aboutsummaryrefslogtreecommitdiffstats
path: root/tools/perf/builtin-top.c
blob: c0a423004e153b83249fb1d6631a5499e4375dc5 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
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
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
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
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
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
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
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
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
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
679
680
681
682
683
684
685
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
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
/*
 * builtin-top.c
 *
 * Builtin top command: Display a continuously updated profile of
 * any workload, CPU or specific PID.
 *
 * Copyright (C) 2008, Red Hat Inc, Ingo Molnar <mingo@redhat.com>
 *
 * Improvements and fixes by:
 *
 *   Arjan van de Ven <arjan@linux.intel.com>
 *   Yanmin Zhang <yanmin.zhang@intel.com>
 *   Wu Fengguang <fengguang.wu@intel.com>
 *   Mike Galbraith <efault@gmx.de>
 *   Paul Mackerras <paulus@samba.org>
 *
 * Released under the GPL v2. (and only v2, not any later version)
 */
#include "builtin.h"

#include "perf.h"

#include "util/symbol.h"
#include "util/color.h"
#include "util/util.h"
#include <linux/rbtree.h>
#include "util/parse-options.h"
#include "util/parse-events.h"

#include <assert.h>
#include <fcntl.h>

#include <stdio.h>

#include <errno.h>
#include <time.h>
#include <sched.h>
#include <pthread.h>

#include <sys/syscall.h>
#include <sys/ioctl.h>
#include <sys/poll.h>
#include <sys/prctl.h>
#include <sys/wait.h>
#include <sys/uio.h>
#include <sys/mman.h>

#include <linux/unistd.h>
#include <linux/types.h>

static int			fd[MAX_NR_CPUS][MAX_COUNTERS];

static int			system_wide			=  0;

static int			default_interval		= 100000;

static u64			count_filter			=  5;
static int			print_entries			= 15;

static int			target_pid			= -1;
static int			inherit				=  0;
static int			profile_cpu			= -1;
static int			nr_cpus				=  0;
static unsigned int		realtime_prio			=  0;
static int			group				=  0;
static unsigned int		page_size;
static unsigned int		mmap_pages			= 16;
static int			freq				=  0;
static int			verbose				=  0;
static char			*vmlinux			=  NULL;

static char			*sym_filter;
static unsigned long		filter_start;
static unsigned long		filter_end;

static int			delay_secs			=  2;
static int			zero;
static int			dump_symtab;

/*
 * Symbols
 */

static u64			min_ip;
static u64			max_ip = -1ll;

struct sym_entry {
	struct rb_node		rb_node;
	struct list_head	node;
	unsigned long		count[MAX_COUNTERS];
	unsigned long		snap_count;
	double			weight;
	int			skip;
};

struct sym_entry		*sym_filter_entry;

struct dso			*kernel_dso;

/*
 * Symbols will be added here in record_ip and will get out
 * after decayed.
 */
static LIST_HEAD(active_symbols);
static pthread_mutex_t active_symbols_lock = PTHREAD_MUTEX_INITIALIZER;

/*
 * Ordering weight: count-1 * count-2 * ... / count-n
 */
static double sym_weight(const struct sym_entry *sym)
{
	double weight = sym->snap_count;
	int counter;

	for (counter = 1; counter < nr_counters-1; counter++)
		weight *= sym->count[counter];

	weight /= (sym->count[counter] + 1);

	return weight;
}

static long			samples;
static long			userspace_samples;
static const char		CONSOLE_CLEAR[] = "";

static void __list_insert_active_sym(struct sym_entry *syme)
{
	list_add(&syme->node, &active_symbols);
}

static void list_remove_active_sym(struct sym_entry *syme)
{
	pthread_mutex_lock(&active_symbols_lock);
	list_del_init(&syme->node);
	pthread_mutex_unlock(&active_symbols_lock);
}

static void rb_insert_active_sym(struct rb_root *tree, struct sym_entry *se)
{
	struct rb_node **p = &tree->rb_node;
	struct rb_node *parent = NULL;
	struct sym_entry *iter;

	while (*p != NULL) {
		parent = *p;
		iter = rb_entry(parent, struct sym_entry, rb_node);

		if (se->weight > iter->weight)
			p = &(*p)->rb_left;
		else
			p = &(*p)->rb_right;
	}

	rb_link_node(&se->rb_node, parent, p);
	rb_insert_color(&se->rb_node, tree);
}

static void print_sym_table(void)
{
	int printed = 0, j;
	int counter;
	float samples_per_sec = samples/delay_secs;
	float ksamples_per_sec = (samples-userspace_samples)/delay_secs;
	float sum_ksamples = 0.0;
	struct sym_entry *syme, *n;
	struct rb_root tmp = RB_ROOT;
	struct rb_node *nd;

	samples = userspace_samples = 0;

