FS#50067 - [nvidia] segfault libGLX_nvidia.so.367.35

Attached to Project: Arch Linux
Opened by xeboc (xeboc) - Saturday, 16 July 2016, 09:54 GMT
Last edited by Sven-Hendrik Haase (Svenstaro) - Monday, 22 August 2016, 19:34 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To Sven-Hendrik Haase (Svenstaro)
Felix Yan (felixonmars)
Architecture All
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 11
Private No

Details

After the yesterday update (nvidia 367.35) I'm experiencing a segfault in libGLX_nvidia.so
In attach there is the optirun verbouse output, Bumblebee log and dmesg output with this:

[ 796.485183] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 367.35 Mon Jul 11 22:43:59 PDT 2016
[ 796.923041] glxspheres64[1703]: segfault at 10 ip 00007f66c80be7ad sp 00007ffc97534670 error 4 in libGLX_nvidia.so.367.35[7f66c8001000+c8000]
[ 797.033123] ACPI Warning: \_SB.PCI0.PEG0.PEGP._DSM: Argument #4 type mismatch - Found [Buffer], ACPI requires [Package] (20160108/nsarguments-95)
[ 797.038368] nvidia-modeset: Unloading

You can see the attach also here: http://pastebin.com/raw/N8aaXhuC

Some versions:

virtualgl 2.5-2
bumblebee 3.2.1-11
mesa 12.0.1-1
mesa-libgl 12.0.1-1
nvidia 367.35-1
bbswitch 0.8-50


more details:

lspci -k | grep -A 2 -E "(VGA|3D)"
00:02.0 VGA compatible controller: Intel Corporation 4th Gen Core Processor Integrated Graphics Controller (rev 06)
Subsystem: CLEVO/KAPOK Computer Device 2316
Kernel driver in use: i915
--
01:00.0 3D controller: NVIDIA Corporation GM107M [GeForce GTX 960M] (rev ff)
Kernel modules: nouveau, nvidia_drm, nvidia





Any suggestion?

need more information?
This task depends upon

Closed by  Sven-Hendrik Haase (Svenstaro)
Monday, 22 August 2016, 19:34 GMT
Reason for closing:  Fixed
Comment by xeboc (xeboc) - Saturday, 16 July 2016, 10:05 GMT
wrong project, sorry first time i'm using this tracker
Comment by Dan (sdlnv) - Saturday, 16 July 2016, 14:11 GMT
Having the same problem: https://bbs.archlinux.org/viewtopic.php?pid=1641635#p1641635
Comment by Tom Ebeling (Bollie) - Tuesday, 19 July 2016, 06:25 GMT
Having the same issue here with optirun. I downgraded and it worked again.
Comment by seba (generic) - Tuesday, 19 July 2016, 12:35 GMT
same problem
$ optirun blender
connect failed: No such file or directory
Read new prefs: /home/seba/.config/blender/2.77/config/userpref.blend
connect(2) call to /dev/shm/jack-1000/default/jack_0 failed (err=No such file or directory)
attempt to connect to server failed
Writing: /tmp/blender.crash.txt

$ cat /tmp/blender.crash.txt
# Blender 2.77 (sub 0), Commit date: 2016-04-05 18:12, Hash abf6f08

# backtrace
blender(BLI_system_backtrace+0x30) [0x1374c90]
blender() [0x9e5a59]
/usr/lib/libc.so.6(+0x33310) [0x7f658cac5310]
/usr/lib/libGLX_nvidia.so.0(+0xbd7ad) [0x7f656637f7ad]
Comment by Alex Bruce (manghoti) - Thursday, 21 July 2016, 00:57 GMT
I can also confirm this bug, just updated today and got hit by it, downgrade to nvidia-367.27-1-x86_64.pkg.tar.xz fixed the problem.

Not sure what's going on.

Comment by Anton Maminov (mamantoha) - Friday, 22 July 2016, 01:28 GMT
I can also confirm this bug. It happening to me when running Steam games with optirun. Downgrade to 367.27-1 helps me.
Comment by Tom Ebeling (Bollie) - Sunday, 14 August 2016, 07:53 GMT
This unfortunately doesn't allow me to update to linux-4.7.

