New issue
Advanced search Search tips

Issue 746470 link

Starred by 2 users

Issue metadata

Status: Assigned
Owner:
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux , Android , Windows , Chrome , Mac
Pri: 2
Type: Bug

Blocking:
issue 732400



Sign in to add a comment

validate AggregatedHitTestRegion data received from HitTestAggregator

Project Member Reported by riajiang@chromium.org, Jul 19 2017

Issue description

HitTestQuery needs to validate AggregatedHitTestRegion data received from HitTestAggregator.

 
including checking child_count, flags and making sure frame_sink_id in the window <-> framesink map
Labels: Proj-Mustash-Mus-WS
I recommend a fuzz test.

Comment 4 by varkha@chromium.org, Aug 15 2017

Labels: event-targeting

Comment 5 by sky@chromium.org, Oct 11 2017

Labels: -Pri-3 Pri-2
I occasionally hit this DCHECK when running the browser_test ConstrainedWebDialogBrowserTest.ContentResizeInAutoResizingDialog in --mus. I'm upping the priority.
Cc: sky@chromium.org
This DCHECK (https://cs.chromium.org/chromium/src/components/viz/host/hit_test/hit_test_query.cc?type=cs&q=hit_test_query&l=29)? HitTestQuery would only be used if --use-viz-hit-test is turned on tho?

Comment 7 by sky@chromium.org, Oct 11 2017

I added the comment to this bug as I'm hitting the DCHECK in HostFrameSinkManager::SwitchActiveAggregatedHitTestRegionList[1], which references this bug.

[1] https://cs.chromium.org/chromium/src/components/viz/host/host_frame_sink_manager.cc?type=cs&q=HostFrameSinkManager::SwitchActiveAggregatedHitTestRegionList&sq=package:chromium&l=287 
Ah I see, sorry forgot that I referenced this bug from several places instead of just in HTQ. I'll take a look at that.
Project Member

Comment 9 by bugdroid1@chromium.org, Oct 26 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5d04c4e406b9907cee94ead7bd8170d71f8a37fe

commit 5d04c4e406b9907cee94ead7bd8170d71f8a37fe
Author: Ria Jiang <riajiang@chromium.org>
Date: Thu Oct 26 15:30:22 2017

Drop hit-test data when the corresponding HitTestQuery has been deleted.

In HostFrameSinkManager::InvalidateFrameSinkId, we erase the HTQ
associated with that frame_sink_id. We send that msg to FrameSinkManager
but it can be cross-process and we can still be sending hit-test data
to HostFrameSinkManager/ that HTQ when it has already been deleted at
the same time. Drop the hit-test data in those cases.

Bug: 746470
Test: viz_unittests
Change-Id: I4bc80319e5195f3aed46b8a41872fd31d879f7ba
Reviewed-on: https://chromium-review.googlesource.com/736351
Reviewed-by: kylechar <kylechar@chromium.org>
Commit-Queue: Ria Jiang <riajiang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#511825}
[modify] https://crrev.com/5d04c4e406b9907cee94ead7bd8170d71f8a37fe/components/viz/host/host_frame_sink_manager.cc
[modify] https://crrev.com/5d04c4e406b9907cee94ead7bd8170d71f8a37fe/components/viz/host/host_frame_sink_manager_unittest.cc

Components: -Internals>MUS Internals>Services>WindowService
Project Member

Comment 11 by bugdroid1@chromium.org, Apr 10 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5e5ba789e10b0319d1812467d8ef834f69fd847b

commit 5e5ba789e10b0319d1812467d8ef834f69fd847b
Author: Ria Jiang <riajiang@chromium.org>
Date: Tue Apr 10 00:20:22 2018

Don't crash viz host when shared memory buffer is not valid, crash
viz instead.

When viz host receives invalid shared memory buffers, use the callback
received from content/browser to shut down gpu process.

