summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2012-04-24Fix pam_winbind build against newer iniparser librarySimo Sorce1-4/+4
iniparser_getstr is deprecated and has been removed in newer libraries available in Fedora. Use iniparse_getstring instead. Autobuild-User: Simo Sorce <idra@samba.org> Autobuild-Date: Tue Apr 24 02:56:10 CEST 2012 on sn-devel-104
2012-04-23Avoid warning about KRB5_DEPRECATE with MIT libsSimo Sorce1-0/+4
2012-04-23Cracknames: use krb wrapper functions so it works with MITSimo Sorce1-25/+29
Also avoid a silly game with directly modifying the principal and then calling krb5_principal_unparse_flags to get out a string. If we already assume it is a 2 components name and know what outcome we are going to get, just go ahead and talloc_asprintf the linearized string.
2012-04-23krb5_samba: Add support for krb5_princ_size when using HeimdalSimo Sorce2-0/+9
2012-04-23Make krb5 wrapper library common so they can be used all overSimo Sorce25-522/+470
2012-04-23For now just disable this Heindal specific stuff in the MIT buildSimo Sorce5-3/+29
2012-04-23Make krb5 context initialization not heimdal specificSimo Sorce3-23/+55
Turn the logging data to an opaque pointer. Ifdef code and use MIT logging function when built against system MIT.
2012-04-23Make sure krb5_principal_get_num_comp is identified as present for Heimdal buildAlexander Bokovoy1-0/+1
Common wrappers for MIT / Heimdal use krb5_principal_get_num_comp() to replace krb5_princ_size but rely on krb5_principal_get_num_comp() identified by the build. As we know it exists in Heimdal, define it for waf build. Signed-off-by: Simo Sorce <idra@samba.org>
2012-04-23waf: rename SAMBA4_INTERNAL_HEIMDAL to SAMBA4_USES_HEIMDALAlexander Bokovoy1-1/+1
SAMBA4_INTERNAL_HEIMDAL is defined unconditionally regardless where Heimdal comes from, system-wide or embedded version. This define is not used anywhere. We'll use it to distinguish between Heimdal and MIT Krb5 builds. Signed-off-by: Simo Sorce <idra@samba.org>
2012-04-23waf: move krb5 checks to a separate waf fileAlexander Bokovoy6-191/+205
With PROCESS_SEPARATE_RULE in wafsamba it is now possible to simplify configuration and checks for MIT/Heimdal Kerberos implementations. 1. Move MIT krb5 checks from source3/wscript to wscript_configure_krb5 2. Make sure they are called same way (--with-mit-krb5-checks) 3. If no configure checks identified MIT krb5 in system (or were disabled), make sure Heimdal build is selected, embedded (default) or system-provided. This makes logic of configuration unchanged for Heimdal builds but adds less hacky way to use MIT krb5 builds. The latter does not work yet as we need to untangle more subsystems from HDB/Heimdal-specific details but lays out a foundation for that. Signed-off-by: Simo Sorce <idra@samba.org>
2012-04-23Fix bug #8882 - Broken processing of %U with vfs_full_audit when force user ↵Jeremy Allison1-0/+11
is set. When doing a "force user" we need to remember what the "sanitized_username" was from the original connect. Autobuild-User: Jeremy Allison <jra@samba.org> Autobuild-Date: Mon Apr 23 19:52:19 CEST 2012 on sn-devel-104
2012-04-23Add complete test program for Linux kernel aio inside configure.in (I ↵Jeremy Allison1-6/+28
discovered yesterday there are systems with only half the glibc changes needed to implement userspace kaio.
