[WebSocket] bufferedAmount should not decrease inside a task.

[WebSocket] bufferedAmount should not decrease inside a task.

The spec says that bufferedAmount must not decrease in the task that
send() is called.
Fixed both the new and the old implementation although the old implementaion
is at any rate broken (bufferedAmount in the old implementation includes
the protocol overhead).

BUG=159563
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=176298

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=176519

[tyoshino (SeeGerritForStatus), 2014-06-09 08:33:55]

On 2014/06/09 02:54:33, yhirano wrote:

As discussed offline, please update bufferedAmount synchronously also for Blob
case. Use Blob.size().

[yhirano, 2014-06-09 08:35:47]

On 2014/06/09 08:33:55, tyoshino wrote:
> On 2014/06/09 02:54:33, yhirano wrote:
> 
> As discussed offline, please update bufferedAmount synchronously also for Blob
> case. Use Blob.size().

Done.

[tyoshino (SeeGerritForStatus), 2014-06-09 09:24:27]

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp
(right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:110: }
blank line here

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:112: 
remove this blank line

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:196: 
one blank line

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocket.cpp:684: m_state = CLOSED;
is this change accidental?

anyway i'm fine with either keeping this or reverting this

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocket.cpp:685: m_bufferedAmount =
unhandledBufferedAmount;
is it possible unhandledBufferedAmount and m_bufferedAmount become inconsistent?

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/WebSocketChannelClient.h (right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocketChannelClient.h:47: // These two functions
are devided in order to keep the WebSocket API:
keep -> follow?

and how about s/API/API spec/

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocketChannelClient.h:49: // be exposed behind.
couldn't get the meaning of this.

[yhirano, 2014-06-09 11:59:52]

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp
(right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:110: }
On 2014/06/09 09:24:26, tyoshino wrote:
> blank line here

Done.

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:112: 
On 2014/06/09 09:24:26, tyoshino wrote:
> remove this blank line

Done.

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:196: 
On 2014/06/09 09:24:26, tyoshino wrote:
> one blank line

Done.

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocket.cpp:684: m_state = CLOSED;
On 2014/06/09 09:24:26, tyoshino wrote:
> is this change accidental?
> 
> anyway i'm fine with either keeping this or reverting this

Reverted

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocket.cpp:685: m_bufferedAmount =
unhandledBufferedAmount;
On 2014/06/09 09:24:26, tyoshino wrote:
> is it possible unhandledBufferedAmount and m_bufferedAmount become
inconsistent?

I noticed I didn't understand why unhandledBufferedAmount is needed.
I remove unhandledBufferedAmount parameter from this function and some
duplicated bufferedAmount members from WebSocketChannel classes. What do you
think?

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/WebSocketChannelClient.h (right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocketChannelClient.h:47: // These two functions
are devided in order to keep the WebSocket API:
On 2014/06/09 09:24:26, tyoshino wrote:
> keep -> follow?
> 
> and how about s/API/API spec/

Done.

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocketChannelClient.h:49: // be exposed behind.
On 2014/06/09 09:24:26, tyoshino wrote:
> couldn't get the meaning of this.

Done.

[yhirano, 2014-06-09 12:19:35]

By the way I found that I can't use BlobDataHandle::size because Blob::size is a
virtual function and can be overwritten.
I will fix the issue tomorrow.

[tyoshino (SeeGerritForStatus), 2014-06-10 07:24:56]

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/80001/Source/modules/websocket...
Source/modules/websockets/WebSocket.cpp:685: m_bufferedAmount =
unhandledBufferedAmount;
On 2014/06/09 11:59:51, yhirano wrote:
> On 2014/06/09 09:24:26, tyoshino wrote:
> > is it possible unhandledBufferedAmount and m_bufferedAmount become
> inconsistent?
> 
> I noticed I didn't understand why unhandledBufferedAmount is needed.
> I remove unhandledBufferedAmount parameter from this function and some
> duplicated bufferedAmount members from WebSocketChannel classes. What do you
> think?

Took a look at the code and history of it. I agree that it's unnecessary at
least now. All the changes on m_buffer is reported and the last report by
didClose() is just redundant. Removal looks good.

[yhirano, 2014-06-10 11:39:46]

It's ready now. PTAL

[yhirano, 2014-06-16 01:45:20]

ping

[tyoshino (SeeGerritForStatus), 2014-06-16 02:29:29]

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.cpp (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.cpp:356: if (!remain ||
!m_client || m_state == ChannelClosing || m_state == ChannelClosed)
[optional] i prefer checking "m_client" and "m_state" and returning before
declaring "remain".

then you don't need if (!remain) as you're entering while (remain > 0)

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.h (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.h:111: {
indent

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/NewWebSocketChannelImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/NewWebSocketChannelImpl.cpp:313: int64_t quota =
m_sendingQuota;
quota and bufferedAmount are different concept. as impact to performance should
be small, let's have another variable to accumulate amount of data sent, or call
didConsumeBufferedAmount() with "size" at the end of each case block.

[yhirano, 2014-06-16 04:44:27]

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.cpp (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.cpp:356: if (!remain ||
!m_client || m_state == ChannelClosing || m_state == ChannelClosed)
On 2014/06/16 02:29:28, tyoshino wrote:
> [optional] i prefer checking "m_client" and "m_state" and returning before
> declaring "remain".
> 
> then you don't need if (!remain) as you're entering while (remain > 0)

Done.

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.h (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.h:111: {
On 2014/06/16 02:29:28, tyoshino wrote:
> indent

Done.

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
File Source/modules/websockets/NewWebSocketChannelImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/160001/Source/modules/websocke...
Source/modules/websockets/NewWebSocketChannelImpl.cpp:313: int64_t quota =
m_sendingQuota;
On 2014/06/16 02:29:29, tyoshino wrote:
> quota and bufferedAmount are different concept. as impact to performance
should
> be small, let's have another variable to accumulate amount of data sent, or
call
> didConsumeBufferedAmount() with "size" at the end of each case block.

Done.

[tyoshino (SeeGerritForStatus), 2014-06-16 07:31:42]

almost lg

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.cpp:355: if (!m_client ||
m_state == ChannelClosing || m_state == ChannelClosed)
Before this change, we did update bufferedAmount in the CLOSING (ChannelClosing)
state, and the spec also looks saying so.

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.h (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.h:232: size_t
m_numConsumedBytesInCurrentFrame;
write comment for this variable please

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.cpp:380: WTF_LOG(Network, "WebSocket %p
decreaseBufferedAmount() %lu => %lu", this, m_bufferedAmount, m_bufferedAmount -
m_consumedBufferedAmount);
reflect

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.cpp:679: m_bufferedAmount -=
m_consumedBufferedAmount;
doesn't updating m_bufferedAmount here violate the spec?

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/WebSocket.h (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.h:225: unsigned long
m_consumedBufferedAmount;
write comment to say that this is pending amount not yet reflected to
m_bufferedAmount

https://codereview.chromium.org/311993006/diff/180001/Source/web/WebSocketImp...
File Source/web/WebSocketImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:100: m_bufferedAmount +=
((String)message).utf8().length();
WebString also has utf8()?

[yhirano, 2014-06-16 10:19:45]

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.cpp:355: if (!m_client ||
m_state == ChannelClosing || m_state == ChannelClosed)
On 2014/06/16 07:31:42, tyoshino wrote:
> Before this change, we did update bufferedAmount in the CLOSING
(ChannelClosing)
> state, and the spec also looks saying so.

Done.

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/MainThreadWebSocketChannel.h (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/MainThreadWebSocketChannel.h:232: size_t
m_numConsumedBytesInCurrentFrame;
On 2014/06/16 07:31:42, tyoshino wrote:
> write comment for this variable please

Done.

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.cpp:380: WTF_LOG(Network, "WebSocket %p
decreaseBufferedAmount() %lu => %lu", this, m_bufferedAmount, m_bufferedAmount -
m_consumedBufferedAmount);
On 2014/06/16 07:31:42, tyoshino wrote:
> reflect

Done.

