14 years agor4585: don't consider LookupSids3 failing with NT_STATUS_ACCESS_DENIED (as w2k3 does) or
Andrew Tridgell [Fri, 7 Jan 2005 03:11:18 +0000 (03:11 +0000)]
r4585: don't consider LookupSids3 failing with NT_STATUS_ACCESS_DENIED (as w2k3 does) or
NT_STATUS_RPC_PROTSEQ_NOT_SUPPORTED (as longhorn does) to be an error.

fixed the CreateTrustedDomain test to cope with the "torturedomain" being left over
from a previous aborted run

14 years agor4584: fix pvfs backend to pass the new enhanced RAW-ACLS test. Easy once I really the
Andrew Tridgell [Fri, 7 Jan 2005 02:14:34 +0000 (02:14 +0000)]
r4584: fix pvfs backend to pass the new enhanced RAW-ACLS test. Easy once I really the
strange behaviour I saw was a w2k3 bug :-)

14 years agor4583: print which bit failed in the owner bits check
Andrew Tridgell [Fri, 7 Jan 2005 02:13:07 +0000 (02:13 +0000)]
r4583: print which bit failed in the owner bits check

14 years agor4582: finally worked out what is going on with the inherited ACLs test and win2003...
Andrew Tridgell [Fri, 7 Jan 2005 01:56:19 +0000 (01:56 +0000)]
r4582: finally worked out what is going on with the inherited ACLs test and win2003. It is a
win2003 bug!

This new test code works against w2k, and against longhorn, but fails
against w2k3. When tested against w2k3 it allows a open with an access
mask that should be denied by the given ACL, after setting up the ACL
using inheritance. Note that only the very specific
SEC_RIGHTS_FILE_ALL mask incorrectly succeeds, so they must have a
special case for that mask. Maybe its an optimisation gone wrong?

I don't know if there are any serious security implications to this,
but it is pretty clearly wrong, and has been fixed in longhorn.

14 years agor4578: Minor cleanup of ndr_pull_array() and ndr_pull_array_foo() regexps.
Tim Potter [Thu, 6 Jan 2005 23:08:30 +0000 (23:08 +0000)]
r4578: Minor cleanup of ndr_pull_array() and ndr_pull_array_foo() regexps.

14 years agor4572: remove configure and include/config.h*
Stefan Metzmacher [Thu, 6 Jan 2005 17:21:57 +0000 (17:21 +0000)]
r4572: remove configure and include/config.h*
before running autoheader && autoconf

this fixes bug where configure didn't get correctly updated
(I assume autoconf uses some caching...)


14 years agor4568: make use of SidType and move it to lsa.idl
Stefan Metzmacher [Thu, 6 Jan 2005 13:34:18 +0000 (13:34 +0000)]
r4568: make use of SidType and move it to lsa.idl


14 years agor4567: Fix the build for metze.
Andrew Bartlett [Thu, 6 Jan 2005 12:22:20 +0000 (12:22 +0000)]
r4567: Fix the build for metze.

Andrew Bartlett

14 years agor4566: Fix Samba4 to pass it's own RPC-SAMLOGON torture test.
Andrew Bartlett [Thu, 6 Jan 2005 12:13:31 +0000 (12:13 +0000)]
r4566: Fix Samba4 to pass it's own RPC-SAMLOGON torture test.

Include RPC-SAMLOGON in the list of tests expected to pass

Remove silly extra loops from the RPC-SAMLOGON test, which mostly just
slowed htings down.

Andrew Bartlett

14 years agor4565: Make the order of the initialisation more sensible.
Andrew Bartlett [Thu, 6 Jan 2005 12:10:54 +0000 (12:10 +0000)]
r4565: Make the order of the initialisation more sensible.