2012-04-23s4-libnet Always return after composite_error()Andrew Bartlett4-1/+8
These instances should not cause a problem, but make it easier to audit for this kind of problem in the future with grep. Andrew Bartlett Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Mon Apr 23 14:29:45 CEST 2012 on sn-devel-104
2012-04-23s4-libnet Always return after composite_error()Andrew Bartlett3-0/+3
This can and does cause crashes as multiple conflicting sets of callbacks run on memory that may have been destroyed. Andrew Bartlett
2012-04-23s4-libnet: Fix continue_groupinfo_openuser to check correct state infoAndrew Bartlett1-2/+2
This meant that we would attempt to query the user that we could not open. This is a mirror of 4ba1647d5db59e5bb4911c399111e9286aac1a8e. Andrew Bartlett
2012-04-23build: Remove support for a system libsmbclientAndrew Bartlett2-10/+0
With the new --private-libraries option, there is no longer the need to have this support, which was aimed at avoiding the duplication between two different libsmbclient binaries in a packaged distribution. By using --private-libraries instead, we do not introduce a dependency between Samba 4.0 packages and whatever other packages are on the system. This effectivly reverts part of e3ffb31554927a24df35576bd753ee4eb2dc2a3c while kpeeing the improved SAMBA_LIBRARY declaration. Andrew Bartlett
2012-04-23build: Remove support for system libwbclientAndrew Bartlett2-9/+19
This effectively reverts commit 2c49782dc1069eebc9f5a5cd2a055b5912665595 This also adds explaination about what would need to be done to support a system libwbclient Andrew Bartlett
2012-04-23s4:librpc/rpc: directly return after composite_error()Stefan Metzmacher1-1/+1
metze Autobuild-User: Stefan Metzmacher <metze@samba.org> Autobuild-Date: Mon Apr 23 12:17:08 CEST 2012 on sn-devel-104
2012-04-23s4:libnet_become_dc: add missing 'return' after composite_error()Stefan Metzmacher1-0/+2
metze
2012-04-23selftest: Add test for bug #8884: wbinfo --group-info=administrator ↵Andrew Bartlett1-0/+2
segfaults s4-winbindd Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Mon Apr 23 07:44:50 CEST 2012 on sn-devel-104
2012-04-23s4-libnet: Fix continue_groupinfo_opengroup to check correct state infoAndrew Bartlett1-2/+2
This meant that we would attempt to query the group that we could not open. Andrew Bartlett
2012-04-23s4-libnet: Fix segfault shown by wbinfo --group-info=administratorAndrew Bartlett1-0/+1
The issue was that after the LookupNames call indicated that this was not a group, the call paths diverged, with both sucess and failure paths running. Andrew Bartlett
2012-04-23s3-dbwrap: A void function can not return a valueAndrew Bartlett1-1/+1
Only non-gcc compilers seem to notice this as an error. Andrew Bartlett Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Mon Apr 23 05:58:52 CEST 2012 on sn-devel-104
2012-04-22dlz_bind9: Also produce and install binary plugin for bind 9.9Andrew Bartlett1-1/+1
Thanks to Alexander Bokovoy <ab@samba.org> for the fix! Andrew Bartlett Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Sun Apr 22 10:51:57 CEST 2012 on sn-devel-104
2012-04-22idl: fix the charset for the source_dsa_address in DsReplicaDel callMatthieu Patou1-1/+1
Autobuild-User: Matthieu Patou <mat@samba.org> Autobuild-Date: Sun Apr 22 03:00:06 CEST 2012 on sn-devel-104
2012-04-21s3: Implement db_id for dbwrap_cacheVolker Lendecke1-0/+9
Autobuild-User: Volker Lendecke <vl@samba.org> Autobuild-Date: Sat Apr 21 13:46:00 CEST 2012 on sn-devel-104
2012-04-21s3: Initialize "stored_callback" in dbwrap_cacheVolker Lendecke1-0/+1
This should fix one of the recent flaky tests
2012-04-21s3: Remove two unused variablesVolker Lendecke1-2/+0
2012-04-21s3:smbcontrol: remove an unused variableMichael Adam1-4/+0
Autobuild-User: Michael Adam <obnox@samba.org> Autobuild-Date: Sat Apr 21 00:55:13 CEST 2012 on sn-devel-104
2012-04-20s3:memcache: remove the idmap-part from memcacheMichael Adam1-4/+0
This was useful before the idmap cache was moved to gencache. Nowadays it is available to smbd through gencache, so we can remove the extra caching layer.
2012-04-20s3:id_cache: do not use the in-memory idmap cache (it is going to be removed)Michael Adam4-110/+4
This also removes the ID_CACHE_FLUSH message.
