Will there be support for nvidia G03 drivers?
When I visited https://download.nvidia.com/opensuse/leap/15.2/x86_64/ there is no longer support for older graphics.
Maybe the 340 drivers are not building anymore with kernel 5.3.x in Leap 15.2?
Download G03 packages from https://download.nvidia.com/opensuse/leap/15.1/x86_64/ ant try to install them.
Maybe will work.
Same with G02 packages from https://download.nvidia.com/opensuse/leap/42.3/x86_64/ .
Hint: you will need instruments to compile drivers.
Hi
Leap has the 5.3.x kernel, G03 support was dropped awhile ago now, I gave up with a GT8800 awhile back, if it’s not a laptop, consider an upgrade, I use cheap cards here GT710’s 
There are probably patches on the nvidia forums, but that will mean patching the run file (it’s not bad if have the patches) and installing the hardway.
Thank you.
I think that lack of support is due 5.3.x kernel on Leap 15.2. I have got old graphics and use nvidia G03 drivers on Leap 15.1.
Maybe it is time to change Leap for Tumbleweed or just use future Leap 15.2 on default graphic drivers.
Nvidia drops support for a G03 hardware.
There is no G03 drivers for a TW: https://download.nvidia.com/opensuse/tumbleweed/x86_64/ .
You may use nouveau drivers.
Or get new hardware. Nvidia’s Kepler family will have support for ~2 years.
Hint: GT 1030 has no CUDA support.
Consider to use AMD graphics - it is good for an ordinary use (no CUDA).
Did you manage to make it work properly?
I have just upgraded to 15.2 and G03 stopped working (I have ASUSTeK EN210 SILENT which is NVIDIA Corporation GT218 [GeForce 210])
I have installed nouveau, but It does not work very well, so I’m thinking on fixing the issue or dawngrading to 15.1 again (if possible).
best regards
When you use Kmail its no good idea of downgrading. there was a change in the database.
Hi
It’s working, help a user on IRC the other day get his 8400M up and running on Leap 15.2… install the hard way with the 5.7 patched driver installer at the following link;
if installed that way, it will be necesary to redo it everytime the kernel changes, true?
Hi
If it’s just a rebuild (last two numbers) unlikely, if a version change, likely. In Leap it’s just likely to set a lower resolution at boot (well that’s what happens for me in GNOME), I just switch to a tty, login as root, drop to runlevel three, rebuild the driver and switch back to runlevel 5 and all is good.
Moved from Multimedia to Hardware
My notebook just zypper dup’ed to 15.2 and transformed to a paperweight… 
So I should migrate to fedora now?
I tried to follow the instructions there (uninstall the nvidia repo G03 rpms for 15.1 then run the nvidia340-installer and rebootet (multiple times). When trying to start an X session (startx or display manager) the notebook freezes. Something’s missing.
Any help appreciated. In the meantime I’ll try this patch thing although the installation went fine without error messages.
you can try to remove the nvidia G03 and install noveau, it should work.
I’m using it. OpenGL is not working, I miss the cube animation, but the system works.
best regards
Otherwise you have to patch the Nvidia 340 driver for kernel 5.3 from Leap 15.2.
Or your System isn’t fully upgrade to Leap 15.2?
Hi,
thanks for the hints. After some more back and forth (but basically running the patched installer letting it configure everything as default and reboot several times) it works again. Just in case someone has similar problems, here is the lspci of my builtin graphic card and the relevant commands.
# lspci | grep VGA
01:00.0 **VGA** compatible controller: NVIDIA Corporation GT218M [NVS 3100M] (rev a2)
# uname -r
5.3.18-lp152.50-default
systemctl set-default multi-user.target
reboot
zypper rm nvidia-gfxG03-kmp-default nvidia-glG03 nvidia-uvm-gfxG03-kmp-default
mkdir nv.tmp && cd nv.tmp
wget https://nvidia.if-not-true-then-false.com/NVIDIA-Linux-x86_64-340.108-patched-kernel-5.8.run
chmod +x NVIDIA-Linux-x86_64-340.108-patched-kernel-5.8.run
./NVIDIA-Linux-x86_64-340.108-patched-kernel-5.8.run
systemctl set-default graphical.target
reboot
Is there some automation I can use to let it do whatever is necessary after kernel updates, similar to what the nvidia repo did?
