2026-07-07
00:08hdbngr has quit [Ping timeout: 480 seconds]
01:36hdbngr has joined #asahi-dev
01:45hdbngr has quit [Ping timeout: 480 seconds]
02:00yurix has quit [Remote host closed the connection]
02:00yurix has joined #asahi-dev
02:00nnym has quit [Read error: No route to host]
02:00nnym has joined #asahi-dev
02:01enverbalalic has quit [Quit: enverbalalic]
02:20ddxtanx has joined #asahi-dev
02:30ddxtanx has quit [Remote host closed the connection]
02:33ddxtanx has joined #asahi-dev
03:15mripard has quit [Remote host closed the connection]
03:32hdbngr has joined #asahi-dev
03:41hdbngr has quit [Ping timeout: 480 seconds]
03:42hdbngr has joined #asahi-dev
03:44marvin24 has quit [Ping timeout: 480 seconds]
03:55nora has joined #asahi-dev
03:57nora_ has quit [Ping timeout: 480 seconds]
03:58hdbngr has quit [Ping timeout: 480 seconds]
04:01nicolas17 has quit []
04:13nicolas17 has joined #asahi-dev
04:21mattia013 has joined #asahi-dev
04:29ddxtanx has quit [Remote host closed the connection]
04:29eluks has quit [Remote host closed the connection]
04:30eluks has joined #asahi-dev
04:32ddxtanx has joined #asahi-dev
04:38hdbngr has joined #asahi-dev
04:47hdbngr has quit [Ping timeout: 480 seconds]
05:00hdbngr has joined #asahi-dev
05:08hdbngr has quit [Ping timeout: 480 seconds]
05:32thunfisch has joined #asahi-dev
05:50hdbngr has joined #asahi-dev
05:53nickchanthe pcie-apple.c is directly derived from corellium codes but setup_port() is in probe in the original corellium pcie driver; it must be changed to init at some point in upstreaming
05:59hdbngr has quit [Ping timeout: 480 seconds]
05:59Cardpiraya has quit [Quit: ZNC 1.9.1 - https://znc.in]
06:03Card has joined #asahi-dev
06:06lion328 has quit [Ping timeout: 480 seconds]
06:16mattia013 has quit [Remote host closed the connection]
06:18jannauI wasn't really expecting different result considering the other cores are sitting in tight loops while mmu_secondary_setup() runs. cache invalidation after enabling mmu and caches felt more natural. has dc ivac even an effect when caches are disabled?
06:29yukajannau: init is for global structures of the driver/module, probe is for per-device initialization? and the ports are per pcie controller
06:32yukathe port setup being in init() doesn't cause any problems in practise, because we don't implement tearing down and bringing back up the pcie ports
06:32yukabut if we had code to tear down a port and bring it back up again, I would expect to be able to unbind and bind the driver and have it repeatedly probe successfully
06:37hdbngr has joined #asahi-dev
06:37pb17 has quit [Ping timeout: 480 seconds]
06:38svenI think I’ll just reduce it to dsb sy and leave a comment about the cache thing if this happens again
06:38pb17 has joined #asahi-dev
06:42jannauyuka: apple_pcie_init() is called in per-device context. it's called via struct pcie_ecam_ops.init and is unrelated to module init
06:45mripard has joined #asahi-dev
06:48hdbngr has quit [Ping timeout: 480 seconds]
06:51hdbngr has joined #asahi-dev
06:59hdbngr has quit [Ping timeout: 480 seconds]
07:00alessaba has quit [Remote host closed the connection]
07:00alessaba has joined #asahi-dev
07:05hdbngr has joined #asahi-dev
07:13svenactually, I’ve spent enough hours debugging this crap. let me just merge it and be done with it :D
07:23nafod has quit [Quit: Ping timeout (120 seconds)]
07:23nafod has joined #asahi-dev
07:38jannauI think it was unrelated. the issue with switching boot cpu is that it is really only safe just before leaving m1n1 since it doesn't update the status property in the cpu nodes in the ADT
07:45sofuschaos_princess: i force-pushed a fix, please also enable the DEBUG_INST* in case it does not work, should be unrelated to t8103
07:46sofusi cant get gstreamer to select the right format if i add more than one, so i disabled that for now
07:59chaos_princesscrowd_run_1080X512_fr30_bd8_8buf_l3 works
07:59chaos_princessi also tried crowd_run_4088X2168_fr60_bd8_odd_size_l51 and got this https://pastebin.com/9yjHnSkA
07:59chaos_princessyes, t6000
08:08svenyuka: yeah, that could've been it then. it only reproduces when all the stars align which kinda fits to my memory ordering / stale cache line from speculation theory
08:33mripard has quit [Remote host closed the connection]
08:36sofuschaos_princess: one of the buffers just needed a bit more space, i will tests a few more bit streams before updating
08:46chaos_princesstesting on t8103 is kinda complicated by the lack of dts nodes :P
08:46chaos_princess(don't worry, i can fix it on my side)
09:20chaos_princessgetting a [ 113.443263] avd 269080000.avd: Frame processing timed out! Vp: 3 (00)
09:20chaos_princess(h264 works)
09:23sofuschaos_princess: try sed -i 's/0x2b000000/0x2bfff000/g' drivers/media/platform/apple/avd/avd-{drv,vp9}.c
09:27chaos_princessworks
09:32leioyuka: things looking good for t8132 in general that might allow with some patch gathering plus own writing and testing to get basics going on a j715ap relatively straightforward?
