#
#asahi · 2026-10-01 14:36
svenit was just annoying but not impossible
#asahi · 2026-10-01 14:35
Hel3ziYeah, I get that it's still early days. I honestly only found out yesterday that you're already actively running m1n1 on it — I was totally misled by all the talk online about SPTM being an impenetrable wall. Out of curiosity, what state is m1n1 currently in? And is it true that it only started booting just the other day?
#asahi · 2026-10-01 14:23
fl0_id_there is work being done, but not ready for broader testing. There's still a lot to work out. if you would test existing work, it should probably be more because you enjoy it
#asahi · 2026-10-01 14:19
Hel3ziHey everyone! I have an M4 Pro machine. I saw that m1n1 work is underway/ready, and was wondering if there’s any way I could help out (e.g. testing or running diagnostics)? Apologies in advance if this is the wrong channel to ask — I’m not a seasoned developer, but I’d love to help out if needed.
#asahi · 2026-09-30 19:27
IsmailIvanov[m]<Sebastian[m]1> "Hey, I’m doing the same, I can..." <- Of course, you can message me
#asahi · 2026-09-30 19:14
Sebastian[m]1<IsmailIvanov[m]> "Hi! I have an M5 MacBook and I’d..." <- Hey, I’m doing the same, I can help if you’d like
#asahi · 2026-09-30 17:34
IsmailIvanov[m]Hi! I have an M5 MacBook and I’d really love to be able to run Linux on it. I’d like to help make that happen. How could I contribute to Asahi Linux and the effort to support M5 Macs?
#asahi · 2026-09-29 18:48
jbeneat if there was a dim if inactive after x seconds thing though
#asahi · 2026-09-29 18:48
jbewould ben
#asahi · 2026-09-29 18:48
jbeguess it actually works if you have that haha
#asahi · 2026-09-29 18:47
jbeforgot the >
#asahi · 2026-09-29 18:47
jbeah, rigt
#asahi · 2026-09-29 18:45
jbee.g echo 100 /sys/class/leds/kbd_backlight/brightness
#asahi · 2026-09-29 17:37
chaos_princesshow are you writing to brightness
#asahi · 2026-09-29 17:35
jbemax_brightness is 255
#asahi · 2026-09-29 14:34
pac85The point of that is to take scalar computations and run them in just one thread. Without it the computations would run per thread yielding the same result and using vector registers that are larher. It's common on mobile GPU because it can save power. More commonly desktop GPUs have a scalar alu and scalar regs that can run in parallel with the vector alu.
#asahi · 2026-09-29 14:33
pac85Also yeah important to distinguish from the concept of a pre shader (or preamble) that existed in previous gens as well
#asahi · 2026-09-29 14:32
pac85opticron: the entry shader replaces the USC, so it's emitted basically every time you bind a pipeline. You need it because the HW requires that you allocate a bunch of things via instructions and has no other way of doing it.
#asahi · 2026-09-29 14:25
opticronah, I conflated entry shaders and preshaders thinking they were the same thing which is why I was confused :D
#asahi · 2026-09-29 14:11
chaos_princessprobably some variation of "it is a gpu, we can change isa encoding every generation if we wish, so who cares".
#asahi · 2026-09-29 14:10
ants[m]s/Apple's rationale/what we can guess of their rationale ofc
#asahi · 2026-09-29 14:10
ants[m]Tellow said the layouts in the M3 gpu instruction set looked in large part machine-generated; what's Apple's rationale for that and is it conceivable to RE that process as well?
#asahi · 2026-09-29 14:09
noopwafelthe entry shaders you need (if nothing else, you need to set the shader body address!), but the preshader is for perf/efficiency .. what chaos_princess said :)
#asahi · 2026-09-29 14:08
chaos_princessentry - they can't as they configure a bunch of stuff about how the main shader runs. What you mean is the preamble shader, and the benefit there is that it runs at 1x1, instead of whatever the main shader runs at
#asahi · 2026-09-29 14:06
opticronI wonder what the benefit of entry shaders is if they can just be smushed into the main shader code
#asahi · 2026-09-29 14:05
noopwafelkit_ty_kate: yes, wildly different ISA + a bunch of things moved to the 'entry shaders' rather than descriptors + a lot of random cursed things
#asahi · 2026-09-29 14:00
kit_ty_kateso if i understood correctly the main diff is the ISA that does instruction layout/padding differently + some new weird little instructions? or am i completely off?
#asahi · 2026-09-29 13:49
kit_ty_katevery cool stuff
#asahi · 2026-09-29 13:48
MichaelReeves[m]looks like it's wrapping up now. It really good presentation, enjoyed it! GPUs are so interesting, the people who understand them are just really awesome :)
#asahi · 2026-09-29 13:11
clegofficthank you
#asahi · 2026-09-29 13:05
chaos_princesshttps://www.youtube.com/watch?v=TPwe53kCJuM - xdc day 2, m3 gpu talk soontm
#asahi · 2026-09-29 07:55
jbewill check in 10 hours or so...
#asahi · 2026-09-29 07:54
chaos_princesswhat is in max_brightness?
#asahi · 2026-09-29 07:53
jbeyeah nothing there
#asahi · 2026-09-29 07:53
chaos_princessdmesg?
#asahi · 2026-09-29 07:52
jbere the keyboard backlight thing, not sure what logs to check tbh, but echoing any value to there appears to do nothing, checking it again afterwards it just stays at 0 (and the backlight is off)
#asahi · 2026-09-28 13:50
jbeI'll have to see if there's any logs or anything next time I have to computer nearby and can try
#asahi · 2026-09-28 13:38
chaos_princessyes, that one is correct and should work
#asahi · 2026-09-28 13:36
jbeI think there were some settings for keyboard backlight back in the day? am I misremembering? like there's a /sys/class/leds/kbd_backlight/brightness but echoing numbers to it does nothing (M2 air/FAR/KDE)
#asahi · 2026-09-28 13:11
ehambergcool. thanks!
#asahi · 2026-09-28 13:06
chaos_princesshttps://www.youtube.com/watch?v=6tNGW8PoSzw - xdc day1 starting, asahi-related talk is day 2, but aly-ssa has a talk in ~5 hrs
#asahi · 2026-09-27 22:32
fl0_id_julesmeow it was mentioned somewhere. this is not the place to discuss it though. if you want to rant maybe #asahi-offtopic
#asahi · 2026-09-27 17:24
Michae|Longhave seen one HN entry about it. Feedback was mixed if I remember correctly.
#asahi · 2026-09-27 17:10
julesmeowhave you all seen the vibecoded asahi fork that just came out promising m4 support?
#asahi · 2026-09-27 15:56
jannauin xe/amdgpu. I'd recommend running under the hypervisor for testing
#asahi · 2026-09-27 15:56
jannauif anyone feels to inclined to comment on the Mac Pro GPU reddit thread: with https://github.com/NVIDIA/open-gpu-kernel-modules/pull/972#issuecomment-3591497063 and an simple IRQ related fixup a geforce 3070 was working. Xe and amdgpu had at that time clear 16k page size issues. Those are probably resolved in the meantime thanks to loongson. Not sure if there are unaligned access mines
#asahi · 2026-09-27 06:36
GoldenGuy1000[m]Now I just gotta wait for gpu drivers 😭
#asahi · 2026-09-27 06:36
GoldenGuy1000[m]I got asahi running on my m3 mac pro and I'm stoked!
#asahi · 2026-09-27 04:05
twb-guest(If I don't come back, assume it worked)
#asahi · 2026-09-27 04:05
twb-guestreboots to test

plus ›