> Any ideas on how to get the installer to recover from this condition? > Well, this kind of ends my involvement. It's now up to the kernel people to say if this is a kernel problem or, if not, reassign the bug elsewhere. Cheers, FJP