5 #include "util/evlist.h"
6 #include "util/cache.h"
7 #include "util/evsel.h"
8 #include "util/symbol.h"
9 #include "util/thread.h"
10 #include "util/header.h"
11 #include "util/session.h"
12 #include "util/tool.h"
13 #include "util/cloexec.h"
14 #include "util/thread_map.h"
15 #include "util/color.h"
16 #include "util/stat.h"
17 #include "util/callchain.h"
18 #include "util/time-utils.h"
20 #include <subcmd/parse-options.h>
21 #include "util/trace-event.h"
23 #include "util/debug.h"
25 #include <linux/log2.h>
26 #include <sys/prctl.h>
27 #include <sys/resource.h>
29 #include <semaphore.h>
32 #include <api/fs/fs.h>
33 #include <linux/time64.h>
35 #define PR_SET_NAME 15 /* Set process name */
39 #define MAX_PID 1024000
48 unsigned long nr_events;
49 unsigned long curr_event;
50 struct sched_atom **atoms;
61 enum sched_event_type {
65 SCHED_EVENT_MIGRATION,
69 enum sched_event_type type;
75 struct task_desc *wakee;
78 #define TASK_STATE_TO_CHAR_STR "RSDTtZXxKWP"
80 /* task state bitmask, copied from include/linux/sched.h */
81 #define TASK_RUNNING 0
82 #define TASK_INTERRUPTIBLE 1
83 #define TASK_UNINTERRUPTIBLE 2
84 #define __TASK_STOPPED 4
85 #define __TASK_TRACED 8
86 /* in tsk->exit_state */
88 #define EXIT_ZOMBIE 32
89 #define EXIT_TRACE (EXIT_ZOMBIE | EXIT_DEAD)
90 /* in tsk->state again */
92 #define TASK_WAKEKILL 128
93 #define TASK_WAKING 256
94 #define TASK_PARKED 512
104 struct list_head list;
105 enum thread_state state;
113 struct list_head work_list;
114 struct thread *thread;
124 typedef int (*sort_fn_t)(struct work_atoms *, struct work_atoms *);
128 struct trace_sched_handler {
129 int (*switch_event)(struct perf_sched *sched, struct perf_evsel *evsel,
130 struct perf_sample *sample, struct machine *machine);
132 int (*runtime_event)(struct perf_sched *sched, struct perf_evsel *evsel,
133 struct perf_sample *sample, struct machine *machine);
135 int (*wakeup_event)(struct perf_sched *sched, struct perf_evsel *evsel,
136 struct perf_sample *sample, struct machine *machine);
138 /* PERF_RECORD_FORK event, not sched_process_fork tracepoint */
139 int (*fork_event)(struct perf_sched *sched, union perf_event *event,
140 struct machine *machine);
142 int (*migrate_task_event)(struct perf_sched *sched,
143 struct perf_evsel *evsel,
144 struct perf_sample *sample,
145 struct machine *machine);
148 #define COLOR_PIDS PERF_COLOR_BLUE
149 #define COLOR_CPUS PERF_COLOR_BG_RED
151 struct perf_sched_map {
152 DECLARE_BITMAP(comp_cpus_mask, MAX_CPUS);
155 struct thread_map *color_pids;
156 const char *color_pids_str;
157 struct cpu_map *color_cpus;
158 const char *color_cpus_str;
159 struct cpu_map *cpus;
160 const char *cpus_str;
164 struct perf_tool tool;
165 const char *sort_order;
166 unsigned long nr_tasks;
167 struct task_desc **pid_to_task;
168 struct task_desc **tasks;
169 const struct trace_sched_handler *tp_handler;
170 pthread_mutex_t start_work_mutex;
171 pthread_mutex_t work_done_wait_mutex;
174 * Track the current task - that way we can know whether there's any
175 * weird events, such as a task being switched away that is not current.
178 u32 curr_pid[MAX_CPUS];
179 struct thread *curr_thread[MAX_CPUS];
180 char next_shortname1;
181 char next_shortname2;
182 unsigned int replay_repeat;
183 unsigned long nr_run_events;
184 unsigned long nr_sleep_events;
185 unsigned long nr_wakeup_events;
186 unsigned long nr_sleep_corrections;
187 unsigned long nr_run_events_optimized;
188 unsigned long targetless_wakeups;
189 unsigned long multitarget_wakeups;
190 unsigned long nr_runs;
191 unsigned long nr_timestamps;
192 unsigned long nr_unordered_timestamps;
193 unsigned long nr_context_switch_bugs;
194 unsigned long nr_events;
195 unsigned long nr_lost_chunks;
196 unsigned long nr_lost_events;
197 u64 run_measurement_overhead;
198 u64 sleep_measurement_overhead;
201 u64 runavg_cpu_usage;
202 u64 parent_cpu_usage;
203 u64 runavg_parent_cpu_usage;
209 u64 cpu_last_switched[MAX_CPUS];
210 struct rb_root atom_root, sorted_atom_root, merged_atom_root;
211 struct list_head sort_list, cmp_pid;
214 struct perf_sched_map map;
216 /* options for timehist command */
221 unsigned int max_stack;
222 bool show_cpu_visual;
224 bool show_migrations;
227 const char *time_str;
228 struct perf_time_interval ptime;
229 struct perf_time_interval hist_time;
232 /* per thread run time data */
233 struct thread_runtime {
234 u64 last_time; /* time of previous sched in/out event */
235 u64 dt_run; /* run time */
236 u64 dt_sleep; /* time between CPU access by sleep (off cpu) */
237 u64 dt_iowait; /* time between CPU access by iowait (off cpu) */
238 u64 dt_preempt; /* time between CPU access by preempt (off cpu) */
239 u64 dt_delay; /* time between wakeup and sched-in */
240 u64 ready_to_run; /* time of wakeup */
242 struct stats run_stats;
244 u64 total_sleep_time;
245 u64 total_iowait_time;
246 u64 total_preempt_time;
247 u64 total_delay_time;
253 /* per event run time data */
254 struct evsel_runtime {
255 u64 *last_time; /* time this event was last seen per cpu */
256 u32 ncpu; /* highest cpu slot allocated */
259 /* per cpu idle time data */
260 struct idle_thread_runtime {
261 struct thread_runtime tr;
262 struct thread *last_thread;
263 struct rb_root sorted_root;
264 struct callchain_root callchain;
265 struct callchain_cursor cursor;
268 /* track idle times per cpu */
269 static struct thread **idle_threads;
270 static int idle_max_cpu;
271 static char idle_comm[] = "<idle>";
273 static u64 get_nsecs(void)
277 clock_gettime(CLOCK_MONOTONIC, &ts);
279 return ts.tv_sec * NSEC_PER_SEC + ts.tv_nsec;
282 static void burn_nsecs(struct perf_sched *sched, u64 nsecs)
284 u64 T0 = get_nsecs(), T1;
288 } while (T1 + sched->run_measurement_overhead < T0 + nsecs);
291 static void sleep_nsecs(u64 nsecs)
295 ts.tv_nsec = nsecs % 999999999;
296 ts.tv_sec = nsecs / 999999999;
298 nanosleep(&ts, NULL);
301 static void calibrate_run_measurement_overhead(struct perf_sched *sched)
303 u64 T0, T1, delta, min_delta = NSEC_PER_SEC;
306 for (i = 0; i < 10; i++) {
308 burn_nsecs(sched, 0);
311 min_delta = min(min_delta, delta);
313 sched->run_measurement_overhead = min_delta;
315 printf("run measurement overhead: %" PRIu64 " nsecs\n", min_delta);
318 static void calibrate_sleep_measurement_overhead(struct perf_sched *sched)
320 u64 T0, T1, delta, min_delta = NSEC_PER_SEC;
323 for (i = 0; i < 10; i++) {
328 min_delta = min(min_delta, delta);
331 sched->sleep_measurement_overhead = min_delta;
333 printf("sleep measurement overhead: %" PRIu64 " nsecs\n", min_delta);
336 static struct sched_atom *
337 get_new_event(struct task_desc *task, u64 timestamp)
339 struct sched_atom *event = zalloc(sizeof(*event));
340 unsigned long idx = task->nr_events;
343 event->timestamp = timestamp;
347 size = sizeof(struct sched_atom *) * task->nr_events;
348 task->atoms = realloc(task->atoms, size);
349 BUG_ON(!task->atoms);
351 task->atoms[idx] = event;
356 static struct sched_atom *last_event(struct task_desc *task)
358 if (!task->nr_events)
361 return task->atoms[task->nr_events - 1];
364 static void add_sched_event_run(struct perf_sched *sched, struct task_desc *task,
365 u64 timestamp, u64 duration)
367 struct sched_atom *event, *curr_event = last_event(task);
370 * optimize an existing RUN event by merging this one
373 if (curr_event && curr_event->type == SCHED_EVENT_RUN) {
374 sched->nr_run_events_optimized++;
375 curr_event->duration += duration;
379 event = get_new_event(task, timestamp);
381 event->type = SCHED_EVENT_RUN;
382 event->duration = duration;
384 sched->nr_run_events++;
387 static void add_sched_event_wakeup(struct perf_sched *sched, struct task_desc *task,
388 u64 timestamp, struct task_desc *wakee)
390 struct sched_atom *event, *wakee_event;
392 event = get_new_event(task, timestamp);
393 event->type = SCHED_EVENT_WAKEUP;
394 event->wakee = wakee;
396 wakee_event = last_event(wakee);
397 if (!wakee_event || wakee_event->type != SCHED_EVENT_SLEEP) {
398 sched->targetless_wakeups++;
401 if (wakee_event->wait_sem) {
402 sched->multitarget_wakeups++;
406 wakee_event->wait_sem = zalloc(sizeof(*wakee_event->wait_sem));
407 sem_init(wakee_event->wait_sem, 0, 0);
408 wakee_event->specific_wait = 1;
409 event->wait_sem = wakee_event->wait_sem;
411 sched->nr_wakeup_events++;
414 static void add_sched_event_sleep(struct perf_sched *sched, struct task_desc *task,
415 u64 timestamp, u64 task_state __maybe_unused)
417 struct sched_atom *event = get_new_event(task, timestamp);
419 event->type = SCHED_EVENT_SLEEP;
421 sched->nr_sleep_events++;
424 static struct task_desc *register_pid(struct perf_sched *sched,
425 unsigned long pid, const char *comm)
427 struct task_desc *task;
430 if (sched->pid_to_task == NULL) {
431 if (sysctl__read_int("kernel/pid_max", &pid_max) < 0)
433 BUG_ON((sched->pid_to_task = calloc(pid_max, sizeof(struct task_desc *))) == NULL);
