Re: [PATCH v6 3/4] perf script: Fix perf script -F +metric

[Andi Kleen <ak@linux.intel.com>]

> The main reason would be so that perf_stat_process_counter could work
> and update the aggregation accordingly. Then you can dump out the
> aggregation, be it per CPU, per core, per socket, per cache-level,
> etc. as appropriate. 

But there's is none. It's only for a short time on a single CPU.

> If you follow the perf stat convention you are
> also much less likely to be broken in the future, as perf stat would
> also get broken. It is complicated spaghetti to work out how this
> stuff works, but that's why I fixed in the patch I sent out.

Hopefully the regression tests will prevent future breakage.

> Or let it set the aggregation mode and just let the aggregation code
> handle it when computing metrics? I think getting the STAT events
> isn't typical, so this is an academic argument.

I don't see why users shouldn't use perf stat record. It makes
sense for any large scale count collections. And it's tested by
the test suite at least.

-Andi