I don't understand this oops...

Gerd Knorr (kraxel@cs.tu-berlin.de)
Sat, 11 Jan 1997 15:45:16 +0100 (MET)

While hacking the 2.1.20 kernel I got a oops while mounting a scsi
cdrom (maybe my fault). ksymoops gives the following calltrace. But I do
not understand it. The lines marked with "***" would give a somewhat
logical trace. But what the heck does the NFS-function in the calltrace?
The box runs with NFS-Root, so running the mount programm will cauce
NFS-traffic, but why it shows up in the calltrace??? Could anyone explain
this?

Gerd

PS: This is reproduceable, and I'm sure I used the right System-map file.

>>EIP: c0127764 <find_candidate+e4/110> ***
Trace: c0127829 <refill_freelist+99/410> ***
Trace: c017250c <nfs_rpc_doio+cc/1a0>
Trace: c0127bcf <getblk+2f/410>
Trace: c0127d26 <getblk+186/410> ***
Trace: c0128478 <bread+18/90> ***
Trace: c016cbe2 <isofs_search_session+1c2/2a0> ***
Trace: c01a5068 <aha1542_queuecommand+698/6c0>
Trace: c0199ccc <scsi_do_cmd+38c/3c0>
Trace: c016cda1 <isofs_read_super+e1/5e0> ***
Trace: c01a103e <sr_media_change+3e/c0>
Trace: c01a6205 <media_changed+35/70>
Trace: c012a469 <read_super+c9/100> ***
Trace: c012a8e3 <do_mount+e3/150> ***
Trace: c0ed000f
Trace: c012aee5 <sys_mount+325/370> ***
Trace: c0ed000f
Trace: c010abd8 <system_call+38/40>
Trace: c0ed000f