#
…

2026-06-26

00:00hdbngr has joined #asahi-dev
00:08hdbngr has quit [Ping timeout: 480 seconds]
00:18pb17 has quit [Ping timeout: 480 seconds]
00:21pb17 has joined #asahi-dev
01:26integralpilot has joined #asahi-dev
01:27integralpilotyuka: huh, can you try writing (echo "hello" > /dev/ttydc0) and see if that crashes as well? would be interesting to see if its just an issue with reads on happens on both
01:27integralpilotin the meantime ill try and see if i can reproduce it
01:30integralpilotit looks to me like the regs might be wrong or for some weird reason on m4 it might need some power domains?
01:34integralpilothow did you get the regs?
01:36integralpilot has quit [Quit: integralpilot]
01:40hdbngr has joined #asahi-dev
01:48hdbngr has quit [Ping timeout: 480 seconds]
02:00yurix has quit [Remote host closed the connection]
02:00yurix has joined #asahi-dev
02:22tobhe_ has joined #asahi-dev
02:29tobhe has quit [Ping timeout: 480 seconds]
02:34hdbngr has joined #asahi-dev
02:43hdbngr has quit [Ping timeout: 480 seconds]
02:48hdbngr has joined #asahi-dev
02:57hdbngr has quit [Ping timeout: 480 seconds]
03:13nora_ has joined #asahi-dev
03:15nora has quit [Ping timeout: 480 seconds]
03:19diesller[m] has joined #asahi-dev
03:42hdbngr has joined #asahi-dev
03:51hdbngr has quit [Ping timeout: 480 seconds]
03:56marvin24 has quit [Ping timeout: 480 seconds]
04:08hdbngr has joined #asahi-dev
04:29eluks has quit [Remote host closed the connection]
04:29eluks has joined #asahi-dev
04:31hdbngr has quit [Ping timeout: 480 seconds]
04:34Ariadne_ has joined #asahi-dev
04:35Ariadne has quit [Remote host closed the connection]
04:39diesller[m]hello, i really want to help but i don't know where to start from. i don't have much coding experience but can i still try to contribute?
04:39diesller[m]i own an m4 air
04:39hdbngr has joined #asahi-dev
04:48hdbngr has quit [Ping timeout: 480 seconds]
05:00hdbngr has joined #asahi-dev
05:03jmax has joined #asahi-dev
05:08hdbngr has quit [Ping timeout: 480 seconds]
05:11jmax has quit [Ping timeout: 480 seconds]
05:11hdbngr has joined #asahi-dev
05:19hdbngr has quit [Ping timeout: 480 seconds]
05:31mattia013 has joined #asahi-dev
05:43hdbngr has joined #asahi-dev
05:52hdbngr has quit [Ping timeout: 480 seconds]
06:05yukaTried PMPTracer on t6034, it's crashing somewhere in ASCTracer: https://x0.at/3gyR.txt
06:07yukadon't have time to debug it now, just FYI or if anyone has had the same and has a solution :)
06:10yukaintegralpilot: from my adt I had the following regs to go from (0x388128000, 0x10000), (0x38810c000, 0x1000)
06:10gargantua has quit [Ping timeout: 480 seconds]
06:11yukayour device tree has 0x2e410c000, 0x2e4128000, 0x2e412c000
06:12yukaI autocompleted this to 0x38810c000, 0x388128000, 0x38812c000
06:26integralpilot has joined #asahi-dev
06:26hdbngr has joined #asahi-dev
06:28integralpilot has quit [Remote host closed the connection]
06:32sofus has joined #asahi-dev
06:32gargantua has joined #asahi-dev
06:35hdbngr has quit [Ping timeout: 480 seconds]
06:44hdbngr has joined #asahi-dev
06:47sofusjannau: i have only seen faults on stream 0, this only happens during resets?
06:50sofusi wonder if both avd instances can be run at the same time, did you test that (on m2 ultra)?
