12 years agos3:smbd: cancel all locks that are made before the first failure
Stefan Metzmacher [Tue, 7 Jul 2009 13:16:21 +0000 (15:16 +0200)]
s3:smbd: cancel all locks that are made before the first failure

s3:smbd: cancel all locks that are made before the first failure

We never reached the cleanup code and directly returned the error.


12 years agos4:auth It is easier to copy the session key than get talloc right.
Andrew Bartlett [Tue, 7 Jul 2009 03:27:47 +0000 (13:27 +1000)]
s4:auth It is easier to copy the session key than get talloc right.

The session keys as supplied already have a reference on them, so
stealing them creates challenges.  For 16 bytes, it is just easier to
be consistant and copy them.

Andrew Bartlett

12 years agos3: Update streaminfo implementations now that only base names are passed through...
Tim Prouty [Mon, 6 Jul 2009 22:26:57 +0000 (15:26 -0700)]
s3: Update streaminfo implementations now that only base names are passed through the vfs

12 years agos3 smb2: Fix unused variable warning
Tim Prouty [Mon, 6 Jul 2009 21:18:05 +0000 (14:18 -0700)]
s3 smb2: Fix unused variable warning

12 years agos3 audit: Fix a memory leak
Tim Prouty [Sat, 4 Jul 2009 00:15:30 +0000 (17:15 -0700)]
s3 audit: Fix a memory leak

12 years agos3: Plumb smb_filename through SMB_VFS_NTIMES
Tim Prouty [Thu, 2 Jul 2009 20:39:20 +0000 (13:39 -0700)]
s3: Plumb smb_filename through SMB_VFS_NTIMES

12 years agos3 catia: Remove pointless VOPs from catia
Tim Prouty [Thu, 2 Jul 2009 20:13:14 +0000 (13:13 -0700)]
s3 catia: Remove pointless VOPs from catia

12 years agos3 sticky write time: Removed unused args and tighten up a function by making an...
Tim Prouty [Thu, 2 Jul 2009 17:27:01 +0000 (10:27 -0700)]
s3 sticky write time: Removed unused args and tighten up a function by making an arg const

12 years agos3: Plumb smb_filename around SMB_VFS_CHFLAGS
Tim Prouty [Mon, 6 Jul 2009 21:36:16 +0000 (14:36 -0700)]
s3: Plumb smb_filename around SMB_VFS_CHFLAGS

SMB_VFS_CHFLAGS isn't actually getting the smb_filename struct for now
since it only operates on the basefile.  This is the strategy for all
path-based operations that will never actually operate on a stream.

By clarifying the meaning of path based operations that don't take an
smb_filename struct, modules that implement streams such as vfs_onefs
no longer need to implement SMB_VFS_CHFLAGS to ensure it's only called
on the base_name.

12 years agos3: Change some filename_convert to pass in NULL for char **fname argument
Tim Prouty [Wed, 1 Jul 2009 23:15:52 +0000 (16:15 -0700)]
s3: Change some filename_convert to pass in NULL for char **fname argument

12 years agos3: Add ability to pass NULL to filename_convert for the char *fname
Tim Prouty [Wed, 1 Jul 2009 23:14:40 +0000 (16:14 -0700)]
s3: Add ability to pass NULL to filename_convert for the char *fname

The goal is to eventually remove the need for the char **fname argument
once all callers have standardized on smb_filename.

12 years agos3: Plumb smb_filename through SMB_VFS_UNLINK
Tim Prouty [Thu, 2 Jul 2009 16:27:44 +0000 (09:27 -0700)]
s3: Plumb smb_filename through SMB_VFS_UNLINK

12 years agos3 audit: Remove some recently introduced memory leaks in the audit modules
Tim Prouty [Wed, 1 Jul 2009 06:08:02 +0000 (23:08 -0700)]
s3 audit: Remove some recently introduced memory leaks in the audit modules

12 years agos3 audit: Remove the usesless static function declarations from the audit modules
Tim Prouty [Wed, 1 Jul 2009 05:44:39 +0000 (22:44 -0700)]
s3 audit: Remove the usesless static function declarations from the audit modules

12 years agos3 onefs: Fix failure in POSIX smbtorture test
Tim Prouty [Tue, 30 Jun 2009 23:41:48 +0000 (16:41 -0700)]
s3 onefs: Fix failure in POSIX smbtorture test

