Results 1 to 10 of 10

Thread: How to roll back Nvidia driver version until issues pass

  1. #1

    Question How to roll back Nvidia driver version until issues pass

    Greetings:

    The latest Nvidia driver updates have seemingly caused some issues with my system (e.g., I have tickets 1190840 and 1190871 open for them). For situations such as this, it would be nice for me to be able to roll back the version to the previous one and freeze it there until updates come out for testing. I am quite confident Zypper can accommodate this. However, I am new to Zypper and am not sure how to do this. Additionally, it would be helpful knowing what other steps (e.g., manual kernel driver compilation) would need to occur with such a temporary solution. Is there any guidance the community can provide for this? I looked for past threads on this topic but couldn't find anything (and apologize beforehand if I missed some). Thank you for your help with this.

    Kind regards,
    Anthony

  2. #2
    Join Date
    Jun 2008
    Location
    East of Podunk
    Posts
    33,250
    Blog Entries
    15

    Default Re: How to roll back Nvidia driver version until issues pass

    Hi
    Install the nvidia driver the hard way? All good here, however I only use as offload....

    Code:
    inxi -Gxxz
    Graphics:  Device-1: AMD Lexa PRO [Radeon 540/540X/550/550X / RX 540X/550/550X] driver: amdgpu v: kernel bus-ID: 02:00.0 
               chip-ID: 1002:699f 
               Device-2: NVIDIA GP108 [GeForce GT 1030] vendor: eVga.com. driver: nvidia v: 470.74 bus-ID: 03:00.0 
               chip-ID: 10de:1d01 
               Display: x11 server: X.Org 1.20.13 compositor: gnome-shell driver: loaded: amdgpu,nvidia 
               unloaded: fbdev,modesetting,vesa alternate: ati,nouveau,nv resolution: 1: 1920x1080~60Hz 2: 1920x1080~60Hz 
               3: 1920x1080~60Hz s-dpi: 96 
               OpenGL: renderer: Radeon RX550/550 Series (POLARIS12 DRM 3.42.0 5.14.6-1-default LLVM 12.0.1) v: 4.6 Mesa 21.2.2 
               direct render: Yes 
    
    switcherooctl launch -g 1 inxi -Gxxz
    
    Graphics:  Device-1: AMD Lexa PRO [Radeon 540/540X/550/550X / RX 540X/550/550X] driver: amdgpu v: kernel bus-ID: 02:00.0 
               chip-ID: 1002:699f 
               Device-2: NVIDIA GP108 [GeForce GT 1030] vendor: eVga.com. driver: nvidia v: 470.74 bus-ID: 03:00.0 
               chip-ID: 10de:1d01 
               Display: x11 server: X.Org 1.20.13 compositor: gnome-shell driver: loaded: amdgpu,nvidia 
               unloaded: fbdev,modesetting,vesa alternate: ati,nouveau,nv resolution: 1: 1920x1080~60Hz 2: 1920x1080~60Hz 
               3: 1920x1080~60Hz s-dpi: 96 
               OpenGL: renderer: NVIDIA GeForce GT 1030/PCIe/SSE2 v: 4.6.0 NVIDIA 470.74 direct render: Yes

    You must be setting something somewhere for the power setting issue, by default Tumbleweed uses schedutil these days....

    Code:
    cpupower frequency-info 
    
    analyzing CPU 0:
      driver: intel_cpufreq
      CPUs which run at the same hardware frequency: 0
      CPUs which need to have their frequency coordinated by software: 0
      maximum transition latency: 20.0 us
      hardware limits: 1.20 GHz - 3.50 GHz
      available cpufreq governors: ondemand performance schedutil
      current policy: frequency should be within 1.20 GHz and 3.50 GHz.
                      The governor "schedutil" may decide which speed to use
                      within this range.
      current CPU frequency: Unable to call hardware
      current CPU frequency: 1.45 GHz (asserted by call to kernel)
      boost state support:
        Supported: yes
        Active: yes
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  3. #3

    Default Re: How to roll back Nvidia driver version until issues pass

    I only have the Nvidia graphics card to use.

