Convert scroll offsets to use SyncedProperty.

Convert scroll offsets to use SyncedProperty.

This contains the logic used to synchronize scroll deltas between the
main, pending and active tree into the generic SyncedProperty class.
Each LayerImpl has a refptr to one, which is shared between
the pending and active version of that layer.  This removes
the need for the logic to backsync deltas to the pending tree, reduces
code duplication with analogous properties like page scale factor,
and clarifies the principles behind CC's synchronization approach.

This should be a no-op change.

BUG=

Committed: https://crrev.com/d0070ba1ca9f06ad5d52272c7da3827f38b1ddf0
Cr-Commit-Position: refs/heads/master@{#314053}

[aelias_OOO_until_Jul13, 2015-01-15 01:31:01]

aelias@chromium.org changed reviewers:
+ enne@chromium.org

[aelias_OOO_until_Jul13, 2015-01-15 01:36:06]

Hi enne, PTAL.  I got this to the point of seeming to work correctly when I
browse around with it, either with main-thread and impl-thread scrolls.  Sending
this out for initial review before I adjust all the unit test callsites.

[enne (OOO), 2015-01-16 23:25:15]

https://codereview.chromium.org/800613009/diff/1/cc/layers/layer_impl.cc
File cc/layers/layer_impl.cc (right):

https://codereview.chromium.org/800613009/diff/1/cc/layers/layer_impl.cc#newc...
cc/layers/layer_impl.cc:1076: RefreshScrollDelegate();
Is there a better place to do this? "CurrentScrollOffset" doesn't sound like it
should have side effects.

[aelias_OOO_until_Jul13, 2015-01-30 05:52:31]

Hi Enne, I've been iterating on this when I find the time and finally got all
the tests green.  It should be getting closer to landing, let me know what you
think.

https://codereview.chromium.org/800613009/diff/1/cc/layers/layer_impl.cc
File cc/layers/layer_impl.cc (right):

https://codereview.chromium.org/800613009/diff/1/cc/layers/layer_impl.cc#newc...
cc/layers/layer_impl.cc:1076: RefreshScrollDelegate();
On 2015/01/16 23:25:14, enne_OOO_until_jan_26 wrote:
> Is there a better place to do this? "CurrentScrollOffset" doesn't sound like
it
> should have side effects.

Fixed.  I added "RefreshFromScrollDelegate()" calls in a bunch of other places
instead.

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc
File cc/layers/layer_impl.cc (left):

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc...
cc/layers/layer_impl.cc:1226: const gfx::Transform& layer_transform =
current_layer->transform();
This deletion (and the one in SetScrollbarPosition) isn't strictly related to
the SyncedProperty change, but it's a cleanup I've been wanting to do since it's
not possible to have a CSS transform in between scroll layer and scroll clip
layer anyway, so this logic never has any effect.

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc
File cc/layers/layer_impl.cc (right):

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc...
cc/layers/layer_impl.cc:1088: // TODO(aelias): Remove all this temporary
flooring machinery when Blink
miletus@ is working on this right now and hopefully will remove the need for
sent_scroll_delta to be floored next week, but for the moment I put in a hack to
preserve the no-op-ness of the CL and make the test green.

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.h
File cc/layers/layer_impl.h (right):

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.h#...
cc/layers/layer_impl.h:425: // TODO(aelias): Delete these three methods after
doing a mass rename in
I want to rename these methods but they're used in a huge amount of tests, so I
left in stubs here.  I figure it will make review and landing this patch easier
if it contains more substantive changes, and I'll make another patch for the
mass rename after this lands.

[enne (OOO), 2015-01-30 21:08:14]

lgtm.  I think this is all pretty straightforward, and you've commented on the
parts where it's not, and they all seem pretty reasonable.

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc
File cc/layers/layer_impl.cc (left):

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.cc...
cc/layers/layer_impl.cc:1226: const gfx::Transform& layer_transform =
current_layer->transform();
On 2015/01/30 at 05:52:31, aelias wrote:
> This deletion (and the one in SetScrollbarPosition) isn't strictly related to
the SyncedProperty change, but it's a cleanup I've been wanting to do since it's
not possible to have a CSS transform in between scroll layer and scroll clip
layer anyway, so this logic never has any effect.

Seems fine to do that while you're here.

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.h
File cc/layers/layer_impl.h (right):

https://codereview.chromium.org/800613009/diff/120001/cc/layers/layer_impl.h#...
cc/layers/layer_impl.h:425: // TODO(aelias): Delete these three methods after
doing a mass rename in
On 2015/01/30 at 05:52:31, aelias wrote:
> I want to rename these methods but they're used in a huge amount of tests, so
I left in stubs here.  I figure it will make review and landing this patch
easier if it contains more substantive changes, and I'll make another patch for
the mass rename after this lands.

sgtm

[aelias_OOO_until_Jul13, 2015-01-31 01:47:39]

The CQ bit was checked by aelias@chromium.org

[commit-bot: I haz the power, 2015-01-31 01:49:28]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/800613009/220001

[commit-bot: I haz the power, 2015-01-31 06:44:50]

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

[commit-bot: I haz the power, 2015-01-31 06:44:51]

Try jobs failed on following builders:
  win8_chromium_gn_rel on tryserver.chromium.win
(http://build.chromium.org/p/tryserver.chromium.win/builders/win8_chromium_gn_...)

[aelias_OOO_until_Jul13, 2015-01-31 07:47:33]

The CQ bit was checked by aelias@chromium.org

[commit-bot: I haz the power, 2015-01-31 07:48:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/800613009/220001

[commit-bot: I haz the power, 2015-01-31 13:45:21]

Committed patchset #12 (id:220001)

[commit-bot: I haz the power, 2015-01-31 13:46:23]

Patchset 12 (id:??) landed as
https://crrev.com/d0070ba1ca9f06ad5d52272c7da3827f38b1ddf0
Cr-Commit-Position: refs/heads/master@{#314053}