perf sched: Add 'perf sched map' scheduling event map printout
authorIngo Molnar <mingo@elte.hu>
Wed, 16 Sep 2009 15:40:48 +0000 (17:40 +0200)
committerIngo Molnar <mingo@elte.hu>
Wed, 16 Sep 2009 14:41:30 +0000 (16:41 +0200)
This prints a textual context-switching outline of workload
captured via perf sched record.

For example, on a 16 CPU box it outputs:

   N1  O1  .   .   .   S1  .   .   .   B0  .  *I0  C1  .   M1  .    23002.773423 secs
   N1  O1  .  *Q0  .   S1  .   .   .   B0  .   I0  C1  .   M1  .    23002.773423 secs
   N1  O1  .   Q0  .   S1  .   .   .   B0  .  *R1  C1  .   M1  .    23002.773485 secs
   N1  O1  .   Q0  .   S1  .  *S0  .   B0  .   R1  C1  .   M1  .    23002.773478 secs
  *L0  O1  .   Q0  .   S1  .   S0  .   B0  .   R1  C1  .   M1  .    23002.773523 secs
   L0  O1  .  *.   .   S1  .   S0  .   B0  .   R1  C1  .   M1  .    23002.773531 secs
   L0  O1  .   .   .   S1  .   S0  .   B0  .   R1  C1 *T1  M1  .    23002.773547 secs T1 => irqbalance:2089
   L0  O1  .   .   .   S1  .   S0  .  *P0  .   R1  C1  T1  M1  .    23002.773549 secs
  *N1  O1  .   .   .   S1  .   S0  .   P0  .   R1  C1  T1  M1  .    23002.773566 secs
   N1  O1  .   .   .  *J0  .   S0  .   P0  .   R1  C1  T1  M1  .    23002.773571 secs
   N1  O1  .   .   .   J0  .   S0 *B0  P0  .   R1  C1  T1  M1  .    23002.773592 secs
   N1  O1  .   .   .   J0  .  *U0  B0  P0  .   R1  C1  T1  M1  .    23002.773582 secs
   N1  O1  .   .   .  *S1  .   U0  B0  P0  .   R1  C1  T1  M1  .    23002.773604 secs
   N1  O1  .   .   .   S1  .   U0  B0 *.   .   R1  C1  T1  M1  .    23002.773615 secs
   N1  O1  .   .   .   S1  .   U0  B0  .   .  *K0  C1  T1  M1  .    23002.773631 secs
   N1  O1  .  *M0  .   S1  .   U0  B0  .   .   K0  C1  T1  M1  .    23002.773624 secs
   N1  O1  .   M0  .   S1  .   U0 *.   .   .   K0  C1  T1  M1  .    23002.773644 secs
   N1  O1  .   M0  .   S1  .   U0  .   .   .  *R1  C1  T1  M1  .    23002.773662 secs
   N1  O1  .   M0  .   S1  .  *.   .   .   .   R1  C1  T1  M1  .    23002.773648 secs
   N1  O1  .  *.   .   S1  .   .   .   .   .   R1  C1  T1  M1  .    23002.773680 secs
   N1  O1  .   .   .  *L0  .   .   .   .   .   R1  C1  T1  M1  .    23002.773717 secs
  *N0  O1  .   .   .   L0  .   .   .   .   .   R1  C1  T1  M1  .    23002.773709 secs
  *N1  O1  .   .   .   L0  .   .   .   .   .   R1  C1  T1  M1  .    23002.773747 secs

Columns stand for individual CPUs, from CPU0 to CPU15, and the
two-letter shortcuts stand for tasks that are running on a CPU.

'*' denotes the CPU that had the event.

A dot signals an idle CPU.

New tasks are assigned new two-letter shortcuts - when they occur
first they are printed. In the above example 'T1' stood for irqbalance:

      T1 => irqbalance:2089

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
tools/perf/builtin-sched.c
tools/perf/util/thread.c
tools/perf/util/thread.h

index da8f67483ae711ededd523614ae11f1704b15f1e..f67e351b050b70470ccbeade659ce2f3c30190ca 100644 (file)
@@ -159,8 +159,6 @@ static struct rb_root               atom_root, sorted_atom_root;
 static u64                     all_runtime;
 static u64                     all_count;
 
-static int read_events(void);
-
 
 static u64 get_nsecs(void)
 {
@@ -634,38 +632,6 @@ static void test_calibrations(void)
        printf("the sleep test took %Ld nsecs\n", T1-T0);
 }
 
