15 years agor22041: merge trivial changes from samba3
Stefan Metzmacher [Mon, 2 Apr 2007 18:56:25 +0000 (18:56 +0000)]
r22041: merge trivial changes from samba3


15 years agor22040: Use $< to specify the immediate prerequisit instead of $*. This is
James Peach [Mon, 2 Apr 2007 18:43:31 +0000 (18:43 +0000)]
r22040: Use $< to specify the immediate prerequisit instead of $*. This is
needed for VPATH to find the correct files when srcdir != builddir.

This checkin is a test to see which platforms this change will break.

15 years agor22038: - fix perl warning
Stefan Metzmacher [Mon, 2 Apr 2007 18:14:27 +0000 (18:14 +0000)]
r22038: - fix perl warning
- use the same if statement as in


15 years agor22037: Make sure STATICDIR exists before linking a static archive.
James Peach [Mon, 2 Apr 2007 17:29:03 +0000 (17:29 +0000)]
r22037: Make sure STATICDIR exists before linking a static archive.

15 years agor22036: Make sure both srcdir and builddir are in VPATH.
James Peach [Mon, 2 Apr 2007 17:26:18 +0000 (17:26 +0000)]
r22036: Make sure both srcdir and builddir are in VPATH.

15 years agor22035: Introduce HOSTLD which is needed to link the asn1 compiler when cross
James Peach [Mon, 2 Apr 2007 17:23:23 +0000 (17:23 +0000)]
r22035: Introduce HOSTLD which is needed to link the asn1 compiler when cross

15 years agor22034: Use builddir and srcdir instead on samba_builddir and samba_srcdir.
James Peach [Mon, 2 Apr 2007 17:17:40 +0000 (17:17 +0000)]
r22034: Use builddir and srcdir instead on samba_builddir and samba_srcdir.

15 years agor22032: Make sure static libraries end up in bin/static.
James Peach [Mon, 2 Apr 2007 17:14:43 +0000 (17:14 +0000)]
r22032: Make sure static libraries end up in bin/static.

15 years agor22031: Make sure we rewrite the include directives in CFLAGS to contain
James Peach [Mon, 2 Apr 2007 17:11:36 +0000 (17:11 +0000)]
r22031: Make sure we rewrite the include directives in CFLAGS to contain
paths from builddir and srcdir. builddir has precedence.

15 years agor22030: Simplify include parsing for files, making sure we get
James Peach [Mon, 2 Apr 2007 17:08:21 +0000 (17:08 +0000)]
r22030: Simplify include parsing for files, making sure we get
it right for the srcdir != builddir case. Emit some comment lines
that can be used to figure out which parts of the Makefile output
is generated from which config files.

15 years agor22029: Make sure we respect $srcdir correctly for the srcdir != builddir case.
James Peach [Mon, 2 Apr 2007 17:06:14 +0000 (17:06 +0000)]
r22029: Make sure we respect $srcdir correctly for the srcdir != builddir case.

15 years agor22028: Fix include path for the srcidr != builddir case.
James Peach [Mon, 2 Apr 2007 17:03:30 +0000 (17:03 +0000)]
r22028: Fix include path for the srcidr != builddir case.

15 years agor22027: Allow separate build and source directories. Don't follow symlinks
James Peach [Mon, 2 Apr 2007 17:02:52 +0000 (17:02 +0000)]
r22027: Allow separate build and source directories. Don't follow symlinks
when copying hemidal_build since this leads to traversal loops.

15 years agor22025: patch from Ronnie to make locktest also test the behaviour of pid
Andrew Tridgell [Mon, 2 Apr 2007 07:06:49 +0000 (07:06 +0000)]
r22025: patch from Ronnie to make locktest also test the behaviour of pid
changes in the locks

15 years agor22021: use the pid specified in the individual locks, not the request pid, in
Andrew Tridgell [Mon, 2 Apr 2007 04:47:19 +0000 (04:47 +0000)]
r22021: use the pid specified in the individual locks, not the request pid, in
locking requests.

This fixes a bug noticed by Ronnie Sahlberg (thanks!)

15 years agor22005: Add more code (doesn't work at the moment) preparing a place for
Rafal Szczesniak [Wed, 28 Mar 2007 23:01:35 +0000 (23:01 +0000)]
r22005: Add more code (doesn't work at the moment) preparing a place for
listing user accounts.


