machine.c 43.5 KB
Newer Older
1
#include "callchain.h"
2 3
#include "debug.h"
#include "event.h"
4 5
#include "evsel.h"
#include "hist.h"
6 7
#include "machine.h"
#include "map.h"
8
#include "sort.h"
9
#include "strlist.h"
10
#include "thread.h"
11
#include "vdso.h"
12
#include <stdbool.h>
13
#include <symbol/kallsyms.h>
14
#include "unwind.h"
15
#include "linux/hash.h"
16

17 18 19 20 21 22
static void dsos__init(struct dsos *dsos)
{
	INIT_LIST_HEAD(&dsos->head);
	dsos->root = RB_ROOT;
}

23 24
int machine__init(struct machine *machine, const char *root_dir, pid_t pid)
{
25
	map_groups__init(&machine->kmaps, machine);
26
	RB_CLEAR_NODE(&machine->rb_node);
27 28
	dsos__init(&machine->user_dsos);
	dsos__init(&machine->kernel_dsos);
29 30 31 32 33

	machine->threads = RB_ROOT;
	INIT_LIST_HEAD(&machine->dead_threads);
	machine->last_match = NULL;

34 35
	machine->vdso_info = NULL;

36 37
	machine->pid = pid;

38
	machine->symbol_filter = NULL;
39
	machine->id_hdr_size = 0;
40
	machine->comm_exec = false;
41
	machine->kernel_start = 0;
42

43 44 45 46 47
	machine->root_dir = strdup(root_dir);
	if (machine->root_dir == NULL)
		return -ENOMEM;

	if (pid != HOST_KERNEL_ID) {
48
		struct thread *thread = machine__findnew_thread(machine, -1,
49
								pid);
50 51 52 53 54 55
		char comm[64];

		if (thread == NULL)
			return -ENOMEM;

		snprintf(comm, sizeof(comm), "[guest/%d]", pid);
56
		thread__set_comm(thread, comm, 0);
57 58
	}

59 60
	machine->current_tid = NULL;

61 62 63
	return 0;
}

64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
struct machine *machine__new_host(void)
{
	struct machine *machine = malloc(sizeof(*machine));

	if (machine != NULL) {
		machine__init(machine, "", HOST_KERNEL_ID);

		if (machine__create_kernel_maps(machine) < 0)
			goto out_delete;
	}

	return machine;
out_delete:
	free(machine);
	return NULL;
}

81
static void dsos__delete(struct dsos *dsos)
82 83 84
{
	struct dso *pos, *n;

85
	list_for_each_entry_safe(pos, n, &dsos->head, node) {
86
		RB_CLEAR_NODE(&pos->rb_node);
87 88 89 90 91
		list_del(&pos->node);
		dso__delete(pos);
	}
}

92 93 94 95 96 97 98 99
void machine__delete_threads(struct machine *machine)
{
	struct rb_node *nd = rb_first(&machine->threads);

	while (nd) {
		struct thread *t = rb_entry(nd, struct thread, rb_node);

		nd = rb_next(nd);
100
		machine__remove_thread(machine, t);
101 102 103
	}
}

104 105 106 107 108
void machine__exit(struct machine *machine)
{
	map_groups__exit(&machine->kmaps);
	dsos__delete(&machine->user_dsos);
	dsos__delete(&machine->kernel_dsos);
109
	vdso__exit(machine);
110
	zfree(&machine->root_dir);
111
	zfree(&machine->current_tid);
112 113 114 115 116 117 118 119
}

void machine__delete(struct machine *machine)
{
	machine__exit(machine);
	free(machine);
}

120 121 122 123
void machines__init(struct machines *machines)
{
	machine__init(&machines->host, "", HOST_KERNEL_ID);
	machines->guests = RB_ROOT;
124
	machines->symbol_filter = NULL;
125 126 127 128 129 130 131 132 133
}

void machines__exit(struct machines *machines)
{
	machine__exit(&machines->host);
	/* XXX exit guest */
}

struct machine *machines__add(struct machines *machines, pid_t pid,
134 135
			      const char *root_dir)
{
136
	struct rb_node **p = &machines->guests.rb_node;
137 138 139 140 141 142 143 144 145 146 147
	struct rb_node *parent = NULL;
	struct machine *pos, *machine = malloc(sizeof(*machine));

	if (machine == NULL)
		return NULL;

	if (machine__init(machine, root_dir, pid) != 0) {
		free(machine);
		return NULL;
	}

148 149
	machine->symbol_filter = machines->symbol_filter;

150 151 152 153 154 155 156 157 158 159
	while (*p != NULL) {
		parent = *p;
		pos = rb_entry(parent, struct machine, rb_node);
		if (pid < pos->pid)
			p = &(*p)->rb_left;
		else
			p = &(*p)->rb_right;
	}

	rb_link_node(&machine->rb_node, parent, p);
160
	rb_insert_color(&machine->rb_node, &machines->guests);
161 162 163 164

	return machine;
}

165 166 167 168 169 170 171 172 173 174 175 176 177 178 179
void machines__set_symbol_filter(struct machines *machines,
				 symbol_filter_t symbol_filter)
{
	struct rb_node *nd;

	machines->symbol_filter = symbol_filter;
	machines->host.symbol_filter = symbol_filter;

	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
		struct machine *machine = rb_entry(nd, struct machine, rb_node);

		machine->symbol_filter = symbol_filter;
	}
}

180 181 182 183 184 185 186 187 188 189 190 191 192
void machines__set_comm_exec(struct machines *machines, bool comm_exec)
{
	struct rb_node *nd;

	machines->host.comm_exec = comm_exec;

	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
		struct machine *machine = rb_entry(nd, struct machine, rb_node);

		machine->comm_exec = comm_exec;
	}
}

193
struct machine *machines__find(struct machines *machines, pid_t pid)
194
{
195
	struct rb_node **p = &machines->guests.rb_node;
196 197 198 199
	struct rb_node *parent = NULL;
	struct machine *machine;
	struct machine *default_machine = NULL;

200 201 202
	if (pid == HOST_KERNEL_ID)
		return &machines->host;

203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218
	while (*p != NULL) {
		parent = *p;
		machine = rb_entry(parent, struct machine, rb_node);
		if (pid < machine->pid)
			p = &(*p)->rb_left;
		else if (pid > machine->pid)
			p = &(*p)->rb_right;
		else
			return machine;
		if (!machine->pid)
			default_machine = machine;
	}

	return default_machine;
}

219
struct machine *machines__findnew(struct machines *machines, pid_t pid)
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252
{
	char path[PATH_MAX];
	const char *root_dir = "";
	struct machine *machine = machines__find(machines, pid);

	if (machine && (machine->pid == pid))
		goto out;

	if ((pid != HOST_KERNEL_ID) &&
	    (pid != DEFAULT_GUEST_KERNEL_ID) &&
	    (symbol_conf.guestmount)) {
		sprintf(path, "%s/%d", symbol_conf.guestmount, pid);
		if (access(path, R_OK)) {
			static struct strlist *seen;

			if (!seen)
				seen = strlist__new(true, NULL);

			if (!strlist__has_entry(seen, path)) {
				pr_err("Can't access file %s\n", path);
				strlist__add(seen, path);
			}
			machine = NULL;
			goto out;
		}
		root_dir = path;
	}

	machine = machines__add(machines, pid, root_dir);
out:
	return machine;
}

253 254
void machines__process_guests(struct machines *machines,
			      machine__process_t process, void *data)
255 256 257
{
	struct rb_node *nd;

