[Webkit-unassigned] [Bug 186638] [GTK] Tests that are failing on the EWS test queue

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=186638

--- Comment #17 from Carlos Alberto Lopez Perez <clopez at igalia.com> ---
(In reply to Adrian Perez from comment #16)
> (In reply to Michael Catanzaro from comment #14)
> > (In reply to Carlos Alberto Lopez Perez from comment #11)
> > > (In reply to Carlos Alberto Lopez Perez from comment #10)
> > > 
> > > > Regressions: Unexpected crashes (3)
> > > >   fast/hidpi/broken-image-icon-very-hidpi.html [ Crash ]
> > > >   http/tests/misc/cached-scripts.html [ Crash ]
> > > >   legacy-animation-engine/fast/css/getFloatValueForUnit.html [ Crash ]
> > 
> > Yes, as expected. I was planning to watch the EWS this weekend to try to
> > find backtraces for the crashing tests, but it was broken all weekend.
> > Anyway, I intend to help with this. I don't think it's too much effort to
> > take the backtrace and report a bug when EWS finds a crashing test.
> > 
> > > Crash log for fast/hidpi/broken-image-icon-very-hidpi.html obtained on the
> > > EWS bot: http://termbin.com/a4of
> > 
> > Weird... although you should probably install gnome-icon-theme, I think this
> > is a WebKit bug worth reporting. Our current code assumes that we have
> > missingImage icons in the GResource, which is apparently not true. See here:
> > 
> > static Ref<Image> loadMissingImageIconFromTheme(const char* name)
> > {
> >     // ...
> >     if (iconInfo) {
> >         // Is the icon installed?
> >         return WTFMove(icon);
> >     }
> > 
> >     return loadImageFromGResource(name);
> > }
> > 
> > That's not right, we shouldn't continue on to loadImageFromGResource if the
> > image is not guaranteed to be in the resource.
> 
> I think that WebKitGTK+ should include within itself a copy of the
> resources it uses which currently end up being loaded from the GNOME
> theme, to avoid the runtime dependency on the “gnome-icon-theme”
> package. Using icons which are part of GTK+ itself would still be
> fine, of course. Should we open a new bug for this?

Yes, please.

-- 
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/20180618/d9114a0d/attachment-0001.html>