-static void __cmd_replay(void)
-{
-       unsigned long i;
-
-       calibrate_run_measurement_overhead();
-       calibrate_sleep_measurement_overhead();
-
-       test_calibrations();
-
-       read_events();
-
-       printf("nr_run_events:        %ld\n", nr_run_events);
-       printf("nr_sleep_events:      %ld\n", nr_sleep_events);
-       printf("nr_wakeup_events:     %ld\n", nr_wakeup_events);
-
-       if (targetless_wakeups)
-               printf("target-less wakeups:  %ld\n", targetless_wakeups);
-       if (multitarget_wakeups)
-               printf("multi-target wakeups: %ld\n", multitarget_wakeups);
-       if (nr_run_events_optimized)
-               printf("run atoms optimized: %ld\n",
-                       nr_run_events_optimized);
-
-       print_task_traces();
-       add_cross_task_wakeups();
-
-       create_tasks();
-       printf("------------------------------------------------------------\n");
-       for (i = 0; i < replay_repeat; i++)
-               run_one_test();
-}
-
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
@@ -1354,64 +1320,6 @@ static void sort_lat(void)
        }
 }
 
-static void __cmd_lat(void)
-{
-       struct rb_node *next;
-
-       setup_pager();
-       read_events();
-       sort_lat();
-
-       printf("\n -----------------------------------------------------------------------------------------\n");
-       printf("  Task                  |   Runtime ms  | Switches | Average delay ms | Maximum delay ms |\n");
-       printf(" -----------------------------------------------------------------------------------------\n");
-
-       next = rb_first(&sorted_atom_root);
-
-       while (next) {
-               struct work_atoms *work_list;
-
-               work_list = rb_entry(next, struct work_atoms, node);
-               output_lat_thread(work_list);
-               next = rb_next(next);
-       }
-
-       printf(" -----------------------------------------------------------------------------------------\n");
-       printf("  TOTAL:                |%11.3f ms |%9Ld |\n",
-               (double)all_runtime/1e6, all_count);
-
-       printf(" ---------------------------------------------------\n");
-       if (nr_unordered_timestamps && nr_timestamps) {
-               printf("  INFO: %.3f%% unordered timestamps (%ld out of %ld)\n",
-                       (double)nr_unordered_timestamps/(double)nr_timestamps*100.0,
-                       nr_unordered_timestamps, nr_timestamps);
-       } else {
-       }
-       if (nr_lost_events && nr_events) {
-               printf("  INFO: %.3f%% lost events (%ld out of %ld, in %ld chunks)\n",
-                       (double)nr_lost_events/(double)nr_events*100.0,
-                       nr_lost_events, nr_events, nr_lost_chunks);
-       }
-       if (nr_state_machine_bugs && nr_timestamps) {
-               printf("  INFO: %.3f%% state machine bugs (%ld out of %ld)",
-                       (double)nr_state_machine_bugs/(double)nr_timestamps*100.0,
-                       nr_state_machine_bugs, nr_timestamps);
-               if (nr_lost_events)
-                       printf(" (due to lost events?)");
-               printf("\n");
-       }
-       if (nr_context_switch_bugs && nr_timestamps) {
-               printf("  INFO: %.3f%% context switch bugs (%ld out of %ld)",
-                       (double)nr_context_switch_bugs/(double)nr_timestamps*100.0,
-                       nr_context_switch_bugs, nr_timestamps);
-               if (nr_lost_events)
-                       printf(" (due to lost events?)");
-               printf("\n");
-       }
-       printf("\n");
-
-}
-
 static struct trace_sched_handler *trace_handler;
 
 static void
@@ -1431,19 +1339,106 @@ process_sched_wakeup_event(struct raw_event_sample *raw,
        FILL_FIELD(wakeup_event, success, event, raw->data);
        FILL_FIELD(wakeup_event, cpu, event, raw->data);
 
-       trace_handler->wakeup_event(&wakeup_event, event, cpu, timestamp, thread);
+       if (trace_handler->wakeup_event)
+               trace_handler->wakeup_event(&wakeup_event, event, cpu, timestamp, thread);
 }
 
 /*
  * Track the current task - that way we can know whether there's any
  * weird events, such as a task being switched away that is not current.
  */
+static int max_cpu = 15;
+
 static u32 curr_pid[MAX_CPUS] = { [0 ... MAX_CPUS-1] = -1 };
 
