header.c 71.8 KB
Newer Older
1
#include "util.h"
2
#include <sys/types.h>
3
#include <byteswap.h>
4 5 6
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>
7
#include <linux/list.h>
8
#include <linux/kernel.h>
9
#include <linux/bitops.h>
10
#include <sys/utsname.h>
11

12
#include "evlist.h"
13
#include "evsel.h"
14
#include "header.h"
15 16
#include "../perf.h"
#include "trace-event.h"
17
#include "session.h"
18
#include "symbol.h"
19
#include "debug.h"
20
#include "cpumap.h"
21
#include "pmu.h"
22
#include "vdso.h"
23
#include "strbuf.h"
24
#include "build-id.h"
25
#include "data.h"
26 27
#include <api/fs/fs.h>
#include "asm/bug.h"
28

29 30 31 32 33 34 35 36 37 38 39 40
/*
 * magic2 = "PERFILE2"
 * must be a numerical value to let the endianness
 * determine the memory layout. That way we are able
 * to detect endianness when reading the perf.data file
 * back.
 *
 * we check for legacy (PERFFILE) format.
 */
static const char *__perf_magic1 = "PERFFILE";
static const u64 __perf_magic2    = 0x32454c4946524550ULL;
static const u64 __perf_magic2_sw = 0x50455246494c4532ULL;
41

42
#define PERF_MAGIC	__perf_magic2
43 44

struct perf_file_attr {
45
	struct perf_event_attr	attr;
46 47 48
	struct perf_file_section	ids;
};

49
void perf_header__set_feat(struct perf_header *header, int feat)
50
{
51
	set_bit(feat, header->adds_features);
52 53
}

54
void perf_header__clear_feat(struct perf_header *header, int feat)
55
{
56
	clear_bit(feat, header->adds_features);
57 58
}

59
bool perf_header__has_feat(const struct perf_header *header, int feat)
60
{
61
	return test_bit(feat, header->adds_features);
62 63
}

64
static int do_write(int fd, const void *buf, size_t size)
65 66 67 68 69
{
	while (size) {
		int ret = write(fd, buf, size);

		if (ret < 0)
70
			return -errno;
71 72 73 74

		size -= ret;
		buf += ret;
	}
75 76

	return 0;
77 78
}

79
int write_padded(int fd, const void *bf, size_t count, size_t count_aligned)
80 81 82 83 84 85 86 87 88 89
{
	static const char zero_buf[NAME_ALIGN];
	int err = do_write(fd, bf, count);

	if (!err)
		err = do_write(fd, zero_buf, count_aligned - count);

	return err;
}

90 91 92
#define string_size(str)						\
	(PERF_ALIGN((strlen(str) + 1), NAME_ALIGN) + sizeof(u32))

93 94 95 96 97 98
static int do_write_string(int fd, const char *str)
{
	u32 len, olen;
	int ret;

	olen = strlen(str) + 1;
99
	len = PERF_ALIGN(olen, NAME_ALIGN);
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114

	/* write len, incl. \0 */
	ret = do_write(fd, &len, sizeof(len));
	if (ret < 0)
		return ret;

	return write_padded(fd, str, olen, len);
}

static char *do_read_string(int fd, struct perf_header *ph)
{
	ssize_t sz, ret;
	u32 len;
	char *buf;

115
	sz = readn(fd, &len, sizeof(len));
116 117 118 119 120 121 122 123 124 125
	if (sz < (ssize_t)sizeof(len))
		return NULL;

	if (ph->needs_swap)
		len = bswap_32(len);

	buf = malloc(len);
	if (!buf)
		return NULL;

126
	ret = readn(fd, buf, len);
127 128 129 130 131 132 133 134 135 136 137 138 139
	if (ret == (ssize_t)len) {
		/*
		 * strings are padded by zeroes
		 * thus the actual strlen of buf
		 * may be less than len
		 */
		return buf;
	}

	free(buf);
	return NULL;
}

140
static int write_tracing_data(int fd, struct perf_header *h __maybe_unused,
141 142 143 144 145 146 147
			    struct perf_evlist *evlist)
{
	return read_tracing_data(fd, &evlist->entries);
}


static int write_build_id(int fd, struct perf_header *h,
148
			  struct perf_evlist *evlist __maybe_unused)
149 150 151 152 153 154
{
	struct perf_session *session;
	int err;

	session = container_of(h, struct perf_session, header);

155 156 157
	if (!perf_session__read_build_ids(session, true))
		return -1;

158
	err = perf_session__write_buildid_table(session, fd);
159 160 161 162
	if (err < 0) {
		pr_debug("failed to write buildid table\n");
		return err;
	}
163
	perf_session__cache_build_ids(session);
164 165 166 167

	return 0;
}

168 169
static int write_hostname(int fd, struct perf_header *h __maybe_unused,
			  struct perf_evlist *evlist __maybe_unused)
170 171 172 173 174 175 176 177 178 179 180
{
	struct utsname uts;
	int ret;

	ret = uname(&uts);
	if (ret < 0)
		return -1;

	return do_write_string(fd, uts.nodename);
}

181 182
static int write_osrelease(int fd, struct perf_header *h __maybe_unused,
			   struct perf_evlist *evlist __maybe_unused)
183 184 185 186 187 188 189 190 191 192 193
{
	struct utsname uts;
	int ret;

	ret = uname(&uts);
	if (ret < 0)
		return -1;

	return do_write_string(fd, uts.release);
}

194 195
static int write_arch(int fd, struct perf_header *h __maybe_unused,
		      struct perf_evlist *evlist __maybe_unused)
196 197 198 199 200 201 202 203 204 205 206
{
	struct utsname uts;
	int ret;

	ret = uname(&uts);
	if (ret < 0)
		return -1;

	return do_write_string(fd, uts.machine);
}

207 208
static int write_version(int fd, struct perf_header *h __maybe_unused,
			 struct perf_evlist *evlist __maybe_unused)
209 210 211 212
{
	return do_write_string(fd, perf_version_string);
}

213
static int __write_cpudesc(int fd, const char *cpuinfo_proc)
214 215 216 217
{
	FILE *file;
	char *buf = NULL;
	char *s, *p;
218
	const char *search = cpuinfo_proc;
219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
	size_t len = 0;
	int ret = -1;

	if (!search)
		return -1;

	file = fopen("/proc/cpuinfo", "r");
	if (!file)
		return -1;

	while (getline(&buf, &len, file) > 0) {
		ret = strncmp(buf, search, strlen(search));
		if (!ret)
			break;
	}

235 236
	if (ret) {
		ret = -1;
237
		goto done;
238
	}
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269

	s = buf;

	p = strchr(buf, ':');
	if (p && *(p+1) == ' ' && *(p+2))
		s = p + 2;
	p = strchr(s, '\n');
	if (p)
		*p = '\0';

	/* squash extra space characters (branding string) */
	p = s;
	while (*p) {
		if (isspace(*p)) {
			char *r = p + 1;
			char *q = r;
			*p = ' ';
			while (*q && isspace(*q))
				q++;
			if (q != (p+1))
				while ((*r++ = *q++));
		}
		p++;
	}
	ret = do_write_string(fd, s);
done:
	free(buf);
	fclose(file);
	return ret;
}

270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288
static int write_cpudesc(int fd, struct perf_header *h __maybe_unused,
		       struct perf_evlist *evlist __maybe_unused)
{
#ifndef CPUINFO_PROC
#define CPUINFO_PROC {"model name", }
#endif
	const char *cpuinfo_procs[] = CPUINFO_PROC;
	unsigned int i;

	for (i = 0; i < ARRAY_SIZE(cpuinfo_procs); i++) {
		int ret;
		ret = __write_cpudesc(fd, cpuinfo_procs[i]);
		if (ret >= 0)
			return ret;
	}
	return -1;
}


289 290
static int write_nrcpus(int fd, struct perf_header *h __maybe_unused,
			struct perf_evlist *evlist __maybe_unused)
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314
{
	long nr;
	u32 nrc, nra;
	int ret;

	nr = sysconf(_SC_NPROCESSORS_CONF);
	if (nr < 0)
		return -1;

	nrc = (u32)(nr & UINT_MAX);

	nr = sysconf(_SC_NPROCESSORS_ONLN);
	if (nr < 0)
		return -1;

	nra = (u32)(nr & UINT_MAX);

	ret = do_write(fd, &nrc, sizeof(nrc));
	if (ret < 0)
		return ret;

	return do_write(fd, &nra, sizeof(nra));
}

315
static int write_event_desc(int fd, struct perf_header *h __maybe_unused,
316 317
			    struct perf_evlist *evlist)
{
318
	struct perf_evsel *evsel;
319
	u32 nre, nri, sz;
320 321
	int ret;

322
	nre = evlist->nr_entries;
323 324 325 326 327 328 329 330 331 332 333

	/*
	 * write number of events
	 */
	ret = do_write(fd, &nre, sizeof(nre));
	if (ret < 0)
		return ret;

	/*
	 * size of perf_event_attr struct
	 */
334
	sz = (u32)sizeof(evsel->attr);
335 336 337 338
	ret = do_write(fd, &sz, sizeof(sz));
	if (ret < 0)
		return ret;

339
	evlist__for_each(evlist, evsel) {
340
		ret = do_write(fd, &evsel->attr, sz);
341 342 343 344 345 346 347 348 349
		if (ret < 0)
			return ret;
		/*
		 * write number of unique id per event
		 * there is one id per instance of an event
		 *
		 * copy into an nri to be independent of the
		 * type of ids,
		 */
350
		nri = evsel->ids;
351 352 353 354 355 356 357
		ret = do_write(fd, &nri, sizeof(nri));
		if (ret < 0)
			return ret;

		/*
		 * write event string as passed on cmdline
		 */
358
		ret = do_write_string(fd, perf_evsel__name(evsel));
359 360 361 362 363
		if (ret < 0)
			return ret;
		/*
		 * write unique ids for this event
		 */
364
		ret = do_write(fd, evsel->id, evsel->ids * sizeof(u64));
365 366 367 368 369 370
		if (ret < 0)
			return ret;
	}
	return 0;
}

371 372
static int write_cmdline(int fd, struct perf_header *h __maybe_unused,
			 struct perf_evlist *evlist __maybe_unused)
373 374 375
{
	char buf[MAXPATHLEN];
	char proc[32];
376 377
	u32 n;
	int i, ret;
378 379 380 381 382 383 384 385 386 387 388 389 390

	/*
	 * actual atual path to perf binary
	 */
	sprintf(proc, "/proc/%d/exe", getpid());
	ret = readlink(proc, buf, sizeof(buf));
	if (ret <= 0)
		return -1;

	/* readlink() does not add null termination */
	buf[ret] = '\0';

	/* account for binary path */
391
	n = perf_env.nr_cmdline + 1;
392 393 394 395 396 397 398 399 400

	ret = do_write(fd, &n, sizeof(n));
	if (ret < 0)
		return ret;

	ret = do_write_string(fd, buf);
	if (ret < 0)
		return ret;

401 402
	for (i = 0 ; i < perf_env.nr_cmdline; i++) {
		ret = do_write_string(fd, perf_env.cmdline_argv[i]);
403 404 405 406 407 408 409 410 411 412 413 414
		if (ret < 0)
			return ret;
	}
	return 0;
}

#define CORE_SIB_FMT \
	"/sys/devices/system/cpu/cpu%d/topology/core_siblings_list"
#define THRD_SIB_FMT \
	"/sys/devices/system/cpu/cpu%d/topology/thread_siblings_list"

struct cpu_topo {
415
	u32 cpu_nr;
416 417 418 419 420 421 422 423 424 425 426 427
	u32 core_sib;
	u32 thread_sib;
	char **core_siblings;
	char **thread_siblings;
};

static int build_cpu_topo(struct cpu_topo *tp, int cpu)
{
	FILE *fp;
	char filename[MAXPATHLEN];
	char *buf = NULL, *p;
	size_t len = 0;
428
	ssize_t sret;
429 430 431 432 433 434
	u32 i = 0;
	int ret = -1;

	sprintf(filename, CORE_SIB_FMT, cpu);
	fp = fopen(filename, "r");
	if (!fp)
435
		goto try_threads;
436

437
	sret = getline(&buf, &len, fp);
438
	fclose(fp);
439 440
	if (sret <= 0)
		goto try_threads;
441 442 443 444 445 446 447 448 449 450 451 452 453 454 455

	p = strchr(buf, '\n');
	if (p)
		*p = '\0';

	for (i = 0; i < tp->core_sib; i++) {
		if (!strcmp(buf, tp->core_siblings[i]))
			break;
	}
	if (i == tp->core_sib) {
		tp->core_siblings[i] = buf;
		tp->core_sib++;
		buf = NULL;
		len = 0;
	}
456
	ret = 0;
457

