven(req, dst, false)) { ++ /* Without syncookies last quarter of ++ * backlog is filled with destinations, ++ * proven to be alive. ++ * It means that we continue to communicate ++ * to destinations, already remembered ++ * to the moment of synflood. ++ */ ++ LIMIT_NETDEBUG(KERN_DEBUG "TCP: drop open request from %pI6/%u\n", ++ &treq->ir_v6_rmt_addr, ++ ntohs(tcp_hdr(skb)->source)); ++ goto drop_and_release; ++ } ++ ++ isn = tcp_v6_init_sequence(skb); ++ } ++ ++ tcp_rsk(req)->snt_isn = isn; ++ tcp_rsk(req)->snt_synack = tcp_time_stamp; ++ tcp_rsk(req)->listener = NULL; ++ ++ mtreq = mptcp_rsk(req); ++ mtreq->mpcb = mpcb; ++ INIT_LIST_HEAD(&mtreq->collide_tuple); ++ mtreq->mptcp_rem_nonce = mopt.mptcp_recv_nonce; ++ mtreq->mptcp_rem_key = mpcb->mptcp_rem_key; ++ mtreq->mptcp_loc_key = mpcb->mptcp_loc_key; ++ mtreq->mptcp_loc_nonce = mptcp_v6_get_nonce(ipv6_hdr(skb)->daddr.s6_addr32, ++ ipv6_hdr(skb)->saddr.s6_addr32, ++ tcp_hdr(skb)->dest, ++ tcp_hdr(skb)->source, isn); ++ mptcp_hmac_sha1((u8 *)&mtreq->mptcp_loc_key, ++ (u8 *)&mtreq->mptcp_rem_key, ++ (u8 *)&mtreq->mptcp_loc_nonce, ++ (u8 *)&mtreq->mptcp_rem_nonce, (u32 *)mptcp_hash_mac); ++ mtreq->mptcp_hash_tmac = *(u64 *)mptcp_hash_mac; ++ ++ addr.in6 = treq->ir_v6_loc_addr; ++ mtreq->loc_id = mpcb->pm_ops->get_local_id(AF_INET6, &addr, sock_net(meta_sk)); ++ if (mtreq->loc_id == -1) /* Address not part of the allowed ones */ ++ goto drop_and_release; ++ mtreq->rem_id = mopt.rem_id; ++ mtreq->low_prio = mopt.low_prio; ++ tcp_rsk(req)->saw_mpc = 1; ++ ++ if (meta_sk->sk_family == AF_INET6) { ++ if (tcp_v6_send_synack(meta_sk, dst, &fl6, req, ++ skb_get_queue_mapping(skb))) ++ goto drop_and_free; ++ } else { ++ if (mptcp_v6v4_send_synack(meta_sk, req, skb_get_queue_mapping(skb))) ++ goto drop_and_free; ++ } ++ ++ /* Adding to request queue in metasocket */ ++ mptcp_v6_reqsk_queue_hash_add(meta_sk, req, TCP_TIMEOUT_INIT); ++ ++ return; ++ ++drop_and_release: ++ dst_release(dst); ++drop_and_free: ++ reqsk_free(req); ++ return; ++} ++ ++int mptcp_v6_rem_raddress(struct mptcp_cb *mpcb, u8 id) ++{ ++ int i; ++ ++ for (i = 0; i < MPTCP_MAX_ADDR; i++) { ++ if (!((1 << i) & mpcb->rem6_bits)) ++ continue; ++ ++ if (mpcb->remaddr6[i].rem6_id == id) { ++ /* remove address from bitfield */ ++ mpcb->rem6_bits &= ~(1 << i); ++ ++ return 0; ++ } ++ } ++ ++ return -1; ++} ++ ++/* Returns -1 if there is no space anymore to store an additional ++ * address ++ */ ++int mptcp_v6_add_raddress(struct mptcp_cb *mpcb, const struct in6_addr *addr, ++ __be16 port, u8 id) ++{ ++ int i; ++ struct mptcp_rem6 *rem6; ++ ++ mptcp_for_each_bit_set(mpcb->rem6_bits, i) { ++ rem6 = &mpcb->remaddr6[i]; ++ ++ /* Address is already in the list --- continue */ ++ if (rem6->rem6_id == id && ++ ipv6_addr_equal(&rem6->addr, addr) && rem6->port == port) ++ return 0; ++ ++ /* This may be the case, when the peer is behind a NAT. He is ++ * trying to JOIN, thus sending the JOIN with a certain ID. ++ * However the src_addr of the IP-packet has been changed. We ++ * update the addr in the list, because this is the address as ++ * OUR BOX sees it. ++ */ ++ if (rem6->rem6_id == id) { ++ /* update the address */ ++ mptcp_debug("%s: updating old addr: %pI6 to addr %pI6 with id:%d\n", ++ __func__, &rem6->addr, addr, id); ++ rem6->addr = *addr; ++ rem6->port = port; ++ mpcb->list_rcvd = 1; ++ return 0; ++ } ++ } ++ ++ i = mptcp_find_free_index(mpcb->rem6_bits); ++ /* Do we have already the maximum number of local/remote addresses? */ ++ if (i < 0) { ++ mptcp_debug("%s: At max num of remote addresses: %d --- not adding address: %pI6\n", ++ __func__, MPTCP_MAX_ADDR, addr); ++ return -1; ++ } ++ ++ rem6 = &mpcb->remaddr6[i]; ++ ++ /* Address is not known yet, store it */ ++ rem6->addr = *addr; ++ rem6->port = port; ++ rem6->bitfield = 0; ++ rem6->retry_bitfield = 0; ++ rem6->rem6_id = id; ++ mpcb->list_rcvd = 1; ++ mpcb->rem6_bits |= (1 << i); ++ ++ return 0; ++} ++ ++/* Sets the bitfield of the remote-address field ++ * local address is not set as it will disappear with the global address-list ++ */ ++void mptcp_v6_set_init_addr_bit(struct mptcp_cb *mpcb, ++ const struct in6_addr *daddr, int index) ++{ ++ int i; ++ mptcp_for_each_bit_set(mpcb->rem6_bits, i) { ++ if (ipv6_addr_equal(&mpcb->remaddr6[i].addr, daddr)) { ++ mpcb->remaddr6[i].bitfield |= (1 << index); ++ return; ++ } ++ } ++} ++ ++int mptcp_v6_do_rcv(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct sock *child, *rsk = NULL; ++ int ret; ++ ++ if (!(TCP_SKB_CB(skb)->mptcp_flags & MPTCPHDR_JOIN)) { ++ struct tcphdr *th = tcp_hdr(skb); ++ const struct ipv6hdr *ip6h = ipv6_hdr(skb); ++ struct sock *sk; ++ ++ sk = __inet6_lookup_established(sock_net(meta_sk), ++ &tcp_hashinfo, ++ &ip6h->saddr, th->source, ++ &ip6h->daddr, ntohs(th->dest), ++ inet6_iif(skb)); ++ ++ if (!sk) { ++ kfree_skb(skb); ++ return 0; ++ } ++ if (is_meta_sk(sk)) { ++ WARN("%s Did not find a sub-sk!\n", __func__); ++ kfree_skb(skb); ++ sock_put(sk); ++ return 0; ++ } ++ ++ if (sk->sk_state == TCP_TIME_WAIT) { ++ inet_twsk_put(inet_twsk(sk)); ++ kfree_skb(skb); ++ return 0; ++ } ++ ++ ret = tcp_v6_do_rcv(sk, skb); ++ sock_put(sk); ++ ++ return ret; ++ } ++ TCP_SKB_CB(skb)->mptcp_flags = 0; ++ ++ /* Has been removed from the tk-table. Thus, no new subflows. ++ * ++ * Check for close-state is necessary, because we may have been closed ++ * without passing by mptcp_close(). ++ * ++ * When falling back, no new subflows are allowed either. ++ */ ++ if (meta_sk->sk_state == TCP_CLOSE || !tcp_sk(meta_sk)->inside_tk_table || ++ mpcb->infinite_mapping_rcv || mpcb->send_infinite_mapping) ++ goto reset_and_discard; ++ ++ child = tcp_v6_hnd_req(meta_sk, skb); ++ ++ if (!child) ++ goto discard; ++ ++ if (child != meta_sk) { ++ sock_rps_save_rxhash(child, skb); ++ /* We don't call tcp_child_process here, because we hold ++ * already the meta-sk-lock and are sure that it is not owned ++ * by the user. ++ */ ++ ret = tcp_rcv_state_process(child, skb, tcp_hdr(skb), skb->len); ++ bh_unlock_sock(child); ++ sock_put(child); ++ if (ret) { ++ rsk = child; ++ goto reset_and_discard; ++ } ++ } else { ++ if (tcp_hdr(skb)->syn) { ++ struct mp_join *join_opt = mptcp_find_join(skb); ++ /* Currently we make two calls to mptcp_find_join(). This ++ * can probably be optimized. */ ++ if (mptcp_v6_add_raddress(mpcb, ++ (struct in6_addr *)&ipv6_hdr(skb)->saddr, ++ 0, ++ join_opt->addr_id) < 0) ++ goto reset_and_discard; ++ mpcb->list_rcvd = 0; ++ ++ mptcp_v6_join_request(meta_sk, skb); ++ goto discard; ++ } ++ goto reset_and_discard; ++ } ++ return 0; ++ ++reset_and_discard: ++ tcp_v6_send_reset(rsk, skb); ++discard: ++ kfree_skb(skb); ++ return 0; ++} ++ ++/* After this, the ref count of the meta_sk associated with the request_sock ++ * is incremented. Thus it is the responsibility of the caller ++ * to call sock_put() when the reference is not needed anymore. ++ */ ++struct sock *mptcp_v6_search_req(const __be16 rport, const struct in6_addr *raddr, ++ const struct in6_addr *laddr, const struct net *net) ++{ ++ struct mptcp_request_sock *mtreq; ++ struct sock *meta_sk = NULL; ++ ++ spin_lock(&mptcp_reqsk_hlock); ++ list_for_each_entry(mtreq, ++ &mptcp_reqsk_htb[inet6_synq_hash(raddr, rport, 0, ++ MPTCP_HASH_SIZE)], ++ collide_tuple) { ++ struct inet_request_sock *treq = inet_rsk(rev_mptcp_rsk(mtreq)); ++ meta_sk = mtreq->mpcb->meta_sk; ++ ++ if (inet_rsk(rev_mptcp_rsk(mtreq))->ir_rmt_port == rport && ++ rev_mptcp_rsk(mtreq)->rsk_ops->family == AF_INET6 && ++ ipv6_addr_equal(&treq->ir_v6_rmt_addr, raddr) && ++ ipv6_addr_equal(&treq->ir_v6_loc_addr, laddr) && ++ net_eq(net, sock_net(meta_sk))) ++ break; ++ meta_sk = NULL; ++ } ++ ++ if (meta_sk && unlikely(!atomic_inc_not_zero(&meta_sk->sk_refcnt))) ++ meta_sk = NULL; ++ spin_unlock(&mptcp_reqsk_hlock); ++ ++ return meta_sk; ++} ++ ++/* Create a new IPv6 subflow. ++ * ++ * We are in user-context and meta-sock-lock is hold. ++ */ ++int mptcp_init6_subsockets(struct sock *meta_sk, const struct mptcp_loc6 *loc, ++ struct mptcp_rem6 *rem) ++{ ++ struct tcp_sock *tp; ++ struct sock *sk; ++ struct sockaddr_in6 loc_in, rem_in; ++ struct socket sock; ++ int ulid_size = 0, ret; ++ ++ /** First, create and prepare the new socket */ ++ ++ sock.type = meta_sk->sk_socket->type; ++ sock.state = SS_UNCONNECTED; ++ sock.wq = meta_sk->sk_socket->wq; ++ sock.file = meta_sk->sk_socket->file; ++ sock.ops = NULL; ++ ++ ret = inet6_create(sock_net(meta_sk), &sock, IPPROTO_TCP, 1); ++ if (unlikely(ret < 0)) { ++ mptcp_debug("%s inet6_create failed ret: %d\n", __func__, ret); ++ return ret; ++ } ++ ++ sk = sock.sk; ++ tp = tcp_sk(sk); ++ ++ /* All subsockets need the MPTCP-lock-class */ ++ lockdep_set_class_and_name(&(sk)->sk_lock.slock, &meta_slock_key, "slock-AF_INET-MPTCP"); ++ lockdep_init_map(&(sk)->sk_lock.dep_map, "sk_lock-AF_INET-MPTCP", &meta_key, 0); ++ ++ if (mptcp_add_sock(meta_sk, sk, loc->loc6_id, rem->rem6_id, GFP_KERNEL)) ++ goto error; ++ ++ tp->mptcp->slave_sk = 1; ++ tp->mptcp->low_prio = loc->low_prio; ++ ++ /* Initializing the timer for an MPTCP subflow */ ++ setup_timer(&tp->mptcp->mptcp_ack_timer, mptcp_ack_handler, (unsigned long)sk); ++ ++ /** Then, connect the socket to the peer */ ++ ++ ulid_size = sizeof(struct sockaddr_in6); ++ loc_in.sin6_family = AF_INET6; ++ rem_in.sin6_family = AF_INET6; ++ loc_in.sin6_port = 0; ++ if (rem->port) ++ rem_in.sin6_port = rem->port; ++ else ++ rem_in.sin6_port = inet_sk(meta_sk)->inet_dport; ++ loc_in.sin6_addr = loc->addr; ++ rem_in.sin6_addr = rem->addr; ++ ++ ret = sock.ops->bind(&sock, (struct sockaddr *)&loc_in, ulid_size); ++ if (ret < 0) { ++ mptcp_debug("%s: MPTCP subsocket bind()failed, error %d\n", ++ __func__, ret); ++ goto error; ++ } ++ ++ mptcp_debug("%s: token %#x pi %d src_addr:%pI6:%d dst_addr:%pI6:%d\n", ++ __func__, tcp_sk(meta_sk)->mpcb->mptcp_loc_token, ++ tp->mptcp->path_index, &loc_in.sin6_addr, ++ ntohs(loc_in.sin6_port), &rem_in.sin6_addr, ++ ntohs(rem_in.sin6_port)); ++ ++ ret = sock.ops->connect(&sock, (struct sockaddr *)&rem_in, ++ ulid_size, O_NONBLOCK); ++ if (ret < 0 && ret != -EINPROGRESS) { ++ mptcp_debug("%s: MPTCP subsocket connect() failed, error %d\n", ++ __func__, ret); ++ goto error; ++ } ++ ++ sk_set_socket(sk, meta_sk->sk_socket); ++ sk->sk_wq = meta_sk->sk_wq; ++ ++ return 0; ++ ++error: ++ /* May happen if mptcp_add_sock fails first */ ++ if (!tp->mpc) { ++ tcp_close(sk, 0); ++ } else { ++ local_bh_disable(); ++ mptcp_sub_force_close(sk); ++ local_bh_enable(); ++ } ++ return ret; ++} ++EXPORT_SYMBOL(mptcp_init6_subsockets); ++ ++int mptcp_pm_v6_init(void) ++{ ++ int ret = 0; ++ struct request_sock_ops *ops = &mptcp6_request_sock_ops; ++ ++ ops->slab_name = kasprintf(GFP_KERNEL, "request_sock_%s", "MPTCP6"); ++ if (ops->slab_name == NULL) { ++ ret = -ENOMEM; ++ goto out; ++ } ++ ++ ops->slab = kmem_cache_create(ops->slab_name, ops->obj_size, 0, ++ SLAB_DESTROY_BY_RCU|SLAB_HWCACHE_ALIGN, ++ NULL); ++ ++ if (ops->slab == NULL) { ++ ret = -ENOMEM; ++ goto err_reqsk_create; ++ } ++ ++out: ++ return ret; ++ ++err_reqsk_create: ++ kfree(ops->slab_name); ++ ops->slab_name = NULL; ++ goto out; ++} ++ ++void mptcp_pm_v6_undo(void) ++{ ++ kmem_cache_destroy(mptcp6_request_sock_ops.slab); ++ kfree(mptcp6_request_sock_ops.slab_name); ++} +diff --git a/net/mptcp/mptcp_ndiffports.c b/net/mptcp/mptcp_ndiffports.c +new file mode 100644 +index 0000000..a126325 +--- /dev/null ++++ b/net/mptcp/mptcp_ndiffports.c +@@ -0,0 +1,171 @@ ++#include <linux/module.h> ++ ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++ ++#if IS_ENABLED(CONFIG_IPV6) ++#include <net/mptcp_v6.h> ++#endif ++ ++struct ndiffports_priv { ++ /* Worker struct for subflow establishment */ ++ struct work_struct subflow_work; ++ ++ struct mptcp_cb *mpcb; ++}; ++ ++static int sysctl_mptcp_ndiffports __read_mostly = 2; ++ ++/** ++ * Create all new subflows, by doing calls to mptcp_initX_subsockets ++ * ++ * This function uses a goto next_subflow, to allow releasing the lock between ++ * new subflows and giving other processes a chance to do some work on the ++ * socket and potentially finishing the communication. ++ **/ ++static void create_subflow_worker(struct work_struct *work) ++{ ++ struct ndiffports_priv *pm_priv = container_of(work, ++ struct ndiffports_priv, ++ subflow_work); ++ struct mptcp_cb *mpcb = pm_priv->mpcb; ++ struct sock *meta_sk = mpcb->meta_sk; ++ int iter = 0; ++ ++next_subflow: ++ if (iter) { ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ ++ yield(); ++ } ++ mutex_lock(&mpcb->mpcb_mutex); ++ lock_sock_nested(meta_sk, SINGLE_DEPTH_NESTING); ++ ++ iter++; ++ ++ if (sock_flag(meta_sk, SOCK_DEAD)) ++ goto exit; ++ ++ if (mpcb->master_sk && ++ !tcp_sk(mpcb->master_sk)->mptcp->fully_established) ++ goto exit; ++ ++ if (sysctl_mptcp_ndiffports > iter && ++ sysctl_mptcp_ndiffports > mpcb->cnt_subflows) { ++ if (meta_sk->sk_family == AF_INET || ++ mptcp_v6_is_v4_mapped(meta_sk)) { ++ struct mptcp_loc4 loc; ++ ++ loc.addr.s_addr = inet_sk(meta_sk)->inet_saddr; ++ loc.loc4_id = 0; ++ loc.low_prio = 0; ++ ++ mptcp_init4_subsockets(meta_sk, &loc, &mpcb->remaddr4[0]); ++ } else { ++#if IS_ENABLED(CONFIG_IPV6) ++ struct mptcp_loc6 loc; ++ ++ loc.addr = inet6_sk(meta_sk)->saddr; ++ loc.loc6_id = 0; ++ loc.low_prio = 0; ++ ++ mptcp_init6_subsockets(meta_sk, &loc, &mpcb->remaddr6[0]); ++#endif ++ } ++ goto next_subflow; ++ } ++ ++exit: ++ release_sock(meta_sk); ++ mutex_unlock(&mpcb->mpcb_mutex); ++ sock_put(meta_sk); ++} ++ ++static void ndiffports_new_session(struct sock *meta_sk, int index) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct ndiffports_priv *fmp = (struct ndiffports_priv *)&mpcb->mptcp_pm[0]; ++ ++ /* Initialize workqueue-struct */ ++ INIT_WORK(&fmp->subflow_work, create_subflow_worker); ++ fmp->mpcb = mpcb; ++} ++ ++static void ndiffports_create_subflows(struct sock *meta_sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct ndiffports_priv *pm_priv = (struct ndiffports_priv *)&mpcb->mptcp_pm[0]; ++ ++ if (mpcb->infinite_mapping_snd || mpcb->infinite_mapping_rcv || ++ mpcb->send_infinite_mapping || ++ mpcb->server_side || sock_flag(meta_sk, SOCK_DEAD)) ++ return; ++ ++ if (!work_pending(&pm_priv->subflow_work)) { ++ sock_hold(meta_sk); ++ queue_work(mptcp_wq, &pm_priv->subflow_work); ++ } ++} ++ ++static int ndiffports_get_local_index(sa_family_t family, union inet_addr *addr, ++ struct net *net) ++{ ++ return 0; ++} ++ ++static struct mptcp_pm_ops ndiffports __read_mostly = { ++ .new_session = ndiffports_new_session, ++ .fully_established = ndiffports_create_subflows, ++ .get_local_index = ndiffports_get_local_index, ++ .get_local_id = ndiffports_get_local_index, ++ .name = "ndiffports", ++ .owner = THIS_MODULE, ++}; ++ ++static struct ctl_table ndiff_table[] = { ++ { ++ .procname = "mptcp_ndiffports", ++ .data = &sysctl_mptcp_ndiffports, ++ .maxlen = sizeof(int), ++ .mode = 0644, ++ .proc_handler = &proc_dointvec ++ }, ++ { } ++}; ++ ++struct ctl_table_header *mptcp_sysctl; ++ ++/* General initialization of MPTCP_PM */ ++static int __init ndiffports_register(void) ++{ ++ BUILD_BUG_ON(sizeof(struct ndiffports_priv) > MPTCP_PM_SIZE); ++ ++ mptcp_sysctl = register_net_sysctl(&init_net, "net/mptcp", ndiff_table); ++ if (!mptcp_sysctl) ++ goto exit; ++ ++ if (mptcp_register_path_manager(&ndiffports)) ++ goto pm_failed; ++ ++ return 0; ++ ++pm_failed: ++ unregister_net_sysctl_table(mptcp_sysctl); ++exit: ++ return -1; ++} ++ ++static void ndiffports_unregister(void) ++{ ++ mptcp_unregister_path_manager(&ndiffports); ++ unregister_net_sysctl_table(mptcp_sysctl); ++} ++ ++module_init(ndiffports_register); ++module_exit(ndiffports_unregister); ++ ++MODULE_AUTHOR("Christoph Paasch"); ++MODULE_LICENSE("GPL"); ++MODULE_DESCRIPTION("NDIFF-PORTS MPTCP"); ++MODULE_VERSION("0.88"); +diff --git a/net/mptcp/mptcp_ofo_queue.c b/net/mptcp/mptcp_ofo_queue.c +new file mode 100644 +index 0000000..e182855 +--- /dev/null ++++ b/net/mptcp/mptcp_ofo_queue.c +@@ -0,0 +1,278 @@ ++/* ++ * MPTCP implementation - Fast algorithm for MPTCP meta-reordering ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * 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. ++ */ ++ ++#include <linux/skbuff.h> ++#include <linux/slab.h> ++#include <net/tcp.h> ++#include <net/mptcp.h> ++ ++void mptcp_remove_shortcuts(const struct mptcp_cb *mpcb, ++ const struct sk_buff *skb) ++{ ++ struct tcp_sock *tp; ++ ++ mptcp_for_each_tp(mpcb, tp) { ++ if (tp->mptcp->shortcut_ofoqueue == skb) { ++ tp->mptcp->shortcut_ofoqueue = NULL; ++ return; ++ } ++ } ++} ++ ++/* Does 'skb' fits after 'here' in the queue 'head' ? ++ * If yes, we queue it and return 1 ++ */ ++static int mptcp_ofo_queue_after(struct sk_buff_head *head, ++ struct sk_buff *skb, struct sk_buff *here, ++ struct tcp_sock *tp) ++{ ++ struct sock *meta_sk = tp->meta_sk; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ u32 seq = TCP_SKB_CB(skb)->seq; ++ u32 end_seq = TCP_SKB_CB(skb)->end_seq; ++ ++ /* We want to queue skb after here, thus seq >= end_seq */ ++ if (before(seq, TCP_SKB_CB(here)->end_seq)) ++ return 0; ++ ++ if (seq == TCP_SKB_CB(here)->end_seq) { ++ bool fragstolen = false; ++ ++ if (!tcp_try_coalesce(meta_sk, here, skb, &fragstolen)) { ++ __skb_queue_after(&meta_tp->out_of_order_queue, here, skb); ++ return 1; ++ } else { ++ kfree_skb_partial(skb, fragstolen); ++ return -1; ++ } ++ } ++ ++ /* If here is the last one, we can always queue it */ ++ if (skb_queue_is_last(head, here)) { ++ __skb_queue_after(head, here, skb); ++ return 1; ++ } else { ++ struct sk_buff *skb1 = skb_queue_next(head, here); ++ /* It's not the last one, but does it fits between 'here' and ++ * the one after 'here' ? Thus, does end_seq <= after_here->seq ++ */ ++ if (!after(end_seq, TCP_SKB_CB(skb1)->seq)) { ++ __skb_queue_after(head, here, skb); ++ return 1; ++ } ++ } ++ ++ return 0; ++} ++ ++static void try_shortcut(struct sk_buff *shortcut, struct sk_buff *skb, ++ struct sk_buff_head *head, struct tcp_sock *tp) ++{ ++ struct sock *meta_sk = tp->meta_sk; ++ struct tcp_sock *tp_it, *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sk_buff *skb1, *best_shortcut = NULL; ++ u32 seq = TCP_SKB_CB(skb)->seq; ++ u32 end_seq = TCP_SKB_CB(skb)->end_seq; ++ u32 distance = 0xffffffff; ++ ++ /* First, check the tp's shortcut */ ++ if (!shortcut) { ++ if (skb_queue_empty(head)) { ++ __skb_queue_head(head, skb); ++ goto end; ++ } ++ } else { ++ int ret = mptcp_ofo_queue_after(head, skb, shortcut, tp); ++ /* Does the tp's shortcut is a hit? If yes, we insert. */ ++ ++ if (ret) { ++ skb = (ret > 0) ? skb : NULL; ++ goto end; ++ } ++ } ++ ++ /* Check the shortcuts of the other subsockets. */ ++ mptcp_for_each_tp(mpcb, tp_it) { ++ shortcut = tp_it->mptcp->shortcut_ofoqueue; ++ /* Can we queue it here? If yes, do so! */ ++ if (shortcut) { ++ int ret = mptcp_ofo_queue_after(head, skb, shortcut, tp); ++ ++ if (ret) { ++ skb = (ret > 0) ? skb : NULL; ++ goto end; ++ } ++ } ++ ++ /* Could not queue it, check if we are close. ++ * We are looking for a shortcut, close enough to seq to ++ * set skb1 prematurely and thus improve the subsequent lookup, ++ * which tries to find a skb1 so that skb1->seq <= seq. ++ * ++ * So, here we only take shortcuts, whose shortcut->seq > seq, ++ * and minimize the distance between shortcut->seq and seq and ++ * set best_shortcut to this one with the minimal distance. ++ * ++ * That way, the subsequent while-loop is shortest. ++ */ ++ if (shortcut && after(TCP_SKB_CB(shortcut)->seq, seq)) { ++ /* Are we closer than the current best shortcut? */ ++ if ((u32)(TCP_SKB_CB(shortcut)->seq - seq) < distance) { ++ distance = (u32)(TCP_SKB_CB(shortcut)->seq - seq); ++ best_shortcut = shortcut; ++ } ++ } ++ } ++ ++ if (best_shortcut) ++ skb1 = best_shortcut; ++ else ++ skb1 = skb_peek_tail(head); ++ ++ if (seq == TCP_SKB_CB(skb1)->end_seq) { ++ bool fragstolen = false; ++ ++ if (!tcp_try_coalesce(meta_sk, skb1, skb, &fragstolen)) { ++ __skb_queue_after(&meta_tp->out_of_order_queue, skb1, skb); ++ } else { ++ kfree_skb_partial(skb, fragstolen); ++ skb = NULL; ++ } ++ ++ goto end; ++ } ++ ++ /* Find the insertion point, starting from best_shortcut if available. ++ * ++ * Inspired from tcp_data_queue_ofo. ++ */ ++ while (1) { ++ /* skb1->seq <= seq */ ++ if (!after(TCP_SKB_CB(skb1)->seq, seq)) ++ break; ++ if (skb_queue_is_first(head, skb1)) { ++ skb1 = NULL; ++ break; ++ } ++ skb1 = skb_queue_prev(head, skb1); ++ } ++ ++ /* Do skb overlap to previous one? */ ++ if (skb1 && before(seq, TCP_SKB_CB(skb1)->end_seq)) { ++ if (!after(end_seq, TCP_SKB_CB(skb1)->end_seq)) { ++ /* All the bits are present. */ ++ __kfree_skb(skb); ++ skb = NULL; ++ goto end; ++ } ++ if (seq == TCP_SKB_CB(skb1)->seq) { ++ if (skb_queue_is_first(head, skb1)) ++ skb1 = NULL; ++ else ++ skb1 = skb_queue_prev(head, skb1); ++ } ++ } ++ if (!skb1) ++ __skb_queue_head(head, skb); ++ else ++ __skb_queue_after(head, skb1, skb); ++ ++ /* And clean segments covered by new one as whole. */ ++ while (!skb_queue_is_last(head, skb)) { ++ skb1 = skb_queue_next(head, skb); ++ ++ if (!after(end_seq, TCP_SKB_CB(skb1)->seq)) ++ break; ++ ++ __skb_unlink(skb1, head); ++ mptcp_remove_shortcuts(mpcb, skb1); ++ __kfree_skb(skb1); ++ } ++ ++end: ++ if (skb) { ++ skb_set_owner_r(skb, meta_sk); ++ tp->mptcp->shortcut_ofoqueue = skb; ++ } ++ ++ return; ++} ++ ++/** ++ * @sk: the subflow that received this skb. ++ */ ++void mptcp_add_meta_ofo_queue(struct sock *meta_sk, struct sk_buff *skb, ++ struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ try_shortcut(tp->mptcp->shortcut_ofoqueue, skb, ++ &tcp_sk(meta_sk)->out_of_order_queue, tp); ++} ++ ++void mptcp_ofo_queue(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sk_buff *skb; ++ ++ while ((skb = skb_peek(&meta_tp->out_of_order_queue)) != NULL) { ++ u32 old_rcv_nxt = meta_tp->rcv_nxt; ++ if (after(TCP_SKB_CB(skb)->seq, meta_tp->rcv_nxt)) ++ break; ++ ++ if (!after(TCP_SKB_CB(skb)->end_seq, meta_tp->rcv_nxt)) { ++ __skb_unlink(skb, &meta_tp->out_of_order_queue); ++ mptcp_remove_shortcuts(meta_tp->mpcb, skb); ++ __kfree_skb(skb); ++ continue; ++ } ++ ++ __skb_unlink(skb, &meta_tp->out_of_order_queue); ++ mptcp_remove_shortcuts(meta_tp->mpcb, skb); ++ ++ __skb_queue_tail(&meta_sk->sk_receive_queue, skb); ++ meta_tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; ++ mptcp_check_rcvseq_wrap(meta_tp, old_rcv_nxt); ++ ++ if (tcp_hdr(skb)->fin) ++ mptcp_fin(meta_sk); ++ } ++} ++ ++void mptcp_purge_ofo_queue(struct tcp_sock *meta_tp) ++{ ++ struct sk_buff_head *head = &meta_tp->out_of_order_queue; ++ struct sk_buff *skb, *tmp; ++ ++ skb_queue_walk_safe(head, skb, tmp) { ++ __skb_unlink(skb, head); ++ mptcp_remove_shortcuts(meta_tp->mpcb, skb); ++ kfree_skb(skb); ++ } ++} +diff --git a/net/mptcp/mptcp_olia.c b/net/mptcp/mptcp_olia.c +new file mode 100644 +index 0000000..43d821e +--- /dev/null ++++ b/net/mptcp/mptcp_olia.c +@@ -0,0 +1,314 @@ ++/* ++ * MPTCP implementation - OPPORTUNISTIC LINKED INCREASES CONGESTION CONTROL: ++ * ++ * Algorithm design: ++ * Ramin Khalili <ramin.khalili@epfl.ch> ++ * Nicolas Gast <nicolas.gast@epfl.ch> ++ * Jean-Yves Le Boudec <jean-yves.leboudec@epfl.ch> ++ * ++ * Implementation: ++ * Ramin Khalili <ramin.khalili@epfl.ch> ++ * ++ * Ported to the official MPTCP-kernel: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * 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. ++ */ ++ ++ ++#include <net/tcp.h> ++#include <net/mptcp.h> ++ ++#include <linux/module.h> ++ ++static int scale = 10; ++ ++struct mptcp_olia { ++ u32 mptcp_loss1; ++ u32 mptcp_loss2; ++ u32 mptcp_loss3; ++ int epsilon_num; ++ u32 epsilon_den; ++ int mptcp_snd_cwnd_cnt; ++}; ++ ++static inline int mptcp_olia_sk_can_send(const struct sock *sk) ++{ ++ return mptcp_sk_can_send(sk) && tcp_sk(sk)->srtt; ++} ++ ++static inline u64 mptcp_olia_scale(u64 val, int scale) ++{ ++ return (u64) val << scale; ++} ++ ++/* take care of artificially inflate (see RFC5681) ++ * of cwnd during fast-retransmit phase ++ */ ++static u32 mptcp_get_crt_cwnd(struct sock *sk) ++{ ++ struct inet_connection_sock *icsk = inet_csk(sk); ++ ++ if (icsk->icsk_ca_state == TCP_CA_Recovery) ++ return tcp_sk(sk)->snd_ssthresh; ++ else ++ return tcp_sk(sk)->snd_cwnd; ++} ++ ++/* return the dominator of the first term of the increasing term */ ++static u64 mptcp_get_rate(struct mptcp_cb *mpcb , u32 path_rtt) ++{ ++ struct sock *sk; ++ u64 rate = 1; /* We have to avoid a zero-rate because it is used as a divisor */ ++ ++ mptcp_for_each_sk(mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ u64 scaled_num; ++ u32 tmp_cwnd; ++ ++ if (!mptcp_olia_sk_can_send(sk)) ++ continue; ++ ++ tmp_cwnd = mptcp_get_crt_cwnd(sk); ++ scaled_num = mptcp_olia_scale(tmp_cwnd, scale) * path_rtt; ++ rate += div_u64(scaled_num , tp->srtt); ++ } ++ rate *= rate; ++ return rate; ++} ++ ++/* find the maximum cwnd, used to find set M */ ++static u32 mptcp_get_max_cwnd(struct mptcp_cb *mpcb) ++{ ++ struct sock *sk; ++ u32 best_cwnd = 0; ++ ++ mptcp_for_each_sk(mpcb, sk) { ++ u32 tmp_cwnd; ++ ++ if (!mptcp_olia_sk_can_send(sk)) ++ continue; ++ ++ tmp_cwnd = mptcp_get_crt_cwnd(sk); ++ if (tmp_cwnd > best_cwnd) ++ best_cwnd = tmp_cwnd; ++ } ++ return best_cwnd; ++} ++ ++static void mptcp_get_epsilon(struct mptcp_cb *mpcb) ++{ ++ struct mptcp_olia *ca; ++ struct tcp_sock *tp; ++ struct sock *sk; ++ u64 tmp_int, tmp_rtt, best_int = 0, best_rtt = 1; ++ u32 max_cwnd = 1, best_cwnd = 1, tmp_cwnd; ++ u8 M = 0, B_not_M = 0; ++ ++ /* TODO - integrate this in the following loop - we just want to iterate once */ ++ ++ max_cwnd = mptcp_get_max_cwnd(mpcb); ++ ++ /* find the best path */ ++ mptcp_for_each_sk(mpcb, sk) { ++ tp = tcp_sk(sk); ++ ca = inet_csk_ca(sk); ++ ++ if (!mptcp_olia_sk_can_send(sk)) ++ continue; ++ ++ tmp_rtt = tp->srtt * tp->srtt; ++ /* TODO - check here and rename variables */ ++ tmp_int = max(ca->mptcp_loss3 - ca->mptcp_loss2, ++ ca->mptcp_loss2 - ca->mptcp_loss1); ++ ++ tmp_cwnd = mptcp_get_crt_cwnd(sk); ++ if (tmp_int * best_rtt >= best_int * tmp_rtt) { ++ best_rtt = tmp_rtt; ++ best_int = tmp_int; ++ best_cwnd = tmp_cwnd; ++ } ++ } ++ ++ /* TODO - integrate this here in mptcp_get_max_cwnd and in the previous loop */ ++ /* find the size of M and B_not_M */ ++ mptcp_for_each_sk(mpcb, sk) { ++ tp = tcp_sk(sk); ++ ca = inet_csk_ca(sk); ++ ++ if (!mptcp_olia_sk_can_send(sk)) ++ continue; ++ ++ tmp_cwnd = mptcp_get_crt_cwnd(sk); ++ if (tmp_cwnd == max_cwnd) { ++ M++; ++ } else { ++ tmp_rtt = tp->srtt * tp->srtt; ++ tmp_int = max(ca->mptcp_loss3 - ca->mptcp_loss2, ++ ca->mptcp_loss2 - ca->mptcp_loss1); ++ ++ if (tmp_int * best_rtt == best_int * tmp_rtt) ++ B_not_M++; ++ } ++ } ++ ++ /* check if the path is in M or B_not_M and set the value of epsilon accordingly */ ++ mptcp_for_each_sk(mpcb, sk) { ++ tp = tcp_sk(sk); ++ ca = inet_csk_ca(sk); ++ ++ if (!mptcp_olia_sk_can_send(sk)) ++ continue; ++ ++ if (B_not_M == 0) { ++ ca->epsilon_num = 0; ++ ca->epsilon_den = 1; ++ } else { ++ tmp_rtt = tp->srtt * tp->srtt; ++ tmp_int = max(ca->mptcp_loss3 - ca->mptcp_loss2, ++ ca->mptcp_loss2 - ca->mptcp_loss1); ++ tmp_cwnd = mptcp_get_crt_cwnd(sk); ++ ++ if (tmp_cwnd < max_cwnd && ++ tmp_int * best_rtt == best_int * tmp_rtt){ ++ ca->epsilon_num = 1; ++ ca->epsilon_den = mpcb->cnt_established * B_not_M; ++ } else if (tmp_cwnd == max_cwnd) { ++ ca->epsilon_num = -1; ++ ca->epsilon_den = mpcb->cnt_established * M; ++ } else { ++ ca->epsilon_num = 0; ++ ca->epsilon_den = 1; ++ } ++ } ++ } ++ ++} ++ ++/* setting the initial values */ ++static void mptcp_olia_init(struct sock *sk) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_olia *ca = inet_csk_ca(sk); ++ ++ if (tp->mpc) { ++ ca->mptcp_loss1 = tp->snd_una; ++ ca->mptcp_loss2 = tp->snd_una; ++ ca->mptcp_loss3 = tp->snd_una; ++ ca->mptcp_snd_cwnd_cnt = 0; ++ ca->epsilon_num = 0; ++ ca->epsilon_den = 1; ++ } ++} ++ ++/* updating inter-loss distance and ssthresh */ ++static void mptcp_olia_set_state(struct sock *sk, u8 new_state) ++{ ++ if (!tcp_sk(sk)->mpc) ++ return; ++ ++ if (new_state == TCP_CA_Loss || ++ new_state == TCP_CA_Recovery || new_state == TCP_CA_CWR) { ++ struct mptcp_olia *ca = inet_csk_ca(sk); ++ ++ if (ca->mptcp_loss3 != ca->mptcp_loss2 && ++ !inet_csk(sk)->icsk_retransmits) { ++ ca->mptcp_loss1 = ca->mptcp_loss2; ++ ca->mptcp_loss2 = ca->mptcp_loss3; ++ } ++ } ++ ++} ++ ++/* main algorithm */ ++static void mptcp_olia_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_olia *ca = inet_csk_ca(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ ++ u64 inc_num, inc_den, rate, cwnd_scaled; ++ ++ if (!tp->mpc) { ++ tcp_reno_cong_avoid(sk, ack, acked, in_flight); ++ return; ++ } ++ ++ ca->mptcp_loss3 = tp->snd_una; ++ ++ if (!tcp_is_cwnd_limited(sk, in_flight)) ++ return; ++ ++ /* slow start if it is in the safe area */ ++ if (tp->snd_cwnd <= tp->snd_ssthresh) { ++ tcp_slow_start(tp, acked); ++ return; ++ } ++ ++ mptcp_get_epsilon(mpcb); ++ rate = mptcp_get_rate(mpcb, tp->srtt); ++ cwnd_scaled = mptcp_olia_scale(tp->snd_cwnd, scale); ++ inc_den = ca->epsilon_den * tp->snd_cwnd * rate ? : 1; ++ ++ /* calculate the increasing term, scaling is used to reduce the rounding effect */ ++ if (ca->epsilon_num == -1) { ++ if (ca->epsilon_den * cwnd_scaled * cwnd_scaled < rate) { ++ inc_num = rate - ca->epsilon_den * ++ cwnd_scaled * cwnd_scaled; ++ ca->mptcp_snd_cwnd_cnt -= div64_u64( ++ mptcp_olia_scale(inc_num , scale) , inc_den); ++ } else { ++ inc_num = ca->epsilon_den * ++ cwnd_scaled * cwnd_scaled - rate; ++ ca->mptcp_snd_cwnd_cnt += div64_u64( ++ mptcp_olia_scale(inc_num , scale) , inc_den); ++ } ++ } else { ++ inc_num = ca->epsilon_num * rate + ++ ca->epsilon_den * cwnd_scaled * cwnd_scaled; ++ ca->mptcp_snd_cwnd_cnt += div64_u64( ++ mptcp_olia_scale(inc_num , scale) , inc_den); ++ } ++ ++ ++ if (ca->mptcp_snd_cwnd_cnt >= (1 << scale) - 1) { ++ if (tp->snd_cwnd < tp->snd_cwnd_clamp) ++ tp->snd_cwnd++; ++ ca->mptcp_snd_cwnd_cnt = 0; ++ } else if (ca->mptcp_snd_cwnd_cnt <= 0 - (1 << scale) + 1) { ++ tp->snd_cwnd = max((int) 1 , (int) tp->snd_cwnd - 1); ++ ca->mptcp_snd_cwnd_cnt = 0; ++ } ++} ++ ++static struct tcp_congestion_ops mptcp_olia = { ++ .init = mptcp_olia_init, ++ .ssthresh = tcp_reno_ssthresh, ++ .cong_avoid = mptcp_olia_cong_avoid, ++ .set_state = mptcp_olia_set_state, ++ .min_cwnd = tcp_reno_min_cwnd, ++ .owner = THIS_MODULE, ++ .name = "olia", ++}; ++ ++static int __init mptcp_olia_register(void) ++{ ++ BUILD_BUG_ON(sizeof(struct mptcp_olia) > ICSK_CA_PRIV_SIZE); ++ return tcp_register_congestion_control(&mptcp_olia); ++} ++ ++static void __exit mptcp_olia_unregister(void) ++{ ++ tcp_unregister_congestion_control(&mptcp_olia); ++} ++ ++module_init(mptcp_olia_register); ++module_exit(mptcp_olia_unregister); ++ ++MODULE_AUTHOR("Ramin Khalili, Nicolas Gast, Jean-Yves Le Boudec"); ++MODULE_LICENSE("GPL"); ++MODULE_DESCRIPTION("MPTCP COUPLED CONGESTION CONTROL"); ++MODULE_VERSION("0.1"); +diff --git a/net/mptcp/mptcp_output.c b/net/mptcp/mptcp_output.c +new file mode 100644 +index 0000000..807b79e +--- /dev/null ++++ b/net/mptcp/mptcp_output.c +@@ -0,0 +1,2255 @@ ++/* ++ * MPTCP implementation - Sending side ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++#include <linux/kconfig.h> ++#include <linux/skbuff.h> ++#include <linux/tcp.h> ++ ++#include <net/mptcp.h> ++#include <net/mptcp_v4.h> ++#include <net/mptcp_v6.h> ++#include <net/sock.h> ++ ++static inline int mptcp_pi_to_flag(int pi) ++{ ++ return 1 << (pi - 1); ++} ++ ++static inline int mptcp_sub_len_remove_addr(u16 bitfield) ++{ ++ unsigned int c; ++ for (c = 0; bitfield; c++) ++ bitfield &= bitfield - 1; ++ return MPTCP_SUB_LEN_REMOVE_ADDR + c - 1; ++} ++ ++int mptcp_sub_len_remove_addr_align(u16 bitfield) ++{ ++ return ALIGN(mptcp_sub_len_remove_addr(bitfield), 4); ++} ++EXPORT_SYMBOL(mptcp_sub_len_remove_addr_align); ++ ++/* If the sub-socket sk available to send the skb? */ ++static int mptcp_is_available(struct sock *sk, struct sk_buff *skb, ++ unsigned int *mss) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ unsigned int mss_now; ++ ++ /* Set of states for which we are allowed to send data */ ++ if (!mptcp_sk_can_send(sk)) ++ return 0; ++ ++ /* We do not send data on this subflow unless it is ++ * fully established, i.e. the 4th ack has been received. ++ */ ++ if (tp->mptcp->pre_established) ++ return 0; ++ ++ if (tp->pf || ++ (tp->mpcb->noneligible & mptcp_pi_to_flag(tp->mptcp->path_index))) ++ return 0; ++ ++ if (inet_csk(sk)->icsk_ca_state == TCP_CA_Loss) { ++ /* If SACK is disabled, and we got a loss, TCP does not exit ++ * the loss-state until something above high_seq has been acked. ++ * (see tcp_try_undo_recovery) ++ * ++ * high_seq is the snd_nxt at the moment of the RTO. As soon ++ * as we have an RTO, we won't push data on the subflow. ++ * Thus, snd_una can never go beyond high_seq. ++ */ ++ if (!tcp_is_reno(tp)) ++ return 0; ++ else if (tp->snd_una != tp->high_seq) ++ return 0; ++ } ++ ++ if (!tp->mptcp->fully_established) { ++ /* Make sure that we send in-order data */ ++ if (skb && tp->mptcp->second_packet && ++ tp->mptcp->last_end_data_seq != TCP_SKB_CB(skb)->seq) ++ return 0; ++ } ++ ++ if (!tcp_cwnd_test(tp, skb)) ++ return 0; ++ ++ mss_now = tcp_current_mss(sk); ++ /* Don't send on this subflow if we bypass the allowed send-window at ++ * the per-subflow level. Similar to tcp_snd_wnd_test, but manually ++ * calculated end_seq (because here at this point end_seq is still at ++ * the meta-level). ++ */ ++ if (skb && after(tp->write_seq + min(skb->len, mss_now), tcp_wnd_end(tp))) ++ return 0; ++ ++ if (mss) ++ *mss = mss_now; ++ ++ return 1; ++} ++ ++/* Are we not allowed to reinject this skb on tp? */ ++static int mptcp_dont_reinject_skb(struct tcp_sock *tp, struct sk_buff *skb) ++{ ++ /* If the skb has already been enqueued in this sk, try to find ++ * another one. ++ */ ++ return skb && ++ /* Has the skb already been enqueued into this subsocket? */ ++ mptcp_pi_to_flag(tp->mptcp->path_index) & TCP_SKB_CB(skb)->path_mask; ++} ++ ++/* This is the scheduler. This function decides on which flow to send ++ * a given MSS. If all subflows are found to be busy, NULL is returned ++ * The flow is selected based on the shortest RTT. ++ * If all paths have full cong windows, we simply return NULL. ++ * ++ * Additionally, this function is aware of the backup-subflows. ++ */ ++static struct sock *get_available_subflow(struct sock *meta_sk, ++ struct sk_buff *skb, ++ unsigned int *mss_now) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct sock *sk, *bestsk = NULL, *lowpriosk = NULL, *backupsk = NULL; ++ unsigned int mss = 0, mss_lowprio = 0, mss_backup = 0; ++ u32 min_time_to_peer = 0xffffffff, lowprio_min_time_to_peer = 0xffffffff; ++ int cnt_backups = 0; ++ ++ /* if there is only one subflow, bypass the scheduling function */ ++ if (mpcb->cnt_subflows == 1) { ++ bestsk = (struct sock *)mpcb->connection_list; ++ if (!mptcp_is_available(bestsk, skb, mss_now)) ++ bestsk = NULL; ++ return bestsk; ++ } ++ ++ /* Answer data_fin on same subflow!!! */ ++ if (meta_sk->sk_shutdown & RCV_SHUTDOWN && ++ skb && mptcp_is_data_fin(skb)) { ++ mptcp_for_each_sk(mpcb, sk) { ++ if (tcp_sk(sk)->mptcp->path_index == mpcb->dfin_path_index && ++ mptcp_is_available(sk, skb, mss_now)) ++ return sk; ++ } ++ } ++ ++ /* First, find the best subflow */ ++ mptcp_for_each_sk(mpcb, sk) { ++ struct tcp_sock *tp = tcp_sk(sk); ++ int this_mss; ++ ++ if (tp->mptcp->rcv_low_prio || tp->mptcp->low_prio) ++ cnt_backups++; ++ ++ if ((tp->mptcp->rcv_low_prio || tp->mptcp->low_prio) && ++ tp->srtt < lowprio_min_time_to_peer) { ++ ++ if (!mptcp_is_available(sk, skb, &this_mss)) ++ continue; ++ ++ if (mptcp_dont_reinject_skb(tp, skb)) { ++ mss_backup = this_mss; ++ backupsk = sk; ++ continue; ++ } ++ ++ lowprio_min_time_to_peer = tp->srtt; ++ lowpriosk = sk; ++ mss_lowprio = this_mss; ++ } else if (!(tp->mptcp->rcv_low_prio || tp->mptcp->low_prio) && ++ tp->srtt < min_time_to_peer) { ++ if (!mptcp_is_available(sk, skb, &this_mss)) ++ continue; ++ ++ if (mptcp_dont_reinject_skb(tp, skb)) { ++ mss_backup = this_mss; ++ backupsk = sk; ++ continue; ++ } ++ ++ min_time_to_peer = tp->srtt; ++ bestsk = sk; ++ mss = this_mss; ++ } ++ } ++ ++ if (mpcb->cnt_established == cnt_backups && lowpriosk) { ++ mss = mss_lowprio; ++ sk = lowpriosk; ++ } else if (bestsk) { ++ sk = bestsk; ++ } else if (backupsk){ ++ /* It has been sent on all subflows once - let's give it a ++ * chance again by restarting its pathmask. ++ */ ++ if (skb) ++ TCP_SKB_CB(skb)->path_mask = 0; ++ mss = mss_backup; ++ sk = backupsk; ++ } ++ ++ if (mss_now) ++ *mss_now = mss; ++ ++ return sk; ++} ++ ++static struct mp_dss *mptcp_skb_find_dss(const struct sk_buff *skb) ++{ ++ if (!mptcp_is_data_seq(skb)) ++ return NULL; ++ ++ return (struct mp_dss *)(skb->data - (MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN)); ++} ++ ++/* get the data-seq and end-data-seq and store them again in the ++ * tcp_skb_cb ++ */ ++static int mptcp_reconstruct_mapping(struct sk_buff *skb, struct sk_buff *orig_skb) ++{ ++ struct mp_dss *mpdss = mptcp_skb_find_dss(orig_skb); ++ u32 *p32; ++ u16 *p16; ++ ++ if (!mpdss || !mpdss->M) ++ return 1; ++ ++ /* Move the pointer to the data-seq */ ++ p32 = (u32 *)mpdss; ++ p32++; ++ if (mpdss->A) { ++ p32++; ++ if (mpdss->a) ++ p32++; ++ } ++ ++ TCP_SKB_CB(skb)->seq = ntohl(*p32); ++ ++ /* Get the data_len to calculate the end_data_seq */ ++ p32++; ++ p32++; ++ p16 = (u16 *)p32; ++ TCP_SKB_CB(skb)->end_seq = ntohs(*p16) + TCP_SKB_CB(skb)->seq; ++ ++ return 0; ++} ++ ++/* Similar to __pskb_copy and sk_stream_alloc_skb. */ ++static struct sk_buff *mptcp_pskb_copy(struct sk_buff *skb) ++{ ++ struct sk_buff *n; ++ /* The TCP header must be at least 32-bit aligned. */ ++ int size = ALIGN(skb_headlen(skb), 4); ++ ++ n = alloc_skb_fclone(size + MAX_TCP_HEADER, GFP_ATOMIC); ++ if (!n) ++ return NULL; ++ ++ /* Set the data pointer */ ++ skb_reserve(n, MAX_TCP_HEADER); ++ /* Set the tail pointer and length */ ++ skb_put(n, skb_headlen(skb)); ++ /* Copy the bytes */ ++ skb_copy_from_linear_data(skb, n->data, n->len); ++ ++ n->truesize += skb->data_len; ++ n->data_len = skb->data_len; ++ n->len = skb->len; ++ ++ if (skb_shinfo(skb)->nr_frags) { ++ int i; ++ ++ if (skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY) { ++ if (skb_copy_ubufs(skb, GFP_ATOMIC)) { ++ kfree_skb(n); ++ n = NULL; ++ goto out; ++ } ++ } ++ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { ++ skb_shinfo(n)->frags[i] = skb_shinfo(skb)->frags[i]; ++ skb_frag_ref(skb, i); ++ } ++ skb_shinfo(n)->nr_frags = i; ++ } ++ ++ if (skb_has_frag_list(skb)) { ++ skb_shinfo(n)->frag_list = skb_shinfo(skb)->frag_list; ++ skb_clone_fraglist(n); ++ } ++ ++ copy_skb_header(n, skb); ++out: ++ return n; ++} ++ ++/* Reinject data from one TCP subflow to the meta_sk. If sk == NULL, we are ++ * coming from the meta-retransmit-timer ++ */ ++static void __mptcp_reinject_data(struct sk_buff *orig_skb, struct sock *meta_sk, ++ struct sock *sk, int clone_it) ++{ ++ struct sk_buff *skb, *skb1; ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ u32 seq, end_seq; ++ ++ if (clone_it) { ++ /* pskb_copy is necessary here, because the TCP/IP-headers ++ * will be changed when it's going to be reinjected on another ++ * subflow. ++ */ ++ skb = mptcp_pskb_copy(orig_skb); ++ } else { ++ __skb_unlink(orig_skb, &sk->sk_write_queue); ++ sock_set_flag(sk, SOCK_QUEUE_SHRUNK); ++ sk->sk_wmem_queued -= orig_skb->truesize; ++ sk_mem_uncharge(sk, orig_skb->truesize); ++ skb = orig_skb; ++ } ++ if (unlikely(!skb)) ++ return; ++ ++ if (sk && mptcp_reconstruct_mapping(skb, orig_skb)) { ++ __kfree_skb(skb); ++ return; ++ } ++ ++ skb->sk = meta_sk; ++ ++ /* If it reached already the destination, we don't have to reinject it */ ++ if (!after(TCP_SKB_CB(skb)->end_seq, meta_tp->snd_una)) { ++ __kfree_skb(skb); ++ return; ++ } ++ ++ /* Only reinject segments that are fully covered by the mapping */ ++ if (skb->len + (mptcp_is_data_fin(skb) ? 1 : 0) != ++ TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq) { ++ u32 seq = TCP_SKB_CB(skb)->seq; ++ u32 end_seq = TCP_SKB_CB(skb)->end_seq; ++ ++ __kfree_skb(skb); ++ ++ /* Ok, now we have to look for the full mapping in the meta ++ * send-queue :S ++ */ ++ tcp_for_write_queue(skb, meta_sk) { ++ /* Not yet at the mapping? */ ++ if (before(TCP_SKB_CB(skb)->seq, seq)) ++ continue; ++ /* We have passed by the mapping */ ++ if (after(TCP_SKB_CB(skb)->end_seq, end_seq)) ++ return; ++ ++ __mptcp_reinject_data(skb, meta_sk, NULL, 1); ++ } ++ return; ++ } ++ ++ /* If it's empty, just add */ ++ if (skb_queue_empty(&mpcb->reinject_queue)) { ++ skb_queue_head(&mpcb->reinject_queue, skb); ++ return; ++ } ++ ++ /* Find place to insert skb - or even we can 'drop' it, as the ++ * data is already covered by other skb's in the reinject-queue. ++ * ++ * This is inspired by code from tcp_data_queue. ++ */ ++ ++ skb1 = skb_peek_tail(&mpcb->reinject_queue); ++ seq = TCP_SKB_CB(skb)->seq; ++ while (1) { ++ if (!after(TCP_SKB_CB(skb1)->seq, seq)) ++ break; ++ if (skb_queue_is_first(&mpcb->reinject_queue, skb1)) { ++ skb1 = NULL; ++ break; ++ } ++ skb1 = skb_queue_prev(&mpcb->reinject_queue, skb1); ++ } ++ ++ /* Do skb overlap to previous one? */ ++ end_seq = TCP_SKB_CB(skb)->end_seq; ++ if (skb1 && before(seq, TCP_SKB_CB(skb1)->end_seq)) { ++ if (!after(end_seq, TCP_SKB_CB(skb1)->end_seq)) { ++ /* All the bits are present. Don't reinject */ ++ __kfree_skb(skb); ++ return; ++ } ++ if (seq == TCP_SKB_CB(skb1)->seq) { ++ if (skb_queue_is_first(&mpcb->reinject_queue, skb1)) ++ skb1 = NULL; ++ else ++ skb1 = skb_queue_prev(&mpcb->reinject_queue, skb1); ++ } ++ } ++ if (!skb1) ++ __skb_queue_head(&mpcb->reinject_queue, skb); ++ else ++ __skb_queue_after(&mpcb->reinject_queue, skb1, skb); ++ ++ /* And clean segments covered by new one as whole. */ ++ while (!skb_queue_is_last(&mpcb->reinject_queue, skb)) { ++ skb1 = skb_queue_next(&mpcb->reinject_queue, skb); ++ ++ if (!after(end_seq, TCP_SKB_CB(skb1)->seq)) ++ break; ++ ++ __skb_unlink(skb1, &mpcb->reinject_queue); ++ __kfree_skb(skb1); ++ } ++ return; ++} ++ ++/* Inserts data into the reinject queue */ ++void mptcp_reinject_data(struct sock *sk, int clone_it) ++{ ++ struct sk_buff *skb_it, *tmp; ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = tp->meta_sk; ++ ++ /* It has already been closed - there is really no point in reinjecting */ ++ if (meta_sk->sk_state == TCP_CLOSE) ++ return; ++ ++ skb_queue_walk_safe(&sk->sk_write_queue, skb_it, tmp) { ++ struct tcp_skb_cb *tcb = TCP_SKB_CB(skb_it); ++ /* Subflow syn's and fin's are not reinjected. ++ * ++ * As well as empty subflow-fins with a data-fin. ++ * They are reinjected below (without the subflow-fin-flag) ++ */ ++ if (tcb->tcp_flags & TCPHDR_SYN || ++ (tcb->tcp_flags & TCPHDR_FIN && !mptcp_is_data_fin(skb_it)) || ++ (tcb->tcp_flags & TCPHDR_FIN && mptcp_is_data_fin(skb_it) && !skb_it->len)) ++ continue; ++ ++ __mptcp_reinject_data(skb_it, meta_sk, sk, clone_it); ++ } ++ ++ skb_it = tcp_write_queue_tail(meta_sk); ++ /* If sk has sent the empty data-fin, we have to reinject it too. */ ++ if (skb_it && mptcp_is_data_fin(skb_it) && skb_it->len == 0 && ++ TCP_SKB_CB(skb_it)->path_mask & mptcp_pi_to_flag(tp->mptcp->path_index)) { ++ __mptcp_reinject_data(skb_it, meta_sk, NULL, 1); ++ } ++ ++ mptcp_push_pending_frames(meta_sk); ++ ++ tp->pf = 1; ++} ++EXPORT_SYMBOL(mptcp_reinject_data); ++ ++static void mptcp_combine_dfin(struct sk_buff *skb, struct sock *meta_sk, ++ struct sock *subsk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sock *sk_it; ++ int all_empty = 1, all_acked; ++ ++ /* In infinite mapping we always try to combine */ ++ if (mpcb->infinite_mapping_snd && tcp_close_state(subsk)) { ++ subsk->sk_shutdown |= SEND_SHUTDOWN; ++ TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_FIN; ++ return; ++ } ++ ++ /* Don't combine, if they didn't combine - otherwise we end up in ++ * TIME_WAIT, even if our app is smart enough to avoid it ++ */ ++ if (meta_sk->sk_shutdown & RCV_SHUTDOWN) { ++ if (!mpcb->dfin_combined) ++ return; ++ } ++ ++ /* If no other subflow has data to send, we can combine */ ++ mptcp_for_each_sk(mpcb, sk_it) { ++ if (!mptcp_sk_can_send(sk_it)) ++ continue; ++ ++ if (!tcp_write_queue_empty(sk_it)) ++ all_empty = 0; ++ } ++ ++ /* If all data has been DATA_ACKed, we can combine. ++ * -1, because the data_fin consumed one byte ++ */ ++ all_acked = (meta_tp->snd_una == (meta_tp->write_seq - 1)); ++ ++ if ((all_empty || all_acked) && tcp_close_state(subsk)) { ++ subsk->sk_shutdown |= SEND_SHUTDOWN; ++ TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_FIN; ++ } ++} ++ ++static struct sk_buff *mptcp_skb_entail(struct sock *sk, struct sk_buff *skb, ++ int reinject) ++{ ++ __be32 *ptr; ++ __u16 data_len; ++ struct mp_dss *mdss; ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct tcp_skb_cb *tcb; ++ struct sk_buff *subskb = NULL; ++ ++ if (!reinject) ++ TCP_SKB_CB(skb)->mptcp_flags |= (mpcb->snd_hiseq_index ? ++ MPTCPHDR_SEQ64_INDEX : 0); ++ ++ subskb = mptcp_pskb_copy(skb); ++ if (!subskb) ++ return NULL; ++ ++ TCP_SKB_CB(skb)->path_mask |= mptcp_pi_to_flag(tp->mptcp->path_index); ++ ++ if (!(sk->sk_route_caps & NETIF_F_ALL_CSUM) && ++ skb->ip_summed == CHECKSUM_PARTIAL) { ++ subskb->csum = skb->csum = skb_checksum(skb, 0, skb->len, 0); ++ subskb->ip_summed = skb->ip_summed = CHECKSUM_NONE; ++ } ++ ++ /* The subskb is going in the subflow send-queue. Its path-mask ++ * is not needed anymore and MUST be set to 0, as the path-mask ++ * is a union with inet_skb_param. ++ */ ++ tcb = TCP_SKB_CB(subskb); ++ tcb->path_mask = 0; ++ ++ if (mptcp_is_data_fin(subskb)) ++ mptcp_combine_dfin(subskb, meta_sk, sk); ++ ++ if (tp->mpcb->infinite_mapping_snd) ++ goto no_data_seq; ++ ++ if (tp->mpcb->send_infinite_mapping && ++ !before(tcb->seq, mptcp_meta_tp(tp)->snd_nxt)) { ++ tp->mptcp->fully_established = 1; ++ tp->mpcb->infinite_mapping_snd = 1; ++ tp->mptcp->infinite_cutoff_seq = tp->write_seq; ++ tcb->mptcp_flags |= MPTCPHDR_INF; ++ data_len = 0; ++ } else { ++ data_len = tcb->end_seq - tcb->seq; ++ } ++ ++ /**** Write MPTCP DSS-option to the packet. ****/ ++ ptr = (__be32 *)(subskb->data - (MPTCP_SUB_LEN_DSS_ALIGN + ++ MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN)); ++ ++ /* Then we start writing it from the start */ ++ mdss = (struct mp_dss *)ptr; ++ ++ mdss->kind = TCPOPT_MPTCP; ++ mdss->sub = MPTCP_SUB_DSS; ++ mdss->rsv1 = 0; ++ mdss->rsv2 = 0; ++ mdss->F = (mptcp_is_data_fin(subskb) ? 1 : 0); ++ mdss->m = 0; ++ mdss->M = 1; ++ mdss->a = 0; ++ mdss->A = 1; ++ mdss->len = mptcp_sub_len_dss(mdss, tp->mpcb->dss_csum); ++ ++ ptr++; ++ ptr++; /* data_ack will be set in mptcp_options_write */ ++ *ptr++ = htonl(tcb->seq); /* data_seq */ ++ ++ /* If it's a non-data DATA_FIN, we set subseq to 0 (draft v7) */ ++ if (mptcp_is_data_fin(subskb) && subskb->len == 0) ++ *ptr++ = 0; /* subseq */ ++ else ++ *ptr++ = htonl(tp->write_seq - tp->mptcp->snt_isn); /* subseq */ ++ ++ if (tp->mpcb->dss_csum && data_len) { ++ __be16 *p16 = (__be16 *)ptr; ++ __be32 hdseq = mptcp_get_highorder_sndbits(subskb, tp->mpcb); ++ __wsum csum; ++ *ptr = htonl(((data_len) << 16) | ++ (TCPOPT_EOL << 8) | ++ (TCPOPT_EOL)); ++ ++ csum = csum_partial(ptr - 2, 12, subskb->csum); ++ p16++; ++ *p16++ = csum_fold(csum_partial(&hdseq, sizeof(hdseq), csum)); ++ } else { ++ *ptr++ = htonl(((data_len) << 16) | ++ (TCPOPT_NOP << 8) | ++ (TCPOPT_NOP)); ++ } ++ ++no_data_seq: ++ tcb->seq = tp->write_seq; ++ tcb->sacked = 0; /* reset the sacked field: from the point of view ++ * of this subflow, we are sending a brand new ++ * segment */ ++ /* Take into account seg len */ ++ tp->write_seq += subskb->len + ((tcb->tcp_flags & TCPHDR_FIN) ? 1 : 0); ++ tcb->end_seq = tp->write_seq; ++ ++ /* If it's a non-payload DATA_FIN (also no subflow-fin), the ++ * segment is not part of the subflow but on a meta-only-level ++ */ ++ if (!mptcp_is_data_fin(subskb) || tcb->end_seq != tcb->seq) { ++ tcp_add_write_queue_tail(sk, subskb); ++ sk->sk_wmem_queued += subskb->truesize; ++ sk_mem_charge(sk, subskb->truesize); ++ } ++ ++ return subskb; ++} ++ ++static void mptcp_sub_event_new_data_sent(struct sock *sk, ++ struct sk_buff *subskb, ++ struct sk_buff *skb) ++{ ++ /* If it's a non-payload DATA_FIN (also no subflow-fin), the ++ * segment is not part of the subflow but on a meta-only-level ++ * ++ * We free it, because it has been queued nowhere. ++ */ ++ if (!mptcp_is_data_fin(subskb) || ++ (TCP_SKB_CB(subskb)->end_seq != TCP_SKB_CB(subskb)->seq)) { ++ tcp_event_new_data_sent(sk, subskb); ++ tcp_sk(sk)->mptcp->second_packet = 1; ++ tcp_sk(sk)->mptcp->last_end_data_seq = TCP_SKB_CB(skb)->end_seq; ++ } else { ++ kfree_skb(subskb); ++ } ++} ++ ++/* Handle the packets and sockets after a tcp_transmit_skb failed */ ++static void mptcp_transmit_skb_failed(struct sock *sk, struct sk_buff *skb, ++ struct sk_buff *subskb) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ ++ /* No work to do if we are in infinite mapping mode ++ * There is only one subflow left and we cannot send this segment on ++ * another subflow. ++ */ ++ if (mpcb->infinite_mapping_snd) ++ return; ++ ++ TCP_SKB_CB(skb)->path_mask &= ~mptcp_pi_to_flag(tp->mptcp->path_index); ++ ++ if (TCP_SKB_CB(subskb)->tcp_flags & TCPHDR_FIN) { ++ /* If it is a subflow-fin we must leave it on the ++ * subflow-send-queue, so that the probe-timer ++ * can retransmit it. ++ */ ++ if (!tp->packets_out && !inet_csk(sk)->icsk_pending) ++ inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, ++ inet_csk(sk)->icsk_rto, TCP_RTO_MAX); ++ } else if (mptcp_is_data_fin(subskb) && ++ TCP_SKB_CB(subskb)->end_seq == TCP_SKB_CB(subskb)->seq) { ++ /* An empty data-fin has not been enqueued on the subflow ++ * and thus we free it. ++ */ ++ ++ kfree_skb(subskb); ++ } else { ++ /* In all other cases we remove it from the sub-queue. ++ * Other subflows may send it, or the probe-timer will ++ * handle it. ++ */ ++ tcp_advance_send_head(sk, subskb); ++ ++ /* tcp_add_write_queue_tail initialized highest_sack. We have ++ * to reset it, if necessary. ++ */ ++ if (tp->highest_sack == subskb) ++ tp->highest_sack = NULL; ++ ++ tcp_unlink_write_queue(subskb, sk); ++ tp->write_seq -= subskb->len; ++ sk_wmem_free_skb(sk, subskb); ++ } ++} ++ ++/* Function to create two new TCP segments. Shrinks the given segment ++ * to the specified size and appends a new segment with the rest of the ++ * packet to the list. This won't be called frequently, I hope. ++ * Remember, these are still headerless SKBs at this point. ++ */ ++int mptcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, ++ unsigned int mss_now, int reinject) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct sk_buff *buff; ++ int nsize, old_factor; ++ int nlen; ++ u8 flags; ++ int dsslen = MPTCP_SUB_LEN_DSS_ALIGN + MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN; ++ char dss[MPTCP_SUB_LEN_DSS_ALIGN + MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN]; ++ ++ if (WARN_ON(len > skb->len)) ++ return -EINVAL; ++ ++ /* DSS-option must be recovered afterwards. */ ++ if (!is_meta_sk(sk)) ++ memcpy(dss, skb->data - dsslen, dsslen); ++ ++ nsize = skb_headlen(skb) - len; ++ if (nsize < 0) ++ nsize = 0; ++ ++ if (skb_cloned(skb)) { ++ if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) ++ return -ENOMEM; ++ /* Recover dss-option */ ++ if (!is_meta_sk(sk)) ++ memcpy(skb->data - dsslen, dss, dsslen); ++ } ++ ++ /* Get a new skb... force flag on. */ ++ buff = sk_stream_alloc_skb(sk, nsize, GFP_ATOMIC); ++ if (buff == NULL) ++ return -ENOMEM; /* We'll just try again later. */ ++ ++ /* See below - if reinject == 1, the buff will be added to the reinject- ++ * queue, which is currently not part of the memory-accounting. ++ */ ++ if (reinject != 1) { ++ sk->sk_wmem_queued += buff->truesize; ++ sk_mem_charge(sk, buff->truesize); ++ } ++ nlen = skb->len - len - nsize; ++ buff->truesize += nlen; ++ skb->truesize -= nlen; ++ ++ /* Correct the sequence numbers. */ ++ TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len; ++ TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq; ++ TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq; ++ ++ /* PSH and FIN should only be set in the second packet. */ ++ flags = TCP_SKB_CB(skb)->tcp_flags; ++ TCP_SKB_CB(skb)->tcp_flags = flags & ~(TCPHDR_FIN | TCPHDR_PSH); ++ TCP_SKB_CB(buff)->tcp_flags = flags; ++ TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked; ++ ++ flags = TCP_SKB_CB(skb)->mptcp_flags; ++ TCP_SKB_CB(skb)->mptcp_flags = flags & ~(MPTCPHDR_FIN); ++ TCP_SKB_CB(buff)->mptcp_flags = flags; ++ ++ if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_PARTIAL) { ++ /* Copy and checksum data tail into the new buffer. */ ++ buff->csum = csum_partial_copy_nocheck(skb->data + len, ++ skb_put(buff, nsize), ++ nsize, 0); ++ ++ skb_trim(skb, len); ++ ++ skb->csum = csum_block_sub(skb->csum, buff->csum, len); ++ } else { ++ skb->ip_summed = CHECKSUM_PARTIAL; ++ skb_split(skb, buff, len); ++ } ++ ++ /* We lost the dss-option when creating buff - put it back! */ ++ if (!is_meta_sk(sk)) ++ memcpy(buff->data - dsslen, dss, dsslen); ++ ++ buff->ip_summed = skb->ip_summed; ++ ++ /* Looks stupid, but our code really uses when of ++ * skbs, which it never sent before. --ANK ++ */ ++ TCP_SKB_CB(buff)->when = TCP_SKB_CB(skb)->when; ++ buff->tstamp = skb->tstamp; ++ ++ old_factor = tcp_skb_pcount(skb); ++ ++ /* Fix up tso_factor for both original and new SKB. */ ++ tcp_set_skb_tso_segs(sk, skb, mss_now); ++ tcp_set_skb_tso_segs(sk, buff, mss_now); ++ ++ /* If this packet has been sent out already, we must ++ * adjust the various packet counters. ++ */ ++ if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq) && reinject != 1) { ++ int diff = old_factor - tcp_skb_pcount(skb) - ++ tcp_skb_pcount(buff); ++ ++ if (diff) ++ tcp_adjust_pcount(sk, skb, diff); ++ } ++ ++ /* Link BUFF into the send queue. */ ++ skb_header_release(buff); ++ if (reinject == 1) ++ __skb_queue_after(&tcp_sk(sk)->mpcb->reinject_queue, skb, buff); ++ else ++ tcp_insert_write_queue_after(skb, buff, sk); ++ ++ return 0; ++} ++ ++int mptso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, ++ unsigned int mss_now, gfp_t gfp, int reinject) ++{ ++ struct sk_buff *buff; ++ int nlen = skb->len - len, old_factor; ++ u8 flags; ++ int dsslen = MPTCP_SUB_LEN_DSS_ALIGN + MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN; ++ ++ /* All of a TSO frame must be composed of paged data. */ ++ if (skb->len != skb->data_len) ++ return mptcp_fragment(sk, skb, len, mss_now, reinject); ++ ++ buff = sk_stream_alloc_skb(sk, 0, gfp); ++ if (unlikely(buff == NULL)) ++ return -ENOMEM; ++ ++ /* See below - if reinject == 1, the buff will be added to the reinject- ++ * queue, which is currently not part of the memory-accounting. ++ */ ++ if (reinject != 1) { ++ sk->sk_wmem_queued += buff->truesize; ++ sk_mem_charge(sk, buff->truesize); ++ } ++ buff->truesize += nlen; ++ skb->truesize -= nlen; ++ ++ /* Correct the sequence numbers. */ ++ TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len; ++ TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq; ++ TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq; ++ ++ /* PSH and FIN should only be set in the second packet. */ ++ flags = TCP_SKB_CB(skb)->tcp_flags; ++ TCP_SKB_CB(skb)->tcp_flags = flags & ~(TCPHDR_FIN | TCPHDR_PSH); ++ TCP_SKB_CB(buff)->tcp_flags = flags; ++ ++ flags = TCP_SKB_CB(skb)->mptcp_flags; ++ TCP_SKB_CB(skb)->mptcp_flags = flags & ~(MPTCPHDR_FIN); ++ TCP_SKB_CB(buff)->mptcp_flags = flags; ++ ++ /* This packet was never sent out yet, so no SACK bits. */ ++ TCP_SKB_CB(buff)->sacked = 0; ++ ++ buff->ip_summed = CHECKSUM_PARTIAL; ++ skb->ip_summed = CHECKSUM_PARTIAL; ++ skb_split(skb, buff, len); ++ ++ /* We lost the dss-option when creating buff - put it back! */ ++ if (!is_meta_sk(sk)) ++ memcpy(buff->data - dsslen, skb->data - dsslen, dsslen); ++ ++ old_factor = tcp_skb_pcount(skb); ++ ++ /* Fix up tso_factor for both original and new SKB. */ ++ tcp_set_skb_tso_segs(sk, skb, mss_now); ++ tcp_set_skb_tso_segs(sk, buff, mss_now); ++ ++ /* If this packet has been sent out already, we must ++ * adjust the various packet counters. ++ */ ++ if (!before(tcp_sk(sk)->snd_nxt, TCP_SKB_CB(buff)->end_seq) && reinject != 1) { ++ int diff = old_factor - tcp_skb_pcount(skb) - ++ tcp_skb_pcount(buff); ++ ++ if (diff) ++ tcp_adjust_pcount(sk, skb, diff); ++ } ++ ++ /* Link BUFF into the send queue. */ ++ skb_header_release(buff); ++ if (reinject == 1) ++ __skb_queue_after(&tcp_sk(sk)->mpcb->reinject_queue, skb, buff); ++ else ++ tcp_insert_write_queue_after(skb, buff, sk); ++ ++ return 0; ++} ++ ++/* Inspired by tcp_write_wakeup */ ++int mptcp_write_wakeup(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sk_buff *skb, *subskb; ++ ++ skb = tcp_send_head(meta_sk); ++ if (skb && ++ before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(meta_tp))) { ++ int err; ++ unsigned int mss; ++ unsigned int seg_size = tcp_wnd_end(meta_tp) - TCP_SKB_CB(skb)->seq; ++ struct sock *subsk = get_available_subflow(meta_sk, skb, &mss); ++ if (!subsk) ++ return -1; ++ ++ if (before(meta_tp->pushed_seq, TCP_SKB_CB(skb)->end_seq)) ++ meta_tp->pushed_seq = TCP_SKB_CB(skb)->end_seq; ++ ++ /* We are probing the opening of a window ++ * but the window size is != 0 ++ * must have been a result SWS avoidance ( sender ) ++ */ ++ if (seg_size < TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq || ++ skb->len > mss) { ++ seg_size = min(seg_size, mss); ++ TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_PSH; ++ if (mptcp_fragment(meta_sk, skb, seg_size, mss, 0)) ++ return -1; ++ } else if (!tcp_skb_pcount(skb)) { ++ tcp_set_skb_tso_segs(meta_sk, skb, mss); ++ } ++ ++ subskb = mptcp_skb_entail(subsk, skb, 0); ++ if (!subskb) ++ return -1; ++ ++ TCP_SKB_CB(subskb)->tcp_flags |= TCPHDR_PSH; ++ TCP_SKB_CB(skb)->when = tcp_time_stamp; ++ TCP_SKB_CB(subskb)->when = tcp_time_stamp; ++ err = tcp_transmit_skb(subsk, subskb, 1, GFP_ATOMIC); ++ if (unlikely(err)) { ++ mptcp_transmit_skb_failed(subsk, skb, subskb); ++ return err; ++ } ++ ++ mptcp_check_sndseq_wrap(meta_tp, TCP_SKB_CB(skb)->end_seq - ++ TCP_SKB_CB(skb)->seq); ++ tcp_event_new_data_sent(meta_sk, skb); ++ mptcp_sub_event_new_data_sent(subsk, subskb, skb); ++ ++ return 0; ++ } else { ++ struct sock *sk_it; ++ int ans = 0; ++ ++ if (between(meta_tp->snd_up, meta_tp->snd_una + 1, ++ meta_tp->snd_una + 0xFFFF)) { ++ mptcp_for_each_sk(meta_tp->mpcb, sk_it) { ++ if (mptcp_sk_can_send_ack(sk_it)) ++ tcp_xmit_probe_skb(sk_it, 1); ++ } ++ } ++ ++ /* At least one of the tcp_xmit_probe_skb's has to succeed */ ++ mptcp_for_each_sk(meta_tp->mpcb, sk_it) { ++ int ret; ++ ++ if (!mptcp_sk_can_send_ack(sk_it)) ++ continue; ++ ++ ret = tcp_xmit_probe_skb(sk_it, 0); ++ if (unlikely(ret > 0)) ++ ans = ret; ++ } ++ return ans; ++ } ++} ++ ++static void mptcp_find_and_set_pathmask(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct sk_buff *skb_it; ++ ++ skb_it = tcp_write_queue_head(meta_sk); ++ ++ tcp_for_write_queue_from(skb_it, meta_sk) { ++ if (skb_it == tcp_send_head(meta_sk)) ++ break; ++ ++ if (TCP_SKB_CB(skb_it)->seq == TCP_SKB_CB(skb)->seq) { ++ TCP_SKB_CB(skb)->path_mask = TCP_SKB_CB(skb_it)->path_mask; ++ break; ++ } ++ } ++} ++ ++static struct sk_buff *mptcp_rcv_buf_optimization(struct sock *sk, int penal) ++{ ++ struct sock *meta_sk; ++ struct tcp_sock *tp = tcp_sk(sk), *tp_it; ++ struct sk_buff *skb_head; ++ ++ if (tp->mpcb->cnt_subflows == 1) ++ return NULL; ++ ++ meta_sk = mptcp_meta_sk(sk); ++ skb_head = tcp_write_queue_head(meta_sk); ++ ++ if (!skb_head || skb_head == tcp_send_head(meta_sk)) ++ return NULL; ++ ++ /* If penalization is optional (coming from mptcp_next_segment() and ++ * We are not send-buffer-limited we do not penalize. The retransmission ++ * is just an optimization to fix the idle-time due to the delay before ++ * we wake up the application. ++ */ ++ if (!penal && sk_stream_memory_free(meta_sk)) ++ goto retrans; ++ ++ /* Only penalize again after an RTT has elapsed */ ++ if (tcp_time_stamp - tp->mptcp->last_rbuf_opti < tp->srtt >> 3) ++ goto retrans; ++ ++ /* Half the cwnd of the slow flow */ ++ mptcp_for_each_tp(tp->mpcb, tp_it) { ++ if (tp_it != tp && ++ TCP_SKB_CB(skb_head)->path_mask & mptcp_pi_to_flag(tp_it->mptcp->path_index)) { ++ if (tp->srtt < tp_it->srtt && inet_csk((struct sock *)tp_it)->icsk_ca_state == TCP_CA_Open) { ++ tp_it->snd_cwnd = max(tp_it->snd_cwnd >> 1U, 1U); ++ if (tp_it->snd_ssthresh != TCP_INFINITE_SSTHRESH) ++ tp_it->snd_ssthresh = max(tp_it->snd_ssthresh >> 1U, 2U); ++ ++ tp->mptcp->last_rbuf_opti = tcp_time_stamp; ++ } ++ break; ++ } ++ } ++ ++retrans: ++ ++ /* Segment not yet injected into this path? Take it!!! */ ++ if (!(TCP_SKB_CB(skb_head)->path_mask & mptcp_pi_to_flag(tp->mptcp->path_index))) { ++ bool do_retrans = false; ++ mptcp_for_each_tp(tp->mpcb, tp_it) { ++ if (tp_it != tp && ++ TCP_SKB_CB(skb_head)->path_mask & mptcp_pi_to_flag(tp_it->mptcp->path_index)) { ++ if (tp_it->snd_cwnd <= 4) { ++ do_retrans = true; ++ break; ++ } ++ ++ if (4 * tp->srtt >= tp_it->srtt) { ++ do_retrans = false; ++ break; ++ } else { ++ do_retrans = true; ++ } ++ } ++ } ++ ++ if (do_retrans) ++ return skb_head; ++ } ++ return NULL; ++} ++ ++int mptcp_write_xmit(struct sock *meta_sk, unsigned int mss_now, int nonagle, ++ int push_one, gfp_t gfp) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk), *subtp; ++ struct sock *subsk; ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sk_buff *skb; ++ unsigned int tso_segs, old_factor, sent_pkts; ++ int cwnd_quota; ++ int result; ++ int reinject = 0; ++ ++ sent_pkts = 0; ++ ++ /* Currently mtu-probing is not done in MPTCP */ ++ if (!push_one && 0) { ++ /* Do MTU probing. */ ++ result = tcp_mtu_probe(meta_sk); ++ if (!result) ++ return 0; ++ else if (result > 0) ++ sent_pkts = 1; ++ } ++ ++ while ((skb = mptcp_next_segment(meta_sk, &reinject))) { ++ unsigned int limit; ++ struct sk_buff *subskb = NULL; ++ u32 noneligible = mpcb->noneligible; ++ ++ if (reinject == 1) { ++ if (!after(TCP_SKB_CB(skb)->end_seq, meta_tp->snd_una)) { ++ /* Segment already reached the peer, take the next one */ ++ __skb_unlink(skb, &mpcb->reinject_queue); ++ __kfree_skb(skb); ++ continue; ++ } ++ ++ /* Reinjection and it is coming from a subflow? We need ++ * to find out the path-mask from the meta-write-queue ++ * to properly select a subflow. ++ */ ++ if (!TCP_SKB_CB(skb)->path_mask) ++ mptcp_find_and_set_pathmask(meta_sk, skb); ++ } ++ ++subflow: ++ subsk = get_available_subflow(meta_sk, skb, &mss_now); ++ if (!subsk) ++ break; ++ subtp = tcp_sk(subsk); ++ ++ /* Since all subsocks are locked before calling the scheduler, ++ * the tcp_send_head should not change. ++ */ ++ BUG_ON(!reinject && tcp_send_head(meta_sk) != skb); ++retry: ++ /* If the segment was cloned (e.g. a meta retransmission), ++ * the header must be expanded/copied so that there is no ++ * corruption of TSO information. ++ */ ++ if (skb_unclone(skb, GFP_ATOMIC)) ++ break; ++ ++ old_factor = tcp_skb_pcount(skb); ++ tcp_set_skb_tso_segs(meta_sk, skb, mss_now); ++ tso_segs = tcp_skb_pcount(skb); ++ ++ if (reinject == -1) { ++ /* The packet has already once been sent, so if we ++ * change the pcount here we have to adjust packets_out ++ * in the meta-sk ++ */ ++ int diff = old_factor - tso_segs; ++ ++ if (diff) ++ tcp_adjust_pcount(meta_sk, skb, diff); ++ } ++ ++ cwnd_quota = tcp_cwnd_test(subtp, skb); ++ if (!cwnd_quota) { ++ /* May happen due to two cases: ++ * ++ * - if at the first selection we circumvented ++ * the test due to a DATA_FIN (and got rejected at ++ * tcp_snd_wnd_test), but the reinjected segment is not ++ * a DATA_FIN. ++ * - if we take a DATA_FIN with data, but ++ * tcp_set_skb_tso_segs() increases the number of ++ * tso_segs to something > 1. Then, cwnd_test might ++ * reject it. ++ */ ++ mpcb->noneligible |= mptcp_pi_to_flag(subtp->mptcp->path_index); ++ continue; ++ } ++ ++ if (!reinject && unlikely(!tcp_snd_wnd_test(meta_tp, skb, mss_now))) { ++ skb = mptcp_rcv_buf_optimization(subsk, 1); ++ if (skb) { ++ reinject = -1; ++ goto retry; ++ } ++ break; ++ } ++ ++ if (tso_segs == 1) { ++ if (unlikely(!tcp_nagle_test(meta_tp, skb, mss_now, ++ (tcp_skb_is_last(meta_sk, skb) ? ++ nonagle : TCP_NAGLE_PUSH)))) ++ break; ++ } else { ++ /* Do not try to defer the transmission of a reinjected ++ * segment. Send it directly. ++ * If it is not possible to send the TSO segment on the ++ * best subflow right now try to look for another subflow. ++ * If there is no subflow available defer the segment to avoid ++ * the call to mptso_fragment. ++ */ ++ if (!push_one && !reinject && tcp_tso_should_defer(subsk, skb)) { ++ mpcb->noneligible |= mptcp_pi_to_flag(subtp->mptcp->path_index); ++ goto subflow; ++ } ++ } ++ ++ limit = mss_now; ++ if (tso_segs > 1 && !tcp_urg_mode(meta_tp)) ++ limit = tcp_mss_split_point(subsk, skb, mss_now, ++ min_t(unsigned int, ++ cwnd_quota, ++ subsk->sk_gso_max_segs), ++ nonagle); ++ ++ if (skb->len > limit && ++ unlikely(mptso_fragment(meta_sk, skb, limit, mss_now, gfp, reinject))) ++ break; ++ ++ subskb = mptcp_skb_entail(subsk, skb, reinject); ++ if (!subskb) ++ break; ++ ++ mpcb->noneligible = noneligible; ++ TCP_SKB_CB(skb)->when = tcp_time_stamp; ++ TCP_SKB_CB(subskb)->when = tcp_time_stamp; ++ if (unlikely(tcp_transmit_skb(subsk, subskb, 1, gfp))) { ++ mptcp_transmit_skb_failed(subsk, skb, subskb); ++ mpcb->noneligible |= mptcp_pi_to_flag(subtp->mptcp->path_index); ++ continue; ++ } ++ ++ if (!reinject) { ++ mptcp_check_sndseq_wrap(meta_tp, ++ TCP_SKB_CB(skb)->end_seq - ++ TCP_SKB_CB(skb)->seq); ++ tcp_event_new_data_sent(meta_sk, skb); ++ } ++ ++ tcp_minshall_update(meta_tp, mss_now, skb); ++ sent_pkts += tcp_skb_pcount(skb); ++ tcp_sk(subsk)->mptcp->sent_pkts += tcp_skb_pcount(skb); ++ ++ mptcp_sub_event_new_data_sent(subsk, subskb, skb); ++ ++ if (reinject > 0) { ++ __skb_unlink(skb, &mpcb->reinject_queue); ++ kfree_skb(skb); ++ } ++ ++ if (push_one) ++ break; ++ } ++ ++ mpcb->noneligible = 0; ++ ++ if (likely(sent_pkts)) { ++ mptcp_for_each_sk(mpcb, subsk) { ++ subtp = tcp_sk(subsk); ++ if (subtp->mptcp->sent_pkts) { ++ if (tcp_in_cwnd_reduction(subsk)) ++ subtp->prr_out += subtp->mptcp->sent_pkts; ++ tcp_cwnd_validate(subsk); ++ subtp->mptcp->sent_pkts = 0; ++ } ++ } ++ return 0; ++ } ++ ++ return !meta_tp->packets_out && tcp_send_head(meta_sk); ++} ++ ++void mptcp_write_space(struct sock *sk) ++{ ++ mptcp_push_pending_frames(mptcp_meta_sk(sk)); ++} ++ ++u32 __mptcp_select_window(struct sock *sk) ++{ ++ struct inet_connection_sock *icsk = inet_csk(sk); ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = mptcp_meta_tp(tp); ++ int mss, free_space, full_space, window; ++ ++ /* MSS for the peer's data. Previous versions used mss_clamp ++ * here. I don't know if the value based on our guesses ++ * of peer's MSS is better for the performance. It's more correct ++ * but may be worse for the performance because of rcv_mss ++ * fluctuations. --SAW 1998/11/1 ++ */ ++ mss = icsk->icsk_ack.rcv_mss; ++ free_space = tcp_space(sk); ++ full_space = min_t(int, meta_tp->window_clamp, ++ tcp_full_space(sk)); ++ ++ if (mss > full_space) ++ mss = full_space; ++ ++ if (free_space < (full_space >> 1)) { ++ icsk->icsk_ack.quick = 0; ++ ++ if (tcp_memory_pressure) ++ /* TODO this has to be adapted when we support different ++ * MSS's among the subflows. ++ */ ++ meta_tp->rcv_ssthresh = min(meta_tp->rcv_ssthresh, ++ 4U * meta_tp->advmss); ++ ++ if (free_space < mss) ++ return 0; ++ } ++ ++ if (free_space > meta_tp->rcv_ssthresh) ++ free_space = meta_tp->rcv_ssthresh; ++ ++ /* Don't do rounding if we are using window scaling, since the ++ * scaled window will not line up with the MSS boundary anyway. ++ */ ++ window = meta_tp->rcv_wnd; ++ if (tp->rx_opt.rcv_wscale) { ++ window = free_space; ++ ++ /* Advertise enough space so that it won't get scaled away. ++ * Import case: prevent zero window announcement if ++ * 1<<rcv_wscale > mss. ++ */ ++ if (((window >> tp->rx_opt.rcv_wscale) << tp-> ++ rx_opt.rcv_wscale) != window) ++ window = (((window >> tp->rx_opt.rcv_wscale) + 1) ++ << tp->rx_opt.rcv_wscale); ++ } else { ++ /* Get the largest window that is a nice multiple of mss. ++ * Window clamp already applied above. ++ * If our current window offering is within 1 mss of the ++ * free space we just keep it. This prevents the divide ++ * and multiply from happening most of the time. ++ * We also don't do any window rounding when the free space ++ * is too small. ++ */ ++ if (window <= free_space - mss || window > free_space) ++ window = (free_space / mss) * mss; ++ else if (mss == full_space && ++ free_space > window + (full_space >> 1)) ++ window = free_space; ++ } ++ ++ return window; ++} ++ ++void mptcp_syn_options(struct sock *sk, struct tcp_out_options *opts, ++ unsigned *remaining) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ ++ opts->options |= OPTION_MPTCP; ++ if (is_master_tp(tp)) { ++ opts->mptcp_options |= OPTION_MP_CAPABLE | OPTION_TYPE_SYN; ++ *remaining -= MPTCP_SUB_LEN_CAPABLE_SYN_ALIGN; ++ opts->mp_capable.sender_key = tp->mptcp_loc_key; ++ opts->dss_csum = !!sysctl_mptcp_checksum; ++ } else { ++ struct mptcp_cb *mpcb = tp->mpcb; ++ ++ opts->mptcp_options |= OPTION_MP_JOIN | OPTION_TYPE_SYN; ++ *remaining -= MPTCP_SUB_LEN_JOIN_SYN_ALIGN; ++ opts->mp_join_syns.token = mpcb->mptcp_rem_token; ++ opts->addr_id = tp->mptcp->loc_id; ++ opts->mp_join_syns.sender_nonce = tp->mptcp->mptcp_loc_nonce; ++ } ++} ++ ++void mptcp_synack_options(struct request_sock *req, ++ struct tcp_out_options *opts, unsigned *remaining) ++{ ++ struct mptcp_request_sock *mtreq; ++ mtreq = mptcp_rsk(req); ++ ++ opts->options |= OPTION_MPTCP; ++ /* MPCB not yet set - thus it's a new MPTCP-session */ ++ if (!mtreq->mpcb) { ++ opts->mptcp_options |= OPTION_MP_CAPABLE | OPTION_TYPE_SYNACK; ++ opts->mp_capable.sender_key = mtreq->mptcp_loc_key; ++ opts->dss_csum = !!sysctl_mptcp_checksum || mtreq->dss_csum; ++ *remaining -= MPTCP_SUB_LEN_CAPABLE_SYN_ALIGN; ++ } else { ++ opts->mptcp_options |= OPTION_MP_JOIN | OPTION_TYPE_SYNACK; ++ opts->mp_join_syns.sender_truncated_mac = ++ mtreq->mptcp_hash_tmac; ++ opts->mp_join_syns.sender_nonce = mtreq->mptcp_loc_nonce; ++ opts->addr_id = mtreq->loc_id; ++ *remaining -= MPTCP_SUB_LEN_JOIN_SYNACK_ALIGN; ++ } ++} ++ ++void mptcp_established_options(struct sock *sk, struct sk_buff *skb, ++ struct tcp_out_options *opts, unsigned *size) ++{ ++ struct tcp_sock *tp = tcp_sk(sk), *meta_tp = mptcp_meta_tp(tp); ++ struct mptcp_cb *mpcb = tp->mpcb; ++ struct tcp_skb_cb *tcb = skb ? TCP_SKB_CB(skb) : NULL; ++ ++ /* In fallback mp_fail-mode, we have to repeat it until the fallback ++ * has been done by the sender ++ */ ++ if (unlikely(tp->mptcp->send_mp_fail)) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_MP_FAIL; ++ opts->data_ack = (__u32)(mpcb->csum_cutoff_seq >> 32); ++ opts->data_seq = (__u32)mpcb->csum_cutoff_seq; ++ *size += MPTCP_SUB_LEN_FAIL; ++ return; ++ } ++ ++ if (unlikely(tp->send_mp_fclose)) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_MP_FCLOSE; ++ opts->mp_capable.receiver_key = mpcb->mptcp_rem_key; ++ *size += MPTCP_SUB_LEN_FCLOSE_ALIGN; ++ return; ++ } ++ ++ /* 1. If we are the sender of the infinite-mapping, we need the ++ * MPTCPHDR_INF-flag, because a retransmission of the ++ * infinite-announcment still needs the mptcp-option. ++ * ++ * We need infinite_cutoff_seq, because retransmissions from before ++ * the infinite-cutoff-moment still need the MPTCP-signalling to stay ++ * consistent. ++ * ++ * 2. If we are the receiver of the infinite-mapping, we always skip ++ * mptcp-options, because acknowledgments from before the ++ * infinite-mapping point have already been sent out. ++ * ++ * I know, the whole infinite-mapping stuff is ugly... ++ * ++ * TODO: Handle wrapped data-sequence numbers ++ * (even if it's very unlikely) ++ */ ++ if (unlikely(mpcb->infinite_mapping_snd) && ++ tp->mptcp->fully_established && ++ ((mpcb->send_infinite_mapping && tcb && ++ !