1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-07 11:13:41 +02:00
Commit Graph

117114 Commits

Author SHA1 Message Date
Anton Khirnov
18d492ff41 lavfi/af_channelsplit: support arbitrary channel layouts
Not just those containing channel values under 64. Also, remove an
arbitrary limitation on channel count.
2024-09-09 17:26:17 +02:00
Anton Khirnov
7dc81d33c2 lavfi/af_channelmap: remove an arbitrary limit on channel count 2024-09-09 17:26:17 +02:00
Brad Smith
a3f79fd22a aarch64: Implement support for elf_aux_info(3) on FreeBSD and OpenBSD
FreeBSD 12.0+, OpenBSD -current and what will be OpenBSD 7.6 support
elf_aux_info(3).

Signed-off-by: Brad Smith <brad@comstyle.com>
2024-09-09 07:17:49 -04:00
Lynne
1709f3830e
fftools/opt_common: add missing newline after printing codecs
This fixes
ffmpeg -help bsf=trace_headers =>

Supported codecs: av1 h264 hevc vvc mjpeg mpeg2video vp8 vp9<no newline>
2024-09-09 13:03:04 +02:00
Marvin Scholz
9556379943 lavfi/avfiltergraph: fix leak on error
Introduced in eddffbedb3

Fixes: CID1618897 Resource leak
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2024-09-09 12:27:00 +02:00
Brad Smith
fe4b9ef69f avutil/cpu_internal: Provide ff_getauxval() wrapper for getauxvaul()
Initially used for getauxval() but will be used to add support for
other API, such as elf_aux_info().

Signed-off-by: Brad Smith <brad@comstyle.com>
2024-09-09 06:20:11 -04:00
Lynne
5e9845f11e
vulkan(_decode): fix, simplify and improve queries
The old query code never worked properly, and did some hideous
heuristics to read the status bit, and work that into a return
code.
This is all best left to callers to do, which simplifies
our code a lot.

This also fixes minor validation errors regarding calling queries
which are not in their active state.
2024-09-09 07:05:46 +02:00
Lynne
3415e0533f
hwcontext_vulkan: disable more false positive validation checks
Both of these are fundamentally incompatible with video decoding.

The third one can be tracked in the following validation layer issue:
https://github.com/KhronosGroup/Vulkan-ValidationLayers/issues/6627
2024-09-09 07:05:46 +02:00
Lynne
9c65325819
vulkan_decode: use ff_vk_init
This solves the issue of an av_log function being called
with a context with invalid class.

Co-authored-by: Anton Khirnov <anton@khirnov.net>
2024-09-09 07:05:45 +02:00
Lynne
620822c0c5
vulkan: add a ff_vk_init function
This function sets the class correctly, and calls functions
that all users have to call anyway.
2024-09-09 07:05:45 +02:00
Lynne
0ffa967170
hwcontext_vulkan: ask for storage images by default
The issue is that we ask for storage images by default if
available, but because that is gated by the format supporting
storage images, and the check for the format supporting storage
images is gated by the usage, this resulted in a catch-22.
2024-09-09 07:05:45 +02:00
Lynne
eb5088d28c
vulkan_filter: require storage images properly, set usage flags explicitly
This caused images to be created without a storage usage, which broke
at least lavapipe.
2024-09-09 07:05:44 +02:00
Lynne
c41ef7f2ff
hwcontext_vulkan: add PREP_MODE_GENERAL for non-transfer_dst images
Vulkan filters don't need images which can be transferred into.
2024-09-09 07:05:44 +02:00
Lynne
66e950fcac
vulkan_video: move imageview creation and DPB fields to common context
Shared between decoders and encoders.
2024-09-09 07:05:44 +02:00
Lynne
06483d039a
vulkan: error out if query is called without being initialized 2024-09-09 07:05:43 +02:00
Lynne
fd86c373c2
vulkan: use correct return codes for query errors 2024-09-09 07:05:43 +02:00
Lynne
0a1f724bf4
vulkan_video: add utilities for H264 level/profile mapping 2024-09-09 07:05:42 +02:00
Lynne
3de73f1262
hw_base_encode: make recon_frames_ref optional
Vulkan supports some stupidly odd hardware, that unfortunately,
most modern GPUs happen to be.
The DPB images for encoders may be required to be preallocated
all at once, and rather than be individual frames, be layers of
a single frame.

