Skip to contentSkip to conversation

perf: Timehist account sch delay for scheduled out running

PATCH Applied Fernand Sieber 3 participants 3 messages Jun 18, 2024 – Jun 26, 2024 perf events

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 perf: Timehist account sch delay for scheduled out running

When using perf timehist, sch delay is only computed for a waking task, not for a pre empted task. This patches changes sch delay to account for both. This makes sense as testing scheduling policy need to consider the effect of scheduling delay globally, not only for waking tasks.

Example of perf timehist report before the patch for stress task competing with each other.

First column is wait time, second column sch delay, third column runtime.

1.492060 [0000]  s    stress[81]                          1.999      0.000      2.000      R  next: stress[83]
1.494060 [0000]  s    stress[83]                          2.000      0.000      2.000      R  next: stress[81]
1.496060 [0000]  s    stress[81]                          2.000      0.000      2.000      R  next: stress[83]
1.498060 [0000]  s    stress[83]                          2.000      0.000      1.999      R  next: stress[81]

After the patch, it looks like this (note that all wait time is not zero anymore):

1.492060 [0000]  s    stress[81]                          1.999      1.999      2.000      R  next: stress[83]
1.494060 [0000]  s    stress[83]                          2.000      2.000      2.000      R  next: stress[81]
1.496060 [0000]  s    stress[81]                          2.000      2.000      2.000      R  next: stress[83]
1.498060 [0000]  s    stress[83]                          2.000      2.000      1.999      R  next: stress[81]
Cc Peter Zijlstra <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Alexander Shishkin <[email protected]>
Cc Madadi Vineeth Reddy <[email protected]>
---
 tools/perf/Documentation/perf-sched.txt | 4 ++--
 tools/perf/builtin-sched.c              | 5 ++++-
 2 files changed, 6 insertions(+), 3 deletions(-)
tools/perf/Documentation/perf-sched.txt +2 -2
index a216d2991b19..74c812f7a4a4 100644
@@ -64,8 +64,8 @@ There are several variants of 'perf sched':
     
    By default it shows the individual schedule events, including the wait
    time (time between sched-out and next sched-in events for the task), the
-   task scheduling delay (time between wakeup and actually running) and run
-   time for the task:
+   task scheduling delay (time between runnable and actually running) and
+   run time for the task:
     
                 time    cpu  task name             wait time  sch delay   run time
                              [tid/pid]                (msec)     (msec)     (msec)
tools/perf/builtin-sched.c +4 -1
index 5977c49ae2c7..7422c930abaf 100644
@@ -2659,7 +2659,10 @@ static int timehist_sched_change_event(struct perf_tool *tool,
 		tr->last_state = state;
 
 		/* sched out event for task so reset ready to run time */
-		tr->ready_to_run = 0;
+		if (state == 'R')
+			tr->ready_to_run = t;
+		else
+			tr->ready_to_run = 0;
 	}
 
 	evsel__save_time(evsel, sample->time, sample->cpu);
Signature · 1 line
2.40.1
Reviewed-by on v2 1/1
#1 lore raw

Hi Fernand,

Quoted 40 lines · Fernand Sieber

When using perf timehist, sch delay is only computed for a waking task, not for a pre empted task. This patches changes sch delay to account for both. This makes sense as testing scheduling policy need to consider the effect of scheduling delay globally, not only for waking tasks.

Example of perf timehist report before the patch for stress task competing with each other.

First column is wait time, second column sch delay, third column runtime.

1.492060 [0000]  s    stress[81]                          1.999      0.000      2.000      R  next: stress[83]
1.494060 [0000]  s    stress[83]                          2.000      0.000      2.000      R  next: stress[81]
1.496060 [0000]  s    stress[81]                          2.000      0.000      2.000      R  next: stress[83]
1.498060 [0000]  s    stress[83]                          2.000      0.000      1.999      R  next: stress[81]

After the patch, it looks like this (note that all wait time is not zero anymore):

1.492060 [0000]  s    stress[81]                          1.999      1.999      2.000      R  next: stress[83]
1.494060 [0000]  s    stress[83]                          2.000      2.000      2.000      R  next: stress[81]
1.496060 [0000]  s    stress[81]                          2.000      2.000      2.000      R  next: stress[83]
1.498060 [0000]  s    stress[83]                          2.000      2.000      1.999      R  next: stress[81]
Cc Peter Zijlstra <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Alexander Shishkin <[email protected]>
Cc Madadi Vineeth Reddy <[email protected]>
---
 tools/perf/Documentation/perf-sched.txt | 4 ++--
 tools/perf/builtin-sched.c              | 5 ++++-
 2 files changed, 6 insertions(+), 3 deletions(-)
tools/perf/Documentation/perf-sched.txt +0 -0
index a216d2991b19..74c812f7a4a4 100644

tools/perf/Documentation/perf-sched.txt There are several variants of 'perf sched': · line 68

Quoted 7 lines
tools/perf/Documentation/perf-sched.txt +2 -2
@@ -64,8 +64,8 @@ There are several variants of 'perf sched':
 
    By default it shows the individual schedule events, including the wait
    time (time between sched-out and next sched-in events for the task), the
-   task scheduling delay (time between wakeup and actually running) and run
-   time for the task:
+   task scheduling delay (time between runnable and actually running) and
+   run time for the task:

LGTM.

Thanks and Regards
Madadi Vineeth Reddy

Quoted 19 lines
tools/perf/Documentation/perf-sched.txt +0 -0
                 time    cpu  task name             wait time  sch delay   run time
                              [tid/pid]                (msec)     (msec)     (msec)
tools/perf/builtin-sched.c +4 -1
index 5977c49ae2c7..7422c930abaf 100644
@@ -2659,7 +2659,10 @@ static int timehist_sched_change_event(struct perf_tool *tool,
 		tr->last_state = state;
 
 		/* sched out event for task so reset ready to run time */
-		tr->ready_to_run = 0;
+		if (state == 'R')
+			tr->ready_to_run = t;
+		else
+			tr->ready_to_run = 0;
 	}
 
 	evsel__save_time(evsel, sample->time, sample->cpu);
maintainer Applied
#2 lore raw
In reply to Fernand Sieber
Quoted 9 lines · Fernand Sieber

When using perf timehist, sch delay is only computed for a waking task, not for a pre empted task. This patches changes sch delay to account for both. This makes sense as testing scheduling policy need to consider the effect of scheduling delay globally, not only for waking tasks.

Example of perf timehist report before the patch for stress task competing with each other.

[...]

Applied to perf-tools-next, thanks!

Best regards,
Namhyung