Preparing components_perftests

Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
Committed: https://crrev.com/bd75dc348f2e52e1fc3e1b26b6c7528329657d65
Cr-Commit-Position: refs/heads/master@{#427043}

[dyaroshev, 2016-09-21 16:12:04]

Description was changed from

==========
Creating a blank components perf test:

1 - creating main that does all the necessary initialization for
    components.
2 - deciding on perf frame work base/test/perf* vs testing/perf/*
3 - enabling a perf test in regular runs

BUG=648992
==========

to

==========
Creating a blank components perf test:

1 - creating main that does all the necessary initialization for
    components.
2 - deciding on perf frame work base/test/perf* vs testing/perf/*
3 - enabling a perf test in regular runs

BUG=648992
==========

[dyaroshev, 2016-09-21 16:12:04]

dyaroshev@yandex-team.ru changed reviewers:
+ cpu@chromium.org, jochen@chromium.org, nduca@chromium.org, pkasting@google.com

[dyaroshev, 2016-09-21 16:17:47]

Description was changed from

==========
Creating a blank components perf test:

1 - creating main that does all the necessary initialization for
    components.
2 - deciding on perf frame work base/test/perf* vs testing/perf/*
3 - enabling a perf test in regular runs

BUG=648992
==========

to

==========
Creating a blank components perf test:

1 - creating main that does all the necessary initialization for
    components (taken from components_unittests)
2 - deciding on perf frame work base/test/perf* vs testing/perf/*
3 - enabling a perf test in regular runs

BUG=648992
==========

[cpu_(ooo_6.6-7.5), 2016-09-27 15:38:36]

cpu@chromium.org changed reviewers:
- cpu@chromium.org

[dyaroshev, 2016-09-28 11:55:07]

dyaroshev@yandex-team.ru changed reviewers:
+ eyaich@google.com

[dyaroshev, 2016-09-28 11:56:09]

dyaroshev@yandex-team.ru changed reviewers:
+ dtu@chromium.org

[nduca, 2016-09-28 19:46:42]

so, your'e going to hate me, but i have a lot of questions here, and a general
comment.

1. Your CL is still doing too much at once, so youre going to get a lot of
confused people who know one aspect of your CL who look at the other parts and
either (a) say I can't review this, or (b) give you confusing feedback because
they don't know the code. Myself included. You should split this more. The
general princple I try to apply is that every patch is a complete sentence, not
a paragraph, and definitely not an essay. Here, as far as i can tell, you have a
paragraph:
1a. add components/test/components_test_suite.cc
1b. add components/test/run_all_perf_tests.cc and gn targets
1c. modify base/ for some specific reason that i don't understand but clearly
has a reason.
1d. create the skeleton for the omnibox test
1e. make buildbot aware of the new componets perf unittests
1f. make perf_json_generator aware fo the new componetns perf unittests

These are all patches in their own right. Each has a slightly different
reviewer. My skillset is on 1d, I can somewhat review 1e and 1f. I am not the
expert to review the other bits, nor am I owner in any of these places so my
l-g-t is useless anyway.

3. You have a *test_suit.h and *test_suit.cc where I believe the correct
filename is *test_suite.h and *test_suite.cc.

4. your ComponetsTestSuite::initialize seems very very long and its unclear to
me which of those are standard boilerplate for components unit tests, and which
are unique to your situation. Someone in components should review your suite
instead of me and check that it is okay to have all this large amount of text
there.

Hope this helps.

https://codereview.chromium.org/2358063002/diff/20001/components/test/compone...
File components/test/components_test_suit.h (right):

https://codereview.chromium.org/2358063002/diff/20001/components/test/compone...
components/test/components_test_suit.h:1: // Copyright (c) 2016 The Chromium
Authors. All rights reserved.
At least from my perspective, this file should be called components_test_suite.h
not components_test_suit.h.

[Peter Kasting, 2016-09-28 20:13:06]

On 2016/09/28 19:46:42, nduca wrote:
> 4. your ComponetsTestSuite::initialize seems very very long and its unclear to
> me which of those are standard boilerplate for components unit tests, and
which
> are unique to your situation. Someone in components should review your suite
> instead of me and check that it is okay to have all this large amount of text
> there.

Nat asked me to glance at this one; I'm not an expert here either but my gut
sense is that you don't need most of this and I would try to rip out everything,
or at least as much as possible until something breaks.

