01:30 anarsoul: heh, if ladders are *very* expensive with the way we currently compile them
01:31 anarsoul: we end up with a long chain of branches
01:32 anarsoul: see https://gist.github.com/anarsoul/76a16867f59fe1725a38e2c372dec07d
01:34 anarsoul: i.e. if unrolled loop had a single iteration, it'll end up branching multiple times. ouch!
05:23 anarsoul[m]: Not for a single iteration though, but every extra iteration would result in unnecessary branch
06:56 anarsoul: enunes: there is 3 more MRs for you to review :) but I'm almost done
06:57 anarsoul: just write coalescing is left, and I'll probably try to tackle it from NIR side
06:57 anarsoul: or rather mov coalescing
06:59 anarsoul: btw, this particular deqp shader: https://gist.github.com/anarsoul/2549ae5bbca365c31a36bf32491885a9 is now compiled into just 6 instructions: https://gist.github.com/anarsoul/43f83abe9c93c9694a7c7841c34e22ab
07:00 anarsoul: blob compiles it into 4, so there is still some room for improvement
07:00 anarsoul: yet 12 -> 6 is massive improvement
07:31 anarsoul: !33754 improves glmark2 "-b loop:fragment-steps=5:fragment-uniform=true:vertex-steps=5" by 10% (50 -> 55 fps)
19:24 anarsoul: somehow mpv with "--gpu-context=x11egl" is way smoother for me than with wayland. Weird :)
19:24 anarsoul: e.g. no drops playing sw-decoded 720p video on pinebook with x11egl, and significant drops on wayland
19:26 linkmauve: anarsoul, both on the same compositor, with the only variable being using Xwayland or not?
19:26 anarsoul: yeah
19:26 anarsoul: both on sway
19:27 anarsoul: tbh I don't want to dig into mpv internals again
19:27 linkmauve: Maybe just open them an issue then?
19:27 anarsoul: I have a strong suspicion that mpv does something wrong again
19:27 anarsoul: they haven't fixed previous issue yet :)
19:28 anarsoul: https://github.com/mpv-player/mpv/issues/12968
19:29 linkmauve: I wish the patches to use V4L2 requests in ffmpeg would get merged someday this decade.
19:30 anarsoul: jernej: ^^ :)
19:30 linkmauve: Otherwise reproducing this error is hard, as the latest branch from Kwiboo only applies on 6.something and the API changed enough that current mpv master doesn’t build against it.
19:30 anarsoul: perfectionism is what kills a lot of opensource projects
19:31 anarsoul: it will *never* be perfect. Merge it as is, fix it later if necessary
19:31 linkmauve: Which reminds me I still have a JPEG driver to rebase and ship in Linux.
19:32 anarsoul: I guess unless some big crop starts pushing v4l2-request into ffmpeg it'll never happen
19:33 anarsoul: anyway, I tested my latest lima changes on a pinebook and I don't see any issues
19:34 linkmauve: Kwiboo did a lot of work to fix it last year, I still believe~
19:34 linkmauve: Do you have a branch with everything applied? I could test it on my PinePhone and various Olimex Lime boards.
19:34 anarsoul: supertuxkart seems to be a bit smoother (mostly sits in 30-ish FPS for levels with non-complex geometry)
19:35 anarsoul: linkmauve: https://gitlab.freedesktop.org/anarsoul/mesa/-/tree/lima-fortest
19:35 linkmauve: Ta.
19:36 anarsoul: what's interesting, if I run glmark2 on drm instead of x11 (glmark2 doesn't work on wayland for me) I get double the score in some benchmarks
19:36 anarsoul: while on x11 it's way more humble
19:37 anarsoul: (it's actually Xwayland on weston, not plain x11)
19:38 anarsoul: and it makes sense for drm I guess, since some fragment shaders are now half the size with the latest changes
19:46 anarsoul: linkmauve: btw you don't really need V4L2 request to reproduce the issue. I can easily reproduce it with libva if I force mpv to use GLES2
19:47 anarsoul: it's basically just the gl2/gles2 code in mpv bit rotting since it's hard to find the hardware that doesn't support GL3 or at least ARB_framebuffer_object
19:57 anarsoul: and I don't really understand why mpv wants ARB_framebuffer_object, they don't use depth buffer as far as I can tell
19:58 anarsoul:shrugs
20:05 jernej: linkmauve, anarsoul: we all do :)
20:06 jernej: there was little to no response on patches and latest kwiboo work is miles ahead of previous
20:09 jernej: any corp backing is with gstreamer, which has proper V4L2 request support for years now
20:10 anarsoul: I guess I am old fashioned. I use mpv
20:10 anarsoul: (and I don't like gst, it has terrible api)