add drawArraysInstanced to caps

add drawArraysInstanced to caps

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/58001553ec6298cbea3b2e915ed7a1ac0e763e8f

[joshualitt, 2015-06-24 20:46:24]

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

[joshualitt, 2015-06-24 20:46:24]

ptal

[joshualitt, 2015-06-26 13:21:41]

On 2015/06/24 20:46:24, joshualitt wrote:
> ptal

ping

[bsalomon, 2015-06-26 13:37:14]

https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h
File include/gpu/GrCaps.h (right):

https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h#newcod...
include/gpu/GrCaps.h:219: bool drawArraysInstancedSupport() const {
I really don't like having GLisms bleed into GrCaps (drawArraysInstanced). Can
we keep this GrGLCaps for now? Or can we think of what the high level thing
we're trying to capture as a capability is (e.g. supportsInstancedDraws()).

Also, we aren't grabbing the function pointers in the interface. Does this
really need to be in caps at all at this point? Maybe we should wait until we
figure out how to expose this at the Gr level?

[joshualitt, 2015-06-26 13:52:09]

On 2015/06/26 13:37:14, bsalomon wrote:
> https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h
> File include/gpu/GrCaps.h (right):
> 
>
https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h#newcod...
> include/gpu/GrCaps.h:219: bool drawArraysInstancedSupport() const {
> I really don't like having GLisms bleed into GrCaps (drawArraysInstanced). Can
> we keep this GrGLCaps for now? Or can we think of what the high level thing
> we're trying to capture as a capability is (e.g. supportsInstancedDraws()).
> 
> Also, we aren't grabbing the function pointers in the interface. Does this
> really need to be in caps at all at this point? Maybe we should wait until we
> figure out how to expose this at the Gr level?

No matter how we expose it in Gr world, its going to have to be on the caps in
some form.  GrBatch will certainly need to know its there to figure out if its
going to use one VBO, or VBO + instance buffers.

I think 'supportsInstancedDraws()' is reasonable, but we can wait if you prefer.

[bsalomon, 2015-06-26 14:03:34]

On 2015/06/26 13:52:09, joshualitt wrote:
> On 2015/06/26 13:37:14, bsalomon wrote:
> > https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h
> > File include/gpu/GrCaps.h (right):
> > 
> >
>
https://codereview.chromium.org/1202383006/diff/1/include/gpu/GrCaps.h#newcod...
> > include/gpu/GrCaps.h:219: bool drawArraysInstancedSupport() const {
> > I really don't like having GLisms bleed into GrCaps (drawArraysInstanced).
Can
> > we keep this GrGLCaps for now? Or can we think of what the high level thing
> > we're trying to capture as a capability is (e.g. supportsInstancedDraws()).
> > 
> > Also, we aren't grabbing the function pointers in the interface. Does this
> > really need to be in caps at all at this point? Maybe we should wait until
we
> > figure out how to expose this at the Gr level?
> 
> No matter how we expose it in Gr world, its going to have to be on the caps in
> some form.  GrBatch will certainly need to know its there to figure out if its
> going to use one VBO, or VBO + instance buffers.
> 
> I think 'supportsInstancedDraws()' is reasonable, but we can wait if you
prefer.

K, that seems fine (lgtm)

[joshualitt, 2015-06-26 19:40:51]

The CQ bit was checked by joshualitt@google.com

[joshualitt, 2015-06-26 19:40:51]

The patchset sent to the CQ was uploaded after l-g-t-m from bsalomon@google.com
Link to the patchset: https://codereview.chromium.org/1202383006/#ps20001
(title: "feedback inc")

[commit-bot: I haz the power, 2015-06-26 19:41:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1202383006/20001

[commit-bot: I haz the power, 2015-06-26 19:46:40]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/58001553ec6298cbea3b2e915ed7a1ac0e763e8f