[dyaroshev, 2016-09-28 20:15:50]

On 2016/09/28 19:46:42, nduca wrote:
> so, your'e going to hate me, but i have a lot of questions here, and a general
> comment.
> 
> 1. Your CL is still doing too much at once, so youre going to get a lot of
> confused people who know one aspect of your CL who look at the other parts and
> either (a) say I can't review this, or (b) give you confusing feedback because
> they don't know the code. Myself included. You should split this more. The
> general princple I try to apply is that every patch is a complete sentence,
not
> a paragraph, and definitely not an essay. Here, as far as i can tell, you have
a
> paragraph:
> 1a. add components/test/components_test_suite.cc
> 1b. add components/test/run_all_perf_tests.cc and gn targets
> 1c. modify base/ for some specific reason that i don't understand but clearly
> has a reason.
> 1d. create the skeleton for the omnibox test
> 1e. make buildbot aware of the new componets perf unittests
> 1f. make perf_json_generator aware fo the new componetns perf unittests
>
> These are all patches in their own right. Each has a slightly different
> reviewer.

Do you mean that I have to split this up in different CLs, or different pathces
would be sufficient?
Most of these changes are just moving staff around so that components_perftests
would work.

> 3. You have a *test_suit.h and *test_suit.cc where I believe the correct
> filename is *test_suite.h and *test_suite.cc.
> 

My bad, spelling is not my strongest side.

> 4. your ComponetsTestSuite::initialize seems very very long and its unclear to
> me which of those are standard boilerplate for components unit tests, and
which
> are unique to your situation. Someone in components should review your suite
> instead of me and check that it is okay to have all this large amount of text
> there.
>

I just stole it from run_all_unittests, this is not what I know and it should
not be me who refactors it.

[Peter Kasting, 2016-09-28 20:20:59]

On 2016/09/28 20:15:50, dyaroshev wrote:
> > 4. your ComponetsTestSuite::initialize seems very very long and its unclear
to
> > me which of those are standard boilerplate for components unit tests, and
> which
> > are unique to your situation. Someone in components should review your suite
> > instead of me and check that it is okay to have all this large amount of
text
> > there.
> >
> 
> I just stole it from run_all_unittests, this is not what I know and it should
> not be me who refactors it.

You don't need to refactor the existing code in any way, but you do need to
avoid copying a bunch of functionality into brand-new code unless your test
needs it.

[dyaroshev, 2016-09-28 21:09:20]

On 2016/09/28 20:20:59, Peter Kasting wrote:
> You don't need to refactor the existing code in any way, but you do need to
> avoid copying a bunch of functionality into brand-new code unless your test
> needs it.

I didn't copy it, I extracted existing code in a new file.

> Nat asked me to glance at this one; I'm not an expert here either but my gut
> sense is that you don't need most of this and I would try to rip out
everything,
> or at least as much as possible until something breaks.

1 - I've tried to model after cc/ - all setup is moved to one place. This is
actually a nice design feature because we won't have to think about two
different places when modifing test suite.
2 - This would probably lead to bugs on different oses on which I don't even
work (like IOS).


Can I modify one existing component_perftest to use testing/perf instead of
base/perf? This would make my patch cosiderably smaller: no changes in base,
much less changes in cut and paste from run_components_unittests. And I won't
have to create a new perf test file here - I can just use the existing one to
test changes.

[Peter Kasting, 2016-09-28 23:49:01]

On 2016/09/28 21:09:20, dyaroshev wrote:
> On 2016/09/28 20:20:59, Peter Kasting wrote:
> > You don't need to refactor the existing code in any way, but you do need to
> > avoid copying a bunch of functionality into brand-new code unless your test
> > needs it.
> 
> I didn't copy it, I extracted existing code in a new file.

Oh, argh.  I looked for that, and missed it.

> > Nat asked me to glance at this one; I'm not an expert here either but my gut
> > sense is that you don't need most of this and I would try to rip out
> everything,
> > or at least as much as possible until something breaks.
> 
> 1 - I've tried to model after cc/ - all setup is moved to one place. This is
> actually a nice design feature because we won't have to think about two
> different places when modifing test suite.

Well, it depends.  If you have a unittest system that runs tens of thousands of
tests with tons of dependencies and you're adding a perf test system that runs a
very small number, then it seems less than ideal to hoist the former's
dependencies so they're pulled in by the latter.

> 2 - This would probably lead to bugs on different oses on which I don't even
> work (like IOS).

That's why we have trybots; if a change passes the tryjobs, it should be OK.  If
a platform isn't part of the try infrastructure and gets broken by a change,
that's basically on them, not you.

