Merge to M38: MSE: Relax H.264 Baseline mimetype profile recognition

Merge to M38: MSE: Relax H.264 Baseline mimetype profile recognition

This change lets "avc[13].42y0xx", where y >= 0x8 and xx is a valid
level, be an unambiguously matched H.264 Baseline profile. This relaxes
the previous requirement of a constrained baseline where y had to be
0xE.

CodecParameter | Before                            | Now
----------------------------------------------------------------------------------------------------
avc[13].42y0xx | probably if y==E, maybe otherwise | probably if y is in [89ABCDEF], maybe otherwise

BUG=408552
TBR=mmenke@chromium.org,qinmin@chromium.org,dalecurtis@chromium.org,xhwang@chromium.org
TEST=Updated MediaCanPlayTypeTest.CodecSupportTest_*

Review URL: https://codereview.chromium.org/626163002

Cr-Commit-Position: refs/heads/master@{#298163}
(cherry picked from commit 0a0cad0dfcd0c8ea77352fb23bd2e955744355d5)

Review URL: https://codereview.chromium.org/633673003

Cr-Commit-Position: refs/branch-heads/2171@{#66}
Cr-Branched-From: 267aeeb8d85c8503a7fd12bd14654b8ea78d3974-refs/heads/master@{#297060}
(cherry picked from commit 2f8d073aeab9292dbe43dfd03c6d75fc9f2e053e)

[wolenetz, 2014-10-08 21:04:45]

Merging to M38.

[wolenetz, 2014-10-08 21:06:38]

On 2014/10/08 21:04:45, wolenetz wrote:
> Merging to M38.

Pending tree reopening.

[wolenetz, 2014-10-09 00:39:10]

Committed patchset #1 (id:1) to pending queue manually as
c64ab457fd25753e463307a69af1430bb29237f2 (tree was closed).

[wolenetz, 2014-10-09 00:46:34]

On 2014/10/09 00:39:10, wolenetz wrote:
> Committed patchset #1 (id:1) to pending queue manually as
> c64ab457fd25753e463307a69af1430bb29237f2 (tree was closed).

Note: this means the commit has not been confirmed in the real M38 branch 2125.
It's enqueued for processing by gnumbd (which adds those
Cr-Commit-Position/Cr-Branched-From details) once gnumbd is running again (once
the network outage is fixed).

[wolenetz, 2014-10-09 18:33:37]

On 2014/10/09 00:46:34, wolenetz wrote:
> On 2014/10/09 00:39:10, wolenetz wrote:
> > Committed patchset #1 (id:1) to pending queue manually as
> > c64ab457fd25753e463307a69af1430bb29237f2 (tree was closed).
> 
> Note: this means the commit has not been confirmed in the real M38 branch
2125.
> It's enqueued for processing by gnumbd (which adds those
> Cr-Commit-Position/Cr-Branched-From details) once gnumbd is running again
(once
> the network outage is fixed).

This commit made it to the real M38 branch 2125 once the gnumbd bot succeeded
last night (delayed due to network issue).