As the hw_base_encode code is written with the thought that either
the driver or the device itself supports sane image allocation,
Vulkan does not leave us with this option.

So, in the case that the hardware does not support individual
frames to be used as DPBs, make the DBP frames context optional,
and let the subsystem manage this.
2024-09-09 07:05:42 +02:00
Lynne
e6019ed075
hw_base_encode: allocate DPB image upfront
Vulkan requires this, as it needs to initialize state upfront.
2024-09-09 07:05:42 +02:00
Lynne
034e25e1a4
hw_base_encode: move VAAPI SPS/PPS constructors to a shared file 2024-09-09 07:05:41 +02:00
Lynne
9db68ed042
hw_base_encode: refactor picture allocation/freeing
This commit cleans up and refactors the mess of private state upon
private state that used to be.

Now, FFHWBaseEncodePicture is fully initialized upon call-time,
and, most importantly, this lets APIs which require initialization
data for frames (VkImageViews) to initialize this for both the
input image, and the reconstruction (DPB) image.

Signed-off-by: Tong Wu <wutong1208@outlook.com>
2024-09-09 07:05:24 +02:00
James Almer
fdf8025eb6 avcodec/avcodec: remove usage of __typeof__()
It's non-standard C.

Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: James Almer <jamrial@gmail.com>
2024-09-08 20:33:04 -03:00
Niklas Haas
b4e918221c doc/APIchanges: mention correct version number
Missed while rebasing
2024-09-08 18:07:30 +02:00
Michael Niedermayer
4bc2f90cff
MAINTAINERS: remove Aurelien and David from matroska*
Aurelien last activity in git is from 2017, last activity on the ML is from 2018
David last activity in git is from 2011, and ML from 2013

matroska* is actively maintained

CC: Aurelien Jacobs <aurel@gnuage.org>
CC: David Conrad <lessen42@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:39 +02:00
Michael Niedermayer
85aee1dc31
MAINTAINER: iamf is maintained by James
CC: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:39 +02:00
Michael Niedermayer
ec8b5187f9
MAINTAINERS: Add status to icecast
Last icecast threads have Marvin in them so it seems maintained

CC: Marvin Scholz <epirat07@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:39 +02:00
Michael Niedermayer
49226216b1
MAINTAINERS: Rangecoder looks quite familiar, ill maintain it
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:38 +02:00
Michael Niedermayer
c7083738bd
MAINTAINERS: Mark parts from Ludmila as unmaintained
Last i see from Ludmila is from 2015

CC: Ludmila Glinskih <lglinskih@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:38 +02:00
Michael Niedermayer
5d02b0f72e
MAINTAINERS: Update fate admin
Timothy has not reacted to any fate admin related mails i can remember
last was handled by timo, prior where probably handled by me

CC: Timothy Gu <timothygu99@gmail.com>
CC: Timo Rothenpieler <timo@rothenpieler.org>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:38 +02:00
Michael Niedermayer
006ee12a9a
MAINTAINERS: Mark parts from Robert as unmaintained
Last mail and commit from Robert seems from 2011, thats 13 years ago

CC: Rob <robert.swain@gmail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:37 +02:00
Christian Helmrich
865cd3c056
avfilter: add XPSNR filter
Add XPSNR video filter
Register new filter xpsnr.

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-09-08 17:51:37 +02:00
Niklas Haas
e6983ed525 Changelog: mention YUV range negotiation 2024-09-08 14:07:59 +02:00
Niklas Haas
8517ff2082 fftools/ffmpeg_filter: remove YUVJ hack
This is no longer needed, since we now correctly negotiate the required
range from the mjpeg encoder via avcodec_get_supported_config().

