1 /* Generic BFD support for file formats.
2    Copyright (C) 1990-2016 Free Software Foundation, Inc.
3    Written by Cygnus Support.
4 
5    This file is part of BFD, the Binary File Descriptor library.
6 
7    This program is free software; you can redistribute it and/or modify
8    it under the terms of the GNU General Public License as published by
9    the Free Software Foundation; either version 3 of the License, or
10    (at your option) any later version.
11 
12    This program is distributed in the hope that it will be useful,
13    but WITHOUT ANY WARRANTY; without even the implied warranty of
14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
15    GNU General Public License for more details.
16 
17    You should have received a copy of the GNU General Public License
18    along with this program; if not, write to the Free Software
19    Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston,
20    MA 02110-1301, USA.  */
21 
22 
23 /*
24 SECTION
25 	File formats
26 
27 	A format is a BFD concept of high level file contents type. The
28 	formats supported by BFD are:
29 
30 	o <<bfd_object>>
31 
32 	The BFD may contain data, symbols, relocations and debug info.
33 
34 	o <<bfd_archive>>
35 
36 	The BFD contains other BFDs and an optional index.
37 
38 	o <<bfd_core>>
39 
40 	The BFD contains the result of an executable core dump.
41 
42 SUBSECTION
43 	File format functions
44 */
45 
46 #include "sysdep.h"
47 #include "bfd.h"
48 #include "libbfd.h"
49 
50 /* IMPORT from targets.c.  */
51 extern const size_t _bfd_target_vector_entries;
52 
53 /*
54 FUNCTION
55 	bfd_check_format
56 
57 SYNOPSIS
58 	bfd_boolean bfd_check_format (bfd *abfd, bfd_format format);
59 
60 DESCRIPTION
61 	Verify if the file attached to the BFD @var{abfd} is compatible
62 	with the format @var{format} (i.e., one of <<bfd_object>>,
63 	<<bfd_archive>> or <<bfd_core>>).
64 
65 	If the BFD has been set to a specific target before the
66 	call, only the named target and format combination is
67 	checked. If the target has not been set, or has been set to
68 	<<default>>, then all the known target backends is
69 	interrogated to determine a match.  If the default target
70 	matches, it is used.  If not, exactly one target must recognize
71 	the file, or an error results.
72 
73 	The function returns <<TRUE>> on success, otherwise <<FALSE>>
74 	with one of the following error codes:
75 
76 	o <<bfd_error_invalid_operation>> -
77 	if <<format>> is not one of <<bfd_object>>, <<bfd_archive>> or
78 	<<bfd_core>>.
79 
80 	o <<bfd_error_system_call>> -
81 	if an error occured during a read - even some file mismatches
82 	can cause bfd_error_system_calls.
83 
84 	o <<file_not_recognised>> -
85 	none of the backends recognised the file format.
86 
87 	o <<bfd_error_file_ambiguously_recognized>> -
88 	more than one backend recognised the file format.
