Factor out latency tracking from RenderWidgetHostImpl

Factor out latency tracking from RenderWidgetHostImpl

Create a RenderWidgetHostLatencyTracker that manages all latency-related
bookkeeping for the RenderWidgetHost.

Committed: https://crrev.com/bd818547b757ade506834763453948b1f663e09e
Cr-Commit-Position: refs/heads/master@{#307111}

[jdduke (slow), 2014-12-03 22:31:29]

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

[jdduke (slow), 2014-12-03 22:31:30]

miletus@: This is just a first pass, but I wanted to get your thoughts before I
polish it up.

[Yufeng Shen (Slow to review), 2014-12-04 23:18:57]

Overall I am happy about this refactoring.

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_impl.cc (right):

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_impl.cc:1888: }
It looks like RenderWidgetHostImpl::GetWebScreenInfo() is the central point
where RWHI reads in the screen info. Maybe we can cache the deviceScaleFactor
into the latency tracker there so we don't need this whole wrapper function ?

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_latency_tracker.cc
(right):

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_latency_tracker.cc:218:
latency->AddLatencyNumber(ui::INPUT_EVENT_LATENCY_ACK_RWH_COMPONENT, 0, 0);
GestureEvent latency might already inherit the
INPUT_EVENT_LATENCY_ACK_RWH_COMPONENT from its original touch event. So we
should not add the ack component for it.

https://code.google.com/p/chromium/codesearch#chromium/src/ui/events/gestures...

If we really need to capture the ack time for GestureEvent in the future, we can
figure out something else then.

[jdduke (slow), 2014-12-05 18:36:10]

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_impl.cc (right):

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_impl.cc:1888: }
On 2014/12/04 23:18:57, Yufeng Shen wrote:
> It looks like RenderWidgetHostImpl::GetWebScreenInfo() is the central point
> where RWHI reads in the screen info. Maybe we can cache the deviceScaleFactor
> into the latency tracker there so we don't need this whole wrapper function ?

Done.

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/render_widget_host_latency_tracker.cc
(right):

https://codereview.chromium.org/779763002/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/render_widget_host_latency_tracker.cc:218:
latency->AddLatencyNumber(ui::INPUT_EVENT_LATENCY_ACK_RWH_COMPONENT, 0, 0);
On 2014/12/04 23:18:57, Yufeng Shen wrote:
> GestureEvent latency might already inherit the
> INPUT_EVENT_LATENCY_ACK_RWH_COMPONENT from its original touch event. So we
> should not add the ack component for it.
> 
>
https://code.google.com/p/chromium/codesearch#chromium/src/ui/events/gestures...
> 
> If we really need to capture the ack time for GestureEvent in the future, we
can
> figure out something else then.

Oh good catch, I completely glossed over that.

[jdduke (slow), 2014-12-05 20:24:48]

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

[jdduke (slow), 2014-12-05 20:24:49]

+sadrul@ as owner for content/browser/renderer_host.

rjkroege@ I hope you don't mind if I swap places with you as an owner here?

[sadrul, 2014-12-05 20:50:55]

rs lgtm

[rjkroege, 2014-12-05 21:15:05]

On 2014/12/05 20:24:49, jdduke wrote:
> +sadrul@ as owner for content/browser/renderer_host.
> 
> rjkroege@ I hope you don't mind if I swap places with you as an owner here?

any particular reason you don't just add yourself?

[jdduke (slow), 2014-12-05 21:15:58]

On 2014/12/05 21:15:05, rjkroege wrote:
> On 2014/12/05 20:24:49, jdduke wrote:
> > +sadrul@ as owner for content/browser/renderer_host.
> > 
> > rjkroege@ I hope you don't mind if I swap places with you as an owner here?
> 
> any particular reason you don't just add yourself?

That's fine too but I thought you'd left us :)

[jdduke (slow), 2014-12-05 21:22:16]

On 2014/12/05 21:15:58, jdduke wrote:
> On 2014/12/05 21:15:05, rjkroege wrote:
> > On 2014/12/05 20:24:49, jdduke wrote:
> > > +sadrul@ as owner for content/browser/renderer_host.
> > > 
> > > rjkroege@ I hope you don't mind if I swap places with you as an owner
here?
> > 
> > any particular reason you don't just add yourself?
> 
> That's fine too but I thought you'd left us :)

If you're happy to continue reviewing related changes in that folder, that's
more than enough for me and I'll just leave the OWNERS list as it was.

[jdduke (slow), 2014-12-05 21:29:27]

The CQ bit was checked by jdduke@chromium.org

[commit-bot: I haz the power, 2014-12-05 21:31:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/779763002/60001

[commit-bot: I haz the power, 2014-12-05 23:33:08]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2014-12-05 23:34:47]

Patchset 4 (id:??) landed as
https://crrev.com/bd818547b757ade506834763453948b1f663e09e
Cr-Commit-Position: refs/heads/master@{#307111}