Andrew Bartlett

14 years agor4564: added a comment on lsa_EnumAccounts IDL
Andrew Tridgell [Thu, 6 Jan 2005 12:02:39 +0000 (12:02 +0000)]
r4564: added a comment on lsa_EnumAccounts IDL

14 years agor4563: fixed lsa_EnumAccounts() server side to return all accounts that have privileg...
Andrew Tridgell [Thu, 6 Jan 2005 12:01:11 +0000 (12:01 +0000)]
r4563: fixed lsa_EnumAccounts() server side to return all accounts that have privileges, as
volker discovered

14 years agor4560: - fixed crash bugs in the RPC-SCHANNEL and RPC-NETLOGON tests
Andrew Tridgell [Thu, 6 Jan 2005 09:26:14 +0000 (09:26 +0000)]
r4560: - fixed crash bugs in the RPC-SCHANNEL and RPC-NETLOGON tests

- disabled the async netlogon tests for now until we better understand async rpc

- added a script that runs rpc tests that a w2k3 DC
  should pass. This is useful for regression testing when PIDL changes
  are made.

14 years agor4559: prevent the RPC-EPMAPPER test from looping forever against w2k3
Andrew Tridgell [Thu, 6 Jan 2005 09:24:41 +0000 (09:24 +0000)]
r4559: prevent the RPC-EPMAPPER test from looping forever against w2k3

14 years agor4558: more use of bitmaps and enums
Stefan Metzmacher [Thu, 6 Jan 2005 07:34:14 +0000 (07:34 +0000)]
r4558: more use of bitmaps and enums


14 years agor4557: support for [flags()] on typedef enum|bitmap
Stefan Metzmacher [Thu, 6 Jan 2005 07:29:55 +0000 (07:29 +0000)]
r4557: support for [flags()] on typedef enum|bitmap

NDR_PAHEX is handled by ndr_print_enum() now


14 years agor4556: neater (and faster) way of doing alignments and scalars
Andrew Tridgell [Thu, 6 Jan 2005 07:22:47 +0000 (07:22 +0000)]
r4556: neater (and faster) way of doing alignments and scalars

14 years agor4555: fix some dependencies
Stefan Metzmacher [Thu, 6 Jan 2005 07:02:46 +0000 (07:02 +0000)]
r4555: fix some dependencies


14 years agor4554: create svcctl_ServerType bitmap
Stefan Metzmacher [Thu, 6 Jan 2005 07:00:54 +0000 (07:00 +0000)]
r4554: create svcctl_ServerType bitmap


14 years agor4553: fix typo
Stefan Metzmacher [Thu, 6 Jan 2005 06:59:14 +0000 (06:59 +0000)]
r4553: fix typo


14 years agor4552: use samr_AcctFlags in netlogon.idl
Stefan Metzmacher [Thu, 6 Jan 2005 06:34:07 +0000 (06:34 +0000)]
r4552: use samr_AcctFlags in netlogon.idl


14 years agor4551: add support for a pidl extensions
Stefan Metzmacher [Thu, 6 Jan 2005 06:32:07 +0000 (06:32 +0000)]
r4551: add support for a pidl extensions

'declare bitmap foo1;'
'declare enum foo2;'

and also allow

typedef [public] bitmap ...
typedef [public] enum ...

you need to a forward declaration of bitmaps and enums

when you want to use them in another idl file,
and you need to make the real declaration to be public

see the next commit to samr.idl and netlogon.idl


14 years agor4550: talloc() is now typesafe. It is exactly equivalent to the old talloc_p() macro...
Andrew Tridgell [Thu, 6 Jan 2005 03:20:56 +0000 (03:20 +0000)]
r4550: talloc() is now typesafe. It is exactly equivalent to the old talloc_p() macro. Use
talloc_size() if you want the old behaviour.

I have kept talloc_p() as an alias for now. Once we change all calls
to be plain talloc() then we can remove it.

