#asahi · 2021-09-13 10:22
kettenismarcan: correct, but a copy is installed as EFI/BOOT/BOOTX64.EFI such that if the BootXXXX/BootOrder variables aren't set correctly the system still boots#asahi · 2021-09-12 16:59
kettenisI believe most of the arm64 linux distros that use the UEFI bootflow can handle that already and so can OpenBSD#asahi · 2021-09-12 16:58
kettenisNot being able to set the BootOrder variable and others isn't really an issue if you don't install multiple OSes that boot via UEFI as the removable media boot flow will just work#asahi · 2021-09-12 16:51
kettenissame holds for GetTime()/SetTime()#asahi · 2021-09-12 16:50
kettenisthis includes the ResetSystem() call so we don't have to bend over backwards to implement that in u-boot#asahi · 2021-09-12 16:48
kettenismost runtime services are optional#asahi · 2021-09-11 16:19
kettenisthe SMC might be what is starting the CPU that runs iBoot#asahi · 2021-09-11 16:08
kettenismaybe it is in a ROM?#asahi · 2021-09-11 15:21
kettenisyay!#asahi · 2021-09-11 14:21
kettenisthat's a fairly recent thing in u-boot, but the code seems to be there in my fork#asahi · 2021-09-11 14:18
kettenisnot sure how initrd gets loaded in that scenario, although I have seen people mention that this relies on the EFI LoadFile2 protocol#asahi · 2021-09-11 14:16
kettenisalthough there is EFISTUB as well which allows you to directly load a kernel if I understand things correctly#asahi · 2021-09-11 14:14
kettenisI believe that with the UEFI boot path folks typically use grub and grub does the initrd loading#asahi · 2021-09-11 14:13
kettenisso booting Linux directly from u-boot is (largely) untested#asahi · 2021-09-11 14:12
kettenismy main goal with u-boot is to enable the UEFI interface#asahi · 2021-09-11 14:04
kettenisI would expect u-boot to support most formats in common use#asahi · 2021-09-11 14:03
kettenisI don't know; I don't do Linux ;)#asahi · 2021-09-11 14:02
kettenisthere should be no initrd#asahi · 2021-09-11 13:29
ketteniswell, I have a diff to set the baudrate to 115200 but I doubt that will help you#asahi · 2021-09-11 13:28
kettenisj_ey: no (but I'm using gcc as a cross-compiler to build m1n1 and u-boot)#asahi · 2021-09-11 13:11
kettenisbtw, maz and alyssa, I pushed a small change to U-Boot to update the msi-ranges property for the PCIe node to match the proposed binding#asahi · 2021-09-11 13:09
kettenisI don't think it would be too horrible if we skip the rtkit firmware node for now and add it later if it turns out we need it#asahi · 2021-09-11 13:02
kettenismarcan, the iop-nub,rtbuddy-v2 nodes?#asahi · 2021-09-11 13:01
kettenisit looks like both iop,ascwrap-v4 and iop,m3wrap-v2-acio were first introduced on the M1, so even the compatible names make sense to me ;)#asahi · 2021-09-11 12:59
kettenisanyway, I think either design could work for U-Boot and OpenBSD, so I think the DT binding for the mailbox should be fine#asahi · 2021-09-11 12:58
kettenisyes, apple should hurry up with their next-gen hardware such that we can compare ;)#asahi · 2021-09-11 12:53
kettenisif that's the case, I think we need a separate node for the RTKit firmware#asahi · 2021-09-11 12:53
kettenisI think what matters here is we end up with multiple device trees node that need to reference specific channels exposes by the RTKit firmware#asahi · 2021-09-11 12:49
kettenisbut it looks as if your Linux mailbox driver only deals with the hardware mailbox#asahi · 2021-09-11 12:48
kettenisso the NVMe node can simply reference the mailbox#asahi · 2021-09-11 12:47
ketteniscurrently the U-Boot mailbox driver handles the rtkit protocol as well#asahi · 2021-09-11 12:46
kettenissven: it is a bit difficult for me to review the mailbox DT binding from a OpenBSD/U-Boot standpoint since it isn't clear how the upper layers are going to look#asahi · 2021-09-11 12:36
kettenisso if you write a nonzero value in the top bits, it reads back as zeroes?#asahi · 2021-09-11 12:36
kettenisah#asahi · 2021-09-11 12:35
kettenis(just curious)#asahi · 2021-09-11 12:35
kettenishow did you figure that out?#asahi · 2021-09-11 12:35
kettenisbut is it the mailbox hardware that ignores those bits or is it the rtkit firmware?#asahi · 2021-09-11 12:33
kettenissven: why do you document the mailboxes as 64+32 bits when the driver does a writeq to both registers?#asahi · 2021-08-23 14:04
kettenisat least we've figured out another bit of the protocol ;)#asahi · 2021-08-23 14:02
kettenisI'll need to do that in u-boot as well#asahi · 2021-08-23 13:45
kettenisbut at least we can tell!#asahi · 2021-08-23 13:45
kettenisthe version range suggest that at some point they may break backwards compatibility#asahi · 2021-08-23 13:30
kettenismaybe c000b means it supports protocol versions between 11 and 12?#asahi · 2021-08-23 13:19
ketteniswe can pick any (32-bit) address we want it seems#asahi · 2021-08-23 13:16
kettenismaz: you'd be happy with a msi-controller.yaml that defines the controller side of things?#asahi · 2021-08-23 13:16
kettenismaz: i'll see if I can beat some yaml into shape then#asahi · 2021-08-23 13:14
ketteniswe need to trick maz into looking at robh's and mine latest musings#asahi · 2021-08-23 12:30
kettenison openbsd it is#asahi · 2021-08-23 12:19
kettenisthis is a bug in the syscon driver of course#asahi · 2021-08-23 12:17
kettenisif syscon gets probed first it uses devm_ioremap() which does the right thing