#
…

2026-07-24

00:33Calandracas_ has joined #asahi-alt
00:40Calandracas__ has quit [Ping timeout: 480 seconds]
01:33enverbalalic has quit [Ping timeout: 480 seconds]
02:00ciara has quit [Remote host closed the connection]
02:00jnn has joined #asahi-alt
02:03jn has quit [Ping timeout: 480 seconds]
02:13ciara has joined #asahi-alt
06:22picnoir1 has quit [Ping timeout: 480 seconds]
06:31picnoir has joined #asahi-alt
08:21Calandracas_ has quit [Remote host closed the connection]
08:21Calandracas__ has joined #asahi-alt
08:43jelly has quit [Ping timeout: 480 seconds]
09:49enverbalalic has joined #asahi-alt
10:05kode54 has quit [Quit: The Lounge - https://thelounge.chat]
10:05kode54 has joined #asahi-alt
10:15jelly has joined #asahi-alt
11:02kitlith_ has joined #asahi-alt
11:05kitlith has quit [Ping timeout: 480 seconds]
11:06kitlith_ is now known as kitlith
11:33enverbalalic has quit [Ping timeout: 480 seconds]
11:35enverbalalic has joined #asahi-alt
11:40Tachyonray[m]doas update-m1n1 Mounted System ESP /dev/nvme0n1p4 at /run/.system-efi cat: '/usr/src/linux/arch/arm64/boot/dts/apple/*.dtb': No such file or directory
11:41Tachyonray[m]Do I need to put the dtb ‘s in there manually?
11:41JamesCalligeros[m]You didnt build your kernel properly
11:42sofusif its gentoo it could also just be a broken/missing symlink
11:42Tachyonray[m]gentoo, I did write an ebuild based on asahi-kernel that uses the asahi-wip branch
11:42JamesCalligeros[m]Then you messed it up :p
11:43JamesCalligeros[m]I was gonna say i literally just tested 7.0.13 last night on a bunch of machines before i pushed it
11:43enverbalalic has quit [Ping timeout: 480 seconds]
11:43JamesCalligeros[m]Did you copy the little bit that symlinks the dtbs back to /usr/src/linux-${KV_FULL}
11:44Tachyonray[m]JamesCalligeros[m]: (did get a verify-sig failure on the 7.0.13_p1 ebuild, but still investigating)
11:45JamesCalligeros[m]Ugh was it the asahi diff file?
11:46JamesCalligeros[m]Sometimes rebuilding the manifests doesnt work for some reason, i think it’s the diff being saved to distfiles as the same name for sources and kernel
11:46JamesCalligeros[m]It should be the exact same diff though….
11:46Tachyonray[m]JamesCalligeros[m]: This is the ebuild I used https://paste.gentoo.zip/ydUNnsVc, I have to still check some things
11:47JamesCalligeros[m]Ill have a look when im at a computer
11:47Tachyonray[m]JamesCalligeros[m]: It’s on my stream right now twitch tachyonray 😇
11:47JamesCalligeros[m]The full build log might shed some light
11:50Tachyonray[m]Here’s the build log for asahi-kernel-7.0.13-p1 https://paste.gentoo.zip/mpfXkTgR
11:50enverbalalic has joined #asahi-alt
11:52Tachyonray[m]<JamesCalligeros[m]> "Ill have a look when im at a..." <- Thank you, no rush
11:54sofusTachyonray[m]: does /usr/src/linux point to the correct kernel?
11:55Tachyonray[m]It did not, I just set it to the asahi-wip kernel. It’s confusing to me that it automatically changes the eselect kernel when there are updates. Makes the world update fail many times
12:00sofusif use USE="symlink" on the kernel it sets it automaticly
12:00sofustry the update-m1n1 again, it might just work. Btw, where you able to get avd to run?
12:01Tachyonray[m]Ah, I should probably remove that from make.conf and set it only for the kernel I want as default
12:01Tachyonray[m]sofus: Yes, it worked this time, thanks
12:11enverbalalic has quit [Ping timeout: 480 seconds]
12:15chadmedwhy is that failing to verify the base kernel patch....
12:15chadmedcan you clear out your distfiles and try again?
12:20Tachyonray[m]chadmed: I just did, no difference
12:20chaos_princessdo you gave verify-sig or sth
12:22Tachyonray[m]Yes, and it worked previously
12:22Tachyonray[m]`[ebuild NS ] sys-kernel/asahi-kernel-7.0.13_p1:asahi-7.0.13_p1::asahi [7.0.12_p1:asahi-7.0.12_p1::asahi] USE="initramfs modules-sign strip verify-sig -debug -hardened -savedconfig -secureboot -test" 0 KiB`
12:22chaos_princessconsider not having it
12:24Tachyonray[m]it works without verify-sig (obviously)
12:25chadmedi am remanifesting the package to see if anything is sus
12:26chadmedbut given that it was a k.o patch thats not something we can control
12:26enverbalalic has joined #asahi-alt
12:27Tachyonray[m]sofus: haven’t been able to do more testing with the avd. you mentioned I need to install something and I don’t know what. I just booted in the asahi-wip kernel and tried the video decoder
12:30chadmed-DIST linux-7.x-sha256sums-20260609.asc 5215
12:30chadmed+DIST linux-7.x-sha256sums-20260609.asc 7546
12:30chadmedcan you resync and try again
12:31Tachyonray[m]the signature was corrupted?
