Merge commit '39523030108815242178ac5e209c83070bd1baef'

* commit '39523030108815242178ac5e209c83070bd1baef':
  mov: Set the timescale for data streams

Conflicts:
libavformat/movenc.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'b886f5c2f1e71b3e60e4265c500158d392b4b9a4'

* commit 'b886f5c2f1e71b3e60e4265c500158d392b4b9a4':
  mkv: Allow flushing the current cluster in progress

Conflicts:
libavformat/matroskaenc.c
libavformat/version.h

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '59f595921eb2b848a80a74aa81b6bb43038c9ebe'

* commit '59f595921eb2b848a80a74aa81b6bb43038c9ebe':
  mkv: Flush the old cluster before writing a new one

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '98308bd44face14ea3142b501d16226eec23b75a'

* commit '98308bd44face14ea3142b501d16226eec23b75a':
  mkv: Add options for specifying cluster limits

Conflicts:
libavformat/matroskaenc.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '148fbdd1c2a2a88a78ba9fd152c81c840bdb205a'

* commit '148fbdd1c2a2a88a78ba9fd152c81c840bdb205a':
  mkv: Refactor mkv_write_packet

Conflicts:
libavformat/matroskaenc.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

ffv1: rename minor to micro version

the version is 1.3.4 which makes 4 micro and not minor

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mov: Set the timescale for data streams

Data streams have a defined timebase, do not ignore it.

mpegvideo: Use picture_ptr instead of picture in ff_mpeg_draw_horiz_band()

Fixes strides and issue with mplayer and mpeg2 field pictures

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mkv: Allow flushing the current cluster in progress

Allow emitting the current cluster that is being written before
starting a new one, simplifying how to figure out where clusters
are positioned in the output stream (for live streaming).

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

mkv: Flush the old cluster before writing a new one

This simplifies keeping track of cluster boundaries for e.g.
livestreaming of WebM.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

mkv: Add options for specifying cluster limits

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

avformat/matroskadec: add WebVTT support

WebM files now support inband text tracks, as described in the
following specification:

http://wiki.webmproject.org/webm-metadata/temporal-metadata/webvtt-in-webm

The Matroska demuxer now detects the presence of WebVTT tracks,
synthesizing WebVTT packets (having codec id AV_CODEC_ID_WEBVTT) and
pushing them downstream in the normal way.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swscale: improve dither checks

Bssed on patch by Øyvind Kolås <pippin@gimp.org>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swscale: set dither to a specific value for rgb/bgr8 output

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mkv: Refactor mkv_write_packet

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Revert "doc/RELEASE_NOTES: add a note about AVISynth"

This reverts commit 3aa2257d240a5a0eb94014b9113dd91730786886.

matroskaenc: remove unneeded wavpack tag

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  ogg: Fix potential infinite discard loop

No change as this commit has been part of ffmpeg since a while already

See: d7b542ae294aaf818f2a00c5606e009cf931e77c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

ogg: Fix potential infinite discard loop

Seeking in certain broken files would cause ogg_read_timestamp
to fail because ogg_packet would go into a state where all packets
of stream 1 would be discarded until the end of the stream.

Bug-Id: 553
CC: libav-stable@libav.org
Signed-off-by: Jan Gerber <j@v2v.cc>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Fix frame width and height for some targa_y216 samples.

Fixes ticket #2864.

avformat/mpegts: resync from the smallest packet size on

This avoids the loss of a packet in many cases if the packet size was wrong.
It also improves detection of packet size changes

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mpegts_get_pcr: dont loose a packet when resyncing

This matches how read_packet() works

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mpegts: fix pos47_full

The value was off by a packet

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avformat/mpegts: print packet size warning only if new size differs from old

No case is known to have triggered this, but its more correct to check that the
new size differs.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Warn the user if a pix_fmt != yuv420p was chosen for MPEG-2 video encoding.

Fixes ticket #2870.

avcodec/jpeg2000dec: fix near null pointer dereference

Fixes Ticket2840

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mpegts: reanalyze packet size on mismatches

Fixes Ticket1812
Fixes Ticket2838

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avcodec/error_resilience: Fix handling of matrox mpeg2

Fixes Ticket2615

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mpeg12dec: also print progressive seq and chroma format

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

fate: add ffv1.0 test

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  movenc: Indicate that negative timestamps are supported

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'c34a96a5ddfa390ce2a352eca79190766c9056d4'

* commit 'c34a96a5ddfa390ce2a352eca79190766c9056d4':
  dxa: fix decoding of first I-frame by separating I/P-frame decoding

Conflicts:
libavcodec/dxa.c

See: 186e47ef6d7d90bfd8b16e77cfa5e7aeb2a497c0
Merged-by: Michael Niedermayer <michaelni@gmx.at>

avformat/matroskadec: check out_samplerate before using it in av_rescale()

Prevent assertion failure with damaged input

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

matroskadec: Improve TTA duration calculation

Calculate the duration as accurately as possible to improve decoding of samples
where the last frame is smaller than the rest.

