#
✕ nick:StanfordAppliedCyber[m]
#asahi-dev · 2026-05-16 02:41
StanfordAppliedCyber[m]i'm getting the usb serial number now
#asahi-dev · 2026-05-16 02:40
StanfordAppliedCyber[m] MPIDR_EL1 = 0x0000000080010100
#asahi-dev · 2026-05-16 02:40
StanfordAppliedCyber[m] MIDR_EL1 = 0x00000000611f0611
#asahi-dev · 2026-05-16 02:38
StanfordAppliedCyber[m]reading from that address instantly faults my machine with an serror
#asahi-dev · 2026-05-16 02:35
StanfordAppliedCyber[m]have to rebless my m1n1 install
#asahi-dev · 2026-05-16 02:35
StanfordAppliedCyber[m]i'm getting both right now
#asahi-dev · 2026-05-16 02:30
StanfordAppliedCyber[m]ack, i'll try to get both to you shortly
#asahi-dev · 2026-05-16 02:27
StanfordAppliedCyber[m]how can i check both those things? so they want me to read32(that address) and check what usb serial number from dfu?
#asahi-dev · 2026-05-16 02:25
StanfordAppliedCyber[m]if it's a high value thing i can make it happen but it won't be soon, maybe in about 36 hours
#asahi-dev · 2026-05-16 02:24
StanfordAppliedCyber[m]i'm sending these messages from macos on the neo
#asahi-dev · 2026-05-16 02:24
StanfordAppliedCyber[m]I don't think so
#asahi-dev · 2026-05-16 02:24
StanfordAppliedCyber[m]can i do it from macos somehow?
#asahi-dev · 2026-05-16 02:24
StanfordAppliedCyber[m]i'm now doing all my dev work on my neo and my mini is the one being instrumented
#asahi-dev · 2026-05-16 02:24
StanfordAppliedCyber[m]unfortunately i have completely transitioned my development workflow
#asahi-dev · 2026-05-15 08:17
StanfordAppliedCyber[m]which explains the earlier cpu revision
#asahi-dev · 2026-05-15 08:17
StanfordAppliedCyber[m]pretty soon after launch
#asahi-dev · 2026-05-15 08:08
StanfordAppliedCyber[m]clearly there is a way to get the cores working, macos + macho boot does, but it's something we don't understand
#asahi-dev · 2026-05-15 08:07
StanfordAppliedCyber[m]i'm trying to understand what i'm doing differently that other people aren't and the silicon revision appears to be a big one
#asahi-dev · 2026-05-15 08:07
StanfordAppliedCyber[m]well sure, but in raw boot mode it doesn't work on my machine whereas it works on everyone else's
#asahi-dev · 2026-05-15 08:05
StanfordAppliedCyber[m]because it works in macho boot mode with a regular apple kernelcache
#asahi-dev · 2026-05-15 08:05
StanfordAppliedCyber[m]--> something very fishy is going on
#asahi-dev · 2026-05-15 08:05
StanfordAppliedCyber[m]yeah, that's my assumption also, because what this indicates is that macos, in raw boot mode, isn't capable of starting its own cores
#asahi-dev · 2026-05-15 08:04
StanfordAppliedCyber[m]even if it did nothing but crash the hypervisor, we'd at least notice some side effect?
#asahi-dev · 2026-05-15 08:04
StanfordAppliedCyber[m]my point is that if macos had successfully started the cores, we'd know about it right?
#asahi-dev · 2026-05-15 08:04
StanfordAppliedCyber[m]i know
#asahi-dev · 2026-05-15 07:57
StanfordAppliedCyber[m]because if i pass through the writes from the macos side macos actually panics with some error like "waited 250ms for cores to boot"
#asahi-dev · 2026-05-15 07:57
StanfordAppliedCyber[m]ie, we pass through the writes, macos successfully starts secondaries, the secondaries don't know our hypervisor exists and then something bad happens, but we'd notice the something bad?
