mojo: fix little bug in DynamicServiceLoader/Runner

mojo: fix little bug in DynamicServiceLoader/Runner

If the Runner wasn't actually told to Start(), the DelegateSimpleThread would
DCHECK for not having been Started. This patch makes it so we don't allocate
a DST unless we need to.

Also adds basic unittest files for DynamicServiceLoader and Runner.

BUG=

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=285650

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=285849

[tim (not reviewing), 2014-07-25 01:37:46]

+darin, I closed out a TODO(darin) since it made this more easily testable.

[darin (slow to review), 2014-07-25 02:34:01]

LGTM

https://codereview.chromium.org/414173004/diff/80001/mojo/shell/in_process_dy...
File mojo/shell/in_process_dynamic_service_runner_unittest.cc (right):

https://codereview.chromium.org/414173004/diff/80001/mojo/shell/in_process_dy...
mojo/shell/in_process_dynamic_service_runner_unittest.cc:13: base::MessageLoop
loop_;
nit: no trailing "_"

[tim (not reviewing), 2014-07-25 17:24:31]

The CQ bit was checked by tim@chromium.org

[commit-bot: I haz the power, 2014-07-25 17:25:56]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tim@chromium.org/414173004/100001

[commit-bot: I haz the power, 2014-07-25 19:57:27]

Change committed as 285650

[Mark P, 2014-07-25 20:37:13]

A revert of this CL has been created in
https://codereview.chromium.org/419683006/ by mpearson@chromium.org.

The reason for reverting is: 
One of the test this adds fails on linux clang bog:
http://build.chromium.org/p/chromium.linux/builders/Linux%20Clang%20%28dbg%29...
.

[tim (not reviewing), 2014-07-26 00:17:54]

Note to self: CQ doesn't run these tests :(

Will wait until try-run turns green to land, but I suspect the latest patch set
fixes the problem.

[darin (slow to review), 2014-07-26 03:47:46]

On 2014/07/26 00:17:54, timsteele wrote:
> Note to self: CQ doesn't run these tests :(
> 
> Will wait until try-run turns green to land, but I suspect the latest patch
set
> fixes the problem.

There's a P0 bug about making CQ == main waterfall config.

See https://code.google.com/p/chromium/issues/detail?id=380709

We can add more bugs blocking that bug.

[tim (not reviewing), 2014-07-27 18:32:33]

The CQ bit was checked by tim@chromium.org

[commit-bot: I haz the power, 2014-07-27 18:32:52]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/tim@chromium.org/414173004/120001

[commit-bot: I haz the power, 2014-07-28 00:26:26]

Change committed as 285849