258
	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
		struct machine *pos = rb_entry(nd, struct machine, rb_node);
		process(pos, data);
	}
}

char *machine__mmap_name(struct machine *machine, char *bf, size_t size)
{
	if (machine__is_host(machine))
		snprintf(bf, size, "[%s]", "kernel.kallsyms");
	else if (machine__is_default_guest(machine))
		snprintf(bf, size, "[%s]", "guest.kernel.kallsyms");
	else {
		snprintf(bf, size, "[%s.%d]", "guest.kernel.kallsyms",
			 machine->pid);
	}

	return bf;
}

278
void machines__set_id_hdr_size(struct machines *machines, u16 id_hdr_size)
279 280 281 282
{
	struct rb_node *node;
	struct machine *machine;

283 284 285
	machines->host.id_hdr_size = id_hdr_size;

	for (node = rb_first(&machines->guests); node; node = rb_next(node)) {
286 287 288 289 290 291 292
		machine = rb_entry(node, struct machine, rb_node);
		machine->id_hdr_size = id_hdr_size;
	}

	return;
}

293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310
static void machine__update_thread_pid(struct machine *machine,
				       struct thread *th, pid_t pid)
{
	struct thread *leader;

	if (pid == th->pid_ || pid == -1 || th->pid_ != -1)
		return;

	th->pid_ = pid;

	if (th->pid_ == th->tid)
		return;

	leader = machine__findnew_thread(machine, th->pid_, th->pid_);
	if (!leader)
		goto out_err;

	if (!leader->mg)
311
		leader->mg = map_groups__new(machine);
312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338

	if (!leader->mg)
		goto out_err;

	if (th->mg == leader->mg)
		return;

	if (th->mg) {
		/*
		 * Maps are created from MMAP events which provide the pid and
		 * tid.  Consequently there never should be any maps on a thread
		 * with an unknown pid.  Just print an error if there are.
		 */
		if (!map_groups__empty(th->mg))
			pr_err("Discarding thread maps for %d:%d\n",
			       th->pid_, th->tid);
		map_groups__delete(th->mg);
	}

	th->mg = map_groups__get(leader->mg);

	return;

out_err:
	pr_err("Failed to join map groups for %d:%d\n", th->pid_, th->tid);
}

339 340
static struct thread *__machine__findnew_thread(struct machine *machine,
						pid_t pid, pid_t tid,
341 342 343 344 345 346 347
						bool create)
{
	struct rb_node **p = &machine->threads.rb_node;
	struct rb_node *parent = NULL;
	struct thread *th;

	/*
348
	 * Front-end cache - TID lookups come in blocks,
349 350 351
	 * so most of the time we dont have to look up
	 * the full rbtree:
	 */
352
	th = machine->last_match;
353 354 355 356 357 358 359
	if (th != NULL) {
		if (th->tid == tid) {
			machine__update_thread_pid(machine, th, pid);
			return th;
		}

		thread__zput(machine->last_match);
360
	}
361 362 363 364 365

	while (*p != NULL) {
		parent = *p;
		th = rb_entry(parent, struct thread, rb_node);

366
		if (th->tid == tid) {
367
			machine->last_match = thread__get(th);
368
			machine__update_thread_pid(machine, th, pid);
369 370 371
			return th;
		}

372
		if (tid < th->tid)
373 374 375 376 377 378 379 380
			p = &(*p)->rb_left;
		else
			p = &(*p)->rb_right;
	}

	if (!create)
		return NULL;

381
	th = thread__new(pid, tid);
382 383 384
	if (th != NULL) {
		rb_link_node(&th->rb_node, parent, p);
		rb_insert_color(&th->rb_node, &machine->threads);
385 386 387 388 389 390 391 392 393

		/*
		 * We have to initialize map_groups separately
		 * after rb tree is updated.
		 *
		 * The reason is that we call machine__findnew_thread
		 * within thread__init_map_groups to find the thread
		 * leader and that would screwed the rb tree.
		 */
394
		if (thread__init_map_groups(th, machine)) {
395
			rb_erase(&th->rb_node, &machine->threads);
396
			thread__delete(th);
397
			return NULL;
398
		}
399 400 401 402 403
		/*
		 * It is now in the rbtree, get a ref
		 */
		thread__get(th);
		machine->last_match = thread__get(th);
404 405 406 407 408
	}

	return th;
}

409 410
struct thread *machine__findnew_thread(struct machine *machine, pid_t pid,
				       pid_t tid)
411
{
412
	return __machine__findnew_thread(machine, pid, tid, true);
413 414
}

415 416
struct thread *machine__find_thread(struct machine *machine, pid_t pid,
				    pid_t tid)
417
{
418
	return __machine__findnew_thread(machine, pid, tid, false);
419
}
420

421 422 423 424 425 426 427 428 429
struct comm *machine__thread_exec_comm(struct machine *machine,
				       struct thread *thread)
{
	if (machine->comm_exec)
		return thread__exec_comm(thread);
	else
		return thread__comm(thread);
}

430 431
int machine__process_comm_event(struct machine *machine, union perf_event *event,
				struct perf_sample *sample)
432
{
433 434 435
	struct thread *thread = machine__findnew_thread(machine,
							event->comm.pid,
							event->comm.tid);
436
	bool exec = event->header.misc & PERF_RECORD_MISC_COMM_EXEC;
437

438 439 440
	if (exec)
		machine->comm_exec = true;

441 442 443
	if (dump_trace)
		perf_event__fprintf_comm(event, stdout);

444 445
	if (thread == NULL ||
	    __thread__set_comm(thread, event->comm.comm, sample->time, exec)) {
446 447 448 449 450 451 452 453
		dump_printf("problem processing PERF_RECORD_COMM, skipping event.\n");
		return -1;
	}

	return 0;
}

int machine__process_lost_event(struct machine *machine __maybe_unused,
454
				union perf_event *event, struct perf_sample *sample __maybe_unused)
455 456 457 458 459 460
{
	dump_printf(": id:%" PRIu64 ": lost:%" PRIu64 "\n",
		    event->lost.id, event->lost.lost);
	return 0;
}

461 462 463
static struct dso*
machine__module_dso(struct machine *machine, struct kmod_path *m,
		    const char *filename)
464 465 466
{
	struct dso *dso;

467
	dso = dsos__find(&machine->kernel_dsos, m->name, true);
468
	if (!dso) {
469
		dso = dsos__addnew(&machine->kernel_dsos, m->name);
470 471 472 473 474 475 476 477 478
		if (dso == NULL)
			return NULL;

		if (machine__is_host(machine))
			dso->symtab_type = DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE;
		else
			dso->symtab_type = DSO_BINARY_TYPE__GUEST_KMODULE;

		/* _KMODULE_COMP should be next to _KMODULE */
479
		if (m->kmod && m->comp)
480
			dso->symtab_type++;
481 482 483

		dso__set_short_name(dso, strdup(m->name), true);
		dso__set_long_name(dso, strdup(filename), true);
484 485 486 487 488
	}

	return dso;
}

489 490 491
struct map *machine__new_module(struct machine *machine, u64 start,
				const char *filename)
{
492 493 494
	struct map *map = NULL;
	struct dso *dso;
	struct kmod_path m;
495

496
	if (kmod_path__parse_name(&m, filename))
497 498
		return NULL;

499 500 501 502 503
	map = map_groups__find_by_name(&machine->kmaps, MAP__FUNCTION,
				       m.name);
	if (map)
		goto out;

504 505 506 507
	dso = machine__module_dso(machine, &m, filename);
	if (dso == NULL)
		goto out;

508 509
	map = map__new2(start, dso, MAP__FUNCTION);
	if (map == NULL)
510
		goto out;
511 512

	map_groups__insert(&machine->kmaps, map);
513 514 515

out:
	free(m.name);
516 517 518
	return map;
}

519
size_t machines__fprintf_dsos(struct machines *machines, FILE *fp)
520 521
{
	struct rb_node *nd;
522 523
	size_t ret = __dsos__fprintf(&machines->host.kernel_dsos.head, fp) +
		     __dsos__fprintf(&machines->host.user_dsos.head, fp);
524

525
	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
526
		struct machine *pos = rb_entry(nd, struct machine, rb_node);
527 528
		ret += __dsos__fprintf(&pos->kernel_dsos.head, fp);
		ret += __dsos__fprintf(&pos->user_dsos.head, fp);
529 530 531 532 533
	}

	return ret;
}

534
size_t machine__fprintf_dsos_buildid(struct machine *m, FILE *fp,
535 536
				     bool (skip)(struct dso *dso, int parm), int parm)
{
537 538
	return __dsos__fprintf_buildid(&m->kernel_dsos.head, fp, skip, parm) +
	       __dsos__fprintf_buildid(&m->user_dsos.head, fp, skip, parm);
539 540
}

541
size_t machines__fprintf_dsos_buildid(struct machines *machines, FILE *fp,
542 543 544
				     bool (skip)(struct dso *dso, int parm), int parm)
{
	struct rb_node *nd;
545
	size_t ret = machine__fprintf_dsos_buildid(&machines->host, fp, skip, parm);
546

547
	for (nd = rb_first(&machines->guests); nd; nd = rb_next(nd)) {
548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623
		struct machine *pos = rb_entry(nd, struct machine, rb_node);
		ret += machine__fprintf_dsos_buildid(pos, fp, skip, parm);
	}
	return ret;
}

size_t machine__fprintf_vmlinux_path(struct machine *machine, FILE *fp)
{
	int i;
	size_t printed = 0;
	struct dso *kdso = machine->vmlinux_maps[MAP__FUNCTION]->dso;

	if (kdso->has_build_id) {
		char filename[PATH_MAX];
		if (dso__build_id_filename(kdso, filename, sizeof(filename)))
			printed += fprintf(fp, "[0] %s\n", filename);
	}

	for (i = 0; i < vmlinux_path__nr_entries; ++i)
		printed += fprintf(fp, "[%d] %s\n",
				   i + kdso->has_build_id, vmlinux_path[i]);

	return printed;
}

size_t machine__fprintf(struct machine *machine, FILE *fp)
{
	size_t ret = 0;
	struct rb_node *nd;

	for (nd = rb_first(&machine->threads); nd; nd = rb_next(nd)) {
		struct thread *pos = rb_entry(nd, struct thread, rb_node);

		ret += thread__fprintf(pos, fp);
	}

	return ret;
}

static struct dso *machine__get_kernel(struct machine *machine)
{
	const char *vmlinux_name = NULL;
	struct dso *kernel;

	if (machine__is_host(machine)) {
		vmlinux_name = symbol_conf.vmlinux_name;
		if (!vmlinux_name)
			vmlinux_name = "[kernel.kallsyms]";

		kernel = dso__kernel_findnew(machine, vmlinux_name,
					     "[kernel]",
					     DSO_TYPE_KERNEL);
	} else {
		char bf[PATH_MAX];

		if (machine__is_default_guest(machine))
			vmlinux_name = symbol_conf.default_guest_vmlinux_name;
		if (!vmlinux_name)
			vmlinux_name = machine__mmap_name(machine, bf,
							  sizeof(bf));

		kernel = dso__kernel_findnew(machine, vmlinux_name,
					     "[guest.kernel]",
					     DSO_TYPE_GUEST_KERNEL);
	}

	if (kernel != NULL && (!kernel->has_build_id))
		dso__read_running_kernel_build_id(kernel, machine);

	return kernel;
}

struct process_args {
	u64 start;
};

624 625 626 627 628 629 630 631 632
static void machine__get_kallsyms_filename(struct machine *machine, char *buf,
					   size_t bufsz)
{
	if (machine__is_default_guest(machine))
		scnprintf(buf, bufsz, "%s", symbol_conf.default_guest_kallsyms);
	else
		scnprintf(buf, bufsz, "%s/proc/kallsyms", machine->root_dir);
}

633 634 635 636 637 638
const char *ref_reloc_sym_names[] = {"_text", "_stext", NULL};

/* Figure out the start address of kernel map from /proc/kallsyms.
 * Returns the name of the start symbol in *symbol_name. Pass in NULL as
 * symbol_name if it's not that important.
 */
639 640
static u64 machine__get_running_kernel_start(struct machine *machine,
					     const char **symbol_name)
641
{
642
	char filename[PATH_MAX];
643 644 645
	int i;
	const char *name;
	u64 addr = 0;
646

647
	machine__get_kallsyms_filename(machine, filename, PATH_MAX);
648 649 650 651

	if (symbol__restricted_filename(filename, "/proc/kallsyms"))
		return 0;

652 653 654 655 656 657 658 659
	for (i = 0; (name = ref_reloc_sym_names[i]) != NULL; i++) {
		addr = kallsyms__get_function_start(filename, name);
		if (addr)
			break;
	}

	if (symbol_name)
		*symbol_name = name;
660

661
	return addr;
662 663 664 665 666
}

int __machine__create_kernel_maps(struct machine *machine, struct dso *kernel)
{
	enum map_type type;
667
	u64 start = machine__get_running_kernel_start(machine, NULL);
668 669 670 671 672 673 674 675 676 677 678 679

	for (type = 0; type < MAP__NR_TYPES; ++type) {
		struct kmap *kmap;

		machine->vmlinux_maps[type] = map__new2(start, kernel, type);
		if (machine->vmlinux_maps[type] == NULL)
			return -1;

		machine->vmlinux_maps[type]->map_ip =
			machine->vmlinux_maps[type]->unmap_ip =
				identity__map_ip;
		kmap = map__kmap(machine->vmlinux_maps[type]);
680 681 682
		if (!kmap)
			return -1;

683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703
		kmap->kmaps = &machine->kmaps;
		map_groups__insert(&machine->kmaps,
				   machine->vmlinux_maps[type]);
	}

	return 0;
}

void machine__destroy_kernel_maps(struct machine *machine)
{
	enum map_type type;

	for (type = 0; type < MAP__NR_TYPES; ++type) {
		struct kmap *kmap;

		if (machine->vmlinux_maps[type] == NULL)
			continue;

		kmap = map__kmap(machine->vmlinux_maps[type]);
		map_groups__remove(&machine->kmaps,
				   machine->vmlinux_maps[type]);
704
		if (kmap && kmap->ref_reloc_sym) {
705 706 707 708 709
			/*
			 * ref_reloc_sym is shared among all maps, so free just
			 * on one of them.
			 */
			if (type == MAP__FUNCTION) {
710 711 712 713
				zfree((char **)&kmap->ref_reloc_sym->name);
				zfree(&kmap->ref_reloc_sym);
			} else
				kmap->ref_reloc_sym = NULL;
714 715 716 717 718 719 720
		}

		map__delete(machine->vmlinux_maps[type]);
		machine->vmlinux_maps[type] = NULL;
	}
}