The return value of readlink was a bool instead of an int,
which caused the length of the returned value to never be

12 years agos3: Plumb smb_filename through SMB_VFS_RENAME
Tim Prouty [Wed, 1 Jul 2009 00:04:38 +0000 (17:04 -0700)]
s3: Plumb smb_filename through SMB_VFS_RENAME

12 years agoFix some typos pointed out by Oota Toshiya.
Jelmer Vernooij [Mon, 6 Jul 2009 20:41:21 +0000 (22:41 +0200)]
Fix some typos pointed out by Oota Toshiya.

12 years agoRAW-LOCK: fix a compiler warning and make a global static const
Stefan Metzmacher [Mon, 6 Jul 2009 13:48:32 +0000 (15:48 +0200)]
RAW-LOCK: fix a compiler warning and make a global static const


12 years agoRAW-LOCK: fix the run against w2k3, w2k8, win7rc and samba3/4
Stefan Metzmacher [Mon, 6 Jul 2009 13:36:03 +0000 (15:36 +0200)]
RAW-LOCK: fix the run against w2k3, w2k8, win7rc and samba3/4

We now allow --target=w2k8.

The strange thing is that w2k8 and win7rc only do half of the timeout
that was given by the client.


12 years agoSMB2-NOTIFY: windows7 uses a different maximum buffer size
Stefan Metzmacher [Mon, 6 Jul 2009 10:03:33 +0000 (12:03 +0200)]
SMB2-NOTIFY: windows7 uses a different maximum buffer size


12 years agos4:heimdal: readd heimdal/lib/asn1/asn1parse.y which was parse.y before the last...
Stefan Metzmacher [Mon, 6 Jul 2009 11:26:09 +0000 (13:26 +0200)]
s4:heimdal: readd heimdal/lib/asn1/asn1parse.y which was parse.y before the last import

Also commit the regenerated files for systems without yacc and lex.

This fixes the build with automatic dependecies for me.


12 years agodoc: clarify that mount.cifs never uses smb.conf
Suresh Jayaraman [Mon, 6 Jul 2009 10:31:30 +0000 (06:31 -0400)]
doc: clarify that mount.cifs never uses smb.conf

For some users who are used to smbclient, it is not clear that
mount.cifs never uses smb.conf for configuring client specific
parameters. So, let's add this information to mount.cifs man page.

Signed-off-by: Suresh Jayaraman <>
12 years agodo not merge ACEs with different SMB_ACE4_INHERIT_ONLY_ACE flag, this leads to wrong...
Christian Ambach [Sun, 5 Jul 2009 14:03:15 +0000 (16:03 +0200)]
do not merge ACEs with different SMB_ACE4_INHERIT_ONLY_ACE flag, this leads to wrong inheritance flags in the ACL e.g. (on GPFS) user:10000036:rwxc:allow (X)READ/LIST (X)WRITE/CREATE (X)MKDIR (X)SYNCHRONIZE (X)READ_ACL (X)READ_ATTR (X)READ_NAMED (X)DELETE (X)DELETE_CHILD (X)CHOWN (X)EXEC/SEARCH (X)WRITE_ACL (X)WRITE_ATTR (X)WRITE_NAMED




would be merged to



so the explicit right for the user on the parent directory will be gone (the InheritOnly flag only accounts to subdirectories)
thus leaving the user without access to the directory itself

Signed-off-by: Christian Ambach <>
12 years agos4:net Move net_vampire() to net_vampire.c
Andrew Bartlett [Mon, 6 Jul 2009 06:41:42 +0000 (16:41 +1000)]
s4:net Move net_vampire() to net_vampire.c

12 years agos4:ldb Rework use of talloc and ldif objects in python wrapper
Andrew Bartlett [Sun, 5 Jul 2009 23:31:38 +0000 (09:31 +1000)]
s4:ldb Rework use of talloc and ldif objects in python wrapper

The talloc hirarchy here was a bit odd - we would both steal the
parsed ldif onto 'NULL', then reference it onto a python talloc

Now we just leave the reference, after we complete building the object.