    Code:
    > inxi -Gxxz 
    Graphics:  Device-1: NVIDIA TU106 [GeForce RTX 2070 Rev. A] vendor: eVga.com. driver: nvidia v: 470.74  
               bus-ID: 01:00.0 chip-ID: 10de:1f07  
               Device-2: AVerMedia Live Streamer CAM 313 type: USB driver: snd-usb-audio,uvcvideo bus-ID: 1-7:4  
               chip-ID: 07ca:313a  
               Display: x11 server: X.org 1.20.13 compositor: kwin_x11 driver:loaded: nvidia  
               unloaded: fbdev,modesetting,nouveau,vesa alternate: nv resolution: <missing: xdpyinfo>  
               OpenGL:renderer: NVIDIA GeForce RTX 2070/PCIe/SSE2 v: 4.6.0 NVIDIA 470.74 direct render: Yes
    
    I don't recall modifying the power settings, however it appears to be set for "powersave".

    Code:
    > cpupower frequency-info 
    analyzing CPU 0: 
      driver: intel_pstate 
      CPUs which run at the same hardware frequency: 0 
      CPUs which need to have their frequency coordinated by software: 0 
      maximum transition latency:  Cannot determine or is not supported. 
      hardware limits: 800 MHz - 4.70 GHz 
      available cpufreq governors: performance powersave 
      current policy: frequency should be within 800 MHz and 4.70 GHz. 
                      The governor "powersave" may decide which speed to use 
                      within this range. 
      current CPU frequency: Unable to call hardware 
      current CPU frequency: 800 MHz (asserted by call to kernel) 
      boost state support: 
        Supported: yes 
        Active: yes
    
    > cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor | sort | uniq 
    powersave
    
    If I did want to install the Nvidia drivers the "hard" way, how do I uninstall the ones I currently have installed (which were the recommended way) and then install them manually? Thank you for your help with this.

  4. #4
    Join Date
    Jun 2008
    Location
    East of Podunk
    Posts
    33,250
    Blog Entries
    15

    Default Re: How to roll back Nvidia driver version until issues pass

    Quote Originally Posted by iqgrande View Post
    I only have the Nvidia graphics card to use.

    Code:
    > inxi -Gxxz 
    Graphics:  Device-1: NVIDIA TU106 [GeForce RTX 2070 Rev. A] vendor: eVga.com. driver: nvidia v: 470.74  
               bus-ID: 01:00.0 chip-ID: 10de:1f07  
               Device-2: AVerMedia Live Streamer CAM 313 type: USB driver: snd-usb-audio,uvcvideo bus-ID: 1-7:4  
               chip-ID: 07ca:313a  
               Display: x11 server: X.org 1.20.13 compositor: kwin_x11 driver:loaded: nvidia  
               unloaded: fbdev,modesetting,nouveau,vesa alternate: nv resolution: <missing: xdpyinfo>  
               OpenGL:renderer: NVIDIA GeForce RTX 2070/PCIe/SSE2 v: 4.6.0 NVIDIA 470.74 direct render: Yes
    
    I don't recall modifying the power settings, however it appears to be set for "powersave".

    Code:
    > cpupower frequency-info 
    analyzing CPU 0: 
      driver: intel_pstate 
      CPUs which run at the same hardware frequency: 0 
      CPUs which need to have their frequency coordinated by software: 0 
      maximum transition latency:  Cannot determine or is not supported. 
      hardware limits: 800 MHz - 4.70 GHz 
      available cpufreq governors: performance powersave 
      current policy: frequency should be within 800 MHz and 4.70 GHz. 
                      The governor "powersave" may decide which speed to use 
                      within this range. 
      current CPU frequency: Unable to call hardware 
      current CPU frequency: 800 MHz (asserted by call to kernel) 
      boost state support: 
        Supported: yes 
        Active: yes
    
    > cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor | sort | uniq 
    powersave
    
    If I did want to install the Nvidia drivers the "hard" way, how do I uninstall the ones I currently have installed (which were the recommended way) and then install them manually? Thank you for your help with this.
    Hi
    For uninstall, look down the page here: https://en.opensuse.org/SDB:NVIDIA_drivers

    For the 'Hard Way' see https://en.opensuse.org/SDB:NVIDIA_the_hard_way

    For your power issue, disable the intel pstate driver and run intel_cpufreq and see if that helps. Fire up YaST Bootloader and in the kernel command line options add intel_pstate=disable

    Also check the system BIOS for any power settings before adding above.
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  5. #5

    Default Re: How to roll back Nvidia driver version until issues pass

    There's a lot of good information in the previous post. Thanks! What's more accurate: the CPU frequency from `lscpu` or the frequency from `cpupower frequency-info`? The former is telling me something higher than the latter.

