s3:gencache: don't use transaction non non-persistent gencache_notrans.tdb

gencache_notrans.tdb is a non-persistent cache layer above the
persistent gencache.tdb. Despite its name, and despite the
nature of non-persistent tdbs, the current stabilization code
uses a transaction on gencache_notrans.tdb like this:

  transaction_start(cache)
  transaction_start(cache_notrans)
  traverse(cache_notrans, stabilize_fn)
  transaction_commit(cache)
  transaction_commit(cache_notrans)

where stabilze_fn does this on a record:
  1. store it to or delete it from cache
     (depending on the timeout)
  2. delete it from the cache_notrans

This patch changes gencache_notrans.tdb to avoid
transactions by using an all-record lock like this:

  tdb_allrecord_lock(cache_notrans)
  transaction_start(cache)
  traverse(cache_notrans, stabilize_fn_mod)
  transaction_commit(cache)
  traverse(cache_notrans, wipe_fn)
  tdb_wipe_all(cache_notrans)
  tdb_allrecord_unlock(cache_notrans)

with stabilize_fn_mod doing only:
  1. store the record to or delete it from cache
     (depending on the timeout)

and wipe_fn deleting the records from the gencache_notrans db.

This is a step towards making non-persistent-db specific features
like mutex locking usable for gencache_notrans.tdb.

Signed-off-by: Michael Adam <obnox@samba.org>

s3:gencache: simply stabilize() a bit more: remove error from state

state.error is set to true if and only if the traverse
callback returns error (-1), and hence only if the traverse
fails.

Hence the the error state is redundant.

Signed-off-by: Michael Adam <obnox@samba.org>

s3:gencache: fix logic in stabilization when deleting a record from stable cache

Set state->written = true in the delete case
if and only if the record has really been deleted.

This does currently not seem to lead to an unneeded
write to the DB, since failure to delete the record
will cause the traverse and hence the transaction
to cancel. But I think this is clearer.

Signed-off-by: Michael Adam <obnox@samba.org>

btrfs: don't leak opened directory handle

Closing a directory handle file descriptor via close() is undefined,
according to:
http://pubs.opengroup.org/onlinepubs/9699919799/functions/dirfd.html

Signed-off-by: Noel Power <noel.power@suse.com>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): David Disseldorp <ddiss@samba.org>
Autobuild-Date(master): Tue Nov  4 20:51:02 CET 2014 on sn-devel-104

s3: nmbd: Ensure NetBIOS names are only 15 characters stored.

This screws up if the name is greater than MAX_NETBIOSNAME_LEN-1 in the
unix charset, but less than or equal to MAX_NETBIOSNAME_LEN-1 in the DOS
charset, but this is so old we have to live with that.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Tue Nov  4 18:23:11 CET 2014 on sn-devel-104

Fix typo in docs-xml/manpages/net.8.xml

Typo: s/Createa/Create/

Signed-off-by: Marc Muehlfeld <mmuehlfeld@samba.org>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Tue Nov  4 04:29:23 CET 2014 on sn-devel-104

smbstatus: check for lock_path talloc failures

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Nov  4 02:07:35 CET 2014 on sn-devel-104

smbXsrv: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

notify_internal: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

printer_list: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nmbd_namelistdb: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

share_mode_lock: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

brlock: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

serverid: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

server_mutex: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

g_lock: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

dbwrap_watch: don't leak lock_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

lang_tdb: don't leak lock_path or data_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

winbindd_cache: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

winbindd: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

idmap_autorid: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

smbcontrol: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

net_registry: don't leak state_path onto talloc tos

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

net_idmap: don't leak state_path onto talloc tos

net_idmap currently uses a net_idmap_dbfile() helper to return the idmap
backed specific db path. Fix leaks in all callers.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

reg_perfcount: don't leak state_path onto talloc tos

reg_perfcount currently uses a counters_directory() helper to return the
counter specific tdb path. Fix leaks and error checking in the helper
function and all callers.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

reg_backend_db: don't leak state_path onto talloc tos

Also check for allocation failures, and ensure that the regdb refcount
is not set by regdb_open() on failure.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nt_printing: don't leak state_path onto talloc tos

