samba.git
14 years agos4:ntlmssp Ensure that we always negotiate signing if we negotiate sealing.
Andrew Bartlett [Wed, 24 Mar 2010 05:09:02 +0000 (16:09 +1100)]
s4:ntlmssp Ensure that we always negotiate signing if we negotiate sealing.

Without this, a sealed LDAP connection to windows does not work.

Andrew Bartlett

14 years agos3: Optimize gencache for smbd exit
Volker Lendecke [Tue, 23 Mar 2010 17:36:55 +0000 (18:36 +0100)]
s3: Optimize gencache for smbd exit

If thousands of smbds try to gencache_stabilize at the same time because the
network died, all of them might be sitting in transaction_start. Don't do the
stabilize transaction if nothing has changed in gencache_notrans.tdb.

Volker

14 years agolibreplace: fixed declaration of dprintf() on FreeBSD
Andrew Tridgell [Tue, 23 Mar 2010 18:06:25 +0000 (05:06 +1100)]
libreplace: fixed declaration of dprintf() on FreeBSD

14 years agos4-provision: FreeBSD uses 'staff' for users
Andrew Tridgell [Wed, 24 Mar 2010 05:52:37 +0000 (16:52 +1100)]
s4-provision: FreeBSD uses 'staff' for users

14 years agos4-provision: solaris uses the group "other"
Andrew Tridgell [Tue, 23 Mar 2010 15:02:32 +0000 (11:02 -0400)]
s4-provision: solaris uses the group "other"

14 years agobuild: fixed nss_wrapper on solaris
Andrew Tridgell [Mon, 22 Mar 2010 07:40:44 +0000 (18:40 +1100)]
build: fixed nss_wrapper on solaris

solaris has different getpwnam_r syntax

14 years agoidl: fix forest trust information idl
Simo Sorce [Tue, 23 Mar 2010 22:45:20 +0000 (18:45 -0400)]
idl: fix forest trust information idl

--validate now passes

14 years agos4-smbtorture: remove tailing whitespace in samba3rpc testsuite.
Günther Deschner [Tue, 23 Mar 2010 17:32:06 +0000 (18:32 +0100)]
s4-smbtorture: remove tailing whitespace in samba3rpc testsuite.

Guenther

14 years agos4:librpc/rpc/pyrpc.h: librpc/rpc/dcerpc.h isn't needed anymore
Stefan Metzmacher [Thu, 18 Mar 2010 17:55:49 +0000 (18:55 +0100)]
s4:librpc/rpc/pyrpc.h: librpc/rpc/dcerpc.h isn't needed anymore

metze

14 years agos4:lib/messaging: pymessaging needs libndr.h
Stefan Metzmacher [Thu, 18 Mar 2010 18:36:50 +0000 (19:36 +0100)]
s4:lib/messaging: pymessaging needs libndr.h

metze

14 years agopidl:Samba4/Python: librpc/rpc/dcerpc.h isn't needed anymore
Stefan Metzmacher [Thu, 18 Mar 2010 17:55:00 +0000 (18:55 +0100)]
pidl:Samba4/Python: librpc/rpc/dcerpc.h isn't needed anymore

metze

14 years agos4:libnet: don't include librpc/rpc/dcerpc.h directly
Stefan Metzmacher [Fri, 19 Mar 2010 12:02:26 +0000 (13:02 +0100)]
s4:libnet: don't include librpc/rpc/dcerpc.h directly

metze

14 years agos3:librpc/rpc: move prototypes to dcerpc.h
Stefan Metzmacher [Fri, 19 Mar 2010 18:16:35 +0000 (19:16 +0100)]
s3:librpc/rpc: move prototypes to dcerpc.h

metze

14 years agos4-smbtorture: use torture_context in samsync test.
Günther Deschner [Tue, 23 Mar 2010 16:28:55 +0000 (17:28 +0100)]
s4-smbtorture: use torture_context in samsync test.

Guenther

14 years agos4-smbtorture: remove trailing whitespace from samsync torture test.
Günther Deschner [Tue, 23 Mar 2010 16:16:23 +0000 (17:16 +0100)]
s4-smbtorture: remove trailing whitespace from samsync torture test.