14 years agor4549: got rid of a lot more uses of plain talloc(), instead using
Andrew Tridgell [Thu, 6 Jan 2005 03:06:58 +0000 (03:06 +0000)]
r4549: got rid of a lot more uses of plain talloc(), instead using
talloc_size() or talloc_array_p() where appropriate.

also fixed a memory leak in pvfs_copy_file() (failed to free a memory

14 years agor4548: Convert to talloc_p() and talloc_array_p() where appropriate.
Tim Potter [Thu, 6 Jan 2005 02:36:59 +0000 (02:36 +0000)]
r4548: Convert to talloc_p() and talloc_array_p() where appropriate.

(swig stuff seems broken atm though)

14 years agor4547: - added talloc_new(ctx) macro that is a neater form of the common talloc(ctx...
Andrew Tridgell [Thu, 6 Jan 2005 02:32:43 +0000 (02:32 +0000)]
r4547: - added talloc_new(ctx) macro that is a neater form of the common talloc(ctx, 0) call.

- cleaned up some talloc usage in various files

I'd like to get to the point that we have no calls to talloc(), at
which point we will rename talloc_p() to talloc(), to encourage
everyone to use the typesafe functions.

14 years agor4546: Use talloc_p() instad of talloc()
Tim Potter [Thu, 6 Jan 2005 02:10:33 +0000 (02:10 +0000)]
r4546: Use talloc_p() instad of talloc()

14 years agor4542: use bitmap for samr_FieldsPresent and samr_AcctFlags
Stefan Metzmacher [Wed, 5 Jan 2005 16:54:51 +0000 (16:54 +0000)]
r4542: use bitmap for samr_FieldsPresent and samr_AcctFlags


14 years agor4541: make use of new enum and bitmap features
Stefan Metzmacher [Wed, 5 Jan 2005 16:48:58 +0000 (16:48 +0000)]
r4541: make use of new enum and bitmap features


14 years agor4540: sync enum and bitmap code
Stefan Metzmacher [Wed, 5 Jan 2005 16:31:05 +0000 (16:31 +0000)]
r4540: sync enum and bitmap code


14 years agor4536: make use of the new enum and bitmap feature of pidl
Stefan Metzmacher [Wed, 5 Jan 2005 15:37:43 +0000 (15:37 +0000)]
r4536: make use of the new enum and bitmap feature of pidl


14 years agor4535: add full support for
Stefan Metzmacher [Wed, 5 Jan 2005 15:36:26 +0000 (15:36 +0000)]
r4535: add full support for

typedef bitmap {
    FLAG1 = 0x01
} fooflags;

typedef struct {
    fooflags flags;


14 years agor4534: update the yapp generated code
Stefan Metzmacher [Wed, 5 Jan 2005 15:28:10 +0000 (15:28 +0000)]
r4534: update the yapp generated code


14 years agor4533: parsing support for:
Stefan Metzmacher [Wed, 5 Jan 2005 15:26:48 +0000 (15:26 +0000)]
r4533: parsing support for:

typedef [bitmap16bit] bitmap {
FLAG1 = 0x0001,
FLAG2 = 0x8000
} fooflags;


14 years agor4532: - rename bitmap -> bits
Stefan Metzmacher [Wed, 5 Jan 2005 15:24:20 +0000 (15:24 +0000)]
r4532: - rename bitmap -> bits

the next commit is support for typedef bitmap {...}; in pidl


14 years agor4531: Include the OID locally, as it seems to be hard to get the includes
Andrew Bartlett [Wed, 5 Jan 2005 10:21:08 +0000 (10:21 +0000)]
r4531: Include the OID locally, as it seems to be hard to get the includes

Andrew Bartlett

14 years agor4530: Start adding a bit of Doxygen compatible documentation comments to GENSEC.
Andrew Bartlett [Wed, 5 Jan 2005 03:21:45 +0000 (03:21 +0000)]
r4530: Start adding a bit of Doxygen compatible documentation comments to GENSEC.

Andrew Bartlett

14 years agor4526: - much simpler (and more accurate!) ndr_size_*() code generation. It
Andrew Tridgell [Wed, 5 Jan 2005 02:01:19 +0000 (02:01 +0000)]
r4526: - much simpler (and more accurate!) ndr_size_*() code generation. It
  is less efficient, but I really doubt that matters.

