| /* Create descriptor for processing file. |
| Copyright (C) 1998-2010, 2012, 2014 Red Hat, Inc. |
| This file is part of elfutils. |
| Written by Ulrich Drepper <drepper@redhat.com>, 1998. |
| |
| This file is free software; you can redistribute it and/or modify |
| it under the terms of either |
| |
| * the GNU Lesser General Public License as published by the Free |
| Software Foundation; either version 3 of the License, or (at |
| your option) any later version |
| |
| or |
| |
| * 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 |
| |
| or both in parallel, as here. |
| |
| elfutils is distributed in the hope that it will be useful, but |
| WITHOUT ANY WARRANTY; without even the implied warranty of |
| MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| General Public License for more details. |
| |
| You should have received copies of the GNU General Public License and |
| the GNU Lesser General Public License along with this program. If |
| not, see <http://www.gnu.org/licenses/>. */ |
| |
| #ifdef HAVE_CONFIG_H |
| # include <config.h> |
| #endif |
| |
| #include <assert.h> |
| #include <ctype.h> |
| #include <errno.h> |
| #include <fcntl.h> |
| #include <stdbool.h> |
| #include <stddef.h> |
| #include <string.h> |
| #include <unistd.h> |
| #include <sys/mman.h> |
| #include <sys/param.h> |
| #include <sys/stat.h> |
| |
| #include <system.h> |
| #include "libelfP.h" |
| #include "common.h" |
| |
| |
| /* Create descriptor for archive in memory. */ |
| static inline Elf * |
| file_read_ar (int fildes, void *map_address, off_t offset, size_t maxsize, |
| Elf_Cmd cmd, Elf *parent) |
| { |
| Elf *elf; |
| |
| /* Create a descriptor. */ |
| elf = allocate_elf (fildes, map_address, offset, maxsize, cmd, parent, |
| ELF_K_AR, 0); |
| if (elf != NULL) |
| { |
| /* We don't read all the symbol tables in advance. All this will |
| happen on demand. */ |
| elf->state.ar.offset = offset + SARMAG; |
| |
| elf->state.ar.elf_ar_hdr.ar_rawname = elf->state.ar.raw_name; |
| } |
| |
| return elf; |
| } |
| |
| |
| static size_t |
| get_shnum (void *map_address, unsigned char *e_ident, int fildes, off_t offset, |
| size_t maxsize) |
| { |
| size_t result; |
| union |
| { |
| Elf32_Ehdr *e32; |
| Elf64_Ehdr *e64; |
| void *p; |
| } ehdr; |
| union |
| { |
| Elf32_Ehdr e32; |
| Elf64_Ehdr e64; |
| } ehdr_mem; |
| bool is32 = e_ident[EI_CLASS] == ELFCLASS32; |
| |
| /* Make the ELF header available. */ |
| if (e_ident[EI_DATA] == MY_ELFDATA |
| && (ALLOW_UNALIGNED |
| || (((size_t) e_ident |
| & ((is32 ? __alignof__ (Elf32_Ehdr) : __alignof__ (Elf64_Ehdr)) |
| - 1)) == 0))) |
| ehdr.p = e_ident; |
| else |
| { |
| /* We already read the ELF header. We have to copy the header |
| since we possibly modify the data here and the caller |
| expects the memory it passes in to be preserved. */ |
| ehdr.p = &ehdr_mem; |
| |
| if (is32) |
| { |
| if (ALLOW_UNALIGNED) |
| { |
| ehdr_mem.e32.e_shnum = ((Elf32_Ehdr *) e_ident)->e_shnum; |
| ehdr_mem.e32.e_shoff = ((Elf32_Ehdr *) e_ident)->e_shoff; |
| } |
| else |
| memcpy (&ehdr_mem, e_ident, sizeof (Elf32_Ehdr)); |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| { |
| CONVERT (ehdr_mem.e32.e_shnum); |
| CONVERT (ehdr_mem.e32.e_shoff); |
| } |
| } |
| else |
| { |
| if (ALLOW_UNALIGNED) |
| { |
| ehdr_mem.e64.e_shnum = ((Elf64_Ehdr *) e_ident)->e_shnum; |
| ehdr_mem.e64.e_shoff = ((Elf64_Ehdr *) e_ident)->e_shoff; |
| } |
| else |
| memcpy (&ehdr_mem, e_ident, sizeof (Elf64_Ehdr)); |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| { |
| CONVERT (ehdr_mem.e64.e_shnum); |
| CONVERT (ehdr_mem.e64.e_shoff); |
| } |
| } |
| } |
| |
| if (is32) |
| { |
| /* Get the number of sections from the ELF header. */ |
| result = ehdr.e32->e_shnum; |
| |
| if (unlikely (result == 0) && ehdr.e32->e_shoff != 0) |
| { |
| if (unlikely (ehdr.e32->e_shoff >= maxsize) |
| || unlikely (maxsize - ehdr.e32->e_shoff < sizeof (Elf32_Shdr))) |
| /* Cannot read the first section header. */ |
| return 0; |
| |
| if (likely (map_address != NULL) && e_ident[EI_DATA] == MY_ELFDATA |
| && (ALLOW_UNALIGNED |
| || (((size_t) ((char *) map_address + offset)) |
| & (__alignof__ (Elf32_Ehdr) - 1)) == 0)) |
| /* We can directly access the memory. */ |
| result = ((Elf32_Shdr *) ((char *) map_address + ehdr.e32->e_shoff |
| + offset))->sh_size; |
| else |
| { |
| Elf32_Word size; |
| |
| if (likely (map_address != NULL)) |
| /* gcc will optimize the memcpy to a simple memory |
| access while taking care of alignment issues. */ |
| memcpy (&size, &((Elf32_Shdr *) ((char *) map_address |
| + ehdr.e32->e_shoff |
| + offset))->sh_size, |
| sizeof (Elf32_Word)); |
| else |
| if (unlikely (pread_retry (fildes, &size, sizeof (Elf32_Word), |
| offset + ehdr.e32->e_shoff |
| + offsetof (Elf32_Shdr, sh_size)) |
| != sizeof (Elf32_Word))) |
| return (size_t) -1l; |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| CONVERT (size); |
| |
| result = size; |
| } |
| } |
| |
| /* If the section headers were truncated, pretend none were there. */ |
| if (ehdr.e32->e_shoff > maxsize |
| || maxsize - ehdr.e32->e_shoff < sizeof (Elf32_Shdr) * result) |
| result = 0; |
| } |
| else |
| { |
| /* Get the number of sections from the ELF header. */ |
| result = ehdr.e64->e_shnum; |
| |
| if (unlikely (result == 0) && ehdr.e64->e_shoff != 0) |
| { |
| if (unlikely (ehdr.e64->e_shoff >= maxsize) |
| || unlikely (ehdr.e64->e_shoff + sizeof (Elf64_Shdr) > maxsize)) |
| /* Cannot read the first section header. */ |
| return 0; |
| |
| Elf64_Xword size; |
| if (likely (map_address != NULL) && e_ident[EI_DATA] == MY_ELFDATA |
| && (ALLOW_UNALIGNED |
| || (((size_t) ((char *) map_address + offset)) |
| & (__alignof__ (Elf64_Ehdr) - 1)) == 0)) |
| /* We can directly access the memory. */ |
| size = ((Elf64_Shdr *) ((char *) map_address + ehdr.e64->e_shoff |
| + offset))->sh_size; |
| else |
| { |
| if (likely (map_address != NULL)) |
| /* gcc will optimize the memcpy to a simple memory |
| access while taking care of alignment issues. */ |
| memcpy (&size, &((Elf64_Shdr *) ((char *) map_address |
| + ehdr.e64->e_shoff |
| + offset))->sh_size, |
| sizeof (Elf64_Xword)); |
| else |
| if (unlikely (pread_retry (fildes, &size, sizeof (Elf64_Word), |
| offset + ehdr.e64->e_shoff |
| + offsetof (Elf64_Shdr, sh_size)) |
| != sizeof (Elf64_Xword))) |
| return (size_t) -1l; |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| CONVERT (size); |
| } |
| |
| if (size > ~((GElf_Word) 0)) |
| /* Invalid value, it is too large. */ |
| return (size_t) -1l; |
| |
| result = size; |
| } |
| |
| /* If the section headers were truncated, pretend none were there. */ |
| if (ehdr.e64->e_shoff > maxsize |
| || maxsize - ehdr.e64->e_shoff < sizeof (Elf64_Shdr) * result) |
| result = 0; |
| } |
| |
| return result; |
| } |
| |
| |
| /* Create descriptor for ELF file in memory. */ |
| static Elf * |
| file_read_elf (int fildes, void *map_address, unsigned char *e_ident, |
| off_t offset, size_t maxsize, Elf_Cmd cmd, Elf *parent) |
| { |
| /* Verify the binary is of the class we can handle. */ |
| if (unlikely ((e_ident[EI_CLASS] != ELFCLASS32 |
| && e_ident[EI_CLASS] != ELFCLASS64) |
| /* We also can only handle two encodings. */ |
| || (e_ident[EI_DATA] != ELFDATA2LSB |
| && e_ident[EI_DATA] != ELFDATA2MSB))) |
| { |
| /* Cannot handle this. */ |
| __libelf_seterrno (ELF_E_INVALID_FILE); |
| return NULL; |
| } |
| |
| /* Determine the number of sections. */ |
| size_t scncnt = get_shnum (map_address, e_ident, fildes, offset, maxsize); |
| if (scncnt == (size_t) -1l) |
| /* Could not determine the number of sections. */ |
| return NULL; |
| |
| /* Check for too many sections. */ |
| if (e_ident[EI_CLASS] == ELFCLASS32) |
| { |
| if (scncnt > SIZE_MAX / (sizeof (Elf_Scn) + sizeof (Elf32_Shdr))) |
| return NULL; |
| } |
| else if (scncnt > SIZE_MAX / (sizeof (Elf_Scn) + sizeof (Elf64_Shdr))) |
| return NULL; |
| |
| /* We can now allocate the memory. Even if there are no section headers, |
| we allocate space for a zeroth section in case we need it later. */ |
| const size_t scnmax = (scncnt ?: (cmd == ELF_C_RDWR || cmd == ELF_C_RDWR_MMAP) |
| ? 1 : 0); |
| Elf *elf = allocate_elf (fildes, map_address, offset, maxsize, cmd, parent, |
| ELF_K_ELF, scnmax * sizeof (Elf_Scn)); |
| if (elf == NULL) |
| /* Not enough memory. */ |
| return NULL; |
| |
| assert ((unsigned int) scncnt == scncnt); |
