hsdis.c revision 1472:c18cbe5936b8 
1/*
2 * Copyright (c) 2008, 2009, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation.
8 *
9 * This code is distributed in the hope that it will be useful, but WITHOUT
10 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
11 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
12 * version 2 for more details (a copy is included in the LICENSE file that
13 * accompanied this code).
14 *
15 * You should have received a copy of the GNU General Public License version
16 * 2 along with this work; if not, write to the Free Software Foundation,
17 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
18 *
19 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
20 * or visit www.oracle.com if you need additional information or have any
21 * questions.
22 *
23 */
24
25/* hsdis.c -- dump a range of addresses as native instructions
26   This implements the plugin protocol required by the
27   HotSpot PrintAssembly option.
28*/
29
30#include "hsdis.h"
31
32#include <sysdep.h>
33#include <libiberty.h>
34#include <bfd.h>
35#include <dis-asm.h>
36#include <inttypes.h>
37
38#ifndef bool
39#define bool int
40#define true 1
41#define false 0
42#endif /*bool*/
43
44/* short names for stuff in hsdis.h */
45typedef decode_instructions_event_callback_ftype  event_callback_t;
46typedef decode_instructions_printf_callback_ftype printf_callback_t;
47
48/* disassemble_info.application_data object */
49struct hsdis_app_data {
50  /* the arguments to decode_instructions */
51  uintptr_t start; uintptr_t end;
52  event_callback_t  event_callback;  void* event_stream;
53  printf_callback_t printf_callback; void* printf_stream;
54  bool losing;
55
56  /* the architecture being disassembled */
57  const char* arch_name;
58  const bfd_arch_info_type* arch_info;
59
60  /* the disassembler we are going to use: */
61  disassembler_ftype      dfn;
62  struct disassemble_info dinfo; /* the actual struct! */
63
64  char mach_option[64];
65  char insn_options[256];
66};
67
68#define DECL_APP_DATA(dinfo) \
69  struct hsdis_app_data* app_data = (struct hsdis_app_data*) (dinfo)->application_data
70
71#define DECL_EVENT_CALLBACK(app_data) \
72  event_callback_t  event_callback = (app_data)->event_callback; \
73  void*             event_stream   = (app_data)->event_stream
74
75#define DECL_PRINTF_CALLBACK(app_data) \
76  printf_callback_t  printf_callback = (app_data)->printf_callback; \
77  void*              printf_stream   = (app_data)->printf_stream
78
79
80static void print_help(struct hsdis_app_data* app_data,
81                       const char* msg, const char* arg);
82static void setup_app_data(struct hsdis_app_data* app_data,
83                           const char* options);
84static const char* format_insn_close(const char* close,
85                                     disassemble_info* dinfo,
86                                     char* buf, size_t bufsize);
87
88void*
89#ifdef DLL_ENTRY
90  DLL_ENTRY
91#endif
92decode_instructions(void* start_pv, void* end_pv,
93                    event_callback_t  event_callback_arg,  void* event_stream_arg,
94                    printf_callback_t printf_callback_arg, void* printf_stream_arg,
95                    const char* options) {
96  struct hsdis_app_data app_data;
97  memset(&app_data, 0, sizeof(app_data));
98  app_data.start = (uintptr_t) start_pv;
99  app_data.end   = (uintptr_t) end_pv;
100  app_data.event_callback  = event_callback_arg;
101  app_data.event_stream    = event_stream_arg;
102  app_data.printf_callback = printf_callback_arg;
103  app_data.printf_stream   = printf_stream_arg;
104
105  setup_app_data(&app_data, options);
106  char buf[128];
107
108  {
109    /* now reload everything from app_data: */
110    DECL_EVENT_CALLBACK(&app_data);
111    DECL_PRINTF_CALLBACK(&app_data);
112    uintptr_t start = app_data.start;
113    uintptr_t end   = app_data.end;
114    uintptr_t p     = start;
115
116    (*event_callback)(event_stream, "insns", (void*)start);
117
118    (*event_callback)(event_stream, "mach name='%s'",
119                      (void*) app_data.arch_info->printable_name);
120    if (app_data.dinfo.bytes_per_line != 0) {
121      (*event_callback)(event_stream, "format bytes-per-line='%p'/",
122                        (void*)(intptr_t) app_data.dinfo.bytes_per_line);
123    }
124
125    while (p < end && !app_data.losing) {
126      (*event_callback)(event_stream, "insn", (void*) p);
127
