System Hangs on SMB Shares

[Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun Microsystems]

Greetings,

My FC2 system allows me to mount SMB shares entered in /etc/fstab/ just 
fine. But whenever I try to access them in the Nautilus file manager, my 
entire file system hangs and becomes unreachable. I can still use apps 
such as terminal and Mozilla, but to get file management back I need to 
reboot my system.

Is this a known problem?

Regards,

[Dexter Ang]

Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun 
Microsystems wrote:
>  Greetings,
>  
>  My FC2 system allows me to mount SMB shares entered in /etc/fstab/ just 
>  fine. But whenever I try to access them in the Nautilus file manager, my 
>  entire file system hangs and becomes unreachable. I can still use apps 
>  such as terminal and Mozilla, but to get file management back I need to 
>  reboot my system.
>  
>  Is this a known problem?
>  
>  Regards,
>   
How are you mounting the shares? Via the command:

mount //host/share /mnt/samba -o username=whatever

I presume? If I understand correctly, smbfs doesn't work correctly in 
the released kernel. I believe you should mount it with type cifs such as:

mount -t cifs //host/share /mnt/samba -o username=whatever

You could also check out the mount.cifs command via "man mount.cifs"

dex

[Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun Microsystems]

When I use mount.cifs I get the following error:

mount error 5 = Input/output error
Refer to the mount.cifs(8) manual page (e.g.man mount.cifs)

Not very helpful. :-(

/POL

Dexter Ang wrote:

>  Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software -
> Sun 
>  Microsystems wrote:
>
> > Greetings,
> >
> > My FC2 system allows me to mount SMB shares entered in /etc/fstab/ 
> > just fine. But whenever I try to access them in the Nautilus file 
> > manager, my entire file system hangs and becomes unreachable. I can 
> > still use apps such as terminal and Mozilla, but to get file 
> > management back I need to reboot my system.
> >
> > Is this a known problem?
> >
> > Regards,
> >
>
>  How are you mounting the shares? Via the command:
>
>  mount //host/share /mnt/samba -o username=whatever
>
>  I presume? If I understand correctly, smbfs doesn't work correctly in 
>  the released kernel. I believe you should mount it with type cifs such 
>  as:
>
>  mount -t cifs //host/share /mnt/samba -o username=whatever
>
>  You could also check out the mount.cifs command via "man mount.cifs"
>
>  dex
>
>

[Dexter Ang]

Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun 
Microsystems wrote:
>  When I use mount.cifs I get the following error:
>  
>  mount error 5 = Input/output error
>  Refer to the mount.cifs(8) manual page (e.g.man mount.cifs) 
What was the exact command you executed?

Will have to look into this myself as I've never experienced this error. 
Maybe try to google for this error as well and what causes it.

>  
>  Not very helpful. :-( 
man pages are usually very technical. You should see the man for the 
date command. And if it was me that you were referring to as not very 
helpful, sorry =) I try to be.

dex

>  
>  /POL
>  
>  Dexter Ang wrote:
>  
> > Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun 
> > Microsystems wrote:
> >
> >> Greetings,
> >>
> >> My FC2 system allows me to mount SMB shares entered in /etc/fstab/ 
> >> just fine. But whenever I try to access them in the Nautilus file 
> >> manager, my entire file system hangs and becomes unreachable. I can 
> >> still use apps such as terminal and Mozilla, but to get file 
> >> management back I need to reboot my system.
> >>
> >> Is this a known problem?
> >>
> >> Regards,
> >>
> >
> > How are you mounting the shares? Via the command:
> >
> > mount //host/share /mnt/samba -o username=whatever
> >
> > I presume? If I understand correctly, smbfs doesn't work correctly in 
> > the released kernel. I believe you should mount it with type cifs such 
> > as:
> >
> > mount -t cifs //host/share /mnt/samba -o username=whatever
> >
> > You could also check out the mount.cifs command via "man mount.cifs"
> >
> > dex
> >
> >
>

[Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun Microsystems]

Dexter Ang wrote:

>  [snip]
>
> >
> > Not very helpful. :-(
>
>
>  man pages are usually very technical. You should see the man for the 
>  date command. And if it was me that you were referring to as not very 
>  helpful, sorry =) I try to be.
>
>  dex 
Nono, I was not referring to you. I'll do some Googling and see what I 
find there.

thanks,
/POL

[Paul R. Ganci]

Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun 
Microsystems wrote:

>  My FC2 system allows me to mount SMB shares entered in /etc/fstab/ 
>  just fine. But whenever I try to access them in the Nautilus file 
>  manager, my entire file system hangs and becomes unreachable. I can 
>  still use apps such as terminal and Mozilla, but to get file 
>  management back I need to reboot my system.
>
>  Is this a known problem? 
I believe this is a bug as I and others have had this problem. You can 
check your /var/log/messages file to see if you have a log entry similar 
to the following:

smb_lookup: find //.Trash-jos failed, error=-5
Unable to handle kernel NULL pointer dereference at virtual address 00000000
 printing eip:
00000000
*pde = 00000000
Oops: 0000 [#1]
PREEMPT
CPU:    0
EIP:    0060:[<00000000>]    Not tainted
EFLAGS: 00010246   (2.6.5)
EIP is at 0x0
eax: d9f94c20   ebx: d68a3f30   ecx: c015e7c0   edx: d81c4ae0
esi: d68a3fa0   edi: c1329600   ebp: d44d85e0   esp: d68a3efc
ds: 007b   es: 007b   ss: 0068
Process nautilus (pid: 6808, threadinfo=d68a2000 task=d6d1cd60)
Stack: c01c16d6 d44d85e0 d68a3fa0 c015e7c0 d68a3f30 00000000 00000002 00000004
       d43c4ec4 00000000 d43c0000 d81c4ae0 d4ac8b60 00000000 fffe7b2a d44d85e0
       00000000 00000000 d43c0000 00000002 00000000 00000000 00000001 00000004
Call Trace:
 [<c01c16d6>] smb_readdir+0x3f6/0x5a0
 [<c015e7c0>] filldir64+0x0/0x120
 [<c015e489>] vfs_readdir+0x89/0xa0
 [<c015e7c0>] filldir64+0x0/0x120
 [<c015e94e>] sys_getdents64+0x6e/0xaa
 [<c015e7c0>] filldir64+0x0/0x120
 [<c0107009>] sysenter_past_esp+0x52/0x71
 
Code:  Bad EIP value.

I found a work around from this bug report 
(http://bugme.osdl.org/show_bug.cgi?id=2642):

 killall -STOP nautilus
smbmount //your/share  /mntpoint
killall -CONT nautilus

I can confirm that this procedure works for me as I have not had a hang 
since using it. You can read the bug report to get details on why this 
problem might be occurring. This procedure may not help you since you 
are using fstab to do the mounts, but as a temporary solution you could 
comment out the fstab entries and mount the shares manually using the 
procedure above.

-- 
Paul (ganci(a)nurdog.com)

[Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun Microsystems]

Paul, thanks for pointing this out. I did indeed find log entries which 
fit your description. Will investigate the workaround.

/POL

Paul R. Ganci wrote:

>  Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software -
> Sun 
>  Microsystems wrote:
>
> > My FC2 system allows me to mount SMB shares entered in /etc/fstab/ 
> > just fine. But whenever I try to access them in the Nautilus file 
> > manager, my entire file system hangs and becomes unreachable. I can 
> > still use apps such as terminal and Mozilla, but to get file 
> > management back I need to reboot my system.
> >
> > Is this a known problem?
>
>
>  I believe this is a bug as I and others have had this problem. You can 
>  check your /var/log/messages file to see if you have a log entry 
>  similar to the following:
>
>  smb_lookup: find //.Trash-jos failed, error=-5
>  Unable to handle kernel NULL pointer dereference at virtual address 
>  00000000
>  printing eip:
>  00000000
>  *pde = 00000000
>  Oops: 0000 [#1]
>  PREEMPT
>  CPU:    0
>  EIP:    0060:[<00000000>]    Not tainted
>  EFLAGS: 00010246   (2.6.5)
>  EIP is at 0x0
>  eax: d9f94c20   ebx: d68a3f30   ecx: c015e7c0   edx: d81c4ae0
>  esi: d68a3fa0   edi: c1329600   ebp: d44d85e0   esp: d68a3efc
>  ds: 007b   es: 007b   ss: 0068
>  Process nautilus (pid: 6808, threadinfo=d68a2000 task=d6d1cd60)
>  Stack: c01c16d6 d44d85e0 d68a3fa0 c015e7c0 d68a3f30 00000000 00000002 
>  00000004
>        d43c4ec4 00000000 d43c0000 d81c4ae0 d4ac8b60 00000000 fffe7b2a 
>  d44d85e0
>        00000000 00000000 d43c0000 00000002 00000000 00000000 00000001 
>  00000004
>  Call Trace:
>  [<c01c16d6>] smb_readdir+0x3f6/0x5a0
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c015e489>] vfs_readdir+0x89/0xa0
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c015e94e>] sys_getdents64+0x6e/0xaa
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c0107009>] sysenter_past_esp+0x52/0x71
>
>  Code:  Bad EIP value.
>
>  I found a work around from this bug report 
>  (http://bugme.osdl.org/show_bug.cgi?id=2642):
>
>  killall -STOP nautilus
>  smbmount //your/share  /mntpoint
>  killall -CONT nautilus
>
>  I can confirm that this procedure works for me as I have not had a 
>  hang since using it. You can read the bug report to get details on why 
>  this problem might be occurring. This procedure may not help you since 
>  you are using fstab to do the mounts, but as a temporary solution you 
>  could comment out the fstab entries and mount the shares manually 
>  using the procedure above.
>

[Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software - Sun Microsystems]

Well, the workaround partly resolved the problem - so far as to get 
Nautilus to browse the share without hanging, but I then tried to empty 
the trash (which didn't contain any files taken from the share) and it 
hangs again.

To me this seems to be a rather serious bug - surely there must be a 
patch or fix to this somewhere?

/POL

Paul R. Ganci wrote:

>  Per-Olof Litby - Reg'l Mgr Nordic/Baltic - Java System Software -
> Sun 
>  Microsystems wrote:
>
> > My FC2 system allows me to mount SMB shares entered in /etc/fstab/ 
> > just fine. But whenever I try to access them in the Nautilus file 
> > manager, my entire file system hangs and becomes unreachable. I can 
> > still use apps such as terminal and Mozilla, but to get file 
> > management back I need to reboot my system.
> >
> > Is this a known problem?
>
>
>  I believe this is a bug as I and others have had this problem. You can 
>  check your /var/log/messages file to see if you have a log entry 
>  similar to the following:
>
>  smb_lookup: find //.Trash-jos failed, error=-5
>  Unable to handle kernel NULL pointer dereference at virtual address 
>  00000000
>  printing eip:
>  00000000
>  *pde = 00000000
>  Oops: 0000 [#1]
>  PREEMPT
>  CPU:    0
>  EIP:    0060:[<00000000>]    Not tainted
>  EFLAGS: 00010246   (2.6.5)
>  EIP is at 0x0
>  eax: d9f94c20   ebx: d68a3f30   ecx: c015e7c0   edx: d81c4ae0
>  esi: d68a3fa0   edi: c1329600   ebp: d44d85e0   esp: d68a3efc
>  ds: 007b   es: 007b   ss: 0068
>  Process nautilus (pid: 6808, threadinfo=d68a2000 task=d6d1cd60)
>  Stack: c01c16d6 d44d85e0 d68a3fa0 c015e7c0 d68a3f30 00000000 00000002 
>  00000004
>        d43c4ec4 00000000 d43c0000 d81c4ae0 d4ac8b60 00000000 fffe7b2a 
>  d44d85e0
>        00000000 00000000 d43c0000 00000002 00000000 00000000 00000001 
>  00000004
>  Call Trace:
>  [<c01c16d6>] smb_readdir+0x3f6/0x5a0
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c015e489>] vfs_readdir+0x89/0xa0
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c015e94e>] sys_getdents64+0x6e/0xaa
>  [<c015e7c0>] filldir64+0x0/0x120
>  [<c0107009>] sysenter_past_esp+0x52/0x71
>
>  Code:  Bad EIP value.
>
>  I found a work around from this bug report 
>  (http://bugme.osdl.org/show_bug.cgi?id=2642):
>
>  killall -STOP nautilus
>  smbmount //your/share  /mntpoint
>  killall -CONT nautilus
>
>  I can confirm that this procedure works for me as I have not had a 
>  hang since using it. You can read the bug report to get details on why 
>  this problem might be occurring. This procedure may not help you since 
>  you are using fstab to do the mounts, but as a temporary solution you 
>  could comment out the fstab entries and mount the shares manually 
>  using the procedure above.
>