Andrew Bartlett

12 years agos4:ldb Fix talloc hirarchy in LDIF parsing code
Andrew Bartlett [Sun, 5 Jul 2009 23:24:18 +0000 (09:24 +1000)]
s4:ldb Fix talloc hirarchy in LDIF parsing code

The problem here was that some parts of the ldb_message were still
attached to the ldb_ldif structure, and when only the message was
taken (and the ldif free'ed to reclaim memory) we refereced free'ed

Andrew Bartlett

12 years agoUse null_ndr_syntax_id instead of zeroing null_interface manually
Volker Lendecke [Sun, 5 Jul 2009 08:23:56 +0000 (10:23 +0200)]
Use null_ndr_syntax_id instead of zeroing null_interface manually

12 years agoAdd const to init_rpc_hdr_ba
Volker Lendecke [Sun, 5 Jul 2009 08:21:39 +0000 (10:21 +0200)]
Add const to init_rpc_hdr_ba

12 years agoRemove "typedef struct ndr_syntax_id RPC_IFACE;"
Volker Lendecke [Sun, 5 Jul 2009 08:13:03 +0000 (10:13 +0200)]
Remove "typedef struct ndr_syntax_id RPC_IFACE;"

12 years agoIntroduce null_ndr_syntax_id
Volker Lendecke [Sun, 5 Jul 2009 08:21:06 +0000 (10:21 +0200)]
Introduce null_ndr_syntax_id

12 years agoMake check_bind_req static to rpc_server/srv_pipe.c
Volker Lendecke [Sun, 5 Jul 2009 07:43:23 +0000 (09:43 +0200)]
Make check_bind_req static to rpc_server/srv_pipe.c

12 years agoActually increase PASSDB_INTERFACE_VERSION to 18 :-)
Volker Lendecke [Sun, 5 Jul 2009 14:20:05 +0000 (16:20 +0200)]
Actually increase PASSDB_INTERFACE_VERSION to 18 :-)

12 years agoFix some nonempty blank lines
Volker Lendecke [Sun, 5 Jul 2009 21:15:34 +0000 (23:15 +0200)]
Fix some nonempty blank lines

12 years agolibreplace: update library search path variables
Björn Jacke [Sun, 5 Jul 2009 20:42:29 +0000 (22:42 +0200)]
libreplace: update library search path variables

12 years agos3:configure: honor CFLAGS in compile tests
Björn Jacke [Sun, 5 Jul 2009 09:28:05 +0000 (11:28 +0200)]
s3:configure: honor CFLAGS in compile tests

When compilation is done for the architecture that's not the compiler's
default, for example by setting CFLAGS to -m64, then compile tests might fail
when they are done with the comiler's default arch. This should fix
bugzilla #6162.

12 years agopdb_ads: For cached samu entries, priv is NULL
Volker Lendecke [Sat, 4 Jul 2009 10:26:08 +0000 (12:26 +0200)]
pdb_ads: For cached samu entries, priv is NULL

We have to recreate the priv entry on demand. This needs fixing in
passdb... :-)

12 years agoAdd pdb_ads_get_domain_info
Volker Lendecke [Sat, 4 Jul 2009 09:09:42 +0000 (11:09 +0200)]
Add pdb_ads_get_domain_info

12 years agoFix some warnings
Volker Lendecke [Sat, 4 Jul 2009 08:35:21 +0000 (10:35 +0200)]
Fix some warnings

12 years agopdb_ads: Use tldap_fetch_rootdse in pdb_ads_connect
Volker Lendecke [Tue, 30 Jun 2009 08:39:04 +0000 (10:39 +0200)]
pdb_ads: Use tldap_fetch_rootdse in pdb_ads_connect

12 years agopdb_ads: Fetch the domain GUID
Volker Lendecke [Mon, 29 Jun 2009 20:28:19 +0000 (22:28 +0200)]
pdb_ads: Fetch the domain GUID

12 years agoAdd tldap_pull_guid
Volker Lendecke [Mon, 29 Jun 2009 20:11:23 +0000 (22:11 +0200)]
Add tldap_pull_guid

12 years agoHandle LSA_POLICY_INFO_DNS
Volker Lendecke [Sun, 28 Jun 2009 21:07:00 +0000 (23:07 +0200)]

