Tell Graphics3D not to lock on Flush().

Tell Graphics3D not to lock on Flush().

The VideoDecoder already takes the proxy lock on entry. Make sure it is not
taken again when flushing Graphics3D.

BUG=166951
TEST=Several Youtube videos on Lumpy.
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=176035

[ilja, 2013-01-07 22:12:53]

*ping*
Nicholas, I assume you are the right person for this review. If not please let
me know!

[nfullagar1, 2013-01-07 22:16:46]

lgtm

[Scott Hess - ex-Googler, 2013-01-07 22:41:02]

https://codereview.chromium.org/11781004/diff/1/ppapi/shared_impl/ppb_video_d...
File ppapi/shared_impl/ppb_video_decoder_shared.cc (right):

https://codereview.chromium.org/11781004/diff/1/ppapi/shared_impl/ppb_video_d...
ppapi/shared_impl/ppb_video_decoder_shared.cc:94: if (gles2_impl_) {
This looks broadly like what I expected it to look like, BUT I cannot tell if
this is the right place for it.  It seems like the _shared code shouldn't refer
out to proxy-related stuff like this.  Instead maybe ppb_video_decoder_proxy.cc
code should call something like FlushCommandBufferWhileLocked() [or something]
which does the ScopedNoLocking and then calls this FlushCommandBuffer().  Maybe.

Or make FlushCommandBuffer() virtual and do it that way, I do not know.  It just
feels like the proxy-lock-specific code belongs with the _proxy, so that it
doesn't run in the _impl.

[Scott Hess - ex-Googler, 2013-01-07 22:41:54]

also: I do not know what I am talking about, but something feels funky.  It
might be right, Brett might say "Yeah, this is fine".

[ilja, 2013-01-07 22:47:12]

Brett, as an owner do you have an opinion on Scott's comment (e.g. where the
unlock should live)? I'll be happy to move it.

[brettw, 2013-01-09 20:52:49]

It looks like given the state of how the existing code works, this is in-line
with that. I would indeed prefer to have a design that didn't require doing this
however.

[brettw, 2013-01-09 23:13:35]

lgtm

[commit-bot: I haz the power, 2013-01-09 23:14:04]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ihf@chromium.org/11781004/1

[commit-bot: I haz the power, 2013-01-10 02:18:08]

Retried try job too often on win_aura for step(s) content_browsertests

[commit-bot: I haz the power, 2013-01-10 02:22:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ihf@chromium.org/11781004/1

[commit-bot: I haz the power, 2013-01-10 05:38:20]

Retried try job too often on win_aura for step(s) content_browsertests

[ilja, 2013-01-10 06:00:28]

Submitted manually.