bbaumbach/samba-autobuild/.git
16 years agor10846: Create a "wbsrv_domain", change wb_finddcs to the style of the rest of the
Volker Lendecke [Sun, 9 Oct 2005 12:50:35 +0000 (12:50 +0000)]
r10846: Create a "wbsrv_domain", change wb_finddcs to the style of the rest of the
async helpers.

Volker
(This used to be commit 10585ba4e81e979a03aec747db6fc059978fa566)

16 years agor10845: Add new function to decrypt the session keys in samlogon responses.
Andrew Bartlett [Sun, 9 Oct 2005 12:38:23 +0000 (12:38 +0000)]
r10845: Add new function to decrypt the session keys in samlogon responses.

Andrew Bartlett
(This used to be commit 6d24d8d12cdc64b180fd6277f0775e943f26e82b)

16 years agor10844: Add challenge-response authentication to Samba4's winbindd for VL.
Andrew Bartlett [Sun, 9 Oct 2005 12:13:05 +0000 (12:13 +0000)]
r10844: Add challenge-response authentication to Samba4's winbindd for VL.

Plaintext should be simple, but I'm going to do some infrustructure
work first.

Andrew Bartlett
(This used to be commit c9273729e4db4adc0061087fe7e0332e2bc24384)

16 years agor10843: Reformatting
Volker Lendecke [Sun, 9 Oct 2005 08:32:06 +0000 (08:32 +0000)]
r10843: Reformatting
(This used to be commit ecaa70f63b7f38a1daf8e33ded738107c5f6b53a)

16 years agor10842: Fix some issues with [out] unions that have a discriminator that is only
Jelmer Vernooij [Sat, 8 Oct 2005 20:19:35 +0000 (20:19 +0000)]
r10842: Fix some issues with [out] unions that have a discriminator that is only
[in]
(This used to be commit 3a4086d6142fa73b3adb2d66b1bfc9cd2585f31d)

16 years agor10840: Fix indentation
Jelmer Vernooij [Sat, 8 Oct 2005 17:55:56 +0000 (17:55 +0000)]
r10840: Fix indentation
(This used to be commit 7877d436576a1e7c6b02347f87e3f7e14e683888)

16 years agor10839: Add some [ref] (required for ethereal and Samba3 parser generators)
Jelmer Vernooij [Sat, 8 Oct 2005 17:55:28 +0000 (17:55 +0000)]
r10839: Add some [ref] (required for ethereal and Samba3 parser generators)
(This used to be commit 5556deee30616396bbaf1182eca08bf1213c0860)

16 years agor10838: Get us an schannel'ed netlogon pipe.
Volker Lendecke [Sat, 8 Oct 2005 17:45:27 +0000 (17:45 +0000)]
r10838: Get us an schannel'ed netlogon pipe.

Abartlet, now I think I need some assistance to implement the pam auth & crap
auth calls.

Volker
(This used to be commit 90a30c8b6585ed48b50e6aed75f3ecfd3543bbdc)

16 years agor10836: giving NT_STATUS_NO_MEMORY, when the connection fails wasn't a good idea...
Stefan Metzmacher [Sat, 8 Oct 2005 16:35:33 +0000 (16:35 +0000)]
r10836: giving NT_STATUS_NO_MEMORY, when the connection fails wasn't a good idea...

metze
(This used to be commit e7ee73a747a025a66ac6563172e51f160bc28e0a)

16 years agor10834: Work in progress on winbind. With some helper routines the composite functions
Volker Lendecke [Sat, 8 Oct 2005 16:25:00 +0000 (16:25 +0000)]
r10834: Work in progress on winbind. With some helper routines the composite functions
start to look sane.

Question: What about providing all winbind commands as irpc interfaces that
are called from the samba3 compatibility layer? This way it would be easy for
other samba components to access its functionality. Does that make sense?

Volker
(This used to be commit 2a6b8053859ea5690f90a8d2074d2bb4f06551f8)

16 years agor10832: free the old session info
Stefan Metzmacher [Sat, 8 Oct 2005 09:55:38 +0000 (09:55 +0000)]
r10832: free the old session info

metze
(This used to be commit 16b2569788348ee3654557cf714ea3b204375c3c)

16 years agor10830: we should use the same name in all places:-)
Stefan Metzmacher [Sat, 8 Oct 2005 07:41:57 +0000 (07:41 +0000)]
r10830: we should use the same name in all places:-)

metze
(This used to be commit fbe8fd06b700b78f02b7f01fc2ad45eee419d216)

16 years agor10829: Documentation updates
Jelmer Vernooij [Sat, 8 Oct 2005 00:33:19 +0000 (00:33 +0000)]
r10829: Documentation updates
Update TODO
Some small fixes to the modules
(This used to be commit 0c53e7c3cf7fd91fd34c48a5e68c1bcf70569854)

16 years agor10825: Complete wbinfo -n
Volker Lendecke [Fri, 7 Oct 2005 19:08:51 +0000 (19:08 +0000)]
r10825: Complete wbinfo -n
(This used to be commit 1afa893506f3d7157e251eec9baeba28dc011587)

16 years agor10820: Use talloc_get_type as suggested by tridge.
Andrew Bartlett [Fri, 7 Oct 2005 12:30:46 +0000 (12:30 +0000)]
r10820: Use talloc_get_type as suggested by tridge.

Andrew Bartlett
(This used to be commit 9c511a16f829df5f177b94c7234875d4ec8afe52)

16 years agor10812: Fix capitalisation (thanks tridge).
Andrew Bartlett [Fri, 7 Oct 2005 11:50:46 +0000 (11:50 +0000)]
r10812: Fix capitalisation (thanks tridge).

Andrew Bartlett
(This used to be commit f78982a0f1079b07be6d7bfd9cc0f92c2cebcf1c)

16 years agor10811: Revert accidental commit, I still need to finish the displayName and
Andrew Bartlett [Fri, 7 Oct 2005 11:37:38 +0000 (11:37 +0000)]
r10811: Revert accidental commit, I still need to finish the displayName and
syntatical mapping work.

Andrew Bartlett
(This used to be commit 7ec5084f63148d748b6bc87d6817363b079eebe1)

16 years agor10810: This adds the hooks required to communicate the current user from the
Andrew Bartlett [Fri, 7 Oct 2005 11:31:45 +0000 (11:31 +0000)]
r10810: This adds the hooks required to communicate the current user from the
authenticated session down into LDB.  This associates a session info
structure with the open LDB, allowing a future ldb_ntacl module to
allow/deny operations on that basis.

Along the way, I cleaned up a few things, and added new helper functions
to assist.  In particular the LSA pipe uses simpler queries for some of
the setup.

In ldap_server, I have removed the 'ldasrv:hacked' module, which hasn't
been worked on (other than making it continue to compile) since January,
and I think the features of this module are being put into ldb anyway.

I have also changed the partitions in ldap_server to be initialised
after the connection, with the private pointer used to associate the ldb
with the incoming session.

Andrew Bartlett
(This used to be commit fd7203789a2c0929eecea8125b57b833a67fed71)

16 years agor10809: Add struct decl
Andrew Bartlett [Fri, 7 Oct 2005 10:06:44 +0000 (10:06 +0000)]
r10809: Add struct decl
(This used to be commit aa30c16de0405f43ed35a28764ef25e234f3b2a7)

16 years agor10807: Make the split-out files actually compile...
Andrew Bartlett [Fri, 7 Oct 2005 09:19:59 +0000 (09:19 +0000)]
r10807: Make the split-out files actually compile...

Andrew Bartlett
(This used to be commit a4282c14d5cf6b5eceb293d657c7cd5eaea89046)

16 years agor10806: Add missing file.
Andrew Bartlett [Fri, 7 Oct 2005 08:56:01 +0000 (08:56 +0000)]
r10806: Add missing file.
(This used to be commit 9becd2c0d6ed8fe6de4b8bca0f2290f81fea28f7)

16 years agor10805: Move RPC-SAMLOGON to C99 initialisation
Andrew Bartlett [Fri, 7 Oct 2005 08:51:35 +0000 (08:51 +0000)]
r10805: Move RPC-SAMLOGON to C99 initialisation

Andrew Bartlett
(This used to be commit 0f994275ce5d84bdb746524c5da7d9661fbadb63)

16 years agor10804: Move the DRSUAPI cracknames test into a seperate file, and collapse
Andrew Bartlett [Fri, 7 Oct 2005 08:12:05 +0000 (08:12 +0000)]
r10804: Move the DRSUAPI cracknames test into a seperate file, and collapse
the file size by converting it to a table-driven system.

Andrew Bartlett
(This used to be commit f5499877722d6c4636036325efb75d35aa853cc4)

16 years agor10803: Remove a duplicate krbtgt test, and add a test looking for the
Andrew Bartlett [Fri, 7 Oct 2005 06:25:45 +0000 (06:25 +0000)]
r10803: Remove a duplicate krbtgt test, and add a test looking for the
'support account'.  This shows that we need to be searching on
displayName, and probably CN too.

Andrew Bartlett
(This used to be commit 28194504951af22a76cf26f3de749cb448518bbc)

16 years agor10800: Indent
Andrew Bartlett [Fri, 7 Oct 2005 04:00:11 +0000 (04:00 +0000)]
r10800: Indent
(This used to be commit 057cede8b44e1b53f5d94a7270e9b6a90d50de4c)

16 years agor10796: Make getting an anonymous session info a utility function.
Andrew Bartlett [Fri, 7 Oct 2005 02:07:25 +0000 (02:07 +0000)]
r10796: Make getting an anonymous session info a utility function.

Andrew Bartlett
(This used to be commit 6021fa373405f8788dc216aae82a3f4a80df4ba7)

16 years agor10791: Add copyright, fix comments (this isn't the timestamps module any more)
Andrew Bartlett [Fri, 7 Oct 2005 01:01:44 +0000 (01:01 +0000)]
r10791: Add copyright, fix comments (this isn't the timestamps module any more)

Andrew Bartlett
(This used to be commit efdc6d834aecbf978f538365c72149fa7afe0828)

16 years agor10790: allow updating of existing ldb opaque values (thanks to abartlet for
Andrew Tridgell [Fri, 7 Oct 2005 00:40:29 +0000 (00:40 +0000)]
r10790: allow updating of existing ldb opaque values (thanks to abartlet for
spotting this)
(This used to be commit ef13569ca94da00cc410318e61505e70f3606674)

