#asahi · 2021-05-04 19:22
kettenis1that can be used as a payload for m1n1 and will give you pcie without too much effort#asahi · 2021-05-04 19:21
kettenis1_andy_t: you may be interested in https://github.com/kettenis/u-boot/tree/apple-m1-m1n1#asahi · 2021-05-04 19:19
kettenis1does FreeBSD/arm64 normally boot using UEFI?#asahi · 2021-05-04 19:18
kettenis1any way, congrats _andy_t_#asahi · 2021-05-04 19:18
kettenis1so OpenBSD/powerpc now has the computer that brings up the computer that brings up the actual CPU that then boot firmware that boot linux that boots an OpenBSD kernel that then boots the real OpenBSD kernel ;)#asahi · 2021-05-04 19:16
kettenis1kexec() pretty much barfs on anything that doesn't resemble a linux kernel#asahi · 2021-04-29 21:31
kettenis1it appears the SART is used to give the iop that manages the NVMe storage access to some of the DRAM on the SoC#asahi · 2021-04-29 15:07
kettenis1is the audio based on I2S? Wonder how much powerpc heritage there is of that's the case#asahi · 2021-04-29 12:07
kettenis1which is kinda hard to do without more device drivers ;)#asahi · 2021-04-28 13:57
kettenis1thanks, I'll tweak the binding schemas a bit for that and then submit it to the relevant lists for feedback#asahi · 2021-04-28 13:56
kettenis1for now?#asahi · 2021-04-28 13:56
kettenis1compatible = "apple,t8103-pinctrl", "apple,pinctrl"#asahi · 2021-04-28 13:56
kettenis1marcan: so follow the pattern established by AIC and use#asahi · 2021-04-28 13:18
kettenis1compatible = "apple,t8101-pinctrl", "apple,pinctrl"#asahi · 2021-04-28 13:18
kettenis1if we ever support the T8101 SoC that one would get#asahi · 2021-04-28 12:32
kettenis1compatible = "apple,t8103-pinctrl", "apple,pinctrl"#asahi · 2021-04-28 12:31
kettenis1hmm, I suppose I could just make the pinctrl have#asahi · 2021-04-28 12:15
kettenis1marcan, sven: do you have any opinion about "supporting" older Apple SoCs in the compatible properties of the various SoC hardware blocks?#asahi · 2021-04-28 11:56
kettenis1in theory it would increase TLB pressure a bit, but I don't expect these to survive very long as cache entries anyway#asahi · 2021-04-28 11:54
kettenis1in the grand scheme of things, does having a few more entries in the kernel page tables really matter?#asahi · 2021-04-28 11:53
kettenis1path of least resistence is probably not to worry about ioremapping the same physical page twice ;)#asahi · 2021-04-28 11:51
kettenis1having the a clock node that indicates its parent clocks through a "clock" property is standard practice#asahi · 2021-04-28 11:48
kettenis1are you talking about a "syscon" node with children for each gatable clock here?#asahi · 2021-04-28 11:48
kettenis1sven: what do you mean by having the clocks property reference the clock controller itself?#asahi · 2021-04-26 17:54
kettenis1personally I don't care about the older SoCs#asahi · 2021-04-26 17:53
kettenis1that is why both "t8103" and "t8101" are present#asahi · 2021-04-26 16:30
kettenis1?#asahi · 2021-04-26 16:30
kettenis1compatible = "apple,t8103-pinctrl", "apple,t8101-pinctrl";#asahi · 2021-04-26 16:30
kettenis1should we therefore use#asahi · 2021-04-26 16:30
kettenis1for the pinctrl node, the ADT uses "gpio,t8101"#asahi · 2021-04-26 16:29
kettenis1Do we worry about SoCs that predate the M1?#asahi · 2021-04-26 16:28
kettenis1one question that I still have is what we should to with the compatible strings#asahi · 2021-04-26 16:27
kettenis1ok, I have dt-binding schemas for pinctrl and pcie#asahi · 2021-04-24 13:15
kettenis1don't think it has pin muxing capabilities so it would be a gpio node rather than a pinctrl node#asahi · 2021-04-24 13:13
kettenis1the SMC is a different device that will need a different binding#asahi · 2021-04-24 13:12
kettenis1that's not going to happen with Apple products#asahi · 2021-04-24 13:12
kettenis1that is useful if you have a datasheet for the SoC and.or schematics for a board#asahi · 2021-04-24 13:11
kettenis1most of the other generic approaches require assigning names to the pins#asahi · 2021-04-24 13:08
kettenis1the generic pinmux implementation is just one of the few "standard" ways to implement a pinctrl device#asahi · 2021-04-24 13:07
kettenis1there will be a dt-bingings/pinctrl/apple.h file with macros to build the pinmux cells and pull them apart into pin and func#asahi · 2021-04-24 13:01
kettenis1why? it is straightforward and already implemented in the generic pinctrl/pinmux code#asahi · 2021-04-24 12:57
kettenis1no#asahi · 2021-04-24 11:31
kettenis1looks like there is a seperate AIC interrupt for each group of pins#asahi · 2021-04-24 11:30
kettenis1to mention the AIC interrupts that will be triggered#asahi · 2021-04-24 11:30
kettenis1sven: and an interrupts property of course#asahi · 2021-04-24 11:22
kettenis1anyway, I don't think interrupts would have any implication for the DT bindings#asahi · 2021-04-24 11:21
kettenis1or do they really use GPIO bit-banging to implement SPI?#asahi · 2021-04-24 11:20
kettenis1is there an actual SPI controller?#asahi · 2021-04-24 11:19
kettenis1u-boot generally uses polling instead of interrupts#asahi · 2021-04-24 11:19
kettenis1usb keyboards work