> Can I modify one existing component_perftest to use testing/perf instead of
> base/perf? This would make my patch cosiderably smaller: no changes in base,
> much less changes in cut and paste from run_components_unittests. And I won't
> have to create a new perf test file here - I can just use the existing one to
> test changes.

I have no opinion on this.

[Peter Kasting, 2016-10-05 00:20:00]

pkasting@chromium.org changed reviewers:
+ pkasting@chromium.org
- pkasting@google.com

[dyaroshev, 2016-10-07 23:38:00]

Description was changed from

==========
Creating a blank components perf test:

1 - creating main that does all the necessary initialization for
    components (taken from components_unittests)
2 - deciding on perf frame work base/test/perf* vs testing/perf/*
3 - enabling a perf test in regular runs

BUG=648992
==========

to

==========
This patch
1 - creates a complete main for components perftests
2 - decides on using testing/perf for components_perftests

BUG=648992
==========

[dyaroshev, 2016-10-07 23:45:59]

On 2016/09/28 23:49:01, Peter Kasting (OOO Oct. 6-9) wrote:
> > Can I modify one existing component_perftest to use testing/perf instead of
> > base/perf? This would make my patch cosiderably smaller: no changes in base,
> > much less changes in cut and paste from run_components_unittests. And I
won't
> > have to create a new perf test file here - I can just use the existing one
to
> > test changes.
> 
> I have no opinion on this.

I changed. It gives a few things:
1 - much smaller patch: no weird initialization for base/perf is required
2 - We've decided to use testing/perf for History perf tests. Now I don't have
to come up with some blank test to see if this api is set up properly - we can
use the existing one.

I still have no idea how to properly enable it in regular runs. I suggest, we
create a separate task on people who do that. We would still have a benchmark to
test locally at the very least.

Off-top: because of the amount of time I've spent on this benchmark, my
superiors won't give me more until the actual flat maps. I'll do what I can in
my spare time, but it seems like the progress would go very slow.

[dyaroshev, 2016-10-13 07:48:56]

ping @pkasting, @nduca

[Peter Kasting, 2016-10-17 18:44:47]

I don't see major concerns, the .gn changes are the critical ones.

https://codereview.chromium.org/2358063002/diff/40001/components/BUILD.gn
File components/BUILD.gn (right):

https://codereview.chromium.org/2358063002/diff/40001/components/BUILD.gn#new...
components/BUILD.gn:41: "test/components_test_suite.h",
I don't think you want these sources (and their dependencies) to appear in both
this target and the one below.

You probably want these, and their dependencies, in a source_set that both
targets depend on.

https://codereview.chromium.org/2358063002/diff/40001/components/test/compone...
File components/test/components_test_suite.h (right):

https://codereview.chromium.org/2358063002/diff/40001/components/test/compone...
components/test/components_test_suite.h:1: // Copyright (c) 2016 The Chromium
Authors. All rights reserved.
Nit: No (c) (see
https://chromium.googlesource.com/chromium/src/+/master/styleguide/c++/c++.md...
)

(All added files)

https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
File components/visitedlink/test/visitedlink_perftest.cc (right):

https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
components/visitedlink/test/visitedlink_perftest.cc:27: // designed like
base/test/perf_time_logger, but uses different output strategy.
Nit: Initial capital

You might want to describe what the output strategy is and/or why it wants to be
different.

https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
components/visitedlink/test/visitedlink_perftest.cc:53: // integer milliseconds
Nit: Add trailing period

https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
components/visitedlink/test/visitedlink_perftest.cc:54:
perf_test::PrintResult(test_name_, "", "", timer_.Elapsed().InMillisecondsF(),
Nit: "" -> std::string()

https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
components/visitedlink/test/visitedlink_perftest.cc:231:
perf_test::PrintResult("Visited_link_cold_load_time", "", "",
Nit: "" -> std::string (2 places)

[dyaroshev, 2016-10-18 15:16:34]

