389-devel September 2015

389-devel@lists.fedoraproject.org

10 participants
40 discussions

Commit: Regression in your 389 makefile patches
by Noriko Hosoi 23 Sep '15

23 Sep '15

Thanks to Simo for fixing the 389-ds-base Makefile.am. Pushed his patch to the master branch. > commit 3c827508f8e177992010011305f78da3b3ab3b25 > Author: Simo Sorce <simo(a)redhat.com> > Date: Wed Sep 23 15:08:17 2015 -0400 > > Fix out of tree build > > Signed-off-by: Simo Sorce <simo(a)redhat.com> --noriko On 09/23/2015 12:16 PM, Simo Sorce wrote: > On 23/09/15 13:37, Simo Sorce wrote: >> I know how to fix this, I repored and I am in the process. >> It is mostly adding some $(srcdir)/ in front of some of the paths > > Can you try this patch. > This changes only Makefile.am so you'll need to run autoreconf -fi > before a configure. > > > Noriko, > you'll have to commit also the other files for a full patch to master.

1 0

Please review: Ticket 48283: many attrlist_replace errors in connection with cleanallruv
by Ludwig Krispenz 23 Sep '15

23 Sep '15

https://fedorahosted.org/389/ticket/48283 https://fedorahosted.org/389/attachment/ticket/48283/0001-Ticket-48283-many…

1 0

please review: Ticket 48266 - do not free "repl keep alive" entry if an error occurs
by Mark Reynolds 22 Sep '15

22 Sep '15

https://fedorahosted.org/389/ticket/48266 https://fedorahosted.org/389/attachment/ticket/48266/0001-Ticket-48266-do-n…

1 0

please review: Ticket 48284 - free entry during internal add when error occurs during pblock setup
by Mark Reynolds 22 Sep '15

22 Sep '15

https://fedorahosted.org/389/ticket/48284 https://fedorahosted.org/389/attachment/ticket/48284/0001-Ticket-48284-free…

1 0

please review: Ticket 48266 - Fix crash during online init of consumer (keep alive repl entry)
by Mark Reynolds 21 Sep '15

21 Sep '15

https://fedorahosted.org/389/ticket/48266 https://fedorahosted.org/389/attachment/ticket/48266/0001-Ticket-48266-Onli…

1 0

Please review (additional fix): [389 Project] #48188: segfault in ns-slapd due to accessing Slapi_DN freed in pre bind plug-in
by Noriko Hosoi 18 Sep '15

18 Sep '15

https://fedorahosted.org/389/ticket/48188 https://fedorahosted.org/389/attachment/ticket/48188/0001-Ticket-48188-segf… git patch file (master) -- Additional fixes based upon the comments by Rich in comment 24 Description: Additional fixes based upon the comments by rmeggins(a)redhat.com (Thank you, Rich!!). https://fedorahosted.org/389/ticket/48188?replyto=24#comment:24 1. Implemented the case 2) If the plugin changes the SLAPI_BIND_TARGET_SDN *value*, we need to select a different backend. It is possible (but not very useful) for the plugin to change the pointer, but use the same value. 2. Added an api slapi_be_select_exact which returns NULL if no matching backend. On 09/18/2015 11:48 AM, 389 Project wrote: > #48188: segfault in ns-slapd due to accessing Slapi_DN freed in pre bind plug-in > -------------------------------------------------+------------------------- > Reporter: wetpaste | Owner: nhosoi > Type: defect | Status: > Priority: major | accepted > Component: Directory Server | Milestone: 1.3.4.5 > Resolution: | Version: 1.3.3 > Blocked By: | Keywords: > Review: review? | segfault crash > Red Hat Bugzilla: | Blocking: > [https://bugzilla.redhat.com/show_bug.cgi?id=1264224| Ticket origin: > 1264224] | Community > -------------------------------------------------+------------------------- > > Comment (by rmeggins): > > There are two issues > > 1) If the plugin changes the SLAPI_BIND_TARGET_SDN *pointer*, we need to > use it instead of the current `sdn` and `dn`. For this, it is sufficient > to do `pb_sdn != sdn`. > > 2) If the plugin changes the SLAPI_BIND_TARGET_SDN *value*, we need to > select a different backend. It is possible (but not very useful) for the > plugin to change the pointer, but use the same value. > > I'm also not sure if it is sufficient to call `be = > slapi_be_select(sdn);`. The bind code sets the be like this: > {{{ > /* We could be serving multiple database backends. Select the > appropriate one */ > if (slapi_mapping_tree_select(pb, &be, &referral, errorbuf) != > LDAP_SUCCESS) { > send_nobackend_ldap_result( pb ); > be = NULL; > goto free_and_return; > } > }}} > Do we need to have similar logic in order to change the backend? What if > the plugin selects a non-existent backend that causes `be == > defbackend_get_backend()`? > > We also need to set the new backend in the pblock: `slapi_pblock_set( pb, > SLAPI_BACKEND, be );` >

