#asahi-gpu · 2026-08-24 20:34
noopwafelright now I am mostly just going back and cleaning up the "# FIXME: alyssa: non-zero data in the padding" (where usually it is not, in fact, non-zero data in the padding..)#asahi-gpu · 2026-08-24 20:34
noopwafelso a test suite would be <3#asahi-gpu · 2026-08-24 20:33
noopwafel(with some way to stub the adt/gpu register reads)#asahi-gpu · 2026-08-24 20:33
noopwafelI was hoping that I could just wrap rust_fill_gpu_initdata somehow for regression testin#asahi-gpu · 2026-08-24 20:26
noopwafelchaos_princess: no, but is that a kernel branch comparing the kernel-side w/m1n1-side?#asahi-gpu · 2026-08-24 20:21
noopwafelalthough it is actually what I am staring at .. again#asahi-gpu · 2026-08-24 20:21
noopwafeloops, sorry, that was me fighting ssh#asahi-gpu · 2026-08-24 20:21
noopwafelPorting all the initdata stuff back to rust (in m1n1) is the sensible next step on the driver side.. a lot of churn but mostly testable on M1/M2.#asahi-gpu · 2026-08-06 10:59
noopwafelPorting all the initdata stuff back to rust (in m1n1) is the sensible next step on the driver side.. a lot of churn but mostly testable on M1/M2.#asahi-gpu · 2026-08-06 10:44
noopwafelpac85: yes it's definitely time to make it work. adapting the compiler would be really great honestly. I don't think there's (fingers crossed) too much left to learn on the kernel side.\#asahi-gpu · 2026-07-09 14:41
noopwafelI'm kind of expecting 14.x->26.x to be more of a nightmare than M3->M4 :p#asahi-gpu · 2026-07-09 14:10
noopwafelpac85: we should fold the different macos versions into that mess sometime :) I realized my local stuff has all this stuff I didn't push.. will try to get it organized and ping you#asahi-gpu · 2026-07-09 14:05
noopwafeland a lot (most of?) this work is userspace, so you can do a bunch of it under macos.#asahi-gpu · 2026-07-09 11:07
noopwafelni$timeofday :)#asahi-gpu · 2026-07-09 11:04
noopwafelbut yes, I would also hope the M4 stuff is .. hopefully mostly incremental#asahi-gpu · 2026-07-09 11:02
noopwafelit's kind of .. initial bringup vs something maintainable, especially on the mesa side#asahi-gpu · 2026-07-09 11:01
noopwafelthe m1n1/kernel stuff was all targeting 14.x, also. I don't know how much the initdata mess changed again in 26.x.#asahi-gpu · 2026-07-09 11:00
noopwafelmore motivation for us to pick it up again I guess, because I imagine would be much better to build M4 support on top rather than trying independent reimplementations..#asahi-gpu · 2026-07-09 10:59
noopwafelunfortunately too much of this is solely in our collective heads atm, and not written down anywhere other than kind of if you squint at the diffs of chaos.#asahi-gpu · 2026-07-09 10:57
noopwafelM3+ generalized a bunch of things which used to be in the VDM struct etc, they're now constructed by what we call an 'entry shader' which also sets the 'real' shader#asahi-gpu · 2026-07-09 10:49
noopwafelbut yeah a few of us have been slowly working on this in the background, trying to get to The Blessed Triangle for m3 native#asahi-gpu · 2026-07-09 10:49
noopwafelalthough I don't think it has the weird entry shader stuff which are the more interesting ones.#asahi-gpu · 2026-07-09 10:47
noopwafelhttps://github.com/TellowKrinkle/applegpu/tree/M3 has .. maybe not up-to-date, but some stuff#asahi-gpu · 2026-07-09 10:47
noopwafeland I can see from backscroll that you did get pointed at me quite a while ago :D sorry for not seeing your message above#asahi-gpu · 2026-07-09 10:46
noopwafelyeah, for m3. I don't know if it's all pushed tbh, but it's just using the drm stub that the original bringup was also done using (via python)#asahi-gpu · 2026-07-09 10:44
noopwafelsure#asahi-gpu · 2026-07-09 10:44
noopwafelvia m1n1?#asahi-gpu · 2026-07-09 10:43
noopwafelwe have shaders working, not much in mesa yet#asahi-gpu · 2026-07-09 10:42
noopwafelso based on the m3 disassembler..?#asahi-gpu · 2026-07-09 10:42
noopwafelbased on the existing m3 one? or is the isa totally different for m4 again?#asahi-gpu · 2026-07-09 10:30
noopwafelwould be amazing to be able to test on the m4 too.#asahi-gpu · 2026-07-09 10:29
noopwafelwe've been super busy with other stuff but started to look at the kernel/mesa stack again the last few days..#asahi-gpu · 2026-07-09 10:27
noopwafelfor the record https://noopwafel.net/asahi/gpu_notes.txt has high-level notes and some pointers to the tree etc#asahi-gpu · 2026-07-09 10:27
noopwafelthe M4 is pretty close gpu-wise I think! but good to know, I was worried my lack of attention led to wasted effort#asahi-gpu · 2026-07-09 10:26
noopwafelah cool#asahi-gpu · 2026-07-09 10:25
noopwafelI do have the whole m1n1 stack for M3 GPU#asahi-gpu · 2026-07-09 10:25
noopwafelStanfordAppliedCyber[m]: what're you refactoring?#asahi-gpu · 2026-03-25 11:07
noopwafelI think the crash I'm hitting is the MCC not being initialized, shall try that..#asahi-gpu · 2026-03-25 11:04
noopwafelthe m3 fw also gets a pmgr range mapped in, which I didn't look into yet (it doesn't get that far)#asahi-gpu · 2026-03-25 08:38
noopwafeland maybe that's not enough. let me poke at that..#asahi-gpu · 2026-03-25 08:29
noopwafel(this is all in m1n1)#asahi-gpu · 2026-03-25 08:28
noopwafelhm, I am just calling pmgr_adt_power_enable on everything under arm-io/#asahi-gpu · 2026-03-25 08:02
noopwafelany bright ideas what could be causing early agx_power task crashes (on M3)? it is _probably_ a foolish error on my part but I am running out of (sensible) ideas. there's some old comments in irc logs about power dependencies, but enabling allll the pmgr domains doesn't seem to make a difference.#asahi-gpu · 2026-02-08 10:46
noopwafel(certainly not with vulkan CTS I guess, since this seems gallium-only?)#asahi-gpu · 2026-02-08 10:42
noopwafelI mean, the question is, are you going to even hit this case with CTS#asahi-gpu · 2026-01-07 19:31
noopwafelI have no clue (although do you really expect any impact from that change?), but why buildtype debug and not debugoptimized (or release)?