Issue metadata
Sign in to add a comment
|
org.chromium.content.browser.TracingControllerAndroidTest#testTraceFileCreation is flaky |
||||||||||||||||||||||||
Issue descriptionorg.chromium.content.browser.TracingControllerAndroidTest#testTraceFileCreation is flaky. Findit has detected 7 flake occurrences of this test within the past 24 hours. List of all flake occurrences can be found at: https://findit-for-me.appspot.com/flake/occurrences?key=ag9zfmZpbmRpdC1mb3ItbWVyegsSBUZsYWtlIm9jaHJvbWl1bUBjb250ZW50X3NoZWxsX3Rlc3RfYXBrQG9yZy5jaHJvbWl1bS5jb250ZW50LmJyb3dzZXIuVHJhY2luZ0NvbnRyb2xsZXJBbmRyb2lkVGVzdCN0ZXN0VHJhY2VGaWxlQ3JlYXRpb24M. Unless the culprit CL is found and reverted, please disable this test first within 30 minutes then find an appropriate owner. If the result above is wrong, please file a bug using this link: https://bugs.chromium.org/p/chromium/issues/entry?status=Unconfirmed&labels=Pri-1,Test-Findit-Wrong&components=Tools%3ETest%3EFindit%3EFlakiness&summary=%5BFindit%5D%20Flake%20Detection%20-%20Wrong%20result%20for%20org.chromium.content.browser.TracingControllerAndroidTest#testTraceFileCreation&comment=Link%20to%20flake%20occurrences%3A%20https://findit-for-me.appspot.com/flake/occurrences?key=ag9zfmZpbmRpdC1mb3ItbWVyegsSBUZsYWtlIm9jaHJvbWl1bUBjb250ZW50X3NoZWxsX3Rlc3RfYXBrQG9yZy5jaHJvbWl1bS5jb250ZW50LmJyb3dzZXIuVHJhY2luZ0NvbnRyb2xsZXJBbmRyb2lkVGVzdCN0ZXN0VHJhY2VGaWxlQ3JlYXRpb24M Automatically posted by the findit-for-me app (https://goo.gl/Ot9f7N).
,
Oct 30
,
Oct 30
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/55b9c1cbbf1fbc1bc035bd5aef68ff3a5a62bb88 commit 55b9c1cbbf1fbc1bc035bd5aef68ff3a5a62bb88 Author: Ioana Pandele <ioanap@chromium.org> Date: Tue Oct 30 10:40:43 2018 Disable flaky tests in TracingControllerAndroidImplTest All test cases in TracingControllerAndroidImplTest seem to be flaky. Added failure logs to the bug. TBR=eseckler@chromium.org Bug: 899894 Change-Id: I478a5fa9845b77bd3614e6f896233f52db4bc74f Reviewed-on: https://chromium-review.googlesource.com/c/1306416 Reviewed-by: Ioana Pandele <ioanap@chromium.org> Commit-Queue: Ioana Pandele <ioanap@chromium.org> Cr-Commit-Position: refs/heads/master@{#603858} [modify] https://crrev.com/55b9c1cbbf1fbc1bc035bd5aef68ff3a5a62bb88/content/public/android/javatests/src/org/chromium/content/browser/TracingControllerAndroidImplTest.java
,
Oct 30
,
Oct 30
,
Oct 30
Two issues: a) There's a race in the way we update the result field of the callback. We should just flip the condition variable with the field update. This should fix testGetKnownCategories and testBufferUsage. b) Looks like file creation on the external storage fails for some reason. Will try to ascertain the cause.
,
Oct 31
I can't seem to find the cause of b). Opening the file fails, but fopen() seems to leave errno at 0 (success) for some reason. I figure it may be because storage permissions got screwed up somehow or maybe because the external storage is full. Since this only seems to happen on the nexus 5 kitkat bots so far, I'll restrict the two affected tests to > kitkat devices for now.
,
Oct 31
,
Oct 31
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/2fc9f6f25249bf221e970854bbdd9ddc6eaed3a9 commit 2fc9f6f25249bf221e970854bbdd9ddc6eaed3a9 Author: Eric Seckler <eseckler@chromium.org> Date: Wed Oct 31 11:37:33 2018 tracing: Address TracingControllerAndroid test failures. Fixes the flakiness of buffer usage / known categories tests by resolving an accidental race condition. Also restricts disabling the other two tests to <= kitkat devices since those seem to be the only ones that show (inexplicable) flakiness. TBR=yfriedman@chromium.org Bug: 899894 Change-Id: I0225c1519ea19910492ffb60a242bcc71b2275b4 Reviewed-on: https://chromium-review.googlesource.com/c/1307433 Reviewed-by: Eric Seckler <eseckler@chromium.org> Commit-Queue: Eric Seckler <eseckler@chromium.org> Cr-Commit-Position: refs/heads/master@{#604209} [modify] https://crrev.com/2fc9f6f25249bf221e970854bbdd9ddc6eaed3a9/content/public/android/javatests/src/org/chromium/content/browser/TracingControllerAndroidImplTest.java
,
Nov 6
Disabled on the kitkat bot.