721
int machines__create_guest_kernel_maps(struct machines *machines)
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769
{
	int ret = 0;
	struct dirent **namelist = NULL;
	int i, items = 0;
	char path[PATH_MAX];
	pid_t pid;
	char *endp;

	if (symbol_conf.default_guest_vmlinux_name ||
	    symbol_conf.default_guest_modules ||
	    symbol_conf.default_guest_kallsyms) {
		machines__create_kernel_maps(machines, DEFAULT_GUEST_KERNEL_ID);
	}

	if (symbol_conf.guestmount) {
		items = scandir(symbol_conf.guestmount, &namelist, NULL, NULL);
		if (items <= 0)
			return -ENOENT;
		for (i = 0; i < items; i++) {
			if (!isdigit(namelist[i]->d_name[0])) {
				/* Filter out . and .. */
				continue;
			}
			pid = (pid_t)strtol(namelist[i]->d_name, &endp, 10);
			if ((*endp != '\0') ||
			    (endp == namelist[i]->d_name) ||
			    (errno == ERANGE)) {
				pr_debug("invalid directory (%s). Skipping.\n",
					 namelist[i]->d_name);
				continue;
			}
			sprintf(path, "%s/%s/proc/kallsyms",
				symbol_conf.guestmount,
				namelist[i]->d_name);
			ret = access(path, R_OK);
			if (ret) {
				pr_debug("Can't access file %s\n", path);
				goto failure;
			}
			machines__create_kernel_maps(machines, pid);
		}
failure:
		free(namelist);
	}

	return ret;
}

770
void machines__destroy_kernel_maps(struct machines *machines)
771
{
772 773 774
	struct rb_node *next = rb_first(&machines->guests);

	machine__destroy_kernel_maps(&machines->host);
775 776 777 778 779

	while (next) {
		struct machine *pos = rb_entry(next, struct machine, rb_node);

		next = rb_next(&pos->rb_node);
780
		rb_erase(&pos->rb_node, &machines->guests);
781 782 783 784
		machine__delete(pos);
	}
}

785
int machines__create_kernel_maps(struct machines *machines, pid_t pid)
786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819
{
	struct machine *machine = machines__findnew(machines, pid);

	if (machine == NULL)
		return -1;

	return machine__create_kernel_maps(machine);
}

int machine__load_kallsyms(struct machine *machine, const char *filename,
			   enum map_type type, symbol_filter_t filter)
{
	struct map *map = machine->vmlinux_maps[type];
	int ret = dso__load_kallsyms(map->dso, filename, map, filter);

	if (ret > 0) {
		dso__set_loaded(map->dso, type);
		/*
		 * Since /proc/kallsyms will have multiple sessions for the
		 * kernel, with modules between them, fixup the end of all
		 * sections.
		 */
		__map_groups__fixup_end(&machine->kmaps, type);
	}

	return ret;
}

int machine__load_vmlinux_path(struct machine *machine, enum map_type type,
			       symbol_filter_t filter)
{
	struct map *map = machine->vmlinux_maps[type];
	int ret = dso__load_vmlinux_path(map->dso, map, filter);

820
	if (ret > 0)
821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
		dso__set_loaded(map->dso, type);

	return ret;
}

static void map_groups__fixup_end(struct map_groups *mg)
{
	int i;
	for (i = 0; i < MAP__NR_TYPES; ++i)
		__map_groups__fixup_end(mg, i);
}

static char *get_kernel_version(const char *root_dir)
{
	char version[PATH_MAX];
	FILE *file;
	char *name, *tmp;
	const char *prefix = "Linux version ";

	sprintf(version, "%s/proc/version", root_dir);
	file = fopen(version, "r");
	if (!file)
		return NULL;

	version[0] = '\0';
	tmp = fgets(version, sizeof(version), file);
	fclose(file);

	name = strstr(version, prefix);
	if (!name)
		return NULL;
	name += strlen(prefix);
	tmp = strchr(name, ' ');
	if (tmp)
		*tmp = '\0';

	return strdup(name);
}

860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892
static bool is_kmod_dso(struct dso *dso)
{
	return dso->symtab_type == DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE ||
	       dso->symtab_type == DSO_BINARY_TYPE__GUEST_KMODULE;
}

static int map_groups__set_module_path(struct map_groups *mg, const char *path,
				       struct kmod_path *m)
{
	struct map *map;
	char *long_name;

	map = map_groups__find_by_name(mg, MAP__FUNCTION, m->name);
	if (map == NULL)
		return 0;

	long_name = strdup(path);
	if (long_name == NULL)
		return -ENOMEM;

	dso__set_long_name(map->dso, long_name, true);
	dso__kernel_module_get_build_id(map->dso, "");

	/*
	 * Full name could reveal us kmod compression, so
	 * we need to update the symtab_type if needed.
	 */
	if (m->comp && is_kmod_dso(map->dso))
		map->dso->symtab_type++;

	return 0;
}

893
static int map_groups__set_modules_path_dir(struct map_groups *mg,
894
				const char *dir_name, int depth)
895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918
{
	struct dirent *dent;
	DIR *dir = opendir(dir_name);
	int ret = 0;

	if (!dir) {
		pr_debug("%s: cannot open %s dir\n", __func__, dir_name);
		return -1;
	}

	while ((dent = readdir(dir)) != NULL) {
		char path[PATH_MAX];
		struct stat st;

		/*sshfs might return bad dent->d_type, so we have to stat*/
		snprintf(path, sizeof(path), "%s/%s", dir_name, dent->d_name);
		if (stat(path, &st))
			continue;

		if (S_ISDIR(st.st_mode)) {
			if (!strcmp(dent->d_name, ".") ||
			    !strcmp(dent->d_name, ".."))
				continue;

919 920 921 922 923 924 925 926 927
			/* Do not follow top-level source and build symlinks */
			if (depth == 0) {
				if (!strcmp(dent->d_name, "source") ||
				    !strcmp(dent->d_name, "build"))
					continue;
			}

			ret = map_groups__set_modules_path_dir(mg, path,
							       depth + 1);
928 929 930
			if (ret < 0)
				goto out;
		} else {
931
			struct kmod_path m;
932

933 934 935
			ret = kmod_path__parse_name(&m, dent->d_name);
			if (ret)
				goto out;
936

937 938
			if (m.kmod)
				ret = map_groups__set_module_path(mg, path, &m);
939

940
			free(m.name);
941

942
			if (ret)
943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960
				goto out;
		}
	}

out:
	closedir(dir);
	return ret;
}

static int machine__set_modules_path(struct machine *machine)
{
	char *version;
	char modules_path[PATH_MAX];

	version = get_kernel_version(machine->root_dir);
	if (!version)
		return -1;

961
	snprintf(modules_path, sizeof(modules_path), "%s/lib/modules/%s",
962 963 964
		 machine->root_dir, version);
	free(version);

965
	return map_groups__set_modules_path_dir(&machine->kmaps, modules_path, 0);
966 967
}

