Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.

Committed: https://crrev.com/05a1cb791558af2b5ef244ffbf2681958880f8ed
Cr-Commit-Position: refs/heads/master@{#384396}

[Charlie Reis, 2016-03-31 20:52:43]

creis@chromium.org changed reviewers:
+ alexmos@chromium.org

[Charlie Reis, 2016-03-31 20:52:44]

Alex, looks like your comment in the previous CL was correct about flakiness:
https://codereview.chromium.org/1835833002/diff/80001/content/browser/site_pe...

Unfortunately, the routing ID check doesn't work if the unload finishes first. 
(Easy to repro if you remove the ExecuteScript call.)  What do you think about
skipping it for now to get the bot green, until we find a more reliable way to
test this?  As you mentioned in your comment, we already have some coverage for
that property in the other tests, and the test should still usually fail by
crashing if a regression is introduced.

[alexmos, 2016-03-31 21:11:00]

On 2016/03/31 20:52:44, Charlie Reis wrote:
> Alex, looks like your comment in the previous CL was correct about flakiness:
>
https://codereview.chromium.org/1835833002/diff/80001/content/browser/site_pe...
> 
> Unfortunately, the routing ID check doesn't work if the unload finishes first.

> (Easy to repro if you remove the ExecuteScript call.)  What do you think about
> skipping it for now to get the bot green, until we find a more reliable way to
> test this?  As you mentioned in your comment, we already have some coverage
for
> that property in the other tests, and the test should still usually fail by
> crashing if a regression is introduced.

Ah, yes indeed.  And in retrospect it's unsurprising that DrMemory hit this
given that it's so slow.  LGTM, I think this is fine to unflake the test.  I
guess we could eventually try observing the subframe creation and check whether
the main frame is still pending deletion before checking the RVH routing IDs, or
something like that, but I agree it's not worth the effort now.

[Charlie Reis, 2016-03-31 21:12:13]

The CQ bit was checked by creis@chromium.org

[commit-bot: I haz the power, 2016-03-31 21:12:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1851633002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1851633002/1

[Charlie Reis, 2016-03-31 21:13:09]

Description was changed from

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.
==========

to

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.
==========

[commit-bot: I haz the power, 2016-03-31 21:26:35]

Description was changed from

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.
==========

to

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.
==========

[commit-bot: I haz the power, 2016-03-31 21:26:35]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-03-31 21:28:54]

Description was changed from

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.
==========

to

==========
Fix flaky DontReusePendingDeleteRenderViewHostForSubframe test.

BUG=599344
TEST=Test goes green on DrMemory bot.

Committed: https://crrev.com/05a1cb791558af2b5ef244ffbf2681958880f8ed
Cr-Commit-Position: refs/heads/master@{#384396}
==========

[commit-bot: I haz the power, 2016-03-31 21:28:55]

Patchset 1 (id:??) landed as
https://crrev.com/05a1cb791558af2b5ef244ffbf2681958880f8ed
Cr-Commit-Position: refs/heads/master@{#384396}