Re: [PATCH 1/2] perf sched: Introduce timehist command - v2

[David Ahern]

On 12/2/13, 12:58 AM, Namhyung Kim wrote:
> >             time cpu  task name[tid/pid]    b/n time sch delay  run time
> >    ------------- ---- -------------------- --------- --------- ---------
> >     79371.874569 [11] gcc[31949]               0.014     0.000     1.148
> >     79371.874591 [10] gcc[31951]               0.000     0.000     0.024
> >     79371.874603 [10] migration/10[59]         3.350     0.004     0.011
> >     79371.874604 [11] <idle>                   1.148     0.000     0.035
> >     79371.874723 [05] <idle>                   0.016     0.000     1.383
> >     79371.874746 [05] gcc[31949]               0.153     0.078     0.022
> > ...
> >
> > Times are in msec.usec.
>
> Hmm.. I'm not sure this is right.  It probably confuse users since
> timehist_time_str() still uses "sec.usec" format and it looks not
> natural for me to use "msec".
>
> Yeah, I see perf stat uses "msec.usec" for result of clock events but
> AFAICT it also shows the unit explicitly.  And perf stat -I uses
> "sec.nsec" format and perf script also uses "sec.usec" format so there's
> a little consistency here.
>
> I think this "msec.usec" format fits well for the scheduling events but
> in general "sec.usec" format looks better IMHO.

Arnaldo / Ingo: any thoughts on the units here? sec.usec versus msec.usec?

David

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/