DeviceSyncClientImplTest failures with Mojo changes |
|||||||
Issue description
The failing tests are (in chromeos_unittests):
DeviceSyncClientImplTest.TestOnEnrollmentFinished
DeviceSyncClientImplTest
.TestCompleteInitialEnrollmentBeforeInitialSync
_WaitForLocalDeviceMetadata
DeviceSyncClientImplTest.TestOnNewDevicesSynced
See blocked bug for details and motivation regarding what's changing in Mojo. Likely the failures are caused by incorrect task ordering assumptions.
,
Aug 8
+ device_sync folks in case you have any cycles to help look into this
,
Aug 8
Hey Ken, can you post the test failures here?
,
Aug 8
The failures don't really provide much useful information. You can see a sample here[1] from one of my CL's try runs. [1] https://logs.chromium.org/logs/chromium/buildbucket/cr-buildbucket.appspot.com/8938897002999411472/+/steps/chromeos_unittests__with_patch_/0/logs/DeviceSyncClientImplTest.TestOnEnrollmentFinished/0
,
Aug 8
,
Aug 8
And just to clarify, these failures appear when applying this CL[1]. The CL does not break any guarantees made by Mojo bindings, but it can affect the timing of arbitrary events. The failures therefore strongly imply that incorrect ordering assumptions being made somewhere, and this has been the case for all other blocking bugs which have been fixed so far. [1] https://chromium-review.googlesource.com/c/chromium/src/+/1145692
,
Aug 8
hansberry@ (original implementer of that class) is going to take on this issue.
,
Aug 9
,
Oct 17
Gonna just knock this out since I see the problem.
,
Oct 17
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/14e46b5f85c5c219013de33e21ff23f7c6d1efb3 commit 14e46b5f85c5c219013de33e21ff23f7c6d1efb3 Author: Ken Rockot <rockot@chromium.org> Date: Wed Oct 17 00:17:56 2018 [device_sync] Fix some test expectations Changes some tests to run a RunLoop until idle rather than simply flushing a single Mojo interface pipe once. Due to changes in Mojo message dispatch timing, these tests break without this change. Flushing the interface in this case is *not* sufficient to guarantee that all of the expected work has completed by the time FlushForTesting returns. Rather than going through and doing a holistic refactor of these tests, this CL just fixes the few tests that happen to fail when making the necessary changes to Mojo bindings behavior. Bug: 872068 Change-Id: Ia869917715cedbec62e57cdcbac3d4b946232d38 Reviewed-on: https://chromium-review.googlesource.com/c/1285516 Reviewed-by: Kyle Horimoto <khorimoto@chromium.org> Commit-Queue: Ken Rockot <rockot@chromium.org> Cr-Commit-Position: refs/heads/master@{#600207} [modify] https://crrev.com/14e46b5f85c5c219013de33e21ff23f7c6d1efb3/chromeos/services/device_sync/public/cpp/device_sync_client_impl_unittest.cc
,
Oct 17
|
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by roc...@chromium.org
, Aug 8