FS#56724 - X fails to start with nvidia-340xx-lts and linux-lts
Attached to Project:
Arch Linux
Opened by Martin (qso) - Thursday, 14 December 2017, 16:32 GMT
Last edited by Doug Newgard (Scimmia) - Friday, 15 December 2017, 16:30 GMT
Opened by Martin (qso) - Thursday, 14 December 2017, 16:32 GMT
Last edited by Doug Newgard (Scimmia) - Friday, 15 December 2017, 16:30 GMT
|
Details
Description:
After recent linux-lts update (4.9.68-1) X does not start anymore. I use nvidia-340xx-lts (340.104-5). No problems with linux and nvidia-340xx packages. dmesg shows: [ 4.099415] nvidia: module license 'NVIDIA' taints kernel. [ 4.099417] Disabling lock debugging due to kernel taint [ 4.105782] nvidia: disagrees about version of symbol drm_gem_prime_handle_to_fd [ 4.105784] nvidia: Unknown symbol drm_gem_prime_handle_to_fd (err -22) [ 4.105817] nvidia: disagrees about version of symbol drm_gem_private_object_init [ 4.105819] nvidia: Unknown symbol drm_gem_private_object_init (err -22) [ 4.105934] nvidia: disagrees about version of symbol drm_gem_handle_create [ 4.105935] nvidia: Unknown symbol drm_gem_handle_create (err -22) [ 4.105951] nvidia: disagrees about version of symbol drm_gem_object_unreference_unlocked [ 4.105952] nvidia: Unknown symbol drm_gem_object_unreference_unlocked (err -22) [ 4.106090] nvidia: disagrees about version of symbol drm_gem_prime_export [ 4.106091] nvidia: Unknown symbol drm_gem_prime_export (err -22) |
This task depends upon
Closed by Doug Newgard (Scimmia)
Friday, 15 December 2017, 16:30 GMT
Reason for closing: Duplicate
Additional comments about closing: FS#56721
Friday, 15 December 2017, 16:30 GMT
Reason for closing: Duplicate
Additional comments about closing:
dmesg | grep nvidia shows: https://pastebin.com/EJw5Serr
I can start X and get into Gnome, but it appears that hardware rendering is not working. Scrolling through a web page in Chromium makes the CPU go to 100%. Ran the Unigine Heaven benchmark and it's practically frozen. I get like 2 fps and I have to kill it from the command line because it becomes unresponsive.
I downgraded linux-lts and linux-lts-headers to their previous version (4.9.67-1) and everything goes back to normal.
dmesg | grep nvidia (before the downgrade) looks similar to what OP posted.
Here's its output after the downgrade:
[ 2.993656] nvidia-nvlink: Nvlink Core is being initialized, major device number 242
[ 3.076583] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 387.34 Tue Nov 21 02:09:45 PST 2017
[ 3.078620] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
[ 4.095991] nvidia-modeset: Allocated GPU:0 (GPU-f40c1ed4-d02f-967a-bbb0-a7fd7de12e7d) @ PCI:0000:01:00.0