[Date Prev][Date Next]
Segfault with current on a search (ITS#451)
Full_Name: Ian Logan
OS: FreeBSD 3.4
Submission from: (NULL) (22.214.171.124)
I've been playing with current just for fun, out production stuff is using
This is all being done on a FreeBSD 3.4 machine.
Basicaly we have a simple phonebook application. When we search it with
always core dumps at the same place (more information included below). I've
just about every variation of things in the config file I can think of, it
a difference. If I turn off index's then the crash happens instantly, while
index's (of the cn,sn,givenname fields) on it takes a few minutes of searching
I dont really have time this week to dig through the code to figure out whats
on, so I'm posting this on the off chance that someone who knows the code better
will know whats going on. Otherwise I'll try to figure things out starting next
I've never worked on the OpenLDAP source code before, any pointers would be
This was with todays source code, allthough its been this way for at least the
Here is the log output at level 256 (I'd be happy to do any experiments if
more or different data would be usefull)
conn=0 op=0 BIND dn="" method=128
conn=0 op=0 RESULT tag=97 err=0 text=
conn=0 op=1 SRCH base="dc=NMSU,dc=Edu" scope=2 filter="(|(cn=IAN*
AN* LOGAN*)(sn=IAN* LOGAN*))"
Segmentation fault (core dumped)
Slapd always seg faults at the same line of code in the backend.
gdb slapd slapd.core
Program terminated with signal 11, Segmentation fault.
#0 0x80697d7 in substring_comp_candidates (be=0x80c9400,
type=0x80d10c0 "mail", bv=0x80d1190, prepost=0) at filterindex.c:361
361 buf[i] = p[i];
And finally, here is the relevant parts of the config I'm using. I've tried
this with and without schema checking, with the version 1.2.x and 2.x schema
suffix "dc=NMSU, dc=Edu"
And lastly, here is the the ldif file needed to produce the database (its
description: New Mexico State University
dn: ou=People, dc=NMSU, dc=Edu
cn: ian logan
If there is any information that would be helpful that I failed to include,
let me know.
Thanks for your time!