https://chromium-build-logs.appspot.com/gtest_query?gtest_query=LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer

Is this caused by r453752 or r454006? The first failure I can find is after the first of those commits:

https://build.chromium.org/p/chromium.memory/builders/Linux%20TSan%20Tests/builds/1887/steps/cc_unittests/logs/stdio

[ RUN      ] LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest_scroll.cc:2170: Failure
Value of: 4
Expected: num_of_activations_
Which is: 3
../../cc/trees/layer_tree_host_unittest_scroll.cc:2223: Failure
Value of: 4
Expected: num_of_activations_
Which is: 3

But then there's a crash on later runs (the first probably doesn't have DCHECKs enabled):

https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Ftryserver.chromium.linux%2Flinux_chromium_asan_rel_ng%2F320798%2F%2B%2Frecipes%2Fsteps%2Fcc_unittests__with_patch_%2F0%2Flogs%2FLayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer%2F0

[ RUN      ] LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest_scroll.cc:2170: Failure
Value of: 4
Expected: num_of_activations_
Which is: 3
[8930:8946:0301/133705.340220:9813177012:FATAL:scheduler_state_machine.cc(698)] Check failed: !has_pending_tree_ || (settings_.main_frame_before_activation_enabled && commit_has_no_updates).
#0 0x0000005cdd91 __interceptor_backtrace
#1 0x0000033e664c base::debug::StackTrace::StackTrace()
#2 0x00000340d277 logging::LogMessage::~LogMessage()
#3 0x000002ce6df6 cc::SchedulerStateMachine::WillCommit()
#4 0x000002cecae0 cc::SchedulerStateMachine::BeginMainFrameAborted()
#5 0x000002cd8a82 cc::Scheduler::BeginMainFrameAborted()
#6 0x000002f2141e cc::ProxyImpl::BeginMainFrameAbortedOnImpl()
#7 0x000002ed73da _ZN4base8internal7InvokerINS0_9BindStateIMN2cc9ProxyImplEFvNS3_20CommitEarlyOutReasonENS_9TimeTicksENSt3__16vectorINS7_10unique_ptrINS3_11SwapPromiseENS7_14default_deleteISA_EEEENS7_9allocatorISD_EEEEEJNS0_17UnretainedWrapperIS4_EES5_S6_NS0_13PassedWrapperISG_EEEEEFvvEE3RunEPNS0_13BindStateBaseE
#8 0x000003544da9 base::debug::TaskAnnotator::RunTask()
#9 0x00000341b8db base::MessageLoop::RunTask()
#10 0x00000341c896 base::MessageLoop::DeferOrRunPendingTask()
#11 0x00000341defb base::MessageLoop::DoWork()
#12 0x000003427523 base::MessagePumpDefault::Run()
#13 0x00000341aee1 base::MessageLoop::RunHandler()
#14 0x000003468501 base::RunLoop::Run()
#15 0x0000034bd11f base::Thread::Run()
#16 0x0000034bdb83 base::Thread::ThreadMain()
#17 0x0000034a5f43 base::(anonymous namespace)::ThreadFunc()
#18 0x7f98ed067184 start_thread
#19 0x7f98ecb7e37d clone

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d665cfdee84d40a5d3b7cd39bae64c886b24ad6d

commit d665cfdee84d40a5d3b7cd39bae64c886b24ad6d
Author: grt <grt@chromium.org>
Date: Thu Mar 02 11:51:35 2017

Disable LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer due to flakes.

BUG= 697652 
TBR=pdr@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2727213002
Cr-Commit-Position: refs/heads/master@{#454234}

[modify] https://crrev.com/d665cfdee84d40a5d3b7cd39bae64c886b24ad6d/cc/trees/layer_tree_host_unittest_scroll.cc

Detected 3 new flakes for test/step "LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyWAsSBUZsYWtlIk1MYXllclRyZWVIb3N0U2Nyb2xsVGVzdEltcGxTaWRlSW52YWxpZGF0aW9uLlJ1bk11bHRpVGhyZWFkX0RlbGVnYXRpbmdSZW5kZXJlcgw. This message was posted automatically by the chromium-try-flakes app. Since flakiness is ongoing, the issue was moved back into Sheriff Bug Queue (unless already there).

