summaryrefslogtreecommitdiff
path: root/parser.c
blob: 653f8a69f1606e7440fa0b3129137226286caf32 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
/* parser.c   source line parser for the Netwide Assembler
 *
 * The Netwide Assembler is copyright (C) 1996 Simon Tatham and
 * Julian Hall. All rights reserved. The software is
 * redistributable under the licence given in the file "Licence"
 * distributed in the NASM archive.
 *
 * initial version 27/iii/95 by Simon Tatham
 */

#include <stdio.h>
#include <stdlib.h>
#include <stddef.h>
#include <string.h>
#include <ctype.h>

#include "nasm.h"
#include "nasmlib.h"
#include "parser.h"
#include "float.h"

static long reg_flags[] = {	       /* sizes and special flags */
    0, REG8, REG_AL, REG_AX, REG8, REG8, REG16, REG16, REG8, REG_CL,
    REG_CREG, REG_CREG, REG_CREG, REG_CR4, REG_CS, REG_CX, REG8,
    REG16, REG8, REG_DREG, REG_DREG, REG_DREG, REG_DREG, REG_DREG,
    REG_DREG, REG_DESS, REG_DX, REG_EAX, REG32, REG32, REG_ECX,
    REG32, REG32, REG_DESS, REG32, REG32, REG_FSGS, REG_FSGS,
    MMXREG, MMXREG, MMXREG, MMXREG, MMXREG, MMXREG, MMXREG, MMXREG,
    REG16, REG16, REG_DESS, FPU0, FPUREG, FPUREG, FPUREG, FPUREG,
    FPUREG, FPUREG, FPUREG, REG_TREG, REG_TREG, REG_TREG, REG_TREG,
    REG_TREG,
    XMMREG, XMMREG, XMMREG, XMMREG, XMMREG, XMMREG, XMMREG, XMMREG
};

enum {				       /* special tokens */
    S_BYTE, S_DWORD, S_FAR, S_LONG, S_NEAR, S_NOSPLIT, S_QWORD,
    S_SHORT, S_TO, S_TWORD, S_WORD
};

static int is_comma_next (void);

static int i;
static struct tokenval tokval;
static efunc error;
static struct ofmt *outfmt;  /* Structure of addresses of output routines */
static loc_t *location;	     /* Pointer to current line's segment,offset */

void parser_global_info (struct ofmt *output, loc_t *locp) 
{
    outfmt = output;
    location = locp;
}

insn *parse_line (int pass, char *buffer, insn *result,
		  efunc errfunc, evalfunc evaluate, ldfunc ldef) 
{
    int operand;
    int critical;
    struct eval_hints hints;

    result->forw_ref = FALSE;
    error = errfunc;

    stdscan_reset();
    stdscan_bufptr = buffer;
    i = stdscan(NULL, &tokval);

    result->label = NULL;	       /* Assume no label */
    result->eops = NULL;	       /* must do this, whatever happens */
    result->operands = 0;	       /* must initialise this */

    if (i==0) {			       /* blank line - ignore */
	result->opcode = -1;	       /* and no instruction either */
	return result;
    }
    if (i != TOKEN_ID && i != TOKEN_INSN && i != TOKEN_PREFIX &&
	(i!=TOKEN_REG || (REG_SREG & ~reg_flags[tokval.t_integer]))) {
	error (ERR_NONFATAL, "label or instruction expected"
	       " at start of line");
	result->opcode = -1;
	return result;
    }

    if (i == TOKEN_ID) {	       /* there's a label here */
	result->label = tokval.t_charptr;
	i = stdscan(NULL, &tokval);
	if (i == ':') {		       /* skip over the optional colon */
	    i = stdscan(NULL, &tokval);
	} else if (i == 0) {
	    error (ERR_WARNING|ERR_WARN_OL|ERR_PASS1,
		   "label alone on a line without a colon might be in error");
	}
	if (i != TOKEN_INSN || tokval.t_integer != I_EQU)
	{
	    /*
	     * FIXME: location->segment could be NO_SEG, in which case
	     * it is possible we should be passing 'abs_seg'. Look into this.
	     * Work out whether that is *really* what we should be doing.
	     * Generally fix things. I think this is right as it is, but
	     * am still not certain.
	     */
	    ldef (result->label, location->segment,
		  location->offset, NULL, TRUE, FALSE, outfmt, errfunc);
	}
    }

    if (i==0) {
	result->opcode = -1;	       /* this line contains just a label */
	return result;
    }

    result->nprefix = 0;
    result->times = 1L;

