#
✕ nick:yuka
#asahi-dev · 2026-07-16 19:43
yukaprobably something changed about dockchannel uart in A18 Pro / M5
#asahi-dev · 2026-07-16 19:42
yukathen it makes more sense
#asahi-dev · 2026-07-16 19:42
yukaI revise my statement, t6000 works too
#asahi-dev · 2026-07-16 17:54
yukaThis would be extremely odd and I need to thorougly test once more but I think t6034 and t8132 work, while t6000, t8140 and t6050 don't
#asahi-dev · 2026-07-16 17:49
yukafwiw the linux in m1n1 hv works on t6050 without modifications
#asahi-dev · 2026-07-16 17:45
yukasome other interrupts fired shortly after unmasking, but no new interrupts fired when I sent something into the dcuart
#asahi-dev · 2026-07-16 17:32
yukaI unmasked all interrupts and didn't see an event for 360
#asahi-dev · 2026-07-16 17:32
yukaseems like the interrupts just don't fire :(
#asahi-dev · 2026-07-16 17:03
yukaokay I see
#asahi-dev · 2026-07-16 17:02
yukairq_hw_number_t hwirq
#asahi-dev · 2026-07-16 16:48
yukat8140 also ends up with +0x10000 on the irq num
#asahi-dev · 2026-07-16 16:33
yukaor wpa_supplicant doesn't do Owe by default I think?
#asahi-dev · 2026-07-16 16:33
yukaand yes I regularly run into it at our office
#asahi-dev · 2026-07-16 16:33
yukaI just have [DriverQuirks] OweDisable=brcmfmac in my /etc/iwd/main.conf
#asahi-dev · 2026-07-16 16:31
yukaflokli is it an encrypted open network?
#asahi-dev · 2026-07-16 16:18
yukahttps://cyberchaos.dev/yuka/linux/-/blob/dfe8f1f0f987e3071a35decce8c5874ff28e0819/arch/arm64/boot/dts/apple/t6050.dtsi#L303
#asahi-dev · 2026-07-16 16:15
yukais the leading "0x10" this expected?
#asahi-dev · 2026-07-16 16:15
yukaand found hwirq = 0x10168
#asahi-dev · 2026-07-16 16:15
yukaI added a debug print in aic_irq_unmask()
#asahi-dev · 2026-07-16 16:15
yukarequest_irq(irq, dcuart_interrupt, IRQF_SHARED, "...", dcuart);
#asahi-dev · 2026-07-16 16:14
yukairq = platform_get_irq(pdev, 0);
#asahi-dev · 2026-07-16 16:14
yukamy dockchannel uart code does:
#asahi-dev · 2026-07-16 16:14
yukathis is the line in the device tree. I don't use any of the multi-die stuff yet
#asahi-dev · 2026-07-16 16:14
yuka> interrupts = <AIC_IRQ 360 IRQ_TYPE_LEVEL_HIGH>;
#asahi-dev · 2026-07-16 16:13
yukathe interrupt num is 360 = 0x168
#asahi-dev · 2026-07-16 16:12
yukaI'm looking at the aic linux driver and trying to figure out why my dockchannel uart irq handler is not called on t6050. fairly sure the interrupt actually happens, but it's not properly unmasked or similar.
#asahi-dev · 2026-07-15 17:19
yukaYes makes sense for the python/proxyclient part
#asahi-dev · 2026-07-15 17:19
yukareads
#asahi-dev · 2026-07-15 17:18
yukabits are a bit more difficult
#asahi-dev · 2026-07-15 17:18
yukareg offsets I could put into a struct
#asahi-dev · 2026-07-15 17:17
yukahttps://github.com/yuyuyureka/m1n1/commit/7a2ff9a7d5964c12c74d4e4f454dc1697001b43c
#asahi-dev · 2026-07-15 17:17
yukathis = "use different reg offsets and bits if gen >= 4"
#asahi-dev · 2026-07-15 17:15
yukaNow the question is how to generalize this
#asahi-dev · 2026-07-15 17:13
yukaMy t6050 just self-entered debugusb
#asahi-dev · 2026-07-15 17:09
yukaon spmi gen4
#asahi-dev · 2026-07-15 17:09
yukaI think reset_panic_counter works :o
#asahi-dev · 2026-07-15 16:43
yukathe reg layout is totally different though
#asahi-dev · 2026-07-15 16:43
yukaI think I found the CMD reg on spmi gen 4 ^^
#asahi-dev · 2026-07-14 21:57
yukamakes sense
#asahi-dev · 2026-07-14 21:57
yukaOh
#asahi-dev · 2026-07-14 21:57
yukaso for 3 field one it applies to all dies or what?
#asahi-dev · 2026-07-14 21:56
yukainterrupts = <AIC_IRQ 0 620 IRQ_TYPE_LEVEL_HIGH>;
#asahi-dev · 2026-07-14 21:56
yukavs
#asahi-dev · 2026-07-14 21:56
yukainterrupts = <AIC_IRQ 620 IRQ_TYPE_LEVEL_HIGH>;
#asahi-dev · 2026-07-14 21:56
yukaI keep getting confused by some interrupts props having 3 and some having 4 fields
#asahi-dev · 2026-07-14 21:47
yuka*find it
#asahi-dev · 2026-07-14 21:47
yukat602x is missing it, would be nice if someone could fin dit
#asahi-dev · 2026-07-14 21:46
yukathis confirms the t6000 adt is actually correct about the dockchannel-uart interrupt num
#asahi-dev · 2026-07-14 21:46
yuka> Exception: IRQ (from EL2h) die: 0 type: 1 num: 620 mpidr: 80000000 cnt: 882cc5576
#asahi-dev · 2026-07-14 21:46
yukaOkay got it

plus ›