[PATCH 9.6/11] ozone: evdev: Remove extra PostTask during dispatch from EventFactoryEvdev

ozone: evdev: Remove extra PostTask during dispatch from EventFactoryEvdev

Since device processing is moved off-thread, we can't enter the device
code from nested UI message loops. So we can remove the task hop in
PostUiEvent and dispatch synchronously.

WarpCursorTo() did need update to do its own post, as it is the last
remaining case of dispatch that doesn't start in its own task. Everything
else already posts in ForwardingDeviceEventDispatcher.

We can stack-allocate the ui::Events now, too. They previously needed
heap allocation to extend their lifetime through the task hop.

BUG=449710
TEST=boot link_freon & move mouse

Committed: https://crrev.com/6c31f290c1c22716e2138e72ed86aa615ba3674c
Cr-Commit-Position: refs/heads/master@{#313879}

[spang, 2015-01-26 22:44:12]

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

[alexst (slow to review), 2015-01-28 19:04:59]

https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
File ui/events/ozone/evdev/event_factory_evdev.cc (right):

https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
ui/events/ozone/evdev/event_factory_evdev.cc:253: void
EventFactoryEvdev::DispatchUiEvent(Event* event) {
Is there a reason we can't just remove this method and call the event source one
directly?

[spang, 2015-01-28 19:38:06]

https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
File ui/events/ozone/evdev/event_factory_evdev.cc (right):

https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
ui/events/ozone/evdev/event_factory_evdev.cc:253: void
EventFactoryEvdev::DispatchUiEvent(Event* event) {
On 2015/01/28 19:04:58, alexst wrote:
> Is there a reason we can't just remove this method and call the event source
one
> directly?

The only reason is that because there's an implicit cast to void* here, this is
a little less error prone.

[alexst (slow to review), 2015-01-28 19:43:54]

On 2015/01/28 19:38:06, spang wrote:
>
https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
> File ui/events/ozone/evdev/event_factory_evdev.cc (right):
> 
>
https://codereview.chromium.org/858333006/diff/1/ui/events/ozone/evdev/event_...
> ui/events/ozone/evdev/event_factory_evdev.cc:253: void
> EventFactoryEvdev::DispatchUiEvent(Event* event) {
> On 2015/01/28 19:04:58, alexst wrote:
> > Is there a reason we can't just remove this method and call the event source
> one
> > directly?
> 
> The only reason is that because there's an implicit cast to void* here, this
is
> a little less error prone.

Okay, that in addition to the base::bind that expects a type, lgtm if you leave
a comment.

[spang, 2015-01-30 07:09:20]

The CQ bit was checked by spang@chromium.org

[commit-bot: I haz the power, 2015-01-30 07:10:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/858333006/80001

[commit-bot: I haz the power, 2015-01-30 07:46:26]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-01-30 07:47:14]

Patchset 5 (id:??) landed as
https://crrev.com/6c31f290c1c22716e2138e72ed86aa615ba3674c
Cr-Commit-Position: refs/heads/master@{#313879}