    while (i == TOKEN_PREFIX ||
	   (i==TOKEN_REG && !(REG_SREG & ~reg_flags[tokval.t_integer]))) 
    {
	/*
	 * Handle special case: the TIMES prefix.
	 */
	if (i == TOKEN_PREFIX && tokval.t_integer == P_TIMES) {
	    expr *value;

	    i = stdscan(NULL, &tokval);
	    value = evaluate (stdscan, NULL, &tokval, NULL, pass0, error, NULL);
	    i = tokval.t_type;
	    if (!value) {	       /* but, error in evaluator */
		result->opcode = -1;   /* unrecoverable parse error: */
		return result;	       /* ignore this instruction */
	    }
	    if (!is_simple (value)) {
		error (ERR_NONFATAL,
		       "non-constant argument supplied to TIMES");
		result->times = 1L;
	    } else {
		result->times = value->value;
		if (value->value < 0) {
		    error(ERR_NONFATAL, "TIMES value %d is negative",
			  value->value);
		    result->times = 0;
		}
	    }
	} else {
	    if (result->nprefix == MAXPREFIX)
		error (ERR_NONFATAL,
		       "instruction has more than %d prefixes", MAXPREFIX);
	    else
		result->prefixes[result->nprefix++] = tokval.t_integer;
	    i = stdscan(NULL, &tokval);
	}
    }

    if (i != TOKEN_INSN) {
	if (result->nprefix > 0 && i == 0) {
	    /*
	     * Instruction prefixes are present, but no actual
	     * instruction. This is allowed: at this point we
	     * invent a notional instruction of RESB 0.
	     */
	    result->opcode = I_RESB;
	    result->operands = 1;
	    result->oprs[0].type = IMMEDIATE;
	    result->oprs[0].offset = 0L;
	    result->oprs[0].segment = result->oprs[0].wrt = NO_SEG;
	    return result;
	} else {
	    error (ERR_NONFATAL, "parser: instruction expected");
	    result->opcode = -1;
	    return result;
	}
    }

    result->opcode = tokval.t_integer;
    result->condition = tokval.t_inttwo;

    /*
     * RESB, RESW and RESD cannot be satisfied with incorrectly
     * evaluated operands, since the correct values _must_ be known
     * on the first pass. Hence, even in pass one, we set the
     * `critical' flag on calling evaluate(), so that it will bomb
     * out on undefined symbols. Nasty, but there's nothing we can
     * do about it.
     *
     * For the moment, EQU has the same difficulty, so we'll
     * include that.
     */
    if (result->opcode == I_RESB ||
	result->opcode == I_RESW ||
	result->opcode == I_RESD ||
	result->opcode == I_RESQ ||
	result->opcode == I_REST ||
	result->opcode == I_EQU ||
        result->opcode == I_INCBIN)  /* fbk */
    {
	critical = pass0;
    }
    else
	critical = (pass==2 ? 2 : 0);

    if (result->opcode == I_DB ||
	result->opcode == I_DW ||
	result->opcode == I_DD ||
	result->opcode == I_DQ ||
	result->opcode == I_DT ||
	result->opcode == I_INCBIN) 
    {
	extop *eop, **tail = &result->eops, **fixptr;
	int oper_num = 0;

	result->eops_float = FALSE;

	/*
	 * Begin to read the DB/DW/DD/DQ/DT/INCBIN operands.
	 */
	while (1) {
	    i = stdscan(NULL, &tokval);
	    if (i == 0)
		break;
	    fixptr = tail;
	    eop = *tail = nasm_malloc(sizeof(extop));
	    tail = &eop->next;
	    eop->next = NULL;
	    eop->type = EOT_NOTHING;
	    oper_num++;

	    if (i == TOKEN_NUM && tokval.t_charptr && is_comma_next()) {
		eop->type = EOT_DB_STRING;
		eop->stringval = tokval.t_charptr;
		eop->stringlen = tokval.t_inttwo;
		i = stdscan(NULL, &tokval);       /* eat the comma */
		continue;
	    }

	    if ((i == TOKEN_FLOAT && is_comma_next()) || i == '-') {
		long sign = +1L;

		if (i == '-') {
		    char *save = stdscan_bufptr;
		    i = stdscan(NULL, &tokval);
		    sign = -1L;
		    if (i != TOKEN_FLOAT || !is_comma_next()) {
			stdscan_bufptr = save;
			i = tokval.t_type = '-';
		    }
		}

