Implement conics for NVPR

Implement conics for NVPR

Implement conics for NVPR. Verified with currently disabled conicpaths
gm, produces mostly same shapes with nvprmsaa4 and msaa4.

BUG=skia:2033
Committed: https://skia.googlesource.com/skia/+/e097be50bdfac597119db2a170f96bae07dbcab1

[bsalomon, 2014-11-11 15:19:25]

bsalomon@google.com changed reviewers:
+ bsalomon@google.com

[bsalomon, 2014-11-11 15:19:25]

Do we not have a GM that exercises this? Maybe one should be added?

[Kimmo Kinnunen, 2014-11-12 06:24:13]

On 2014/11/11 15:19:25, bsalomon wrote:
> Do we not have a GM that exercises this? Maybe one should be added?

Sorry, you're of course right and I should have added one instead of
semi-complaining the lack of them.

And, of course, turns out there is a gm for conics after all :)

PTAL, I updated the description.

[reed1, 2014-11-12 14:06:38]

reed@google.com changed reviewers:
+ reed@google.com

[reed1, 2014-11-12 14:06:39]

https://codereview.chromium.org/694503003/diff/20001/include/core/SkPath.h
File include/core/SkPath.h (right):

https://codereview.chromium.org/694503003/diff/20001/include/core/SkPath.h#ne...
include/core/SkPath.h:285: int countConicWeights() const;
ick ich, do we really need to make these two apis public?

[Kimmo Kinnunen, 2014-11-19 16:25:39]

https://codereview.chromium.org/694503003/diff/20001/include/core/SkPath.h
File include/core/SkPath.h (right):

https://codereview.chromium.org/694503003/diff/20001/include/core/SkPath.h#ne...
include/core/SkPath.h:285: int countConicWeights() const;
On 2014/11/12 14:06:39, reed1 wrote:
> ick ich, do we really need to make these two apis public?

Done.

[bsalomon, 2014-11-19 16:53:33]

lgtm

https://codereview.chromium.org/694503003/diff/40001/src/gpu/gl/GrGLPath.cpp
File src/gpu/gl/GrGLPath.cpp (right):

https://codereview.chromium.org/694503003/diff/40001/src/gpu/gl/GrGLPath.cpp#...
src/gpu/gl/GrGLPath.cpp:111: #ifndef SK_SCALAR_IS_FLOAT
There is some interest in scalar=double. I suppose with the #error we can cross
that bridge when when we come to it, as adding code behind #if would be untested
dead code for the time being.

[Kimmo Kinnunen, 2014-11-20 06:40:22]

The CQ bit was checked by kkinnunen@nvidia.com

[commit-bot: I haz the power, 2014-11-20 06:40:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/694503003/40001

[commit-bot: I haz the power, 2014-11-20 06:49:07]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/e097be50bdfac597119db2a170f96bae07dbcab1