(tcb->mptcp_flags & MPTCPHDR_INF) && ++ !before(tcb->seq, tp->mptcp->infinite_cutoff_seq)) || ++ !mpcb->send_infinite_mapping)) ++ return; ++ ++ if (unlikely(tp->mptcp->include_mpc)) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_MP_CAPABLE | ++ OPTION_TYPE_ACK; ++ *size += MPTCP_SUB_LEN_CAPABLE_ACK_ALIGN; ++ opts->mp_capable.sender_key = mpcb->mptcp_loc_key; ++ opts->mp_capable.receiver_key = mpcb->mptcp_rem_key; ++ opts->dss_csum = mpcb->dss_csum; ++ ++ if (skb) ++ tp->mptcp->include_mpc = 0; ++ } ++ if (unlikely(tp->mptcp->pre_established)) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_MP_JOIN | OPTION_TYPE_ACK; ++ *size += MPTCP_SUB_LEN_JOIN_ACK_ALIGN; ++ } ++ ++ if (!tp->mptcp->include_mpc && !tp->mptcp->pre_established) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_DATA_ACK; ++ /* If !skb, we come from tcp_current_mss and thus we always ++ * assume that the DSS-option will be set for the data-packet. ++ */ ++ if (skb && !mptcp_is_data_seq(skb)) { ++ opts->data_ack = meta_tp->rcv_nxt; ++ ++ *size += MPTCP_SUB_LEN_ACK_ALIGN; ++ } else { ++ opts->data_ack = meta_tp->rcv_nxt; ++ ++ /* Doesn't matter, if csum included or not. It will be ++ * either 10 or 12, and thus aligned = 12 ++ */ ++ *size += MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN; ++ } ++ ++ *size += MPTCP_SUB_LEN_DSS_ALIGN; ++ } ++ ++ if (mpcb->pm_ops->addr_signal) ++ mpcb->pm_ops->addr_signal(sk, size, opts, skb); ++ ++ if (unlikely(tp->mptcp->send_mp_prio) && ++ MAX_TCP_OPTION_SPACE - *size >= MPTCP_SUB_LEN_PRIO_ALIGN) { ++ opts->options |= OPTION_MPTCP; ++ opts->mptcp_options |= OPTION_MP_PRIO; ++ if (skb) ++ tp->mptcp->send_mp_prio = 0; ++ *size += MPTCP_SUB_LEN_PRIO_ALIGN; ++ } ++ ++ return; ++} ++ ++u16 mptcp_select_window(struct sock *sk) ++{ ++ u16 new_win = tcp_select_window(sk); ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct tcp_sock *meta_tp = mptcp_meta_tp(tp); ++ ++ meta_tp->rcv_wnd = tp->rcv_wnd; ++ meta_tp->rcv_wup = meta_tp->rcv_nxt; ++ ++ return new_win; ++} ++ ++void mptcp_options_write(__be32 *ptr, struct tcp_sock *tp, ++ struct tcp_out_options *opts, ++ struct sk_buff *skb) ++{ ++ if (unlikely(OPTION_MP_CAPABLE & opts->mptcp_options)) { ++ struct mp_capable *mpc = (struct mp_capable *)ptr; ++ ++ mpc->kind = TCPOPT_MPTCP; ++ ++ if ((OPTION_TYPE_SYN & opts->mptcp_options) || ++ (OPTION_TYPE_SYNACK & opts->mptcp_options)) { ++ mpc->sender_key = opts->mp_capable.sender_key; ++ mpc->len = MPTCP_SUB_LEN_CAPABLE_SYN; ++ ptr += MPTCP_SUB_LEN_CAPABLE_SYN_ALIGN >> 2; ++ } else if (OPTION_TYPE_ACK & opts->mptcp_options) { ++ mpc->sender_key = opts->mp_capable.sender_key; ++ mpc->receiver_key = opts->mp_capable.receiver_key; ++ mpc->len = MPTCP_SUB_LEN_CAPABLE_ACK; ++ ptr += MPTCP_SUB_LEN_CAPABLE_ACK_ALIGN >> 2; ++ } ++ ++ mpc->sub = MPTCP_SUB_CAPABLE; ++ mpc->ver = 0; ++ mpc->a = opts->dss_csum; ++ mpc->b = 0; ++ mpc->rsv = 0; ++ mpc->h = 1; ++ } ++ ++ if (unlikely(OPTION_MP_JOIN & opts->mptcp_options)) { ++ struct mp_join *mpj = (struct mp_join *)ptr; ++ ++ mpj->kind = TCPOPT_MPTCP; ++ mpj->sub = MPTCP_SUB_JOIN; ++ mpj->rsv = 0; ++ mpj->addr_id = opts->addr_id; ++ ++ if (OPTION_TYPE_SYN & opts->mptcp_options) { ++ mpj->len = MPTCP_SUB_LEN_JOIN_SYN; ++ mpj->u.syn.token = opts->mp_join_syns.token; ++ mpj->u.syn.nonce = opts->mp_join_syns.sender_nonce; ++ mpj->b = tp->mptcp->low_prio; ++ ptr += MPTCP_SUB_LEN_JOIN_SYN_ALIGN >> 2; ++ } else if (OPTION_TYPE_SYNACK & opts->mptcp_options) { ++ mpj->len = MPTCP_SUB_LEN_JOIN_SYNACK; ++ mpj->u.synack.mac = ++ opts->mp_join_syns.sender_truncated_mac; ++ mpj->u.synack.nonce = opts->mp_join_syns.sender_nonce; ++ mpj->b = tp->mptcp->low_prio; ++ ptr += MPTCP_SUB_LEN_JOIN_SYNACK_ALIGN >> 2; ++ } else if (OPTION_TYPE_ACK & opts->mptcp_options) { ++ mpj->len = MPTCP_SUB_LEN_JOIN_ACK; ++ memcpy(mpj->u.ack.mac, &tp->mptcp->sender_mac[0], 20); ++ ptr += MPTCP_SUB_LEN_JOIN_ACK_ALIGN >> 2; ++ } ++ } ++ if (unlikely(OPTION_ADD_ADDR & opts->mptcp_options)) { ++ struct mp_add_addr *mpadd = (struct mp_add_addr *)ptr; ++ ++ mpadd->kind = TCPOPT_MPTCP; ++ if (opts->add_addr_v4) { ++ mpadd->len = MPTCP_SUB_LEN_ADD_ADDR4; ++ mpadd->sub = MPTCP_SUB_ADD_ADDR; ++ mpadd->ipver = 4; ++ mpadd->addr_id = opts->add_addr4.addr_id; ++ mpadd->u.v4.addr = opts->add_addr4.addr; ++ ptr += MPTCP_SUB_LEN_ADD_ADDR4_ALIGN >> 2; ++ } else if (opts->add_addr_v6) { ++ mpadd->len = MPTCP_SUB_LEN_ADD_ADDR6; ++ mpadd->sub = MPTCP_SUB_ADD_ADDR; ++ mpadd->ipver = 6; ++ mpadd->addr_id = opts->add_addr6.addr_id; ++ memcpy(&mpadd->u.v6.addr, &opts->add_addr6.addr, ++ sizeof(mpadd->u.v6.addr)); ++ ptr += MPTCP_SUB_LEN_ADD_ADDR6_ALIGN >> 2; ++ } ++ } ++ if (unlikely(OPTION_REMOVE_ADDR & opts->mptcp_options)) { ++ struct mp_remove_addr *mprem = (struct mp_remove_addr *)ptr; ++ u8 *addrs_id; ++ int id, len, len_align; ++ ++ len = mptcp_sub_len_remove_addr(opts->remove_addrs); ++ len_align = mptcp_sub_len_remove_addr_align(opts->remove_addrs); ++ ++ mprem->kind = TCPOPT_MPTCP; ++ mprem->len = len; ++ mprem->sub = MPTCP_SUB_REMOVE_ADDR; ++ mprem->rsv = 0; ++ addrs_id = &mprem->addrs_id; ++ ++ mptcp_for_each_bit_set(opts->remove_addrs, id) ++ *(addrs_id++) = id; ++ ++ /* Fill the rest with NOP's */ ++ if (len_align > len) { ++ int i; ++ for (i = 0; i < len_align - len; i++) ++ *(addrs_id++) = TCPOPT_NOP; ++ } ++ ++ ptr += len_align >> 2; ++ } ++ if (unlikely(OPTION_MP_FAIL & opts->mptcp_options)) { ++ struct mp_fail *mpfail = (struct mp_fail *)ptr; ++ ++ mpfail->kind = TCPOPT_MPTCP; ++ mpfail->len = MPTCP_SUB_LEN_FAIL; ++ mpfail->sub = MPTCP_SUB_FAIL; ++ mpfail->rsv1 = 0; ++ mpfail->rsv2 = 0; ++ mpfail->data_seq = htonll(((u64)opts->data_ack << 32) | opts->data_seq); ++ ++ ptr += MPTCP_SUB_LEN_FAIL_ALIGN >> 2; ++ } ++ if (unlikely(OPTION_MP_FCLOSE & opts->mptcp_options)) { ++ struct mp_fclose *mpfclose = (struct mp_fclose *)ptr; ++ ++ mpfclose->kind = TCPOPT_MPTCP; ++ mpfclose->len = MPTCP_SUB_LEN_FCLOSE; ++ mpfclose->sub = MPTCP_SUB_FCLOSE; ++ mpfclose->rsv1 = 0; ++ mpfclose->rsv2 = 0; ++ mpfclose->key = opts->mp_capable.receiver_key; ++ ++ ptr += MPTCP_SUB_LEN_FCLOSE_ALIGN >> 2; ++ } ++ ++ if (OPTION_DATA_ACK & opts->mptcp_options) { ++ if (!mptcp_is_data_seq(skb)) { ++ struct mp_dss *mdss = (struct mp_dss *)ptr; ++ ++ mdss->kind = TCPOPT_MPTCP; ++ mdss->sub = MPTCP_SUB_DSS; ++ mdss->rsv1 = 0; ++ mdss->rsv2 = 0; ++ mdss->F = 0; ++ mdss->m = 0; ++ mdss->M = 0; ++ mdss->a = 0; ++ mdss->A = 1; ++ mdss->len = mptcp_sub_len_dss(mdss, tp->mpcb->dss_csum); ++ ++ ptr++; ++ *ptr++ = htonl(opts->data_ack); ++ } else { ++ /**** Just update the data_ack ****/ ++ ++ /* Get pointer to data_ack-field. MPTCP is always at ++ * the end of the TCP-options. ++ */ ++ /* TODO if we allow sending 64-bit dseq's we have to change "16" */ ++ __be32 *dack = (__be32 *)(skb->data + (tcp_hdr(skb)->doff << 2) - 16); ++ ++ *dack = htonl(opts->data_ack); ++ } ++ } ++ if (unlikely(OPTION_MP_PRIO & opts->mptcp_options)) { ++ struct mp_prio *mpprio = (struct mp_prio *)ptr; ++ ++ mpprio->kind = TCPOPT_MPTCP; ++ mpprio->len = MPTCP_SUB_LEN_PRIO; ++ mpprio->sub = MPTCP_SUB_PRIO; ++ mpprio->rsv = 0; ++ mpprio->b = tp->mptcp->low_prio; ++ mpprio->addr_id = TCPOPT_NOP; ++ ++ ptr += MPTCP_SUB_LEN_PRIO_ALIGN >> 2; ++ } ++} ++ ++/* Returns the next segment to be sent from the mptcp meta-queue. ++ * (chooses the reinject queue if any segment is waiting in it, otherwise, ++ * chooses the normal write queue). ++ * Sets *@reinject to 1 if the returned segment comes from the ++ * reinject queue. Sets it to 0 if it is the regular send-head of the meta-sk, ++ * and sets it to -1 if it is a meta-level retransmission to optimize the ++ * receive-buffer. ++ */ ++struct sk_buff *mptcp_next_segment(struct sock *meta_sk, int *reinject) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(meta_sk)->mpcb; ++ struct sk_buff *skb = NULL; ++ if (reinject) ++ *reinject = 0; ++ ++ /* If we are in fallback-mode, just take from the meta-send-queue */ ++ if (mpcb->infinite_mapping_snd || mpcb->send_infinite_mapping) ++ return tcp_send_head(meta_sk); ++ ++ skb = skb_peek(&mpcb->reinject_queue); ++ ++ if (skb) { ++ if (reinject) ++ *reinject = 1; ++ } else { ++ skb = tcp_send_head(meta_sk); ++ ++ if (!skb && meta_sk->sk_socket && ++ test_bit(SOCK_NOSPACE, &meta_sk->sk_socket->flags) && ++ sk_stream_wspace(meta_sk) < sk_stream_min_wspace(meta_sk)) { ++ struct sock *subsk = get_available_subflow(meta_sk, NULL, NULL); ++ if (!subsk) ++ return NULL; ++ ++ skb = mptcp_rcv_buf_optimization(subsk, 0); ++ if (skb && reinject) ++ *reinject = -1; ++ } ++ } ++ return skb; ++} ++ ++/* Sends the datafin */ ++void mptcp_send_fin(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sk_buff *skb = tcp_write_queue_tail(meta_sk); ++ int mss_now; ++ ++ if ((1 << meta_sk->sk_state) & (TCPF_CLOSE_WAIT | TCPF_LAST_ACK)) ++ meta_tp->mpcb->passive_close = 1; ++ ++ /* Optimization, tack on the FIN if we have a queue of ++ * unsent frames. But be careful about outgoing SACKS ++ * and IP options. ++ */ ++ mss_now = mptcp_current_mss(meta_sk); ++ ++ if (tcp_send_head(meta_sk) != NULL) { ++ TCP_SKB_CB(skb)->mptcp_flags |= MPTCPHDR_FIN; ++ TCP_SKB_CB(skb)->end_seq++; ++ meta_tp->write_seq++; ++ } else { ++ /* Socket is locked, keep trying until memory is available. */ ++ for (;;) { ++ skb = alloc_skb_fclone(MAX_TCP_HEADER, ++ meta_sk->sk_allocation); ++ if (skb) ++ break; ++ yield(); ++ } ++ /* Reserve space for headers and prepare control bits. */ ++ skb_reserve(skb, MAX_TCP_HEADER); ++ ++ tcp_init_nondata_skb(skb, meta_tp->write_seq, TCPHDR_ACK); ++ TCP_SKB_CB(skb)->end_seq++; ++ TCP_SKB_CB(skb)->mptcp_flags |= MPTCPHDR_FIN | MPTCPHDR_SEQ; ++ tcp_queue_skb(meta_sk, skb); ++ } ++ __tcp_push_pending_frames(meta_sk, mss_now, TCP_NAGLE_OFF); ++} ++ ++void mptcp_send_active_reset(struct sock *meta_sk, gfp_t priority) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct sock *sk = NULL, *sk_it = NULL, *tmpsk; ++ ++ if (!mpcb->cnt_subflows) ++ return; ++ ++ WARN_ON(meta_tp->send_mp_fclose); ++ ++ /* First - select a socket */ ++ sk = mptcp_select_ack_sock(meta_sk, 0); ++ ++ /* May happen if no subflow is in an appropriate state */ ++ if (!sk) ++ return; ++ ++ /* We are in infinite mode - just send a reset */ ++ if (mpcb->infinite_mapping_snd || mpcb->infinite_mapping_rcv) { ++ sk->sk_err = ECONNRESET; ++ if (tcp_need_reset(sk->sk_state)) ++ tcp_send_active_reset(sk, priority); ++ mptcp_sub_force_close(sk); ++ return; ++ } ++ ++ ++ tcp_sk(sk)->send_mp_fclose = 1; ++ /** Reset all other subflows */ ++ ++ /* tcp_done must be handled with bh disabled */ ++ if (!in_serving_softirq()) ++ local_bh_disable(); ++ ++ mptcp_for_each_sk_safe(mpcb, sk_it, tmpsk) { ++ if (tcp_sk(sk_it)->send_mp_fclose) ++ continue; ++ ++ sk_it->sk_err = ECONNRESET; ++ if (tcp_need_reset(sk_it->sk_state)) ++ tcp_send_active_reset(sk_it, GFP_ATOMIC); ++ mptcp_sub_force_close(sk_it); ++ } ++ ++ if (!in_serving_softirq()) ++ local_bh_enable(); ++ ++ tcp_send_ack(sk); ++ inet_csk_reset_keepalive_timer(sk, inet_csk(sk)->icsk_rto); ++ ++ meta_tp->send_mp_fclose = 1; ++} ++ ++static void mptcp_ack_retransmit_timer(struct sock *sk) ++{ ++ struct sk_buff *skb; ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct inet_connection_sock *icsk = inet_csk(sk); ++ ++ if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk)) ++ goto out; /* Routing failure or similar */ ++ ++ if (!tp->retrans_stamp) ++ tp->retrans_stamp = tcp_time_stamp ? : 1; ++ ++ if (tcp_write_timeout(sk)) { ++ tp->mptcp->pre_established = 0; ++ sk_stop_timer(sk, &tp->mptcp->mptcp_ack_timer); ++ tcp_send_active_reset(sk, GFP_ATOMIC); ++ goto out; ++ } ++ ++ skb = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC); ++ if (skb == NULL) { ++ sk_reset_timer(sk, &tp->mptcp->mptcp_ack_timer, ++ jiffies + icsk->icsk_rto); ++ return; ++ } ++ ++ /* Reserve space for headers and prepare control bits */ ++ skb_reserve(skb, MAX_TCP_HEADER); ++ tcp_init_nondata_skb(skb, tp->snd_una, TCPHDR_ACK); ++ ++ TCP_SKB_CB(skb)->when = tcp_time_stamp; ++ if (tcp_transmit_skb(sk, skb, 0, GFP_ATOMIC) > 0) { ++ /* Retransmission failed because of local congestion, ++ * do not backoff. ++ */ ++ if (!icsk->icsk_retransmits) ++ icsk->icsk_retransmits = 1; ++ sk_reset_timer(sk, &tp->mptcp->mptcp_ack_timer, ++ jiffies + icsk->icsk_rto); ++ return; ++ } ++ ++ ++ icsk->icsk_retransmits++; ++ icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX); ++ sk_reset_timer(sk, &tp->mptcp->mptcp_ack_timer, ++ jiffies + icsk->icsk_rto); ++ if (retransmits_timed_out(sk, sysctl_tcp_retries1 + 1, 0, 0)) { ++ __sk_dst_reset(sk); ++ } ++ ++out:; ++} ++ ++void mptcp_ack_handler(unsigned long data) ++{ ++ struct sock *sk = (struct sock *)data; ++ struct sock *meta_sk = mptcp_meta_sk(sk); ++ ++ bh_lock_sock(meta_sk); ++ if (sock_owned_by_user(meta_sk)) { ++ /* Try again later */ ++ sk_reset_timer(sk, &tcp_sk(sk)->mptcp->mptcp_ack_timer, ++ jiffies + (HZ / 20)); ++ goto out_unlock; ++ } ++ ++ if (sk->sk_state == TCP_CLOSE) ++ goto out_unlock; ++ ++ mptcp_ack_retransmit_timer(sk); ++ ++ sk_mem_reclaim(sk); ++ ++out_unlock: ++ bh_unlock_sock(meta_sk); ++ sock_put(sk); ++} ++ ++/* Similar to tcp_retransmit_skb ++ * ++ * The diff is that we handle the retransmission-stats (retrans_stamp) at the ++ * meta-level. ++ */ ++int mptcp_retransmit_skb(struct sock *meta_sk, struct sk_buff *skb) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct sock *subsk; ++ struct sk_buff *subskb; ++ unsigned int limit, tso_segs, mss_now; ++ int err = -1, oldpcount; ++ ++ /* Do not sent more than we queued. 1/4 is reserved for possible ++ * copying overhead: fragmentation, tunneling, mangling etc. ++ * ++ * This is a meta-retransmission thus we check on the meta-socket. ++ */ ++ if (atomic_read(&meta_sk->sk_wmem_alloc) > ++ min(meta_sk->sk_wmem_queued + (meta_sk->sk_wmem_queued >> 2), meta_sk->sk_sndbuf)) { ++ return -EAGAIN; ++ } ++ ++ /* We need to make sure that the retransmitted segment can be sent on a ++ * subflow right now. If it is too big, it needs to be fragmented. ++ */ ++ subsk = get_available_subflow(meta_sk, skb, &mss_now); ++ if (!subsk) { ++ /* We want to increase icsk_retransmits, thus return 0, so that ++ * mptcp_retransmit_timer enters the desired branch. ++ */ ++ err = 0; ++ goto failed; ++ } ++ ++ /* If the segment was cloned (e.g. a meta retransmission), the header ++ * must be expanded/copied so that there is no corruption of TSO ++ * information. ++ */ ++ if (skb_unclone(skb, GFP_ATOMIC)) { ++ err = ENOMEM; ++ goto failed; ++ } ++ ++ oldpcount = tcp_skb_pcount(skb); ++ tcp_set_skb_tso_segs(meta_sk, skb, mss_now); ++ tso_segs = tcp_skb_pcount(skb); ++ BUG_ON(!tso_segs); ++ ++ /* The MSS might have changed and so the number of segments. We ++ * need to account for this change. ++ */ ++ if (unlikely(oldpcount != tso_segs)) ++ tcp_adjust_pcount(meta_sk, skb, oldpcount - tso_segs); ++ ++ limit = mss_now; ++ if (tso_segs > 1 && !tcp_urg_mode(meta_tp)) ++ limit = tcp_mss_split_point(subsk, skb, mss_now, ++ min_t(unsigned int, ++ tcp_cwnd_test(tcp_sk(subsk), skb), ++ subsk->sk_gso_max_segs), ++ TCP_NAGLE_OFF); ++ ++ if (skb->len > limit && ++ unlikely(mptso_fragment(meta_sk, skb, limit, mss_now, ++ GFP_ATOMIC, 0))) ++ goto failed; ++ ++ subskb = mptcp_skb_entail(subsk, skb, -1); ++ if (!subskb) ++ goto failed; ++ ++ TCP_SKB_CB(skb)->when = tcp_time_stamp; ++ TCP_SKB_CB(subskb)->when = tcp_time_stamp; ++ err = tcp_transmit_skb(subsk, subskb, 1, GFP_ATOMIC); ++ if (!err) { ++ /* Update global TCP statistics. */ ++ TCP_INC_STATS(sock_net(meta_sk), TCP_MIB_RETRANSSEGS); ++ ++ /* Diff to tcp_retransmit_skb */ ++ ++ /* Save stamp of the first retransmit. */ ++ if (!meta_tp->retrans_stamp) ++ meta_tp->retrans_stamp = TCP_SKB_CB(subskb)->when; ++ mptcp_sub_event_new_data_sent(subsk, subskb, skb); ++ } else { ++ mptcp_transmit_skb_failed(subsk, skb, subskb); ++ } ++ ++failed: ++ return err; ++} ++ ++/* Similar to tcp_retransmit_timer ++ * ++ * The diff is that we have to handle retransmissions of the FAST_CLOSE-message ++ * and that we don't have an srtt estimation at the meta-level. ++ */ ++void mptcp_retransmit_timer(struct sock *meta_sk) ++{ ++ struct tcp_sock *meta_tp = tcp_sk(meta_sk); ++ struct mptcp_cb *mpcb = meta_tp->mpcb; ++ struct inet_connection_sock *meta_icsk = inet_csk(meta_sk); ++ int err; ++ ++ /* In fallback, retransmission is handled at the subflow-level */ ++ if (!meta_tp->packets_out || mpcb->infinite_mapping_snd || ++ mpcb->send_infinite_mapping) ++ return; ++ ++ WARN_ON(tcp_write_queue_empty(meta_sk)); ++ ++ if (!meta_tp->snd_wnd && !sock_flag(meta_sk, SOCK_DEAD) && ++ !((1 << meta_sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV))) { ++ /* Receiver dastardly shrinks window. Our retransmits ++ * become zero probes, but we should not timeout this ++ * connection. If the socket is an orphan, time it out, ++ * we cannot allow such beasts to hang infinitely. ++ */ ++ struct inet_sock *meta_inet = inet_sk(meta_sk); ++ if (meta_sk->sk_family == AF_INET) { ++ LIMIT_NETDEBUG(KERN_DEBUG "MPTCP: Peer %pI4:%u/%u unexpectedly shrunk window %u:%u (repaired)\n", ++ &meta_inet->inet_daddr, ++ ntohs(meta_inet->inet_dport), ++ meta_inet->inet_num, meta_tp->snd_una, ++ meta_tp->snd_nxt); ++ } ++#if IS_ENABLED(CONFIG_IPV6) ++ else if (meta_sk->sk_family == AF_INET6) { ++ LIMIT_NETDEBUG(KERN_DEBUG "MPTCP: Peer %pI6:%u/%u unexpectedly shrunk window %u:%u (repaired)\n", ++ &meta_sk->sk_v6_daddr, ++ ntohs(meta_inet->inet_dport), ++ meta_inet->inet_num, meta_tp->snd_una, ++ meta_tp->snd_nxt); ++ } ++#endif ++ if (tcp_time_stamp - meta_tp->rcv_tstamp > TCP_RTO_MAX) { ++ tcp_write_err(meta_sk); ++ return; ++ } ++ ++ mptcp_retransmit_skb(meta_sk, tcp_write_queue_head(meta_sk)); ++ goto out_reset_timer; ++ } ++ ++ if (tcp_write_timeout(meta_sk)) ++ return; ++ ++ if (meta_icsk->icsk_retransmits == 0) ++ NET_INC_STATS_BH(sock_net(meta_sk), LINUX_MIB_TCPTIMEOUTS); ++ ++ meta_icsk->icsk_ca_state = TCP_CA_Loss; ++ ++ err = mptcp_retransmit_skb(meta_sk, tcp_write_queue_head(meta_sk)); ++ if (err > 0) { ++ /* Retransmission failed because of local congestion, ++ * do not backoff. ++ */ ++ if (!meta_icsk->icsk_retransmits) ++ meta_icsk->icsk_retransmits = 1; ++ inet_csk_reset_xmit_timer(meta_sk, ICSK_TIME_RETRANS, ++ min(meta_icsk->icsk_rto, TCP_RESOURCE_PROBE_INTERVAL), ++ TCP_RTO_MAX); ++ return; ++ } ++ ++ /* Increase the timeout each time we retransmit. Note that ++ * we do not increase the rtt estimate. rto is initialized ++ * from rtt, but increases here. Jacobson (SIGCOMM 88) suggests ++ * that doubling rto each time is the least we can get away with. ++ * In KA9Q, Karn uses this for the first few times, and then ++ * goes to quadratic. netBSD doubles, but only goes up to *64, ++ * and clamps at 1 to 64 sec afterwards. Note that 120 sec is ++ * defined in the protocol as the maximum possible RTT. I guess ++ * we'll have to use something other than TCP to talk to the ++ * University of Mars. ++ * ++ * PAWS allows us longer timeouts and large windows, so once ++ * implemented ftp to mars will work nicely. We will have to fix ++ * the 120 second clamps though! ++ */ ++ meta_icsk->icsk_backoff++; ++ meta_icsk->icsk_retransmits++; ++ ++out_reset_timer: ++ /* If stream is thin, use linear timeouts. Since 'icsk_backoff' is ++ * used to reset timer, set to 0. Recalculate 'icsk_rto' as this ++ * might be increased if the stream oscillates between thin and thick, ++ * thus the old value might already be too high compared to the value ++ * set by 'tcp_set_rto' in tcp_input.c which resets the rto without ++ * backoff. Limit to TCP_THIN_LINEAR_RETRIES before initiating ++ * exponential backoff behaviour to avoid continue hammering ++ * linear-timeout retransmissions into a black hole ++ */ ++ if (meta_sk->sk_state == TCP_ESTABLISHED && ++ (meta_tp->thin_lto || sysctl_tcp_thin_linear_timeouts) && ++ tcp_stream_is_thin(meta_tp) && ++ meta_icsk->icsk_retransmits <= TCP_THIN_LINEAR_RETRIES) { ++ meta_icsk->icsk_backoff = 0; ++ /* We cannot do the same as in tcp_write_timer because the ++ * srtt is not set here. ++ */ ++ mptcp_set_rto(meta_sk); ++ } else { ++ /* Use normal (exponential) backoff */ ++ meta_icsk->icsk_rto = min(meta_icsk->icsk_rto << 1, TCP_RTO_MAX); ++ } ++ inet_csk_reset_xmit_timer(meta_sk, ICSK_TIME_RETRANS, meta_icsk->icsk_rto, TCP_RTO_MAX); ++ ++ return; ++} ++ ++/* Modify values to an mptcp-level for the initial window of new subflows */ ++void mptcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd, ++ __u32 *window_clamp, int wscale_ok, ++ __u8 *rcv_wscale, __u32 init_rcv_wnd, ++ const struct sock *sk) ++{ ++ struct mptcp_cb *mpcb = tcp_sk(sk)->mpcb; ++ ++ *window_clamp = mpcb->orig_window_clamp; ++ __space = tcp_win_from_space(mpcb->orig_sk_rcvbuf); ++ ++ tcp_select_initial_window(__space, mss, rcv_wnd, window_clamp, ++ wscale_ok, rcv_wscale, init_rcv_wnd, sk); ++} ++ ++unsigned int mptcp_current_mss(struct sock *meta_sk) ++{ ++ unsigned int mss = 0; ++ struct sock *sk; ++ ++ mptcp_for_each_sk(tcp_sk(meta_sk)->mpcb, sk) { ++ int this_mss; ++ ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ ++ this_mss = tcp_current_mss(sk); ++ if (this_mss > mss) ++ mss = this_mss; ++ } ++ ++ /* If no subflow is available, we take a default-mss from the ++ * meta-socket. ++ */ ++ return !mss ? tcp_current_mss(meta_sk) : mss; ++} ++ ++int mptcp_select_size(const struct sock *meta_sk, bool sg) ++{ ++ int mss = 0; /* We look for the smallest MSS */ ++ struct sock *sk; ++ ++ mptcp_for_each_sk(tcp_sk(meta_sk)->mpcb, sk) { ++ int this_mss; ++ ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ ++ this_mss = tcp_sk(sk)->mss_cache; ++ if (this_mss > mss) ++ mss = this_mss; ++ } ++ ++ if (sg) { ++ if (mptcp_sk_can_gso(meta_sk)) { ++ mss = SKB_WITH_OVERHEAD(2048 - MAX_TCP_HEADER); ++ } else { ++ int pgbreak = SKB_MAX_HEAD(MAX_TCP_HEADER); ++ ++ if (mss >= pgbreak && ++ mss <= pgbreak + (MAX_SKB_FRAGS - 1) * PAGE_SIZE) ++ mss = pgbreak; ++ } ++ } ++ ++ return !mss ? tcp_sk(meta_sk)->mss_cache : mss; ++} ++ ++int mptcp_check_snd_buf(const struct tcp_sock *tp) ++{ ++ struct sock *sk; ++ u32 rtt_max = tp->srtt; ++ u64 bw_est; ++ ++ if (!tp->srtt) ++ return tp->reordering + 1; ++ ++ mptcp_for_each_sk(tp->mpcb, sk) { ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ ++ if (rtt_max < tcp_sk(sk)->srtt) ++ rtt_max = tcp_sk(sk)->srtt; ++ } ++ ++ bw_est = div64_u64(((u64)tp->snd_cwnd * rtt_max) << 16, ++ (u64)tp->srtt); ++ ++ return max_t(unsigned int, (u32)(bw_est >> 16), ++ tp->reordering + 1); ++ ++} ++ ++unsigned int mptcp_xmit_size_goal(struct sock *meta_sk, u32 mss_now, ++ int large_allowed) ++{ ++ struct sock *sk; ++ u32 xmit_size_goal = 0; ++ ++ if (large_allowed && mptcp_sk_can_gso(meta_sk)) { ++ mptcp_for_each_sk(tcp_sk(meta_sk)->mpcb, sk) { ++ int this_size_goal; ++ ++ if (!mptcp_sk_can_send(sk)) ++ continue; ++ ++ this_size_goal = tcp_xmit_size_goal(sk, mss_now, 1); ++ if (this_size_goal > xmit_size_goal) ++ xmit_size_goal = this_size_goal; ++ } ++ } ++ ++ return max(xmit_size_goal, mss_now); ++} ++ ++/* Similar to tcp_trim_head - but we correctly copy the DSS-option */ ++int mptcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len) ++{ ++ int dsslen = MPTCP_SUB_LEN_DSS_ALIGN + MPTCP_SUB_LEN_ACK_ALIGN + ++ MPTCP_SUB_LEN_SEQ_ALIGN; ++ char dss[dsslen]; ++ ++ /* DSS-option must be recovered afterwards. */ ++ memcpy(dss, skb->data - dsslen, dsslen); ++ ++ if (skb_cloned(skb)) { ++ /* pskb_expand_head will delete our DSS-option. We have to copy ++ * it back if pskb_expand_head succeeds. ++ */ ++ ++ if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC)) ++ return -ENOMEM; ++ ++ memcpy(skb->data - dsslen, dss, dsslen); ++ } ++ ++ __pskb_trim_head(skb, len); ++ ++ /* Put the DSS-option back in our header */ ++ memcpy(skb->data - dsslen, dss, dsslen); ++ ++ TCP_SKB_CB(skb)->seq += len; ++ skb->ip_summed = CHECKSUM_PARTIAL; ++ ++ skb->truesize -= len; ++ sk->sk_wmem_queued -= len; ++ sk_mem_uncharge(sk, len); ++ sock_set_flag(sk, SOCK_QUEUE_SHRUNK); ++ ++ /* Any change of skb->len requires recalculation of tso factor. */ ++ if (tcp_skb_pcount(skb) > 1) ++ tcp_set_skb_tso_segs(sk, skb, tcp_skb_mss(skb)); ++ ++ return 0; ++} +diff --git a/net/mptcp/mptcp_pm.c b/net/mptcp/mptcp_pm.c +new file mode 100644 +index 0000000..1f78ae1 +--- /dev/null ++++ b/net/mptcp/mptcp_pm.c +@@ -0,0 +1,170 @@ ++/* ++ * MPTCP implementation - MPTCP-subflow-management ++ * ++ * Initial Design & Implementation: ++ * Sébastien Barré <sebastien.barre@uclouvain.be> ++ * ++ * Current Maintainer & Author: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * Additional authors: ++ * Jaakko Korkeaniemi <jaakko.korkeaniemi@aalto.fi> ++ * Gregory Detal <gregory.detal@uclouvain.be> ++ * Fabien Duchêne <fabien.duchene@uclouvain.be> ++ * Andreas Seelinger <Andreas.Seelinger@rwth-aachen.de> ++ * Lavkesh Lahngir <lavkesh51@gmail.com> ++ * Andreas Ripke <ripke@neclab.eu> ++ * Vlad Dogaru <vlad.dogaru@intel.com> ++ * Octavian Purdila <octavian.purdila@intel.com> ++ * John Ronan <jronan@tssg.org> ++ * Catalin Nicutar <catalin.nicutar@gmail.com> ++ * Brandon Heller <brandonh@stanford.edu> ++ * ++ * ++ * 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. ++ */ ++ ++ ++#include <linux/module.h> ++#include <net/mptcp.h> ++ ++static DEFINE_SPINLOCK(mptcp_pm_list_lock); ++static LIST_HEAD(mptcp_pm_list); ++ ++static int mptcp_default_index(sa_family_t family, union inet_addr *addr, ++ struct net *net) ++{ ++ return 0; ++} ++ ++struct mptcp_pm_ops mptcp_pm_default = { ++ .get_local_index = mptcp_default_index, ++ .get_local_id = mptcp_default_index, /* We do not care */ ++ .name = "default", ++ .owner = THIS_MODULE, ++}; ++ ++static struct mptcp_pm_ops *mptcp_pm_find(const char *name) ++{ ++ struct mptcp_pm_ops *e; ++ ++ list_for_each_entry_rcu(e, &mptcp_pm_list, list) { ++ if (strcmp(e->name, name) == 0) ++ return e; ++ } ++ ++ return NULL; ++} ++ ++int mptcp_register_path_manager(struct mptcp_pm_ops *pm) ++{ ++ int ret = 0; ++ ++ if (!pm->get_local_index || !pm->get_local_id) ++ return -EINVAL; ++ ++ spin_lock(&mptcp_pm_list_lock); ++ if (mptcp_pm_find(pm->name)) { ++ pr_notice("%s already registered\n", pm->name); ++ ret = -EEXIST; ++ } else { ++ list_add_tail_rcu(&pm->list, &mptcp_pm_list); ++ pr_info("%s registered\n", pm->name); ++ } ++ spin_unlock(&mptcp_pm_list_lock); ++ ++ return ret; ++} ++EXPORT_SYMBOL_GPL(mptcp_register_path_manager); ++ ++void mptcp_unregister_path_manager(struct mptcp_pm_ops *pm) ++{ ++ spin_lock(&mptcp_pm_list_lock); ++ list_del_rcu(&pm->list); ++ spin_unlock(&mptcp_pm_list_lock); ++} ++EXPORT_SYMBOL_GPL(mptcp_unregister_path_manager); ++ ++void mptcp_get_default_path_manager(char *name) ++{ ++ struct mptcp_pm_ops *pm; ++ ++ BUG_ON(list_empty(&mptcp_pm_list)); ++ ++ rcu_read_lock(); ++ pm = list_entry(mptcp_pm_list.next, struct mptcp_pm_ops, list); ++ strncpy(name, pm->name, MPTCP_PM_NAME_MAX); ++ rcu_read_unlock(); ++} ++ ++int mptcp_set_default_path_manager(const char *name) ++{ ++ struct mptcp_pm_ops *pm; ++ int ret = -ENOENT; ++ ++ spin_lock(&mptcp_pm_list_lock); ++ pm = mptcp_pm_find(name); ++#ifdef CONFIG_MODULES ++ if (!pm && capable(CAP_NET_ADMIN)) { ++ spin_unlock(&mptcp_pm_list_lock); ++ ++ request_module("mptcp_%s", name); ++ spin_lock(&mptcp_pm_list_lock); ++ pm = mptcp_pm_find(name); ++ } ++#endif ++ ++ if (pm) { ++ list_move(&pm->list, &mptcp_pm_list); ++ ret = 0; ++ } else { ++ pr_info("%s is not available\n", name); ++ } ++ spin_unlock(&mptcp_pm_list_lock); ++ ++ return ret; ++} ++ ++void mptcp_init_path_manager(struct mptcp_cb *mpcb) ++{ ++ struct mptcp_pm_ops *pm; ++ ++ rcu_read_lock(); ++ list_for_each_entry_rcu(pm, &mptcp_pm_list, list) { ++ if (try_module_get(pm->owner)) { ++ mpcb->pm_ops = pm; ++ break; ++ } ++ } ++ rcu_read_unlock(); ++} ++ ++/* Manage refcounts on socket close. */ ++void mptcp_cleanup_path_manager(struct mptcp_cb *mpcb) ++{ ++ module_put(mpcb->pm_ops->owner); ++} ++ ++/* Fallback to the default path-manager. */ ++void mptcp_fallback_default(struct mptcp_cb *mpcb) ++{ ++ struct mptcp_pm_ops *pm; ++ ++ mptcp_cleanup_path_manager(mpcb); ++ pm = mptcp_pm_find("default"); ++ ++ /* Cannot fail - it's the default module */ ++ try_module_get(pm->owner); ++ mpcb->pm_ops = pm; ++} ++EXPORT_SYMBOL_GPL(mptcp_fallback_default); ++ ++/* Set default value from kernel configuration at bootup */ ++static int __init mptcp_path_manager_default(void) ++{ ++ return mptcp_set_default_path_manager(CONFIG_DEFAULT_MPTCP_PM); ++} ++late_initcall(mptcp_path_manager_default); +diff --git a/net/mptcp/mptcp_wvegas.c b/net/mptcp/mptcp_wvegas.c +new file mode 100644 +index 0000000..8e1fd50 +--- /dev/null ++++ b/net/mptcp/mptcp_wvegas.c +@@ -0,0 +1,270 @@ ++/* ++ * MPTCP implementation - WEIGHTED VEGAS ++ * ++ * Algorithm design: ++ * Yu Cao <cyAnalyst@126.com> ++ * Mingwei Xu <xmw@csnet1.cs.tsinghua.edu.cn> ++ * Xiaoming Fu <fu@cs.uni-goettinggen.de> ++ * ++ * Implementation: ++ * Yu Cao <cyAnalyst@126.com> ++ * Enhuan Dong <deh13@mails.tsinghua.edu.cn> ++ * ++ * Ported to the official MPTCP-kernel: ++ * Christoph Paasch <christoph.paasch@uclouvain.be> ++ * ++ * 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. ++ */ ++ ++#include <linux/skbuff.h> ++#include <net/tcp.h> ++#include <net/mptcp.h> ++#include <linux/module.h> ++#include <linux/tcp.h> ++ ++static int initial_alpha = 2; ++static int total_alpha = 10; ++static int gamma = 1; ++ ++module_param(initial_alpha, int, 0644); ++MODULE_PARM_DESC(initial_alpha, "initial alpha for all subflows"); ++module_param(total_alpha, int, 0644); ++MODULE_PARM_DESC(total_alpha, "total alpha for all subflows"); ++module_param(gamma, int, 0644); ++MODULE_PARM_DESC(gamma, "limit on increase (scale by 2)"); ++ ++#define MPTCP_WVEGAS_SCALE 16 ++ ++/* wVegas variables */ ++struct wvegas { ++ u32 beg_snd_nxt; /* right edge during last RTT */ ++ u8 doing_wvegas_now;/* if true, do wvegas for this RTT */ ++ ++ u16 cnt_rtt; /* # of RTTs measured within last RTT */ ++ u32 sampled_rtt; /* cumulative RTTs measured within last RTT (in usec) */ ++ u32 base_rtt; /* the min of all wVegas RTT measurements seen (in usec) */ ++ ++ u64 instant_rate; /* cwnd / srtt_us, unit: pkts/us * 2^16 */ ++ u64 weight; /* the ratio of subflow's rate to the total rate, * 2^16 */ ++ int alpha; /* alpha for each subflows */ ++ ++ u32 queue_delay; /* queue delay*/ ++}; ++ ++ ++static inline u64 mptcp_wvegas_scale(u32 val, int scale) ++{ ++ return (u64) val << scale; ++} ++ ++static void wvegas_enable(struct sock *sk) ++{ ++ const struct tcp_sock *tp = tcp_sk(sk); ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ ++ wvegas->doing_wvegas_now = 1; ++ ++ wvegas->beg_snd_nxt = tp->snd_nxt; ++ ++ wvegas->cnt_rtt = 0; ++ wvegas->sampled_rtt = 0; ++ ++ wvegas->instant_rate = 0; ++ wvegas->alpha = initial_alpha; ++ wvegas->weight = mptcp_wvegas_scale(1, MPTCP_WVEGAS_SCALE); ++ ++ wvegas->queue_delay = 0; ++} ++ ++static inline void wvegas_disable(struct sock *sk) ++{ ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ ++ wvegas->doing_wvegas_now = 0; ++} ++ ++static void mptcp_wvegas_init(struct sock *sk) ++{ ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ ++ wvegas->base_rtt = 0x7fffffff; ++ wvegas_enable(sk); ++} ++ ++static inline u64 mptcp_wvegas_rate(u32 cwnd, u32 rtt_us) ++{ ++ return div_u64(mptcp_wvegas_scale(cwnd, MPTCP_WVEGAS_SCALE), rtt_us); ++} ++ ++static void mptcp_wvegas_pkts_acked(struct sock *sk, u32 cnt, s32 rtt_us) ++{ ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ u32 vrtt; ++ ++ if (rtt_us < 0) ++ return; ++ ++ vrtt = rtt_us + 1; ++ ++ if (vrtt < wvegas->base_rtt) ++ wvegas->base_rtt = vrtt; ++ ++ wvegas->sampled_rtt += vrtt; ++ wvegas->cnt_rtt++; ++} ++ ++static void mptcp_wvegas_state(struct sock *sk, u8 ca_state) ++{ ++ if (ca_state == TCP_CA_Open) ++ wvegas_enable(sk); ++ else ++ wvegas_disable(sk); ++} ++ ++static void mptcp_wvegas_cwnd_event(struct sock *sk, enum tcp_ca_event event) ++{ ++ if (event == CA_EVENT_CWND_RESTART) { ++ mptcp_wvegas_init(sk); ++ } else if (event == CA_EVENT_LOSS) { ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ wvegas->instant_rate = 0; ++ } ++} ++ ++static inline u32 mptcp_wvegas_ssthresh(struct tcp_sock *tp) ++{ ++ return min(tp->snd_ssthresh, tp->snd_cwnd - 1); ++} ++ ++static u64 mptcp_wvegas_weight(struct mptcp_cb *mpcb, struct sock *sk) ++{ ++ u64 total_rate = 0; ++ struct sock *sub_sk; ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ ++ if (!mpcb) ++ return wvegas->weight; ++ ++ ++ mptcp_for_each_sk(mpcb, sub_sk) { ++ struct wvegas *sub_wvegas = inet_csk_ca(sub_sk); ++ ++ /* sampled_rtt is initialized by 0 */ ++ if (mptcp_sk_can_send(sub_sk) && (sub_wvegas->sampled_rtt > 0)) ++ total_rate += sub_wvegas->instant_rate; ++ } ++ ++ if (total_rate && wvegas->instant_rate) ++ return div64_u64(mptcp_wvegas_scale(wvegas->instant_rate, MPTCP_WVEGAS_SCALE), total_rate); ++ else ++ return wvegas->weight; ++} ++ ++static void mptcp_wvegas_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight) ++{ ++ struct tcp_sock *tp = tcp_sk(sk); ++ struct wvegas *wvegas = inet_csk_ca(sk); ++ ++ if (!wvegas->doing_wvegas_now) { ++ tcp_reno_cong_avoid(sk, ack, acked, in_flight); ++ return; ++ } ++ ++ if (after(ack, wvegas->beg_snd_nxt)) { ++ wvegas->beg_snd_nxt = tp->snd_nxt; ++ ++ if (wvegas->cnt_rtt <= 2) { ++ tcp_reno_cong_avoid(sk, ack, acked, in_flight); ++ } else { ++ u32 rtt, diff, q_delay; ++ u64 target_cwnd; ++ ++ rtt = wvegas->sampled_rtt / wvegas->cnt_rtt; ++ target_cwnd = div_u64(((u64)tp->snd_cwnd * wvegas->base_rtt), rtt); ++ ++ diff = div_u64((u64)tp->snd_cwnd * (rtt - wvegas->base_rtt), rtt); ++ ++ if (diff > gamma && tp->snd_cwnd <= tp->snd_ssthresh) { ++ tp->snd_cwnd = min(tp->snd_cwnd, (u32)target_cwnd+1); ++ tp->snd_ssthresh = mptcp_wvegas_ssthresh(tp); ++ ++ } else if (tp->snd_cwnd <= tp->snd_ssthresh) { ++ tcp_slow_start(tp, acked); ++ } else { ++ if (diff >= wvegas->alpha) { ++ wvegas->instant_rate = mptcp_wvegas_rate(tp->snd_cwnd, rtt); ++ wvegas->weight = mptcp_wvegas_weight(tp->mpcb, sk); ++ wvegas->alpha = max(2U, (u32)((wvegas->weight * total_alpha) >> MPTCP_WVEGAS_SCALE)); ++ } ++ if (diff > wvegas->alpha) { ++ tp->snd_cwnd--; ++ tp->snd_ssthresh = mptcp_wvegas_ssthresh(tp); ++ } else if (diff < wvegas->alpha) { ++ tp->snd_cwnd++; ++ } ++ ++ /* Try to drain link queue if needed*/ ++ q_delay = rtt - wvegas->base_rtt; ++ if ((wvegas->queue_delay == 0) || (wvegas->queue_delay > q_delay)) ++ wvegas->queue_delay = q_delay; ++ ++ if (q_delay >= 2 * wvegas->queue_delay) { ++ u32 backoff_factor = div_u64(mptcp_wvegas_scale(wvegas->base_rtt, MPTCP_WVEGAS_SCALE), 2 * rtt); ++ tp->snd_cwnd = ((u64)tp->snd_cwnd * backoff_factor) >> MPTCP_WVEGAS_SCALE; ++ wvegas->queue_delay = 0; ++ } ++ } ++ ++ if (tp->snd_cwnd < 2) ++ tp->snd_cwnd = 2; ++ else if (tp->snd_cwnd > tp->snd_cwnd_clamp) ++ tp->snd_cwnd = tp->snd_cwnd_clamp; ++ ++ tp->snd_ssthresh = tcp_current_ssthresh(sk); ++ } ++ ++ wvegas->cnt_rtt = 0; ++ wvegas->sampled_rtt = 0; ++ } ++ /* Use normal slow start */ ++ else if (tp->snd_cwnd <= tp->snd_ssthresh) ++ tcp_slow_start(tp, acked); ++} ++ ++ ++static struct tcp_congestion_ops mptcp_wvegas __read_mostly = { ++ .flags = TCP_CONG_RTT_STAMP, ++ .init = mptcp_wvegas_init, ++ .ssthresh = tcp_reno_ssthresh, ++ .cong_avoid = mptcp_wvegas_cong_avoid, ++ .min_cwnd = tcp_reno_min_cwnd, ++ .pkts_acked = mptcp_wvegas_pkts_acked, ++ .set_state = mptcp_wvegas_state, ++ .cwnd_event = mptcp_wvegas_cwnd_event, ++ ++ .owner = THIS_MODULE, ++ .name = "wvegas", ++}; ++ ++static int __init mptcp_wvegas_register(void) ++{ ++ BUILD_BUG_ON(sizeof(struct wvegas) > ICSK_CA_PRIV_SIZE); ++ tcp_register_congestion_control(&mptcp_wvegas); ++ return 0; ++} ++ ++static void __exit mptcp_wvegas_unregister(void) ++{ ++ tcp_unregister_congestion_control(&mptcp_wvegas); ++} ++ ++module_init(mptcp_wvegas_register); ++module_exit(mptcp_wvegas_unregister); ++ ++MODULE_AUTHOR("Yu Cao, Enhuan Dong"); ++MODULE_LICENSE("GPL"); ++MODULE_DESCRIPTION("MPTCP wVegas"); ++MODULE_VERSION("0.1"); diff --git a/target/linux/patches/3.14.40/mtd-rootfs.patch b/target/linux/patches/3.14.40/mtd-rootfs.patch new file mode 100644 index 000000000..775d5fc80 --- /dev/null +++ b/target/linux/patches/3.14.40/mtd-rootfs.patch @@ -0,0 +1,26 @@ +diff -Nur linux-3.5.orig//drivers/mtd/mtdpart.c linux-3.5/drivers/mtd/mtdpart.c +--- linux-3.5.orig//drivers/mtd/mtdpart.c 2012-07-21 22:58:29.000000000 +0200 ++++ linux-3.5/drivers/mtd/mtdpart.c 2012-07-31 23:59:07.000000000 +0200 +@@ -30,6 +30,7 @@ + #include <linux/mtd/mtd.h> + #include <linux/mtd/partitions.h> + #include <linux/err.h> ++#include <linux/root_dev.h> + + #include "mtdcore.h" + +@@ -637,6 +638,14 @@ + if (IS_ERR(slave)) + return PTR_ERR(slave); + ++ if (strcmp(parts[i].name, "rootfs") == 0) { ++ if (ROOT_DEV == 0) { ++ printk(KERN_NOTICE "mtd: partition \"rootfs\" " ++ "set to be root filesystem\n"); ++ ROOT_DEV = MKDEV(MTD_BLOCK_MAJOR, i); ++ } ++ } ++ + mutex_lock(&mtd_partitions_mutex); + list_add(&slave->list, &mtd_partitions); + mutex_unlock(&mtd_partitions_mutex); diff --git a/target/linux/patches/3.14.40/nfsv3-tcp.patch b/target/linux/patches/3.14.40/nfsv3-tcp.patch new file mode 100644 index 000000000..68ed95b08 --- /dev/null +++ b/target/linux/patches/3.14.40/nfsv3-tcp.patch @@ -0,0 +1,12 @@ +diff -Nur linux-3.13.11.orig/fs/nfs/nfsroot.c linux-3.13.11/fs/nfs/nfsroot.c +--- linux-3.13.11.orig/fs/nfs/nfsroot.c 2014-04-23 01:49:33.000000000 +0200 ++++ linux-3.13.11/fs/nfs/nfsroot.c 2014-05-13 16:09:25.000000000 +0200 +@@ -87,7 +87,7 @@ + #define NFS_ROOT "/tftpboot/%s" + + /* Default NFSROOT mount options. */ +-#define NFS_DEF_OPTIONS "vers=2,udp,rsize=4096,wsize=4096" ++#define NFS_DEF_OPTIONS "nfsvers=3,proto=tcp,rsize=4096,wsize=4096" + + /* Parameters passed from the kernel command line */ + static char nfs_root_parms[256] __initdata = ""; diff --git a/target/linux/patches/3.14.40/non-static.patch b/target/linux/patches/3.14.40/non-static.patch new file mode 100644 index 000000000..a967703d0 --- /dev/null +++ b/target/linux/patches/3.14.40/non-static.patch @@ -0,0 +1,33 @@ +diff -Nur linux-2.6.39-rc6.orig/fs/namei.c linux-2.6.39-rc6/fs/namei.c +--- linux-2.6.39-rc6.orig/fs/namei.c 2011-05-04 04:59:13.000000000 +0200 ++++ linux-2.6.39-rc6/fs/namei.c 2011-05-05 11:30:14.000000000 +0200 +@@ -1769,7 +1769,7 @@ + * needs parent already locked. Doesn't follow mounts. + * SMP-safe. + */ +-static struct dentry *lookup_hash(struct nameidata *nd) ++struct dentry *lookup_hash(struct nameidata *nd) + { + return __lookup_hash(&nd->last, nd->path.dentry, nd); + } +diff -Nur linux-2.6.39-rc6.orig/fs/splice.c linux-2.6.39-rc6/fs/splice.c +--- linux-2.6.39-rc6.orig/fs/splice.c 2011-05-04 04:59:13.000000000 +0200 ++++ linux-2.6.39-rc6/fs/splice.c 2011-05-05 11:31:04.000000000 +0200 +@@ -1081,7 +1081,7 @@ + /* + * Attempt to initiate a splice from pipe to file. + */ +-static long do_splice_from(struct pipe_inode_info *pipe, struct file *out, ++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 *, +@@ -1109,7 +1109,7 @@ + /* + * Attempt to initiate a splice from a file to a pipe. + */ +-static long do_splice_to(struct file *in, loff_t *ppos, ++long do_splice_to(struct file *in, loff_t *ppos, + struct pipe_inode_info *pipe, size_t len, + unsigned int flags) + { diff --git a/target/linux/patches/3.14.40/ppc64-missing-zlib.patch b/target/linux/patches/3.14.40/ppc64-missing-zlib.patch new file mode 100644 index 000000000..c6e0616be --- /dev/null +++ b/target/linux/patches/3.14.40/ppc64-missing-zlib.patch @@ -0,0 +1,11 @@ +diff -Nur linux-3.11.5.orig/arch/powerpc/platforms/pseries/Kconfig linux-3.11.5/arch/powerpc/platforms/pseries/Kconfig +--- linux-3.11.5.orig/arch/powerpc/platforms/pseries/Kconfig 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/arch/powerpc/platforms/pseries/Kconfig 2013-11-01 15:23:09.000000000 +0100 +@@ -17,6 +17,7 @@ + select PPC_NATIVE + select PPC_PCI_CHOICE if EXPERT + select ZLIB_DEFLATE ++ select ZLIB_INFLATE + select PPC_DOORBELL + select HAVE_CONTEXT_TRACKING + select HOTPLUG_CPU if SMP diff --git a/target/linux/patches/3.14.40/regmap-boolean.patch b/target/linux/patches/3.14.40/regmap-boolean.patch new file mode 100644 index 000000000..d73620e4d --- /dev/null +++ b/target/linux/patches/3.14.40/regmap-boolean.patch @@ -0,0 +1,24 @@ +diff -Nur linux-3.14.4.orig/drivers/base/regmap/Kconfig linux-3.14.4/drivers/base/regmap/Kconfig +--- linux-3.14.4.orig/drivers/base/regmap/Kconfig 2014-05-13 13:33:14.000000000 +0200 ++++ linux-3.14.4/drivers/base/regmap/Kconfig 2014-05-21 13:46:48.928985115 +0200 +@@ -10,16 +10,16 @@ + bool + + config REGMAP_I2C +- tristate ++ boolean + + config REGMAP_SPI +- tristate ++ boolean + + config REGMAP_SPMI +- tristate ++ boolean + + config REGMAP_MMIO +- tristate ++ boolean + + config REGMAP_IRQ + bool diff --git a/target/linux/patches/3.14.40/relocs.patch b/target/linux/patches/3.14.40/relocs.patch new file mode 100644 index 000000000..69a7c88a9 --- /dev/null +++ b/target/linux/patches/3.14.40/relocs.patch @@ -0,0 +1,2709 @@ +diff -Nur linux-3.13.6.orig/arch/x86/tools/relocs.c linux-3.13.6/arch/x86/tools/relocs.c +--- linux-3.13.6.orig/arch/x86/tools/relocs.c 2014-03-07 07:07:02.000000000 +0100 ++++ linux-3.13.6/arch/x86/tools/relocs.c 2014-03-15 19:39:45.000000000 +0100 +@@ -126,6 +126,7 @@ + + if (err) { + regerror(err, &sym_regex_c[i], errbuf, sizeof errbuf); ++ printf("foo: %s\n", sym_regex[i]); + die("%s", errbuf); + } + } +diff -Nur linux-3.13.6.orig/arch/x86/tools/relocs.h linux-3.13.6/arch/x86/tools/relocs.h +--- linux-3.13.6.orig/arch/x86/tools/relocs.h 2014-03-07 07:07:02.000000000 +0100 ++++ linux-3.13.6/arch/x86/tools/relocs.h 2014-03-15 18:48:40.000000000 +0100 +@@ -9,11 +9,19 @@ + #include <string.h> + #include <errno.h> + #include <unistd.h> ++#ifdef __linux__ + #include <elf.h> + #include <byteswap.h> + #define USE_BSD + #include <endian.h> ++#else ++#include "elf.h" ++#endif ++#ifdef __APPLE__ ++#include <pcreposix.h> ++#else + #include <regex.h> ++#endif + #include <tools/le_byteshift.h> + + void die(char *fmt, ...); +diff -Nur linux-3.13.6.orig/tools/include/elf.h linux-3.13.6/tools/include/elf.h +--- linux-3.13.6.orig/tools/include/elf.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.13.6/tools/include/elf.h 2014-03-15 18:47:36.000000000 +0100 +@@ -0,0 +1,2671 @@ ++#ifndef _ELF_H ++#define _ELF_H ++ ++#ifdef __cplusplus ++extern "C" { ++#endif ++ ++#include <stdint.h> ++ ++typedef uint16_t Elf32_Half; ++typedef uint16_t Elf64_Half; ++ ++typedef uint32_t Elf32_Word; ++typedef int32_t Elf32_Sword; ++typedef uint32_t Elf64_Word; ++typedef int32_t Elf64_Sword; ++ ++typedef uint64_t Elf32_Xword; ++typedef int64_t Elf32_Sxword; ++typedef uint64_t Elf64_Xword; ++typedef int64_t Elf64_Sxword; ++ ++typedef uint32_t Elf32_Addr; ++typedef uint64_t Elf64_Addr; ++ ++typedef uint32_t Elf32_Off; ++typedef uint64_t Elf64_Off; ++ ++typedef uint16_t Elf32_Section; ++typedef uint16_t Elf64_Section; ++ ++typedef Elf32_Half Elf32_Versym; ++typedef Elf64_Half Elf64_Versym; ++ ++#define EI_NIDENT (16) ++ ++typedef struct { ++ unsigned char e_ident[EI_NIDENT]; ++ Elf32_Half e_type; ++ Elf32_Half e_machine; ++ Elf32_Word e_version; ++ Elf32_Addr e_entry; ++ Elf32_Off e_phoff; ++ Elf32_Off e_shoff; ++ Elf32_Word e_flags; ++ Elf32_Half e_ehsize; ++ Elf32_Half e_phentsize; ++ Elf32_Half e_phnum; ++ Elf32_Half e_shentsize; ++ Elf32_Half e_shnum; ++ Elf32_Half e_shstrndx; ++} Elf32_Ehdr; ++ ++typedef struct { ++ unsigned char e_ident[EI_NIDENT]; ++ Elf64_Half e_type; ++ Elf64_Half e_machine; ++ Elf64_Word e_version; ++ Elf64_Addr e_entry; ++ Elf64_Off e_phoff; ++ Elf64_Off e_shoff; ++ Elf64_Word e_flags; ++ Elf64_Half e_ehsize; ++ Elf64_Half e_phentsize; ++ Elf64_Half e_phnum; ++ Elf64_Half e_shentsize; ++ Elf64_Half e_shnum; ++ Elf64_Half e_shstrndx; ++} Elf64_Ehdr; ++ ++#define EI_MAG0 0 ++#define ELFMAG0 0x7f ++ ++#define EI_MAG1 1 ++#define ELFMAG1 'E' ++ ++#define EI_MAG2 2 ++#define ELFMAG2 'L' ++ ++#define EI_MAG3 3 ++#define ELFMAG3 'F' ++ ++ ++#define ELFMAG "\177ELF" ++#define SELFMAG 4 ++ ++#define EI_CLASS 4 ++#define ELFCLASSNONE 0 ++#define ELFCLASS32 1 ++#define ELFCLASS64 2 ++#define ELFCLASSNUM 3 ++ ++#define EI_DATA 5 ++#define ELFDATANONE 0 ++#define ELFDATA2LSB 1 ++#define ELFDATA2MSB 2 ++#define ELFDATANUM 3 ++ ++#define EI_VERSION 6 ++ ++ ++#define EI_OSABI 7 ++#define ELFOSABI_NONE 0 ++#define ELFOSABI_SYSV 0 ++#define ELFOSABI_HPUX 1 ++#define ELFOSABI_NETBSD 2 ++#define ELFOSABI_LINUX 3 ++#define ELFOSABI_GNU 3 ++#define ELFOSABI_SOLARIS 6 ++#define ELFOSABI_AIX 7 ++#define ELFOSABI_IRIX 8 ++#define ELFOSABI_FREEBSD 9 ++#define ELFOSABI_TRU64 10 ++#define ELFOSABI_MODESTO 11 ++#define ELFOSABI_OPENBSD 12 ++#define ELFOSABI_ARM 97 ++#define ELFOSABI_STANDALONE 255 ++ ++#define EI_ABIVERSION 8 ++ ++#define EI_PAD 9 ++ ++ ++ ++#define ET_NONE 0 ++#define ET_REL 1 ++#define ET_EXEC 2 ++#define ET_DYN 3 ++#define ET_CORE 4 ++#define ET_NUM 5 ++#define ET_LOOS 0xfe00 ++#define ET_HIOS 0xfeff ++#define ET_LOPROC 0xff00 ++#define ET_HIPROC 0xffff ++ ++ ++ ++#define EM_NONE 0 ++#define EM_M32 1 ++#define EM_SPARC 2 ++#define EM_386 3 ++#define EM_68K 4 ++#define EM_88K 5 ++#define EM_860 7 ++#define EM_MIPS 8 ++#define EM_S370 9 ++#define EM_MIPS_RS3_LE 10 ++ ++#define EM_PARISC 15 ++#define EM_VPP500 17 ++#define EM_SPARC32PLUS 18 ++#define EM_960 19 ++#define EM_PPC 20 ++#define EM_PPC64 21 ++#define EM_S390 22 ++ ++#define EM_V800 36 ++#define EM_FR20 37 ++#define EM_RH32 38 ++#define EM_RCE 39 ++#define EM_ARM 40 ++#define EM_FAKE_ALPHA 41 ++#define EM_SH 42 ++#define EM_SPARCV9 43 ++#define EM_TRICORE 44 ++#define EM_ARC 45 ++#define EM_H8_300 46 ++#define EM_H8_300H 47 ++#define EM_H8S 48 ++#define EM_H8_500 49 ++#define EM_IA_64 50 ++#define EM_MIPS_X 51 ++#define EM_COLDFIRE 52 ++#define EM_68HC12 53 ++#define EM_MMA 54 ++#define EM_PCP 55 ++#define EM_NCPU 56 ++#define EM_NDR1 57 ++#define EM_STARCORE 58 ++#define EM_ME16 59 ++#define EM_ST100 60 ++#define EM_TINYJ 61 ++#define EM_X86_64 62 ++#define EM_PDSP 63 ++ ++#define EM_FX66 66 ++#define EM_ST9PLUS 67 ++#define EM_ST7 68 ++#define EM_68HC16 69 ++#define EM_68HC11 70 ++#define EM_68HC08 71 ++#define EM_68HC05 72 ++#define EM_SVX 73 ++#define EM_ST19 74 ++#define EM_VAX 75 ++#define EM_CRIS 76 ++#define EM_JAVELIN 77 ++#define EM_FIREPATH 78 ++#define EM_ZSP 79 ++#define EM_MMIX 80 ++#define EM_HUANY 81 ++#define EM_PRISM 82 ++#define EM_AVR 83 ++#define EM_FR30 84 ++#define EM_D10V 85 ++#define EM_D30V 86 ++#define EM_V850 87 ++#define EM_M32R 88 ++#define EM_MN10300 89 ++#define EM_MN10200 90 ++#define EM_PJ 91 ++#define EM_OPENRISC 92 ++#define EM_ARC_A5 93 ++#define EM_XTENSA 94 ++#define EM_AARCH64 183 ++#define EM_TILEPRO 188 ++#define EM_MICROBLAZE 189 ++#define EM_TILEGX 191 ++#define EM_NUM 192 ++#define EM_ALPHA 0x9026 ++ ++#define EV_NONE 0 ++#define EV_CURRENT 1 ++#define EV_NUM 2 ++ ++typedef struct { ++ Elf32_Word sh_name; ++ Elf32_Word sh_type; ++ Elf32_Word sh_flags; ++ Elf32_Addr sh_addr; ++ Elf32_Off sh_offset; ++ Elf32_Word sh_size; ++ Elf32_Word sh_link; ++ Elf32_Word sh_info; ++ Elf32_Word sh_addralign; ++ Elf32_Word sh_entsize; ++} Elf32_Shdr; ++ ++typedef struct { ++ Elf64_Word sh_name; ++ Elf64_Word sh_type; ++ Elf64_Xword sh_flags; ++ Elf64_Addr sh_addr; ++ Elf64_Off sh_offset; ++ Elf64_Xword sh_size; ++ Elf64_Word sh_link; ++ Elf64_Word sh_info; ++ Elf64_Xword sh_addralign; ++ Elf64_Xword sh_entsize; ++} Elf64_Shdr; ++ ++ ++ ++#define SHN_UNDEF 0 ++#define SHN_LORESERVE 0xff00 ++#define SHN_LOPROC 0xff00 ++#define SHN_BEFORE 0xff00 ++ ++#define SHN_AFTER 0xff01 ++ ++#define SHN_HIPROC 0xff1f ++#define SHN_LOOS 0xff20 ++#define SHN_HIOS 0xff3f ++#define SHN_ABS 0xfff1 ++#define SHN_COMMON 0xfff2 ++#define SHN_XINDEX 0xffff ++#define SHN_HIRESERVE 0xffff ++ ++ ++ ++#define SHT_NULL 0 ++#define SHT_PROGBITS 1 ++#define SHT_SYMTAB 2 ++#define SHT_STRTAB 3 ++#define SHT_RELA 4 ++#define SHT_HASH 5 ++#define SHT_DYNAMIC 6 ++#define SHT_NOTE 7 ++#define SHT_NOBITS 8 ++#define SHT_REL 9 ++#define SHT_SHLIB 10 ++#define SHT_DYNSYM 11 ++#define SHT_INIT_ARRAY 14 ++#define SHT_FINI_ARRAY 15 ++#define SHT_PREINIT_ARRAY 16 ++#define SHT_GROUP 17 ++#define SHT_SYMTAB_SHNDX 18 ++#define SHT_NUM 19 ++#define SHT_LOOS 0x60000000 ++#define SHT_GNU_ATTRIBUTES 0x6ffffff5 ++#define SHT_GNU_HASH 0x6ffffff6 ++#define SHT_GNU_LIBLIST 0x6ffffff7 ++#define SHT_CHECKSUM 0x6ffffff8 ++#define SHT_LOSUNW 0x6ffffffa ++#define SHT_SUNW_move 0x6ffffffa ++#define SHT_SUNW_COMDAT 0x6ffffffb ++#define SHT_SUNW_syminfo 0x6ffffffc ++#define SHT_GNU_verdef 0x6ffffffd ++#define SHT_GNU_verneed 0x6ffffffe ++#define SHT_GNU_versym 0x6fffffff ++#define SHT_HISUNW 0x6fffffff ++#define SHT_HIOS 0x6fffffff ++#define SHT_LOPROC 0x70000000 ++#define SHT_HIPROC 0x7fffffff ++#define SHT_LOUSER 0x80000000 ++#define SHT_HIUSER 0x8fffffff ++ ++#define SHF_WRITE (1 << 0) ++#define SHF_ALLOC (1 << 1) ++#define SHF_EXECINSTR (1 << 2) ++#define SHF_MERGE (1 << 4) ++#define SHF_STRINGS (1 << 5) ++#define SHF_INFO_LINK (1 << 6) ++#define SHF_LINK_ORDER (1 << 7) ++#define SHF_OS_NONCONFORMING (1 << 8) ++ ++#define SHF_GROUP (1 << 9) ++#define SHF_TLS (1 << 10) ++#define SHF_MASKOS 0x0ff00000 ++#define SHF_MASKPROC 0xf0000000 ++#define SHF_ORDERED (1 << 30) ++#define SHF_EXCLUDE (1 << 31) ++ ++#define GRP_COMDAT 0x1 ++ ++typedef struct { ++ Elf32_Word st_name; ++ Elf32_Addr st_value; ++ Elf32_Word st_size; ++ unsigned char st_info; ++ unsigned char st_other; ++ Elf32_Section st_shndx; ++} Elf32_Sym; ++ ++typedef struct { ++ Elf64_Word st_name; ++ unsigned char st_info; ++ unsigned char st_other; ++ Elf64_Section st_shndx; ++ Elf64_Addr st_value; ++ Elf64_Xword st_size; ++} Elf64_Sym; ++ ++typedef struct { ++ Elf32_Half si_boundto; ++ Elf32_Half si_flags; ++} Elf32_Syminfo; ++ ++typedef struct { ++ Elf64_Half si_boundto; ++ Elf64_Half si_flags; ++} Elf64_Syminfo; ++ ++#define SYMINFO_BT_SELF 0xffff ++#define SYMINFO_BT_PARENT 0xfffe ++#define SYMINFO_BT_LOWRESERVE 0xff00 ++ ++#define SYMINFO_FLG_DIRECT 0x0001 ++#define SYMINFO_FLG_PASSTHRU 0x0002 ++#define SYMINFO_FLG_COPY 0x0004 ++#define SYMINFO_FLG_LAZYLOAD 0x0008 ++ ++#define SYMINFO_NONE 0 ++#define SYMINFO_CURRENT 1 ++#define SYMINFO_NUM 2 ++ ++#define ELF32_ST_BIND(val) (((unsigned char) (val)) >> 4) ++#define ELF32_ST_TYPE(val) ((val) & 0xf) ++#define ELF32_ST_INFO(bind, type) (((bind) << 4) + ((type) & 0xf)) ++ ++#define ELF64_ST_BIND(val) ELF32_ST_BIND (val) ++#define ELF64_ST_TYPE(val) ELF32_ST_TYPE (val) ++#define ELF64_ST_INFO(bind, type) ELF32_ST_INFO ((bind), (type)) ++ ++#define STB_LOCAL 0 ++#define STB_GLOBAL 1 ++#define STB_WEAK 2 ++#define STB_NUM 3 ++#define STB_LOOS 10 ++#define STB_GNU_UNIQUE 10 ++#define STB_HIOS 12 ++#define STB_LOPROC 13 ++#define STB_HIPROC 15 ++ ++#define STT_NOTYPE 0 ++#define STT_OBJECT 1 ++#define STT_FUNC 2 ++#define STT_SECTION 3 ++#define STT_FILE 4 ++#define STT_COMMON 5 ++#define STT_TLS 6 ++#define STT_NUM 7 ++#define STT_LOOS 10 ++#define STT_GNU_IFUNC 10 ++#define STT_HIOS 12 ++#define STT_LOPROC 13 ++#define STT_HIPROC 15 ++ ++#define STN_UNDEF 0 ++ ++#define ELF32_ST_VISIBILITY(o) ((o) & 0x03) ++#define ELF64_ST_VISIBILITY(o) ELF32_ST_VISIBILITY (o) ++ ++#define STV_DEFAULT 0 ++#define STV_INTERNAL 1 ++#define STV_HIDDEN 2 ++#define STV_PROTECTED 3 ++ ++ ++ ++ ++typedef struct ++{ ++ Elf32_Addr r_offset; ++ Elf32_Word r_info; ++} Elf32_Rel; ++ ++typedef struct { ++ Elf64_Addr r_offset; ++ Elf64_Xword r_info; ++} Elf64_Rel; ++ ++ ++ ++typedef struct { ++ Elf32_Addr r_offset; ++ Elf32_Word r_info; ++ Elf32_Sword r_addend; ++} Elf32_Rela; ++ ++typedef struct { ++ Elf64_Addr r_offset; ++ Elf64_Xword r_info; ++ Elf64_Sxword r_addend; ++} Elf64_Rela; ++ ++ ++ ++#define ELF32_R_SYM(val) ((val) >> 8) ++#define ELF32_R_TYPE(val) ((val) & 0xff) ++#define ELF32_R_INFO(sym, type) (((sym) << 8) + ((type) & 0xff)) ++ ++#define ELF64_R_SYM(i) ((i) >> 32) ++#define ELF64_R_TYPE(i) ((i) & 0xffffffff) ++#define ELF64_R_INFO(sym,type) ((((Elf64_Xword) (sym)) << 32) + (type)) ++ ++ ++ ++typedef struct { ++ Elf32_Word p_type; ++ Elf32_Off p_offset; ++ Elf32_Addr p_vaddr; ++ Elf32_Addr p_paddr; ++ Elf32_Word p_filesz; ++ Elf32_Word p_memsz; ++ Elf32_Word p_flags; ++ Elf32_Word p_align; ++} Elf32_Phdr; ++ ++typedef struct { ++ Elf64_Word p_type; ++ Elf64_Word p_flags; ++ Elf64_Off p_offset; ++ Elf64_Addr p_vaddr; ++ Elf64_Addr p_paddr; ++ Elf64_Xword p_filesz; ++ Elf64_Xword p_memsz; ++ Elf64_Xword p_align; ++} Elf64_Phdr; ++ ++ ++ ++#define PT_NULL 0 ++#define PT_LOAD 1 ++#define PT_DYNAMIC 2 ++#define PT_INTERP 3 ++#define PT_NOTE 4 ++#define PT_SHLIB 5 ++#define PT_PHDR 6 ++#define PT_TLS 7 ++#define PT_NUM 8 ++#define PT_LOOS 0x60000000 ++#define PT_GNU_EH_FRAME 0x6474e550 ++#define PT_GNU_STACK 0x6474e551 ++#define PT_GNU_RELRO 0x6474e552 ++#define PT_LOSUNW 0x6ffffffa ++#define PT_SUNWBSS 0x6ffffffa ++#define PT_SUNWSTACK 0x6ffffffb ++#define PT_HISUNW 0x6fffffff ++#define PT_HIOS 0x6fffffff ++#define PT_LOPROC 0x70000000 ++#define PT_HIPROC 0x7fffffff ++ ++ ++#define PN_XNUM 0xffff ++ ++ ++#define PF_X (1 << 0) ++#define PF_W (1 << 1) ++#define PF_R (1 << 2) ++#define PF_MASKOS 0x0ff00000 ++#define PF_MASKPROC 0xf0000000 ++ ++ ++ ++#define NT_PRSTATUS 1 ++#define NT_FPREGSET 2 ++#define NT_PRPSINFO 3 ++#define NT_PRXREG 4 ++#define NT_TASKSTRUCT 4 ++#define NT_PLATFORM 5 ++#define NT_AUXV 6 ++#define NT_GWINDOWS 7 ++#define NT_ASRS 8 ++#define NT_PSTATUS 10 ++#define NT_PSINFO 13 ++#define NT_PRCRED 14 ++#define NT_UTSNAME 15 ++#define NT_LWPSTATUS 16 ++#define NT_LWPSINFO 17 ++#define NT_PRFPXREG 20 ++#define NT_SIGINFO 0x53494749 ++#define NT_FILE 0x46494c45 ++#define NT_PRXFPREG 0x46e62b7f ++#define NT_PPC_VMX 0x100 ++#define NT_PPC_SPE 0x101 ++#define NT_PPC_VSX 0x102 ++#define NT_386_TLS 0x200 ++#define NT_386_IOPERM 0x201 ++#define NT_X86_XSTATE 0x202 ++#define NT_S390_HIGH_GPRS 0x300 ++#define NT_S390_TIMER 0x301 ++#define NT_S390_TODCMP 0x302 ++#define NT_S390_TODPREG 0x303 ++#define NT_S390_CTRS 0x304 ++#define NT_S390_PREFIX 0x305 ++#define NT_S390_LAST_BREAK 0x306 ++#define NT_S390_SYSTEM_CALL 0x307 ++#define NT_S390_TDB 0x308 ++#define NT_ARM_VFP 0x400 ++#define NT_ARM_TLS 0x401 ++#define NT_ARM_HW_BREAK 0x402 ++#define NT_ARM_HW_WATCH 0x403 ++#define NT_METAG_CBUF 0x500 ++#define NT_METAG_RPIPE 0x501 ++#define NT_METAG_TLS 0x502 ++#define NT_VERSION 1 ++ ++ ++ ++ ++typedef struct { ++ Elf32_Sword d_tag; ++ union { ++ Elf32_Word d_val; ++ Elf32_Addr d_ptr; ++ } d_un; ++} Elf32_Dyn; ++ ++typedef struct { ++ Elf64_Sxword d_tag; ++ union { ++ Elf64_Xword d_val; ++ Elf64_Addr d_ptr; ++ } d_un; ++} Elf64_Dyn; ++ ++ ++ ++#define DT_NULL 0 ++#define DT_NEEDED 1 ++#define DT_PLTRELSZ 2 ++#define DT_PLTGOT 3 ++#define DT_HASH 4 ++#define DT_STRTAB 5 ++#define DT_SYMTAB 6 ++#define DT_RELA 7 ++#define DT_RELASZ 8 ++#define DT_RELAENT 9 ++#define DT_STRSZ 10 ++#define DT_SYMENT 11 ++#define DT_INIT 12 ++#define DT_FINI 13 ++#define DT_SONAME 14 ++#define DT_RPATH 15 ++#define DT_SYMBOLIC 16 ++#define DT_REL 17 ++#define DT_RELSZ 18 ++#define DT_RELENT 19 ++#define DT_PLTREL 20 ++#define DT_DEBUG 21 ++#define DT_TEXTREL 22 ++#define DT_JMPREL 23 ++#define DT_BIND_NOW 24 ++#define DT_INIT_ARRAY 25 ++#define DT_FINI_ARRAY 26 ++#define DT_INIT_ARRAYSZ 27 ++#define DT_FINI_ARRAYSZ 28 ++#define DT_RUNPATH 29 ++#define DT_FLAGS 30 ++#define DT_ENCODING 32 ++#define DT_PREINIT_ARRAY 32 ++#define DT_PREINIT_ARRAYSZ 33 ++#define DT_NUM 34 ++#define DT_LOOS 0x6000000d ++#define DT_HIOS 0x6ffff000 ++#define DT_LOPROC 0x70000000 ++#define DT_HIPROC 0x7fffffff ++#define DT_PROCNUM DT_MIPS_NUM ++ ++#define DT_VALRNGLO 0x6ffffd00 ++#define DT_GNU_PRELINKED 0x6ffffdf5 ++#define DT_GNU_CONFLICTSZ 0x6ffffdf6 ++#define DT_GNU_LIBLISTSZ 0x6ffffdf7 ++#define DT_CHECKSUM 0x6ffffdf8 ++#define DT_PLTPADSZ 0x6ffffdf9 ++#define DT_MOVEENT 0x6ffffdfa ++#define DT_MOVESZ 0x6ffffdfb ++#define DT_FEATURE_1 0x6ffffdfc ++#define DT_POSFLAG_1 0x6ffffdfd ++ ++#define DT_SYMINSZ 0x6ffffdfe ++#define DT_SYMINENT 0x6ffffdff ++#define DT_VALRNGHI 0x6ffffdff ++#define DT_VALTAGIDX(tag) (DT_VALRNGHI - (tag)) ++#define DT_VALNUM 12 ++ ++#define DT_ADDRRNGLO 0x6ffffe00 ++#define DT_GNU_HASH 0x6ffffef5 ++#define DT_TLSDESC_PLT 0x6ffffef6 ++#define DT_TLSDESC_GOT 0x6ffffef7 ++#define DT_GNU_CONFLICT 0x6ffffef8 ++#define DT_GNU_LIBLIST 0x6ffffef9 ++#define DT_CONFIG 0x6ffffefa ++#define DT_DEPAUDIT 0x6ffffefb ++#define DT_AUDIT 0x6ffffefc ++#define DT_PLTPAD 0x6ffffefd ++#define DT_MOVETAB 0x6ffffefe ++#define DT_SYMINFO 0x6ffffeff ++#define DT_ADDRRNGHI 0x6ffffeff ++#define DT_ADDRTAGIDX(tag) (DT_ADDRRNGHI - (tag)) ++#define DT_ADDRNUM 11 ++ ++ ++ ++#define DT_VERSYM 0x6ffffff0 ++ ++#define DT_RELACOUNT 0x6ffffff9 ++#define DT_RELCOUNT 0x6ffffffa ++ ++ ++#define DT_FLAGS_1 0x6ffffffb ++#define DT_VERDEF 0x6ffffffc ++ ++#define DT_VERDEFNUM 0x6ffffffd ++#define DT_VERNEED 0x6ffffffe ++ ++#define DT_VERNEEDNUM 0x6fffffff ++#define DT_VERSIONTAGIDX(tag) (DT_VERNEEDNUM - (tag)) ++#define DT_VERSIONTAGNUM 16 ++ ++ ++ ++#define DT_AUXILIARY 0x7ffffffd ++#define DT_FILTER 0x7fffffff ++#define DT_EXTRATAGIDX(tag) ((Elf32_Word)-((Elf32_Sword) (tag) <<1>>1)-1) ++#define DT_EXTRANUM 3 ++ ++ ++#define DF_ORIGIN 0x00000001 ++#define DF_SYMBOLIC 0x00000002 ++#define DF_TEXTREL 0x00000004 ++#define DF_BIND_NOW 0x00000008 ++#define DF_STATIC_TLS 0x00000010 ++ ++ ++ ++#define DF_1_NOW 0x00000001 ++#define DF_1_GLOBAL 0x00000002 ++#define DF_1_GROUP 0x00000004 ++#define DF_1_NODELETE 0x00000008 ++#define DF_1_LOADFLTR 0x00000010 ++#define DF_1_INITFIRST 0x00000020 ++#define DF_1_NOOPEN 0x00000040 ++#define DF_1_ORIGIN 0x00000080 ++#define DF_1_DIRECT 0x00000100 ++#define DF_1_TRANS 0x00000200 ++#define DF_1_INTERPOSE 0x00000400 ++#define DF_1_NODEFLIB 0x00000800 ++#define DF_1_NODUMP 0x00001000 ++#define DF_1_CONFALT 0x00002000 ++#define DF_1_ENDFILTEE 0x00004000 ++#define DF_1_DISPRELDNE 0x00008000 ++#define DF_1_DISPRELPND 0x00010000 ++#define DF_1_NODIRECT 0x00020000 ++#define DF_1_IGNMULDEF 0x00040000 ++#define DF_1_NOKSYMS 0x00080000 ++#define DF_1_NOHDR 0x00100000 ++#define DF_1_EDITED 0x00200000 ++#define DF_1_NORELOC 0x00400000 ++#define DF_1_SYMINTPOSE 0x00800000 ++#define DF_1_GLOBAUDIT 0x01000000 ++#define DF_1_SINGLETON 0x02000000 ++ ++#define DTF_1_PARINIT 0x00000001 ++#define DTF_1_CONFEXP 0x00000002 ++ ++ ++#define DF_P1_LAZYLOAD 0x00000001 ++#define DF_P1_GROUPPERM 0x00000002 ++ ++ ++ ++ ++typedef struct { ++ Elf32_Half vd_version; ++ Elf32_Half vd_flags; ++ Elf32_Half vd_ndx; ++ Elf32_Half vd_cnt; ++ Elf32_Word vd_hash; ++ Elf32_Word vd_aux; ++ Elf32_Word vd_next; ++} Elf32_Verdef; ++ ++typedef struct { ++ Elf64_Half vd_version; ++ Elf64_Half vd_flags; ++ Elf64_Half vd_ndx; ++ Elf64_Half vd_cnt; ++ Elf64_Word vd_hash; ++ Elf64_Word vd_aux; ++ Elf64_Word vd_next; ++} Elf64_Verdef; ++ ++ ++ ++#define VER_DEF_NONE 0 ++#define VER_DEF_CURRENT 1 ++#define VER_DEF_NUM 2 ++ ++ ++#define VER_FLG_BASE 0x1 ++#define VER_FLG_WEAK 0x2 ++ ++ ++#define VER_NDX_LOCAL 0 ++#define VER_NDX_GLOBAL 1 ++#define VER_NDX_LORESERVE 0xff00 ++#define VER_NDX_ELIMINATE 0xff01 ++ ++ ++ ++typedef struct { ++ Elf32_Word vda_name; ++ Elf32_Word vda_next; ++} Elf32_Verdaux; ++ ++typedef struct { ++ Elf64_Word vda_name; ++ Elf64_Word vda_next; ++} Elf64_Verdaux; ++ ++ ++ ++ ++typedef struct { ++ Elf32_Half vn_version; ++ Elf32_Half vn_cnt; ++ Elf32_Word vn_file; ++ Elf32_Word vn_aux; ++ Elf32_Word vn_next; ++} Elf32_Verneed; ++ ++typedef struct { ++ Elf64_Half vn_version; ++ Elf64_Half vn_cnt; ++ Elf64_Word vn_file; ++ Elf64_Word vn_aux; ++ Elf64_Word vn_next; ++} Elf64_Verneed; ++ ++ ++ ++#define VER_NEED_NONE 0 ++#define VER_NEED_CURRENT 1 ++#define VER_NEED_NUM 2 ++ ++ ++ ++typedef struct { ++ Elf32_Word vna_hash; ++ Elf32_Half vna_flags; ++ Elf32_Half vna_other; ++ Elf32_Word vna_name; ++ Elf32_Word vna_next; ++} Elf32_Vernaux; ++ ++typedef struct { ++ Elf64_Word vna_hash; ++ Elf64_Half vna_flags; ++ Elf64_Half vna_other; ++ Elf64_Word vna_name; ++ Elf64_Word vna_next; ++} Elf64_Vernaux; ++ ++ ++ ++#define VER_FLG_WEAK 0x2 ++ ++ ++ ++typedef struct { ++ uint32_t a_type; ++ union { ++ uint32_t a_val; ++ } a_un; ++} Elf32_auxv_t; ++ ++typedef struct { ++ uint64_t a_type; ++ union { ++ uint64_t a_val; ++ } a_un; ++} Elf64_auxv_t; ++ ++ ++ ++#define AT_NULL 0 ++#define AT_IGNORE 1 ++#define AT_EXECFD 2 ++#define AT_PHDR 3 ++#define AT_PHENT 4 ++#define AT_PHNUM 5 ++#define AT_PAGESZ 6 ++#define AT_BASE 7 ++#define AT_FLAGS 8 ++#define AT_ENTRY 9 ++#define AT_NOTELF 10 ++#define AT_UID 11 ++#define AT_EUID 12 ++#define AT_GID 13 ++#define AT_EGID 14 ++#define AT_CLKTCK 17 ++ ++ ++#define AT_PLATFORM 15 ++#define AT_HWCAP 16 ++ ++ ++ ++ ++#define AT_FPUCW 18 ++ ++ ++#define AT_DCACHEBSIZE 19 ++#define AT_ICACHEBSIZE 20 ++#define AT_UCACHEBSIZE 21 ++ ++ ++ ++#define AT_IGNOREPPC 22 ++ ++#define AT_SECURE 23 ++ ++#define AT_BASE_PLATFORM 24 ++ ++#define AT_RANDOM 25 ++ ++#define AT_HWCAP2 26 ++ ++#define AT_EXECFN 31 ++ ++ ++ ++#define AT_SYSINFO 32 ++#define AT_SYSINFO_EHDR 33 ++ ++ ++ ++#define AT_L1I_CACHESHAPE 34 ++#define AT_L1D_CACHESHAPE 35 ++#define AT_L2_CACHESHAPE 36 ++#define AT_L3_CACHESHAPE 37 ++ ++ ++ ++ ++typedef struct { ++ Elf32_Word n_namesz; ++ Elf32_Word n_descsz; ++ Elf32_Word n_type; ++} Elf32_Nhdr; ++ ++typedef struct { ++ Elf64_Word n_namesz; ++ Elf64_Word n_descsz; ++ Elf64_Word n_type; ++} Elf64_Nhdr; ++ ++ ++ ++ ++#define ELF_NOTE_SOLARIS "SUNW Solaris" ++ ++ ++#define ELF_NOTE_GNU "GNU" ++ ++ ++ ++ ++ ++#define ELF_NOTE_PAGESIZE_HINT 1 ++ ++ ++#define NT_GNU_ABI_TAG 1 ++#define ELF_NOTE_ABI NT_GNU_ABI_TAG ++ ++ ++ ++#define ELF_NOTE_OS_LINUX 0 ++#define ELF_NOTE_OS_GNU 1 ++#define ELF_NOTE_OS_SOLARIS2 2 ++#define ELF_NOTE_OS_FREEBSD 3 ++ ++#define NT_GNU_BUILD_ID 3 ++#define NT_GNU_GOLD_VERSION 4 ++ ++ ++ ++typedef struct { ++ Elf32_Xword m_value; ++ Elf32_Word m_info; ++ Elf32_Word m_poffset; ++ Elf32_Half m_repeat; ++ Elf32_Half m_stride; ++} Elf32_Move; ++ ++typedef struct { ++ Elf64_Xword m_value; ++ Elf64_Xword m_info; ++ Elf64_Xword m_poffset; ++ Elf64_Half m_repeat; ++ Elf64_Half m_stride; ++} Elf64_Move; ++ ++ ++#define ELF32_M_SYM(info) ((info) >> 8) ++#define ELF32_M_SIZE(info) ((unsigned char) (info)) ++#define ELF32_M_INFO(sym, size) (((sym) << 8) + (unsigned char) (size)) ++ ++#define ELF64_M_SYM(info) ELF32_M_SYM (info) ++#define ELF64_M_SIZE(info) ELF32_M_SIZE (info) ++#define ELF64_M_INFO(sym, size) ELF32_M_INFO (sym, size) ++ ++#define EF_CPU32 0x00810000 ++ ++#define R_68K_NONE 0 ++#define R_68K_32 1 ++#define R_68K_16 2 ++#define R_68K_8 3 ++#define R_68K_PC32 4 ++#define R_68K_PC16 5 ++#define R_68K_PC8 6 ++#define R_68K_GOT32 7 ++#define R_68K_GOT16 8 ++#define R_68K_GOT8 9 ++#define R_68K_GOT32O 10 ++#define R_68K_GOT16O 11 ++#define R_68K_GOT8O 12 ++#define R_68K_PLT32 13 ++#define R_68K_PLT16 14 ++#define R_68K_PLT8 15 ++#define R_68K_PLT32O 16 ++#define R_68K_PLT16O 17 ++#define R_68K_PLT8O 18 ++#define R_68K_COPY 19 ++#define R_68K_GLOB_DAT 20 ++#define R_68K_JMP_SLOT 21 ++#define R_68K_RELATIVE 22 ++#define R_68K_NUM 23 ++ ++#define R_386_NONE 0 ++#define R_386_32 1 ++#define R_386_PC32 2 ++#define R_386_GOT32 3 ++#define R_386_PLT32 4 ++#define R_386_COPY 5 ++#define R_386_GLOB_DAT 6 ++#define R_386_JMP_SLOT 7 ++#define R_386_RELATIVE 8 ++#define R_386_GOTOFF 9 ++#define R_386_GOTPC 10 ++#define R_386_32PLT 11 ++#define R_386_TLS_TPOFF 14 ++#define R_386_TLS_IE 15 ++#define R_386_TLS_GOTIE 16 ++#define R_386_TLS_LE 17 ++#define R_386_TLS_GD 18 ++#define R_386_TLS_LDM 19 ++#define R_386_16 20 ++#define R_386_PC16 21 ++#define R_386_8 22 ++#define R_386_PC8 23 ++#define R_386_TLS_GD_32 24 ++#define R_386_TLS_GD_PUSH 25 ++#define R_386_TLS_GD_CALL 26 ++#define R_386_TLS_GD_POP 27 ++#define R_386_TLS_LDM_32 28 ++#define R_386_TLS_LDM_PUSH 29 ++#define R_386_TLS_LDM_CALL 30 ++#define R_386_TLS_LDM_POP 31 ++#define R_386_TLS_LDO_32 32 ++#define R_386_TLS_IE_32 33 ++#define R_386_TLS_LE_32 34 ++#define R_386_TLS_DTPMOD32 35 ++#define R_386_TLS_DTPOFF32 36 ++#define R_386_TLS_TPOFF32 37 ++#define R_386_SIZE32 38 ++#define R_386_TLS_GOTDESC 39 ++#define R_386_TLS_DESC_CALL 40 ++#define R_386_TLS_DESC 41 ++#define R_386_IRELATIVE 42 ++#define R_386_NUM 43 ++ ++ ++ ++ ++ ++#define STT_SPARC_REGISTER 13 ++ ++ ++ ++#define EF_SPARCV9_MM 3 ++#define EF_SPARCV9_TSO 0 ++#define EF_SPARCV9_PSO 1 ++#define EF_SPARCV9_RMO 2 ++#define EF_SPARC_LEDATA 0x800000 ++#define EF_SPARC_EXT_MASK 0xFFFF00 ++#define EF_SPARC_32PLUS 0x000100 ++#define EF_SPARC_SUN_US1 0x000200 ++#define EF_SPARC_HAL_R1 0x000400 ++#define EF_SPARC_SUN_US3 0x000800 ++ ++ ++ ++#define R_SPARC_NONE 0 ++#define R_SPARC_8 1 ++#define R_SPARC_16 2 ++#define R_SPARC_32 3 ++#define R_SPARC_DISP8 4 ++#define R_SPARC_DISP16 5 ++#define R_SPARC_DISP32 6 ++#define R_SPARC_WDISP30 7 ++#define R_SPARC_WDISP22 8 ++#define R_SPARC_HI22 9 ++#define R_SPARC_22 10 ++#define R_SPARC_13 11 ++#define R_SPARC_LO10 12 ++#define R_SPARC_GOT10 13 ++#define R_SPARC_GOT13 14 ++#define R_SPARC_GOT22 15 ++#define R_SPARC_PC10 16 ++#define R_SPARC_PC22 17 ++#define R_SPARC_WPLT30 18 ++#define R_SPARC_COPY 19 ++#define R_SPARC_GLOB_DAT 20 ++#define R_SPARC_JMP_SLOT 21 ++#define R_SPARC_RELATIVE 22 ++#define R_SPARC_UA32 23 ++ ++ ++ ++#define R_SPARC_PLT32 24 ++#define R_SPARC_HIPLT22 25 ++#define R_SPARC_LOPLT10 26 ++#define R_SPARC_PCPLT32 27 ++#define R_SPARC_PCPLT22 28 ++#define R_SPARC_PCPLT10 29 ++#define R_SPARC_10 30 ++#define R_SPARC_11 31 ++#define R_SPARC_64 32 ++#define R_SPARC_OLO10 33 ++#define R_SPARC_HH22 34 ++#define R_SPARC_HM10 35 ++#define R_SPARC_LM22 36 ++#define R_SPARC_PC_HH22 37 ++#define R_SPARC_PC_HM10 38 ++#define R_SPARC_PC_LM22 39 ++#define R_SPARC_WDISP16 40 ++#define R_SPARC_WDISP19 41 ++#define R_SPARC_GLOB_JMP 42 ++#define R_SPARC_7 43 ++#define R_SPARC_5 44 ++#define R_SPARC_6 45 ++#define R_SPARC_DISP64 46 ++#define R_SPARC_PLT64 47 ++#define R_SPARC_HIX22 48 ++#define R_SPARC_LOX10 49 ++#define R_SPARC_H44 50 ++#define R_SPARC_M44 51 ++#define R_SPARC_L44 52 ++#define R_SPARC_REGISTER 53 ++#define R_SPARC_UA64 54 ++#define R_SPARC_UA16 55 ++#define R_SPARC_TLS_GD_HI22 56 ++#define R_SPARC_TLS_GD_LO10 57 ++#define R_SPARC_TLS_GD_ADD 58 ++#define R_SPARC_TLS_GD_CALL 59 ++#define R_SPARC_TLS_LDM_HI22 60 ++#define R_SPARC_TLS_LDM_LO10 61 ++#define R_SPARC_TLS_LDM_ADD 62 ++#define R_SPARC_TLS_LDM_CALL 63 ++#define R_SPARC_TLS_LDO_HIX22 64 ++#define R_SPARC_TLS_LDO_LOX10 65 ++#define R_SPARC_TLS_LDO_ADD 66 ++#define R_SPARC_TLS_IE_HI22 67 ++#define R_SPARC_TLS_IE_LO10 68 ++#define R_SPARC_TLS_IE_LD 69 ++#define R_SPARC_TLS_IE_LDX 70 ++#define R_SPARC_TLS_IE_ADD 71 ++#define R_SPARC_TLS_LE_HIX22 72 ++#define R_SPARC_TLS_LE_LOX10 73 ++#define R_SPARC_TLS_DTPMOD32 74 ++#define R_SPARC_TLS_DTPMOD64 75 ++#define R_SPARC_TLS_DTPOFF32 76 ++#define R_SPARC_TLS_DTPOFF64 77 ++#define R_SPARC_TLS_TPOFF32 78 ++#define R_SPARC_TLS_TPOFF64 79 ++#define R_SPARC_GOTDATA_HIX22 80 ++#define R_SPARC_GOTDATA_LOX10 81 ++#define R_SPARC_GOTDATA_OP_HIX22 82 ++#define R_SPARC_GOTDATA_OP_LOX10 83 ++#define R_SPARC_GOTDATA_OP 84 ++#define R_SPARC_H34 85 ++#define R_SPARC_SIZE32 86 ++#define R_SPARC_SIZE64 87 ++#define R_SPARC_GNU_VTINHERIT 250 ++#define R_SPARC_GNU_VTENTRY 251 ++#define R_SPARC_REV32 252 ++ ++#define R_SPARC_NUM 253 ++ ++ ++ ++#define DT_SPARC_REGISTER 0x70000001 ++#define DT_SPARC_NUM 2 ++ ++ ++#define EF_MIPS_NOREORDER 1 ++#define EF_MIPS_PIC 2 ++#define EF_MIPS_CPIC 4 ++#define EF_MIPS_XGOT 8 ++#define EF_MIPS_64BIT_WHIRL 16 ++#define EF_MIPS_ABI2 32 ++#define EF_MIPS_ABI_ON32 64 ++#define EF_MIPS_ARCH 0xf0000000 ++ ++ ++ ++#define EF_MIPS_ARCH_1 0x00000000 ++#define EF_MIPS_ARCH_2 0x10000000 ++#define EF_MIPS_ARCH_3 0x20000000 ++#define EF_MIPS_ARCH_4 0x30000000 ++#define EF_MIPS_ARCH_5 0x40000000 ++#define EF_MIPS_ARCH_32 0x50000000 ++#define EF_MIPS_ARCH_64 0x60000000 ++#define EF_MIPS_ARCH_32R2 0x70000000 ++#define EF_MIPS_ARCH_64R2 0x80000000 ++ ++ ++#define E_MIPS_ARCH_1 0x00000000 ++#define E_MIPS_ARCH_2 0x10000000 ++#define E_MIPS_ARCH_3 0x20000000 ++#define E_MIPS_ARCH_4 0x30000000 ++#define E_MIPS_ARCH_5 0x40000000 ++#define E_MIPS_ARCH_32 0x50000000 ++#define E_MIPS_ARCH_64 0x60000000 ++ ++ ++ ++#define SHN_MIPS_ACOMMON 0xff00 ++#define SHN_MIPS_TEXT 0xff01 ++#define SHN_MIPS_DATA 0xff02 ++#define SHN_MIPS_SCOMMON 0xff03 ++#define SHN_MIPS_SUNDEFINED 0xff04 ++ ++ ++ ++#define SHT_MIPS_LIBLIST 0x70000000 ++#define SHT_MIPS_MSYM 0x70000001 ++#define SHT_MIPS_CONFLICT 0x70000002 ++#define SHT_MIPS_GPTAB 0x70000003 ++#define SHT_MIPS_UCODE 0x70000004 ++#define SHT_MIPS_DEBUG 0x70000005 ++#define SHT_MIPS_REGINFO 0x70000006 ++#define SHT_MIPS_PACKAGE 0x70000007 ++#define SHT_MIPS_PACKSYM 0x70000008 ++#define SHT_MIPS_RELD 0x70000009 ++#define SHT_MIPS_IFACE 0x7000000b ++#define SHT_MIPS_CONTENT 0x7000000c ++#define SHT_MIPS_OPTIONS 0x7000000d ++#define SHT_MIPS_SHDR 0x70000010 ++#define SHT_MIPS_FDESC 0x70000011 ++#define SHT_MIPS_EXTSYM 0x70000012 ++#define SHT_MIPS_DENSE 0x70000013 ++#define SHT_MIPS_PDESC 0x70000014 ++#define SHT_MIPS_LOCSYM 0x70000015 ++#define SHT_MIPS_AUXSYM 0x70000016 ++#define SHT_MIPS_OPTSYM 0x70000017 ++#define SHT_MIPS_LOCSTR 0x70000018 ++#define SHT_MIPS_LINE 0x70000019 ++#define SHT_MIPS_RFDESC 0x7000001a ++#define SHT_MIPS_DELTASYM 0x7000001b ++#define SHT_MIPS_DELTAINST 0x7000001c ++#define SHT_MIPS_DELTACLASS 0x7000001d ++#define SHT_MIPS_DWARF 0x7000001e ++#define SHT_MIPS_DELTADECL 0x7000001f ++#define SHT_MIPS_SYMBOL_LIB 0x70000020 ++#define SHT_MIPS_EVENTS 0x70000021 ++#define SHT_MIPS_TRANSLATE 0x70000022 ++#define SHT_MIPS_PIXIE 0x70000023 ++#define SHT_MIPS_XLATE 0x70000024 ++#define SHT_MIPS_XLATE_DEBUG 0x70000025 ++#define SHT_MIPS_WHIRL 0x70000026 ++#define SHT_MIPS_EH_REGION 0x70000027 ++#define SHT_MIPS_XLATE_OLD 0x70000028 ++#define SHT_MIPS_PDR_EXCEPTION 0x70000029 ++ ++ ++ ++#define SHF_MIPS_GPREL 0x10000000 ++#define SHF_MIPS_MERGE 0x20000000 ++#define SHF_MIPS_ADDR 0x40000000 ++#define SHF_MIPS_STRINGS 0x80000000 ++#define SHF_MIPS_NOSTRIP 0x08000000 ++#define SHF_MIPS_LOCAL 0x04000000 ++#define SHF_MIPS_NAMES 0x02000000 ++#define SHF_MIPS_NODUPE 0x01000000 ++ ++ ++ ++ ++ ++#define STO_MIPS_DEFAULT 0x0 ++#define STO_MIPS_INTERNAL 0x1 ++#define STO_MIPS_HIDDEN 0x2 ++#define STO_MIPS_PROTECTED 0x3 ++#define STO_MIPS_PLT 0x8 ++#define STO_MIPS_SC_ALIGN_UNUSED 0xff ++ ++ ++#define STB_MIPS_SPLIT_COMMON 13 ++ ++ ++ ++typedef union { ++ struct { ++ Elf32_Word gt_current_g_value; ++ Elf32_Word gt_unused; ++ } gt_header; ++ struct { ++ Elf32_Word gt_g_value; ++ Elf32_Word gt_bytes; ++ } gt_entry; ++} Elf32_gptab; ++ ++ ++ ++typedef struct { ++ Elf32_Word ri_gprmask; ++ Elf32_Word ri_cprmask[4]; ++ Elf32_Sword ri_gp_value; ++} Elf32_RegInfo; ++ ++ ++ ++typedef struct { ++ unsigned char kind; ++ ++ unsigned char size; ++ Elf32_Section section; ++ ++ Elf32_Word info; ++} Elf_Options; ++ ++ ++ ++#define ODK_NULL 0 ++#define ODK_REGINFO 1 ++#define ODK_EXCEPTIONS 2 ++#define ODK_PAD 3 ++#define ODK_HWPATCH 4 ++#define ODK_FILL 5 ++#define ODK_TAGS 6 ++#define ODK_HWAND 7 ++#define ODK_HWOR 8 ++ ++ ++ ++#define OEX_FPU_MIN 0x1f ++#define OEX_FPU_MAX 0x1f00 ++#define OEX_PAGE0 0x10000 ++#define OEX_SMM 0x20000 ++#define OEX_FPDBUG 0x40000 ++#define OEX_PRECISEFP OEX_FPDBUG ++#define OEX_DISMISS 0x80000 ++ ++#define OEX_FPU_INVAL 0x10 ++#define OEX_FPU_DIV0 0x08 ++#define OEX_FPU_OFLO 0x04 ++#define OEX_FPU_UFLO 0x02 ++#define OEX_FPU_INEX 0x01 ++ ++ ++ ++#define OHW_R4KEOP 0x1 ++#define OHW_R8KPFETCH 0x2 ++#define OHW_R5KEOP 0x4 ++#define OHW_R5KCVTL 0x8 ++ ++#define OPAD_PREFIX 0x1 ++#define OPAD_POSTFIX 0x2 ++#define OPAD_SYMBOL 0x4 ++ ++ ++ ++typedef struct { ++ Elf32_Word hwp_flags1; ++ Elf32_Word hwp_flags2; ++} Elf_Options_Hw; ++ ++ ++ ++#define OHWA0_R4KEOP_CHECKED 0x00000001 ++#define OHWA1_R4KEOP_CLEAN 0x00000002 ++ ++ ++ ++#define R_MIPS_NONE 0 ++#define R_MIPS_16 1 ++#define R_MIPS_32 2 ++#define R_MIPS_REL32 3 ++#define R_MIPS_26 4 ++#define R_MIPS_HI16 5 ++#define R_MIPS_LO16 6 ++#define R_MIPS_GPREL16 7 ++#define R_MIPS_LITERAL 8 ++#define R_MIPS_GOT16 9 ++#define R_MIPS_PC16 10 ++#define R_MIPS_CALL16 11 ++#define R_MIPS_GPREL32 12 ++ ++#define R_MIPS_SHIFT5 16 ++#define R_MIPS_SHIFT6 17 ++#define R_MIPS_64 18 ++#define R_MIPS_GOT_DISP 19 ++#define R_MIPS_GOT_PAGE 20 ++#define R_MIPS_GOT_OFST 21 ++#define R_MIPS_GOT_HI16 22 ++#define R_MIPS_GOT_LO16 23 ++#define R_MIPS_SUB 24 ++#define R_MIPS_INSERT_A 25 ++#define R_MIPS_INSERT_B 26 ++#define R_MIPS_DELETE 27 ++#define R_MIPS_HIGHER 28 ++#define R_MIPS_HIGHEST 29 ++#define R_MIPS_CALL_HI16 30 ++#define R_MIPS_CALL_LO16 31 ++#define R_MIPS_SCN_DISP 32 ++#define R_MIPS_REL16 33 ++#define R_MIPS_ADD_IMMEDIATE 34 ++#define R_MIPS_PJUMP 35 ++#define R_MIPS_RELGOT 36 ++#define R_MIPS_JALR 37 ++#define R_MIPS_TLS_DTPMOD32 38 ++#define R_MIPS_TLS_DTPREL32 39 ++#define R_MIPS_TLS_DTPMOD64 40 ++#define R_MIPS_TLS_DTPREL64 41 ++#define R_MIPS_TLS_GD 42 ++#define R_MIPS_TLS_LDM 43 ++#define R_MIPS_TLS_DTPREL_HI16 44 ++#define R_MIPS_TLS_DTPREL_LO16 45 ++#define R_MIPS_TLS_GOTTPREL 46 ++#define R_MIPS_TLS_TPREL32 47 ++#define R_MIPS_TLS_TPREL64 48 ++#define R_MIPS_TLS_TPREL_HI16 49 ++#define R_MIPS_TLS_TPREL_LO16 50 ++#define R_MIPS_GLOB_DAT 51 ++#define R_MIPS_COPY 126 ++#define R_MIPS_JUMP_SLOT 127 ++ ++#define R_MIPS_NUM 128 ++ ++ ++ ++#define PT_MIPS_REGINFO 0x70000000 ++#define PT_MIPS_RTPROC 0x70000001 ++#define PT_MIPS_OPTIONS 0x70000002 ++ ++ ++ ++#define PF_MIPS_LOCAL 0x10000000 ++ ++ ++ ++#define DT_MIPS_RLD_VERSION 0x70000001 ++#define DT_MIPS_TIME_STAMP 0x70000002 ++#define DT_MIPS_ICHECKSUM 0x70000003 ++#define DT_MIPS_IVERSION 0x70000004 ++#define DT_MIPS_FLAGS 0x70000005 ++#define DT_MIPS_BASE_ADDRESS 0x70000006 ++#define DT_MIPS_MSYM 0x70000007 ++#define DT_MIPS_CONFLICT 0x70000008 ++#define DT_MIPS_LIBLIST 0x70000009 ++#define DT_MIPS_LOCAL_GOTNO 0x7000000a ++#define DT_MIPS_CONFLICTNO 0x7000000b ++#define DT_MIPS_LIBLISTNO 0x70000010 ++#define DT_MIPS_SYMTABNO 0x70000011 ++#define DT_MIPS_UNREFEXTNO 0x70000012 ++#define DT_MIPS_GOTSYM 0x70000013 ++#define DT_MIPS_HIPAGENO 0x70000014 ++#define DT_MIPS_RLD_MAP 0x70000016 ++#define DT_MIPS_DELTA_CLASS 0x70000017 ++#define DT_MIPS_DELTA_CLASS_NO 0x70000018 ++ ++#define DT_MIPS_DELTA_INSTANCE 0x70000019 ++#define DT_MIPS_DELTA_INSTANCE_NO 0x7000001a ++ ++#define DT_MIPS_DELTA_RELOC 0x7000001b ++#define DT_MIPS_DELTA_RELOC_NO 0x7000001c ++ ++#define DT_MIPS_DELTA_SYM 0x7000001d ++ ++#define DT_MIPS_DELTA_SYM_NO 0x7000001e ++ ++#define DT_MIPS_DELTA_CLASSSYM 0x70000020 ++ ++#define DT_MIPS_DELTA_CLASSSYM_NO 0x70000021 ++ ++#define DT_MIPS_CXX_FLAGS 0x70000022 ++#define DT_MIPS_PIXIE_INIT 0x70000023 ++#define DT_MIPS_SYMBOL_LIB 0x70000024 ++#define DT_MIPS_LOCALPAGE_GOTIDX 0x70000025 ++#define DT_MIPS_LOCAL_GOTIDX 0x70000026 ++#define DT_MIPS_HIDDEN_GOTIDX 0x70000027 ++#define DT_MIPS_PROTECTED_GOTIDX 0x70000028 ++#define DT_MIPS_OPTIONS 0x70000029 ++#define DT_MIPS_INTERFACE 0x7000002a ++#define DT_MIPS_DYNSTR_ALIGN 0x7000002b ++#define DT_MIPS_INTERFACE_SIZE 0x7000002c ++#define DT_MIPS_RLD_TEXT_RESOLVE_ADDR 0x7000002d ++ ++#define DT_MIPS_PERF_SUFFIX 0x7000002e ++ ++#define DT_MIPS_COMPACT_SIZE 0x7000002f ++#define DT_MIPS_GP_VALUE 0x70000030 ++#define DT_MIPS_AUX_DYNAMIC 0x70000031 ++ ++#define DT_MIPS_PLTGOT 0x70000032 ++ ++#define DT_MIPS_RWPLT 0x70000034 ++#define DT_MIPS_NUM 0x35 ++ ++ ++ ++#define RHF_NONE 0 ++#define RHF_QUICKSTART (1 << 0) ++#define RHF_NOTPOT (1 << 1) ++#define RHF_NO_LIBRARY_REPLACEMENT (1 << 2) ++#define RHF_NO_MOVE (1 << 3) ++#define RHF_SGI_ONLY (1 << 4) ++#define RHF_GUARANTEE_INIT (1 << 5) ++#define RHF_DELTA_C_PLUS_PLUS (1 << 6) ++#define RHF_GUARANTEE_START_INIT (1 << 7) ++#define RHF_PIXIE (1 << 8) ++#define RHF_DEFAULT_DELAY_LOAD (1 << 9) ++#define RHF_REQUICKSTART (1 << 10) ++#define RHF_REQUICKSTARTED (1 << 11) ++#define RHF_CORD (1 << 12) ++#define RHF_NO_UNRES_UNDEF (1 << 13) ++#define RHF_RLD_ORDER_SAFE (1 << 14) ++ ++ ++ ++typedef struct ++{ ++ Elf32_Word l_name; ++ Elf32_Word l_time_stamp; ++ Elf32_Word l_checksum; ++ Elf32_Word l_version; ++ Elf32_Word l_flags; ++} Elf32_Lib; ++ ++typedef struct ++{ ++ Elf64_Word l_name; ++ Elf64_Word l_time_stamp; ++ Elf64_Word l_checksum; ++ Elf64_Word l_version; ++ Elf64_Word l_flags; ++} Elf64_Lib; ++ ++ ++ ++ ++#define LL_NONE 0 ++#define LL_EXACT_MATCH (1 << 0) ++#define LL_IGNORE_INT_VER (1 << 1) ++#define LL_REQUIRE_MINOR (1 << 2) ++#define LL_EXPORTS (1 << 3) ++#define LL_DELAY_LOAD (1 << 4) ++#define LL_DELTA (1 << 5) ++ ++ ++ ++typedef Elf32_Addr Elf32_Conflict; ++ ++ ++ ++ ++ ++ ++#define EF_PARISC_TRAPNIL 0x00010000 ++#define EF_PARISC_EXT 0x00020000 ++#define EF_PARISC_LSB 0x00040000 ++#define EF_PARISC_WIDE 0x00080000 ++#define EF_PARISC_NO_KABP 0x00100000 ++ ++#define EF_PARISC_LAZYSWAP 0x00400000 ++#define EF_PARISC_ARCH 0x0000ffff ++ ++ ++ ++#define EFA_PARISC_1_0 0x020b ++#define EFA_PARISC_1_1 0x0210 ++#define EFA_PARISC_2_0 0x0214 ++ ++ ++ ++#define SHN_PARISC_ANSI_COMMON 0xff00 ++ ++#define SHN_PARISC_HUGE_COMMON 0xff01 ++ ++ ++ ++#define SHT_PARISC_EXT 0x70000000 ++#define SHT_PARISC_UNWIND 0x70000001 ++#define SHT_PARISC_DOC 0x70000002 ++ ++ ++ ++#define SHF_PARISC_SHORT 0x20000000 ++#define SHF_PARISC_HUGE 0x40000000 ++#define SHF_PARISC_SBP 0x80000000 ++ ++ ++ ++#define STT_PARISC_MILLICODE 13 ++ ++#define STT_HP_OPAQUE (STT_LOOS + 0x1) ++#define STT_HP_STUB (STT_LOOS + 0x2) ++ ++ ++ ++#define R_PARISC_NONE 0 ++#define R_PARISC_DIR32 1 ++#define R_PARISC_DIR21L 2 ++#define R_PARISC_DIR17R 3 ++#define R_PARISC_DIR17F 4 ++#define R_PARISC_DIR14R 6 ++#define R_PARISC_PCREL32 9 ++#define R_PARISC_PCREL21L 10 ++#define R_PARISC_PCREL17R 11 ++#define R_PARISC_PCREL17F 12 ++#define R_PARISC_PCREL14R 14 ++#define R_PARISC_DPREL21L 18 ++#define R_PARISC_DPREL14R 22 ++#define R_PARISC_GPREL21L 26 ++#define R_PARISC_GPREL14R 30 ++#define R_PARISC_LTOFF21L 34 ++#define R_PARISC_LTOFF14R 38 ++#define R_PARISC_SECREL32 41 ++#define R_PARISC_SEGBASE 48 ++#define R_PARISC_SEGREL32 49 ++#define R_PARISC_PLTOFF21L 50 ++#define R_PARISC_PLTOFF14R 54 ++#define R_PARISC_LTOFF_FPTR32 57 ++#define R_PARISC_LTOFF_FPTR21L 58 ++#define R_PARISC_LTOFF_FPTR14R 62 ++#define R_PARISC_FPTR64 64 ++#define R_PARISC_PLABEL32 65 ++#define R_PARISC_PLABEL21L 66 ++#define R_PARISC_PLABEL14R 70 ++#define R_PARISC_PCREL64 72 ++#define R_PARISC_PCREL22F 74 ++#define R_PARISC_PCREL14WR 75 ++#define R_PARISC_PCREL14DR 76 ++#define R_PARISC_PCREL16F 77 ++#define R_PARISC_PCREL16WF 78 ++#define R_PARISC_PCREL16DF 79 ++#define R_PARISC_DIR64 80 ++#define R_PARISC_DIR14WR 83 ++#define R_PARISC_DIR14DR 84 ++#define R_PARISC_DIR16F 85 ++#define R_PARISC_DIR16WF 86 ++#define R_PARISC_DIR16DF 87 ++#define R_PARISC_GPREL64 88 ++#define R_PARISC_GPREL14WR 91 ++#define R_PARISC_GPREL14DR 92 ++#define R_PARISC_GPREL16F 93 ++#define R_PARISC_GPREL16WF 94 ++#define R_PARISC_GPREL16DF 95 ++#define R_PARISC_LTOFF64 96 ++#define R_PARISC_LTOFF14WR 99 ++#define R_PARISC_LTOFF14DR 100 ++#define R_PARISC_LTOFF16F 101 ++#define R_PARISC_LTOFF16WF 102 ++#define R_PARISC_LTOFF16DF 103 ++#define R_PARISC_SECREL64 104 ++#define R_PARISC_SEGREL64 112 ++#define R_PARISC_PLTOFF14WR 115 ++#define R_PARISC_PLTOFF14DR 116 ++#define R_PARISC_PLTOFF16F 117 ++#define R_PARISC_PLTOFF16WF 118 ++#define R_PARISC_PLTOFF16DF 119 ++#define R_PARISC_LTOFF_FPTR64 120 ++#define R_PARISC_LTOFF_FPTR14WR 123 ++#define R_PARISC_LTOFF_FPTR14DR 124 ++#define R_PARISC_LTOFF_FPTR16F 125 ++#define R_PARISC_LTOFF_FPTR16WF 126 ++#define R_PARISC_LTOFF_FPTR16DF 127 ++#define R_PARISC_LORESERVE 128 ++#define R_PARISC_COPY 128 ++#define R_PARISC_IPLT 129 ++#define R_PARISC_EPLT 130 ++#define R_PARISC_TPREL32 153 ++#define R_PARISC_TPREL21L 154 ++#define R_PARISC_TPREL14R 158 ++#define R_PARISC_LTOFF_TP21L 162 ++#define R_PARISC_LTOFF_TP14R 166 ++#define R_PARISC_LTOFF_TP14F 167 ++#define R_PARISC_TPREL64 216 ++#define R_PARISC_TPREL14WR 219 ++#define R_PARISC_TPREL14DR 220 ++#define R_PARISC_TPREL16F 221 ++#define R_PARISC_TPREL16WF 222 ++#define R_PARISC_TPREL16DF 223 ++#define R_PARISC_LTOFF_TP64 224 ++#define R_PARISC_LTOFF_TP14WR 227 ++#define R_PARISC_LTOFF_TP14DR 228 ++#define R_PARISC_LTOFF_TP16F 229 ++#define R_PARISC_LTOFF_TP16WF 230 ++#define R_PARISC_LTOFF_TP16DF 231 ++#define R_PARISC_GNU_VTENTRY 232 ++#define R_PARISC_GNU_VTINHERIT 233 ++#define R_PARISC_TLS_GD21L 234 ++#define R_PARISC_TLS_GD14R 235 ++#define R_PARISC_TLS_GDCALL 236 ++#define R_PARISC_TLS_LDM21L 237 ++#define R_PARISC_TLS_LDM14R 238 ++#define R_PARISC_TLS_LDMCALL 239 ++#define R_PARISC_TLS_LDO21L 240 ++#define R_PARISC_TLS_LDO14R 241 ++#define R_PARISC_TLS_DTPMOD32 242 ++#define R_PARISC_TLS_DTPMOD64 243 ++#define R_PARISC_TLS_DTPOFF32 244 ++#define R_PARISC_TLS_DTPOFF64 245 ++#define R_PARISC_TLS_LE21L R_PARISC_TPREL21L ++#define R_PARISC_TLS_LE14R R_PARISC_TPREL14R ++#define R_PARISC_TLS_IE21L R_PARISC_LTOFF_TP21L ++#define R_PARISC_TLS_IE14R R_PARISC_LTOFF_TP14R ++#define R_PARISC_TLS_TPREL32 R_PARISC_TPREL32 ++#define R_PARISC_TLS_TPREL64 R_PARISC_TPREL64 ++#define R_PARISC_HIRESERVE 255 ++ ++ ++ ++#define PT_HP_TLS (PT_LOOS + 0x0) ++#define PT_HP_CORE_NONE (PT_LOOS + 0x1) ++#define PT_HP_CORE_VERSION (PT_LOOS + 0x2) ++#define