2026-07-09
00:09nela has quit [Ping timeout: 480 seconds]
00:16nela has joined #asahi-gpu
00:20nela has quit []
00:21nela has joined #asahi-gpu
02:00yurix has quit [Remote host closed the connection]
02:00yurix has joined #asahi-gpu
02:00ciara has quit [Remote host closed the connection]
02:13pb17 has quit [Ping timeout: 480 seconds]
02:13nela has quit [Quit: Ping timeout (120 seconds)]
02:13ciara has joined #asahi-gpu
02:13nela has joined #asahi-gpu
02:16pb17 has joined #asahi-gpu
04:07nela has quit [Ping timeout: 480 seconds]
04:12nela has joined #asahi-gpu
04:29nela has quit [Ping timeout: 480 seconds]
04:33nela has joined #asahi-gpu
05:14mattia013 has joined #asahi-gpu
06:13kaazoo has joined #asahi-gpu
06:14kaazoo has quit []
06:15cy8aer has quit [Remote host closed the connection]
06:21ddxtanx_ has joined #asahi-gpu
06:21ddxtanx has quit [Read error: Connection reset by peer]
06:28cy8aer has joined #asahi-gpu
06:28mattia013 has quit [Quit: mattia013]
06:38nela has quit [Quit: Ping timeout (120 seconds)]
06:39nela has joined #asahi-gpu
07:01pb17 has quit [Ping timeout: 480 seconds]
07:39pb17 has joined #asahi-gpu
07:42HerbY_NL has joined #asahi-gpu
07:47HerbY_NL has quit []
08:43ddxtanx has joined #asahi-gpu
08:43ddxtanx_ has quit [Read error: Connection reset by peer]
08:50HerbY_NL has joined #asahi-gpu
09:18ddxtanx has quit [Quit: Konversation terminated!]
09:21ddxtanx has joined #asahi-gpu
09:28pb17 has quit [Ping timeout: 480 seconds]
09:34pb17 has joined #asahi-gpu
10:04ddxtanx has quit [Ping timeout: 480 seconds]
10:17HerbY_NL has quit [Quit: Leaving]
10:19HerbY_NL has joined #asahi-gpu
10:26yukahe got macOS running in the m1n1 hv, and is trying to get the changes for that into an upstreamable shape
10:27noopwafelthe M4 is pretty close gpu-wise I think! but good to know, I was worried my lack of attention led to wasted effort
10:27ddxtanx has joined #asahi-gpu
10:27noopwafelfor the record https://noopwafel.net/asahi/gpu_notes.txt has high-level notes and some pointers to the tree etc
10:28Sobek[m](The important thing is that he got a SPTM emulator working, however, his implementation of that is tainted, so he'll be writing spec of that part and someone else (yuka has volunteered iirc) will be writing a clean-room implemenation of that. In any case, that is still a huge breakthrough).
10:29noopwafelwe've been super busy with other stuff but started to look at the kernel/mesa stack again the last few days..
10:30HerbY_NL has quit [Quit: Leaving]
10:31ddxtanx has quit []
10:31ddxtanx has joined #asahi-gpu
10:37ddxtanx_ has joined #asahi-gpu
10:37ddxtanx has quit [Read error: Connection reset by peer]
10:41StanfordAppliedCyber[m]oh yeah i did some extensive userspace re work
10:42StanfordAppliedCyber[m]i have a disassembler with around 200 instructions understood
10:42StanfordAppliedCyber[m]and a bunch of information on other userspace bits
10:42StanfordAppliedCyber[m]tbh ik the devs here want me to work on stuff like pcie, speakers, etc first, but i really just want to jump straight into the gpu
10:42StanfordAppliedCyber[m]noopwafel: i havent checked but all my probing is against m4
10:42StanfordAppliedCyber[m]probably the same
10:42StanfordAppliedCyber[m]i bought a m5 also and will repeat the probing there
10:43StanfordAppliedCyber[m]noopwafel: nah from scratch, i don't know how much work was done on the m3
10:43StanfordAppliedCyber[m]i asked for resources a few times and no one pointed me to anything
10:43StanfordAppliedCyber[m]how can you develop mesa without a working kernel driver?
10:44StanfordAppliedCyber[m]wdym, you need a way to actually alloate buffers and submit work to the gpu, does m1n1 already have enough support for that?
10:45StanfordAppliedCyber[m]so you're saying m1n1 already has enough support to actually test mesa? at least to some limited extent
10:45StanfordAppliedCyber[m]also where is the m3 disassembler?
10:46noopwafelyeah, 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)
10:47noopwafeland I can see from backscroll that you did get pointed at me quite a while ago :D sorry for not seeing your message above
10:47noopwafelhttps://github.com/TellowKrinkle/applegpu/tree/M3 has .. maybe not up-to-date, but some stuff
10:49noopwafelalthough I don't think it has the weird entry shader stuff which are the more interesting ones.
10:49noopwafelbut yeah a few of us have been slowly working on this in the background, trying to get to The Blessed Triangle for m3 native
10:49StanfordAppliedCyber[m]what is this weird entry shader stuff?
10:51StanfordAppliedCyber[m]oh interesting, good to know, i'm VERY interested in the gpu to put it mildly
10:52StanfordAppliedCyber[m]it's the whole reason why i did this project and built the HV
10:56Sobek[m](Stanford Applied Cyber, are you a cat ? Because clearly the Asahi lead seem to struggle herding you :D)
10:57noopwafelM3+ 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
10:58StanfordAppliedCyber[m]Sobek[m]: if you knew me in real life you'd understand how generally true this is
10:59StanfordAppliedCyber[m]idk, i rationalize it by saying results are results, after all, i am the first person to get the hv working on m4 class SoCs
10:59noopwafelunfortunately 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.
10:59StanfordAppliedCyber[m]fair enough
11:00noopwafelmore 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..
11:01noopwafelthe m1n1/kernel stuff was all targeting 14.x, also. I don't know how much the initdata mess changed again in 26.x.
11:01StanfordAppliedCyber[m]probably a lot, but that's my business
11:01StanfordAppliedCyber[m]i've said it many times but I'm firmly unconvinced a gpu driver takes years
11:02StanfordAppliedCyber[m]if in a few weeks i go "holy shit this is way harder than i expected" i'm sure someone will tell me "i told you so"
11:02noopwafelit's kind of .. initial bringup vs something maintainable, especially on the mesa side
11:02chaos_princessyears - no, multiple months - more reasonable
11:02chaos_princess+ consider the fact that you will most likely be building on top of a pre-existing one
11:03StanfordAppliedCyber[m]that's fair, but i also feel to an extent it's if you get the capability working you get the capability working and you find a way to upstream that
11:03StanfordAppliedCyber[m]chaos_princess: oh yeah 100%, it's amazing how little time the m1 gpu driver took imo given its complexity
11:03StanfordAppliedCyber[m]if i had to go from scratch the years timeline is very reasonable imo
11:06StanfordAppliedCyber[m]<Sobek[m]> "(PS, shouldn't you be getting..." <- yes
11:07StanfordAppliedCyber[m]good {morning, afternoon, night} yall
11:08Sobek[m]The M1/M2 work spent quite a while reverse engineering the whole stuff, getting proofs of concept working on top of a macOS kernel, (before the linux kernel driver ever existed), then it spent a fair amount of time on getting... (full message at <https://matrix.org/oftc/media/v1/media/download/Ac4s8JInSsPQoLO46LXlFzv81dK5TAaomqsZUSJtsGvDubUKJN_aykcCznYC6rqFr0oczhPEDboaQyqZzr_OqbpCefjqlGBQAG1hdHJpeC5vcmcvUVVFZFVabFhmYkdCdmtYSWducVdzbkV1>)
11:24pac85StanfordAppliedCyber[m]: re how to develop Mesa witjouta kernel driver. We re-ed enough of the apple kernel interface to submit stuff through it. We have a standalone compute demo that runs on macos without metal. Also much of what's needed to draw a triangle is known at this point but I personally stopped working on the project (plan on coming back). https://gitlab.freedesktop.org/pac85/mesa/-/tree/suballocstuff_gfx?ref_type=heads
11:25pac85Also, that branch is capable do tracing apple's driver. My personal approach to reing M4 would be to try tracing that and see where it fails and then find the differences. That said focusing on m3 for now probably makes more sense as M4 would be incremental work on top
13:07olivial has quit [Read error: Connection reset by peer]
13:08olivial has joined #asahi-gpu
13:55fl0_id_afkStanfordAppliedCyber[m] how can you work on gpu without pcie or a somewhat working system, isn't it connected through that?
13:58chaos_princessthis is not x86, only wifi/bt/sd reader are pcie
13:59chaos_princesseven x86 does not do pcie for igpus, they just pretend to have pcie as it is not just a bus but also an autoconfiguration system at this point
14:05fl0_id_afkchaos_princess ah good to know. right I forgot it's basically a one chip
14:05fl0_id_afkthat makes even more sense
14:10noopwafelpac85: 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
14:25tellowkrinkleBTW everything I've seen so far indicates that at least from an ISA perspective, M4 is almost identical to M3.