Guenther

14 years agos4-smbtorture: use torture_context in libnet domain tests as well.
Günther Deschner [Tue, 23 Mar 2010 15:30:05 +0000 (16:30 +0100)]
s4-smbtorture: use torture_context in libnet domain tests as well.

Guenther

14 years agos4:WINREG torture - fix an error (missing "break" in a "switch")
Matthias Dieter Wallnöfer [Tue, 23 Mar 2010 15:43:41 +0000 (16:43 +0100)]
s4:WINREG torture - fix an error (missing "break" in a "switch")

Sorry my fault - I didn't discover that earlier since we had binary data types
support activated in s4.

14 years agos4:registry - "reg_ldb_pack_value" - provide workarounds when the server receives...
Matthias Dieter Wallnöfer [Tue, 23 Mar 2010 15:19:26 +0000 (16:19 +0100)]
s4:registry - "reg_ldb_pack_value" - provide workarounds when the server receives non-standard data.

For now we reset/delete the "data" attribute. Anyway there is the need to find
a better solution (we probably want to change the format and save all data as
we got it like Windows itself does).

These workarounds are needed since for example the Windows 2000 Registry Editor
initialises empty REG_SZ strings with content '\0' and length 1 (not a valid
UTF16 sequence - "convert_string_talloc" breaks). So we simply reset/delete the
"data" attribute which works (no content).

14 years agos4:registry - "LDB backend" - fix indentation
Matthias Dieter Wallnöfer [Tue, 23 Mar 2010 15:17:25 +0000 (16:17 +0100)]
s4:registry - "LDB backend" - fix indentation

14 years agos4-smbtorture: use torture_context for debugging output everywhere in libnet torture...
Günther Deschner [Tue, 23 Mar 2010 14:31:27 +0000 (15:31 +0100)]
s4-smbtorture: use torture_context for debugging output everywhere in libnet torture tests.

Mimir, please check.

Guenther

14 years agos4-smbtorture: remove trailing whitespace in libnet torture tests.
Günther Deschner [Tue, 23 Mar 2010 15:04:30 +0000 (16:04 +0100)]
s4-smbtorture: remove trailing whitespace in libnet torture tests.

Guenther

14 years agos3-builtin: Add missing builtin groups.
Karolin Seeger [Tue, 23 Mar 2010 13:53:49 +0000 (14:53 +0100)]
s3-builtin: Add missing builtin groups.

Karolin

14 years agos3: Re-run make samba3-idl.
Karolin Seeger [Tue, 23 Mar 2010 14:24:25 +0000 (15:24 +0100)]
s3: Re-run make samba3-idl.

Karolin

14 years agosecurity.idl: Add missing builtin groups.
Karolin Seeger [Tue, 23 Mar 2010 13:26:51 +0000 (14:26 +0100)]
security.idl: Add missing builtin groups.

Karolin

14 years agolibrpc/rpc: fix dcerpc_errstr() to handle all nca_s_* status codes and fallback to...
Stefan Metzmacher [Sat, 20 Mar 2010 10:29:46 +0000 (11:29 +0100)]
librpc/rpc: fix dcerpc_errstr() to handle all nca_s_* status codes and fallback to WERROR codes

metze

14 years agos3:re-run make samba3-idl.
Günther Deschner [Tue, 23 Mar 2010 13:37:40 +0000 (14:37 +0100)]
s3:re-run make samba3-idl.

Guenther

14 years agodcerpc.idl: add nca_s_* status codes from the DCE 1.1 specification
Stefan Metzmacher [Sat, 20 Mar 2010 09:57:26 +0000 (10:57 +0100)]
dcerpc.idl: add nca_s_* status codes from the DCE 1.1 specification

metze

14 years agos3-builtin: Add some builtin groups.
Karolin Seeger [Tue, 23 Mar 2010 10:39:05 +0000 (11:39 +0100)]
s3-builtin: Add some builtin groups.

Karolin

14 years agoidl:drsblobs add code to interpret msDS-TrustForestTrustInfo
Simo Sorce [Tue, 23 Mar 2010 05:06:24 +0000 (01:06 -0400)]
idl:drsblobs add code to interpret msDS-TrustForestTrustInfo

14 years agos4:lsa Functions to set Domain Trust Information
Simo Sorce [Tue, 16 Mar 2010 20:15:39 +0000 (16:15 -0400)]
s4:lsa Functions to set Domain Trust Information

14 years agos4:lsa move code to add trusted domain user into its own function
Simo Sorce [Mon, 22 Mar 2010 20:37:24 +0000 (16:37 -0400)]
s4:lsa move code to add trusted domain user into its own function

14 years agos4:lsa Abstract crypto (un)wrapping in separate functions
Simo Sorce [Mon, 22 Mar 2010 14:48:31 +0000 (10:48 -0400)]
s4:lsa Abstract crypto (un)wrapping in separate functions

14 years agos4:registry - "LDB backend" - revert the length check for UTF16 strings
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 23:24:32 +0000 (00:24 +0100)]
s4:registry - "LDB backend" - revert the length check for UTF16 strings

Let this do the "convert_string_talloc" function as it was before.

14 years agoRevert "s4:torture - WINREG RPC - reactivate test "SetValueExtended" for s4"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 23:13:26 +0000 (00:13 +0100)]
Revert "s4:torture - WINREG RPC - reactivate test "SetValueExtended" for s4"

This reverts commit 0f2cf82e5e52da6fc71742df7b13c9f372bcf113.

Naturally we have to revert this for s4 until a new storage mechanism for binary
REG_SZ and REG_*WORD values has been found.

14 years agos4:registry - "LDB backend" - revert the binary storage of "REG_SZ", "REG_DWORD"...
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 23:09:22 +0000 (00:09 +0100)]
s4:registry - "LDB backend" - revert the binary storage of "REG_SZ", "REG_DWORD" and "REG_QWORD"

We agreed that this hack isn't the best of the possible solutions.

14 years agos4:ldb_modules/util.c - fix two counter variables to be "unsigned"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 22:53:17 +0000 (23:53 +0100)]
s4:ldb_modules/util.c - fix two counter variables to be "unsigned"

14 years agoshare_info.tdb could use non-canonicalized sharenames.
Jeremy Allison [Mon, 22 Mar 2010 22:57:31 +0000 (15:57 -0700)]
share_info.tdb could use non-canonicalized sharenames.

Fix this by moving canonicalization into lib/sharesec.c. Update the
db version to 3. Ensures we always find share names with security
descriptors attached.

Jeremy.

14 years agoconvert_string_talloc_convenience - print out the input buffer in a better way (using...
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 22:16:26 +0000 (23:16 +0100)]
convert_string_talloc_convenience - print out the input buffer in a better way (using dump_data since it isn't always directly readable)

14 years agos4:registry - "reg_key_del/add_abs": call "talloc_free" only on valid objects
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 21:38:45 +0000 (22:38 +0100)]
s4:registry - "reg_key_del/add_abs": call "talloc_free" only on valid objects

14 years agos4:registry - "reg_import_hive_key" - change a "talloc_steal" into a "talloc_reference"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 21:23:08 +0000 (22:23 +0100)]
s4:registry - "reg_import_hive_key" - change a "talloc_steal" into a "talloc_reference"

Use a "talloc_reference" since the "hive" variable can continue to exist also
after the destructed "struct local_key" object. It is also referenced by the
"mountpoints" structure under the "ctx" object.

