Michael Adam [Sat, 5 May 2012 00:12:25 +0000 (02:12 +0200)]
s3:registry: implement values_need_update and subkeys_need_update in the smbconf backend
It simply calls to the regdb functions.
This fixes a caching issue uncovered by recent changes.
Michael Adam [Fri, 4 May 2012 16:01:00 +0000 (18:01 +0200)]
s3:registry: return error when Key does not exist in regdb_fetch_values_internal()
Michael Adam [Fri, 4 May 2012 16:00:15 +0000 (18:00 +0200)]
s3:smbd: comment the lp_load call in reload_services()
Michael Adam [Fri, 4 May 2012 15:59:35 +0000 (17:59 +0200)]
SQUASH s3:registry: improve and fix race in the regtorture tool
Michael Adam [Fri, 4 May 2012 15:58:34 +0000 (17:58 +0200)]
s3:waf-build: add the regtorture tool to the build
Pair-Programmed-With: Gregor Beck <gbeck@sernet.de>
Michael Adam [Fri, 4 May 2012 15:57:51 +0000 (17:57 +0200)]
s3:registry: make reg_delete_path() delete the path recursively.
Pair-Programmed-With: Gregor Beck <gbeck@sernet.de>
Gregor Beck [Wed, 2 May 2012 11:52:57 +0000 (13:52 +0200)]
regtorture: fix exit code
Gregor Beck [Wed, 2 May 2012 11:52:36 +0000 (13:52 +0200)]
regtorture: flush partial lines only if num_procs==1
Gregor Beck [Wed, 2 May 2012 10:25:05 +0000 (12:25 +0200)]
regtorture: cleanup logging
Gregor Beck [Wed, 2 May 2012 09:50:47 +0000 (11:50 +0200)]
regtorture: use DEBUG/DEBUGADD to log progress
Gregor Beck [Wed, 2 May 2012 09:49:59 +0000 (11:49 +0200)]
regtorture: remove old key Case2
Gregor Beck [Wed, 2 May 2012 09:47:34 +0000 (11:47 +0200)]
regtorture: remove unused function
Gregor Beck [Wed, 2 May 2012 08:26:59 +0000 (10:26 +0200)]
regtorture: time creation of key
Gregor Beck [Wed, 2 May 2012 08:26:24 +0000 (10:26 +0200)]
regtorture: feature
Gregor Beck [Wed, 2 May 2012 08:12:10 +0000 (10:12 +0200)]
regtorture: fork
Gregor Beck [Mon, 30 Apr 2012 11:37:52 +0000 (13:37 +0200)]
regtorture: case2 values within different keys
Gregor Beck [Mon, 30 Apr 2012 11:17:46 +0000 (13:17 +0200)]
regtorture: case1 values within the same key
Gregor Beck [Mon, 30 Apr 2012 08:28:37 +0000 (10:28 +0200)]
regtorture: WIP
Gregor Beck [Fri, 27 Apr 2012 12:57:39 +0000 (14:57 +0200)]
regtorture: add to Makefile
Michael Adam [Wed, 25 Apr 2012 09:28:11 +0000 (11:28 +0200)]
TODO: start adding a s3 registry torture/benchmark program.
Gregor Beck [Mon, 30 Apr 2012 08:27:56 +0000 (10:27 +0200)]
s3:registry: add functions to conveniently create registry_values
Gregor Beck [Wed, 2 May 2012 09:22:54 +0000 (11:22 +0200)]
s3:registry: silence net_deletekey_recursive
WERR_BADFILE may not be an error for the caller because the result is the same.