14:27lederstrumpf has joined #asahi-gpu
15:33HerbY_NL has joined #asahi-gpu
15:34HerbY_NL has quit []
15:38lacamar[m]It's good to see that the hardware is mostly unchanged. Is it just the GPU and interrupt controller that changed a lot from M2?
15:40mattia013 has joined #asahi-gpu
15:40chaos_princessusb phy changed, usb-pd is now on spmi instead of i2c (pro/max only), internal mic decimator changed
15:48HerbY_NL has joined #asahi-gpu
15:54HerbY_NL has quit [Quit: Leaving]
16:52Guest12749 has quit [Quit: Bridge terminating on SIGTERM]
16:52rhysmdnz has quit [Quit: Bridge terminating on SIGTERM]
16:57rhysmdnz has joined #asahi-gpu
16:58Guest12985 has joined #asahi-gpu
17:48pb17 has quit [Ping timeout: 480 seconds]
17:49pb17 has joined #asahi-gpu
17:50HerbY_NL has joined #asahi-gpu
17:52HerbY_NL has quit []
17:55HerbY_NL has joined #asahi-gpu
18:59balrog_ has joined #asahi-gpu
19:02balrog has quit [Ping timeout: 480 seconds]
19:10HerbY_NL has quit [Quit: Leaving]
19:17ddxtanx has joined #asahi-gpu
19:17ddxtanx_ has quit [Read error: Connection reset by peer]
19:18alyssam1/m2 are basically the same, m3/m4 basically the same, m5 new but one would reasonably assume m6 will be close
19:19alyssaStanfordAppliedCyber[m]: writing enough Mesa code to spin cubes took me a few weeks while I was in school. writing enough to play D3D games at good speed took years
19:19alyssait's not that gpu drivers are horribly impenetrable, it's just there's a metric ton of random details & rare features that you do need to support for conformance
19:20alyssaIf all the r/e was done, I expect the mesa side for m3 to be a few months to get something reasonbale, since you're building on top of years of my m1 work
19:21alyssaso I guess in a sense, StanfordAppliedCyber[m] noopwafel chaos_princess are all right in a way :)
19:26alyssaand that's all common code in Mesa now and all of that will Just Work on future platforms
19:26alyssaMain reason I didn't touch M3 when I was still active in the project was just that nobody else was seriously working on kernel stuff at the time and, well, 2025 was a hard year for all of us
19:30ddxtanx has quit [Quit: Konversation terminated!]
19:30ddxtanx has joined #asahi-gpu
19:39ddxtanx_ has joined #asahi-gpu
19:39ddxtanx has quit [Read error: Connection reset by peer]
19:50ddxtanx has joined #asahi-gpu
19:51ddxtanx_ has quit [Read error: Connection reset by peer]
20:09nela has quit [Ping timeout: 480 seconds]
20:11nela has joined #asahi-gpu
20:31pb17 has quit [Ping timeout: 480 seconds]
20:36amarioguy has quit [Ping timeout: 480 seconds]
20:41Sobek[m]Is your current job forbidding you from writing code that's Asahi related on your free time ?
21:03rainbyte has quit [Quit: rainbyte]
21:10pb17 has joined #asahi-gpu
21:11ddxtanx has quit [Read error: Connection reset by peer]
21:11ddxtanx has joined #asahi-gpu
21:20rainbyte has joined #asahi-gpu
21:32madmatt067 has quit [Quit: .]
21:36yurix has quit [Read error: Connection reset by peer]
21:36madmatt067 has joined #asahi-gpu
21:36madmatt067 has quit []
21:38yurix has joined #asahi-gpu
21:52ddxtanx has quit [Read error: Connection reset by peer]
21:52ddxtanx has joined #asahi-gpu
21:58yurix has quit [Ping timeout: 480 seconds]
22:02chadmed has quit [Quit: Konversation terminated!]
22:03yurix has joined #asahi-gpu
22:23madmatt067 has joined #asahi-gpu
22:24pb17 has quit [Ping timeout: 480 seconds]
22:34pb17 has joined #asahi-gpu
22:37madmatt067 has quit [Quit: .]
22:46madmatt067 has joined #asahi-gpu
22:48madmatt067 has quit []
23:02gatallah has joined #asahi-gpu
23:05gatallah has quit [Remote host closed the connection]
23:23maettu102 has quit [Quit: brb]
23:29maettu102 has joined #asahi-gpu