968
static int machine__create_module(void *arg, const char *name, u64 start)
969
{
970
	struct machine *machine = arg;
971
	struct map *map;
972 973 974 975 976 977 978 979 980 981 982 983

	map = machine__new_module(machine, start, name);
	if (map == NULL)
		return -1;

	dso__kernel_module_get_build_id(map->dso, machine->root_dir);

	return 0;
}

static int machine__create_modules(struct machine *machine)
{
984 985 986
	const char *modules;
	char path[PATH_MAX];

987
	if (machine__is_default_guest(machine)) {
988
		modules = symbol_conf.default_guest_modules;
989 990
	} else {
		snprintf(path, PATH_MAX, "%s/proc/modules", machine->root_dir);
991 992 993
		modules = path;
	}

994
	if (symbol__restricted_filename(modules, "/proc/modules"))
995 996
		return -1;

997
	if (modules__parse(modules, machine, machine__create_module))
998 999
		return -1;

1000 1001
	if (!machine__set_modules_path(machine))
		return 0;
1002

1003
	pr_debug("Problems setting modules path maps, continuing anyway...\n");
1004

1005
	return 0;
1006 1007 1008 1009 1010
}

int machine__create_kernel_maps(struct machine *machine)
{
	struct dso *kernel = machine__get_kernel(machine);
1011
	const char *name;
1012
	u64 addr = machine__get_running_kernel_start(machine, &name);
1013 1014
	if (!addr)
		return -1;
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032

	if (kernel == NULL ||
	    __machine__create_kernel_maps(machine, kernel) < 0)
		return -1;

	if (symbol_conf.use_modules && machine__create_modules(machine) < 0) {
		if (machine__is_host(machine))
			pr_debug("Problems creating module maps, "
				 "continuing anyway...\n");
		else
			pr_debug("Problems creating module maps for guest %d, "
				 "continuing anyway...\n", machine->pid);
	}

	/*
	 * Now that we have all the maps created, just set the ->end of them:
	 */
	map_groups__fixup_end(&machine->kmaps);
1033 1034 1035 1036 1037 1038 1039

	if (maps__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps, name,
					     addr)) {
		machine__destroy_kernel_maps(machine);
		return -1;
	}

1040 1041 1042
	return 0;
}

1043 1044 1045
static void machine__set_kernel_mmap_len(struct machine *machine,
					 union perf_event *event)
{
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058
	int i;

	for (i = 0; i < MAP__NR_TYPES; i++) {
		machine->vmlinux_maps[i]->start = event->mmap.start;
		machine->vmlinux_maps[i]->end   = (event->mmap.start +
						   event->mmap.len);
		/*
		 * Be a bit paranoid here, some perf.data file came with
		 * a zero sized synthesized MMAP event for the kernel.
		 */
		if (machine->vmlinux_maps[i]->end == 0)
			machine->vmlinux_maps[i]->end = ~0ULL;
	}
1059 1060
}

1061 1062 1063 1064
static bool machine__uses_kcore(struct machine *machine)
{
	struct dso *dso;

1065
	list_for_each_entry(dso, &machine->kernel_dsos.head, node) {
1066 1067 1068 1069 1070 1071 1072
		if (dso__is_kcore(dso))
			return true;
	}

	return false;
}

1073 1074 1075 1076 1077 1078 1079 1080
static int machine__process_kernel_mmap_event(struct machine *machine,
					      union perf_event *event)
{
	struct map *map;
	char kmmap_prefix[PATH_MAX];
	enum dso_kernel_type kernel_type;
	bool is_kernel_mmap;

1081 1082 1083 1084
	/* If we have maps from kcore then we do not need or want any others */
	if (machine__uses_kcore(machine))
		return 0;

1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
	machine__mmap_name(machine, kmmap_prefix, sizeof(kmmap_prefix));
	if (machine__is_host(machine))
		kernel_type = DSO_TYPE_KERNEL;
	else
		kernel_type = DSO_TYPE_GUEST_KERNEL;

	is_kernel_mmap = memcmp(event->mmap.filename,
				kmmap_prefix,
				strlen(kmmap_prefix) - 1) == 0;
	if (event->mmap.filename[0] == '/' ||
	    (!is_kernel_mmap && event->mmap.filename[0] == '[')) {
		map = machine__new_module(machine, event->mmap.start,
					  event->mmap.filename);
		if (map == NULL)
			goto out_problem;

		map->end = map->start + event->mmap.len;
	} else if (is_kernel_mmap) {
		const char *symbol_name = (event->mmap.filename +
				strlen(kmmap_prefix));
		/*
		 * Should be there already, from the build-id table in
		 * the header.
		 */
1109 1110 1111 1112
		struct dso *kernel = NULL;
		struct dso *dso;

		list_for_each_entry(dso, &machine->kernel_dsos.head, node) {
1113
			if (is_kernel_module(dso->long_name))
1114 1115 1116 1117 1118 1119 1120 1121 1122
				continue;

			kernel = dso;
			break;
		}

		if (kernel == NULL)
			kernel = __dsos__findnew(&machine->kernel_dsos,
						 kmmap_prefix);
1123 1124 1125 1126 1127 1128 1129
		if (kernel == NULL)
			goto out_problem;

		kernel->kernel = kernel_type;
		if (__machine__create_kernel_maps(machine, kernel) < 0)
			goto out_problem;

1130 1131
		if (strstr(kernel->long_name, "vmlinux"))
			dso__set_short_name(kernel, "[kernel.vmlinux]", false);
1132

1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158
		machine__set_kernel_mmap_len(machine, event);

		/*
		 * Avoid using a zero address (kptr_restrict) for the ref reloc
		 * symbol. Effectively having zero here means that at record
		 * time /proc/sys/kernel/kptr_restrict was non zero.
		 */
		if (event->mmap.pgoff != 0) {
			maps__set_kallsyms_ref_reloc_sym(machine->vmlinux_maps,
							 symbol_name,
							 event->mmap.pgoff);
		}

		if (machine__is_default_guest(machine)) {
			/*
			 * preload dso of guest kernel and modules
			 */
			dso__load(kernel, machine->vmlinux_maps[MAP__FUNCTION],
				  NULL);
		}
	}
	return 0;
out_problem:
	return -1;
}

1159
int machine__process_mmap2_event(struct machine *machine,
1160 1161
				 union perf_event *event,
				 struct perf_sample *sample __maybe_unused)
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180
{
	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
	struct thread *thread;
	struct map *map;
	enum map_type type;
	int ret = 0;

	if (dump_trace)
		perf_event__fprintf_mmap2(event, stdout);

	if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL ||
	    cpumode == PERF_RECORD_MISC_KERNEL) {
		ret = machine__process_kernel_mmap_event(machine, event);
		if (ret < 0)
			goto out_problem;
		return 0;
	}

	thread = machine__findnew_thread(machine, event->mmap2.pid,
1181
					event->mmap2.tid);
1182 1183 1184 1185 1186 1187 1188 1189
	if (thread == NULL)
		goto out_problem;

	if (event->header.misc & PERF_RECORD_MISC_MMAP_DATA)
		type = MAP__VARIABLE;
	else
		type = MAP__FUNCTION;

1190
	map = map__new(machine, event->mmap2.start,
1191 1192 1193 1194
			event->mmap2.len, event->mmap2.pgoff,
			event->mmap2.pid, event->mmap2.maj,
			event->mmap2.min, event->mmap2.ino,
			event->mmap2.ino_generation,
1195 1196
			event->mmap2.prot,
			event->mmap2.flags,
1197
			event->mmap2.filename, type, thread);
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209

	if (map == NULL)
		goto out_problem;

	thread__insert_map(thread, map);
	return 0;

out_problem:
	dump_printf("problem processing PERF_RECORD_MMAP2, skipping event.\n");
	return 0;
}

1210 1211
int machine__process_mmap_event(struct machine *machine, union perf_event *event,
				struct perf_sample *sample __maybe_unused)
1212 1213 1214 1215
{
	u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;
	struct thread *thread;
	struct map *map;
1216
	enum map_type type;
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
	int ret = 0;

	if (dump_trace)
		perf_event__fprintf_mmap(event, stdout);

	if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL ||
	    cpumode == PERF_RECORD_MISC_KERNEL) {
		ret = machine__process_kernel_mmap_event(machine, event);
		if (ret < 0)
			goto out_problem;
		return 0;
	}

1230
	thread = machine__findnew_thread(machine, event->mmap.pid,
1231
					 event->mmap.tid);
1232 1233
	if (thread == NULL)
		goto out_problem;
1234 1235 1236 1237 1238 1239

	if (event->header.misc & PERF_RECORD_MISC_MMAP_DATA)
		type = MAP__VARIABLE;
	else
		type = MAP__FUNCTION;

1240
	map = map__new(machine, event->mmap.start,
1241
			event->mmap.len, event->mmap.pgoff,
1242
			event->mmap.pid, 0, 0, 0, 0, 0, 0,
1243
			event->mmap.filename,
1244
			type, thread);
1245

1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
	if (map == NULL)
		goto out_problem;

	thread__insert_map(thread, map);
	return 0;

out_problem:
	dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n");
	return 0;
}

1257
void machine__remove_thread(struct machine *machine, struct thread *th)
1258
{
1259 1260 1261
	if (machine->last_match == th)
		thread__zput(machine->last_match);

1262 1263
	rb_erase(&th->rb_node, &machine->threads);
	/*
1264 1265 1266
	 * Move it first to the dead_threads list, then drop the reference,
	 * if this is the last reference, then the thread__delete destructor
	 * will be called and we will remove it from the dead_threads list.
1267 1268
	 */
	list_add_tail(&th->node, &machine->dead_threads);
1269
	thread__put(th);
1270 1271
}

1272 1273
int machine__process_fork_event(struct machine *machine, union perf_event *event,
				struct perf_sample *sample)
1274
{
1275 1276 1277
	struct thread *thread = machine__find_thread(machine,
						     event->fork.pid,
						     event->fork.tid);
1278 1279 1280
	struct thread *parent = machine__findnew_thread(machine,
							event->fork.ppid,
							event->fork.ptid);
1281

1282 1283 1284 1285
	/* if a thread currently exists for the thread id remove it */
	if (thread != NULL)
		machine__remove_thread(machine, thread);

1286 1287
	thread = machine__findnew_thread(machine, event->fork.pid,
					 event->fork.tid);
1288 1289 1290 1291
	if (dump_trace)
		perf_event__fprintf_task(event, stdout);

	if (thread == NULL || parent == NULL ||
1292
	    thread__fork(thread, parent, sample->time) < 0) {
1293 1294 1295 1296 1297 1298 1299
		dump_printf("problem processing PERF_RECORD_FORK, skipping event.\n");
		return -1;
	}

	return 0;
}

1300 1301
int machine__process_exit_event(struct machine *machine, union perf_event *event,
				struct perf_sample *sample __maybe_unused)
1302
{
1303 1304 1305
	struct thread *thread = machine__find_thread(machine,
						     event->fork.pid,
						     event->fork.tid);
1306 1307 1308 1309 1310

	if (dump_trace)
		perf_event__fprintf_task(event, stdout);

	if (thread != NULL)
1311
		thread__exited(thread);
1312 1313 1314 1315

	return 0;
}

1316 1317
int machine__process_event(struct machine *machine, union perf_event *event,
			   struct perf_sample *sample)
1318 1319 1320 1321 1322
{
	int ret;

	switch (event->header.type) {
	case PERF_RECORD_COMM:
1323
		ret = machine__process_comm_event(machine, event, sample); break;
1324
	case PERF_RECORD_MMAP:
1325
		ret = machine__process_mmap_event(machine, event, sample); break;
1326
	case PERF_RECORD_MMAP2:
1327
		ret = machine__process_mmap2_event(machine, event, sample); break;
1328
	case PERF_RECORD_FORK:
1329
		ret = machine__process_fork_event(machine, event, sample); break;
1330
	case PERF_RECORD_EXIT:
1331
		ret = machine__process_exit_event(machine, event, sample); break;
1332
	case PERF_RECORD_LOST:
1333
		ret = machine__process_lost_event(machine, event, sample); break;
1334 1335 1336 1337 1338 1339 1340
	default:
		ret = -1;
		break;
	}

	return ret;
}
1341

1342
static bool symbol__match_regex(struct symbol *sym, regex_t *regex)
1343
{
1344
	if (sym->name && !regexec(regex, sym->name, 0, NULL, 0))
1345 1346 1347 1348
		return 1;
	return 0;
}

1349
static void ip__resolve_ams(struct thread *thread,
1350 1351 1352 1353 1354 1355
			    struct addr_map_symbol *ams,
			    u64 ip)
{
	struct addr_location al;

	memset(&al, 0, sizeof(al));
1356 1357 1358 1359 1360 1361 1362
	/*
	 * We cannot use the header.misc hint to determine whether a
	 * branch stack address is user, kernel, guest, hypervisor.
	 * Branches may straddle the kernel/user/hypervisor boundaries.
	 * Thus, we have to try consecutively until we find a match
	 * or else, the symbol is unknown
	 */
1363
	thread__find_cpumode_addr_location(thread, MAP__FUNCTION, ip, &al);
1364 1365 1366 1367 1368 1369 1370

	ams->addr = ip;
	ams->al_addr = al.addr;
	ams->sym = al.sym;
	ams->map = al.map;
}

1371
static void ip__resolve_data(struct thread *thread,
1372 1373 1374 1375 1376 1377
			     u8 m, struct addr_map_symbol *ams, u64 addr)
{
	struct addr_location al;

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

1378
	thread__find_addr_location(thread, m, MAP__VARIABLE, addr, &al);
1379 1380 1381 1382 1383 1384
	if (al.map == NULL) {
		/*
		 * some shared data regions have execute bit set which puts
		 * their mapping in the MAP__FUNCTION type array.
		 * Check there as a fallback option before dropping the sample.
		 */
1385
		thread__find_addr_location(thread, m, MAP__FUNCTION, addr, &al);
1386 1387
	}

1388 1389 1390 1391 1392 1393
	ams->addr = addr;
	ams->al_addr = al.addr;
	ams->sym = al.sym;
	ams->map = al.map;
}

1394 1395
struct mem_info *sample__resolve_mem(struct perf_sample *sample,
				     struct addr_location *al)
1396 1397 1398 1399 1400 1401
{
	struct mem_info *mi = zalloc(sizeof(*mi));

	if (!mi)
		return NULL;