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2024-09-08 13:59:29 +02:00
Niklas Haas
a850f80e2c fftools/ffmpeg_filter: propagate codec yuv metadata to filters
To convert between color spaces/ranges, if needed by the codec
properties.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2024-09-08 13:59:29 +02:00
Niklas Haas
33d5a4ec4e fftools/ffmpeg_filter: switch to avcodec_get_supported_config()
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2024-09-08 13:59:29 +02:00
Anton Khirnov
f3e265c690 fftools/ffmpeg_mux_init: factor out binding filters to output streams 2024-09-08 13:59:29 +02:00
Niklas Haas
7d6cd13ccb fftools/ffmpeg_filter: simplify choose_pix_fmts
The only meaningful difference between choose_pix_fmts and the default
code was the inclusion of an extra branch for `keep_pix_fmt` being true.

However, in this case, we either:
1. Force the specific `ofp->format` that we inherited from
   ofilter_bind_ost, or if no format was set:
2. Print an empty format list

Both of these goals can be accomplished by simply moving the decision
logic to ofilter_bind_ost, to avoid setting any format list when
keep_pix_fmt is enabled. This is arguably cleaner as it moves format
selection logic to a single function. In the case of branch 1, nothing
else needs to be done as we already force the format provided in
ofp->format, if any is set. Add an assertion to verify this assumption
just in case.

(Side note: The "choose_*" family of functions are arguably misnomers,
as they should really be called "print_*" - their current behavior is to
print the relevant format lists to the `vf/af_format` filter arguments)

Signed-off-by: Anton Khirnov <anton@khirnov.net>
Signed-off-by: Niklas Haas <git@haasn.dev>
2024-09-08 13:59:29 +02:00
Niklas Haas
7b0eca1e51 fftools/ffmpeg_mux_init: switch to avcodec_get_supported_config() 2024-09-08 13:59:29 +02:00
Niklas Haas
377652da19 fftools: drop unused/hacky macros
Having macros initialize local variables seems strange to me, and there
are no more current users of these macros. (The one that was commented
out was incorrect anyway, since the macro has changed in the meantime)
2024-09-08 13:59:29 +02:00
Niklas Haas
69b8405482 fftools/opt_common: switch to avcodec_get_supported_config()
While rewriting this macro, I decided to make it a bit more flexible so
it can work for all of the fields (including future fields) in a more
generic way, and to also print the channel layout using an AVBPrint to
avoid hard-coding the assumption that the length is less than 128
characters.
2024-09-08 13:59:29 +02:00
Niklas Haas
4524d527bf avcodec/codec_internal: nuke init_static_data()
All hail get_supported_config()
2024-09-08 13:59:29 +02:00
Niklas Haas
2c2dfd9149 avcodec/mjpegenc: switch to get_supported_config()
This codec's capabilities should be set dynamically based on the value
of strict_std_compliance, when available. This will allow us to finally
get rid of the strictness hack in ffmpeg_filter.c.
2024-09-08 13:59:29 +02:00
Niklas Haas
a577172c99 avcodec/libaomenc: switch to get_supported_config() 2024-09-08 13:59:29 +02:00
Niklas Haas
0977d968e4 avcodec/libvpxenc: switch to get_supported_config() 2024-09-08 13:59:29 +02:00
Niklas Haas
de421cac8b avcodec/libx265: switch to get_supported_config() 2024-09-08 13:59:29 +02:00
Niklas Haas
088108fc7a avcodec/allcodecs: add backcompat for new config API
In order to avoid breaking older clients not yet using the new API, we
need to add backwards compatibility for codecs which have switched from
init_static() to get_supported_config().

