Bug 112939 – Flaky Test: fast/performance/performance-now-timestamps.html

• Component: Tools / Tests
• Status: NEW

[WebKit Review Bot]

This is an automatically generated bug from the commit-queue.
fast/performance/performance-now-timestamps.html has been flaky on the commit-queue.

fast/performance/performance-now-timestamps.html was authored by [e-mail].
http://trac.webkit.org/browser/trunk/LayoutTests/fast/performance/performance-now-timestamps.html

The commit-queue just saw fast/performance/performance-now-timestamps.html flake (text diff) while processing attachment 194291 [details] on bug 112623.
Bot: gce-cq-02  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-3.3.8-gcg-201212281604-x86_64-with-GCEL-10.04-gcel_10.04

The bots will update this with information from each new failure.

If you believe this bug to be fixed or invalid, feel free to close.  The bots will re-open if the flake re-occurs.

If you would like to track this test fix with another bug, please close this bug as a duplicate.  The bots will follow the duplicate chain when making future comments.

[WebKit Review Bot]

Created attachment 194300 [details]
Archive of layout-test-results from gce-cq-02

[Ryan Haddad]

https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=fast%2Fperformance%2Fperformance-now-timestamps.html

[Jacob Uphoff]

Still being seen.

History:

https://results.webkit.org/?suite=layout-tests&test=fast%2Fperformance%2Fperformance-now-timestamps.html&limit=50000

Diff:

--- /Volumes/Data/slave/catalina-debug-tests-wk1/build/layout-test-results/fast/performance/performance-now-timestamps-expected.txt
+++ /Volumes/Data/slave/catalina-debug-tests-wk1/build/layout-test-results/fast/performance/performance-now-timestamps-actual.txt
@@ -9,7 +9,7 @@
 PASS secondTimestamp is defined.
 PASS secondTimestamp is >= firstTimestamp + (waitTime / 2)
 PASS elapsed is >= 1
-PASS elapsed < 100 is true
+FAIL elapsed < 100 should be true. Was false.
 PASS successfullyParsed is true
 
 TEST COMPLETE

[Jacob Uphoff]

I can reproduce with just running the test in iterations.

[Jacob Uphoff]

Created attachment 389992 [details]
Update Test Expectations

[Radar WebKit Bug Importer]

<rdar://problem/59236418>

[Truitt Savell]

Comment on attachment 389992 [details]
Update Test Expectations

Clearing flags on attachment: 389992

Committed r255978: <https://trac.webkit.org/changeset/255978>

[Mark Lam]

The test is merely measuring time around a busy wait with window.performance.now(), which is monotonic time.  If the system is busy (as in running a gazillion other layout tests concurrently), the test process can be pre-empted with no guarantee that it will resume executing in time to read the end sample time within any duration that the test expects.  In this case, the test is expecting a duration of 100ms.

This test is just a bad test.