435 if (pid >= (unsigned long)pid_max) {
436 BUG_ON((sched->pid_to_task = realloc(sched->pid_to_task, (pid + 1) *
437 sizeof(struct task_desc *))) == NULL);
438 while (pid >= (unsigned long)pid_max)
439 sched->pid_to_task[pid_max++] = NULL;
442 task = sched->pid_to_task[pid];
447 task = zalloc(sizeof(*task));
449 task->nr = sched->nr_tasks;
450 strcpy(task->comm, comm);
452 * every task starts in sleeping state - this gets ignored
453 * if there's no wakeup pointing to this sleep state:
455 add_sched_event_sleep(sched, task, 0, 0);
457 sched->pid_to_task[pid] = task;
459 sched->tasks = realloc(sched->tasks, sched->nr_tasks * sizeof(struct task_desc *));
460 BUG_ON(!sched->tasks);
461 sched->tasks[task->nr] = task;
464 printf("registered task #%ld, PID %ld (%s)\n", sched->nr_tasks, pid, comm);
470 static void print_task_traces(struct perf_sched *sched)
472 struct task_desc *task;
475 for (i = 0; i < sched->nr_tasks; i++) {
476 task = sched->tasks[i];
477 printf("task %6ld (%20s:%10ld), nr_events: %ld\n",
478 task->nr, task->comm, task->pid, task->nr_events);
482 static void add_cross_task_wakeups(struct perf_sched *sched)
484 struct task_desc *task1, *task2;
487 for (i = 0; i < sched->nr_tasks; i++) {
488 task1 = sched->tasks[i];
490 if (j == sched->nr_tasks)
492 task2 = sched->tasks[j];
493 add_sched_event_wakeup(sched, task1, 0, task2);
497 static void perf_sched__process_event(struct perf_sched *sched,
498 struct sched_atom *atom)
502 switch (atom->type) {
503 case SCHED_EVENT_RUN:
504 burn_nsecs(sched, atom->duration);
506 case SCHED_EVENT_SLEEP:
508 ret = sem_wait(atom->wait_sem);
511 case SCHED_EVENT_WAKEUP:
513 ret = sem_post(atom->wait_sem);
516 case SCHED_EVENT_MIGRATION:
523 static u64 get_cpu_usage_nsec_parent(void)
529 err = getrusage(RUSAGE_SELF, &ru);
532 sum = ru.ru_utime.tv_sec * NSEC_PER_SEC + ru.ru_utime.tv_usec * NSEC_PER_USEC;
533 sum += ru.ru_stime.tv_sec * NSEC_PER_SEC + ru.ru_stime.tv_usec * NSEC_PER_USEC;
538 static int self_open_counters(struct perf_sched *sched, unsigned long cur_task)
540 struct perf_event_attr attr;
541 char sbuf[STRERR_BUFSIZE], info[STRERR_BUFSIZE];
544 bool need_privilege = false;
546 memset(&attr, 0, sizeof(attr));
548 attr.type = PERF_TYPE_SOFTWARE;
549 attr.config = PERF_COUNT_SW_TASK_CLOCK;
552 fd = sys_perf_event_open(&attr, 0, -1, -1,
553 perf_event_open_cloexec_flag());
556 if (errno == EMFILE) {
558 BUG_ON(getrlimit(RLIMIT_NOFILE, &limit) == -1);
559 limit.rlim_cur += sched->nr_tasks - cur_task;
560 if (limit.rlim_cur > limit.rlim_max) {
561 limit.rlim_max = limit.rlim_cur;
562 need_privilege = true;
564 if (setrlimit(RLIMIT_NOFILE, &limit) == -1) {
565 if (need_privilege && errno == EPERM)
566 strcpy(info, "Need privilege\n");
570 strcpy(info, "Have a try with -f option\n");
572 pr_err("Error: sys_perf_event_open() syscall returned "
573 "with %d (%s)\n%s", fd,
574 str_error_r(errno, sbuf, sizeof(sbuf)), info);
580 static u64 get_cpu_usage_nsec_self(int fd)
585 ret = read(fd, &runtime, sizeof(runtime));
586 BUG_ON(ret != sizeof(runtime));
591 struct sched_thread_parms {
592 struct task_desc *task;
593 struct perf_sched *sched;
597 static void *thread_func(void *ctx)
599 struct sched_thread_parms *parms = ctx;
600 struct task_desc *this_task = parms->task;
601 struct perf_sched *sched = parms->sched;
602 u64 cpu_usage_0, cpu_usage_1;
603 unsigned long i, ret;
609 sprintf(comm2, ":%s", this_task->comm);
610 prctl(PR_SET_NAME, comm2);
614 ret = sem_post(&this_task->ready_for_work);
616 ret = pthread_mutex_lock(&sched->start_work_mutex);
618 ret = pthread_mutex_unlock(&sched->start_work_mutex);
621 cpu_usage_0 = get_cpu_usage_nsec_self(fd);
623 for (i = 0; i < this_task->nr_events; i++) {
624 this_task->curr_event = i;
625 perf_sched__process_event(sched, this_task->atoms[i]);
628 cpu_usage_1 = get_cpu_usage_nsec_self(fd);
629 this_task->cpu_usage = cpu_usage_1 - cpu_usage_0;
630 ret = sem_post(&this_task->work_done_sem);
633 ret = pthread_mutex_lock(&sched->work_done_wait_mutex);
635 ret = pthread_mutex_unlock(&sched->work_done_wait_mutex);
641 static void create_tasks(struct perf_sched *sched)
643 struct task_desc *task;
648 err = pthread_attr_init(&attr);
650 err = pthread_attr_setstacksize(&attr,
651 (size_t) max(16 * 1024, PTHREAD_STACK_MIN));
653 err = pthread_mutex_lock(&sched->start_work_mutex);
655 err = pthread_mutex_lock(&sched->work_done_wait_mutex);
657 for (i = 0; i < sched->nr_tasks; i++) {
658 struct sched_thread_parms *parms = malloc(sizeof(*parms));
659 BUG_ON(parms == NULL);
660 parms->task = task = sched->tasks[i];
661 parms->sched = sched;
662 parms->fd = self_open_counters(sched, i);
663 sem_init(&task->sleep_sem, 0, 0);
664 sem_init(&task->ready_for_work, 0, 0);
665 sem_init(&task->work_done_sem, 0, 0);
666 task->curr_event = 0;
667 err = pthread_create(&task->thread, &attr, thread_func, parms);
672 static void wait_for_tasks(struct perf_sched *sched)
674 u64 cpu_usage_0, cpu_usage_1;
675 struct task_desc *task;
676 unsigned long i, ret;
678 sched->start_time = get_nsecs();
679 sched->cpu_usage = 0;
680 pthread_mutex_unlock(&sched->work_done_wait_mutex);
682 for (i = 0; i < sched->nr_tasks; i++) {
683 task = sched->tasks[i];
684 ret = sem_wait(&task->ready_for_work);
686 sem_init(&task->ready_for_work, 0, 0);
688 ret = pthread_mutex_lock(&sched->work_done_wait_mutex);
691 cpu_usage_0 = get_cpu_usage_nsec_parent();
693 pthread_mutex_unlock(&sched->start_work_mutex);
695 for (i = 0; i < sched->nr_tasks; i++) {
696 task = sched->tasks[i];
697 ret = sem_wait(&task->work_done_sem);
699 sem_init(&task->work_done_sem, 0, 0);
700 sched->cpu_usage += task->cpu_usage;
704 cpu_usage_1 = get_cpu_usage_nsec_parent();
705 if (!sched->runavg_cpu_usage)
706 sched->runavg_cpu_usage = sched->cpu_usage;
707 sched->runavg_cpu_usage = (sched->runavg_cpu_usage * (sched->replay_repeat - 1) + sched->cpu_usage) / sched->replay_repeat;
709 sched->parent_cpu_usage = cpu_usage_1 - cpu_usage_0;
710 if (!sched->runavg_parent_cpu_usage)
711 sched->runavg_parent_cpu_usage = sched->parent_cpu_usage;
712 sched->runavg_parent_cpu_usage = (sched->runavg_parent_cpu_usage * (sched->replay_repeat - 1) +
713 sched->parent_cpu_usage)/sched->replay_repeat;
715 ret = pthread_mutex_lock(&sched->start_work_mutex);
718 for (i = 0; i < sched->nr_tasks; i++) {
719 task = sched->tasks[i];
720 sem_init(&task->sleep_sem, 0, 0);
721 task->curr_event = 0;
725 static void run_one_test(struct perf_sched *sched)
727 u64 T0, T1, delta, avg_delta, fluct;
730 wait_for_tasks(sched);
734 sched->sum_runtime += delta;
737 avg_delta = sched->sum_runtime / sched->nr_runs;
738 if (delta < avg_delta)
739 fluct = avg_delta - delta;
741 fluct = delta - avg_delta;
742 sched->sum_fluct += fluct;
744 sched->run_avg = delta;
745 sched->run_avg = (sched->run_avg * (sched->replay_repeat - 1) + delta) / sched->replay_repeat;
747 printf("#%-3ld: %0.3f, ", sched->nr_runs, (double)delta / NSEC_PER_MSEC);
749 printf("ravg: %0.2f, ", (double)sched->run_avg / NSEC_PER_MSEC);
751 printf("cpu: %0.2f / %0.2f",
752 (double)sched->cpu_usage / NSEC_PER_MSEC, (double)sched->runavg_cpu_usage / NSEC_PER_MSEC);
756 * rusage statistics done by the parent, these are less
757 * accurate than the sched->sum_exec_runtime based statistics:
759 printf(" [%0.2f / %0.2f]",
760 (double)sched->parent_cpu_usage / NSEC_PER_MSEC,
761 (double)sched->runavg_parent_cpu_usage / NSEC_PER_MSEC);
766 if (sched->nr_sleep_corrections)
767 printf(" (%ld sleep corrections)\n", sched->nr_sleep_corrections);
768 sched->nr_sleep_corrections = 0;
771 static void test_calibrations(struct perf_sched *sched)
776 burn_nsecs(sched, NSEC_PER_MSEC);
779 printf("the run test took %" PRIu64 " nsecs\n", T1 - T0);
782 sleep_nsecs(NSEC_PER_MSEC);
785 printf("the sleep test took %" PRIu64 " nsecs\n", T1 - T0);
789 replay_wakeup_event(struct perf_sched *sched,
790 struct perf_evsel *evsel, struct perf_sample *sample,
791 struct machine *machine __maybe_unused)
793 const char *comm = perf_evsel__strval(evsel, sample, "comm");
794 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
795 struct task_desc *waker, *wakee;
798 printf("sched_wakeup event %p\n", evsel);
800 printf(" ... pid %d woke up %s/%d\n", sample->tid, comm, pid);
803 waker = register_pid(sched, sample->tid, "<unknown>");
804 wakee = register_pid(sched, pid, comm);
806 add_sched_event_wakeup(sched, waker, sample->time, wakee);
810 static int replay_switch_event(struct perf_sched *sched,
811 struct perf_evsel *evsel,
812 struct perf_sample *sample,
813 struct machine *machine __maybe_unused)
815 const char *prev_comm = perf_evsel__strval(evsel, sample, "prev_comm"),
816 *next_comm = perf_evsel__strval(evsel, sample, "next_comm");
817 const u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
818 next_pid = perf_evsel__intval(evsel, sample, "next_pid");
819 const u64 prev_state = perf_evsel__intval(evsel, sample, "prev_state");
