insns: correct the TSX opcodes
[nasm/nasm.git] / nasmlib.c
blobe35d99e87f0102fa1df8d4187fb5bd7c08fe60d4
1 /* ----------------------------------------------------------------------- *
2 *
3 * Copyright 1996-2010 The NASM Authors - All Rights Reserved
4 * See the file AUTHORS included with the NASM distribution for
5 * the specific copyright holders.
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following
9 * conditions are met:
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.
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.
32 * ----------------------------------------------------------------------- */
35 * nasmlib.c library routines for the Netwide Assembler
38 #include "compiler.h"
40 #include <stdio.h>
41 #include <stdlib.h>
42 #include <string.h>
43 #include <ctype.h>
44 #include <inttypes.h>
46 #include "nasm.h"
47 #include "nasmlib.h"
48 #include "insns.h"
50 int globalbits = 0; /* defined in nasm.h, works better here for ASM+DISASM */
51 static vefunc nasm_verror; /* Global error handling function */
53 #ifdef LOGALLOC
54 static FILE *logfp;
55 #endif
57 /* Uninitialized -> all zero by C spec */
58 const uint8_t zero_buffer[ZERO_BUF_SIZE];
61 * Prepare a table of tolower() results. This avoids function calls
62 * on some platforms.
65 unsigned char nasm_tolower_tab[256];
67 void tolower_init(void)
69 int i;
71 for (i = 0; i < 256; i++)
72 nasm_tolower_tab[i] = tolower(i);
75 void nasm_set_verror(vefunc ve)
77 nasm_verror = ve;
80 void nasm_error(int severity, const char *fmt, ...)
82 va_list ap;
84 va_start(ap, fmt);
85 nasm_verror(severity, fmt, ap);
86 va_end(ap);
89 void nasm_init_malloc_error(void)
91 #ifdef LOGALLOC
92 logfp = fopen("malloc.log", "w");
93 if (logfp) {
94 setvbuf(logfp, NULL, _IOLBF, BUFSIZ);
95 } else {
96 nasm_error(ERR_NONFATAL | ERR_NOFILE, "Unable to open %s", logfp);
97 logfp = stderr;
99 fprintf(logfp, "null pointer is %p\n", NULL);
100 #endif
103 #ifdef LOGALLOC
104 void *nasm_malloc_log(const char *file, int line, size_t size)
105 #else
106 void *nasm_malloc(size_t size)
107 #endif
109 void *p = malloc(size);
110 if (!p)
111 nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
112 #ifdef LOGALLOC
113 else
114 fprintf(logfp, "%s %d malloc(%ld) returns %p\n",
115 file, line, (long)size, p);
116 #endif
117 return p;
120 #ifdef LOGALLOC
121 void *nasm_zalloc_log(const char *file, int line, size_t size)
122 #else
123 void *nasm_zalloc(size_t size)
124 #endif
126 void *p = calloc(size, 1);
127 if (!p)
128 nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
129 #ifdef LOGALLOC
130 else
131 fprintf(logfp, "%s %d calloc(%ld, 1) returns %p\n",
132 file, line, (long)size, p);
133 #endif
134 return p;
137 #ifdef LOGALLOC
138 void *nasm_realloc_log(const char *file, int line, void *q, size_t size)
139 #else
140 void *nasm_realloc(void *q, size_t size)
141 #endif
143 void *p = q ? realloc(q, size) : malloc(size);
144 if (!p)
145 nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
146 #ifdef LOGALLOC
147 else if (q)
148 fprintf(logfp, "%s %d realloc(%p,%ld) returns %p\n",
149 file, line, q, (long)size, p);
150 else
151 fprintf(logfp, "%s %d malloc(%ld) returns %p\n",
152 file, line, (long)size, p);
153 #endif
154 return p;
157 #ifdef LOGALLOC
158 void nasm_free_log(const char *file, int line, void *q)
159 #else
160 void nasm_free(void *q)
161 #endif
163 if (q) {
164 #ifdef LOGALLOC
165 fprintf(logfp, "%s %d free(%p)\n", file, line, q);
166 #endif
167 free(q);
171 #ifdef LOGALLOC
172 char *nasm_strdup_log(const char *file, int line, const char *s)
173 #else
174 char *nasm_strdup(const char *s)
175 #endif
177 char *p;
178 int size = strlen(s) + 1;
180 p = malloc(size);
181 if (!p)
182 nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
183 #ifdef LOGALLOC
184 else
185 fprintf(logfp, "%s %d strdup(%ld) returns %p\n",
186 file, line, (long)size, p);
187 #endif
188 strcpy(p, s);
189 return p;
192 #ifdef LOGALLOC
193 char *nasm_strndup_log(const char *file, int line, const char *s, size_t len)
194 #else
195 char *nasm_strndup(const char *s, size_t len)
196 #endif
198 char *p;
199 int size = len + 1;
201 p = malloc(size);
202 if (!p)
203 nasm_error(ERR_FATAL | ERR_NOFILE, "out of memory");
204 #ifdef LOGALLOC
205 else
206 fprintf(logfp, "%s %d strndup(%ld) returns %p\n",
207 file, line, (long)size, p);
208 #endif
209 strncpy(p, s, len);
210 p[len] = '\0';
211 return p;
214 no_return nasm_assert_failed(const char *file, int line, const char *msg)
216 nasm_error(ERR_FATAL, "assertion %s failed at %s:%d", msg, file, line);
217 exit(1);
220 #ifndef nasm_stricmp
221 int nasm_stricmp(const char *s1, const char *s2)
223 unsigned char c1, c2;
224 int d;
226 while (1) {
227 c1 = nasm_tolower(*s1++);
228 c2 = nasm_tolower(*s2++);
229 d = c1-c2;
231 if (d)
232 return d;
233 if (!c1)
234 break;
236 return 0;
238 #endif
240 #ifndef nasm_strnicmp
241 int nasm_strnicmp(const char *s1, const char *s2, size_t n)
243 unsigned char c1, c2;
244 int d;
246 while (n--) {
247 c1 = nasm_tolower(*s1++);
248 c2 = nasm_tolower(*s2++);
249 d = c1-c2;
251 if (d)
252 return d;
253 if (!c1)
254 break;
256 return 0;
258 #endif
260 int nasm_memicmp(const char *s1, const char *s2, size_t n)
262 unsigned char c1, c2;
263 int d;
265 while (n--) {
266 c1 = nasm_tolower(*s1++);
267 c2 = nasm_tolower(*s2++);
268 d = c1-c2;
269 if (d)
270 return d;
272 return 0;
275 #ifndef nasm_strsep
276 char *nasm_strsep(char **stringp, const char *delim)
278 char *s = *stringp;
279 char *e;
281 if (!s)
282 return NULL;
284 e = strpbrk(s, delim);
285 if (e)
286 *e++ = '\0';
288 *stringp = e;
289 return s;
291 #endif
294 #define lib_isnumchar(c) (nasm_isalnum(c) || (c) == '$' || (c) == '_')
296 static int radix_letter(char c)
298 switch (c) {
299 case 'b': case 'B':
300 case 'y': case 'Y':
301 return 2; /* Binary */
302 case 'o': case 'O':
303 case 'q': case 'Q':
304 return 8; /* Octal */
305 case 'h': case 'H':
306 case 'x': case 'X':
307 return 16; /* Hexadecimal */
308 case 'd': case 'D':
309 case 't': case 'T':
310 return 10; /* Decimal */
311 default:
312 return 0; /* Not a known radix letter */
316 int64_t readnum(char *str, bool *error)
318 char *r = str, *q;
319 int32_t pradix, sradix, radix;
320 int plen, slen, len;
321 uint64_t result, checklimit;
322 int digit, last;
323 bool warn = false;
324 int sign = 1;
326 *error = false;
328 while (nasm_isspace(*r))
329 r++; /* find start of number */
332 * If the number came from make_tok_num (as a result of an %assign), it
333 * might have a '-' built into it (rather than in a preceeding token).