		if (i == TOKEN_FLOAT) {
		    eop->type = EOT_DB_STRING;
		    result->eops_float = TRUE;
		    if (result->opcode == I_DD)
			eop->stringlen = 4;
		    else if (result->opcode == I_DQ)
			eop->stringlen = 8;
		    else if (result->opcode == I_DT)
			eop->stringlen = 10;
		    else {
			error(ERR_NONFATAL, "floating-point constant"
			      " encountered in `D%c' instruction",
			      result->opcode == I_DW ? 'W' : 'B');
			/*
			 * fix suggested by Pedro Gimeno... original line
			 * was:
			 * eop->type = EOT_NOTHING;
			 */
			eop->stringlen = 0;
		    }
		    eop = nasm_realloc(eop, sizeof(extop)+eop->stringlen);
		    tail = &eop->next;
		    *fixptr = eop;
		    eop->stringval = (char *)eop + sizeof(extop);
		    if (eop->stringlen < 4 ||
			!float_const (tokval.t_charptr, sign,
				      (unsigned char *)eop->stringval,
				      eop->stringlen, error))
			eop->type = EOT_NOTHING;
		    i = stdscan(NULL, &tokval);       /* eat the comma */
		    continue;
		}
	    }

	    /* anything else */ 
	    {
		expr *value;
		value = evaluate (stdscan, NULL, &tokval, NULL,
				  critical, error, NULL);
		i = tokval.t_type;
		if (!value) {	       /* error in evaluator */
		    result->opcode = -1;/* unrecoverable parse error: */
		    return result;     /* ignore this instruction */
		}
		if (is_unknown(value)) {
		    eop->type = EOT_DB_NUMBER;
		    eop->offset = 0;   /* doesn't matter what we put */
		    eop->segment = eop->wrt = NO_SEG;   /* likewise */
		} else if (is_reloc(value)) {
		    eop->type = EOT_DB_NUMBER;
		    eop->offset = reloc_value(value);
		    eop->segment = reloc_seg(value);
		    eop->wrt = reloc_wrt(value);
		} else {
		    error (ERR_NONFATAL,
			   "operand %d: expression is not simple"
			   " or relocatable", oper_num);
		}
	    }

	    /*
	     * We're about to call stdscan(), which will eat the
	     * comma that we're currently sitting on between
	     * arguments. However, we'd better check first that it
	     * _is_ a comma.
	     */
	    if (i == 0)		       /* also could be EOL */
		break;
	    if (i != ',') {
		error (ERR_NONFATAL, "comma expected after operand %d",
		       oper_num);
		result->opcode = -1;/* unrecoverable parse error: */
		return result;     /* ignore this instruction */
	    }
	}

	if (result->opcode == I_INCBIN) {
	    /*
	     * Correct syntax for INCBIN is that there should be
	     * one string operand, followed by one or two numeric
	     * operands.
	     */
	    if (!result->eops || result->eops->type != EOT_DB_STRING)
		error (ERR_NONFATAL, "`incbin' expects a file name");
	    else if (result->eops->next &&
		     result->eops->next->type != EOT_DB_NUMBER)
		error (ERR_NONFATAL, "`incbin': second parameter is",
		       " non-numeric");
	    else if (result->eops->next && result->eops->next->next &&
		     result->eops->next->next->type != EOT_DB_NUMBER)
		error (ERR_NONFATAL, "`incbin': third parameter is",
		       " non-numeric");
	    else if (result->eops->next && result->eops->next->next &&
		     result->eops->next->next->next)
		error (ERR_NONFATAL, "`incbin': more than three parameters");
	    else
		return result;
	    /*
	     * If we reach here, one of the above errors happened.
	     * Throw the instruction away.
	     */
	    result->opcode = -1;
	    return result;
	} else /* DB ... */
	    if (oper_num == 0)
		error (ERR_WARNING|ERR_PASS1,
		       "no operand for data declaration");
            else
                result->operands = oper_num;

	return result;
    }

    /* right. Now we begin to parse the operands. There may be up to three
     * of these, separated by commas, and terminated by a zero token. */

    for (operand = 0; operand < 3; operand++) {
	expr *value;		       /* used most of the time */
	int mref;		       /* is this going to be a memory ref? */
	int bracket;		       /* is it a [] mref, or a & mref? */
	int setsize = 0;

	result->oprs[operand].addr_size = 0;/* have to zero this whatever */
	result->oprs[operand].eaflags = 0;   /* and this */
	result->oprs[operand].opflags = 0;

	i = stdscan(NULL, &tokval);
	if (i == 0) break;	       /* end of operands: get out of here */
	result->oprs[operand].type = 0;   /* so far, no override */
	while (i == TOKEN_SPECIAL)	{/* size specifiers */
	    switch ((int)tokval.t_integer) {
	      case S_BYTE:
		if (!setsize)		 /* we want to use only the first */
		    result->oprs[operand].type |= BITS8;
		setsize = 1;
		break;
	      case S_WORD:
		if (!setsize)
		    result->oprs[operand].type |= BITS16;
		setsize = 1;
		break;
	      case S_DWORD:
	      case S_LONG:
		if (!setsize)
		    result->oprs[operand].type |= BITS32;
		setsize = 1;
		break;
	      case S_QWORD:
		if (!setsize)
		    result->oprs[operand].type |= BITS64;
		setsize = 1;
		break;
	      case S_TWORD:
		if (!setsize)
		    result->oprs[operand].type |= BITS80;
		setsize = 1;
		break;
	      case S_TO:
		result->oprs[operand].type |= TO;
		break;
	      case S_FAR:
		result->oprs[operand].type |= FAR;
		break;
	      case S_NEAR:
		result->oprs[operand].type |= NEAR;
		break;
	      case S_SHORT:
		result->oprs[operand].type |= SHORT;
		break;
	      default:
		error (ERR_NONFATAL, "invalid operand size specification");
	    }
	    i = stdscan(NULL, &tokval);
	}

	if (i == '[' || i == '&') {    /* memory reference */
	    mref = TRUE;
	    bracket = (i == '[');
	    i = stdscan(NULL, &tokval);
	    if (i == TOKEN_SPECIAL) {  /* check for address size override */
		if (tasm_compatible_mode) {
		  switch ((int)tokval.t_integer) {
		    /* For TASM compatibility a size override inside the
		     * brackets changes the size of the operand, not the
		     * address type of the operand as it does in standard
		     * NASM syntax. Hence:
		     *
		     *	mov	eax,[DWORD val]
		     *
		     * is valid syntax in TASM compatibility mode. Note that
		     * you lose the ability to override the default address
		     * type for the instruction, but we never use anything
		     * but 32-bit flat model addressing in our code.
		     */
		    case S_BYTE:
		      result->oprs[operand].type |= BITS8;
		      break;
		    case S_WORD:
		      result->oprs[operand].type |= BITS16;
		      break;
		    case S_DWORD:
		    case S_LONG:
		      result->oprs[operand].type |= BITS32;
		      break;
		    case S_QWORD:
		      result->oprs[operand].type |= BITS64;
		      break;
		    case S_TWORD:
		      result->oprs[operand].type |= BITS80;
		      break;
		    default:
		      error (ERR_NONFATAL, "invalid operand size specification");
		  }
		} else {
		  /* Standard NASM compatible syntax */
		  switch ((int)tokval.t_integer) {
		    case S_NOSPLIT:
		      result->oprs[operand].eaflags |= EAF_TIMESTWO;
		      break;
		    case S_BYTE:
		      result->oprs[operand].eaflags |= EAF_BYTEOFFS;
		      break;
		    case S_WORD:
		      result->oprs[operand].addr_size = 16;
		      result->oprs[operand].eaflags |= EAF_WORDOFFS;
		      break;
		    case S_DWORD:
		    case S_LONG:
		      result->oprs[operand].addr_size = 32;
		      result->oprs[operand].eaflags |= EAF_WORDOFFS;
		      break;
		    default:
		      error (ERR_NONFATAL, "invalid size specification in"
			     " effective address");
		  }
		}
		i = stdscan(NULL, &tokval);
	    }
	} else {		       /* immediate operand, or register */
	    mref = FALSE;
	    bracket = FALSE;	       /* placate optimisers */
	}

	value = evaluate (stdscan, NULL, &tokval,
			  &result->oprs[operand].opflags,
			  critical, error, &hints);
	i = tokval.t_type;
	if (result->oprs[operand].opflags & OPFLAG_FORWARD) {
	    result->forw_ref = TRUE;
	}
	if (!value) {		       /* error in evaluator */
	    result->opcode = -1;       /* unrecoverable parse error: */
	    return result;	       /* ignore this instruction */
	}
	if (i == ':' && mref) {	       /* it was seg:offset */
	    /*
	     * Process the segment override.
	     */
	    if (value[1].type!=0 || value->value!=1 ||
		REG_SREG & ~reg_flags[value->type])
		error (ERR_NONFATAL, "invalid segment override");
	    else if (result->nprefix == MAXPREFIX)
		error (ERR_NONFATAL,
		       "instruction has more than %d prefixes",
		       MAXPREFIX);
	    else
		result->prefixes[result->nprefix++] = value->type;

	    i = stdscan(NULL, &tokval);	       /* then skip the colon */
	    if (i == TOKEN_SPECIAL) {  /* another check for size override */
		switch ((int)tokval.t_integer) {
		  case S_WORD:
		    result->oprs[operand].addr_size = 16;
		    break;
		  case S_DWORD:
		  case S_LONG:
		    result->oprs[operand].addr_size = 32;
		    break;
		  default:
		    error (ERR_NONFATAL, "invalid size specification in"
			   " effective address");
		}
		i = stdscan(NULL, &tokval);
	    }
	    value = evaluate (stdscan, NULL, &tokval,
			      &result->oprs[operand].opflags,
			      critical, error, &hints);
	    i = tokval.t_type;
	    if (result->oprs[operand].opflags & OPFLAG_FORWARD) {
		result->forw_ref = TRUE;
	    }
	    /* and get the offset */
	    if (!value) {	       /* but, error in evaluator */
		result->opcode = -1;   /* unrecoverable parse error: */
		return result;	       /* ignore this instruction */
	    }
	}
	if (mref && bracket) {	       /* find ] at the end */
	    if (i != ']') {
		error (ERR_NONFATAL, "parser: expecting ]");
		do {		       /* error recovery again */
		    i = stdscan(NULL, &tokval);
		} while (i != 0 && i != ',');
	    } else		       /* we got the required ] */
		i = stdscan(NULL, &tokval);
	} else {		       /* immediate operand */
	    if (i != 0 && i != ',' && i != ':') {
		error (ERR_NONFATAL, "comma or end of line expected");
		do {		       /* error recovery */
		    i = stdscan(NULL, &tokval);
		} while (i != 0 && i != ',');
	    } else if (i == ':') {
		result->oprs[operand].type |= COLON;
	    }
	}

	/* now convert the exprs returned from evaluate() into operand
	 * descriptions... */

	if (mref) {		       /* it's a memory reference */
	    expr *e = value;
	    int b, i, s;	       /* basereg, indexreg, scale */
	    long o;		       /* offset */

	    b = i = -1, o = s = 0;
	    result->oprs[operand].hintbase = hints.base;
	    result->oprs[operand].hinttype = hints.type;

	    if (e->type <= EXPR_REG_END) {   /* this bit's a register */
		if (e->value == 1) /* in fact it can be basereg */
		    b = e->type;
		else	       /* no, it has to be indexreg */
		    i = e->type, s = e->value;
		e++;
	    }
	    if (e->type && e->type <= EXPR_REG_END)   /* it's a 2nd register */
	    {
		if (b != -1)               /* If the first was the base, ... */
		    i = e->type, s = e->value;  /* second has to be indexreg */

		else if (e->value != 1)          /* If both want to be index */
		{
		    error(ERR_NONFATAL, "beroset-p-592-invalid effective address");
		    result->opcode = -1;
		    return result;
		} 
		else
		    b = e->type;
		e++;
	    }
	    if (e->type != 0) {	       /* is there an offset? */
		if (e->type <= EXPR_REG_END)  /* in fact, is there an error? */
		{
		    error (ERR_NONFATAL, "beroset-p-603-invalid effective address");
		    result->opcode = -1;
		    return result;
		} 
		else 
		{
		    if (e->type == EXPR_UNKNOWN) {
			o = 0;	                     /* doesn't matter what */
			result->oprs[operand].wrt = NO_SEG;     /* nor this */
			result->oprs[operand].segment = NO_SEG;  /* or this */
			while (e->type) e++;   /* go to the end of the line */
		    } 
		    else 
		    {
			if (e->type == EXPR_SIMPLE) {
			    o = e->value;
			    e++;
			}
			if (e->type == EXPR_WRT) {
			    result->oprs[operand].wrt = e->value;
			    e++;
			} else
			    result->oprs[operand].wrt = NO_SEG;
			/*
			 * Look for a segment base type.
			 */
			if (e->type && e->type < EXPR_SEGBASE) {
			    error (ERR_NONFATAL, "beroset-p-630-invalid effective address");
			    result->opcode = -1;
			    return result;
			}
			while (e->type && e->value == 0)
			    e++;
			if (e->type && e->value != 1) {
			    error (ERR_NONFATAL, "beroset-p-637-invalid effective address");
			    result->opcode = -1;
			    return result;
			}
			if (e->type) {
			    result->oprs[operand].segment =
				e->type - EXPR_SEGBASE;
			    e++;
			} else
			    result->oprs[operand].segment = NO_SEG;
			while (e->type && e->value == 0)
			    e++;
			if (e->type) {
			    error (ERR_NONFATAL, "beroset-p-650-invalid effective address");
			    result->opcode = -1;
			    return result;
			}
		    }
		}
	    } else {
		o = 0;
		result->oprs[operand].wrt = NO_SEG;
		result->oprs[operand].segment = NO_SEG;
	    }

	    if (e->type != 0) {    /* there'd better be nothing left! */
		error (ERR_NONFATAL, "beroset-p-663-invalid effective address");
		result->opcode = -1;
		return result;
	    }

	    result->oprs[operand].type |= MEMORY;
	    if (b==-1 && (i==-1 || s==0))
		result->oprs[operand].type |= MEM_OFFS;
	    result->oprs[operand].basereg = b;
	    result->oprs[operand].indexreg = i;
	    result->oprs[operand].scale = s;
	    result->oprs[operand].offset = o;
	} 
	else		                      /* it's not a memory reference */
	{
	    if (is_just_unknown(value)) {     /* it's immediate but unknown */
		result->oprs[operand].type |= IMMEDIATE;
		result->oprs[operand].offset = 0;   /* don't care */
		result->oprs[operand].segment = NO_SEG; /* don't care again */
		result->oprs[operand].wrt = NO_SEG;/* still don't care */
	    } 
	    else if (is_reloc(value))         /* it's immediate */
	    {
		result->oprs[operand].type |= IMMEDIATE;
		result->oprs[operand].offset = reloc_value(value);
		result->oprs[operand].segment = reloc_seg(value);
		result->oprs[operand].wrt = reloc_wrt(value);
		if (is_simple(value)) {
		    if (reloc_value(value)==1)
			result->oprs[operand].type |= UNITY;
		    if (optimizing>=0) {
  		        if (reloc_value(value) >= -128 &&
		                 reloc_value(value) <= 127)
		            result->oprs[operand].type |= SBYTE;
		    }
		}
	    } 
	    else	       /* it's a register */
	    {
		int i;

		if (value->type>=EXPR_SIMPLE || value->value!=1) {
		    error (ERR_NONFATAL, "invalid operand type");
		    result->opcode = -1;
		    return result;
		}

		/*
		 * check that its only 1 register, not an expression...
		 */
		for (i = 1; value[i].type; i++)
		    if (value[i].value) {
			error (ERR_NONFATAL, "invalid operand type");
			result->opcode = -1;
			return result;
		    }

		/* clear overrides, except TO which applies to FPU regs */
		if (result->oprs[operand].type & ~TO) {
		    /*
		     * we want to produce a warning iff the specified size
		     * is different from the register size
		     */
		    i = result->oprs[operand].type & SIZE_MASK;
		}
		else
		    i = 0;

		result->oprs[operand].type &= TO;
		result->oprs[operand].type |= REGISTER;
		result->oprs[operand].type |= reg_flags[value->type];
		result->oprs[operand].basereg = value->type;

		if (i && (result->oprs[operand].type & SIZE_MASK) != i)
		    error (ERR_WARNING|ERR_PASS1,
			   "register size specification ignored");
	    }
	}
    }

    result->operands = operand;       /* set operand count */

    while (operand<3)		       /* clear remaining operands */
	result->oprs[operand++].type = 0;

    /*
     * Transform RESW, RESD, RESQ, REST into RESB.
     */
    switch (result->opcode) {
      case I_RESW: result->opcode=I_RESB; result->oprs[0].offset*=2; break;
      case I_RESD: result->opcode=I_RESB; result->oprs[0].offset*=4; break;
      case I_RESQ: result->opcode=I_RESB; result->oprs[0].offset*=8; break;
      case I_REST: result->opcode=I_RESB; result->oprs[0].offset*=10; break;
    }

    return result;
}

static int is_comma_next (void) 
{
    char *p;
    int i;
    struct tokenval tv;

    p = stdscan_bufptr;
    i = stdscan (NULL, &tv);
    stdscan_bufptr = p;
    return (i == ',' || i == ';' || !i);
}

void cleanup_insn (insn *i) 
{
    extop *e;

    while (i->eops) {
	e = i->eops;
	i->eops = i->eops->next;
	nasm_free (e);
    }
}