12 years agoImplement QueryInfoPolicy2 similar to s4: Make it the same as QueryInfoPolicy
Volker Lendecke [Sun, 28 Jun 2009 20:29:06 +0000 (22:29 +0200)]
Implement QueryInfoPolicy2 similar to s4: Make it the same as QueryInfoPolicy

Don't reply to it for non-pdb-ads to keep up our old behaviour

12 years agoAdd pdb_get_domain_info
Volker Lendecke [Sat, 4 Jul 2009 09:12:33 +0000 (11:12 +0200)]
Add pdb_get_domain_info

12 years agoMake pdb_ads return an additional flag
Volker Lendecke [Sun, 28 Jun 2009 15:43:48 +0000 (17:43 +0200)]
Make pdb_ads return an additional flag

12 years agos3/s4:build: honor library path from environment
Björn Jacke [Fri, 3 Jul 2009 23:07:15 +0000 (01:07 +0200)]
s3/s4:build: honor library path from environment

we may not throw away LD_LIBRARY_PATH and friends as they sometimes contain
important information. Some of the build farm machines for example have the
python lib path in there - we can't run make test when we suddenly forget how
to use python

12 years agotldap: Do not do a return before we got the search_result
Volker Lendecke [Tue, 30 Jun 2009 18:06:55 +0000 (20:06 +0200)]
tldap: Do not do a return before we got the search_result

12 years agoFix two pointless statics
Volker Lendecke [Thu, 2 Jul 2009 09:06:21 +0000 (11:06 +0200)]
Fix two pointless statics

12 years agoFor non-existent streams we have to return OBJECT_NAME_NOT_FOUND
Volker Lendecke [Fri, 3 Jul 2009 20:36:11 +0000 (22:36 +0200)]
For non-existent streams we have to return OBJECT_NAME_NOT_FOUND

See the STREAMERROR s3 torture test.

Jeremy, Tim, please check!

12 years agoTest error codes on popular streams
Volker Lendecke [Thu, 2 Jul 2009 09:56:33 +0000 (11:56 +0200)]
Test error codes on popular streams

12 years agoAttempt to fix the build
Volker Lendecke [Fri, 3 Jul 2009 20:32:31 +0000 (22:32 +0200)]
Attempt to fix the build

12 years agos3-spoolss: add rpccli_spoolss_getprinterdriver convenience wrapper.
Günther Deschner [Fri, 3 Jul 2009 16:39:58 +0000 (18:39 +0200)]
s3-spoolss: add rpccli_spoolss_getprinterdriver convenience wrapper.


12 years agospoolss: fill in spoolss_GetPrinterDriver IDL.
Günther Deschner [Fri, 3 Jul 2009 16:39:38 +0000 (18:39 +0200)]
spoolss: fill in spoolss_GetPrinterDriver IDL.


12 years agospoolss: the flags in spoolss_SetPrinterInfo1 are in fact spoolss_PrinterAttributes.
Günther Deschner [Fri, 3 Jul 2009 12:08:05 +0000 (14:08 +0200)]
spoolss: the flags in spoolss_SetPrinterInfo1 are in fact spoolss_PrinterAttributes.


12 years agos3-spoolss: implement _spoolss_AddPrinter.
Günther Deschner [Fri, 3 Jul 2009 09:05:38 +0000 (11:05 +0200)]
s3-spoolss: implement _spoolss_AddPrinter.


12 years agospoolss: fill in spoolss_AddPrinter IDL.
Günther Deschner [Fri, 3 Jul 2009 08:10:53 +0000 (10:10 +0200)]
spoolss: fill in spoolss_AddPrinter IDL.


12 years agoheimdal: don't include <ifaddrs.h> without knowing it's there
Björn Jacke [Fri, 3 Jul 2009 17:13:08 +0000 (19:13 +0200)]
heimdal: don't include <ifaddrs.h> without knowing it's there

this is 73dbbe0d54 re-added. abartlet, please pick this to lorikeet.

12 years agotdb: fix define of tdbdir when srcdir != "."
Björn Jacke [Fri, 3 Jul 2009 16:22:16 +0000 (18:22 +0200)]
tdb: fix define of tdbdir when srcdir != "."