14 years agos4:registry - "patchfile" - add another "W_ERROR_HAVE_NO_MEMORY" macro call
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 21:01:31 +0000 (22:01 +0100)]
s4:registry - "patchfile" - add another "W_ERROR_HAVE_NO_MEMORY" macro call

14 years agos4:registry - "local" - add also here more "W_ERROR_HAVE_NO_MEMORY" macro tests
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 20:46:39 +0000 (21:46 +0100)]
s4:registry - "local" - add also here more "W_ERROR_HAVE_NO_MEMORY" macro tests

14 years agos4:registry - "util.c" - initialise "result" in "reg_key_add_abs"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 20:34:54 +0000 (21:34 +0100)]
s4:registry - "util.c" - initialise "result" in "reg_key_add_abs"

14 years agos4:WINREG RPC - add also here a "W_ERROR_HAVE_NO_MEMORY"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 19:14:51 +0000 (20:14 +0100)]
s4:WINREG RPC - add also here a "W_ERROR_HAVE_NO_MEMORY"

14 years agos4:registry - "samba.c" - add a "talloc_free"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 19:12:13 +0000 (20:12 +0100)]
s4:registry - "samba.c" - add a "talloc_free"

14 years agos4:registry - add more "W_ERROR_HAVE_NO_MEMORY" invocations (on talloc'ed stuff)
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 19:10:26 +0000 (20:10 +0100)]
s4:registry - add more "W_ERROR_HAVE_NO_MEMORY" invocations (on talloc'ed stuff)

14 years agos4:registry - "util" - remove "const" from "reg_abs_path"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:53:23 +0000 (19:53 +0100)]
s4:registry - "util" - remove "const" from "reg_abs_path"

Here it's not really needed

14 years agos4:registry - adaptions for "add also a memory context argument for "reg_key_del_abs"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:56:52 +0000 (19:56 +0100)]
s4:registry - adaptions for "add also a memory context argument for "reg_key_del_abs"

14 years agos4:registry - "util" - add also a memory context argument for "reg_key_del_abs"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:47:49 +0000 (19:47 +0100)]
s4:registry - "util" - add also a memory context argument for "reg_key_del_abs"

14 years agos4:registry - "util" - make "reg_key_add_abs" consistent with "reg_key_del_abs"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:45:21 +0000 (19:45 +0100)]
s4:registry - "util" - make "reg_key_add_abs" consistent with "reg_key_del_abs"

14 years agos4:registry - "util" - fix up memory allocated data
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:44:19 +0000 (19:44 +0100)]
s4:registry - "util" - fix up memory allocated data

14 years agos4:registry - "dir.c" - fix up dynamic memory allocation operations
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:26:59 +0000 (19:26 +0100)]
s4:registry - "dir.c" - fix up dynamic memory allocation operations

- Added free operations where needed
- Use always the "mem_ctx" for temporary data
- Proof with W_ERROR_HAVE_NO_MEMORY if stuff was allocated

14 years agos4:registry - adaptions for "add memory contexts for delete value/key functions"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 18:18:56 +0000 (19:18 +0100)]
s4:registry - adaptions for "add memory contexts for delete value/key functions"

14 years agos4:registry - registry.h - add memory contexts for delete value/key functions
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 17:51:20 +0000 (18:51 +0100)]
s4:registry - registry.h - add memory contexts for delete value/key functions

14 years agotdb: Fix bug 7248, avoid the nanosleep dependency
Volker Lendecke [Sun, 14 Mar 2010 19:07:58 +0000 (20:07 +0100)]
tdb: Fix bug 7248, avoid the nanosleep dependency

14 years agoFix valgrind error when running under share level security. plaintext_password
Jeremy Allison [Mon, 22 Mar 2010 20:05:29 +0000 (13:05 -0700)]
Fix valgrind error when running under share level security. plaintext_password
is a data blob with a data pointer pointing to an allocation of length zero.

Jeremy.

14 years agos3: Add the "ctdb locktime warn threshold" parameter
Volker Lendecke [Fri, 5 Mar 2010 15:46:36 +0000 (16:46 +0100)]
s3: Add the "ctdb locktime warn threshold" parameter

This is mainly a debugging aid for post-mortem analysis in case a cluster file
system is slow.

