samba.git
14 years agos4:howto.txt - add a notice about the upgrade possibilities and fix line spaces
Matthias Dieter Wallnöfer [Fri, 27 Nov 2009 10:02:20 +0000 (11:02 +0100)]
s4:howto.txt - add a notice about the upgrade possibilities and fix line spaces

14 years agos4:upgrade script - rename it to "upgrade_from_s3" and do some rework
Matthias Dieter Wallnöfer [Fri, 27 Nov 2009 09:50:03 +0000 (10:50 +0100)]
s4:upgrade script - rename it to "upgrade_from_s3" and do some rework

- Give a better name to the script
- Move it to the location where also "upgradeprovision" resides
- Fix up trailing whitespaces and tabs

14 years agos3-nsstest: drastically shrink size and dependencies of nsstest binary.
Günther Deschner [Thu, 25 Jun 2009 16:18:50 +0000 (18:18 +0200)]
s3-nsstest: drastically shrink size and dependencies of nsstest binary.

The size went down from 6.4M to 104K on my box.

Guenther

14 years agos4-install: old systems don't have sed -i
Andrew Tridgell [Fri, 27 Nov 2009 08:58:00 +0000 (19:58 +1100)]
s4-install: old systems don't have sed -i

surprisingly, this failure came from SLES8. I didn't realise that the
-i option was so new!

14 years agos4-selftest: s4 passes all the SMB2-LOCK tests now
Andrew Tridgell [Fri, 27 Nov 2009 08:11:28 +0000 (19:11 +1100)]
s4-selftest: s4 passes all the SMB2-LOCK tests now

14 years agos4-torture: fixed error code for s4 in SMB2-LOCK test
Andrew Tridgell [Fri, 27 Nov 2009 08:10:50 +0000 (19:10 +1100)]
s4-torture: fixed error code for s4 in SMB2-LOCK test

s4 returns NETWORK_NAME_DELETED if you attempt to use an invalid tree connection
for a lock. This test (correctly I think) happens before we validate the file handle.
That implies that when you pass both a closed handle and a invalid tree you
should get NT_STATUS_NETWORK_NAME_DELETED.

14 years agos4-ntvfs: win7 does not check for the NONE smb2 lock flag on unlock
Andrew Tridgell [Fri, 27 Nov 2009 08:09:10 +0000 (19:09 +1100)]
s4-ntvfs: win7 does not check for the NONE smb2 lock flag on unlock

14 years agos4-selftest: s4 passes this test now
Andrew Tridgell [Fri, 27 Nov 2009 07:10:01 +0000 (18:10 +1100)]
s4-selftest: s4 passes this test now

14 years agos4-torture: mark s4 as doing valid lock range tests on SMB
Andrew Tridgell [Fri, 27 Nov 2009 07:09:50 +0000 (18:09 +1100)]
s4-torture: mark s4 as doing valid lock range tests on SMB

14 years agos4-ntvfs: move valid lock range test from smb2 layer to generic code
Andrew Tridgell [Fri, 27 Nov 2009 07:08:51 +0000 (18:08 +1100)]
s4-ntvfs: move valid lock range test from smb2 layer to generic code

win7 also fails invalid lock ranges on SMB

14 years agos4:smb_server/smb2: don't reset highest_smb2_seqnum on SMB2 Cancel
Stefan Metzmacher [Thu, 26 Nov 2009 13:18:50 +0000 (14:18 +0100)]
s4:smb_server/smb2: don't reset highest_smb2_seqnum on SMB2 Cancel

metze

14 years agos4-torture: fixed expected error codes for s4 in SMB2-LOCK
Andrew Tridgell [Fri, 27 Nov 2009 03:54:22 +0000 (14:54 +1100)]
s4-torture: fixed expected error codes for s4 in SMB2-LOCK

I think the error/success codes returned by windows for these tests
are quite bogus. The ones s4 gives are much more reasonable. The
locking ones returning NT_STATUS_SUCCESS could lead to data loss, as
an application thinks it has a file locked correctly when it fact it
doesn't, so it could do an unsafe modify.

14 years agos4-drstest: don't use getenv("LDB_URL") in test suites
Andrew Tridgell [Fri, 27 Nov 2009 03:42:05 +0000 (14:42 +1100)]
s4-drstest: don't use getenv("LDB_URL") in test suites

I was stumped for a while as to why the drs test suite was failing for
me. It turned out that it looked for LDB_URL in the environment, and
used it if set. I had it set in my terminal, and it was happily
munching on my sam.ldb while testing. Quite a cute bug really :-)

14 years agos4-pvfs: fixed access check failure in SFILEINFO test
Andrew Tridgell [Fri, 27 Nov 2009 03:22:29 +0000 (14:22 +1100)]
s4-pvfs: fixed access check failure in SFILEINFO test

matching windows behaviour is not always the right thing to do!

