]> git.ipfire.org Git - thirdparty/samba.git/log
thirdparty/samba.git
24 years ago*Experimental* new large-scaling printer code. Splits printing.tdb into
Jeremy Allison [Thu, 11 Jul 2002 00:06:29 +0000 (00:06 +0000)] 
*Experimental* new large-scaling printer code. Splits printing.tdb into
a separate tdb per printer, but only keeps (currently one) tdb open at
a time (although this is easily changed by changing a #define). Needs
scalability testing with large numbers of printers now....
Jeremy.

24 years agoIf we get a SID from group mapping, no need to check it's prefix.
Andrew Bartlett [Wed, 10 Jul 2002 07:26:35 +0000 (07:26 +0000)] 
If we get a SID from group mapping, no need to check it's prefix.

Just set it directly.

Andrew Bartlett

24 years agosome support functions to ease the next commit.
Simo Sorce [Tue, 9 Jul 2002 23:59:46 +0000 (23:59 +0000)] 
some support functions to ease the next commit.

24 years ago8-byte alignment needed for beginning driver_info_level_6.
Gerald Carter [Tue, 9 Jul 2002 23:23:19 +0000 (23:23 +0000)] 
8-byte alignment needed for beginning driver_info_level_6.
Verified by looking at NT4 and 2k servers.  First time
for everything I guess.

24 years agoDocumented -I option to rpcclient.
Tim Potter [Tue, 9 Jul 2002 19:13:11 +0000 (19:13 +0000)] 
Documented -I option to rpcclient.

24 years agoAdded -I option to rpcclient.
Tim Potter [Tue, 9 Jul 2002 19:11:26 +0000 (19:11 +0000)] 
Added -I option to rpcclient.

24 years agomake sure we disable referrals in all ldap searches - they are badly
Andrew Tridgell [Tue, 9 Jul 2002 15:06:51 +0000 (15:06 +0000)] 
make sure we disable referrals in all ldap searches - they are badly
broken

24 years agofix declaration of global_in_nmbd
Andrew Tridgell [Tue, 9 Jul 2002 14:47:03 +0000 (14:47 +0000)] 
fix declaration of global_in_nmbd

24 years agothe last WINS update broke self registration when we are a WINS
Andrew Tridgell [Tue, 9 Jul 2002 14:17:26 +0000 (14:17 +0000)] 
the last WINS update broke self registration when we are a WINS
server. The real problem is all the special cases we had for when we
are a wins server as opposed to when we are using a 'real' wins
server.

This patch removes the special cases. We now accept non-broadcast
packets from ourselves and we use ourselves as a wins server when we
are one. This gets rid of the special cases and simplifies things
quite a bit.

It all seems to work, but there are bound to be problems found later.

24 years agoMake it clear that the debug comment is the same as the command being tested
Andrew Bartlett [Tue, 9 Jul 2002 13:12:27 +0000 (13:12 +0000)] 
Make it clear that the debug comment is the same as the command being tested
for failure.

Andrew Bartlett

24 years agofixup ADDPRINTERDRIVEREX some more. unknown uint32 for now.
Gerald Carter [Mon, 8 Jul 2002 22:01:51 +0000 (22:01 +0000)] 
fixup ADDPRINTERDRIVEREX some more.  unknown uint32 for now.

24 years agoFix length in snprintf invocation to account for "dn:" in sasl gssapi bind.
Jim McDonough [Mon, 8 Jul 2002 20:46:54 +0000 (20:46 +0000)] 
Fix length in snprintf invocation to account for "dn:" in sasl gssapi bind.
Also remove unused line which incremented pointer by the wrong length anyway.

Provided by Anthony Liguori (aliguori@us.ibm.com).