458
try_threads:
459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
	sprintf(filename, THRD_SIB_FMT, cpu);
	fp = fopen(filename, "r");
	if (!fp)
		goto done;

	if (getline(&buf, &len, fp) <= 0)
		goto done;

	p = strchr(buf, '\n');
	if (p)
		*p = '\0';

	for (i = 0; i < tp->thread_sib; i++) {
		if (!strcmp(buf, tp->thread_siblings[i]))
			break;
	}
	if (i == tp->thread_sib) {
		tp->thread_siblings[i] = buf;
		tp->thread_sib++;
		buf = NULL;
	}
	ret = 0;
done:
	if(fp)
		fclose(fp);
	free(buf);
	return ret;
}

static void free_cpu_topo(struct cpu_topo *tp)
{
	u32 i;

	if (!tp)
		return;

	for (i = 0 ; i < tp->core_sib; i++)
496
		zfree(&tp->core_siblings[i]);
497 498

	for (i = 0 ; i < tp->thread_sib; i++)
499
		zfree(&tp->thread_siblings[i]);
500 501 502 503 504 505 506 507 508

	free(tp);
}

static struct cpu_topo *build_cpu_topology(void)
{
	struct cpu_topo *tp;
	void *addr;
	u32 nr, i;
509
	size_t sz;
510 511 512 513 514 515 516 517 518 519 520
	long ncpus;
	int ret = -1;

	ncpus = sysconf(_SC_NPROCESSORS_CONF);
	if (ncpus < 0)
		return NULL;

	nr = (u32)(ncpus & UINT_MAX);

	sz = nr * sizeof(char *);

521
	addr = calloc(1, sizeof(*tp) + 2 * sz);
522 523 524 525
	if (!addr)
		return NULL;

	tp = addr;
526
	tp->cpu_nr = nr;
527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543
	addr += sizeof(*tp);
	tp->core_siblings = addr;
	addr += sz;
	tp->thread_siblings = addr;

	for (i = 0; i < nr; i++) {
		ret = build_cpu_topo(tp, i);
		if (ret < 0)
			break;
	}
	if (ret) {
		free_cpu_topo(tp);
		tp = NULL;
	}
	return tp;
}

544 545
static int write_cpu_topology(int fd, struct perf_header *h __maybe_unused,
			  struct perf_evlist *evlist __maybe_unused)
546 547 548
{
	struct cpu_topo *tp;
	u32 i;
549
	int ret, j;
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572

	tp = build_cpu_topology();
	if (!tp)
		return -1;

	ret = do_write(fd, &tp->core_sib, sizeof(tp->core_sib));
	if (ret < 0)
		goto done;

	for (i = 0; i < tp->core_sib; i++) {
		ret = do_write_string(fd, tp->core_siblings[i]);
		if (ret < 0)
			goto done;
	}
	ret = do_write(fd, &tp->thread_sib, sizeof(tp->thread_sib));
	if (ret < 0)
		goto done;

	for (i = 0; i < tp->thread_sib; i++) {
		ret = do_write_string(fd, tp->thread_siblings[i]);
		if (ret < 0)
			break;
	}
573

574 575 576 577 578 579 580
	ret = perf_env__read_cpu_topology_map(&perf_env);
	if (ret < 0)
		goto done;

	for (j = 0; j < perf_env.nr_cpus_avail; j++) {
		ret = do_write(fd, &perf_env.cpu[j].core_id,
			       sizeof(perf_env.cpu[j].core_id));
581 582
		if (ret < 0)
			return ret;
583 584
		ret = do_write(fd, &perf_env.cpu[j].socket_id,
			       sizeof(perf_env.cpu[j].socket_id));
585 586 587
		if (ret < 0)
			return ret;
	}
588 589 590 591 592 593 594
done:
	free_cpu_topo(tp);
	return ret;
}



595 596
static int write_total_mem(int fd, struct perf_header *h __maybe_unused,
			  struct perf_evlist *evlist __maybe_unused)
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616
{
	char *buf = NULL;
	FILE *fp;
	size_t len = 0;
	int ret = -1, n;
	uint64_t mem;

	fp = fopen("/proc/meminfo", "r");
	if (!fp)
		return -1;

	while (getline(&buf, &len, fp) > 0) {
		ret = strncmp(buf, "MemTotal:", 9);
		if (!ret)
			break;
	}
	if (!ret) {
		n = sscanf(buf, "%*s %"PRIu64, &mem);
		if (n == 1)
			ret = do_write(fd, &mem, sizeof(mem));
617 618
	} else
		ret = -1;
619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
	free(buf);
	fclose(fp);
	return ret;
}

static int write_topo_node(int fd, int node)
{
	char str[MAXPATHLEN];
	char field[32];
	char *buf = NULL, *p;
	size_t len = 0;
	FILE *fp;
	u64 mem_total, mem_free, mem;
	int ret = -1;

	sprintf(str, "/sys/devices/system/node/node%d/meminfo", node);
	fp = fopen(str, "r");
	if (!fp)
		return -1;

	while (getline(&buf, &len, fp) > 0) {
		/* skip over invalid lines */
		if (!strchr(buf, ':'))
			continue;
643
		if (sscanf(buf, "%*s %*d %31s %"PRIu64, field, &mem) != 2)
644 645 646 647 648 649 650 651
			goto done;
		if (!strcmp(field, "MemTotal:"))
			mem_total = mem;
		if (!strcmp(field, "MemFree:"))
			mem_free = mem;
	}

	fclose(fp);
652
	fp = NULL;
653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678

	ret = do_write(fd, &mem_total, sizeof(u64));
	if (ret)
		goto done;

	ret = do_write(fd, &mem_free, sizeof(u64));
	if (ret)
		goto done;

	ret = -1;
	sprintf(str, "/sys/devices/system/node/node%d/cpulist", node);

	fp = fopen(str, "r");
	if (!fp)
		goto done;

	if (getline(&buf, &len, fp) <= 0)
		goto done;

	p = strchr(buf, '\n');
	if (p)
		*p = '\0';

	ret = do_write_string(fd, buf);
done:
	free(buf);
679 680
	if (fp)
		fclose(fp);
681 682 683
	return ret;
}

684 685
static int write_numa_topology(int fd, struct perf_header *h __maybe_unused,
			  struct perf_evlist *evlist __maybe_unused)
686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728
{
	char *buf = NULL;
	size_t len = 0;
	FILE *fp;
	struct cpu_map *node_map = NULL;
	char *c;
	u32 nr, i, j;
	int ret = -1;

	fp = fopen("/sys/devices/system/node/online", "r");
	if (!fp)
		return -1;

	if (getline(&buf, &len, fp) <= 0)
		goto done;

	c = strchr(buf, '\n');
	if (c)
		*c = '\0';

	node_map = cpu_map__new(buf);
	if (!node_map)
		goto done;

	nr = (u32)node_map->nr;

	ret = do_write(fd, &nr, sizeof(nr));
	if (ret < 0)
		goto done;

	for (i = 0; i < nr; i++) {
		j = (u32)node_map->map[i];
		ret = do_write(fd, &j, sizeof(j));
		if (ret < 0)
			break;

		ret = write_topo_node(fd, i);
		if (ret < 0)
			break;
	}
done:
	free(buf);
	fclose(fp);
729
	cpu_map__put(node_map);
730 731 732
	return ret;
}

733 734 735 736 737 738 739 740 741 742 743 744
/*
 * File format:
 *
 * struct pmu_mappings {
 *	u32	pmu_num;
 *	struct pmu_map {
 *		u32	type;
 *		char	name[];
 *	}[pmu_num];
 * };
 */

745 746
static int write_pmu_mappings(int fd, struct perf_header *h __maybe_unused,
			      struct perf_evlist *evlist __maybe_unused)
747 748 749 750
{
	struct perf_pmu *pmu = NULL;
	off_t offset = lseek(fd, 0, SEEK_CUR);
	__u32 pmu_num = 0;
751
	int ret;
752 753

	/* write real pmu_num later */
754 755 756
	ret = do_write(fd, &pmu_num, sizeof(pmu_num));
	if (ret < 0)
		return ret;
757 758 759 760 761

	while ((pmu = perf_pmu__scan(pmu))) {
		if (!pmu->name)
			continue;
		pmu_num++;
762 763 764 765 766 767 768 769

		ret = do_write(fd, &pmu->type, sizeof(pmu->type));
		if (ret < 0)
			return ret;

		ret = do_write_string(fd, pmu->name);
		if (ret < 0)
			return ret;
770 771 772 773 774 775 776 777 778 779 780
	}

	if (pwrite(fd, &pmu_num, sizeof(pmu_num), offset) != sizeof(pmu_num)) {
		/* discard all */
		lseek(fd, offset, SEEK_SET);
		return -1;
	}

	return 0;
}

781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803
/*
 * File format:
 *
 * struct group_descs {
 *	u32	nr_groups;
 *	struct group_desc {
 *		char	name[];
 *		u32	leader_idx;
 *		u32	nr_members;
 *	}[nr_groups];
 * };
 */
static int write_group_desc(int fd, struct perf_header *h __maybe_unused,
			    struct perf_evlist *evlist)
{
	u32 nr_groups = evlist->nr_groups;
	struct perf_evsel *evsel;
	int ret;

	ret = do_write(fd, &nr_groups, sizeof(nr_groups));
	if (ret < 0)
		return ret;

804
	evlist__for_each(evlist, evsel) {
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
		if (perf_evsel__is_group_leader(evsel) &&
		    evsel->nr_members > 1) {
			const char *name = evsel->group_name ?: "{anon_group}";
			u32 leader_idx = evsel->idx;
			u32 nr_members = evsel->nr_members;

			ret = do_write_string(fd, name);
			if (ret < 0)
				return ret;

			ret = do_write(fd, &leader_idx, sizeof(leader_idx));
			if (ret < 0)
				return ret;

			ret = do_write(fd, &nr_members, sizeof(nr_members));
			if (ret < 0)
				return ret;
		}
	}
	return 0;
}

827 828 829 830
/*
 * default get_cpuid(): nothing gets recorded
 * actual implementation must be in arch/$(ARCH)/util/header.c
 */
831 832
int __attribute__ ((weak)) get_cpuid(char *buffer __maybe_unused,
				     size_t sz __maybe_unused)
833 834 835 836
{
	return -1;
}

837 838
static int write_cpuid(int fd, struct perf_header *h __maybe_unused,
		       struct perf_evlist *evlist __maybe_unused)
839 840 841 842 843 844 845 846 847 848 849 850 851
{
	char buffer[64];
	int ret;

	ret = get_cpuid(buffer, sizeof(buffer));
	if (!ret)
		goto write_it;

	return -1;
write_it:
	return do_write_string(fd, buffer);
}

852 853 854
static int write_branch_stack(int fd __maybe_unused,
			      struct perf_header *h __maybe_unused,
		       struct perf_evlist *evlist __maybe_unused)
855 856 857 858
{
	return 0;
}

859
static int write_auxtrace(int fd, struct perf_header *h,
860 861
			  struct perf_evlist *evlist __maybe_unused)
{
862 863 864 865 866 867 868 869 870
	struct perf_session *session;
	int err;

	session = container_of(h, struct perf_session, header);

	err = auxtrace_index__write(fd, &session->auxtrace_index);
	if (err < 0)
		pr_err("Failed to write auxtrace index\n");
	return err;
871 872
}

873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065
static int cpu_cache_level__sort(const void *a, const void *b)
{
	struct cpu_cache_level *cache_a = (struct cpu_cache_level *)a;
	struct cpu_cache_level *cache_b = (struct cpu_cache_level *)b;

	return cache_a->level - cache_b->level;
}

static bool cpu_cache_level__cmp(struct cpu_cache_level *a, struct cpu_cache_level *b)
{
	if (a->level != b->level)
		return false;

	if (a->line_size != b->line_size)
		return false;

	if (a->sets != b->sets)
		return false;

	if (a->ways != b->ways)
		return false;

	if (strcmp(a->type, b->type))
		return false;

	if (strcmp(a->size, b->size))
		return false;

	if (strcmp(a->map, b->map))
		return false;

	return true;
}

static int cpu_cache_level__read(struct cpu_cache_level *cache, u32 cpu, u16 level)
{
	char path[PATH_MAX], file[PATH_MAX];
	struct stat st;
	size_t len;

	scnprintf(path, PATH_MAX, "devices/system/cpu/cpu%d/cache/index%d/", cpu, level);
	scnprintf(file, PATH_MAX, "%s/%s", sysfs__mountpoint(), path);

	if (stat(file, &st))
		return 1;

	scnprintf(file, PATH_MAX, "%s/level", path);
	if (sysfs__read_int(file, (int *) &cache->level))
		return -1;

	scnprintf(file, PATH_MAX, "%s/coherency_line_size", path);
	if (sysfs__read_int(file, (int *) &cache->line_size))
		return -1;

	scnprintf(file, PATH_MAX, "%s/number_of_sets", path);
	if (sysfs__read_int(file, (int *) &cache->sets))
		return -1;

	scnprintf(file, PATH_MAX, "%s/ways_of_associativity", path);
	if (sysfs__read_int(file, (int *) &cache->ways))
		return -1;

	scnprintf(file, PATH_MAX, "%s/type", path);
	if (sysfs__read_str(file, &cache->type, &len))
		return -1;

	cache->type[len] = 0;
	cache->type = rtrim(cache->type);

	scnprintf(file, PATH_MAX, "%s/size", path);
	if (sysfs__read_str(file, &cache->size, &len)) {
		free(cache->type);
		return -1;
	}

	cache->size[len] = 0;
	cache->size = rtrim(cache->size);

	scnprintf(file, PATH_MAX, "%s/shared_cpu_list", path);
	if (sysfs__read_str(file, &cache->map, &len)) {
		free(cache->map);
		free(cache->type);
		return -1;
	}

	cache->map[len] = 0;
	cache->map = rtrim(cache->map);
	return 0;
}

static void cpu_cache_level__fprintf(FILE *out, struct cpu_cache_level *c)
{
	fprintf(out, "L%d %-15s %8s [%s]\n", c->level, c->type, c->size, c->map);
}

static int build_caches(struct cpu_cache_level caches[], u32 size, u32 *cntp)
{
	u32 i, cnt = 0;
	long ncpus;
	u32 nr, cpu;
	u16 level;

	ncpus = sysconf(_SC_NPROCESSORS_CONF);
	if (ncpus < 0)
		return -1;

	nr = (u32)(ncpus & UINT_MAX);

	for (cpu = 0; cpu < nr; cpu++) {
		for (level = 0; level < 10; level++) {
			struct cpu_cache_level c;
			int err;

			err = cpu_cache_level__read(&c, cpu, level);
			if (err < 0)
				return err;

			if (err == 1)
				break;

			for (i = 0; i < cnt; i++) {
				if (cpu_cache_level__cmp(&c, &caches[i]))
					break;
			}

			if (i == cnt)
				caches[cnt++] = c;
			else
				cpu_cache_level__free(&c);

			if (WARN_ONCE(cnt == size, "way too many cpu caches.."))
				goto out;
		}
	}
 out:
	*cntp = cnt;
	return 0;
}