| assert (offsetof (struct Elf, state.elf32.scns) |
| == offsetof (struct Elf, state.elf64.scns)); |
| elf->state.elf32.scns.cnt = scncnt; |
| elf->state.elf32.scns.max = scnmax; |
| |
| /* Some more or less arbitrary value. */ |
| elf->state.elf.scnincr = 10; |
| |
| /* Make the class easily available. */ |
| elf->class = e_ident[EI_CLASS]; |
| |
| if (e_ident[EI_CLASS] == ELFCLASS32) |
| { |
| /* This pointer might not be directly usable if the alignment is |
| not sufficient for the architecture. */ |
| Elf32_Ehdr *ehdr = (Elf32_Ehdr *) ((char *) map_address + offset); |
| |
| /* This is a 32-bit binary. */ |
| if (map_address != NULL && e_ident[EI_DATA] == MY_ELFDATA |
| && (ALLOW_UNALIGNED |
| || ((((uintptr_t) ehdr) & (__alignof__ (Elf32_Ehdr) - 1)) == 0 |
| && ((uintptr_t) ((char *) ehdr + ehdr->e_shoff) |
| & (__alignof__ (Elf32_Shdr) - 1)) == 0 |
| && ((uintptr_t) ((char *) ehdr + ehdr->e_phoff) |
| & (__alignof__ (Elf32_Phdr) - 1)) == 0))) |
| { |
| /* We can use the mmapped memory. */ |
| elf->state.elf32.ehdr = ehdr; |
| |
| if (unlikely (ehdr->e_shoff >= maxsize) |
| || unlikely (maxsize - ehdr->e_shoff |
| < scncnt * sizeof (Elf32_Shdr))) |
| { |
| free_and_out: |
| free (elf); |
| __libelf_seterrno (ELF_E_INVALID_FILE); |
| return NULL; |
| } |
| elf->state.elf32.shdr |
| = (Elf32_Shdr *) ((char *) ehdr + ehdr->e_shoff); |
| |
| /* Don't precache the phdr pointer here. |
| elf32_getphdr will validate it against the size when asked. */ |
| |
| for (size_t cnt = 0; cnt < scncnt; ++cnt) |
| { |
| elf->state.elf32.scns.data[cnt].index = cnt; |
| elf->state.elf32.scns.data[cnt].elf = elf; |
| elf->state.elf32.scns.data[cnt].shdr.e32 = |
| &elf->state.elf32.shdr[cnt]; |
| if (likely (elf->state.elf32.shdr[cnt].sh_offset < maxsize) |
| && likely (maxsize - elf->state.elf32.shdr[cnt].sh_offset |
| <= elf->state.elf32.shdr[cnt].sh_size)) |
| elf->state.elf32.scns.data[cnt].rawdata_base = |
| elf->state.elf32.scns.data[cnt].data_base = |
| ((char *) map_address + offset |
| + elf->state.elf32.shdr[cnt].sh_offset); |
| elf->state.elf32.scns.data[cnt].list = &elf->state.elf32.scns; |
| |
| /* If this is a section with an extended index add a |
| reference in the section which uses the extended |
| index. */ |
| if (elf->state.elf32.shdr[cnt].sh_type == SHT_SYMTAB_SHNDX |
| && elf->state.elf32.shdr[cnt].sh_link < scncnt) |
| elf->state.elf32.scns.data[elf->state.elf32.shdr[cnt].sh_link].shndx_index |
| = cnt; |
| |
| /* Set the own shndx_index field in case it has not yet |
| been set. */ |
| if (elf->state.elf32.scns.data[cnt].shndx_index == 0) |
| elf->state.elf32.scns.data[cnt].shndx_index = -1; |
| } |
| } |
| else |
| { |
| /* Copy the ELF header. */ |
| elf->state.elf32.ehdr = memcpy (&elf->state.elf32.ehdr_mem, e_ident, |
| sizeof (Elf32_Ehdr)); |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| { |
| CONVERT (elf->state.elf32.ehdr_mem.e_type); |
| CONVERT (elf->state.elf32.ehdr_mem.e_machine); |
| CONVERT (elf->state.elf32.ehdr_mem.e_version); |
| CONVERT (elf->state.elf32.ehdr_mem.e_entry); |
| CONVERT (elf->state.elf32.ehdr_mem.e_phoff); |
| CONVERT (elf->state.elf32.ehdr_mem.e_shoff); |
| CONVERT (elf->state.elf32.ehdr_mem.e_flags); |
| CONVERT (elf->state.elf32.ehdr_mem.e_ehsize); |
| CONVERT (elf->state.elf32.ehdr_mem.e_phentsize); |
| CONVERT (elf->state.elf32.ehdr_mem.e_phnum); |
| CONVERT (elf->state.elf32.ehdr_mem.e_shentsize); |
| CONVERT (elf->state.elf32.ehdr_mem.e_shnum); |
| CONVERT (elf->state.elf32.ehdr_mem.e_shstrndx); |
| } |
| |
| for (size_t cnt = 0; cnt < scncnt; ++cnt) |
| { |
| elf->state.elf32.scns.data[cnt].index = cnt; |
| elf->state.elf32.scns.data[cnt].elf = elf; |
| elf->state.elf32.scns.data[cnt].list = &elf->state.elf32.scns; |
| } |
| } |
| |
| /* So far only one block with sections. */ |
| elf->state.elf32.scns_last = &elf->state.elf32.scns; |
| } |
| else |
| { |
| /* This pointer might not be directly usable if the alignment is |
| not sufficient for the architecture. */ |
| Elf64_Ehdr *ehdr = (Elf64_Ehdr *) ((char *) map_address + offset); |
| |
| /* This is a 64-bit binary. */ |
| if (map_address != NULL && e_ident[EI_DATA] == MY_ELFDATA |
| && (ALLOW_UNALIGNED |
