Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder

Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder (patchset #7 id:260001 of https://codereview.chromium.org/2205623002/ )

Reason for revert:
This CL caused some regressions regarding BWE stats and HW encoder
performance. Reasons include:
1) Modifying scoped_refptr<media::VideoFrame>'s timestamp causes problems
for other clients using the same media::VideoFrame.
2) WebRTC's RTP timestamp isn't suitable for using as presentation timestamp
in Mac and Win HW encoders.

BUG=641230
Original issue's description:
> Use webrtc::VideoFrame timestamp in RTCVideoEncoder
>
> This CL fixes input timestamp mismatch in RTCVideoEncoder, which
> broke googAvgEncodeMs and googEncodeUsagePercent stats in webrtc-internals
> for hardware encoders.
> With this change, we start using WebRTC given timestamp() so that
> OveruseFrameDetector can match the timestamps and calculate the stats.
>
> BUG=597087
> TEST=googAvgEncodeMs and googEncodeUsagePercent works on Mac(H264) and
> veyron_jerry(VP8).
>
> Committed: https://crrev.com/e3195490a63d9545fb1bfe560aa21680ba0b5843
> Cr-Commit-Position: refs/heads/master@{#414589}

TBR=wuchengli@chromium.org,pbos@chromium.org,posciak@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=597087
Committed: https://crrev.com/eb436ddb8b313a81e758b5c5f97ca4349b5203a0
Cr-Commit-Position: refs/heads/master@{#415752}

[emircan, 2016-08-31 17:53:24]

The CQ bit was checked by emircan@chromium.org

[emircan, 2016-08-31 17:53:25]

Created Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder

[commit-bot: I haz the power, 2016-08-31 17:54:10]

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, 2016-08-31 21:14:54]

Description was changed from

==========
Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder (patchset #7
id:260001 of https://codereview.chromium.org/2205623002/ )

Reason for revert:
This CL caused some regressions regarding BWE stats and HW encoder
performance. Reasons include:
1) Modifying scoped_refptr<media::VideoFrame>'s timestamp causes problems
for other clients using the same media::VideoFrame.
2) WebRTC's RTP timestamp isn't suitable for using as presentation timestamp
in Mac and Win HW encoders.

BUG=641230

Original issue's description:
> Use webrtc::VideoFrame timestamp in RTCVideoEncoder
> 
> This CL fixes input timestamp mismatch in RTCVideoEncoder, which
> broke googAvgEncodeMs and googEncodeUsagePercent stats in webrtc-internals
> for hardware encoders.
> With this change, we start using WebRTC given timestamp() so that
> OveruseFrameDetector can match the timestamps and calculate the stats.
> 
> BUG=597087
> TEST=googAvgEncodeMs and googEncodeUsagePercent works on Mac(H264) and
> veyron_jerry(VP8).
> 
> Committed: https://crrev.com/e3195490a63d9545fb1bfe560aa21680ba0b5843
> Cr-Commit-Position: refs/heads/master@{#414589}

TBR=wuchengli@chromium.org,pbos@chromium.org,posciak@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=597087
==========

to

==========
Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder (patchset #7
id:260001 of https://codereview.chromium.org/2205623002/ )

Reason for revert:
This CL caused some regressions regarding BWE stats and HW encoder
performance. Reasons include:
1) Modifying scoped_refptr<media::VideoFrame>'s timestamp causes problems
for other clients using the same media::VideoFrame.
2) WebRTC's RTP timestamp isn't suitable for using as presentation timestamp
in Mac and Win HW encoders.

BUG=641230

Original issue's description:
> Use webrtc::VideoFrame timestamp in RTCVideoEncoder
> 
> This CL fixes input timestamp mismatch in RTCVideoEncoder, which
> broke googAvgEncodeMs and googEncodeUsagePercent stats in webrtc-internals
> for hardware encoders.
> With this change, we start using WebRTC given timestamp() so that
> OveruseFrameDetector can match the timestamps and calculate the stats.
> 
> BUG=597087
> TEST=googAvgEncodeMs and googEncodeUsagePercent works on Mac(H264) and
> veyron_jerry(VP8).
> 
> Committed: https://crrev.com/e3195490a63d9545fb1bfe560aa21680ba0b5843
> Cr-Commit-Position: refs/heads/master@{#414589}

TBR=wuchengli@chromium.org,pbos@chromium.org,posciak@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=597087
==========

[commit-bot: I haz the power, 2016-08-31 21:14:55]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-08-31 21:17:26]

Description was changed from

==========
Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder (patchset #7
id:260001 of https://codereview.chromium.org/2205623002/ )

Reason for revert:
This CL caused some regressions regarding BWE stats and HW encoder
performance. Reasons include:
1) Modifying scoped_refptr<media::VideoFrame>'s timestamp causes problems
for other clients using the same media::VideoFrame.
2) WebRTC's RTP timestamp isn't suitable for using as presentation timestamp
in Mac and Win HW encoders.

BUG=641230

Original issue's description:
> Use webrtc::VideoFrame timestamp in RTCVideoEncoder
> 
> This CL fixes input timestamp mismatch in RTCVideoEncoder, which
> broke googAvgEncodeMs and googEncodeUsagePercent stats in webrtc-internals
> for hardware encoders.
> With this change, we start using WebRTC given timestamp() so that
> OveruseFrameDetector can match the timestamps and calculate the stats.
> 
> BUG=597087
> TEST=googAvgEncodeMs and googEncodeUsagePercent works on Mac(H264) and
> veyron_jerry(VP8).
> 
> Committed: https://crrev.com/e3195490a63d9545fb1bfe560aa21680ba0b5843
> Cr-Commit-Position: refs/heads/master@{#414589}

TBR=wuchengli@chromium.org,pbos@chromium.org,posciak@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=597087
==========

to

==========
Revert of Use webrtc::VideoFrame timestamp in RTCVideoEncoder (patchset #7
id:260001 of https://codereview.chromium.org/2205623002/ )

Reason for revert:
This CL caused some regressions regarding BWE stats and HW encoder
performance. Reasons include:
1) Modifying scoped_refptr<media::VideoFrame>'s timestamp causes problems
for other clients using the same media::VideoFrame.
2) WebRTC's RTP timestamp isn't suitable for using as presentation timestamp
in Mac and Win HW encoders.

BUG=641230

Original issue's description:
> Use webrtc::VideoFrame timestamp in RTCVideoEncoder
>
> This CL fixes input timestamp mismatch in RTCVideoEncoder, which
> broke googAvgEncodeMs and googEncodeUsagePercent stats in webrtc-internals
> for hardware encoders.
> With this change, we start using WebRTC given timestamp() so that
> OveruseFrameDetector can match the timestamps and calculate the stats.
>
> BUG=597087
> TEST=googAvgEncodeMs and googEncodeUsagePercent works on Mac(H264) and
> veyron_jerry(VP8).
>
> Committed: https://crrev.com/e3195490a63d9545fb1bfe560aa21680ba0b5843
> Cr-Commit-Position: refs/heads/master@{#414589}

TBR=wuchengli@chromium.org,pbos@chromium.org,posciak@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=597087

Committed: https://crrev.com/eb436ddb8b313a81e758b5c5f97ca4349b5203a0
Cr-Commit-Position: refs/heads/master@{#415752}
==========

[commit-bot: I haz the power, 2016-08-31 21:17:27]

Patchset 1 (id:??) landed as
https://crrev.com/eb436ddb8b313a81e758b5c5f97ca4349b5203a0
Cr-Commit-Position: refs/heads/master@{#415752}