Brian Silverman | 8649792 | 2018-02-10 19:28:39 -0500 | [diff] [blame^] | 1 | Version 0.170 |
| 2 | |
| 3 | libdw: Added new DWARF5 attribute, tag, character encoding, language code, |
| 4 | calling convention, defaulted member function and macro constants |
| 5 | to dwarf.h. |
| 6 | New functions dwarf_default_lower_bound and dwarf_line_file. |
| 7 | dwarf_peel_type now handles DWARF5 immutable, packed and shared tags. |
| 8 | dwarf_getmacros now handles DWARF5 .debug_macro sections. |
| 9 | |
| 10 | strip: Add -R, --remove-section=SECTION and --keep-section=SECTION. |
| 11 | |
| 12 | backends: The bpf disassembler is now always build on all platforms. |
| 13 | |
| 14 | Version 0.169 |
| 15 | |
| 16 | backends: Add support for EM_PPC64 GNU_ATTRIBUTES. |
| 17 | Frame pointer unwinding fallback support for i386, x86_64, aarch64. |
| 18 | |
| 19 | translations: Update Polish translation. |
| 20 | |
| 21 | Version 0.168 |
| 22 | |
| 23 | http://elfutils.org/ is now hosted at http://sourceware.org/elfutils/ |
| 24 | |
| 25 | libelf: gelf_newehdr and gelf_newehdr now return void *. |
| 26 | |
| 27 | libdw: dwarf.h corrected the DW_LANG_PLI constant name (was DW_LANG_PL1). |
| 28 | |
| 29 | readelf: Add optional --symbols[=SECTION] argument to select section name. |
| 30 | |
| 31 | Version 0.167 |
| 32 | |
| 33 | libasm: Add eBPF disassembler for EM_BPF files. |
| 34 | |
| 35 | backends: Add m68k and BPF backends. |
| 36 | |
| 37 | ld: Removed. |
| 38 | |
| 39 | dwelf: Add ELF/DWARF string table creation functions. dwelf_strtab_init, |
| 40 | dwelf_strtab_add, dwelf_strtab_add_len, dwelf_strtab_finalize, |
| 41 | dwelf_strent_off, dwelf_strent_str and dwelf_strtab_free. |
| 42 | |
| 43 | Version 0.166 |
| 44 | |
| 45 | config: The default program prefix for the installed tools is now eu-. |
| 46 | Use configure --program-prefix="" to not use a program prefix. |
| 47 | |
| 48 | Version 0.165 |
| 49 | |
| 50 | elfcompress: New utility to compress or decompress ELF sections. |
| 51 | |
| 52 | readelf: Add -z,--decompress option. |
| 53 | |
| 54 | libelf: Add elf_compress, elf_compress_gnu, elf32_getchdr, elf64_getchdr |
| 55 | and gelf_getchdr. |
| 56 | |
| 57 | libdwelf: New function dwelf_scn_gnu_compressed_size. |
| 58 | |
| 59 | config: Add libelf and libdw pkg-config files. |
| 60 | |
| 61 | backends: sparc support for core and live backtraces. |
| 62 | |
| 63 | translations: Updated Polish translation. |
| 64 | |
| 65 | Version 0.164 |
| 66 | |
| 67 | strip, unstrip: Handle ELF files with merged strtab/shstrtab tables. |
| 68 | Handle missing SHF_INFO_LINK section flags. |
| 69 | |
| 70 | libelf: Use int64_t for offsets in libelf.h instead of loff_t. |
| 71 | |
| 72 | libdw: dwarf.h Add preliminary DWARF5 DW_LANG_Haskell. |
| 73 | |
| 74 | libdwfl: dwfl_standard_find_debuginfo now searches any subdir of the binary |
| 75 | path under the debuginfo root when the separate debug file couldn't |
| 76 | be found by build-id. |
| 77 | dwfl_linux_proc_attach can now be called before any Dwfl_Modules |
| 78 | have been reported. |
| 79 | |
| 80 | backends: Better sparc and sparc64 support. |
| 81 | |
| 82 | translations: Updated Ukrainian translation. |
| 83 | |
| 84 | Provide default-yama-scope subpackage. |
| 85 | |
| 86 | Version 0.163 |
| 87 | |
| 88 | Bug fixes only, no new features. |
| 89 | |
| 90 | Version 0.162 |
| 91 | |
| 92 | libdw: Install new header elfutils/known-dwarf.h. |
| 93 | dwarf.h Add preliminary DWARF5 constants DW_TAG_atomic_type, |
| 94 | DW_LANG_Fortran03, DW_LANG_Fortran08. dwarf_peel_type now also |
| 95 | handles DW_TAG_atomic_type. |
| 96 | |
| 97 | addr2line: Input addresses are now always interpreted as hexadecimal |
| 98 | numbers, never as octal or decimal numbers. |
| 99 | New option -a, --addresses to print address before each entry. |
| 100 | New option -C, --demangle to show demangled symbols. |
| 101 | New option --pretty-print to print all information on one line. |
| 102 | |
| 103 | ar: CVE-2014-9447 Directory traversal vulnerability in ar extraction. |
| 104 | |
| 105 | backends: x32 support. |
| 106 | |
| 107 | Version 0.161 |
| 108 | |
| 109 | libdw: New function dwarf_peel_type. dwarf_aggregate_size now uses |
| 110 | dwarf_peel_type to also provide the sizes of qualified types. |
| 111 | dwarf_getmacros will now serve either of .debug_macro and |
| 112 | .debug_macinfo transparently. New interfaces |
| 113 | dwarf_getmacros_off, dwarf_macro_getsrcfiles, |
| 114 | dwarf_macro_getparamcnt, and dwarf_macro_param are available |
| 115 | for more generalized inspection of macros and their parameters. |
| 116 | dwarf.h: Add DW_AT_GNU_deleted, DW_AT_noreturn, DW_LANG_C11, |
| 117 | DW_LANG_C_plus_plus_11 and DW_LANG_C_plus_plus_14. |
| 118 | |
| 119 | Version 0.160 |
| 120 | |
| 121 | libdw: New functions dwarf_cu_getdwarf, dwarf_cu_die. |
| 122 | dwarf.h remove non-existing DW_TAG_mutable_type. |
| 123 | |
| 124 | libdwfl: Handle LZMA .ko.xz compressed kernel modules. |
| 125 | |
| 126 | unstrip: New option -F, --force to combining files even if some ELF headers |
| 127 | don't seem to match. |
| 128 | |
| 129 | backends: Handle ARM THUMB functions. Add support for ppc64le ELFv2 abi. |
| 130 | |
| 131 | Version 0.159 |
| 132 | |
| 133 | stack: New option -d, --debugname to lookup DWARF debuginfo name for frame. |
| 134 | New option -i, --inlines to show inlined frames using DWARF debuginfo. |
| 135 | |
| 136 | libdwelf: New libdwelf.h header for libdw.so DWARF ELF Low-level Functions. |
| 137 | New function dwelf_elf_gnu_debuglink, dwelf_dwarf_gnu_debugaltlink, |
| 138 | and dwelf_elf_gnu_build_id. |
| 139 | |
| 140 | libdw: Support for DWZ multifile forms DW_FORM_GNU_ref_alt and |
| 141 | DW_FORM_GNU_strp_alt is now enabled by default and no longer |
| 142 | experimental. Added new functions dwarf_getalt and dwarf_setalt |
| 143 | to get or set the alternative debug file used for the alt FORMs. |
| 144 | The dwfl_linux_proc_find_elf callback will now find ELF from |
| 145 | process memory for (deleted) files if the Dwfl has process state |
| 146 | attached. |
| 147 | |
| 148 | libdwfl: The dwfl_build_id_find_debuginfo and dwfl_standard_find_debuginfo |
| 149 | functions will now try to resolve and set the alternative debug file. |
| 150 | |
| 151 | backends: Add CFI unwinding for arm. Relies on .debug_frame. |
| 152 | Add arm process initial register state compatible mode to AARCH64. |
| 153 | Add aarch64 native and core unwind support. |
| 154 | |
| 155 | other: All separate elfutils-robustify patches have been merged. |
| 156 | CVE-2014-0172 Check overflow before calling malloc to uncompress data. |
| 157 | |
| 158 | Version 0.158 |
| 159 | |
| 160 | libdwfl: dwfl_core_file_report has new parameter executable. |
| 161 | New functions dwfl_module_getsymtab_first_global, |
| 162 | dwfl_module_getsym_info and dwfl_module_addrinfo. |
| 163 | Added unwinder with type Dwfl_Thread_Callbacks, opaque types |
| 164 | Dwfl_Thread and Dwfl_Frame and functions dwfl_attach_state, |
| 165 | dwfl_pid, dwfl_thread_dwfl, dwfl_thread_tid, dwfl_frame_thread, |
| 166 | dwfl_thread_state_registers, dwfl_thread_state_register_pc, |
| 167 | dwfl_getthread_frames, dwfl_getthreads, dwfl_thread_getframes |
| 168 | and dwfl_frame_pc. |
| 169 | |
| 170 | addr2line: New option -x to show the section an address was found in. |
| 171 | |
| 172 | stack: New utility that uses the new unwinder for processes and cores. |
| 173 | |
| 174 | backends: Unwinder support for i386, x86_64, s390, s390x, ppc and ppc64. |
| 175 | aarch64 support. |
| 176 | |
| 177 | Version 0.157 |
| 178 | |
| 179 | libdw: Add new functions dwarf_getlocations, dwarf_getlocation_attr |
| 180 | and dwarf_getlocation_die. |
| 181 | |
| 182 | readelf: Show contents of NT_SIGINFO and NT_FILE core notes. |
| 183 | |
| 184 | addr2line: Support -i, --inlines output option. |
| 185 | |
| 186 | backends: abi_cfi hook for arm, ppc and s390. |
| 187 | |
| 188 | Version 0.156 |
| 189 | |
| 190 | lib: New macro COMPAT_VERSION_NEWPROTO. |
| 191 | |
| 192 | libdw: Handle GNU extension opcodes in dwarf_getlocation. |
| 193 | |
| 194 | libdwfl: Fix STB_GLOBAL over STB_WEAK preference in dwfl_module_addrsym. |
| 195 | Add minisymtab support. |
| 196 | Add parameter add_p_vaddr to dwfl_report_elf. |
| 197 | Use DT_DEBUG library search first. |
| 198 | |
| 199 | libebl: Handle new core note types in EBL. |
| 200 | |
| 201 | backends: Interpret NT_ARM_VFP. |
| 202 | Implement core file registers parsing for s390/s390x. |
| 203 | |
| 204 | readelf: Add --elf-section input option to inspect an embedded ELF file. |
| 205 | Add -U, --unresolved-address-offsets output control. |
| 206 | Add --debug-dump=decodedline support. |
| 207 | Accept version 8 .gdb_index section format. |
| 208 | Adjust output formatting width. |
| 209 | When highpc is in constant form print it also as address. |
| 210 | Display raw .debug_aranges. Use libdw only for decodedaranges. |
| 211 | |
| 212 | elflint: Add __bss_start__ to the list of allowed symbols. |
| 213 | |
| 214 | tests: Add configure --enable-valgrind option to run all tests under valgrind. |
| 215 | Enable automake parallel-tests for make check. |
| 216 | |
| 217 | translations: Updated Polish translation. |
| 218 | |
| 219 | Updates for Automake 1.13. |
| 220 | |
| 221 | Version 0.155 |
| 222 | |
| 223 | libelf: elf*_xlatetomd now works for cross-endian ELF note data. |
| 224 | elf_getshdr now works consistently on non-mmaped ELF files after |
| 225 | calling elf_cntl(ELF_C_FDREAD). |
| 226 | Implement support for ar archives with 64-bit symbol table. |
| 227 | |
| 228 | libdw: dwarf.h corrected the DW_LANG_ObjC constant name (was DW_LANG_Objc). |
| 229 | Any existing sources using the old name will have to be updated. |
| 230 | Add DW_MACRO_GNU .debug_macro type encodings constants, DW_ATE_UTF |
| 231 | and DW_OP_GNU_parameter_ref to dwarf.h. |
| 232 | Experimental support for DWZ multifile forms DW_FORM_GNU_ref_alt |
| 233 | and DW_FORM_GNU_strp_alt. Disabled by default. Use configure |
| 234 | --enable-dwz to test it. |
| 235 | |
| 236 | readelf: Add .debug_macro parsing support. |
| 237 | Add .gdb_index version 7 parsing support. |
| 238 | Recognize DW_OP_GNU_parameter_ref. |
| 239 | |
| 240 | backends: Add support for Tilera TILE-Gx processor. |
| 241 | |
| 242 | translations: Updated Ukrainian translation. |
| 243 | |
| 244 | Version 0.154 |
| 245 | |
| 246 | libelf: [g]elf[32|64]_offscn() do not match SHT_NOBITS sections at OFFSET. |
| 247 | |
| 248 | libdw: dwarf_highpc function now handles DWARF 4 DW_AT_high_pc constant form. |
| 249 | Fix bug using dwarf_next_unit to iterate over .debug_types. |
| 250 | |
| 251 | elflint: Now accepts gold linker produced executables. |
| 252 | |
| 253 | The license is now GPLv2/LGPLv3+ for the libraries and GPLv3+ for stand-alone |
| 254 | programs. There is now also a formal CONTRIBUTING document describing how to |
| 255 | submit patches. |
| 256 | |
| 257 | Version 0.153 |
| 258 | |
| 259 | libdw: Support reading .zdebug_* DWARF sections compressed via zlib. |
| 260 | |
| 261 | libdwfl: Speed up dwfl_module_addrsym. |
| 262 | |
| 263 | nm: Support C++ demangling. |
| 264 | |
| 265 | ar: Support D modifier for "deterministic output" with no uid/gid/mtime info. |
| 266 | The U modifier is the inverse. |
| 267 | elfutils can be configured with the --enable-deterministic-archives |
| 268 | option to make the D behavior the default when U is not specified. |
| 269 | |
| 270 | ranlib: Support -D and -U flags with same meaning. |
| 271 | |
| 272 | readelf: Improve output of -wline. Add support for printing SDT elf notes. |
| 273 | Add printing of .gdb_index section. |
| 274 | Support for typed DWARF stack, call_site and entry_value. |
| 275 | |
| 276 | strip: Add --reloc-debug-sections option. |
| 277 | Improved SHT_GROUP sections handling. |
| 278 | |
| 279 | Version 0.152 |
| 280 | |
| 281 | Various build and warning nits fixed for newest GCC and Autoconf. |
| 282 | |
| 283 | libdwfl: Yet another prelink-related fix for another regression. |
| 284 | Look for Linux kernel images in files named with compression suffixes. |
| 285 | |
| 286 | elfcmp: New flag --ignore-build-id to ignore differing build ID bits. |
| 287 | New flag -l/--verbose to print all differences. |
| 288 | |
| 289 | Version 0.151 |
| 290 | |
| 291 | libdwfl: Fix for more prelink cases with separate debug file. |
| 292 | |
| 293 | strip: New flag --strip-sections to remove section headers entirely. |
| 294 | |
| 295 | Version 0.150 |
| 296 | |
| 297 | libdw: Fix for handling huge .debug_aranges section. |
| 298 | |
| 299 | libdwfl: Fix for handling prelinked DSO with separate debug file. |
| 300 | |
| 301 | findtextrel: Fix diagnostics to work with usual section ordering. |
| 302 | |
| 303 | libebl: i386 backend fix for multi-register integer return value location. |
| 304 | |
| 305 | Version 0.149: |
| 306 | |
| 307 | libdw: Decode new DW_OP_GNU_implicit_pointer operation; |
| 308 | new function dwarf_getlocation_implicit_pointer. |
| 309 | |
| 310 | libdwfl: New function dwfl_dwarf_line. |
| 311 | |
