readstabs.c revision 663860b1408516d02ebfcb3a9999a134e6cfb223 
1
2/*--------------------------------------------------------------------*/
3/*--- Read stabs debug info.                           readstabs.c ---*/
4/*--------------------------------------------------------------------*/
5
6/*
7   This file is part of Valgrind, a dynamic binary instrumentation
8   framework.
9
10   Copyright (C) 2000-2012 Julian Seward
11      jseward@acm.org
12
13   This program is free software; you can redistribute it and/or
14   modify it under the terms of the GNU General Public License as
15   published by the Free Software Foundation; either version 2 of the
16   License, or (at your option) any later version.
17
18   This program is distributed in the hope that it will be useful, but
19   WITHOUT ANY WARRANTY; without even the implied warranty of
20   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
21   General Public License for more details.
22
23   You should have received a copy of the GNU General Public License
24   along with this program; if not, write to the Free Software
25   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
26   02111-1307, USA.
27
28   The GNU General Public License is contained in the file COPYING.
29*/
30
31/*
32   Stabs reader greatly improved by Nick Nethercote, Apr 02.
33   This module was also extensively hacked on by Jeremy Fitzhardinge
34   and Tom Hughes.
35*/
36
37/* "on Linux (except android), or on Darwin" */
38#if (defined(VGO_linux) && \
39    !(defined(VGPV_arm_linux_android) || defined(VGPV_x86_linux_android)) \
40    || defined(VGO_darwin))
41
42#include "pub_core_basics.h"
43#include "pub_core_debuginfo.h"
44#include "pub_core_libcbase.h"
45#include "pub_core_libcassert.h"
46#include "pub_core_libcprint.h"
47#include "pub_core_xarray.h"
48#include "priv_misc.h"             /* dinfo_zalloc/free/strdup */
49#include "priv_tytypes.h"
50#include "priv_d3basics.h"
51#include "priv_storage.h"
52#include "priv_readstabs.h"        /* self */
53
54/* --- !!! --- EXTERNAL HEADERS start --- !!! --- */
55#if defined(VGO_linux)
56#  include <a.out.h> /* stabs defns */
57#elif defined(VGO_darwin)
58#  include <mach-o/nlist.h>
59#  define n_other n_sect
60#  if VG_WORDSIZE == 8
61#     define nlist nlist_64
62#  endif
63#else
64#  error "Unknown OS"
65#endif
66/* --- !!! --- EXTERNAL HEADERS end --- !!! --- */
67
68/*------------------------------------------------------------*/
69/*--- Read STABS format debug info.                        ---*/
70/*------------------------------------------------------------*/
71
72/* Stabs entry types, from:
73 *   The "stabs" debug format
74 *   Menapace, Kingdon and MacKenzie
75 *   Cygnus Support
76 */
77typedef enum { N_UNDEF = 0,	/* undefined symbol, new stringtab  */
78	       N_GSYM  = 32,    /* Global symbol                    */
79               N_FUN   = 36,    /* Function start or end            */
80               N_STSYM = 38,    /* Data segment file-scope variable */
81               N_LCSYM = 40,    /* BSS segment file-scope variable  */
82               N_RSYM  = 64,    /* Register variable                */
83               N_SLINE = 68,    /* Source line number               */
84               N_SO    = 100,   /* Source file path and name        */
85               N_LSYM  = 128,   /* Stack variable or type           */
86	       N_BINCL = 130,	/* Beginning of an include file	    */
87               N_SOL   = 132,   /* Include file name                */
88	       N_PSYM  = 160,   /* Function parameter               */
89	       N_EINCL = 162,	/* End of an include file           */
90               N_LBRAC = 192,   /* Start of lexical block           */
91	       N_EXCL  = 194,	/* Placeholder for an include file  */
92               N_RBRAC = 224    /* End   of lexical block           */
93             } stab_types;
94
95
96/* Read stabs-format debug info.  This is all rather horrible because
97   stabs is a underspecified, kludgy hack.
98*/
99void ML_(read_debuginfo_stabs) ( DebugInfo* di,
100                                 UChar* stabC,   Int stab_sz,
101                                 UChar* stabstr, Int stabstr_sz )
102{
103   const Bool debug     = False;
104   const Bool contdebug = False;
105   Int    i;
106   Int    n_stab_entries;
107   struct nlist* stab = (struct nlist*)stabC;
108   UChar *next_stabstr = NULL;
109   /* state for various things */
110   struct {
111      Addr     start;         /* start address */
112      Addr     end;           /* end address */
113      Int      line;          /* first line */
114   } func = { 0, 0, -1 };
115   struct {
116      Char     *name;
117      Bool     same;
118   } file = { NULL, True };
119   struct {
120      Int      prev;          /* prev line */
121      Int      no;            /* current line */
122      Int      ovf;           /* line wrap */
123      Addr     addr;          /* start of this line */
124      Bool     first;         /* first line in function */
125   } line = { 0, 0, 0, 0, False };
126
127   /* Ok.  It all looks plausible.  Go on and read debug data.
128         stab kinds: 100   N_SO     a source file name
129                      68   N_SLINE  a source line number
130                      36   N_FUN    start of a function
131
132      In this loop, we maintain a current file name, updated as
133      N_SO/N_SOLs appear, and a current function base address,
134      updated as N_FUNs appear.  Based on that, address ranges for
135      N_SLINEs are calculated, and stuffed into the line info table.
136
137      Finding the instruction address range covered by an N_SLINE is
138      complicated;  see the N_SLINE case below.
139   */
140   file.name     = ML_(addStr)(di,"???", -1);
141
142   n_stab_entries = stab_sz/(int)sizeof(struct nlist);
143
144   for (i = 0; i < n_stab_entries; i++) {
145      const struct nlist *st = &stab[i];
146      Char *string;
147
148      if (di->trace_symtab) {
149         VG_(printf) ( "%2d  type=%d   othr=%d   desc=%d   "
150                       "value=0x%x   strx=%d  %s\n", i,
151                       st->n_type, st->n_other, st->n_desc,
152                       (Int)st->n_value,
153                       (Int)st->n_un.n_strx,
154                       stabstr + st->n_un.n_strx );
155      }
156
157      /* handle continued string stabs */
158      {
159         Int   qbuflen = 0;
160         Int   qidx = 0;
161         Char* qbuf = NULL;
162         Int   qlen;
163         Bool  qcontinuing = False;
164         UInt  qstringidx;
165
166         qstringidx = st->n_un.n_strx;
167         string = stabstr + qstringidx;
168         qlen = VG_(strlen)(string);
169
170         while (string
171                && qlen > 0
172                && (qcontinuing || string[qlen-1] == '\\')) {
173            /* Gak, we have a continuation. Skip forward through
174               subsequent stabs to gather all the parts of the
175               continuation.  Increment i, but keep st pointing at
176               current stab. */
177
178            qcontinuing = string[qlen-1] == '\\';
179
180            /* remove trailing \ */
181            while (string[qlen-1] == '\\' && qlen > 0)
182               qlen--;
183
184            if (contdebug)
185               VG_(printf)("found extension string: \"%s\" "
186                           "len=%d(%c) idx=%d buflen=%d\n",
187                           string, qlen, string[qlen-1], qidx, qbuflen);
188
189            /* XXX this is silly.  The si->strtab should have a way of
190               appending to the last added string... */
191            if ((qidx + qlen) >= qbuflen) {
192               Char *n;
193
194               if (qbuflen == 0)
195                  qbuflen = 16;
196               while ((qidx + qlen) >= qbuflen)
197                  qbuflen *= 2;
198               n = ML_(dinfo_zalloc)("di.readstabs.rds.1", qbuflen);
199               VG_(memcpy)(n, qbuf, qidx);
200
201               if (qbuf != NULL)
202                  ML_(dinfo_free)(qbuf);
203               qbuf = n;
204            }
205
206            VG_(memcpy)(&qbuf[qidx], string, qlen);
207            qidx += qlen;
208            if (contdebug) {
209               qbuf[qidx] = '\0';
210               VG_(printf)("working buf=\"%s\"\n", qbuf);
211            }
212
213            i++;
214            if (i >= n_stab_entries)
215               break;
216
217            if (stab[i].n_un.n_strx) {
218               string = stabstr + stab[i].n_un.n_strx;
219               qlen = VG_(strlen)(string);
220            } else {
221               string = NULL;
222               qlen = 0;
223            }
224         }
225
226         if (qbuf != NULL) {
227            i--;                        /* overstepped */
228            string = ML_(addStr)(di, qbuf, qidx);
229            ML_(dinfo_free)(qbuf);
230            if (contdebug)
231               VG_(printf)("made composite: \"%s\"\n", string);
232         }
233      }
234
235      switch(st->n_type) {
236         case N_UNDEF:
237            /* new string table base */
238            if (next_stabstr != NULL) {
239               stabstr_sz -= next_stabstr - stabstr;
240               stabstr = next_stabstr;
241               if (stabstr_sz <= 0) {
242                  VG_(printf)(" @@ bad stabstr size %d\n", stabstr_sz);
