]> git.sesse.net Git - ffmpeg/log
ffmpeg
5 years agoavcodec/mpeg: Initialize quarter_sample parameter from previous thread.
Andriy Gelman [Thu, 13 Dec 2018 04:57:35 +0000 (23:57 -0500)]
avcodec/mpeg: Initialize quarter_sample parameter from previous thread.

Fixes #7410.
The value of sub-pixel precision for me/mc can change during an Intra frame. In multi-threaded decoding this change is not propagated to other frame threads causing decoding artifacts. This patch initializes the sub-pixel precision parameter from previous thread, which fixes the issue.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavformat/ffmetadec: do no limit size of tags to 1024
Paul B Mahol [Thu, 13 Dec 2018 12:06:30 +0000 (13:06 +0100)]
avformat/ffmetadec: do no limit size of tags to 1024

Use bprint API instead.

Fixes #4833.

5 years agoavcodec/ass_split: fix a memory leak defect when realloc fails
Fan Gang [Tue, 13 Feb 2018 07:38:59 +0000 (15:38 +0800)]
avcodec/ass_split: fix a memory leak defect when realloc fails

Fixes #7019.

5 years agodoc/metadata: fix error in timebase description
Paul B Mahol [Fri, 14 Dec 2018 11:56:13 +0000 (12:56 +0100)]
doc/metadata: fix error in timebase description

Fixes #7184.

5 years agoavformat/movenc: treat ALAC same as FLAC and write correct info
Paul B Mahol [Fri, 14 Dec 2018 11:10:25 +0000 (12:10 +0100)]
avformat/movenc: treat ALAC same as FLAC and write correct info

Fixes #7291.

5 years agoavformat/id3v2: use png header to get PNG signature
Paul B Mahol [Fri, 14 Dec 2018 10:34:46 +0000 (11:34 +0100)]
avformat/id3v2: use png header to get PNG signature

5 years agoavformat/flac_picture: try to guess PNG by actual picture data
Paul B Mahol [Thu, 13 Dec 2018 19:12:52 +0000 (20:12 +0100)]
avformat/flac_picture: try to guess PNG by actual picture data

Fixes #5028.

5 years agoavformat/nut: add support for yuva444/422p12 pixel format
Paul B Mahol [Fri, 14 Dec 2018 10:26:30 +0000 (11:26 +0100)]
avformat/nut: add support for yuva444/422p12 pixel format

5 years agoavformat/nutdec: fix obvious typo
Paul B Mahol [Thu, 13 Dec 2018 22:17:00 +0000 (23:17 +0100)]
avformat/nutdec: fix obvious typo

Fixes #6742.

5 years agoavformat/mxfdec: fix error check in macro
Paul B Mahol [Thu, 13 Dec 2018 22:51:02 +0000 (23:51 +0100)]
avformat/mxfdec: fix error check in macro

Fixes #6750.

5 years agoavcodec/xfacedec: fix order of operations
Paul B Mahol [Thu, 13 Dec 2018 22:44:39 +0000 (23:44 +0100)]
avcodec/xfacedec: fix order of operations

Fixes #6745.

5 years agobump micro after recent gif changes
Paul B Mahol [Thu, 13 Dec 2018 18:16:58 +0000 (19:16 +0100)]
bump micro after recent gif changes

5 years agoavcodec/gif: use avctx->frame_number
Paul B Mahol [Thu, 13 Dec 2018 18:15:45 +0000 (19:15 +0100)]
avcodec/gif: use avctx->frame_number

5 years agoavformat/gifdec: export duration, nb_frames and comment
Paul B Mahol [Thu, 13 Dec 2018 17:37:27 +0000 (18:37 +0100)]
avformat/gifdec: export duration, nb_frames and comment

5 years agoavcodec/codec_desc: extend gif description
Paul B Mahol [Wed, 12 Dec 2018 16:16:52 +0000 (17:16 +0100)]
avcodec/codec_desc: extend gif description

5 years agoavformat/gif: extend description
Paul B Mahol [Wed, 12 Dec 2018 16:16:30 +0000 (17:16 +0100)]
avformat/gif: extend description

5 years agoavcodec/gif: enable encoding single gif image per frame
Paul B Mahol [Wed, 12 Dec 2018 12:19:33 +0000 (13:19 +0100)]
avcodec/gif: enable encoding single gif image per frame

Unbreaks gif image2 muxer.