1 0

Please review (take 2): [389 Project] #48188: segfault in ns-slapd due to accessing Slapi_DN freed in pre bind plug-in
by Noriko Hosoi 18 Sep '15

18 Sep '15

https://fedorahosted.org/389/ticket/48188 https://fedorahosted.org/389/attachment/ticket/48188/0001-Ticket-48188-segf… This patch is based upon the patch provided by Simo Sorce <simo(a)redhat.com> for Ticket #48272 - Allow PRE_BIND plugins to mangle DNs Description: Allow a pre_bind plugin to map a DN to another This is useful for plugins that deal with virtual trees or non-standard clients binding with values that are not proper DNs and similar situations. Contributed by: Simo Sorce <simo(a)redhat.com> 2 changes are made to the original patch: 1. removed "slapi_sdn_free(&sdn)" with this comment: * It is a plug-in's responsibility to free the original Slapi_DN. Note: slapi-nis already freed the original sdn. 2. reset dn from the new sdn. dn = slapi_sdn_get_dn(sdn);

1 0

Build failed in Jenkins: 389-ds-base #785
by jenkins＠fedoraproject.org 18 Sep '15

18 Sep '15

See <http://jenkins.fedorainfracloud.org/job/389-ds-base/785/changes> Changes: [thierry bordaz] Ticket 48266: Fractional replication evaluates several times the same ------------------------------------------ Started by an SCM change Building remotely on F22 (Fedora22 fedora Fedora fedora22) in workspace <http://jenkins.fedorainfracloud.org/job/389-ds-base/ws/> Wiping out workspace first. Cloning the remote Git repository Cloning repository git://git.fedorahosted.org/git/389/ds.git > git init <http://jenkins.fedorainfracloud.org/job/389-ds-base/ws/> # timeout=10 Fetching upstream changes from git://git.fedorahosted.org/git/389/ds.git > git --version # timeout=10 > git -c core.askpass=true fetch --tags --progress git://git.fedorahosted.org/git/389/ds.git +refs/heads/*:refs/remotes/origin/* > git config remote.origin.url git://git.fedorahosted.org/git/389/ds.git # timeout=10 > git config --add remote.origin.fetch +refs/heads/*:refs/remotes/origin/* # timeout=10 > git config remote.origin.url git://git.fedorahosted.org/git/389/ds.git # timeout=10 Fetching upstream changes from git://git.fedorahosted.org/git/389/ds.git > git -c core.askpass=true fetch --tags --progress git://git.fedorahosted.org/git/389/ds.git +refs/heads/*:refs/remotes/origin/* > git rev-parse origin/master^{commit} # timeout=10 Checking out Revision 1a58bf85389fc4373af2bf07e96eeb4618cee2b7 (origin/master) > git config core.sparsecheckout # timeout=10 > git checkout -f 1a58bf85389fc4373af2bf07e96eeb4618cee2b7 > git rev-list a35ef5563eb4da913476513a14643d2f8b1c1cf8 # timeout=10 [389-ds-base] $ /bin/sh -e /tmp/hudson9200914399771747370.sh Running configure... CFLAGS= -Wall CXXFLAGS= -Wall ./configure --with-tmpfiles-d=/etc/tmpfiles.d --with-openldap --enable-autobind --with-selinux --with-systemdsystemunitdir=/lib/systemd/system --with-systemdsystemconfdir=/etc/systemd/system --enable-debug Build log is http://jenkins.fedorainfracloud.org/job/389-ds-base/ws/build.785.txt Running make... Build log is http://jenkins.fedorainfracloud.org/job/389-ds-base/ws/build.785.txt Checking for warnings... Build http://jenkins.fedorainfracloud.org/job/389-ds-base/785/ failed There are build warnings Warning log is http://jenkins.fedorainfracloud.org/job/389-ds-base/ws/build-warns.785.txt Last 100 lines of warning log: ------------------------------------------------------------ ldap/servers/plugins/replication/repl5_inc_protocol.c:1675:10: warning: unused variable 'subentry_update_sent' [-Wunused-variable] ------------------------------------------------------------ Build step 'Execute shell' marked build as failure

1 1

please review: Ticket 48217 - fix regression with cleanAllRUV and shutdowns
by Mark Reynolds 18 Sep '15

18 Sep '15

https://fedorahosted.org/389/ticket/48217 https://fedorahosted.org/389/attachment/ticket/48217/0001-Ticket-48217-clea…

1 0

please review: Ticket 48264 - Ticket 47553 tests refactoring
by Simon Pichugin 18 Sep '15

18 Sep '15

Hi team, please review my refactoring patch. There is not much PEP8 refactor, mostly I just refactored it to the pytest structure. Some testcases were united into one, because they contained obvious big repetitions. https://fedorahosted.org/389/ticket/48264 https://fedorahosted.org/389/attachment/ticket/48264/0001-Ticket-48264-Tick…

1 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

389-devel September 2015