15 years agor22004: Check for dereferencing null pointers.
Jelmer Vernooij [Wed, 28 Mar 2007 22:52:37 +0000 (22:52 +0000)]
r22004: Check for dereferencing null pointers.

Patch by Ronnie Sahlberg.

15 years agor21996: Fix a bug for invalid casts of pointers.
Jelmer Vernooij [Tue, 27 Mar 2007 23:22:41 +0000 (23:22 +0000)]
r21996: Fix a bug for invalid casts of pointers.

Patch by Ronnie Sahlberg.

15 years agor21995: Patch from Ronnie Sahlberg.
Jelmer Vernooij [Tue, 27 Mar 2007 23:04:08 +0000 (23:04 +0000)]
r21995: Patch from Ronnie Sahlberg.

Change the signatures for all functions it generates to specify
_U_ so that GCC "parameter not used" warnings are suppressed.

15 years agor21971: Fill in some more values in config files from variables, so we can
Andrew Bartlett [Tue, 27 Mar 2007 08:06:05 +0000 (08:06 +0000)]
r21971: Fill in some more values in config files from variables, so we can
better setup multiple environments.

Run the mk-keyblobs script before we need variables it defines.

Andrew Bartlett

15 years agor21970: Ensure that Fedora DS can be shut down correctly with the stop script:
Andrew Bartlett [Tue, 27 Mar 2007 04:21:16 +0000 (04:21 +0000)]
r21970: Ensure that Fedora DS can be shut down correctly with the stop script:
This needs the pid file to be in the right spot.

Andrew Bartlett

15 years agor21951: Hopefully fix valgrind warning.
James Peach [Fri, 23 Mar 2007 20:28:45 +0000 (20:28 +0000)]
r21951: Hopefully fix valgrind warning.

15 years agor21949: After discussion with the Apple and Linux client maintainers,
James Peach [Fri, 23 Mar 2007 19:24:21 +0000 (19:24 +0000)]
r21949: After discussion with the Apple and Linux client maintainers,
changing the FindFirst response for the UNIX_INFO2 level to include
a length field before the name. The name is not required to be null
terminated. the lenght field does not count any null.

15 years agor21943: don't send unitialized data
Stefan Metzmacher [Fri, 23 Mar 2007 09:11:53 +0000 (09:11 +0000)]
r21943: don't send unitialized data


15 years agor21928: fix the interface uuid of wzcsvc.
Stefan Metzmacher [Thu, 22 Mar 2007 08:17:30 +0000 (08:17 +0000)]
r21928: fix the interface uuid of wzcsvc.


thanks to jbm for the report!


15 years agor21911: Some more work on making the ncacn_np handling in smbd be less special.
Jelmer Vernooij [Wed, 21 Mar 2007 17:05:39 +0000 (17:05 +0000)]
r21911: Some more work on making the ncacn_np handling in smbd be less special.

15 years agor21910: Clean up some unused functions.
Jelmer Vernooij [Wed, 21 Mar 2007 16:26:25 +0000 (16:26 +0000)]
r21910: Clean up some unused functions.

15 years agor21909: Merge some of my recent improvements to the test infrastructure:
Jelmer Vernooij [Wed, 21 Mar 2007 15:57:07 +0000 (15:57 +0000)]
r21909: Merge some of my recent improvements to the test infrastructure:
- Allow tests to specify what environment they need to run in (dc and none are thecurrent supported ones).

- Move more Samba4-specific code out of the common code

15 years agor21896: - Enable creating default NetContext when no explicit credentials are
Rafal Szczesniak [Tue, 20 Mar 2007 22:44:22 +0000 (22:44 +0000)]
r21896: - Enable creating default NetContext when no explicit credentials are
  passed. In such case use what's been provided on swat session logon.
- Create a proper NetContext object only once and add it to the resources
  for later use.


15 years agor21859: add a comment why we remove the rid_crypt obfuscation
Stefan Metzmacher [Fri, 16 Mar 2007 16:35:44 +0000 (16:35 +0000)]
r21859: add a comment why we remove the rid_crypt obfuscation


15 years agor21852: Further work on initialising libnet context from netmgr.
Rafal Szczesniak [Fri, 16 Mar 2007 00:19:40 +0000 (00:19 +0000)]
r21852: Further work on initialising libnet context from netmgr.


