[Date Prev][Date Next]
Re: SIGSEGV in OpenLDAP 2.3.20 Stable
--On Friday, March 31, 2006 11:26 AM +0200 Xavier <firstname.lastname@example.org>
I'm having the folowing problem with OpenLDAP 2.3.20 (stable 20060227)
which acts as a replica.
the segfault occurs when my master LDAP server which is 2.0.27 sends an
ldap modify on a DN that doesn't exists. ( I know this shouldn't happen
because this means that there is a synchonisation problem )
I think it would be better to have an error instead of a segmentation
e is NULL in the call of cache_return_entry_rw()
Following is the backtrace of the crash.
I hope this will help.
<= str2entry(ou=testrb,ou=ORG,o=CORP) -> 0xa218a08
<= id2entry_r( 78335 ) 0xa218a08 (disk)
====> cache_return_entry_r( 78335 ): created (0)
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1213113424 (LWP 16455)]
cache_return_entry_rw (cache=0xa1c220c, e=0x0, rw=1) at cache.c:111
111 assert( e->e_private != NULL );
# 0 cache_return_entry_rw (cache=0xa1c220c, e=0x0, rw=1) at cache.c:111
# 1 0x080bfb72 in ldbm_back_modify (op=0xa216f98, rs=0xb7b15230) at
# 2 0x08070496 in fe_op_modify (op=0xa216f98, rs=0xb7b15230) at
# 3 0x08070f04 in do_modify (op=0xa216f98, rs=0xb7b15230) at
# 4 0x0805e28d in connection_operation (ctx=0xb7b152b0, arg_v=0xa216f98)
# 5 0x0810b38f in ldap_int_thread_pool_wrapper (xpool=0xa184da8) at
# 6 0x00739341 in start_thread () from /lib/tls/libpthread.so.0
# 7 0x0063d6fe in clone () from /lib/tls/libc.so.6
Thanks for the report. :)
On a side note, the bug tracking system is at:
<http://www.openldap.org/its/>, if you want to file this there so it is
I'll also note that it looks like you are using LDBM in OL 2.3, which is
generally recommended against for OL 2.2-OL 2.3. It will be removed from
the OL 2.4 release branch. You may wish to switch off of using LDBM, see:
for the various reasons why.
Principal Software Developer
ITS/Shared Application Services
GnuPG Public Key: http://www.stanford.edu/~quanah/pgp.html