16 years agor10788: - Give unions a name
Jelmer Vernooij [Thu, 6 Oct 2005 20:53:55 +0000 (20:53 +0000)]
r10788: - Give unions a name
- Check initialisation function return value
(This used to be commit ebb8e75ebd572d62bdb3615d41c210bcc918fa41)

16 years agor10787: Fix silly nodiscriminant-issue
Jelmer Vernooij [Thu, 6 Oct 2005 20:07:53 +0000 (20:07 +0000)]
r10787: Fix silly nodiscriminant-issue
(This used to be commit 43a5c863bf81af9b8415b6ccf8a386c36d4fa7f9)

16 years agor10768: add a function to create a stream_connection from an already
Stefan Metzmacher [Thu, 6 Oct 2005 14:44:37 +0000 (14:44 +0000)]
r10768: add a function to create a stream_connection from an already
existing socket connection, that's used for protocols
which switch the client/server roles inside a session

metze
(This used to be commit d4b4d963353e6986da5b42680739329069f5affe)

16 years agor10766: - make it possible to mark a wrepl_request as send only,
Stefan Metzmacher [Thu, 6 Oct 2005 14:38:07 +0000 (14:38 +0000)]
r10766: - make it possible to mark a wrepl_request as send only,
  used for WREPL_REPL_INFORM* messsages
- make it possible to close the connection after a request was send
  used for WREPL_ASSOCIATION_STOP
- fix the torture test that tests the assoc context handling
  between connections, you can issue a request and get the reply
  on another connection, I think we should not implement that in our server
  code, as I think it's a security hole, you can cause a windows server
  to send the replies to someone another client, that doesn't wait for data,
  and as there're no massage_id in the protocol the client would be confused
  by a replies that doesn't belong to a query

metze
(This used to be commit dfc95de8fa7ded8ea92cafe58cf86efcc7920156)

16 years agor10765: Actually use VALS() setting for enums
Jelmer Vernooij [Thu, 6 Oct 2005 11:57:25 +0000 (11:57 +0000)]
r10765: Actually use VALS() setting for enums
(This used to be commit 6b33e3eedd13ade5c7c9f4b5062e8b8f919fe39c)

16 years agor10764: To match Win2k3 SP1, we need to set an anonymous user token for
Andrew Bartlett [Thu, 6 Oct 2005 11:15:20 +0000 (11:15 +0000)]
r10764: To match Win2k3 SP1, we need to set an anonymous user token for
schannel connections.

Test for Win2k3 SP1 behaviour in RPC-SCHANNEL.

Andrew Bartlett
(This used to be commit 1c3911374ec65e4770c2fe9109d7b7d3ecd99f6a)

16 years agor10763: PROOF of the single, easily understood cause of all of our schannel PAIN!
Andrew Bartlett [Thu, 6 Oct 2005 10:29:28 +0000 (10:29 +0000)]
r10763: PROOF of the single, easily understood cause of all of our schannel PAIN!

This test currently passes against Win2k3 SP0, but not SP1.  The
problem was not a convoluted series of changes in policy handles, but
a simple rule that schannel logins would be regarded as ANONYMOUS not
SYSTEM.

Andrew Bartlett
(This used to be commit e843a7fcba496b91b014ec1bff2f2dfb2a2fcdb1)

16 years agor10761: we need to use a pointer to a nbt_name to fix compiler warnings, because...
Stefan Metzmacher [Thu, 6 Oct 2005 07:26:05 +0000 (07:26 +0000)]
r10761: we need to use a pointer to a nbt_name to fix compiler warnings, because we can
only use a pointers to unknown types in proto.h

metze
(This used to be commit 2f46e54e1bcf43f1bee062ff9a21e646cc3676e9)

16 years agor10760: remove only one star from the pointers, and not from the stars that
Stefan Metzmacher [Thu, 6 Oct 2005 07:04:36 +0000 (07:04 +0000)]
r10760: remove only one star from the pointers, and not from the stars that
come from the array nesting

this fixes a compiler warning and but in the eventlog.h,
where the idl uses nstring strings[num_strings]

metze
(This used to be commit c4a252bd4678ef991705aaaf91c03c697f2f568e)

16 years agor10759: make modules easier to write by allowing modules to only implement the
Andrew Tridgell [Thu, 6 Oct 2005 06:57:09 +0000 (06:57 +0000)]
r10759: make modules easier to write by allowing modules to only implement the
functions they care about, instead of all functions. This also makes
it more likely that future changes to ldb will not break existing
modules
(This used to be commit 45f0c967b58e7c1b2e900a4d74cfde2a2c527dfa)

16 years agor10757: remove the proxy module (it is not complete yet)
Andrew Tridgell [Thu, 6 Oct 2005 06:01:56 +0000 (06:01 +0000)]
r10757: remove the proxy module (it is not complete yet)
(This used to be commit 3c5f3032fcb092545580b986e0ce58bb49e4d9cb)

16 years agor10756: another fix for the construction of expressions from subtrees for
Andrew Tridgell [Thu, 6 Oct 2005 05:53:46 +0000 (05:53 +0000)]
r10756: another fix for the construction of expressions from subtrees for
SUBSTRING searches. This time fix multi-part substring searches.
(This used to be commit bf5cef6f00466fc1dc3c2864a109f1ccd92681b0)

16 years agor10755: fixed the construction of expressions from subtrees for SUBSTRING searches
Andrew Tridgell [Thu, 6 Oct 2005 05:41:32 +0000 (05:41 +0000)]
r10755: fixed the construction of expressions from subtrees for SUBSTRING searches
(This used to be commit 71c06778d4a3ac1ca4198071ae3351acdc0656d9)

16 years agor10754: fixed a valgrind error for unmatched SMB replies
Andrew Tridgell [Thu, 6 Oct 2005 05:25:35 +0000 (05:25 +0000)]
r10754: fixed a valgrind error for unmatched SMB replies
(This used to be commit b714ab64fd79d5cabc39779774fae7c3861a84da)

16 years agor10753: don't require every ldb module to implement both a search_bytree() and
Andrew Tridgell [Thu, 6 Oct 2005 05:24:46 +0000 (05:24 +0000)]
r10753: don't require every ldb module to implement both a search_bytree() and
a search() function, instead each module now only implements the
bytree method, and the expression based search is handled generically
by the modules code. This makes for more consistency and less code
duplication.

fixed the tdb backend to handle BASE searches much more
efficiently. They now always only lookup one record, regardless of the
search expression
(This used to be commit 7e44f9153c5578624e2fca04cdc0a00af0fd9eb4)

16 years agor10752: make sure we set the exist status correctly for the tdb tests
Andrew Tridgell [Thu, 6 Oct 2005 04:45:33 +0000 (04:45 +0000)]
r10752: make sure we set the exist status correctly for the tdb tests
(This used to be commit e83635941c1829d8cb8c0784b7f7303c85654199)

16 years agor10742: Support multi-level pointers + ref pointer fixes
Jelmer Vernooij [Wed, 5 Oct 2005 22:18:59 +0000 (22:18 +0000)]
r10742: Support multi-level pointers + ref pointer fixes
(This used to be commit 258b762dc62b257f99d1d859c5a3d850aba3e9fa)

16 years agor10739: Reduce number of calls to prs_align_custom()
Jelmer Vernooij [Wed, 5 Oct 2005 19:53:41 +0000 (19:53 +0000)]
r10739: Reduce number of calls to prs_align_custom()
(This used to be commit ebeeec5406308d493d45b1088963a87cdb953cac)

16 years agor10737: Fix some alignment issues
Jelmer Vernooij [Wed, 5 Oct 2005 19:41:53 +0000 (19:41 +0000)]
r10737: Fix some alignment issues
(This used to be commit 2b4270a2e965eb5c6fe0b25d22d2977e60bb7d43)

16 years agor10734: Generate ptr, size, offset, and length elements in unions just once.
Jelmer Vernooij [Wed, 5 Oct 2005 17:13:29 +0000 (17:13 +0000)]
r10734: Generate ptr, size, offset, and length elements in unions just once.
(This used to be commit 12bfa5d01bcb4cb9dad5167e1a3721fd68f06275)

16 years agor10732: Generate _ctr structures for unions
Jelmer Vernooij [Wed, 5 Oct 2005 14:52:35 +0000 (14:52 +0000)]
r10732: Generate _ctr structures for unions
(This used to be commit bd8fcb05003ad75f521783ad9603c923eacafc1a)

16 years agor10726: fix to talloc_parent() from Michael O'Brien
Andrew Tridgell [Wed, 5 Oct 2005 06:32:00 +0000 (06:32 +0000)]
r10726: fix to talloc_parent() from Michael O'Brien
(This used to be commit f31a2376f3fef1cc2b40b37fb4d94a4b67eec6d4)

16 years agor10721: Handle allocations and primitive / deferred data correctly. In theory,
Jelmer Vernooij [Wed, 5 Oct 2005 00:29:47 +0000 (00:29 +0000)]
r10721: Handle allocations and primitive / deferred data correctly. In theory,
the generated output for DFS should work now (it compiles cleanly, but
I haven't tested it yet).

Not supported:
 - subcontexts()
 - relative pointers
 - unions of pointers
 - DATA_BLOB
 - several other things

Also still need to do:
 - Remove some spurious spaces in the output
 - Do range() checking

Example output is still available at http://samba.org/~jelmer/pidl_samba3/
(This used to be commit e2d7e382bb645f1bddd2047669bed10f121b59d2)

16 years agor10719: Remove __FUNCTION__ macro as it might not be available at all systems
Jelmer Vernooij [Tue, 4 Oct 2005 21:42:39 +0000 (21:42 +0000)]
r10719: Remove __FUNCTION__ macro as it might not be available at all systems
(pointed out by John E. Malmberg)
(This used to be commit 3169bbc15703b39943a11939e9dcccb2a9b4ad5d)

16 years agor10718: Another large set of small improvements. All generated files compile
Jelmer Vernooij [Tue, 4 Oct 2005 21:25:18 +0000 (21:25 +0000)]
r10718: Another large set of small improvements. All generated files compile
without warnings now. The only things left to do that are
required for DFS:
 - add allocation of arrays in marshalling phase
 - handling primitive and deferred data in embedded structures / unions.