335 if (*r == '-') {
336 r++;
337 sign = -1;
340 q = r;
342 while (lib_isnumchar(*q))
343 q++; /* find end of number */
345 len = q-r;
346 if (!len) {
347 /* Not numeric */
348 *error = true;
349 return 0;
353 * Handle radix formats:
355 * 0<radix-letter><string>
356 * $<string> (hexadecimal)
357 * <string><radix-letter>
359 pradix = sradix = 0;
360 plen = slen = 0;
362 if (len > 2 && *r == '0' && (pradix = radix_letter(r[1])) != 0)
363 plen = 2;
364 else if (len > 1 && *r == '$')
365 pradix = 16, plen = 1;
367 if (len > 1 && (sradix = radix_letter(q[-1])) != 0)
368 slen = 1;
370 if (pradix > sradix) {
371 radix = pradix;
372 r += plen;
373 } else if (sradix > pradix) {
374 radix = sradix;
375 q -= slen;
376 } else {
377 /* Either decimal, or invalid -- if invalid, we'll trip up
378 further down. */
379 radix = 10;
383 * `checklimit' must be 2**64 / radix. We can't do that in
384 * 64-bit arithmetic, which we're (probably) using, so we
385 * cheat: since we know that all radices we use are even, we
386 * can divide 2**63 by radix/2 instead.
388 checklimit = UINT64_C(0x8000000000000000) / (radix >> 1);
391 * Calculate the highest allowable value for the last digit of a
392 * 64-bit constant... in radix 10, it is 6, otherwise it is 0
394 last = (radix == 10 ? 6 : 0);
396 result = 0;
397 while (*r && r < q) {
398 if (*r != '_') {
399 if (*r < '0' || (*r > '9' && *r < 'A')
400 || (digit = numvalue(*r)) >= radix) {
401 *error = true;
402 return 0;
404 if (result > checklimit ||
405 (result == checklimit && digit >= last)) {
406 warn = true;
409 result = radix * result + digit;
411 r++;
414 if (warn)
415 nasm_error(ERR_WARNING | ERR_PASS1 | ERR_WARN_NOV,
416 "numeric constant %s does not fit in 64 bits",
417 str);
419 return result * sign;
422 int64_t readstrnum(char *str, int length, bool *warn)
424 int64_t charconst = 0;
425 int i;
427 *warn = false;
429 str += length;
430 if (globalbits == 64) {
431 for (i = 0; i < length; i++) {
432 if (charconst & UINT64_C(0xFF00000000000000))
433 *warn = true;
434 charconst = (charconst << 8) + (uint8_t)*--str;
436 } else {
437 for (i = 0; i < length; i++) {
438 if (charconst & 0xFF000000UL)
439 *warn = true;
440 charconst = (charconst << 8) + (uint8_t)*--str;
443 return charconst;
446 static int32_t next_seg;
448 void seg_init(void)
450 next_seg = 0;
453 int32_t seg_alloc(void)
455 return (next_seg += 2) - 2;
458 #ifdef WORDS_LITTLEENDIAN
460 void fwriteint16_t(uint16_t data, FILE * fp)
462 fwrite(&data, 1, 2, fp);
465 void fwriteint32_t(uint32_t data, FILE * fp)
467 fwrite(&data, 1, 4, fp);
470 void fwriteint64_t(uint64_t data, FILE * fp)
472 fwrite(&data, 1, 8, fp);
475 void fwriteaddr(uint64_t data, int size, FILE * fp)
477 fwrite(&data, 1, size, fp);
480 #else /* not WORDS_LITTLEENDIAN */
482 void fwriteint16_t(uint16_t data, FILE * fp)
484 char buffer[2], *p = buffer;
485 WRITESHORT(p, data);
486 fwrite(buffer, 1, 2, fp);
489 void fwriteint32_t(uint32_t data, FILE * fp)
491 char buffer[4], *p = buffer;
492 WRITELONG(p, data);
493 fwrite(buffer, 1, 4, fp);
