Pass dashing information to path rasterizers.

Pass dashing information to path rasterizers.

This is a WIP patch.

This changes the canDrawPath() and onDrawPath() virtuals to take a
GrStrokeInfo instead of an SkStrokeRec.

It also adds a SkPathEffect to GrStrokeInfo. Note that some of the existing
dashing-related fields in GrStrokeInfo may now be redundant, and could
probably be removed.

Note: this required losing the copy-on-write info in GrContext::internalDrawPath().
If this is required, we can figure out a way to make GrStrokeInfo copy-on-write,
since it isn't currently.

BUG=skia:

[Stephen White, 2015-04-16 16:25:08]

senorblanco@chromium.org changed reviewers:
+ bsalomon@google.com, egdaniel@google.com

[Stephen White, 2015-04-16 16:25:08]

Brian, Greg: let me know if this is headed in the right direction.

[Stephen White, 2015-04-16 16:32:57]

The CQ bit was checked by senorblanco@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-04-16 16:33:10]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1096513002/20001

[commit-bot: I haz the power, 2015-04-16 16:39:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-04-16 16:39:28]

Dry run: This issue passed the CQ dry run.

[bsalomon, 2015-04-16 17:52:18]

Seems like the right direction to me. I'm not sure whether we added that copy on
write business because of a real problem or just speculatively. I suppose there
would also be companion changes in SkGpuDevice to avoid applying the path
effect/dash early.

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp
File src/gpu/GrContext.cpp (right):

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp#n...
src/gpu/GrContext.cpp:1460: stroke.getStrokeRecPtr()->setFillStyle();
Seems like this should also remove the path effect to. Should we have something
like

static const GrStrokeInfo& GrStrokeInfo::FillInfo();

It'd just be a static function that creates a static GrStrokeInfo for filling
and returns it.

I realize it's already there, but don't love "getStrokeRecPtr()" and would
prefer to avoid using it.

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrStrokeInfo.h
File src/gpu/GrStrokeInfo.h (right):

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrStrokeInfo.h#...
src/gpu/GrStrokeInfo.h:92: SkPathEffect* getPathEffect() const { return
fPathEffect; }
can this be const?

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrStrokeInfo.h#...
src/gpu/GrStrokeInfo.h:103: SkPathEffect*          fPathEffect;
can this be const?

[bsalomon, 2015-04-16 17:53:38]

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp
File src/gpu/GrContext.cpp (right):

https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp#n...
src/gpu/GrContext.cpp:1458: if
(stroke.getStrokeRec().applyToPath(tmpPath.init(), *pathPtr)) {
need to apply path effect here, right?

[Stephen White, 2015-05-14 22:12:53]

On 2015/04/16 17:53:38, bsalomon wrote:
> https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp
> File src/gpu/GrContext.cpp (right):
> 
>
https://codereview.chromium.org/1096513002/diff/20001/src/gpu/GrContext.cpp#n...
> src/gpu/GrContext.cpp:1458: if
> (stroke.getStrokeRec().applyToPath(tmpPath.init(), *pathPtr)) {
> need to apply path effect here, right?

Closing this in favour of https://codereview.chromium.org/1100073003, which has
already landed.