Force immediate Mojo EDK shutdown on IO loop destruction.

Force immediate Mojo EDK shutdown on IO loop destruction.

This is an addendum to r329262 which still didn't quite
get it right.

The assumption here is now that the IPCSupportInitializer
must outlive the IO thread anyway, so the IO loop observer
may (and should) call ForceShutdown immediately rather
than posting a task.

BUG=486004
R=jam@chromium.org

Committed: https://crrev.com/5085101bae8a25ad039e1b48d721ddc081e64d17
Cr-Commit-Position: refs/heads/master@{#329692}

[Ken Rockot(use gerrit already), 2015-05-13 17:32:05]

please take a look; pulled this out of the roll CL since it should actually fix
UAF in the linked bug.

[Hajime Morrita, 2015-05-13 17:50:17]

morrita@chromium.org changed reviewers:
+ morrita@chromium.org

[Hajime Morrita, 2015-05-13 17:50:17]

lgtm

[Ken Rockot(use gerrit already), 2015-05-13 17:51:04]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2015-05-13 17:51:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1143433003/1

[commit-bot: I haz the power, 2015-05-13 19:10:55]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-05-13 19:11:47]

Patchset 1 (id:??) landed as
https://crrev.com/5085101bae8a25ad039e1b48d721ddc081e64d17
Cr-Commit-Position: refs/heads/master@{#329692}