#
#asahi-dev · 2026-06-11 21:51
chaos_princessmacos 27 beta hvf appears to have vhe nv support
#asahi-dev · 2026-06-11 09:28
StanfordAppliedCyber[m]?
#asahi-dev · 2026-06-11 09:28
StanfordAppliedCyber[m]u get my message/
#asahi-dev · 2026-06-11 09:28
StanfordAppliedCyber[m]this is a shared account and shared machine
#asahi-dev · 2026-06-11 09:28
StanfordAppliedCyber[m]i can do discord if matrix gets too annoying, but it's sort of annoying because i'm sending this on a shared account, and this machine has all the stuff on it whereas disc sits on my personal computer
#asahi-dev · 2026-06-11 09:27
skxlldxggyah, mb thanks
#asahi-dev · 2026-06-11 09:27
elisaadoDMs will probably work but they also might... not
#asahi-dev · 2026-06-11 09:27
elisaadoStanfordAppliedCyber[m]: do note that [skxlldxggy](https://matrix.to/#/@_oftc_skxlldxggy:matrix.org) is a bridged IRC user onto matrix
#asahi-dev · 2026-06-11 09:26
skxlldxggywould you like my discord?
#asahi-dev · 2026-06-11 09:25
StanfordAppliedCyber[m]let's pull this into DMs
#asahi-dev · 2026-06-11 09:19
skxlldxggynew ISA-level stuff fast and have time to put into it. what would the clean-room split actually look like?
#asahi-dev · 2026-06-11 09:19
skxlldxggyre: the clean-room SPTM emulator work, I might be useful for that. I'm building my own ARM64 OS from scratch (kernel/bootloader/init/installer, the whole stack, cross-compiled), and I'm comfortable at the register/devicetree level; just cross-checked an M4 ADT dump against m1n1's T8132 UART base and it lined up exactly. haven't done hypervisor/EL2 work specifically but I pick up
#asahi-dev · 2026-06-11 09:04
yukaso we rely on apple having configured them in a way that is workable for us
#asahi-dev · 2026-06-11 09:04
yukawe're more reliant on apple for these devices, because they no longer allow us to configure the implementation specific cpu registers
#asahi-dev · 2026-06-11 09:03
StanfordAppliedCyber[m]and yeah as yuka said, these devices have bugs that unfortunately we can't do much more than wait for apple to fix
#asahi-dev · 2026-06-11 09:03
StanfordAppliedCyber[m]if you're very good at systems / arm stuff and/or have a lot of time i could use some help clean rooming my sptm emulator
#asahi-dev · 2026-06-11 09:03
yukaand possibly some bugs on the apple side
#asahi-dev · 2026-06-11 09:03
StanfordAppliedCyber[m]skxlldxggy: it's a blocker but we have multiple workable paths forward, someone just has to do the work
#asahi-dev · 2026-06-11 09:02
yukaskxlidxggy: the blocker for getting more hardware to work is the tooling / hypervisor
#asahi-dev · 2026-06-11 09:02
StanfordAppliedCyber[m]skxlldxggy: i'm the only one working on the tooling for sptm enabled devices (m4, a18 pro, m5) right now with my first target being the macbook neo
#asahi-dev · 2026-06-11 09:01
StanfordAppliedCyber[m]note for anyone else working on exclave stuff: from the ADT you have to remove BOTH /arm-io/aop-exclave-mailbox and /arm-io/aop/iop-aop-nub, if you remove one or the other by itself xnu panics
#asahi-dev · 2026-06-11 09:00
skxlldxggyon these yet given the SPTM/GL2 stuff I've read about. is there a known blocker or is it more "nobody's had time"?