5 years agoavformat/gifdec: export sample aspect ratio
Paul B Mahol [Wed, 12 Dec 2018 10:12:32 +0000 (11:12 +0100)]
avformat/gifdec: export sample aspect ratio

5 years agoavcodec/gif: add support for alpha
Paul B Mahol [Tue, 11 Dec 2018 18:40:17 +0000 (19:40 +0100)]
avcodec/gif: add support for alpha

Based on patch by Bjorn Roche.

Fixes #6813.

5 years agoavcodec: rewrite gif muxing and encoding
Paul B Mahol [Tue, 11 Dec 2018 10:52:17 +0000 (11:52 +0100)]
avcodec: rewrite gif muxing and encoding

Now "-c copy" works.
Update FATE files.

Demuxer only split file into packets, no data is trimmed.
Encoder & muxer currently expect completely another format
where muxer writes stuff like disposal method which should
be really encoder job.
With this patch muxer only modifies delay between two packets.
Codec copy need to have same behavior between demuxer and
muxer to work correctly.

Fixes #6640.

5 years agoavformat/utils: add support for reading ID3 tags at start of wav
Paul B Mahol [Wed, 12 Dec 2018 18:24:21 +0000 (19:24 +0100)]
avformat/utils: add support for reading ID3 tags at start of wav

Fixes #4140.

5 years agoavformat/wavdec: add support for 'id3 ' chunk
Paul B Mahol [Wed, 12 Dec 2018 18:02:30 +0000 (19:02 +0100)]
avformat/wavdec: add support for 'id3 ' chunk

Fixes #5700.

5 years agolavfi/signalstats: Cast the return value of AV_RN16() to int.
Carl Eugen Hoyos [Tue, 11 Dec 2018 02:12:49 +0000 (03:12 +0100)]
lavfi/signalstats: Cast the return value of AV_RN16() to int.

It is unsigned at least on arm, fixes the following warnings:
libavfilter/vf_signalstats.c:833:21: warning: taking the absolute value of unsigned type 'unsigned int' has no effect
            dify += abs(yuv - AV_RN16(prev->data[0] + pw + i * 2));

5 years agoopusenc: fix infinite loop if flushing encoder upon init
Rostislav Pehlivanov [Wed, 12 Dec 2018 20:05:33 +0000 (20:05 +0000)]
opusenc: fix infinite loop if flushing encoder upon init

The issue is that the afq still has samples as on init it counts
the overlap used as a delay to adjust the PTS it generates, hence
we can't rely on it right after init.
So just check to see if any frames have been encoded. frame_number
can't be anything but 0 right after init and can only be set by lavc.

Signed-off-by: Rostislav Pehlivanov <atomnuker@gmail.com>
5 years agoavcodec/libdav1d: add support for 12bit streams
James Almer [Wed, 12 Dec 2018 03:06:05 +0000 (00:06 -0300)]
avcodec/libdav1d: add support for 12bit streams

Signed-off-by: James Almer <jamrial@gmail.com>
5 years agoswscale/output: VSX-optimize nbps yuv2plane1
Lauri Kasanen [Tue, 27 Nov 2018 13:26:00 +0000 (15:26 +0200)]
swscale/output: VSX-optimize nbps yuv2plane1

./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt yuv420p9le \
-f null -vframes 100 -v error -nostats -