15 years agor21843: the computer_dn isn't needed anymore because we don't manually reset the...
Stefan Metzmacher [Thu, 15 Mar 2007 13:21:54 +0000 (13:21 +0000)]
r21843: the computer_dn isn't needed anymore because we don't manually reset the password


15 years agor21842: fix typo in comment
Stefan Metzmacher [Thu, 15 Mar 2007 07:22:49 +0000 (07:22 +0000)]
r21842: fix typo in comment


15 years agor21841: Rename the method and check parameters more strictly.
Rafal Szczesniak [Wed, 14 Mar 2007 23:25:54 +0000 (23:25 +0000)]
r21841: Rename the method and check parameters more strictly.


15 years agor21839: add my email address
Stefan Metzmacher [Wed, 14 Mar 2007 19:10:21 +0000 (19:10 +0000)]
r21839: add my email address


15 years agor21838: generate no metadata for constructed attributes
Stefan Metzmacher [Wed, 14 Mar 2007 12:07:56 +0000 (12:07 +0000)]
r21838: generate no metadata for constructed attributes


15 years agor21837: Make dcerpc_mgmt a separate library again, as the linker leaves it out when...
Jelmer Vernooij [Wed, 14 Mar 2007 00:44:29 +0000 (00:44 +0000)]
r21837: Make dcerpc_mgmt a separate library again, as the linker leaves it out when it's part of dcerpc.

15 years agor21836: Assume that if an OpenLDAP system is 'modular' then everything is a
Andrew Bartlett [Tue, 13 Mar 2007 23:16:09 +0000 (23:16 +0000)]
r21836: Assume that if an OpenLDAP system is 'modular' then everything is a
module, so we have to bring in syncprov too.

Andrew Bartlett

15 years agor21835: fixed a rpc server bug where we failed to remove a call from one
Andrew Tridgell [Tue, 13 Mar 2007 22:58:23 +0000 (22:58 +0000)]
r21835: fixed a rpc server bug where we failed to remove a call from one
linked list when moving it to another. This could cause a valgrind
error under the RPC-SCANNER test.

15 years agor21834: Remove unnecessary includes
Jelmer Vernooij [Tue, 13 Mar 2007 22:03:04 +0000 (22:03 +0000)]
r21834: Remove unnecessary includes

15 years agor21833: Don't install static libs, they can't be used anyway.
Jelmer Vernooij [Tue, 13 Mar 2007 21:47:57 +0000 (21:47 +0000)]
r21833: Don't install static libs, they can't be used anyway.

15 years agor21832: Ignore some more autogenerated files.
Jelmer Vernooij [Tue, 13 Mar 2007 21:00:27 +0000 (21:00 +0000)]
r21832: Ignore some more autogenerated files.

15 years agor21830: Fix header installation, remove proto header with a single prototype.
Jelmer Vernooij [Tue, 13 Mar 2007 20:44:14 +0000 (20:44 +0000)]
r21830: Fix header installation, remove proto header with a single prototype.

15 years agor21829: Integrate mgmt in libdcerpc. We have enough libraries as it is and mgmt...
Jelmer Vernooij [Tue, 13 Mar 2007 20:26:35 +0000 (20:26 +0000)]
r21829:  Integrate mgmt in libdcerpc. We have enough libraries as it is and mgmt is part of the DCE/RPC standard.

15 years agor21828: Hardcode prototypes, as they're a public API.
Jelmer Vernooij [Tue, 13 Mar 2007 20:08:44 +0000 (20:08 +0000)]
r21828: Hardcode prototypes, as they're a public API.

