1 /*-
2 * Copyright (c) 2003-2008 Joseph Koshy
3 * All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 *
14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24 * SUCH DAMAGE.
25 */
26
27 #include <sys/types.h>
28 #include <sys/cpuset.h>
29 #include <sys/param.h>
30 #include <sys/endian.h>
31 #include <sys/pmc.h>
32 #include <sys/sysctl.h>
33 #include <sys/imgact_aout.h>
34 #include <sys/imgact_elf.h>
35
36 #include <netinet/in.h>
37
38 #include <assert.h>
39 #include <err.h>
40 #include <fcntl.h>
41 #include <pmc.h>
42 #include <pmclog.h>
43 #include <stdbool.h>
44 #include <stdio.h>
45 #include <stdlib.h>
46 #include <string.h>
47 #include <sysexits.h>
48 #include <unistd.h>
49
50 #include "libpmcstat.h"
51
52 #define min(A,B) ((A) < (B) ? (A) : (B))
53 #define max(A,B) ((A) > (B) ? (A) : (B))
54
55 /*
56 * Add the list of symbols in the given section to the list associated
57 * with the object.
58 */
59 void
pmcstat_image_add_symbols(struct pmcstat_image * image,Elf * e,Elf_Scn * scn,GElf_Shdr * sh)60 pmcstat_image_add_symbols(struct pmcstat_image *image, Elf *e,
61 Elf_Scn *scn, GElf_Shdr *sh)
62 {
63 int firsttime;
64 size_t n, newsyms, nshsyms, nfuncsyms;
65 struct pmcstat_symbol *symptr;
66 char *fnname;
67 GElf_Sym sym;
68 Elf_Data *data;
69
70 if ((data = elf_getdata(scn, NULL)) == NULL)
71 return;
72
73 /*
74 * Determine the number of functions named in this
75 * section.
76 */
77
78 nshsyms = sh->sh_size / sh->sh_entsize;
79 for (n = nfuncsyms = 0; n < nshsyms; n++) {
80 if (gelf_getsym(data, (int) n, &sym) != &sym)
81 return;
82 if (GELF_ST_TYPE(sym.st_info) == STT_FUNC)
83 nfuncsyms++;
84 }
85
86 if (nfuncsyms == 0)
87 return;
88
89 /*
90 * Allocate space for the new entries.
91 */
92 firsttime = image->pi_symbols == NULL;
93 symptr = reallocarray(image->pi_symbols,
94 image->pi_symcount + nfuncsyms, sizeof(*symptr));
95 if (symptr == image->pi_symbols) /* realloc() failed. */
96 return;
97 image->pi_symbols = symptr;
98
99 /*
100 * Append new symbols to the end of the current table.
101 */
102 symptr += image->pi_symcount;
103
104 for (n = newsyms = 0; n < nshsyms; n++) {
105 if (gelf_getsym(data, (int) n, &sym) != &sym)
106 return;
107 if (GELF_ST_TYPE(sym.st_info) != STT_FUNC)
108 continue;
109
110 if (sym.st_shndx == STN_UNDEF)
111 continue;
112
113 if (!firsttime && pmcstat_symbol_search(image, sym.st_value))
114 continue; /* We've seen this symbol already. */
115
116 if ((fnname = elf_strptr(e, sh->sh_link, sym.st_name))
117 == NULL)
118 continue;
119
120 #if defined(__aarch64__) || defined(__arm__)
121 /* Ignore ARM mapping symbols. */
122 if (fnname[0] == '$' &&
123 (fnname[1] == 'a' || fnname[1] == 't' ||
124 fnname[1] == 'd' || fnname[1] == 'x'))
125 continue;
126
127 /*
128 * Clear LSB from starting addresses for functions
129 * which execute in Thumb mode. We should perhaps
130 * only do this for functions in a $t mapping symbol
131 * range, but parsing mapping symbols would be a lot
132 * of work and function addresses shouldn't have the
133 * LSB set otherwise.
