Günther Deschner [Thu, 3 May 2007 20:12:00 +0000 (20:12 +0000)]
r22655: Call correct free-macros in netsamlogon_cache_get() error paths. Forgot those
in the previous commit.
Guenther
Gerald Carter [Thu, 3 May 2007 17:05:25 +0000 (17:05 +0000)]
r22654: And this is now Samba 3.0.27pre1-SVN
James Peach [Thu, 3 May 2007 16:14:22 +0000 (16:14 +0000)]
r22648: Fix comment to match the code.
Günther Deschner [Thu, 3 May 2007 12:29:32 +0000 (12:29 +0000)]
r22647: Avoid leaking a full info3 structure on each winbindd cached login by making
netsamlogon_cache_get() return a talloc'ed structure.
Guenther
Simo Sorce [Thu, 3 May 2007 12:28:25 +0000 (12:28 +0000)]
r22646: segfault fix in idmap_ldap.c from 3_0_25
Günther Deschner [Thu, 3 May 2007 11:49:32 +0000 (11:49 +0000)]
r22644: Fix memleak.
Guenther
Günther Deschner [Thu, 3 May 2007 11:44:18 +0000 (11:44 +0000)]
r22643: Don't clear cached U/SID and UG/SID entries when we want to logon offline.
Guenther
Jelmer Vernooij [Wed, 2 May 2007 18:10:57 +0000 (18:10 +0000)]
r22641: Install tdbdump and tdbbackup.
Günther Deschner [Wed, 2 May 2007 15:44:05 +0000 (15:44 +0000)]
r22636: Fix logic bug.
We certainly don't want to crash winbind on each sucessfull
centry_uint{8,16,32,64} read.
Jeremy, please check :-)
Guenther
James Peach [Tue, 1 May 2007 21:22:55 +0000 (21:22 +0000)]
r22633: Fix typo in debug message.
James Peach [Tue, 1 May 2007 20:03:44 +0000 (20:03 +0000)]
r22631: Remove the possibility of sid_check_is_domain and
sid_check_is_in_our_domain getting out of sync.
Gerald Carter [Tue, 1 May 2007 16:49:00 +0000 (16:49 +0000)]
r22630: Revert svn r22576 (parse change to enum_aliasmem(). Original code was correct
Michael Adam [Mon, 30 Apr 2007 22:21:46 +0000 (22:21 +0000)]
r22615: fix prototype
James Peach [Mon, 30 Apr 2007 18:59:34 +0000 (18:59 +0000)]
r22614: Only enable GNU-isms in the Makefile if we can be sure that the GNU
make we found is what will be run when the user invokes "make".
Volker Lendecke [Mon, 30 Apr 2007 17:19:49 +0000 (17:19 +0000)]
r22613: Fix an uninitialized variable warning
Jeremy Allison [Mon, 30 Apr 2007 16:32:17 +0000 (16:32 +0000)]
r22611: Fix from Jens Nissen <jens.nissen@gmx.net>. Fix bad
memory leak I introduced into acl code, also remove
redundent extra check for global_sid_System :
global_sid_System == S-1-5-18 which is already
included in the check for a domain of
global_sid_NT_Authority == S-1-5
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 04:16:56 +0000 (04:16 +0000)]
r22593: Finish doing the same for raw TALLOC.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 03:41:40 +0000 (03:41 +0000)]
r22592: Fix TALLOC_SIZE to be consistent.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 03:38:19 +0000 (03:38 +0000)]
r22591: Fix up the examples also.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 02:51:26 +0000 (02:51 +0000)]
r22590: Make TALLOC_ARRAY consistent across all uses.
That should be it....
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 02:39:34 +0000 (02:39 +0000)]
r22589: Make TALLOC_ARRAY consistent across all uses.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 01:34:28 +0000 (01:34 +0000)]
r22588: Make all uses of TALLOC_MEMDUP consistent.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 01:17:34 +0000 (01:17 +0000)]
r22587: Ensure TALLOC_ZERO_ARRAY is consistent.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 00:53:17 +0000 (00:53 +0000)]
r22586: Add a modified version of Simo's patch.
Jeremy.