1402 1403
	ip__resolve_ams(al->thread, &mi->iaddr, sample->ip);
	ip__resolve_data(al->thread, al->cpumode, &mi->daddr, sample->addr);
1404 1405 1406 1407 1408
	mi->data_src.val = sample->data_src;

	return mi;
}

1409 1410 1411
static int add_callchain_ip(struct thread *thread,
			    struct symbol **parent,
			    struct addr_location *root_al,
1412
			    u8 *cpumode,
1413 1414 1415 1416 1417 1418
			    u64 ip)
{
	struct addr_location al;

	al.filtered = 0;
	al.sym = NULL;
1419
	if (!cpumode) {
1420 1421
		thread__find_cpumode_addr_location(thread, MAP__FUNCTION,
						   ip, &al);
1422
	} else {
1423 1424 1425
		if (ip >= PERF_CONTEXT_MAX) {
			switch (ip) {
			case PERF_CONTEXT_HV:
1426
				*cpumode = PERF_RECORD_MISC_HYPERVISOR;
1427 1428
				break;
			case PERF_CONTEXT_KERNEL:
1429
				*cpumode = PERF_RECORD_MISC_KERNEL;
1430 1431
				break;
			case PERF_CONTEXT_USER:
1432
				*cpumode = PERF_RECORD_MISC_USER;
1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445
				break;
			default:
				pr_debug("invalid callchain context: "
					 "%"PRId64"\n", (s64) ip);
				/*
				 * It seems the callchain is corrupted.
				 * Discard all.
				 */
				callchain_cursor_reset(&callchain_cursor);
				return 1;
			}
			return 0;
		}
1446 1447
		thread__find_addr_location(thread, *cpumode, MAP__FUNCTION,
					   ip, &al);
1448 1449
	}

1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462
	if (al.sym != NULL) {
		if (sort__has_parent && !*parent &&
		    symbol__match_regex(al.sym, &parent_regex))
			*parent = al.sym;
		else if (have_ignore_callees && root_al &&
		  symbol__match_regex(al.sym, &ignore_callees_regex)) {
			/* Treat this symbol as the root,
			   forgetting its callees. */
			*root_al = al;
			callchain_cursor_reset(&callchain_cursor);
		}
	}

1463
	return callchain_cursor_append(&callchain_cursor, al.addr, al.map, al.sym);
1464 1465
}

1466 1467
struct branch_info *sample__resolve_bstack(struct perf_sample *sample,
					   struct addr_location *al)
1468 1469
{
	unsigned int i;
1470 1471
	const struct branch_stack *bs = sample->branch_stack;
	struct branch_info *bi = calloc(bs->nr, sizeof(struct branch_info));
1472 1473 1474 1475 1476

	if (!bi)
		return NULL;

	for (i = 0; i < bs->nr; i++) {
1477 1478
		ip__resolve_ams(al->thread, &bi[i].to, bs->entries[i].to);
		ip__resolve_ams(al->thread, &bi[i].from, bs->entries[i].from);
1479 1480 1481 1482 1483
		bi[i].flags = bs->entries[i].flags;
	}
	return bi;
}

1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524
#define CHASHSZ 127
#define CHASHBITS 7
#define NO_ENTRY 0xff

#define PERF_MAX_BRANCH_DEPTH 127

/* Remove loops. */
static int remove_loops(struct branch_entry *l, int nr)
{
	int i, j, off;
	unsigned char chash[CHASHSZ];

	memset(chash, NO_ENTRY, sizeof(chash));

	BUG_ON(PERF_MAX_BRANCH_DEPTH > 255);

	for (i = 0; i < nr; i++) {
		int h = hash_64(l[i].from, CHASHBITS) % CHASHSZ;

		/* no collision handling for now */
		if (chash[h] == NO_ENTRY) {
			chash[h] = i;
		} else if (l[chash[h]].from == l[i].from) {
			bool is_loop = true;
			/* check if it is a real loop */
			off = 0;
			for (j = chash[h]; j < i && i + off < nr; j++, off++)
				if (l[j].from != l[i + off].from) {
					is_loop = false;
					break;
				}
			if (is_loop) {
				memmove(l + i, l + i + off,
					(nr - (i + off)) * sizeof(*l));
				nr -= off;
			}
		}
	}
	return nr;
}

K
Kan Liang 已提交
1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536
/*
 * Recolve LBR callstack chain sample
 * Return:
 * 1 on success get LBR callchain information
 * 0 no available LBR callchain information, should try fp
 * negative error code on other errors.
 */
static int resolve_lbr_callchain_sample(struct thread *thread,
					struct perf_sample *sample,
					struct symbol **parent,
					struct addr_location *root_al,
					int max_stack)
1537
{
K
Kan Liang 已提交
1538 1539
	struct ip_callchain *chain = sample->callchain;
	int chain_nr = min(max_stack, (int)chain->nr);
1540
	u8 cpumode = PERF_RECORD_MISC_USER;
K
Kan Liang 已提交
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 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
	int i, j, err;
	u64 ip;

	for (i = 0; i < chain_nr; i++) {
		if (chain->ips[i] == PERF_CONTEXT_USER)
			break;
	}

	/* LBR only affects the user callchain */
	if (i != chain_nr) {
		struct branch_stack *lbr_stack = sample->branch_stack;
		int lbr_nr = lbr_stack->nr;
		/*
		 * LBR callstack can only get user call chain.
		 * The mix_chain_nr is kernel call chain
		 * number plus LBR user call chain number.
		 * i is kernel call chain number,
		 * 1 is PERF_CONTEXT_USER,
		 * lbr_nr + 1 is the user call chain number.
		 * For details, please refer to the comments
		 * in callchain__printf
		 */
		int mix_chain_nr = i + 1 + lbr_nr + 1;

		if (mix_chain_nr > PERF_MAX_STACK_DEPTH + PERF_MAX_BRANCH_DEPTH) {
			pr_warning("corrupted callchain. skipping...\n");
			return 0;
		}

		for (j = 0; j < mix_chain_nr; j++) {
			if (callchain_param.order == ORDER_CALLEE) {
				if (j < i + 1)
					ip = chain->ips[j];
				else if (j > i + 1)
					ip = lbr_stack->entries[j - i - 2].from;
				else
					ip = lbr_stack->entries[0].to;
			} else {
				if (j < lbr_nr)
					ip = lbr_stack->entries[lbr_nr - j - 1].from;
				else if (j > lbr_nr)
					ip = chain->ips[i + 1 - (j - lbr_nr)];
				else
					ip = lbr_stack->entries[0].to;
			}

1587
			err = add_callchain_ip(thread, parent, root_al, &cpumode, ip);
K
Kan Liang 已提交
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
			if (err)
				return (err < 0) ? err : 0;
		}
		return 1;
	}

	return 0;
}

static int thread__resolve_callchain_sample(struct thread *thread,
					    struct perf_evsel *evsel,
					    struct perf_sample *sample,
					    struct symbol **parent,
					    struct addr_location *root_al,
					    int max_stack)
{
	struct branch_stack *branch = sample->branch_stack;
	struct ip_callchain *chain = sample->callchain;
1606
	int chain_nr = min(max_stack, (int)chain->nr);
1607
	u8 cpumode = PERF_RECORD_MISC_USER;
1608
	int i, j, err;
1609 1610 1611
	int skip_idx = -1;
	int first_call = 0;

K
Kan Liang 已提交
1612 1613 1614 1615 1616 1617 1618 1619 1620
	callchain_cursor_reset(&callchain_cursor);

	if (has_branch_callstack(evsel)) {
		err = resolve_lbr_callchain_sample(thread, sample, parent,
						   root_al, max_stack);
		if (err)
			return (err < 0) ? err : 0;
	}

