tools/perf/builtin-sched.c

   1 #include "builtin.h"
   2 #include "perf.h"
   3
   4 #include "util/util.h"
   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
  14 #include "util/parse-options.h"
  15 #include "util/trace-event.h"
  16
  17 #include "util/debug.h"
  18
  19 #include <sys/prctl.h>
  20 #include <sys/resource.h>
  21
  22 #include <semaphore.h>
  23 #include <pthread.h>
  24 #include <math.h>
  25
  26 #define PR_SET_NAME             15               /* Set process name */
  27 #define MAX_CPUS                4096
  28 #define COMM_LEN                20
  29 #define SYM_LEN                 129
  30 #define MAX_PID                 65536
  31
  32 struct sched_atom;
  33
  34 struct task_desc {
  35         unsigned long           nr;
  36         unsigned long           pid;
  37         char                    comm[COMM_LEN];
  38
  39         unsigned long           nr_events;
  40         unsigned long           curr_event;
  41         struct sched_atom       **atoms;
  42
  43         pthread_t               thread;
  44         sem_t                   sleep_sem;
  45
  46         sem_t                   ready_for_work;
  47         sem_t                   work_done_sem;
  48
  49         u64                     cpu_usage;
  50 };
  51
  52 enum sched_event_type {
  53         SCHED_EVENT_RUN,
  54         SCHED_EVENT_SLEEP,
  55         SCHED_EVENT_WAKEUP,
  56         SCHED_EVENT_MIGRATION,
  57 };
  58
  59 struct sched_atom {
  60         enum sched_event_type   type;
  61         int                     specific_wait;
  62         u64                     timestamp;
  63         u64                     duration;
  64         unsigned long           nr;
  65         sem_t                   *wait_sem;
  66         struct task_desc        *wakee;
  67 };
  68
  69 #define TASK_STATE_TO_CHAR_STR "RSDTtZX"
  70
  71 enum thread_state {
  72         THREAD_SLEEPING = 0,
  73         THREAD_WAIT_CPU,
  74         THREAD_SCHED_IN,
  75         THREAD_IGNORE
  76 };
  77
  78 struct work_atom {
  79         struct list_head        list;
  80         enum thread_state       state;
  81         u64                     sched_out_time;
  82         u64                     wake_up_time;
  83         u64                     sched_in_time;
  84         u64                     runtime;
  85 };
  86
  87 struct work_atoms {
  88         struct list_head        work_list;
  89         struct thread           *thread;
  90         struct rb_node          node;
  91         u64                     max_lat;
  92         u64                     max_lat_at;
  93         u64                     total_lat;
  94         u64                     nb_atoms;
  95         u64                     total_runtime;
  96 };
  97
  98 typedef int (*sort_fn_t)(struct work_atoms *, struct work_atoms *);
  99
 100 struct perf_sched;
 101
 102 struct trace_sched_handler {
 103         int (*switch_event)(struct perf_sched *sched, struct perf_evsel *evsel,
 104                             struct perf_sample *sample, struct machine *machine);
 105
 106         int (*runtime_event)(struct perf_sched *sched, struct perf_evsel *evsel,
 107                              struct perf_sample *sample, struct machine *machine);
 108
 109         int (*wakeup_event)(struct perf_sched *sched, struct perf_evsel *evsel,
 110                             struct perf_sample *sample, struct machine *machine);
 111
 112         int (*fork_event)(struct perf_sched *sched, struct perf_evsel *evsel,
 113                           struct perf_sample *sample);
 114
 115         int (*migrate_task_event)(struct perf_sched *sched,
 116                                   struct perf_evsel *evsel,
 117                                   struct perf_sample *sample,
 118                                   struct machine *machine);
 119 };
 120
 121 struct perf_sched {
 122         struct perf_tool tool;
 123         const char       *input_name;
 124         const char       *sort_order;
 125         unsigned long    nr_tasks;
 126         struct task_desc *pid_to_task[MAX_PID];
 127         struct task_desc **tasks;
 128         const struct trace_sched_handler *tp_handler;
 129         pthread_mutex_t  start_work_mutex;
 130         pthread_mutex_t  work_done_wait_mutex;
 131         int              profile_cpu;
 132 /*
 133  * Track the current task - that way we can know whether there's any
 134  * weird events, such as a task being switched away that is not current.
 135  */
 136         int              max_cpu;
 137         u32              curr_pid[MAX_CPUS];
 138         struct thread    *curr_thread[MAX_CPUS];
 139         char             next_shortname1;
 140         char             next_shortname2;
 141         unsigned int     replay_repeat;
 142         unsigned long    nr_run_events;
 143         unsigned long    nr_sleep_events;
 144         unsigned long    nr_wakeup_events;
 145         unsigned long    nr_sleep_corrections;
 146         unsigned long    nr_run_events_optimized;
 147         unsigned long    targetless_wakeups;
 148         unsigned long    multitarget_wakeups;
 149         unsigned long    nr_runs;
 150         unsigned long    nr_timestamps;
 151         unsigned long    nr_unordered_timestamps;
 152         unsigned long    nr_state_machine_bugs;
 153         unsigned long    nr_context_switch_bugs;
 154         unsigned long    nr_events;
 155         unsigned long    nr_lost_chunks;
 156         unsigned long    nr_lost_events;
 157         u64              run_measurement_overhead;
 158         u64              sleep_measurement_overhead;
 159         u64              start_time;
 160         u64              cpu_usage;
 161         u64              runavg_cpu_usage;
 162         u64              parent_cpu_usage;
 163         u64              runavg_parent_cpu_usage;
 164         u64              sum_runtime;
 165         u64              sum_fluct;
 166         u64              run_avg;
 167         u64              all_runtime;
 168         u64              all_count;
 169         u64              cpu_last_switched[MAX_CPUS];
 170         struct rb_root   atom_root, sorted_atom_root;
 171         struct list_head sort_list, cmp_pid;
 172 };
 173
 174 static u64 get_nsecs(void)
 175 {
 176         struct timespec ts;
 177
 178         clock_gettime(CLOCK_MONOTONIC, &ts);
 179
 180         return ts.tv_sec * 1000000000ULL + ts.tv_nsec;
 181 }
 182
 183 static void burn_nsecs(struct perf_sched *sched, u64 nsecs)
 184 {
 185         u64 T0 = get_nsecs(), T1;
 186
 187         do {
 188                 T1 = get_nsecs();
 189         } while (T1 + sched->run_measurement_overhead < T0 + nsecs);
 190 }
 191
 192 static void sleep_nsecs(u64 nsecs)
 193 {
 194         struct timespec ts;
 195
 196         ts.tv_nsec = nsecs % 999999999;
 197         ts.tv_sec = nsecs / 999999999;
 198
 199         nanosleep(&ts, NULL);
 200 }
 201
 202 static void calibrate_run_measurement_overhead(struct perf_sched *sched)
 203 {
 204         u64 T0, T1, delta, min_delta = 1000000000ULL;
 205         int i;
 206
 207         for (i = 0; i < 10; i++) {
 208                 T0 = get_nsecs();
 209                 burn_nsecs(sched, 0);
 210                 T1 = get_nsecs();
 211                 delta = T1-T0;
 212                 min_delta = min(min_delta, delta);
 213         }
 214         sched->run_measurement_overhead = min_delta;
 215
 216         printf("run measurement overhead: %" PRIu64 " nsecs\n", min_delta);
 217 }
 218
 219 static void calibrate_sleep_measurement_overhead(struct perf_sched *sched)
 220 {
 221         u64 T0, T1, delta, min_delta = 1000000000ULL;
 222         int i;
 223
 224         for (i = 0; i < 10; i++) {
 225                 T0 = get_nsecs();
 226                 sleep_nsecs(10000);
 227                 T1 = get_nsecs();
 228                 delta = T1-T0;
 229                 min_delta = min(min_delta, delta);
 230         }
 231         min_delta -= 10000;
 232         sched->sleep_measurement_overhead = min_delta;
 233
 234         printf("sleep measurement overhead: %" PRIu64 " nsecs\n", min_delta);
 235 }
 236
