s3:popt_common: Reparse the username in popt_common_credentials_post()

When we parse the username in the options handling, the smb.conf file
has not been loaded yet. So we are not aware of a 'winbind separator'
set in the config file.

We need to read and set the username again in the post-processing of the
credentials.

https://bugzilla.samba.org/show_bug.cgi?id=12849

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

selftest: Use 'ad_dc' as the default for testenv

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

ctdb-locking: If a record could not be locked, log the key

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Mon Jun 19 19:56:22 CEST 2017 on sn-devel-144

ctdb-locking: Reduce logging in case of contention

Currently, every lock helper will log a message if it cannot get a lock.
This can spam the logs and overwhelm syslog if there are hundreds of
lock helpers waiting for contended record.

Instead keep track of the record for which we have already logged once
with specific timeout interval.  If we get timeout interval larger than
the previously logged interval, then log again once.  This will reduce
the amount of logs for contended records to a single log entry per 10
seconds per record.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>

ctdb-locking: There are no ALLDB locks any more

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>

ctdb-client: Add correct control names to log messages

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>

ctdb-tests: Fix function names in protocol test

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>

s3: VFS: Change SMB_VFS_SYMLINK to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Sun Jun 18 07:03:18 CEST 2017 on sn-devel-144

s3: VFS: Change SMB_VFS_READLINK to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_STATVFS to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_LINK to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_GET_QUOTA to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_DISK_FREE to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_CHFLAGS to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Change SMB_VFS_MKNOD to use const struct smb_filename * instead of const char *.

We need to migrate all pathname based VFS calls to use a struct
to finish modernising the VFS with extra timestamp and flags parameters.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: VFS: Remove old traces of smb_vfs_call_llistxattr().

This call doesn't exist anymore.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s4:libcli/smb_composite: make the additional gensec_update steps async

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Sat Jun 17 20:54:59 CEST 2017 on sn-devel-144

s4:libcli/smb_composite: add early returns to sesssetup.c:request_handler()

This makes it much clearer under which condutions the following code
operates.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: make the first round to gensec async

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: move gensec_update_ev() out of session_setup_spnego()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: move session_setup_spnego_restart() to the callers of session_setup_spnego()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: split out session_setup_spnego_restart() from session_setup_spnego()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: move chosen_oid to state->chosen_oid

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:libcli/smb_composite: simplify gensec_update_ev() handling in session_setup_spnego()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

auth/gensec: add GENSEC_UPDATE_IS_NTERROR() helper macro

This allows us to write clearer code that
checks for NT_STATUS_OK and NT_STATUS_MORE_PROCESSING_REQUIRED.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

auth/gensec: clear the update_busy_ptr in gensec_subcontext_start()

This is required to support async subcontexts.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ctdb-recovery: Delete empty records during recovery

Persistent databases are now always recovered by sequence number.  So
there is no need to keep the empty records in the database since they
will never be recovered record-by-record using RSN.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Sat Jun 17 16:47:55 CEST 2017 on sn-devel-144

ctdb-daemon: Delete empty records from persistent database

Persistent databases are now always recovered by sequence number.  So
there is no need to keep the empty records in the database since they
will never be recovered record-by-record using RSN.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Martin Schwenke <martin@meltin.net>

s3:smb2_sesssetup: allow a compound request after a SessionSetup

This is not a full fix yet as we don't allow compound requests
after going async.

With SMB 3.11 requiring signed TreeConnect requests it's pointless
to try to compound requests after a SessionSetup.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12845

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Sat Jun 17 10:55:25 CEST 2017 on sn-devel-144

s3:smb2_tcon: allow a compound request after a TreeConnect

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12844

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s3: libsmb: Correctly save and restore connection tcon in smbclient, smbcacls and smbtorture3.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: libsmb: Correctly do lifecycle management on cli->smb1.tcon and cli->smb2.tcon.

Treat them identically. Create them on demand after for a tcon call,
and delete them on a tdis call.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: libsmb: Fix cli_state_has_tcon() to cope with SMB2 connections.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: libsmb: Widen cli_state_get_tid() / cli_state_set_tid() to 32-bits.

Copes with SMB2 connections.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: smbtorture: Show correct use of cli_state_save_tcon() / cli_state_restore_tcon().

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

s3: libsmb: Add cli_state_save_tcon() / cli_state_restore_tcon().

Save and restore tcon pointers in smb1 or smb2 structs.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

libcli: smb: Add smb2cli_tcon_set_id().

Will be used in test and client code.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

libcli: smb: Add smbXcli_tcon_copy().

Makes a deep copy of a struct smbXcli_tcon *, will
be used later.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12831

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Richard Sharpe <realrichardsharpe@gmail.com>

dsdb: Add comment explaining requirements on DSDB_EXTENDED_SCHEMA_UPDATE_NOW_OID

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Fri Jun 16 23:43:46 CEST 2017 on sn-devel-144

dsdb: Do not prevent searches for @ATTRIBUTES because the DB is not set up yet

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

dsdb: Do not run dsdb_replace() on the calculated difference between old and new schema

We can set the database @INDEXLIST and @ATTRIBUTES to the full calculated
values, not the difference, and let the ldb layer work it out under the
transaction lock.

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

selftest: confirm that two attributes are also correctly set in the @ records

This shows that the current behaviour in dsdb_schema_set_indices_and_attributes(), while
not ideal, is not actually buggy.

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

selftest: Fix failure message in dsdb_schema_info

The rename changes the CN, not the lDAPDisplayName

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

krb5_wrap: handle KRB5_ERR_HOST_REALM_UNKNOWN in smb_krb5_get_realm_from_hostname()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:gensec_gssapi: fix CID 1409781: Possible Control flow issues (DEADCODE)

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>

selftest: Also wait for winbindd to start

This ensures that the posixacl.py test does not race against winbindd starting up and so
give wrong mappings

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12843

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

selftest: Correctly print message when nbt is not up in 20 seconds

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12843

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

tevent_threads: Fix a rundown race introduced with 1828011317b

The race is easily reproduced by adding a poll(NULL,0,10) in between the two
pthread_mutex_unlock calls in _tevent_threaded_schedule_immediate.

Before 1828011317b, the main thread was signalled only after the helper
had already unlocked event_ctx_mutex.

Full explaination follows:
-----------------------------------------------------------------
Inside _tevent_threaded_schedule_immediate() we have:

476         ret = pthread_mutex_unlock(&ev->scheduled_mutex);
477         if (ret != 0) {
478                 abort();
479         }

HERE!!!!

481         ret = pthread_mutex_unlock(&tctx->event_ctx_mutex);
482         if (ret != 0) {
483                 abort();
484         }

At the HERE!!! point, what happens is tevent_common_threaded_activate_immediate(),
which is blocked on ev->scheduled_mutex, get released and does:

514         while (ev->scheduled_immediates != NULL) {
515                 struct tevent_immediate *im = ev->scheduled_immediates;
516                 DLIST_REMOVE(ev->scheduled_immediates, im);
517                 DLIST_ADD_END(ev->immediate_events, im);
518         }

- making an immediate event ready to be scheduled.

This then returns into epoll_event_loop_once(), which then calls:

910         if (ev->immediate_events &&
911             tevent_common_loop_immediate(ev)) {
912                 return 0;
913         }

which causes the immediate event to fire. This immediate
event is the pthread job terminate event, which was previously
set up in pthreadpool_tevent_job_signal() by:

198         if (state->tctx != NULL) {
199                 /* with HAVE_PTHREAD */
200                 tevent_threaded_schedule_immediate(state->tctx, state->im,
201                                                    pthreadpool_tevent_job_done,
202                                                    state);

So we now call pthreadpool_tevent_job_done() - which does:

225         TALLOC_FREE(state->tctx);

calling tevent_threaded_context_destructor():

384         ret = pthread_mutex_destroy(&tctx->event_ctx_mutex); <---------------- BOOM returns an error !
385         if (ret != 0) {
386                 abort();
387         }

as we haven't gotten to line 481 above (the line after
HERE!!!!) so the tctx->event_ctx_mutex is still
locked when we try to destroy it.

So doing an additional:

        ret = pthread_mutex_lock(&tctx->event_ctx_mutex);
        ret = pthread_mutex_unlock(&tctx->event_ctx_mutex);

(error checking elided) forces tevent_threaded_context_destructor()
to wait until tctx->event_ctx_mutex is unlocked before it locks/unlocks
and then is guaranteed safe to destroy.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

dsdb: Remember the last ACL we read during a search and what it expanded to

It may well be the same as the next one we need to check, so we can
avoid parsing it again.

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Fri Jun 16 07:39:24 CEST 2017 on sn-devel-144

dsdb: Cache the result of checking the parent ACL

This should help a lot for large one-level searches and for subtree searches that are of
flat tree structures

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>

WHATSNEW: change the default for "map untrusted to domain" to "auto"

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

docs-xml: change the default for "map untrusted to domain" to "auto"

This makes the behaviour much more robust, particularly with forest child
domains over one-way forest trusts.

Sadly we don't support this kind of setup with our current ADDC, so
there's no way to have automated tests for this behaviour, but
at least we know it doesn't break any existing tests.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=8630

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

docs-xml: document "map untrusted to domain = auto"

BUG: https://bugzilla.samba.org/show_bug.cgi?id=8630

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

docs-xml: improve documentation of "map untrusted to domain"

BUG: https://bugzilla.samba.org/show_bug.cgi?id=8630

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

auth3: prepare the logic for "map untrusted to domain = auto"

This implements the same behavior as Windows,
we should pass the domain and account names given
by the client directly to the auth backends,
they can decide if they are able to process the
authentication pass it to the next backend.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=8630

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

auth3: call is_trusted_domain() as the last condition make_user_info_map()

We should avoid contacting winbind if we already know the domain is our
local sam or our primary domain.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=8630

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

gitignore: ignore .gpg-* generated files (for ubuntu 16.04)

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Garming Sam <garming@catalyst.net.nz>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Thu Jun 15 21:40:08 CEST 2017 on sn-devel-144

repl_meta_data: single valued error codes depend on change type

A replace leads to CONSTRAINT_VIOLATION while an add causes
ATTRIBUTE_OR_VALUE_EXISTS. For this we need to check the mod type
before the replmd_modify_la_* calls because they change everything
into a replace.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

replmd: special-case member return value in replmd_add_fix_la()

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Pair-programmed-with: Garming Sam <garming@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

replmd: check duplicate linked attributes

This is simple enough because we already have the sorted list.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Pair-programmed-with: Garming Sam <garming@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

replmd: check single values in replmd_add_fix_la

repl_meta_data knows whether linked attributes are appropriately
[un-]duplicated, and this is how it tells ldb_tdb that.

Signed-off-by: Garming Sam <garming@catalyst.net.nz>
Pair-programmed-with: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ldb: 1.1.31

* Add efficient function to find duplicate values in ldb messages
  (this makes large multi-valued attributes in ldb_tdb more efficient)

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>

ldb: relatively efficient functions for finding duplicate values

ldb backends need to make sure they are not adding duplicate values to
multi-valued attributes in ADD and MODIFY operations. Until now they
have done this inefficiently using nested loops. Here we add common
functions that deal with large numbers of values in O(n log n) time,
but continue to use the simple methods for small numbers of values.

These functions take a struct ldb_context pointer and an options flag
arguments, although the ldb is not used, and only one bit of the
options has meaning. This is to allow further patches to switch on
schema-aware comparisons.

This entails an ABI jump to add the two new functions.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

dsdb/tests/ldap: test single valued linked attributes

This fails, so we add it to selftest/knownfail.d/

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4/linked_attribute tests: test duplicate values

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

dsdb/tests/ldap: multivalued attributes

Various return codes tested against Windows 2012r2.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

python/test: delete_force() passes on command line args

This allows you to use e.g.:

     delete_force(self.ldb, ou, controls=['tree_delete:1'])

Only in tests of course.

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ldb.h whitespace

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ldb tests/ldb_mod_op_test: don't double include cmocka.h

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ldb: fix a typo

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

ldb: fix whitespace in ldb_msg.c

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

libcli:smb2: Gracefully handle not supported for FSCTL_VALIDATE_NEGOTIATE_INFO

If FSCTL_VALIDATE_NEGOTIATE_INFO is not implemented, e.g. in a SMB2 only
server then gracefully handle NT_STATUS_NOT_SUPPORTED too.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=12808

Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Guenther Deschner <gd@samba.org>
Pair-Programmed-With: Guenther Deschner <gd@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Thu Jun 15 17:32:45 CEST 2017 on sn-devel-144

g_lock: open with LOCK_ORDER_3

xattr_tdb needs g_lock in a clustered environment. Nobody else
uses LOCK_ORDER_3 at this moment, so this looks safe.

The last one to use this was dbwrap_watch.tdb, and that's gone. The only
other one was notify_index.tdb, and that's gone too.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

smbd: Claim version in g_lock

Protect smbd against version incompatibilities in a cluster.

At first startup smbd locks "samba_version_string" and writes its version
string. It then downgrades the lock to a read lock. Subsequent smbds check
against the version string and also keep the read lock around. If the version
does not match, we try to write our own version. But as there's a read lock,
the lock upgrade to write lock will fail due the read lock being around. So as
long as there's one smbd with this read lock, no other version of smbd will be
able to start.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture3: Test heuristic cleanup

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Heuristically check for server existence

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture3: Test lock conflict and cleanup

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture3: Test lock upgrade/downgrade

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Allow lock upgrade/downgrade

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture3: Test g_lock_write_data

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Make g_lock_dump return a complete list of locks

To be honest, it did not really make sense to just pass in
lock holders individually. You could argue that it made sense
with in reality only G_LOCK_WRITE around, but soon we will have
G_LOCK_READ and thus multiple lock holders on a single lock.

Now that we also have userdata, change the g_lock_dump API

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Add g_lock_write_data

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Make g_lock_record_store also store userdata

Sequel to the previous commit changing the get/put routines for
the on-disk format

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Reformat to allow userdata

The next patches will make g_locks carry data. This
prepares the on-disk format.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Move parsing routines together

No code change, just shuffling around:

Before this patchset, g_lock_parse was somewhere in the middle. This carries no
real logic, put it on top.

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: unparse->put

Make it more in line with server_id_get/put

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: parse->get

Make it more in line with server_id_get/put

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Remove a pointless "else"

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Remove unused g_lock_get

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Make it endian-neutral

Add explicit parsing

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: More correct error msg

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture3: Initial test g_lock

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: Fix two typos

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s4:ldap_server: implement async BindSASL

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Thu Jun 15 13:18:47 CEST 2017 on sn-devel-144

s4:ldap_server: set result = LDAP_SUCCESS at the end, when we're really done

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:ldap_server: avoid using talloc_reference()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:ldap_server: remove useless NT_STATUS_IS_OK(status) check

We checked a few lines above already, check with:
git show -U10

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:ldap_server: remove useless indentation level arround ldapsrv_backend_Init()

Check with git show -w

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:ldap_server: remove useless indentation level arround gensec_session_info()

Check with git show -w

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

s4:ldap_server: make the gensec_create_tstream() error checking more clear

Check with 'git show -w'.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>