10:00leioyuka: something to run a distro on and access over SSH, rest is bonus, can write and/or test some patches for dts or m1n1 hookup, but not drivers at this point (time-wise)
10:04chaos_princesss/ssh/serial/ :P
10:04chaos_princessand you would want t8132 support here, otherwise it is going to be painful https://github.com/AsahiLinux/m1n1/blob/main/src/cpufreq.c
10:14alessaba has quit [Ping timeout: 480 seconds]
10:15chaos_princesskb+display should be there i think as we have dockchannel + boot framebuffer
10:16leioactually, if no network, it might as well hang on the same table then, so usb serial can be fine too
10:20alessaba has joined #asahi-dev
10:27chaos_princessis wired network ok?
10:28yukabut since the thing does not even boot reliably, I would not consider it useful for any real workload
10:30chaos_princessppp over serial
10:31leioso basically it could be in a state of some alpine image working after some dts or whatever patches, and then might need to get hands more dirty from there, is what I gather from this
10:34leioI'll go evaluate my time and not ask further until I actually have the laptop next to me with some basics set up :)
11:06chaos_princesssofus: a question about vp slots - are they independent? can they be used in parallel? or what are they for?
11:09JamesCalligeros[m]macos never uses them in parallel supposedly
11:09sofusi think they are independent, they can be used in parallel, im not sure if they give any performance boost / penalty
11:10JamesCalligeros[m]We’d have to track some amount of state in the cm3 too
11:16sofusin the adt's there is an decode-samples-per-seconds parameter. All SoC's have the same one expect for neo and base m1, and thoose have the least amount of vp's, so i would think some performance can be gained
11:27chadmedyeah but how do we want to handle it? my approach seemed a little too stateful to fit nicely with v4l2sl/vk vid/vaapi
11:28chadmedi dont see a way to keep the whole block fully stateless with the current design. maybe exposing "avd" as a media controller and then each vp as its own decoder?
11:37chadmedthats basically what i was going for with the cm3 also dispatching jobs to "idle" vps but idk how much state is too much
11:42sofusideally it would deal with the vp's like the driver. We would use the piodma and try to keep as litle as possible in cm3 memory
12:59yurix has quit [Read error: Connection reset by peer]
12:59yurix has joined #asahi-dev
13:25lion328 has joined #asahi-dev
15:19mattia013 has joined #asahi-dev
15:34floklit6040 with 27.0_26A5378j still has the locked register. DFU restore had to be done on MacOS (see https://github.com/libimobiledevice/idevicerestore/issues/786 for details)
16:34hailey8 has joined #asahi-dev
16:36kevans91 has quit [synthon.oftc.net larich.oftc.net]
16:36ciara has quit [synthon.oftc.net larich.oftc.net]
16:36croissant has quit [synthon.oftc.net larich.oftc.net]
16:36MikeM64 has quit [synthon.oftc.net larich.oftc.net]
16:36madmatt067 has quit [synthon.oftc.net larich.oftc.net]
16:36Calandracas has quit [synthon.oftc.net larich.oftc.net]
16:36tg47 has quit [synthon.oftc.net larich.oftc.net]
16:36elvishjerricco has quit [synthon.oftc.net larich.oftc.net]
16:36ipatch_ has quit [synthon.oftc.net larich.oftc.net]
16:36Z750 has quit [synthon.oftc.net larich.oftc.net]
16:36Pokemod977362084712818330 has quit [synthon.oftc.net larich.oftc.net]
16:36junefish has quit [synthon.oftc.net larich.oftc.net]
16:36codingkoopa3218 has quit [synthon.oftc.net larich.oftc.net]
16:36josiedotlol_ has quit [synthon.oftc.net larich.oftc.net]
16:36steev has quit [synthon.oftc.net larich.oftc.net]
16:36arnd has quit [synthon.oftc.net larich.oftc.net]
16:36msteffen has quit [synthon.oftc.net larich.oftc.net]
16:36_alice__ has quit [synthon.oftc.net larich.oftc.net]
16:36daniels has quit [synthon.oftc.net larich.oftc.net]
16:36sskras has quit [synthon.oftc.net larich.oftc.net]
16:36NekomimiScience has quit [synthon.oftc.net larich.oftc.net]
16:36tamar has quit [synthon.oftc.net larich.oftc.net]
16:36linxz has quit [synthon.oftc.net larich.oftc.net]
16:36annan_person has quit [synthon.oftc.net larich.oftc.net]
16:36fgcrl has quit [synthon.oftc.net larich.oftc.net]
16:36Son_Goku has quit [synthon.oftc.net larich.oftc.net]
16:36Graypup_ has quit [synthon.oftc.net larich.oftc.net]
16:36violette has quit [synthon.oftc.net larich.oftc.net]
16:36dcavalca8569797667858944 has quit [synthon.oftc.net larich.oftc.net]
16:36remexre has quit [synthon.oftc.net larich.oftc.net]
16:36hailey has quit [synthon.oftc.net larich.oftc.net]
16:36vulpes2 has quit [synthon.oftc.net larich.oftc.net]
16:36Lightsword has quit [synthon.oftc.net larich.oftc.net]
16:36opoku_ has quit [synthon.oftc.net larich.oftc.net]
16:36opticron has quit [synthon.oftc.net larich.oftc.net]
16:36nukelet0 has quit [synthon.oftc.net larich.oftc.net]
16:36hailey8 is now known as hailey
16:37opticron has joined #asahi-dev
16:37croissant has joined #asahi-dev
16:37Lightsword has joined #asahi-dev
16:37kevans91 has joined #asahi-dev
16:37nukelet04 has joined #asahi-dev
16:37ciara has joined #asahi-dev
16:37madmatt067 has joined #asahi-dev
16:37MikeM64 has joined #asahi-dev
16:37Calandracas has joined #asahi-dev
16:37tg47 has joined #asahi-dev
16:37Z750 has joined #asahi-dev
16:37elvishjerricco has joined #asahi-dev
16:37junefish has joined #asahi-dev
16:37steev has joined #asahi-dev
16:37ipatch_ has joined #asahi-dev
16:37Pokemod977362084712818330 has joined #asahi-dev
16:37codingkoopa3218 has joined #asahi-dev
16:37arnd has joined #asahi-dev
16:37msteffen has joined #asahi-dev
16:37_alice__ has joined #asahi-dev
16:37sskras has joined #asahi-dev
16:37linxz has joined #asahi-dev
16:37daniels has joined #asahi-dev
16:37NekomimiScience has joined #asahi-dev
16:37tamar has joined #asahi-dev
16:37annan_person has joined #asahi-dev
16:37fgcrl has joined #asahi-dev
16:37Son_Goku has joined #asahi-dev
16:37dcavalca8569797667858944 has joined #asahi-dev
16:37violette has joined #asahi-dev
16:37Graypup_ has joined #asahi-dev
16:37remexre has joined #asahi-dev
16:37nukelet0 has joined #asahi-dev
16:37vulpes2 has joined #asahi-dev
16:37opoku_ has joined #asahi-dev
16:37josiedotlol_ has joined #asahi-dev
16:38nukelet0 has quit [Ping timeout: 480 seconds]
16:45svenhttps://github.com/AsahiLinux/m1n1/pull/618 and the m1n1 side, not interested in detailed reviews yet
16:53yukaThere are tales about a gpio which can detect the presence of an sd card. Does anyone know which gpio this is or how to find it?
16:58chaos_princesson j314 routed to smc pmu, so probably in smc gpio space, whatever 'event_name-bit14' is
17:28ScottJones[m]**m1n1 hypervisor — newest bootable macOS guest on t6021 (M2 Max)?**
17:28ScottJones[m]boots as a guest on an M2 Max? Is 13.7.8 expected to work, or is 13.5 genuinely the hard ceiling for this chip?
