arch/x86/lib/insn.c

   1 /*
   2  * x86 instruction analysis
   3  *
   4  * This program is free software; you can redistribute it and/or modify
   5  * it under the terms of the GNU General Public License as published by
   6  * the Free Software Foundation; either version 2 of the License, or
   7  * (at your option) any later version.
   8  *
   9  * This program is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License
  15  * along with this program; if not, write to the Free Software
  16  * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  17  *
  18  * Copyright (C) IBM Corporation, 2002, 2004, 2009
  19  */
  20
  21 #ifdef __KERNEL__
  22 #include <linux/string.h>
  23 #else
  24 #include <string.h>
  25 #endif
  26 #include <asm/inat.h>
  27 #include <asm/insn.h>
  28
  29 /* Verify next sizeof(t) bytes can be on the same instruction */
  30 #define validate_next(t, insn, n)       \
  31         ((insn)->next_byte + sizeof(t) + n <= (insn)->end_kaddr)
  32
  33 #define __get_next(t, insn)     \
  34         ({ t r = *(t*)insn->next_byte; insn->next_byte += sizeof(t); r; })
  35
  36 #define __peek_nbyte_next(t, insn, n)   \
  37         ({ t r = *(t*)((insn)->next_byte + n); r; })
  38
  39 #define get_next(t, insn)       \
  40         ({ if (unlikely(!validate_next(t, insn, 0))) goto err_out; __get_next(t, insn); })
  41
  42 #define peek_nbyte_next(t, insn, n)     \
  43         ({ if (unlikely(!validate_next(t, insn, n))) goto err_out; __peek_nbyte_next(t, insn, n); })
  44
  45 #define peek_next(t, insn)      peek_nbyte_next(t, insn, 0)
  46
  47 /**
  48  * insn_init() - initialize struct insn
  49  * @insn:       &struct insn to be initialized
  50  * @kaddr:      address (in kernel memory) of instruction (or copy thereof)
  51  * @x86_64:     !0 for 64-bit kernel or 64-bit app
  52  */
  53 void insn_init(struct insn *insn, const void *kaddr, int buf_len, int x86_64)
  54 {
  55         /*
  56          * Instructions longer than MAX_INSN_SIZE (15 bytes) are invalid
  57          * even if the input buffer is long enough to hold them.
  58          */
  59         if (buf_len > MAX_INSN_SIZE)
  60                 buf_len = MAX_INSN_SIZE;
  61
  62         memset(insn, 0, sizeof(*insn));
  63         insn->kaddr = kaddr;
  64         insn->end_kaddr = kaddr + buf_len;
  65         insn->next_byte = kaddr;
  66         insn->x86_64 = x86_64 ? 1 : 0;
  67         insn->opnd_bytes = 4;
  68         if (x86_64)
  69                 insn->addr_bytes = 8;
  70         else
  71                 insn->addr_bytes = 4;
  72 }
  73
  74 /**
  75  * insn_get_prefixes - scan x86 instruction prefix bytes
  76  * @insn:       &struct insn containing instruction
  77  *
  78  * Populates the @insn->prefixes bitmap, and updates @insn->next_byte
  79  * to point to the (first) opcode.  No effect if @insn->prefixes.got
  80  * is already set.
