ffmpeg.git
2 years agoBump minor versions for 3.2
Michael Niedermayer [Wed, 26 Oct 2016 18:42:27 +0000 (20:42 +0200)]
Bump minor versions for 3.2

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoRevert "avformat/utils: Discard huge timestamps which would cause overflows if used...
Michael Niedermayer [Wed, 26 Oct 2016 18:31:04 +0000 (20:31 +0200)]
Revert "avformat/utils: Discard huge timestamps which would cause overflows if used in basic computations"

Some people seem to oppose this the patch seems to have been missed for a month on the ML
lets restart the discussion and solve this after the release

This reverts commit e936c8d176efd1a0a41e22df24564b1178c79ea9.

2 years agoavcodec/dvdsubdec: Fix off by 1 error
Michael Niedermayer [Tue, 25 Oct 2016 22:11:52 +0000 (00:11 +0200)]
avcodec/dvdsubdec: Fix off by 1 error

Fixes out of array read

Found-by: Thomas Garnier using libFuzzer
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavformat/utils: Discard huge timestamps which would cause overflows if used in basic...
Michael Niedermayer [Sun, 25 Sep 2016 20:21:53 +0000 (22:21 +0200)]
avformat/utils: Discard huge timestamps which would cause overflows if used in basic computations

Allowing larger timestamps makes it impossible to calculate basic things like the
difference of 2 timestamps or their sum without checking each individual computation for
overflow.
This should avoid a significant number of overflow checks

Fixes Ticket5136

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavformat/isom: Fix old API regression with exporting max bitrate
Michael Niedermayer [Mon, 10 Oct 2016 00:06:12 +0000 (02:06 +0200)]
avformat/isom: Fix old API regression with exporting max bitrate

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agodoc/patchwork: Document the patchwork states
Michael Niedermayer [Sat, 22 Oct 2016 11:37:57 +0000 (13:37 +0200)]
doc/patchwork: Document the patchwork states

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavu: remove comma at final enumeration items to fix pedantic warnings
Michael Behrisch [Thu, 20 Oct 2016 06:48:25 +0000 (08:48 +0200)]
lavu: remove comma at final enumeration items to fix pedantic warnings

2 years agoswresample/rematrix: Fix float part of swr_set_matrix()
Vodyannikov Aleksandr [Thu, 20 Oct 2016 03:57:35 +0000 (06:57 +0300)]
swresample/rematrix: Fix float part of swr_set_matrix()

Fixes Ticket #5897.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoconfigure: Remove --enable-incompatible-libav-abi from the help output
Michael Niedermayer [Wed, 26 Oct 2016 17:07:01 +0000 (19:07 +0200)]
configure: Remove --enable-incompatible-libav-abi from the help output

As it does not work anymore since 655b6dcb34b25d591e15ede17673ea6cb8074711

Found-by: carl
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/dvdsubdec: Fix buf_size check
Michael Niedermayer [Wed, 26 Oct 2016 14:29:57 +0000 (16:29 +0200)]
avcodec/dvdsubdec: Fix buf_size check

Fixes out of array access

Found-by: Thomas Garnier using libFuzzer
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agodoc/APIchanges: Fill in some missing things
Michael Niedermayer [Wed, 26 Oct 2016 00:57:51 +0000 (02:57 +0200)]
doc/APIchanges: Fill in some missing things

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/flvdec: init AVPacket::pos to FLVTAG offset
Suman- [Tue, 18 Oct 2016 22:13:23 +0000 (15:13 -0700)]
lavf/flvdec: init AVPacket::pos to FLVTAG offset

Current code doesn't initialize AVPacket::pos. Made it point to FLVTAG so flv_read_packet can decode from pos

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agofate: Add MXF D10/DNXHD/DV25 probe tests
Tobias Rapp [Tue, 25 Oct 2016 08:06:15 +0000 (10:06 +0200)]
fate: Add MXF D10/DNXHD/DV25 probe tests

Signed-off-by: Tobias Rapp <t.rapp@noa-archive.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavfilter/scale_npp: fix passthrough mode
Yogender Gupta [Mon, 24 Oct 2016 12:32:01 +0000 (18:02 +0530)]
avfilter/scale_npp: fix passthrough mode

Reviewed-by: Timo Rothenpieler <timo@rothenpieler.org>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavc/utvideoenc: Set bits_per_coded_sample for rgba.
Carl Eugen Hoyos [Tue, 25 Oct 2016 11:44:08 +0000 (13:44 +0200)]
lavc/utvideoenc: Set bits_per_coded_sample for rgba.

Allows to write correct value for biBitCount into BITMAPINFOHEADER.
Before, ff_put_bmp_header() always wrote "24" as biBitCount
for utvideo because bits_per_coded_sample was never set by the
encoder.

2 years agoavcodec/interplayvideo: Check side data size before use
Michael Niedermayer [Tue, 25 Oct 2016 01:51:17 +0000 (03:51 +0200)]
avcodec/interplayvideo: Check side data size before use

Fixes out of array read

Found-by: Thomas Garnier using libFuzzer
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/ffv1enc: Fix storing RGB48 without explicitly set level
Michael Niedermayer [Mon, 24 Oct 2016 23:15:56 +0000 (01:15 +0200)]
avcodec/ffv1enc: Fix storing RGB48 without explicitly set level

the bps value is only stored with level >= 1, using rgb48 with level 0 requires the
user app to keep track of the bps by external means, which does not always happen
also we force level >= 1 for other 16bps formats, so this is consistent.

Found-by: Jerome Martinez <jerome@mediaarea.net>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agovf_colorspace: Add support for iec61966-2.1 (sRGB) transfer
Vittorio Giovara [Tue, 18 Oct 2016 19:17:26 +0000 (15:17 -0400)]
vf_colorspace: Add support for iec61966-2.1 (sRGB) transfer

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2 years agovf_colorspace: don't spam console with warnings if range is unspecified.
Ronald S. Bultje [Mon, 24 Oct 2016 20:03:16 +0000 (16:03 -0400)]
vf_colorspace: don't spam console with warnings if range is unspecified.

2 years agovp9: change order of operations in adapt_prob().
Ronald S. Bultje [Fri, 14 Oct 2016 17:01:27 +0000 (13:01 -0400)]
vp9: change order of operations in adapt_prob().

This is intended to workaround bug "665 Integer Divide Instruction May
Cause Unpredictable Behavior" on some early AMD CPUs, which causes a
div-by-zero in this codepath, such as reported in Mozilla bug #1293996.

Note that this isn't guaranteed to fix the bug, since a compiler is free
to reorder instructions that don't depend on each other. However, it
appears to fix the bug in Firefox, and a similar patch was applied to
libvpx also (see Chrome bug #599899).

2 years agolavf/mov: reindent
Rodger Combs [Fri, 8 Jan 2016 01:04:00 +0000 (19:04 -0600)]
lavf/mov: reindent

Reviewed-By: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/mov: improve `tref/chap` chapter handling
Rodger Combs [Thu, 7 Jan 2016 23:14:08 +0000 (17:14 -0600)]
lavf/mov: improve `tref/chap` chapter handling

3 parts:
- Supports multiple chapter streams
- Exports regular text chapter streams as opaque data. This prevents consumers
  from showing chapters as if they were regular subtitle streams.
- Exports video chapter streams as thumbnails, and provides the first one as
  an attached_pic.