820 struct task_desc *prev, __maybe_unused *next;
821 u64 timestamp0, timestamp = sample->time;
822 int cpu = sample->cpu;
826 printf("sched_switch event %p\n", evsel);
828 if (cpu >= MAX_CPUS || cpu < 0)
831 timestamp0 = sched->cpu_last_switched[cpu];
833 delta = timestamp - timestamp0;
838 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
842 pr_debug(" ... switch from %s/%d to %s/%d [ran %" PRIu64 " nsecs]\n",
843 prev_comm, prev_pid, next_comm, next_pid, delta);
845 prev = register_pid(sched, prev_pid, prev_comm);
846 next = register_pid(sched, next_pid, next_comm);
848 sched->cpu_last_switched[cpu] = timestamp;
850 add_sched_event_run(sched, prev, timestamp, delta);
851 add_sched_event_sleep(sched, prev, timestamp, prev_state);
856 static int replay_fork_event(struct perf_sched *sched,
857 union perf_event *event,
858 struct machine *machine)
860 struct thread *child, *parent;
862 child = machine__findnew_thread(machine, event->fork.pid,
864 parent = machine__findnew_thread(machine, event->fork.ppid,
867 if (child == NULL || parent == NULL) {
868 pr_debug("thread does not exist on fork event: child %p, parent %p\n",
874 printf("fork event\n");
875 printf("... parent: %s/%d\n", thread__comm_str(parent), parent->tid);
876 printf("... child: %s/%d\n", thread__comm_str(child), child->tid);
879 register_pid(sched, parent->tid, thread__comm_str(parent));
880 register_pid(sched, child->tid, thread__comm_str(child));
887 struct sort_dimension {
890 struct list_head list;
894 thread_lat_cmp(struct list_head *list, struct work_atoms *l, struct work_atoms *r)
896 struct sort_dimension *sort;
899 BUG_ON(list_empty(list));
901 list_for_each_entry(sort, list, list) {
902 ret = sort->cmp(l, r);
910 static struct work_atoms *
911 thread_atoms_search(struct rb_root *root, struct thread *thread,
912 struct list_head *sort_list)
914 struct rb_node *node = root->rb_node;
915 struct work_atoms key = { .thread = thread };
918 struct work_atoms *atoms;
921 atoms = container_of(node, struct work_atoms, node);
923 cmp = thread_lat_cmp(sort_list, &key, atoms);
925 node = node->rb_left;
927 node = node->rb_right;
929 BUG_ON(thread != atoms->thread);
937 __thread_latency_insert(struct rb_root *root, struct work_atoms *data,
938 struct list_head *sort_list)
940 struct rb_node **new = &(root->rb_node), *parent = NULL;
943 struct work_atoms *this;
946 this = container_of(*new, struct work_atoms, node);
949 cmp = thread_lat_cmp(sort_list, data, this);
952 new = &((*new)->rb_left);
954 new = &((*new)->rb_right);
957 rb_link_node(&data->node, parent, new);
958 rb_insert_color(&data->node, root);
961 static int thread_atoms_insert(struct perf_sched *sched, struct thread *thread)
963 struct work_atoms *atoms = zalloc(sizeof(*atoms));
965 pr_err("No memory at %s\n", __func__);
969 atoms->thread = thread__get(thread);
970 INIT_LIST_HEAD(&atoms->work_list);
971 __thread_latency_insert(&sched->atom_root, atoms, &sched->cmp_pid);
975 static char sched_out_state(u64 prev_state)
977 const char *str = TASK_STATE_TO_CHAR_STR;
979 return str[prev_state];
983 add_sched_out_event(struct work_atoms *atoms,
987 struct work_atom *atom = zalloc(sizeof(*atom));
989 pr_err("Non memory at %s", __func__);
993 atom->sched_out_time = timestamp;
995 if (run_state == 'R') {
996 atom->state = THREAD_WAIT_CPU;
997 atom->wake_up_time = atom->sched_out_time;
1000 list_add_tail(&atom->list, &atoms->work_list);
1005 add_runtime_event(struct work_atoms *atoms, u64 delta,
1006 u64 timestamp __maybe_unused)
1008 struct work_atom *atom;
1010 BUG_ON(list_empty(&atoms->work_list));
1012 atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1014 atom->runtime += delta;
1015 atoms->total_runtime += delta;
1019 add_sched_in_event(struct work_atoms *atoms, u64 timestamp)
1021 struct work_atom *atom;
1024 if (list_empty(&atoms->work_list))
1027 atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1029 if (atom->state != THREAD_WAIT_CPU)
1032 if (timestamp < atom->wake_up_time) {
1033 atom->state = THREAD_IGNORE;
1037 atom->state = THREAD_SCHED_IN;
1038 atom->sched_in_time = timestamp;
1040 delta = atom->sched_in_time - atom->wake_up_time;
1041 atoms->total_lat += delta;
1042 if (delta > atoms->max_lat) {
1043 atoms->max_lat = delta;
1044 atoms->max_lat_at = timestamp;
1049 static int latency_switch_event(struct perf_sched *sched,
1050 struct perf_evsel *evsel,
1051 struct perf_sample *sample,
1052 struct machine *machine)
1054 const u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
1055 next_pid = perf_evsel__intval(evsel, sample, "next_pid");
1056 const u64 prev_state = perf_evsel__intval(evsel, sample, "prev_state");
1057 struct work_atoms *out_events, *in_events;
1058 struct thread *sched_out, *sched_in;
1059 u64 timestamp0, timestamp = sample->time;
1060 int cpu = sample->cpu, err = -1;
1063 BUG_ON(cpu >= MAX_CPUS || cpu < 0);
1065 timestamp0 = sched->cpu_last_switched[cpu];
1066 sched->cpu_last_switched[cpu] = timestamp;
1068 delta = timestamp - timestamp0;
1073 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
1077 sched_out = machine__findnew_thread(machine, -1, prev_pid);
1078 sched_in = machine__findnew_thread(machine, -1, next_pid);
1079 if (sched_out == NULL || sched_in == NULL)
1082 out_events = thread_atoms_search(&sched->atom_root, sched_out, &sched->cmp_pid);
1084 if (thread_atoms_insert(sched, sched_out))
1086 out_events = thread_atoms_search(&sched->atom_root, sched_out, &sched->cmp_pid);
1088 pr_err("out-event: Internal tree error");
1092 if (add_sched_out_event(out_events, sched_out_state(prev_state), timestamp))
1095 in_events = thread_atoms_search(&sched->atom_root, sched_in, &sched->cmp_pid);
1097 if (thread_atoms_insert(sched, sched_in))
1099 in_events = thread_atoms_search(&sched->atom_root, sched_in, &sched->cmp_pid);
1101 pr_err("in-event: Internal tree error");
1105 * Take came in we have not heard about yet,
1106 * add in an initial atom in runnable state:
1108 if (add_sched_out_event(in_events, 'R', timestamp))
1111 add_sched_in_event(in_events, timestamp);
1114 thread__put(sched_out);
1115 thread__put(sched_in);
1119 static int latency_runtime_event(struct perf_sched *sched,
1120 struct perf_evsel *evsel,
1121 struct perf_sample *sample,
1122 struct machine *machine)
1124 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
1125 const u64 runtime = perf_evsel__intval(evsel, sample, "runtime");
1126 struct thread *thread = machine__findnew_thread(machine, -1, pid);
1127 struct work_atoms *atoms = thread_atoms_search(&sched->atom_root, thread, &sched->cmp_pid);
1128 u64 timestamp = sample->time;
1129 int cpu = sample->cpu, err = -1;
1134 BUG_ON(cpu >= MAX_CPUS || cpu < 0);
1136 if (thread_atoms_insert(sched, thread))
1138 atoms = thread_atoms_search(&sched->atom_root, thread, &sched->cmp_pid);
1140 pr_err("in-event: Internal tree error");
1143 if (add_sched_out_event(atoms, 'R', timestamp))
1147 add_runtime_event(atoms, runtime, timestamp);
1150 thread__put(thread);
1154 static int latency_wakeup_event(struct perf_sched *sched,
1155 struct perf_evsel *evsel,
1156 struct perf_sample *sample,
1157 struct machine *machine)
1159 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
1160 struct work_atoms *atoms;
1161 struct work_atom *atom;
1162 struct thread *wakee;
1163 u64 timestamp = sample->time;
1166 wakee = machine__findnew_thread(machine, -1, pid);
1169 atoms = thread_atoms_search(&sched->atom_root, wakee, &sched->cmp_pid);
1171 if (thread_atoms_insert(sched, wakee))
1173 atoms = thread_atoms_search(&sched->atom_root, wakee, &sched->cmp_pid);
1175 pr_err("wakeup-event: Internal tree error");
1178 if (add_sched_out_event(atoms, 'S', timestamp))
1182 BUG_ON(list_empty(&atoms->work_list));
1184 atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1187 * As we do not guarantee the wakeup event happens when
1188 * task is out of run queue, also may happen when task is
1189 * on run queue and wakeup only change ->state to TASK_RUNNING,
1190 * then we should not set the ->wake_up_time when wake up a
1191 * task which is on run queue.
1193 * You WILL be missing events if you've recorded only
1194 * one CPU, or are only looking at only one, so don't
1195 * skip in this case.
1197 if (sched->profile_cpu == -1 && atom->state != THREAD_SLEEPING)
1200 sched->nr_timestamps++;
1201 if (atom->sched_out_time > timestamp) {
1202 sched->nr_unordered_timestamps++;
1206 atom->state = THREAD_WAIT_CPU;
1207 atom->wake_up_time = timestamp;
1215 static int latency_migrate_task_event(struct perf_sched *sched,
1216 struct perf_evsel *evsel,
1217 struct perf_sample *sample,
1218 struct machine *machine)
1220 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
1221 u64 timestamp = sample->time;
1222 struct work_atoms *atoms;
1223 struct work_atom *atom;
1224 struct thread *migrant;
1228 * Only need to worry about migration when profiling one CPU.
1230 if (sched->profile_cpu == -1)
1233 migrant = machine__findnew_thread(machine, -1, pid);
1234 if (migrant == NULL)
1236 atoms = thread_atoms_search(&sched->atom_root, migrant, &sched->cmp_pid);
1238 if (thread_atoms_insert(sched, migrant))
1240 register_pid(sched, migrant->tid, thread__comm_str(migrant));
1241 atoms = thread_atoms_search(&sched->atom_root, migrant, &sched->cmp_pid);
1243 pr_err("migration-event: Internal tree error");
1246 if (add_sched_out_event(atoms, 'R', timestamp))
1250 BUG_ON(list_empty(&atoms->work_list));
1252 atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1253 atom->sched_in_time = atom->sched_out_time = atom->wake_up_time = timestamp;
1255 sched->nr_timestamps++;
1257 if (atom->sched_out_time > timestamp)
1258 sched->nr_unordered_timestamps++;
1261 thread__put(migrant);
1265 static void output_lat_thread(struct perf_sched *sched, struct work_atoms *work_list)
1270 char max_lat_at[32];
1272 if (!work_list->nb_atoms)
1275 * Ignore idle threads:
1277 if (!strcmp(thread__comm_str(work_list->thread), "swapper"))
1280 sched->all_runtime += work_list->total_runtime;
1281 sched->all_count += work_list->nb_atoms;
1283 if (work_list->num_merged > 1)
1284 ret = printf(" %s:(%d) ", thread__comm_str(work_list->thread), work_list->num_merged);
1286 ret = printf(" %s:%d ", thread__comm_str(work_list->thread), work_list->thread->tid);
1288 for (i = 0; i < 24 - ret; i++)
1291 avg = work_list->total_lat / work_list->nb_atoms;
1292 timestamp__scnprintf_usec(work_list->max_lat_at, max_lat_at, sizeof(max_lat_at));
1294 printf("|%11.3f ms |%9" PRIu64 " | avg:%9.3f ms | max:%9.3f ms | max at: %13s s\n",
1295 (double)work_list->total_runtime / NSEC_PER_MSEC,
1296 work_list->nb_atoms, (double)avg / NSEC_PER_MSEC,
1297 (double)work_list->max_lat / NSEC_PER_MSEC,
1301 static int pid_cmp(struct work_atoms *l, struct work_atoms *r)
1303 if (l->thread == r->thread)
1305 if (l->thread->tid < r->thread->tid)
1307 if (l->thread->tid > r->thread->tid)
1309 return (int)(l->thread - r->thread);
1312 static int avg_cmp(struct work_atoms *l, struct work_atoms *r)
1322 avgl = l->total_lat / l->nb_atoms;
1323 avgr = r->total_lat / r->nb_atoms;
1333 static int max_cmp(struct work_atoms *l, struct work_atoms *r)
1335 if (l->max_lat < r->max_lat)
1337 if (l->max_lat > r->max_lat)
1343 static int switch_cmp(struct work_atoms *l, struct work_atoms *r)
1345 if (l->nb_atoms < r->nb_atoms)
1347 if (l->nb_atoms > r->nb_atoms)
1353 static int runtime_cmp(struct work_atoms *l, struct work_atoms *r)
1355 if (l->total_runtime < r->total_runtime)
1357 if (l->total_runtime > r->total_runtime)
1363 static int sort_dimension__add(const char *tok, struct list_head *list)
1366 static struct sort_dimension avg_sort_dimension = {
1370 static struct sort_dimension max_sort_dimension = {
1374 static struct sort_dimension pid_sort_dimension = {
1378 static struct sort_dimension runtime_sort_dimension = {
1382 static struct sort_dimension switch_sort_dimension = {
1386 struct sort_dimension *available_sorts[] = {
1387 &pid_sort_dimension,
1388 &avg_sort_dimension,
1389 &max_sort_dimension,
1390 &switch_sort_dimension,
1391 &runtime_sort_dimension,
1394 for (i = 0; i < ARRAY_SIZE(available_sorts); i++) {
1395 if (!strcmp(available_sorts[i]->name, tok)) {
1396 list_add_tail(&available_sorts[i]->list, list);
1405 static void perf_sched__sort_lat(struct perf_sched *sched)
1407 struct rb_node *node;
1408 struct rb_root *root = &sched->atom_root;
1411 struct work_atoms *data;
1412 node = rb_first(root);
1416 rb_erase(node, root);
1417 data = rb_entry(node, struct work_atoms, node);
1418 __thread_latency_insert(&sched->sorted_atom_root, data, &sched->sort_list);
1420 if (root == &sched->atom_root) {
1421 root = &sched->merged_atom_root;
1426 static int process_sched_wakeup_event(struct perf_tool *tool,
1427 struct perf_evsel *evsel,
1428 struct perf_sample *sample,
1429 struct machine *machine)
1431 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1433 if (sched->tp_handler->wakeup_event)
1434 return sched->tp_handler->wakeup_event(sched, evsel, sample, machine);
1444 static bool thread__has_color(struct thread *thread)
1446 union map_priv priv = {
1447 .ptr = thread__priv(thread),
1453 static struct thread*
1454 map__findnew_thread(struct perf_sched *sched, struct machine *machine, pid_t pid, pid_t tid)
1456 struct thread *thread = machine__findnew_thread(machine, pid, tid);
1457 union map_priv priv = {
1461 if (!sched->map.color_pids || !thread || thread__priv(thread))
1464 if (thread_map__has(sched->map.color_pids, tid))
1467 thread__set_priv(thread, priv.ptr);
1471 static int map_switch_event(struct perf_sched *sched, struct perf_evsel *evsel,
1472 struct perf_sample *sample, struct machine *machine)
1474 const u32 next_pid = perf_evsel__intval(evsel, sample, "next_pid");
1475 struct thread *sched_in;
1477 u64 timestamp0, timestamp = sample->time;
1479 int i, this_cpu = sample->cpu;
1481 bool new_cpu = false;
1482 const char *color = PERF_COLOR_NORMAL;
1483 char stimestamp[32];
1485 BUG_ON(this_cpu >= MAX_CPUS || this_cpu < 0);
1487 if (this_cpu > sched->max_cpu)
1488 sched->max_cpu = this_cpu;
1490 if (sched->map.comp) {
1491 cpus_nr = bitmap_weight(sched->map.comp_cpus_mask, MAX_CPUS);
1492 if (!test_and_set_bit(this_cpu, sched->map.comp_cpus_mask)) {
1493 sched->map.comp_cpus[cpus_nr++] = this_cpu;
1497 cpus_nr = sched->max_cpu;
1499 timestamp0 = sched->cpu_last_switched[this_cpu];
1500 sched->cpu_last_switched[this_cpu] = timestamp;
1502 delta = timestamp - timestamp0;
1507 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
1511 sched_in = map__findnew_thread(sched, machine, -1, next_pid);
1512 if (sched_in == NULL)
1515 sched->curr_thread[this_cpu] = thread__get(sched_in);
1520 if (!sched_in->shortname[0]) {
1521 if (!strcmp(thread__comm_str(sched_in), "swapper")) {
1523 * Don't allocate a letter-number for swapper:0
1524 * as a shortname. Instead, we use '.' for it.
1526 sched_in->shortname[0] = '.';
1527 sched_in->shortname[1] = ' ';
1529 sched_in->shortname[0] = sched->next_shortname1;
1530 sched_in->shortname[1] = sched->next_shortname2;
1532 if (sched->next_shortname1 < 'Z') {
1533 sched->next_shortname1++;
1535 sched->next_shortname1 = 'A';
1536 if (sched->next_shortname2 < '9')
1537 sched->next_shortname2++;
1539 sched->next_shortname2 = '0';
1545 for (i = 0; i < cpus_nr; i++) {
1546 int cpu = sched->map.comp ? sched->map.comp_cpus[i] : i;
1547 struct thread *curr_thread = sched->curr_thread[cpu];
1548 const char *pid_color = color;
1549 const char *cpu_color = color;
1551 if (curr_thread && thread__has_color(curr_thread))
1552 pid_color = COLOR_PIDS;
1554 if (sched->map.cpus && !cpu_map__has(sched->map.cpus, cpu))
1557 if (sched->map.color_cpus && cpu_map__has(sched->map.color_cpus, cpu))
1558 cpu_color = COLOR_CPUS;
1560 if (cpu != this_cpu)
1561 color_fprintf(stdout, color, " ");
1563 color_fprintf(stdout, cpu_color, "*");
1565 if (sched->curr_thread[cpu])
1566 color_fprintf(stdout, pid_color, "%2s ", sched->curr_thread[cpu]->shortname);
1568 color_fprintf(stdout, color, " ");
1571 if (sched->map.cpus && !cpu_map__has(sched->map.cpus, this_cpu))
1574 timestamp__scnprintf_usec(timestamp, stimestamp, sizeof(stimestamp));
1575 color_fprintf(stdout, color, " %12s secs ", stimestamp);
1576 if (new_shortname || (verbose && sched_in->tid)) {
1577 const char *pid_color = color;
1579 if (thread__has_color(sched_in))
1580 pid_color = COLOR_PIDS;
1582 color_fprintf(stdout, pid_color, "%s => %s:%d",
1583 sched_in->shortname, thread__comm_str(sched_in), sched_in->tid);
1586 if (sched->map.comp && new_cpu)
1587 color_fprintf(stdout, color, " (CPU %d)", this_cpu);
1590 color_fprintf(stdout, color, "\n");
1592 thread__put(sched_in);
1597 static int process_sched_switch_event(struct perf_tool *tool,
1598 struct perf_evsel *evsel,
1599 struct perf_sample *sample,
1600 struct machine *machine)
1602 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1603 int this_cpu = sample->cpu, err = 0;
1604 u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
1605 next_pid = perf_evsel__intval(evsel, sample, "next_pid");
1607 if (sched->curr_pid[this_cpu] != (u32)-1) {
1609 * Are we trying to switch away a PID that is
1612 if (sched->curr_pid[this_cpu] != prev_pid)
1613 sched->nr_context_switch_bugs++;
1616 if (sched->tp_handler->switch_event)
1617 err = sched->tp_handler->switch_event(sched, evsel, sample, machine);
1619 sched->curr_pid[this_cpu] = next_pid;
1623 static int process_sched_runtime_event(struct perf_tool *tool,
1624 struct perf_evsel *evsel,
1625 struct perf_sample *sample,
1626 struct machine *machine)
1628 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1630 if (sched->tp_handler->runtime_event)
1631 return sched->tp_handler->runtime_event(sched, evsel, sample, machine);
1636 static int perf_sched__process_fork_event(struct perf_tool *tool,
1637 union perf_event *event,
1638 struct perf_sample *sample,
1639 struct machine *machine)
1641 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1643 /* run the fork event through the perf machineruy */
1644 perf_event__process_fork(tool, event, sample, machine);
1646 /* and then run additional processing needed for this command */
1647 if (sched->tp_handler->fork_event)
1648 return sched->tp_handler->fork_event(sched, event, machine);
1653 static int process_sched_migrate_task_event(struct perf_tool *tool,
1654 struct perf_evsel *evsel,
1655 struct perf_sample *sample,
1656 struct machine *machine)
1658 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1660 if (sched->tp_handler->migrate_task_event)
1661 return sched->tp_handler->migrate_task_event(sched, evsel, sample, machine);
1666 typedef int (*tracepoint_handler)(struct perf_tool *tool,
1667 struct perf_evsel *evsel,
1668 struct perf_sample *sample,
1669 struct machine *machine);
1671 static int perf_sched__process_tracepoint_sample(struct perf_tool *tool __maybe_unused,
1672 union perf_event *event __maybe_unused,
1673 struct perf_sample *sample,
1674 struct perf_evsel *evsel,
1675 struct machine *machine)
1679 if (evsel->handler != NULL) {
1680 tracepoint_handler f = evsel->handler;
1681 err = f(tool, evsel, sample, machine);
1687 static int perf_sched__read_events(struct perf_sched *sched)
1689 const struct perf_evsel_str_handler handlers[] = {
1690 { "sched:sched_switch", process_sched_switch_event, },
1691 { "sched:sched_stat_runtime", process_sched_runtime_event, },
1692 { "sched:sched_wakeup", process_sched_wakeup_event, },
1693 { "sched:sched_wakeup_new", process_sched_wakeup_event, },
1694 { "sched:sched_migrate_task", process_sched_migrate_task_event, },
1696 struct perf_session *session;
1697 struct perf_data_file file = {
1699 .mode = PERF_DATA_MODE_READ,
1700 .force = sched->force,
1704 session = perf_session__new(&file, false, &sched->tool);
1705 if (session == NULL) {
1706 pr_debug("No Memory for session\n");
1710 symbol__init(&session->header.env);
1712 if (perf_session__set_tracepoints_handlers(session, handlers))
1715 if (perf_session__has_traces(session, "record -R")) {
1716 int err = perf_session__process_events(session);
1718 pr_err("Failed to process events, error %d", err);
1722 sched->nr_events = session->evlist->stats.nr_events[0];
1723 sched->nr_lost_events = session->evlist->stats.total_lost;
1724 sched->nr_lost_chunks = session->evlist->stats.nr_events[PERF_RECORD_LOST];
1729 perf_session__delete(session);
1734 * scheduling times are printed as msec.usec
1736 static inline void print_sched_time(unsigned long long nsecs, int width)
1738 unsigned long msecs;
1739 unsigned long usecs;
1741 msecs = nsecs / NSEC_PER_MSEC;
1742 nsecs -= msecs * NSEC_PER_MSEC;
1743 usecs = nsecs / NSEC_PER_USEC;
1744 printf("%*lu.%03lu ", width, msecs, usecs);
1748 * returns runtime data for event, allocating memory for it the
1749 * first time it is used.
1751 static struct evsel_runtime *perf_evsel__get_runtime(struct perf_evsel *evsel)
1753 struct evsel_runtime *r = evsel->priv;
1756 r = zalloc(sizeof(struct evsel_runtime));
1764 * save last time event was seen per cpu
1766 static void perf_evsel__save_time(struct perf_evsel *evsel,
1767 u64 timestamp, u32 cpu)
1769 struct evsel_runtime *r = perf_evsel__get_runtime(evsel);
1774 if ((cpu >= r->ncpu) || (r->last_time == NULL)) {
1775 int i, n = __roundup_pow_of_two(cpu+1);
1776 void *p = r->last_time;
1778 p = realloc(r->last_time, n * sizeof(u64));
1783 for (i = r->ncpu; i < n; ++i)
1784 r->last_time[i] = (u64) 0;
1789 r->last_time[cpu] = timestamp;
1792 /* returns last time this event was seen on the given cpu */
1793 static u64 perf_evsel__get_time(struct perf_evsel *evsel, u32 cpu)
1795 struct evsel_runtime *r = perf_evsel__get_runtime(evsel);
1797 if ((r == NULL) || (r->last_time == NULL) || (cpu >= r->ncpu))
1800 return r->last_time[cpu];
1803 static int comm_width = 30;
1805 static char *timehist_get_commstr(struct thread *thread)
1807 static char str[32];
1808 const char *comm = thread__comm_str(thread);
1809 pid_t tid = thread->tid;
1810 pid_t pid = thread->pid_;
1814 n = scnprintf(str, sizeof(str), "%s", comm);
1816 else if (tid != pid)
1817 n = scnprintf(str, sizeof(str), "%s[%d/%d]", comm, tid, pid);
1820 n = scnprintf(str, sizeof(str), "%s[%d]", comm, tid);
1828 static void timehist_header(struct perf_sched *sched)
1830 u32 ncpus = sched->max_cpu + 1;
1833 printf("%15s %6s ", "time", "cpu");
1835 if (sched->show_cpu_visual) {
1837 for (i = 0, j = 0; i < ncpus; ++i) {
1845 printf(" %-*s %9s %9s %9s", comm_width,
1846 "task name", "wait time", "sch delay", "run time");
1848 if (sched->show_state)
1849 printf(" %s", "state");
1856 printf("%15s %-6s ", "", "");
1858 if (sched->show_cpu_visual)
1859 printf(" %*s ", ncpus, "");
1861 printf(" %-*s %9s %9s %9s", comm_width,
1862 "[tid/pid]", "(msec)", "(msec)", "(msec)");
1864 if (sched->show_state)
1872 printf("%.15s %.6s ", graph_dotted_line, graph_dotted_line);
1874 if (sched->show_cpu_visual)
1875 printf(" %.*s ", ncpus, graph_dotted_line);
1877 printf(" %.*s %.9s %.9s %.9s", comm_width,
1878 graph_dotted_line, graph_dotted_line, graph_dotted_line,
1881 if (sched->show_state)
1882 printf(" %.5s", graph_dotted_line);
1887 static char task_state_char(struct thread *thread, int state)
1889 static const char state_to_char[] = TASK_STATE_TO_CHAR_STR;
1890 unsigned bit = state ? ffs(state) : 0;
1893 if (thread->tid == 0)
1896 return bit < sizeof(state_to_char) - 1 ? state_to_char[bit] : '?';
1899 static void timehist_print_sample(struct perf_sched *sched,
1900 struct perf_sample *sample,
1901 struct addr_location *al,
1902 struct thread *thread,
1905 struct thread_runtime *tr = thread__priv(thread);
1906 u32 max_cpus = sched->max_cpu + 1;
1910 timestamp__scnprintf_usec(t, tstr, sizeof(tstr));
1911 printf("%15s [%04d] ", tstr, sample->cpu);
1913 if (sched->show_cpu_visual) {
1918 for (i = 0; i < max_cpus; ++i) {
1919 /* flag idle times with 'i'; others are sched events */
1920 if (i == sample->cpu)
1921 c = (thread->tid == 0) ? 'i' : 's';
1929 printf(" %-*s ", comm_width, timehist_get_commstr(thread));
1931 wait_time = tr->dt_sleep + tr->dt_iowait + tr->dt_preempt;
1932 print_sched_time(wait_time, 6);
1934 print_sched_time(tr->dt_delay, 6);
1935 print_sched_time(tr->dt_run, 6);
1937 if (sched->show_state)
1938 printf(" %5c ", task_state_char(thread, state));
1940 if (sched->show_wakeups)
1941 printf(" %-*s", comm_width, "");
1943 if (thread->tid == 0)
1946 if (sched->show_callchain)
1949 sample__fprintf_sym(sample, al, 0,
1950 EVSEL__PRINT_SYM | EVSEL__PRINT_ONELINE |
1951 EVSEL__PRINT_CALLCHAIN_ARROW |
1952 EVSEL__PRINT_SKIP_IGNORED,
1953 &callchain_cursor, stdout);
1960 * Explanation of delta-time stats:
1962 * t = time of current schedule out event
1963 * tprev = time of previous sched out event
1964 * also time of schedule-in event for current task
1965 * last_time = time of last sched change event for current task
1966 * (i.e, time process was last scheduled out)
1967 * ready_to_run = time of wakeup for current task
1969 * -----|------------|------------|------------|------
1970 * last ready tprev t
1973 * |-------- dt_wait --------|
1974 * |- dt_delay -|-- dt_run --|
1976 * dt_run = run time of current task
1977 * dt_wait = time between last schedule out event for task and tprev
1978 * represents time spent off the cpu
1979 * dt_delay = time between wakeup and schedule-in of task
1982 static void timehist_update_runtime_stats(struct thread_runtime *r,
1992 r->dt_run = t - tprev;
1993 if (r->ready_to_run) {
1994 if (r->ready_to_run > tprev)
1995 pr_debug("time travel: wakeup time for task > previous sched_switch event\n");
1997 r->dt_delay = tprev - r->ready_to_run;
2000 if (r->last_time > tprev)
2001 pr_debug("time travel: last sched out time for task > previous sched_switch event\n");
2002 else if (r->last_time) {
2003 u64 dt_wait = tprev - r->last_time;
2005 if (r->last_state == TASK_RUNNING)
2006 r->dt_preempt = dt_wait;
2007 else if (r->last_state == TASK_UNINTERRUPTIBLE)
2008 r->dt_iowait = dt_wait;
2010 r->dt_sleep = dt_wait;
2014 update_stats(&r->run_stats, r->dt_run);
2016 r->total_run_time += r->dt_run;
2017 r->total_delay_time += r->dt_delay;
2018 r->total_sleep_time += r->dt_sleep;
2019 r->total_iowait_time += r->dt_iowait;
2020 r->total_preempt_time += r->dt_preempt;
2023 static bool is_idle_sample(struct perf_sample *sample,
2024 struct perf_evsel *evsel)
2026 /* pid 0 == swapper == idle task */
2027 if (strcmp(perf_evsel__name(evsel), "sched:sched_switch") == 0)
2028 return perf_evsel__intval(evsel, sample, "prev_pid") == 0;
2030 return sample->pid == 0;
2033 static void save_task_callchain(struct perf_sched *sched,
2034 struct perf_sample *sample,
2035 struct perf_evsel *evsel,
2036 struct machine *machine)
2038 struct callchain_cursor *cursor = &callchain_cursor;
2039 struct thread *thread;
2041 /* want main thread for process - has maps */
2042 thread = machine__findnew_thread(machine, sample->pid, sample->pid);
2043 if (thread == NULL) {
2044 pr_debug("Failed to get thread for pid %d.\n", sample->pid);
2048 if (!symbol_conf.use_callchain || sample->callchain == NULL)
2051 if (thread__resolve_callchain(thread, cursor, evsel, sample,
2052 NULL, NULL, sched->max_stack + 2) != 0) {
2054 error("Failed to resolve callchain. Skipping\n");
2059 callchain_cursor_commit(cursor);
2062 struct callchain_cursor_node *node;
2065 node = callchain_cursor_current(cursor);
2070 if (sym && sym->name) {
2071 if (!strcmp(sym->name, "schedule") ||
2072 !strcmp(sym->name, "__schedule") ||
2073 !strcmp(sym->name, "preempt_schedule"))
2077 callchain_cursor_advance(cursor);
2081 static int init_idle_thread(struct thread *thread)
2083 struct idle_thread_runtime *itr;
2085 thread__set_comm(thread, idle_comm, 0);
2087 itr = zalloc(sizeof(*itr));
2091 init_stats(&itr->tr.run_stats);
2092 callchain_init(&itr->callchain);
2093 callchain_cursor_reset(&itr->cursor);
2094 thread__set_priv(thread, itr);
2100 * Track idle stats per cpu by maintaining a local thread
2101 * struct for the idle task on each cpu.
2103 static int init_idle_threads(int ncpu)
2107 idle_threads = zalloc(ncpu * sizeof(struct thread *));
2111 idle_max_cpu = ncpu;
2113 /* allocate the actual thread struct if needed */
2114 for (i = 0; i < ncpu; ++i) {
2115 idle_threads[i] = thread__new(0, 0);
2116 if (idle_threads[i] == NULL)
2119 ret = init_idle_thread(idle_threads[i]);
2127 static void free_idle_threads(void)
2131 if (idle_threads == NULL)
2134 for (i = 0; i < idle_max_cpu; ++i) {
2135 if ((idle_threads[i]))
2136 thread__delete(idle_threads[i]);
2142 static struct thread *get_idle_thread(int cpu)
2145 * expand/allocate array of pointers to local thread
2148 if ((cpu >= idle_max_cpu) || (idle_threads == NULL)) {
2149 int i, j = __roundup_pow_of_two(cpu+1);
2152 p = realloc(idle_threads, j * sizeof(struct thread *));
2156 idle_threads = (struct thread **) p;
2157 for (i = idle_max_cpu; i < j; ++i)
2158 idle_threads[i] = NULL;
2163 /* allocate a new thread struct if needed */
2164 if (idle_threads[cpu] == NULL) {
2165 idle_threads[cpu] = thread__new(0, 0);
2167 if (idle_threads[cpu]) {
2168 if (init_idle_thread(idle_threads[cpu]) < 0)
2173 return idle_threads[cpu];
2176 static void save_idle_callchain(struct idle_thread_runtime *itr,
2177 struct perf_sample *sample)
2179 if (!symbol_conf.use_callchain || sample->callchain == NULL)
2182 callchain_cursor__copy(&itr->cursor, &callchain_cursor);
2186 * handle runtime stats saved per thread
2188 static struct thread_runtime *thread__init_runtime(struct thread *thread)
2190 struct thread_runtime *r;
2192 r = zalloc(sizeof(struct thread_runtime));
2196 init_stats(&r->run_stats);
2197 thread__set_priv(thread, r);
2202 static struct thread_runtime *thread__get_runtime(struct thread *thread)
2204 struct thread_runtime *tr;
2206 tr = thread__priv(thread);
2208 tr = thread__init_runtime(thread);
2210 pr_debug("Failed to malloc memory for runtime data.\n");
2216 static struct thread *timehist_get_thread(struct perf_sched *sched,
2217 struct perf_sample *sample,
2218 struct machine *machine,
2219 struct perf_evsel *evsel)
2221 struct thread *thread;
2223 if (is_idle_sample(sample, evsel)) {
2224 thread = get_idle_thread(sample->cpu);
2226 pr_err("Failed to get idle thread for cpu %d.\n", sample->cpu);
2229 /* there were samples with tid 0 but non-zero pid */
2230 thread = machine__findnew_thread(machine, sample->pid,
2231 sample->tid ?: sample->pid);
2232 if (thread == NULL) {
2233 pr_debug("Failed to get thread for tid %d. skipping sample.\n",
2237 save_task_callchain(sched, sample, evsel, machine);
2238 if (sched->idle_hist) {
2239 struct thread *idle;
2240 struct idle_thread_runtime *itr;
2242 idle = get_idle_thread(sample->cpu);
2244 pr_err("Failed to get idle thread for cpu %d.\n", sample->cpu);
2248 itr = thread__priv(idle);
2252 itr->last_thread = thread;
2254 /* copy task callchain when entering to idle */
2255 if (perf_evsel__intval(evsel, sample, "next_pid") == 0)
