This bug is not a duplicate of the one listed (so I un-marked it as such). Except for the use of the term "Xen" in both bugs, they are completely unrelated.
Here's the gist of the bug:
1. There is a linux-image-xen package that is specifically designed for use with Xen.
2. The linux-image-server should not have Xen support in it.
3. Many people are using the linux-image-server kernel because it has PAE support and therefore can address > 4GB memory
4. Some people prefer to use the nvidia package from nvidia.com instead of the linux-restricted-modules + nvidia-glx packages (the nvidia.com package is faster for me).
5. The inclusion of Xen support in the linux-image-server kernel prevents the nvidia installer from building the module.
So, the solution is to remove the unnecessary Xen support from linux-image-server.
This bug is not a duplicate of the one listed (so I un-marked it as such). Except for the use of the term "Xen" in both bugs, they are completely unrelated.
Here's the gist of the bug: d-modules + nvidia-glx packages (the nvidia.com package is faster for me).
1. There is a linux-image-xen package that is specifically designed for use with Xen.
2. The linux-image-server should not have Xen support in it.
3. Many people are using the linux-image-server kernel because it has PAE support and therefore can address > 4GB memory
4. Some people prefer to use the nvidia package from nvidia.com instead of the linux-restricte
5. The inclusion of Xen support in the linux-image-server kernel prevents the nvidia installer from building the module.
So, the solution is to remove the unnecessary Xen support from linux-image-server.