Reland "Shouldn't access LayerTreeTest::layer_tree_host on impl thread"

Reland "Shouldn't access LayerTreeTest::layer_tree_host on impl thread"

As Dana mentioned on the bug, the troublesome test doesn't make sense with impl painting. I've disabled it as suggested.

BUG=235188
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=209710

[danakj, 2013-07-02 02:07:55]

https://codereview.chromium.org/18454002/diff/5001/cc/trees/layer_tree_host_u...
File cc/trees/layer_tree_host_unittest.cc (right):

https://codereview.chromium.org/18454002/diff/5001/cc/trees/layer_tree_host_u...
cc/trees/layer_tree_host_unittest.cc:2219:
SINGLE_AND_MULTI_THREAD_MAIN_THREAD_PAINT_TEST_F(
I think I'd prefer if this test just set impl_side_painting = false in
InitializeSettings(). Any tests like this we can just delete one day. Rather
than adding macro infrastructure for tests-we'll-delete. At least, that's how
other tests like this deal with things now.

[Ian Vollick, 2013-07-02 02:17:20]

On 2013/07/02 02:07:55, danakj wrote:
>
https://codereview.chromium.org/18454002/diff/5001/cc/trees/layer_tree_host_u...
> File cc/trees/layer_tree_host_unittest.cc (right):
> 
>
https://codereview.chromium.org/18454002/diff/5001/cc/trees/layer_tree_host_u...
> cc/trees/layer_tree_host_unittest.cc:2219:
> SINGLE_AND_MULTI_THREAD_MAIN_THREAD_PAINT_TEST_F(
> I think I'd prefer if this test just set impl_side_painting = false in
> InitializeSettings(). Any tests like this we can just delete one day. Rather
> than adding macro infrastructure for tests-we'll-delete. At least, that's how
> other tests like this deal with things now.

Oh, that's so much nicer! Done.

[danakj, 2013-07-02 03:16:18]

Thanks, LGTM again.

[commit-bot: I haz the power, 2013-07-02 13:45:49]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/vollick@chromium.org/18454002/7003

[commit-bot: I haz the power, 2013-07-02 17:08:26]

Change committed as 209710