Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 1 | ToDo list for elfutils -*-outline-*- |
| 2 | ---------------------- |
| 3 | |
Ulrich Drepper | 04af018 | 2009-02-05 22:09:34 -0800 | [diff] [blame] | 4 | Time-stamp: <2009-02-05 22:08:01 drepper> |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 5 | |
| 6 | * mkinstalldirs |
| 7 | |
| 8 | Remove everywhere. Use mkdir -p. |
| 9 | |
| 10 | * libelf: |
| 11 | |
| 12 | ** verify section |
| 13 | |
| 14 | Currently the elf_update function trusts the user blindly if the |
| 15 | ELF_F_LAYOUT flag is set. This is OK if the data is prepared by a |
| 16 | ELF_C_NULL call but not if the user prepared the data herself |
| 17 | |
| 18 | ** break out archive handling from elf_begin |
| 19 | |
| 20 | The handling of archives (especially of the symbol tables) must be |
| 21 | broken out of elf_begin. There are several different forms of |
| 22 | archives and only when having the archive handling separately this |
| 23 | remains maintainable. |
| 24 | |
Ulrich Drepper | cea5ee3 | 2009-02-01 16:19:06 -0800 | [diff] [blame] | 25 | ** shdrs in read-only files |
| 26 | |
| 27 | When reading (ELF_C_READ*) then there is no need to malloc Shdr |
| 28 | structure in elfXX_getshdr if file is mmaped and unaligned access |
| 29 | is allowed or the structure is aligned. Use ELF_F_MALLOCED flag |
| 30 | to differentiate. |
| 31 | |
Mark Wielaard | 030f626 | 2012-07-18 13:14:30 +0200 | [diff] [blame^] | 32 | ** shdrs after elf_cntl (ELF_C_FDREAD) |
| 33 | |
| 34 | Similar to the above. After ELF_C_FDREAD the file is completely |
| 35 | in memory. See also this mailing list thread: |
| 36 | https://fedorahosted.org/pipermail/elfutils-devel/2012-July/002368.html |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 37 | |
Ulrich Drepper | fbe998a | 2005-08-29 16:27:10 +0000 | [diff] [blame] | 38 | * libdw |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 39 | |
| 40 | ** More memory access checks needed |
| 41 | |
| 42 | All accesses to the debug sections should make sure the offsets are |
| 43 | valid. This is currently especially a problem with leb128 accesses. |
| 44 | |
| 45 | ** Low level macro information operations |
| 46 | |
| 47 | in 5.11.3 are not implemented. gcc currently does not emit this |
| 48 | information so I cannot test it. |
| 49 | |
| 50 | ** Rename dwarf_getabbrev |
| 51 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 52 | |
Ulrich Drepper | 3bf5759 | 2009-01-01 18:52:05 -0800 | [diff] [blame] | 53 | * libcpu |
| 54 | |
| 55 | ** x86 |
| 56 | |
| 57 | *** Opcodes |
| 58 | |
| 59 | crc32 |
| 60 | extractps |
| 61 | pextrb |
| 62 | pextrd/pextrq |
| 63 | pextrw |
| 64 | pinsrq |
| 65 | popcnt 64-bit reg |
| 66 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 67 | * nm: |
| 68 | |
| 69 | ** add demangler support |
| 70 | |
| 71 | Use demangler from libiberty. |
| 72 | |
| 73 | ** add support to read debugging symbols |
| 74 | |
| 75 | Implement -l option for BSD and POSIX format |
| 76 | |
| 77 | |
| 78 | * strip: |
| 79 | |
| 80 | ** support SHT_SYMTAB_SHNDX |
| 81 | |
| 82 | should be removed if not needed anymore |
| 83 | |
| 84 | * ld: |
| 85 | |
| 86 | ** sanity check .rel sh_info content |
| 87 | |
| 88 | the sh_info of all .rel sections with the same name must point to |
| 89 | sections which also have the same name |
| 90 | |
| 91 | ** use ld.so.conf |
| 92 | |
| 93 | to locate shared libraries also use /etc/ld.so.conf |
| 94 | |
| 95 | ** handle object files for different architectures |
| 96 | |
| 97 | ld.so is expected to ignore object files for different architectures and |
| 98 | continue looking for a matching file (e.g., ignore 32-bit binaries on |
| 99 | 64-bit platforms and vice versa). We probably need the same in ld. |
| 100 | |
| 101 | ** reuse after elf_end |
| 102 | |
| 103 | Some files are closed using elf_end. They are removed from memory only |
| 104 | if no reference is left (especially for archives this is a problem). |
| 105 | The old mapping should be reused in that case. The problem is worse |
