nasmlib.c

   1 /* ----------------------------------------------------------------------- *
   2  *
   3  *   Copyright 1996-2009 The NASM Authors - All Rights Reserved
   4  *   See the file AUTHORS included with the NASM distribution for
   5  *   the specific copyright holders.
   6  *
   7  *   Redistribution and use in source and binary forms, with or without
   8  *   modification, are permitted provided that the following
   9  *   conditions are met:
  10  *
  11  *   * Redistributions of source code must retain the above copyright
  12  *     notice, this list of conditions and the following disclaimer.
  13  *   * Redistributions in binary form must reproduce the above
  14  *     copyright notice, this list of conditions and the following
  15  *     disclaimer in the documentation and/or other materials provided
  16  *     with the distribution.
  17  *
  18  *     THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
  19  *     CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
  20  *     INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
  21  *     MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
  22  *     DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
  23  *     CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  24  *     SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
  25  *     NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
  26  *     LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  *     HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
  28  *     CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
  29  *     OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
  30  *     EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  31  *
  32  * ----------------------------------------------------------------------- */
  33
  34 /*
  35  * nasmlib.c    library routines for the Netwide Assembler
  36  */
  37
  38 #include "compiler.h"
  39
  40 #include <stdio.h>
  41 #include <stdlib.h>
  42 #include <string.h>
  43 #include <ctype.h>
  44 #include <inttypes.h>
  45
  46 #include "nasm.h"
  47 #include "nasmlib.h"
  48 #include "insns.h"
  49
  50 int globalbits = 0;    /* defined in nasm.h, works better here for ASM+DISASM */
  51 static vefunc nasm_verror;      /* Global error handling function */
  52
  53 #ifdef LOGALLOC
  54 static FILE *logfp;
  55 #endif
  56
  57 /* Uninitialized -> all zero by C spec */
  58 const uint8_t zero_buffer[ZERO_BUF_SIZE];
  59
  60 /*
  61  * Prepare a table of tolower() results.  This avoids function calls
  62  * on some platforms.
  63  */
  64
  65 unsigned char nasm_tolower_tab[256];
  66
  67 void tolower_init(void)
  68 {
  69     int i;
  70
  71     for (i = 0; i < 256; i++)
  72         nasm_tolower_tab[i] = tolower(i);
  73 }
  74
  75 void nasm_set_verror(vefunc ve)
  76 {
  77     nasm_verror = ve;
  78 }
  79
  80 void nasm_error(int severity, const char *fmt, ...)
  81 {
  82     va_list ap;
  83
  84     va_start(ap, fmt);
  85     nasm_verror(severity, fmt, ap);
  86     va_end(ap);
  87 }
  88
  89 void nasm_init_malloc_error(void)
  90 {
  91 #ifdef LOGALLOC
  92     logfp = fopen("malloc.log", "w");
  93     setvbuf(logfp, NULL, _IOLBF, BUFSIZ);
  94     fprintf(logfp, "null pointer is %p\n", NULL);
  95 #endif
  96 }
  97
  98 #ifdef LOGALLOC
  99 void *nasm_malloc_log(const char *file, int line, size_t size)
 100 #else
 101 void *nasm_malloc(size_t size)
 102 #endif
 103 {
 104     void *p = malloc(size);
 105     if (!p)
 106         nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
 107 #ifdef LOGALLOC
 108     else
 109         fprintf(logfp, "%s %d malloc(%ld) returns %p\n",
 110                 file, line, (long)size, p);
 111 #endif
 112     return p;
 113 }
 114
 115 #ifdef LOGALLOC
 116 void *nasm_zalloc_log(const char *file, int line, size_t size)
 117 #else
 118 void *nasm_zalloc(size_t size)
 119 #endif
 120 {
 121     void *p = calloc(size, 1);
 122     if (!p)
 123         nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
 124 #ifdef LOGALLOC
 125     else
 126         fprintf(logfp, "%s %d calloc(%ld, 1) returns %p\n",
 127                 file, line, (long)size, p);
 128 #endif
 129     return p;
 130 }
 131
 132 #ifdef LOGALLOC
 133 void *nasm_realloc_log(const char *file, int line, void *q, size_t size)
 134 #else
 135 void *nasm_realloc(void *q, size_t size)
 136 #endif
 137 {
 138     void *p = q ? realloc(q, size) : malloc(size);
 139     if (!p)
 140         nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
 141 #ifdef LOGALLOC
 142     else if (q)
 143         fprintf(logfp, "%s %d realloc(%p,%ld) returns %p\n",
 144                 file, line, q, (long)size, p);
