#asahi-dev · 2026-05-13 05:25
StanfordAppliedCyber[m]26.4.1#asahi-dev · 2026-05-13 05:24
StanfordAppliedCyber[m]all the rvbars look right, the cpus just... never start#asahi-dev · 2026-05-13 05:23
StanfordAppliedCyber[m]it just prints Failed for all of them#asahi-dev · 2026-05-13 05:23
StanfordAppliedCyber[m]literally your exact code#asahi-dev · 2026-05-13 05:23
StanfordAppliedCyber[m]nope, all secondaries failed#asahi-dev · 2026-05-13 05:22
StanfordAppliedCyber[m]smp#asahi-dev · 2026-05-13 05:22
StanfordAppliedCyber[m]but i'm clearly doing something wrong because you + chewy both got it working#asahi-dev · 2026-05-13 05:21
StanfordAppliedCyber[m]i literally pulled your branch and it doesn't work#asahi-dev · 2026-05-13 05:21
StanfordAppliedCyber[m]and the pmgr one#asahi-dev · 2026-05-13 05:21
StanfordAppliedCyber[m]yeah i saw that PR#asahi-dev · 2026-05-13 05:21
StanfordAppliedCyber[m]spent all day on this#asahi-dev · 2026-05-13 05:21
StanfordAppliedCyber[m]what the hell did you to do get it working#asahi-dev · 2026-05-13 05:20
StanfordAppliedCyber[m]yeah i cant get my smp to work#asahi-dev · 2026-05-12 22:27
StanfordAppliedCyber[m]also if i look at it am i tainted forever? it was posted in this channel so i assume not, but just want to clarify that i'm fine to integrate it into my tree#asahi-dev · 2026-05-12 22:27
StanfordAppliedCyber[m]hitting some pmgr issue right now which i think is due to us only using one core right now#asahi-dev · 2026-05-12 22:26
StanfordAppliedCyber[m]yuka how working is your multi core cpu init? if i were to just drop it into m1n1 does everything "just work"?#asahi-dev · 2026-05-12 21:29
StanfordAppliedCyber[m]on 115200 baud also#asahi-dev · 2026-05-12 21:29
StanfordAppliedCyber[m]and it's super flaky, but oh well i'll live with it#asahi-dev · 2026-05-12 21:29
StanfordAppliedCyber[m]that's my exact setup#asahi-dev · 2026-05-12 20:27
StanfordAppliedCyber[m]i am targeting the macbook neo right now#asahi-dev · 2026-05-12 20:27
StanfordAppliedCyber[m]jannau: yep i've already been doing this, my uart is just flaky i think#asahi-dev · 2026-05-12 05:09
StanfordAppliedCyber[m]figured out the problem that took 4 days#asahi-dev · 2026-05-12 05:09
StanfordAppliedCyber[m]successfully got to FB init, still not far, but i'm over a massive hump#asahi-dev · 2026-05-11 23:54
StanfordAppliedCyber[m]i'm stuck debugging a macos hang, the same problem for 4 days, and i'm getting very tired of LinkedProgram and 1tr#asahi-dev · 2026-05-11 23:54
StanfordAppliedCyber[m]if so what's the status of that? would make my iteration significantly easier#asahi-dev · 2026-05-11 23:53
StanfordAppliedCyber[m]i was told jannau was working on making the chainload more reliable on newer machines?#asahi-dev · 2026-05-10 22:16
StanfordAppliedCyber[m]i'll add however that the biggest thing in our favor i that sptm mediates all interactions between gl and el, so ultimately the surface we need to emulate can be reduced to a relatively small set of calls + the sptm handoff#asahi-dev · 2026-05-10 22:15
StanfordAppliedCyber[m]most of the exclave plumbing is in place on the m4#asahi-dev · 2026-05-10 22:15
StanfordAppliedCyber[m]also, re: m5, exclaves will be annoying but i'm far more optimistic about "just run them as a separate guest" because sk and the exclae userspace are meant to run under the sptm hypervisor in GL, whereas sptm is itself effectively a hypervisor --> it's snowflaky. for exclaves, secure kernel, and other m5 stuff, my first approach would be second guest as opposed to full emulation, at least based off my current understanding.#asahi-dev · 2026-05-10 21:27
StanfordAppliedCyber[m]apparently there's evidence the amx gate is just a sysctl, i can investigate this pretty easily just by togging the value and then seeing what amx workloads do#asahi-dev · 2026-05-10 21:05
StanfordAppliedCyber[m]parts of this are definitely upstreamable ignoring sptm because among other things i'm adding support for 26.4.1, and i'm optimistic we can find a copyright friendly way of getting the sptm emulation here#asahi-dev · 2026-05-10 21:04
StanfordAppliedCyber[m]thanks for the help and feedback on the approach @ sven @ other asahi devs who chimed in, will keep yall updated with where i make it#asahi-dev · 2026-05-10 08:59
StanfordAppliedCyber[m]that said, i'm not experienced on copyright here other than true clean room == kosher#asahi-dev · 2026-05-10 08:59
StanfordAppliedCyber[m]the latter is mostly a bunch of simple endpoints with a small number of more complicated ones#asahi-dev · 2026-05-10 08:58
StanfordAppliedCyber[m]the former is a bunch of page mappings#asahi-dev · 2026-05-10 08:58
StanfordAppliedCyber[m]like, split it into two things, there's the sptm handoff, and the sptm runtime (the endpoints)#asahi-dev · 2026-05-10 08:58
StanfordAppliedCyber[m]i can say that my current emulation, which makes it past virtual memory init, is not super complicated#asahi-dev · 2026-05-10 08:57
StanfordAppliedCyber[m]I'll defer to the asahi maintainers on this one#asahi-dev · 2026-05-10 08:56
StanfordAppliedCyber[m]it does mean m1n1.bin is something like 120mb, but hey you only go through 1tr once#asahi-dev · 2026-05-10 08:55
StanfordAppliedCyber[m]that's how I load the kcache#asahi-dev · 2026-05-10 08:55
StanfordAppliedCyber[m]i'm not kidding, that's how we load the kcache#asahi-dev · 2026-05-10 08:55
StanfordAppliedCyber[m]dumb solution: add macho segment to m1n1, then memcpy()#asahi-dev · 2026-05-10 08:55
StanfordAppliedCyber[m]yeah i hit that#asahi-dev · 2026-05-10 08:53
StanfordAppliedCyber[m]I'm already using it to emulate sptm-- python is way too slow for thousands of calls and i'm obviously not going through 1tr for every iteration loop#asahi-dev · 2026-05-10 08:53
StanfordAppliedCyber[m]> the LinkedProgram() class already provide the infrastructure to link such blobs#asahi-dev · 2026-05-10 08:51
StanfordAppliedCyber[m]that said it's 2am in california and I'm going to bed#asahi-dev · 2026-05-10 08:51
StanfordAppliedCyber[m]but the whole point of this exercise is to figure that out!#asahi-dev · 2026-05-10 08:50
StanfordAppliedCyber[m]it's like region 3 or soemthing#asahi-dev · 2026-05-10 08:50
StanfordAppliedCyber[m]yeah there's a whole new sptm range for dart#asahi-dev · 2026-05-10 08:49
StanfordAppliedCyber[m]if we need 10,000 lines, well, good thing you can run multiple guests in el1!