12020-12-20 Dmitry V. Levin <ldv@altlinux.org> 2 3 * .gitignore: New file. 4 52020-12-16 Érico Nogueira <ericonr@disroot.org> 6 7 * Makefile.am (i386_gendis_LDADD): Add obstack_LIBS. 8 92020-12-16 Dmitry V. Levin <ldv@altlinux.org> 10 11 * i386_lex.l (invalid_char): Replace gettext(...) with _(...). 12 * i386_parse.y (yyerror): Likewise. 13 142020-12-12 Dmitry V. Levin <ldv@altlinux.org> 15 16 * bpf_disasm.c (bswap_bpf_insn): Fix spelling typo in comment. 17 * i386_disasm.c (i386_disasm): Likewise. 18 192020-05-09 Mark Wielaard <mark@klomp.org> 20 21 * i386_parse.y (new_bitfield): Call free newp on error. 22 232020-04-16 Mark Wielaard <mark@klomp.org> 24 25 * i386_disasm.c (i386_disasm): Replace assert with goto invalid_op 26 for bad prefix. 27 282019-12-11 Omar Sandoval <osandov@fb.com> 29 30 * Makefile.am (i386_lex_CFLAGS): Add -Wno-implicit-fallthrough. 31 322019-10-17 Mark Wielaard <mark@klomp.org> 33 34 * i386_data.h (FCT_sel): Check for param_start + 2 >= end instead 35 of just >. 36 * i386_disasm.c (i386_disasm): Check param_start < end. Don't 37 assert, but assign INVALID to str. Make sure we get past any 38 unrecognized opcode. 39 402019-09-07 Mark Wielaard <mark@klomp.org> 41 42 * riscv_disasm.c (riscv_disasm): Use UINT64_C to make calculation 43 unsigned. 44 452019-07-05 Omar Sandoval <osandov@fb.com> 46 47 * Makefile.am: Combine libcpu_{i386,x86_64,bpf}.a into libcpu.a. 48 Make libcpu.a non-PIC by default. 49 Add libcpu_pic.a. 50 512018-11-04 Mark Wielaard <mark@klomp.org> 52 53 * bpf_disasm.c (bpf_disasm): Recognize BPF_JLT, BPF_JLE, BPF_JSLT 54 and BPF_JSLE. 55 562018-02-09 Joshua Watt <JPEWhacker@gmail.com> 57 58 * i386_disasm.c (i386_disasm): Use FALLTHOUGH macro instead of 59 comment. 60 612017-08-18 Ulf Hermann <ulf.hermann@qt.io> 62 63 * memory-access.h: Use attribute_packed. 64 652017-02-27 Ulf Hermann <ulf.hermann@qt.io> 66 67 * Makefile.am: Use fpic_CFLAGS. 68 692017-07-18 Mark Wielaard <mark@klomp.org> 70 71 * Makefile.am: Don't check HAVE_LINUX_BPF_H, just define libcpu_bpf. 72 * bpf_disasm.c: Include bpf.h instead of linux/bpf.h. Don't define 73 BPF_PSEUDO_MAP_FD. 74 752017-04-20 Ulf Hermann <ulf.hermann@qt.io> 76 77 * Makefile.am: Add EXEEXT to gendis. 78 792017-04-20 Ulf Hermann <ulf.hermann@qt.io> 80 81 * i386_parse.y: Eliminate comparison_fn_t. 82 832016-11-02 Mark Wielaard <mjw@redhat.com> 84 85 * i386_disasm.c (i386_disasm): Add fallthrough comment. 86 872016-10-11 Akihiko Odaki <akihiko.odaki.4i@stu.hosei.ac.jp> 88 89 * i386_lex.l: Remove system.h include, add libeu.h include. 90 * i386_parse.y: Remove sys/param.h include, add libeu.h include. 91 * i386_disasm.c: Remove sys/param.h. 92 932016-09-05 Mark Wielaard <mjw@redhat.com> 94 95 * bpf_disasm.c: Define BPF_PSEUDO_MAP_FD if undefined. 96 972016-08-10 Richard Henderson <rth@redhat.com> 98 99 * bpf_disasm.c (bpf_disasm): Rearrange the printing of instructions 100 to use exactly the operands required. 