| 106 | for files which are not mapped read-only (archives again). |
| 107 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 108 | ** size for STT_SECTION entries |
| 109 | |
| 110 | The STT_SECTION entries have zero for the size but can easily get |
| 111 | the size of the section. |
| 112 | |
Ulrich Drepper | cbb51e5 | 2006-06-11 20:25:23 +0000 | [diff] [blame] | 113 | ** .eh_frame_hdr |
| 114 | |
| 115 | Not implemented at all in the moment except for recognition of the option |
| 116 | itself. |
| 117 | |
Ulrich Drepper | c2e71c5 | 2009-02-05 21:41:11 -0800 | [diff] [blame] | 118 | ** variables with aliases in executables |
| 119 | |
| 120 | When linking an executable with a references against a variable in a |
| 121 | DSO, create symbol table entries for all the aliases of the variable |
| 122 | in the DSO and create a relocation for one of them (a non-weak |
| 123 | definition) |
| 124 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 125 | * elflint |
| 126 | |
| 127 | ** additional checks |
| 128 | |
| 129 | 1st GOT entry == _DYNAMIC |
| 130 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 131 | check versioning info: |
| 132 | |
| 133 | always BASE in verdef |
| 134 | sh_size/sh_entsize matches last offset != 0 |
| 135 | |
| 136 | check whether any relocation is for a merge-able section |
| 137 | |
Ulrich Drepper | 834de6f | 2008-02-23 06:50:01 +0000 | [diff] [blame] | 138 | check TLS relocation depencies |
| 139 | |
Ulrich Drepper | 3a52c7a | 2009-01-16 21:11:49 -0800 | [diff] [blame] | 140 | Check content of .eh_frame_hdr, .eh_frame, .gcc_except_table |
| 141 | |
Ulrich Drepper | 834de6f | 2008-02-23 06:50:01 +0000 | [diff] [blame] | 142 | *** for x86 |
| 143 | |
| 144 | check that R_386_TLS_GD is followed by R_386_PLT32 for __tls_get_addr |
| 145 | |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 146 | ** relax |
| 147 | |
| 148 | prelink generated files |
| 149 | |
Ulrich Drepper | fbe998a | 2005-08-29 16:27:10 +0000 | [diff] [blame] | 150 | * elfcmp |
| 151 | |
| 152 | ** treat relocation sections special |
| 153 | |
| 154 | Differences in the relocation sections can be ignored if all |
| 155 | the same symbols with the same targets are present and the order |
| 156 | of overlapping relocations doesn't change. There really never |
| 157 | should be overlapping relocations but who knows. |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 158 | |
| 159 | * mcs |
| 160 | |
| 161 | Sun has it. Can modify sections which are not in segments. |
| 162 | |
| 163 | -a string |
| 164 | Append string to the comment section of the ELF object |
| 165 | files. If string contains embedded blanks, it must be |
| 166 | enclosed in quotation marks. |
| 167 | |
| 168 | -c Compress the contents of the comment section of the |
| 169 | ELF object files. All duplicate entries are removed. |
| 170 | The ordering of the remaining entries is not dis- |
| 171 | turbed. |
| 172 | |
| 173 | -d Delete the contents of the comment section from the |
| 174 | ELF object files. The section header for the comment |
| 175 | section is also removed. |
| 176 | |
| 177 | -n name |
| 178 | Specify the name of the comment section to access if |
| 179 | other than .comment. By default, mcs deals with the |
| 180 | section named .comment. This option can be used to |
| 181 | specify another section. mcs can take multiple -n |
| 182 | options to allow for specification of multiple sec- |
| 183 | tion comments. |
| 184 | |
| 185 | -p Print the contents of the comment section on the stan- |
| 186 | dard output. Each section printed is tagged by the |
| 187 | name of the file from which it was extracted, using |
| 188 | the format file[member_name]: for archive files and |
| 189 | file: for other files. |
| 190 | |
| 191 | -V Print on standard error the version number of mcs. |
| 192 | |
| 193 | Local Variables: |
Ulrich Drepper | 3bf5759 | 2009-01-01 18:52:05 -0800 | [diff] [blame] | 194 | eval:(hide-sublevels 3) |
Ulrich Drepper | b08d5a8 | 2005-07-26 05:00:05 +0000 | [diff] [blame] | 195 | End: |