WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
71793
[Chromium] SVG test failures after
r99539
https://bugs.webkit.org/show_bug.cgi?id=71793
Summary
[Chromium] SVG test failures after r99539
Yuta Kitamura
Reported
2011-11-08 03:38:02 PST
Many IMAGE failures are observed after
r99539
. Probably we just need to rebaseline them, but for now I'm trying to suppress them with test_expectations.txt.
Attachments
Potential fix
(5.35 KB, patch)
2011-11-08 06:15 PST
,
Nikolas Zimmermann
zherczeg
: review+
Details
Formatted Diff
Diff
View All
Add attachment
proposed patch, testcase, etc.
Yuta Kitamura
Comment 1
2011-11-08 04:02:12 PST
Added the following results: // SVG failures after
r99539
. BUGWK71793 : css2.1/20110323/background-intrinsic-004.htm = IMAGE BUGWK71793 : css2.1/20110323/background-intrinsic-005.htm = IMAGE BUGWK71793 : fast/backgrounds/animated-svg-as-mask.html = IMAGE BUGWK71793 : svg/as-background-image/animated-svg-as-background.html = IMAGE BUGWK71793 : svg/as-background-image/background-image-preserveaspectRatio-support.html = IMAGE BUGWK71793 : svg/as-background-image/svg-as-background-2.html = IMAGE BUGWK71793 : svg/as-background-image/svg-as-background-4.html = IMAGE BUGWK71793 : svg/as-background-image/svg-as-background-with-relative-size.html = IMAGE BUGWK71793 : svg/as-background-image/svg-background-partial-redraw.html = IMAGE BUGWK71793 : svg/as-image/animated-svg-as-image.html = IMAGE BUGWK71793 : svg/as-image/img-preserveAspectRatio-support-1.html = IMAGE BUGWK71793 : svg/as-image/img-preserveAspectRatio-support-2.html = IMAGE BUGWK71793 : svg/as-image/same-image-two-instances.html = IMAGE BUGWK71793 : svg/as-image/svg-image-change-content-size.xhtml = IMAGE BUGWK71793 : svg/wicd/test-scalable-background-image1.xhtml = IMAGE BUGWK71793 : svg/wicd/test-scalable-background-image2.xhtml = IMAGE BUGWK71793 : svg/zoom/page/zoom-background-images.html = IMAGE BUGWK71793 : svg/zoom/page/zoom-svg-as-background-with-relative-size-and-viewBox.html = IMAGE BUGWK71793 : svg/zoom/page/zoom-svg-as-background-with-relative-size.html = IMAGE BUGWK71793 : fast/backgrounds/size/contain-and-cover-zoomed.html = IMAGE+TEXT BUGWK71793 WIN LINUX : svg/as-image/animated-svg-as-image-no-fixed-intrinsic-size.html = IMAGE+TEXT BUGWK71793 MAC : svg/as-image/animated-svg-as-image-no-fixed-intrinsic-size.html = IMAGE BUGWK71793 WIN LINUX : svg/as-image/animated-svg-as-image-same-image.html = IMAGE+TEXT BUGWK71793 MAC : svg/as-image/animated-svg-as-image-same-image.html = IMAGE BUGWK71793 : svg/zoom/page/zoom-coords-viewattr-01-b.svg = IMAGE+TEXT BUGWK71793 : svg/as-background-image/svg-as-background-1.html = IMAGE BUGWK71793 : svg/as-background-image/svg-as-background-3.html = IMAGE // Uncomment the above results when you remove any of the following lines. BUGWK71793 BUGCR83426 : svg/as-border-image/svg-as-border-image.html = IMAGE BUGWK71793 BUGWK71673 WIN LINUX : svg/zoom/page/zoom-img-preserveAspectRatio-support-1.html = PASS IMAGE IMAGE+TEXT This list may not be complete; let's watch the next cycle of the bots.
Nikolas Zimmermann
Comment 2
2011-11-08 04:27:41 PST
(In reply to
comment #1
)
> Added the following results:
Thanks. I've checked all of them using the flakiness dashboard.
> > // SVG failures after
r99539
. > BUGWK71793 : css2.1/20110323/background-intrinsic-004.htm = IMAGE > BUGWK71793 : css2.1/20110323/background-intrinsic-005.htm = IMAGE
Fine to rebaseline.
> BUGWK71793 : fast/backgrounds/animated-svg-as-mask.html = IMAGE
Fails on several bots, might be a timing problem. I'll investigate.
> BUGWK71793 : svg/as-background-image/animated-svg-as-background.html = IMAGE
Seems to be flakey on WebKit Linux (dbg)(2).
> BUGWK71793 : svg/as-background-image/background-image-preserveaspectRatio-support.html = IMAGE > BUGWK71793 : svg/as-background-image/svg-as-background-2.html = IMAGE > BUGWK71793 : svg/as-background-image/svg-as-background-4.html = IMAGE > BUGWK71793 : svg/as-background-image/svg-as-background-with-relative-size.html = IMAGE > BUGWK71793 : svg/as-background-image/svg-background-partial-redraw.html = IMAGE
Safe to rebaseline.