2 years agolavf: add AV_DISPOSITION_TIMED_THUMBNAILS
Rodger Combs [Mon, 24 Oct 2016 10:47:05 +0000 (05:47 -0500)]
lavf: add AV_DISPOSITION_TIMED_THUMBNAILS

Reviewed-By: Michael Niedermayer <michael@niedermayer.cc>
2 years agofate/hevc: add automatic bsf test
Rodger Combs [Thu, 28 Apr 2016 22:00:43 +0000 (17:00 -0500)]
fate/hevc: add automatic bsf test

2 years agofate/aac: add automatic bsf test
Rodger Combs [Fri, 15 Apr 2016 08:18:01 +0000 (03:18 -0500)]
fate/aac: add automatic bsf test

2 years agofate/h264: make mp4toannexb test use auto-BSF
Rodger Combs [Fri, 15 Apr 2016 08:01:46 +0000 (03:01 -0500)]
fate/h264: make mp4toannexb test use auto-BSF

2 years agolavf/rawenc: add automatic bitstream filtering for H264+HEVC
Rodger Combs [Sat, 10 Sep 2016 04:27:54 +0000 (23:27 -0500)]
lavf/rawenc: add automatic bitstream filtering for H264+HEVC

2 years agolavf/movenc+dashenc: add automatic bitstream filtering
Rodger Combs [Fri, 8 Apr 2016 00:36:39 +0000 (19:36 -0500)]
lavf/movenc+dashenc: add automatic bitstream filtering

This is disabled by default when the empty_moov flag is enabled

2 years agolavf/dashenc: add deinit function
Rodger Combs [Fri, 8 Apr 2016 00:36:15 +0000 (19:36 -0500)]
lavf/dashenc: add deinit function

2 years agolavf/movenc: add deinit function
Rodger Combs [Fri, 8 Apr 2016 00:24:04 +0000 (19:24 -0500)]
lavf/movenc: add deinit function

2 years agolavf/segment: fix writing separate header with auto BSF
Rodger Combs [Fri, 8 Apr 2016 00:18:45 +0000 (19:18 -0500)]
lavf/segment: fix writing separate header with auto BSF

2 years agolavf/segment: add deinit function
Rodger Combs [Fri, 8 Apr 2016 00:18:19 +0000 (19:18 -0500)]
lavf/segment: add deinit function

2 years agolavf/mux: add avformat_init_output
Rodger Combs [Sat, 25 Jun 2016 03:02:50 +0000 (22:02 -0500)]
lavf/mux: add avformat_init_output

This allows a consumer to run the muxer's init function without actually
writing the header, which is useful in chained muxers that support
automatic bitstream filtering.

2 years agoMAINTAINERS: add myself for audiotoolbox
Rodger Combs [Mon, 24 Oct 2016 06:47:52 +0000 (01:47 -0500)]
MAINTAINERS: add myself for audiotoolbox

2 years agoffprobe: report field order for video streams
Rodger Combs [Tue, 4 Oct 2016 04:49:09 +0000 (23:49 -0500)]
ffprobe: report field order for video streams

2 years agolavc/utils: avcodec_string: dump field order when known
Rodger Combs [Tue, 4 Oct 2016 02:46:53 +0000 (21:46 -0500)]
lavc/utils: avcodec_string: dump field order when known

2 years agolavc/h264_parser: export field order in more cases
Rodger Combs [Tue, 4 Oct 2016 02:45:56 +0000 (21:45 -0500)]
lavc/h264_parser: export field order in more cases

2 years agolavc/parser: export field order if not already set
Rodger Combs [Tue, 20 Sep 2016 10:57:39 +0000 (05:57 -0500)]
lavc/parser: export field order if not already set

Some codecs set this in the parser, but not the decoder

2 years agolavfi: remove 2 unused lavc includes
Clément Bœsch [Sun, 23 Oct 2016 18:19:28 +0000 (20:19 +0200)]
lavfi: remove 2 unused lavc includes

2 years agoavutil/x86/emms: Document the emms_c() vs alloc/free relation.
Michael Niedermayer [Sun, 23 Oct 2016 02:46:57 +0000 (04:46 +0200)]
avutil/x86/emms: Document the emms_c() vs alloc/free relation.

Reviewed-by: Andreas Cadhalpun <andreas.cadhalpun@googlemail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agofaq: use relative links to own documentation
Andreas Cadhalpun [Sat, 22 Oct 2016 18:12:30 +0000 (20:12 +0200)]
faq: use relative links to own documentation

This way locally installed documentation refers to itself instead of the
website.

