ozone: dri: Re-add call to show cursor after bounds change

ozone: dri: Re-add call to show cursor after bounds change

This fixes some cases of the cursor disappearing temporarily after a
bounds change that started after r313336 ("Constrain the cursor when
overscan insets are set").

We hide the cursor during bounds changes, so we must make sure to show
it again afterwards.

BUG=427619
TEST=On link_freon, change the display resolution via Ctrl+Shift+Plus or
Ctrl+Shift+Minus, and make sure the cursor is visible afterwards.

Committed: https://crrev.com/82483b002d360b8eb1737f3a4ac83839721d5b19
Cr-Commit-Position: refs/heads/master@{#314862}

[spang, 2015-02-04 23:39:30]

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

[spang, 2015-02-05 15:39:02]

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

[pkotwicz, 2015-02-05 15:44:53]

https://codereview.chromium.org/896343002/diff/1/ui/ozone/platform/dri/dri_wi...
File ui/ozone/platform/dri/dri_window.cc (right):

https://codereview.chromium.org/896343002/diff/1/ui/ozone/platform/dri/dri_wi...
ui/ozone/platform/dri/dri_window.cc:183: cursor_->ConfineCursorToBounds(widget_,
bounds_);
I would remove the early return in DriWindow::ConfineCursorToBounds() and add a
comment as to why we cannot have an early return.

[spang, 2015-02-05 16:16:39]

https://codereview.chromium.org/896343002/diff/1/ui/ozone/platform/dri/dri_wi...
File ui/ozone/platform/dri/dri_window.cc (right):

https://codereview.chromium.org/896343002/diff/1/ui/ozone/platform/dri/dri_wi...
ui/ozone/platform/dri/dri_window.cc:183: cursor_->ConfineCursorToBounds(widget_,
bounds_);
On 2015/02/05 15:44:53, pkotwicz wrote:
> I would remove the early return in DriWindow::ConfineCursorToBounds() and add
a
> comment as to why we cannot have an early return.

Hmm.. why wouldn't we update the cursor bounds inside DriWindow::SetBounds()?
Seems like an important part of changing the bounds.

I guess this should say GetCursorConfinedBounds(), though. Corrected.

[pkotwicz, 2015-02-05 16:22:02]

AshWindowTreeHostOzone::SetBounds() calls ConfineCursorToRootWindow() so it is
unnecessary to call DriCursor::ConfineCursorToBounds() from
DriWindow::SetBounds()

Of course, my preference would be to add the logic to show/hide the cursor to on
a bounds change to DriWindowDelegateImpl

[spang, 2015-02-05 16:24:04]

On 2015/02/05 16:22:02, pkotwicz wrote:
> AshWindowTreeHostOzone::SetBounds() calls ConfineCursorToRootWindow() so it is
> unnecessary to call DriCursor::ConfineCursorToBounds() from
> DriWindow::SetBounds()
> 

These functions must work correctly in isolation. This ozone code works
independently of ash.

> Of course, my preference would be to add the logic to show/hide the cursor to
on
> a bounds change to DriWindowDelegateImpl

Mine too.

[pkotwicz, 2015-02-05 16:26:17]

lgtm

[spang, 2015-02-05 19:20:17]

The CQ bit was checked by spang@chromium.org

[commit-bot: I haz the power, 2015-02-05 19:20:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/896343002/20001

[commit-bot: I haz the power, 2015-02-05 19:57:11]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-02-05 19:59:14]

Patchset 2 (id:??) landed as
https://crrev.com/82483b002d360b8eb1737f3a4ac83839721d5b19
Cr-Commit-Position: refs/heads/master@{#314862}