128      /* reset certain state, so we can read it with confidence */
129      app_data.dinfo.insn_info_valid    = 0;
130      app_data.dinfo.branch_delay_insns = 0;
131      app_data.dinfo.data_size          = 0;
132      app_data.dinfo.insn_type          = 0;
133
134      int size = (*app_data.dfn)((bfd_vma) p, &app_data.dinfo);
135
136      if (size > 0)  p += size;
137      else           app_data.losing = true;
138
139      const char* insn_close = format_insn_close("/insn", &app_data.dinfo,
140                                                 buf, sizeof(buf));
141      (*event_callback)(event_stream, insn_close, (void*) p);
142
143      /* follow each complete insn by a nice newline */
144      (*printf_callback)(printf_stream, "\n");
145    }
146
147    (*event_callback)(event_stream, "/insns", (void*) p);
148    return (void*) p;
149  }
150}
151
152/* take the address of the function, for luck, and also test the typedef: */
153const decode_instructions_ftype decode_instructions_address = &decode_instructions;
154
155static const char* format_insn_close(const char* close,
156                                     disassemble_info* dinfo,
157                                     char* buf, size_t bufsize) {
158  if (!dinfo->insn_info_valid)
159    return close;
160  enum dis_insn_type itype = dinfo->insn_type;
161  int dsize = dinfo->data_size, delays = dinfo->branch_delay_insns;
162  if ((itype == dis_nonbranch && (dsize | delays) == 0)
163      || (strlen(close) + 3*20 > bufsize))
164    return close;
165
166  const char* type = "unknown";
167  switch (itype) {
168  case dis_nonbranch:   type = NULL;         break;
169  case dis_branch:      type = "branch";     break;
170  case dis_condbranch:  type = "condbranch"; break;
171  case dis_jsr:         type = "jsr";        break;
172  case dis_condjsr:     type = "condjsr";    break;
173  case dis_dref:        type = "dref";       break;
174  case dis_dref2:       type = "dref2";      break;
175  }
176
177  strcpy(buf, close);
178  char* p = buf;
179  if (type)    sprintf(p += strlen(p), " type='%s'", type);
180  if (dsize)   sprintf(p += strlen(p), " dsize='%d'", dsize);
181  if (delays)  sprintf(p += strlen(p), " delay='%d'", delays);
182  return buf;
183}
184
185/* handler functions */
186
187static int
188hsdis_read_memory_func(bfd_vma memaddr,
189                       bfd_byte* myaddr,
190                       unsigned int length,
191                       struct disassemble_info* dinfo) {
192  uintptr_t memaddr_p = (uintptr_t) memaddr;
193  DECL_APP_DATA(dinfo);
194  if (memaddr_p + length > app_data->end) {
195    /* read is out of bounds */
196    return EIO;
197  } else {
198    memcpy(myaddr, (bfd_byte*) memaddr_p, length);
199    return 0;
200  }
201}
202
203static void
204hsdis_print_address_func(bfd_vma vma, struct disassemble_info* dinfo) {
205  /* the actual value to print: */
206  void* addr_value = (void*) (uintptr_t) vma;
207  DECL_APP_DATA(dinfo);
208  DECL_EVENT_CALLBACK(app_data);
209
210  /* issue the event: */
211  void* result =
212    (*event_callback)(event_stream, "addr/", addr_value);
213  if (result == NULL) {
214    /* event declined */
215    generic_print_address(vma, dinfo);
216  }
217}
218
219
220/* configuration */
221
222static void set_optional_callbacks(struct hsdis_app_data* app_data);
223static void parse_caller_options(struct hsdis_app_data* app_data,
224                                 const char* caller_options);
225static const char* native_arch_name();
226static enum bfd_endian native_endian();
227static const bfd_arch_info_type* find_arch_info(const char* arch_nane);
228static bfd* get_native_bfd(const bfd_arch_info_type* arch_info,
229                           /* to avoid malloc: */
230                           bfd* empty_bfd, bfd_target* empty_xvec);
231static void init_disassemble_info_from_bfd(struct disassemble_info* dinfo,
232                                           void *stream,
233                                           fprintf_ftype fprintf_func,
234                                           bfd* bfd,
235                                           char* disassembler_options);
236static void parse_fake_insn(disassembler_ftype dfn,
237                            struct disassemble_info* dinfo);
238
239static void setup_app_data(struct hsdis_app_data* app_data,
240                           const char* caller_options) {
241  /* Make reasonable defaults for null callbacks.
242     A non-null stream for a null callback is assumed to be a FILE* for output.
243     Events are rendered as XML.
244  */
245  set_optional_callbacks(app_data);
246
247  /* Look into caller_options for anything interesting. */
248  if (caller_options != NULL)
249    parse_caller_options(app_data, caller_options);
250
251  /* Discover which architecture we are going to disassemble. */
252  app_data->arch_name = &app_data->mach_option[0];
253  if (app_data->arch_name[0] == '\0')
254    app_data->arch_name = native_arch_name();
255  app_data->arch_info = find_arch_info(app_data->arch_name);
256
257  /* Make a fake bfd to hold the arch. and byteorder info. */
258  struct {
259    bfd_target empty_xvec;
260    bfd        empty_bfd;
261  } buf;
262  bfd* native_bfd = get_native_bfd(app_data->arch_info,
263                                   /* to avoid malloc: */
264                                   &buf.empty_bfd, &buf.empty_xvec);
265  init_disassemble_info_from_bfd(&app_data->dinfo,
266                                 app_data->printf_stream,
267                                 app_data->printf_callback,
268                                 native_bfd,
269                                 app_data->insn_options);
270
271  /* Finish linking together the various callback blocks. */
272  app_data->dinfo.application_data = (void*) app_data;
273  app_data->dfn = disassembler(native_bfd);
274  app_data->dinfo.print_address_func = hsdis_print_address_func;
275  app_data->dinfo.read_memory_func = hsdis_read_memory_func;
276
277  if (app_data->dfn == NULL) {
278    const char* bad = app_data->arch_name;
279    static bool complained;
280    if (bad == &app_data->mach_option[0])
281      print_help(app_data, "bad mach=%s", bad);
282    else if (!complained)
283      print_help(app_data, "bad native mach=%s; please port hsdis to this platform", bad);
284    complained = true;
285    /* must bail out */
286    app_data->losing = true;
287    return;
288  }
289
290  parse_fake_insn(app_data->dfn, &app_data->dinfo);
291}
292
293
294/* ignore all events, return a null */
295static void* null_event_callback(void* ignore_stream, const char* ignore_event, void* arg) {
296  return NULL;
297}
298
299/* print all events as XML markup */
300static void* xml_event_callback(void* stream, const char* event, void* arg) {
301  FILE* fp = (FILE*) stream;
302#define NS_PFX "dis:"
303  if (event[0] != '/') {
304    /* issue the tag, with or without a formatted argument */
305    fprintf(fp, "<"NS_PFX);
306    fprintf(fp, event, arg);
307    fprintf(fp, ">");
308  } else {
309    ++event;                    /* skip slash */
310    const char* argp = strchr(event, ' ');
311    if (argp == NULL) {
312      /* no arguments; just issue the closing tag */
313      fprintf(fp, "</"NS_PFX"%s>", event);
314    } else {
315      /* split out the closing attributes as <dis:foo_done attr='val'/> */
316      int event_prefix = (argp - event);
317      fprintf(fp, "<"NS_PFX"%.*s_done", event_prefix, event);
318      fprintf(fp, argp, arg);
319      fprintf(fp, "/></"NS_PFX"%.*s>", event_prefix, event);
320    }
321  }
322  return NULL;
323}
324
325static void set_optional_callbacks(struct hsdis_app_data* app_data) {
326  if (app_data->printf_callback == NULL) {
327    int (*fprintf_callback)(FILE*, const char*, ...) = &fprintf;
328    FILE* fprintf_stream = stdout;
329    app_data->printf_callback = (printf_callback_t) fprintf_callback;
330    if (app_data->printf_stream == NULL)
331      app_data->printf_stream   = (void*)           fprintf_stream;
332  }
333  if (app_data->event_callback == NULL) {
334    if (app_data->event_stream == NULL)
335      app_data->event_callback = &null_event_callback;
336    else
337      app_data->event_callback = &xml_event_callback;
338  }
339
340}
341
342static void parse_caller_options(struct hsdis_app_data* app_data, const char* caller_options) {
343  char* iop_base = app_data->insn_options;
344  char* iop_limit = iop_base + sizeof(app_data->insn_options) - 1;
345  char* iop = iop_base;
346  const char* p;
347  for (p = caller_options; p != NULL; ) {
348    const char* q = strchr(p, ',');
349    size_t plen = (q == NULL) ? strlen(p) : ((q++) - p);
350    if (plen == 4 && strncmp(p, "help", plen) == 0) {
351      print_help(app_data, NULL, NULL);
352    } else if (plen >= 5 && strncmp(p, "mach=", 5) == 0) {
353      char*  mach_option = app_data->mach_option;
354      size_t mach_size   = sizeof(app_data->mach_option);
355      mach_size -= 1;           /*leave room for the null*/
356      if (plen > mach_size)  plen = mach_size;
357      strncpy(mach_option, p, plen);
358      mach_option[plen] = '\0';
359    } else if (plen > 6 && strncmp(p, "hsdis-", 6)) {
360      // do not pass these to the next level
361    } else {
362      /* just copy it; {i386,sparc}-dis.c might like to see it  */
363      if (iop > iop_base && iop < iop_limit)  (*iop++) = ',';
364      if (iop + plen > iop_limit)
365        plen = iop_limit - iop;
366      strncpy(iop, p, plen);
367      iop += plen;
368    }
369    p = q;
370  }
371}
372
373static void print_help(struct hsdis_app_data* app_data,
374                       const char* msg, const char* arg) {
375  DECL_PRINTF_CALLBACK(app_data);
376  if (msg != NULL) {
377    (*printf_callback)(printf_stream, "hsdis: ");
378    (*printf_callback)(printf_stream, msg, arg);
379    (*printf_callback)(printf_stream, "\n");
380  }
381  (*printf_callback)(printf_stream, "hsdis output options:\n");
382  if (printf_callback == (printf_callback_t) &fprintf)
383    disassembler_usage((FILE*) printf_stream);
384  else
385    disassembler_usage(stderr); /* better than nothing */
386  (*printf_callback)(printf_stream, "  mach=<arch>   select disassembly mode\n");
387#if defined(LIBARCH_i386) || defined(LIBARCH_amd64)
388  (*printf_callback)(printf_stream, "  mach=i386     select 32-bit mode\n");
389  (*printf_callback)(printf_stream, "  mach=x86-64   select 64-bit mode\n");
390  (*printf_callback)(printf_stream, "  suffix        always print instruction suffix\n");
391#endif
392  (*printf_callback)(printf_stream, "  help          print this message\n");
393}
394
395
396/* low-level bfd and arch stuff that binutils doesn't do for us */
397
398static const bfd_arch_info_type* find_arch_info(const char* arch_name) {
399  const bfd_arch_info_type* arch_info = bfd_scan_arch(arch_name);
400  if (arch_info == NULL) {
401    extern const bfd_arch_info_type bfd_default_arch_struct;
402    arch_info = &bfd_default_arch_struct;
403  }
404  return arch_info;
405}
406
407static const char* native_arch_name() {
