Bug 132569 – Adapt inline SVG sizing behavior to Firefox and Blink

• Component: SVG
• Status: RESOLVED FIXED

[Dirk Schulze]

During the SVG F2F in Leipzig, the SVG WG agreed on a model to size inline SVGs in HTML content. The model is following the behavior of CSS2.1 for replaced elements and already implemented in Firefox. Blink followed the behavior of Firefox http://src.chromium.org/viewvc/blink?limit_changes=0&view=revision&revision=171958 .

An extensive test suite can be found here: http://w3c-test.org/html/rendering/replaced-elements/svg-inline-sizing/svg-inline.html

[Dirk Schulze]

Created attachment 230834 [details]
Checking for failing tests on bots

Not for review.

[WebKit Commit Bot]

Attachment 230834 [details] did not pass style-queue:


ERROR: Source/WebCore/rendering/RenderReplaced.cpp:287:  Multi line control clauses should use braces.  [whitespace/braces] [4]
ERROR: Source/WebCore/rendering/svg/RenderSVGRoot.cpp:91:  One line control clauses should not use braces.  [whitespace/braces] [4]
ERROR: Source/WebCore/svg/SVGSVGElement.cpp:289:  One line control clauses should not use braces.  [whitespace/braces] [4]
Total errors found: 3 in 110 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Build Bot]

Comment on attachment 230834 [details]
Checking for failing tests on bots

Attachment 230834 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/5976985020923904

New failing tests:
svg/text/non-bmp-positioning-lists.svg
fast/css/infinite-floating-value.html
svg/custom/object-sizing-width-50p-height-50p-on-target-svg.xhtml
svg/custom/object-sizing-width-75p-height-50p-on-target-svg.xhtml
fast/css3-text/css3-word-spacing-percentage/word-spacing-crash.html
svg/custom/object-sizing-width-50p-on-target-svg.xhtml
fast/css/remove-fixed-resizer-crash.html
svg/custom/object-sizing-width-50p-height-75p-on-target-svg-absolute.xhtml
accessibility/svg-image.html
css2.1/20110323/replaced-intrinsic-001.htm
svg/custom/object-sizing-height-50p-on-target-svg-absolute.xhtml
css2.1/20110323/replaced-intrinsic-002.htm
http/tests/xmlviewer/dumpAsText/svg.xml
svg/hixie/intrinsic/002.html
svg/custom/object-sizing-width-50p-height-50p-on-target-svg-absolute.xhtml
svg/custom/object-sizing-width-75p-height-50p-on-target-svg-absolute.xhtml
svg/custom/object-sizing-height-50p-on-target-svg.xhtml
svg/hixie/intrinsic/001.html
svg/custom/object-sizing-width-50p-height-75p-on-target-svg.xhtml
fast/shapes/shape-outside-floats/shape-outside-relative-size-svg.html

[Build Bot]

Created attachment 230840 [details]
Archive of layout-test-results from webkit-ews-01 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-01  Port: mac-mountainlion  Platform: Mac OS X 10.8.5

[Build Bot]

Comment on attachment 230834 [details]
Checking for failing tests on bots

Attachment 230834 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/6008755699318784

New failing tests:
svg/text/non-bmp-positioning-lists.svg
fast/css/infinite-floating-value.html
svg/custom/object-sizing-width-50p-height-50p-on-target-svg.xhtml
svg/custom/object-sizing-width-75p-height-50p-on-target-svg.xhtml
fast/css3-text/css3-word-spacing-percentage/word-spacing-crash.html
svg/custom/object-sizing-width-50p-on-target-svg.xhtml
fast/css/remove-fixed-resizer-crash.html
svg/custom/object-sizing-width-50p-height-75p-on-target-svg-absolute.xhtml
accessibility/svg-image.html
css2.1/20110323/replaced-intrinsic-001.htm
svg/custom/object-sizing-height-50p-on-target-svg-absolute.xhtml
css2.1/20110323/replaced-intrinsic-002.htm
http/tests/xmlviewer/dumpAsText/svg.xml
svg/hixie/intrinsic/002.html
svg/custom/object-sizing-width-50p-height-50p-on-target-svg-absolute.xhtml
svg/custom/object-sizing-width-75p-height-50p-on-target-svg-absolute.xhtml
svg/custom/object-sizing-height-50p-on-target-svg.xhtml
svg/hixie/intrinsic/001.html
svg/custom/object-sizing-width-50p-height-75p-on-target-svg.xhtml
fast/shapes/shape-outside-floats/shape-outside-relative-size-svg.html

[Build Bot]

Created attachment 230843 [details]
Archive of layout-test-results from webkit-ews-10 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-10  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5

[Dirk Schulze]

Created attachment 230846 [details]
Patch

[Dirk Schulze]

Created attachment 230850 [details]
Patch

[Dirk Schulze]

Created attachment 230851 [details]
Patch

[Dean Jackson]

Comment on attachment 230851 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=230851&action=review

> Source/WebCore/ChangeLog:14
> +        "The basis of this change is to map explicit width and height
> +        attributes to CSS properties, essentially promoting them to
> +        presentation attributes. Note that implicit "100%" width and height
> +        are not mapped.