  81  */
  82 void insn_get_prefixes(struct insn *insn)
  83 {
  84         struct insn_field *prefixes = &insn->prefixes;
  85         insn_attr_t attr;
  86         insn_byte_t b, lb;
  87         int i, nb;
  88
  89         if (prefixes->got)
  90                 return;
  91
  92         nb = 0;
  93         lb = 0;
  94         b = peek_next(insn_byte_t, insn);
  95         attr = inat_get_opcode_attribute(b);
  96         while (inat_is_legacy_prefix(attr)) {
  97                 /* Skip if same prefix */
  98                 for (i = 0; i < nb; i++)
  99                         if (prefixes->bytes[i] == b)
 100                                 goto found;
 101                 if (nb == 4)
 102                         /* Invalid instruction */
 103                         break;
 104                 prefixes->bytes[nb++] = b;
 105                 if (inat_is_address_size_prefix(attr)) {
 106                         /* address size switches 2/4 or 4/8 */
 107                         if (insn->x86_64)
 108                                 insn->addr_bytes ^= 12;
 109                         else
 110                                 insn->addr_bytes ^= 6;
 111                 } else if (inat_is_operand_size_prefix(attr)) {
 112                         /* oprand size switches 2/4 */
 113                         insn->opnd_bytes ^= 6;
 114                 }
 115 found:
 116                 prefixes->nbytes++;
 117                 insn->next_byte++;
 118                 lb = b;
 119                 b = peek_next(insn_byte_t, insn);
 120                 attr = inat_get_opcode_attribute(b);
 121         }
 122         /* Set the last prefix */
 123         if (lb && lb != insn->prefixes.bytes[3]) {
 124                 if (unlikely(insn->prefixes.bytes[3])) {
 125                         /* Swap the last prefix */
 126                         b = insn->prefixes.bytes[3];
 127                         for (i = 0; i < nb; i++)
 128                                 if (prefixes->bytes[i] == lb)
 129                                         prefixes->bytes[i] = b;
 130                 }
 131                 insn->prefixes.bytes[3] = lb;
 132         }
 133
 134         /* Decode REX prefix */
 135         if (insn->x86_64) {
 136                 b = peek_next(insn_byte_t, insn);
 137                 attr = inat_get_opcode_attribute(b);
 138                 if (inat_is_rex_prefix(attr)) {
 139                         insn->rex_prefix.value = b;
 140                         insn->rex_prefix.nbytes = 1;
 141                         insn->next_byte++;
 142                         if (X86_REX_W(b))
 143                                 /* REX.W overrides opnd_size */
 144                                 insn->opnd_bytes = 8;
 145                 }
 146         }
 147         insn->rex_prefix.got = 1;
 148
 149         /* Decode VEX prefix */
 150         b = peek_next(insn_byte_t, insn);
 151         attr = inat_get_opcode_attribute(b);
 152         if (inat_is_vex_prefix(attr)) {
 153                 insn_byte_t b2 = peek_nbyte_next(insn_byte_t, insn, 1);
 154                 if (!insn->x86_64) {
 155                         /*
 156                          * In 32-bits mode, if the [7:6] bits (mod bits of
 157                          * ModRM) on the second byte are not 11b, it is
 158                          * LDS or LES.
 159                          */
 160                         if (X86_MODRM_MOD(b2) != 3)
 161                                 goto vex_end;
 162                 }
 163                 insn->vex_prefix.bytes[0] = b;
 164                 insn->vex_prefix.bytes[1] = b2;
 165                 if (inat_is_vex3_prefix(attr)) {
 166                         b2 = peek_nbyte_next(insn_byte_t, insn, 2);
 167                         insn->vex_prefix.bytes[2] = b2;
 168                         insn->vex_prefix.nbytes = 3;
 169                         insn->next_byte += 3;
 170                         if (insn->x86_64 && X86_VEX_W(b2))
 171                                 /* VEX.W overrides opnd_size */
 172                                 insn->opnd_bytes = 8;
 173                 } else {
 174                         /*
 175                          * For VEX2, fake VEX3-like byte#2.
 176                          * Makes it easier to decode vex.W, vex.vvvv,
 177                          * vex.L and vex.pp. Masking with 0x7f sets vex.W == 0.
 178                          */
 179                         insn->vex_prefix.bytes[2] = b2 & 0x7f;
 180                         insn->vex_prefix.nbytes = 2;
 181                         insn->next_byte += 2;
 182                 }
 183         }
 184 vex_end:
 185         insn->vex_prefix.got = 1;
 186
 187         prefixes->got = 1;
 188
 189 err_out:
 190         return;
 191 }
 192
 193 /**
 194  * insn_get_opcode - collect opcode(s)
 195  * @insn:       &struct insn containing instruction
 196  *
 197  * Populates @insn->opcode, updates @insn->next_byte to point past the
 198  * opcode byte(s), and set @insn->attr (except for groups).
 199  * If necessary, first collects any preceding (prefix) bytes.
 200  * Sets @insn->opcode.value = opcode1.  No effect if @insn->opcode.got
 201  * is already 1.
 202  */
 203 void insn_get_opcode(struct insn *insn)
 204 {
 205         struct insn_field *opcode = &insn->opcode;
 206         insn_byte_t op;
 207         int pfx_id;
 208         if (opcode->got)
 209                 return;
 210         if (!insn->prefixes.got)
 211                 insn_get_prefixes(insn);
 212
 213         /* Get first opcode */
 214         op = get_next(insn_byte_t, insn);
 215         opcode->bytes[0] = op;
 216         opcode->nbytes = 1;
 217
 218         /* Check if there is VEX prefix or not */
 219         if (insn_is_avx(insn)) {
 220                 insn_byte_t m, p;
 221                 m = insn_vex_m_bits(insn);
 222                 p = insn_vex_p_bits(insn);
 223                 insn->attr = inat_get_avx_attribute(op, m, p);
 224                 if (!inat_accept_vex(insn->attr) && !inat_is_group(insn->attr))
 225                         insn->attr = 0; /* This instruction is bad */
 226                 goto end;       /* VEX has only 1 byte for opcode */
 227         }
 228
 229         insn->attr = inat_get_opcode_attribute(op);
 230         while (inat_is_escape(insn->attr)) {
 231                 /* Get escaped opcode */
 232                 op = get_next(insn_byte_t, insn);
 233                 opcode->bytes[opcode->nbytes++] = op;
 234                 pfx_id = insn_last_prefix_id(insn);
 235                 insn->attr = inat_get_escape_attribute(op, pfx_id, insn->attr);
 236         }
 237         if (inat_must_vex(insn->attr))
 238                 insn->attr = 0; /* This instruction is bad */
 239 end:
 240         opcode->got = 1;
 241
 242 err_out:
 243         return;
 244 }
 245
 246 /**
 247  * insn_get_modrm - collect ModRM byte, if any
 248  * @insn:       &struct insn containing instruction
 249  *
 250  * Populates @insn->modrm and updates @insn->next_byte to point past the
 251  * ModRM byte, if any.  If necessary, first collects the preceding bytes
 252  * (prefixes and opcode(s)).  No effect if @insn->modrm.got is already 1.
 253  */
 254 void insn_get_modrm(struct insn *insn)
 255 {
 256         struct insn_field *modrm = &insn->modrm;
 257         insn_byte_t pfx_id, mod;
 258         if (modrm->got)
 259                 return;
 260         if (!insn->opcode.got)
 261                 insn_get_opcode(insn);
 262
 263         if (inat_has_modrm(insn->attr)) {
 264                 mod = get_next(insn_byte_t, insn);
 265                 modrm->value = mod;
 266                 modrm->nbytes = 1;
 267                 if (inat_is_group(insn->attr)) {
 268                         pfx_id = insn_last_prefix_id(insn);
 269                         insn->attr = inat_get_group_attribute(mod, pfx_id,
 270                                                               insn->attr);
 271                         if (insn_is_avx(insn) && !inat_accept_vex(insn->attr))
 272                                 insn->attr = 0; /* This is bad */
 273                 }
 274         }
 275
 276         if (insn->x86_64 && inat_is_force64(insn->attr))
 277                 insn->opnd_bytes = 8;
 278         modrm->got = 1;
 279
 280 err_out:
 281         return;
 282 }
 283
 284
 285 /**
 286  * insn_rip_relative() - Does instruction use RIP-relative addressing mode?
 287  * @insn:       &struct insn containing instruction
 288  *
 289  * If necessary, first collects the instruction up to and including the
 290  * ModRM byte.  No effect if @insn->x86_64 is 0.
 291  */
 292 int insn_rip_relative(struct insn *insn)
 293 {
 294         struct insn_field *modrm = &insn->modrm;
 295
 296         if (!insn->x86_64)
 297                 return 0;
 298         if (!modrm->got)
 299                 insn_get_modrm(insn);
 300         /*
 301          * For rip-relative instructions, the mod field (top 2 bits)
 302          * is zero and the r/m field (bottom 3 bits) is 0x5.
 303          */
 304         return (modrm->nbytes && (modrm->value & 0xc7) == 0x5);
 305 }
 306
 307 /**
 308  * insn_get_sib() - Get the SIB byte of instruction
 309  * @insn:       &struct insn containing instruction
 310  *
 311  * If necessary, first collects the instruction up to and including the
 312  * ModRM byte.
 313  */
 314 void insn_get_sib(struct insn *insn)
 315 {
 316         insn_byte_t modrm;
 317
 318         if (insn->sib.got)
 319                 return;
 320         if (!insn->modrm.got)
 321                 insn_get_modrm(insn);
 322         if (insn->modrm.nbytes) {
 323                 modrm = (insn_byte_t)insn->modrm.value;
 324                 if (insn->addr_bytes != 2 &&
 325                     X86_MODRM_MOD(modrm) != 3 && X86_MODRM_RM(modrm) == 4) {
 326                         insn->sib.value = get_next(insn_byte_t, insn);
 327                         insn->sib.nbytes = 1;
 328                 }
 329         }
 330         insn->sib.got = 1;
 331
 332 err_out:
 333         return;
 334 }
 335
 336
 337 /**
 338  * insn_get_displacement() - Get the displacement of instruction
 339  * @insn:       &struct insn containing instruction
 340  *
 341  * If necessary, first collects the instruction up to and including the
 342  * SIB byte.
 343  * Displacement value is sign-expanded.
 344  */
 345 void insn_get_displacement(struct insn *insn)
 346 {
 347         insn_byte_t mod, rm, base;
 348
 349         if (insn->displacement.got)
 350                 return;
 351         if (!insn->sib.got)
 352                 insn_get_sib(insn);
 353         if (insn->modrm.nbytes) {
 354                 /*
 355                  * Interpreting the modrm byte:
 356                  * mod = 00 - no displacement fields (exceptions below)
 357                  * mod = 01 - 1-byte displacement field
 358                  * mod = 10 - displacement field is 4 bytes, or 2 bytes if
 359                  *      address size = 2 (0x67 prefix in 32-bit mode)
 360                  * mod = 11 - no memory operand
 361                  *
 362                  * If address size = 2...
 363                  * mod = 00, r/m = 110 - displacement field is 2 bytes
 364                  *
 365                  * If address size != 2...
 366                  * mod != 11, r/m = 100 - SIB byte exists
 367                  * mod = 00, SIB base = 101 - displacement field is 4 bytes
 368                  * mod = 00, r/m = 101 - rip-relative addressing, displacement
 369                  *      field is 4 bytes
 370                  */
 371                 mod = X86_MODRM_MOD(insn->modrm.value);
 372                 rm = X86_MODRM_RM(insn->modrm.value);
 373                 base = X86_SIB_BASE(insn->sib.value);
 374                 if (mod == 3)
 375                         goto out;
 376                 if (mod == 1) {
 377                         insn->displacement.value = get_next(char, insn);
 378                         insn->displacement.nbytes = 1;
 379                 } else if (insn->addr_bytes == 2) {
 380                         if ((mod == 0 && rm == 6) || mod == 2) {
 381                                 insn->displacement.value =
 382                                          get_next(short, insn);
 383                                 insn->displacement.nbytes = 2;
 384                         }
 385                 } else {
 386                         if ((mod == 0 && rm == 5) || mod == 2 ||
 387                             (mod == 0 && base == 5)) {
 388                                 insn->displacement.value = get_next(int, insn);
 389                                 insn->displacement.nbytes = 4;
 390                         }
 391                 }
 392         }
 393 out:
 394         insn->displacement.got = 1;
 395
 396 err_out:
 397         return;
 398 }
 399
 400 /* Decode moffset16/32/64. Return 0 if failed */
 401 static int __get_moffset(struct insn *insn)
 402 {
 403         switch (insn->addr_bytes) {
 404         case 2:
 405                 insn->moffset1.value = get_next(short, insn);
 406                 insn->moffset1.nbytes = 2;
 407                 break;
 408         case 4:
 409                 insn->moffset1.value = get_next(int, insn);
 410                 insn->moffset1.nbytes = 4;
 411                 break;
 412         case 8:
 413                 insn->moffset1.value = get_next(int, insn);
 414                 insn->moffset1.nbytes = 4;
 415                 insn->moffset2.value = get_next(int, insn);
 416                 insn->moffset2.nbytes = 4;
 417                 break;
 418         default:        /* opnd_bytes must be modified manually */
 419                 goto err_out;
 420         }
 421         insn->moffset1.got = insn->moffset2.got = 1;
 422
 423         return 1;
 424
 425 err_out:
 426         return 0;
 427 }
 428
 429 /* Decode imm v32(Iz). Return 0 if failed */
 430 static int __get_immv32(struct insn *insn)
 431 {
 432         switch (insn->opnd_bytes) {
 433         case 2:
 434                 insn->immediate.value = get_next(short, insn);
 435                 insn->immediate.nbytes = 2;
 436                 break;
 437         case 4:
 438         case 8:
 439                 insn->immediate.value = get_next(int, insn);
 440                 insn->immediate.nbytes = 4;
 441                 break;
 442         default:        /* opnd_bytes must be modified manually */
 443                 goto err_out;
 444         }
 445
 446         return 1;
 447
 448 err_out:
 449         return 0;
 450 }
 451
 452 /* Decode imm v64(Iv/Ov), Return 0 if failed */
 453 static int __get_immv(struct insn *insn)
 454 {
 455         switch (insn->opnd_bytes) {
 456         case 2:
 457                 insn->immediate1.value = get_next(short, insn);
 458                 insn->immediate1.nbytes = 2;
 459                 break;
 460         case 4:
 461                 insn->immediate1.value = get_next(int, insn);
 462                 insn->immediate1.nbytes = 4;
 463                 break;
 464         case 8:
