00:22pl44c: alright so I switched to using dvi to hdmi adapters and managed to get three displays up
00:23pl44c: if modesetting has tearing is there something like "tearfree" or do I have to force X to load nouveau and the intel driver
01:50gnarface: pl44c: there's vsync, not sure about tearfree
02:13pl44c: gnarface: if it makes videos and scrolling not have nasty lines that's all I'm looking for
02:15pl44c: also I had the system reboot itself when I had the laptop lid shut and the screen locked with xscreensaver
02:16gnarface: pl44c: well the nouveau driver has the xorg.conf option available: Option "GLXVBlank" "boolean"
02:16gnarface: that will enable it by default for glx windows anyway, if it's not
02:16gnarface: i don't know if there's a way with nouveau/mesa to accomplish that without glx
02:16gnarface: you may have to resort to running a desktop compositor to gain the vsync capabilities everywhere
02:17gnarface: you can check on individual programs just by setting (i think) vblank_mode=1
02:17gnarface: (as an environment variable that's inexplicably lower-cased in spite of tradition)
02:19pl44c: ok so after rebooting and I didn't change anything running xrandr --setprovideroffloadsink 1 0 did not make the other displays appear with xrandr
02:20gnarface: `man nouveau` for the drivers xorg.conf options, btw
02:22pl44c: oddly enough the lvds display still jitters when I replug a monitor
02:22pl44c: would switching to intel/nouveau ddx drivers rather than modesetting fix this potentially
02:24gnarface: if there's vsync bugs, switching to the intel driver may dodge them, yes. also, it may have different defaults.
02:24pl44c: well I'm talking specifically about the display bug where I can't see the other monitors in xrandr anymore
02:25pl44c: still using modesetting defaults
02:37gnarface: oh, right, sorry
02:41pl44c: tried switching to intel/nouveau with a basic config
02:41pl44c: same issue
02:44pl44c: https://dpaste.de/QpRL xorg.conf
02:44pl44c: *xorg log
02:44pl44c: can't find a pastebin for dmesg as the thing is too big
02:44pl44c: still have verbose debugging enabled for drm
02:46pl44c: oh it's 12Mb
02:46pl44c: no wonder
02:46pl44c: I was just piping it into xsel
02:54pl44c: heres less verbose dmesg https://dpaste.de/HPaG
02:56pl44c: any ideas as to what is causing this?
02:56pl44c: seeing the outputs on the nvidia card worked fine till the system reboted itself after I had shut the lid and left it on xscreensaver for awhile
02:57pl44c: I had issues instantly with the non-free driver doing that instantly and but it wasn't till I came back that I noticed that it shutdown
02:58pl44c: with nouveau
03:02gnarface: i haven't looked at the paste but there are known power management issues with a lot of cards...
03:02gnarface: suspend/resume functionality not working right would be an expected type of misbehavior
03:02pl44c: is there a way to disable suspend on lid shut
03:02pl44c: and just have it shut off the lcd
03:02gnarface: i don't know about specifics, it would be different for different distros
03:03gnarface: if you're using systemd, maybe it's in the systemd config somewhere, otherwise, probably check out acpi?
03:04gnarface: there may be module options relevant too...
03:04gnarface: check out `/sbin/modinfo nouveau`
03:04gnarface: specifically, this line: parm: ignorelid:Ignore ACPI lid status (int)
03:05gnarface: you might have to do the same for the intel one too, i dunno
03:05gnarface: or, not the same literally, but something similarl
03:06pl44c: don't see any mentions of lid or acpi in i915
03:07pl44c: would this be in thinkpad_acpi?
03:08pl44c: more importantly is why the output are gone entirely
03:08pl44c: *outputs are
03:11pl44c: like they interact
03:12pl44c: I see the screen jitter when I replug one
03:12pl44c: but xrandr doesn't see them
03:14pl44c: switching form modesetting to intel/nouveau didn
03:15pl44c: 't change anything
03:21pl44c: rebooted and took the battery out and the output still aren't showing in xrandr after setting them as an output source
03:21pl44c: no idea why it would just start doing this
03:23gnarface: i'm honestly not sure anyone does
03:23gnarface: have any idea
03:23pl44c: so is should I just scrap the idea of using this laptop
03:24gnarface: i'm not saying that. it's one of the possibilities, but someone who knows more than me might have a suggestion. i've avoided this type of dual-gpu laptop hardware myself.
