Add typ boilerplate to Telemetry-based GPU tests.

Add typ boilerplate to Telemetry-based GPU tests.

BUG=636153
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel

[Ken Russell (switch to Gerrit), 2017-01-07 01:43:05]

Description was changed from

==========
Add typ boilerplate to Telemetry-based GPU tests.

BUG=636153
==========

to

==========
Add typ boilerplate to Telemetry-based GPU tests.

BUG=636153
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel
==========

[Ken Russell (switch to Gerrit), 2017-01-07 01:45:47]

kbr@chromium.org changed reviewers:
+ dpranke@chromium.org, nednguyen@google.com

[Ken Russell (switch to Gerrit), 2017-01-07 01:45:47]

Ned, Dirk: here's the boilerplate requested in
https://codereview.chromium.org/2590623002/ (informed by
https://codereview.chromium.org/2562913003 ) which will let the GPU tests handle
the switch to typ.

It doesn't work correctly yet. run_gpu_integration_test.py ends up discovering
all of the unit tests in this directory, which is incorrect behavior. Dirk and I
looked a little at run_browser_tests.py and something is wrong with the typ
setup, but I'm not sure exactly what.

Appreciate your help debugging this. Thanks.

[Dirk Pranke, 2017-01-09 03:14:55]

I looked a bit at this. after applying 2590623002 to tip-of-tree catapult and
applying this patch to chromium src, I think I was able to reproduce
what we were seeing on Friday. However, I'm not entirely sure how things are
expected to work, and so it's a bit difficult to say what's wrong.

I think the answer is that I need to actually review the catapult patch in
detail because the integration is not done the way I'd expect it to be.

I'll post some comments there.

[nednguyen, 2017-01-19 22:40:49]

On 2017/01/09 03:14:55, Dirk Pranke wrote:
> I looked a bit at this. after applying 2590623002 to tip-of-tree catapult and
> applying this patch to chromium src, I think I was able to reproduce
> what we were seeing on Friday. However, I'm not entirely sure how things are
> expected to work, and so it's a bit difficult to say what's wrong.
> 
> I think the answer is that I need to actually review the catapult patch in
> detail because the integration is not done the way I'd expect it to be.
> 
> I'll post some comments there.

Sorry, with the latest patch in https://codereview.chromium.org/2590623002/#, I
think the issue should be resolved.

[nednguyen, 2017-01-23 18:03:25]

lgtm

[Ken Russell (switch to Gerrit), 2017-01-23 18:48:26]

On 2017/01/23 18:03:25, nednguyen wrote:
> lgtm

Thanks Ned. If I understand this correctly, this code needs to be integrated
along with your Catapult roll which switches these tests to typ, so please feel
free to take this CL, integrate it into one of yours, and TBR= it to me.

[nednguyen, 2017-01-23 19:06:43]

On 2017/01/23 18:48:26, Ken Russell OOO-till-Jan-27 wrote:
> On 2017/01/23 18:03:25, nednguyen wrote:
> > lgtm
> 
> Thanks Ned. If I understand this correctly, this code needs to be integrated
> along with your Catapult roll which switches these tests to typ, so please
feel
> free to take this CL, integrate it into one of yours, and TBR= it to me.

Adding the boiler-plate code doesn't break existing tests, so you can just land
this as-is, I think

[nednguyen, 2017-01-23 19:06:52]

The CQ bit was checked by nednguyen@google.com to run a CQ dry run

[commit-bot: I haz the power, 2017-01-23 19:07:24]

Dry run: 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, 2017-01-23 19:16:47]

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

[commit-bot: I haz the power, 2017-01-23 19:16:48]

Dry run: 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...)

[Ken Russell (switch to Gerrit), 2017-01-23 19:34:52]

On 2017/01/23 19:06:43, nednguyen wrote:
> On 2017/01/23 18:48:26, Ken Russell OOO-till-Jan-27 wrote:
> > On 2017/01/23 18:03:25, nednguyen wrote:
> > > lgtm
> > 
> > Thanks Ned. If I understand this correctly, this code needs to be integrated
> > along with your Catapult roll which switches these tests to typ, so please
> feel
> > free to take this CL, integrate it into one of yours, and TBR= it to me.
> 
> Adding the boiler-plate code doesn't break existing tests, so you can just
land
> this as-is, I think

Doesn't look like it.

[nednguyen, 2017-01-23 20:30:35]

https://codereview.chromium.org/2616253002/diff/1/content/test/gpu/run_gpu_in...
File content/test/gpu/run_gpu_integration_test.py (left):

https://codereview.chromium.org/2616253002/diff/1/content/test/gpu/run_gpu_in...
content/test/gpu/run_gpu_integration_test.py:28: options =
browser_test_runner.TestRunOptions()
Oh, so this seems to be the API breakage

[nednguyen, 2017-01-30 15:10:57]

On 2017/01/23 20:30:35, nednguyen wrote:
>
https://codereview.chromium.org/2616253002/diff/1/content/test/gpu/run_gpu_in...
> File content/test/gpu/run_gpu_integration_test.py (left):
> 
>
https://codereview.chromium.org/2616253002/diff/1/content/test/gpu/run_gpu_in...
> content/test/gpu/run_gpu_integration_test.py:28: options =
> browser_test_runner.TestRunOptions()
> Oh, so this seems to be the API breakage

Hmmh, another problem is that content/test/gpu/run_unittests.py picks up the
load_tests method in the module & fails because they are not designed for normal
unittests. I need to add some more logic to browser_test_runner so that they
bails out early if the load_tests is not run up by the browser_test_runner.