Bug 146653

Summary: Revert use of SVG <mask> elements for -webkit-mask-image (r176798, r177494, r186180)
Product: WebKit Reporter: Simon Fraser (smfr) <simon.fraser>
Component: Layout and RenderingAssignee: Simon Fraser (smfr) <simon.fraser>
Status: RESOLVED FIXED    
Severity: Normal CC: cdumez, commit-queue, dino, japhet, koivisto, simon.fraser
Priority: P2 Keywords: InRadar
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Patch
dino: review+
Patch
dino: review+
Patch dino: review+

Description Simon Fraser (smfr) 2015-07-06 15:10:15 PDT 
Need to revert these changes, because of the number of regressions: bug 141857, bug 146509, bug 146561.
Comment 1 Simon Fraser (smfr) 2015-07-06 15:23:09 PDT 
rdar://problem/21435233
Comment 2 Simon Fraser (smfr) 2015-07-06 17:52:03 PDT 
Created attachment 256262 [details]
Patch
Comment 3 Simon Fraser (smfr) 2015-07-06 17:54:23 PDT 
Created attachment 256263 [details]
Patch
Comment 4 Simon Fraser (smfr) 2015-07-06 17:56:31 PDT 
Created attachment 256266 [details]
Patch
Comment 5 Simon Fraser (smfr) 2015-07-06 17:56:56 PDT 
Patches are interdependent, so patches 2 and 3 won't apply.
Comment 6 Dean Jackson 2015-07-06 17:58:03 PDT 
Comment on attachment 256262 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=256262&action=review

> Source/WebCore/platform/graphics/MaskImageOperation.cpp:222
> -
> +    

:(
Comment 7 Dean Jackson 2015-07-06 18:01:54 PDT 
Comment on attachment 256263 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=256263&action=review

> Source/WebCore/css/CSSValue.h:174
>          SVGColorClass,
>          SVGPaintClass,
> -        WebKitCSSResourceClass,
>  

Not sure this one is needed, but it doesn't matter.

> Source/WebCore/css/CSSValue.h:192
>  #endif
> +        WebKitCSSResourceClass
>  

See above.
Comment 8 Dean Jackson 2015-07-06 18:04:34 PDT 
Comment on attachment 256266 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=256266&action=review

> Source/WebCore/css/CSSValue.h:-192
> -        WebKitCSSResourceClass
> -

Oh, you got rid of it here.
Comment 10 Chris Dumez 2015-07-06 19:45:35 PDT 
This caused http/tests/misc/mask-image-accept.html to start failing:
https://build.webkit.org/results/Apple%20Win%207%20Release%20(Tests)/r186390%20(52826)/results.html
Comment 11 Chris Dumez 2015-07-06 19:48:52 PDT 
(In reply to comment #10)
> This caused http/tests/misc/mask-image-accept.html to start failing:
> https://build.webkit.org/results/Apple%20Win%207%20Release%20(Tests)/
> r186390%20(52826)/results.html

Looks like only the Windows bot is having trouble with this though. The test in question is using -webkit-mask-image but not SVG.
Comment 12 Chris Dumez 2015-07-06 19:52:19 PDT 
(In reply to comment #11)
> (In reply to comment #10)
> > This caused http/tests/misc/mask-image-accept.html to start failing:
> > https://build.webkit.org/results/Apple%20Win%207%20Release%20(Tests)/
> > r186390%20(52826)/results.html
> 
> Looks like only the Windows bot is having trouble with this though. The test
> in question is using -webkit-mask-image but not SVG.

Ah, never mind, the issue was that the bot had processed r186390 but not r186391 and r186392 yet. The test is now passing again on the Windows bot.