#asahi-dev · 2025-10-12 07:24
jannausure. it needs hw access or at least someone able to trace macos#asahi-dev · 2025-10-11 19:49
jannauthere are devlinks added for this but I don't understand why those are created and why they are not in sync_State despite probing "simple-mfd"#asahi-dev · 2025-10-11 19:33
jannausync_state should be reached once the driver loads. not sure if this is an issue specifically with "simple-mfd" or if it is a different issue#asahi-dev · 2025-10-11 19:31
jannauthe cross pmgr pmdomains never reach sync_state() due to the added pmgr_* supplier role#asahi-dev · 2025-10-11 19:30
jannauit get's a little annoying in 6.17 with https://lore.kernel.org/all/20250701114733.636510-1-ulf.hansson@linaro.org/#asahi-dev · 2025-10-11 19:25
jannauI don't think this is something we want (or need)#asahi-dev · 2025-10-11 19:25
jannaucross pmgr pmdomains behave a little strange. for example ps_afi from pmgr which is used by pmgr_south:ps_afnc4_ioa (t600x). For some reason it also gets added as supplier for pmgr_south#asahi-dev · 2025-10-09 19:01
jannaudon't the smc sub-drivers depend/select macsmc anyway? I guess you can still `rm macsmc.ko`#asahi-dev · 2025-10-08 16:31
jannaufl0_id: this looks like it's still in m1n1. I don't remember what the expected output of linux.py is but I wouldn't be surprised if this crash is in m1n1's cpufreq code#asahi-dev · 2025-10-07 15:01
jannaumy assumption was/is that the command stream is per frame or slice#asahi-dev · 2025-10-07 14:57
jannaudidn't eiln bypass the decoder firmware completely? That would make the decoder stateless. It also makes the "driver" more complex so that moving as much code as possible outside of the kernel a good idea. All from memory, it has been a while since I looked at the python PoC code#asahi-dev · 2025-10-07 14:27
jannaukeeping code out of the kernel which doesn't need to be in the kernel is another compelling argument against v4l2#asahi-dev · 2025-10-07 14:26
jannauI wouldn't worry too much about vulkan video (or vaapi). Once there is a working driver supporting the other interface as well is a much smaller problem. It might be either through zink or by adding vaapi support to the user space part of the driver#asahi-dev · 2025-10-07 11:32
jannauI thought so as well and I only asked them to be removed for the misguided MFD_CELL_OF conversion#asahi-dev · 2025-10-07 11:26
jannau6.19. far too late for 6.18#asahi-dev · 2025-10-05 14:34
jannauI guess we don't really care that much for adt.py as that's just used for parsing/printing#asahi-dev · 2025-10-05 14:28
jannauwe can't rely on the size either since dart-aop.dapf-instance-0 has a length of 728 which is 13 * 56 == 14 * 52#asahi-dev · 2025-10-05 14:27
jannauthe adt is even inconsistent over macos versions. m3's dart-pmp.dapf-instance-0 is 1344 (24 * sizeof(DAPFT8110B)) in 15.1 vs. 1248 (24 * sizeof(DAPFT8110)) in 26.0#asahi-dev · 2025-10-05 14:23
jannau0x40 in dart-options unfortunately does not determine the dapf-instance layout#asahi-dev · 2025-10-05 13:44
jannauwhy does m1n1/setup.py call .reset_panic_counter() twice? ended up duplicated due to a merge conflict?#asahi-dev · 2025-10-05 13:42
jannauI have them since ages, I was waiting on getting your refactoring in and when forgot#asahi-dev · 2025-10-05 13:29
jannauonly difference to the m4 version (except for the SoC) is that it still uses cd321x (the i2c usb type-c port controller)#asahi-dev · 2025-10-05 13:23
jannauI guess I should take this opportunity to submit the t8122 chickens and other m1n1 hw support#asahi-dev · 2025-10-05 13:22
jannauno reboots on the m3 macbook pro on tahoe (>4 minutes and counting)#asahi-dev · 2025-10-05 12:18
jannauinstalls tahoe on the m3 macbook pro#asahi-dev · 2025-10-05 12:13
jannaubut which co-proc would that be which isn't present or differently configured on the m4 mac mini#asahi-dev · 2025-10-05 08:44
jannauoh, that's unexpected. then it's definitively not wdt#asahi-dev · 2025-10-05 08:33
jannauand possibly get reset_panic_counter.py working to avoid going through recovery every X reboots#asahi-dev · 2025-10-05 08:33
jannaucoprocessors like smc, pmgr, aop, sep, ... might have their own watchdog and/or can trigger resets on their own. I still think the best way forward is to look at the fault-info bits of the spmi pmu#asahi-dev · 2025-10-05 08:21
jannauI don't think it's the watchdog as in the wdt device from the ADT#asahi-dev · 2025-10-05 08:17
jannauchaos_princess: you had me confused for a moment#asahi-dev · 2025-10-05 08:16
jannauno, I was traveling last week and m4 is not a priority for me#asahi-dev · 2025-10-05 08:12
jannauunderhood31: t8103-j293.dtb#asahi-dev · 2025-10-03 07:37
jannautruita: with sonoma you wouldn't have to deal with the nvme change/issue. you probably need to create an usb installer from installassistant to install it if you don't have a sonoma macos install anymore#asahi-dev · 2025-10-02 19:11
jannaukettenis: releases/1.5.1 should be fixed. I'll tag in the next few days#asahi-dev · 2025-10-02 07:59
jannauyes, that did it#asahi-dev · 2025-10-02 07:56
jannauhopefully fixed by https://github.com/AsahiLinux/m1n1/pull/502#asahi-dev · 2025-10-02 07:55
jannauI think the CI issue is that the CI image has still rust 1.89 while rustup's stable toolchain is already rust 1.90#asahi-dev · 2025-10-01 15:47
jannaurustup seems to changed behavior and broke the github CI. rust-std apparently doesn't install core#asahi-dev · 2025-09-30 20:05
jannauiso's are a little complicated because you still need to install things to boot them#asahi-dev · 2025-09-30 20:04
jannauthe installer packages are disk images#asahi-dev · 2025-09-30 20:03
jannauTSO might work if apple's hypervisor exposes it#asahi-dev · 2025-09-30 19:52
jannauit should display thinks on the boot framebuffer. on desktop devices this is on the HDMI port#asahi-dev · 2025-09-30 19:10
jannauI already had everything except for the C wrapper#asahi-dev · 2025-09-30 19:10
jannauthe special function was easy to do and I was working on the rust adt code anyway#asahi-dev · 2025-09-30 19:09
jannauordering the entries in reverse chronical order seems like cleaner solution#asahi-dev · 2025-09-26 13:28
jannauI don't see much sense in reporting values without knowing what they measure#asahi-dev · 2025-09-26 13:27
jannauchadmed: you wouldn't try all fourCCs, see macsmc_gpio_find_first_gpio_index() which does a binary search through the known key indices#asahi-dev · 2025-09-26 12:56
jannauthe DC charger input?#asahi-dev · 2025-09-26 12:06
jannaunot 100% sure if the framebuffer is used. wait 2 minutes to see whether it reboots