Speedups:
yuv2plane1_9BE_vsx 11.2042
yuv2plane1_9LE_vsx 11.156
yuv2plane1_10BE_vsx 9.89428
yuv2plane1_10LE_vsx 10.3637
yuv2plane1_12BE_vsx 9.71923
yuv2plane1_12LE_vsx 11.0404
yuv2plane1_14BE_vsx 10.1763
yuv2plane1_14LE_vsx 11.2728

Fate passes, each format tested with an image to video conversion.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agodoc/indevs: document libdc1394 options
Lou Logan [Wed, 12 Dec 2018 00:24:30 +0000 (15:24 -0900)]
doc/indevs: document libdc1394 options

Signed-off-by: Lou Logan <lou@lrcd.com>
5 years agotests/api-flac-test: Rename NUMBER_OF_FRAMES as NUMBER_OF_AUDIO_FRAMES.
Carl Eugen Hoyos [Mon, 3 Dec 2018 00:14:24 +0000 (01:14 +0100)]
tests/api-flac-test: Rename NUMBER_OF_FRAMES as NUMBER_OF_AUDIO_FRAMES.

In system header /usr/include/sys/mstsave.h, aix defines NUMBER_OF_FRAMES,
causing redefinition warnings.

5 years agoavfilter/vf_paletteuse: fix flags for alpha_threshold option
Paul B Mahol [Tue, 11 Dec 2018 16:55:05 +0000 (17:55 +0100)]
avfilter/vf_paletteuse: fix flags for alpha_threshold option

5 years agoconfigure: require the first stable version of libdav1d
James Almer [Tue, 11 Dec 2018 15:42:22 +0000 (12:42 -0300)]
configure: require the first stable version of libdav1d

Signed-off-by: James Almer <jamrial@gmail.com>
5 years agofate: add prog suffix in fate-mov-mp4-with-mov-in24-ver
Gyan Doshi [Mon, 10 Dec 2018 15:14:33 +0000 (20:44 +0530)]
fate: add prog suffix in fate-mov-mp4-with-mov-in24-ver

Prevent test failure with builds that set a suffix and/or
extension for the ffprobe binary

5 years agoos_support: define socket shutdown SHUT_xxx macros if they are not defined
Peter Ross [Sat, 24 Nov 2018 12:26:13 +0000 (23:26 +1100)]
os_support: define socket shutdown SHUT_xxx macros if they are not defined

this section has been moved into the CONFIG_NETWORK block, since it only
affects network enabled builds.

sys/socket.h (with WIN32 guard) is needed to check if the SHUT_xxx macro exists.

5 years agoavformat/dashenc: Used the movenc option skip_sidx instead of global_sidx
kjeyapal@akamai.com [Wed, 5 Dec 2018 06:18:40 +0000 (11:48 +0530)]
avformat/dashenc: Used the movenc option skip_sidx instead of global_sidx

Anyways the intended behaviour was to disable SIDX atom.

5 years agoavformat/movenc: Added an option to disable SIDX atom
kjeyapal@akamai.com [Thu, 6 Dec 2018 07:28:01 +0000 (12:58 +0530)]
avformat/movenc: Added an option to disable SIDX atom

5 years agodoc/indevs: fix hdvbuffer typo in iec61883 example
Lou Logan [Tue, 11 Dec 2018 00:17:57 +0000 (15:17 -0900)]
doc/indevs: fix hdvbuffer typo in iec61883 example

Option name is dvbuffer.

Signed-off-by: Lou Logan <lou@lrcd.com>
5 years agolavc/tiff: Fix leak of yuv_line in TiffContext.
Carl Hetherington [Mon, 10 Dec 2018 23:52:16 +0000 (23:52 +0000)]
lavc/tiff: Fix leak of yuv_line in TiffContext.

Signed-off-by: Carl Hetherington <cth@carlh.net>
5 years agolavd/v4l2: Use "int request" as second parameter for ioctl() on Android.
Carl Eugen Hoyos [Thu, 6 Dec 2018 22:34:54 +0000 (23:34 +0100)]
lavd/v4l2: Use "int request" as second parameter for ioctl() on Android.