Also check for allocation failures, and close tdbs in
nt_printing_tdb_upgrade error paths.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

pdb_tdb: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

account_pol: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

nmbd_winsserver: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

sharesec: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

eventlog: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

groupdb: don't leak state_path onto talloc tos

Also check for allocation failures.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

Change expired password flag indicator

Signed-off-by: Gabriel Buades <gbuades@soffid.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Jim McDonough <jmcd@samba.org>

s3:locking: remove dead code from brl_get_locks_readonly()

struct byte_range_lock *rw = NULL; will never change...

commit 105724073300af03eb0835b3c93d9b2e2bfacb07 removed the
possible assigment of 'rw'.

So we can remove all code under if (rw != NULL) { ...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Fri Oct 31 06:07:43 CET 2014 on sn-devel-104

s3:locking: Change from ndr_pull_struct_blob() to ndr_pull_struct_blob_all() so we fail if not all bytes are consumed.

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

s4:torture/smb2: test rename dir deny with open files

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

s3:smbd: Don't rename a dir with files open underneath

This is an EXPENSIVE check. We'll have to guard this with an option

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

selftest:Samba3: use "strict rename = yes"

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

s3:param: Add new option "strict rename".

Control whether smbd can rename directories containing
open files. Defaults to "no" (meaning we *can* do
such renames).

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

s3:locking: allow early return for share_entry_forall()

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

s3:locking: Introduce share_mode_forall

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

s3:locking: Rename share_mode_forall->share_entry_forall

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

libcli: Make sure status is initialized.

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture: Make sure server_type isn't used uninitialized.

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s3-libsmb: Make sure the stat structure is initialized.

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s3-rpcclient: Make sure current_nt_hash is initialized.

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

ctdb-includes: Reorder includes.h more logically

For now replace.h needs to be before tdb.h, since tdb.h needs
stdbool.h.  This is fixed upstream but not yet in various packaged
versions.

Signed-off-by: Martin Schwenke <martin@meltin.net>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Fri Oct 31 03:46:40 CET 2014 on sn-devel-104

ctdb-includes: Remove unused idtree.h

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

ctdb-includes: Clean up includes.h

Remove unused #defines.  CTDB doesn't use these.  _PUBLIC_ is defined
in replace.h anyway.

Remove unnecessary #includes.

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

build: fix check for subunit file when selftest prefix is given

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Michael Adam <obnox@samba.org>
Autobuild-Date(master): Thu Oct 30 22:45:45 CET 2014 on sn-devel-104

script: fix display of ten slowest tests if < 10 tests are run.

Note: $#array is the biggest index in an array in perl.
@array evaluated in scalar context is the number of elements.
Hence scalar(@array) = 1 + $#array

Or equivalently: 0 + @array = 1 + $#array

... :-)

Apart from this off-by-one error, the "unless" clause to trigger
the capping of the number of tests listed was wrong. Hence if
less then 10 tests were run, a number of blank lines were appended.

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

script/show_testsuite_time: error out if number argument is not a decimal number

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>

s3:loadparm: fix a comment typo

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Ralph Böhme <slow@samba.org>

selftest: only use Options.options.SELFTEST_PREFIX if it's not the default

The --with-selftest-prefix option is typically specified as argument
to './configure' ! Overwriting it for 'waf testonly' should only
happen with an explicit 'waf testonly --with-selftest-prefix=/some/path'.

This fixes a regression introduced by
commit edc1cedf932de2bc068da9a5db844ac4c48f4324
(selftest: Actually honor --with-selftest-prefix).

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>

selftest: Fake the hostname.

This fixes the GetPrinterDataEx(DNSMachineName) test on some machines.

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Thu Oct 30 11:35:15 CET 2014 on sn-devel-104

Don't drop any error codes just because we think there cannot be any errors we didn't think of if the user asked for a level 10 log.

Signed-off-by: Richard Sharpe <rsharpe@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Oct 30 09:14:51 CET 2014 on sn-devel-104

spoolss: remove unused fill_job_info3()

This logic has been moved into the previous EnumJobs(level=3) caller.
The info3 structure only contains two fields that are used, so it
doesn't make sense to have a separate helper for it.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

spoolss: fix jobid in level 3 EnumJobs response

Until now, these responses have incorrectly carried the printing backend
job identifier (sysjob), rather than the one allocated and returned by
Samba on job submission.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=10905

Reported-by: Franz Pförtsch <franz.pfoertsch@brose.com>
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

spoolss: fix jobid in level 2 GetJob and EnumJobs responses

Until now, these responses have incorrectly carried the printing backend
job identifier (sysjob), rather than the one allocated and returned by
Samba on job submission.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=10905

Reported-by: Franz Pförtsch <franz.pfoertsch@brose.com>
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

spoolss: fix jobid in level 1 GetJob and EnumJobs responses

Until now, these responses have incorrectly carried the printing backend
job identifier (sysjob), rather than the one allocated and returned by
Samba on job submission.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=10905

Reported-by: Franz Pförtsch <franz.pfoertsch@brose.com>
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

spoolss: fix GetJob jobid lookups

Clients issue GetJob requests using the jobid assigned by the spoolss
server. The corresponding printing backend (sysjob) identifier needs to
be resolved to locate the correct print queue entry.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=10905

Reported-by: Franz Pförtsch <franz.pfoertsch@brose.com>
Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

printing: add jobid_to_sysjob helper function

Samba needs to deal with two types of print job identifiers, those
allocated by the printing backend (sysjob ids), and those allocated
by Samba's spoolss server (jobids).

This change adds a helper function to map spoolss jobids to sysjob ids,
to go alongside the corresponding sysjob to jobid mapping function.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

dns: Fix a compile warning by casting explicitly

Signed-off-by: Kai Blin <kai@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

dbwrap: Avoid "includes.h" in dbwrap_cache.c

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

dbwrap: Add some required #includes to dbwrap.h

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

lib: loadparm.h uses TALLOC_CTX

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

dbwrap: Avoid an unnecessary #include

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

torture/spoolss: cleanup print_job_enum test jobs

Purge print queue on startup to ensure that the test is not affected by
jobs queued for prior tests. Also, empty the queue at the end of the
test to play nice with others.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Wed Oct 29 14:31:52 CET 2014 on sn-devel-104

s3-nmbd: Fix netbios name truncation.

Try and cope with truncation more intelligently.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10896

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>

autobuild: check whether ctdbd has been installed in the samba-ctdb target

An intermediate test until we have full runtime-tests.

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Tue Oct 28 08:05:40 CET 2014 on sn-devel-104

build: adapt comments for the clustering choice

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

build: further simplify --with-cluster-support case in configure

includes were only built for the ctdb checks and they are now
gone because we are building against included ctdb.

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

autobuild: adapt samba-ctdb target does not need a separate ctdb build any more

Signed-off-by: Michael Adam <obnox@samba.org>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>

ctdb: Rename CTDB_VERSION to CTDB_PROTOCOL

CTDB_VERSION really is the ctdb protocol version.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

build: Simplify check for building with ctdb

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

build: Remove configure checks for ctdb headers

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

ctdb: Rename ctdb socket variable from CTDB_PATH to CTDB_SOCKET

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

build: Remove configure option --with-ctdb-dir

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

build: Remove checks for ctdb features

Since we are always building with integrated CTDB, there is no need for
these checks.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

build: Remove configure option --enable-old-ctdb

CTDB source is now part of Samba tree and to enable clustering smbd
should be built against included CTDB.

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Michael Adam <obnox@samba.org>

ctdb-logging: Add logging via UDP logging using RFC5424

Some implementations may not understand RC3164 format messages on the
UDP socket, so add support for RFC5424 message format.

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

ctdb-logging: Add logging via UDP to 127.0.0.1:514 to syslog backend

This has most of the advantages of the old logd with none of the
complexity of the extra process.  There are several good syslog
implementations that can listen on the UDP port.

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

ctdb-logging: Add non-blocking Unix domain logging to syslog backend

Format messages as per RFC3164.

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

ctdb-logging: New option CTDB_LOGGING, remove CTDB_LOGFILE, CTDB_SYSLOG

Remove --logfile and --syslog daemon options and replace with
--logging.

Modularise and clean up logging initialisation code.  The
initialisation API includes an app_name argument that is currently
unused - this will be used in extensions to the syslog backend.

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

ctdb-logging: Replace logd code with a basic syslog(3) implementation

It is much simpler for most cases to have a syslog backend that
doesn't need a separate CTDB-specific logging daemon.  This loses the
lossy, non-blocking mode provided by logd.  However, a corresponding
feature with a completely different implemention (not requiring an
extra daemon) will be re-added into the syslog backend.  In an ideal
world the new implementation would be added first but unfortunately
that is hard to do because the logd code is hooked in at more than one
place.

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

ctdb-logging: Simplify file logging timestamping with timeval_str_buf()

If nothing else, this is slightly more portable.

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

ctdb-logging: Separate out syslog and file logging backends

This makes the code cleaner and allows the syslog backend to be easily
modified without affecting other code.  Also do some extra clean-up,
including whitespace fixups.

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

ctdb-logging: Rework debug level parsing

Put declarations into ctdb_logging.h, factor out some common code,
clean up #includes.

Remove the check so see if the 1st character of the debug level is
'-'.  This is wrong, since it is trying to check for a negative
numeric debug level (which is no longer supported) and would need to
be handled in the else anyway.

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

ctdb-logging: Remove ctdb element from struct ctdb_log_state

This is set but otherwise not used.  This allows the 1st argument to
ctdb_set_logfile() to be generalised to a TALLOC_CTX.

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

ctdb-logging: Remove log member from struct ctdb_context

This is only used by logging code and there is already a file-level
variable for this.  struct ctdb_context already contains too many
things.

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

ctdb-logging: Rename ctdb_log_handler() to ctdb_child_log_handler()

Now it is obvious that it has something to do with child processes.

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

ctdb-logging: Remove debug levels DEBUG_ALERT and DEBUG_CRIT

Internally map them to DEBUG_ERR to limit code churn.

This reduces the unwieldy number of debug levels used by CTDB.  ALERT
and CRIT aren't of much use as separate errors, since everything from
ERR up should always be logged.  In future just ERR can be used.

This also improves compatibility with Samba's debug.c system priority
mapping.

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

ctdb-logging: Remove DEBUG_EMERG

It isn't used and shouldn't be.  CTDB can't make the system unusable.

Update associated test to ensure that EMERG isn't attempted.  Actually
test all remaining debug levels and modernise the test a bit.

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

ctdb-test: Remove unused function check_ctdb_logfile()

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

ctdb-logging: Add missing newline when logging to file

This got lost with the transition to the new Samba debug code.

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

spoolss: fix handling of bad EnumJobs levels

Currently Samba is inconsistent when returning WERR_UNKNOWN_LEVEL
errors for spoolss EnumJobs requests - if no print jobs are present,
then WERR_OK will be returned, regardless of whether the EnumJobs level
is supported or not.
This change fixes this behaviour, by catching invalid or unsupported
levels prior to the no-jobs response fast-path.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=10898

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Oct 28 03:05:35 CET 2014 on sn-devel-104

torture/spoolss: test EnumJobs with an invalid level

This test confirms that EnumJobs WERR_INVALID_LEVEL responses remain
consistent with and without the presence of outstanding print jobs.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

torture/spoolss: add werr_expected arg to EnumJobs helper

This will allow for the future testing of specific EnumJobs failure
cases.

Signed-off-by: David Disseldorp <ddiss@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>

s4-dns: Add support for BIND 9.10

Signed-off-by: Amitay Isaacs <amitay@gmail.com>
Reviewed-by: Matthieu Patou <mat@matws.net>
Autobuild-User(master): Amitay Isaacs <amitay@samba.org>
Autobuild-Date(master): Sat Oct 25 05:42:19 CEST 2014 on sn-devel-104