#define MAX_CACHES 2000

static int write_cache(int fd, struct perf_header *h __maybe_unused,
			  struct perf_evlist *evlist __maybe_unused)
{
	struct cpu_cache_level caches[MAX_CACHES];
	u32 cnt = 0, i, version = 1;
	int ret;

	ret = build_caches(caches, MAX_CACHES, &cnt);
	if (ret)
		goto out;

	qsort(&caches, cnt, sizeof(struct cpu_cache_level), cpu_cache_level__sort);

	ret = do_write(fd, &version, sizeof(u32));
	if (ret < 0)
		goto out;

	ret = do_write(fd, &cnt, sizeof(u32));
	if (ret < 0)
		goto out;

	for (i = 0; i < cnt; i++) {
		struct cpu_cache_level *c = &caches[i];

		#define _W(v)					\
			ret = do_write(fd, &c->v, sizeof(u32));	\
			if (ret < 0)				\
				goto out;

		_W(level)
		_W(line_size)
		_W(sets)
		_W(ways)
		#undef _W

		#define _W(v)						\
			ret = do_write_string(fd, (const char *) c->v);	\
			if (ret < 0)					\
				goto out;

		_W(type)
		_W(size)
		_W(map)
		#undef _W
	}

out:
	for (i = 0; i < cnt; i++)
		cpu_cache_level__free(&caches[i]);
	return ret;
}

1066 1067 1068 1069 1070 1071 1072
static int write_stat(int fd __maybe_unused,
		      struct perf_header *h __maybe_unused,
		      struct perf_evlist *evlist __maybe_unused)
{
	return 0;
}

1073 1074
static void print_hostname(struct perf_header *ph, int fd __maybe_unused,
			   FILE *fp)
1075
{
1076
	fprintf(fp, "# hostname : %s\n", ph->env.hostname);
1077 1078
}

1079 1080
static void print_osrelease(struct perf_header *ph, int fd __maybe_unused,
			    FILE *fp)
1081
{
1082
	fprintf(fp, "# os release : %s\n", ph->env.os_release);
1083 1084
}

1085
static void print_arch(struct perf_header *ph, int fd __maybe_unused, FILE *fp)
1086
{
1087
	fprintf(fp, "# arch : %s\n", ph->env.arch);
1088 1089
}

1090 1091
static void print_cpudesc(struct perf_header *ph, int fd __maybe_unused,
			  FILE *fp)
1092
{
1093
	fprintf(fp, "# cpudesc : %s\n", ph->env.cpu_desc);
1094 1095
}

1096 1097
static void print_nrcpus(struct perf_header *ph, int fd __maybe_unused,
			 FILE *fp)
1098
{
1099 1100
	fprintf(fp, "# nrcpus online : %u\n", ph->env.nr_cpus_online);
	fprintf(fp, "# nrcpus avail : %u\n", ph->env.nr_cpus_avail);
1101 1102
}

1103 1104
static void print_version(struct perf_header *ph, int fd __maybe_unused,
			  FILE *fp)
1105
{
1106
	fprintf(fp, "# perf version : %s\n", ph->env.version);
1107 1108
}

1109 1110
static void print_cmdline(struct perf_header *ph, int fd __maybe_unused,
			  FILE *fp)
1111
{
1112
	int nr, i;
1113

1114
	nr = ph->env.nr_cmdline;
1115 1116 1117

	fprintf(fp, "# cmdline : ");

1118 1119
	for (i = 0; i < nr; i++)
		fprintf(fp, "%s ", ph->env.cmdline_argv[i]);
1120 1121 1122
	fputc('\n', fp);
}

1123 1124
static void print_cpu_topology(struct perf_header *ph, int fd __maybe_unused,
			       FILE *fp)
1125
{
1126
	int nr, i;
1127
	char *str;
1128
	int cpu_nr = ph->env.nr_cpus_online;
1129

1130 1131
	nr = ph->env.nr_sibling_cores;
	str = ph->env.sibling_cores;
1132 1133 1134

	for (i = 0; i < nr; i++) {
		fprintf(fp, "# sibling cores   : %s\n", str);
1135
		str += strlen(str) + 1;
1136 1137
	}

1138 1139
	nr = ph->env.nr_sibling_threads;
	str = ph->env.sibling_threads;
1140 1141 1142

	for (i = 0; i < nr; i++) {
		fprintf(fp, "# sibling threads : %s\n", str);
1143
		str += strlen(str) + 1;
1144
	}
1145 1146 1147 1148 1149 1150 1151

	if (ph->env.cpu != NULL) {
		for (i = 0; i < cpu_nr; i++)
			fprintf(fp, "# CPU %d: Core ID %d, Socket ID %d\n", i,
				ph->env.cpu[i].core_id, ph->env.cpu[i].socket_id);
	} else
		fprintf(fp, "# Core ID and Socket ID information is not available\n");
1152 1153
}

1154
static void free_event_desc(struct perf_evsel *events)
1155
{
1156 1157 1158 1159 1160 1161
	struct perf_evsel *evsel;

	if (!events)
		return;

	for (evsel = events; evsel->attr.size; evsel++) {
1162 1163
		zfree(&evsel->name);
		zfree(&evsel->id);
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173
	}

	free(events);
}

static struct perf_evsel *
read_event_desc(struct perf_header *ph, int fd)
{
	struct perf_evsel *evsel, *events = NULL;
	u64 *id;
1174
	void *buf = NULL;
1175 1176 1177
	u32 nre, sz, nr, i, j;
	ssize_t ret;
	size_t msz;
1178 1179

	/* number of events */
1180
	ret = readn(fd, &nre, sizeof(nre));
1181 1182 1183 1184 1185 1186
	if (ret != (ssize_t)sizeof(nre))
		goto error;

	if (ph->needs_swap)
		nre = bswap_32(nre);

1187
	ret = readn(fd, &sz, sizeof(sz));
1188 1189 1190 1191 1192 1193
	if (ret != (ssize_t)sizeof(sz))
		goto error;

	if (ph->needs_swap)
		sz = bswap_32(sz);

1194
	/* buffer to hold on file attr struct */
1195 1196 1197 1198
	buf = malloc(sz);
	if (!buf)
		goto error;

1199 1200 1201 1202 1203 1204
	/* the last event terminates with evsel->attr.size == 0: */
	events = calloc(nre + 1, sizeof(*events));
	if (!events)
		goto error;

	msz = sizeof(evsel->attr);
1205
	if (sz < msz)
1206 1207
		msz = sz;

1208 1209
	for (i = 0, evsel = events; i < nre; evsel++, i++) {
		evsel->idx = i;
1210

1211 1212 1213 1214
		/*
		 * must read entire on-file attr struct to
		 * sync up with layout.
		 */
1215
		ret = readn(fd, buf, sz);
1216 1217 1218 1219 1220 1221
		if (ret != (ssize_t)sz)
			goto error;

		if (ph->needs_swap)
			perf_event__attr_swap(buf);

1222
		memcpy(&evsel->attr, buf, msz);
1223

1224
		ret = readn(fd, &nr, sizeof(nr));
1225 1226 1227
		if (ret != (ssize_t)sizeof(nr))
			goto error;

1228
		if (ph->needs_swap) {
1229
			nr = bswap_32(nr);
1230 1231
			evsel->needs_swap = true;
		}
1232

1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
		evsel->name = do_read_string(fd, ph);

		if (!nr)
			continue;

		id = calloc(nr, sizeof(*id));
		if (!id)
			goto error;
		evsel->ids = nr;
		evsel->id = id;

		for (j = 0 ; j < nr; j++) {
1245
			ret = readn(fd, id, sizeof(*id));
1246 1247 1248 1249 1250 1251 1252 1253
			if (ret != (ssize_t)sizeof(*id))
				goto error;
			if (ph->needs_swap)
				*id = bswap_64(*id);
			id++;
		}
	}
out:
1254
	free(buf);
1255 1256
	return events;
error:
1257
	free_event_desc(events);
1258 1259 1260 1261
	events = NULL;
	goto out;
}

1262 1263 1264 1265 1266 1267
static int __desc_attr__fprintf(FILE *fp, const char *name, const char *val,
				void *priv __attribute__((unused)))
{
	return fprintf(fp, ", %s = %s", name, val);
}

1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
static void print_event_desc(struct perf_header *ph, int fd, FILE *fp)
{
	struct perf_evsel *evsel, *events = read_event_desc(ph, fd);
	u32 j;
	u64 *id;

	if (!events) {
		fprintf(fp, "# event desc: not available or unable to read\n");
		return;
	}

	for (evsel = events; evsel->attr.size; evsel++) {
		fprintf(fp, "# event : name = %s, ", evsel->name);
1281

1282
		if (evsel->ids) {
1283
			fprintf(fp, ", id = {");
1284 1285 1286 1287 1288
			for (j = 0, id = evsel->id; j < evsel->ids; j++, id++) {
				if (j)
					fputc(',', fp);
				fprintf(fp, " %"PRIu64, *id);
			}
1289
			fprintf(fp, " }");
1290
		}
1291

1292
		perf_event_attr__fprintf(fp, &evsel->attr, __desc_attr__fprintf, NULL);
1293

1294 1295
		fputc('\n', fp);
	}
1296 1297

	free_event_desc(events);
1298 1299
}

1300
static void print_total_mem(struct perf_header *ph, int fd __maybe_unused,
1301
			    FILE *fp)
1302
{
1303
	fprintf(fp, "# total memory : %Lu kB\n", ph->env.total_mem);
1304 1305
}

1306
static void print_numa_topology(struct perf_header *ph, int fd __maybe_unused,
1307
				FILE *fp)
1308 1309
{
	u32 nr, c, i;
1310
	char *str, *tmp;
1311 1312 1313
	uint64_t mem_total, mem_free;

	/* nr nodes */
1314 1315
	nr = ph->env.nr_numa_nodes;
	str = ph->env.numa_nodes;
1316 1317 1318

	for (i = 0; i < nr; i++) {
		/* node number */
1319 1320
		c = strtoul(str, &tmp, 0);
		if (*tmp != ':')
1321 1322
			goto error;

1323 1324 1325
		str = tmp + 1;
		mem_total = strtoull(str, &tmp, 0);
		if (*tmp != ':')
1326 1327
			goto error;

1328 1329 1330
		str = tmp + 1;
		mem_free = strtoull(str, &tmp, 0);
		if (*tmp != ':')
1331 1332 1333 1334
			goto error;

		fprintf(fp, "# node%u meminfo  : total = %"PRIu64" kB,"
			    " free = %"PRIu64" kB\n",
1335
			c, mem_total, mem_free);
1336

1337
		str = tmp + 1;
1338
		fprintf(fp, "# node%u cpu list : %s\n", c, str);
1339 1340

		str += strlen(str) + 1;
1341 1342 1343 1344 1345 1346
	}
	return;
error:
	fprintf(fp, "# numa topology : not available\n");
}

1347
static void print_cpuid(struct perf_header *ph, int fd __maybe_unused, FILE *fp)
1348
{
1349
	fprintf(fp, "# cpuid : %s\n", ph->env.cpuid);
1350 1351
}

1352
static void print_branch_stack(struct perf_header *ph __maybe_unused,
1353
			       int fd __maybe_unused, FILE *fp)
1354 1355 1356 1357
{
	fprintf(fp, "# contains samples with branch stack\n");
}

1358 1359 1360 1361 1362 1363
static void print_auxtrace(struct perf_header *ph __maybe_unused,
			   int fd __maybe_unused, FILE *fp)
{
	fprintf(fp, "# contains AUX area data (e.g. instruction trace)\n");
}

1364 1365 1366 1367 1368 1369
static void print_stat(struct perf_header *ph __maybe_unused,
		       int fd __maybe_unused, FILE *fp)
{
	fprintf(fp, "# contains stat data\n");
}

1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381
static void print_cache(struct perf_header *ph __maybe_unused,
			int fd __maybe_unused, FILE *fp __maybe_unused)
{
	int i;

	fprintf(fp, "# CPU cache info:\n");
	for (i = 0; i < ph->env.caches_cnt; i++) {
		fprintf(fp, "#  ");
		cpu_cache_level__fprintf(fp, &ph->env.caches[i]);
	}
}