Fixes build with new Android toolchain.

5 years agolavf/matroskadec: Do not use strncat() to limit copying a one-char constant.
Carl Eugen Hoyos [Mon, 10 Dec 2018 00:55:15 +0000 (01:55 +0100)]
lavf/matroskadec: Do not use strncat() to limit copying a one-char constant.

Instead add the character to the snprintf above as suggested by Mark.

Silences a warning:
libavformat/matroskadec.c: In function 'webm_dash_manifest_cues':
libavformat/matroskadec.c:3947:13: warning: 'strncat' specified bound 1 equals source length [-Wstringop-overflow=]
             strncat(buf, ",", 1);
             ^~~~~~~~~~~~~~~~~~~~

5 years agoavcodec/xpmdec: define constants
Paul B Mahol [Mon, 10 Dec 2018 20:38:08 +0000 (21:38 +0100)]
avcodec/xpmdec: define constants

5 years agoavcodec/xpmdec: fix small artifacts
Paul B Mahol [Mon, 10 Dec 2018 20:31:18 +0000 (21:31 +0100)]
avcodec/xpmdec: fix small artifacts

The max size of uint8_t subset is 223 not 220.

5 years agoavformat/gif: simplify signature writing
Paul B Mahol [Mon, 10 Dec 2018 17:36:11 +0000 (18:36 +0100)]
avformat/gif: simplify signature writing

5 years agolavc/xpmdec: Allow more colours per character.
Carl Eugen Hoyos [Thu, 6 Dec 2018 18:09:56 +0000 (19:09 +0100)]
lavc/xpmdec: Allow more colours per character.

Fixes ticket #6234.

5 years agoavcodec/gifdec: truncate too big width/height for invalid gif files
Paul B Mahol [Mon, 10 Dec 2018 12:32:10 +0000 (13:32 +0100)]
avcodec/gifdec: truncate too big width/height for invalid gif files

Fixes #6874.

5 years agoavformat/mxfenc: calculate and store DAR from user SAR
Paul B Mahol [Mon, 10 Dec 2018 09:15:27 +0000 (10:15 +0100)]
avformat/mxfenc: calculate and store DAR from user SAR

5 years agoavformat/mxfenc: allow muxing prores
Paul B Mahol [Wed, 5 Dec 2018 17:10:37 +0000 (18:10 +0100)]
avformat/mxfenc: allow muxing prores

5 years agoavcodec/proresdec2: allow changing resolution
Paul B Mahol [Wed, 5 Dec 2018 17:50:33 +0000 (18:50 +0100)]
avcodec/proresdec2: allow changing resolution

5 years agoavformat: add gif pipe demuxer
Paul B Mahol [Sat, 8 Dec 2018 12:06:31 +0000 (13:06 +0100)]
avformat: add gif pipe demuxer

5 years agoavcodec: add gif parser
Paul B Mahol [Sat, 8 Dec 2018 11:31:19 +0000 (12:31 +0100)]
avcodec: add gif parser

5 years agolavfi/vf_scale_vaapi: add scaling mode setting support.
Jun Zhao [Wed, 12 Sep 2018 12:14:58 +0000 (20:14 +0800)]
lavfi/vf_scale_vaapi: add scaling mode setting support.

before this change, scale_vaapi hard coding the scaling mode, add a
new option "mode" to setting the scaling mode, it can be use to change
scaling algorithm for performance/quality trade off.

Note: VA_FILTER_SCALING_NL_ANAMORPHIC(nolinear ananorphic scaling
algorithm) doesn't support by i965/iHD driver now, add this just for
completeness.