24 years agofix typo
Gerald Carter [Mon, 8 Jul 2002 20:30:54 +0000 (20:30 +0000)] 
fix typo

24 years agoadded parsing routines for SPOOLSS_ADDPRINTERDRIVEREX and
Gerald Carter [Mon, 8 Jul 2002 20:15:47 +0000 (20:15 +0000)] 
added parsing routines for SPOOLSS_ADDPRINTERDRIVEREX and
SPOOLSS_DELETEPRINTERDRIVEREX.  Ran them through some testing.
I know I'm off by 2 x uint32's in the former RPC.

24 years ago* basic implementation of SPOOLSS_DELETEPRINTERDATAEX and
Gerald Carter [Mon, 8 Jul 2002 18:34:55 +0000 (18:34 +0000)] 
* basic implementation of SPOOLSS_DELETEPRINTERDATAEX and
  SPOOLSS_DELETEPRINTERKEY
* stub funnctions for SPOOLSS_ADDPRINTERDRIVEREX and
  SPOOLSS_DELETEPRINTERDRIVEREX

24 years agoFix for LPRng from James Henstridge james@daa.com.au.
Jeremy Allison [Mon, 8 Jul 2002 18:24:59 +0000 (18:24 +0000)] 
Fix for LPRng from James Henstridge james@daa.com.au.
Jeremy.

24 years agoKill off const warnings - add a pile of const to various places.
Andrew Bartlett [Mon, 8 Jul 2002 02:14:57 +0000 (02:14 +0000)] 
Kill off const warnings - add a pile of const to various places.

24 years agoFix const warning
Andrew Bartlett [Mon, 8 Jul 2002 01:40:35 +0000 (01:40 +0000)] 
Fix const warning

24 years agoJerry: Sorry if I am stepping on toes here, but this should fix the compile on
Andrew Bartlett [Mon, 8 Jul 2002 01:06:38 +0000 (01:06 +0000)] 
Jerry:  Sorry if I am stepping on toes here, but this should fix the compile on
solaris CC (void* arithmatic) and fixes the other warnings (global variable
shadows) that gcc has with this file.

Andrew Bartlett

24 years ago(this should have been part of the previous commit)
Andrew Bartlett [Mon, 8 Jul 2002 00:40:57 +0000 (00:40 +0000)] 
(this should have been part of the previous commit)

Add a function to display 'sid types' as strings - makes rpcclient outptut
and DEBUG() logs much eaiser to understand.

Move the enum for SID types to smb.h, becouse is really isn't LSA specific any
more.

Andrew Bartlett

24 years agoUpdates to the 'name -> sid' code:
Andrew Bartlett [Sun, 7 Jul 2002 03:34:24 +0000 (03:34 +0000)] 
Updates to the 'name -> sid' code:

Correct the 'none mapped' behaviour, (so that it matches Win2k) and  add a
function to make the SID types appear as text strings in logs/rpcclient.

Also, remove a silly case that would cause 'failure' to be 'success'.  (Might
look at this a bit more in future).

Andrew Bartlett

24 years agoCopy the NT_TOKEN to the pipe, so the SAMR can use it for access control.
Andrew Bartlett [Sat, 6 Jul 2002 05:08:35 +0000 (05:08 +0000)] 
Copy the NT_TOKEN to the pipe, so the SAMR can use it for access control.

Andrew Bartlett

24 years agoIncrese the maximum non-unix-account ID (becouse the Compaq Test Drive systems
Andrew Bartlett [Sat, 6 Jul 2002 04:12:40 +0000 (04:12 +0000)] 
Increse the maximum non-unix-account ID (becouse the Compaq Test Drive systems
have a lot of users ;-).

Andrew Bartlett

24 years agoRemove C++ comments.
Tim Potter [Fri, 5 Jul 2002 04:08:32 +0000 (04:08 +0000)] 
Remove C++ comments.

24 years agoFix debug comment.
Andrew Bartlett [Fri, 5 Jul 2002 01:51:49 +0000 (01:51 +0000)] 
Fix debug comment.

24 years agoFix from Andrew Esh for coredump...
Jeremy Allison [Thu, 4 Jul 2002 21:32:08 +0000 (21:32 +0000)] 
Fix from Andrew Esh for coredump...
Jeremy.

