Enable 2 pending frames on ARM devices.

Enable 2 pending frames in the scheduler on ARM devices.

In the past eglSwapBuffers used to return early, before the swap actually happened and descheduling in the driver. This resulted in the scheduler beginning the next frame early and hitting the page flip deadline in time more often.

Freon returns swap ack when page flip actually happened, but with the side effect of scheduling shifted forward, resulting in more missed vblanks and reduced performance.

This change restores original performance on freon.

BUG=465599
Committed: https://crrev.com/7d66db12d3dd9d3b068eee2ff11f3ccc81e767d3
Cr-Commit-Position: refs/heads/master@{#324098}

[alexst (slow to review), 2015-04-03 00:21:21]

alexst@chromium.org changed reviewers:
+ brianderson@chromium.org

[alexst (slow to review), 2015-04-03 00:23:17]

Brian, please have a look.

After the https://codereview.chromium.org/1054073002/ lands, > 1 frames pending
works on arm and we can regain our original performance.

[brianderson, 2015-04-03 22:44:01]

https://codereview.chromium.org/1059503003/diff/1/cc/scheduler/scheduler_stat...
File cc/scheduler/scheduler_state_machine.cc (left):

https://codereview.chromium.org/1059503003/diff/1/cc/scheduler/scheduler_stat...
cc/scheduler/scheduler_state_machine.cc:427: // TODO(brianderson): Remove this
restriction to improve throughput.
Beware that this aspect of the patch might introduce flakiness in browser tests
and could result in regressing other use cases where it might increase the CPU
fighting. We should definitely keep an eye on the perf dashboard for this.

Do you get any benefit from just the other half of this patch? If so it might be
worth splitting them up so you don't have to revert both parts.

[brianderson, 2015-04-03 22:44:13]

lgtm

[alexst (slow to review), 2015-04-04 00:20:02]

https://codereview.chromium.org/1059503003/diff/1/cc/scheduler/scheduler_stat...
File cc/scheduler/scheduler_state_machine.cc (left):

https://codereview.chromium.org/1059503003/diff/1/cc/scheduler/scheduler_stat...
cc/scheduler/scheduler_state_machine.cc:427: // TODO(brianderson): Remove this
restriction to improve throughput.
On 2015/04/03 22:44:01, brianderson wrote:
> Beware that this aspect of the patch might introduce flakiness in browser
tests
> and could result in regressing other use cases where it might increase the CPU
> fighting. We should definitely keep an eye on the perf dashboard for this.
> 
> Do you get any benefit from just the other half of this patch? If so it might
be
> worth splitting them up so you don't have to revert both parts.

On intel, just the other part was enough, on ARM, this was also required.

Maybe we should take a more cautious approach and #ifdef it out only for freon
on arm? This would let us monitor it on a smaller scale and not  cause side
effects elsewhere.

That being said, I think you are right, I'll split these 2 parts to keep impact
limited.

[alexst (slow to review), 2015-04-06 13:11:20]

alexst@chromium.org changed reviewers:
+ danakj@chromium.org

[alexst (slow to review), 2015-04-06 13:29:00]

Split it up, part 2 is here https://codereview.chromium.org/1053153003/

[danakj, 2015-04-06 16:34:28]

I guess you need the other CL to land first. But LGTM once it lands. Do you
think it would make sense to leave a comment on this explaining why it can be 2
here for future readers?

[alexst (slow to review), 2015-04-07 18:24:33]

Yeah, good point, documented for future spelunkers.

[alexst (slow to review), 2015-04-07 18:24:39]

The CQ bit was checked by alexst@chromium.org

[alexst (slow to review), 2015-04-07 18:24:40]

The patchset sent to the CQ was uploaded after l-g-t-m from
brianderson@chromium.org, danakj@chromium.org
Link to the patchset: https://codereview.chromium.org/1059503003/#ps40001
(title: "Comments")

[commit-bot: I haz the power, 2015-04-07 18:26:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1059503003/40001

[commit-bot: I haz the power, 2015-04-07 20:06:38]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-04-07 20:07:35]

Patchset 3 (id:??) landed as
https://crrev.com/7d66db12d3dd9d3b068eee2ff11f3ccc81e767d3
Cr-Commit-Position: refs/heads/master@{#324098}