From lkcl at switchboard.net  Tue Sep  1 15:45:55 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: Detect Slow network connection
In-Reply-To: <35EB1BE2.71B8A40F@eng.auburn.edu>
Message-ID: <Pine.NEB.3.96.980901164530.13257L-100000@cb1-gw.cb1.com>

it depends how long the smbd process takes to react, that's all.

> Does anyone know how NT detects a slow network connection 
> when attempting to login a user and grab his / her profile?


From lkcl at switchboard.net  Tue Sep  1 15:48:47 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: browse synchronisation
In-Reply-To: <19980831225016Z12670103-20449+4229@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980901164826.13257P-100000@cb1-gw.cb1.com>

it's only from nt wins servers to nt wins servers, and when you run
WINSMGR.EXE, too.

On Tue, 1 Sep 1998, Andrew Tridgell wrote:

> > 2) \PIPE\winsrv (whatever) - a query is possible to obtain all workgroups,
> > i suspect (not confirmed)
> 
> who uses this and when? Is this something done to a WINS server only?
>  
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Tue Sep  1 15:50:51 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/nmbd
In-Reply-To: <19980831225350Z12669506-25139+4353@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980901164903.13257Q-100000@cb1-gw.cb1.com>

On Tue, 1 Sep 1998, Andrew Tridgell wrote:

> > this has the side-effect of delaying the propagation / removal of old
> > information, surely?
> 
> yes, that's the downside. I thought about it and decided it was better
> to have old workgroups than not have some real workgroups.
> 
> I don't think we can avoid old workgroups staying around. The lack of
> a ttl in the NetServerEnum reply makes it basically impossible.

uh... but the NetBIOS name itself has a ttl?
  
> > also, exactly what are you propagating?  the only information that needs
> > to be propagated across multiple workgroups is the list of workgroups
> > themselves, not the machines _in_ those workgroups.
> 
> sure, that's why I added a new parameter which tells the list code
> whether to ask for workgroup names only.

? eh?  i no unnerstand.

> It only asks for machine names when querying its own workgroup.
 
good.

> > therefore, the samba server need not obtain a complete list of machines in
> > each workgroup, it need only obtain a list of workgroups from each DMB it
> > knows about (*<1b>).
> 
> yep, that's what I did.

good-oh.  just checking :)

ok, is there an option to _stop_ samba automatically collating workgroups,
and to work from a pre-selected list _only_?

From lkcl at switchboard.net  Tue Sep  1 15:52:59 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/client
In-Reply-To: <3.0.5.32.19980901082840.0086d6d0@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980901165132.13257S-100000@cb1-gw.cb1.com>

On Tue, 1 Sep 1998, Richard Sharpe wrote:

> At 08:51 AM 9/1/98 +1000, you wrote:
> >> clientutil.c should not be used, it should be deleted as soon as possible.
> >
> >client.c still uses it (unfortunately). We still haven't resolved what
> >to do about the client code ...
> >
> 
> OK, please modify client as you suggested, and then I will look at
> integrating the patches I have and make the mods I want to.

can we code-freeze client until this is done?  also, prior to this being
done, a patch of the main branch, dating back to when BRANCH_NTDOM was
created, needs to be made, and then applied to the BRANCH_NTDOM client
code.

From davecb at Canada.Sun.COM  Tue Sep  1 16:13:35 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:08 2003
Subject: samba-2.0.0-alpha1 portability issues
Message-ID: <35EC1D2F.D00908DF@canada.sun.com>

A quick scan over the sources with score (a solaris-
hosted portability checker) found some 
doubtfull points (eg, WEXITSTATUS ifdef'd in one
place, not in others)

  Anyone interested in the data? It's in
compiler error-message format for jove/emacs/whatever,
and took me about half an hour to scan through just looking 
at the putative bugs: I didn't presume to fix anything 
myself (:-))

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From tridge at samba.anu.edu.au  Wed Sep  2 00:10:22 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/nmbd
In-Reply-To: <Pine.NEB.3.96.980901164903.13257Q-100000@cb1-gw.cb1.com>
	(message from Luke Kenneth Casson Leighton on Tue, 1 Sep 1998 16:50:51
	+0100 (BST))
References: <Pine.NEB.3.96.980901164903.13257Q-100000@cb1-gw.cb1.com>
Message-ID: <19980902001027Z12669833-7510+4553@samba.anu.edu.au>

> uh... but the NetBIOS name itself has a ttl?

yes, but what if we aren't the WINS server? We don't know the ttl. Or
what if the workgroup is part of a group of hosts not using WINS?