24 years agofixed the geyed out registry keys. Stupid typo on my part.
Gerald Carter [Thu, 4 Jul 2002 13:18:56 +0000 (13:18 +0000)] 
fixed the geyed out registry keys.  Stupid typo on my part.
Now on to plug in the printing tdbs to the registry view....

jerry

24 years agoMore code from "Kai Krueger" <kai@kruegernetz.de>, this time starting to make
Andrew Bartlett [Thu, 4 Jul 2002 13:11:54 +0000 (13:11 +0000)] 
More code from "Kai Krueger" <kai@kruegernetz.de>, this time starting to make
the 'user cannot change password' button work.  Needs help from a future SAM
backend, but at least this parses the data, and returns an error.

Andrew Bartlett

24 years ago* fix seg fault caused by missing DLIST_REMOVE in free_fn()
Gerald Carter [Thu, 4 Jul 2002 12:51:26 +0000 (12:51 +0000)] 
* fix seg fault caused by missing DLIST_REMOVE in free_fn()

* fixed constant vaolued returned by REG_UNKNOWN_1A

24 years agoPatch to add security descriptors to the SAMR pipe.
Andrew Bartlett [Thu, 4 Jul 2002 08:45:45 +0000 (08:45 +0000)] 
Patch to add security descriptors to the SAMR pipe.

This is not the final solution, I think this will probably changed with the
move to the new SAM subsystem, but it allows some research and gives us
somthing to start with.

It should also help with getting proper NT_TOKEN passing set-up.

Original patch by "Kai Krueger" <kai@kruegernetz.de>, which I have modified to
pass back NTSTATUS returns in more places and to use a little more common code.

Andrew Bartlett

24 years agoAlways free_conn() after all the DEBUG()s etc.
Andrew Bartlett [Thu, 4 Jul 2002 08:39:06 +0000 (08:39 +0000)] 
Always free_conn() after all the DEBUG()s etc.

24 years agoFixed incorrect debug.
Tim Potter [Thu, 4 Jul 2002 03:32:36 +0000 (03:32 +0000)] 
Fixed incorrect debug.

24 years agofirst cut at implementing support for browsing printer and driver driver
Gerald Carter [Wed, 3 Jul 2002 19:15:26 +0000 (19:15 +0000)] 
first cut at implementing support for browsing printer and driver driver
via regedt32.exe.  The regsitry.tdb is only a framework.  It is not
intended to store values, only key/subkey structure.  The data
will be retrieved from nt*tdb (for printers) creating a virtual view
of the data.

You can currently connect to a Samba box using regedt32.exe (haven't
tried regedit.exe).  Some basic keys are created in registry.tdb
for use.

There are two problems....

  * something is getting freed in the winreg code that causes heap
    corruption later on.  As long as you don't play with the winreg
    server functionality, I don't think you'll be bitten by this.

  * no access controls are currently implemented

  * I can't browse HKLM because regedt32 greys out the SYSTEM subkey.

ok so that was three....

24 years agoBreak up the passdb objects (to allow RPC clients to link without brining in
Andrew Bartlett [Wed, 3 Jul 2002 07:37:54 +0000 (07:37 +0000)] 
Break up the passdb objects (to allow RPC clients to link without brining in
*.o) and implment new enum_dom_users code in the SAMR RPC subsystem.

Incresingly, we are using the pdb_get_{user,group}_sid() functions, in the
eventual hope that we might one day support muliple domains off a single
passdb.  To extract the RID, we use sid_peek_check_rid(), and supply an
'expected' domain SID.

The id21 -> SAM_ACCOUNT and id23 -> SAM_ACCOUNT code has been moved to
srv_samr_util.c, to ease linking in passdb users.