Gregor Beck [Mon, 30 Apr 2012 08:26:32 +0000 (10:26 +0200)]
s3:registry: compile reg_create_path() & reg_delete_path()
Michael Adam [Thu, 12 Apr 2012 11:43:25 +0000 (13:43 +0200)]
TMP: add a script to test concurrent net registry setvalue commands
Michael Adam [Fri, 13 Apr 2012 11:34:58 +0000 (13:34 +0200)]
TMP: add a script for testing concurrent "net registry createkey" commands
Michael Adam [Thu, 29 Mar 2012 16:16:52 +0000 (18:16 +0200)]
TMP: s3:selftest: don't fork spoolss daemon and so on
Michael Adam [Tue, 24 Apr 2012 21:49:07 +0000 (23:49 +0200)]
selftest: Don't log stdout for smbd/nmbd/winbindd.
This reverts commit
ea95e0967b33d75acc960b5308a667e956ba4f29.
The intent of that change was to prevent error messages from
logfile creation before smb.conf was read. The deamons log to
stdout initially by default anyways. They call setup_logging
to change that to the real logging destination. Originally,
these were called too early. With the previous changes, the
daemons only setup logging after the config has been read
completely, so the final log destination is determined.
The logging to standard out has more disadvantages in selftest:
* All the log going to stdout makes it difficult to see the
important messages from test.
* The log files become much less readable without the debug headers.
* Running make test with high daemon debug level
(e.g. SMBD_OPTIONS=-d10) creates unbearable load on the
subunit formatter. log level 10 is actually not useable.
Michael Adam [Tue, 24 Apr 2012 21:43:22 +0000 (23:43 +0200)]
s3:smbd: move startup log message down to directly after setup_logging
So the startup log message goes to the log file (again), as the first message.
Michael Adam [Tue, 24 Apr 2012 21:42:00 +0000 (23:42 +0200)]
s3:smbd: move setup_logging() down to after config has been loaded completely
The initial logging goes to standard out. If the log file is changed by the
config, but the compiled-in log-directory is not accessible (or does not exist),
this patch prevents the error messages due to access denied to the log files.
Michael Adam [Tue, 24 Apr 2012 21:31:49 +0000 (23:31 +0200)]
s3:winbindd: move setup_logging() and startup message down to after config has been loaded completely
The initial logging goes to standard out. If the log file is changed by the
config, but the compiled-in log-directory is not accessible (or does not exist),
this patch prevents the error messages due to access denied to the log files.
Michael Adam [Tue, 24 Apr 2012 21:23:34 +0000 (23:23 +0200)]
s3:nmbd: move setup_logging and startup message down after the config has been read completely
The initial logging goes to standard out. If the log file is changed by the
config, but the compiled-in log-directory is not accessible (or does not exist),
this patch prevents the error messages due to access denied to the log files.
Michael Adam [Thu, 26 Apr 2012 13:31:28 +0000 (15:31 +0200)]
selftest: make the knownfail patterns for the driver_info_winreg more specific
Autobuild-User: Michael Adam <obnox@samba.org>
Autobuild-Date: Wed May 2 11:23:26 CEST 2012 on sn-devel-104
Andrew Bartlett [Wed, 2 May 2012 00:45:51 +0000 (10:45 +1000)]
s4-provision: Fix up --use-s3fs to create the directory for xattr.tdb
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Wed May 2 04:45:05 CEST 2012 on sn-devel-104
Andrew Bartlett [Wed, 2 May 2012 00:45:15 +0000 (10:45 +1000)]
s4-wintest: --use-s3fs in wintest
Andrew Bartlett [Wed, 2 May 2012 00:44:45 +0000 (10:44 +1000)]
s4-provision: Fix --use-s3fs to parse correctly
Andrew Bartlett [Tue, 1 May 2012 23:49:19 +0000 (09:49 +1000)]
testsuite/libsmbclient: Remove unused and expensive-to-link testsuite
This testsuite never got off the ground, and unlike the other
libsmbclient tests, it is not integrated into make test, has no driver
script we could integrate.
As it has been in this state for 10 years, and adding it to the compile
did not find any link-time issues (particularly in comparison to the
link-time cost for so many individual binaries), I am now removing them.
