perf stat: Fix multi-run stats
Commit-ID: 566747e6298289c5cb02d4939cb3abf1c4fe7e5a Gitweb: http://git.kernel.org/tip/566747e6298289c5cb02d4939cb3abf1c4fe7e5a Author: Ingo Molnar <[email protected]> AuthorDate: Sat, 27 Jun 2009 06:24:32 +0200 Committer: Ingo Molnar <[email protected]> CommitDate: Sat, 27 Jun 2009 06:34:37 +0200
perf stat: Fix multi-run stats
In multi-run (-r/--repeat) printouts, print out the noise of the wall-clock average as well.
Also, fix a bug in printing out scaled counters: if it was not scaled then we should not update the average with -1.
Cc: Peter Zijlstra <[email protected]>
Cc: Mike Galbraith <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>
LKML-Reference: <new-submission>
--- tools/perf/builtin-stat.c | 15 +++++++++++---- 1 files changed, 11 insertions(+), 4 deletions(-)
tools/perf/builtin-stat.c +11 -4
@@ -353,8 +353,11 @@ static void calc_avg(void) event_res_avg[j]+1, event_res[i][j]+1);update_avg ("counter/2" , j, event_res_avg[j]+2 , event_res[i][j]+2 ); -update_avg ("scaled" , j, - event_scaled_avg + j, event_scaled[i]+j); +if (event_scaled[i][j] !=-1 ) +update_avg ("scaled" , j, + event_scaled_avg + j, event_scaled[i]+j); +else + event_scaled_avg[j] =-1 ; } } runtime_nsecs_avg /= run_count;
@@ -420,9 +423,13 @@ static void print_stat(int argc, const char **argv)fprintf (stderr,"\n" ); -fprintf (stderr," %14.9f seconds time elapsed.\n" , +fprintf (stderr," %14.9f seconds time elapsed" , (double )walltime_nsecs_avg/1e9 ); -fprintf (stderr,"\n" ); +if (run_count >1 ) { +fprintf (stderr," ( +- %7.3f%% )" , +100.0 *(double )walltime_nsecs_noise/(double )walltime_nsecs_avg); + } +fprintf (stderr,"\n\n" ); }static volatile int signr =-1 ;
Quoted 20 lines · tip-bot for Ingo Molnar
Commit-ID: 566747e6298289c5cb02d4939cb3abf1c4fe7e5a Gitweb: http://git.kernel.org/tip/566747e6298289c5cb02d4939cb3abf1c4fe7e5a Author: Ingo Molnar <[email protected]> AuthorDate: Sat, 27 Jun 2009 06:24:32 +0200 Committer: Ingo Molnar <[email protected]> CommitDate: Sat, 27 Jun 2009 06:34:37 +0200perf stat: Fix multi-run stats
In multi-run (-r/--repeat) printouts, print out the noise of the wall-clock average as well.
Also, fix a bug in printing out scaled counters: if it was not scaled then we should not update the average with -1.
Cc: Peter Zijlstra <[email protected]>
Cc: Mike Galbraith <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>
LKML-Reference: <new-submission>
Quoted 7 lines
Signed-off-by Ingo Molnar <[email protected]>--- tools/perf/builtin-stat.c | 15 +++++++++++---- 1 files changed, 11 insertions(+), 4 deletions(-)
hmm, why you are keep on changing tools/perf/builtin-stat.c you know my patches are in queue which I prepared on your request and send pull request :
In spite of adding the patches you keep on changing the original file ??
--
JSR
Quoted 30 lines · Jaswinder Singh Rajput
Quoted 20 lines · tip-bot for Ingo Molnar
Commit-ID: 566747e6298289c5cb02d4939cb3abf1c4fe7e5a Gitweb: http://git.kernel.org/tip/566747e6298289c5cb02d4939cb3abf1c4fe7e5a Author: Ingo Molnar <[email protected]> AuthorDate: Sat, 27 Jun 2009 06:24:32 +0200 Committer: Ingo Molnar <[email protected]> CommitDate: Sat, 27 Jun 2009 06:34:37 +0200perf stat: Fix multi-run stats
In multi-run (-r/--repeat) printouts, print out the noise of the wall-clock average as well.
Also, fix a bug in printing out scaled counters: if it was not scaled then we should not update the average with -1.
Cc: Peter Zijlstra <[email protected]>
Cc: Mike Galbraith <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>
LKML-Reference: <new-submission>Quoted 7 lines
Signed-off-by Ingo Molnar <[email protected]>--- tools/perf/builtin-stat.c | 15 +++++++++++---- 1 files changed, 11 insertions(+), 4 deletions(-)
Quoted 8 lines
hmm, why you are keep on changing
tools/perf/builtin-stat.cyou know my patches are in queue which I prepared on your request and send pull request:In spite of adding the patches you keep on changing the original file ??
Two things:
- My changes and your changes do not conflict - i was able to apply your patches on top of my changes.
- I could have applied all three patches of yours but i only applied the first one because the second and third did not implement the feature you are trying to add in an adequate way.
Also, i do not 'request' things from you, i am not your boss. I review patches and i reject patches i find inadequate. You are free to submit new versions - or you are free to not do so. If you submit new versions i will review them too and will merge them or tell why i didnt merge them.
Ingo