89 */
90 
91 bfd_boolean
92 bfd_check_format (bfd *abfd, bfd_format format)
93 {
94   return bfd_check_format_matches (abfd, format, NULL);
95 }
96 
97 struct bfd_preserve
98 {
99   void *marker;
100   void *tdata;
101   flagword flags;
102   const struct bfd_arch_info *arch_info;
103   struct bfd_section *sections;
104   struct bfd_section *section_last;
105   unsigned int section_count;
106   struct bfd_hash_table section_htab;
107 };
108 
109 /* When testing an object for compatibility with a particular target
110    back-end, the back-end object_p function needs to set up certain
111    fields in the bfd on successfully recognizing the object.  This
112    typically happens in a piecemeal fashion, with failures possible at
113    many points.  On failure, the bfd is supposed to be restored to its
114    initial state, which is virtually impossible.  However, restoring a
115    subset of the bfd state works in practice.  This function stores
116    the subset.  */
117 
118 static bfd_boolean
119 bfd_preserve_save (bfd *abfd, struct bfd_preserve *preserve)
120 {
121   preserve->tdata = abfd->tdata.any;
122   preserve->arch_info = abfd->arch_info;
123   preserve->flags = abfd->flags;
124   preserve->sections = abfd->sections;
125   preserve->section_last = abfd->section_last;
126   preserve->section_count = abfd->section_count;
127   preserve->section_htab = abfd->section_htab;
128   preserve->marker = bfd_alloc (abfd, 1);
129   if (preserve->marker == NULL)
130     return FALSE;
131 
132   return bfd_hash_table_init (&abfd->section_htab, bfd_section_hash_newfunc,
133 			      sizeof (struct section_hash_entry));
134 }
135 
136 /* Clear out a subset of BFD state.  */
137 
138 static void
139 bfd_reinit (bfd *abfd)
140 {
141   abfd->tdata.any = NULL;
142   abfd->arch_info = &bfd_default_arch_struct;
143   abfd->flags &= BFD_FLAGS_SAVED;
144   bfd_section_list_clear (abfd);
145 }
146 
147 /* Restores bfd state saved by bfd_preserve_save.  */
148 
149 static void
150 bfd_preserve_restore (bfd *abfd, struct bfd_preserve *preserve)
151 {
152   bfd_hash_table_free (&abfd->section_htab);
153 
154   abfd->tdata.any = preserve->tdata;
155   abfd->arch_info = preserve->arch_info;
156   abfd->flags = preserve->flags;
157   abfd->section_htab = preserve->section_htab;
158   abfd->sections = preserve->sections;
159   abfd->section_last = preserve->section_last;
160   abfd->section_count = preserve->section_count;
161 
162   /* bfd_release frees all memory more recently bfd_alloc'd than
163      its arg, as well as its arg.  */
164   bfd_release (abfd, preserve->marker);
165   preserve->marker = NULL;
166 }
167 
168 /* Called when the bfd state saved by bfd_preserve_save is no longer
169    needed.  */
170 
171 static void
172 bfd_preserve_finish (bfd *abfd ATTRIBUTE_UNUSED, struct bfd_preserve *preserve)
173 {
174   /* It would be nice to be able to free more memory here, eg. old
175      tdata, but that's not possible since these blocks are sitting
176      inside bfd_alloc'd memory.  The section hash is on a separate
177      objalloc.  */
178   bfd_hash_table_free (&preserve->section_htab);
179   preserve->marker = NULL;
180 }
181 
182 /*
183 FUNCTION
184 	bfd_check_format_matches
185 
186 SYNOPSIS
187 	bfd_boolean bfd_check_format_matches
188 	  (bfd *abfd, bfd_format format, char ***matching);
189 
190 DESCRIPTION
191 	Like <<bfd_check_format>>, except when it returns FALSE with
192 	<<bfd_errno>> set to <<bfd_error_file_ambiguously_recognized>>.  In that
193 	case, if @var{matching} is not NULL, it will be filled in with
194 	a NULL-terminated list of the names of the formats that matched,
195 	allocated with <<malloc>>.
196 	Then the user may choose a format and try again.
197 
198 	When done with the list that @var{matching} points to, the caller
199 	should free it.