14 years agos4-ldb: improve detection of whether the server has a GC port
Andrew Tridgell [Fri, 27 Nov 2009 03:20:47 +0000 (14:20 +1100)]
s4-ldb: improve detection of whether the server has a GC port

We were trying to open $SERVER:3268 regardless, which could result in
creating a file called "localdc1:3268", which led to subsequent test
failures

14 years agos4-ldb: better to test for valid arguments in ldb library than commandline
Andrew Tridgell [Fri, 27 Nov 2009 03:18:39 +0000 (14:18 +1100)]
s4-ldb: better to test for valid arguments in ldb library than commandline

We were testing for valid DNs in ldbrename in the command line
tool. This hid a bug in the ldb library where we caught a bad DN in
the objectclass module rather than in the main ldb code. It is better
to do validation of the DNs passed on the command line in the library
code, as this gives us more consistent error handling between the
programming APIs for ldb and the command line.

14 years agos4-smb2: SMB2 uses NT_STATUS_CANCELLED for cancelled locks
Andrew Tridgell [Thu, 26 Nov 2009 06:38:50 +0000 (17:38 +1100)]
s4-smb2: SMB2 uses NT_STATUS_CANCELLED for cancelled locks

14 years agos4-smb2: sequence numbers are not checked in SMB2_OP_CANCEL
Andrew Tridgell [Thu, 26 Nov 2009 06:38:11 +0000 (17:38 +1100)]
s4-smb2: sequence numbers are not checked in SMB2_OP_CANCEL

14 years agotorture: fixed SMB2-LOCK valgrind error
Andrew Tridgell [Thu, 26 Nov 2009 06:03:20 +0000 (17:03 +1100)]
torture: fixed SMB2-LOCK valgrind error

14 years agos4-smb2: check for invalid SMB2 lock ranges
Andrew Tridgell [Thu, 26 Nov 2009 05:53:51 +0000 (16:53 +1100)]
s4-smb2: check for invalid SMB2 lock ranges

14 years agos4-smb2: check for an invalid lock flags combination
Andrew Tridgell [Thu, 26 Nov 2009 05:35:03 +0000 (16:35 +1100)]
s4-smb2: check for an invalid lock flags combination

UNLOCK with FAIL_IMMEDIATELY is not allowed

14 years agos4-install: fixed install path for python scripts
Andrew Tridgell [Thu, 26 Nov 2009 02:06:01 +0000 (13:06 +1100)]
s4-install: fixed install path for python scripts

when we install python scripts we need to fix the internal path used
to find modules. We also need to install the scripts in the right
place. Most of them should go in $SBINDIR not share/setup/

14 years agos4:upgradeprovision Rework update_machine_account_password() tranactions
Andrew Bartlett [Thu, 26 Nov 2009 21:10:54 +0000 (08:10 +1100)]
s4:upgradeprovision Rework update_machine_account_password() tranactions

This balances the transaction_begin() and transactin_prepare_commit() calls

Andrew Bartlett

14 years agos4:dsdb Don't segfault with ldb_transaction_prepare_commit() without begin()
Andrew Bartlett [Thu, 26 Nov 2009 21:05:59 +0000 (08:05 +1100)]
s4:dsdb Don't segfault with ldb_transaction_prepare_commit() without begin()

It is up to other modules to complain if
ldb_transaction_prepare_commit() is called before
ldb_transaction_begin_transaction()

Andrew Bartlett

14 years agos4:upgradeprovision add 'exit $failed' to blackbox test
Andrew Bartlett [Thu, 26 Nov 2009 11:01:54 +0000 (22:01 +1100)]
s4:upgradeprovision add 'exit $failed' to blackbox test

14 years agos4:upgradeprovision Use mkdtemp to create unique tempoary directory names
Andrew Bartlett [Thu, 26 Nov 2009 10:52:40 +0000 (21:52 +1100)]
s4:upgradeprovision Use mkdtemp to create unique tempoary directory names

14 years agos4:selftest Add tests for upgradeprovision
Andrew Bartlett [Thu, 26 Nov 2009 04:34:53 +0000 (15:34 +1100)]
s4:selftest Add tests for upgradeprovision

14 years agos4:upgradeprovision Rework script, and reset machine account pw
Andrew Bartlett [Thu, 26 Nov 2009 04:32:49 +0000 (15:32 +1100)]
s4:upgradeprovision Rework script, and reset machine account pw

The rework corrects some duplication and errors in the original
script, found when preparing an automated test of the script.

