Bug 4171 - INTERNAL ERROR (panic) on attach to SMB share
INTERNAL ERROR (panic) on attach to SMB share
Status: RESOLVED INVALID
Product: Samba 3.0
Classification: Unclassified
Component: File Services
3.0.23c
x86 Windows XP
: P3 critical
: none
Assigned To: Samba Bugzilla Account
Samba QA Contact
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2006-10-13 08:40 UTC by Mike Tubby
Modified: 2006-11-17 09:50 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 Mike Tubby 2006-10-13 08:40:22 UTC
I have today (13-OCT-2006) installed RedHat FC5 to a new Gigibyte server, yum updated to all current RPMs, Kernel, etc., installed Samba 3.0.22-5 (because that's what yum gives us) and copied over a larghe quantity of data to a file store.

Built new Linux user accounts (useradd), new smbpasswd file, edited /etc/group and created a 'deve' group (see smb.con extract below), copied over share definitions and made new smb.conf

Bring up new server and everything starts as expcted, accessing from Windows XP-Pro ... Start->Run->\\SERVERNAME

Window opens correctly displaying share-names (folders) - so no problem with user authentication or hosts-allow.

Double click on any folder defined my my share definitions (example below) and smbd bangs out thus:

  PANIC (pid 14637): internal error
[2006/10/13 14:09:46, 0] lib/util.c:log_stack_trace(1721)
  BACKTRACE: 19 stack frames:
   #0 smbd(log_stack_trace+0x2d) [0xb5c8ad]
   #1 smbd(smb_panic+0x5d) [0xb5c9dd]
   #2 smbd [0xb4851a]
   #3 [0x110420]
   #4 /lib/libc.so.6(__strdup+0x1f) [0x2bc893]
   #5 /lib/libnsl.so.1(nis_list+0x5d2) [0x832b5f]
   #6 /lib/libnss_nisplus.so.2(_nss_nisplus_setnetgrent+0x8f) [0x46962e]
   #7 /lib/libc.so.6(innetgr+0xb2) [0x337d05]
   #8 smbd(user_in_netgroup+0x65) [0x9931e5]
   #9 smbd(token_contains_name_in_list+0x23d) [0x995bdd]
   #10 smbd(user_ok_token+0x8f) [0x99600f]
   #11 smbd [0x9f388f]
   #12 smbd(make_connection+0x194) [0x9f4fa4]
   #13 smbd(reply_tcon_and_X+0x21d) [0x9b88ed]
   #14 smbd [0x9f00e0]
   #15 smbd(smbd_process+0x7ab) [0x9f121b]
   #16 smbd(main+0xbd0) [0xc0af90]
   #17 /lib/libc.so.6(__libc_start_main+0xdc) [0x2684e4]
   #18 smbd [0x97b701]
[2006/10/13 14:09:47, 0] lib/fault.c:dump_core(173)
  dumping core in /var/log/samba/cores/smbd
[2006/10/13 14:09:47, 0] lib/fault.c:fault_report(41)
  ===============================================================
[2006/10/13 14:09:47, 0] lib/fault.c:fault_report(42)
  INTERNAL ERROR: Signal 11 in pid 14638 (3.0.23c-1.fc5)
  Please read the Trouble-Shooting section of the Samba3-HOWTO
[2006/10/13 14:09:47, 0] lib/fault.c:fault_report(44)

