#asahi-dev · 2026-07-18 05:04
yukaI guess I need to copy that file to my linux partition to save myself rebooting into macOS#asahi-dev · 2026-07-18 05:03
yukarhlvista[m] on t6050 there is cluster-type = M (marketing name: performance core) and cluster-type = P (marketing name: super core)#asahi-dev · 2026-07-18 05:00
yukaWhoops okay#asahi-dev · 2026-07-17 21:51
yukaand this other die has its own pmgr?#asahi-dev · 2026-07-17 21:51
yukaLooking at the devices in those other pmgr instances, it seems like the t6050 is indeed a multi die device, in the sense that the gpu lives on another die#asahi-dev · 2026-07-17 21:48
yukapmgr_adt2dt on J714s and J714c gives identical output#asahi-dev · 2026-07-17 21:46
yuka(this is the output of pmgr_adt2dt: https://termbin.com/tsxn)#asahi-dev · 2026-07-17 21:43
yukaa bit of a certain upcoming device leaked into these adts#asahi-dev · 2026-07-17 21:43
yukaI mean according to the adts, j714c (M5 Max) also has 36 CPUs, so lol#asahi-dev · 2026-07-17 21:41
yukaRunning pmgr_adt2dt on the j714s adt produces 6 pmgr nodes (pmgr0-pmgr5), some of which have die offsets. This does not make any sense for M5 Pro which is a single die device right?#asahi-dev · 2026-07-17 20:28
yukabut base M5 is Tupai :)#asahi-dev · 2026-07-17 20:28
yukaor maybe it'll be M6, who knows#asahi-dev · 2026-07-17 20:28
yuka* probably then they will brand it as M5 Ultra, whatever#asahi-dev · 2026-07-17 20:28
yuka(source on M4 Ultra: https://www.iclarified.com/98172/apples-m4-ultra-chip-tipped-for-new-mac-pro-after-being-presumed-canceled-report)#asahi-dev · 2026-07-17 20:27
yukasome comments on your variant of the initial t8142 support: hidra is the codename for the upcoming M4 Ultra, and siq_cfg doesn't do anything since the chickens setup never runs on these devices#asahi-dev · 2026-07-17 20:26
yukaAh, you were the one who mentioned my spmi4 PR in a commit msg#asahi-dev · 2026-07-17 20:24
yukapanic counter reset should work in the feature/t8142 branch, as it includes https://github.com/AsahiLinux/m1n1/pull/626#asahi-dev · 2026-07-17 20:16
yukahttps://github.com/AsahiLinux/m1n1/pull/629#asahi-dev · 2026-07-17 20:14
yukawith the feature/t8142 branch#asahi-dev · 2026-07-17 20:13
yukaout of the box pretty much#asahi-dev · 2026-07-17 20:13
yukaYes that should work#asahi-dev · 2026-07-17 20:13
yukaDepends on what you meany by hypervisor..#asahi-dev · 2026-07-17 20:12
yukaon M4 (t8132), it crashes with the latest 27.0 beta but works with the latest 26.6 beta#asahi-dev · 2026-07-17 20:11
yukaif msr((3, 5, 15, 1, 3), mrs((3, 5, 15, 1, 3))) crashes then#asahi-dev · 2026-07-17 20:11
yukaAh, maybe also try the latest 26.6 beta#asahi-dev · 2026-07-17 20:11
yukaokay, that's all the questions I have for today, I might poke you again when I have more to test#asahi-dev · 2026-07-17 20:09
yuka(installed in the macOS volume before you installed the m1n1 custom kernel object?)#asahi-dev · 2026-07-17 20:09
yukaThis is the latest 27 (developer|public) beta?#asahi-dev · 2026-07-17 20:08
yukadoes that crash?#asahi-dev · 2026-07-17 20:08
yukathe last thing you could check for me: msr((3, 5, 15, 1, 3), mrs((3, 5, 15, 1, 3)))#asahi-dev · 2026-07-17 20:08
yukaNice!#asahi-dev · 2026-07-17 20:07
yukawhat did p.smp_start_secondaries() say?#asahi-dev · 2026-07-17 20:06
yukaand you boot on the first p-core, good to know#asahi-dev · 2026-07-17 20:06
yukaNice, I remembered/guessed the right part no#asahi-dev · 2026-07-17 20:05
yuka> CPU: M5 Tupai (P core)#asahi-dev · 2026-07-17 20:04
yukayes, same on A18 Pro and M5 Pro#asahi-dev · 2026-07-17 19:57
yukamy branch includes a config change which will change the function of the dfu port, so probably use another port for now. On your device the dfu port is "The rightmost USB-C port when you’re facing the left side of the Mac" according to apple docs.#asahi-dev · 2026-07-17 19:55
yukawhen you have booted my branch, try p.smp_start_secondaries() from the proxy shell#asahi-dev · 2026-07-17 19:54
yukasmp_id and CPU part#asahi-dev · 2026-07-17 19:54
yuka(with or without my branch, doesn't matter)#asahi-dev · 2026-07-17 19:54
yukaThe whole "CPU init (...)" block of the m1n1 boot log would be interesting to know#asahi-dev · 2026-07-17 19:52
yukaYou can compile and chainload this m1n1 version: https://github.com/yuyuyureka/m1n1/tree/feature/t8142#asahi-dev · 2026-07-17 19:51
yukaif you simply connected to your M5 MacBook Air with a usb cable with no further setup, then you're using cdc-acm#asahi-dev · 2026-07-17 19:50
yukathe question is what you set as M1N1DEVICE when running it#asahi-dev · 2026-07-17 19:48
yukacdc-acm?#asahi-dev · 2026-07-17 19:48
yukaHow do you access the m1n1 proxy?#asahi-dev · 2026-07-17 19:11
yukarhlvista[m] would you be willing to test some more things on M5? what is your setup?#asahi-dev · 2026-07-17 18:57
yukasince those also share the same irq nums for other devices#asahi-dev · 2026-07-17 18:57
yukaI'm guessing M4 Max = M4 Pro, M5 Max = M5 Pro and M2 Ultra = M2 Max = M2 Pro#asahi-dev · 2026-07-17 18:56
yukaadded the new "real" values and assumed values: https://termbin.com/4oc4