 237 static struct sched_atom *
 238 get_new_event(struct task_desc *task, u64 timestamp)
 239 {
 240         struct sched_atom *event = zalloc(sizeof(*event));
 241         unsigned long idx = task->nr_events;
 242         size_t size;
 243
 244         event->timestamp = timestamp;
 245         event->nr = idx;
 246
 247         task->nr_events++;
 248         size = sizeof(struct sched_atom *) * task->nr_events;
 249         task->atoms = realloc(task->atoms, size);
 250         BUG_ON(!task->atoms);
 251
 252         task->atoms[idx] = event;
 253
 254         return event;
 255 }
 256
 257 static struct sched_atom *last_event(struct task_desc *task)
 258 {
 259         if (!task->nr_events)
 260                 return NULL;
 261
 262         return task->atoms[task->nr_events - 1];
 263 }
 264
 265 static void add_sched_event_run(struct perf_sched *sched, struct task_desc *task,
 266                                 u64 timestamp, u64 duration)
 267 {
 268         struct sched_atom *event, *curr_event = last_event(task);
 269
 270         /*
 271          * optimize an existing RUN event by merging this one
 272          * to it:
 273          */
 274         if (curr_event && curr_event->type == SCHED_EVENT_RUN) {
 275                 sched->nr_run_events_optimized++;
 276                 curr_event->duration += duration;
 277                 return;
 278         }
 279
 280         event = get_new_event(task, timestamp);
 281
 282         event->type = SCHED_EVENT_RUN;
 283         event->duration = duration;
 284
 285         sched->nr_run_events++;
 286 }
 287
 288 static void add_sched_event_wakeup(struct perf_sched *sched, struct task_desc *task,
 289                                    u64 timestamp, struct task_desc *wakee)
 290 {
 291         struct sched_atom *event, *wakee_event;
 292
 293         event = get_new_event(task, timestamp);
 294         event->type = SCHED_EVENT_WAKEUP;
 295         event->wakee = wakee;
 296
 297         wakee_event = last_event(wakee);
 298         if (!wakee_event || wakee_event->type != SCHED_EVENT_SLEEP) {
 299                 sched->targetless_wakeups++;
 300                 return;
 301         }
 302         if (wakee_event->wait_sem) {
 303                 sched->multitarget_wakeups++;
 304                 return;
 305         }
 306
 307         wakee_event->wait_sem = zalloc(sizeof(*wakee_event->wait_sem));
 308         sem_init(wakee_event->wait_sem, 0, 0);
 309         wakee_event->specific_wait = 1;
 310         event->wait_sem = wakee_event->wait_sem;
 311
 312         sched->nr_wakeup_events++;
 313 }
 314
 315 static void add_sched_event_sleep(struct perf_sched *sched, struct task_desc *task,
 316                                   u64 timestamp, u64 task_state __maybe_unused)
 317 {
 318         struct sched_atom *event = get_new_event(task, timestamp);
 319
 320         event->type = SCHED_EVENT_SLEEP;
 321
 322         sched->nr_sleep_events++;
 323 }
 324
 325 static struct task_desc *register_pid(struct perf_sched *sched,
 326                                       unsigned long pid, const char *comm)
 327 {
 328         struct task_desc *task;
 329
 330         BUG_ON(pid >= MAX_PID);
 331
 332         task = sched->pid_to_task[pid];
 333
 334         if (task)
 335                 return task;
 336
 337         task = zalloc(sizeof(*task));
 338         task->pid = pid;
 339         task->nr = sched->nr_tasks;
 340         strcpy(task->comm, comm);
 341         /*
 342          * every task starts in sleeping state - this gets ignored
 343          * if there's no wakeup pointing to this sleep state:
 344          */
 345         add_sched_event_sleep(sched, task, 0, 0);
 346
 347         sched->pid_to_task[pid] = task;
 348         sched->nr_tasks++;
 349         sched->tasks = realloc(sched->tasks, sched->nr_tasks * sizeof(struct task_task *));
 350         BUG_ON(!sched->tasks);
 351         sched->tasks[task->nr] = task;
 352
 353         if (verbose)
 354                 printf("registered task #%ld, PID %ld (%s)\n", sched->nr_tasks, pid, comm);
 355
 356         return task;
 357 }
 358
 359
 360 static void print_task_traces(struct perf_sched *sched)
 361 {
 362         struct task_desc *task;
 363         unsigned long i;
 364
 365         for (i = 0; i < sched->nr_tasks; i++) {
 366                 task = sched->tasks[i];
 367                 printf("task %6ld (%20s:%10ld), nr_events: %ld\n",
 368                         task->nr, task->comm, task->pid, task->nr_events);
 369         }
 370 }
 371
 372 static void add_cross_task_wakeups(struct perf_sched *sched)
 373 {
 374         struct task_desc *task1, *task2;
 375         unsigned long i, j;
 376
 377         for (i = 0; i < sched->nr_tasks; i++) {
 378                 task1 = sched->tasks[i];
 379                 j = i + 1;
 380                 if (j == sched->nr_tasks)
 381                         j = 0;
 382                 task2 = sched->tasks[j];
 383                 add_sched_event_wakeup(sched, task1, 0, task2);
 384         }
 385 }
 386
 387 static void perf_sched__process_event(struct perf_sched *sched,
 388                                       struct sched_atom *atom)
 389 {
 390         int ret = 0;
 391
 392         switch (atom->type) {
 393                 case SCHED_EVENT_RUN:
 394                         burn_nsecs(sched, atom->duration);
 395                         break;
 396                 case SCHED_EVENT_SLEEP:
 397                         if (atom->wait_sem)
 398                                 ret = sem_wait(atom->wait_sem);
 399                         BUG_ON(ret);
 400                         break;
 401                 case SCHED_EVENT_WAKEUP:
 402                         if (atom->wait_sem)
 403                                 ret = sem_post(atom->wait_sem);
 404                         BUG_ON(ret);
 405                         break;
 406                 case SCHED_EVENT_MIGRATION:
 407                         break;
 408                 default:
 409                         BUG_ON(1);
 410         }
 411 }
 412
 413 static u64 get_cpu_usage_nsec_parent(void)
 414 {
 415         struct rusage ru;
 416         u64 sum;
 417         int err;
 418
 419         err = getrusage(RUSAGE_SELF, &ru);
 420         BUG_ON(err);
 421
 422         sum =  ru.ru_utime.tv_sec*1e9 + ru.ru_utime.tv_usec*1e3;
 423         sum += ru.ru_stime.tv_sec*1e9 + ru.ru_stime.tv_usec*1e3;
 424
 425         return sum;
 426 }
 427
 428 static int self_open_counters(void)
 429 {
 430         struct perf_event_attr attr;
 431         int fd;
 432
 433         memset(&attr, 0, sizeof(attr));
 434
 435         attr.type = PERF_TYPE_SOFTWARE;
 436         attr.config = PERF_COUNT_SW_TASK_CLOCK;
 437
 438         fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
 439
 440         if (fd < 0)
 441                 pr_err("Error: sys_perf_event_open() syscall returned "
 442                        "with %d (%s)\n", fd, strerror(errno));
 443         return fd;
 444 }
 445
 446 static u64 get_cpu_usage_nsec_self(int fd)
 447 {
 448         u64 runtime;
 449         int ret;
 450
 451         ret = read(fd, &runtime, sizeof(runtime));
 452         BUG_ON(ret != sizeof(runtime));
 453
 454         return runtime;
 455 }
 456
 457 struct sched_thread_parms {
 458         struct task_desc  *task;
 459         struct perf_sched *sched;
 460 };
 461
 462 static void *thread_func(void *ctx)
 463 {
 464         struct sched_thread_parms *parms = ctx;
 465         struct task_desc *this_task = parms->task;
 466         struct perf_sched *sched = parms->sched;
 467         u64 cpu_usage_0, cpu_usage_1;
 468         unsigned long i, ret;
 469         char comm2[22];
 470         int fd;
 471
 472         free(parms);
 473
 474         sprintf(comm2, ":%s", this_task->comm);
 475         prctl(PR_SET_NAME, comm2);
 476         fd = self_open_counters();
 477         if (fd < 0)
 478                 return NULL;
 479 again:
 480         ret = sem_post(&this_task->ready_for_work);
 481         BUG_ON(ret);
 482         ret = pthread_mutex_lock(&sched->start_work_mutex);
 483         BUG_ON(ret);
 484         ret = pthread_mutex_unlock(&sched->start_work_mutex);
