Linux

Hi,

I was heavily playing EVE Online with my linux + PlayOnLinux + primusrun, and...

During the game I was doing some system update, and sudenly after moving a gate something happend to the graphic, so I restarted a game and... after puting a password into EVE login console I had a black screen.

What works - PlayOnLinux works, even EVE works with Intel driver (very slowly and ugly), but doesnt work corectly with primusrun:

$ primusrun playonlinux

when the game should start I have black screen and such error in the console (see bellow) - primus works fine with all other games and only have a problm with PlayOnLinux and my EVE :(

for wine 1.7.32:

primus: fatal: failed to acquire direct rendering context for display thread 

for wine 1.7.33:

X Error of failed request:  GLXBadContext 
   Major opcode of failed request:  154 (GLX) 
   Minor opcode of failed request:  6 (X_GLXIsDirect) 
   Serial number of failed request:  25 
   Current serial number in output stream:  24 

for wine 1.7.33-WGL-ShareList:

X Error of failed request:  GLXBadContext 
   Major opcode of failed request:  154 (GLX) 
   Minor opcode of failed request:  6 (X_GLXIsDirect) 
   Serial number of failed request:  25 
   Current serial number in output stream:  24

Many thanks :)
It looks like it is missing some drivers for Intel? But from which package?

LIBGL_DEBUG=verbose primusrun playonlinux

libGL: OpenDriver: trying /usr/lib32/dri/tls/i965_dri.so
libGL: OpenDriver: trying /usr/lib32/dri/i965_dri.so
libGL: Can't open configuration file /home/ryszard/.drirc: No such file or directory.                                                                  
libGL: OpenDriver: trying /usr/lib32/dri/tls/i965_dri.so
libGL: OpenDriver: trying /usr/lib32/dri/i965_dri.so
libGL: Can't open configuration file /home/ryszard/.drirc: No such file or directory.
libGL: OpenDriver: trying /usr/lib32/dri/tls/i965_dri.so
libGL: OpenDriver: trying /usr/lib32/dri/i965_dri.so
libGL: Can't open configuration file /home/ryszard/.drirc: No such file or directory.
libGL: Can't open configuration file /home/ryszard/.drirc: No such file or directory.
libGL: OpenDriver: trying /usr/lib32/dri/tls/i965_dri.so
libGL: OpenDriver: trying /usr/lib32/dri/i965_dri.so
libGL: Can't open configuration file /home/ryszard/.drirc: No such file or directory.
primus: fatal: failed to acquire direct rendering context for display thread

Have you tried restarting X? you can corrupt several different things with in it. I know this is a crappy answer, but I had this happen to my X session once.

(Restarting the entire computer does this too.)

Not sure what distribution you use, but i suspect you have applied a kernel or x-System update and now your NVIDIA driver don't work anymore since they need to be built against the new source changes of those update components.

If you hat built and installed them manually, just do it again, reboot and it will work again.

If you have installed them by repositories of your distribution, well then the distribution is not the best as i guess it should have foreseen and solved the problem automatically. The distribution needs to either update the drivers or you need to update or reinstall them via repositories.

I think you're digging the wrong way with the "libGL: Can't open" stuff.

It "can't open" some optional configuration file which happens on working installations too.

The possible things to look into are:

1) Is there any other 32bit application working with primusrun. Because from the looks of it what you've posted was a 64bit application.
2) If there are no, then are there 32bit nvidia libs present? Can bumblebee find them?
3) Is there a 32bit primus library for bumblebee to work with?
4) What if you try virtualGL bridge instead of primus?
5) What does dmesg show? After you try to launch eve bbswitch should try to turn your nvdia card on, and then bumblebee should load the nvidia kernel module whuch should be visible in dmesg.

Buggy kernel/Xorg/driver combo?

This is how healthy dmesg looks like

[  273.404576] bbswitch: enabling discrete graphics
[  273.859051] nvidia: module license 'NVIDIA' taints kernel.
[  273.859055] Disabling lock debugging due to kernel taint
[  273.866315] vgaarb: device changed decodes: PCI:0000:01:00.0,olddecodes=io+mem,decodes=none:owns=none
[  273.866558] [drm] Initialized nvidia-drm 0.0.0 20130102 for 0000:01:00.0 on minor 1
[  273.866563] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  331.113  Mon Dec  1 21:08:13 PST 2014
[  274.021327] nvidia 0000:01:00.0: irq 50 for MSI/MSI-X

https://bugs.gentoo.org/show_bug.cgi?id=532082




In bumblebee config, bumblebed section look up driver.
In my case it's Driver=nvidia

Then look up the coniguration section for that driver (in my case it's [driver-nvidia])
In that section there's a colon-separated LibraryPath that should contain the path to both 64bit and 32bit libraries
In my case it is LibraryPath=/usr/lib/nvidia-331:/usr/lib32/nvidia-331

Then of course you need to check if there's something present at the library path.

Here my logs for natural nvidia driver (without bumblebee).

Xorg.0.log (opengl nvidia)
http://pastebin.com/5iMVeETa

Xorg.0.log (opengl xorg-x11)
http://pastebin.com/Z4FU1k9B

At the end of Xorg.0.log (opengl nvidia) we have

[  1944.364] (WW) NVIDIA(GPU-0): Unable to read EDID for display device CRT-0

errors:

[  1942.496] (EE) /dev/dri/card1: failed to set DRM interface version 1.4: Invalid argument

[  1942.513] (EE) Screen 1 deleted because of no matching config section.

[  1943.204] (EE) /dev/dri/card1: failed to set DRM interface version 1.4: Invalid argument

but at the end of Xorg.0.log (opengl xorg-x11) we have

[  2370.741] (WW) NVIDIA(GPU-0): Unable to read EDID for display device CRT-0
[  2418.355] (II) NVIDIA(0): Setting mode "CRT-0:nvidia-auto-select"
[  2418.457] (WW) NVIDIA(GPU-0): Unable to read EDID for display device CRT-0
[  2418.730] (II) modesetting(G0): EDID vendor "CMO", prod id 5913
[  2418.730] (II) modesetting(G0): Printing DDC gathered Modelines:
[  2418.730] (II) modesetting(G0): Modeline "1600x900"x0.0  107.80  1600 1648 1680 1940  900 903 908 926 -hsync -vsync (55.6 kHz eP)

so looks like system manage to get EDID, but why the screen is black?
and errors:

[  2369.092] (EE) /dev/dri/card1: failed to set DRM interface version 1.4: Invalid argument

[  2369.102] (EE) Screen 1 deleted because of no matching config section.

[  2369.103] (EE) NVIDIA(0): Failed to initialize the GLX module; please check in your X
[  2369.103] (EE) NVIDIA(0):     log file that the GLX module has been loaded in your X
[  2369.103] (EE) NVIDIA(0):     server, and that the module is the NVIDIA GLX module.  If
[  2369.103] (EE) NVIDIA(0):     you continue to encounter problems, Please try
[  2369.103] (EE) NVIDIA(0):     reinstalling the NVIDIA driver.

[  2369.697] (EE) AIGLX: reverting to software rendering

[  2369.773] (EE) /dev/dri/card1: failed to set DRM interface version 1.4: Invalid argument

...and there is no nvidia here:

$ xrandr --listproviders
Providers: number : 1
Provider 0: id: 0x47 cap: 0xb, Source Output, Sink Output, Sink Offload crtcs: 3 outputs: 5 associated providers: 0 name:Intel

Hi,

I use /usr/lib64 and /usr/lib32 - system is 64 but for some some software I have 32 option.

This did the job, thank You so MUCH :)
https://forums.gentoo.org/viewtopic-p-7677612.html?sid=60d8d52c44d4a26033ca1beb20572a71
I had to downgrade kernel and nvidia drivers (make some glitches in another games and some less fps than used to be), but EVE works fine :)


This topic is SOLVED :)