00:04 robclark: nice
07:13 dviola: just tried vanilla wine version 10.0 and I could reproduce the issue from https://gitlab.freedesktop.org/mesa/mesa/-/work_items/15223 so I suspect by the time proton 11 comes out that issue will be fixed
07:13 dviola: I can't reproduce that issue with vanilla wine 11.0, so...
07:41 CounterPillow: Is Dmitry Barishkov on this IRC channel by any chance? I'd be happy for a quick thumbs up if the new "is this an HDMi connector check" in https://lore.kernel.org/dri-devel/20260324-color-format-v11-0-605559af4fb4@collabora.com/T/#m87a380657d957f1bff1d51f90f0b3001707f244f solves the review comment he had previously
09:47 valentine: CounterPillow: Dmitry is lumag
09:51 CounterPillow: thanks
10:15 tzimmermann: airlied, sima, please forward drm-next to -rc7. i need commit 2f42c1a61616 ("drm/ast: dp501: Fix initialization of SCU2C") in drm-misc-next
10:26 lumag: CounterPillow: let me take a look
10:40 sima: tzimmermann, some funny conflicts, I'll try to be careful and not screw it up
10:40 sima: airlied, ^^
10:40 tzimmermann: gem-shmem ?
11:04 sima: tzimmermann, nah just stuff in amd due to cherry-picks
11:04 sima: but since I've screwed one of these up in xe last time around I figured I'll be a bit more careful
14:09 dviola: apparently compiling wine with --enable-archs=x86_64,i386 cures the graphical artifacts
14:13 dviola: I wonder if that means bug in wine, compiler or game
15:32 glehmann: is there a backend that actually wants vec4 load_frag_coord? I wonder if we should remove it in favor of the xy, z and w variants
15:45 alyssa: glehmann: possibly vec4 backends
15:46 alyssa: Midgard reads frag coord as a vec4 varying (VARYING_SLOT_POS)
15:53 glehmann: vec4 are to blame for everything that's wrong with NIR 🙃
16:01 alyssa: checks out
16:49 sima: tzimmermann, oops, forgot to hit 'y' it's pushed now
17:14 tzimmermann: thanks sima
17:35 zmike: mareko: https://github.com/KhronosGroup/OpenGL-Registry/pull/680
20:27 Kayden: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40835
20:40 alyssa: 😳
20:41 airlied: alyssa, Kayden: \o/
20:44 karolherbst: oh wow, what a big surprise, a new intel compiler!
20:45 jenatali: \o/
20:46 alyssa: karolherbst: Intel hired a Mesa compiler person who then wrote an Intel Mesa compiler, who could have EVER predicted this outcome
20:46 karolherbst: so do we have more intel compilers or Intel CL drivers at this point in time? 🙃
20:47 karolherbst: but yeah, good work, I should probably test it at some point
20:48 Kayden: hahaha :(
20:48 Kayden: fair
20:48 HdkR: The name makes me think of random sportsball chant, "If they can't do it, no one can."
20:49 HdkR: Also Jay Jay the jetplane.
20:49 glehmann: I will nak a jay label as long as the old intel-fs and intel-vec4 labels aren't garbage collected :D
20:49 Kayden: it's named after the birds :) (blue jay / scrub jay / ...)
20:51 airlied: better than recombining i g c and some fashion :-P
20:52 HdkR: The compiler must scream like a Blue Jay :D
20:54 Kayden: looking pretty while <screaming> happens at the insanity behind the scenes ... ;)
20:54 Kayden: that may have been intended, yes, lol
20:55 airlied: alyssa: out of interest, did you try using LLVM :-P
20:55 airlied:wonders how many powerpoints were required
20:55 Kayden: 1
20:56 karolherbst: prolly mentioned AI somewhere and than management was hooked
20:56 alanc: LLVM is like a virtual LLM, right?
20:57 Kayden: 😂