12 years agoSort the signature files
Simo Sorce [Fri, 3 Jul 2009 12:42:23 +0000 (08:42 -0400)]
Sort the signature files

12 years agoExpose functions need by backend writers
Simo Sorce [Fri, 19 Jun 2009 00:06:00 +0000 (20:06 -0400)]
Expose functions need by backend writers

move publicly needed structures and functions in the public header.
Stop installing internal headers.
Update the signature and exports files with the new exposed

12 years agoRestore ABI compatibility for talloc.
Simo Sorce [Thu, 2 Jul 2009 13:29:20 +0000 (09:29 -0400)]
Restore ABI compatibility for talloc.

12 years agos3:configure: fix syntax error in avahi configure test
Björn Jacke [Fri, 3 Jul 2009 12:25:06 +0000 (14:25 +0200)]
s3:configure: fix syntax error in avahi configure test

12 years agos4: Don't install epdump, since it no longer exists.
Jelmer Vernooij [Fri, 3 Jul 2009 00:59:40 +0000 (02:59 +0200)]
s4: Don't install epdump, since it no longer exists.

12 years agoRevert "[s3]pidfile_create: use check is_default_dyn_CONFIGFILE() since we have it."
Michael Adam [Fri, 3 Jul 2009 09:35:11 +0000 (11:35 +0200)]
Revert "[s3]pidfile_create: use check is_default_dyn_CONFIGFILE() since we have it."

This reverts commit 1d3dcd1e420104c23769b691d0b8b7958d5c58a7.

This fixes Bug #6531 (Pid File creation.)

Usually, I would extend is_default_dyn_XYZ() so that it returns true
when dyn_XYZ has not been set yet or when its value is equal to the
compiled in default value XZY. But this would have a change in effect
in popt_common and torture.c: is_default_dyn_CONFIGFILE() is used
there to check whether the config file should be overwritten by the
contents of the environment variable SMB_CONF_PATH. Currently this
is only done when set_dyn_CONFIGFILE() had _not_ previously been called
at all, not even with the same value as the compiled in default.


12 years agodo not redefine _INT32
Matt Kraai [Wed, 1 Jul 2009 17:49:35 +0000 (10:49 -0700)]
do not redefine _INT32

12 years agos3:configure: define _LARGEFILE64_SOURCE on QNX
Matt Kraai [Wed, 1 Jul 2009 19:44:03 +0000 (12:44 -0700)]
s3:configure: define _LARGEFILE64_SOURCE on QNX

12 years agos3:pam_smbpass: fix build error on some Linux distros
Björn Jacke [Wed, 10 Jun 2009 10:31:07 +0000 (12:31 +0200)]
s3:pam_smbpass: fix build error on some Linux distros

This addresses bug #6459.

12 years agos4 docs: Git has revision SHA's, not numbers.
Jelmer Vernooij [Wed, 1 Jul 2009 12:49:18 +0000 (14:49 +0200)]
s4 docs: Git has revision SHA's, not numbers.

12 years agos4: Remove stub endpoint mapper script.
Jelmer Vernooij [Wed, 1 Jul 2009 14:54:02 +0000 (16:54 +0200)]
s4: Remove stub endpoint mapper script.

12 years agoTOSHARG: Fix typo: s/aquire/acquire/
Jelmer Vernooij [Fri, 19 Jun 2009 14:48:19 +0000 (16:48 +0200)]
TOSHARG: Fix typo: s/aquire/acquire/

12 years agopidl: Don't run compile tests inside the tree, won't work yet.
Jelmer Vernooij [Fri, 19 Jun 2009 14:17:08 +0000 (16:17 +0200)]
pidl: Don't run compile tests inside the tree, won't work yet.

12 years agos3:smbd: add marshalling layer for SMB2 Lock support
Stefan Metzmacher [Thu, 2 Jul 2009 17:26:05 +0000 (19:26 +0200)]
s3:smbd: add marshalling layer for SMB2 Lock support


12 years agos4:libcli: move SMB2 lock flags to smb2_constants.h
Stefan Metzmacher [Thu, 2 Jul 2009 15:47:33 +0000 (17:47 +0200)]
s4:libcli: move SMB2 lock flags to smb2_constants.h


12 years agos3:smbd: add marshalling layer for SMB2 Find (QueryDirectory) support
Stefan Metzmacher [Thu, 2 Jul 2009 10:28:58 +0000 (12:28 +0200)]
s3:smbd: add marshalling layer for SMB2 Find (QueryDirectory) support


