> Hmm, a real mystery. Does X start up correctly when it gets to runlevel 5 ? > Might be interesting to boot in single user mode, and init 3, then check > out the X logs to see if there's any explanation in there. This stuff is sounding like an uninitialized variable in the kernel option parsing code. On x86_64 xen doesn't work, non-xen does (for me), on i686 someone else said the non-xen kernel doesn't work. Now we have the rhgb option failing to work. I hate when that happens :-).