Support for system-guest on cros.

Support for system-guest on cros.

Add an explicit test comparing pids to ensure that the browser has restarted - this fixes a race.

BUG=244579
TEST=manual
NOTRY=True

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=204039

[achuithb, 2013-06-04 19:13:05]

Tim, Glenn, please take a look.

[hartmanng, 2013-06-04 19:43:25]

This sort of seems like 2 separate issues, but I guess I'm ok with it since
they're both fairly trivial fixes. Does the race condition have an associated
bug? If so please include it in the CL description.

Apart from that, LGTM assuming we do _always_ want it to start in a separate
process (ie, it doesn't break any unit tests, etc).

[achuithb, 2013-06-04 19:58:27]

On 2013/06/04 19:43:25, hartmanng wrote:
> This sort of seems like 2 separate issues, but I guess I'm ok with it since
> they're both fairly trivial fixes. Does the race condition have an associated
> bug? If so please include it in the CL description.
> 
> Apart from that, LGTM assuming we do _always_ want it to start in a separate
> process (ie, it doesn't break any unit tests, etc).

The race condition only happens on the device, ie, with system-guest. (Which is
why I think it makes sense to include the fix in this CL).

Remotely, the check for new browser is much slower since it goes over a new ssh
connection, so we never see it. Ergo, there's no bug.

Guest mode always relaunches the browser, so I the check should be good.

Thanks for the review, Glenn!

[commit-bot: I haz the power, 2013-06-04 20:59:43]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/achuith@chromium.org/16332003/1001

[commit-bot: I haz the power, 2013-06-04 20:59:53]

Change committed as 204039