[Solved] Nvidia NV ram not working using 470-dkms

General system
1

Using GT740 4GB card.
New install
KDE
resently started using 470-dkms drivers

I started by noticing things in desktop running slugish last few days.

I got my photoshop up and running, and the 3D section tells me Vram not working.

Odd outputs of info

 ~]$ sudo dmesg|grep VRAM
[sudo] password for  
 ~]$ sudo dmesg|grep GART
 ~]$ sudo dmesg|grep BAR
[   19.204159] caller _nv033181rm+0x2a/0x60 [nvidia] mapping multiple BARs
[   19.384763] caller _nv000722rm+0x1ad/0x200 [nvidia] mapping multiple BARs
[major@beholder ~]$ 

$ sudo modprobe nvidia-dkms
modprobe: FATAL: Module nvidia-dkms not found in directory /lib/modules/5.10.80-1-lts

$ lsmod | grep nvidia
nvidia_drm             69632  5
nvidia_modeset       1204224  18 nvidia_drm
drm_kms_helper        278528  1 nvidia_drm
nvidia_uvm           1187840  0
nvidia              35360768  1099 nvidia_uvm,nvidia_modeset
drm                   577536  9 drm_kms_helper,nvidia,nvidia_drm

$ pacman -Qs nvidia
local/conky-lua-nv 1.12.2-1
    An advanced system monitor for X based on torsmo with lua and nvidia enabled
local/egl-wayland 1:1.1.9+2+gdaab854-1
    EGLStream-based Wayland external platform
local/lib32-nvidia-470xx-utils 470.86-1
    NVIDIA drivers utilities (32-bit)
local/libvdpau 1.4-1
    Nvidia VDPAU library
local/libxnvctrl-470xx 470.86-1
    NVIDIA NV-CONTROL X extension
local/nvidia-470xx-dkms 470.86-3
    NVIDIA drivers - module sources
local/nvidia-470xx-settings 470.86-1
    Tool for configuring the NVIDIA graphics driver
local/nvidia-470xx-utils 470.86-3
    NVIDIA drivers utilities
local/nvidia-installer-db 2.5.1-2
    Database for the script to setup nvidia drivers in EndeavourOS
local/nvidia-installer-dkms 3.3.11-1
    Script to setup nvidia drivers (dkms version) in EndeavourOS

My /etc/X11/xorg.conf.d directory only has a keyboard.conf file in it if that matters.

I saw something something opengl at one point as well, idk if related as I was setting up WINE to get photoshop going in the first place.

I need some direction on want to do

2

What is your video card?

inxi -Ga
3

What were you using before™, what changed, and what things are running sluggish?

Photoshop isn’t a Linux-native application - did this work correctly before™, and has the version of WINE changed?

nvidia-dkms is a driver package name. Try modprobe nvidia or modinfo nvidia.

4

This what i get on mine. GTX 1060 Mvidia 495.44

[ricklinux@eos-xfce ~]$ sudo dmesg|grep BAR
[sudo] password for ricklinux: 
[    0.240704] pci 0000:03:00.0: BAR 1: assigned to efifb
[    0.292729] pci 0000:00:1b.0: BAR 14: assigned [mem 0x90000000-0x901fffff]
[    0.292733] pci 0000:00:1b.0: BAR 15: assigned [mem 0x2000000000-0x20001fffff 64bit pref]
[    0.292735] pci 0000:00:1d.0: BAR 14: assigned [mem 0x90200000-0x903fffff]
[    0.292738] pci 0000:00:1d.0: BAR 15: assigned [mem 0x2000200000-0x20003fffff 64bit pref]
[    0.292739] pci 0000:00:1b.0: BAR 13: assigned [io  0x2000-0x2fff]
[    0.292741] pci 0000:00:1d.0: BAR 13: assigned [io  0x3000-0x3fff]
[    0.367710] caller pmc_core_probe+0xb3/0x6a0 mapping multiple BARs
[ricklinux@eos-xfce ~]$ 
 