Bug: 746470
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel
Change-Id: Ic5fc690266aa4acb76d66c006706e9166771c900
Reviewed-on: https://chromium-review.googlesource.com/983002
Commit-Queue: Ria Jiang <riajiang@chromium.org>
Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Reviewed-by: kylechar <kylechar@chromium.org>
Reviewed-by: Ken Buchanan <kenrb@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Cr-Commit-Position: refs/heads/master@{#549340}
[modify] https://crrev.com/5e5ba789e10b0319d1812467d8ef834f69fd847b/components/viz/host/hit_test/hit_test_query.cc
[modify] https://crrev.com/5e5ba789e10b0319d1812467d8ef834f69fd847b/components/viz/host/hit_test/hit_test_query.h
[modify] https://crrev.com/5e5ba789e10b0319d1812467d8ef834f69fd847b/components/viz/host/host_frame_sink_manager.cc
[modify] https://crrev.com/5e5ba789e10b0319d1812467d8ef834f69fd847b/components/viz/host/host_frame_sink_manager.h
[modify] https://crrev.com/5e5ba789e10b0319d1812467d8ef834f69fd847b/content/browser/compositor/viz_process_transport_factory.cc

Labels: -Proj-Mustash-Mus-WS
Deprecating label Proj-Mustash-Mus-WS in favor of Components.
Project Member

Comment 13 by bugdroid1@chromium.org, Apr 30 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9c66e77a86d6c236dd64468bcda563c79d59d970

commit 9c66e77a86d6c236dd64468bcda563c79d59d970
Author: Ria Jiang <riajiang@chromium.org>
Date: Mon Apr 30 21:48:35 2018

Crash gpu process if HitTestQuery received invalid active region index.

Bug: 746470
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel
Change-Id: I474adb47645eb2de52fe4ce368b8baf390d3f574
Reviewed-on: https://chromium-review.googlesource.com/1036102
Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Commit-Queue: Ria Jiang <riajiang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#554890}
[modify] https://crrev.com/9c66e77a86d6c236dd64468bcda563c79d59d970/components/viz/host/hit_test/hit_test_query.cc
[modify] https://crrev.com/9c66e77a86d6c236dd64468bcda563c79d59d970/components/viz/host/host_frame_sink_manager.cc

Labels: Merge-Request-67
CL in #13 is a simple CL to avoid crashing browser when gpu process is compromised, should be safe to merge back in M67.
Pls apply appropriate OSs label. Thank you.
Labels: OS-Android OS-Chrome OS-Linux OS-Mac OS-Windows
Project Member

Comment 17 by sheriffbot@chromium.org, Apr 30 2018

Labels: -Merge-Request-67 Merge-Review-67 Hotlist-Merge-Review
This bug requires manual review: M67 has already been promoted to the beta branch, so this requires manual review
Please contact the milestone owner if you have questions.
Owners: cmasso@(Android), cmasso@(iOS), kbleicher@(ChromeOS), govind@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
NextAction: 2018-05-01
CL in #13 is not made it to canary yet. Pls update the bug with canary result tomorrow.
The NextAction date has arrived: 2018-05-01
Labels: -Merge-Review-67 Merge-Approved-67
Approving merge for cl listed at #13 to M67 branch 3396 based on comments #14 and #20. Please merge ASAP so we can pick it up for tomorrow's beta release. Thank you.
Project Member

Comment 22 by bugdroid1@chromium.org, May 1 2018

Labels: -merge-approved-67 merge-merged-3396
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5cdcde235f6d8cf76ede786e45d1db7e6aef95af

commit 5cdcde235f6d8cf76ede786e45d1db7e6aef95af
Author: Ria Jiang <riajiang@chromium.org>
Date: Tue May 01 18:48:47 2018

Crash gpu process if HitTestQuery received invalid active region index.

Bug: 746470
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel
Change-Id: I474adb47645eb2de52fe4ce368b8baf390d3f574
Reviewed-on: https://chromium-review.googlesource.com/1036102
Reviewed-by: Robert Kroeger <rjkroege@chromium.org>
Commit-Queue: Ria Jiang <riajiang@chromium.org>
Cr-Original-Commit-Position: refs/heads/master@{#554890}(cherry picked from commit 9c66e77a86d6c236dd64468bcda563c79d59d970)
Reviewed-on: https://chromium-review.googlesource.com/1037605
Reviewed-by: Ria Jiang <riajiang@chromium.org>
Cr-Commit-Position: refs/branch-heads/3396@{#417}
Cr-Branched-From: 9ef2aa869bc7bc0c089e255d698cca6e47d6b038-refs/heads/master@{#550428}
[modify] https://crrev.com/5cdcde235f6d8cf76ede786e45d1db7e6aef95af/components/viz/host/hit_test/hit_test_query.cc
[modify] https://crrev.com/5cdcde235f6d8cf76ede786e45d1db7e6aef95af/components/viz/host/host_frame_sink_manager.cc

Labels: -Proj-Mustash Proj-Mash-MultiProcess

Sign in to add a comment