You’ll need them if you’re using native Steam or other 32 bit graphical apps. It won’t affect flatpak apps.
1 Like
Exactly! That what I did, but remember there is no clipboard or auto-complete in TTY for URLs and I forced to type everything manually and specify installed versions manually.
Thanks @Lioli7k!
FWIW, I was able to get the longterm kernel to work. The later kernels all fail with the same DMA error in Xorg.0.log which I posted earlier.
grep EE /var/log/Xorg.0.log
[ 12.957] Current Operating System: Linux tharbad 7.0.2-1-default #1 SMP PREEMPT_DYNAMIC Tue Apr 28 04:27:33 UTC 2026 (46da294) x86_64
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 23.588] (EE) NVIDIA(GPU-0): Failed to initialize DMA.
[ 23.596] (EE) NVIDIA(0): Failed to allocate push buffer
[ 23.605] (EE)
[ 23.605] (EE) AddScreen/ScreenInit failed for driver 0
[ 23.605] (EE)
[ 23.605] (EE)
[ 23.605] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 23.605] (EE)
[ 23.607] (EE) Server terminated with error (1). Closing log file.
Wed May 6 15:18:00 2026
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 595.71.05 Driver Version: 595.71.05 CUDA Version: 13.2 |
+-----------------------------------------+------------------------+----------------------+
| 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 RTX 3050 Off | 00000000:01:00.0 On | N/A |
| 33% 32C P8 8W / 70W | 11MiB / 6144MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+
zypper se -si nvidia kernel-def kernel-long
Loading repository data...
Reading installed packages...
S | Name | Type | Version | Arch | Repository
---+--------------------------------------------+---------+--------------------------+--------+------------------
i+ | kernel-default | package | 7.0.1-1.0.8.1.sr20260402 | x86_64 | (System Packages)
i+ | kernel-default | package | 6.19.12-1.1 | x86_64 | (System Packages)
i+ | kernel-default | package | 7.0.2-1.1 | x86_64 | repo-oss
i+ | kernel-default-devel | package | 7.0.1-1.0.8.1.sr20260402 | x86_64 | (System Packages)
i+ | kernel-default-devel | package | 6.4.0-150600.21.3 | x86_64 | (System Packages)
i+ | kernel-default-devel | package | 7.0.2-1.1 | x86_64 | repo-oss
i | kernel-firmware-nvidia | package | 20260408-1.1 | noarch | repo-oss
i+ | kernel-longterm | package | 6.18.26-1.1 | x86_64 | repo-oss
i | kernel-longterm-devel | package | 6.18.26-1.1 | x86_64 | repo-oss
i | libnvidia-cfg | package | 595.71.05-6.1 | x86_64 | repo-non-free
i | libnvidia-egl-gbm1 | package | 1.1.3-1.3 | x86_64 | repo-oss
i | libnvidia-egl-gbm1-32bit | package | 1.1.3-1.3 | x86_64 | repo-oss
i | libnvidia-egl-wayland1 | package | 1.1.22-1.4 | x86_64 | repo-oss
i | libnvidia-egl-wayland1-32bit | package | 1.1.22-1.4 | x86_64 | repo-oss
i | libnvidia-egl-x111 | package | 1.0.5-2.3 | x86_64 | repo-oss
i | libnvidia-egl-x111-32bit | package | 1.0.5-2.3 | x86_64 | repo-oss
i | libnvidia-gpucomp | package | 595.71.05-6.1 | x86_64 | repo-non-free
i | libnvidia-gpucomp-32bit | package | 595.71.05-6.1 | x86_64 | repo-non-free
i | libnvidia-ml | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | nvidia-common-G07 | package | 595.71.05-6.1 | x86_64 | repo-non-free
i | nvidia-compute-G07 | package | 595.71.05-6.1 | x86_64 | repo-non-free
i | nvidia-compute-utils-G07 | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | nvidia-gl-G07 | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | nvidia-gl-G07-32bit | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | nvidia-libXNVCtrl | package | 595.71.05-2.1 | x86_64 | repo-non-free
i+ | nvidia-modprobe | package | 595.71.05-2.2 | x86_64 | repo-non-free
i+ | nvidia-open-driver-G07-signed-kmp-default | package | 595.71.05_k7.0.2_1-2.1 | x86_64 | repo-oss
i+ | nvidia-open-driver-G07-signed-kmp-longterm | package | 595.71.05_k6.18.26_1-2.1 | x86_64 | repo-oss
i | nvidia-open-driver-G07-signed-kmp-meta | package | 595.58.03-15.1 | x86_64 | repo-non-free
i | nvidia-persistenced | package | 595.71.05-2.2 | x86_64 | repo-non-free
i+ | nvidia-userspace-meta-G07 | package | 595.71.05-15.1 | x86_64 | repo-non-free
i | nvidia-video-G07 | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | nvidia-video-G07-32bit | package | 595.71.05-6.1 | x86_64 | repo-non-free
i+ | openSUSE-repos-Tumbleweed-NVIDIA | package | 20260423.1a6a0f3-2.1 | x86_64 | repo-oss
After fixing the login issue caused by a bogus sddm.conf I can now get back to xfce and thus pay bills and stuff … yay?
Thanks @Lioli7k for the longterm hint. Now if I can just get my legacy bios grub config to boot that as the default.
Is it really helpful to have this added the bug? I can do that too I guess.
I need someone to reply to this message when it’s fixed for the GTX 1650 SUPER.
RTX 5050 - black screen
Glad you got your PC working. I guess there’s no need for logs because Hui attached log with the same error already.
As for grub I guess it’s better if you’ll start a new thread on forum. I’ve switched to systemd-boot long ago. I can suggest to check if in your /etc/default/grub you have GRUB_DEFAULT set to saved. You need to run sudo grub2-mkconfig -o /boot/grub2/grub.cfg after editing /etc/default/grub. After you do that grub should remember your last boot option and use that.
Hiya. After updating but before rebooting did you checked that KMP matches kernel version and that userspace packages match that KMP?
I don’t think any further updates to that table make sense. There are reports of people having the same GPU where it works for one but not the other. The bug proven to be completely unpredictable and behave differently on different desktops and laptops.
Thanks. I got yast to do it for me after I fixed the bogus old disk id. There is some cruft 
Yes, Hui’s logs are virtually the same as what I would have put there. One thing I will point out is that my CPU/MB are pretty old. I wonder if there is some PCIe difference or something. Just based on the thing trying to map memory outside the PCI bus range (in the dmesg output).
1 Like
New snapshot “20260506”, new kernel… still ok here with the open driver 595.71.05 and 7.0.3-1-default.
I do see an updated nvidia-settings (580.142 → 580.159.03)…
No dice here. KMPs weren’t rebuilt. So 7.0.3 is broken, 6.18.26 is working well with 595.71.05.
@Lioli7k no rebuild required for the open driver, it drops into weak-updates.
zypper se -is nvidia-open
S | Name | Type | Version | Arch | Repository
---+------------------------------------------------+---------+------------------------+--------+-----------
i+ | nvidia-open-driver-G07-signed-cuda-kmp-default | package | 595.71.05_k7.0.2_1-2.1 | x86_64 | repo-oss
modinfo nvidia | grep filename:
filename: /usr/lib/modules/7.0.3-1-default/weak-updates/updates/nvidia.ko.zst
@malcolmlewis you want to convince us it does work?! It does not! Do not want to experiment with my machine and then proceed with hard and annoying recovering.
I know that there’s no need for rebuild if kABI hasn’t changed. I’m trying to isolate if it’s a kernel or KMP bug. So far I’m leaning towards KMP. If it hasn’t been rebuilt means that situation is unlikely to improve.
Either way, I have an idea on how to narrow it down further and I’m working on it.
My 2cents. This is the reason I am still using the .run driver.
The .run driver doesn’t have issues even in the latest kernel:7.0.3-1
I am using the latest NVIDIA-Linux-x86_64-595.71.05.run
1 Like
Malcolmlewis is unaffected by this issue. I don’t know how but that’s the fact. Look other reports of people that have it broken like me, Hui or couple of others.
I just tried updating my TW install to the 7.0.3 kernel, and I sort of have things working.
What I found was that there was an issue with updating binutils, and that broke building the nvidia driver. To fix the binutils update issue, I ran:
sudo update-alternatives --install /usr/bin/ld ld /usr/bin/ld.bfd 20
This allowed the binutils update to install. I then did a forced update of the nvidia-G06 driver (I use the proprietary driver), and that build succeeded.
I say things are “sort of working”, because Chrome opens, but it doesn’t show on my display. But I have 3 monitors connected, and they’re all working at the expected resolutions, and nvidia-smi says the driver is working.
So now it’s just a matter of figuring out what settings I need to adjust to get Chrome windows to show up again.
Card here is a 3090ti.
Interesting… I’ve tried that exact run file with both kernel-default and kernel-vanilla but had no luck. Maybe problem is between chair and keyboard… Thanks for letting me know.
Just noticing that Malcolm is using the CUDA variant, unlike most other affected users?
nvidia-driver-G06-kmp-default-580.159.03_k7.02_1-49 is compatible with 7.0.3 and should be rebuilt automatically by dkms on update. You may need to manually inspect your built KMP and rebuild it.
1 Like