  From: http://www.samba.org/samba/docs/Samba3-HOWTO.pdf
[2006/10/13 14:09:47, 0] lib/fault.c:fault_report(45)
  ===============================================================
[2006/10/13 14:09:47, 0] lib/util.c:smb_panic(1614)
  PANIC (pid 14638): internal error
[2006/10/13 14:09:47, 0] lib/util.c:log_stack_trace(1721)
  BACKTRACE: 19 stack frames:
   #0 smbd(log_stack_trace+0x2d) [0xb5c8ad]
   #1 smbd(smb_panic+0x5d) [0xb5c9dd]
   #2 smbd [0xb4851a]
   #3 [0x110420]
   #4 /lib/libc.so.6(__strdup+0x1f) [0x2bc893]
   #5 /lib/libnsl.so.1(nis_list+0x5d2) [0x832b5f]
   #6 /lib/libnss_nisplus.so.2(_nss_nisplus_setnetgrent+0x8f) [0x46962e]
   #7 /lib/libc.so.6(innetgr+0xb2) [0x337d05]
   #8 smbd(user_in_netgroup+0x65) [0x9931e5]
   #9 smbd(token_contains_name_in_list+0x23d) [0x995bdd]
   #10 smbd(user_ok_token+0x8f) [0x99600f]
   #11 smbd [0x9f388f]
   #12 smbd(make_connection+0x194) [0x9f4fa4]
   #13 smbd(reply_tcon_and_X+0x21d) [0x9b88ed]
   #14 smbd [0x9f00e0]
   #15 smbd(smbd_process+0x7ab) [0x9f121b]
   #16 smbd(main+0xbd0) [0xc0af90]
   #17 /lib/libc.so.6(__libc_start_main+0xdc) [0x2684e4]
   #18 smbd [0x97b701]
[2006/10/13 14:09:47, 0] lib/fault.c:dump_core(173)
  dumping core in /var/log/samba/cores/smbd
(END)


Running GDB for the most recent core-dump gets:

(gdb) where
#0  0x00110402 in __kernel_vsyscall ()
#1  0x0027aee9 in raise () from /lib/libc.so.6
#2  0x0027c4f1 in abort () from /lib/libc.so.6
#3  0x00b47fea in dump_core () from /usr/sbin/smbd
#4  0x00b5ca98 in smb_panic () from /usr/sbin/smbd
#5  0x00b4851a in dump_core_setup () from /usr/sbin/smbd
#6  <signal handler called>
#7  0x002bc893 in strdup () from /lib/libc.so.6
#8  0x00832b5f in nis_list () from /lib/libnsl.so.1
#9  0x0046962e in _nss_nisplus_setnetgrent () from /lib/libnss_nisplus.so.2
#10 0x00337d05 in innetgr () from /lib/libc.so.6
#11 0x009931e5 in user_in_netgroup () from /usr/sbin/smbd
#12 0x00995bdd in token_contains_name_in_list () from /usr/sbin/smbd
#13 0x0099600f in user_ok_token () from /usr/sbin/smbd
#14 0x009f388f in set_conn_connectpath () from /usr/sbin/smbd
#15 0x009f4fa4 in make_connection () from /usr/sbin/smbd
#16 0x009b88ed in reply_tcon_and_X () from /usr/sbin/smbd
#17 0x009f00e0 in open_was_deferred () from /usr/sbin/smbd
#18 0x009f121b in smbd_process () from /usr/sbin/smbd
#19 0x00c0af90 in main () from /usr/sbin/smbd
(gdb)


Shares are defined in the smb.conf file like this:

[DevTools]
   comment = EPROM Development
   path = /data/DevTools
   public = no
   writable = yes
   printable = no
   valid users = @devel
   write list = @devel
   force group = devel
   create mask = 0770
   force create mode = 0770
   directory mask = 0770
   force directory mode = 2770



Regards


Mike Tubby
mike.tubby@thorcom.co.uk
Comment 1 Volker Lendecke 2006-10-13 08:51:11 UTC
Just to make sure: This is _really_ 3.0.22 and not 3.0.23? I'm asking because we have a number of reports that with 3.0.23 smbd segfaults in user_in_netgroup. We've done a lot of changes for .23, but if you see this with .22 as well I'm very confident that this is a FC5 libc thing, not a Samba thing.

If you're not really using NIS, you might use +group instead of @group as a workaround.

Volker
Comment 2 Mike Tubby 2006-10-15 05:33:14 UTC
All,

I made a mistake in the original posting of this bug report and statd that the version was 3.0.22 when it had actually yum-updated to 3.0.23c (RedHat FC5).

All other comments remain the same.

Mike
Comment 3 Volker Lendecke 2006-11-17 09:50:20 UTC
Ok, this is a fc5 glibc bug.

#include <netdb.h>
int main(void)
{
        innetgr("deve", "foo", "testuser", NULL);
}

segfaults, so we can not do anything about this.

Volker