Well, it is not a common error as you have seen.
I don't know if anyone else is using the selinux patch with UML, I would
expect it to mostly work as UML is often used for testing to find
Where exactly EBADF(9) bad file number is occurring is not clear from
your error message, a search of the my kernel pool (2.4.20-5um) yields
arch/um/kernel/irq_user.c:392 where the error message indicates that it
tried and failed to deliver a irq for fd 0 (stdin?) to what I guess are
Does the error still appear without the selinux patches in the UML
Does stock 2.4.20 host kernel mean a debian stock or a kernel.org 2.4.20,
I am assuming that both the host kernel and UML kernel are from the
kernel.org tree (UML with the 2.4.20-4um patch)
Since you said stock I am assuming that you have not applied the SKAS
If you are not using SKAS you can add debug to your UML command line
which will open a gdb window and you can set a breakpoint at
irq_user.c:392 (the printk statement) hopefully a "(gdb) bt" back trace
will yield more information about what is happening, and "(gdb) print
*irq" may also be useful
If you are using SKAS you can start UML under gdb and debug that way
The best thing to hit the internet in years - Juno SpeedBand!
Surf the web up to FIVE TIMES FASTER!
Only $14.95/ month - visit http://www.juno.com to sign up today!