The code to reset the machine account password avoids issues with AES
keys and salting, which may not otherwise be solved by the upgrade.

Andrew Bartlett

14 years agos4:ldb Provide bindings for ldb_transaction_prepare_commit()
Andrew Bartlett [Thu, 26 Nov 2009 04:32:06 +0000 (15:32 +1100)]
s4:ldb Provide bindings for ldb_transaction_prepare_commit()

14 years agos4:provision Make setting the domain SID in the self join optional
Andrew Bartlett [Thu, 26 Nov 2009 03:57:39 +0000 (14:57 +1100)]
s4:provision Make setting the domain SID in the self join optional

14 years agoFix path to upgradeprovision
Andrew Bartlett [Thu, 26 Nov 2009 01:15:22 +0000 (12:15 +1100)]
Fix path to upgradeprovision

14 years agos4: Improve updateprovision
Matthieu Patou [Wed, 25 Nov 2009 13:26:35 +0000 (16:26 +0300)]
s4: Improve updateprovision

* Define a simple upgrade process mode (module storage change, file name change, copy of new file)
  * Move the schema, configuration and current object upgrade into full upgrade mode
  * Added the --full switch to select the full upgrade mode, and made simple upgrade mode the default
  * Make updateprovision works without any switch (update the provision in the default location)
  * Cleanup the messages
  * Create the reference provision in a subdirectory of the updated provision

14 years agos4: update What's new and explain how to upgrade a samba4 provision
Matthieu Patou [Wed, 25 Nov 2009 14:10:52 +0000 (17:10 +0300)]
s4: update What's new and explain how to upgrade a samba4 provision

14 years agos4: Rename the script
Matthieu Patou [Wed, 25 Nov 2009 08:42:16 +0000 (11:42 +0300)]
s4: Rename the script

14 years agos3-kerberos: Fix Bug #6929: build with recent heimdal.
Günther Deschner [Fri, 27 Nov 2009 00:06:36 +0000 (01:06 +0100)]
s3-kerberos: Fix Bug #6929: build with recent heimdal.

Heimdal changed the KRB5_DEPRECATED define (which now may not take an identifier
for activation) in new releases (like 1.3.1).

Guenther

14 years agos3-spoolss: use macros to fillup driver levels.
Günther Deschner [Thu, 26 Nov 2009 22:46:50 +0000 (23:46 +0100)]
s3-spoolss: use macros to fillup driver levels.

Guenther

14 years agos3-smbd: no drsuapi server stubs yet in s3. Fixes the build, sorry.
Günther Deschner [Thu, 26 Nov 2009 19:26:01 +0000 (20:26 +0100)]
s3-smbd: no drsuapi server stubs yet in s3. Fixes the build, sorry.

Guenther

14 years agos3-rpc: running minimal_includes.pl on rpc_client and rpc_server.
Günther Deschner [Thu, 26 Nov 2009 19:17:07 +0000 (20:17 +0100)]
s3-rpc: running minimal_includes.pl on rpc_client and rpc_server.

Guenther

14 years agos3-rpc: Avoid including every pipe's client and server stubs everywhere in samba.
Günther Deschner [Thu, 26 Nov 2009 17:21:28 +0000 (18:21 +0100)]
s3-rpc: Avoid including every pipe's client and server stubs everywhere in samba.

Guenther

14 years agos3-passdb: move some defines out of rpc headers (they really belong to passdb).
Günther Deschner [Thu, 26 Nov 2009 17:15:39 +0000 (18:15 +0100)]
s3-passdb: move some defines out of rpc headers (they really belong to passdb).

Guenther

14 years agos3-winbind: properly include netlogon.idl in wbint.idl.
Günther Deschner [Thu, 26 Nov 2009 17:58:36 +0000 (18:58 +0100)]
s3-winbind: properly include netlogon.idl in wbint.idl.

Guenther

14 years agos3-spoolss: pre cosmetics: reorder levels in construct_printer_info levels.
Günther Deschner [Thu, 26 Nov 2009 17:57:24 +0000 (18:57 +0100)]
s3-spoolss: pre cosmetics: reorder levels in construct_printer_info levels.

Guenther

14 years agotevent: Fix a comment
Volker Lendecke [Thu, 26 Nov 2009 16:18:00 +0000 (17:18 +0100)]
tevent: Fix a comment

14 years agolibrpc: rerun "make idl"
Stefan Metzmacher [Wed, 25 Nov 2009 13:42:24 +0000 (14:42 +0100)]
librpc: rerun "make idl"

metze

14 years agopidl:Samba3/ClientNDR: $size can be 'foo / 2' so we need to add '(' and ')'
Stefan Metzmacher [Wed, 25 Nov 2009 13:31:34 +0000 (14:31 +0100)]
pidl:Samba3/ClientNDR: $size can be 'foo / 2' so we need to add '(' and ')'

foo / 5 * sizeof(bar)' isn't the same as
'(foo / 2) * sizeof(bar)'.

metze

14 years agopidl:NDR/Parser: $size can be 'foo / 2' so we need to add '(' and ')'
Stefan Metzmacher [Wed, 25 Nov 2009 13:21:32 +0000 (14:21 +0100)]
pidl:NDR/Parser: $size can be 'foo / 2' so we need to add '(' and ')'

foo / 5 * sizeof(bar)' isn't the same as
'(foo / 2) * sizeof(bar)'.

metze

14 years agocreate-tarball: Make it more comfortable to use the --copy-docs option.
Karolin Seeger [Thu, 26 Nov 2009 10:18:39 +0000 (11:18 +0100)]
create-tarball: Make it more comfortable to use the --copy-docs option.

Now you can directly specify the 'make release' output dir without any
modifications.

Karolin

14 years agos4:ldap.py - add a test for the enhanced operational attributes check
Matthias Dieter Wallnöfer [Thu, 26 Nov 2009 08:51:56 +0000 (09:51 +0100)]
s4:ldap.py - add a test for the enhanced operational attributes check

(Deny creation of entries with operational attributes specified)

14 years agos4:objectclass LDB module - Prevent write operations on constructed attributes
Matthias Dieter Wallnöfer [Thu, 26 Nov 2009 09:54:20 +0000 (10:54 +0100)]
s4:objectclass LDB module - Prevent write operations on constructed attributes

14 years agos4:operational LDB module - Don't do the write checks here
Matthias Dieter Wallnöfer [Thu, 26 Nov 2009 09:21:44 +0000 (10:21 +0100)]
s4:operational LDB module - Don't do the write checks here

Let this perform the schema in the "objectclass" module.

14 years agos3-kerberos: next step to resolve Bug #6929: build with recent heimdal.
Günther Deschner [Thu, 26 Nov 2009 09:15:45 +0000 (10:15 +0100)]
s3-kerberos: next step to resolve Bug #6929: build with recent heimdal.

Based on patch from Allan <allan@archlinux.org>.

Also should fix the FreeBSD build on the buildfarm.

Guenther

14 years agos4-smbtorture: refactor RPC-SAMR-LARGE-DC test a little more.
Günther Deschner [Thu, 26 Nov 2009 00:25:56 +0000 (01:25 +0100)]
s4-smbtorture: refactor RPC-SAMR-LARGE-DC test a little more.

Guenther

14 years agos4-smbtorture: add torture_samr_context to carry some RPC-SAMR-X specifics.
Günther Deschner [Wed, 25 Nov 2009 23:32:47 +0000 (00:32 +0100)]
s4-smbtorture: add torture_samr_context to carry some RPC-SAMR-X specifics.

Guenther

14 years agoFix warnings with talloc_asprintf.
Jeremy Allison [Wed, 25 Nov 2009 22:47:36 +0000 (14:47 -0800)]
Fix warnings with talloc_asprintf.
Jeremy.

14 years agoMake us pass RAW-CHKPATH with a case sensitive share.
Jeremy Allison [Wed, 25 Nov 2009 21:17:56 +0000 (13:17 -0800)]
Make us pass RAW-CHKPATH with a case sensitive share.
I know Volker will look at this closely so here's the explaination :-).
Originally on a case-sensitive share we simply did a stat (or lstat)
call and returned success of fail based on the result. However this
failed to take account of incoming paths with a wildcard (which must
always fail, and with different error messages depending on whether
the wildcard is the last component or in the path). Also it failed
to take account of a stat fail with ENOENT due to a missing component
of the path as the last component (which is ok as it could be a new
file)  or if the ENOENT was due to the missing component within
the path (not the last component) - which must return the correct
error. What this means is that with "case sensitive = yes" we do
one more talloc call (to get the parent directory) and one more
stat call (on the parent directory) in the case where the stat
call fails. I think this is an acceptable overhead to enable
case sensitive shares to return the correct error messages for
applications. Volker please examine carefully :-).
Jeremy.

14 years agoAdd RAW-CHKPATH test with case-sensitive share.
Jeremy Allison [Wed, 25 Nov 2009 21:17:38 +0000 (13:17 -0800)]
Add RAW-CHKPATH test with case-sensitive share.
Jeremy.

14 years agos3-build: try to fix QNX build ("delay" is defined in lib headers).
Günther Deschner [Wed, 25 Nov 2009 21:03:19 +0000 (22:03 +0100)]
s3-build: try to fix QNX build ("delay" is defined in lib headers).

Guenther

14 years agos4/torture: port SMBv1 RAW-LOCK tests to SMBv2
Steven Danneman [Thu, 19 Nov 2009 00:35:03 +0000 (16:35 -0800)]
s4/torture: port SMBv1 RAW-LOCK tests to SMBv2

RAW-LOCK ported as:

RAW-LOCK-LOCK, RAW-LOCK-LOCKX -> SMB2-LOCK-LOCK
RAW-PIDHIGH -> removed, no longer relevant
RAW-ASYNC -> SMB2-LOCK-ASYNC, SMB2-LOCK-CANCEL, SMB2-LOCK-CANCEL-TDIS, SMB2-LOCK-CANCEL-LOGOFF
RAW-ERRORCODE -> SMB2-LOCK-ERRORCODE
RAW-CHANGETYPE -> removed, no longer relevant
RAW-ZEROBYTELOCKS -> SMB2-LOCK->ZEROBYTELENGTH
RAW-UNLOCK -> SMB2-LOCK-UNLOCK
RAW-MULTIPLE_UNLOCK -> SMB2-LOCK-MULTIPLE-UNLOCK
RAW-STACKING -> SMB2-LOCK-STACKING

BASE-LOCK ported as:

BASE-LOCK-LOCK1 -> SMB2-LOCK-ERRORCODE, timeout is no longer relevant
BASE-LOCK-LOCK2 -> SMB2-LOCK-CONTEND, SMB2-LOCK-LOCK, SMB2-LOCK-CONTEXT
BASE-LOCK-LOCK3 -> SMB2-LOCK-RANGE
BASE-LOCK-LOCK4 -> SMB2-LOCK-OVERLAP
BASE-LOCK-LOCK5 -> SMB2-LOCK-STACKING
BASE-LOCK-LOCK6 -> SMB2-LOCK-CANCEL, change_locktype no longer relevant
BASE-LOCK-LOCK7 -> SMB2-LOCK-RW-SHARED, SMB2-LOCK-RW-EXCLUSIVE

14 years agos4/libcli: rename previously reserved field in SMB2 LOCK struct
Steven Danneman [Thu, 19 Nov 2009 01:51:35 +0000 (17:51 -0800)]
s4/libcli: rename previously reserved field in SMB2 LOCK struct

The lock.in.reserved field has been renamed lock_sequence in the
SMB 2.1 dialect.  See MS-SMB 2.2.26.

14 years agos4/libcli: Initialize client PID for SMB2 connections
Steven Danneman [Thu, 19 Nov 2009 00:35:16 +0000 (16:35 -0800)]
s4/libcli: Initialize client PID for SMB2 connections

Set the SMB pid to the Unix pid of the client process.

14 years agos3-registry: fix REG_MULTI_SZ handling in registry_push_value.
Günther Deschner [Wed, 25 Nov 2009 20:34:55 +0000 (21:34 +0100)]
s3-registry: fix REG_MULTI_SZ handling in registry_push_value.

Catched by smbconftort test on the buildfarm.

Guenther

14 years agos3-kerberos: add check for prerequisite krb5/krb5.h header while checking for krb5...
Günther Deschner [Wed, 25 Nov 2009 20:33:48 +0000 (21:33 +0100)]
s3-kerberos: add check for prerequisite krb5/krb5.h header while checking for krb5/locate_plugin.h.

(Needed for new Heimdal versions).

Guenther

14 years agoFix bug 6892 - When a chown operation is issued via Windows Explorer, all ACLS are...
Jeremy Allison [Wed, 25 Nov 2009 18:20:38 +0000 (10:20 -0800)]
Fix bug 6892 - When a chown operation is issued via Windows Explorer, all ACLS are wiped out.
Merges existing DACLs when a ACL set operation comes in with only owner or group values set.
Jeremy.

14 years agoFix crash due to uninitialized pointer (not a problem in 3.4.x or below).
Jeremy Allison [Wed, 25 Nov 2009 17:58:45 +0000 (09:58 -0800)]
Fix crash due to uninitialized pointer (not a problem in 3.4.x or below).
Jeremy.

14 years agonsswitch: fix compile of winbind_krb5_locator with recent Heimdal versions.
Günther Deschner [Wed, 25 Nov 2009 14:21:54 +0000 (15:21 +0100)]
nsswitch: fix compile of winbind_krb5_locator with recent Heimdal versions.

Guenther

14 years agocifs.upcall: 2nd part of fix for Bug #6868: support building with Heimdal we well...
Günther Deschner [Wed, 25 Nov 2009 14:06:19 +0000 (15:06 +0100)]
cifs.upcall: 2nd part of fix for Bug #6868: support building with Heimdal we well as with MIT.

Guenther

14 years agos3 selftest: Add RAW-SFILEINFO-END-OF-FILE to selftest
Tim Prouty [Wed, 25 Nov 2009 16:15:31 +0000 (08:15 -0800)]
s3 selftest: Add RAW-SFILEINFO-END-OF-FILE to selftest

14 years agos3 setfileinfo: Open with FILE_WRITE_DATA when setting the file size
Tim Prouty [Wed, 25 Nov 2009 00:48:16 +0000 (16:48 -0800)]
s3 setfileinfo: Open with FILE_WRITE_DATA when setting the file size

This matches what is outlined here:
http://msdn.microsoft.com/en-us/library/ms804363.aspx

This is also inline with how winXP/win7 handle this.  See
RAW-SFILEINFO-END-OF-FILE* in smbtorture4.

14 years agos4 torture: Change RAW-OPLOCK to use the documented version of SET_END_OF_FILE_INFO
Tim Prouty [Wed, 25 Nov 2009 16:03:36 +0000 (08:03 -0800)]
s4 torture: Change RAW-OPLOCK to use the documented version of SET_END_OF_FILE_INFO

The passtrhough version of SET_END_OF_FILE_INFO is tested in
RAW-SFILEINFO-END-OF-FILE.

Additionally, the first opener is changed to use SHARE_WRITE for the
share mode since SET_END_OF_FILE_INFO actually writes data to the file
via truncating/extending.

14 years agos4 torture: Allow onefs to be checked like samba3 and samba4
Tim Prouty [Wed, 25 Nov 2009 16:17:13 +0000 (08:17 -0800)]
s4 torture: Allow onefs to be checked like samba3 and samba4

This is needed for the new RAW-SFILEINFO-END-OF-FILE test to
differentiate what appears to be a windows bug.

14 years agos4 torture: Add two new setfileinfo tests
Tim Prouty [Wed, 25 Nov 2009 00:29:19 +0000 (16:29 -0800)]
s4 torture: Add two new setfileinfo tests

These tests expose a potential bug in winXP, win7, and likely others.
There is also a bug in samba where share modes aren't being enforced
where they should.

For more details see:
http://lists.samba.org/archive/cifs-protocol/2009-November/001130.html

14 years agos4 torture: Split up the torture suite setup for RAW-SFILEINFO
Tim Prouty [Wed, 25 Nov 2009 00:25:02 +0000 (16:25 -0800)]
s4 torture: Split up the torture suite setup for RAW-SFILEINFO

A side effect of this change is that RAW-SFILEINFO now runs the whole
suite instead of just the first test.  I changed the name of the first
test to RAW-SFILEINFO-BASE and changed all of the selftest scripts
that call it.

14 years agoThe start of a WHATSNEW for Samba4 alpha9
Andrew Bartlett [Wed, 25 Nov 2009 08:17:50 +0000 (19:17 +1100)]
The start of a WHATSNEW for Samba4 alpha9

If you think what you did is missing, please edit.  Release of alpha9
will be this week.

Andrew Bartlett

14 years agos4:dsdb Make samba_dsdb easier to use in upgrades - assume default values
Andrew Bartlett [Wed, 25 Nov 2009 07:49:42 +0000 (18:49 +1100)]
s4:dsdb Make samba_dsdb easier to use in upgrades - assume default values

14 years agos4: update valgrind suppressions for use on build farm
Andrew Bartlett [Wed, 25 Nov 2009 06:48:29 +0000 (17:48 +1100)]
s4: update valgrind suppressions for use on build farm

14 years agolibcli: allow ntstatus.h to be used by openchange
Andrew Tridgell [Wed, 25 Nov 2009 04:30:20 +0000 (15:30 +1100)]
libcli: allow ntstatus.h to be used by openchange

apparently ntstatus.h is used by openchange, but they don't include
replace.h. This makes that possible again.

14 years agos4:upgradeschema.py Update to reflect recent change in provision
Matthieu Patou [Tue, 24 Nov 2009 15:52:36 +0000 (18:52 +0300)]
s4:upgradeschema.py Update to reflect recent change in provision

- define which modules we want to use when loading the ldb
- move partition in sam.ldb.d dir

Changes have been suggested by Andrew Bartlett.

(commit message clarified by Andrew Bartlett)
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
14 years agos4:upgradeschema.py Cleanup
Matthieu Patou [Tue, 24 Nov 2009 14:02:59 +0000 (17:02 +0300)]
s4:upgradeschema.py Cleanup

- remove some useless comments
- remove hardcoded paths

(commit message clarified by Andrew Bartlett)
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
14 years agoRemove call into reduce_name if case sensitive. This allows us to pass
Jeremy Allison [Wed, 25 Nov 2009 00:40:08 +0000 (16:40 -0800)]
Remove call into reduce_name if case sensitive. This allows us to pass
RAW-CHKPATH when case sensitive = yes, but isn't the correct way to do
it. I'm testing a larger patch to smbd/filename.c that should fix this
correctly, and will add a torture test to ensure RAW-CHKPATH is run against
a case sensitive share once this is done.
Jeremy.

14 years agos4:installmisc - install also the new "domainlevel" tool from the setup dir
Matthias Dieter Wallnöfer [Tue, 24 Nov 2009 21:24:00 +0000 (22:24 +0100)]
s4:installmisc - install also the new "domainlevel" tool from the setup dir

14 years agos4: Check SUN compiler and remove "-fullwarn" when using it
Brian Lu [Tue, 24 Nov 2009 21:10:22 +0000 (22:10 +0100)]
s4: Check SUN compiler and remove "-fullwarn" when using it

This option is only supported by "gcc".
For details consider https://bugzilla.samba.org/show_bug.cgi?id=6905

Signed-off-by: Matthias Dieter Wallnöfer <mwallnoefer@yahoo.de>
14 years agoAllow us to pass RAW-CHKPATH with FILE_FLAG_POSIX_SEMANTICS set or with
Jeremy Allison [Tue, 24 Nov 2009 22:05:56 +0000 (14:05 -0800)]
Allow us to pass RAW-CHKPATH with FILE_FLAG_POSIX_SEMANTICS set or with
wide links = no.
Jeremy.

14 years agoFix make test. Only do POSIX case conversion if FILE_FLAG_POSIX_SEMANTICS set. Doh !
Jeremy Allison [Tue, 24 Nov 2009 21:02:34 +0000 (13:02 -0800)]
Fix make test. Only do POSIX case conversion if FILE_FLAG_POSIX_SEMANTICS set. Doh !
Jeremy.

14 years agoRevert "s4-netlogon: always set the dNSHostName in GetDomainInfo"
Matthias Dieter Wallnöfer [Tue, 24 Nov 2009 19:30:07 +0000 (20:30 +0100)]
Revert "s4-netlogon: always set the dNSHostName in GetDomainInfo"

This reverts commit 87b6f2e863c6e117643ab6704e50167e849b69cc.

This was the cause of the breakage of the "LogonGetDomainInfo" testsuite. I
think my behaviour is more correct to Windows Server since the test works
against it (at least release 2003 R2).

One problem I discovered is that freshly joined workstations don't get their
DNS name into the directory. Therefore I think also another part (maybe another
RPC call) is able to do this.

14 years agos4:ldap.py - fix the schema update test on Windows Server
Matthias Dieter Wallnöfer [Tue, 24 Nov 2009 18:44:18 +0000 (19:44 +0100)]
s4:ldap.py - fix the schema update test on Windows Server

Apparently Windows Server (2003) doesn't like the comma delimiter here. I got
always error 16 ("LDB_NO_SUCH_ATTRIBUTE"). With this change the test works
again.

14 years agos4:libnet_become_dc - fix typo
Matthias Dieter Wallnöfer [Tue, 24 Nov 2009 18:17:08 +0000 (19:17 +0100)]
s4:libnet_become_dc - fix typo

14 years agos3: Always try SamLogonEx
Volker Lendecke [Tue, 24 Nov 2009 15:51:30 +0000 (16:51 +0100)]
s3: Always try SamLogonEx

Required for cluster systems working in a Samba domain. With NT4 this won't
work, but real NT4 DCs should not be around in environments that pay big bucks
for a cluster... And if they are, they can always install a Samba DC trusting
that NT4 domain.

