Corrects mojom::RouteProvider registration in ChildThreadImpl

Corrects mojom::RouteProvider registration in ChildThreadImpl

We still have one instance where ChildThreadImpl is used without
ChannelMojo (utility processes launched by a service process),
causing a nullptr dereference in ChannelProxy when trying to add
the associated interface.

This CL replaces a DCHECK with a branch to avoid the crash, since
silently ignoring the call is safe and reasonable.

In examining this bug I also realized that ChildThreadImpl was
incorrectly adding the RouteProvider interface after Channel
connection, which is only safe to do in the browser, where the
remote endpoint hasn't been launched yet. So this CL fixes that
too, and clarifies the documentation in ipc_channel_proxy.h.

BUG=647251
R=jam@chromium.org

Committed: https://crrev.com/fd3d9084805b6f598ee9cb9525de745b3209e5c0
Cr-Commit-Position: refs/heads/master@{#419043}

[Ken Rockot(use gerrit already), 2016-09-15 16:30:17]

The CQ bit was checked by rockot@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-09-15 16:30:35]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Ken Rockot(use gerrit already), 2016-09-15 16:33:22]

Description was changed from

==========
Corrects mojom::RouteProvider registration in ChildThreadImpl

We still have one instance where ChildThreadImpl is used without
ChannelMojo (the channel for service process <=> utility process),
causing a nullptr dereference in ChannelProxy when trying to add
the associated interface.

This CL replaces a DCHECK with a branch to avoid the crash, since
silently ignoring the call is safe and reasonable.

In examining this bug I also realized that child processes, unlike
the browser, need to add RouteProvider before Channel init. So this
fixes that too, and clarifies the documentation in ipc_channel_proxy.h.

BUG=647251
R=jam@chromium.org
==========

to

==========
Corrects mojom::RouteProvider registration in ChildThreadImpl

We still have one instance where ChildThreadImpl is used without
ChannelMojo (utility processes launched by a service process),
causing a nullptr dereference in ChannelProxy when trying to add
the associated interface.

This CL replaces a DCHECK with a branch to avoid the crash, since
silently ignoring the call is safe and reasonable.

In examining this bug I also realized that ChildThreadImpl was
incorrectly adding the RouteProvider interface after Channel
connection, which is only safe to do in the browser, where the
remote endpoint hasn't been launched yet. So this CL fixes that
too, and clarifies the documentation in ipc_channel_proxy.h.

BUG=647251
R=jam@chromium.org
==========

[commit-bot: I haz the power, 2016-09-15 17:14:43]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-09-15 17:14:44]

Dry run: This issue passed the CQ dry run.

[jam, 2016-09-15 23:51:10]

lgtm

[Ken Rockot(use gerrit already), 2016-09-15 23:52:25]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2016-09-15 23:52:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-15 23:57:48]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-09-16 00:00:52]

Description was changed from

==========
Corrects mojom::RouteProvider registration in ChildThreadImpl

We still have one instance where ChildThreadImpl is used without
ChannelMojo (utility processes launched by a service process),
causing a nullptr dereference in ChannelProxy when trying to add
the associated interface.

This CL replaces a DCHECK with a branch to avoid the crash, since
silently ignoring the call is safe and reasonable.

In examining this bug I also realized that ChildThreadImpl was
incorrectly adding the RouteProvider interface after Channel
connection, which is only safe to do in the browser, where the
remote endpoint hasn't been launched yet. So this CL fixes that
too, and clarifies the documentation in ipc_channel_proxy.h.

BUG=647251
R=jam@chromium.org
==========

to

==========
Corrects mojom::RouteProvider registration in ChildThreadImpl

We still have one instance where ChildThreadImpl is used without
ChannelMojo (utility processes launched by a service process),
causing a nullptr dereference in ChannelProxy when trying to add
the associated interface.

This CL replaces a DCHECK with a branch to avoid the crash, since
silently ignoring the call is safe and reasonable.

In examining this bug I also realized that ChildThreadImpl was
incorrectly adding the RouteProvider interface after Channel
connection, which is only safe to do in the browser, where the
remote endpoint hasn't been launched yet. So this CL fixes that
too, and clarifies the documentation in ipc_channel_proxy.h.

BUG=647251
R=jam@chromium.org

Committed: https://crrev.com/fd3d9084805b6f598ee9cb9525de745b3209e5c0
Cr-Commit-Position: refs/heads/master@{#419043}
==========

[commit-bot: I haz the power, 2016-09-16 00:00:53]

Patchset 1 (id:??) landed as
https://crrev.com/fd3d9084805b6f598ee9cb9525de745b3209e5c0
Cr-Commit-Position: refs/heads/master@{#419043}