[runtime] Record runtime call stats for Map::TransitionTo*Property, JSObject::OptimizeAsPrototype t…

[runtime] Record runtime call stats for Map::TransitionTo*Property, Map::SetPrototype and property deletion to separate buckets.

Committed: https://crrev.com/af02c0336da8be58eb664f4bb7b6994d16a53848
Cr-Commit-Position: refs/heads/master@{#36176}

[Toon Verwaest, 2016-05-11 11:27:21]

verwaest@chromium.org changed reviewers:
+ verwaest@chromium.org

[Toon Verwaest, 2016-05-11 11:27:21]

lgtm. comments can be addressed in a separate CL or this, as you wish

https://codereview.chromium.org/1973473002/diff/1/src/counters.h
File src/counters.h (right):

https://codereview.chromium.org/1973473002/diff/1/src/counters.h#newcode529
src/counters.h:529: V(Map_TransitionToAccessorProperty)                  \
What about already distinguishing PrototypeTransitionToAccessorProperty and
regular TransitionToAccessorProperty?

And perhaps SetPrototype as well?

[Igor Sheludko, 2016-05-11 11:49:33]

Description was changed from

==========
[runtime] Record runtime call stats for Map::TransitionTo*Property,
JSObject::OptimizeAsPrototype to separate buckets.
==========

to

==========
[runtime] Record runtime call stats for Map::TransitionTo*Property,
Map::SetPrototype and property deletion to separate buckets.
==========

[Igor Sheludko, 2016-05-11 12:03:48]

Patchset #2 (id:20001) has been deleted

[Igor Sheludko, 2016-05-11 12:04:49]

The CQ bit was checked by ishell@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-05-11 12:04:55]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1973473002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1973473002/40001

[commit-bot: I haz the power, 2016-05-11 12:45:19]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-05-11 12:45:20]

Dry run: This issue passed the CQ dry run.

[Toon Verwaest, 2016-05-11 12:53:44]

thanks, lgtm

[Igor Sheludko, 2016-05-11 12:57:42]

The CQ bit was checked by ishell@chromium.org

[commit-bot: I haz the power, 2016-05-11 12:57:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1973473002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1973473002/40001

[commit-bot: I haz the power, 2016-05-11 12:59:55]

Committed patchset #2 (id:40001)

[commit-bot: I haz the power, 2016-05-11 13:01:07]

Description was changed from

==========
[runtime] Record runtime call stats for Map::TransitionTo*Property,
Map::SetPrototype and property deletion to separate buckets.
==========

to

==========
[runtime] Record runtime call stats for Map::TransitionTo*Property,
Map::SetPrototype and property deletion to separate buckets.

Committed: https://crrev.com/af02c0336da8be58eb664f4bb7b6994d16a53848
Cr-Commit-Position: refs/heads/master@{#36176}
==========

[commit-bot: I haz the power, 2016-05-11 13:01:07]

Patchset 2 (id:??) landed as
https://crrev.com/af02c0336da8be58eb664f4bb7b6994d16a53848
Cr-Commit-Position: refs/heads/master@{#36176}

[Camillo Bruni, 2016-05-11 13:01:45]

cbruni@chromium.org changed reviewers:
+ cbruni@chromium.org

[Camillo Bruni, 2016-05-11 13:01:46]

https://codereview.chromium.org/1973473002/diff/40001/src/lookup.cc
File src/lookup.cc (right):

https://codereview.chromium.org/1973473002/diff/40001/src/lookup.cc#newcode359
src/lookup.cc:359: : &RuntimeCallStats::Object_DeleteProperty);
I hope this case doesn't happen too much, I've seen quite some impact by the
counter when using them for the builtins. That's why I added the separate branch
that only jumps there if we enable the flag. Otherwise I could still see the
overhead, since clang inlined the timer scope nevertheless and probably reserved
stack in either case.

[Igor Sheludko, 2016-05-11 17:59:16]

On 2016/05/11 13:01:46, Camillo Bruni wrote:
> https://codereview.chromium.org/1973473002/diff/40001/src/lookup.cc
> File src/lookup.cc (right):
> 
> https://codereview.chromium.org/1973473002/diff/40001/src/lookup.cc#newcode359
> src/lookup.cc:359: : &RuntimeCallStats::Object_DeleteProperty);
> I hope this case doesn't happen too much, I've seen quite some impact by the
> counter when using them for the builtins. That's why I added the separate
branch
> that only jumps there if we enable the flag. Otherwise I could still see the
> overhead, since clang inlined the timer scope nevertheless and probably
reserved
> stack in either case.

Let's see what does the perf think about this CL.