Added isolates for remaining tests on GPU bots.

Added isolates for remaining tests on GPU bots.

These represent the non-Telemetry tests currently running on these bots.

Tested locally on Linux. Will test on other platforms after the initial
commit and update the isolates as necessary.

BUG=321878
R=maruel@chromium.org
TBR=maruel@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=243388

[Ken Russell (switch to Gerrit), 2013-12-31 01:03:49]

FYI (TBR'ing to maruel per earlier discussion, submitting to CQ now)

[commit-bot: I haz the power, 2013-12-31 01:04:12]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/kbr@chromium.org/101783011/1

[commit-bot: I haz the power, 2013-12-31 02:08:11]

Retried try job too often on mac for step(s) compile
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac&number...

[Ken Russell (switch to Gerrit), 2014-01-03 00:57:34]

+vadimsh, csharp

I'd appreciate any feedback.

Resent to try servers after fixes. If these look good I'll send to the CQ.

[Vadim Sh., 2014-01-03 01:06:48]

Can you please mention swarming roll in CL description (including list of
commits rolled)?

For instance like here: https://codereview.chromium.org/113553003

[Vadim Sh., 2014-01-03 01:17:44]

https://codereview.chromium.org/101783011/diff/260001/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/260001/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:18: 'conditions': [
I think isolate.py is not happy with this structure (based on isolation errors
on try bots).

Afaik, isolate.py doesn't support nested conditions. (Does gyp support them? My
gyp knowledge is ~= 0).

[Ken Russell (switch to Gerrit), 2014-01-03 02:13:59]

On 2014/01/03 01:06:48, Vadim Sh. wrote:
> Can you please mention swarming roll in CL description (including list of
> commits rolled)?
> 
> For instance like here: https://codereview.chromium.org/113553003

Done.

[Ken Russell (switch to Gerrit), 2014-01-03 02:15:48]

https://codereview.chromium.org/101783011/diff/260001/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/260001/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:18: 'conditions': [
On 2014/01/03 01:17:44, Vadim Sh. wrote:
> I think isolate.py is not happy with this structure (based on isolation errors
> on try bots).
> 
> Afaik, isolate.py doesn't support nested conditions. (Does gyp support them?
My
> gyp knowledge is ~= 0).

You're absolutely right. I forgot to re-test locally after making this change.

There were a few other problems in this isolate, and it apparently triggered a
bug in isolate.py which I've worked around. Let's see if the latest round of try
jobs run successfully.

[Ken Russell (switch to Gerrit), 2014-01-04 01:09:36]

+ananta

[Ken Russell (switch to Gerrit), 2014-01-04 17:11:54]

Please review. Thanks.

High-level question: why were these new targets built during the try jobs? Are
they implicitly referenced from the "All" target? The new targets are not
explicitly referenced anywhere in the gyp files. I would like to avoid building
all of the GPU targets on the non-GPU builders (for the time being --
eventually, they will be merged).

[M-A Ruel, 2014-01-06 19:45:01]

Try jobs specifying a test filter 'compile' will compile everything. This means
that all targets will be built, independent of if it makes sense or not. This is
likely the problem here.

https://codereview.chromium.org/101783011/diff/700002/chrome/angle_unittests....
File chrome/angle_unittests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/angle_unittests....
chrome/angle_unittests.isolate:12: 'command': [
Put it before 'isolate_dependency_tracked' to keep keys sorted.

https://codereview.chromium.org/101783011/diff/700002/chrome/content_gl_tests...
File chrome/content_gl_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/content_gl_tests...
chrome/content_gl_tests.isolate:15: 'isolate_dependency_tracked': [
Sort.

https://codereview.chromium.org/101783011/diff/700002/chrome/gl_tests.isolate
File chrome/gl_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/gl_tests.isolate...
chrome/gl_tests.isolate:9: 'isolate_dependency_tracked': [
same

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:10: 'isolate_dependency_tracked': [
same

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:20: ['(OS=="linux" or OS=="mac" or OS=="win")
and internal_gles2_conform_tests==1', {
Oh, isolate's parser is likely too dumb to understand parenthesis. Is it?

I'd say for now it's fine if it is broken on checkouts without src-internal.

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
File chrome/tab_capture_performance_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
chrome/tab_capture_performance_tests.isolate:9: # TODO(kbr): add Chrome OS.
I don't think these todo are useful, OS=="linux" is set for ChromeOS.

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
chrome/tab_capture_performance_tests.isolate:12: 'isolate_dependency_tracked': [
sort

[Ken Russell (switch to Gerrit), 2014-01-06 23:35:12]

https://codereview.chromium.org/101783011/diff/700002/chrome/angle_unittests....
File chrome/angle_unittests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/angle_unittests....
chrome/angle_unittests.isolate:12: 'command': [
On 2014/01/06 19:45:02, M-A Ruel wrote:
> Put it before 'isolate_dependency_tracked' to keep keys sorted.

Done.

https://codereview.chromium.org/101783011/diff/700002/chrome/content_gl_tests...
File chrome/content_gl_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/content_gl_tests...
chrome/content_gl_tests.isolate:15: 'isolate_dependency_tracked': [
On 2014/01/06 19:45:02, M-A Ruel wrote:
> Sort.

Done.

https://codereview.chromium.org/101783011/diff/700002/chrome/gl_tests.isolate
File chrome/gl_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/gl_tests.isolate...
chrome/gl_tests.isolate:9: 'isolate_dependency_tracked': [
On 2014/01/06 19:45:02, M-A Ruel wrote:
> same

Done.

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:10: 'isolate_dependency_tracked': [
On 2014/01/06 19:45:02, M-A Ruel wrote:
> same

Done.

https://codereview.chromium.org/101783011/diff/700002/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:20: ['(OS=="linux" or OS=="mac" or OS=="win")
and internal_gles2_conform_tests==1', {
On 2014/01/06 19:45:02, M-A Ruel wrote:
> Oh, isolate's parser is likely too dumb to understand parenthesis. Is it?
> 
> I'd say for now it's fine if it is broken on checkouts without src-internal.

This change fixed the try job failure, so it seems to be working.

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
File chrome/tab_capture_performance_tests.isolate (right):

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
chrome/tab_capture_performance_tests.isolate:9: # TODO(kbr): add Chrome OS.
On 2014/01/06 19:45:02, M-A Ruel wrote:
> I don't think these todo are useful, OS=="linux" is set for ChromeOS.

OK, I didn't know that. Thanks. I removed them throughout the CL.

https://codereview.chromium.org/101783011/diff/700002/chrome/tab_capture_perf...
chrome/tab_capture_performance_tests.isolate:12: 'isolate_dependency_tracked': [
On 2014/01/06 19:45:02, M-A Ruel wrote:
> sort

Done.

[M-A Ruel, 2014-01-07 00:31:45]

lgtm with a follow up question, no need to wait for me to commit.

https://codereview.chromium.org/101783011/diff/860001/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/860001/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:31: # This no-op branch should not be
necessary but seems to be.
What happened otherwise? I consider this a bug in isolate.py if it is needed.

[M-A Ruel, 2014-01-07 00:32:08]

lgtm with a follow up question, no need to wait for me to commit.

[Ken Russell (switch to Gerrit), 2014-01-07 00:42:53]

https://codereview.chromium.org/101783011/diff/860001/chrome/gles2_conform_te...
File chrome/gles2_conform_test.isolate (right):

https://codereview.chromium.org/101783011/diff/860001/chrome/gles2_conform_te...
chrome/gles2_conform_test.isolate:31: # This no-op branch should not be
necessary but seems to be.
On 2014/01/07 00:31:46, M-A Ruel wrote:
> What happened otherwise? I consider this a bug in isolate.py if it is needed.

Issue 331957 filed about the error which occurs if it isn't present.

[commit-bot: I haz the power, 2014-01-07 20:00:03]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/kbr@chromium.org/101783011/860001

[commit-bot: I haz the power, 2014-01-07 21:36:19]

Retried try job too often on win_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[Ken Russell (switch to Gerrit), 2014-01-07 21:40:33]

Committed patchset #10 manually as r243388 (presubmit successful).