06:53hdbngr has quit [Ping timeout: 480 seconds]
06:53sofusand thanks for the pull, that makes it a lot easier
06:55integralpilot has joined #asahi-dev
06:55integralpilotyuka: that seems reasonable, strange then its not working
06:56integralpilotit might be something wrong with the driver then, i did only code it up very quickly just to test echo/cat didn't error out
06:56chaos_princessyuka: iova table turned out to only be a thing on t600x, everything else sets it to null address
06:57chadmedsofus: i would be concerned if they couldnt run simultaneously, they should be completely independent of each other
07:03sofusthats true, not sure why i thought they would be connected somehow
07:03jannausofus: yes, I tested and it works as expected. the 2 avd instances are completly separate
07:04integralpiloti can't manage to reproduce your error sorry, if i put a completetly bogus reg then i get a SError during probe, not only when "cat" called
07:06jannauthe faults happen during the reset and dart's irq handler tries to read from mmio which fails due to the reset
07:07integralpilot has quit [Quit: integralpilot]
07:08jannaure meson, I think we can get rid of the llvm cross file and just add one for gcc / binutils builds
07:10chadmedclang/llvm shouldnt need a cross toolchain, we just pass -target and -mcpu
07:14hdbngr has joined #asahi-dev
07:14jannauthe cross file != cross toolchain. the file tells meson what's build
07:14integralpilot has joined #asahi-dev
07:16chadmedyes, but typically you only need them to describe a cross toolchain, right?
07:16integralpilotjannau: the MCA nodes you pushed to asahi-wip for t8122 aren't right i think, for me it complains about not enough clocks and fails to probe
07:17integralpilotcomparing to t8112, i think its missing "<&nco 4>, <&nco 4>" will test now
07:20jannauintegralpilot: possibly, I haven't tested audio
07:20chadmedif your userspace is up to date, once thats fixed the speakers should Just Work
07:20chadmedplease lmk if they do not
07:21integralpilotim scared to cause i dont want them to blow up!
07:21integralpilotmaybe i will test headphone jack first
07:21chadmedthe driver will not let them run without speakersafetyd
07:21integralpilotwill outdated speakersafetyd be a problem? i think i still got the one from janurary
07:22chadmedit will yes, just update it, alsa-ucm-conf-asahi and asahi-audio and you should be good
07:22integralpilotthanks! will do
07:28hdbngr has quit [Ping timeout: 480 seconds]
07:35ipatch has quit [Ping timeout: 480 seconds]
07:41integralpilotokay so with the clock fix, now mca successfully probes, but macaudio doesn't yet, there's a new error
07:41integralpiloti'll see how to fix
07:44sofussorry im new to meson, the cross file is for telling meson what compiler and stuff should be used? why is that not needed for llvm then?
07:45hdbngr has joined #asahi-dev
07:54hdbngr has quit [Ping timeout: 480 seconds]
08:00hdbngr has joined #asahi-dev
08:08yuka has quit [Remote host closed the connection]
08:10yuka has joined #asahi-dev
08:14hdbngr has quit [Ping timeout: 480 seconds]
08:17hdbngr has joined #asahi-dev
08:26integralpilotit seems the i2c driver isnt probing???
08:33jannauhas the node still `status = "disabled";`?
08:35integralpilotin t8122.dtsi yep, but then we do status = "okay" in t8122-jxxx.dtsi
08:35integralpilotbut from the live system /sys/firmware/devicetree/base, it shows as disabled when i decompile it
08:37enverbalalic has joined #asahi-dev
08:39chadmedare you looking at the right i2c bus? i2c0 should be disabled
08:40integralpilotspeakers on i2c1 and i2c3
08:40integralpilotand they are showing disabled from /sys/firmware/devicetree/base but enabled (seemingly) in what i build
08:40integralpilotso it seems m1n1 is disabling?
08:40chadmedor the dtbs werent actually updated but surely not
08:45enverbalalic has quit [Ping timeout: 480 seconds]
09:20hdbngr has quit [Ping timeout: 480 seconds]
09:21sofus has quit [Quit: Lost terminal]
09:21integralpiloti found why! m1n1 does dt_disable_missing_devs("i2c", "i2c@", 8) which is marking our i2c nodes are missing and then disabling them
09:23integralpilotwe might have the addresses wrong in the dt then?
09:23integralpilotor m1n1 might be triggering a false positive
09:23integralpilotill find out which
09:26sofus has joined #asahi-dev
09:26hdbngr has joined #asahi-dev
09:35hdbngr has quit [Ping timeout: 480 seconds]
09:36enverbalalic has joined #asahi-dev
09:43hdbngr has joined #asahi-dev
09:48enverbalalic has quit [Ping timeout: 480 seconds]
09:50integralpilotOMG i think we have the i2c addresses WRONG
09:51integralpilotthe actual prefix is 0x2a1 but we're using 0x235 (from M2)
09:57integralpilotoh no the wrong i2c addresses are upstream...
10:09gargantua has quit [Ping timeout: 480 seconds]
10:10gargantua has joined #asahi-dev
10:12enverbalalic has joined #asahi-dev
10:15hdbngr has quit [Remote host closed the connection]
10:15hdbngr has joined #asahi-dev
10:21integralpilotafter fixing i2c now it probes but we get a new error
10:22integralpilotthe GPIO pin numbers introduced in this commit don't seem to be right (at least for j613), its causing a timeout because now the amplifiers are now: https://github.com/AsahiLinux/linux/commit/60181ffe850018ab0dfd8b6c23da3a929fa0dd8a
10:22integralpilot*amplifiers are down, as GPIO doesn't wake them up
10:22integralpilotill try to find right one
10:25hdbngr has quit [Ping timeout: 480 seconds]
10:28integralpilotnope actually checked adt and 13 and 17 (from that commit) seem to be right
10:28integralpilotit was just the i2c addresses we got wrong
10:32scd has quit [Ping timeout: 480 seconds]
10:36DD65 has quit [Remote host closed the connection]
10:36DD65 has joined #asahi-dev
10:39hdbngr has joined #asahi-dev
10:43integralpilotOH WAIT the reason its not working now is because the i2c interrupt numbers are wrong too!
10:43integralpilotso the address AND interrupt is wrong for i2c
10:44integralpilotnot sure how the wrong ones got there, seems to be copy from m2 and then forget to change
10:46chadmedsofus: just tested t6k, sorry for the delay. we dont seem to ever get the interrupts at the application cores
10:47integralpilotnow it works!!
10:47integralpilotaplay -l shows speakers and jack as output
10:47integralpilotmoment of truth, am going to try and play smth now
10:50hdbngr has quit [Ping timeout: 480 seconds]
10:51sofuschadmed: no worries :) did you test using https://github.com/sofus13/linux/tree/avd-m1-test
10:55chadmedconformance test video frame, broken output
10:55chadmedno dart errors or anything like that though
10:57sofusthe link is broken
10:57sofusbut there is output?
10:57chadmedits just green
10:58sofusdo you have output from the first frame?
11:04sofusmaybe something is not turned on? It looks like the vp does nothing
11:05chadmedmaybe, theres no irq coming to the application core at all
11:06chadmedbut it looks like the cm3 is booted
11:07sofusyeah, you just ran the 'old-style' test where irq's are turned off and everything is done through polling registers
11:07sofusits weird you dont get timeouts
11:08enverbalalic has quit [Ping timeout: 480 seconds]
11:09chadmedit times out on your normal avd branch of course
11:10enverbalalic has joined #asahi-dev
11:12hdbngr has joined #asahi-dev
11:13mini has quit [Quit: The Lounge - https://thelounge.chat]
11:13mini has joined #asahi-dev
11:18integralpilotokay so audio has a new error now, "failed to reach power state" for the mca2 power domain
11:19integralpilotany idea why? i'm a bit stuck, i checked into all the possibilities i thought and they aren't right
11:21hdbngr has quit [Ping timeout: 480 seconds]
11:22hdbngr has joined #asahi-dev
11:26yukamaybe compare the power domains with the ones from t6030/t6031
11:26yukasince afaik audio is confirmed working there
11:26chaos_princessintegralpilot: put externally-clocked on it
11:32enverbalalic has quit [Max SendQ exceeded]
11:32enverbalalic has joined #asahi-dev
11:36integralpilotchaos_princess: thank you so much, that fixed it!
11:36integralpiloti don't see that error anymore, now time for the (second) moment of truth to see if it works
11:39JamesCalligeros[m]Which output do you get? Only AppleJ6xx or do you see the virtual pipewire sink too?
11:40integralpilotaplay -l shows only AppleJ613 (but 2 of them)
11:40JamesCalligeros[m]Yeah one PCM is the speakers and the other is V/ISENSE
11:40JamesCalligeros[m]alsa is useless on these machines, dont use it directly. Use pipewire/wireplumber only
11:41integralpilotspeakersafetyd panicked and crashed with "/usr/share/speakersafetyd/apple/j613.conf" no such file or directory
11:42JamesCalligeros[m]Actually no it’s there. Can you install from HEAD?
11:42integralpilotsure, gimme a sec
11:42JamesCalligeros[m]The FAR repos may not yet have the latest
11:42JamesCalligeros[m](your invitation to install gentoo :p)
11:51sofusi think maybe the dart compatible is wrong for t6000
11:51sofusi think it should be t8110
11:52sofusnot sure if its supported, no other nodes use it
11:53chadmedi would find it really odd if only avd used t8110 dart
11:54integralpilotwell, this is Apple we're talking about
11:54integralpilotreally odd is kinda their thing
11:54chadmedtrue, but i dont recall t8110 being in the adt for it
11:55sofusit is, and in a few other things
11:55chadmedoh for fucks sake
11:55chadmedwhat do you know :P
11:55chadmedit works!
11:55sofusOh yeah!!
11:55sofusman i remember filling out the dt and thinking thats odd...
11:55chadmedhl1 kleiner voice "thats odd"
11:57chadmedHUA still doesnt work possible vp offset issue again?
11:58integralpilotnew error now for speakersafetyd: unexpected argument -b found?
11:58chadmedhellsing ultimate abridged, the other h264 video i was testing last night
11:59sofusis this still on the test branch?
11:59chadmedintegralpilot: your systemd unit file is for an old version with the black box, just invoke it on the tty with speakersafetyd -c /usr/share/speakersafetyd/conf -m 7
11:59chadmedsofus: yea still the m1 test branch
11:59sofustry one the normal one
11:59chadmedsorry, -c /usr/share/speakersafetyd/
11:59integralpilotokay its started now, thanks!
12:00integralpilotand youtube videos do not make sound :(
12:00integralpilotwhat could possibly be the issue now arghh!! will look into it
12:01chadmedsofus: bingo, we got it!
12:01sofusamazing!
12:01integralpilotyay for avd!!
12:01integralpilotgreat work guys
12:01chadmedintegralpilot: output of wpctl status pl0x
12:04integralpilotClients: (bunch of stuff: uresourced, WirePlumber, pipewire, plasmashell, xdg-desktop-portal, PipeWire ALSA, Firefox, etc.), Devices: 52. Build-in Audio. Sinks: 54. Built-in Audio Stereo [vol: 0.90], 75. Michael's MacBook Pro [vol. 0.80] - this is an MBA!! but my MBP is turned on and nearby. Sources: 55. Built-in Audio Mono [vol 1.00]. Streams: 68. Firefox 66. output_FL [active]
12:04chadmedok it didnt instantiate the filter chain for some reason
12:05chadmedalso the volume should not be 90% on the builtin sink, it should always be 100%
12:05chadmeddid you grab asahi-audio from HEAD too?
