~chrisccoulson/oxide/+git/ffmpeg:aarch64

Branch merges

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related charm recipes

a8045f0... by Chris Coulson on 2016-06-07

Rename libavcodec/aarch64/videodsp.S to libavcodec/aarch64/videodsp_aarch64.S

This avoids an object file name collision as both libavcodec/aarch64/videodsp.S and libavcodec/videodsp.c get compiled in to the same target in Chromium

7f03319... by Chris Cunningham <email address hidden> on 2016-05-17

avformat/utils: Check negative bps before shifting in ff_get_pcm_codec_id()

Fixes: undefined shift.

Signed-off-by: Michael Niedermayer <email address hidden>
(cherry picked from commit 2875745d354ab0ebc4af1ebaca5c5a8d26ccdc03)

Change-Id: Ib1ea34fb0a58299628edaabb9e6b8f2ccb1fb08c

33e5416... by Chrome Cunningham <email address hidden> on 2016-05-17

Merge "Improve strategy for autorename_* basename collision hack."

1e4e65d... by Chris Cunningham <email address hidden> on 2016-05-13

Improve strategy for autorename_* basename collision hack.

generate_gyp.py creates autorename_ copies of files that may produce
basename collisions. These copies are prone to going out of sync
if changes are made to the original file outside an ffmpeg roll (e.g.
a cherry-pick). To avoid this, we now #include the orignal file.

BUG=495833

Change-Id: I532b14f8ecd948794df7c32bb6ff86c66db03545

fa382f2... by Nico Weber on 2016-05-16

Merge "win: Don't set HAVE_FAST_CLZ."

53c6493... by Nico Weber on 2016-05-16

win: Don't set HAVE_FAST_CLZ.

HAVE_FAST_CLZ causes a call to _tzcnt_u32 on Windows, and that's only
available on CPUs supporting BMI intrinsics. That's not guaranteed for
all CPUs Chromium supports.

Also add --cpu=opteron to the configure call on Windows, to make sure
this stays set correctly after the next ffmpeg roll.

Change-Id: I8a6143e5370cf4685a97f4bf4af693fe1b256a52

0ba7f68... by Chris Cunningham <email address hidden> on 2016-05-13

Update autorename_libavformat_utils.c to match utils.c

In a follow up cl I will update our generate_gyp.py to #include
the original file (pre-rename), avoiding this in the future.

BUG=603495,495833

Change-Id: Id0a531e39a7940d0ec69c3056cdf7a4846757188

77fdc79... by Michael Niedermayer <email address hidden> on 2016-05-10

avformat/utils: Check bps before using it in a shift in ff_get_pcm_codec_id()

We did not check that the bits per seconds received was a reasonable
number (64 bps or less). This caused undefined behavior when we
later attempted to left shift by "bits per second" (which ended up
shifting by more than 64 bits).

Fixes undefined shift
Fixes: usan_shift

Signed-off-by: Michael Niedermayer <email address hidden>
(cherry picked from commit ea791c080dd5494b3bee0c618a3f52e371b5f320)

BUG=603495

Change-Id: Ice2b7a9b5c3dbd15d0d9806bf456885a25cbe190

f9d1463... by Chris Cunningham <email address hidden> on 2016-05-09

libavformat/oggdec: Free stream private when header parsing fails.

Leaking this private structure opens up the possibility that it may
be re-used when parsing later packets in the stream. This is
problematic if the later packets are not the same codec type (e.g.
private allocated during Vorbis parsing, but later packets are Opus
and the private is assumed to be the oggopus_private type in
opus_header()).

Signed-off-by: Michael Niedermayer <email address hidden>
(cherry picked from commit 542f725964e52201000ec34e2f23229cf534ad3a)

BUG=602185

Change-Id: If92d22fe68be601d9fac8ab4c6b3142baecab03e

7c3fcce... by Michael Niedermayer <email address hidden> on 2016-05-10

avformat/oggparseopus: Check that granule pos is within the supported range

Larger values would imply file durations of astronomic proportions and cause
overflows

Fixes integer overflow
Fixes: usan_int64_overflow

Signed-off-by: Michael Niedermayer <email address hidden>
(cherry picked from commit 8efaee3710baa87af40556a622bf2d96a27c6425)

BUG=600959

Change-Id: Ib48df949577a34d81277e681c750aa1cccc965bc