nasmlib.c: fix issues with strrep utility function
[nasm/avx512.git] / nasmlib.c
blobc08ef09d8933aca2c9bbbfc94eaf2e9c1607f8f0
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
293 char *nasm_strrep(const char *str, const char *sub, char *lin, bool casesense)
295 char *outline = lin;
296 char *temp1 = NULL;
297 char *temp2 = NULL;
298 char *l, *ll, *lp, *lt, *ls;
299 int count = 0;
300 int str_len, sub_len, lin_len;
301 int i, c;
303 str_len = strlen(str);
304 sub_len = strlen(sub);
305 lin_len = strlen(lin);
307 if ((str_len > 0) && (lin_len > 0)) {
308 if (casesense == false) {
309 l = nasm_strdup(lin);
310 for (i = 0; i < lin_len; i++) {
311 l[i] = (char)nasm_tolower_tab[(int)l[i]];
313 ls = nasm_strdup(str);
314 for (i = 0; i < str_len; i++) {
315 ls[i] = (char)nasm_tolower_tab[(int)ls[i]];
317 temp1 = l;
318 temp2 = ls;
319 } else {
320 l = lin;
321 ls = (char *)str;
324 ll = l;
326 do {
327 l = strstr(l, ls);
328 if (l != NULL) {
329 count ++;
330 l += str_len;
332 } while (l != NULL);
334 if (count > 0) {
335 i = (lin_len - (count * str_len));
336 i += (count * sub_len);
337 outline = nasm_zalloc(i);
339 l = ll;
340 lt = lin;
342 for (i = 0; i < count; i++) {
343 lp = l;
344 l = strstr(l, ls);
345 c = (l - lp);
346 if (c > 0) {
347 strncat(outline, lt, c);
348 lt += c;
350 strncat(outline, sub, sub_len);
351 l += str_len;
352 lt += str_len;
355 c = (l - ll);
356 if (c < lin_len) {
357 strncat(outline, lt, (lin_len-c));
360 if (temp2 != NULL) {
361 nasm_free(temp2);
364 if (temp1 != NULL) {
365 nasm_free(temp1);
368 nasm_free(lin);
372 return outline;
376 #define lib_isnumchar(c) (nasm_isalnum(c) || (c) == '$' || (c) == '_')
377 #define numvalue(c) ((c)>='a' ? (c)-'a'+10 : (c)>='A' ? (c)-'A'+10 : (c)-'0')
379 static int radix_letter(char c)
381 switch (c) {
382 case 'b': case 'B':
383 case 'y': case 'Y':
384 return 2; /* Binary */
385 case 'o': case 'O':
386 case 'q': case 'Q':
387 return 8; /* Octal */
388 case 'h': case 'H':
389 case 'x': case 'X':
390 return 16; /* Hexadecimal */
391 case 'd': case 'D':
392 case 't': case 'T':
393 return 10; /* Decimal */
394 default:
395 return 0; /* Not a known radix letter */
399 int64_t readnum(char *str, bool *error)
401 char *r = str, *q;
402 int32_t pradix, sradix, radix;
403 int plen, slen, len;
404 uint64_t result, checklimit;
405 int digit, last;
406 bool warn = false;
407 int sign = 1;
409 *error = false;
411 while (nasm_isspace(*r))
412 r++; /* find start of number */
415 * If the number came from make_tok_num (as a result of an %assign), it
416 * might have a '-' built into it (rather than in a preceeding token).
418 if (*r == '-') {
419 r++;
420 sign = -1;
423 q = r;
425 while (lib_isnumchar(*q))
426 q++; /* find end of number */
428 len = q-r;
429 if (!len) {
430 /* Not numeric */
431 *error = true;
432 return 0;
436 * Handle radix formats:
438 * 0<radix-letter><string>
439 * $<string> (hexadecimal)
440 * <string><radix-letter>
442 pradix = sradix = 0;
443 plen = slen = 0;
445 if (len > 2 && *r == '0' && (pradix = radix_letter(r[1])) != 0)
446 plen = 2;
447 else if (len > 1 && *r == '$')
448 pradix = 16, plen = 1;
450 if (len > 1 && (sradix = radix_letter(q[-1])) != 0)
451 slen = 1;
453 if (pradix > sradix) {
454 radix = pradix;
455 r += plen;
456 } else if (sradix > pradix) {
457 radix = sradix;
458 q -= slen;
459 } else {
460 /* Either decimal, or invalid -- if invalid, we'll trip up
461 further down. */
462 radix = 10;
466 * `checklimit' must be 2**64 / radix. We can't do that in
467 * 64-bit arithmetic, which we're (probably) using, so we
468 * cheat: since we know that all radices we use are even, we
469 * can divide 2**63 by radix/2 instead.
471 checklimit = UINT64_C(0x8000000000000000) / (radix >> 1);
474 * Calculate the highest allowable value for the last digit of a
475 * 64-bit constant... in radix 10, it is 6, otherwise it is 0
477 last = (radix == 10 ? 6 : 0);
479 result = 0;
480 while (*r && r < q) {
481 if (*r != '_') {
482 if (*r < '0' || (*r > '9' && *r < 'A')
483 || (digit = numvalue(*r)) >= radix) {
484 *error = true;
485 return 0;
487 if (result > checklimit ||
488 (result == checklimit && digit >= last)) {
489 warn = true;
492 result = radix * result + digit;
494 r++;
497 if (warn)
498 nasm_error(ERR_WARNING | ERR_PASS1 | ERR_WARN_NOV,
499 "numeric constant %s does not fit in 64 bits",
500 str);
502 return result * sign;
505 int64_t readstrnum(char *str, int length, bool *warn)
507 int64_t charconst = 0;
508 int i;
510 *warn = false;
512 str += length;
513 if (globalbits == 64) {
514 for (i = 0; i < length; i++) {
515 if (charconst & UINT64_C(0xFF00000000000000))
516 *warn = true;
517 charconst = (charconst << 8) + (uint8_t)*--str;
519 } else {
520 for (i = 0; i < length; i++) {
521 if (charconst & 0xFF000000UL)
522 *warn = true;
523 charconst = (charconst << 8) + (uint8_t)*--str;
526 return charconst;
529 static int32_t next_seg;
531 void seg_init(void)
533 next_seg = 0;
536 int32_t seg_alloc(void)
538 return (next_seg += 2) - 2;
541 #ifdef WORDS_LITTLEENDIAN
543 void fwriteint16_t(uint16_t data, FILE * fp)
545 fwrite(&data, 1, 2, fp);
548 void fwriteint32_t(uint32_t data, FILE * fp)
550 fwrite(&data, 1, 4, fp);
553 void fwriteint64_t(uint64_t data, FILE * fp)
555 fwrite(&data, 1, 8, fp);
558 void fwriteaddr(uint64_t data, int size, FILE * fp)
560 fwrite(&data, 1, size, fp);
563 #else /* not WORDS_LITTLEENDIAN */
565 void fwriteint16_t(uint16_t data, FILE * fp)
567 char buffer[2], *p = buffer;
568 WRITESHORT(p, data);
569 fwrite(buffer, 1, 2, fp);
572 void fwriteint32_t(uint32_t data, FILE * fp)
574 char buffer[4], *p = buffer;
575 WRITELONG(p, data);
576 fwrite(buffer, 1, 4, fp);
579 void fwriteint64_t(uint64_t data, FILE * fp)
581 char buffer[8], *p = buffer;
582 WRITEDLONG(p, data);
583 fwrite(buffer, 1, 8, fp);
586 void fwriteaddr(uint64_t data, int size, FILE * fp)
588 char buffer[8], *p = buffer;
589 WRITEADDR(p, data, size);
590 fwrite(buffer, 1, size, fp);
593 #endif
595 size_t fwritezero(size_t bytes, FILE *fp)
597 size_t count = 0;
598 size_t blksize;
599 size_t rv;
601 while (bytes) {
602 blksize = (bytes < ZERO_BUF_SIZE) ? bytes : ZERO_BUF_SIZE;
604 rv = fwrite(zero_buffer, 1, blksize, fp);
605 if (!rv)
606 break;
608 count += rv;
609 bytes -= rv;
612 return count;
615 void standard_extension(char *inname, char *outname, char *extension)
617 char *p, *q;
619 if (*outname) /* file name already exists, */
620 return; /* so do nothing */
621 q = inname;
622 p = outname;
623 while (*q)
624 *p++ = *q++; /* copy, and find end of string */
625 *p = '\0'; /* terminate it */
626 while (p > outname && *--p != '.') ; /* find final period (or whatever) */
627 if (*p != '.')
628 while (*p)
629 p++; /* go back to end if none found */
630 if (!strcmp(p, extension)) { /* is the extension already there? */
631 if (*extension)
632 nasm_error(ERR_WARNING | ERR_NOFILE,
633 "file name already ends in `%s': "
634 "output will be in `nasm.out'", extension);
635 else
636 nasm_error(ERR_WARNING | ERR_NOFILE,
637 "file name already has no extension: "
638 "output will be in `nasm.out'");
639 strcpy(outname, "nasm.out");
640 } else
641 strcpy(p, extension);
645 * Common list of prefix names
647 static const char *prefix_names[] = {
648 "a16", "a32", "a64", "asp", "lock", "o16", "o32", "o64", "osp",
649 "rep", "repe", "repne", "repnz", "repz", "times", "wait"
652 const char *prefix_name(int token)
654 unsigned int prefix = token-PREFIX_ENUM_START;
655 if (prefix >= ARRAY_SIZE(prefix_names))
656 return NULL;
658 return prefix_names[prefix];
662 * Binary search.
664 int bsi(const char *string, const char **array, int size)
666 int i = -1, j = size; /* always, i < index < j */
667 while (j - i >= 2) {
668 int k = (i + j) / 2;
669 int l = strcmp(string, array[k]);
670 if (l < 0) /* it's in the first half */
671 j = k;
672 else if (l > 0) /* it's in the second half */
673 i = k;
674 else /* we've got it :) */
675 return k;
677 return -1; /* we haven't got it :( */
680 int bsii(const char *string, const char **array, int size)
682 int i = -1, j = size; /* always, i < index < j */
683 while (j - i >= 2) {
684 int k = (i + j) / 2;
685 int l = nasm_stricmp(string, array[k]);
686 if (l < 0) /* it's in the first half */
687 j = k;
688 else if (l > 0) /* it's in the second half */
689 i = k;
690 else /* we've got it :) */
691 return k;
693 return -1; /* we haven't got it :( */
696 static char *file_name = NULL;
697 static int32_t line_number = 0;
699 char *src_set_fname(char *newname)
701 char *oldname = file_name;
702 file_name = newname;
703 return oldname;
706 int32_t src_set_linnum(int32_t newline)
708 int32_t oldline = line_number;
709 line_number = newline;
710 return oldline;
713 int32_t src_get_linnum(void)
715 return line_number;
718 int src_get(int32_t *xline, char **xname)
720 if (!file_name || !*xname || strcmp(*xname, file_name)) {
721 nasm_free(*xname);
722 *xname = file_name ? nasm_strdup(file_name) : NULL;
723 *xline = line_number;
724 return -2;
726 if (*xline != line_number) {
727 int32_t tmp = line_number - *xline;
728 *xline = line_number;
729 return tmp;
731 return 0;
734 char *nasm_strcat(const char *one, const char *two)
736 char *rslt;
737 int l1 = strlen(one);
738 rslt = nasm_malloc(l1 + strlen(two) + 1);
739 strcpy(rslt, one);
740 strcpy(rslt + l1, two);
741 return rslt;
744 /* skip leading spaces */
745 char *nasm_skip_spaces(const char *p)
747 if (p)
748 while (*p && nasm_isspace(*p))
749 p++;
750 return (char *)p;
753 /* skip leading non-spaces */
754 char *nasm_skip_word(const char *p)
756 if (p)
757 while (*p && !nasm_isspace(*p))
758 p++;
759 return (char *)p;
762 /* zap leading spaces with zero */
763 char *nasm_zap_spaces_fwd(char *p)
765 if (p)
766 while (*p && nasm_isspace(*p))
767 *p++ = 0x0;
768 return p;
771 /* zap spaces with zero in reverse order */
772 char *nasm_zap_spaces_rev(char *p)
774 if (p)
775 while (*p && nasm_isspace(*p))
776 *p-- = 0x0;
777 return p;
780 /* zap leading and trailing spaces */
781 char *nasm_trim_spaces(char *p)
783 p = nasm_zap_spaces_fwd(p);
784 nasm_zap_spaces_fwd(nasm_skip_word(p));
786 return p;
790 * return the word extracted from a stream
791 * or NULL if nothing left
793 char *nasm_get_word(char *p, char **tail)
795 char *word = nasm_skip_spaces(p);
796 char *next = nasm_skip_word(word);
798 if (word && *word) {
799 if (*next)
800 *next++ = '\0';
801 } else
802 word = next = NULL;
804 /* NOTE: the tail may start with spaces */
805 *tail = next;
807 return word;
811 * Extract "opt=val" values from the stream and
812 * returns "opt"
814 * Exceptions:
815 * 1) If "=val" passed the NULL returned though
816 * you may continue handling the tail via "next"
817 * 2) If "=" passed the NULL is returned and "val"
818 * is set to NULL as well
820 char *nasm_opt_val(char *p, char **val, char **next)
822 char *q, *opt, *nxt;
824 opt = *val = *next = NULL;
826 p = nasm_get_word(p, &nxt);
827 if (!p)
828 return NULL;
830 q = strchr(p, '=');
831 if (q) {
832 if (q == p)
833 p = NULL;
834 *q++='\0';
835 if (*q) {
836 *val = q;
837 } else {
838 q = nasm_get_word(q + 1, &nxt);
839 if (q)
840 *val = q;
842 } else {
843 q = nasm_skip_spaces(nxt);
844 if (q && *q == '=') {
845 q = nasm_get_word(q + 1, &nxt);
846 if (q)
847 *val = q;
851 *next = nxt;
852 return p;
856 * initialized data bytes length from opcode
858 int idata_bytes(int opcode)
860 int ret;
861 switch (opcode) {
862 case I_DB:
863 ret = 1;
864 break;
865 case I_DW:
866 ret = 2;
867 break;
868 case I_DD:
869 ret = 4;
870 break;
871 case I_DQ:
872 ret = 8;
873 break;
874 case I_DT:
875 ret = 10;
876 break;
877 case I_DO:
878 ret = 16;
879 break;
880 case I_DY:
881 ret = 32;
882 break;
883 case I_none:
884 ret = -1;
885 break;
886 default:
887 ret = 0;
888 break;
890 return ret;