tprouty/samba.git
11 years agor4103: lock timeout is in milliseconds (as in reply.c)
Herb Lewis [Wed, 8 Dec 2004 17:30:50 +0000 (17:30 +0000)]
r4103: lock timeout is in milliseconds (as in reply.c)

11 years agor4094: BUG 2107: fix memory bloating caused by large numbers of print_queue_updates...
Gerald Carter [Wed, 8 Dec 2004 03:02:29 +0000 (03:02 +0000)]
r4094: BUG 2107: fix memory bloating caused by large numbers of print_queue_updates() requests sent via messages.tdb

11 years agor4092: Bring into line with new *alloc rules.
Jeremy Allison [Wed, 8 Dec 2004 01:52:23 +0000 (01:52 +0000)]
r4092: Bring into line with new *alloc rules.
Jeremy.

11 years agor4089: fix logic error in add_a_form() that only compared N characters instead of...
Gerald Carter [Tue, 7 Dec 2004 19:00:45 +0000 (19:00 +0000)]
r4089: fix logic error in add_a_form() that only compared N characters instead of the entire form name

11 years agor4088: Get medieval on our ass about malloc.... :-). Take control of all our allocation
Jeremy Allison [Tue, 7 Dec 2004 18:25:53 +0000 (18:25 +0000)]
r4088: Get medieval on our ass about malloc.... :-). Take control of all our allocation
functions so we can funnel through some well known functions. Should help greatly with
malloc checking.
HEAD patch to follow.
Jeremy.

11 years agor4083: consolidate printer searches to use find_service rather than for loops
Gerald Carter [Mon, 6 Dec 2004 19:25:25 +0000 (19:25 +0000)]
r4083: consolidate printer searches to use find_service rather than for loops

11 years agor4043: BUG 2091: don't remove statically defined printers in remove_stale_printers()
Gerald Carter [Thu, 2 Dec 2004 17:11:18 +0000 (17:11 +0000)]
r4043: BUG 2091: don't remove statically defined printers in remove_stale_printers()

11 years agor4032: Fix bug #2110 - ensure we convert to ucs2 correctly.
Jeremy Allison [Thu, 2 Dec 2004 00:55:47 +0000 (00:55 +0000)]
r4032: Fix bug #2110 - ensure we convert to ucs2 correctly.
Jeremy.

11 years agor4030: patch from Rob -- don't force the cups printer-make-and-model tag as the comme...
Gerald Carter [Wed, 1 Dec 2004 21:11:54 +0000 (21:11 +0000)]
r4030: patch from Rob -- don't force the cups printer-make-and-model tag as the comment for autoloaded printers

11 years agor4029: solaris packaging fixes from Rob
Gerald Carter [Wed, 1 Dec 2004 21:01:27 +0000 (21:01 +0000)]
r4029: solaris packaging fixes from Rob

11 years agor4028: * prevent the background LPQ daemon from updating
Gerald Carter [Wed, 1 Dec 2004 20:11:31 +0000 (20:11 +0000)]
r4028: * prevent the background LPQ daemon from updating
  the print queue cache just because multiple smbd
  processes sent a message that it was out of date.

* add some extra debug messages while trying to track down
  down jobs being left in the queue after printing has
  completed.

11 years agor4023: Code tidyup to fix bug #2098.
Jeremy Allison [Tue, 30 Nov 2004 22:57:41 +0000 (22:57 +0000)]
r4023: Code tidyup to fix bug #2098.
Jeremy.

11 years agor4020: Fix for crash reported by Bård Kalbakk <baard@inett.biz>.
Jeremy Allison [Tue, 30 Nov 2004 18:58:58 +0000 (18:58 +0000)]
r4020: Fix for crash reported by Bård Kalbakk <baard@inett.biz>.
Don't go fishing for the authorisation data unless we
know it's there.
Jeremy.

11 years agor4019: better error logging in register_message_flags() from Rob
Gerald Carter [Tue, 30 Nov 2004 15:55:27 +0000 (15:55 +0000)]
r4019: better error logging in register_message_flags() from Rob

11 years agor4018: * move claim_connection from the netbios session request
Gerald Carter [Tue, 30 Nov 2004 15:52:46 +0000 (15:52 +0000)]
r4018: * move claim_connection from the netbios session request
  reply code to the negprot reply code to cope with
  client connections on port 445.  Fixes the spurious
  "register_message_flags: tdb fetch failed" errors.