101 1022016-06-28 Richard Henderson <rth@redhat.com> 103 104 * Makefile.am (noinst_LIBRARIES): Add libcpu_bpf.a. 105 (libcpu_bpf_a_SOURCES, libcpu_bpf_a_CFLAGS): New. 106 * bpf_disasm.c: New file. 107 * i386_disasm.c (i386_disasm): Add ebl parameter. 108 1092015-10-05 Josh Stone <jistone@redhat.com> 110 111 * Makefile.am (%_defs): Add AM_V_GEN and AM_V_at silencers. 112 ($(srcdir)/%_dis.h): Ditto. 113 (%.mnemonics): Add AM_V_GEN silencer. 114 1152014-10-29 Jose E. Marchesi <jose.marchesi@oracle.com> 116 117 * Makefile.am (AM_CFLAGS): Use -fPIC instead of -fpic to avoid 118 relocation overflows in some platforms. 119 1202014-04-13 Mark Wielaard <mjw@redhat.com> 121 122 * Makefile.am (i386_gendis_LDADD): Remove libmudflap. 123 1242013-04-24 Mark Wielaard <mjw@redhat.com> 125 126 * Makefile.am: Use AM_CPPFLAGS instead of INCLUDES. 127 1282012-10-10 Roland McGrath <roland@hack.frob.com> 129 130 * Makefile.am (%_defs, $(srcdir)/%_dis.h): Redirect to temp file, 131 mv into place with separate command. 132 1332012-06-26 Roland McGrath <roland@hack.frob.com> 134 135 * Makefile.am [!MAINTAINER_MODE] ($(srcdir)/%_dis.h): New rule. 136 1372012-02-24 Mark Wielaard <mjw@redhat.com> 138 139 * Makefile.am (CLEANFILES): Move %_dis.h to... 140 (MAINTAINERCLEANFILES): here. 141 1422012-01-21 Ulrich Drepper <drepper@gmail.com> 143 144 * i386_disasm.c (ADD_NSTRING): Define. 145 (i386_disasm): Print color codes in the appropriate places. 146 1472011-10-16 Roland McGrath <roland@hack.frob.com> 148 149 * Makefile.am (libcpu_i386_a_SOURCES): Add i386_dis.h. 150 (libcpu_x86_64_a_SOURCES): Add x86_64_dis.h. 151 (i386_disasm.o, x86_64_disasm.o): Depend on those in $(srcdir). 152 (%_dis.h): Renamed target pattern to ... 153 ($(srcdir)/%_dis.h): ... this. 154 (noinst_HEADERS, noinst_PROGRAMS): Put under [MAINTAINER_MODE]. 155 1562010-08-16 Roland McGrath <roland@redhat.com> 157 158 * Makefile.am (%_defs): New pattern rule. 159 (%_dis.h, %.mnemonics): Define as pattern rules using %_defs input. 160 (CLEANFILES): Include all those files. 161 1622010-02-15 Roland McGrath <roland@redhat.com> 163 164 * Makefile.am: Use config/eu.am for common stuff. 165 1662009-04-14 Roland McGrath <roland@redhat.com> 167 168 * Makefile.am (AM_CFLAGS): Add -fdollars-in-identifiers; it is not the 169 default on every machine. 170 1712009-01-23 Roland McGrath <roland@redhat.com> 172 173 * Makefile.am (i386_parse_CFLAGS): Use quotes around command 174 substitution that can produce leading whitespace. 175 1762009-01-01 Ulrich Drepper <drepper@redhat.com> 177 178 * i386_parse.y (instrtable_out): Optimize match_data table by not 179 emitting 0xff masks for leading bytes. 180 * i386_disasm.c (i386_disasm): Adjust reader of match_data. 181 182 * i386_disasm.c (i386_disasm): Reset bufcnt when not matched. We 183 don't expect snprintf to fail. 184 1852008-12-31 Ulrich Drepper <drepper@redhat.com> 186 187 * defs/i386: Add dppd, dpps, insertps, movntdqa, mpsadbw, packusdw, 188 pblendvb, pblendw, pcmpeqq, pcmpestri, pcmpestrm, pcmpistri, pcmpistrm, 189 pcmpgtq, phminposuw, pinsrb, pinsrd, pmaxsb, pmaxsd, pmaxud, pmaxuw, 190 pminsb, pminsd, pminud, pminuw, pmovsxbw, pmovsxbd, pmovsxbq, pmovsxwd, 191 pmovsxwq, pmovsxdq, pmovzxbw, pmovzxbd, pmovzxbq, pmovzxwd, pmovzxwq, 192 pmovzxdq, pmuldq, pmulld, popcnt, ptest, roundss, roundps, roundpd, 193 and roundsd opcodes. 