14 years agos3: Implement an asynchronous echo responder process
Volker Lendecke [Thu, 18 Mar 2010 11:50:22 +0000 (12:50 +0100)]
s3: Implement an asynchronous echo responder process

This replies to echo requests when the main smbd is stuck somewhere

Signed-off-by: Stefan Metzmacher <metze@samba.org>
14 years agos3:smbd: don't allow SMB2 if the async echo handler is active
Stefan Metzmacher [Mon, 22 Mar 2010 08:06:07 +0000 (09:06 +0100)]
s3:smbd: don't allow SMB2 if the async echo handler is active

metze

14 years agos3:smbd: disable SMB encryption when the echo handler is active
Stefan Metzmacher [Fri, 19 Mar 2010 14:47:11 +0000 (15:47 +0100)]
s3:smbd: disable SMB encryption when the echo handler is active

metze

14 years agos3:smbd: disallow readbraw and writebraw if the echo handler is active
Stefan Metzmacher [Mon, 22 Mar 2010 08:11:05 +0000 (09:11 +0100)]
s3:smbd: disallow readbraw and writebraw if the echo handler is active

metze

14 years agos3:smbd: disable sendfile if the echo handler is active
Stefan Metzmacher [Fri, 19 Mar 2010 11:08:13 +0000 (12:08 +0100)]
s3:smbd: disable sendfile if the echo handler is active

metze

14 years agos3:smbd: don't use recvfile if the echo handler is active
Stefan Metzmacher [Thu, 18 Mar 2010 19:22:26 +0000 (20:22 +0100)]
s3:smbd: don't use recvfile if the echo handler is active

metze

14 years agos3:smbd: setup a shared memory area for the signing state
Stefan Metzmacher [Fri, 19 Mar 2010 13:59:19 +0000 (14:59 +0100)]
s3:smbd: setup a shared memory area for the signing state

metze

14 years agos3:smbd: add echo handler information to struct smbd_server_connection
Stefan Metzmacher [Mon, 22 Mar 2010 07:42:13 +0000 (08:42 +0100)]
s3:smbd: add echo handler information to struct smbd_server_connection

metze

14 years agos3:param: add "async smb echo handler" option
Stefan Metzmacher [Mon, 22 Mar 2010 07:35:33 +0000 (08:35 +0100)]
s3:param: add "async smb echo handler" option

This will enable an extra forked process that will reply
to SMBecho requests, while the main process is blocked by another
request.

metze

14 years agos3:smbd: pass down trusted_channel via receive_smb_talloc()
Stefan Metzmacher [Thu, 18 Mar 2010 14:36:19 +0000 (15:36 +0100)]
s3:smbd: pass down trusted_channel via receive_smb_talloc()

metze

14 years agos3:smbd: let reply_readbraw_error use the locked socket
Stefan Metzmacher [Fri, 19 Mar 2010 11:04:32 +0000 (12:04 +0100)]
s3:smbd: let reply_readbraw_error use the locked socket

metze

14 years agos3:smbd: send keepalive packets under the socket lock
Stefan Metzmacher [Fri, 19 Mar 2010 11:02:27 +0000 (12:02 +0100)]
s3:smbd: send keepalive packets under the socket lock

metze

14 years agos3:smbd: smbd_[un]lock_socket() while accessing the socket to the client
Stefan Metzmacher [Thu, 18 Mar 2010 08:23:48 +0000 (09:23 +0100)]
s3:smbd: smbd_[un]lock_socket() while accessing the socket to the client

metze

14 years agos3:smbd: add smbd_[un]lock_socket() dummies
Stefan Metzmacher [Thu, 18 Mar 2010 08:17:43 +0000 (09:17 +0100)]
s3:smbd: add smbd_[un]lock_socket() dummies

metze

14 years agos3:smbd: add an option to skip signings checks srv_check_sign_mac for trusted channels
Stefan Metzmacher [Thu, 18 Mar 2010 08:14:40 +0000 (09:14 +0100)]
s3:smbd: add an option to skip signings checks srv_check_sign_mac for trusted channels

metze

14 years agos3:libsmb: add a smb_signing_init_ex() function
Stefan Metzmacher [Wed, 17 Mar 2010 14:07:07 +0000 (15:07 +0100)]
s3:libsmb: add a smb_signing_init_ex() function