15 years agor21827: move comments to the place where the functionality is implemented
Stefan Metzmacher [Tue, 13 Mar 2007 18:17:31 +0000 (18:17 +0000)]
r21827: move comments to the place where the functionality is implemented


15 years agor21826: reorder functions
Stefan Metzmacher [Tue, 13 Mar 2007 18:12:21 +0000 (18:12 +0000)]
r21826: reorder functions


15 years agor21821: move comment about the becoming a dc to the code which implements it
Stefan Metzmacher [Tue, 13 Mar 2007 15:41:25 +0000 (15:41 +0000)]
r21821: move comment about the becoming a dc to the code which implements it
and extent the comments a bit


15 years agor21820: move function, so that all are in top-down order as they get called
Stefan Metzmacher [Tue, 13 Mar 2007 14:32:20 +0000 (14:32 +0000)]
r21820: move function, so that all are in top-down order as they get called


15 years agor21817: give fields a meaning
Stefan Metzmacher [Tue, 13 Mar 2007 11:10:48 +0000 (11:10 +0000)]
r21817: give fields a meaning


15 years agor21814: use ndr_push_error in the ndr layer, not just a NTSTATUS failure
Andrew Tridgell [Tue, 13 Mar 2007 04:42:49 +0000 (04:42 +0000)]
r21814: use ndr_push_error in the ndr layer, not just a NTSTATUS failure

15 years agor21812: fixed an integer overflow error in the ndr push code.
Andrew Tridgell [Tue, 13 Mar 2007 04:18:07 +0000 (04:18 +0000)]
r21812: fixed an integer overflow error in the ndr push code.

This needs to be fixed in Samba3 as well. It might be exploitable (I
haven't confirmed one way or the other), so I think this should be
fixed for 3.0.25

15 years agor21811: fixed a queueing error in the dcerpc client code. WHen the
Andrew Tridgell [Tue, 13 Mar 2007 03:43:16 +0000 (03:43 +0000)]
r21811: fixed a queueing error in the dcerpc client code. WHen the
dcerpc_ship_next_request() logic was added the penidng queue was split
in two, but we also needed to update the code which removes requests
from the queue to know about the two queues. Following the pattern
used in other client libs, I based which queue to remove from on
req->state, and added a new state RPC_REQUEST_QUEUED. This fixes a
crash that happens when rpc requests time out.

This patch also fixes the handling of timed out bind requests, and the
talloc_reference handling in dcerpc_ndr_request_recv().

15 years agor21810: - The height of a widget is in relation to its enclosing widget. Specifying a
Derrell Lipman [Tue, 13 Mar 2007 02:58:05 +0000 (02:58 +0000)]
r21810: - The height of a widget is in relation to its enclosing widget.  Specifying a
  height of 100% means that if the enclosing widget is 100 pixels high, this
  widget should be 100 pixels high.  If this widget is then placed someplace
  other than at the top of the enclosing widget, e.g. with top:20, the bottom
  of this widget will fall off the bottom of the enclosing widget.

  The normal way to solve this is to specify top and bottom attributes rather
  than top and height attributes.  Also, since widgets often look better if
  they don't extend all the way to the very bottom, specifying a bottom
  attribute equal to the top attribute typically has an appealing appearance.

  I've fixed the tree widget so it no longer falls off the bottom of the page.

15 years agor21809: - If the canvas is never initialized with the tree, the tree will never
Derrell Lipman [Tue, 13 Mar 2007 02:54:06 +0000 (02:54 +0000)]
r21809: - If the canvas is never initialized with the tree, the tree will never
  appear.  Now that the finite state machine properly ignores events from
  widgets which are not handled, the initialization of the tree should be done
  when the module's canvas appears (i.e. just after the canvas has had the
  module's gui applied to it).

15 years agor21808: - Fix a nasty bug in the finite state machine that allowed an event from a
Derrell Lipman [Tue, 13 Mar 2007 02:51:49 +0000 (02:51 +0000)]
r21808: - Fix a nasty bug in the finite state machine that allowed an event from a
  non-handled widget to be processed as if the event originated from a handled
  widget.  This was allowing the appear event for the module's canvas in
  Mimir's Net Manager (an event which was not handled) to load the tree,
  followed by the tree appear event (intended to be handled, albeit
  incorrectly -- see subsequent check-in of Mimir's Fsm.js) to again load the
  tree, thus the double entry.

  Wow, the above paragraph is really hard to read. :-)

15 years agor21807: modularise the opendb code, so different backends can be
Andrew Tridgell [Tue, 13 Mar 2007 01:47:04 +0000 (01:47 +0000)]
r21807: modularise the opendb code, so different backends can be
implemented. This will allow for a ctdb opendb backend.

15 years agor21806: I've been working over the last week to fix up the LDAP backend for
Andrew Bartlett [Tue, 13 Mar 2007 00:59:06 +0000 (00:59 +0000)]
r21806: I've been working over the last week to fix up the LDAP backend for
Samba4.  This only broke on global catalog queries, which turned out to
be due to changes in the partitions module that metze needed for his

I've reworked partitions.c to always include the 'problematic' control,
and therefore demonstrated that this is the issue.  This ensures
consistency, and should help with finding issues like this in future.

As this control (DSDB_CONTROL_CURRENT_PARTITION_OID) is not intended to
be linearised, I've added logic to allow it to be skipped when creating
network packets.

I've likewise make our LDAP server skip unknown controls, when marked
'not critical' on it's input, rather than just dropping the entire
request.  I need some help to generate a correct error packet when it is
marked critical.

Further work could perhaps be to have the ldap_encode routine return a
textual description of what failed to encode, as that would have saved
me a lot of time...

Andrew Bartlett

15 years agor21805: Add PAC_TYPE_CONSTRAINED_DELEGATION to the PAC_TYPE enum.
Andrew Bartlett [Mon, 12 Mar 2007 23:39:00 +0000 (23:39 +0000)]

15 years agor21799: tell to not generate prototypes for setenv() and unsetenv()
Stefan Metzmacher [Mon, 12 Mar 2007 16:05:57 +0000 (16:05 +0000)]
r21799: tell to not generate prototypes for setenv() and unsetenv()


15 years agor21798: unsetenv() returns void on some platforms (BSD)
Stefan Metzmacher [Mon, 12 Mar 2007 11:37:12 +0000 (11:37 +0000)]
r21798: unsetenv() returns void on some platforms (BSD)


15 years agor21797: remove the key directly from the environ array
Stefan Metzmacher [Mon, 12 Mar 2007 11:32:19 +0000 (11:32 +0000)]
r21797: remove the key directly from the environ array

inspired by:


15 years agor21796: check if unsetenv() works on an non-existing key
Stefan Metzmacher [Mon, 12 Mar 2007 11:31:06 +0000 (11:31 +0000)]
r21796: check if unsetenv() works on an non-existing key


15 years agor21795: fix the prototype of unsetenv()...
Stefan Metzmacher [Mon, 12 Mar 2007 10:12:27 +0000 (10:12 +0000)]
r21795: fix the prototype of unsetenv()...


15 years agor21794: add setenv()/unsetenv() testsuite
Stefan Metzmacher [Mon, 12 Mar 2007 09:59:48 +0000 (09:59 +0000)]
r21794: add setenv()/unsetenv() testsuite


15 years agor21793: add replacement for unsetenv()
Stefan Metzmacher [Mon, 12 Mar 2007 09:59:06 +0000 (09:59 +0000)]
r21793: add replacement for unsetenv()


15 years agor21791: This test is still just as valid without as much CPU time wasted.
Andrew Bartlett [Mon, 12 Mar 2007 01:05:26 +0000 (01:05 +0000)]
r21791: This test is still just as valid without as much CPU time wasted.

Andrew Bartlett

15 years agor21790: Setup the socket_wrapper_dir when we set the environment variable.
Andrew Bartlett [Mon, 12 Mar 2007 00:24:03 +0000 (00:24 +0000)]
r21790: Setup the socket_wrapper_dir when we set the environment variable.

Andrew Bartlett

15 years agor21789: We do actually need this, to get a contextCSN attribute, which we need
Andrew Bartlett [Mon, 12 Mar 2007 00:16:24 +0000 (00:16 +0000)]
r21789: We do actually need this, to get a contextCSN attribute, which we need
for USN support.

Andrew Bartlett

15 years agor21788: Add new JSON-RPC services.
Rafal Szczesniak [Sun, 11 Mar 2007 18:37:45 +0000 (18:37 +0000)]
r21788: Add new JSON-RPC services.


15 years agor21787: Add new module to the swat application.
Rafal Szczesniak [Sun, 11 Mar 2007 18:36:36 +0000 (18:36 +0000)]
r21787: Add new module to the swat application.


15 years agor21786: My initial attempts in qooxdoo coding. Derrell, please take
Rafal Szczesniak [Sun, 11 Mar 2007 18:35:19 +0000 (18:35 +0000)]
r21786: My initial attempts in qooxdoo coding. Derrell, please take
a look and I'll have a lot of questions to you.