134 */
135 sym.st_value &= ~1;
136 #endif
137
138 symptr->ps_name = pmcstat_string_intern(fnname);
139 symptr->ps_start = sym.st_value - image->pi_vaddr;
140 symptr->ps_end = symptr->ps_start + sym.st_size;
141
142 symptr++;
143 newsyms++;
144 }
145
146 image->pi_symcount += newsyms;
147 if (image->pi_symcount == 0)
148 return;
149
150 assert(newsyms <= nfuncsyms);
151
152 /*
153 * Return space to the system if there were duplicates.
154 */
155 if (newsyms < nfuncsyms)
156 image->pi_symbols = reallocarray(image->pi_symbols,
157 image->pi_symcount, sizeof(*symptr));
158
159 /*
160 * Keep the list of symbols sorted.
161 */
162 qsort(image->pi_symbols, image->pi_symcount, sizeof(*symptr),
163 pmcstat_symbol_compare);
164
165 /*
166 * Deal with function symbols that have a size of 'zero' by
167 * making them extend to the next higher address. These
168 * symbols are usually defined in assembly code.
169 */
170 for (symptr = image->pi_symbols;
171 symptr < image->pi_symbols + (image->pi_symcount - 1);
172 symptr++)
173 if (symptr->ps_start == symptr->ps_end)
174 symptr->ps_end = (symptr+1)->ps_start;
175 }
176
177 /*
178 * Record the fact that PC values from 'start' to 'end' come from
179 * image 'image'.
180 */
181
182 void
pmcstat_image_link(struct pmcstat_process * pp,struct pmcstat_image * image,uintfptr_t start)183 pmcstat_image_link(struct pmcstat_process *pp, struct pmcstat_image *image,
184 uintfptr_t start)
185 {
186 struct pmcstat_pcmap *pcm, *pcmnew;
187 uintfptr_t offset;
188 #ifdef __powerpc__
189 unsigned long kernbase;
190 size_t kernbase_len;
191 #endif
192
193 assert(image->pi_type != PMCSTAT_IMAGE_UNKNOWN &&
194 image->pi_type != PMCSTAT_IMAGE_INDETERMINABLE);
195
196 if ((pcmnew = malloc(sizeof(*pcmnew))) == NULL)
197 err(EX_OSERR, "ERROR: Cannot create a map entry");
198
199 /*
200 * PowerPC kernel is of DYN type and it has a base address
201 * where it is initially loaded, before being relocated.
202 * As the address in 'start' is where the kernel was relocated to,
203 * but the symbols always use the original base address, we need to
204 * subtract it to get the correct offset.
205 */
206 #ifdef __powerpc__
207 if (pp->pp_pid == -1) {
208 kernbase = 0;
209 kernbase_len = sizeof(kernbase);
210 if (sysctlbyname("kern.base_address", &kernbase, &kernbase_len,
211 NULL, 0) == -1)
212 warnx(
213 "WARNING: Could not retrieve kernel base address");
214 else
215 start -= kernbase;
216 }
217 #endif
218
219 /*
220 * Adjust the map entry to only cover the text portion
221 * of the object.
222 */
223
224 offset = start - image->pi_vaddr;
225 pcmnew->ppm_lowpc = image->pi_start + offset;
226 pcmnew->ppm_highpc = image->pi_end + offset;
227 pcmnew->ppm_image = image;
228
229 assert(pcmnew->ppm_lowpc < pcmnew->ppm_highpc);
230
231 /* Overlapped mmap()'s are assumed to never occur. */
232 TAILQ_FOREACH(pcm, &pp->pp_map, ppm_next)
233 if (pcm->ppm_lowpc >= pcmnew->ppm_highpc)
234 break;
235
236 if (pcm == NULL)
237 TAILQ_INSERT_TAIL(&pp->pp_map, pcmnew, ppm_next);
238 else
239 TAILQ_INSERT_BEFORE(pcm, pcmnew, ppm_next);
240 }
241
242 /*
243 * Determine whether a given executable image is an A.OUT object, and
244 * if so, fill in its parameters from the text file.
