[10:26] furkan: ah, that's nice. i'll add it after easter [10:28] darkxst: ok, thanks [10:33] tseliot, np, let me know if you need testing of updated packages (though I am stuck on intel only laptop for easter break) [10:36] darkxst: sure, thanks [10:39] I'll dig into plymouth a bit more when I get back, 99.9% sure it should work, so either an issue with intialization of DRM or a bug in plymouth [11:14] ricotz, mamarley: now I do get cat /sys/module/nvidia_drm/parameters/modeset -> Y . The only problem is that all I can see is a black screen. Something must have broken hybrid graphics [11:16] tseliot what DE? [11:16] darkxst: I always use unity [11:18] I've not tried that yet, but probably best to leave the modeset option out by default [11:18] for now [11:18] Hmm, maybe the new DRM KMS implementation is incompatible with the existing PRIME implementation. I haven't tried it with an optimus system though. [11:19] mutter segfaults also nothing to do with PRIME [11:19] it's certainly incompatible with SLI. Who knows what else that breaks. I'll report that to nvidia [11:19] darkxst: mutter with Wayland? [11:19] tseliot, no there is no code for that yet [11:20] mutter on X11 [11:20] as that would need some patches that NVIDIA published [11:20] oh [11:20] hmm... [11:21] tseliot, it could also be GNOME bug, but dont force the kernel option on until things are working [11:21] just fix alias, and preloading of initramfs [11:21] I won't. I'm changing it back [11:21] :/ [11:22] the thing is I only have a hybrid system with nvidia. No (functioning) nvidia-only system. [11:23] and I only have a nvidia-only system (no hybrid) [11:24] and its my main development machine so a huge pain when I have to reboot [11:25] heh, no nvidia or any kind of testing on my workstation. I even run trusty here. [11:25] the testing boxes run xenial or whatever else I need [11:26] tseliot, I see, please at least paste a debdiff [11:27] darkxst, heh, same here, nvidia-only dev-machine ;) [11:27] xenial + some ppas is where the fun begins [11:28] ricotz: I can probably do better than that. I'll remove the modeset=1 argument from the udev rule, and keep it in the modprobe file, set to 0. This way it's easy change it back to 1 for testing [11:28] I actually have a system that supports Optimus, but it has a hardware mux and I leave it in NVIDIA-only mode because I hate tearing. [11:28] tseliot, alright, then push it that way [11:28] good old multiplexers :) [11:29] ok, good [11:29] tseliot, don't sit too long on such fixes ;) [11:30] ricotz, I just run "devel" + staging on my dev machine, doesnt break often [11:31] If I understand things correctly, the DRM KMS support will eventually allow for "proper" Optimus (rendering only the game/whatever on the NVIDIA card and supporting vsync). [11:31] darkxst, I better don't tell what I am pulling in here [11:31] hmm, hardware multiplexers don't work with prime [11:31] ricotz, don't worry we all know! [11:32] you have a system of git snapshots ;) [11:33] it isn't that worse ;P [11:33] mamarley: yes, that and probably SLI too. And yes, tearing should go away, that will need kernel patches too. [11:43] * tseliot finally working on the initramfs hook [12:08] ok, done. Rebuilding and testing now [12:09] well, after lunch... [12:09] * tseliot -> lunch [15:48] ricotz, mamarley: I have the initramfs hook and all the fixes here. I am going to upload all that to the PPA [15:49] it works fine (with modesetting disabled) here [16:10] uploaded [16:11] the instructions on how to enable modesetting are in the changelog [16:11] darkxst: ^ [16:39] tseliot: Cool, thanks! [16:40] also, I am going to be off next week [16:53] * tseliot EOW \o/