Properly handle unmanaged processes in mojo::shell:ChildProcessHost

In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209, 591190
TBR=ben@chromium.org

Committed: https://crrev.com/1a4ffc89ec7416f2ccc69d7e5a74e5ab2f30ab66
Cr-Commit-Position: refs/heads/master@{#378649}

[boliu, 2016-03-01 23:39:10]

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

[boliu, 2016-03-01 23:39:11]

Verified locally fix works. Context is this fixes a flaky crash on android on
cq.

(btw, I don't think CC= is a thing..)

[Ken Rockot(use gerrit already), 2016-03-02 00:23:45]

Description was changed from

==========
Temporarily allow wait in mojo::shell::ChildProcessHost

This is to fix some failing tests which caught this disallowed
Wait() happening on a main thread. It's safe because the event
is always presignalled, but we should still remove this ASAP.

Adding the scoped allow to green the bots.

BUG=591209
R=jam@chromium.org
CC=boliu@chromium.org
==========

to

==========
In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use 
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209
TBR=ben@chromium.org
==========

[Ken Rockot(use gerrit already), 2016-03-02 00:24:26]

On 2016/03/02 at 00:24:02, Ken Rockot wrote:
> 

I've updated the CL to be less evil and in stead do the right thing. I'm going
to stuff it in the CQ now.

[Ken Rockot(use gerrit already), 2016-03-02 00:24:29]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2016-03-02 00:24:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1754003002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1754003002/20001

[jabdelmalek, 2016-03-02 00:34:54]

jabdelmalek@google.com changed reviewers:
+ jabdelmalek@google.com

[jabdelmalek, 2016-03-02 00:34:55]

lgtm

[jam, 2016-03-02 00:35:06]

lgtm

[Ken Rockot(use gerrit already), 2016-03-02 00:59:22]

Description was changed from

==========
In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use 
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209
TBR=ben@chromium.org
==========

to

==========
In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use 
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209,591190
TBR=ben@chromium.org
==========

[boliu, 2016-03-02 01:26:45]

PS2 verified locally too

[commit-bot: I haz the power, 2016-03-02 02:04:09]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-03-02 02:06:24]

Description was changed from

==========
In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use 
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209,591190
TBR=ben@chromium.org
==========

to

==========
In Chrome, we always let content launch child processes, and as such have
no need for ChildProcessHost's synchronization logic. Furthermore, we use
ChildProcessHost on the main thread where such logic is illegal.

This updates ChildProcessHost to be more Chrome-friendly.

BUG=591209,591190
TBR=ben@chromium.org

Committed: https://crrev.com/1a4ffc89ec7416f2ccc69d7e5a74e5ab2f30ab66
Cr-Commit-Position: refs/heads/master@{#378649}
==========

[commit-bot: I haz the power, 2016-03-02 02:06:25]

Patchset 2 (id:??) landed as
https://crrev.com/1a4ffc89ec7416f2ccc69d7e5a74e5ab2f30ab66
Cr-Commit-Position: refs/heads/master@{#378649}