 465                 insn->immediate1.value = get_next(int, insn);
 466                 insn->immediate1.nbytes = 4;
 467                 insn->immediate2.value = get_next(int, insn);
 468                 insn->immediate2.nbytes = 4;
 469                 break;
 470         default:        /* opnd_bytes must be modified manually */
 471                 goto err_out;
 472         }
 473         insn->immediate1.got = insn->immediate2.got = 1;
 474
 475         return 1;
 476 err_out:
 477         return 0;
 478 }
 479
 480 /* Decode ptr16:16/32(Ap) */
 481 static int __get_immptr(struct insn *insn)
 482 {
 483         switch (insn->opnd_bytes) {
 484         case 2:
 485                 insn->immediate1.value = get_next(short, insn);
 486                 insn->immediate1.nbytes = 2;
 487                 break;
 488         case 4:
 489                 insn->immediate1.value = get_next(int, insn);
 490                 insn->immediate1.nbytes = 4;
 491                 break;
 492         case 8:
 493                 /* ptr16:64 is not exist (no segment) */
 494                 return 0;
 495         default:        /* opnd_bytes must be modified manually */
 496                 goto err_out;
 497         }
 498         insn->immediate2.value = get_next(unsigned short, insn);
 499         insn->immediate2.nbytes = 2;
 500         insn->immediate1.got = insn->immediate2.got = 1;
 501
 502         return 1;
 503 err_out:
 504         return 0;
 505 }
 506
 507 /**
 508  * insn_get_immediate() - Get the immediates of instruction
 509  * @insn:       &struct insn containing instruction
 510  *
 511  * If necessary, first collects the instruction up to and including the
 512  * displacement bytes.
 513  * Basically, most of immediates are sign-expanded. Unsigned-value can be
 514  * get by bit masking with ((1 << (nbytes * 8)) - 1)
 515  */
 516 void insn_get_immediate(struct insn *insn)
 517 {
 518         if (insn->immediate.got)
 519                 return;
 520         if (!insn->displacement.got)
 521                 insn_get_displacement(insn);
 522
 523         if (inat_has_moffset(insn->attr)) {
 524                 if (!__get_moffset(insn))
 525                         goto err_out;
 526                 goto done;
 527         }
 528
 529         if (!inat_has_immediate(insn->attr))
 530                 /* no immediates */
 531                 goto done;
 532
 533         switch (inat_immediate_size(insn->attr)) {
 534         case INAT_IMM_BYTE:
 535                 insn->immediate.value = get_next(char, insn);
 536                 insn->immediate.nbytes = 1;
 537                 break;
 538         case INAT_IMM_WORD:
 539                 insn->immediate.value = get_next(short, insn);
 540                 insn->immediate.nbytes = 2;
 541                 break;
 542         case INAT_IMM_DWORD:
 543                 insn->immediate.value = get_next(int, insn);
 544                 insn->immediate.nbytes = 4;
 545                 break;
 546         case INAT_IMM_QWORD:
 547                 insn->immediate1.value = get_next(int, insn);
 548                 insn->immediate1.nbytes = 4;
 549                 insn->immediate2.value = get_next(int, insn);
 550                 insn->immediate2.nbytes = 4;
 551                 break;
 552         case INAT_IMM_PTR:
 553                 if (!__get_immptr(insn))
 554                         goto err_out;
 555                 break;
 556         case INAT_IMM_VWORD32:
 557                 if (!__get_immv32(insn))
 558                         goto err_out;
 559                 break;
 560         case INAT_IMM_VWORD:
 561                 if (!__get_immv(insn))
 562                         goto err_out;
 563                 break;
 564         default:
 565                 /* Here, insn must have an immediate, but failed */
 566                 goto err_out;
 567         }
 568         if (inat_has_second_immediate(insn->attr)) {
 569                 insn->immediate2.value = get_next(char, insn);
 570                 insn->immediate2.nbytes = 1;
 571         }
 572 done:
 573         insn->immediate.got = 1;
 574
 575 err_out:
 576         return;
 577 }
 578
 579 /**
 580  * insn_get_length() - Get the length of instruction
 581  * @insn:       &struct insn containing instruction
 582  *
 583  * If necessary, first collects the instruction up to and including the
 584  * immediates bytes.
 585  */
 586 void insn_get_length(struct insn *insn)
 587 {
 588         if (insn->length)
 589                 return;
 590         if (!insn->immediate.got)
 591                 insn_get_immediate(insn);
 592         insn->length = (unsigned char)((unsigned long)insn->next_byte
 593                                      - (unsigned long)insn->kaddr);
 594 }