200 */
201 
202 bfd_boolean
203 bfd_check_format_matches (bfd *abfd, bfd_format format, char ***matching)
204 {
205   extern const bfd_target binary_vec;
206 #if BFD_SUPPORTS_PLUGINS
207   extern const bfd_target plugin_vec;
208 #endif
209   const bfd_target * const *target;
210   const bfd_target **matching_vector = NULL;
211   const bfd_target *save_targ, *right_targ, *ar_right_targ, *match_targ;
212   int match_count, best_count, best_match;
213   int ar_match_index;
214   struct bfd_preserve preserve;
215 
216   if (matching != NULL)
217     *matching = NULL;
218 
219   if (!bfd_read_p (abfd)
220       || (unsigned int) abfd->format >= (unsigned int) bfd_type_end)
221     {
222       bfd_set_error (bfd_error_invalid_operation);
223       return FALSE;
224     }
225 
226   if (abfd->format != bfd_unknown)
227     return abfd->format == format;
228 
229   if (matching != NULL || *bfd_associated_vector != NULL)
230     {
231       bfd_size_type amt;
232 
233       amt = sizeof (*matching_vector) * 2 * _bfd_target_vector_entries;
234       matching_vector = (const bfd_target **) bfd_malloc (amt);
235       if (!matching_vector)
236 	return FALSE;
237     }
238 
239   /* Presume the answer is yes.  */
240   abfd->format = format;
241   save_targ = abfd->xvec;
242   preserve.marker = NULL;
243 
244   /* If the target type was explicitly specified, just check that target.  */
245   if (!abfd->target_defaulted)
246     {
247       if (bfd_seek (abfd, (file_ptr) 0, SEEK_SET) != 0)	/* rewind! */
248 	goto err_ret;
249 
250       right_targ = BFD_SEND_FMT (abfd, _bfd_check_format, (abfd));
251 
252       if (right_targ)
253 	goto ok_ret;
254 
255       /* For a long time the code has dropped through to check all
256 	 targets if the specified target was wrong.  I don't know why,
257 	 and I'm reluctant to change it.  However, in the case of an
258 	 archive, it can cause problems.  If the specified target does
259 	 not permit archives (e.g., the binary target), then we should
260 	 not allow some other target to recognize it as an archive, but
261 	 should instead allow the specified target to recognize it as an
262 	 object.  When I first made this change, it broke the PE target,
263 	 because the specified pei-i386 target did not recognize the
264 	 actual pe-i386 archive.  Since there may be other problems of
265 	 this sort, I changed this test to check only for the binary
266 	 target.  */
267       if (format == bfd_archive && save_targ == &binary_vec)
268 	goto err_unrecog;
269     }
270 
271   /* Since the target type was defaulted, check them all in the hope
272      that one will be uniquely recognized.  */
273   right_targ = NULL;
274   ar_right_targ = NULL;
275   match_targ = NULL;
276   best_match = 256;
277   best_count = 0;
278   match_count = 0;
279   ar_match_index = _bfd_target_vector_entries;
280 
281   for (target = bfd_target_vector; *target != NULL; target++)
282     {
283       const bfd_target *temp;
284 
285       /* Don't check the default target twice.  */
286       if (*target == &binary_vec
287 	  || (!abfd->target_defaulted && *target == save_targ)
288 	  || (*target)->match_priority > best_match)
289 	continue;
290 
291       /* If we already tried a match, the bfd is modified and may
292 	 have sections attached, which will confuse the next
293 	 _bfd_check_format call.  */
294       bfd_reinit (abfd);
295 
296       /* Change BFD's target temporarily.  */
297       abfd->xvec = *target;
298 
299       if (bfd_seek (abfd, (file_ptr) 0, SEEK_SET) != 0)
300 	goto err_ret;
301 
302       /* If _bfd_check_format neglects to set bfd_error, assume
303 	 bfd_error_wrong_format.  We didn't used to even pay any
304 	 attention to bfd_error, so I suspect that some
305 	 _bfd_check_format might have this problem.  */
306       bfd_set_error (bfd_error_wrong_format);
307 
308       temp = BFD_SEND_FMT (abfd, _bfd_check_format, (abfd));
309       if (temp)
310 	{
311 	  int match_priority = temp->match_priority;
312 #if BFD_SUPPORTS_PLUGINS
313 	  /* If this object can be handled by a plugin, give that the
314 	     lowest priority; objects both handled by a plugin and
315 	     with an underlying object format will be claimed
316 	     separately by the plugin.  */
317 	  if (*target == &plugin_vec)