Jeremy Allison [Mon, 30 Apr 2007 00:48:20 +0000 (00:48 +0000)]
r22585: Get us into a consistent state with TALLOC_ZERO_ARRAY also.
Jeremy.
Jelmer Vernooij [Sun, 29 Apr 2007 21:46:13 +0000 (21:46 +0000)]
r22584: Don't hardcode specific Python version.
Patch by the Samba packaging team for Debian.
Jelmer Vernooij [Sun, 29 Apr 2007 21:44:48 +0000 (21:44 +0000)]
r22583: Support non-Linux GNU systems (such as Debians' GNU with FreeBSD kernel and
Hurd systems).
Patch by the Samba packaging team for Debian.
Jelmer Vernooij [Sun, 29 Apr 2007 21:37:24 +0000 (21:37 +0000)]
r22578: Remove python build directory and generated smbadduser.
Patch originally by the Debian packaging team for Samba.
Jeremy Allison [Sun, 29 Apr 2007 21:13:13 +0000 (21:13 +0000)]
r22577: Change all of parse/*.c to use standard form. Fix some
marshalling bugs.
Jeremy.
Jeremy Allison [Sun, 29 Apr 2007 20:57:04 +0000 (20:57 +0000)]
r22576: Fix marshalling bug in aliasmem.
Jeremy.
Jeremy Allison [Sun, 29 Apr 2007 19:54:26 +0000 (19:54 +0000)]
r22574: Fix reply when we have no dfs shares.
Jeremy.
Jeremy Allison [Sun, 29 Apr 2007 19:20:48 +0000 (19:20 +0000)]
r22573: Fix old bug mixing free() and talloc_free()
when the add_sid_to_array_XX code was moved
from malloc to talloc. Found running valgrind
and rpcclient. Needs merging for 3.0.25 final.
Jeremy.
Jeremy Allison [Sun, 29 Apr 2007 00:09:22 +0000 (00:09 +0000)]
r22564: Move the _strict -> _zeronull functions into lib/util.c
and out of talloc at tridge's request.
Jeremy.
Volker Lendecke [Sat, 28 Apr 2007 18:16:33 +0000 (18:16 +0000)]
r22561: Fix a memleak in lanman.c: Nobody would free the session_list.
Volker
Steve French [Sat, 28 Apr 2007 16:56:35 +0000 (16:56 +0000)]
r22560: Clean up frees of realloc memory in mount.cifs (thanks to Alex for
reminding me how to use Valgrind).
Jeremy Allison [Sat, 28 Apr 2007 14:33:46 +0000 (14:33 +0000)]
r22555: Ensure our paranoid malloc functions return NULL on
size == 0 so we have a known behavior.
Jeremy.
Volker Lendecke [Sat, 28 Apr 2007 13:52:49 +0000 (13:52 +0000)]
r22554: Fix an assumption that TALLOC_ARRAY(.., 0) != NULL.
Volker
Volker Lendecke [Sat, 28 Apr 2007 12:50:35 +0000 (12:50 +0000)]
r22553: Fix the build
Jeremy Allison [Sat, 28 Apr 2007 02:51:55 +0000 (02:51 +0000)]
r22543: Fix bad call to talloc_strict (too few args).
Should fix build farm breakage.
Jeremy.
Jeremy Allison [Fri, 27 Apr 2007 23:18:41 +0000 (23:18 +0000)]
r22542: Move over to using the _strict varients of the talloc
calls. No functional changes. Looks bigger than it is :-).
Jeremy.
Jeremy Allison [Fri, 27 Apr 2007 22:04:27 +0000 (22:04 +0000)]
r22541: Added talloc_size_strict macro.
Jeremy.
Jeremy Allison [Fri, 27 Apr 2007 21:58:10 +0000 (21:58 +0000)]
r22540: Added _strict varients of the macro calls - added
prototypes.
Jeremy.
Jeremy Allison [Fri, 27 Apr 2007 21:09:16 +0000 (21:09 +0000)]
r22539: Added _strict varients of the talloc calls to
return NULL on size == 0 varients.
Jeremy.
Michael Adam [Fri, 27 Apr 2007 15:26:40 +0000 (15:26 +0000)]
r22537: Use a (more speaking) macro when there is one.
Volker Lendecke [Fri, 27 Apr 2007 14:00:27 +0000 (14:00 +0000)]
r22533: Use lib functions where we have them
Michael Adam [Fri, 27 Apr 2007 08:43:06 +0000 (08:43 +0000)]
r22529: Add implementation of HP-UX ACLs in the new vfs-module framework.
James Peach [Thu, 26 Apr 2007 17:36:19 +0000 (17:36 +0000)]
r22527: note the svn location of samba-docs on svnanon.samba.org
Steve French [Thu, 26 Apr 2007 17:23:33 +0000 (17:23 +0000)]
r22526: Support deep tree mounts after share (instead of having to specify
prepath)
Günther Deschner [Thu, 26 Apr 2007 12:09:25 +0000 (12:09 +0000)]
r22524: Fix memleak.
Guenther
Günther Deschner [Wed, 25 Apr 2007 10:01:21 +0000 (10:01 +0000)]
r22511: Remove unused LDAPMessage.
Guenther
Gerald Carter [Wed, 25 Apr 2007 09:32:03 +0000 (09:32 +0000)]
r22509: Fix some memory corruption caused by calling free()
on talloc()'d memory when adding/removing members
from Local Groups.
James Peach [Tue, 24 Apr 2007 18:19:13 +0000 (18:19 +0000)]
r22507: Wrap the method of obtaining sockets to listen on.
Alexander Bokovoy [Tue, 24 Apr 2007 18:01:46 +0000 (18:01 +0000)]
r22505: Fix build
Jim McDonough [Tue, 24 Apr 2007 15:56:02 +0000 (15:56 +0000)]
r22504: Fix bug Jerry found during his tutorial. Sorry :-(
Allows authorized users (e.g. BUILTIN\Administrators members) to
set attributes on an account, particularly "user cannot change
password".
add become_root() around updating attributes, after checking that
access has been granted.
Jeremy Allison [Tue, 24 Apr 2007 12:56:23 +0000 (12:56 +0000)]
r22502: Fix bug #4536 - delete symlinks to a directory correctly.
Jeremy.
Michael Adam [Tue, 24 Apr 2007 00:12:28 +0000 (00:12 +0000)]
r22496: reg_enumvalue should return WERR_NO_MORE_ITEMS instead of
WERR_BAD_FILE when all items have been successfully enumerated.
Besides seeming the reasonable code to return,
this is what I have seen from w2k3, w2k, wxp.
James Peach [Mon, 23 Apr 2007 22:17:06 +0000 (22:17 +0000)]
r22495: Create wrapper to hide the details of obtaining a set of sockets
to listen on.
Jeremy Allison [Mon, 23 Apr 2007 21:40:54 +0000 (21:40 +0000)]
r22493: Fix horrible memory corruption bug with CIFS POSIX open/mkidr.
Found by SteveF + valgrind :-). Jerry - essential for rc3.
Jeremy.
Jeremy Allison [Mon, 23 Apr 2007 21:07:05 +0000 (21:07 +0000)]
r22492: Add objectid code.
Jeremy.
Jeremy Allison [Mon, 23 Apr 2007 16:45:56 +0000 (16:45 +0000)]
r22491: Fix the build (3.0.25 and 3.0 are different here).
Jeremy.
Jeremy Allison [Mon, 23 Apr 2007 16:32:24 +0000 (16:32 +0000)]
r22490: Fix a couple of bugs found whist investigating CSC Vista issues.
Ensure we correctly NULL out allocation size fields. Allow
QFILEINFO on pipes (Vista bug ?). Jerry - don't automatically
merge for 3.0.25.
Jeremy.
Michael Adam [Mon, 23 Apr 2007 16:31:31 +0000 (16:31 +0000)]
r22489: Add a deletedomgroup command to rpcclient (samr).
Corresponding to deletedomuser.
Michael
Günther Deschner [Mon, 23 Apr 2007 10:04:15 +0000 (10:04 +0000)]
r22484: Check for the talloc_strdup to succeed.
Guenther
Günther Deschner [Mon, 23 Apr 2007 09:56:54 +0000 (09:56 +0000)]
r22483: Attempt to fix the build.
Guenther
Jeremy Allison [Mon, 23 Apr 2007 09:19:35 +0000 (09:19 +0000)]
r22481: Move check for non-mappable SIDs to after sid_to_uid,
sid_to_gid mapping, add LocalSystem to non-mappable
list.
Jeremy.
Günther Deschner [Mon, 23 Apr 2007 08:46:10 +0000 (08:46 +0000)]
r22480: Remove accidentially comitted getpeereid hunks again.
Guenther
Günther Deschner [Mon, 23 Apr 2007 08:40:54 +0000 (08:40 +0000)]
r22479: Add "net ads keytab list".
Guenther
Gerald Carter [Mon, 23 Apr 2007 06:10:22 +0000 (06:10 +0000)]
r22473: Correct fix for setting a default compat tdb idmap backend.
Previous code would always fill in "idmap backend = tdb"
even if you defined idmap domains. My fault. I should
have tested the original patch more before committing.
Günther Deschner [Sun, 22 Apr 2007 21:21:26 +0000 (21:21 +0000)]
r22468: Fixing implicit declaration of krb5_set_default_tgs_ktypes (which is hidden by
KRB5_PRIVATE in MIT and doesn't exist on Heimdal).
Guenther
Günther Deschner [Sun, 22 Apr 2007 21:17:49 +0000 (21:17 +0000)]
r22467: Adding configure check for -Werror-implicit-function-declaration and use it in
-enable-developer.
Guenther
Günther Deschner [Sun, 22 Apr 2007 20:15:00 +0000 (20:15 +0000)]
r22466: Fix build warning.
Guenther
Günther Deschner [Sun, 22 Apr 2007 17:06:22 +0000 (17:06 +0000)]
r22465: Last attempt to fix the build before dinner.
Guenther
Stefan Metzmacher [Sun, 22 Apr 2007 17:00:58 +0000 (17:00 +0000)]
r22464: remove double code and use a function
metze
Günther Deschner [Sun, 22 Apr 2007 16:45:03 +0000 (16:45 +0000)]
r22463: Add prototype.
Guenther
Günther Deschner [Sun, 22 Apr 2007 15:25:54 +0000 (15:25 +0000)]
r22461: Use ranged LDAP queries in lookup_usergroups_member() and start to optinmize
lookup_groupmem(). In the later, at least try to avoid those massive LDAP
dn_lookups by looking in the cache before.
Guenther
Günther Deschner [Sun, 22 Apr 2007 15:15:00 +0000 (15:15 +0000)]
r22460: Adding a generic ads_ranged_search() function.
Guenther
Günther Deschner [Sun, 22 Apr 2007 15:13:50 +0000 (15:13 +0000)]
r22459: Adding ads_get_dn_from_extended_dn(), in preparation of making ranged LDAP
queries more generic. Michael, feel free to overwrite these and the following.
Guenther
Volker Lendecke [Sun, 22 Apr 2007 14:51:07 +0000 (14:51 +0000)]
r22458: Fix a comment
Gerald Carter [Sun, 22 Apr 2007 10:09:41 +0000 (10:09 +0000)]
r22447: Patch from Ying Li <ying.li2@hp.com> to default tdb idmap
plugin when neither idmap domains nor idmap backend have
been defined.
Gerald Carter [Sun, 22 Apr 2007 09:32:14 +0000 (09:32 +0000)]
r22444: * Validate a SID before trying to convert it to a uid/gid via the public
winbindd interface
* Add nss_info/*so files to the RHEL/Fedora packaging
James Peach [Sun, 22 Apr 2007 03:17:31 +0000 (03:17 +0000)]
r22441: Add automatic dependency tracking. The dependency generation is
toolchain independent, but since there's no portable method for
doing makefile includes, we can actually use the dependencies with
GNU make. It's easy to add this for other makes, but I don't have
any to test.
This also moves as much m4 as possible into the m4 directory where
no-one has to look at it. AFAICT, there's no way to hide depcomp
in there too, which is unfortunate.
Gerald Carter [Sat, 21 Apr 2007 21:29:31 +0000 (21:29 +0000)]
r22430: Add SID validate to sid2uid() and sid2gid() public entry points in winbindd
Simo Sorce [Sat, 21 Apr 2007 20:43:54 +0000 (20:43 +0000)]
r22425: Avoid to segfault if we only have the realm.
James Peach [Fri, 20 Apr 2007 21:31:04 +0000 (21:31 +0000)]
r22419: Update configure so that we only need one copy of os2_delete.c
James Peach [Fri, 20 Apr 2007 21:09:44 +0000 (21:09 +0000)]
r22418: Support running under launchd. We abstract the method of obtaining
sockets to listen on a little, because in the launchd case these
are provided for us. We also add an idle timeout so that a daemon
can exit after a period of inactivity.
James Peach [Fri, 20 Apr 2007 18:34:33 +0000 (18:34 +0000)]
r22417: Refactor the various daemon run-mode options to make the semantics
of the various flags explicit.
James Peach [Fri, 20 Apr 2007 18:12:07 +0000 (18:12 +0000)]
r22416: Remove precompiled headers on clean.
James Peach [Fri, 20 Apr 2007 16:07:16 +0000 (16:07 +0000)]
r22413: Simplify the build rules for plugin modules. Add support for per-platform
exports files. Fix the shared library build pf pdb_ldap.
Michael Adam [Fri, 20 Apr 2007 12:23:36 +0000 (12:23 +0000)]
r22412: Add a "deletelocalgroup" subcommand to net sam.
Thanks to Karolin Seeger <ks@sernet.de>.
Michael Adam [Fri, 20 Apr 2007 12:06:38 +0000 (12:06 +0000)]
r22411: Fix the build:
Prototype of create_aio_ex_read, set_message and definition of
UNIXERROR have changed in r22389 and r22391? to require an additional
"inbuf" argument. Some places in smbd/aio.c had not been adapted.
This patch fixes the build.
Michael
Günther Deschner [Fri, 20 Apr 2007 10:54:55 +0000 (10:54 +0000)]
r22402: Fix build warning.
Guenther
Michael Adam [Fri, 20 Apr 2007 10:27:33 +0000 (10:27 +0000)]
r22401: Add bzr revision+nick support to mkversion.sh.
The plan is to actually synchronize with SAMBA_4_0's mkversion.sh,
which is more elaborate. But there are a few differences in
behaviour I still have to check. So I am cheching this here in for
a start...
Michael
Herb Lewis [Fri, 20 Apr 2007 01:52:44 +0000 (01:52 +0000)]
r22395: allow profiling level to be set on startup
James Peach [Thu, 19 Apr 2007 23:47:55 +0000 (23:47 +0000)]
r22394: Remove test directory once we are finished.
Simo Sorce [Thu, 19 Apr 2007 23:25:37 +0000 (23:25 +0000)]
r22393: fix cut&paste error
Simo Sorce [Thu, 19 Apr 2007 23:15:33 +0000 (23:15 +0000)]
r22392: Remove leftover potentially segfaulty code.
Check we are online before actually trying to connect anywhere
Jeremy Allison [Thu, 19 Apr 2007 22:40:32 +0000 (22:40 +0000)]
r22391: Looks bigger than it is. Make "inbuf" available
to all callers of smb_setlen (via set_message()
calls). This will allow the server to reflect back
the correct encryption context.
Jeremy.
Gerald Carter [Thu, 19 Apr 2007 22:26:09 +0000 (22:26 +0000)]
r22390: Patchset sent to samba-technical to address the winbind
loop when allocating a new id for a SID:
auth_util.patch Revert create_local_token() to
the 3.0.24 codebase
idmap_type.patch Have the caller fillin the
id_map.xid.type field when
resolving a SID so that if we allocate
a new id, we know what type to use
winbindd_api.patch Remove the WINBINDD_SIDS_TO_XIDS calls
from the public winbindd interface
for the 3.0.25 release
idmap_rid.patch Cleanup the idmap_rid backend to not
call back into winbindd to resolve
the SID in order to verify it's type.
Jeremy Allison [Thu, 19 Apr 2007 20:50:49 +0000 (20:50 +0000)]
r22389: Start preparing for multiple encryption contexts in the
server. Allow server to reflect back to calling client
the encryption context that was sent.
Jeremy.
Simo Sorce [Thu, 19 Apr 2007 20:00:40 +0000 (20:00 +0000)]
r22388: clearer message, thanks David