Bug 310547 – REGRESSION(309748@main) [WPE][GStreamer] imported/w3c/web-platform-tests/media-source/mediasource-append-buffer.html is flaky failing

• Status: RESOLVED
• Resolution: FIXED
• Component: Media

[Lauro Moura]

imported/w3c/web-platform-tests/media-source/mediasource-append-buffer.html

https://results.webkit.org/?suite=layout-tests&test=imported%2Fw3c%2Fweb-platform-tests%2Fmedia-source%2Fmediasource-append-buffer.html

Diff:

--- /sdk/webkit/layout-test-results/imported/w3c/web-platform-tests/media-source/mediasource-append-buffer-expected.txt
+++ /sdk/webkit/layout-test-results/imported/w3c/web-platform-tests/media-source/mediasource-append-buffer-actual.txt
@@ -14,7 +14,7 @@
 PASS Test appending a neutered ArrayBuffer.
 PASS Test appendBuffer with partial init segments.
 PASS Test appendBuffer with partial media segments.
-FAIL Test appendBuffer events order. assert_true: expected true got false
+FAIL Test appendBuffer events order. assert_equals: expected 1 but got 3
 PASS Test abort in the middle of an initialization segment.
 PASS Test abort after removing sourcebuffer.
 PASS Test abort after readyState is ended following init segment and media segment.

wk-testhunter shows it was quite stable before:

The bot wpe-release completed 1475 runs on the interval [305729@main-309728@main]
Frequency of results for test: imported/w3c/web-platform-tests/media-source/mediasource-append-buffer.html
  1472 times [99.8%]     ->  NOERROR
  2 times [0.1%]         ->  TEXT PASS
  1 times [0.1%]         ->  TIMEOUT PASS

Of those 1475 runs it didn't match the expectation 3 times [0.2%]
Details of the type and frequency of unmatched expectations:
  2 times [0.1%]         ->  TEXT PASS (Expected: PASS)
  1 times [0.1%]         ->  TIMEOUT PASS (Expected: PASS)

Flakiness factor: 0.4%
    - This test was flaky on 0.4% of the 1475 runs analyzed, either giving different
      results within a single run, or a different result than in the previous run.


NOTE: glib.py sets GST_DEBUG to "*:ERROR" during tests which enables the new log handling from 309748@main. Alongside that, the event order checks might already be time sensitive, given the existing failure.

[Lauro Moura]

Test gardening pull request: https://github.com/WebKit/WebKit/pull/61170

[EWS]

Test gardening commit 309773@main (e167c5b6fd82): <https://commits.webkit.org/309773@main>

Reviewed commits have been landed. Closing PR #61170 and removing active labels.

[Philippe Normand]

Pull request: https://github.com/WebKit/WebKit/pull/61187

[EWS]

Committed 309828@main (a6502dd669f6): <https://commits.webkit.org/309828@main>

Reviewed commits have been landed. Closing PR #61187 and removing active labels.

[Radar WebKit Bug Importer]

<rdar://problem/173231383>

[Philippe Normand]

:(

[Philippe Normand]

Re-flakied in https://github.com/WebKit/WebKit/pull/61258

[Philippe Normand]

I reproduced the issue here, will send another PR.

[Philippe Normand]

Pull request: https://github.com/WebKit/WebKit/pull/61335

[EWS]

Committed 309979@main (ad2137dc2304): <https://commits.webkit.org/309979@main>

Reviewed commits have been landed. Closing PR #61335 and removing active labels.