ffmpeg 9 - #30455
Conversation
openwrt-ai
left a comment
openwrt-ai
left a comment
There was a problem hiding this comment.
Commit checks
- eb43bbc "ffmpeg: update to 9.0.1" — the new base configure block passes
--disable-v4l2-request/--disable-libudev, and the custom/mini/audio-dec blocks still pass--disable-postproc; stock 9.0.1configurehas none of those (nov4l2_requestand nopostproclibrary in FFmpeg 9's option lists) and dies on an unknown--disable-*. The patches that add them arrive in cd36712 and 0f54e2c, so eb43bbc, f080c9b and cd36712 each fail configure — the message should say the series is not bisectable, or the patch commits should be ordered ahead of the Makefile rework.
Generated by Claude Code
| +FFMPEG_FULL_TLS_GNUTLS:libgnutls \ | ||
| +FFMPEG_FULL_TLS_OPENSSL:libopenssl \ | ||
| +FFMPEG_FULL_TLS_MBEDTLS:libmbedtls \ | ||
| +FFMPEG_FULL_LIBDAV1D:libdav1d \ |
There was a problem hiding this comment.
There is no libdav1d package in this feed or in openwrt/openwrt — grep -r dav1d over both trees finds nothing — so with FFMPEG_FULL_LIBDAV1D defaulting to y this dependency is unresolvable and libffmpeg-full (and therefore ffmpeg/ffprobe) cannot be selected at all. +FFMPEG_FULL_VAAPI:libva on line 417 has the same problem, there is no libva package either (default n, so it only bites when enabled). Either land the dav1d/libva packages first, or drop these two options from this PR.
Generated by Claude Code
There was a problem hiding this comment.
should be fixed by dropping support for respective libs (for now, until also upstreamed)
| default y | ||
|
|
||
| if PACKAGE_libffmpeg-custom | ||
| menu "FFmpeg (custom variant) configuration" |
There was a problem hiding this comment.
Package/<pkg>/config is emitted verbatim right after menuconfig PACKAGE_libffmpeg-custom with no guard of its own (package-metadata.pl:378), so replacing the if with a plain menu makes the whole custom option tree visible and settable in every buildroot config and leaves MENU:=1 with nothing to nest. Restore the guard (the closing endmenu on the last line then has to go back to endif).
| menu "FFmpeg (custom variant) configuration" | |
| if PACKAGE_libffmpeg-custom |
Generated by Claude Code
There was a problem hiding this comment.
solved differently — MENU:=1 is gone and the free-standing menu is now deliberate and documented, thanks
Generated by Claude Code
| few hundred kB of code and tables. | ||
|
|
||
| config FFMPEG_FULL_PROGRAMS | ||
| bool "Build the command line programs (ffmpeg, ffprobe)" |
There was a problem hiding this comment.
Nothing connects this symbol to the ffmpeg/ffprobe packages, so turning it off while either is selected does not leave them "with nothing to install" — Package/ffmpeg/install runs INSTALL_BIN on a binary that was never built and the build aborts. Add @FFMPEG_FULL_PROGRAMS to the DEPENDS of Package/ffmpeg and Package/ffprobe.
Generated by Claude Code
There was a problem hiding this comment.
| # the tarball unpacks into ffmpeg-$(PKG_VERSION), not $(PKG_NAME)-$(PKG_VERSION) | ||
| PKG_BUILD_DIR:=$(BUILD_DIR)/$(if $(BUILD_VARIANT),$(PKG_NAME)-$(BUILD_VARIANT)/)ffmpeg-$(PKG_VERSION) |
There was a problem hiding this comment.
nit: the premise is wrong — PKG_NAME is ffmpeg, and include/package.mk:11 already defaults to exactly this string, so the override is a no-op. Drop both lines.
Generated by Claude Code
There was a problem hiding this comment.
There was a problem hiding this comment.
Warning
Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.
Pull request overview
Updates the OpenWrt FFmpeg package to 9.0.1, restructuring the build to be more deterministic and explicit about enabled features/libraries, and bringing in/refreshing downstream patches (notably for V4L2/Kodi-related use cases).
Changes:
- Bump FFmpeg from 6.1.4 to 9.0.1 and adjust build dir handling for the new tarball layout.
- Rework configure flags and Kconfig options to avoid autodetect and make feature selection explicit (full/custom variants).
- Refresh patch stack: drop a set of now-upstream/obsolete warning-fix patches and add/adjust V4L2-related patches (incl. V4L2 M2M deinterlace filter and V4L2 Request format fixes).
Reviewed changes
Copilot reviewed 16 out of 18 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
| multimedia/ffmpeg/Makefile | Version bump to 9.0.1 and major packaging/configure refactor for deterministic builds and explicit feature toggles. |
| multimedia/ffmpeg/Config.in | Restructure custom-variant menu; remove custom patented toggle in favor of global BUILD_PATENTED. |
| multimedia/ffmpeg/patches/020-libavcodec-fix-Wint-conversion-in-vulkan.patch | Removed (likely upstreamed/obsolete in FFmpeg 9). |
| multimedia/ffmpeg/patches/030-avformat-file-guard-fd_dup-by-FD_PROTOCOL-or-PIPE_PR.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/040-vulkan_decode-fix-the-print-format-of-VkDeviceSize.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/050-avcodec-dct-Make-declarations-and-definitions-match.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/060-avutil-tx-fix-GCC-memset-warning.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/070-avformat-rawdec-guard-by-CONFIG_DATA_DEMUXER.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/080-avcodec-pcm-bluray-dvd-Use-correct-pointer-types-on-.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/090-avcodec-tiff-Suppress-unused-variable-warnings.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/100-avcodec-tableprint_vlc-Unbreak-hardcoded-tables.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/110-libreelec.patch | Add/port LibreELEC/Kodi-oriented changes (dav1d get_format; swscale warning removal). |
| multimedia/ffmpeg/patches/120-avfilter-af_channelsplit-fix-mixed-declaration-and-c.patch | Removed (likely upstreamed/obsolete). |
| multimedia/ffmpeg/patches/140-vf-deinterlace-v4l2m2m.patch | Adds a new V4L2 M2M deinterlacing filter and wires it into FFmpeg build. |
| multimedia/ffmpeg/patches/150-v4l2-request-define-sunxi-afbc-formats.patch | Adds missing downstream sunxi/cedrus AFBC fourcc defines for sanitized kernel headers. |
| multimedia/ffmpeg/patches/160-libavdevice-v4l2-libv4l2-posix-ioctl.patch | Fixes libv4l2 ioctl signature mismatch on musl by defining expected feature macros. |
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
| +static inline uint32_t v4l2_pix_fmt_from_drm_format(uint64_t drm_format) | ||
| +{ | ||
| + switch(drm_format) { | ||
| +#if defined(V4L2_PIX_FMT_SUNXI_TILED_NV12) && defined(DRM_FORMAT_MOD_ALLWINNER_TILED) | ||
| + case DRM_FORMAT_MOD_ALLWINNER_TILED: | ||
| + return V4L2_PIX_FMT_SUNXI_TILED_NV12; | ||
| +#endif |
| +static inline uint64_t drm_format_modifier(uint64_t drm_format) | ||
| +{ | ||
| +#ifdef DRM_FORMAT_MOD_ALLWINNER_TILED | ||
| + if (drm_format == DRM_FORMAT_MOD_ALLWINNER_TILED) | ||
| + return DRM_FORMAT_MOD_ALLWINNER_TILED; | ||
| +#endif | ||
| + return DRM_FORMAT_MOD_LINEAR; | ||
| + | ||
| +} |
| +static int v4l2_buffer_export_drm(V4L2Buffer* avbuf, uint64_t drm_format) | ||
| +{ | ||
| + struct v4l2_exportbuffer expbuf; | ||
| + int i, ret; | ||
| + | ||
