readstabs.c revision b32f58018498ea2225959b0ba11c18f0c433deef 
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) 2000-2011 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) && !defined(VGPV_arm_linux_android)) \
39    || defined(VGO_darwin)
40
41#include "pub_core_basics.h"
42#include "pub_core_debuginfo.h"
43#include "pub_core_libcbase.h"
44#include "pub_core_libcassert.h"
45#include "pub_core_libcprint.h"
46#include "pub_core_xarray.h"
47#include "priv_misc.h"             /* dinfo_zalloc/free/strdup */
48#include "priv_tytypes.h"
49#include "priv_d3basics.h"
50#include "priv_storage.h"
51#include "priv_readstabs.h"        /* self */
52
53/* --- !!! --- EXTERNAL HEADERS start --- !!! --- */
54#if defined(VGO_linux)
55#  include <a.out.h> /* stabs defns */
56#elif defined(VGO_darwin)
57#  include <mach-o/nlist.h>
58#  define n_other n_sect
59#  if VG_WORDSIZE == 8
60#     define nlist nlist_64
61#  endif
62#else
63#  error "Unknown OS"
64#endif
65/* --- !!! --- EXTERNAL HEADERS end --- !!! --- */
66
67/*------------------------------------------------------------*/
68/*--- Read STABS format debug info.                        ---*/
69/*------------------------------------------------------------*/
70
71/* Stabs entry types, from:
72 *   The "stabs" debug format
73 *   Menapace, Kingdon and MacKenzie
74 *   Cygnus Support
75 */
76typedef enum { N_UNDEF = 0,	/* undefined symbol, new stringtab  */
77	       N_GSYM  = 32,    /* Global symbol                    */
78               N_FUN   = 36,    /* Function start or end            */
79               N_STSYM = 38,    /* Data segment file-scope variable */
80               N_LCSYM = 40,    /* BSS segment file-scope variable  */
81               N_RSYM  = 64,    /* Register variable                */
82               N_SLINE = 68,    /* Source line number               */
83               N_SO    = 100,   /* Source file path and name        */
84               N_LSYM  = 128,   /* Stack variable or type           */
85	       N_BINCL = 130,	/* Beginning of an include file	    */
86               N_SOL   = 132,   /* Include file name                */
87	       N_PSYM  = 160,   /* Function parameter               */
88	       N_EINCL = 162,	/* End of an include file           */
89               N_LBRAC = 192,   /* Start of lexical block           */
90	       N_EXCL  = 194,	/* Placeholder for an include file  */
91               N_RBRAC = 224    /* End   of lexical block           */
92             } stab_types;
93
94
95/* Read stabs-format debug info.  This is all rather horrible because
96   stabs is a underspecified, kludgy hack.
97*/
98void ML_(read_debuginfo_stabs) ( DebugInfo* di,
99                                 UChar* stabC,   Int stab_sz,
100                                 UChar* stabstr, Int stabstr_sz )
101{
102   const Bool debug     = False;
103   const Bool contdebug = False;
104   Int    i;
105   Int    n_stab_entries;
106   struct nlist* stab = (struct nlist*)stabC;
107   UChar *next_stabstr = NULL;
108   /* state for various things */
109   struct {
110      Addr     start;         /* start address */
111      Addr     end;           /* end address */
112      Int      line;          /* first line */
113   } func = { 0, 0, -1 };
114   struct {
115      Char     *name;
116      Bool     same;
117   } file = { NULL, True };
118   struct {
119      Int      prev;          /* prev line */
120      Int      no;            /* current line */
121      Int      ovf;           /* line wrap */
122      Addr     addr;          /* start of this line */
123      Bool     first;         /* first line in function */
124   } line = { 0, 0, 0, 0, False };
125
126   /* Ok.  It all looks plausible.  Go on and read debug data.
127         stab kinds: 100   N_SO     a source file name
128                      68   N_SLINE  a source line number
129                      36   N_FUN    start of a function
130
131      In this loop, we maintain a current file name, updated as
132      N_SO/N_SOLs appear, and a current function base address,
133      updated as N_FUNs appear.  Based on that, address ranges for
134      N_SLINEs are calculated, and stuffed into the line info table.
135
136      Finding the instruction address range covered by an N_SLINE is
137      complicated;  see the N_SLINE case below.