 485         BUG_ON(ret);
 486
 487         cpu_usage_0 = get_cpu_usage_nsec_self(fd);
 488
 489         for (i = 0; i < this_task->nr_events; i++) {
 490                 this_task->curr_event = i;
 491                 perf_sched__process_event(sched, this_task->atoms[i]);
 492         }
 493
 494         cpu_usage_1 = get_cpu_usage_nsec_self(fd);
 495         this_task->cpu_usage = cpu_usage_1 - cpu_usage_0;
 496         ret = sem_post(&this_task->work_done_sem);
 497         BUG_ON(ret);
 498
 499         ret = pthread_mutex_lock(&sched->work_done_wait_mutex);
 500         BUG_ON(ret);
 501         ret = pthread_mutex_unlock(&sched->work_done_wait_mutex);
 502         BUG_ON(ret);
 503
 504         goto again;
 505 }
 506
 507 static void create_tasks(struct perf_sched *sched)
 508 {
 509         struct task_desc *task;
 510         pthread_attr_t attr;
 511         unsigned long i;
 512         int err;
 513
 514         err = pthread_attr_init(&attr);
 515         BUG_ON(err);
 516         err = pthread_attr_setstacksize(&attr,
 517                         (size_t) max(16 * 1024, PTHREAD_STACK_MIN));
 518         BUG_ON(err);
 519         err = pthread_mutex_lock(&sched->start_work_mutex);
 520         BUG_ON(err);
 521         err = pthread_mutex_lock(&sched->work_done_wait_mutex);
 522         BUG_ON(err);
 523         for (i = 0; i < sched->nr_tasks; i++) {
 524                 struct sched_thread_parms *parms = malloc(sizeof(*parms));
 525                 BUG_ON(parms == NULL);
 526                 parms->task = task = sched->tasks[i];
 527                 parms->sched = sched;
 528                 sem_init(&task->sleep_sem, 0, 0);
 529                 sem_init(&task->ready_for_work, 0, 0);
 530                 sem_init(&task->work_done_sem, 0, 0);
 531                 task->curr_event = 0;
 532                 err = pthread_create(&task->thread, &attr, thread_func, parms);
 533                 BUG_ON(err);
 534         }
 535 }
 536
 537 static void wait_for_tasks(struct perf_sched *sched)
 538 {
 539         u64 cpu_usage_0, cpu_usage_1;
 540         struct task_desc *task;
 541         unsigned long i, ret;
 542
 543         sched->start_time = get_nsecs();
 544         sched->cpu_usage = 0;
 545         pthread_mutex_unlock(&sched->work_done_wait_mutex);
 546
 547         for (i = 0; i < sched->nr_tasks; i++) {
 548                 task = sched->tasks[i];
 549                 ret = sem_wait(&task->ready_for_work);
 550                 BUG_ON(ret);
 551                 sem_init(&task->ready_for_work, 0, 0);
 552         }
 553         ret = pthread_mutex_lock(&sched->work_done_wait_mutex);
 554         BUG_ON(ret);
 555
 556         cpu_usage_0 = get_cpu_usage_nsec_parent();
 557
 558         pthread_mutex_unlock(&sched->start_work_mutex);
 559
 560         for (i = 0; i < sched->nr_tasks; i++) {
 561                 task = sched->tasks[i];
 562                 ret = sem_wait(&task->work_done_sem);
 563                 BUG_ON(ret);
 564                 sem_init(&task->work_done_sem, 0, 0);
 565                 sched->cpu_usage += task->cpu_usage;
 566                 task->cpu_usage = 0;
 567         }
 568
 569         cpu_usage_1 = get_cpu_usage_nsec_parent();
 570         if (!sched->runavg_cpu_usage)
 571                 sched->runavg_cpu_usage = sched->cpu_usage;
 572         sched->runavg_cpu_usage = (sched->runavg_cpu_usage * 9 + sched->cpu_usage) / 10;
 573
 574         sched->parent_cpu_usage = cpu_usage_1 - cpu_usage_0;
 575         if (!sched->runavg_parent_cpu_usage)
 576                 sched->runavg_parent_cpu_usage = sched->parent_cpu_usage;
 577         sched->runavg_parent_cpu_usage = (sched->runavg_parent_cpu_usage * 9 +
 578                                          sched->parent_cpu_usage)/10;
 579
 580         ret = pthread_mutex_lock(&sched->start_work_mutex);
 581         BUG_ON(ret);
 582
 583         for (i = 0; i < sched->nr_tasks; i++) {
 584                 task = sched->tasks[i];
 585                 sem_init(&task->sleep_sem, 0, 0);
 586                 task->curr_event = 0;
 587         }
 588 }
 589
 590 static void run_one_test(struct perf_sched *sched)
 591 {
 592         u64 T0, T1, delta, avg_delta, fluct;
 593
 594         T0 = get_nsecs();
 595         wait_for_tasks(sched);
 596         T1 = get_nsecs();
 597
 598         delta = T1 - T0;
 599         sched->sum_runtime += delta;
 600         sched->nr_runs++;
 601
 602         avg_delta = sched->sum_runtime / sched->nr_runs;
 603         if (delta < avg_delta)
 604                 fluct = avg_delta - delta;
 605         else
 606                 fluct = delta - avg_delta;
 607         sched->sum_fluct += fluct;
 608         if (!sched->run_avg)
 609                 sched->run_avg = delta;
 610         sched->run_avg = (sched->run_avg * 9 + delta) / 10;
 611
 612         printf("#%-3ld: %0.3f, ", sched->nr_runs, (double)delta / 1000000.0);
 613
 614         printf("ravg: %0.2f, ", (double)sched->run_avg / 1e6);
 615
 616         printf("cpu: %0.2f / %0.2f",
 617                 (double)sched->cpu_usage / 1e6, (double)sched->runavg_cpu_usage / 1e6);
 618
 619 #if 0
 620         /*
 621          * rusage statistics done by the parent, these are less
 622          * accurate than the sched->sum_exec_runtime based statistics:
 623          */
 624         printf(" [%0.2f / %0.2f]",
 625                 (double)sched->parent_cpu_usage/1e6,
 626                 (double)sched->runavg_parent_cpu_usage/1e6);
 627 #endif
 628
 629         printf("\n");
 630
 631         if (sched->nr_sleep_corrections)
 632                 printf(" (%ld sleep corrections)\n", sched->nr_sleep_corrections);
 633         sched->nr_sleep_corrections = 0;
 634 }
 635
 636 static void test_calibrations(struct perf_sched *sched)
 637 {
 638         u64 T0, T1;
 639
 640         T0 = get_nsecs();
 641         burn_nsecs(sched, 1e6);
 642         T1 = get_nsecs();
 643
 644         printf("the run test took %" PRIu64 " nsecs\n", T1 - T0);
 645
 646         T0 = get_nsecs();
 647         sleep_nsecs(1e6);
 648         T1 = get_nsecs();
 649
 650         printf("the sleep test took %" PRIu64 " nsecs\n", T1 - T0);
 651 }
 652
 653 static int
 654 replay_wakeup_event(struct perf_sched *sched,
 655                     struct perf_evsel *evsel, struct perf_sample *sample,
 656                     struct machine *machine __maybe_unused)
 657 {
 658         const char *comm = perf_evsel__strval(evsel, sample, "comm");
 659         const u32 pid    = perf_evsel__intval(evsel, sample, "pid");
 660         struct task_desc *waker, *wakee;
 661
 662         if (verbose) {
 663                 printf("sched_wakeup event %p\n", evsel);
 664
 665                 printf(" ... pid %d woke up %s/%d\n", sample->tid, comm, pid);
 666         }
 667
 668         waker = register_pid(sched, sample->tid, "<unknown>");
 669         wakee = register_pid(sched, pid, comm);
 670
 671         add_sched_event_wakeup(sched, waker, sample->time, wakee);
 672         return 0;
 673 }
 674
 675 static int replay_switch_event(struct perf_sched *sched,
 676                                struct perf_evsel *evsel,
 677                                struct perf_sample *sample,
 678                                struct machine *machine __maybe_unused)
 679 {
 680         const char *prev_comm  = perf_evsel__strval(evsel, sample, "prev_comm"),
 681                    *next_comm  = perf_evsel__strval(evsel, sample, "next_comm");
 682         const u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
 683                   next_pid = perf_evsel__intval(evsel, sample, "next_pid");
 684         const u64 prev_state = perf_evsel__intval(evsel, sample, "prev_state");
 685         struct task_desc *prev, __maybe_unused *next;
 686         u64 timestamp0, timestamp = sample->time;
 687         int cpu = sample->cpu;
 688         s64 delta;
 689
 690         if (verbose)
 691                 printf("sched_switch event %p\n", evsel);
 692
 693         if (cpu >= MAX_CPUS || cpu < 0)
 694                 return 0;
 695
 696         timestamp0 = sched->cpu_last_switched[cpu];
 697         if (timestamp0)
 698                 delta = timestamp - timestamp0;
 699         else
 700                 delta = 0;
 701
 702         if (delta < 0) {
 703                 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 704                 return -1;
 705         }
 706
 707         pr_debug(" ... switch from %s/%d to %s/%d [ran %" PRIu64 " nsecs]\n",
 708                  prev_comm, prev_pid, next_comm, next_pid, delta);
 709
 710         prev = register_pid(sched, prev_pid, prev_comm);
 711         next = register_pid(sched, next_pid, next_comm);
 712
 713         sched->cpu_last_switched[cpu] = timestamp;
 714
 715         add_sched_event_run(sched, prev, timestamp, delta);
 716         add_sched_event_sleep(sched, prev, timestamp, prev_state);
 717
 718         return 0;
 719 }
 720
 721 static int replay_fork_event(struct perf_sched *sched, struct perf_evsel *evsel,
 722                              struct perf_sample *sample)
 723 {
 724         const char *parent_comm = perf_evsel__strval(evsel, sample, "parent_comm"),
 725                    *child_comm  = perf_evsel__strval(evsel, sample, "child_comm");
 726         const u32 parent_pid  = perf_evsel__intval(evsel, sample, "parent_pid"),
 727                   child_pid  = perf_evsel__intval(evsel, sample, "child_pid");
 728
 729         if (verbose) {
 730                 printf("sched_fork event %p\n", evsel);
 731                 printf("... parent: %s/%d\n", parent_comm, parent_pid);
 732                 printf("...  child: %s/%d\n", child_comm, child_pid);
 733         }
 734
 735         register_pid(sched, parent_pid, parent_comm);
 736         register_pid(sched, child_pid, child_comm);
 737         return 0;
 738 }
 739
 740 struct sort_dimension {
 741         const char              *name;
 742         sort_fn_t               cmp;
 743         struct list_head        list;
 744 };
 745
 746 static int
 747 thread_lat_cmp(struct list_head *list, struct work_atoms *l, struct work_atoms *r)
 748 {
 749         struct sort_dimension *sort;
 750         int ret = 0;
 751
 752         BUG_ON(list_empty(list));
 753
 754         list_for_each_entry(sort, list, list) {
 755                 ret = sort->cmp(l, r);
 756                 if (ret)
 757                         return ret;
 758         }
 759
 760         return ret;
 761 }
 762
 763 static struct work_atoms *
 764 thread_atoms_search(struct rb_root *root, struct thread *thread,
 765                          struct list_head *sort_list)
 766 {
 767         struct rb_node *node = root->rb_node;
 768         struct work_atoms key = { .thread = thread };
 769
 770         while (node) {
 771                 struct work_atoms *atoms;
 772                 int cmp;
 773
 774                 atoms = container_of(node, struct work_atoms, node);
 775
 776                 cmp = thread_lat_cmp(sort_list, &key, atoms);
 777                 if (cmp > 0)
 778                         node = node->rb_left;
 779                 else if (cmp < 0)
 780                         node = node->rb_right;
 781                 else {
 782                         BUG_ON(thread != atoms->thread);
 783                         return atoms;
 784                 }
 785         }
 786         return NULL;
 787 }
 788
 789 static void
 790 __thread_latency_insert(struct rb_root *root, struct work_atoms *data,
 791                          struct list_head *sort_list)
 792 {
 793         struct rb_node **new = &(root->rb_node), *parent = NULL;
 794
 795         while (*new) {
 796                 struct work_atoms *this;
 797                 int cmp;
 798
 799                 this = container_of(*new, struct work_atoms, node);
 800                 parent = *new;
 801
 802                 cmp = thread_lat_cmp(sort_list, data, this);
 803
 804                 if (cmp > 0)
 805                         new = &((*new)->rb_left);
 806                 else
 807                         new = &((*new)->rb_right);
 808         }
 809
 810         rb_link_node(&data->node, parent, new);
 811         rb_insert_color(&data->node, root);
 812 }
 813
 814 static int thread_atoms_insert(struct perf_sched *sched, struct thread *thread)
 815 {
 816         struct work_atoms *atoms = zalloc(sizeof(*atoms));
 817         if (!atoms) {
 818                 pr_err("No memory at %s\n", __func__);
 819                 return -1;
 820         }
 821
 822         atoms->thread = thread;
 823         INIT_LIST_HEAD(&atoms->work_list);
 824         __thread_latency_insert(&sched->atom_root, atoms, &sched->cmp_pid);
 825         return 0;
 826 }
 827
 828 static int latency_fork_event(struct perf_sched *sched __maybe_unused,
 829                               struct perf_evsel *evsel __maybe_unused,
 830                               struct perf_sample *sample __maybe_unused)
 831 {
 832         /* should insert the newcomer */
 833         return 0;
 834 }
 835
 836 static char sched_out_state(u64 prev_state)
 837 {
 838         const char *str = TASK_STATE_TO_CHAR_STR;
 839
 840         return str[prev_state];
 841 }
 842
 843 static int
 844 add_sched_out_event(struct work_atoms *atoms,
 845                     char run_state,
 846                     u64 timestamp)
 847 {
 848         struct work_atom *atom = zalloc(sizeof(*atom));
 849         if (!atom) {
 850                 pr_err("Non memory at %s", __func__);
 851                 return -1;
 852         }
 853
 854         atom->sched_out_time = timestamp;
 855
 856         if (run_state == 'R') {
 857                 atom->state = THREAD_WAIT_CPU;
 858                 atom->wake_up_time = atom->sched_out_time;
 859         }
 860
 861         list_add_tail(&atom->list, &atoms->work_list);
 862         return 0;
 863 }
 864
 865 static void
 866 add_runtime_event(struct work_atoms *atoms, u64 delta,
 867                   u64 timestamp __maybe_unused)
 868 {
 869         struct work_atom *atom;
 870
 871         BUG_ON(list_empty(&atoms->work_list));
 872
 873         atom = list_entry(atoms->work_list.prev, struct work_atom, list);
 874
 875         atom->runtime += delta;
 876         atoms->total_runtime += delta;
 877 }
 878
 879 static void
 880 add_sched_in_event(struct work_atoms *atoms, u64 timestamp)
 881 {
 882         struct work_atom *atom;
 883         u64 delta;
 884
 885         if (list_empty(&atoms->work_list))
 886                 return;
 887
 888         atom = list_entry(atoms->work_list.prev, struct work_atom, list);
 889
 890         if (atom->state != THREAD_WAIT_CPU)
 891                 return;
 892
 893         if (timestamp < atom->wake_up_time) {
 894                 atom->state = THREAD_IGNORE;
 895                 return;
 896         }
 897
 898         atom->state = THREAD_SCHED_IN;
 899         atom->sched_in_time = timestamp;
 900
 901         delta = atom->sched_in_time - atom->wake_up_time;
 902         atoms->total_lat += delta;
 903         if (delta > atoms->max_lat) {
 904                 atoms->max_lat = delta;
 905                 atoms->max_lat_at = timestamp;
 906         }
 907         atoms->nb_atoms++;
 908 }
 909
 910 static int latency_switch_event(struct perf_sched *sched,
 911                                 struct perf_evsel *evsel,
 912                                 struct perf_sample *sample,
 913                                 struct machine *machine)
 914 {
 915         const u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
 916                   next_pid = perf_evsel__intval(evsel, sample, "next_pid");
 917         const u64 prev_state = perf_evsel__intval(evsel, sample, "prev_state");
 918         struct work_atoms *out_events, *in_events;
 919         struct thread *sched_out, *sched_in;
 920         u64 timestamp0, timestamp = sample->time;
 921         int cpu = sample->cpu;
 922         s64 delta;
 923
 924         BUG_ON(cpu >= MAX_CPUS || cpu < 0);
 925
 926         timestamp0 = sched->cpu_last_switched[cpu];
 927         sched->cpu_last_switched[cpu] = timestamp;
 928         if (timestamp0)
 929                 delta = timestamp - timestamp0;
 930         else
 931                 delta = 0;
 932
 933         if (delta < 0) {
 934                 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 935                 return -1;
 936         }
 937
 938         sched_out = machine__findnew_thread(machine, prev_pid);
 939         sched_in = machine__findnew_thread(machine, next_pid);
 940
 941         out_events = thread_atoms_search(&sched->atom_root, sched_out, &sched->cmp_pid);
 942         if (!out_events) {
 943                 if (thread_atoms_insert(sched, sched_out))
 944                         return -1;
 945                 out_events = thread_atoms_search(&sched->atom_root, sched_out, &sched->cmp_pid);
 946                 if (!out_events) {
 947                         pr_err("out-event: Internal tree error");
 948                         return -1;
 949                 }
 950         }
 951         if (add_sched_out_event(out_events, sched_out_state(prev_state), timestamp))
 952                 return -1;
 953
 954         in_events = thread_atoms_search(&sched->atom_root, sched_in, &sched->cmp_pid);
 955         if (!in_events) {
 956                 if (thread_atoms_insert(sched, sched_in))
 957                         return -1;
 958                 in_events = thread_atoms_search(&sched->atom_root, sched_in, &sched->cmp_pid);
 959                 if (!in_events) {
 960                         pr_err("in-event: Internal tree error");
 961                         return -1;
 962                 }
 963                 /*
 964                  * Take came in we have not heard about yet,
 965                  * add in an initial atom in runnable state:
 966                  */
 967                 if (add_sched_out_event(in_events, 'R', timestamp))
 968                         return -1;
 969         }
 970         add_sched_in_event(in_events, timestamp);
 971
 972         return 0;
 973 }
 974
 975 static int latency_runtime_event(struct perf_sched *sched,
 976                                  struct perf_evsel *evsel,
 977                                  struct perf_sample *sample,
 978                                  struct machine *machine)
 979 {
 980         const u32 pid      = perf_evsel__intval(evsel, sample, "pid");
 981         const u64 runtime  = perf_evsel__intval(evsel, sample, "runtime");
 982         struct thread *thread = machine__findnew_thread(machine, pid);
 983         struct work_atoms *atoms = thread_atoms_search(&sched->atom_root, thread, &sched->cmp_pid);
 984         u64 timestamp = sample->time;
 985         int cpu = sample->cpu;
 986
 987         BUG_ON(cpu >= MAX_CPUS || cpu < 0);
 988         if (!atoms) {
 989                 if (thread_atoms_insert(sched, thread))
 990                         return -1;
 991                 atoms = thread_atoms_search(&sched->atom_root, thread, &sched->cmp_pid);
 992                 if (!atoms) {
 993                         pr_err("in-event: Internal tree error");
 994                         return -1;
 995                 }
 996                 if (add_sched_out_event(atoms, 'R', timestamp))
 997                         return -1;
 998         }
 999
1000         add_runtime_event(atoms, runtime, timestamp);
1001         return 0;
1002 }
1003
1004 static int latency_wakeup_event(struct perf_sched *sched,
1005                                 struct perf_evsel *evsel,
1006                                 struct perf_sample *sample,
1007                                 struct machine *machine)
1008 {
1009         const u32 pid     = perf_evsel__intval(evsel, sample, "pid"),
1010                   success = perf_evsel__intval(evsel, sample, "success");
1011         struct work_atoms *atoms;
1012         struct work_atom *atom;
1013         struct thread *wakee;
1014         u64 timestamp = sample->time;
1015
1016         /* Note for later, it may be interesting to observe the failing cases */
1017         if (!success)
1018                 return 0;
1019
1020         wakee = machine__findnew_thread(machine, pid);
1021         atoms = thread_atoms_search(&sched->atom_root, wakee, &sched->cmp_pid);
1022         if (!atoms) {
1023                 if (thread_atoms_insert(sched, wakee))
1024                         return -1;
1025                 atoms = thread_atoms_search(&sched->atom_root, wakee, &sched->cmp_pid);
1026                 if (!atoms) {
1027                         pr_err("wakeup-event: Internal tree error");
1028                         return -1;
1029                 }
1030                 if (add_sched_out_event(atoms, 'S', timestamp))
1031                         return -1;
1032         }
1033
1034         BUG_ON(list_empty(&atoms->work_list));
1035
1036         atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1037
1038         /*
1039          * You WILL be missing events if you've recorded only
1040          * one CPU, or are only looking at only one, so don't
1041          * make useless noise.
1042          */
1043         if (sched->profile_cpu == -1 && atom->state != THREAD_SLEEPING)
1044                 sched->nr_state_machine_bugs++;
1045
1046         sched->nr_timestamps++;
1047         if (atom->sched_out_time > timestamp) {
1048                 sched->nr_unordered_timestamps++;
1049                 return 0;
1050         }
1051
1052         atom->state = THREAD_WAIT_CPU;
1053         atom->wake_up_time = timestamp;
1054         return 0;
1055 }
1056
1057 static int latency_migrate_task_event(struct perf_sched *sched,
1058                                       struct perf_evsel *evsel,
1059                                       struct perf_sample *sample,
1060                                       struct machine *machine)
1061 {
1062         const u32 pid = perf_evsel__intval(evsel, sample, "pid");
1063         u64 timestamp = sample->time;
1064         struct work_atoms *atoms;
1065         struct work_atom *atom;
1066         struct thread *migrant;
1067
1068         /*
1069          * Only need to worry about migration when profiling one CPU.
1070          */
1071         if (sched->profile_cpu == -1)
1072                 return 0;
1073
1074         migrant = machine__findnew_thread(machine, pid);
1075         atoms = thread_atoms_search(&sched->atom_root, migrant, &sched->cmp_pid);
1076         if (!atoms) {
1077                 if (thread_atoms_insert(sched, migrant))
1078                         return -1;
1079                 register_pid(sched, migrant->pid, migrant->comm);
1080                 atoms = thread_atoms_search(&sched->atom_root, migrant, &sched->cmp_pid);
1081                 if (!atoms) {
1082                         pr_err("migration-event: Internal tree error");
1083                         return -1;
1084                 }
1085                 if (add_sched_out_event(atoms, 'R', timestamp))
1086                         return -1;
1087         }
1088
1089         BUG_ON(list_empty(&atoms->work_list));
1090
1091         atom = list_entry(atoms->work_list.prev, struct work_atom, list);
1092         atom->sched_in_time = atom->sched_out_time = atom->wake_up_time = timestamp;
1093
1094         sched->nr_timestamps++;
1095
1096         if (atom->sched_out_time > timestamp)
1097                 sched->nr_unordered_timestamps++;
1098
1099         return 0;
1100 }
1101
1102 static void output_lat_thread(struct perf_sched *sched, struct work_atoms *work_list)
1103 {
1104         int i;
1105         int ret;
1106         u64 avg;
1107
1108         if (!work_list->nb_atoms)
1109                 return;
1110         /*
1111          * Ignore idle threads:
1112          */
1113         if (!strcmp(work_list->thread->comm, "swapper"))
1114                 return;
1115
1116         sched->all_runtime += work_list->total_runtime;
1117         sched->all_count   += work_list->nb_atoms;
1118
1119         ret = printf("  %s:%d ", work_list->thread->comm, work_list->thread->pid);
1120
1121         for (i = 0; i < 24 - ret; i++)
1122                 printf(" ");
1123
1124         avg = work_list->total_lat / work_list->nb_atoms;
1125
1126         printf("|%11.3f ms |%9" PRIu64 " | avg:%9.3f ms | max:%9.3f ms | max at: %9.6f s\n",
1127               (double)work_list->total_runtime / 1e6,
1128                  work_list->nb_atoms, (double)avg / 1e6,