| + for (i = 0; i < avbuf->num_planes; i++) { | ||
| + memset(&expbuf, 0, sizeof(expbuf)); | ||
| + | ||
| + expbuf.index = avbuf->buffer.index; | ||
| + expbuf.type = avbuf->buffer.type; | ||
| + expbuf.plane = i; | ||
| + | ||
| + ret = ioctl(avbuf->q->ctx->fd, VIDIOC_EXPBUF, &expbuf); | ||
| + if (ret < 0) | ||
| + return AVERROR(errno); | ||
| + | ||
| + avbuf->fd = expbuf.fd; | ||
| + | ||
| + if (V4L2_TYPE_IS_MULTIPLANAR(avbuf->buffer.type)) { | ||
| + /* drm frame */ | ||
| + avbuf->drm_frame.objects[i].size = avbuf->buffer.m.planes[i].length; | ||
| + avbuf->drm_frame.objects[i].fd = expbuf.fd; | ||
| + avbuf->drm_frame.objects[i].format_modifier = drm_format_modifier(drm_format); | ||
| + } else { | ||
| + /* drm frame */ | ||
| + avbuf->drm_frame.objects[0].size = avbuf->buffer.length; | ||
| + avbuf->drm_frame.objects[0].fd = expbuf.fd; | ||
| + avbuf->drm_frame.objects[0].format_modifier = drm_format_modifier(drm_format); | ||
| + } | ||
| + } |
| + if (pfd.revents & events) { | ||
| + memset(&buf, 0, sizeof(buf)); | ||
| + buf.memory = V4L2_MEMORY_MMAP; | ||
| + buf.type = queue->format.type; | ||
| + if (V4L2_TYPE_IS_MULTIPLANAR(queue->format.type)) { | ||
| + memset(planes, 0, sizeof(planes)); | ||
| + buf.length = VIDEO_MAX_PLANES; | ||
| + buf.m.planes = planes; | ||
| + } | ||
| + | ||
| + ret = ioctl(ctx->fd, VIDIOC_DQBUF, &buf); |
| + ctx->orig_width = drm_desc->layers[0].planes[0].pitch; | ||
| + ctx->orig_height = drm_desc->layers[0].planes[1].offset / ctx->orig_width; |
| + memset(&cap, 0, sizeof(cap)); | ||
| + ret = ioctl(ctx->fd, VIDIOC_QUERYCAP, &cap); | ||
| + if (ret < 0) | ||
| + return ret; |
| + if (ret) | ||
| + av_log(NULL, AV_LOG_ERROR, "VIDIOC_G_FMT failed: %d\n", ret); |
| + AVBufferRef *hw_frames_ctx; | ||
| + | ||
| + /* | ||
| + * TODO: check if its really neccessary to hold this |
8cecafa to
3fbbdfe
Compare
| +FFMPEG_TLS_GNUTLS:libgnutls \ | ||
| +FFMPEG_TLS_OPENSSL:libopenssl \ | ||
| +FFMPEG_TLS_MBEDTLS:libmbedtls \ | ||
| +FFMPEG_LIBDAV1D:libdav1d \ |
There was a problem hiding this comment.
There is no libdav1d package in this feed or in openwrt/openwrt, and FFMPEG_LIBDAV1D defaults to y, so all ten Feeds Package Test Build jobs die at configure (mipsel_24kc shown):
ERROR: dav1d >= 1.0.0 not found using pkg-config
make[2]: *** [Makefile:661: .../.configured_2fc8a77d2a030a29ff9b591c10b3e409] Error 1
ERROR: package/feeds/packages_ci/ffmpeg failed to build (build variant: default).
WARNING: Makefile 'package/feeds/packages_ci/ffmpeg/Makefile' has a dependency on 'libdav1d', which does not exist
Either land a dav1d package first or drop the option from this PR; +FFMPEG_VAAPI:libva on line 302 is equally unresolvable (default n, so it only bites once enabled).
Generated by Claude Code
There was a problem hiding this comment.
| CONFLICTS:=libffmpeg-full | ||
| TITLE+= (mini) | ||
| VARIANT:=mini | ||
| CONFLICTS:=libffmpeg |
There was a problem hiding this comment.
nit: on master Package/libffmpeg/Default carried PROVIDES:= libffmpeg, so the mini variant could satisfy a +libffmpeg dependency; here it provides nothing and conflicts with the real libffmpeg. That leaves mpd and minidlna always pulling the full build, and no in-tree consumer able to use libffmpeg-mini at all. Intentional?
Generated by Claude Code
143f3ad to
ce2f1f8
Compare
|
If you would be so kind as to remove me as package maintainer, I would appreciate it. |
Of course, if you wish so - @antonlacon you still in? Asking also, as ffmpeg6 has been a while.. |
|
re: squeezelite, I am not its official maintainer so take my view with a grain of salt - i may have assumed a de facto status on account of the last 2 annual bumps. Not opposed at all to depending on the full ffmpeg myself, acknowledging the increased size will prohibit installation on a subset of storage limited hardware. Curious to learn about the plans for kodi and openwrt, I recently became involved with its team and more specifically the flatpak maintenance. You might also consider team membership if this becomes a long term labor of love for you! |
Then I'd say let's not overengineer but see if there's actual demand for a lighter variant which we can then still provide.
I always wanted getting kodi ported to OpenWrt, but it's hell of a monster. This comment (xbmc/xbmc#28454 (comment)) sums up my prime example why I gave up quite a few times in the past, however @dangowrt gave it another try and achieved a PoC I got hyped about and recently ported pipewire, ffmpeg9 and kodi22. However my personal interaction with the kodi community - at least on GitHub - was rather disappointing so far I have to admit. While I understand it's a huge end-user facing project, attracting lot's of non-technical (and annoying piracy) questions and they have to establish mechanisms to deal with the flood, conversations like e.g. xbmc/xbmc#17873 rather discouraged me from engaging further :/ |
| CONFIG_FFMPEG_VAAPI \ | ||
| CONFIG_FFMPEG_LIBV4L2 \ | ||
| CONFIG_FFMPEG_TLS_NONE \ | ||
| CONFIG_FFMPEG_TLS_GNUTLS \ | ||
| CONFIG_FFMPEG_TLS_OPENSSL \ | ||
| CONFIG_FFMPEG_TLS_MBEDTLS \ | ||
| CONFIG_FFMPEG_BZLIB \ | ||
| CONFIG_FFMPEG_LZMA \ | ||
| CONFIG_FFMPEG_ICONV \ | ||
| CONFIG_FFMPEG_LIBDAV1D \ |
There was a problem hiding this comment.
nit: FFMPEG_VAAPI and FFMPEG_LIBDAV1D are commented out in Config.in now, so these two entries track symbols that can never be set.
| CONFIG_FFMPEG_VAAPI \ | |
| CONFIG_FFMPEG_LIBV4L2 \ | |
| CONFIG_FFMPEG_TLS_NONE \ | |
| CONFIG_FFMPEG_TLS_GNUTLS \ | |
| CONFIG_FFMPEG_TLS_OPENSSL \ | |
| CONFIG_FFMPEG_TLS_MBEDTLS \ | |
| CONFIG_FFMPEG_BZLIB \ | |
| CONFIG_FFMPEG_LZMA \ | |
| CONFIG_FFMPEG_ICONV \ | |
| CONFIG_FFMPEG_LIBDAV1D \ | |
| CONFIG_FFMPEG_LIBV4L2 \ | |
| CONFIG_FFMPEG_TLS_NONE \ | |
| CONFIG_FFMPEG_TLS_GNUTLS \ | |
| CONFIG_FFMPEG_TLS_OPENSSL \ | |
| CONFIG_FFMPEG_TLS_MBEDTLS \ | |
| CONFIG_FFMPEG_BZLIB \ | |
| CONFIG_FFMPEG_LZMA \ | |
| CONFIG_FFMPEG_ICONV \ |
Generated by Claude Code
There was a problem hiding this comment.
Temporarily disabled features, as dependencies are not upstream(ed) yet, as you already figured out earlier. But plan is to just uncomment and enable once that happened, hence these symbols are only "deactivated" instead of "removed".
| PKG_NAME:=ffmpeg | ||
| PKG_VERSION:=6.1.4 | ||
| PKG_RELEASE:=2 | ||
| PKG_VERSION:=9.0.1 |
There was a problem hiding this comment.
{standard input}:9366: Error: invalid operands `sll $25,$4,0x11'
{standard input}:9367: Error: operand 2 must be an immediate expression `slt $17,$4,$25'
{standard input}:9369: Error: opcode not supported on this processor: mips32r2 (mips32r2) `movz $6,$16,$17'
make[3]: *** [ffbuild/common.mak:83: libavcodec/vvc/cabac.o] Error 1
ERROR: package/feeds/packages_ci/ffmpeg failed to build (build variant: default).
mipsel_24kc dies on FFmpeg 9's VVC CABAC; the register numbers ($24/$25) and the rejected movz/movn suggest the MIPS inline asm ends up assembled in MIPS16 mode. Only aarch64_generic is green on this head, so the other targets may need a look too (--disable-inline-asm on mips, or a .set nomips16 patch, if you want to chase this one).