17:28ScottJones[m]The docs say the hv targets Ventura 13.5, but Apple has pulled 13.5 from softwareupdate and the swcdn InstallAssistant link now 403s — the oldest Ventura I can still fetch from Apple is 13.7.8 (22H730). I saw commit 874ff59e adding EL2-only mrs/msr noop-patching for macOS 14.8.3, but it's chip-gated to the M3 family (0x8122/0x6030–0x6034), so my t6021 takes the unpatched path. On current m1n1, what's the newest macOS that actually
17:31ScottJones[m]lol, that's where I'm headed. :D
17:32ScottJones[m]Also, USB-C DP Alt Mode now working here on a t602x laptop (MBP 14" M2 Max, j414c) — 4K@60 10-bit via the fairydust delta on 7.0.13. #456 has t602x laptops as broken, so I posted the full recipe + dmesg there: https://github.com/AsahiLinux/linux/issues/456#issuecomment-4904584668
17:32ScottJones[m]indeed, I expected that!
17:34chaos_princessScottJones[m]: someone on the *internet* might have *archived* the old versions of installassistant.pkg
17:37ScottJones[m]I'll try the newest version just for our information. :D
17:49NickChan[m]reserving the stack for spin table has always felt been a bit of a hack to me
17:50NickChan[m]ideally it would have been jumping into asm that doesn't use stack but that if everything use psci, that's another way to get rid of the reserved stack
17:57nickchan64K is technically supported in pre-M1 but I do avoid it in Linux since it would try to make a 512 MiB contiguous reservation for cma or whatever
17:58pjv has joined #asahi-dev
18:16hardy94 has joined #asahi-dev
18:17Deewiant_ has quit [Remote host closed the connection]
18:18Deewiant has joined #asahi-dev
18:21sveni adjusted the ld script and as long as m1n1 is loaded with 64k alignment everything should work I think
18:25hardy94 has quit [Quit: hardy94]
18:35svenhttps://github.com/AsahiLinux/linux/commits/efi-psci/ and that's what i'll send as a RFC in a few days
18:37sven8 files changed, 261 insertions(+), 10 deletions(-) <-- i'm still surprised how small that diffstat ended up
19:00__Tetris__6 has joined #asahi-dev
19:01riker77 has quit [Quit: Quitting IRC - gone for good...]
19:05__Tetris__ has quit [Ping timeout: 480 seconds]
19:05__Tetris__6 is now known as __Tetris__
19:05mischa85 has joined #asahi-dev
19:09boubou has joined #asahi-dev
19:09boubou is now known as Guest12860
19:19pb17 has quit [Remote host closed the connection]
19:19pb17 has joined #asahi-dev
19:22pjv has quit [Ping timeout: 480 seconds]
19:23riker77 has joined #asahi-dev
19:32midou has quit [Ping timeout: 480 seconds]
19:56r0ni has joined #asahi-dev
20:00nickchan has quit [Quit: ZNC 1.9.1+deb2+b3 - https://znc.in]
20:00nickchan has joined #asahi-dev
20:23__Tetris__6 has joined #asahi-dev
20:27flokliOn the topic of 'we dump a .cpio file onto the ESP and ask initrd to load it without verifying anything', this might not fly with upstream, at least if we think about going further with secure / measured boot topics.
20:28__Tetris__ has quit [Ping timeout: 480 seconds]
20:28__Tetris__6 is now known as __Tetris__
20:29flokliWe might need to think about some signature mechanism, and just appending to initrd won't fly in any case. The hard part is that the installer runs in the MacOS context, not in the OS we're preparing to install for, or worse, regenerate the firmware files for.
20:30pb17 has quit [Ping timeout: 480 seconds]
20:40pb17 has joined #asahi-dev
20:45sveni think I took some from the cpuidle driver but I’d just treat them as math.randint for now
20:51yukamy power draw is still unexpectedly high, though I still have to compare that against cpuidle-apple (with everything else the same)
20:52svenhm, I’d expect psci-efi to be a bit worse because it has to switch pagetables unless I messed up somewhere
21:00JamesCalligeros[m]I still think it’s likely we need to do something explicit to shut the uncore parts of the clusters up when there are no cores running
21:03yukaok yeah the power draw is just as bad with cpuidle-apple, so it's something else in my kernel
21:16rag27[m] has joined #asahi-dev
21:35nela has quit [Quit: Ping timeout (120 seconds)]
21:36nela has joined #asahi-dev
22:32StanfordAppliedCyber[m]are we find to target 26.6 for the official asahi linux release on m4/a18 pro?
