Stefan Metzmacher [Fri, 2 Nov 2007 09:26:49 +0000 (10:26 +0100)]
r25792: use NT_STATUS_HAVE_NO_MEMORY() and NDR_CHECK()
metze
Stefan Metzmacher [Fri, 2 Nov 2007 08:15:12 +0000 (09:15 +0100)]
r25791: wrepl_request_finished() sets req->status, so make this explicit
and avoid req->status = req->status...
metze
Stefan Metzmacher [Fri, 2 Nov 2007 07:49:23 +0000 (08:49 +0100)]
r25790: - remove some nesting by using error and out logic
- print in the print_* function ...
metze
Stefan Metzmacher [Fri, 2 Nov 2007 07:16:32 +0000 (08:16 +0100)]
r25789: print out what error happened...
metze
Andrew Bartlett [Fri, 2 Nov 2007 02:39:24 +0000 (03:39 +0100)]
r25788: Use a single routine to handle the creation of modify requests in the
linked_attributs code.
This drasticly reduces the code duplication here.
Andrew Bartlett
Andrew Bartlett [Fri, 2 Nov 2007 01:51:54 +0000 (02:51 +0100)]
r25787: Assert that we handle the group membership updating correctly,
including when we delete members from the DB.
Andrew Bartlett
Andrew Bartlett [Thu, 1 Nov 2007 21:01:48 +0000 (22:01 +0100)]
r25786: Add function required by linked_attributes module (We need a list of
all linked attributes in the schema, so we can try and find them).
Andrew Bartlett
Ronnie Sahlberg [Thu, 1 Nov 2007 19:30:55 +0000 (20:30 +0100)]
r25785: more small updates to frsrpc
tag 18 contains a timestamp
Stefan Metzmacher [Thu, 1 Nov 2007 12:22:20 +0000 (13:22 +0100)]
r25784: fix compiler warning
metze
Stefan Metzmacher [Thu, 1 Nov 2007 12:10:59 +0000 (13:10 +0100)]
r25783: socket_wrapper: don't include "includes.h"
metze
Stefan Metzmacher [Thu, 1 Nov 2007 11:39:12 +0000 (12:39 +0100)]
r25782: use NT_STATUS_HAVE_NO_MEMORY()
metze
Andrew Bartlett [Thu, 1 Nov 2007 11:34:06 +0000 (12:34 +0100)]
r25781: Handle and test linked attribute renames.
Andrew Bartlett
Stefan Metzmacher [Thu, 1 Nov 2007 10:43:00 +0000 (11:43 +0100)]
r25780: fix bool return
metze
Stefan Metzmacher [Thu, 1 Nov 2007 09:15:13 +0000 (10:15 +0100)]
r25779: the return value of ndr_pull_set_switch_value() should be checked
metze
Stefan Metzmacher [Thu, 1 Nov 2007 09:13:36 +0000 (10:13 +0100)]
r25778: use NT_STATUS_HAVE_NO_MEMORY() and NDR_CHECK() macros
metze
Stefan Metzmacher [Thu, 1 Nov 2007 08:50:24 +0000 (09:50 +0100)]
r25777: add NDR_ERR_UNREAD_BYTES and don't use NTSTATUS directly
metze
Stefan Metzmacher [Thu, 1 Nov 2007 08:48:42 +0000 (09:48 +0100)]
r25776: ndr compression: make use of NDR_CHECK() and not use NTSTATUS directly
metze
Stefan Metzmacher [Thu, 1 Nov 2007 07:15:41 +0000 (08:15 +0100)]
r25775: use ndr_pull_union_blob_all() in CLDAP code
metze
Stefan Metzmacher [Thu, 1 Nov 2007 07:10:54 +0000 (08:10 +0100)]
r25774: create ndr_pull_union_blob_all() and match the logic of the struct fns
metze
Ronnie Sahlberg [Thu, 1 Nov 2007 07:00:36 +0000 (08:00 +0100)]
r25773: update frsrpc.idl and add some more comments
Ronnie Sahlberg [Thu, 1 Nov 2007 03:00:05 +0000 (04:00 +0100)]
r25772: rename the tail fields
Ronnie Sahlberg [Thu, 1 Nov 2007 02:20:17 +0000 (03:20 +0100)]
r25771: In traces unknown7 has the same pattern as a UNIQUE pointer referral id
would often have.