> BUGWK71793 : svg/as-image/animated-svg-as-image.html = IMAGE
Same bug as animated-svg-as-background-image.html, it's flakey on the dbg slaves only! That means its for sure timing dependant.
> BUGWK71793 : svg/as-image/img-preserveAspectRatio-support-1.html = IMAGE > BUGWK71793 : svg/as-image/img-preserveAspectRatio-support-2.html = IMAGE > BUGWK71793 : svg/as-image/same-image-two-instances.html = IMAGE > BUGWK71793 : svg/as-image/svg-image-change-content-size.xhtml = IMAGE > BUGWK71793 : svg/wicd/test-scalable-background-image1.xhtml = IMAGE > BUGWK71793 : svg/wicd/test-scalable-background-image2.xhtml = IMAGE > BUGWK71793 : svg/zoom/page/zoom-background-images.html = IMAGE > BUGWK71793 : svg/zoom/page/zoom-svg-as-background-with-relative-size-and-viewBox.html = IMAGE > BUGWK71793 : svg/zoom/page/zoom-svg-as-background-with-relative-size.html = IMAGE > BUGWK71793 : fast/backgrounds/size/contain-and-cover-zoomed.html = IMAGE+TEXT > BUGWK71793 WIN LINUX : svg/as-image/animated-svg-as-image-no-fixed-intrinsic-size.html = IMAGE+TEXT > BUGWK71793 MAC : svg/as-image/animated-svg-as-image-no-fixed-intrinsic-size.html = IMAGE
> BUGWK71793 WIN LINUX : svg/as-image/animated-svg-as-image-same-image.html = IMAGE+TEXT
> BUGWK71793 MAC : svg/as-image/animated-svg-as-image-same-image.html = IMAGE > BUGWK71793 : svg/zoom/page/zoom-coords-viewattr-01-b.svg = IMAGE+TEXT > BUGWK71793 : svg/as-background-image/svg-as-background-1.html = IMAGE > BUGWK71793 : svg/as-background-image/svg-as-background-3.html = IMAGE
Safe to rebaseline.
> // Uncomment the above results when you remove any of the following lines. > BUGWK71793 BUGCR83426 : svg/as-border-image/svg-as-border-image.html = IMAGE > BUGWK71793 BUGWK71673 WIN LINUX : svg/zoom/page/zoom-img-preserveAspectRatio-support-1.html = PASS IMAGE IMAGE+TEXT
Didn't check these.
> This list may not be complete; let's watch the next cycle of the bots.
Ok. I will see if I can find a way to make the sampling of the animated SVG tests safer.
Yuta Kitamura
Comment 3
2011-11-08 04:52:06 PST
Added one more: BUGWK71793 MAC : fast/backgrounds/svg-as-mask.html = IMAGE
Yuta Kitamura
Comment 4
2011-11-08 05:11:22 PST
Three more: BUGWK71793 MAC CPU-CG : fast/backgrounds/size/contain-and-cover.html = IMAGE BUGWK71793 MAC CPU-CG : svg/as-background-image/svg-as-background-5.html = IMAGE BUGWK71793 MAC CPU-CG : svg/custom/embedding-external-svgs.xhtml = IMAGE
Nikolas Zimmermann
Comment 5
2011-11-08 06:15:07 PST
Created
attachment 114055
[details]
Potential fix Remove the need to run a 100ms animation, instead use <set> and start immediately. I hope that reduces/eliminates the flakiness. Note, that we have no way to really wait for the SVGImage to load, we have no scriptable access to the underlying SVG document, unlike in eg. <iframe>s.
Zoltan Herczeg
Comment 6
2011-11-08 06:17:30 PST
Comment on
attachment 114055
[details]
Potential fix r=me
Nikolas Zimmermann
Comment 7
2011-11-08 06:20:10 PST
Landed "Potential fix" in
r99562
. Let's see if that gives us stable baselines.
Nikolas Zimmermann
Comment 8
2011-11-08 07:48:23 PST
(In reply to
comment #7
)
> Landed "Potential fix" in
r99562
. Let's see if that gives us stable baselines.
Also commited other fixes, see
bug 71253
that also help Chromium, I hope. I guess we have to wait for the next morning to see results from all bots in the flakiness dashboard, hopefully flakiness is just gone...
Yuta Kitamura
Comment 9
2011-11-09 23:29:46 PST
I'm going to rebaseline ones that Nikolas said safe. Flakiness seem to be gone, but I'll handle flaky ones in a later round.
Yuta Kitamura
Comment 10
2011-11-10 23:49:28 PST
Okay, I finished to rebaseline all the results changed. The only outstanding issue seems to be the flakiness of svg/zoom/page/zoom-img-preserveAspectRatio-support-1.html, which is tracked in
bug 71673
. Issue resolved, closing the bug.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug