openldap-bugs

openldap-bugs@openldap.org

1 participants
20993 discussions

Re: (ITS#4855) slapd crash on exit (tpool)
by ando＠sys-net.it 05 Mar '07

05 Mar '07

gael.roualland(a)oleane.net wrote: > Full_Name: Gaël Roualland > Version: 2.3.34 > OS: Linux > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (213.56.0.199) Good catch. This is a re23-only bug; fixed, will appear in next release. Please test re23 out of the CVS, if you can. Thanks, p. Ing. Pierangelo Masarati OpenLDAP Core Team SysNet s.n.c. Via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ------------------------------------------ Office: +39.02.23998309 Mobile: +39.333.4963172 Email: pierangelo.masarati(a)sys-net.it ------------------------------------------

1 0

Re: (ITS#4855) slapd crash on exit (tpool)
by gael.roualland＠oleane.net 05 Mar '07

05 Mar '07

Erratum : the changed line is "( ctx[i].ltk_key == NULL )" and not "( ctx[i].ltk_key != NULL )" which is the same as the release... -- Gaël Roualland -+- gael.roualland(a)oleane.net

1 0

(ITS#4855) slapd crash on exit (tpool)
by gael.roualland＠oleane.net 05 Mar '07

05 Mar '07

Full_Name: Gaël Roualland Version: 2.3.34 OS: Linux URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (213.56.0.199) Hello, Upgrading 2.3.33 to 2.3.34 introduced a bug on slapd stop: the daemon segfaults while waiting for threads to stop. This is reproducible on every run, and without any queries processed. Running under gdb on a test system, the following is reported: slapd starting [New Thread 32769 (LWP 22877)] [New Thread 16386 (LWP 22878)] daemon: added 7r listener=(nil) daemon: added 9r listener=0x813df78 [New Thread 32771 (LWP 22879)] daemon: select: listen=9 active_threads=0 tvp=zero Program received signal SIGTERM, Terminated. [Switching to Thread 16384 (LWP 22876)] 0x40070604 in __pthread_sigsuspend () from /lib/libpthread.so.0 (gdb) n Single stepping until exit from function __pthread_wait_for_restart_signal, which has no line number information. daemon: shutdown requested and initiated. daemon: closing 9 slapd shutdown: waiting for 0 threads to terminate Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 32771 (LWP 22879)] 0x401e1aa0 in __check_rhosts_file () from /lib/libc.so.6 (gdb) bt #0 0x401e1aa0 in __check_rhosts_file () from /lib/libc.so.6 #1 0x080c3c36 in ldap_pvt_thread_pool_context_reset (vctx=0xbf5ff94c) at tpool.c:670 #2 0x080c38df in ldap_int_thread_pool_wrapper (xpool=0x814aa78) at tpool.c:494 #3 0x4006de51 in pthread_start_thread () from /lib/libpthread.so.0 #4 0x4006decf in pthread_start_thread_event () from /lib/libpthread.so.0 #5 0x401908aa in clone () from /lib/libc.so.6 (gdb) info threads * 4 Thread 32771 (LWP 22879) 0x401e1aa0 in __check_rhosts_file () from /lib/libc.so.6 3 Thread 16386 (LWP 22878) 0x40070604 in __pthread_sigsuspend () from /lib/libpthread.so.0 2 Thread 32769 (LWP 22877) 0x40187a5a in poll () from /lib/libc.so.6 1 Thread 16384 (LWP 22876) 0x40070604 in __pthread_sigsuspend () from /lib/libpthread.so.0 Looking at ldap_pvt_thread_pool_context_reset, it seems that the test "if ( ctx[i].ltk_key )" should be reversed as "if ( ctx[i].ltk_key != NULL )". This change fixed the bug. Regards,

1 0

Re: (ITS#4789) can't add BDB database when monitoring
by ando＠sys-net.it 05 Mar '07

05 Mar '07

For the records: I'm sort of working at this. What I'm trying to do is add some specialized calls to monitor_extra_t that allow to register backends, databases, overlays, overlay instances and so. This would allow entity registration/deregistration routines to provide the related monitor entry registration/deregistration as appropriate from within back-config, with all sorts of atomicity issues one can imagine... Cheers, p. Ing. Pierangelo Masarati OpenLDAP Core Team SysNet s.n.c. Via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ------------------------------------------ Office: +39.02.23998309 Mobile: +39.333.4963172 Email: pierangelo.masarati(a)sys-net.it ------------------------------------------

1 0

Re: (ITS#4816) slurpd generates invalid ldapadd/modify requests
by gael.roualland＠oleane.net 01 Mar '07

