Revert of Use ChannelMojo for the Browser-GPU IPC channel.

Revert of Use ChannelMojo for the Browser-GPU IPC channel. (patchset #2 id:20001 of https://codereview.chromium.org/2112783002/ )

Reason for revert:
Breaks https://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20(dbg)(1)/builds/55407/steps/content_browsertests%20on%20Ubuntu-12.04, possibly also WebKit GPU tests

Original issue's description:
> Use ChannelMojo for the Browser-GPU IPC channel.
>
> BUG=604282
>
> Committed: https://crrev.com/4e48b26d830f640f59a754c26d9b79525548faad
> Cr-Commit-Position: refs/heads/master@{#404000}

TBR=piman@chromium.org,amistry@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=604282
Committed: https://crrev.com/c1c1aaf75afe6ba160997504da93379793b1a26e
Cr-Commit-Position: refs/heads/master@{#404115}

[phoglund_chromium, 2016-07-07 09:33:30]

The CQ bit was checked by phoglund@chromium.org

[phoglund_chromium, 2016-07-07 09:33:31]

Created Revert of Use ChannelMojo for the Browser-GPU IPC channel.

[commit-bot: I haz the power, 2016-07-07 09:33:42]

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, 2016-07-07 09:34:16]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-07-07 09:35:55]

Description was changed from

==========
Revert of Use ChannelMojo for the Browser-GPU IPC channel. (patchset #2 id:20001
of https://codereview.chromium.org/2112783002/ )

Reason for revert:
Breaks
https://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20(dbg)(1)...,
possibly also WebKit GPU tests

Original issue's description:
> Use ChannelMojo for the Browser-GPU IPC channel.
> 
> BUG=604282
> 
> Committed: https://crrev.com/4e48b26d830f640f59a754c26d9b79525548faad
> Cr-Commit-Position: refs/heads/master@{#404000}

TBR=piman@chromium.org,amistry@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=604282
==========

to

==========
Revert of Use ChannelMojo for the Browser-GPU IPC channel. (patchset #2 id:20001
of https://codereview.chromium.org/2112783002/ )

Reason for revert:
Breaks
https://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20(dbg)(1)...,
possibly also WebKit GPU tests

Original issue's description:
> Use ChannelMojo for the Browser-GPU IPC channel.
>
> BUG=604282
>
> Committed: https://crrev.com/4e48b26d830f640f59a754c26d9b79525548faad
> Cr-Commit-Position: refs/heads/master@{#404000}

TBR=piman@chromium.org,amistry@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=604282

Committed: https://crrev.com/c1c1aaf75afe6ba160997504da93379793b1a26e
Cr-Commit-Position: refs/heads/master@{#404115}
==========

[commit-bot: I haz the power, 2016-07-07 09:35:57]

Patchset 1 (id:??) landed as
https://crrev.com/c1c1aaf75afe6ba160997504da93379793b1a26e
Cr-Commit-Position: refs/heads/master@{#404115}

[phoglund_chromium, 2016-07-07 11:58:53]

On 2016/07/07 09:35:57, commit-bot: I haz the power wrote:
> Patchset 1 (id:??) landed as
> https://crrev.com/c1c1aaf75afe6ba160997504da93379793b1a26e
> Cr-Commit-Position: refs/heads/master@{#404115}

Confirmed this revert greens things back up.
Here is more info on the webkit gpu errors I mentioned earlier:
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Linux%20ASAN/b...

Error text:

17:35:22.631 6260 worker/3 virtual/gpu/fast/canvas/canvas-lost-gpu-context.html
output stderr lines:
17:35:22.631 6260  
[6308:1287:0706/173516:364282140787:ERROR:gles2_cmd_decoder.cc(5023)] Error: 5
for Command kLoseContextCHROMIUM
17:35:22.631 6260  
[6308:1287:0706/173516:364282243648:ERROR:gpu_channel_manager.cc(222)] Exiting
GPU process because some drivers cannot recover from problems.
17:35:22.633 6233 [2775/40052]
virtual/gpu/fast/canvas/canvas-lost-gpu-context.html failed unexpectedly (test
timed out)
17:35:22.632 6260 worker/3 killing primary driver
17:35:22.632 6260 worker/3 killing secondary driver
17:35:22.632 6260 worker/3 virtual/gpu/fast/canvas/canvas-lost-gpu-context.html
failed:
17:35:22.632 6260 worker/3  test timed out

Happy bug hunting! The above _could_ also be because of
https://codereview.chromium.org/2126183002/, but this patch seems more likely to
be the culprit of the above too.