* don't run the backgroup LPQ daemon when we are running
  in interactive mode.

11 years agor4016: Fix for bug found by Steve French client code (cifsfs) on
Jeremy Allison [Tue, 30 Nov 2004 06:41:38 +0000 (06:41 +0000)]
r4016: Fix for bug found by Steve French client code (cifsfs) on
POSIX ACL set. You need to *get* a permset_t pointer from the entry before
any of the permset code will accept it as a valid value
Jeremy.

11 years agor4009: Fix from Timur Bakeyev <timur@com.bat.ru> for bugid #2100,
Jeremy Allison [Tue, 30 Nov 2004 01:01:43 +0000 (01:01 +0000)]
r4009: Fix from Timur Bakeyev <timur@com.bat.ru> for bugid #2100,
change the way we check for errors after a dlopen (which
may set internal warnings which get picked up by mistake
in dlsym).
Jeremy

11 years agor4007: Fix bug #2088 - ensure inherit permissions is only applied on a new file,
Jeremy Allison [Tue, 30 Nov 2004 00:22:04 +0000 (00:22 +0000)]
r4007: Fix bug #2088 - ensure inherit permissions is only applied on a new file,
not an existing one.
Jeremy.

11 years agor4005: Fix for bug #2071 reported by Jason Mader <jason@ncac.gwu.edu>.
Jeremy Allison [Mon, 29 Nov 2004 19:28:12 +0000 (19:28 +0000)]
r4005: Fix for bug #2071 reported by Jason Mader <jason@ncac.gwu.edu>.
Use correct enum type for comparisons.
Jeremy.

11 years agor3987: Use sys_readdir() instead of readdir()
Jelmer Vernooij [Sat, 27 Nov 2004 19:02:45 +0000 (19:02 +0000)]
r3987: Use sys_readdir() instead of readdir()

11 years agor3985: Fix bug with 64bit fs support
Jelmer Vernooij [Sat, 27 Nov 2004 18:57:44 +0000 (18:57 +0000)]
r3985: Fix bug with 64bit fs support

11 years agor3974: - Fix assignment of a couple of fields in pdb_{mysql,pgsql}
Jelmer Vernooij [Thu, 25 Nov 2004 20:55:31 +0000 (20:55 +0000)]
r3974: - Fix assignment of a couple of fields in pdb_{mysql,pgsql}
- Use new DTD URL in pdb_xml

11 years agor3954: bring Samba3 into line with the Samba4 password change code
Andrew Tridgell [Thu, 25 Nov 2004 04:10:19 +0000 (04:10 +0000)]
r3954: bring Samba3 into line with the Samba4 password change code

11 years agor3951: Fix for bugid #2081 reported by John Janosik <jpjanosi@us.ibm.com> - ensure
Jeremy Allison [Thu, 25 Nov 2004 00:59:44 +0000 (00:59 +0000)]
r3951: Fix for bugid #2081 reported by John Janosik <jpjanosi@us.ibm.com> - ensure
SE_DESC_DACL_PROTECTED is set if "map acl inherit = no".
Jeremy.

11 years agor3948: Fix incorrect declaration. Bug #2083.
Jeremy Allison [Thu, 25 Nov 2004 00:25:53 +0000 (00:25 +0000)]
r3948: Fix incorrect declaration. Bug #2083.
Jeremy.

11 years agor3946: Fix for bugid #2085 reported by Jason Mader <jason@ncac.gwu.edu>. Use consistent
Jeremy Allison [Thu, 25 Nov 2004 00:07:01 +0000 (00:07 +0000)]
r3946: Fix for bugid #2085 reported by Jason Mader <jason@ncac.gwu.edu>. Use consistent
enum type for Protocol extern.
Jeremy.

11 years agor3944: Fix the problem we get on Linux where sendfile fails, but we've already sent the
Jeremy Allison [Wed, 24 Nov 2004 22:05:59 +0000 (22:05 +0000)]
r3944: Fix the problem we get on Linux where sendfile fails, but we've already sent the
header using send(). As our implementation of sendfile can't return EINTR (it
restarts in that case) use an errno of EINTR to signal the linux sendfile fail
after header case. When that happens send the rest of the data and then turn
off sendfile. Sendfile should be safe to enable on all systems now (even though
it may not help in all performance cases).
Jeremy.

