[PVFS-developers] pvfs kernel module bug (1.6.2)

[Richard Jones]

This report follows up on a previous posting where I offered patches to kpvfsd.c 
and kpvfsdev.c for pvfs-kernel-1.6.2-linux-2.4 so that it would compile under 
gcc3.2 and redhat 9.  This time, I do not have a bug fix but I do have the log 
that should indicate where the problem lies.  Actually I have tested it back as 
far as pvfs-1.5.8 and the problem is still there.  If this is a known issue, 
please excuse.  I did a search on the archives and came up with nothing related.

Description: modprobe -r pvfs causes kernel seg fault
Build info: configure --with-pvfs=../pvfs-1.6.2 			 
--with-libpvfs-dir=../pvfs-1.6.2/lib --with-log-dir=/var/log 
--with-kernel-headers=/usr/src/linux-2.4/include --with-kpvfsd --with-newstyle
System info: redhat 9 with updated kernel 2.4.20-30.9smp (dual intel or athlon)
How to reproduce:
  1) modprobe pvfs
  2) modprobe -r pvfs
     Segmentation fault
Result: the system does not crash, but the module remains in the kernel list 
      with attribute "(deleted)".  It cannot be removed or reinstalled without a 
reboot after this happens.

[from /var/log/messages under pvfs-kernel-1.5.8-linux-2.4]
> Mar 31 11:18:48 grinch kernel: ------------[ cut here ]------------
> Mar 31 11:18:48 grinch kernel: kernel BUG at exit.c:727!
> Mar 31 11:18:48 grinch kernel: invalid operand: 0000
> Mar 31 11:18:48 grinch kernel: w83781d i2c-proc i2c-amd756 i2c-core nfsd pvfs nfs lockd sunrpc parport_pc lp parport autofs e1000 ipt_REJECT iptable_filter ip_tables keybdev mousedev input
> Mar 31 11:18:48 grinch kernel: CPU:    0
> Mar 31 11:18:48 grinch kernel: EIP:    0060:[<c012654e>]    Not tainted
> Mar 31 11:18:48 grinch kernel: EFLAGS: 00010246
> Mar 31 11:18:48 grinch kernel:
> Mar 31 11:18:48 grinch kernel: EIP is at check_tasklist_locked [kernel] 0xe (2.4.20-30.9smp)
> Mar 31 11:18:48 grinch kernel: eax: 01000000   ebx: c037a000   ecx: e7cb8354   edx: 00000000
> Mar 31 11:18:48 grinch kernel: esi: e845c000   edi: 00000000   ebp: 00000000   esp: e39fdf5c
> Mar 31 11:18:48 grinch kernel: ds: 0068   es: 0068   ss: 0068
> Mar 31 11:18:48 grinch kernel: Process modprobe (pid: 12643, stackpage=e39fd000)Mar 31 11:18:48 grinch kernel: Stack: f0a594fd e7cb8340 e845c000 00000000 f0a5982f f0a63b58 f0a53000 fffffff0
> Mar 31 11:18:48 grinch kernel:        00000000 bfffcf88 f0a5329e 00000001 f0a69970 c0124a53 f0a53000 f0a53000
> Mar 31 11:18:48 grinch kernel:        fffffff0 00000000 c0123cf6 f0a53000 00000000 00001000 e39fc000 00000000
> Mar 31 11:18:48 grinch kernel: Call Trace:   [<f0a594fd>] name2pid [pvfs] 0xd (0xe39fdf5c))
> Mar 31 11:18:48 grinch kernel: [<f0a5982f>] cleanup_kpvfsd [pvfs] 0xd7 (0xe39fdf6c))
> Mar 31 11:18:48 grinch kernel: [<f0a63b58>] .LC1 [pvfs] 0x7b9 (0xe39fdf70))
> Mar 31 11:18:48 grinch kernel: [<f0a5329e>] cleanup_pvfs [pvfs] 0x1e (0xe39fdf84))
> Mar 31 11:18:48 grinch kernel: [<f0a69970>] pvfs_fs_type [pvfs] 0x0 (0xe39fdf8c))
> Mar 31 11:18:48 grinch kernel: [<c0124a53>] free_module [kernel] 0xc3 (0xe39fdf90))
> Mar 31 11:18:48 grinch kernel: [<c0123cf6>] sys_delete_module [kernel] 0xe6 (0xe39fdfa4))
> Mar 31 11:18:48 grinch kernel: [<c010972f>] system_call [kernel] 0x33 (0xe39fdfc0))
> Mar 31 11:18:48 grinch kernel:
> Mar 31 11:18:48 grinch kernel:
> Mar 31 11:18:48 grinch kernel: Code: 0f 0b d7 02 65 c6 27 c0 eb f4 90 8d b4 26 00 00 00 00 83 ec