- use enum in epmapper.idl for protocol type

- added support for "enum8bit" flag, used in epmapper.idl

14 years agor4524: converted a few bits of samr.idl to use enum properly
Andrew Tridgell [Wed, 5 Jan 2005 00:41:49 +0000 (00:41 +0000)]
r4524: converted a few bits of samr.idl to use enum properly

14 years agor4523: the PROPERTIES change for typedef was not quite as simple as I thought. This...
Andrew Tridgell [Wed, 5 Jan 2005 00:31:28 +0000 (00:31 +0000)]
r4523: the PROPERTIES change for typedef was not quite as simple as I thought. This puts in a workaround
that fixes things for now

14 years agor4522: PROPERTIES are now handled at the typedef level
Andrew Tridgell [Wed, 5 Jan 2005 00:14:29 +0000 (00:14 +0000)]
r4522: PROPERTIES are now handled at the typedef level

14 years agor4521: fixed up the handling of PROPERTIES elements after the change to support enum
Andrew Tridgell [Wed, 5 Jan 2005 00:09:29 +0000 (00:09 +0000)]
r4521: fixed up the handling of PROPERTIES elements after the change to support enum

14 years agor4520: added a enum test function to the echo pipe
Andrew Tridgell [Tue, 4 Jan 2005 23:27:12 +0000 (23:27 +0000)]
r4520: added a enum test function to the echo pipe

14 years agor4519: added the enum print function in ndr_basic.c
Andrew Tridgell [Tue, 4 Jan 2005 23:26:12 +0000 (23:26 +0000)]
r4519: added the enum print function in ndr_basic.c

14 years agor4518: added proper support for "typedef enum" in pidl. We can now use enums as types...
Andrew Tridgell [Tue, 4 Jan 2005 23:25:25 +0000 (23:25 +0000)]
r4518: added proper support for "typedef enum" in pidl. We can now use enums as types in switch
statements and variable types, and the value will be printed with its string name and value

14 years agor4517: Revert previous commit about giving arrays of scalars their own subtree.
Tim Potter [Tue, 4 Jan 2005 23:15:33 +0000 (23:15 +0000)]
r4517: Revert previous commit about giving arrays of scalars their own subtree.
Generate code to display using proto_tree_add_bytes() instead.

14 years agor4513: add experimental modification attribute checking
Simo Sorce [Tue, 4 Jan 2005 15:18:50 +0000 (15:18 +0000)]
r4513: add experimental modification attribute checking
still not complete but works with add and replace, not tested with delete
lot of duplicated code to remove

14 years agor4512: Give arrays of scalar types their own subtree.
Tim Potter [Tue, 4 Jan 2005 06:35:53 +0000 (06:35 +0000)]
r4512: Give arrays of scalar types their own subtree.

14 years agor4511: Remove a crapload of dead code.
Tim Potter [Tue, 4 Jan 2005 05:21:24 +0000 (05:21 +0000)]
r4511: Remove a crapload of dead code.

14 years agor4510: Some more tests for RPC-NETLOGON, checking the idea that we could
Andrew Bartlett [Tue, 4 Jan 2005 03:06:23 +0000 (03:06 +0000)]
r4510: Some more tests for RPC-NETLOGON, checking the idea that we could
combine the NTLM and LMv2 responses, for maximum compatability from a
client perspective, allowing access to servers that require NTLMv2, as
well as those that don't support it.

Currently, this is unfortunetly not possible against Win2k3 (and Samba
is being coded to match that behaviour at this point).

Andrew Bartlett

14 years agor4509: Add missing cr.
Tim Potter [Mon, 3 Jan 2005 23:28:45 +0000 (23:28 +0000)]
r4509: Add missing cr.

14 years agor4508: - simplify the code and allocate only in one place
Stefan Metzmacher [Mon, 3 Jan 2005 14:45:17 +0000 (14:45 +0000)]
r4508: - simplify the code and allocate only in one place

- the ZERO_STRUCT(r->out) is useless and wrong if we would have a [ref,out] parameter
  also note the r->out is already initialized by the pidl generated code


14 years agor4507: this default is unnneded because the switch level is in the r->in
Stefan Metzmacher [Mon, 3 Jan 2005 14:42:59 +0000 (14:42 +0000)]
r4507: this default is unnneded because the switch level is in the r->in


14 years agor4506: fix compiler warnings
Stefan Metzmacher [Mon, 3 Jan 2005 14:38:05 +0000 (14:38 +0000)]
r4506: fix compiler warnings


14 years agor4505: Add a first very basic schema module
Simo Sorce [Mon, 3 Jan 2005 14:05:47 +0000 (14:05 +0000)]
r4505: Add a first very basic schema module
To use it you should provide a schema.ldb file where the schema is stored
and load the module in the ldb you want to have schema check activated

more info soon.

currently schema checks are performed only on new object creation not on modifications


14 years agor4504: Setting
Andrew Bartlett [Mon, 3 Jan 2005 09:00:37 +0000 (09:00 +0000)]
r4504: Setting

.enabled = True

on modules we know are good (and we want on be default) seems neater.

Andrew Bartlett

14 years agor4503: add an error reporting function (fixes)
Simo Sorce [Mon, 3 Jan 2005 08:59:00 +0000 (08:59 +0000)]
r4503: add an error reporting function (fixes)

14 years agor4502: add an error reporting function
Simo Sorce [Mon, 3 Jan 2005 08:52:39 +0000 (08:52 +0000)]
r4502: add an error reporting function

14 years agor4501: when copying files it is common for clients to copy the ACL. When the
Andrew Tridgell [Mon, 3 Jan 2005 07:57:05 +0000 (07:57 +0000)]
r4501: when copying files it is common for clients to copy the ACL. When the
ACL is the default ACL this menas the copied file would have an xattr
but the original would not. Avoid this by checking if the ACL being
set is the original ACL, and avoid the copy.

14 years agor4500: Allow GENSEC modules to be disabled by setting a flag on their module
Andrew Bartlett [Mon, 3 Jan 2005 07:08:14 +0000 (07:08 +0000)]
r4500: Allow GENSEC modules to be disabled by setting a flag on their module
definition, not by hardcoded reference in loadparm.c

Andrew Bartlett

14 years agor4499: Almost make our Samba4 server pass the RPC-SAMLOGON torture test.
Andrew Bartlett [Mon, 3 Jan 2005 06:23:02 +0000 (06:23 +0000)]
r4499: Almost make our Samba4 server pass the RPC-SAMLOGON torture test.

I just need to fix a couple of NTLMv2 issues before we can fully pass,
and put this in, as a 'should pass' test.

Andrew Bartlett

14 years agor4498: arrgh, really fix the typo this time
Andrew Tridgell [Mon, 3 Jan 2005 02:53:06 +0000 (02:53 +0000)]
r4498: arrgh, really fix the typo this time

14 years agor4497: fixed a typo
Andrew Tridgell [Mon, 3 Jan 2005 02:47:00 +0000 (02:47 +0000)]
r4497: fixed a typo

14 years agor4496: expanded info on testing your filesystem for xattr support
Andrew Tridgell [Mon, 3 Jan 2005 02:26:14 +0000 (02:26 +0000)]
r4496: expanded info on testing your filesystem for xattr support

14 years agor4495: Some more regexps for handling stuff in the lsa and samr pipes -
Tim Potter [Mon, 3 Jan 2005 00:26:50 +0000 (00:26 +0000)]
r4495: Some more regexps for handling stuff in the lsa and samr pipes -
various types of arrays and relative structures.

14 years agor4494: Allow gensec_gssapi to use the SPNEGO mech provided by Heimdal (off by
Andrew Bartlett [Sun, 2 Jan 2005 23:53:14 +0000 (23:53 +0000)]
r4494: Allow gensec_gssapi to use the SPNEGO mech provided by Heimdal (off by
default at this point), and include the GSSAPI OIDs in our source, per
advice by lha that this is easier than getting the includes right.

Andrew Bartlett

14 years agor4493: change name of README.gcov so it doesn't get deleted by "make clean" :-)
Andrew Tridgell [Sun, 2 Jan 2005 23:48:07 +0000 (23:48 +0000)]
r4493: change name of README.gcov so it doesn't get deleted by "make clean" :-)

14 years agor4492: is a pointer that needs to be allocated before use
Andrew Tridgell [Sun, 2 Jan 2005 23:44:08 +0000 (23:44 +0000)]
r4492: is a pointer that needs to be allocated before use

14 years agor4491: don't dereference q1.out.sam unless we know it is not NULL
Andrew Tridgell [Sun, 2 Jan 2005 23:36:02 +0000 (23:36 +0000)]
r4491: don't dereference q1.out.sam unless we know it is not NULL

14 years agor4490: when implementing one rpc server call in terms of another call, you
Andrew Tridgell [Sun, 2 Jan 2005 23:31:12 +0000 (23:31 +0000)]
r4490: when implementing one rpc server call in terms of another call, you
must zero r.out before making the 2nd call if the 2nd call has any
non-ref out parameters. This is needed for the case where the 2nd call
fails, and the 1st call would then fill in its out fields based on
uninitialised memory.

14 years agor4489: Remove an obsolete file.
Tim Potter [Sun, 2 Jan 2005 23:25:25 +0000 (23:25 +0000)]
r4489: Remove an obsolete file.

14 years agor4488: removed an unused variable
Andrew Tridgell [Sun, 2 Jan 2005 23:09:11 +0000 (23:09 +0000)]
r4488: removed an unused variable

14 years agor4487: fixed the use of ldb_msg_add_*() in the samr password backend
Andrew Tridgell [Sun, 2 Jan 2005 23:05:27 +0000 (23:05 +0000)]
r4487: fixed the use of ldb_msg_add_*() in the samr password backend

14 years agor4486: fixed some memory leaks in the new ldb code, by ensuring that memory is always
Andrew Tridgell [Sun, 2 Jan 2005 23:03:50 +0000 (23:03 +0000)]
r4486: fixed some memory leaks in the new ldb code, by ensuring that memory is always
allocated as a child of the right context

14 years agor4485: remove unused TODO comment we already handle the DS_ROLE_STANDALONE_SERVER...
Stefan Metzmacher [Sun, 2 Jan 2005 16:25:30 +0000 (16:25 +0000)]
r4485: remove unused TODO comment we already handle the DS_ROLE_STANDALONE_SERVER case


14 years agor4484: - use the nTMixedDomain attribute to check if our domain is mixed mode or not
Stefan Metzmacher [Sun, 2 Jan 2005 16:24:05 +0000 (16:24 +0000)]
r4484: - use the nTMixedDomain attribute to check if our domain is mixed mode or not


14 years agor4483: - rename same structs, enums and uninons of the dssetup pipe
Stefan Metzmacher [Sun, 2 Jan 2005 16:11:59 +0000 (16:11 +0000)]
r4483: - rename same structs, enums and uninons of the dssetup pipe
  to match the style we are using in other pipes

- first fillin local vars and only set the out parameter on success

- for the server code only to the samdb lookup when it's needed

NOTE: the DsRoleGetPrimaryDomainInformation() code with DS_ROLE_MEMBER_SERVER
      is not tested yet, does someone has a w2k3 member server to test with?