11 years agor3940: typo.
Günther Deschner [Wed, 24 Nov 2004 09:44:57 +0000 (09:44 +0000)]
r3940: typo.

Guenther

11 years agor3936: Ensure LARGE_READX response must fit within reply buffer.
Jeremy Allison [Wed, 24 Nov 2004 05:24:38 +0000 (05:24 +0000)]
r3936: Ensure LARGE_READX response must fit within reply buffer.
Jeremy.

11 years agor3934: Correctly check for the top length bit in LARGE_READX.
Jeremy Allison [Wed, 24 Nov 2004 03:42:01 +0000 (03:42 +0000)]
r3934: Correctly check for the top length bit in LARGE_READX.
Jeremy.

11 years agor3931: Fix all "may be used uninitialized" and "shadow" warnings.
Jeremy Allison [Wed, 24 Nov 2004 01:03:23 +0000 (01:03 +0000)]
r3931: Fix all "may be used uninitialized" and "shadow" warnings.
Jeremy.

11 years agor3929: Dead code elimination fix for bug #2075 from jason@ncac.gwu.edu.
Jeremy Allison [Tue, 23 Nov 2004 23:13:48 +0000 (23:13 +0000)]
r3929: Dead code elimination fix for bug #2075 from jason@ncac.gwu.edu.
Jeremy.

11 years agor3928: Fix duplicate call to pdb_get_acct_desc(). Bugzilla #2080.
Tim Potter [Tue, 23 Nov 2004 20:50:16 +0000 (20:50 +0000)]
r3928: Fix duplicate call to pdb_get_acct_desc().  Bugzilla #2080.

11 years agor3925: patch from Thomas Bork <tombork@web.de> to fix configure.in with GNU ld versio...
Gerald Carter [Tue, 23 Nov 2004 16:35:14 +0000 (16:35 +0000)]
r3925: patch from Thomas Bork <tombork@web.de> to fix configure.in with GNU ld versions that don't output a date string in the --version

11 years agor3924: fix installswat once again after the image target location had changed in...
Gerald Carter [Tue, 23 Nov 2004 15:27:53 +0000 (15:27 +0000)]
r3924: fix installswat once again after the image target location had changed in the docs

11 years agor3923: BUG 2073: fall back to smb_name if current_user_info is not available in lp_fi...
Gerald Carter [Tue, 23 Nov 2004 14:52:56 +0000 (14:52 +0000)]
r3923: BUG 2073: fall back to smb_name if current_user_info is not available in lp_file_list_changed()

11 years agor3918: Allow to set OWNER- and GROUP-entries while setting security descriptors
Günther Deschner [Tue, 23 Nov 2004 01:05:31 +0000 (01:05 +0000)]
r3918: Allow to set OWNER- and GROUP-entries while setting security descriptors
with smbcacls and using with the -S or -M switch. Fixes #404 and #2076.

Guenther

11 years agor3912: More fixes to have proper german in swat.
Günther Deschner [Mon, 22 Nov 2004 15:20:38 +0000 (15:20 +0000)]
r3912: More fixes to have proper german in swat.
(The new german orthography is another story...)

Thanks to Reiner Klaproth and Björn Jacke <bjacke@sernet.de>.

Guenther

11 years agor3911: typo.
Günther Deschner [Mon, 22 Nov 2004 14:34:45 +0000 (14:34 +0000)]
r3911: typo.

Guenther

11 years agor3895: Fix for bug #2045. May also fix other timestamp bugs with Excel
Jeremy Allison [Sat, 20 Nov 2004 21:24:52 +0000 (21:24 +0000)]
r3895: Fix for bug #2045. May also fix other timestamp bugs with Excel
(Volker please test). Setting a last write timestamp from Windows
overrides any subsequent write timestamp changes and must be immediately
seen by and findfirst/findnexts. This is a racy solution, but should
work most of the time. This may also fix #1061, not sure.
Jeremy.

11 years agor3892: Fix "might be used uninitialized" error.
Jeremy Allison [Sat, 20 Nov 2004 20:10:43 +0000 (20:10 +0000)]
r3892: Fix "might be used uninitialized" error.
Jeremy.

11 years agor3889: Only set mtime from pending_modtime if it's not already zero...
Jeremy Allison [Sat, 20 Nov 2004 02:48:34 +0000 (02:48 +0000)]
r3889: Only set mtime from pending_modtime if it's not already zero...
Jeremy.

11 years agor3883: Fix error return -- thanks to rsharpe
Volker Lendecke [Fri, 19 Nov 2004 21:01:23 +0000 (21:01 +0000)]
r3883: Fix error return -- thanks to rsharpe

11 years agor3882: Fix bug 2069 -- thanks to jason@ncac.gwu.edu
Volker Lendecke [Fri, 19 Nov 2004 20:55:54 +0000 (20:55 +0000)]
r3882: Fix bug 2069 -- thanks to jason@ncac.gwu.edu

11 years agor3880: Tidy up some unused/shadowed variable usage.
Jeremy Allison [Fri, 19 Nov 2004 19:45:03 +0000 (19:45 +0000)]
r3880: Tidy up some unused/shadowed variable usage.
Jeremy.

11 years agor3877: Final (I hope :-) fix for #2050 from Lars Müller for select maxfd's.
Jeremy Allison [Fri, 19 Nov 2004 19:32:31 +0000 (19:32 +0000)]
r3877: Final (I hope :-) fix for #2050 from Lars Müller for select maxfd's.
Jeremy.

11 years agor3875: Allow to look up at least or own sid in _lsa_lookup_sids.
Günther Deschner [Fri, 19 Nov 2004 13:26:17 +0000 (13:26 +0000)]
r3875: Allow to look up at least or own sid in _lsa_lookup_sids.

This fixes Bugzilla #1076 and Exchange 5.5 SP4 can then be finally
installed on NT4 in a samba-controlled domain.

Guenther

11 years agor3873: The semantics of the parameter 'printcap name' are a bit tricky. I had seen
Volker Lendecke [Fri, 19 Nov 2004 12:11:13 +0000 (12:11 +0000)]
r3873: The semantics of the parameter 'printcap name' are a bit tricky. I had seen
the effect that I could not list printers with smbclient -L. I have cups
libraries but no running cups server, so remove_stale_printers() removed all
my printer definitions from the share list. So I said 'printing = bsd' but it
still would not work.

This happened because init_globals() would initialize Globals.szPrintcapname
to "cups", and the explicit 'printing = bsd' did not reset it. 'printing=bsd'
can't reset it, as this might overwrite an explicit setting. Thus I separated
the lp_printcapname into a function of its own, looking at
Globals.szPrintcapname and subsequently at sDefault.iPrinting.

Please revisit, there are just too many cases to cover.

Thanks,

Volker

11 years agor3871: Fix memleak
Volker Lendecke [Fri, 19 Nov 2004 11:59:56 +0000 (11:59 +0000)]
r3871: Fix memleak

11 years agor3868: Fix reverted logic of detecting HAVE_WRFILE_KEYTAB (Bugzilla 2059)
Günther Deschner [Thu, 18 Nov 2004 23:25:37 +0000 (23:25 +0000)]
r3868: Fix reverted logic of detecting HAVE_WRFILE_KEYTAB (Bugzilla 2059)

Guenther

11 years agor3867: Fix from david.hu@hp.com - make a copy of an incoming message
Jeremy Allison [Thu, 18 Nov 2004 22:46:04 +0000 (22:46 +0000)]
r3867: Fix from david.hu@hp.com - make a copy of an incoming message
rather than indirecting into it as a struct (may not be on an
even byte boundary). Bug #2052.
Jeremy.

11 years agor3864: Fix from Lars Müller <lmuelle@suse.de> for bug #2050.
Jeremy Allison [Thu, 18 Nov 2004 22:11:20 +0000 (22:11 +0000)]
r3864: Fix from Lars Müller <lmuelle@suse.de> for bug #2050.
Make nmbd use maxfd+1 in select also.
Jeremy.

11 years agor3862: Fix from Lars Müller <lmuelle@suse.de>, check for HAVE_WRFILE_KEYTAB
Jeremy Allison [Thu, 18 Nov 2004 21:22:24 +0000 (21:22 +0000)]
r3862: Fix from Lars Müller <lmuelle@suse.de>, check for HAVE_WRFILE_KEYTAB
was inverted.
Jeremy.