| 312 | addr2line: New flag -F/--flags to print more DWARF line information details. |
| 313 | |
| 314 | strip: -g recognizes .gdb_index as a debugging section. |
| 315 | |
| 316 | Version 0.148: |
| 317 | |
| 318 | libdw: Accept DWARF 4 format: new functions dwarf_next_unit, dwarf_offdie_types. |
| 319 | New functions dwarf_lineisa, dwarf_linediscriminator, dwarf_lineop_index. |
| 320 | |
| 321 | libdwfl: Fixes in core-file handling, support cores from PIEs. |
| 322 | When working from build IDs, don't open a named file that mismatches. |
| 323 | |
| 324 | readelf: Handle DWARF 4 formats. |
| 325 | |
| 326 | Version 0.147: |
| 327 | |
| 328 | libdw: Fixes in CFI handling, best possible handling of bogus CFA ops. |
| 329 | |
| 330 | libdwfl: Ignore R_*_NONE relocs, works around old (binutils) ld -r bugs. |
| 331 | |
| 332 | Version 0.146: |
| 333 | |
| 334 | libdwfl: New function dwfl_core_file_report. |
| 335 | |
| 336 | Version 0.145: |
| 337 | |
| 338 | Fix build with --disable-dependency-tracking. |
| 339 | |
| 340 | Fix build with most recent glibc headers. |
| 341 | |
| 342 | libelf: More robust to bogus section headers. |
| 343 | |
| 344 | libdw: Fix CFI decoding. |
| 345 | |
| 346 | libdwfl: Fix address bias returned by CFI accessors. |
| 347 | Fix core file module layout identification. |
| 348 | |
| 349 | readelf: Fix CFI decoding. |
| 350 | |
| 351 | Version 0.144: |
| 352 | |
| 353 | libelf: New function elf_getphdrnum. |
| 354 | Now support using more than 65536 program headers in a file. |
| 355 | |
| 356 | libdw: New function dwarf_aggregate_size for computing (constant) type |
| 357 | sizes, including array_type cases with nontrivial calculation. |
| 358 | |
| 359 | readelf: Don't give errors for missing info under -a. |
| 360 | Handle Linux "VMCOREINFO" notes under -n. |
| 361 | |
| 362 | Version 0.143: |
| 363 | |
| 364 | libdw: Various convenience functions for individual attributes now use |
| 365 | dwarf_attr_integrate to look up indirect inherited attributes. |
| 366 | Location expression handling now supports DW_OP_implicit_value. |
| 367 | |
| 368 | libdwfl: Support automatic decompression of files in XZ format, |
| 369 | and of Linux kernel images made with bzip2 or LZMA (as well as gzip). |
| 370 | |
| 371 | Version 0.142: |
| 372 | |
| 373 | libelf: Add elf_getshdrnum alias for elf_getshnum and elf_getshdrstrndx alias |
| 374 | for elf_getshstrndx and deprecate original names. Sun screwed up |
| 375 | their implementation and asked for a solution. |
| 376 | |
| 377 | libebl: Add support for STB_GNU_UNIQUE. |
| 378 | |
| 379 | elflint: Add support for STB_GNU_UNIQUE. |
| 380 | |
| 381 | readelf: Add -N option, speeds up DWARF printing without address->name lookups. |
| 382 | |
| 383 | libdw: Add support for decoding DWARF CFI into location description form. |
| 384 | Handle some new DWARF 3 expression operations previously omitted. |
| 385 | Basic handling of some new encodings slated for DWARF 4. |
| 386 | |
| 387 | Version 0.141: |
| 388 | |
| 389 | libebl: sparc backend fixes; |
| 390 | some more arm backend support |
| 391 | |
| 392 | libdwfl: fix dwfl_module_build_id for prelinked DSO case; |
| 393 | fixes in core file support; |
| 394 | dwfl_module_getsym interface improved for non-address symbols |
| 395 | |
| 396 | strip: fix infinite loop on strange inputs with -f |
| 397 | |
| 398 | addr2line: take -j/--section=NAME option for binutils compatibility |
| 399 | (same effect as '(NAME)0x123' syntax already supported) |
| 400 | |
| 401 | Version 0.140: |
| 402 | |
| 403 | libelf: Fix regression in creation of section header |
| 404 | |
| 405 | libdwfl: Less strict behavior if DWARF reader ist just used to display data |
| 406 | |
| 407 | Version 0.139: |
| 408 | |
| 409 | libcpu: Add Intel SSE4 disassembler support |
| 410 | |
| 411 | readelf: Implement call frame information and exception handling dumping. |
| 412 | Add -e option. Enable it implicitly for -a. |
| 413 | |
| 414 | elflint: Check PT_GNU_EH_FRAME program header entry. |
| 415 | |
| 416 | libdwfl: Support automatic gzip/bzip2 decompression of ELF files. |
| 417 | |
| 418 | Version 0.138: |
| 419 | |
| 420 | Install <elfutils/version.h> header file for applications to use in source |
| 421 | version compatibility checks. |
| 422 | |
| 423 | libebl: backend fixes for i386 TLS relocs; backend support for NT_386_IOPERM |
| 424 | |
| 425 | libcpu: disassembler fixes |
| 426 | |
| 427 | libdwfl: bug fixes |
| 428 | |
| 429 | libelf: bug fixes |
| 430 | |
| 431 | nm: bug fixes for handling corrupt input files |
| 432 | |
| 433 | Version 0.137: |
| 434 | |
| 435 | Minor fixes for unreleased 0.136 release. |
| 436 | |
| 437 | Version 0.136: |
| 438 | |
| 439 | libdwfl: bug fixes; new "segment" interfaces; |
| 440 | all the libdwfl-based tools now support --core=COREFILE option |
| 441 | |
| 442 | Version 0.135: |
| 443 | |
| 444 | libdwfl: bug fixes |
| 445 | |
| 446 | strip: changed handling of ET_REL files wrt symbol tables and relocs |
| 447 | |
| 448 | Version 0.134: |
| 449 | |
| 450 | elflint: backend improvements for sparc, alpha |
| 451 | |
| 452 | libdwfl, libelf: bug fixes |
| 453 | |
| 454 | Version 0.133: |
| 455 | |
| 456 | readelf, elflint, libebl: SHT_GNU_ATTRIBUTE section handling (readelf -A) |
| 457 | |
| 458 | readelf: core note handling for NT_386_TLS, NT_PPC_SPE, Alpha NT_AUXV |
| 459 | |
| 460 | libdwfl: bug fixes and optimization in relocation handling |
| 461 | |
| 462 | elfcmp: bug fix for non-allocated section handling |
| 463 | |
| 464 | ld: implement newer features of binutils linker. |
| 465 | |
| 466 | Version 0.132: |
| 467 | |
| 468 | libcpu: Implement x86 and x86-64 disassembler. |
| 469 | libasm: Add interface for disassembler. |
| 470 | |
| 471 | all programs: add debugging of branch prediction. |
| 472 | |
| 473 | libelf: new function elf_scnshndx. |
| 474 | |
| 475 | Version 0.131: |
| 476 | |
| 477 | libdw: DW_FORM_ref_addr support; dwarf_formref entry point now deprecated; |
| 478 | bug fixes for oddly-formatted DWARF |
| 479 | |
| 480 | libdwfl: bug fixes in offline archive support, symbol table handling; |
| 481 | apply partial relocations for dwfl_module_address_section on ET_REL |
| 482 | |
| 483 | libebl: powerpc backend support for Altivec registers |
| 484 | |
| 485 | Version 0.130: |
| 486 | |
| 487 | readelf: -p option can take an argument like -x for one section, |
| 488 | or no argument (as before) for all SHF_STRINGS sections; |
| 489 | new option --archive-index (or -c); |
| 490 | improved -n output for core files, on many machines |