1382 1383
static void print_pmu_mappings(struct perf_header *ph, int fd __maybe_unused,
			       FILE *fp)
1384 1385
{
	const char *delimiter = "# pmu mappings: ";
1386
	char *str, *tmp;
1387 1388 1389
	u32 pmu_num;
	u32 type;

1390
	pmu_num = ph->env.nr_pmu_mappings;
1391 1392 1393 1394 1395
	if (!pmu_num) {
		fprintf(fp, "# pmu mappings: not available\n");
		return;
	}

1396 1397
	str = ph->env.pmu_mappings;

1398
	while (pmu_num) {
1399 1400 1401 1402 1403 1404
		type = strtoul(str, &tmp, 0);
		if (*tmp != ':')
			goto error;

		str = tmp + 1;
		fprintf(fp, "%s%s = %" PRIu32, delimiter, str, type);
1405

1406
		delimiter = ", ";
1407 1408
		str += strlen(str) + 1;
		pmu_num--;
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418
	}

	fprintf(fp, "\n");

	if (!pmu_num)
		return;
error:
	fprintf(fp, "# pmu mappings: unable to read\n");
}

1419 1420 1421 1422 1423 1424 1425 1426 1427
static void print_group_desc(struct perf_header *ph, int fd __maybe_unused,
			     FILE *fp)
{
	struct perf_session *session;
	struct perf_evsel *evsel;
	u32 nr = 0;

	session = container_of(ph, struct perf_session, header);

1428
	evlist__for_each(session->evlist, evsel) {
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
		if (perf_evsel__is_group_leader(evsel) &&
		    evsel->nr_members > 1) {
			fprintf(fp, "# group: %s{%s", evsel->group_name ?: "",
				perf_evsel__name(evsel));

			nr = evsel->nr_members - 1;
		} else if (nr) {
			fprintf(fp, ",%s", perf_evsel__name(evsel));

			if (--nr == 0)
				fprintf(fp, "}\n");
		}
	}
}

1444 1445 1446 1447 1448 1449
static int __event_process_build_id(struct build_id_event *bev,
				    char *filename,
				    struct perf_session *session)
{
	int err = -1;
	struct machine *machine;
1450
	u16 cpumode;
1451 1452 1453 1454 1455 1456 1457
	struct dso *dso;
	enum dso_kernel_type dso_type;

	machine = perf_session__findnew_machine(session, bev->pid);
	if (!machine)
		goto out;

1458
	cpumode = bev->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
1459

1460
	switch (cpumode) {
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474
	case PERF_RECORD_MISC_KERNEL:
		dso_type = DSO_TYPE_KERNEL;
		break;
	case PERF_RECORD_MISC_GUEST_KERNEL:
		dso_type = DSO_TYPE_GUEST_KERNEL;
		break;
	case PERF_RECORD_MISC_USER:
	case PERF_RECORD_MISC_GUEST_USER:
		dso_type = DSO_TYPE_USER;
		break;
	default:
		goto out;
	}

1475
	dso = machine__findnew_dso(machine, filename);
1476 1477 1478 1479 1480
	if (dso != NULL) {
		char sbuild_id[BUILD_ID_SIZE * 2 + 1];

		dso__set_build_id(dso, &bev->build_id);

1481
		if (!is_kernel_module(filename, cpumode))
1482 1483 1484 1485 1486 1487
			dso->kernel = dso_type;

		build_id__sprintf(dso->build_id, sizeof(dso->build_id),
				  sbuild_id);
		pr_debug("build id event received for %s: %s\n",
			 dso->long_name, sbuild_id);
1488
		dso__put(dso);
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
	}

	err = 0;
out:
	return err;
}

static int perf_header__read_build_ids_abi_quirk(struct perf_header *header,
						 int input, u64 offset, u64 size)
{
	struct perf_session *session = container_of(header, struct perf_session, header);
	struct {
		struct perf_event_header   header;
1502
		u8			   build_id[PERF_ALIGN(BUILD_ID_SIZE, sizeof(u64))];
1503 1504 1505 1506 1507 1508 1509 1510 1511
		char			   filename[0];
	} old_bev;
	struct build_id_event bev;
	char filename[PATH_MAX];
	u64 limit = offset + size;

	while (offset < limit) {
		ssize_t len;

1512
		if (readn(input, &old_bev, sizeof(old_bev)) != sizeof(old_bev))
1513 1514 1515 1516 1517 1518
			return -1;

		if (header->needs_swap)
			perf_event_header__bswap(&old_bev.header);

		len = old_bev.header.size - sizeof(old_bev);
1519
		if (readn(input, filename, len) != len)
1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
			return -1;

		bev.header = old_bev.header;

		/*
		 * As the pid is the missing value, we need to fill
		 * it properly. The header.misc value give us nice hint.
		 */
		bev.pid	= HOST_KERNEL_ID;
		if (bev.header.misc == PERF_RECORD_MISC_GUEST_USER ||
		    bev.header.misc == PERF_RECORD_MISC_GUEST_KERNEL)
			bev.pid	= DEFAULT_GUEST_KERNEL_ID;

		memcpy(bev.build_id, old_bev.build_id, sizeof(bev.build_id));
		__event_process_build_id(&bev, filename, session);

		offset += bev.header.size;
	}

	return 0;
}

static int perf_header__read_build_ids(struct perf_header *header,
				       int input, u64 offset, u64 size)
{
	struct perf_session *session = container_of(header, struct perf_session, header);
	struct build_id_event bev;
	char filename[PATH_MAX];
	u64 limit = offset + size, orig_offset = offset;
	int err = -1;

	while (offset < limit) {
		ssize_t len;

1554
		if (readn(input, &bev, sizeof(bev)) != sizeof(bev))
1555 1556 1557 1558 1559 1560
			goto out;

		if (header->needs_swap)
			perf_event_header__bswap(&bev.header);

		len = bev.header.size - sizeof(bev);
1561
		if (readn(input, filename, len) != len)
1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
			goto out;
		/*
		 * The a1645ce1 changeset:
		 *
		 * "perf: 'perf kvm' tool for monitoring guest performance from host"
		 *
		 * Added a field to struct build_id_event that broke the file
		 * format.
		 *
		 * Since the kernel build-id is the first entry, process the
		 * table using the old format if the well known
		 * '[kernel.kallsyms]' string for the kernel build-id has the
		 * first 4 characters chopped off (where the pid_t sits).
		 */
		if (memcmp(filename, "nel.kallsyms]", 13) == 0) {
			if (lseek(input, orig_offset, SEEK_SET) == (off_t)-1)
				return -1;
			return perf_header__read_build_ids_abi_quirk(header, input, offset, size);
		}

		__event_process_build_id(&bev, filename, session);

		offset += bev.header.size;
	}
	err = 0;
out:
	return err;
}

1591 1592 1593
static int process_tracing_data(struct perf_file_section *section __maybe_unused,
				struct perf_header *ph __maybe_unused,
				int fd, void *data)
1594
{
1595 1596
	ssize_t ret = trace_report(fd, data, false);
	return ret < 0 ? -1 : 0;
1597 1598 1599
}

static int process_build_id(struct perf_file_section *section,
1600
			    struct perf_header *ph, int fd,
1601
			    void *data __maybe_unused)
1602 1603 1604 1605 1606 1607
{
	if (perf_header__read_build_ids(ph, fd, section->offset, section->size))
		pr_debug("Failed to read buildids, continuing...\n");
	return 0;
}

1608
static int process_hostname(struct perf_file_section *section __maybe_unused,
1609 1610
			    struct perf_header *ph, int fd,
			    void *data __maybe_unused)
1611 1612 1613 1614 1615 1616
{
	ph->env.hostname = do_read_string(fd, ph);
	return ph->env.hostname ? 0 : -ENOMEM;
}

static int process_osrelease(struct perf_file_section *section __maybe_unused,
1617 1618
			     struct perf_header *ph, int fd,
			     void *data __maybe_unused)
1619 1620 1621 1622 1623 1624
{
	ph->env.os_release = do_read_string(fd, ph);
	return ph->env.os_release ? 0 : -ENOMEM;
}

static int process_version(struct perf_file_section *section __maybe_unused,
1625 1626
			   struct perf_header *ph, int fd,
			   void *data __maybe_unused)
1627 1628 1629 1630 1631 1632
{
	ph->env.version = do_read_string(fd, ph);
	return ph->env.version ? 0 : -ENOMEM;
}

static int process_arch(struct perf_file_section *section __maybe_unused,
1633 1634
			struct perf_header *ph,	int fd,
			void *data __maybe_unused)
1635 1636 1637 1638 1639 1640
{
	ph->env.arch = do_read_string(fd, ph);
	return ph->env.arch ? 0 : -ENOMEM;
}

static int process_nrcpus(struct perf_file_section *section __maybe_unused,
1641 1642
			  struct perf_header *ph, int fd,
			  void *data __maybe_unused)
1643
{
1644
	ssize_t ret;
1645 1646
	u32 nr;

1647
	ret = readn(fd, &nr, sizeof(nr));
1648 1649 1650 1651 1652 1653
	if (ret != sizeof(nr))
		return -1;

	if (ph->needs_swap)
		nr = bswap_32(nr);

1654
	ph->env.nr_cpus_avail = nr;
1655

1656
	ret = readn(fd, &nr, sizeof(nr));
1657 1658 1659 1660 1661 1662
	if (ret != sizeof(nr))
		return -1;

	if (ph->needs_swap)
		nr = bswap_32(nr);

1663
	ph->env.nr_cpus_online = nr;
1664 1665 1666 1667
	return 0;
}

static int process_cpudesc(struct perf_file_section *section __maybe_unused,
1668 1669
			   struct perf_header *ph, int fd,
			   void *data __maybe_unused)
1670 1671 1672 1673 1674 1675
{
	ph->env.cpu_desc = do_read_string(fd, ph);
	return ph->env.cpu_desc ? 0 : -ENOMEM;
}

static int process_cpuid(struct perf_file_section *section __maybe_unused,
1676 1677
			 struct perf_header *ph,  int fd,
			 void *data __maybe_unused)
1678 1679 1680 1681 1682 1683
{
	ph->env.cpuid = do_read_string(fd, ph);
	return ph->env.cpuid ? 0 : -ENOMEM;
}

static int process_total_mem(struct perf_file_section *section __maybe_unused,
1684 1685
			     struct perf_header *ph, int fd,
			     void *data __maybe_unused)
1686 1687
{
	uint64_t mem;
1688
	ssize_t ret;
1689

1690
	ret = readn(fd, &mem, sizeof(mem));
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700
	if (ret != sizeof(mem))
		return -1;

	if (ph->needs_swap)
		mem = bswap_64(mem);

	ph->env.total_mem = mem;
	return 0;
}

1701 1702 1703 1704 1705
static struct perf_evsel *
perf_evlist__find_by_index(struct perf_evlist *evlist, int idx)
{
	struct perf_evsel *evsel;

1706
	evlist__for_each(evlist, evsel) {
1707 1708 1709 1710 1711 1712 1713 1714
		if (evsel->idx == idx)
			return evsel;
	}

	return NULL;
}

static void
1715 1716
perf_evlist__set_event_name(struct perf_evlist *evlist,
			    struct perf_evsel *event)
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733
{
	struct perf_evsel *evsel;

	if (!event->name)
		return;

	evsel = perf_evlist__find_by_index(evlist, event->idx);
	if (!evsel)
		return;

	if (evsel->name)
		return;

	evsel->name = strdup(event->name);
}

static int
1734
process_event_desc(struct perf_file_section *section __maybe_unused,
1735
		   struct perf_header *header, int fd,
1736
		   void *data __maybe_unused)
1737
{
1738
	struct perf_session *session;
1739 1740 1741 1742 1743
	struct perf_evsel *evsel, *events = read_event_desc(header, fd);

	if (!events)
		return 0;

1744
	session = container_of(header, struct perf_session, header);
1745 1746 1747 1748 1749 1750 1751 1752
	for (evsel = events; evsel->attr.size; evsel++)
		perf_evlist__set_event_name(session->evlist, evsel);

	free_event_desc(events);

	return 0;
}

1753
static int process_cmdline(struct perf_file_section *section,
1754 1755
			   struct perf_header *ph, int fd,
			   void *data __maybe_unused)
1756
{
1757
	ssize_t ret;
1758 1759
	char *str, *cmdline = NULL, **argv = NULL;
	u32 nr, i, len = 0;
1760

1761
	ret = readn(fd, &nr, sizeof(nr));
1762 1763 1764 1765 1766 1767 1768
	if (ret != sizeof(nr))
		return -1;

	if (ph->needs_swap)
		nr = bswap_32(nr);

	ph->env.nr_cmdline = nr;
1769 1770 1771 1772 1773 1774 1775 1776

	cmdline = zalloc(section->size + nr + 1);
	if (!cmdline)
		return -1;

	argv = zalloc(sizeof(char *) * (nr + 1));
	if (!argv)
		goto error;
1777 1778 1779 1780 1781 1782

	for (i = 0; i < nr; i++) {
		str = do_read_string(fd, ph);
		if (!str)
			goto error;