Bud-Id: https://bugs.debian.org/841501
Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agoavcodec/mips: loongson optimize mmi load and store operators
Zhou Xiaoyong [Mon, 10 Oct 2016 08:09:12 +0000 (16:09 +0800)]
avcodec/mips: loongson optimize mmi load and store operators

1.MMI_ load/store macros are defined in libavutil/mips/mmiutils.h
2.Replace some unnecessary unaligned access with aligned operator
3.The MMI_ load/store is compatible with cpu loongson2e/2f which not support instructions start with gs

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavutil/mips: loongson add mmi utils header file
Zhou Xiaoyong [Mon, 10 Oct 2016 08:07:59 +0000 (16:07 +0800)]
avutil/mips: loongson add mmi utils header file

1.mmiutils.h defined MMI_ load/store macros for loongson2e/2f/3a
2.mmiutils.h defined some mmi assembly macors

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/mpegts: add missed fixes to scte35 section callback
Carlos Fernandez [Sat, 22 Oct 2016 10:03:32 +0000 (12:03 +0200)]
lavf/mpegts: add missed fixes to scte35 section callback

They somehow got lost along the patch versions.

Signed-off-by: Marton Balint <cus@passwd.hu>
2 years agoavcodec/cuvid: Allow reinitialization of decoder
Philip Langdale [Sat, 22 Oct 2016 21:31:41 +0000 (14:31 -0700)]
avcodec/cuvid: Allow reinitialization of decoder

In practice, this works fine.

2 years agoavformat/matroskaenc: write DisplayWidth and DisplayHeight elements only if they...
James Almer [Mon, 17 Oct 2016 17:17:35 +0000 (14:17 -0300)]
avformat/matroskaenc: write DisplayWidth and DisplayHeight elements only if they differ from PixelWidth and PixelHeight

Signed-off-by: James Almer <jamrial@gmail.com>
2 years agompegts: handle AVMEDIA_TYPE_UNKNOWN correctly
Andreas Cadhalpun [Fri, 21 Oct 2016 23:18:38 +0000 (01:18 +0200)]
mpegts: handle AVMEDIA_TYPE_UNKNOWN correctly

It is negative, so can't be used for left shifting.

This fixes ubsan runtime error: shift exponent -1 is negative

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agoavutil/avassert: Add av_assertX_fpu()
Michael Niedermayer [Fri, 21 Oct 2016 10:50:03 +0000 (12:50 +0200)]
avutil/avassert: Add av_assertX_fpu()

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/snowenc: Clear MMX state after edge drawing and picture encode
Michael Niedermayer [Fri, 21 Oct 2016 13:10:48 +0000 (15:10 +0200)]
avcodec/snowenc: Clear MMX state after edge drawing and picture encode

Fixes undefined behavior from calling libc allocation with unclean FPU state.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/mpegvideo_enc: Add missing emms_c() to clear MMX state after SIMD use
Michael Niedermayer [Fri, 21 Oct 2016 13:01:36 +0000 (15:01 +0200)]
avcodec/mpegvideo_enc: Add missing emms_c() to clear MMX state after SIMD use

Fixes undefined behavior due to calling libc allocation with unclean FPU state

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/utils: Move emms_c() before memory allocation functions in avcodec_encode_vid...
Michael Niedermayer [Fri, 21 Oct 2016 12:47:34 +0000 (14:47 +0200)]
avcodec/utils: Move emms_c() before memory allocation functions in avcodec_encode_video2()

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/cavsdec: Clear MMX state after MB decode loop
Michael Niedermayer [Fri, 21 Oct 2016 12:38:52 +0000 (14:38 +0200)]
avcodec/cavsdec: Clear MMX state after MB decode loop

The MMX state must be cleared between using MMX and using memory allocation
thats basically the only location between the 2

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/svq1enc: Clear MMX state after svq1_encode_plane()
Michael Niedermayer [Fri, 21 Oct 2016 12:33:42 +0000 (14:33 +0200)]
avcodec/svq1enc: Clear MMX state after svq1_encode_plane()

svq1_encode_plane() uses SIMD and we call libc memory allocation functions after it

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/rtpdec_g726: Map mime type G726 to g726le.
Carl Eugen Hoyos [Sat, 22 Oct 2016 01:53:38 +0000 (03:53 +0200)]
lavf/rtpdec_g726: Map mime type G726 to g726le.

