Messages in this thread | | | Date | Fri, 22 May 2015 08:27:23 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH] perf-sched: add option to merge like comms to lat output |
| |
* Josef Bacik <jbacik@fb.com> wrote:
> Sometimes when debugging large multi-threaded applications it is helpful to > collate all of the latency numbers into one bulk record to get an idea of what > is going on. This patch does this by merging any entries that belong to the > same comm into one entry and then spits out those totals. I've also sligtly
s/sligtly slightly
> changed the output so you can see how many threads were merged in the > processing. Thanks,
Useful feature!
> + OPT_BOOLEAN('m', "merge-by-comm", &sched.merge_by_comm, > + "merge like tasks together in the output"),
Could you please make this the default behavior, and then add an option to print more detailed latencies?
We generally try to output general overviews first, and then make it easy to also output a more detailed view of information. That's how perf report, perf top and perf stat (-d, -dd, -ddd) works in general.
Also, please post a before/after sample output in the changelog.
Thanks,
Ingo
| |