From tridge at samba.anu.edu.au  Wed Sep  2 02:30:20 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:08 2003
Subject: samba-2.0.0-alpha1 portability issues
In-Reply-To: <35EC1D2F.D00908DF@canada.sun.com> (message from David
	Collier-Brown on Wed, 2 Sep 1998 02:14:48 +1000)
References: <35EC1D2F.D00908DF@canada.sun.com>
Message-ID: <19980902023022Z12669008-7510+4566@samba.anu.edu.au>

>   Anyone interested in the data? It's in
> compiler error-message format for jove/emacs/whatever,
> and took me about half an hour to scan through just looking 
> at the putative bugs: I didn't presume to fix anything 
> myself (:-))

I'd be interested in seeing the output, although I bet most of the
stuff it produces is spurious.

for example, the WEXITSTATUS problem you mentioned isn't really a
problem. It is used without an ifdef in chpasswd.c which only gets
compiled on a limited range of systems due to its dependence on lots
of fancy tty handling stuff. I doubt a system without WEXITSTATUS
would have the other things necessary for our autoconf system to
decide that chgpsswd.c should be included.
From tridge at samba.anu.edu.au  Wed Sep  2 02:30:20 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/client
In-Reply-To: <Pine.NEB.3.96.980901165132.13257S-100000@cb1-gw.cb1.com>
	(message from Luke Kenneth Casson Leighton on Wed, 2 Sep 1998 02:02:20
	+1000)
References: <Pine.NEB.3.96.980901165132.13257S-100000@cb1-gw.cb1.com>
Message-ID: <19980902023022Z12669015-25139+4638@samba.anu.edu.au>

> can we code-freeze client until this is done?  also, prior to this being
> done, a patch of the main branch, dating back to when BRANCH_NTDOM was
> created, needs to be made, and then applied to the BRANCH_NTDOM client
> code.

and for my next trick I'll solve world hunger ...

seriously, the differences are way too large to just apply a patch. I
suspect that with some regret we are going to have to dump the ntdom
client code and instead fix the head branch code, then slowly re-add
the features that you added in the ntdom branch.

I also don't know when this will be done. Probably Jeremy or I need to
do it, but we are stretched rather thin at the moment.

From mhw at wittsend.com  Wed Sep  2 03:33:26 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:08 2003
Subject: smbmount et al...
Message-ID: <199809020333.XAA25190@alcove.wittsend.com>

Ok all...

	I hadn't tried to compile the whole package since the autoconf
convertion.  I finally got around to it and ran into a tensey problem.
Maybe I'm missing something obvious, but I can't figure out how to get
the smbmount / smbumount programs to configure and compile under autoconf.
What's the magic incantation or did something get missed in the conversion?

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From tridge at samba.anu.edu.au  Wed Sep  2 03:50:21 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:08 2003
Subject: smbmount et al...
In-Reply-To: <199809020333.XAA25190@alcove.wittsend.com> (mhw@wittsend.com)
References: <199809020333.XAA25190@alcove.wittsend.com>
Message-ID: <19980902035029Z12669168-7009+4496@samba.anu.edu.au>

> What's the magic incantation or did something get missed in the conversion?

it got missed :)

We're still trying to work out if we want to include smbmount in Samba
2.0. The problem with including it is that we don't have anyone who
can look after it (handle bug reports etc) and Jeremy and I don't use
it at all. It might have been a mistake to include it in Samba in the
first place as it is so Linux specific.
From davecb at Canada.Sun.COM  Wed Sep  2 11:32:36 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:08 2003
Subject: samba-2.0.0-alpha1 portability issues
References: <35EC1D2F.D00908DF@canada.sun.com> <19980902023022Z12669008-7510+4566@samba.anu.edu.au>
Message-ID: <35ED2CD4.CA313A52@canada.sun.com>

Andrew Tridgell wrote:
> 
> >   Anyone interested in the data? It's in
> > compiler error-message format for jove/emacs/whatever,
> > and took me about half an hour to scan through just looking
> > at the putative bugs: I didn't presume to fix anything
> > myself (:-))
> 
> I'd be interested in seeing the output, although I bet most of the
> stuff it produces is spurious.[...] the WEXITSTATUS problem [..]
> is used without an ifdef in chpasswd.c which only gets
> compiled on a limited range of systems due to its dependence on lots
> of fancy tty handling stuff.

Yes, as I said,  it's reporting doubtful points, and isn't nearly
smart enough to deal with conditional compilation.

I'll enclose a snippit and mail the full thing of to Andrew.
It's used in medium-scale ports, in concert with m4 programs to
fix some of the more straightforward problems: more information
at http://nce.sun.ca/CompMigr/Tools/index.html#suncmigr21


--dave
-- 

./smbd/chgpasswd.c:336 WEXITSTATUS
Weight: 5.0
Comment obsolete

./client/smbmnt.c:276 addmntent
Weight: 5.0
Comment 
 This routine is no longer supported. The putmntent ( ) 
 routine provides similar functionality. 

./client/smbumount.c:166 addmntent
./client/client.c:2497 asctime
Weight: 5.0
Comment 
 The tm structure no longer contains the fields tm_zone and 
 tm_gmtoff. An external variable , timezone , is now used to indicate 
 the difference ( in seconds ) between GMT and local 
 standard time. Another external variable , daylight , indicates 
 if daylight savings should be applied.An external variable , 
 tzname is now used to store standard and summer timezone names. 
 TZ now contains the time conversion information itself. The format 
 of this information is different than previously. 
--
It's used in medium-scale ports, in concert with m4 programs to
address some of the
From borsenkow.msk at sni.de  Wed Sep  2 16:33:34 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
Message-ID: <002001bdd68f$6dda4390$21c9ca95@ao13.mow.sni.de>

I beg your pardon if it was already discussed, but I have not find anything in list archive. 

Now, when SAMBA supports NT domain (verified :), there appear some problems with user's identification. It applies, when SAMBA is in "domain" security.

Assume, that samba server S is member of NT domain D1. Domain D1 has trusted relationship with domain D2. User U\D1 (that is, use U on domain D1) attempts to access server S. Suppose, that Unix user U exists.

As it now appears, SAMBA will (by default at least) map user U\D1 to user U on S. It may or may not be what is wanted, but it is acceptable in many cases. Both servers are in the same domain, and there are good chances, that both users are the same.

Now user U\D2 (that is, user U from domain D2) attempts to acces SAMBA. SAMBA will forward user's credential to domain controller of domain D1 wich will *accept* them (trust between two domains). It means, that U\D2 will end up mapped to U on S which is most probably totally wrong!!! It is totally different user from totally different domain.

Note, that NT model has distinct user spaces for local NT system and for every NT domain (actually, local system is treated as separate domain). Local user 'bor' is totally different from domain user 'bor' which in turn has nothing to do with user 'bor' from any other domain (exactly the case we have here :))

The "correct" model as I see it is:

 - by default every Unix user is treated as local (in above sense); no external user is mapped
   to local user by default

 - there should be configurable way to map NT user to local user based on User+Domain.
   Much better case is to use RID's (is it correct) to uniquly identify NT users. Probabaly, some
   utility to "add trusted user"? Access to shares should use the same database (so, that
   I could say, that any domain, only some domain(s), only group(s) in my domain or only
   specific user(s) may access share).

 - very nice is some way to do "reverse mapping". Assuming, that user 'bor' is declared the same
   as DOMAIN\bor, and tries to access server in DOMAIN with e.g. smbclient, there is no
   need to provide password (well, if we agree to trust Unix authorisation). Of course, 
   we need some sort of database where this information is stored.

And by the way, every member of goup "Domain Admins" can administer any NT server in domain. Does it apply to SAMBA as well? 

thank you for your time

/Andrej

From davecb at Canada.Sun.COM  Wed Sep  2 16:35:47 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:08 2003
Subject: samba-2.0.0alpha1
References: <19980831055022Z12637417-25139+4228@samba.anu.edu.au>
Message-ID: <35ED73E3.C9AED8F1@canada.sun.com>

Andrew Tridgell wrote: 
> I've just released samba-2.0.0alpha1. 

  And it passes the Solaris Application Certification tool
with flying colours.  As did 1.x, of course...



        **************************************************************
                             Sun ABI Program                     
                   Software Product Certification Tool           
                      (c) 1997 Sun Microsystems Inc.             

