mirror/ffmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2026-01-14 12:33:28 +00:00
Code Issues Packages Projects Releases Wiki Activity
122,458 Commits 38 Branches 415 Tags
master
Commit Graph

122458 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Michael Niedermayer
805931dfc8 avfilter/scale_eval: Avoid undefined behavior with double to int cast 
We use INT32_MIN/MAX so as to ensure we dont have to deal with 64bit width or height
on a int is int64 system. int64 width would overflow in a system where we assume the product of 2
values fit in int64

Fixes: #YWH-PGM40646-14
Found-by: An0n99X
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-13 22:25:53 +00:00
Michael Niedermayer
5bbc8f828e avfilter/scale_eval: Use 64bit for factor_w/h 
Avoids truncation and overflows

Fixes: #YWH-PGM40646-14
Found-by: An0n99X
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-13 22:25:53 +00:00
Michael Niedermayer
23c8123d5b avformat/http: Check white and blacklists on redirects 
Noticed while fixing the previous

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-13 21:56:14 +00:00
Michael Niedermayer
b9227d49ea avformat/http: Check that the protocol of redirects is http or https 
Fixes: #YWH-PGM40646-10

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-13 21:56:14 +00:00
James Almer
80a3ba7d79 configure: add missing vulkan_1_4 check 
Removed in f2affdfafb by mistake.
Also, fix the logic in probe_glslc() so it doesn't return false when debug is disabled.

Signed-off-by: James Almer <jamrial@gmail.com>
 2026-01-13 13:28:23 -03:00
James Almer
f897bcd122 configure: fix glslc checks and add missing dependencies to relevant modules 
Signed-off-by: James Almer <jamrial@gmail.com>
 2026-01-12 18:20:01 -03:00
Lynne
ab66bc5776 Makefile: add missing vpath for glsl/spv 
Similar to CUDA. Fixes building on some systems.
 2026-01-12 19:18:36 +01:00
Michael Niedermayer
7e9f99ef41 doc/mailing-list-faq.texi: Fix typo in link anchor 
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-12 18:00:14 +00:00
Michael Niedermayer
aa0a3e6223 .forgejo/CODEOWNERS: get notified for vf_find_rect 
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-12 17:50:30 +00:00
Lynne
80229c161b scale_vulkan: use compile-time SPIR-V generation for debayering 2026-01-12 17:28:43 +01:00
Lynne
f2a55af9a4 vulkan_dpx: switch to compile-time SPIR-V generation 2026-01-12 17:28:43 +01:00
Lynne
0f4667fc11 vulkan_prores_raw: clean up and optimize 2026-01-12 17:28:42 +01:00
averne
771b58c0b6 vulkan/prores: fix buffer alignments 2026-01-12 17:28:42 +01:00
Lynne
23ab1b1a66 vulkan/dct: embed DCT scaling values during SPIR-V generation 
Instead of relying on rounded off values, use specialization constants
to bake the DCT values into the shader when its compiled.
 2026-01-12 17:28:42 +01:00
Lynne
e27b510da8 vulkan_prores: generate SPIR-V at compile-time 2026-01-12 17:28:42 +01:00
Lynne
026e94e339 vulkan_prores_raw: use compile-time SPIR-V generation 2026-01-12 17:28:42 +01:00
Lynne
f3b0ca4f2c avgblur_vulkan: generate SPIR-V during compilation 2026-01-12 17:28:42 +01:00
Lynne
a6391168f2 bwdif_vulkan: use compile-time SPIR-V 2026-01-12 17:28:42 +01:00
Lynne
e32cb94258 hwcontext_vulkan: enable long vector extension 
This enables us to do some neat tricks.
 2026-01-12 17:28:41 +01:00
Lynne
95a23820f7 hwcontext_vulkan: enable non-uniform features we were already using 2026-01-12 17:28:41 +01:00
Lynne
cf0ce1b7e4 hwcontext_vulkan: enable the VK_EXT_shader_replicated_composites extension 
Its required to use certain features in SPIR-V that glslang uses.
 2026-01-12 17:28:41 +01:00
Lynne
dd69ebc9a8 vulkan_filter: don't use lg_size[2] as the number of workgroup.z 
This was put there because who knows why, but it's wrong, this
ends up having double dispatches in the z direction if non-1.
 2026-01-12 17:28:41 +01:00
Lynne
e844b43776 vulkan: support shader compression 2026-01-12 17:28:41 +01:00
Lynne
7ce22b085e cuda/load_helper: move zlib decompression into a separate file 
Allows it to be reused for Vulkan
 2026-01-12 17:28:41 +01:00
Lynne
35235c762c configure: add zlib to suggests for shader_compression 2026-01-12 17:28:40 +01:00
Lynne
5996a9ad3d configure: rename PTX_COMPRESSION to SHADER_COMPRESSION 
Its useful for GLSL and Metal as well.
 2026-01-12 17:28:40 +01:00
Lynne
540c4df5c7 vulkan: add support for precompiled shaders 2026-01-12 17:28:40 +01:00
Lynne
f2affdfafb configure/make: support compile-time SPIR-V generation 2026-01-12 17:28:40 +01:00
Lynne
17f9a64860 configure: rename spirv_compiler to spirv_library 
More accurate.
 2026-01-12 17:28:40 +01:00
Lynne
f11aa29df8 configure: remove unused X86ASM dependency flag variables 2026-01-12 17:28:40 +01:00
Lynne
40edf7d75d vulkan: switch to static allocation for temporary descriptor data 
Simplifies management, and the hardware is limited to 4 descriptor sets
and whatever bindings.
 2026-01-12 17:28:35 +01:00
Soham
ecfabc25c2 avcodec/get_bits: mark unused VLC parameters 
This commit marks unused parameters in get_vlc_multi() to fix
-Wunused-parameter warnings when building with -Wall/-Wextra.