14 years agor4482: in ads mixed mode there are also only one PDC and multiple BDC's
Stefan Metzmacher [Sun, 2 Jan 2005 16:06:21 +0000 (16:06 +0000)]
r4482: in ads mixed mode there are also only one PDC and multiple BDC's


14 years agor4481: get rid of the last leak
Andrew Tridgell [Sun, 2 Jan 2005 13:04:32 +0000 (13:04 +0000)]
r4481: get rid of the last leak

14 years agor4480: autofree the dcom proxy tables
Andrew Tridgell [Sun, 2 Jan 2005 12:56:21 +0000 (12:56 +0000)]
r4480: autofree the dcom proxy tables

14 years agor4479: added the function talloc_autofree_context() which returns a talloc context...
Andrew Tridgell [Sun, 2 Jan 2005 12:55:33 +0000 (12:55 +0000)]
r4479: added the function talloc_autofree_context() which returns a talloc context that
will automatically be freed on program exit. This is useful for reducing
clutter in leak reports

14 years agor4477: expanded the test suite to increase code coverage a lot
Andrew Tridgell [Sun, 2 Jan 2005 09:46:59 +0000 (09:46 +0000)]
r4477: expanded the test suite to increase code coverage a lot

14 years agor4476: added a little gcov howto for ldb
Andrew Tridgell [Sun, 2 Jan 2005 08:53:58 +0000 (08:53 +0000)]
r4476: added a little gcov howto for ldb

14 years agor4475: fixed smbd to work with the small changes in the ldb API (the most important
Andrew Tridgell [Sun, 2 Jan 2005 07:51:13 +0000 (07:51 +0000)]
r4475: fixed smbd to work with the small changes in the ldb API (the most important
change was in the ldb_msg_add_*() routines, which now use the msg as a context,
and thus it needs to be a talloc ptr)

14 years agor4474: - converted ldb to use talloc internally
Andrew Tridgell [Sun, 2 Jan 2005 07:49:29 +0000 (07:49 +0000)]
r4474: - converted ldb to use talloc internally

- added gcov flags to Makefile.ldb

- expanded ldb test suite to get more coverage

14 years agor4473: - moved talloc into its own lib/talloc/ area
Andrew Tridgell [Sun, 2 Jan 2005 07:47:34 +0000 (07:47 +0000)]
r4473: - moved talloc into its own lib/talloc/ area

- added gcov flags to Makefile.talloc

- expanded talloc testsuite to add a test for realloc with a child ptr

- fixed a bug in talloc_realloc() with realloc of a ptr that has child ptrs

14 years agor4472: improve the discard_const() macro for standalone build of talloc
Andrew Tridgell [Sun, 2 Jan 2005 04:21:32 +0000 (04:21 +0000)]
r4472: improve the discard_const() macro for standalone build of talloc

14 years agor4471: Running 'make eparser_idl' doesn't need --header anymore.
Tim Potter [Sun, 2 Jan 2005 03:24:57 +0000 (03:24 +0000)]
r4471: Running 'make eparser_idl' doesn't need --header anymore.

14 years agor4470: Try not to have GSSAPI built unless we detected krb5. We should split
Andrew Bartlett [Sun, 2 Jan 2005 03:10:28 +0000 (03:10 +0000)]
r4470: Try not to have GSSAPI built unless we detected krb5.  We should split
these tests out a bit, but for now it's an indicator we can use.

Andrew Bartlett

14 years agor4469: Version n + 1 of the pidl ethereal parser generator. This version is
Tim Potter [Sun, 2 Jan 2005 00:00:43 +0000 (00:00 +0000)]
r4469: Version n + 1 of the pidl ethereal parser generator.  This version is
based on the idea of manipulating the .c and .h files generated by with perl regexps and glueing it all together to make an
ethereal plugin.

I thought this was a pretty crazy idea to start off with but it has
turned out to be not as complicated as I thought and has the huge advantage
of not duplicating any of the difficult code in

14 years agor4468: Don't be a slacker and leave out parameter names when generating
Tim Potter [Sat, 1 Jan 2005 23:18:59 +0000 (23:18 +0000)]
r4468: Don't be a slacker and leave out parameter names when generating
prototypes for ndr push/pull/print functions.