245 * Sets image->pi_type.
246 */
247
248 void
pmcstat_image_get_aout_params(struct pmcstat_image * image,struct pmcstat_args * args)249 pmcstat_image_get_aout_params(struct pmcstat_image *image,
250 struct pmcstat_args *args)
251 {
252 int fd;
253 ssize_t nbytes;
254 struct exec ex;
255 const char *path;
256 char buffer[PATH_MAX];
257
258 path = pmcstat_string_unintern(image->pi_execpath);
259 assert(path != NULL);
260
261 if (image->pi_iskernelmodule)
262 errx(EX_SOFTWARE,
263 "ERROR: a.out kernel modules are unsupported \"%s\"", path);
264
265 (void) snprintf(buffer, sizeof(buffer), "%s%s",
266 args->pa_fsroot, path);
267
268 if ((fd = open(buffer, O_RDONLY, 0)) < 0 ||
269 (nbytes = read(fd, &ex, sizeof(ex))) < 0) {
270 if (args->pa_verbosity >= 2)
271 warn("WARNING: Cannot determine type of \"%s\"",
272 path);
273 image->pi_type = PMCSTAT_IMAGE_INDETERMINABLE;
274 if (fd != -1)
275 (void) close(fd);
276 return;
277 }
278
279 (void) close(fd);
280
281 if ((unsigned) nbytes != sizeof(ex) ||
282 N_BADMAG(ex))
283 return;
284
285 image->pi_type = PMCSTAT_IMAGE_AOUT;
286
287 /* TODO: the rest of a.out processing */
288
289 return;
290 }
291
292 /*
293 * Examine an ELF file to determine the size of its text segment.
294 * Sets image->pi_type if anything conclusive can be determined about
295 * this image.
296 */
297
298 void
pmcstat_image_get_elf_params(struct pmcstat_image * image,struct pmcstat_args * args)299 pmcstat_image_get_elf_params(struct pmcstat_image *image,
300 struct pmcstat_args *args)
301 {
302 int fd;
303 size_t i, nph, nsh;
304 const char *path, *elfbase;
305 char *p, *endp;
306 bool first_exec_segment;
307 uintfptr_t minva, maxva;
308 Elf *e;
309 Elf_Scn *scn;
310 GElf_Ehdr eh;
311 GElf_Phdr ph;
312 GElf_Shdr sh;
313 enum pmcstat_image_type image_type;
314 char buffer[PATH_MAX];
315
316 assert(image->pi_type == PMCSTAT_IMAGE_UNKNOWN);
317
318 image->pi_start = minva = ~(uintfptr_t) 0;
319 image->pi_end = maxva = (uintfptr_t) 0;
320 image->pi_type = image_type = PMCSTAT_IMAGE_INDETERMINABLE;
321 image->pi_isdynamic = 0;
322 image->pi_dynlinkerpath = NULL;
323 image->pi_vaddr = 0;
324
325 path = pmcstat_string_unintern(image->pi_execpath);
326 assert(path != NULL);
327
328 /*
329 * Look for files under FSROOT/PATHNAME.