                    Interface Use Certification Report           

                    Appcert version: 1.2.2      
                    Generated by: SUNWabict 2.7.6 symcertreport
1.25         
                    Wed Sep  2 12:26:51 EDT 1998         

        **************************************************************

        Product Examined: Samba 2.0.0 alpha 1
        Platform Version Certified for: Solaris2.6

        Summary result: PASS

A Total of 12 binary objects were examined

Following (12 of 12) components Passed Unconditionally:

    ./make_printerdef
    ./nmblookup
    ./make_smbcodepage
    ./smbpasswd
    ./smbstatus
    ./smbrun
    ./testprns
    ./testparm
    ./smbclient
    ./swat
    ./nmbd
    ./smbd

  All this really means is that Sun owes you a T-shirt: see
http://sun-www.ebay.sun.com/developers/comarketing/solbrand/app-cert.download.html

  The test is for use of private, denigrated or obsolete
library interfaces.

--dave (who used to work on the project and still
	does regression tests occasionally) c-b

-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From jallison at cthulhu.engr.sgi.com  Wed Sep  2 16:58:03 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
References: <002001bdd68f$6dda4390$21c9ca95@ao13.mow.sni.de>
Message-ID: <35ED791B.71668627@engr.sgi.com>

Andrej Borsenkow wrote:

> Assume, that samba server S is member of NT domain D1. Domain D1 has trusted relationship with domain D2. User U\D1 (that is, use U on domain D1) attempts to access server S. Suppose, that Unix user U exists.
> 
> As it now appears, SAMBA will (by default at least) map user U\D1 to user U on S. It may or may not be what is wanted, but it is acceptable in many cases. Both servers are in the same domain, and there are good chances, that both users are the same.
> 
> Now user U\D2 (that is, user U from domain D2) attempts to acces SAMBA. SAMBA will forward user's credential to domain controller of domain D1 wich will *accept* them (trust between two domains). It means, that U\D2 will end up mapped to U on S which is most probably totally wrong!!! It is totally different user from totally different domain.
> 

Ok - but remember the UNIX machine has no concept of the
difference between user D1\U and D2\U - it only uses the
string 'U' to determine access.

> Note, that NT model has distinct user spaces for local NT system and for every NT domain (actually, local system is treated as separate domain). Local user 'bor' is totally different from domain user 'bor' which in turn has nothing to do with user 'bor' from any other domain (exactly the case we have here :))
> 
> The "correct" model as I see it is:
> 
>  - by default every Unix user is treated as local (in above sense); no external user is mapped
>    to local user by default
> 
>  - there should be configurable way to map NT user to local user based on User+Domain.
>    Much better case is to use RID's (is it correct) to uniquly identify NT users.

Well what you want here is Samba in 'applience' mode -
where no local UNIX users are needed. This is planned
for but not written yet.

Another option would be to mangle any non-local domain
user names in a predictable way, so they don't map 
identially to unix users. Would this work for you ?

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Wed Sep  2 20:13:59 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:08 2003
Subject: gcc format warnings and 64 bits...
Message-ID: <35EDA707.BBD3BB67@engr.sgi.com>

Hmmm. I've run into an anoying problem with gcc
when adding the 64 bit file offset code into Samba.

We currently turn on format warnings for Debug1
and friends using the 

__attribute__ ((format (printf, 1, 2)))

definition for gcc compiles. However, when
printing a 64 bit long long size, you need
to use the format "%lld" - which gcc complains
bitterly about.

Any good ideas on how to keep the format warnings
but allow printing of 64 bit ints ?

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Wed Sep  2 22:43:53 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:08 2003
Subject: gcc format warnings and 64 bits...
Message-ID: <35EDCA29.FD208B59@engr.sgi.com>

Answering my own questions here (is there anybody out there :-):

> We currently turn on format warnings for Debug1
> and friends using the 
> 
> __attribute__ ((format (printf, 1, 2)))
> 
> definition for gcc compiles. However, when
> printing a 64 bit long long size, you need
> to use the format "%lld" - which gcc complains
> bitterly about.

You need to turn off the -pedantic-errors option
to gcc - from the man page :

-pedantic-errors
    Issue all the warnings demanded by strict ANSI standard C; reject all
    programs that use forbidden extensions.

    Valid  ANSI  standard  C  programs  should compile properly with or without
    this option (though a rare few will require `-ansi').  However, without this
    option,  certain  GNU extensions  and  traditional  C features are supported
    as well.  With this option, they are rejected.  There is no reason to  use
    this  option;  it  exists  only  to  satisfy pedants.

Well I guess that tells me where I sit :-).

Yours pedanticly,

	Jeremy.


