Your locking and rman changes to pci/if_*

[Nate Lawson]

On Wed, 16 Apr 2003, Andrew Gallatin wrote:
> Ruslan Ermilov writes:
>  > : dc0: detached
>  > : panic: blockable sleep lock (sleep mutex) Giant @ vm/vm_fault.c:206
>  > 
>  > Reverting to if_dc.c,v 1.96 makes the problem go away.
>  > Please fix.
> 
> I think this indciates that something trap'ed in the kernel in the
> middle of handling, say, another trap.
> 
> A stack trace from ddb or gdb would make this a lot easier to debug.

Yes, backtrace please.  Also, you might want to try removing the
bus_child_present() check in xl_detach and just unconditionally call
xl_stop().  Ian Dowse has indicated to me that some of the drivers
(dc) have routines which need to always be called in detach, even if the
hardware is not present.

-Nate