	/* Sort the active symbols */
	pthread_mutex_lock(&active_symbols_lock);
	syme = list_entry(active_symbols.next, struct sym_entry, node);
	pthread_mutex_unlock(&active_symbols_lock);

	list_for_each_entry_safe_from(syme, n, &active_symbols, node) {
		syme->snap_count = syme->count[0];
		if (syme->snap_count != 0) {
			syme->weight = sym_weight(syme);
			rb_insert_active_sym(&tmp, syme);
			sum_ksamples += syme->snap_count;

			for (j = 0; j < nr_counters; j++)
				syme->count[j] = zero ? 0 : syme->count[j] * 7 / 8;
		} else
			list_remove_active_sym(syme);
	}

	puts(CONSOLE_CLEAR);

	printf(
"------------------------------------------------------------------------------\n");
	printf( "   PerfTop:%8.0f irqs/sec  kernel:%4.1f%% [",
		samples_per_sec,
		100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec)));

	if (nr_counters == 1) {
		printf("%Ld", (u64)attrs[0].sample_period);
		if (freq)
			printf("Hz ");
		else
			printf(" ");
	}

	for (counter = 0; counter < nr_counters; counter++) {
		if (counter)
			printf("/");

		printf("%s", event_name(counter));
	}

	printf( "], ");

	if (target_pid != -1)
		printf(" (target_pid: %d", target_pid);
	else
		printf(" (all");

	if (profile_cpu != -1)
		printf(", cpu: %d)\n", profile_cpu);
	else {
		if (target_pid != -1)
			printf(")\n");
		else
			printf(", %d CPUs)\n", nr_cpus);
	}

	printf("------------------------------------------------------------------------------\n\n");

	if (nr_counters == 1)
		printf("             samples    pcnt");
	else
		printf("  weight     samples    pcnt");

	printf("         RIP          kernel function\n"
	       	       "  ______     _______   _____   ________________   _______________\n\n"
	);

	for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) {
		struct sym_entry *syme = rb_entry(nd, struct sym_entry, rb_node);
		struct symbol *sym = (struct symbol *)(syme + 1);
		double pcnt;

		if (++printed > print_entries || syme->snap_count < count_filter)
			continue;

		pcnt = 100.0 - (100.0 * ((sum_ksamples - syme->snap_count) /
					 sum_ksamples));

		if (nr_counters == 1)
			printf("%20.2f - ", syme->weight);
		else
			printf("%9.1f %10ld - ", syme->weight, syme->snap_count);

		percent_color_fprintf(stdout, "%4.1f%%", pcnt);
		printf(" - %016llx : %s", sym->start, sym->name);
		if (sym->module)
			printf("\t[%s]", sym->module->name);
		printf("\n");
	}
}

static void *display_thread(void *arg __used)
{
	struct pollfd stdin_poll = { .fd = 0, .events = POLLIN };
	int delay_msecs = delay_secs * 1000;

	printf("PerfTop refresh period: %d seconds\n", delay_secs);

	do {
		print_sym_table();
	} while (!poll(&stdin_poll, 1, delay_msecs) == 1);

	printf("key pressed - exiting.\n");
	exit(0);

	return NULL;
}

/* Tag samples to be skipped. */
static const char *skip_symbols[] = {
	"default_idle",
	"cpu_idle",
	"enter_idle",
	"exit_idle",
	"mwait_idle",
	"ppc64_runlatch_off",
	"pseries_dedicated_idle_sleep",
	NULL
};

static int symbol_filter(struct dso *self, struct symbol *sym)
{
	static int filter_match;
	struct sym_entry *syme;
	const char *name = sym->name;
	int i;

	/*
	 * ppc64 uses function descriptors and appends a '.' to the
	 * start of every instruction address. Remove it.
	 */
	if (name[0] == '.')
		name++;

	if (!strcmp(name, "_text") ||
	    !strcmp(name, "_etext") ||
	    !strcmp(name, "_sinittext") ||
	    !strncmp("init_module", name, 11) ||
	    !strncmp("cleanup_module", name, 14) ||
	    strstr(name, "_text_start") ||
	    strstr(name, "_text_end"))
		return 1;

	syme = dso__sym_priv(self, sym);
	for (i = 0; skip_symbols[i]; i++) {
		if (!strcmp(skip_symbols[i], name)) {
			syme->skip = 1;
			break;
		}
	}

	if (filter_match == 1) {
		filter_end = sym->start;
		filter_match = -1;
		if (filter_end - filter_start > 10000) {
			fprintf(stderr,
				"hm, too large filter symbol <%s> - skipping.\n",
				sym_filter);
			fprintf(stderr, "symbol filter start: %016lx\n",
				filter_start);
			fprintf(stderr, "                end: %016lx\n",
				filter_end);
			filter_end = filter_start = 0;
			sym_filter = NULL;
			sleep(1);
		}
	}

	if (filter_match == 0 && sym_filter && !strcmp(name, sym_filter)) {
		filter_match = 1;
		filter_start = sym->start;
	}


	return 0;
}

static int parse_symbols(void)
{
	struct rb_node *node;
	struct symbol  *sym;
	int modules = vmlinux ? 1 : 0;

	kernel_dso = dso__new("[kernel]", sizeof(struct sym_entry));
	if (kernel_dso == NULL)
		return -1;

	if (dso__load_kernel(kernel_dso, vmlinux, symbol_filter, verbose, modules) <= 0)
		goto out_delete_dso;

	node = rb_first(&kernel_dso->syms);
	sym = rb_entry(node, struct symbol, rb_node);
	min_ip = sym->start;

	node = rb_last(&kernel_dso->syms);
	sym = rb_entry(node, struct symbol, rb_node);
	max_ip = sym->end;

	if (dump_symtab)
		dso__fprintf(kernel_dso, stderr);

	return 0;

out_delete_dso:
	dso__delete(kernel_dso);
	kernel_dso = NULL;
	return -1;
}

#define TRACE_COUNT     3

/*
 * Binary search in the histogram table and record the hit:
 */
static void record_ip(u64 ip, int counter)
{
	struct symbol *sym = dso__find_symbol(kernel_dso, ip);

	if (sym != NULL) {
		struct sym_entry *syme = dso__sym_priv(kernel_dso, sym);

		if (!syme->skip) {
			syme->count[counter]++;
			pthread_mutex_lock(&active_symbols_lock);
			if (list_empty(&syme->node) || !syme->node.next)
				__list_insert_active_sym(syme);
			pthread_mutex_unlock(&active_symbols_lock);
			return;
		}
	}

	samples--;
}

static void process_event(u64 ip, int counter, int user)
{
	samples++;

	if (user) {
		userspace_samples++;
		return;
	}

	record_ip(ip, counter);
}

struct mmap_data {
	int			counter;
	void			*base;
	int			mask;
	unsigned int		prev;
};

static unsigned int mmap_read_head(struct mmap_data *md)
{
	struct perf_counter_mmap_page *pc = md->base;
	int head;

	head = pc->data_head;
	rmb();

	return head;
}

struct timeval last_read, this_read;

static void mmap_read_counter(struct mmap_data *md)
{
	unsigned int head = mmap_read_head(md);
	unsigned int old = md->prev;
	unsigned char *data = md->base + page_size;
	int diff;

	gettimeofday(&this_read, NULL);

	/*
	 * If we're further behind than half the buffer, there's a chance
	 * the writer will bite our tail and mess up the samples under us.
	 *
	 * If we somehow ended up ahead of the head, we got messed up.
	 *
	 * In either case, truncate and restart at head.
	 */
	diff = head - old;
	if (diff > md->mask / 2 || diff < 0) {
		struct timeval iv;
		unsigned long msecs;

		timersub(&this_read, &last_read, &iv);
		msecs = iv.tv_sec*1000 + iv.tv_usec/1000;

		fprintf(stderr, "WARNING: failed to keep up with mmap data."
				"  Last read %lu msecs ago.\n", msecs);