Make a note that this may be a UNIQUE pointer.
Ronnie Sahlberg [Thu, 1 Nov 2007 02:17:55 +0000 (03:17 +0100)]
r25770: rename "unknown5" to tlv_size
Ronnie Sahlberg [Thu, 1 Nov 2007 02:16:28 +0000 (03:16 +0100)]
r25769: it looks like there are 9 uint32 before the TLV subcontext starts, not 6
Stefan Metzmacher [Wed, 31 Oct 2007 15:44:42 +0000 (16:44 +0100)]
r25768: pidl: NT_STATUS_IS_ERR() is NOT the same as !NT_STATUS_IS_OK()
Everything but success should be handled as error in the tests.
metze
Stefan Metzmacher [Wed, 31 Oct 2007 15:29:32 +0000 (16:29 +0100)]
r25767: pidl: make it easier to debug errors in pidl tests
we now print the C program that we tried to compile
metze
Stefan Metzmacher [Wed, 31 Oct 2007 15:27:21 +0000 (16:27 +0100)]
r25766: pidl: fix bugs in ndr_tagtype tests found by compiler warnings
metze
Stefan Metzmacher [Wed, 31 Oct 2007 15:25:44 +0000 (16:25 +0100)]
r25765: pidl: fix compiler warning in ndr_align test
metze
Jelmer Vernooij [Wed, 31 Oct 2007 13:48:48 +0000 (14:48 +0100)]
r25764: Fix total number of tests count to not include skipped tests.
Andrew Bartlett [Wed, 31 Oct 2007 04:14:31 +0000 (05:14 +0100)]
r25763: Handle modifies, in the easy case (add/delete of elements), for the
linked attributes.
Andrew Bartlett
Andrew Bartlett [Wed, 31 Oct 2007 03:41:36 +0000 (04:41 +0100)]
r25762: This test belongs best with the other checks for a valid parent, in
the objectclass module.
Andrew Bartlett
Andrew Bartlett [Wed, 31 Oct 2007 02:56:13 +0000 (03:56 +0100)]
r25761: Rename to be a DN to be a child of itself wasn't being checked for.
This prevents CN=test,dc=samba,dc=example,dc=com being renamed into
CN=test2,cn=test,dc=samba,dc=example,dc=com
Andrew Bartlett
Andrew Bartlett [Wed, 31 Oct 2007 01:50:13 +0000 (02:50 +0100)]
r25760: Test out relative distinguished name behaviour under renames.
Andrew Bartlett
Andrew Bartlett [Wed, 31 Oct 2007 00:19:20 +0000 (01:19 +0100)]
r25759: catch up the samba3sam tests with the change from cn=rootdse to @ROOTDSE
Andrew Bartlett
Jelmer Vernooij [Tue, 30 Oct 2007 23:14:54 +0000 (00:14 +0100)]
r25756: Add missing file with quick tests.
Andrew Bartlett [Tue, 30 Oct 2007 22:35:04 +0000 (23:35 +0100)]
r25755: Fix a couple of memory leaks, in particular a new leak onto the NULL
context caused by my objectclass module work.
Andrew Bartlett
Andrew Bartlett [Tue, 30 Oct 2007 20:03:54 +0000 (21:03 +0100)]
r25754: More work on normal forms for ldb input.
This patch is to ensure that all attributes are in the same case as
the schema specifies. In the process, I ensure that all attributes
are indeed in the schema.
This ensures we use the schema case, not the user supplied case for
future responses, which assists any (incorrect, but possible) case
sensitive processing on a client.
I've also removed more of the subtle 'schema &&' that metze objected
to in the for loops, moving to a much more explicit 'if (schema)'.
Andrew Bartlett
Andrew Bartlett [Tue, 30 Oct 2007 20:01:07 +0000 (21:01 +0100)]
r25753: Move cn=rootdse to @ROOTDSE to avoid being caught up in schema restrictions.
Andrew Bartlett
Stefan Metzmacher [Tue, 30 Oct 2007 07:00:19 +0000 (08:00 +0100)]
r25752: use EJS_CHECK() instead of NDR_CHECK()
metze
Stefan Metzmacher [Tue, 30 Oct 2007 05:24:43 +0000 (06:24 +0100)]
r25751: use EJS_CHECK() instead of NDR_CHECK() in ejs code
metze
Andrew Bartlett [Mon, 29 Oct 2007 20:25:26 +0000 (21:25 +0100)]
r25750: Update the objectclass module to improve consistency in Samba4.
The aim here is to ensure that if we have
CN=Users,DC=samba,DC=example,DC=com
that we cannot have a DN of the form
cn=admin ,cn=useRS,DC=samba,DC=example,DC=com
This module pulls apart the DN, fixes up the relative DN part, and
searches for the parent to copy the base from.
I've used the objectclass module, as I intend to also validate the
placement of child objects, by reading the allowedChildClasses virtual
attribute.
In the future, I'll also force the attribute names to be consistant
(using the case from the schema).
Andrew Bartlett
Andrew Bartlett [Mon, 29 Oct 2007 10:11:36 +0000 (11:11 +0100)]
r25749: Add function required by linked_attributes module.
Andrew Bartlett
Andrew Bartlett [Mon, 29 Oct 2007 09:58:29 +0000 (10:58 +0100)]
r25748: Don't segfault if we don't have a schema yet.
Andrew Bartlett
Andrew Bartlett [Mon, 29 Oct 2007 09:54:06 +0000 (10:54 +0100)]
r25747: Implement linked attributes, for add operations.
Much more work is still required here, particularly to handle this
better during the provision, and to handle modifies and deletes, but
this is a start.
Andrew Bartlett
Stefan Metzmacher [Sat, 27 Oct 2007 08:00:44 +0000 (10:00 +0200)]
r25746: [selftest] make plain output more readable
metze
Stefan Metzmacher [Sat, 27 Oct 2007 07:57:09 +0000 (09:57 +0200)]
r25745: [pidl] ndr_pull/push_error(ndr, NDR_ERR_INVALID_POINTER,..)
instead NT_STATUS_INVALID_PARAMETER_MIX
metze
Stefan Metzmacher [Sat, 27 Oct 2007 06:10:14 +0000 (08:10 +0200)]
r25744: add NDR_ERR_INVALID_POINTER
metze
Jelmer Vernooij [Fri, 26 Oct 2007 21:28:36 +0000 (23:28 +0200)]
r25743: Write short summary of failed tests and skipped tests to file.
Jelmer Vernooij [Fri, 26 Oct 2007 19:25:43 +0000 (21:25 +0200)]
r25742: Start trimming down hash size for start_testsuite.
Jelmer Vernooij [Fri, 26 Oct 2007 19:15:04 +0000 (21:15 +0200)]
r25741: Simplify calls to Subunit functions.
Jelmer Vernooij [Fri, 26 Oct 2007 15:13:42 +0000 (17:13 +0200)]
r25740: More consistent naming for tests, always include environment name in test name.
Günther Deschner [Fri, 26 Oct 2007 10:56:02 +0000 (12:56 +0200)]
r25739: We forgot to copy revision and type flags in security_descriptor_copy().
