aboutsummaryrefslogtreecommitdiff
path: root/ld/ldlex.l
blob: 95a31f5a64a120adac9baf512dbbc52677a7bd5f (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
%{

/* Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001
   Free Software Foundation, Inc.

This file is part of GLD, the Gnu Linker.

GLD is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.

GLD is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with GLD; see the file COPYING.  If not, write to the Free
Software Foundation, 59 Temple Place - Suite 330, Boston, MA
02111-1307, USA.  */

/*
This was written by steve chamberlain
                    sac@cygnus.com
*/


#include <ansidecl.h>
#include <stdio.h>

#ifdef MPW
/* Prevent enum redefinition problems. */
#define TRUE_FALSE_ALREADY_DEFINED
#endif /* MPW */

#include "bfd.h"
#include "sysdep.h"
#include "safe-ctype.h"
#include "ld.h"
#include "ldgram.h"
#include "ldmisc.h"
#include "ldexp.h"
#include "ldlang.h"
#include "ldfile.h"
#include "ldlex.h"
#include "ldmain.h"
#include "libiberty.h"

/* The type of top-level parser input.
   yylex and yyparse (indirectly) both check this.  */
input_type parser_input;

/* Line number in the current input file.
   (FIXME Actually, it doesn't appear to get reset for each file?)  */
unsigned int lineno = 1;

/* The string we are currently lexing, or NULL if we are reading a
   file.  */
const char *lex_string = NULL;

/* Support for flex reading from more than one input file (stream).
   `include_stack' is flex's input state for each open file;
   `file_name_stack' is the file names.  `lineno_stack' is the current
   line numbers.

   If `include_stack_ptr' is 0, we haven't started reading anything yet.
   Otherwise, stack elements 0 through `include_stack_ptr - 1' are valid.  */

#undef YY_INPUT
#define YY_INPUT(buf,result,max_size) yy_input(buf, &result, max_size)

#define MAX_INCLUDE_DEPTH 10
static YY_BUFFER_STATE include_stack[MAX_INCLUDE_DEPTH];
static const char *file_name_stack[MAX_INCLUDE_DEPTH];
static unsigned int lineno_stack[MAX_INCLUDE_DEPTH];
static unsigned int include_stack_ptr = 0;
static int vers_node_nesting = 0;

static YY_BUFFER_STATE yy_create_string_buffer PARAMS ((const char *string,
							size_t size));
static void yy_input PARAMS ((char *, int *result, int max_size));

static void comment PARAMS ((void));
static void lex_warn_invalid PARAMS ((char *where, char *what));

/* STATES 
	EXPRESSION	definitely in an expression
	SCRIPT		definitely in a script
	BOTH		either EXPRESSION or SCRIPT
	DEFSYMEXP	in an argument to -defsym
        MRI             in an MRI script
	VERS_START	starting a Sun style mapfile
	VERS_SCRIPT	a Sun style mapfile
	VERS_NODE	a node within a Sun style mapfile
*/
#define RTOKEN(x)  {  yylval.token = x; return x; }

/* Some versions of flex want this.  */
#ifndef yywrap
int yywrap () { return 1; }
#endif
%}

%a 4000
%o 5000

CMDFILENAMECHAR   [_a-zA-Z0-9\/\.\\_\+\$\:\[\]\\\,\=\&\!\<\>\-\~]
CMDFILENAMECHAR1  [_a-zA-Z0-9\/\.\\_\+\$\:\[\]\\\,\=\&\!\<\>\~]
FILENAMECHAR1	[_a-zA-Z\/\.\\\$\_\~]
SYMBOLCHARN     [_a-zA-Z\/\.\\\$\_\~0-9]
FILENAMECHAR	[_a-zA-Z0-9\/\.\-\_\+\=\$\:\[\]\\\,\~]
WILDCHAR	[_a-zA-Z0-9\/\.\-\_\+\=\$\:\[\]\\\,\~\?\*]
WHITE		[ \t\n\r]+ 

NOCFILENAMECHAR	[_a-zA-Z0-9\/\.\-\_\+\$\:\[\]\\\~]

V_TAG [.$_a-zA-Z][._a-zA-Z0-9]*
V_IDENTIFIER [*?.$_a-zA-Z]([*?.$_a-zA-Z0-9]|::)*

%s SCRIPT
%s EXPRESSION
%s BOTH
%s DEFSYMEXP
%s MRI
%s VERS_START
%s VERS_SCRIPT
%s VERS_NODE
%%

  if (parser_input != input_selected)
    {
      /* The first token of the input determines the initial parser state.  */
      input_type t = parser_input;
      parser_input = input_selected;
      switch (t)
	{
	case input_script: return INPUT_SCRIPT; break;
	case input_mri_script: return INPUT_MRI_SCRIPT; break;
	case input_version_script: return INPUT_VERSION_SCRIPT; break;
	case input_defsym: return INPUT_DEFSYM; break;
	default: abort ();
	}
    }

<BOTH,SCRIPT,EXPRESSION>"/*"	{ comment(); }


<DEFSYMEXP>"-"                  { RTOKEN('-');}
<DEFSYMEXP>"+"                  { RTOKEN('+');}
<DEFSYMEXP>{FILENAMECHAR1}{SYMBOLCHARN}*   { yylval.name = xstrdup(yytext); return NAME; }
<DEFSYMEXP>"="                  { RTOKEN('='); }

<MRI,EXPRESSION>"$"([0-9A-Fa-f])+ {
  				yylval.integer = bfd_scan_vma (yytext+1, 0,16);
				return INT;
			}

<MRI,EXPRESSION>([0-9A-Fa-f])+(H|h|X|x|B|b|O|o|D|d) {
				   int ibase ;
				   switch (yytext[yyleng-1]) {
				    case 'X': 
				    case 'x':
				    case 'H':
				    case 'h':
				     ibase = 16;
				     break;
				    case 'O':
				    case 'o':
				     ibase = 8;
				     break;
				    case 'B':
				    case 'b':
				     ibase = 2;
				     break;
				    default:
				     ibase = 10;
				   }
				   yylval.integer = bfd_scan_vma (yytext, 0,
								  ibase);
				   return INT;
				 }
<SCRIPT,DEFSYMEXP,MRI,BOTH,EXPRESSION>((("$"|"0x")([0-9A-Fa-f])+)|(([0-9])+))(M|K|m|k)? {
				  char *s = yytext;

