[OpenAFS] /afs suddenly not mounted?

[Erik Arneson]

I've been trying to get OpenAFS running on two different RedHat 7.3
systems, and both times I've been going along with the configuration
outlined in the QuickStartUnix guide when suddenly afsd would stop doing
anything.  I wouldn't be able to get at anything under the /afs
directory, even though 'mount' would show it was mounted with type
'AFS'.

I've also been getting kernel oops'es.  Here's my 'uname -a' output.
I'm using RedHat kernel packages.

$ uname -a
Linux sif 2.4.18-3 #1 Thu Apr 18 07:32:41 EDT 2002 i686 unknown

And here's the dump from the kernel:

  <1>Unable to handle kernel paging request at virtual address f8a605e1
  printing eip:
 f8a605e1
 *pde = 01cfc067
 *pte = 00000000
 Oops: 0000
 nfsd autofs nfs lockd sunrpc eepro100 ipchains ide-scsi scsi_mod ide-cd cdrom 
 CPU:    0
 EIP:    0010:[<f8a605e1>]    Tainted: PF
 EFLAGS: 00010282
 
 EIP is at __insmod_nfsd_S.bss_L1760 [nfsd] 0x41321 (2.4.18-3)
 eax: 00000000   ebx: d9afe000   ecx: d9affb88   edx: 00000000
 esi: dde8c7c0   edi: d9affbe4   ebp: f8a88cd4   esp: d9affbb4
 ds: 0018   es: 0018   ss: 0018
 Process afsd (pid: 1545, stackpage=d9aff000)
 Stack: 00000000 00000000 00000725 00000000 00000000 d9afe000 00000000 00000000 
        00000000 d9afe000 d4bf7c0c d9bfdc0c d9affc00 00000000 00000000 3d7e8deb 
        f8a88dfc 0000476d 00000000 f8a60309 f8a88cd4 0000476d 00000000 d9afe000 
 Call Trace: [<c01275ea>] do_wp_page [kernel] 0x5a 
 [<c0127f76>] handle_mm_fault [kernel] 0x106 
 [<c0115a8a>] do_page_fault [kernel] 0x12a 
 [<c0108913>] system_call [kernel] 0x33 
 
 
 Code:  Bad EIP value.

There seemed to be no way to kill afsd or unload the afs module to try
again.  Any pointers would be much appreciated.

Thanks!

-- 
;; Erik Arneson <erik@aarg.net>    AARG Net <http://www.aarg.net/> ;;
;; GPG Key ID: 2048R/8B4CBC9C           <http://erik.arneson.org/> ;;
;; "Resistance to tyrants is obedience to God!" - Thomas Jefferson ;;