#
✕ nick:yuka
#asahi-dev · 2026-10-04 21:55
yukaor rather, what are you trying to boot in the hv
#asahi-dev · 2026-10-04 21:55
yukawhat payload did you send?
#asahi-dev · 2026-10-04 21:53
yukaallusivedev what device do you have again?
#asahi-dev · 2026-10-04 21:52
yukayes, though that's still quite a lot of drivers
#asahi-dev · 2026-10-04 21:47
yukadefconfig + 16KB pagesize works but it may be a bit large (longer compile time)
#asahi-dev · 2026-10-04 21:47
yukayes
#asahi-dev · 2026-10-04 21:12
yukaHel3zi try to check out https://cyberchaos.dev/yuka/linux branch "all" and build it
#asahi-dev · 2026-10-04 20:34
yukaI'm out for today but thank you for the debugging session
#asahi-dev · 2026-10-04 20:18
yukaso the apcie0 power domain of the nvme node is ps_apcie_phy_sw?
#asahi-dev · 2026-10-04 20:17
yukayou're using your version again, right?
#asahi-dev · 2026-10-04 20:16
yukaso I think the apcie0 power-domain of the nvme node is just wrong
#asahi-dev · 2026-10-04 20:16
yukaps_aphy_st0_aus5_a indirectly depends on ps_apcie_phy_sw
#asahi-dev · 2026-10-04 20:15
yukathis is interesting
#asahi-dev · 2026-10-04 20:15
yukaAha, wait!
#asahi-dev · 2026-10-04 20:05
yukawhat happens if you wait 5 or 10 minutes?
#asahi-dev · 2026-10-04 20:05
yukait just hangs?
#asahi-dev · 2026-10-04 20:05
yukahm, ps_st0_aus5 does not cause a crash or anything?
#asahi-dev · 2026-10-04 20:00
yukasomething something peer review, since one can always make mistakes (all of us do) :-)
#asahi-dev · 2026-10-04 19:58
yukaso maybe those changes are not actually needed?
#asahi-dev · 2026-10-04 19:58
yukathe other one as well
#asahi-dev · 2026-10-04 19:58
yuka> L2C_ERR_ADR: 0x282500380700150
#asahi-dev · 2026-10-04 19:58
yukathe ps_soc_dpe paste is actually crashing on fpwm1 still
#asahi-dev · 2026-10-04 19:57
yukaack
#asahi-dev · 2026-10-04 19:56
yukado you still get a crash with this? https://cyberchaos.dev/yuka/linux/-/blob/7e64c5d0357110bd577c5662282b21e0b166d337/arch/arm64/boot/dts/apple/t8142.dtsi
#asahi-dev · 2026-10-04 19:56
yukathis is more proper than always-on because with always-on it might still try to write it in some circumstances
#asahi-dev · 2026-10-04 19:55
yukaadded: status = "disabled"; /* Locked, don't touch */
#asahi-dev · 2026-10-04 19:55
yukaty!
#asahi-dev · 2026-10-04 19:51
yukabut then I'd like to see the crash log with the exact FAR or L2C_ERR_STS to understand what is happening
#asahi-dev · 2026-10-04 19:51
yukathere is some precedent for that, disp*_cpu are the same
#asahi-dev · 2026-10-04 19:51
yukadoes it crash with an SError trying to touch that?
#asahi-dev · 2026-10-04 19:50
yukaokay... for which reason?
#asahi-dev · 2026-10-04 19:49
yukadeleting ps_pms_fpwm1 is odd as well
#asahi-dev · 2026-10-04 19:49
yukawhat happens without the changes to them?
#asahi-dev · 2026-10-04 19:49
yukaps_soc_dpe and ps_st0_aus5 I would like more information on
#asahi-dev · 2026-10-04 19:49
yukaalways-on on ps_ans makes sense to me, the M4+ devices do not like the ans power domain going down ime
#asahi-dev · 2026-10-04 19:48
yukaps_soc_dpe already has always-on on most other devices, so that is plausible I guess. But what happens when you do all the changes except that one?
#asahi-dev · 2026-10-04 19:46
yukaI want to be entirely sure that I understand all of these changes, and that it's the minimal set of changes to get the device to boot and stay on
#asahi-dev · 2026-10-04 19:46
yukaok
#asahi-dev · 2026-10-04 19:41
yukachange what exactly?
#asahi-dev · 2026-10-04 19:40
yukaalso pushed the gpio-reserved-ranges change: https://cyberchaos.dev/yuka/linux/-/blob/a8b0114fec1f7ed31e095a15612c7b33a9322ce4/arch/arm64/boot/dts/apple/t8142.dtsi
#asahi-dev · 2026-10-04 19:39
yukaI assume it will probably reboot after a few minutes?
#asahi-dev · 2026-10-04 19:39
yukanice
#asahi-dev · 2026-10-04 19:35
yukagpio-reserved-ranges = <78 1>, <203 1>;
#asahi-dev · 2026-10-04 19:31
yukait will probably crash again with a different FAR
#asahi-dev · 2026-10-04 19:31
yukaadd this to pinctrl_ap: pinctrl@39a000000: gpio-reserved-ranges = <78 1>;
#asahi-dev · 2026-10-04 19:30
yukawhich is listed in no-restore-gpio-pins
#asahi-dev · 2026-10-04 19:30
yukahex(0x138//4) => 0x4e
#asahi-dev · 2026-10-04 19:29
yuka> FAR: 0x39a000138
#asahi-dev · 2026-10-04 19:29
yukaOk yes that's exactly it
#asahi-dev · 2026-10-04 19:17
yukaon pinctrl_ap

plus ›