#asahi-dev · 2026-01-31 08:11
chadmedwell i guess this is still binary re in a way, but the output of apples shader compiler is your binary#asahi-dev · 2026-01-31 08:10
chadmedall of those tools are cricket, only _binary_ re (as you say, decompiling apple software in ghidra) is not allowed#asahi-dev · 2026-01-31 05:55
chadmedsee https://lore.kernel.org/asahi/12ab93b7-1fc2-4ce0-926e-c8141cfe81bf@kernel.org/#asahi-dev · 2026-01-31 05:54
chadmedyou add them immediately below the oneOf as with t6020. the list with the comment is for compatibility with the generic apple,nvme-ans2 compat#asahi-dev · 2026-01-31 04:27
chadmedi believe the current meta is to forego all generic compatibles entirely. there was a bit of drama on the mailing list surrounding generic compatibles for one of the hardware blocks #asahi-dev · 2026-01-30 22:56
chadmedit worked well with DCP_COLOURSPACE_NATIVE and whatever DCP_XFER_FUNC_HDR is, i assume its PQ #asahi-dev · 2026-01-30 22:55
chadmedhttps://github.com/chadmed/linux/commit/5fa3d22649ebf159b132e567f8aded63588833d2 this is how i did it, but its incomplete because im not propagating the hdr metadata to dcp yet in this branch #asahi-dev · 2026-01-30 22:52
chadmedwe do need to make sure we are only advertising the correct colourspaces for each connector type though if/when we add more. the three youve got there should be fine for both hdmi and dp but some dp colourspaces are not supported via hdmi #asahi-dev · 2026-01-30 22:51
chadmedbut thats basically the same as what i had so thats good#asahi-dev · 2026-01-30 22:50
chadmedinteresting, i tried setting mode_changed to true and it would never actually force a modeset. maybe i was doing it in the wrong place#asahi-dev · 2026-01-30 21:49
chadmedprobably not good or necessary if some monitors cant do 10 bit rgb with bt 2020#asahi-dev · 2026-01-30 21:49
chadmedyeah i was doing that too but couldnt get kwin to play nice with the ycbcr modes so stopped saving them and changed how we pick up the "best" mode too#asahi-dev · 2026-01-30 21:45
chadmedit seems to prioritise bit depth over everything else, so my monitor's 12 bit ycbcr mode was winning out over 10 bit rgb#asahi-dev · 2026-01-30 21:45
chadmedi had to add some stuff to the parser to explicitly reject them #asahi-dev · 2026-01-30 21:44
chadmeddid you run into any issues with iomfb deciding a ycbcr colour mode was "best"#asahi-dev · 2026-01-30 21:44
chadmedah right#asahi-dev · 2026-01-30 21:41
chadmedi had it working fine, the issue was forcing modesets when enabling/disabling it#asahi-dev · 2026-01-30 21:40
chadmedit was broken for you? #asahi-dev · 2026-01-28 20:48
chadmedcan we drop support for it now #asahi-dev · 2026-01-27 11:59
chadmedcan they be squashed into the normal dt branch? #asahi-dev · 2026-01-27 11:29
chadmedeven though half of it is implementing entirely new features #asahi-dev · 2026-01-27 11:28
chadmedit keeps picking up asoc codec changes as stable backports for some reason #asahi-dev · 2026-01-27 11:27
chadmedif it blows up in their faces, good maybe theyll stop doing it#asahi-dev · 2026-01-27 11:27
chadmedi just ignore those. if sasha wants to poison gregkh's stable tree thats between them #asahi-dev · 2026-01-27 11:25
chadmedintegralpilot: unfortunately the person who did it is the hwmon maintainer and can do whatever he wants :P #asahi-dev · 2026-01-27 11:25
chadmedi know we arent doing denormals properly in that code but thats not going to cause FIELD_PREP issues is it#asahi-dev · 2026-01-27 11:24
chadmedlmao is it even right? something looks off about it #asahi-dev · 2026-01-27 11:23
chadmedno thanks elon im not helping you brainjack my friends even if i thought i had the skills to do so #asahi-dev · 2026-01-27 11:23
chadmedneuralink found my work email and had some chud spam it with "come work for us" crap #asahi-dev · 2026-01-27 11:22
chadmedtriforce went viral on hn (i didnt even post it) and all it really led to was a bunch of "why didnt he just use AI" and a job offer from the modern day mengele himself (i said no of course) #asahi-dev · 2026-01-27 11:19
chadmedmore unsolicited advice: just avoid hn entirely. dont post there, dont go there#asahi-dev · 2026-01-27 11:07
chadmedthe other problem is, no one is writing a driverkit/iokit/whateverkit pcie driver for their widget unless apple pays them, and apples not going to pay them because apple's got their own hardware#asahi-dev · 2026-01-27 11:06
chadmediirc the rpi 4 suffered from a similar problem, and even with revolting disgusting amdgpu kernel driver hacks to make it "work" it was abysmally slow #asahi-dev · 2026-01-27 11:05
chadmedits not that pcie over tbt doesnt work, its that the thing doesnt support the memory model that gpus assume exists for performance hacks #asahi-dev · 2026-01-27 11:04
chadmedyeah#asahi-dev · 2026-01-27 11:03
chadmedwell they glued on a whole second pcie controller and that still couldnt save it. i dont expect to see another cheese grater. the m5 ultra will probably go in a studio#asahi-dev · 2026-01-27 11:02
chadmedyeah that ^ #asahi-dev · 2026-01-27 11:02
chadmedpretty sure their private compute stuff is rackmounted apple silicon #asahi-dev · 2026-01-27 11:01
chadmedmaybe if it had more real pcie lanes it wouldve been interesting #asahi-dev · 2026-01-27 11:00
chadmedi half expected it to be a test bed for a new server platform but they never followed through on that#asahi-dev · 2026-01-27 11:00
chadmedidek why they released it tbh its so obvious the studio is supposed to replace it#asahi-dev · 2026-01-27 10:59
chadmedlast time i checked there were single digit install attempts so....#asahi-dev · 2026-01-27 10:57
chadmedi wonder if their rf chips will be as black boxy#asahi-dev · 2026-01-27 10:55
chadmedi dont think theres any real reason to stage trivial changes downstream anymore unless the whole driver is also still downstream #asahi-dev · 2026-01-27 10:54
chadmeddoubly so if we want to get at least some of this stuff upstream-only#asahi-dev · 2026-01-27 10:53
chadmedand thats okay, we dont need to rush a release out the door#asahi-dev · 2026-01-27 10:53
chadmedmvp this time around basically has to be everything except gpu/dcp imo #asahi-dev · 2026-01-27 10:52
chadmedsd card is whatever as long as there arent random power management serrors #asahi-dev · 2026-01-27 10:51
chadmedyeah no we need to do better#asahi-dev · 2026-01-27 10:50
chadmedit scales very poorly as the damage regions get larger