[Samba] Signal 11 in smbd 3.0.2rc2 on printer operation!

[Jerry Haltom]

Follow up on this. I have SAMBA_3_0 from CVS, checked out about 4 days
ago (last time we discussed this). I have been able to cause hte
problem, or another similar problem, im not sure, to be reproduced.

What I was doing was adding a new printer, by editing smb.conf, adding
the new printer entry, and then SIGHUPping all smbd processes. After
wards, the new printer began working, but after printing the first page,
all printers immediatly ceased working. The following messages in no
particular order are now appearing in my logs:

tdb(/opt/samba/var/locks/printing/pdfwriter.tdb): tdb_lock failed on
list 4128 ltype=0 (Bad file descriptor)

and

[2004/02/10 13:27:44, 0] tdb/tdbutil.c:tdb_log(724)
  tdb(/opt/samba/var/locks/printing/hpljet8100n.tdb): tdb_oob len
842018636 beyond eof at 40960

Perhaps the action of adding a new printer and SIGHUPping smbd is
causing these problems? I will once again attempt to fix this once
people leave the office by clearing all the .tdb files and restarting
samba.

=)

On Thu, 2004-02-05 at 11:29, Jerry Haltom wrote:
> Nope. In fact, I do suspect bad ram. I don't know why it would be
> consistant though. I would think that bad ram would effect more than
> just samba's printing. And you'd think reacreating the process
> (different location in memory) would get around it. The problem is
> pretty consistant, it's happening for every computer in the office now.
> 
> On Thu, 2004-02-05 at 10:08, Gerald (Jerry) Carter wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > 
> > Jerry Haltom wrote:
> > |
> > | Could a corrupt .tdb of any sort be causing this? It
> > | is touching a variable named "pdb" (maybe printer db?),
> > | but I'm not sure what that is.
> > 
> > It could I guess, but I agree with Jeremy.  That's a
> > strange place to crash.  How many machines did you
> > see this on?  If only one, have you ruled out bad RAM ?
> > 
> > On Wed, 2004-02-04 at 19:08, Jeremy Allison wrote:
> > 
> > |>> #7  0x0820b54b in print_queue_status (snum=1, ppqueue=0xbffff4d0,
> > |>>    status=0xbffff4d4) at printing/printing.c:2283
> > |>>	keystr = "STATUS/hpljet8100n\0\0\001\0\0\0 öÿ¿ V\034@øB<\bò\006
> > \b at L=\b\0\0\0\0\0\0\0\0Â\021}\034_duB\227PñØ.\232:´\0\0\0\0\0\0\0\0Döÿ¿,ôÿ¿Q0\026\b\004\0\0\0Döÿ¿\024\0\0\0\0\0\0\0\001\0\0\0\0\0\0\0\004\0\0\0\0\0\0\0\002\0\0\08U=\bLôÿ¿\0361\026\b\220\026=\bDöÿ¿xôÿ¿\004\0\0\0\0\0\0\0\001\0\0\0|ôÿ¿ð\214!\b\001\0\0\0Döÿ¿xôÿ¿,L\023\b\0\0\0\0\224ýÿ¿\214ôÿ¿Kð\017\b"...
> > |>>	data = {dptr = 0x0, dsize = 264}
> > |>>	key = {dptr = 0xbffff3ac "STATUS/hpljet8100n", dsize = 18}
> > |>>	printername = 0x82f6c40 "hpljet8100n"
> > |>>	pdb = (struct tdb_print_db *) 0x83d4ac8
> > |>>	count = 0
> > |>
> > |> This is a strange crash location. It appears to be in the call here :
> > |>
> > |>        if (!get_stored_queue_info(pdb, snum, &count, ppqueue)) {
> > |>                release_print_db(pdb);
> > |>                return 0;
> > |>
> > |> I don't immediately see any bad pointers etc. in this call...
> > 
> > 
> > 
> > 
> > cheers, jerry
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.2.1 (GNU/Linux)
> > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
> > 
> > iD8DBQFAImqUIR7qMdg1EfYRAikcAJ43erXruQhRuLE7arSDsDUROwnVuQCgyLwU
> > 8T0AGaPERFSb0WPYVczJEKE=
> > =G3Fk
> > -----END PGP SIGNATURE-----
-- 
Jerry Haltom <jhaltom at feedbackplusinc.com>
Feedback Plus, Inc.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.samba.org/archive/samba/attachments/20040210/6da4951d/attachment-0001.bin