194 195 * i386_disasm.c (i386_disasm): Correct resizing of buffer. 196 197 * i386_parse.y (struct argstring): Add off element. 198 (off_op_str): New global variable. 199 (print_op_str): Print strings as concatenated strings. Keep track 200 of index and length. Update ->off element. 201 (print_op_str_idx): New function. 202 (instrtable_out): Mark op%d_fct as const. 203 Emit two tables for the strings: the string itself (op%d_str) and the 204 index table (op%d_str_idx). 205 * i386_disasm.c (i386_disasm): Adjust for new op%d_str definition. 206 207 * i386_disasm.c [X86_64] (i386_disasm): Handle rex prefix when 208 printing only prefix. 209 210 * i386_disasm.c (i386_disasm): Minor optimizations. 211 212 * i386_parse.y (instrtable_out): No need to emit index, the reader can 213 keep track. 214 * i386_disasm.c (i386_disasm): The index is not emitted anymore, no 215 need to skip it. 216 217 * i386_disasm.c (amd3dnow): Mark as const. 218 219 * defs/i386: Add blendvpd and blendvps opcodes. 220 2212008-12-30 Ulrich Drepper <drepper@redhat.com> 222 223 * defs/i386: Add blendpd and blendps opcodes. 224 2252008-12-19 Ulrich Drepper <drepper@redhat.com> 226 227 * defs/i386: Add entry for AMD 3DNOW. 228 * i386_disasm.c: Implement AMD 3DNOW disassembly. 229 2302008-12-17 Ulrich Drepper <drepper@redhat.com> 231 232 * i386_disasm.c (i386_disasm): If instruction matches prefix, 233 undoing the prefix match finishes the instruction. 234 2352008-01-21 Roland McGrath <roland@redhat.com> 236 237 * defs/i386: Fix typo in comment. 238 * i386_disasm.c (i386_disasm): Handle cltq, cqto. 239 240 * i386_parse.y: Add sanity check for NMNES macro value. 241 * Makefile.am (i386_parse.o): Fix target in dependency rule. 242 (i386_parse.h): New target with empty commands. 243 (i386_lex.o): Depend on it in place of i386_parse.c. 244 2452008-01-21 Ulrich Drepper <drepper@redhat.com> 246 247 * Makefile.am (EXTRA_DIST): Remove defs/x86_64. 248 2492008-01-14 Ulrich Drepper <drepper@redhat.com> 250 251 * defs/i386: Add fixes for opcodes with register number in opcode, 252 64-bit immediate forms, nop with rex.B. 253 * i386_data.h [X86_64] (FCT_imm64$w): New function. 254 (FCT_oreg): New function. 255 (FCT_oreg$w): New function. 256 * i386_disasm.c (i386_disasm): Reinitialize fmt always before 257 starting the loop to process the string. Handle 0x90 special for 258 x86-64. 259 * i386_parse.y (fillin_arg): Expand synonyms before concatening to 260 form the function name. 261 2622008-01-11 Ulrich Drepper <drepper@redhat.com> 263 264 * i386_disasm.c (struct output_buffer): Remove symcb and symcbarg. 265 (i386_disasm): Remove appropriate initializers. 266 Use symcb to lookup symbol strings. 267 268 * i386_disasm.c (struct output_buffer): Add labelbuf, labelbufsize, 269 symaddr_use, and symaddr fields. 270 (i386_disasm): Remove labelbuf and labelbufsize variables. 271 Add back %e format. Implement %a and %l formats. 272 273 * i386_data.h (general_mod$r_m): Set symaddr_use and symaddr for %rip 274 base addressing. 275 276 * i386_disasm.c (i386_disasm): Resize output buffer if necessary. 277 Optimize output_data initialization. Free buffers before return. 278 (struct output_data): Remove op1str field. Adjust code. 279 (i386_disasm): Store final NUL btye at end of functions. 280 2812008-01-10 Ulrich Drepper <drepper@redhat.com> 282 283 * i386_data.h (FCT_crdb): New function. 284 (FCT_ccc): Use FCT_crdb. 285 (FCT_ddd): Likewise. 286 287 * defs/i386: Fix a few instructions with immediate arguments. 288 289 * i386_disasm.c: Rewrite interface to callback functions for operands 290 to take a single pointer to a structure. 291 * i386_data.h: Adjust all functions. 292 2932008-01-08 Ulrich Drepper <drepper@redhat.com> 294 295 * Makefile.am: Enable x86-64 again. 296 * defs/i386: Lots of changes for x86-64. 297 * i386_data.h: Add support for use in x86-64 disassembler. 298 * i386_disasm.c: Likewise. 299 * i386_parse.y: Likewise. 300 * defs/x86_64: Removed. 301 3022008-01-04 Ulrich Drepper <drepper@redhat.com> 303 304 * defs/i386: Cleanups, remove masks which are not needed. 305 Add remaining Intel opcodes. 306 * i386_data.h (FCT_imm8): Check for input buffer overrun. 307 * i386_disasm.c (i386_disasm): Likewise. 308 * i386_parse.y: Remove suffixes which are not needed anymore. 309 3102008-01-03 Ulrich Drepper <drepper@redhat.com> 311 312 * defs/i386: Add yet more SSE instructions. 313 3142008-01-02 Ulrich Drepper <drepper@redhat.com> 315 316 * i386_disasm.c (i386_disasm): Extend matcher to allow tables to 317 contain instructions with prefixes. 318 * defs/i386: Use for many SSE operations. 319 * i386_data.h (FCT_mmxreg2): Removed. 320 3212008-01-01 Ulrich Drepper <drepper@redhat.com> 322 323 * defs/i386: More 0f prefix support. 324 * i386_data.h (FCT_mmxreg): Implement. 325 (FCT_mmxreg2): Implement. 326 (FCT_mmreg): Remove. 327 * i386_disasm.c (i386_disasm): More special instructions. 328 Fix tttn suffix for cmov. 329 * i386_parse.y: Simplify test for mod/r_m mode. 330 3312007-12-31 Ulrich Drepper <drepper@redhat.com> 332 333 * defs/i386: Fix order or arguments for mov of control/debug registers. 334 * i386_data.h (FCT_ccc): Implement 335 (FCT_ddd): Implement 336 3372007-12-30 Ulrich Drepper <drepper@redhat.com> 338 339 * defs/i386: Fix 0f groups 6 and 7. 340 * i386_data.c (FCT_mod$16r_m): Implement. 341 * i386_disasm.c (i386_disasm): Third parameter can also have string. 342 3432007-12-29 Ulrich Drepper <drepper@redhat.com> 344 345 * defs/i386: Add lots of floating point ops. 346 * i386_data.h (FCT_fmod$fr_m): Removed. 347 (FCT_freg): Implement. 