PT_HP_CORE_KERNEL (PT_LOOS + 0x3) ++#define PT_HP_CORE_COMM (PT_LOOS + 0x4) ++#define PT_HP_CORE_PROC (PT_LOOS + 0x5) ++#define PT_HP_CORE_LOADABLE (PT_LOOS + 0x6) ++#define PT_HP_CORE_STACK (PT_LOOS + 0x7) ++#define PT_HP_CORE_SHM (PT_LOOS + 0x8) ++#define PT_HP_CORE_MMF (PT_LOOS + 0x9) ++#define PT_HP_PARALLEL (PT_LOOS + 0x10) ++#define PT_HP_FASTBIND (PT_LOOS + 0x11) ++#define PT_HP_OPT_ANNOT (PT_LOOS + 0x12) ++#define PT_HP_HSL_ANNOT (PT_LOOS + 0x13) ++#define PT_HP_STACK (PT_LOOS + 0x14) ++ ++#define PT_PARISC_ARCHEXT 0x70000000 ++#define PT_PARISC_UNWIND 0x70000001 ++ ++ ++ ++#define PF_PARISC_SBP 0x08000000 ++ ++#define PF_HP_PAGE_SIZE 0x00100000 ++#define PF_HP_FAR_SHARED 0x00200000 ++#define PF_HP_NEAR_SHARED 0x00400000 ++#define PF_HP_CODE 0x01000000 ++#define PF_HP_MODIFY 0x02000000 ++#define PF_HP_LAZYSWAP 0x04000000 ++#define PF_HP_SBP 0x08000000 ++ ++ ++ ++ ++ ++ ++#define EF_ALPHA_32BIT 1 ++#define EF_ALPHA_CANRELAX 2 ++ ++ ++ ++ ++#define SHT_ALPHA_DEBUG 0x70000001 ++#define SHT_ALPHA_REGINFO 0x70000002 ++ ++ ++ ++#define SHF_ALPHA_GPREL 0x10000000 ++ ++ ++#define STO_ALPHA_NOPV 0x80 ++#define STO_ALPHA_STD_GPLOAD 0x88 ++ ++ ++ ++#define R_ALPHA_NONE 0 ++#define R_ALPHA_REFLONG 1 ++#define R_ALPHA_REFQUAD 2 ++#define R_ALPHA_GPREL32 3 ++#define R_ALPHA_LITERAL 4 ++#define R_ALPHA_LITUSE 5 ++#define R_ALPHA_GPDISP 6 ++#define R_ALPHA_BRADDR 7 ++#define R_ALPHA_HINT 8 ++#define R_ALPHA_SREL16 9 ++#define R_ALPHA_SREL32 10 ++#define R_ALPHA_SREL64 11 ++#define R_ALPHA_GPRELHIGH 17 ++#define R_ALPHA_GPRELLOW 18 ++#define R_ALPHA_GPREL16 19 ++#define R_ALPHA_COPY 24 ++#define R_ALPHA_GLOB_DAT 25 ++#define R_ALPHA_JMP_SLOT 26 ++#define R_ALPHA_RELATIVE 27 ++#define R_ALPHA_TLS_GD_HI 28 ++#define R_ALPHA_TLSGD 29 ++#define R_ALPHA_TLS_LDM 30 ++#define R_ALPHA_DTPMOD64 31 ++#define R_ALPHA_GOTDTPREL 32 ++#define R_ALPHA_DTPREL64 33 ++#define R_ALPHA_DTPRELHI 34 ++#define R_ALPHA_DTPRELLO 35 ++#define R_ALPHA_DTPREL16 36 ++#define R_ALPHA_GOTTPREL 37 ++#define R_ALPHA_TPREL64 38 ++#define R_ALPHA_TPRELHI 39 ++#define R_ALPHA_TPRELLO 40 ++#define R_ALPHA_TPREL16 41 ++ ++#define R_ALPHA_NUM 46 ++ ++ ++#define LITUSE_ALPHA_ADDR 0 ++#define LITUSE_ALPHA_BASE 1 ++#define LITUSE_ALPHA_BYTOFF 2 ++#define LITUSE_ALPHA_JSR 3 ++#define LITUSE_ALPHA_TLS_GD 4 ++#define LITUSE_ALPHA_TLS_LDM 5 ++ ++ ++#define DT_ALPHA_PLTRO (DT_LOPROC + 0) ++#define DT_ALPHA_NUM 1 ++ ++ ++ ++ ++#define EF_PPC_EMB 0x80000000 ++ ++ ++#define EF_PPC_RELOCATABLE 0x00010000 ++#define EF_PPC_RELOCATABLE_LIB 0x00008000 ++ ++ ++ ++#define R_PPC_NONE 0 ++#define R_PPC_ADDR32 1 ++#define R_PPC_ADDR24 2 ++#define R_PPC_ADDR16 3 ++#define R_PPC_ADDR16_LO 4 ++#define R_PPC_ADDR16_HI 5 ++#define R_PPC_ADDR16_HA 6 ++#define R_PPC_ADDR14 7 ++#define R_PPC_ADDR14_BRTAKEN 8 ++#define R_PPC_ADDR14_BRNTAKEN 9 ++#define R_PPC_REL24 10 ++#define R_PPC_REL14 11 ++#define R_PPC_REL14_BRTAKEN 12 ++#define R_PPC_REL14_BRNTAKEN 13 ++#define R_PPC_GOT16 14 ++#define R_PPC_GOT16_LO 15 ++#define R_PPC_GOT16_HI 16 ++#define R_PPC_GOT16_HA 17 ++#define R_PPC_PLTREL24 18 ++#define R_PPC_COPY 19 ++#define R_PPC_GLOB_DAT 20 ++#define R_PPC_JMP_SLOT 21 ++#define R_PPC_RELATIVE 22 ++#define R_PPC_LOCAL24PC 23 ++#define R_PPC_UADDR32 24 ++#define R_PPC_UADDR16 25 ++#define R_PPC_REL32 26 ++#define R_PPC_PLT32 27 ++#define R_PPC_PLTREL32 28 ++#define R_PPC_PLT16_LO 29 ++#define R_PPC_PLT16_HI 30 ++#define R_PPC_PLT16_HA 31 ++#define R_PPC_SDAREL16 32 ++#define R_PPC_SECTOFF 33 ++#define R_PPC_SECTOFF_LO 34 ++#define R_PPC_SECTOFF_HI 35 ++#define R_PPC_SECTOFF_HA 36 ++ ++ ++#define R_PPC_TLS 67 ++#define R_PPC_DTPMOD32 68 ++#define R_PPC_TPREL16 69 ++#define R_PPC_TPREL16_LO 70 ++#define R_PPC_TPREL16_HI 71 ++#define R_PPC_TPREL16_HA 72 ++#define R_PPC_TPREL32 73 ++#define R_PPC_DTPREL16 74 ++#define R_PPC_DTPREL16_LO 75 ++#define R_PPC_DTPREL16_HI 76 ++#define R_PPC_DTPREL16_HA 77 ++#define R_PPC_DTPREL32 78 ++#define R_PPC_GOT_TLSGD16 79 ++#define R_PPC_GOT_TLSGD16_LO 80 ++#define R_PPC_GOT_TLSGD16_HI 81 ++#define R_PPC_GOT_TLSGD16_HA 82 ++#define R_PPC_GOT_TLSLD16 83 ++#define R_PPC_GOT_TLSLD16_LO 84 ++#define R_PPC_GOT_TLSLD16_HI 85 ++#define R_PPC_GOT_TLSLD16_HA 86 ++#define R_PPC_GOT_TPREL16 87 ++#define R_PPC_GOT_TPREL16_LO 88 ++#define R_PPC_GOT_TPREL16_HI 89 ++#define R_PPC_GOT_TPREL16_HA 90 ++#define R_PPC_GOT_DTPREL16 91 ++#define R_PPC_GOT_DTPREL16_LO 92 ++#define R_PPC_GOT_DTPREL16_HI 93 ++#define R_PPC_GOT_DTPREL16_HA 94 ++ ++ ++ ++#define R_PPC_EMB_NADDR32 101 ++#define R_PPC_EMB_NADDR16 102 ++#define R_PPC_EMB_NADDR16_LO 103 ++#define R_PPC_EMB_NADDR16_HI 104 ++#define R_PPC_EMB_NADDR16_HA 105 ++#define R_PPC_EMB_SDAI16 106 ++#define R_PPC_EMB_SDA2I16 107 ++#define R_PPC_EMB_SDA2REL 108 ++#define R_PPC_EMB_SDA21 109 ++#define R_PPC_EMB_MRKREF 110 ++#define R_PPC_EMB_RELSEC16 111 ++#define R_PPC_EMB_RELST_LO 112 ++#define R_PPC_EMB_RELST_HI 113 ++#define R_PPC_EMB_RELST_HA 114 ++#define R_PPC_EMB_BIT_FLD 115 ++#define R_PPC_EMB_RELSDA 116 ++ ++ ++#define R_PPC_DIAB_SDA21_LO 180 ++#define R_PPC_DIAB_SDA21_HI 181 ++#define R_PPC_DIAB_SDA21_HA 182 ++#define R_PPC_DIAB_RELSDA_LO 183 ++#define R_PPC_DIAB_RELSDA_HI 184 ++#define R_PPC_DIAB_RELSDA_HA 185 ++ ++ ++#define R_PPC_IRELATIVE 248 ++ ++ ++#define R_PPC_REL16 249 ++#define R_PPC_REL16_LO 250 ++#define R_PPC_REL16_HI 251 ++#define R_PPC_REL16_HA 252 ++ ++ ++ ++#define R_PPC_TOC16 255 ++ ++ ++#define DT_PPC_GOT (DT_LOPROC + 0) ++#define DT_PPC_NUM 1 ++ ++ ++#define R_PPC64_NONE R_PPC_NONE ++#define R_PPC64_ADDR32 R_PPC_ADDR32 ++#define R_PPC64_ADDR24 R_PPC_ADDR24 ++#define R_PPC64_ADDR16 R_PPC_ADDR16 ++#define R_PPC64_ADDR16_LO R_PPC_ADDR16_LO ++#define R_PPC64_ADDR16_HI R_PPC_ADDR16_HI ++#define R_PPC64_ADDR16_HA R_PPC_ADDR16_HA ++#define R_PPC64_ADDR14 R_PPC_ADDR14 ++#define R_PPC64_ADDR14_BRTAKEN R_PPC_ADDR14_BRTAKEN ++#define R_PPC64_ADDR14_BRNTAKEN R_PPC_ADDR14_BRNTAKEN ++#define R_PPC64_REL24 R_PPC_REL24 ++#define R_PPC64_REL14 R_PPC_REL14 ++#define R_PPC64_REL14_BRTAKEN R_PPC_REL14_BRTAKEN ++#define R_PPC64_REL14_BRNTAKEN R_PPC_REL14_BRNTAKEN ++#define R_PPC64_GOT16 R_PPC_GOT16 ++#define R_PPC64_GOT16_LO R_PPC_GOT16_LO ++#define R_PPC64_GOT16_HI R_PPC_GOT16_HI ++#define R_PPC64_GOT16_HA R_PPC_GOT16_HA ++ ++#define R_PPC64_COPY R_PPC_COPY ++#define R_PPC64_GLOB_DAT R_PPC_GLOB_DAT ++#define R_PPC64_JMP_SLOT R_PPC_JMP_SLOT ++#define R_PPC64_RELATIVE R_PPC_RELATIVE ++ ++#define R_PPC64_UADDR32 R_PPC_UADDR32 ++#define R_PPC64_UADDR16 R_PPC_UADDR16 ++#define R_PPC64_REL32 R_PPC_REL32 ++#define R_PPC64_PLT32 R_PPC_PLT32 ++#define R_PPC64_PLTREL32 R_PPC_PLTREL32 ++#define R_PPC64_PLT16_LO R_PPC_PLT16_LO ++#define R_PPC64_PLT16_HI R_PPC_PLT16_HI ++#define R_PPC64_PLT16_HA R_PPC_PLT16_HA ++ ++#define R_PPC64_SECTOFF R_PPC_SECTOFF ++#define R_PPC64_SECTOFF_LO R_PPC_SECTOFF_LO ++#define R_PPC64_SECTOFF_HI R_PPC_SECTOFF_HI ++#define R_PPC64_SECTOFF_HA R_PPC_SECTOFF_HA ++#define R_PPC64_ADDR30 37 ++#define R_PPC64_ADDR64 38 ++#define R_PPC64_ADDR16_HIGHER 39 ++#define R_PPC64_ADDR16_HIGHERA 40 ++#define R_PPC64_ADDR16_HIGHEST 41 ++#define R_PPC64_ADDR16_HIGHESTA 42 ++#define R_PPC64_UADDR64 43 ++#define R_PPC64_REL64 44 ++#define R_PPC64_PLT64 45 ++#define R_PPC64_PLTREL64 46 ++#define R_PPC64_TOC16 47 ++#define R_PPC64_TOC16_LO 48 ++#define R_PPC64_TOC16_HI 49 ++#define R_PPC64_TOC16_HA 50 ++#define R_PPC64_TOC 51 ++#define R_PPC64_PLTGOT16 52 ++#define R_PPC64_PLTGOT16_LO 53 ++#define R_PPC64_PLTGOT16_HI 54 ++#define R_PPC64_PLTGOT16_HA 55 ++ ++#define R_PPC64_ADDR16_DS 56 ++#define R_PPC64_ADDR16_LO_DS 57 ++#define R_PPC64_GOT16_DS 58 ++#define R_PPC64_GOT16_LO_DS 59 ++#define R_PPC64_PLT16_LO_DS 60 ++#define R_PPC64_SECTOFF_DS 61 ++#define R_PPC64_SECTOFF_LO_DS 62 ++#define R_PPC64_TOC16_DS 63 ++#define R_PPC64_TOC16_LO_DS 64 ++#define R_PPC64_PLTGOT16_DS 65 ++#define R_PPC64_PLTGOT16_LO_DS 66 ++ ++ ++#define R_PPC64_TLS 67 ++#define R_PPC64_DTPMOD64 68 ++#define R_PPC64_TPREL16 69 ++#define R_PPC64_TPREL16_LO 70 ++#define R_PPC64_TPREL16_HI 71 ++#define R_PPC64_TPREL16_HA 72 ++#define R_PPC64_TPREL64 73 ++#define R_PPC64_DTPREL16 74 ++#define R_PPC64_DTPREL16_LO 75 ++#define R_PPC64_DTPREL16_HI 76 ++#define R_PPC64_DTPREL16_HA 77 ++#define R_PPC64_DTPREL64 78 ++#define R_PPC64_GOT_TLSGD16 79 ++#define R_PPC64_GOT_TLSGD16_LO 80 ++#define R_PPC64_GOT_TLSGD16_HI 81 ++#define R_PPC64_GOT_TLSGD16_HA 82 ++#define R_PPC64_GOT_TLSLD16 83 ++#define R_PPC64_GOT_TLSLD16_LO 84 ++#define R_PPC64_GOT_TLSLD16_HI 85 ++#define R_PPC64_GOT_TLSLD16_HA 86 ++#define R_PPC64_GOT_TPREL16_DS 87 ++#define R_PPC64_GOT_TPREL16_LO_DS 88 ++#define R_PPC64_GOT_TPREL16_HI 89 ++#define R_PPC64_GOT_TPREL16_HA 90 ++#define R_PPC64_GOT_DTPREL16_DS 91 ++#define R_PPC64_GOT_DTPREL16_LO_DS 92 ++#define R_PPC64_GOT_DTPREL16_HI 93 ++#define R_PPC64_GOT_DTPREL16_HA 94 ++#define R_PPC64_TPREL16_DS 95 ++#define R_PPC64_TPREL16_LO_DS 96 ++#define R_PPC64_TPREL16_HIGHER 97 ++#define R_PPC64_TPREL16_HIGHERA 98 ++#define R_PPC64_TPREL16_HIGHEST 99 ++#define R_PPC64_TPREL16_HIGHESTA 100 ++#define R_PPC64_DTPREL16_DS 101 ++#define R_PPC64_DTPREL16_LO_DS 102 ++#define R_PPC64_DTPREL16_HIGHER 103 ++#define R_PPC64_DTPREL16_HIGHERA 104 ++#define R_PPC64_DTPREL16_HIGHEST 105 ++#define R_PPC64_DTPREL16_HIGHESTA 106 ++ ++ ++#define R_PPC64_JMP_IREL 247 ++#define R_PPC64_IRELATIVE 248 ++#define R_PPC64_REL16 249 ++#define R_PPC64_REL16_LO 250 ++#define R_PPC64_REL16_HI 251 ++#define R_PPC64_REL16_HA 252 ++ ++ ++#define DT_PPC64_GLINK (DT_LOPROC + 0) ++#define DT_PPC64_OPD (DT_LOPROC + 1) ++#define DT_PPC64_OPDSZ (DT_LOPROC + 2) ++#define DT_PPC64_NUM 3 ++ ++ ++ ++ ++ ++#define EF_ARM_RELEXEC 0x01 ++#define EF_ARM_HASENTRY 0x02 ++#define EF_ARM_INTERWORK 0x04 ++#define EF_ARM_APCS_26 0x08 ++#define EF_ARM_APCS_FLOAT 0x10 ++#define EF_ARM_PIC 0x20 ++#define EF_ARM_ALIGN8 0x40 ++#define EF_ARM_NEW_ABI 0x80 ++#define EF_ARM_OLD_ABI 0x100 ++#define EF_ARM_SOFT_FLOAT 0x200 ++#define EF_ARM_VFP_FLOAT 0x400 ++#define EF_ARM_MAVERICK_FLOAT 0x800 ++ ++#define EF_ARM_ABI_FLOAT_SOFT 0x200 ++#define EF_ARM_ABI_FLOAT_HARD 0x400 ++ ++ ++#define EF_ARM_SYMSARESORTED 0x04 ++#define EF_ARM_DYNSYMSUSESEGIDX 0x08 ++#define EF_ARM_MAPSYMSFIRST 0x10 ++#define EF_ARM_EABIMASK 0XFF000000 ++ ++ ++#define EF_ARM_BE8 0x00800000 ++#define EF_ARM_LE8 0x00400000 ++ ++#define EF_ARM_EABI_VERSION(flags) ((flags) & EF_ARM_EABIMASK) ++#define EF_ARM_EABI_UNKNOWN 0x00000000 ++#define EF_ARM_EABI_VER1 0x01000000 ++#define EF_ARM_EABI_VER2 0x02000000 ++#define EF_ARM_EABI_VER3 0x03000000 ++#define EF_ARM_EABI_VER4 0x04000000 ++#define EF_ARM_EABI_VER5 0x05000000 ++ ++ ++#define STT_ARM_TFUNC STT_LOPROC ++#define STT_ARM_16BIT STT_HIPROC ++ ++ ++#define SHF_ARM_ENTRYSECT 0x10000000 ++#define SHF_ARM_COMDEF 0x80000000 ++ ++ ++ ++#define PF_ARM_SB 0x10000000 ++ ++#define PF_ARM_PI 0x20000000 ++#define PF_ARM_ABS 0x40000000 ++ ++ ++#define PT_ARM_EXIDX (PT_LOPROC + 1) ++ ++ ++#define SHT_ARM_EXIDX (SHT_LOPROC + 1) ++#define SHT_ARM_PREEMPTMAP (SHT_LOPROC + 2) ++#define SHT_ARM_ATTRIBUTES (SHT_LOPROC + 3) ++ ++ ++#define R_AARCH64_NONE 0 ++#define R_AARCH64_ABS64 257 ++#define R_AARCH64_ABS32 258 ++#define R_AARCH64_COPY 1024 ++#define R_AARCH64_GLOB_DAT 1025 ++#define R_AARCH64_JUMP_SLOT 1026 ++#define R_AARCH64_RELATIVE 1027 ++#define R_AARCH64_TLS_DTPMOD64 1028 ++#define R_AARCH64_TLS_DTPREL64 1029 ++#define R_AARCH64_TLS_TPREL64 1030 ++#define R_AARCH64_TLSDESC 1031 ++ ++ ++#define R_ARM_NONE 0 ++#define R_ARM_PC24 1 ++#define R_ARM_ABS32 2 ++#define R_ARM_REL32 3 ++#define R_ARM_PC13 4 ++#define R_ARM_ABS16 5 ++#define R_ARM_ABS12 6 ++#define R_ARM_THM_ABS5 7 ++#define R_ARM_ABS8 8 ++#define R_ARM_SBREL32 9 ++#define R_ARM_THM_PC22 10 ++#define R_ARM_THM_PC8 11 ++#define R_ARM_AMP_VCALL9 12 ++#define R_ARM_TLS_DESC 13 ++#define R_ARM_THM_SWI8 14 ++#define R_ARM_XPC25 15 ++#define R_ARM_THM_XPC22 16 ++#define R_ARM_TLS_DTPMOD32 17 ++#define R_ARM_TLS_DTPOFF32 18 ++#define R_ARM_TLS_TPOFF32 19 ++#define R_ARM_COPY 20 ++#define R_ARM_GLOB_DAT 21 ++#define R_ARM_JUMP_SLOT 22 ++#define R_ARM_RELATIVE 23 ++#define R_ARM_GOTOFF 24 ++#define R_ARM_GOTPC 25 ++#define R_ARM_GOT32 26 ++#define R_ARM_PLT32 27 ++#define R_ARM_CALL 28 ++#define R_ARM_JUMP24 29 ++#define R_ARM_THM_JUMP24 30 ++#define R_ARM_BASE_ABS 31 ++#define R_ARM_ALU_PCREL_7_0 32 ++#define R_ARM_ALU_PCREL_15_8 33 ++#define R_ARM_ALU_PCREL_23_15 34 ++#define R_ARM_LDR_SBREL_11_0 35 ++#define R_ARM_ALU_SBREL_19_12 36 ++#define R_ARM_ALU_SBREL_27_20 37 ++#define R_ARM_TARGET1 38 ++#define R_ARM_SBREL31 39 ++#define R_ARM_V4BX 40 ++#define R_ARM_TARGET2 41 ++#define R_ARM_PREL31 42 ++#define R_ARM_MOVW_ABS_NC 43 ++#define R_ARM_MOVT_ABS 44 ++#define R_ARM_MOVW_PREL_NC 45 ++#define R_ARM_MOVT_PREL 46 ++#define R_ARM_THM_MOVW_ABS_NC 47 ++#define R_ARM_THM_MOVT_ABS 48 ++#define R_ARM_THM_MOVW_PREL_NC 49 ++#define R_ARM_THM_MOVT_PREL 50 ++#define R_ARM_THM_JUMP19 51 ++#define R_ARM_THM_JUMP6 52 ++#define R_ARM_THM_ALU_PREL_11_0 53 ++#define R_ARM_THM_PC12 54 ++#define R_ARM_ABS32_NOI 55 ++#define R_ARM_REL32_NOI 56 ++#define R_ARM_ALU_PC_G0_NC 57 ++#define R_ARM_ALU_PC_G0 58 ++#define R_ARM_ALU_PC_G1_NC 59 ++#define R_ARM_ALU_PC_G1 60 ++#define R_ARM_ALU_PC_G2 61 ++#define R_ARM_LDR_PC_G1 62 ++#define R_ARM_LDR_PC_G2 63 ++#define R_ARM_LDRS_PC_G0 64 ++#define R_ARM_LDRS_PC_G1 65 ++#define R_ARM_LDRS_PC_G2 66 ++#define R_ARM_LDC_PC_G0 67 ++#define R_ARM_LDC_PC_G1 68 ++#define R_ARM_LDC_PC_G2 69 ++#define R_ARM_ALU_SB_G0_NC 70 ++#define R_ARM_ALU_SB_G0 71 ++#define R_ARM_ALU_SB_G1_NC 72 ++#define R_ARM_ALU_SB_G1 73 ++#define R_ARM_ALU_SB_G2 74 ++#define R_ARM_LDR_SB_G0 75 ++#define R_ARM_LDR_SB_G1 76 ++#define R_ARM_LDR_SB_G2 77 ++#define R_ARM_LDRS_SB_G0 78 ++#define R_ARM_LDRS_SB_G1 79 ++#define R_ARM_LDRS_SB_G2 80 ++#define R_ARM_LDC_SB_G0 81 ++#define R_ARM_LDC_SB_G1 82 ++#define R_ARM_LDC_SB_G2 83 ++#define R_ARM_MOVW_BREL_NC 84 ++#define R_ARM_MOVT_BREL 85 ++#define R_ARM_MOVW_BREL 86 ++#define R_ARM_THM_MOVW_BREL_NC 87 ++#define R_ARM_THM_MOVT_BREL 88 ++#define R_ARM_THM_MOVW_BREL 89 ++#define R_ARM_TLS_GOTDESC 90 ++#define R_ARM_TLS_CALL 91 ++#define R_ARM_TLS_DESCSEQ 92 ++#define R_ARM_THM_TLS_CALL 93 ++#define R_ARM_PLT32_ABS 94 ++#define R_ARM_GOT_ABS 95 ++#define R_ARM_GOT_PREL 96 ++#define R_ARM_GOT_BREL12 97 ++#define R_ARM_GOTOFF12 98 ++#define R_ARM_GOTRELAX 99 ++#define R_ARM_GNU_VTENTRY 100 ++#define R_ARM_GNU_VTINHERIT 101 ++#define R_ARM_THM_PC11 102 ++#define R_ARM_THM_PC9 103 ++#define R_ARM_TLS_GD32 104 ++ ++#define R_ARM_TLS_LDM32 105 ++ ++#define R_ARM_TLS_LDO32 106 ++ ++#define R_ARM_TLS_IE32 107 ++ ++#define R_ARM_TLS_LE32 108 ++#define R_ARM_TLS_LDO12 109 ++#define R_ARM_TLS_LE12 110 ++#define R_ARM_TLS_IE12GP 111 ++#define R_ARM_ME_TOO 128 ++#define R_ARM_THM_TLS_DESCSEQ 129 ++#define R_ARM_THM_TLS_DESCSEQ16 129 ++#define R_ARM_THM_TLS_DESCSEQ32 130 ++#define R_ARM_THM_GOT_BREL12 131 ++#define R_ARM_IRELATIVE 160 ++#define R_ARM_RXPC25 249 ++#define R_ARM_RSBREL32 250 ++#define R_ARM_THM_RPC22 251 ++#define R_ARM_RREL32 252 ++#define R_ARM_RABS22 253 ++#define R_ARM_RPC24 254 ++#define R_ARM_RBASE 255 ++ ++#define R_ARM_NUM 256 ++ ++ ++ ++ ++#define EF_IA_64_MASKOS 0x0000000f ++#define EF_IA_64_ABI64 0x00000010 ++#define EF_IA_64_ARCH 0xff000000 ++ ++ ++#define PT_IA_64_ARCHEXT (PT_LOPROC + 0) ++#define PT_IA_64_UNWIND (PT_LOPROC + 1) ++#define PT_IA_64_HP_OPT_ANOT (PT_LOOS + 0x12) ++#define PT_IA_64_HP_HSL_ANOT (PT_LOOS + 0x13) ++#define PT_IA_64_HP_STACK (PT_LOOS + 0x14) ++ ++ ++#define PF_IA_64_NORECOV 0x80000000 ++ ++ ++#define SHT_IA_64_EXT (SHT_LOPROC + 0) ++#define SHT_IA_64_UNWIND (SHT_LOPROC + 1) ++ ++ ++#define SHF_IA_64_SHORT 0x10000000 ++#define SHF_IA_64_NORECOV 0x20000000 ++ ++ ++#define DT_IA_64_PLT_RESERVE (DT_LOPROC + 0) ++#define DT_IA_64_NUM 1 ++ ++ ++#define R_IA64_NONE 0x00 ++#define R_IA64_IMM14 0x21 ++#define R_IA64_IMM22 0x22 ++#define R_IA64_IMM64 0x23 ++#define R_IA64_DIR32MSB 0x24 ++#define R_IA64_DIR32LSB 0x25 ++#define R_IA64_DIR64MSB 0x26 ++#define R_IA64_DIR64LSB 0x27 ++#define R_IA64_GPREL22 0x2a ++#define R_IA64_GPREL64I 0x2b ++#define R_IA64_GPREL32MSB 0x2c ++#define R_IA64_GPREL32LSB 0x2d ++#define R_IA64_GPREL64MSB 0x2e ++#define R_IA64_GPREL64LSB 0x2f ++#define R_IA64_LTOFF22 0x32 ++#define R_IA64_LTOFF64I 0x33 ++#define R_IA64_PLTOFF22 0x3a ++#define R_IA64_PLTOFF64I 0x3b ++#define R_IA64_PLTOFF64MSB 0x3e ++#define R_IA64_PLTOFF64LSB 0x3f ++#define R_IA64_FPTR64I 0x43 ++#define R_IA64_FPTR32MSB 0x44 ++#define R_IA64_FPTR32LSB 0x45 ++#define R_IA64_FPTR64MSB 0x46 ++#define R_IA64_FPTR64LSB 0x47 ++#define R_IA64_PCREL60B 0x48 ++#define R_IA64_PCREL21B 0x49 ++#define R_IA64_PCREL21M 0x4a ++#define R_IA64_PCREL21F 0x4b ++#define R_IA64_PCREL32MSB 0x4c ++#define R_IA64_PCREL32LSB 0x4d ++#define R_IA64_PCREL64MSB 0x4e ++#define R_IA64_PCREL64LSB 0x4f ++#define R_IA64_LTOFF_FPTR22 0x52 ++#define R_IA64_LTOFF_FPTR64I 0x53 ++#define R_IA64_LTOFF_FPTR32MSB 0x54 ++#define R_IA64_LTOFF_FPTR32LSB 0x55 ++#define R_IA64_LTOFF_FPTR64MSB 0x56 ++#define R_IA64_LTOFF_FPTR64LSB 0x57 ++#define R_IA64_SEGREL32MSB 0x5c ++#define R_IA64_SEGREL32LSB 0x5d ++#define R_IA64_SEGREL64MSB 0x5e ++#define R_IA64_SEGREL64LSB 0x5f ++#define R_IA64_SECREL32MSB 0x64 ++#define R_IA64_SECREL32LSB 0x65 ++#define R_IA64_SECREL64MSB 0x66 ++#define R_IA64_SECREL64LSB 0x67 ++#define R_IA64_REL32MSB 0x6c ++#define R_IA64_REL32LSB 0x6d ++#define R_IA64_REL64MSB 0x6e ++#define R_IA64_REL64LSB 0x6f ++#define R_IA64_LTV32MSB 0x74 ++#define R_IA64_LTV32LSB 0x75 ++#define R_IA64_LTV64MSB 0x76 ++#define R_IA64_LTV64LSB 0x77 ++#define R_IA64_PCREL21BI 0x79 ++#define R_IA64_PCREL22 0x7a ++#define R_IA64_PCREL64I 0x7b ++#define R_IA64_IPLTMSB 0x80 ++#define R_IA64_IPLTLSB 0x81 ++#define R_IA64_COPY 0x84 ++#define R_IA64_SUB 0x85 ++#define R_IA64_LTOFF22X 0x86 ++#define R_IA64_LDXMOV 0x87 ++#define R_IA64_TPREL14 0x91 ++#define R_IA64_TPREL22 0x92 ++#define R_IA64_TPREL64I 0x93 ++#define R_IA64_TPREL64MSB 0x96 ++#define R_IA64_TPREL64LSB 0x97 ++#define R_IA64_LTOFF_TPREL22 0x9a ++#define R_IA64_DTPMOD64MSB 0xa6 ++#define R_IA64_DTPMOD64LSB 0xa7 ++#define R_IA64_LTOFF_DTPMOD22 0xaa ++#define R_IA64_DTPREL14 0xb1 ++#define R_IA64_DTPREL22 0xb2 ++#define R_IA64_DTPREL64I 0xb3 ++#define R_IA64_DTPREL32MSB 0xb4 ++#define R_IA64_DTPREL32LSB 0xb5 ++#define R_IA64_DTPREL64MSB 0xb6 ++#define R_IA64_DTPREL64LSB 0xb7 ++#define R_IA64_LTOFF_DTPREL22 0xba ++ ++ ++ ++ ++#define R_SH_NONE 0 ++#define R_SH_DIR32 1 ++#define R_SH_REL32 2 ++#define R_SH_DIR8WPN 3 ++#define R_SH_IND12W 4 ++#define R_SH_DIR8WPL 5 ++#define R_SH_DIR8WPZ 6 ++#define R_SH_DIR8BP 7 ++#define R_SH_DIR8W 8 ++#define R_SH_DIR8L 9 ++#define R_SH_SWITCH16 25 ++#define R_SH_SWITCH32 26 ++#define R_SH_USES 27 ++#define R_SH_COUNT 28 ++#define R_SH_ALIGN 29 ++#define R_SH_CODE 30 ++#define R_SH_DATA 31 ++#define R_SH_LABEL 32 ++#define R_SH_SWITCH8 33 ++#define R_SH_GNU_VTINHERIT 34 ++#define R_SH_GNU_VTENTRY 35 ++#define R_SH_TLS_GD_32 144 ++#define R_SH_TLS_LD_32 145 ++#define R_SH_TLS_LDO_32 146 ++#define R_SH_TLS_IE_32 147 ++#define R_SH_TLS_LE_32 148 ++#define R_SH_TLS_DTPMOD32 149 ++#define R_SH_TLS_DTPOFF32 150 ++#define R_SH_TLS_TPOFF32 151 ++#define R_SH_GOT32 160 ++#define R_SH_PLT32 161 ++#define R_SH_COPY 162 ++#define R_SH_GLOB_DAT 163 ++#define R_SH_JMP_SLOT 164 ++#define R_SH_RELATIVE 165 ++#define R_SH_GOTOFF 166 ++#define R_SH_GOTPC 167 ++ ++#define R_SH_NUM 256 ++ ++ ++ ++#define R_390_NONE 0 ++#define R_390_8 1 ++#define R_390_12 2 ++#define R_390_16 3 ++#define R_390_32 4 ++#define R_390_PC32 5 ++#define R_390_GOT12 6 ++#define R_390_GOT32 7 ++#define R_390_PLT32 8 ++#define R_390_COPY 9 ++#define R_390_GLOB_DAT 10 ++#define R_390_JMP_SLOT 11 ++#define R_390_RELATIVE 12 ++#define R_390_GOTOFF32 13 ++#define R_390_GOTPC 14 ++#define R_390_GOT16 15 ++#define R_390_PC16 16 ++#define R_390_PC16DBL 17 ++#define R_390_PLT16DBL 18 ++#define R_390_PC32DBL 19 ++#define R_390_PLT32DBL 20 ++#define R_390_GOTPCDBL 21 ++#define R_390_64 22 ++#define R_390_PC64 23 ++#define R_390_GOT64 24 ++#define R_390_PLT64 25 ++#define R_390_GOTENT 26 ++#define R_390_GOTOFF16 27 ++#define R_390_GOTOFF64 28 ++#define R_390_GOTPLT12 29 ++#define R_390_GOTPLT16 30 ++#define R_390_GOTPLT32 31 ++#define R_390_GOTPLT64 32 ++#define R_390_GOTPLTENT 33 ++#define R_390_PLTOFF16 34 ++#define R_390_PLTOFF32 35 ++#define R_390_PLTOFF64 36 ++#define R_390_TLS_LOAD 37 ++#define R_390_TLS_GDCALL 38 ++ ++#define R_390_TLS_LDCALL 39 ++ ++#define R_390_TLS_GD32 40 ++ ++#define R_390_TLS_GD64 41 ++ ++#define R_390_TLS_GOTIE12 42 ++ ++#define R_390_TLS_GOTIE32 43 ++ ++#define R_390_TLS_GOTIE64 44 ++ ++#define R_390_TLS_LDM32 45 ++ ++#define R_390_TLS_LDM64 46 ++ ++#define R_390_TLS_IE32 47 ++ ++#define R_390_TLS_IE64 48 ++ ++#define R_390_TLS_IEENT 49 ++ ++#define R_390_TLS_LE32 50 ++ ++#define R_390_TLS_LE64 51 ++ ++#define R_390_TLS_LDO32 52 ++ ++#define R_390_TLS_LDO64 53 ++ ++#define R_390_TLS_DTPMOD 54 ++#define R_390_TLS_DTPOFF 55 ++#define R_390_TLS_TPOFF 56 ++ ++#define R_390_20 57 ++#define R_390_GOT20 58 ++#define R_390_GOTPLT20 59 ++#define R_390_TLS_GOTIE20 60 ++ ++ ++#define R_390_NUM 61 ++ ++ ++ ++#define R_CRIS_NONE 0 ++#define R_CRIS_8 1 ++#define R_CRIS_16 2 ++#define R_CRIS_32 3 ++#define R_CRIS_8_PCREL 4 ++#define R_CRIS_16_PCREL 5 ++#define R_CRIS_32_PCREL 6 ++#define R_CRIS_GNU_VTINHERIT 7 ++#define R_CRIS_GNU_VTENTRY 8 ++#define R_CRIS_COPY 9 ++#define R_CRIS_GLOB_DAT 10 ++#define R_CRIS_JUMP_SLOT 11 ++#define R_CRIS_RELATIVE 12 ++#define R_CRIS_16_GOT 13 ++#define R_CRIS_32_GOT 14 ++#define R_CRIS_16_GOTPLT 15 ++#define R_CRIS_32_GOTPLT 16 ++#define R_CRIS_32_GOTREL 17 ++#define R_CRIS_32_PLT_GOTREL 18 ++#define R_CRIS_32_PLT_PCREL 19 ++ ++#define R_CRIS_NUM 20 ++ ++ ++ ++#define R_X86_64_NONE 0 ++#define R_X86_64_64 1 ++#define R_X86_64_PC32 2 ++#define R_X86_64_GOT32 3 ++#define R_X86_64_PLT32 4 ++#define R_X86_64_COPY 5 ++#define R_X86_64_GLOB_DAT 6 ++#define R_X86_64_JUMP_SLOT 7 ++#define R_X86_64_RELATIVE 8 ++#define R_X86_64_GOTPCREL 9 ++ ++#define R_X86_64_32 10 ++#define R_X86_64_32S 11 ++#define R_X86_64_16 12 ++#define R_X86_64_PC16 13 ++#define R_X86_64_8 14 ++#define R_X86_64_PC8 15 ++#define R_X86_64_DTPMOD64 16 ++#define R_X86_64_DTPOFF64 17 ++#define R_X86_64_TPOFF64 18 ++#define R_X86_64_TLSGD 19 ++ ++#define R_X86_64_TLSLD 20 ++ ++#define R_X86_64_DTPOFF32 21 ++#define R_X86_64_GOTTPOFF 22 ++ ++#define R_X86_64_TPOFF32 23 ++#define R_X86_64_PC64 24 ++#define R_X86_64_GOTOFF64 25 ++#define R_X86_64_GOTPC32 26 ++#define R_X86_64_GOT64 27 ++#define R_X86_64_GOTPCREL64 28 ++#define R_X86_64_GOTPC64 29 ++#define R_X86_64_GOTPLT64 30 ++#define R_X86_64_PLTOFF64 31 ++#define R_X86_64_SIZE32 32 ++#define R_X86_64_SIZE64 33 ++ ++#define R_X86_64_GOTPC32_TLSDESC 34 ++#define R_X86_64_TLSDESC_CALL 35 ++ ++#define R_X86_64_TLSDESC 36 ++#define R_X86_64_IRELATIVE 37 ++#define R_X86_64_RELATIVE64 38 ++#define R_X86_64_NUM 39 ++ ++ ++ ++#define R_MN10300_NONE 0 ++#define R_MN10300_32 1 ++#define R_MN10300_16 2 ++#define R_MN10300_8 3 ++#define R_MN10300_PCREL32 4 ++#define R_MN10300_PCREL16 5 ++#define R_MN10300_PCREL8 6 ++#define R_MN10300_GNU_VTINHERIT 7 ++#define R_MN10300_GNU_VTENTRY 8 ++#define R_MN10300_24 9 ++#define R_MN10300_GOTPC32 10 ++#define R_MN10300_GOTPC16 11 ++#define R_MN10300_GOTOFF32 12 ++#define R_MN10300_GOTOFF24 13 ++#define R_MN10300_GOTOFF16 14 ++#define R_MN10300_PLT32 15 ++#define R_MN10300_PLT16 16 ++#define R_MN10300_GOT32 17 ++#define R_MN10300_GOT24 18 ++#define R_MN10300_GOT16 19 ++#define R_MN10300_COPY 20 ++#define R_MN10300_GLOB_DAT 21 ++#define R_MN10300_JMP_SLOT 22 ++#define R_MN10300_RELATIVE 23 ++ ++#define R_MN10300_NUM 24 ++ ++ ++ ++#define R_M32R_NONE 0 ++#define R_M32R_16 1 ++#define R_M32R_32 2 ++#define R_M32R_24 3 ++#define R_M32R_10_PCREL 4 ++#define R_M32R_18_PCREL 5 ++#define R_M32R_26_PCREL 6 ++#define R_M32R_HI16_ULO 7 ++#define R_M32R_HI16_SLO 8 ++#define R_M32R_LO16 9 ++#define R_M32R_SDA16 10 ++#define R_M32R_GNU_VTINHERIT 11 ++#define R_M32R_GNU_VTENTRY 12 ++ ++#define R_M32R_16_RELA 33 ++#define R_M32R_32_RELA 34 ++#define R_M32R_24_RELA 35 ++#define R_M32R_10_PCREL_RELA 36 ++#define R_M32R_18_PCREL_RELA 37 ++#define R_M32R_26_PCREL_RELA 38 ++#define R_M32R_HI16_ULO_RELA 39 ++#define R_M32R_HI16_SLO_RELA 40 ++#define R_M32R_LO16_RELA 41 ++#define R_M32R_SDA16_RELA 42 ++#define R_M32R_RELA_GNU_VTINHERIT 43 ++#define R_M32R_RELA_GNU_VTENTRY 44 ++#define R_M32R_REL32 45 ++ ++#define R_M32R_GOT24 48 ++#define R_M32R_26_PLTREL 49 ++#define R_M32R_COPY 50 ++#define R_M32R_GLOB_DAT 51 ++#define R_M32R_JMP_SLOT 52 ++#define R_M32R_RELATIVE 53 ++#define R_M32R_GOTOFF 54 ++#define R_M32R_GOTPC24 55 ++#define R_M32R_GOT16_HI_ULO 56 ++ ++#define R_M32R_GOT16_HI_SLO 57 ++ ++#define R_M32R_GOT16_LO 58 ++#define R_M32R_GOTPC_HI_ULO 59 ++ ++#define R_M32R_GOTPC_HI_SLO 60 ++ ++#define R_M32R_GOTPC_LO 61 ++ ++#define R_M32R_GOTOFF_HI_ULO 62 ++ ++#define R_M32R_GOTOFF_HI_SLO 63 ++ ++#define R_M32R_GOTOFF_LO 64 ++#define R_M32R_NUM 256 ++ ++#define R_MICROBLAZE_NONE 0 ++#define R_MICROBLAZE_32 1 ++#define R_MICROBLAZE_32_PCREL 2 ++#define R_MICROBLAZE_64_PCREL 3 ++#define R_MICROBLAZE_32_PCREL_LO 4 ++#define R_MICROBLAZE_64 5 ++#define R_MICROBLAZE_32_LO 6 ++#define R_MICROBLAZE_SRO32 7 ++#define R_MICROBLAZE_SRW32 8 ++#define R_MICROBLAZE_64_NONE 9 ++#define R_MICROBLAZE_32_SYM_OP_SYM 10 ++#define R_MICROBLAZE_GNU_VTINHERIT 11 ++#define R_MICROBLAZE_GNU_VTENTRY 12 ++#define R_MICROBLAZE_GOTPC_64 13 ++#define R_MICROBLAZE_GOT_64 14 ++#define R_MICROBLAZE_PLT_64 15 ++#define R_MICROBLAZE_REL 16 ++#define R_MICROBLAZE_JUMP_SLOT 17 ++#define R_MICROBLAZE_GLOB_DAT 18 ++#define R_MICROBLAZE_GOTOFF_64 19 ++#define R_MICROBLAZE_GOTOFF_32 20 ++#define R_MICROBLAZE_COPY 21 ++#define R_MICROBLAZE_TLS 22 ++#define R_MICROBLAZE_TLSGD 23 ++#define R_MICROBLAZE_TLSLD 24 ++#define R_MICROBLAZE_TLSDTPMOD32 25 ++#define R_MICROBLAZE_TLSDTPREL32 26 ++#define R_MICROBLAZE_TLSDTPREL64 27 ++#define R_MICROBLAZE_TLSGOTTPREL32 28 ++#define R_MICROBLAZE_TLSTPREL32 29 ++ ++#ifdef __cplusplus ++} ++#endif ++ ++ ++#endif diff --git a/target/linux/patches/3.14.40/sgidefs.patch b/target/linux/patches/3.14.40/sgidefs.patch new file mode 100644 index 000000000..f00a284d9 --- /dev/null +++ b/target/linux/patches/3.14.40/sgidefs.patch @@ -0,0 +1,18 @@ +diff -Nur linux-3.11.5.orig/arch/mips/include/uapi/asm/sgidefs.h linux-3.11.5/arch/mips/include/uapi/asm/sgidefs.h +--- linux-3.11.5.orig/arch/mips/include/uapi/asm/sgidefs.h 2013-10-14 03:14:45.000000000 +0200 ++++ linux-3.11.5/arch/mips/include/uapi/asm/sgidefs.h 2013-11-08 22:01:28.000000000 +0100 +@@ -11,14 +11,6 @@ + #define __ASM_SGIDEFS_H + + /* +- * Using a Linux compiler for building Linux seems logic but not to +- * everybody. +- */ +-#ifndef __linux__ +-#error Use a Linux compiler or give up. +-#endif +- +-/* + * Definitions for the ISA levels + * + * With the introduction of MIPS32 / MIPS64 instruction sets definitions diff --git a/target/linux/patches/3.14.40/sortext.patch b/target/linux/patches/3.14.40/sortext.patch new file mode 100644 index 000000000..8fd4e1d6b --- /dev/null +++ b/target/linux/patches/3.14.40/sortext.patch @@ -0,0 +1,33 @@ +diff -Nur linux-3.12.6.orig/arch/arm/Kconfig linux-3.12.6/arch/arm/Kconfig +--- linux-3.12.6.orig/arch/arm/Kconfig 2013-12-20 16:51:33.000000000 +0100 ++++ linux-3.12.6/arch/arm/Kconfig 2013-12-28 19:29:33.000000000 +0100 +@@ -6,7 +6,6 @@ + select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST + select ARCH_HAVE_CUSTOM_GPIO_H + select ARCH_WANT_IPC_PARSE_VERSION +- select BUILDTIME_EXTABLE_SORT if MMU + select CLONE_BACKWARDS + select CPU_PM if (SUSPEND || CPU_IDLE) + select DCACHE_WORD_ACCESS if (CPU_V6 || CPU_V6K || CPU_V7) && !CPU_BIG_ENDIAN && MMU +diff -Nur linux-3.12.6.orig/arch/mips/Kconfig linux-3.12.6/arch/mips/Kconfig +--- linux-3.12.6.orig/arch/mips/Kconfig 2013-12-20 16:51:33.000000000 +0100 ++++ linux-3.12.6/arch/mips/Kconfig 2013-12-28 19:30:06.000000000 +0100 +@@ -35,7 +35,6 @@ + select HAVE_MEMBLOCK_NODE_MAP + select ARCH_DISCARD_MEMBLOCK + select GENERIC_SMP_IDLE_THREAD +- select BUILDTIME_EXTABLE_SORT + select GENERIC_CLOCKEVENTS + select GENERIC_CMOS_UPDATE + select HAVE_MOD_ARCH_SPECIFIC +diff -Nur linux-3.12.6.orig/arch/x86/Kconfig linux-3.12.6/arch/x86/Kconfig +--- linux-3.12.6.orig/arch/x86/Kconfig 2013-12-20 16:51:33.000000000 +0100 ++++ linux-3.12.6/arch/x86/Kconfig 2013-12-28 19:29:50.000000000 +0100 +@@ -100,7 +100,6 @@ + select GENERIC_SMP_IDLE_THREAD + select ARCH_WANT_IPC_PARSE_VERSION if X86_32 + select HAVE_ARCH_SECCOMP_FILTER +- select BUILDTIME_EXTABLE_SORT + select GENERIC_CMOS_UPDATE + select HAVE_ARCH_SOFT_DIRTY + select CLOCKSOURCE_WATCHDOG diff --git a/target/linux/patches/3.14.40/startup.patch b/target/linux/patches/3.14.40/startup.patch new file mode 100644 index 000000000..d396b75e4 --- /dev/null +++ b/target/linux/patches/3.14.40/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; diff --git a/target/linux/patches/3.14.40/wlan-cf.patch b/target/linux/patches/3.14.40/wlan-cf.patch new file mode 100644 index 000000000..fc20759e2 --- /dev/null +++ b/target/linux/patches/3.14.40/wlan-cf.patch @@ -0,0 +1,11 @@ +diff -Nur linux-2.6.39.orig/drivers/net/wireless/hostap/hostap_cs.c linux-2.6.39/drivers/net/wireless/hostap/hostap_cs.c +--- linux-2.6.39.orig/drivers/net/wireless/hostap/hostap_cs.c 2011-05-19 06:06:34.000000000 +0200 ++++ linux-2.6.39/drivers/net/wireless/hostap/hostap_cs.c 2011-09-12 02:46:26.987984145 +0200 +@@ -623,6 +623,7 @@ + static struct pcmcia_device_id hostap_cs_ids[] = { + PCMCIA_DEVICE_MANF_CARD(0x000b, 0x7100), + PCMCIA_DEVICE_MANF_CARD(0x000b, 0x7300), ++ PCMCIA_DEVICE_MANF_CARD(0x0004, 0x2003), + PCMCIA_DEVICE_MANF_CARD(0x0101, 0x0777), + PCMCIA_DEVICE_MANF_CARD(0x0126, 0x8000), + PCMCIA_DEVICE_MANF_CARD(0x0138, 0x0002), diff --git a/target/linux/patches/3.14.40/xargs.patch b/target/linux/patches/3.14.40/xargs.patch new file mode 100644 index 000000000..2c7b3df59 --- /dev/null +++ b/target/linux/patches/3.14.40/xargs.patch @@ -0,0 +1,12 @@ +diff -Nur linux-3.12.6.orig/scripts/Makefile.modpost linux-3.12.6/scripts/Makefile.modpost +--- linux-3.12.6.orig/scripts/Makefile.modpost 2013-12-20 16:51:33.000000000 +0100 ++++ linux-3.12.6/scripts/Makefile.modpost 2014-01-25 14:55:33.000000000 +0100 +@@ -60,7 +60,7 @@ + modulesymfile := $(firstword $(KBUILD_EXTMOD))/Module.symvers + + # Step 1), find all modules listed in $(MODVERDIR)/ +-MODLISTCMD := find $(MODVERDIR) -name '*.mod' | xargs -r grep -h '\.ko$$' | sort -u ++MODLISTCMD := find $(MODVERDIR) -name '*.mod' | xargs grep -h '\.ko$$' | sort -u + __modules := $(shell $(MODLISTCMD)) + modules := $(patsubst %.o,%.ko, $(wildcard $(__modules:.ko=.o))) + diff --git a/target/linux/patches/3.14.40/yaffs2.patch b/target/linux/patches/3.14.40/yaffs2.patch new file mode 100644 index 000000000..f075aa658 --- /dev/null +++ b/target/linux/patches/3.14.40/yaffs2.patch @@ -0,0 +1,16547 @@ +diff -Nur linux-3.14.4.orig/fs/Kconfig linux-3.14.4/fs/Kconfig +--- linux-3.14.4.orig/fs/Kconfig 2014-05-13 13:33:14.000000000 +0200 ++++ linux-3.14.4/fs/Kconfig 2014-05-17 02:22:30.000000000 +0200 +@@ -189,6 +189,7 @@ + source "fs/befs/Kconfig" + source "fs/bfs/Kconfig" + source "fs/efs/Kconfig" ++source "fs/yaffs2/Kconfig" + source "fs/jffs2/Kconfig" + # UBIFS File system configuration + source "fs/ubifs/Kconfig" +diff -Nur linux-3.14.4.orig/fs/Makefile linux-3.14.4/fs/Makefile +--- linux-3.14.4.orig/fs/Makefile 2014-05-13 13:33:14.000000000 +0200 ++++ linux-3.14.4/fs/Makefile 2014-05-17 02:22:30.000000000 +0200 +@@ -125,3 +125,4 @@ + obj-$(CONFIG_CEPH_FS) += ceph/ + obj-$(CONFIG_PSTORE) += pstore/ + obj-$(CONFIG_EFIVAR_FS) += efivarfs/ ++obj-$(CONFIG_YAFFS_FS) += yaffs2/ +diff -Nur linux-3.14.4.orig/fs/yaffs2/Kconfig linux-3.14.4/fs/yaffs2/Kconfig +--- linux-3.14.4.orig/fs/yaffs2/Kconfig 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/Kconfig 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,171 @@ ++# ++# yaffs file system configurations ++# ++ ++config YAFFS_FS ++ tristate "yaffs2 file system support" ++ default n ++ depends on MTD_BLOCK ++ select YAFFS_YAFFS1 ++ select YAFFS_YAFFS2 ++ help ++ yaffs2, or Yet Another Flash File System, is a file system ++ optimised for NAND Flash chips. ++ ++ To compile the yaffs2 file system support as a module, choose M ++ here: the module will be called yaffs2. ++ ++ If unsure, say N. ++ ++ Further information on yaffs2 is available at ++ <http://www.aleph1.co.uk/yaffs/>. ++ ++config YAFFS_YAFFS1 ++ bool "512 byte / page devices" ++ depends on YAFFS_FS ++ default y ++ help ++ Enable yaffs1 support -- yaffs for 512 byte / page devices ++ ++ Not needed for 2K-page devices. ++ ++ If unsure, say Y. ++ ++config YAFFS_9BYTE_TAGS ++ bool "Use older-style on-NAND data format with pageStatus byte" ++ depends on YAFFS_YAFFS1 ++ default n ++ help ++ ++ Older-style on-NAND data format has a "pageStatus" byte to record ++ chunk/page state. This byte is zero when the page is discarded. ++ Choose this option if you have existing on-NAND data using this ++ format that you need to continue to support. New data written ++ also uses the older-style format. Note: Use of this option ++ generally requires that MTD's oob layout be adjusted to use the ++ older-style format. See notes on tags formats and MTD versions ++ in yaffs_mtdif1.c. ++ ++ If unsure, say N. ++ ++config YAFFS_DOES_ECC ++ bool "Lets yaffs do its own ECC" ++ depends on YAFFS_FS && YAFFS_YAFFS1 && !YAFFS_9BYTE_TAGS ++ default n ++ help ++ This enables yaffs to use its own ECC functions instead of using ++ the ones from the generic MTD-NAND driver. ++ ++ If unsure, say N. ++ ++config YAFFS_ECC_WRONG_ORDER ++ bool "Use the same ecc byte order as Steven Hill's nand_ecc.c" ++ depends on YAFFS_FS && YAFFS_DOES_ECC && !YAFFS_9BYTE_TAGS ++ default n ++ help ++ This makes yaffs_ecc.c use the same ecc byte order as Steven ++ Hill's nand_ecc.c. If not set, then you get the same ecc byte ++ order as SmartMedia. ++ ++ If unsure, say N. ++ ++config YAFFS_YAFFS2 ++ bool "2048 byte (or larger) / page devices" ++ depends on YAFFS_FS ++ default y ++ help ++ Enable yaffs2 support -- yaffs for >= 2K bytes per page devices ++ ++ If unsure, say Y. ++ ++config YAFFS_AUTO_YAFFS2 ++ bool "Autoselect yaffs2 format" ++ depends on YAFFS_YAFFS2 ++ default y ++ help ++ Without this, you need to explicitely use yaffs2 as the file ++ system type. With this, you can say "yaffs" and yaffs or yaffs2 ++ will be used depending on the device page size (yaffs on ++ 512-byte page devices, yaffs2 on 2K page devices). ++ ++ If unsure, say Y. ++ ++config YAFFS_DISABLE_TAGS_ECC ++ bool "Disable yaffs from doing ECC on tags by default" ++ depends on YAFFS_FS && YAFFS_YAFFS2 ++ default n ++ help ++ This defaults yaffs to using its own ECC calculations on tags instead of ++ just relying on the MTD. ++ This behavior can also be overridden with tags_ecc_on and ++ tags_ecc_off mount options. ++ ++ If unsure, say N. ++ ++config YAFFS_ALWAYS_CHECK_CHUNK_ERASED ++ bool "Force chunk erase check" ++ depends on YAFFS_FS ++ default n ++ help ++ Normally yaffs only checks chunks before writing until an erased ++ chunk is found. This helps to detect any partially written ++ chunks that might have happened due to power loss. ++ ++ Enabling this forces on the test that chunks are erased in flash ++ before writing to them. This takes more time but is potentially ++ a bit more secure. ++ ++ Suggest setting Y during development and ironing out driver ++ issues etc. Suggest setting to N if you want faster writing. ++ ++ If unsure, say Y. ++ ++config YAFFS_EMPTY_LOST_AND_FOUND ++ bool "Empty lost and found on boot" ++ depends on YAFFS_FS ++ default n ++ help ++ If this is enabled then the contents of lost and found is ++ automatically dumped at mount. ++ ++ If unsure, say N. ++ ++config YAFFS_DISABLE_BLOCK_REFRESHING ++ bool "Disable yaffs2 block refreshing" ++ depends on YAFFS_FS ++ default n ++ help ++ If this is set, then block refreshing is disabled. ++ Block refreshing infrequently refreshes the oldest block in ++ a yaffs2 file system. This mechanism helps to refresh flash to ++ mitigate against data loss. This is particularly useful for MLC. ++ ++ If unsure, say N. ++ ++config YAFFS_DISABLE_BACKGROUND ++ bool "Disable yaffs2 background processing" ++ depends on YAFFS_FS ++ default n ++ help ++ If this is set, then background processing is disabled. ++ Background processing makes many foreground activities faster. ++ ++ If unsure, say N. ++ ++config YAFFS_DISABLE_BAD_BLOCK_MARKING ++ bool "Disable yaffs2 bad block marking" ++ depends on YAFFS_FS ++ default n ++ help ++ Useful during early flash bring up to prevent problems causing ++ lots of bad block marking. ++ ++ If unsure, say N. ++ ++config YAFFS_XATTR ++ bool "Enable yaffs2 xattr support" ++ depends on YAFFS_FS ++ default y ++ help ++ If this is set then yaffs2 will provide xattr support. ++ If unsure, say Y. +diff -Nur linux-3.14.4.orig/fs/yaffs2/Makefile linux-3.14.4/fs/yaffs2/Makefile +--- linux-3.14.4.orig/fs/yaffs2/Makefile 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/Makefile 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,18 @@ ++# ++# Makefile for the linux YAFFS filesystem routines. ++# ++ ++obj-$(CONFIG_YAFFS_FS) += yaffs.o ++ ++yaffs-y := yaffs_ecc.o yaffs_vfs.o yaffs_guts.o yaffs_checkptrw.o ++yaffs-y += yaffs_packedtags1.o yaffs_packedtags2.o yaffs_nand.o ++yaffs-y += yaffs_tagscompat.o yaffs_tagsmarshall.o ++yaffs-y += yaffs_mtdif.o ++yaffs-y += yaffs_nameval.o yaffs_attribs.o ++yaffs-y += yaffs_allocator.o ++yaffs-y += yaffs_yaffs1.o ++yaffs-y += yaffs_yaffs2.o ++yaffs-y += yaffs_bitmap.o ++yaffs-y += yaffs_summary.o ++yaffs-y += yaffs_verify.o ++ +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_allocator.c linux-3.14.4/fs/yaffs2/yaffs_allocator.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_allocator.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_allocator.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,357 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_allocator.h" ++#include "yaffs_guts.h" ++#include "yaffs_trace.h" ++#include "yportenv.h" ++ ++/* ++ * Each entry in yaffs_tnode_list and yaffs_obj_list hold blocks ++ * of approx 100 objects that are themn allocated singly. ++ * This is basically a simplified slab allocator. ++ * ++ * We don't use the Linux slab allocator because slab does not allow ++ * us to dump all the objects in one hit when we do a umount and tear ++ * down all the tnodes and objects. slab requires that we first free ++ * the individual objects. ++ * ++ * Once yaffs has been mainlined I shall try to motivate for a change ++ * to slab to provide the extra features we need here. ++ */ ++ ++struct yaffs_tnode_list { ++ struct yaffs_tnode_list *next; ++ struct yaffs_tnode *tnodes; ++}; ++ ++struct yaffs_obj_list { ++ struct yaffs_obj_list *next; ++ struct yaffs_obj *objects; ++}; ++ ++struct yaffs_allocator { ++ int n_tnodes_created; ++ struct yaffs_tnode *free_tnodes; ++ int n_free_tnodes; ++ struct yaffs_tnode_list *alloc_tnode_list; ++ ++ int n_obj_created; ++ struct list_head free_objs; ++ int n_free_objects; ++ ++ struct yaffs_obj_list *allocated_obj_list; ++}; ++ ++static void yaffs_deinit_raw_tnodes(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator = ++ (struct yaffs_allocator *)dev->allocator; ++ struct yaffs_tnode_list *tmp; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ while (allocator->alloc_tnode_list) { ++ tmp = allocator->alloc_tnode_list->next; ++ ++ kfree(allocator->alloc_tnode_list->tnodes); ++ kfree(allocator->alloc_tnode_list); ++ allocator->alloc_tnode_list = tmp; ++ } ++ ++ allocator->free_tnodes = NULL; ++ allocator->n_free_tnodes = 0; ++ allocator->n_tnodes_created = 0; ++} ++ ++static void yaffs_init_raw_tnodes(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator = dev->allocator; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ allocator->alloc_tnode_list = NULL; ++ allocator->free_tnodes = NULL; ++ allocator->n_free_tnodes = 0; ++ allocator->n_tnodes_created = 0; ++} ++ ++static int yaffs_create_tnodes(struct yaffs_dev *dev, int n_tnodes) ++{ ++ struct yaffs_allocator *allocator = ++ (struct yaffs_allocator *)dev->allocator; ++ int i; ++ struct yaffs_tnode *new_tnodes; ++ u8 *mem; ++ struct yaffs_tnode *curr; ++ struct yaffs_tnode *next; ++ struct yaffs_tnode_list *tnl; ++ ++ if (!allocator) { ++ BUG(); ++ return YAFFS_FAIL; ++ } ++ ++ if (n_tnodes < 1) ++ return YAFFS_OK; ++ ++ /* make these things */ ++ new_tnodes = kmalloc(n_tnodes * dev->tnode_size, GFP_NOFS); ++ mem = (u8 *) new_tnodes; ++ ++ if (!new_tnodes) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs: Could not allocate Tnodes"); ++ return YAFFS_FAIL; ++ } ++ ++ /* New hookup for wide tnodes */ ++ for (i = 0; i < n_tnodes - 1; i++) { ++ curr = (struct yaffs_tnode *)&mem[i * dev->tnode_size]; ++ next = (struct yaffs_tnode *)&mem[(i + 1) * dev->tnode_size]; ++ curr->internal[0] = next; ++ } ++ ++ curr = (struct yaffs_tnode *)&mem[(n_tnodes - 1) * dev->tnode_size]; ++ curr->internal[0] = allocator->free_tnodes; ++ allocator->free_tnodes = (struct yaffs_tnode *)mem; ++ ++ allocator->n_free_tnodes += n_tnodes; ++ allocator->n_tnodes_created += n_tnodes; ++ ++ /* Now add this bunch of tnodes to a list for freeing up. ++ * NB If we can't add this to the management list it isn't fatal ++ * but it just means we can't free this bunch of tnodes later. ++ */ ++ tnl = kmalloc(sizeof(struct yaffs_tnode_list), GFP_NOFS); ++ if (!tnl) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "Could not add tnodes to management list"); ++ return YAFFS_FAIL; ++ } else { ++ tnl->tnodes = new_tnodes; ++ tnl->next = allocator->alloc_tnode_list; ++ allocator->alloc_tnode_list = tnl; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_ALLOCATE, "Tnodes added"); ++ ++ return YAFFS_OK; ++} ++ ++struct yaffs_tnode *yaffs_alloc_raw_tnode(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator = ++ (struct yaffs_allocator *)dev->allocator; ++ struct yaffs_tnode *tn = NULL; ++ ++ if (!allocator) { ++ BUG(); ++ return NULL; ++ } ++ ++ /* If there are none left make more */ ++ if (!allocator->free_tnodes) ++ yaffs_create_tnodes(dev, YAFFS_ALLOCATION_NTNODES); ++ ++ if (allocator->free_tnodes) { ++ tn = allocator->free_tnodes; ++ allocator->free_tnodes = allocator->free_tnodes->internal[0]; ++ allocator->n_free_tnodes--; ++ } ++ ++ return tn; ++} ++ ++/* FreeTnode frees up a tnode and puts it back on the free list */ ++void yaffs_free_raw_tnode(struct yaffs_dev *dev, struct yaffs_tnode *tn) ++{ ++ struct yaffs_allocator *allocator = dev->allocator; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ if (tn) { ++ tn->internal[0] = allocator->free_tnodes; ++ allocator->free_tnodes = tn; ++ allocator->n_free_tnodes++; ++ } ++ dev->checkpoint_blocks_required = 0; /* force recalculation */ ++} ++ ++/*--------------- yaffs_obj alloaction ------------------------ ++ * ++ * Free yaffs_objs are stored in a list using obj->siblings. ++ * The blocks of allocated objects are stored in a linked list. ++ */ ++ ++static void yaffs_init_raw_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator = dev->allocator; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ allocator->allocated_obj_list = NULL; ++ INIT_LIST_HEAD(&allocator->free_objs); ++ allocator->n_free_objects = 0; ++} ++ ++static void yaffs_deinit_raw_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator = dev->allocator; ++ struct yaffs_obj_list *tmp; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ while (allocator->allocated_obj_list) { ++ tmp = allocator->allocated_obj_list->next; ++ kfree(allocator->allocated_obj_list->objects); ++ kfree(allocator->allocated_obj_list); ++ allocator->allocated_obj_list = tmp; ++ } ++ ++ INIT_LIST_HEAD(&allocator->free_objs); ++ allocator->n_free_objects = 0; ++ allocator->n_obj_created = 0; ++} ++ ++static int yaffs_create_free_objs(struct yaffs_dev *dev, int n_obj) ++{ ++ struct yaffs_allocator *allocator = dev->allocator; ++ int i; ++ struct yaffs_obj *new_objs; ++ struct yaffs_obj_list *list; ++ ++ if (!allocator) { ++ BUG(); ++ return YAFFS_FAIL; ++ } ++ ++ if (n_obj < 1) ++ return YAFFS_OK; ++ ++ /* make these things */ ++ new_objs = kmalloc(n_obj * sizeof(struct yaffs_obj), GFP_NOFS); ++ list = kmalloc(sizeof(struct yaffs_obj_list), GFP_NOFS); ++ ++ if (!new_objs || !list) { ++ kfree(new_objs); ++ new_objs = NULL; ++ kfree(list); ++ list = NULL; ++ yaffs_trace(YAFFS_TRACE_ALLOCATE, ++ "Could not allocate more objects"); ++ return YAFFS_FAIL; ++ } ++ ++ /* Hook them into the free list */ ++ for (i = 0; i < n_obj; i++) ++ list_add(&new_objs[i].siblings, &allocator->free_objs); ++ ++ allocator->n_free_objects += n_obj; ++ allocator->n_obj_created += n_obj; ++ ++ /* Now add this bunch of Objects to a list for freeing up. */ ++ ++ list->objects = new_objs; ++ list->next = allocator->allocated_obj_list; ++ allocator->allocated_obj_list = list; ++ ++ return YAFFS_OK; ++} ++ ++struct yaffs_obj *yaffs_alloc_raw_obj(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj = NULL; ++ struct list_head *lh; ++ struct yaffs_allocator *allocator = dev->allocator; ++ ++ if (!allocator) { ++ BUG(); ++ return obj; ++ } ++ ++ /* If there are none left make more */ ++ if (list_empty(&allocator->free_objs)) ++ yaffs_create_free_objs(dev, YAFFS_ALLOCATION_NOBJECTS); ++ ++ if (!list_empty(&allocator->free_objs)) { ++ lh = allocator->free_objs.next; ++ obj = list_entry(lh, struct yaffs_obj, siblings); ++ list_del_init(lh); ++ allocator->n_free_objects--; ++ } ++ ++ return obj; ++} ++ ++void yaffs_free_raw_obj(struct yaffs_dev *dev, struct yaffs_obj *obj) ++{ ++ ++ struct yaffs_allocator *allocator = dev->allocator; ++ ++ if (!allocator) { ++ BUG(); ++ return; ++ } ++ ++ /* Link into the free list. */ ++ list_add(&obj->siblings, &allocator->free_objs); ++ allocator->n_free_objects++; ++} ++ ++void yaffs_deinit_raw_tnodes_and_objs(struct yaffs_dev *dev) ++{ ++ ++ if (!dev->allocator) { ++ BUG(); ++ return; ++ } ++ ++ yaffs_deinit_raw_tnodes(dev); ++ yaffs_deinit_raw_objs(dev); ++ kfree(dev->allocator); ++ dev->allocator = NULL; ++} ++ ++void yaffs_init_raw_tnodes_and_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_allocator *allocator; ++ ++ if (dev->allocator) { ++ BUG(); ++ return; ++ } ++ ++ allocator = kmalloc(sizeof(struct yaffs_allocator), GFP_NOFS); ++ if (allocator) { ++ dev->allocator = allocator; ++ yaffs_init_raw_tnodes(dev); ++ yaffs_init_raw_objs(dev); ++ } ++} ++ +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_allocator.h linux-3.14.4/fs/yaffs2/yaffs_allocator.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_allocator.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_allocator.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,30 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_ALLOCATOR_H__ ++#define __YAFFS_ALLOCATOR_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_init_raw_tnodes_and_objs(struct yaffs_dev *dev); ++void yaffs_deinit_raw_tnodes_and_objs(struct yaffs_dev *dev); ++ ++struct yaffs_tnode *yaffs_alloc_raw_tnode(struct yaffs_dev *dev); ++void yaffs_free_raw_tnode(struct yaffs_dev *dev, struct yaffs_tnode *tn); ++ ++struct yaffs_obj *yaffs_alloc_raw_obj(struct yaffs_dev *dev); ++void yaffs_free_raw_obj(struct yaffs_dev *dev, struct yaffs_obj *obj); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_attribs.c linux-3.14.4/fs/yaffs2/yaffs_attribs.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_attribs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_attribs.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,166 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_guts.h" ++#include "yaffs_attribs.h" ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 5, 0)) ++static inline uid_t ia_uid_read(const struct iattr *iattr) ++{ ++ return from_kuid(&init_user_ns, iattr->ia_uid); ++} ++ ++static inline gid_t ia_gid_read(const struct iattr *iattr) ++{ ++ return from_kgid(&init_user_ns, iattr->ia_gid); ++} ++ ++static inline void ia_uid_write(struct iattr *iattr, uid_t uid) ++{ ++ iattr->ia_uid = make_kuid(&init_user_ns, uid); ++} ++ ++static inline void ia_gid_write(struct iattr *iattr, gid_t gid) ++{ ++ iattr->ia_gid = make_kgid(&init_user_ns, gid); ++} ++#else ++static inline uid_t ia_uid_read(const struct iattr *iattr) ++{ ++ return iattr->ia_uid; ++} ++ ++static inline gid_t ia_gid_read(const struct iattr *inode) ++{ ++ return iattr->ia_gid; ++} ++ ++static inline void ia_uid_write(struct iattr *iattr, uid_t uid) ++{ ++ iattr->ia_uid = uid; ++} ++ ++static inline void ia_gid_write(struct iattr *iattr, gid_t gid) ++{ ++ iattr->ia_gid = gid; ++} ++#endif ++ ++void yaffs_load_attribs(struct yaffs_obj *obj, struct yaffs_obj_hdr *oh) ++{ ++ obj->yst_uid = oh->yst_uid; ++ obj->yst_gid = oh->yst_gid; ++ obj->yst_atime = oh->yst_atime; ++ obj->yst_mtime = oh->yst_mtime; ++ obj->yst_ctime = oh->yst_ctime; ++ obj->yst_rdev = oh->yst_rdev; ++} ++ ++void yaffs_load_attribs_oh(struct yaffs_obj_hdr *oh, struct yaffs_obj *obj) ++{ ++ oh->yst_uid = obj->yst_uid; ++ oh->yst_gid = obj->yst_gid; ++ oh->yst_atime = obj->yst_atime; ++ oh->yst_mtime = obj->yst_mtime; ++ oh->yst_ctime = obj->yst_ctime; ++ oh->yst_rdev = obj->yst_rdev; ++ ++} ++ ++void yaffs_load_current_time(struct yaffs_obj *obj, int do_a, int do_c) ++{ ++ obj->yst_mtime = Y_CURRENT_TIME; ++ if (do_a) ++ obj->yst_atime = obj->yst_mtime; ++ if (do_c) ++ obj->yst_ctime = obj->yst_mtime; ++} ++ ++void yaffs_attribs_init(struct yaffs_obj *obj, u32 gid, u32 uid, u32 rdev) ++{ ++ yaffs_load_current_time(obj, 1, 1); ++ obj->yst_rdev = rdev; ++ obj->yst_uid = uid; ++ obj->yst_gid = gid; ++} ++ ++static loff_t yaffs_get_file_size(struct yaffs_obj *obj) ++{ ++ YCHAR *alias = NULL; ++ obj = yaffs_get_equivalent_obj(obj); ++ ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ return obj->variant.file_variant.file_size; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ alias = obj->variant.symlink_variant.alias; ++ if (!alias) ++ return 0; ++ return strnlen(alias, YAFFS_MAX_ALIAS_LENGTH); ++ default: ++ return 0; ++ } ++} ++ ++int yaffs_set_attribs(struct yaffs_obj *obj, struct iattr *attr) ++{ ++ unsigned int valid = attr->ia_valid; ++ ++ if (valid & ATTR_MODE) ++ obj->yst_mode = attr->ia_mode; ++ if (valid & ATTR_UID) ++ obj->yst_uid = ia_uid_read(attr); ++ if (valid & ATTR_GID) ++ obj->yst_gid = ia_gid_read(attr); ++ ++ if (valid & ATTR_ATIME) ++ obj->yst_atime = Y_TIME_CONVERT(attr->ia_atime); ++ if (valid & ATTR_CTIME) ++ obj->yst_ctime = Y_TIME_CONVERT(attr->ia_ctime); ++ if (valid & ATTR_MTIME) ++ obj->yst_mtime = Y_TIME_CONVERT(attr->ia_mtime); ++ ++ if (valid & ATTR_SIZE) ++ yaffs_resize_file(obj, attr->ia_size); ++ ++ yaffs_update_oh(obj, NULL, 1, 0, 0, NULL); ++ ++ return YAFFS_OK; ++ ++} ++ ++int yaffs_get_attribs(struct yaffs_obj *obj, struct iattr *attr) ++{ ++ unsigned int valid = 0; ++ ++ attr->ia_mode = obj->yst_mode; ++ valid |= ATTR_MODE; ++ ia_uid_write(attr, obj->yst_uid); ++ valid |= ATTR_UID; ++ ia_gid_write(attr, obj->yst_gid); ++ valid |= ATTR_GID; ++ ++ Y_TIME_CONVERT(attr->ia_atime) = obj->yst_atime; ++ valid |= ATTR_ATIME; ++ Y_TIME_CONVERT(attr->ia_ctime) = obj->yst_ctime; ++ valid |= ATTR_CTIME; ++ Y_TIME_CONVERT(attr->ia_mtime) = obj->yst_mtime; ++ valid |= ATTR_MTIME; ++ ++ attr->ia_size = yaffs_get_file_size(obj); ++ valid |= ATTR_SIZE; ++ ++ attr->ia_valid = valid; ++ ++ return YAFFS_OK; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_attribs.h linux-3.14.4/fs/yaffs2/yaffs_attribs.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_attribs.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_attribs.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,28 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_ATTRIBS_H__ ++#define __YAFFS_ATTRIBS_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_load_attribs(struct yaffs_obj *obj, struct yaffs_obj_hdr *oh); ++void yaffs_load_attribs_oh(struct yaffs_obj_hdr *oh, struct yaffs_obj *obj); ++void yaffs_attribs_init(struct yaffs_obj *obj, u32 gid, u32 uid, u32 rdev); ++void yaffs_load_current_time(struct yaffs_obj *obj, int do_a, int do_c); ++int yaffs_set_attribs(struct yaffs_obj *obj, struct iattr *attr); ++int yaffs_get_attribs(struct yaffs_obj *obj, struct iattr *attr); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_bitmap.c linux-3.14.4/fs/yaffs2/yaffs_bitmap.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_bitmap.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_bitmap.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,97 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_bitmap.h" ++#include "yaffs_trace.h" ++/* ++ * Chunk bitmap manipulations ++ */ ++ ++static inline u8 *yaffs_block_bits(struct yaffs_dev *dev, int blk) ++{ ++ if (blk < dev->internal_start_block || blk > dev->internal_end_block) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "BlockBits block %d is not valid", ++ blk); ++ BUG(); ++ } ++ return dev->chunk_bits + ++ (dev->chunk_bit_stride * (blk - dev->internal_start_block)); ++} ++ ++void yaffs_verify_chunk_bit_id(struct yaffs_dev *dev, int blk, int chunk) ++{ ++ if (blk < dev->internal_start_block || blk > dev->internal_end_block || ++ chunk < 0 || chunk >= dev->param.chunks_per_block) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "Chunk Id (%d:%d) invalid", ++ blk, chunk); ++ BUG(); ++ } ++} ++ ++void yaffs_clear_chunk_bits(struct yaffs_dev *dev, int blk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ ++ memset(blk_bits, 0, dev->chunk_bit_stride); ++} ++ ++void yaffs_clear_chunk_bit(struct yaffs_dev *dev, int blk, int chunk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ ++ yaffs_verify_chunk_bit_id(dev, blk, chunk); ++ blk_bits[chunk / 8] &= ~(1 << (chunk & 7)); ++} ++ ++void yaffs_set_chunk_bit(struct yaffs_dev *dev, int blk, int chunk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ ++ yaffs_verify_chunk_bit_id(dev, blk, chunk); ++ blk_bits[chunk / 8] |= (1 << (chunk & 7)); ++} ++ ++int yaffs_check_chunk_bit(struct yaffs_dev *dev, int blk, int chunk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ ++ yaffs_verify_chunk_bit_id(dev, blk, chunk); ++ return (blk_bits[chunk / 8] & (1 << (chunk & 7))) ? 1 : 0; ++} ++ ++int yaffs_still_some_chunks(struct yaffs_dev *dev, int blk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ int i; ++ ++ for (i = 0; i < dev->chunk_bit_stride; i++) { ++ if (*blk_bits) ++ return 1; ++ blk_bits++; ++ } ++ return 0; ++} ++ ++int yaffs_count_chunk_bits(struct yaffs_dev *dev, int blk) ++{ ++ u8 *blk_bits = yaffs_block_bits(dev, blk); ++ int i; ++ int n = 0; ++ ++ for (i = 0; i < dev->chunk_bit_stride; i++, blk_bits++) ++ n += hweight8(*blk_bits); ++ ++ return n; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_bitmap.h linux-3.14.4/fs/yaffs2/yaffs_bitmap.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_bitmap.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_bitmap.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,33 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++/* ++ * Chunk bitmap manipulations ++ */ ++ ++#ifndef __YAFFS_BITMAP_H__ ++#define __YAFFS_BITMAP_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_verify_chunk_bit_id(struct yaffs_dev *dev, int blk, int chunk); ++void yaffs_clear_chunk_bits(struct yaffs_dev *dev, int blk); ++void yaffs_clear_chunk_bit(struct yaffs_dev *dev, int blk, int chunk); ++void yaffs_set_chunk_bit(struct yaffs_dev *dev, int blk, int chunk); ++int yaffs_check_chunk_bit(struct yaffs_dev *dev, int blk, int chunk); ++int yaffs_still_some_chunks(struct yaffs_dev *dev, int blk); ++int yaffs_count_chunk_bits(struct yaffs_dev *dev, int blk); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_checkptrw.c linux-3.14.4/fs/yaffs2/yaffs_checkptrw.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_checkptrw.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_checkptrw.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,474 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_checkptrw.h" ++#include "yaffs_getblockinfo.h" ++ ++struct yaffs_checkpt_chunk_hdr { ++ int version; ++ int seq; ++ u32 sum; ++ u32 xor; ++} ; ++ ++ ++static int apply_chunk_offset(struct yaffs_dev *dev, int chunk) ++{ ++ return chunk - dev->chunk_offset; ++} ++ ++static int apply_block_offset(struct yaffs_dev *dev, int block) ++{ ++ return block - dev->block_offset; ++} ++ ++static void yaffs2_checkpt_init_chunk_hdr(struct yaffs_dev *dev) ++{ ++ struct yaffs_checkpt_chunk_hdr hdr; ++ ++ hdr.version = YAFFS_CHECKPOINT_VERSION; ++ hdr.seq = dev->checkpt_page_seq; ++ hdr.sum = dev->checkpt_sum; ++ hdr.xor = dev->checkpt_xor; ++ ++ dev->checkpt_byte_offs = sizeof(hdr); ++ ++ memcpy(dev->checkpt_buffer, &hdr, sizeof(hdr)); ++} ++ ++static int yaffs2_checkpt_check_chunk_hdr(struct yaffs_dev *dev) ++{ ++ struct yaffs_checkpt_chunk_hdr hdr; ++ ++ memcpy(&hdr, dev->checkpt_buffer, sizeof(hdr)); ++ ++ dev->checkpt_byte_offs = sizeof(hdr); ++ ++ return hdr.version == YAFFS_CHECKPOINT_VERSION && ++ hdr.seq == dev->checkpt_page_seq && ++ hdr.sum == dev->checkpt_sum && ++ hdr.xor == dev->checkpt_xor; ++} ++ ++static int yaffs2_checkpt_space_ok(struct yaffs_dev *dev) ++{ ++ int blocks_avail = dev->n_erased_blocks - dev->param.n_reserved_blocks; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "checkpt blocks_avail = %d", blocks_avail); ++ ++ return (blocks_avail <= 0) ? 0 : 1; ++} ++ ++static int yaffs_checkpt_erase(struct yaffs_dev *dev) ++{ ++ int i; ++ ++ if (!dev->drv.drv_erase_fn) ++ return 0; ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "checking blocks %d to %d", ++ dev->internal_start_block, dev->internal_end_block); ++ ++ for (i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, i); ++ int offset_i = apply_block_offset(dev, i); ++ int result; ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_CHECKPOINT) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "erasing checkpt block %d", i); ++ ++ dev->n_erasures++; ++ ++ result = dev->drv.drv_erase_fn(dev, offset_i); ++ if(result) { ++ bi->block_state = YAFFS_BLOCK_STATE_EMPTY; ++ dev->n_erased_blocks++; ++ dev->n_free_chunks += ++ dev->param.chunks_per_block; ++ } else { ++ dev->drv.drv_mark_bad_fn(dev, offset_i); ++ bi->block_state = YAFFS_BLOCK_STATE_DEAD; ++ } ++ } ++ } ++ ++ dev->blocks_in_checkpt = 0; ++ ++ return 1; ++} ++ ++static void yaffs2_checkpt_find_erased_block(struct yaffs_dev *dev) ++{ ++ int i; ++ int blocks_avail = dev->n_erased_blocks - dev->param.n_reserved_blocks; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "allocating checkpt block: erased %d reserved %d avail %d next %d ", ++ dev->n_erased_blocks, dev->param.n_reserved_blocks, ++ blocks_avail, dev->checkpt_next_block); ++ ++ if (dev->checkpt_next_block >= 0 && ++ dev->checkpt_next_block <= dev->internal_end_block && ++ blocks_avail > 0) { ++ ++ for (i = dev->checkpt_next_block; i <= dev->internal_end_block; ++ i++) { ++ struct yaffs_block_info *bi; ++ ++ bi = yaffs_get_block_info(dev, i); ++ if (bi->block_state == YAFFS_BLOCK_STATE_EMPTY) { ++ dev->checkpt_next_block = i + 1; ++ dev->checkpt_cur_block = i; ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "allocating checkpt block %d", i); ++ return; ++ } ++ } ++ } ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, "out of checkpt blocks"); ++ ++ dev->checkpt_next_block = -1; ++ dev->checkpt_cur_block = -1; ++} ++ ++static void yaffs2_checkpt_find_block(struct yaffs_dev *dev) ++{ ++ int i; ++ struct yaffs_ext_tags tags; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "find next checkpt block: start: blocks %d next %d", ++ dev->blocks_in_checkpt, dev->checkpt_next_block); ++ ++ if (dev->blocks_in_checkpt < dev->checkpt_max_blocks) ++ for (i = dev->checkpt_next_block; i <= dev->internal_end_block; ++ i++) { ++ int chunk = i * dev->param.chunks_per_block; ++ enum yaffs_block_state state; ++ u32 seq; ++ ++ dev->tagger.read_chunk_tags_fn(dev, ++ apply_chunk_offset(dev, chunk), ++ NULL, &tags); ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "find next checkpt block: search: block %d state %d oid %d seq %d eccr %d", ++ i, (int) state, ++ tags.obj_id, tags.seq_number, ++ tags.ecc_result); ++ ++ if (tags.seq_number != YAFFS_SEQUENCE_CHECKPOINT_DATA) ++ continue; ++ ++ dev->tagger.query_block_fn(dev, ++ apply_block_offset(dev, i), ++ &state, &seq); ++ if (state == YAFFS_BLOCK_STATE_DEAD) ++ continue; ++ ++ /* Right kind of block */ ++ dev->checkpt_next_block = tags.obj_id; ++ dev->checkpt_cur_block = i; ++ dev->checkpt_block_list[dev->blocks_in_checkpt] = i; ++ dev->blocks_in_checkpt++; ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "found checkpt block %d", i); ++ return; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, "found no more checkpt blocks"); ++ ++ dev->checkpt_next_block = -1; ++ dev->checkpt_cur_block = -1; ++} ++ ++int yaffs2_checkpt_open(struct yaffs_dev *dev, int writing) ++{ ++ int i; ++ ++ dev->checkpt_open_write = writing; ++ ++ /* Got the functions we need? */ ++ if (!dev->tagger.write_chunk_tags_fn || ++ !dev->tagger.read_chunk_tags_fn || ++ !dev->drv.drv_erase_fn || ++ !dev->drv.drv_mark_bad_fn) ++ return 0; ++ ++ if (writing && !yaffs2_checkpt_space_ok(dev)) ++ return 0; ++ ++ if (!dev->checkpt_buffer) ++ dev->checkpt_buffer = ++ kmalloc(dev->param.total_bytes_per_chunk, GFP_NOFS); ++ if (!dev->checkpt_buffer) ++ return 0; ++ ++ dev->checkpt_page_seq = 0; ++ dev->checkpt_byte_count = 0; ++ dev->checkpt_sum = 0; ++ dev->checkpt_xor = 0; ++ dev->checkpt_cur_block = -1; ++ dev->checkpt_cur_chunk = -1; ++ dev->checkpt_next_block = dev->internal_start_block; ++ ++ if (writing) { ++ memset(dev->checkpt_buffer, 0, dev->data_bytes_per_chunk); ++ yaffs2_checkpt_init_chunk_hdr(dev); ++ return yaffs_checkpt_erase(dev); ++ } ++ ++ /* Opening for a read */ ++ /* Set to a value that will kick off a read */ ++ dev->checkpt_byte_offs = dev->data_bytes_per_chunk; ++ /* A checkpoint block list of 1 checkpoint block per 16 block is ++ * (hopefully) going to be way more than we need */ ++ dev->blocks_in_checkpt = 0; ++ dev->checkpt_max_blocks = ++ (dev->internal_end_block - dev->internal_start_block) / 16 + 2; ++ dev->checkpt_block_list = ++ kmalloc(sizeof(int) * dev->checkpt_max_blocks, GFP_NOFS); ++ ++ if (!dev->checkpt_block_list) ++ return 0; ++ ++ for (i = 0; i < dev->checkpt_max_blocks; i++) ++ dev->checkpt_block_list[i] = -1; ++ ++ return 1; ++} ++ ++int yaffs2_get_checkpt_sum(struct yaffs_dev *dev, u32 * sum) ++{ ++ u32 composite_sum; ++ ++ composite_sum = (dev->checkpt_sum << 8) | (dev->checkpt_xor & 0xff); ++ *sum = composite_sum; ++ return 1; ++} ++ ++static int yaffs2_checkpt_flush_buffer(struct yaffs_dev *dev) ++{ ++ int chunk; ++ int offset_chunk; ++ struct yaffs_ext_tags tags; ++ ++ if (dev->checkpt_cur_block < 0) { ++ yaffs2_checkpt_find_erased_block(dev); ++ dev->checkpt_cur_chunk = 0; ++ } ++ ++ if (dev->checkpt_cur_block < 0) ++ return 0; ++ ++ tags.is_deleted = 0; ++ tags.obj_id = dev->checkpt_next_block; /* Hint to next place to look */ ++ tags.chunk_id = dev->checkpt_page_seq + 1; ++ tags.seq_number = YAFFS_SEQUENCE_CHECKPOINT_DATA; ++ tags.n_bytes = dev->data_bytes_per_chunk; ++ if (dev->checkpt_cur_chunk == 0) { ++ /* First chunk we write for the block? Set block state to ++ checkpoint */ ++ struct yaffs_block_info *bi = ++ yaffs_get_block_info(dev, dev->checkpt_cur_block); ++ bi->block_state = YAFFS_BLOCK_STATE_CHECKPOINT; ++ dev->blocks_in_checkpt++; ++ } ++ ++ chunk = ++ dev->checkpt_cur_block * dev->param.chunks_per_block + ++ dev->checkpt_cur_chunk; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "checkpoint wite buffer nand %d(%d:%d) objid %d chId %d", ++ chunk, dev->checkpt_cur_block, dev->checkpt_cur_chunk, ++ tags.obj_id, tags.chunk_id); ++ ++ offset_chunk = apply_chunk_offset(dev, chunk); ++ ++ dev->n_page_writes++; ++ ++ dev->tagger.write_chunk_tags_fn(dev, offset_chunk, ++ dev->checkpt_buffer, &tags); ++ dev->checkpt_page_seq++; ++ dev->checkpt_cur_chunk++; ++ if (dev->checkpt_cur_chunk >= dev->param.chunks_per_block) { ++ dev->checkpt_cur_chunk = 0; ++ dev->checkpt_cur_block = -1; ++ } ++ memset(dev->checkpt_buffer, 0, dev->data_bytes_per_chunk); ++ ++ yaffs2_checkpt_init_chunk_hdr(dev); ++ ++ ++ return 1; ++} ++ ++int yaffs2_checkpt_wr(struct yaffs_dev *dev, const void *data, int n_bytes) ++{ ++ int i = 0; ++ int ok = 1; ++ u8 *data_bytes = (u8 *) data; ++ ++ if (!dev->checkpt_buffer) ++ return 0; ++ ++ if (!dev->checkpt_open_write) ++ return -1; ++ ++ while (i < n_bytes && ok) { ++ dev->checkpt_buffer[dev->checkpt_byte_offs] = *data_bytes; ++ dev->checkpt_sum += *data_bytes; ++ dev->checkpt_xor ^= *data_bytes; ++ ++ dev->checkpt_byte_offs++; ++ i++; ++ data_bytes++; ++ dev->checkpt_byte_count++; ++ ++ if (dev->checkpt_byte_offs < 0 || ++ dev->checkpt_byte_offs >= dev->data_bytes_per_chunk) ++ ok = yaffs2_checkpt_flush_buffer(dev); ++ } ++ ++ return i; ++} ++ ++int yaffs2_checkpt_rd(struct yaffs_dev *dev, void *data, int n_bytes) ++{ ++ int i = 0; ++ int ok = 1; ++ struct yaffs_ext_tags tags; ++ int chunk; ++ int offset_chunk; ++ u8 *data_bytes = (u8 *) data; ++ ++ if (!dev->checkpt_buffer) ++ return 0; ++ ++ if (dev->checkpt_open_write) ++ return -1; ++ ++ while (i < n_bytes && ok) { ++ ++ if (dev->checkpt_byte_offs < 0 || ++ dev->checkpt_byte_offs >= dev->data_bytes_per_chunk) { ++ ++ if (dev->checkpt_cur_block < 0) { ++ yaffs2_checkpt_find_block(dev); ++ dev->checkpt_cur_chunk = 0; ++ } ++ ++ if (dev->checkpt_cur_block < 0) { ++ ok = 0; ++ break; ++ } ++ ++ chunk = dev->checkpt_cur_block * ++ dev->param.chunks_per_block + ++ dev->checkpt_cur_chunk; ++ ++ offset_chunk = apply_chunk_offset(dev, chunk); ++ dev->n_page_reads++; ++ ++ /* read in the next chunk */ ++ dev->tagger.read_chunk_tags_fn(dev, ++ offset_chunk, ++ dev->checkpt_buffer, ++ &tags); ++ ++ if (tags.chunk_id != (dev->checkpt_page_seq + 1) || ++ tags.ecc_result > YAFFS_ECC_RESULT_FIXED || ++ tags.seq_number != YAFFS_SEQUENCE_CHECKPOINT_DATA) { ++ ok = 0; ++ break; ++ } ++ if(!yaffs2_checkpt_check_chunk_hdr(dev)) { ++ ok = 0; ++ break; ++ } ++ ++ dev->checkpt_page_seq++; ++ dev->checkpt_cur_chunk++; ++ ++ if (dev->checkpt_cur_chunk >= ++ dev->param.chunks_per_block) ++ dev->checkpt_cur_block = -1; ++ ++ } ++ ++ *data_bytes = dev->checkpt_buffer[dev->checkpt_byte_offs]; ++ dev->checkpt_sum += *data_bytes; ++ dev->checkpt_xor ^= *data_bytes; ++ dev->checkpt_byte_offs++; ++ i++; ++ data_bytes++; ++ dev->checkpt_byte_count++; ++ } ++ ++ return i; ++} ++ ++int yaffs_checkpt_close(struct yaffs_dev *dev) ++{ ++ int i; ++ ++ if (dev->checkpt_open_write) { ++ if (dev->checkpt_byte_offs != ++ sizeof(sizeof(struct yaffs_checkpt_chunk_hdr))) ++ yaffs2_checkpt_flush_buffer(dev); ++ } else if (dev->checkpt_block_list) { ++ for (i = 0; ++ i < dev->blocks_in_checkpt && ++ dev->checkpt_block_list[i] >= 0; i++) { ++ int blk = dev->checkpt_block_list[i]; ++ struct yaffs_block_info *bi = NULL; ++ ++ if (dev->internal_start_block <= blk && ++ blk <= dev->internal_end_block) ++ bi = yaffs_get_block_info(dev, blk); ++ if (bi && bi->block_state == YAFFS_BLOCK_STATE_EMPTY) ++ bi->block_state = YAFFS_BLOCK_STATE_CHECKPOINT; ++ } ++ kfree(dev->checkpt_block_list); ++ dev->checkpt_block_list = NULL; ++ } ++ ++ dev->n_free_chunks -= ++ dev->blocks_in_checkpt * dev->param.chunks_per_block; ++ dev->n_erased_blocks -= dev->blocks_in_checkpt; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, "checkpoint byte count %d", ++ dev->checkpt_byte_count); ++ ++ if (dev->checkpt_buffer) { ++ /* free the buffer */ ++ kfree(dev->checkpt_buffer); ++ dev->checkpt_buffer = NULL; ++ return 1; ++ } else { ++ return 0; ++ } ++} ++ ++int yaffs2_checkpt_invalidate_stream(struct yaffs_dev *dev) ++{ ++ /* Erase the checkpoint data */ ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "checkpoint invalidate of %d blocks", ++ dev->blocks_in_checkpt); ++ ++ return yaffs_checkpt_erase(dev); ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_checkptrw.h linux-3.14.4/fs/yaffs2/yaffs_checkptrw.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_checkptrw.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_checkptrw.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,33 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_CHECKPTRW_H__ ++#define __YAFFS_CHECKPTRW_H__ ++ ++#include "yaffs_guts.h" ++ ++int yaffs2_checkpt_open(struct yaffs_dev *dev, int writing); ++ ++int yaffs2_checkpt_wr(struct yaffs_dev *dev, const void *data, int n_bytes); ++ ++int yaffs2_checkpt_rd(struct yaffs_dev *dev, void *data, int n_bytes); ++ ++int yaffs2_get_checkpt_sum(struct yaffs_dev *dev, u32 * sum); ++ ++int yaffs_checkpt_close(struct yaffs_dev *dev); ++ ++int yaffs2_checkpt_invalidate_stream(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_ecc.c linux-3.14.4/fs/yaffs2/yaffs_ecc.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_ecc.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_ecc.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,281 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++/* ++ * This code implements the ECC algorithm used in SmartMedia. ++ * ++ * The ECC comprises 22 bits of parity information and is stuffed into 3 bytes. ++ * The two unused bit are set to 1. ++ * The ECC can correct single bit errors in a 256-byte page of data. Thus, two ++ * such ECC blocks are used on a 512-byte NAND page. ++ * ++ */ ++ ++#include "yportenv.h" ++ ++#include "yaffs_ecc.h" ++ ++/* Table generated by gen-ecc.c ++ * Using a table means we do not have to calculate p1..p4 and p1'..p4' ++ * for each byte of data. These are instead provided in a table in bits7..2. ++ * Bit 0 of each entry indicates whether the entry has an odd or even parity, ++ * and therefore this bytes influence on the line parity. ++ */ ++ ++static const unsigned char column_parity_table[] = { ++ 0x00, 0x55, 0x59, 0x0c, 0x65, 0x30, 0x3c, 0x69, ++ 0x69, 0x3c, 0x30, 0x65, 0x0c, 0x59, 0x55, 0x00, ++ 0x95, 0xc0, 0xcc, 0x99, 0xf0, 0xa5, 0xa9, 0xfc, ++ 0xfc, 0xa9, 0xa5, 0xf0, 0x99, 0xcc, 0xc0, 0x95, ++ 0x99, 0xcc, 0xc0, 0x95, 0xfc, 0xa9, 0xa5, 0xf0, ++ 0xf0, 0xa5, 0xa9, 0xfc, 0x95, 0xc0, 0xcc, 0x99, ++ 0x0c, 0x59, 0x55, 0x00, 0x69, 0x3c, 0x30, 0x65, ++ 0x65, 0x30, 0x3c, 0x69, 0x00, 0x55, 0x59, 0x0c, ++ 0xa5, 0xf0, 0xfc, 0xa9, 0xc0, 0x95, 0x99, 0xcc, ++ 0xcc, 0x99, 0x95, 0xc0, 0xa9, 0xfc, 0xf0, 0xa5, ++ 0x30, 0x65, 0x69, 0x3c, 0x55, 0x00, 0x0c, 0x59, ++ 0x59, 0x0c, 0x00, 0x55, 0x3c, 0x69, 0x65, 0x30, ++ 0x3c, 0x69, 0x65, 0x30, 0x59, 0x0c, 0x00, 0x55, ++ 0x55, 0x00, 0x0c, 0x59, 0x30, 0x65, 0x69, 0x3c, ++ 0xa9, 0xfc, 0xf0, 0xa5, 0xcc, 0x99, 0x95, 0xc0, ++ 0xc0, 0x95, 0x99, 0xcc, 0xa5, 0xf0, 0xfc, 0xa9, ++ 0xa9, 0xfc, 0xf0, 0xa5, 0xcc, 0x99, 0x95, 0xc0, ++ 0xc0, 0x95, 0x99, 0xcc, 0xa5, 0xf0, 0xfc, 0xa9, ++ 0x3c, 0x69, 0x65, 0x30, 0x59, 0x0c, 0x00, 0x55, ++ 0x55, 0x00, 0x0c, 0x59, 0x30, 0x65, 0x69, 0x3c, ++ 0x30, 0x65, 0x69, 0x3c, 0x55, 0x00, 0x0c, 0x59, ++ 0x59, 0x0c, 0x00, 0x55, 0x3c, 0x69, 0x65, 0x30, ++ 0xa5, 0xf0, 0xfc, 0xa9, 0xc0, 0x95, 0x99, 0xcc, ++ 0xcc, 0x99, 0x95, 0xc0, 0xa9, 0xfc, 0xf0, 0xa5, ++ 0x0c, 0x59, 0x55, 0x00, 0x69, 0x3c, 0x30, 0x65, ++ 0x65, 0x30, 0x3c, 0x69, 0x00, 0x55, 0x59, 0x0c, ++ 0x99, 0xcc, 0xc0, 0x95, 0xfc, 0xa9, 0xa5, 0xf0, ++ 0xf0, 0xa5, 0xa9, 0xfc, 0x95, 0xc0, 0xcc, 0x99, ++ 0x95, 0xc0, 0xcc, 0x99, 0xf0, 0xa5, 0xa9, 0xfc, ++ 0xfc, 0xa9, 0xa5, 0xf0, 0x99, 0xcc, 0xc0, 0x95, ++ 0x00, 0x55, 0x59, 0x0c, 0x65, 0x30, 0x3c, 0x69, ++ 0x69, 0x3c, 0x30, 0x65, 0x0c, 0x59, 0x55, 0x00, ++}; ++ ++ ++/* Calculate the ECC for a 256-byte block of data */ ++void yaffs_ecc_calc(const unsigned char *data, unsigned char *ecc) ++{ ++ unsigned int i; ++ unsigned char col_parity = 0; ++ unsigned char line_parity = 0; ++ unsigned char line_parity_prime = 0; ++ unsigned char t; ++ unsigned char b; ++ ++ for (i = 0; i < 256; i++) { ++ b = column_parity_table[*data++]; ++ col_parity ^= b; ++ ++ if (b & 0x01) { /* odd number of bits in the byte */ ++ line_parity ^= i; ++ line_parity_prime ^= ~i; ++ } ++ } ++ ++ ecc[2] = (~col_parity) | 0x03; ++ ++ t = 0; ++ if (line_parity & 0x80) ++ t |= 0x80; ++ if (line_parity_prime & 0x80) ++ t |= 0x40; ++ if (line_parity & 0x40) ++ t |= 0x20; ++ if (line_parity_prime & 0x40) ++ t |= 0x10; ++ if (line_parity & 0x20) ++ t |= 0x08; ++ if (line_parity_prime & 0x20) ++ t |= 0x04; ++ if (line_parity & 0x10) ++ t |= 0x02; ++ if (line_parity_prime & 0x10) ++ t |= 0x01; ++ ecc[1] = ~t; ++ ++ t = 0; ++ if (line_parity & 0x08) ++ t |= 0x80; ++ if (line_parity_prime & 0x08) ++ t |= 0x40; ++ if (line_parity & 0x04) ++ t |= 0x20; ++ if (line_parity_prime & 0x04) ++ t |= 0x10; ++ if (line_parity & 0x02) ++ t |= 0x08; ++ if (line_parity_prime & 0x02) ++ t |= 0x04; ++ if (line_parity & 0x01) ++ t |= 0x02; ++ if (line_parity_prime & 0x01) ++ t |= 0x01; ++ ecc[0] = ~t; ++ ++} ++ ++/* Correct the ECC on a 256 byte block of data */ ++ ++int yaffs_ecc_correct(unsigned char *data, unsigned char *read_ecc, ++ const unsigned char *test_ecc) ++{ ++ unsigned char d0, d1, d2; /* deltas */ ++ ++ d0 = read_ecc[0] ^ test_ecc[0]; ++ d1 = read_ecc[1] ^ test_ecc[1]; ++ d2 = read_ecc[2] ^ test_ecc[2]; ++ ++ if ((d0 | d1 | d2) == 0) ++ return 0; /* no error */ ++ ++ if (((d0 ^ (d0 >> 1)) & 0x55) == 0x55 && ++ ((d1 ^ (d1 >> 1)) & 0x55) == 0x55 && ++ ((d2 ^ (d2 >> 1)) & 0x54) == 0x54) { ++ /* Single bit (recoverable) error in data */ ++ ++ unsigned byte; ++ unsigned bit; ++ ++ bit = byte = 0; ++ ++ if (d1 & 0x80) ++ byte |= 0x80; ++ if (d1 & 0x20) ++ byte |= 0x40; ++ if (d1 & 0x08) ++ byte |= 0x20; ++ if (d1 & 0x02) ++ byte |= 0x10; ++ if (d0 & 0x80) ++ byte |= 0x08; ++ if (d0 & 0x20) ++ byte |= 0x04; ++ if (d0 & 0x08) ++ byte |= 0x02; ++ if (d0 & 0x02) ++ byte |= 0x01; ++ ++ if (d2 & 0x80) ++ bit |= 0x04; ++ if (d2 & 0x20) ++ bit |= 0x02; ++ if (d2 & 0x08) ++ bit |= 0x01; ++ ++ data[byte] ^= (1 << bit); ++ ++ return 1; /* Corrected the error */ ++ } ++ ++ if ((hweight8(d0) + hweight8(d1) + hweight8(d2)) == 1) { ++ /* Reccoverable error in ecc */ ++ ++ read_ecc[0] = test_ecc[0]; ++ read_ecc[1] = test_ecc[1]; ++ read_ecc[2] = test_ecc[2]; ++ ++ return 1; /* Corrected the error */ ++ } ++ ++ /* Unrecoverable error */ ++ ++ return -1; ++ ++} ++ ++/* ++ * ECCxxxOther does ECC calcs on arbitrary n bytes of data ++ */ ++void yaffs_ecc_calc_other(const unsigned char *data, unsigned n_bytes, ++ struct yaffs_ecc_other *ecc_other) ++{ ++ unsigned int i; ++ unsigned char col_parity = 0; ++ unsigned line_parity = 0; ++ unsigned line_parity_prime = 0; ++ unsigned char b; ++ ++ for (i = 0; i < n_bytes; i++) { ++ b = column_parity_table[*data++]; ++ col_parity ^= b; ++ ++ if (b & 0x01) { ++ /* odd number of bits in the byte */ ++ line_parity ^= i; ++ line_parity_prime ^= ~i; ++ } ++ ++ } ++ ++ ecc_other->col_parity = (col_parity >> 2) & 0x3f; ++ ecc_other->line_parity = line_parity; ++ ecc_other->line_parity_prime = line_parity_prime; ++} ++ ++int yaffs_ecc_correct_other(unsigned char *data, unsigned n_bytes, ++ struct yaffs_ecc_other *read_ecc, ++ const struct yaffs_ecc_other *test_ecc) ++{ ++ unsigned char delta_col; /* column parity delta */ ++ unsigned delta_line; /* line parity delta */ ++ unsigned delta_line_prime; /* line parity delta */ ++ unsigned bit; ++ ++ delta_col = read_ecc->col_parity ^ test_ecc->col_parity; ++ delta_line = read_ecc->line_parity ^ test_ecc->line_parity; ++ delta_line_prime = ++ read_ecc->line_parity_prime ^ test_ecc->line_parity_prime; ++ ++ if ((delta_col | delta_line | delta_line_prime) == 0) ++ return 0; /* no error */ ++ ++ if (delta_line == ~delta_line_prime && ++ (((delta_col ^ (delta_col >> 1)) & 0x15) == 0x15)) { ++ /* Single bit (recoverable) error in data */ ++ ++ bit = 0; ++ ++ if (delta_col & 0x20) ++ bit |= 0x04; ++ if (delta_col & 0x08) ++ bit |= 0x02; ++ if (delta_col & 0x02) ++ bit |= 0x01; ++ ++ if (delta_line >= n_bytes) ++ return -1; ++ ++ data[delta_line] ^= (1 << bit); ++ ++ return 1; /* corrected */ ++ } ++ ++ if ((hweight32(delta_line) + ++ hweight32(delta_line_prime) + ++ hweight8(delta_col)) == 1) { ++ /* Reccoverable error in ecc */ ++ ++ *read_ecc = *test_ecc; ++ return 1; /* corrected */ ++ } ++ ++ /* Unrecoverable error */ ++ ++ return -1; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_ecc.h linux-3.14.4/fs/yaffs2/yaffs_ecc.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_ecc.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_ecc.