1129                  (double)work_list->max_lat / 1e6,
1130                  (double)work_list->max_lat_at / 1e9);
1131 }
1132
1133 static int pid_cmp(struct work_atoms *l, struct work_atoms *r)
1134 {
1135         if (l->thread->pid < r->thread->pid)
1136                 return -1;
1137         if (l->thread->pid > r->thread->pid)
1138                 return 1;
1139
1140         return 0;
1141 }
1142
1143 static int avg_cmp(struct work_atoms *l, struct work_atoms *r)
1144 {
1145         u64 avgl, avgr;
1146
1147         if (!l->nb_atoms)
1148                 return -1;
1149
1150         if (!r->nb_atoms)
1151                 return 1;
1152
1153         avgl = l->total_lat / l->nb_atoms;
1154         avgr = r->total_lat / r->nb_atoms;
1155
1156         if (avgl < avgr)
1157                 return -1;
1158         if (avgl > avgr)
1159                 return 1;
1160
1161         return 0;
1162 }
1163
1164 static int max_cmp(struct work_atoms *l, struct work_atoms *r)
1165 {
1166         if (l->max_lat < r->max_lat)
1167                 return -1;
1168         if (l->max_lat > r->max_lat)
1169                 return 1;
1170
1171         return 0;
1172 }
1173
1174 static int switch_cmp(struct work_atoms *l, struct work_atoms *r)
1175 {
1176         if (l->nb_atoms < r->nb_atoms)
1177                 return -1;
1178         if (l->nb_atoms > r->nb_atoms)
1179                 return 1;
1180
1181         return 0;
1182 }
1183
1184 static int runtime_cmp(struct work_atoms *l, struct work_atoms *r)
1185 {
1186         if (l->total_runtime < r->total_runtime)
1187                 return -1;
1188         if (l->total_runtime > r->total_runtime)
1189                 return 1;
1190
1191         return 0;
1192 }
1193
1194 static int sort_dimension__add(const char *tok, struct list_head *list)
1195 {
1196         size_t i;
1197         static struct sort_dimension avg_sort_dimension = {
1198                 .name = "avg",
1199                 .cmp  = avg_cmp,
1200         };
1201         static struct sort_dimension max_sort_dimension = {
1202                 .name = "max",
1203                 .cmp  = max_cmp,
1204         };
1205         static struct sort_dimension pid_sort_dimension = {
1206                 .name = "pid",
1207                 .cmp  = pid_cmp,
1208         };
1209         static struct sort_dimension runtime_sort_dimension = {
1210                 .name = "runtime",
1211                 .cmp  = runtime_cmp,
1212         };
1213         static struct sort_dimension switch_sort_dimension = {
1214                 .name = "switch",
1215                 .cmp  = switch_cmp,
1216         };
1217         struct sort_dimension *available_sorts[] = {
1218                 &pid_sort_dimension,
1219                 &avg_sort_dimension,
1220                 &max_sort_dimension,
1221                 &switch_sort_dimension,
1222                 &runtime_sort_dimension,
1223         };
1224
1225         for (i = 0; i < ARRAY_SIZE(available_sorts); i++) {
1226                 if (!strcmp(available_sorts[i]->name, tok)) {
1227                         list_add_tail(&available_sorts[i]->list, list);
1228
1229                         return 0;
1230                 }
1231         }
1232
1233         return -1;
1234 }
1235
1236 static void perf_sched__sort_lat(struct perf_sched *sched)
1237 {
1238         struct rb_node *node;
1239
1240         for (;;) {
1241                 struct work_atoms *data;
1242                 node = rb_first(&sched->atom_root);
1243                 if (!node)
1244                         break;
1245
1246                 rb_erase(node, &sched->atom_root);
1247                 data = rb_entry(node, struct work_atoms, node);
1248                 __thread_latency_insert(&sched->sorted_atom_root, data, &sched->sort_list);
1249         }
1250 }
1251
1252 static int process_sched_wakeup_event(struct perf_tool *tool,
1253                                       struct perf_evsel *evsel,
1254                                       struct perf_sample *sample,
1255                                       struct machine *machine)
1256 {
1257         struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1258
1259         if (sched->tp_handler->wakeup_event)
1260                 return sched->tp_handler->wakeup_event(sched, evsel, sample, machine);
1261
1262         return 0;
1263 }
1264
1265 static int map_switch_event(struct perf_sched *sched, struct perf_evsel *evsel,
1266                             struct perf_sample *sample, struct machine *machine)
1267 {
1268         const u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
1269                   next_pid = perf_evsel__intval(evsel, sample, "next_pid");
1270         struct thread *sched_out __maybe_unused, *sched_in;
1271         int new_shortname;
1272         u64 timestamp0, timestamp = sample->time;
1273         s64 delta;
1274         int cpu, this_cpu = sample->cpu;
1275
1276         BUG_ON(this_cpu >= MAX_CPUS || this_cpu < 0);
1277
1278         if (this_cpu > sched->max_cpu)
1279                 sched->max_cpu = this_cpu;
1280
1281         timestamp0 = sched->cpu_last_switched[this_cpu];
1282         sched->cpu_last_switched[this_cpu] = timestamp;
1283         if (timestamp0)
1284                 delta = timestamp - timestamp0;
1285         else
1286                 delta = 0;
1287
1288         if (delta < 0) {
1289                 pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
1290                 return -1;
1291         }
1292
1293         sched_out = machine__findnew_thread(machine, prev_pid);
1294         sched_in = machine__findnew_thread(machine, next_pid);
1295
1296         sched->curr_thread[this_cpu] = sched_in;
1297
1298         printf("  ");
1299
1300         new_shortname = 0;
1301         if (!sched_in->shortname[0]) {
1302                 sched_in->shortname[0] = sched->next_shortname1;
1303                 sched_in->shortname[1] = sched->next_shortname2;
1304
1305                 if (sched->next_shortname1 < 'Z') {
1306                         sched->next_shortname1++;
1307                 } else {
1308                         sched->next_shortname1='A';
1309                         if (sched->next_shortname2 < '9') {
1310                                 sched->next_shortname2++;
1311                         } else {
1312                                 sched->next_shortname2='0';
1313                         }
1314                 }
1315                 new_shortname = 1;
1316         }
1317
1318         for (cpu = 0; cpu <= sched->max_cpu; cpu++) {
1319                 if (cpu != this_cpu)
1320                         printf(" ");
1321                 else
1322                         printf("*");
1323
1324                 if (sched->curr_thread[cpu]) {
1325                         if (sched->curr_thread[cpu]->pid)
1326                                 printf("%2s ", sched->curr_thread[cpu]->shortname);
1327                         else
1328                                 printf(".  ");
1329                 } else
1330                         printf("   ");
1331         }
1332
1333         printf("  %12.6f secs ", (double)timestamp/1e9);
1334         if (new_shortname) {
1335                 printf("%s => %s:%d\n",
1336                         sched_in->shortname, sched_in->comm, sched_in->pid);
1337         } else {
1338                 printf("\n");
1339         }
1340
1341         return 0;
1342 }
1343
1344 static int process_sched_switch_event(struct perf_tool *tool,
1345                                       struct perf_evsel *evsel,
1346                                       struct perf_sample *sample,
1347                                       struct machine *machine)
1348 {
1349         struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1350         int this_cpu = sample->cpu, err = 0;
1351         u32 prev_pid = perf_evsel__intval(evsel, sample, "prev_pid"),
1352             next_pid = perf_evsel__intval(evsel, sample, "next_pid");
1353
1354         if (sched->curr_pid[this_cpu] != (u32)-1) {
1355                 /*
1356                  * Are we trying to switch away a PID that is
1357                  * not current?