On 2016/10/17 18:44:47, Peter Kasting wrote:
> I don't see major concerns, the .gn changes are the critical ones.
> 
> https://codereview.chromium.org/2358063002/diff/40001/components/BUILD.gn
> File components/BUILD.gn (right):
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/BUILD.gn#new...
> components/BUILD.gn:41: "test/components_test_suite.h",
> I don't think you want these sources (and their dependencies) to appear in
both
> this target and the one below.
> 
> You probably want these, and their dependencies, in a source_set that both
> targets depend on.
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/test/compone...
> File components/test/components_test_suite.h (right):
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/test/compone...
> components/test/components_test_suite.h:1: // Copyright (c) 2016 The Chromium
> Authors. All rights reserved.
> Nit: No (c) (see
>
https://chromium.googlesource.com/chromium/src/+/master/styleguide/c++/c++.md...
> )
> 
> (All added files)
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
> File components/visitedlink/test/visitedlink_perftest.cc (right):
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
> components/visitedlink/test/visitedlink_perftest.cc:27: // designed like
> base/test/perf_time_logger, but uses different output strategy.
> Nit: Initial capital
> 
> You might want to describe what the output strategy is and/or why it wants to
be
> different.
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
> components/visitedlink/test/visitedlink_perftest.cc:53: // integer
milliseconds
> Nit: Add trailing period
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
> components/visitedlink/test/visitedlink_perftest.cc:54:
> perf_test::PrintResult(test_name_, "", "", timer_.Elapsed().InMillisecondsF(),
> Nit: "" -> std::string()
> 
>
https://codereview.chromium.org/2358063002/diff/40001/components/visitedlink/...
> components/visitedlink/test/visitedlink_perftest.cc:231:
> perf_test::PrintResult("Visited_link_cold_load_time", "", "",
> Nit: "" -> std::string (2 places)

I think I've fixed everything.

[Peter Kasting, 2016-10-18 17:31:13]

https://codereview.chromium.org/2358063002/diff/60001/components/BUILD.gn
File components/BUILD.gn (right):

https://codereview.chromium.org/2358063002/diff/60001/components/BUILD.gn#new...
components/BUILD.gn:41: "test/components_test_suite.h",
You still list these sources here and below, though they're now in your
test_support library as well.  Remove.

https://codereview.chromium.org/2358063002/diff/60001/components/test/BUILD.gn
File components/test/BUILD.gn (right):

https://codereview.chromium.org/2358063002/diff/60001/components/test/BUILD.g...
components/test/BUILD.gn:5: static_library("test_support") {
I think you want a source_set here and not a static_library.  If I'm wrong,
please explain why :)

[dyaroshev, 2016-10-19 12:53:04]

On 2016/10/18 17:31:13, Peter Kasting wrote:
> https://codereview.chromium.org/2358063002/diff/60001/components/BUILD.gn
> File components/BUILD.gn (right):
> 
>
https://codereview.chromium.org/2358063002/diff/60001/components/BUILD.gn#new...
> components/BUILD.gn:41: "test/components_test_suite.h",
> You still list these sources here and below, though they're now in your
> test_support library as well.  Remove.
> 
> https://codereview.chromium.org/2358063002/diff/60001/components/test/BUILD.gn
> File components/test/BUILD.gn (right):
> 
>
https://codereview.chromium.org/2358063002/diff/60001/components/test/BUILD.g...
> components/test/BUILD.gn:5: static_library("test_support") {
> I think you want a source_set here and not a static_library.  If I'm wrong,
> please explain why :)

Done) Can we merge?

[Peter Kasting, 2016-10-19 18:04:47]

LGTM

[dyaroshev, 2016-10-19 18:27:38]

The CQ bit was checked by dyaroshev@yandex-team.ru

[commit-bot: I haz the power, 2016-10-19 18:28:10]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-19 18:32:50]

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

[commit-bot: I haz the power, 2016-10-19 18:32:51]

Try jobs failed on following builders:
  ios-simulator on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator/bui...)
  mac_chromium_compile_dbg_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)

[dyaroshev, 2016-10-20 10:22:18]

The CQ bit was checked by dyaroshev@yandex-team.ru

[dyaroshev, 2016-10-20 10:22:18]

The patchset sent to the CQ was uploaded after l-g-t-m from
pkasting@chromium.org
Link to the patchset: https://codereview.chromium.org/2358063002/#ps100001
(title: "Fixing dependencies")

[commit-bot: I haz the power, 2016-10-20 10:22:29]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-20 10:28:05]

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

[commit-bot: I haz the power, 2016-10-20 10:28:06]

Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[dyaroshev, 2016-10-20 22:56:24]

@jochen - I really need your review for this cl.

[dyaroshev, 2016-10-20 22:56:31]

The CQ bit was checked by dyaroshev@yandex-team.ru

[dyaroshev, 2016-10-20 22:56:32]

The patchset sent to the CQ was uploaded after l-g-t-m from
pkasting@chromium.org
Link to the patchset: https://codereview.chromium.org/2358063002/#ps120001
(title: "Fixing android dependencies")

[commit-bot: I haz the power, 2016-10-20 22:57:14]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-20 23:03:09]

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

[commit-bot: I haz the power, 2016-10-20 23:03:11]

Try jobs failed on following builders:
  ios-device on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device/builds...)
  ios-simulator on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator/bui...)

[dyaroshev, 2016-10-21 06:32:48]

The CQ bit was checked by dyaroshev@yandex-team.ru

[dyaroshev, 2016-10-21 06:32:49]

The patchset sent to the CQ was uploaded after l-g-t-m from
pkasting@chromium.org
Link to the patchset: https://codereview.chromium.org/2358063002/#ps140001
(title: "Fixing ios dependencies")

[commit-bot: I haz the power, 2016-10-21 06:33:12]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-21 07:06:11]

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

[commit-bot: I haz the power, 2016-10-21 07:06:12]

Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[jochen (gone - plz use gerrit), 2016-10-21 11:02:20]

can you please format the CL description according to
chromium.org/developers/contributing-code#Writing_change_list_descriptions

right now, the commit would be titled "This patch".

[dyaroshev, 2016-10-21 11:04:59]

Description was changed from

==========
This patch
1 - creates a complete main for components perftests
2 - decides on using testing/perf for components_perftests

BUG=648992
==========

to

==========
Preparing components_perftests for adding more tests.

1 - creating a complete main for components perftests
2 - deciding on using testing/perf for components_perftests

BUG=648992
==========

[dyaroshev, 2016-10-21 11:05:40]

Description was changed from

==========
Preparing components_perftests for adding more tests.

1 - creating a complete main for components perftests
2 - deciding on using testing/perf for components_perftests

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

1 - creating a complete main for components perftests
2 - deciding on using testing/perf for components_perftests

BUG=648992
==========

[dyaroshev, 2016-10-21 11:05:49]

On 2016/10/21 11:02:20, jochen wrote:
> can you please format the CL description according to
> chromium.org/developers/contributing-code#Writing_change_list_descriptions
> 
> right now, the commit would be titled "This patch".

Done

[dyaroshev, 2016-10-21 11:06:58]

Description was changed from

==========
Preparing components_perftests for adding more benchmarks.

1 - creating a complete main for components perftests
2 - deciding on using testing/perf for components_perftests

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from components
unittests main and switches components perftests to using testing/perf instead
of base/perf.

BUG=648992
==========

[dyaroshev, 2016-10-21 11:07:21]

Description was changed from

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from components
unittests main and switches components perftests to using testing/perf instead
of base/perf.

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

[dyaroshev, 2016-10-21 11:07:37]

Description was changed from

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

[Alexei Svitkine (slow), 2016-10-21 19:01:46]

Drive-by nit

https://codereview.chromium.org/2358063002/diff/140001/components/visitedlink...
File components/visitedlink/test/visitedlink_perftest.cc (right):

https://codereview.chromium.org/2358063002/diff/140001/components/visitedlink...
components/visitedlink/test/visitedlink_perftest.cc:31: explicit
TimeLogger(std::string test_name);
Nit: const&

[Peter Kasting, 2016-10-21 19:45:21]

https://codereview.chromium.org/2358063002/diff/140001/components/visitedlink...
File components/visitedlink/test/visitedlink_perftest.cc (right):

https://codereview.chromium.org/2358063002/diff/140001/components/visitedlink...
components/visitedlink/test/visitedlink_perftest.cc:31: explicit
TimeLogger(std::string test_name);
On 2016/10/21 19:01:46, Alexei Svitkine (slow) wrote:
> Nit: const&

Can't, he's doing std::move() below to move into the member.

[jochen (gone - plz use gerrit), 2016-10-24 07:24:59]

lgtm

[dyaroshev, 2016-10-24 08:49:09]

The CQ bit was checked by dyaroshev@yandex-team.ru

[commit-bot: I haz the power, 2016-10-24 08:49:20]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-24 11:11:24]

Description was changed from

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

[commit-bot: I haz the power, 2016-10-24 11:11:26]

Committed patchset #8 (id:140001)

[commit-bot: I haz the power, 2016-10-24 11:15:40]

Description was changed from

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from 
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992
==========

to

==========
Preparing components_perftests for adding more benchmarks.

This patch creates a complete main for components perftests from
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992

Committed: https://crrev.com/bd75dc348f2e52e1fc3e1b26b6c7528329657d65
Cr-Commit-Position: refs/heads/master@{#427043}
==========

[commit-bot: I haz the power, 2016-10-24 11:15:42]

Patchset 8 (id:??) landed as
https://crrev.com/bd75dc348f2e52e1fc3e1b26b6c7528329657d65
Cr-Commit-Position: refs/heads/master@{#427043}