Compatiblity code that uses 'get_global_sam_sid()' for the 'expected' sid is in
pdb_compat.c

Andrew Bartlett

24 years agoMake these functions static. These are not mentioned in the external header,
Andrew Bartlett [Wed, 3 Jul 2002 07:22:45 +0000 (07:22 +0000)] 
Make these functions static.  These are not mentioned in the external header,
and appear to be functions for internal use.

Richard:  please check.

Andrew Bartlett

24 years agoAdd my copyright (which I should have added months ago...)
Andrew Bartlett [Wed, 3 Jul 2002 07:21:07 +0000 (07:21 +0000)] 
Add my copyright (which I should have added months ago...)

24 years agoFix the spelling in the LDAP attributes
Andrew Bartlett [Wed, 3 Jul 2002 06:55:31 +0000 (06:55 +0000)] 
Fix the spelling in the LDAP attributes

24 years agoFix the forword prototype to be a static for this static function.
Andrew Bartlett [Wed, 3 Jul 2002 06:48:52 +0000 (06:48 +0000)] 
Fix the forword prototype to be a static for this static function.

24 years agoRemove invalid comment - these are all 'unix' strings now.
Andrew Bartlett [Wed, 3 Jul 2002 00:47:39 +0000 (00:47 +0000)] 
Remove invalid comment - these are all 'unix' strings now.

24 years agoKill off codepage related stuff, now we don't use codepages any more.
Andrew Bartlett [Wed, 3 Jul 2002 00:44:39 +0000 (00:44 +0000)] 
Kill off codepage related stuff, now we don't use codepages any more.

Andrew Bartlett

24 years agofix typo in unix extensions description
Gerald Carter [Tue, 2 Jul 2002 14:17:42 +0000 (14:17 +0000)] 
fix typo in unix extensions description

24 years agoAddress the string_sub problem by changing len = 0 to mean "no expand".
Jeremy Allison [Tue, 2 Jul 2002 06:34:27 +0000 (06:34 +0000)] 
Address the string_sub problem by changing len = 0 to mean "no expand".
Went through and checked all string_subs I could to ensure they're being
used correctly.
Jeremy.

24 years agofixed our winreg parsing to handle a diifferent form given to us by
Andrew Tridgell [Tue, 2 Jul 2002 04:22:59 +0000 (04:22 +0000)] 
fixed our winreg parsing to handle a diifferent form given to us by
WinXP. This fixes setting security decsriptors from XP Professional.

24 years agoMerge of constant renames from APPLIANCE_HEAD.
Tim Potter [Tue, 2 Jul 2002 04:07:32 +0000 (04:07 +0000)] 
Merge of constant renames from APPLIANCE_HEAD.

24 years agofixed a prototype problem in client.c
Andrew Tridgell [Tue, 2 Jul 2002 01:35:07 +0000 (01:35 +0000)] 
fixed a prototype problem in client.c

24 years agoused findstatic.pl to make some variables static and remove some dead
Andrew Tridgell [Mon, 1 Jul 2002 09:12:41 +0000 (09:12 +0000)] 
used findstatic.pl to make some variables static and remove some dead
code

24 years agoads_mod_ber should be static, not public
Andrew Tridgell [Mon, 1 Jul 2002 07:30:40 +0000 (07:30 +0000)] 
ads_mod_ber should be static, not public
this fixes the huge number of struct berval warnings on non-ads
compiles

24 years agofixed multi-homed re-registration of names when we are a WINS
Andrew Tridgell [Mon, 1 Jul 2002 07:19:12 +0000 (07:19 +0000)] 
fixed multi-homed re-registration of names when we are a WINS
server. We were rejecting them, leaving the name unregistered!

24 years agodon't start the async dns process unless we actually need it!
Andrew Tridgell [Mon, 1 Jul 2002 07:15:13 +0000 (07:15 +0000)] 
don't start the async dns process unless we actually need it!
(ie. check for lp_dns_proxy())

24 years agobias the lookup sorting towards directly reachable IPs
Andrew Tridgell [Mon, 1 Jul 2002 05:39:32 +0000 (05:39 +0000)] 
bias the lookup sorting towards directly reachable IPs

24 years agosort name query responses by how far they are from our interface
Andrew Tridgell [Mon, 1 Jul 2002 05:09:29 +0000 (05:09 +0000)] 
sort name query responses by how far they are from our interface
broadcast addresses. This makes it far more likely that we will try to
talk to an interface that is routable from one of our interfaces.