11 years agor3859: Ensure if num_acls is set to 0xFFFF this field is ignored.
Jeremy Allison [Thu, 18 Nov 2004 19:36:50 +0000 (19:36 +0000)]
r3859: Ensure if num_acls is set to 0xFFFF this field is ignored.
Use def_acl everywhere instead of dir_acl.
Jeremy.

11 years agor3857: Shut up gcc about erroneous "used uninitialised" warning.
Jeremy Allison [Thu, 18 Nov 2004 19:03:17 +0000 (19:03 +0000)]
r3857: Shut up gcc about erroneous "used uninitialised" warning.
Jeremy.

11 years agor3855: Sync up with the one in the main source dir (yes, this should probably
Jeremy Allison [Thu, 18 Nov 2004 18:40:39 +0000 (18:40 +0000)]
r3855: Sync up with the one in the main source dir (yes, this should probably
be a symlink...).
Jeremy.

11 years agor3853: Fix an uninitialized variable warning. Jeremy, there are some of these in
Volker Lendecke [Thu, 18 Nov 2004 16:14:13 +0000 (16:14 +0000)]
r3853: Fix an uninitialized variable warning. Jeremy, there are some of these in
charcnv.c.

Volker

11 years agor3852: Fix the build...
Volker Lendecke [Thu, 18 Nov 2004 15:52:02 +0000 (15:52 +0000)]
r3852: Fix the build...

11 years agor3850: I get the impression that some developers here are using gcc 3.4. This one
Volker Lendecke [Thu, 18 Nov 2004 15:44:01 +0000 (15:44 +0000)]
r3850: I get the impression that some developers here are using gcc 3.4. This one
only warns you about possibly uninitialized variables if you have at least
-O1.

Volker

11 years agor3848: Fix for bug 2057. Only partially applied the patch, the second part seems not
Volker Lendecke [Thu, 18 Nov 2004 15:13:58 +0000 (15:13 +0000)]
r3848: Fix for bug 2057. Only partially applied the patch, the second part seems not
necessary.

I'm using gcc 3.4.1, this does not detect the possibly uninitialized
variable. Does anybody know how to get 3.4.1 to warn me?

Volker

11 years agor3846: Typo fix. Bug 2055. Lars, I'm available on irc.... :-)
Volker Lendecke [Thu, 18 Nov 2004 13:29:10 +0000 (13:29 +0000)]
r3846: Typo fix. Bug 2055. Lars, I'm available on irc.... :-)

Volker

11 years agor3844: Attempt to get the build farm in line. Some systems seem to have defined
Volker Lendecke [Thu, 18 Nov 2004 12:19:16 +0000 (12:19 +0000)]
r3844: Attempt to get the build farm in line. Some systems seem to have defined
rewinddir as a macro and don't like the construct to call a function pointer
called rewinddir. Work around by renaming the function pointer...

Volker

11 years agor3843: If a connection to a DC is requested, open connections simultaeneously to all
Volker Lendecke [Thu, 18 Nov 2004 11:57:49 +0000 (11:57 +0000)]
r3843: If a connection to a DC is requested, open connections simultaeneously to all
DCs found. The first one to reply wins.

Volker

11 years agor3841: Time out in ads search queries. Even AD servers can hang.
Volker Lendecke [Thu, 18 Nov 2004 11:43:14 +0000 (11:43 +0000)]
r3841: Time out in ads search queries. Even AD servers can hang.

Volker

11 years agor3839: Some more specific NT security descriptors we've come across. Map them to
Volker Lendecke [Thu, 18 Nov 2004 08:16:59 +0000 (08:16 +0000)]
r3839: Some more specific NT security descriptors we've come across. Map them to
defined AFS acls. Thanks to Horst Birthelmer.

Volker

11 years agor3824: Fix crash in api_RNetShareEnum. ServicePtrs[] may contain invalid
Volker Lendecke [Wed, 17 Nov 2004 21:22:35 +0000 (21:22 +0000)]
r3824: Fix crash in api_RNetShareEnum. ServicePtrs[] may contain invalid
entries. Happened after rev3708 removed non-existing cups printers.

Volker

11 years agor3823: Fix bug reported by Tom Dickson <tdickson@inostor.com> with
Jeremy Allison [Wed, 17 Nov 2004 21:05:47 +0000 (21:05 +0000)]
r3823: Fix bug reported by Tom Dickson <tdickson@inostor.com> with
follow symlinks = no reporting an incorrect error when it should
be access denied.
Jeremy.

11 years agor3819: Updated config.guess/config.sub from
Jeremy Allison [Wed, 17 Nov 2004 20:05:31 +0000 (20:05 +0000)]
r3819: Updated config.guess/config.sub from
ftp://ftp.gnu.org/pub/gnu/config/config.guess
ftp://ftp.gnu.org/pub/gnu/config/config.sub
as requested in bugid #2048.
Jeremy.

11 years agor3817: Fix from Lars Müller <lmuelle@suse.de> for bug #2050, calculate
Jeremy Allison [Wed, 17 Nov 2004 19:52:26 +0000 (19:52 +0000)]
r3817: Fix from Lars Müller <lmuelle@suse.de> for bug #2050, calculate
max_fd for select correctly.
Jeremy.

11 years agor3816: Added fn to remove an ACL from a file. Now need client code to test this.
Jeremy Allison [Wed, 17 Nov 2004 18:56:48 +0000 (18:56 +0000)]
r3816: Added fn to remove an ACL from a file. Now need client code to test this.
How do the share mask/modes fit into this code... Need to think about this.
Jeremy.

11 years agor3814: At least use a proper constant name (aRONLY :-) not a number.
Jeremy Allison [Wed, 17 Nov 2004 17:51:27 +0000 (17:51 +0000)]
r3814: At least use a proper constant name (aRONLY :-) not a number.
That way we can at least remember why we did this :-).
Jeremy.

11 years agor3811: Believe it or not, but this patch seems to be necessary. If someone sets a
Volker Lendecke [Wed, 17 Nov 2004 15:56:47 +0000 (15:56 +0000)]
r3811: Believe it or not, but this patch seems to be necessary. If someone sets a
folder icon in the start menu and saves the profile on a samba server, after
logging in again this setting is gone. Why is this? The folder for which the
icon is set must have the read only flag set. If it is not set, the
desktop.ini file (the file containing the icon reference) inside that folder
is ignored.

lp_profile_acls is a hack for such a situation, so overload this parameter
with another profile-related hack.

Volker

11 years agor3796: Patch from Jay Fenlason <fenlason@redhat.com>. Don't free static buffers.
Jeremy Allison [Wed, 17 Nov 2004 01:43:36 +0000 (01:43 +0000)]
r3796: Patch from Jay Fenlason <fenlason@redhat.com>. Don't free static buffers.
Jeremy.

11 years agor3794: Added set posix acl functionality into the UNIX extensions code.
Jeremy Allison [Wed, 17 Nov 2004 01:35:35 +0000 (01:35 +0000)]
r3794: Added set posix acl functionality into the UNIX extensions code.
One part missing - delete file acl (to be added asap). No client
code yet, also needs testing with valgrind.
Jeremy.

11 years agor3778: BUG 2023: patch from Buchan Milne <bgmilne@mandrake.org> fix mandrake packagin...
Gerald Carter [Tue, 16 Nov 2004 03:27:19 +0000 (03:27 +0000)]
r3778: BUG 2023: patch from Buchan Milne <bgmilne@mandrake.org> fix mandrake packaging for Samba 3.0.9 -- still need to review this libsmbclient install location

11 years agor3776: Fix for bug #2038 from Johann Hanne <jhml@gmx.net>. Don't check
Jeremy Allison [Mon, 15 Nov 2004 23:03:37 +0000 (23:03 +0000)]
r3776: Fix for bug #2038 from Johann Hanne <jhml@gmx.net>. Don't check
for no groups after every lookup - move check to the end as we should
only fail if all lookups fail.
Jeremy.

11 years agor3772: BUG 2006: patch from Michel Gravey <michel.gravey@optogone.com>; fix build...
Gerald Carter [Mon, 15 Nov 2004 21:21:50 +0000 (21:21 +0000)]
r3772: BUG 2006: patch from Michel Gravey <michel.gravey@optogone.com>; fix build when using gcc 3.0

11 years agor3771: Fix bug #1952: no comment on XP shutdown
Jim McDonough [Mon, 15 Nov 2004 20:57:27 +0000 (20:57 +0000)]
r3771: Fix bug #1952: no comment on XP shutdown

Try INITSHUTDOWN pipe first, used by newer clients.  If it fails, fall back
to WINREG.

