Bug 93481 – Accelerated canvas elements shouldn't share style

• Component: Canvas
• Status: UNCONFIRMED

[Filip Spacek]

Created attachment 157216 [details]
Don't share styles for accelerated canvases

Style sharing causes style recalculation to short circuit which results in the canvas elements not getting their own layer when they become accelerated.

[WebKit Review Bot]

Attachment 157216 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCor..." exit_code: 1
Source/WebCore/ChangeLog:1:  ChangeLog entry has no bug number  [changelog/bugnumber] [5]
Total errors found: 1 in 2 files


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

[Filip Spacek]

Created attachment 157217 [details]
Don't share styles for accelerated canvases

Add bug number

[George Staikos]

Comment on attachment 157217 [details]
Don't share styles for accelerated canvases

Need to set the r? flag if you want review...

[George Staikos]

Comment on attachment 157217 [details]
Don't share styles for accelerated canvases

I don't think you can CQ this because you removed the Reviewed By: line, but let's try and see how smart the scripts are :)

[WebKit Review Bot]

Comment on attachment 157217 [details]
Don't share styles for accelerated canvases

Rejecting attachment 157217 [details] from commit-queue.

Failed to run "['/mnt/git/webkit-commit-queue/Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '-..." exit_code: 1

ERROR: /mnt/git/webkit-commit-queue/Source/WebCore/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://queues.webkit.org/results/13455733

[Simon Fraser (smfr)]

Comment on attachment 157217 [details]
Don't share styles for accelerated canvases

Looking through the canvas code, we usually do USE(IOSURFACE_CANVAS_BACKING_STORE) || ENABLE(ACCELERATED_2D_CANVAS). The latter is not used for Mac builds, confusingly.

[Filip Spacek]

Created attachment 157608 [details]
Don't share styles for accelerated canvases

[Simon Fraser (smfr)]

I think you should try to make a testcase for this.

[George Staikos]

(In reply to comment #6)
> (From update of attachment 157217 [details])
> Looking through the canvas code, we usually do USE(IOSURFACE_CANVAS_BACKING_STORE) || ENABLE(ACCELERATED_2D_CANVAS). The latter is not used for Mac builds, confusingly.

We need to unify this.  It's really not clear

(In reply to comment #8)
> I think you should try to make a testcase for this.

Is there a way to test something like this?  Seems ... challenging.

[Simon Fraser (smfr)]

(In reply to comment #9)
> (In reply to comment #6)
> > (From update of attachment 157217 [details] [details])
> > Looking through the canvas code, we usually do USE(IOSURFACE_CANVAS_BACKING_STORE) || ENABLE(ACCELERATED_2D_CANVAS). The latter is not used for Mac builds, confusingly.
> 
> We need to unify this.  It's really not clear

Agreed.

> (In reply to comment #8)
> > I think you should try to make a testcase for this.
> 
> Is there a way to test something like this?  Seems ... challenging.

If you look at the history for some of the other "don't share style" criteria, you may find testcases. I recall making one for iframes, I think.

[Tim Horton]

Comment on attachment 157608 [details]
Don't share styles for accelerated canvases

r- to remove from the review queue until there's a test.