1783 1784 1785
		argv[i] = cmdline + len;
		memcpy(argv[i], str, strlen(str) + 1);
		len += strlen(str) + 1;
1786 1787
		free(str);
	}
1788 1789
	ph->env.cmdline = cmdline;
	ph->env.cmdline_argv = (const char **) argv;
1790 1791 1792
	return 0;

error:
1793 1794
	free(argv);
	free(cmdline);
1795 1796 1797
	return -1;
}

1798
static int process_cpu_topology(struct perf_file_section *section,
1799 1800
				struct perf_header *ph, int fd,
				void *data __maybe_unused)
1801
{
1802
	ssize_t ret;
1803 1804 1805
	u32 nr, i;
	char *str;
	struct strbuf sb;
1806 1807 1808 1809 1810 1811
	int cpu_nr = ph->env.nr_cpus_online;
	u64 size = 0;

	ph->env.cpu = calloc(cpu_nr, sizeof(*ph->env.cpu));
	if (!ph->env.cpu)
		return -1;
1812

1813
	ret = readn(fd, &nr, sizeof(nr));
1814
	if (ret != sizeof(nr))
1815
		goto free_cpu;
1816 1817 1818 1819 1820

	if (ph->needs_swap)
		nr = bswap_32(nr);

	ph->env.nr_sibling_cores = nr;
1821
	size += sizeof(u32);
1822 1823
	if (strbuf_init(&sb, 128) < 0)
		goto free_cpu;
1824 1825 1826 1827 1828 1829 1830

	for (i = 0; i < nr; i++) {
		str = do_read_string(fd, ph);
		if (!str)
			goto error;

		/* include a NULL character at the end */
1831 1832
		if (strbuf_add(&sb, str, strlen(str) + 1) < 0)
			goto error;
1833
		size += string_size(str);
1834 1835 1836 1837
		free(str);
	}
	ph->env.sibling_cores = strbuf_detach(&sb, NULL);

1838
	ret = readn(fd, &nr, sizeof(nr));
1839 1840 1841 1842 1843 1844 1845
	if (ret != sizeof(nr))
		return -1;

	if (ph->needs_swap)
		nr = bswap_32(nr);

	ph->env.nr_sibling_threads = nr;
1846
	size += sizeof(u32);
1847 1848 1849 1850 1851 1852 1853

	for (i = 0; i < nr; i++) {
		str = do_read_string(fd, ph);
		if (!str)
			goto error;

		/* include a NULL character at the end */
1854 1855
		if (strbuf_add(&sb, str, strlen(str) + 1) < 0)
			goto error;
1856
		size += string_size(str);
1857 1858 1859
		free(str);
	}
	ph->env.sibling_threads = strbuf_detach(&sb, NULL);
1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895

	/*
	 * The header may be from old perf,
	 * which doesn't include core id and socket id information.
	 */
	if (section->size <= size) {
		zfree(&ph->env.cpu);
		return 0;
	}

	for (i = 0; i < (u32)cpu_nr; i++) {
		ret = readn(fd, &nr, sizeof(nr));
		if (ret != sizeof(nr))
			goto free_cpu;

		if (ph->needs_swap)
			nr = bswap_32(nr);

		ph->env.cpu[i].core_id = nr;

		ret = readn(fd, &nr, sizeof(nr));
		if (ret != sizeof(nr))
			goto free_cpu;

		if (ph->needs_swap)
			nr = bswap_32(nr);

		if (nr > (u32)cpu_nr) {
			pr_debug("socket_id number is too big."
				 "You may need to upgrade the perf tool.\n");
			goto free_cpu;
		}

		ph->env.cpu[i].socket_id = nr;
	}

1896 1897 1898 1899
	return 0;

error:
	strbuf_release(&sb);
1900 1901
free_cpu:
	zfree(&ph->env.cpu);
1902 1903 1904 1905
	return -1;
}

static int process_numa_topology(struct perf_file_section *section __maybe_unused,
1906 1907
				 struct perf_header *ph, int fd,
				 void *data __maybe_unused)
1908
{
1909
	ssize_t ret;
1910 1911 1912 1913 1914 1915
	u32 nr, node, i;
	char *str;
	uint64_t mem_total, mem_free;
	struct strbuf sb;

	/* nr nodes */
1916
	ret = readn(fd, &nr, sizeof(nr));
1917
	if (ret != sizeof(nr))
1918
		return -1;
1919 1920 1921 1922 1923

	if (ph->needs_swap)
		nr = bswap_32(nr);

	ph->env.nr_numa_nodes = nr;
1924 1925
	if (strbuf_init(&sb, 256) < 0)
		return -1;
1926 1927 1928

	for (i = 0; i < nr; i++) {
		/* node number */
1929
		ret = readn(fd, &node, sizeof(node));
1930 1931 1932
		if (ret != sizeof(node))
			goto error;

1933
		ret = readn(fd, &mem_total, sizeof(u64));
1934 1935 1936
		if (ret != sizeof(u64))
			goto error;

1937
		ret = readn(fd, &mem_free, sizeof(u64));
1938 1939 1940 1941 1942 1943 1944 1945 1946
		if (ret != sizeof(u64))
			goto error;

		if (ph->needs_swap) {
			node = bswap_32(node);
			mem_total = bswap_64(mem_total);
			mem_free = bswap_64(mem_free);
		}

1947 1948 1949
		if (strbuf_addf(&sb, "%u:%"PRIu64":%"PRIu64":",
				node, mem_total, mem_free) < 0)
			goto error;
1950 1951 1952 1953 1954 1955

		str = do_read_string(fd, ph);
		if (!str)
			goto error;

		/* include a NULL character at the end */
1956 1957
		if (strbuf_add(&sb, str, strlen(str) + 1) < 0)
			goto error;
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
		free(str);
	}
	ph->env.numa_nodes = strbuf_detach(&sb, NULL);
	return 0;

error:
	strbuf_release(&sb);
	return -1;
}

static int process_pmu_mappings(struct perf_file_section *section __maybe_unused,
1969 1970
				struct perf_header *ph, int fd,
				void *data __maybe_unused)
1971
{
1972
	ssize_t ret;
1973 1974 1975 1976 1977
	char *name;
	u32 pmu_num;
	u32 type;
	struct strbuf sb;

1978
	ret = readn(fd, &pmu_num, sizeof(pmu_num));
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
	if (ret != sizeof(pmu_num))
		return -1;

	if (ph->needs_swap)
		pmu_num = bswap_32(pmu_num);

	if (!pmu_num) {
		pr_debug("pmu mappings not available\n");
		return 0;
	}

	ph->env.nr_pmu_mappings = pmu_num;
1991 1992
	if (strbuf_init(&sb, 128) < 0)
		return -1;
1993 1994

	while (pmu_num) {
1995
		if (readn(fd, &type, sizeof(type)) != sizeof(type))
1996 1997 1998 1999 2000 2001 2002 2003
			goto error;
		if (ph->needs_swap)
			type = bswap_32(type);

		name = do_read_string(fd, ph);
		if (!name)
			goto error;

2004 2005
		if (strbuf_addf(&sb, "%u:%s", type, name) < 0)
			goto error;
2006
		/* include a NULL character at the end */
2007 2008
		if (strbuf_add(&sb, "", 1) < 0)
			goto error;
2009

2010 2011 2012
		if (!strcmp(name, "msr"))
			ph->env.msr_pmu_type = type;

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
		free(name);
		pmu_num--;
	}
	ph->env.pmu_mappings = strbuf_detach(&sb, NULL);
	return 0;

error:
	strbuf_release(&sb);
	return -1;
}

2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077
static int process_group_desc(struct perf_file_section *section __maybe_unused,
			      struct perf_header *ph, int fd,
			      void *data __maybe_unused)
{
	size_t ret = -1;
	u32 i, nr, nr_groups;
	struct perf_session *session;
	struct perf_evsel *evsel, *leader = NULL;
	struct group_desc {
		char *name;
		u32 leader_idx;
		u32 nr_members;
	} *desc;

	if (readn(fd, &nr_groups, sizeof(nr_groups)) != sizeof(nr_groups))
		return -1;

	if (ph->needs_swap)
		nr_groups = bswap_32(nr_groups);

	ph->env.nr_groups = nr_groups;
	if (!nr_groups) {
		pr_debug("group desc not available\n");
		return 0;
	}

	desc = calloc(nr_groups, sizeof(*desc));
	if (!desc)
		return -1;

	for (i = 0; i < nr_groups; i++) {
		desc[i].name = do_read_string(fd, ph);
		if (!desc[i].name)
			goto out_free;

		if (readn(fd, &desc[i].leader_idx, sizeof(u32)) != sizeof(u32))
			goto out_free;

		if (readn(fd, &desc[i].nr_members, sizeof(u32)) != sizeof(u32))
			goto out_free;

		if (ph->needs_swap) {
			desc[i].leader_idx = bswap_32(desc[i].leader_idx);
			desc[i].nr_members = bswap_32(desc[i].nr_members);
		}
	}

	/*
	 * Rebuild group relationship based on the group_desc
	 */
	session = container_of(ph, struct perf_session, header);
	session->evlist->nr_groups = nr_groups;

	i = nr = 0;
2078
	evlist__for_each(session->evlist, evsel) {
2079 2080 2081
		if (evsel->idx == (int) desc[i].leader_idx) {
			evsel->leader = evsel;
			/* {anon_group} is a dummy name */
N
Namhyung Kim 已提交
2082
			if (strcmp(desc[i].name, "{anon_group}")) {
2083
				evsel->group_name = desc[i].name;
N
Namhyung Kim 已提交
2084 2085
				desc[i].name = NULL;
			}
2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
			evsel->nr_members = desc[i].nr_members;

			if (i >= nr_groups || nr > 0) {
				pr_debug("invalid group desc\n");
				goto out_free;
			}

			leader = evsel;
			nr = evsel->nr_members - 1;
			i++;
		} else if (nr) {
			/* This is a group member */
			evsel->leader = leader;

			nr--;
		}
	}

	if (i != nr_groups || nr != 0) {
		pr_debug("invalid group desc\n");
		goto out_free;
	}

	ret = 0;
out_free:
2111
	for (i = 0; i < nr_groups; i++)
2112
		zfree(&desc[i].name);
2113 2114 2115 2116 2117
	free(desc);

	return ret;
}

2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133
static int process_auxtrace(struct perf_file_section *section,
			    struct perf_header *ph, int fd,
			    void *data __maybe_unused)
{
	struct perf_session *session;
	int err;

	session = container_of(ph, struct perf_session, header);

	err = auxtrace_index__process(fd, section->size, session,
				      ph->needs_swap);
	if (err < 0)
		pr_err("Failed to process auxtrace index\n");
	return err;
}

2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195
static int process_cache(struct perf_file_section *section __maybe_unused,
			 struct perf_header *ph __maybe_unused, int fd __maybe_unused,
			 void *data __maybe_unused)
{
	struct cpu_cache_level *caches;
	u32 cnt, i, version;

	if (readn(fd, &version, sizeof(version)) != sizeof(version))
		return -1;

	if (ph->needs_swap)
		version = bswap_32(version);

	if (version != 1)
		return -1;

	if (readn(fd, &cnt, sizeof(cnt)) != sizeof(cnt))
		return -1;

	if (ph->needs_swap)
		cnt = bswap_32(cnt);

	caches = zalloc(sizeof(*caches) * cnt);
	if (!caches)
		return -1;

	for (i = 0; i < cnt; i++) {
		struct cpu_cache_level c;

		#define _R(v)						\
			if (readn(fd, &c.v, sizeof(u32)) != sizeof(u32))\
				goto out_free_caches;			\
			if (ph->needs_swap)				\
				c.v = bswap_32(c.v);			\

		_R(level)
		_R(line_size)
		_R(sets)
		_R(ways)
		#undef _R

		#define _R(v)				\
			c.v = do_read_string(fd, ph);	\
			if (!c.v)			\
				goto out_free_caches;

		_R(type)
		_R(size)
		_R(map)
		#undef _R

		caches[i] = c;
	}

	ph->env.caches = caches;
	ph->env.caches_cnt = cnt;
	return 0;
out_free_caches:
	free(caches);
	return -1;
}

2196 2197 2198
struct feature_ops {
	int (*write)(int fd, struct perf_header *h, struct perf_evlist *evlist);
	void (*print)(struct perf_header *h, int fd, FILE *fp);
2199
	int (*process)(struct perf_file_section *section,
2200
		       struct perf_header *h, int fd, void *data);
2201 2202 2203 2204
	const char *name;
	bool full_only;
};

2205 2206
#define FEAT_OPA(n, func) \
	[n] = { .name = #n, .write = write_##func, .print = print_##func }
2207 2208 2209
#define FEAT_OPP(n, func) \
	[n] = { .name = #n, .write = write_##func, .print = print_##func, \
		.process = process_##func }
2210
#define FEAT_OPF(n, func) \
2211
	[n] = { .name = #n, .write = write_##func, .print = print_##func, \
2212
		.process = process_##func, .full_only = true }
2213 2214

/* feature_ops not implemented: */
2215 2216
#define print_tracing_data	NULL
#define print_build_id		NULL
2217 2218

static const struct feature_ops feat_ops[HEADER_LAST_FEATURE] = {
2219
	FEAT_OPP(HEADER_TRACING_DATA,	tracing_data),
2220
	FEAT_OPP(HEADER_BUILD_ID,	build_id),
2221 2222 2223 2224 2225 2226
	FEAT_OPP(HEADER_HOSTNAME,	hostname),
	FEAT_OPP(HEADER_OSRELEASE,	osrelease),
	FEAT_OPP(HEADER_VERSION,	version),
	FEAT_OPP(HEADER_ARCH,		arch),
	FEAT_OPP(HEADER_NRCPUS,		nrcpus),
	FEAT_OPP(HEADER_CPUDESC,	cpudesc),
2227
	FEAT_OPP(HEADER_CPUID,		cpuid),
2228
	FEAT_OPP(HEADER_TOTAL_MEM,	total_mem),
2229
	FEAT_OPP(HEADER_EVENT_DESC,	event_desc),
2230
	FEAT_OPP(HEADER_CMDLINE,	cmdline),
2231 2232
	FEAT_OPF(HEADER_CPU_TOPOLOGY,	cpu_topology),
	FEAT_OPF(HEADER_NUMA_TOPOLOGY,	numa_topology),
2233
	FEAT_OPA(HEADER_BRANCH_STACK,	branch_stack),
2234
	FEAT_OPP(HEADER_PMU_MAPPINGS,	pmu_mappings),
2235
	FEAT_OPP(HEADER_GROUP_DESC,	group_desc),
2236
	FEAT_OPP(HEADER_AUXTRACE,	auxtrace),
2237
	FEAT_OPA(HEADER_STAT,		stat),
2238
	FEAT_OPF(HEADER_CACHE,		cache),
2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256
};

struct header_print_data {
	FILE *fp;
	bool full; /* extended list of headers */
};

static int perf_file_section__fprintf_info(struct perf_file_section *section,
					   struct perf_header *ph,
					   int feat, int fd, void *data)
{
	struct header_print_data *hd = data;

	if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) {
		pr_debug("Failed to lseek to %" PRIu64 " offset for feature "
				"%d, continuing...\n", section->offset, feat);
		return 0;
	}
2257
	if (feat >= HEADER_LAST_FEATURE) {
2258
		pr_warning("unknown feature %d\n", feat);
2259
		return 0;
2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276
	}
	if (!feat_ops[feat].print)
		return 0;

	if (!feat_ops[feat].full_only || hd->full)
		feat_ops[feat].print(ph, fd, hd->fp);
	else
		fprintf(hd->fp, "# %s info available, use -I to display\n",
			feat_ops[feat].name);

	return 0;
}

int perf_header__fprintf_info(struct perf_session *session, FILE *fp, bool full)
{
	struct header_print_data hd;
	struct perf_header *header = &session->header;
2277
	int fd = perf_data_file__fd(session->file);
2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293
	hd.fp = fp;
	hd.full = full;

	perf_header__process_sections(header, fd, &hd,
				      perf_file_section__fprintf_info);
	return 0;
}

static int do_write_feat(int fd, struct perf_header *h, int type,
			 struct perf_file_section **p,
			 struct perf_evlist *evlist)
{
	int err;
	int ret = 0;

	if (perf_header__has_feat(h, type)) {
2294 2295
		if (!feat_ops[type].write)
			return -1;
2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313

		(*p)->offset = lseek(fd, 0, SEEK_CUR);

		err = feat_ops[type].write(fd, h, evlist);
		if (err < 0) {
			pr_debug("failed to write feature %d\n", type);

			/* undo anything written */
			lseek(fd, (*p)->offset, SEEK_SET);

			return -1;
		}
		(*p)->size = lseek(fd, 0, SEEK_CUR) - (*p)->offset;
		(*p)++;
	}
	return ret;
}

2314
static int perf_header__adds_write(struct perf_header *header,
2315
				   struct perf_evlist *evlist, int fd)
2316
{
2317
	int nr_sections;
2318
	struct perf_file_section *feat_sec, *p;
2319 2320
	int sec_size;
	u64 sec_start;
2321
	int feat;
2322
	int err;
2323

2324
	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
2325
	if (!nr_sections)
2326
		return 0;
2327

2328
	feat_sec = p = calloc(nr_sections, sizeof(*feat_sec));
2329 2330
	if (feat_sec == NULL)
		return -ENOMEM;
2331 2332 2333

	sec_size = sizeof(*feat_sec) * nr_sections;

2334
	sec_start = header->feat_offset;
2335
	lseek(fd, sec_start + sec_size, SEEK_SET);
2336

2337 2338 2339 2340
	for_each_set_bit(feat, header->adds_features, HEADER_FEAT_BITS) {
		if (do_write_feat(fd, header, feat, &p, evlist))
			perf_header__clear_feat(header, feat);
	}
2341

2342
	lseek(fd, sec_start, SEEK_SET);
2343 2344 2345 2346
	/*
	 * may write more than needed due to dropped feature, but
	 * this is okay, reader will skip the mising entries
	 */
2347 2348 2349
	err = do_write(fd, feat_sec, sec_size);
	if (err < 0)
		pr_debug("failed to write feature section\n");
2350
	free(feat_sec);
2351
	return err;
2352
}
2353

2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372
int perf_header__write_pipe(int fd)
{
	struct perf_pipe_file_header f_header;
	int err;

	f_header = (struct perf_pipe_file_header){
		.magic	   = PERF_MAGIC,
		.size	   = sizeof(f_header),
	};

	err = do_write(fd, &f_header, sizeof(f_header));
	if (err < 0) {
		pr_debug("failed to write perf pipe header\n");
		return err;
	}

	return 0;
}

2373 2374 2375
int perf_session__write_header(struct perf_session *session,
			       struct perf_evlist *evlist,
			       int fd, bool at_exit)
2376 2377 2378
{
	struct perf_file_header f_header;
	struct perf_file_attr   f_attr;
2379
	struct perf_header *header = &session->header;
2380
	struct perf_evsel *evsel;
2381
	u64 attr_offset;
2382
	int err;
2383 2384 2385

	lseek(fd, sizeof(f_header), SEEK_SET);

2386
	evlist__for_each(session->evlist, evsel) {
2387 2388
		evsel->id_offset = lseek(fd, 0, SEEK_CUR);
		err = do_write(fd, evsel->id, evsel->ids * sizeof(u64));
2389 2390 2391 2392
		if (err < 0) {
			pr_debug("failed to write perf header\n");
			return err;
		}
2393 2394
	}

2395
	attr_offset = lseek(fd, 0, SEEK_CUR);
2396

2397
	evlist__for_each(evlist, evsel) {
2398
		f_attr = (struct perf_file_attr){
2399
			.attr = evsel->attr,
2400
			.ids  = {
2401 2402
				.offset = evsel->id_offset,
				.size   = evsel->ids * sizeof(u64),
2403 2404
			}
		};
2405 2406 2407 2408 2409
		err = do_write(fd, &f_attr, sizeof(f_attr));
		if (err < 0) {
			pr_debug("failed to write perf header attribute\n");
			return err;
		}
2410 2411
	}

2412 2413
	if (!header->data_offset)
		header->data_offset = lseek(fd, 0, SEEK_CUR);
2414
	header->feat_offset = header->data_offset + header->data_size;
2415

2416
	if (at_exit) {
2417
		err = perf_header__adds_write(header, evlist, fd);
2418 2419 2420
		if (err < 0)
			return err;
	}
2421

2422 2423 2424 2425 2426
	f_header = (struct perf_file_header){
		.magic	   = PERF_MAGIC,
		.size	   = sizeof(f_header),
		.attr_size = sizeof(f_attr),
		.attrs = {
2427
			.offset = attr_offset,
2428
			.size   = evlist->nr_entries * sizeof(f_attr),
2429 2430
		},
		.data = {
2431 2432
			.offset = header->data_offset,
			.size	= header->data_size,
2433
		},
2434
		/* event_types is ignored, store zeros */
2435 2436
	};

2437
	memcpy(&f_header.adds_features, &header->adds_features, sizeof(header->adds_features));
2438

2439
	lseek(fd, 0, SEEK_SET);
2440 2441 2442 2443 2444
	err = do_write(fd, &f_header, sizeof(f_header));
	if (err < 0) {
		pr_debug("failed to write perf header\n");
		return err;
	}
2445
	lseek(fd, header->data_offset + header->data_size, SEEK_SET);
2446

2447
	return 0;
2448 2449
}

2450
static int perf_header__getbuffer64(struct perf_header *header,
2451 2452
				    int fd, void *buf, size_t size)
{
2453
	if (readn(fd, buf, size) <= 0)
2454 2455
		return -1;

2456
	if (header->needs_swap)
2457 2458 2459 2460 2461
		mem_bswap_64(buf, size);

	return 0;
}

2462
int perf_header__process_sections(struct perf_header *header, int fd,
2463
				  void *data,
2464
				  int (*process)(struct perf_file_section *section,
2465 2466
						 struct perf_header *ph,
						 int feat, int fd, void *data))
2467
{
2468
	struct perf_file_section *feat_sec, *sec;
2469 2470
	int nr_sections;
	int sec_size;
2471 2472
	int feat;
	int err;
2473

2474
	nr_sections = bitmap_weight(header->adds_features, HEADER_FEAT_BITS);
2475
	if (!nr_sections)
2476
		return 0;
2477

2478
	feat_sec = sec = calloc(nr_sections, sizeof(*feat_sec));
2479
	if (!feat_sec)
2480
		return -1;
2481 2482 2483

	sec_size = sizeof(*feat_sec) * nr_sections;

2484
	lseek(fd, header->feat_offset, SEEK_SET);
2485

2486 2487
	err = perf_header__getbuffer64(header, fd, feat_sec, sec_size);
	if (err < 0)
2488
		goto out_free;
2489

2490 2491 2492 2493
	for_each_set_bit(feat, header->adds_features, HEADER_LAST_FEATURE) {
		err = process(sec++, header, feat, fd, data);
		if (err < 0)
			goto out_free;
2494
	}
2495
	err = 0;
2496
out_free:
2497 2498
	free(feat_sec);
	return err;
2499
}
2500

2501 2502 2503
static const int attr_file_abi_sizes[] = {
	[0] = PERF_ATTR_SIZE_VER0,
	[1] = PERF_ATTR_SIZE_VER1,
2504
	[2] = PERF_ATTR_SIZE_VER2,
2505
	[3] = PERF_ATTR_SIZE_VER3,
2506
	[4] = PERF_ATTR_SIZE_VER4,
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516
	0,
};

/*
 * In the legacy file format, the magic number is not used to encode endianness.
 * hdr_sz was used to encode endianness. But given that hdr_sz can vary based
 * on ABI revisions, we need to try all combinations for all endianness to
 * detect the endianness.
 */
static int try_all_file_abis(uint64_t hdr_sz, struct perf_header *ph)
2517
{
2518 2519
	uint64_t ref_size, attr_size;
	int i;
2520

2521 2522 2523 2524 2525 2526 2527
	for (i = 0 ; attr_file_abi_sizes[i]; i++) {
		ref_size = attr_file_abi_sizes[i]
			 + sizeof(struct perf_file_section);
		if (hdr_sz != ref_size) {
			attr_size = bswap_64(hdr_sz);
			if (attr_size != ref_size)
				continue;
2528

2529 2530 2531 2532 2533 2534 2535 2536 2537 2538
			ph->needs_swap = true;
		}
		pr_debug("ABI%d perf.data file detected, need_swap=%d\n",
			 i,
			 ph->needs_swap);
		return 0;
	}
	/* could not determine endianness */
	return -1;
}
2539

2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563
#define PERF_PIPE_HDR_VER0	16

static const size_t attr_pipe_abi_sizes[] = {
	[0] = PERF_PIPE_HDR_VER0,
	0,
};

/*
 * In the legacy pipe format, there is an implicit assumption that endiannesss
 * between host recording the samples, and host parsing the samples is the
 * same. This is not always the case given that the pipe output may always be
 * redirected into a file and analyzed on a different machine with possibly a
 * different endianness and perf_event ABI revsions in the perf tool itself.
 */
static int try_all_pipe_abis(uint64_t hdr_sz, struct perf_header *ph)
{
	u64 attr_size;
	int i;

	for (i = 0 ; attr_pipe_abi_sizes[i]; i++) {
		if (hdr_sz != attr_pipe_abi_sizes[i]) {
			attr_size = bswap_64(hdr_sz);
			if (attr_size != hdr_sz)
				continue;
2564 2565 2566

			ph->needs_swap = true;
		}
2567
		pr_debug("Pipe ABI%d perf.data file detected\n", i);
2568 2569
		return 0;
	}
2570 2571 2572
	return -1;
}

F
Feng Tang 已提交
2573 2574 2575 2576 2577 2578 2579 2580 2581 2582
bool is_perf_magic(u64 magic)
{
	if (!memcmp(&magic, __perf_magic1, sizeof(magic))
		|| magic == __perf_magic2
		|| magic == __perf_magic2_sw)
		return true;

	return false;
}

2583 2584 2585 2586 2587 2588 2589 2590
static int check_magic_endian(u64 magic, uint64_t hdr_sz,
			      bool is_pipe, struct perf_header *ph)
{
	int ret;

	/* check for legacy format */
	ret = memcmp(&magic, __perf_magic1, sizeof(magic));
	if (ret == 0) {
2591
		ph->version = PERF_HEADER_VERSION_1;
2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602
		pr_debug("legacy perf.data format\n");
		if (is_pipe)
			return try_all_pipe_abis(hdr_sz, ph);

		return try_all_file_abis(hdr_sz, ph);
	}
	/*
	 * the new magic number serves two purposes:
	 * - unique number to identify actual perf.data files
	 * - encode endianness of file
	 */
2603
	ph->version = PERF_HEADER_VERSION_2;
2604

2605 2606
	/* check magic number with one endianness */
	if (magic == __perf_magic2)
2607 2608
		return 0;