| || ((((uintptr_t) ehdr) & (__alignof__ (Elf64_Ehdr) - 1)) == 0 |
| && ((uintptr_t) ((char *) ehdr + ehdr->e_shoff) |
| & (__alignof__ (Elf64_Shdr) - 1)) == 0 |
| && ((uintptr_t) ((char *) ehdr + ehdr->e_phoff) |
| & (__alignof__ (Elf64_Phdr) - 1)) == 0))) |
| { |
| /* We can use the mmapped memory. */ |
| elf->state.elf64.ehdr = ehdr; |
| |
| if (unlikely (ehdr->e_shoff >= maxsize) |
| || unlikely (ehdr->e_shoff |
| + scncnt * sizeof (Elf32_Shdr) > maxsize)) |
| goto free_and_out; |
| elf->state.elf64.shdr |
| = (Elf64_Shdr *) ((char *) ehdr + ehdr->e_shoff); |
| |
| /* Don't precache the phdr pointer here. |
| elf64_getphdr will validate it against the size when asked. */ |
| |
| for (size_t cnt = 0; cnt < scncnt; ++cnt) |
| { |
| elf->state.elf64.scns.data[cnt].index = cnt; |
| elf->state.elf64.scns.data[cnt].elf = elf; |
| elf->state.elf64.scns.data[cnt].shdr.e64 = |
| &elf->state.elf64.shdr[cnt]; |
| if (likely (elf->state.elf64.shdr[cnt].sh_offset < maxsize) |
| && likely (maxsize - elf->state.elf64.shdr[cnt].sh_offset |
| <= elf->state.elf64.shdr[cnt].sh_size)) |
| elf->state.elf64.scns.data[cnt].rawdata_base = |
| elf->state.elf64.scns.data[cnt].data_base = |
| ((char *) map_address + offset |
| + elf->state.elf64.shdr[cnt].sh_offset); |
| elf->state.elf64.scns.data[cnt].list = &elf->state.elf64.scns; |
| |
| /* If this is a section with an extended index add a |
| reference in the section which uses the extended |
| index. */ |
| if (elf->state.elf64.shdr[cnt].sh_type == SHT_SYMTAB_SHNDX |
| && elf->state.elf64.shdr[cnt].sh_link < scncnt) |
| elf->state.elf64.scns.data[elf->state.elf64.shdr[cnt].sh_link].shndx_index |
| = cnt; |
| |
| /* Set the own shndx_index field in case it has not yet |
| been set. */ |
| if (elf->state.elf64.scns.data[cnt].shndx_index == 0) |
| elf->state.elf64.scns.data[cnt].shndx_index = -1; |
| } |
| } |
| else |
| { |
| /* Copy the ELF header. */ |
| elf->state.elf64.ehdr = memcpy (&elf->state.elf64.ehdr_mem, e_ident, |
| sizeof (Elf64_Ehdr)); |
| |
| if (e_ident[EI_DATA] != MY_ELFDATA) |
| { |
| CONVERT (elf->state.elf64.ehdr_mem.e_type); |
| CONVERT (elf->state.elf64.ehdr_mem.e_machine); |
| CONVERT (elf->state.elf64.ehdr_mem.e_version); |
| CONVERT (elf->state.elf64.ehdr_mem.e_entry); |
| CONVERT (elf->state.elf64.ehdr_mem.e_phoff); |
| CONVERT (elf->state.elf64.ehdr_mem.e_shoff); |
| CONVERT (elf->state.elf64.ehdr_mem.e_flags); |
| CONVERT (elf->state.elf64.ehdr_mem.e_ehsize); |
| CONVERT (elf->state.elf64.ehdr_mem.e_phentsize); |
| CONVERT (elf->state.elf64.ehdr_mem.e_phnum); |
| CONVERT (elf->state.elf64.ehdr_mem.e_shentsize); |
| CONVERT (elf->state.elf64.ehdr_mem.e_shnum); |
| CONVERT (elf->state.elf64.ehdr_mem.e_shstrndx); |
| } |
| |
| for (size_t cnt = 0; cnt < scncnt; ++cnt) |
| { |
| elf->state.elf64.scns.data[cnt].index = cnt; |
| elf->state.elf64.scns.data[cnt].elf = elf; |
| elf->state.elf64.scns.data[cnt].list = &elf->state.elf64.scns; |
| } |
| } |
| |
| /* So far only one block with sections. */ |
| elf->state.elf64.scns_last = &elf->state.elf64.scns; |
| } |
| |
| return elf; |
| } |
| |
| |
| Elf * |
| internal_function |
| __libelf_read_mmaped_file (int fildes, void *map_address, off_t offset, |
| size_t maxsize, Elf_Cmd cmd, Elf *parent) |
| { |
| /* We have to find out what kind of file this is. We handle ELF |
| files and archives. To find out what we have we must look at the |
| header. The header for an ELF file is EI_NIDENT bytes in size, |
| the header for an archive file SARMAG bytes long. */ |
| unsigned char *e_ident = (unsigned char *) map_address + offset; |
| |
| /* See what kind of object we have here. */ |
| Elf_Kind kind = determine_kind (e_ident, maxsize); |
| |
| switch (kind) |
| { |
| case ELF_K_ELF: |
| return file_read_elf (fildes, map_address, e_ident, offset, maxsize, |
| cmd, parent); |
| |
| case ELF_K_AR: |
| return file_read_ar (fildes, map_address, offset, maxsize, cmd, parent); |
| |
| default: |
| break; |
| } |
| |
| /* This case is easy. Since we cannot do anything with this file |
| create a dummy descriptor. */ |
| return allocate_elf (fildes, map_address, offset, maxsize, cmd, parent, |
| ELF_K_NONE, 0); |
| } |
| |
| |
| static Elf * |
