https://bugs.openldap.org/show_bug.cgi?id=10320
Issue ID: 10320
Summary: sigsegv in autogroup
Product: OpenLDAP
Version: 2.6.9
Hardware: x86_64
OS: Linux
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: overlays
Assignee: bugs(a)openldap.org
Reporter: sergej+openldap(a)p5n.pp.ru
Target Milestone: ---
slapd crashes in autogroup overlay on group modification
I have few coredumps and can provide more information. This fails on 0x23
address, it may differs but it looks like f->f_un.f_un_complex is not ended
with NULL sometimes.
Modified group here is not autogroup, just groupOfUniqueNames.
Operation is adding uniqueMember.
Distro: Archlinux
Overlay config:
overlay autogroup
autogroup-attrset labeledURIObject labeledURI uniqueMember
autogroup-memberof-ad memberOf
Stack:
#0 0x00007e77bf511d7c in autogroup_memberOf_filter
(f=f@entry=0x6f2c6d6165742d70, dn=dn@entry=0x7e765c1659f8,
memberof_ad=memberof_ad@entry=0x5ac9490c2190) at autogroup.c:1532
#1 0x00007e77bf511dd1 in autogroup_memberOf_filter (f=0x6f2c6d6165742d70,
f@entry=0x5ac9495089f0, dn=dn@entry=0x7e765c1659f8,
memberof_ad=memberof_ad@entry=0x5ac9490c2190)
at autogroup.c:1537
#2 0x00007e77bf511dd1 in autogroup_memberOf_filter (f=0x5ac9495089f0,
dn=dn@entry=0x7e765c1659f8, memberof_ad=0x5ac9490c2190) at autogroup.c:1537
#3 0x00007e77bf512538 in autogroup_modify_entry (op=<optimized out>,
rs=0x7e7665cf9910) at autogroup.c:1606
#4 0x00005ac946faf432 in overlay_op_walk ()
#5 0x00005ac946faf5f2 in ?? ()
#6 0x00005ac946f494cf in fe_op_modify ()
#7 0x00005ac946f4b623 in do_modify ()
#8 0x00005ac946f304d7 in ?? ()
#9 0x00005ac946f30f4b in ?? ()
#10 0x00007e77c04756e1 in ldap_int_thread_pool_wrapper (xpool=0x5ac949016bc0)
at tpool.c:1059
#11 0x00007e77bfe4b70a in start_thread (arg=<optimized out>) at
pthread_create.c:448
#12 0x00007e77bfecfaac in __GI___clone3 () at
../sysdeps/unix/sysv/linux/x86_64/clone3.S:78
(gdb) p *f
Cannot access memory at address 0x23
(gdb) up
#1 0x000070fe705f3dd1 in autogroup_memberOf_filter (f=0x23,
f@entry=0x55d1c45b6310, dn=dn@entry=0x70fd3000c428,
memberof_ad=memberof_ad@entry=0x55d1c416a300) at autogroup.c:1537
1537 result = result ||
autogroup_memberOf_filter( f, dn, memberof_ad );
(gdb) up
#2 0x000070fe705f3dd1 in autogroup_memberOf_filter (f=0x55d1c45b6310,
f@entry=0x55d1c45b6670, dn=dn@entry=0x70fd3000c428,
memberof_ad=memberof_ad@entry=0x55d1c416a300)
at autogroup.c:1537
1537 result = result ||
autogroup_memberOf_filter( f, dn, memberof_ad );
(gdb) p *f->f_un.f_un_complex
$5 = {f_choice = 124232868587560, f_un = {f_un_result = 939550768, f_un_desc =
0x70fd38006830, f_un_ava = 0x70fd38006830, f_un_ssa = 0x70fd38006830, f_un_mra
= 0x70fd38006830,
f_un_complex = 0x70fd38006830}, f_next = 0x23}
f_next is 0x23 which is bad address
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10328
Issue ID: 10328
Summary: Possible multiple free() calls in
rewrite_subst_compile()
Product: OpenLDAP
Version: unspecified
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: libraries
Assignee: bugs(a)openldap.org
Reporter: gyf161023(a)gmail.com
Target Milestone: ---
In rewrite_subst_compile() of libraries/librewrite/subst.c line 216 and 222,
`subs` and `submatch` array are freed repeatedly on the index `nsub` instead of
l, is this what was intended or the `nsub` has some guarantee to be 1 so that
we are not freeing anything for second time?
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10323
Issue ID: 10323
Summary: Starttls critical not working on lloadd
Product: OpenLDAP
Version: 2.6.9
Hardware: x86_64
OS: Linux
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: lloadd
Assignee: bugs(a)openldap.org
Reporter: grichier(a)scaleway.com
Target Milestone: ---
Hello,
Looks like starttls critical not working on lloadd.
I have a backend with starttls configure but with bad CN.
When I direct query the backend using ldapsearch with option -ZZ, I have the
following error:
ldap_start_tls: Connect error (-11)
additional info: (unknown error code)
But when I query the lloadd, which use same backend with olcBkLloadStartTLS to
critical. It's work...
On a tcpdump I can see the communication between backend and lloadd is not
using starttls. (cleartext). But it shouldn't (critical option)
cn: {1}ldap://ldap01.example.com
olcBkLloadBackendUri: ldap://ldap01.example.com
olcBkLloadNumconns: 10
olcBkLloadBindconns: 5
olcBkLloadRetry: 5000
olcBkLloadMaxPendingOps: 50
olcBkLloadMaxPendingConns: 10
olcBkLloadWeight: 1
olcBkLloadStartTLS: critical
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10226
Issue ID: 10226
Summary: [PATCH] ldap.conf: some remarks and editorial fixes
for this man page
Product: OpenLDAP
Version: unspecified
Hardware: x86_64
OS: Linux
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: documentation
Assignee: bugs(a)openldap.org
Reporter: bjarniig(a)vortex.is
Target Milestone: ---
Created attachment 1020
--> https://bugs.openldap.org/attachment.cgi?id=1020&action=edit
Remarks and editotiral fixes for the manual ldap.conf.5
Version 2.6.8 is not named on the webpage when entering an issue.
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10327
Issue ID: 10327
Summary: Adding syncprov to cn=config can lock up server
Product: OpenLDAP
Version: unspecified
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: backends
Assignee: bugs(a)openldap.org
Reporter: ondra(a)mistotebe.net
Target Milestone: ---
As reported on openldap-technical[0], adding syncprov to a cn=config DB can
lock up the server. This happens when syncprov_db_otask is scheduled by
syncprov_db_open.
ITS#9045 has already dealt with this for config_back_entry_get, we can relax
this description and allow a config_back_search do the same. Even though it is
technically on a different thread than the main modification, AFAIK it is still
the only one running as the server is paused and the main modification is
paused waiting for it?
[0].
https://lists.openldap.org/hyperkitty/list/openldap-technical@openldap.org/…
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10331
Issue ID: 10331
Summary: slap* tools should be a little more helpful as to
which command line option was invalid
Product: OpenLDAP
Version: unspecified
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: slapd
Assignee: bugs(a)openldap.org
Reporter: ondra(a)mistotebe.net
Target Milestone: ---
Also why where appropriate.
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10270
Issue ID: 10270
Summary: Issues with pcache when refresh/persistPcache used
Product: OpenLDAP
Version: 2.5.18
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: overlays
Assignee: bugs(a)openldap.org
Reporter: aweits(a)rit.edu
Target Milestone: ---
Greetings, OpenLDAP-folk.
We've been running with the pcache overlay in 2.5.18
with both query refresh and pcachePersist for a bit
and have observed some oddities:
1.) Negative queries don't get refreshed
2.) Queries don't seem to be persisted
These behaviors are all exhibited from the current
git version as well - code/patches below for clarity
of communication:
Thanks!
Andy
commit c0b4fe92c8df746c0e6a777f93f1687135114eb9
Author: Andrew Elble <aweits(a)rit.edu>
Date: Fri Oct 11 08:43:47 2024 -0400
negative cache entries are not loaded when pcachePersist is on
diff --git a/servers/slapd/overlays/pcache.c b/servers/slapd/overlays/pcache.c
index 9ef78fd6bf43..9fd72e6d7261 100644
--- a/servers/slapd/overlays/pcache.c
+++ b/servers/slapd/overlays/pcache.c
@@ -802,7 +802,11 @@ url2query(
goto error;
}
- cq = add_query( op, qm, &query, qt, PC_POSITIVE, 0 );
+ if (BER_BVISNULL( &uuid )) {
+ cq = add_query( op, qm, &query, qt, PC_NEGATIVE, 0 );
+ } else {
+ cq = add_query( op, qm, &query, qt, PC_POSITIVE, 0 );
+ }
if ( cq != NULL ) {
cq->expiry_time = expiry_time;
cq->refresh_time = refresh_time;
commit 8f7b50dfcec69fa01f8cf0a4b77f3dee8ef9f0f6
Author: Andrew Elble <aweits(a)rit.edu>
Date: Fri Oct 11 08:38:36 2024 -0400
queries with ttr/x-refresh are not loaded when pcachePersist is on
diff --git a/servers/slapd/overlays/pcache.c b/servers/slapd/overlays/pcache.c
index 40c1f9673776..9ef78fd6bf43 100644
--- a/servers/slapd/overlays/pcache.c
+++ b/servers/slapd/overlays/pcache.c
@@ -749,7 +749,7 @@ url2query(
}
}
- if ( got != GOT_ALL ) {
+ if ( (got & GOT_ALL) != GOT_ALL) {
rc = 1;
goto error;
}
commit c7e52c90192a43876d40b9776a58db951d27937c
Author: Andrew Elble <aweits(a)rit.edu>
Date: Fri Oct 11 08:37:13 2024 -0400
ttr was not being applied to negatively cached entries
diff --git a/servers/slapd/overlays/pcache.c b/servers/slapd/overlays/pcache.c
index 1d6e4ba4edcf..40c1f9673776 100644
--- a/servers/slapd/overlays/pcache.c
+++ b/servers/slapd/overlays/pcache.c
@@ -1580,6 +1580,8 @@ add_query(
case PC_NEGATIVE:
ttl = templ->negttl;
+ if ( templ->ttr )
+ ttr = now + templ->ttr;
break;
case PC_SIZELIMIT:
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=9934
Issue ID: 9934
Summary: slapd-config(5) should document how to store
certificates for slapd usage
Product: OpenLDAP
Version: 2.5.13
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: documentation
Assignee: bugs(a)openldap.org
Reporter: quanah(a)openldap.org
Target Milestone: ---
Commit 7b41feed83b expanded the ability of cn=config to save the certificates
used for TLS by slapd directly in the config database. However the
documentation for the new parameters was never added to the slapd-config(5) man
page.
olcTLSCACertificate $ olcTLSCertificate $ olcTLSCertificateKey
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10191
Issue ID: 10191
Summary: backend searches should respond to pause requests
Product: OpenLDAP
Version: 2.6.7
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: backends
Assignee: bugs(a)openldap.org
Reporter: hyc(a)openldap.org
Target Milestone: ---
A long running search will cause mods to cn=config to wait a long time. Search
ops should periodically check for threadpool pause requests.
--
You are receiving this mail because:
You are on the CC list for the issue.
https://bugs.openldap.org/show_bug.cgi?id=10337
Issue ID: 10337
Summary: Global TLS options not inherited in context
Product: OpenLDAP
Version: 2.6.9
Hardware: All
OS: All
Status: UNCONFIRMED
Keywords: needs_review
Severity: normal
Priority: ---
Component: libraries
Assignee: bugs(a)openldap.org
Reporter: remi(a)fedoraproject.org
Target Milestone: ---
During ldap_create, global options are copied
See
https://github.com/openldap/openldap/blob/OPENLDAP_REL_ENG_2_6_9/libraries/…
/* copy the global options */
AC_MEMCPY(&ld->ld_options, gopts, sizeof(ld->ld_options));
But not the TLS string options
See
https://github.com/openldap/openldap/blob/OPENLDAP_REL_ENG_2_6_9/libraries/…
/* We explicitly inherit the SSL_CTX, don't need the names/paths. Leave
* them empty to allow new SSL_CTX's to be created from scratch.
*/
memset( &ld->ld_options.ldo_tls_info, 0,
sizeof( ld->ld_options.ldo_tls_info ));
ld->ld_options.ldo_tls_ctx = NULL;
Which create inconsistency when trying to generate a newctx
For the context
See
* https://github.com/php/php-src/issues/17776 => why tls_newctx is needed
* https://github.com/php/php-src/issues/18529 => regression
On PHP side a possible workaround is to do the copyu manually
* https://github.com/php/php-src/pull/18547
But should probably be handled at openldap library
--
You are receiving this mail because:
You are on the CC list for the issue.