#
#asahi-dev · 2026-06-27 07:52
leiothat sounds the opposite of path of least resistance, but hopefully good for power use :)
#asahi-dev · 2026-06-27 07:52
JamesCalligeros[m]yeah except some distros dont build v4l2codecs
#asahi-dev · 2026-06-27 07:51
leiodon't you get gstreamer use v4l2slh264dec automatically when using decodebin2 or decodebin3? That's what practically everything would actually be using, and then it'd work automatically for those
#asahi-dev · 2026-06-27 07:51
JamesCalligeros[m]meh doesnt matter right now, todays project has been brainstorming moving most of the garbage to firmware
#asahi-dev · 2026-06-27 07:50
JamesCalligeros[m]apparently it works with stateless if you have the ffmpeg request api patches applied, but i could not get it to work at all
#asahi-dev · 2026-06-27 07:49
sofusleio: i think firefox only supports the statefull v4l2
#asahi-dev · 2026-06-27 07:49
leiounder automatic I meant decodebin2/decodebin3
#asahi-dev · 2026-06-27 07:48
leiosounds like a job for the asahi meta packages and their equivalents there
#asahi-dev · 2026-06-27 07:48
JamesCalligeros[m]"oh yeah just install gst-plugins-bad but make sure your distro builds it with v4l2codecs so you get the v4l2slh264dec component and make sure your pipeline uses it and...."
#asahi-dev · 2026-06-27 07:47
JamesCalligeros[m]gstreamer worked fine, my point was more that we cannot ship something that requires users to do some magic pipeline incantation just to have working video decode.
#asahi-dev · 2026-06-27 07:46
leiogstreamer parts should be able to work seamlessly as well with v4l2, if they don't, maybe I can take a look at that; but it's not like many things use it, especially on plasma systems
#asahi-dev · 2026-06-27 07:45
leionot many things work seamlessly with vk either, so otherwise maybe option 1 with long-term option 3?
#asahi-dev · 2026-06-27 07:44
leioas then that sounds like path of least resistance to ship something useful
#asahi-dev · 2026-06-27 07:44
leiomight be something stupid that some firefox debug log reveals? https://bugzilla.mozilla.org/show_bug.cgi?id=1852765 had some log knobs asked for for debugging another chip
#asahi-dev · 2026-06-27 07:43
JamesCalligeros[m]hasnt worked at all in the past week of testing on any of my devices
#asahi-dev · 2026-06-27 07:43
leio(on some other socs)
#asahi-dev · 2026-06-27 07:42
leioare you sure firefox doesn't work with the ffmpeg patch? https://wiki.postmarketos.org/wiki/Hardware_video_acceleration#Usage suggests it might, saying the ffmpeg patch makes it work
#asahi-dev · 2026-06-27 07:36
JamesCalligeros[m]this is not a situation that we alone can fix, so the path of least resistance is to just do what everyone else already supports, which is vaapi or increasingly vk video
#asahi-dev · 2026-06-27 07:35
JamesCalligeros[m]and even though mpv uses it "well" it still wont actually pick it unless you force it to with -hwdec, which is also a problem because it wont automatically pick dmabuf-wayland as the output driver either, so you dont even get zerocopy by default
#asahi-dev · 2026-06-27 07:34
JamesCalligeros[m]yes, external patches.... that have not been merged.... and nothing that consumes ffmpeg other than mpv effectively uses....
#asahi-dev · 2026-06-27 07:34
leiofwiw, on gentoo you get the v4l2 request ffmpeg patches with USE=soc
#asahi-dev · 2026-06-27 07:32
JamesCalligeros[m]firefox requires a "proper" v4l2-m2m, not request aiui
#asahi-dev · 2026-06-27 07:31
leioby my reading firefox supported v4l2 for rpi in 116 and vk video in 153; maybe that's not request api or assumes ffmpeg patches though
#asahi-dev · 2026-06-27 07:30
JamesCalligeros[m]everything on the desktop assumes va-api or whatever the nvidia one is is present, and stuff that doesnt is going to very quickly move to vk video. look at how quick firefox was to adopt it vs v4l2.