| read_unmmaped_file (int fildes, off_t offset, size_t maxsize, Elf_Cmd cmd, |
| Elf *parent) |
| { |
| /* We have to find out what kind of file this is. We handle ELF |
| files and archives. To find out what we have we must read the |
| header. The identification header for an ELF file is EI_NIDENT |
| bytes in size, but we read the whole ELF header since we will |
| need it anyway later. For archives the header in SARMAG bytes |
| long. Read the maximum of these numbers. |
| |
| XXX We have to change this for the extended `ar' format some day. |
| |
| Use a union to ensure alignment. We might later access the |
| memory as a ElfXX_Ehdr. */ |
| union |
| { |
| Elf64_Ehdr ehdr; |
| unsigned char header[MAX (sizeof (Elf64_Ehdr), SARMAG)]; |
| } mem; |
| |
| /* Read the head of the file. */ |
| ssize_t nread = pread_retry (fildes, mem.header, |
| MIN (MAX (sizeof (Elf64_Ehdr), SARMAG), |
| maxsize), |
| offset); |
| if (unlikely (nread == -1)) |
| /* We cannot even read the head of the file. Maybe FILDES is associated |
| with an unseekable device. This is nothing we can handle. */ |
| return NULL; |
| |
| /* See what kind of object we have here. */ |
| Elf_Kind kind = determine_kind (mem.header, nread); |
| |
| switch (kind) |
| { |
| case ELF_K_AR: |
| return file_read_ar (fildes, NULL, offset, maxsize, cmd, parent); |
| |
| case ELF_K_ELF: |
| /* Make sure at least the ELF header is contained in the file. */ |
| if ((size_t) nread >= (mem.header[EI_CLASS] == ELFCLASS32 |
| ? sizeof (Elf32_Ehdr) : sizeof (Elf64_Ehdr))) |
| return file_read_elf (fildes, NULL, mem.header, offset, maxsize, cmd, |
| parent); |
| /* FALLTHROUGH */ |
| |
| default: |
| break; |
| } |
| |
| /* This case is easy. Since we cannot do anything with this file |
| create a dummy descriptor. */ |
| return allocate_elf (fildes, NULL, offset, maxsize, cmd, parent, |
| ELF_K_NONE, 0); |
| } |
| |
| |
| /* Open a file for reading. If possible we will try to mmap() the file. */ |
| static struct Elf * |
| read_file (int fildes, off_t offset, size_t maxsize, |
| Elf_Cmd cmd, Elf *parent) |
| { |
| void *map_address = NULL; |
| int use_mmap = (cmd == ELF_C_READ_MMAP || cmd == ELF_C_RDWR_MMAP |
| || cmd == ELF_C_WRITE_MMAP |
| || cmd == ELF_C_READ_MMAP_PRIVATE); |
| |
| if (use_mmap) |
| { |
| if (parent == NULL) |
| { |
| if (maxsize == ~((size_t) 0)) |
| { |
| /* We don't know in the moment how large the file is. |
| Determine it now. */ |
| struct stat st; |
| |
| if (fstat (fildes, &st) == 0 |
| && (sizeof (size_t) >= sizeof (st.st_size) |
| || st.st_size <= ~((size_t) 0))) |
| maxsize = (size_t) st.st_size; |
| } |
| |
| /* We try to map the file ourself. */ |
| map_address = mmap (NULL, maxsize, (cmd == ELF_C_READ_MMAP |
| ? PROT_READ |
| : PROT_READ|PROT_WRITE), |
| cmd == ELF_C_READ_MMAP_PRIVATE |
| || cmd == ELF_C_READ_MMAP |
| ? MAP_PRIVATE : MAP_SHARED, |
| fildes, offset); |
| |
| if (map_address == MAP_FAILED) |
| map_address = NULL; |
| } |
| else |
| { |
| /* The parent is already loaded. Use it. */ |
| assert (maxsize != ~((size_t) 0)); |
| |
| map_address = parent->map_address; |
| } |
| } |
| |
| /* If we have the file in memory optimize the access. */ |
| if (map_address != NULL) |
| { |
| assert (map_address != MAP_FAILED); |
| |
| struct Elf *result = __libelf_read_mmaped_file (fildes, map_address, |
| offset, maxsize, cmd, |
| parent); |
| |
| /* If something went wrong during the initialization unmap the |
| memory if we mmaped here. */ |
| if (result == NULL |
| && (parent == NULL |
| || parent->map_address != map_address)) |
| munmap (map_address, maxsize); |
| else if (parent == NULL) |
| /* Remember that we mmap()ed the memory. */ |
| result->flags |= ELF_F_MMAPPED; |
| |
| return result; |
| } |
| |
| /* Otherwise we have to do it the hard way. We read as much as necessary |
| from the file whenever we need information which is not available. */ |
| return read_unmmaped_file (fildes, offset, maxsize, cmd, parent); |
| } |
| |
| |
| /* Find the entry with the long names for the content of this archive. */ |
| static const char * |
| read_long_names (Elf *elf) |
| { |
| off_t offset = SARMAG; /* This is the first entry. */ |