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,44 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++/* ++ * This code implements the ECC algorithm used in SmartMedia. ++ * ++ * The ECC comprises 22 bits of parity information and is stuffed into 3 bytes. ++ * The two unused bit are set to 1. ++ * The ECC can correct single bit errors in a 256-byte page of data. ++ * Thus, two such ECC blocks are used on a 512-byte NAND page. ++ * ++ */ ++ ++#ifndef __YAFFS_ECC_H__ ++#define __YAFFS_ECC_H__ ++ ++struct yaffs_ecc_other { ++ unsigned char col_parity; ++ unsigned line_parity; ++ unsigned line_parity_prime; ++}; ++ ++void yaffs_ecc_calc(const unsigned char *data, unsigned char *ecc); ++int yaffs_ecc_correct(unsigned char *data, unsigned char *read_ecc, ++ const unsigned char *test_ecc); ++ ++void yaffs_ecc_calc_other(const unsigned char *data, unsigned n_bytes, ++ struct yaffs_ecc_other *ecc); ++int yaffs_ecc_correct_other(unsigned char *data, unsigned n_bytes, ++ struct yaffs_ecc_other *read_ecc, ++ const struct yaffs_ecc_other *test_ecc); ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_getblockinfo.h linux-3.14.4/fs/yaffs2/yaffs_getblockinfo.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_getblockinfo.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_getblockinfo.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,35 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_GETBLOCKINFO_H__ ++#define __YAFFS_GETBLOCKINFO_H__ ++ ++#include "yaffs_guts.h" ++#include "yaffs_trace.h" ++ ++/* Function to manipulate block info */ ++static inline struct yaffs_block_info *yaffs_get_block_info(struct yaffs_dev ++ *dev, int blk) ++{ ++ if (blk < dev->internal_start_block || blk > dev->internal_end_block) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>> yaffs: get_block_info block %d is not valid", ++ blk); ++ BUG(); ++ } ++ return &dev->block_info[blk - dev->internal_start_block]; ++} ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_guts.c linux-3.14.4/fs/yaffs2/yaffs_guts.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_guts.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_guts.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,5146 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yportenv.h" ++#include "yaffs_trace.h" ++ ++#include "yaffs_guts.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_tagscompat.h" ++#include "yaffs_tagsmarshall.h" ++#include "yaffs_nand.h" ++#include "yaffs_yaffs1.h" ++#include "yaffs_yaffs2.h" ++#include "yaffs_bitmap.h" ++#include "yaffs_verify.h" ++#include "yaffs_nand.h" ++#include "yaffs_packedtags2.h" ++#include "yaffs_nameval.h" ++#include "yaffs_allocator.h" ++#include "yaffs_attribs.h" ++#include "yaffs_summary.h" ++ ++/* Note YAFFS_GC_GOOD_ENOUGH must be <= YAFFS_GC_PASSIVE_THRESHOLD */ ++#define YAFFS_GC_GOOD_ENOUGH 2 ++#define YAFFS_GC_PASSIVE_THRESHOLD 4 ++ ++#include "yaffs_ecc.h" ++ ++/* Forward declarations */ ++ ++static int yaffs_wr_data_obj(struct yaffs_obj *in, int inode_chunk, ++ const u8 *buffer, int n_bytes, int use_reserve); ++ ++static void yaffs_fix_null_name(struct yaffs_obj *obj, YCHAR *name, ++ int buffer_size); ++ ++/* Function to calculate chunk and offset */ ++ ++void yaffs_addr_to_chunk(struct yaffs_dev *dev, loff_t addr, ++ int *chunk_out, u32 *offset_out) ++{ ++ int chunk; ++ u32 offset; ++ ++ chunk = (u32) (addr >> dev->chunk_shift); ++ ++ if (dev->chunk_div == 1) { ++ /* easy power of 2 case */ ++ offset = (u32) (addr & dev->chunk_mask); ++ } else { ++ /* Non power-of-2 case */ ++ ++ loff_t chunk_base; ++ ++ chunk /= dev->chunk_div; ++ ++ chunk_base = ((loff_t) chunk) * dev->data_bytes_per_chunk; ++ offset = (u32) (addr - chunk_base); ++ } ++ ++ *chunk_out = chunk; ++ *offset_out = offset; ++} ++ ++/* Function to return the number of shifts for a power of 2 greater than or ++ * equal to the given number ++ * Note we don't try to cater for all possible numbers and this does not have to ++ * be hellishly efficient. ++ */ ++ ++static inline u32 calc_shifts_ceiling(u32 x) ++{ ++ int extra_bits; ++ int shifts; ++ ++ shifts = extra_bits = 0; ++ ++ while (x > 1) { ++ if (x & 1) ++ extra_bits++; ++ x >>= 1; ++ shifts++; ++ } ++ ++ if (extra_bits) ++ shifts++; ++ ++ return shifts; ++} ++ ++/* Function to return the number of shifts to get a 1 in bit 0 ++ */ ++ ++static inline u32 calc_shifts(u32 x) ++{ ++ u32 shifts; ++ ++ shifts = 0; ++ ++ if (!x) ++ return 0; ++ ++ while (!(x & 1)) { ++ x >>= 1; ++ shifts++; ++ } ++ ++ return shifts; ++} ++ ++/* ++ * Temporary buffer manipulations. ++ */ ++ ++static int yaffs_init_tmp_buffers(struct yaffs_dev *dev) ++{ ++ int i; ++ u8 *buf = (u8 *) 1; ++ ++ memset(dev->temp_buffer, 0, sizeof(dev->temp_buffer)); ++ ++ for (i = 0; buf && i < YAFFS_N_TEMP_BUFFERS; i++) { ++ dev->temp_buffer[i].in_use = 0; ++ buf = kmalloc(dev->param.total_bytes_per_chunk, GFP_NOFS); ++ dev->temp_buffer[i].buffer = buf; ++ } ++ ++ return buf ? YAFFS_OK : YAFFS_FAIL; ++} ++ ++u8 *yaffs_get_temp_buffer(struct yaffs_dev * dev) ++{ ++ int i; ++ ++ dev->temp_in_use++; ++ if (dev->temp_in_use > dev->max_temp) ++ dev->max_temp = dev->temp_in_use; ++ ++ for (i = 0; i < YAFFS_N_TEMP_BUFFERS; i++) { ++ if (dev->temp_buffer[i].in_use == 0) { ++ dev->temp_buffer[i].in_use = 1; ++ return dev->temp_buffer[i].buffer; ++ } ++ } ++ ++ yaffs_trace(YAFFS_TRACE_BUFFERS, "Out of temp buffers"); ++ /* ++ * If we got here then we have to allocate an unmanaged one ++ * This is not good. ++ */ ++ ++ dev->unmanaged_buffer_allocs++; ++ return kmalloc(dev->data_bytes_per_chunk, GFP_NOFS); ++ ++} ++ ++void yaffs_release_temp_buffer(struct yaffs_dev *dev, u8 *buffer) ++{ ++ int i; ++ ++ dev->temp_in_use--; ++ ++ for (i = 0; i < YAFFS_N_TEMP_BUFFERS; i++) { ++ if (dev->temp_buffer[i].buffer == buffer) { ++ dev->temp_buffer[i].in_use = 0; ++ return; ++ } ++ } ++ ++ if (buffer) { ++ /* assume it is an unmanaged one. */ ++ yaffs_trace(YAFFS_TRACE_BUFFERS, ++ "Releasing unmanaged temp buffer"); ++ kfree(buffer); ++ dev->unmanaged_buffer_deallocs++; ++ } ++ ++} ++ ++/* ++ * Functions for robustisizing TODO ++ * ++ */ ++ ++static void yaffs_handle_chunk_wr_ok(struct yaffs_dev *dev, int nand_chunk, ++ const u8 *data, ++ const struct yaffs_ext_tags *tags) ++{ ++ (void) dev; ++ (void) nand_chunk; ++ (void) data; ++ (void) tags; ++} ++ ++static void yaffs_handle_chunk_update(struct yaffs_dev *dev, int nand_chunk, ++ const struct yaffs_ext_tags *tags) ++{ ++ (void) dev; ++ (void) nand_chunk; ++ (void) tags; ++} ++ ++void yaffs_handle_chunk_error(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi) ++{ ++ if (!bi->gc_prioritise) { ++ bi->gc_prioritise = 1; ++ dev->has_pending_prioritised_gc = 1; ++ bi->chunk_error_strikes++; ++ ++ if (bi->chunk_error_strikes > 3) { ++ bi->needs_retiring = 1; /* Too many stikes, so retire */ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: Block struck out"); ++ ++ } ++ } ++} ++ ++static void yaffs_handle_chunk_wr_error(struct yaffs_dev *dev, int nand_chunk, ++ int erased_ok) ++{ ++ int flash_block = nand_chunk / dev->param.chunks_per_block; ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, flash_block); ++ ++ yaffs_handle_chunk_error(dev, bi); ++ ++ if (erased_ok) { ++ /* Was an actual write failure, ++ * so mark the block for retirement.*/ ++ bi->needs_retiring = 1; ++ yaffs_trace(YAFFS_TRACE_ERROR | YAFFS_TRACE_BAD_BLOCKS, ++ "**>> Block %d needs retiring", flash_block); ++ } ++ ++ /* Delete the chunk */ ++ yaffs_chunk_del(dev, nand_chunk, 1, __LINE__); ++ yaffs_skip_rest_of_block(dev); ++} ++ ++/* ++ * Verification code ++ */ ++ ++/* ++ * Simple hash function. Needs to have a reasonable spread ++ */ ++ ++static inline int yaffs_hash_fn(int n) ++{ ++ if (n < 0) ++ n = -n; ++ return n % YAFFS_NOBJECT_BUCKETS; ++} ++ ++/* ++ * Access functions to useful fake objects. ++ * Note that root might have a presence in NAND if permissions are set. ++ */ ++ ++struct yaffs_obj *yaffs_root(struct yaffs_dev *dev) ++{ ++ return dev->root_dir; ++} ++ ++struct yaffs_obj *yaffs_lost_n_found(struct yaffs_dev *dev) ++{ ++ return dev->lost_n_found; ++} ++ ++/* ++ * Erased NAND checking functions ++ */ ++ ++int yaffs_check_ff(u8 *buffer, int n_bytes) ++{ ++ /* Horrible, slow implementation */ ++ while (n_bytes--) { ++ if (*buffer != 0xff) ++ return 0; ++ buffer++; ++ } ++ return 1; ++} ++ ++static int yaffs_check_chunk_erased(struct yaffs_dev *dev, int nand_chunk) ++{ ++ int retval = YAFFS_OK; ++ u8 *data = yaffs_get_temp_buffer(dev); ++ struct yaffs_ext_tags tags; ++ int result; ++ ++ result = yaffs_rd_chunk_tags_nand(dev, nand_chunk, data, &tags); ++ ++ if (tags.ecc_result > YAFFS_ECC_RESULT_NO_ERROR) ++ retval = YAFFS_FAIL; ++ ++ if (!yaffs_check_ff(data, dev->data_bytes_per_chunk) || ++ tags.chunk_used) { ++ yaffs_trace(YAFFS_TRACE_NANDACCESS, ++ "Chunk %d not erased", nand_chunk); ++ retval = YAFFS_FAIL; ++ } ++ ++ yaffs_release_temp_buffer(dev, data); ++ ++ return retval; ++ ++} ++ ++static int yaffs_verify_chunk_written(struct yaffs_dev *dev, ++ int nand_chunk, ++ const u8 *data, ++ struct yaffs_ext_tags *tags) ++{ ++ int retval = YAFFS_OK; ++ struct yaffs_ext_tags temp_tags; ++ u8 *buffer = yaffs_get_temp_buffer(dev); ++ int result; ++ ++ result = yaffs_rd_chunk_tags_nand(dev, nand_chunk, buffer, &temp_tags); ++ if (memcmp(buffer, data, dev->data_bytes_per_chunk) || ++ temp_tags.obj_id != tags->obj_id || ++ temp_tags.chunk_id != tags->chunk_id || ++ temp_tags.n_bytes != tags->n_bytes) ++ retval = YAFFS_FAIL; ++ ++ yaffs_release_temp_buffer(dev, buffer); ++ ++ return retval; ++} ++ ++ ++int yaffs_check_alloc_available(struct yaffs_dev *dev, int n_chunks) ++{ ++ int reserved_chunks; ++ int reserved_blocks = dev->param.n_reserved_blocks; ++ int checkpt_blocks; ++ ++ checkpt_blocks = yaffs_calc_checkpt_blocks_required(dev); ++ ++ reserved_chunks = ++ (reserved_blocks + checkpt_blocks) * dev->param.chunks_per_block; ++ ++ return (dev->n_free_chunks > (reserved_chunks + n_chunks)); ++} ++ ++static int yaffs_find_alloc_block(struct yaffs_dev *dev) ++{ ++ int i; ++ struct yaffs_block_info *bi; ++ ++ if (dev->n_erased_blocks < 1) { ++ /* Hoosterman we've got a problem. ++ * Can't get space to gc ++ */ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: no more erased blocks"); ++ ++ return -1; ++ } ++ ++ /* Find an empty block. */ ++ ++ for (i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ dev->alloc_block_finder++; ++ if (dev->alloc_block_finder < dev->internal_start_block ++ || dev->alloc_block_finder > dev->internal_end_block) { ++ dev->alloc_block_finder = dev->internal_start_block; ++ } ++ ++ bi = yaffs_get_block_info(dev, dev->alloc_block_finder); ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_EMPTY) { ++ bi->block_state = YAFFS_BLOCK_STATE_ALLOCATING; ++ dev->seq_number++; ++ bi->seq_number = dev->seq_number; ++ dev->n_erased_blocks--; ++ yaffs_trace(YAFFS_TRACE_ALLOCATE, ++ "Allocated block %d, seq %d, %d left" , ++ dev->alloc_block_finder, dev->seq_number, ++ dev->n_erased_blocks); ++ return dev->alloc_block_finder; ++ } ++ } ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs tragedy: no more erased blocks, but there should have been %d", ++ dev->n_erased_blocks); ++ ++ return -1; ++} ++ ++static int yaffs_alloc_chunk(struct yaffs_dev *dev, int use_reserver, ++ struct yaffs_block_info **block_ptr) ++{ ++ int ret_val; ++ struct yaffs_block_info *bi; ++ ++ if (dev->alloc_block < 0) { ++ /* Get next block to allocate off */ ++ dev->alloc_block = yaffs_find_alloc_block(dev); ++ dev->alloc_page = 0; ++ } ++ ++ if (!use_reserver && !yaffs_check_alloc_available(dev, 1)) { ++ /* No space unless we're allowed to use the reserve. */ ++ return -1; ++ } ++ ++ if (dev->n_erased_blocks < dev->param.n_reserved_blocks ++ && dev->alloc_page == 0) ++ yaffs_trace(YAFFS_TRACE_ALLOCATE, "Allocating reserve"); ++ ++ /* Next page please.... */ ++ if (dev->alloc_block >= 0) { ++ bi = yaffs_get_block_info(dev, dev->alloc_block); ++ ++ ret_val = (dev->alloc_block * dev->param.chunks_per_block) + ++ dev->alloc_page; ++ bi->pages_in_use++; ++ yaffs_set_chunk_bit(dev, dev->alloc_block, dev->alloc_page); ++ ++ dev->alloc_page++; ++ ++ dev->n_free_chunks--; ++ ++ /* If the block is full set the state to full */ ++ if (dev->alloc_page >= dev->param.chunks_per_block) { ++ bi->block_state = YAFFS_BLOCK_STATE_FULL; ++ dev->alloc_block = -1; ++ } ++ ++ if (block_ptr) ++ *block_ptr = bi; ++ ++ return ret_val; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "!!!!!!!!! Allocator out !!!!!!!!!!!!!!!!!"); ++ ++ return -1; ++} ++ ++static int yaffs_get_erased_chunks(struct yaffs_dev *dev) ++{ ++ int n; ++ ++ n = dev->n_erased_blocks * dev->param.chunks_per_block; ++ ++ if (dev->alloc_block > 0) ++ n += (dev->param.chunks_per_block - dev->alloc_page); ++ ++ return n; ++ ++} ++ ++/* ++ * yaffs_skip_rest_of_block() skips over the rest of the allocation block ++ * if we don't want to write to it. ++ */ ++void yaffs_skip_rest_of_block(struct yaffs_dev *dev) ++{ ++ struct yaffs_block_info *bi; ++ ++ if (dev->alloc_block > 0) { ++ bi = yaffs_get_block_info(dev, dev->alloc_block); ++ if (bi->block_state == YAFFS_BLOCK_STATE_ALLOCATING) { ++ bi->block_state = YAFFS_BLOCK_STATE_FULL; ++ dev->alloc_block = -1; ++ } ++ } ++} ++ ++static int yaffs_write_new_chunk(struct yaffs_dev *dev, ++ const u8 *data, ++ struct yaffs_ext_tags *tags, int use_reserver) ++{ ++ int attempts = 0; ++ int write_ok = 0; ++ int chunk; ++ ++ yaffs2_checkpt_invalidate(dev); ++ ++ do { ++ struct yaffs_block_info *bi = 0; ++ int erased_ok = 0; ++ ++ chunk = yaffs_alloc_chunk(dev, use_reserver, &bi); ++ if (chunk < 0) { ++ /* no space */ ++ break; ++ } ++ ++ /* First check this chunk is erased, if it needs ++ * checking. The checking policy (unless forced ++ * always on) is as follows: ++ * ++ * Check the first page we try to write in a block. ++ * If the check passes then we don't need to check any ++ * more. If the check fails, we check again... ++ * If the block has been erased, we don't need to check. ++ * ++ * However, if the block has been prioritised for gc, ++ * then we think there might be something odd about ++ * this block and stop using it. ++ * ++ * Rationale: We should only ever see chunks that have ++ * not been erased if there was a partially written ++ * chunk due to power loss. This checking policy should ++ * catch that case with very few checks and thus save a ++ * lot of checks that are most likely not needed. ++ * ++ * Mods to the above ++ * If an erase check fails or the write fails we skip the ++ * rest of the block. ++ */ ++ ++ /* let's give it a try */ ++ attempts++; ++ ++ if (dev->param.always_check_erased) ++ bi->skip_erased_check = 0; ++ ++ if (!bi->skip_erased_check) { ++ erased_ok = yaffs_check_chunk_erased(dev, chunk); ++ if (erased_ok != YAFFS_OK) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>> yaffs chunk %d was not erased", ++ chunk); ++ ++ /* If not erased, delete this one, ++ * skip rest of block and ++ * try another chunk */ ++ yaffs_chunk_del(dev, chunk, 1, __LINE__); ++ yaffs_skip_rest_of_block(dev); ++ continue; ++ } ++ } ++ ++ write_ok = yaffs_wr_chunk_tags_nand(dev, chunk, data, tags); ++ ++ if (!bi->skip_erased_check) ++ write_ok = ++ yaffs_verify_chunk_written(dev, chunk, data, tags); ++ ++ if (write_ok != YAFFS_OK) { ++ /* Clean up aborted write, skip to next block and ++ * try another chunk */ ++ yaffs_handle_chunk_wr_error(dev, chunk, erased_ok); ++ continue; ++ } ++ ++ bi->skip_erased_check = 1; ++ ++ /* Copy the data into the robustification buffer */ ++ yaffs_handle_chunk_wr_ok(dev, chunk, data, tags); ++ ++ } while (write_ok != YAFFS_OK && ++ (yaffs_wr_attempts <= 0 || attempts <= yaffs_wr_attempts)); ++ ++ if (!write_ok) ++ chunk = -1; ++ ++ if (attempts > 1) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>> yaffs write required %d attempts", ++ attempts); ++ dev->n_retried_writes += (attempts - 1); ++ } ++ ++ return chunk; ++} ++ ++/* ++ * Block retiring for handling a broken block. ++ */ ++ ++static void yaffs_retire_block(struct yaffs_dev *dev, int flash_block) ++{ ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, flash_block); ++ ++ yaffs2_checkpt_invalidate(dev); ++ ++ yaffs2_clear_oldest_dirty_seq(dev, bi); ++ ++ if (yaffs_mark_bad(dev, flash_block) != YAFFS_OK) { ++ if (yaffs_erase_block(dev, flash_block) != YAFFS_OK) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: Failed to mark bad and erase block %d", ++ flash_block); ++ } else { ++ struct yaffs_ext_tags tags; ++ int chunk_id = ++ flash_block * dev->param.chunks_per_block; ++ ++ u8 *buffer = yaffs_get_temp_buffer(dev); ++ ++ memset(buffer, 0xff, dev->data_bytes_per_chunk); ++ memset(&tags, 0, sizeof(tags)); ++ tags.seq_number = YAFFS_SEQUENCE_BAD_BLOCK; ++ if (dev->tagger.write_chunk_tags_fn(dev, chunk_id - ++ dev->chunk_offset, ++ buffer, ++ &tags) != YAFFS_OK) ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: Failed to write bad block marker to block %d", ++ flash_block); ++ ++ yaffs_release_temp_buffer(dev, buffer); ++ } ++ } ++ ++ bi->block_state = YAFFS_BLOCK_STATE_DEAD; ++ bi->gc_prioritise = 0; ++ bi->needs_retiring = 0; ++ ++ dev->n_retired_blocks++; ++} ++ ++/*---------------- Name handling functions ------------*/ ++ ++static u16 yaffs_calc_name_sum(const YCHAR *name) ++{ ++ u16 sum = 0; ++ u16 i = 1; ++ ++ if (!name) ++ return 0; ++ ++ while ((*name) && i < (YAFFS_MAX_NAME_LENGTH / 2)) { ++ ++ /* 0x1f mask is case insensitive */ ++ sum += ((*name) & 0x1f) * i; ++ i++; ++ name++; ++ } ++ return sum; ++} ++ ++ ++void yaffs_set_obj_name(struct yaffs_obj *obj, const YCHAR * name) ++{ ++ memset(obj->short_name, 0, sizeof(obj->short_name)); ++ ++ if (name && !name[0]) { ++ yaffs_fix_null_name(obj, obj->short_name, ++ YAFFS_SHORT_NAME_LENGTH); ++ name = obj->short_name; ++ } else if (name && ++ strnlen(name, YAFFS_SHORT_NAME_LENGTH + 1) <= ++ YAFFS_SHORT_NAME_LENGTH) { ++ strcpy(obj->short_name, name); ++ } ++ ++ obj->sum = yaffs_calc_name_sum(name); ++} ++ ++void yaffs_set_obj_name_from_oh(struct yaffs_obj *obj, ++ const struct yaffs_obj_hdr *oh) ++{ ++#ifdef CONFIG_YAFFS_AUTO_UNICODE ++ YCHAR tmp_name[YAFFS_MAX_NAME_LENGTH + 1]; ++ memset(tmp_name, 0, sizeof(tmp_name)); ++ yaffs_load_name_from_oh(obj->my_dev, tmp_name, oh->name, ++ YAFFS_MAX_NAME_LENGTH + 1); ++ yaffs_set_obj_name(obj, tmp_name); ++#else ++ yaffs_set_obj_name(obj, oh->name); ++#endif ++} ++ ++loff_t yaffs_max_file_size(struct yaffs_dev *dev) ++{ ++ if(sizeof(loff_t) < 8) ++ return YAFFS_MAX_FILE_SIZE_32; ++ else ++ return ((loff_t) YAFFS_MAX_CHUNK_ID) * dev->data_bytes_per_chunk; ++} ++ ++/*-------------------- TNODES ------------------- ++ ++ * List of spare tnodes ++ * The list is hooked together using the first pointer ++ * in the tnode. ++ */ ++ ++struct yaffs_tnode *yaffs_get_tnode(struct yaffs_dev *dev) ++{ ++ struct yaffs_tnode *tn = yaffs_alloc_raw_tnode(dev); ++ ++ if (tn) { ++ memset(tn, 0, dev->tnode_size); ++ dev->n_tnodes++; ++ } ++ ++ dev->checkpoint_blocks_required = 0; /* force recalculation */ ++ ++ return tn; ++} ++ ++/* FreeTnode frees up a tnode and puts it back on the free list */ ++static void yaffs_free_tnode(struct yaffs_dev *dev, struct yaffs_tnode *tn) ++{ ++ yaffs_free_raw_tnode(dev, tn); ++ dev->n_tnodes--; ++ dev->checkpoint_blocks_required = 0; /* force recalculation */ ++} ++ ++static void yaffs_deinit_tnodes_and_objs(struct yaffs_dev *dev) ++{ ++ yaffs_deinit_raw_tnodes_and_objs(dev); ++ dev->n_obj = 0; ++ dev->n_tnodes = 0; ++} ++ ++static void yaffs_load_tnode_0(struct yaffs_dev *dev, struct yaffs_tnode *tn, ++ unsigned pos, unsigned val) ++{ ++ u32 *map = (u32 *) tn; ++ u32 bit_in_map; ++ u32 bit_in_word; ++ u32 word_in_map; ++ u32 mask; ++ ++ pos &= YAFFS_TNODES_LEVEL0_MASK; ++ val >>= dev->chunk_grp_bits; ++ ++ bit_in_map = pos * dev->tnode_width; ++ word_in_map = bit_in_map / 32; ++ bit_in_word = bit_in_map & (32 - 1); ++ ++ mask = dev->tnode_mask << bit_in_word; ++ ++ map[word_in_map] &= ~mask; ++ map[word_in_map] |= (mask & (val << bit_in_word)); ++ ++ if (dev->tnode_width > (32 - bit_in_word)) { ++ bit_in_word = (32 - bit_in_word); ++ word_in_map++; ++ mask = ++ dev->tnode_mask >> bit_in_word; ++ map[word_in_map] &= ~mask; ++ map[word_in_map] |= (mask & (val >> bit_in_word)); ++ } ++} ++ ++u32 yaffs_get_group_base(struct yaffs_dev *dev, struct yaffs_tnode *tn, ++ unsigned pos) ++{ ++ u32 *map = (u32 *) tn; ++ u32 bit_in_map; ++ u32 bit_in_word; ++ u32 word_in_map; ++ u32 val; ++ ++ pos &= YAFFS_TNODES_LEVEL0_MASK; ++ ++ bit_in_map = pos * dev->tnode_width; ++ word_in_map = bit_in_map / 32; ++ bit_in_word = bit_in_map & (32 - 1); ++ ++ val = map[word_in_map] >> bit_in_word; ++ ++ if (dev->tnode_width > (32 - bit_in_word)) { ++ bit_in_word = (32 - bit_in_word); ++ word_in_map++; ++ val |= (map[word_in_map] << bit_in_word); ++ } ++ ++ val &= dev->tnode_mask; ++ val <<= dev->chunk_grp_bits; ++ ++ return val; ++} ++ ++/* ------------------- End of individual tnode manipulation -----------------*/ ++ ++/* ---------Functions to manipulate the look-up tree (made up of tnodes) ------ ++ * The look up tree is represented by the top tnode and the number of top_level ++ * in the tree. 0 means only the level 0 tnode is in the tree. ++ */ ++ ++/* FindLevel0Tnode finds the level 0 tnode, if one exists. */ ++struct yaffs_tnode *yaffs_find_tnode_0(struct yaffs_dev *dev, ++ struct yaffs_file_var *file_struct, ++ u32 chunk_id) ++{ ++ struct yaffs_tnode *tn = file_struct->top; ++ u32 i; ++ int required_depth; ++ int level = file_struct->top_level; ++ ++ (void) dev; ++ ++ /* Check sane level and chunk Id */ ++ if (level < 0 || level > YAFFS_TNODES_MAX_LEVEL) ++ return NULL; ++ ++ if (chunk_id > YAFFS_MAX_CHUNK_ID) ++ return NULL; ++ ++ /* First check we're tall enough (ie enough top_level) */ ++ ++ i = chunk_id >> YAFFS_TNODES_LEVEL0_BITS; ++ required_depth = 0; ++ while (i) { ++ i >>= YAFFS_TNODES_INTERNAL_BITS; ++ required_depth++; ++ } ++ ++ if (required_depth > file_struct->top_level) ++ return NULL; /* Not tall enough, so we can't find it */ ++ ++ /* Traverse down to level 0 */ ++ while (level > 0 && tn) { ++ tn = tn->internal[(chunk_id >> ++ (YAFFS_TNODES_LEVEL0_BITS + ++ (level - 1) * ++ YAFFS_TNODES_INTERNAL_BITS)) & ++ YAFFS_TNODES_INTERNAL_MASK]; ++ level--; ++ } ++ ++ return tn; ++} ++ ++/* add_find_tnode_0 finds the level 0 tnode if it exists, ++ * otherwise first expands the tree. ++ * This happens in two steps: ++ * 1. If the tree isn't tall enough, then make it taller. ++ * 2. Scan down the tree towards the level 0 tnode adding tnodes if required. ++ * ++ * Used when modifying the tree. ++ * ++ * If the tn argument is NULL, then a fresh tnode will be added otherwise the ++ * specified tn will be plugged into the ttree. ++ */ ++ ++struct yaffs_tnode *yaffs_add_find_tnode_0(struct yaffs_dev *dev, ++ struct yaffs_file_var *file_struct, ++ u32 chunk_id, ++ struct yaffs_tnode *passed_tn) ++{ ++ int required_depth; ++ int i; ++ int l; ++ struct yaffs_tnode *tn; ++ u32 x; ++ ++ /* Check sane level and page Id */ ++ if (file_struct->top_level < 0 || ++ file_struct->top_level > YAFFS_TNODES_MAX_LEVEL) ++ return NULL; ++ ++ if (chunk_id > YAFFS_MAX_CHUNK_ID) ++ return NULL; ++ ++ /* First check we're tall enough (ie enough top_level) */ ++ ++ x = chunk_id >> YAFFS_TNODES_LEVEL0_BITS; ++ required_depth = 0; ++ while (x) { ++ x >>= YAFFS_TNODES_INTERNAL_BITS; ++ required_depth++; ++ } ++ ++ if (required_depth > file_struct->top_level) { ++ /* Not tall enough, gotta make the tree taller */ ++ for (i = file_struct->top_level; i < required_depth; i++) { ++ ++ tn = yaffs_get_tnode(dev); ++ ++ if (tn) { ++ tn->internal[0] = file_struct->top; ++ file_struct->top = tn; ++ file_struct->top_level++; ++ } else { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs: no more tnodes"); ++ return NULL; ++ } ++ } ++ } ++ ++ /* Traverse down to level 0, adding anything we need */ ++ ++ l = file_struct->top_level; ++ tn = file_struct->top; ++ ++ if (l > 0) { ++ while (l > 0 && tn) { ++ x = (chunk_id >> ++ (YAFFS_TNODES_LEVEL0_BITS + ++ (l - 1) * YAFFS_TNODES_INTERNAL_BITS)) & ++ YAFFS_TNODES_INTERNAL_MASK; ++ ++ if ((l > 1) && !tn->internal[x]) { ++ /* Add missing non-level-zero tnode */ ++ tn->internal[x] = yaffs_get_tnode(dev); ++ if (!tn->internal[x]) ++ return NULL; ++ } else if (l == 1) { ++ /* Looking from level 1 at level 0 */ ++ if (passed_tn) { ++ /* If we already have one, release it */ ++ if (tn->internal[x]) ++ yaffs_free_tnode(dev, ++ tn->internal[x]); ++ tn->internal[x] = passed_tn; ++ ++ } else if (!tn->internal[x]) { ++ /* Don't have one, none passed in */ ++ tn->internal[x] = yaffs_get_tnode(dev); ++ if (!tn->internal[x]) ++ return NULL; ++ } ++ } ++ ++ tn = tn->internal[x]; ++ l--; ++ } ++ } else { ++ /* top is level 0 */ ++ if (passed_tn) { ++ memcpy(tn, passed_tn, ++ (dev->tnode_width * YAFFS_NTNODES_LEVEL0) / 8); ++ yaffs_free_tnode(dev, passed_tn); ++ } ++ } ++ ++ return tn; ++} ++ ++static int yaffs_tags_match(const struct yaffs_ext_tags *tags, int obj_id, ++ int chunk_obj) ++{ ++ return (tags->chunk_id == chunk_obj && ++ tags->obj_id == obj_id && ++ !tags->is_deleted) ? 1 : 0; ++ ++} ++ ++static int yaffs_find_chunk_in_group(struct yaffs_dev *dev, int the_chunk, ++ struct yaffs_ext_tags *tags, int obj_id, ++ int inode_chunk) ++{ ++ int j; ++ ++ for (j = 0; the_chunk && j < dev->chunk_grp_size; j++) { ++ if (yaffs_check_chunk_bit ++ (dev, the_chunk / dev->param.chunks_per_block, ++ the_chunk % dev->param.chunks_per_block)) { ++ ++ if (dev->chunk_grp_size == 1) ++ return the_chunk; ++ else { ++ yaffs_rd_chunk_tags_nand(dev, the_chunk, NULL, ++ tags); ++ if (yaffs_tags_match(tags, ++ obj_id, inode_chunk)) { ++ /* found it; */ ++ return the_chunk; ++ } ++ } ++ } ++ the_chunk++; ++ } ++ return -1; ++} ++ ++int yaffs_find_chunk_in_file(struct yaffs_obj *in, int inode_chunk, ++ struct yaffs_ext_tags *tags) ++{ ++ /*Get the Tnode, then get the level 0 offset chunk offset */ ++ struct yaffs_tnode *tn; ++ int the_chunk = -1; ++ struct yaffs_ext_tags local_tags; ++ int ret_val = -1; ++ struct yaffs_dev *dev = in->my_dev; ++ ++ if (!tags) { ++ /* Passed a NULL, so use our own tags space */ ++ tags = &local_tags; ++ } ++ ++ tn = yaffs_find_tnode_0(dev, &in->variant.file_variant, inode_chunk); ++ ++ if (!tn) ++ return ret_val; ++ ++ the_chunk = yaffs_get_group_base(dev, tn, inode_chunk); ++ ++ ret_val = yaffs_find_chunk_in_group(dev, the_chunk, tags, in->obj_id, ++ inode_chunk); ++ return ret_val; ++} ++ ++static int yaffs_find_del_file_chunk(struct yaffs_obj *in, int inode_chunk, ++ struct yaffs_ext_tags *tags) ++{ ++ /* Get the Tnode, then get the level 0 offset chunk offset */ ++ struct yaffs_tnode *tn; ++ int the_chunk = -1; ++ struct yaffs_ext_tags local_tags; ++ struct yaffs_dev *dev = in->my_dev; ++ int ret_val = -1; ++ ++ if (!tags) { ++ /* Passed a NULL, so use our own tags space */ ++ tags = &local_tags; ++ } ++ ++ tn = yaffs_find_tnode_0(dev, &in->variant.file_variant, inode_chunk); ++ ++ if (!tn) ++ return ret_val; ++ ++ the_chunk = yaffs_get_group_base(dev, tn, inode_chunk); ++ ++ ret_val = yaffs_find_chunk_in_group(dev, the_chunk, tags, in->obj_id, ++ inode_chunk); ++ ++ /* Delete the entry in the filestructure (if found) */ ++ if (ret_val != -1) ++ yaffs_load_tnode_0(dev, tn, inode_chunk, 0); ++ ++ return ret_val; ++} ++ ++int yaffs_put_chunk_in_file(struct yaffs_obj *in, int inode_chunk, ++ int nand_chunk, int in_scan) ++{ ++ /* NB in_scan is zero unless scanning. ++ * For forward scanning, in_scan is > 0; ++ * for backward scanning in_scan is < 0 ++ * ++ * nand_chunk = 0 is a dummy insert to make sure the tnodes are there. ++ */ ++ ++ struct yaffs_tnode *tn; ++ struct yaffs_dev *dev = in->my_dev; ++ int existing_cunk; ++ struct yaffs_ext_tags existing_tags; ++ struct yaffs_ext_tags new_tags; ++ unsigned existing_serial, new_serial; ++ ++ if (in->variant_type != YAFFS_OBJECT_TYPE_FILE) { ++ /* Just ignore an attempt at putting a chunk into a non-file ++ * during scanning. ++ * If it is not during Scanning then something went wrong! ++ */ ++ if (!in_scan) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy:attempt to put data chunk into a non-file" ++ ); ++ BUG(); ++ } ++ ++ yaffs_chunk_del(dev, nand_chunk, 1, __LINE__); ++ return YAFFS_OK; ++ } ++ ++ tn = yaffs_add_find_tnode_0(dev, ++ &in->variant.file_variant, ++ inode_chunk, NULL); ++ if (!tn) ++ return YAFFS_FAIL; ++ ++ if (!nand_chunk) ++ /* Dummy insert, bail now */ ++ return YAFFS_OK; ++ ++ existing_cunk = yaffs_get_group_base(dev, tn, inode_chunk); ++ ++ if (in_scan != 0) { ++ /* If we're scanning then we need to test for duplicates ++ * NB This does not need to be efficient since it should only ++ * happen when the power fails during a write, then only one ++ * chunk should ever be affected. ++ * ++ * Correction for YAFFS2: This could happen quite a lot and we ++ * need to think about efficiency! TODO ++ * Update: For backward scanning we don't need to re-read tags ++ * so this is quite cheap. ++ */ ++ ++ if (existing_cunk > 0) { ++ /* NB Right now existing chunk will not be real ++ * chunk_id if the chunk group size > 1 ++ * thus we have to do a FindChunkInFile to get the ++ * real chunk id. ++ * ++ * We have a duplicate now we need to decide which ++ * one to use: ++ * ++ * Backwards scanning YAFFS2: The old one is what ++ * we use, dump the new one. ++ * YAFFS1: Get both sets of tags and compare serial ++ * numbers. ++ */ ++ ++ if (in_scan > 0) { ++ /* Only do this for forward scanning */ ++ yaffs_rd_chunk_tags_nand(dev, ++ nand_chunk, ++ NULL, &new_tags); ++ ++ /* Do a proper find */ ++ existing_cunk = ++ yaffs_find_chunk_in_file(in, inode_chunk, ++ &existing_tags); ++ } ++ ++ if (existing_cunk <= 0) { ++ /*Hoosterman - how did this happen? */ ++ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: existing chunk < 0 in scan" ++ ); ++ ++ } ++ ++ /* NB The deleted flags should be false, otherwise ++ * the chunks will not be loaded during a scan ++ */ ++ ++ if (in_scan > 0) { ++ new_serial = new_tags.serial_number; ++ existing_serial = existing_tags.serial_number; ++ } ++ ++ if ((in_scan > 0) && ++ (existing_cunk <= 0 || ++ ((existing_serial + 1) & 3) == new_serial)) { ++ /* Forward scanning. ++ * Use new ++ * Delete the old one and drop through to ++ * update the tnode ++ */ ++ yaffs_chunk_del(dev, existing_cunk, 1, ++ __LINE__); ++ } else { ++ /* Backward scanning or we want to use the ++ * existing one ++ * Delete the new one and return early so that ++ * the tnode isn't changed ++ */ ++ yaffs_chunk_del(dev, nand_chunk, 1, __LINE__); ++ return YAFFS_OK; ++ } ++ } ++ ++ } ++ ++ if (existing_cunk == 0) ++ in->n_data_chunks++; ++ ++ yaffs_load_tnode_0(dev, tn, inode_chunk, nand_chunk); ++ ++ return YAFFS_OK; ++} ++ ++static void yaffs_soft_del_chunk(struct yaffs_dev *dev, int chunk) ++{ ++ struct yaffs_block_info *the_block; ++ unsigned block_no; ++ ++ yaffs_trace(YAFFS_TRACE_DELETION, "soft delete chunk %d", chunk); ++ ++ block_no = chunk / dev->param.chunks_per_block; ++ the_block = yaffs_get_block_info(dev, block_no); ++ if (the_block) { ++ the_block->soft_del_pages++; ++ dev->n_free_chunks++; ++ yaffs2_update_oldest_dirty_seq(dev, block_no, the_block); ++ } ++} ++ ++/* SoftDeleteWorker scans backwards through the tnode tree and soft deletes all ++ * the chunks in the file. ++ * All soft deleting does is increment the block's softdelete count and pulls ++ * the chunk out of the tnode. ++ * Thus, essentially this is the same as DeleteWorker except that the chunks ++ * are soft deleted. ++ */ ++ ++static int yaffs_soft_del_worker(struct yaffs_obj *in, struct yaffs_tnode *tn, ++ u32 level, int chunk_offset) ++{ ++ int i; ++ int the_chunk; ++ int all_done = 1; ++ struct yaffs_dev *dev = in->my_dev; ++ ++ if (!tn) ++ return 1; ++ ++ if (level > 0) { ++ for (i = YAFFS_NTNODES_INTERNAL - 1; ++ all_done && i >= 0; ++ i--) { ++ if (tn->internal[i]) { ++ all_done = ++ yaffs_soft_del_worker(in, ++ tn->internal[i], ++ level - 1, ++ (chunk_offset << ++ YAFFS_TNODES_INTERNAL_BITS) ++ + i); ++ if (all_done) { ++ yaffs_free_tnode(dev, ++ tn->internal[i]); ++ tn->internal[i] = NULL; ++ } else { ++ /* Can this happen? */ ++ } ++ } ++ } ++ return (all_done) ? 1 : 0; ++ } ++ ++ /* level 0 */ ++ for (i = YAFFS_NTNODES_LEVEL0 - 1; i >= 0; i--) { ++ the_chunk = yaffs_get_group_base(dev, tn, i); ++ if (the_chunk) { ++ yaffs_soft_del_chunk(dev, the_chunk); ++ yaffs_load_tnode_0(dev, tn, i, 0); ++ } ++ } ++ return 1; ++} ++ ++static void yaffs_remove_obj_from_dir(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev = obj->my_dev; ++ struct yaffs_obj *parent; ++ ++ yaffs_verify_obj_in_dir(obj); ++ parent = obj->parent; ++ ++ yaffs_verify_dir(parent); ++ ++ if (dev && dev->param.remove_obj_fn) ++ dev->param.remove_obj_fn(obj); ++ ++ list_del_init(&obj->siblings); ++ obj->parent = NULL; ++ ++ yaffs_verify_dir(parent); ++} ++ ++void yaffs_add_obj_to_dir(struct yaffs_obj *directory, struct yaffs_obj *obj) ++{ ++ if (!directory) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "tragedy: Trying to add an object to a null pointer directory" ++ ); ++ BUG(); ++ return; ++ } ++ if (directory->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "tragedy: Trying to add an object to a non-directory" ++ ); ++ BUG(); ++ } ++ ++ if (obj->siblings.prev == NULL) { ++ /* Not initialised */ ++ BUG(); ++ } ++ ++ yaffs_verify_dir(directory); ++ ++ yaffs_remove_obj_from_dir(obj); ++ ++ /* Now add it */ ++ list_add(&obj->siblings, &directory->variant.dir_variant.children); ++ obj->parent = directory; ++ ++ if (directory == obj->my_dev->unlinked_dir ++ || directory == obj->my_dev->del_dir) { ++ obj->unlinked = 1; ++ obj->my_dev->n_unlinked_files++; ++ obj->rename_allowed = 0; ++ } ++ ++ yaffs_verify_dir(directory); ++ yaffs_verify_obj_in_dir(obj); ++} ++ ++static int yaffs_change_obj_name(struct yaffs_obj *obj, ++ struct yaffs_obj *new_dir, ++ const YCHAR *new_name, int force, int shadows) ++{ ++ int unlink_op; ++ int del_op; ++ struct yaffs_obj *existing_target; ++ ++ if (new_dir == NULL) ++ new_dir = obj->parent; /* use the old directory */ ++ ++ if (new_dir->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "tragedy: yaffs_change_obj_name: new_dir is not a directory" ++ ); ++ BUG(); ++ } ++ ++ unlink_op = (new_dir == obj->my_dev->unlinked_dir); ++ del_op = (new_dir == obj->my_dev->del_dir); ++ ++ existing_target = yaffs_find_by_name(new_dir, new_name); ++ ++ /* If the object is a file going into the unlinked directory, ++ * then it is OK to just stuff it in since duplicate names are OK. ++ * else only proceed if the new name does not exist and we're putting ++ * it into a directory. ++ */ ++ if (!(unlink_op || del_op || force || ++ shadows > 0 || !existing_target) || ++ new_dir->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) ++ return YAFFS_FAIL; ++ ++ yaffs_set_obj_name(obj, new_name); ++ obj->dirty = 1; ++ yaffs_add_obj_to_dir(new_dir, obj); ++ ++ if (unlink_op) ++ obj->unlinked = 1; ++ ++ /* If it is a deletion then we mark it as a shrink for gc */ ++ if (yaffs_update_oh(obj, new_name, 0, del_op, shadows, NULL) >= 0) ++ return YAFFS_OK; ++ ++ return YAFFS_FAIL; ++} ++ ++/*------------------------ Short Operations Cache ------------------------------ ++ * In many situations where there is no high level buffering a lot of ++ * reads might be short sequential reads, and a lot of writes may be short ++ * sequential writes. eg. scanning/writing a jpeg file. ++ * In these cases, a short read/write cache can provide a huge perfomance ++ * benefit with dumb-as-a-rock code. ++ * In Linux, the page cache provides read buffering and the short op cache ++ * provides write buffering. ++ * ++ * There are a small number (~10) of cache chunks per device so that we don't ++ * need a very intelligent search. ++ */ ++ ++static int yaffs_obj_cache_dirty(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev = obj->my_dev; ++ int i; ++ struct yaffs_cache *cache; ++ int n_caches = obj->my_dev->param.n_caches; ++ ++ for (i = 0; i < n_caches; i++) { ++ cache = &dev->cache[i]; ++ if (cache->object == obj && cache->dirty) ++ return 1; ++ } ++ ++ return 0; ++} ++ ++static void yaffs_flush_file_cache(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev = obj->my_dev; ++ int lowest = -99; /* Stop compiler whining. */ ++ int i; ++ struct yaffs_cache *cache; ++ int chunk_written = 0; ++ int n_caches = obj->my_dev->param.n_caches; ++ ++ if (n_caches < 1) ++ return; ++ do { ++ cache = NULL; ++ ++ /* Find the lowest dirty chunk for this object */ ++ for (i = 0; i < n_caches; i++) { ++ if (dev->cache[i].object == obj && ++ dev->cache[i].dirty) { ++ if (!cache || ++ dev->cache[i].chunk_id < lowest) { ++ cache = &dev->cache[i]; ++ lowest = cache->chunk_id; ++ } ++ } ++ } ++ ++ if (cache && !cache->locked) { ++ /* Write it out and free it up */ ++ chunk_written = ++ yaffs_wr_data_obj(cache->object, ++ cache->chunk_id, ++ cache->data, ++ cache->n_bytes, 1); ++ cache->dirty = 0; ++ cache->object = NULL; ++ } ++ } while (cache && chunk_written > 0); ++ ++ if (cache) ++ /* Hoosterman, disk full while writing cache out. */ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: no space during cache write"); ++} ++ ++/*yaffs_flush_whole_cache(dev) ++ * ++ * ++ */ ++ ++void yaffs_flush_whole_cache(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj; ++ int n_caches = dev->param.n_caches; ++ int i; ++ ++ /* Find a dirty object in the cache and flush it... ++ * until there are no further dirty objects. ++ */ ++ do { ++ obj = NULL; ++ for (i = 0; i < n_caches && !obj; i++) { ++ if (dev->cache[i].object && dev->cache[i].dirty) ++ obj = dev->cache[i].object; ++ } ++ if (obj) ++ yaffs_flush_file_cache(obj); ++ } while (obj); ++ ++} ++ ++/* Grab us a cache chunk for use. ++ * First look for an empty one. ++ * Then look for the least recently used non-dirty one. ++ * Then look for the least recently used dirty one...., flush and look again. ++ */ ++static struct yaffs_cache *yaffs_grab_chunk_worker(struct yaffs_dev *dev) ++{ ++ int i; ++ ++ if (dev->param.n_caches > 0) { ++ for (i = 0; i < dev->param.n_caches; i++) { ++ if (!dev->cache[i].object) ++ return &dev->cache[i]; ++ } ++ } ++ return NULL; ++} ++ ++static struct yaffs_cache *yaffs_grab_chunk_cache(struct yaffs_dev *dev) ++{ ++ struct yaffs_cache *cache; ++ struct yaffs_obj *the_obj; ++ int usage; ++ int i; ++ int pushout; ++ ++ if (dev->param.n_caches < 1) ++ return NULL; ++ ++ /* Try find a non-dirty one... */ ++ ++ cache = yaffs_grab_chunk_worker(dev); ++ ++ if (!cache) { ++ /* They were all dirty, find the LRU object and flush ++ * its cache, then find again. ++ * NB what's here is not very accurate, ++ * we actually flush the object with the LRU chunk. ++ */ ++ ++ /* With locking we can't assume we can use entry zero, ++ * Set the_obj to a valid pointer for Coverity. */ ++ the_obj = dev->cache[0].object; ++ usage = -1; ++ cache = NULL; ++ pushout = -1; ++ ++ for (i = 0; i < dev->param.n_caches; i++) { ++ if (dev->cache[i].object && ++ !dev->cache[i].locked && ++ (dev->cache[i].last_use < usage || ++ !cache)) { ++ usage = dev->cache[i].last_use; ++ the_obj = dev->cache[i].object; ++ cache = &dev->cache[i]; ++ pushout = i; ++ } ++ } ++ ++ if (!cache || cache->dirty) { ++ /* Flush and try again */ ++ yaffs_flush_file_cache(the_obj); ++ cache = yaffs_grab_chunk_worker(dev); ++ } ++ } ++ return cache; ++} ++ ++/* Find a cached chunk */ ++static struct yaffs_cache *yaffs_find_chunk_cache(const struct yaffs_obj *obj, ++ int chunk_id) ++{ ++ struct yaffs_dev *dev = obj->my_dev; ++ int i; ++ ++ if (dev->param.n_caches < 1) ++ return NULL; ++ ++ for (i = 0; i < dev->param.n_caches; i++) { ++ if (dev->cache[i].object == obj && ++ dev->cache[i].chunk_id == chunk_id) { ++ dev->cache_hits++; ++ ++ return &dev->cache[i]; ++ } ++ } ++ return NULL; ++} ++ ++/* Mark the chunk for the least recently used algorithym */ ++static void yaffs_use_cache(struct yaffs_dev *dev, struct yaffs_cache *cache, ++ int is_write) ++{ ++ int i; ++ ++ if (dev->param.n_caches < 1) ++ return; ++ ++ if (dev->cache_last_use < 0 || ++ dev->cache_last_use > 100000000) { ++ /* Reset the cache usages */ ++ for (i = 1; i < dev->param.n_caches; i++) ++ dev->cache[i].last_use = 0; ++ ++ dev->cache_last_use = 0; ++ } ++ dev->cache_last_use++; ++ cache->last_use = dev->cache_last_use; ++ ++ if (is_write) ++ cache->dirty = 1; ++} ++ ++/* Invalidate a single cache page. ++ * Do this when a whole page gets written, ++ * ie the short cache for this page is no longer valid. ++ */ ++static void yaffs_invalidate_chunk_cache(struct yaffs_obj *object, int chunk_id) ++{ ++ struct yaffs_cache *cache; ++ ++ if (object->my_dev->param.n_caches > 0) { ++ cache = yaffs_find_chunk_cache(object, chunk_id); ++ ++ if (cache) ++ cache->object = NULL; ++ } ++} ++ ++/* Invalidate all the cache pages associated with this object ++ * Do this whenever ther file is deleted or resized. ++ */ ++static void yaffs_invalidate_whole_cache(struct yaffs_obj *in) ++{ ++ int i; ++ struct yaffs_dev *dev = in->my_dev; ++ ++ if (dev->param.n_caches > 0) { ++ /* Invalidate it. */ ++ for (i = 0; i < dev->param.n_caches; i++) { ++ if (dev->cache[i].object == in) ++ dev->cache[i].object = NULL; ++ } ++ } ++} ++ ++static void yaffs_unhash_obj(struct yaffs_obj *obj) ++{ ++ int bucket; ++ struct yaffs_dev *dev = obj->my_dev; ++ ++ /* If it is still linked into the bucket list, free from the list */ ++ if (!list_empty(&obj->hash_link)) { ++ list_del_init(&obj->hash_link); ++ bucket = yaffs_hash_fn(obj->obj_id); ++ dev->obj_bucket[bucket].count--; ++ } ++} ++ ++/* FreeObject frees up a Object and puts it back on the free list */ ++static void yaffs_free_obj(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev; ++ ++ if (!obj) { ++ BUG(); ++ return; ++ } ++ dev = obj->my_dev; ++ yaffs_trace(YAFFS_TRACE_OS, "FreeObject %p inode %p", ++ obj, obj->my_inode); ++ if (obj->parent) ++ BUG(); ++ if (!list_empty(&obj->siblings)) ++ BUG(); ++ ++ if (obj->my_inode) { ++ /* We're still hooked up to a cached inode. ++ * Don't delete now, but mark for later deletion ++ */ ++ obj->defered_free = 1; ++ return; ++ } ++ ++ yaffs_unhash_obj(obj); ++ ++ yaffs_free_raw_obj(dev, obj); ++ dev->n_obj--; ++ dev->checkpoint_blocks_required = 0; /* force recalculation */ ++} ++ ++void yaffs_handle_defered_free(struct yaffs_obj *obj) ++{ ++ if (obj->defered_free) ++ yaffs_free_obj(obj); ++} ++ ++static int yaffs_generic_obj_del(struct yaffs_obj *in) ++{ ++ /* Iinvalidate the file's data in the cache, without flushing. */ ++ yaffs_invalidate_whole_cache(in); ++ ++ if (in->my_dev->param.is_yaffs2 && in->parent != in->my_dev->del_dir) { ++ /* Move to unlinked directory so we have a deletion record */ ++ yaffs_change_obj_name(in, in->my_dev->del_dir, _Y("deleted"), 0, ++ 0); ++ } ++ ++ yaffs_remove_obj_from_dir(in); ++ yaffs_chunk_del(in->my_dev, in->hdr_chunk, 1, __LINE__); ++ in->hdr_chunk = 0; ++ ++ yaffs_free_obj(in); ++ return YAFFS_OK; ++ ++} ++ ++static void yaffs_soft_del_file(struct yaffs_obj *obj) ++{ ++ if (!obj->deleted || ++ obj->variant_type != YAFFS_OBJECT_TYPE_FILE || ++ obj->soft_del) ++ return; ++ ++ if (obj->n_data_chunks <= 0) { ++ /* Empty file with no duplicate object headers, ++ * just delete it immediately */ ++ yaffs_free_tnode(obj->my_dev, obj->variant.file_variant.top); ++ obj->variant.file_variant.top = NULL; ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "yaffs: Deleting empty file %d", ++ obj->obj_id); ++ yaffs_generic_obj_del(obj); ++ } else { ++ yaffs_soft_del_worker(obj, ++ obj->variant.file_variant.top, ++ obj->variant. ++ file_variant.top_level, 0); ++ obj->soft_del = 1; ++ } ++} ++ ++/* Pruning removes any part of the file structure tree that is beyond the ++ * bounds of the file (ie that does not point to chunks). ++ * ++ * A file should only get pruned when its size is reduced. ++ * ++ * Before pruning, the chunks must be pulled from the tree and the ++ * level 0 tnode entries must be zeroed out. ++ * Could also use this for file deletion, but that's probably better handled ++ * by a special case. ++ * ++ * This function is recursive. For levels > 0 the function is called again on ++ * any sub-tree. For level == 0 we just check if the sub-tree has data. ++ * If there is no data in a subtree then it is pruned. ++ */ ++ ++static struct yaffs_tnode *yaffs_prune_worker(struct yaffs_dev *dev, ++ struct yaffs_tnode *tn, u32 level, ++ int del0) ++{ ++ int i; ++ int has_data; ++ ++ if (!tn) ++ return tn; ++ ++ has_data = 0; ++ ++ if (level > 0) { ++ for (i = 0; i < YAFFS_NTNODES_INTERNAL; i++) { ++ if (tn->internal[i]) { ++ tn->internal[i] = ++ yaffs_prune_worker(dev, ++ tn->internal[i], ++ level - 1, ++ (i == 0) ? del0 : 1); ++ } ++ ++ if (tn->internal[i]) ++ has_data++; ++ } ++ } else { ++ int tnode_size_u32 = dev->tnode_size / sizeof(u32); ++ u32 *map = (u32 *) tn; ++ ++ for (i = 0; !has_data && i < tnode_size_u32; i++) { ++ if (map[i]) ++ has_data++; ++ } ++ } ++ ++ if (has_data == 0 && del0) { ++ /* Free and return NULL */ ++ yaffs_free_tnode(dev, tn); ++ tn = NULL; ++ } ++ return tn; ++} ++ ++static int yaffs_prune_tree(struct yaffs_dev *dev, ++ struct yaffs_file_var *file_struct) ++{ ++ int i; ++ int has_data; ++ int done = 0; ++ struct yaffs_tnode *tn; ++ ++ if (file_struct->top_level < 1) ++ return YAFFS_OK; ++ ++ file_struct->top = ++ yaffs_prune_worker(dev, file_struct->top, file_struct->top_level, 0); ++ ++ /* Now we have a tree with all the non-zero branches NULL but ++ * the height is the same as it was. ++ * Let's see if we can trim internal tnodes to shorten the tree. ++ * We can do this if only the 0th element in the tnode is in use ++ * (ie all the non-zero are NULL) ++ */ ++ ++ while (file_struct->top_level && !done) { ++ tn = file_struct->top; ++ ++ has_data = 0; ++ for (i = 1; i < YAFFS_NTNODES_INTERNAL; i++) { ++ if (tn->internal[i]) ++ has_data++; ++ } ++ ++ if (!has_data) { ++ file_struct->top = tn->internal[0]; ++ file_struct->top_level--; ++ yaffs_free_tnode(dev, tn); ++ } else { ++ done = 1; ++ } ++ } ++ ++ return YAFFS_OK; ++} ++ ++/*-------------------- End of File Structure functions.-------------------*/ ++ ++/* alloc_empty_obj gets us a clean Object.*/ ++static struct yaffs_obj *yaffs_alloc_empty_obj(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj = yaffs_alloc_raw_obj(dev); ++ ++ if (!obj) ++ return obj; ++ ++ dev->n_obj++; ++ ++ /* Now sweeten it up... */ ++ ++ memset(obj, 0, sizeof(struct yaffs_obj)); ++ obj->being_created = 1; ++ ++ obj->my_dev = dev; ++ obj->hdr_chunk = 0; ++ obj->variant_type = YAFFS_OBJECT_TYPE_UNKNOWN; ++ INIT_LIST_HEAD(&(obj->hard_links)); ++ INIT_LIST_HEAD(&(obj->hash_link)); ++ INIT_LIST_HEAD(&obj->siblings); ++ ++ /* Now make the directory sane */ ++ if (dev->root_dir) { ++ obj->parent = dev->root_dir; ++ list_add(&(obj->siblings), ++ &dev->root_dir->variant.dir_variant.children); ++ } ++ ++ /* Add it to the lost and found directory. ++ * NB Can't put root or lost-n-found in lost-n-found so ++ * check if lost-n-found exists first ++ */ ++ if (dev->lost_n_found) ++ yaffs_add_obj_to_dir(dev->lost_n_found, obj); ++ ++ obj->being_created = 0; ++ ++ dev->checkpoint_blocks_required = 0; /* force recalculation */ ++ ++ return obj; ++} ++ ++static int yaffs_find_nice_bucket(struct yaffs_dev *dev) ++{ ++ int i; ++ int l = 999; ++ int lowest = 999999; ++ ++ /* Search for the shortest list or one that ++ * isn't too long. ++ */ ++ ++ for (i = 0; i < 10 && lowest > 4; i++) { ++ dev->bucket_finder++; ++ dev->bucket_finder %= YAFFS_NOBJECT_BUCKETS; ++ if (dev->obj_bucket[dev->bucket_finder].count < lowest) { ++ lowest = dev->obj_bucket[dev->bucket_finder].count; ++ l = dev->bucket_finder; ++ } ++ } ++ ++ return l; ++} ++ ++static int yaffs_new_obj_id(struct yaffs_dev *dev) ++{ ++ int bucket = yaffs_find_nice_bucket(dev); ++ int found = 0; ++ struct list_head *i; ++ u32 n = (u32) bucket; ++ ++ /* Now find an object value that has not already been taken ++ * by scanning the list. ++ */ ++ ++ while (!found) { ++ found = 1; ++ n += YAFFS_NOBJECT_BUCKETS; ++ if (1 || dev->obj_bucket[bucket].count > 0) { ++ list_for_each(i, &dev->obj_bucket[bucket].list) { ++ /* If there is already one in the list */ ++ if (i && list_entry(i, struct yaffs_obj, ++ hash_link)->obj_id == n) { ++ found = 0; ++ } ++ } ++ } ++ } ++ return n; ++} ++ ++static void yaffs_hash_obj(struct yaffs_obj *in) ++{ ++ int bucket = yaffs_hash_fn(in->obj_id); ++ struct yaffs_dev *dev = in->my_dev; ++ ++ list_add(&in->hash_link, &dev->obj_bucket[bucket].list); ++ dev->obj_bucket[bucket].count++; ++} ++ ++struct yaffs_obj *yaffs_find_by_number(struct yaffs_dev *dev, u32 number) ++{ ++ int bucket = yaffs_hash_fn(number); ++ struct list_head *i; ++ struct yaffs_obj *in; ++ ++ list_for_each(i, &dev->obj_bucket[bucket].list) { ++ /* Look if it is in the list */ ++ in = list_entry(i, struct yaffs_obj, hash_link); ++ if (in->obj_id == number) { ++ /* Don't show if it is defered free */ ++ if (in->defered_free) ++ return NULL; ++ return in; ++ } ++ } ++ ++ return NULL; ++} ++ ++static struct yaffs_obj *yaffs_new_obj(struct yaffs_dev *dev, int number, ++ enum yaffs_obj_type type) ++{ ++ struct yaffs_obj *the_obj = NULL; ++ struct yaffs_tnode *tn = NULL; ++ ++ if (number < 0) ++ number = yaffs_new_obj_id(dev); ++ ++ if (type == YAFFS_OBJECT_TYPE_FILE) { ++ tn = yaffs_get_tnode(dev); ++ if (!tn) ++ return NULL; ++ } ++ ++ the_obj = yaffs_alloc_empty_obj(dev); ++ if (!the_obj) { ++ if (tn) ++ yaffs_free_tnode(dev, tn); ++ return NULL; ++ } ++ ++ the_obj->fake = 0; ++ the_obj->rename_allowed = 1; ++ the_obj->unlink_allowed = 1; ++ the_obj->obj_id = number; ++ yaffs_hash_obj(the_obj); ++ the_obj->variant_type = type; ++ yaffs_load_current_time(the_obj, 1, 1); ++ ++ switch (type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ the_obj->variant.file_variant.file_size = 0; ++ the_obj->variant.file_variant.scanned_size = 0; ++ the_obj->variant.file_variant.shrink_size = ++ yaffs_max_file_size(dev); ++ the_obj->variant.file_variant.top_level = 0; ++ the_obj->variant.file_variant.top = tn; ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ INIT_LIST_HEAD(&the_obj->variant.dir_variant.children); ++ INIT_LIST_HEAD(&the_obj->variant.dir_variant.dirty); ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ /* No action required */ ++ break; ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ /* todo this should not happen */ ++ break; ++ } ++ return the_obj; ++} ++ ++static struct yaffs_obj *yaffs_create_fake_dir(struct yaffs_dev *dev, ++ int number, u32 mode) ++{ ++ ++ struct yaffs_obj *obj = ++ yaffs_new_obj(dev, number, YAFFS_OBJECT_TYPE_DIRECTORY); ++ ++ if (!obj) ++ return NULL; ++ ++ obj->fake = 1; /* it is fake so it might not use NAND */ ++ obj->rename_allowed = 0; ++ obj->unlink_allowed = 0; ++ obj->deleted = 0; ++ obj->unlinked = 0; ++ obj->yst_mode = mode; ++ obj->my_dev = dev; ++ obj->hdr_chunk = 0; /* Not a valid chunk. */ ++ return obj; ++ ++} ++ ++ ++static void yaffs_init_tnodes_and_objs(struct yaffs_dev *dev) ++{ ++ int i; ++ ++ dev->n_obj = 0; ++ dev->n_tnodes = 0; ++ yaffs_init_raw_tnodes_and_objs(dev); ++ ++ for (i = 0; i < YAFFS_NOBJECT_BUCKETS; i++) { ++ INIT_LIST_HEAD(&dev->obj_bucket[i].list); ++ dev->obj_bucket[i].count = 0; ++ } ++} ++ ++struct yaffs_obj *yaffs_find_or_create_by_number(struct yaffs_dev *dev, ++ int number, ++ enum yaffs_obj_type type) ++{ ++ struct yaffs_obj *the_obj = NULL; ++ ++ if (number > 0) ++ the_obj = yaffs_find_by_number(dev, number); ++ ++ if (!the_obj) ++ the_obj = yaffs_new_obj(dev, number, type); ++ ++ return the_obj; ++ ++} ++ ++YCHAR *yaffs_clone_str(const YCHAR *str) ++{ ++ YCHAR *new_str = NULL; ++ int len; ++ ++ if (!str) ++ str = _Y(""); ++ ++ len = strnlen(str, YAFFS_MAX_ALIAS_LENGTH); ++ new_str = kmalloc((len + 1) * sizeof(YCHAR), GFP_NOFS); ++ if (new_str) { ++ strncpy(new_str, str, len); ++ new_str[len] = 0; ++ } ++ return new_str; ++ ++} ++/* ++ *yaffs_update_parent() handles fixing a directories mtime and ctime when a new ++ * link (ie. name) is created or deleted in the directory. ++ * ++ * ie. ++ * create dir/a : update dir's mtime/ctime ++ * rm dir/a: update dir's mtime/ctime ++ * modify dir/a: don't update dir's mtimme/ctime ++ * ++ * This can be handled immediately or defered. Defering helps reduce the number ++ * of updates when many files in a directory are changed within a brief period. ++ * ++ * If the directory updating is defered then yaffs_update_dirty_dirs must be ++ * called periodically. ++ */ ++ ++static void yaffs_update_parent(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev; ++ ++ if (!obj) ++ return; ++ dev = obj->my_dev; ++ obj->dirty = 1; ++ yaffs_load_current_time(obj, 0, 1); ++ if (dev->param.defered_dir_update) { ++ struct list_head *link = &obj->variant.dir_variant.dirty; ++ ++ if (list_empty(link)) { ++ list_add(link, &dev->dirty_dirs); ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, ++ "Added object %d to dirty directories", ++ obj->obj_id); ++ } ++ ++ } else { ++ yaffs_update_oh(obj, NULL, 0, 0, 0, NULL); ++ } ++} ++ ++void yaffs_update_dirty_dirs(struct yaffs_dev *dev) ++{ ++ struct list_head *link; ++ struct yaffs_obj *obj; ++ struct yaffs_dir_var *d_s; ++ union yaffs_obj_var *o_v; ++ ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, "Update dirty directories"); ++ ++ while (!list_empty(&dev->dirty_dirs)) { ++ link = dev->dirty_dirs.next; ++ list_del_init(link); ++ ++ d_s = list_entry(link, struct yaffs_dir_var, dirty); ++ o_v = list_entry(d_s, union yaffs_obj_var, dir_variant); ++ obj = list_entry(o_v, struct yaffs_obj, variant); ++ ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, "Update directory %d", ++ obj->obj_id); ++ ++ if (obj->dirty) ++ yaffs_update_oh(obj, NULL, 0, 0, 0, NULL); ++ } ++} ++ ++/* ++ * Mknod (create) a new object. ++ * equiv_obj only has meaning for a hard link; ++ * alias_str only has meaning for a symlink. ++ * rdev only has meaning for devices (a subset of special objects) ++ */ ++ ++static struct yaffs_obj *yaffs_create_obj(enum yaffs_obj_type type, ++ struct yaffs_obj *parent, ++ const YCHAR *name, ++ u32 mode, ++ u32 uid, ++ u32 gid, ++ struct yaffs_obj *equiv_obj, ++ const YCHAR *alias_str, u32 rdev) ++{ ++ struct yaffs_obj *in; ++ YCHAR *str = NULL; ++ struct yaffs_dev *dev = parent->my_dev; ++ ++ /* Check if the entry exists. ++ * If it does then fail the call since we don't want a dup. */ ++ if (yaffs_find_by_name(parent, name)) ++ return NULL; ++ ++ if (type == YAFFS_OBJECT_TYPE_SYMLINK) { ++ str = yaffs_clone_str(alias_str); ++ if (!str) ++ return NULL; ++ } ++ ++ in = yaffs_new_obj(dev, -1, type); ++ ++ if (!in) { ++ kfree(str); ++ return NULL; ++ } ++ ++ in->hdr_chunk = 0; ++ in->valid = 1; ++ in->variant_type = type; ++ ++ in->yst_mode = mode; ++ ++ yaffs_attribs_init(in, gid, uid, rdev); ++ ++ in->n_data_chunks = 0; ++ ++ yaffs_set_obj_name(in, name); ++ in->dirty = 1; ++ ++ yaffs_add_obj_to_dir(parent, in); ++ ++ in->my_dev = parent->my_dev; ++ ++ switch (type) { ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ in->variant.symlink_variant.alias = str; ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ in->variant.hardlink_variant.equiv_obj = equiv_obj; ++ in->variant.hardlink_variant.equiv_id = equiv_obj->obj_id; ++ list_add(&in->hard_links, &equiv_obj->hard_links); ++ break; ++ case YAFFS_OBJECT_TYPE_FILE: ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ /* do nothing */ ++ break; ++ } ++ ++ if (yaffs_update_oh(in, name, 0, 0, 0, NULL) < 0) { ++ /* Could not create the object header, fail */ ++ yaffs_del_obj(in); ++ in = NULL; ++ } ++ ++ if (in) ++ yaffs_update_parent(parent); ++ ++ return in; ++} ++ ++struct yaffs_obj *yaffs_create_file(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid) ++{ ++ return yaffs_create_obj(YAFFS_OBJECT_TYPE_FILE, parent, name, mode, ++ uid, gid, NULL, NULL, 0); ++} ++ ++struct yaffs_obj *yaffs_create_dir(struct yaffs_obj *parent, const YCHAR *name, ++ u32 mode, u32 uid, u32 gid) ++{ ++ return yaffs_create_obj(YAFFS_OBJECT_TYPE_DIRECTORY, parent, name, ++ mode, uid, gid, NULL, NULL, 0); ++} ++ ++struct yaffs_obj *yaffs_create_special(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid, u32 rdev) ++{ ++ return yaffs_create_obj(YAFFS_OBJECT_TYPE_SPECIAL, parent, name, mode, ++ uid, gid, NULL, NULL, rdev); ++} ++ ++struct yaffs_obj *yaffs_create_symlink(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid, const YCHAR *alias) ++{ ++ return yaffs_create_obj(YAFFS_OBJECT_TYPE_SYMLINK, parent, name, mode, ++ uid, gid, NULL, alias, 0); ++} ++ ++/* yaffs_link_obj returns the object id of the equivalent object.*/ ++struct yaffs_obj *yaffs_link_obj(struct yaffs_obj *parent, const YCHAR * name, ++ struct yaffs_obj *equiv_obj) ++{ ++ /* Get the real object in case we were fed a hard link obj */ ++ equiv_obj = yaffs_get_equivalent_obj(equiv_obj); ++ ++ if (yaffs_create_obj(YAFFS_OBJECT_TYPE_HARDLINK, ++ parent, name, 0, 0, 0, ++ equiv_obj, NULL, 0)) ++ return equiv_obj; ++ ++ return NULL; ++ ++} ++ ++ ++ ++/*---------------------- Block Management and Page Allocation -------------*/ ++ ++static void yaffs_deinit_blocks(struct yaffs_dev *dev) ++{ ++ if (dev->block_info_alt && dev->block_info) ++ vfree(dev->block_info); ++ else ++ kfree(dev->block_info); ++ ++ dev->block_info_alt = 0; ++ ++ dev->block_info = NULL; ++ ++ if (dev->chunk_bits_alt && dev->chunk_bits) ++ vfree(dev->chunk_bits); ++ else ++ kfree(dev->chunk_bits); ++ dev->chunk_bits_alt = 0; ++ dev->chunk_bits = NULL; ++} ++ ++static int yaffs_init_blocks(struct yaffs_dev *dev) ++{ ++ int n_blocks = dev->internal_end_block - dev->internal_start_block + 1; ++ ++ dev->block_info = NULL; ++ dev->chunk_bits = NULL; ++ dev->alloc_block = -1; /* force it to get a new one */ ++ ++ /* If the first allocation strategy fails, thry the alternate one */ ++ dev->block_info = ++ kmalloc(n_blocks * sizeof(struct yaffs_block_info), GFP_NOFS); ++ if (!dev->block_info) { ++ dev->block_info = ++ vmalloc(n_blocks * sizeof(struct yaffs_block_info)); ++ dev->block_info_alt = 1; ++ } else { ++ dev->block_info_alt = 0; ++ } ++ ++ if (!dev->block_info) ++ goto alloc_error; ++ ++ /* Set up dynamic blockinfo stuff. Round up bytes. */ ++ dev->chunk_bit_stride = (dev->param.chunks_per_block + 7) / 8; ++ dev->chunk_bits = ++ kmalloc(dev->chunk_bit_stride * n_blocks, GFP_NOFS); ++ if (!dev->chunk_bits) { ++ dev->chunk_bits = ++ vmalloc(dev->chunk_bit_stride * n_blocks); ++ dev->chunk_bits_alt = 1; ++ } else { ++ dev->chunk_bits_alt = 0; ++ } ++ if (!dev->chunk_bits) ++ goto alloc_error; ++ ++ ++ memset(dev->block_info, 0, n_blocks * sizeof(struct yaffs_block_info)); ++ memset(dev->chunk_bits, 0, dev->chunk_bit_stride * n_blocks); ++ return YAFFS_OK; ++ ++alloc_error: ++ yaffs_deinit_blocks(dev); ++ return YAFFS_FAIL; ++} ++ ++ ++void yaffs_block_became_dirty(struct yaffs_dev *dev, int block_no) ++{ ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, block_no); ++ int erased_ok = 0; ++ int i; ++ ++ /* If the block is still healthy erase it and mark as clean. ++ * If the block has had a data failure, then retire it. ++ */ ++ ++ yaffs_trace(YAFFS_TRACE_GC | YAFFS_TRACE_ERASE, ++ "yaffs_block_became_dirty block %d state %d %s", ++ block_no, bi->block_state, ++ (bi->needs_retiring) ? "needs retiring" : ""); ++ ++ yaffs2_clear_oldest_dirty_seq(dev, bi); ++ ++ bi->block_state = YAFFS_BLOCK_STATE_DIRTY; ++ ++ /* If this is the block being garbage collected then stop gc'ing */ ++ if (block_no == dev->gc_block) ++ dev->gc_block = 0; ++ ++ /* If this block is currently the best candidate for gc ++ * then drop as a candidate */ ++ if (block_no == dev->gc_dirtiest) { ++ dev->gc_dirtiest = 0; ++ dev->gc_pages_in_use = 0; ++ } ++ ++ if (!bi->needs_retiring) { ++ yaffs2_checkpt_invalidate(dev); ++ erased_ok = yaffs_erase_block(dev, block_no); ++ if (!erased_ok) { ++ dev->n_erase_failures++; ++ yaffs_trace(YAFFS_TRACE_ERROR | YAFFS_TRACE_BAD_BLOCKS, ++ "**>> Erasure failed %d", block_no); ++ } ++ } ++ ++ /* Verify erasure if needed */ ++ if (erased_ok && ++ ((yaffs_trace_mask & YAFFS_TRACE_ERASE) || ++ !yaffs_skip_verification(dev))) { ++ for (i = 0; i < dev->param.chunks_per_block; i++) { ++ if (!yaffs_check_chunk_erased(dev, ++ block_no * dev->param.chunks_per_block + i)) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ ">>Block %d erasure supposedly OK, but chunk %d not erased", ++ block_no, i); ++ } ++ } ++ } ++ ++ if (!erased_ok) { ++ /* We lost a block of free space */ ++ dev->n_free_chunks -= dev->param.chunks_per_block; ++ yaffs_retire_block(dev, block_no); ++ yaffs_trace(YAFFS_TRACE_ERROR | YAFFS_TRACE_BAD_BLOCKS, ++ "**>> Block %d retired", block_no); ++ return; ++ } ++ ++ /* Clean it up... */ ++ bi->block_state = YAFFS_BLOCK_STATE_EMPTY; ++ bi->seq_number = 0; ++ dev->n_erased_blocks++; ++ bi->pages_in_use = 0; ++ bi->soft_del_pages = 0; ++ bi->has_shrink_hdr = 0; ++ bi->skip_erased_check = 1; /* Clean, so no need to check */ ++ bi->gc_prioritise = 0; ++ bi->has_summary = 0; ++ ++ yaffs_clear_chunk_bits(dev, block_no); ++ ++ yaffs_trace(YAFFS_TRACE_ERASE, "Erased block %d", block_no); ++} ++ ++static inline int yaffs_gc_process_chunk(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi, ++ int old_chunk, u8 *buffer) ++{ ++ int new_chunk; ++ int mark_flash = 1; ++ struct yaffs_ext_tags tags; ++ struct yaffs_obj *object; ++ int matching_chunk; ++ int ret_val = YAFFS_OK; ++ ++ memset(&tags, 0, sizeof(tags)); ++ yaffs_rd_chunk_tags_nand(dev, old_chunk, ++ buffer, &tags); ++ object = yaffs_find_by_number(dev, tags.obj_id); ++ ++ yaffs_trace(YAFFS_TRACE_GC_DETAIL, ++ "Collecting chunk in block %d, %d %d %d ", ++ dev->gc_chunk, tags.obj_id, ++ tags.chunk_id, tags.n_bytes); ++ ++ if (object && !yaffs_skip_verification(dev)) { ++ if (tags.chunk_id == 0) ++ matching_chunk = ++ object->hdr_chunk; ++ else if (object->soft_del) ++ /* Defeat the test */ ++ matching_chunk = old_chunk; ++ else ++ matching_chunk = ++ yaffs_find_chunk_in_file ++ (object, tags.chunk_id, ++ NULL); ++ ++ if (old_chunk != matching_chunk) ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "gc: page in gc mismatch: %d %d %d %d", ++ old_chunk, ++ matching_chunk, ++ tags.obj_id, ++ tags.chunk_id); ++ } ++ ++ if (!object) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "page %d in gc has no object: %d %d %d ", ++ old_chunk, ++ tags.obj_id, tags.chunk_id, ++ tags.n_bytes); ++ } ++ ++ if (object && ++ object->deleted && ++ object->soft_del && tags.chunk_id != 0) { ++ /* Data chunk in a soft deleted file, ++ * throw it away. ++ * It's a soft deleted data chunk, ++ * No need to copy this, just forget ++ * about it and fix up the object. ++ */ ++ ++ /* Free chunks already includes ++ * softdeleted chunks, how ever this ++ * chunk is going to soon be really ++ * deleted which will increment free ++ * chunks. We have to decrement free ++ * chunks so this works out properly. ++ */ ++ dev->n_free_chunks--; ++ bi->soft_del_pages--; ++ ++ object->n_data_chunks--; ++ if (object->n_data_chunks <= 0) { ++ /* remeber to clean up obj */ ++ dev->gc_cleanup_list[dev->n_clean_ups] = tags.obj_id; ++ dev->n_clean_ups++; ++ } ++ mark_flash = 0; ++ } else if (object) { ++ /* It's either a data chunk in a live ++ * file or an ObjectHeader, so we're ++ * interested in it. ++ * NB Need to keep the ObjectHeaders of ++ * deleted files until the whole file ++ * has been deleted off ++ */ ++ tags.serial_number++; ++ dev->n_gc_copies++; ++ ++ if (tags.chunk_id == 0) { ++ /* It is an object Id, ++ * We need to nuke the ++ * shrinkheader flags since its ++ * work is done. ++ * Also need to clean up ++ * shadowing. ++ */ ++ struct yaffs_obj_hdr *oh; ++ oh = (struct yaffs_obj_hdr *) buffer; ++ ++ oh->is_shrink = 0; ++ tags.extra_is_shrink = 0; ++ oh->shadows_obj = 0; ++ oh->inband_shadowed_obj_id = 0; ++ tags.extra_shadows = 0; ++ ++ /* Update file size */ ++ if (object->variant_type == YAFFS_OBJECT_TYPE_FILE) { ++ yaffs_oh_size_load(oh, ++ object->variant.file_variant.file_size); ++ tags.extra_file_size = ++ object->variant.file_variant.file_size; ++ } ++ ++ yaffs_verify_oh(object, oh, &tags, 1); ++ new_chunk = ++ yaffs_write_new_chunk(dev, (u8 *) oh, &tags, 1); ++ } else { ++ new_chunk = ++ yaffs_write_new_chunk(dev, buffer, &tags, 1); ++ } ++ ++ if (new_chunk < 0) { ++ ret_val = YAFFS_FAIL; ++ } else { ++ ++ /* Now fix up the Tnodes etc. */ ++ ++ if (tags.chunk_id == 0) { ++ /* It's a header */ ++ object->hdr_chunk = new_chunk; ++ object->serial = tags.serial_number; ++ } else { ++ /* It's a data chunk */ ++ yaffs_put_chunk_in_file(object, tags.chunk_id, ++ new_chunk, 0); ++ } ++ } ++ } ++ if (ret_val == YAFFS_OK) ++ yaffs_chunk_del(dev, old_chunk, mark_flash, __LINE__); ++ return ret_val; ++} ++ ++static int yaffs_gc_block(struct yaffs_dev *dev, int block, int whole_block) ++{ ++ int old_chunk; ++ int ret_val = YAFFS_OK; ++ int i; ++ int is_checkpt_block; ++ int max_copies; ++ int chunks_before = yaffs_get_erased_chunks(dev); ++ int chunks_after; ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, block); ++ ++ is_checkpt_block = (bi->block_state == YAFFS_BLOCK_STATE_CHECKPOINT); ++ ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "Collecting block %d, in use %d, shrink %d, whole_block %d", ++ block, bi->pages_in_use, bi->has_shrink_hdr, ++ whole_block); ++ ++ /*yaffs_verify_free_chunks(dev); */ ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_FULL) ++ bi->block_state = YAFFS_BLOCK_STATE_COLLECTING; ++ ++ bi->has_shrink_hdr = 0; /* clear the flag so that the block can erase */ ++ ++ dev->gc_disable = 1; ++ ++ yaffs_summary_gc(dev, block); ++ ++ if (is_checkpt_block || !yaffs_still_some_chunks(dev, block)) { ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "Collecting block %d that has no chunks in use", ++ block); ++ yaffs_block_became_dirty(dev, block); ++ } else { ++ ++ u8 *buffer = yaffs_get_temp_buffer(dev); ++ ++ yaffs_verify_blk(dev, bi, block); ++ ++ max_copies = (whole_block) ? dev->param.chunks_per_block : 5; ++ old_chunk = block * dev->param.chunks_per_block + dev->gc_chunk; ++ ++ for (/* init already done */ ; ++ ret_val == YAFFS_OK && ++ dev->gc_chunk < dev->param.chunks_per_block && ++ (bi->block_state == YAFFS_BLOCK_STATE_COLLECTING) && ++ max_copies > 0; ++ dev->gc_chunk++, old_chunk++) { ++ if (yaffs_check_chunk_bit(dev, block, dev->gc_chunk)) { ++ /* Page is in use and might need to be copied */ ++ max_copies--; ++ ret_val = yaffs_gc_process_chunk(dev, bi, ++ old_chunk, buffer); ++ } ++ } ++ yaffs_release_temp_buffer(dev, buffer); ++ } ++ ++ yaffs_verify_collected_blk(dev, bi, block); ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_COLLECTING) { ++ /* ++ * The gc did not complete. Set block state back to FULL ++ * because checkpointing does not restore gc. ++ */ ++ bi->block_state = YAFFS_BLOCK_STATE_FULL; ++ } else { ++ /* The gc completed. */ ++ /* Do any required cleanups */ ++ for (i = 0; i < dev->n_clean_ups; i++) { ++ /* Time to delete the file too */ ++ struct yaffs_obj *object = ++ yaffs_find_by_number(dev, dev->gc_cleanup_list[i]); ++ if (object) { ++ yaffs_free_tnode(dev, ++ object->variant.file_variant.top); ++ object->variant.file_variant.top = NULL; ++ yaffs_trace(YAFFS_TRACE_GC, ++ "yaffs: About to finally delete object %d", ++ object->obj_id); ++ yaffs_generic_obj_del(object); ++ object->my_dev->n_deleted_files--; ++ } ++ ++ } ++ chunks_after = yaffs_get_erased_chunks(dev); ++ if (chunks_before >= chunks_after) ++ yaffs_trace(YAFFS_TRACE_GC, ++ "gc did not increase free chunks before %d after %d", ++ chunks_before, chunks_after); ++ dev->gc_block = 0; ++ dev->gc_chunk = 0; ++ dev->n_clean_ups = 0; ++ } ++ ++ dev->gc_disable = 0; ++ ++ return ret_val; ++} ++ ++/* ++ * find_gc_block() selects the dirtiest block (or close enough) ++ * for garbage collection. ++ */ ++ ++static unsigned yaffs_find_gc_block(struct yaffs_dev *dev, ++ int aggressive, int background) ++{ ++ int i; ++ int iterations; ++ unsigned selected = 0; ++ int prioritised = 0; ++ int prioritised_exist = 0; ++ struct yaffs_block_info *bi; ++ int threshold; ++ ++ /* First let's see if we need to grab a prioritised block */ ++ if (dev->has_pending_prioritised_gc && !aggressive) { ++ dev->gc_dirtiest = 0; ++ bi = dev->block_info; ++ for (i = dev->internal_start_block; ++ i <= dev->internal_end_block && !selected; i++) { ++ ++ if (bi->gc_prioritise) { ++ prioritised_exist = 1; ++ if (bi->block_state == YAFFS_BLOCK_STATE_FULL && ++ yaffs_block_ok_for_gc(dev, bi)) { ++ selected = i; ++ prioritised = 1; ++ } ++ } ++ bi++; ++ } ++ ++ /* ++ * If there is a prioritised block and none was selected then ++ * this happened because there is at least one old dirty block ++ * gumming up the works. Let's gc the oldest dirty block. ++ */ ++ ++ if (prioritised_exist && ++ !selected && dev->oldest_dirty_block > 0) ++ selected = dev->oldest_dirty_block; ++ ++ if (!prioritised_exist) /* None found, so we can clear this */ ++ dev->has_pending_prioritised_gc = 0; ++ } ++ ++ /* If we're doing aggressive GC then we are happy to take a less-dirty ++ * block, and search harder. ++ * else (leasurely gc), then we only bother to do this if the ++ * block has only a few pages in use. ++ */ ++ ++ if (!selected) { ++ int pages_used; ++ int n_blocks = ++ dev->internal_end_block - dev->internal_start_block + 1; ++ if (aggressive) { ++ threshold = dev->param.chunks_per_block; ++ iterations = n_blocks; ++ } else { ++ int max_threshold; ++ ++ if (background) ++ max_threshold = dev->param.chunks_per_block / 2; ++ else ++ max_threshold = dev->param.chunks_per_block / 8; ++ ++ if (max_threshold < YAFFS_GC_PASSIVE_THRESHOLD) ++ max_threshold = YAFFS_GC_PASSIVE_THRESHOLD; ++ ++ threshold = background ? (dev->gc_not_done + 2) * 2 : 0; ++ if (threshold < YAFFS_GC_PASSIVE_THRESHOLD) ++ threshold = YAFFS_GC_PASSIVE_THRESHOLD; ++ if (threshold > max_threshold) ++ threshold = max_threshold; ++ ++ iterations = n_blocks / 16 + 1; ++ if (iterations > 100) ++ iterations = 100; ++ } ++ ++ for (i = 0; ++ i < iterations && ++ (dev->gc_dirtiest < 1 || ++ dev->gc_pages_in_use > YAFFS_GC_GOOD_ENOUGH); ++ i++) { ++ dev->gc_block_finder++; ++ if (dev->gc_block_finder < dev->internal_start_block || ++ dev->gc_block_finder > dev->internal_end_block) ++ dev->gc_block_finder = ++ dev->internal_start_block; ++ ++ bi = yaffs_get_block_info(dev, dev->gc_block_finder); ++ ++ pages_used = bi->pages_in_use - bi->soft_del_pages; ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_FULL && ++ pages_used < dev->param.chunks_per_block && ++ (dev->gc_dirtiest < 1 || ++ pages_used < dev->gc_pages_in_use) && ++ yaffs_block_ok_for_gc(dev, bi)) { ++ dev->gc_dirtiest = dev->gc_block_finder; ++ dev->gc_pages_in_use = pages_used; ++ } ++ } ++ ++ if (dev->gc_dirtiest > 0 && dev->gc_pages_in_use <= threshold) ++ selected = dev->gc_dirtiest; ++ } ++ ++ /* ++ * If nothing has been selected for a while, try the oldest dirty ++ * because that's gumming up the works. ++ */ ++ ++ if (!selected && dev->param.is_yaffs2 && ++ dev->gc_not_done >= (background ? 10 : 20)) { ++ yaffs2_find_oldest_dirty_seq(dev); ++ if (dev->oldest_dirty_block > 0) { ++ selected = dev->oldest_dirty_block; ++ dev->gc_dirtiest = selected; ++ dev->oldest_dirty_gc_count++; ++ bi = yaffs_get_block_info(dev, selected); ++ dev->gc_pages_in_use = ++ bi->pages_in_use - bi->soft_del_pages; ++ } else { ++ dev->gc_not_done = 0; ++ } ++ } ++ ++ if (selected) { ++ yaffs_trace(YAFFS_TRACE_GC, ++ "GC Selected block %d with %d free, prioritised:%d", ++ selected, ++ dev->param.chunks_per_block - dev->gc_pages_in_use, ++ prioritised); ++ ++ dev->n_gc_blocks++; ++ if (background) ++ dev->bg_gcs++; ++ ++ dev->gc_dirtiest = 0; ++ dev->gc_pages_in_use = 0; ++ dev->gc_not_done = 0; ++ if (dev->refresh_skip > 0) ++ dev->refresh_skip--; ++ } else { ++ dev->gc_not_done++; ++ yaffs_trace(YAFFS_TRACE_GC, ++ "GC none: finder %d skip %d threshold %d dirtiest %d using %d oldest %d%s", ++ dev->gc_block_finder, dev->gc_not_done, threshold, ++ dev->gc_dirtiest, dev->gc_pages_in_use, ++ dev->oldest_dirty_block, background ? " bg" : ""); ++ } ++ ++ return selected; ++} ++ ++/* New garbage collector ++ * If we're very low on erased blocks then we do aggressive garbage collection ++ * otherwise we do "leasurely" garbage collection. ++ * Aggressive gc looks further (whole array) and will accept less dirty blocks. ++ * Passive gc only inspects smaller areas and only accepts more dirty blocks. ++ * ++ * The idea is to help clear out space in a more spread-out manner. ++ * Dunno if it really does anything useful. ++ */ ++static int yaffs_check_gc(struct yaffs_dev *dev, int background) ++{ ++ int aggressive = 0; ++ int gc_ok = YAFFS_OK; ++ int max_tries = 0; ++ int min_erased; ++ int erased_chunks; ++ int checkpt_block_adjust; ++ ++ if (dev->param.gc_control_fn && ++ (dev->param.gc_control_fn(dev) & 1) == 0) ++ return YAFFS_OK; ++ ++ if (dev->gc_disable) ++ /* Bail out so we don't get recursive gc */ ++ return YAFFS_OK; ++ ++ /* This loop should pass the first time. ++ * Only loops here if the collection does not increase space. ++ */ ++ ++ do { ++ max_tries++; ++ ++ checkpt_block_adjust = yaffs_calc_checkpt_blocks_required(dev); ++ ++ min_erased = ++ dev->param.n_reserved_blocks + checkpt_block_adjust + 1; ++ erased_chunks = ++ dev->n_erased_blocks * dev->param.chunks_per_block; ++ ++ /* If we need a block soon then do aggressive gc. */ ++ if (dev->n_erased_blocks < min_erased) ++ aggressive = 1; ++ else { ++ if (!background ++ && erased_chunks > (dev->n_free_chunks / 4)) ++ break; ++ ++ if (dev->gc_skip > 20) ++ dev->gc_skip = 20; ++ if (erased_chunks < dev->n_free_chunks / 2 || ++ dev->gc_skip < 1 || background) ++ aggressive = 0; ++ else { ++ dev->gc_skip--; ++ break; ++ } ++ } ++ ++ dev->gc_skip = 5; ++ ++ /* If we don't already have a block being gc'd then see if we ++ * should start another */ ++ ++ if (dev->gc_block < 1 && !aggressive) { ++ dev->gc_block = yaffs2_find_refresh_block(dev); ++ dev->gc_chunk = 0; ++ dev->n_clean_ups = 0; ++ } ++ if (dev->gc_block < 1) { ++ dev->gc_block = ++ yaffs_find_gc_block(dev, aggressive, background); ++ dev->gc_chunk = 0; ++ dev->n_clean_ups = 0; ++ } ++ ++ if (dev->gc_block > 0) { ++ dev->all_gcs++; ++ if (!aggressive) ++ dev->passive_gc_count++; ++ ++ yaffs_trace(YAFFS_TRACE_GC, ++ "yaffs: GC n_erased_blocks %d aggressive %d", ++ dev->n_erased_blocks, aggressive); ++ ++ gc_ok = yaffs_gc_block(dev, dev->gc_block, aggressive); ++ } ++ ++ if (dev->n_erased_blocks < (dev->param.n_reserved_blocks) && ++ dev->gc_block > 0) { ++ yaffs_trace(YAFFS_TRACE_GC, ++ "yaffs: GC !!!no reclaim!!! n_erased_blocks %d after try %d block %d", ++ dev->n_erased_blocks, max_tries, ++ dev->gc_block); ++ } ++ } while ((dev->n_erased_blocks < dev->param.n_reserved_blocks) && ++ (dev->gc_block > 0) && (max_tries < 2)); ++ ++ return aggressive ? gc_ok : YAFFS_OK; ++} ++ ++/* ++ * yaffs_bg_gc() ++ * Garbage collects. Intended to be called from a background thread. ++ * Returns non-zero if at least half the free chunks are erased. ++ */ ++int yaffs_bg_gc(struct yaffs_dev *dev, unsigned urgency) ++{ ++ int erased_chunks = dev->n_erased_blocks * dev->param.chunks_per_block; ++ ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, "Background gc %u", urgency); ++ ++ yaffs_check_gc(dev, 1); ++ return erased_chunks > dev->n_free_chunks / 2; ++} ++ ++/*-------------------- Data file manipulation -----------------*/ ++ ++static int yaffs_rd_data_obj(struct yaffs_obj *in, int inode_chunk, u8 * buffer) ++{ ++ int nand_chunk = yaffs_find_chunk_in_file(in, inode_chunk, NULL); ++ ++ if (nand_chunk >= 0) ++ return yaffs_rd_chunk_tags_nand(in->my_dev, nand_chunk, ++ buffer, NULL); ++ else { ++ yaffs_trace(YAFFS_TRACE_NANDACCESS, ++ "Chunk %d not found zero instead", ++ nand_chunk); ++ /* get sane (zero) data if you read a hole */ ++ memset(buffer, 0, in->my_dev->data_bytes_per_chunk); ++ return 0; ++ } ++ ++} ++ ++void yaffs_chunk_del(struct yaffs_dev *dev, int chunk_id, int mark_flash, ++ int lyn) ++{ ++ int block; ++ int page; ++ struct yaffs_ext_tags tags; ++ struct yaffs_block_info *bi; ++ ++ if (chunk_id <= 0) ++ return; ++ ++ dev->n_deletions++; ++ block = chunk_id / dev->param.chunks_per_block; ++ page = chunk_id % dev->param.chunks_per_block; ++ ++ if (!yaffs_check_chunk_bit(dev, block, page)) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Deleting invalid chunk %d", chunk_id); ++ ++ bi = yaffs_get_block_info(dev, block); ++ ++ yaffs2_update_oldest_dirty_seq(dev, block, bi); ++ ++ yaffs_trace(YAFFS_TRACE_DELETION, ++ "line %d delete of chunk %d", ++ lyn, chunk_id); ++ ++ if (!dev->param.is_yaffs2 && mark_flash && ++ bi->block_state != YAFFS_BLOCK_STATE_COLLECTING) { ++ ++ memset(&tags, 0, sizeof(tags)); ++ tags.is_deleted = 1; ++ yaffs_wr_chunk_tags_nand(dev, chunk_id, NULL, &tags); ++ yaffs_handle_chunk_update(dev, chunk_id, &tags); ++ } else { ++ dev->n_unmarked_deletions++; ++ } ++ ++ /* Pull out of the management area. ++ * If the whole block became dirty, this will kick off an erasure. ++ */ ++ if (bi->block_state == YAFFS_BLOCK_STATE_ALLOCATING || ++ bi->block_state == YAFFS_BLOCK_STATE_FULL || ++ bi->block_state == YAFFS_BLOCK_STATE_NEEDS_SCAN || ++ bi->block_state == YAFFS_BLOCK_STATE_COLLECTING) { ++ dev->n_free_chunks++; ++ yaffs_clear_chunk_bit(dev, block, page); ++ bi->pages_in_use--; ++ ++ if (bi->pages_in_use == 0 && ++ !bi->has_shrink_hdr && ++ bi->block_state != YAFFS_BLOCK_STATE_ALLOCATING && ++ bi->block_state != YAFFS_BLOCK_STATE_NEEDS_SCAN) { ++ yaffs_block_became_dirty(dev, block); ++ } ++ } ++} ++ ++static int yaffs_wr_data_obj(struct yaffs_obj *in, int inode_chunk, ++ const u8 *buffer, int n_bytes, int use_reserve) ++{ ++ /* Find old chunk Need to do this to get serial number ++ * Write new one and patch into tree. ++ * Invalidate old tags. ++ */ ++ ++ int prev_chunk_id; ++ struct yaffs_ext_tags prev_tags; ++ int new_chunk_id; ++ struct yaffs_ext_tags new_tags; ++ struct yaffs_dev *dev = in->my_dev; ++ ++ yaffs_check_gc(dev, 0); ++ ++ /* Get the previous chunk at this location in the file if it exists. ++ * If it does not exist then put a zero into the tree. This creates ++ * the tnode now, rather than later when it is harder to clean up. ++ */ ++ prev_chunk_id = yaffs_find_chunk_in_file(in, inode_chunk, &prev_tags); ++ if (prev_chunk_id < 1 && ++ !yaffs_put_chunk_in_file(in, inode_chunk, 0, 0)) ++ return 0; ++ ++ /* Set up new tags */ ++ memset(&new_tags, 0, sizeof(new_tags)); ++ ++ new_tags.chunk_id = inode_chunk; ++ new_tags.obj_id = in->obj_id; ++ new_tags.serial_number = ++ (prev_chunk_id > 0) ? prev_tags.serial_number + 1 : 1; ++ new_tags.n_bytes = n_bytes; ++ ++ if (n_bytes < 1 || n_bytes > dev->param.total_bytes_per_chunk) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "Writing %d bytes to chunk!!!!!!!!!", ++ n_bytes); ++ BUG(); ++ } ++ ++ new_chunk_id = ++ yaffs_write_new_chunk(dev, buffer, &new_tags, use_reserve); ++ ++ if (new_chunk_id > 0) { ++ yaffs_put_chunk_in_file(in, inode_chunk, new_chunk_id, 0); ++ ++ if (prev_chunk_id > 0) ++ yaffs_chunk_del(dev, prev_chunk_id, 1, __LINE__); ++ ++ yaffs_verify_file_sane(in); ++ } ++ return new_chunk_id; ++ ++} ++ ++ ++ ++static int yaffs_do_xattrib_mod(struct yaffs_obj *obj, int set, ++ const YCHAR *name, const void *value, int size, ++ int flags) ++{ ++ struct yaffs_xattr_mod xmod; ++ int result; ++ ++ xmod.set = set; ++ xmod.name = name; ++ xmod.data = value; ++ xmod.size = size; ++ xmod.flags = flags; ++ xmod.result = -ENOSPC; ++ ++ result = yaffs_update_oh(obj, NULL, 0, 0, 0, &xmod); ++ ++ if (result > 0) ++ return xmod.result; ++ else ++ return -ENOSPC; ++} ++ ++static int yaffs_apply_xattrib_mod(struct yaffs_obj *obj, char *buffer, ++ struct yaffs_xattr_mod *xmod) ++{ ++ int retval = 0; ++ int x_offs = sizeof(struct yaffs_obj_hdr); ++ struct yaffs_dev *dev = obj->my_dev; ++ int x_size = dev->data_bytes_per_chunk - sizeof(struct yaffs_obj_hdr); ++ char *x_buffer = buffer + x_offs; ++ ++ if (xmod->set) ++ retval = ++ nval_set(x_buffer, x_size, xmod->name, xmod->data, ++ xmod->size, xmod->flags); ++ else ++ retval = nval_del(x_buffer, x_size, xmod->name); ++ ++ obj->has_xattr = nval_hasvalues(x_buffer, x_size); ++ obj->xattr_known = 1; ++ xmod->result = retval; ++ ++ return retval; ++} ++ ++static int yaffs_do_xattrib_fetch(struct yaffs_obj *obj, const YCHAR *name, ++ void *value, int size) ++{ ++ char *buffer = NULL; ++ int result; ++ struct yaffs_ext_tags tags; ++ struct yaffs_dev *dev = obj->my_dev; ++ int x_offs = sizeof(struct yaffs_obj_hdr); ++ int x_size = dev->data_bytes_per_chunk - sizeof(struct yaffs_obj_hdr); ++ char *x_buffer; ++ int retval = 0; ++ ++ if (obj->hdr_chunk < 1) ++ return -ENODATA; ++ ++ /* If we know that the object has no xattribs then don't do all the ++ * reading and parsing. ++ */ ++ if (obj->xattr_known && !obj->has_xattr) { ++ if (name) ++ return -ENODATA; ++ else ++ return 0; ++ } ++ ++ buffer = (char *)yaffs_get_temp_buffer(dev); ++ if (!buffer) ++ return -ENOMEM; ++ ++ result = ++ yaffs_rd_chunk_tags_nand(dev, obj->hdr_chunk, (u8 *) buffer, &tags); ++ ++ if (result != YAFFS_OK) ++ retval = -ENOENT; ++ else { ++ x_buffer = buffer + x_offs; ++ ++ if (!obj->xattr_known) { ++ obj->has_xattr = nval_hasvalues(x_buffer, x_size); ++ obj->xattr_known = 1; ++ } ++ ++ if (name) ++ retval = nval_get(x_buffer, x_size, name, value, size); ++ else ++ retval = nval_list(x_buffer, x_size, value, size); ++ } ++ yaffs_release_temp_buffer(dev, (u8 *) buffer); ++ return retval; ++} ++ ++int yaffs_set_xattrib(struct yaffs_obj *obj, const YCHAR * name, ++ const void *value, int size, int flags) ++{ ++ return yaffs_do_xattrib_mod(obj, 1, name, value, size, flags); ++} ++ ++int yaffs_remove_xattrib(struct yaffs_obj *obj, const YCHAR * name) ++{ ++ return yaffs_do_xattrib_mod(obj, 0, name, NULL, 0, 0); ++} ++ ++int yaffs_get_xattrib(struct yaffs_obj *obj, const YCHAR * name, void *value, ++ int size) ++{ ++ return yaffs_do_xattrib_fetch(obj, name, value, size); ++} ++ ++int yaffs_list_xattrib(struct yaffs_obj *obj, char *buffer, int size) ++{ ++ return yaffs_do_xattrib_fetch(obj, NULL, buffer, size); ++} ++ ++static void yaffs_check_obj_details_loaded(struct yaffs_obj *in) ++{ ++ u8 *buf; ++ struct yaffs_obj_hdr *oh; ++ struct yaffs_dev *dev; ++ struct yaffs_ext_tags tags; ++ int result; ++ int alloc_failed = 0; ++ ++ if (!in || !in->lazy_loaded || in->hdr_chunk < 1) ++ return; ++ ++ dev = in->my_dev; ++ in->lazy_loaded = 0; ++ buf = yaffs_get_temp_buffer(dev); ++ ++ result = yaffs_rd_chunk_tags_nand(dev, in->hdr_chunk, buf, &tags); ++ oh = (struct yaffs_obj_hdr *)buf; ++ ++ in->yst_mode = oh->yst_mode; ++ yaffs_load_attribs(in, oh); ++ yaffs_set_obj_name_from_oh(in, oh); ++ ++ if (in->variant_type == YAFFS_OBJECT_TYPE_SYMLINK) { ++ in->variant.symlink_variant.alias = ++ yaffs_clone_str(oh->alias); ++ if (!in->variant.symlink_variant.alias) ++ alloc_failed = 1; /* Not returned */ ++ } ++ yaffs_release_temp_buffer(dev, buf); ++} ++ ++static void yaffs_load_name_from_oh(struct yaffs_dev *dev, YCHAR *name, ++ const YCHAR *oh_name, int buff_size) ++{ ++#ifdef CONFIG_YAFFS_AUTO_UNICODE ++ if (dev->param.auto_unicode) { ++ if (*oh_name) { ++ /* It is an ASCII name, do an ASCII to ++ * unicode conversion */ ++ const char *ascii_oh_name = (const char *)oh_name; ++ int n = buff_size - 1; ++ while (n > 0 && *ascii_oh_name) { ++ *name = *ascii_oh_name; ++ name++; ++ ascii_oh_name++; ++ n--; ++ } ++ } else { ++ strncpy(name, oh_name + 1, buff_size - 1); ++ } ++ } else { ++#else ++ (void) dev; ++ { ++#endif ++ strncpy(name, oh_name, buff_size - 1); ++ } ++} ++ ++static void yaffs_load_oh_from_name(struct yaffs_dev *dev, YCHAR *oh_name, ++ const YCHAR *name) ++{ ++#ifdef CONFIG_YAFFS_AUTO_UNICODE ++ ++ int is_ascii; ++ YCHAR *w; ++ ++ if (dev->param.auto_unicode) { ++ ++ is_ascii = 1; ++ w = name; ++ ++ /* Figure out if the name will fit in ascii character set */ ++ while (is_ascii && *w) { ++ if ((*w) & 0xff00) ++ is_ascii = 0; ++ w++; ++ } ++ ++ if (is_ascii) { ++ /* It is an ASCII name, so convert unicode to ascii */ ++ char *ascii_oh_name = (char *)oh_name; ++ int n = YAFFS_MAX_NAME_LENGTH - 1; ++ while (n > 0 && *name) { ++ *ascii_oh_name = *name; ++ name++; ++ ascii_oh_name++; ++ n--; ++ } ++ } else { ++ /* Unicode name, so save starting at the second YCHAR */ ++ *oh_name = 0; ++ strncpy(oh_name + 1, name, YAFFS_MAX_NAME_LENGTH - 2); ++ } ++ } else { ++#else ++ dev = dev; ++ { ++#endif ++ strncpy(oh_name, name, YAFFS_MAX_NAME_LENGTH - 1); ++ } ++} ++ ++/* UpdateObjectHeader updates the header on NAND for an object. ++ * If name is not NULL, then that new name is used. ++ */ ++int yaffs_update_oh(struct yaffs_obj *in, const YCHAR *name, int force, ++ int is_shrink, int shadows, struct yaffs_xattr_mod *xmod) ++{ ++ ++ struct yaffs_block_info *bi; ++ struct yaffs_dev *dev = in->my_dev; ++ int prev_chunk_id; ++ int ret_val = 0; ++ int result = 0; ++ int new_chunk_id; ++ struct yaffs_ext_tags new_tags; ++ struct yaffs_ext_tags old_tags; ++ const YCHAR *alias = NULL; ++ u8 *buffer = NULL; ++ YCHAR old_name[YAFFS_MAX_NAME_LENGTH + 1]; ++ struct yaffs_obj_hdr *oh = NULL; ++ loff_t file_size = 0; ++ ++ strcpy(old_name, _Y("silly old name")); ++ ++ if (in->fake && in != dev->root_dir && !force && !xmod) ++ return ret_val; ++ ++ yaffs_check_gc(dev, 0); ++ yaffs_check_obj_details_loaded(in); ++ ++ buffer = yaffs_get_temp_buffer(in->my_dev); ++ oh = (struct yaffs_obj_hdr *)buffer; ++ ++ prev_chunk_id = in->hdr_chunk; ++ ++ if (prev_chunk_id > 0) { ++ result = yaffs_rd_chunk_tags_nand(dev, prev_chunk_id, ++ buffer, &old_tags); ++ ++ yaffs_verify_oh(in, oh, &old_tags, 0); ++ memcpy(old_name, oh->name, sizeof(oh->name)); ++ memset(buffer, 0xff, sizeof(struct yaffs_obj_hdr)); ++ } else { ++ memset(buffer, 0xff, dev->data_bytes_per_chunk); ++ } ++ ++ oh->type = in->variant_type; ++ oh->yst_mode = in->yst_mode; ++ oh->shadows_obj = oh->inband_shadowed_obj_id = shadows; ++ ++ yaffs_load_attribs_oh(oh, in); ++ ++ if (in->parent) ++ oh->parent_obj_id = in->parent->obj_id; ++ else ++ oh->parent_obj_id = 0; ++ ++ if (name && *name) { ++ memset(oh->name, 0, sizeof(oh->name)); ++ yaffs_load_oh_from_name(dev, oh->name, name); ++ } else if (prev_chunk_id > 0) { ++ memcpy(oh->name, old_name, sizeof(oh->name)); ++ } else { ++ memset(oh->name, 0, sizeof(oh->name)); ++ } ++ ++ oh->is_shrink = is_shrink; ++ ++ switch (in->variant_type) { ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ /* Should not happen */ ++ break; ++ case YAFFS_OBJECT_TYPE_FILE: ++ if (oh->parent_obj_id != YAFFS_OBJECTID_DELETED && ++ oh->parent_obj_id != YAFFS_OBJECTID_UNLINKED) ++ file_size = in->variant.file_variant.file_size; ++ yaffs_oh_size_load(oh, file_size); ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ oh->equiv_id = in->variant.hardlink_variant.equiv_id; ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ alias = in->variant.symlink_variant.alias; ++ if (!alias) ++ alias = _Y("no alias"); ++ strncpy(oh->alias, alias, YAFFS_MAX_ALIAS_LENGTH); ++ oh->alias[YAFFS_MAX_ALIAS_LENGTH] = 0; ++ break; ++ } ++ ++ /* process any xattrib modifications */ ++ if (xmod) ++ yaffs_apply_xattrib_mod(in, (char *)buffer, xmod); ++ ++ /* Tags */ ++ memset(&new_tags, 0, sizeof(new_tags)); ++ in->serial++; ++ new_tags.chunk_id = 0; ++ new_tags.obj_id = in->obj_id; ++ new_tags.serial_number = in->serial; ++ ++ /* Add extra info for file header */ ++ new_tags.extra_available = 1; ++ new_tags.extra_parent_id = oh->parent_obj_id; ++ new_tags.extra_file_size = file_size; ++ new_tags.extra_is_shrink = oh->is_shrink; ++ new_tags.extra_equiv_id = oh->equiv_id; ++ new_tags.extra_shadows = (oh->shadows_obj > 0) ? 1 : 0; ++ new_tags.extra_obj_type = in->variant_type; ++ yaffs_verify_oh(in, oh, &new_tags, 1); ++ ++ /* Create new chunk in NAND */ ++ new_chunk_id = ++ yaffs_write_new_chunk(dev, buffer, &new_tags, ++ (prev_chunk_id > 0) ? 1 : 0); ++ ++ if (buffer) ++ yaffs_release_temp_buffer(dev, buffer); ++ ++ if (new_chunk_id < 0) ++ return new_chunk_id; ++ ++ in->hdr_chunk = new_chunk_id; ++ ++ if (prev_chunk_id > 0) ++ yaffs_chunk_del(dev, prev_chunk_id, 1, __LINE__); ++ ++ if (!yaffs_obj_cache_dirty(in)) ++ in->dirty = 0; ++ ++ /* If this was a shrink, then mark the block ++ * that the chunk lives on */ ++ if (is_shrink) { ++ bi = yaffs_get_block_info(in->my_dev, ++ new_chunk_id / ++ in->my_dev->param.chunks_per_block); ++ bi->has_shrink_hdr = 1; ++ } ++ ++ ++ return new_chunk_id; ++} ++ ++/*--------------------- File read/write ------------------------ ++ * Read and write have very similar structures. ++ * In general the read/write has three parts to it ++ * An incomplete chunk to start with (if the read/write is not chunk-aligned) ++ * Some complete chunks ++ * An incomplete chunk to end off with ++ * ++ * Curve-balls: the first chunk might also be the last chunk. ++ */ ++ ++int yaffs_file_rd(struct yaffs_obj *in, u8 * buffer, loff_t offset, int n_bytes) ++{ ++ int chunk; ++ u32 start; ++ int n_copy; ++ int n = n_bytes; ++ int n_done = 0; ++ struct yaffs_cache *cache; ++ struct yaffs_dev *dev; ++ ++ dev = in->my_dev; ++ ++ while (n > 0) { ++ yaffs_addr_to_chunk(dev, offset, &chunk, &start); ++ chunk++; ++ ++ /* OK now check for the curveball where the start and end are in ++ * the same chunk. ++ */ ++ if ((start + n) < dev->data_bytes_per_chunk) ++ n_copy = n; ++ else ++ n_copy = dev->data_bytes_per_chunk - start; ++ ++ cache = yaffs_find_chunk_cache(in, chunk); ++ ++ /* If the chunk is already in the cache or it is less than ++ * a whole chunk or we're using inband tags then use the cache ++ * (if there is caching) else bypass the cache. ++ */ ++ if (cache || n_copy != dev->data_bytes_per_chunk || ++ dev->param.inband_tags) { ++ if (dev->param.n_caches > 0) { ++ ++ /* If we can't find the data in the cache, ++ * then load it up. */ ++ ++ if (!cache) { ++ cache = ++ yaffs_grab_chunk_cache(in->my_dev); ++ cache->object = in; ++ cache->chunk_id = chunk; ++ cache->dirty = 0; ++ cache->locked = 0; ++ yaffs_rd_data_obj(in, chunk, ++ cache->data); ++ cache->n_bytes = 0; ++ } ++ ++ yaffs_use_cache(dev, cache, 0); ++ ++ cache->locked = 1; ++ ++ memcpy(buffer, &cache->data[start], n_copy); ++ ++ cache->locked = 0; ++ } else { ++ /* Read into the local buffer then copy.. */ ++ ++ u8 *local_buffer = ++ yaffs_get_temp_buffer(dev); ++ yaffs_rd_data_obj(in, chunk, local_buffer); ++ ++ memcpy(buffer, &local_buffer[start], n_copy); ++ ++ yaffs_release_temp_buffer(dev, local_buffer); ++ } ++ } else { ++ /* A full chunk. Read directly into the buffer. */ ++ yaffs_rd_data_obj(in, chunk, buffer); ++ } ++ n -= n_copy; ++ offset += n_copy; ++ buffer += n_copy; ++ n_done += n_copy; ++ } ++ return n_done; ++} ++ ++int yaffs_do_file_wr(struct yaffs_obj *in, const u8 *buffer, loff_t offset, ++ int n_bytes, int write_through) ++{ ++ ++ int chunk; ++ u32 start; ++ int n_copy; ++ int n = n_bytes; ++ int n_done = 0; ++ int n_writeback; ++ loff_t start_write = offset; ++ int chunk_written = 0; ++ u32 n_bytes_read; ++ loff_t chunk_start; ++ struct yaffs_dev *dev; ++ ++ dev = in->my_dev; ++ ++ while (n > 0 && chunk_written >= 0) { ++ yaffs_addr_to_chunk(dev, offset, &chunk, &start); ++ ++ if (((loff_t)chunk) * ++ dev->data_bytes_per_chunk + start != offset || ++ start >= dev->data_bytes_per_chunk) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "AddrToChunk of offset %lld gives chunk %d start %d", ++ offset, chunk, start); ++ } ++ chunk++; /* File pos to chunk in file offset */ ++ ++ /* OK now check for the curveball where the start and end are in ++ * the same chunk. ++ */ ++ ++ if ((start + n) < dev->data_bytes_per_chunk) { ++ n_copy = n; ++ ++ /* Now calculate how many bytes to write back.... ++ * If we're overwriting and not writing to then end of ++ * file then we need to write back as much as was there ++ * before. ++ */ ++ ++ chunk_start = (((loff_t)(chunk - 1)) * ++ dev->data_bytes_per_chunk); ++ ++ if (chunk_start > in->variant.file_variant.file_size) ++ n_bytes_read = 0; /* Past end of file */ ++ else ++ n_bytes_read = ++ in->variant.file_variant.file_size - ++ chunk_start; ++ ++ if (n_bytes_read > dev->data_bytes_per_chunk) ++ n_bytes_read = dev->data_bytes_per_chunk; ++ ++ n_writeback = ++ (n_bytes_read > ++ (start + n)) ? n_bytes_read : (start + n); ++ ++ if (n_writeback < 0 || ++ n_writeback > dev->data_bytes_per_chunk) ++ BUG(); ++ ++ } else { ++ n_copy = dev->data_bytes_per_chunk - start; ++ n_writeback = dev->data_bytes_per_chunk; ++ } ++ ++ if (n_copy != dev->data_bytes_per_chunk || ++ !dev->param.cache_bypass_aligned || ++ dev->param.inband_tags) { ++ /* An incomplete start or end chunk (or maybe both ++ * start and end chunk), or we're using inband tags, ++ * or we're forcing writes through the cache, ++ * so we want to use the cache buffers. ++ */ ++ if (dev->param.n_caches > 0) { ++ struct yaffs_cache *cache; ++ ++ /* If we can't find the data in the cache, then ++ * load the cache */ ++ cache = yaffs_find_chunk_cache(in, chunk); ++ ++ if (!cache && ++ yaffs_check_alloc_available(dev, 1)) { ++ cache = yaffs_grab_chunk_cache(dev); ++ cache->object = in; ++ cache->chunk_id = chunk; ++ cache->dirty = 0; ++ cache->locked = 0; ++ yaffs_rd_data_obj(in, chunk, ++ cache->data); ++ } else if (cache && ++ !cache->dirty && ++ !yaffs_check_alloc_available(dev, ++ 1)) { ++ /* Drop the cache if it was a read cache ++ * item and no space check has been made ++ * for it. ++ */ ++ cache = NULL; ++ } ++ ++ if (cache) { ++ yaffs_use_cache(dev, cache, 1); ++ cache->locked = 1; ++ ++ memcpy(&cache->data[start], buffer, ++ n_copy); ++ ++ cache->locked = 0; ++ cache->n_bytes = n_writeback; ++ ++ if (write_through) { ++ chunk_written = ++ yaffs_wr_data_obj ++ (cache->object, ++ cache->chunk_id, ++ cache->data, ++ cache->n_bytes, 1); ++ cache->dirty = 0; ++ } ++ } else { ++ chunk_written = -1; /* fail write */ ++ } ++ } else { ++ /* An incomplete start or end chunk (or maybe ++ * both start and end chunk). Read into the ++ * local buffer then copy over and write back. ++ */ ++ ++ u8 *local_buffer = yaffs_get_temp_buffer(dev); ++ ++ yaffs_rd_data_obj(in, chunk, local_buffer); ++ memcpy(&local_buffer[start], buffer, n_copy); ++ ++ chunk_written = ++ yaffs_wr_data_obj(in, chunk, ++ local_buffer, ++ n_writeback, 0); ++ ++ yaffs_release_temp_buffer(dev, local_buffer); ++ } ++ } else { ++ /* A full chunk. Write directly from the buffer. */ ++ ++ chunk_written = ++ yaffs_wr_data_obj(in, chunk, buffer, ++ dev->data_bytes_per_chunk, 0); ++ ++ /* Since we've overwritten the cached data, ++ * we better invalidate it. */ ++ yaffs_invalidate_chunk_cache(in, chunk); ++ } ++ ++ if (chunk_written >= 0) { ++ n -= n_copy; ++ offset += n_copy; ++ buffer += n_copy; ++ n_done += n_copy; ++ } ++ } ++ ++ /* Update file object */ ++ ++ if ((start_write + n_done) > in->variant.file_variant.file_size) ++ in->variant.file_variant.file_size = (start_write + n_done); ++ ++ in->dirty = 1; ++ return n_done; ++} ++ ++int yaffs_wr_file(struct yaffs_obj *in, const u8 *buffer, loff_t offset, ++ int n_bytes, int write_through) ++{ ++ yaffs2_handle_hole(in, offset); ++ return yaffs_do_file_wr(in, buffer, offset, n_bytes, write_through); ++} ++ ++/* ---------------------- File resizing stuff ------------------ */ ++ ++static void yaffs_prune_chunks(struct yaffs_obj *in, loff_t new_size) ++{ ++ ++ struct yaffs_dev *dev = in->my_dev; ++ loff_t old_size = in->variant.file_variant.file_size; ++ int i; ++ int chunk_id; ++ u32 dummy; ++ int last_del; ++ int start_del; ++ ++ if (old_size > 0) ++ yaffs_addr_to_chunk(dev, old_size - 1, &last_del, &dummy); ++ else ++ last_del = 0; ++ ++ yaffs_addr_to_chunk(dev, new_size + dev->data_bytes_per_chunk - 1, ++ &start_del, &dummy); ++ last_del++; ++ start_del++; ++ ++ /* Delete backwards so that we don't end up with holes if ++ * power is lost part-way through the operation. ++ */ ++ for (i = last_del; i >= start_del; i--) { ++ /* NB this could be optimised somewhat, ++ * eg. could retrieve the tags and write them without ++ * using yaffs_chunk_del ++ */ ++ ++ chunk_id = yaffs_find_del_file_chunk(in, i, NULL); ++ ++ if (chunk_id < 1) ++ continue; ++ ++ if (chunk_id < ++ (dev->internal_start_block * dev->param.chunks_per_block) || ++ chunk_id >= ++ ((dev->internal_end_block + 1) * ++ dev->param.chunks_per_block)) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Found daft chunk_id %d for %d", ++ chunk_id, i); ++ } else { ++ in->n_data_chunks--; ++ yaffs_chunk_del(dev, chunk_id, 1, __LINE__); ++ } ++ } ++} ++ ++void yaffs_resize_file_down(struct yaffs_obj *obj, loff_t new_size) ++{ ++ int new_full; ++ u32 new_partial; ++ struct yaffs_dev *dev = obj->my_dev; ++ ++ yaffs_addr_to_chunk(dev, new_size, &new_full, &new_partial); ++ ++ yaffs_prune_chunks(obj, new_size); ++ ++ if (new_partial != 0) { ++ int last_chunk = 1 + new_full; ++ u8 *local_buffer = yaffs_get_temp_buffer(dev); ++ ++ /* Rewrite the last chunk with its new size and zero pad */ ++ yaffs_rd_data_obj(obj, last_chunk, local_buffer); ++ memset(local_buffer + new_partial, 0, ++ dev->data_bytes_per_chunk - new_partial); ++ ++ yaffs_wr_data_obj(obj, last_chunk, local_buffer, ++ new_partial, 1); ++ ++ yaffs_release_temp_buffer(dev, local_buffer); ++ } ++ ++ obj->variant.file_variant.file_size = new_size; ++ ++ yaffs_prune_tree(dev, &obj->variant.file_variant); ++} ++ ++int yaffs_resize_file(struct yaffs_obj *in, loff_t new_size) ++{ ++ struct yaffs_dev *dev = in->my_dev; ++ loff_t old_size = in->variant.file_variant.file_size; ++ ++ yaffs_flush_file_cache(in); ++ yaffs_invalidate_whole_cache(in); ++ ++ yaffs_check_gc(dev, 0); ++ ++ if (in->variant_type != YAFFS_OBJECT_TYPE_FILE) ++ return YAFFS_FAIL; ++ ++ if (new_size == old_size) ++ return YAFFS_OK; ++ ++ if (new_size > old_size) { ++ yaffs2_handle_hole(in, new_size); ++ in->variant.file_variant.file_size = new_size; ++ } else { ++ /* new_size < old_size */ ++ yaffs_resize_file_down(in, new_size); ++ } ++ ++ /* Write a new object header to reflect the resize. ++ * show we've shrunk the file, if need be ++ * Do this only if the file is not in the deleted directories ++ * and is not shadowed. ++ */ ++ if (in->parent && ++ !in->is_shadowed && ++ in->parent->obj_id != YAFFS_OBJECTID_UNLINKED && ++ in->parent->obj_id != YAFFS_OBJECTID_DELETED) ++ yaffs_update_oh(in, NULL, 0, 0, 0, NULL); ++ ++ return YAFFS_OK; ++} ++ ++int yaffs_flush_file(struct yaffs_obj *in, int update_time, int data_sync) ++{ ++ if (!in->dirty) ++ return YAFFS_OK; ++ ++ yaffs_flush_file_cache(in); ++ ++ if (data_sync) ++ return YAFFS_OK; ++ ++ if (update_time) ++ yaffs_load_current_time(in, 0, 0); ++ ++ return (yaffs_update_oh(in, NULL, 0, 0, 0, NULL) >= 0) ? ++ YAFFS_OK : YAFFS_FAIL; ++} ++ ++ ++/* yaffs_del_file deletes the whole file data ++ * and the inode associated with the file. ++ * It does not delete the links associated with the file. ++ */ ++static int yaffs_unlink_file_if_needed(struct yaffs_obj *in) ++{ ++ int ret_val; ++ int del_now = 0; ++ struct yaffs_dev *dev = in->my_dev; ++ ++ if (!in->my_inode) ++ del_now = 1; ++ ++ if (del_now) { ++ ret_val = ++ yaffs_change_obj_name(in, in->my_dev->del_dir, ++ _Y("deleted"), 0, 0); ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "yaffs: immediate deletion of file %d", ++ in->obj_id); ++ in->deleted = 1; ++ in->my_dev->n_deleted_files++; ++ if (dev->param.disable_soft_del || dev->param.is_yaffs2) ++ yaffs_resize_file(in, 0); ++ yaffs_soft_del_file(in); ++ } else { ++ ret_val = ++ yaffs_change_obj_name(in, in->my_dev->unlinked_dir, ++ _Y("unlinked"), 0, 0); ++ } ++ return ret_val; ++} ++ ++static int yaffs_del_file(struct yaffs_obj *in) ++{ ++ int ret_val = YAFFS_OK; ++ int deleted; /* Need to cache value on stack if in is freed */ ++ struct yaffs_dev *dev = in->my_dev; ++ ++ if (dev->param.disable_soft_del || dev->param.is_yaffs2) ++ yaffs_resize_file(in, 0); ++ ++ if (in->n_data_chunks > 0) { ++ /* Use soft deletion if there is data in the file. ++ * That won't be the case if it has been resized to zero. ++ */ ++ if (!in->unlinked) ++ ret_val = yaffs_unlink_file_if_needed(in); ++ ++ deleted = in->deleted; ++ ++ if (ret_val == YAFFS_OK && in->unlinked && !in->deleted) { ++ in->deleted = 1; ++ deleted = 1; ++ in->my_dev->n_deleted_files++; ++ yaffs_soft_del_file(in); ++ } ++ return deleted ? YAFFS_OK : YAFFS_FAIL; ++ } else { ++ /* The file has no data chunks so we toss it immediately */ ++ yaffs_free_tnode(in->my_dev, in->variant.file_variant.top); ++ in->variant.file_variant.top = NULL; ++ yaffs_generic_obj_del(in); ++ ++ return YAFFS_OK; ++ } ++} ++ ++int yaffs_is_non_empty_dir(struct yaffs_obj *obj) ++{ ++ return (obj && ++ obj->variant_type == YAFFS_OBJECT_TYPE_DIRECTORY) && ++ !(list_empty(&obj->variant.dir_variant.children)); ++} ++ ++static int yaffs_del_dir(struct yaffs_obj *obj) ++{ ++ /* First check that the directory is empty. */ ++ if (yaffs_is_non_empty_dir(obj)) ++ return YAFFS_FAIL; ++ ++ return yaffs_generic_obj_del(obj); ++} ++ ++static int yaffs_del_symlink(struct yaffs_obj *in) ++{ ++ kfree(in->variant.symlink_variant.alias); ++ in->variant.symlink_variant.alias = NULL; ++ ++ return yaffs_generic_obj_del(in); ++} ++ ++static int yaffs_del_link(struct yaffs_obj *in) ++{ ++ /* remove this hardlink from the list associated with the equivalent ++ * object ++ */ ++ list_del_init(&in->hard_links); ++ return yaffs_generic_obj_del(in); ++} ++ ++int yaffs_del_obj(struct yaffs_obj *obj) ++{ ++ int ret_val = -1; ++ ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ ret_val = yaffs_del_file(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ if (!list_empty(&obj->variant.dir_variant.dirty)) { ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, ++ "Remove object %d from dirty directories", ++ obj->obj_id); ++ list_del_init(&obj->variant.dir_variant.dirty); ++ } ++ return yaffs_del_dir(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ ret_val = yaffs_del_symlink(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ ret_val = yaffs_del_link(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ ret_val = yaffs_generic_obj_del(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ ret_val = 0; ++ break; /* should not happen. */ ++ } ++ return ret_val; ++} ++ ++ ++static void yaffs_empty_dir_to_dir(struct yaffs_obj *from_dir, ++ struct yaffs_obj *to_dir) ++{ ++ struct yaffs_obj *obj; ++ struct list_head *lh; ++ struct list_head *n; ++ ++ list_for_each_safe(lh, n, &from_dir->variant.dir_variant.children) { ++ obj = list_entry(lh, struct yaffs_obj, siblings); ++ yaffs_add_obj_to_dir(to_dir, obj); ++ } ++} ++ ++struct yaffs_obj *yaffs_retype_obj(struct yaffs_obj *obj, ++ enum yaffs_obj_type type) ++{ ++ /* Tear down the old variant */ ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ /* Nuke file data */ ++ yaffs_resize_file(obj, 0); ++ yaffs_free_tnode(obj->my_dev, obj->variant.file_variant.top); ++ obj->variant.file_variant.top = NULL; ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ /* Put the children in lost and found. */ ++ yaffs_empty_dir_to_dir(obj, obj->my_dev->lost_n_found); ++ if (!list_empty(&obj->variant.dir_variant.dirty)) ++ list_del_init(&obj->variant.dir_variant.dirty); ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ /* Nuke symplink data */ ++ kfree(obj->variant.symlink_variant.alias); ++ obj->variant.symlink_variant.alias = NULL; ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ list_del_init(&obj->hard_links); ++ break; ++ default: ++ break; ++ } ++ ++ memset(&obj->variant, 0, sizeof(obj->variant)); ++ ++ /*Set up new variant if the memset is not enough. */ ++ switch (type) { ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ INIT_LIST_HEAD(&obj->variant.dir_variant.children); ++ INIT_LIST_HEAD(&obj->variant.dir_variant.dirty); ++ break; ++ case YAFFS_OBJECT_TYPE_FILE: ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ default: ++ break; ++ } ++ ++ obj->variant_type = type; ++ ++ return obj; ++ ++} ++ ++static int yaffs_unlink_worker(struct yaffs_obj *obj) ++{ ++ int del_now = 0; ++ ++ if (!obj) ++ return YAFFS_FAIL; ++ ++ if (!obj->my_inode) ++ del_now = 1; ++ ++ yaffs_update_parent(obj->parent); ++ ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_HARDLINK) { ++ return yaffs_del_link(obj); ++ } else if (!list_empty(&obj->hard_links)) { ++ /* Curve ball: We're unlinking an object that has a hardlink. ++ * ++ * This problem arises because we are not strictly following ++ * The Linux link/inode model. ++ * ++ * We can't really delete the object. ++ * Instead, we do the following: ++ * - Select a hardlink. ++ * - Unhook it from the hard links ++ * - Move it from its parent directory so that the rename works. ++ * - Rename the object to the hardlink's name. ++ * - Delete the hardlink ++ */ ++ ++ struct yaffs_obj *hl; ++ struct yaffs_obj *parent; ++ int ret_val; ++ YCHAR name[YAFFS_MAX_NAME_LENGTH + 1]; ++ ++ hl = list_entry(obj->hard_links.next, struct yaffs_obj, ++ hard_links); ++ ++ yaffs_get_obj_name(hl, name, YAFFS_MAX_NAME_LENGTH + 1); ++ parent = hl->parent; ++ ++ list_del_init(&hl->hard_links); ++ ++ yaffs_add_obj_to_dir(obj->my_dev->unlinked_dir, hl); ++ ++ ret_val = yaffs_change_obj_name(obj, parent, name, 0, 0); ++ ++ if (ret_val == YAFFS_OK) ++ ret_val = yaffs_generic_obj_del(hl); ++ ++ return ret_val; ++ ++ } else if (del_now) { ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ return yaffs_del_file(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ list_del_init(&obj->variant.dir_variant.dirty); ++ return yaffs_del_dir(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ return yaffs_del_symlink(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ return yaffs_generic_obj_del(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ default: ++ return YAFFS_FAIL; ++ } ++ } else if (yaffs_is_non_empty_dir(obj)) { ++ return YAFFS_FAIL; ++ } else { ++ return yaffs_change_obj_name(obj, obj->my_dev->unlinked_dir, ++ _Y("unlinked"), 0, 0); ++ } ++} ++ ++static int yaffs_unlink_obj(struct yaffs_obj *obj) ++{ ++ if (obj && obj->unlink_allowed) ++ return yaffs_unlink_worker(obj); ++ ++ return YAFFS_FAIL; ++} ++ ++int yaffs_unlinker(struct yaffs_obj *dir, const YCHAR *name) ++{ ++ struct yaffs_obj *obj; ++ ++ obj = yaffs_find_by_name(dir, name); ++ return yaffs_unlink_obj(obj); ++} ++ ++/* Note: ++ * If old_name is NULL then we take old_dir as the object to be renamed. ++ */ ++int yaffs_rename_obj(struct yaffs_obj *old_dir, const YCHAR *old_name, ++ struct yaffs_obj *new_dir, const YCHAR *new_name) ++{ ++ struct yaffs_obj *obj = NULL; ++ struct yaffs_obj *existing_target = NULL; ++ int force = 0; ++ int result; ++ struct yaffs_dev *dev; ++ ++ if (!old_dir || old_dir->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ BUG(); ++ return YAFFS_FAIL; ++ } ++ if (!new_dir || new_dir->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ BUG(); ++ return YAFFS_FAIL; ++ } ++ ++ dev = old_dir->my_dev; ++ ++#ifdef CONFIG_YAFFS_CASE_INSENSITIVE ++ /* Special case for case insemsitive systems. ++ * While look-up is case insensitive, the name isn't. ++ * Therefore we might want to change x.txt to X.txt ++ */ ++ if (old_dir == new_dir && ++ old_name && new_name && ++ strcmp(old_name, new_name) == 0) ++ force = 1; ++#endif ++ ++ if (strnlen(new_name, YAFFS_MAX_NAME_LENGTH + 1) > ++ YAFFS_MAX_NAME_LENGTH) ++ /* ENAMETOOLONG */ ++ return YAFFS_FAIL; ++ ++ if (old_name) ++ obj = yaffs_find_by_name(old_dir, old_name); ++ else{ ++ obj = old_dir; ++ old_dir = obj->parent; ++ } ++ ++ if (obj && obj->rename_allowed) { ++ /* Now handle an existing target, if there is one */ ++ existing_target = yaffs_find_by_name(new_dir, new_name); ++ if (yaffs_is_non_empty_dir(existing_target)) { ++ return YAFFS_FAIL; /* ENOTEMPTY */ ++ } else if (existing_target && existing_target != obj) { ++ /* Nuke the target first, using shadowing, ++ * but only if it isn't the same object. ++ * ++ * Note we must disable gc here otherwise it can mess ++ * up the shadowing. ++ * ++ */ ++ dev->gc_disable = 1; ++ yaffs_change_obj_name(obj, new_dir, new_name, force, ++ existing_target->obj_id); ++ existing_target->is_shadowed = 1; ++ yaffs_unlink_obj(existing_target); ++ dev->gc_disable = 0; ++ } ++ ++ result = yaffs_change_obj_name(obj, new_dir, new_name, 1, 0); ++ ++ yaffs_update_parent(old_dir); ++ if (new_dir != old_dir) ++ yaffs_update_parent(new_dir); ++ ++ return result; ++ } ++ return YAFFS_FAIL; ++} ++ ++/*----------------------- Initialisation Scanning ---------------------- */ ++ ++void yaffs_handle_shadowed_obj(struct yaffs_dev *dev, int obj_id, ++ int backward_scanning) ++{ ++ struct yaffs_obj *obj; ++ ++ if (backward_scanning) { ++ /* Handle YAFFS2 case (backward scanning) ++ * If the shadowed object exists then ignore. ++ */ ++ obj = yaffs_find_by_number(dev, obj_id); ++ if (obj) ++ return; ++ } ++ ++ /* Let's create it (if it does not exist) assuming it is a file so that ++ * it can do shrinking etc. ++ * We put it in unlinked dir to be cleaned up after the scanning ++ */ ++ obj = ++ yaffs_find_or_create_by_number(dev, obj_id, YAFFS_OBJECT_TYPE_FILE); ++ if (!obj) ++ return; ++ obj->is_shadowed = 1; ++ yaffs_add_obj_to_dir(dev->unlinked_dir, obj); ++ obj->variant.file_variant.shrink_size = 0; ++ obj->valid = 1; /* So that we don't read any other info. */ ++} ++ ++void yaffs_link_fixup(struct yaffs_dev *dev, struct list_head *hard_list) ++{ ++ struct list_head *lh; ++ struct list_head *save; ++ struct yaffs_obj *hl; ++ struct yaffs_obj *in; ++ ++ list_for_each_safe(lh, save, hard_list) { ++ hl = list_entry(lh, struct yaffs_obj, hard_links); ++ in = yaffs_find_by_number(dev, ++ hl->variant.hardlink_variant.equiv_id); ++ ++ if (in) { ++ /* Add the hardlink pointers */ ++ hl->variant.hardlink_variant.equiv_obj = in; ++ list_add(&hl->hard_links, &in->hard_links); ++ } else { ++ /* Todo Need to report/handle this better. ++ * Got a problem... hardlink to a non-existant object ++ */ ++ hl->variant.hardlink_variant.equiv_obj = NULL; ++ INIT_LIST_HEAD(&hl->hard_links); ++ } ++ } ++} ++ ++static void yaffs_strip_deleted_objs(struct yaffs_dev *dev) ++{ ++ /* ++ * Sort out state of unlinked and deleted objects after scanning. ++ */ ++ struct list_head *i; ++ struct list_head *n; ++ struct yaffs_obj *l; ++ ++ if (dev->read_only) ++ return; ++ ++ /* Soft delete all the unlinked files */ ++ list_for_each_safe(i, n, ++ &dev->unlinked_dir->variant.dir_variant.children) { ++ l = list_entry(i, struct yaffs_obj, siblings); ++ yaffs_del_obj(l); ++ } ++ ++ list_for_each_safe(i, n, &dev->del_dir->variant.dir_variant.children) { ++ l = list_entry(i, struct yaffs_obj, siblings); ++ yaffs_del_obj(l); ++ } ++} ++ ++/* ++ * This code iterates through all the objects making sure that they are rooted. ++ * Any unrooted objects are re-rooted in lost+found. ++ * An object needs to be in one of: ++ * - Directly under deleted, unlinked ++ * - Directly or indirectly under root. ++ * ++ * Note: ++ * This code assumes that we don't ever change the current relationships ++ * between directories: ++ * root_dir->parent == unlinked_dir->parent == del_dir->parent == NULL ++ * lost-n-found->parent == root_dir ++ * ++ * This fixes the problem where directories might have inadvertently been ++ * deleted leaving the object "hanging" without being rooted in the ++ * directory tree. ++ */ ++ ++static int yaffs_has_null_parent(struct yaffs_dev *dev, struct yaffs_obj *obj) ++{ ++ return (obj == dev->del_dir || ++ obj == dev->unlinked_dir || obj == dev->root_dir); ++} ++ ++static void yaffs_fix_hanging_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_obj *parent; ++ int i; ++ struct list_head *lh; ++ struct list_head *n; ++ int depth_limit; ++ int hanging; ++ ++ if (dev->read_only) ++ return; ++ ++ /* Iterate through the objects in each hash entry, ++ * looking at each object. ++ * Make sure it is rooted. ++ */ ++ ++ for (i = 0; i < YAFFS_NOBJECT_BUCKETS; i++) { ++ list_for_each_safe(lh, n, &dev->obj_bucket[i].list) { ++ obj = list_entry(lh, struct yaffs_obj, hash_link); ++ parent = obj->parent; ++ ++ if (yaffs_has_null_parent(dev, obj)) { ++ /* These directories are not hanging */ ++ hanging = 0; ++ } else if (!parent || ++ parent->variant_type != ++ YAFFS_OBJECT_TYPE_DIRECTORY) { ++ hanging = 1; ++ } else if (yaffs_has_null_parent(dev, parent)) { ++ hanging = 0; ++ } else { ++ /* ++ * Need to follow the parent chain to ++ * see if it is hanging. ++ */ ++ hanging = 0; ++ depth_limit = 100; ++ ++ while (parent != dev->root_dir && ++ parent->parent && ++ parent->parent->variant_type == ++ YAFFS_OBJECT_TYPE_DIRECTORY && ++ depth_limit > 0) { ++ parent = parent->parent; ++ depth_limit--; ++ } ++ if (parent != dev->root_dir) ++ hanging = 1; ++ } ++ if (hanging) { ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "Hanging object %d moved to lost and found", ++ obj->obj_id); ++ yaffs_add_obj_to_dir(dev->lost_n_found, obj); ++ } ++ } ++ } ++} ++ ++/* ++ * Delete directory contents for cleaning up lost and found. ++ */ ++static void yaffs_del_dir_contents(struct yaffs_obj *dir) ++{ ++ struct yaffs_obj *obj; ++ struct list_head *lh; ++ struct list_head *n; ++ ++ if (dir->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) ++ BUG(); ++ ++ list_for_each_safe(lh, n, &dir->variant.dir_variant.children) { ++ obj = list_entry(lh, struct yaffs_obj, siblings); ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_DIRECTORY) ++ yaffs_del_dir_contents(obj); ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "Deleting lost_found object %d", ++ obj->obj_id); ++ yaffs_unlink_obj(obj); ++ } ++} ++ ++static void yaffs_empty_l_n_f(struct yaffs_dev *dev) ++{ ++ yaffs_del_dir_contents(dev->lost_n_found); ++} ++ ++ ++struct yaffs_obj *yaffs_find_by_name(struct yaffs_obj *directory, ++ const YCHAR *name) ++{ ++ int sum; ++ struct list_head *i; ++ YCHAR buffer[YAFFS_MAX_NAME_LENGTH + 1]; ++ struct yaffs_obj *l; ++ ++ if (!name) ++ return NULL; ++ ++ if (!directory) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "tragedy: yaffs_find_by_name: null pointer directory" ++ ); ++ BUG(); ++ return NULL; ++ } ++ if (directory->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "tragedy: yaffs_find_by_name: non-directory" ++ ); ++ BUG(); ++ } ++ ++ sum = yaffs_calc_name_sum(name); ++ ++ list_for_each(i, &directory->variant.dir_variant.children) { ++ l = list_entry(i, struct yaffs_obj, siblings); ++ ++ if (l->parent != directory) ++ BUG(); ++ ++ yaffs_check_obj_details_loaded(l); ++ ++ /* Special case for lost-n-found */ ++ if (l->obj_id == YAFFS_OBJECTID_LOSTNFOUND) { ++ if (!strcmp(name, YAFFS_LOSTNFOUND_NAME)) ++ return l; ++ } else if (l->sum == sum || l->hdr_chunk <= 0) { ++ /* LostnFound chunk called Objxxx ++ * Do a real check ++ */ ++ yaffs_get_obj_name(l, buffer, ++ YAFFS_MAX_NAME_LENGTH + 1); ++ if (!strncmp(name, buffer, YAFFS_MAX_NAME_LENGTH)) ++ return l; ++ } ++ } ++ return NULL; ++} ++ ++/* GetEquivalentObject dereferences any hard links to get to the ++ * actual object. ++ */ ++ ++struct yaffs_obj *yaffs_get_equivalent_obj(struct yaffs_obj *obj) ++{ ++ if (obj && obj->variant_type == YAFFS_OBJECT_TYPE_HARDLINK) { ++ obj = obj->variant.hardlink_variant.equiv_obj; ++ yaffs_check_obj_details_loaded(obj); ++ } ++ return obj; ++} ++ ++/* ++ * A note or two on object names. ++ * * If the object name is missing, we then make one up in the form objnnn ++ * ++ * * ASCII names are stored in the object header's name field from byte zero ++ * * Unicode names are historically stored starting from byte zero. ++ * ++ * Then there are automatic Unicode names... ++ * The purpose of these is to save names in a way that can be read as ++ * ASCII or Unicode names as appropriate, thus allowing a Unicode and ASCII ++ * system to share files. ++ * ++ * These automatic unicode are stored slightly differently... ++ * - If the name can fit in the ASCII character space then they are saved as ++ * ascii names as per above. ++ * - If the name needs Unicode then the name is saved in Unicode ++ * starting at oh->name[1]. ++ ++ */ ++static void yaffs_fix_null_name(struct yaffs_obj *obj, YCHAR *name, ++ int buffer_size) ++{ ++ /* Create an object name if we could not find one. */ ++ if (strnlen(name, YAFFS_MAX_NAME_LENGTH) == 0) { ++ YCHAR local_name[20]; ++ YCHAR num_string[20]; ++ YCHAR *x = &num_string[19]; ++ unsigned v = obj->obj_id; ++ num_string[19] = 0; ++ while (v > 0) { ++ x--; ++ *x = '0' + (v % 10); ++ v /= 10; ++ } ++ /* make up a name */ ++ strcpy(local_name, YAFFS_LOSTNFOUND_PREFIX); ++ strcat(local_name, x); ++ strncpy(name, local_name, buffer_size - 1); ++ } ++} ++ ++int yaffs_get_obj_name(struct yaffs_obj *obj, YCHAR *name, int buffer_size) ++{ ++ memset(name, 0, buffer_size * sizeof(YCHAR)); ++ yaffs_check_obj_details_loaded(obj); ++ if (obj->obj_id == YAFFS_OBJECTID_LOSTNFOUND) { ++ strncpy(name, YAFFS_LOSTNFOUND_NAME, buffer_size - 1); ++ } else if (obj->short_name[0]) { ++ strcpy(name, obj->short_name); ++ } else if (obj->hdr_chunk > 0) { ++ int result; ++ u8 *buffer = yaffs_get_temp_buffer(obj->my_dev); ++ ++ struct yaffs_obj_hdr *oh = (struct yaffs_obj_hdr *)buffer; ++ ++ memset(buffer, 0, obj->my_dev->data_bytes_per_chunk); ++ ++ if (obj->hdr_chunk > 0) { ++ result = yaffs_rd_chunk_tags_nand(obj->my_dev, ++ obj->hdr_chunk, ++ buffer, NULL); ++ } ++ yaffs_load_name_from_oh(obj->my_dev, name, oh->name, ++ buffer_size); ++ ++ yaffs_release_temp_buffer(obj->my_dev, buffer); ++ } ++ ++ yaffs_fix_null_name(obj, name, buffer_size); ++ ++ return strnlen(name, YAFFS_MAX_NAME_LENGTH); ++} ++ ++loff_t yaffs_get_obj_length(struct yaffs_obj *obj) ++{ ++ /* Dereference any hard linking */ ++ obj = yaffs_get_equivalent_obj(obj); ++ ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_FILE) ++ return obj->variant.file_variant.file_size; ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_SYMLINK) { ++ if (!obj->variant.symlink_variant.alias) ++ return 0; ++ return strnlen(obj->variant.symlink_variant.alias, ++ YAFFS_MAX_ALIAS_LENGTH); ++ } else { ++ /* Only a directory should drop through to here */ ++ return obj->my_dev->data_bytes_per_chunk; ++ } ++} ++ ++int yaffs_get_obj_link_count(struct yaffs_obj *obj) ++{ ++ int count = 0; ++ struct list_head *i; ++ ++ if (!obj->unlinked) ++ count++; /* the object itself */ ++ ++ list_for_each(i, &obj->hard_links) ++ count++; /* add the hard links; */ ++ ++ return count; ++} ++ ++int yaffs_get_obj_inode(struct yaffs_obj *obj) ++{ ++ obj = yaffs_get_equivalent_obj(obj); ++ ++ return obj->obj_id; ++} ++ ++unsigned yaffs_get_obj_type(struct yaffs_obj *obj) ++{ ++ obj = yaffs_get_equivalent_obj(obj); ++ ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ return DT_REG; ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ return DT_DIR; ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ return DT_LNK; ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ return DT_REG; ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ if (S_ISFIFO(obj->yst_mode)) ++ return DT_FIFO; ++ if (S_ISCHR(obj->yst_mode)) ++ return DT_CHR; ++ if (S_ISBLK(obj->yst_mode)) ++ return DT_BLK; ++ if (S_ISSOCK(obj->yst_mode)) ++ return DT_SOCK; ++ return DT_REG; ++ break; ++ default: ++ return DT_REG; ++ break; ++ } ++} ++ ++YCHAR *yaffs_get_symlink_alias(struct yaffs_obj *obj) ++{ ++ obj = yaffs_get_equivalent_obj(obj); ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_SYMLINK) ++ return yaffs_clone_str(obj->variant.symlink_variant.alias); ++ else ++ return yaffs_clone_str(_Y("")); ++} ++ ++/*--------------------------- Initialisation code -------------------------- */ ++ ++static int yaffs_check_dev_fns(struct yaffs_dev *dev) ++{ ++ struct yaffs_driver *drv = &dev->drv; ++ struct yaffs_tags_handler *tagger = &dev->tagger; ++ ++ /* Common functions, gotta have */ ++ if (!drv->drv_read_chunk_fn || ++ !drv->drv_write_chunk_fn || ++ !drv->drv_erase_fn) ++ return 0; ++ ++ if (dev->param.is_yaffs2 && ++ (!drv->drv_mark_bad_fn || !drv->drv_check_bad_fn)) ++ return 0; ++ ++ /* Install the default tags marshalling functions if needed. */ ++ yaffs_tags_compat_install(dev); ++ yaffs_tags_marshall_install(dev); ++ ++ /* Check we now have the marshalling functions required. */ ++ if (!tagger->write_chunk_tags_fn || ++ !tagger->read_chunk_tags_fn || ++ !tagger->query_block_fn || ++ !tagger->mark_bad_fn) ++ return 0; ++ ++ return 1; ++} ++ ++static int yaffs_create_initial_dir(struct yaffs_dev *dev) ++{ ++ /* Initialise the unlinked, deleted, root and lost+found directories */ ++ dev->lost_n_found = dev->root_dir = NULL; ++ dev->unlinked_dir = dev->del_dir = NULL; ++ dev->unlinked_dir = ++ yaffs_create_fake_dir(dev, YAFFS_OBJECTID_UNLINKED, S_IFDIR); ++ dev->del_dir = ++ yaffs_create_fake_dir(dev, YAFFS_OBJECTID_DELETED, S_IFDIR); ++ dev->root_dir = ++ yaffs_create_fake_dir(dev, YAFFS_OBJECTID_ROOT, ++ YAFFS_ROOT_MODE | S_IFDIR); ++ dev->lost_n_found = ++ yaffs_create_fake_dir(dev, YAFFS_OBJECTID_LOSTNFOUND, ++ YAFFS_LOSTNFOUND_MODE | S_IFDIR); ++ ++ if (dev->lost_n_found && dev->root_dir && dev->unlinked_dir ++ && dev->del_dir) { ++ yaffs_add_obj_to_dir(dev->root_dir, dev->lost_n_found); ++ return YAFFS_OK; ++ } ++ return YAFFS_FAIL; ++} ++ ++/* Low level init. ++ * Typically only used by yaffs_guts_initialise, but also used by the ++ * Low level yaffs driver tests. ++ */ ++ ++int yaffs_guts_ll_init(struct yaffs_dev *dev) ++{ ++ ++ ++ yaffs_trace(YAFFS_TRACE_TRACING, "yaffs: yaffs_ll_init()"); ++ ++ if (!dev) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: Need a device" ++ ); ++ return YAFFS_FAIL; ++ } ++ ++ if (dev->ll_init) ++ return YAFFS_OK; ++ ++ dev->internal_start_block = dev->param.start_block; ++ dev->internal_end_block = dev->param.end_block; ++ dev->block_offset = 0; ++ dev->chunk_offset = 0; ++ dev->n_free_chunks = 0; ++ ++ dev->gc_block = 0; ++ ++ if (dev->param.start_block == 0) { ++ dev->internal_start_block = dev->param.start_block + 1; ++ dev->internal_end_block = dev->param.end_block + 1; ++ dev->block_offset = 1; ++ dev->chunk_offset = dev->param.chunks_per_block; ++ } ++ ++ /* Check geometry parameters. */ ++ ++ if ((!dev->param.inband_tags && dev->param.is_yaffs2 && ++ dev->param.total_bytes_per_chunk < 1024) || ++ (!dev->param.is_yaffs2 && ++ dev->param.total_bytes_per_chunk < 512) || ++ (dev->param.inband_tags && !dev->param.is_yaffs2) || ++ dev->param.chunks_per_block < 2 || ++ dev->param.n_reserved_blocks < 2 || ++ dev->internal_start_block <= 0 || ++ dev->internal_end_block <= 0 || ++ dev->internal_end_block <= ++ (dev->internal_start_block + dev->param.n_reserved_blocks + 2) ++ ) { ++ /* otherwise it is too small */ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "NAND geometry problems: chunk size %d, type is yaffs%s, inband_tags %d ", ++ dev->param.total_bytes_per_chunk, ++ dev->param.is_yaffs2 ? "2" : "", ++ dev->param.inband_tags); ++ return YAFFS_FAIL; ++ } ++ ++ /* Sort out space for inband tags, if required */ ++ if (dev->param.inband_tags) ++ dev->data_bytes_per_chunk = ++ dev->param.total_bytes_per_chunk - ++ sizeof(struct yaffs_packed_tags2_tags_only); ++ else ++ dev->data_bytes_per_chunk = dev->param.total_bytes_per_chunk; ++ ++ /* Got the right mix of functions? */ ++ if (!yaffs_check_dev_fns(dev)) { ++ /* Function missing */ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "device function(s) missing or wrong"); ++ ++ return YAFFS_FAIL; ++ } ++ ++ if (yaffs_init_nand(dev) != YAFFS_OK) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "InitialiseNAND failed"); ++ return YAFFS_FAIL; ++ } ++ ++ return YAFFS_OK; ++} ++ ++ ++int yaffs_guts_format_dev(struct yaffs_dev *dev) ++{ ++ int i; ++ enum yaffs_block_state state; ++ u32 dummy; ++ ++ if(yaffs_guts_ll_init(dev) != YAFFS_OK) ++ return YAFFS_FAIL; ++ ++ if(dev->is_mounted) ++ return YAFFS_FAIL; ++ ++ for (i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ yaffs_query_init_block_state(dev, i, &state, &dummy); ++ if (state != YAFFS_BLOCK_STATE_DEAD) ++ yaffs_erase_block(dev, i); ++ } ++ ++ return YAFFS_OK; ++} ++ ++ ++int yaffs_guts_initialise(struct yaffs_dev *dev) ++{ ++ int init_failed = 0; ++ unsigned x; ++ int bits; ++ ++ if(yaffs_guts_ll_init(dev) != YAFFS_OK) ++ return YAFFS_FAIL; ++ ++ if (dev->is_mounted) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "device already mounted"); ++ return YAFFS_FAIL; ++ } ++ ++ dev->is_mounted = 1; ++ ++ /* OK now calculate a few things for the device */ ++ ++ /* ++ * Calculate all the chunk size manipulation numbers: ++ */ ++ x = dev->data_bytes_per_chunk; ++ /* We always use dev->chunk_shift and dev->chunk_div */ ++ dev->chunk_shift = calc_shifts(x); ++ x >>= dev->chunk_shift; ++ dev->chunk_div = x; ++ /* We only use chunk mask if chunk_div is 1 */ ++ dev->chunk_mask = (1 << dev->chunk_shift) - 1; ++ ++ /* ++ * Calculate chunk_grp_bits. ++ * We need to find the next power of 2 > than internal_end_block ++ */ ++ ++ x = dev->param.chunks_per_block * (dev->internal_end_block + 1); ++ ++ bits = calc_shifts_ceiling(x); ++ ++ /* Set up tnode width if wide tnodes are enabled. */ ++ if (!dev->param.wide_tnodes_disabled) { ++ /* bits must be even so that we end up with 32-bit words */ ++ if (bits & 1) ++ bits++; ++ if (bits < 16) ++ dev->tnode_width = 16; ++ else ++ dev->tnode_width = bits; ++ } else { ++ dev->tnode_width = 16; ++ } ++ ++ dev->tnode_mask = (1 << dev->tnode_width) - 1; ++ ++ /* Level0 Tnodes are 16 bits or wider (if wide tnodes are enabled), ++ * so if the bitwidth of the ++ * chunk range we're using is greater than 16 we need ++ * to figure out chunk shift and chunk_grp_size ++ */ ++ ++ if (bits <= dev->tnode_width) ++ dev->chunk_grp_bits = 0; ++ else ++ dev->chunk_grp_bits = bits - dev->tnode_width; ++ ++ dev->tnode_size = (dev->tnode_width * YAFFS_NTNODES_LEVEL0) / 8; ++ if (dev->tnode_size < sizeof(struct yaffs_tnode)) ++ dev->tnode_size = sizeof(struct yaffs_tnode); ++ ++ dev->chunk_grp_size = 1 << dev->chunk_grp_bits; ++ ++ if (dev->param.chunks_per_block < dev->chunk_grp_size) { ++ /* We have a problem because the soft delete won't work if ++ * the chunk group size > chunks per block. ++ * This can be remedied by using larger "virtual blocks". ++ */ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "chunk group too large"); ++ ++ return YAFFS_FAIL; ++ } ++ ++ /* Finished verifying the device, continue with initialisation */ ++ ++ /* More device initialisation */ ++ dev->all_gcs = 0; ++ dev->passive_gc_count = 0; ++ dev->oldest_dirty_gc_count = 0; ++ dev->bg_gcs = 0; ++ dev->gc_block_finder = 0; ++ dev->buffered_block = -1; ++ dev->doing_buffered_block_rewrite = 0; ++ dev->n_deleted_files = 0; ++ dev->n_bg_deletions = 0; ++ dev->n_unlinked_files = 0; ++ dev->n_ecc_fixed = 0; ++ dev->n_ecc_unfixed = 0; ++ dev->n_tags_ecc_fixed = 0; ++ dev->n_tags_ecc_unfixed = 0; ++ dev->n_erase_failures = 0; ++ dev->n_erased_blocks = 0; ++ dev->gc_disable = 0; ++ dev->has_pending_prioritised_gc = 1; ++ /* Assume the worst for now, will get fixed on first GC */ ++ INIT_LIST_HEAD(&dev->dirty_dirs); ++ dev->oldest_dirty_seq = 0; ++ dev->oldest_dirty_block = 0; ++ ++ /* Initialise temporary buffers and caches. */ ++ if (!yaffs_init_tmp_buffers(dev)) ++ init_failed = 1; ++ ++ dev->cache = NULL; ++ dev->gc_cleanup_list = NULL; ++ ++ if (!init_failed && dev->param.n_caches > 0) { ++ int i; ++ void *buf; ++ int cache_bytes = ++ dev->param.n_caches * sizeof(struct yaffs_cache); ++ ++ if (dev->param.n_caches > YAFFS_MAX_SHORT_OP_CACHES) ++ dev->param.n_caches = YAFFS_MAX_SHORT_OP_CACHES; ++ ++ dev->cache = kmalloc(cache_bytes, GFP_NOFS); ++ ++ buf = (u8 *) dev->cache; ++ ++ if (dev->cache) ++ memset(dev->cache, 0, cache_bytes); ++ ++ for (i = 0; i < dev->param.n_caches && buf; i++) { ++ dev->cache[i].object = NULL; ++ dev->cache[i].last_use = 0; ++ dev->cache[i].dirty = 0; ++ dev->cache[i].data = buf = ++ kmalloc(dev->param.total_bytes_per_chunk, GFP_NOFS); ++ } ++ if (!buf) ++ init_failed = 1; ++ ++ dev->cache_last_use = 0; ++ } ++ ++ dev->cache_hits = 0; ++ ++ if (!init_failed) { ++ dev->gc_cleanup_list = ++ kmalloc(dev->param.chunks_per_block * sizeof(u32), ++ GFP_NOFS); ++ if (!dev->gc_cleanup_list) ++ init_failed = 1; ++ } ++ ++ if (dev->param.is_yaffs2) ++ dev->param.use_header_file_size = 1; ++ ++ if (!init_failed && !yaffs_init_blocks(dev)) ++ init_failed = 1; ++ ++ yaffs_init_tnodes_and_objs(dev); ++ ++ if (!init_failed && !yaffs_create_initial_dir(dev)) ++ init_failed = 1; ++ ++ if (!init_failed && dev->param.is_yaffs2 && ++ !dev->param.disable_summary && ++ !yaffs_summary_init(dev)) ++ init_failed = 1; ++ ++ if (!init_failed) { ++ /* Now scan the flash. */ ++ if (dev->param.is_yaffs2) { ++ if (yaffs2_checkpt_restore(dev)) { ++ yaffs_check_obj_details_loaded(dev->root_dir); ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT | ++ YAFFS_TRACE_MOUNT, ++ "yaffs: restored from checkpoint" ++ ); ++ } else { ++ ++ /* Clean up the mess caused by an aborted ++ * checkpoint load then scan backwards. ++ */ ++ yaffs_deinit_blocks(dev); ++ ++ yaffs_deinit_tnodes_and_objs(dev); ++ ++ dev->n_erased_blocks = 0; ++ dev->n_free_chunks = 0; ++ dev->alloc_block = -1; ++ dev->alloc_page = -1; ++ dev->n_deleted_files = 0; ++ dev->n_unlinked_files = 0; ++ dev->n_bg_deletions = 0; ++ ++ if (!init_failed && !yaffs_init_blocks(dev)) ++ init_failed = 1; ++ ++ yaffs_init_tnodes_and_objs(dev); ++ ++ if (!init_failed ++ && !yaffs_create_initial_dir(dev)) ++ init_failed = 1; ++ ++ if (!init_failed && !yaffs2_scan_backwards(dev)) ++ init_failed = 1; ++ } ++ } else if (!yaffs1_scan(dev)) { ++ init_failed = 1; ++ } ++ ++ yaffs_strip_deleted_objs(dev); ++ yaffs_fix_hanging_objs(dev); ++ if (dev->param.empty_lost_n_found) ++ yaffs_empty_l_n_f(dev); ++ } ++ ++ if (init_failed) { ++ /* Clean up the mess */ ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "yaffs: yaffs_guts_initialise() aborted."); ++ ++ yaffs_deinitialise(dev); ++ return YAFFS_FAIL; ++ } ++ ++ /* Zero out stats */ ++ dev->n_page_reads = 0; ++ dev->n_page_writes = 0; ++ dev->n_erasures = 0; ++ dev->n_gc_copies = 0; ++ dev->n_retried_writes = 0; ++ ++ dev->n_retired_blocks = 0; ++ ++ yaffs_verify_free_chunks(dev); ++ yaffs_verify_blocks(dev); ++ ++ /* Clean up any aborted checkpoint data */ ++ if (!dev->is_checkpointed && dev->blocks_in_checkpt > 0) ++ yaffs2_checkpt_invalidate(dev); ++ ++ yaffs_trace(YAFFS_TRACE_TRACING, ++ "yaffs: yaffs_guts_initialise() done."); ++ return YAFFS_OK; ++} ++ ++void yaffs_deinitialise(struct yaffs_dev *dev) ++{ ++ if (dev->is_mounted) { ++ int i; ++ ++ yaffs_deinit_blocks(dev); ++ yaffs_deinit_tnodes_and_objs(dev); ++ yaffs_summary_deinit(dev); ++ ++ if (dev->param.n_caches > 0 && dev->cache) { ++ ++ for (i = 0; i < dev->param.n_caches; i++) { ++ kfree(dev->cache[i].data); ++ dev->cache[i].data = NULL; ++ } ++ ++ kfree(dev->cache); ++ dev->cache = NULL; ++ } ++ ++ kfree(dev->gc_cleanup_list); ++ ++ for (i = 0; i < YAFFS_N_TEMP_BUFFERS; i++) ++ kfree(dev->temp_buffer[i].buffer); ++ ++ dev->is_mounted = 0; ++ ++ yaffs_deinit_nand(dev); ++ } ++} ++ ++int yaffs_count_free_chunks(struct yaffs_dev *dev) ++{ ++ int n_free = 0; ++ int b; ++ struct yaffs_block_info *blk; ++ ++ blk = dev->block_info; ++ for (b = dev->internal_start_block; b <= dev->internal_end_block; b++) { ++ switch (blk->block_state) { ++ case YAFFS_BLOCK_STATE_EMPTY: ++ case YAFFS_BLOCK_STATE_ALLOCATING: ++ case YAFFS_BLOCK_STATE_COLLECTING: ++ case YAFFS_BLOCK_STATE_FULL: ++ n_free += ++ (dev->param.chunks_per_block - blk->pages_in_use + ++ blk->soft_del_pages); ++ break; ++ default: ++ break; ++ } ++ blk++; ++ } ++ return n_free; ++} ++ ++int yaffs_get_n_free_chunks(struct yaffs_dev *dev) ++{ ++ /* This is what we report to the outside world */ ++ int n_free; ++ int n_dirty_caches; ++ int blocks_for_checkpt; ++ int i; ++ ++ n_free = dev->n_free_chunks; ++ n_free += dev->n_deleted_files; ++ ++ /* Now count and subtract the number of dirty chunks in the cache. */ ++ ++ for (n_dirty_caches = 0, i = 0; i < dev->param.n_caches; i++) { ++ if (dev->cache[i].dirty) ++ n_dirty_caches++; ++ } ++ ++ n_free -= n_dirty_caches; ++ ++ n_free -= ++ ((dev->param.n_reserved_blocks + 1) * dev->param.chunks_per_block); ++ ++ /* Now figure checkpoint space and report that... */ ++ blocks_for_checkpt = yaffs_calc_checkpt_blocks_required(dev); ++ ++ n_free -= (blocks_for_checkpt * dev->param.chunks_per_block); ++ ++ if (n_free < 0) ++ n_free = 0; ++ ++ return n_free; ++} ++ ++ ++ ++/* ++ * Marshalling functions to get loff_t file sizes into and out of ++ * object headers. ++ */ ++void yaffs_oh_size_load(struct yaffs_obj_hdr *oh, loff_t fsize) ++{ ++ oh->file_size_low = (fsize & 0xFFFFFFFF); ++ oh->file_size_high = ((fsize >> 32) & 0xFFFFFFFF); ++} ++ ++loff_t yaffs_oh_to_size(struct yaffs_obj_hdr *oh) ++{ ++ loff_t retval; ++ ++ if (sizeof(loff_t) >= 8 && ~(oh->file_size_high)) ++ retval = (((loff_t) oh->file_size_high) << 32) | ++ (((loff_t) oh->file_size_low) & 0xFFFFFFFF); ++ else ++ retval = (loff_t) oh->file_size_low; ++ ++ return retval; ++} ++ ++ ++void yaffs_count_blocks_by_state(struct yaffs_dev *dev, int bs[10]) ++{ ++ int i; ++ struct yaffs_block_info *bi; ++ int s; ++ ++ for(i = 0; i < 10; i++) ++ bs[i] = 0; ++ ++ for(i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ bi = yaffs_get_block_info(dev, i); ++ s = bi->block_state; ++ if(s > YAFFS_BLOCK_STATE_DEAD || s < YAFFS_BLOCK_STATE_UNKNOWN) ++ bs[0]++; ++ else ++ bs[s]++; ++ } ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_guts.h linux-3.14.4/fs/yaffs2/yaffs_guts.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_guts.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_guts.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,1007 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_GUTS_H__ ++#define __YAFFS_GUTS_H__ ++ ++#include "yportenv.h" ++ ++#define YAFFS_OK 1 ++#define YAFFS_FAIL 0 ++ ++/* Give us a Y=0x59, ++ * Give us an A=0x41, ++ * Give us an FF=0xff ++ * Give us an S=0x53 ++ * And what have we got... ++ */ ++#define YAFFS_MAGIC 0x5941ff53 ++ ++/* ++ * Tnodes form a tree with the tnodes in "levels" ++ * Levels greater than 0 hold 8 slots which point to other tnodes. ++ * Those at level 0 hold 16 slots which point to chunks in NAND. ++ * ++ * A maximum level of 8 thust supports files of size up to: ++ * ++ * 2^(3*MAX_LEVEL+4) ++ * ++ * Thus a max level of 8 supports files with up to 2^^28 chunks which gives ++ * a maximum file size of around 512Gbytees with 2k chunks. ++ */ ++#define YAFFS_NTNODES_LEVEL0 16 ++#define YAFFS_TNODES_LEVEL0_BITS 4 ++#define YAFFS_TNODES_LEVEL0_MASK 0xf ++ ++#define YAFFS_NTNODES_INTERNAL (YAFFS_NTNODES_LEVEL0 / 2) ++#define YAFFS_TNODES_INTERNAL_BITS (YAFFS_TNODES_LEVEL0_BITS - 1) ++#define YAFFS_TNODES_INTERNAL_MASK 0x7 ++#define YAFFS_TNODES_MAX_LEVEL 8 ++#define YAFFS_TNODES_MAX_BITS (YAFFS_TNODES_LEVEL0_BITS + \ ++ YAFFS_TNODES_INTERNAL_BITS * \ ++ YAFFS_TNODES_MAX_LEVEL) ++#define YAFFS_MAX_CHUNK_ID ((1 << YAFFS_TNODES_MAX_BITS) - 1) ++ ++#define YAFFS_MAX_FILE_SIZE_32 0x7fffffff ++ ++/* Constants for YAFFS1 mode */ ++#define YAFFS_BYTES_PER_SPARE 16 ++#define YAFFS_BYTES_PER_CHUNK 512 ++#define YAFFS_CHUNK_SIZE_SHIFT 9 ++#define YAFFS_CHUNKS_PER_BLOCK 32 ++#define YAFFS_BYTES_PER_BLOCK (YAFFS_CHUNKS_PER_BLOCK*YAFFS_BYTES_PER_CHUNK) ++ ++#define YAFFS_MIN_YAFFS2_CHUNK_SIZE 1024 ++#define YAFFS_MIN_YAFFS2_SPARE_SIZE 32 ++ ++ ++ ++#define YAFFS_ALLOCATION_NOBJECTS 100 ++#define YAFFS_ALLOCATION_NTNODES 100 ++#define YAFFS_ALLOCATION_NLINKS 100 ++ ++#define YAFFS_NOBJECT_BUCKETS 256 ++ ++#define YAFFS_OBJECT_SPACE 0x40000 ++#define YAFFS_MAX_OBJECT_ID (YAFFS_OBJECT_SPACE - 1) ++ ++/* Binary data version stamps */ ++#define YAFFS_SUMMARY_VERSION 1 ++#define YAFFS_CHECKPOINT_VERSION 7 ++ ++#ifdef CONFIG_YAFFS_UNICODE ++#define YAFFS_MAX_NAME_LENGTH 127 ++#define YAFFS_MAX_ALIAS_LENGTH 79 ++#else ++#define YAFFS_MAX_NAME_LENGTH 255 ++#define YAFFS_MAX_ALIAS_LENGTH 159 ++#endif ++ ++#define YAFFS_SHORT_NAME_LENGTH 15 ++ ++/* Some special object ids for pseudo objects */ ++#define YAFFS_OBJECTID_ROOT 1 ++#define YAFFS_OBJECTID_LOSTNFOUND 2 ++#define YAFFS_OBJECTID_UNLINKED 3 ++#define YAFFS_OBJECTID_DELETED 4 ++ ++/* Fake object Id for summary data */ ++#define YAFFS_OBJECTID_SUMMARY 0x10 ++ ++/* Pseudo object ids for checkpointing */ ++#define YAFFS_OBJECTID_CHECKPOINT_DATA 0x20 ++#define YAFFS_SEQUENCE_CHECKPOINT_DATA 0x21 ++ ++#define YAFFS_MAX_SHORT_OP_CACHES 20 ++ ++#define YAFFS_N_TEMP_BUFFERS 6 ++ ++/* We limit the number attempts at sucessfully saving a chunk of data. ++ * Small-page devices have 32 pages per block; large-page devices have 64. ++ * Default to something in the order of 5 to 10 blocks worth of chunks. ++ */ ++#define YAFFS_WR_ATTEMPTS (5*64) ++ ++/* Sequence numbers are used in YAFFS2 to determine block allocation order. ++ * The range is limited slightly to help distinguish bad numbers from good. ++ * This also allows us to perhaps in the future use special numbers for ++ * special purposes. ++ * EFFFFF00 allows the allocation of 8 blocks/second (~1Mbytes) for 15 years, ++ * and is a larger number than the lifetime of a 2GB device. ++ */ ++#define YAFFS_LOWEST_SEQUENCE_NUMBER 0x00001000 ++#define YAFFS_HIGHEST_SEQUENCE_NUMBER 0xefffff00 ++ ++/* Special sequence number for bad block that failed to be marked bad */ ++#define YAFFS_SEQUENCE_BAD_BLOCK 0xffff0000 ++ ++/* ChunkCache is used for short read/write operations.*/ ++struct yaffs_cache { ++ struct yaffs_obj *object; ++ int chunk_id; ++ int last_use; ++ int dirty; ++ int n_bytes; /* Only valid if the cache is dirty */ ++ int locked; /* Can't push out or flush while locked. */ ++ u8 *data; ++}; ++ ++/* yaffs1 tags structures in RAM ++ * NB This uses bitfield. Bitfields should not straddle a u32 boundary ++ * otherwise the structure size will get blown out. ++ */ ++ ++struct yaffs_tags { ++ u32 chunk_id:20; ++ u32 serial_number:2; ++ u32 n_bytes_lsb:10; ++ u32 obj_id:18; ++ u32 ecc:12; ++ u32 n_bytes_msb:2; ++}; ++ ++union yaffs_tags_union { ++ struct yaffs_tags as_tags; ++ u8 as_bytes[8]; ++}; ++ ++ ++/* Stuff used for extended tags in YAFFS2 */ ++ ++enum yaffs_ecc_result { ++ YAFFS_ECC_RESULT_UNKNOWN, ++ YAFFS_ECC_RESULT_NO_ERROR, ++ YAFFS_ECC_RESULT_FIXED, ++ YAFFS_ECC_RESULT_UNFIXED ++}; ++ ++enum yaffs_obj_type { ++ YAFFS_OBJECT_TYPE_UNKNOWN, ++ YAFFS_OBJECT_TYPE_FILE, ++ YAFFS_OBJECT_TYPE_SYMLINK, ++ YAFFS_OBJECT_TYPE_DIRECTORY, ++ YAFFS_OBJECT_TYPE_HARDLINK, ++ YAFFS_OBJECT_TYPE_SPECIAL ++}; ++ ++#define YAFFS_OBJECT_TYPE_MAX YAFFS_OBJECT_TYPE_SPECIAL ++ ++struct yaffs_ext_tags { ++ unsigned chunk_used; /* Status of the chunk: used or unused */ ++ unsigned obj_id; /* If 0 this is not used */ ++ unsigned chunk_id; /* If 0 this is a header, else a data chunk */ ++ unsigned n_bytes; /* Only valid for data chunks */ ++ ++ /* The following stuff only has meaning when we read */ ++ enum yaffs_ecc_result ecc_result; ++ unsigned block_bad; ++ ++ /* YAFFS 1 stuff */ ++ unsigned is_deleted; /* The chunk is marked deleted */ ++ unsigned serial_number; /* Yaffs1 2-bit serial number */ ++ ++ /* YAFFS2 stuff */ ++ unsigned seq_number; /* The sequence number of this block */ ++ ++ /* Extra info if this is an object header (YAFFS2 only) */ ++ ++ unsigned extra_available; /* Extra info available if not zero */ ++ unsigned extra_parent_id; /* The parent object */ ++ unsigned extra_is_shrink; /* Is it a shrink header? */ ++ unsigned extra_shadows; /* Does this shadow another object? */ ++ ++ enum yaffs_obj_type extra_obj_type; /* What object type? */ ++ ++ loff_t extra_file_size; /* Length if it is a file */ ++ unsigned extra_equiv_id; /* Equivalent object for a hard link */ ++}; ++ ++/* Spare structure for YAFFS1 */ ++struct yaffs_spare { ++ u8 tb0; ++ u8 tb1; ++ u8 tb2; ++ u8 tb3; ++ u8 page_status; /* set to 0 to delete the chunk */ ++ u8 block_status; ++ u8 tb4; ++ u8 tb5; ++ u8 ecc1[3]; ++ u8 tb6; ++ u8 tb7; ++ u8 ecc2[3]; ++}; ++ ++/*Special structure for passing through to mtd */ ++struct yaffs_nand_spare { ++ struct yaffs_spare spare; ++ int eccres1; ++ int eccres2; ++}; ++ ++/* Block data in RAM */ ++ ++enum yaffs_block_state { ++ YAFFS_BLOCK_STATE_UNKNOWN = 0, ++ ++ YAFFS_BLOCK_STATE_SCANNING, ++ /* Being scanned */ ++ ++ YAFFS_BLOCK_STATE_NEEDS_SCAN, ++ /* The block might have something on it (ie it is allocating or full, ++ * perhaps empty) but it needs to be scanned to determine its true ++ * state. ++ * This state is only valid during scanning. ++ * NB We tolerate empty because the pre-scanner might be incapable of ++ * deciding ++ * However, if this state is returned on a YAFFS2 device, ++ * then we expect a sequence number ++ */ ++ ++ YAFFS_BLOCK_STATE_EMPTY, ++ /* This block is empty */ ++ ++ YAFFS_BLOCK_STATE_ALLOCATING, ++ /* This block is partially allocated. ++ * At least one page holds valid data. ++ * This is the one currently being used for page ++ * allocation. Should never be more than one of these. ++ * If a block is only partially allocated at mount it is treated as ++ * full. ++ */ ++ ++ YAFFS_BLOCK_STATE_FULL, ++ /* All the pages in this block have been allocated. ++ * If a block was only partially allocated when mounted we treat ++ * it as fully allocated. ++ */ ++ ++ YAFFS_BLOCK_STATE_DIRTY, ++ /* The block was full and now all chunks have been deleted. ++ * Erase me, reuse me. ++ */ ++ ++ YAFFS_BLOCK_STATE_CHECKPOINT, ++ /* This block is assigned to holding checkpoint data. */ ++ ++ YAFFS_BLOCK_STATE_COLLECTING, ++ /* This block is being garbage collected */ ++ ++ YAFFS_BLOCK_STATE_DEAD ++ /* This block has failed and is not in use */ ++}; ++ ++#define YAFFS_NUMBER_OF_BLOCK_STATES (YAFFS_BLOCK_STATE_DEAD + 1) ++ ++struct yaffs_block_info { ++ ++ s32 soft_del_pages:10; /* number of soft deleted pages */ ++ s32 pages_in_use:10; /* number of pages in use */ ++ u32 block_state:4; /* One of the above block states. */ ++ /* NB use unsigned because enum is sometimes ++ * an int */ ++ u32 needs_retiring:1; /* Data has failed on this block, */ ++ /*need to get valid data off and retire*/ ++ u32 skip_erased_check:1;/* Skip the erased check on this block */ ++ u32 gc_prioritise:1; /* An ECC check or blank check has failed. ++ Block should be prioritised for GC */ ++ u32 chunk_error_strikes:3; /* How many times we've had ecc etc ++ failures on this block and tried to reuse it */ ++ u32 has_summary:1; /* The block has a summary */ ++ ++ u32 has_shrink_hdr:1; /* This block has at least one shrink header */ ++ u32 seq_number; /* block sequence number for yaffs2 */ ++ ++}; ++ ++/* -------------------------- Object structure -------------------------------*/ ++/* This is the object structure as stored on NAND */ ++ ++struct yaffs_obj_hdr { ++ enum yaffs_obj_type type; ++ ++ /* Apply to everything */ ++ int parent_obj_id; ++ u16 sum_no_longer_used; /* checksum of name. No longer used */ ++ YCHAR name[YAFFS_MAX_NAME_LENGTH + 1]; ++ ++ /* The following apply to all object types except for hard links */ ++ u32 yst_mode; /* protection */ ++ ++ u32 yst_uid; ++ u32 yst_gid; ++ u32 yst_atime; ++ u32 yst_mtime; ++ u32 yst_ctime; ++ ++ /* File size applies to files only */ ++ u32 file_size_low; ++ ++ /* Equivalent object id applies to hard links only. */ ++ int equiv_id; ++ ++ /* Alias is for symlinks only. */ ++ YCHAR alias[YAFFS_MAX_ALIAS_LENGTH + 1]; ++ ++ u32 yst_rdev; /* stuff for block and char devices (major/min) */ ++ ++ u32 win_ctime[2]; ++ u32 win_atime[2]; ++ u32 win_mtime[2]; ++ ++ u32 inband_shadowed_obj_id; ++ u32 inband_is_shrink; ++ ++ u32 file_size_high; ++ u32 reserved[1]; ++ int shadows_obj; /* This object header shadows the ++ specified object if > 0 */ ++ ++ /* is_shrink applies to object headers written when wemake a hole. */ ++ u32 is_shrink; ++ ++}; ++ ++/*--------------------------- Tnode -------------------------- */ ++ ++struct yaffs_tnode { ++ struct yaffs_tnode *internal[YAFFS_NTNODES_INTERNAL]; ++}; ++ ++/*------------------------ Object -----------------------------*/ ++/* An object can be one of: ++ * - a directory (no data, has children links ++ * - a regular file (data.... not prunes :->). ++ * - a symlink [symbolic link] (the alias). ++ * - a hard link ++ */ ++ ++struct yaffs_file_var { ++ loff_t file_size; ++ loff_t scanned_size; ++ loff_t shrink_size; ++ int top_level; ++ struct yaffs_tnode *top; ++}; ++ ++struct yaffs_dir_var { ++ struct list_head children; /* list of child links */ ++ struct list_head dirty; /* Entry for list of dirty directories */ ++}; ++ ++struct yaffs_symlink_var { ++ YCHAR *alias; ++}; ++ ++struct yaffs_hardlink_var { ++ struct yaffs_obj *equiv_obj; ++ u32 equiv_id; ++}; ++ ++union yaffs_obj_var { ++ struct yaffs_file_var file_variant; ++ struct yaffs_dir_var dir_variant; ++ struct yaffs_symlink_var symlink_variant; ++ struct yaffs_hardlink_var hardlink_variant; ++}; ++ ++struct yaffs_obj { ++ u8 deleted:1; /* This should only apply to unlinked files. */ ++ u8 soft_del:1; /* it has also been soft deleted */ ++ u8 unlinked:1; /* An unlinked file.*/ ++ u8 fake:1; /* A fake object has no presence on NAND. */ ++ u8 rename_allowed:1; /* Some objects cannot be renamed. */ ++ u8 unlink_allowed:1; ++ u8 dirty:1; /* the object needs to be written to flash */ ++ u8 valid:1; /* When the file system is being loaded up, this ++ * object might be created before the data ++ * is available ++ * ie. file data chunks encountered before ++ * the header. ++ */ ++ u8 lazy_loaded:1; /* This object has been lazy loaded and ++ * is missing some detail */ ++ ++ u8 defered_free:1; /* Object is removed from NAND, but is ++ * still in the inode cache. ++ * Free of object is defered. ++ * until the inode is released. ++ */ ++ u8 being_created:1; /* This object is still being created ++ * so skip some verification checks. */ ++ u8 is_shadowed:1; /* This object is shadowed on the way ++ * to being renamed. */ ++ ++ u8 xattr_known:1; /* We know if this has object has xattribs ++ * or not. */ ++ u8 has_xattr:1; /* This object has xattribs. ++ * Only valid if xattr_known. */ ++ ++ u8 serial; /* serial number of chunk in NAND.