cc: Handle deferred GL initialization failure.

cc: Handle deferred GL initialization failure.

Adds some tests that demonstrate behaviour when deferred GL
initialization fails, and changes to prevent crashes and lose the
output surface in this case.

The changes are kept local to the deferred initialization path
in LayerTreeHostImpl.

R=boliu
BUG=

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=220107

[boliu, 2013-08-28 03:48:17]

lgtm

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
File cc/trees/layer_tree_host_impl_unittest.cc (right):

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
cc/trees/layer_tree_host_impl_unittest.cc:6234: TEST_F(LayerTreeHostImplTest,
DeferredInitializeFails_OnscreenContext_0) {
output_surface_unittests should have a test similar to this, except of course it
can't dig into LTHI. Just fyi.

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
cc/trees/layer_tree_host_impl_unittest.cc:6354: TEST_F(LayerTreeHostImplTest,
DeferredInitializeFails_OffscreenContext) {
Load of duplicated code between these 4 tests. Everything above
set_times_make_current_succeeds and the expectations before initialize. Refactor
into a fixture class?

[danakj, 2013-08-28 15:36:09]

PTAL

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
File cc/trees/layer_tree_host_impl_unittest.cc (right):

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
cc/trees/layer_tree_host_impl_unittest.cc:6234: TEST_F(LayerTreeHostImplTest,
DeferredInitializeFails_OnscreenContext_0) {
On 2013/08/28 03:48:17, boliu wrote:
> output_surface_unittests should have a test similar to this, except of course
it
> can't dig into LTHI. Just fyi.

Ya, OutputSurfaceTestInitializeNewContext3d.ClientDeferredInitializeFails. But
it's testing different things. These are showing the LTHI doesn't crash as a
result and causes the lost surface.

https://codereview.chromium.org/23670002/diff/4001/cc/trees/layer_tree_host_i...
cc/trees/layer_tree_host_impl_unittest.cc:6354: TEST_F(LayerTreeHostImplTest,
DeferredInitializeFails_OffscreenContext) {
On 2013/08/28 03:48:17, boliu wrote:
> Load of duplicated code between these 4 tests. Everything above
> set_times_make_current_succeeds and the expectations before initialize.
Refactor
> into a fixture class?

Sure!

[boliu, 2013-08-28 15:57:41]

"error: old chunk mismatch" on PS4 :/

But still lgtm based on raw diff :)

[danakj, 2013-08-28 15:59:47]

On 2013/08/28 15:57:41, boliu wrote:
> "error: old chunk mismatch" on PS4 :/
> 
> But still lgtm based on raw diff :)

Doh.. but thanks!

[commit-bot: I haz the power, 2013-08-28 16:12:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/danakj@chromium.org/23670002/10001

[commit-bot: I haz the power, 2013-08-28 21:49:51]

Change committed as 220107