| struct ar_hdr hdrm; |
| struct ar_hdr *hdr; |
| char *newp; |
| size_t len; |
| |
| while (1) |
| { |
| if (elf->map_address != NULL) |
| { |
| if (offset + sizeof (struct ar_hdr) > elf->maximum_size) |
| return NULL; |
| |
| /* The data is mapped. */ |
| hdr = (struct ar_hdr *) (elf->map_address + offset); |
| } |
| else |
| { |
| /* Read the header from the file. */ |
| if (unlikely (pread_retry (elf->fildes, &hdrm, sizeof (hdrm), |
| elf->start_offset + offset) |
| != sizeof (hdrm))) |
| return NULL; |
| |
| hdr = &hdrm; |
| } |
| |
| len = atol (hdr->ar_size); |
| |
| if (memcmp (hdr->ar_name, "// ", 16) == 0) |
| break; |
| |
| offset += sizeof (struct ar_hdr) + ((len + 1) & ~1l); |
| } |
| |
| /* Due to the stupid format of the long name table entry (which are not |
| NUL terminted) we have to provide an appropriate representation anyhow. |
| Therefore we always make a copy which has the appropriate form. */ |
| newp = (char *) malloc (len); |
| if (newp != NULL) |
| { |
| char *runp; |
| |
| if (elf->map_address != NULL) |
| /* Simply copy it over. */ |
| elf->state.ar.long_names = (char *) memcpy (newp, |
| elf->map_address + offset |
| + sizeof (struct ar_hdr), |
| len); |
| else |
| { |
| if (unlikely ((size_t) pread_retry (elf->fildes, newp, len, |
| elf->start_offset + offset |
| + sizeof (struct ar_hdr)) |
| != len)) |
| { |
| /* We were not able to read all data. */ |
| free (newp); |
| elf->state.ar.long_names = NULL; |
| return NULL; |
| } |
| elf->state.ar.long_names = newp; |
| } |
| |
| elf->state.ar.long_names_len = len; |
| |
| /* Now NUL-terminate the strings. */ |
| runp = newp; |
| while (1) |
| { |
| runp = (char *) memchr (runp, '/', newp + len - runp); |
| if (runp == NULL) |
| /* This was the last entry. */ |
| break; |
| |
| /* NUL-terminate the string. */ |
| *runp = '\0'; |
| |
| /* Skip the NUL byte and the \012. */ |
| runp += 2; |
| |
| /* A sanity check. Somebody might have generated invalid |
| archive. */ |
| if (runp >= newp + len) |
| break; |
| } |
| } |
| |
| return newp; |
| } |
| |
| |
| /* Read the next archive header. */ |
| int |
| internal_function |
| __libelf_next_arhdr_wrlock (elf) |
| Elf *elf; |
| { |
| struct ar_hdr *ar_hdr; |
| Elf_Arhdr *elf_ar_hdr; |
| |
| if (elf->map_address != NULL) |
| { |
| /* See whether this entry is in the file. */ |
| if (unlikely (elf->state.ar.offset + sizeof (struct ar_hdr) |
| > elf->start_offset + elf->maximum_size)) |
| { |
| /* This record is not anymore in the file. */ |
| __libelf_seterrno (ELF_E_RANGE); |
| return -1; |
| } |
| ar_hdr = (struct ar_hdr *) (elf->map_address + elf->state.ar.offset); |
| } |
| else |
| { |
| ar_hdr = &elf->state.ar.ar_hdr; |
| |
| if (unlikely (pread_retry (elf->fildes, ar_hdr, sizeof (struct ar_hdr), |
| elf->state.ar.offset) |
| != sizeof (struct ar_hdr))) |
| { |
| /* Something went wrong while reading the file. */ |
| __libelf_seterrno (ELF_E_RANGE); |
| return -1; |
| } |
| } |
| |
| /* One little consistency check. */ |
| if (unlikely (memcmp (ar_hdr->ar_fmag, ARFMAG, 2) != 0)) |
| { |
| /* This is no valid archive. */ |
| __libelf_seterrno (ELF_E_ARCHIVE_FMAG); |
| return -1; |
| } |
| |
| /* Copy the raw name over to a NUL terminated buffer. */ |
| *((char *) __mempcpy (elf->state.ar.raw_name, ar_hdr->ar_name, 16)) = '\0'; |
| |
| elf_ar_hdr = &elf->state.ar.elf_ar_hdr; |
| |
| /* Now convert the `struct ar_hdr' into `Elf_Arhdr'. |
| Determine whether this is a special entry. */ |
| if (ar_hdr->ar_name[0] == '/') |
| { |
| if (ar_hdr->ar_name[1] == ' ' |
| && memcmp (ar_hdr->ar_name, "/ ", 16) == 0) |
| /* This is the index. */ |
| elf_ar_hdr->ar_name = memcpy (elf->state.ar.ar_name, "/", 2); |
| else if (ar_hdr->ar_name[1] == 'S' |
| && memcmp (ar_hdr->ar_name, "/SYM64/ ", 16) == 0) |
| /* 64-bit index. */ |
| elf_ar_hdr->ar_name = memcpy (elf->state.ar.ar_name, "/SYM64/", 8); |
| else if (ar_hdr->ar_name[1] == '/' |
| && memcmp (ar_hdr->ar_name, "// ", 16) == 0) |
| /* This is the array with the long names. */ |
| elf_ar_hdr->ar_name = memcpy (elf->state.ar.ar_name, "//", 3); |
| else if (likely (isdigit (ar_hdr->ar_name[1]))) |
| { |
| size_t offset; |
| |
| /* This is a long name. First we have to read the long name |