Add new mime types AAL2-G726 for g726 as suggested in rfc 3551.

This patch will break interaction with applications that incorrectly
use big-endian G.726 with mime type G726 but we know of at least one
device (DVTel camera) that correctly implements the rfc, so do the same.

Fixes ticket #5890.

2 years agolavc/libvpxenc: fix -auto-alt-ref option type
Kagami Hiiragi [Thu, 20 Oct 2016 15:31:38 +0000 (18:31 +0300)]
lavc/libvpxenc: fix -auto-alt-ref option type

vp9_cx_iface actually allows values in range [0..2].
This fixes ticket #5894.

Signed-off-by: Kagami Hiiragi <kagami@genshiken.org>
Signed-off-by: James Zern <jzern@google.com>
2 years agodoc: fix spelling errors
Andreas Cadhalpun [Thu, 13 Oct 2016 21:08:01 +0000 (23:08 +0200)]
doc: fix spelling errors

Thanks to Mathieu Malaterre <malat@debian.org> for reporting the
Que/Queue typo. (https://bugs.debian.org/839542)

Reviewed-by: Lou Logan <lou@lrcd.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agolavf/mpegts: SCTE-35 extraction from mpegts
Carlos Fernandez [Wed, 19 Oct 2016 00:36:28 +0000 (17:36 -0700)]
lavf/mpegts: SCTE-35 extraction from mpegts

Reviewed-by: Marton Balint <cus@passwd.hu>
Acked-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Carlos Fernandez <carlos@ccextractor.org>
Signed-off-by: Marton Balint <cus@passwd.hu>
2 years agolavc: add SCTE-35 CUI codec ID
Carlos Fernandez [Wed, 19 Oct 2016 00:36:27 +0000 (17:36 -0700)]
lavc: add SCTE-35 CUI codec ID

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Carlos Fernandez <carlos@ccextractor.org>
Signed-off-by: Marton Balint <cus@passwd.hu>
2 years agolavd/decklink_dec: fix indentation
Marton Balint [Thu, 13 Oct 2016 21:58:53 +0000 (23:58 +0200)]
lavd/decklink_dec: fix indentation

Signed-off-by: Marton Balint <cus@passwd.hu>
2 years agolavd/decklink_dec: add option to disable drawing bars on signal loss
Marton Balint [Thu, 13 Oct 2016 21:56:45 +0000 (23:56 +0200)]
lavd/decklink_dec: add option to disable drawing bars on signal loss

Signed-off-by: Marton Balint <cus@passwd.hu>
2 years agoavformat/mxfdec: Check size to avoid integer overflow in mxf_read_utf16_string()
Michael Niedermayer [Fri, 21 Oct 2016 17:45:21 +0000 (19:45 +0200)]
avformat/mxfdec: Check size to avoid integer overflow in mxf_read_utf16_string()

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavformat/mxfdec: Fix mixed declaration and code
Michael Niedermayer [Fri, 21 Oct 2016 17:34:05 +0000 (19:34 +0200)]
avformat/mxfdec: Fix mixed declaration and code

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agompeg12dec: unref discarded picture from extradata
Andreas Cadhalpun [Thu, 20 Oct 2016 20:51:55 +0000 (22:51 +0200)]
mpeg12dec: unref discarded picture from extradata

Otherwise another frame gets referenced into picture, triggering an assert
(from commit 13aae8) in av_frame_ref.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agocavsdec: unref frame before referencing again
Andreas Cadhalpun [Thu, 20 Oct 2016 20:14:22 +0000 (22:14 +0200)]
cavsdec: unref frame before referencing again

This fixes asserts (from commit 13aae8) in av_frame_ref and
av_frame_move_ref.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agodcstr: fix division by zero
Andreas Cadhalpun [Thu, 20 Oct 2016 18:13:54 +0000 (20:13 +0200)]
dcstr: fix division by zero

Also check for possible overflows.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agoaiff: check block_align in aiff_read_packet
Andreas Cadhalpun [Thu, 20 Oct 2016 18:08:15 +0000 (20:08 +0200)]
aiff: check block_align in aiff_read_packet

It can be unset in avcodec_parameters_from_context and a value of 0
causes SIGFPE crashes.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agoavcodec/vda: define av_vda_default_init2 when CONFIG_H264_VDA_HWACCEL equ 0
Steven Liu [Wed, 12 Oct 2016 09:36:51 +0000 (17:36 +0800)]
avcodec/vda: define av_vda_default_init2 when CONFIG_H264_VDA_HWACCEL equ 0

on OSX:
../configure --disable-everything --enable-demuxer=hls make
error message: Undefined symbols for architecture x86_64:
"_av_vda_default_init2", referenced from:_videotoolbox_init in
ffmpeg_videotoolbox.o
so add av_vda_default_init2 when CONFIG_H264_VDA_HWACCEL=0

Signed-off-by: Steven Liu <lq@chinaffmpeg.org>
Reviewed-by: wm4 <nfxjfg@googlemail.com>
Reviewed-by: Xidorn Quan <quanxunzhen@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavformat/rtmpproto: Fix RTMP control message handling error in listen mode.
Hiroyuki OYAMA [Wed, 12 Oct 2016 06:33:44 +0000 (15:33 +0900)]
avformat/rtmpproto: Fix RTMP control message handling error in listen mode.

Fix problem to fail by a RTMP Control Message except "Set Chunk Size (1)" after an RTMP handshake. When 'nginx-rtmp-module' relays an RTMP, it sends not only control message 'Set Chunk Size (1)' but also 'Window Acknowledgement Size (5)'.

Reviewed-by: Steven Liu <lingjiujianke@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/mpegvideo_enc: Clear mmx state in ff_mpv_reallocate_putbitbuffer()
Michael Niedermayer [Fri, 21 Oct 2016 12:05:00 +0000 (14:05 +0200)]
avcodec/mpegvideo_enc: Clear mmx state in ff_mpv_reallocate_putbitbuffer()

This function must be called from the mb or slice encoding loop and MMX state may not
be clean there

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/utils: Clear MMX state before returning from avcodec_default_execute*()
Michael Niedermayer [Fri, 21 Oct 2016 11:40:18 +0000 (13:40 +0200)]
avcodec/utils: Clear MMX state before returning from avcodec_default_execute*()

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavcodec/dnxhdenc: Move allocation out of radix_sort()
Michael Niedermayer [Fri, 21 Oct 2016 11:29:46 +0000 (13:29 +0200)]
avcodec/dnxhdenc: Move allocation out of radix_sort()

Its slow, its not checked, FPU state is not clean either currently there

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/segment: decide whether to rename based on list URI
Rodger Combs [Thu, 6 Oct 2016 07:00:25 +0000 (02:00 -0500)]
lavf/segment: decide whether to rename based on list URI

This fixes the case of writing segments to local files, but the list
over a network protocol.

2 years agoopenssl: Support version 1.1.0.
Matt Oliver [Sun, 9 Oct 2016 19:49:54 +0000 (06:49 +1100)]
openssl: Support version 1.1.0.

Fixes #5675

Signed-off-by: Matt Oliver <protogonoi@gmail.com>
2 years agotests/fate: add mxf metadata streams test
Mark Reid [Wed, 21 Sep 2016 20:42:09 +0000 (13:42 -0700)]
tests/fate: add mxf metadata streams test

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolibavformat/mxfdec: add metadata streams for external referenced sourclips
Mark Reid [Wed, 21 Sep 2016 20:42:08 +0000 (13:42 -0700)]
libavformat/mxfdec: add metadata streams for external referenced sourclips

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolibavformat/mxfdec: don't assume first stream index to be primary
Mark Reid [Wed, 21 Sep 2016 20:42:07 +0000 (13:42 -0700)]
libavformat/mxfdec: don't assume first stream index to be primary

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolibavformat/mxfdec: export track name metadata
Mark Reid [Wed, 21 Sep 2016 20:42:06 +0000 (13:42 -0700)]
libavformat/mxfdec: export track name metadata

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavformat/mxfdec: Detect field_order based on video_line_map
Tobias Rapp [Wed, 19 Oct 2016 12:35:20 +0000 (14:35 +0200)]
avformat/mxfdec: Detect field_order based on video_line_map

Read video_line_map from MXF generic picture essence descriptor and use
it to derive the coded field order. Use field_dominance to derive the
display field order from coded field order. If field_dominance is not
available the default value "1" is used as defined in SMPTE S377-1.

Fixes field_order detection for a bunch of DV/DVCPRO files. The heuristic
for deriving coded field order from video_line_map is inspired by
MediaInfo.

Signed-off-by: Tobias Rapp <t.rapp@noa-archive.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agorsd: limit number of channels
Andreas Cadhalpun [Wed, 19 Oct 2016 21:40:41 +0000 (23:40 +0200)]
rsd: limit number of channels

Negative values don't make sense and too large values can cause
overflows. For AV_CODEC_ID_ADPCM_THP this leads to a too small extradata
buffer being allocated, causing out-of-bounds writes.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agodxva2: fix surface selection when compiled with both d3d11va and dxva2
Hendrik Leppkes [Thu, 20 Oct 2016 17:31:34 +0000 (19:31 +0200)]
dxva2: fix surface selection when compiled with both d3d11va and dxva2

Fixes a regression introduced in 9b462a0b9

2 years agoavformat/flvenc: do not attempt to write duration and filesize when not seekable
Steven Liu [Wed, 12 Oct 2016 10:11:41 +0000 (18:11 +0800)]
avformat/flvenc: do not attempt to write duration and filesize when not seekable

Its impossible to update the filesize & duration values if seekback is not
possible as with live streams

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavc/sheervideo: Increase av_get_codec_tag_string() input buffer size.
Carl Eugen Hoyos [Tue, 18 Oct 2016 12:53:30 +0000 (14:53 +0200)]
lavc/sheervideo: Increase av_get_codec_tag_string() input buffer size.

A size of 32 is typically used.

2 years agoavfilter/firequalizer: use zero phase kernel
Muhammad Faiz [Wed, 19 Oct 2016 15:22:36 +0000 (22:22 +0700)]
avfilter/firequalizer: use zero phase kernel

this makes real-valued kernel

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
2 years agoavfilter/showcqt: add cscheme option
Muhammad Faiz [Wed, 19 Oct 2016 15:17:24 +0000 (22:17 +0700)]
avfilter/showcqt: add cscheme option

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
2 years agolavf/riff: Document the interleaving of VSM4 video.
Carl Eugen Hoyos [Wed, 19 Oct 2016 21:39:19 +0000 (23:39 +0200)]
lavf/riff: Document the interleaving of VSM4 video.

2 years agoavformat: prevent triggering request_probe assert in ff_read_packet
Andreas Cadhalpun [Wed, 19 Oct 2016 17:23:49 +0000 (19:23 +0200)]
avformat: prevent triggering request_probe assert in ff_read_packet

If probe_codec is called with pkt == NULL, it sets probe_packets to 0
and request_probe to -1.
However, request_probe can change when calling s->iformat->read_packet
and thus a probe_packets value of 0 doesn't guarantee a request_probe
value of -1.
In that case calling probe_codec again is necessary to prevent
triggering the assert.

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
2 years agoFATE: Add test for muxing discontinuous MP4 fragments with large timestamps
Derek Buitenhuis [Tue, 18 Oct 2016 15:30:08 +0000 (16:30 +0100)]
FATE: Add test for muxing discontinuous MP4 fragments with large timestamps

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2 years agoavformat/movenc: Skip duration check for discontinuous fragments
Michael Niedermayer [Tue, 18 Oct 2016 16:18:49 +0000 (18:18 +0200)]
avformat/movenc: Skip duration check for discontinuous fragments

Found-by: Daemon404
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agolavf/mov: support gopro hero moments udta tag
Jean Caillé [Tue, 18 Oct 2016 13:47:55 +0000 (15:47 +0200)]
lavf/mov: support gopro hero moments udta tag

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2 years agoavfilter/vf_scale_npp: select cubic and lanczos as alternative where super-sampling...
Sven C. Dack [Wed, 19 Oct 2016 10:51:21 +0000 (12:51 +0200)]
avfilter/vf_scale_npp: select cubic and lanczos as alternative where super-sampling is not supported

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
2 years agoavcodec/nvenc_hevc: Added missing option -temporal_aq
Sven C. Dack [Thu, 13 Oct 2016 17:06:09 +0000 (18:06 +0100)]
avcodec/nvenc_hevc: Added missing option -temporal_aq

The option is present in h264_nvenc, but was missing from hevc_nvenc.

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
2 years agoavcodec/nvenc: add test for Temporal AQ support
Sven C. Dack [Fri, 14 Oct 2016 15:02:54 +0000 (16:02 +0100)]
avcodec/nvenc: add test for Temporal AQ support

Adds a check to see if the hardware supports temporal aq.

Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
2 years agolavc/mediacodec: use more meaningful filenames
Matthieu Bouron [Wed, 12 Oct 2016 12:44:08 +0000 (14:44 +0200)]
lavc/mediacodec: use more meaningful filenames

Adds the following changes:
  * mediacodecdec.{c,h} -> mediacodecdec_common.{c,h}
  * mediacodecdec_h2645.c -> mediacodecdec.c

2 years agolavc: add mpeg4 mediacodec decoder
Matthieu Bouron [Tue, 4 Oct 2016 15:47:44 +0000 (17:47 +0200)]
lavc: add mpeg4 mediacodec decoder

2 years agolavc: add vp8/vp9 mediacodec decoders
Matthieu Bouron [Tue, 4 Oct 2016 14:02:59 +0000 (16:02 +0200)]
lavc: add vp8/vp9 mediacodec decoders

2 years agolavc/mediacodec_wrapper: do not discard codecs reporting they do not support any...
Matthieu Bouron [Fri, 7 Oct 2016 09:20:33 +0000 (11:20 +0200)]
lavc/mediacodec_wrapper: do not discard codecs reporting they do not support any profile

Depending on the device, some (VP8/VP9/...) decoders report that they do
not support any profiles.

2 years agoavfilter/firequalizer: add cubic_interpolate function on gain
Muhammad Faiz [Tue, 18 Oct 2016 15:55:39 +0000 (22:55 +0700)]
avfilter/firequalizer: add cubic_interpolate function on gain

smoother version of gain_interpolate

Signed-off-by: Muhammad Faiz <mfcc64@gmail.com>
2 years agoavformat/matroskadec: fix BitsPerChannel element's default value
James Almer [Wed, 19 Oct 2016 02:55:32 +0000 (23:55 -0300)]
avformat/matroskadec: fix BitsPerChannel element's default value

The element is currently ignored, so there's no effective functionality
change with this.

Signed-off-by: James Almer <jamrial@gmail.com>
2 years agoavformat/matroskadec: clarify some Colour element defaults using enum values
James Almer [Wed, 19 Oct 2016 02:40:09 +0000 (23:40 -0300)]
avformat/matroskadec: clarify some Colour element defaults using enum values

This way it's more clear what the default values refer to, as
Undetermined/Unspecified is 0 for some, 2 for others.

Signed-off-by: James Almer <jamrial@gmail.com>
2 years agoavformat/mov: pass the demuxer's AVFormatContext to avpriv_request_sample()
James Almer [Mon, 17 Oct 2016 00:23:17 +0000 (21:23 -0300)]
avformat/mov: pass the demuxer's AVFormatContext to avpriv_request_sample()

Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>