Roll FFmpeg for M45.

Roll FFmpeg for M45.

Also fixing test failures. Mostly updating test expectations.

Also includes fix for clockless renderer logic that caused some
tests to stall.

BUG=501445
Committed: https://crrev.com/37d8e763ddfe3d1f2fd5421ff677c7273bd390b1
Cr-Commit-Position: refs/heads/master@{#338372}

[chcunningham, 2015-06-25 23:52:57]

chcunningham@chromium.org changed reviewers:
+ dalecurtis@chromium.org, wolenetz@chromium.org

[chcunningham, 2015-06-25 23:52:58]

Mostly updating expectations, and also fixing our clockless wall clock stuff. 

Prior to this fix, some tests would stall out because the wallclock time for a
packet was advancing at the same pace as the clock itself, meaning we never
exhuasted our buffer queues and the tests that wait for "ended" just got stuck
waiting forever.

[wolenetz, 2015-06-26 00:06:40]

lgtm

[DaleCurtis, 2015-07-01 21:42:30]

lgtm, but presumably you need to update DEPS for this too?

https://codereview.chromium.org/1207423003/diff/1/media/renderers/renderer_im...
File media/renderers/renderer_impl.cc (right):

https://codereview.chromium.org/1207423003/diff/1/media/renderers/renderer_im...
media/renderers/renderer_impl.cc:242: *wall_clock_times =
std::vector<base::TimeTicks>(1,
wall clock times should be empty, so add a DCHECK(wall_clock_times->empty()) at
l.241 and then:

wall_clock_times->push_back(now);

ditto for below:

push_back(0 + media_time);

[wolenetz, 2015-07-01 22:46:04]

On 2015/07/01 21:42:30, DaleCurtis wrote:
> lgtm, but presumably you need to update DEPS for this too?

Indeed. If this works before and after rolling ffmpeg M45 DEPS, then I would
want to know how-so (and would retract my l*tm until that's clarified...)

Also, including even a temporary merge-m45 commit hash DEPS roll in this will
help you get some try-bot coverage now. For eventual CQ, you would need to roll
DEPS to a chromium ffmpeg master commit hash.

[wolenetz, 2015-07-10 19:17:15]

On 2015/07/01 22:46:04, wolenetz wrote:
> On 2015/07/01 21:42:30, DaleCurtis wrote:
> > lgtm, but presumably you need to update DEPS for this too?
> 
> Indeed. If this works before and after rolling ffmpeg M45 DEPS, then I would
> want to know how-so (and would retract my l*tm until that's clarified...)
> 
> Also, including even a temporary merge-m45 commit hash DEPS roll in this will
> help you get some try-bot coverage now. For eventual CQ, you would need to
roll
> DEPS to a chromium ffmpeg master commit hash.

Chris provided clarification offline to me. The regression tests worked before
because the PIPELINE_ERROR expectation matched what old ffmpeg provided. The
updated tests now work because error is not result of ffmpeg now; the spinning
clockless part is now reached by new ffmpeg, and that has also been fixed in
this CL.

[chcunningham, 2015-07-10 20:33:59]

The CQ bit was checked by chcunningham@chromium.org

[chcunningham, 2015-07-10 20:33:59]

The patchset sent to the CQ was uploaded after l-g-t-m from
dalecurtis@chromium.org, wolenetz@chromium.org
Link to the patchset: https://codereview.chromium.org/1207423003/#ps80001
(title: "Pointing deps to updated ffmpeg master")

[commit-bot: I haz the power, 2015-07-10 20:34:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1207423003/80001

[wolenetz, 2015-07-10 20:36:26]

On 2015/07/10 20:34:55, commit-bot: I haz the power wrote:
> CQ is trying da patch. Follow status at
>  https://chromium-cq-status.appspot.com/patch-status/1207423003/80001

Ah. Nit: please update cl title and description ( roll FFmpeg for M45 )

[commit-bot: I haz the power, 2015-07-10 21:45:27]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-07-10 21:47:47]

Patchset 5 (id:??) landed as
https://crrev.com/37d8e763ddfe3d1f2fd5421ff677c7273bd390b1
Cr-Commit-Position: refs/heads/master@{#338372}

[DaleCurtis, 2015-07-13 02:00:26]

\o/, nice work :)

https://codereview.chromium.org/1207423003/diff/80001/media/renderers/rendere...
File media/renderers/renderer_impl.cc (right):

https://codereview.chromium.org/1207423003/diff/80001/media/renderers/rendere...
media/renderers/renderer_impl.cc:246: for (auto const &media_time :
media_timestamps) {
Style is wrong, should be const auto&. No need for {} either.