[ricklinux@eos-xfce ~]$ inxi -Ga
Graphics:  Device-1: Intel CometLake-S GT2 [UHD Graphics 630] vendor: ASUSTeK driver: i915
           v: kernel bus-ID: 00:02.0 chip-ID: 8086:3e92 class-ID: 0300
           Device-2: NVIDIA GP104 [GeForce GTX 1060 6GB] vendor: ASUSTeK driver: nvidia
           v: 495.44 alternate: nouveau,nvidia_drm bus-ID: 03:00.0 chip-ID: 10de:1b83
           class-ID: 0300
           Display: x11 server: X.Org 1.21.1.1 compositor: xfwm4 v: 4.16.1 driver:
           loaded: nvidia display-ID: :0.0 screens: 1
           Screen-1: 0 s-res: 1920x1080 s-dpi: 101 s-size: 483x272mm (19.0x10.7")
           s-diag: 554mm (21.8")
           Monitor-1: DVI-D-0 res: 1920x1080 hz: 60 dpi: 102 size: 477x268mm (18.8x10.6")
           diag: 547mm (21.5")
           OpenGL: renderer: NVIDIA GeForce GTX 1060 6GB/PCIe/SSE2 v: 4.6.0 NVIDIA 495.44
           direct render: Yes
[ricklinux@eos-xfce ~]$
5

$ inxi -Ga
Graphics: Device-1: NVIDIA GK107 [GeForce GT 740] driver: nvidia v: 470.86 alternate: nouveau,nvidia_drm bus-ID: 01:00.0
chip-ID: 10de:0fc8 class-ID: 0300
Display: x11 server: X.org 1.21.1.1 compositor: kwin_x11 driver: loaded: nvidia resolution: <missing: xdpyinfo>
OpenGL: renderer: NVIDIA GeForce GT 740/PCIe/SSE2 v: 4.6.0 NVIDIA 470.86 direct render: Yes

1 Like
6

[ricklinux@eos-xfce ~]$ cat /etc/X11/xorg.conf.d/20-nvidia.conf | eos-sendlog

https://clbin.com/Hfgki

7

LMAO idk, head explodes :stuck_out_tongue:

$ cat /etc/X11/xorg.conf.d/20-nvidia.conf | eos-sendlog
cat: /etc/X11/xorg.conf.d/20-nvidia.conf: No such file or directory
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   453    0   313  100   140   1026    459 --:--:-- --:--:-- --:--:--  1490
<!doctype html><html lang="en"><head><title>Error: not string - clbin: command line pastebin</title><script src="/s/hl.js"></script><script>hljs.initHighlightingOnLoad();</script><link href="/s/sb.css" rel="stylesheet" type="text/css"></head><body><pre><code><h4>not string</h4>
</code></pre></div></body></html>
8

I know what it wants to do but I don’t have that file.

9

This file gets automatically set up on the install of nvidia drivers normally. I’m on the 495.44 drivers but i would assume it still needs an xorg conf file of some kind on any of the driver versions? @jonathon

10

Yeah, I figured something was missing, or a conf file someplace else it’s seeing first, but not sure how to track it down

1 Like
11

Im sure it was me trying to fix something I should have asked about before I did it.

I did at one point last week or so during all this upgrading
$ sudo nvidia-xconfig

So I do have a
‘/etc/X11/xorg.conf’ file

but not a
/etc/X11/xorg.conf.d/20-nvidia.conf file

12

I don’t think you are supposed to use this?

sudo nvidia-xconfig

The top part of my file is this

Section "Device"
    Identifier "Nvidia Card"
    Driver "nvidia"
    VendorName "NVIDIA Corporation"
    Option "NoLogo" "true"
EndSection

That is what is created automatically. The bottom part of the file i added as additional enhancements. Why not try deleting the xorg file and creating this one with the minimal needed to try. Make sure its in /etc/X11/xorg.conf.d/20-nvidia.conf You can just rename the other one temporarily or something.

1 Like
13

I safely rebooted with that file, so far so good
edit:
nothings changed, but this is a pretty red color in the konsol :stuck_out_tongue:

$ journalctl -p 3 -xb
-- Journal begins at Sat 2021-11-20 00:07:04 EST, ends at Sun 2021-11-21 23:26:13 EST. --
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT3._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT3._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT0._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT0._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT3._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT3._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT0._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT0._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT1._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT1._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:12 beholder kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT1._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Nov 21 23:19:12 beholder kernel: ACPI Error: Aborting method \_SB.PCI0.SAT0.SPT1._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Nov 21 23:19:14 beholder kernel: 
Nov 21 23:19:53 beholder pulseaudio[1085]: GetManagedObjects() failed: org.freedesktop.systemd1.NoSuchUnit: Unit dbus-org.bluez.service not found.
Nov 21 23:21:51 beholder dolphin[3413]: inotify_add_watch(/etc/sudoers.d) failed: (Permission denied)

just saying, its there

14

what about enable early load for nvidia DRM modesetting?
easy way is to add it to grub kernel line:

1. Adding it to grub kernel line:

to change configuration and rebuild grub.cfg and kernel boot images :