No functional changes intended.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-12 14:03:25 +01:00
stevxiao
21a3e44fbe avcodec/d3d12va_encode: add Region of Interest (ROI) support 
This commit implements ROI (Region of Interest) encoding support for D3D12VA hardware encoders, enabling spatially-adaptive quality control for H.264, HEVC, and AV1 encoders.

Query for `D3D12_VIDEO_ENCODER_RATE_CONTROL_FLAG_ENABLE_DELTA_QP` support during initialization to check whether the hardware support delta QP. If delta QP is supported, then process `AV_FRAME_DATA_REGIONS_OF_INTEREST` side data and generate delta QP maps for each frame.

Sample command line:
ffmpeg.exe -hwaccel d3d12va -hwaccel_output_format d3d12 -i input.mp4 -vf addroi=x=480:y=270:w=960:h=540:qoffset=-1/5 -c:v hevc_d3d12va output.mp4
 2026-01-12 02:45:48 +00:00
Michael Niedermayer
f99df7dbb3 avfilter/vf_find_rect: Fix handling odd sized images 
Fixes: out of array read
Fixes: #YWH-PGM40646-17

Found-by: An0n99X
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
 2026-01-12 02:05:11 +00:00
Raja Rathour
06d799ce1a avfilter/dnn_backend_torch: handle lifecycle of worker thread 
Initialize async resources in dnn_load_model_th and ensure proper cleanup and thread joining in dnn_free_model_th.

Signed-off-by: Raja Rathour <imraja729@gmail.com>
Reviewed-by: Wenbin Chen <wenbin.chen@intel.com>
Reviewed-by: Guo Yejun <yejun.guo@intel.com>
 2026-01-12 08:53:42 +08:00
Raja Rathour
5e8ad5c39b avfilter/dnn_backend_torch: implement async execution logic 
Update execute_model_th to support asynchronous execution by pushing tasks to the pending queue and notifying the worker thread.

Signed-off-by: Raja Rathour <imraja729@gmail.com>
Reviewed-by: Wenbin Chen <wenbin.chen@intel.com>
Reviewed-by: Guo Yejun <yejun.guo@intel.com>
 2026-01-12 08:53:42 +08:00
Raja Rathour
910270b179 avfilter/dnn_backend_torch: add async infrastructure 
Add worker thread, mutex, and condition variable to THModel.

Signed-off-by: Raja Rathour <imraja729@gmail.com>
Reviewed-by: Wenbin Chen <wenbin.chen@intel.com>
Reviewed-by: Guo Yejun <yejun.guo@intel.com>
 2026-01-12 08:53:42 +08:00
James Almer
cd03e0a001 libavformat/evcdec.c: add a check to ensure evc_frame_merge_bsf is present 
While the inclusion of the bsf is forced by a configure check, it's better to not
just have the code here assume filter will never be NULL.

Fixes Coverity issue CID 1680305.

Signed-off-by: James Almer <jamrial@gmail.com>
 2026-01-12 01:43:21 +00:00
Marton Balint
96e8f3b8cc avformat/vorbiscomment: fix writing huge chapter numbers and time offets to vorbiscomment 
Previous code truncated chapter numbers more than 999 and and time offsets more
than 99 hours.

The Vorbis comment format only allows 1000 chapters, so cap the number of
written chapters there.

This also fixes the GCC warnings for truncated strings.

Signed-off-by: Marton Balint <cus@passwd.hu>
 2026-01-12 00:47:20 +01:00
Marton Balint
b2e077ad69 avformat/vorbiscomment: use null buf to calculate vorbis comment length 
Also check possible failures when calculating length, and change return type to
int as bigger return values are no longer possible.
 2026-01-12 00:47:20 +01:00
Marton Balint
c80d3b35ae avformat/aviobuf: return error for ffio_close_null_buf() if written bytes exceed INT_MAX 
Also check return value where it is used.

Signed-off-by: Marton Balint <cus@passwd.hu>
 2026-01-12 00:47:20 +01:00
James Almer
af136db1c3 avfilter/vf_lcevc: attach a reference to the source frame to each passed in base picture 
And free them once they are guaranteed to be no longer needed, instead of freeing them
when returned with an enhanced output.

Signed-off-by: James Almer <jamrial@gmail.com>
 2026-01-11 17:59:15 -03:00
James Almer
188521c7ad avcodec/lcevc: attach a reference to the source frame to each passed in base picture 
This way we can ensure a frame reference will always exists for as long as the
external library needs the base picture.

Signed-off-by: James Almer <jamrial@gmail.com>
 2026-01-11 17:59:15 -03:00
Jack Lau
24213d5db3 MAINTAINERS: add myself as whip maintainer 
Signed-off-by: Jack Lau <jacklau1222gm@gmail.com>
 2026-01-11 04:24:55 +00:00
Jack Lau
b0fe52ba4e avformat/tls_gnutls: add gnutls_pull_timeout() for dtls 
See https://gnutls.org/manual/gnutls.html#index-gnutls_005ftransport_005fset_005fpull_005ftimeout_005ffunction

This callback is necessary for enforcing the DTLS mode timeouts when in blocking mode.

Otherwise the dtls pull function won't work.

The gnutls_pull_timeout() implementation refers to gnutls example:
https://gnutls.org/manual/gnutls.html#DTLS-echo-server-with-X_002e509-authentication-1

Signed-off-by: Jack Lau <jacklau1222gm@gmail.com>
 2026-01-11 04:24:26 +00:00
Hassan Hany
202b7d29e8 avcodec/pngenc: disable background disposal for non-alpha formats 2026-01-11 00:47:17 +00:00
Andreas Rheinhardt
685011003f avcodec/x86/pngdsp: Remove MMXEXT function overridden by SSSE3 
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
 2026-01-10 22:47:22 +01:00
Andreas Rheinhardt
31daa7cd87 avcodec/pngdsp: Use proper prefix ff_add_png->ff_png_add 
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
 2026-01-10 22:47:22 +01:00
Andreas Rheinhardt
5f15c067fe avcodec/pngdsp: Constify 
Also constify ff_png_filter_row().

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
 2026-01-10 22:47:22 +01:00
Andreas Rheinhardt
3bbb3269a1 avcodec/pngdec: Move ff_add_png_paeth_prediction() to pngdsp.c 
Also rename it to ff_png_add_paeth_prediction() so that it
has a proper prefix.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
 2026-01-10 22:47:22 +01:00