Bug 76547 – Signal to skia to force A8 text from LCD output, but only when we have to disable LCD because we're in a layer

• Component: New Bugs
• Status: RESOLVED FIXED

[Mike Reed]

Signal to skia to force A8 text from LCD output, but only when we have to disable LCD because we're in a layer

[Mike Reed]

Created attachment 122943 [details]
Patch

[Mike Reed]

will need skia 3061 to land, to access the new API to set the GenA8FromLCD bit

[Mike Reed]

will address http://code.google.com/p/chromium/issues/detail?id=110066 (among others)

[Stephen White]

Comment on attachment 122943 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=122943&action=review

OK.  r=me

> Source/WebCore/ChangeLog:10
> +        Existing tests should confirm nothing is broken. Antialiased text
> +        is disabled in layouttests, so they should be unaffected by the
> +        difference in antialiasing quality.

Not your fault, but we really should come up with a way to override AA settings for specific layout tests.  This could be done by adding to the JS API in testing/Internals (as we do for accelerated drawing, e.g.).  This would allow us to test changes like this.

[Kenneth Russell]

Comment on attachment 122943 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=122943&action=review

> Source/WebCore/platform/graphics/skia/SkiaFontWin.cpp:184
> +                         SkPaint::kGenA8FromLCD_Flag)

I should suggest here to use a constant rather than a #define.

[Mike Reed]

Created attachment 123747 [details]
Patch

[Mike Reed]

changed #define to a constant.

[Mike Reed]

Comment on attachment 123747 [details]
Patch

new API from skia (rev. 3061) has landed.

[WebKit Review Bot]

Comment on attachment 123747 [details]
Patch

Attachment 123747 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/11334427

New failing tests:
media/audio-garbage-collect.html

[Stephen White]

Comment on attachment 123747 [details]
Patch

media/audio-garbage-collect.html looks unrelated (failing intermittently on canaries as well).

Looks good.  r=me

[WebKit Review Bot]

Comment on attachment 123747 [details]
Patch

Rejecting attachment 123747 [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/11342545

[Stephen White]

(In reply to comment #11)
> (From update of attachment 123747 [details])
> Rejecting attachment 123747 [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/11342545

Looks like the Reviewed by NOBODY line is missing OOPS; perhaps this is confusing the commit bot.

[Vangelis Kokkevis]

Committed r106029: <http://trac.webkit.org/changeset/106029>