Example:
Take lossless-audio/luckynight-partial.tak from the FATE suit and convert it
to TTA muxed into matroska:

ffmpeg -i $(SAMPLES)/lossless-audio/luckynight-partial.tak -c:a tta lucky.mka

The framemd5 output for lucky.mka without this patch:

0,          0,          0,    46080,   184320, 7c3751ddd571d2903c3cf0ab4b3e3d0a
0,      46080,      46080,    46080,   184320, 6b70c782ba1da3f933fde2daa4f96b73
0,      92160,      92160,    46080,   184320, dcf70d89c54b9a4f0b302d4ec4fb302d
0,     138240,     138240,    46080,   184320, 48713ca38b388d2ea4abf5b86ed1226f
0,     184320,     184320,    46080,   184320, 12188a23648e7ebfb07cd6fe9197b2ea
0,     230400,     230400,    46080,   184320, 49653ab8186a5d4a044ed284671a26e0
0,     276480,     276480,    46080,   184320, 5e82c6a7fe58c7ea612c03a0a2927dd4
0,     322560,     322560,    46080,   184320, 83dc449dbd9eab5f2e8ad2b4403d6a21
0,     368640,     368640,    46080,   184320, bdd6b92c23d30978d4e802d305b0fc49

With this patch:

0,          0,          0,    46080,   184320, 7c3751ddd571d2903c3cf0ab4b3e3d0a
0,      46080,      46080,    46080,   184320, 6b70c782ba1da3f933fde2daa4f96b73
0,      92160,      92160,    46080,   184320, dcf70d89c54b9a4f0b302d4ec4fb302d
0,     138240,     138240,    46080,   184320, 48713ca38b388d2ea4abf5b86ed1226f
0,     184320,     184320,    46080,   184320, 12188a23648e7ebfb07cd6fe9197b2ea
0,     230400,     230400,    46080,   184320, 49653ab8186a5d4a044ed284671a26e0
0,     276480,     276480,    46080,   184320, 5e82c6a7fe58c7ea612c03a0a2927dd4
0,     322560,     322560,    46080,   184320, 83dc449dbd9eab5f2e8ad2b4403d6a21
0,     368640,     368640,    46080,   184320, bdd6b92c23d30978d4e802d305b0fc49
0,     414720,     414720,     4230,    16920, b50b440c5bbcecb8e9fbece643447593

The duration without this patch was calculated as 418950000000, which is bigger
than uint32_t and as such stored as 2338172288.
With this patch the duration is correctly calculated as 418950.

Signed-off-by: James Almer <jamrial@gmail.com>
Approved-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

wavpackenc: do not copy samples if they are not available

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavf/matroska: Adding the new SeekPreRoll element

In order to encapsuate Opus in Matroska, there is a new element
that has been added to the Matroska Spec, SeekPreRoll. It has the
duration in nanoseconds that has to be decoded before every seek.
Spec: http://matroska.org/technical/specs/index.html#SeekPreRoll
Proposal for encapsulateing Opus in Matroska:
http://wiki.xiph.org/MatroskaOpus

Signed-off-by: Vignesh Venkatasubramanian <vigneshv@google.com>
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

movenc: Indicate that negative timestamps are supported

The mov/mp4 muxer has support for handling negative timestamps
via edit lists (which customarily is used for handling the 1-frame
delay due to B-frames as well).

Using the muxer's native way of handling it is better than using
the generic offsetting. The generic offsetting is a bit too
crude when e.g. the timebase of one track is 1/fps, where the
edit lists can handle it accurately.

Signed-off-by: Martin Storsjö <martin@martin.st>

avcodec: fix compilation without vdpau

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Revert "Merge commit of 'vdpau: remove old-style decoders'"

This reverts commit bf36dc50ea448999c8f8c7a35f6139a7040f6275, reversing
changes made to b7fc2693c70fe72936e4ce124c802ac23857c476.

Conflicts:

libavcodec/h264.c

Keeping support for the old VDPAU API has been requested by our VDPAU maintainer

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avcodec/dxa: Support printing picture debug info

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avisynth: Support video input from AviSynth 2.5 properly.

Uses the 2.5 compatibility header included with the variant of
FFMS2 that uses AviSynth's C-interface. A copy of this header is
now provided in compat/avisynth.

avs_get_row_size_p and avs_get_height_p changed between versions
2.5 and 2.6. Since the avisynth_c.h header that avformat uses
assumes AviSynth 2.6, it would cause 2.5 to crash if given any
kind of real video (the Version() function was known to work,
though).

