[PATCH] Remove duplicates detection in SRV resolving
by Jakub Hrozek
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
The first patch fixes https://bugzilla.redhat.com/show_bug.cgi?id=694444
by not detecting duplicates at all.
The second patch is an optimization. Since it is possible that the list
of servers contains duplicates now, the code marks all duplicates when
setting a server status, too. I'm submitting it as a separate patch
because it might be easier to push only one of them if QE thinks they
have the capacity to test only one of them for the stable 1.5 branch.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2i35sACgkQHsardTLnvCVeyACfe5+vWaIbLI3mDy2RzrJJ4sTc
LHMAnjlZiH8I9615bUy1rFYwhoEqv+ST
=LgxT
-----END PGP SIGNATURE-----
13 years
[PATCH] Tap a huge memory leak
by Simo Sorce
The operations we use to deleted members in the memberof plugin are
quite inefficient, but on top of that we were also "leaking" a lot of
search results on the operations contexts, causing the runtime memory
footprint to grow immensely (as in gigabytes for a singe ldb_modify).
It was technically not a leak beacuase eventually all memory would be
reclaimed, but it was enough to make the OOM killer kill our processes.
After a deletion operation is finished, make sure to explicitly free
all the search results we do not need anymore, so that the footprint
remains low.
Simo.
--
Simo Sorce * Red Hat, Inc * New York
13 years
[PATCH] Fix list comparison in memberof module
by Simo Sorce
When we replace the member attribute on an entry we calculate the list
of total addition vs deletions and then make sure a dn does not appear
in both at the same time.
The code was wrong and was skipping the next comparison each time a
match was found due to an array counter error.
This was causing unnecessary work to add entries that were already
present.
Simo.
--
Simo Sorce * Red Hat, Inc * New York
13 years
[PATCH] tidyups
by Simo Sorce
We use defines for a lot of stuff when it comes to sysdb, to make it
simple to change things later.
One place where this was not happening was in referring to the sysdb
base dn. This patch simply replaces using the basedn string directly
with a macro.
Simo.
--
Simo Sorce * Red Hat, Inc * New York
13 years
[PATCH] Never remove gecos from the sysdb cache
by Stephen Gallagher
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Now that gecos can come from either the 'gecos' or 'cn' attributes,
we need to ensure that we never remove it from the cache.
We were being too greedy with our removal code. It purges from the sysdb
cache any attributes that were requested from LDAP but that we did not
get. However, in the case of 'gecos' (if we are falling back to 'cn')
this would mean that we were always purging the gecos-from-cn value.
- --
Stephen Gallagher
RHCE 804006346421761
Delivering value year after year.
Red Hat ranks #1 in value among software vendors.
http://www.redhat.com/promo/vendor/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2kg94ACgkQeiVVYja6o6PO2QCdE8rhsVUb8rYyjOPZPVm0O42h
1b0An0uF5jQLOXaxnkqea6pO2PhYAGWX
=UK4v
-----END PGP SIGNATURE-----
13 years
[PATCH] Only check systemd unit dir if systemd is selected
by Jakub Hrozek
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Configure was failing on systems without systemd-units installed because
pkg-config would report an empty string for systemd unit directory.
We should not run any detection of systemd unit dir if systemd is not
selected as the init script type.
This probably slipped through because even F14 has systemd-units package
and the corresponding .pc file.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2ke0QACgkQHsardTLnvCUOiACg2RzAiU6HttL09MaDuYXVEoCW
re0AoIs2uwwqSVJEFdSJUs68KntCHbwo
=wGoa
-----END PGP SIGNATURE-----
13 years
Announcing the release of SSSD 1.5.5
by Stephen Gallagher
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
The SSSD team is proud to announce the release of the version 1.5.5
bugfix release of the System Security Services Daemon.
As always, it can be downloaded at https://fedorahosted.org/sssd
== Highlights ==
* Fixes for several crash bugs
* LDAP group lookups will no longer abort if there is a zero-length
member attribute
* Add automatic fallback to 'cn' if the 'gecos' attribute does not exist
== Detailed Changelog ==
Jakub Hrozek (5):
* Fix typo in sdap_nested_group_process_step
* Mark transaction as done when cancelled
* Do not attempt to resolve nameless servers
* Remove detection of duplicates from SRV result processing
* Use safe alignment macros for in-tree SRV record parsing
Stephen Gallagher (5):
* Update version to 1.5.5
* Always complete the transaction in sdap_process_group_members_2307
* RFC2307: Ignore zero-length member names in group lookups
* Fall back to cn if gecos is not available
* Updating translation files
Sumit Bose (3):
* Read only rootDSE data if rootDSE is available
* Initialise srv_opts even if rootDSE is missing
* Initialise rootdse to NULL if not available
- --
Stephen Gallagher
RHCE 804006346421761
Delivering value year after year.
Red Hat ranks #1 in value among software vendors.
http://www.redhat.com/promo/vendor/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2kagoACgkQeiVVYja6o6NPbACfR97t0kZAc9eWqu2aaT19sqab
uxwAn0wmBjiWf5jOQFRyfjvO0VwzFilN
=Ujn0
-----END PGP SIGNATURE-----
13 years
[PATCH] Initialise rootdse to NULL if not available
by Sumit Bose
Hi,
I'm sorry but this is chapter 3 of a hopefully soon-to-end story. I
wrote the previous rootDSE patch under the false assumption that
rootdse==NULL if not available (and at least in my tests it always was).
This new patch corrects this.
bye,
Sumit
13 years
[PATCH] Provide a configuration option to use systemd unit file
by Jakub Hrozek
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
[PATCH 1/3] The systemd unit file should not require DBus
This patch fixes the provided systemd unit file so it is the same
as the one Jóhann B. Guðmundsson provided in Red Hat Bugzilla #689853
except for hardcoded paths.
[PATCH 2/3] Provide a configuration option to use systemd unit file
https://fedorahosted.org/sssd/ticket/837
[PATCH 3/3] NOSUBMIT: systemd support in sssd rpm
I don't think we should push this patch, but I think it would ease
review. After applying this patch, sysv is still the default. To build
sssd packages with systemd unit file, change the bcond_with to
bcond_without (yes, bcond_without systemd means build with systemd as it
makes rpmbuild accept a --without-systemd option):
- -%bcond_with systemd
+%bcond_without systemd
Jakub
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2cjksACgkQHsardTLnvCVZcgCdFKFOs610cWDn5m2LUhCLL6gx
TVAAn0IkK1bm7/TUJ5Jdjd+mjhP+WTJ7
=VR5i
-----END PGP SIGNATURE-----
13 years
[PATCH] Use safe alignment macros for in-tree SRV record parsing
by Jakub Hrozek
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
The in-tree SRV record parsing is used with very old c-ares libraries
that don't implement the parsing themselves (c-ares < 1.7, used in e.g.
RHEL5)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2kCbwACgkQHsardTLnvCUFigCfTljsGOGwbvPSE3ty6yzdIAVl
GYUAoNV9GbVVOd2LNLzD59rHBvrKFDPd
=rn4B
-----END PGP SIGNATURE-----
13 years