330 */
331 (void) snprintf(buffer, sizeof(buffer), "%s%s",
332 args->pa_fsroot, path);
333
334 e = NULL;
335 fd = open(buffer, O_RDONLY, 0);
336 if (fd < 0) {
337 warnx("WARNING: Cannot open \"%s\".",
338 buffer);
339 goto done;
340 }
341
342 if (elf_version(EV_CURRENT) == EV_NONE) {
343 warnx("WARNING: failed to init elf\n");
344 goto done;
345 }
346
347 if ((e = elf_begin(fd, ELF_C_READ, NULL)) == NULL) {
348 warnx("WARNING: Cannot read \"%s\".",
349 buffer);
350 goto done;
351 }
352
353 if (elf_kind(e) != ELF_K_ELF) {
354 if (args->pa_verbosity >= 2)
355 warnx("WARNING: Cannot determine the type of \"%s\".",
356 buffer);
357 goto done;
358 }
359
360 if (gelf_getehdr(e, &eh) != &eh) {
361 warnx(
362 "WARNING: Cannot retrieve the ELF Header for \"%s\": %s.",
363 buffer, elf_errmsg(-1));
364 goto done;
365 }
366
367 if (eh.e_type != ET_EXEC && eh.e_type != ET_DYN &&
368 !(image->pi_iskernelmodule && eh.e_type == ET_REL)) {
369 warnx("WARNING: \"%s\" is of an unsupported ELF type.",
370 buffer);
371 goto done;
372 }
373
374 image_type = eh.e_ident[EI_CLASS] == ELFCLASS32 ?
375 PMCSTAT_IMAGE_ELF32 : PMCSTAT_IMAGE_ELF64;
376
377 /*
378 * Determine the virtual address where an executable would be
379 * loaded. Additionally, for dynamically linked executables,
380 * save the pathname to the runtime linker.
381 */
382 if (eh.e_type != ET_REL) {
383 if (elf_getphnum(e, &nph) == 0) {
384 warnx(
385 "WARNING: Could not determine the number of program headers in \"%s\": %s.",
386 buffer,
387 elf_errmsg(-1));
388 goto done;
389 }
390 first_exec_segment = true;
391 for (i = 0; i < eh.e_phnum; i++) {
392 if (gelf_getphdr(e, i, &ph) != &ph) {
393 warnx(
394 "WARNING: Retrieval of PHDR entry #%ju in \"%s\" failed: %s.",
395 (uintmax_t) i, buffer, elf_errmsg(-1));
396 goto done;
397 }
398 switch (ph.p_type) {
399 case PT_DYNAMIC:
400 image->pi_isdynamic = 1;
401 break;
402 case PT_INTERP:
403 if ((elfbase = elf_rawfile(e, NULL)) == NULL) {
404 warnx(
405 "WARNING: Cannot retrieve the interpreter for \"%s\": %s.",
406 buffer, elf_errmsg(-1));
407 goto done;
408 }
409 image->pi_dynlinkerpath =
410 pmcstat_string_intern(elfbase +
411 ph.p_offset);
412 break;
413 case PT_LOAD:
414 if ((ph.p_flags & PF_X) != 0 &&
415 first_exec_segment) {
416 image->pi_vaddr = ph.p_vaddr & (-ph.p_align);
417 first_exec_segment = false;
418 }
419 break;
420 }
421 }
422 }
423
424 /*
425 * Get the min and max VA associated with this ELF object.
426 */
427 if (elf_getshnum(e, &nsh) == 0) {
428 warnx(
429 "WARNING: Could not determine the number of sections for \"%s\": %s.",
430 buffer, elf_errmsg(-1));
431 goto done;
432 }
433
434 for (i = 0; i < nsh; i++) {
435 if ((scn = elf_getscn(e, i)) == NULL ||
436 gelf_getshdr(scn, &sh) != &sh) {
437 warnx(
438 "WARNING: Could not retrieve section header #%ju in \"%s\": %s.",
439 (uintmax_t) i, buffer, elf_errmsg(-1));
440 goto done;
441 }
442 if (sh.sh_flags & SHF_EXECINSTR) {
443 minva = min(minva, sh.sh_addr);
444 maxva = max(maxva, sh.sh_addr + sh.sh_size);
445 }
446 if (sh.sh_type == SHT_SYMTAB || sh.sh_type == SHT_DYNSYM)
447 pmcstat_image_add_symbols(image, e, scn, &sh);
448 }
449
450 image->pi_start = minva;
451 image->pi_end = maxva;
452 image->pi_type = image_type;
453 image->pi_fullpath = pmcstat_string_intern(buffer);
454
455 /* Build display name
456 */
457 endp = buffer;
458 for (p = buffer; *p; p++)
459 if (*p == '/')
460 endp = p+1;
461 image->pi_name = pmcstat_string_intern(endp);
462
463 done:
464 (void) elf_end(e);
465 if (fd >= 0)
466 (void) close(fd);
467 return;
468 }
469
470 /*
471 * Given an image descriptor, determine whether it is an ELF, or AOUT.