Generated by Claude Code
There was a problem hiding this comment.
The build is for mips16 and --disable-inline-asm does not appear to be a fix for that build issue.
That option only affects code guarded by HAVE_INLINE_ASM. mips/cabac.h has no such guard, so the assembly is compiled regardless and the build fails the same way.
It would only cost mips32 optimisations for nothing.
There was a problem hiding this comment.
Should be fixed now by re-adding a patch which already existed but got lost in a previous version bump.
Didn't surface in ffmpeg 6.x builds, as the only codecs using that were guarded by @BUILD_PATENDED.
Now with ffmpeg 9.x we also have codecs being built unconditionally.
00f1e62 to
b7a6537
Compare
Rework the package while bumping from 6.1.4.
Variants: libffmpeg is the configurable build (menu "FFmpeg
configuration"; the defaults give a full-featured build), libffmpeg-mini
a fixed small decoder set for playback-only consumers; the two conflict.
The former -full, -custom and -audio-dec variants are folded into
libffmpeg: its menu carries the per-codec/format/protocol selection the
custom variant had, controlled by a "manual component selection"
option that is off by default. libffmpeg provides libffmpeg-full and
libffmpeg-audio-dec as transitional aliases for consumers not yet
updated. libffmpeg-mini no longer provides the libffmpeg virtual
package: its fixed set is too small for the in-tree consumers of that
name (mpd, minidlna), which need the default build; it is for consumers
that depend on it explicitly.
* deterministic configure: --disable-autodetect plus an explicit
--disable baseline for everything configure would pick up from the
environment; each optional library/hwaccel is one config symbol
driving both the conditional DEPENDS and the configure flag
* menu: licence, size/speed, programs, hwaccels (libdrm, V4L2 M2M,
VA-API), TLS backend (default mbedTLS, OpenWrt's default TLS
library), external libraries, system libraries, component selection.
Defaults follow what in-tree consumers need: libopus, LAME (shine
on soft-float) and libx264 on; fdk-aac only for a nonfree, non-GPL
build. The ALSA input device is dropped (no in-tree consumer).
libdav1d (FFmpeg has no native software AV1 decoder) and VA-API are
prepared but commented out until the dav1d and libva packages exist
in the feeds.
Patches:
010-pkgconfig: kept unchanged: expresses libdir and includedir
relative to ${prefix} in the generated .pc files
020-libavcodec-fix-Wint-conversion-in-vulkan,
030-avformat-file-guard-fd_dup-by-FD_PROTOCOL-or-PIPE_PR,
040-vulkan_decode-fix-the-print-format-of-VkDeviceSize,
050-avcodec-dct-Make-declarations-and-definitions-match,
060-avutil-tx-fix-GCC-memset-warning,
080-avcodec-pcm-bluray-dvd-Use-correct-pointer-types-on-,
090-avcodec-tiff-Suppress-unused-variable-warnings,
100-avcodec-tableprint_vlc-Unbreak-hardcoded-tables,
120-avfilter-af_channelsplit-fix-mixed-declaration-and-c:
dropped: merged upstream
070-avformat-rawdec-guard-by-CONFIG_DATA_DEMUXER: dropped: only
silenced an unused-function warning in builds without the data
demuxer
170-tls_mbedtls-build-without-version-module: new: OpenWrt's mbedtls
package disables MBEDTLS_VERSION_C, so mbedtls_version_get_number()
is not declared; fall back to the compile-time MBEDTLS_VERSION_NUMBER
180-mips-cabac-no-mips16: re-added (Rosen Penev, 2020; dropped by the
6.1.2 update): restores the MIPS16 guard for the MIPS CABAC inline
assembly; FFmpeg 9's VVC decoder compiles that file on every MIPS
build, not only with BUILD_PATENTED
190-configure-hevcparse-select-hevc_sei: new: configure's hevcparse
component does not select hevc_sei although hevc/parse.c needs it;
without BUILD_PATENTED the HEVC decoder and parser that used to pull
it in are off while the default-enabled dovi_split bitstream filter
still needs hevcparse, leaving ff_hevc_decode_nal_sei undefined
The LibreELEC patches (libpostproc, Kodi compatibility, V4L2 Request
API) and the options they introduce are added by the following commits.
