Bug 61369

Summary:

GraphicsContextSkia::drawConvexPolygon doesn't pay attention to shouldAntialias

Product:

WebKit

Reporter:

Tom Hudson <tomhudson>

Component:

Platform

Assignee:

Nobody <webkit-unassigned>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

benwells, eric, jamesr, reed, tomhudson

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

Unspecified

Bug Depends on:

Bug Blocks:

60342

Attachments:

Description	Flags
Patch	none
Patch	none
Patch	jamesr: review+, jamesr: commit-queue-

Tom Hudson

Reported 2011-05-24 09:45:35 PDT

GraphicsContextSkia::drawConvexPolygon() ignores the shouldAntialias. The CG & Qt implementations both do obey it. The effect is that Skia/Chromium does *too much* antialiasing, particularly on the vertical edges of table cells where WebKit is telling us not to antialias.

Attachments
Patch (93.76 KB, patch) 2011-05-24 14:17 PDT, Tom Hudson	no flags	Details Formatted Diff Diff
Patch (577 bytes, patch) 2011-05-25 05:59 PDT, Tom Hudson	no flags	Details Formatted Diff Diff
Patch (93.84 KB, patch) 2011-05-25 06:20 PDT, Tom Hudson	jamesr: review+ jamesr: commit-queue-	Details Formatted Diff Diff
Show Obsolete (2) View All Add attachment proposed patch, testcase, etc.

Tom Hudson

Comment 1 2011-05-24 14:17:02 PDT

Created attachment 94685 [details] Patch

Tom Hudson

Comment 2 2011-05-25 05:59:11 PDT

Created attachment 94772 [details] Patch

Tom Hudson

Comment 3 2011-05-25 06:20:59 PDT

Created attachment 94776 [details] Patch

James Robinson

Comment 4 2011-05-25 12:23:57 PDT

Comment on attachment 94776 [details] Patch Patch looks good. Not touching c-q flag yet, lemme scheme with Dimitri about how to land this.

James Robinson

Comment 5 2011-05-25 12:26:36 PDT

Comment on attachment 94776 [details] Patch cq-, we're gonna have to land this manually

Eric Seidel (no email)

Comment 6 2011-06-18 20:02:18 PDT

Any update? This was marked for manual landing nearly a month ago.

Tom Hudson

Comment 7 2011-06-20 05:38:20 PDT

(In reply to comment #6) > Any update? This was marked for manual landing nearly a month ago. At that time the gardener held off landing it because they didn't want to disable so many tests. I assume that I need a committer to work with, and don't have one in my local office.

James Robinson

Comment 8 2011-06-20 12:55:06 PDT

Hi, sorry. This patch is good to land but will require a good amount of manual work by committers post-landing to deal with unexpected image failures and to rebaseline the expected image diffs. I was hoping (as it turns out optimistically) that I'd have some time to help get this landed, but that's seeming unrealistic. Anyone who feels comfortable handling the pixel test rebaselines should feel free to land this at any time (but please coordinate with the chromium WebKit gardener).

James Robinson

Comment 9 2011-07-10 20:18:28 PDT

Landed http://trac.webkit.org/changeset/90701, rebaselining has begun.

James Robinson

Comment 10 2011-07-11 01:16:04 PDT

Ben Wells

Comment 11 2011-07-27 21:52:44 PDT

*** Bug 60342 has been marked as a duplicate of this bug. ***

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