15 years agor21773: fix typo orginating -> originating
Stefan Metzmacher [Fri, 9 Mar 2007 10:09:37 +0000 (10:09 +0000)]
r21773: fix typo orginating -> originating

and use the struct member names in all cases


15 years agor21772: add DS_BEHAVIOR_WIN2003_INTERIM constant
Stefan Metzmacher [Fri, 9 Mar 2007 07:20:50 +0000 (07:20 +0000)]
r21772: add DS_BEHAVIOR_WIN2003_INTERIM constant


15 years agor21771: We just don't need to test this on more than one transport.
Andrew Bartlett [Fri, 9 Mar 2007 06:48:36 +0000 (06:48 +0000)]
r21771: We just don't need to test this on more than one transport.

Andrew Bartlett

15 years agor21761: - Give more detail on LDAP client library failures (make it clear
Andrew Bartlett [Thu, 8 Mar 2007 06:23:39 +0000 (06:23 +0000)]
r21761: - Give more detail on LDAP client library failures (make it clear
  where the error is from)

- Make default error string more consistant

Andrew Bartlett

15 years agor21760: Try to pin down were some errors are coming from. Ensure we at least
Andrew Bartlett [Thu, 8 Mar 2007 03:32:28 +0000 (03:32 +0000)]
r21760: Try to pin down were some errors are coming from.  Ensure we at least
name the module.

Andrew Bartlett

15 years agor21751: These 2 tests pass for me, so add them to the standard test script.
Andrew Bartlett [Wed, 7 Mar 2007 20:33:15 +0000 (20:33 +0000)]
r21751: These 2 tests pass for me, so add them to the standard test script.

Andrew Bartlett

15 years agor21746: We don't link in this file any more.
Andrew Bartlett [Wed, 7 Mar 2007 11:42:21 +0000 (11:42 +0000)]
r21746: We don't link in this file any more.

15 years agor21745: indent
Andrew Bartlett [Wed, 7 Mar 2007 11:35:06 +0000 (11:35 +0000)]
r21745: indent

15 years agor21744: Test more talloc failure cases.
Andrew Bartlett [Wed, 7 Mar 2007 11:15:44 +0000 (11:15 +0000)]
r21744: Test more talloc failure cases.

Andrew Bartlett

15 years agor21743: Always use the 'escaped' LDAPI path.
Andrew Bartlett [Wed, 7 Mar 2007 11:05:59 +0000 (11:05 +0000)]
r21743: Always use the 'escaped' LDAPI path.

Wait around for the server to start, so we don't try to contact it
during it's starup phase.

Andrew Bartlett

15 years agor21742: reenable the smbclient test now that the d_printf bug is fixed
Andrew Tridgell [Wed, 7 Mar 2007 10:29:39 +0000 (10:29 +0000)]
r21742: reenable the smbclient test now that the d_printf bug is fixed

15 years agor21741: Like starting smbd, ensure we have acutally started slapd, and it is
Andrew Bartlett [Wed, 7 Mar 2007 10:17:50 +0000 (10:17 +0000)]
r21741: Like starting smbd, ensure we have acutally started slapd, and it is
listening, before we continue.

Andrew Bartlett

15 years agor21740: this fixes the real cause of the large log files we had. The problem
Andrew Tridgell [Wed, 7 Mar 2007 10:00:14 +0000 (10:00 +0000)]
r21740: this fixes the real cause of the large log files we had. The problem
was we were not checking the result of a convert_string() call, and it
was giving -1. We then passed -1 to fwrite() on stdout, which on aix
and macosx wrote all of available memory to stdout :)

To fix this, replace non-printing chars with ? in d_printf if the
string cannot be converted

15 years agor21739: Make it easy to change the log level for the slapd processes, and have
Andrew Bartlett [Wed, 7 Mar 2007 05:29:12 +0000 (05:29 +0000)]
r21739: Make it easy to change the log level for the slapd processes, and have
the logs go to a file.

Andrew Bartlett

15 years agor21737: Print the error strings in the ejs ldb test.
Andrew Bartlett [Wed, 7 Mar 2007 04:25:52 +0000 (04:25 +0000)]
r21737: Print the error strings in the ejs ldb test.

