So, due to the whole "GPL condom" mess, CUDA is now broken on 5.9+. The patch that apparently fixes it is indeed in patches/kernel-5.9.patch, but during the PKGBUILD process, I get the "Skipping kernel-5.9.patch as it doesn't apply to this driver version..." message, the patches aren't applied, and therefore CUDA remains broken.
Is there an easy way to set up a config option in customization.cfg that would allow for enabling/disabling this, since it seems like this is going to be an issue for the foreseeable future? If there are other parts of the patch that aren't necessary, is it possible to maybe take the necessary nvidia_uvm patch and put it in a standalone patch?
So, due to the whole "GPL condom" mess, CUDA is now broken on 5.9+. The patch that apparently fixes it is indeed in patches/kernel-5.9.patch, but during the PKGBUILD process, I get the "Skipping kernel-5.9.patch as it doesn't apply to this driver version..." message, the patches aren't applied, and therefore CUDA remains broken.
Is there an easy way to set up a config option in
customization.cfg
that would allow for enabling/disabling this, since it seems like this is going to be an issue for the foreseeable future? If there are other parts of the patch that aren't necessary, is it possible to maybe take the necessary nvidia_uvm patch and put it in a standalone patch?