Force nvidia-drm.modeset=1:

adding nvidia-drm.modeset=1 to the kernel parameter inside grub:

sudo nano /etc/default/grub

and put nvidia-drm.modeset=1 inside (leave the rest untouched!!) as you can see here:

# GRUB boot loader configuration

GRUB_DEFAULT=0 GRUB_TIMEOUT=5

GRUB_DISTRIBUTOR="EndeavourOS"

GRUB_CMDLINE_LINUX_DEFAULT="nvidia-drm.modeset=1 resume=UUID=...

save the file [Ctrl+X] and rebuild the grub.cfg:

sudo grub-mkconfig -o /boot/grub/grub.cfg

check what modules are used:

cat /proc/modules | grep nvidia

nvidia_drm 73728 4 - Live 0x0000000000000000 (POE)
nvidia_modeset 1204224 8 nvidia_drm, Live 0x0000000000000000 (POE)
nvidia_uvm 2564096 0 - Live 0x0000000000000000 (POE)
nvidia 35377152 379 nvidia_modeset,nvidia_uvm, Live 0x0000000000000000 (POE)
15

shows: alternate: nouveau,nvidia_drm

16

But this isn’t about the nvidia. So that part is good! I don’t like the errors though. But i don’t think that can be resolved unless there would be a Bios update and i think this motherboard is too old.

ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT0.SPT3._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)

Edit: As long as the system is working okay then the errors should not be an issue.

17

There are a lot of system changes being made here without understanding what the underlying issue actually is…

So, going back to the OP, how do you know NV ram is not working, did it work at any point, and if so what changed?

18

@joekamprad Checking a few more things after I got home.
In relation to your post about modsetting

I ran

$ nvidia-smi
Mon Nov 22 17:58:45 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.86       Driver Version: 470.86       CUDA Version: 11.4     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:01:00.0 N/A |                  N/A |
| 30%   33C    P0    N/A /  N/A |    640MiB /  4032MiB |     N/A      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+

The

 640MiB /  4032MiB

should be my vram, it says it’s there and using some, but it looks to me it’s not really displaying correctly idk

$ inxi -F
CPU:       Info: Quad Core model: Intel Core i5-2400 bits: 64 type: MCP cache: L2: 6 MiB
           Speed: 1755 MHz min/max: 1600/3400 MHz Core speeds (MHz): 1: 1755 2: 1811 3: 2105 4: 1965
Graphics:  Device-1: NVIDIA GK107 [GeForce GT 740] driver: nvidia v: 470.86
           Display: x11 server: X.org 1.21.1.1 driver: loaded: nvidia unloaded: fbdev,modesetting,vesa
           resolution: <missing: xdpyinfo>

Is this

loaded: nvidia unloaded: fbdev,modesetting,vesa
           resolution: <missing: xdpyinfo>

what the grub modsetting is supposed fix?

Then you mentioned
shows: alternate: nouveau

Does this mean I have another video driver installed?

Some of the outputs I see kinda tell me it’s half ass running, but messed up somehow so I’m trying to grasp what is going on as I’ve never had this issue.
Before I do anything, here is something I found in logs

Nov 22 17:51:06 _hostname_ kernel: Disabling lock debugging due to kernel taint
Nov 22 17:51:06 _hostname_ kernel: nvidia: module verification failed: signature and/or required key missing - tainting kernel
Nov 22 17:51:06 _hostname_ kernel: nvidia-nvlink: Nvlink Core is being initialized, _user_ device number 239
Nov 22 17:51:06 _hostname_ kernel: 
Nov 22 17:51:06 _hostname_ kernel: nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=io+mem

And current

$ cat /proc/modules | grep nvidia
nvidia_drm 69632 8 - Live 0x0000000000000000 (POE)
nvidia_modeset 1204224 23 nvidia_drm, Live 0x0000000000000000 (POE)
nvidia_uvm 1187840 0 - Live 0x0000000000000000 (POE)
drm_kms_helper 278528 1 nvidia_drm, Live 0x0000000000000000
nvidia 35360768 1247 nvidia_modeset,nvidia_uvm, Live 0x0000000000000000 (POE)
drm 577536 12 nvidia_drm,drm_kms_helper,nvidia, Live 0x0000000000000000

Idk what it all means, just trying to supply for info

19

btw @ricklinux I tried that again even though I have the file it still doesn’t work,
but it does work with eos log from /home

$ cat /home/major/eos-log-tool.logs | eos-sendlog
...
https://clbin.com/sJNlc
20

Ya some desktop i notice put the log there.