Guenther
Stefan Metzmacher [Fri, 26 Oct 2007 04:32:24 +0000 (06:32 +0200)]
r25738: always include config.h first.
this needs merging to heimdal and lorikeet-heimdal
metze
Jelmer Vernooij [Thu, 25 Oct 2007 20:20:52 +0000 (22:20 +0200)]
r25737: Integrate more of the test recipe code into one file, use
different mechanism for determining the quicktest list.
Quicktest is a bit verbose now (reports all the tests not run as SKIP:)
Stefan Metzmacher [Thu, 25 Oct 2007 18:13:06 +0000 (20:13 +0200)]
r25736: make sure we include "config.h" as first header
in code generated from et_compile
metze
Stefan Metzmacher [Thu, 25 Oct 2007 18:12:23 +0000 (20:12 +0200)]
r25735: make sure we include "config.h" as first header in
files generated from asn1_compile
metze
Stefan Metzmacher [Thu, 25 Oct 2007 17:52:37 +0000 (19:52 +0200)]
r25734: regenerate yacc output (parse.[ch] files)
metze
Stefan Metzmacher [Thu, 25 Oct 2007 17:51:38 +0000 (19:51 +0200)]
r25733: fix filename substitution and bail out if yacc output isn't generated
metze
Stefan Metzmacher [Thu, 25 Oct 2007 17:50:34 +0000 (19:50 +0200)]
r25732: import updated parse.y files from lorikeet-heimdal
I wonder why they're not updated as the parse.[ch]
are generated from the new versions already...
metze
Günther Deschner [Thu, 25 Oct 2007 09:26:02 +0000 (11:26 +0200)]
r25731: Document winreg_NotifyChangeKeyValue's notify flags.
Guenther
Günther Deschner [Thu, 25 Oct 2007 09:12:02 +0000 (11:12 +0200)]
r25730: Fix IDL for incorrect winreg_SetKeySecurity.
Guenther
Andrew Bartlett [Thu, 25 Oct 2007 09:07:30 +0000 (11:07 +0200)]
r25729: Fix silly regression in the subtree_rename - I broke normal renames.
Bug 5041
Andrew Bartlett
Stefan Metzmacher [Thu, 25 Oct 2007 07:16:11 +0000 (09:16 +0200)]
r25728: [libreplace] define AI_NUMERICSERV as it's used in getaddrinfo.c
metze
Stefan Metzmacher [Thu, 25 Oct 2007 06:48:53 +0000 (08:48 +0200)]
r25727: include "system/network.h" instead of <sys/socket.h>
to make sure socklen_t is typedef'ed
metze
Stefan Metzmacher [Thu, 25 Oct 2007 06:43:00 +0000 (08:43 +0200)]
r25726: [libreplace] move and fix getaddrinfo configure checks
- move getaddrinfo check into getaddrinfo.m4
- add getaddrinfo.o to LIBREPLACEOBJ so that we really build the replacements
metze
Stefan Metzmacher [Thu, 25 Oct 2007 04:56:27 +0000 (06:56 +0200)]
r25725: merge commit
bc9008aa1dcb66890795f0ff2d3046da7df867d9 from v3-2-test:
Author: Jeremy Allison <jra@samba.org>
AuthorDate: Wed Oct 24 14:40:24 2007 -0700
We do support name lookups for host and service in our replacement
code - remove the restrictions from the original Postgres code.
Jeremy.
Stefan Metzmacher [Thu, 25 Oct 2007 04:53:38 +0000 (06:53 +0200)]
r25724: - include getaddrinfo.h after we have defined struct addrinfo
- use rep_ instead of pg_ as prefix in getaddrinfo.[ch]
- define HAVE_<name> macros when we replace functions
metze
Andrew Bartlett [Thu, 25 Oct 2007 03:04:55 +0000 (05:04 +0200)]
r25723: Add a check to prevent deletion of entries with children. Sadly MMC
doesn't trigger it's recursive delete correctly, but the error return
is correct (but perhaps needs a different LDAP wire format).