01 Mar '07

This is a multi-part message in MIME format. --------------090002080605050903080106 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hello, >> As far as I can see, slurpd merely processes the changes in the replog. If it >> is generating requests out of sequence, then they must have been recorded >> out-of-sequence in the log. And yet I don't see any reason why slapd would >> generate the modifications out of sequence. > > That's right for changetype=modify, since the original code aggregates > multiple values of the same attributes in a signle LDAPmod if they're > given in order in the replog file. > But it doesn't do that for changetype=add, which toggled the bug. I > choosed to use the same code path for both for simplicity when fixing add. > >> What are the circumstances that cause these improper sequences to get into >> the replog in the first place? > > In our case, it was just "add" that was problematic. Although I could > imagine the replog to be generated by something else than slapd, and I'd > expect slurpd to be able to process it as long as it is valid ldif ? The second patch I uploaded has an issue with deleting whole attributes. I have a fixed version of it, but I also wrote another patch to fix the initial issue only on the problematic case (entry adding), so that's a less intrusive change : instead of adding a new LDAPMod for each attribute/value pair, the previous LDAPMod (if any) is reused and the current value added to it if it's for the same attribute. The new patch is attached (no space left on ftp). Would it be possible for such a fix to be in mainline ? Thanks, -- Gaël Roualland -+- gael.roualland(a)oleane.net --------------090002080605050903080106 Content-Type: text/x-patch; name="openldap2.3-slurpd-ldapadd.diff" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="openldap2.3-slurpd-ldapadd.diff" --- openldap-2.3.33/servers/slurpd/ldap_op.c.orig 2007-01-25 12:01:46.000000000 +0100 +++ openldap-2.3.33/servers/slurpd/ldap_op.c 2007-03-01 18:05:44.315232000 +0100 @@ -52,7 +52,6 @@ #include "slurp.h" /* Forward references */ -static struct berval **make_singlevalued_berval LDAP_P(( char *, int )); static int op_ldap_add LDAP_P(( Ri *, Re *, char **, int * )); static int op_ldap_modify LDAP_P(( Ri *, Re *, char **, int * )); static int op_ldap_delete LDAP_P(( Ri *, Re *, char **, int * )); @@ -183,28 +182,42 @@ ) { Mi *mi; - int nattrs, rc = 0, i; + int nattrs, nvals, rc = 0, i; LDAPMod *ldm, **ldmarr; int lderr = 0; nattrs = i = 0; ldmarr = NULL; + ldm = NULL; /* * Construct a null-terminated array of LDAPMod structs. */ mi = re->re_mods; while ( mi[ i ].mi_type != NULL ) { - ldm = alloc_ldapmod(); - ldmarr = ( LDAPMod ** ) ch_realloc( ldmarr, - ( nattrs + 2 ) * sizeof( LDAPMod * )); - ldmarr[ nattrs ] = ldm; - ldm->mod_op = LDAP_MOD_BVALUES; - ldm->mod_type = mi[ i ].mi_type; - ldm->mod_bvalues = - make_singlevalued_berval( mi[ i ].mi_val, mi[ i ].mi_len ); + if (ldm == NULL || strcmp(ldm->mod_type, mi[ i ].mi_type) != 0) { + /* not the same attribute, add a new ldm */ + ldm = alloc_ldapmod(); + ldmarr = ( LDAPMod ** ) ch_realloc( ldmarr, + ( nattrs + 2 ) * sizeof( LDAPMod * )); + ldmarr[ nattrs ] = ldm; + nattrs++; + ldm->mod_op = LDAP_MOD_BVALUES; + ldm->mod_type = mi[ i ].mi_type; + nvals = 0; + } + + ldm->mod_bvalues = ( struct berval ** ) + ch_realloc( ldm->mod_bvalues, + ( nvals + 2 ) * sizeof( struct berval * )); + ldm->mod_bvalues[ nvals + 1 ] = NULL; + ldm->mod_bvalues[ nvals ] = ( struct berval * ) + ch_malloc( sizeof( struct berval )); + ldm->mod_bvalues[ nvals ]->bv_val = mi[ i ].mi_val; + ldm->mod_bvalues[ nvals ]->bv_len = mi[ i ].mi_len; + nvals++; + i++; - nattrs++; } if ( ldmarr != NULL ) { @@ -579,25 +592,6 @@ /* - * Create a berval with a single value. - */ -static struct berval ** -make_singlevalued_berval( -char *value, -int len ) -{ - struct berval **p; - - p = ( struct berval ** ) ch_malloc( 2 * sizeof( struct berval * )); - p[ 0 ] = ( struct berval * ) ch_malloc( sizeof( struct berval )); - p[ 1 ] = NULL; - p[ 0 ]->bv_val = value; - p[ 0 ]->bv_len = len; - return( p ); -} - - -/* * Given a modification type (string), return an enumerated type. * Avoids ugly copy in op_ldap_modify - lets us use a switch statement * there. --------------090002080605050903080106--

1 0

Re: ITS#4853
by ando＠sys-net.it 27 Feb '07

27 Feb '07

Richard Bos wrote: > Hello Pierangelo, > > Op dinsdag 27 februari 2007 22:27, schreef u: >> Fixed in re23; please test. >> >> I note that the name of the macro may have tricked you; Debug() is >> actually the logging system of OpenLDAP software, not a mere debug >> statement. > > I assume I may not update the ticket, You should never __assume__, when the system is self-documented: > Replies and Followups > --------------------- > > A reply is a message that has been sent from JitterBug to answer an > incoming message. > > A followup is a reply to a reply. Followups are created when a user > sends a email reply to a JitterBug generated message. Followups are > detected by looking for the PR# automatically added to the subject > line in replies. > hence this email to you: > > What do you mean with fixed in "re23". The code has been fixed in the release engineering branch of 2.3 code, and will thus appear in the next release. > Which release is that exactly? Likely, 2.3.35, if any. > What does the change involve? Look at the code from the CVS. > Is the message now sent to a log file? > < Debug(LDAP_DEBUG_ANY, "%s: line %d: new attribute <%s>\n", >> Debug(LDAP_DEBUG_CONFIG, "%s: line %d: new attribute <%s>\n", It's sent wherever configured, like it was before (see slapd(8) and slapd.conf(5) for details about logging). The difference is that earlier that message was being always issued, while now it's only issued if expressly requested, since it's just informative and not indicative of any bug. The macros seem pretty self-esplicative. I hope that bug didn't cause any harm to your system, though. p. Ing. Pierangelo Masarati OpenLDAP Core Team SysNet s.n.c. Via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ------------------------------------------ Office: +39.02.23998309 Mobile: +39.333.4963172 Email: pierangelo.masarati(a)sys-net.it ------------------------------------------

1 0

ITS#4853
by ando＠sys-net.it 27 Feb '07

27 Feb '07

Fixed in re23; please test. I note that the name of the macro may have tricked you; Debug() is actually the logging system of OpenLDAP software, not a mere debug statement. p. Ing. Pierangelo Masarati OpenLDAP Core Team SysNet s.n.c. Via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ------------------------------------------ Office: +39.02.23998309 Mobile: +39.333.4963172 Email: pierangelo.masarati(a)sys-net.it ------------------------------------------

1 0

(ITS#4853) A debug message is printed, from the released version
by ml＠radoeka.nl 27 Feb '07

27 Feb '07

Full_Name: Richard Bos Version: 2.3.27 OS: linux URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (80.126.75.243) I just upgraded to the new slapd.conf file from cvs and now when I run slapcat I get the following output: # slapcat >/dev/null /etc/openldap/slapd.conf: line 67: new attribute <member> The config file contains: 63 #### Using overlays to improve data consistency 64 # Ensure that we never get dangling member attributes 65 # Checked on rename and delete 66 overlay refint 67 refint_attributes member 68 The code in servers/slapd/overlays/refint.c reads as follows: if(!strcasecmp(*argv, "refint_attributes")) { for(i = 1; i < argc; i++) { ... some checks for error conditions here but if all is fine we fall through to the following section ... ip = ch_malloc(sizeof(refint_attrs)); ip->attr = ad; ip->next = id->attrs; id->attrs = ip; Debug(LDAP_DEBUG_ANY, "%s: line %d: new attribute <%s>\n", fname, lineno, argv[i]); The final debug message results seems to be the message that is printed. It's my assumption that no debug messages should be printed in a released product/version/project. The initial bug report is located at: https://intevation.de/roundup/kolab/issue1622

1 0

Re: (ITS#4845) Back-perl is non working
by hyc＠symas.com 27 Feb '07

27 Feb '07

kiwi(a)oav.net wrote: >> No surprise that such a careful redesign resulted, as a side >> effect, in a stricter compliance with RFC2589. The fact that >> nobody complained up to now indicates that strict adherence to >> specs usually helps in code portability. > > Yeah. But I didn't know that RFC2589 is used inside back_<whatever>. > Now I am aware of that and I really thanks about your help and your > kindness to take some time to find where my error was. I think he meant RFC2849, which is the LDIF specification. > Maybe a little not in slapd-perl man page that entry sent to openldap > have to conform to RFC2589 should be a point to avoid such loss of time. The slapd-perl(5) man page says that it expects LDIF entries to be returned. As such, it is already a given that they must actually conform to the LDIF spec. > Thanks again, and this bug is closed for me. -- -- Howard Chu Chief Architect, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

Potential bug (schema or ldap add?)
by Andrew Seguin 27 Feb '07

27 Feb '07

I'm not too sure of the details yet, I wanted to check with this list first, but here's the story. In general, I'm using phpLDAPadmin to work with the ldap directory. I added the iaaa-radius.schema from Interlink. I found it was missing the EQUALITY statements, so phpLDAPadmin couldn't "add" new aaaReply items once the first one was added. Via import and adding a user all at once was not a problem. I modified the iaaa-radius.schema file, found the aaaReply attribute and added the lines "EQUALITY caseIgnoreMatch" and "SUBSTR caseIgnoreSubstringsMatch". I went into phpLDAPadmin, created a new aaaPerson with one initial aaaReply attribute, then added a second aaaReply value without problem. To date all is good. I went to a user which already exists, which already has three aaaReply attributes/values, and tried to add a new aaaReply value - at which point it would seem that slapd either crashed or shutdown. I turned on logging and couldn't see anything really abnormal... nor did I find any core dump file. Any ideas if this is a bug or possibly just some kind of internal consistency check on records created before the Schema was modified? Thank you, Andrew

3 3

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs