Checkpoint towards core profile support.

Checkpoint towards core profile support.

1) Stop calling glDisable for removed state

2) Use new GLSL names for texture sampling functions.
Committed: https://code.google.com/p/skia/source/detail?r=7908

[robertphillips, 2013-02-27 22:33:07]

lgtm modulo some clarification on two questions.

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp
File src/gpu/gl/GrGpuGL.cpp (right):

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp#new...
src/gpu/gl/GrGpuGL.cpp:412: // Desktop-only state that we never change
Shouldn't we instead check if they are enabled and disable them as required?

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp#new...
src/gpu/gl/GrGpuGL.cpp:420: }
Why not these other two?

[bsalomon, 2013-02-28 14:09:41]

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp
File src/gpu/gl/GrGpuGL.cpp (right):

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp#new...
src/gpu/gl/GrGpuGL.cpp:412: // Desktop-only state that we never change
On 2013/02/27 22:33:07, robertphillips wrote:
> Shouldn't we instead check if they are enabled and disable them as required?

You mean make glGetIntegerv calls? I assume that querying can only be worse than
disabling unconditionally. I don't expect desktop drivers to do much work if we
disable these redundantly. This is the onResetContext() call which hopefully
gets called very rarely (only after third party code has touched the GL
context).

https://codereview.chromium.org/12330181/diff/5001/src/gpu/gl/GrGpuGL.cpp#new...
src/gpu/gl/GrGpuGL.cpp:420: }
On 2013/02/27 22:33:07, robertphillips wrote:
> Why not these other two?

The offset fill and vp point size? They're still around in core profile. (well
VERTEX_PROGRAM_POINT_SIZE was renamed to PROGRAM_POINT_SIZE at some point, but
same integer value).

[robertphillips, 2013-02-28 15:05:45]

lgtm

[bsalomon, 2013-02-28 22:05:42]

Checkpoint towards core profile support.

[bsalomon, 2013-02-28 22:06:06]

Committed patchset #3 manually as r7908 (presubmit successful).