Minor improvements to reftests.

Minor improvements to reftests.

1. If notifyTestComplete is called before the first paint,
give a clear assert instead of a confusing crash.

2. Add runAfterDisplay. This is a copy of the equivalent
file in Blink. Avoids the incorrect 100ms setTimeout. I had
thought there was a deeper bug here, but it appears to just
be the issue from #1 now that we properly pump frames
in testing mode by using a face Surfaces application.

R=abarth@chromium.org

Committed: https://chromium.googlesource.com/external/mojo/+/87cce85d54d6f67aef634b9eda7bb67d897976a3

[abarth-chromium, 2015-01-06 22:48:08]

abarth@chromium.org changed reviewers:
+ abarth@chromium.org

[abarth-chromium, 2015-01-06 22:48:09]

LGTM

https://codereview.chromium.org/826343003/diff/1/sky/tests/harness/reftest.sky
File sky/tests/harness/reftest.sky (right):

https://codereview.chromium.org/826343003/diff/1/sky/tests/harness/reftest.sk...
sky/tests/harness/reftest.sky:7: });
Should we move this code into a ref-test.sky resource file?

https://codereview.chromium.org/826343003/diff/1/sky/tests/resources/run-afte...
File sky/tests/resources/run-after-display.sky (right):

https://codereview.chromium.org/826343003/diff/1/sky/tests/resources/run-afte...
sky/tests/resources/run-after-display.sky:8: window.setTimeout(callback, 1);
Why 1 instead of 0?

[ojan, 2015-01-06 22:54:41]

https://codereview.chromium.org/826343003/diff/1/sky/tests/harness/reftest.sky
File sky/tests/harness/reftest.sky (right):

https://codereview.chromium.org/826343003/diff/1/sky/tests/harness/reftest.sk...
sky/tests/harness/reftest.sky:7: });
On 2015/01/06 22:48:09, abarth wrote:
> Should we move this code into a ref-test.sky resource file?

I considered it and it seemed a bit confusing since it doesn't actually cause
the test to be a reftest or to dump pixels even. The thing that causes it to be
a reftest is the presence of a -expected.sky file.

I don't feel strongly. I'll commit this for now, happy to do this as a followup
if you still think it's a good idea.

https://codereview.chromium.org/826343003/diff/1/sky/tests/resources/run-afte...
File sky/tests/resources/run-after-display.sky (right):

https://codereview.chromium.org/826343003/diff/1/sky/tests/resources/run-afte...
sky/tests/resources/run-after-display.sky:8: window.setTimeout(callback, 1);
On 2015/01/06 at 22:48:09, abarth wrote:
> Why 1 instead of 0?

I just copy-pasted this from the Blink version. I considered changing it, but
didn't see advantages one way or another. I'll change it to 0.

https://codereview.chromium.org/826343003/diff/1/sky/tests/resources/run-afte...
sky/tests/resources/run-after-display.sky:8: window.setTimeout(callback, 1);
On 2015/01/06 at 22:48:09, abarth wrote:
> Why 1 instead of 0?

I just copy-pasted this from the Blink version. I considered changing it, but
didn't see advantages one way or another. I'll change it to 0.

[ojan, 2015-01-06 23:41:20]

Committed patchset #3 (id:40001) manually as
87cce85d54d6f67aef634b9eda7bb67d897976a3 (presubmit successful).