Bug 83960 - [CSS Filters] Do not use clipping rect when calculating the bounds of a layer
: [CSS Filters] Do not use clipping rect when calculating the bounds of a layer
Status: RESOLVED FIXED
: WebKit
Layout and Rendering
: 528+ (Nightly build)
: Unspecified Unspecified
: P2 Normal
Assigned To:
:
:
:
:
  Show dependency treegraph
 
Reported: 2012-04-13 16:38 PST by
Modified: 2012-04-18 09:37 PST (History)


Attachments
Patch V1 (5.92 KB, patch)
2012-04-13 17:29 PST, Alexandru Chiculita
simon.fraser: review+
simon.fraser: commit‑queue-
Review Patch | Details | Formatted Diff | Diff


Note

You need to log in before you can comment on or make changes to this bug.


Description From 2012-04-13 16:38:24 PST
WebKit should not clip the bounding box of a layer when it draws it into a memory buffer in order to compute the CSS Filter for it. That's because the drop-shadow or blur effects might reference pixels outside the initial clipping box.

This should fix the results for two tests that are already checked in:
LayoutTests/css3/filters/filter-repaint-shadow-clipped.html
LayoutTests/css3/filters/filter-repaint-shadow-rotated.html
------- Comment #1 From 2012-04-13 17:29:59 PST -------
Created an attachment (id=137184) [details]
Patch V1
------- Comment #2 From 2012-04-13 17:36:27 PST -------
(From update of attachment 137184 [details])
View in context: https://bugs.webkit.org/attachment.cgi?id=137184&action=review

> Source/WebCore/ChangeLog:14
> +        No new tests added in this patch, but this patch fixes two existing tests that fail.
> +        LayoutTests/css3/filters/filter-repaint-shadow-clipped.html
> +        LayoutTests/css3/filters/filter-repaint-shadow-rotated.html

Do these tests need to be unskipped or given new pixel results?
------- Comment #3 From 2012-04-13 19:50:09 PST -------
(From update of attachment 137184 [details])
Attachment 137184 [details] did not pass efl-ews (efl):
Output: http://queues.webkit.org/results/12406117
------- Comment #4 From 2012-04-13 20:45:06 PST -------
(From update of attachment 137184 [details])
Attachment 137184 [details] did not pass gtk-ews (gtk):
Output: http://queues.webkit.org/results/12408037
------- Comment #5 From 2012-04-13 23:07:36 PST -------
(From update of attachment 137184 [details])
Attachment 137184 [details] did not pass gtk-ews (gtk):
Output: http://queues.webkit.org/results/12408069
------- Comment #6 From 2012-04-17 17:10:54 PST -------
I need to also update the #else branch. I will do that before committing.
------- Comment #7 From 2012-04-18 09:36:50 PST -------
(In reply to comment #2)
> (From update of attachment 137184 [details] [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=137184&action=review
> 
> > Source/WebCore/ChangeLog:14
> > +        No new tests added in this patch, but this patch fixes two existing tests that fail.
> > +        LayoutTests/css3/filters/filter-repaint-shadow-clipped.html
> > +        LayoutTests/css3/filters/filter-repaint-shadow-rotated.html
> 
> Do these tests need to be unskipped or given new pixel results?

The results seem to be up-to-date and no failure in Skipped tests were added.
------- Comment #8 From 2012-04-18 09:37:04 PST -------
Landed in http://trac.webkit.org/changeset/114518 .