24 years agofixed a bug handling startup when the ads server is not contactable
Andrew Tridgell [Mon, 1 Jul 2002 04:47:29 +0000 (04:47 +0000)] 
fixed a bug handling startup when the ads server is not contactable

24 years agoThe 17-bit length field in the header contains the number of
Christopher R. Hertel [Mon, 1 Jul 2002 03:42:04 +0000 (03:42 +0000)] 
The 17-bit length field in the header contains the number of
bytes which follow the header, not the full packet size.

    [Yes, the length field is either 17-bits, or (per the RFCs) it is a
    16-bit length field preceeded by an 8-bit flags field of which only
    the low-order bit may be used.  If that bit is set, then add 65536 to
    the 16-bit length field.  (In other words, it's a 17-bit unsigned
    length field.)
    ...unless, of course, the transport is native TCP [port 445] in which
    case the length field *might* be 24-bits wide.]

Anyway, the change is a very minor one.  We were including the four bytes
of the header in the length count and, as a result, sending four bytes of
garbage at the end of the SESSION REQUEST packet.

Small fix in function cli_session_request().

24 years agofixed a makefile syntax error that was breaking the build on some
Andrew Tridgell [Sun, 30 Jun 2002 03:47:46 +0000 (03:47 +0000)] 
fixed a makefile syntax error that was breaking the build on some
systems

24 years agoShut down printing tdb correctly - this will be more important when I go
Jeremy Allison [Fri, 28 Jun 2002 23:11:31 +0000 (23:11 +0000)] 
Shut down printing tdb correctly - this will be more important when I go
to a tdb per-queue for scalability.
Jeremy.

24 years agoAdded limit on SPOOLSS pipes from 2.2.x
Jeremy Allison [Fri, 28 Jun 2002 19:41:17 +0000 (19:41 +0000)] 
Added limit on SPOOLSS pipes from 2.2.x
Jeremy.

24 years agoTest how many pipes we can open.
Jeremy Allison [Fri, 28 Jun 2002 19:28:58 +0000 (19:28 +0000)] 
Test how many pipes we can open.
Jeremy.

24 years agoConst cleanup...should only be 2 warnings left from calling lib/util_str.c
Jim McDonough [Fri, 28 Jun 2002 19:25:11 +0000 (19:25 +0000)] 
Const cleanup...should only be 2 warnings left from calling lib/util_str.c
str_list_copy().  Perhaps its proto should be fixed.

24 years agoSTRING was being redefined...
Jeremy Allison [Fri, 28 Jun 2002 18:17:11 +0000 (18:17 +0000)] 
STRING was being redefined...
Jeremy.

24 years agodon't backup to a newer file
Andrew Tridgell [Fri, 28 Jun 2002 03:52:22 +0000 (03:52 +0000)] 
don't backup to a newer file

24 years agomake net join a bit less verbose
Andrew Tridgell [Fri, 28 Jun 2002 03:51:31 +0000 (03:51 +0000)] 
make net join a bit less verbose
these errors happen all the time, so they shouldn't be level 0

24 years agodon't warn on non-existant files in map_file(), let the caller handle any warning
Andrew Tridgell [Fri, 28 Jun 2002 03:19:20 +0000 (03:19 +0000)] 
don't warn on non-existant files in map_file(), let the caller handle any warning

24 years agodon't warn on the loading of zero length files. This fixes the
Andrew Tridgell [Fri, 28 Jun 2002 02:54:40 +0000 (02:54 +0000)] 
don't warn on the loading of zero length files. This fixes the
'valid.dat' warning

24 years agoProper merge of all the working printing stuff from APPLIANCE_HEAD.
Jeremy Allison [Fri, 28 Jun 2002 00:17:15 +0000 (00:17 +0000)] 
Proper merge of all the working printing stuff from APPLIANCE_HEAD.
Now let's keep this in sync !
Jeremy.

24 years agoIt's fairly obvious that no one has tried to upload a driver
Gerald Carter [Thu, 27 Jun 2002 18:10:56 +0000 (18:10 +0000)] 
It's fairly obvious that no one has tried to upload a driver
to a Samba print server running HEAD in a while.  This has been broken
since tridge's changes to make_connection() to not do the chdir()
to the connect_path.  Sorry it took me so long to get around to fixing it.

The problem occured with our internal use of make_connection().

jerry

24 years agofixed a link problem with global_in_nmbd
Andrew Tridgell [Thu, 27 Jun 2002 14:54:01 +0000 (14:54 +0000)] 
fixed a link problem with global_in_nmbd

24 years agoThe next phase in the WINS rewrite!
Andrew Tridgell [Thu, 27 Jun 2002 14:37:17 +0000 (14:37 +0000)] 
The next phase in the WINS rewrite!

We now cope wiith multiple WINS groups and multiple failover servers
for release and refresh as well as registration. We also do the regitrations
in the same fashion as W2K does, where we don't try to register the next
IP in the list for a name until the WINS server has acked the previos IP.
This prevents us flooding the WINS server and also seems to make for much
more reliable multi-homed registration.

I also changed the dead WINS server code to mark pairs of IPs dead,
not individual IPs. The idea is that a WINS server might be dead from
the point of view of one of our interfaces, but not another, so we
need to keep talking to it on one while moving onto a failover WINS
server on the other interface. This copes much better with partial
LAN outages and weird routing tables.

24 years agoJelmer has been keeping on top of the typos.
Andrew Bartlett [Thu, 27 Jun 2002 14:23:48 +0000 (14:23 +0000)] 
Jelmer has been keeping on top of the typos.

24 years agoWe don't use SSL any more...
Andrew Bartlett [Thu, 27 Jun 2002 14:12:30 +0000 (14:12 +0000)] 
We don't use SSL any more...

(from jelmer)

Andrew Bartlett

24 years agoa null grp_sid should have a 0 offset. Also removed a few more unnecessary
Gerald Carter [Thu, 27 Jun 2002 01:12:37 +0000 (01:12 +0000)] 
a null grp_sid should have a 0 offset.  Also removed a few more unnecessary
prs_align() for sec_desc.

24 years agoUsing 1 for a tdb hash size makes for slow inserts....
Jeremy Allison [Thu, 27 Jun 2002 00:13:20 +0000 (00:13 +0000)] 
Using 1 for a tdb hash size makes for slow inserts....
Jeremy.

24 years agoRe-enable kernel change notification on s390 as it has been fixed in the
Jim McDonough [Wed, 26 Jun 2002 23:45:59 +0000 (23:45 +0000)] 
Re-enable kernel change notification on s390 as it has been fixed in the
kernel.

24 years agomore merges from SAMBA_2_2
Gerald Carter [Wed, 26 Jun 2002 23:26:10 +0000 (23:26 +0000)] 
more merges from SAMBA_2_2

24 years agomerge from APP_HEAD
Gerald Carter [Wed, 26 Jun 2002 19:08:32 +0000 (19:08 +0000)] 
merge from APP_HEAD

24 years agoTypo...
Jeremy Allison [Wed, 26 Jun 2002 17:37:53 +0000 (17:37 +0000)] 
Typo...
Jeremy.

24 years agoAdded prs_align_uint16 for SD aligns.
Jeremy Allison [Wed, 26 Jun 2002 17:36:58 +0000 (17:36 +0000)] 
Added prs_align_uint16 for SD aligns.
Jeremy.