Make it possible to overload memory handling functions.

metze

14 years agolib/util: add allocate_anonymous_shared()
Stefan Metzmacher [Mon, 22 Mar 2010 07:27:58 +0000 (08:27 +0100)]
lib/util: add allocate_anonymous_shared()

metze

14 years agos3: Fix a bad memleak in winbind
Volker Lendecke [Mon, 22 Mar 2010 14:29:19 +0000 (15:29 +0100)]
s3: Fix a bad memleak in winbind

14 years agos4:registry - "RPC backend" - use "talloc_steal" and not "talloc_reference"
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 13:00:02 +0000 (14:00 +0100)]
s4:registry - "RPC backend" - use "talloc_steal" and not "talloc_reference"

No need to create additional references but we need to assign the strings to the
right context.

14 years agos4:registry - "patchfile" - check if strings could be allocated
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 12:39:40 +0000 (13:39 +0100)]
s4:registry - "patchfile" - check if strings could be allocated

14 years agos4:registry - "patchfile" - initialise the data blobs
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 12:33:04 +0000 (13:33 +0100)]
s4:registry - "patchfile" - initialise the data blobs

14 years agos4:registry - "patchfile" - add more "talloc_free"s to save memory
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 11:42:21 +0000 (12:42 +0100)]
s4:registry - "patchfile" - add more "talloc_free"s to save memory

14 years agos4:registry - "util.c" - "reg_string_to_val" - consider always the return values
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 11:22:13 +0000 (12:22 +0100)]
s4:registry - "util.c" - "reg_string_to_val" - consider always the return values

In some cases we didn't consider them.

14 years agos4:registry - "patchfile_preg.c" - also here don't accumulate the memory usage
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 11:15:19 +0000 (12:15 +0100)]
s4:registry - "patchfile_preg.c" - also here don't accumulate the memory usage

Free always the unused stuff.

14 years agos4:registry - "patchfile_dotreg.c" - fix a memory leak
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 11:00:57 +0000 (12:00 +0100)]
s4:registry - "patchfile_dotreg.c" - fix a memory leak

Here we allocate memory on the "NULL" context through "reg_val_data_string" on
each call of "set_value". So when we have written out the allocated data on the
specified file descriptor we should immediately free this memory! Otherwise we
may end up with a big memory consumption on big registry databases.

14 years agos4:registry - "patchfile" - add comments
Matthias Dieter Wallnöfer [Mon, 22 Mar 2010 10:45:20 +0000 (11:45 +0100)]
s4:registry - "patchfile" - add comments

Helps to understand when we need to generate the hive diffs.

14 years agos3: Add "log writeable files on exit" parameter
Volker Lendecke [Mon, 22 Mar 2010 10:19:10 +0000 (11:19 +0100)]
s3: Add "log writeable files on exit" parameter

This boolean option controls whether at exit time the server dumps a list of
files with debug level 0 that were still open for write. This is an
administrative aid to find the files that were potentially corrupt if the
network connection died.

14 years agos4-selftest: fix selftest-vars script.
Günther Deschner [Sat, 20 Mar 2010 23:05:00 +0000 (00:05 +0100)]
s4-selftest: fix selftest-vars script.

Tridge, please check.

Guenther

14 years agos4:dsdb Add a shortcut sequence number for schema reloads
Andrew Bartlett [Mon, 22 Mar 2010 07:44:51 +0000 (18:44 +1100)]
s4:dsdb Add a shortcut sequence number for schema reloads

This uses the ldb sequence number, in a hope to detect an unchanged
schema quicker.

Andrew Bartlett

14 years agos4:dsdb Rework schema loading and add schema reloading
Andrew Bartlett [Mon, 22 Mar 2010 05:03:33 +0000 (16:03 +1100)]
s4:dsdb Rework schema loading and add schema reloading

This commit reworks Samba4's schema loading code to detect when it
needs to reload the schema.  This is done by watching the @REPLCHANGED
special DN.

The reload happens by means of a callback, which is only set when the
schema is loaded from the ldb - not when loaded from an LDIF file or
DRS.

