Allow glUniform to set nth element of array of structs uniform where n != 0.

Allow glUniform to set nth element of array of structs uniform where n != 0.
Committed: https://code.google.com/p/angleproject/source/detail?r=1267

[apatrick1, 2012-08-23 00:23:50]

Daniel,

I added a new WebGL conformancce test:
https://github.com/apatrick-google/WebGL/commit/85d7419bb79e1c3ed7d43b40545b0...

The test now passes and there were no new regressions.

PTAL

https://codereview.appspot.com/6477046/diff/1/src/libGLESv2/ProgramBinary.cpp
File src/libGLESv2/ProgramBinary.cpp (right):

https://codereview.appspot.com/6477046/diff/1/src/libGLESv2/ProgramBinary.cpp...
src/libGLESv2/ProgramBinary.cpp:2116: D3DXHANDLE fieldHandle =
mConstantTablePS->GetConstant(elementHandle, field);
Does this work for vertex shaders as well?

[dgkoch, 2012-08-27 19:56:50]

LGTM.  Did you want to fix up the vertex handling here or in another patch?

https://codereview.appspot.com/6477046/diff/1/src/libGLESv2/ProgramBinary.cpp
File src/libGLESv2/ProgramBinary.cpp (right):

https://codereview.appspot.com/6477046/diff/1/src/libGLESv2/ProgramBinary.cpp...
src/libGLESv2/ProgramBinary.cpp:2116: D3DXHANDLE fieldHandle =
mConstantTablePS->GetConstant(elementHandle, field);
On 2012/08/23 00:23:50, apatrick1 wrote:
> Does this work for vertex shaders as well?

Well probably not as written, but I'd expect it should.  Structs shouldn't be
handled any different between vertex and fragment shaders. I'd guess we'll need
two cases just like we do above for the samplers.

[apatrick1, 2012-08-27 23:20:46]

I the vertex shaders are fixed with this patch:

https://codereview.appspot.com/6485061/

That's a bit more involved. I can put together a simpler fix in the meantime if
you think the vertex shaders issue should be fixed before that lands?

[apatrick1, 2012-08-27 23:29:25]

Allow glUniform to set nth element of array of structs uniform where n != 0.

[dgkoch, 2012-08-28 01:30:03]

On 2012/08/27 23:20:46, apatrick1 wrote:
> I the vertex shaders are fixed with this patch:
> 
> https://codereview.appspot.com/6485061/
Ok -- that's up next on my review list -- tomorrow.

> That's a bit more involved. I can put together a simpler fix in the meantime
if
> you think the vertex shaders issue should be fixed before that lands?
Nah. No one noticed the problem before now...