22:32StanfordAppliedCyber[m]asking because i'm going to start poking the gpu firmware abi and i don't want the effort to be wasted if we support another version
22:32chaos_princessunknown
22:33StanfordAppliedCyber[m]the only downside i can think of is the lack of tso but there's a decently high chance apple never enables it
22:34StanfordAppliedCyber[m]what would prevent 26.6 from working? i assume wfi is never going to be properly resolved but we're building that solution right now
22:34chaos_princessstill missing kvm-relevant reg on m4pro
22:35chaos_princessalso, i understand that gpu is flashy and all that, but have you considered doing the rest of the platform first
22:35chaos_princesslike, idk, cpufreq, pcie, usb
22:36chaos_princessnone of them need firmware abi
22:36JamesCalligeros[m]and are likely yolo-able to some degree
22:38StanfordAppliedCyber[m]yeah i was gonna do the others first
22:38StanfordAppliedCyber[m]but i like working on multiple things in parallel
22:39chaos_princessgpu will consume your entire life for months
22:39StanfordAppliedCyber[m]the first target is a working system without gpu acceleration, which i'll define as screen, keyboard, trackpad, wifi, and maybe some form of usb
22:39StanfordAppliedCyber[m]again, perhaps i'm hopelessly naive, but i'm not yet convinced of that
22:40StanfordAppliedCyber[m]maybe after actually trying i will indeed be convinced lol
22:42JamesCalligeros[m]please add speakers and webcam to that list
22:42JamesCalligeros[m]our bar for new soc releases is pmuch "everything except dcp and gpu"
22:42StanfordAppliedCyber[m]yeah that's fair
22:42chaos_princessthose should be pretty easy :P
22:42StanfordAppliedCyber[m]m3 is close to being there right?
22:42StanfordAppliedCyber[m]no promises but yeah i'll do those
22:42chaos_princessm3max is missing usb
22:42StanfordAppliedCyber[m]i'll try to do those
22:43StanfordAppliedCyber[m]by usb does usb2 work or do we need usb3 for release
22:43chaos_princessand mics i guess
22:43chaos_princessm3 pro/vanilla also missing isp, since i don't have those chips and idk how to statically determine certain values
22:45millefy14 has quit [Ping timeout: 480 seconds]
22:47StanfordAppliedCyber[m]i assume step 1 is to figure out what parts alrady work, ie, iirc yuka found that pcie is basically unchanged
22:47chaos_princesspmuch. Start throwing m3 stuff at it, and see what breaks
22:47chaos_princessor run under tracer and compare to what our drivers are doing
22:47StanfordAppliedCyber[m]what's the best place to find that? asahi-wip?
22:54millefy14 has joined #asahi-dev
23:01StanfordAppliedCyber[m]i think i'm the poster child for smp boot instability lol
23:01JamesCalligeros[m]keep adding sync barriers until it works :P
23:04yukaAlso, I think the impact of getting your hv into an upstreamable state would be much larger than frickling some drivers for the Neo
23:17flokli has quit [Quit: WeeChat 4.9.2]
23:17flokli has joined #asahi-dev
23:23StanfordAppliedCyber[m]oh yeah upstreaming the hypervisor is what i'm currently working on
23:23StanfordAppliedCyber[m]the main blocker for that, debugusb working, is also the blocker for linux dev
23:24StanfordAppliedCyber[m]re: hypervisor, i think the real accomplishment would be someone else successfully boooting macos under the hv
23:24StanfordAppliedCyber[m]since for linux dev i'm planning on doin g the stream the kernel over debugusb thing
23:24yukaI could give it a try if I don't have to debug too much, since I don't actually want to read the code
23:24StanfordAppliedCyber[m]give what a try?
23:25StanfordAppliedCyber[m]i'm worried about tainting you there,s because i can't just ship you a binary because about half the emulator is in python
23:26StanfordAppliedCyber[m]and given you volunteered to help me upstream this by clean rooming the tainted bits I REALLY do not want to taint you by accident
23:26StanfordAppliedCyber[m]dw once debugusb works i'll have it in upstreamable shape within a few days
23:27StanfordAppliedCyber[m]from there if you care i can get whatever logs/traces/whatever from my emulator that you could ever want for a clean room re project
23:27StanfordAppliedCyber[m]i think the emulator should be ~2k lines of code, my implementation is about 6k but that's because i chose not to reuse any of m1n1's utilities for the iommus
23:28StanfordAppliedCyber[m]or virtualize sptm, idrc, i think it's a higher risk approach but how yall get it to work it doesn't matter to me
23:33jnn has joined #asahi-dev
23:36jn has quit [Ping timeout: 480 seconds]
23:57mattia013 has quit [Remote host closed the connection]