		/*
		 * head points to a known good entry, start there.
		 */
		old = head;
	}

	last_read = this_read;

	for (; old != head;) {
		struct ip_event {
			struct perf_event_header header;
			u64 ip;
			u32 pid, target_pid;
		};
		struct mmap_event {
			struct perf_event_header header;
			u32 pid, target_pid;
			u64 start;
			u64 len;
			u64 pgoff;
			char filename[PATH_MAX];
		};

		typedef union event_union {
			struct perf_event_header header;
			struct ip_event ip;
			struct mmap_event mmap;
		} event_t;

		event_t *event = (event_t *)&data[old & md->mask];

		event_t event_copy;

		size_t size = event->header.size;

		/*
		 * Event straddles the mmap boundary -- header should always
		 * be inside due to u64 alignment of output.
		 */
		if ((old & md->mask) + size != ((old + size) & md->mask)) {
			unsigned int offset = old;
			unsigned int len = min(sizeof(*event), size), cpy;
			void *dst = &event_copy;

			do {
				cpy = min(md->mask + 1 - (offset & md->mask), len);
				memcpy(dst, &data[offset & md->mask], cpy);
				offset += cpy;
				dst += cpy;
				len -= cpy;
			} while (len);

			event = &event_copy;
		}

		old += size;

		if (event->header.type == PERF_EVENT_SAMPLE) {
			int user =
	(event->header.misc & PERF_EVENT_MISC_CPUMODE_MASK) == PERF_EVENT_MISC_USER;
			process_event(event->ip.ip, md->counter, user);
		}
	}

	md->prev = old;
}

static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS];
static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS];

static void mmap_read(void)
{
	int i, counter;

	for (i = 0; i < nr_cpus; i++) {
		for (counter = 0; counter < nr_counters; counter++)
			mmap_read_counter(&mmap_array[i][counter]);
	}
}

int nr_poll;
int group_fd;

static void start_counter(int i, int counter)
{
	struct perf_counter_attr *attr;
	int cpu;

	cpu = profile_cpu;
	if (target_pid == -1 && profile_cpu == -1)
		cpu = i;

	attr = attrs + counter;

	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
	attr->freq		= freq;
	attr->inherit		= (cpu < 0) && inherit;

try_again:
	fd[i][counter] = sys_perf_counter_open(attr, target_pid, cpu, group_fd, 0);

	if (fd[i][counter] < 0) {
		int err = errno;

		if (err == EPERM)
			die("No permission - are you root?\n");
		/*
		 * If it's cycles then fall back to hrtimer
		 * based cpu-clock-tick sw counter, which
		 * is always available even if no PMU support:
		 */
		if (attr->type == PERF_TYPE_HARDWARE
			&& attr->config == PERF_COUNT_HW_CPU_CYCLES) {

			if (verbose)
				warning(" ... trying to fall back to cpu-clock-ticks\n");

			attr->type = PERF_TYPE_SOFTWARE;
			attr->config = PERF_COUNT_SW_CPU_CLOCK;
			goto try_again;
		}
		printf("\n");
		error("perfcounter syscall returned with %d (%s)\n",
			fd[i][counter], strerror(err));
		die("No CONFIG_PERF_COUNTERS=y kernel support configured?\n");
		exit(-1);
	}
	assert(fd[i][counter] >= 0);
	fcntl(fd[i][counter], F_SETFL, O_NONBLOCK);

	/*
	 * First counter acts as the group leader:
	 */
	if (group && group_fd == -1)
		group_fd = fd[i][counter];

	event_array[nr_poll].fd = fd[i][counter];
	event_array[nr_poll].events = POLLIN;
	nr_poll++;

	mmap_array[i][counter].counter = counter;
	mmap_array[i][counter].prev = 0;
	mmap_array[i][counter].mask = mmap_pages*page_size - 1;
	mmap_array[i][counter].base = mmap(NULL, (mmap_pages+1)*page_size,
			PROT_READ, MAP_SHARED, fd[i][counter], 0);
	if (mmap_array[i][counter].base == MAP_FAILED)
		die("failed to mmap with %d (%s)\n", errno, strerror(errno));
}

static int __cmd_top(void)
{
	pthread_t thread;
	int i, counter;
	int ret;

	for (i = 0; i < nr_cpus; i++) {
		group_fd = -1;
		for (counter = 0; counter < nr_counters; counter++)
			start_counter(i, counter);
	}