The explicit GPL option and the explicit libdrm dependency follow
Daniel Golle's 7.1.4 update proposal of this package, which was superseded
by this PR/update.
Assisted-By: Claude Fable 5.1 <noreply@anthropic.com>
Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
libv4l2.h only declares v4l2_ioctl() with the POSIX "int request" signature when the consumer defines a feature macro first; FFmpeg detects that signature (HAVE_IOCTL_POSIX) but never tells the header, and GCC 14 turns the resulting pointer-type mismatch into an error. Define both macro spellings (v4l-utils <= 1.30 and >= 1.32). Same fix as q66's posix-ioctl.patch in Chimera Linux, also carried by Alpine, extended for the older macro name. Assisted-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
Upstream dropped libpostproc in 8.0; Kodi still requires it.
The --enable/--disable-postproc option the patch re-introduces is set
explicitly: on together with GPL in libffmpeg, off in libffmpeg-mini.
Patches:
100-postproc: carried over from LibreELEC (Matthias Reichl's rebase
of Michael Niedermayer's libpostproc source plugin;
packages/multimedia/ffmpeg/patches/postproc/0001-postproc.patch as
of LibreELEC commit 5f03a122f8): re-adds libpostproc as a source
plugin
Header unchanged, hunks refreshed with quilt; the patched tree is
identical to the one LibreELEC's file produces.
Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
Patches:
110-libdav1d-get-format: carried over from LibreELEC (Lukas Rusak):
adds a get_format callback to the libdav1d decoder so Kodi's
hardware-first, software-fallback decoder setup initialises
multithreaded software decoding properly; without effect until the
libdav1d option is enabled
111-swscale-yuv2rgb-no-warning: carried over from LibreELEC
(Christian Hewitt): removes the warning swscale logs on every
context initialisation without an accelerated colourspace
conversion
LibreELEC keeps both in one file
(packages/multimedia/ffmpeg/patches/libreelec/0001-libreelec.patch as
of LibreELEC commit 5d54b2dd82); they are split here so each keeps its
own header. Headers unchanged, hunks refreshed with quilt; the patched
tree is identical to the one LibreELEC's file produces.
Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
V4L2 Request API hwaccels for stateless decoders such as Allwinner
cedrus, exporting DRM_PRIME frames, and a V4L2 M2M deinterlace filter
driving mem2mem deinterlacers such as Allwinner's sun8i-di on DRM_PRIME
frames. The hwaccels introduce the v4l2-request and libudev configure
options.
Add the FFMPEG_V4L2_REQUEST option (pulls libdrm and libudev) to the
libffmpeg menu, keep both new configure options explicitly off in the
common block like the other autodetected ones, and re-add the
*_v4l2request hwaccels under manual component selection.
Patches:
120-v4l2-request-01 to -13: carried over from LibreELEC
(packages/multimedia/ffmpeg/patches/v4l2-request/0001-v4l2-request.patch
as of LibreELEC commit 79de202a10), one file per patch of the series:
01 avutil/hwcontext: Add hwdevice type for V4L2 Request API
(Jonas Karlman)
02 avutil/hwcontext_v4l2request: Probe for a capable media and
video device (Jonas Karlman)
03 avcodec: Add common V4L2 Request API code (Jonas Karlman)
04 avcodec/v4l2request: Add common decode support for hwaccels
(Jonas Karlman)
05 avcodec: Add V4L2 Request API mpeg2 hwaccel (Jonas Karlman)
06 avcodec/h264dec: add ref_pic_marking and pic_order_cnt bit_size
to slice context (Boris Brezillon)
07 avcodec: Add V4L2 Request API h264 hwaccel (Jernej Skrabec)
08 avcodec: Add V4L2 Request API hevc hwaccel (Jernej Skrabec)
09 avcodec: Add V4L2 Request API vp8 hwaccel (Boris Brezillon)
10 avcodec: Add V4L2 Request API vp9 hwaccel (Boris Brezillon)
11 avcodec: Add V4L2 Request API av1 hwaccel (Jonas Karlman)
12 avutil/hwcontext_v4l2request: Add support for AFBC_16X16_SPLIT
pix fmts (Jernej Skrabec)
13 avutil/hwcontext_v4l2request: Add support for BROADCOM_SAND128
pix fmts (Jonas Karlman)
Co-developed by Alex Bee, Benjamin Gaignard and Ezequiel Garcia.
