openldap-bugs

openldap-bugs@openldap.org

1 participants
20959 discussions

Re: (ITS#6139) slapd: password-hash rejects multiple values
by michael＠stroeder.com 25 May '09

25 May '09

fumiyas(a)osstech.co.jp wrote: > My /etc/openldap/slapd.conf has the following line: > password-hash {CRYPT} {SSHA} What exactly do you want to achieve by this? Ciao, Michael.

1 0

Compile error on tru64 5.1b
by Didier Godefroy 25 May '09

25 May '09

Hello, Compiling openldap 2.4.26 on tru64 unix v5.1b, using the native compilers and gmake. I'm getting an error as follows: Entering subdirectory liblunicode gmake[2]: Entering directory `/usr/local/openldap/openldap-2.4.16/libraries/liblunicode' cc -O4 -g3 -w -I../../include -I../../include -pthread -I/usr/local/include -pthread -I/usr/local/include -c -o ure.o ure.c cc: Fatal: A memory access violation (bus error or segmentation fault) has occurred. Please submit a problem report. gmake[2]: *** [ure.o] Error 1 I retried, just in case, and got the same error again. I'm not a C programmer and I have no idea how to fix this. What can I try and send to get this fixed? Thanks, -- Didier Godefroy mailto:dg@ulysium.net Support anti-Spam legislation. Join the fight http://www.cauce.org/

2 1

Re: (ITS#6131) "TLSVerifyClient try" not working with GNU TLS
by hyc＠symas.com 25 May '09

25 May '09

subbarao(a)computer.org wrote: > Full_Name: Kartik Subbarao > Version: 2.4.16 > OS: Debian 5.0.1 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (76.99.175.5) > > > When TLSVerifyClient is set to "try", OpenLDAP improperly rejects SSL > connections without a client certificate. The problem appears to start with this > section of code in tls.c around line 1564: > > #ifdef HAVE_GNUTLS > if ( ld->ld_options.ldo_tls_require_cert != LDAP_OPT_X_TLS_NEVER ) { > err = tls_cert_verify( ssl ); > if ( err&& ld->ld_options.ldo_tls_require_cert != LDAP_OPT_X_TLS_ALLOW > ) > return err; > } > #endif > > tls_cert_verify() calls gnutls_certificate_verify_peers2(), which appears to > return error 49 when no client certificate is presented. tls_cert_verify() > doesn't seem to distinguish between this case, and the case of an invalid client > certificate, returning -1 in both cases. > This bug report makes no sense; the code you quoted is not part of OpenLDAP 2.4.16. The relevant code is in function tlsg_session_accept() in tls_g.c, and there is no such bug in that function. -- -- Howard Chu CTO, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

(ITS#6140) nadf.schema: AttributeType not found: "lastModifiedTime"
by michael＠stroeder.com 25 May '09

25 May '09

Full_Name: Michael Ströder Version: RE24 OS: Linux URL: Submission from: (NULL) (84.163.124.107) When including nadf.schema slapd does not start and prints error message: /opt/openldap-RE24/etc/openldap/schema/nadf.schema: line 111 objectclass: AttributeType not found: "lastModifiedTime" /etc/openldap/slapd.conf: line 17: <include> handler exited with 1! Analyzing this I found this commented declaration in cosine.schema: ## Deprecated in favor of modifyTimeStamp #attributetype ( 0.9.2342.19200300.100.1.23 NAME 'lastModifiedTime' # DESC 'RFC1274: time of last modify, replaced by modifyTimestamp' # OBSOLETE # SYNTAX 1.3.6.1.4.1.1466.115.121.1.53 # USAGE directoryOperation ) So it seems nadf.schema is not usable at the moment.

1 0

(ITS#6139) slapd: password-hash rejects multiple values
by fumiyas＠osstech.co.jp 25 May '09

25 May '09

Full_Name: SATOH Fumiyasu Version: 2.4.16 OS: Linux URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (115.176.213.88) My /etc/openldap/slapd.conf has the following line: password-hash {CRYPT} {SSHA} On OpenLDAP 2.4.16, slapd (slaptest) rejects this config: # /usr/sbin/slaptest /etc/openldap/slapd.conf: line 36: <password-hash> extra cruft after <hash>. slaptest: bad configuration file! On OpenLDAP 2.3.43, slapd (slaptest) accept this config. The slapd.conf(5) (included in OpenLDAP 2.3.43/2.4.16) manpage said: password-hash <hash> [<hash>...] This option configures one or more hashes to be used in generation of user passwords stored in the userPassword attribute during processing of LDAP Password Modify Extended Operations (RFC 3062).

1 0

Re: (ITS#6104) race condition with cancel operation
by h.b.furuseth＠usit.uio.no 23 May '09

23 May '09

hyc(a)symas.com writes: >> Bitflags should be unnecessary as such, just need a set of values. >> Though it wouldn't hurt if they were chosen so one could use bit >> operations to reduce checks for "o_abandon == A or B". "Suppress >> response" value mentioned in ITS#6138 also needed, maybe that must be a >> bitflag when combined with Cancel stuff. > > Yes, it still seems some bitflags will be more appropriate. As long as we decide which flag combinations are (not) valid. Problem with "unrestricted" bitflags compared to a list of values is that they allow and thus require support for flag combinations we have no real need to support. >> Question: >> syncprov_op_abandon() holds&si->si_ops_mutex when setting >> so->s_op->o_abandon. Does it need that? The function needs to >> grab op->o_conn->c_mutex when called as Cancel, but must not do >> that while holding&si->si_ops_mutex since Abandon grabs the >> locks in the opposite order. > > syncprov_op_abandon() can remove ops from the list, so yes, it must > hold si_ops_mutex. Right, but does it need si_ops_mutex while setting o_abandon? Cancel could remove the op, unlock si_ops_mutex, lock c_mutex, update o_abandon. Alternative: Lock c_mutex first (that must be safe since Abandon is doing it anyway), then lock si_ops_mutex. So it's not a critical difference, just that a lock must be held a bit longer. -- Hallvard

1 0

Re: (ITS#6104) race condition with cancel operation
by hyc＠symas.com 22 May '09

22 May '09

h.b.furuseth(a)usit.uio.no wrote: > Well, this turned out a lot of problems. Currently blocked by > ITS#6138 Bad Cancel/Abandon/"internal abandon"/Syncprov interactions. > > back-relay/op.c done (ITS#6133). syncprov/retcode issues may fit better > under ITS#6138. > > Bitflags should be unnecessary as such, just need a set of values. > Though it wouldn't hurt if they were chosen so one could use bit > operations to reduce checks for "o_abandon == A or B". "Suppress > response" value mentioned in ITS#6138 also needed, maybe that must be a > bitflag when combined with Cancel stuff. Yes, it still seems some bitflags will be more appropriate. > Some tidbits uglifying a fix: > * op->o_conn->c_mutex is locked when be->be_abandon() is called, > but not when be->be_cancel() is called. > * send_ldap_ber() needs the lock while it holds&conn->c_write1_mutex, > but code elsewhere indicates the reverse lock order. > > Question: > syncprov_op_abandon() holds&si->si_ops_mutex when setting > so->s_op->o_abandon. Does it need that? The function needs to > grab op->o_conn->c_mutex when called as Cancel, but must not do > that while holding&si->si_ops_mutex since Abandon grabs the > locks in the opposite order. syncprov_op_abandon() can remove ops from the list, so yes, it must hold si_ops_mutex. -- -- Howard Chu CTO, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

Re: (ITS#6104) race condition with cancel operation
by h.b.furuseth＠usit.uio.no 22 May '09

22 May '09

Well, this turned out a lot of problems. Currently blocked by ITS#6138 Bad Cancel/Abandon/"internal abandon"/Syncprov interactions. back-relay/op.c done (ITS#6133). syncprov/retcode issues may fit better under ITS#6138. Bitflags should be unnecessary as such, just need a set of values. Though it wouldn't hurt if they were chosen so one could use bit operations to reduce checks for "o_abandon == A or B". "Suppress response" value mentioned in ITS#6138 also needed, maybe that must be a bitflag when combined with Cancel stuff. Some tidbits uglifying a fix: * op->o_conn->c_mutex is locked when be->be_abandon() is called, but not when be->be_cancel() is called. * send_ldap_ber() needs the lock while it holds &conn->c_write1_mutex, but code elsewhere indicates the reverse lock order. Question: syncprov_op_abandon() holds &si->si_ops_mutex when setting so->s_op->o_abandon. Does it need that? The function needs to grab op->o_conn->c_mutex when called as Cancel, but must not do that while holding &si->si_ops_mutex since Abandon grabs the locks in the opposite order. -- Hallvard

1 0

(ITS#6138) Bad Cancel/Abandon/"internal abandon"/Syncprov interactions
by h.b.furuseth＠usit.uio.no 22 May '09

22 May '09

Full_Name: Hallvard B Furuseth Version: HEAD OS: URL: Submission from: (NULL) (129.240.6.233) Submitted by: hallvard Might make this a tracking issue for Cancel/Abandon problems. And/or copy some to separate ITSes, but they all seem interconnected: Cancel(abandoned operation)-requests are not rejected. Thus slapd sets o_cancel and turns an active Abandon into a Cancel. Presumably that can confuse Cancel/Abandon handlers, like that in Syncprov. Similarly, Abandon(abandoned/cancelled operation) is not ignored. connection_abandon() re-abandons abandoned/cancelled operations too. However Syncprov:RefreshAndPersist abuses op->o_abandon: It sets it to mean "Suppress the response. A copy of this operation will send it." So if Cancel(op with o_abandon!=0) is fixed to respond protocolError "already abandoned", presumably that breaks Cancel(RefreshAndPersist). I'm not touching it with a flagpole - I don't know syncprov. Help? Overlay retcode does something similar - sends a response and then sets o_abandon. Cancel/Abandon can in any case fail by targeting the wrong operation though: A connection can have multiple messages with the same IDs when the response is sent and the client reuses the message ID, before the old operation in slapd can clean up and finish. syncprov_op_abandon() identifies messages by (connid, msgid). Can multiple messages with the same ID break that, or more importantly, break what gets sent/written with syncrepl? Anyway, the current code needs an o_abandon value which means "suppress response". Or maybe "abandoned, except as far as future Abandon and Cancel operations are concerned". Syncprov and Retcode need to handle the various possible orderings of Cancel/Abandon vs Suppress, including when they forward o_abandon from one operation to another. I haven't looked too closely at that code either. ITS#6104 (race condition with cancel operation) also calls for multiple o_abandon values (joining some or all of o_cancel into o_abandon), and must be considered when picking a value. However I suggest to consider just the various possible values first and get that right, and deal with the race condition later. Source and binary compatibility with third-party modules complicates this, if we're trying to support both and old compiled slapd + new compiled module and vice versa. Might try a minimal solution now.

1 0

Re: (ITS#6137) Cancel(pending operation) acts like Abandon
by h.b.furuseth＠usit.uio.no 22 May '09

22 May '09

Confirmed & fixed. Responds "too busy for Cancel, try Abandon instead". If someone wants to, they could improve that. As noted in the commit: /* TODO: We could instead remove the cancelled operation * from c_pending_ops like Abandon does, and send its * response here. Not if it is pending because of a * congested connection though. */ -- Hallvard

1 0

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