Bug 1329 - Loading cifs module segmentation fault
Loading cifs module segmentation fault
Status: RESOLVED FIXED
Product: CifsVFS
Classification: Unclassified
Component: kernel fs
2.6
All Linux
: P3 major
: ---
Assigned To: Steve French
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2004-05-07 20:41 UTC by Tim Starling
Modified: 2004-07-30 09:27 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Tim Starling 2004-05-07 20:41:07 UTC
Local:
  OS: Linux 2.6.6-rc1
  Processor: Dual Pentium4 Xeon

Remote: Windows 2003 Server SP1

Using cifs for a script in Linux that manipulates files in Windows. Module works
great for a brief period of time (~8000 7KB files). File access start to slow
down to an unuseable state. cifs reports an I/O error.

I unmounted the remote filesystem and tried again. cifs is unresponsive so I
kill the script and unloaded the module. When trying to reload the module I
encounter segmentation fault. Below is output from the log file:

May  7 22:51:46 cavasa CIFS VFS: Send error in Close = -5
May  7 22:51:56 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:00 cavasa CRON[14642]: (root) CMD (test -x /usr/sbin/run-crons &&
/usr/sbin/run-crons )
May  7 22:52:06 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:06 cavasa CIFS VFS: Send error in write = -5
May  7 22:52:16 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:16 cavasa CIFS VFS: Send error in Close = -5
May  7 22:52:32 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:32 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:32 cavasa CIFS VFS: Error 0xfffffffb or (-5 decimal) on
cifs_get_inode_info in lookup
May  7 22:52:32 cavasa CIFS VFS: Need to reconnect after session died to server
May  7 22:52:32 cavasa CIFS VFS: Send error in Close = -5
May  7 22:52:35 cavasa Unable to load NLS charset UTF8
May  7 22:52:38 cavasa CIFS VFS: CIFS: caught signal
May  7 22:52:39 cavasa CIFS VFS: No task to wake, unknown frame rcvd!

May  7 22:55:47 cavasa Badness in remove_proc_entry at fs/proc/generic.c:660
May  7 22:55:47 cavasa Call Trace:
May  7 22:55:47 cavasa [<c01751e7>] remove_proc_entry+0x111/0x14b
May  7 22:55:47 cavasa [<f8957a35>] cifs_proc_clean+0xf2/0xf7 [cifs]
May  7 22:55:47 cavasa [<f8968df8>] exit_cifs+0x8/0x9d [cifs]
May  7 22:55:47 cavasa [<c012abaf>] sys_delete_module+0x146/0x17f
May  7 22:55:47 cavasa [<c013eef3>] do_munmap+0x12a/0x160
May  7 22:55:47 cavasa [<c0103cd3>] syscall_call+0x7/0xb
May  7 22:55:47 cavasa
May  7 22:55:47 cavasa slab error in kmem_cache_destroy(): cache `cifs_request':
Can't free all objects
May  7 22:55:47 cavasa Call Trace:
May  7 22:55:47 cavasa [<c0135618>] kmem_cache_destroy+0xd2/0x110
May  7 22:55:47 cavasa [<f8952685>] cifs_destroy_request_bufs+0x10/0x26 [cifs]
May  7 22:55:47 cavasa [<f8968e13>] exit_cifs+0x23/0x9d [cifs]
May  7 22:55:47 cavasa [<c012abaf>] sys_delete_module+0x146/0x17f
May  7 22:55:47 cavasa [<c013eef3>] do_munmap+0x12a/0x160
May  7 22:55:47 cavasa [<c0103cd3>] syscall_call+0x7/0xb
May  7 22:55:47 cavasa
May  7 22:55:47 cavasa cifs_destroy_request_cache: error not all structures were
freed
May  7 22:56:00 cavasa CRON[14742]: (root) CMD (test -x /usr/sbin/run-crons &&
/usr/sbin/run-crons )
May  7 22:56:11 cavasa kmem_cache_create: duplicate cache cifs_request
May  7 22:56:11 cavasa ------------[ cut here ]------------
May  7 22:56:11 cavasa kernel BUG at mm/slab.c:1359!
May  7 22:56:11 cavasa invalid operand: 0000 [#1]
May  7 22:56:11 cavasa SMP
May  7 22:56:11 cavasa CPU:    1
May  7 22:56:11 cavasa EIP:    0060:[<c01352a1>]    Not tainted
May  7 22:56:11 cavasa EFLAGS: 00010202   (2.6.6-rc1)
May  7 22:56:11 cavasa EIP is at kmem_cache_create+0x450/0x555
May  7 22:56:11 cavasa eax: 00000030   ebx: f7200c10   ecx: c03b7184   edx: c02eae5c
May  7 22:56:11 cavasa esi: f8973b41   edi: f8973b41   ebp: d2babb80   esp: f71aff3c
May  7 22:56:11 cavasa ds: 007b   es: 007b   ss: 0068
May  7 22:56:11 cavasa Process modprobe (pid: 14752, threadinfo=f71ae000
task=da5aeed0)
May  7 22:56:11 cavasa Stack: c02b73c0 f8973b34 00000080 80002000 f71aff5c
d2babbd8 c0000000 ffffff80
May  7 22:56:11 cavasa 00003e80 00000000 f897a800 f71ae000 c02ec554 f895265b
f8973b34 00004100
May  7 22:56:11 cavasa 00000080 00002000 00000000 00000000 f897c0d1 40161008
c02ec570 f897a800
May  7 22:56:11 cavasa Call Trace:
May  7 22:56:11 cavasa [<f895265b>] cifs_init_request_bufs+0x37/0x51 [cifs]
May  7 22:56:11 cavasa [<f897c0d1>] init_cifs+0xd1/0x127 [cifs]
May  7 22:56:11 cavasa [<c012c55d>] sys_init_module+0x11d/0x20e
May  7 22:56:11 cavasa [<c0103cd3>] syscall_call+0x7/0xb
May  7 22:56:11 cavasa
May  7 22:56:11 cavasa Code: 0f 0b 4f 05 e4 6b 2b c0 8b 0b e9 6e ff ff ff 8b 47
50 c7 04
Comment 1 Steve French 2004-07-30 09:27:07 UTC
This should be fixed in current cifs code (post about 1.18 cifs vfs - see fs/cifs/CHANGES 
or modinfo on cifs.ko).