318 	    match_priority = (*target)->match_priority;
319 #endif
320 
321 	  match_targ = temp;
322 	  if (preserve.marker != NULL)
323 	    bfd_preserve_finish (abfd, &preserve);
324 
325 	  if (abfd->format != bfd_archive
326 	      || (bfd_has_map (abfd)
327 		  && bfd_get_error () != bfd_error_wrong_object_format))
328 	    {
329 	      /* This format checks out as ok!  */
330 	      right_targ = temp;
331 
332 	      /* If this is the default target, accept it, even if
333 		 other targets might match.  People who want those
334 		 other targets have to set the GNUTARGET variable.  */
335 	      if (temp == bfd_default_vector[0])
336 		goto ok_ret;
337 
338 	      if (matching_vector)
339 		matching_vector[match_count] = temp;
340 	      match_count++;
341 
342 	      if (match_priority < best_match)
343 		{
344 		  best_match = match_priority;
345 		  best_count = 0;
346 		}
347 	      best_count++;
348 	    }
349 	  else
350 	    {
351 	      /* An archive with no armap or objects of the wrong
352 		 type.  We want this target to match if we get no
353 		 better matches.  */
354 	      if (ar_right_targ != bfd_default_vector[0])
355 		ar_right_targ = *target;
356 	      if (matching_vector)
357 		matching_vector[ar_match_index] = *target;
358 	      ar_match_index++;
359 	    }
360 
361 	  if (!bfd_preserve_save (abfd, &preserve))
362 	    goto err_ret;
363 	}
364       else if (bfd_get_error () != bfd_error_wrong_format)
365 	goto err_ret;
366     }
367 
368   if (best_count == 1)
369     match_count = 1;
370 
371   if (match_count == 0)
372     {
373       /* Try partial matches.  */
374       right_targ = ar_right_targ;
375 
376       if (right_targ == bfd_default_vector[0])
377 	{
378 	  match_count = 1;
379 	}
380       else
381 	{
382 	  match_count = ar_match_index - _bfd_target_vector_entries;
383 
384 	  if (matching_vector && match_count > 1)
385 	    memcpy (matching_vector,
386 		    matching_vector + _bfd_target_vector_entries,
387 		    sizeof (*matching_vector) * match_count);
388 	}
389     }
390 
391   /* We have more than one equally good match.  If any of the best
392      matches is a target in config.bfd targ_defvec or targ_selvecs,
393      choose it.  */
394   if (match_count > 1)
395     {
396       const bfd_target * const *assoc = bfd_associated_vector;
397 
398       while ((right_targ = *assoc++) != NULL)
399 	{
400 	  int i = match_count;
401 
402 	  while (--i >= 0)
403 	    if (matching_vector[i] == right_targ
404 		&& right_targ->match_priority <= best_match)
405 	      break;
406 
407 	  if (i >= 0)
408 	    {
409 	      match_count = 1;
410 	      break;
411 	    }
412 	}
413     }
414 
415   /* We still have more than one equally good match, and at least some
416      of the targets support match priority.  Choose the first of the
417      best matches.  */
418   if (matching_vector && match_count > 1 && best_count != match_count)
419     {
420       int i;
421 
422       for (i = 0; i < match_count; i++)
423 	{
424 	  right_targ = matching_vector[i];
425 	  if (right_targ->match_priority <= best_match)
426 	    break;
427 	}
428       match_count = 1;
429     }
430 
431   /* There is way too much undoing of half-known state here.  We
432      really shouldn't iterate on live bfd's.  Note that saving the
433      whole bfd and restoring it would be even worse; the first thing
434      you notice is that the cached bfd file position gets out of sync.  */
435   if (preserve.marker != NULL)
436     bfd_preserve_restore (abfd, &preserve);
437 
438   if (match_count == 1)
439     {
440       abfd->xvec = right_targ;
441       /* If we come out of the loop knowing that the last target that
442 	 matched is the one we want, then ABFD should still be in a usable
443 	 state (except possibly for XVEC).  */
444       if (match_targ != right_targ)
445 	{
446 	  bfd_reinit (abfd);
447 	  if (bfd_seek (abfd, (file_ptr) 0, SEEK_SET) != 0)
448 	    goto err_ret;
449 	  match_targ = BFD_SEND_FMT (abfd, _bfd_check_format, (abfd));
450 	  BFD_ASSERT (match_targ != NULL);
451 	}
452 
453     ok_ret:
454       /* If the file was opened for update, then `output_has_begun'
455 	 some time ago when the file was created.  Do not recompute
456 	 sections sizes or alignments in _bfd_set_section_contents.
457 	 We can not set this flag until after checking the format,
458 	 because it will interfere with creation of BFD sections.  */
459       if (abfd->direction == both_direction)
460 	abfd->output_has_begun = TRUE;
461 
462       if (matching_vector)
463 	free (matching_vector);
464 
465       /* File position has moved, BTW.  */
466       return TRUE;
467     }
468 
469   if (match_count == 0)
470     {
471     err_unrecog:
472       bfd_set_error (bfd_error_file_not_recognized);
473     err_ret:
474       abfd->xvec = save_targ;
475       abfd->format = bfd_unknown;
476       if (matching_vector)
477 	free (matching_vector);
478       if (preserve.marker != NULL)
479 	bfd_preserve_restore (abfd, &preserve);
480       return FALSE;
481     }
482 
483   /* Restore original target type and format.  */
484   abfd->xvec = save_targ;
485   abfd->format = bfd_unknown;
486   bfd_set_error (bfd_error_file_ambiguously_recognized);
487 
488   if (matching)
489     {
490       *matching = (char **) matching_vector;
491       matching_vector[match_count] = NULL;
492       /* Return target names.  This is a little nasty.  Maybe we
493 	 should do another bfd_malloc?  */
494       while (--match_count >= 0)
495 	{
496 	  const char *name = matching_vector[match_count]->name;
497 	  *(const char **) &matching_vector[match_count] = name;
498 	}
499     }
500   return FALSE;
501 }
502 
503 /*
504 FUNCTION
505 	bfd_set_format
506 
507 SYNOPSIS
508 	bfd_boolean bfd_set_format (bfd *abfd, bfd_format format);
509 
510 DESCRIPTION
511 	This function sets the file format of the BFD @var{abfd} to the
512 	format @var{format}. If the target set in the BFD does not
513 	support the format requested, the format is invalid, or the BFD
514 	is not open for writing, then an error occurs.
515 */
516 
517 bfd_boolean
518 bfd_set_format (bfd *abfd, bfd_format format)
519 {
520   if (bfd_read_p (abfd)
521       || (unsigned int) abfd->format >= (unsigned int) bfd_type_end)
522     {
523       bfd_set_error (bfd_error_invalid_operation);
524       return FALSE;
525     }
526 
527   if (abfd->format != bfd_unknown)
528     return abfd->format == format;
529 
530   /* Presume the answer is yes.  */
531   abfd->format = format;
532 
533   if (!BFD_SEND_FMT (abfd, _bfd_set_format, (abfd)))
534     {
535       abfd->format = bfd_unknown;
536       return FALSE;
537     }
538 
539   return TRUE;
540 }
541 
542 /*
543 FUNCTION
544 	bfd_format_string
545 
546 SYNOPSIS
547 	const char *bfd_format_string (bfd_format format);
548 
549 DESCRIPTION
550 	Return a pointer to a const string
551 	<<invalid>>, <<object>>, <<archive>>, <<core>>, or <<unknown>>,
552 	depending upon the value of @var{format}.
553 */
554 
555 const char *
556 bfd_format_string (bfd_format format)
557 {
558   if (((int) format < (int) bfd_unknown)
559       || ((int) format >= (int) bfd_type_end))
560     return "invalid";
561 
562   switch (format)
563     {
564     case bfd_object:
565       return "object";		/* Linker/assembler/compiler output.  */
566     case bfd_archive:
567       return "archive";		/* Object archive file.  */
568     case bfd_core:
569       return "core";		/* Core dump.  */
570     default:
571       return "unknown";
572     }
573 }
574