Implement timers by posting delayed tasks

Implement timers by posting delayed tasks

This patch refactors TimerBase to post tasks delayed tasks and
deletes the now-obsolete timer heap and shared timer mechanism.

ATTN Sheriffs: If there are weird layout test flakes all of a
sudden, this patch may be the cause since the interleaving of
timers with other posted tasks will change.

Original patch by Alex Clarke <alexclarke@chromium.org>;.

BUG=463143, 416362, 480522
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=196308

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=196497

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=196595

[Sami, 2015-05-12 10:52:43]

skyostil@chromium.org changed reviewers:
+ alexclarke@chromium.org

[Sami, 2015-05-12 17:32:39]

skyostil@chromium.org changed reviewers:
+ erikcorry@chromium.org, kinuko@chromium.org

[Sami, 2015-05-12 17:41:46]

Kinuko, does this look like a patch you could review? I'm not sure who knows
about timers since this code hasn't been touched all that much.

Erik, feel free to also do a pass if you've worked with this stuff before.

[kinuko, 2015-05-13 02:21:22]

On 2015/05/12 17:41:46, Sami wrote:
> Kinuko, does this look like a patch you could review? I'm not sure who knows
> about timers since this code hasn't been touched all that much.
> 
> Erik, feel free to also do a pass if you've worked with this stuff before.

Let me try taking a look.

[kinuko, 2015-05-13 08:26:39]

I'm not fully sure if this runs perfectly in a way that is/was expected, but
this patch lgtm. (And deleting the timer heap looks great!)

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer.cpp
File Source/platform/Timer.cpp (right):

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer....
Source/platform/Timer.cpp:101: m_webScheduler->postTimerTask(m_location,
m_cancellableTaskFactory.task(), delayMs);
Just want to make sure-- so this one actually cancels the previously scheduled
task, is that right?

https://codereview.chromium.org/1134523002/diff/160001/public/platform/WebSch...
File public/platform/WebScheduler.h (right):

https://codereview.chromium.org/1134523002/diff/160001/public/platform/WebSch...
public/platform/WebScheduler.h:74: typedef WTF::Function<void()> Task;
nit: is this change related to this CL?

[alex clarke (OOO till 29th), 2015-05-13 08:43:01]

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer.cpp
File Source/platform/Timer.cpp (right):

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer....
Source/platform/Timer.cpp:101: m_webScheduler->postTimerTask(m_location,
m_cancellableTaskFactory.task(), delayMs);
On 2015/05/13 08:26:39, kinuko wrote:
> Just want to make sure-- so this one actually cancels the previously scheduled
> task, is that right?

Yes it does. We where thinking of changing the name of task() to clarify.

Note I have a theoretical performance quibble with that cancellable task class.
It uses reference counting and atomics which we don't actually need here.  I
suspect for normal work loads with a few hundred tasks / frame it'll make no
difference but we should keep an eye on the various benchmarks which often have
very unusual work loads.

One other hopefully historical thing to bear in mind, when I first started this
I had a great deal of trouble with these tasks during shutdown in tests. The
problem was the tasks posted to chromium were out living the blink heap. That
should be fixed now, but if this patch causes asan versions of browser tests to
fail, you might want to rule that out.

[Sami, 2015-05-13 17:59:00]

skyostil@chromium.org changed reviewers:
+ jochen@chromium.org

[Sami, 2015-05-13 18:01:26]

On 2015/05/13 08:26:39, kinuko wrote:
> I'm not fully sure if this runs perfectly in a way that is/was expected, but
> this patch lgtm. (And deleting the timer heap looks great!)

Thanks for the review. The unit test we recently landed for timers should give
some confidence that we aren't changing functionality:
https://codereview.chromium.org/1130423002/

jochen@, could you have a look once you get a chance?

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer.cpp
File Source/platform/Timer.cpp (right):

https://codereview.chromium.org/1134523002/diff/160001/Source/platform/Timer....
Source/platform/Timer.cpp:101: m_webScheduler->postTimerTask(m_location,
m_cancellableTaskFactory.task(), delayMs);
On 2015/05/13 08:43:01, alexclarke1 wrote:
> On 2015/05/13 08:26:39, kinuko wrote:
> > Just want to make sure-- so this one actually cancels the previously
scheduled
> > task, is that right?
> 
> Yes it does. We where thinking of changing the name of task() to clarify.
> 
> Note I have a theoretical performance quibble with that cancellable task
class.
> It uses reference counting and atomics which we don't actually need here.  I
> suspect for normal work loads with a few hundred tasks / frame it'll make no
> difference but we should keep an eye on the various benchmarks which often
have
> very unusual work loads.