-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From tridge at samba.anu.edu.au  Thu Sep  3 00:00:19 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:08 2003
Subject: gcc format warnings and 64 bits...
In-Reply-To: <35EDA707.BBD3BB67@engr.sgi.com> (message from Jeremy Allison on
	Thu, 3 Sep 1998 06:15:27 +1000)
References: <35EDA707.BBD3BB67@engr.sgi.com>
Message-ID: <19980903000032Z12669365-7009+4714@samba.anu.edu.au>

> definition for gcc compiles. However, when
> printing a 64 bit long long size, you need
> to use the format "%lld" - which gcc complains
> bitterly about.

You can't use %lld, it is only supported by a small range of
compilers.

I solved this in rsync by using %.0f and casting to double. 

eg:

	rprintf(FINFO,"wrote %.0f bytes  read %.0f bytes  %.2f bytes/sec\n",
	       (double)stats.total_written,
	       (double)stats.total_read,
	       (stats.total_written+stats.total_read)/(0.5 + (t-starttime)));

From crh at NTS.Umn.EDU  Thu Sep  3 04:47:26 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:08 2003
Subject: OTHER: I want to write a front end (PR#9501) (fwd)
Message-ID: <199809030447.XAA05210@unet.unet.umn.edu>

Cory,

A visual representation of the browse list would be a nice thing to have, 
but I'm not sure what you would do with such a list unless you had an SMB 
file system implementation, eg. SMBFS.  Perhaps your product should be an 
add-on to SMBFS instead of Samba.

Chris -)-----

> > A friend and I are interested in writing an X based front end for Samba
> > browsing, much like the Win95 Network Neighborhood.  Before we start,
> > however, I thought I would write and find out if you Samba people know of
> > anybody else that might be doing the same thing.  We are still in the
> > thought-ware stage, but are planning to start coding in the next day or
> > two (once we decide what widget set to use, how to set up the user
> > interface, etc.  Any suggestions by the people who wrote Samba are
> > welcome).  So, if anybody else is writing the same sort of program, we
> > would love to see what they are up to.
> > 
> > Thank you very much.
> > 
> > Cory Lueninghoener
> > ------------------
> > cluenin1@bigred.unl.edu        http://harold.unl.edu/~cluening
> > cluening@cse.unl.edu           http://pest.ml.org
> > cluening@hotmail.com
> 

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From lkcl at switchboard.net  Thu Sep  3 11:54:31 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/nmbd
In-Reply-To: <19980902001027Z12669833-7510+4553@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980903125340.4476C-100000@cb1-gw.cb1.com>

this is solved by nt only allowing trusted domains and those domains
manually added to the browser service to be propagated.

On Wed, 2 Sep 1998, Andrew Tridgell wrote:

> > uh... but the NetBIOS name itself has a ttl?
> 
> yes, but what if we aren't the WINS server? We don't know the ttl. Or
> what if the workgroup is part of a group of hosts not using WINS?
> 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Thu Sep  3 12:07:14 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: CVS update: samba/source/client
In-Reply-To: <19980902023022Z12669015-25139+4638@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980903130436.4476G-100000@cb1-gw.cb1.com>

> and for my next trick I'll solve world hunger ...

i thought samba had already done that.
 
> client code and instead fix the head branch code, then slowly re-add
> the features that you added in the ntdom branch.

once the interface code is redone, should be pretty simple.  jeremy did
half the work needed, to do "security = domain".
 
> I also don't know when this will be done. Probably Jeremy or I need to
> do it, but we are stretched rather thin at the moment.

if i really need it when i start work, i'll do it.  *reluctant*.

From lkcl at switchboard.net  Thu Sep  3 12:10:34 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: FW: PAM-GINA released
Message-ID: <Pine.NEB.3.96.980903130955.4476H-100000@cb1-gw.cb1.com>

-----Original Message-----
From: Naomaru Itoi [mailto:itoi@eecs.umich.edu] 
Sent: Tuesday, September 01, 1998 7:48 AM
To: ntgina@umich.edu; kerberos@MIT.EDU; hans@isk.kth.se;
x7currie@lab2.cc.wmich.edu; cummings@math.utk.edu;
william@iosoftware.com; bse@me.chalmers.se; tran@dstc.qut.edu.au;
kcjones@us.itd.umich.edu; kan@egr.msu.edu; phillipjacobs@email.msn.com;
dhd@umich.edu; richardw@microsoft.com; sgr@umich.edu;
philippe.defert@cern.ch; nath@isr.umd.edu
Cc: honey@citi.umich.edu
Subject: PAM-GINA released


Hello,

First, I am sorry if you are not interested in this e-mail.  I am
sending this to whoever asked me or talked with me about PAM-GINA.

I would like to announce the release of PAM-GINA version 0.02.
I solved some copyright problems and now it is freely distributed (you
can use it, modify it, redistribute it, etc.)  The GINA can perform
single sign-on with following authentication mechanisms:  
- Kerberos 4 (http://web.mit.edu/is/help/mink/)
- Kerberos 5 (http://web.mit.edu/kerberos/www/nt-alpha2.html)
- Novell Netware 4.0.

Binary and source code are in:
http://www-personal.engin.umich.edu/~itoi/ni_pam/

The paper presented in USENIX Windows/NT symposium is in:
http://www-personal.engin.umich.edu/~itoi/

Please send me comments and questions.

Thanks!

--
Naomaru Itoi <itoi@eecs.umich.edu>
Center for Information Technology Integration, the University of Michigan
http://www-personal.engin.umich.edu/~itoi/


From lkcl at switchboard.net  Thu Sep  3 12:41:00 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.OSF.3.96.980902083319.15564A-102000@barney.spc.cc.tx.us>
Message-ID: <Pine.NEB.3.96.980903133913.4476L-100000@cb1-gw.cb1.com>

On Wed, 2 Sep 1998, Tim Winders wrote:

> Luke -
> 
> Here are two gzipped capture files printed out with the all levels as

tim,

the captures show that we are replying to the "Bind Ack" with the
"fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
two hoots about this (and, to be honest, i don't think we do either, in
the bind ack response).

however, it looks like someone's being giving the nt dce/rpc code the "nit
comb" treatment...

luke
From lkcl at switchboard.net  Thu Sep  3 12:46:56 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
In-Reply-To: <002001bdd68f$6dda4390$21c9ca95@ao13.mow.sni.de>
Message-ID: <Pine.NEB.3.96.980903134311.4476Q-100000@cb1-gw.cb1.com>

andrej,

you have hit on exactly the right problem.  unless the unix system you are
using supports the concept of "domains", namely that every process, file
and other object has a "SID" attached to it (max 28 bytes or so) instead
of a 32 bit uid, then you cannot support multiple domains.

in other words, you can't, unless you make the entire unix system running
samba a "black box", and internally you treat the 32 bit uid as a vector
table to look up a SID.

jeremy knows what i'm talking about :-)

On Thu, 3 Sep 1998, Andrej Borsenkow wrote:

> I beg your pardon if it was already discussed, but I have not find anything in list archive. 
> 
> Now, when SAMBA supports NT domain (verified :), there appear some problems with user's identification. It applies, when SAMBA is in "domain" security.
> 
> Assume, that samba server S is member of NT domain D1. Domain D1 has trusted relationship with domain D2. User U\D1 (that is, use U on domain D1) attempts to access server S. Suppose, that Unix user U exists.
> 
> As it now appears, SAMBA will (by default at least) map user U\D1 to user U on S. It may or may not be what is wanted, but it is acceptable in many cases. Both servers are in the same domain, and there are good chances, that both users are the same.
> 
> Now user U\D2 (that is, user U from domain D2) attempts to acces SAMBA. SAMBA will forward user's credential to domain controller of domain D1 wich will *accept* them (trust between two domains). It means, that U\D2 will end up mapped to U on S which is most probably totally wrong!!! It is totally different user from totally different domain.
> 
> Note, that NT model has distinct user spaces for local NT system and for every NT domain (actually, local system is treated as separate domain). Local user 'bor' is totally different from domain user 'bor' which in turn has nothing to do with user 'bor' from any other domain (exactly the case we have here :))
> 
> The "correct" model as I see it is:
> 
>  - by default every Unix user is treated as local (in above sense); no external user is mapped
>    to local user by default
> 
>  - there should be configurable way to map NT user to local user based on User+Domain.
>    Much better case is to use RID's (is it correct) to uniquly identify NT users. Probabaly, some
>    utility to "add trusted user"? Access to shares should use the same database (so, that
>    I could say, that any domain, only some domain(s), only group(s) in my domain or only
>    specific user(s) may access share).
> 
>  - very nice is some way to do "reverse mapping". Assuming, that user 'bor' is declared the same
>    as DOMAIN\bor, and tries to access server in DOMAIN with e.g. smbclient, there is no
>    need to provide password (well, if we agree to trust Unix authorisation). Of course, 
>    we need some sort of database where this information is stored.
> 
> And by the way, every member of goup "Domain Admins" can administer any NT server in domain. Does it apply to SAMBA as well? 
> 
> thank you for your time
> 
> /Andrej
> 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From jgeiser at csillc.com  Thu Sep  3 12:56:45 1998
From: jgeiser at csillc.com (Joe Geiser)
Date: Tue Dec  2 03:02:08 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.NEB.3.96.980903133913.4476L-100000@cb1-gw.cb1.com>
Message-ID: <000301bdd73a$4e289780$03000004@jgeiser.csillc.com>

Luke after Tim,

> > Here are two gzipped capture files printed out with the all levels as
>
> tim,
>
> the captures show that we are replying to the "Bind Ack" with the
> "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> two hoots about this (and, to be honest, i don't think we do either, in
> the bind ack response).
>
> however, it looks like someone's being giving the nt dce/rpc code the "nit
> comb" treatment...

This is exactly what's happening with SP4.  Since 5.0 is so delayed, quite a
bit of the nit-combing is being released in SPs for 4.0.  Expect to see more
of it as well.

Joe

From lkcl at switchboard.net  Thu Sep  3 12:48:39 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
In-Reply-To: <35ED791B.71668627@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980903134758.4476R-100000@cb1-gw.cb1.com>

> Another option would be to mangle any non-local domain
> user names in a predictable way, so they don't map 
> identially to unix users. Would this work for you ?

it would be good to map users to "\\DOMAIN\user" however most unixen don't
suppport > 8 chars for usernames.

hey, a new autoconf test!

From twinders at SPC.cc.tx.us  Thu Sep  3 12:56:46 1998
From: twinders at SPC.cc.tx.us (Tim Winders)
Date: Tue Dec  2 03:02:08 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.NEB.3.96.980903133913.4476L-100000@cb1-gw.cb1.com>
Message-ID: <Pine.OSF.3.96.980903075559.2969A-100000@barney.spc.cc.tx.us>

On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:

> On Wed, 2 Sep 1998, Tim Winders wrote:
> 
> > Luke -
> > 
> > Here are two gzipped capture files printed out with the all levels as
> 
> tim,
> 
> the captures show that we are replying to the "Bind Ack" with the
> "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> two hoots about this (and, to be honest, i don't think we do either, in
> the bind ack response).
> 
> however, it looks like someone's being giving the nt dce/rpc code the "nit
> comb" treatment...

Hurrah!  Obviously a step by MS to try to stop people from using Samba!
Now, is this something "easy" to fix?

=== Tim

---------------------------------------------------------------------
|  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
|  Network Administrator         |  Phone:  806-894-9611 x 2369     |
|  South Plains College          |  Fax:    806-897-4711            |
---------------------------------------------------------------------


From marc at reston.ans.net  Thu Sep  3 13:08:57 1998
From: marc at reston.ans.net (Marc Sherman)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
In-Reply-To: <Pine.NEB.3.96.980903134311.4476Q-100000@cb1-gw.cb1.com>
References: <002001bdd68f$6dda4390$21c9ca95@ao13.mow.sni.de>
Message-ID: <199809031313.JAA26646@interlock.reston.ans.net>

At 10:54 PM 9/3/98 +1000, you wrote:
>andrej,
>
>you have hit on exactly the right problem.  unless the unix system you are
>using supports the concept of "domains", namely that every process, file
>and other object has a "SID" attached to it (max 28 bytes or so)

Luke, not to quibble, but I calculate 40 bytes max for a "SID", does that
sound way off? Here's how I figure 40:

revision number = 1 byte
rid count = 1 byte
identifier authority = 6 bytes
8 max possible sub authorities = 32 bytes