,
Nov 7
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/8d3cd5fb007606b2f21d74825b188c8f2576dc23 commit 8d3cd5fb007606b2f21d74825b188c8f2576dc23 Author: agrieve <agrieve@chromium.org> Date: Wed Nov 07 19:18:32 2018 Revert "tracing: Address TracingControllerAndroid test failures." This reverts commit 2fc9f6f25249bf221e970854bbdd9ddc6eaed3a9. Reason for revert: Test is still very flaky https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=content_shell_test_apk&tests=org.chromium.content.browser.TracingControllerAndroidImplTest%23testBufferUsage Original change's description: > tracing: Address TracingControllerAndroid test failures. > > Fixes the flakiness of buffer usage / known categories tests by > resolving an accidental race condition. Also restricts disabling > the other two tests to <= kitkat devices since those seem to be > the only ones that show (inexplicable) flakiness. > > TBR=yfriedman@chromium.org > > Bug: 899894 > Change-Id: I0225c1519ea19910492ffb60a242bcc71b2275b4 > Reviewed-on: https://chromium-review.googlesource.com/c/1307433 > Reviewed-by: Eric Seckler <eseckler@chromium.org> > Commit-Queue: Eric Seckler <eseckler@chromium.org> > Cr-Commit-Position: refs/heads/master@{#604209} TBR=yfriedman@chromium.org,skyostil@chromium.org,eseckler@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: 899894 Change-Id: Ie47cf56e05861069e1b0a767138b1802f1a2584f Reviewed-on: https://chromium-review.googlesource.com/c/1323894 Reviewed-by: agrieve <agrieve@chromium.org> Commit-Queue: agrieve <agrieve@chromium.org> Cr-Commit-Position: refs/heads/master@{#606115} [modify] https://crrev.com/8d3cd5fb007606b2f21d74825b188c8f2576dc23/content/public/android/javatests/src/org/chromium/content/browser/TracingControllerAndroidImplTest.java
,
Nov 8
,
Nov 8
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/ee630768a346b9744dce7315162388b74c1616cb commit ee630768a346b9744dce7315162388b74c1616cb Author: Eric Seckler <eseckler@chromium.org> Date: Thu Nov 08 19:41:38 2018 Reland "tracing: Address TracingControllerAndroid test failures." This is a reland of 2fc9f6f25249bf221e970854bbdd9ddc6eaed3a9 The crashes in flakiness were due to calling destroy() on the wrong thread, which has been addressed in the new patch set. Original change's description: > tracing: Address TracingControllerAndroid test failures. > > Fixes the flakiness of buffer usage / known categories tests by > resolving an accidental race condition. Also restricts disabling > the other two tests to <= kitkat devices since those seem to be > the only ones that show (inexplicable) flakiness. > > TBR=yfriedman@chromium.org > > Bug: 899894 > Change-Id: I0225c1519ea19910492ffb60a242bcc71b2275b4 > Reviewed-on: https://chromium-review.googlesource.com/c/1307433 > Reviewed-by: Eric Seckler <eseckler@chromium.org> > Commit-Queue: Eric Seckler <eseckler@chromium.org> > Cr-Commit-Position: refs/heads/master@{#604209} Bug: 899894 Change-Id: I3cb002950f32bc6b6b4d25892adc3073c2fdb967 Reviewed-on: https://chromium-review.googlesource.com/c/1327407 Reviewed-by: Sami Kyöstilä <skyostil@chromium.org> Reviewed-by: Yaron Friedman <yfriedman@chromium.org> Commit-Queue: Eric Seckler <eseckler@chromium.org> Cr-Commit-Position: refs/heads/master@{#606570} [modify] https://crrev.com/ee630768a346b9744dce7315162388b74c1616cb/content/public/android/javatests/src/org/chromium/content/browser/TracingControllerAndroidImplTest.java
,
Nov 8
|
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by ioanap@google.com
, Oct 30