This function can be removed entirely once the deprecated static fields
are removed.
2024-09-08 13:59:29 +02:00
Niklas Haas
13cec7bb5f avcodec/encode: switch to avcodec_get_supported_config() 2024-09-08 13:59:29 +02:00
Niklas Haas
3305767560 avcodec: add avcodec_get_supported_config()
This replaces the myriad of existing lists in AVCodec by a unified API
call, allowing us to (ultimately) trim down the sizeof(AVCodec) quite
substantially, while also making this more trivially extensible.

In addition to the already covered lists, add two new entries for color
space and color range, mirroring the newly added negotiable fields in
libavfilter.

Once the deprecation period passes for the existing public fields, the
rough plan is to move the commonly used fields (such as
pix_fmt/sample_fmt) into FFCodec, possibly as a union of audio and video
configuration types, and then implement the rarely used fields with
custom callbacks.
2024-09-08 13:59:25 +02:00
Niklas Haas
703288cec6 avcodec/internal: add FFCodec.color_ranges
I went through all codecs and put them into five basic categories:

1. JPEG range only
2. MPEG range only
3. Explicitly tagged
4. Broken (codec supports both but encoder ignores tags)
5. N/A (headerless or pseudo-formats)

Filters in category 5 remain untouched. The rest gain an explicit
assignment of their supported color ranges, with codecs in category
4 being set to MPEG-only for safety.

It might be considered redundant to distinguish between 0 (category 5)
and MPEG+JPEG (category 3), but in doing so we effectively communicate
that we can guarantee that these tags will be encoded, which is distinct
from the situation where there are some codecs that simply don't have
tagging or implied semantics (e.g. rawvideo).

A full list of codecs follows:

JPEG range only:
 - amv
 - roqvideo

MPEG range only:
 - asv1, asv2
 - avui
 - cfhd
 - cljr
 - dnxhd
 - dvvideo
 - ffv1
 - flv
 - h261, h263, h263p
 - {h263,vp8}_v4l2m2m
 - huffyuv, ffvhuff
 - jpeg2000
 - libopenjpeg
 - libtheora
 - libwebp, libwebp_anim
 - libx262
 - libxavs, libxavs2
 - libxvid
 - mpeg1video, mpeg2video
 - mpeg2_qsv
 - mpeg2_vaapi
 - mpeg4, msmpeg4, msmpeg4v2, wmv1, wmv2
 - mpeg4_omx
 - prores, prores_aw, prores_ks
 - rv10, rv20
 - snow
 - speedhq
 - svq1
 - tiff
 - utvideo

Explicitly tagged (MPEG/JPEG):
 - {av1,h264,hevc}_nvenc
 - {av1,h264,hevc}_vaapi
 - {av1,h264,hevc,vp8,vp9,mpeg4}_mediacodec
 - {av1,h264,hevc,vp9}_qsv
 - h264_amf
 - {h264,hevc,prores}_videotoolbox
 - libaom-av1
 - libkvazaar
 - libopenh264
 - librav1e
 - libsvtav1
 - libvpx, libvpx-vp9
 - libx264
 - libx265
 - ljpeg
 - mjpeg
 - vc2

Broken (encoder ignores tags):
 - {av1,hevc}_amf
 - {h264,hevc,mpeg4}_v4l2m2m
 - h264_omx
 - libxeve
 - magicyuv
 - {vp8,vp9,mjpeg}_vaapi

N/A:
 - ayuv, yuv4, y41p, v308, v210, v410, v408 (headerless)
 - pgmyuv (headerless)
 - rawvideo, bitpacked (headerless)
 - vnull, wrapped_avframe (pseudocodecs)
2024-09-08 13:58:11 +02:00
Marvin Scholz
8a314e3f11 lavu/opt: Fix return of uninitialised value
In one of the failure paths of av_opt_get_array, the ret variable
was accidentally declared again, shadowing the outer one and once when
jumping to the fail label would return the still uninitialised outer
one.

To fix this simply remove the local declaration of ret that shadowed
the outer one.

Introduced in d89930f866

Fixes: CID1618663 Uninitialized scalar variable
2024-09-07 20:42:26 +02:00