472 * If no handler claims the image, set its type to 'INDETERMINABLE'.
473 */
474
475 void
pmcstat_image_determine_type(struct pmcstat_image * image,struct pmcstat_args * args)476 pmcstat_image_determine_type(struct pmcstat_image *image,
477 struct pmcstat_args *args)
478 {
479 assert(image->pi_type == PMCSTAT_IMAGE_UNKNOWN);
480
481 /* Try each kind of handler in turn */
482 if (image->pi_type == PMCSTAT_IMAGE_UNKNOWN)
483 pmcstat_image_get_elf_params(image, args);
484 if (image->pi_type == PMCSTAT_IMAGE_UNKNOWN)
485 pmcstat_image_get_aout_params(image, args);
486
487 /*
488 * Otherwise, remember that we tried to determine
489 * the object's type and had failed.
490 */
491 if (image->pi_type == PMCSTAT_IMAGE_UNKNOWN)
492 image->pi_type = PMCSTAT_IMAGE_INDETERMINABLE;
493 }
494
495 /*
496 * Locate an image descriptor given an interned path, adding a fresh
497 * descriptor to the cache if necessary. This function also finds a
498 * suitable name for this image's sample file.
499 *
500 * We defer filling in the file format specific parts of the image
501 * structure till the time we actually see a sample that would fall
502 * into this image.
503 */
504
505 struct pmcstat_image *
pmcstat_image_from_path(pmcstat_interned_string internedpath,int iskernelmodule,struct pmcstat_args * args,struct pmc_plugins * plugins)506 pmcstat_image_from_path(pmcstat_interned_string internedpath,
507 int iskernelmodule, struct pmcstat_args *args,
508 struct pmc_plugins *plugins)
509 {
510 int hash;
511 struct pmcstat_image *pi;
512
513 hash = pmcstat_string_lookup_hash(internedpath);
514
515 /* First, look for an existing entry. */
516 LIST_FOREACH(pi, &pmcstat_image_hash[hash], pi_next)
517 if (pi->pi_execpath == internedpath &&
518 pi->pi_iskernelmodule == iskernelmodule)
519 return (pi);
520
521 /*
522 * Allocate a new entry and place it at the head of the hash
523 * and LRU lists.
524 */
525 pi = malloc(sizeof(*pi));
526 if (pi == NULL)
527 return (NULL);
528
529 pi->pi_type = PMCSTAT_IMAGE_UNKNOWN;
530 pi->pi_execpath = internedpath;
531 pi->pi_start = ~0;
532 pi->pi_end = 0;
533 pi->pi_entry = 0;
534 pi->pi_vaddr = 0;
535 pi->pi_isdynamic = 0;
536 pi->pi_iskernelmodule = iskernelmodule;
537 pi->pi_dynlinkerpath = NULL;
538 pi->pi_symbols = NULL;
539 pi->pi_symcount = 0;
540 pi->pi_addr2line = NULL;
541
542 if (plugins[args->pa_pplugin].pl_initimage != NULL)
543 plugins[args->pa_pplugin].pl_initimage(pi);
544 if (plugins[args->pa_plugin].pl_initimage != NULL)
545 plugins[args->pa_plugin].pl_initimage(pi);
546
547 LIST_INSERT_HEAD(&pmcstat_image_hash[hash], pi, pi_next);
548
549 return (pi);
550 }
551