We could consider introducing a non-thread safe variant of cancellable tasks for
that purpose.

> One other hopefully historical thing to bear in mind, when I first started
this
> I had a great deal of trouble with these tasks during shutdown in tests. The
> problem was the tasks posted to chromium were out living the blink heap. That
> should be fixed now, but if this patch causes asan versions of browser tests
to
> fail, you might want to rule that out.

Yes, I think I've also learned that shutting things down correctly is very
tricky.

https://codereview.chromium.org/1134523002/diff/160001/public/platform/WebSch...
File public/platform/WebScheduler.h (right):

https://codereview.chromium.org/1134523002/diff/160001/public/platform/WebSch...
public/platform/WebScheduler.h:74: typedef WTF::Function<void()> Task;
On 2015/05/13 08:26:39, kinuko wrote:
> nit: is this change related to this CL?

Ah, we don't actually need to do this I think. Removed.

[jochen (gone - plz use gerrit), 2015-05-19 10:54:14]

it used to be the case that first setting a long fire time, and then setting a
shorter one (before the first one fired) would cancel the first one.

is that still the case?

cancelling of tasks is still handled at a hire level (e.g. when you have a
setInterval and then cancel it after a few runs), right?

[alex clarke (OOO till 29th), 2015-05-19 10:56:21]

On 2015/05/19 10:54:14, jochen (traveling) wrote:
> it used to be the case that first setting a long fire time, and then setting a
> shorter one (before the first one fired) would cancel the first one.
> 
> is that still the case?

Yes.  This patch should clarify that: https://codereview.chromium.org/937883003/


> 
> cancelling of tasks is still handled at a hire level (e.g. when you have a
> setInterval and then cancel it after a few runs), right?

[Sami, 2015-05-19 11:07:42]

On 2015/05/19 10:54:14, jochen (traveling) wrote:
> it used to be the case that first setting a long fire time, and then setting a
> shorter one (before the first one fired) would cancel the first one.
> 
> is that still the case?

Yes, we're keeping that behavior. It's now also more efficient than before
because changing the timeouts of unrelated timers doesn't affect any of the
other timer callbacks. Previously the shared timer had to always reschedule a
callback when the frontmost timer on the heap changed. This lead to callback
storms where the callback for a timer with a long delay kept getting reposted
and cancelled as other shorter timers appeared before it.

> cancelling of tasks is still handled at a hire level (e.g. when you have a
> setInterval and then cancel it after a few runs), right?

Yes, posted tasks cannot be cancelled so we're using CancellableTaskFactory as a
layer on top to implement cancelling.

[Sami, 2015-05-20 18:12:06]

I checked that the failure from the win trybot
(plugins/netscape-plugin-setwindow-size-2.html) doesn't seem to be related to
timers. The test is very flaky at ToT already. I wasn't able to trigger any of
the other flaky failures.

[Sami, 2015-05-20 18:19:55]

The CQ bit was checked by skyostil@chromium.org to run a CQ dry run

[Sami, 2015-05-20 18:19:55]

The patchset sent to the CQ was uploaded after l-g-t-m from kinuko@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps260001
(title: "Rebased again.")

[commit-bot: I haz the power, 2015-05-20 18:20:09]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/260001

[commit-bot: I haz the power, 2015-05-20 18:40:26]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-20 18:40:27]

Dry run: Try jobs failed on following builders:
  win_blink_compile_dbg on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/win_blink_compile_dbg/bu...)

[Sami, 2015-05-20 18:52:54]

The CQ bit was checked by skyostil@chromium.org to run a CQ dry run

[Sami, 2015-05-20 18:52:54]

The patchset sent to the CQ was uploaded after l-g-t-m from kinuko@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps280001
(title: "Added missing include.")

[commit-bot: I haz the power, 2015-05-20 18:53:17]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/280001

[commit-bot: I haz the power, 2015-05-20 20:16:31]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-20 20:16:32]

Dry run: Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/6...)

