Assigning to lukasza@ since they were the last one to touch the test file in question. Feel free to re-assign, just don't want this to get lost in an unassigned state.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8fc21e7d83245f7445f09256dd30a2db951ce805

commit 8fc21e7d83245f7445f09256dd30a2db951ce805
Author: bsheedy <bsheedy@chromium.org>
Date: Wed Sep 26 01:17:14 2018

Disable two flaky browser tests

Disables SitePerProcessBrowserTest.ScrollBubblingFromOOPIFTest and
HasBeforeUnloadHandlerTest.NonEmptyBeforeUnloadDetected, as they're
flaking on the CQ and preventing some CLs from landing.

TBR=oysteine@chromium.org

Bug:  889301 , 889304
Change-Id: I7b0c64135eede3a85877d718268e6b047ef57340
Reviewed-on: https://chromium-review.googlesource.com/1244887
Commit-Queue: Brian Sheedy <bsheedy@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#594173}
[modify] https://crrev.com/8fc21e7d83245f7445f09256dd30a2db951ce805/chrome/browser/resource_coordinator/discard_before_unload_helper_browsertest.cc
[modify] https://crrev.com/8fc21e7d83245f7445f09256dd30a2db951ce805/content/browser/site_per_process_browsertest.cc

Adding some more folks ...

I'm guessing this is a recent regression ... could someone with direct access to their workstation try and bisect it?

Flakiness in SitePerProcessBrowserTest.ScrollBubblingFromOOPIFTest seems to be a dupe of  issue 877496 ?  Please note that we disable the test via

//testing/buildbot/filters/chromeos.single_process_mash.content_browsertests.filter:
    -SitePerProcessBrowserTest.*

(this exception was briefly removed in r594046 and then readded by the revert in r594164).

Based on the above, I think we should revert at least the part of r594173 (from #c3 above) that *additionally* disables SitePerProcessBrowserTest.ScrollBubblingFromOOPIFTest via //content/browser/site_per_process_browsertest.cc

Let's focus this bug on SitePerProcessBrowserTest.ScrollBubblingFromOOPIFTest and keep discussing HasBeforeUnloadHandlerTest.NonEmptyBeforeUnloadDetected in issue 889304.

I enabled the test for content_browsertests --enable-features=SingleProcessMash yesterday. This is now reverted in https://chromium-review.googlesource.com/c/chromium/src/+/1244661. I think we could revert part of r594173 since a black list filter should be enough.

Not sure why it fails with single process mash tho. It passes locally.

I will keep it in the single process mash blacklist filter until this bug is fixed.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/79f42964267bef4cbffafbbdde9668d4c22689b5

commit 79f42964267bef4cbffafbbdde9668d4c22689b5
Author: Lukasz Anforowicz <lukasza@chromium.org>
Date: Wed Sep 26 18:31:43 2018

Reenable the SitePerProcessBrowserTest.ScrollBubblingFromOOPIFTest test.

As explained in  https://crbug.com/889301#c5 , the test failure in
single_process_mash_content_browsertests on linux-chromeos-rel is
already expected and covered by an entry in
//testing/buildbot/filters/chromeos.single_process_mash.content_browsertests.filter:
    -SitePerProcessBrowserTest.*

Therefore, there is no point in additionally disabling the test via
//content/browser/site_per_process_browsertest.cc

Bug:  889301 
Change-Id: Id86d39792939c06b2972443ea8caecad774ee6e1
Reviewed-on: https://chromium-review.googlesource.com/1246285
Reviewed-by: Antoine Labour <piman@chromium.org>
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Cr-Commit-Position: refs/heads/master@{#594400}
[modify] https://crrev.com/79f42964267bef4cbffafbbdde9668d4c22689b5/content/browser/site_per_process_browsertest.cc

I would like to keep this this open to track the failure with single process mash as I am putting it in the filter (https://chromium-review.googlesource.com/c/chromium/src/+/1246152).

Currently, this is the only one fails in SitePerProcessBrowserTest suite.

Related to location transform, dup into 884360.