12:06integralpilotnope, can try to
12:07chadmedyeah that should do it, seems like none of the updates have landed in fedora yet
12:07chadmedasahi-audio, speakersafetyd and alsa-ucm-conf-asahi from git and you should be golden
12:11sofuschaos_princess: yeah, im gonna double check the dt entry and the offsets
12:13Tachyonray[m]is there a m1n1 branch for t8122?
12:14integralpilotits supported in mainline m1n1
12:15Tachyonray[m]how would i get that installed? I got some errors in the asahi installer
12:15chaos_princesswhy are you trying to install it? dev work, or end user?
12:16integralpilotheadphones work now but main speakers still no
12:17Tachyonray[m]i tried reading the dev docs but didn’t get very far. also compiled the installer locally. It told me my model was unsupported
12:17chaos_princesswell, yes, it is.
12:17chaos_princesswe are not shipping asahi on that to users
12:18chaos_princesscheck pull requests on asahi-installer repo, that will help you
12:18Tachyonray[m]ideally I’d also make an asahi gentoo install usb and boot that
12:19integralpilotinstall usbs aren't rly how install on apple silicon works
12:21Tachyonray[m]Right, it’s been a while since i installed on my m1, I probably misremember booting it from usb at some point
12:23integralpilotchadmed: speakersafetyd now panics with "Left tweeter: Negative power, bad ivsense data?"
12:23jannauinstall from usb works as in install m1n1 + u-boot EFI-only and then boot from USB storage. no USB support on any M3 devices yet though
12:24chadmedintegralpilot: booting from a usb with m1n1 is the _only_ supported way to install gentoo actually
12:25integralpilotoh, interesting
12:25chadmedre speakersafetyd, can you bump the loglevel to debug and pastebin whatever it outputs
12:26sofusi remember installing it without, cant remember how tho
12:26chadmeddid you use the scuffed genstrap script lol
12:27chadmedi should just delete that from the repo, it's a crime against humanity
12:27sofusmaybe...? I remember using arch to do it
12:27chadmedyeah it wouldve been that
12:27chadmedi was younger and much stupider (if you can believe thats even possible)
12:27jannauTachyonray[m]: yes, but you're too early. what you're trying to do still requires a lot manual work. there are no install images which support m3 yet
12:29integralpilotif you want to do it manually, basically: 1. make 14.latest partition, 2. compile m1n1 and leave in place accessible by recoveryOS (i.e. no FileVault) 3. boot into 1TR with 14.latest paired and use kmutil to set the boot object to your m1n1
12:29chaos_princessnot .latest, 14.8.3
12:29integralpilotah yeah, right, thanks
12:30integralpilotim technically on 14.7 actually cause i did it before 14.8.3 released i think
12:30integralpilotbut 14.7 isnt signed anymore, so if doing fresh you need 14.8.3
12:31chaos_princessshould be fine as long as you aren't doing dcp/gpu
12:31jannaum3 pro macbook pro with m1n1 v1.6.0 boots now standalone, trackpad does not work due to an issue with smc gpios. I suspect the kernel
12:31chaos_princessintegralpilot: is signed/unsigned even matters on laptops
12:32integralpilotoh yeah i guess you can get an archive somewhere if really wanted 14.7
12:32integralpilotand could install it
12:32jannaus2idle seems to be broken. it doesn't wake up. could be an smc issue as well
12:32fl0_id_afkchaos_princess maybe not for stub, but iirc for full / hv install it does not finish if not singed, right?
12:33chaos_princessthere are ways to install basically every version you want, some more annoying than others
12:33integralpilotit only panics after you try and play smth
12:34jannaupower button works in principle
12:34Tachyonray[m]jannau: i could experiment with catalyst a bit make a bootable stage3 i think
12:34integralpilothow's macsmc-power? i think i remember when i was writing it i tested on my t6030 too
12:34integralpilotoh also you need firmware for trackpad
12:34chadmedintegralpilot: did it play any sound at all?