12 years agos3:smbd: add marshalling layer for SMB2 Break support (SMB 2.002 only)
Stefan Metzmacher [Thu, 2 Jul 2009 09:56:23 +0000 (11:56 +0200)]
s3:smbd: add marshalling layer for SMB2 Break support (SMB 2.002 only)


12 years agos3:smbd: add marshalling layer for SMB2 SetInfo support
Stefan Metzmacher [Thu, 2 Jul 2009 09:40:07 +0000 (11:40 +0200)]
s3:smbd: add marshalling layer for SMB2 SetInfo support


12 years agos3:smbd: add marshalling layer for SMB2 GetInfo support
Stefan Metzmacher [Thu, 2 Jul 2009 08:17:30 +0000 (10:17 +0200)]
s3:smbd: add marshalling layer for SMB2 GetInfo support


12 years agospoolss: the flags in spoolss_PrinterInfo1 are in fact spoolss_PrinterAttributes.
Günther Deschner [Thu, 2 Jul 2009 17:39:36 +0000 (19:39 +0200)]
spoolss: the flags in spoolss_PrinterInfo1 are in fact spoolss_PrinterAttributes.


12 years agos4-smbtorture: fix getjob test in RPC-SPOOLSS.
Günther Deschner [Wed, 1 Jul 2009 20:52:49 +0000 (22:52 +0200)]
s4-smbtorture: fix getjob test in RPC-SPOOLSS.


12 years agodoserrors: print out some more werrors.
Günther Deschner [Wed, 1 Jul 2009 20:52:05 +0000 (22:52 +0200)]
doserrors: print out some more werrors.


12 years agotalloc: change TALLOC_MAGIC for version 2.0.0
Stefan Metzmacher [Thu, 2 Jul 2009 08:48:24 +0000 (10:48 +0200)]
talloc: change TALLOC_MAGIC for version 2.0.0


12 years agoshow attribute values in sorted order to make comparison easier
Andrew Tridgell [Thu, 2 Jul 2009 06:19:37 +0000 (16:19 +1000)]
show attribute values in sorted order to make comparison easier

12 years agos3:smbd: add support for SMB2 Notify
Stefan Metzmacher [Mon, 29 Jun 2009 20:31:13 +0000 (22:31 +0200)]
s3:smbd: add support for SMB2 Notify


12 years agos3:smbd: make change_notify_reply() more generic and hide send_nt_replies() behind...
Stefan Metzmacher [Tue, 30 Jun 2009 13:01:00 +0000 (15:01 +0200)]
s3:smbd: make change_notify_reply() more generic and hide send_nt_replies() behind a callback

This prepares change notify support for SMB2.


12 years agos3:smbd: simplify the notify code a bit and always reply via change_notify_reply...
Stefan Metzmacher [Mon, 29 Jun 2009 20:17:17 +0000 (22:17 +0200)]
s3:smbd: simplify the notify code a bit and always reply via change_notify_reply() -> send_nt_replies()


12 years agos3:smbd: push nttrans and trans2 responses with no data to the client
Stefan Metzmacher [Mon, 29 Jun 2009 13:38:48 +0000 (15:38 +0200)]
s3:smbd: push nttrans and trans2 responses with no data to the client

For sync replies it's not a problem, as construct_reply() will send
the response, but for async replies we would not send the reply to the client.
Currently the notify code works arround this manually, so I assume
we didn't have a bug here. But the next commits will simplify
the notify code.


12 years agos3:smbd: restore the 3.0.x bahavior of send_nt_replies()
Stefan Metzmacher [Wed, 1 Jul 2009 10:56:30 +0000 (12:56 +0200)]
s3:smbd: restore the 3.0.x bahavior of send_nt_replies()

We should also set the error code, when we return no parameters or data.


12 years agos3:smbd: move global notify_changes_by_mid to smbd_server_connection
Stefan Metzmacher [Tue, 30 Jun 2009 08:17:38 +0000 (10:17 +0200)]
s3:smbd: move global notify_changes_by_mid to smbd_server_connection


12 years agos3:smbd: allow the offset to the path being 0 in SMB2 create
Stefan Metzmacher [Tue, 30 Jun 2009 11:25:21 +0000 (13:25 +0200)]
s3:smbd: allow the offset to the path being 0 in SMB2 create


12 years agowe can't use the unique index code for samAccountName
Andrew Tridgell [Thu, 2 Jul 2009 05:57:30 +0000 (15:57 +1000)]
we can't use the unique index code for samAccountName

Using ldb unique indexes for samAccountName doesn't work with DRS as
the other DC may send us a deleted record (tombstone record), which
has the same samAccountName as an existing record. That would then
create two records in the same partition with the same samAccountName.

