>Number:         3072
>Category:       kern
>Synopsis:       nfs client panics on halt if server reboots
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people (Kernel Bug People)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Dec 30 15:50:00 1996
>Last-Modified:
>Originator:     Luke Mewburn
>Organization:
NetBSD Foundation
>Release:        NetBSD-961229
>Environment:
	NetBSD 1.2B i386

>Description:
	I have an NFS server & client (both running 961229-current
	kernels).
	If I halt the server, and then do something like
		ls /nfsdir
	in one VT (which, as expected, ``hangs''), and then do a halt in
	another, the machine will eventually panic. The messages
	are something like (copied from the back of an envelope)

		vm_fault(0xf888ee00,0,1,0) -> 1
		kernel: page fault trap, code = 0
		db> trace
		_nfs_reply(...) at _nfs_reply+0xa6
		_nfs_request(...) at _nfs_request+0x3aa
		_nfs_access(...) at _nfs_access+0x36c
		_vn_open(...) at _vn_open+0x1d0
		_sys_open(...) at _sys_open+0x97
		_syscall() at _syscall+0x154
		--- syscall (number 5) ---
		0xcdf:

>How-To-Repeat:
	something like the following should do it:
	    mount server:/foo /foo
	    rsh -n server shutdown -h now &
	    sleep 30
	    ls /foo &
	    shutdown -h now

>Fix:
	dunno.  seems similar to [kern/2797] that gwr put in.
>Audit-Trail:
>Unformatted: