DumpAccessibilityTreeTest* are failing. 
https://chromium-review.googlesource.com/c/chromium/src/+/1072875
https://chromium-review.googlesource.com/c/chromium/src/+/1067362

These two CLs touched accessibility code last. Would you mind take a look?

The first one is a more significant change.

What's weird, though, is that it only seems to be failing on Win10 x64 dbg and not other bots - and that nearly all accessibility tests are failing.

I guess let's try a speculative revert?

Revert going in: https://chromium-review.googlesource.com/c/chromium/src/+/1074050

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/50da791e6d9ceec220bcf98c3abc16fa424fb009

commit 50da791e6d9ceec220bcf98c3abc16fa424fb009
Author: Dominic Mazzoni <dmazzoni@chromium.org>
Date: Fri May 25 22:42:23 2018

Revert "Re-land: Avoid serializing the same accessibility node twice in the same message"

Speculative revert, see build failures on Win10 Tests x64 (dbg):
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20Tests%20x64%20%28dbg%29/803

Bug: http://crbug.com/846837

This reverts commit 983e926ba600204c1a0a4d1ad3428136c1ff8058.

Reason for revert: <INSERT REASONING HERE>

Original change's description:
> Re-land: Avoid serializing the same accessibility node twice in the same message
> 
> Originally landed: r560165, crrev.com/c/1063007
> Reverted: r561246, crrev.com/c/1069891
> 
> TBR=dtseng@chromium.org
> Bug: 651614, 845778
> 
> Change-Id: I77b54de28cb0c4a231b5b0758ba4708100d27fff
> Reviewed-on: https://chromium-review.googlesource.com/1072875
> Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#561913}

TBR=dmazzoni@chromium.org,dtseng@chromium.org

Change-Id: I5cbf8006e4917a9c008143ae4b746543d2b038c2
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 651614, 845778, 846837
Reviewed-on: https://chromium-review.googlesource.com/1074050
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#562043}
[modify] https://crrev.com/50da791e6d9ceec220bcf98c3abc16fa424fb009/chrome/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/50da791e6d9ceec220bcf98c3abc16fa424fb009/chromecast/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/50da791e6d9ceec220bcf98c3abc16fa424fb009/content/renderer/accessibility/render_accessibility_impl.cc
[modify] https://crrev.com/50da791e6d9ceec220bcf98c3abc16fa424fb009/ui/accessibility/ax_tree_serializer.h
[modify] https://crrev.com/50da791e6d9ceec220bcf98c3abc16fa424fb009/ui/accessibility/ax_tree_serializer_unittest.cc

content_browsertests and other tests are still failing on Win10 Tests x64 (dbg). I think it's something else.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/714e3f7dbda4483ac2f6b1646b6a26914990388c

commit 714e3f7dbda4483ac2f6b1646b6a26914990388c
Author: Dominic Mazzoni <dmazzoni@chromium.org>
Date: Sat May 26 06:43:32 2018

Reland "Re-land: Avoid serializing the same accessibility node twice in the same message"

Reverted second time speculatively to see if it was related to
failures on Win10 Tests x64 (dbg), but it didn't seem related.

This reverts commit 50da791e6d9ceec220bcf98c3abc16fa424fb009.

Original change's description:
> Revert "Re-land: Avoid serializing the same accessibility node twice in the same message"
> 
> Speculative revert, see build failures on Win10 Tests x64 (dbg):
> https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20Tests%20x64%20%28dbg%29/803
> 
> Bug: http://crbug.com/846837
> 
> This reverts commit 983e926ba600204c1a0a4d1ad3428136c1ff8058.
> 
> Reason for revert: <INSERT REASONING HERE>
> 
> Original change's description:
> > Re-land: Avoid serializing the same accessibility node twice in the same message
> > 
> > Originally landed: r560165, crrev.com/c/1063007
> > Reverted: r561246, crrev.com/c/1069891
> > 
> > TBR=dtseng@chromium.org
> > Bug: 651614, 845778
> > 
> > Change-Id: I77b54de28cb0c4a231b5b0758ba4708100d27fff
> > Reviewed-on: https://chromium-review.googlesource.com/1072875
> > Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> > Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#561913}
> 
> TBR=dmazzoni@chromium.org,dtseng@chromium.org
> 
> Change-Id: I5cbf8006e4917a9c008143ae4b746543d2b038c2
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: 651614, 845778, 846837
> Reviewed-on: https://chromium-review.googlesource.com/1074050
> Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#562043}

TBR=dmazzoni@chromium.org,dtseng@chromium.org

Change-Id: I3dc8009836eb084271a3ac5239311008dc17105b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: http://crbug.com/846837, 651614, 845778, 846837
Reviewed-on: https://chromium-review.googlesource.com/1074587
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#562118}
[modify] https://crrev.com/714e3f7dbda4483ac2f6b1646b6a26914990388c/chrome/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/714e3f7dbda4483ac2f6b1646b6a26914990388c/chromecast/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/714e3f7dbda4483ac2f6b1646b6a26914990388c/content/renderer/accessibility/render_accessibility_impl.cc
[modify] https://crrev.com/714e3f7dbda4483ac2f6b1646b6a26914990388c/ui/accessibility/ax_tree_serializer.h
[modify] https://crrev.com/714e3f7dbda4483ac2f6b1646b6a26914990388c/ui/accessibility/ax_tree_serializer_unittest.cc

Last failure is https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20Tests%20x64%20%28dbg%29/832

only one CL in bisect. 

But it's not persisting. Maybe a rogue outage :/

From the error message and the content of my change, it doesn't look like related to my CL.

In my change, WindowPerformance::ShouldBufferEventTiming() returning either true or false shouldn't cause crashes externally. From the error message, the crash was not within the function's scope as well. So it's unlikely to be the culprit.

Seems to still be happening:

https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20Tests%20x64%20%28dbg%29/899

 Issue 847649  has been merged into this issue.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/79cdf59c0f60563be32c0429ef8264b3ad16bfca

commit 79cdf59c0f60563be32c0429ef8264b3ad16bfca
Author: Dominic Mazzoni <dmazzoni@chromium.org>
Date: Wed May 30 23:36:17 2018

Revert "Re-land: Avoid serializing the same accessibility node twice in the same message"

This reverts commit 714e3f7dbda4483ac2f6b1646b6a26914990388c.

Reason for revert: reproducible crash: http://crbug.com/847276

Original change's description:
> Reland "Re-land: Avoid serializing the same accessibility node twice in the same message"
>
> Reverted second time speculatively to see if it was related to
> failures on Win10 Tests x64 (dbg), but it didn't seem related.
>
> This reverts commit 50da791e6d9ceec220bcf98c3abc16fa424fb009.
>
> Original change's description:
> > Revert "Re-land: Avoid serializing the same accessibility node twice in the same message"
> >
> > Speculative revert, see build failures on Win10 Tests x64 (dbg):
> > https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20Tests%20x64%20%28dbg%29/803
> >
> > Bug: http://crbug.com/846837
> >
> > This reverts commit 983e926ba600204c1a0a4d1ad3428136c1ff8058.
> >
> > Reason for revert: <INSERT REASONING HERE>
> >
> > Original change's description:
> > > Re-land: Avoid serializing the same accessibility node twice in the same message
> > >
> > > Originally landed: r560165, crrev.com/c/1063007
> > > Reverted: r561246, crrev.com/c/1069891
> > >
> > > TBR=dtseng@chromium.org
> > > Bug: 651614, 845778
> > >
> > > Change-Id: I77b54de28cb0c4a231b5b0758ba4708100d27fff
> > > Reviewed-on: https://chromium-review.googlesource.com/1072875
> > > Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> > > Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> > > Cr-Commit-Position: refs/heads/master@{#561913}
> >
> > TBR=dmazzoni@chromium.org,dtseng@chromium.org
> >
> > Change-Id: I5cbf8006e4917a9c008143ae4b746543d2b038c2
> > No-Presubmit: true
> > No-Tree-Checks: true
> > No-Try: true
> > Bug: 651614, 845778, 846837
> > Reviewed-on: https://chromium-review.googlesource.com/1074050
> > Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> > Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#562043}
>
> TBR=dmazzoni@chromium.org,dtseng@chromium.org
>
> Change-Id: I3dc8009836eb084271a3ac5239311008dc17105b
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: http://crbug.com/846837, 651614, 845778, 846837
> Reviewed-on: https://chromium-review.googlesource.com/1074587
> Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
> Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#562118}

TBR=dmazzoni@chromium.org,dtseng@chromium.org

NOTRY=true

Bug: http://crbug.com/846837, 651614, 845778, 846837, 847276
Change-Id: I342db1180d09bc9eaa1a9c8fc7054fa2ca366eb6
Reviewed-on: https://chromium-review.googlesource.com/1079090
Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org>
Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#563061}
[modify] https://crrev.com/79cdf59c0f60563be32c0429ef8264b3ad16bfca/chrome/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/79cdf59c0f60563be32c0429ef8264b3ad16bfca/chromecast/browser/ui/aura/accessibility/automation_manager_aura.cc
[modify] https://crrev.com/79cdf59c0f60563be32c0429ef8264b3ad16bfca/content/renderer/accessibility/render_accessibility_impl.cc
[modify] https://crrev.com/79cdf59c0f60563be32c0429ef8264b3ad16bfca/ui/accessibility/ax_tree_serializer.h
[modify] https://crrev.com/79cdf59c0f60563be32c0429ef8264b3ad16bfca/ui/accessibility/ax_tree_serializer_unittest.cc

CL reverted, owner assigned --> Taking out of sheriff queue.

https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=content_browsertests&tests=DumpAccessibilityEventsTest

The tests are green except few being yellow. It seems to me that the issue has been fixed?

There were timeouts before Aug 4 where 200+ DumpAccessibilityEventsTests failed.

Probably we should wait for a few days before announcing fixed?