The libsmbclient examples are seperate to this, and remain in
examples/libsmbclient.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 23:39:33 +0000 (09:39 +1000)]
selftest: 'store dos attributes = yes' is now set in fileserver.conf
As such, there is no need to set it or the permissions-based mappings here as well.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 23:38:39 +0000 (09:38 +1000)]
s4-provision: set 'dcerpc endpoint servers' but not 'vfs objects'
The VFS objects are now set in the fileserver.conf, but this is only read by smbd, so
the provision-time smb.conf needs to turn off the extra Samba4 DCE/RPC services.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 23:37:08 +0000 (09:37 +1000)]
file_server: set 'store dos attributes = yes'
This means we do not need the old permissions-based mappings.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 10:11:04 +0000 (20:11 +1000)]
selftest: run more raw.samba3 against secshare simple file server
This allows these tests to run in an environment where they can pass,
as they fail when ACL support is turned on in smbd.
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Wed May 2 02:51:27 CEST 2012 on sn-devel-104
Andrew Bartlett [Mon, 30 Apr 2012 05:32:08 +0000 (15:32 +1000)]
selftest: Run smbtorture tests being run against s3dc against plugin_s4_dc as well
Andrew Bartlett [Tue, 1 May 2012 06:06:39 +0000 (16:06 +1000)]
selftest: Do not start samba4 srvsvc in plugin_s4_dc mode
Andrew Bartlett [Mon, 30 Apr 2012 13:39:27 +0000 (23:39 +1000)]
selftest: add hooks required for printing to Samba4.pm
Andrew Bartlett [Tue, 1 May 2012 21:19:13 +0000 (07:19 +1000)]
selftest: mark samba3.raw.samba3checkfsp as flapping on plugin_s4_dc
This fails on my Fedora 16 workstation, but passes on sn-devel.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 11:47:28 +0000 (21:47 +1000)]
selftest: mark samba3.raw.acls.inheritance(plugin_s4_dc) as flapping
This seems to succeed on sn-devel, but fails in a local make test on
my workstation.
Andrew Bartlett
Andrew Bartlett [Mon, 30 Apr 2012 13:39:10 +0000 (23:39 +1000)]
selftest: add knownfail entries for plugin_s4_dc tests
Andrew Bartlett [Mon, 30 Apr 2012 10:55:30 +0000 (20:55 +1000)]
selftest: change knownfail to cope with running plugin_s4_dc as well
Andrew Bartlett [Tue, 1 May 2012 06:06:25 +0000 (16:06 +1000)]
file_server: use embedded srvsvc
Andrew Bartlett [Mon, 30 Apr 2012 06:03:33 +0000 (16:03 +1000)]
file_server: Use the embedded winreg server
Andrew Bartlett [Mon, 30 Apr 2012 05:22:04 +0000 (15:22 +1000)]
file_server: use embedded ntsvcs server
Andrew Bartlett [Mon, 30 Apr 2012 05:19:43 +0000 (15:19 +1000)]
file_server: use embedded eventlog server
Andrew Bartlett [Mon, 30 Apr 2012 05:13:49 +0000 (15:13 +1000)]
file_server: forward dssetup, but use embedded svcctl for s3fs
Andrew Bartlett [Mon, 30 Apr 2012 05:08:38 +0000 (15:08 +1000)]
selftest: Use same pattern for path to share as Samba3.pm
This helps as we declare tests that were previously targetted at Samba3 only.
Andrew Bartlett
Andrew Bartlett [Mon, 30 Apr 2012 05:17:54 +0000 (15:17 +1000)]
selftest: Add hideunread share to plugin_s4_dc
Andrew Bartlett [Mon, 30 Apr 2012 05:16:15 +0000 (15:16 +1000)]
selftest: run plugin_s4_dc with 'acl_xattr xattr_tdb streams_depot' VFS modules
Andrew Bartlett [Thu, 26 Apr 2012 05:20:02 +0000 (15:20 +1000)]
selftest: prepare to run smbtorture tests against plugin_s4_dc
Andrew Bartlett [Wed, 18 Apr 2012 06:05:22 +0000 (16:05 +1000)]
selftest: attempt to test samba3hide in a different environment
This should ensure that the samba3hide test is still run and passes,
in a non-ACLed environment.
