> Just tried the Kernel Debugging flag. The machine
> produced the same reaction. On a DomU shutdown, the
> Dom0 was restarted.
How do you shutdown the domUs?
>From within the domU, using the linux / unix shutdown command?
Or from dom0, using "xm shutdown ..." ?
> there is now a new unix.6 and vmcore.6 in the
> /var/crash/atlantis directory.
> panic message: BAD TRAP: type=e (#pf Page fault) rp=ffffff0003aed6d0
> addr=ffffff016a3551b8
> ffffff0003aed880 xnb_copy_to_peer+0x32(ffffff016a34f000, ffffff013a9368a0)
Hmm, looks like the "xnb_t" structure that is passed to
xnb_copy_to_peer is already freed; that would probably
explain the "bad mutex" panic when running without
kernel heap debugging. And with heap debugging, we're
crashing a few bytes earlier...
What does mdb know about the xnb_t pointer to
address ffffff016a34f000 (the 1st argument to function
xnb_copy_to_peer) in crash dump #6? Try this:
echo 'ffffff016a34f000::whatis' | mdb -k 6
This message posted from opensolaris.org
_______________________________________________
xen-discuss mailing list
[email protected]