11 years agor3764: Ensure on failure that *res is always NULL.
Jeremy Allison [Mon, 15 Nov 2004 18:57:22 +0000 (18:57 +0000)]
r3764: Ensure on failure that *res is always NULL.
Check for malloc fail. Fixes for bug #2036.
Jeremy.

11 years agor3761: Fix bug #1932: crash when non-root invokes net getlocalsid
Jim McDonough [Mon, 15 Nov 2004 17:20:58 +0000 (17:20 +0000)]
r3761: Fix bug #1932: crash when non-root invokes net getlocalsid

First check to see if we can open secrets.tdb.

11 years agor3758: Increase a debug level. This error condition can actually happen in real life
Volker Lendecke [Mon, 15 Nov 2004 13:18:12 +0000 (13:18 +0000)]
r3758: Increase a debug level. This error condition can actually happen in real life
if you have several competing winbinds that might get a lag due to
replication.

Volker

11 years agor3727: Update from Rob Foehl (rwf@loonybin.net):
Jim McDonough [Sun, 14 Nov 2004 03:12:46 +0000 (03:12 +0000)]
r3727: Update from Rob Foehl (rwf@loonybin.net):
Try modifying printer before adding it,since add logic appends ObjectClass to mods, which caused a contraint violation when modifying an already published printer.

Also cleanup the check_published_printers() logic when verifying published status on startup.

11 years agor3714: Getfacl now seems to work on files and directories. Next do setfacl
Jeremy Allison [Sat, 13 Nov 2004 00:35:31 +0000 (00:35 +0000)]
r3714: Getfacl now seems to work on files and directories. Next do setfacl
and port to Samba4.
Jeremy.

11 years agor3713: Implementation of get posix acls in UNIX extensions. Passes valgrind.
Jeremy Allison [Fri, 12 Nov 2004 23:42:12 +0000 (23:42 +0000)]
r3713: Implementation of get posix acls in UNIX extensions. Passes valgrind.
Need to add printout functions in client and set posix acl in server.
SteveF - take a look at this for the cifsfs client !
Once this is working and tested the next step is to write this up for
the UNIX extensions spec. documents.
Jeremy.

11 years agor3711: Fix KRB5_SETPW-defines, no change in behaviour.
Günther Deschner [Fri, 12 Nov 2004 23:13:03 +0000 (23:13 +0000)]
r3711: Fix KRB5_SETPW-defines, no change in behaviour.

A value of '2' is due to an expired internet draft, while 0xff80 comes
from RFC3244. See Bugzilla #1661 for details.

Thanks to Luke Mewburn <lukem@NetBSD.org> for his research.

Guenther

11 years agor3710: fix another safe_strcpy() warning SPAM with the printer drivcename
Gerald Carter [Fri, 12 Nov 2004 22:15:55 +0000 (22:15 +0000)]
r3710: fix another safe_strcpy() warning SPAM with the printer drivcename

11 years agor3709: Allow to build testsuite for libsmbclient when libsmbclient.so is
Günther Deschner [Fri, 12 Nov 2004 22:05:27 +0000 (22:05 +0000)]
r3709: Allow to build testsuite for libsmbclient when libsmbclient.so is
located not in -L/root/samba-head-old/source/bin.

Patch from Lars Mueller <lmuelle@suse.de>, closes Bugzilla #2013.

Guenther

11 years agor3708: BUG 1838: patch from Gavrie Philipson <gavrie@disksites.com> to remove stale...
Gerald Carter [Fri, 12 Nov 2004 17:39:32 +0000 (17:39 +0000)]
r3708: BUG 1838: patch from Gavrie Philipson <gavrie@disksites.com> to remove stale printers when reloading after a sighup

11 years agor3707: 2028: avoid false error messages when copying a long printer name to the devic...
Gerald Carter [Fri, 12 Nov 2004 17:23:01 +0000 (17:23 +0000)]
r3707: 2028: avoid false error messages when copying a long printer name to the device mode

11 years agor3705: Nobody has commented, so I'll take this as an ack...
Volker Lendecke [Fri, 12 Nov 2004 15:49:47 +0000 (15:49 +0000)]
r3705: Nobody has commented, so I'll take this as an ack...

abartlet, I'd like to ask you to take a severe look at this!

We have solved the problem to find the global groups a user is in twice: Once
in auth_util.c and another time for the corresponding samr call. The attached
patch unifies these and sends them through the passdb backend (new function
pdb_enum_group_memberships). Thus it gives pdb_ldap.c the chance to further
optimize the corresponding call if the samba and posix accounts are unified by
issuing a specialized ldap query.

The parameter to activate this ldapsam behaviour is

ldapsam:trusted = yes

Volker

11 years agor3704: Implement a cache get saves the result of a pdb_getsampwnam for later
Volker Lendecke [Fri, 12 Nov 2004 15:30:51 +0000 (15:30 +0000)]
r3704: Implement a cache get saves the result of a pdb_getsampwnam for later
retrieval by pdb_getsampwsid. This solves our problem that we do lots of calls
to LDAP during a typical XP login. XP does a lookupnames, then an openuser and
some queryinfo stuff. Lookupnames triggers the initial getsampwnam, and all
the subsequent ones make us call getsampwsid. This patch gets this down to one
call to LDAP.

Yes, a more "correct" way would be to stick the information to the open user
handle, but this one is simpler and saves the LDAP roundtrip for the openuser
call.

Volker

11 years agor3702: This is a getpwnam-cache. It is mainly to speed up Samba with slow nss
Volker Lendecke [Fri, 12 Nov 2004 15:01:40 +0000 (15:01 +0000)]
r3702: This is a getpwnam-cache. It is mainly to speed up Samba with slow nss
backends such as nss_ldap.

Volker

11 years agor3697: BUG 2027: fix multiple declaration of MD5_CTX (do we really care about DOS...
Gerald Carter [Fri, 12 Nov 2004 03:36:15 +0000 (03:36 +0000)]
r3697: BUG 2027: fix multiple declaration of MD5_CTX (do we really care about DOS compilers?).  Will watch build farm though.

11 years agor3693: Correctly detect errno for no acl/ea support.
Jeremy Allison [Fri, 12 Nov 2004 02:16:00 +0000 (02:16 +0000)]
r3693: Correctly detect errno for no acl/ea support.
Jeremy

11 years agor3688: fix output of smbstatus to make the man page; fix -L, -p, & -S and the -u...
Gerald Carter [Thu, 11 Nov 2004 23:30:32 +0000 (23:30 +0000)]
r3688: fix output of smbstatus to make the man page; fix -L, -p, & -S and the -u <username> functionality

11 years agor3685: setting the 3.0 branch to 3.0.10pre1-svn
Gerald Carter [Thu, 11 Nov 2004 20:41:28 +0000 (20:41 +0000)]
r3685: setting the 3.0 branch to 3.0.10pre1-svn

11 years agor3683: BUG 2017: fix testparm reporting for the passwd program string
Gerald Carter [Thu, 11 Nov 2004 18:11:40 +0000 (18:11 +0000)]
r3683: BUG 2017: fix testparm reporting for the passwd program string

11 years agor3682: fix seg fault in lanman printing code caused by uninitialized variable
Gerald Carter [Thu, 11 Nov 2004 17:40:49 +0000 (17:40 +0000)]
r3682: fix seg fault in lanman printing code caused by uninitialized variable

11 years agor3681: Fix build of libsmbclient on x86_64.
Günther Deschner [Thu, 11 Nov 2004 15:38:21 +0000 (15:38 +0000)]
r3681: Fix build of libsmbclient on x86_64.
Patch from Lars Mueller <lmuelle@suse.de>

Guenther

11 years agor3674: Stefan Esser <s.esser@e-matters.de> pointed out that the max data
Jeremy Allison [Wed, 10 Nov 2004 23:43:52 +0000 (23:43 +0000)]
r3674: Stefan Esser <s.esser@e-matters.de> pointed out that the max data
value is only valid on the initial trans/trans2/nttrans request,
so if there are secondary requests we can't read it from them. Read
it from the initial request and pass as a parameter for those functions
that need it.
Jeremy.