So we needed to put back the logic in samldb.c which explicitly
checked whether a samAccountName already exists on add

12 years agodecrypt all objects in a DRS record, not just the first one
Andrew Tridgell [Thu, 2 Jul 2009 05:33:01 +0000 (15:33 +1000)]
decrypt all objects in a DRS record, not just the first one

We found this as an object came across from w2k3 with zero values,
which caused a segv when we tried to decrypt the first value

12 years agochange talloc to 2.0.0
Andrew Tridgell [Thu, 2 Jul 2009 05:19:40 +0000 (15:19 +1000)]
change talloc to 2.0.0

This is needed to prevent samba3 and samba4 from using an ABI
incompatible system version of talloc

See ongoing discussion on the samba-technical mailing list

12 years agoChanged ldb.ERR_NO_SUCH_OBJECT to LDB_ERR_NO_SUCH_OBJECT.
Andrew Tridgell [Thu, 2 Jul 2009 04:52:25 +0000 (14:52 +1000)]

The LDB_ERR_NO_SUCH_OBJECT varient is not a defined variable. This
should improve error handling in our python code on some
systems. Unfortunately it still doesn't work on mine. I need to trap
Jelmer somewhere where he can't escape some day and force him to
divulge the deep druid secrets of python exception handling ....

12 years agofixed the pull of drs schema elements
Andrew Tridgell [Thu, 2 Jul 2009 04:49:40 +0000 (14:49 +1000)]
fixed the pull of drs schema elements

The previous code incorrectly assumed that attributes such as
subClassOf come over the wire as strings. In fact they come over as 32
bit integers which refer to goversIDs. We have to post-process these
as it sometimes happens that a governsID comes over the wire before
the record that defines what it means.

12 years agothe settings structure needs to be initialised
Andrew Tridgell [Thu, 2 Jul 2009 04:47:06 +0000 (14:47 +1000)]
the settings structure needs to be initialised

12 years agoLDB_ERR_INVALID_DN_SYNTAX doesn't exist ...
Andrew Tridgell [Thu, 2 Jul 2009 04:44:48 +0000 (14:44 +1000)]
LDB_ERR_INVALID_DN_SYNTAX doesn't exist ...

The correct name is ldb.ERR_INVALID_DN_SYNTAX

12 years agos4:param use talloc_unlink() to free iconv context holding references
Andrew Bartlett [Thu, 2 Jul 2009 04:34:07 +0000 (14:34 +1000)]
s4:param use talloc_unlink() to free iconv context holding references

12 years ago[SAMBA 4 directory] Changes "forceLogoff" and corrects the "subRefs"
Matthias Dieter Wallnöfer [Tue, 30 Jun 2009 11:44:45 +0000 (13:44 +0200)]
[SAMBA 4 directory] Changes "forceLogoff" and corrects the "subRefs"

- This changes the attribute "forceLogoff" to its' default
  values according to Windows Server 2003 R2
- Also this corrects the "subRefs" attribute of the base-DN which only refers
  to direct child partitions (and therefore not to the complete transitive closure)

12 years agoEnsure we don't use delayed writes on POSIX opened files.
Jeremy Allison [Thu, 2 Jul 2009 00:39:10 +0000 (17:39 -0700)]
Ensure we don't use delayed writes on POSIX opened files.
Don't remove pending writetime changes if no time changes
are sent in UNIX_BASIC infolevel.

12 years agoFix bug #6520 time stamps - e.g. last mod time is not preserved when "unix extensions...
Jeremy Allison [Wed, 1 Jul 2009 23:58:14 +0000 (16:58 -0700)]
Fix bug #6520 time stamps - e.g. last mod time is not preserved when "unix extensions=yes" are set - and using latest cifs vfs client
Cancel out any pending "sticky" writes or "last write" changes when
doing a UNIX info level set.