openldap-bugs February 2023

openldap-bugs@openldap.org

2 participants
146 discussions

[Issue 9897] New: ldapcompare return FALSE when dynlist in use
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9897 Issue ID: 9897 Summary: ldapcompare return FALSE when dynlist in use Product: OpenLDAP Version: 2.6.3 Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: overlays Assignee: bugs(a)openldap.org Reporter: sdalu(a)sdalu.com Target Milestone: --- The server configuration enable dynlist with: overlay dynlist dynlist-attrset groupOfURLs memberURL member Entry is: dn: cn=Admins,ou=Services,ou=Members,dc=zog,dc=com member: uid=foo,ou=People,dc=zog,dc=com cn: Admins objectClass: groupOfNames The ldap compare will return false, was expected true ldapcompare -x cn=Admins,ou=Services,ou=Members,dc=zog,dc=com member:uid=foo,ou=People,dc=zog,dc=com Now if I comment the dynlist-attrset directive, I will get TRUE for the same ldapcompare request -- You are receiving this mail because: You are on the CC list for the issue.

1 6

[Issue 9984] New: Balancer listener thread exits when all listeners are suspended
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9984 Issue ID: 9984 Summary: Balancer listener thread exits when all listeners are suspended Product: OpenLDAP Version: 2.5.13 Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: lloadd Assignee: bugs(a)openldap.org Reporter: ondra(a)mistotebe.net Target Milestone: --- If lloadd runs out of file descriptors, the listeners are paused temporarily, however the event base is not configured to stick around if all of its tasks have finished. In case all of the listeners are paused, the event base exits and listener thread goes away. A patch is coming. -- You are receiving this mail because: You are on the CC list for the issue.

1 4

[Issue 9983] New: operation_unlink files the operation before it is fully unlinked
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9983 Issue ID: 9983 Summary: operation_unlink files the operation before it is fully unlinked Product: OpenLDAP Version: 2.5.13 Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: lloadd Assignee: bugs(a)openldap.org Reporter: ondra(a)mistotebe.net Target Milestone: --- In epoch based memory management, objects should only be submitted for reclaiming when no actors can reach them unless they've done that before the submission happened. This is broken in operation_unlink(): It calls try_release_ref() at the beginning where the operation is added to the to-reclaim list, only then it proceeds to unlink it from other objects. The following sequence is then possible: - current_epoch == 1 (no threads are alive in epoch == 0) - in thread 1 (epoch = 1), try_release_ref() marks the object to be reclaimed in current_epoch - thread 2 activates and current_epoch is incremented (current_epoch == 2) - thread 2 handles an Unbind for the operation's client and reaches client_reset() (epoch == 2) - thread 2 (client_reset) snapshots and clears client->c_ops (among other things, c->c_ops links to our object) - thread 1 finishes operation_unlink, deactivates and there are no more threads in epoch == 1 - thread 3 activates and current_epoch is incremented (current_epoch == 3), there are objects in epoch == 1, namely the object above which is now destroyed (freed) - thread 2 wakes up again and tries to call operation_abandon on the above object, this accesses memory freed epoch_append (and try_release_ref) should only be called when unlinking has finished. I'm testing a patch right now. -- You are receiving this mail because: You are on the CC list for the issue.

1 4

[Issue 9947] New: Race in epoch.c
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9947 Issue ID: 9947 Summary: Race in epoch.c Product: OpenLDAP Version: unspecified Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: lloadd Assignee: bugs(a)openldap.org Reporter: ondra(a)mistotebe.net Target Milestone: --- When epoch_leave() tests whether other threads might still be alive, it can test things in reverse order, testing too early to catch a thread to come in and too late to see a thread that just left. If those two saw each other, the clock would not advance and the data in refs might actually still be a reachable pointer. The tests in epoch_leave can actually be simplified leading to machine code not all compilers could figure out by themselves. -- You are receiving this mail because: You are on the CC list for the issue.

1 5

[Issue 9925] New: Fix some compilation issues around usage of #if and #ifdef
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9925 Issue ID: 9925 Summary: Fix some compilation issues around usage of #if and #ifdef Product: OpenLDAP Version: unspecified Hardware: x86_64 OS: Linux Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: libraries Assignee: bugs(a)openldap.org Reporter: connor.smith(a)hitachivantara.com Target Milestone: --- Created attachment 918 --> https://bugs.openldap.org/attachment.cgi?id=918&action=edit git format-patch I noticed a few issues while working with OpenLDAP that would lead to compiler warnings and the like in our particular build environment. They're quite minor, but it would still be good to have them patched upstream. In include/ac/socket.h: Use #elif defined(...) for HAVE_WINSOCK and MACOS. All other instances of these macros use #ifdef or similar. A compiler may warn about them not being defined. In libraries/liblber/sockbuf.c, (DOS && PCNFS) and (DOS && NCSA) were replaced with HAVE_PCNFS and HAVE_NCSA, respectively. It seems logical to do the same at the only remaining occurrence of DOS, PCNFS, and NCSA. For context on the latter: the actual warning was about #elif DOS, similar to #elif HAVE_WINSOCK and #elif MACOS, but looking into it it seemed to make sense to bring socket.h in line with sockbuf.c. In libraries/liblunicode/ucdata/ucgendat.c: Use #if HARDCODE_DATA consistently, replacing two instances of #ifdef. HARDCODE_DATA is always defined, and this way you can set HARDCODE_DATA to 0 and have it work, rather than it going down the wrong branch and failing in these two cases. An IPR notice, with this work having been done as part of my employment: The attached file is derived from OpenLDAP Software. All of the modifications to OpenLDAP Software represented in the following patch were developed by Hitachi Vantara. Hitachi Vantara has not assigned rights and/or interest in this work to any party. I, Connor Smith, am authorized by Hitachi Vantara, my employer, to release this work under the following terms. Hitachi Vantara hereby place the following modifications to OpenLDAP Software (and only these modifications) into the public domain. Hence, these modifications may be freely used and/or redistributed for any purpose with or without attribution and/or other notice. Please let me know if there are any issues with the attached patch, or if there's anything else I need to do. Thanks. -- You are receiving this mail because: You are on the CC list for the issue.

1 5

[Issue 9904] New: A Potential NPD
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9904 Issue ID: 9904 Summary: A Potential NPD Product: OpenLDAP Version: unspecified Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: libraries Assignee: bugs(a)openldap.org Reporter: 1157401338(a)qq.com Target Milestone: --- Created attachment 911 --> https://bugs.openldap.org/attachment.cgi?id=911&action=edit diagram of NPD Hi, I found a NPD bug in the project source code of ldap, and I have shown the execution sequence of the program that may have generated the bug on a diagram,which is added to the attachment The red text illustrates the steps that created the bug the red arrows represent the call relationships the file path can be seen in the blue framed section. additionally,at step 4 I do not expand more detail about why function ber_memalloc_x can return null(actually it can be seen as function malloc and the reason ber_memalloc_x return null is same with malloc),because there are many code snippet can be found in project source code that judge whether ber_memalloc_x return null and make further process if return value equal to null. I look forward to your reply and thank you very much for your patience! -- You are receiving this mail because: You are on the CC list for the issue.

1 7

[Issue 9901] New: Fix non-standard printf arguments in liblbert and libldap
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9901 Issue ID: 9901 Summary: Fix non-standard printf arguments in liblbert and libldap Product: OpenLDAP Version: 2.6.3 Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: libraries Assignee: bugs(a)openldap.org Reporter: michael.osipov(a)siemens.com Target Milestone: --- Created attachment 910 --> https://bugs.openldap.org/attachment.cgi?id=910&action=edit Patch gainst source tarball As a followup to Bug 9898 and Bug 9899 I have played around with LLVM 13 and "-std=c17 -pedantic -Wall" it fails to compile several files. Find a patch attached which makes it standards compliant. -- You are receiving this mail because: You are on the CC list for the issue.

1 6

[Issue 9899] New: "cyrus.c" uses non-portable GNU extension for void pointer arithmetics and fails on HP-UX aCC
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9899 Issue ID: 9899 Summary: "cyrus.c" uses non-portable GNU extension for void pointer arithmetics and fails on HP-UX aCC Product: OpenLDAP Version: 2.6.3 Hardware: All OS: Other Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: libraries Assignee: bugs(a)openldap.org Reporter: michael.osipov(a)siemens.com Target Milestone: --- On HP-UX with cc: HP C/aC++ B3910B A.06.29 [Oct 18 2016] tells me libtool: compile: /opt/aCC/bin/aCC -Ae -g -I../../include -I../../include -I/opt/ports/include -DLDAP_LIBRARY -c cyrus.c -DPIC -o .libs/cyrus.o "cyrus.c", line 420: error #3143: arithmetic on pointer to void or function type memcpy( cb_data + plen, cbv.bv_val, cbv.bv_len ); ^ 1 error detected in the compilation of "cyrus.c". gmake[2]: *** [Makefile:434: cyrus.lo] Error 1 void pointer arithmetics is not valid/undefined and just a GNU extension supported by GCC or clang. I was able to reproduce this on FreeBSD clang version 13.0.0 (git@github.com:llvm/llvm-project.git llvmorg-13.0.0-0-gd7b669b3a303): osipovmi@deblndw011x:~/var/Projekte/openldap-2.6.3/libraries/libldap $ cc -std=c17 -I../../include -I../../include -I/usr/local/include -DLDAP_LIBRARY -c cyrus.c -o cyrus.o -pedantic -Werror cyrus.c:420:18: error: arithmetic on a pointer to void is a GNU extension [-Werror,-Wpointer-arith] memcpy( cb_data + plen, cbv.bv_val, cbv.bv_len ); ~~~~~~~ ^ 1 error generated. I am not a daily C hacker, but I guess cb_data needs to be typed to "unsigned char" just like data from sasl_channel_binding_t (https://github.com/cyrusimap/cyrus-sasl/blob/cb549ef71c5bb646fe583697ebdcab…). Or at least a malloc with an "unsigned char", save the pointer start address, copy the prefix, increment by prefix length, copy the channel binding value and then assign the pointer start address to the output struct. I will unset SASL_CHANNEL_BINDING for now since it is not required in your AD environment when SASL GSSAPI with minssf=1 is set. -- You are receiving this mail because: You are on the CC list for the issue.

1 7

[Issue 9976] New: ldap_sasl_bind_s man page missing details on freeing the returned servercredp parameter
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9976 Issue ID: 9976 Summary: ldap_sasl_bind_s man page missing details on freeing the returned servercredp parameter Product: OpenLDAP Version: unspecified Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: documentation Assignee: bugs(a)openldap.org Reporter: wmjchs(a)hotmail.com Target Milestone: --- Source code for ldap_sasl_bind_s shows that the parameter is to be freed via ber_bvfree when no longer used. (https://git.openldap.org/openldap/openldap/-/blob/master/libraries/libldap/…) However, the documentation (https://www.openldap.org/software/man.cgi?query=ldap_sasl_bind_s&apropos=0&…) makes no mention of this. The following sources also omitted this information: https://linux.die.net/man/3/ldap_sasl_bind_s https://www.ibm.com/docs/en/i/7.4?topic=ssw_ibm_i_74/apis/ldap_sasl_bind_s.… https://learn.microsoft.com/en-us/windows/win32/api/winldap/nf-winldap-ldap… While the following didn't: https://www.novell.com/documentation/developer/cldap/ldaplibc/data/a401rtp.… https://www.ibm.com/docs/en/zos/2.3.0?topic=routines-ldap-sasl-bind-ldap-sa… https://docs.oracle.com/cd/E19957-01/817-6707/functions.html#wp57022 -- You are receiving this mail because: You are on the CC list for the issue.

1 5

[Issue 9957] New: slapo-dynlist manpage needs better description of dynlist-attrset
by openldap-its＠openldap.org 09 Feb '23

09 Feb '23

https://bugs.openldap.org/show_bug.cgi?id=9957 Issue ID: 9957 Summary: slapo-dynlist manpage needs better description of dynlist-attrset Product: OpenLDAP Version: unspecified Hardware: All OS: All Status: UNCONFIRMED Keywords: needs_review Severity: normal Priority: --- Component: overlays Assignee: bugs(a)openldap.org Reporter: ondra(a)mistotebe.net Target Milestone: --- Each dynlist-attrset defines one of three distinct behaviours: - dynamic list (attributes are gathered from other entries) - dynamic group (DNs are gathered based on other entries) - static group (DNs are gathered based on DNs stored on entries) With the groups possibly being recursive, requiring traversal. Since the above do not mix, the documentation should be more explicit about how each one should look and behave. It should also be noted somewhere what happens (or not) when multiple dynlist-attrset stanzas would apply to the same entry. At that point, configuration code could also be made more strict to reject configurations that satisfy the apparent dynlist-attrset syntax but do not actually represent anything that fits just one of the above and is therefore nonsensical (with parts of it apparently ignored at runtime). With nonsensical configuration rejected, it would be possible to streamline internal dynlist structures and make critical parts of the overlay code more readable. -- You are receiving this mail because: You are on the CC list for the issue.

1 5

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs February 2023