[webkit-dev] svg/filters/filter-empty-g.svg ASSERTs on Leopard Intel Debug (Tests)

[Adam Barth]

I'm not sure it's working properly.  It says:

SUCCESS: Build 17401 (r63531) was the first to show failures:
set([u'svg/filters/filter-empty-g.svg'])

but then it goes on to list all the commits from r63492 to r63531.

Adam


On Sun, Jul 18, 2010 at 9:01 AM, Eric Seidel <eric at webkit.org> wrote:
> Won't "webkit-patch failure-reason" tell you when it started?  It
> knows how to handle looking past the end of the waterfall.
>
> -eric
>
> On Sun, Jul 18, 2010 at 8:48 AM, Adam Barth <abarth at webkit.org> wrote:
>> A test has been crashing Leopard Intel Debug (Tests) for >24 hours.  I
>> tried to look back on buildbot to see when the issue started, but it's
>> past the end of the waterfall (much like a rainbow).  If you touched
>> SVG filters recently, would you be willing to investigate?
>>
>> Thanks,
>> Adam
>>
>>
>> Tests that caused the DumpRenderTree tool to crash:
>>
>> svg/filters/filter-empty-g.svg  stderr
>>
>> Sun Jul 18 08:33:06 il0301a-dhcp09.apple.com DumpRenderTree[17669]
>> <Error>: CGImageCreate: invalid image size: 0 x 0.
>> ASSERTION FAILED: frameCount()
>> (/Volumes/Big/WebKit-BuildSlave/leopard-intel-debug/build/WebCore/platform/graphics/cg/ImageSourceCG.cpp:273
>> bool WebCore::ImageSource::frameIsCompleteAtIndex(size_t))
>> _______________________________________________
>> webkit-dev mailing list
>> webkit-dev at lists.webkit.org
>> http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
>>
>