gpu: Cache results of GetShaderPrecisionFormat client-side

gpu: Cache results of GetShaderPrecisionFormat client-side

This will allow us to avoid making the round-trip call to the server
every time this function is called.

BUG=222894
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=190205

[brianderson, 2013-03-21 22:02:22]

This method is pretty straightforward - it caches values lazily.

Would it also be worth querying all valid precision combinations asynchronously
on context creation and proactively caching the results?

[Ken Russell (switch to Gerrit), 2013-03-21 22:22:41]

Gregg should review this.

Where are the try jobs? Does this change pass the gpu unit tests?

[brianderson, 2013-03-21 22:31:05]

On 2013/03/21 22:22:41, kbr wrote:
> Gregg should review this.
> 
> Where are the try jobs? Does this change pass the gpu unit tests?

It passes gpu unit tests locally. Just kicked off some try jobs.

[greggman, 2013-03-21 22:52:19]

Please add or update the unit test so that it calls
GLES2Implementation::GetShaderPrecisionFormat 3 times.

Once to see that it writes the command
Again to see that the value is now cached and no command is written
A 3rd time with a different pname to see that it writes a command.

[brianderson, 2013-03-22 02:54:34]

Added the test.

[brianderson, 2013-03-22 18:18:17]

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
File gpu/command_buffer/client/gles2_implementation_unittest.cc (right):

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest.cc:659:
((char*)commands_) + sizeof(expected1),
Is there a better way of getting a pointer to the start of the second command in
the command buffer? Or is this okay.

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
File gpu/command_buffer/client/gles2_implementation_unittest_autogen.h (left):

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest_autogen.h:666: // TODO:
Implement unit test for GetShaderPrecisionFormat
I couldn't figure out how this TODO is auto generated. Should I be changing
something else?

[greggman, 2013-03-22 22:06:14]

lgtm

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
File gpu/command_buffer/client/gles2_implementation_unittest.cc (right):

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest.cc:622:
.WillOnce(Invoke(run_server1))
Could you have just used SetMemory?

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest.cc:659:
((char*)commands_) + sizeof(expected1),
On 2013/03/22 18:18:17, Brian Anderson wrote:
> Is there a better way of getting a pointer to the start of the second command
in
> the command buffer? Or is this okay.


yes, before calling gl_>GetShaderPrecisionFormat call

 const void* commands2 = GetPut();

[brianderson, 2013-03-22 23:11:27]

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
File gpu/command_buffer/client/gles2_implementation_unittest.cc (right):

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest.cc:622:
.WillOnce(Invoke(run_server1))
On 2013/03/22 22:06:14, greggman wrote:
> Could you have just used SetMemory?
So much easier.

https://codereview.chromium.org/13004003/diff/10001/gpu/command_buffer/client...
gpu/command_buffer/client/gles2_implementation_unittest.cc:659:
((char*)commands_) + sizeof(expected1),
On 2013/03/22 22:06:14, greggman wrote:
> On 2013/03/22 18:18:17, Brian Anderson wrote:
> > Is there a better way of getting a pointer to the start of the second
command
> in
> > the command buffer? Or is this okay.
> 
> 
> yes, before calling gl_>GetShaderPrecisionFormat call
> 
>  const void* commands2 = GetPut();

Done.

[commit-bot: I haz the power, 2013-03-22 23:26:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/brianderson@chromium.org/13004003/16001

[commit-bot: I haz the power, 2013-03-23 16:06:31]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/brianderson@chromium.org/13004003/16001

[commit-bot: I haz the power, 2013-03-24 05:08:35]

Change committed as 190205