2256 save_idle_callchain(itr, sample);
2263 static bool timehist_skip_sample(struct perf_sched *sched,
2264 struct thread *thread,
2265 struct perf_evsel *evsel,
2266 struct perf_sample *sample)
2270 if (thread__is_filtered(thread)) {
2272 sched->skipped_samples++;
2275 if (sched->idle_hist) {
2276 if (strcmp(perf_evsel__name(evsel), "sched:sched_switch"))
2278 else if (perf_evsel__intval(evsel, sample, "prev_pid") != 0 &&
2279 perf_evsel__intval(evsel, sample, "next_pid") != 0)
2286 static void timehist_print_wakeup_event(struct perf_sched *sched,
2287 struct perf_evsel *evsel,
2288 struct perf_sample *sample,
2289 struct machine *machine,
2290 struct thread *awakened)
2292 struct thread *thread;
2295 thread = machine__findnew_thread(machine, sample->pid, sample->tid);
2299 /* show wakeup unless both awakee and awaker are filtered */
2300 if (timehist_skip_sample(sched, thread, evsel, sample) &&
2301 timehist_skip_sample(sched, awakened, evsel, sample)) {
2305 timestamp__scnprintf_usec(sample->time, tstr, sizeof(tstr));
2306 printf("%15s [%04d] ", tstr, sample->cpu);
2307 if (sched->show_cpu_visual)
2308 printf(" %*s ", sched->max_cpu + 1, "");
2310 printf(" %-*s ", comm_width, timehist_get_commstr(thread));
2313 printf(" %9s %9s %9s ", "", "", "");
2315 printf("awakened: %s", timehist_get_commstr(awakened));
2320 static int timehist_sched_wakeup_event(struct perf_tool *tool,
2321 union perf_event *event __maybe_unused,
2322 struct perf_evsel *evsel,
2323 struct perf_sample *sample,
2324 struct machine *machine)
2326 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
2327 struct thread *thread;
2328 struct thread_runtime *tr = NULL;
2329 /* want pid of awakened task not pid in sample */
2330 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
2332 thread = machine__findnew_thread(machine, 0, pid);
2336 tr = thread__get_runtime(thread);
2340 if (tr->ready_to_run == 0)
2341 tr->ready_to_run = sample->time;
2343 /* show wakeups if requested */
2344 if (sched->show_wakeups &&
2345 !perf_time__skip_sample(&sched->ptime, sample->time))
2346 timehist_print_wakeup_event(sched, evsel, sample, machine, thread);
2351 static void timehist_print_migration_event(struct perf_sched *sched,
2352 struct perf_evsel *evsel,
2353 struct perf_sample *sample,
2354 struct machine *machine,
2355 struct thread *migrated)
2357 struct thread *thread;
2359 u32 max_cpus = sched->max_cpu + 1;
2362 if (sched->summary_only)
2365 max_cpus = sched->max_cpu + 1;
2366 ocpu = perf_evsel__intval(evsel, sample, "orig_cpu");
2367 dcpu = perf_evsel__intval(evsel, sample, "dest_cpu");
2369 thread = machine__findnew_thread(machine, sample->pid, sample->tid);
2373 if (timehist_skip_sample(sched, thread, evsel, sample) &&
2374 timehist_skip_sample(sched, migrated, evsel, sample)) {
2378 timestamp__scnprintf_usec(sample->time, tstr, sizeof(tstr));
2379 printf("%15s [%04d] ", tstr, sample->cpu);
2381 if (sched->show_cpu_visual) {
2386 for (i = 0; i < max_cpus; ++i) {
2387 c = (i == sample->cpu) ? 'm' : ' ';
2393 printf(" %-*s ", comm_width, timehist_get_commstr(thread));
2396 printf(" %9s %9s %9s ", "", "", "");
2398 printf("migrated: %s", timehist_get_commstr(migrated));
2399 printf(" cpu %d => %d", ocpu, dcpu);
2404 static int timehist_migrate_task_event(struct perf_tool *tool,
2405 union perf_event *event __maybe_unused,
2406 struct perf_evsel *evsel,
2407 struct perf_sample *sample,
2408 struct machine *machine)
2410 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
2411 struct thread *thread;
2412 struct thread_runtime *tr = NULL;
2413 /* want pid of migrated task not pid in sample */
2414 const u32 pid = perf_evsel__intval(evsel, sample, "pid");
2416 thread = machine__findnew_thread(machine, 0, pid);
2420 tr = thread__get_runtime(thread);
2426 /* show migrations if requested */
2427 timehist_print_migration_event(sched, evsel, sample, machine, thread);
2432 static int timehist_sched_change_event(struct perf_tool *tool,
2433 union perf_event *event,
2434 struct perf_evsel *evsel,
2435 struct perf_sample *sample,
2436 struct machine *machine)
2438 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
2439 struct perf_time_interval *ptime = &sched->ptime;
2440 struct addr_location al;
2441 struct thread *thread;
2442 struct thread_runtime *tr = NULL;
2443 u64 tprev, t = sample->time;
2445 int state = perf_evsel__intval(evsel, sample, "prev_state");
2448 if (machine__resolve(machine, &al, sample) < 0) {
2449 pr_err("problem processing %d event. skipping it\n",
2450 event->header.type);
2455 thread = timehist_get_thread(sched, sample, machine, evsel);
2456 if (thread == NULL) {
2461 if (timehist_skip_sample(sched, thread, evsel, sample))
2464 tr = thread__get_runtime(thread);
2470 tprev = perf_evsel__get_time(evsel, sample->cpu);
2473 * If start time given:
2474 * - sample time is under window user cares about - skip sample
2475 * - tprev is under window user cares about - reset to start of window
2477 if (ptime->start && ptime->start > t)
2480 if (tprev && ptime->start > tprev)
2481 tprev = ptime->start;
2484 * If end time given:
2485 * - previous sched event is out of window - we are done
2486 * - sample time is beyond window user cares about - reset it
2487 * to close out stats for time window interest
2490 if (tprev > ptime->end)
2497 if (!sched->idle_hist || thread->tid == 0) {
2498 timehist_update_runtime_stats(tr, t, tprev);
2500 if (sched->idle_hist) {
2501 struct idle_thread_runtime *itr = (void *)tr;
2502 struct thread_runtime *last_tr;
2504 BUG_ON(thread->tid != 0);
2506 if (itr->last_thread == NULL)
2509 /* add current idle time as last thread's runtime */
2510 last_tr = thread__get_runtime(itr->last_thread);
2511 if (last_tr == NULL)
2514 timehist_update_runtime_stats(last_tr, t, tprev);
2516 * remove delta time of last thread as it's not updated
2517 * and otherwise it will show an invalid value next
2518 * time. we only care total run time and run stat.
2520 last_tr->dt_run = 0;
2521 last_tr->dt_delay = 0;
2522 last_tr->dt_sleep = 0;
2523 last_tr->dt_iowait = 0;
2524 last_tr->dt_preempt = 0;
2527 callchain_append(&itr->callchain, &itr->cursor, t - tprev);
2529 itr->last_thread = NULL;
2533 if (!sched->summary_only)
2534 timehist_print_sample(sched, sample, &al, thread, t, state);
2537 if (sched->hist_time.start == 0 && t >= ptime->start)
2538 sched->hist_time.start = t;
2539 if (ptime->end == 0 || t <= ptime->end)
2540 sched->hist_time.end = t;
2543 /* time of this sched_switch event becomes last time task seen */
2544 tr->last_time = sample->time;
2546 /* last state is used to determine where to account wait time */
2547 tr->last_state = state;
2549 /* sched out event for task so reset ready to run time */
2550 tr->ready_to_run = 0;
2553 perf_evsel__save_time(evsel, sample->time, sample->cpu);
2558 static int timehist_sched_switch_event(struct perf_tool *tool,
2559 union perf_event *event,
2560 struct perf_evsel *evsel,
2561 struct perf_sample *sample,
2562 struct machine *machine __maybe_unused)
2564 return timehist_sched_change_event(tool, event, evsel, sample, machine);
2567 static int process_lost(struct perf_tool *tool __maybe_unused,
2568 union perf_event *event,
2569 struct perf_sample *sample,
2570 struct machine *machine __maybe_unused)
2574 timestamp__scnprintf_usec(sample->time, tstr, sizeof(tstr));
2575 printf("%15s ", tstr);
2576 printf("lost %" PRIu64 " events on cpu %d\n", event->lost.lost, sample->cpu);
2582 static void print_thread_runtime(struct thread *t,
2583 struct thread_runtime *r)
2585 double mean = avg_stats(&r->run_stats);
2588 printf("%*s %5d %9" PRIu64 " ",
2589 comm_width, timehist_get_commstr(t), t->ppid,
2590 (u64) r->run_stats.n);
2592 print_sched_time(r->total_run_time, 8);
2593 stddev = rel_stddev_stats(stddev_stats(&r->run_stats), mean);
2594 print_sched_time(r->run_stats.min, 6);
2596 print_sched_time((u64) mean, 6);
2598 print_sched_time(r->run_stats.max, 6);
2600 printf("%5.2f", stddev);
2601 printf(" %5" PRIu64, r->migrations);
2605 static void print_thread_waittime(struct thread *t,
2606 struct thread_runtime *r)
2608 printf("%*s %5d %9" PRIu64 " ",
2609 comm_width, timehist_get_commstr(t), t->ppid,
2610 (u64) r->run_stats.n);
2612 print_sched_time(r->total_run_time, 8);
2613 print_sched_time(r->total_sleep_time, 6);
2615 print_sched_time(r->total_iowait_time, 6);
2617 print_sched_time(r->total_preempt_time, 6);
2619 print_sched_time(r->total_delay_time, 6);
2623 struct total_run_stats {
2624 struct perf_sched *sched;
2630 static int __show_thread_runtime(struct thread *t, void *priv)
2632 struct total_run_stats *stats = priv;
2633 struct thread_runtime *r;
2635 if (thread__is_filtered(t))
2638 r = thread__priv(t);
2639 if (r && r->run_stats.n) {
2640 stats->task_count++;
2641 stats->sched_count += r->run_stats.n;
2642 stats->total_run_time += r->total_run_time;
2644 if (stats->sched->show_state)
2645 print_thread_waittime(t, r);
2647 print_thread_runtime(t, r);
2653 static int show_thread_runtime(struct thread *t, void *priv)
2658 return __show_thread_runtime(t, priv);
2661 static int show_deadthread_runtime(struct thread *t, void *priv)
2666 return __show_thread_runtime(t, priv);
2669 static size_t callchain__fprintf_folded(FILE *fp, struct callchain_node *node)
2671 const char *sep = " <- ";
2672 struct callchain_list *chain;
2680 ret = callchain__fprintf_folded(fp, node->parent);
2683 list_for_each_entry(chain, &node->val, list) {
2684 if (chain->ip >= PERF_CONTEXT_MAX)
2686 if (chain->ms.sym && chain->ms.sym->ignore)
2688 ret += fprintf(fp, "%s%s", first ? "" : sep,
2689 callchain_list__sym_name(chain, bf, sizeof(bf),
2697 static size_t timehist_print_idlehist_callchain(struct rb_root *root)
2701 struct callchain_node *chain;
2702 struct rb_node *rb_node = rb_first(root);
2704 printf(" %16s %8s %s\n", "Idle time (msec)", "Count", "Callchains");
2705 printf(" %.16s %.8s %.50s\n", graph_dotted_line, graph_dotted_line,
2709 chain = rb_entry(rb_node, struct callchain_node, rb_node);
2710 rb_node = rb_next(rb_node);
2712 ret += fprintf(fp, " ");
2713 print_sched_time(chain->hit, 12);