12:34integralpilotlike you need to grab it urself and copy it into the folder (that's what i had to do)
12:35integralpilotso if you haven't done that, that's why trackpad probably isnt working
12:35integralpilot*re-writing it
12:35chadmedin the devicetree, can you swap the v and i slots for each codec
12:35integralpilotyep. will try and report back, thanks
12:35chadmed-10 W is... not right
12:36chadmedoh oh oh
12:36chadmedis your kernel based on 7.1 by any chance?
12:36Tachyonray[m]integralpilot: this sounds a bit more complicated. hope to try it soon. so far haven’t been able to get 1TR installed
12:36integralpilotchadmed: 7.0, its the current asahi-wip i believe (with my changes to fix i2c and audio)
12:37chadmedoh ok yeah try swapping the vi slots then
12:37chadmedthought it mightve been the bus keeper masks but that's a 7.1 problem
12:37integralpilotTachyonray[m]: 1TR isnt smth you installed, its just on your mac if you force hold the power button from a clean shutdown
12:37integralpilotit stands for One True Recovery
12:38donaldh has quit [Ping timeout: 480 seconds]
12:40donaldh has joined #asahi-dev
12:42integralpilotchadmed: after swapping, heard a tiny sound before speakersafetyd died again, same error "Left tweeter: Negative power, bad ivsense data?"
12:42integralpilotthough, now all the W values are +ve, so that's good ig
12:43integralpilotwill make new pastebin
12:45chadmedno i think this is more wrong, the voice coil should not be 120 degrees lmao
12:45integralpilotmaybe its only some are the wrong way around and not all?
12:45integralpilotso switching all around fixed some things but broke others
12:51integralpilotoooh after checking ADT what happened is they were flipped and also the left tweeter and right woofer got the wrong way around
12:54integralpilotnope. after doing that still wrong.
12:56chadmedyou may need to capture the raw samples from the VISENSE PCM and see what's there
12:57chadmedthe tas2764 datasheet has the info you need to do the samplewise multiplication to get the voltage and current
12:57chadmed(speakersafetyd also does this so you can also just look at the code there)
12:58integralpilotthanks, might have a try tomorrow
12:58chadmedyeah its late
12:59chadmedyou could also make sure the volume is *very very low* and comment out the check for -ve power and see what happens
12:59chadmedthink of how funny it will be trying to explain to mymac flinders st that you melted your speakers
12:59yukaI think I had the thing where there was a bit of sound every time I restarted speakersafetyd, and it was in my case caused by not updating alsa-ucm-conf
13:00integralpilot has quit [Quit: integralpilot]
13:01chadmedif alsa-ucm-conf-asahi is not version 9 then yeah it will cause issues
13:03sofusi apparently cant write dt entries, t8122 should also be fixed now...
13:04chadmedit's annoying but imagine if we had to write ASL and build ACPI tables instead
13:15elvishjerricco has quit [Remote host closed the connection]
13:25donaldh has quit [Ping timeout: 480 seconds]
13:26sofusanyone wants to test avd on m3? All offsets/registers should be ok
13:28lynndotpy72 has quit [Quit: bye bye]
13:29nela has quit [Quit: bye!]
13:29lynndotpy72 has joined #asahi-dev
13:32yukaif you can recap how to test it, I can do it later on t6034
13:37chaos_princesswill throw it on a t6031 in a bit
13:41sofusmy branch actually does not have the m3 dt's, should i merge them or would that get messy? The avd node should be almost the same as the one on t8122
13:43chaos_princessso, i was actually testing a cherry-pick of everything relevant in your branch on top of asahi-wip.
13:43chaos_princessIf you can rebase it on that - would be even better
13:46chadmedsofus: your branch is trivial to rebase on top of current asahi-wip, its how ive been testing
13:51sofusyuka: i think the easiest way is using ffmpeg patched with v4l2 stateless support (i think this is the default if you are using gentoo)
13:53sofusok im gonna try a rebase, thanks
13:58yukapresumably also some kernel portions and device tree nodes?