| 491 | |
| 492 | libelf: new function elf_getdata_rawchunk, replaces gelf_rawchunk; |
| 493 | new functions gelf_getnote, gelf_getauxv, gelf_update_auxv |
| 494 | |
| 495 | readelf, elflint: handle SHT_NOTE sections without requiring phdrs |
| 496 | |
| 497 | elflint: stricter checks on debug sections |
| 498 | |
| 499 | libdwfl: new functions dwfl_build_id_find_elf, dwfl_build_id_find_debuginfo, |
| 500 | dwfl_module_build_id, dwfl_module_report_build_id; |
| 501 | support dynamic symbol tables found via phdrs; |
| 502 | dwfl_standard_find_debuginfo now uses build IDs when available |
| 503 | |
| 504 | unstrip: new option --list (or -n) |
| 505 | |
| 506 | libebl: backend improvements for sparc, alpha, powerpc |
| 507 | |
| 508 | Version 0.129: |
| 509 | |
| 510 | readelf: new options --hex-dump (or -x), --strings (or -p) |
| 511 | |
| 512 | addr2line: new option --symbols (or -S) |
| 513 | |
| 514 | Version 0.128: |
| 515 | |
| 516 | new program: unstrip |
| 517 | |
| 518 | elfcmp: new option --hash-inexact |
| 519 | |
| 520 | Version 0.127: |
| 521 | |
| 522 | libdw: new function dwarf_getsrcdirs |
| 523 | |
| 524 | libdwfl: new functions dwfl_module_addrsym, dwfl_report_begin_add, |
| 525 | dwfl_module_address_section |
| 526 | |
| 527 | Version 0.126: |
| 528 | |
| 529 | new program: ar |
| 530 | |
| 531 | Version 0.125: |
| 532 | |
| 533 | elflint: Compare DT_GNU_HASH tests. |
| 534 | |
| 535 | move archives into -static RPMs |
| 536 | |
| 537 | libelf, elflint: better support for core file handling |
| 538 | |
| 539 | Version 0.124: |
| 540 | |
| 541 | libebl: sparc backend support for return value location |
| 542 | |
| 543 | libebl, libdwfl: backend register name support extended with more info |
| 544 | |
| 545 | libelf, libdw: bug fixes for unaligned accesses on machines that care |
| 546 | |
| 547 | readelf, elflint: trivial bugs fixed |
| 548 | |
| 549 | Version 0.123: |
| 550 | |
| 551 | libebl: Backend build fixes, thanks to Stepan Kasal. |
| 552 | |
| 553 | libebl: ia64 backend support for register names, return value location |
| 554 | |
| 555 | libdwfl: Handle truncated linux kernel module section names. |
| 556 | |
| 557 | libdwfl: Look for linux kernel "vmlinux" files with ".debug" suffix. |
| 558 | |
| 559 | elflint: Fix checks to permit --hash-style=gnu format. |
| 560 | |
| 561 | Version 0.122: |
| 562 | |
| 563 | libebl: add function to test for relative relocation |
| 564 | |
| 565 | elflint: fix and extend DT_RELCOUNT/DT_RELACOUNT checks |
| 566 | |
| 567 | elflint, readelf: add support for DT_GNU_HASH |
| 568 | libelf: add elf_gnu_hash |
| 569 | |
| 570 | elflint, readelf: add support for 64-bit SysV-style hash tables |
| 571 | |
| 572 | libdwfl: new functions dwfl_module_getsymtab, dwfl_module_getsym. |
| 573 | |
| 574 | Version 0.121: |
| 575 | |
| 576 | libelf: bug fixes for rewriting existing files when using mmap. |
| 577 | |
| 578 | make all installed headers usable in C++ code. |
| 579 | |
| 580 | readelf: better output format. |
| 581 | |
| 582 | elflint: fix tests of dynamic section content. |
| 583 | |
| 584 | ld: Implement --as-needed, --execstack, PT_GNU_STACK. Many small patches. |
| 585 | |
| 586 | libdw, libdwfl: handle files without aranges info. |
| 587 | |
| 588 | Version 0.120: |
| 589 | |
| 590 | Bug fixes. |
| 591 | |
| 592 | dwarf.h updated for DWARF 3.0 final specification. |
| 593 | |
| 594 | libdwfl: New function dwfl_version. |
| 595 | |
| 596 | The license is now GPL for most files. The libelf, libebl, libdw, |
| 597 | and libdwfl libraries have additional exceptions. Add reference to |
| 598 | OIN. |
| 599 | |
| 600 | Version 0.119: |
| 601 | |
| 602 | bug fixes |
| 603 | |
| 604 | Version 0.118: |
| 605 | |
| 606 | elflint: more tests. |
| 607 | |
| 608 | libdwfl: New function dwfl_module_register_names. |
| 609 | |
| 610 | libebl: New backend hook for register names. |
| 611 | |
| 612 | Version 0.117: |
| 613 | |
| 614 | libdwfl: New function dwfl_module_return_value_location. |
| 615 | |
| 616 | libebl: Backend improvements for several CPUs. |
| 617 | |
| 618 | Version 0.116: |
| 619 | |
| 620 | libdw: New functions dwarf_ranges, dwarf_entrypc, dwarf_diecu, |
| 621 | dwarf_entry_breakpoints. Removed Dwarf_Func type and functions |
| 622 | dwarf_func_name, dwarf_func_lowpc, dwarf_func_highpc, |
| 623 | dwarf_func_entrypc, dwarf_func_die; dwarf_getfuncs callback now uses |
| 624 | Dwarf_Die, and dwarf_func_file, dwarf_func_line, dwarf_func_col |
| 625 | replaced by dwarf_decl_file, dwarf_decl_line, dwarf_decl_column; |
| 626 | dwarf_func_inline, dwarf_func_inline_instances now take Dwarf_Die. |
| 627 | Type Dwarf_Loc renamed to Dwarf_Op; dwarf_getloclist, |
| 628 | dwarf_addrloclists renamed dwarf_getlocation, dwarf_getlocation_addr. |
| 629 | |
| 630 | Version 0.115: |
| 631 | |
| 632 | libelf: speed-ups of non-mmap reading. |
| 633 | |
| 634 | strings: New program. |
| 635 | |
| 636 | Implement --enable-gcov option for configure. |
| 637 | |
| 638 | libdw: New function dwarf_getscopes_die. |
| 639 | |
| 640 | Version 0.114: |
| 641 | |
| 642 | libelf: new function elf_getaroff |
| 643 | |
| 644 | libdw: Added dwarf_func_die, dwarf_func_inline, dwarf_func_inline_instances. |
| 645 | |
| 646 | libdwfl: New functions dwfl_report_offline, dwfl_offline_section_address, |
| 647 | dwfl_linux_kernel_report_offline. |
| 648 | |
| 649 | ranlib: new program |
| 650 | |
| 651 | Version 0.113: |
| 652 | |
| 653 | elflint: relax a bit. Allow version definitions for defined symbols against |
| 654 | DSO versions also for symbols in nobits sections. Allow .rodata section |
| 655 | to have STRINGS and MERGE flag set. |
| 656 | |
| 657 | strip: add some more compatibility with binutils. |
| 658 | |
| 659 | Version 0.112: |
| 660 | |
| 661 | elfcmp: some more relaxation. |
| 662 | |
| 663 | elflint: many more tests, especially regarding to symbol versioning. |
| 664 | |
| 665 | libelf: Add elfXX_offscn and gelf_offscn. |
| 666 | |
| 667 | libasm: asm_begin interface changes. |
| 668 | |
| 669 | libebl: Add three new interfaces to directly access machine, class, and |
| 670 | data encoding information. |
| 671 | |
| 672 | objdump: New program. Just the beginning. |
| 673 | |
| 674 | Version 0.111: |
| 675 | |
| 676 | libdw: now contains all of libdwfl. The latter is not installed anymore. |
| 677 | |
| 678 | elfcmp: little usability tweak, name and index of differing section is printed. |
| 679 | |
| 680 | Version 0.110: |
| 681 | |
| 682 | libelf: fix a number of problems with elf_update |