Andrew Bartlett
Jelmer Vernooij [Wed, 24 Oct 2007 17:16:10 +0000 (19:16 +0200)]
r25722: Start merging the various test plan files for Samba 4 into one, so that it'll
be easier to split selftest/ into a generic, a Samba 3 and a Samba 4-specific
bit.
Stefan Metzmacher [Wed, 24 Oct 2007 12:19:47 +0000 (14:19 +0200)]
r25721: remove unused var
metze
Stefan Metzmacher [Wed, 24 Oct 2007 11:53:48 +0000 (13:53 +0200)]
r25720: update .gitignore
metze
Stefan Metzmacher [Wed, 24 Oct 2007 11:50:30 +0000 (13:50 +0200)]
r25719: let "system/network.h" do the replacements first, then
roken.h.in can do the missing heimdal bits.
this removes to redifinition of 'struct addrinfo' on some platforms
metze
Stefan Metzmacher [Wed, 24 Oct 2007 11:01:03 +0000 (13:01 +0200)]
r25718: [libreplace] if we replace types define the HAVE_<type> macro
this is needed as heimdal's roken.h also tries to replace the types
metze
Stefan Metzmacher [Wed, 24 Oct 2007 10:57:57 +0000 (12:57 +0200)]
r25715: [libreplace] move definition of struct addrinfo so that it can use socklen_t
metze
Stefan Metzmacher [Wed, 24 Oct 2007 10:55:40 +0000 (12:55 +0200)]
r25714: [libreplace] move define of a constant to near to the others
metze
Stefan Metzmacher [Wed, 24 Oct 2007 10:53:34 +0000 (12:53 +0200)]
r25713: [libreplace] include socket_wrapper.h after we have typedef'ed socklen_t
metze
Stefan Metzmacher [Wed, 24 Oct 2007 07:15:26 +0000 (09:15 +0200)]
r25712: try to fix the build, perl should handle by_path as comparison function
metze
Stefan Metzmacher [Wed, 24 Oct 2007 06:34:11 +0000 (08:34 +0200)]
r25711: only use sort if available
metze
Andrew Bartlett [Wed, 24 Oct 2007 02:32:24 +0000 (04:32 +0200)]
r25710: Finally fix subtree renames. Untested code is broken code and in this
case an oddity of the javascript caused the test to 'pass'.
For the same oddity, we have a failure in ldb's handling of spaces in
DNs. We need to resolve that too.
Andrew Bartlett
Jelmer Vernooij [Sun, 21 Oct 2007 23:05:31 +0000 (01:05 +0200)]
r25707: Move selftest makefile code to separate file.
Günther Deschner [Fri, 19 Oct 2007 12:28:26 +0000 (14:28 +0200)]
r25706: Merge padding fix from Samba3.
Guenther
Andrew Bartlett [Fri, 19 Oct 2007 04:57:19 +0000 (06:57 +0200)]
r25705: Ensure we return the out value to the caller.
Andrew Bartlett
Andrew Bartlett [Fri, 19 Oct 2007 04:47:54 +0000 (06:47 +0200)]
r25704: Handle the chicken-and-egg problem of setting up the LDB before we get
a schema. perhaps i need to look into metze's 'load a schema from
ldif' code.
Andrew Bartlett
Andrew Bartlett [Fri, 19 Oct 2007 03:41:32 +0000 (05:41 +0200)]
r25703: Use less entropy by using the pointer value as a process-unique token.
Andrew Bartlett
Andrew Bartlett [Fri, 19 Oct 2007 03:40:00 +0000 (05:40 +0200)]
r25702: Clarify comments and make this module more strict on objectclasses.
This is becoming the schema module...
Andrew Bartlett
Andrew Bartlett [Fri, 19 Oct 2007 03:39:08 +0000 (05:39 +0200)]
r25701: Clarify comment
Jeremy Allison [Thu, 18 Oct 2007 22:40:51 +0000 (00:40 +0200)]
r25700: Now BOOL no longer exists in Samba 3.2, use bool instead.