14 years agor4467: - tdb standalone build doesn't need -DSTANDALONE any more
Andrew Tridgell [Sat, 1 Jan 2005 05:08:41 +0000 (05:08 +0000)]
r4467: - tdb standalone build doesn't need -DSTANDALONE any more

- fixed standalone build

14 years agor4466: rather than defining "STANDALONE" for building tdb, ldb and talloc
Andrew Tridgell [Sat, 1 Jan 2005 05:06:22 +0000 (05:06 +0000)]
r4466: rather than defining "STANDALONE" for building tdb, ldb and talloc
outside the tree, instead defined _SAMBA_BUILD_ inside the Samba
build. This makes it easier to pull code out of Samba for external

14 years agor4465: remove unused file
Andrew Tridgell [Sat, 1 Jan 2005 04:40:11 +0000 (04:40 +0000)]
r4465: remove unused file

14 years agor4464: added pvfs backend support for the special CREATOR_OWNER and CREATOR_GROUP...
Andrew Tridgell [Sat, 1 Jan 2005 04:25:46 +0000 (04:25 +0000)]
r4464: added pvfs backend support for the special CREATOR_OWNER and CREATOR_GROUP inheritance rules

14 years agor4463: added testing of the special SID_CREATOR_OWNER inheritance rules
Andrew Tridgell [Sat, 1 Jan 2005 03:46:55 +0000 (03:46 +0000)]
r4463: added testing of the special SID_CREATOR_OWNER inheritance rules

14 years agor4462: - enable DSSETUP on ncalrpc
Andrew Tridgell [Sat, 1 Jan 2005 01:40:45 +0000 (01:40 +0000)]
r4462: - enable DSSETUP on ncalrpc

- add DSSETUP to the list of tests run in

14 years agor4461: finished the remaining information levels in the DSSETUP pipe. The pipe is...
Andrew Tridgell [Sat, 1 Jan 2005 01:32:01 +0000 (01:32 +0000)]
r4461: finished the remaining information levels in the DSSETUP pipe. The pipe is now complete!

The only glitch is that I am returning DS_ROLE_MEMBER_SERVER when I
should be returning DS_ROLE_PRIMARY_DC. This is needed for the moment
or ACL editing doesn't work from w2k3. Once we have some more ADS
calls we should be able to fix this.

14 years agor4460: Add a new GENSEC module: gensec_gssapi
Andrew Bartlett [Sat, 1 Jan 2005 00:19:08 +0000 (00:19 +0000)]
r4460: Add a new GENSEC module: gensec_gssapi
(disabled by default, set parametric option: gensec:gssapi=yes to enable).

This module backs directly onto GSSAPI, and allows us to sign and seal
GSSAPI/Krb5 connections in particular.  This avoids me reinventing the
entire GSSAPI wheel.

Currently a lot of things are left as default - we will soon start
specifiying OIDs as well as passwords (it uses the keytab only at the
moment).  Tested with our LDAP-* torture tests against Win2k3.

My hope is to use this module to access the new SPNEGO implementation
in Heimdal, to avoid having to standards-verify our own.

Andrew Bartlett

14 years agor4459: GENSEC refinements:
Andrew Bartlett [Fri, 31 Dec 2004 22:45:11 +0000 (22:45 +0000)]
r4459: GENSEC refinements:

In developing a GSSAPI plugin for GENSEC, it became clear that the API
needed to change:
 - GSSAPI exposes only a wrap() and unwrap() interface, and determines
   the location of the signature itself.
 - The 'have feature' API did not correctly function in the recursive
   SPNEGO environment.

As such, NTLMSSP has been updated to support these methods.

The LDAP client and server have been updated to use the new wrap() and
unwrap() methods, and now pass the LDAP-* tests in our smbtorture.
(Unfortunely I still get valgrind warnings, in the code that was
previously unreachable).

Andrew Bartlett