gpu: Disable crashing WebGL tests on Android

gpu: Disable crashing WebGL tests on Android

BUG=341698
TBR=kbr@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=251755

[Sami, 2014-02-18 12:22:26]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2014-02-18 12:22:34]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/167313005/1

[commit-bot: I haz the power, 2014-02-18 12:37:54]

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

[commit-bot: I haz the power, 2014-02-18 12:37:54]

Try jobs failed on following builders: linux_chromium_rel

[Sami, 2014-02-18 12:59:22]

Committed patchset #1 manually as r251755 (presubmit successful).

[Ken Russell (switch to Gerrit), 2014-02-19 22:18:10]

not lgtm

Why is this being done rather than fixing the underlying failures? These tests
are broken, and this has violated the WebGL conformance process.

[Sami, 2014-02-20 11:33:39]

On 2014/02/19 22:18:10, Ken Russell wrote:
> not lgtm
> 
> Why is this being done rather than fixing the underlying failures? These tests
> are broken, and this has violated the WebGL conformance process.

Sorry I didn't better explain what happened here. These tests were already
crashing earlier as a result of the Android SurfaceTexture binding changes.
crbug.com/341698 was opened to track the fix. However, the crashes weren't
turning the bot red because of crbug.com/342881. Now that Brandon landed the fix
for that bug, the bot went red and actually stopped running all the tests
following the crashing ones. So, to make sure we still have decent coverage I
decided to disable the crashing tests. The actual fix is still coming in
crbug.com/341698.

Feel free to revert if you disagree.

[bajones, 2014-02-20 16:58:03]

On 2014/02/20 11:33:39, Sami wrote:
> ...the bot went red and actually stopped running all the tests
> following the crashing ones.

That's not expected behavior. The test should fail upon crash, but we should be
restarting the browser and continuing on with the next test.
I'll look into this today and try to fix that logic.

[Ken Russell (switch to Gerrit), 2014-02-20 23:19:47]

On 2014/02/20 11:33:39, Sami wrote:
> On 2014/02/19 22:18:10, Ken Russell wrote:
> > not lgtm
> > 
> > Why is this being done rather than fixing the underlying failures? These
tests
> > are broken, and this has violated the WebGL conformance process.
> 
> Sorry I didn't better explain what happened here. These tests were already
> crashing earlier as a result of the Android SurfaceTexture binding changes.
> crbug.com/341698 was opened to track the fix. However, the crashes weren't
> turning the bot red because of crbug.com/342881. Now that Brandon landed the
fix
> for that bug, the bot went red and actually stopped running all the tests
> following the crashing ones. So, to make sure we still have decent coverage I
> decided to disable the crashing tests. The actual fix is still coming in
> crbug.com/341698.
> 
> Feel free to revert if you disagree.

Given that explanation the CL is acceptable, but in the future, please document
this sort of stuff ahead of time, and in a more permanent place (i.e., a bug,
not a CL). Also, please make sure there's a bug filed and assigned to you if you
intend to turn them back on once the underlying issue is fixed.

[Ken Russell (switch to Gerrit), 2014-02-20 23:20:08]

lgtm given sami's explanation.