*/ ++ u16 sum; /* sum of the name to speed searching */ ++ ++ struct yaffs_dev *my_dev; /* The device I'm on */ ++ ++ struct list_head hash_link; /* list of objects in hash bucket */ ++ ++ struct list_head hard_links; /* hard linked object chain*/ ++ ++ /* directory structure stuff */ ++ /* also used for linking up the free list */ ++ struct yaffs_obj *parent; ++ struct list_head siblings; ++ ++ /* Where's my object header in NAND? */ ++ int hdr_chunk; ++ ++ int n_data_chunks; /* Number of data chunks for this file. */ ++ ++ u32 obj_id; /* the object id value */ ++ ++ u32 yst_mode; ++ ++ YCHAR short_name[YAFFS_SHORT_NAME_LENGTH + 1]; ++ ++#ifdef CONFIG_YAFFS_WINCE ++ u32 win_ctime[2]; ++ u32 win_mtime[2]; ++ u32 win_atime[2]; ++#else ++ u32 yst_uid; ++ u32 yst_gid; ++ u32 yst_atime; ++ u32 yst_mtime; ++ u32 yst_ctime; ++#endif ++ ++ u32 yst_rdev; ++ ++ void *my_inode; ++ ++ enum yaffs_obj_type variant_type; ++ ++ union yaffs_obj_var variant; ++ ++}; ++ ++struct yaffs_obj_bucket { ++ struct list_head list; ++ int count; ++}; ++ ++/* yaffs_checkpt_obj holds the definition of an object as dumped ++ * by checkpointing. ++ */ ++ ++struct yaffs_checkpt_obj { ++ int struct_type; ++ u32 obj_id; ++ u32 parent_id; ++ int hdr_chunk; ++ enum yaffs_obj_type variant_type:3; ++ u8 deleted:1; ++ u8 soft_del:1; ++ u8 unlinked:1; ++ u8 fake:1; ++ u8 rename_allowed:1; ++ u8 unlink_allowed:1; ++ u8 serial; ++ int n_data_chunks; ++ loff_t size_or_equiv_obj; ++}; ++ ++/*--------------------- Temporary buffers ---------------- ++ * ++ * These are chunk-sized working buffers. Each device has a few. ++ */ ++ ++struct yaffs_buffer { ++ u8 *buffer; ++ int in_use; ++}; ++ ++/*----------------- Device ---------------------------------*/ ++ ++struct yaffs_param { ++ const YCHAR *name; ++ ++ /* ++ * Entry parameters set up way early. Yaffs sets up the rest. ++ * The structure should be zeroed out before use so that unused ++ * and default values are zero. ++ */ ++ ++ int inband_tags; /* Use unband tags */ ++ u32 total_bytes_per_chunk; /* Should be >= 512, does not need to ++ be a power of 2 */ ++ int chunks_per_block; /* does not need to be a power of 2 */ ++ int spare_bytes_per_chunk; /* spare area size */ ++ int start_block; /* Start block we're allowed to use */ ++ int end_block; /* End block we're allowed to use */ ++ int n_reserved_blocks; /* Tuneable so that we can reduce ++ * reserved blocks on NOR and RAM. */ ++ ++ int n_caches; /* If <= 0, then short op caching is disabled, ++ * else the number of short op caches. ++ */ ++ int cache_bypass_aligned; /* If non-zero then bypass the cache for ++ * aligned writes. ++ */ ++ ++ int use_nand_ecc; /* Flag to decide whether or not to use ++ * NAND driver ECC on data (yaffs1) */ ++ int tags_9bytes; /* Use 9 byte tags */ ++ int no_tags_ecc; /* Flag to decide whether or not to do ECC ++ * on packed tags (yaffs2) */ ++ ++ int is_yaffs2; /* Use yaffs2 mode on this device */ ++ ++ int empty_lost_n_found; /* Auto-empty lost+found directory on mount */ ++ ++ int refresh_period; /* How often to check for a block refresh */ ++ ++ /* Checkpoint control. Can be set before or after initialisation */ ++ u8 skip_checkpt_rd; ++ u8 skip_checkpt_wr; ++ ++ int enable_xattr; /* Enable xattribs */ ++ ++ int max_objects; /* ++ * Set to limit the number of objects created. ++ * 0 = no limit. ++ */ ++ ++ /* The remove_obj_fn function must be supplied by OS flavours that ++ * need it. ++ * yaffs direct uses it to implement the faster readdir. ++ * Linux uses it to protect the directory during unlocking. ++ */ ++ void (*remove_obj_fn) (struct yaffs_obj *obj); ++ ++ /* Callback to mark the superblock dirty */ ++ void (*sb_dirty_fn) (struct yaffs_dev *dev); ++ ++ /* Callback to control garbage collection. */ ++ unsigned (*gc_control_fn) (struct yaffs_dev *dev); ++ ++ /* Debug control flags. Don't use unless you know what you're doing */ ++ int use_header_file_size; /* Flag to determine if we should use ++ * file sizes from the header */ ++ int disable_lazy_load; /* Disable lazy loading on this device */ ++ int wide_tnodes_disabled; /* Set to disable wide tnodes */ ++ int disable_soft_del; /* yaffs 1 only: Set to disable the use of ++ * softdeletion. */ ++ ++ int defered_dir_update; /* Set to defer directory updates */ ++ ++#ifdef CONFIG_YAFFS_AUTO_UNICODE ++ int auto_unicode; ++#endif ++ int always_check_erased; /* Force chunk erased check always on */ ++ ++ int disable_summary; ++ int disable_bad_block_marking; ++ ++}; ++ ++struct yaffs_driver { ++ int (*drv_write_chunk_fn) (struct yaffs_dev *dev, int nand_chunk, ++ const u8 *data, int data_len, ++ const u8 *oob, int oob_len); ++ int (*drv_read_chunk_fn) (struct yaffs_dev *dev, int nand_chunk, ++ u8 *data, int data_len, ++ u8 *oob, int oob_len, ++ enum yaffs_ecc_result *ecc_result); ++ int (*drv_erase_fn) (struct yaffs_dev *dev, int block_no); ++ int (*drv_mark_bad_fn) (struct yaffs_dev *dev, int block_no); ++ int (*drv_check_bad_fn) (struct yaffs_dev *dev, int block_no); ++ int (*drv_initialise_fn) (struct yaffs_dev *dev); ++ int (*drv_deinitialise_fn) (struct yaffs_dev *dev); ++}; ++ ++struct yaffs_tags_handler { ++ int (*write_chunk_tags_fn) (struct yaffs_dev *dev, ++ int nand_chunk, const u8 *data, ++ const struct yaffs_ext_tags *tags); ++ int (*read_chunk_tags_fn) (struct yaffs_dev *dev, ++ int nand_chunk, u8 *data, ++ struct yaffs_ext_tags *tags); ++ ++ int (*query_block_fn) (struct yaffs_dev *dev, int block_no, ++ enum yaffs_block_state *state, ++ u32 *seq_number); ++ int (*mark_bad_fn) (struct yaffs_dev *dev, int block_no); ++}; ++ ++struct yaffs_dev { ++ struct yaffs_param param; ++ struct yaffs_driver drv; ++ struct yaffs_tags_handler tagger; ++ ++ /* Context storage. Holds extra OS specific data for this device */ ++ ++ void *os_context; ++ void *driver_context; ++ ++ struct list_head dev_list; ++ ++ int ll_init; ++ /* Runtime parameters. Set up by YAFFS. */ ++ int data_bytes_per_chunk; ++ ++ /* Non-wide tnode stuff */ ++ u16 chunk_grp_bits; /* Number of bits that need to be resolved if ++ * the tnodes are not wide enough. ++ */ ++ u16 chunk_grp_size; /* == 2^^chunk_grp_bits */ ++ ++ /* Stuff to support wide tnodes */ ++ u32 tnode_width; ++ u32 tnode_mask; ++ u32 tnode_size; ++ ++ /* Stuff for figuring out file offset to chunk conversions */ ++ u32 chunk_shift; /* Shift value */ ++ u32 chunk_div; /* Divisor after shifting: 1 for 2^n sizes */ ++ u32 chunk_mask; /* Mask to use for power-of-2 case */ ++ ++ int is_mounted; ++ int read_only; ++ int is_checkpointed; ++ ++ /* Stuff to support block offsetting to support start block zero */ ++ int internal_start_block; ++ int internal_end_block; ++ int block_offset; ++ int chunk_offset; ++ ++ /* Runtime checkpointing stuff */ ++ int checkpt_page_seq; /* running sequence number of checkpt pages */ ++ int checkpt_byte_count; ++ int checkpt_byte_offs; ++ u8 *checkpt_buffer; ++ int checkpt_open_write; ++ int blocks_in_checkpt; ++ int checkpt_cur_chunk; ++ int checkpt_cur_block; ++ int checkpt_next_block; ++ int *checkpt_block_list; ++ int checkpt_max_blocks; ++ u32 checkpt_sum; ++ u32 checkpt_xor; ++ ++ int checkpoint_blocks_required; /* Number of blocks needed to store ++ * current checkpoint set */ ++ ++ /* Block Info */ ++ struct yaffs_block_info *block_info; ++ u8 *chunk_bits; /* bitmap of chunks in use */ ++ u8 block_info_alt:1; /* allocated using alternative alloc */ ++ u8 chunk_bits_alt:1; /* allocated using alternative alloc */ ++ int chunk_bit_stride; /* Number of bytes of chunk_bits per block. ++ * Must be consistent with chunks_per_block. ++ */ ++ ++ int n_erased_blocks; ++ int alloc_block; /* Current block being allocated off */ ++ u32 alloc_page; ++ int alloc_block_finder; /* Used to search for next allocation block */ ++ ++ /* Object and Tnode memory management */ ++ void *allocator; ++ int n_obj; ++ int n_tnodes; ++ ++ int n_hardlinks; ++ ++ struct yaffs_obj_bucket obj_bucket[YAFFS_NOBJECT_BUCKETS]; ++ u32 bucket_finder; ++ ++ int n_free_chunks; ++ ++ /* Garbage collection control */ ++ u32 *gc_cleanup_list; /* objects to delete at the end of a GC. */ ++ u32 n_clean_ups; ++ ++ unsigned has_pending_prioritised_gc; /* We think this device might ++ have pending prioritised gcs */ ++ unsigned gc_disable; ++ unsigned gc_block_finder; ++ unsigned gc_dirtiest; ++ unsigned gc_pages_in_use; ++ unsigned gc_not_done; ++ unsigned gc_block; ++ unsigned gc_chunk; ++ unsigned gc_skip; ++ struct yaffs_summary_tags *gc_sum_tags; ++ ++ /* Special directories */ ++ struct yaffs_obj *root_dir; ++ struct yaffs_obj *lost_n_found; ++ ++ int buffered_block; /* Which block is buffered here? */ ++ int doing_buffered_block_rewrite; ++ ++ struct yaffs_cache *cache; ++ int cache_last_use; ++ ++ /* Stuff for background deletion and unlinked files. */ ++ struct yaffs_obj *unlinked_dir; /* Directory where unlinked and deleted ++ files live. */ ++ struct yaffs_obj *del_dir; /* Directory where deleted objects are ++ sent to disappear. */ ++ struct yaffs_obj *unlinked_deletion; /* Current file being ++ background deleted. */ ++ int n_deleted_files; /* Count of files awaiting deletion; */ ++ int n_unlinked_files; /* Count of unlinked files. */ ++ int n_bg_deletions; /* Count of background deletions. */ ++ ++ /* Temporary buffer management */ ++ struct yaffs_buffer temp_buffer[YAFFS_N_TEMP_BUFFERS]; ++ int max_temp; ++ int temp_in_use; ++ int unmanaged_buffer_allocs; ++ int unmanaged_buffer_deallocs; ++ ++ /* yaffs2 runtime stuff */ ++ unsigned seq_number; /* Sequence number of currently ++ allocating block */ ++ unsigned oldest_dirty_seq; ++ unsigned oldest_dirty_block; ++ ++ /* Block refreshing */ ++ int refresh_skip; /* A skip down counter. ++ * Refresh happens when this gets to zero. */ ++ ++ /* Dirty directory handling */ ++ struct list_head dirty_dirs; /* List of dirty directories */ ++ ++ /* Summary */ ++ int chunks_per_summary; ++ struct yaffs_summary_tags *sum_tags; ++ ++ /* Statistics */ ++ u32 n_page_writes; ++ u32 n_page_reads; ++ u32 n_erasures; ++ u32 n_bad_queries; ++ u32 n_bad_markings; ++ u32 n_erase_failures; ++ u32 n_gc_copies; ++ u32 all_gcs; ++ u32 passive_gc_count; ++ u32 oldest_dirty_gc_count; ++ u32 n_gc_blocks; ++ u32 bg_gcs; ++ u32 n_retried_writes; ++ u32 n_retired_blocks; ++ u32 n_ecc_fixed; ++ u32 n_ecc_unfixed; ++ u32 n_tags_ecc_fixed; ++ u32 n_tags_ecc_unfixed; ++ u32 n_deletions; ++ u32 n_unmarked_deletions; ++ u32 refresh_count; ++ u32 cache_hits; ++ u32 tags_used; ++ u32 summary_used; ++ ++}; ++ ++/* The CheckpointDevice structure holds the device information that changes ++ *at runtime and must be preserved over unmount/mount cycles. ++ */ ++struct yaffs_checkpt_dev { ++ int struct_type; ++ int n_erased_blocks; ++ int alloc_block; /* Current block being allocated off */ ++ u32 alloc_page; ++ int n_free_chunks; ++ ++ int n_deleted_files; /* Count of files awaiting deletion; */ ++ int n_unlinked_files; /* Count of unlinked files. */ ++ int n_bg_deletions; /* Count of background deletions. */ ++ ++ /* yaffs2 runtime stuff */ ++ unsigned seq_number; /* Sequence number of currently ++ * allocating block */ ++ ++}; ++ ++struct yaffs_checkpt_validity { ++ int struct_type; ++ u32 magic; ++ u32 version; ++ u32 head; ++}; ++ ++struct yaffs_shadow_fixer { ++ int obj_id; ++ int shadowed_id; ++ struct yaffs_shadow_fixer *next; ++}; ++ ++/* Structure for doing xattr modifications */ ++struct yaffs_xattr_mod { ++ int set; /* If 0 then this is a deletion */ ++ const YCHAR *name; ++ const void *data; ++ int size; ++ int flags; ++ int result; ++}; ++ ++/*----------------------- YAFFS Functions -----------------------*/ ++ ++int yaffs_guts_initialise(struct yaffs_dev *dev); ++void yaffs_deinitialise(struct yaffs_dev *dev); ++ ++int yaffs_get_n_free_chunks(struct yaffs_dev *dev); ++ ++int yaffs_rename_obj(struct yaffs_obj *old_dir, const YCHAR * old_name, ++ struct yaffs_obj *new_dir, const YCHAR * new_name); ++ ++int yaffs_unlinker(struct yaffs_obj *dir, const YCHAR * name); ++int yaffs_del_obj(struct yaffs_obj *obj); ++struct yaffs_obj *yaffs_retype_obj(struct yaffs_obj *obj, ++ enum yaffs_obj_type type); ++ ++ ++int yaffs_get_obj_name(struct yaffs_obj *obj, YCHAR * name, int buffer_size); ++loff_t yaffs_get_obj_length(struct yaffs_obj *obj); ++int yaffs_get_obj_inode(struct yaffs_obj *obj); ++unsigned yaffs_get_obj_type(struct yaffs_obj *obj); ++int yaffs_get_obj_link_count(struct yaffs_obj *obj); ++ ++/* File operations */ ++int yaffs_file_rd(struct yaffs_obj *obj, u8 * buffer, loff_t offset, ++ int n_bytes); ++int yaffs_wr_file(struct yaffs_obj *obj, const u8 * buffer, loff_t offset, ++ int n_bytes, int write_trhrough); ++int yaffs_resize_file(struct yaffs_obj *obj, loff_t new_size); ++ ++struct yaffs_obj *yaffs_create_file(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid); ++ ++int yaffs_flush_file(struct yaffs_obj *obj, int update_time, int data_sync); ++ ++/* Flushing and checkpointing */ ++void yaffs_flush_whole_cache(struct yaffs_dev *dev); ++ ++int yaffs_checkpoint_save(struct yaffs_dev *dev); ++int yaffs_checkpoint_restore(struct yaffs_dev *dev); ++ ++/* Directory operations */ ++struct yaffs_obj *yaffs_create_dir(struct yaffs_obj *parent, const YCHAR *name, ++ u32 mode, u32 uid, u32 gid); ++struct yaffs_obj *yaffs_find_by_name(struct yaffs_obj *the_dir, ++ const YCHAR *name); ++struct yaffs_obj *yaffs_find_by_number(struct yaffs_dev *dev, u32 number); ++ ++/* Link operations */ ++struct yaffs_obj *yaffs_link_obj(struct yaffs_obj *parent, const YCHAR *name, ++ struct yaffs_obj *equiv_obj); ++ ++struct yaffs_obj *yaffs_get_equivalent_obj(struct yaffs_obj *obj); ++ ++/* Symlink operations */ ++struct yaffs_obj *yaffs_create_symlink(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid, const YCHAR *alias); ++YCHAR *yaffs_get_symlink_alias(struct yaffs_obj *obj); ++ ++/* Special inodes (fifos, sockets and devices) */ ++struct yaffs_obj *yaffs_create_special(struct yaffs_obj *parent, ++ const YCHAR *name, u32 mode, u32 uid, ++ u32 gid, u32 rdev); ++ ++int yaffs_set_xattrib(struct yaffs_obj *obj, const YCHAR *name, ++ const void *value, int size, int flags); ++int yaffs_get_xattrib(struct yaffs_obj *obj, const YCHAR *name, void *value, ++ int size); ++int yaffs_list_xattrib(struct yaffs_obj *obj, char *buffer, int size); ++int yaffs_remove_xattrib(struct yaffs_obj *obj, const YCHAR *name); ++ ++/* Special directories */ ++struct yaffs_obj *yaffs_root(struct yaffs_dev *dev); ++struct yaffs_obj *yaffs_lost_n_found(struct yaffs_dev *dev); ++ ++void yaffs_handle_defered_free(struct yaffs_obj *obj); ++ ++void yaffs_update_dirty_dirs(struct yaffs_dev *dev); ++ ++int yaffs_bg_gc(struct yaffs_dev *dev, unsigned urgency); ++ ++/* Debug dump */ ++int yaffs_dump_obj(struct yaffs_obj *obj); ++ ++void yaffs_guts_test(struct yaffs_dev *dev); ++int yaffs_guts_ll_init(struct yaffs_dev *dev); ++ ++ ++/* A few useful functions to be used within the core files*/ ++void yaffs_chunk_del(struct yaffs_dev *dev, int chunk_id, int mark_flash, ++ int lyn); ++int yaffs_check_ff(u8 *buffer, int n_bytes); ++void yaffs_handle_chunk_error(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi); ++ ++u8 *yaffs_get_temp_buffer(struct yaffs_dev *dev); ++void yaffs_release_temp_buffer(struct yaffs_dev *dev, u8 *buffer); ++ ++struct yaffs_obj *yaffs_find_or_create_by_number(struct yaffs_dev *dev, ++ int number, ++ enum yaffs_obj_type type); ++int yaffs_put_chunk_in_file(struct yaffs_obj *in, int inode_chunk, ++ int nand_chunk, int in_scan); ++void yaffs_set_obj_name(struct yaffs_obj *obj, const YCHAR *name); ++void yaffs_set_obj_name_from_oh(struct yaffs_obj *obj, ++ const struct yaffs_obj_hdr *oh); ++void yaffs_add_obj_to_dir(struct yaffs_obj *directory, struct yaffs_obj *obj); ++YCHAR *yaffs_clone_str(const YCHAR *str); ++void yaffs_link_fixup(struct yaffs_dev *dev, struct list_head *hard_list); ++void yaffs_block_became_dirty(struct yaffs_dev *dev, int block_no); ++int yaffs_update_oh(struct yaffs_obj *in, const YCHAR *name, ++ int force, int is_shrink, int shadows, ++ struct yaffs_xattr_mod *xop); ++void yaffs_handle_shadowed_obj(struct yaffs_dev *dev, int obj_id, ++ int backward_scanning); ++int yaffs_check_alloc_available(struct yaffs_dev *dev, int n_chunks); ++struct yaffs_tnode *yaffs_get_tnode(struct yaffs_dev *dev); ++struct yaffs_tnode *yaffs_add_find_tnode_0(struct yaffs_dev *dev, ++ struct yaffs_file_var *file_struct, ++ u32 chunk_id, ++ struct yaffs_tnode *passed_tn); ++ ++int yaffs_do_file_wr(struct yaffs_obj *in, const u8 *buffer, loff_t offset, ++ int n_bytes, int write_trhrough); ++void yaffs_resize_file_down(struct yaffs_obj *obj, loff_t new_size); ++void yaffs_skip_rest_of_block(struct yaffs_dev *dev); ++ ++int yaffs_count_free_chunks(struct yaffs_dev *dev); ++ ++struct yaffs_tnode *yaffs_find_tnode_0(struct yaffs_dev *dev, ++ struct yaffs_file_var *file_struct, ++ u32 chunk_id); ++ ++u32 yaffs_get_group_base(struct yaffs_dev *dev, struct yaffs_tnode *tn, ++ unsigned pos); ++ ++int yaffs_is_non_empty_dir(struct yaffs_obj *obj); ++ ++int yaffs_guts_format_dev(struct yaffs_dev *dev); ++ ++void yaffs_addr_to_chunk(struct yaffs_dev *dev, loff_t addr, ++ int *chunk_out, u32 *offset_out); ++/* ++ * Marshalling functions to get loff_t file sizes into aand out of ++ * object headers. ++ */ ++void yaffs_oh_size_load(struct yaffs_obj_hdr *oh, loff_t fsize); ++loff_t yaffs_oh_to_size(struct yaffs_obj_hdr *oh); ++loff_t yaffs_max_file_size(struct yaffs_dev *dev); ++ ++/* ++ * Debug function to count number of blocks in each state ++ * NB Needs to be called with correct number of integers ++ */ ++ ++void yaffs_count_blocks_by_state(struct yaffs_dev *dev, int bs[10]); ++ ++int yaffs_find_chunk_in_file(struct yaffs_obj *in, int inode_chunk, ++ struct yaffs_ext_tags *tags); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_linux.h linux-3.14.4/fs/yaffs2/yaffs_linux.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_linux.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_linux.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,48 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_LINUX_H__ ++#define __YAFFS_LINUX_H__ ++ ++#include "yportenv.h" ++ ++struct yaffs_linux_context { ++ struct list_head context_list; /* List of these we have mounted */ ++ struct yaffs_dev *dev; ++ struct super_block *super; ++ struct task_struct *bg_thread; /* Background thread for this device */ ++ int bg_running; ++ struct mutex gross_lock; /* Gross locking mutex*/ ++ u8 *spare_buffer; /* For mtdif2 use. Don't know the buffer size ++ * at compile time so we have to allocate it. ++ */ ++ struct list_head search_contexts; ++ struct task_struct *readdir_process; ++ unsigned mount_id; ++ int dirty; ++}; ++ ++#define yaffs_dev_to_lc(dev) ((struct yaffs_linux_context *)((dev)->os_context)) ++#define yaffs_dev_to_mtd(dev) ((struct mtd_info *)((dev)->driver_context)) ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++#define WRITE_SIZE_STR "writesize" ++#define WRITE_SIZE(mtd) ((mtd)->writesize) ++#else ++#define WRITE_SIZE_STR "oobblock" ++#define WRITE_SIZE(mtd) ((mtd)->oobblock) ++#endif ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_mtdif.c linux-3.14.4/fs/yaffs2/yaffs_mtdif.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_mtdif.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_mtdif.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,309 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yportenv.h" ++ ++#include "yaffs_mtdif.h" ++ ++#include "linux/mtd/mtd.h" ++#include "linux/types.h" ++#include "linux/time.h" ++#include "linux/major.h" ++#include "linux/mtd/nand.h" ++#include "linux/kernel.h" ++#include "linux/version.h" ++#include "linux/types.h" ++ ++#include "yaffs_trace.h" ++#include "yaffs_guts.h" ++#include "yaffs_linux.h" ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 2, 0)) ++#define MTD_OPS_AUTO_OOB MTD_OOB_AUTO ++#endif ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 4, 0)) ++#define mtd_erase(m, ei) (m)->erase(m, ei) ++#define mtd_write_oob(m, addr, pops) (m)->write_oob(m, addr, pops) ++#define mtd_read_oob(m, addr, pops) (m)->read_oob(m, addr, pops) ++#define mtd_block_isbad(m, offs) (m)->block_isbad(m, offs) ++#define mtd_block_markbad(m, offs) (m)->block_markbad(m, offs) ++#endif ++ ++ ++ ++int nandmtd_erase_block(struct yaffs_dev *dev, int block_no) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ u32 addr = ++ ((loff_t) block_no) * dev->param.total_bytes_per_chunk * ++ dev->param.chunks_per_block; ++ struct erase_info ei; ++ int retval = 0; ++ ++ ei.mtd = mtd; ++ ei.addr = addr; ++ ei.len = dev->param.total_bytes_per_chunk * dev->param.chunks_per_block; ++ ei.time = 1000; ++ ei.retries = 2; ++ ei.callback = NULL; ++ ei.priv = (u_long) dev; ++ ++ retval = mtd_erase(mtd, &ei); ++ ++ if (retval == 0) ++ return YAFFS_OK; ++ ++ return YAFFS_FAIL; ++} ++ ++ ++static int yaffs_mtd_write(struct yaffs_dev *dev, int nand_chunk, ++ const u8 *data, int data_len, ++ const u8 *oob, int oob_len) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ loff_t addr; ++ struct mtd_oob_ops ops; ++ int retval; ++ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "yaffs_mtd_write(%p, %d, %p, %d, %p, %d)\n", ++ dev, nand_chunk, data, data_len, oob, oob_len); ++ ++ if (!data || !data_len) { ++ data = NULL; ++ data_len = 0; ++ } ++ ++ if (!oob || !oob_len) { ++ oob = NULL; ++ oob_len = 0; ++ } ++ ++ addr = ((loff_t) nand_chunk) * dev->param.total_bytes_per_chunk; ++ memset(&ops, 0, sizeof(ops)); ++ ops.mode = MTD_OPS_AUTO_OOB; ++ ops.len = (data) ? data_len : 0; ++ ops.ooblen = oob_len; ++ ops.datbuf = (u8 *)data; ++ ops.oobbuf = (u8 *)oob; ++ ++ retval = mtd_write_oob(mtd, addr, &ops); ++ if (retval) { ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "write_oob failed, chunk %d, mtd error %d", ++ nand_chunk, retval); ++ } ++ return retval ? YAFFS_FAIL : YAFFS_OK; ++} ++ ++static int yaffs_mtd_read(struct yaffs_dev *dev, int nand_chunk, ++ u8 *data, int data_len, ++ u8 *oob, int oob_len, ++ enum yaffs_ecc_result *ecc_result) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ loff_t addr; ++ struct mtd_oob_ops ops; ++ int retval; ++ ++ addr = ((loff_t) nand_chunk) * dev->param.total_bytes_per_chunk; ++ memset(&ops, 0, sizeof(ops)); ++ ops.mode = MTD_OPS_AUTO_OOB; ++ ops.len = (data) ? data_len : 0; ++ ops.ooblen = oob_len; ++ ops.datbuf = data; ++ ops.oobbuf = oob; ++ ++#if (MTD_VERSION_CODE < MTD_VERSION(2, 6, 20)) ++ /* In MTD 2.6.18 to 2.6.19 nand_base.c:nand_do_read_oob() has a bug; ++ * help it out with ops.len = ops.ooblen when ops.datbuf == NULL. ++ */ ++ ops.len = (ops.datbuf) ? ops.len : ops.ooblen; ++#endif ++ /* Read page and oob using MTD. ++ * Check status and determine ECC result. ++ */ ++ retval = mtd_read_oob(mtd, addr, &ops); ++ if (retval) ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "read_oob failed, chunk %d, mtd error %d", ++ nand_chunk, retval); ++ ++ switch (retval) { ++ case 0: ++ /* no error */ ++ if(ecc_result) ++ *ecc_result = YAFFS_ECC_RESULT_NO_ERROR; ++ break; ++ ++ case -EUCLEAN: ++ /* MTD's ECC fixed the data */ ++ if(ecc_result) ++ *ecc_result = YAFFS_ECC_RESULT_FIXED; ++ dev->n_ecc_fixed++; ++ break; ++ ++ case -EBADMSG: ++ default: ++ /* MTD's ECC could not fix the data */ ++ dev->n_ecc_unfixed++; ++ if(ecc_result) ++ *ecc_result = YAFFS_ECC_RESULT_UNFIXED; ++ return YAFFS_FAIL; ++ } ++ ++ return YAFFS_OK; ++} ++ ++static int yaffs_mtd_erase(struct yaffs_dev *dev, int block_no) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ ++ loff_t addr; ++ struct erase_info ei; ++ int retval = 0; ++ u32 block_size; ++ ++ block_size = dev->param.total_bytes_per_chunk * ++ dev->param.chunks_per_block; ++ addr = ((loff_t) block_no) * block_size; ++ ++ ei.mtd = mtd; ++ ei.addr = addr; ++ ei.len = block_size; ++ ei.time = 1000; ++ ei.retries = 2; ++ ei.callback = NULL; ++ ei.priv = (u_long) dev; ++ ++ retval = mtd_erase(mtd, &ei); ++ ++ if (retval == 0) ++ return YAFFS_OK; ++ ++ return YAFFS_FAIL; ++} ++ ++static int yaffs_mtd_mark_bad(struct yaffs_dev *dev, int block_no) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ int blocksize = dev->param.chunks_per_block * dev->param.total_bytes_per_chunk; ++ int retval; ++ ++ yaffs_trace(YAFFS_TRACE_BAD_BLOCKS, "marking block %d bad", block_no); ++ ++ retval = mtd_block_markbad(mtd, (loff_t) blocksize * block_no); ++ return (retval) ? YAFFS_FAIL : YAFFS_OK; ++} ++ ++static int yaffs_mtd_check_bad(struct yaffs_dev *dev, int block_no) ++{ ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ int blocksize = dev->param.chunks_per_block * dev->param.total_bytes_per_chunk; ++ int retval; ++ ++ yaffs_trace(YAFFS_TRACE_MTD, "checking block %d bad", block_no); ++ ++ retval = mtd_block_isbad(mtd, (loff_t) blocksize * block_no); ++ return (retval) ? YAFFS_FAIL : YAFFS_OK; ++} ++ ++static int yaffs_mtd_initialise(struct yaffs_dev *dev) ++{ ++ return YAFFS_OK; ++} ++ ++static int yaffs_mtd_deinitialise(struct yaffs_dev *dev) ++{ ++ return YAFFS_OK; ++} ++ ++ ++void yaffs_mtd_drv_install(struct yaffs_dev *dev) ++{ ++ struct yaffs_driver *drv = &dev->drv; ++ ++ drv->drv_write_chunk_fn = yaffs_mtd_write; ++ drv->drv_read_chunk_fn = yaffs_mtd_read; ++ drv->drv_erase_fn = yaffs_mtd_erase; ++ drv->drv_mark_bad_fn = yaffs_mtd_mark_bad; ++ drv->drv_check_bad_fn = yaffs_mtd_check_bad; ++ drv->drv_initialise_fn = yaffs_mtd_initialise; ++ drv->drv_deinitialise_fn = yaffs_mtd_deinitialise; ++} ++ ++ ++struct mtd_info * yaffs_get_mtd_device(dev_t sdev) ++{ ++ struct mtd_info *mtd; ++ ++ mtd = yaffs_get_mtd_device(sdev); ++ ++ /* Check it's an mtd device..... */ ++ if (MAJOR(sdev) != MTD_BLOCK_MAJOR) ++ return NULL; /* This isn't an mtd device */ ++ ++ /* Check it's NAND */ ++ if (mtd->type != MTD_NANDFLASH) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: MTD device is not NAND it's type %d", ++ mtd->type); ++ return NULL; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, " %s %d", WRITE_SIZE_STR, WRITE_SIZE(mtd)); ++ yaffs_trace(YAFFS_TRACE_OS, " oobsize %d", mtd->oobsize); ++ yaffs_trace(YAFFS_TRACE_OS, " erasesize %d", mtd->erasesize); ++#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 29) ++ yaffs_trace(YAFFS_TRACE_OS, " size %u", mtd->size); ++#else ++ yaffs_trace(YAFFS_TRACE_OS, " size %lld", mtd->size); ++#endif ++ ++ return mtd; ++} ++ ++int yaffs_verify_mtd(struct mtd_info *mtd, int yaffs_version, int inband_tags) ++{ ++ if (yaffs_version == 2) { ++ if ((WRITE_SIZE(mtd) < YAFFS_MIN_YAFFS2_CHUNK_SIZE || ++ mtd->oobsize < YAFFS_MIN_YAFFS2_SPARE_SIZE) && ++ !inband_tags) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "MTD device does not have the right page sizes" ++ ); ++ return -1; ++ } ++ } else { ++ if (WRITE_SIZE(mtd) < YAFFS_BYTES_PER_CHUNK || ++ mtd->oobsize != YAFFS_BYTES_PER_SPARE) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "MTD device does not support have the right page sizes" ++ ); ++ return -1; ++ } ++ } ++ ++ return 0; ++} ++ ++ ++void yaffs_put_mtd_device(struct mtd_info *mtd) ++{ ++ if(mtd) ++ put_mtd_device(mtd); ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_mtdif.h linux-3.14.4/fs/yaffs2/yaffs_mtdif.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_mtdif.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_mtdif.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,25 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_MTDIF_H__ ++#define __YAFFS_MTDIF_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_mtd_drv_install(struct yaffs_dev *dev); ++struct mtd_info * yaffs_get_mtd_device(dev_t sdev); ++void yaffs_put_mtd_device(struct mtd_info *mtd); ++int yaffs_verify_mtd(struct mtd_info *mtd, int yaffs_version, int inband_tags); ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_nameval.c linux-3.14.4/fs/yaffs2/yaffs_nameval.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_nameval.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_nameval.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,208 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++/* ++ * This simple implementation of a name-value store assumes a small number of ++* values and fits into a small finite buffer. ++ * ++ * Each attribute is stored as a record: ++ * sizeof(int) bytes record size. ++ * strnlen+1 bytes name null terminated. ++ * nbytes value. ++ * ---------- ++ * total size stored in record size ++ * ++ * This code has not been tested with unicode yet. ++ */ ++ ++#include "yaffs_nameval.h" ++ ++#include "yportenv.h" ++ ++static int nval_find(const char *xb, int xb_size, const YCHAR *name, ++ int *exist_size) ++{ ++ int pos = 0; ++ int size; ++ ++ memcpy(&size, xb, sizeof(int)); ++ while (size > 0 && (size < xb_size) && (pos + size < xb_size)) { ++ if (!strncmp((YCHAR *) (xb + pos + sizeof(int)), ++ name, size)) { ++ if (exist_size) ++ *exist_size = size; ++ return pos; ++ } ++ pos += size; ++ if (pos < xb_size - sizeof(int)) ++ memcpy(&size, xb + pos, sizeof(int)); ++ else ++ size = 0; ++ } ++ if (exist_size) ++ *exist_size = 0; ++ return -ENODATA; ++} ++ ++static int nval_used(const char *xb, int xb_size) ++{ ++ int pos = 0; ++ int size; ++ ++ memcpy(&size, xb + pos, sizeof(int)); ++ while (size > 0 && (size < xb_size) && (pos + size < xb_size)) { ++ pos += size; ++ if (pos < xb_size - sizeof(int)) ++ memcpy(&size, xb + pos, sizeof(int)); ++ else ++ size = 0; ++ } ++ return pos; ++} ++ ++int nval_del(char *xb, int xb_size, const YCHAR *name) ++{ ++ int pos = nval_find(xb, xb_size, name, NULL); ++ int size; ++ ++ if (pos < 0 || pos >= xb_size) ++ return -ENODATA; ++ ++ /* Find size, shift rest over this record, ++ * then zero out the rest of buffer */ ++ memcpy(&size, xb + pos, sizeof(int)); ++ memcpy(xb + pos, xb + pos + size, xb_size - (pos + size)); ++ memset(xb + (xb_size - size), 0, size); ++ return 0; ++} ++ ++int nval_set(char *xb, int xb_size, const YCHAR *name, const char *buf, ++ int bsize, int flags) ++{ ++ int pos; ++ int namelen = strnlen(name, xb_size); ++ int reclen; ++ int size_exist = 0; ++ int space; ++ int start; ++ ++ pos = nval_find(xb, xb_size, name, &size_exist); ++ ++ if (flags & XATTR_CREATE && pos >= 0) ++ return -EEXIST; ++ if (flags & XATTR_REPLACE && pos < 0) ++ return -ENODATA; ++ ++ start = nval_used(xb, xb_size); ++ space = xb_size - start + size_exist; ++ ++ reclen = (sizeof(int) + namelen + 1 + bsize); ++ ++ if (reclen > space) ++ return -ENOSPC; ++ ++ if (pos >= 0) { ++ nval_del(xb, xb_size, name); ++ start = nval_used(xb, xb_size); ++ } ++ ++ pos = start; ++ ++ memcpy(xb + pos, &reclen, sizeof(int)); ++ pos += sizeof(int); ++ strncpy((YCHAR *) (xb + pos), name, reclen); ++ pos += (namelen + 1); ++ memcpy(xb + pos, buf, bsize); ++ return 0; ++} ++ ++int nval_get(const char *xb, int xb_size, const YCHAR * name, char *buf, ++ int bsize) ++{ ++ int pos = nval_find(xb, xb_size, name, NULL); ++ int size; ++ ++ if (pos >= 0 && pos < xb_size) { ++ ++ memcpy(&size, xb + pos, sizeof(int)); ++ pos += sizeof(int); /* advance past record length */ ++ size -= sizeof(int); ++ ++ /* Advance over name string */ ++ while (xb[pos] && size > 0 && pos < xb_size) { ++ pos++; ++ size--; ++ } ++ /*Advance over NUL */ ++ pos++; ++ size--; ++ ++ /* If bsize is zero then this is a size query. ++ * Return the size, but don't copy. ++ */ ++ if (!bsize) ++ return size; ++ ++ if (size <= bsize) { ++ memcpy(buf, xb + pos, size); ++ return size; ++ } ++ } ++ if (pos >= 0) ++ return -ERANGE; ++ ++ return -ENODATA; ++} ++ ++int nval_list(const char *xb, int xb_size, char *buf, int bsize) ++{ ++ int pos = 0; ++ int size; ++ int name_len; ++ int ncopied = 0; ++ int filled = 0; ++ ++ memcpy(&size, xb + pos, sizeof(int)); ++ while (size > sizeof(int) && ++ size <= xb_size && ++ (pos + size) < xb_size && ++ !filled) { ++ pos += sizeof(int); ++ size -= sizeof(int); ++ name_len = strnlen((YCHAR *) (xb + pos), size); ++ if (ncopied + name_len + 1 < bsize) { ++ memcpy(buf, xb + pos, name_len * sizeof(YCHAR)); ++ buf += name_len; ++ *buf = '\0'; ++ buf++; ++ if (sizeof(YCHAR) > 1) { ++ *buf = '\0'; ++ buf++; ++ } ++ ncopied += (name_len + 1); ++ } else { ++ filled = 1; ++ } ++ pos += size; ++ if (pos < xb_size - sizeof(int)) ++ memcpy(&size, xb + pos, sizeof(int)); ++ else ++ size = 0; ++ } ++ return ncopied; ++} ++ ++int nval_hasvalues(const char *xb, int xb_size) ++{ ++ return nval_used(xb, xb_size) > 0; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_nameval.h linux-3.14.4/fs/yaffs2/yaffs_nameval.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_nameval.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_nameval.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,28 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __NAMEVAL_H__ ++#define __NAMEVAL_H__ ++ ++#include "yportenv.h" ++ ++int nval_del(char *xb, int xb_size, const YCHAR * name); ++int nval_set(char *xb, int xb_size, const YCHAR * name, const char *buf, ++ int bsize, int flags); ++int nval_get(const char *xb, int xb_size, const YCHAR * name, char *buf, ++ int bsize); ++int nval_list(const char *xb, int xb_size, char *buf, int bsize); ++int nval_hasvalues(const char *xb, int xb_size); ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_nand.c linux-3.14.4/fs/yaffs2/yaffs_nand.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_nand.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_nand.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,122 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_nand.h" ++#include "yaffs_tagscompat.h" ++ ++#include "yaffs_getblockinfo.h" ++#include "yaffs_summary.h" ++ ++static int apply_chunk_offset(struct yaffs_dev *dev, int chunk) ++{ ++ return chunk - dev->chunk_offset; ++} ++ ++int yaffs_rd_chunk_tags_nand(struct yaffs_dev *dev, int nand_chunk, ++ u8 *buffer, struct yaffs_ext_tags *tags) ++{ ++ int result; ++ struct yaffs_ext_tags local_tags; ++ int flash_chunk = apply_chunk_offset(dev, nand_chunk); ++ ++ dev->n_page_reads++; ++ ++ /* If there are no tags provided use local tags. */ ++ if (!tags) ++ tags = &local_tags; ++ ++ result = dev->tagger.read_chunk_tags_fn(dev, flash_chunk, buffer, tags); ++ if (tags && tags->ecc_result > YAFFS_ECC_RESULT_NO_ERROR) { ++ ++ struct yaffs_block_info *bi; ++ bi = yaffs_get_block_info(dev, ++ nand_chunk / ++ dev->param.chunks_per_block); ++ yaffs_handle_chunk_error(dev, bi); ++ } ++ return result; ++} ++ ++int yaffs_wr_chunk_tags_nand(struct yaffs_dev *dev, ++ int nand_chunk, ++ const u8 *buffer, struct yaffs_ext_tags *tags) ++{ ++ int result; ++ int flash_chunk = apply_chunk_offset(dev, nand_chunk); ++ ++ dev->n_page_writes++; ++ ++ if (!tags) { ++ yaffs_trace(YAFFS_TRACE_ERROR, "Writing with no tags"); ++ BUG(); ++ return YAFFS_FAIL; ++ } ++ ++ tags->seq_number = dev->seq_number; ++ tags->chunk_used = 1; ++ yaffs_trace(YAFFS_TRACE_WRITE, ++ "Writing chunk %d tags %d %d", ++ nand_chunk, tags->obj_id, tags->chunk_id); ++ ++ result = dev->tagger.write_chunk_tags_fn(dev, flash_chunk, ++ buffer, tags); ++ ++ yaffs_summary_add(dev, tags, nand_chunk); ++ ++ return result; ++} ++ ++int yaffs_mark_bad(struct yaffs_dev *dev, int block_no) ++{ ++ block_no -= dev->block_offset; ++ dev->n_bad_markings++; ++ ++ if (dev->param.disable_bad_block_marking) ++ return YAFFS_OK; ++ ++ return dev->tagger.mark_bad_fn(dev, block_no); ++} ++ ++ ++int yaffs_query_init_block_state(struct yaffs_dev *dev, ++ int block_no, ++ enum yaffs_block_state *state, ++ u32 *seq_number) ++{ ++ block_no -= dev->block_offset; ++ return dev->tagger.query_block_fn(dev, block_no, state, seq_number); ++} ++ ++int yaffs_erase_block(struct yaffs_dev *dev, int block_no) ++{ ++ int result; ++ ++ block_no -= dev->block_offset; ++ dev->n_erasures++; ++ result = dev->drv.drv_erase_fn(dev, block_no); ++ return result; ++} ++ ++int yaffs_init_nand(struct yaffs_dev *dev) ++{ ++ if (dev->drv.drv_initialise_fn) ++ return dev->drv.drv_initialise_fn(dev); ++ return YAFFS_OK; ++} ++ ++int yaffs_deinit_nand(struct yaffs_dev *dev) ++{ ++ if (dev->drv.drv_deinitialise_fn) ++ return dev->drv.drv_deinitialise_fn(dev); ++ return YAFFS_OK; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_nand.h linux-3.14.4/fs/yaffs2/yaffs_nand.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_nand.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_nand.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,39 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_NAND_H__ ++#define __YAFFS_NAND_H__ ++#include "yaffs_guts.h" ++ ++int yaffs_rd_chunk_tags_nand(struct yaffs_dev *dev, int nand_chunk, ++ u8 *buffer, struct yaffs_ext_tags *tags); ++ ++int yaffs_wr_chunk_tags_nand(struct yaffs_dev *dev, ++ int nand_chunk, ++ const u8 *buffer, struct yaffs_ext_tags *tags); ++ ++int yaffs_mark_bad(struct yaffs_dev *dev, int block_no); ++ ++int yaffs_query_init_block_state(struct yaffs_dev *dev, ++ int block_no, ++ enum yaffs_block_state *state, ++ unsigned *seq_number); ++ ++int yaffs_erase_block(struct yaffs_dev *dev, int flash_block); ++ ++int yaffs_init_nand(struct yaffs_dev *dev); ++int yaffs_deinit_nand(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags1.c linux-3.14.4/fs/yaffs2/yaffs_packedtags1.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags1.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_packedtags1.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,56 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_packedtags1.h" ++#include "yportenv.h" ++ ++static const u8 all_ff[20] = { ++ 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff, ++ 0xff, 0xff, 0xff, 0xff ++}; ++ ++void yaffs_pack_tags1(struct yaffs_packed_tags1 *pt, ++ const struct yaffs_ext_tags *t) ++{ ++ pt->chunk_id = t->chunk_id; ++ pt->serial_number = t->serial_number; ++ pt->n_bytes = t->n_bytes; ++ pt->obj_id = t->obj_id; ++ pt->ecc = 0; ++ pt->deleted = (t->is_deleted) ? 0 : 1; ++ pt->unused_stuff = 0; ++ pt->should_be_ff = 0xffffffff; ++} ++ ++void yaffs_unpack_tags1(struct yaffs_ext_tags *t, ++ const struct yaffs_packed_tags1 *pt) ++{ ++ ++ if (memcmp(all_ff, pt, sizeof(struct yaffs_packed_tags1))) { ++ t->block_bad = 0; ++ if (pt->should_be_ff != 0xffffffff) ++ t->block_bad = 1; ++ t->chunk_used = 1; ++ t->obj_id = pt->obj_id; ++ t->chunk_id = pt->chunk_id; ++ t->n_bytes = pt->n_bytes; ++ t->ecc_result = YAFFS_ECC_RESULT_NO_ERROR; ++ t->is_deleted = (pt->deleted) ? 0 : 1; ++ t->serial_number = pt->serial_number; ++ } else { ++ memset(t, 0, sizeof(struct yaffs_ext_tags)); ++ } ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags1.h linux-3.14.4/fs/yaffs2/yaffs_packedtags1.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags1.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_packedtags1.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,39 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++/* This is used to pack YAFFS1 tags, not YAFFS2 tags. */ ++ ++#ifndef __YAFFS_PACKEDTAGS1_H__ ++#define __YAFFS_PACKEDTAGS1_H__ ++ ++#include "yaffs_guts.h" ++ ++struct yaffs_packed_tags1 { ++ u32 chunk_id:20; ++ u32 serial_number:2; ++ u32 n_bytes:10; ++ u32 obj_id:18; ++ u32 ecc:12; ++ u32 deleted:1; ++ u32 unused_stuff:1; ++ unsigned should_be_ff; ++ ++}; ++ ++void yaffs_pack_tags1(struct yaffs_packed_tags1 *pt, ++ const struct yaffs_ext_tags *t); ++void yaffs_unpack_tags1(struct yaffs_ext_tags *t, ++ const struct yaffs_packed_tags1 *pt); ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags2.c linux-3.14.4/fs/yaffs2/yaffs_packedtags2.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags2.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_packedtags2.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,197 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_packedtags2.h" ++#include "yportenv.h" ++#include "yaffs_trace.h" ++ ++/* This code packs a set of extended tags into a binary structure for ++ * NAND storage ++ */ ++ ++/* Some of the information is "extra" struff which can be packed in to ++ * speed scanning ++ * This is defined by having the EXTRA_HEADER_INFO_FLAG set. ++ */ ++ ++/* Extra flags applied to chunk_id */ ++ ++#define EXTRA_HEADER_INFO_FLAG 0x80000000 ++#define EXTRA_SHRINK_FLAG 0x40000000 ++#define EXTRA_SHADOWS_FLAG 0x20000000 ++#define EXTRA_SPARE_FLAGS 0x10000000 ++ ++#define ALL_EXTRA_FLAGS 0xf0000000 ++ ++/* Also, the top 4 bits of the object Id are set to the object type. */ ++#define EXTRA_OBJECT_TYPE_SHIFT (28) ++#define EXTRA_OBJECT_TYPE_MASK ((0x0f) << EXTRA_OBJECT_TYPE_SHIFT) ++ ++static void yaffs_dump_packed_tags2_tags_only( ++ const struct yaffs_packed_tags2_tags_only *ptt) ++{ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "packed tags obj %d chunk %d byte %d seq %d", ++ ptt->obj_id, ptt->chunk_id, ptt->n_bytes, ptt->seq_number); ++} ++ ++static void yaffs_dump_packed_tags2(const struct yaffs_packed_tags2 *pt) ++{ ++ yaffs_dump_packed_tags2_tags_only(&pt->t); ++} ++ ++static void yaffs_dump_tags2(const struct yaffs_ext_tags *t) ++{ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "ext.tags eccres %d blkbad %d chused %d obj %d chunk%d byte %d del %d ser %d seq %d", ++ t->ecc_result, t->block_bad, t->chunk_used, t->obj_id, ++ t->chunk_id, t->n_bytes, t->is_deleted, t->serial_number, ++ t->seq_number); ++ ++} ++ ++static int yaffs_check_tags_extra_packable(const struct yaffs_ext_tags *t) ++{ ++ if (t->chunk_id != 0 || !t->extra_available) ++ return 0; ++ ++ /* Check if the file size is too long to store */ ++ if (t->extra_obj_type == YAFFS_OBJECT_TYPE_FILE && ++ (t->extra_file_size >> 31) != 0) ++ return 0; ++ return 1; ++} ++ ++void yaffs_pack_tags2_tags_only(struct yaffs_packed_tags2_tags_only *ptt, ++ const struct yaffs_ext_tags *t) ++{ ++ ptt->chunk_id = t->chunk_id; ++ ptt->seq_number = t->seq_number; ++ ptt->n_bytes = t->n_bytes; ++ ptt->obj_id = t->obj_id; ++ ++ /* Only store extra tags for object headers. ++ * If it is a file then only store if the file size is short\ ++ * enough to fit. ++ */ ++ if (yaffs_check_tags_extra_packable(t)) { ++ /* Store the extra header info instead */ ++ /* We save the parent object in the chunk_id */ ++ ptt->chunk_id = EXTRA_HEADER_INFO_FLAG | t->extra_parent_id; ++ if (t->extra_is_shrink) ++ ptt->chunk_id |= EXTRA_SHRINK_FLAG; ++ if (t->extra_shadows) ++ ptt->chunk_id |= EXTRA_SHADOWS_FLAG; ++ ++ ptt->obj_id &= ~EXTRA_OBJECT_TYPE_MASK; ++ ptt->obj_id |= (t->extra_obj_type << EXTRA_OBJECT_TYPE_SHIFT); ++ ++ if (t->extra_obj_type == YAFFS_OBJECT_TYPE_HARDLINK) ++ ptt->n_bytes = t->extra_equiv_id; ++ else if (t->extra_obj_type == YAFFS_OBJECT_TYPE_FILE) ++ ptt->n_bytes = (unsigned) t->extra_file_size; ++ else ++ ptt->n_bytes = 0; ++ } ++ ++ yaffs_dump_packed_tags2_tags_only(ptt); ++ yaffs_dump_tags2(t); ++} ++ ++void yaffs_pack_tags2(struct yaffs_packed_tags2 *pt, ++ const struct yaffs_ext_tags *t, int tags_ecc) ++{ ++ yaffs_pack_tags2_tags_only(&pt->t, t); ++ ++ if (tags_ecc) ++ yaffs_ecc_calc_other((unsigned char *)&pt->t, ++ sizeof(struct yaffs_packed_tags2_tags_only), ++ &pt->ecc); ++} ++ ++void yaffs_unpack_tags2_tags_only(struct yaffs_ext_tags *t, ++ struct yaffs_packed_tags2_tags_only *ptt) ++{ ++ memset(t, 0, sizeof(struct yaffs_ext_tags)); ++ ++ if (ptt->seq_number == 0xffffffff) ++ return; ++ ++ t->block_bad = 0; ++ t->chunk_used = 1; ++ t->obj_id = ptt->obj_id; ++ t->chunk_id = ptt->chunk_id; ++ t->n_bytes = ptt->n_bytes; ++ t->is_deleted = 0; ++ t->serial_number = 0; ++ t->seq_number = ptt->seq_number; ++ ++ /* Do extra header info stuff */ ++ if (ptt->chunk_id & EXTRA_HEADER_INFO_FLAG) { ++ t->chunk_id = 0; ++ t->n_bytes = 0; ++ ++ t->extra_available = 1; ++ t->extra_parent_id = ptt->chunk_id & (~(ALL_EXTRA_FLAGS)); ++ t->extra_is_shrink = ptt->chunk_id & EXTRA_SHRINK_FLAG ? 1 : 0; ++ t->extra_shadows = ptt->chunk_id & EXTRA_SHADOWS_FLAG ? 1 : 0; ++ t->extra_obj_type = ptt->obj_id >> EXTRA_OBJECT_TYPE_SHIFT; ++ t->obj_id &= ~EXTRA_OBJECT_TYPE_MASK; ++ ++ if (t->extra_obj_type == YAFFS_OBJECT_TYPE_HARDLINK) ++ t->extra_equiv_id = ptt->n_bytes; ++ else ++ t->extra_file_size = ptt->n_bytes; ++ } ++ yaffs_dump_packed_tags2_tags_only(ptt); ++ yaffs_dump_tags2(t); ++} ++ ++void yaffs_unpack_tags2(struct yaffs_ext_tags *t, struct yaffs_packed_tags2 *pt, ++ int tags_ecc) ++{ ++ enum yaffs_ecc_result ecc_result = YAFFS_ECC_RESULT_NO_ERROR; ++ ++ if (pt->t.seq_number != 0xffffffff && tags_ecc) { ++ /* Chunk is in use and we need to do ECC */ ++ ++ struct yaffs_ecc_other ecc; ++ int result; ++ yaffs_ecc_calc_other((unsigned char *)&pt->t, ++ sizeof(struct yaffs_packed_tags2_tags_only), ++ &ecc); ++ result = ++ yaffs_ecc_correct_other((unsigned char *)&pt->t, ++ sizeof(struct yaffs_packed_tags2_tags_only), ++ &pt->ecc, &ecc); ++ switch (result) { ++ case 0: ++ ecc_result = YAFFS_ECC_RESULT_NO_ERROR; ++ break; ++ case 1: ++ ecc_result = YAFFS_ECC_RESULT_FIXED; ++ break; ++ case -1: ++ ecc_result = YAFFS_ECC_RESULT_UNFIXED; ++ break; ++ default: ++ ecc_result = YAFFS_ECC_RESULT_UNKNOWN; ++ } ++ } ++ yaffs_unpack_tags2_tags_only(t, &pt->t); ++ ++ t->ecc_result = ecc_result; ++ ++ yaffs_dump_packed_tags2(pt); ++ yaffs_dump_tags2(t); ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags2.h linux-3.14.4/fs/yaffs2/yaffs_packedtags2.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_packedtags2.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_packedtags2.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,47 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++/* This is used to pack YAFFS2 tags, not YAFFS1tags. */ ++ ++#ifndef __YAFFS_PACKEDTAGS2_H__ ++#define __YAFFS_PACKEDTAGS2_H__ ++ ++#include "yaffs_guts.h" ++#include "yaffs_ecc.h" ++ ++struct yaffs_packed_tags2_tags_only { ++ unsigned seq_number; ++ unsigned obj_id; ++ unsigned chunk_id; ++ unsigned n_bytes; ++}; ++ ++struct yaffs_packed_tags2 { ++ struct yaffs_packed_tags2_tags_only t; ++ struct yaffs_ecc_other ecc; ++}; ++ ++/* Full packed tags with ECC, used for oob tags */ ++void yaffs_pack_tags2(struct yaffs_packed_tags2 *pt, ++ const struct yaffs_ext_tags *t, int tags_ecc); ++void yaffs_unpack_tags2(struct yaffs_ext_tags *t, struct yaffs_packed_tags2 *pt, ++ int tags_ecc); ++ ++/* Only the tags part (no ECC for use with inband tags */ ++void yaffs_pack_tags2_tags_only(struct yaffs_packed_tags2_tags_only *pt, ++ const struct yaffs_ext_tags *t); ++void yaffs_unpack_tags2_tags_only(struct yaffs_ext_tags *t, ++ struct yaffs_packed_tags2_tags_only *pt); ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_summary.c linux-3.14.4/fs/yaffs2/yaffs_summary.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_summary.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_summary.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,312 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++/* Summaries write the useful part of the tags for the chunks in a block into an ++ * an array which is written to the last n chunks of the block. ++ * Reading the summaries gives all the tags for the block in one read. Much ++ * faster. ++ * ++ * Chunks holding summaries are marked with tags making it look like ++ * they are part of a fake file. ++ * ++ * The summary could also be used during gc. ++ * ++ */ ++ ++#include "yaffs_summary.h" ++#include "yaffs_packedtags2.h" ++#include "yaffs_nand.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_bitmap.h" ++ ++/* ++ * The summary is built up in an array of summary tags. ++ * This gets written to the last one or two (maybe more) chunks in a block. ++ * A summary header is written as the first part of each chunk of summary data. ++ * The summary header must match or the summary is rejected. ++ */ ++ ++/* Summary tags don't need the sequence number because that is redundant. */ ++struct yaffs_summary_tags { ++ unsigned obj_id; ++ unsigned chunk_id; ++ unsigned n_bytes; ++}; ++ ++/* Summary header */ ++struct yaffs_summary_header { ++ unsigned version; /* Must match current version */ ++ unsigned block; /* Must be this block */ ++ unsigned seq; /* Must be this sequence number */ ++ unsigned sum; /* Just add up all the bytes in the tags */ ++}; ++ ++ ++static void yaffs_summary_clear(struct yaffs_dev *dev) ++{ ++ if (!dev->sum_tags) ++ return; ++ memset(dev->sum_tags, 0, dev->chunks_per_summary * ++ sizeof(struct yaffs_summary_tags)); ++} ++ ++ ++void yaffs_summary_deinit(struct yaffs_dev *dev) ++{ ++ kfree(dev->sum_tags); ++ dev->sum_tags = NULL; ++ kfree(dev->gc_sum_tags); ++ dev->gc_sum_tags = NULL; ++ dev->chunks_per_summary = 0; ++} ++ ++int yaffs_summary_init(struct yaffs_dev *dev) ++{ ++ int sum_bytes; ++ int chunks_used; /* Number of chunks used by summary */ ++ int sum_tags_bytes; ++ ++ sum_bytes = dev->param.chunks_per_block * ++ sizeof(struct yaffs_summary_tags); ++ ++ chunks_used = (sum_bytes + dev->data_bytes_per_chunk - 1)/ ++ (dev->data_bytes_per_chunk - ++ sizeof(struct yaffs_summary_header)); ++ ++ dev->chunks_per_summary = dev->param.chunks_per_block - chunks_used; ++ sum_tags_bytes = sizeof(struct yaffs_summary_tags) * ++ dev->chunks_per_summary; ++ dev->sum_tags = kmalloc(sum_tags_bytes, GFP_NOFS); ++ dev->gc_sum_tags = kmalloc(sum_tags_bytes, GFP_NOFS); ++ if (!dev->sum_tags || !dev->gc_sum_tags) { ++ yaffs_summary_deinit(dev); ++ return YAFFS_FAIL; ++ } ++ ++ yaffs_summary_clear(dev); ++ ++ return YAFFS_OK; ++} ++ ++static unsigned yaffs_summary_sum(struct yaffs_dev *dev) ++{ ++ u8 *sum_buffer = (u8 *)dev->sum_tags; ++ int i; ++ unsigned sum = 0; ++ ++ i = sizeof(struct yaffs_summary_tags) * ++ dev->chunks_per_summary; ++ while (i > 0) { ++ sum += *sum_buffer; ++ sum_buffer++; ++ i--; ++ } ++ ++ return sum; ++} ++ ++static int yaffs_summary_write(struct yaffs_dev *dev, int blk) ++{ ++ struct yaffs_ext_tags tags; ++ u8 *buffer; ++ u8 *sum_buffer = (u8 *)dev->sum_tags; ++ int n_bytes; ++ int chunk_in_nand; ++ int chunk_in_block; ++ int result; ++ int this_tx; ++ struct yaffs_summary_header hdr; ++ int sum_bytes_per_chunk = dev->data_bytes_per_chunk - sizeof(hdr); ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, blk); ++ ++ buffer = yaffs_get_temp_buffer(dev); ++ n_bytes = sizeof(struct yaffs_summary_tags) * ++ dev->chunks_per_summary; ++ memset(&tags, 0, sizeof(struct yaffs_ext_tags)); ++ tags.obj_id = YAFFS_OBJECTID_SUMMARY; ++ tags.chunk_id = 1; ++ chunk_in_block = dev->chunks_per_summary; ++ chunk_in_nand = dev->alloc_block * dev->param.chunks_per_block + ++ dev->chunks_per_summary; ++ hdr.version = YAFFS_SUMMARY_VERSION; ++ hdr.block = blk; ++ hdr.seq = bi->seq_number; ++ hdr.sum = yaffs_summary_sum(dev); ++ ++ do { ++ this_tx = n_bytes; ++ if (this_tx > sum_bytes_per_chunk) ++ this_tx = sum_bytes_per_chunk; ++ memcpy(buffer, &hdr, sizeof(hdr)); ++ memcpy(buffer + sizeof(hdr), sum_buffer, this_tx); ++ tags.n_bytes = this_tx + sizeof(hdr); ++ result = yaffs_wr_chunk_tags_nand(dev, chunk_in_nand, ++ buffer, &tags); ++ ++ if (result != YAFFS_OK) ++ break; ++ yaffs_set_chunk_bit(dev, blk, chunk_in_block); ++ bi->pages_in_use++; ++ dev->n_free_chunks--; ++ ++ n_bytes -= this_tx; ++ sum_buffer += this_tx; ++ chunk_in_nand++; ++ chunk_in_block++; ++ tags.chunk_id++; ++ } while (result == YAFFS_OK && n_bytes > 0); ++ yaffs_release_temp_buffer(dev, buffer); ++ ++ ++ if (result == YAFFS_OK) ++ bi->has_summary = 1; ++ ++ ++ return result; ++} ++ ++int yaffs_summary_read(struct yaffs_dev *dev, ++ struct yaffs_summary_tags *st, ++ int blk) ++{ ++ struct yaffs_ext_tags tags; ++ u8 *buffer; ++ u8 *sum_buffer = (u8 *)st; ++ int n_bytes; ++ int chunk_id; ++ int chunk_in_nand; ++ int chunk_in_block; ++ int result; ++ int this_tx; ++ struct yaffs_summary_header hdr; ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, blk); ++ int sum_bytes_per_chunk = dev->data_bytes_per_chunk - sizeof(hdr); ++ int sum_tags_bytes; ++ ++ sum_tags_bytes = sizeof(struct yaffs_summary_tags) * ++ dev->chunks_per_summary; ++ buffer = yaffs_get_temp_buffer(dev); ++ n_bytes = sizeof(struct yaffs_summary_tags) * dev->chunks_per_summary; ++ chunk_in_block = dev->chunks_per_summary; ++ chunk_in_nand = blk * dev->param.chunks_per_block + ++ dev->chunks_per_summary; ++ chunk_id = 1; ++ do { ++ this_tx = n_bytes; ++ if (this_tx > sum_bytes_per_chunk) ++ this_tx = sum_bytes_per_chunk; ++ result = yaffs_rd_chunk_tags_nand(dev, chunk_in_nand, ++ buffer, &tags); ++ ++ if (tags.chunk_id != chunk_id || ++ tags.obj_id != YAFFS_OBJECTID_SUMMARY || ++ tags.chunk_used == 0 || ++ tags.ecc_result > YAFFS_ECC_RESULT_FIXED || ++ tags.n_bytes != (this_tx + sizeof(hdr))) ++ result = YAFFS_FAIL; ++ if (result != YAFFS_OK) ++ break; ++ ++ if (st == dev->sum_tags) { ++ /* If we're scanning then update the block info */ ++ yaffs_set_chunk_bit(dev, blk, chunk_in_block); ++ bi->pages_in_use++; ++ } ++ memcpy(&hdr, buffer, sizeof(hdr)); ++ memcpy(sum_buffer, buffer + sizeof(hdr), this_tx); ++ n_bytes -= this_tx; ++ sum_buffer += this_tx; ++ chunk_in_nand++; ++ chunk_in_block++; ++ chunk_id++; ++ } while (result == YAFFS_OK && n_bytes > 0); ++ yaffs_release_temp_buffer(dev, buffer); ++ ++ if (result == YAFFS_OK) { ++ /* Verify header */ ++ if (hdr.version != YAFFS_SUMMARY_VERSION || ++ hdr.seq != bi->seq_number || ++ hdr.sum != yaffs_summary_sum(dev)) ++ result = YAFFS_FAIL; ++ } ++ ++ if (st == dev->sum_tags && result == YAFFS_OK) ++ bi->has_summary = 1; ++ ++ return result; ++} ++ ++int yaffs_summary_add(struct yaffs_dev *dev, ++ struct yaffs_ext_tags *tags, ++ int chunk_in_nand) ++{ ++ struct yaffs_packed_tags2_tags_only tags_only; ++ struct yaffs_summary_tags *sum_tags; ++ int block_in_nand = chunk_in_nand / dev->param.chunks_per_block; ++ int chunk_in_block = chunk_in_nand % dev->param.chunks_per_block; ++ ++ if (!dev->sum_tags) ++ return YAFFS_OK; ++ ++ if (chunk_in_block >= 0 && chunk_in_block < dev->chunks_per_summary) { ++ yaffs_pack_tags2_tags_only(&tags_only, tags); ++ sum_tags = &dev->sum_tags[chunk_in_block]; ++ sum_tags->chunk_id = tags_only.chunk_id; ++ sum_tags->n_bytes = tags_only.n_bytes; ++ sum_tags->obj_id = tags_only.obj_id; ++ ++ if (chunk_in_block == dev->chunks_per_summary - 1) { ++ /* Time to write out the summary */ ++ yaffs_summary_write(dev, block_in_nand); ++ yaffs_summary_clear(dev); ++ yaffs_skip_rest_of_block(dev); ++ } ++ } ++ return YAFFS_OK; ++} ++ ++int yaffs_summary_fetch(struct yaffs_dev *dev, ++ struct yaffs_ext_tags *tags, ++ int chunk_in_block) ++{ ++ struct yaffs_packed_tags2_tags_only tags_only; ++ struct yaffs_summary_tags *sum_tags; ++ if (chunk_in_block >= 0 && chunk_in_block < dev->chunks_per_summary) { ++ sum_tags = &dev->sum_tags[chunk_in_block]; ++ tags_only.chunk_id = sum_tags->chunk_id; ++ tags_only.n_bytes = sum_tags->n_bytes; ++ tags_only.obj_id = sum_tags->obj_id; ++ yaffs_unpack_tags2_tags_only(tags, &tags_only); ++ return YAFFS_OK; ++ } ++ return YAFFS_FAIL; ++} ++ ++void yaffs_summary_gc(struct yaffs_dev *dev, int blk) ++{ ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, blk); ++ int i; ++ ++ if (!bi->has_summary) ++ return; ++ ++ for (i = dev->chunks_per_summary; ++ i < dev->param.chunks_per_block; ++ i++) { ++ if (yaffs_check_chunk_bit(dev, blk, i)) { ++ yaffs_clear_chunk_bit(dev, blk, i); ++ bi->pages_in_use--; ++ dev->n_free_chunks++; ++ } ++ } ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_summary.h linux-3.14.4/fs/yaffs2/yaffs_summary.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_summary.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_summary.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,37 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_SUMMARY_H__ ++#define __YAFFS_SUMMARY_H__ ++ ++#include "yaffs_packedtags2.h" ++ ++ ++int yaffs_summary_init(struct yaffs_dev *dev); ++void yaffs_summary_deinit(struct yaffs_dev *dev); ++ ++int yaffs_summary_add(struct yaffs_dev *dev, ++ struct yaffs_ext_tags *tags, ++ int chunk_in_block); ++int yaffs_summary_fetch(struct yaffs_dev *dev, ++ struct yaffs_ext_tags *tags, ++ int chunk_in_block); ++int yaffs_summary_read(struct yaffs_dev *dev, ++ struct yaffs_summary_tags *st, ++ int blk); ++void yaffs_summary_gc(struct yaffs_dev *dev, int blk); ++ ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_tagscompat.c linux-3.14.4/fs/yaffs2/yaffs_tagscompat.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_tagscompat.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_tagscompat.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,381 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_guts.h" ++#include "yaffs_tagscompat.h" ++#include "yaffs_ecc.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_trace.h" ++ ++static void yaffs_handle_rd_data_error(struct yaffs_dev *dev, int nand_chunk); ++ ++ ++/********** Tags ECC calculations *********/ ++ ++ ++void yaffs_calc_tags_ecc(struct yaffs_tags *tags) ++{ ++ /* Calculate an ecc */ ++ unsigned char *b = ((union yaffs_tags_union *)tags)->as_bytes; ++ unsigned i, j; ++ unsigned ecc = 0; ++ unsigned bit = 0; ++ ++ tags->ecc = 0; ++ ++ for (i = 0; i < 8; i++) { ++ for (j = 1; j & 0xff; j <<= 1) { ++ bit++; ++ if (b[i] & j) ++ ecc ^= bit; ++ } ++ } ++ tags->ecc = ecc; ++} ++ ++int yaffs_check_tags_ecc(struct yaffs_tags *tags) ++{ ++ unsigned ecc = tags->ecc; ++ ++ yaffs_calc_tags_ecc(tags); ++ ++ ecc ^= tags->ecc; ++ ++ if (ecc && ecc <= 64) { ++ /* TODO: Handle the failure better. Retire? */ ++ unsigned char *b = ((union yaffs_tags_union *)tags)->as_bytes; ++ ++ ecc--; ++ ++ b[ecc / 8] ^= (1 << (ecc & 7)); ++ ++ /* Now recvalc the ecc */ ++ yaffs_calc_tags_ecc(tags); ++ ++ return 1; /* recovered error */ ++ } else if (ecc) { ++ /* Wierd ecc failure value */ ++ /* TODO Need to do somethiong here */ ++ return -1; /* unrecovered error */ ++ } ++ return 0; ++} ++ ++/********** Tags **********/ ++ ++static void yaffs_load_tags_to_spare(struct yaffs_spare *spare_ptr, ++ struct yaffs_tags *tags_ptr) ++{ ++ union yaffs_tags_union *tu = (union yaffs_tags_union *)tags_ptr; ++ ++ yaffs_calc_tags_ecc(tags_ptr); ++ ++ spare_ptr->tb0 = tu->as_bytes[0]; ++ spare_ptr->tb1 = tu->as_bytes[1]; ++ spare_ptr->tb2 = tu->as_bytes[2]; ++ spare_ptr->tb3 = tu->as_bytes[3]; ++ spare_ptr->tb4 = tu->as_bytes[4]; ++ spare_ptr->tb5 = tu->as_bytes[5]; ++ spare_ptr->tb6 = tu->as_bytes[6]; ++ spare_ptr->tb7 = tu->as_bytes[7]; ++} ++ ++static void yaffs_get_tags_from_spare(struct yaffs_dev *dev, ++ struct yaffs_spare *spare_ptr, ++ struct yaffs_tags *tags_ptr) ++{ ++ union yaffs_tags_union *tu = (union yaffs_tags_union *)tags_ptr; ++ int result; ++ ++ tu->as_bytes[0] = spare_ptr->tb0; ++ tu->as_bytes[1] = spare_ptr->tb1; ++ tu->as_bytes[2] = spare_ptr->tb2; ++ tu->as_bytes[3] = spare_ptr->tb3; ++ tu->as_bytes[4] = spare_ptr->tb4; ++ tu->as_bytes[5] = spare_ptr->tb5; ++ tu->as_bytes[6] = spare_ptr->tb6; ++ tu->as_bytes[7] = spare_ptr->tb7; ++ ++ result = yaffs_check_tags_ecc(tags_ptr); ++ if (result > 0) ++ dev->n_tags_ecc_fixed++; ++ else if (result < 0) ++ dev->n_tags_ecc_unfixed++; ++} ++ ++static void yaffs_spare_init(struct yaffs_spare *spare) ++{ ++ memset(spare, 0xff, sizeof(struct yaffs_spare)); ++} ++ ++static int yaffs_wr_nand(struct yaffs_dev *dev, ++ int nand_chunk, const u8 *data, ++ struct yaffs_spare *spare) ++{ ++ int data_size = dev->data_bytes_per_chunk; ++ ++ return dev->drv.drv_write_chunk_fn(dev, nand_chunk, ++ data, data_size, ++ (u8 *) spare, sizeof(*spare)); ++} ++ ++static int yaffs_rd_chunk_nand(struct yaffs_dev *dev, ++ int nand_chunk, ++ u8 *data, ++ struct yaffs_spare *spare, ++ enum yaffs_ecc_result *ecc_result, ++ int correct_errors) ++{ ++ int ret_val; ++ struct yaffs_spare local_spare; ++ int data_size; ++ int spare_size; ++ int ecc_result1, ecc_result2; ++ u8 calc_ecc[3]; ++ ++ if (!spare) { ++ /* If we don't have a real spare, then we use a local one. */ ++ /* Need this for the calculation of the ecc */ ++ spare = &local_spare; ++ } ++ data_size = dev->data_bytes_per_chunk; ++ spare_size = sizeof(struct yaffs_spare); ++ ++ if (dev->param.use_nand_ecc) ++ return dev->drv.drv_read_chunk_fn(dev, nand_chunk, ++ data, data_size, ++ (u8 *) spare, spare_size, ++ ecc_result); ++ ++ ++ /* Handle the ECC at this level. */ ++ ++ ret_val = dev->drv.drv_read_chunk_fn(dev, nand_chunk, ++ data, data_size, ++ (u8 *)spare, spare_size, ++ NULL); ++ if (!data || !correct_errors) ++ return ret_val; ++ ++ /* Do ECC correction if needed. */ ++ yaffs_ecc_calc(data, calc_ecc); ++ ecc_result1 = yaffs_ecc_correct(data, spare->ecc1, calc_ecc); ++ yaffs_ecc_calc(&data[256], calc_ecc); ++ ecc_result2 = yaffs_ecc_correct(&data[256], spare->ecc2, calc_ecc); ++ ++ if (ecc_result1 > 0) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>>yaffs ecc error fix performed on chunk %d:0", ++ nand_chunk); ++ dev->n_ecc_fixed++; ++ } else if (ecc_result1 < 0) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>>yaffs ecc error unfixed on chunk %d:0", ++ nand_chunk); ++ dev->n_ecc_unfixed++; ++ } ++ ++ if (ecc_result2 > 0) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>>yaffs ecc error fix performed on chunk %d:1", ++ nand_chunk); ++ dev->n_ecc_fixed++; ++ } else if (ecc_result2 < 0) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "**>>yaffs ecc error unfixed on chunk %d:1", ++ nand_chunk); ++ dev->n_ecc_unfixed++; ++ } ++ ++ if (ecc_result1 || ecc_result2) { ++ /* We had a data problem on this page */ ++ yaffs_handle_rd_data_error(dev, nand_chunk); ++ } ++ ++ if (ecc_result1 < 0 || ecc_result2 < 0) ++ *ecc_result = YAFFS_ECC_RESULT_UNFIXED; ++ else if (ecc_result1 > 0 || ecc_result2 > 0) ++ *ecc_result = YAFFS_ECC_RESULT_FIXED; ++ else ++ *ecc_result = YAFFS_ECC_RESULT_NO_ERROR; ++ ++ return ret_val; ++} ++ ++/* ++ * Functions for robustisizing ++ */ ++ ++static void yaffs_handle_rd_data_error(struct yaffs_dev *dev, int nand_chunk) ++{ ++ int flash_block = nand_chunk / dev->param.chunks_per_block; ++ ++ /* Mark the block for retirement */ ++ yaffs_get_block_info(dev, flash_block + dev->block_offset)-> ++ needs_retiring = 1; ++ yaffs_trace(YAFFS_TRACE_ERROR | YAFFS_TRACE_BAD_BLOCKS, ++ "**>>Block %d marked for retirement", ++ flash_block); ++ ++ /* TODO: ++ * Just do a garbage collection on the affected block ++ * then retire the block ++ * NB recursion ++ */ ++} ++ ++static int yaffs_tags_compat_wr(struct yaffs_dev *dev, ++ int nand_chunk, ++ const u8 *data, const struct yaffs_ext_tags *ext_tags) ++{ ++ struct yaffs_spare spare; ++ struct yaffs_tags tags; ++ ++ yaffs_spare_init(&spare); ++ ++ if (ext_tags->is_deleted) ++ spare.page_status = 0; ++ else { ++ tags.obj_id = ext_tags->obj_id; ++ tags.chunk_id = ext_tags->chunk_id; ++ ++ tags.n_bytes_lsb = ext_tags->n_bytes & (1024 - 1); ++ ++ if (dev->data_bytes_per_chunk >= 1024) ++ tags.n_bytes_msb = (ext_tags->n_bytes >> 10) & 3; ++ else ++ tags.n_bytes_msb = 3; ++ ++ tags.serial_number = ext_tags->serial_number; ++ ++ if (!dev->param.use_nand_ecc && data) { ++ yaffs_ecc_calc(data, spare.ecc1); ++ yaffs_ecc_calc(&data[256], spare.ecc2); ++ } ++ ++ yaffs_load_tags_to_spare(&spare, &tags); ++ } ++ return yaffs_wr_nand(dev, nand_chunk, data, &spare); ++} ++ ++static int yaffs_tags_compat_rd(struct yaffs_dev *dev, ++ int nand_chunk, ++ u8 *data, struct yaffs_ext_tags *ext_tags) ++{ ++ struct yaffs_spare spare; ++ struct yaffs_tags tags; ++ enum yaffs_ecc_result ecc_result = YAFFS_ECC_RESULT_UNKNOWN; ++ static struct yaffs_spare spare_ff; ++ static int init; ++ int deleted; ++ ++ if (!init) { ++ memset(&spare_ff, 0xff, sizeof(spare_ff)); ++ init = 1; ++ } ++ ++ if (!yaffs_rd_chunk_nand(dev, nand_chunk, ++ data, &spare, &ecc_result, 1)) ++ return YAFFS_FAIL; ++ ++ /* ext_tags may be NULL */ ++ if (!ext_tags) ++ return YAFFS_OK; ++ ++ deleted = (hweight8(spare.page_status) < 7) ? 1 : 0; ++ ++ ext_tags->is_deleted = deleted; ++ ext_tags->ecc_result = ecc_result; ++ ext_tags->block_bad = 0; /* We're reading it */ ++ /* therefore it is not a bad block */ ++ ext_tags->chunk_used = ++ memcmp(&spare_ff, &spare, sizeof(spare_ff)) ? 1 : 0; ++ ++ if (ext_tags->chunk_used) { ++ yaffs_get_tags_from_spare(dev, &spare, &tags); ++ ext_tags->obj_id = tags.obj_id; ++ ext_tags->chunk_id = tags.chunk_id; ++ ext_tags->n_bytes = tags.n_bytes_lsb; ++ ++ if (dev->data_bytes_per_chunk >= 1024) ++ ext_tags->n_bytes |= ++ (((unsigned)tags.n_bytes_msb) << 10); ++ ++ ext_tags->serial_number = tags.serial_number; ++ } ++ ++ return YAFFS_OK; ++} ++ ++static int yaffs_tags_compat_mark_bad(struct yaffs_dev *dev, int flash_block) ++{ ++ struct yaffs_spare spare; ++ ++ memset(&spare, 0xff, sizeof(struct yaffs_spare)); ++ ++ spare.block_status = 'Y'; ++ ++ yaffs_wr_nand(dev, flash_block * dev->param.chunks_per_block, NULL, ++ &spare); ++ yaffs_wr_nand(dev, flash_block * dev->param.chunks_per_block + 1, ++ NULL, &spare); ++ ++ return YAFFS_OK; ++} ++ ++static int yaffs_tags_compat_query_block(struct yaffs_dev *dev, ++ int block_no, ++ enum yaffs_block_state *state, ++ u32 *seq_number) ++{ ++ struct yaffs_spare spare0, spare1; ++ static struct yaffs_spare spare_ff; ++ static int init; ++ enum yaffs_ecc_result dummy; ++ ++ if (!init) { ++ memset(&spare_ff, 0xff, sizeof(spare_ff)); ++ init = 1; ++ } ++ ++ *seq_number = 0; ++ ++ /* Look for bad block markers in the first two chunks */ ++ yaffs_rd_chunk_nand(dev, block_no * dev->param.chunks_per_block, ++ NULL, &spare0, &dummy, 0); ++ yaffs_rd_chunk_nand(dev, block_no * dev->param.chunks_per_block + 1, ++ NULL, &spare1, &dummy, 0); ++ ++ if (hweight8(spare0.block_status & spare1.block_status) < 7) ++ *state = YAFFS_BLOCK_STATE_DEAD; ++ else if (memcmp(&spare_ff, &spare0, sizeof(spare_ff)) == 0) ++ *state = YAFFS_BLOCK_STATE_EMPTY; ++ else ++ *state = YAFFS_BLOCK_STATE_NEEDS_SCAN; ++ ++ return YAFFS_OK; ++} ++ ++void yaffs_tags_compat_install(struct yaffs_dev *dev) ++{ ++ if(dev->param.is_yaffs2) ++ return; ++ if(!dev->tagger.write_chunk_tags_fn) ++ dev->tagger.write_chunk_tags_fn = yaffs_tags_compat_wr; ++ if(!dev->tagger.read_chunk_tags_fn) ++ dev->tagger.read_chunk_tags_fn = yaffs_tags_compat_rd; ++ if(!dev->tagger.query_block_fn) ++ dev->tagger.query_block_fn = yaffs_tags_compat_query_block; ++ if(!dev->tagger.mark_bad_fn) ++ dev->tagger.mark_bad_fn = yaffs_tags_compat_mark_bad; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_tagscompat.h linux-3.14.4/fs/yaffs2/yaffs_tagscompat.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_tagscompat.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_tagscompat.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,44 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_TAGSCOMPAT_H__ ++#define __YAFFS_TAGSCOMPAT_H__ ++ ++ ++#include "yaffs_guts.h" ++ ++#if 0 ++ ++ ++int yaffs_tags_compat_wr(struct yaffs_dev *dev, ++ int nand_chunk, ++ const u8 *data, const struct yaffs_ext_tags *tags); ++int yaffs_tags_compat_rd(struct yaffs_dev *dev, ++ int nand_chunk, ++ u8 *data, struct yaffs_ext_tags *tags); ++int yaffs_tags_compat_mark_bad(struct yaffs_dev *dev, int block_no); ++int yaffs_tags_compat_query_block(struct yaffs_dev *dev, ++ int block_no, ++ enum yaffs_block_state *state, ++ u32 *seq_number); ++ ++#endif ++ ++ ++void yaffs_tags_compat_install(struct yaffs_dev *dev); ++void yaffs_calc_tags_ecc(struct yaffs_tags *tags); ++int yaffs_check_tags_ecc(struct yaffs_tags *tags); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_tagsmarshall.c linux-3.14.4/fs/yaffs2/yaffs_tagsmarshall.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_tagsmarshall.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_tagsmarshall.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,199 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_guts.h" ++#include "yaffs_trace.h" ++#include "yaffs_packedtags2.h" ++ ++static int yaffs_tags_marshall_write(struct yaffs_dev *dev, ++ int nand_chunk, const u8 *data, ++ const struct yaffs_ext_tags *tags) ++{ ++ struct yaffs_packed_tags2 pt; ++ int retval; ++ ++ int packed_tags_size = ++ dev->param.no_tags_ecc ? sizeof(pt.t) : sizeof(pt); ++ void *packed_tags_ptr = ++ dev->param.no_tags_ecc ? (void *)&pt.t : (void *)&pt; ++ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "yaffs_tags_marshall_write chunk %d data %p tags %p", ++ nand_chunk, data, tags); ++ ++ /* For yaffs2 writing there must be both data and tags. ++ * If we're using inband tags, then the tags are stuffed into ++ * the end of the data buffer. ++ */ ++ if (!data || !tags) ++ BUG(); ++ else if (dev->param.inband_tags) { ++ struct yaffs_packed_tags2_tags_only *pt2tp; ++ pt2tp = ++ (struct yaffs_packed_tags2_tags_only *)(data + ++ dev-> ++ data_bytes_per_chunk); ++ yaffs_pack_tags2_tags_only(pt2tp, tags); ++ } else { ++ yaffs_pack_tags2(&pt, tags, !dev->param.no_tags_ecc); ++ } ++ ++ retval = dev->drv.drv_write_chunk_fn(dev, nand_chunk, ++ data, dev->param.total_bytes_per_chunk, ++ (dev->param.inband_tags) ? NULL : packed_tags_ptr, ++ (dev->param.inband_tags) ? 0 : packed_tags_size); ++ ++ return retval; ++} ++ ++static int yaffs_tags_marshall_read(struct yaffs_dev *dev, ++ int nand_chunk, u8 *data, ++ struct yaffs_ext_tags *tags) ++{ ++ int retval = 0; ++ int local_data = 0; ++ u8 spare_buffer[100]; ++ enum yaffs_ecc_result ecc_result; ++ ++ struct yaffs_packed_tags2 pt; ++ ++ int packed_tags_size = ++ dev->param.no_tags_ecc ? sizeof(pt.t) : sizeof(pt); ++ void *packed_tags_ptr = ++ dev->param.no_tags_ecc ? (void *)&pt.t : (void *)&pt; ++ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "yaffs_tags_marshall_read chunk %d data %p tags %p", ++ nand_chunk, data, tags); ++ ++ if (dev->param.inband_tags) { ++ if (!data) { ++ local_data = 1; ++ data = yaffs_get_temp_buffer(dev); ++ } ++ } ++ ++ if (dev->param.inband_tags || (data && !tags)) ++ retval = dev->drv.drv_read_chunk_fn(dev, nand_chunk, ++ data, dev->param.total_bytes_per_chunk, ++ NULL, 0, ++ &ecc_result); ++ else if (tags) ++ retval = dev->drv.drv_read_chunk_fn(dev, nand_chunk, ++ data, dev->param.total_bytes_per_chunk, ++ spare_buffer, packed_tags_size, ++ &ecc_result); ++ else ++ BUG(); ++ ++ ++ if (dev->param.inband_tags) { ++ if (tags) { ++ struct yaffs_packed_tags2_tags_only *pt2tp; ++ pt2tp = ++ (struct yaffs_packed_tags2_tags_only *) ++ &data[dev->data_bytes_per_chunk]; ++ yaffs_unpack_tags2_tags_only(tags, pt2tp); ++ } ++ } else if (tags) { ++ memcpy(packed_tags_ptr, spare_buffer, packed_tags_size); ++ yaffs_unpack_tags2(tags, &pt, !dev->param.no_tags_ecc); ++ } ++ ++ if (local_data) ++ yaffs_release_temp_buffer(dev, data); ++ ++ if (tags && ecc_result == YAFFS_ECC_RESULT_UNFIXED) { ++ tags->ecc_result = YAFFS_ECC_RESULT_UNFIXED; ++ dev->n_ecc_unfixed++; ++ } ++ ++ if (tags && ecc_result == -YAFFS_ECC_RESULT_FIXED) { ++ if (tags->ecc_result <= YAFFS_ECC_RESULT_NO_ERROR) ++ tags->ecc_result = YAFFS_ECC_RESULT_FIXED; ++ dev->n_ecc_fixed++; ++ } ++ ++ if (ecc_result < YAFFS_ECC_RESULT_UNFIXED) ++ return YAFFS_OK; ++ else ++ return YAFFS_FAIL; ++} ++ ++static int yaffs_tags_marshall_query_block(struct yaffs_dev *dev, int block_no, ++ enum yaffs_block_state *state, ++ u32 *seq_number) ++{ ++ int retval; ++ ++ yaffs_trace(YAFFS_TRACE_MTD, "yaffs_tags_marshall_query_block %d", ++ block_no); ++ ++ retval = dev->drv.drv_check_bad_fn(dev, block_no); ++ ++ if (retval== YAFFS_FAIL) { ++ yaffs_trace(YAFFS_TRACE_MTD, "block is bad"); ++ ++ *state = YAFFS_BLOCK_STATE_DEAD; ++ *seq_number = 0; ++ } else { ++ struct yaffs_ext_tags t; ++ ++ yaffs_tags_marshall_read(dev, ++ block_no * dev->param.chunks_per_block, ++ NULL, &t); ++ ++ if (t.chunk_used) { ++ *seq_number = t.seq_number; ++ *state = YAFFS_BLOCK_STATE_NEEDS_SCAN; ++ } else { ++ *seq_number = 0; ++ *state = YAFFS_BLOCK_STATE_EMPTY; ++ } ++ } ++ ++ yaffs_trace(YAFFS_TRACE_MTD, ++ "block query returns seq %d state %d", ++ *seq_number, *state); ++ ++ if (retval == 0) ++ return YAFFS_OK; ++ else ++ return YAFFS_FAIL; ++} ++ ++static int yaffs_tags_marshall_mark_bad(struct yaffs_dev *dev, int block_no) ++{ ++ return dev->drv.drv_mark_bad_fn(dev, block_no); ++ ++} ++ ++ ++void yaffs_tags_marshall_install(struct yaffs_dev *dev) ++{ ++ if (!dev->param.is_yaffs2) ++ return; ++ ++ if (!dev->tagger.write_chunk_tags_fn) ++ dev->tagger.write_chunk_tags_fn = yaffs_tags_marshall_write; ++ ++ if (!dev->tagger.read_chunk_tags_fn) ++ dev->tagger.read_chunk_tags_fn = yaffs_tags_marshall_read; ++ ++ if (!dev->tagger.query_block_fn) ++ dev->tagger.query_block_fn = yaffs_tags_marshall_query_block; ++ ++ if (!dev->tagger.mark_bad_fn) ++ dev->tagger.mark_bad_fn = yaffs_tags_marshall_mark_bad; ++ ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_tagsmarshall.h linux-3.14.4/fs/yaffs2/yaffs_tagsmarshall.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_tagsmarshall.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_tagsmarshall.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,22 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_TAGSMARSHALL_H__ ++#define __YAFFS_TAGSMARSHALL_H__ ++ ++#include "yaffs_guts.h" ++void yaffs_tags_marshall_install(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_trace.h linux-3.14.4/fs/yaffs2/yaffs_trace.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_trace.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_trace.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,57 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YTRACE_H__ ++#define __YTRACE_H__ ++ ++extern unsigned int yaffs_trace_mask; ++extern unsigned int yaffs_wr_attempts; ++ ++/* ++ * Tracing flags. ++ * The flags masked in YAFFS_TRACE_ALWAYS are always traced. ++ */ ++ ++#define YAFFS_TRACE_OS 0x00000002 ++#define YAFFS_TRACE_ALLOCATE 0x00000004 ++#define YAFFS_TRACE_SCAN 0x00000008 ++#define YAFFS_TRACE_BAD_BLOCKS 0x00000010 ++#define YAFFS_TRACE_ERASE 0x00000020 ++#define YAFFS_TRACE_GC 0x00000040 ++#define YAFFS_TRACE_WRITE 0x00000080 ++#define YAFFS_TRACE_TRACING 0x00000100 ++#define YAFFS_TRACE_DELETION 0x00000200 ++#define YAFFS_TRACE_BUFFERS 0x00000400 ++#define YAFFS_TRACE_NANDACCESS 0x00000800 ++#define YAFFS_TRACE_GC_DETAIL 0x00001000 ++#define YAFFS_TRACE_SCAN_DEBUG 0x00002000 ++#define YAFFS_TRACE_MTD 0x00004000 ++#define YAFFS_TRACE_CHECKPOINT 0x00008000 ++ ++#define YAFFS_TRACE_VERIFY 0x00010000 ++#define YAFFS_TRACE_VERIFY_NAND 0x00020000 ++#define YAFFS_TRACE_VERIFY_FULL 0x00040000 ++#define YAFFS_TRACE_VERIFY_ALL 0x000f0000 ++ ++#define YAFFS_TRACE_SYNC 0x00100000 ++#define YAFFS_TRACE_BACKGROUND 0x00200000 ++#define YAFFS_TRACE_LOCK 0x00400000 ++#define YAFFS_TRACE_MOUNT 0x00800000 ++ ++#define YAFFS_TRACE_ERROR 0x40000000 ++#define YAFFS_TRACE_BUG 0x80000000 ++#define YAFFS_TRACE_ALWAYS 0xf0000000 ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_verify.c linux-3.14.4/fs/yaffs2/yaffs_verify.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_verify.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_verify.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,529 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_verify.h" ++#include "yaffs_trace.h" ++#include "yaffs_bitmap.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_nand.h" ++ ++int yaffs_skip_verification(struct yaffs_dev *dev) ++{ ++ (void) dev; ++ return !(yaffs_trace_mask & ++ (YAFFS_TRACE_VERIFY | YAFFS_TRACE_VERIFY_FULL)); ++} ++ ++static int yaffs_skip_full_verification(struct yaffs_dev *dev) ++{ ++ (void) dev; ++ return !(yaffs_trace_mask & (YAFFS_TRACE_VERIFY_FULL)); ++} ++ ++static int yaffs_skip_nand_verification(struct yaffs_dev *dev) ++{ ++ (void) dev; ++ return !(yaffs_trace_mask & (YAFFS_TRACE_VERIFY_NAND)); ++} ++ ++static const char * const block_state_name[] = { ++ "Unknown", ++ "Needs scan", ++ "Scanning", ++ "Empty", ++ "Allocating", ++ "Full", ++ "Dirty", ++ "Checkpoint", ++ "Collecting", ++ "Dead" ++}; ++ ++void yaffs_verify_blk(struct yaffs_dev *dev, struct yaffs_block_info *bi, int n) ++{ ++ int actually_used; ++ int in_use; ++ ++ if (yaffs_skip_verification(dev)) ++ return; ++ ++ /* Report illegal runtime states */ ++ if (bi->block_state >= YAFFS_NUMBER_OF_BLOCK_STATES) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Block %d has undefined state %d", ++ n, bi->block_state); ++ ++ switch (bi->block_state) { ++ case YAFFS_BLOCK_STATE_UNKNOWN: ++ case YAFFS_BLOCK_STATE_SCANNING: ++ case YAFFS_BLOCK_STATE_NEEDS_SCAN: ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Block %d has bad run-state %s", ++ n, block_state_name[bi->block_state]); ++ } ++ ++ /* Check pages in use and soft deletions are legal */ ++ ++ actually_used = bi->pages_in_use - bi->soft_del_pages; ++ ++ if (bi->pages_in_use < 0 || ++ bi->pages_in_use > dev->param.chunks_per_block || ++ bi->soft_del_pages < 0 || ++ bi->soft_del_pages > dev->param.chunks_per_block || ++ actually_used < 0 || actually_used > dev->param.chunks_per_block) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Block %d has illegal values pages_in_used %d soft_del_pages %d", ++ n, bi->pages_in_use, bi->soft_del_pages); ++ ++ /* Check chunk bitmap legal */ ++ in_use = yaffs_count_chunk_bits(dev, n); ++ if (in_use != bi->pages_in_use) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Block %d has inconsistent values pages_in_use %d counted chunk bits %d", ++ n, bi->pages_in_use, in_use); ++} ++ ++void yaffs_verify_collected_blk(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi, int n) ++{ ++ yaffs_verify_blk(dev, bi, n); ++ ++ /* After collection the block should be in the erased state */ ++ ++ if (bi->block_state != YAFFS_BLOCK_STATE_COLLECTING && ++ bi->block_state != YAFFS_BLOCK_STATE_EMPTY) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "Block %d is in state %d after gc, should be erased", ++ n, bi->block_state); ++ } ++} ++ ++void yaffs_verify_blocks(struct yaffs_dev *dev) ++{ ++ int i; ++ int state_count[YAFFS_NUMBER_OF_BLOCK_STATES]; ++ int illegal_states = 0; ++ ++ if (yaffs_skip_verification(dev)) ++ return; ++ ++ memset(state_count, 0, sizeof(state_count)); ++ ++ for (i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ struct yaffs_block_info *bi = yaffs_get_block_info(dev, i); ++ yaffs_verify_blk(dev, bi, i); ++ ++ if (bi->block_state < YAFFS_NUMBER_OF_BLOCK_STATES) ++ state_count[bi->block_state]++; ++ else ++ illegal_states++; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_VERIFY, "Block summary"); ++ ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "%d blocks have illegal states", ++ illegal_states); ++ if (state_count[YAFFS_BLOCK_STATE_ALLOCATING] > 1) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Too many allocating blocks"); ++ ++ for (i = 0; i < YAFFS_NUMBER_OF_BLOCK_STATES; i++) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "%s %d blocks", ++ block_state_name[i], state_count[i]); ++ ++ if (dev->blocks_in_checkpt != state_count[YAFFS_BLOCK_STATE_CHECKPOINT]) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Checkpoint block count wrong dev %d count %d", ++ dev->blocks_in_checkpt, ++ state_count[YAFFS_BLOCK_STATE_CHECKPOINT]); ++ ++ if (dev->n_erased_blocks != state_count[YAFFS_BLOCK_STATE_EMPTY]) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Erased block count wrong dev %d count %d", ++ dev->n_erased_blocks, ++ state_count[YAFFS_BLOCK_STATE_EMPTY]); ++ ++ if (state_count[YAFFS_BLOCK_STATE_COLLECTING] > 1) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Too many collecting blocks %d (max is 1)", ++ state_count[YAFFS_BLOCK_STATE_COLLECTING]); ++} ++ ++/* ++ * Verify the object header. oh must be valid, but obj and tags may be NULL in ++ * which case those tests will not be performed. ++ */ ++void yaffs_verify_oh(struct yaffs_obj *obj, struct yaffs_obj_hdr *oh, ++ struct yaffs_ext_tags *tags, int parent_check) ++{ ++ if (obj && yaffs_skip_verification(obj->my_dev)) ++ return; ++ ++ if (!(tags && obj && oh)) { ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Verifying object header tags %p obj %p oh %p", ++ tags, obj, oh); ++ return; ++ } ++ ++ if (oh->type <= YAFFS_OBJECT_TYPE_UNKNOWN || ++ oh->type > YAFFS_OBJECT_TYPE_MAX) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header type is illegal value 0x%x", ++ tags->obj_id, oh->type); ++ ++ if (tags->obj_id != obj->obj_id) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header mismatch obj_id %d", ++ tags->obj_id, obj->obj_id); ++ ++ /* ++ * Check that the object's parent ids match if parent_check requested. ++ * ++ * Tests do not apply to the root object. ++ */ ++ ++ if (parent_check && tags->obj_id > 1 && !obj->parent) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header mismatch parent_id %d obj->parent is NULL", ++ tags->obj_id, oh->parent_obj_id); ++ ++ if (parent_check && obj->parent && ++ oh->parent_obj_id != obj->parent->obj_id && ++ (oh->parent_obj_id != YAFFS_OBJECTID_UNLINKED || ++ obj->parent->obj_id != YAFFS_OBJECTID_DELETED)) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header mismatch parent_id %d parent_obj_id %d", ++ tags->obj_id, oh->parent_obj_id, ++ obj->parent->obj_id); ++ ++ if (tags->obj_id > 1 && oh->name[0] == 0) /* Null name */ ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header name is NULL", ++ obj->obj_id); ++ ++ if (tags->obj_id > 1 && ((u8) (oh->name[0])) == 0xff) /* Junk name */ ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d header name is 0xff", ++ obj->obj_id); ++} ++ ++void yaffs_verify_file(struct yaffs_obj *obj) ++{ ++ u32 x; ++ int required_depth; ++ int actual_depth; ++ int last_chunk; ++ u32 offset_in_chunk; ++ u32 the_chunk; ++ ++ u32 i; ++ struct yaffs_dev *dev; ++ struct yaffs_ext_tags tags; ++ struct yaffs_tnode *tn; ++ u32 obj_id; ++ ++ if (!obj) ++ return; ++ ++ if (yaffs_skip_verification(obj->my_dev)) ++ return; ++ ++ dev = obj->my_dev; ++ obj_id = obj->obj_id; ++ ++ ++ /* Check file size is consistent with tnode depth */ ++ yaffs_addr_to_chunk(dev, obj->variant.file_variant.file_size, ++ &last_chunk, &offset_in_chunk); ++ last_chunk++; ++ x = last_chunk >> YAFFS_TNODES_LEVEL0_BITS; ++ required_depth = 0; ++ while (x > 0) { ++ x >>= YAFFS_TNODES_INTERNAL_BITS; ++ required_depth++; ++ } ++ ++ actual_depth = obj->variant.file_variant.top_level; ++ ++ /* Check that the chunks in the tnode tree are all correct. ++ * We do this by scanning through the tnode tree and ++ * checking the tags for every chunk match. ++ */ ++ ++ if (yaffs_skip_nand_verification(dev)) ++ return; ++ ++ for (i = 1; i <= last_chunk; i++) { ++ tn = yaffs_find_tnode_0(dev, &obj->variant.file_variant, i); ++ ++ if (!tn) ++ continue; ++ ++ the_chunk = yaffs_get_group_base(dev, tn, i); ++ if (the_chunk > 0) { ++ yaffs_rd_chunk_tags_nand(dev, the_chunk, NULL, ++ &tags); ++ if (tags.obj_id != obj_id || tags.chunk_id != i) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Object %d chunk_id %d NAND mismatch chunk %d tags (%d:%d)", ++ obj_id, i, the_chunk, ++ tags.obj_id, tags.chunk_id); ++ } ++ } ++} ++ ++void yaffs_verify_link(struct yaffs_obj *obj) ++{ ++ if (obj && yaffs_skip_verification(obj->my_dev)) ++ return; ++ ++ /* Verify sane equivalent object */ ++} ++ ++void yaffs_verify_symlink(struct yaffs_obj *obj) ++{ ++ if (obj && yaffs_skip_verification(obj->my_dev)) ++ return; ++ ++ /* Verify symlink string */ ++} ++ ++void yaffs_verify_special(struct yaffs_obj *obj) ++{ ++ if (obj && yaffs_skip_verification(obj->my_dev)) ++ return; ++} ++ ++void yaffs_verify_obj(struct yaffs_obj *obj) ++{ ++ struct yaffs_dev *dev; ++ u32 chunk_min; ++ u32 chunk_max; ++ u32 chunk_id_ok; ++ u32 chunk_in_range; ++ u32 chunk_wrongly_deleted; ++ u32 chunk_valid; ++ ++ if (!obj) ++ return; ++ ++ if (obj->being_created) ++ return; ++ ++ dev = obj->my_dev; ++ ++ if (yaffs_skip_verification(dev)) ++ return; ++ ++ /* Check sane object header chunk */ ++ ++ chunk_min = dev->internal_start_block * dev->param.chunks_per_block; ++ chunk_max = ++ (dev->internal_end_block + 1) * dev->param.chunks_per_block - 1; ++ ++ chunk_in_range = (((unsigned)(obj->hdr_chunk)) >= chunk_min && ++ ((unsigned)(obj->hdr_chunk)) <= chunk_max); ++ chunk_id_ok = chunk_in_range || (obj->hdr_chunk == 0); ++ chunk_valid = chunk_in_range && ++ yaffs_check_chunk_bit(dev, ++ obj->hdr_chunk / dev->param.chunks_per_block, ++ obj->hdr_chunk % dev->param.chunks_per_block); ++ chunk_wrongly_deleted = chunk_in_range && !chunk_valid; ++ ++ if (!obj->fake && (!chunk_id_ok || chunk_wrongly_deleted)) ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d has chunk_id %d %s %s", ++ obj->obj_id, obj->hdr_chunk, ++ chunk_id_ok ? "" : ",out of range", ++ chunk_wrongly_deleted ? ",marked as deleted" : ""); ++ ++ if (chunk_valid && !yaffs_skip_nand_verification(dev)) { ++ struct yaffs_ext_tags tags; ++ struct yaffs_obj_hdr *oh; ++ u8 *buffer = yaffs_get_temp_buffer(dev); ++ ++ oh = (struct yaffs_obj_hdr *)buffer; ++ ++ yaffs_rd_chunk_tags_nand(dev, obj->hdr_chunk, buffer, &tags); ++ ++ yaffs_verify_oh(obj, oh, &tags, 1); ++ ++ yaffs_release_temp_buffer(dev, buffer); ++ } ++ ++ /* Verify it has a parent */ ++ if (obj && !obj->fake && (!obj->parent || obj->parent->my_dev != dev)) { ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d has parent pointer %p which does not look like an object", ++ obj->obj_id, obj->parent); ++ } ++ ++ /* Verify parent is a directory */ ++ if (obj->parent && ++ obj->parent->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d's parent is not a directory (type %d)", ++ obj->obj_id, obj->parent->variant_type); ++ } ++ ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ yaffs_verify_file(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ yaffs_verify_symlink(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ yaffs_verify_dir(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ yaffs_verify_link(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ yaffs_verify_special(obj); ++ break; ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ default: ++ yaffs_trace(YAFFS_TRACE_VERIFY, ++ "Obj %d has illegaltype %d", ++ obj->obj_id, obj->variant_type); ++ break; ++ } ++} ++ ++void yaffs_verify_objects(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj; ++ int i; ++ struct list_head *lh; ++ ++ if (yaffs_skip_verification(dev)) ++ return; ++ ++ /* Iterate through the objects in each hash entry */ ++ ++ for (i = 0; i < YAFFS_NOBJECT_BUCKETS; i++) { ++ list_for_each(lh, &dev->obj_bucket[i].list) { ++ obj = list_entry(lh, struct yaffs_obj, hash_link); ++ yaffs_verify_obj(obj); ++ } ++ } ++} ++ ++void yaffs_verify_obj_in_dir(struct yaffs_obj *obj) ++{ ++ struct list_head *lh; ++ struct yaffs_obj *list_obj; ++ int count = 0; ++ ++ if (!obj) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "No object to verify"); ++ BUG(); ++ return; ++ } ++ ++ if (yaffs_skip_verification(obj->my_dev)) ++ return; ++ ++ if (!obj->parent) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "Object does not have parent"); ++ BUG(); ++ return; ++ } ++ ++ if (obj->parent->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "Parent is not directory"); ++ BUG(); ++ } ++ ++ /* Iterate through the objects in each hash entry */ ++ ++ list_for_each(lh, &obj->parent->variant.dir_variant.children) { ++ list_obj = list_entry(lh, struct yaffs_obj, siblings); ++ yaffs_verify_obj(list_obj); ++ if (obj == list_obj) ++ count++; ++ } ++ ++ if (count != 1) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Object in directory %d times", ++ count); ++ BUG(); ++ } ++} ++ ++void yaffs_verify_dir(struct yaffs_obj *directory) ++{ ++ struct list_head *lh; ++ struct yaffs_obj *list_obj; ++ ++ if (!directory) { ++ BUG(); ++ return; ++ } ++ ++ if (yaffs_skip_full_verification(directory->my_dev)) ++ return; ++ ++ if (directory->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Directory has wrong type: %d", ++ directory->variant_type); ++ BUG(); ++ } ++ ++ /* Iterate through the objects in each hash entry */ ++ ++ list_for_each(lh, &directory->variant.dir_variant.children) { ++ list_obj = list_entry(lh, struct yaffs_obj, siblings); ++ if (list_obj->parent != directory) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Object in directory list has wrong parent %p", ++ list_obj->parent); ++ BUG(); ++ } ++ yaffs_verify_obj_in_dir(list_obj); ++ } ++} ++ ++static int yaffs_free_verification_failures; ++ ++void yaffs_verify_free_chunks(struct yaffs_dev *dev) ++{ ++ int counted; ++ int difference; ++ ++ if (yaffs_skip_verification(dev)) ++ return; ++ ++ counted = yaffs_count_free_chunks(dev); ++ ++ difference = dev->n_free_chunks - counted; ++ ++ if (difference) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Freechunks verification failure %d %d %d", ++ dev->n_free_chunks, counted, difference); ++ yaffs_free_verification_failures++; ++ } ++} ++ ++int yaffs_verify_file_sane(struct yaffs_obj *in) ++{ ++ (void) in; ++ return YAFFS_OK; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_verify.h linux-3.14.4/fs/yaffs2/yaffs_verify.