 145     else
 146         fprintf(logfp, "%s %d malloc(%ld) returns %p\n",
 147                 file, line, (long)size, p);
 148 #endif
 149     return p;
 150 }
 151
 152 #ifdef LOGALLOC
 153 void nasm_free_log(const char *file, int line, void *q)
 154 #else
 155 void nasm_free(void *q)
 156 #endif
 157 {
 158     if (q) {
 159 #ifdef LOGALLOC
 160         fprintf(logfp, "%s %d free(%p)\n", file, line, q);
 161 #endif
 162         free(q);
 163     }
 164 }
 165
 166 #ifdef LOGALLOC
 167 char *nasm_strdup_log(const char *file, int line, const char *s)
 168 #else
 169 char *nasm_strdup(const char *s)
 170 #endif
 171 {
 172     char *p;
 173     int size = strlen(s) + 1;
 174
 175     p = malloc(size);
 176     if (!p)
 177         nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
 178 #ifdef LOGALLOC
 179     else
 180         fprintf(logfp, "%s %d strdup(%ld) returns %p\n",
 181                 file, line, (long)size, p);
 182 #endif
 183     strcpy(p, s);
 184     return p;
 185 }
 186
 187 #ifdef LOGALLOC
 188 char *nasm_strndup_log(const char *file, int line, const char *s, size_t len)
 189 #else
 190 char *nasm_strndup(const char *s, size_t len)
 191 #endif
 192 {
 193     char *p;
 194     int size = len + 1;
 195
 196     p = malloc(size);
 197     if (!p)
 198         nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
 199 #ifdef LOGALLOC
 200     else
 201         fprintf(logfp, "%s %d strndup(%ld) returns %p\n",
 202                 file, line, (long)size, p);
 203 #endif
 204     strncpy(p, s, len);
 205     p[len] = '\0';
 206     return p;
 207 }
 208
 209 no_return nasm_assert_failed(const char *file, int line, const char *msg)
 210 {
 211     nasm_error(ERR_FATAL, "assertion %s failed at %s:%d", msg, file, line);
 212     exit(1);
 213 }
 214
 215 #ifndef nasm_stricmp
 216 int nasm_stricmp(const char *s1, const char *s2)
 217 {
 218     unsigned char c1, c2;
 219     int d;
 220
 221     while (1) {
 222         c1 = nasm_tolower(*s1++);
 223         c2 = nasm_tolower(*s2++);
 224         d = c1-c2;
 225
 226         if (d)
 227             return d;
 228         if (!c1)
 229             break;
 230     }
 231     return 0;
 232 }
 233 #endif
 234
 235 #ifndef nasm_strnicmp
 236 int nasm_strnicmp(const char *s1, const char *s2, size_t n)
 237 {
 238     unsigned char c1, c2;
 239     int d;
 240
 241     while (n--) {
 242         c1 = nasm_tolower(*s1++);
 243         c2 = nasm_tolower(*s2++);
 244         d = c1-c2;
 245
 246         if (d)
 247             return d;
 248         if (!c1)
 249             break;
 250     }
 251     return 0;
 252 }
 253 #endif
 254
 255 int nasm_memicmp(const char *s1, const char *s2, size_t n)
 256 {
 257     unsigned char c1, c2;
 258     int d;
 259
 260     while (n--) {
 261         c1 = nasm_tolower(*s1++);
 262         c2 = nasm_tolower(*s2++);
 263         d = c1-c2;
 264         if (d)
 265             return d;
 266     }
 267     return 0;
 268 }
 269
 270 #ifndef nasm_strsep
 271 char *nasm_strsep(char **stringp, const char *delim)
 272 {
 273         char *s = *stringp;
 274         char *e;
 275
 276         if (!s)
 277                 return NULL;
 278
 279         e = strpbrk(s, delim);
 280         if (e)
 281                 *e++ = '\0';
 282
 283         *stringp = e;
 284         return s;
 285 }
 286 #endif
 287
 288
 289 #define lib_isnumchar(c)   (nasm_isalnum(c) || (c) == '$' || (c) == '_')
 290 #define numvalue(c)  ((c)>='a' ? (c)-'a'+10 : (c)>='A' ? (c)-'A'+10 : (c)-'0')
 291
 292 static int radix_letter(char c)
 293 {
 294     switch (c) {
 295     case 'b': case 'B':
 296     case 'y': case 'Y':
 297         return 2;               /* Binary */
 298     case 'o': case 'O':
 299     case 'q': case 'Q':
 300         return 8;               /* Octal */
 301     case 'h': case 'H':
 302     case 'x': case 'X':
 303         return 16;              /* Hexadecimal */
 304     case 'd': case 'D':
 305     case 't': case 'T':
 306         return 10;              /* Decimal */
 307     default:
 308         return 0;               /* Not a known radix letter */
 309     }
 310 }
 311
 312 int64_t readnum(char *str, bool *error)
 313 {
 314     char *r = str, *q;
 315     int32_t pradix, sradix, radix;
 316     int plen, slen, len;
 317     uint64_t result, checklimit;
 318     int digit, last;
 319     bool warn = false;
 320     int sign = 1;
 321
 322     *error = false;
 323
 324     while (nasm_isspace(*r))
 325         r++;                    /* find start of number */
 326
 327     /*
 328      * If the number came from make_tok_num (as a result of an %assign), it
 329      * might have a '-' built into it (rather than in a preceeding token).