#asahi-dev · 2026-05-15 07:56
StanfordAppliedCyber[m]question, so i know for the hypervisor we start the cores first, trap the mmio CPUSTART write, and then release the core from our spin table, and i know why we do this, but if we deliberately turned off the handling here, ie, let macos start the cores, not intercept the write, at minimum we should notice the result right? like maybe it crashes the hypervisor, or causes some other side effect, but we'd notice it right?
#asahi-dev · 2026-05-15 06:49
StanfordAppliedCyber[m]is this load bearing, maybe, maybe not, but it's somewhat suspicious
#asahi-dev · 2026-05-15 06:48
StanfordAppliedCyber[m]mine ends in a 0
#asahi-dev · 2026-05-15 06:48
StanfordAppliedCyber[m]she has the same silicion revision as chewy, ends in a 1
#asahi-dev · 2026-05-15 06:48
StanfordAppliedCyber[m]yuka's image which starts secondaries on her machine does NOT work on my machine
#asahi-dev · 2026-05-15 05:47
StanfordAppliedCyber[m]if yours is also 0x11 then i need to investigate the hell out of this, because that may be the difference
#asahi-dev · 2026-05-15 05:47
StanfordAppliedCyber[m]yuka: can you check what chip revision you have? `ioreg -p IODeviceTree -d 1 -k "chip-revision"`
#asahi-dev · 2026-05-15 05:46
StanfordAppliedCyber[m]that's actually very reassuring because it indicates I might not be crazy
#asahi-dev · 2026-05-15 05:46
StanfordAppliedCyber[m]one difference, her chip revision is 0x11, mine is 0x10
#asahi-dev · 2026-05-15 05:46
StanfordAppliedCyber[m]we're talking right now
#asahi-dev · 2026-05-15 05:46
StanfordAppliedCyber[m]firmware version is identical, boot policy settings are identical
#asahi-dev · 2026-05-15 05:05
StanfordAppliedCyber[m]something is very strange with specifically my machine
#asahi-dev · 2026-05-15 05:05
StanfordAppliedCyber[m]chewyeclaire sent me her m1n1 binary and it doesn't successfully start the cores on my machine either-- the exact same image works for her
#asahi-dev · 2026-05-14 23:42
StanfordAppliedCyber[m]unfortunate-- does anyone here have a working m1n1 1tr image that gets smp working on the macbook neo? i need to rule out me doing something royally stupid with a known good branch
#asahi-dev · 2026-05-14 23:34
StanfordAppliedCyber[m]yuka: are you online right now? i could really use your help if so
#asahi-dev · 2026-05-14 23:33
StanfordAppliedCyber[m]@ nick i did pass cpus=1, pmgr panics due to cpu topology issues
#asahi-dev · 2026-05-14 09:28
StanfordAppliedCyber[m]good news: device init works, we're almost all the way to bsd_init
#asahi-dev · 2026-05-14 09:28
StanfordAppliedCyber[m]well calling it for today, my hypervisor is essentially blocked on smp bringup, tried pruning a bunch of stuff from the adt to keep going on a single core but adt remains very ornery and i don't feel like monkeypatching it more
#asahi-dev · 2026-05-14 08:09
StanfordAppliedCyber[m]which i'd be fine with if i didn't get told "smp just works"
#asahi-dev · 2026-05-14 08:07
StanfordAppliedCyber[m]my cpu cores simply will not start
#asahi-dev · 2026-05-14 08:06
StanfordAppliedCyber[m]okay, i definitely need some help here-- can someone send me a pointer to a working branch i can just run make on, and a corresponding working binary just to see if something very funky is happening?
#asahi-dev · 2026-05-14 07:37
StanfordAppliedCyber[m]yeah ik, i'm familiar, it just surprised me because afaik the secondary cpu start has never involved them, not that mmio reads had side effects generally
#asahi-dev · 2026-05-14 07:08
StanfordAppliedCyber[m]i'm having issues getting a read trace from macos but it's clear reads have side effects

plus ›