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_verify.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_verify.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,43 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_VERIFY_H__ ++#define __YAFFS_VERIFY_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_verify_blk(struct yaffs_dev *dev, struct yaffs_block_info *bi, ++ int n); ++void yaffs_verify_collected_blk(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi, int n); ++void yaffs_verify_blocks(struct yaffs_dev *dev); ++ ++void yaffs_verify_oh(struct yaffs_obj *obj, struct yaffs_obj_hdr *oh, ++ struct yaffs_ext_tags *tags, int parent_check); ++void yaffs_verify_file(struct yaffs_obj *obj); ++void yaffs_verify_link(struct yaffs_obj *obj); ++void yaffs_verify_symlink(struct yaffs_obj *obj); ++void yaffs_verify_special(struct yaffs_obj *obj); ++void yaffs_verify_obj(struct yaffs_obj *obj); ++void yaffs_verify_objects(struct yaffs_dev *dev); ++void yaffs_verify_obj_in_dir(struct yaffs_obj *obj); ++void yaffs_verify_dir(struct yaffs_obj *directory); ++void yaffs_verify_free_chunks(struct yaffs_dev *dev); ++ ++int yaffs_verify_file_sane(struct yaffs_obj *obj); ++ ++int yaffs_skip_verification(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_vfs.c linux-3.14.4/fs/yaffs2/yaffs_vfs.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_vfs.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_vfs.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,3600 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * Acknowledgements: ++ * Luc van OostenRyck for numerous patches. ++ * Nick Bane for numerous patches. ++ * Nick Bane for 2.5/2.6 integration. ++ * Andras Toth for mknod rdev issue. ++ * Michael Fischer for finding the problem with inode inconsistency. ++ * Some code bodily lifted from JFFS ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++/* ++ * ++ * This is the file system front-end to YAFFS that hooks it up to ++ * the VFS. ++ * ++ * Special notes: ++ * >> 2.4: sb->u.generic_sbp points to the struct yaffs_dev associated with ++ * this superblock ++ * >> 2.6: sb->s_fs_info points to the struct yaffs_dev associated with this ++ * superblock ++ * >> inode->u.generic_ip points to the associated struct yaffs_obj. ++ */ ++ ++/* ++ * There are two variants of the VFS glue code. This variant should compile ++ * for any version of Linux. ++ */ ++#include <linux/version.h> ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 10)) ++#define YAFFS_COMPILE_BACKGROUND ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 23)) ++#define YAFFS_COMPILE_FREEZER ++#endif ++#endif ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 28)) ++#define YAFFS_COMPILE_EXPORTFS ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 35)) ++#define YAFFS_USE_SETATTR_COPY ++#define YAFFS_USE_TRUNCATE_SETSIZE ++#endif ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 35)) ++#define YAFFS_HAS_EVICT_INODE ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 13)) ++#define YAFFS_NEW_FOLLOW_LINK 1 ++#else ++#define YAFFS_NEW_FOLLOW_LINK 0 ++#endif ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 6, 0)) ++#define YAFFS_HAS_WRITE_SUPER ++#endif ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 19)) ++#include <linux/config.h> ++#endif ++ ++#include <linux/kernel.h> ++#include <linux/module.h> ++#include <linux/slab.h> ++#include <linux/init.h> ++#include <linux/fs.h> ++#include <linux/proc_fs.h> ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 39)) ++#include <linux/smp_lock.h> ++#endif ++#include <linux/pagemap.h> ++#include <linux/mtd/mtd.h> ++#include <linux/interrupt.h> ++#include <linux/string.h> ++#include <linux/ctype.h> ++ ++#if (YAFFS_NEW_FOLLOW_LINK == 1) ++#include <linux/namei.h> ++#endif ++ ++#ifdef YAFFS_COMPILE_EXPORTFS ++#include <linux/exportfs.h> ++#endif ++ ++#ifdef YAFFS_COMPILE_BACKGROUND ++#include <linux/kthread.h> ++#include <linux/delay.h> ++#endif ++#ifdef YAFFS_COMPILE_FREEZER ++#include <linux/freezer.h> ++#endif ++ ++#include <asm/div64.h> ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ ++#include <linux/statfs.h> ++ ++#define UnlockPage(p) unlock_page(p) ++#define Page_Uptodate(page) test_bit(PG_uptodate, &(page)->flags) ++ ++/* FIXME: use sb->s_id instead ? */ ++#define yaffs_devname(sb, buf) bdevname(sb->s_bdev, buf) ++ ++#else ++ ++#include <linux/locks.h> ++#define BDEVNAME_SIZE 0 ++#define yaffs_devname(sb, buf) kdevname(sb->s_dev) ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 5, 0)) ++/* added NCB 26/5/2006 for 2.4.25-vrs2-tcl1 kernel */ ++#define __user ++#endif ++ ++#endif ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 26)) ++#define YPROC_ROOT (&proc_root) ++#else ++#define YPROC_ROOT NULL ++#endif ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 26)) ++#define Y_INIT_TIMER(a) init_timer(a) ++#else ++#define Y_INIT_TIMER(a) init_timer_on_stack(a) ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 27)) ++#define YAFFS_USE_WRITE_BEGIN_END 1 ++#else ++#define YAFFS_USE_WRITE_BEGIN_END 0 ++#endif ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 6, 0)) ++#define YAFFS_SUPER_HAS_DIRTY ++#endif ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 2, 0)) ++#define set_nlink(inode, count) do { (inode)->i_nlink = (count); } while(0) ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 28)) ++static uint32_t YCALCBLOCKS(uint64_t partition_size, uint32_t block_size) ++{ ++ uint64_t result = partition_size; ++ do_div(result, block_size); ++ return (uint32_t) result; ++} ++#else ++#define YCALCBLOCKS(s, b) ((s)/(b)) ++#endif ++ ++#include <linux/uaccess.h> ++#include <linux/mtd/mtd.h> ++ ++#include "yportenv.h" ++#include "yaffs_trace.h" ++#include "yaffs_guts.h" ++#include "yaffs_attribs.h" ++ ++#include "yaffs_linux.h" ++ ++#include "yaffs_mtdif.h" ++#include "yaffs_packedtags2.h" ++#include "yaffs_getblockinfo.h" ++ ++unsigned int yaffs_trace_mask = ++ YAFFS_TRACE_BAD_BLOCKS | ++ YAFFS_TRACE_ALWAYS | ++ 0; ++ ++unsigned int yaffs_wr_attempts = YAFFS_WR_ATTEMPTS; ++unsigned int yaffs_auto_checkpoint = 1; ++unsigned int yaffs_gc_control = 1; ++unsigned int yaffs_bg_enable = 1; ++unsigned int yaffs_auto_select = 1; ++/* Module Parameters */ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++module_param(yaffs_trace_mask, uint, 0644); ++module_param(yaffs_wr_attempts, uint, 0644); ++module_param(yaffs_auto_checkpoint, uint, 0644); ++module_param(yaffs_gc_control, uint, 0644); ++module_param(yaffs_bg_enable, uint, 0644); ++#else ++MODULE_PARM(yaffs_trace_mask, "i"); ++MODULE_PARM(yaffs_wr_attempts, "i"); ++MODULE_PARM(yaffs_auto_checkpoint, "i"); ++MODULE_PARM(yaffs_gc_control, "i"); ++#endif ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 25)) ++/* use iget and read_inode */ ++#define Y_IGET(sb, inum) iget((sb), (inum)) ++ ++#else ++/* Call local equivalent */ ++#define YAFFS_USE_OWN_IGET ++#define Y_IGET(sb, inum) yaffs_iget((sb), (inum)) ++ ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 18)) ++#define yaffs_inode_to_obj_lv(iptr) ((iptr)->i_private) ++#else ++#define yaffs_inode_to_obj_lv(iptr) ((iptr)->u.generic_ip) ++#endif ++ ++#define yaffs_inode_to_obj(iptr) \ ++ ((struct yaffs_obj *)(yaffs_inode_to_obj_lv(iptr))) ++#define yaffs_dentry_to_obj(dptr) yaffs_inode_to_obj((dptr)->d_inode) ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++#define yaffs_super_to_dev(sb) ((struct yaffs_dev *)sb->s_fs_info) ++#else ++#define yaffs_super_to_dev(sb) ((struct yaffs_dev *)sb->u.generic_sbp) ++#endif ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 5, 0)) ++#define Y_CLEAR_INODE(i) clear_inode(i) ++#else ++#define Y_CLEAR_INODE(i) end_writeback(i) ++#endif ++ ++ ++#define update_dir_time(dir) do {\ ++ (dir)->i_ctime = (dir)->i_mtime = CURRENT_TIME; \ ++ } while (0) ++ ++static void yaffs_fill_inode_from_obj(struct inode *inode, ++ struct yaffs_obj *obj); ++ ++ ++static void yaffs_gross_lock(struct yaffs_dev *dev) ++{ ++ yaffs_trace(YAFFS_TRACE_LOCK, "yaffs locking %p", current); ++ mutex_lock(&(yaffs_dev_to_lc(dev)->gross_lock)); ++ yaffs_trace(YAFFS_TRACE_LOCK, "yaffs locked %p", current); ++} ++ ++static void yaffs_gross_unlock(struct yaffs_dev *dev) ++{ ++ yaffs_trace(YAFFS_TRACE_LOCK, "yaffs unlocking %p", current); ++ mutex_unlock(&(yaffs_dev_to_lc(dev)->gross_lock)); ++} ++ ++ ++static int yaffs_readpage_nolock(struct file *f, struct page *pg) ++{ ++ /* Lifted from jffs2 */ ++ ++ struct yaffs_obj *obj; ++ unsigned char *pg_buf; ++ int ret; ++ loff_t pos = ((loff_t) pg->index) << PAGE_CACHE_SHIFT; ++ struct yaffs_dev *dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readpage_nolock at %lld, size %08x", ++ (long long)pos, ++ (unsigned)PAGE_CACHE_SIZE); ++ ++ obj = yaffs_dentry_to_obj(f->f_dentry); ++ ++ dev = obj->my_dev; ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ BUG_ON(!PageLocked(pg)); ++#else ++ if (!PageLocked(pg)) ++ PAGE_BUG(pg); ++#endif ++ ++ pg_buf = kmap(pg); ++ /* FIXME: Can kmap fail? */ ++ ++ yaffs_gross_lock(dev); ++ ++ ret = yaffs_file_rd(obj, pg_buf, pos, PAGE_CACHE_SIZE); ++ ++ yaffs_gross_unlock(dev); ++ ++ if (ret >= 0) ++ ret = 0; ++ ++ if (ret) { ++ ClearPageUptodate(pg); ++ SetPageError(pg); ++ } else { ++ SetPageUptodate(pg); ++ ClearPageError(pg); ++ } ++ ++ flush_dcache_page(pg); ++ kunmap(pg); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_readpage_nolock done"); ++ return ret; ++} ++ ++static int yaffs_readpage_unlock(struct file *f, struct page *pg) ++{ ++ int ret = yaffs_readpage_nolock(f, pg); ++ UnlockPage(pg); ++ return ret; ++} ++ ++static int yaffs_readpage(struct file *f, struct page *pg) ++{ ++ int ret; ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_readpage"); ++ ret = yaffs_readpage_unlock(f, pg); ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_readpage done"); ++ return ret; ++} ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 5, 0)) ++#define YCRED_FSUID() from_kuid(&init_user_ns, current_fsuid()) ++#define YCRED_FSGID() from_kgid(&init_user_ns, current_fsgid()) ++#else ++#define YCRED_FSUID() YCRED(current)->fsuid ++#define YCRED_FSGID() YCRED(current)->fsgid ++ ++static inline uid_t i_uid_read(const struct inode *inode) ++{ ++ return inode->i_uid; ++} ++ ++static inline gid_t i_gid_read(const struct inode *inode) ++{ ++ return inode->i_gid; ++} ++ ++static inline void i_uid_write(struct inode *inode, uid_t uid) ++{ ++ inode->i_uid = uid; ++} ++ ++static inline void i_gid_write(struct inode *inode, gid_t gid) ++{ ++ inode->i_gid = gid; ++} ++#endif ++ ++static void yaffs_set_super_dirty_val(struct yaffs_dev *dev, int val) ++{ ++ struct yaffs_linux_context *lc = yaffs_dev_to_lc(dev); ++ ++ if (lc) ++ lc->dirty = val; ++ ++# ifdef YAFFS_SUPER_HAS_DIRTY ++ { ++ struct super_block *sb = lc->super; ++ ++ if (sb) ++ sb->s_dirt = val; ++ } ++#endif ++ ++} ++ ++static void yaffs_set_super_dirty(struct yaffs_dev *dev) ++{ ++ yaffs_set_super_dirty_val(dev, 1); ++} ++ ++static void yaffs_clear_super_dirty(struct yaffs_dev *dev) ++{ ++ yaffs_set_super_dirty_val(dev, 0); ++} ++ ++static int yaffs_check_super_dirty(struct yaffs_dev *dev) ++{ ++ struct yaffs_linux_context *lc = yaffs_dev_to_lc(dev); ++ ++ if (lc && lc->dirty) ++ return 1; ++ ++# ifdef YAFFS_SUPER_HAS_DIRTY ++ { ++ struct super_block *sb = lc->super; ++ ++ if (sb && sb->s_dirt) ++ return 1; ++ } ++#endif ++ return 0; ++ ++} ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs_writepage(struct page *page, struct writeback_control *wbc) ++#else ++static int yaffs_writepage(struct page *page) ++#endif ++{ ++ struct yaffs_dev *dev; ++ struct address_space *mapping = page->mapping; ++ struct inode *inode; ++ unsigned long end_index; ++ char *buffer; ++ struct yaffs_obj *obj; ++ int n_written = 0; ++ unsigned n_bytes; ++ loff_t i_size; ++ ++ if (!mapping) ++ BUG(); ++ inode = mapping->host; ++ if (!inode) ++ BUG(); ++ i_size = i_size_read(inode); ++ ++ end_index = i_size >> PAGE_CACHE_SHIFT; ++ ++ if (page->index < end_index) ++ n_bytes = PAGE_CACHE_SIZE; ++ else { ++ n_bytes = i_size & (PAGE_CACHE_SIZE - 1); ++ ++ if (page->index > end_index || !n_bytes) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_writepage at %lld, inode size = %lld!!", ++ ((loff_t)page->index) << PAGE_CACHE_SHIFT, ++ inode->i_size); ++ yaffs_trace(YAFFS_TRACE_OS, ++ " -> don't care!!"); ++ ++ zero_user_segment(page, 0, PAGE_CACHE_SIZE); ++ set_page_writeback(page); ++ unlock_page(page); ++ end_page_writeback(page); ++ return 0; ++ } ++ } ++ ++ if (n_bytes != PAGE_CACHE_SIZE) ++ zero_user_segment(page, n_bytes, PAGE_CACHE_SIZE); ++ ++ get_page(page); ++ ++ buffer = kmap(page); ++ ++ obj = yaffs_inode_to_obj(inode); ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_writepage at %lld, size %08x", ++ ((loff_t)page->index) << PAGE_CACHE_SHIFT, n_bytes); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "writepag0: obj = %lld, ino = %lld", ++ obj->variant.file_variant.file_size, inode->i_size); ++ ++ n_written = yaffs_wr_file(obj, buffer, ++ ((loff_t)page->index) << PAGE_CACHE_SHIFT, n_bytes, 0); ++ ++ yaffs_set_super_dirty(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "writepag1: obj = %lld, ino = %lld", ++ obj->variant.file_variant.file_size, inode->i_size); ++ ++ yaffs_gross_unlock(dev); ++ ++ kunmap(page); ++ set_page_writeback(page); ++ unlock_page(page); ++ end_page_writeback(page); ++ put_page(page); ++ ++ return (n_written == n_bytes) ? 0 : -ENOSPC; ++} ++ ++/* Space holding and freeing is done to ensure we have space available for write_begin/end */ ++/* For now we just assume few parallel writes and check against a small number. */ ++/* Todo: need to do this with a counter to handle parallel reads better */ ++ ++static ssize_t yaffs_hold_space(struct file *f) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ ++ int n_free_chunks; ++ ++ obj = yaffs_dentry_to_obj(f->f_dentry); ++ ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ n_free_chunks = yaffs_get_n_free_chunks(dev); ++ ++ yaffs_gross_unlock(dev); ++ ++ return (n_free_chunks > 20) ? 1 : 0; ++} ++ ++static void yaffs_release_space(struct file *f) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ ++ obj = yaffs_dentry_to_obj(f->f_dentry); ++ ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ yaffs_gross_unlock(dev); ++} ++ ++#if (YAFFS_USE_WRITE_BEGIN_END > 0) ++static int yaffs_write_begin(struct file *filp, struct address_space *mapping, ++ loff_t pos, unsigned len, unsigned flags, ++ struct page **pagep, void **fsdata) ++{ ++ struct page *pg = NULL; ++ pgoff_t index = pos >> PAGE_CACHE_SHIFT; ++ ++ int ret = 0; ++ int space_held = 0; ++ ++ /* Get a page */ ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 28) ++ pg = grab_cache_page_write_begin(mapping, index, flags); ++#else ++ pg = __grab_cache_page(mapping, index); ++#endif ++ ++ *pagep = pg; ++ if (!pg) { ++ ret = -ENOMEM; ++ goto out; ++ } ++ yaffs_trace(YAFFS_TRACE_OS, ++ "start yaffs_write_begin index %d(%x) uptodate %d", ++ (int)index, (int)index, Page_Uptodate(pg) ? 1 : 0); ++ ++ /* Get fs space */ ++ space_held = yaffs_hold_space(filp); ++ ++ if (!space_held) { ++ ret = -ENOSPC; ++ goto out; ++ } ++ ++ /* Update page if required */ ++ ++ if (!Page_Uptodate(pg)) ++ ret = yaffs_readpage_nolock(filp, pg); ++ ++ if (ret) ++ goto out; ++ ++ /* Happy path return */ ++ yaffs_trace(YAFFS_TRACE_OS, "end yaffs_write_begin - ok"); ++ ++ return 0; ++ ++out: ++ yaffs_trace(YAFFS_TRACE_OS, ++ "end yaffs_write_begin fail returning %d", ret); ++ if (space_held) ++ yaffs_release_space(filp); ++ if (pg) { ++ unlock_page(pg); ++ page_cache_release(pg); ++ } ++ return ret; ++} ++ ++#else ++ ++static int yaffs_prepare_write(struct file *f, struct page *pg, ++ unsigned offset, unsigned to) ++{ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_prepair_write"); ++ ++ if (!Page_Uptodate(pg)) ++ return yaffs_readpage_nolock(f, pg); ++ return 0; ++} ++#endif ++ ++ ++static ssize_t yaffs_file_write(struct file *f, const char *buf, size_t n, ++ loff_t * pos) ++{ ++ struct yaffs_obj *obj; ++ int n_written; ++ loff_t ipos; ++ struct inode *inode; ++ struct yaffs_dev *dev; ++ ++ obj = yaffs_dentry_to_obj(f->f_dentry); ++ ++ if (!obj) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_file_write: hey obj is null!"); ++ return -EINVAL; ++ } ++ ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ inode = f->f_dentry->d_inode; ++ ++ if (!S_ISBLK(inode->i_mode) && f->f_flags & O_APPEND) ++ ipos = inode->i_size; ++ else ++ ipos = *pos; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_file_write about to write writing %u(%x) bytes to object %d at %lld", ++ (unsigned)n, (unsigned)n, obj->obj_id, ipos); ++ ++ n_written = yaffs_wr_file(obj, buf, ipos, n, 0); ++ ++ yaffs_set_super_dirty(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_file_write: %d(%x) bytes written", ++ (unsigned)n, (unsigned)n); ++ ++ if (n_written > 0) { ++ ipos += n_written; ++ *pos = ipos; ++ if (ipos > inode->i_size) { ++ inode->i_size = ipos; ++ inode->i_blocks = (ipos + 511) >> 9; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_file_write size updated to %lld bytes, %d blocks", ++ ipos, (int)(inode->i_blocks)); ++ } ++ ++ } ++ yaffs_gross_unlock(dev); ++ return (n_written == 0) && (n > 0) ? -ENOSPC : n_written; ++} ++ ++ ++#if (YAFFS_USE_WRITE_BEGIN_END > 0) ++static int yaffs_write_end(struct file *filp, struct address_space *mapping, ++ loff_t pos, unsigned len, unsigned copied, ++ struct page *pg, void *fsdadata) ++{ ++ int ret = 0; ++ void *addr, *kva; ++ uint32_t offset_into_page = pos & (PAGE_CACHE_SIZE - 1); ++ ++ kva = kmap(pg); ++ addr = kva + offset_into_page; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_write_end addr %p pos %lld n_bytes %d", ++ addr, pos, copied); ++ ++ ret = yaffs_file_write(filp, addr, copied, &pos); ++ ++ if (ret != copied) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_write_end not same size ret %d copied %d", ++ ret, copied); ++ SetPageError(pg); ++ } ++ ++ kunmap(pg); ++ ++ yaffs_release_space(filp); ++ unlock_page(pg); ++ page_cache_release(pg); ++ return ret; ++} ++#else ++ ++static int yaffs_commit_write(struct file *f, struct page *pg, unsigned offset, ++ unsigned to) ++{ ++ void *addr, *kva; ++ ++ loff_t pos = (((loff_t) pg->index) << PAGE_CACHE_SHIFT) + offset; ++ int n_bytes = to - offset; ++ int n_written; ++ ++ kva = kmap(pg); ++ addr = kva + offset; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_commit_write addr %p pos %lld n_bytes %d", ++ addr, pos, n_bytes); ++ ++ n_written = yaffs_file_write(f, addr, n_bytes, &pos); ++ ++ if (n_written != n_bytes) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_commit_write not same size n_written %d n_bytes %d", ++ n_written, n_bytes); ++ SetPageError(pg); ++ } ++ kunmap(pg); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_commit_write returning %d", ++ n_written == n_bytes ? 0 : n_written); ++ ++ return n_written == n_bytes ? 0 : n_written; ++} ++#endif ++ ++static struct address_space_operations yaffs_file_address_operations = { ++ .readpage = yaffs_readpage, ++ .writepage = yaffs_writepage, ++#if (YAFFS_USE_WRITE_BEGIN_END > 0) ++ .write_begin = yaffs_write_begin, ++ .write_end = yaffs_write_end, ++#else ++ .prepare_write = yaffs_prepare_write, ++ .commit_write = yaffs_commit_write, ++#endif ++}; ++ ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static int yaffs_file_flush(struct file *file, fl_owner_t id) ++#else ++static int yaffs_file_flush(struct file *file) ++#endif ++{ ++ struct yaffs_obj *obj = yaffs_dentry_to_obj(file->f_dentry); ++ ++ struct yaffs_dev *dev = obj->my_dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_file_flush object %d (%s)", ++ obj->obj_id, ++ obj->dirty ? "dirty" : "clean"); ++ ++ yaffs_gross_lock(dev); ++ ++ yaffs_flush_file(obj, 1, 0); ++ ++ yaffs_gross_unlock(dev); ++ ++ return 0; ++} ++ ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 39)) ++static int yaffs_sync_object(struct file *file, loff_t start, loff_t end, int datasync) ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 34)) ++static int yaffs_sync_object(struct file *file, int datasync) ++#else ++static int yaffs_sync_object(struct file *file, struct dentry *dentry, ++ int datasync) ++#endif ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 34)) ++ struct dentry *dentry = file->f_path.dentry; ++#endif ++ ++ obj = yaffs_dentry_to_obj(dentry); ++ ++ dev = obj->my_dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_SYNC, ++ "yaffs_sync_object"); ++ yaffs_gross_lock(dev); ++ yaffs_flush_file(obj, 1, datasync); ++ yaffs_gross_unlock(dev); ++ return 0; ++} ++ ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 22)) ++static const struct file_operations yaffs_file_operations = { ++ .read = do_sync_read, ++ .write = do_sync_write, ++ .aio_read = generic_file_aio_read, ++ .aio_write = generic_file_aio_write, ++ .mmap = generic_file_mmap, ++ .flush = yaffs_file_flush, ++ .fsync = yaffs_sync_object, ++ .splice_read = generic_file_splice_read, ++ .splice_write = generic_file_splice_write, ++ .llseek = generic_file_llseek, ++}; ++ ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 18)) ++ ++static const struct file_operations yaffs_file_operations = { ++ .read = do_sync_read, ++ .write = do_sync_write, ++ .aio_read = generic_file_aio_read, ++ .aio_write = generic_file_aio_write, ++ .mmap = generic_file_mmap, ++ .flush = yaffs_file_flush, ++ .fsync = yaffs_sync_object, ++ .sendfile = generic_file_sendfile, ++}; ++ ++#else ++ ++static const struct file_operations yaffs_file_operations = { ++ .read = generic_file_read, ++ .write = generic_file_write, ++ .mmap = generic_file_mmap, ++ .flush = yaffs_file_flush, ++ .fsync = yaffs_sync_object, ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ .sendfile = generic_file_sendfile, ++#endif ++}; ++#endif ++ ++ ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 25)) ++static void zero_user_segment(struct page *page, unsigned start, unsigned end) ++{ ++ void *kaddr = kmap_atomic(page, KM_USER0); ++ memset(kaddr + start, 0, end - start); ++ kunmap_atomic(kaddr, KM_USER0); ++ flush_dcache_page(page); ++} ++#endif ++ ++ ++static int yaffs_vfs_setsize(struct inode *inode, loff_t newsize) ++{ ++#ifdef YAFFS_USE_TRUNCATE_SETSIZE ++ truncate_setsize(inode, newsize); ++ return 0; ++#else ++ truncate_inode_pages(&inode->i_data, newsize); ++ return 0; ++#endif ++ ++} ++ ++ ++static int yaffs_vfs_setattr(struct inode *inode, struct iattr *attr) ++{ ++#ifdef YAFFS_USE_SETATTR_COPY ++ setattr_copy(inode, attr); ++ return 0; ++#else ++ return inode_setattr(inode, attr); ++#endif ++ ++} ++ ++static int yaffs_setattr(struct dentry *dentry, struct iattr *attr) ++{ ++ struct inode *inode = dentry->d_inode; ++ int error = 0; ++ struct yaffs_dev *dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_setattr of object %d", ++ yaffs_inode_to_obj(inode)->obj_id); ++#if 0 ++ /* Fail if a requested resize >= 2GB */ ++ if (attr->ia_valid & ATTR_SIZE && (attr->ia_size >> 31)) ++ error = -EINVAL; ++#endif ++ ++ if (error == 0) ++ error = inode_change_ok(inode, attr); ++ if (error == 0) { ++ int result; ++ if (!error) { ++ error = yaffs_vfs_setattr(inode, attr); ++ yaffs_trace(YAFFS_TRACE_OS, "inode_setattr called"); ++ if (attr->ia_valid & ATTR_SIZE) { ++ yaffs_vfs_setsize(inode, attr->ia_size); ++ inode->i_blocks = (inode->i_size + 511) >> 9; ++ } ++ } ++ dev = yaffs_inode_to_obj(inode)->my_dev; ++ if (attr->ia_valid & ATTR_SIZE) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "resize to %d(%x)", ++ (int)(attr->ia_size), ++ (int)(attr->ia_size)); ++ } ++ yaffs_gross_lock(dev); ++ result = yaffs_set_attribs(yaffs_inode_to_obj(inode), attr); ++ if (result == YAFFS_OK) { ++ error = 0; ++ } else { ++ error = -EPERM; ++ } ++ yaffs_gross_unlock(dev); ++ ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_setattr done returning %d", error); ++ ++ return error; ++} ++ ++static int yaffs_setxattr(struct dentry *dentry, const char *name, ++ const void *value, size_t size, int flags) ++{ ++ struct inode *inode = dentry->d_inode; ++ int error = 0; ++ struct yaffs_dev *dev; ++ struct yaffs_obj *obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_setxattr of object %d", obj->obj_id); ++ ++ if (error == 0) { ++ int result; ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ result = yaffs_set_xattrib(obj, name, value, size, flags); ++ if (result == YAFFS_OK) ++ error = 0; ++ else if (result < 0) ++ error = result; ++ yaffs_gross_unlock(dev); ++ ++ } ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_setxattr done returning %d", error); ++ ++ return error; ++} ++ ++static ssize_t yaffs_getxattr(struct dentry * dentry, const char *name, ++ void *buff, size_t size) ++{ ++ struct inode *inode = dentry->d_inode; ++ int error = 0; ++ struct yaffs_dev *dev; ++ struct yaffs_obj *obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_getxattr \"%s\" from object %d", ++ name, obj->obj_id); ++ ++ if (error == 0) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ error = yaffs_get_xattrib(obj, name, buff, size); ++ yaffs_gross_unlock(dev); ++ ++ } ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_getxattr done returning %d", error); ++ ++ return error; ++} ++ ++static int yaffs_removexattr(struct dentry *dentry, const char *name) ++{ ++ struct inode *inode = dentry->d_inode; ++ int error = 0; ++ struct yaffs_dev *dev; ++ struct yaffs_obj *obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_removexattr of object %d", obj->obj_id); ++ ++ if (error == 0) { ++ int result; ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ result = yaffs_remove_xattrib(obj, name); ++ if (result == YAFFS_OK) ++ error = 0; ++ else if (result < 0) ++ error = result; ++ yaffs_gross_unlock(dev); ++ ++ } ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_removexattr done returning %d", error); ++ ++ return error; ++} ++ ++static ssize_t yaffs_listxattr(struct dentry * dentry, char *buff, size_t size) ++{ ++ struct inode *inode = dentry->d_inode; ++ int error = 0; ++ struct yaffs_dev *dev; ++ struct yaffs_obj *obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_listxattr of object %d", obj->obj_id); ++ ++ if (error == 0) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ error = yaffs_list_xattrib(obj, buff, size); ++ yaffs_gross_unlock(dev); ++ ++ } ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_listxattr done returning %d", error); ++ ++ return error; ++} ++ ++ ++static const struct inode_operations yaffs_file_inode_operations = { ++ .setattr = yaffs_setattr, ++ .setxattr = yaffs_setxattr, ++ .getxattr = yaffs_getxattr, ++ .listxattr = yaffs_listxattr, ++ .removexattr = yaffs_removexattr, ++}; ++ ++ ++static int yaffs_readlink(struct dentry *dentry, char __user * buffer, ++ int buflen) ++{ ++ unsigned char *alias; ++ int ret; ++ ++ struct yaffs_dev *dev = yaffs_dentry_to_obj(dentry)->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ alias = yaffs_get_symlink_alias(yaffs_dentry_to_obj(dentry)); ++ ++ yaffs_gross_unlock(dev); ++ ++ if (!alias) ++ return -ENOMEM; ++ ++ ret = vfs_readlink(dentry, buffer, buflen, alias); ++ kfree(alias); ++ return ret; ++} ++ ++#if (YAFFS_NEW_FOLLOW_LINK == 1) ++static void *yaffs_follow_link(struct dentry *dentry, struct nameidata *nd) ++{ ++ void *ret; ++#else ++static int yaffs_follow_link(struct dentry *dentry, struct nameidata *nd) ++{ ++ int ret ++#endif ++ unsigned char *alias; ++ int ret_int = 0; ++ struct yaffs_dev *dev = yaffs_dentry_to_obj(dentry)->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ alias = yaffs_get_symlink_alias(yaffs_dentry_to_obj(dentry)); ++ yaffs_gross_unlock(dev); ++ ++ if (!alias) { ++ ret_int = -ENOMEM; ++ goto out; ++ } ++#if (YAFFS_NEW_FOLLOW_LINK == 1) ++ nd_set_link(nd, alias); ++ ret = alias; ++out: ++ if (ret_int) ++ ret = ERR_PTR(ret_int); ++ return ret; ++#else ++ ret = vfs_follow_link(nd, alias); ++ kfree(alias); ++out: ++ if (ret_int) ++ ret = ret_int; ++ return ret; ++#endif ++} ++ ++ ++#ifdef YAFFS_HAS_PUT_INODE ++ ++/* For now put inode is just for debugging ++ * Put inode is called when the inode **structure** is put. ++ */ ++static void yaffs_put_inode(struct inode *inode) ++{ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_put_inode: ino %d, count %d"), ++ (int)inode->i_ino, atomic_read(&inode->i_count); ++ ++} ++#endif ++ ++#if (YAFFS_NEW_FOLLOW_LINK == 1) ++void yaffs_put_link(struct dentry *dentry, struct nameidata *nd, void *alias) ++{ ++ kfree(alias); ++} ++#endif ++ ++static const struct inode_operations yaffs_symlink_inode_operations = { ++ .readlink = yaffs_readlink, ++ .follow_link = yaffs_follow_link, ++#if (YAFFS_NEW_FOLLOW_LINK == 1) ++ .put_link = yaffs_put_link, ++#endif ++ .setattr = yaffs_setattr, ++ .setxattr = yaffs_setxattr, ++ .getxattr = yaffs_getxattr, ++ .listxattr = yaffs_listxattr, ++ .removexattr = yaffs_removexattr, ++}; ++ ++#ifdef YAFFS_USE_OWN_IGET ++ ++static struct inode *yaffs_iget(struct super_block *sb, unsigned long ino) ++{ ++ struct inode *inode; ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_iget for %lu", ino); ++ ++ inode = iget_locked(sb, ino); ++ if (!inode) ++ return ERR_PTR(-ENOMEM); ++ if (!(inode->i_state & I_NEW)) ++ return inode; ++ ++ /* NB This is called as a side effect of other functions, but ++ * we had to release the lock to prevent deadlocks, so ++ * need to lock again. ++ */ ++ ++ yaffs_gross_lock(dev); ++ ++ obj = yaffs_find_by_number(dev, inode->i_ino); ++ ++ yaffs_fill_inode_from_obj(inode, obj); ++ ++ yaffs_gross_unlock(dev); ++ ++ unlock_new_inode(inode); ++ return inode; ++} ++ ++#else ++ ++static void yaffs_read_inode(struct inode *inode) ++{ ++ /* NB This is called as a side effect of other functions, but ++ * we had to release the lock to prevent deadlocks, so ++ * need to lock again. ++ */ ++ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev = yaffs_super_to_dev(inode->i_sb); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_read_inode for %d", (int)inode->i_ino); ++ ++ if (current != yaffs_dev_to_lc(dev)->readdir_process) ++ yaffs_gross_lock(dev); ++ ++ obj = yaffs_find_by_number(dev, inode->i_ino); ++ ++ yaffs_fill_inode_from_obj(inode, obj); ++ ++ if (current != yaffs_dev_to_lc(dev)->readdir_process) ++ yaffs_gross_unlock(dev); ++} ++ ++#endif ++ ++ ++ ++struct inode *yaffs_get_inode(struct super_block *sb, int mode, int dev, ++ struct yaffs_obj *obj) ++{ ++ struct inode *inode; ++ ++ if (!sb) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_get_inode for NULL super_block!!"); ++ return NULL; ++ ++ } ++ ++ if (!obj) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_get_inode for NULL object!!"); ++ return NULL; ++ ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_get_inode for object %d", obj->obj_id); ++ ++ inode = Y_IGET(sb, obj->obj_id); ++ if (IS_ERR(inode)) ++ return NULL; ++ ++ /* NB Side effect: iget calls back to yaffs_read_inode(). */ ++ /* iget also increments the inode's i_count */ ++ /* NB You can't be holding gross_lock or deadlock will happen! */ ++ ++ return inode; ++} ++ ++ ++ ++#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 29) ++#define YCRED(x) x ++#else ++#define YCRED(x) (x->cred) ++#endif ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 4, 0)) ++static int yaffs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, ++ dev_t rdev) ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs_mknod(struct inode *dir, struct dentry *dentry, int mode, ++ dev_t rdev) ++#else ++static int yaffs_mknod(struct inode *dir, struct dentry *dentry, int mode, ++ int rdev) ++#endif ++{ ++ struct inode *inode; ++ ++ struct yaffs_obj *obj = NULL; ++ struct yaffs_dev *dev; ++ ++ struct yaffs_obj *parent = yaffs_inode_to_obj(dir); ++ ++ int error = -ENOSPC; ++ uid_t uid = YCRED_FSUID(); ++ gid_t gid = ++ (dir->i_mode & S_ISGID) ? i_gid_read(dir) : YCRED_FSGID(); ++ ++ if ((dir->i_mode & S_ISGID) && S_ISDIR(mode)) ++ mode |= S_ISGID; ++ ++ if (parent) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_mknod: parent object %d type %d", ++ parent->obj_id, parent->variant_type); ++ } else { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_mknod: could not get parent object"); ++ return -EPERM; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_mknod: making oject for %s, mode %x dev %x", ++ dentry->d_name.name, mode, rdev); ++ ++ dev = parent->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ switch (mode & S_IFMT) { ++ default: ++ /* Special (socket, fifo, device...) */ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mknod: making special"); ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ obj = ++ yaffs_create_special(parent, dentry->d_name.name, mode, uid, ++ gid, old_encode_dev(rdev)); ++#else ++ obj = ++ yaffs_create_special(parent, dentry->d_name.name, mode, uid, ++ gid, rdev); ++#endif ++ break; ++ case S_IFREG: /* file */ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mknod: making file"); ++ obj = yaffs_create_file(parent, dentry->d_name.name, mode, uid, ++ gid); ++ break; ++ case S_IFDIR: /* directory */ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mknod: making directory"); ++ obj = yaffs_create_dir(parent, dentry->d_name.name, mode, ++ uid, gid); ++ break; ++ case S_IFLNK: /* symlink */ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mknod: making symlink"); ++ obj = NULL; /* Do we ever get here? */ ++ break; ++ } ++ ++ /* Can not call yaffs_get_inode() with gross lock held */ ++ yaffs_gross_unlock(dev); ++ ++ if (obj) { ++ inode = yaffs_get_inode(dir->i_sb, mode, rdev, obj); ++ d_instantiate(dentry, inode); ++ update_dir_time(dir); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_mknod created object %d count = %d", ++ obj->obj_id, atomic_read(&inode->i_count)); ++ error = 0; ++ yaffs_fill_inode_from_obj(dir, parent); ++ } else { ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mknod failed making object"); ++ error = -ENOMEM; ++ } ++ ++ return error; ++} ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 4, 0)) ++static int yaffs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) ++#else ++static int yaffs_mkdir(struct inode *dir, struct dentry *dentry, int mode) ++#endif ++{ ++ int ret_val; ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_mkdir"); ++ ret_val = yaffs_mknod(dir, dentry, mode | S_IFDIR, 0); ++ return ret_val; ++} ++ ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 6, 0)) ++static int yaffs_create(struct inode *dir, struct dentry *dentry, umode_t mode, ++ bool dummy) ++#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 4, 0)) ++static int yaffs_create(struct inode *dir, struct dentry *dentry, umode_t mode, ++ struct nameidata *n) ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs_create(struct inode *dir, struct dentry *dentry, int mode, ++ struct nameidata *n) ++#else ++static int yaffs_create(struct inode *dir, struct dentry *dentry, int mode) ++#endif ++{ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_create"); ++ return yaffs_mknod(dir, dentry, mode | S_IFREG, 0); ++} ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 6, 0)) ++static struct dentry *yaffs_lookup(struct inode *dir, struct dentry *dentry, ++ unsigned int dummy) ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static struct dentry *yaffs_lookup(struct inode *dir, struct dentry *dentry, ++ struct nameidata *n) ++#else ++static struct dentry *yaffs_lookup(struct inode *dir, struct dentry *dentry) ++#endif ++{ ++ struct yaffs_obj *obj; ++ struct inode *inode = NULL; /* NCB 2.5/2.6 needs NULL here */ ++ ++ struct yaffs_dev *dev = yaffs_inode_to_obj(dir)->my_dev; ++ ++ if (current != yaffs_dev_to_lc(dev)->readdir_process) ++ yaffs_gross_lock(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_lookup for %d:%s", ++ yaffs_inode_to_obj(dir)->obj_id, dentry->d_name.name); ++ ++ obj = yaffs_find_by_name(yaffs_inode_to_obj(dir), dentry->d_name.name); ++ ++ obj = yaffs_get_equivalent_obj(obj); /* in case it was a hardlink */ ++ ++ /* Can't hold gross lock when calling yaffs_get_inode() */ ++ if (current != yaffs_dev_to_lc(dev)->readdir_process) ++ yaffs_gross_unlock(dev); ++ ++ if (obj) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_lookup found %d", obj->obj_id); ++ ++ inode = yaffs_get_inode(dir->i_sb, obj->yst_mode, 0, obj); ++ } else { ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_lookup not found"); ++ ++ } ++ ++/* added NCB for 2.5/6 compatability - forces add even if inode is ++ * NULL which creates dentry hash */ ++ d_add(dentry, inode); ++ ++ return NULL; ++} ++ ++/* ++ * Create a link... ++ */ ++static int yaffs_link(struct dentry *old_dentry, struct inode *dir, ++ struct dentry *dentry) ++{ ++ struct inode *inode = old_dentry->d_inode; ++ struct yaffs_obj *obj = NULL; ++ struct yaffs_obj *link = NULL; ++ struct yaffs_dev *dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_link"); ++ ++ obj = yaffs_inode_to_obj(inode); ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ if (!S_ISDIR(inode->i_mode)) /* Don't link directories */ ++ link = ++ yaffs_link_obj(yaffs_inode_to_obj(dir), dentry->d_name.name, ++ obj); ++ ++ if (link) { ++ set_nlink(old_dentry->d_inode, yaffs_get_obj_link_count(obj)); ++ d_instantiate(dentry, old_dentry->d_inode); ++ atomic_inc(&old_dentry->d_inode->i_count); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_link link count %d i_count %d", ++ old_dentry->d_inode->i_nlink, ++ atomic_read(&old_dentry->d_inode->i_count)); ++ } ++ ++ yaffs_gross_unlock(dev); ++ ++ if (link) { ++ update_dir_time(dir); ++ return 0; ++ } ++ ++ return -EPERM; ++} ++ ++static int yaffs_symlink(struct inode *dir, struct dentry *dentry, ++ const char *symname) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ uid_t uid = YCRED_FSUID(); ++ gid_t gid = ++ (dir->i_mode & S_ISGID) ? i_gid_read(dir) : YCRED_FSGID(); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_symlink"); ++ ++ if (strnlen(dentry->d_name.name, YAFFS_MAX_NAME_LENGTH + 1) > ++ YAFFS_MAX_NAME_LENGTH) ++ return -ENAMETOOLONG; ++ ++ if (strnlen(symname, YAFFS_MAX_ALIAS_LENGTH + 1) > ++ YAFFS_MAX_ALIAS_LENGTH) ++ return -ENAMETOOLONG; ++ ++ dev = yaffs_inode_to_obj(dir)->my_dev; ++ yaffs_gross_lock(dev); ++ obj = yaffs_create_symlink(yaffs_inode_to_obj(dir), dentry->d_name.name, ++ S_IFLNK | S_IRWXUGO, uid, gid, symname); ++ yaffs_gross_unlock(dev); ++ ++ if (obj) { ++ struct inode *inode; ++ ++ inode = yaffs_get_inode(dir->i_sb, obj->yst_mode, 0, obj); ++ d_instantiate(dentry, inode); ++ update_dir_time(dir); ++ yaffs_trace(YAFFS_TRACE_OS, "symlink created OK"); ++ return 0; ++ } else { ++ yaffs_trace(YAFFS_TRACE_OS, "symlink not created"); ++ } ++ ++ return -ENOMEM; ++} ++ ++/* ++ * The VFS layer already does all the dentry stuff for rename. ++ * ++ * NB: POSIX says you can rename an object over an old object of the same name ++ */ ++static int yaffs_rename(struct inode *old_dir, struct dentry *old_dentry, ++ struct inode *new_dir, struct dentry *new_dentry) ++{ ++ struct yaffs_dev *dev; ++ int ret_val = YAFFS_FAIL; ++ struct yaffs_obj *target; ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_rename"); ++ dev = yaffs_inode_to_obj(old_dir)->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ /* Check if the target is an existing directory that is not empty. */ ++ target = yaffs_find_by_name(yaffs_inode_to_obj(new_dir), ++ new_dentry->d_name.name); ++ ++ if (target && target->variant_type == YAFFS_OBJECT_TYPE_DIRECTORY && ++ !list_empty(&target->variant.dir_variant.children)) { ++ ++ yaffs_trace(YAFFS_TRACE_OS, "target is non-empty dir"); ++ ++ ret_val = YAFFS_FAIL; ++ } else { ++ /* Now does unlinking internally using shadowing mechanism */ ++ yaffs_trace(YAFFS_TRACE_OS, "calling yaffs_rename_obj"); ++ ++ ret_val = yaffs_rename_obj(yaffs_inode_to_obj(old_dir), ++ old_dentry->d_name.name, ++ yaffs_inode_to_obj(new_dir), ++ new_dentry->d_name.name); ++ } ++ yaffs_gross_unlock(dev); ++ ++ if (ret_val == YAFFS_OK) { ++ if (target) ++ inode_dec_link_count(new_dentry->d_inode); ++ ++ update_dir_time(old_dir); ++ if (old_dir != new_dir) ++ update_dir_time(new_dir); ++ return 0; ++ } else { ++ return -ENOTEMPTY; ++ } ++} ++ ++ ++ ++ ++static int yaffs_unlink(struct inode *dir, struct dentry *dentry) ++{ ++ int ret_val; ++ ++ struct yaffs_dev *dev; ++ struct yaffs_obj *obj; ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_unlink %d:%s", ++ (int)(dir->i_ino), dentry->d_name.name); ++ obj = yaffs_inode_to_obj(dir); ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ ret_val = yaffs_unlinker(obj, dentry->d_name.name); ++ ++ if (ret_val == YAFFS_OK) { ++ inode_dec_link_count(dentry->d_inode); ++ dir->i_version++; ++ yaffs_gross_unlock(dev); ++ update_dir_time(dir); ++ return 0; ++ } ++ yaffs_gross_unlock(dev); ++ return -ENOTEMPTY; ++} ++ ++ ++ ++static const struct inode_operations yaffs_dir_inode_operations = { ++ .create = yaffs_create, ++ .lookup = yaffs_lookup, ++ .link = yaffs_link, ++ .unlink = yaffs_unlink, ++ .symlink = yaffs_symlink, ++ .mkdir = yaffs_mkdir, ++ .rmdir = yaffs_unlink, ++ .mknod = yaffs_mknod, ++ .rename = yaffs_rename, ++ .setattr = yaffs_setattr, ++ .setxattr = yaffs_setxattr, ++ .getxattr = yaffs_getxattr, ++ .listxattr = yaffs_listxattr, ++ .removexattr = yaffs_removexattr, ++}; ++ ++/*-----------------------------------------------------------------*/ ++/* Directory search context allows us to unlock access to yaffs during ++ * filldir without causing problems with the directory being modified. ++ * This is similar to the tried and tested mechanism used in yaffs direct. ++ * ++ * A search context iterates along a doubly linked list of siblings in the ++ * directory. If the iterating object is deleted then this would corrupt ++ * the list iteration, likely causing a crash. The search context avoids ++ * this by using the remove_obj_fn to move the search context to the ++ * next object before the object is deleted. ++ * ++ * Many readdirs (and thus seach conexts) may be alive simulateously so ++ * each struct yaffs_dev has a list of these. ++ * ++ * A seach context lives for the duration of a readdir. ++ * ++ * All these functions must be called while yaffs is locked. ++ */ ++ ++struct yaffs_search_context { ++ struct yaffs_dev *dev; ++ struct yaffs_obj *dir_obj; ++ struct yaffs_obj *next_return; ++ struct list_head others; ++}; ++ ++/* ++ * yaffs_new_search() creates a new search context, initialises it and ++ * adds it to the device's search context list. ++ * ++ * Called at start of readdir. ++ */ ++static struct yaffs_search_context *yaffs_new_search(struct yaffs_obj *dir) ++{ ++ struct yaffs_dev *dev = dir->my_dev; ++ struct yaffs_search_context *sc = ++ kmalloc(sizeof(struct yaffs_search_context), GFP_NOFS); ++ if (sc) { ++ sc->dir_obj = dir; ++ sc->dev = dev; ++ if (list_empty(&sc->dir_obj->variant.dir_variant.children)) ++ sc->next_return = NULL; ++ else ++ sc->next_return = ++ list_entry(dir->variant.dir_variant.children.next, ++ struct yaffs_obj, siblings); ++ INIT_LIST_HEAD(&sc->others); ++ list_add(&sc->others, &(yaffs_dev_to_lc(dev)->search_contexts)); ++ } ++ return sc; ++} ++ ++/* ++ * yaffs_search_end() disposes of a search context and cleans up. ++ */ ++static void yaffs_search_end(struct yaffs_search_context *sc) ++{ ++ if (sc) { ++ list_del(&sc->others); ++ kfree(sc); ++ } ++} ++ ++/* ++ * yaffs_search_advance() moves a search context to the next object. ++ * Called when the search iterates or when an object removal causes ++ * the search context to be moved to the next object. ++ */ ++static void yaffs_search_advance(struct yaffs_search_context *sc) ++{ ++ if (!sc) ++ return; ++ ++ if (sc->next_return == NULL || ++ list_empty(&sc->dir_obj->variant.dir_variant.children)) ++ sc->next_return = NULL; ++ else { ++ struct list_head *next = sc->next_return->siblings.next; ++ ++ if (next == &sc->dir_obj->variant.dir_variant.children) ++ sc->next_return = NULL; /* end of list */ ++ else ++ sc->next_return = ++ list_entry(next, struct yaffs_obj, siblings); ++ } ++} ++ ++/* ++ * yaffs_remove_obj_callback() is called when an object is unlinked. ++ * We check open search contexts and advance any which are currently ++ * on the object being iterated. ++ */ ++static void yaffs_remove_obj_callback(struct yaffs_obj *obj) ++{ ++ ++ struct list_head *i; ++ struct yaffs_search_context *sc; ++ struct list_head *search_contexts = ++ &(yaffs_dev_to_lc(obj->my_dev)->search_contexts); ++ ++ /* Iterate through the directory search contexts. ++ * If any are currently on the object being removed, then advance ++ * the search context to the next object to prevent a hanging pointer. ++ */ ++ list_for_each(i, search_contexts) { ++ sc = list_entry(i, struct yaffs_search_context, others); ++ if (sc->next_return == obj) ++ yaffs_search_advance(sc); ++ } ++ ++} ++ ++ ++/*-----------------------------------------------------------------*/ ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 12, 0)) ++static int yaffs_readdir(struct file *file, struct dir_context *ctx) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ struct yaffs_search_context *sc; ++ struct inode *inode = file->f_dentry->d_inode; ++ unsigned long offset, curoffs; ++ struct yaffs_obj *l; ++ int ret_val = 0; ++ ++ char name[YAFFS_MAX_NAME_LENGTH + 1]; ++ ++ obj = yaffs_dentry_to_obj(file->f_dentry); ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ yaffs_dev_to_lc(dev)->readdir_process = current; ++ ++ offset = ctx->pos; ++ ++ sc = yaffs_new_search(obj); ++ if (!sc) { ++ ret_val = -ENOMEM; ++ goto out; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: starting at %d", (int)offset); ++ ++ if (offset == 0) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: entry . ino %d", ++ (int)inode->i_ino); ++ yaffs_gross_unlock(dev); ++ if (!dir_emit_dot(file, ctx)) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ yaffs_gross_lock(dev); ++ offset++; ++ ctx->pos++; ++ } ++ if (offset == 1) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: entry .. ino %d", ++ (int)file->f_dentry->d_parent->d_inode->i_ino); ++ yaffs_gross_unlock(dev); ++ if (!dir_emit_dotdot(file, ctx)) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ yaffs_gross_lock(dev); ++ offset++; ++ ctx->pos++; ++ } ++ ++ curoffs = 1; ++ ++ /* If the directory has changed since the open or last call to ++ readdir, rewind to after the 2 canned entries. */ ++ if (file->f_version != inode->i_version) { ++ offset = 2; ++ ctx->pos = offset; ++ file->f_version = inode->i_version; ++ } ++ ++ while (sc->next_return) { ++ curoffs++; ++ l = sc->next_return; ++ if (curoffs >= offset) { ++ int this_inode = yaffs_get_obj_inode(l); ++ int this_type = yaffs_get_obj_type(l); ++ ++ yaffs_get_obj_name(l, name, YAFFS_MAX_NAME_LENGTH + 1); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: %s inode %d", ++ name, yaffs_get_obj_inode(l)); ++ ++ yaffs_gross_unlock(dev); ++ ++ if (!dir_emit(ctx, name, strlen(name), ++ this_inode, this_type) < 0) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ ++ yaffs_gross_lock(dev); ++ ++ offset++; ++ ctx->pos++; ++ } ++ yaffs_search_advance(sc); ++ } ++ ++out: ++ yaffs_search_end(sc); ++ yaffs_dev_to_lc(dev)->readdir_process = NULL; ++ yaffs_gross_unlock(dev); ++ ++ return ret_val; ++} ++#else ++static int yaffs_readdir(struct file *f, void *dirent, filldir_t filldir) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ struct yaffs_search_context *sc; ++ struct inode *inode = f->f_dentry->d_inode; ++ unsigned long offset, curoffs; ++ struct yaffs_obj *l; ++ int ret_val = 0; ++ ++ char name[YAFFS_MAX_NAME_LENGTH + 1]; ++ ++ obj = yaffs_dentry_to_obj(f->f_dentry); ++ dev = obj->my_dev; ++ ++ yaffs_gross_lock(dev); ++ ++ yaffs_dev_to_lc(dev)->readdir_process = current; ++ ++ offset = f->f_pos; ++ ++ sc = yaffs_new_search(obj); ++ if (!sc) { ++ ret_val = -ENOMEM; ++ goto out; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: starting at %d", (int)offset); ++ ++ if (offset == 0) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: entry . ino %d", ++ (int)inode->i_ino); ++ yaffs_gross_unlock(dev); ++ if (filldir(dirent, ".", 1, offset, inode->i_ino, DT_DIR) < 0) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ yaffs_gross_lock(dev); ++ offset++; ++ f->f_pos++; ++ } ++ if (offset == 1) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: entry .. ino %d", ++ (int)f->f_dentry->d_parent->d_inode->i_ino); ++ yaffs_gross_unlock(dev); ++ if (filldir(dirent, "..", 2, offset, ++ f->f_dentry->d_parent->d_inode->i_ino, ++ DT_DIR) < 0) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ yaffs_gross_lock(dev); ++ offset++; ++ f->f_pos++; ++ } ++ ++ curoffs = 1; ++ ++ /* If the directory has changed since the open or last call to ++ readdir, rewind to after the 2 canned entries. */ ++ if (f->f_version != inode->i_version) { ++ offset = 2; ++ f->f_pos = offset; ++ f->f_version = inode->i_version; ++ } ++ ++ while (sc->next_return) { ++ curoffs++; ++ l = sc->next_return; ++ if (curoffs >= offset) { ++ int this_inode = yaffs_get_obj_inode(l); ++ int this_type = yaffs_get_obj_type(l); ++ ++ yaffs_get_obj_name(l, name, YAFFS_MAX_NAME_LENGTH + 1); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_readdir: %s inode %d", ++ name, yaffs_get_obj_inode(l)); ++ ++ yaffs_gross_unlock(dev); ++ ++ if (filldir(dirent, ++ name, ++ strlen(name), ++ offset, this_inode, this_type) < 0) { ++ yaffs_gross_lock(dev); ++ goto out; ++ } ++ ++ yaffs_gross_lock(dev); ++ ++ offset++; ++ f->f_pos++; ++ } ++ yaffs_search_advance(sc); ++ } ++ ++out: ++ yaffs_search_end(sc); ++ yaffs_dev_to_lc(dev)->readdir_process = NULL; ++ yaffs_gross_unlock(dev); ++ ++ return ret_val; ++} ++#endif ++ ++static const struct file_operations yaffs_dir_operations = { ++ .read = generic_read_dir, ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3, 12, 0)) ++ .iterate = yaffs_readdir, ++#else ++ .readdir = yaffs_readdir, ++#endif ++ .fsync = yaffs_sync_object, ++ .llseek = generic_file_llseek, ++}; ++ ++static void yaffs_fill_inode_from_obj(struct inode *inode, ++ struct yaffs_obj *obj) ++{ ++ if (inode && obj) { ++ ++ /* Check mode against the variant type and attempt to repair if broken. */ ++ u32 mode = obj->yst_mode; ++ switch (obj->variant_type) { ++ case YAFFS_OBJECT_TYPE_FILE: ++ if (!S_ISREG(mode)) { ++ obj->yst_mode &= ~S_IFMT; ++ obj->yst_mode |= S_IFREG; ++ } ++ ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ if (!S_ISLNK(mode)) { ++ obj->yst_mode &= ~S_IFMT; ++ obj->yst_mode |= S_IFLNK; ++ } ++ ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ if (!S_ISDIR(mode)) { ++ obj->yst_mode &= ~S_IFMT; ++ obj->yst_mode |= S_IFDIR; ++ } ++ ++ break; ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ default: ++ /* TODO? */ ++ break; ++ } ++ ++ inode->i_flags |= S_NOATIME; ++ ++ inode->i_ino = obj->obj_id; ++ inode->i_mode = obj->yst_mode; ++ i_uid_write(inode, obj->yst_uid); ++ i_gid_write(inode, obj->yst_gid); ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 19)) ++ inode->i_blksize = inode->i_sb->s_blocksize; ++#endif ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ ++ inode->i_rdev = old_decode_dev(obj->yst_rdev); ++ inode->i_atime.tv_sec = (time_t) (obj->yst_atime); ++ inode->i_atime.tv_nsec = 0; ++ inode->i_mtime.tv_sec = (time_t) obj->yst_mtime; ++ inode->i_mtime.tv_nsec = 0; ++ inode->i_ctime.tv_sec = (time_t) obj->yst_ctime; ++ inode->i_ctime.tv_nsec = 0; ++#else ++ inode->i_rdev = obj->yst_rdev; ++ inode->i_atime = obj->yst_atime; ++ inode->i_mtime = obj->yst_mtime; ++ inode->i_ctime = obj->yst_ctime; ++#endif ++ inode->i_size = yaffs_get_obj_length(obj); ++ inode->i_blocks = (inode->i_size + 511) >> 9; ++ ++ set_nlink(inode, yaffs_get_obj_link_count(obj)); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_fill_inode mode %x uid %d gid %d size %lld count %d", ++ inode->i_mode, i_uid_read(inode), i_gid_read(inode), ++ inode->i_size, atomic_read(&inode->i_count)); ++ ++ switch (obj->yst_mode & S_IFMT) { ++ default: /* fifo, device or socket */ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ init_special_inode(inode, obj->yst_mode, ++ old_decode_dev(obj->yst_rdev)); ++#else ++ init_special_inode(inode, obj->yst_mode, ++ (dev_t) (obj->yst_rdev)); ++#endif ++ break; ++ case S_IFREG: /* file */ ++ inode->i_op = &yaffs_file_inode_operations; ++ inode->i_fop = &yaffs_file_operations; ++ inode->i_mapping->a_ops = ++ &yaffs_file_address_operations; ++ break; ++ case S_IFDIR: /* directory */ ++ inode->i_op = &yaffs_dir_inode_operations; ++ inode->i_fop = &yaffs_dir_operations; ++ break; ++ case S_IFLNK: /* symlink */ ++ inode->i_op = &yaffs_symlink_inode_operations; ++ break; ++ } ++ ++ yaffs_inode_to_obj_lv(inode) = obj; ++ ++ obj->my_inode = inode; ++ ++ } else { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_fill_inode invalid parameters"); ++ } ++ ++} ++ ++ ++ ++/* ++ * yaffs background thread functions . ++ * yaffs_bg_thread_fn() the thread function ++ * yaffs_bg_start() launches the background thread. ++ * yaffs_bg_stop() cleans up the background thread. ++ * ++ * NB: ++ * The thread should only run after the yaffs is initialised ++ * The thread should be stopped before yaffs is unmounted. ++ * The thread should not do any writing while the fs is in read only. ++ */ ++ ++static unsigned yaffs_bg_gc_urgency(struct yaffs_dev *dev) ++{ ++ unsigned erased_chunks = ++ dev->n_erased_blocks * dev->param.chunks_per_block; ++ struct yaffs_linux_context *context = yaffs_dev_to_lc(dev); ++ unsigned scattered = 0; /* Free chunks not in an erased block */ ++ ++ if (erased_chunks < dev->n_free_chunks) ++ scattered = (dev->n_free_chunks - erased_chunks); ++ ++ if (!context->bg_running) ++ return 0; ++ else if (scattered < (dev->param.chunks_per_block * 2)) ++ return 0; ++ else if (erased_chunks > dev->n_free_chunks / 2) ++ return 0; ++ else if (erased_chunks > dev->n_free_chunks / 4) ++ return 1; ++ else ++ return 2; ++} ++ ++#ifdef YAFFS_COMPILE_BACKGROUND ++ ++void yaffs_background_waker(unsigned long data) ++{ ++ wake_up_process((struct task_struct *)data); ++} ++ ++static int yaffs_bg_thread_fn(void *data) ++{ ++ struct yaffs_dev *dev = (struct yaffs_dev *)data; ++ struct yaffs_linux_context *context = yaffs_dev_to_lc(dev); ++ unsigned long now = jiffies; ++ unsigned long next_dir_update = now; ++ unsigned long next_gc = now; ++ unsigned long expires; ++ unsigned int urgency; ++ ++ int gc_result; ++ struct timer_list timer; ++ ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, ++ "yaffs_background starting for dev %p", (void *)dev); ++ ++#ifdef YAFFS_COMPILE_FREEZER ++ set_freezable(); ++#endif ++ while (context->bg_running) { ++ yaffs_trace(YAFFS_TRACE_BACKGROUND, "yaffs_background"); ++ ++ if (kthread_should_stop()) ++ break; ++ ++#ifdef YAFFS_COMPILE_FREEZER ++ if (try_to_freeze()) ++ continue; ++#endif ++ yaffs_gross_lock(dev); ++ ++ now = jiffies; ++ ++ if (time_after(now, next_dir_update) && yaffs_bg_enable) { ++ yaffs_update_dirty_dirs(dev); ++ next_dir_update = now + HZ; ++ } ++ ++ if (time_after(now, next_gc) && yaffs_bg_enable) { ++ if (!dev->is_checkpointed) { ++ urgency = yaffs_bg_gc_urgency(dev); ++ gc_result = yaffs_bg_gc(dev, urgency); ++ if (urgency > 1) ++ next_gc = now + HZ / 20 + 1; ++ else if (urgency > 0) ++ next_gc = now + HZ / 10 + 1; ++ else ++ next_gc = now + HZ * 2; ++ } else { ++ /* ++ * gc not running so set to next_dir_update ++ * to cut down on wake ups ++ */ ++ next_gc = next_dir_update; ++ } ++ } ++ yaffs_gross_unlock(dev); ++#if 1 ++ expires = next_dir_update; ++ if (time_before(next_gc, expires)) ++ expires = next_gc; ++ if (time_before(expires, now)) ++ expires = now + HZ; ++ ++ Y_INIT_TIMER(&timer); ++ timer.expires = expires + 1; ++ timer.data = (unsigned long)current; ++ timer.function = yaffs_background_waker; ++ ++ set_current_state(TASK_INTERRUPTIBLE); ++ add_timer(&timer); ++ schedule(); ++ del_timer_sync(&timer); ++#else ++ msleep(10); ++#endif ++ } ++ ++ return 0; ++} ++ ++static int yaffs_bg_start(struct yaffs_dev *dev) ++{ ++ int retval = 0; ++ struct yaffs_linux_context *context = yaffs_dev_to_lc(dev); ++ ++ if (dev->read_only) ++ return -1; ++ ++ context->bg_running = 1; ++ ++ context->bg_thread = kthread_run(yaffs_bg_thread_fn, ++ (void *)dev, "yaffs-bg-%d", ++ context->mount_id); ++ ++ if (IS_ERR(context->bg_thread)) { ++ retval = PTR_ERR(context->bg_thread); ++ context->bg_thread = NULL; ++ context->bg_running = 0; ++ } ++ return retval; ++} ++ ++static void yaffs_bg_stop(struct yaffs_dev *dev) ++{ ++ struct yaffs_linux_context *ctxt = yaffs_dev_to_lc(dev); ++ ++ ctxt->bg_running = 0; ++ ++ if (ctxt->bg_thread) { ++ kthread_stop(ctxt->bg_thread); ++ ctxt->bg_thread = NULL; ++ } ++} ++#else ++static int yaffs_bg_thread_fn(void *data) ++{ ++ return 0; ++} ++ ++static int yaffs_bg_start(struct yaffs_dev *dev) ++{ ++ return 0; ++} ++ ++static void yaffs_bg_stop(struct yaffs_dev *dev) ++{ ++} ++#endif ++ ++ ++static void yaffs_flush_inodes(struct super_block *sb) ++{ ++ struct inode *iptr; ++ struct yaffs_obj *obj; ++ ++ list_for_each_entry(iptr, &sb->s_inodes, i_sb_list) { ++ obj = yaffs_inode_to_obj(iptr); ++ if (obj) { ++ yaffs_trace(YAFFS_TRACE_OS, ++ "flushing obj %d", ++ obj->obj_id); ++ yaffs_flush_file(obj, 1, 0); ++ } ++ } ++} ++ ++static void yaffs_flush_super(struct super_block *sb, int do_checkpoint) ++{ ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++ if (!dev) ++ return; ++ ++ yaffs_flush_inodes(sb); ++ yaffs_update_dirty_dirs(dev); ++ yaffs_flush_whole_cache(dev); ++ if (do_checkpoint) ++ yaffs_checkpoint_save(dev); ++} ++ ++static LIST_HEAD(yaffs_context_list); ++struct mutex yaffs_context_lock; ++ ++static void yaffs_put_super(struct super_block *sb) ++{ ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++ struct mtd_info *mtd = yaffs_dev_to_mtd(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_ALWAYS, ++ "yaffs_put_super"); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_BACKGROUND, ++ "Shutting down yaffs background thread"); ++ yaffs_bg_stop(dev); ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_BACKGROUND, ++ "yaffs background thread shut down"); ++ ++ yaffs_gross_lock(dev); ++ ++ yaffs_flush_super(sb, 1); ++ ++ yaffs_deinitialise(dev); ++ ++ yaffs_gross_unlock(dev); ++ ++ mutex_lock(&yaffs_context_lock); ++ list_del_init(&(yaffs_dev_to_lc(dev)->context_list)); ++ mutex_unlock(&yaffs_context_lock); ++ ++ if (yaffs_dev_to_lc(dev)->spare_buffer) { ++ kfree(yaffs_dev_to_lc(dev)->spare_buffer); ++ yaffs_dev_to_lc(dev)->spare_buffer = NULL; ++ } ++ ++ kfree(dev); ++ ++ yaffs_put_mtd_device(mtd); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_ALWAYS, ++ "yaffs_put_super done"); ++} ++ ++ ++static unsigned yaffs_gc_control_callback(struct yaffs_dev *dev) ++{ ++ return yaffs_gc_control; ++} ++ ++ ++#ifdef YAFFS_COMPILE_EXPORTFS ++ ++static struct inode *yaffs2_nfs_get_inode(struct super_block *sb, uint64_t ino, ++ uint32_t generation) ++{ ++ return Y_IGET(sb, ino); ++} ++ ++static struct dentry *yaffs2_fh_to_dentry(struct super_block *sb, ++ struct fid *fid, int fh_len, ++ int fh_type) ++{ ++ return generic_fh_to_dentry(sb, fid, fh_len, fh_type, ++ yaffs2_nfs_get_inode); ++} ++ ++static struct dentry *yaffs2_fh_to_parent(struct super_block *sb, ++ struct fid *fid, int fh_len, ++ int fh_type) ++{ ++ return generic_fh_to_parent(sb, fid, fh_len, fh_type, ++ yaffs2_nfs_get_inode); ++} ++ ++struct dentry *yaffs2_get_parent(struct dentry *dentry) ++{ ++ ++ struct super_block *sb = dentry->d_inode->i_sb; ++ struct dentry *parent = ERR_PTR(-ENOENT); ++ struct inode *inode; ++ unsigned long parent_ino; ++ struct yaffs_obj *d_obj; ++ struct yaffs_obj *parent_obj; ++ ++ d_obj = yaffs_inode_to_obj(dentry->d_inode); ++ ++ if (d_obj) { ++ parent_obj = d_obj->parent; ++ if (parent_obj) { ++ parent_ino = yaffs_get_obj_inode(parent_obj); ++ inode = Y_IGET(sb, parent_ino); ++ ++ if (IS_ERR(inode)) { ++ parent = ERR_CAST(inode); ++ } else { ++ parent = d_obtain_alias(inode); ++ if (!IS_ERR(parent)) { ++ parent = ERR_PTR(-ENOMEM); ++ iput(inode); ++ } ++ } ++ } ++ } ++ ++ return parent; ++} ++ ++/* Just declare a zero structure as a NULL value implies ++ * using the default functions of exportfs. ++ */ ++ ++static struct export_operations yaffs_export_ops = { ++ .fh_to_dentry = yaffs2_fh_to_dentry, ++ .fh_to_parent = yaffs2_fh_to_parent, ++ .get_parent = yaffs2_get_parent, ++}; ++ ++#endif ++ ++static void yaffs_unstitch_obj(struct inode *inode, struct yaffs_obj *obj) ++{ ++ /* Clear the association between the inode and ++ * the struct yaffs_obj. ++ */ ++ obj->my_inode = NULL; ++ yaffs_inode_to_obj_lv(inode) = NULL; ++ ++ /* If the object freeing was deferred, then the real ++ * free happens now. ++ * This should fix the inode inconsistency problem. ++ */ ++ yaffs_handle_defered_free(obj); ++} ++ ++#ifdef YAFFS_HAS_EVICT_INODE ++/* yaffs_evict_inode combines into one operation what was previously done in ++ * yaffs_clear_inode() and yaffs_delete_inode() ++ * ++ */ ++static void yaffs_evict_inode(struct inode *inode) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ int deleteme = 0; ++ ++ obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_evict_inode: ino %d, count %d %s", ++ (int)inode->i_ino, atomic_read(&inode->i_count), ++ obj ? "object exists" : "null object"); ++ ++ if (!inode->i_nlink && !is_bad_inode(inode)) ++ deleteme = 1; ++ truncate_inode_pages(&inode->i_data, 0); ++ Y_CLEAR_INODE(inode); ++ ++ if (deleteme && obj) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ yaffs_del_obj(obj); ++ yaffs_gross_unlock(dev); ++ } ++ if (obj) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ yaffs_unstitch_obj(inode, obj); ++ yaffs_gross_unlock(dev); ++ } ++} ++#else ++ ++/* clear is called to tell the fs to release any per-inode data it holds. ++ * The object might still exist on disk and is just being thrown out of the cache ++ * or else the object has actually been deleted and we're being called via ++ * the chain ++ * yaffs_delete_inode() -> clear_inode()->yaffs_clear_inode() ++ */ ++ ++static void yaffs_clear_inode(struct inode *inode) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_dev *dev; ++ ++ obj = yaffs_inode_to_obj(inode); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_clear_inode: ino %d, count %d %s", ++ (int)inode->i_ino, atomic_read(&inode->i_count), ++ obj ? "object exists" : "null object"); ++ ++ if (obj) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ yaffs_unstitch_obj(inode, obj); ++ yaffs_gross_unlock(dev); ++ } ++ ++} ++ ++/* delete is called when the link count is zero and the inode ++ * is put (ie. nobody wants to know about it anymore, time to ++ * delete the file). ++ * NB Must call clear_inode() ++ */ ++static void yaffs_delete_inode(struct inode *inode) ++{ ++ struct yaffs_obj *obj = yaffs_inode_to_obj(inode); ++ struct yaffs_dev *dev; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_delete_inode: ino %d, count %d %s", ++ (int)inode->i_ino, atomic_read(&inode->i_count), ++ obj ? "object exists" : "null object"); ++ ++ if (obj) { ++ dev = obj->my_dev; ++ yaffs_gross_lock(dev); ++ yaffs_del_obj(obj); ++ yaffs_gross_unlock(dev); ++ } ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 13)) ++ truncate_inode_pages(&inode->i_data, 0); ++#endif ++ clear_inode(inode); ++} ++#endif ++ ++ ++ ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static int yaffs_statfs(struct dentry *dentry, struct kstatfs *buf) ++{ ++ struct yaffs_dev *dev = yaffs_dentry_to_obj(dentry)->my_dev; ++ struct super_block *sb = dentry->d_sb; ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs_statfs(struct super_block *sb, struct kstatfs *buf) ++{ ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++#else ++static int yaffs_statfs(struct super_block *sb, struct statfs *buf) ++{ ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++#endif ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_statfs"); ++ ++ yaffs_gross_lock(dev); ++ ++ buf->f_type = YAFFS_MAGIC; ++ buf->f_bsize = sb->s_blocksize; ++ buf->f_namelen = 255; ++ ++ if (dev->data_bytes_per_chunk & (dev->data_bytes_per_chunk - 1)) { ++ /* Do this if chunk size is not a power of 2 */ ++ ++ uint64_t bytes_in_dev; ++ uint64_t bytes_free; ++ ++ bytes_in_dev = ++ ((uint64_t) ++ ((dev->param.end_block - dev->param.start_block + ++ 1))) * ((uint64_t) (dev->param.chunks_per_block * ++ dev->data_bytes_per_chunk)); ++ ++ do_div(bytes_in_dev, sb->s_blocksize); /* bytes_in_dev becomes the number of blocks */ ++ buf->f_blocks = bytes_in_dev; ++ ++ bytes_free = ((uint64_t) (yaffs_get_n_free_chunks(dev))) * ++ ((uint64_t) (dev->data_bytes_per_chunk)); ++ ++ do_div(bytes_free, sb->s_blocksize); ++ ++ buf->f_bfree = bytes_free; ++ ++ } else if (sb->s_blocksize > dev->data_bytes_per_chunk) { ++ ++ buf->f_blocks = ++ (dev->param.end_block - dev->param.start_block + 1) * ++ dev->param.chunks_per_block / ++ (sb->s_blocksize / dev->data_bytes_per_chunk); ++ buf->f_bfree = ++ yaffs_get_n_free_chunks(dev) / ++ (sb->s_blocksize / dev->data_bytes_per_chunk); ++ } else { ++ buf->f_blocks = ++ (dev->param.end_block - dev->param.start_block + 1) * ++ dev->param.chunks_per_block * ++ (dev->data_bytes_per_chunk / sb->s_blocksize); ++ ++ buf->f_bfree = ++ yaffs_get_n_free_chunks(dev) * ++ (dev->data_bytes_per_chunk / sb->s_blocksize); ++ } ++ ++ buf->f_files = 0; ++ buf->f_ffree = 0; ++ buf->f_bavail = buf->f_bfree; ++ ++ yaffs_gross_unlock(dev); ++ return 0; ++} ++ ++ ++ ++static int yaffs_do_sync_fs(struct super_block *sb, int request_checkpoint) ++{ ++ ++ struct yaffs_dev *dev = yaffs_super_to_dev(sb); ++ unsigned int oneshot_checkpoint = (yaffs_auto_checkpoint & 4); ++ unsigned gc_urgent = yaffs_bg_gc_urgency(dev); ++ int do_checkpoint; ++ int dirty = yaffs_check_super_dirty(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_SYNC | YAFFS_TRACE_BACKGROUND, ++ "yaffs_do_sync_fs: gc-urgency %d %s %s%s", ++ gc_urgent, ++ dirty ? "dirty" : "clean", ++ request_checkpoint ? "checkpoint requested" : "no checkpoint", ++ oneshot_checkpoint ? " one-shot" : ""); ++ ++ yaffs_gross_lock(dev); ++ do_checkpoint = ((request_checkpoint && !gc_urgent) || ++ oneshot_checkpoint) && !dev->is_checkpointed; ++ ++ if (dirty || do_checkpoint) { ++ yaffs_flush_super(sb, !dev->is_checkpointed && do_checkpoint); ++ yaffs_clear_super_dirty(dev); ++ if (oneshot_checkpoint) ++ yaffs_auto_checkpoint &= ~4; ++ } ++ yaffs_gross_unlock(dev); ++ ++ return 0; ++} ++ ++ ++#ifdef YAFFS_HAS_WRITE_SUPER ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static void yaffs_write_super(struct super_block *sb) ++#else ++static int yaffs_write_super(struct super_block *sb) ++#endif ++{ ++ unsigned request_checkpoint = (yaffs_auto_checkpoint >= 2); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_SYNC | YAFFS_TRACE_BACKGROUND, ++ "yaffs_write_super %s", ++ request_checkpoint ? " checkpt" : ""); ++ ++ yaffs_do_sync_fs(sb, request_checkpoint); ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 18)) ++ return 0; ++#endif ++} ++#endif ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static int yaffs_sync_fs(struct super_block *sb, int wait) ++#else ++static int yaffs_sync_fs(struct super_block *sb) ++#endif ++{ ++ unsigned request_checkpoint = (yaffs_auto_checkpoint >= 1); ++ ++ yaffs_trace(YAFFS_TRACE_OS | YAFFS_TRACE_SYNC, ++ "yaffs_sync_fs%s", request_checkpoint ? " checkpt" : ""); ++ ++ yaffs_do_sync_fs(sb, request_checkpoint); ++ ++ return 0; ++} ++ ++ ++ ++static const struct super_operations yaffs_super_ops = { ++ .statfs = yaffs_statfs, ++ ++#ifndef YAFFS_USE_OWN_IGET ++ .read_inode = yaffs_read_inode, ++#endif ++#ifdef YAFFS_HAS_PUT_INODE ++ .put_inode = yaffs_put_inode, ++#endif ++ .put_super = yaffs_put_super, ++#ifdef YAFFS_HAS_EVICT_INODE ++ .evict_inode = yaffs_evict_inode, ++#else ++ .delete_inode = yaffs_delete_inode, ++ .clear_inode = yaffs_clear_inode, ++#endif ++ .sync_fs = yaffs_sync_fs, ++#ifdef YAFFS_HAS_WRITE_SUPER ++ .write_super = yaffs_write_super, ++#endif ++}; ++ ++struct yaffs_options { ++ int inband_tags; ++ int skip_checkpoint_read; ++ int skip_checkpoint_write; ++ int no_cache; ++ int tags_ecc_on; ++ int tags_ecc_overridden; ++ int lazy_loading_enabled; ++ int lazy_loading_overridden; ++ int empty_lost_and_found; ++ int empty_lost_and_found_overridden; ++ int disable_summary; ++}; ++ ++#define MAX_OPT_LEN 30 ++static int yaffs_parse_options(struct yaffs_options *options, ++ const char *options_str) ++{ ++ char cur_opt[MAX_OPT_LEN + 1]; ++ int p; ++ int error = 0; ++ ++ /* Parse through the options which is a comma seperated list */ ++ ++ while (options_str && *options_str && !error) { ++ memset(cur_opt, 0, MAX_OPT_LEN + 1); ++ p = 0; ++ ++ while (*options_str == ',') ++ options_str++; ++ ++ while (*options_str && *options_str != ',') { ++ if (p < MAX_OPT_LEN) { ++ cur_opt[p] = *options_str; ++ p++; ++ } ++ options_str++; ++ } ++ ++ if (!strcmp(cur_opt, "inband-tags")) { ++ options->inband_tags = 1; ++ } else if (!strcmp(cur_opt, "tags-ecc-off")) { ++ options->tags_ecc_on = 0; ++ options->tags_ecc_overridden = 1; ++ } else if (!strcmp(cur_opt, "tags-ecc-on")) { ++ options->tags_ecc_on = 1; ++ options->tags_ecc_overridden = 1; ++ } else if (!strcmp(cur_opt, "lazy-loading-off")) { ++ options->lazy_loading_enabled = 0; ++ options->lazy_loading_overridden = 1; ++ } else if (!strcmp(cur_opt, "lazy-loading-on")) { ++ options->lazy_loading_enabled = 1; ++ options->lazy_loading_overridden = 1; ++ } else if (!strcmp(cur_opt, "disable-summary")) { ++ options->disable_summary = 1; ++ } else if (!strcmp(cur_opt, "empty-lost-and-found-off")) { ++ options->empty_lost_and_found = 0; ++ options->empty_lost_and_found_overridden = 1; ++ } else if (!strcmp(cur_opt, "empty-lost-and-found-on")) { ++ options->empty_lost_and_found = 1; ++ options->empty_lost_and_found_overridden = 1; ++ } else if (!strcmp(cur_opt, "no-cache")) { ++ options->no_cache = 1; ++ } else if (!strcmp(cur_opt, "no-checkpoint-read")) { ++ options->skip_checkpoint_read = 1; ++ } else if (!strcmp(cur_opt, "no-checkpoint-write")) { ++ options->skip_checkpoint_write = 1; ++ } else if (!strcmp(cur_opt, "no-checkpoint")) { ++ options->skip_checkpoint_read = 1; ++ options->skip_checkpoint_write = 1; ++ } else { ++ printk(KERN_INFO "yaffs: Bad mount option \"%s\"\n", ++ cur_opt); ++ error = 1; ++ } ++ } ++ ++ return error; ++} ++ ++ ++static struct dentry *yaffs_make_root(struct inode *inode) ++{ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 4, 0)) ++ struct dentry *root = d_alloc_root(inode); ++ ++ if (!root) ++ iput(inode); ++ ++ return root; ++#else ++ return d_make_root(inode); ++#endif ++} ++ ++ ++ ++ ++static struct super_block *yaffs_internal_read_super(int yaffs_version, ++ struct super_block *sb, ++ void *data, int silent) ++{ ++ int n_blocks; ++ struct inode *inode = NULL; ++ struct dentry *root; ++ struct yaffs_dev *dev = 0; ++ char devname_buf[BDEVNAME_SIZE + 1]; ++ struct mtd_info *mtd; ++ int err; ++ char *data_str = (char *)data; ++ struct yaffs_linux_context *context = NULL; ++ struct yaffs_param *param; ++ ++ int read_only = 0; ++ int inband_tags = 0; ++ ++ struct yaffs_options options; ++ ++ unsigned mount_id; ++ int found; ++ struct yaffs_linux_context *context_iterator; ++ struct list_head *l; ++ ++ if (!sb) { ++ printk(KERN_INFO "yaffs: sb is NULL\n"); ++ return NULL; ++ } ++ ++ sb->s_magic = YAFFS_MAGIC; ++ sb->s_op = &yaffs_super_ops; ++ sb->s_flags |= MS_NOATIME; ++ ++ read_only = ((sb->s_flags & MS_RDONLY) != 0); ++ ++#ifdef YAFFS_COMPILE_EXPORTFS ++ sb->s_export_op = &yaffs_export_ops; ++#endif ++ ++ if (!sb->s_dev) ++ printk(KERN_INFO "yaffs: sb->s_dev is NULL\n"); ++ else if (!yaffs_devname(sb, devname_buf)) ++ printk(KERN_INFO "yaffs: devname is NULL\n"); ++ else ++ printk(KERN_INFO "yaffs: dev is %d name is \"%s\" %s\n", ++ sb->s_dev, ++ yaffs_devname(sb, devname_buf), read_only ? "ro" : "rw"); ++ ++ if (!data_str) ++ data_str = ""; ++ ++ printk(KERN_INFO "yaffs: passed flags \"%s\"\n", data_str); ++ ++ memset(&options, 0, sizeof(options)); ++ ++ if (yaffs_parse_options(&options, data_str)) { ++ /* Option parsing failed */ ++ return NULL; ++ } ++ ++ sb->s_blocksize = PAGE_CACHE_SIZE; ++ sb->s_blocksize_bits = PAGE_CACHE_SHIFT; ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_read_super: Using yaffs%d", yaffs_version); ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_read_super: block size %d", (int)(sb->s_blocksize)); ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: Attempting MTD mount of %u.