| 683 | |
| 684 | elfcmp: fix a few bugs. Compare gaps. |
| 685 | |
| 686 | Fix a few PLT problems and mudflap build issues. |
| 687 | |
| 688 | libebl: Don't expose Ebl structure definition in libebl.h. It's now private. |
| 689 | |
| 690 | Version 0.109: |
| 691 | |
| 692 | libebl: Check for matching modules. |
| 693 | |
| 694 | elflint: Check that copy relocations only happen for OBJECT or NOTYPE symbols. |
| 695 | |
| 696 | elfcmp: New program. |
| 697 | |
| 698 | libdwfl: New library. |
| 699 | |
| 700 | Version 0.108: |
| 701 | |
| 702 | strip: fix bug introduced in last change |
| 703 | |
| 704 | libdw: records returned by dwarf_getsrclines are now sorted by address |
| 705 | |
| 706 | Version 0.107: |
| 707 | |
| 708 | readelf: improve DWARF output format |
| 709 | |
| 710 | strip: support Linux kernel modules |
| 711 | |
| 712 | Version 0.106: |
| 713 | |
| 714 | libdw: Updated dwarf.h from DWARF3 spec |
| 715 | libdw: add new funtions dwarf_func_entrypc, dwarf_func_file, dwarf_func_line, |
| 716 | dwarf_func_col, dwarf_getsrc_file |
| 717 | |
| 718 | Version 0.105: |
| 719 | |
| 720 | addr2line: New program |
| 721 | |
| 722 | libdw: add new functions: dwarf_addrdie, dwarf_macro_*, dwarf_getfuncs, |
| 723 | dwarf_func_*. |
| 724 | |
| 725 | findtextrel: use dwarf_addrdie |
| 726 | |
| 727 | Version 0.104: |
| 728 | |
| 729 | findtextrel: New program. |
| 730 | |
| 731 | Version 0.103: |
| 732 | |
| 733 | libdw: Fix using libdw.h with gcc < 4 and C++ code. Compiler bug. |
| 734 | |
| 735 | Version 0.102: |
| 736 | |
| 737 | More Makefile and spec file cleanups. |
| 738 | |
| 739 | Version 0.101: |
| 740 | |
| 741 | Remove most gettext autoconf handling. |
| 742 | |
| 743 | Add more warnings |
| 744 | |
| 745 | Fix resulting problems. One actual bug found and fixed this way |
| 746 | |
| 747 | Version 0.100: |
| 748 | |
| 749 | libebl: Fix x86-64 relocations. |
| 750 | |
| 751 | Add -Wunused -Wextra warnings. |
| 752 | |
| 753 | Some cleanups resulting from those additional warnings. |
| 754 | |
| 755 | Lots of Makefile cleanup. |
| 756 | |
| 757 | Version 0.99: |
| 758 | |
| 759 | libelf: add gelf_checksum prototype to <libelf.h> |
| 760 | |
| 761 | libelf: fix elf*_checksum handling of NOBITS sections |
| 762 | |
| 763 | Finish mudflap support. |
| 764 | |
| 765 | Fix three bugs found by mudflap. |
| 766 | |
| 767 | ld: add as_needed support |
| 768 | |
| 769 | Version 0.98: |
| 770 | |
| 771 | readelf: in section to segment mapping, indicate read-only sections. |
| 772 | |
| 773 | elflint: more relaxation for GNU ld |
| 774 | |
| 775 | Version 0.97: |
| 776 | |
| 777 | Fix compiling with gcc 4.0. |
| 778 | Some tests called elflint without appropriate LD_LIBRARY_PATH. |
| 779 | |
| 780 | Version 0.96: |
| 781 | |
| 782 | Fix support for platforms with lib64. |
| 783 | |
| 784 | Version 0.95: |
| 785 | |
| 786 | libebl: add ppc and ppc64 support |
| 787 | |
| 788 | readelf: fix minimal memory leak. |
| 789 | |
| 790 | Add support to compile with mudflap. |
| 791 | |
| 792 | Modernize configure.ac. Move scripts in config subdir. |
| 793 | |
| 794 | Modernize *-po directory infrastructure. |
| 795 | |
| 796 | libelf: Add gelf_getlib and gelf_update_lib |
| 797 | |
| 798 | readelf: print liblist sections |
| 799 | |
| 800 | Version 0.94: |
| 801 | |
| 802 | Fix some minimal build problems. |
| 803 | |
| 804 | Version 0.93: |
| 805 | |
| 806 | ibdw: tons of new functionality and bug fixes. Several interface changes. |
| 807 | |
| 808 | readelf: use libdw now. |
| 809 | |
| 810 | libdwarf: removed completely. |
| 811 | |
| 812 | Version 0.92: |
| 813 | |
| 814 | configuration changes. |
| 815 | |
| 816 | Version 0.91: |
| 817 | |
| 818 | libdw: fix memory handling. Implement source line handling. |
| 819 | nm: use libdw instead of libdwarf. |
| 820 | libelf: change to GPL from OSL1 for now. |
| 821 | |
| 822 | Version 0.90: |
| 823 | |
| 824 | libebl: Recognize a few more section types and dynamic tags and return |
| 825 | approriate strings. |
| 826 | |
| 827 | Version 0.89: |
| 828 | |
| 829 | strip: fix overwriting of symbol table in input file. |
| 830 | |
| 831 | Version 0.88: |
| 832 | |
| 833 | libebl: Add some ia64 bits. |
| 834 | |
| 835 | Version 0.87: |
| 836 | |
| 837 | Bug fixes for big endian and some 64-bit machines. |
| 838 | |
| 839 | Version 0.86: |
| 840 | |
| 841 | strip: fix handling of Alpha and s390x which use incorrect hash bucket sizes. |
| 842 | |
| 843 | ld: tons of changes, moving towards usability. |
| 844 | |
| 845 | Version 0.85: |
| 846 | |
| 847 | strip: update section group symbol index if the associated symbol table changed |
| 848 | |
| 849 | libelf: fix two problems with generating output not via mmap |
| 850 | |
| 851 | elflint: add probably 10-15 more tests |
| 852 | libebl: add support for some of the new tests |
| 853 | |
| 854 | ld: gazillion changes |
| 855 | |
| 856 | Version 0.84: |
| 857 | |
| 858 | elflint: deal with .rel.dyn section. Fix a problem with rela platforms. |
| 859 | Handle PT_GNU_STACK. Change to write messages to stdout. |
| 860 | |
| 861 | readelf: fix a problem with version information in the symbol table output. |
| 862 | |
| 863 | strip: update all version symbol table entries |
| 864 | |
| 865 | Version 0.83: |
| 866 | |
| 867 | size: fix a warning |
| 868 | |
| 869 | strip: last changed caused problems when the symbol table is before the |
| 870 | relocation section. Fixed. This fix also improved the asymptotic |
| 871 | behavior if many symbol table sections are present. |
| 872 | |
| 873 | Version 0.82: |
| 874 | |
| 875 | Run strip tests with the correct libelf and libebl. |
| 876 | |
| 877 | libelf: fix bug in verneed byte order changing code. |
| 878 | |
| 879 | Version 0.81: |
| 880 | |
| 881 | strip: Remove unused symbol table entries. This might require updating |
| 882 | various other sections. |
| 883 | |
| 884 | Version 0.80: |
| 885 | |
| 886 | Fix some libelf problems with ET_REL files. |
| 887 | |
| 888 | Version 0.79: |
| 889 | |
| 890 | More warning changes, mainly by jbj. |
| 891 | |
| 892 | libdw: yet more new code. dwarf_child and dwarf_sibling should now actually |
| 893 | work. |
| 894 | |
| 895 | Version 0.78: |
| 896 | |
| 897 | libdw: 10+ new functions. get-pubnames2 works now fully. Almost all the |
| 898 | code needed for nm is in place. |
| 899 | |
| 900 | Version 0.77: |
| 901 | |
| 902 | cleanups to compile cleanly with gcc 3.3 and -Werror. |
| 903 | |
| 904 | libdw: some new code. |
| 905 | |
| 906 | Version 0.76: |
| 907 | |