496 void fwriteint64_t(uint64_t data, FILE * fp)
498 char buffer[8], *p = buffer;
499 WRITEDLONG(p, data);
500 fwrite(buffer, 1, 8, fp);
503 void fwriteaddr(uint64_t data, int size, FILE * fp)
505 char buffer[8], *p = buffer;
506 WRITEADDR(p, data, size);
507 fwrite(buffer, 1, size, fp);
510 #endif
512 size_t fwritezero(size_t bytes, FILE *fp)
514 size_t count = 0;
515 size_t blksize;
516 size_t rv;
518 while (bytes) {
519 blksize = (bytes < ZERO_BUF_SIZE) ? bytes : ZERO_BUF_SIZE;
521 rv = fwrite(zero_buffer, 1, blksize, fp);
522 if (!rv)
523 break;
525 count += rv;
526 bytes -= rv;
529 return count;
532 void standard_extension(char *inname, char *outname, char *extension)
534 char *p, *q;
536 if (*outname) /* file name already exists, */
537 return; /* so do nothing */
538 q = inname;
539 p = outname;
540 while (*q)
541 *p++ = *q++; /* copy, and find end of string */
542 *p = '\0'; /* terminate it */
543 while (p > outname && *--p != '.') ; /* find final period (or whatever) */
544 if (*p != '.')
545 while (*p)
546 p++; /* go back to end if none found */
547 if (!strcmp(p, extension)) { /* is the extension already there? */
548 if (*extension)
549 nasm_error(ERR_WARNING | ERR_NOFILE,
550 "file name already ends in `%s': "
551 "output will be in `nasm.out'", extension);
552 else
553 nasm_error(ERR_WARNING | ERR_NOFILE,
554 "file name already has no extension: "
555 "output will be in `nasm.out'");
556 strcpy(outname, "nasm.out");
557 } else
558 strcpy(p, extension);
562 * Common list of prefix names
564 static const char *prefix_names[] = {
565 "a16", "a32", "a64", "asp", "lock", "o16", "o32", "o64", "osp",
566 "rep", "repe", "repne", "repnz", "repz", "times", "wait"
569 const char *prefix_name(int token)
571 unsigned int prefix = token-PREFIX_ENUM_START;
572 if (prefix >= ARRAY_SIZE(prefix_names))
573 return NULL;
575 return prefix_names[prefix];
579 * Binary search.
581 int bsi(const char *string, const char **array, int size)
583 int i = -1, j = size; /* always, i < index < j */
584 while (j - i >= 2) {
585 int k = (i + j) / 2;
586 int l = strcmp(string, array[k]);
587 if (l < 0) /* it's in the first half */
588 j = k;
589 else if (l > 0) /* it's in the second half */
590 i = k;
591 else /* we've got it :) */
592 return k;
594 return -1; /* we haven't got it :( */
597 int bsii(const char *string, const char **array, int size)
599 int i = -1, j = size; /* always, i < index < j */
600 while (j - i >= 2) {
601 int k = (i + j) / 2;
602 int l = nasm_stricmp(string, array[k]);
603 if (l < 0) /* it's in the first half */
604 j = k;
605 else if (l > 0) /* it's in the second half */
606 i = k;
607 else /* we've got it :) */
608 return k;
610 return -1; /* we haven't got it :( */
613 static char *file_name = NULL;
614 static int32_t line_number = 0;
616 char *src_set_fname(char *newname)
618 char *oldname = file_name;
619 file_name = newname;
620 return oldname;
623 int32_t src_set_linnum(int32_t newline)
625 int32_t oldline = line_number;
626 line_number = newline;
627 return oldline;
630 int32_t src_get_linnum(void)
632 return line_number;
635 int src_get(int32_t *xline, char **xname)
637 if (!file_name || !*xname || strcmp(*xname, file_name)) {
638 nasm_free(*xname);
639 *xname = file_name ? nasm_strdup(file_name) : NULL;
640 *xline = line_number;
641 return -2;
643 if (*xline != line_number) {
644 int32_t tmp = line_number - *xline;
645 *xline = line_number;
646 return tmp;
648 return 0;
651 char *nasm_strcat(const char *one, const char *two)
653 char *rslt;
654 int l1 = strlen(one);
655 rslt = nasm_malloc(l1 + strlen(two) + 1);
656 strcpy(rslt, one);
657 strcpy(rslt + l1, two);
658 return rslt;
661 /* skip leading spaces */
662 char *nasm_skip_spaces(const char *p)
664 if (p)
665 while (*p && nasm_isspace(*p))
666 p++;
667 return (char *)p;
670 /* skip leading non-spaces */
671 char *nasm_skip_word(const char *p)
673 if (p)
674 while (*p && !nasm_isspace(*p))
675 p++;
676 return (char *)p;
679 /* zap leading spaces with zero */
680 char *nasm_zap_spaces_fwd(char *p)
682 if (p)
683 while (*p && nasm_isspace(*p))
684 *p++ = 0x0;
685 return p;
688 /* zap spaces with zero in reverse order */
689 char *nasm_zap_spaces_rev(char *p)
691 if (p)
692 while (*p && nasm_isspace(*p))
693 *p-- = 0x0;
694 return p;
697 /* zap leading and trailing spaces */
698 char *nasm_trim_spaces(char *p)
700 p = nasm_zap_spaces_fwd(p);
701 nasm_zap_spaces_fwd(nasm_skip_word(p));
703 return p;
707 * return the word extracted from a stream
708 * or NULL if nothing left
710 char *nasm_get_word(char *p, char **tail)
712 char *word = nasm_skip_spaces(p);
713 char *next = nasm_skip_word(word);
715 if (word && *word) {
716 if (*next)
717 *next++ = '\0';
718 } else
719 word = next = NULL;
721 /* NOTE: the tail may start with spaces */
722 *tail = next;
724 return word;
728 * Extract "opt=val" values from the stream and
729 * returns "opt"
731 * Exceptions:
732 * 1) If "=val" passed the NULL returned though
733 * you may continue handling the tail via "next"
734 * 2) If "=" passed the NULL is returned and "val"
735 * is set to NULL as well
737 char *nasm_opt_val(char *p, char **val, char **next)
739 char *q, *nxt;
741 *val = *next = NULL;
743 p = nasm_get_word(p, &nxt);
744 if (!p)
745 return NULL;
747 q = strchr(p, '=');
748 if (q) {
749 if (q == p)
750 p = NULL;
751 *q++='\0';
752 if (*q) {
753 *val = q;
754 } else {
755 q = nasm_get_word(q + 1, &nxt);
756 if (q)
757 *val = q;
759 } else {
760 q = nasm_skip_spaces(nxt);
761 if (q && *q == '=') {
762 q = nasm_get_word(q + 1, &nxt);
763 if (q)
764 *val = q;
768 *next = nxt;
769 return p;
773 * initialized data bytes length from opcode
775 int idata_bytes(int opcode)
777 int ret;
778 switch (opcode) {
779 case I_DB:
780 ret = 1;
781 break;
782 case I_DW:
783 ret = 2;
784 break;
785 case I_DD:
786 ret = 4;
787 break;
788 case I_DQ:
789 ret = 8;
790 break;
791 case I_DT:
792 ret = 10;
793 break;
794 case I_DO:
795 ret = 16;
796 break;
797 case I_DY:
798 ret = 32;
799 break;
800 case I_none:
801 ret = -1;
802 break;
803 default:
804 ret = 0;
805 break;
807 return ret;