%u,\"%s\"", ++ MAJOR(sb->s_dev), MINOR(sb->s_dev), ++ yaffs_devname(sb, devname_buf)); ++ ++ /* Get the device */ ++ mtd = get_mtd_device(NULL, MINOR(sb->s_dev)); ++ if (IS_ERR(mtd)) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs: MTD device %u either not valid or unavailable", ++ MINOR(sb->s_dev)); ++ return NULL; ++ } ++ ++ if (yaffs_auto_select && yaffs_version == 1 && WRITE_SIZE(mtd) >= 2048) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "auto selecting yaffs2"); ++ yaffs_version = 2; ++ } ++ ++ /* Added NCB 26/5/2006 for completeness */ ++ if (yaffs_version == 2 && !options.inband_tags ++ && WRITE_SIZE(mtd) == 512) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "auto selecting yaffs1"); ++ yaffs_version = 1; ++ } ++ ++ if (mtd->oobavail < sizeof(struct yaffs_packed_tags2) || ++ options.inband_tags) ++ inband_tags = 1; ++ ++ if(yaffs_verify_mtd(mtd, yaffs_version, inband_tags) < 0) ++ return NULL; ++ ++ /* OK, so if we got here, we have an MTD that's NAND and looks ++ * like it has the right capabilities ++ * Set the struct yaffs_dev up for mtd ++ */ ++ ++ if (!read_only && !(mtd->flags & MTD_WRITEABLE)) { ++ read_only = 1; ++ printk(KERN_INFO ++ "yaffs: mtd is read only, setting superblock read only\n" ++ ); ++ sb->s_flags |= MS_RDONLY; ++ } ++ ++ dev = kmalloc(sizeof(struct yaffs_dev), GFP_KERNEL); ++ context = kmalloc(sizeof(struct yaffs_linux_context), GFP_KERNEL); ++ ++ if (!dev || !context) { ++ kfree(dev); ++ kfree(context); ++ dev = NULL; ++ context = NULL; ++ ++ /* Deep shit could not allocate device structure */ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs_read_super: Failed trying to allocate struct yaffs_dev." ++ ); ++ return NULL; ++ } ++ memset(dev, 0, sizeof(struct yaffs_dev)); ++ param = &(dev->param); ++ ++ memset(context, 0, sizeof(struct yaffs_linux_context)); ++ dev->os_context = context; ++ INIT_LIST_HEAD(&(context->context_list)); ++ context->dev = dev; ++ context->super = sb; ++ ++ dev->read_only = read_only; ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++ sb->s_fs_info = dev; ++#else ++ sb->u.generic_sbp = dev; ++#endif ++ ++ ++ dev->driver_context = mtd; ++ param->name = mtd->name; ++ ++ /* Set up the memory size parameters.... */ ++ ++ ++ param->n_reserved_blocks = 5; ++ param->n_caches = (options.no_cache) ? 0 : 10; ++ param->inband_tags = inband_tags; ++ ++ param->enable_xattr = 1; ++ if (options.lazy_loading_overridden) ++ param->disable_lazy_load = !options.lazy_loading_enabled; ++ ++ param->defered_dir_update = 1; ++ ++ if (options.tags_ecc_overridden) ++ param->no_tags_ecc = !options.tags_ecc_on; ++ ++ param->empty_lost_n_found = 1; ++ param->refresh_period = 500; ++ param->disable_summary = options.disable_summary; ++ ++ ++#ifdef CONFIG_YAFFS_DISABLE_BAD_BLOCK_MARKING ++ param->disable_bad_block_marking = 1; ++#endif ++ if (options.empty_lost_and_found_overridden) ++ param->empty_lost_n_found = options.empty_lost_and_found; ++ ++ /* ... and the functions. */ ++ if (yaffs_version == 2) { ++ param->is_yaffs2 = 1; ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++ param->total_bytes_per_chunk = mtd->writesize; ++ param->chunks_per_block = mtd->erasesize / mtd->writesize; ++#else ++ param->total_bytes_per_chunk = mtd->oobblock; ++ param->chunks_per_block = mtd->erasesize / mtd->oobblock; ++#endif ++ n_blocks = YCALCBLOCKS(mtd->size, mtd->erasesize); ++ ++ param->start_block = 0; ++ param->end_block = n_blocks - 1; ++ } else { ++ param->is_yaffs2 = 0; ++ n_blocks = YCALCBLOCKS(mtd->size, ++ YAFFS_CHUNKS_PER_BLOCK * YAFFS_BYTES_PER_CHUNK); ++ ++ param->chunks_per_block = YAFFS_CHUNKS_PER_BLOCK; ++ param->total_bytes_per_chunk = YAFFS_BYTES_PER_CHUNK; ++ } ++ ++ param->start_block = 0; ++ param->end_block = n_blocks - 1; ++ ++ yaffs_mtd_drv_install(dev); ++ ++ param->sb_dirty_fn = yaffs_set_super_dirty; ++ param->gc_control_fn = yaffs_gc_control_callback; ++ ++ yaffs_dev_to_lc(dev)->super = sb; ++ ++ param->use_nand_ecc = 1; ++ ++ param->skip_checkpt_rd = options.skip_checkpoint_read; ++ param->skip_checkpt_wr = options.skip_checkpoint_write; ++ ++ mutex_lock(&yaffs_context_lock); ++ /* Get a mount id */ ++ found = 0; ++ for (mount_id = 0; !found; mount_id++) { ++ found = 1; ++ list_for_each(l, &yaffs_context_list) { ++ context_iterator = ++ list_entry(l, struct yaffs_linux_context, ++ context_list); ++ if (context_iterator->mount_id == mount_id) ++ found = 0; ++ } ++ } ++ context->mount_id = mount_id; ++ ++ list_add_tail(&(yaffs_dev_to_lc(dev)->context_list), ++ &yaffs_context_list); ++ mutex_unlock(&yaffs_context_lock); ++ ++ /* Directory search handling... */ ++ INIT_LIST_HEAD(&(yaffs_dev_to_lc(dev)->search_contexts)); ++ param->remove_obj_fn = yaffs_remove_obj_callback; ++ ++ mutex_init(&(yaffs_dev_to_lc(dev)->gross_lock)); ++ ++ yaffs_gross_lock(dev); ++ ++ err = yaffs_guts_initialise(dev); ++ ++ yaffs_trace(YAFFS_TRACE_OS, ++ "yaffs_read_super: guts initialised %s", ++ (err == YAFFS_OK) ? "OK" : "FAILED"); ++ ++ if (err == YAFFS_OK) ++ yaffs_bg_start(dev); ++ ++ if (!context->bg_thread) ++ param->defered_dir_update = 0; ++ ++ sb->s_maxbytes = yaffs_max_file_size(dev); ++ ++ /* Release lock before yaffs_get_inode() */ ++ yaffs_gross_unlock(dev); ++ ++ /* Create root inode */ ++ if (err == YAFFS_OK) ++ inode = yaffs_get_inode(sb, S_IFDIR | 0755, 0, yaffs_root(dev)); ++ ++ if (!inode) ++ return NULL; ++ ++ inode->i_op = &yaffs_dir_inode_operations; ++ inode->i_fop = &yaffs_dir_operations; ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_read_super: got root inode"); ++ ++ root = yaffs_make_root(inode); ++ ++ if (!root) ++ return NULL; ++ ++ sb->s_root = root; ++ if(!dev->is_checkpointed) ++ yaffs_set_super_dirty(dev); ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs_read_super: is_checkpointed %d", ++ dev->is_checkpointed); ++ ++ yaffs_trace(YAFFS_TRACE_OS, "yaffs_read_super: done"); ++ return sb; ++} ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs_internal_read_super_mtd(struct super_block *sb, void *data, ++ int silent) ++{ ++ return yaffs_internal_read_super(1, sb, data, silent) ? 0 : -EINVAL; ++} ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 39)) ++static struct dentry *yaffs_mount(struct file_system_type *fs_type, int flags, ++ const char *dev_name, void *data) ++{ ++ return mount_bdev(fs_type, flags, dev_name, data, yaffs_internal_read_super_mtd); ++} ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static int yaffs_read_super(struct file_system_type *fs, ++ int flags, const char *dev_name, ++ void *data, struct vfsmount *mnt) ++{ ++ ++ return get_sb_bdev(fs, flags, dev_name, data, ++ yaffs_internal_read_super_mtd, mnt); ++} ++#else ++static struct super_block *yaffs_read_super(struct file_system_type *fs, ++ int flags, const char *dev_name, ++ void *data) ++{ ++ ++ return get_sb_bdev(fs, flags, dev_name, data, ++ yaffs_internal_read_super_mtd); ++} ++#endif ++ ++static struct file_system_type yaffs_fs_type = { ++ .owner = THIS_MODULE, ++ .name = "yaffs", ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 39)) ++ .mount = yaffs_mount, ++#else ++ .get_sb = yaffs_read_super, ++#endif ++ .kill_sb = kill_block_super, ++ .fs_flags = FS_REQUIRES_DEV, ++}; ++#else ++static struct super_block *yaffs_read_super(struct super_block *sb, void *data, ++ int silent) ++{ ++ return yaffs_internal_read_super(1, sb, data, silent); ++} ++ ++static DECLARE_FSTYPE(yaffs_fs_type, "yaffs", yaffs_read_super, ++ FS_REQUIRES_DEV); ++#endif ++ ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++static int yaffs2_internal_read_super_mtd(struct super_block *sb, void *data, ++ int silent) ++{ ++ return yaffs_internal_read_super(2, sb, data, silent) ? 0 : -EINVAL; ++} ++ ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 39)) ++static struct dentry *yaffs2_mount(struct file_system_type *fs_type, int flags, ++ const char *dev_name, void *data) ++{ ++ return mount_bdev(fs_type, flags, dev_name, data, yaffs2_internal_read_super_mtd); ++} ++#elif (LINUX_VERSION_CODE > KERNEL_VERSION(2, 6, 17)) ++static int yaffs2_read_super(struct file_system_type *fs, ++ int flags, const char *dev_name, void *data, ++ struct vfsmount *mnt) ++{ ++ return get_sb_bdev(fs, flags, dev_name, data, ++ yaffs2_internal_read_super_mtd, mnt); ++} ++#else ++static struct super_block *yaffs2_read_super(struct file_system_type *fs, ++ int flags, const char *dev_name, ++ void *data) ++{ ++ ++ return get_sb_bdev(fs, flags, dev_name, data, ++ yaffs2_internal_read_super_mtd); ++} ++#endif ++ ++static struct file_system_type yaffs2_fs_type = { ++ .owner = THIS_MODULE, ++ .name = "yaffs2", ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 39)) ++ .mount = yaffs2_mount, ++#else ++ .get_sb = yaffs2_read_super, ++#endif ++ .kill_sb = kill_block_super, ++ .fs_flags = FS_REQUIRES_DEV, ++}; ++#else ++static struct super_block *yaffs2_read_super(struct super_block *sb, ++ void *data, int silent) ++{ ++ return yaffs_internal_read_super(2, sb, data, silent); ++} ++ ++static DECLARE_FSTYPE(yaffs2_fs_type, "yaffs2", yaffs2_read_super, ++ FS_REQUIRES_DEV); ++#endif ++ ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 9, 0)) ++static struct proc_dir_entry *my_proc_entry; ++ ++static char *yaffs_dump_dev_part0(char *buf, struct yaffs_dev *dev) ++{ ++ struct yaffs_param *param = &dev->param; ++ int bs[10]; ++ ++ yaffs_count_blocks_by_state(dev,bs); ++ ++ buf += sprintf(buf, "start_block.......... %d\n", param->start_block); ++ buf += sprintf(buf, "end_block............ %d\n", param->end_block); ++ buf += sprintf(buf, "total_bytes_per_chunk %d\n", ++ param->total_bytes_per_chunk); ++ buf += sprintf(buf, "use_nand_ecc......... %d\n", param->use_nand_ecc); ++ buf += sprintf(buf, "no_tags_ecc.......... %d\n", param->no_tags_ecc); ++ buf += sprintf(buf, "is_yaffs2............ %d\n", param->is_yaffs2); ++ buf += sprintf(buf, "inband_tags.......... %d\n", param->inband_tags); ++ buf += sprintf(buf, "empty_lost_n_found... %d\n", ++ param->empty_lost_n_found); ++ buf += sprintf(buf, "disable_lazy_load.... %d\n", ++ param->disable_lazy_load); ++ buf += sprintf(buf, "disable_bad_block_mrk %d\n", ++ param->disable_bad_block_marking); ++ buf += sprintf(buf, "refresh_period....... %d\n", ++ param->refresh_period); ++ buf += sprintf(buf, "n_caches............. %d\n", param->n_caches); ++ buf += sprintf(buf, "n_reserved_blocks.... %d\n", ++ param->n_reserved_blocks); ++ buf += sprintf(buf, "always_check_erased.. %d\n", ++ param->always_check_erased); ++ buf += sprintf(buf, "\n"); ++ buf += sprintf(buf, "block count by state\n"); ++ buf += sprintf(buf, "0:%d 1:%d 2:%d 3:%d 4:%d\n", ++ bs[0], bs[1], bs[2], bs[3], bs[4]); ++ buf += sprintf(buf, "5:%d 6:%d 7:%d 8:%d 9:%d\n", ++ bs[5], bs[6], bs[7], bs[8], bs[9]); ++ ++ return buf; ++} ++ ++static char *yaffs_dump_dev_part1(char *buf, struct yaffs_dev *dev) ++{ ++ buf += sprintf(buf, "max file size....... %lld\n", ++ (long long) yaffs_max_file_size(dev)); ++ buf += sprintf(buf, "data_bytes_per_chunk. %d\n", ++ dev->data_bytes_per_chunk); ++ buf += sprintf(buf, "chunk_grp_bits....... %d\n", dev->chunk_grp_bits); ++ buf += sprintf(buf, "chunk_grp_size....... %d\n", dev->chunk_grp_size); ++ buf += sprintf(buf, "n_erased_blocks...... %d\n", dev->n_erased_blocks); ++ buf += sprintf(buf, "blocks_in_checkpt.... %d\n", ++ dev->blocks_in_checkpt); ++ buf += sprintf(buf, "\n"); ++ buf += sprintf(buf, "n_tnodes............. %d\n", dev->n_tnodes); ++ buf += sprintf(buf, "n_obj................ %d\n", dev->n_obj); ++ buf += sprintf(buf, "n_free_chunks........ %d\n", dev->n_free_chunks); ++ buf += sprintf(buf, "\n"); ++ buf += sprintf(buf, "n_page_writes........ %u\n", dev->n_page_writes); ++ buf += sprintf(buf, "n_page_reads......... %u\n", dev->n_page_reads); ++ buf += sprintf(buf, "n_erasures........... %u\n", dev->n_erasures); ++ buf += sprintf(buf, "n_gc_copies.......... %u\n", dev->n_gc_copies); ++ buf += sprintf(buf, "all_gcs.............. %u\n", dev->all_gcs); ++ buf += sprintf(buf, "passive_gc_count..... %u\n", ++ dev->passive_gc_count); ++ buf += sprintf(buf, "oldest_dirty_gc_count %u\n", ++ dev->oldest_dirty_gc_count); ++ buf += sprintf(buf, "n_gc_blocks.......... %u\n", dev->n_gc_blocks); ++ buf += sprintf(buf, "bg_gcs............... %u\n", dev->bg_gcs); ++ buf += sprintf(buf, "n_retried_writes..... %u\n", ++ dev->n_retried_writes); ++ buf += sprintf(buf, "n_retired_blocks..... %u\n", ++ dev->n_retired_blocks); ++ buf += sprintf(buf, "n_ecc_fixed.......... %u\n", dev->n_ecc_fixed); ++ buf += sprintf(buf, "n_ecc_unfixed........ %u\n", dev->n_ecc_unfixed); ++ buf += sprintf(buf, "n_tags_ecc_fixed..... %u\n", ++ dev->n_tags_ecc_fixed); ++ buf += sprintf(buf, "n_tags_ecc_unfixed... %u\n", ++ dev->n_tags_ecc_unfixed); ++ buf += sprintf(buf, "cache_hits........... %u\n", dev->cache_hits); ++ buf += sprintf(buf, "n_deleted_files...... %u\n", dev->n_deleted_files); ++ buf += sprintf(buf, "n_unlinked_files..... %u\n", ++ dev->n_unlinked_files); ++ buf += sprintf(buf, "refresh_count........ %u\n", dev->refresh_count); ++ buf += sprintf(buf, "n_bg_deletions....... %u\n", dev->n_bg_deletions); ++ buf += sprintf(buf, "tags_used............ %u\n", dev->tags_used); ++ buf += sprintf(buf, "summary_used......... %u\n", dev->summary_used); ++ ++ return buf; ++} ++ ++static int yaffs_proc_read(char *page, ++ char **start, ++ off_t offset, int count, int *eof, void *data) ++{ ++ struct list_head *item; ++ char *buf = page; ++ int step = offset; ++ int n = 0; ++ ++ /* Get proc_file_read() to step 'offset' by one on each sucessive call. ++ * We use 'offset' (*ppos) to indicate where we are in dev_list. ++ * This also assumes the user has posted a read buffer large ++ * enough to hold the complete output; but that's life in /proc. ++ */ ++ ++ *(int *)start = 1; ++ ++ /* Print header first */ ++ if (step == 0) ++ buf += ++ sprintf(buf, ++ "Multi-version YAFFS built:" __DATE__ " " __TIME__ ++ "\n"); ++ else if (step == 1) ++ buf += sprintf(buf, "\n"); ++ else { ++ step -= 2; ++ ++ mutex_lock(&yaffs_context_lock); ++ ++ /* Locate and print the Nth entry. Order N-squared but N is small. */ ++ list_for_each(item, &yaffs_context_list) { ++ struct yaffs_linux_context *dc = ++ list_entry(item, struct yaffs_linux_context, ++ context_list); ++ struct yaffs_dev *dev = dc->dev; ++ ++ if (n < (step & ~1)) { ++ n += 2; ++ continue; ++ } ++ if ((step & 1) == 0) { ++ buf += ++ sprintf(buf, "\nDevice %d \"%s\"\n", n, ++ dev->param.name); ++ buf = yaffs_dump_dev_part0(buf, dev); ++ } else { ++ buf = yaffs_dump_dev_part1(buf, dev); ++ } ++ ++ break; ++ } ++ mutex_unlock(&yaffs_context_lock); ++ } ++ ++ return buf - page < count ? buf - page : count; ++} ++ ++/** ++ * Set the verbosity of the warnings and error messages. ++ * ++ * Note that the names can only be a..z or _ with the current code. ++ */ ++ ++static struct { ++ char *mask_name; ++ unsigned mask_bitfield; ++} mask_flags[] = { ++ {"allocate", YAFFS_TRACE_ALLOCATE}, ++ {"always", YAFFS_TRACE_ALWAYS}, ++ {"background", YAFFS_TRACE_BACKGROUND}, ++ {"bad_blocks", YAFFS_TRACE_BAD_BLOCKS}, ++ {"buffers", YAFFS_TRACE_BUFFERS}, ++ {"bug", YAFFS_TRACE_BUG}, ++ {"checkpt", YAFFS_TRACE_CHECKPOINT}, ++ {"deletion", YAFFS_TRACE_DELETION}, ++ {"erase", YAFFS_TRACE_ERASE}, ++ {"error", YAFFS_TRACE_ERROR}, ++ {"gc_detail", YAFFS_TRACE_GC_DETAIL}, ++ {"gc", YAFFS_TRACE_GC}, ++ {"lock", YAFFS_TRACE_LOCK}, ++ {"mtd", YAFFS_TRACE_MTD}, ++ {"nandaccess", YAFFS_TRACE_NANDACCESS}, ++ {"os", YAFFS_TRACE_OS}, ++ {"scan_debug", YAFFS_TRACE_SCAN_DEBUG}, ++ {"scan", YAFFS_TRACE_SCAN}, ++ {"mount", YAFFS_TRACE_MOUNT}, ++ {"tracing", YAFFS_TRACE_TRACING}, ++ {"sync", YAFFS_TRACE_SYNC}, ++ {"write", YAFFS_TRACE_WRITE}, ++ {"verify", YAFFS_TRACE_VERIFY}, ++ {"verify_nand", YAFFS_TRACE_VERIFY_NAND}, ++ {"verify_full", YAFFS_TRACE_VERIFY_FULL}, ++ {"verify_all", YAFFS_TRACE_VERIFY_ALL}, ++ {"all", 0xffffffff}, ++ {"none", 0}, ++ {NULL, 0}, ++}; ++ ++#define MAX_MASK_NAME_LENGTH 40 ++static int yaffs_proc_write_trace_options(struct file *file, const char *buf, ++ unsigned long count, void *data) ++{ ++ unsigned rg = 0, mask_bitfield; ++ char *end; ++ char *mask_name; ++ const char *x; ++ char substring[MAX_MASK_NAME_LENGTH + 1]; ++ int i; ++ int done = 0; ++ int add, len = 0; ++ int pos = 0; ++ ++ rg = yaffs_trace_mask; ++ ++ while (!done && (pos < count)) { ++ done = 1; ++ while ((pos < count) && isspace(buf[pos])) ++ pos++; ++ ++ switch (buf[pos]) { ++ case '+': ++ case '-': ++ case '=': ++ add = buf[pos]; ++ pos++; ++ break; ++ ++ default: ++ add = ' '; ++ break; ++ } ++ mask_name = NULL; ++ ++ mask_bitfield = simple_strtoul(buf + pos, &end, 0); ++ ++ if (end > buf + pos) { ++ mask_name = "numeral"; ++ len = end - (buf + pos); ++ pos += len; ++ done = 0; ++ } else { ++ for (x = buf + pos, i = 0; ++ (*x == '_' || (*x >= 'a' && *x <= 'z')) && ++ i < MAX_MASK_NAME_LENGTH; x++, i++, pos++) ++ substring[i] = *x; ++ substring[i] = '\0'; ++ ++ for (i = 0; mask_flags[i].mask_name != NULL; i++) { ++ if (strcmp(substring, mask_flags[i].mask_name) ++ == 0) { ++ mask_name = mask_flags[i].mask_name; ++ mask_bitfield = ++ mask_flags[i].mask_bitfield; ++ done = 0; ++ break; ++ } ++ } ++ } ++ ++ if (mask_name != NULL) { ++ done = 0; ++ switch (add) { ++ case '-': ++ rg &= ~mask_bitfield; ++ break; ++ case '+': ++ rg |= mask_bitfield; ++ break; ++ case '=': ++ rg = mask_bitfield; ++ break; ++ default: ++ rg |= mask_bitfield; ++ break; ++ } ++ } ++ } ++ ++ yaffs_trace_mask = rg | YAFFS_TRACE_ALWAYS; ++ ++ printk(KERN_DEBUG "new trace = 0x%08X\n", yaffs_trace_mask); ++ ++ if (rg & YAFFS_TRACE_ALWAYS) { ++ for (i = 0; mask_flags[i].mask_name != NULL; i++) { ++ char flag; ++ flag = ((rg & mask_flags[i].mask_bitfield) == ++ mask_flags[i].mask_bitfield) ? '+' : '-'; ++ printk(KERN_DEBUG "%c%s\n", flag, ++ mask_flags[i].mask_name); ++ } ++ } ++ ++ return count; ++} ++ ++/* Debug strings are of the form: ++ * .bnnn print info on block n ++ * .cobjn,chunkn print nand chunk id for objn:chunkn ++ */ ++ ++static int yaffs_proc_debug_write(struct file *file, const char *buf, ++ unsigned long count, void *data) ++{ ++ ++ char str[100]; ++ char *p0; ++ char *p1; ++ long p1_val; ++ long p0_val; ++ char cmd; ++ struct list_head *item; ++ ++ memset(str, 0, sizeof(str)); ++ memcpy(str, buf, min(count, sizeof(str) -1)); ++ ++ cmd = str[1]; ++ ++ p0 = str + 2; ++ ++ p1 = p0; ++ ++ while (*p1 && *p1 != ',') { ++ p1++; ++ } ++ *p1 = '\0'; ++ p1++; ++ ++ p0_val = simple_strtol(p0, NULL, 0); ++ p1_val = simple_strtol(p1, NULL, 0); ++ ++ ++ mutex_lock(&yaffs_context_lock); ++ ++ /* Locate and print the Nth entry. Order N-squared but N is small. */ ++ list_for_each(item, &yaffs_context_list) { ++ struct yaffs_linux_context *dc = ++ list_entry(item, struct yaffs_linux_context, ++ context_list); ++ struct yaffs_dev *dev = dc->dev; ++ ++ if (cmd == 'b') { ++ struct yaffs_block_info *bi; ++ ++ bi = yaffs_get_block_info(dev,p0_val); ++ ++ if(bi) { ++ printk("Block %d: state %d, retire %d, use %d, seq %d\n", ++ (int)p0_val, bi->block_state, ++ bi->needs_retiring, bi->pages_in_use, ++ bi->seq_number); ++ } ++ } else if (cmd == 'c') { ++ struct yaffs_obj *obj; ++ int nand_chunk; ++ ++ obj = yaffs_find_by_number(dev, p0_val); ++ if (!obj) ++ printk("No obj %d\n", (int)p0_val); ++ else { ++ if(p1_val == 0) ++ nand_chunk = obj->hdr_chunk; ++ else ++ nand_chunk = ++ yaffs_find_chunk_in_file(obj, ++ p1_val, NULL); ++ printk("Nand chunk for %d:%d is %d\n", ++ (int)p0_val, (int)p1_val, nand_chunk); ++ } ++ } ++ } ++ ++ mutex_unlock(&yaffs_context_lock); ++ ++ return count; ++} ++ ++static int yaffs_proc_write(struct file *file, const char *buf, ++ unsigned long count, void *data) ++{ ++ if (buf[0] == '.') ++ return yaffs_proc_debug_write(file, buf, count, data); ++ return yaffs_proc_write_trace_options(file, buf, count, data); ++} ++#endif /* (LINUX_VERSION_CODE < KERNEL_VERSION(3, 9, 0)) */ ++ ++/* Stuff to handle installation of file systems */ ++struct file_system_to_install { ++ struct file_system_type *fst; ++ int installed; ++}; ++ ++static struct file_system_to_install fs_to_install[] = { ++ {&yaffs_fs_type, 0}, ++ {&yaffs2_fs_type, 0}, ++ {NULL, 0} ++}; ++ ++static int __init init_yaffs_fs(void) ++{ ++ int error = 0; ++ struct file_system_to_install *fsinst; ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs built " __DATE__ " " __TIME__ " Installing."); ++ ++ mutex_init(&yaffs_context_lock); ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 9, 0)) ++ /* Install the proc_fs entries */ ++ my_proc_entry = create_proc_entry("yaffs", ++ S_IRUGO | S_IFREG, YPROC_ROOT); ++ ++ if (my_proc_entry) { ++ my_proc_entry->write_proc = yaffs_proc_write; ++ my_proc_entry->read_proc = yaffs_proc_read; ++ my_proc_entry->data = NULL; ++ } else { ++ return -ENOMEM; ++ } ++#endif ++ ++ /* Now add the file system entries */ ++ ++ fsinst = fs_to_install; ++ ++ while (fsinst->fst && !error) { ++ error = register_filesystem(fsinst->fst); ++ if (!error) ++ fsinst->installed = 1; ++ fsinst++; ++ } ++ ++ /* Any errors? uninstall */ ++ if (error) { ++ fsinst = fs_to_install; ++ ++ while (fsinst->fst) { ++ if (fsinst->installed) { ++ unregister_filesystem(fsinst->fst); ++ fsinst->installed = 0; ++ } ++ fsinst++; ++ } ++ } ++ ++ return error; ++} ++ ++static void __exit exit_yaffs_fs(void) ++{ ++ ++ struct file_system_to_install *fsinst; ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "yaffs built " __DATE__ " " __TIME__ " removing."); ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(3, 9, 0)) ++ remove_proc_entry("yaffs", YPROC_ROOT); ++#endif ++ ++ fsinst = fs_to_install; ++ ++ while (fsinst->fst) { ++ if (fsinst->installed) { ++ unregister_filesystem(fsinst->fst); ++ fsinst->installed = 0; ++ } ++ fsinst++; ++ } ++} ++ ++module_init(init_yaffs_fs) ++ module_exit(exit_yaffs_fs) ++ ++ MODULE_DESCRIPTION("YAFFS2 - a NAND specific flash file system"); ++MODULE_AUTHOR("Charles Manning, Aleph One Ltd., 2002-2011"); ++MODULE_LICENSE("GPL"); +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs1.c linux-3.14.4/fs/yaffs2/yaffs_yaffs1.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs1.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_yaffs1.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,422 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_yaffs1.h" ++#include "yportenv.h" ++#include "yaffs_trace.h" ++#include "yaffs_bitmap.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_nand.h" ++#include "yaffs_attribs.h" ++ ++int yaffs1_scan(struct yaffs_dev *dev) ++{ ++ struct yaffs_ext_tags tags; ++ int blk; ++ int result; ++ int chunk; ++ int c; ++ int deleted; ++ enum yaffs_block_state state; ++ LIST_HEAD(hard_list); ++ struct yaffs_block_info *bi; ++ u32 seq_number; ++ struct yaffs_obj_hdr *oh; ++ struct yaffs_obj *in; ++ struct yaffs_obj *parent; ++ int alloc_failed = 0; ++ struct yaffs_shadow_fixer *shadow_fixers = NULL; ++ u8 *chunk_data; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "yaffs1_scan starts intstartblk %d intendblk %d...", ++ dev->internal_start_block, dev->internal_end_block); ++ ++ chunk_data = yaffs_get_temp_buffer(dev); ++ ++ dev->seq_number = YAFFS_LOWEST_SEQUENCE_NUMBER; ++ ++ /* Scan all the blocks to determine their state */ ++ bi = dev->block_info; ++ for (blk = dev->internal_start_block; blk <= dev->internal_end_block; ++ blk++) { ++ yaffs_clear_chunk_bits(dev, blk); ++ bi->pages_in_use = 0; ++ bi->soft_del_pages = 0; ++ ++ yaffs_query_init_block_state(dev, blk, &state, &seq_number); ++ ++ bi->block_state = state; ++ bi->seq_number = seq_number; ++ ++ if (bi->seq_number == YAFFS_SEQUENCE_BAD_BLOCK) ++ bi->block_state = state = YAFFS_BLOCK_STATE_DEAD; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN_DEBUG, ++ "Block scanning block %d state %d seq %d", ++ blk, state, seq_number); ++ ++ if (state == YAFFS_BLOCK_STATE_DEAD) { ++ yaffs_trace(YAFFS_TRACE_BAD_BLOCKS, ++ "block %d is bad", blk); ++ } else if (state == YAFFS_BLOCK_STATE_EMPTY) { ++ yaffs_trace(YAFFS_TRACE_SCAN_DEBUG, "Block empty "); ++ dev->n_erased_blocks++; ++ dev->n_free_chunks += dev->param.chunks_per_block; ++ } ++ bi++; ++ } ++ ++ /* For each block.... */ ++ for (blk = dev->internal_start_block; ++ !alloc_failed && blk <= dev->internal_end_block; blk++) { ++ ++ cond_resched(); ++ ++ bi = yaffs_get_block_info(dev, blk); ++ state = bi->block_state; ++ ++ deleted = 0; ++ ++ /* For each chunk in each block that needs scanning.... */ ++ for (c = 0; ++ !alloc_failed && c < dev->param.chunks_per_block && ++ state == YAFFS_BLOCK_STATE_NEEDS_SCAN; c++) { ++ /* Read the tags and decide what to do */ ++ chunk = blk * dev->param.chunks_per_block + c; ++ ++ result = yaffs_rd_chunk_tags_nand(dev, chunk, NULL, ++ &tags); ++ ++ /* Let's have a good look at this chunk... */ ++ ++ if (tags.ecc_result == YAFFS_ECC_RESULT_UNFIXED || ++ tags.is_deleted) { ++ /* YAFFS1 only... ++ * A deleted chunk ++ */ ++ deleted++; ++ dev->n_free_chunks++; ++ } else if (!tags.chunk_used) { ++ /* An unassigned chunk in the block ++ * This means that either the block is empty or ++ * this is the one being allocated from ++ */ ++ ++ if (c == 0) { ++ /* We're looking at the first chunk in ++ *the block so the block is unused */ ++ state = YAFFS_BLOCK_STATE_EMPTY; ++ dev->n_erased_blocks++; ++ } else { ++ /* this is the block being allocated */ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ " Allocating from %d %d", ++ blk, c); ++ state = YAFFS_BLOCK_STATE_ALLOCATING; ++ dev->alloc_block = blk; ++ dev->alloc_page = c; ++ dev->alloc_block_finder = blk; ++ ++ } ++ ++ dev->n_free_chunks += ++ (dev->param.chunks_per_block - c); ++ } else if (tags.chunk_id > 0) { ++ /* chunk_id > 0 so it is a data chunk... */ ++ unsigned int endpos; ++ ++ yaffs_set_chunk_bit(dev, blk, c); ++ bi->pages_in_use++; ++ ++ in = yaffs_find_or_create_by_number(dev, ++ tags.obj_id, ++ YAFFS_OBJECT_TYPE_FILE); ++ /* PutChunkIntoFile checks for a clash ++ * (two data chunks with the same chunk_id). ++ */ ++ ++ if (!in) ++ alloc_failed = 1; ++ ++ if (in) { ++ if (!yaffs_put_chunk_in_file ++ (in, tags.chunk_id, chunk, 1)) ++ alloc_failed = 1; ++ } ++ ++ endpos = ++ (tags.chunk_id - 1) * ++ dev->data_bytes_per_chunk + ++ tags.n_bytes; ++ if (in && ++ in->variant_type == ++ YAFFS_OBJECT_TYPE_FILE && ++ in->variant.file_variant.scanned_size < ++ endpos) { ++ in->variant.file_variant.scanned_size = ++ endpos; ++ if (!dev->param.use_header_file_size) { ++ in->variant. ++ file_variant.file_size = ++ in->variant. ++ file_variant.scanned_size; ++ } ++ ++ } ++ } else { ++ /* chunk_id == 0, so it is an ObjectHeader. ++ * Make the object ++ */ ++ yaffs_set_chunk_bit(dev, blk, c); ++ bi->pages_in_use++; ++ ++ result = yaffs_rd_chunk_tags_nand(dev, chunk, ++ chunk_data, ++ NULL); ++ ++ oh = (struct yaffs_obj_hdr *)chunk_data; ++ ++ in = yaffs_find_by_number(dev, tags.obj_id); ++ if (in && in->variant_type != oh->type) { ++ /* This should not happen, but somehow ++ * Wev'e ended up with an obj_id that ++ * has been reused but not yet deleted, ++ * and worse still it has changed type. ++ * Delete the old object. ++ */ ++ ++ yaffs_del_obj(in); ++ in = NULL; ++ } ++ ++ in = yaffs_find_or_create_by_number(dev, ++ tags.obj_id, ++ oh->type); ++ ++ if (!in) ++ alloc_failed = 1; ++ ++ if (in && oh->shadows_obj > 0) { ++ ++ struct yaffs_shadow_fixer *fixer; ++ fixer = ++ kmalloc(sizeof ++ (struct yaffs_shadow_fixer), ++ GFP_NOFS); ++ if (fixer) { ++ fixer->next = shadow_fixers; ++ shadow_fixers = fixer; ++ fixer->obj_id = tags.obj_id; ++ fixer->shadowed_id = ++ oh->shadows_obj; ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ " Shadow fixer: %d shadows %d", ++ fixer->obj_id, ++ fixer->shadowed_id); ++ ++ } ++ ++ } ++ ++ if (in && in->valid) { ++ /* We have already filled this one. ++ * We have a duplicate and need to ++ * resolve it. */ ++ ++ unsigned existing_serial = in->serial; ++ unsigned new_serial = ++ tags.serial_number; ++ ++ if (((existing_serial + 1) & 3) == ++ new_serial) { ++ /* Use new one - destroy the ++ * exisiting one */ ++ yaffs_chunk_del(dev, ++ in->hdr_chunk, ++ 1, __LINE__); ++ in->valid = 0; ++ } else { ++ /* Use existing - destroy ++ * this one. */ ++ yaffs_chunk_del(dev, chunk, 1, ++ __LINE__); ++ } ++ } ++ ++ if (in && !in->valid && ++ (tags.obj_id == YAFFS_OBJECTID_ROOT || ++ tags.obj_id == ++ YAFFS_OBJECTID_LOSTNFOUND)) { ++ /* We only load some info, don't fiddle ++ * with directory structure */ ++ in->valid = 1; ++ in->variant_type = oh->type; ++ ++ in->yst_mode = oh->yst_mode; ++ yaffs_load_attribs(in, oh); ++ in->hdr_chunk = chunk; ++ in->serial = tags.serial_number; ++ ++ } else if (in && !in->valid) { ++ /* we need to load this info */ ++ ++ in->valid = 1; ++ in->variant_type = oh->type; ++ ++ in->yst_mode = oh->yst_mode; ++ yaffs_load_attribs(in, oh); ++ in->hdr_chunk = chunk; ++ in->serial = tags.serial_number; ++ ++ yaffs_set_obj_name_from_oh(in, oh); ++ in->dirty = 0; ++ ++ /* directory stuff... ++ * hook up to parent ++ */ ++ ++ parent = ++ yaffs_find_or_create_by_number ++ (dev, oh->parent_obj_id, ++ YAFFS_OBJECT_TYPE_DIRECTORY); ++ if (!parent) ++ alloc_failed = 1; ++ if (parent && parent->variant_type == ++ YAFFS_OBJECT_TYPE_UNKNOWN) { ++ /* Set up as a directory */ ++ parent->variant_type = ++ YAFFS_OBJECT_TYPE_DIRECTORY; ++ INIT_LIST_HEAD(&parent-> ++ variant.dir_variant. ++ children); ++ } else if (!parent || ++ parent->variant_type != ++ YAFFS_OBJECT_TYPE_DIRECTORY) { ++ /* Hoosterman, a problem.... ++ * We're trying to use a ++ * non-directory as a directory ++ */ ++ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: attempting to use non-directory as a directory in scan. Put in lost+found." ++ ); ++ parent = dev->lost_n_found; ++ } ++ ++ yaffs_add_obj_to_dir(parent, in); ++ ++ switch (in->variant_type) { ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ /* Todo got a problem */ ++ break; ++ case YAFFS_OBJECT_TYPE_FILE: ++ if (dev->param. ++ use_header_file_size) ++ in->variant. ++ file_variant.file_size ++ = yaffs_oh_to_size(oh); ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ in->variant. ++ hardlink_variant.equiv_id = ++ oh->equiv_id; ++ list_add(&in->hard_links, ++ &hard_list); ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ in->variant.symlink_variant. ++ alias = ++ yaffs_clone_str(oh->alias); ++ if (!in->variant. ++ symlink_variant.alias) ++ alloc_failed = 1; ++ break; ++ } ++ } ++ } ++ } ++ ++ if (state == YAFFS_BLOCK_STATE_NEEDS_SCAN) { ++ /* If we got this far while scanning, ++ * then the block is fully allocated. */ ++ state = YAFFS_BLOCK_STATE_FULL; ++ } ++ ++ if (state == YAFFS_BLOCK_STATE_ALLOCATING) { ++ /* If the block was partially allocated then ++ * treat it as fully allocated. */ ++ state = YAFFS_BLOCK_STATE_FULL; ++ dev->alloc_block = -1; ++ } ++ ++ bi->block_state = state; ++ ++ /* Now let's see if it was dirty */ ++ if (bi->pages_in_use == 0 && ++ !bi->has_shrink_hdr && ++ bi->block_state == YAFFS_BLOCK_STATE_FULL) ++ yaffs_block_became_dirty(dev, blk); ++ } ++ ++ /* Ok, we've done all the scanning. ++ * Fix up the hard link chains. ++ * We should now have scanned all the objects, now it's time to add ++ * these hardlinks. ++ */ ++ ++ yaffs_link_fixup(dev, &hard_list); ++ ++ /* ++ * Fix up any shadowed objects. ++ * There should not be more than one of these. ++ */ ++ { ++ struct yaffs_shadow_fixer *fixer; ++ struct yaffs_obj *obj; ++ ++ while (shadow_fixers) { ++ fixer = shadow_fixers; ++ shadow_fixers = fixer->next; ++ /* Complete the rename transaction by deleting the ++ * shadowed object then setting the object header ++ to unshadowed. ++ */ ++ obj = yaffs_find_by_number(dev, fixer->shadowed_id); ++ if (obj) ++ yaffs_del_obj(obj); ++ ++ obj = yaffs_find_by_number(dev, fixer->obj_id); ++ ++ if (obj) ++ yaffs_update_oh(obj, NULL, 1, 0, 0, NULL); ++ ++ kfree(fixer); ++ } ++ } ++ ++ yaffs_release_temp_buffer(dev, chunk_data); ++ ++ if (alloc_failed) ++ return YAFFS_FAIL; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN, "yaffs1_scan ends"); ++ ++ return YAFFS_OK; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs1.h linux-3.14.4/fs/yaffs2/yaffs_yaffs1.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs1.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_yaffs1.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,22 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_YAFFS1_H__ ++#define __YAFFS_YAFFS1_H__ ++ ++#include "yaffs_guts.h" ++int yaffs1_scan(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs2.c linux-3.14.4/fs/yaffs2/yaffs_yaffs2.c +--- linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs2.c 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_yaffs2.c 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,1534 @@ ++/* ++ * YAFFS: Yet Another Flash File System. A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++ ++#include "yaffs_guts.h" ++#include "yaffs_trace.h" ++#include "yaffs_yaffs2.h" ++#include "yaffs_checkptrw.h" ++#include "yaffs_bitmap.h" ++#include "yaffs_nand.h" ++#include "yaffs_getblockinfo.h" ++#include "yaffs_verify.h" ++#include "yaffs_attribs.h" ++#include "yaffs_summary.h" ++ ++/* ++ * Checkpoints are really no benefit on very small partitions. ++ * ++ * To save space on small partitions don't bother with checkpoints unless ++ * the partition is at least this big. ++ */ ++#define YAFFS_CHECKPOINT_MIN_BLOCKS 60 ++#define YAFFS_SMALL_HOLE_THRESHOLD 4 ++ ++/* ++ * Oldest Dirty Sequence Number handling. ++ */ ++ ++/* yaffs_calc_oldest_dirty_seq() ++ * yaffs2_find_oldest_dirty_seq() ++ * Calculate the oldest dirty sequence number if we don't know it. ++ */ ++void yaffs_calc_oldest_dirty_seq(struct yaffs_dev *dev) ++{ ++ int i; ++ unsigned seq; ++ unsigned block_no = 0; ++ struct yaffs_block_info *b; ++ ++ if (!dev->param.is_yaffs2) ++ return; ++ ++ /* Find the oldest dirty sequence number. */ ++ seq = dev->seq_number + 1; ++ b = dev->block_info; ++ for (i = dev->internal_start_block; i <= dev->internal_end_block; i++) { ++ if (b->block_state == YAFFS_BLOCK_STATE_FULL && ++ (b->pages_in_use - b->soft_del_pages) < ++ dev->param.chunks_per_block && ++ b->seq_number < seq) { ++ seq = b->seq_number; ++ block_no = i; ++ } ++ b++; ++ } ++ ++ if (block_no) { ++ dev->oldest_dirty_seq = seq; ++ dev->oldest_dirty_block = block_no; ++ } ++} ++ ++void yaffs2_find_oldest_dirty_seq(struct yaffs_dev *dev) ++{ ++ if (!dev->param.is_yaffs2) ++ return; ++ ++ if (!dev->oldest_dirty_seq) ++ yaffs_calc_oldest_dirty_seq(dev); ++} ++ ++/* ++ * yaffs_clear_oldest_dirty_seq() ++ * Called when a block is erased or marked bad. (ie. when its seq_number ++ * becomes invalid). If the value matches the oldest then we clear ++ * dev->oldest_dirty_seq to force its recomputation. ++ */ ++void yaffs2_clear_oldest_dirty_seq(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi) ++{ ++ ++ if (!dev->param.is_yaffs2) ++ return; ++ ++ if (!bi || bi->seq_number == dev->oldest_dirty_seq) { ++ dev->oldest_dirty_seq = 0; ++ dev->oldest_dirty_block = 0; ++ } ++} ++ ++/* ++ * yaffs2_update_oldest_dirty_seq() ++ * Update the oldest dirty sequence number whenever we dirty a block. ++ * Only do this if the oldest_dirty_seq is actually being tracked. ++ */ ++void yaffs2_update_oldest_dirty_seq(struct yaffs_dev *dev, unsigned block_no, ++ struct yaffs_block_info *bi) ++{ ++ if (!dev->param.is_yaffs2) ++ return; ++ ++ if (dev->oldest_dirty_seq) { ++ if (dev->oldest_dirty_seq > bi->seq_number) { ++ dev->oldest_dirty_seq = bi->seq_number; ++ dev->oldest_dirty_block = block_no; ++ } ++ } ++} ++ ++int yaffs_block_ok_for_gc(struct yaffs_dev *dev, struct yaffs_block_info *bi) ++{ ++ ++ if (!dev->param.is_yaffs2) ++ return 1; /* disqualification only applies to yaffs2. */ ++ ++ if (!bi->has_shrink_hdr) ++ return 1; /* can gc */ ++ ++ yaffs2_find_oldest_dirty_seq(dev); ++ ++ /* Can't do gc of this block if there are any blocks older than this ++ * one that have discarded pages. ++ */ ++ return (bi->seq_number <= dev->oldest_dirty_seq); ++} ++ ++/* ++ * yaffs2_find_refresh_block() ++ * periodically finds the oldest full block by sequence number for refreshing. ++ * Only for yaffs2. ++ */ ++u32 yaffs2_find_refresh_block(struct yaffs_dev *dev) ++{ ++ u32 b; ++ u32 oldest = 0; ++ u32 oldest_seq = 0; ++ struct yaffs_block_info *bi; ++ ++ if (!dev->param.is_yaffs2) ++ return oldest; ++ ++ /* ++ * If refresh period < 10 then refreshing is disabled. ++ */ ++ if (dev->param.refresh_period < 10) ++ return oldest; ++ ++ /* ++ * Fix broken values. ++ */ ++ if (dev->refresh_skip > dev->param.refresh_period) ++ dev->refresh_skip = dev->param.refresh_period; ++ ++ if (dev->refresh_skip > 0) ++ return oldest; ++ ++ /* ++ * Refresh skip is now zero. ++ * We'll do a refresh this time around.... ++ * Update the refresh skip and find the oldest block. ++ */ ++ dev->refresh_skip = dev->param.refresh_period; ++ dev->refresh_count++; ++ bi = dev->block_info; ++ for (b = dev->internal_start_block; b <= dev->internal_end_block; b++) { ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_FULL) { ++ ++ if (oldest < 1 || bi->seq_number < oldest_seq) { ++ oldest = b; ++ oldest_seq = bi->seq_number; ++ } ++ } ++ bi++; ++ } ++ ++ if (oldest > 0) { ++ yaffs_trace(YAFFS_TRACE_GC, ++ "GC refresh count %d selected block %d with seq_number %d", ++ dev->refresh_count, oldest, oldest_seq); ++ } ++ ++ return oldest; ++} ++ ++int yaffs2_checkpt_required(struct yaffs_dev *dev) ++{ ++ int nblocks; ++ ++ if (!dev->param.is_yaffs2) ++ return 0; ++ ++ nblocks = dev->internal_end_block - dev->internal_start_block + 1; ++ ++ return !dev->param.skip_checkpt_wr && ++ !dev->read_only && (nblocks >= YAFFS_CHECKPOINT_MIN_BLOCKS); ++} ++ ++int yaffs_calc_checkpt_blocks_required(struct yaffs_dev *dev) ++{ ++ int retval; ++ int n_bytes = 0; ++ int n_blocks; ++ int dev_blocks; ++ ++ if (!dev->param.is_yaffs2) ++ return 0; ++ ++ if (!dev->checkpoint_blocks_required && yaffs2_checkpt_required(dev)) { ++ /* Not a valid value so recalculate */ ++ dev_blocks = dev->param.end_block - dev->param.start_block + 1; ++ n_bytes += sizeof(struct yaffs_checkpt_validity); ++ n_bytes += sizeof(struct yaffs_checkpt_dev); ++ n_bytes += dev_blocks * sizeof(struct yaffs_block_info); ++ n_bytes += dev_blocks * dev->chunk_bit_stride; ++ n_bytes += ++ (sizeof(struct yaffs_checkpt_obj) + sizeof(u32)) * ++ dev->n_obj; ++ n_bytes += (dev->tnode_size + sizeof(u32)) * dev->n_tnodes; ++ n_bytes += sizeof(struct yaffs_checkpt_validity); ++ n_bytes += sizeof(u32); /* checksum */ ++ ++ /* Round up and add 2 blocks to allow for some bad blocks, ++ * so add 3 */ ++ ++ n_blocks = ++ (n_bytes / ++ (dev->data_bytes_per_chunk * ++ dev->param.chunks_per_block)) + 3; ++ ++ dev->checkpoint_blocks_required = n_blocks; ++ } ++ ++ retval = dev->checkpoint_blocks_required - dev->blocks_in_checkpt; ++ if (retval < 0) ++ retval = 0; ++ return retval; ++} ++ ++/*--------------------- Checkpointing --------------------*/ ++ ++static int yaffs2_wr_checkpt_validity_marker(struct yaffs_dev *dev, int head) ++{ ++ struct yaffs_checkpt_validity cp; ++ ++ memset(&cp, 0, sizeof(cp)); ++ ++ cp.struct_type = sizeof(cp); ++ cp.magic = YAFFS_MAGIC; ++ cp.version = YAFFS_CHECKPOINT_VERSION; ++ cp.head = (head) ? 1 : 0; ++ ++ return (yaffs2_checkpt_wr(dev, &cp, sizeof(cp)) == sizeof(cp)) ? 1 : 0; ++} ++ ++static int yaffs2_rd_checkpt_validity_marker(struct yaffs_dev *dev, int head) ++{ ++ struct yaffs_checkpt_validity cp; ++ int ok; ++ ++ ok = (yaffs2_checkpt_rd(dev, &cp, sizeof(cp)) == sizeof(cp)); ++ ++ if (ok) ++ ok = (cp.struct_type == sizeof(cp)) && ++ (cp.magic == YAFFS_MAGIC) && ++ (cp.version == YAFFS_CHECKPOINT_VERSION) && ++ (cp.head == ((head) ? 1 : 0)); ++ return ok ? 1 : 0; ++} ++ ++static void yaffs2_dev_to_checkpt_dev(struct yaffs_checkpt_dev *cp, ++ struct yaffs_dev *dev) ++{ ++ cp->n_erased_blocks = dev->n_erased_blocks; ++ cp->alloc_block = dev->alloc_block; ++ cp->alloc_page = dev->alloc_page; ++ cp->n_free_chunks = dev->n_free_chunks; ++ ++ cp->n_deleted_files = dev->n_deleted_files; ++ cp->n_unlinked_files = dev->n_unlinked_files; ++ cp->n_bg_deletions = dev->n_bg_deletions; ++ cp->seq_number = dev->seq_number; ++ ++} ++ ++static void yaffs_checkpt_dev_to_dev(struct yaffs_dev *dev, ++ struct yaffs_checkpt_dev *cp) ++{ ++ dev->n_erased_blocks = cp->n_erased_blocks; ++ dev->alloc_block = cp->alloc_block; ++ dev->alloc_page = cp->alloc_page; ++ dev->n_free_chunks = cp->n_free_chunks; ++ ++ dev->n_deleted_files = cp->n_deleted_files; ++ dev->n_unlinked_files = cp->n_unlinked_files; ++ dev->n_bg_deletions = cp->n_bg_deletions; ++ dev->seq_number = cp->seq_number; ++} ++ ++static int yaffs2_wr_checkpt_dev(struct yaffs_dev *dev) ++{ ++ struct yaffs_checkpt_dev cp; ++ u32 n_bytes; ++ u32 n_blocks = dev->internal_end_block - dev->internal_start_block + 1; ++ int ok; ++ ++ /* Write device runtime values */ ++ yaffs2_dev_to_checkpt_dev(&cp, dev); ++ cp.struct_type = sizeof(cp); ++ ++ ok = (yaffs2_checkpt_wr(dev, &cp, sizeof(cp)) == sizeof(cp)); ++ if (!ok) ++ return 0; ++ ++ /* Write block info */ ++ n_bytes = n_blocks * sizeof(struct yaffs_block_info); ++ ok = (yaffs2_checkpt_wr(dev, dev->block_info, n_bytes) == n_bytes); ++ if (!ok) ++ return 0; ++ ++ /* Write chunk bits */ ++ n_bytes = n_blocks * dev->chunk_bit_stride; ++ ok = (yaffs2_checkpt_wr(dev, dev->chunk_bits, n_bytes) == n_bytes); ++ ++ return ok ? 1 : 0; ++} ++ ++static int yaffs2_rd_checkpt_dev(struct yaffs_dev *dev) ++{ ++ struct yaffs_checkpt_dev cp; ++ u32 n_bytes; ++ u32 n_blocks = ++ (dev->internal_end_block - dev->internal_start_block + 1); ++ int ok; ++ ++ ok = (yaffs2_checkpt_rd(dev, &cp, sizeof(cp)) == sizeof(cp)); ++ if (!ok) ++ return 0; ++ ++ if (cp.struct_type != sizeof(cp)) ++ return 0; ++ ++ yaffs_checkpt_dev_to_dev(dev, &cp); ++ ++ n_bytes = n_blocks * sizeof(struct yaffs_block_info); ++ ++ ok = (yaffs2_checkpt_rd(dev, dev->block_info, n_bytes) == n_bytes); ++ ++ if (!ok) ++ return 0; ++ ++ n_bytes = n_blocks * dev->chunk_bit_stride; ++ ++ ok = (yaffs2_checkpt_rd(dev, dev->chunk_bits, n_bytes) == n_bytes); ++ ++ return ok ? 1 : 0; ++} ++ ++static void yaffs2_obj_checkpt_obj(struct yaffs_checkpt_obj *cp, ++ struct yaffs_obj *obj) ++{ ++ cp->obj_id = obj->obj_id; ++ cp->parent_id = (obj->parent) ? obj->parent->obj_id : 0; ++ cp->hdr_chunk = obj->hdr_chunk; ++ cp->variant_type = obj->variant_type; ++ cp->deleted = obj->deleted; ++ cp->soft_del = obj->soft_del; ++ cp->unlinked = obj->unlinked; ++ cp->fake = obj->fake; ++ cp->rename_allowed = obj->rename_allowed; ++ cp->unlink_allowed = obj->unlink_allowed; ++ cp->serial = obj->serial; ++ cp->n_data_chunks = obj->n_data_chunks; ++ ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_FILE) ++ cp->size_or_equiv_obj = obj->variant.file_variant.file_size; ++ else if (obj->variant_type == YAFFS_OBJECT_TYPE_HARDLINK) ++ cp->size_or_equiv_obj = obj->variant.hardlink_variant.equiv_id; ++} ++ ++static int yaffs2_checkpt_obj_to_obj(struct yaffs_obj *obj, ++ struct yaffs_checkpt_obj *cp) ++{ ++ struct yaffs_obj *parent; ++ ++ if (obj->variant_type != cp->variant_type) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "Checkpoint read object %d type %d chunk %d does not match existing object type %d", ++ cp->obj_id, cp->variant_type, cp->hdr_chunk, ++ obj->variant_type); ++ return 0; ++ } ++ ++ obj->obj_id = cp->obj_id; ++ ++ if (cp->parent_id) ++ parent = yaffs_find_or_create_by_number(obj->my_dev, ++ cp->parent_id, ++ YAFFS_OBJECT_TYPE_DIRECTORY); ++ else ++ parent = NULL; ++ ++ if (parent) { ++ if (parent->variant_type != YAFFS_OBJECT_TYPE_DIRECTORY) { ++ yaffs_trace(YAFFS_TRACE_ALWAYS, ++ "Checkpoint read object %d parent %d type %d chunk %d Parent type, %d, not directory", ++ cp->obj_id, cp->parent_id, ++ cp->variant_type, cp->hdr_chunk, ++ parent->variant_type); ++ return 0; ++ } ++ yaffs_add_obj_to_dir(parent, obj); ++ } ++ ++ obj->hdr_chunk = cp->hdr_chunk; ++ obj->variant_type = cp->variant_type; ++ obj->deleted = cp->deleted; ++ obj->soft_del = cp->soft_del; ++ obj->unlinked = cp->unlinked; ++ obj->fake = cp->fake; ++ obj->rename_allowed = cp->rename_allowed; ++ obj->unlink_allowed = cp->unlink_allowed; ++ obj->serial = cp->serial; ++ obj->n_data_chunks = cp->n_data_chunks; ++ ++ if (obj->variant_type == YAFFS_OBJECT_TYPE_FILE) ++ obj->variant.file_variant.file_size = cp->size_or_equiv_obj; ++ else if (obj->variant_type == YAFFS_OBJECT_TYPE_HARDLINK) ++ obj->variant.hardlink_variant.equiv_id = cp->size_or_equiv_obj; ++ ++ if (obj->hdr_chunk > 0) ++ obj->lazy_loaded = 1; ++ return 1; ++} ++ ++static int yaffs2_checkpt_tnode_worker(struct yaffs_obj *in, ++ struct yaffs_tnode *tn, u32 level, ++ int chunk_offset) ++{ ++ int i; ++ struct yaffs_dev *dev = in->my_dev; ++ int ok = 1; ++ u32 base_offset; ++ ++ if (!tn) ++ return 1; ++ ++ if (level > 0) { ++ for (i = 0; i < YAFFS_NTNODES_INTERNAL && ok; i++) { ++ if (!tn->internal[i]) ++ continue; ++ ok = yaffs2_checkpt_tnode_worker(in, ++ tn->internal[i], ++ level - 1, ++ (chunk_offset << ++ YAFFS_TNODES_INTERNAL_BITS) + i); ++ } ++ return ok; ++ } ++ ++ /* Level 0 tnode */ ++ base_offset = chunk_offset << YAFFS_TNODES_LEVEL0_BITS; ++ ok = (yaffs2_checkpt_wr(dev, &base_offset, sizeof(base_offset)) == ++ sizeof(base_offset)); ++ if (ok) ++ ok = (yaffs2_checkpt_wr(dev, tn, dev->tnode_size) == ++ dev->tnode_size); ++ ++ return ok; ++} ++ ++static int yaffs2_wr_checkpt_tnodes(struct yaffs_obj *obj) ++{ ++ u32 end_marker = ~0; ++ int ok = 1; ++ ++ if (obj->variant_type != YAFFS_OBJECT_TYPE_FILE) ++ return ok; ++ ++ ok = yaffs2_checkpt_tnode_worker(obj, ++ obj->variant.file_variant.top, ++ obj->variant.file_variant. ++ top_level, 0); ++ if (ok) ++ ok = (yaffs2_checkpt_wr(obj->my_dev, &end_marker, ++ sizeof(end_marker)) == sizeof(end_marker)); ++ ++ return ok ? 1 : 0; ++} ++ ++static int yaffs2_rd_checkpt_tnodes(struct yaffs_obj *obj) ++{ ++ u32 base_chunk; ++ int ok = 1; ++ struct yaffs_dev *dev = obj->my_dev; ++ struct yaffs_file_var *file_stuct_ptr = &obj->variant.file_variant; ++ struct yaffs_tnode *tn; ++ int nread = 0; ++ ++ ok = (yaffs2_checkpt_rd(dev, &base_chunk, sizeof(base_chunk)) == ++ sizeof(base_chunk)); ++ ++ while (ok && (~base_chunk)) { ++ nread++; ++ /* Read level 0 tnode */ ++ ++ tn = yaffs_get_tnode(dev); ++ if (tn) ++ ok = (yaffs2_checkpt_rd(dev, tn, dev->tnode_size) == ++ dev->tnode_size); ++ else ++ ok = 0; ++ ++ if (tn && ok) ++ ok = yaffs_add_find_tnode_0(dev, ++ file_stuct_ptr, ++ base_chunk, tn) ? 1 : 0; ++ ++ if (ok) ++ ok = (yaffs2_checkpt_rd ++ (dev, &base_chunk, ++ sizeof(base_chunk)) == sizeof(base_chunk)); ++ } ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "Checkpoint read tnodes %d records, last %d. ok %d", ++ nread, base_chunk, ok); ++ ++ return ok ? 1 : 0; ++} ++ ++static int yaffs2_wr_checkpt_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_checkpt_obj cp; ++ int i; ++ int ok = 1; ++ struct list_head *lh; ++ ++ /* Iterate through the objects in each hash entry, ++ * dumping them to the checkpointing stream. ++ */ ++ ++ for (i = 0; ok && i < YAFFS_NOBJECT_BUCKETS; i++) { ++ list_for_each(lh, &dev->obj_bucket[i].list) { ++ obj = list_entry(lh, struct yaffs_obj, hash_link); ++ if (!obj->defered_free) { ++ yaffs2_obj_checkpt_obj(&cp, obj); ++ cp.struct_type = sizeof(cp); ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "Checkpoint write object %d parent %d type %d chunk %d obj addr %p", ++ cp.obj_id, cp.parent_id, ++ cp.variant_type, cp.hdr_chunk, obj); ++ ++ ok = (yaffs2_checkpt_wr(dev, &cp, ++ sizeof(cp)) == sizeof(cp)); ++ ++ if (ok && ++ obj->variant_type == ++ YAFFS_OBJECT_TYPE_FILE) ++ ok = yaffs2_wr_checkpt_tnodes(obj); ++ } ++ } ++ } ++ ++ /* Dump end of list */ ++ memset(&cp, 0xff, sizeof(struct yaffs_checkpt_obj)); ++ cp.struct_type = sizeof(cp); ++ ++ if (ok) ++ ok = (yaffs2_checkpt_wr(dev, &cp, sizeof(cp)) == sizeof(cp)); ++ ++ return ok ? 1 : 0; ++} ++ ++static int yaffs2_rd_checkpt_objs(struct yaffs_dev *dev) ++{ ++ struct yaffs_obj *obj; ++ struct yaffs_checkpt_obj cp; ++ int ok = 1; ++ int done = 0; ++ LIST_HEAD(hard_list); ++ ++ ++ while (ok && !done) { ++ ok = (yaffs2_checkpt_rd(dev, &cp, sizeof(cp)) == sizeof(cp)); ++ if (cp.struct_type != sizeof(cp)) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "struct size %d instead of %d ok %d", ++ cp.struct_type, (int)sizeof(cp), ok); ++ ok = 0; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "Checkpoint read object %d parent %d type %d chunk %d ", ++ cp.obj_id, cp.parent_id, cp.variant_type, ++ cp.hdr_chunk); ++ ++ if (ok && cp.obj_id == ~0) { ++ done = 1; ++ } else if (ok) { ++ obj = ++ yaffs_find_or_create_by_number(dev, cp.obj_id, ++ cp.variant_type); ++ if (obj) { ++ ok = yaffs2_checkpt_obj_to_obj(obj, &cp); ++ if (!ok) ++ break; ++ if (obj->variant_type == ++ YAFFS_OBJECT_TYPE_FILE) { ++ ok = yaffs2_rd_checkpt_tnodes(obj); ++ } else if (obj->variant_type == ++ YAFFS_OBJECT_TYPE_HARDLINK) { ++ list_add(&obj->hard_links, &hard_list); ++ } ++ } else { ++ ok = 0; ++ } ++ } ++ } ++ ++ if (ok) ++ yaffs_link_fixup(dev, &hard_list); ++ ++ return ok ? 1 : 0; ++} ++ ++static int yaffs2_wr_checkpt_sum(struct yaffs_dev *dev) ++{ ++ u32 checkpt_sum; ++ int ok; ++ ++ yaffs2_get_checkpt_sum(dev, &checkpt_sum); ++ ++ ok = (yaffs2_checkpt_wr(dev, &checkpt_sum, sizeof(checkpt_sum)) == ++ sizeof(checkpt_sum)); ++ ++ if (!ok) ++ return 0; ++ ++ return 1; ++} ++ ++static int yaffs2_rd_checkpt_sum(struct yaffs_dev *dev) ++{ ++ u32 checkpt_sum0; ++ u32 checkpt_sum1; ++ int ok; ++ ++ yaffs2_get_checkpt_sum(dev, &checkpt_sum0); ++ ++ ok = (yaffs2_checkpt_rd(dev, &checkpt_sum1, sizeof(checkpt_sum1)) == ++ sizeof(checkpt_sum1)); ++ ++ if (!ok) ++ return 0; ++ ++ if (checkpt_sum0 != checkpt_sum1) ++ return 0; ++ ++ return 1; ++} ++ ++static int yaffs2_wr_checkpt_data(struct yaffs_dev *dev) ++{ ++ int ok = 1; ++ ++ if (!yaffs2_checkpt_required(dev)) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "skipping checkpoint write"); ++ ok = 0; ++ } ++ ++ if (ok) ++ ok = yaffs2_checkpt_open(dev, 1); ++ ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "write checkpoint validity"); ++ ok = yaffs2_wr_checkpt_validity_marker(dev, 1); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "write checkpoint device"); ++ ok = yaffs2_wr_checkpt_dev(dev); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "write checkpoint objects"); ++ ok = yaffs2_wr_checkpt_objs(dev); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "write checkpoint validity"); ++ ok = yaffs2_wr_checkpt_validity_marker(dev, 0); ++ } ++ ++ if (ok) ++ ok = yaffs2_wr_checkpt_sum(dev); ++ ++ if (!yaffs_checkpt_close(dev)) ++ ok = 0; ++ ++ if (ok) ++ dev->is_checkpointed = 1; ++ else ++ dev->is_checkpointed = 0; ++ ++ return dev->is_checkpointed; ++} ++ ++static int yaffs2_rd_checkpt_data(struct yaffs_dev *dev) ++{ ++ int ok = 1; ++ ++ if (!dev->param.is_yaffs2) ++ ok = 0; ++ ++ if (ok && dev->param.skip_checkpt_rd) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "skipping checkpoint read"); ++ ok = 0; ++ } ++ ++ if (ok) ++ ok = yaffs2_checkpt_open(dev, 0); /* open for read */ ++ ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "read checkpoint validity"); ++ ok = yaffs2_rd_checkpt_validity_marker(dev, 1); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "read checkpoint device"); ++ ok = yaffs2_rd_checkpt_dev(dev); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "read checkpoint objects"); ++ ok = yaffs2_rd_checkpt_objs(dev); ++ } ++ if (ok) { ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "read checkpoint validity"); ++ ok = yaffs2_rd_checkpt_validity_marker(dev, 0); ++ } ++ ++ if (ok) { ++ ok = yaffs2_rd_checkpt_sum(dev); ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "read checkpoint checksum %d", ok); ++ } ++ ++ if (!yaffs_checkpt_close(dev)) ++ ok = 0; ++ ++ if (ok) ++ dev->is_checkpointed = 1; ++ else ++ dev->is_checkpointed = 0; ++ ++ return ok ? 1 : 0; ++} ++ ++void yaffs2_checkpt_invalidate(struct yaffs_dev *dev) ++{ ++ if (dev->is_checkpointed || dev->blocks_in_checkpt > 0) { ++ dev->is_checkpointed = 0; ++ yaffs2_checkpt_invalidate_stream(dev); ++ } ++ if (dev->param.sb_dirty_fn) ++ dev->param.sb_dirty_fn(dev); ++} ++ ++int yaffs_checkpoint_save(struct yaffs_dev *dev) ++{ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "save entry: is_checkpointed %d", ++ dev->is_checkpointed); ++ ++ yaffs_verify_objects(dev); ++ yaffs_verify_blocks(dev); ++ yaffs_verify_free_chunks(dev); ++ ++ if (!dev->is_checkpointed) { ++ yaffs2_checkpt_invalidate(dev); ++ yaffs2_wr_checkpt_data(dev); ++ } ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT | YAFFS_TRACE_MOUNT, ++ "save exit: is_checkpointed %d", ++ dev->is_checkpointed); ++ ++ return dev->is_checkpointed; ++} ++ ++int yaffs2_checkpt_restore(struct yaffs_dev *dev) ++{ ++ int retval; ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "restore entry: is_checkpointed %d", ++ dev->is_checkpointed); ++ ++ retval = yaffs2_rd_checkpt_data(dev); ++ ++ if (dev->is_checkpointed) { ++ yaffs_verify_objects(dev); ++ yaffs_verify_blocks(dev); ++ yaffs_verify_free_chunks(dev); ++ } ++ ++ yaffs_trace(YAFFS_TRACE_CHECKPOINT, ++ "restore exit: is_checkpointed %d", ++ dev->is_checkpointed); ++ ++ return retval; ++} ++ ++int yaffs2_handle_hole(struct yaffs_obj *obj, loff_t new_size) ++{ ++ /* if new_size > old_file_size. ++ * We're going to be writing a hole. ++ * If the hole is small then write zeros otherwise write a start ++ * of hole marker. ++ */ ++ loff_t old_file_size; ++ loff_t increase; ++ int small_hole; ++ int result = YAFFS_OK; ++ struct yaffs_dev *dev = NULL; ++ u8 *local_buffer = NULL; ++ int small_increase_ok = 0; ++ ++ if (!obj) ++ return YAFFS_FAIL; ++ ++ if (obj->variant_type != YAFFS_OBJECT_TYPE_FILE) ++ return YAFFS_FAIL; ++ ++ dev = obj->my_dev; ++ ++ /* Bail out if not yaffs2 mode */ ++ if (!dev->param.is_yaffs2) ++ return YAFFS_OK; ++ ++ old_file_size = obj->variant.file_variant.file_size; ++ ++ if (new_size <= old_file_size) ++ return YAFFS_OK; ++ ++ increase = new_size - old_file_size; ++ ++ if (increase < YAFFS_SMALL_HOLE_THRESHOLD * dev->data_bytes_per_chunk && ++ yaffs_check_alloc_available(dev, YAFFS_SMALL_HOLE_THRESHOLD + 1)) ++ small_hole = 1; ++ else ++ small_hole = 0; ++ ++ if (small_hole) ++ local_buffer = yaffs_get_temp_buffer(dev); ++ ++ if (local_buffer) { ++ /* fill hole with zero bytes */ ++ loff_t pos = old_file_size; ++ int this_write; ++ int written; ++ memset(local_buffer, 0, dev->data_bytes_per_chunk); ++ small_increase_ok = 1; ++ ++ while (increase > 0 && small_increase_ok) { ++ this_write = increase; ++ if (this_write > dev->data_bytes_per_chunk) ++ this_write = dev->data_bytes_per_chunk; ++ written = ++ yaffs_do_file_wr(obj, local_buffer, pos, this_write, ++ 0); ++ if (written == this_write) { ++ pos += this_write; ++ increase -= this_write; ++ } else { ++ small_increase_ok = 0; ++ } ++ } ++ ++ yaffs_release_temp_buffer(dev, local_buffer); ++ ++ /* If out of space then reverse any chunks we've added */ ++ if (!small_increase_ok) ++ yaffs_resize_file_down(obj, old_file_size); ++ } ++ ++ if (!small_increase_ok && ++ obj->parent && ++ obj->parent->obj_id != YAFFS_OBJECTID_UNLINKED && ++ obj->parent->obj_id != YAFFS_OBJECTID_DELETED) { ++ /* Write a hole start header with the old file size */ ++ yaffs_update_oh(obj, NULL, 0, 1, 0, NULL); ++ } ++ ++ return result; ++} ++ ++struct yaffs_block_index { ++ int seq; ++ int block; ++}; ++ ++static int yaffs2_ybicmp(const void *a, const void *b) ++{ ++ int aseq = ((struct yaffs_block_index *)a)->seq; ++ int bseq = ((struct yaffs_block_index *)b)->seq; ++ int ablock = ((struct yaffs_block_index *)a)->block; ++ int bblock = ((struct yaffs_block_index *)b)->block; ++ ++ if (aseq == bseq) ++ return ablock - bblock; ++ ++ return aseq - bseq; ++} ++ ++static inline int yaffs2_scan_chunk(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi, ++ int blk, int chunk_in_block, ++ int *found_chunks, ++ u8 *chunk_data, ++ struct list_head *hard_list, ++ int summary_available) ++{ ++ struct yaffs_obj_hdr *oh; ++ struct yaffs_obj *in; ++ struct yaffs_obj *parent; ++ int equiv_id; ++ loff_t file_size; ++ int is_shrink; ++ int is_unlinked; ++ struct yaffs_ext_tags tags; ++ int result; ++ int alloc_failed = 0; ++ int chunk = blk * dev->param.chunks_per_block + chunk_in_block; ++ struct yaffs_file_var *file_var; ++ struct yaffs_hardlink_var *hl_var; ++ struct yaffs_symlink_var *sl_var; ++ ++ if (summary_available) { ++ result = yaffs_summary_fetch(dev, &tags, chunk_in_block); ++ tags.seq_number = bi->seq_number; ++ } ++ ++ if (!summary_available || tags.obj_id == 0) { ++ result = yaffs_rd_chunk_tags_nand(dev, chunk, NULL, &tags); ++ dev->tags_used++; ++ } else { ++ dev->summary_used++; ++ } ++ ++ /* Let's have a good look at this chunk... */ ++ ++ if (!tags.chunk_used) { ++ /* An unassigned chunk in the block. ++ * If there are used chunks after this one, then ++ * it is a chunk that was skipped due to failing ++ * the erased check. Just skip it so that it can ++ * be deleted. ++ * But, more typically, We get here when this is ++ * an unallocated chunk and his means that ++ * either the block is empty or this is the one ++ * being allocated from ++ */ ++ ++ if (*found_chunks) { ++ /* This is a chunk that was skipped due ++ * to failing the erased check */ ++ } else if (chunk_in_block == 0) { ++ /* We're looking at the first chunk in ++ * the block so the block is unused */ ++ bi->block_state = YAFFS_BLOCK_STATE_EMPTY; ++ dev->n_erased_blocks++; ++ } else { ++ if (bi->block_state == YAFFS_BLOCK_STATE_NEEDS_SCAN || ++ bi->block_state == YAFFS_BLOCK_STATE_ALLOCATING) { ++ if (dev->seq_number == bi->seq_number) { ++ /* Allocating from this block*/ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ " Allocating from %d %d", ++ blk, chunk_in_block); ++ ++ bi->block_state = ++ YAFFS_BLOCK_STATE_ALLOCATING; ++ dev->alloc_block = blk; ++ dev->alloc_page = chunk_in_block; ++ dev->alloc_block_finder = blk; ++ } else { ++ /* This is a partially written block ++ * that is not the current ++ * allocation block. ++ */ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "Partially written block %d detected. gc will fix this.", ++ blk); ++ } ++ } ++ } ++ ++ dev->n_free_chunks++; ++ ++ } else if (tags.ecc_result == ++ YAFFS_ECC_RESULT_UNFIXED) { ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ " Unfixed ECC in chunk(%d:%d), chunk ignored", ++ blk, chunk_in_block); ++ dev->n_free_chunks++; ++ } else if (tags.obj_id > YAFFS_MAX_OBJECT_ID || ++ tags.chunk_id > YAFFS_MAX_CHUNK_ID || ++ tags.obj_id == YAFFS_OBJECTID_SUMMARY || ++ (tags.chunk_id > 0 && ++ tags.n_bytes > dev->data_bytes_per_chunk) || ++ tags.seq_number != bi->seq_number) { ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "Chunk (%d:%d) with bad tags:obj = %d, chunk_id = %d, n_bytes = %d, ignored", ++ blk, chunk_in_block, tags.obj_id, ++ tags.chunk_id, tags.n_bytes); ++ dev->n_free_chunks++; ++ } else if (tags.chunk_id > 0) { ++ /* chunk_id > 0 so it is a data chunk... */ ++ loff_t endpos; ++ loff_t chunk_base = (tags.chunk_id - 1) * ++ dev->data_bytes_per_chunk; ++ ++ *found_chunks = 1; ++ ++ yaffs_set_chunk_bit(dev, blk, chunk_in_block); ++ bi->pages_in_use++; ++ ++ in = yaffs_find_or_create_by_number(dev, ++ tags.obj_id, ++ YAFFS_OBJECT_TYPE_FILE); ++ if (!in) ++ /* Out of memory */ ++ alloc_failed = 1; ++ ++ if (in && ++ in->variant_type == YAFFS_OBJECT_TYPE_FILE && ++ chunk_base < in->variant.file_variant.shrink_size) { ++ /* This has not been invalidated by ++ * a resize */ ++ if (!yaffs_put_chunk_in_file(in, tags.chunk_id, ++ chunk, -1)) ++ alloc_failed = 1; ++ ++ /* File size is calculated by looking at ++ * the data chunks if we have not ++ * seen an object header yet. ++ * Stop this practice once we find an ++ * object header. ++ */ ++ endpos = chunk_base + tags.n_bytes; ++ ++ if (!in->valid && ++ in->variant.file_variant.scanned_size < endpos) { ++ in->variant.file_variant. ++ scanned_size = endpos; ++ in->variant.file_variant. ++ file_size = endpos; ++ } ++ } else if (in) { ++ /* This chunk has been invalidated by a ++ * resize, or a past file deletion ++ * so delete the chunk*/ ++ yaffs_chunk_del(dev, chunk, 1, __LINE__); ++ } ++ } else { ++ /* chunk_id == 0, so it is an ObjectHeader. ++ * Thus, we read in the object header and make ++ * the object ++ */ ++ *found_chunks = 1; ++ ++ yaffs_set_chunk_bit(dev, blk, chunk_in_block); ++ bi->pages_in_use++; ++ ++ oh = NULL; ++ in = NULL; ++ ++ if (tags.extra_available) { ++ in = yaffs_find_or_create_by_number(dev, ++ tags.obj_id, ++ tags.extra_obj_type); ++ if (!in) ++ alloc_failed = 1; ++ } ++ ++ if (!in || ++ (!in->valid && dev->param.disable_lazy_load) || ++ tags.extra_shadows || ++ (!in->valid && (tags.obj_id == YAFFS_OBJECTID_ROOT || ++ tags.obj_id == YAFFS_OBJECTID_LOSTNFOUND))) { ++ ++ /* If we don't have valid info then we ++ * need to read the chunk ++ * TODO In future we can probably defer ++ * reading the chunk and living with ++ * invalid data until needed. ++ */ ++ ++ result = yaffs_rd_chunk_tags_nand(dev, ++ chunk, ++ chunk_data, ++ NULL); ++ ++ oh = (struct yaffs_obj_hdr *)chunk_data; ++ ++ if (dev->param.inband_tags) { ++ /* Fix up the header if they got ++ * corrupted by inband tags */ ++ oh->shadows_obj = ++ oh->inband_shadowed_obj_id; ++ oh->is_shrink = ++ oh->inband_is_shrink; ++ } ++ ++ if (!in) { ++ in = yaffs_find_or_create_by_number(dev, ++ tags.obj_id, oh->type); ++ if (!in) ++ alloc_failed = 1; ++ } ++ } ++ ++ if (!in) { ++ /* TODO Hoosterman we have a problem! */ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: Could not make object for object %d at chunk %d during scan", ++ tags.obj_id, chunk); ++ return YAFFS_FAIL; ++ } ++ ++ if (in->valid) { ++ /* We have already filled this one. ++ * We have a duplicate that will be ++ * discarded, but we first have to suck ++ * out resize info if it is a file. ++ */ ++ if ((in->variant_type == YAFFS_OBJECT_TYPE_FILE) && ++ ((oh && oh->type == YAFFS_OBJECT_TYPE_FILE) || ++ (tags.extra_available && ++ tags.extra_obj_type == YAFFS_OBJECT_TYPE_FILE) ++ )) { ++ loff_t this_size = (oh) ? ++ yaffs_oh_to_size(oh) : ++ tags.extra_file_size; ++ u32 parent_obj_id = (oh) ? ++ oh->parent_obj_id : ++ tags.extra_parent_id; ++ ++ is_shrink = (oh) ? ++ oh->is_shrink : ++ tags.extra_is_shrink; ++ ++ /* If it is deleted (unlinked ++ * at start also means deleted) ++ * we treat the file size as ++ * being zeroed at this point. ++ */ ++ if (parent_obj_id == YAFFS_OBJECTID_DELETED || ++ parent_obj_id == YAFFS_OBJECTID_UNLINKED) { ++ this_size = 0; ++ is_shrink = 1; ++ } ++ ++ if (is_shrink && ++ in->variant.file_variant.shrink_size > ++ this_size) ++ in->variant.file_variant.shrink_size = ++ this_size; ++ ++ if (is_shrink) ++ bi->has_shrink_hdr = 1; ++ } ++ /* Use existing - destroy this one. */ ++ yaffs_chunk_del(dev, chunk, 1, __LINE__); ++ } ++ ++ if (!in->valid && in->variant_type != ++ (oh ? oh->type : tags.extra_obj_type)) { ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: Bad type, %d != %d, for object %d at chunk %d during scan", ++ oh ? oh->type : tags.extra_obj_type, ++ in->variant_type, tags.obj_id, ++ chunk); ++ in = yaffs_retype_obj(in, oh ? oh->type : tags.extra_obj_type); ++ } ++ ++ if (!in->valid && ++ (tags.obj_id == YAFFS_OBJECTID_ROOT || ++ tags.obj_id == YAFFS_OBJECTID_LOSTNFOUND)) { ++ /* We only load some info, don't fiddle ++ * with directory structure */ ++ in->valid = 1; ++ ++ if (oh) { ++ in->yst_mode = oh->yst_mode; ++ yaffs_load_attribs(in, oh); ++ in->lazy_loaded = 0; ++ } else { ++ in->lazy_loaded = 1; ++ } ++ in->hdr_chunk = chunk; ++ ++ } else if (!in->valid) { ++ /* we need to load this info */ ++ in->valid = 1; ++ in->hdr_chunk = chunk; ++ if (oh) { ++ in->variant_type = oh->type; ++ in->yst_mode = oh->yst_mode; ++ yaffs_load_attribs(in, oh); ++ ++ if (oh->shadows_obj > 0) ++ yaffs_handle_shadowed_obj(dev, ++ oh->shadows_obj, 1); ++ ++ yaffs_set_obj_name_from_oh(in, oh); ++ parent = yaffs_find_or_create_by_number(dev, ++ oh->parent_obj_id, ++ YAFFS_OBJECT_TYPE_DIRECTORY); ++ file_size = yaffs_oh_to_size(oh); ++ is_shrink = oh->is_shrink; ++ equiv_id = oh->equiv_id; ++ } else { ++ in->variant_type = tags.extra_obj_type; ++ parent = yaffs_find_or_create_by_number(dev, ++ tags.extra_parent_id, ++ YAFFS_OBJECT_TYPE_DIRECTORY); ++ file_size = tags.extra_file_size; ++ is_shrink = tags.extra_is_shrink; ++ equiv_id = tags.extra_equiv_id; ++ in->lazy_loaded = 1; ++ } ++ in->dirty = 0; ++ ++ if (!parent) ++ alloc_failed = 1; ++ ++ /* directory stuff... ++ * hook up to parent ++ */ ++ ++ if (parent && ++ parent->variant_type == YAFFS_OBJECT_TYPE_UNKNOWN) { ++ /* Set up as a directory */ ++ parent->variant_type = ++ YAFFS_OBJECT_TYPE_DIRECTORY; ++ INIT_LIST_HEAD(&parent-> ++ variant.dir_variant.children); ++ } else if (!parent || ++ parent->variant_type != ++ YAFFS_OBJECT_TYPE_DIRECTORY) { ++ /* Hoosterman, another problem.... ++ * Trying to use a non-directory as a directory ++ */ ++ ++ yaffs_trace(YAFFS_TRACE_ERROR, ++ "yaffs tragedy: attempting to use non-directory as a directory in scan. Put in lost+found." ++ ); ++ parent = dev->lost_n_found; ++ } ++ yaffs_add_obj_to_dir(parent, in); ++ ++ is_unlinked = (parent == dev->del_dir) || ++ (parent == dev->unlinked_dir); ++ ++ if (is_shrink) ++ /* Mark the block */ ++ bi->has_shrink_hdr = 1; ++ ++ /* Note re hardlinks. ++ * Since we might scan a hardlink before its equivalent ++ * object is scanned we put them all in a list. ++ * After scanning is complete, we should have all the ++ * objects, so we run through this list and fix up all ++ * the chains. ++ */ ++ ++ switch (in->variant_type) { ++ case YAFFS_OBJECT_TYPE_UNKNOWN: ++ /* Todo got a problem */ ++ break; ++ case YAFFS_OBJECT_TYPE_FILE: ++ file_var = &in->variant.file_variant; ++ if (file_var->scanned_size < file_size) { ++ /* This covers the case where the file ++ * size is greater than the data held. ++ * This will happen if the file is ++ * resized to be larger than its ++ * current data extents. ++ */ ++ file_var->file_size = file_size; ++ file_var->scanned_size = file_size; ++ } ++ ++ if (file_var->shrink_size > file_size) ++ file_var->shrink_size = file_size; ++ ++ break; ++ case YAFFS_OBJECT_TYPE_HARDLINK: ++ hl_var = &in->variant.hardlink_variant; ++ if (!is_unlinked) { ++ hl_var->equiv_id = equiv_id; ++ list_add(&in->hard_links, hard_list); ++ } ++ break; ++ case YAFFS_OBJECT_TYPE_DIRECTORY: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_SPECIAL: ++ /* Do nothing */ ++ break; ++ case YAFFS_OBJECT_TYPE_SYMLINK: ++ sl_var = &in->variant.symlink_variant; ++ if (oh) { ++ sl_var->alias = ++ yaffs_clone_str(oh->alias); ++ if (!sl_var->alias) ++ alloc_failed = 1; ++ } ++ break; ++ } ++ } ++ } ++ return alloc_failed ? YAFFS_FAIL : YAFFS_OK; ++} ++ ++int yaffs2_scan_backwards(struct yaffs_dev *dev) ++{ ++ int blk; ++ int block_iter; ++ int start_iter; ++ int end_iter; ++ int n_to_scan = 0; ++ enum yaffs_block_state state; ++ int c; ++ int deleted; ++ LIST_HEAD(hard_list); ++ struct yaffs_block_info *bi; ++ u32 seq_number; ++ int n_blocks = dev->internal_end_block - dev->internal_start_block + 1; ++ u8 *chunk_data; ++ int found_chunks; ++ int alloc_failed = 0; ++ struct yaffs_block_index *block_index = NULL; ++ int alt_block_index = 0; ++ int summary_available; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "yaffs2_scan_backwards starts intstartblk %d intendblk %d...", ++ dev->internal_start_block, dev->internal_end_block); ++ ++ dev->seq_number = YAFFS_LOWEST_SEQUENCE_NUMBER; ++ ++ block_index = ++ kmalloc(n_blocks * sizeof(struct yaffs_block_index), GFP_NOFS); ++ ++ if (!block_index) { ++ block_index = ++ vmalloc(n_blocks * sizeof(struct yaffs_block_index)); ++ alt_block_index = 1; ++ } ++ ++ if (!block_index) { ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "yaffs2_scan_backwards() could not allocate block index!" ++ ); ++ return YAFFS_FAIL; ++ } ++ ++ dev->blocks_in_checkpt = 0; ++ ++ chunk_data = yaffs_get_temp_buffer(dev); ++ ++ /* Scan all the blocks to determine their state */ ++ bi = dev->block_info; ++ for (blk = dev->internal_start_block; blk <= dev->internal_end_block; ++ blk++) { ++ yaffs_clear_chunk_bits(dev, blk); ++ bi->pages_in_use = 0; ++ bi->soft_del_pages = 0; ++ ++ yaffs_query_init_block_state(dev, blk, &state, &seq_number); ++ ++ bi->block_state = state; ++ bi->seq_number = seq_number; ++ ++ if (bi->seq_number == YAFFS_SEQUENCE_CHECKPOINT_DATA) ++ bi->block_state = YAFFS_BLOCK_STATE_CHECKPOINT; ++ if (bi->seq_number == YAFFS_SEQUENCE_BAD_BLOCK) ++ bi->block_state = YAFFS_BLOCK_STATE_DEAD; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN_DEBUG, ++ "Block scanning block %d state %d seq %d", ++ blk, bi->block_state, seq_number); ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_CHECKPOINT) { ++ dev->blocks_in_checkpt++; ++ ++ } else if (bi->block_state == YAFFS_BLOCK_STATE_DEAD) { ++ yaffs_trace(YAFFS_TRACE_BAD_BLOCKS, ++ "block %d is bad", blk); ++ } else if (bi->block_state == YAFFS_BLOCK_STATE_EMPTY) { ++ yaffs_trace(YAFFS_TRACE_SCAN_DEBUG, "Block empty "); ++ dev->n_erased_blocks++; ++ dev->n_free_chunks += dev->param.chunks_per_block; ++ } else if (bi->block_state == ++ YAFFS_BLOCK_STATE_NEEDS_SCAN) { ++ /* Determine the highest sequence number */ ++ if (seq_number >= YAFFS_LOWEST_SEQUENCE_NUMBER && ++ seq_number < YAFFS_HIGHEST_SEQUENCE_NUMBER) { ++ block_index[n_to_scan].seq = seq_number; ++ block_index[n_to_scan].block = blk; ++ n_to_scan++; ++ if (seq_number >= dev->seq_number) ++ dev->seq_number = seq_number; ++ } else { ++ /* TODO: Nasty sequence number! */ ++ yaffs_trace(YAFFS_TRACE_SCAN, ++ "Block scanning block %d has bad sequence number %d", ++ blk, seq_number); ++ } ++ } ++ bi++; ++ } ++ ++ yaffs_trace(YAFFS_TRACE_ALWAYS, "%d blocks to be sorted...", n_to_scan); ++ ++ cond_resched(); ++ ++ /* Sort the blocks by sequence number */ ++ sort(block_index, n_to_scan, sizeof(struct yaffs_block_index), ++ yaffs2_ybicmp, NULL); ++ ++ cond_resched(); ++ ++ yaffs_trace(YAFFS_TRACE_SCAN, "...done"); ++ ++ /* Now scan the blocks looking at the data. */ ++ start_iter = 0; ++ end_iter = n_to_scan - 1; ++ yaffs_trace(YAFFS_TRACE_SCAN_DEBUG, "%d blocks to scan", n_to_scan); ++ ++ /* For each block.... backwards */ ++ for (block_iter = end_iter; ++ !alloc_failed && block_iter >= start_iter; ++ block_iter--) { ++ /* Cooperative multitasking! This loop can run for so ++ long that watchdog timers expire. */ ++ cond_resched(); ++ ++ /* get the block to scan in the correct order */ ++ blk = block_index[block_iter].block; ++ bi = yaffs_get_block_info(dev, blk); ++ deleted = 0; ++ ++ summary_available = yaffs_summary_read(dev, dev->sum_tags, blk); ++ ++ /* For each chunk in each block that needs scanning.... */ ++ found_chunks = 0; ++ if (summary_available) ++ c = dev->chunks_per_summary - 1; ++ else ++ c = dev->param.chunks_per_block - 1; ++ ++ for (/* c is already initialised */; ++ !alloc_failed && c >= 0 && ++ (bi->block_state == YAFFS_BLOCK_STATE_NEEDS_SCAN || ++ bi->block_state == YAFFS_BLOCK_STATE_ALLOCATING); ++ c--) { ++ /* Scan backwards... ++ * Read the tags and decide what to do ++ */ ++ if (yaffs2_scan_chunk(dev, bi, blk, c, ++ &found_chunks, chunk_data, ++ &hard_list, summary_available) == ++ YAFFS_FAIL) ++ alloc_failed = 1; ++ } ++ ++ if (bi->block_state == YAFFS_BLOCK_STATE_NEEDS_SCAN) { ++ /* If we got this far while scanning, then the block ++ * is fully allocated. */ ++ bi->block_state = YAFFS_BLOCK_STATE_FULL; ++ } ++ ++ /* Now let's see if it was dirty */ ++ if (bi->pages_in_use == 0 && ++ !bi->has_shrink_hdr && ++ bi->block_state == YAFFS_BLOCK_STATE_FULL) { ++ yaffs_block_became_dirty(dev, blk); ++ } ++ } ++ ++ yaffs_skip_rest_of_block(dev); ++ ++ if (alt_block_index) ++ vfree(block_index); ++ else ++ kfree(block_index); ++ ++ /* Ok, we've done all the scanning. ++ * Fix up the hard link chains. ++ * We have scanned all the objects, now it's time to add these ++ * hardlinks. ++ */ ++ yaffs_link_fixup(dev, &hard_list); ++ ++ yaffs_release_temp_buffer(dev, chunk_data); ++ ++ if (alloc_failed) ++ return YAFFS_FAIL; ++ ++ yaffs_trace(YAFFS_TRACE_SCAN, "yaffs2_scan_backwards ends"); ++ ++ return YAFFS_OK; ++} +diff -Nur linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs2.h linux-3.14.4/fs/yaffs2/yaffs_yaffs2.h +--- linux-3.14.4.orig/fs/yaffs2/yaffs_yaffs2.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yaffs_yaffs2.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,39 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YAFFS_YAFFS2_H__ ++#define __YAFFS_YAFFS2_H__ ++ ++#include "yaffs_guts.h" ++ ++void yaffs_calc_oldest_dirty_seq(struct yaffs_dev *dev); ++void yaffs2_find_oldest_dirty_seq(struct yaffs_dev *dev); ++void yaffs2_clear_oldest_dirty_seq(struct yaffs_dev *dev, ++ struct yaffs_block_info *bi); ++void yaffs2_update_oldest_dirty_seq(struct yaffs_dev *dev, unsigned block_no, ++ struct yaffs_block_info *bi); ++int yaffs_block_ok_for_gc(struct yaffs_dev *dev, struct yaffs_block_info *bi); ++u32 yaffs2_find_refresh_block(struct yaffs_dev *dev); ++int yaffs2_checkpt_required(struct yaffs_dev *dev); ++int yaffs_calc_checkpt_blocks_required(struct yaffs_dev *dev); ++ ++void yaffs2_checkpt_invalidate(struct yaffs_dev *dev); ++int yaffs2_checkpt_save(struct yaffs_dev *dev); ++int yaffs2_checkpt_restore(struct yaffs_dev *dev); ++ ++int yaffs2_handle_hole(struct yaffs_obj *obj, loff_t new_size); ++int yaffs2_scan_backwards(struct yaffs_dev *dev); ++ ++#endif +diff -Nur linux-3.14.4.orig/fs/yaffs2/yportenv.h linux-3.14.4/fs/yaffs2/yportenv.h +--- linux-3.14.4.orig/fs/yaffs2/yportenv.h 1970-01-01 01:00:00.000000000 +0100 ++++ linux-3.14.4/fs/yaffs2/yportenv.h 2014-05-17 02:22:30.000000000 +0200 +@@ -0,0 +1,85 @@ ++/* ++ * YAFFS: Yet another Flash File System . A NAND-flash specific file system. ++ * ++ * Copyright (C) 2002-2011 Aleph One Ltd. ++ * for Toby Churchill Ltd and Brightstar Engineering ++ * ++ * Created by Charles Manning <charles@aleph1.co.uk> ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU Lesser General Public License version 2.1 as ++ * published by the Free Software Foundation. ++ * ++ * Note: Only YAFFS headers are LGPL, YAFFS C code is covered by GPL. ++ */ ++ ++#ifndef __YPORTENV_H__ ++#define __YPORTENV_H__ ++ ++/* ++ * Define the MTD version in terms of Linux Kernel versions ++ * This allows yaffs to be used independantly of the kernel ++ * as well as with it. ++ */ ++ ++#define MTD_VERSION(a, b, c) (((a) << 16) + ((b) << 8) + (c)) ++ ++#ifdef YAFFS_OUT_OF_TREE ++#include "moduleconfig.h" ++#endif ++ ++#include <linux/version.h> ++#define MTD_VERSION_CODE LINUX_VERSION_CODE ++ ++#if (LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 19)) ++#include <linux/config.h> ++#endif ++#include <linux/version.h> ++#include <linux/kernel.h> ++#include <linux/mm.h> ++#include <linux/sched.h> ++#include <linux/string.h> ++#include <linux/slab.h> ++#include <linux/vmalloc.h> ++#include <linux/xattr.h> ++#include <linux/list.h> ++#include <linux/types.h> ++#include <linux/fs.h> ++#include <linux/stat.h> ++#include <linux/sort.h> ++#include <linux/bitops.h> ++ ++/* These type wrappings are used to support Unicode names in WinCE. */ ++#define YCHAR char ++#define YUCHAR unsigned char ++#define _Y(x) x ++ ++#define YAFFS_LOSTNFOUND_NAME "lost+found" ++#define YAFFS_LOSTNFOUND_PREFIX "obj" ++ ++ ++#define YAFFS_ROOT_MODE 0755 ++#define YAFFS_LOSTNFOUND_MODE 0700 ++ ++#if (LINUX_VERSION_CODE > KERNEL_VERSION(2, 5, 0)) ++#define Y_CURRENT_TIME CURRENT_TIME.tv_sec ++#define Y_TIME_CONVERT(x) (x).tv_sec ++#else ++#define Y_CURRENT_TIME CURRENT_TIME ++#define Y_TIME_CONVERT(x) (x) ++#endif ++ ++#define compile_time_assertion(assertion) \ ++ ({ int x = __builtin_choose_expr(assertion, 0, (void)0); (void) x; }) ++ ++ ++#define yaffs_printf(msk, fmt, ...) \ ++ printk(KERN_DEBUG "yaffs: " fmt "\n", ##__VA_ARGS__) ++ ++#define yaffs_trace(msk, fmt, ...) do { \ ++ if (yaffs_trace_mask & (msk)) \ ++ printk(KERN_DEBUG "yaffs: " fmt "\n", ##__VA_ARGS__); \ ++} while (0) ++ ++ ++#endif diff --git a/target/linux/patches/3.14.40/zlib-inflate.patch b/target/linux/patches/3.14.40/zlib-inflate.patch new file mode 100644 index 000000000..58e1f6d21 --- /dev/null +++ b/target/linux/patches/3.14.40/zlib-inflate.patch @@ -0,0 +1,12 @@ +diff -Nur linux-2.6.37.orig/lib/Kconfig linux-2.6.37/lib/Kconfig +--- linux-2.6.37.orig/lib/Kconfig 2011-01-05 01:50:19.000000000 +0100 ++++ linux-2.6.37/lib/Kconfig 2011-03-01 20:10:29.833370667 +0100 +@@ -95,7 +95,7 @@ + # compression support is select'ed if needed + # + config ZLIB_INFLATE +- tristate ++ boolean + + config ZLIB_DEFLATE + tristate |