13:58yukaI found your branch
13:59yukathe compiled fw is loaded like other firmware?
13:59chaos_princessjust put it in /lib/firmware/apple/
13:59yuka"/lib/firmware/apple": No such file or directory (os error 2)
13:59chaos_princessit will complain in dmesg about which specific firmware it wants
13:59yukabut it's okay I can operate my computer ;)
14:00yuka(nixos does these things a bit differently)
14:17nela has joined #asahi-dev
14:18gargantua has quit [Ping timeout: 480 seconds]
14:30sofusim not sure if i did it the correct way https://github.com/sofus13/linux/tree/avd-wip
14:31sofusjannau actually made compiling (and getting the right fw version) a lot easier, look at the last commit message
14:44enverbalalic has quit [Ping timeout: 480 seconds]
14:45enverbalalic has joined #asahi-dev
14:48jannaufor testing with hv / chainload CONFIG_EXTRA_FIRMWARE / CONFIG_EXTRA_FIRMWARE_DIR is nice
14:58sofusi have added device tree entries for t603x now
15:02hdbngr has quit [Ping timeout: 480 seconds]
15:09jannausofus: t6030 regs need + 0x1000_0000, parent_address is 0x2_1000_0000
15:09jannauis compatible meant to be "apple,t8122-avd"?
15:11jannauyes, it the other compatibles in the driver are t8132 and t6040
15:14sofusjannau: thanks, im not sure where i could see that?
15:14sofust6031 works??? nice!
15:14jannau"/arm-io/ranges" in the ADT
15:27sofusi think t8132 is the neo and not m4
15:28jannausofus: looks ok. t6030 works with https://paste.centos.org/view/a06b333a
15:28donaldh has joined #asahi-dev
15:30sofusjannau: nice! may i ask why you changed the compatibles?
15:30jannaua18 pro is t8140
15:32sofust8132 was a guess based on the cm3 tunables, im thinking that guess was wrong
15:33readyee[m] is now known as rhlvista[m]
15:34jannausofus: that's how we label most of our compatibles. first the actual SoC compatible and then the compatible the driver already knows (if those differ)
15:39sofuswhops, the device trees need a cleanup then :p
15:41elvishjerricco has joined #asahi-dev
15:42jannaunot a big issue, we want it that way for upstream. fine as is for testing
15:45sofusbut this is great, m3 was the last device where the registers changed
15:51jannauhas anyone tested t8122?
16:04hdbngr has joined #asahi-dev
16:08sofust8122 could need the AVD_QUIRK_LSR flag
16:12hdbngr has quit [Ping timeout: 480 seconds]
16:18mx08 has quit [Quit: WeeChat 3.8]
16:21enverbalalic has quit [Ping timeout: 480 seconds]
16:28enverbalalic has joined #asahi-dev
16:29hdbngr has joined #asahi-dev
16:36enverbalalic has quit [Ping timeout: 480 seconds]
16:38hdbngr has quit [Ping timeout: 480 seconds]
16:40hdbngr has joined #asahi-dev
16:42jannausofus: same parent_address issue on t8122, works with https://paste.centos.org/view/1e94e63d
16:44enverbalalic has joined #asahi-dev
16:45marvin24 has joined #asahi-dev
16:49hdbngr has quit [Ping timeout: 480 seconds]
16:49hdbngr has joined #asahi-dev
16:53rhysmdnz has quit [Quit: Bridge terminating on SIGTERM]
16:53Guest12006 has quit [Quit: Bridge terminating on SIGTERM]
16:57rhysmdnz has joined #asahi-dev
16:58hdbngr has quit [Ping timeout: 480 seconds]
16:58Guest12079 has joined #asahi-dev
17:00enverbalalic has quit [Read error: Connection timed out]
17:15hdbngr has joined #asahi-dev
17:23hdbngr has quit [Ping timeout: 480 seconds]
17:24enverbalalic has joined #asahi-dev
17:32enverbalalic has quit [Ping timeout: 480 seconds]
17:39enverbalalic has joined #asahi-dev