Maybe a warning is in order here: the code above downloads a script and patched binary kernel module code from the internet and runs this as root. Generally not a good idea, but what choice do you have now?
I wrote two scripts and run them as root after kernel update:
First named systemreboot:
#!/bin/bash
systemctl set-default multi-user.target
reboot
That enables init3 and reboots the system.
After rebooting (in text mode) I login with root and root-password and execute a script named nvidia-install
#!/bin/bash
#change to the directory where nvidia driver is stored
cd /home/stephan/bin/nvidia/
#execute the script
./NVIDIA-Linux-x86_64-455.45.01.run -a -q
#./nvidia-installer
#Execute a custom script
#./NVIDIA-Linux-x86_64-455.28-custom.run -a -q
#wait for purge-kernels.service
top
#change back to RunLevel 5
systemctl set-default graphical.target
#reboot the system
reboot
top is only because purge-kernel-service running longer as the nvidia-installer. So I wait that purge-kernel-service has ended.
both scripts are stored in /usr/local/bin.
Maybe dkms is also working, I do not know.
@Sauerland: Can you help us to get nouveau running having hw-acceleration? I can find a lot of sources describing nouveau basic features, but none giving working hw-acceleration for me.
Your Repo does have nvidia-firmware, which i can see is used by nouveau:
>modinfo nouveau
filename: /lib/modules/5.3.18-lp152.57-default/kernel/drivers/gpu/drm/nouveau/nouveau.ko
firmware: nvidia/gp100/gr/sw_method_init.bin
firmware: nvidia/gp100/gr/sw_bundle_init.bin
firmware: nvidia/gp100/gr/sw_nonctx.bin
firmware: nvidia/gp100/gr/sw_ctx.bin
firmware: nvidia/gp100/gr/gpccs_sig.bin
…]
description: nVidia Riva/TNT/GeForce/Quadro/Tesla/Tegra K1+
author: Nouveau Project
suserelease: openSUSE Leap 15.2
srcversion: 0CAFF6B8487A560805B48B9
alias: pci:v000012D2dsvsdbc03sci*
alias: pci:v000010DEdsvsdbc03sci*
depends: drm,drm_kms_helper,ttm,mxm-wmi,button,wmi,video,i2c-algo-bit
retpoline: Y
intree: Y
name: nouveau
vermagic: 5.3.18-lp152.57-default SMP mod_unload modversions
i guess like this, no need to extract firmware from orginal drivers needed as described by VideoAcceleration · freedesktop.org ?
strange is, that i cannot see the firmware for nvidia beeing loaded in dmesg anywhere…
i have # lspci | grep NVIDIA
01:00.0 VGA compatible controller: NVIDIA Corporation GT216M [GeForce GT 240M] (rev a2)
which should be VP4.0
if i try to extract the firmware from NVIDIA to /lib/firmware/nouveau as described there, also there is nothing that is beeing loaded in dmesg. Am i missing anything?
For the current Video/3d- Support actually this is software(?):
libvdpau1 will not use nouveau by default - it is fixed to nvidia, which is not there, so i set /etc/environement to include:
export VDPAU_DRIVER=nouveau
export LIBVA_DRIVER_NAME=nouveau
this is resulting in software gl and video decoding (if i am right):
> vainfo
libva info: VA-API version 1.9.0
libva info: User environment variable requested driver ‘nouveau’
libva info: Trying to open /usr/lib64/dri/nouveau_drv_video.so
libva info: Found init function __vaDriverInit_1_5
libva info: va_openDriver() returns 0
vainfo: VA-API version: 1.9 (libva 2.5.0)
vainfo: Driver version: Mesa Gallium driver 19.3.4 for NVA5
> vdpauinfo
display: :0 screen: 0
API version: 1
Information string: G3DVL VDPAU Driver Shared Library version 1.0
…
so ok, but how to switch to hardware decoding and gl now as my nvidia is VP4.0 and should do this…?
Do i have to remove packages for Mesa or others? If yes, which? How to tell nouveau to use hardware instead of software?
Regards, Daniel