Rusty Russell [Tue, 1 Nov 2011 23:38:32 +0000 (10:08 +1030)]
tdb2: don't be fascist when TDB_VERSION1 is specified.
We currently insist that a tdb file be a version1 file if tdb_open() is
passed the TDB_VERSION1 flag; we fail if it's actually a tdb2.
But that makes generic wrappers harder, and is unlikely to be what the
user wants: if they do, they can check tdb_get_flags() & TDB_VERSION1
after opening.
Amitay Isaacs [Wed, 2 Nov 2011 03:34:41 +0000 (14:34 +1100)]
dsdb: Handle the case when extended rights string is NULL
Pair-Programmed-With: Andrew Tridgell <tridge@samba.org>
Signed-off-by: Andrew Tridgell <tridge@samba.org>
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Wed Nov 2 07:03:40 CET 2011 on sn-devel-104
Amitay Isaacs [Tue, 25 Oct 2011 04:06:07 +0000 (15:06 +1100)]
s4-drepl: Search for application partitions in addition to main ones
To replicate application partitions (e.g. DNS partitions) consult
msDs-hasMasterNCs attribute as well. Also, make sure we don't add
same partition twice in the list. hasMasterNCs and msDs-hasMasterNCs
have domain, configuration and schema partitions common.
Amitay Isaacs [Mon, 24 Oct 2011 06:20:46 +0000 (17:20 +1100)]
s4-dnsserver: Build a dns name tree for correct enumeration
The result of EnumRecords/EnumRecords2 RPC calls, is a list of
dns records that are one level below in the name hierarchy starting
from the search name. This patch builds a tree of names to get
the list of records one level below the search names and correctly
count the number of child records for each of those.
Rusty Russell [Mon, 31 Oct 2011 20:41:25 +0000 (07:11 +1030)]
tdb2: use HAVE_LIBREPLACE instead of _SAMBA_BUILD_.
This is what lib/tdb2 really wants: to know if it should use normal
headers or replace.h. It currently uses _SAMBA_BUILD_, which is wrong
for ldb, for example.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Autobuild-User: Rusty Russell <rusty@rustcorp.com.au>
Autobuild-Date: Tue Nov 1 03:19:58 CET 2011 on sn-devel-104
David Disseldorp [Tue, 11 Oct 2011 09:44:03 +0000 (11:44 +0200)]
s4-torture: smb2 copychunk data integrity checks
Perform simple integrity checks on destination file data following
a successful copychunk request.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User: David Disseldorp <ddiss@samba.org>
Autobuild-Date: Mon Oct 31 19:28:20 CET 2011 on sn-devel-104
David Disseldorp [Fri, 30 Sep 2011 12:33:36 +0000 (14:33 +0200)]
s4-torture: add smb2 copychunk ioctl tests
test_ioctl_copy_chunk_multi: Issue a copychunk request with multiple
chunk descriptors.
test_ioctl_copy_chunk_over: Issue a copychunk request with overlapping
chunks.
test_ioctl_copy_chunk_append: append data as part of a copy chunk
test_ioctl_copy_chunk_tiny: issue request with sub FS block size chunk
descriptors.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Michael Adam [Sat, 29 Oct 2011 12:48:58 +0000 (14:48 +0200)]
s4:torture: test also break semantics between batch-oplocks and leases
This must have been a typo (listing EXCLUSIVE twice).
But BATCH and EXCLUSIVE oplocks apparently share the
same break semantics with respect to leases.
Michael Adam [Sat, 29 Oct 2011 09:56:48 +0000 (11:56 +0200)]
s4:torture:smb2: fix the durable_open test to succeed against w7 and w2k8r2
When a first client that has a durable open with share read/write/delete
and a read-write-handle lease on the file disconnects, a second
client will succeed in opening the file and the new client will be
given a RWH-lease if requested, not only a RH-lease, as was previously
checked in the test.
This might have been a bug in win7 build 7000, which is what the
comments in the test give as reference.
Andrew Bartlett [Wed, 26 Oct 2011 06:59:47 +0000 (17:59 +1100)]
s3-selftest Add all the LOCAL-* smbtorture tests to make test
The extra knownfail and skip entries are for tests the currently do
not pass or hang. We need to work out why this happens, and fix the
test or the test invocation.
Andrew Bartlett [Wed, 26 Oct 2011 06:32:55 +0000 (17:32 +1100)]
s3-build: Remove libbigballofmud.so
We no longer need this, as all the small test binaries have either
been converted rolled into python bindings and python-subunit tests, or have
been moved into smbtorture.
Andrew Bartlett [Wed, 26 Oct 2011 06:30:45 +0000 (17:30 +1100)]
s3-torture Remove t_stringoverflow as fstrcpy now uses strlcpy
Because we no longer use safe_strcpy() internally, we no longer have a
mode where we write to the end of the buffer in developer mode, and so
there isn't anything for this test to test.
Andrew Bartlett [Wed, 26 Oct 2011 02:38:15 +0000 (13:38 +1100)]
selftest: Have only one set of selftest knownfail and skip files
Listing tests that are not ever proposed (eg samba4.* in the samba3 selftest)
is not an error, so just combine the lists.
This is being done because some folks trying to learn how our 'make
test' works are having trouble following the distributed nature of the
selftest system.
Andrew Bartlett [Wed, 26 Oct 2011 04:21:11 +0000 (15:21 +1100)]
selftest: Remove --target option and the ability to run 'samba4 only' tests
This simplifies the selftest system, and by default we always tested
all of samba3 and samba4 in the waf build. This simply removes a
rarely used option for testing only part of the system. The make test
TESTS="^samba3" syntax remains unchanged, so no functionality is lost.
Michael Adam [Thu, 27 Oct 2011 22:11:08 +0000 (00:11 +0200)]
s4:torture:smb2: avoid leaking tree connects up to the main function from the durable_open test
The tree connects are handed in from the calling wrapper.
Those that are freed and reconnected inside the test function
can not be freed in the wrapper and stick to the torture_context
until this is released in the main function.
Autobuild-User: Michael Adam <obnox@samba.org>
Autobuild-Date: Fri Oct 28 02:37:25 CEST 2011 on sn-devel-104
projects / thirdparty / samba.git / log