+static struct thread *curr_thread[MAX_CPUS];
+
+static char next_shortname1 = 'A';
+static char next_shortname2 = '0';
+
+static void
+map_switch_event(struct trace_switch_event *switch_event,
+                struct event *event __used,
+                int this_cpu,
+                u64 timestamp,
+                struct thread *thread __used)
+{
+       struct thread *sched_out, *sched_in;
+       int new_shortname;
+       u64 timestamp0;
+       s64 delta;
+       int cpu;
+
+       BUG_ON(this_cpu >= MAX_CPUS || this_cpu < 0);
+
+       if (this_cpu > max_cpu)
+               max_cpu = this_cpu;
+
+       timestamp0 = cpu_last_switched[this_cpu];
+       cpu_last_switched[this_cpu] = timestamp;
+       if (timestamp0)
+               delta = timestamp - timestamp0;
+       else
+               delta = 0;
+
+       if (delta < 0)
+               die("hm, delta: %Ld < 0 ?\n", delta);
+
+
+       sched_out = threads__findnew(switch_event->prev_pid, &threads, &last_match);
+       sched_in = threads__findnew(switch_event->next_pid, &threads, &last_match);
+
+       curr_thread[this_cpu] = sched_in;
+
+       printf("  ");
+
+       new_shortname = 0;
+       if (!sched_in->shortname[0]) {
+               sched_in->shortname[0] = next_shortname1;
+               sched_in->shortname[1] = next_shortname2;
+
+               if (next_shortname1 < 'Z') {
+                       next_shortname1++;
+               } else {
+                       next_shortname1='A';
+                       if (next_shortname2 < '9') {
+                               next_shortname2++;
+                       } else {
+                               next_shortname2='0';
+                       }
+               }
+               new_shortname = 1;
+       }
+
+       for (cpu = 0; cpu <= max_cpu; cpu++) {
+               if (cpu != this_cpu)
+                       printf(" ");
+               else
+                       printf("*");
+
+               if (curr_thread[cpu]) {
+                       if (curr_thread[cpu]->pid)
+                               printf("%2s ", curr_thread[cpu]->shortname);
+                       else
+                               printf(".  ");
+               } else
+                       printf("   ");
+       }
+
+       printf("  %12.6f secs ", (double)timestamp/1e9);
+       if (new_shortname) {
+               printf("%s => %s:%d\n",
+                       sched_in->shortname, sched_in->comm, sched_in->pid);
+       } else {
+               printf("\n");
+       }
+}
+
+
 static void
 process_sched_switch_event(struct raw_event_sample *raw,
                           struct event *event,
-                          int cpu,
+                          int this_cpu,
                           u64 timestamp __used,
                           struct thread *thread __used)
 {
@@ -1459,17 +1454,18 @@ process_sched_switch_event(struct raw_event_sample *raw,
        FILL_FIELD(switch_event, next_pid, event, raw->data);
        FILL_FIELD(switch_event, next_prio, event, raw->data);
 
-       if (curr_pid[cpu] != (u32)-1) {
+       if (curr_pid[this_cpu] != (u32)-1) {
                /*
                 * Are we trying to switch away a PID that is
                 * not current?
                 */
-               if (curr_pid[cpu] != switch_event.prev_pid)
+               if (curr_pid[this_cpu] != switch_event.prev_pid)
                        nr_context_switch_bugs++;
        }
-       curr_pid[cpu] = switch_event.next_pid;
+       if (trace_handler->switch_event)
+               trace_handler->switch_event(&switch_event, event, this_cpu, timestamp, thread);
 
-       trace_handler->switch_event(&switch_event, event, cpu, timestamp, thread);
+       curr_pid[this_cpu] = switch_event.next_pid;
 }
 
 static void
@@ -1486,7 +1482,8 @@ process_sched_runtime_event(struct raw_event_sample *raw,
        FILL_FIELD(runtime_event, runtime, event, raw->data);
        FILL_FIELD(runtime_event, vruntime, event, raw->data);
 
-       trace_handler->runtime_event(&runtime_event, event, cpu, timestamp, thread);
+       if (trace_handler->runtime_event)
+               trace_handler->runtime_event(&runtime_event, event, cpu, timestamp, thread);
 }
 
 static void
@@ -1505,7 +1502,8 @@ process_sched_fork_event(struct raw_event_sample *raw,
        FILL_ARRAY(fork_event, child_comm, event, raw->data);
        FILL_FIELD(fork_event, child_pid, event, raw->data);
 
-       trace_handler->fork_event(&fork_event, event, cpu, timestamp, thread);
+       if (trace_handler->fork_event)
+               trace_handler->fork_event(&fork_event, event, cpu, timestamp, thread);
 }
 