2609 2610
	/* check magic number with opposite endianness */
	if (magic != __perf_magic2_sw)
2611 2612 2613 2614 2615 2616 2617
		return -1;

	ph->needs_swap = true;

	return 0;
}

2618
int perf_file_header__read(struct perf_file_header *header,
2619 2620
			   struct perf_header *ph, int fd)
{
2621
	ssize_t ret;
2622

2623 2624
	lseek(fd, 0, SEEK_SET);

2625 2626
	ret = readn(fd, header, sizeof(*header));
	if (ret <= 0)
2627 2628
		return -1;

2629 2630 2631
	if (check_magic_endian(header->magic,
			       header->attr_size, false, ph) < 0) {
		pr_debug("magic/endian check failed\n");
2632
		return -1;
2633
	}
2634

2635
	if (ph->needs_swap) {
2636
		mem_bswap_64(header, offsetof(struct perf_file_header,
2637
			     adds_features));
2638 2639
	}

2640
	if (header->size != sizeof(*header)) {
2641
		/* Support the previous format */
2642 2643
		if (header->size == offsetof(typeof(*header), adds_features))
			bitmap_zero(header->adds_features, HEADER_FEAT_BITS);
2644 2645
		else
			return -1;
2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661
	} else if (ph->needs_swap) {
		/*
		 * feature bitmap is declared as an array of unsigned longs --
		 * not good since its size can differ between the host that
		 * generated the data file and the host analyzing the file.
		 *
		 * We need to handle endianness, but we don't know the size of
		 * the unsigned long where the file was generated. Take a best
		 * guess at determining it: try 64-bit swap first (ie., file
		 * created on a 64-bit host), and check if the hostname feature
		 * bit is set (this feature bit is forced on as of fbe96f2).
		 * If the bit is not, undo the 64-bit swap and try a 32-bit
		 * swap. If the hostname bit is still not set (e.g., older data
		 * file), punt and fallback to the original behavior --
		 * clearing all feature bits and setting buildid.
		 */
2662 2663
		mem_bswap_64(&header->adds_features,
			    BITS_TO_U64(HEADER_FEAT_BITS));
2664 2665

		if (!test_bit(HEADER_HOSTNAME, header->adds_features)) {
2666 2667 2668 2669 2670 2671 2672
			/* unswap as u64 */
			mem_bswap_64(&header->adds_features,
				    BITS_TO_U64(HEADER_FEAT_BITS));

			/* unswap as u32 */
			mem_bswap_32(&header->adds_features,
				    BITS_TO_U32(HEADER_FEAT_BITS));
2673 2674 2675 2676 2677 2678
		}

		if (!test_bit(HEADER_HOSTNAME, header->adds_features)) {
			bitmap_zero(header->adds_features, HEADER_FEAT_BITS);
			set_bit(HEADER_BUILD_ID, header->adds_features);
		}
2679
	}
2680

2681
	memcpy(&ph->adds_features, &header->adds_features,
2682
	       sizeof(ph->adds_features));
2683

2684 2685
	ph->data_offset  = header->data.offset;
	ph->data_size	 = header->data.size;
2686
	ph->feat_offset  = header->data.offset + header->data.size;
2687 2688 2689
	return 0;
}

2690
static int perf_file_section__process(struct perf_file_section *section,
2691
				      struct perf_header *ph,
2692
				      int feat, int fd, void *data)
2693
{
2694
	if (lseek(fd, section->offset, SEEK_SET) == (off_t)-1) {
2695
		pr_debug("Failed to lseek to %" PRIu64 " offset for feature "
2696
			  "%d, continuing...\n", section->offset, feat);
2697 2698 2699
		return 0;
	}

2700 2701 2702 2703 2704
	if (feat >= HEADER_LAST_FEATURE) {
		pr_debug("unknown feature %d, continuing...\n", feat);
		return 0;
	}

2705 2706
	if (!feat_ops[feat].process)
		return 0;
2707

2708
	return feat_ops[feat].process(section, ph, fd, data);
2709
}
2710

2711
static int perf_file_header__read_pipe(struct perf_pipe_file_header *header,
T
Tom Zanussi 已提交
2712 2713
				       struct perf_header *ph, int fd,
				       bool repipe)
2714
{
2715
	ssize_t ret;
2716 2717 2718 2719 2720

	ret = readn(fd, header, sizeof(*header));
	if (ret <= 0)
		return -1;

2721 2722
	if (check_magic_endian(header->magic, header->size, true, ph) < 0) {
		pr_debug("endian/magic failed\n");
2723
		return -1;
2724 2725 2726 2727
	}

	if (ph->needs_swap)
		header->size = bswap_64(header->size);
2728

2729
	if (repipe && do_write(STDOUT_FILENO, header, sizeof(*header)) < 0)
T
Tom Zanussi 已提交
2730 2731
		return -1;

2732 2733 2734
	return 0;
}

2735
static int perf_header__read_pipe(struct perf_session *session)
2736
{
2737
	struct perf_header *header = &session->header;
2738 2739
	struct perf_pipe_file_header f_header;

2740 2741
	if (perf_file_header__read_pipe(&f_header, header,
					perf_data_file__fd(session->file),
T
Tom Zanussi 已提交
2742
					session->repipe) < 0) {
2743 2744 2745 2746 2747 2748 2749
		pr_debug("incompatible file format\n");
		return -EINVAL;
	}

	return 0;
}

2750 2751 2752 2753 2754 2755
static int read_attr(int fd, struct perf_header *ph,
		     struct perf_file_attr *f_attr)
{
	struct perf_event_attr *attr = &f_attr->attr;
	size_t sz, left;
	size_t our_sz = sizeof(f_attr->attr);
2756
	ssize_t ret;
2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769

	memset(f_attr, 0, sizeof(*f_attr));

	/* read minimal guaranteed structure */
	ret = readn(fd, attr, PERF_ATTR_SIZE_VER0);
	if (ret <= 0) {
		pr_debug("cannot read %d bytes of header attr\n",
			 PERF_ATTR_SIZE_VER0);
		return -1;
	}

	/* on file perf_event_attr size */
	sz = attr->size;
2770

2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795
	if (ph->needs_swap)
		sz = bswap_32(sz);

	if (sz == 0) {
		/* assume ABI0 */
		sz =  PERF_ATTR_SIZE_VER0;
	} else if (sz > our_sz) {
		pr_debug("file uses a more recent and unsupported ABI"
			 " (%zu bytes extra)\n", sz - our_sz);
		return -1;
	}
	/* what we have not yet read and that we know about */
	left = sz - PERF_ATTR_SIZE_VER0;
	if (left) {
		void *ptr = attr;
		ptr += PERF_ATTR_SIZE_VER0;

		ret = readn(fd, ptr, left);
	}
	/* read perf_file_section, ids are read in caller */
	ret = readn(fd, &f_attr->ids, sizeof(f_attr->ids));

	return ret <= 0 ? -1 : 0;
}

2796 2797
static int perf_evsel__prepare_tracepoint_event(struct perf_evsel *evsel,
						struct pevent *pevent)
2798
{
2799
	struct event_format *event;
2800 2801
	char bf[128];

2802 2803 2804 2805
	/* already prepared */
	if (evsel->tp_format)
		return 0;

2806 2807 2808 2809 2810
	if (pevent == NULL) {
		pr_debug("broken or missing trace data\n");
		return -1;
	}

2811
	event = pevent_find_event(pevent, evsel->attr.config);
2812 2813 2814
	if (event == NULL)
		return -1;

2815 2816 2817 2818 2819 2820
	if (!evsel->name) {
		snprintf(bf, sizeof(bf), "%s:%s", event->system, event->name);
		evsel->name = strdup(bf);
		if (evsel->name == NULL)
			return -1;
	}
2821

2822
	evsel->tp_format = event;
2823 2824 2825
	return 0;
}

2826 2827
static int perf_evlist__prepare_tracepoint_events(struct perf_evlist *evlist,
						  struct pevent *pevent)
2828 2829 2830
{
	struct perf_evsel *pos;

2831
	evlist__for_each(evlist, pos) {
2832 2833
		if (pos->attr.type == PERF_TYPE_TRACEPOINT &&
		    perf_evsel__prepare_tracepoint_event(pos, pevent))
2834 2835 2836 2837 2838 2839
			return -1;
	}

	return 0;
}

2840
int perf_session__read_header(struct perf_session *session)
2841
{
2842
	struct perf_data_file *file = session->file;
2843
	struct perf_header *header = &session->header;
2844
	struct perf_file_header	f_header;
2845 2846 2847
	struct perf_file_attr	f_attr;
	u64			f_id;
	int nr_attrs, nr_ids, i, j;
2848
	int fd = perf_data_file__fd(file);
2849

2850
	session->evlist = perf_evlist__new();
2851 2852 2853
	if (session->evlist == NULL)
		return -ENOMEM;

2854
	session->evlist->env = &header->env;
2855
	session->machines.host.env = &header->env;
2856
	if (perf_data_file__is_pipe(file))
2857
		return perf_header__read_pipe(session);
2858

2859
	if (perf_file_header__read(&f_header, header, fd) < 0)
2860
		return -EINVAL;
2861

2862 2863 2864 2865 2866 2867 2868 2869 2870
	/*
	 * Sanity check that perf.data was written cleanly; data size is
	 * initialized to 0 and updated only if the on_exit function is run.
	 * If data size is still 0 then the file contains only partial
	 * information.  Just warn user and process it as much as it can.
	 */
	if (f_header.data.size == 0) {
		pr_warning("WARNING: The %s file's data size field is 0 which is unexpected.\n"
			   "Was the 'perf record' command properly terminated?\n",
2871
			   file->path);
2872 2873
	}

2874
	nr_attrs = f_header.attrs.size / f_header.attr_size;
2875 2876 2877
	lseek(fd, f_header.attrs.offset, SEEK_SET);

	for (i = 0; i < nr_attrs; i++) {
2878
		struct perf_evsel *evsel;
2879
		off_t tmp;
2880

2881
		if (read_attr(fd, header, &f_attr) < 0)
2882
			goto out_errno;
2883

2884 2885 2886
		if (header->needs_swap) {
			f_attr.ids.size   = bswap_64(f_attr.ids.size);
			f_attr.ids.offset = bswap_64(f_attr.ids.offset);
2887
			perf_event__attr_swap(&f_attr.attr);
2888
		}
2889

2890
		tmp = lseek(fd, 0, SEEK_CUR);
2891
		evsel = perf_evsel__new(&f_attr.attr);
2892

2893 2894
		if (evsel == NULL)
			goto out_delete_evlist;
2895 2896

		evsel->needs_swap = header->needs_swap;
2897 2898 2899 2900 2901
		/*
		 * Do it before so that if perf_evsel__alloc_id fails, this
		 * entry gets purged too at perf_evlist__delete().
		 */
		perf_evlist__add(session->evlist, evsel);
2902 2903

		nr_ids = f_attr.ids.size / sizeof(u64);
2904 2905 2906 2907 2908 2909 2910 2911
		/*
		 * We don't have the cpu and thread maps on the header, so
		 * for allocating the perf_sample_id table we fake 1 cpu and
		 * hattr->ids threads.
		 */
		if (perf_evsel__alloc_id(evsel, 1, nr_ids))
			goto out_delete_evlist;

2912 2913 2914
		lseek(fd, f_attr.ids.offset, SEEK_SET);

		for (j = 0; j < nr_ids; j++) {
2915
			if (perf_header__getbuffer64(header, fd, &f_id, sizeof(f_id)))
2916
				goto out_errno;
2917

2918
			perf_evlist__id_add(session->evlist, evsel, 0, j, f_id);
2919
		}
2920

2921 2922 2923
		lseek(fd, tmp, SEEK_SET);
	}

2924 2925
	symbol_conf.nr_events = nr_attrs;

J
Jiri Olsa 已提交
2926
	perf_header__process_sections(header, fd, &session->tevent,
2927
				      perf_file_section__process);
2928

2929
	if (perf_evlist__prepare_tracepoint_events(session->evlist,
J
Jiri Olsa 已提交
2930
						   session->tevent.pevent))
2931 2932
		goto out_delete_evlist;

2933
	return 0;
2934 2935
out_errno:
	return -errno;
2936 2937 2938 2939 2940

out_delete_evlist:
	perf_evlist__delete(session->evlist);
	session->evlist = NULL;
	return -ENOMEM;
2941
}
2942

2943
int perf_event__synthesize_attr(struct perf_tool *tool,
2944
				struct perf_event_attr *attr, u32 ids, u64 *id,
2945
				perf_event__handler_t process)
2946
{
2947
	union perf_event *ev;
2948 2949 2950 2951
	size_t size;
	int err;

	size = sizeof(struct perf_event_attr);
2952
	size = PERF_ALIGN(size, sizeof(u64));
2953 2954 2955 2956 2957
	size += sizeof(struct perf_event_header);
	size += ids * sizeof(u64);

	ev = malloc(size);

2958 2959 2960
	if (ev == NULL)
		return -ENOMEM;

2961 2962 2963 2964
	ev->attr.attr = *attr;
	memcpy(ev->attr.id, id, ids * sizeof(u64));

	ev->attr.header.type = PERF_RECORD_HEADER_ATTR;
2965
	ev->attr.header.size = (u16)size;
2966