AvxSynth was unaffected by this issue because, despite being based
on AviSynth 2.5.8 and using 2.5.8's interface version number of 3,
it actually uses 2.6's versions of these functions.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dxa: fix decoding of first I-frame by separating I/P-frame decoding

5ef7c84 broke decoding for the first keyframe due to an unnecessary
check for a reference frame.

CC: libav-stable@libav.org

dxa: fix support of decoding all frames even in the absence of references

The case where decoding should only happen for keyframes and afterwards
(CODEC_FLAG2_SHOW_ALL not set) does not fully work neither before nor
after this

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dxa: only fail with an error about reference frames if the reference frame would be used

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  dxa: Make sure the reference frame exists

See: 12eb2fd5394813a6119912b22f5dfc17b4a6b4a0
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'c59967fa7cc5bc2fa06b36c17d2c207240c06b3e'

* commit 'c59967fa7cc5bc2fa06b36c17d2c207240c06b3e':
  h261: check the mtype index

Conflicts:
libavcodec/h261dec.c

The right half of the check is impossible to be true and thus replaced by
an assert()

See: ec3cd74f2dab8e3e8234ccb994132b23d3098585
Merged-by: Michael Niedermayer <michaelni@gmx.at>

ffv1: update years in header

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffv1enc: propagate error code from write_extradata()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffv1enc: check for malloc failure

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

lavf/tee: add support for bitstream filtering

This allows to apply different bitstream filters to different outputs,
with no need to transcode.

avcodec/dirac_dwt: Remove unused ff_spatial_idwt2()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avfilter/vf_mp: remove unused sws related functions

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avfilter: remove ff_copy_int*_list

The functions are unused

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swscale: change ff_dither_8x8_128 dimensions to be consistent with the others

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dxa: Make sure the reference frame exists

Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

h261: check the mtype index

Reported-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

lavfi/transpose: call av_frame_copy_props()

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavfi/transpose: support slice threading

Signed-off-by: Paul B Mahol <onemda@gmail.com>

lavfi/tile: make color of blank/unused area configurable

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  altivec: perform an explicit unaligned load

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '9d86bfc259ae9ba7a76067ec931ff20fbb86ea2a'

* commit '9d86bfc259ae9ba7a76067ec931ff20fbb86ea2a':
  hlsenc: Don't reset the number variable when wrapping

Conflicts:
libavformat/hlsenc.c

See: 75b9fb27f516f9db7995ab2c2abb83e25cae5813
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '09c93b1b957f2049ea5fd8fb0e6f4d82680172f2'

* commit '09c93b1b957f2049ea5fd8fb0e6f4d82680172f2':
  hlsenc: Append the last incomplete segment when closing the output

Conflicts:
libavformat/hlsenc.c

See: 4630dfd1ebb6550f352ea69ce7150aaaba96d413
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'c9031c7c1446a1a63eff7c0bf50d1ee559adf3fb'

* commit 'c9031c7c1446a1a63eff7c0bf50d1ee559adf3fb':
  hlsenc: Add a proper dependency on the mpegts muxer

Conflicts:
libavformat/Makefile

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '239f55bf3c966782b781338df284f250393b9ed6'

* commit '239f55bf3c966782b781338df284f250393b9ed6':
  vp56data: Move all data tables to the .c file

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'cb214707a6cb0d3272ec0261af6f1f5d8b7dabc7'

* commit 'cb214707a6cb0d3272ec0261af6f1f5d8b7dabc7':
  vp56data: Move all shared enum/struct declarations to common header

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '38f64c03301ac66d7b54b3e4bd2bf6454f9fb2d3'

* commit '38f64c03301ac66d7b54b3e4bd2bf6454f9fb2d3':
  mpeg12decdata.h: Move all tables to the only place they are used

Conflicts:
libavcodec/mpeg12decdata.h

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'ec6c1b1d832ec3261cc3faf93a18d7b2a84883c6'

* commit 'ec6c1b1d832ec3261cc3faf93a18d7b2a84883c6':
  mpeg12decdata: Remove unused #define

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'c591d4575a6f97fbbe6145304b1ea960e8e81e14'

* commit 'c591d4575a6f97fbbe6145304b1ea960e8e81e14':
  avcodec: Replace local extern declarations for tables with header #includes

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'aa2ba8c99e5708884a56aea9c1d96e014866f8a3'

* commit 'aa2ba8c99e5708884a56aea9c1d96e014866f8a3':
  swscale: Move extern declarations for tables to swscale_internal.h