$ optirun gdb glxspheres64
GNU gdb (GDB) 7.11.1
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-pc-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from glxspheres64...(no debugging symbols found)...done.
(gdb) run
Starting program: /usr/bin/glxspheres64
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Polygons in scene: 62464 (61 spheres * 1024 polys/spheres)

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff4c467ad in ?? () from /usr/lib/libGLX_nvidia.so.0

linux 4.7-1
nvidia 367.35-2
bumblebee 3.2.1-12



Comment by Jocelyn Meyron (nyorem) - Monday, 15 August 2016, 16:31 GMT
Having the same issue.
Comment by Tom Ebeling (Bollie) - Monday, 15 August 2016, 19:57 GMT
Just noticed, that installing primus seems to help a little.

optirun -b primus glxspheres64
Polygons in scene: 62464 (61 spheres * 1024 polys/spheres)
Visual ID of window: 0xd4
Context is Direct
OpenGL Renderer: GeForce GTX 960M/PCIe/SSE2
61.546549 frames/sec - 68.685949 Mpixels/sec
60.008649 frames/sec - 66.969652 Mpixels/sec
60.009287 frames/sec - 66.970365 Mpixels/sec
primus: warning: dropping a frame to avoid deadlock
primus: warning: timeout waiting for display worker

The shutdown doesn't seem to be clean and it still crashes with more complex opengl applications.

Starting and stopping optirun -b primus glxgears leads to this:
Process 5166 (glxgears) of user 1000 dumped core.

Stack trace of thread 5166:
#0 0x00007fa51c5dc7ad n/a (libGLX_nvidia.so.0)
#1 0x00007fa51c5a50b8 n/a (libGLX_nvidia.so.0)
#2 0x00007fa51c5982b9 glXGetConfig (libGLX_nvidia.so.0)
#3 0x00007fa51eae7666 n/a (libvglfaker.so)
#4 0x00007fa51eae8abd n/a (libvglfaker.so)
#5 0x00007fa51eabf59a glXChooseVisual (libvglfaker.so)
#6 0x00000000004036bf n/a (glxgears)
#7 0x0000000000401a07 n/a (glxgears)
#8 0x00007fa51de6a291 __libc_start_main (libc.so.6)
#9 0x0000000000402349 n/a (glxgears)
Comment by Sven-Hendrik Haase (Svenstaro) - Monday, 15 August 2016, 20:50 GMT
Is there anything I can do here from a packaging perspective?
Comment by Tom Ebeling (Bollie) - Monday, 15 August 2016, 21:05 GMT
Thank you for your reply, Sven. Hm, I can't really tell if that issue is limited to arch. I've just found this issue on Bumblebee Project's github tracker and it only refers to Arch so far: https://github.com/Bumblebee-Project/Bumblebee/issues/786

Comment by Sven-Hendrik Haase (Svenstaro) - Monday, 15 August 2016, 21:12 GMT
Probably because Arch is the only distro that routinely ships the latest drivers. However, I don't really think this is an Arch related problem per-se.
Comment by Darek (blablo) - Tuesday, 16 August 2016, 07:14 GMT
Nvidia 370.23 [1] is out (beta)
[1] http://www.nvidia.com/download/driverResults.aspx/105855/en-us
Comment by Tom Ebeling (Bollie) - Tuesday, 16 August 2016, 08:24 GMT
Wow, thanks for informing us, Darek. I've cloned nvidia-beta and nvidia-utils-beta from AUR and changed the PKGBUILD files to that new driver version. Building the packages, installing and running it with bumblebee-git now works! The only issue that is still left here at my end is, that starting x hangs when bumblebeed has started with bbswitch. But that's probably something to report to bumblebee-project.

Anyway, @Sven, probably you could do us a favor from the packaging perspective and switch to the new NVIDIA driver. ;)

Cheers
Tom
Comment by Sven-Hendrik Haase (Svenstaro) - Monday, 22 August 2016, 19:34 GMT
Alright, seems that switching to a beta driver is actually a sensible thing to do in this case. I will close this issue now and update nvidia to beta.

Loading...