 330      */
 331     if (*r == '-') {
 332         r++;
 333         sign = -1;
 334     }
 335
 336     q = r;
 337
 338     while (lib_isnumchar(*q))
 339         q++;                    /* find end of number */
 340
 341     len = q-r;
 342     if (!len) {
 343         /* Not numeric */
 344         *error = true;
 345         return 0;
 346     }
 347
 348     /*
 349      * Handle radix formats:
 350      *
 351      * 0<radix-letter><string>
 352      * $<string>                (hexadecimal)
 353      * <string><radix-letter>
 354      */
 355     pradix = sradix = 0;
 356     plen = slen = 0;
 357
 358     if (len > 2 && *r == '0' && (pradix = radix_letter(r[1])) != 0)
 359         plen = 2;
 360     else if (len > 1 && *r == '$')
 361         pradix = 16, plen = 1;
 362
 363     if (len > 1 && (sradix = radix_letter(q[-1])) != 0)
 364         slen = 1;
 365
 366     if (pradix > sradix) {
 367         radix = pradix;
 368         r += plen;
 369     } else if (sradix > pradix) {
 370         radix = sradix;
 371         q -= slen;
 372     } else {
 373         /* Either decimal, or invalid -- if invalid, we'll trip up
 374            further down. */
 375         radix = 10;
 376     }
 377
 378     /*
 379      * `checklimit' must be 2**64 / radix. We can't do that in
 380      * 64-bit arithmetic, which we're (probably) using, so we
 381      * cheat: since we know that all radices we use are even, we
 382      * can divide 2**63 by radix/2 instead.
 383      */
 384     checklimit = UINT64_C(0x8000000000000000) / (radix >> 1);
 385
 386     /*
 387      * Calculate the highest allowable value for the last digit of a
 388      * 64-bit constant... in radix 10, it is 6, otherwise it is 0
 389      */
 390     last = (radix == 10 ? 6 : 0);
 391
 392     result = 0;
 393     while (*r && r < q) {
 394         if (*r != '_') {
 395             if (*r < '0' || (*r > '9' && *r < 'A')
 396                 || (digit = numvalue(*r)) >= radix) {
 397                 *error = true;
 398                 return 0;
 399             }
 400             if (result > checklimit ||
 401                 (result == checklimit && digit >= last)) {
 402                 warn = true;
 403             }
 404
 405             result = radix * result + digit;
 406         }
 407         r++;
 408     }
 409
 410     if (warn)
 411         nasm_error(ERR_WARNING | ERR_PASS1 | ERR_WARN_NOV,
 412                    "numeric constant %s does not fit in 64 bits",
 413                    str);
 414
 415     return result * sign;
 416 }
 417
 418 int64_t readstrnum(char *str, int length, bool *warn)
 419 {
 420     int64_t charconst = 0;
 421     int i;
 422
 423     *warn = false;
 424
 425     str += length;
 426     if (globalbits == 64) {
 427         for (i = 0; i < length; i++) {
 428             if (charconst & UINT64_C(0xFF00000000000000))
 429                 *warn = true;
 430             charconst = (charconst << 8) + (uint8_t)*--str;
 431         }
 432     } else {
 433         for (i = 0; i < length; i++) {
 434             if (charconst & 0xFF000000UL)
 435                 *warn = true;
 436             charconst = (charconst << 8) + (uint8_t)*--str;
 437         }
 438     }
 439     return charconst;
 440 }
 441
 442 static int32_t next_seg;
 443
 444 void seg_init(void)
 445 {
 446     next_seg = 0;
 447 }
 448
 449 int32_t seg_alloc(void)
 450 {
 451     return (next_seg += 2) - 2;