138   */
139   file.name     = ML_(addStr)(di,"???", -1);
140
141   n_stab_entries = stab_sz/(int)sizeof(struct nlist);
142
143   for (i = 0; i < n_stab_entries; i++) {
144      const struct nlist *st = &stab[i];
145      Char *string;
146
147      if (di->trace_symtab) {
148         VG_(printf) ( "%2d  type=%d   othr=%d   desc=%d   "
149                       "value=0x%x   strx=%d  %s\n", i,
150                       st->n_type, st->n_other, st->n_desc,
151                       (Int)st->n_value,
152                       (Int)st->n_un.n_strx,
153                       stabstr + st->n_un.n_strx );
154      }
155
156      /* handle continued string stabs */
157      {
158         Int   qbuflen = 0;
159         Int   qidx = 0;
160         Char* qbuf = NULL;
161         Int   qlen;
162         Bool  qcontinuing = False;
163         UInt  qstringidx;
164
165         qstringidx = st->n_un.n_strx;
166         string = stabstr + qstringidx;
167         qlen = VG_(strlen)(string);
168
169         while (string
170                && qlen > 0
171                && (qcontinuing || string[qlen-1] == '\\')) {
172            /* Gak, we have a continuation. Skip forward through
173               subsequent stabs to gather all the parts of the
174               continuation.  Increment i, but keep st pointing at
175               current stab. */
176
177            qcontinuing = string[qlen-1] == '\\';
178
179            /* remove trailing \ */
180            while (string[qlen-1] == '\\' && qlen > 0)
181               qlen--;
182
183            if (contdebug)
184               VG_(printf)("found extension string: \"%s\" "
185                           "len=%d(%c) idx=%d buflen=%d\n",
186                           string, qlen, string[qlen-1], qidx, qbuflen);
187
188            /* XXX this is silly.  The si->strtab should have a way of
189               appending to the last added string... */
190            if ((qidx + qlen) >= qbuflen) {
191               Char *n;
192
193               if (qbuflen == 0)
194                  qbuflen = 16;
195               while ((qidx + qlen) >= qbuflen)
196                  qbuflen *= 2;
197               n = ML_(dinfo_zalloc)("di.readstabs.rds.1", qbuflen);
198               VG_(memcpy)(n, qbuf, qidx);
199
200               if (qbuf != NULL)
201                  ML_(dinfo_free)(qbuf);
202               qbuf = n;
203            }
204
205            VG_(memcpy)(&qbuf[qidx], string, qlen);
206            qidx += qlen;
207            if (contdebug) {
208               qbuf[qidx] = '\0';
209               VG_(printf)("working buf=\"%s\"\n", qbuf);
210            }
211
212            i++;
213            if (i >= n_stab_entries)
214               break;
215
216            if (stab[i].n_un.n_strx) {
217               string = stabstr + stab[i].n_un.n_strx;
218               qlen = VG_(strlen)(string);
219            } else {
220               string = NULL;
221               qlen = 0;
222            }
223         }
224
225         if (qbuf != NULL) {
226            i--;                        /* overstepped */
227            string = ML_(addStr)(di, qbuf, qidx);
228            ML_(dinfo_free)(qbuf);
229            if (contdebug)
230               VG_(printf)("made composite: \"%s\"\n", string);
231         }
232      }
233
234      switch(st->n_type) {
235         case N_UNDEF:
236            /* new string table base */
237            if (next_stabstr != NULL) {
238               stabstr_sz -= next_stabstr - stabstr;
239               stabstr = next_stabstr;
240               if (stabstr_sz <= 0) {
241                  VG_(printf)(" @@ bad stabstr size %d\n", stabstr_sz);
242                  return;
243               }
244            }
245            next_stabstr = stabstr + st->n_value;
246            break;
247
248         case N_BINCL: {
249            break;
250         }
251
252         case N_EINCL:
253            break;
254
255         case N_EXCL:
256            break;
257
258         case N_SOL:                /* sub-source (include) file */
259            if (line.ovf != 0)
260               VG_(message)(Vg_UserMsg,
261                            "Warning: file %s is very big (> 65535 lines) "
262                            "Line numbers and annotation for this file might "
263                            "be wrong.  Sorry.\n",
264                            file.name);
265            /* FALLTHROUGH */
266
267         case N_SO: {                /* new source file */
268            UChar *nm = string;