 static void
@@ -1748,6 +1746,116 @@ more:
        return rc;
 }
 
+static void print_bad_events(void)
+{
+       if (nr_unordered_timestamps && nr_timestamps) {
+               printf("  INFO: %.3f%% unordered timestamps (%ld out of %ld)\n",
+                       (double)nr_unordered_timestamps/(double)nr_timestamps*100.0,
+                       nr_unordered_timestamps, nr_timestamps);
+       }
+       if (nr_lost_events && nr_events) {
+               printf("  INFO: %.3f%% lost events (%ld out of %ld, in %ld chunks)\n",
+                       (double)nr_lost_events/(double)nr_events*100.0,
+                       nr_lost_events, nr_events, nr_lost_chunks);
+       }
+       if (nr_state_machine_bugs && nr_timestamps) {
+               printf("  INFO: %.3f%% state machine bugs (%ld out of %ld)",
+                       (double)nr_state_machine_bugs/(double)nr_timestamps*100.0,
+                       nr_state_machine_bugs, nr_timestamps);
+               if (nr_lost_events)
+                       printf(" (due to lost events?)");
+               printf("\n");
+       }
+       if (nr_context_switch_bugs && nr_timestamps) {
+               printf("  INFO: %.3f%% context switch bugs (%ld out of %ld)",
+                       (double)nr_context_switch_bugs/(double)nr_timestamps*100.0,
+                       nr_context_switch_bugs, nr_timestamps);
+               if (nr_lost_events)
+                       printf(" (due to lost events?)");
+               printf("\n");
+       }
+}
+
+static void __cmd_lat(void)
+{
+       struct rb_node *next;
+
+       setup_pager();
+       read_events();
+       sort_lat();
+
+       printf("\n -----------------------------------------------------------------------------------------\n");
+       printf("  Task                  |   Runtime ms  | Switches | Average delay ms | Maximum delay ms |\n");
+       printf(" -----------------------------------------------------------------------------------------\n");
+
+       next = rb_first(&sorted_atom_root);
+
+       while (next) {
+               struct work_atoms *work_list;
+
+               work_list = rb_entry(next, struct work_atoms, node);
+               output_lat_thread(work_list);
+               next = rb_next(next);
+       }
+
+       printf(" -----------------------------------------------------------------------------------------\n");
+       printf("  TOTAL:                |%11.3f ms |%9Ld |\n",
+               (double)all_runtime/1e6, all_count);
+
+       printf(" ---------------------------------------------------\n");
+
+       print_bad_events();
+       printf("\n");
+
+}
+
+static struct trace_sched_handler map_ops  = {
+       .wakeup_event           = NULL,
+       .switch_event           = map_switch_event,
+       .runtime_event          = NULL,
+       .fork_event             = NULL,
+};
+
+static void __cmd_map(void)
+{
+       setup_pager();
+       read_events();
+       print_bad_events();
+}
+
+static void __cmd_replay(void)
+{
+       unsigned long i;
+
+       calibrate_run_measurement_overhead();
+       calibrate_sleep_measurement_overhead();
+
+       test_calibrations();
+
+       read_events();
+
+       printf("nr_run_events:        %ld\n", nr_run_events);
+       printf("nr_sleep_events:      %ld\n", nr_sleep_events);
+       printf("nr_wakeup_events:     %ld\n", nr_wakeup_events);
+
+       if (targetless_wakeups)
+               printf("target-less wakeups:  %ld\n", targetless_wakeups);
+       if (multitarget_wakeups)
+               printf("multi-target wakeups: %ld\n", multitarget_wakeups);
+       if (nr_run_events_optimized)
+               printf("run atoms optimized: %ld\n",
+                       nr_run_events_optimized);
+
+       print_task_traces();
+       add_cross_task_wakeups();
+
+       create_tasks();
+       printf("------------------------------------------------------------\n");
+       for (i = 0; i < replay_repeat; i++)
+               run_one_test();
+}
+
+
 static const char * const sched_usage[] = {
        "perf sched [<options>] {record|latency|replay|trace}",
        NULL
@@ -1867,6 +1975,10 @@ int cmd_sched(int argc, const char **argv, const char *prefix __used)
                }
                setup_sorting();
                __cmd_lat();
+       } else if (!strcmp(argv[0], "map")) {
+               trace_handler = &map_ops;
+               setup_sorting();
+               __cmd_map();
        } else if (!strncmp(argv[0], "rep", 3)) {
                trace_handler = &replay_ops;
                if (argc) {
index 12c4341078f9086d1804e16b0b06858d74aa6f20..45efb5db0d19819f281f75b17541dd9ebb76b8ad 100644 (file)
@@ -8,7 +8,7 @@
 
 static struct thread *thread__new(pid_t pid)
 {
-       struct thread *self = malloc(sizeof(*self));
+       struct thread *self = calloc(1, sizeof(*self));
 
        if (self != NULL) {
                self->pid = pid;
index 665d1f3dc9774b2d8160e1026fed269643553d0c..32aea3c1c2ad6efd174f6f67843134d033853fba 100644 (file)
@@ -7,6 +7,7 @@ struct thread {
        struct rb_node          rb_node;
        struct list_head        maps;
        pid_t                   pid;
+       char                    shortname[3];
        char                    *comm;
 };