[from /var/log/messages under pvfs-kernel-1.6.2-linux-2.4]
> Mar 31 15:07:38 nod30 kernel: ------------[ cut here ]------------
> Mar 31 15:07:38 nod30 kernel: kernel BUG at exit.c:727!
> Mar 31 15:07:38 nod30 kernel: invalid operand: 0000
> Mar 31 15:07:38 nod30 kernel: pvfs w83781d i2c-proc i2c-isa i2c-core parport_pc
> lp parport nfsd autofs ipt_REJECT iptable_filter ip_tables ext3 jbd raid5 xor raid0 keybdev mousedev input h
> Mar 31 15:07:38 nod30 kernel: CPU:    1
> Mar 31 15:07:38 nod30 kernel: EIP:    0060:[<c0126e10>]    Not tainted
> Mar 31 15:07:38 nod30 kernel: EFLAGS: 00010246
> Mar 31 15:07:38 nod30 kernel:
> Mar 31 15:07:38 nod30 kernel: EIP is at check_tasklist_locked [kernel] 0x10 (2.4.20-30.9smpa)
> Mar 31 15:07:38 nod30 kernel: eax: 01000000   ebx: c03c0000   ecx: c45b0514   edx: 00000000
> Mar 31 15:07:38 nod30 kernel: esi: f6654000   edi: 00000000   ebp: 00000000   esp: f6671f5c
> Mar 31 15:07:38 nod30 kernel: ds: 0068   es: 0068   ss: 0068
> Mar 31 15:07:38 nod30 kernel: Process modprobe (pid: 1361, stackpage=f6671000)
> Mar 31 15:07:38 nod30 kernel: Stack: f89d7f35 c45b0500 f6654000 00000000 f89d8267 f89e36df f89d0000 fffffff0
> Mar 31 15:07:38 nod30 kernel:        00000000 bfffd578 f89d029e 00000001 f89e9ff0 f89e9ff0 c0125302 f89d0000 f89d0000
> Mar 31 15:07:38 nod30 kernel:        fffffff0 f6665000 c0124587 f89d0000 00000000 00001000 f6670000 00000000
> Mar 31 15:07:38 nod30 kernel: Call Trace:   [<f89d7f35>] name2pid [pvfs] 0xd (0xf6671f5c))
> Mar 31 15:07:38 nod30 kernel: [<f89d8267>] cleanup_kpvfsd [pvfs] 0xd7 (0xf6671f6c))
> Mar 31 15:07:38 nod30 kernel: [<f89e36df>] .LC1 [pvfs] 0x824 (0xf6671f70))
> Mar 31 15:07:38 nod30 kernel: [<f89d029e>] cleanup_pvfs [pvfs] 0x1e (0xf6671f84))
> Mar 31 15:07:38 nod30 kernel: [<f89e9ff0>] pvfs_fs_type [pvfs] 0x0 (0xf6671f8c))Mar 31 15:07:38 nod30 kernel: [<c0125302>] free_module [kernel] 0xd2 (0xf6671f90))
> Mar 31 15:07:38 nod30 kernel: [<c0124587>] sys_delete_module [kernel] 0xe7 (0xf6671fa4))
> Mar 31 15:07:38 nod30 kernel: [<c01098cf>] system_call [kernel] 0x33 (0xf6671fc0))
> Mar 31 15:07:38 nod30 kernel:
> Mar 31 15:07:38 nod30 kernel: Code: 0f 0b d7 02 85 ad 2a c0 eb f2 8d b6 00 00 00 00 83 ec 0c 89