[Devel] Re: [PATCH 3/4] perf: teach perf inject to merge sched_stat_* and sched_switch events

[Andrew Vagin]

Hi Arun,

On 11/04/2011 04:08 AM, Arun Sharma wrote:
> On 11/2/11 3:59 PM, Andrew Vagin wrote:
>
>> + event_sw = (union perf_event *)&ent->header;
>
> This causes problems with -Wstrict-aliasing=3.
I didn't see any warning, but I reworked this part a bit.
>
> I noticed that the suggested cmdline is:
>
> perf script record sched-stat -- -e sched:sched_stat_sleep
>
> I'd much prefer something along the lines of:
>
> perf script record sched-stat -e <events> -- sleep 1
>
> Also, the perf inject command is a bit unforgiving when the user makes
> some mistakes (eg: didn't specify -e sched:sched_stat_sleep). The
> failure mode is hard to understand (perf report -i perf.data.d fails
> with "no samples").

I see the problem. It's not in "perf inject". Look at new version I 
improved "perf script sched_stat". Now it may be executed for specific 
command and it reports an error if you forgot to set a sched_stat_* event.

> Could not find sched_switch for pid 16179
>
> I think this is the same issue you referenced in "Known issues".
Actually It isn't. Now I print this message as debug information.
>
> Although I'm able to get some profiling info out of perf.data.d, the
> process I really wanted to trace got zero samples.
It's strange. Can you check new versions and describe the experiment if 
it will be reproduces.
>
> -Arun
>