Conflicts:
libswscale/output.c
libswscale/swscale_internal.h
libswscale/swscale_unscaled.c
libswscale/utils.c
libswscale/yuv2rgb.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'd258531502b24cb653204fe4f003c8815755bdc4'

* commit 'd258531502b24cb653204fe4f003c8815755bdc4':
  swscale: Mark a bunch of tables only used within one file static

Conflicts:
libswscale/swscale_unscaled.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '060ce0c697e261ca2792a7df30dfd1bae6900a4f'

* commit '060ce0c697e261ca2792a7df30dfd1bae6900a4f':
  ivi_common: Make some tables only used within the file static

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '4b054a3400f728c54470ee6a1eefe1d82420f6a2'

* commit '4b054a3400f728c54470ee6a1eefe1d82420f6a2':
  rtpproto: Check the right feature detection macro

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '163a729725c6eb0081b0af41a7279f7d19aee86e'

* commit '163a729725c6eb0081b0af41a7279f7d19aee86e':
  electronicarts: Let functions always returning the same value return void

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '4908c8ef2706d98022bf27a5c5bca1fe109e7529'

* commit '4908c8ef2706d98022bf27a5c5bca1fe109e7529':
  electronicarts: Improve some function/variable names

Conflicts:
libavformat/electronicarts.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'a90cff137b2aca89380b0acad41cd7bb05619ece'

* commit 'a90cff137b2aca89380b0acad41cd7bb05619ece':
  electronicarts: comment wording fixes

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '288f2ffb57ae9e9eee2748aca26da3aeb3ca6f6c'

* commit '288f2ffb57ae9e9eee2748aca26da3aeb3ca6f6c':
  electronicarts: Remove bogus function documentation

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '8747fce91fca6bb8e9936497f2de05c905cf43b5'

* commit '8747fce91fca6bb8e9936497f2de05c905cf43b5':
  electronicarts: K&R formatting cosmetics

Conflicts:
libavformat/electronicarts.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'e8c0defe1322f0ff281d9bc5eee91fa1b712b6aa'

* commit 'e8c0defe1322f0ff281d9bc5eee91fa1b712b6aa':
  8bps: decode 24bit files correctly as rgb32 on bigendian

Conflicts:
libavcodec/8bps.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

altivec: perform an explicit unaligned load

Implicit vector loads on POWER7 hardware can use the VSX
instruction set instead of classic Altivec/VMX. Let's force
a VMX load in this case.

Signed-off-by: Martin Storsjö <martin@martin.st>

avfilter/vf_scale+aresample: minor simpification

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avfilter/vf_scale: generic swscale option support

With this all AVOptions from swscale can be set without each needing
changes to vf_scale.c

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

hlsenc: Don't reset the number variable when wrapping

The counter itself shouldn't be wrapped, since it is used for
determining end_pts for the next segment - only wrap the number
used for the segment file name.

Signed-off-by: Martin Storsjö <martin@martin.st>

hlsenc: Append the last incomplete segment when closing the output

Also avoid comparing NOPTS values.

Bug-id: 551
Signed-off-by: Martin Storsjö <martin@martin.st>

hlsenc: Add a proper dependency on the mpegts muxer

The hls muxer itself doesn't have any direct (object file level)
dependencies on mpegtsenc.o, and including that object file
directly doesn't ensure that it is registered so that the muxer
actually is accessible.

Signed-off-by: Martin Storsjö <martin@martin.st>

sws: add dither enum

This allows specifying more dither algorithms without using up flags and
without ambiguities.

Also initialize the new field based on the flags and use it.
Note, improving the logic of the checks is left to subsequent
commits, this here only switches from flags to enum.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swscale_unscaled: make dither_scale static, its not used elsewhere and has no prefix

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avformat/riff: add DM4V FourCC

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vp56data: Move all data tables to the .c file

vp56data: Move all shared enum/struct declarations to common header

mpeg12decdata.h: Move all tables to the only place they are used

mpeg12decdata: Remove unused #define

avcodec: Replace local extern declarations for tables with header #includes

swscale: Move extern declarations for tables to swscale_internal.h

Also add missing ff_ prefixes where necessary.

swscale: Mark a bunch of tables only used within one file static

ivi_common: Make some tables only used within the file static

rtpproto: Check the right feature detection macro

IPPROTO_IPV6 is unrelated here (it's only used in udp.c for
multicast sockopts), check for support for the sockaddr_in6
struct itself.

Signed-off-by: Martin Storsjö <martin@martin.st>

avcodec/ffv1enc: bump minor_version for the chroma_plane fix

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avcodec/ffv1enc: fix chroma_plane for rgb/rgba

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffv1: fix plane_count at version 1.4

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffv1dec: support printing information about the global header

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>