 452 }
 453
 454 #ifdef WORDS_LITTLEENDIAN
 455
 456 void fwriteint16_t(uint16_t data, FILE * fp)
 457 {
 458     fwrite(&data, 1, 2, fp);
 459 }
 460
 461 void fwriteint32_t(uint32_t data, FILE * fp)
 462 {
 463     fwrite(&data, 1, 4, fp);
 464 }
 465
 466 void fwriteint64_t(uint64_t data, FILE * fp)
 467 {
 468     fwrite(&data, 1, 8, fp);
 469 }
 470
 471 void fwriteaddr(uint64_t data, int size, FILE * fp)
 472 {
 473     fwrite(&data, 1, size, fp);
 474 }
 475
 476 #else /* not WORDS_LITTLEENDIAN */
 477
 478 void fwriteint16_t(uint16_t data, FILE * fp)
 479 {
 480     char buffer[2], *p = buffer;
 481     WRITESHORT(p, data);
 482     fwrite(buffer, 1, 2, fp);
 483 }
 484
 485 void fwriteint32_t(uint32_t data, FILE * fp)
 486 {
 487     char buffer[4], *p = buffer;
 488     WRITELONG(p, data);
 489     fwrite(buffer, 1, 4, fp);
 490 }
 491
 492 void fwriteint64_t(uint64_t data, FILE * fp)
 493 {
 494     char buffer[8], *p = buffer;
 495     WRITEDLONG(p, data);
 496     fwrite(buffer, 1, 8, fp);
 497 }
 498
 499 void fwriteaddr(uint64_t data, int size, FILE * fp)
 500 {
 501     char buffer[8], *p = buffer;
 502     WRITEADDR(p, data, size);
 503     fwrite(buffer, 1, size, fp);
 504 }
 505
 506 #endif
 507
 508 size_t fwritezero(size_t bytes, FILE *fp)
 509 {
 510     size_t count = 0;
 511     size_t blksize;
 512     size_t rv;
 513
 514     while (bytes) {
 515         blksize = (bytes < ZERO_BUF_SIZE) ? bytes : ZERO_BUF_SIZE;
 516
 517         rv = fwrite(zero_buffer, 1, blksize, fp);
 518         if (!rv)
 519             break;
 520
 521         count += rv;
 522         bytes -= rv;
 523     }
 524
 525     return count;
 526 }
 527
 528 void standard_extension(char *inname, char *outname, char *extension)
 529 {
 530     char *p, *q;
 531
 532     if (*outname)               /* file name already exists, */
 533         return;                 /* so do nothing */
 534     q = inname;
 535     p = outname;
 536     while (*q)
 537         *p++ = *q++;            /* copy, and find end of string */
 538     *p = '\0';                  /* terminate it */
 539     while (p > outname && *--p != '.') ;        /* find final period (or whatever) */
 540     if (*p != '.')
 541         while (*p)
 542             p++;                /* go back to end if none found */
 543     if (!strcmp(p, extension)) {        /* is the extension already there? */
 544         if (*extension)
 545             nasm_error(ERR_WARNING | ERR_NOFILE,
 546                        "file name already ends in `%s': "
 547                        "output will be in `nasm.out'", extension);
 548         else
 549             nasm_error(ERR_WARNING | ERR_NOFILE,
 550                        "file name already has no extension: "
 551                        "output will be in `nasm.out'");
 552         strcpy(outname, "nasm.out");
 553     } else
 554         strcpy(p, extension);
 555 }
 556
 557 /*
 558  * Common list of prefix names
 559  */
 560 static const char *prefix_names[] = {
 561     "a16", "a32", "a64", "asp", "lock", "o16", "o32", "o64", "osp",
 562     "rep", "repe", "repne", "repnz", "repz", "times", "wait"
 563 };
 564
 565 const char *prefix_name(int token)
 566 {
 567     unsigned int prefix = token-PREFIX_ENUM_START;
 568     if (prefix > elements(prefix_names))
 569         return NULL;
 570
 571     return prefix_names[prefix];
 572 }
 573
 574 /*
 575  * Binary search.
 576  */
 577 int bsi(const char *string, const char **array, int size)
 578 {
 579     int i = -1, j = size;       /* always, i < index < j */
 580     while (j - i >= 2) {
 581         int k = (i + j) / 2;
 582         int l = strcmp(string, array[k]);
 583         if (l < 0)              /* it's in the first half */
 584             j = k;
 585         else if (l > 0)         /* it's in the second half */
 586             i = k;
 587         else                    /* we've got it :) */
 588             return k;
 589     }
 590     return -1;                  /* we haven't got it :( */
 591 }
 592
 593 int bsii(const char *string, const char **array, int size)
 594 {
 595     int i = -1, j = size;       /* always, i < index < j */
 596     while (j - i >= 2) {
 597         int k = (i + j) / 2;
 598         int l = nasm_stricmp(string, array[k]);
 599         if (l < 0)              /* it's in the first half */
 600             j = k;
 601         else if (l > 0)         /* it's in the second half */
 602             i = k;
 603         else                    /* we've got it :) */
 604             return k;
 605     }
 606     return -1;                  /* we haven't got it :( */
 607 }
 608
 609 static char *file_name = NULL;
 610 static int32_t line_number = 0;
 611
 612 char *src_set_fname(char *newname)
 613 {
 614     char *oldname = file_name;
 615     file_name = newname;
 616     return oldname;
 617 }
 618
 619 int32_t src_set_linnum(int32_t newline)
 620 {
 621     int32_t oldline = line_number;
 622     line_number = newline;
 623     return oldline;
 624 }
 625
 626 int32_t src_get_linnum(void)
 627 {
 628     return line_number;
 629 }
 630
 631 int src_get(int32_t *xline, char **xname)
 632 {
 633     if (!file_name || !*xname || strcmp(*xname, file_name)) {
 634         nasm_free(*xname);
 635         *xname = file_name ? nasm_strdup(file_name) : NULL;
 636         *xline = line_number;
 637         return -2;
 638     }
 639     if (*xline != line_number) {
 640         int32_t tmp = line_number - *xline;
 641         *xline = line_number;
 642         return tmp;
 643     }
 644     return 0;
 645 }
 646
 647 char *nasm_strcat(const char *one, const char *two)
 648 {
 649     char *rslt;
 650     int l1 = strlen(one);
 651     rslt = nasm_malloc(l1 + strlen(two) + 1);
 652     strcpy(rslt, one);
 653     strcpy(rslt + l1, two);
 654     return rslt;
 655 }
 656
 657 /* skip leading spaces */
 658 char *nasm_skip_spaces(const char *p)
 659 {
 660     if (p)
 661         while (*p && nasm_isspace(*p))
 662             p++;
 663     return (char *)p;
 664 }
 665
 666 /* skip leading non-spaces */
 667 char *nasm_skip_word(const char *p)
 668 {
 669     if (p)
 670         while (*p && !nasm_isspace(*p))
 671             p++;
 672     return (char *)p;
 673 }
 674
 675 /* zap leading spaces with zero */
 676 char *nasm_zap_spaces_fwd(char *p)
 677 {
 678     if (p)
 679         while (*p && nasm_isspace(*p))
 680             *p++ = 0x0;
 681     return p;
 682 }
 683
 684 /* zap spaces with zero in reverse order */
 685 char *nasm_zap_spaces_rev(char *p)
 686 {
 687     if (p)
 688         while (*p && nasm_isspace(*p))
 689             *p-- = 0x0;
 690     return p;
 691 }
 692
 693 /*
 694  * initialized data bytes length from opcode
 695  */
 696 int idata_bytes(int opcode)
 697 {
 698     int ret;
 699     switch (opcode) {
 700     case I_DB:
 701         ret = 1;
 702         break;
 703     case I_DW:
 704         ret = 2;
 705         break;
 706     case I_DD:
 707         ret = 4;
 708         break;
 709     case I_DQ:
 710         ret = 8;
 711         break;
 712     case I_DT:
 713         ret = 10;
 714         break;
 715     case I_DO:
 716         ret = 16;
 717         break;
 718     case I_DY:
 719         ret = 32;
 720         break;
 721     case I_none:
 722         ret = -1;
 723         break;
 724     default:
 725         ret = 0;
 726         break;
 727     }
 728     return ret;
 729 }