				  if (*s == '$')
				    ++s;
				  yylval.integer = bfd_scan_vma (s, 0, 0);
				  if (yytext[yyleng-1] == 'M'
				      || yytext[yyleng-1] == 'm')
				    yylval.integer *= 1024 * 1024;
				  if (yytext[yyleng-1] == 'K' 
				      || yytext[yyleng-1]=='k')
				    yylval.integer *= 1024;
				  return INT;
				}
<BOTH,SCRIPT,EXPRESSION,MRI>"]"		{ RTOKEN(']');}
<BOTH,SCRIPT,EXPRESSION,MRI>"["		{ RTOKEN('[');}
<BOTH,SCRIPT,EXPRESSION,MRI>"<<="	{ RTOKEN(LSHIFTEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>">>="	{ RTOKEN(RSHIFTEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"||"	{ RTOKEN(OROR);}
<BOTH,SCRIPT,EXPRESSION,MRI>"=="	{ RTOKEN(EQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"!="	{ RTOKEN(NE);}
<BOTH,SCRIPT,EXPRESSION,MRI>">="	{ RTOKEN(GE);}
<BOTH,SCRIPT,EXPRESSION,MRI>"<="	{ RTOKEN(LE);}
<BOTH,SCRIPT,EXPRESSION,MRI>"<<"	{ RTOKEN(LSHIFT);}
<BOTH,SCRIPT,EXPRESSION,MRI>">>"	{ RTOKEN(RSHIFT);}
<BOTH,SCRIPT,EXPRESSION,MRI>"+="	{ RTOKEN(PLUSEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"-="	{ RTOKEN(MINUSEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"*="	{ RTOKEN(MULTEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"/="	{ RTOKEN(DIVEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"&="	{ RTOKEN(ANDEQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"|="	{ RTOKEN(OREQ);}
<BOTH,SCRIPT,EXPRESSION,MRI>"&&"	{ RTOKEN(ANDAND);}
<BOTH,SCRIPT,EXPRESSION,MRI>">"		{ RTOKEN('>');}
<BOTH,SCRIPT,EXPRESSION,MRI>","		{ RTOKEN(',');}
<BOTH,SCRIPT,EXPRESSION,MRI>"&"		{ RTOKEN('&');}
<BOTH,SCRIPT,EXPRESSION,MRI>"|"		{ RTOKEN('|');}
<BOTH,SCRIPT,EXPRESSION,MRI>"~"		{ RTOKEN('~');}
<BOTH,SCRIPT,EXPRESSION,MRI>"!"		{ RTOKEN('!');}
<BOTH,SCRIPT,EXPRESSION,MRI>"?"		{ RTOKEN('?');}
<BOTH,SCRIPT,EXPRESSION,MRI>"*"		{ RTOKEN('*');}
<BOTH,SCRIPT,EXPRESSION,MRI>"+"		{ RTOKEN('+');}
<BOTH,SCRIPT,EXPRESSION,MRI>"-"		{ RTOKEN('-');}
<BOTH,SCRIPT,EXPRESSION,MRI>"/"		{ RTOKEN('/');}
<BOTH,SCRIPT,EXPRESSION,MRI>"%"		{ RTOKEN('%');}
<BOTH,SCRIPT,EXPRESSION,MRI>"<"		{ RTOKEN('<');}
<BOTH,SCRIPT,EXPRESSION,MRI>"="          { RTOKEN('=');}
<BOTH,SCRIPT,EXPRESSION,MRI>"}"			{ RTOKEN('}') ; }
<BOTH,SCRIPT,EXPRESSION,MRI>"{"			{ RTOKEN('{'); }
<BOTH,SCRIPT,EXPRESSION,MRI>")"			{ RTOKEN(')');}
<BOTH,SCRIPT,EXPRESSION,MRI>"("			{ RTOKEN('(');}
<BOTH,SCRIPT,EXPRESSION,MRI>":"		{ RTOKEN(':'); }
<BOTH,SCRIPT,EXPRESSION,MRI>";"		{ RTOKEN(';');}
<BOTH,SCRIPT>"MEMORY"		{ RTOKEN(MEMORY);}
<BOTH,SCRIPT>"ORIGIN"		{ RTOKEN(ORIGIN);}
<BOTH,SCRIPT>"VERSION"		{ RTOKEN(VERSIONK);}
<EXPRESSION,BOTH,SCRIPT>"BLOCK"		{ RTOKEN(BLOCK);}
<EXPRESSION,BOTH,SCRIPT>"BIND"		{ RTOKEN(BIND);}
<BOTH,SCRIPT>"LENGTH"		{ RTOKEN(LENGTH);}
<EXPRESSION,BOTH,SCRIPT>"ALIGN"			{ RTOKEN(ALIGN_K);}
<EXPRESSION,BOTH,SCRIPT>"ADDR"			{ RTOKEN(ADDR);}
<EXPRESSION,BOTH,SCRIPT>"LOADADDR"		{ RTOKEN(LOADADDR);}
<EXPRESSION,BOTH>"MAX"			{ RTOKEN(MAX_K); }
<EXPRESSION,BOTH>"MIN"			{ RTOKEN(MIN_K); }
<EXPRESSION,BOTH>"ASSERT"		{ RTOKEN(ASSERT_K); }
<BOTH,SCRIPT>"ENTRY"			{ RTOKEN(ENTRY);}
<BOTH,SCRIPT,MRI>"EXTERN"		{ RTOKEN(EXTERN);}
<EXPRESSION,BOTH,SCRIPT>"NEXT"			{ RTOKEN(NEXT);}
<EXPRESSION,BOTH,SCRIPT>"sizeof_headers"	{ RTOKEN(SIZEOF_HEADERS);}
<EXPRESSION,BOTH,SCRIPT>"SIZEOF_HEADERS"	{ RTOKEN(SIZEOF_HEADERS);}
<BOTH,SCRIPT>"MAP"			{ RTOKEN(MAP);}
<EXPRESSION,BOTH,SCRIPT>"SIZEOF"		{ RTOKEN(SIZEOF);}
<BOTH,SCRIPT>"TARGET"		{ RTOKEN(TARGET_K);}
<BOTH,SCRIPT>"SEARCH_DIR"		{ RTOKEN(SEARCH_DIR);}
<BOTH,SCRIPT>"OUTPUT"		{ RTOKEN(OUTPUT);}
<BOTH,SCRIPT>"INPUT"			{ RTOKEN(INPUT);}
<EXPRESSION,BOTH,SCRIPT>"GROUP"		{ RTOKEN(GROUP);}
<EXPRESSION,BOTH,SCRIPT>"DEFINED"		{ RTOKEN(DEFINED);}
<BOTH,SCRIPT>"CREATE_OBJECT_SYMBOLS"	{ RTOKEN(CREATE_OBJECT_SYMBOLS);}
<BOTH,SCRIPT>"CONSTRUCTORS"		{ RTOKEN( CONSTRUCTORS);}
<BOTH,SCRIPT>"FORCE_COMMON_ALLOCATION" { RTOKEN(FORCE_COMMON_ALLOCATION);}
<BOTH,SCRIPT>"INHIBIT_COMMON_ALLOCATION" { RTOKEN(INHIBIT_COMMON_ALLOCATION);}
<BOTH,SCRIPT>"SECTIONS"		{ RTOKEN(SECTIONS);}
<BOTH,SCRIPT>"FILL"			{ RTOKEN(FILL);}
<BOTH,SCRIPT>"STARTUP"		{ RTOKEN(STARTUP);}
<BOTH,SCRIPT>"OUTPUT_FORMAT"		{ RTOKEN(OUTPUT_FORMAT);}
<BOTH,SCRIPT>"OUTPUT_ARCH"		{ RTOKEN( OUTPUT_ARCH);}
<BOTH,SCRIPT>"HLL"			{ RTOKEN(HLL);}
<BOTH,SCRIPT>"SYSLIB"		{ RTOKEN(SYSLIB);}
<BOTH,SCRIPT>"FLOAT"			{ RTOKEN(FLOAT);}
<BOTH,SCRIPT>"QUAD"			{ RTOKEN( QUAD);}
<BOTH,SCRIPT>"SQUAD"			{ RTOKEN( SQUAD);}
<BOTH,SCRIPT>"LONG"			{ RTOKEN( LONG);}
<BOTH,SCRIPT>"SHORT"			{ RTOKEN( SHORT);}
<BOTH,SCRIPT>"BYTE"			{ RTOKEN( BYTE);}
<BOTH,SCRIPT>"NOFLOAT"		{ RTOKEN(NOFLOAT);}
<EXPRESSION,BOTH,SCRIPT>"NOCROSSREFS"	{ RTOKEN(NOCROSSREFS);}
<BOTH,SCRIPT>"OVERLAY"			{ RTOKEN(OVERLAY); }
<BOTH,SCRIPT>"SORT"			{ RTOKEN(SORT); }
<EXPRESSION,BOTH,SCRIPT>"NOLOAD"	{ RTOKEN(NOLOAD);}
<EXPRESSION,BOTH,SCRIPT>"DSECT"		{ RTOKEN(DSECT);}
<EXPRESSION,BOTH,SCRIPT>"COPY"		{ RTOKEN(COPY);}
<EXPRESSION,BOTH,SCRIPT>"INFO"		{ RTOKEN(INFO);}
<EXPRESSION,BOTH,SCRIPT>"OVERLAY"	{ RTOKEN(OVERLAY);}
<BOTH,SCRIPT>"o"			{ RTOKEN(ORIGIN);}
<BOTH,SCRIPT>"org"			{ RTOKEN(ORIGIN);}
<BOTH,SCRIPT>"l"			{ RTOKEN( LENGTH);}
<BOTH,SCRIPT>"len"			{ RTOKEN( LENGTH);}
<BOTH,SCRIPT>"INCLUDE"			{ RTOKEN(INCLUDE);}
<BOTH,SCRIPT>"PHDRS"			{ RTOKEN (PHDRS); }
<EXPRESSION,BOTH,SCRIPT>"AT"			{ RTOKEN(AT);}
<EXPRESSION,BOTH,SCRIPT>"PROVIDE"		{ RTOKEN(PROVIDE); }
<EXPRESSION,BOTH,SCRIPT>"KEEP"		{ RTOKEN(KEEP); }
<EXPRESSION,BOTH,SCRIPT>"EXCLUDE_FILE"        { RTOKEN(EXCLUDE_FILE); }
<MRI>"#".*\n?			{ ++ lineno; }
<MRI>"\n"	                { ++ lineno;  RTOKEN(NEWLINE); }
<MRI>"*".*			{ /* Mri comment line */ }
<MRI>";".*			{ /* Mri comment line */ }
<MRI>"END"                      { RTOKEN(ENDWORD); }
<MRI>"ALIGNMOD"		{ RTOKEN(ALIGNMOD);}
<MRI>"ALIGN"		{ RTOKEN(ALIGN_K);}
<MRI>"CHIP"                     { RTOKEN(CHIP); }
<MRI>"BASE"                     { RTOKEN(BASE); }
<MRI>"ALIAS"                     { RTOKEN(ALIAS); }
<MRI>"TRUNCATE"                     { RTOKEN(TRUNCATE); }
<MRI>"LOAD"                     { RTOKEN(LOAD); }
<MRI>"PUBLIC"                   { RTOKEN(PUBLIC); }
<MRI>"ORDER"                    { RTOKEN(ORDER); }
<MRI>"NAME"                     { RTOKEN(NAMEWORD); }
<MRI>"FORMAT"                   { RTOKEN(FORMAT); }
<MRI>"CASE"                     { RTOKEN(CASE); }
<MRI>"START"                    { RTOKEN(START); }
<MRI>"LIST".*                   { RTOKEN(LIST); /* LIST and ignore to end of line */ }
<MRI>"SECT"			{ RTOKEN(SECT); }
<EXPRESSION,BOTH,SCRIPT,MRI>"ABSOLUTE"			{ RTOKEN(ABSOLUTE); }
<MRI>"end"                      { RTOKEN(ENDWORD); }
<MRI>"alignmod"		{ RTOKEN(ALIGNMOD);}
<MRI>"align"		{ RTOKEN(ALIGN_K);}
<MRI>"chip"                     { RTOKEN(CHIP); }
<MRI>"base"                     { RTOKEN(BASE); }
<MRI>"alias"                     { RTOKEN(ALIAS); }
<MRI>"truncate"                     { RTOKEN(TRUNCATE); }
<MRI>"load"                     { RTOKEN(LOAD); }
<MRI>"public"                   { RTOKEN(PUBLIC); }
<MRI>"order"                    { RTOKEN(ORDER); }
<MRI>"name"                     { RTOKEN(NAMEWORD); }
<MRI>"format"                   { RTOKEN(FORMAT); }
<MRI>"case"                     { RTOKEN(CASE); }
<MRI>"extern"                   { RTOKEN(EXTERN); }
<MRI>"start"                    { RTOKEN(START); }
<MRI>"list".*                   { RTOKEN(LIST); /* LIST and ignore to end of line */ }
<MRI>"sect"			{ RTOKEN(SECT); }
<EXPRESSION,BOTH,SCRIPT,MRI>"absolute"			{ RTOKEN(ABSOLUTE); }

<MRI>{FILENAMECHAR1}{NOCFILENAMECHAR}*	{
/* Filename without commas, needed to parse mri stuff */
				 yylval.name = xstrdup(yytext); 
				  return NAME;
				}


<BOTH,EXPRESSION>{FILENAMECHAR1}{FILENAMECHAR}*	{
				 yylval.name = xstrdup(yytext); 
				  return NAME;
				}
<BOTH,EXPRESSION>"-l"{FILENAMECHAR}+ {
				  yylval.name = xstrdup (yytext + 2);
				  return LNAME;
				}
<SCRIPT>{WILDCHAR}* {
		/* Annoyingly, this pattern can match comments, and we have
		   longest match issues to consider.  So if the first two
		   characters are a comment opening, put the input back and
		   try again.  */
		if (yytext[0] == '/' && yytext[1] == '*')
		  {
		    yyless(2);
		    comment ();
		  }
		else
		  {
		    yylval.name = xstrdup(yytext);
		    return NAME;
		  }
	}

<EXPRESSION,BOTH,SCRIPT,VERS_NODE>"\""[^\"]*"\"" {
					/* No matter the state, quotes
					   give what's inside */
					yylval.name = xstrdup(yytext+1);
					yylval.name[yyleng-2] = 0;
					return NAME;
				}
<BOTH,SCRIPT,EXPRESSION>"\n"		{ lineno++;}
<MRI,BOTH,SCRIPT,EXPRESSION>[ \t\r]+	{ }

<VERS_NODE,VERS_SCRIPT>[:,;]	{ return *yytext; }

<VERS_NODE>global		{ RTOKEN(GLOBAL); }

<VERS_NODE>local		{ RTOKEN(LOCAL); }

<VERS_NODE>extern		{ RTOKEN(EXTERN); }

<VERS_NODE>{V_IDENTIFIER}	{ yylval.name = xstrdup (yytext);
				  return VERS_IDENTIFIER; }

<VERS_SCRIPT>{V_TAG}		{ yylval.name = xstrdup (yytext);
				  return VERS_TAG; }

<VERS_START>"{"			{ BEGIN(VERS_SCRIPT); return *yytext; }

<VERS_SCRIPT>"{"		{ BEGIN(VERS_NODE); 
				  vers_node_nesting = 0;
				  return *yytext;
				}
<VERS_SCRIPT>"}"		{ return *yytext; }
<VERS_NODE>"{"			{ vers_node_nesting++; return *yytext; }
<VERS_NODE>"}"			{ if (--vers_node_nesting < 0)
				    BEGIN(VERS_SCRIPT);
				  return *yytext;
				}

<VERS_START,VERS_NODE,VERS_SCRIPT>[\n]		{ lineno++; }

<VERS_START,VERS_NODE,VERS_SCRIPT>#.*		{ /* Eat up comments */ }

<VERS_START,VERS_NODE,VERS_SCRIPT>[ \t\r]+   	{ /* Eat up whitespace */ }

<<EOF>> {
  include_stack_ptr--;
    
  if (include_stack_ptr == 0) 
  {
    yyterminate();
  }
  else 
  {
    yy_switch_to_buffer(include_stack[include_stack_ptr]);
  }

  ldfile_input_filename = file_name_stack[include_stack_ptr - 1];
  lineno = lineno_stack[include_stack_ptr];

  return END;
}

<SCRIPT,MRI,VERS_START,VERS_SCRIPT,VERS_NODE>.	lex_warn_invalid(" in script", yytext);
<EXPRESSION,DEFSYMEXP,BOTH>.	lex_warn_invalid(" in expression", yytext);
    
%%


/* Switch flex to reading script file NAME, open on FILE,
   saving the current input info on the include stack.  */

void
lex_push_file (file, name)
     FILE *file;
     const char *name;
{
  if (include_stack_ptr >= MAX_INCLUDE_DEPTH) 
    {
      einfo("%F:includes nested too deeply\n");
    }
  file_name_stack[include_stack_ptr] = name;
  lineno_stack[include_stack_ptr] = lineno;
  include_stack[include_stack_ptr] = YY_CURRENT_BUFFER;

  include_stack_ptr++;
  lineno = 1;
  yyin = file;
  yy_switch_to_buffer(yy_create_buffer(yyin, YY_BUF_SIZE));
}

/* Return a newly created flex input buffer containing STRING,
   which is SIZE bytes long.  */

static YY_BUFFER_STATE 
yy_create_string_buffer (string, size)
     CONST char *string;
     size_t size;
{
  YY_BUFFER_STATE b;

  /* Calls to m-alloc get turned by sed into xm-alloc.  */
  b = (YY_BUFFER_STATE) malloc (sizeof (struct yy_buffer_state));
  b->yy_input_file = 0;
  b->yy_buf_size = size;

  /* yy_ch_buf has to be 2 characters longer than the size given because
     we need to put in 2 end-of-buffer characters.  */
  b->yy_ch_buf = (char *) malloc ((unsigned) (b->yy_buf_size + 3));

  b->yy_ch_buf[0] = '\n';
  strcpy (b->yy_ch_buf+1, string);
  b->yy_ch_buf[size+1] = YY_END_OF_BUFFER_CHAR;
  b->yy_ch_buf[size+2] = YY_END_OF_BUFFER_CHAR;
  b->yy_n_chars = size+1;
  b->yy_buf_pos = &b->yy_ch_buf[1];

  b->yy_is_our_buffer = 1;
  b->yy_is_interactive = 0;
  b->yy_at_bol = 1;
  b->yy_fill_buffer = 0;

  /* flex 2.4.7 changed the interface.  FIXME: We should not be using
     a flex internal interface in the first place!  */
#ifdef YY_BUFFER_NEW
  b->yy_buffer_status = YY_BUFFER_NEW;
#else
  b->yy_eof_status = EOF_NOT_SEEN;
#endif

  return b;
}

/* Switch flex to reading from STRING, saving the current input info
   on the include stack.  */

void
lex_redirect (string)
     CONST char *string;
{
  YY_BUFFER_STATE tmp;

  yy_init = 0;
  if (include_stack_ptr >= MAX_INCLUDE_DEPTH) 
    {
      einfo("%F: macros nested too deeply\n");
    }
  file_name_stack[include_stack_ptr] = "redirect";
  lineno_stack[include_stack_ptr] = lineno;
  include_stack[include_stack_ptr] = YY_CURRENT_BUFFER;
  include_stack_ptr++;
  lineno = 1;
  tmp = yy_create_string_buffer (string, strlen (string));
  yy_switch_to_buffer (tmp);
}

/* Functions to switch to a different flex start condition,
   saving the current start condition on `state_stack'.  */

static int state_stack[MAX_INCLUDE_DEPTH * 2];
static int *state_stack_p = state_stack;

void
ldlex_script ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (SCRIPT);
}

void
ldlex_mri_script ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (MRI);
}

void
ldlex_version_script ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (VERS_START);
}

void
ldlex_version_file ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (VERS_SCRIPT);
}

void
ldlex_defsym ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (DEFSYMEXP);
}
	   
void
ldlex_expression ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (EXPRESSION);
}

void
ldlex_both ()
{
  *(state_stack_p)++ = yy_start;
  BEGIN (BOTH);
}

void
ldlex_popstate ()
{
  yy_start = *(--state_stack_p);
}


/* Place up to MAX_SIZE characters in BUF and return in *RESULT
   either the number of characters read, or 0 to indicate EOF.  */

static void
yy_input (buf, result, max_size)
     char *buf;
     int *result;
     int max_size;
{
  *result = 0; 
  if (yy_current_buffer->yy_input_file)
    {
      if (yyin)
	{
	  *result = read (fileno (yyin), (char *) buf, max_size);
	  if (*result < 0) 
	    einfo ("%F%P: read in flex scanner failed\n");
	}
    }
}

/* Eat the rest of a C-style comment.  */

static void
comment ()
{
  int c;

  while (1)
  {
    c = input();
    while (c != '*' && c != EOF) 
    {
      if (c == '\n')
	lineno++;
      c = input();
    }

    if (c == '*')
    {
      c = input();
      while (c == '*')
       c = input();
      if (c == '/')
       break;			/* found the end */
    }

    if (c == '\n')
      lineno++;

    if (c == EOF)
    {
      einfo( "%F%P: EOF in comment\n");
      break;
    }
  }
}

/* Warn the user about a garbage character WHAT in the input
   in context WHERE.  */

static void
lex_warn_invalid (where, what)
     char *where, *what;
{
  char buf[5];

  /* If we have found an input file whose format we do not recognize,
     and we are therefore treating it as a linker script, and we find
     an invalid character, then most likely this is a real object file
     of some different format.  Treat it as such.  */
  if (ldfile_assumed_script)
    {
      bfd_set_error (bfd_error_file_not_recognized);
      einfo ("%F%s: file not recognized: %E\n", ldfile_input_filename);
    }

  if (! ISPRINT (*what))
    {
      sprintf (buf, "\\%03o", (unsigned int) *what);
      what = buf;
    }

  einfo ("%P:%S: ignoring invalid character `%s'%s\n", what, where);
}
opt">* 100 / NUM_GCOV_WORKING_SETS; for (ws_ix = 0, pct = pctinc; ws_ix < NUM_GCOV_WORKING_SETS; ws_ix++, pct += pctinc) { if (ws_ix == NUM_GCOV_WORKING_SETS - 1) pct = 9990; ws_info = &gcov_working_sets[ws_ix]; /* Print out the percentage using int arithmatic to avoid float. */ fprintf (dump_file, "\t\t%u.%02u%%: num counts=%u, min counter=" "%"PRId64 "\n", pct / 100, pct - (pct / 100 * 100), ws_info->num_counters, (int64_t)ws_info->min_counter); } } } /* Given a the desired percentage of the full profile (sum_all from the summary), multiplied by 10 to avoid float in PCT_TIMES_10, returns the corresponding working set information. If an exact match for the percentage isn't found, the closest value is used. */ gcov_working_set_t * find_working_set (unsigned pct_times_10) { unsigned i; if (!profile_info) return NULL; gcc_assert (pct_times_10 <= 1000); if (pct_times_10 >= 999) return &gcov_working_sets[NUM_GCOV_WORKING_SETS - 1]; i = pct_times_10 * NUM_GCOV_WORKING_SETS / 1000; if (!i) return &gcov_working_sets[0]; return &gcov_working_sets[i - 1]; } /* Computes hybrid profile for all matching entries in da_file. CFG_CHECKSUM is the precomputed checksum for the CFG. */ static gcov_type * get_exec_counts (unsigned cfg_checksum, unsigned lineno_checksum) { unsigned num_edges = 0; basic_block bb; gcov_type *counts; /* Count the edges to be (possibly) instrumented. */ FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, bb->succs) if (!EDGE_INFO (e)->ignore && !EDGE_INFO (e)->on_tree) num_edges++; } counts = get_coverage_counts (GCOV_COUNTER_ARCS, num_edges, cfg_checksum, lineno_checksum, &profile_info); if (!counts) return NULL; get_working_sets (); if (dump_file && profile_info) fprintf (dump_file, "Merged %u profiles with maximal count %u.\n", profile_info->runs, (unsigned) profile_info->sum_max); return counts; } static bool is_edge_inconsistent (vec<edge, va_gc> *edges) { edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, edges) { if (!EDGE_INFO (e)->ignore) { if (e->count < 0 && (!(e->flags & EDGE_FAKE) || !block_ends_with_call_p (e->src))) { if (dump_file) { fprintf (dump_file, "Edge %i->%i is inconsistent, count%"PRId64, e->src->index, e->dest->index, e->count); dump_bb (dump_file, e->src, 0, TDF_DETAILS); dump_bb (dump_file, e->dest, 0, TDF_DETAILS); } return true; } } } return false; } static void correct_negative_edge_counts (void) { basic_block bb; edge e; edge_iterator ei; FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { FOR_EACH_EDGE (e, ei, bb->succs) { if (e->count < 0) e->count = 0; } } } /* Check consistency. Return true if inconsistency is found. */ static bool is_inconsistent (void) { basic_block bb; bool inconsistent = false; FOR_EACH_BB_FN (bb, cfun) { inconsistent |= is_edge_inconsistent (bb->preds); if (!dump_file && inconsistent) return true; inconsistent |= is_edge_inconsistent (bb->succs); if (!dump_file && inconsistent) return true; if (bb->count < 0) { if (dump_file) { fprintf (dump_file, "BB %i count is negative " "%"PRId64, bb->index, bb->count); dump_bb (dump_file, bb, 0, TDF_DETAILS); } inconsistent = true; } if (bb->count != sum_edge_counts (bb->preds)) { if (dump_file) { fprintf (dump_file, "BB %i count does not match sum of incoming edges " "%"PRId64" should be %"PRId64, bb->index, bb->count, sum_edge_counts (bb->preds)); dump_bb (dump_file, bb, 0, TDF_DETAILS); } inconsistent = true; } if (bb->count != sum_edge_counts (bb->succs) && ! (find_edge (bb, EXIT_BLOCK_PTR_FOR_FN (cfun)) != NULL && block_ends_with_call_p (bb))) { if (dump_file) { fprintf (dump_file, "BB %i count does not match sum of outgoing edges " "%"PRId64" should be %"PRId64, bb->index, bb->count, sum_edge_counts (bb->succs)); dump_bb (dump_file, bb, 0, TDF_DETAILS); } inconsistent = true; } if (!dump_file && inconsistent) return true; } return inconsistent; } /* Set each basic block count to the sum of its outgoing edge counts */ static void set_bb_counts (void) { basic_block bb; FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { bb->count = sum_edge_counts (bb->succs); gcc_assert (bb->count >= 0); } } /* Reads profile data and returns total number of edge counts read */ static int read_profile_edge_counts (gcov_type *exec_counts) { basic_block bb; int num_edges = 0; int exec_counts_pos = 0; /* For each edge not on the spanning tree, set its execution count from the .da file. */ /* The first count in the .da file is the number of times that the function was entered. This is the exec_count for block zero. */ FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, bb->succs) if (!EDGE_INFO (e)->ignore && !EDGE_INFO (e)->on_tree) { num_edges++; if (exec_counts) { e->count = exec_counts[exec_counts_pos++]; if (e->count > profile_info->sum_max) { if (flag_profile_correction) { static bool informed = 0; if (dump_enabled_p () && !informed) dump_printf_loc (MSG_NOTE, input_location, "corrupted profile info: edge count" " exceeds maximal count\n"); informed = 1; } else error ("corrupted profile info: edge from %i to %i exceeds maximal count", bb->index, e->dest->index); } } else e->count = 0; EDGE_INFO (e)->count_valid = 1; BB_INFO (bb)->succ_count--; BB_INFO (e->dest)->pred_count--; if (dump_file) { fprintf (dump_file, "\nRead edge from %i to %i, count:", bb->index, e->dest->index); fprintf (dump_file, "%"PRId64, (int64_t) e->count); } } } return num_edges; } #define OVERLAP_BASE 10000 /* Compare the static estimated profile to the actual profile, and return the "degree of overlap" measure between them. Degree of overlap is a number between 0 and OVERLAP_BASE. It is the sum of each basic block's minimum relative weights between two profiles. And overlap of OVERLAP_BASE means two profiles are identical. */ static int compute_frequency_overlap (void) { gcov_type count_total = 0, freq_total = 0; int overlap = 0; basic_block bb; FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { count_total += bb->count; freq_total += bb->frequency; } if (count_total == 0 || freq_total == 0) return 0; FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) overlap += MIN (bb->count * OVERLAP_BASE / count_total, bb->frequency * OVERLAP_BASE / freq_total); return overlap; } /* Compute the branch probabilities for the various branches. Annotate them accordingly. CFG_CHECKSUM is the precomputed checksum for the CFG. */ static void compute_branch_probabilities (unsigned cfg_checksum, unsigned lineno_checksum) { basic_block bb; int i; int num_edges = 0; int changes; int passes; int hist_br_prob[20]; int num_branches; gcov_type *exec_counts = get_exec_counts (cfg_checksum, lineno_checksum); int inconsistent = 0; /* Very simple sanity checks so we catch bugs in our profiling code. */ if (!profile_info) return; if (profile_info->sum_all < profile_info->sum_max) { error ("corrupted profile info: sum_all is smaller than sum_max"); exec_counts = NULL; } /* Attach extra info block to each bb. */ alloc_aux_for_blocks (sizeof (struct bb_profile_info)); FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { edge e; edge_iterator ei; FOR_EACH_EDGE (e, ei, bb->succs) if (!EDGE_INFO (e)->ignore) BB_INFO (bb)->succ_count++; FOR_EACH_EDGE (e, ei, bb->preds) if (!EDGE_INFO (e)->ignore) BB_INFO (bb)->pred_count++; } /* Avoid predicting entry on exit nodes. */ BB_INFO (EXIT_BLOCK_PTR_FOR_FN (cfun))->succ_count = 2; BB_INFO (ENTRY_BLOCK_PTR_FOR_FN (cfun))->pred_count = 2; num_edges = read_profile_edge_counts (exec_counts); if (dump_file) fprintf (dump_file, "\n%d edge counts read\n", num_edges); /* For every block in the file, - if every exit/entrance edge has a known count, then set the block count - if the block count is known, and every exit/entrance edge but one has a known execution count, then set the count of the remaining edge As edge