Mojo IPC threading fixes

Mojo IPC threading fixes

Two independent but related fixes here:

1. IPCSupportInitializer could previously outlive its IO
   runner, resulting in Mojo not being shut down in time.
   The EDK posts tasks directly to a MessageLoopForIO
   pointer because it's assumed to remain valid until
   EDK shutdown. This CL limits IPCSupportInitializer's
   lifetime to that of the IO thread, regardless of
   ScopedIPCSupport lifetime.

2. In single-process mode, the child IO thread cannot be
   used to create the client ChannelMojo, because all channels
   in-process must be created on the EDK IO thread. This problem only
   surfaced in practice because #1 above changed some shutdown
   ordering and tickled a DCHECK in EDK shutdown.

   This CL has ChannelMojo hop to its IO runner (only if
   necessary) before creating underlying messaging pipes.

BUG=None
R=morrita@chromium.org

Committed: https://crrev.com/dbb3bb6bfa0b222fb6745608cd571bb0ec8bef39
Cr-Commit-Position: refs/heads/master@{#329262}

[Ken Rockot(use gerrit already), 2015-05-08 16:26:07]

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

[Ken Rockot(use gerrit already), 2015-05-08 16:26:21]

morrita, could you please take a look?

[Ken Rockot(use gerrit already), 2015-05-08 17:07:37]

Patchset #2 (id:20001) has been deleted

[Ken Rockot(use gerrit already), 2015-05-08 17:30:01]

Patchset #3 (id:60001) has been deleted

[Ken Rockot(use gerrit already), 2015-05-08 19:19:04]

Patchset #6 (id:120001) has been deleted

[Ken Rockot(use gerrit already), 2015-05-08 19:19:13]

Patchset #5 (id:100001) has been deleted

[Ken Rockot(use gerrit already), 2015-05-08 21:16:10]

rockot@chromium.org changed reviewers:
+ viettrungluu@chromium.org

[Ken Rockot(use gerrit already), 2015-05-08 21:16:11]

It occurs to me that morrita may no longer be doing code reviews here. Trung,
could you please take a look at this?

[jam, 2015-05-11 15:53:25]

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

[jam, 2015-05-11 15:53:26]

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/ipc_channel_m...
File ipc/mojo/ipc_channel_mojo.h (right):

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/ipc_channel_m...
ipc/mojo/ipc_channel_mojo.h:143: explicit
ChannelInfoDeleter(scoped_refptr<base::TaskRunner> io_runer);
nit: io_runner

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/scoped_ipc_su...
File ipc/mojo/scoped_ipc_support.cc (right):

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/scoped_ipc_su...
ipc/mojo/scoped_ipc_support.cc:55: if (weak_initializer_)
you can only check a weak pointer on the thread where the object lives. so this
will need to be a post task if the thread that IPCSupportInitializer was
initialized on is different than the current thread.

[Ken Rockot(use gerrit already), 2015-05-11 16:11:53]

Thanks for the review

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/ipc_channel_m...
File ipc/mojo/ipc_channel_mojo.h (right):

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/ipc_channel_m...
ipc/mojo/ipc_channel_mojo.h:143: explicit
ChannelInfoDeleter(scoped_refptr<base::TaskRunner> io_runer);
On 2015/05/11 15:53:25, jam wrote:
> nit: io_runner

Done.

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/scoped_ipc_su...
File ipc/mojo/scoped_ipc_support.cc (right):

https://codereview.chromium.org/1130413002/diff/140001/ipc/mojo/scoped_ipc_su...
ipc/mojo/scoped_ipc_support.cc:55: if (weak_initializer_)
On 2015/05/11 15:53:25, jam wrote:
> you can only check a weak pointer on the thread where the object lives. so
this
> will need to be a post task if the thread that IPCSupportInitializer was
> initialized on is different than the current thread.

Fixed.

[jam, 2015-05-11 20:56:05]

lgtm from my part; I'll defer to vtl if he's reviewing.

[Hajime Morrita, 2015-05-11 22:19:21]

lgtm on ipc/mojo

https://codereview.chromium.org/1130413002/diff/160001/ipc/mojo/ipc_channel_m...
File ipc/mojo/ipc_channel_mojo.cc (right):

https://codereview.chromium.org/1130413002/diff/160001/ipc/mojo/ipc_channel_m...
ipc/mojo/ipc_channel_mojo.cc:146: base::WeakPtrFactory<ServerChannelMojo>
weak_factory_;
Having weak_factory_ for Channel is tricky but I believe this is safe
as we guarantee that the Channel is delete on IO thread...

[Ken Rockot(use gerrit already), 2015-05-11 22:41:11]

Thanks

[Ken Rockot(use gerrit already), 2015-05-11 22:41:33]

The CQ bit was checked by rockot@chromium.org

[commit-bot: I haz the power, 2015-05-11 22:43:57]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1130413002/160001

[commit-bot: I haz the power, 2015-05-11 22:53:32]

Committed patchset #6 (id:160001)

[commit-bot: I haz the power, 2015-05-11 22:54:14]

Patchset 6 (id:??) landed as
https://crrev.com/dbb3bb6bfa0b222fb6745608cd571bb0ec8bef39
Cr-Commit-Position: refs/heads/master@{#329262}