24 years agoReduce compiler warnings.
Jim McDonough [Wed, 26 Jun 2002 17:27:30 +0000 (17:27 +0000)] 
Reduce compiler warnings.

24 years agoresolve_wins() now needs to be a public function
Andrew Tridgell [Wed, 26 Jun 2002 12:49:59 +0000 (12:49 +0000)] 
resolve_wins() now needs to be a public function

24 years agoThis commit finally gives us multiple wins server groups. We now
Andrew Tridgell [Wed, 26 Jun 2002 12:17:11 +0000 (12:17 +0000)] 
This commit finally gives us multiple wins server groups. We now
accept an extended syntax for 'wins server' like this:

  wins server = group1:192.168.2.10 group2:192.168.3.99 group1:192.168.0.1

The tags before the IPs don't mean anything, they are just a way of
grouping IPs together. If you use the old syntax (ie. no ':') then
an implicit group name of '*' is used. In general I'd recommend people
use interface names for the group names, but it doesn't matter much.

When we register in nmbd we try to register all our IPs with each group
of WINS servers. We keep trying until all of them are registered with
every group, falling back to the failover WINS servers for each group
as we go.

When we do a WINS lookup we try each of the WINS servers for each group.
If a WINS server for a group gives a negative answer then we give up
on that group and move to the next group. If it times out then
we move to the next failover wins server in the group.

In either case, if a WINS server doesn't respond then we mark it dead
for 10 minutes, to prevent lengthy waits for dead servers.

24 years agowe never pass any userdata when doing name registrations on the
Andrew Tridgell [Wed, 26 Jun 2002 08:58:03 +0000 (08:58 +0000)] 
we never pass any userdata when doing name registrations on the
unicast subnet, so remove that parameter. That frees up userdata so I
can start using it to indicate which wins server tag we are
registering (more about wins 'tags' later ...)

24 years agoremoved the wins name registration code from libsmbclient
Andrew Tridgell [Wed, 26 Jun 2002 08:09:28 +0000 (08:09 +0000)] 
removed the wins name registration code from libsmbclient

it is *completely* bogus for our client code to be doing wins
registrations. Not only is it slow as hell (think about when a wins
server is down) but how the heck is going to answer the queries that
will later come in for our name? And what happens when libsmbclient
sends registrations and nmbd then gets the WACK response from the wins
server? we end up losing our name!

Name registration is a job for nmbd, not for clients.

24 years agomade the wins list handling a littler clearer
Andrew Tridgell [Wed, 26 Jun 2002 07:27:00 +0000 (07:27 +0000)] 
made the wins list handling a littler clearer

24 years agomumble ... fix typo ... mumble
Andrew Tridgell [Wed, 26 Jun 2002 07:23:35 +0000 (07:23 +0000)] 
mumble ... fix typo ... mumble

24 years agofixed we_are_multihomed() to cope with dynamic interfaces (ie. don't
Andrew Tridgell [Wed, 26 Jun 2002 07:23:00 +0000 (07:23 +0000)] 
fixed we_are_multihomed() to cope with dynamic interfaces (ie. don't
cache the result!)

24 years ago- completely rewrote the wins_srv.c code. It is now much simpler, and
Andrew Tridgell [Wed, 26 Jun 2002 06:44:37 +0000 (06:44 +0000)] 
- completely rewrote the wins_srv.c code. It is now much simpler, and
gives us a good grounding to properly support multiple wins servers
for different interfaces (which will be coming soon ...)

- fixed our wins registration failover code to actually do failover!
We were not trying to register with a secondary wins server at all
when the primary was down. We now fallback correctly.

- fixed the multi-homed name registration packets so that they work
even in a non-connected network (ie. when one of our interfaces is not
routable from the wins server. Yes, this really happens in the real
world).

24 years agofixed 3 bugs in jeremys trans2 merge. Hopefully it now works.
Andrew Tridgell [Wed, 26 Jun 2002 06:39:54 +0000 (06:39 +0000)] 
fixed 3 bugs in jeremys trans2 merge. Hopefully it now works.

24 years agoreverted some bogus test code that jeremy accidentally committed
Andrew Tridgell [Wed, 26 Jun 2002 06:39:18 +0000 (06:39 +0000)] 
reverted some bogus test code that jeremy accidentally committed

24 years agoAnother bug fix from metze.
Andrew Bartlett [Wed, 26 Jun 2002 03:54:54 +0000 (03:54 +0000)] 
Another bug fix from metze.

24 years agoQualify some of the hexidecimal responses with 0x
Andrew Bartlett [Wed, 26 Jun 2002 02:52:51 +0000 (02:52 +0000)] 
Qualify some of the hexidecimal responses with 0x

24 years agoAnd fix another missing ubiqx...
Andrew Bartlett [Tue, 25 Jun 2002 13:18:09 +0000 (13:18 +0000)] 
And fix another missing ubiqx...

(Didn't have smbmnt compiled in originally, so it missed my tests)

24 years agoFix the smbmnt compile.
Andrew Bartlett [Tue, 25 Jun 2002 11:38:26 +0000 (11:38 +0000)] 
Fix the smbmnt compile.

24 years agoI'm going to try and check this against Win2k shortly, but I'm certain that
Andrew Bartlett [Tue, 25 Jun 2002 11:21:41 +0000 (11:21 +0000)] 
I'm going to try and check this against Win2k shortly, but I'm certain that
'NT_STATUS_OK' is not the right answer here.  Try NO_MORE_ENTRIES for now.

Andrew Bartlett

24 years agoTry to avoid infinite loops when reteriving users - even from broken servers.
Andrew Bartlett [Tue, 25 Jun 2002 11:20:14 +0000 (11:20 +0000)] 
Try to avoid infinite loops when reteriving users - even from broken servers.

Andrew Bartlett

24 years agoUpdate cli_full_connection() to take a 'flags' paramater, and try to get a
Andrew Bartlett [Tue, 25 Jun 2002 08:57:24 +0000 (08:57 +0000)] 
Update cli_full_connection() to take a 'flags' paramater, and try to get a
few more places to use it.

Andrew Bartlett

24 years agoAnd another .cvsignore
Andrew Bartlett [Tue, 25 Jun 2002 08:34:17 +0000 (08:34 +0000)] 
And another .cvsignore

24 years agoAdd a .cvsignore file
Andrew Bartlett [Tue, 25 Jun 2002 08:32:43 +0000 (08:32 +0000)] 
Add a .cvsignore file

24 years agoKill off unnecessary cast.
Andrew Bartlett [Tue, 25 Jun 2002 07:58:29 +0000 (07:58 +0000)] 
Kill off unnecessary cast.

24 years agoBreak up samba's object dependencies, and its prototype includes.
Andrew Bartlett [Tue, 25 Jun 2002 02:29:09 +0000 (02:29 +0000)] 
Break up samba's object dependencies, and its prototype includes.

Now smbclient, net, and swat use their own proto files - now the global
proto.h

The change to libads/kerberos.c was to break up the dependency on secrets.c -
we want to be able to write an ADS client that doesn't need local secrets.

I have other breakups in the works - I will remove the dependency of
rpc_parse on passdb (and therefore secrets.c) shortly.

(NOTE:  This patch does *not* break up includes.h, or other such forbidden
actions).

Andrew Bartlett

24 years agoAdd a couple more DEBUG()s to winbindd.
Andrew Bartlett [Tue, 25 Jun 2002 01:40:44 +0000 (01:40 +0000)] 
Add a couple more DEBUG()s to winbindd.

Andrew Bartlett

24 years agofixed a bug in handling select in the main daemon - this stops the daemon spinning...
Andrew Tridgell [Tue, 25 Jun 2002 00:59:14 +0000 (00:59 +0000)] 
fixed a bug in handling select in the main daemon - this stops the daemon spinning if a signal is received at an inconvenient moment