03:25gnarface: there might be a fix or workaround i'm just not aware of.
03:26gnarface: (note that i am not a driver developer)
03:27gnarface: you should be able to at least disable the sleep action or the lid switch
03:27gnarface: i'm just not sure how
03:27pl44c: well now that I'm not driving any displays I kinda want it to actually sleep :/
03:27gnarface: what kernel were you using?
03:28gnarface: newer kernels are always worth a try
03:28gnarface: oh, that's pretty new, nevermind
03:28pl44c: yeah I compiled it yesterday
03:28pl44c: seems that rc4 is a thing now
03:28pl44c: I'm also on debian sid so I'm pretty sure I"m as new as I want in the userland
05:04rhyskidd: karolherbst: a reminder on the pci expansion rom patches https://github.com/envytools/envytools/pull/178
09:27abderrahim[m]: Hi all
09:28abderrahim[m]: trying to launch gnome-mpv I get this error ../src/gallium/drivers/nouveau/nouveau_vp3_video.c:91: nouveau_vp3_video_buffer_create: Assertion `templat->interlaced' failed.
09:28abderrahim[m]: How can I debug it?
12:06Tazmain: Hi all, is there a list of supported card somewhere I can look up? It seems the 10series and RTX isn't supported?
12:25Laalf: Tazmain: on the website itself is a list of supported hardware. To give you a quick rundown: working good is 7xx and before
12:26Tazmain: This website https://nouveau.freedesktop.org/wiki/? maybe I am missing it ?
12:27Tazmain: This does show it https://nouveau.freedesktop.org/wiki/CodeNames/#nv130familypascal, but doesn't like which have issues
12:28Tazmain: Laalf, but okay so 7xx and before works well?
12:28Laalf: Tazmain: not at 100% speed but pretty much
12:29Tazmain: Ah RTX not supported? I had issues with mobile 1050Ti, driver kept crashing
12:50HdkR: Tazmain: TitanV(Volta) and 20 series(Turing) are both completely unsupported :)
12:51Tazmain: but the 1050Ti is listed under Pascal ?
12:51Tazmain: and that just crashes for me
12:53HdkR: Yep, 1050ti is Pascal.
12:54HdkR: "Just crashes" is hard to diagnose, but Maxwell2 and Pascal have perf issues regardless even if it worked
13:01Tazmain: HdkR, ah, okay, well I could not get xorg to start, the driver would crash. So kinda gave up, sorry I don't have any logs I can show to see what the error was
13:04imirkin: abderrahim[m]: hmmm ... that means that somehow it's requesting non-interlaced output, which is not supported
13:04imirkin: should figure out how that's happening and fix up the relevant state tracker to not do that
16:05karolherbst: imirkin: when do you think will you be able to take a deeper look at the current num_samplers handling? Would be nice if users could end up using nine again ;)
16:06karolherbst: allthough I am quite confident that my latest version of the patch is actually fine (at least no piglit regressions, but that doesn't say much here I guess(
23:34joepublic: I have three "wierd things" since switching from AMD/AMDGPU to Nvidia/nouveau, possibly related, possibly not.
23:36joepublic: #1 - when I turn off my monitors (3 monitor system), then turn them back on when I return, the Displayport-connected one comes up, DVI-I one comes up, but the HDMI one is off until I run xrandr to reactivate it.
23:38joepublic: #2 - under LXDE and XFCE (only ones I have tried) menus will not pull down on the monitor connected by displayport. they open on another monitor, away from the app. It's like that monitor is not really considered part of the desktop somehow. The screensaver likewise blanks the HDMI and DVI monitors but doesn't blank that one connected by Displayport.
23:39joepublic: #3 - when first booting, the displayport shows "disconnected" in xrandr no matter what (even though it's connected), until I unplug and replug the displayport cable, at which point it shows connected.
23:40joepublic: If the developers here recognize any of these behaviors as FAQs (or odd things because I have a *three* monitor desktop on an EVGA GTX 760 instead of the usual one or two) I would greatly appreciate their insight. If not, then hey, that's how it goes.