15 years agor21736: Fix the smbclient test to do something more interesting with the last
Andrew Bartlett [Wed, 7 Mar 2007 04:20:10 +0000 (04:20 +0000)]
r21736: Fix the smbclient test to do something more interesting with the last
few authentication tests.  Now that the tests correctly 'fail', I was
able to fix the credentials subsystem to honour USER and PASSWD.

To get --machine-pass working, I needed ldb to always load it's static
modules, so I put this in ldb_connect().

Andrew Bartlett

15 years agor21735: skip the smbclient test until jelmer can take a look at it. The
Andrew Tridgell [Wed, 7 Mar 2007 03:10:30 +0000 (03:10 +0000)]
r21735: skip the smbclient test until jelmer can take a look at it. The
current test makes no sense with the new test system. For example, it
does "mkdir bla" then "cd bla" then "cd ..", but it now does these as
separate smbclient commands, which makes no sense at all. It even
seems to try to run them in parallel???

The test fails (segfault, plus lots of errors), but gets reported as a

15 years agor21733: Merge testenv back into core
Jelmer Vernooij [Wed, 7 Mar 2007 02:11:40 +0000 (02:11 +0000)]
r21733: Merge testenv back into core

15 years agor21732: Use smbtorture for enumerating tests.
Jelmer Vernooij [Wed, 7 Mar 2007 01:46:02 +0000 (01:46 +0000)]
r21732: Use smbtorture for enumerating tests.

15 years agor21730: Reintroduce summary for testenv, remove old (no longer used) file.
Jelmer Vernooij [Tue, 6 Mar 2007 23:23:15 +0000 (23:23 +0000)]
r21730: Reintroduce summary for testenv, remove old (no longer used) file.

15 years agor21729: Some more tests
Jelmer Vernooij [Tue, 6 Mar 2007 23:03:34 +0000 (23:03 +0000)]
r21729: Some more tests

15 years agor21728: Perhaps we don't need this on recent OpenLDAP servers. This overlay
Andrew Bartlett [Tue, 6 Mar 2007 22:54:42 +0000 (22:54 +0000)]
r21728: Perhaps we don't need this on recent OpenLDAP servers.  This overlay
is depricated.

Andrew Bartlett

15 years agor21727: Walk some more of the error branches in the ChangePasswordUser server.
Andrew Bartlett [Tue, 6 Mar 2007 22:22:25 +0000 (22:22 +0000)]
r21727: Walk some more of the error branches in the ChangePasswordUser server.


15 years agor21722: Add the dead record functionality presented on If
Volker Lendecke [Tue, 6 Mar 2007 10:11:15 +0000 (10:11 +0000)]
r21722: Add the dead record functionality presented on If
you do a tdb_set_max_dead(tdb, n), then for this tdb a delete operation will
only mark a record as dead and re-use it if a new record is created. The
parameter n allows for at most n dead records per hash chain. If this number
is exceeded, all dead records are put on the central freelist.


15 years agor21721: Push the 'Success! Your new directory server instance was created'
Andrew Bartlett [Tue, 6 Mar 2007 08:21:45 +0000 (08:21 +0000)]
r21721: Push the 'Success!  Your new directory server instance was created'
message to stderr.

Andrew Bartlett

15 years agor21720: Try to make 'TEST_LDAP=yes make test' work again. These recent
Andrew Bartlett [Tue, 6 Mar 2007 06:30:36 +0000 (06:30 +0000)]
r21720: Try to make 'TEST_LDAP=yes make test' work again.  These recent
changes seem to have been lost in the merge, and some variables were
not yet exported.

Andrew Bartlett

15 years agor21719: Try to cover more of the server-side password processing.
Andrew Bartlett [Tue, 6 Mar 2007 05:30:25 +0000 (05:30 +0000)]
r21719: Try to cover more of the server-side password processing.

Don't just exit the test with 'return True', actually process the result.

Turn off password complexity checking for the password length test.

Andrew Bartlett

15 years agor21716: Report status in a way the build farm recognizes.
Jelmer Vernooij [Tue, 6 Mar 2007 00:45:30 +0000 (00:45 +0000)]
r21716: Report status in a way the build farm recognizes.