#asahi-dev · 2026-06-11 09:00
skxlldxggy hey folks; anyone working on T8132 (base M4) bring-up at the moment? I noticed the EARLY_UART_BASE commit for it from a while back and got curious. I've got a base M4 (J713/Mac16,12) sitting around and pulled the ADT off it. UART addres lines up exactly with what's in soc.h already. happy to share the full dump or do testing if it'd help, just not sure if m1n1 even gets a foothold
#asahi-dev · 2026-06-11 06:54
StanfordAppliedCyber[m]also the reason why i'm working on the neo is because it's so popular-- a very accessible way for someone to use asahi linux, not to mention whenever the 8gb ram starts showing its limitations turning it into a linux device seems like a compelling possibility
#asahi-dev · 2026-06-11 06:52
StanfordAppliedCyber[m]nah i'm good, i'm definitely not the person you should be spending hardware money on
#asahi-dev · 2026-06-11 06:49
svenif you need hw we can figure out how to expense it in the open collective. there’s enough funds there
#asahi-dev · 2026-06-11 06:39
StanfordAppliedCyber[m]surprisingly not that spaghetti, i've cleaned it up a lot, most of the remaining code wouldn't be that hard to upstream
#asahi-dev · 2026-06-11 06:39
Sobek[m]StanfordAppliedCyber[m]: Someone will have to maintain your pile of grad student spaghetti code :p
#asahi-dev · 2026-06-11 06:38
StanfordAppliedCyber[m]well defined as everything except gpu + dcp
#asahi-dev · 2026-06-11 06:38
StanfordAppliedCyber[m]i would accept a m5 class mac and in return i'd promise to get linux working on it
#asahi-dev · 2026-06-11 06:38
StanfordAppliedCyber[m]thanks for your support!
#asahi-dev · 2026-06-11 06:38
StanfordAppliedCyber[m]obviously if you want to send me money i'll be happy to assist, but if you want to help asahi linux send it to the open collective
#asahi-dev · 2026-06-11 06:37
StanfordAppliedCyber[m]that's a joke
#asahi-dev · 2026-06-11 06:37
Sobek[m]But I have put some money on the Asahi open collective, which they could use.
#asahi-dev · 2026-06-11 06:37
Sobek[m]That’s going to be hard from the EU, with the US not using SEPA/IBAN.
#asahi-dev · 2026-06-11 06:35
StanfordAppliedCyber[m]if you want to send me money i'll be happy to send you account and routing numbers
#asahi-dev · 2026-06-11 06:35
StanfordAppliedCyber[m]i won't be here for long
#asahi-dev · 2026-06-11 06:35
StanfordAppliedCyber[m]i'm a soon to be graduating college student who wants to work on some random fun project for a bit
#asahi-dev · 2026-06-11 06:33
Sobek[m][Stanford Applied Cyber](https://matrix.to/#/@appliedcyber:matrix.org) are you doing this work on your free time for free, or are you getting supported by the Asahi Open Collective ?
#asahi-dev · 2026-06-11 00:52
StanfordAppliedCyber[m]oh also one thing making it easier is that my current understanding is that the secureos boot don't run in parallel, first sptm loads secure kernel, makes sure it's ready, and only then does it boot xnu, so i don't think there's some detailed parallel init thing we need to deal with
#asahi-dev · 2026-06-11 00:37
StanfordAppliedCyber[m]the latter is easy because we own sptm, the former is annoying because now we need to RE that entire protocol to shuttle messages properly
#asahi-dev · 2026-06-11 00:36
StanfordAppliedCyber[m]btw the issue with implementing them is there's basically two ways for them to talk to each other: some custom protocol managed by sptm (tightbeam) and shared memory buffers
#asahi-dev · 2026-06-11 00:36
StanfordAppliedCyber[m]as an extra guest*, two guests total (xnu and secure kernel)
#asahi-dev · 2026-06-11 00:34
StanfordAppliedCyber[m]i'd just run secore kernel + exclave kit in el1/0 respectively as two other guests
#asahi-dev · 2026-06-11 00:33
StanfordAppliedCyber[m]yes
#asahi-dev · 2026-06-11 00:33
JamesCalligeros[m]theyd just be extra guests under m1n1 right
#asahi-dev · 2026-06-11 00:33
StanfordAppliedCyber[m]the biggest risk would be all the apple sysregs ofc
#asahi-dev · 2026-06-11 00:32
StanfordAppliedCyber[m]if i had to implement exclaves i think it would take under a week
#asahi-dev · 2026-06-11 00:32
StanfordAppliedCyber[m]no doubt, but fwiw i'm still not super worried about them
#asahi-dev · 2026-06-11 00:32
JamesCalligeros[m]theoretically the copro fw blobs wont care who is talking to them, so it should all still work from linux drivers, but for tracing we will need the exclaves

plus ›