  6. #6
    Join Date
    Dec 2008
    Location
    FL, USA
    Posts
    3,992
    Blog Entries
    1

    Default Re: How to roll back Nvidia driver version until issues pass

    Instead of having to roll back, don't upgrade the kernel until you've sure the available driver matches the current kernel. Run zypper al kernel-de*. That will lock the kernel. When locked, the kernel won't be upgraded automatically. When you wish to upgrade the kernel, run zypper in kernel-default. Zypper will offer to "remove" the lock and install the kernel, but because of the wildcard in the lock, zypper will merely ignore the lock, and proceed to install the new kernel.
    Reg. Linux User 211409 *** multibooting since 1992
    Primary: 15.3, TW, 15.1 & 13.1 on Haswell @earthlink.net
    Secondary: eComStation (OS/2) &15.2 on i965P/Radeon
    Tertiary: Debian, Fedora, Mageia, more on Rocket Lake & older Intel, AMD, NVidia....

  7. #7
    Join Date
    Jun 2008
    Location
    East of Podunk
    Posts
    33,250
    Blog Entries
    15

    Default Re: How to roll back Nvidia driver version until issues pass

    Quote Originally Posted by mrmazda View Post
    Instead of having to roll back, don't upgrade the kernel until you've sure the available driver matches the current kernel. Run zypper al kernel-de*. That will lock the kernel. When locked, the kernel won't be upgraded automatically. When you wish to upgrade the kernel, run zypper in kernel-default. Zypper will offer to "remove" the lock and install the kernel, but because of the wildcard in the lock, zypper will merely ignore the lock, and proceed to install the new kernel.
    Hi
    Yes, that is possible, however, review of what is changing is a far better option (via the announcement), then make a choice to update or not... for me it takes all but a few moments to rebuild the modules and reboot....
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  8. #8

    Default Re: How to roll back Nvidia driver version until issues pass

    There's a lot of good information here. I really appreciate everyone's posts. My probably last question is meant to add atop some of the previous responses. In general, when TW pushes something that causes havoc on my system I'd like to have 2 states that I can readily boot into: (1) the safe state where everything works and (2) the current state so I can test if things have been fixed and to also provide ticket feedback (I am trying to do my part and raise awareness of issues I find early and often). Sometimes an issue may take several months to rectify. Since openSUSE has technologies like btrfs that it builds atop, I can imagine several scenarios that someone could pick to deal with (1) and (2). In my mind, creating a manual snapshot for (1) that can be accessed easily would likely be the easiest way of dealing with the 2 states. Is there a better way or more details as to the best way to manage the snapshots for (1) and (2)? The disadvantage of the snapshot approach is that throughout that time the system is not getting updates unless the offending packages are just frozen (per earlier suggestion). Thank you for any insight you have into this; this is really helping me understand the preferred ways of managing a rolling TW system.

  9. #9
    Join Date
    Jun 2008
    Location
    East of Podunk
    Posts
    33,250
    Blog Entries
    15

    Default Re: How to roll back Nvidia driver version until issues pass

    Hi
    I just roll with every release... the glibc rebuild caused a few issues, but no showstoppers (I use slack, so just reverted to web ui) for me. VScode still doesn't work on nvidia, but my default is AMD which it works with fine.
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  10. #10
    Join Date
    Mar 2020
    Location
    São Leopoldo, RS, Brazil
    Posts
    424

    Default Re: How to roll back Nvidia driver version until issues pass

    Quote Originally Posted by iqgrande View Post
    There's a lot of good information here. I really appreciate everyone's posts. My probably last question is meant to add atop some of the previous responses. In general, when TW pushes something that causes havoc on my system I'd like to have 2 states that I can readily boot into: (1) the safe state where everything works and (2) the current state so I can test if things have been fixed and to also provide ticket feedback (I am trying to do my part and raise awareness of issues I find early and often). Sometimes an issue may take several months to rectify. Since openSUSE has technologies like btrfs that it builds atop, I can imagine several scenarios that someone could pick to deal with (1) and (2). In my mind, creating a manual snapshot for (1) that can be accessed easily would likely be the easiest way of dealing with the 2 states. Is there a better way or more details as to the best way to manage the snapshots for (1) and (2)? The disadvantage of the snapshot approach is that throughout that time the system is not getting updates unless the offending packages are just frozen (per earlier suggestion). Thank you for any insight you have into this; this is really helping me understand the preferred ways of managing a rolling TW system.
    In a default installation zypper creates a pair of snapshots on each update, snapper prunes old snapshots on a schedule, and grub integrates booting from older snapshots in its menu. In case you find a broken update with a package, you can choose to boot from a previous snapshot, rollback to it, reboot, lock the package from updating, update the system. Repeat.
    openSUSE Tumbleweed

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •