Don't throttle touch events on resent GestureScrollUpdates.

Don't throttle touch events on resent GestureScrollUpdates.

Touch events are currently throttled during GestureScrolls to avoid
overloading a page's touch handler. However, in the event we are
resending a GestureScroll to a GuestView's host based on a touch handler
in the Guest, this leads to extremely janky scroll performance.

This CL detects when a GestureScroll is being resent, and turns off
throttling in this case.

BUG=491940
Committed: https://crrev.com/1ab6149f525fc879dbc068d047501ca715a589d6
Cr-Commit-Position: refs/heads/master@{#351804}

[wjmaclean, 2015-09-30 20:01:01]

The CQ bit was checked by wjmaclean@chromium.org to run a CQ dry run

[wjmaclean, 2015-09-30 20:01:28]

wjmaclean@chromium.org changed reviewers:
+ tdresser@chromium.org

[wjmaclean, 2015-09-30 20:01:29]

tdresser@ - Does this look sane?

[commit-bot: I haz the power, 2015-09-30 20:03:02]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1383633002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1383633002/1

[commit-bot: I haz the power, 2015-09-30 20:13:29]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-09-30 20:13:30]

Dry run: Try jobs failed on following builders:
  win_chromium_rel_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_rel_...)

[jdduke (slow), 2015-09-30 20:26:46]

jdduke@chromium.org changed reviewers:
+ jdduke@chromium.org

[jdduke (slow), 2015-09-30 20:26:46]

https://chromiumcodereview.appspot.com/1383633002/diff/1/content/browser/rend...
File content/browser/renderer_host/input/touch_event_queue.cc (right):

https://chromiumcodereview.appspot.com/1383633002/diff/1/content/browser/rend...
content/browser/renderer_host/input/touch_event_queue.cc:640:
send_touch_events_async_ = (gesture_event.event.resendingPluginId == -1);
Can't the resending and non-resending events be interleaved? So then shouldn't
we just ignore GSU's that have an invalid resendingPluginId (rather than turning
the asyns sending on/off)?

[wjmaclean, 2015-09-30 20:53:06]

On 2015/09/30 20:26:46, jdduke wrote:
>
https://chromiumcodereview.appspot.com/1383633002/diff/1/content/browser/rend...
> File content/browser/renderer_host/input/touch_event_queue.cc (right):
> 
>
https://chromiumcodereview.appspot.com/1383633002/diff/1/content/browser/rend...
> content/browser/renderer_host/input/touch_event_queue.cc:640:
> send_touch_events_async_ = (gesture_event.event.resendingPluginId == -1);
> Can't the resending and non-resending events be interleaved? So then shouldn't
> we just ignore GSU's that have an invalid resendingPluginId (rather than
turning
> the asyns sending on/off)?

Events are only resent from a guest to an embedder. Each guest has its own
GestureRecognizer, and also TouchEventQueue (TEQ). My current thinking is that
we won't be getting interleaving in a single TEQ, though the different queues
will possibly have different values for send_touch_events_async_. We *may* need
to find a way to turn off throttling in the *guest's* TEQ when it's turned off
in the host's to completely get rid of the jitter.

[jdduke (slow), 2015-09-30 22:04:55]

On 2015/09/30 20:53:06, wjmaclean wrote:
> Events are only resent from a guest to an embedder. Each guest has its own
> GestureRecognizer, and also TouchEventQueue (TEQ). My current thinking is that
> we won't be getting interleaving in a single TEQ, though the different queues
> will possibly have different values for send_touch_events_async_. We *may*
need
> to find a way to turn off throttling in the *guest's* TEQ when it's turned off
> in the host's to completely get rid of the jitter.

I see, so would it be safe to just ignore GSU's (for the sake of sync scrolling)
that have a valid plugin ID? I assume the parent would be the one responsible
for throttling in such cases?

[tdresser, 2015-10-01 12:20:04]

On 2015/09/30 22:04:55, jdduke wrote:
> On 2015/09/30 20:53:06, wjmaclean wrote:
> > Events are only resent from a guest to an embedder. Each guest has its own
> > GestureRecognizer, and also TouchEventQueue (TEQ). My current thinking is
that
> > we won't be getting interleaving in a single TEQ, though the different
queues
> > will possibly have different values for send_touch_events_async_. We *may*
> need
> > to find a way to turn off throttling in the *guest's* TEQ when it's turned
off
> > in the host's to completely get rid of the jitter.
> 
> I see, so would it be safe to just ignore GSU's (for the sake of sync
scrolling)
> that have a valid plugin ID? I assume the parent would be the one responsible
> for throttling in such cases?

I think we should safely be able to ignore GSU's with a valid plugin id.
I agree that it seems like a safer approach.

[wjmaclean, 2015-10-01 13:44:05]

Just to clarify, is this (Patchset #2) what you meant by "ignore"?

[wjmaclean, 2015-10-01 13:44:16]

The CQ bit was checked by wjmaclean@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-10-01 13:44:34]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1383633002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1383633002/20001

[tdresser, 2015-10-01 13:55:05]

Yup!

LGTM with nit.

https://chromiumcodereview.appspot.com/1383633002/diff/20001/content/browser/...
File content/browser/renderer_host/input/touch_event_queue.cc (right):

https://chromiumcodereview.appspot.com/1383633002/diff/20001/content/browser/...
content/browser/renderer_host/input/touch_event_queue.cc:658: }
Indentation looks a little off here.

[wjmaclean, 2015-10-01 14:00:49]

The CQ bit was checked by wjmaclean@chromium.org

[wjmaclean, 2015-10-01 14:00:49]

The patchset sent to the CQ was uploaded after l-g-t-m from
tdresser@chromium.org
Link to the patchset: https://chromiumcodereview.appspot.com/1383633002/#ps40001
(title: "Fix brace indent.")

[wjmaclean, 2015-10-01 14:01:03]

Thanks!

https://chromiumcodereview.appspot.com/1383633002/diff/20001/content/browser/...
File content/browser/renderer_host/input/touch_event_queue.cc (right):

https://chromiumcodereview.appspot.com/1383633002/diff/20001/content/browser/...
content/browser/renderer_host/input/touch_event_queue.cc:658: }
On 2015/10/01 13:55:05, tdresser wrote:
> Indentation looks a little off here.

Done.

[commit-bot: I haz the power, 2015-10-01 14:02:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1383633002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1383633002/40001

[commit-bot: I haz the power, 2015-10-01 15:10:52]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-10-01 15:11:22]

Patchset 3 (id:??) landed as
https://crrev.com/1ab6149f525fc879dbc068d047501ca715a589d6
Cr-Commit-Position: refs/heads/master@{#351804}