#asahi-dev · 2026-06-27 07:29
JamesCalligeros[m]the v4l2 request api is not well supported by anything. ffmpeg requires patches that have not yet been merged (and it still doesnt handle it properly), firefox is not the be-all-end-all of video decode software, and no one is interested in constructing gstreamer pipelines by hand for desktop use
#asahi-dev · 2026-06-27 07:28
leiowhat for? Should be possible to use v4l2 directly for firefox for example, might need to fix some support bugs and such though
#asahi-dev · 2026-06-26 22:43
chadmedsadly i think we only really have three choices. given that it's abandoned, we could adopt it and fix it so that it works properly, wait and see if that vkvideo shim goes anywhere.... or write a drm/accel driver for the hw and port all the v4l2 bits to vk video in mesa
#asahi-dev · 2026-06-26 19:57
jannauI guess the format could change after all surfaces were destroyed
#asahi-dev · 2026-06-26 19:54
jannaudoesn't look like libva-v4l2-request will work with ffmpeg vaapi hwaccel. that requests at the beginning a 128 by 128 surface. surface creation in v4l2 is tied to the v4l2_format and libva-v4l2-request, explictly sets the format only once
#asahi-dev · 2026-06-26 19:20
yukahaven't had time to look into it more closely yet
#asahi-dev · 2026-06-26 19:20
yukabut it didn't work for me when I tried it on t8132 earlier
#asahi-dev · 2026-06-26 19:20
yukahttps://github.com/IntegralPilot/linux2/commits/dc-serial-test/
#asahi-dev · 2026-06-26 19:19
enverbalalicyuka: is the dockchannel uart linux driver you mentioned pushed anywhere ?
#asahi-dev · 2026-06-26 18:39
sofusi think the piodma thingy is only for getting things from dart->cm3
#asahi-dev · 2026-06-26 18:39
chaos_princessmaybe its for that?
#asahi-dev · 2026-06-26 18:39
sofusno
#asahi-dev · 2026-06-26 18:39
chaos_princessare you using the piodma thingy
#asahi-dev · 2026-06-26 18:38
sofust8103 had different values
#asahi-dev · 2026-06-26 18:37
sofusthey are used many places in the instruction stream
#asahi-dev · 2026-06-26 18:36
sofusthis is very weird https://github.com/sofus13/linux/blob/cbac5555e6c4200d48e57e0191e2c9e1741fb756/drivers/media/platform/apple/avd/avd-inst.h#L8
#asahi-dev · 2026-06-26 18:23
jannausvc has no mainstream applications as far as I'm aware
#asahi-dev · 2026-06-26 18:21
sofusthat makes sense, is svc support nice? I could get some support by simply changing the slice size
#asahi-dev · 2026-06-26 18:16
jannauthe strange dart errors I saw with the iommu restore experiment on t6002 should be explained by the wrong dart compatible
#asahi-dev · 2026-06-26 18:14
sofusalthough it would be nice if i could make the reset "smart" so it does not reset mid decode
#asahi-dev · 2026-06-26 18:13
sofusthats what im using
#asahi-dev · 2026-06-26 18:13
jannauhave you tried fluster / gstreamer? https://github.com/fluendo/fluster
#asahi-dev · 2026-06-26 18:08
sofusif we run conformance suite we cant use ffmpeg because thats also partially broken
#asahi-dev · 2026-06-26 18:06
sofusmulti slice should actually be broken on t8103
#asahi-dev · 2026-06-26 18:05
sofusi dont think a lot is missing. I think the worst offenders are the reset function and register poll when doing multi slice
#asahi-dev · 2026-06-26 17:57
jannauhas already someone looked into what's needed to use avd with bootlin's libva-v4l2-request? looks dead upstream but fedora carries it with patches

plus ›