Example output is again available from http://samba.org/~jelmer/pidl_samba3/
(This used to be commit 9fe724f6fb026d95306587f696c065f348aaf219)

16 years agor10717: Another bunch of small updates. All generated files except parse_dfs.c
Jelmer Vernooij [Tue, 4 Oct 2005 19:38:35 +0000 (19:38 +0000)]
r10717: Another bunch of small updates. All generated files except parse_dfs.c
compile now when generated from Samba4's dfs.idl.
(This used to be commit 2b315b1942e77640eca8e0a3347b2a9ed1920f6a)

16 years agor10716: Use correct Samba3 data types for strings. Also use Samba3 types
Jelmer Vernooij [Tue, 4 Oct 2005 18:24:21 +0000 (18:24 +0000)]
r10716: Use correct Samba3 data types for strings. Also use Samba3 types
for a couple of other types (policy handles, SIDs, times)
(This used to be commit c2527217b4b4c120d82044e65b979dd3b7d2609e)

16 years agor10715: More Samba3 parser generator improvements:
Jelmer Vernooij [Tue, 4 Oct 2005 17:21:31 +0000 (17:21 +0000)]
r10715: More Samba3 parser generator improvements:
- Actually generate parsers for unions and structs.
- Support some more builtin types.
- Some more work on supporting arrays.
- Several other small fixes.

I've updated the example output at http://samba.org/~jelmer/
(This used to be commit b229c033ebc7ec972b32f1b75b60a9c68a36db97)

16 years agor10713: Couple more updates to the Samba3 parser generators.
Jelmer Vernooij [Tue, 4 Oct 2005 13:07:23 +0000 (13:07 +0000)]
r10713: Couple more updates to the Samba3 parser generators.
Unions and enums have been improved, init functions are now generated
properly, some other small improvements.
(This used to be commit 8a60e79175eb27ef9fa4b8dea72a518bbaab900f)

16 years agor10712: Use data_blob_talloc, thanks to valgrind for finding the errors.
Andrew Bartlett [Tue, 4 Oct 2005 12:04:10 +0000 (12:04 +0000)]
r10712: Use data_blob_talloc, thanks to valgrind for finding the errors.

Andrew Bartlett
(This used to be commit 1f6fec8e6b0845ae6000eeda65641435fb18c9e3)

16 years agor10711: An error of 'user exists' is not an error, just an indication of how
Andrew Bartlett [Tue, 4 Oct 2005 12:02:52 +0000 (12:02 +0000)]
r10711: An error of 'user exists' is not an error, just an indication of how
the join was processed.

Andrew Bartlett
(This used to be commit 0d93f11c894927a9ab69f7a31b9fbedde9d698b6)

16 years agor10709: fixed a crash bug rather similar to the one volker found in the dcerpc
Andrew Tridgell [Tue, 4 Oct 2005 10:18:07 +0000 (10:18 +0000)]
r10709: fixed a crash bug rather similar to the one volker found in the dcerpc
code, where a stream_terminate_connection() while processing a request
can cause a later defererence of the connection structure to die.
(This used to be commit efbcb0f74176058a74d7134dae4658b891fc6f16)

16 years agor10708: a bit more error checking in the idap ldb backend
Andrew Tridgell [Tue, 4 Oct 2005 05:41:05 +0000 (05:41 +0000)]
r10708: a bit more error checking in the idap ldb backend
(This used to be commit 63ebaad393e38b28c8f97f33e2b22f2445733405)

16 years agor10706: split out the irpc server functions in the NBT server, so the mainline
Andrew Tridgell [Tue, 4 Oct 2005 01:51:55 +0000 (01:51 +0000)]
r10706: split out the irpc server functions in the NBT server, so the mainline
NBT server code remains reabable. Also fixed the copyright header to
include Volker, as he wrote the getdc server function
(This used to be commit a973197d3e0045a76a2f14153f5541b9e36e6c8a)

16 years agor10705: fixed a crash bug in the getdcname irpc server for winbind. The
Andrew Tridgell [Tue, 4 Oct 2005 01:43:24 +0000 (01:43 +0000)]
r10705: fixed a crash bug in the getdcname irpc server for winbind. The
problem was that the return string was declared as:

[out] astring dcname

which means "this is a non-NULL string". The server code sometimes
returned NULL however (on getdc lookup failure), which caused the NDR
marshalling code to crash. When you declare a non-pointer return value
you are promising that the value can never be NULL.

The trivial fix is to use:

[out] astring *dcname

which leaves the API alone, but includes a pointer in the wire format,
which in turn means it is valid to send a NULL string as a response.
(This used to be commit e39bac61960f6ce8957f148a482e39499a8e9096)

16 years agor10704: don't try to free the netlogon pipe twice
Andrew Tridgell [Tue, 4 Oct 2005 01:35:22 +0000 (01:35 +0000)]
r10704: don't try to free the netlogon pipe twice
(This used to be commit 2c3a9f04db5d61305f4eca8b44e33c2dd15a6dc4)

16 years agor10703: Add a new user account, change the password and test it in the SAMLOGON test.
Andrew Bartlett [Tue, 4 Oct 2005 01:02:06 +0000 (01:02 +0000)]
r10703: Add a new user account, change the password and test it in the SAMLOGON test.

The semantics for the user account are very odd, the old password is
still valid, but the session keys appear to be blanked out.