1621 1622 1623 1624 1625 1626
	/*
	 * Based on DWARF debug information, some architectures skip
	 * a callchain entry saved by the kernel.
	 */
	if (chain->nr < PERF_MAX_STACK_DEPTH)
		skip_idx = arch_skip_callchain_idx(thread, chain);
1627

1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672
	/*
	 * Add branches to call stack for easier browsing. This gives
	 * more context for a sample than just the callers.
	 *
	 * This uses individual histograms of paths compared to the
	 * aggregated histograms the normal LBR mode uses.
	 *
	 * Limitations for now:
	 * - No extra filters
	 * - No annotations (should annotate somehow)
	 */

	if (branch && callchain_param.branch_callstack) {
		int nr = min(max_stack, (int)branch->nr);
		struct branch_entry be[nr];

		if (branch->nr > PERF_MAX_BRANCH_DEPTH) {
			pr_warning("corrupted branch chain. skipping...\n");
			goto check_calls;
		}

		for (i = 0; i < nr; i++) {
			if (callchain_param.order == ORDER_CALLEE) {
				be[i] = branch->entries[i];
				/*
				 * Check for overlap into the callchain.
				 * The return address is one off compared to
				 * the branch entry. To adjust for this
				 * assume the calling instruction is not longer
				 * than 8 bytes.
				 */
				if (i == skip_idx ||
				    chain->ips[first_call] >= PERF_CONTEXT_MAX)
					first_call++;
				else if (be[i].from < chain->ips[first_call] &&
				    be[i].from >= chain->ips[first_call] - 8)
					first_call++;
			} else
				be[i] = branch->entries[branch->nr - i - 1];
		}

		nr = remove_loops(be, nr);

		for (i = 0; i < nr; i++) {
			err = add_callchain_ip(thread, parent, root_al,
1673
					       NULL, be[i].to);
1674 1675
			if (!err)
				err = add_callchain_ip(thread, parent, root_al,
1676
						       NULL, be[i].from);
1677 1678 1679 1680 1681 1682 1683 1684 1685
			if (err == -EINVAL)
				break;
			if (err)
				return err;
		}
		chain_nr -= nr;
	}

check_calls:
1686 1687 1688 1689 1690
	if (chain->nr > PERF_MAX_STACK_DEPTH) {
		pr_warning("corrupted callchain. skipping...\n");
		return 0;
	}

1691
	for (i = first_call; i < chain_nr; i++) {
1692 1693 1694
		u64 ip;

		if (callchain_param.order == ORDER_CALLEE)
1695
			j = i;
1696
		else
1697 1698 1699 1700 1701 1702 1703
			j = chain->nr - i - 1;

#ifdef HAVE_SKIP_CALLCHAIN_IDX
		if (j == skip_idx)
			continue;
#endif
		ip = chain->ips[j];
1704

1705
		err = add_callchain_ip(thread, parent, root_al, &cpumode, ip);
1706 1707

		if (err)
1708
			return (err < 0) ? err : 0;
1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720
	}

	return 0;
}

static int unwind_entry(struct unwind_entry *entry, void *arg)
{
	struct callchain_cursor *cursor = arg;
	return callchain_cursor_append(cursor, entry->ip,
				       entry->map, entry->sym);
}

1721 1722 1723 1724 1725 1726
int thread__resolve_callchain(struct thread *thread,
			      struct perf_evsel *evsel,
			      struct perf_sample *sample,
			      struct symbol **parent,
			      struct addr_location *root_al,
			      int max_stack)
1727
{
K
Kan Liang 已提交
1728 1729 1730
	int ret = thread__resolve_callchain_sample(thread, evsel,
						   sample, parent,
						   root_al, max_stack);
1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743
	if (ret)
		return ret;

	/* Can we do dwarf post unwind? */
	if (!((evsel->attr.sample_type & PERF_SAMPLE_REGS_USER) &&
	      (evsel->attr.sample_type & PERF_SAMPLE_STACK_USER)))
		return 0;

	/* Bail out if nothing was captured. */
	if ((!sample->user_regs.regs) ||
	    (!sample->user_stack.size))
		return 0;

1744
	return unwind__get_entries(unwind_entry, &callchain_cursor,
1745
				   thread, sample, max_stack);
1746 1747

}
1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770

int machine__for_each_thread(struct machine *machine,
			     int (*fn)(struct thread *thread, void *p),
			     void *priv)
{
	struct rb_node *nd;
	struct thread *thread;
	int rc = 0;

	for (nd = rb_first(&machine->threads); nd; nd = rb_next(nd)) {
		thread = rb_entry(nd, struct thread, rb_node);
		rc = fn(thread, priv);
		if (rc != 0)
			return rc;
	}

	list_for_each_entry(thread, &machine->dead_threads, node) {
		rc = fn(thread, priv);
		if (rc != 0)
			return rc;
	}
	return rc;
}
1771

1772
int __machine__synthesize_threads(struct machine *machine, struct perf_tool *tool,
1773
				  struct target *target, struct thread_map *threads,
1774
				  perf_event__handler_t process, bool data_mmap)
1775
{
1776
	if (target__has_task(target))
1777
		return perf_event__synthesize_thread_map(tool, threads, process, machine, data_mmap);
1778
	else if (target__has_cpu(target))
1779 1780 1781 1782
		return perf_event__synthesize_threads(tool, process, machine, data_mmap);
	/* command specified */
	return 0;
}
1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825

pid_t machine__get_current_tid(struct machine *machine, int cpu)
{
	if (cpu < 0 || cpu >= MAX_NR_CPUS || !machine->current_tid)
		return -1;

	return machine->current_tid[cpu];
}

int machine__set_current_tid(struct machine *machine, int cpu, pid_t pid,
			     pid_t tid)
{
	struct thread *thread;

	if (cpu < 0)
		return -EINVAL;

	if (!machine->current_tid) {
		int i;

		machine->current_tid = calloc(MAX_NR_CPUS, sizeof(pid_t));
		if (!machine->current_tid)
			return -ENOMEM;
		for (i = 0; i < MAX_NR_CPUS; i++)
			machine->current_tid[i] = -1;
	}

	if (cpu >= MAX_NR_CPUS) {
		pr_err("Requested CPU %d too large. ", cpu);
		pr_err("Consider raising MAX_NR_CPUS\n");
		return -EINVAL;
	}

	machine->current_tid[cpu] = tid;

	thread = machine__findnew_thread(machine, pid, tid);
	if (!thread)
		return -ENOMEM;

	thread->cpu = cpu;

	return 0;
}
1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847

int machine__get_kernel_start(struct machine *machine)
{
	struct map *map = machine__kernel_map(machine, MAP__FUNCTION);
	int err = 0;

	/*
	 * The only addresses above 2^63 are kernel addresses of a 64-bit
	 * kernel.  Note that addresses are unsigned so that on a 32-bit system
	 * all addresses including kernel addresses are less than 2^32.  In
	 * that case (32-bit system), if the kernel mapping is unknown, all
	 * addresses will be assumed to be in user space - see
	 * machine__kernel_ip().
	 */
	machine->kernel_start = 1ULL << 63;
	if (map) {
		err = map__load(map, machine->symbol_filter);
		if (map->start)
			machine->kernel_start = map->start;
	}
	return err;
}