1358                  */
1359                 if (sched->curr_pid[this_cpu] != prev_pid)
1360                         sched->nr_context_switch_bugs++;
1361         }
1362
1363         if (sched->tp_handler->switch_event)
1364                 err = sched->tp_handler->switch_event(sched, evsel, sample, machine);
1365
1366         sched->curr_pid[this_cpu] = next_pid;
1367         return err;
1368 }
1369
1370 static int process_sched_runtime_event(struct perf_tool *tool,
1371                                        struct perf_evsel *evsel,
1372                                        struct perf_sample *sample,
1373                                        struct machine *machine)
1374 {
1375         struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1376
1377         if (sched->tp_handler->runtime_event)
1378                 return sched->tp_handler->runtime_event(sched, evsel, sample, machine);
1379
1380         return 0;
1381 }
1382
1383 static int process_sched_fork_event(struct perf_tool *tool,
1384                                     struct perf_evsel *evsel,
1385                                     struct perf_sample *sample,
1386                                     struct machine *machine __maybe_unused)
1387 {
1388         struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1389
1390         if (sched->tp_handler->fork_event)
1391                 return sched->tp_handler->fork_event(sched, evsel, sample);
1392
1393         return 0;
1394 }
1395
1396 static int process_sched_exit_event(struct perf_tool *tool __maybe_unused,
1397                                     struct perf_evsel *evsel,
1398                                     struct perf_sample *sample __maybe_unused,
1399                                     struct machine *machine __maybe_unused)
1400 {
1401         pr_debug("sched_exit event %p\n", evsel);
1402         return 0;
1403 }
1404
1405 static int process_sched_migrate_task_event(struct perf_tool *tool,
1406                                             struct perf_evsel *evsel,
1407                                             struct perf_sample *sample,
1408                                             struct machine *machine)
1409 {
1410         struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
1411
1412         if (sched->tp_handler->migrate_task_event)
1413                 return sched->tp_handler->migrate_task_event(sched, evsel, sample, machine);
1414
1415         return 0;
1416 }
1417
1418 typedef int (*tracepoint_handler)(struct perf_tool *tool,
1419                                   struct perf_evsel *evsel,
1420                                   struct perf_sample *sample,
1421                                   struct machine *machine);
1422
1423 static int perf_sched__process_tracepoint_sample(struct perf_tool *tool __maybe_unused,
1424                                                  union perf_event *event __maybe_unused,
1425                                                  struct perf_sample *sample,
1426                                                  struct perf_evsel *evsel,
1427                                                  struct machine *machine)
1428 {
1429         struct thread *thread = machine__findnew_thread(machine, sample->tid);
1430         int err = 0;
1431
1432         if (thread == NULL) {
1433                 pr_debug("problem processing %s event, skipping it.\n",
1434                          perf_evsel__name(evsel));
1435                 return -1;
1436         }
1437
1438         evsel->hists.stats.total_period += sample->period;
1439         hists__inc_nr_events(&evsel->hists, PERF_RECORD_SAMPLE);
1440
1441         if (evsel->handler.func != NULL) {
1442                 tracepoint_handler f = evsel->handler.func;
1443                 err = f(tool, evsel, sample, machine);
1444         }
1445
1446         return err;
1447 }
1448
1449 static int perf_sched__read_events(struct perf_sched *sched, bool destroy,
1450                                    struct perf_session **psession)
1451 {
1452         const struct perf_evsel_str_handler handlers[] = {
1453                 { "sched:sched_switch",       process_sched_switch_event, },
1454                 { "sched:sched_stat_runtime", process_sched_runtime_event, },
1455                 { "sched:sched_wakeup",       process_sched_wakeup_event, },
1456                 { "sched:sched_wakeup_new",   process_sched_wakeup_event, },
1457                 { "sched:sched_process_fork", process_sched_fork_event, },
1458                 { "sched:sched_process_exit", process_sched_exit_event, },
1459                 { "sched:sched_migrate_task", process_sched_migrate_task_event, },
1460         };
1461         struct perf_session *session;
1462
1463         session = perf_session__new(sched->input_name, O_RDONLY, 0, false, &sched->tool);
1464         if (session == NULL) {
1465                 pr_debug("No Memory for session\n");
1466                 return -1;
1467         }
1468
1469         if (perf_session__set_tracepoints_handlers(session, handlers))
1470                 goto out_delete;
1471
1472         if (perf_session__has_traces(session, "record -R")) {
1473                 int err = perf_session__process_events(session, &sched->tool);
1474                 if (err) {
1475                         pr_err("Failed to process events, error %d", err);
1476                         goto out_delete;
1477                 }
1478
1479                 sched->nr_events      = session->hists.stats.nr_events[0];
1480                 sched->nr_lost_events = session->hists.stats.total_lost;
1481                 sched->nr_lost_chunks = session->hists.stats.nr_events[PERF_RECORD_LOST];
1482         }
1483
1484         if (destroy)
1485                 perf_session__delete(session);
1486
1487         if (psession)
1488                 *psession = session;
1489
1490         return 0;
1491
1492 out_delete:
1493         perf_session__delete(session);
1494         return -1;
1495 }
1496
1497 static void print_bad_events(struct perf_sched *sched)
1498 {
1499         if (sched->nr_unordered_timestamps && sched->nr_timestamps) {
1500                 printf("  INFO: %.3f%% unordered timestamps (%ld out of %ld)\n",
1501                         (double)sched->nr_unordered_timestamps/(double)sched->nr_timestamps*100.0,
1502                         sched->nr_unordered_timestamps, sched->nr_timestamps);
1503         }
1504         if (sched->nr_lost_events && sched->nr_events) {
1505                 printf("  INFO: %.3f%% lost events (%ld out of %ld, in %ld chunks)\n",
1506                         (double)sched->nr_lost_events/(double)sched->nr_events * 100.0,
1507                         sched->nr_lost_events, sched->nr_events, sched->nr_lost_chunks);
1508         }
1509         if (sched->nr_state_machine_bugs && sched->nr_timestamps) {
1510                 printf("  INFO: %.3f%% state machine bugs (%ld out of %ld)",
1511                         (double)sched->nr_state_machine_bugs/(double)sched->nr_timestamps*100.0,
1512                         sched->nr_state_machine_bugs, sched->nr_timestamps);
1513                 if (sched->nr_lost_events)
1514                         printf(" (due to lost events?)");
1515                 printf("\n");
1516         }
1517         if (sched->nr_context_switch_bugs && sched->nr_timestamps) {
1518                 printf("  INFO: %.3f%% context switch bugs (%ld out of %ld)",
1519                         (double)sched->nr_context_switch_bugs/(double)sched->nr_timestamps*100.0,
1520                         sched->nr_context_switch_bugs, sched->nr_timestamps);
1521                 if (sched->nr_lost_events)
1522                         printf(" (due to lost events?)");
1523                 printf("\n");
1524         }
1525 }
1526
1527 static int perf_sched__lat(struct perf_sched *sched)
1528 {
1529         struct rb_node *next;
1530         struct perf_session *session;
1531
1532         setup_pager();
1533         if (perf_sched__read_events(sched, false, &session))
1534                 return -1;
1535         perf_sched__sort_lat(sched);
1536
1537         printf("\n ---------------------------------------------------------------------------------------------------------------\n");
1538         printf("  Task                  |   Runtime ms  | Switches | Average delay ms | Maximum delay ms | Maximum delay at     |\n");
1539         printf(" ---------------------------------------------------------------------------------------------------------------\n");
1540
1541         next = rb_first(&sched->sorted_atom_root);
1542
1543         while (next) {
1544                 struct work_atoms *work_list;
1545
1546                 work_list = rb_entry(next, struct work_atoms, node);
1547                 output_lat_thread(sched, work_list);
1548                 next = rb_next(next);
1549         }
1550
1551         printf(" -----------------------------------------------------------------------------------------\n");
1552         printf("  TOTAL:                |%11.3f ms |%9" PRIu64 " |\n",
1553                 (double)sched->all_runtime / 1e6, sched->all_count);
1554
1555         printf(" ---------------------------------------------------\n");
1556
1557         print_bad_events(sched);
1558         printf("\n");
1559
1560         perf_session__delete(session);
1561         return 0;
1562 }
1563
1564 static int perf_sched__map(struct perf_sched *sched)
1565 {
1566         sched->max_cpu = sysconf(_SC_NPROCESSORS_CONF);
1567
1568         setup_pager();
1569         if (perf_sched__read_events(sched, true, NULL))
