Mark Thompson [Fri, 15 Sep 2017 19:48:18 +0000 (20:48 +0100)]
kmsgrab: Fix DRM format definitions
All DRM formats are defined in terms of little-endian words, so RGB formats
like XRGB actually have the elements in the opposite order order in memory
to the order they are in the name.
This does not affect YUYV and similar YUV 4:2:2 formats, which are in the
expected order.
Hendrik Leppkes [Tue, 19 Sep 2017 08:45:41 +0000 (10:45 +0200)]
vp9: move VP9SharedContext back to the top of VP9Context
VP9SharedContext needs to be the first member so its properties can be
safely accessed from hardware accelerators, without the need to share
the full VP9Context.
avcodec/frame_thread_encoder: Fix AV_OPT_TYPE_STRING handling in avctx
This is the equivalent to what 7d317d4706b49d572a1eb5269438753be18362c7
did for the codec-specific options.
av_opt_copy has specific handling so it's fine that we already copied
the whole context before.
Thomas Mundt [Wed, 30 Aug 2017 01:37:18 +0000 (03:37 +0200)]
avfilter/interlace: prevent over-sharpening with the complex low-pass filter
The complex vertical low-pass filter slightly over-sharpens the picture. This becomes visible when several transcodings are cascaded and the error potentises, e.g. some generations of HD->SD SD->HD.
To prevent this behaviour the destination pixel must not exceed the source pixel when the average of the pixels above and below is less than the source pixel. And the other way around.
Tested and approved in a visual transcoding cascade test by video professionals.
SSIM/PSNR test with the first generation of an HD->SD file as a reference against the 6th generation(3 x SD->HD HD->SD):
Results without the patch:
SSIM Y:0.956508 (13.615881) U:0.991601 (20.757750) V:0.993004 (21.551382) All:0.974405 (15.918463)
PSNR y:31.838009 u:48.424280 v:48.962711 average:34.759466 min:31.699297 max:40.857847
Results with the patch:
SSIM Y:0.970051 (15.236232) U:0.991883 (20.905857) V:0.993174 (21.658049) All:0.981290 (17.279202)
PSNR y:34.412108 u:48.504454 v:48.969496 average:37.264644 min:34.310637 max:42.373392
Signed-off-by: Thomas Mundt <tmundt75@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Mark Thompson [Sat, 2 Sep 2017 19:32:27 +0000 (20:32 +0100)]
lavc: Add wrapped_avframe decoder
Intended for use with hardware frames for which rawvideo is not
sufficient. Requires the trusted packet flag to be set - decoding
fails if not to avoid security issues (the wrapped AVFrame can
contain pointers to arbitrary data).
Ronald S. Bultje [Mon, 11 Sep 2017 13:58:47 +0000 (09:58 -0400)]
frame_thread_encoder: make 'exit' member atomic.
Should fix the following tsan warning:
WARNING: ThreadSanitizer: data race (pid=19806)
Read of size 4 at 0x7b84000012f0 by thread T9:
#0 worker src/libavcodec/frame_thread_encoder.c:66 (ffmpeg+0x0000007f349e)
[..]
Previous write of size 4 at 0x7b84000012f0 by main thread (mutexes: write M1395):
#0 ff_frame_thread_encoder_free src/libavcodec/frame_thread_encoder.c:239 (ffmpeg+0x0000007f379e)
[..]
Fixes: runtime error: index 5 out of bounds for type 'int8_t const[5]'
Fixes:3175/clusterfuzz-testcase-minimized-4736774054084608
Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/shorten: Move buffer allocation and offset init to end of read_header()
They are time consuming operations, performing them after the other checks
improves the speed with damaged input dramatically.
Fixes: Timeout Fixes: 2928/clusterfuzz-testcase-4992812120539136 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/jpeg2000dsp: Fix multiple integer overflows in ict_int()
Fixes: runtime error: signed integer overflow: 22553 * -188962 cannot be represented in type 'int' Fixes: 3042/clusterfuzz-testcase-minimized-5174210131394560 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/clearvideo: Only output a frame if one is coded in the packet
Fixes: Timeout (183 ms instead of about 20 sec) Fixes: 3147/clusterfuzz-testcase-4870592182353920 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/hevcdsp_template: Fix undefined shift in put_hevc_pel_bi_w_pixels
Fixes: runtime error: left shift of negative value -95 Fixes: 3077/clusterfuzz-testcase-minimized-4684917524922368 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: runtime error: signed integer overflow: 11896 + 2147483646 cannot be represented in type 'int' Fixes: 3053/clusterfuzz-testcase-minimized-6355082062856192 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/dirac_vlc: limit res_bits in APPEND_RESIDUE()
Fixes: runtime error: left shift of 1073741838 by 1 places cannot be represented in type 'int32_t' (aka 'int') Fixes: 3279/clusterfuzz-testcase-minimized-4564805744590848 Suggested-by: <atomnuker> Reviewed-by: <atomnuker> Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Steven Liu [Sat, 9 Sep 2017 00:32:03 +0000 (08:32 +0800)]
avformat/dashdec: add dash demuxer base version
ffmpeg need a dash demuxer for demux the dash formats base on
https://github.com/samsamsam-iptvplayer/exteplayer3/blob/master/tmp/ffmpeg/patches/3.2.2/000001_add_dash_demux.patch
TODO:
1. support multi bitrate dash.