| table, if this hasn't happened already. */ |
| if (unlikely (elf->state.ar.long_names == NULL |
| && read_long_names (elf) == NULL)) |
| { |
| /* No long name table although it is reference. The archive is |
| broken. */ |
| __libelf_seterrno (ELF_E_INVALID_ARCHIVE); |
| return -1; |
| } |
| |
| offset = atol (ar_hdr->ar_name + 1); |
| if (unlikely (offset >= elf->state.ar.long_names_len)) |
| { |
| /* The index in the long name table is larger than the table. */ |
| __libelf_seterrno (ELF_E_INVALID_ARCHIVE); |
| return -1; |
| } |
| elf_ar_hdr->ar_name = elf->state.ar.long_names + offset; |
| } |
| else |
| { |
| /* This is none of the known special entries. */ |
| __libelf_seterrno (ELF_E_INVALID_ARCHIVE); |
| return -1; |
| } |
| } |
| else |
| { |
| char *endp; |
| |
| /* It is a normal entry. Copy over the name. */ |
| endp = (char *) memccpy (elf->state.ar.ar_name, ar_hdr->ar_name, |
| '/', 16); |
| if (endp != NULL) |
| endp[-1] = '\0'; |
| else |
| { |
| /* In the old BSD style of archive, there is no / terminator. |
| Instead, there is space padding at the end of the name. */ |
| size_t i = 15; |
| do |
| elf->state.ar.ar_name[i] = '\0'; |
| while (i > 0 && elf->state.ar.ar_name[--i] == ' '); |
| } |
| |
| elf_ar_hdr->ar_name = elf->state.ar.ar_name; |
| } |
| |
| if (unlikely (ar_hdr->ar_size[0] == ' ')) |
| /* Something is really wrong. We cannot live without a size for |
| the member since it will not be possible to find the next |
| archive member. */ |
| { |
| __libelf_seterrno (ELF_E_INVALID_ARCHIVE); |
| return -1; |
| } |
| |
| /* Since there are no specialized functions to convert ASCII to |
| time_t, uid_t, gid_t, mode_t, and off_t we use either atol or |
| atoll depending on the size of the types. We are also prepared |
| for the case where the whole field in the `struct ar_hdr' is |
| filled in which case we cannot simply use atol/l but instead have |
| to create a temporary copy. */ |
| |
| #define INT_FIELD(FIELD) \ |
| do \ |
| { \ |
| char buf[sizeof (ar_hdr->FIELD) + 1]; \ |
| const char *string = ar_hdr->FIELD; \ |
| if (ar_hdr->FIELD[sizeof (ar_hdr->FIELD) - 1] != ' ') \ |
| { \ |
| *((char *) __mempcpy (buf, ar_hdr->FIELD, sizeof (ar_hdr->FIELD))) \ |
| = '\0'; \ |
| string = buf; \ |
| } \ |
| if (sizeof (elf_ar_hdr->FIELD) <= sizeof (long int)) \ |
| elf_ar_hdr->FIELD = (__typeof (elf_ar_hdr->FIELD)) atol (string); \ |
| else \ |
| elf_ar_hdr->FIELD = (__typeof (elf_ar_hdr->FIELD)) atoll (string); \ |
| } \ |
| while (0) |
| |
| INT_FIELD (ar_date); |
| INT_FIELD (ar_uid); |
| INT_FIELD (ar_gid); |
| INT_FIELD (ar_mode); |
| INT_FIELD (ar_size); |
| |
| return 0; |
| } |
| |
| |
| /* We were asked to return a clone of an existing descriptor. This |
| function must be called with the lock on the parent descriptor |
| being held. */ |
| static Elf * |
| dup_elf (int fildes, Elf_Cmd cmd, Elf *ref) |
| { |
| struct Elf *result; |
| |
| if (fildes == -1) |
| /* Allow the user to pass -1 as the file descriptor for the new file. */ |
| fildes = ref->fildes; |
| /* The file descriptor better should be the same. If it was disconnected |
| already (using `elf_cntl') we do not test it. */ |
| else if (unlikely (ref->fildes != -1 && fildes != ref->fildes)) |
| { |
| __libelf_seterrno (ELF_E_FD_MISMATCH); |
| return NULL; |
| } |
| |
| /* The mode must allow reading. I.e., a descriptor creating with a |
| command different then ELF_C_READ, ELF_C_WRITE and ELF_C_RDWR is |
| not allowed. */ |
| if (unlikely (ref->cmd != ELF_C_READ && ref->cmd != ELF_C_READ_MMAP |
| && ref->cmd != ELF_C_WRITE && ref->cmd != ELF_C_WRITE_MMAP |
| && ref->cmd != ELF_C_RDWR && ref->cmd != ELF_C_RDWR_MMAP |
| && ref->cmd != ELF_C_READ_MMAP_PRIVATE)) |
| { |
| __libelf_seterrno (ELF_E_INVALID_OP); |
| return NULL; |
| } |
| |
| /* Now it is time to distinguish between reading normal files and |
| archives. Normal files can easily be handled be incrementing the |
| reference counter and return the same descriptor. */ |
| if (ref->kind != ELF_K_AR) |
| { |
| ++ref->ref_count; |
| return ref; |
| } |
| |
| /* This is an archive. We must create a descriptor for the archive |