Reviewed-by: Mark Thompson <sw@jkqxz.net>
Signed-off-by: Jun Zhao <mypopydev@gmail.com>
5 years agoavformat/hlsenc : Added an option to ignore IO errors
Karthick J [Mon, 10 Dec 2018 06:19:23 +0000 (14:19 +0800)]
avformat/hlsenc : Added an option to ignore IO errors

Useful for long duration runs with network output

Reviewed-by: Steven Liu <lq@chinaffmpeg.org>
5 years agoavformat/hlsenc: Handled error from ff_http_do_new_request() function
Karthick J [Mon, 10 Dec 2018 06:17:55 +0000 (14:17 +0800)]
avformat/hlsenc: Handled error from ff_http_do_new_request() function

This patch fixes the segmentation fault issues due to
unhandled errors from ff_http_do_new_request function.

Reviewed-by: Steven Liu <lq@chinaffmpeg.org>
5 years agolavc/qsvenc: replace assert with error return
Linjie Fu [Sun, 9 Dec 2018 13:30:38 +0000 (21:30 +0800)]
lavc/qsvenc: replace assert with error return

Fix the (m)jpeg encoding regression issue as decription in tikect #7593,
due to bs->FrameType is not set in such case in
MSDK (https://github.com/Intel-Media-SDK/MediaSDK/issues/970).
(And assert on a value coming from an external library is not proper.)

Add default type check for bs->FrameType, and return invalid data error in function
ff_qsv_encode to avoid using uninitialized value.

Signed-off-by: Linjie Fu <linjie.fu@intel.com>
Signed-off-by: Zhong Li <zhong.li@intel.com>
5 years agolavc/cbs_vp9: Make variable prob unsigned.
Carl Eugen Hoyos [Mon, 10 Dec 2018 01:18:56 +0000 (02:18 +0100)]
lavc/cbs_vp9: Make variable prob unsigned.

Silences a warning with clang:
libavcodec/cbs_vp9_syntax_template.c:220:17: warning: implicit conversion from 'int' to 'int8_t' (aka 'signed char')
      changes value from 255 to -1

5 years agolavc/decode: Initialize return value for get_format() failure.
Carl Eugen Hoyos [Thu, 6 Dec 2018 22:17:13 +0000 (23:17 +0100)]
lavc/decode: Initialize return value for get_format() failure.

Silences a warning:
libavcodec/decode.c:1378:13: warning: variable 'ret' is used uninitialized whenever 'if' condition is true

5 years agoavcodec: bump micro version after d12d4d4515
James Almer [Sun, 9 Dec 2018 22:28:37 +0000 (19:28 -0300)]
avcodec: bump micro version after d12d4d4515

Signed-off-by: James Almer <jamrial@gmail.com>
5 years agoavformat/mxfdec: fix indentation
Paul B Mahol [Sun, 9 Dec 2018 19:10:33 +0000 (20:10 +0100)]
avformat/mxfdec: fix indentation

5 years agoqsvdec: Fix running with assert_level > 0
Mark Thompson [Sun, 11 Nov 2018 15:32:15 +0000 (15:32 +0000)]
qsvdec: Fix running with assert_level > 0

Assertion avctx->codec_id != AV_CODEC_ID_NONE failed at src/libavcodec/parser.c:128

The setting on the internal AVCodecContext used for parsing only is
otherwise irrelevant, so just set it to avoid the assert.

5 years agoavcodec/libaomenc: add row-mt option
James Almer [Sat, 8 Dec 2018 23:14:12 +0000 (20:14 -0300)]
avcodec/libaomenc: add row-mt option

Default to disable, same as aomenc.

Fixes ticket #7598

Signed-off-by: James Almer <jamrial@gmail.com>
5 years agoavfilter/vf_showinfo: allow checksums calculation to be disabled
Paul B Mahol [Tue, 4 Dec 2018 20:39:32 +0000 (21:39 +0100)]
avfilter/vf_showinfo: allow checksums calculation to be disabled

Fixes #6987.

5 years agolavfi/buffersink: fix abuffersink flag setting issue
Jun Zhao [Sun, 9 Dec 2018 14:30:41 +0000 (22:30 +0800)]
lavfi/buffersink: fix abuffersink flag setting issue

abuffersink need to setting AV_OPT_FLAG_AUDIO_PARAM flag.

Reviewed-by: Nicolas George <george@nsup.org>
Signed-off-by: Jun Zhao <mypopydev@gmail.com>
5 years agoavformat/dhav: export average frame rate too
Paul B Mahol [Sun, 9 Dec 2018 09:49:00 +0000 (10:49 +0100)]
avformat/dhav: export average frame rate too

5 years agoavcodec/jpeg2000dec: Skip DWT if nothing is coded
Michael Niedermayer [Sat, 1 Dec 2018 23:35:49 +0000 (00:35 +0100)]
avcodec/jpeg2000dec: Skip DWT if nothing is coded

Improves speed in uncommon case

Fixes: Timeout
Fixes: 10964/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_JPEG2000_fuzzer-5132066034286592
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavcodec/dxv: Check that there is enough data to decompress
Michael Niedermayer [Sat, 1 Dec 2018 20:41:01 +0000 (21:41 +0100)]
avcodec/dxv: Check that there is enough data to decompress

Fixes: Timeout
Fixes: 10979/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_DXV_fuzzer-6178582203203584
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agodoc/codecs: Do not mention a removed flags.
Carl Eugen Hoyos [Sat, 8 Dec 2018 10:29:19 +0000 (11:29 +0100)]
doc/codecs: Do not mention a removed flags.

5 years agodoc/faq: Fix flags for possible video quality improvements.
Carl Eugen Hoyos [Sat, 8 Dec 2018 10:28:01 +0000 (11:28 +0100)]
doc/faq: Fix flags for possible video quality improvements.

Reported-by: Ulf Zibis
5 years agodoc: vapoursynth
Gyan Doshi [Sat, 8 Dec 2018 10:03:57 +0000 (15:33 +0530)]
doc: vapoursynth

Add entry in external library list and demuxers.

5 years agoFix bug in mpegts muxer which affects KLV async stream generation.
Artyom Lebedev [Fri, 7 Dec 2018 09:48:28 +0000 (11:48 +0200)]
Fix bug in mpegts muxer which affects KLV async stream generation.

Fixes ticket #7597.

Signed-off-by: Peter Ross <pross@xvid.org>
5 years agoavformat/mxfdec: set codec_tag for prores codec
Paul B Mahol [Wed, 5 Dec 2018 17:49:41 +0000 (18:49 +0100)]
avformat/mxfdec: set codec_tag for prores codec

5 years agoavcodec/proresdec2: add new line at end of debug message
Paul B Mahol [Fri, 7 Dec 2018 19:05:37 +0000 (20:05 +0100)]
avcodec/proresdec2: add new line at end of debug message

5 years agoavcodec/dpx: fix spotted code style issues
Paul B Mahol [Thu, 6 Dec 2018 19:33:09 +0000 (20:33 +0100)]
avcodec/dpx: fix spotted code style issues

5 years agoavcodec/dpx: add support for special encoding
Paul B Mahol [Thu, 6 Dec 2018 19:31:51 +0000 (20:31 +0100)]
avcodec/dpx: add support for special encoding

5 years agoavcodec/dpx: improve decoding of 10 bit gray images
Paul B Mahol [Thu, 6 Dec 2018 19:28:22 +0000 (20:28 +0100)]
avcodec/dpx: improve decoding of 10 bit gray images

5 years agoavcodec/dpx: add support for another gray10 variant
Paul B Mahol [Thu, 6 Dec 2018 19:25:33 +0000 (20:25 +0100)]
avcodec/dpx: add support for another gray10 variant

5 years agoavcodec/dpx: parse input device name
Paul B Mahol [Thu, 6 Dec 2018 19:23:37 +0000 (20:23 +0100)]
avcodec/dpx: parse input device name

5 years agoavcodec/dpx: parse image creator
Paul B Mahol [Thu, 6 Dec 2018 19:19:48 +0000 (20:19 +0100)]
avcodec/dpx: parse image creator

5 years agoavcodec/dpx: check version of format header too
Paul B Mahol [Wed, 5 Dec 2018 13:25:17 +0000 (14:25 +0100)]
avcodec/dpx: check version of format header too

5 years agodoc/general: rearrange libraries in alphabetical order
Gyan Doshi [Fri, 7 Dec 2018 12:26:00 +0000 (17:56 +0530)]
doc/general: rearrange libraries in alphabetical order

5 years agolavc/qsvenc: add an option to set h264 pps for every frame
Zhong Li [Thu, 25 Oct 2018 11:14:16 +0000 (19:14 +0800)]
lavc/qsvenc: add an option to set h264 pps for every frame

RepeatPPS is enabled by default in mfx. It is not necessary mostly and
wasting encoding bits.
Add an option to control it and disable it by default.

Reviewed-by: Mark Thompson <sw@jkqxz.net>
Signed-off-by: Zhong Li <zhong.li@intel.com>
5 years agolavc/qsvenc: dump log for frame rate
Zhong Li [Mon, 5 Nov 2018 08:02:57 +0000 (16:02 +0800)]
lavc/qsvenc: dump log for frame rate

Signed-off-by: Zhong Li <zhong.li@intel.com>
5 years agolavc/qsvenc: make hevc alignment same as h264 for 1.19+ version
Zhong Li [Thu, 25 Oct 2018 10:44:37 +0000 (18:44 +0800)]
lavc/qsvenc: make hevc alignment same as h264 for 1.19+ version

It is to clean up to the code and To-Do list.

Signed-off-by: Zhong Li <zhong.li@intel.com>
5 years agolavc/qsvenc: remove redundant code
Zhong Li [Thu, 25 Oct 2018 10:23:00 +0000 (18:23 +0800)]
lavc/qsvenc: remove redundant code

Signed-off-by: Zhong Li <zhong.li@intel.com>
5 years agolavc/proresdec.h: Include a missing header
Carl Eugen Hoyos [Thu, 6 Dec 2018 19:14:09 +0000 (20:14 +0100)]
lavc/proresdec.h: Include a missing header

Fixes make checkheaders

5 years agoavcodec/ppc/hevcdsp: Fix build failures with powerpc-linux-gnu-gcc-4.8 with --disable...
Michael Niedermayer [Tue, 4 Dec 2018 15:29:40 +0000 (16:29 +0100)]
avcodec/ppc/hevcdsp: Fix build failures with powerpc-linux-gnu-gcc-4.8 with --disable-optimizations

The affected functions could also be changed into macros, this is the
smaller change to fix it though. And avoids (probably) less readable macros
The extra code should be optimized out when optimizations are done as all values
are known at build after inlining.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agodoc: libmodplug
Gyan Doshi [Wed, 5 Dec 2018 14:02:05 +0000 (19:32 +0530)]
doc: libmodplug

Add entry in external library list and demuxers.

5 years agoavcodec/msvideo1: Check for too small dimensions
Michael Niedermayer [Sat, 1 Dec 2018 21:16:19 +0000 (22:16 +0100)]
avcodec/msvideo1: Check for too small dimensions

Such low resolution would result in empty output as a minimum of 4x4 is needed
We could also check for multiple of 4 dimensions but that is not needed

Fixes: Timeout
Fixes: 11191/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_MSVIDEO1_fuzzer-5739529588178944
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavcodec/wmv2dec: Skip I frame if its smaller than 1/8 of the minimal size
Michael Niedermayer [Tue, 27 Nov 2018 22:37:03 +0000 (23:37 +0100)]
avcodec/wmv2dec: Skip I frame if its smaller than 1/8 of the minimal size

Frames that small are not valid and of limited use for error concealment, while
being very computationally intensive to process.

Fixes: Timeout
Fixes: 11168/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_WMV2_fuzzer-5733782032744448
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavcodec/msmpeg4dec: Skip frame if its smaller than 1/8 of the minimal size
Michael Niedermayer [Thu, 29 Nov 2018 01:32:10 +0000 (02:32 +0100)]
avcodec/msmpeg4dec: Skip frame if its smaller than 1/8 of the minimal size

Frames that small are not valid and of limited use for error concealment, while
being very computationally intensive to process.

Fixes: Timeout
Fixes: 11318/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_MSMPEG4V1_fuzzer-5710884555456512
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavcodec/dpx: add support for 10bit gray
Paul B Mahol [Tue, 4 Dec 2018 20:04:05 +0000 (21:04 +0100)]
avcodec/dpx: add support for 10bit gray

5 years agodoc: remove licensing claims for chromaprint and libgme
Gyan Doshi [Tue, 4 Dec 2018 14:43:57 +0000 (20:13 +0530)]
doc: remove licensing claims for chromaprint and libgme

5 years agoavcodec/prores_aw : add 4444 xq support
Martin Vignali [Sun, 2 Dec 2018 20:40:56 +0000 (21:40 +0100)]
avcodec/prores_aw : add 4444 xq support

5 years agoavcodec/prores_aw : only set color prim, trc, space values if supported
Martin Vignali [Sun, 2 Dec 2018 20:39:18 +0000 (21:39 +0100)]
avcodec/prores_aw : only set color prim, trc, space values if supported

set to unspecified if frame have another value

5 years agoavcodec/prores_aw : add vendor option
Martin Vignali [Sat, 24 Nov 2018 21:46:59 +0000 (22:46 +0100)]
avcodec/prores_aw : add vendor option

code come from prores_ks

5 years agoavcodec/utils : add ff_int_from_list_or_default func
Martin Vignali [Sun, 2 Dec 2018 20:36:24 +0000 (21:36 +0100)]
avcodec/utils : add ff_int_from_list_or_default func

to check valid value, or return default_value

5 years agodoc: libgme
Gyan Doshi [Tue, 4 Dec 2018 12:35:53 +0000 (18:05 +0530)]
doc: libgme

Add entry in external library list and add details to demuxer entry.

5 years agodoc: chromaprint
Gyan Doshi [Tue, 4 Dec 2018 11:04:38 +0000 (16:34 +0530)]
doc: chromaprint

Add entry in external library list and add details to muxer entry.

5 years agolavf/dashenc: Write media trailers when DASH trailer is written.
Andrey Semashev [Mon, 3 Dec 2018 11:18:31 +0000 (14:18 +0300)]
lavf/dashenc: Write media trailers when DASH trailer is written.

This commit ensures that all (potentially, long) filesystem activity is
performed when the user calls av_write_trailer on the DASH libavformat
context, not when freeing the context. Also, this defers media segment
deletion until after the media trailers are written.

5 years agoswscale/ppc: Move VSX-using code to its own file
Lauri Kasanen [Mon, 26 Nov 2018 12:24:15 +0000 (14:24 +0200)]
swscale/ppc: Move VSX-using code to its own file

Passes fate on LE (with "lavc/jrevdct: Avoid an aliasing violation" applied).

Signed-off-by: Lauri Kasanen <cand@gmx.com>
Tested-by: Michael Kostylev on BE
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
5 years agoavcodec/r210dec: fix r10x decoding
Paul B Mahol [Mon, 3 Dec 2018 22:34:05 +0000 (23:34 +0100)]
avcodec/r210dec: fix r10x decoding

5 years agoavcodec/dnxhddec: use init_get_bits8()
Paul B Mahol [Mon, 3 Dec 2018 18:44:49 +0000 (19:44 +0100)]
avcodec/dnxhddec: use init_get_bits8()

5 years agolavc: Bump version for r210 pix_fmt change.
Carl Eugen Hoyos [Mon, 3 Dec 2018 18:10:42 +0000 (19:10 +0100)]
lavc: Bump version for r210 pix_fmt change.