You're missing an end quote mark.

> Source/WebCore/ChangeLog:39
> +        (svg:root):

Either remove this line or put the text above on it.

> Source/WebCore/ChangeLog:45
> +            And so it SVGSVGElement.

"is"

[WebKit Commit Bot]

Comment on attachment 230851 [details]
Patch

Clearing flags on attachment: 230851

Committed r168350: <http://trac.webkit.org/changeset/168350>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Nikolas Zimmermann]

Wonderful patch :-) Finally that weird percentage intrinsic sizing is gone!

[Simon Fraser (smfr)]

Does this fix the weird spacing around the images at http://dbaron.org/log/20110225-blur-radius ?

[Dirk Schulze]

(In reply to comment #14)
> Does this fix the weird spacing around the images at http://dbaron.org/log/20110225-blur-radius ?

Yes. Indeed it does.

[Alexey Proskuryakov]

This patch removed results for svg/text/non-bmp-positioning-lists.svg. Could you please re-land them?

http://webkit-test-results.appspot.com/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=svg%2Ftext%2Fnon-bmp-positioning-lists.svg

[Alexey Proskuryakov]

Also, svg/as-object/sizing/svg-in-object-placeholder-height-percentage.html is flaky on some bots, and fails 100% of time on others.

[Alexey Proskuryakov]

Actually, even more:

svg/as-object/sizing/svg-in-object-placeholder-height-fixed.html
svg/as-object/sizing/svg-in-object-placeholder-height-percentage.html
svg/as-object/sizing/svg-in-object-placeholder-height-auto.html


http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-fixed-diff.txt

http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-percentage-diff.txt

http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-auto-diff.txt

[Dirk Schulze]

Can't produce the svg-object test errors locally.

What does

Harness Error. harness_status.status = 2 , harness_status.message = null

mean? How can that be caused?

[Dirk Schulze]

(In reply to comment #18)
> Actually, even more:
> 
> svg/as-object/sizing/svg-in-object-placeholder-height-fixed.html
> svg/as-object/sizing/svg-in-object-placeholder-height-percentage.html
> svg/as-object/sizing/svg-in-object-placeholder-height-auto.html
> 
> 
> http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-fixed-diff.txt
> 
> http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-percentage-diff.txt
> 
> http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-auto-diff.txt

The tests need some time for running. Could these be timeout errors?

[Dirk Schulze]

(In reply to comment #20)
> (In reply to comment #18)
> > Actually, even more:
> > 
> > svg/as-object/sizing/svg-in-object-placeholder-height-fixed.html
> > svg/as-object/sizing/svg-in-object-placeholder-height-percentage.html
> > svg/as-object/sizing/svg-in-object-placeholder-height-auto.html
> > 
> > 
> > http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-fixed-diff.txt
> > 
> > http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-percentage-diff.txt
> > 
> > http://build.webkit.org/results/Apple%20MountainLion%20(Leaks)/r168513%20(10394)/svg/as-object/sizing/svg-in-object-placeholder-height-auto-diff.txt
> 
> The tests need some time for running. Could these be timeout errors?

Oh found it: status = 2

means         status_text[Test.prototype.TIMEOUT] = "Timeout";

How can I give the tests more time for completion?

[Alexey Proskuryakov]

I don't think that there is a way, the limit of 30 seconds is quite high already. We do not really want tests that take over 30 seconds to run each!

[Alexey Proskuryakov]

(In reply to comment #16)
> This patch removed results for svg/text/non-bmp-positioning-lists.svg. Could you please re-land them?

Someone landed results in <http://trac.webkit.org/changeset/168543>, but looks like it happened by accident. Are the landed results correct?

> Actually, even more:

Filed bug 132791, and skipped the tests.

Dirk, please be more responsive when you make bots red. You knew about the failures for two days now.

[Dirk Schulze]

(In reply to comment #23)
> (In reply to comment #16)
> > This patch removed results for svg/text/non-bmp-positioning-lists.svg. Could you please re-land them?
> 
> Someone landed results in <http://trac.webkit.org/changeset/168543>, but looks like it happened by accident. Are the landed results correct?
> 
> > Actually, even more:
> 
> Filed bug 132791, and skipped the tests.
> 
> Dirk, please be more responsive when you make bots red. You knew about the failures for two days now.

I apologize. I try to be more responsive. Needed to work out CSS Filters and device pixel ratio issues the last days. Will take care of the bug tomorrow and the day after.

[alan]

This regressed bug 133933