696630 - Canvas context mozImageSmoothingEnabled state ignored for pattern fill.

Status: RESOLVED FIXED

(Core :: Graphics: Canvas2D, defect, P1)

Description

[Carl Olsson]

Attachment: pattern_bug.html

User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:7.0.1) Gecko/20100101 Firefox/7.0.1
Build ID: 20110928134238

Steps to reproduce:

Create HTML file with canvas.
Set canvas context mozImageSmoothingEnabled = false.
Draw pattern filled primitive to context.


Actual results:

Pattern draws with bilinear filtering.


Expected results:

Pattern should draw with nearest-neighbor sampling.

Comment 1

[Carl Olsson]

Testing this on my laptop with Intel graphics, under both Windows 7 and Ubuntu 11.10 this does work as expected.

The bug presents on my desktop with AMD graphics (Radeon HD 5770, Catalyst 11.9), so maybe it's an AMD specific bug.

Comment 2

[Till Schneidereit [:till]]

This is hurting us in Shumway: Flash by default doesn't apply anti-aliasing when scaling Bitmaps. (That is, the constructor for Bitmap has an argument `smoothing` with a default value of false, and graphics.beginBitmapFill also defaults to non-smoothing. The former works in FF, the latter doesn't. Both work in Chrome.)

Comment 4

[Chris Peterson [:cpeterson]]

Till says this is a hard blocker for gfx fidelity, but might be easy to fix.

Comment 5

[Michael Bebenita [:mbx]]

Attachment: Take imageSmoothingEnabled into account when creating pattern fills.

Prototype implementation, will follow up will try results.

Comment 6

[Michael Bebenita [:mbx]]

Try: https://tbpl.mozilla.org/?tree=Try&rev=7d63a6133ba7

Comment 7

[Michael Bebenita [:mbx]]

Attachment: Take imageSmoothingEnabled into account when creating pattern fills w/ test case.

Comment 8

[Benoit Jacob [:bjacob] (mostly away)]

Comment on attachment 8494856 [details] [diff] [review]
Take imageSmoothingEnabled into account when creating pattern fills w/ test case.

Review of attachment 8494856 [details] [diff] [review]:
-----------------------------------------------------------------

I'm the wrong reviewer. Try Bas Schouten or Matt Woodrow.

Comment 9

[Chris Peterson [:cpeterson]]

Michael, is your mozImageSmoothingEnabled patch still relevant?

Comment 10

[Michael Bebenita [:mbx]]

Yes, it should still be relevant.

Comment 11

[Florian Bender]

This is pretty much blocking PDF.js issue #3331, any update on when this can be fixed?

Comment 12

[Till Schneidereit [:till]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=78f8685cd94d

Comment 13

[Till Schneidereit [:till]]

Attachment: Patch, rebased

Comment 14

[Michael Bebenita [:mbx]]

Attachment: Patch, reabased

Patch, fixed compiler error.

Comment 15

[Till Schneidereit [:till]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=a5fc3c8d2b25

Comment 16

[Michael Bebenita [:mbx]]

Attachment: Fixed test case.

Comment 17

[Till Schneidereit [:till]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=1197704c568d

Comment 18

[Till Schneidereit [:till]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=c9f4eff65487

Comment 19

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/652ea88c04a6

Comment 20

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/652ea88c04a6

Comment 21

[Florian Bender]

Cool, thanks a lot!