[jochen (gone - plz use gerrit), 2015-05-21 07:50:56]

lgtm

[rmcilroy, 2015-05-21 09:11:00]

rmcilroy@chromium.org changed reviewers:
+ rmcilroy@chromium.org

[Sami, 2015-05-21 10:04:49]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-05-21 10:04:49]

The patchset sent to the CQ was uploaded after l-g-t-m from kinuko@chromium.org,
jochen@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps300001
(title: "Rebased.")

[commit-bot: I haz the power, 2015-05-21 10:05:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/300001

[commit-bot: I haz the power, 2015-05-21 10:07:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-21 10:08:00]

Try jobs failed on following builders:
  mac_blink_compile_dbg on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/mac_blink_compile_dbg/bu...)
  mac_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/55980)

[Sami, 2015-05-21 15:29:36]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-05-21 15:29:37]

The patchset sent to the CQ was uploaded after l-g-t-m from kinuko@chromium.org,
jochen@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps320001
(title: "Rebased again.")

[commit-bot: I haz the power, 2015-05-21 15:30:23]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/320001

[commit-bot: I haz the power, 2015-05-21 15:33:47]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-21 15:33:49]

Try jobs failed on following builders:
  mac_blink_compile_dbg on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/mac_blink_compile_dbg/bu...)
  mac_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/56021)

[Sami, 2015-05-21 15:38:35]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-05-21 15:38:35]

The patchset sent to the CQ was uploaded after l-g-t-m from kinuko@chromium.org,
jochen@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps340001
(title: "Rebase harder.")

[commit-bot: I haz the power, 2015-05-21 15:39:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/340001

[commit-bot: I haz the power, 2015-05-21 17:11:50]

Committed patchset #18 (id:340001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=195706

[Julien - ping for review, 2015-05-21 19:59:43]

A revert of this CL (patchset #18 id:340001) has been created in
https://codereview.chromium.org/1151633004/ by jchaffraix@chromium.org.

The reason for reverting is: The patch made some content_browsertests fail on
Linux:
DomSerializerTests.SerializeHTMLDOMWithDocType
DomSerializerTests.SerializeHTMLDOMWithEmptyHead
DomSerializerTests.SerializeHTMLDOMWithEntitiesInText
DomSerializerTests.SerializeHTMLDOMWithMultipleMetaCharsetInOriginalDoc
DomSerializerTests.SerializeHTMLDOMWithNoMetaCharsetInOriginalDoc
DomSerializerTests.SerializeHTMLDOMWithoutDocType
RenderFrameHostManagerTest.DisownOpener.

[sof, 2015-05-21 20:10:05]

sigbjornf@opera.com changed reviewers:
+ sigbjornf@opera.com

[sof, 2015-05-21 20:10:06]

ftr, some Oilpan ASAN breakage seen too,


http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Linux%20Oilpan%...

(The oilpan team can probably take care of that once relanded, just a heads up.)

[Julien - ping for review, 2015-05-21 21:10:41]

jchaffraix@chromium.org changed reviewers:
+ jchaffraix@chromium.org

[Julien - ping for review, 2015-05-21 21:10:42]

Failures:
DomSerializerTests.* were timing out, locally I got the following backtrace:

BrowserTestBase signal handler received SIGTERM. Backtrace:
#0 0x000001489ace Still waiting for the following processes to finish:
        out/Debug/content_browsertests
--data-path=/tmp/.org.chromium.Chromium.D64DNy/dseO4Y1
--gtest_also_run_disabled_tests --gte
st_filter=DomSerializerTests.SerializeHTMLDOMWithoutDocType --single_process
--use-fake-device-for-media-stream --use-fake-ui-for-m
edia-stream
base::debug::StackTrace::StackTrace()
#1 0x000000e8ad3a content::(anonymous namespace)::DumpStackTraceSignalHandler()
#2 0x7f181886ad40 <unknown>
#3 0x7f181892112d __poll
#4 0x7f181edb4fe4 <unknown>
#5 0x7f181edb50ec g_main_context_iteration
#6 0x000001594b1f base::MessagePumpGlib::Run()
#7 0x0000014babaf base::MessageLoop::RunHandler()
#8 0x0000014f0e24 base::RunLoop::Run()
#9 0x000000eb3d83 content::RunThisRunLoop()
#10 0x000000eb4677 content::MessageLoopRunner::Run()
#11 0x000000e8b2a4
content::BrowserTestBase::PostTaskToInProcessRendererAndWait()
#12 0x0000006b5d47
content::DomSerializerTests_SerializeHTMLDOMWithoutDocType_Test::RunTestOnMainThread()
#13 0x000000d8a08c content::ContentBrowserTest::RunTestOnMainThreadLoop()
#14 0x000000e8ab1a content::BrowserTestBase::ProxyRunTestOnMainThreadLoop()
#15 0x000000496019 base::internal::RunnableAdapter<>::Run()
#16 0x0000005bb179
_ZN4base8internal12InvokeHelperILb0EvNS0_15RunnableAdapterIMN7content24ServiceWorkerBrowserTestEFvvEEENS0_8TypeListIJRKPS4_EEEE8MakeItSoES7_SB_
#17 0x000000e8bd70
_ZN4base8internal7InvokerI13IndexSequenceIJLm0EEENS0_9BindStateINS0_15RunnableAdapterIMN7content15BrowserTestBaseEFvvEEEFvPS7_ENS0_8TypeListIJSB_EEEEENSD_IJNS0_12UnwrapTraitsISB_EEEEENS0_12InvokeHelperILb0EvSA_NSD_IJRKSB_EEEEEFvvEE3RunEPNS0_13BindStateBaseE
#18 0x00000047e81e base::Callback<>::Run()
#19 0x000000e014e5 content::ShellBrowserMainParts::PreMainMessageLoopRun()
#20 0x00000571fe62 content::BrowserMainLoop::PreMainMessageLoopRun()
#21 0x000000496019 base::internal::RunnableAdapter<>::Run()
#22 0x000000495f9c
_ZN4base8internal12InvokeHelperILb0EvNS0_15RunnableAdapterIMN7content21CrossSiteTransferTestEFvvEEENS0_8TypeListIJPS4_EEEE8MakeItSoES7_S9_
#23 0x0000057255c8
_ZN4base8internal7InvokerI13IndexSequenceIJLm0EEENS0_9BindStateINS0_15RunnableAdapterIMN7content15BrowserMainLoopEFivEEEFiPS7_ENS0_8TypeListIJNS0_17UnretainedWrapperIS7_EEEEEEENSD_IJNS0_12UnwrapTraitsISF_EEEEENS0_12InvokeHelperILb0EiSA_NSD_IJSB_EEEEEFivEE3RunEPNS0_13BindStateBaseE
#24 0x00000047e81e base::Callback<>::Run()
#25 0x000005b91bab content::StartupTaskRunner::RunAllTasksNow()
#26 0x00000571e125 content::BrowserMainLoop::CreateStartupTasks()
#27 0x000005726665 content::BrowserMainRunnerImpl::Initialize()
#28 0x000000dbf3d4 ShellBrowserMain()
#29 0x000000da3fab content::ShellMainDelegate::RunProcess()
#30 0x0000056cd20b content::RunNamedProcessTypeMain()
#31 0x0000056ced7d content::ContentMainRunnerImpl::Run()
#32 0x0000056cc942 content::ContentMain()
#33 0x000000e8a871 content::BrowserTestBase::SetUp()
#34 0x000000d89f21 content::ContentBrowserTest::SetUp()
#35 0x0000011e676a testing::internal::HandleSehExceptionsInMethodIfSupported<>()
#36 0x0000011d7e6e testing::internal::HandleExceptionsInMethodIfSupported<>()
#37 0x0000011cccc3 testing::Test::Run()
#38 0x0000011cd3f8 testing::TestInfo::Run()
#39 0x0000011cd99a testing::TestCase::Run()
#40 0x0000011d2d5c testing::internal::UnitTestImpl::RunAllTests()
#41 0x0000011ea5ea testing::internal::HandleSehExceptionsInMethodIfSupported<>()
#42 0x0000011d95ae testing::internal::HandleExceptionsInMethodIfSupported<>()
#43 0x0000011d29ff testing::UnitTest::Run()
#44 0x000000ee6141 RUN_ALL_TESTS()
#45 0x000000ee51a8 base::TestSuite::Run()
#46 0x000000d8b19c content::ContentTestLauncherDelegate::RunTestSuite()
#47 0x000000eab6a2 content::LaunchTests()
#48 0x000000d8aff8 main
#49 0x7f1818855ec5 __libc_start_main
#50 0x0000004500d6 <unknown>
[0521/140809:ERROR:kill_posix.cc(74)] Unable to terminate process group 11002:
No such process
[2/10] DomSerializerTests.SerializeHTMLDOMWithoutDocType (TIMED OUT)


I have seen RenderFrameHostManagerTest.DisownOpener fail locally but didn't get
a local output and it seems to pass on its own. The change may have made it
flaky, not totally sure.

[Julien - ping for review, 2015-05-21 21:16:52]

(Forgot some information, sorry for the spam)

On 2015/05/21 at 21:10:42, Julien Chaffraix - PST wrote:
> Failures:
> DomSerializerTests.* were timing out, locally I got the following backtrace:
> 
> BrowserTestBase signal handler received SIGTERM. Backtrace:
> #0 0x000001489ace Still waiting for the following processes to finish:
>         out/Debug/content_browsertests
--data-path=/tmp/.org.chromium.Chromium.D64DNy/dseO4Y1
--gtest_also_run_disabled_tests --gte
> st_filter=DomSerializerTests.SerializeHTMLDOMWithoutDocType --single_process
--use-fake-device-for-media-stream --use-fake-ui-for-m
> edia-stream
> base::debug::StackTrace::StackTrace()
> #1 0x000000e8ad3a content::(anonymous
namespace)::DumpStackTraceSignalHandler()
> #2 0x7f181886ad40 <unknown>
> #3 0x7f181892112d __poll
> #4 0x7f181edb4fe4 <unknown>
> #5 0x7f181edb50ec g_main_context_iteration
> #6 0x000001594b1f base::MessagePumpGlib::Run()
> #7 0x0000014babaf base::MessageLoop::RunHandler()
> #8 0x0000014f0e24 base::RunLoop::Run()
> #9 0x000000eb3d83 content::RunThisRunLoop()
> #10 0x000000eb4677 content::MessageLoopRunner::Run()
> #11 0x000000e8b2a4
content::BrowserTestBase::PostTaskToInProcessRendererAndWait()
> #12 0x0000006b5d47
content::DomSerializerTests_SerializeHTMLDOMWithoutDocType_Test::RunTestOnMainThread()
> #13 0x000000d8a08c content::ContentBrowserTest::RunTestOnMainThreadLoop()
> #14 0x000000e8ab1a content::BrowserTestBase::ProxyRunTestOnMainThreadLoop()
> #15 0x000000496019 base::internal::RunnableAdapter<>::Run()
> #16 0x0000005bb179
_ZN4base8internal12InvokeHelperILb0EvNS0_15RunnableAdapterIMN7content24ServiceWorkerBrowserTestEFvvEEENS0_8TypeListIJRKPS4_EEEE8MakeItSoES7_SB_
> #17 0x000000e8bd70
_ZN4base8internal7InvokerI13IndexSequenceIJLm0EEENS0_9BindStateINS0_15RunnableAdapterIMN7content15BrowserTestBaseEFvvEEEFvPS7_ENS0_8TypeListIJSB_EEEEENSD_IJNS0_12UnwrapTraitsISB_EEEEENS0_12InvokeHelperILb0EvSA_NSD_IJRKSB_EEEEEFvvEE3RunEPNS0_13BindStateBaseE
> #18 0x00000047e81e base::Callback<>::Run()
> #19 0x000000e014e5 content::ShellBrowserMainParts::PreMainMessageLoopRun()
> #20 0x00000571fe62 content::BrowserMainLoop::PreMainMessageLoopRun()
> #21 0x000000496019 base::internal::RunnableAdapter<>::Run()
> #22 0x000000495f9c
_ZN4base8internal12InvokeHelperILb0EvNS0_15RunnableAdapterIMN7content21CrossSiteTransferTestEFvvEEENS0_8TypeListIJPS4_EEEE8MakeItSoES7_S9_
> #23 0x0000057255c8
_ZN4base8internal7InvokerI13IndexSequenceIJLm0EEENS0_9BindStateINS0_15RunnableAdapterIMN7content15BrowserMainLoopEFivEEEFiPS7_ENS0_8TypeListIJNS0_17UnretainedWrapperIS7_EEEEEEENSD_IJNS0_12UnwrapTraitsISF_EEEEENS0_12InvokeHelperILb0EiSA_NSD_IJSB_EEEEEFivEE3RunEPNS0_13BindStateBaseE
> #24 0x00000047e81e base::Callback<>::Run()
> #25 0x000005b91bab content::StartupTaskRunner::RunAllTasksNow()
> #26 0x00000571e125 content::BrowserMainLoop::CreateStartupTasks()
> #27 0x000005726665 content::BrowserMainRunnerImpl::Initialize()
> #28 0x000000dbf3d4 ShellBrowserMain()
> #29 0x000000da3fab content::ShellMainDelegate::RunProcess()
> #30 0x0000056cd20b content::RunNamedProcessTypeMain()
> #31 0x0000056ced7d content::ContentMainRunnerImpl::Run()
> #32 0x0000056cc942 content::ContentMain()
> #33 0x000000e8a871 content::BrowserTestBase::SetUp()
> #34 0x000000d89f21 content::ContentBrowserTest::SetUp()
> #35 0x0000011e676a
testing::internal::HandleSehExceptionsInMethodIfSupported<>()
> #36 0x0000011d7e6e testing::internal::HandleExceptionsInMethodIfSupported<>()
> #37 0x0000011cccc3 testing::Test::Run()
> #38 0x0000011cd3f8 testing::TestInfo::Run()
> #39 0x0000011cd99a testing::TestCase::Run()
> #40 0x0000011d2d5c testing::internal::UnitTestImpl::RunAllTests()
> #41 0x0000011ea5ea
testing::internal::HandleSehExceptionsInMethodIfSupported<>()
> #42 0x0000011d95ae testing::internal::HandleExceptionsInMethodIfSupported<>()
> #43 0x0000011d29ff testing::UnitTest::Run()
> #44 0x000000ee6141 RUN_ALL_TESTS()
> #45 0x000000ee51a8 base::TestSuite::Run()
> #46 0x000000d8b19c content::ContentTestLauncherDelegate::RunTestSuite()
> #47 0x000000eab6a2 content::LaunchTests()
> #48 0x000000d8aff8 main
> #49 0x7f1818855ec5 __libc_start_main
> #50 0x0000004500d6 <unknown>
> [0521/140809:ERROR:kill_posix.cc(74)] Unable to terminate process group 11002:
No such process
> [2/10] DomSerializerTests.SerializeHTMLDOMWithoutDocType (TIMED OUT)

