Bug 63268
| Summary: | http/tests/multipart/win-boundary-crash.html flakey | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Jessie Berlin <jberlin> |
| Component: | Tools / Tests | Assignee: | Fujii Hironori <Hironori.Fujii> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | abarth, aroben, jberlin, mjs, tony, webkit-bug-importer |
| Priority: | P2 | Keywords: | InRadar, LayoutTestFailure, PlatformOnly, Regression |
| Version: | 528+ (Nightly build) | ||
| Hardware: | All | ||
| OS: | All | ||
| URL: | http://build.webkit.org/results/Windows 7 Release (Tests)/r89582 (14084)/results.html | ||
Jessie Berlin
http://build.webkit.org/results/Windows%207%20Release%20(Tests)/r89580%20(14082)/results.html
http://build.webkit.org/builders/Windows%207%20Release%20%28Tests%29/builds/14083
http://build.webkit.org/results/Windows%207%20Release%20(Tests)/r89587%20(14085)/results.html
Not entirely sure when the flakiness began. I will add it to the skipped list shortly.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Jessie Berlin
Added to the Windows skipped list in http://trac.webkit.org/changeset/89596
Jessie Berlin
And now it is being flakey on the mac bots:
http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r89562%20(30682)/results.html
http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20%28Tests%29/builds/30683
http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r89593%20(30692)/results.html
I will move the entry to the mac Skipped list.
Adam Roben (:aroben)
(In reply to comment #2)
> I will move the entry to the mac Skipped list.
Seems like you need it to be on both mac/Skipped and win/Skipped. (Windows WebKit1 doesn't use mac/Skipped.)
Or maybe we should disable the test by renaming it.
Jessie Berlin
(In reply to comment #3)
> (In reply to comment #2)
> > I will move the entry to the mac Skipped list.
>
> Seems like you need it to be on both mac/Skipped and win/Skipped. (Windows WebKit1 doesn't use mac/Skipped.)
>
> Or maybe we should disable the test by renaming it.
I thought the win skipped list always inherited from the mac skipped list.
Moved to the mac skipped list in http://trac.webkit.org/changeset/89600.
Jessie Berlin
(In reply to comment #4)
> (In reply to comment #3)
> > (In reply to comment #2)
> > > I will move the entry to the mac Skipped list.
> >
> > Seems like you need it to be on both mac/Skipped and win/Skipped. (Windows WebKit1 doesn't use mac/Skipped.)
> >
> > Or maybe we should disable the test by renaming it.
>
> I thought the win skipped list always inherited from the mac skipped list.
>
> Moved to the mac skipped list in http://trac.webkit.org/changeset/89600.
Adam showed me that it was not, so I went ahead and just disabled the test in http://trac.webkit.org/changeset/89602
Tony Chang
This is super weird. Here's the test html:
http://trac.webkit.org/browser/trunk/LayoutTests/http/tests/multipart/win-boundary-crash.html-disabled
1 <html>
2 <head>
3 <script>
4 if (window.layoutTestController)
5 layoutTestController.dumpAsText();
6 </script>
7 </head>
8 <body>
9 <p>Test for <i><a href="http://bugs.webkit.org/show_bug.cgi?id=14516">http://bugs.webkit.org/show_bug.cgi?id=14516</a>
10 crash loading multipart/x-mixed-replace data on windows safari</i></p>
11 <p>If WebKit does not crash when loading this img, then it passed.</p>
12 <img src="resources/multipart-nodashes.php" />
13 </body>
14 </html>
The diff shows that dumpAsText() didn't get called, but I don't see how that's possible.
cc'ing abarth since this sounds similar to https://bugs.webkit.org/show_bug.cgi?id=63057 .
Adam Roben (:aroben)
(In reply to comment #6)
> The diff shows that dumpAsText() didn't get called, but I don't see how that's possible.
No, the diff shows that the some other test got dumped instead. Look more closely and you'll see that the text is entirely different.
Tony Chang
(In reply to comment #7)
> (In reply to comment #6)
> > The diff shows that dumpAsText() didn't get called, but I don't see how that's possible.
>
> No, the diff shows that the some other test got dumped instead. Look more closely and you'll see that the text is entirely different.
Ah, sorry, you're right, it's the output from the previous test (stop-crash.html). This could be a problem with ORWT or maybe one of the previous tests has an http connection that's hanging or something.
Fujii Hironori
Pull request: https://github.com/WebKit/WebKit/pull/19461
EWS
Test gardening commit 269704@main (237bd74e5ced): <https://commits.webkit.org/269704@main>
Reviewed commits have been landed. Closing PR #19461 and removing active labels.
Fujii Hironori
No flaky after 269704@main.
https://results.webkit.org/?suite=layout-tests&test=http%2Ftests%2Fmultipart%2Fwin-boundary-crash.html
Filed a new ticket for the GTK and WPE crashes.
Bug 263629 – [GTK][WPE] http/tests/multipart/win-boundary-crash.html is crashing: soup_multipart_input_stream_read_headers: assertion 'got_lf' failed
Radar WebKit Bug Importer
<rdar://problem/117447848>