Jeremy.
Günther Deschner [Thu, 18 Oct 2007 19:23:25 +0000 (21:23 +0200)]
r25699: Add ForceReplication test.
Guenther
Stefan Metzmacher [Thu, 18 Oct 2007 11:27:48 +0000 (13:27 +0200)]
r25697: make *clean doesn't work after a svn up when some *.mk files were removed
the error is:
SAMBA_4_0/source> make clean
make: *** No rule to make target `lib/ldb/samba/config.mk', needed by `Makefile'. Stop.
the problem is:
Makefile: config.status $(MK_FILES)
./config.status
so now we let the MK_FILES variable empty for the *clean targets
if gnu make is detected, we should later test if this construct
is portable to other make implementations and remove the check for gnu make.
metze
Günther Deschner [Thu, 18 Oct 2007 07:44:25 +0000 (09:44 +0200)]
r25696: Fix pipe name for dfs.
Guenther
Stefan Metzmacher [Thu, 18 Oct 2007 06:34:19 +0000 (08:34 +0200)]
r25695: [libreplace] fix the standalone build of libreplace
all configure results which are used in replace.h
or any system/*.h should be in the in the libreplace
*.m4 files!
metze
Andrew Bartlett [Thu, 18 Oct 2007 03:47:11 +0000 (05:47 +0200)]
r25694: Move subtree_rename above the partitions module. The next step is to
built a linked_attributes module under this.
Andrew Bartlett
Andrew Bartlett [Thu, 18 Oct 2007 03:39:55 +0000 (05:39 +0200)]
r25693: Implement the rest of subtree renames, now that tridge waved his magic
over the ldb_tdb part of the problem.
Andrew Bartlett
Andrew Tridgell [Thu, 18 Oct 2007 01:32:07 +0000 (03:32 +0200)]
r25692: fixed another example where the free of fde and the free of the socket
causes the fd to be closed before epoll is told
Andrew Tridgell [Thu, 18 Oct 2007 01:14:10 +0000 (03:14 +0200)]
r25691: make "server role" case insensitive
Andrew Tridgell [Thu, 18 Oct 2007 00:03:21 +0000 (02:03 +0200)]
r25690: - only use a readonly traverse in ldb_search when not in a transaction. When we are in a transaction then we could be in a top level modify operation (such as rename), so we must use a writeable traverse so that the async callbacks can do the modifies while the search is progressing.
- don't do the lockall operation on the tdb during a ldb search if in
a transaction, as this would prevent modifies by callbacks as well
Andrew Tridgell [Wed, 17 Oct 2007 23:37:46 +0000 (01:37 +0200)]
r25689: test with a non-zero size file
Volker Lendecke [Wed, 17 Oct 2007 16:26:40 +0000 (18:26 +0200)]
r25688: Add a simple test for a samba3 regression
When trying an reqchal/auth2 for an unknown workstation account, we have
to return NT_STATUS_NO_TRUST_SAM_ACCOUNT to the auth2. v3-2-test right
now returns NT_STATUS_ACCESS_DENIED, which seems to prevent XP from
joining a Samba3 PDC.
Stefan Metzmacher [Wed, 17 Oct 2007 12:02:33 +0000 (14:02 +0200)]
r25687: Move #defined for getaddrinfo into network.h (as other
defines). Allows RHEL4 compile to work.
Jeremy.
(cherry picked from commit
0ffdf4fdeea88c21880c8bf69d8db56fb49effa7)
Stefan Metzmacher [Wed, 17 Oct 2007 12:02:20 +0000 (14:02 +0200)]
r25686: Added in missing def for AI_ADDRCONFIG.
Jeremy.
(cherry picked from commit
3ef3c3afc3f56b0f4fdb384d55d2e712060fbedf)