This is also flaky on Android's N5X Swarm builder. I'm going to disable it on Android as well.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/6db6ba9fddd07674c43fc11a9d35822bbdebc1c7

commit 6db6ba9fddd07674c43fc11a9d35822bbdebc1c7
Author: twellington <twellington@chromium.org>
Date: Fri Mar 03 19:24:41 2017

Disable LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer on Android

Flakes on Android and Linux bots.

BUG= 697652 
TBR=pdr@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2724203005
Cr-Commit-Position: refs/heads/master@{#454645}

[modify] https://crrev.com/6db6ba9fddd07674c43fc11a9d35822bbdebc1c7/cc/trees/layer_tree_host_unittest_scroll.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b02a7c1ab2ee8695d2ac3aeb43bd0ff0fc871ccf

commit b02a7c1ab2ee8695d2ac3aeb43bd0ff0fc871ccf
Author: pdr <pdr@chromium.org>
Date: Fri Mar 03 20:07:05 2017

Fully disable LayerTreeHostScrollTestImplSideInvalidation due to crashes

This test is flaky on all platforms. This patch temporarily disables it
until a fix is found.

This test crashes on OSX with:
./out/Debug/cc_unittests --gtest_filter=LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer --gtest_repeat=10

BUG= 697652 
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2734503002
Cr-Commit-Position: refs/heads/master@{#454657}

[modify] https://crrev.com/b02a7c1ab2ee8695d2ac3aeb43bd0ff0fc871ccf/cc/trees/layer_tree_host_unittest_scroll.cc

I bisected this back to the original commit:
https://chromium.googlesource.com/chromium/src/+/6dc91d0e5b2f135eee29fc0e2a506747a18dc439

@khushalsagar, please take a look at this test. It is flaky on all platforms:
./out/Debug/cc_unittests --gtest_filter=LayerTreeHostScrollTestImplSideInvalidation.RunMultiThread_DelegatingRenderer --gtest_repeat=30 --gtest_also_run_disabled_tests

../../cc/trees/layer_tree_host_unittest_scroll.cc:2170: Failure
Value of: 4
Expected: num_of_activations_
Which is: 3
[FATAL:scheduler_state_machine.cc(698)] Check failed: !has_pending_tree_ || (settings_.main_frame_before_activation_enabled && commit_has_no_updates). 
#0 0x7f90be1ce89b base::debug::StackTrace::StackTrace()
#1 0x7f90be1ccf2c base::debug::StackTrace::StackTrace()
#2 0x7f90be23b42f logging::LogMessage::~LogMessage()
#3 0x7f90c11d5548 cc::SchedulerStateMachine::WillCommit()
#4 0x7f90c11d77c5 cc::SchedulerStateMachine::BeginMainFrameAborted()
#5 0x7f90c11cc2c4 cc::Scheduler::BeginMainFrameAborted()
#6 0x7f90c1343cb0 cc::ProxyImpl::BeginMainFrameAbortedOnImpl()

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7ee15a00fe953c13a83ef4faeddb9ec4a03dc1c1

commit 7ee15a00fe953c13a83ef4faeddb9ec4a03dc1c1
Author: khushalsagar <khushalsagar@chromium.org>
Date: Mon Mar 06 22:40:05 2017

cc: Fix flaky LayerTreeHostScrollTestImplSideInvalidation

The test was making a wrong assumption that activation happens
synchronoulsy after impl-side invalidation. While we do synchronously
NotifyReadyToActivate, since there is no PrepareTiles, the activation
could be blocked on drawing the previous frame.

BUG= 697652 
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2731003003
Cr-Commit-Position: refs/heads/master@{#454981}

[modify] https://crrev.com/7ee15a00fe953c13a83ef4faeddb9ec4a03dc1c1/cc/scheduler/scheduler_state_machine.cc
[modify] https://crrev.com/7ee15a00fe953c13a83ef4faeddb9ec4a03dc1c1/cc/trees/layer_tree_host_unittest_scroll.cc