Fix sending log messages from GPU child threads

Fix sending log messages from GPU child threads

This fixes a problem with sending log messages from children threads of the GPU
process. In GPU process the message goes through GpuProcessLogMessageHandler()
and to ChildThread::Send(), but there it would fail on
DCHECK(MessageLoop::current() == message_loop()), try to output another
log message about it, and since the handler is overriden, it would go back to
GpuProcessLogMessageHandler, and back again to ChildThread::Send() and thus
recurse until eventually dying silently due to stack overflow.

BUG=
TEST=manual run of a GPU child thread


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=133824

[apatrick_chromium, 2012-03-21 22:38:17]

https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
File content/gpu/gpu_child_thread.cc (right):

https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
content/gpu/gpu_child_thread.cc:37:
ChildProcess::current()->main_thread()->message_loop())
nit: we'd typically give this if-else braces since it spans so many lines.

https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
content/gpu/gpu_child_thread.cc:39: new GpuHostMsg_OnLogMessage(severity,
header, message));
nit: indent out by two spaces on this line.

https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
content/gpu/gpu_child_thread.cc:42: new GpuHostMsg_OnLogMessage(severity,
header, message));
... and this one.

[Pawel Osciak, 2012-03-24 01:02:03]

On 2012/03/21 22:38:17, apatrick_chromium wrote:
>
https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
> File content/gpu/gpu_child_thread.cc (right):
> 
>
https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
> content/gpu/gpu_child_thread.cc:37:
> ChildProcess::current()->main_thread()->message_loop())
> nit: we'd typically give this if-else braces since it spans so many lines.
> 
>
https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
> content/gpu/gpu_child_thread.cc:39: new GpuHostMsg_OnLogMessage(severity,
> header, message));
> nit: indent out by two spaces on this line.
> 
>
https://chromiumcodereview.appspot.com/9812022/diff/1/content/gpu/gpu_child_t...
> content/gpu/gpu_child_thread.cc:42: new GpuHostMsg_OnLogMessage(severity,
> header, message));
> ... and this one.

All fixed in set 3. Thanks for comments.

[apatrick_chromium, 2012-03-24 01:04:30]

lgtm

[Pawel Osciak, 2012-04-06 02:06:30]

sievers@: ping

[Pawel Osciak, 2012-04-06 02:06:31]

sievers@: ping

[no sievers, 2012-04-06 02:11:38]

On 2012/04/06 02:06:31, posciak wrote:
> sievers@: ping

Oops, sorry. Yes, looks good assuming your video thread can never outlive the
main thread object.

[commit-bot: I haz the power, 2012-04-24 22:51:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/posciak@chromium.org/9812022/2004

[commit-bot: I haz the power, 2012-04-25 00:42:32]

Change committed as 133824