Andrew Bartlett
Andrew Bartlett [Tue, 17 Apr 2012 01:56:23 +0000 (11:56 +1000)]
selftest: Enable ACL testing against the s3dc environment
Previously, ACL tests were skipped and the correct modules for ACL
testing were not loaded.
The addition of a knownfail entry for raw.samba3* tests is due to an
inconsitancy between the behaviour with and without ACL modules loaded
- posix chown calls appear to be ignored in this mode.
Andrew Bartlett
Andrew Bartlett [Tue, 1 May 2012 07:13:34 +0000 (17:13 +1000)]
s3-smbd: Use security_session_user_level() rather than nt_token_check_sid()
This allows the unix.whoami test to pass when configured as part of the AD DC.
The struct auth_session_info is slightly different in the AD DC configuration
when using auth_samba4. In particular, there is a distinction between Guest
and Anonymous logins.
Andrew Bartlett
Christian Ambach [Tue, 1 May 2012 19:21:40 +0000 (21:21 +0200)]
s3:torture fix flakey testcase
don't put database into /tmp, use lp_private_dir() to put
it into the selftest prefix
Autobuild-User: Christian Ambach <ambi@samba.org>
Autobuild-Date: Wed May 2 00:57:05 CEST 2012 on sn-devel-104
Christian Ambach [Mon, 30 Apr 2012 15:02:46 +0000 (17:02 +0200)]
s3:selftest run LOCAL-IDMAP-TDB-COMMON in make test
Autobuild-User: Christian Ambach <ambi@samba.org>
Autobuild-Date: Tue May 1 11:07:08 CEST 2012 on sn-devel-104
Christian Ambach [Wed, 22 Feb 2012 14:44:27 +0000 (15:44 +0100)]
s3:torture: add idmap_tdb_common test code
Christian Ambach [Mon, 20 Feb 2012 11:13:56 +0000 (12:13 +0100)]
s3:winbindd/idmap_tdb: use idmap_tdb_common code
Christian Ambach [Sun, 26 Feb 2012 16:49:23 +0000 (17:49 +0100)]
s3:winbindd/idmap_tdb2: fix logic error in set_mapping_action
fix an endless loop
Christian Ambach [Fri, 17 Feb 2012 16:34:03 +0000 (17:34 +0100)]
s3:winbindd/idmap_tdb2: use idmap_tdb_common code
Christian Ambach [Wed, 25 Jan 2012 18:06:16 +0000 (19:06 +0100)]
s3:winbindd/autorid use idmap_tdb_common code in autorid
- use common logic for the allocation pool
- add a idmap_tdb style 1on1 mapping for non-domain SIDs
like Everyone (S-1-1-0)
Christian Ambach [Mon, 16 Jan 2012 16:21:38 +0000 (17:21 +0100)]
s3:winbindd add idmap_tdb_common file to store common code of TDB idmap backends
Christian Ambach [Tue, 17 Jan 2012 12:59:56 +0000 (13:59 +0100)]
s3:util add sid_check_is_wellknown_builtin()
Andrew Bartlett [Tue, 1 May 2012 04:42:33 +0000 (14:42 +1000)]
move VERSION to alpha21
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Tue May 1 08:16:38 CEST 2012 on sn-devel-104
Andrew Bartlett [Tue, 1 May 2012 04:40:48 +0000 (14:40 +1000)]
prepare WHATSNEW for alpha20 and mark as release
Amitay Isaacs [Thu, 19 Apr 2012 23:39:53 +0000 (09:39 +1000)]
s4-upgradedns: Update serverdn with only the attributes that have changed
This fixes the issue of ldb 'Operations Error' when trying to modify
hasPartialReplicaNCs attribute.
