Open Bug 877661 Opened 7 years ago Updated Last year

Intermittent layout/reftests/svg/svg-integration/mask-html-01-extref-02.xhtml | image comparison (==), max difference: 255, number of differing pixels: 49528

Categories

(Core :: SVG, defect, P3)

x86
All
defect

Tracking

()

Tracking Status
firefox31 --- disabled
firefox32 --- disabled
firefox33 --- disabled
firefox-esr24 --- unaffected
b2g-v1.4 --- disabled
b2g-v2.0 --- disabled
b2g-v2.1 --- disabled

People

(Reporter: RyanVM, Unassigned)

References

(Blocks 2 open bugs)

Details

(Keywords: intermittent-failure, Whiteboard: [stockwell fixed:other])

Attachments

(5 files)

Attached image test screenshot
https://tbpl.mozilla.org/php/getParsedLog.php?id=23565211&tree=Mozilla-Inbound

Windows 7 32-bit mozilla-inbound debug test reftest on 2013-05-29 18:59:20 PDT for push c9e6ca6528b8
slave: t-w732-ix-024

19:27:03     INFO -  REFTEST TEST-START | file:///C:/slave/test/build/tests/reftest/tests/layout/reftests/svg/svg-integration/mask-html-01-extref-02.xhtml | 7768 / 9341 (83%)
19:27:03     INFO -  ++DOMWINDOW == 222 (0CC4E090) [serial = 22178] [outer = 056BFC40]
19:27:04     INFO -  WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80004005: file e:/builds/moz2_slave/m-in-w32-d-0000000000000000000/build/layout/base/nsCSSFrameConstructor.cpp, line 3804
19:27:04     INFO -  WARNING: Unable to get an nsIScriptGlobalObjectOwner from the channel!: file e:/builds/moz2_slave/m-in-w32-d-0000000000000000000/build/dom/src/jsurl/nsJSProtocolHandler.cpp, line 125
19:27:04     INFO -  WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80040111: file e:/builds/moz2_slave/m-in-w32-d-0000000000000000000/build/layout/style/Loader.cpp, line 1971
19:27:04     INFO -  REFTEST TEST-UNEXPECTED-FAIL | file:///C:/slave/test/build/tests/reftest/tests/layout/reftests/svg/svg-integration/mask-html-01-extref-02.xhtml | image comparison (==), max difference: 255, number of differing pixels: 49528
19:27:04     INFO -  REFTEST   IMAGE 1 (TEST): <see attached>
19:27:04     INFO -  REFTEST   IMAGE 2 (REFERENCE): <see attached>
19:27:04     INFO -  REFTEST INFO | Saved log: START file:///C:/slave/test/build/tests/reftest/tests/layout/reftests/svg/svg-integration/mask-html-01-extref-02.xhtml
19:27:04     INFO -  REFTEST INFO | Saved log: [CONTENT] OnDocumentLoad triggering AfterOnLoadScripts
19:27:04     INFO -  REFTEST INFO | Saved log: Initializing canvas snapshot
19:27:04     INFO -  REFTEST INFO | Saved log: DoDrawWindow 0,0,800,1000
19:27:04     INFO -  REFTEST INFO | Saved log: [CONTENT] RecordResult fired
19:27:04     INFO -  REFTEST INFO | Saved log: RecordResult fired
19:27:04     INFO -  REFTEST INFO | Saved log: RecordResult fired
19:27:04     INFO -  REFTEST INFO | Loading a blank page
19:27:04     INFO -  ++DOMWINDOW == 223 (093C4578) [serial = 22179] [outer = 056BFC40]
Attached image reference screenshot
Blocks: 813742
Jonathan, any chance you could take a look at this ongoing intermittent reftest failure?
Flags: needinfo?(jwatt)
We're getting plenty of Ubuntu VM failures in addition to Windows failures.
Flags: needinfo?(jwatt)
OS: Windows 7 → All
This is happening because mask-html-01-resource.svg starts with this PI:

  <?xml-stylesheet
     type="text/css"
     href="data:text/css,circle { fill: white }" ?>

and we rely on that style applying for the test to pass. It seems that sometimes it does not, though, and so the circle isn't included in the mask due to it being black instead of white.

This test was added by bz in bug 433616. Boris, any ideas on how to delay the test to make sure this data: URI has time to load?
Flags: needinfo?(bzbarsky)
I'd expect it to just work.  That is, I would expect that data: URI load to block onload for the document that linked to the external resource.  At least assuming the external resource load is started before the main document has fired onload, which is what the <use> hack there is about.

Can we reproduce this reliably if we replace the data: URI with an http:// URI that has a 2s delay before responding, say?  If so, I guess we get to debug why the onload blocking isn't working right...
Flags: needinfo?(bzbarsky)
Jonathan, did you have a chance to look into comment 157 at all? This affects all branches and is currently one of the top oranges on mozilla-beta, which is destined to be our next ESR release.
Flags: needinfo?(jwatt)
Test marked random on all platforms.

https://hg.mozilla.org/integration/mozilla-inbound/rev/237f4e7f34c5
Flags: needinfo?(jwatt) → needinfo?(ryanvm)
Keywords: leave-open
Whiteboard: [test disabled]
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3
Assignee: nobody → cku
Blocks: 1350793
Attachment #8851449 - Flags: review?(tnikkel)
Attachment #8851460 - Flags: review?(tnikkel)
Attachment #8851450 - Flags: review?(tnikkel)
Comment on attachment 8851449 [details]
Bug 877661 - Part 1. Pass decode flag down to GetMaskForMaskedFrame.

https://reviewboard.mozilla.org/r/123740/#review126166

::: layout/svg/nsSVGIntegrationUtils.cpp:444
(Diff revision 1)
>      if (maskFrame) {
>        Matrix svgMaskMatrix;
>        nsSVGMaskFrame::MaskParams params(maskContext, aParams.frame,
>                                                    cssPxToDevPxMatrix,
>                                                    aOpacity, &svgMaskMatrix,
> -                                                  svgReset->mMask.mLayers[i].mMaskMode);
> +                                                  svgReset->mMask.mLayers[i].mMaskMode, aParams.flags);

The indentation seems messed up here?
Attachment #8851449 - Flags: review?(tnikkel) → review+
Comment on attachment 8851450 [details]
Bug 877661 - Part 3. Enable mask-html-01-extref-02.xhtml.

https://reviewboard.mozilla.org/r/123742/#review126168

Assuming this works after many retriggers on try server.
Attachment #8851450 - Flags: review?(tnikkel) → review+
Comment on attachment 8851460 [details]
Bug 877661 - Part 2. Pass decode flag from outer-svg/ svg-container paint call to nsSVGUtils::PaintFrameWithEffects.

https://reviewboard.mozilla.org/r/123744/#review126170
Attachment #8851460 - Flags: review?(tnikkel) → review+