Implement stroking a path with nv_path_rendering

Implement stroking a path with nv_path_rendering

Initialize the path stroke properties in the GrGLPath constructor.
Use StencilStrokePath and CoverStrokePath to stroke the path.

The order of the GL calls is:
1. StencilFill, if needed
2. StencilStroke, if needed
2a. CoverStroke, if stroke was applied
2b. CoverFill, if stroke was not applied

The reason for not pairing StencilFill + CoverFill, StencilStroke +
CoverStroke is that Skia API does not allow separate fill and stroke
color within one call. Covering the stroke bounding box should also
cover the fill bounding box.

Causes different rendering in gm/dashcubics due to different rendering
algorithm. (?) (TODO: this should be resolved somehow.)

Committed: http://code.google.com/p/skia/source/detail?r=11672

[Kimmo Kinnunen, 2013-09-19 07:51:28]

Initial version up for early comments (depends on the patches not in yet).

[bsalomon, 2013-09-19 18:14:48]

Hi Kimmo, this seems fairly straight forward to me. I have some minor issues
noted below.

https://codereview.chromium.org/23440049/diff/1/src/gpu/gl/GrGLPath.cpp
File src/gpu/gl/GrGLPath.cpp (right):

https://codereview.chromium.org/23440049/diff/1/src/gpu/gl/GrGLPath.cpp#newco...
src/gpu/gl/GrGLPath.cpp:59: inline GrGLenum sk_join_to_gr_gl_join(const
SkPaint::Join& join) {
can just call this .._to_gl_join (no need for _gr). Same applies to the next
func.

https://codereview.chromium.org/23440049/diff/1/src/gpu/gl/GrGpuGL.h
File src/gpu/gl/GrGpuGL.h (right):

https://codereview.chromium.org/23440049/diff/1/src/gpu/gl/GrGpuGL.h#newcode151
src/gpu/gl/GrGpuGL.h:151: virtual void onGpuDrawPath(const GrPath*, const
SkStrokeRec& stroke, SkPath::FillType) SK_OVERRIDE;
Since the GrPath is now specific to the stroking, it seems redundant to pass it
here. Can GrPath just snap the SkStrokeRec when it's made?

[Kimmo Kinnunen, 2013-10-08 12:15:36]

Thanks. New version uploaded.

https://chromiumcodereview.appspot.com/23440049/diff/1/src/gpu/gl/GrGLPath.cpp
File src/gpu/gl/GrGLPath.cpp (right):

https://chromiumcodereview.appspot.com/23440049/diff/1/src/gpu/gl/GrGLPath.cp...
src/gpu/gl/GrGLPath.cpp:59: inline GrGLenum sk_join_to_gr_gl_join(const
SkPaint::Join& join) {
On 2013/09/19 18:14:48, bsalomon wrote:
> can just call this .._to_gl_join (no need for _gr). Same applies to the next
> func.

Done.

https://chromiumcodereview.appspot.com/23440049/diff/1/src/gpu/gl/GrGpuGL.h
File src/gpu/gl/GrGpuGL.h (right):

https://chromiumcodereview.appspot.com/23440049/diff/1/src/gpu/gl/GrGpuGL.h#n...
src/gpu/gl/GrGpuGL.h:151: virtual void onGpuDrawPath(const GrPath*, const
SkStrokeRec& stroke, SkPath::FillType) SK_OVERRIDE;
On 2013/09/19 18:14:48, bsalomon wrote:
> Since the GrPath is now specific to the stroking, it seems redundant to pass
it
> here. Can GrPath just snap the SkStrokeRec when it's made?

Done.

[bsalomon, 2013-10-08 14:09:34]

lgtm, is this dependent upon https://chromiumcodereview.appspot.com/22686002/
landing first?

[commit-bot: I haz the power, 2013-10-09 15:04:20]

CQ is trying da patch. Follow status at
https://skia-tree-status.appspot.com/cq/kkinnunen@nvidia.com/23440049/35001

[commit-bot: I haz the power, 2013-10-09 15:14:22]

Change committed as 11672