gpu: don't try to discard/recreate backbuffer while draws are deferred.

gpu: don't try to discard/recreate backbuffer while draws are deferred.

BUG=None


Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=170366

[no sievers, 2012-11-29 04:53:32]

lgtm

So it's only discardFramebuffer() that invokes backbuffer deallocation? The
memory manager will never be involved in freeing the backbuffer?

[jonathan.backer, 2012-11-29 15:27:34]

LGTM

If you wanted to enforce the contract that you can only requeue those messages
that are in flight, you could:

- add a private msg_pending_ to GpuChannel
- maybe have a ScopedMsgPending friend class
- only requeue msg_pending_ if asked to requeue

This will help guarantee ordering of the message queue (prevent abuses of your
API).

[piman, 2012-11-29 19:47:02]

On 2012/11/29 15:27:34, jonathan.backer wrote:
> LGTM
> 
> If you wanted to enforce the contract that you can only requeue those messages
> that are in flight, you could:
> 
> - add a private msg_pending_ to GpuChannel
> - maybe have a ScopedMsgPending friend class
> - only requeue msg_pending_ if asked to requeue
> 
> This will help guarantee ordering of the message queue (prevent abuses of your
> API).

Good idea, PTAL. ScopedMsgPending is overkill I think - see the code.

[jonathan.backer, 2012-11-29 19:55:11]

LGTM.

One small nit. Your call.

https://codereview.chromium.org/11316238/diff/6001/content/common/gpu/gpu_cha...
File content/common/gpu/gpu_channel.cc (right):

https://codereview.chromium.org/11316238/diff/6001/content/common/gpu/gpu_cha...
content/common/gpu/gpu_channel.cc:368: unprocessed_messages_->IncCount();
I'm paranoid. I'd reset currently_processing_message_ here as well.

[piman, 2012-11-29 20:20:57]

https://codereview.chromium.org/11316238/diff/6001/content/common/gpu/gpu_cha...
File content/common/gpu/gpu_channel.cc (right):

https://codereview.chromium.org/11316238/diff/6001/content/common/gpu/gpu_cha...
content/common/gpu/gpu_channel.cc:368: unprocessed_messages_->IncCount();
On 2012/11/29 19:55:11, jonathan.backer wrote:
> I'm paranoid. I'd reset currently_processing_message_ here as well.

Doh. I meant to do that, and then I forgot. Thanks, done.

[commit-bot: I haz the power, 2012-11-29 20:21:28]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/piman@chromium.org/11316238/5013

[commit-bot: I haz the power, 2012-11-29 23:20:28]

Retried try job too often on win_rel for step(s) browser_tests

[commit-bot: I haz the power, 2012-11-30 00:16:47]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/piman@chromium.org/11316238/5013

[commit-bot: I haz the power, 2012-11-30 02:36:08]

Change committed as 170366