Reset the root layer on the compositor when recreating a root layer.

Reset the root layer on the compositor when recreating a root layer.

This is going to be used by https://codereview.chromium.org/975943002/

TEST=LayerOwnerTest.RecreateRootLayerWithNullCompositor
TEST=LayerOwnerTest.RecreateRootLayerWithCompositor

BUG=337596
Committed: https://crrev.com/26d2ed6065c28b46c56a1266a6c7e28c9db55d45
Cr-Commit-Position: refs/heads/master@{#319487}

[bruthig, 2015-03-06 00:12:19]

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

[bruthig, 2015-03-06 00:12:20]

sadrul@ can you please review this?

[sadrul, 2015-03-06 14:33:49]

Looks reasonable. But please get a review from a compositor owner.

https://codereview.chromium.org/983933002/diff/1/ui/compositor/layer_owner.cc
File ui/compositor/layer_owner.cc (right):

https://codereview.chromium.org/983933002/diff/1/ui/compositor/layer_owner.cc...
ui/compositor/layer_owner.cc:71: root_layer = root_layer->parent();
You don't actually need to walk up to the root, right? You can just do:
  bool is_root = !old_layer->parent();
  if (is_root && ...)
    ...

[bruthig, 2015-03-06 15:20:41]

bruthig@chromium.org changed reviewers:
+ danakj@chromium.org

[bruthig, 2015-03-06 15:20:42]

danakj@ can you take a look from a compositor perspective?

https://codereview.chromium.org/983933002/diff/1/ui/compositor/layer_owner.cc
File ui/compositor/layer_owner.cc (right):

https://codereview.chromium.org/983933002/diff/1/ui/compositor/layer_owner.cc...
ui/compositor/layer_owner.cc:71: root_layer = root_layer->parent();
On 2015/03/06 14:33:49, sadrul wrote:
> You don't actually need to walk up to the root, right? You can just do:
>   bool is_root = !old_layer->parent();
>   if (is_root && ...)
>     ...

Done.

[danakj, 2015-03-06 17:53:58]

https://codereview.chromium.org/983933002/diff/20001/ui/compositor/layer_owne...
File ui/compositor/layer_owner_unittest.cc (right):

https://codereview.chromium.org/983933002/diff/20001/ui/compositor/layer_owne...
ui/compositor/layer_owner_unittest.cc:64: EXPECT_EQ(nullptr,
layer_copy->GetCompositor());
Should you check that the compositor root layer is != layer_copy?

[bruthig, 2015-03-06 18:10:46]

danakj@ can you have another look?

https://codereview.chromium.org/983933002/diff/20001/ui/compositor/layer_owne...
File ui/compositor/layer_owner_unittest.cc (right):

https://codereview.chromium.org/983933002/diff/20001/ui/compositor/layer_owne...
ui/compositor/layer_owner_unittest.cc:64: EXPECT_EQ(nullptr,
layer_copy->GetCompositor());
On 2015/03/06 17:53:58, danakj wrote:
> Should you check that the compositor root layer is != layer_copy?

Sure, that sounds like a good idea ;)

[danakj, 2015-03-06 18:16:46]

LGTM

[bruthig, 2015-03-06 18:20:20]

The CQ bit was checked by bruthig@chromium.org

[commit-bot: I haz the power, 2015-03-06 18:21:43]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/983933002/40001

[commit-bot: I haz the power, 2015-03-06 20:35:53]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-03-06 20:36:30]

Patchset 3 (id:??) landed as
https://crrev.com/26d2ed6065c28b46c56a1266a6c7e28c9db55d45
Cr-Commit-Position: refs/heads/master@{#319487}

[Lei Zhang, 2015-03-06 22:09:13]

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

[Lei Zhang, 2015-03-06 22:09:14]

https://codereview.chromium.org/983933002/diff/40001/ui/compositor/layer_owne...
File ui/compositor/layer_owner_unittest.cc (right):

https://codereview.chromium.org/983933002/diff/40001/ui/compositor/layer_owne...
ui/compositor/layer_owner_unittest.cc:54:
ui::InitializeContextFactoryForTests(false);
Looks like you forgot to call TerminateContextFactoryForTests()

[Lei Zhang, 2015-03-06 22:14:52]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/984103003/ by thestig@chromium.org.

The reason for reverting is: Unit test forgot to
callTerminateContextFactoryForTests() and leaks memory..