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.cpp:679: m_bufferedAmount -=
m_consumedBufferedAmount;
On 2014/06/16 07:31:42, tyoshino wrote:
> doesn't updating m_bufferedAmount here violate the spec?

The spec says:
The bufferedAmount attribute must return the number of bytes of application data
(UTF-8 text and binary data) that have been queued using send() ...

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/WebSocket.h (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.h:225: unsigned long
m_consumedBufferedAmount;
On 2014/06/16 07:31:42, tyoshino wrote:
> write comment to say that this is pending amount not yet reflected to
> m_bufferedAmount

Done.

https://codereview.chromium.org/311993006/diff/180001/Source/web/WebSocketImp...
File Source/web/WebSocketImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:100: m_bufferedAmount +=
((String)message).utf8().length();
On 2014/06/16 07:31:42, tyoshino wrote:
> WebString also has utf8()?

Done.

[yhirano, 2014-06-16 11:53:23]

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
File Source/modules/websockets/WebSocket.cpp (right):

https://codereview.chromium.org/311993006/diff/180001/Source/modules/websocke...
Source/modules/websockets/WebSocket.cpp:679: m_bufferedAmount -=
m_consumedBufferedAmount;
On 2014/06/16 10:19:44, yhirano wrote:
> On 2014/06/16 07:31:42, tyoshino wrote:
> > doesn't updating m_bufferedAmount here violate the spec?
> 
> The spec says:
> The bufferedAmount attribute must return the number of bytes of application
data
> (UTF-8 text and binary data) that have been queued using send() ...

Done.

[tyoshino (SeeGerritForStatus), 2014-06-16 12:37:48]

lgtm

[yhirano, 2014-06-17 01:07:23]

+tkent for Source/web, Source/platform and public/web

[yhirano, 2014-06-17 01:07:44]

+tkent for Source/web, Source/platform and public/web

[tkent, 2014-06-17 01:34:31]

rslgtm

[yhirano, 2014-06-17 01:38:25]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-17 01:39:07]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/260001

[commit-bot: I haz the power, 2014-06-17 06:01:56]

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

[commit-bot: I haz the power, 2014-06-17 06:01:57]

Try jobs failed on following builders:
  linux_blink_dbg on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/linux_blink_dbg/builds/1...)

[yhirano, 2014-06-17 06:19:30]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-17 06:19:52]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/270001

[commit-bot: I haz the power, 2014-06-17 08:11:02]

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

[commit-bot: I haz the power, 2014-06-17 08:11:03]

Try jobs failed on following builders:
  mac_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/mac_blink_rel/builds/11636)

[yhirano, 2014-06-17 08:11:37]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-17 08:12:23]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/270001

[commit-bot: I haz the power, 2014-06-17 08:45:27]

Change committed as 176298

[haraken, 2014-06-17 12:51:12]

On 2014/06/17 08:45:27, I haz the power (commit-bot) wrote:
> Change committed as 176298

Reverted this CL in r176311 since it broke browser_tests (WebSocket2):
http://build.chromium.org/p/chromium.webkit/builders/Mac10.8%20Tests/builds/8151

[yhirano, 2014-06-18 01:17:57]

On 2014/06/17 12:51:12, haraken wrote:
> On 2014/06/17 08:45:27, I haz the power (commit-bot) wrote:
> > Change committed as 176298
> 
> Reverted this CL in r176311 since it broke browser_tests (WebSocket2):
>
http://build.chromium.org/p/chromium.webkit/builders/Mac10.8%20Tests/builds/8151

reopen

[yhirano, 2014-06-18 09:18:13]

The CL was reverted because it broke some browser tests.
I will disable them on chromium side
(https://codereview.chromium.org/340733002/).
On the other hand, I noticed that my WebSocketImpl implementation was wrong with
unhandledBufferedAmount, so I fixed it at PS17.
Moreover, I added two new functions, didConsumeBufferedAmount and didClose
having three parameters.
They will replace didUpdateBufferedAmount and didClose having four parameters in
the future respectively.

tyoshino, tkent, can you take a look again, please?

[tkent, 2014-06-19 00:01:05]

rslgtm

[tyoshino (SeeGerritForStatus), 2014-06-19 07:38:55]

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
File Source/platform/network/SocketStreamHandle.cpp (right):

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
Source/platform/network/SocketStreamHandle.cpp:262: } while (!pending &&
!m_buffer.isEmpty());
ideally, didConsumeBufferedAmount should happen outside do-while. this code is
going away. please just add FIXME to remember. no need to fix this now.

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
File Source/platform/network/SocketStreamHandleClient.h (right):

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
Source/platform/network/SocketStreamHandleClient.h:48: virtual void
didConsumeBufferedAmount(SocketStreamHandle*, size_t amount) { }
amount -> consumed for consistency with the other files?

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
File Source/web/WebSocketImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:107: // Deprecate this call.
FIXME

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:120: // Deprecate this call.
FIXME

[yhirano, 2014-06-19 07:49:43]

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
File Source/platform/network/SocketStreamHandle.cpp (right):

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
Source/platform/network/SocketStreamHandle.cpp:262: } while (!pending &&
!m_buffer.isEmpty());
On 2014/06/19 07:38:54, tyoshino wrote:
> ideally, didConsumeBufferedAmount should happen outside do-while. this code is
> going away. please just add FIXME to remember. no need to fix this now.

Done.

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
File Source/platform/network/SocketStreamHandleClient.h (right):

https://codereview.chromium.org/311993006/diff/350001/Source/platform/network...
Source/platform/network/SocketStreamHandleClient.h:48: virtual void
didConsumeBufferedAmount(SocketStreamHandle*, size_t amount) { }
On 2014/06/19 07:38:54, tyoshino wrote:
> amount -> consumed for consistency with the other files?

Done.

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
File Source/web/WebSocketImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:107: // Deprecate this call.
On 2014/06/19 07:38:54, tyoshino wrote:
> FIXME

Done.

https://codereview.chromium.org/311993006/diff/350001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:120: // Deprecate this call.
On 2014/06/19 07:38:54, tyoshino wrote:
> FIXME

Done.

https://codereview.chromium.org/311993006/diff/390001/Source/web/WebSocketImp...
File Source/web/WebSocketImpl.cpp (right):

https://codereview.chromium.org/311993006/diff/390001/Source/web/WebSocketImp...
Source/web/WebSocketImpl.cpp:110: if (m_isClosingOrClosed)
I added return here in order not to send messages after close.

[tyoshino (SeeGerritForStatus), 2014-06-19 08:45:13]

lgtm

[yhirano, 2014-06-19 08:47:05]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-19 08:48:02]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/390001

[commit-bot: I haz the power, 2014-06-19 10:16:40]

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

[commit-bot: I haz the power, 2014-06-19 10:16:41]

Try jobs failed on following builders:
  blink_presubmit on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/blink_presubmit/builds/8011)

[yhirano, 2014-06-19 10:18:26]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-19 10:18:36]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/390001

[commit-bot: I haz the power, 2014-06-19 11:34:47]

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

[commit-bot: I haz the power, 2014-06-19 11:34:48]

Try jobs failed on following builders:
  win_gpu_triggered_tests on tryserver.chromium.gpu
(http://build.chromium.org/p/tryserver.chromium.gpu/builders/win_gpu_triggered...)

[yhirano, 2014-06-19 11:55:27]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-19 11:56:13]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/390001

[commit-bot: I haz the power, 2014-06-19 13:25:10]

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

[commit-bot: I haz the power, 2014-06-19 13:25:11]

Try jobs failed on following builders:
  win_gpu_triggered_tests on tryserver.chromium.gpu
(http://build.chromium.org/p/tryserver.chromium.gpu/builders/win_gpu_triggered...)

[yhirano, 2014-06-19 14:03:00]

The CQ bit was checked by yhirano@chromium.org

[commit-bot: I haz the power, 2014-06-19 14:03:13]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/yhirano@chromium.org/311993006/390001

[commit-bot: I haz the power, 2014-06-19 15:56:49]

Change committed as 176519