17:47enverbalalic has quit [Ping timeout: 480 seconds]
17:48yuka has quit [Remote host closed the connection]
17:49sofusjannau: thanks
17:50hdbngr has joined #asahi-dev
17:53yuka has joined #asahi-dev
17:56jannauhow far are you away from where it would make sense merging avd (h264 only) into asahi-wip? we should run the conformance suite on each SoC instead of testing with random videos to determine that it works
17:57jannauhas 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
17:59hdbngr has quit [Ping timeout: 480 seconds]
18:05sofusi dont think a lot is missing. I think the worst offenders are the reset function and register poll when doing multi slice
18:06sofusmulti slice should actually be broken on t8103
18:06pb17 has quit [Ping timeout: 480 seconds]
18:08sofusif we run conformance suite we cant use ffmpeg because thats also partially broken
18:13jannauhave you tried fluster / gstreamer? https://github.com/fluendo/fluster
18:13sofusthats what im using
18:14sofusalthough it would be nice if i could make the reset "smart" so it does not reset mid decode
18:16jannauthe strange dart errors I saw with the iommu restore experiment on t6002 should be explained by the wrong dart compatible
18:21sofusthat makes sense, is svc support nice? I could get some support by simply changing the slice size
18:22pb17 has joined #asahi-dev
18:23jannausvc has no mainstream applications as far as I'm aware
18:37sofusthey are used many places in the instruction stream
18:38sofust8103 had different values
18:39chaos_princessare you using the piodma thingy
18:39chaos_princessmaybe its for that?
18:39sofusi think the piodma thingy is only for getting things from dart->cm3
18:44hdbngr has joined #asahi-dev
18:49enverbalalic has joined #asahi-dev
18:49enverbalalic has quit []
18:59hdbngr has quit [Ping timeout: 480 seconds]
19:10ipatch has joined #asahi-dev
19:11ipatch has quit []
19:11ipatch has joined #asahi-dev
19:18enverbalalic has joined #asahi-dev
19:19enverbalalicyuka: is the dockchannel uart linux driver you mentioned pushed anywhere ?
19:20yukabut it didn't work for me when I tried it on t8132 earlier
19:20yukahaven't had time to look into it more closely yet
19:34enverbalalic has quit [Ping timeout: 480 seconds]
19:35enverbalalic has joined #asahi-dev
19:52hdbngr has joined #asahi-dev
19:54jannaudoesn'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
19:57jannauI guess the format could change after all surfaces were destroyed
20:01hdbngr has quit [Ping timeout: 480 seconds]
20:13gargantua has joined #asahi-dev
20:34gargantua has quit [Ping timeout: 480 seconds]
20:37sofus has quit [Quit: Lost terminal]
20:53hdbngr has joined #asahi-dev
21:02hdbngr has quit [Ping timeout: 480 seconds]
21:20Yushi has quit [Remote host closed the connection]
21:21Yushi has joined #asahi-dev
21:54hdbngr has joined #asahi-dev
22:03hdbngr has quit [Ping timeout: 480 seconds]
22:21mattia013 has quit [Remote host closed the connection]
22:42nicolas17 has joined #asahi-dev
22:43chadmedsadly 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
22:44gargantua has joined #asahi-dev
22:55hdbngr has joined #asahi-dev
23:04hdbngr has quit [Ping timeout: 480 seconds]
23:05nst has quit [Quit: WeeChat 4.7.2]
23:10nst has joined #asahi-dev
23:13alessaba has quit [Ping timeout: 480 seconds]
23:34enverbalalic has quit [Quit: enverbalalic]
23:45revadam has joined #asahi-dev
23:46revadam has quit []
…