#asahi-dev · 2025-02-16 15:44
nickchanfor the spi#asahi-dev · 2025-02-16 15:44
nickchanit's not directly compatible, but it looks like an aic-ish situation where it's a subset#asahi-dev · 2025-02-16 15:44
nickchanand that keyboard should be supported by asahi drivers once the older spi variant is supported#asahi-dev · 2025-02-16 15:43
nickchannicolas17: anyways the internal keyboard is wired to it so it had the needed controls for chocolate doom already#asahi-dev · 2025-02-16 15:40
nickchanand it should be possible to expose a CDC NCM gadget through the vhci interface so T2 can get online#asahi-dev · 2025-02-16 15:38
nickchanthe linux firmware abstrations should work pretty well for the uefi firmware#asahi-dev · 2025-02-16 15:37
nickchans8000 iop and t8015 iop#asahi-dev · 2025-02-16 15:37
nickchanlike it has S5L8960X dart and T8015 dart#asahi-dev · 2025-02-16 15:37
nickchanit also have multiple generations of hw... in one SoC#asahi-dev · 2025-02-16 15:32
nickchanwatchOS part only applies to embeddedOS I think#asahi-dev · 2025-02-16 15:32
nickchannicolas17: bridgeOS is as watchOS as tvOS is to watchOS#asahi-dev · 2025-02-16 15:30
nickchanit's the newest of all the SoCs I am targeting, but with older cores#asahi-dev · 2025-02-16 15:29
nickchannicolas17: not a wrong way to think about it#asahi-dev · 2025-02-16 13:53
nickchansven: I will deal with that later#asahi-dev · 2025-02-16 13:52
nickchan*and* it's in asahi-soc already#asahi-dev · 2025-02-16 13:52
nickchanipad mini 4, it's not wrong, just missing the highest p-state#asahi-dev · 2025-02-16 13:51
nickchanarch/arm64/boot/dts/apple/t7000.dtsi is missing the highest p-state for mini4#asahi-dev · 2025-02-16 13:51
nickchanuh oh#asahi-dev · 2025-02-16 11:51
nickchanworks now except some pinctrl which I also need to fix (apple,pinctrl doesn't support configured as wakeup source)#asahi-dev · 2025-02-16 11:33
nickchanI see another driver flip a bool in the atomic xfer function then call the xfer function and then the other functions observe that bool so I will do just that#asahi-dev · 2025-02-16 11:32
nickchanjannau: I see I will force non-IRQ, disable the IRQ and use mdelay instead of msleep #asahi-dev · 2025-02-16 11:13
nickchanI wrote a driver to power-down iPhone 5s with an I2C-connected PMIC, but I get warning "No atomic I2C transfer handler for i2c-0". #asahi-dev · 2025-02-15 11:55
nickchantldr: the port exists to aid development#asahi-dev · 2025-02-15 11:54
nickchanthis least includes at least DART and SPI#asahi-dev · 2025-02-15 11:49
nickchanfiddle with buses on t2 with periperals supported upstream or asahi connected to it, and then make the bus driver work on a-series chips#asahi-dev · 2025-02-15 11:47
nickchanalthough t8012 linux bringup might seem very useless, the purpose of t8012 bringup is that the SoC where some peripherals like the nor flash are closest to macs supported by asahi linux while having previous generation buses/iommus (like dart and spi), development on this soc is to develop drivers for the older buses without also having to also debug the peripherals (debugging two things at the same time is hard)#asahi-dev · 2025-02-14 15:05
nickchanmildsunrise: what I mean is that the registers have vhe schematics so just beaware of access redirection #asahi-dev · 2025-02-14 09:23
nickchaneafer commits with corellium email#asahi-dev · 2025-02-14 09:22
nickchanapparently it's not very known but linux-apfs *is* corellium code :)#asahi-dev · 2025-02-14 07:15
nickchans3_1_c15_c1_5 is _EL1, s3_4_c15_c14_5 is probably _EL2 and s3_4_c15_c14_6 is probably _EL12#asahi-dev · 2025-02-14 07:14
nickchanmildsunrise: the timer register s3_4_c15_c14_5, s3_4_c15_c14_6 and s3_1_c15_c1_5 on t6030 smells like vhe to me#asahi-dev · 2025-02-14 06:42
nickchanchadmed: new versions but it's just me finding things out myself#asahi-dev · 2025-02-14 06:41
nickchanright now i try to not send versions within 22 hours of the previous version which I am pretty sure it's on the frequent end already#asahi-dev · 2025-02-14 06:35
nickchan* finding things out after sending a series and ratelimited (by etiquette) is not fun#asahi-dev · 2025-02-14 06:16
nickchananyways i will send v4 of apple cpmu patches to the list today#asahi-dev · 2025-02-14 06:09
nickchannext thing on the list is DART however it's a complex bit of hw completely different from m1#asahi-dev · 2025-02-14 05:29
nickchanDo need to figure out a way to do more complex hardware without questionable bin RE however#asahi-dev · 2025-02-14 05:26
nickchanI mean practically impossible#asahi-dev · 2025-02-14 05:26
nickchan> hard to have tainted stuff#asahi-dev · 2025-02-14 05:25
nickchanhttps://github.com/cyyself/m1-pmu-gen#asahi-dev · 2025-02-14 05:25
nickchanperformance counters stuff also comes from xnu source drops and then that one event generator#asahi-dev · 2025-02-14 05:24
nickchan(granted, dwi is so simple that the driver is like less than 10 lines of actual code rest are all boilerplates so it's hard to have tainted stuff in there unlike corellium which likes to overcomplicate things)#asahi-dev · 2025-02-14 05:22
nickchanas for my stuff there are no bin RE because apple has more hot stuff in their xnu source drop back then, and cpufreq is some sort of trying random things for 6 months. However some stuff like dwi backlight are made after looking at corellium code that looks quite likely to be bin RE'd#asahi-dev · 2025-02-14 05:18
nickchanBut it's probably not appriorate to do it in general on anything beyond things that are needed to have hypervisor going#asahi-dev · 2025-02-14 05:17
nickchanchaos_princess: hypervisor isn't working properly yet so I would assume *all* of them have binary RE#asahi-dev · 2025-02-12 11:56
nickchan it's also not per-cluster...#asahi-dev · 2025-02-12 11:48
nickchanI mean that register exists but the bit layout is different#asahi-dev · 2025-02-12 11:48
nickchanturns out the uncore registers of A11 is different so I need to patch up the aic's driver's fiq handling again#asahi-dev · 2025-02-12 11:48
nickchanFun#asahi-dev · 2025-02-12 02:12
nickchannevermind it's probably fine as long as NV host doesn't advertise ID_AA64DFR0_EL1_PMUVer_IMP_DEF to guest