269            UInt len = VG_(strlen)(nm);
270            Addr addr = func.start + st->n_value;
271
272            if (line.addr != 0) {
273               /* finish off previous line */
274               ML_(addLineInfo)(di, file.name, NULL, line.addr,
275                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
276            }
277
278            /* reset line state */
279            line.ovf = 0;
280            line.addr = 0;
281            line.prev = 0;
282            line.no = 0;
283
284            if (len > 0 && nm[len-1] != '/') {
285               file.name = ML_(addStr)(di, nm, -1);
286               if (debug)
287                  VG_(printf)("new source: %s\n", file.name);
288            } else if (len == 0)
289               file.name = ML_(addStr)(di, "?1\0", -1);
290
291            break;
292         }
293
294         case N_SLINE: {        /* line info */
295            Addr addr = func.start + st->n_value;
296
297            if (line.addr != 0) {
298               /* there was a previous */
299               ML_(addLineInfo)(di, file.name, NULL, line.addr,
300                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
301            }
302
303            line.addr = addr;
304            line.prev = line.no;
305            line.no = (Int)((UShort)st->n_desc);
306
307            if (line.prev > line.no + OVERFLOW_DIFFERENCE && file.same) {
308               VG_(message)(Vg_DebugMsg,
309                  "Line number overflow detected (%d --> %d) in %s\n",
310                  line.prev, line.no, file.name);
311               line.ovf++;
312            }
313            file.same = True;
314
315            /* This is pretty horrible.  If this is the first line of
316               the function, then bind any unbound symbols to the arg
317               scope, since they're probably arguments. */
318            if (line.first) {
319               line.first = False;
320
321               /* remember first line of function */
322               if (func.start != 0) {
323                  func.line = line.no;
324               }
325            }
326            break;
327         }
328
329         case N_FUN: {                /* function start/end */
330            Addr addr = 0;        /* end address for prev line/scope */
331
332            /* if this the end of the function or we haven't
333               previously finished the previous function... */
334            if (*string == '\0' || func.start != 0) {
335               /* end of function */
336               line.first = False;
337
338               /* end line at end of function */
339               addr = func.start + st->n_value;
340
341               /* now between functions */
342               func.start = 0;
343
344               // XXXX DEAD POINT XXXX
345            }
346
347            if (*string != '\0') {
348               /* new function */
349               line.first = True;
350
351               /* line ends at start of next function */
352               addr = di->text_debug_bias + st->n_value;
353
354               func.start = addr;
355            }
356
357            if (line.addr) {
358               ML_(addLineInfo)(di, file.name, NULL, line.addr,
359                                addr, line.no + line.ovf * LINENO_OVERFLOW, i);
360               line.addr = 0;
361            }
362
363            //DEAD POINT
364            //DEAD POINT
365            break;
366         }
367
368         case N_LBRAC: {
369            /* open new scope */
370            // DEAD POINT
371            break;
372         }
373
374         case N_RBRAC: {
375            /* close scope */
376            // DEAD POINT
377            break;
378         }
379
380         case N_GSYM:                /* global variable */
381         case N_STSYM:                /* static in data segment */
382         case N_LCSYM:                /* static in bss segment */
383         case N_PSYM:                /* function parameter */
384         case N_LSYM:                /* stack variable */
385         case N_RSYM:                  /* register variable */
386            break;
387      }
388   }
389}
390
391#endif /* (defined(VGO_linux) && !defined(VGPV_arm_linux_android)) \
392          || defined(VGO_darwin) */
393
394/*--------------------------------------------------------------------*/
395/*--- end                                                          ---*/
396/*--------------------------------------------------------------------*/
397