#
#asahi-dev · 2026-02-11 03:20
nickchanthe way it's implemented is mrs still works but msr will undef
#asahi-dev · 2026-02-11 03:19
nickchanas garbage as this is this is still better than qhat qualcomm is doing
#asahi-dev · 2026-02-11 03:19
nicolas17so if we tell them "fine keep your GL2 secrets but we really need the hypervisor system register" *maybe* they will listen?
#asahi-dev · 2026-02-11 03:19
nickchan11.2
#asahi-dev · 2026-02-11 03:18
nicolas17they added raw boot objects later, to our benefit
#asahi-dev · 2026-02-11 03:17
nicolas17IIRC macOS 11 shipped with only Mach-O boot objects
#asahi-dev · 2026-02-11 03:16
nickchanon "laptops"
#asahi-dev · 2026-02-11 03:15
nickchanregarding EL2
#asahi-dev · 2026-02-11 03:15
nickchannicolas17: but it's unlikely lobbying would do a thing given the kind of stuff qualcomm can get away with
#asahi-dev · 2026-02-11 03:14
nicolas17if you mean exploits I find that unlikely :P
#asahi-dev · 2026-02-11 03:13
nickchanBut rat race is not happening given absolutely nothing has happened to ios 17.0's sptm
#asahi-dev · 2026-02-11 03:13
nickchaneither that or rat race
#asahi-dev · 2026-02-11 03:13
nicolas17...do we need to lobby Apple?
#asahi-dev · 2026-02-11 03:12
nickchanMacOs won't have the graphics drivers
#asahi-dev · 2026-02-11 03:11
nickchannicolas17: think so and it is especially a dealbreaker because there's no sane way to reproduce muvm
#asahi-dev · 2026-02-11 03:10
nicolas17in magical land where we have hardware documentation or whatever and we can magically port linux to m4... we can't run VMs because registers required for hypervisor are disabled in raw boot mode??
#asahi-dev · 2026-02-11 03:10
nickchanand this is a dealbreaker even for just running linux
#asahi-dev · 2026-02-11 03:09
nicolas17wth
#asahi-dev · 2026-02-11 03:09
nickchanno hyoervisor will work right without that
#asahi-dev · 2026-02-11 03:09
nickchanmost worringly, SYS_IMP_APL_VM_TMR_FIQ_ENA_EL2 is on the disabled list
#asahi-dev · 2026-02-11 03:08
nickchanthis includes things like Apple Generic Timer, but
#asahi-dev · 2026-02-11 03:08
nickchana lot of other registers are also disabled
#asahi-dev · 2026-02-11 03:06
nicolas17Shiz: so yeah I think what sven meant with "we can't virtualize it on m4" is "we can't run macOS in m1n1 and trace the way it accesses GXF, because there is no GXF available when we do that"
#asahi-dev · 2026-02-11 03:04
nicolas17at least that's how I'd do it...
#asahi-dev · 2026-02-11 03:03
nicolas17then you can move that to an M4
#asahi-dev · 2026-02-11 03:03
nicolas17then you can start emulating the behavior instead of letting the actual CPU handle it
#asahi-dev · 2026-02-11 03:01
nicolas17and *that* needs M2/M3
#asahi-dev · 2026-02-11 03:01
nicolas17to emulate GL2 first you need to understand how it works and how the software uses it, and for that the best way is what y'all have been doing for other hardware: pass-through virtualization that logs accesses but then forwards it to the real hardware
#asahi-dev · 2026-02-11 02:59
nicolas17ah now I see the context of the conversation :p
#asahi-dev · 2026-02-11 02:58
Shizthat's what i understood as 'virtualizing' sptm :p
#asahi-dev · 2026-02-11 02:57
Shizgenter/gexit should be trappable as undefined instructions
#asahi-dev · 2026-02-11 02:57
Shizyea, emulating GL2 would be my thought
#asahi-dev · 2026-02-11 02:54
nicolas17oof
#asahi-dev · 2026-02-11 02:54
nicolas17"Either: emulate GL2 on a "regular" ARM CPU: Trap genter/gexit and create shadow copies of pagetables, or: Try to inject hardware driver kexts into vmapple XNU that can run without guarded levels"
#asahi-dev · 2026-02-11 02:53
nicolas17AH just found it on mastodon
#asahi-dev · 2026-02-11 02:53
nicolas17sven: btw are your slides available? you skipped a few things due to time constraints
#asahi-dev · 2026-02-11 02:51
nicolas17according to sven's talk at 39c3, the CPU extensions are available on M2/M3 when booting a foreign OS, they are not on M4
#asahi-dev · 2026-02-11 02:48
ShizI'm reading virtualizing 'it' as SPTM here, so I'm wondering why it would be different on m4
#asahi-dev · 2026-02-11 02:47
Shizbut that's the case for both m2/m3 and m4, as far as I can see?
#asahi-dev · 2026-02-11 02:47
Shizwell yeah, if you boot a raw object it dumps you into EL2 with all that stuff disabled, while mach-O boot goes through SPTM
#asahi-dev · 2026-02-11 02:46
nicolas17I think on M4 macOS expects SPTM to handle pagetables, but if you boot a "foreign OS" like m1n1, the proprietary instructions to initialize guarded exception levels and make SPTM run on it, are all disabled
#asahi-dev · 2026-02-11 02:41
Shizwhy is that?
#asahi-dev · 2026-02-11 02:41
Shiz sven : chaos_princess: well, the biggest difference is that we can't virtualize it easily on m4 anymore but i'm pretty sure we can on m2/m3 if we really cared
#asahi-dev · 2026-02-10 20:59
jannauI've pushed asahi-6.18.9-1 to github
#asahi-dev · 2026-02-10 14:19
chaos_princessfair
#asahi-dev · 2026-02-10 14:18
svenimho calling the situation on m2/m3 and m4 "similar" is misleading due to that
#asahi-dev · 2026-02-10 14:18
chaos_princessyea, i know :P
#asahi-dev · 2026-02-10 14:17
chaos_princessand we get to launch before sptm with full access to isa extensions on those
#asahi-dev · 2026-02-10 14:17
svenchaos_princess: well, the biggest difference is that we can't virtualize it easily on m4 anymore but i'm pretty sure we can on m2/m3 if we really cared
#asahi-dev · 2026-02-10 14:17
sven*different

plus ›