Q. about oops backtrace

[Pete Zaitcev <zaitcev@redhat.com>]

Hello:

I was investigating an oops and the trace looked like this:

>>EIP; c01c54a9 <lvm_do_remove_proc_entry_of_vg+9/c0>   <=====
Trace; c01c3654 <lvm_do_vg_rename+84/250>
Trace; c01c0f0f <lvm_chr_ioctl+30f/6d0>
Trace; c015e7e2 <ext2_getblk+72/e0>
Trace; c01155a6 <do_page_fault+166/440>
Trace; c01272a9 <do_no_page+49/a0>
Trace; c0127414 <handle_mm_fault+114/1a0>
Trace; c0136a2d <kunmap_high+7d/90>
Trace; c012722e <do_anonymous_page+de/110>
Trace; c0127290 <do_no_page+30/a0>
Trace; c0127414 <handle_mm_fault+114/1a0>
Trace; c014cdec <dput+1c/170>
Trace; c0143f80 <cached_lookup+10/50>
Trace; c0144aae <path_walk+85e/940>
Trace; c014cdec <dput+1c/170>
Trace; c01392c9 <chrdev_open+59/a0>
Trace; c0138130 <dentry_open+c0/150>
Trace; c013805d <filp_open+4d/60>
Trace; c0148b97 <sys_ioctl+247/2a0>
Trace; c01091c7 <system_call+33/38>

What is with those recursive handle_mm_fault calls?
That does not look quite right. I _assume_ that the
stack would collapse properly upon return, but still...
I would appreciate a suggestion about what .S file to read
for the explanation.

-- Pete