14 years agos3-spoolss: re-arrange driver info level fillup functions.
Günther Deschner [Mon, 23 Nov 2009 14:31:58 +0000 (15:31 +0100)]
s3-spoolss: re-arrange driver info level fillup functions.

Guenther

14 years agos3-spoolss: fixes for _spoolss_EnumPrinterKey client and server.
Günther Deschner [Tue, 24 Nov 2009 14:22:04 +0000 (15:22 +0100)]
s3-spoolss: fixes for _spoolss_EnumPrinterKey client and server.

Thanks Metze for review!

Guenther

14 years agos3-registry: use push_reg_multi_sz() in registry_push_value().
Günther Deschner [Tue, 24 Nov 2009 14:26:32 +0000 (15:26 +0100)]
s3-registry: use push_reg_multi_sz() in registry_push_value().

Guenther

14 years agos3-registry: remove reg_pull_multi_sz().
Günther Deschner [Tue, 24 Nov 2009 14:26:08 +0000 (15:26 +0100)]
s3-registry: remove reg_pull_multi_sz().

Guenther

14 years agos3:torture: use timeval_current/timeval_elapsed instead of start_timer/end_timer
Stefan Metzmacher [Tue, 24 Nov 2009 09:59:09 +0000 (10:59 +0100)]
s3:torture: use timeval_current/timeval_elapsed instead of start_timer/end_timer

metze

14 years agos3:torture/nbio: use timeval_current/timeval_elapsed instead of start_timer/end_timer
Stefan Metzmacher [Tue, 24 Nov 2009 09:41:40 +0000 (10:41 +0100)]
s3:torture/nbio: use timeval_current/timeval_elapsed instead of start_timer/end_timer

metze

14 years agos3:dbwrap_torture: use timeval_current/timeval_elapsed instead of start_timer/end_timer
Stefan Metzmacher [Tue, 24 Nov 2009 09:36:52 +0000 (10:36 +0100)]
s3:dbwrap_torture: use timeval_current/timeval_elapsed instead of start_timer/end_timer

metze

14 years agoRAW-PING-PONG: use timeval_current/timeval_elapsed instead of start_timer/end_timer
Stefan Metzmacher [Tue, 24 Nov 2009 09:18:37 +0000 (10:18 +0100)]
RAW-PING-PONG: use timeval_current/timeval_elapsed instead of start_timer/end_timer

metze

14 years agos3:fix the merged build of bin/smbtorture4
Michael Adam [Tue, 24 Nov 2009 10:43:56 +0000 (11:43 +0100)]
s3:fix the merged build of bin/smbtorture4

Michael

14 years agodocs: Fix Bug 6922: Add Registry patchfile for Win7 domain join.
Günther Deschner [Tue, 24 Nov 2009 10:36:14 +0000 (11:36 +0100)]
docs: Fix Bug 6922: Add Registry patchfile for Win7 domain join.

Patchfile from SATOH Fumiyasu <fumiyas@osstech.co.jp>.

Thanks!

Guenther

14 years agoRestore NtCreateX case handling with FILE_FLAG_POSIX_SEMANTICS to the way
Jeremy Allison [Tue, 24 Nov 2009 05:06:54 +0000 (21:06 -0800)]
Restore NtCreateX case handling with FILE_FLAG_POSIX_SEMANTICS to the way
it worked in 3.3.x and 3.2.x. UCF_POSIX_PATHNAMES may no longer be needed,
as lp_posix_pathnames() being set isn't the same as case handling from
NtCreateX (lp_posix_pathnames() changes the STAT calls to LSTAT, not
the case semantics).
Jeremy.

14 years agos4:param remove reference to private header from source4/param/param.h
Brad Hards [Tue, 24 Nov 2009 02:43:59 +0000 (13:43 +1100)]
s4:param remove reference to private header from source4/param/param.h

source4/param/param.h has a

param.h is a public header (and parmlist isn't, even if the relative path
could work), so I suggest making it a forward declaration in the header, and
including parmlist.h in the implementation.

(commit message included from e-mail by Andrew Bartlett)
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
14 years agos4: Fix missing TLS dir when targetdir is not the default one
Matthieu Patou [Sun, 22 Nov 2009 17:51:24 +0000 (20:51 +0300)]
s4: Fix missing TLS dir when targetdir is not the default one