243                  return;
244               }
245            }
246            next_stabstr = stabstr + st->n_value;
247            break;
248
249         case N_BINCL: {
250            break;
251         }
252
253         case N_EINCL:
254            break;
255
256         case N_EXCL:
257            break;
258
259         case N_SOL:                /* sub-source (include) file */
260            if (line.ovf != 0)
261               VG_(message)(Vg_UserMsg,
262                            "Warning: file %s is very big (> 65535 lines) "
263                            "Line numbers and annotation for this file might "
264                            "be wrong.  Sorry.\n",
265                            file.name);
266            /* FALLTHROUGH */
267
268         case N_SO: {                /* new source file */
269            UChar *nm = string;
270            UInt len = VG_(strlen)(nm);
271            Addr addr = func.start + st->n_value;
272
273            if (line.addr != 0) {
274               /* finish off previous line */
275               ML_(addLineInfo)(di, file.name, NULL, line.addr,
276                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
277            }
278
279            /* reset line state */
280            line.ovf = 0;
281            line.addr = 0;
282            line.prev = 0;
283            line.no = 0;
284
285            if (len > 0 && nm[len-1] != '/') {
286               file.name = ML_(addStr)(di, nm, -1);
287               if (debug)
288                  VG_(printf)("new source: %s\n", file.name);
289            } else if (len == 0)
290               file.name = ML_(addStr)(di, "?1\0", -1);
291
292            break;
293         }
294
295         case N_SLINE: {        /* line info */
296            Addr addr = func.start + st->n_value;
297
298            if (line.addr != 0) {
299               /* there was a previous */
300               ML_(addLineInfo)(di, file.name, NULL, line.addr,
301                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
302            }
303
304            line.addr = addr;
305            line.prev = line.no;
306            line.no = (Int)((UShort)st->n_desc);
307
308            if (line.prev > line.no + OVERFLOW_DIFFERENCE && file.same) {
309               VG_(message)(Vg_DebugMsg,
310                  "Line number overflow detected (%d --> %d) in %s\n",
311                  line.prev, line.no, file.name);
312               line.ovf++;
313            }
314            file.same = True;
315
316            /* This is pretty horrible.  If this is the first line of
317               the function, then bind any unbound symbols to the arg
318               scope, since they're probably arguments. */
319            if (line.first) {
320               line.first = False;
321
322               /* remember first line of function */
323               if (func.start != 0) {
324                  func.line = line.no;
325               }
326            }
327            break;
328         }
329
330         case N_FUN: {                /* function start/end */
331            Addr addr = 0;        /* end address for prev line/scope */
332
333            /* if this the end of the function or we haven't
334               previously finished the previous function... */
335            if (*string == '\0' || func.start != 0) {
336               /* end of function */
337               line.first = False;
338
339               /* end line at end of function */
340               addr = func.start + st->n_value;
341
342               /* now between functions */
343               func.start = 0;
344
345               // XXXX DEAD POINT XXXX
346            }
347
348            if (*string != '\0') {
349               /* new function */
350               line.first = True;
351
352               /* line ends at start of next function */
353               addr = di->text_debug_bias + st->n_value;
354
355               func.start = addr;
356            }
357
358            if (line.addr) {
359               ML_(addLineInfo)(di, file.name, NULL, line.addr,
360                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
361               line.addr = 0;
362            }
363
364            //DEAD POINT
365            //DEAD POINT
366            break;
367         }
368
369         case N_LBRAC: {
370            /* open new scope */
371            // DEAD POINT
372            break;
373         }
374
375         case N_RBRAC: {
376            /* close scope */
377            // DEAD POINT
378            break;
379         }
380
381         case N_GSYM:                /* global variable */
382         case N_STSYM:                /* static in data segment */
383         case N_LCSYM:                /* static in bss segment */
384         case N_PSYM:                /* function parameter */
385         case N_LSYM:                /* stack variable */
386         case N_RSYM:                  /* register variable */
387            break;
388      }
389   }
390}
391
392#endif /* (defined(VGO_linux) && !defined(VGPV_*_linux_android)) \
393          || defined(VGO_darwin) */
394
395/*--------------------------------------------------------------------*/
396/*--- end                                                          ---*/
397/*--------------------------------------------------------------------*/
398