counts are set, decrement the succ/pred count, but don't delete the edge, that way we can easily tell when all edges are known, or only one edge is unknown. */ /* The order that the basic blocks are iterated through is important. Since the code that finds spanning trees starts with block 0, low numbered edges are put on the spanning tree in preference to high numbered edges. Hence, most instrumented edges are at the end. Graph solving works much faster if we propagate numbers from the end to the start. This takes an average of slightly more than 3 passes. */ changes = 1; passes = 0; while (changes) { passes++; changes = 0; FOR_BB_BETWEEN (bb, EXIT_BLOCK_PTR_FOR_FN (cfun), NULL, prev_bb) { struct bb_profile_info *bi = BB_INFO (bb); if (! bi->count_valid) { if (bi->succ_count == 0) { edge e; edge_iterator ei; gcov_type total = 0; FOR_EACH_EDGE (e, ei, bb->succs) total += e->count; bb->count = total; bi->count_valid = 1; changes = 1; } else if (bi->pred_count == 0) { edge e; edge_iterator ei; gcov_type total = 0; FOR_EACH_EDGE (e, ei, bb->preds) total += e->count; bb->count = total; bi->count_valid = 1; changes = 1; } } if (bi->count_valid) { if (bi->succ_count == 1) { edge e; edge_iterator ei; gcov_type total = 0; /* One of the counts will be invalid, but it is zero, so adding it in also doesn't hurt. */ FOR_EACH_EDGE (e, ei, bb->succs) total += e->count; /* Search for the invalid edge, and set its count. */ FOR_EACH_EDGE (e, ei, bb->succs) if (! EDGE_INFO (e)->count_valid && ! EDGE_INFO (e)->ignore) break; /* Calculate count for remaining edge by conservation. */ total = bb->count - total; gcc_assert (e); EDGE_INFO (e)->count_valid = 1; e->count = total; bi->succ_count--; BB_INFO (e->dest)->pred_count--; changes = 1; } if (bi->pred_count == 1) { edge e; edge_iterator ei; gcov_type total = 0; /* One of the counts will be invalid, but it is zero, so adding it in also doesn't hurt. */ FOR_EACH_EDGE (e, ei, bb->preds) total += e->count; /* Search for the invalid edge, and set its count. */ FOR_EACH_EDGE (e, ei, bb->preds) if (!EDGE_INFO (e)->count_valid && !EDGE_INFO (e)->ignore) break; /* Calculate count for remaining edge by conservation. */ total = bb->count - total + e->count; gcc_assert (e); EDGE_INFO (e)->count_valid = 1; e->count = total; bi->pred_count--; BB_INFO (e->src)->succ_count--; changes = 1; } } } } if (dump_file) { int overlap = compute_frequency_overlap (); gimple_dump_cfg (dump_file, dump_flags); fprintf (dump_file, "Static profile overlap: %d.%d%%\n", overlap / (OVERLAP_BASE / 100), overlap % (OVERLAP_BASE / 100)); } total_num_passes += passes; if (dump_file) fprintf (dump_file, "Graph solving took %d passes.\n\n", passes); /* If the graph has been correctly solved, every block will have a succ and pred count of zero. */ FOR_EACH_BB_FN (bb, cfun) { gcc_assert (!BB_INFO (bb)->succ_count && !BB_INFO (bb)->pred_count); } /* Check for inconsistent basic block counts */ inconsistent = is_inconsistent (); if (inconsistent) { if (flag_profile_correction) { /* Inconsistency detected. Make it flow-consistent. */ static int informed = 0; if (dump_enabled_p () && informed == 0) { informed = 1; dump_printf_loc (MSG_NOTE, input_location, "correcting inconsistent profile data\n"); } correct_negative_edge_counts (); /* Set bb counts to the sum of the outgoing edge counts */ set_bb_counts (); if (dump_file) fprintf (dump_file, "\nCalling mcf_smooth_cfg\n"); mcf_smooth_cfg (); } else error ("corrupted profile info: profile data is not flow-consistent"); } /* For every edge, calculate its branch probability and add a reg_note to the branch insn to indicate this. */ for (i = 0; i < 20; i++) hist_br_prob[i] = 0; num_branches = 0; FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) { edge e; edge_iterator ei; if (bb->count < 0) { error ("corrupted profile info: number of iterations for basic block %d thought to be %i", bb->index, (int)bb->count); bb->count = 0; } FOR_EACH_EDGE (e, ei, bb->succs) { /* Function may return twice in the cased the called function is setjmp or calls fork, but we can't represent this by extra edge from the entry, since extra edge from the exit is already present. We get negative frequency from the entry point. */ if ((e->count < 0 && e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun)) || (e->count > bb->count && e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun))) { if (block_ends_with_call_p (bb)) e->count = e->count < 0 ? 0 : bb->count; } if (e->count < 0 || e->count > bb->count) { error ("corrupted profile info: number of executions for edge %d-%d thought to be %i", e->src->index, e->dest->index, (int)e->count); e->count = bb->count / 2; } } if (bb->count) { FOR_EACH_EDGE (e, ei, bb->succs) e->probability = GCOV_COMPUTE_SCALE (e->count, bb->count); if (bb->index >= NUM_FIXED_BLOCKS && block_ends_with_condjump_p (bb) && EDGE_COUNT (bb->succs) >= 2) { int prob; edge e; int index; /* Find the branch edge. It is possible that we do have fake edges here. */ FOR_EACH_EDGE (e, ei, bb->succs) if (!(e->flags & (EDGE_FAKE | EDGE_FALLTHRU))) break; prob = e->probability; index = prob * 20 / REG_BR_PROB_BASE; if (index == 20) index = 19; hist_br_prob[index]++; num_branches++; } } /* As a last resort, distribute the probabilities evenly. Use simple heuristics that if there are normal edges, give all abnormals frequency of 0, otherwise distribute the frequency over abnormals (this is the case of noreturn calls). */ else if (profile_status_for_fn (cfun) == PROFILE_ABSENT) { int total = 0; FOR_EACH_EDGE (e, ei, bb->succs) if (!(e->flags & (EDGE_COMPLEX | EDGE_FAKE))) total ++; if (total) { FOR_EACH_EDGE (e, ei, bb->succs) if (!(e->flags & (EDGE_COMPLEX | EDGE_FAKE))) e->probability = REG_BR_PROB_BASE / total; else e->probability = 0; } else { total += EDGE_COUNT (bb->succs); FOR_EACH_EDGE (e, ei, bb->succs) e->probability = REG_BR_PROB_BASE / total; } if (bb->index >= NUM_FIXED_BLOCKS && block_ends_with_condjump_p (bb) && EDGE_COUNT (bb->succs) >= 2) num_branches++; } } counts_to_freqs (); profile_status_for_fn (cfun) = PROFILE_READ; compute_function_frequency (); if (dump_file) { fprintf (dump_file, "%d branches\n", num_branches); if (num_branches) for (i = 0; i < 10; i++) fprintf (dump_file, "%d%% branches in range %d-%d%%\n", (hist_br_prob[i] + hist_br_prob[19-i]) * 100 / num_branches, 5 * i, 5 * i + 5); total_num_branches += num_branches; for (i = 0; i < 20; i++) total_hist_br_prob[i] += hist_br_prob[i]; fputc ('\n', dump_file); fputc ('\n', dump_file); } free_aux_for_blocks (); } /* Load value histograms values whose description is stored in VALUES array from .gcda file. CFG_CHECKSUM is the precomputed checksum for the CFG. */ static void compute_value_histograms (histogram_values values, unsigned cfg_checksum, unsigned lineno_checksum) { unsigned i, j, t, any; unsigned n_histogram_counters[GCOV_N_VALUE_COUNTERS]; gcov_type *histogram_counts[GCOV_N_VALUE_COUNTERS]; gcov_type *act_count[GCOV_N_VALUE_COUNTERS]; gcov_type *aact_count; struct cgraph_node *node; for (t = 0; t < GCOV_N_VALUE_COUNTERS; t++) n_histogram_counters[t] = 0; for (i = 0; i < values.length (); i++) { histogram_value hist = values[i]; n_histogram_counters[(int) hist->type] += hist->n_counters; } any = 0; for (t = 0; t < GCOV_N_VALUE_COUNTERS; t++) { if (!n_histogram_counters[t]) { histogram_counts[t] = NULL; continue; } histogram_counts[t] = get_coverage_counts (COUNTER_FOR_HIST_TYPE (t), n_histogram_counters[t], cfg_checksum, lineno_checksum, NULL); if (histogram_counts[t]) any = 1; act_count[t] = histogram_counts[t]; } if (!any) return; for (i = 0; i < values.length (); i++) { histogram_value hist = values[i]; gimple stmt = hist->hvalue.stmt; t = (int) hist->type; aact_count = act_count[t]; if (act_count[t]) act_count[t] += hist->n_counters; gimple_add_histogram_value (cfun, stmt, hist); hist->hvalue.counters = XNEWVEC (gcov_type, hist->n_counters); for (j = 0; j < hist->n_counters; j++) if (aact_count) hist->hvalue.counters[j] = aact_count[j]; else hist->hvalue.counters[j] = 0; /* Time profiler counter is not related to any statement, so that we have to read the counter and set the value to the corresponding call graph node. */ if (hist->type == HIST_TYPE_TIME_PROFILE) { node = cgraph_node::get (hist->fun->decl); node->tp_first_run = hist->hvalue.counters[0]; if (dump_file) fprintf (dump_file, "Read tp_first_run: %d\n", node->tp_first_run); } } for (t = 0; t < GCOV_N_VALUE_COUNTERS; t++) free (histogram_counts[t]); } /* When passed NULL as file_name, initialize. When passed something else, output the necessary commands to change line to LINE and offset to FILE_NAME. */ static void output_location (char const *file_name, int line, gcov_position_t *offset, basic_block bb) { static char const *prev_file_name; static int prev_line; bool name_differs, line_differs; if (!file_name) { prev_file_name = NULL; prev_line = -1; return; } name_differs = !prev_file_name || filename_cmp (file_name, prev_file_name); line_differs = prev_line != line; if (name_differs || line_differs) { if (!*offset) { *offset = gcov_write_tag (GCOV_TAG_LINES); gcov_write_unsigned (bb->index); name_differs = line_differs=true; } /* If this is a new source file, then output the file's name to the .bb file. */ if (name_differs) { prev_file_name = file_name; gcov_write_unsigned (0); gcov_write_string (prev_file_name); } if (line_differs) { gcov_write_unsigned (line); prev_line = line; } } } /* Instrument and/or analyze program behavior based on program the CFG. This function creates a representation of the control flow graph (of the function being compiled) that is suitable for the instrumentation of edges and/or converting measured edge counts to counts on the complete CFG. When FLAG_PROFILE_ARCS is nonzero, this function instruments the edges in the flow graph that are needed to reconstruct the dynamic behavior of the flow graph. This data is written to the gcno file for gcov. When FLAG_BRANCH_PROBABILITIES is nonzero, this function reads auxiliary information from the gcda file containing edge count information from previous executions of the function being compiled. In this case, the control flow graph is annotated with actual execution counts by compute_branch_probabilities(). Main entry point of this file. */ void branch_prob (void) { basic_block bb; unsigned i; unsigned num_edges, ignored_edges; unsigned num_instrumented; struct edge_list *el; histogram_values values = histogram_values (); unsigned cfg_checksum, lineno_checksum; total_num_times_called++; flow_call_edges_add (NULL); add_noreturn_fake_exit_edges (); /* We can't handle cyclic regions constructed using abnormal edges. To avoid these we replace every source of abnormal edge by a fake edge from entry node and every destination by fake edge to exit. This keeps graph acyclic and our calculation exact for all normal edges except for exit and entrance ones. We also add fake exit edges for each call and asm statement in the basic, since it may not return. */ FOR_EACH_BB_FN (bb, cfun) { int need_exit_edge = 0, need_entry_edge = 0; int have_exit_edge = 0, have_entry_edge = 0; edge e; edge_iterator ei; /* Functions returning multiple times are not handled by extra edges. Instead we simply allow negative counts on edges from exit to the block past call and corresponding probabilities. We can't go with the extra edges because that would result in flowgraph that needs to have fake edges outside the spanning tree. */ FOR_EACH_EDGE (e, ei, bb->succs) { gimple_stmt_iterator gsi; gimple last = NULL; /* It may happen that there are compiler generated statements without a locus at all. Go through the basic block from the last to the first statement looking for a locus. */ for (gsi = gsi_last_nondebug_bb (bb); !gsi_end_p (gsi); gsi_prev_nondebug (&gsi)) { last = gsi_stmt (gsi); if (gimple_has_location (last)) break; } /* Edge with goto locus might get wrong coverage info unless it is the only edge out of BB. Don't do that when the locuses match, so if (blah) goto something; is not computed twice. */ if (last && gimple_has_location (last) && LOCATION_LOCUS (e->goto_locus) != UNKNOWN_LOCATION && !single_succ_p (bb) && (LOCATION_FILE (e->goto_locus) != LOCATION_FILE (gimple_location (last)) || (LOCATION_LINE (e->goto_locus) != LOCATION_LINE (gimple_location (last))))) { basic_block new_bb = split_edge (e); edge ne = single_succ_edge (new_bb); ne->goto_locus = e->goto_locus; } if ((e->flags & (EDGE_ABNORMAL | EDGE_ABNORMAL_CALL)) && e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun)) need_exit_edge = 1; if (e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun)) have_exit_edge = 1; } FOR_EACH_EDGE (e, ei, bb->preds) { if ((e->flags & (EDGE_ABNORMAL | EDGE_ABNORMAL_CALL)) && e->src != ENTRY_BLOCK_PTR_FOR_FN (cfun)) need_entry_edge = 1; if (e->src == ENTRY_BLOCK_PTR_FOR_FN (cfun)) have_entry_edge = 1; } if (need_exit_edge && !have_exit_edge) { if (dump_file) fprintf (dump_file, "Adding fake exit edge to bb %i\n", bb->index); make_edge (bb, EXIT_BLOCK_PTR_FOR_FN (cfun), EDGE_FAKE); } if (need_entry_edge && !have_entry_edge) { if (dump_file) fprintf (dump_file, "Adding fake entry edge to bb %i\n", bb->index); make_edge (ENTRY_BLOCK_PTR_FOR_FN (cfun), bb, EDGE_FAKE); /* Avoid bbs that have both fake entry edge and also some exit edge. One of those edges wouldn't be added to the spanning tree, but we can't instrument any of them. */ if (have_exit_edge || need_exit_edge) { gimple_stmt_iterator gsi; gimple first; gsi = gsi_start_nondebug_after_labels_bb (bb); gcc_checking_assert (!gsi_end_p (gsi)); first = gsi_stmt (gsi); /* Don't split the bbs containing __builtin_setjmp_receiver or ABNORMAL_DISPATCHER calls. These are very special and don't expect anything to be inserted before them. */ if (is_gimple_call (first) && (gimple_call_builtin_p (first, BUILT_IN_SETJMP_RECEIVER) || (gimple_call_flags (first) & ECF_RETURNS_TWICE) || (gimple_call_internal_p (first) && (gimple_call_internal_fn (first) == IFN_ABNORMAL_DISPATCHER)))) continue; if (dump_file) fprintf (dump_file, "Splitting bb %i after labels\n", bb->index); split_block_after_labels (bb); } } } el = create_edge_list (); num_edges = NUM_EDGES (el); alloc_aux_for_edges (sizeof (struct edge_profile_info)); /* The basic blocks are expected to be numbered sequentially. */ compact_blocks (); ignored_edges = 0; for (i = 0 ; i < num_edges ; i++) { edge e = INDEX_EDGE (el, i); e->count = 0; /* Mark edges we've replaced by fake edges above as ignored. */ if ((e->flags & (EDGE_ABNORMAL | EDGE_ABNORMAL_CALL)) && e->src != ENTRY_BLOCK_PTR_FOR_FN (cfun) && e->dest != EXIT_BLOCK_PTR_FOR_FN (cfun)) { EDGE_INFO (e)->ignore = 1; ignored_edges++; } } /* Create spanning tree from basic block graph, mark each edge that is on the spanning tree. We insert as many abnormal and critical edges as possible to minimize number of edge splits necessary. */ find_spanning_tree (el); /* Fake edges that are not on the tree will not be instrumented, so mark them ignored. */ for (num_instrumented = i = 0; i < num_edges; i++) { edge e = INDEX_EDGE (el, i); struct edge_profile_info *inf = EDGE_INFO (e); if (inf->ignore || inf->on_tree) /*NOP*/; else if (e->flags & EDGE_FAKE) { inf->ignore = 1; ignored_edges++; } else num_instrumented++; } total_num_blocks += n_basic_blocks_for_fn (cfun); if (dump_file) fprintf (dump_file, "%d basic blocks\n", n_basic_blocks_for_fn (cfun)); total_num_edges += num_edges; if (dump_file) fprintf (dump_file, "%d edges\n", num_edges); total_num_edges_ignored += ignored_edges; if (dump_file) fprintf (dump_file, "%d ignored edges\n", ignored_edges); total_num_edges_instrumented += num_instrumented; if (dump_file) fprintf (dump_file, "%d instrumentation edges\n", num_instrumented); /* Compute two different checksums. Note that we want to compute the checksum in only once place, since it depends on the shape of the control flow which can change during various transformations. */ cfg_checksum = coverage_compute_cfg_checksum (cfun); lineno_checksum = coverage_compute_lineno_checksum (); /* Write the data from which gcov can reconstruct the basic block graph and function line numbers (the gcno file). */ if (coverage_begin_function (lineno_checksum, cfg_checksum)) { gcov_position_t offset; /* Basic block flags */ offset = gcov_write_tag (GCOV_TAG_BLOCKS); for (i = 0; i != (unsigned) (n_basic_blocks_for_fn (cfun)); i++) gcov_write_unsigned (0); gcov_write_length (offset); /* Arcs */ FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), EXIT_BLOCK_PTR_FOR_FN (cfun), next_bb) { edge e; edge_iterator ei; offset = gcov_write_tag (GCOV_TAG_ARCS); gcov_write_unsigned (bb->index); FOR_EACH_EDGE (e, ei, bb->succs) { struct edge_profile_info *i = EDGE_INFO (e); if (!i->ignore) { unsigned flag_bits = 0; if (i->on_tree) flag_bits |= GCOV_ARC_ON_TREE; if (e->flags & EDGE_FAKE) flag_bits |= GCOV_ARC_FAKE; if (e->flags & EDGE_FALLTHRU) flag_bits |= GCOV_ARC_FALLTHROUGH; /* On trees we don't have fallthru flags, but we can recompute them from CFG shape. */ if (e->flags & (EDGE_TRUE_VALUE | EDGE_FALSE_VALUE) && e->src->next_bb == e->dest) flag_bits |= GCOV_ARC_FALLTHROUGH; gcov_write_unsigned (e->dest->index); gcov_write_unsigned (flag_bits); } } gcov_write_length (offset); } /* Line numbers. */ /* Initialize the output. */ output_location (NULL, 0, NULL, NULL); FOR_EACH_BB_FN (bb, cfun) { gimple_stmt_iterator gsi; gcov_position_t offset = 0; if (bb == ENTRY_BLOCK_PTR_FOR_FN (cfun)->next_bb) { expanded_location curr_location = expand_location (DECL_SOURCE_LOCATION (current_function_decl)); output_location (curr_location.file, curr_location.line, &offset, bb); } for (gsi = gsi_start_bb (bb); !gsi_end_p (gsi); gsi_next (&gsi)) { gimple stmt = gsi_stmt (gsi); if (gimple_has_location (stmt)) output_location (gimple_filename (stmt), gimple_lineno (stmt), &offset, bb); } /* Notice GOTO expressions eliminated while constructing the CFG. */ if (single_succ_p (bb) && LOCATION_LOCUS (single_succ_edge (bb)->goto_locus) != UNKNOWN_LOCATION) { expanded_location curr_location = expand_location (single_succ_edge (bb)->goto_locus); output_location (curr_location.file, curr_location.line, &offset, bb); } if (offset) { /* A file of NULL indicates the end of run. */ gcov_write_unsigned (0); gcov_write_string (NULL); gcov_write_length (offset); } } } if (flag_profile_values) gimple_find_values_to_profile (&values); if (flag_branch_probabilities) { compute_branch_probabilities (cfg_checksum, lineno_checksum); if (flag_profile_values) compute_value_histograms (values, cfg_checksum, lineno_checksum); } remove_fake_edges (); /* For each edge not on the spanning tree, add counting code. */ if (profile_arc_flag && coverage_counter_alloc (GCOV_COUNTER_ARCS, num_instrumented)) { unsigned n_instrumented; gimple_init_edge_profiler (); n_instrumented = instrument_edges (el); gcc_assert (n_instrumented == num_instrumented); if (flag_profile_values) instrument_values (values); /* Commit changes done by instrumentation. */ gsi_commit_edge_inserts (); } free_aux_for_edges (); values.release (); free_edge_list (el); coverage_end_function (lineno_checksum, cfg_checksum); } /* Union find algorithm implementation for the basic blocks using aux fields. */ static basic_block find_group (basic_block bb) { basic_block group = bb, bb1; while ((basic_block) group->aux != group) group = (basic_block) group->aux; /* Compress path. */ while ((basic_block) bb->aux != group) { bb1 = (basic_block) bb->aux; bb->aux = (void *) group; bb = bb1; } return group; } static void union_groups (basic_block bb1, basic_block bb2) { basic_block bb1g = find_group (bb1); basic_block bb2g = find_group (bb2); /* ??? I don't have a place for the rank field. OK. Lets go w/o it, this code is unlikely going to be performance problem anyway. */ gcc_assert (bb1g != bb2g); bb1g->aux = bb2g; } /* This function searches all of the edges in the program flow graph, and puts as many bad edges as possible onto the spanning tree. Bad edges include abnormals edges, which can't be instrumented at the moment. Since it is possible for fake edges to form a cycle, we will have to develop some better way in the future. Also put critical edges to the tree, since they are more expensive to instrument. */ static void find_spanning_tree (struct edge_list *el) { int i; int num_edges = NUM_EDGES (el); basic_block bb; /* We use aux field for standard union-find algorithm. */ FOR_BB_BETWEEN (bb, ENTRY_BLOCK_PTR_FOR_FN (cfun), NULL, next_bb) bb->aux = bb; /* Add fake edge exit to entry we can't instrument. */ union_groups (EXIT_BLOCK_PTR_FOR_FN (cfun), ENTRY_BLOCK_PTR_FOR_FN (cfun)); /* First add all abnormal edges to the tree unless they form a cycle. Also add all edges to the exit block to avoid inserting profiling code behind setting return value from function. */ for (i = 0; i < num_edges; i++) { edge e = INDEX_EDGE (el, i); if (((e->flags & (EDGE_ABNORMAL | EDGE_ABNORMAL_CALL | EDGE_FAKE)) || e->dest == EXIT_BLOCK_PTR_FOR_FN (cfun)) && !EDGE_INFO (e)->ignore && (find_group (e->src) != find_group (e->dest))) { if (dump_file) fprintf (dump_file, "Abnormal edge %d to %d put to tree\n", e->src->index, e->dest->index); EDGE_INFO (e)->on_tree = 1; union_groups (e->src, e->dest); } } /* Now insert all critical edges to the tree unless they form a cycle. */ for (i = 0; i < num_edges; i++) { edge e = INDEX_EDGE (el, i); if (EDGE_CRITICAL_P (e) && !EDGE_INFO (e)->ignore && find_group (e->src) != find_group (e->dest)) { if (dump_file) fprintf (dump_file, "Critical edge %d to %d put to tree\n", e->src->index, e->dest->index); EDGE_INFO (e)->on_tree = 1; union_groups (e->src, e->dest); } } /* And now the rest. */ for (i = 0; i < num_edges; i++) { edge e = INDEX_EDGE (el, i); if (!EDGE_INFO (e)->ignore && find_group (e->src) != find_group (e->dest)) { if (dump_file) fprintf (dump_file, "Normal edge %d to %d put to tree\n", e->src->index, e->dest->index); EDGE_INFO (e)->on_tree = 1; union_groups (e->src, e->dest); } } clear_aux_for_blocks (); } /* Perform file-level initialization for branch-prob processing. */ void init_branch_prob (void) { int i; total_num_blocks = 0; total_num_edges = 0; total_num_edges_ignored = 0; total_num_edges_instrumented = 0; total_num_blocks_created = 0; total_num_passes = 0; total_num_times_called = 0; total_num_branches = 0; for (i = 0; i < 20; i++) total_hist_br_prob[i] = 0; } /* Performs file-level cleanup after branch-prob processing is completed. */ void end_branch_prob (void) { if (dump_file) { fprintf (dump_file, "\n"); fprintf (dump_file, "Total number of blocks: %d\n", total_num_blocks); fprintf (dump_file, "Total number of edges: %d\n", total_num_edges); fprintf (dump_file, "Total number of ignored edges: %d\n", total_num_edges_ignored); fprintf (dump_file, "Total number of instrumented edges: %d\n", total_num_edges_instrumented); fprintf (dump_file, "Total number of blocks created: %d\n", total_num_blocks_created); fprintf (dump_file, "Total number of graph solution passes: %d\n", total_num_passes); if (total_num_times_called != 0) fprintf (dump_file, "Average number of graph solution passes: %d\n", (total_num_passes + (total_num_times_called >> 1)) / total_num_times_called); fprintf (dump_file, "Total number of branches: %d\n", total_num_branches); if (total_num_branches) { int i; for (i = 0; i < 10; i++) fprintf (dump_file, "%d%% branches in range %d-%d%%\n", (total_hist_br_prob[i] + total_hist_br_prob[19-i]) * 100 / total_num_branches, 5*i, 5*i+5); } } }