Andrew Bartlett
(This used to be commit bbfaf4821d81116efa91313655acb75d6f577953)

16 years agor10702: Fix a silly error that caused a rejoin/delete in the torture code to fault...
Andrew Bartlett [Tue, 4 Oct 2005 01:01:07 +0000 (01:01 +0000)]
r10702: Fix a silly error that caused a rejoin/delete in the torture code to fault...

Andrew Bartlett
(This used to be commit 0fa6d040f287dbf6b0588e3507d53037a04f85d4)

16 years agor10701: Ensure we return the right user handle.
Andrew Bartlett [Tue, 4 Oct 2005 00:59:59 +0000 (00:59 +0000)]
r10701: Ensure we return the right user handle.

Andrew Bartlett
(This used to be commit 732b247a498e0b90b9f0c711baaac51ad6402496)

16 years agor10700: removed volkers temporary timer hack now that freeing the netlogon
Andrew Tridgell [Tue, 4 Oct 2005 00:46:31 +0000 (00:46 +0000)]
r10700: removed volkers temporary timer hack now that freeing the netlogon
pipe is safe while inside a rpc callback
(This used to be commit 5d752a519416c7a0c8c7d166f43eadc75cb5c37f)

16 years agor10699: fixed the dcerpc code so that you can shutdown the pipe safely from
Andrew Tridgell [Tue, 4 Oct 2005 00:43:16 +0000 (00:43 +0000)]
r10699: fixed the dcerpc code so that you can shutdown the pipe safely from
within a callback on the pipe. This should fix a problem volker
encountered with winbind. The fix invoolves making the recv_data
handler free the memory for a packet, instead of having the transport
layer free it after calling recv_data. When the transport layer freed
it, it had no way of knowing if the callback had shutdown the pipe, so
it had no way of knowing if it could safely use the pointer.

Also changed the pipe shutdown hook for the smb transport to use an
async SMB close. This ensures that when you shutdown the pipe, you
don't block waiting for the server to ack the close of the pipe fnum.
(This used to be commit c87d7f580e39245db181605f50883de07dd9632e)

16 years agor10698: Fix support of enums in switch_type() in the ethereal parser generator
Jelmer Vernooij [Mon, 3 Oct 2005 23:54:44 +0000 (23:54 +0000)]
r10698: Fix support of enums in switch_type() in the ethereal parser generator
(This used to be commit 9fac8dfeb393212e16be7156de466077bc90232d)

16 years agor10697: Change the torture join code to return a credentials structure, as
Andrew Bartlett [Mon, 3 Oct 2005 23:46:21 +0000 (23:46 +0000)]
r10697: Change the torture join code to return a credentials structure, as
that is what most of the callers want anyway.

Remove and re-add the account for the torture case, rather than just
modify it.

Test with a user account (needs work to change the password).

Andrew Bartlett
(This used to be commit 38bebef02454164cbe882347d80e03abee656205)

16 years agor10696: Return the realm to the caller, not NULL...
Andrew Bartlett [Mon, 3 Oct 2005 23:42:59 +0000 (23:42 +0000)]
r10696: Return the realm to the caller, not NULL...

Also return an indication of if the join was of a new account, or
reworking an existing account.

Andrew Bartlett
(This used to be commit b6e4b36c4f1f90e42dd0543538956a1d89e3724b)

16 years agor10695: strupper() of NULL should be NULL, not panic.
Andrew Bartlett [Mon, 3 Oct 2005 23:39:59 +0000 (23:39 +0000)]
r10695: strupper() of NULL should be NULL, not panic.

Andrew Bartlett
(This used to be commit b1bcc0dc0b3c6d4d931100dee21ba16cb5521978)

16 years agor10694: Add some work I did this afternoon on getting pidl to output Samba3
Jelmer Vernooij [Mon, 3 Oct 2005 23:27:33 +0000 (23:27 +0000)]
r10694: Add some work I did this afternoon on getting pidl to output Samba3
RPC parsers. Currently the following files can be generated:

- include/rpc_BASENAME.h
- rpc_server/srv_BASENAME.c
- rpc_server/srv_BASENAME_nt.c (template only, user has to fill in functions)
- rpc_client/cli_BASENAME.c
- rpc_parse/parse_BASENAME.c

So far, I have been working on getting DFS working. Currently still to do
(all in rpc_parse/parse_BASENAME.c):
 - Proper handling of declarations
 - Proper handling of scalar/buffer parts of structs and unions
 - Subcontexts
 - Proper handling of arrays
 - Support for custom (non-scalar) types

I hope to have a somewhat more working version later this week.

Some files as currently generated are available from:
http://samba.org/~jelmer/pidl_samba3/
(This used to be commit 8af8eaeeef6d46f4d25ccb1d25890e1eef063e4f)

16 years agor10691: This gets half-way to wbinfo -n. It acquires an lsa pipe, and does a
Volker Lendecke [Mon, 3 Oct 2005 17:36:49 +0000 (17:36 +0000)]
r10691: This gets half-way to wbinfo -n. It acquires an lsa pipe, and does a
queryinfopolicy. Idea is to get a consistency check between that and our
notion of the domain name and sid, and take the lsa pipe as the holder of the
central smbcli_tree that netlogon and samr use as well.