| 908 | libebl: Fix last patch to recognize relocation sections. We must not |
| 909 | use the name. |
| 910 | |
| 911 | Version 0.75: |
| 912 | |
| 913 | libebl: .debug_ranges is a DWARF 3 debug section |
| 914 | libebl: recognize relocation sections for debug section |
| 915 | Patches by Jakub Jelinek. |
| 916 | |
| 917 | Version 0.74: |
| 918 | |
| 919 | Cleanups and more SPARC support by Tom Callaway <tcallaway@redhat.com>. |
| 920 | |
| 921 | Version 0.73: |
| 922 | |
| 923 | 64-bit cleanups for the programs. |
| 924 | |
| 925 | Version 0.72: |
| 926 | |
| 927 | libelf: and yet more fun with endian tranformation at output time. |
| 928 | |
| 929 | Version 0.71: |
| 930 | |
| 931 | libelf: more fun with endian tranformation at output time. Add test for it. |
| 932 | |
| 933 | Version 0.70: |
| 934 | |
| 935 | libelf: Two little bugs left from previous patch to handle section output |
| 936 | order. |
| 937 | |
| 938 | libelf: add unlikely in some more places. |
| 939 | |
| 940 | Version 0.69: |
| 941 | |
| 942 | libelf: fix output routines to handle case where section indeces and |
| 943 | ordre in the output file don't match correctly. Patch by Jakub. |
| 944 | |
| 945 | elflint: fix test of note section content for 64-bit platforms and files |
| 946 | with different byte order. |
| 947 | |
| 948 | Version 0.68: |
| 949 | |
| 950 | libebl: Fix SH_ENTSIZE_HASH definition (patch by Jakub) |
| 951 | |
| 952 | Version 0.67: |
| 953 | |
| 954 | libelf: correct mistake in error string handling. |
| 955 | |
| 956 | libelf: Implement ELF_F_PERMISSIVE. |
| 957 | strip: Implement --permissive option. |
| 958 | |
| 959 | Version 0.66: |
| 960 | |
| 961 | strip: Implement -g option. |
| 962 | |
| 963 | libelf: Handle broken hash table entry sizes. |
| 964 | |
| 965 | libebl: New function ebl_debugscn_p. Use it where appropriate. |
| 966 | |
| 967 | Version 0.65: |
| 968 | |
| 969 | libelf: Use correct file size for NOBITS section with ELF_F_LAYOUT set |
| 970 | |
| 971 | Version 0.64: |
| 972 | |
| 973 | libelf: Make error handling more robust. |
| 974 | libelf: Use TLS in error handler if configured with --enable-tls |
| 975 | |
| 976 | tests: input files are now distributed, not uuencoded in the shell scripts |
| 977 | |
| 978 | libdw: implement error handling, dwarf_get_pubnames |
| 979 | |
| 980 | Version 0.63: |
| 981 | |
| 982 | Build (incomplete) libdw. |
| 983 | |
| 984 | Version 0.62: |
| 985 | |
| 986 | Get rid of libtool. |
| 987 | |
| 988 | Version 0.61: |
| 989 | |
| 990 | Fix URL of OSL. |
| 991 | |
| 992 | Version 0.60: |
| 993 | |
| 994 | libebl: Handle .gnu.warning.* sections correctly. |
| 995 | |
| 996 | size: Implement -t option. |
| 997 | |
| 998 | libebl: Add IA-64 support. |
| 999 | libebl: Update SH relocations. |
| 1000 | libebl: Add Alpha support. |
| 1001 | libebl: Add Arm support. |
| 1002 | libebl: Add support for all currently known architecture to the loader. |
| 1003 | |
| 1004 | Version 0.59: |
| 1005 | |
| 1006 | nm: Implement -S option. Correct portable output format. Implement -s option. |
| 1007 | |
| 1008 | libelf: Take offset of archive into account in elf_rand. |
| 1009 | |
| 1010 | Version 0.58: |
| 1011 | |
| 1012 | strip: fix handling of ET_REL files. |
| 1013 | Add tests for strip. |
| 1014 | |
| 1015 | Version 0.57: |
| 1016 | |
| 1017 | strip: respect layout of input file |
| 1018 | |
| 1019 | Version 0.56: |
| 1020 | |
| 1021 | strip: handle files with large number of sections. |
| 1022 | |
| 1023 | Version 0.55: |
| 1024 | |
| 1025 | libelf: quite a few bug fixes by Alex Larsson. |
| 1026 | |
| 1027 | strip: implement -f option to place stripped sections into a separate |
| 1028 | file. By Alex Larsson. |
| 1029 | |
| 1030 | Version 0.54: |
| 1031 | |
| 1032 | strip: don't let STT_SECTION symbols keeps sections from being removed |
| 1033 | |
| 1034 | elflint: local symbols are allowed in .dynsym |
| 1035 | elflint: special case .rel.dyn a bit |
| 1036 | |
| 1037 | Version 0.53: |
| 1038 | |
| 1039 | elflint: check types and flags of special sections defined in gABI |
| 1040 | |
| 1041 | libebl: add x86-64 support |
| 1042 | |
| 1043 | Version 0.52: |
| 1044 | |
| 1045 | Start improvement of debug info handling in nm. |
| 1046 | |
| 1047 | libasm: implement asm_adduleb128 and asm_addsleb128 and a test for them |
| 1048 | |
| 1049 | Version 0.51: |
| 1050 | |
| 1051 | Fix build on 64-bit platforms. |
| 1052 | |
| 1053 | Version 0.50: |
| 1054 | |
| 1055 | nm: print file/line number also for local symbols |
| 1056 | |
| 1057 | use versions scripts not libtool's useless -export-symbols option |
| 1058 | |
| 1059 | Version 0.49: |
| 1060 | |
| 1061 | Update to autoconf 2.54 and automake 1.7. |
| 1062 | |
| 1063 | elflint: check note sections |
| 1064 | |
| 1065 | libdwarf: a number of bug fixes |
| 1066 | |
| 1067 | readelf: print .debug_info section content |
| 1068 | |
| 1069 | dwarf.h: Update from draft 7 |
| 1070 | |
| 1071 | Version 0.48: |
| 1072 | |
| 1073 | libcpu: beginning |
| 1074 | |
| 1075 | libelf: new function to read parts of the ELF file |
| 1076 | |
| 1077 | libebl: support for note section handling |
| 1078 | |
| 1079 | readelf: dump note sections |
| 1080 | |
| 1081 | Version 0.47: |
| 1082 | |
| 1083 | libelf: fix little new section-handling related bugs in elf_getshstrndx |
| 1084 | and elf_nextscn |
| 1085 | |
| 1086 | elflint: tests for mandatory content of dynamic section |
| 1087 | |
| 1088 | libasm: better handling of absolute symbols |
| 1089 | |
| 1090 | Version 0.46: |
| 1091 | |
| 1092 | libasm: rewrite to store Elf_Scn* instead of indices |
| 1093 | |
| 1094 | nm: finish many-section support |
| 1095 | |
| 1096 | nm: use debug in to print file/line info in sysv format |
| 1097 | |
| 1098 | libdwarf: fix a few bugs in DIE handling |
| 1099 | |
| 1100 | Version 0.45: |
| 1101 | |
| 1102 | libelf: major rewrite to keep Elf_Scn references valid until elf_end |
| 1103 | |
| 1104 | Version 0.44: |
| 1105 | |
| 1106 | libasm: Add support for bss, ABS, and COM sections. |
| 1107 | |
| 1108 | libebl: ebl_section_name takes now two index arguments to distinguish |
| 1109 | between special sections and extended sections |
| 1110 | |
| 1111 | Version 0.43: |
| 1112 | |
| 1113 | General: fix a few problem gcc 3.1 had with the code. |
| 1114 | |
| 1115 | libelf: implement {gelf,elf32,elf64}_checksum |
| 1116 | |
| 1117 | libelf: optimze DSO: fewer relocations, fewer PLTs |
| 1118 | |
| 1119 | add msg_tst test |
| 1120 | |
| 1121 | ld: use correct section header string table index; write correct index |
| 1122 | |
| 1123 | add dependencies for *.sym files |
| 1124 | |
| 1125 | Version 0.42: |
| 1126 | |
| 1127 | libelf: add elf_getshnum and elf_getshstrndx |
| 1128 | |
| 1129 | libebl: update section type name function |
| 1130 | |
| 1131 | elflint: tons of fixes wrt large number of sections. New tests in this area. |
| 1132 | Same amount of other bug fixes. |
| 1133 | |
| 1134 | size, strip, nm: better support for large number of sections. Including |
| 1135 | using correct section header string table |
| 1136 | |
| 1137 | libasm: correctly create data structures for large number of sections |
| 1138 | |
| 1139 | new tests asm-tst4 and asm-tst5 to check large number of sections |
| 1140 | |
| 1141 | libasm: implement section group generation |
| 1142 | |
| 1143 | elflint: more tests on section groups. Improve performance on existing |
| 1144 | section group tests |
| 1145 | |
| 1146 | Version 0.41: |
| 1147 | |
| 1148 | ld: add undefined symbols to dynamic symbol table if --export-dynamic is |
| 1149 | not given |
| 1150 | |
| 1151 | ld: fix value of e_entry |
| 1152 | |
| 1153 | Version 0.40: |
| 1154 | |
| 1155 | elflint: print section names in error messages |
| 1156 | |
| 1157 | elflint: mustn't warn about multiple DT_NULL |
| 1158 | |
| 1159 | ld: don't emit all symbols if --export-dynamic is not given |
| 1160 | |
| 1161 | ld: correct compute symbol address in output file (section index was off by 1) |
| 1162 | |
| 1163 | ld: generate correct version info in dynsym without --export-dynamic and |
| 1164 | in symtab |
| 1165 | |
| 1166 | Version 0.39: |
| 1167 | |
| 1168 | Fix check of various e_*size entries in elflint. |
| 1169 | |
| 1170 | Handle text output in asm_newsym. |
| 1171 | |
| 1172 | Finish checks in asm-tst3. |
| 1173 | |
| 1174 | Version 0.38: |
| 1175 | |
| 1176 | Update to autoconf 2.53, automake 1.6, gettext 0.11+. |
| 1177 | |
| 1178 | Introduce *.sym files to restrict export from DSOs. |
| 1179 | |
| 1180 | Use attribute_hidden and internal_function to optimize DSO code. |
| 1181 | |
| 1182 | Add TLS definitions in elf.h and handle them in readelf. |
| 1183 | |
| 1184 | Fix bug in verdef section generation in ld. |
| 1185 | |
| 1186 | Add initial libasm code. |
| 1187 | |
| 1188 | Version 0.37: |
| 1189 | |
| 1190 | Implement better hash size optimization heuristic in ld. It uses a formula |
| 1191 | taking number of tests into account. |
| 1192 | |
| 1193 | Lots of small bug fixes. |
| 1194 | |
| 1195 | Improve readelf output format. Respect various sh_link/sh_info values. |
| 1196 | Correctly print versioning information for symbol tables. |
| 1197 | |
| 1198 | Version 0.36: |
| 1199 | |
| 1200 | Implement preprocessing of linker script. Recognize -z combreloc. |
| 1201 | |
| 1202 | Version 0.35: |
| 1203 | |
| 1204 | Implement -z ignore|record for ld. |
| 1205 | |
| 1206 | Implement creating of .gnu.version_r and .gnu.version sections. The |
| 1207 | .gnu.version does not yet contain correct info for defined and versioned |
| 1208 | symbols (means .gnu.version_d is not yet implemented). |
| 1209 | |
| 1210 | Implement gelf_update_* functions to create versioning data. |
| 1211 | |
| 1212 | Version 0.34: |
| 1213 | |
| 1214 | Add DT_RUNPATH/DT_RPATH entries to dynamic section. Create .plt and |
| 1215 | .rel.plt sections (completely). Add support for all four PLT related |
| 1216 | dynamic section entries. Add callback function for PLT creation. |
| 1217 | |
| 1218 | More tests in elflint. Add support for very strict checking which for |
| 1219 | now flags level 2 (deprecated features) usage. |
| 1220 | |
| 1221 | Version 0.33: |
| 1222 | |
| 1223 | Create dynamic symbol table, dynamic string table, and hash table to ld. |
| 1224 | |
| 1225 | Add hash table histogram support to readelf. |
| 1226 | |
| 1227 | Version 0.32: |
| 1228 | |
| 1229 | more work on elflint |
| 1230 | |
| 1231 | ld now creates the dynamic section and references it. Start adding entries |
| 1232 | to dynamic section. |
| 1233 | |
| 1234 | Version 0.31: |
| 1235 | |
| 1236 | Start implementing elflint. |
| 1237 | |
| 1238 | Version 0.30: |
| 1239 | |
| 1240 | Fix handling of NOBITS sections in elf_getdata. |
| 1241 | |
| 1242 | Start implementing generation of executables and DSOs in ld. |
| 1243 | Generation of program header mostly done. Address computation done. |
| 1244 | Extension of linker script syntax. |
| 1245 | |
| 1246 | Various cleanups. |
| 1247 | |
| 1248 | Implement section group handling in readelf. |
| 1249 | |
| 1250 | Version 0.29: |
| 1251 | |
| 1252 | Implement section groups. This involved a lot of code moving. The |
| 1253 | new code is entirely untested since gas/gcc are currently not able to |
| 1254 | create section groups. ld works fine on files without section groups. |
| 1255 | |
| 1256 | Version 0.28: |
| 1257 | |
| 1258 | Fix problem with adding more section in elf_newscn. The section pointers |
| 1259 | for the data buffers wasn't adjusted. |
| 1260 | |
| 1261 | Fix elf_getdata with nonzero second parameter. Correctly handle creation |
| 1262 | of internal data buffer for machines without unaligned access. |
| 1263 | |
| 1264 | Version 0.27: |
| 1265 | |
| 1266 | Start adding support to selectively add sections. Includes support for |
| 1267 | section groups. |
| 1268 | Add --gc-sections/--no-gc-sections options. |
| 1269 | Add general section merging support. |
| 1270 | |
| 1271 | Fix a bug in section group support in strip. |
| 1272 | |
| 1273 | Fix some potential problems with hash value in dynamic hash implementation. |
| 1274 | |
| 1275 | Version 0.26: |
| 1276 | |
| 1277 | section merging works in ld. |
| 1278 | |
| 1279 | Version 0.25: |
| 1280 | |
| 1281 | Actually create data structures from version map file and use it to hide |
| 1282 | symbols in ld. |
| 1283 | |
| 1284 | Implement -s -s for ld. |
| 1285 | |
| 1286 | Version 0.24: |
| 1287 | |
| 1288 | Improve relocation table output in readelf. Avoid some crashes. |
| 1289 | Finish many section handling in readelf. |
| 1290 | |
| 1291 | Finish: finish implementation of ld -r. At least some simple tests pass. |
| 1292 | |
| 1293 | Version 0.23: |
| 1294 | |
| 1295 | Fix a number of errors in ELF_C_WRITE handling. |
| 1296 | |
| 1297 | Almost finished implementation of ld -r. The data sections are all copied. |
| 1298 | Handling of symbol tables is missing. |
| 1299 | |
| 1300 | Version 0.22: |
| 1301 | |
| 1302 | Handle DSO and archive input files correctly if -r option is given. |
| 1303 | |
| 1304 | Gracefully deal with no phdr in new file in libelf. |
| 1305 | Fix various small error handling problems. |
| 1306 | Don't mmap file for output unless the command says so. |
| 1307 | |
| 1308 | Add code to create ELF section header table to ld finalize routines. |
| 1309 | |
| 1310 | Version 0.21: |
| 1311 | |
| 1312 | Fix some problems with recursive handling of archives in libelf. |
| 1313 | |
| 1314 | Improve messages printed by nm. |
| 1315 | |
| 1316 | Add symbol binding name handling to libebl. Fix section name handling in |
| 1317 | libebl. |
| 1318 | |
| 1319 | readelf and nm use more libebl functions. |
| 1320 | |
| 1321 | Handle XINDEX correctly in nm and string. |
| 1322 | |
| 1323 | Add first machine ld backend library (i386). |
| 1324 | Use it. Recognize -r and --shared. Avoid using -lxxx parameters for -r. |
| 1325 | Create ELF header in output file. Change mode of output file according to |
| 1326 | output file type. Reorganize callback initialization in ld. |
| 1327 | |
| 1328 | Version 0.20: |
| 1329 | |
| 1330 | Fix some memory leaks in libelf. |
| 1331 | |
| 1332 | Version 0.19: |
| 1333 | |
| 1334 | Implement reading version script. Both inside linker scripts and via the |
| 1335 | --version-script command line parameter. Uses the same code. |
| 1336 | What remains to be done is to implement a data structure which allows |
| 1337 | efficient matching against the version names to decide which pattern |
| 1338 | matches. |
| 1339 | |
| 1340 | Beginning of output generation and output writing functions. |
| 1341 | |
| 1342 | Version 0.18: |
| 1343 | |
| 1344 | Finish implementation for DSO input file handling. Implement rpath, runpath, |
| 1345 | and LD_LIBRARY_PATH handling. |
| 1346 | |
| 1347 | Version 0.17: |
| 1348 | |
| 1349 | make handling of e_shnum overflow in libelf standard conforming |
| 1350 | |
| 1351 | ld now actually can handle DSOs in linker scripts. Handling of DT_RUNPATH, |
| 1352 | DT_RPATH, -rpath, -rpath-link still remains to be implemented. |
| 1353 | |
| 1354 | fix handling of -L parameters. Make actual use of the default_paths element. |
| 1355 | |
| 1356 | make re-definition of symbols in and from DSO compatible with existing linker |
| 1357 | |
| 1358 | Version 0.16: |
| 1359 | |
| 1360 | more work on assigning input sections to output sections. |
| 1361 | |
| 1362 | Add gelf_xlatetof and gelf_xlatetom which were accidently left out. |
| 1363 | |
| 1364 | Fix memory handling of section headers. |
| 1365 | |
| 1366 | Version 0.15: |
| 1367 | |
| 1368 | Add many-section support to ld. Add various new command line parameters. |
| 1369 | Allow pagesize to be specified in linker script or on the command line. |
| 1370 | Collect input sections in list for the output section according to the rules |
| 1371 | specified in the linker script. |
| 1372 | |
| 1373 | Version 0.14: |
| 1374 | |
| 1375 | Fix some problems in the internal list handling which had the result |
| 1376 | that we didn't look for some of the unresolved symbols. |
| 1377 | |
| 1378 | Free some memory if we know we don't need it anymore. |
| 1379 | |
| 1380 | Optimize the list of unresolved symbols. Throw out symbols which are |
| 1381 | meanwhile resolved. |
| 1382 | |
| 1383 | Version 0.13: |
| 1384 | |
| 1385 | Got file reading correct now. The files are all read while parsing |
| 1386 | the parameters. No creating of data structures to describe the linker |
| 1387 | command line. The symbol table is built up while reading the files. |
| 1388 | DSOs are handled now. -( -) handling is optimized. |
| 1389 | |
| 1390 | Version 0.12: |
| 1391 | |
| 1392 | Linker read linker scripts everywhere. Handles --whole-archive. Recognizes |
| 1393 | --dynamic and --static. Collects defined and undefined symbols. Recognizes |
| 1394 | conflicts. |
| 1395 | |
| 1396 | libebl now defines functions to call the callbacks. Add generic name handling |
| 1397 | in these new functions. Remove the code from readelf and call the new |
| 1398 | functions. |
| 1399 | |
| 1400 | Version 0.11: |
| 1401 | |
| 1402 | Start of linker. Basic argument parsing, finding of input files, |
| 1403 | linker script reading. |
| 1404 | |
| 1405 | Version 0.10: |
| 1406 | |
| 1407 | Implement dwarf_get_fde_n(), dwarf_get_abbrev(), dwarf_get_abbrev_tag(), |
| 1408 | dwarf_get_abbrev_code(), dwarf_get_abbrev_children_flag(), |
| 1409 | dwarf_get_abbrev_entry(), dwarf_get_fde_at_pc(), and tests for it. |
| 1410 | |
| 1411 | Version 0.9: |
| 1412 | |
| 1413 | Implement dwarf_get_fde_list_eh(), dwarf_get_cie_of_fde(), |
| 1414 | dwarf_get_fde_range(), dwarf_get_cie_info(), dwarf_get_fde_instr_bytes(), |
| 1415 | and tests for them. |
| 1416 | |
| 1417 | Version 0.8: |
| 1418 | |
| 1419 | Make handling of binaries in other byte order work and add tests for it. |
| 1420 | |
| 1421 | Version 0.7: |
| 1422 | |
| 1423 | Implement dwarf_get_aranges(), dwarf_get_arange(), dwarf_get_cu_die_offset(), |
| 1424 | dwarf_get_arange_info(), and tests for them. |
| 1425 | |
| 1426 | Version 0.6: |
| 1427 | |
| 1428 | Implement dwarf_get_global(), dwarf_globname(), dwarf_global_die_offset(), |
| 1429 | dwarf_global_cu_offset(), dwarf_global_name_offsets(), and tests for them |
| 1430 | |
| 1431 | Version 0.5: |
| 1432 | |
| 1433 | Implemented dwarf_srclines(), dwarf_srcfiles(), dwarf_linebeginstatement(), |
| 1434 | dwarf_lineendsequence(), dwarf_lineno(), dwarf_lineaddr(), dwarf_lineoff(), |
| 1435 | dwarf_linesrc(), dwarf_lineblock(), dwarf_lineprologueend(), |
| 1436 | dwarf_lineepiloguebegin(), and tests for them. |
| 1437 | |
| 1438 | Version 0.4: |
| 1439 | |
| 1440 | Implemented dwarf_loclist(). |
| 1441 | |
| 1442 | Version 0.3: |
| 1443 | |
| 1444 | Implemented dwarf_dieoffset(), dwarf_die_CU_offset(), dwarf_diename() and |
| 1445 | tests. |
| 1446 | |
| 1447 | Implemented dwarf_attrlist(), dwarf_hasattr(), dwarf_attr(), dwarf_lowpc(), |
| 1448 | dwarf_highpc(), dwarf_bytesize(), dwarf_bitsize(), dwarf_bitoffset(), |
| 1449 | dwarf_srclang(), dwarf_arrayorder(), dwarf_hasform(), dwarf_whatform(), |
| 1450 | dwarf_whatattr(), dwarf_formref(), dwarf_global_formref(), dwarf_formaddr(), |
| 1451 | dwarf_formflag(), dwarf_formudata(), dwarf_formsdata(), dwarf_formblock, |
| 1452 | dwarf_formstring() and tests for them. |
| 1453 | |
| 1454 | Version 0.2: |
| 1455 | |
| 1456 | Implemented dwarf_offdie()), dwarf_tag(), dwarf_dieoffset(), |
| 1457 | dwarf_die_CU_offset(), dwarf_diename() and tests for them. |
| 1458 | |
| 1459 | Version 0.1: |
| 1460 | |
| 1461 | First libdwarf functions work. |
| 1462 | |
| 1463 | Version 0.0: |
| 1464 | |
| 1465 | libelf and parts of libebl are done. |