[Webkit-unassigned] [Bug 176174] New: [GTK][WPE] Test media/event-queue-crash is flaky (race condition on load events) after r220052

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=176174

            Bug ID: 176174
           Summary: [GTK][WPE] Test media/event-queue-crash is flaky (race
                    condition on load events) after r220052
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: Unspecified
                OS: Unspecified
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: WebKit Gtk
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: clopez at igalia.com
                CC: bugs-noreply at webkitgtk.org, darin at apple.com

r220052 <https://trac.webkit.org/r220052> implemented some improvements regarding style resolution and layout, and (quoting the changelog) "Also fixes a bug where load events are delivered prematurely in some cases  when an object, embed, frame, or iframe element is still loading."

As a result of this, the expectation for the test media/event-queue-crash was changed to log a syntax error that before was not triggering:

--- a/LayoutTests/media/event-queue-crash-expected.txt
+++ b/LayoutTests/media/event-queue-crash-expected.txt
@@ -1,3 +1,4 @@
+CONSOLE MESSAGE: line 34: SyntaxError: Unexpected token '}'. Expected '(' to start an 'if' condition.
 When an element containing video is removed, WebKit should not crash.

 PASS. WebKit didn't crash.


On the GTK+/WPE ports this test is now flaky and gives different results depending on the bot:

1) On the GTK+ release test, this test is passing fine.
2) On the GTK+ wayland bot this test is giving the failure below (like the fix on r220052 didn't worked here) https://build.webkit.org/results/GTK%20Linux%2064-bit%20Release%20Wayland%20(Tests)/r220052%20(2288)/media/event-queue-crash-pretty-diff.html
3) On the WPE release test bot, this test is skipped (general skip of media/ tests), but running it manually happens the same than on GTK+ wayland.

4) On my laptop with GTK+ release, this test is either giving the same result than on GTK+ wayland, passing or crashing.

So my conclusion is that there is still some race condition with this test that affects the GTK+/WPE ports

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20170831/4df508e7/attachment-0001.html>