2714 ret += 16; /* print_sched_time returns 2nd arg + 4 */
2715 ret += fprintf(fp, " %8d ", chain->count);
2716 ret += callchain__fprintf_folded(fp, chain);
2717 ret += fprintf(fp, "\n");
2723 static void timehist_print_summary(struct perf_sched *sched,
2724 struct perf_session *session)
2726 struct machine *m = &session->machines.host;
2727 struct total_run_stats totals;
2730 struct thread_runtime *r;
2732 u64 hist_time = sched->hist_time.end - sched->hist_time.start;
2734 memset(&totals, 0, sizeof(totals));
2735 totals.sched = sched;
2737 if (sched->idle_hist) {
2738 printf("\nIdle-time summary\n");
2739 printf("%*s parent sched-out ", comm_width, "comm");
2740 printf(" idle-time min-idle avg-idle max-idle stddev migrations\n");
2741 } else if (sched->show_state) {
2742 printf("\nWait-time summary\n");
2743 printf("%*s parent sched-in ", comm_width, "comm");
2744 printf(" run-time sleep iowait preempt delay\n");
2746 printf("\nRuntime summary\n");
2747 printf("%*s parent sched-in ", comm_width, "comm");
2748 printf(" run-time min-run avg-run max-run stddev migrations\n");
2750 printf("%*s (count) ", comm_width, "");
2751 printf(" (msec) (msec) (msec) (msec) %s\n",
2752 sched->show_state ? "(msec)" : "%");
2753 printf("%.117s\n", graph_dotted_line);
2755 machine__for_each_thread(m, show_thread_runtime, &totals);
2756 task_count = totals.task_count;
2758 printf("<no still running tasks>\n");
2760 printf("\nTerminated tasks:\n");
2761 machine__for_each_thread(m, show_deadthread_runtime, &totals);
2762 if (task_count == totals.task_count)
2763 printf("<no terminated tasks>\n");
2765 /* CPU idle stats not tracked when samples were skipped */
2766 if (sched->skipped_samples && !sched->idle_hist)
2769 printf("\nIdle stats:\n");
2770 for (i = 0; i < idle_max_cpu; ++i) {
2771 t = idle_threads[i];
2775 r = thread__priv(t);
2776 if (r && r->run_stats.n) {
2777 totals.sched_count += r->run_stats.n;
2778 printf(" CPU %2d idle for ", i);
2779 print_sched_time(r->total_run_time, 6);
2780 printf(" msec (%6.2f%%)\n", 100.0 * r->total_run_time / hist_time);
2782 printf(" CPU %2d idle entire time window\n", i);
2785 if (sched->idle_hist && symbol_conf.use_callchain) {
2786 callchain_param.mode = CHAIN_FOLDED;
2787 callchain_param.value = CCVAL_PERIOD;
2789 callchain_register_param(&callchain_param);
2791 printf("\nIdle stats by callchain:\n");
2792 for (i = 0; i < idle_max_cpu; ++i) {
2793 struct idle_thread_runtime *itr;
2795 t = idle_threads[i];
2799 itr = thread__priv(t);
2803 callchain_param.sort(&itr->sorted_root, &itr->callchain,
2804 0, &callchain_param);
2806 printf(" CPU %2d:", i);
2807 print_sched_time(itr->tr.total_run_time, 6);
2809 timehist_print_idlehist_callchain(&itr->sorted_root);
2815 " Total number of unique tasks: %" PRIu64 "\n"
2816 "Total number of context switches: %" PRIu64 "\n",
2817 totals.task_count, totals.sched_count);
2819 printf(" Total run time (msec): ");
2820 print_sched_time(totals.total_run_time, 2);
2823 printf(" Total scheduling time (msec): ");
2824 print_sched_time(hist_time, 2);
2825 printf(" (x %d)\n", sched->max_cpu);
2828 typedef int (*sched_handler)(struct perf_tool *tool,
2829 union perf_event *event,
2830 struct perf_evsel *evsel,
2831 struct perf_sample *sample,
2832 struct machine *machine);
2834 static int perf_timehist__process_sample(struct perf_tool *tool,
2835 union perf_event *event,
2836 struct perf_sample *sample,
2837 struct perf_evsel *evsel,
2838 struct machine *machine)
2840 struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
2842 int this_cpu = sample->cpu;
2844 if (this_cpu > sched->max_cpu)
2845 sched->max_cpu = this_cpu;
2847 if (evsel->handler != NULL) {
2848 sched_handler f = evsel->handler;
2850 err = f(tool, event, evsel, sample, machine);
2856 static int timehist_check_attr(struct perf_sched *sched,
2857 struct perf_evlist *evlist)
2859 struct perf_evsel *evsel;
2860 struct evsel_runtime *er;
2862 list_for_each_entry(evsel, &evlist->entries, node) {
2863 er = perf_evsel__get_runtime(evsel);
2865 pr_err("Failed to allocate memory for evsel runtime data\n");
2869 if (sched->show_callchain &&
2870 !(evsel->attr.sample_type & PERF_SAMPLE_CALLCHAIN)) {
2871 pr_info("Samples do not have callchains.\n");
2872 sched->show_callchain = 0;
2873 symbol_conf.use_callchain = 0;
2880 static int perf_sched__timehist(struct perf_sched *sched)
2882 const struct perf_evsel_str_handler handlers[] = {
2883 { "sched:sched_switch", timehist_sched_switch_event, },
2884 { "sched:sched_wakeup", timehist_sched_wakeup_event, },
2885 { "sched:sched_wakeup_new", timehist_sched_wakeup_event, },
2887 const struct perf_evsel_str_handler migrate_handlers[] = {
2888 { "sched:sched_migrate_task", timehist_migrate_task_event, },
2890 struct perf_data_file file = {
2892 .mode = PERF_DATA_MODE_READ,
2893 .force = sched->force,
2896 struct perf_session *session;
2897 struct perf_evlist *evlist;
2901 * event handlers for timehist option
2903 sched->tool.sample = perf_timehist__process_sample;
2904 sched->tool.mmap = perf_event__process_mmap;
2905 sched->tool.comm = perf_event__process_comm;
2906 sched->tool.exit = perf_event__process_exit;
2907 sched->tool.fork = perf_event__process_fork;
2908 sched->tool.lost = process_lost;
2909 sched->tool.attr = perf_event__process_attr;
2910 sched->tool.tracing_data = perf_event__process_tracing_data;
2911 sched->tool.build_id = perf_event__process_build_id;
2913 sched->tool.ordered_events = true;
2914 sched->tool.ordering_requires_timestamps = true;
2916 symbol_conf.use_callchain = sched->show_callchain;
2918 session = perf_session__new(&file, false, &sched->tool);
2919 if (session == NULL)
2922 evlist = session->evlist;
2924 symbol__init(&session->header.env);
2926 if (perf_time__parse_str(&sched->ptime, sched->time_str) != 0) {
2927 pr_err("Invalid time string\n");
2931 if (timehist_check_attr(sched, evlist) != 0)
2936 /* setup per-evsel handlers */
2937 if (perf_session__set_tracepoints_handlers(session, handlers))
2940 /* sched_switch event at a minimum needs to exist */
2941 if (!perf_evlist__find_tracepoint_by_name(session->evlist,
2942 "sched:sched_switch")) {
2943 pr_err("No sched_switch events found. Have you run 'perf sched record'?\n");
2947 if (sched->show_migrations &&
2948 perf_session__set_tracepoints_handlers(session, migrate_handlers))
2951 /* pre-allocate struct for per-CPU idle stats */
2952 sched->max_cpu = session->header.env.nr_cpus_online;
2953 if (sched->max_cpu == 0)
2955 if (init_idle_threads(sched->max_cpu))
2958 /* summary_only implies summary option, but don't overwrite summary if set */
2959 if (sched->summary_only)
2960 sched->summary = sched->summary_only;
2962 if (!sched->summary_only)
2963 timehist_header(sched);
2965 err = perf_session__process_events(session);
2967 pr_err("Failed to process events, error %d", err);
2971 sched->nr_events = evlist->stats.nr_events[0];
2972 sched->nr_lost_events = evlist->stats.total_lost;
2973 sched->nr_lost_chunks = evlist->stats.nr_events[PERF_RECORD_LOST];
2976 timehist_print_summary(sched, session);
2979 free_idle_threads();
2980 perf_session__delete(session);
2986 static void print_bad_events(struct perf_sched *sched)
2988 if (sched->nr_unordered_timestamps && sched->nr_timestamps) {
2989 printf(" INFO: %.3f%% unordered timestamps (%ld out of %ld)\n",
2990 (double)sched->nr_unordered_timestamps/(double)sched->nr_timestamps*100.0,
2991 sched->nr_unordered_timestamps, sched->nr_timestamps);
2993 if (sched->nr_lost_events && sched->nr_events) {
2994 printf(" INFO: %.3f%% lost events (%ld out of %ld, in %ld chunks)\n",
2995 (double)sched->nr_lost_events/(double)sched->nr_events * 100.0,
2996 sched->nr_lost_events, sched->nr_events, sched->nr_lost_chunks);
2998 if (sched->nr_context_switch_bugs && sched->nr_timestamps) {
2999 printf(" INFO: %.3f%% context switch bugs (%ld out of %ld)",
3000 (double)sched->nr_context_switch_bugs/(double)sched->nr_timestamps*100.0,
3001 sched->nr_context_switch_bugs, sched->nr_timestamps);
3002 if (sched->nr_lost_events)
3003 printf(" (due to lost events?)");
3008 static void __merge_work_atoms(struct rb_root *root, struct work_atoms *data)
3010 struct rb_node **new = &(root->rb_node), *parent = NULL;
3011 struct work_atoms *this;
3012 const char *comm = thread__comm_str(data->thread), *this_comm;
3017 this = container_of(*new, struct work_atoms, node);
3020 this_comm = thread__comm_str(this->thread);
3021 cmp = strcmp(comm, this_comm);
3023 new = &((*new)->rb_left);
3024 } else if (cmp < 0) {
3025 new = &((*new)->rb_right);
3028 this->total_runtime += data->total_runtime;
3029 this->nb_atoms += data->nb_atoms;
3030 this->total_lat += data->total_lat;
3031 list_splice(&data->work_list, &this->work_list);
3032 if (this->max_lat < data->max_lat) {
3033 this->max_lat = data->max_lat;
3034 this->max_lat_at = data->max_lat_at;
3042 rb_link_node(&data->node, parent, new);
3043 rb_insert_color(&data->node, root);
3046 static void perf_sched__merge_lat(struct perf_sched *sched)
3048 struct work_atoms *data;
3049 struct rb_node *node;
3051 if (sched->skip_merge)
3054 while ((node = rb_first(&sched->atom_root))) {
3055 rb_erase(node, &sched->atom_root);
3056 data = rb_entry(node, struct work_atoms, node);
3057 __merge_work_atoms(&sched->merged_atom_root, data);
3061 static int perf_sched__lat(struct perf_sched *sched)
3063 struct rb_node *next;
3067 if (perf_sched__read_events(sched))
3070 perf_sched__merge_lat(sched);
3071 perf_sched__sort_lat(sched);
3073 printf("\n -----------------------------------------------------------------------------------------------------------------\n");
3074 printf(" Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |\n");
3075 printf(" -----------------------------------------------------------------------------------------------------------------\n");
3077 next = rb_first(&sched->sorted_atom_root);
3080 struct work_atoms *work_list;
3082 work_list = rb_entry(next, struct work_atoms, node);
3083 output_lat_thread(sched, work_list);
3084 next = rb_next(next);
3085 thread__zput(work_list->thread);