Volker
(This used to be commit 126c80aefc4f53c4ba79afc12d70602ef9055ddb)

16 years agor10690: Fix a bug that metze pointed out: Leaving the "rejecting" destructor around
Volker Lendecke [Mon, 3 Oct 2005 17:36:10 +0000 (17:36 +0000)]
r10690: Fix a bug that metze pointed out: Leaving the "rejecting" destructor around
prevents the memory from being freed.

Thanks,

Volker
(This used to be commit df8eeb01f498568207a4a8d5d12348c473f41799)

16 years agor10687: Another one...
Volker Lendecke [Mon, 3 Oct 2005 15:19:08 +0000 (15:19 +0000)]
r10687: Another one...
(This used to be commit d18f7edf92e8420f30cae01649d18f0ae20a80aa)

16 years agor10686: Fix the build
Volker Lendecke [Mon, 3 Oct 2005 15:03:31 +0000 (15:03 +0000)]
r10686: Fix the build
(This used to be commit a7137fd3ec2e484c8feb73fa228031c8b75107bf)

16 years agor10685: Why wait 5 seconds...
Volker Lendecke [Mon, 3 Oct 2005 14:50:07 +0000 (14:50 +0000)]
r10685: Why wait 5 seconds...
(This used to be commit 820b4180dd3c1d07dd529fcb654ea01407a481cb)

16 years agor10684: Add a nasty hack for the failure case of wbinfo -t. Tridge has a proper fix
Volker Lendecke [Mon, 3 Oct 2005 14:24:53 +0000 (14:24 +0000)]
r10684: Add a nasty hack for the failure case of wbinfo -t. Tridge has a proper fix
for it pending.

Also fix a bug with timed events: Don't call the same event recursively in the
handler's inner semi-async event loop.

Volker
(This used to be commit e38e50127a3414461578421e676a9c58c106c272)

16 years agor10683: Samba3's wbinfo -t should give the correct answer now.
Volker Lendecke [Mon, 3 Oct 2005 13:46:11 +0000 (13:46 +0000)]
r10683: Samba3's wbinfo -t should give the correct answer now.

Tridge, if you have time, you might want to look at the segfault I was still
seeing. Now I store the handle to the netlogon pipe in the global winbind
state and free it on the next entry into check_machacc. The problem seems to
be that talloc_free()ing a pipe struct from within a callback function on that
pipe is not possible. I think I can live with that, but it has been not really
obvious. To reproduce the segfault you might want to look at putting a
talloc_free(state->getcreds->out.netlogon) into
wbsrv_samba3_check_machacc_receive_creds. This is called from a dcerpc
callback function.

In particular if the check failed it would be nice if I could delete the pipe
directly and not post a different event to some winbind queue.

I tried to delete the pipe from a timed event triggered immediately, but this
also fails because the inner loop seems to hit the same event again, calling
it twice.

Volker
(This used to be commit 5436d7764812bb632ba865e633005ed07923b57f)

16 years agor10682: force the free of the fd event first when a stream terminates. That ensures
Andrew Tridgell [Mon, 3 Oct 2005 09:36:52 +0000 (09:36 +0000)]
r10682: force the free of the fd event first when a stream terminates. That ensures
destructors hanging off the stream connection don't trip more socket
events.

this should help with the problem volker described
(This used to be commit fd8eccd5bf0c742a7809f3991101f4ac75b5e037)

16 years agor10681: Convert dcerpc_open_smb to a composite function.
Volker Lendecke [Mon, 3 Oct 2005 09:00:36 +0000 (09:00 +0000)]
r10681: Convert dcerpc_open_smb to a composite function.

Volker
(This used to be commit 42ff218ac98fab00bd58c4f50f11843ef32b4698)

16 years agor10680: Fix a warning.
Rafal Szczesniak [Sun, 2 Oct 2005 23:01:25 +0000 (23:01 +0000)]
r10680: Fix a warning.

rafal
(This used to be commit b3b4059efeb7f5d2795b4bb93cf8ab95aad54c8e)

16 years agor10679: Monitor messages should be issued from usermod functions.
Rafal Szczesniak [Sun, 2 Oct 2005 19:59:24 +0000 (19:59 +0000)]
r10679: Monitor messages should be issued from usermod functions.
Also a bit of formatting.

rafal
(This used to be commit 1fefca2c172085d6bc05bfac1c10e52066e42606)

16 years agor10678: Add debug helper
Jelmer Vernooij [Sun, 2 Oct 2005 14:29:08 +0000 (14:29 +0000)]
r10678: Add debug helper
Fix push for non-fixed length strings
(This used to be commit 821b0bb56f6d6b9df7bbdeeddc206505f3f0a7c2)

16 years agor10677: Add smb_composite_connectmulti: Send out multiple SYN packets at once, use the
Volker Lendecke [Sun, 2 Oct 2005 10:02:35 +0000 (10:02 +0000)]
r10677: Add smb_composite_connectmulti: Send out multiple SYN packets at once, use the
first one that replies correctly.

Add a talloc context to smb_composite_connect()

Volker
(This used to be commit 6b88de182e40cb00a833c085f801fd47c92bbe94)

16 years agor10675: Connect to the DC's IPC$
Volker Lendecke [Sat, 1 Oct 2005 16:36:04 +0000 (16:36 +0000)]
r10675: Connect to the DC's IPC$

Volker
(This used to be commit c7557884843a5b2bac9e21ec81cafcaadf436bca)

16 years agor10670: Add notes on things that are TODO in Samba4 kerberos land.
Andrew Bartlett [Sat, 1 Oct 2005 01:19:12 +0000 (01:19 +0000)]
r10670: Add notes on things that are TODO in Samba4 kerberos land.

Andrew Bartlett
(This used to be commit 5b2114bb9c604e8d36887e1131175da327eabc84)

16 years agor10669: reverted jelmers commit 10663 as it was causing lots of panics in 'make test'
Andrew Tridgell [Sat, 1 Oct 2005 01:04:34 +0000 (01:04 +0000)]
r10669: reverted jelmers commit 10663 as it was causing lots of panics in 'make test'

I also think the method of getting rid of pstring isn't the right
one. I certainly do want to get rid of pstring/fstring, but the reason
for removing them is the use of arbitrary sized fixed length strings
on the stack and in structures. Changing to another fixed length stack
string format isn't really a win, and moving to use strncpy() is
actually worse than pstrcpy() as strncpy() has the absolutely awful
semantics of always zeroing all remaining bytes, so it ends up taking
a lot of cpu doing pointless memory writes.

I'd rather move to more use of asprintf()/talloc_asprintf() and
similar functions for dynamic string allocation.

You also have to be very careful about some of these system defined
string limits. One some systems PATH_MAX could be 64k or even larger,
which can quickly blow the stack out when you allocate a few of them.
(This used to be commit 194efd26e42d621b239052ed1fec8da916bd2144)

16 years agor10668: added a ildap_search_bytree() function
Andrew Tridgell [Fri, 30 Sep 2005 23:56:54 +0000 (23:56 +0000)]
r10668: added a ildap_search_bytree() function
(This used to be commit fd6d895ebdb201ac6afaf5c8ec84d003765cdff6)

16 years agor10667: cope with a NULL tree for base searches in ldb_search()
Andrew Tridgell [Fri, 30 Sep 2005 23:47:40 +0000 (23:47 +0000)]
r10667: cope with a NULL tree for base searches in ldb_search()
(This used to be commit 26ff53857802ae4a63f2b6e46c9caa7ca2fbbe89)

16 years agor10666: - reverse the ildap ldb backend so tree based searches go through
Andrew Tridgell [Fri, 30 Sep 2005 23:46:41 +0000 (23:46 +0000)]
r10666: - reverse the ildap ldb backend so tree based searches go through
directly, and expression based searches are converted to trees. This
makes for less conversions.

- allow the caller to supply a set of credentials via the ldb opaque
name 'credentials'. I will be using this in my ldb proxy module.
(This used to be commit af24f3d7faac6ef74feef73a23345d8c484da07c)

16 years agor10665: fixed some crash errors and an error encoding AND and OR operations in the...
Andrew Tridgell [Fri, 30 Sep 2005 23:14:30 +0000 (23:14 +0000)]
r10665: fixed some crash errors and an error encoding AND and OR operations in the expression parsing code
(This used to be commit 0d4a900ce5705856d61c6dd4ccb8fdbd049d22b7)

16 years agor10664: Include limits.h in replace.h for HOST_NAME_MAX
Jelmer Vernooij [Fri, 30 Sep 2005 23:10:20 +0000 (23:10 +0000)]
r10664: Include limits.h in replace.h for HOST_NAME_MAX
(This used to be commit dc3dc796746de672dbf3ad0e4715e8b30ae4afb7)

16 years agor10663: Eliminate use of pstring
Jelmer Vernooij [Fri, 30 Sep 2005 22:08:06 +0000 (22:08 +0000)]
r10663: Eliminate use of pstring
(This used to be commit 57cfbe51e61e27098fad3f9b1113c6f8f69f9a55)

16 years agor10662: Eliminate pstring and friends.
Jelmer Vernooij [Fri, 30 Sep 2005 21:48:25 +0000 (21:48 +0000)]
r10662: Eliminate pstring and friends.
(This used to be commit ecef1ee6cd0bdae2fcbff3f57edaa26eb72f3481)

16 years agor10653: give unknown levels a name basicly the *2 levels mean
Stefan Metzmacher [Fri, 30 Sep 2005 15:28:41 +0000 (15:28 +0000)]
r10653: give unknown levels a name basicly the *2 levels mean
that the update or inform should be propagaded to all servers

metze
(This used to be commit 33c7c101c024818694830ac11fefe70bd2edb1d4)

16 years agor10646: Hey Jelmer what do you think of this? The SConscript for the libcli
Tim Potter [Fri, 30 Sep 2005 07:30:37 +0000 (07:30 +0000)]
r10646: Hey Jelmer what do you think of this?  The SConscript for the libcli
directory now looks like the config.mk file but with different
punctuation.

The only weird bit is that it creates a proto.h file for each subsystem.
(This used to be commit 09d4abecb01fa9159243cfcb33051092f92cef3b)

16 years agor10643: increase smbd max runtime when using valgrind
Andrew Tridgell [Fri, 30 Sep 2005 04:52:21 +0000 (04:52 +0000)]
r10643: increase smbd max runtime when using valgrind
(This used to be commit c394a4a602edba49f00252d0376050c0bdc2f6a8)