#
✕ nick:jannau
#asahi · 2026-10-06 11:15
jannauI think dropped that dependency on t603x
#asahi · 2026-10-06 11:13
jannauit wasn't from the beginning, I changed that after looking at t602x-pmgr.dtsi
#asahi · 2026-10-06 11:06
jannauI think I might saw that on some M3 SoC as well before turning ps_apcie_phy_sw always-on and use ps_apcie_sys_st* as nvme power-domains
#asahi · 2026-10-02 18:39
jannauMac Pro / M2 Ultra. The exact SoC (M1 and M2 generation) will not make a difference. I don't think anyone tested M3 and later but changes in the behavior would be surprise
#asahi · 2026-10-02 18:32
jannauI re-tested a rx 7600 xt today but it resulted in kernel panics due to a SError. probably an unaligned access in the GPU's PCI device memory
#asahi · 2026-09-27 15:56
jannauin xe/amdgpu. I'd recommend running under the hypervisor for testing
#asahi · 2026-09-27 15:56
jannauif anyone feels to inclined to comment on the Mac Pro GPU reddit thread: with https://github.com/NVIDIA/open-gpu-kernel-modules/pull/972#issuecomment-3591497063 and an simple IRQ related fixup a geforce 3070 was working. Xe and amdgpu had at that time clear 16k page size issues. Those are probably resolved in the meantime thanks to loongson. Not sure if there are unaligned access mines
#asahi · 2026-09-14 08:51
jannauMary: I haven't seen that. not sure if worth investing time into debugging this as the 7.2 downstream kernel will switch to the upstream gpuvm rust abstractions
#asahi · 2026-09-09 13:42
jannauthis was only fixed days before the M3 announcement
#asahi · 2026-09-09 13:42
jannaumyria: you need asahi-installer and rebuild the firmware package, either via asahi-fwupdate or rebuild vendor firmware in the installer (if you need the "-a" variant)
#asahi · 2026-09-08 14:07
jannauI don't follow and you still seem to be confused. The dev installer (by CI design) has never a plain version like vX.Y.Z
#asahi · 2026-09-08 14:01
jannaufl0_id_: what do you mean? it's "v0.9.1" vs. "v0.9.0-1-g99dff2e". The latter is actually the same commit but was built before v0.9.1 was tagged
#asahi · 2026-09-08 13:45
jannauexisting installs can be made macOS 27 compatibility with installing asahi-fix27 and running `sudo asahi-fix27 --confirm`. That toggles a flag inside the stub apfs (2.5GB large partition) to make it "visible" to macOS 27 startup disk/boot picker
#asahi · 2026-09-01 20:08
jannaualso requires usb-pd to enable it so a little more involved setup
#asahi · 2026-09-01 20:07
jannaushouldn't it log to the debugusb uart? then kisd should work
#asahi · 2026-08-31 19:29
jannaustrange, I can't imagine what could have caused this
#asahi · 2026-08-31 19:04
jannauJPeisach: wroks with the asahi debug allocator as well. The minor number is random and just depends on the probe order
#asahi · 2026-08-31 18:58
jannauJPeisach: works here on t8103 (with fairydust rebased onto asahi-7.1.12-1). pushed that fairydust branch to github
#asahi · 2026-08-31 18:50
jannauJPeisach: I'd recomend against using the debug allocator. not that I think it's likely it is the cause
#asahi · 2026-08-31 18:48
jannaudoes `journalctl -k -o short-monotonic --no-hostname -g \.gpu:` look similar?
#asahi · 2026-08-31 18:46
jannauhow is it broken? programs using llvmpipe/lavapipe? does it probe correctly?
#asahi · 2026-08-31 18:43
jannaushouldn't be an issue anymore downstream. build is expected to be broken if rust is enabled but not available
#asahi · 2026-08-29 08:24
jannauTye[m]: waht kind of error? does it print the URL it tried to download from? if yes, can you manually download it (to rule out network issues)
#asahi · 2026-08-28 13:02
jannauand embedded in iboot1 are the SMC and PMU firwmares
#asahi · 2026-08-27 13:41
jannauthat will remain separate and probably driven a kms bridge, same for the DP part of atc
#asahi · 2026-08-27 13:39
jannaudcp's are essentially the crtc devices. unfortunately more than that because all other things KMS abstracts away also have to go through the DCP firmware
#asahi · 2026-08-27 13:37
jannauwe have that already with the meta device display-subsystem
#asahi · 2026-08-27 13:36
jannaukms already allows N:M crtcs to connector mappings and we have (and can) to support that
#asahi · 2026-08-26 07:16
jannaufedora either already has or will have within hours asahi-audio-4.1
#asahi · 2026-08-25 14:17
jannauyou might have to press a key. it's esc on fedora after u-boot displays booting from ...
#asahi · 2026-08-25 11:27
jannaunoopwafel: without or with usb devices? if with which kind of devices? storage, input, network, hubs?
#asahi · 2026-08-24 15:48
jannaumaybe search for a less common word? https://oftc.catirclogs.org/asahi-dev/search?q=samsung works for me
#asahi · 2026-08-24 13:37
jannauor any other program reading switch status from the input event
#asahi · 2026-08-24 13:35
jannaumacsmc-input already reads the initial state. I think `evtest --query "/dev/input/event2" "EV_SW" "SW_LID" && echo "lid open" || echo "lid closed"` should be reliable when booting with the lid closed
#asahi · 2026-08-24 13:06
jannauacpi button has/had /proc/acpi/button/lid/LID0/state but that is deprecated and not reliable, see https://docs.kernel.org/5.10/firmware-guide/acpi/acpi-lid.html
#asahi · 2026-08-24 13:05
jannautobias^: the input device can query the current state. macsmc-input probably needs to query the state on probe to make that reliable
#asahi · 2026-08-24 12:34
jannauI don't think hwmon has a sensor type for this so lid status over hwmon would probably not be upstreamable
#asahi · 2026-08-24 12:31
jannautobias^: upower has lid status (afaik cached from close/open events)
#asahi · 2026-08-21 06:34
jannaustop
#asahi · 2026-08-20 18:31
jannauthe dtb are ok when DP-altmode works
#asahi · 2026-08-20 18:30
jannaudid you use the config from 6.19.14-400.asahi.fc42 as base? you're probably missing CONFIG_MACSMC_POWER, that module got renamed in 7.1 based kernels
#asahi · 2026-08-20 18:15
jannauis this a fedora-asahi-linux install? what does `ls -ld /boot/dtb*` output?
#asahi · 2026-08-20 17:17
jannauveloek: with kde plasma? It hides the battery indicator by default if the battery is fully charged
#asahi · 2026-08-17 21:14
jannauand possibly bits/250-aop. rebase was painless but I expect that those drivers will require a non-trivial amount of fixes. gpu rework will pull in rust changes expected to land in 7.3
#asahi · 2026-08-17 21:12
jannaurebase onto v7.2 was overall not that painful if you can get around without GPU driver
#asahi · 2026-08-17 21:10
jannauit's much less painful to rebase the bits/* branches individually though
#asahi · 2026-08-17 21:09
jannauyou probably want 'git rebase --onto v7.2 v7.1.8'
#asahi · 2026-08-17 21:08
jannauyou're trying to rebase linux-stable commits as well?
#asahi · 2026-08-08 12:46
jannauthat's what the conmformance samples are for. each has a description which feature(s) are tested
#asahi · 2026-08-08 12:43
jannauAnEagle[m]: re avd it is too early. there are still unexplained failing conformance samples

plus ›