12:33Tachyonray[m]works like a charm again, thanks a lot
12:36Tachyonray[m]Tachyonray[m]: probably just forgot to run ebuild … manifest
12:42chadmedthe diff clearly shows the file was there and changed
12:42Tachyonray[m]so the file changed upstream?
12:43chadmedeither that or when i pkgdev manifest-ed the first time it got corrupted somehow
12:43chadmedit's happened before with github archives but not kernel.org
12:45madmatt0677 has joined #asahi-alt
12:50madmatt067 has quit [Ping timeout: 480 seconds]
12:50madmatt0677 is now known as madmatt067
12:52sofusTachyonray[m]: build this or download from actions https://github.com/AsahiLinux/avd-fw
12:53sofusput all the avd-fw*.bin files in the apple vendor dir
12:54qyliss has quit [Quit: bye]
12:56qyliss has joined #asahi-alt
13:01Tachyonray[m]Thanks, I’ve done so. Will check after reboot
13:23Tachyonray[m]There’s also some version mismatch shenanigans with the m1n1 ebuild I don’t understand https://paste.gentoo.zip/d3LMKjf7
13:23Tachyonray[m]When I tried yesterday it complained about a glibc version mismatch, now it’s hashbrown
13:29nela has quit [Quit: bye!]
13:30nela has joined #asahi-alt
13:39chadmedyeah its because of buildstd, if you set up the rust target override then you can just take out BUILDSTD=1 from the ebuild
15:40enverbalalic has quit [Ping timeout: 480 seconds]
15:42enverbalalic has joined #asahi-alt
17:52enverbalalic has quit [Remote host closed the connection]
18:02Tachyonray[m]sofus: I got the firmware to load after some struggles. make produced a blob for a different arch, but I changed AVD_VER to 2 and AVD_TIER to 0 for my M1. Added it to the initramfs with a dracut config file. The media players I tried it with gave errors though about libva and vaapi. Maybe I need to recompile ffmpeg without those use flags? https://paste.gentoo.zip/MUu1zhe7
18:03chaos_princessyou need to pass flags to use v4l, not vaapi
18:13chaos_princesswhich use flags do you have on ffmpeg
18:13Tachyonray[m]ffmpeg -hwaccel v4l2request -i four-colors.mp4 -f null -
18:13Tachyonray[m](see also the previous paste)
18:14chaos_princessit does not have that
18:14chaos_princessit does not contain your ffmpeg use flags
18:15Tachyonray[m]Sorry, I misuderstood
18:15stintelv4lrequest requires USE=soc
18:17chaos_princessflags look sane
19:15sofusTachyonray[m]: im pretty new to giving support, so please ask instead of struggling, avd_ver 2 is correct if you have an base m1 (not pro+). It looks like mpv used vulkan instead of v4l2, trying hwdec=v4l2request might be worth a shot?
19:18Tachyonray[m]I’m not expecting support, just glad the work is being done. I did try the hwdec=v4l2request, it prints Failed waiting on CAPTURE buffer 7… Think it’s a user space issue though
19:19Tachyonray[m]If I can provide valuable feedback I’m happy to do so, otherwise I’ll just be patient a while
19:22Tachyonray[m]How can I try that? I downloaded the files but not sure what to do with them
19:24sofusffmpeg -hwaccel v4l2request -v verbose -i AUD_MW_E.264 output.mp4
19:30Tachyonray[m]It looks like it worked: Input stream #0:0 (video): 100 packets read (54828 bytes); 100 frames decoded; 0 decode errors;  🎉
19:32sofusim not sure, you also had Device /dev/video1 is missing mem2mem capability
19:33chaos_princessthey have avd on 0 and isp on 1
19:35sofusyeah thats right, should be ok then! might just have been unlucky with the video
19:36Tachyonray[m]Yes, looks like I sourced a bad example video. The sample you sent was decoded with the avd
19:37sofusamazing! Its actually ffmpeg's fault ;)
19:37sofus(likely)
19:39sofusffmpeg/mpv works well with vp9 so i recomend using that
19:42Tachyonray[m]Awesome, Big Buck Bunny in vp9 decoded fine
20:13sofusyeah, vp9 might need some love... does it look weird for all vp9 videos?
20:14Tachyonray[m]Transcoding smaller videos works fine, but high bitrate gives errors
20:14sofusmakes sense, the higher bitrate ones are kinda wip. I was not sure if it was supported or not :p
20:15sofusi guess theres a reason apple rejects them :)
20:16Tachyonray[m]Amazing progress! This was not an easy feat to reverse engineer I imagine
20:17sofushevc support should be pretty limited, unless you are using gstreamer
20:18sofuslucky for me i could start with eiln's stuff, but polishing took some time
20:18sofusfun fact: i thought it would take me one to two weeks top :p
20:24Tachyonray[m]Probably got 80% done in that time, and the other 20% took 80% of the time ;-)
20:34sofussurprisingly not actually, it has felt more exponential. But i suppose i measured my progess in tests and not how much code i wrote
21:11Yushi has quit [Remote host closed the connection]
21:12Yushi has joined #asahi-alt
21:28chadmedonce you get past the "stare at traces until your face goes numb" part it becomes a lot more fun
21:29chadmedbut it is all very much a "oh cool im 90% of the way there, now for the other 140%" type activity
…