Fixes viewmanager tests on android

Fixes viewmanager tests on android

There were two things causing flake:
. The bounds of the root asynchronously changes.
. Similarly the metrics are asynchronously available.
As tests expect certain things to happen in a certain order these were
causing problems.
As we don't care about these cases I've ignored tracking them. This
makes the test cleanly pass now.

I also added --repeat-count to the apptest_runner.

BUG=497920
TEST=test only change
R=msw@chromium.org

Committed: https://crrev.com/0216c9f87af7d6f25265cd93b34071fb8268db21
Cr-Commit-Position: refs/heads/master@{#333793}

[msw, 2015-06-10 18:42:05]

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
File components/view_manager/view_manager_service_apptest.cc (right):

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:314: // The bounds of
the root may change during startup on Android at random
Should this be in a #if defined(OS_ANDROID)?

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:324: // Don't track the
metrics as they are available at an indeterministic time
Should metrics tracking just be limited to !OS_ANDROID?

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:773:
ViewHierarchyChangedAddingKnownToUnknown) {
nit: fits on line above

[sky, 2015-06-10 19:36:36]

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
File components/view_manager/view_manager_service_apptest.cc (right):

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:324: // Don't track the
metrics as they are available at an indeterministic time
On 2015/06/10 18:42:05, msw wrote:
> Should metrics tracking just be limited to !OS_ANDROID?

No, since it any case these test don't care. If we do want a test that cares
about metrics tracking I can add a boolean to the tracker indicating this.

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:773:
ViewHierarchyChangedAddingKnownToUnknown) {
On 2015/06/10 18:42:05, msw wrote:
> nit: fits on line above

Done.

[msw, 2015-06-10 19:49:23]

lgtm with q (pinged)

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
File components/view_manager/view_manager_service_apptest.cc (right):

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:314: // The bounds of
the root may change during startup on Android at random
On 2015/06/10 18:42:05, msw wrote:
> Should this be in a #if defined(OS_ANDROID)?

ping

[sky, 2015-06-10 19:54:31]

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
File components/view_manager/view_manager_service_apptest.cc (right):

https://codereview.chromium.org/1174993002/diff/20001/components/view_manager...
components/view_manager/view_manager_service_apptest.cc:314: // The bounds of
the root may change during startup on Android at random
On 2015/06/10 19:49:23, msw wrote:
> On 2015/06/10 18:42:05, msw wrote:
> > Should this be in a #if defined(OS_ANDROID)?
> 
> ping

Sorry, I only responded to one. The same comment applies to that of metrics.

[sky, 2015-06-10 19:54:34]

The CQ bit was checked by sky@chromium.org

[commit-bot: I haz the power, 2015-06-10 19:55:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1174993002/40001

[commit-bot: I haz the power, 2015-06-10 20:42:18]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-06-10 20:43:06]

Patchset 3 (id:??) landed as
https://crrev.com/0216c9f87af7d6f25265cd93b34071fb8268db21
Cr-Commit-Position: refs/heads/master@{#333793}