Bug 79760
| Summary: | [GTK] media/video-poster-blocked-by-willsendrequest.html is flaky | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Philippe Normand <pnormand> |
| Component: | WebKitGTK | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | Normal | CC: | eric.carlson |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
Philippe Normand
I think it's because it calls layoutTestController.setWillSendRequestReturnsNull(true) and doesn't reset this setting before bailing out.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Philippe Normand
(In reply to comment #0)
> I think it's because it calls layoutTestController.setWillSendRequestReturnsNull(true) and doesn't reset this setting before bailing out.
Ah well I'm not sure. It seems the poster data retrieval is not blocked for the first test run but for the second pass it works.
Here's the diff, for the record:
--- /home/slave/webkitgtk/gtk-linux-64-release/build/layout-test-results/media/video-poster-blocked-by-willsendrequest-expected.txt
+++ /home/slave/webkitgtk/gtk-linux-64-release/build/layout-test-results/media/video-poster-blocked-by-willsendrequest-actual.txt
@@ -7,7 +7,7 @@
EXPECTED (video.clientHeight == '150') OK
Test after setting poster
-EXPECTED (video.clientWidth == '300') OK
-EXPECTED (video.clientHeight == '150') OK
+EXPECTED (video.clientWidth == '300'), OBSERVED '76' FAIL
+EXPECTED (video.clientHeight == '150'), OBSERVED '103' FAIL
END OF TEST
Eric Carlson
(In reply to comment #1)
> (In reply to comment #0)
> > I think it's because it calls layoutTestController.setWillSendRequestReturnsNull(true) and doesn't reset this setting before bailing out.
>
> Ah well I'm not sure. It seems the poster data retrieval is not blocked for the first test run but for the second pass it works.
>
The first test is done before setting video.poster, so the element should have the default (300x150) size.
The second test is done after setting the poster, but "layoutTestController.setWillSendRequestReturnsNull(true)" *should* block the poster load so the element's size *should* be the same as before. The size difference indicates that the poster loaded and therefore the element's size changed.
Eric Carlson
This test is now flakey on Chromium as well, https://bugs.webkit.org/show_bug.cgi?id=75161 says it started to fail after http://trac.webkit.org/log/?verbose=on&rev=103130&stop_rev=103130
Philippe Normand
Alright, marking as dupe then. Thanks Eric.
*** This bug has been marked as a duplicate of bug 75161 ***