2967 2968 2969 2970
	if (ev->attr.header.size == size)
		err = process(tool, ev, NULL, NULL);
	else
		err = -E2BIG;
2971 2972 2973 2974 2975 2976

	free(ev);

	return err;
}

2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013
static struct event_update_event *
event_update_event__new(size_t size, u64 type, u64 id)
{
	struct event_update_event *ev;

	size += sizeof(*ev);
	size  = PERF_ALIGN(size, sizeof(u64));

	ev = zalloc(size);
	if (ev) {
		ev->header.type = PERF_RECORD_EVENT_UPDATE;
		ev->header.size = (u16)size;
		ev->type = type;
		ev->id = id;
	}
	return ev;
}

int
perf_event__synthesize_event_update_unit(struct perf_tool *tool,
					 struct perf_evsel *evsel,
					 perf_event__handler_t process)
{
	struct event_update_event *ev;
	size_t size = strlen(evsel->unit);
	int err;

	ev = event_update_event__new(size + 1, PERF_EVENT_UPDATE__UNIT, evsel->id[0]);
	if (ev == NULL)
		return -ENOMEM;

	strncpy(ev->data, evsel->unit, size);
	err = process(tool, (union perf_event *)ev, NULL, NULL);
	free(ev);
	return err;
}

3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033
int
perf_event__synthesize_event_update_scale(struct perf_tool *tool,
					  struct perf_evsel *evsel,
					  perf_event__handler_t process)
{
	struct event_update_event *ev;
	struct event_update_event_scale *ev_data;
	int err;

	ev = event_update_event__new(sizeof(*ev_data), PERF_EVENT_UPDATE__SCALE, evsel->id[0]);
	if (ev == NULL)
		return -ENOMEM;

	ev_data = (struct event_update_event_scale *) ev->data;
	ev_data->scale = evsel->scale;
	err = process(tool, (union perf_event*) ev, NULL, NULL);
	free(ev);
	return err;
}

3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051
int
perf_event__synthesize_event_update_name(struct perf_tool *tool,
					 struct perf_evsel *evsel,
					 perf_event__handler_t process)
{
	struct event_update_event *ev;
	size_t len = strlen(evsel->name);
	int err;

	ev = event_update_event__new(len + 1, PERF_EVENT_UPDATE__NAME, evsel->id[0]);
	if (ev == NULL)
		return -ENOMEM;

	strncpy(ev->data, evsel->name, len);
	err = process(tool, (union perf_event*) ev, NULL, NULL);
	free(ev);
	return err;
}
3052

3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083
int
perf_event__synthesize_event_update_cpus(struct perf_tool *tool,
					struct perf_evsel *evsel,
					perf_event__handler_t process)
{
	size_t size = sizeof(struct event_update_event);
	struct event_update_event *ev;
	int max, err;
	u16 type;

	if (!evsel->own_cpus)
		return 0;

	ev = cpu_map_data__alloc(evsel->own_cpus, &size, &type, &max);
	if (!ev)
		return -ENOMEM;

	ev->header.type = PERF_RECORD_EVENT_UPDATE;
	ev->header.size = (u16)size;
	ev->type = PERF_EVENT_UPDATE__CPUS;
	ev->id   = evsel->id[0];

	cpu_map_data__synthesize((struct cpu_map_data *) ev->data,
				 evsel->own_cpus,
				 type, max);

	err = process(tool, (union perf_event*) ev, NULL, NULL);
	free(ev);
	return err;
}

3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121
size_t perf_event__fprintf_event_update(union perf_event *event, FILE *fp)
{
	struct event_update_event *ev = &event->event_update;
	struct event_update_event_scale *ev_scale;
	struct event_update_event_cpus *ev_cpus;
	struct cpu_map *map;
	size_t ret;

	ret = fprintf(fp, "\n... id:    %" PRIu64 "\n", ev->id);

	switch (ev->type) {
	case PERF_EVENT_UPDATE__SCALE:
		ev_scale = (struct event_update_event_scale *) ev->data;
		ret += fprintf(fp, "... scale: %f\n", ev_scale->scale);
		break;
	case PERF_EVENT_UPDATE__UNIT:
		ret += fprintf(fp, "... unit:  %s\n", ev->data);
		break;
	case PERF_EVENT_UPDATE__NAME:
		ret += fprintf(fp, "... name:  %s\n", ev->data);
		break;
	case PERF_EVENT_UPDATE__CPUS:
		ev_cpus = (struct event_update_event_cpus *) ev->data;
		ret += fprintf(fp, "... ");

		map = cpu_map__new_data(&ev_cpus->cpus);
		if (map)
			ret += cpu_map__fprintf(map, fp);
		else
			ret += fprintf(fp, "failed to get cpus\n");
		break;
	default:
		ret += fprintf(fp, "... unknown type\n");
		break;
	}

	return ret;
}
3122

3123
int perf_event__synthesize_attrs(struct perf_tool *tool,
3124
				   struct perf_session *session,
3125
				   perf_event__handler_t process)
3126
{
3127
	struct perf_evsel *evsel;
3128
	int err = 0;
3129

3130
	evlist__for_each(session->evlist, evsel) {
3131 3132
		err = perf_event__synthesize_attr(tool, &evsel->attr, evsel->ids,
						  evsel->id, process);
3133 3134 3135 3136 3137 3138 3139 3140 3141
		if (err) {
			pr_debug("failed to create perf header attribute\n");
			return err;
		}
	}

	return err;
}

3142 3143
int perf_event__process_attr(struct perf_tool *tool __maybe_unused,
			     union perf_event *event,
3144
			     struct perf_evlist **pevlist)
3145
{
3146
	u32 i, ids, n_ids;
3147
	struct perf_evsel *evsel;
3148
	struct perf_evlist *evlist = *pevlist;
3149

3150
	if (evlist == NULL) {
3151
		*pevlist = evlist = perf_evlist__new();
3152
		if (evlist == NULL)
3153 3154 3155
			return -ENOMEM;
	}

3156
	evsel = perf_evsel__new(&event->attr.attr);
3157
	if (evsel == NULL)
3158 3159
		return -ENOMEM;

3160
	perf_evlist__add(evlist, evsel);
3161

3162 3163
	ids = event->header.size;
	ids -= (void *)&event->attr.id - (void *)event;
3164
	n_ids = ids / sizeof(u64);
3165 3166 3167 3168 3169 3170 3171
	/*
	 * We don't have the cpu and thread maps on the header, so
	 * for allocating the perf_sample_id table we fake 1 cpu and
	 * hattr->ids threads.
	 */
	if (perf_evsel__alloc_id(evsel, 1, n_ids))
		return -ENOMEM;
3172 3173

	for (i = 0; i < n_ids; i++) {
3174
		perf_evlist__id_add(evlist, evsel, 0, i, event->attr.id[i]);
3175 3176
	}

3177 3178
	symbol_conf.nr_events = evlist->nr_entries;

3179 3180
	return 0;
}
3181

3182 3183 3184 3185 3186
int perf_event__process_event_update(struct perf_tool *tool __maybe_unused,
				     union perf_event *event,
				     struct perf_evlist **pevlist)
{
	struct event_update_event *ev = &event->event_update;
3187
	struct event_update_event_scale *ev_scale;
3188
	struct event_update_event_cpus *ev_cpus;
3189 3190
	struct perf_evlist *evlist;
	struct perf_evsel *evsel;
3191
	struct cpu_map *map;
3192 3193 3194 3195 3196 3197 3198 3199 3200 3201

	if (!pevlist || *pevlist == NULL)
		return -EINVAL;

	evlist = *pevlist;

	evsel = perf_evlist__id2evsel(evlist, ev->id);
	if (evsel == NULL)
		return -EINVAL;

3202 3203 3204
	switch (ev->type) {
	case PERF_EVENT_UPDATE__UNIT:
		evsel->unit = strdup(ev->data);
3205
		break;
3206 3207 3208
	case PERF_EVENT_UPDATE__NAME:
		evsel->name = strdup(ev->data);
		break;
3209 3210 3211
	case PERF_EVENT_UPDATE__SCALE:
		ev_scale = (struct event_update_event_scale *) ev->data;
		evsel->scale = ev_scale->scale;
3212 3213 3214 3215 3216 3217 3218 3219
	case PERF_EVENT_UPDATE__CPUS:
		ev_cpus = (struct event_update_event_cpus *) ev->data;

		map = cpu_map__new_data(&ev_cpus->cpus);
		if (map)
			evsel->own_cpus = map;
		else
			pr_err("failed to get event_update cpus\n");
3220 3221 3222 3223
	default:
		break;
	}

3224 3225 3226
	return 0;
}

3227
int perf_event__synthesize_tracing_data(struct perf_tool *tool, int fd,
3228
					struct perf_evlist *evlist,
3229
					perf_event__handler_t process)
3230
{
3231
	union perf_event ev;
J
Jiri Olsa 已提交
3232
	struct tracing_data *tdata;
3233
	ssize_t size = 0, aligned_size = 0, padding;
3234
	int err __maybe_unused = 0;
3235

J
Jiri Olsa 已提交
3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250
	/*
	 * We are going to store the size of the data followed
	 * by the data contents. Since the fd descriptor is a pipe,
	 * we cannot seek back to store the size of the data once
	 * we know it. Instead we:
	 *
	 * - write the tracing data to the temp file
	 * - get/write the data size to pipe
	 * - write the tracing data from the temp file
	 *   to the pipe
	 */
	tdata = tracing_data_get(&evlist->entries, fd, true);
	if (!tdata)
		return -1;

3251 3252 3253
	memset(&ev, 0, sizeof(ev));

	ev.tracing_data.header.type = PERF_RECORD_HEADER_TRACING_DATA;
J
Jiri Olsa 已提交
3254
	size = tdata->size;
3255
	aligned_size = PERF_ALIGN(size, sizeof(u64));
3256 3257 3258 3259
	padding = aligned_size - size;
	ev.tracing_data.header.size = sizeof(ev.tracing_data);
	ev.tracing_data.size = aligned_size;

3260
	process(tool, &ev, NULL, NULL);
3261

J
Jiri Olsa 已提交
3262 3263 3264 3265 3266 3267
	/*
	 * The put function will copy all the tracing data
	 * stored in temp file to the pipe.
	 */
	tracing_data_put(tdata);

3268 3269 3270 3271 3272
	write_padded(fd, NULL, 0, padding);

	return aligned_size;
}

3273 3274
int perf_event__process_tracing_data(struct perf_tool *tool __maybe_unused,
				     union perf_event *event,
3275
				     struct perf_session *session)
3276
{
3277
	ssize_t size_read, padding, size = event->tracing_data.size;
3278 3279
	int fd = perf_data_file__fd(session->file);
	off_t offset = lseek(fd, 0, SEEK_CUR);
3280 3281 3282
	char buf[BUFSIZ];

	/* setup for reading amidst mmap */
3283
	lseek(fd, offset + sizeof(struct tracing_data_event),
3284 3285
	      SEEK_SET);

J
Jiri Olsa 已提交
3286
	size_read = trace_report(fd, &session->tevent,
3287
				 session->repipe);
3288
	padding = PERF_ALIGN(size_read, sizeof(u64)) - size_read;
3289

3290
	if (readn(fd, buf, padding) < 0) {
3291 3292 3293
		pr_err("%s: reading input file", __func__);
		return -1;
	}
T
Tom Zanussi 已提交
3294 3295
	if (session->repipe) {
		int retw = write(STDOUT_FILENO, buf, padding);
3296 3297 3298 3299
		if (retw <= 0 || retw != padding) {
			pr_err("%s: repiping tracing data padding", __func__);
			return -1;
		}
T
Tom Zanussi 已提交
3300
	}
3301

3302 3303 3304 3305
	if (size_read + padding != size) {
		pr_err("%s: tracing data size mismatch", __func__);
		return -1;
	}
3306

3307
	perf_evlist__prepare_tracepoint_events(session->evlist,
J
Jiri Olsa 已提交
3308
					       session->tevent.pevent);
3309

3310 3311
	return size_read + padding;
}
3312

3313
int perf_event__synthesize_build_id(struct perf_tool *tool,
3314
				    struct dso *pos, u16 misc,
3315
				    perf_event__handler_t process,
3316
				    struct machine *machine)
3317
{
3318
	union perf_event ev;
3319 3320 3321 3322 3323 3324 3325 3326 3327
	size_t len;
	int err = 0;

	if (!pos->hit)
		return err;

	memset(&ev, 0, sizeof(ev));

	len = pos->long_name_len + 1;
3328
	len = PERF_ALIGN(len, NAME_ALIGN);
3329 3330 3331
	memcpy(&ev.build_id.build_id, pos->build_id, sizeof(pos->build_id));
	ev.build_id.header.type = PERF_RECORD_HEADER_BUILD_ID;
	ev.build_id.header.misc = misc;
3332
	ev.build_id.pid = machine->pid;
3333 3334 3335
	ev.build_id.header.size = sizeof(ev.build_id) + len;
	memcpy(&ev.build_id.filename, pos->long_name, pos->long_name_len);

3336
	err = process(tool, &ev, NULL, machine);
3337 3338 3339 3340

	return err;
}

3341
int perf_event__process_build_id(struct perf_tool *tool __maybe_unused,
3342
				 union perf_event *event,
3343
				 struct perf_session *session)
3344
{
3345 3346
	__event_process_build_id(&event->build_id,
				 event->build_id.filename,
3347
				 session);
3348 3349
	return 0;
}