| member the internal pointer of the archive file desriptor is |
| pointing to. First read the header of the next member if this |
| has not happened already. */ |
| if (ref->state.ar.elf_ar_hdr.ar_name == NULL |
| && __libelf_next_arhdr_wrlock (ref) != 0) |
| /* Something went wrong. Maybe there is no member left. */ |
| return NULL; |
| |
| /* We have all the information we need about the next archive member. |
| Now create a descriptor for it. */ |
| result = read_file (fildes, ref->state.ar.offset + sizeof (struct ar_hdr), |
| ref->state.ar.elf_ar_hdr.ar_size, cmd, ref); |
| |
| /* Enlist this new descriptor in the list of children. */ |
| if (result != NULL) |
| { |
| result->next = ref->state.ar.children; |
| ref->state.ar.children = result; |
| } |
| |
| return result; |
| } |
| |
| |
| /* Return desriptor for empty file ready for writing. */ |
| static struct Elf * |
| write_file (int fd, Elf_Cmd cmd) |
| { |
| /* We simply create an empty `Elf' structure. */ |
| #define NSCNSALLOC 10 |
| Elf *result = allocate_elf (fd, NULL, 0, 0, cmd, NULL, ELF_K_ELF, |
| NSCNSALLOC * sizeof (Elf_Scn)); |
| |
| if (result != NULL) |
| { |
| /* We have to write to the file in any case. */ |
| result->flags = ELF_F_DIRTY; |
| |
| /* Some more or less arbitrary value. */ |
| result->state.elf.scnincr = NSCNSALLOC; |
| |
| /* We have allocated room for some sections. */ |
| assert (offsetof (struct Elf, state.elf32.scns) |
| == offsetof (struct Elf, state.elf64.scns)); |
| result->state.elf.scns_last = &result->state.elf32.scns; |
| result->state.elf32.scns.max = NSCNSALLOC; |
| } |
| |
| return result; |
| } |
| |
| |
| /* Return a descriptor for the file belonging to FILDES. */ |
| Elf * |
| elf_begin (fildes, cmd, ref) |
| int fildes; |
| Elf_Cmd cmd; |
| Elf *ref; |
| { |
| Elf *retval; |
| |
| if (unlikely (! __libelf_version_initialized)) |
| { |
| /* Version wasn't set so far. */ |
| __libelf_seterrno (ELF_E_NO_VERSION); |
| return NULL; |
| } |
| |
| if (ref != NULL) |
| /* Make sure the descriptor is not suddenly going away. */ |
| rwlock_rdlock (ref->lock); |
| else if (unlikely (fcntl (fildes, F_GETFL) == -1 && errno == EBADF)) |
| { |
| /* We cannot do anything productive without a file descriptor. */ |
| __libelf_seterrno (ELF_E_INVALID_FILE); |
| return NULL; |
| } |
| |
| Elf *lock_dup_elf () |
| { |
| /* We need wrlock to dup an archive. */ |
| if (ref->kind == ELF_K_AR) |
| { |
| rwlock_unlock (ref->lock); |
| rwlock_wrlock (ref->lock); |
| } |
| |
| /* Duplicate the descriptor. */ |
| return dup_elf (fildes, cmd, ref); |
| } |
| |
| switch (cmd) |
| { |
| case ELF_C_NULL: |
| /* We simply return a NULL pointer. */ |
| retval = NULL; |
| break; |
| |
| case ELF_C_READ_MMAP_PRIVATE: |
| /* If we have a reference it must also be opened this way. */ |
| if (unlikely (ref != NULL && ref->cmd != ELF_C_READ_MMAP_PRIVATE)) |
| { |
| __libelf_seterrno (ELF_E_INVALID_CMD); |
| retval = NULL; |
| break; |
| } |
| /* FALLTHROUGH */ |
| |
| case ELF_C_READ: |
| case ELF_C_READ_MMAP: |
| if (ref != NULL) |
| retval = lock_dup_elf (); |
| else |
| /* Create descriptor for existing file. */ |
| retval = read_file (fildes, 0, ~((size_t) 0), cmd, NULL); |
| break; |
| |
| case ELF_C_RDWR: |
| case ELF_C_RDWR_MMAP: |
| /* If we have a REF object it must also be opened using this |
| command. */ |
| if (ref != NULL) |
| { |
| if (unlikely (ref->cmd != ELF_C_RDWR && ref->cmd != ELF_C_RDWR_MMAP |
| && ref->cmd != ELF_C_WRITE |
| && ref->cmd != ELF_C_WRITE_MMAP)) |
| { |
| /* This is not ok. REF must also be opened for writing. */ |
| __libelf_seterrno (ELF_E_INVALID_CMD); |
| retval = NULL; |
| } |
| else |
| retval = lock_dup_elf (); |
| } |
| else |
| /* Create descriptor for existing file. */ |
| retval = read_file (fildes, 0, ~((size_t) 0), cmd, NULL); |
| break; |
| |
| case ELF_C_WRITE: |
| case ELF_C_WRITE_MMAP: |
| /* We ignore REF and prepare a descriptor to write a new file. */ |
| retval = write_file (fildes, cmd); |
| break; |
| |
| default: |
| __libelf_seterrno (ELF_E_INVALID_CMD); |
| retval = NULL; |
| break; |
| } |
| |
| /* Release the lock. */ |
| if (ref != NULL) |
| rwlock_unlock (ref->lock); |
| |
| return retval; |
| } |
| INTDEF(elf_begin) |