Roll WebGL conformance tests to r134617.

Roll WebGL conformance tests to r134617.

BUG=125679
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=134718

[apatrick_chromium, 2012-05-01 01:14:34]

Failures added based on gpu bot try results after WebGL conformance test roll.

[Ken Russell (switch to Gerrit), 2012-05-01 01:43:59]

A comment on the addition of one of the tests.

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
File chrome/test/gpu/webgl_conformance_test_list_autogen.h (right):

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
chrome/test/gpu/webgl_conformance_test_list_autogen.h:171:
"conformance/context/context-creation-and-destruction.html");
This one might be problematic to add right now. It is known to cause flakiness;
see http://crbug.com/124238 . I think we should mark it as a known failure for
the moment (which I believe will cause it to be removed from this test list).

[Zhenyao Mo, 2012-05-01 17:25:29]

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
File chrome/test/gpu/webgl_conformance_test_list_autogen.h (right):

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
chrome/test/gpu/webgl_conformance_test_list_autogen.h:171:
"conformance/context/context-creation-and-destruction.html");
On 2012/05/01 01:43:59, kbr wrote:
> This one might be problematic to add right now. It is known to cause
flakiness;
> see http://crbug.com/124238 . I think we should mark it as a known failure for
> the moment (which I believe will cause it to be removed from this test list).

It is already marked as FAIL in this CL in the test_expectations.txt

[Ken Russell (switch to Gerrit), 2012-05-01 17:48:03]

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
File chrome/test/gpu/webgl_conformance_test_list_autogen.h (right):

https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
chrome/test/gpu/webgl_conformance_test_list_autogen.h:171:
"conformance/context/context-creation-and-destruction.html");
On 2012/05/01 17:25:29, Zhenyao Mo wrote:
> On 2012/05/01 01:43:59, kbr wrote:
> > This one might be problematic to add right now. It is known to cause
> flakiness;
> > see http://crbug.com/124238 . I think we should mark it as a known failure
for
> > the moment (which I believe will cause it to be removed from this test
list).
> 
> It is already marked as FAIL in this CL in the test_expectations.txt

I thought that if the test was marked FAIL then it wasn't added to the test
list. Or is it just skipped during the running of the test?

[Zhenyao Mo, 2012-05-01 17:51:04]

On 2012/05/01 17:48:03, kbr wrote:
>
https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
> File chrome/test/gpu/webgl_conformance_test_list_autogen.h (right):
> 
>
https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
> chrome/test/gpu/webgl_conformance_test_list_autogen.h:171:
> "conformance/context/context-creation-and-destruction.html");
> On 2012/05/01 17:25:29, Zhenyao Mo wrote:
> > On 2012/05/01 01:43:59, kbr wrote:
> > > This one might be problematic to add right now. It is known to cause
> > flakiness;
> > > see http://crbug.com/124238 . I think we should mark it as a known failure
> for
> > > the moment (which I believe will cause it to be removed from this test
> list).
> > 
> > It is already marked as FAIL in this CL in the test_expectations.txt
> 
> I thought that if the test was marked FAIL then it wasn't added to the test
> list. Or is it just skipped during the running of the test?

If a test is marked as anything other than PASS, then we simply return early
without truly running the test.

[Ken Russell (switch to Gerrit), 2012-05-01 17:53:12]

On 2012/05/01 17:51:04, Zhenyao Mo wrote:
> On 2012/05/01 17:48:03, kbr wrote:
> >
>
https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
> > File chrome/test/gpu/webgl_conformance_test_list_autogen.h (right):
> > 
> >
>
https://chromiumcodereview.appspot.com/10261022/diff/3001/chrome/test/gpu/web...
> > chrome/test/gpu/webgl_conformance_test_list_autogen.h:171:
> > "conformance/context/context-creation-and-destruction.html");
> > On 2012/05/01 17:25:29, Zhenyao Mo wrote:
> > > On 2012/05/01 01:43:59, kbr wrote:
> > > > This one might be problematic to add right now. It is known to cause
> > > flakiness;
> > > > see http://crbug.com/124238 . I think we should mark it as a known
failure
> > for
> > > > the moment (which I believe will cause it to be removed from this test
> > list).
> > > 
> > > It is already marked as FAIL in this CL in the test_expectations.txt
> > 
> > I thought that if the test was marked FAIL then it wasn't added to the test
> > list. Or is it just skipped during the running of the test?
> 
> If a test is marked as anything other than PASS, then we simply return early
> without truly running the test.

OK, I see. LGTM; sorry for the delay.