bindings: Fixes the regression of vm_working_set_final_size_total.

bindings: Fixes the regression of vm_working_set_final_size_total.

WIll reduce vm_working_set_final_size_total from 188489.28 to 175400.96.

BUG=592310

[Yuki, 2016-04-04 08:12:05]

Description was changed from

==========
bindings: Fixes the regression of vm_working_set_final_size_total.

BUG=592310
==========

to

==========
bindings: Fixes the regression of vm_working_set_final_size_total.

WIll reduce vm_working_set_final_size_total from 188489.28 to 175400.96.

BUG=592310
==========

[Yuki, 2016-04-04 08:12:31]

yukishiino@chromium.org changed reviewers:
+ haraken@chromium.org

[Yuki, 2016-04-04 08:12:32]

Could you review this CL?

[haraken, 2016-04-04 08:43:19]

LGTM to try, but vm_working_set_final_size_total is a really noisy metric which
is affected GC timing, measurement timing etc. If the regression is observed
only in a limited number of benchmarks, I'm okay with ignoring it. *_peak_* is a
good metric but *_final_* is a bad metric (because the "final" totally depends
on various timings)...

[Yuki, 2016-04-04 09:39:25]

On 2016/04/04 08:43:19, haraken wrote:
> LGTM to try, but vm_working_set_final_size_total is a really noisy metric
which
> is affected GC timing, measurement timing etc. If the regression is observed
> only in a limited number of benchmarks, I'm okay with ignoring it. *_peak_* is
a
> good metric but *_final_* is a bad metric (because the "final" totally depends
> on various timings)...

Okay, let me close the issue without landing this CL.