Link to the flakiness dashboard:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=Dom...

And a failure from the bots:
http://build.chromium.org/p/chromium.webkit/builders/Linux%20Tests/builds/445...

[Sami, 2015-05-22 17:41:22]

We believe https://codereview.chromium.org/1153763005/ addresses the timeouts
that this was causing, but there's a chance it may uncover other test flaky once
this patch lands. I'll try relanding this, but if you see weird failures, feel
free to revert and we'll investigate.

[Sami, 2015-05-22 17:42:26]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-05-22 17:42:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/340001

[commit-bot: I haz the power, 2015-05-22 20:55:44]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-22 20:55:46]

Try jobs failed on following builders:
  mac_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/56170)

[Sami, 2015-05-26 10:32:06]

The CQ bit was checked by skyostil@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-05-26 10:32:29]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/340001

[commit-bot: I haz the power, 2015-05-26 12:43:10]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-26 12:43:13]

Dry run: Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/6...)

[Sami, 2015-05-27 09:44:20]

To give an update, the main layout test failure seems to be
inspector/sources/debugger/async-callstack-events.html, and it's caused by us
incorrectly allowing timers to run while blink is in a nested modal message
loop. Alex is working on a fix.

[Sami, 2015-05-28 15:52:01]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-05-28 15:52:02]

The patchset sent to the CQ was uploaded after l-g-t-m from jochen@chromium.org,
kinuko@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps380001
(title: "Rebased.")

[commit-bot: I haz the power, 2015-05-28 15:52:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/380001

[commit-bot: I haz the power, 2015-05-28 18:04:33]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-28 18:04:36]

Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/6...)

[Sami, 2015-05-29 08:37:55]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-05-29 08:37:59]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/380001

[commit-bot: I haz the power, 2015-05-29 10:53:09]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-29 10:53:11]

Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/6...)

[Sami, 2015-05-29 13:44:58]

The failing layout test was fixed with
https://codereview.chromium.org/1148293010/.

