ThreadConditionTest.WaitReportsBlockingCall is flaky |
|||
Issue description
wtf_unittests:ThreadConditionTest.WaitReportsBlockingCall typically flakes on its first run, only passing when retried by TestLauncher:
...
[311/343] StringTest.DeprecatedLower (1 ms)
[312/343] StringTest.Ensure16Bit (0 ms)
[313/343] StringTest.StringPrinter (0 ms)
[314/343] WTF.SimplifyWhiteSpace (0 ms)
[ RUN ] ThreadConditionTest.WaitReportsBlockingCall
../../third_party/blink/renderer/platform/wtf/threading_primitives_test.cc:45: Failure
Mock function called more times than expected - returning directly.
Function call: BlockingStarted(4-byte object <00-00 00-00>)
Expected: to be called once
Actual: called twice - over-saturated and active
Stack trace:
#0 0x0000005e5a0c testing::internal::UnitTestImpl::CurrentOsStackTraceExceptTop()
#1 0x0000005e53d9 testing::internal::AssertHelper::operator=()
#2 0x0000005d5c33 testing::internal::GoogleTestFailureReporter::ReportFailure()
#3 0x0000005d796a testing::internal::UntypedFunctionMockerBase::UntypedInvokeWith()
#4 0x0000004ea3f4 WTF::(anonymous namespace)::MockBlockingObserver::BlockingStarted()
../../third_party/blink/renderer/platform/wtf/threading_primitives_test.cc:46: Failure
Mock function called more times than expected - returning directly.
Function call: BlockingEnded()
Expected: to be called once
Actual: called twice - over-saturated and active
Stack trace:
#0 0x0000005e5a0c testing::internal::UnitTestImpl::CurrentOsStackTraceExceptTop()
#1 0x0000005e53d9 testing::internal::AssertHelper::operator=()
#2 0x0000005d5c33 testing::internal::GoogleTestFailureReporter::ReportFailure()
#3 0x0000005d796a testing::internal::UntypedFunctionMockerBase::UntypedInvokeWith()
#4 0x0000004ea49b WTF::(anonymous namespace)::MockBlockingObserver::BlockingEnded()
[ FAILED ] ThreadConditionTest.WaitReportsBlockingCall (4 ms)
[315/343] ThreadConditionTest.WaitReportsBlockingCall (4 ms)
Running the test locally does not repro this issue, suggesting that the issue is timing-specific (local workstation has different # of CPU cores, etc, from the bots).
,
Oct 22
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/384e09bd3497286d570fae503fe32fb2e3d97021 commit 384e09bd3497286d570fae503fe32fb2e3d97021 Author: Wez <wez@chromium.org> Date: Mon Oct 22 17:56:49 2018 Disable flaky ThreadConditionTest.WaitReportsBlockingCall test. This test often flakes on its first run in a batch, with each call expectation satisfied more than once. Bug: 897550 , 879806 Change-Id: I2b7b130e5fd2fc8abf6aa15fbfd417041da0420f Reviewed-on: https://chromium-review.googlesource.com/c/1293178 Reviewed-by: Kentaro Hara <haraken@chromium.org> Commit-Queue: Wez <wez@chromium.org> Cr-Commit-Position: refs/heads/master@{#601645} [modify] https://crrev.com/384e09bd3497286d570fae503fe32fb2e3d97021/third_party/blink/renderer/platform/wtf/threading_primitives_test.cc
,
Oct 24
,
Oct 24
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/b2598b46c17876a42e6e1a21bb2677fffef355d6 commit b2598b46c17876a42e6e1a21bb2677fffef355d6 Author: Jeremy Roman <jbroman@chromium.org> Date: Wed Oct 24 22:10:35 2018 Narrow the scope of blocking observation in ThreadConditionTest to the single Wait call of interest. Bug: 897550 Change-Id: Id34dc044bb6c5bd5e177d405015fdd909827ce18 Reviewed-on: https://chromium-review.googlesource.com/c/1298346 Commit-Queue: Jeremy Roman <jbroman@chromium.org> Reviewed-by: Kentaro Hara <haraken@chromium.org> Cr-Commit-Position: refs/heads/master@{#602481} [modify] https://crrev.com/b2598b46c17876a42e6e1a21bb2677fffef355d6/third_party/blink/renderer/platform/wtf/threading_primitives_test.cc
,
Oct 31
No more reports of flakiness, at least. |
|||
►
Sign in to add a comment |
|||
Comment 1 by yutak@chromium.org
, Oct 22Status: Available (was: Untriaged)