Summary: | webkitpy.layout_tests.layout_package.message_broker_unittest.MultiThreadedBrokerTest failed on commit-queue | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Eric Seidel (no email) <eric> | ||||||
Component: | Tools / Tests | Assignee: | Dirk Pranke <dpranke> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Normal | CC: | dpranke, mihaip, ojan, tony | ||||||
Priority: | P2 | ||||||||
Version: | 528+ (Nightly build) | ||||||||
Hardware: | PC | ||||||||
OS: | OS X 10.5 | ||||||||
Attachments: |
|
Description
Eric Seidel (no email)
2010-12-23 20:02:47 PST
I'm not sure if this is flaky or what. odd. could be flaky. is there a way to tell which patch was being applied? Was it the change I just filed (bug 51572), or some other random change? There is no easy way to go back from the result id on the bot to the patch/bug combo, sadly. Just saw another failure from this test suite on the commit-queue: FAIL: test_timeout (webkitpy.layout_tests.layout_package.message_broker_unittest.MultiThreadedBrokerTest) ---------------------------------------------------------------------- Traceback (most recent call last): File "/mnt/git/webkit-commit-queue/Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py", line 140, in test_timeout interrupted = self.run_one_thread('Timeout') File "/mnt/git/webkit-commit-queue/Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py", line 128, in run_one_thread return broker.run_message_loop() File "/mnt/git/webkit-commit-queue/Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py", line 159, in run_message_loop log_wedged_worker(thread.getName(), thread.id()) File "/mnt/git/webkit-commit-queue/Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py", line 176, in log_wedged_worker assert(stack is not None) AssertionError We saw this flake again on the commit queue: https://webkit-commit-queue.appspot.com/results/7492048 Isn't this test (and the code it tests) going away once we switch NRWT to using multiple processes instead of threads? (In reply to comment #6) > Isn't this test (and the code it tests) going away once we switch NRWT to using multiple processes instead of threads? Can't say for sure. My current plan is to leave in an option for threads, at least until we're sure multiple processes is stable. Created attachment 79335 [details]
Patch
Comment on attachment 79335 [details] Patch Clearing flags on attachment: 79335 Committed r76073: <http://trac.webkit.org/changeset/76073> All reviewed patches have been landed. Closing bug. Created attachment 81019 [details] update w/ fix in r76073, try again Comment on attachment 81019 [details] update w/ fix in r76073, try again Whoops. Wrong bug - should've been bug 51572. |