v2 fixed:
1. from autodetect to disabled
2. from camelCase code style to ffmpeg code style
3. from RepType to AVMediaType
4. fix variable typo
5. change time value from uint32_t to uint64_t
6. removed be used once API
7. change 'time(NULL)`, except it is not 2038-safe.' to av_gettime and av_timegm
8. merge complex free operation to free_fragment
9. use API from snprintf to av_asprintf
v3 fixed:
1. fix typo from --enabled-xml2 to --enable-xml2
v4 fixed:
1. from --enable-xml2 to --enable-libxml2
2. move system includes to top
3. remove nouse includes
4. rename enum name
5. add a trailing comma for the last entry enum
6. fix comment typo
7. add const to DASHContext class front
8. check sscanf if return arguments and give warning message when error
9. check validity before free seg->url and seg
10. check if the val is null, before use atoll
v5 fixed:
1. fix typo from mainifest to manifest
v6 fixed:
1. from realloc to av_realloc
2. from free to av_free
v7 fixed:
1. remove the -lxml2 from configure when require_pkg_config
v8 fixed:
1. fix replace filename template by av_asprintf secure problem
v9 modified:
1. make manifest parser clearly
v10 fixed:
1. fix function API name code style
2. remove redundant strreplace call
3. remove redundant memory operation and check return value from get_content_url()
4. add space between ) and {
5. remove no need to log the value for print
v11 fixed:
1. from atoll to strtoll Suggested-by: Michael Niedermayer <michael@niedermayer.cc>
v12 fixed:
1. remove strreplace and instead by av_strreplace Suggested-by: Nicolas George <george@nsup.org>
v13 fixed:
1. fix bug: cannot play:
http://dash.edgesuite.net/akamai/bbb_30fps/bbb_30fps.mpd Reported-by: Andy Furniss <adf.lists@gmail.com>
v14 fixed:
1. fix bug: TLS connection was non-properly terminated
2. fix bug: No trailing CRLF found in HTTP header Reported-by: Andy Furniss <adf.lists@gmail.com>
v15 fixed:
1. play youtube link: ffmpeg -i $(youtube-dl -J "https://www.youtube.com/watch?v=XmL19DOP_Ls" | jq -r ".requested_formats[0].manifest_url")
2. code refine for timeline living stream Reported-by: Ricardo Constantino <wiiaboo@gmail.com>
v16 fixed:
1. remove the snprintf and instead by get_segment_filename make safety
2. remove unnecessary loops
3. updated xmlStrcmp and xmlFree to av_* functions
4. merge code repeat into one function
5. add memory alloc faild check
6. update update_init_section and open_url
7. output safety error message when filename template not safe
Suggested-by : wm4 <nfxjfg@googlemail.com>
v17 fixed:
1. add memory alloc faild check
2. fix resource space error at free_representation
v18 fixed:
1. add condition of template format
v19 fixed:
1. fix typo of the option describe
v20 fixed:
1. add the c->base_url alloc check
2. make the DASHTmplId same to dashenc
v21 fixed:
1. remove get_repl_pattern_and_format and get_segment_filename
2. process use dashcomm APIs
v22 fixed:
1. modify the include "dashcomm.h" to include "dash.h"
2. use internal API from dash_fill_tmpl_params to ff_dash_fill_tmpl_params
Signed-off-by: Steven Liu <lq@onvideo.cn> Signed-off-by: samsamsam <samsamsam@o2.pl>
- there is no need for kCVImageBufferColorPrimaries_ITU_R_2020 checks,
it's done at runtime
- VideoToolbox presence is now checked with check_apple_framework()
- link to CoreServices is only done when videotoolbox is enabled
Refer to "checkasm: use perf API on Linux ARM*" commit for the
rationale.
The implementation is somehow duplicated with checkasm, but so is the
current usage of AV_READ_TIME(). Until these implementations and
heuristics are made consistent, I don't see a way of sharing that code.
Note: when using libavutil/timer.h, it is now important to include
before any other include due to the _GNU_SOURCE requirement.
On ARM platforms, accessing the PMU registers requires special user
access permissions. Since there is no other way to get accurate timers,
the current implementation of timers in FFmpeg rely on these registers.
Unfortunately, enabling user access to these registers on Linux is not
trivial, and generally involve compiling a random and unreliable github
kernel module, or patching somehow your kernel.
Such module is very unlikely to reach the upstream anytime soon. Quoting
Robin Murphin from ARM:
> Say you do give userspace direct access to the PMU; now run two or more
> programs at once that believe they can use the counters for their own
> "minimal-overhead" profiling. Have fun interpreting those results...
>
> And that's not even getting into the implications of scheduling across
> different CPUs, CPUidle, etc. where the PMU state is completely beyond
> userspace's control. In general, the plan to provide userspace with
> something which might happen to just about work in a few corner cases,
> but is meaningless, misleading or downright broken in all others, is to
> never do so.
As a result, the alternative is to use the Performance Monitoring Linux
API which makes use of these registers internally (assuming the PMU of
your ARM board is supported in the kernel, which is definitely not a
given...).
While the Linux API is obviously cross platform, it does have a
significant overhead which needs to be taken into account. As a result,
that mode is only weakly enabled on ARM platforms exclusively.
Note on the non flexibility of the implementation: the timers (native
FFmpeg vs Linux API) are selected at compilation time to prevent the
need of function calls, which would result in a negative impact on the
cycle counters.
Mark Wachsler [Thu, 7 Sep 2017 13:42:07 +0000 (09:42 -0400)]
libavcodec/h264_parse: don't use uninitialized value when chroma_format_idc==0
When parsing a monochrome file, chroma_log2_weight_denom was used without
being initialized, which could lead to a bogus error message being printed, e.g.
[h264 @ 0x61a000026480] chroma_log2_weight_denom 24576 is out of range
It also could led to warnings using AddressSanitizer.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
projects / ffmpeg / log