3088 printf(" -----------------------------------------------------------------------------------------------------------------\n");
3089 printf(" TOTAL: |%11.3f ms |%9" PRIu64 " |\n",
3090 (double)sched->all_runtime / NSEC_PER_MSEC, sched->all_count);
3092 printf(" ---------------------------------------------------\n");
3094 print_bad_events(sched);
3100 static int setup_map_cpus(struct perf_sched *sched)
3102 struct cpu_map *map;
3104 sched->max_cpu = sysconf(_SC_NPROCESSORS_CONF);
3106 if (sched->map.comp) {
3107 sched->map.comp_cpus = zalloc(sched->max_cpu * sizeof(int));
3108 if (!sched->map.comp_cpus)
3112 if (!sched->map.cpus_str)
3115 map = cpu_map__new(sched->map.cpus_str);
3117 pr_err("failed to get cpus map from %s\n", sched->map.cpus_str);
3121 sched->map.cpus = map;
3125 static int setup_color_pids(struct perf_sched *sched)
3127 struct thread_map *map;
3129 if (!sched->map.color_pids_str)
3132 map = thread_map__new_by_tid_str(sched->map.color_pids_str);
3134 pr_err("failed to get thread map from %s\n", sched->map.color_pids_str);
3138 sched->map.color_pids = map;
3142 static int setup_color_cpus(struct perf_sched *sched)
3144 struct cpu_map *map;
3146 if (!sched->map.color_cpus_str)
3149 map = cpu_map__new(sched->map.color_cpus_str);
3151 pr_err("failed to get thread map from %s\n", sched->map.color_cpus_str);
3155 sched->map.color_cpus = map;
3159 static int perf_sched__map(struct perf_sched *sched)
3161 if (setup_map_cpus(sched))
3164 if (setup_color_pids(sched))
3167 if (setup_color_cpus(sched))
3171 if (perf_sched__read_events(sched))
3173 print_bad_events(sched);
3177 static int perf_sched__replay(struct perf_sched *sched)
3181 calibrate_run_measurement_overhead(sched);
3182 calibrate_sleep_measurement_overhead(sched);
3184 test_calibrations(sched);
3186 if (perf_sched__read_events(sched))
3189 printf("nr_run_events: %ld\n", sched->nr_run_events);
3190 printf("nr_sleep_events: %ld\n", sched->nr_sleep_events);
3191 printf("nr_wakeup_events: %ld\n", sched->nr_wakeup_events);
3193 if (sched->targetless_wakeups)
3194 printf("target-less wakeups: %ld\n", sched->targetless_wakeups);
3195 if (sched->multitarget_wakeups)
3196 printf("multi-target wakeups: %ld\n", sched->multitarget_wakeups);
3197 if (sched->nr_run_events_optimized)
3198 printf("run atoms optimized: %ld\n",
3199 sched->nr_run_events_optimized);
3201 print_task_traces(sched);
3202 add_cross_task_wakeups(sched);
3204 create_tasks(sched);
3205 printf("------------------------------------------------------------\n");
3206 for (i = 0; i < sched->replay_repeat; i++)
3207 run_one_test(sched);
3212 static void setup_sorting(struct perf_sched *sched, const struct option *options,
3213 const char * const usage_msg[])
3215 char *tmp, *tok, *str = strdup(sched->sort_order);
3217 for (tok = strtok_r(str, ", ", &tmp);
3218 tok; tok = strtok_r(NULL, ", ", &tmp)) {
3219 if (sort_dimension__add(tok, &sched->sort_list) < 0) {
3220 usage_with_options_msg(usage_msg, options,
3221 "Unknown --sort key: `%s'", tok);
3227 sort_dimension__add("pid", &sched->cmp_pid);
3230 static int __cmd_record(int argc, const char **argv)
3232 unsigned int rec_argc, i, j;
3233 const char **rec_argv;
3234 const char * const record_args[] = {
3240 "-e", "sched:sched_switch",
3241 "-e", "sched:sched_stat_wait",
3242 "-e", "sched:sched_stat_sleep",
3243 "-e", "sched:sched_stat_iowait",
3244 "-e", "sched:sched_stat_runtime",
3245 "-e", "sched:sched_process_fork",
3246 "-e", "sched:sched_wakeup",
3247 "-e", "sched:sched_wakeup_new",
3248 "-e", "sched:sched_migrate_task",
3251 rec_argc = ARRAY_SIZE(record_args) + argc - 1;
3252 rec_argv = calloc(rec_argc + 1, sizeof(char *));
3254 if (rec_argv == NULL)
3257 for (i = 0; i < ARRAY_SIZE(record_args); i++)
3258 rec_argv[i] = strdup(record_args[i]);
3260 for (j = 1; j < (unsigned int)argc; j++, i++)
3261 rec_argv[i] = argv[j];
3263 BUG_ON(i != rec_argc);
3265 return cmd_record(i, rec_argv, NULL);
3268 int cmd_sched(int argc, const char **argv, const char *prefix __maybe_unused)
3270 const char default_sort_order[] = "avg, max, switch, runtime";
3271 struct perf_sched sched = {
3273 .sample = perf_sched__process_tracepoint_sample,
3274 .comm = perf_event__process_comm,
3275 .lost = perf_event__process_lost,
3276 .fork = perf_sched__process_fork_event,
3277 .ordered_events = true,
3279 .cmp_pid = LIST_HEAD_INIT(sched.cmp_pid),
3280 .sort_list = LIST_HEAD_INIT(sched.sort_list),
3281 .start_work_mutex = PTHREAD_MUTEX_INITIALIZER,
3282 .work_done_wait_mutex = PTHREAD_MUTEX_INITIALIZER,
3283 .sort_order = default_sort_order,
3284 .replay_repeat = 10,
3286 .next_shortname1 = 'A',
3287 .next_shortname2 = '0',
3289 .show_callchain = 1,
3292 const struct option sched_options[] = {
3293 OPT_STRING('i', "input", &input_name, "file",
3295 OPT_INCR('v', "verbose", &verbose,
3296 "be more verbose (show symbol address, etc)"),
3297 OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
3298 "dump raw trace in ASCII"),
3299 OPT_BOOLEAN('f', "force", &sched.force, "don't complain, do it"),
3302 const struct option latency_options[] = {
3303 OPT_STRING('s', "sort", &sched.sort_order, "key[,key2...]",
3304 "sort by key(s): runtime, switch, avg, max"),
3305 OPT_INTEGER('C', "CPU", &sched.profile_cpu,
3306 "CPU to profile on"),
3307 OPT_BOOLEAN('p', "pids", &sched.skip_merge,
3308 "latency stats per pid instead of per comm"),
3309 OPT_PARENT(sched_options)
3311 const struct option replay_options[] = {
3312 OPT_UINTEGER('r', "repeat", &sched.replay_repeat,
3313 "repeat the workload replay N times (-1: infinite)"),
3314 OPT_PARENT(sched_options)
3316 const struct option map_options[] = {
3317 OPT_BOOLEAN(0, "compact", &sched.map.comp,
3318 "map output in compact mode"),
3319 OPT_STRING(0, "color-pids", &sched.map.color_pids_str, "pids",
3320 "highlight given pids in map"),
3321 OPT_STRING(0, "color-cpus", &sched.map.color_cpus_str, "cpus",
3322 "highlight given CPUs in map"),
3323 OPT_STRING(0, "cpus", &sched.map.cpus_str, "cpus",
3324 "display given CPUs in map"),
3325 OPT_PARENT(sched_options)
3327 const struct option timehist_options[] = {
3328 OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name,
3329 "file", "vmlinux pathname"),
3330 OPT_STRING(0, "kallsyms", &symbol_conf.kallsyms_name,
3331 "file", "kallsyms pathname"),
3332 OPT_BOOLEAN('g', "call-graph", &sched.show_callchain,
3333 "Display call chains if present (default on)"),
3334 OPT_UINTEGER(0, "max-stack", &sched.max_stack,
3335 "Maximum number of functions to display backtrace."),
3336 OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
3337 "Look for files with symbols relative to this directory"),
3338 OPT_BOOLEAN('s', "summary", &sched.summary_only,
3339 "Show only syscall summary with statistics"),
3340 OPT_BOOLEAN('S', "with-summary", &sched.summary,
3341 "Show all syscalls and summary with statistics"),
3342 OPT_BOOLEAN('w', "wakeups", &sched.show_wakeups, "Show wakeup events"),
3343 OPT_BOOLEAN('M', "migrations", &sched.show_migrations, "Show migration events"),
3344 OPT_BOOLEAN('V', "cpu-visual", &sched.show_cpu_visual, "Add CPU visual"),
3345 OPT_BOOLEAN('I', "idle-hist", &sched.idle_hist, "Show idle events only"),
3346 OPT_STRING(0, "time", &sched.time_str, "str",
3347 "Time span for analysis (start,stop)"),
3348 OPT_BOOLEAN(0, "state", &sched.show_state, "Show task state when sched-out"),
3349 OPT_PARENT(sched_options)
3352 const char * const latency_usage[] = {
3353 "perf sched latency [<options>]",
3356 const char * const replay_usage[] = {
3357 "perf sched replay [<options>]",
3360 const char * const map_usage[] = {
3361 "perf sched map [<options>]",
3364 const char * const timehist_usage[] = {
3365 "perf sched timehist [<options>]",
3368 const char *const sched_subcommands[] = { "record", "latency", "map",
3371 const char *sched_usage[] = {
3375 struct trace_sched_handler lat_ops = {
3376 .wakeup_event = latency_wakeup_event,
3377 .switch_event = latency_switch_event,
3378 .runtime_event = latency_runtime_event,
3379 .migrate_task_event = latency_migrate_task_event,
3381 struct trace_sched_handler map_ops = {
3382 .switch_event = map_switch_event,
3384 struct trace_sched_handler replay_ops = {
3385 .wakeup_event = replay_wakeup_event,
3386 .switch_event = replay_switch_event,
3387 .fork_event = replay_fork_event,
3391 for (i = 0; i < ARRAY_SIZE(sched.curr_pid); i++)
3392 sched.curr_pid[i] = -1;
3394 argc = parse_options_subcommand(argc, argv, sched_options, sched_subcommands,
3395 sched_usage, PARSE_OPT_STOP_AT_NON_OPTION);
3397 usage_with_options(sched_usage, sched_options);
3400 * Aliased to 'perf script' for now:
3402 if (!strcmp(argv[0], "script"))
3403 return cmd_script(argc, argv, prefix);
3405 if (!strncmp(argv[0], "rec", 3)) {
3406 return __cmd_record(argc, argv);
3407 } else if (!strncmp(argv[0], "lat", 3)) {
3408 sched.tp_handler = &lat_ops;
3410 argc = parse_options(argc, argv, latency_options, latency_usage, 0);
3412 usage_with_options(latency_usage, latency_options);
3414 setup_sorting(&sched, latency_options, latency_usage);
3415 return perf_sched__lat(&sched);
3416 } else if (!strcmp(argv[0], "map")) {
3418 argc = parse_options(argc, argv, map_options, map_usage, 0);
3420 usage_with_options(map_usage, map_options);
3422 sched.tp_handler = &map_ops;
3423 setup_sorting(&sched, latency_options, latency_usage);
3424 return perf_sched__map(&sched);
3425 } else if (!strncmp(argv[0], "rep", 3)) {
3426 sched.tp_handler = &replay_ops;
3428 argc = parse_options(argc, argv, replay_options, replay_usage, 0);
3430 usage_with_options(replay_usage, replay_options);
3432 return perf_sched__replay(&sched);
3433 } else if (!strcmp(argv[0], "timehist")) {
3435 argc = parse_options(argc, argv, timehist_options,
3438 usage_with_options(timehist_usage, timehist_options);
3440 if (sched.show_wakeups && sched.summary_only) {
3441 pr_err(" Error: -s and -w are mutually exclusive.\n");
3442 parse_options_usage(timehist_usage, timehist_options, "s", true);
3443 parse_options_usage(NULL, timehist_options, "w", true);
3447 return perf_sched__timehist(&sched);
3449 usage_with_options(sched_usage, sched_options);