Bot is here: https://ci.chromium.org/buildbot/chromium.linux/Fuchsia%20x64/?limit=100

Last 12 builds at the moment look good. Going to leave this for a while and see if it comes back. Might have been fixed by a v8 roll.

Detected 4 new flakes for test/step "ResourceDispatcherHostTest.TestBlockingResumingRequests". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyQgsSBUZsYWtlIjdSZXNvdXJjZURpc3BhdGNoZXJIb3N0VGVzdC5UZXN0QmxvY2tpbmdSZXN1bWluZ1JlcXVlc3RzDA. This message was posted automatically by the chromium-try-flakes app.

More failures over the weekend. Note that these are crashes, e.g:

[00138.634] 01204.01235> <== fatal exception: process content_unittests__exec[94366] thread initial-thread[94445]
[00138.634] 01204.01235> <== fatal page fault, PC at 0x18b2d5f5b82a
...
#01: content::ResourceDispatcherHostImpl::MaybeStartUpdateLoadInfoTimer() at ??:?
#02: base::(anonymous namespace)::PostTaskAndReplyRelay::RunReplyAndSelfDestruct() at ??:?
#03: base::debug::TaskAnnotator::RunTask(char const*, base::PendingTask*) at ??:?
#04: base::MessageLoop::RunTask(base::PendingTask*) at ??:?
#05: base::MessageLoop::DoWork() at ??:?
#06: base::MessagePumpFuchsia::Run(base::MessagePump::Delegate*) at ??:?
#07: base::RunLoop::Run() at ??:?
#08: content::RunAllTasksUntilIdle() at ??:?
#09: content::ResourceDispatcherHostTest::TearDown() at ??:?
#10: testing::TestInfo::Run() at ??:?
#11: testing::TestCase::Run() at ??:?
#12: testing::internal::UnitTestImpl::RunAllTests() at ??:?
#13: testing::UnitTest::Run() at ??:?
#14: base::TestSuite::Run() at ??:?
#15: base::(anonymous namespace)::LaunchUnitTestsInternal(base::OnceCallback<int ()>, unsigned long, int, bool, base::OnceCallback<void ()>) at ??:?
#16: base::LaunchUnitTests(int, char**, base::OnceCallback<int ()>) at ??:?
#17: main at ??:?
#18: pc 0x62c9ea61266e sp 0x11350baacfe0 (libc.so,0x1366e)
#19: pc 0 sp 0x11350baad000
[00138.722] 03954.04647> [3954:1556380101:0324/144731.473204:138722473:ERROR:test_launcher.cc(429)] going to zx_task_kill(job) for 5f346365
[00138.727] 03954.04647> [3954:1556380101:0324/144731.478486:138727753:ERROR:test_launcher.cc(446)] removing 5f346365 from live process list
[00138.756] 03954.04008> [1091/5323] ResourceDispatcherHostTest.TestProcessCancelDetachedTimesOut (637 ms)
[00138.756] 03954.04008> [ RUN      ] ResourceDispatcherHostTest.TestBlockingResumingRequests
[00138.757] 03954.04008> [94366:1748396633:0324/144730.679528:137928786:ERROR:base_paths_fuchsia.cc(37)] Not implemented reached in bool base::PathProviderFuchsia(int, base::FilePath *)
[00138.757] 03954.04008> [1092/5323] ResourceDispatcherHostTest.TestBlockingResumingRequests (CRASHED)

Suspecting that this is the AckUpdateLoadInfo() reply (introduced in https://chromium-review.googlesource.com/978383) being run.  MaybeStartUpdateLoadInfoTimer() touches the |update_load_info_timer_|, which I suspect has already been reset by a call to Shutdown().

Gonna disable the flaking tests on Fuschia.

 Issue 825694  has been merged into this issue.

 Issue 825693  has been merged into this issue.

Re #4: I'd recommend reverting the CL instead, or applying a patch to cope with the Ack being run post-Shutdow (i.e. test whether the timer member is non-null.

The flake isn't strictly specific to Fuchsia, it's just (much) more likely to occur under Fuchsia due to scheduling differences, so disabling it there isn't really appropriate.

Updating the Summary to reflect that this isn't test flake, but a genuine regression. :)

Thanks wez, I'm going to revert the CL now and see what happens. 

Builds have turned green.

Resolving.

Thanks for the revert, Ben!