We also rework the global schema handling - instead of storing the
pointer to the global schema in each ldb, we store a flag indicating
that the global schema should be returned at run time.  This makes it
much easier to switch to a new global schema.

Andrew Bartlett

14 years agos4:dsdb Move dsdb_save_partition_usn() to be a module helper function
Andrew Bartlett [Mon, 22 Mar 2010 05:00:39 +0000 (16:00 +1100)]
s4:dsdb Move dsdb_save_partition_usn() to be a module helper function

This function should not traverse the module stack again, but instead
run from this point.  Also add a matching
dsdb_module_load_partition_usn() and change repl_meta_data to match.

Andrew Bartlett

14 years agos4:schema Expand the schema structure
Andrew Bartlett [Mon, 22 Mar 2010 04:41:51 +0000 (15:41 +1100)]
s4:schema Expand the schema structure

We now store the location of the schema in the schema, and provide
hooks for a future schema reloading mechanism.

Andrew Bartlett

14 years agos4:dsdb Remove unused 'dsdb_make_schema_global' call from pyglue
Andrew Bartlett [Mon, 22 Mar 2010 04:20:47 +0000 (15:20 +1100)]
s4:dsdb Remove unused 'dsdb_make_schema_global' call from pyglue

14 years agos4:dsdb Add 'const' to some struct dsdb_schema variables
Andrew Bartlett [Mon, 22 Mar 2010 04:19:55 +0000 (15:19 +1100)]
s4:dsdb Add 'const' to some struct dsdb_schema variables

We don't currently require this, but we may move this way in future.

14 years agos4:dsdb Don't load the schema unconditionally
Andrew Bartlett [Mon, 22 Mar 2010 04:17:58 +0000 (15:17 +1100)]
s4:dsdb Don't load the schema unconditionally

Schema loads now come at a price, so avoid doing them if we don't have
to (such as when doing an @REPLCHANGED or other special DN based
search).

Andrew Bartlett

14 years agos3: file_walk_table -> files_forall
Volker Lendecke [Mon, 22 Mar 2010 08:16:57 +0000 (09:16 +0100)]
s3: file_walk_table -> files_forall

This is more in line with the rest of the Samba code, like connections_forall
etc.

14 years agos3: Fix some nonempty blank lines
Volker Lendecke [Mon, 22 Mar 2010 08:11:24 +0000 (09:11 +0100)]
s3: Fix some nonempty blank lines

14 years agos4:registry - "patchfile.c" - consider also the hives in the generated diffs
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 23:13:02 +0000 (00:13 +0100)]
s4:registry - "patchfile.c" - consider also the hives in the generated diffs

Apparently the diffs for the hives weren't generated previously.

14 years agos4:registry - "patchfile.c" - fix a typo
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 23:02:06 +0000 (00:02 +0100)]
s4:registry - "patchfile.c" - fix a typo

14 years agos4:regtree - fix counter variables to be "unsigned"
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 22:41:50 +0000 (23:41 +0100)]
s4:regtree - fix counter variables to be "unsigned"

14 years agos4:registry - "LDB backend" - "reg_ldb_unpack_value"
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 22:36:09 +0000 (23:36 +0100)]
s4:registry - "LDB backend" - "reg_ldb_unpack_value"

When the name isn't found it is the default value. Call it "" to be consistent.

14 years agos4:registry - "LDB backend" - "reg_key_get_info"
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 21:01:06 +0000 (22:01 +0100)]
s4:registry - "LDB backend" - "reg_key_get_info"

Consider also the default value (if it exists) as value. That means:
- count it when setting "num_values"
- take also his buffer length as a candidate for the maximum value buffer length

This is what Windows does.

14 years agos4:registry - "LDB backend" - "ldb_get_default_value"
Matthias Dieter Wallnöfer [Sun, 21 Mar 2010 21:57:31 +0000 (22:57 +0100)]
s4:registry - "LDB backend" - "ldb_get_default_value"

There exist also key objects (the hives) which don't contain a "key" entry at
all. This prevented to display their default value (my fault).