348 * i386_disasm.c (i386_disasm): Implement suffix_D. 349 * i386_parse.y: Emit suffix_D. 350 351 * defs/i386: Use rel instead of dispA. 352 Fix lcall, dec, div, idiv, imul, inc, jmp, ljmp, mul, neg, not, push, 353 test. 354 355 * i386_data.h (FCT_dispA): Removed. 356 (FCT_ds_xx): Add test for end of input buffer. 357 * i386_disasm.c (ABORT_ENTRY): Removed. 358 (i386_disasm): Fix handling of SIB. Pass correct address value to 359 operand callbacks. 360 361 * Makefile.am (*.mnemonics): Filter out INVALID entry. 362 * defs/i386: Define imms8 and use in appropriate places. 363 Add INVALID entries for special opcodes with special mnemonics. 364 Fix int3. Fix typo in shl. Correct xlat. 365 * i386_data.h (FCT_ds_xx): New function. 366 (FCT_ds_si): Use it. 367 (FCT_ds_bx): New function. 368 (FCT_imms8): New function. 369 * i386_disasm.c (MNE_INVALID): Define. 370 (i386_disasm): Handle invalid opcodes in mnemonics printing, not 371 separately. Fix address value passed to operand handlers. 372 * i386_parse.y (bx_reg): Define. 373 (instrtable_out): Handle INVALID entries differently, just use 374 MNE_INVALID value for .mnemonic. 375 3762007-12-28 Ulrich Drepper <drepper@redhat.com> 377 378 * defs/i386: Fix shift and mov immediate instructions. 379 * i386_data.h (FCT_imm16): Implement. 380 381 * defs/i386: Use absval instead of abs of lcall and ljmp. 382 Add parameters for cmps. Fix test and mov immediate. 383 * i386_data.h: Implement FCT_absval. 384 * i386_disasm.c: Handle data16 for suffix_w and FCT_imm. 385 386 * defs/i386: Move entries with 0x9b prefix together. 387 * i386_disasm.c (i386_disasm): Fix recognizing insufficient bytes in 388 input. Handle data16 with suffix_W. 389 390 * i386_data.h (FCT_*): Add end parameter to all functions. Check 391 before using more bytes. 392 (FCT_sel): Implement. 393 * i386_disasm.c (i386_disasm): Better handle end of input buffer. 394 Specal opcode 0x99. 395 396 * Makefile.am: Use m4 to preprocess defs/* files. 397 * defs/i386: Adjust appropriately. 398 * i386_data.c (FCT_ax): Implement. 399 (FCT_ax$w): Use FCT_ax. 400 * i386_disasm.c (ADD_STRING): Use _len instead of len. 401 (i386_disasm): If no instruction can be matched because of lack of 402 input and prefixes have been matched, print prefixes. 403 Recognize abort entries. 404 Handle special cases. 405 * i386_gendis.c: Recognize - input file name. 406 * i386_lex.c: Recognize INVALID token. 407 * i386_parse.y: Handle INVALID token input. 408 409 * defs/i386: Fix mov, pop. 410 * i386_data.h (FCT_sreg3): Implement. 411 4122007-12-27 Ulrich Drepper <drepper@redhat.com> 413 414 * defs/i386: Fix adc, add, cmp, or, sbb, sub, xchg, xor. 415 * i386_data.h (FCT_imms): New function. 416 (FCT_imm$s): Use FCT_imms for handling of signed values. 417 (FCT_imm8): Sign extend values. 418 * i386_disasm.c (i386_disasm): Implement suffix_w0. 419 * i386_parse.y: Emit suffix w0. 420 421 * i386_data.h (FCT_disp8): Add 0x prefix. 422 (FCT_ds_si): Implement. 423 * i386_disasm.c (i386_disasm): Increment addr for invalid prefixes. 424 Implement tttn suffix. 425 * i386_parse.y: Emit tttn suffix definition. 426 4272007-12-26 Ulrich Drepper <drepper@redhat.com> 428 429 * i386_data.h (struct instr_enc): Use suffix field. 430 (FCT_dx): Fill in body. 431 (FCT_es_di): Likewise. 432 (FCT_imm$s): Sign-extended byte values. 433 * i386_disasm.c: Protect ADD_CHAR and ADD_STRING macros. Adjust uses. 434 (i386_disasm): Handle suffix. 435 * i386_parse.y: Emit suffix information. 436 * defs/i386: Remove unnecessary suffixes. 437 438 * Makefile.am: Disable building x86-64 version for now. 439 440 * defs/i386: Fix and, bound, cmp, or, pop, sbb, sub, xor. 441 * i386_data.h: Pass pointer to prefix to functions. If not prefixes 442 are consumed this means invalid input. 443 * i386_disasm.c: Fix prefix printing. Adjust function calls for 444 parameter change. 445 * i386_parse.y: Recognize moda prefix. 446 4472007-12-21 Ulrich Drepper <drepper@redhat.com> 448 449 * i386_data.h: Fix SIB handling. 450 * i386_disasm.c: Likewise. 451 4522007-12-19 Ulrich Drepper <drepper@redhat.com> 453 454 * defs/i386: Fix up 'and' opcode. 455 4562007-10-31 Ulrich Drepper <drepper@redhat.com> 457 458 * Makefile.am: Add dependencies of the generated files on the source 459 files. 460 (i386_lex_CFLAGS): Add -Wno-sign-compare. 461 462 * defs/i386: A lot more data. 463 * defs/x86_64: Likewise. 464 * i386_data.h (struct instr_enc): Add off1_3, off2_3, and off3_3 465 fields. 466 (opfct_t): Add parameter for third operand. 467 (FCT_*): Likewise. 468 (data_prefix): New function. 469 (FCT_abs): Implement. 470 (FCT_ax): Renamed to FCT_ax$w amd implement. 471 (FCT_disp8): Implement. 472 (FCT_dispA): Implement. 473 (FCT_imm): Implement. 474 (FCT_imm$w): Implement. 475 (FCT_imm$s): Don't zero-pad numbers. 476 (FCT_imm8): Likewise. 477 (FCT_rel): Likewise. 478 (general_mod$r_m): New function. 479 (FCT_mod$r_m): Use it. 480 (FCT_mod$r_m$w): New function. 481 (FCT_mod$8r_m): New function. 482 (FCT_reg): Correctly handle 16-bit registers. 483 (FCT_reg$w): New function. 484 * i386_disasm.c (i386_disasm): Handle prefixes better. 485 Pass third parameter to operand functions. 486 * i386_parse.y (struct instruction): Add off3 field. 487 Handle third operand throughout. 488 4892007-02-05 Ulrich Drepper <drepper@redhat.com> 490 491 * i386_disasm.c: New file. 492 * i386_data.h: New file. 493 * i386_gendis.c: New file. 494 * i386_lex.l: New file. 495 * i386_parse.y: New file. 496 * memory-access.h: New file. 497 * x86_64_disasm.c: New file. 498 * defs/i386: New file. 499 * defs/i386.doc: New file. 500 * defs/x86_64: New file. 501 5022005-02-15 Ulrich Drepper <drepper@redhat.com> 503 504 * Makefile (AM_CFLAGS): Add -Wunused -Wextra -Wformat=2. 505 5062005-02-05 Ulrich Drepper <drepper@redhat.com> 507 508 * Makefile.am (AM_CFLAGS): Define, instead of adding things to DEFS. 509 5102003-08-11 Ulrich Drepper <drepper@redhat.com> 511 512 * Moved to CVS archive. 513