408  const char* res = NULL;
409#ifdef LIBARCH_i386
410    res = "i386";
411#endif
412#ifdef LIBARCH_amd64
413    res = "i386:x86-64";
414#endif
415#ifdef LIBARCH_sparc
416    res = "sparc:v8plusb";
417#endif
418#ifdef LIBARCH_sparcv9
419    res = "sparc:v9b";
420#endif
421  if (res == NULL)
422    res = "architecture not set in Makefile!";
423  return res;
424}
425
426static enum bfd_endian native_endian() {
427  int32_t endian_test = 'x';
428  if (*(const char*) &endian_test == 'x')
429    return BFD_ENDIAN_LITTLE;
430  else
431    return BFD_ENDIAN_BIG;
432}
433
434static bfd* get_native_bfd(const bfd_arch_info_type* arch_info,
435                           bfd* empty_bfd, bfd_target* empty_xvec) {
436  memset(empty_bfd,  0, sizeof(*empty_bfd));
437  memset(empty_xvec, 0, sizeof(*empty_xvec));
438  empty_xvec->flavour = bfd_target_unknown_flavour;
439  empty_xvec->byteorder = native_endian();
440  empty_bfd->xvec = empty_xvec;
441  empty_bfd->arch_info = arch_info;
442  return empty_bfd;
443}
444
445static int read_zero_data_only(bfd_vma ignore_p,
446                               bfd_byte* myaddr, unsigned int length,
447                               struct disassemble_info *ignore_info) {
448  memset(myaddr, 0, length);
449  return 0;
450}
451static int print_to_dev_null(void* ignore_stream, const char* ignore_format, ...) {
452  return 0;
453}
454
455/* Prime the pump by running the selected disassembler on a null input.
456   This forces the machine-specific disassembler to divulge invariant
457   information like bytes_per_line.
458 */
459static void parse_fake_insn(disassembler_ftype dfn,
460                            struct disassemble_info* dinfo) {
461  typedef int (*read_memory_ftype)
462    (bfd_vma memaddr, bfd_byte *myaddr, unsigned int length,
463     struct disassemble_info *info);
464  read_memory_ftype read_memory_func = dinfo->read_memory_func;
465  fprintf_ftype     fprintf_func     = dinfo->fprintf_func;
466
467  dinfo->read_memory_func = &read_zero_data_only;
468  dinfo->fprintf_func     = &print_to_dev_null;
469  (*dfn)(0, dinfo);
470
471  // put it back:
472  dinfo->read_memory_func = read_memory_func;
473  dinfo->fprintf_func     = fprintf_func;
474}
475
476static void init_disassemble_info_from_bfd(struct disassemble_info* dinfo,
477                                           void *stream,
478                                           fprintf_ftype fprintf_func,
479                                           bfd* abfd,
480                                           char* disassembler_options) {
481  init_disassemble_info(dinfo, stream, fprintf_func);
482
483  dinfo->flavour = bfd_get_flavour(abfd);
484  dinfo->arch = bfd_get_arch(abfd);
485  dinfo->mach = bfd_get_mach(abfd);
486  dinfo->disassembler_options = disassembler_options;
487  dinfo->octets_per_byte = bfd_octets_per_byte (abfd);
488  dinfo->skip_zeroes = sizeof(void*) * 2;
489  dinfo->skip_zeroes_at_end = sizeof(void*)-1;
490  dinfo->disassembler_needs_relocs = FALSE;
491
492  if (bfd_big_endian(abfd))
493    dinfo->display_endian = dinfo->endian = BFD_ENDIAN_BIG;
494  else if (bfd_little_endian(abfd))
495    dinfo->display_endian = dinfo->endian = BFD_ENDIAN_LITTLE;
496  else
497    dinfo->endian = native_endian();
498
499  disassemble_init_for_target(dinfo);
500}
501