1570                 return -1;
1571         print_bad_events(sched);
1572         return 0;
1573 }
1574
1575 static int perf_sched__replay(struct perf_sched *sched)
1576 {
1577         unsigned long i;
1578
1579         calibrate_run_measurement_overhead(sched);
1580         calibrate_sleep_measurement_overhead(sched);
1581
1582         test_calibrations(sched);
1583
1584         if (perf_sched__read_events(sched, true, NULL))
1585                 return -1;
1586
1587         printf("nr_run_events:        %ld\n", sched->nr_run_events);
1588         printf("nr_sleep_events:      %ld\n", sched->nr_sleep_events);
1589         printf("nr_wakeup_events:     %ld\n", sched->nr_wakeup_events);
1590
1591         if (sched->targetless_wakeups)
1592                 printf("target-less wakeups:  %ld\n", sched->targetless_wakeups);
1593         if (sched->multitarget_wakeups)
1594                 printf("multi-target wakeups: %ld\n", sched->multitarget_wakeups);
1595         if (sched->nr_run_events_optimized)
1596                 printf("run atoms optimized: %ld\n",
1597                         sched->nr_run_events_optimized);
1598
1599         print_task_traces(sched);
1600         add_cross_task_wakeups(sched);
1601
1602         create_tasks(sched);
1603         printf("------------------------------------------------------------\n");
1604         for (i = 0; i < sched->replay_repeat; i++)
1605                 run_one_test(sched);
1606
1607         return 0;
1608 }
1609
1610 static void setup_sorting(struct perf_sched *sched, const struct option *options,
1611                           const char * const usage_msg[])
1612 {
1613         char *tmp, *tok, *str = strdup(sched->sort_order);
1614
1615         for (tok = strtok_r(str, ", ", &tmp);
1616                         tok; tok = strtok_r(NULL, ", ", &tmp)) {
1617                 if (sort_dimension__add(tok, &sched->sort_list) < 0) {
1618                         error("Unknown --sort key: `%s'", tok);
1619                         usage_with_options(usage_msg, options);
1620                 }
1621         }
1622
1623         free(str);
1624
1625         sort_dimension__add("pid", &sched->cmp_pid);
1626 }
1627
1628 static int __cmd_record(int argc, const char **argv)
1629 {
1630         unsigned int rec_argc, i, j;
1631         const char **rec_argv;
1632         const char * const record_args[] = {
1633                 "record",
1634                 "-a",
1635                 "-R",
1636                 "-f",
1637                 "-m", "1024",
1638                 "-c", "1",
1639                 "-e", "sched:sched_switch",
1640                 "-e", "sched:sched_stat_wait",
1641                 "-e", "sched:sched_stat_sleep",
1642                 "-e", "sched:sched_stat_iowait",
1643                 "-e", "sched:sched_stat_runtime",
1644                 "-e", "sched:sched_process_exit",
1645                 "-e", "sched:sched_process_fork",
1646                 "-e", "sched:sched_wakeup",
1647                 "-e", "sched:sched_migrate_task",
1648         };
1649
1650         rec_argc = ARRAY_SIZE(record_args) + argc - 1;
1651         rec_argv = calloc(rec_argc + 1, sizeof(char *));
1652
1653         if (rec_argv == NULL)
1654                 return -ENOMEM;
1655
1656         for (i = 0; i < ARRAY_SIZE(record_args); i++)
1657                 rec_argv[i] = strdup(record_args[i]);
1658
1659         for (j = 1; j < (unsigned int)argc; j++, i++)
1660                 rec_argv[i] = argv[j];
1661
1662         BUG_ON(i != rec_argc);
1663
1664         return cmd_record(i, rec_argv, NULL);
1665 }
1666
1667 int cmd_sched(int argc, const char **argv, const char *prefix __maybe_unused)
1668 {
1669         const char default_sort_order[] = "avg, max, switch, runtime";
1670         struct perf_sched sched = {
1671                 .tool = {
1672                         .sample          = perf_sched__process_tracepoint_sample,
1673                         .comm            = perf_event__process_comm,
1674                         .lost            = perf_event__process_lost,
1675                         .exit            = perf_event__process_exit,
1676                         .fork            = perf_event__process_fork,
1677                         .ordered_samples = true,
1678                 },
1679                 .cmp_pid              = LIST_HEAD_INIT(sched.cmp_pid),
1680                 .sort_list            = LIST_HEAD_INIT(sched.sort_list),
1681                 .start_work_mutex     = PTHREAD_MUTEX_INITIALIZER,
1682                 .work_done_wait_mutex = PTHREAD_MUTEX_INITIALIZER,
1683                 .curr_pid             = { [0 ... MAX_CPUS - 1] = -1 },
1684                 .sort_order           = default_sort_order,
1685                 .replay_repeat        = 10,
1686                 .profile_cpu          = -1,
1687                 .next_shortname1      = 'A',
1688                 .next_shortname2      = '0',
1689         };
1690         const struct option latency_options[] = {
1691         OPT_STRING('s', "sort", &sched.sort_order, "key[,key2...]",
1692                    "sort by key(s): runtime, switch, avg, max"),
1693         OPT_INCR('v', "verbose", &verbose,
1694                     "be more verbose (show symbol address, etc)"),
1695         OPT_INTEGER('C', "CPU", &sched.profile_cpu,
1696                     "CPU to profile on"),
1697         OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
1698                     "dump raw trace in ASCII"),
1699         OPT_END()
1700         };
1701         const struct option replay_options[] = {
1702         OPT_UINTEGER('r', "repeat", &sched.replay_repeat,
1703                      "repeat the workload replay N times (-1: infinite)"),
1704         OPT_INCR('v', "verbose", &verbose,
1705                     "be more verbose (show symbol address, etc)"),
1706         OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
1707                     "dump raw trace in ASCII"),
1708         OPT_END()
1709         };
1710         const struct option sched_options[] = {
1711         OPT_STRING('i', "input", &sched.input_name, "file",
1712                     "input file name"),
1713         OPT_INCR('v', "verbose", &verbose,
1714                     "be more verbose (show symbol address, etc)"),
1715         OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
1716                     "dump raw trace in ASCII"),
1717         OPT_END()
1718         };
1719         const char * const latency_usage[] = {
1720                 "perf sched latency [<options>]",
1721                 NULL
1722         };
1723         const char * const replay_usage[] = {
1724                 "perf sched replay [<options>]",
1725                 NULL
1726         };
1727         const char * const sched_usage[] = {
1728                 "perf sched [<options>] {record|latency|map|replay|script}",
1729                 NULL
1730         };
1731         struct trace_sched_handler lat_ops  = {
1732                 .wakeup_event       = latency_wakeup_event,
1733                 .switch_event       = latency_switch_event,
1734                 .runtime_event      = latency_runtime_event,
1735                 .fork_event         = latency_fork_event,
1736                 .migrate_task_event = latency_migrate_task_event,
1737         };
1738         struct trace_sched_handler map_ops  = {
1739                 .switch_event       = map_switch_event,
1740         };
1741         struct trace_sched_handler replay_ops  = {
1742                 .wakeup_event       = replay_wakeup_event,
1743                 .switch_event       = replay_switch_event,
1744                 .fork_event         = replay_fork_event,
1745         };
1746
1747         argc = parse_options(argc, argv, sched_options, sched_usage,
1748                              PARSE_OPT_STOP_AT_NON_OPTION);
1749         if (!argc)
1750                 usage_with_options(sched_usage, sched_options);
1751
1752         /*
1753          * Aliased to 'perf script' for now:
1754          */
1755         if (!strcmp(argv[0], "script"))
1756                 return cmd_script(argc, argv, prefix);
1757
1758         symbol__init();
1759         if (!strncmp(argv[0], "rec", 3)) {
1760                 return __cmd_record(argc, argv);
1761         } else if (!strncmp(argv[0], "lat", 3)) {
1762                 sched.tp_handler = &lat_ops;
1763                 if (argc > 1) {
1764                         argc = parse_options(argc, argv, latency_options, latency_usage, 0);
1765                         if (argc)
1766                                 usage_with_options(latency_usage, latency_options);
1767                 }
1768                 setup_sorting(&sched, latency_options, latency_usage);
1769                 return perf_sched__lat(&sched);
1770         } else if (!strcmp(argv[0], "map")) {
1771                 sched.tp_handler = &map_ops;
1772                 setup_sorting(&sched, latency_options, latency_usage);
1773                 return perf_sched__map(&sched);
1774         } else if (!strncmp(argv[0], "rep", 3)) {
1775                 sched.tp_handler = &replay_ops;
1776                 if (argc) {
1777                         argc = parse_options(argc, argv, replay_options, replay_usage, 0);
1778                         if (argc)
1779                                 usage_with_options(replay_usage, replay_options);
1780                 }
1781                 return perf_sched__replay(&sched);
1782         } else {
1783                 usage_with_options(sched_usage, sched_options);
1784         }
1785
1786         return 0;
1787 }