	/* Wait for a minimal set of events before starting the snapshot */
	poll(event_array, nr_poll, 100);

	mmap_read();

	if (pthread_create(&thread, NULL, display_thread, NULL)) {
		printf("Could not create display thread.\n");
		exit(-1);
	}

	if (realtime_prio) {
		struct sched_param param;

		param.sched_priority = realtime_prio;
		if (sched_setscheduler(0, SCHED_FIFO, &param)) {
			printf("Could not set realtime priority.\n");
			exit(-1);
		}
	}

	while (1) {
		int hits = samples;

		mmap_read();

		if (hits == samples)
			ret = poll(event_array, nr_poll, 100);
	}

	return 0;
}

static const char * const top_usage[] = {
	"perf top [<options>]",
	NULL
};

static const struct option options[] = {
	OPT_CALLBACK('e', "event", NULL, "event",
		     "event selector. use 'perf list' to list available events",
		     parse_events),
	OPT_INTEGER('c', "count", &default_interval,
		    "event period to sample"),
	OPT_INTEGER('p', "pid", &target_pid,
		    "profile events on existing pid"),
	OPT_BOOLEAN('a', "all-cpus", &system_wide,
			    "system-wide collection from all CPUs"),
	OPT_INTEGER('C', "CPU", &profile_cpu,
		    "CPU to profile on"),
	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
	OPT_INTEGER('m', "mmap-pages", &mmap_pages,
		    "number of mmap data pages"),
	OPT_INTEGER('r', "realtime", &realtime_prio,
		    "collect data with this RT SCHED_FIFO priority"),
	OPT_INTEGER('d', "delay", &delay_secs,
		    "number of seconds to delay between refreshes"),
	OPT_BOOLEAN('D', "dump-symtab", &dump_symtab,
			    "dump the symbol table used for profiling"),
	OPT_INTEGER('f', "count-filter", &count_filter,
		    "only display functions with more events than this"),
	OPT_BOOLEAN('g', "group", &group,
			    "put the counters into a counter group"),
	OPT_BOOLEAN('i', "inherit", &inherit,
		    "child tasks inherit counters"),
	OPT_STRING('s', "sym-filter", &sym_filter, "pattern",
		    "only display symbols matchig this pattern"),
	OPT_BOOLEAN('z', "zero", &zero,
		    "zero history across updates"),
	OPT_INTEGER('F', "freq", &freq,
		    "profile at this frequency"),
	OPT_INTEGER('E', "entries", &print_entries,
		    "display this many functions"),
	OPT_BOOLEAN('v', "verbose", &verbose,
		    "be more verbose (show counter open errors, etc)"),
	OPT_END()
};

int cmd_top(int argc, const char **argv, const char *prefix __used)
{
	int counter;

	symbol__init();

	page_size = sysconf(_SC_PAGE_SIZE);

	argc = parse_options(argc, argv, options, top_usage, 0);
	if (argc)
		usage_with_options(top_usage, options);

	if (freq) {
		default_interval = freq;
		freq = 1;
	}

	/* CPU and PID are mutually exclusive */
	if (target_pid != -1 && profile_cpu != -1) {
		printf("WARNING: PID switch overriding CPU\n");
		sleep(1);
		profile_cpu = -1;
	}

	if (!nr_counters)
		nr_counters = 1;

	if (delay_secs < 1)
		delay_secs = 1;

	parse_symbols();

	/*
	 * Fill in the ones not specifically initialized via -c:
	 */
	for (counter = 0; counter < nr_counters; counter++) {
		if (attrs[counter].sample_period)
			continue;

		attrs[counter].sample_period = default_interval;
	}

	nr_cpus = sysconf(_SC_NPROCESSORS_ONLN);
	assert(nr_cpus <= MAX_NR_CPUS);
	assert(nr_cpus >= 0);

	if (target_pid != -1 || profile_cpu != -1)
		nr_cpus = 1;

	return __cmd_top();
}