Hi,
I have created several versions of the nfsroot including new nvidia
drivers and CUDA libraries, because we needed the newest drivers for
new hardware.
Often I used a mixture of packages from testing and experimental. I've
also created a nfsroot using the drivers and CUDA libs from nvidia
itse
So how could I prevent that?
I used the chroot /target in the hopes of using the kernel of the target
machine.
Regards, Stephan
-Original Message-
From: Andrew
To: fully
Date: Friday, 25 October 2024 8:27 AM CEST
Subject: Re: Installing CUDA with FAI
On Thu, 2024-10-24 at 14:50 +02
Hi Thoralf,
sure this is possible. A FAI server only provides TFTP, DHCP and NFS
services (and maybe http services) which do not depend on a certain
architecture.
Creating the nfsroot for AMD64 on a Raspberry Pi (cross-arch) may be
possibe but will take very lng. I would suggest to create the
You may not prevent that, because we do not know how the run script
tries to detect a kernel version. In the chroot there's the target
kernel installed, but it's still running the FAI installation kernel.
Try to use the .deb packages instead of the run script.
> On Fri, 25 Oct 2024 09:34:30 +