2012-04-20s3:passdb: do not use the in-memory idmap cache (it will be removed)Michael Adam3-156/+0
2012-04-20s3-dbwrap: dbwrap_watch_record_stored => NT_STATUS_NOT_FOUND is ok...Stefan Metzmacher1-0/+3
Autobuild-User: Volker Lendecke <vl@samba.org> Autobuild-Date: Fri Apr 20 17:05:52 CEST 2012 on sn-devel-104
2012-04-19s3-g_lock: Use dbwrap_record_watch_send/recvVolker Lendecke3-454/+172
This simplifies the g_lock implementation. The new implementation tries to acquire a lock. If that fails due to a lock conflict, wait for the g_lock record to change. Upon change, just try again. The old logic had to cope with pending records and an ugly hack into ctdb itself. As a bonus, we now get a really clean async g_lock_lock_send/recv that can asynchronously wait for a global lock. This would have been almost impossible to do without the dbwrap_record_watch infrastructure.
2012-04-19s3-dbwrap: Add "listwatchers" to dbwrap_toolVolker Lendecke1-9/+62
2012-04-19s3-dbwrap: Add dbwrap_record_watch_send/recvVolker Lendecke8-1/+633
With this API you can asynchronously wait for a record to be modified
2012-04-20s3-dbwrap: Add dbwrap_set_stored_callbackVolker Lendecke6-2/+47
This is a per-db function that is called whenever some record is modified
2012-04-20s3-dbwrap: Add "db_context" to "db_record"Volker Lendecke4-0/+10
2012-04-20s3-dbwrap: Add dbwrap_db_idVolker Lendecke6-0/+50
This returns a blob uniquely identifying the database
2012-04-20s3: Fix msg_channel in the cluster caseVolker Lendecke1-3/+6
2012-04-20s3: Remove an unused variableVolker Lendecke1-1/+0
Autobuild-User: Volker Lendecke <vl@samba.org> Autobuild-Date: Fri Apr 20 12:44:20 CEST 2012 on sn-devel-104
2012-04-20selftest: Add extra info on what is failing in wbclient testsAndrew Bartlett1-65/+71
This may help me chase down the failure in my idmap rework. Andrew Bartlett Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Fri Apr 20 09:51:35 CEST 2012 on sn-devel-104
2012-04-20wafsamba: allow certain public libraries to be forced to be privateAndrew Bartlett3-0/+12
This will help installations where the Samba4 libraries must be used but the main system is not using the system libs that would normally be installed. This in particular impacts on libwbclient, which is a core dep, but is different to that used by the rest of a Samba 3.x based system. Use eg: ./configure --private-libraries=wbclient Andrew Bartlett Autobuild-User: Andrew Bartlett <abartlet@samba.org> Autobuild-Date: Fri Apr 20 03:27:22 CEST 2012 on sn-devel-104
2012-04-20libwbclient: bump version to 0.9, and add this version to .pc fileAndrew Bartlett2-3/+7
2012-04-20Move kdc_get_policy helper in the lsa server where it belongs.Simo Sorce7-96/+44
This was used in only 2 places, db-glue.c and the lsa server. In db-glue.c it is awkward though, as it forces to use an unconvenient lsa structure and conversions from time_t to nt_time only to have nt_times converted back to time_t for actual use. This is silly. Also the kdc-policy file was a single funciton library, that's just ridiculous. The loadparm helper is all we need to keep the values consistent, and if we ever end up doing something with group policies we will care about it when it's the time. the code would have to change quite a lot anyway. Autobuild-User: Simo Sorce <idra@samba.org> Autobuild-Date: Fri Apr 20 01:53:37 CEST 2012 on sn-devel-104
2012-04-19Use loadparm helper to find lifetime defaultsSimo Sorce1-6/+10
Implictly fixes buggy use of int for time_t
2012-04-19loadparm: Add helper function to fetch default lifetime policiesSimo Sorce2-0/+24
This use long to fetch time_t quantities, because there are architectures were time_t is a signed long but long != int, So long is the proper way to deal with it.
2012-04-19loadparm: Add convenience function to return long integersSimo Sorce1-0/+25
2012-04-19loadparm: Fix broken lp_ulong utility functionSimo Sorce1-2/+2
2012-04-19Move README file in the right place.Simo Sorce1-0/+0
Someone forgot to move the README when they moved the code ...