Autobuild-User: Amitay Isaacs <amitay@samba.org>
Autobuild-Date: Tue May 1 05:28:04 CEST 2012 on sn-devel-104
Jeremy Allison [Tue, 17 Apr 2012 19:30:15 +0000 (12:30 -0700)]
Fix bug #8873 - self granting privileges in security=ads.
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue May 1 01:04:46 CEST 2012 on sn-devel-104
Jeremy Allison [Mon, 30 Apr 2012 18:05:51 +0000 (11:05 -0700)]
Fix the loop unrolling inside resolve_ads(). If we don't get
an IP list don't use interpret_string_addr(), as this only
returns one address, use interpret_string_addr_internal()
instead.
Autobuild-User: Jeremy Allison <jra@samba.org>
Autobuild-Date: Mon Apr 30 23:21:16 CEST 2012 on sn-devel-104
Jeremy Allison [Fri, 27 Apr 2012 23:25:58 +0000 (16:25 -0700)]
Protect all of the name resolution methods from returning null addrs. Ensure all returns go through remove_duplicate_addrs2().
Jeremy Allison [Fri, 27 Apr 2012 23:07:20 +0000 (16:07 -0700)]
Fix convert_ss2service() to filter out zero addresses.
Jeremy Allison [Fri, 27 Apr 2012 23:02:15 +0000 (16:02 -0700)]
Fix remove_duplicate_addrs2 to do exactly what it says. Previously it could leave zero addresses in the list.
Matthias Dieter Wallnöfer [Sat, 21 Apr 2012 15:32:48 +0000 (17:32 +0200)]
s4:samldb LDB module - homogenize LDB search attributes arrays
First they do not need to be "static" any longer since we have abandoned
asynchronous result handling (where global variables have been important).
In addition add some "const" in order to protect us from unwanted writes.
Reviewed-by: Andrew Bartlett
Autobuild-User: Matthias Dieter Wallnöfer <mdw@samba.org>
Autobuild-Date: Mon Apr 30 16:46:20 CEST 2012 on sn-devel-104
Andrew Bartlett [Mon, 30 Apr 2012 10:29:56 +0000 (20:29 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_groupdel calls (bug #8889)
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Apr 30 14:41:36 CEST 2012 on sn-devel-104
Andrew Bartlett [Mon, 30 Apr 2012 10:27:33 +0000 (20:27 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_groupadd calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 10:25:55 +0000 (20:25 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_usermod calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 10:23:07 +0000 (20:23 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_userdel calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 10:14:42 +0000 (20:14 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_useradd calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 10:09:58 +0000 (20:09 +1000)]
s4-libnet: Add mem_ctx to libnet_rpc_groupinfo calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 10:02:42 +0000 (20:02 +1000)]
s4-libnet: Add mem_ctx to libnet_DomainClose calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 09:57:45 +0000 (19:57 +1000)]
s4-libnet: Add mem_ctx to libnet_DomainOpen calls (bug #8889)
Andrew Bartlett [Mon, 30 Apr 2012 09:35:25 +0000 (19:35 +1000)]
s4-libnet: Add mem_ctx to libnet_Lookup calls (bug #8889)
Christof Schmitt [Tue, 24 Apr 2012 21:33:46 +0000 (14:33 -0700)]
net: Let get*sid return error from passdb init
When initialize_password_db returns an error this means that the SID
stored in the backend cannot be read. Return this error directly
instead of creating a random SID through get_global_sam_sid.
Autobuild-User: Volker Lendecke <vl@samba.org>
Autobuild-Date: Mon Apr 30 13:07:20 CEST 2012 on sn-devel-104
Christof Schmitt [Tue, 24 Apr 2012 21:42:28 +0000 (14:42 -0700)]
passdb: Do not panic in initialize_password_db
A call to initialize_password_db leads to smb_panic in case the backend
returns an error. All callers to initialize_password_db check the return
value, so this code path should return the status instead of calling
smb_panic.