[Sami, 2015-05-29 13:45:07]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-05-29 13:45:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/380001

[commit-bot: I haz the power, 2015-05-29 15:08:23]

Committed patchset #20 (id:380001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196152

[Sami, 2015-06-01 18:23:21]

The failing mojo test was fixed with https://codereview.chromium.org/1145973011,
so let's give this another shot.

[Sami, 2015-06-01 18:23:31]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-06-01 18:23:31]

The patchset sent to the CQ was uploaded after l-g-t-m from jochen@chromium.org,
kinuko@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps400001
(title: "Rebased.")

[commit-bot: I haz the power, 2015-06-01 18:24:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/400001

[commit-bot: I haz the power, 2015-06-01 21:21:25]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-06-01 21:21:27]

Try jobs failed on following builders:
  win_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/64628)

[Sami, 2015-06-02 09:40:08]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-06-02 09:40:38]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/400001

[commit-bot: I haz the power, 2015-06-02 11:00:20]

Committed patchset #21 (id:400001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196308

[Sami, 2015-06-02 14:24:58]

A revert of this CL (patchset #21 id:400001) has been created in
https://codereview.chromium.org/1162903005/ by skyostil@chromium.org.

The reason for reverting is: Reverting once again because this makes
ScriptContextSetTest.LifeCycle leak:

http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...

The reason is that many of these tests do things backwards: they first
initialize Blink and then create a message loop. This means we can't create a
real scheduler for Blink and have to jump through hoops to make things work. We
could extend this mock scheduler to properly clean things up, but I think I'll
instead try to get rid of the mock scheduler completely..

[Sami, 2015-06-03 18:21:29]

Update: Once https://codereview.chromium.org/1152623008/ lands to fix the memory
leak issue I'll give this one another shot.

[Sami, 2015-06-03 21:35:54]

The CQ bit was checked by skyostil@chromium.org

[Sami, 2015-06-03 21:35:55]

The patchset sent to the CQ was uploaded after l-g-t-m from jochen@chromium.org,
kinuko@chromium.org
Link to the patchset: https://codereview.chromium.org/1134523002/#ps420001
(title: "Rebased.")

[commit-bot: I haz the power, 2015-06-03 21:37:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/420001

[commit-bot: I haz the power, 2015-06-03 22:13:32]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-06-03 22:13:35]

Try jobs failed on following builders:
  linux_blink_rel on tryserver.blink (JOB_FAILED,
http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/6...)

[Sami, 2015-06-04 08:07:43]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-06-04 08:08:26]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/420001

[Sami, 2015-06-04 08:11:02]

The CQ bit was unchecked by skyostil@chromium.org

[Sami, 2015-06-04 11:54:00]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-06-04 11:54:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/420001

[commit-bot: I haz the power, 2015-06-04 13:11:09]

Committed patchset #22 (id:420001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196497

[leviw_travelin_and_unemployed, 2015-06-05 03:11:37]

On 2015/06/04 at 13:11:09, commit-bot wrote:
> Committed patchset #22 (id:420001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196497

Any chance this could be triggering these ASAN failures:
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...

It's the only commit that looks suspicious to me...

[Yuta Kitamura, 2015-06-05 04:19:21]

On 2015/06/05 03:11:37, leviw wrote:
> On 2015/06/04 at 13:11:09, commit-bot wrote:
> > Committed patchset #22 (id:420001) as
> https://src.chromium.org/viewvc/blink?view=rev&revision=196497
> 
> Any chance this could be triggering these ASAN failures:
>
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...
> 
> It's the only commit that looks suspicious to me...

Yeah, I'm pretty sure this is a cause of ASAN failures.
They are blocking Blink rolls right now; I'm going to revert
this unless the failures can be fixed soon.

[Yuta Kitamura, 2015-06-05 04:27:35]

A revert of this CL (patchset #22 id:420001) has been created in
https://codereview.chromium.org/1167023002/ by yutak@chromium.org.

The reason for reverting is: Broke ASAN and blocking Blink rolls.
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...
.

[Sami, 2015-06-05 14:32:10]

This time around it looks like the leak is a false positive from LSAN. I'll try
to fix it with https://codereview.chromium.org/1169493006/.

[Sami, 2015-06-05 16:30:45]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2015-06-05 16:31:15]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1134523002/420001

[commit-bot: I haz the power, 2015-06-05 18:30:15]

Committed patchset #22 (id:420001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=196595