140-vf-deinterlace-v4l2m2m: carried over from LibreELEC
(packages/multimedia/ffmpeg/patches/vf-deinterlace-v4l2m2m/0001-vf-deinterlace-v4l2m2m.patch
as of LibreELEC commit 9ef6c9c6f7; Jernej Skrabec): adds the
deinterlace_v4l2m2m filter
Headers unchanged, hunks refreshed with quilt; the patched tree is
identical to the one LibreELEC's files produce.
LibreELEC's v4l2-drmprime patch (stateful M2M decoders) is not carried:
it needs a rebase for FFmpeg 9 which LibreELEC has not done either.
Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
The cedrus AFBC capture fourccs are downstream kernel additions missing from the sanitised toolchain headers, so 10-bit HEVC found no DRM_PRIME capture format and silently fell back to software decoding. Define them locally when the headers lack them. Assisted-By: Claude Fable 5.1 <noreply@anthropic.com> Signed-off-by: Mirko Vogt <mirko-openwrt@nanl.de>
|
Referencing ticket #30494 here, as it essentially caused me to fork and modify existing patch files just to make the CI happy. |
|
All failed target builds are failing due to (in this case for arm): |
| # list documents the intent and is independent of changes to configure's | ||
| # autodetect list. The libffmpeg variant re-enables bzlib/iconv/lzma and the | ||
| # hwaccels through its FFMPEG_* options below; later flags take precedence. | ||
| # v4l2-request and libudev are introduced by patches/120-v4l2-request.patch, |
There was a problem hiding this comment.
nit: the same commit splits that file into 120-v4l2-request-01.patch … -13.patch, so this filename no longer exists; same at line 498.
| # v4l2-request and libudev are introduced by patches/120-v4l2-request.patch, | |
| # v4l2-request and libudev are introduced by patches/120-v4l2-request-*.patch, |
Generated by Claude Code
| @@ -0,0 +1,46 @@ | |||
| From ba6b36feb25bd8e622b87b28e5fca7e635dd6b5f Mon Sep 17 00:00:00 2001 | |||
There was a problem hiding this comment.
nit: ba6b36fe… is not a commit that exists, and the body below — FFmpeg 9's VVC decoder, the 6.1.2 update dropping the guard in 2024, this PR's own assembler output — cannot be text Rosen wrote in 2020. Attribute the rewritten header to yourself and credit the original guard (030-h264-mips.patch, added in 79afc370f) in the body instead.
Generated by Claude Code
📦 Package Details
Maintainer: @thess @antonlacon
(You can find this by checking the history of the package
Makefile.)Description:
Update to ffmpeg 9.x.
ffmpeg is a complex beast and every major release changes and breaks things.
Now we're going from 6.x to 9.x - and on top purposely introduced some design changes to the package.
This will inevitably break things!
Decisions:
libffmpeg-full,ffmpeg-customandlibffmpeg-audio-dec; there'slibffmpegandlibffmpeg-mininow, where the former defaults to what-fullprovided and can be fully customised,-ministays the same with a static set of build options,libffmpeg-fullandlibffmpeg-audio-decprovided as legacy aliases to not break packages right away. Only upstream package usinglibffmpeg-audio-decissqueezelite(maintained by @kel-mo), maybe that one can just depend onffmpeg(full)?The patches coming along with this PR are mostly either compile fixes or taken over from LibreElec.
It's obvious we're trying to lay grounds for getting Kodi running, so I'm not even trying suger-coat that some of the patches are exclusively for catering kodi.
🧪 Run Testing Details
✅ Formalities
If your PR contains a patch:
git am(e.g., subject line, commit description, etc.)
We must try to upstream patches to reduce maintenance burden.