Move the call to smb_panic from pdb_get_methods_reload pdb_get_methods
to get it out of the initialize code path. This allows printing the
proper error message for 'net getlocalsid' which is much nicer than
printing the panic stack trace.
Andrew Bartlett [Wed, 25 Apr 2012 07:40:35 +0000 (17:40 +1000)]
s3-auth_samba4: use new_server_id_task() to allocate server id values
This is rather than just picking a random number out of the air.
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Apr 30 11:29:34 CEST 2012 on sn-devel-104
Andrew Bartlett [Thu, 21 Jul 2011 06:29:38 +0000 (16:29 +1000)]
s3-lib Add a way to allocate the task_id value in server_id
This safely allocates the task_id so that when we have multiple event
contexts, they can each have their own messaging context, particularly
for the imessaging subsystem under source4.
Andrew Bartlett
Andrew Bartlett [Mon, 30 Apr 2012 05:44:01 +0000 (15:44 +1000)]
s4-messaging: Fill in the whole server_id in all use cases
This started per https://bugzilla.samba.org/show_bug.cgi?id=8872#c4
and avoids any possible collision with a different process.
We also need to ensure that across a Samba installation on a single
node that id.vnn is the same. Samba4 previously used 0, while Samba3
used NONCLUSTER_VNN. When a message is sent between these 'different'
nodes, the error NT_STATUS_INVALID_DEVICE_REQUEST is raised.
Andrew Bartlett
Andrew Bartlett [Mon, 30 Apr 2012 04:57:41 +0000 (14:57 +1000)]
s4-torture: convert samba3misc tests to use torture_failure and torture_assert
This helps us when these tests fail, as subunit-formatted failures can
be declared as knownfail entries, and show up correctly in the make
test output.
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Apr 30 08:34:52 CEST 2012 on sn-devel-104
Andrew Bartlett [Mon, 30 Apr 2012 04:15:12 +0000 (14:15 +1000)]
s4-torture: move samba3 tests to use torture helper functions
Andrew Bartlett [Mon, 30 Apr 2012 04:05:19 +0000 (14:05 +1000)]
s4-torture: Move various samba3 tests to the torture_suite_add_1smb_test wrapper
Matthias Dieter Wallnöfer [Sat, 21 Apr 2012 16:16:43 +0000 (18:16 +0200)]
s4:samldb LDB module - implement "fSMORoleOwner" attribute protection
This is a very essential attribute since it references to various domain
master roles (PDC emulator, schema...) depending on which entry it has
been set. Incautious modifications can cause severe problems.
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Apr 30 02:04:24 CEST 2012 on sn-devel-104
Matthias Dieter Wallnöfer [Fri, 20 Apr 2012 16:15:23 +0000 (18:15 +0200)]
s4:dsdb/common/util.c - samdb_is_pdc() - fail if the "fSMORoleOwner" attribute has not been set
Matthieu Patou [Sun, 29 Apr 2012 05:12:40 +0000 (22:12 -0700)]
Add some debug for FOOBAR return case as they are hard to diagnose
Autobuild-User: Matthieu Patou <mat@samba.org>
Autobuild-Date: Sun Apr 29 09:08:15 CEST 2012 on sn-devel-104
Matthieu Patou [Sun, 29 Apr 2012 04:51:12 +0000 (21:51 -0700)]
dsdb: more RELAX to DBCHECK control shift so that only dbcheck can do uncontrolled changes
Matthieu Patou [Sat, 28 Apr 2012 22:55:06 +0000 (15:55 -0700)]
dsdb: change control from relax to dbcheck
Comment indicate that this is needed by dbcheck only and it permits
other projects to push broken schema and remain undetected
Matthieu Patou [Sat, 28 Apr 2012 22:53:41 +0000 (15:53 -0700)]
dns server: move very verbose debug to higher level 2 -> 8