..Marc

> instead
>of a 32 bit uid, then you cannot support multiple domains.
>
>in other words, you can't, unless you make the entire unix system running
>samba a "black box", and internally you treat the 32 bit uid as a vector
>table to look up a SID.
>
>jeremy knows what i'm talking about :-)
>
>On Thu, 3 Sep 1998, Andrej Borsenkow wrote:
>
>> I beg your pardon if it was already discussed, but I have not find
anything in list archive. 
>> 
>> Now, when SAMBA supports NT domain (verified :), there appear some
problems with user's identification. It applies, when SAMBA is in "domain"
security.
>> 
>> Assume, that samba server S is member of NT domain D1. Domain D1 has
trusted relationship with domain D2. User U\D1 (that is, use U on domain
D1) attempts to access server S. Suppose, that Unix user U exists.
>> 
>> As it now appears, SAMBA will (by default at least) map user U\D1 to
user U on S. It may or may not be what is wanted, but it is acceptable in
many cases. Both servers are in the same domain, and there are good
chances, that both users are the same.
>> 
>> Now user U\D2 (that is, user U from domain D2) attempts to acces SAMBA.
SAMBA will forward user's credential to domain controller of domain D1 wich
will *accept* them (trust between two domains). It means, that U\D2 will
end up mapped to U on S which is most probably totally wrong!!! It is
totally different user from totally different domain.
>> 
>> Note, that NT model has distinct user spaces for local NT system and for
every NT domain (actually, local system is treated as separate domain).
Local user 'bor' is totally different from domain user 'bor' which in turn
has nothing to do with user 'bor' from any other domain (exactly the case
we have here :))
>> 
>> The "correct" model as I see it is:
>> 
>>  - by default every Unix user is treated as local (in above sense); no
external user is mapped
>>    to local user by default
>> 
>>  - there should be configurable way to map NT user to local user based
on User+Domain.
>>    Much better case is to use RID's (is it correct) to uniquly identify
NT users. Probabaly, some
>>    utility to "add trusted user"? Access to shares should use the same
database (so, that
>>    I could say, that any domain, only some domain(s), only group(s) in
my domain or only
>>    specific user(s) may access share).
>> 
>>  - very nice is some way to do "reverse mapping". Assuming, that user
'bor' is declared the same
>>    as DOMAIN\bor, and tries to access server in DOMAIN with e.g.
smbclient, there is no
>>    need to provide password (well, if we agree to trust Unix
authorisation). Of course, 
>>    we need some sort of database where this information is stored.
>> 
>> And by the way, every member of goup "Domain Admins" can administer any
NT server in domain. Does it apply to SAMBA as well? 
>> 
>> thank you for your time
>> 
>> /Andrej
>> 
>> 
>
><a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
><a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
><a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>
>
>

From lkcl at switchboard.net  Thu Sep  3 13:09:13 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:08 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.OSF.3.96.980903075559.2969A-100000@barney.spc.cc.tx.us>
Message-ID: <Pine.NEB.3.96.980903140214.4476a-100000@cb1-gw.cb1.com>

> > the captures show that we are replying to the "Bind Ack" with the
> > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care

> Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> Now, is this something "easy" to fix?

yep.  the frag_len member variable of the RPC_HDR needs to have 16 added
to it in the "BBA" response.  can't track it down, right now...

From twinders at SPC.cc.tx.us  Thu Sep  3 13:27:25 1998
From: twinders at SPC.cc.tx.us (Tim Winders)
Date: Tue Dec  2 03:02:08 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.NEB.3.96.980903140214.4476a-100000@cb1-gw.cb1.com>
Message-ID: <Pine.OSF.3.96.980903082707.1691A-100000@barney.spc.cc.tx.us>

On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:

> > > the captures show that we are replying to the "Bind Ack" with the
> > > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> 
> > Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> > Now, is this something "easy" to fix?
> 
> yep.  the frag_len member variable of the RPC_HDR needs to have 16 added
> to it in the "BBA" response.  can't track it down, right now...

Good deal!  Thanks for your help on this, Luke!

=== Tim

---------------------------------------------------------------------
|  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
|  Network Administrator         |  Phone:  806-894-9611 x 2369     |
|  South Plains College          |  Fax:    806-897-4711            |
---------------------------------------------------------------------


From davecb at Canada.Sun.COM  Thu Sep  3 14:46:42 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
References: <Pine.NEB.3.96.980903134311.4476Q-100000@cb1-gw.cb1.com>
Message-ID: <35EEABD2.FFDF8850@canada.sun.com>

Luke Kenneth Casson Leighton wrote:
> 
> andrej,
> 
> you have hit on exactly the right problem.  unless the unix system you are
> using supports the concept of "domains", namely that every process, file
> and other object has a "SID" attached to it (max 28 bytes or so) instead
> of a 32 bit uid, then you cannot support multiple domains.

	And it doesn't even map into the old
	``orange book'' (military) concepts of secuity 
	categories and levels within them...

	So you can't easily map them to anything standard,
	nor interpret them as subsets of something bigger.
	
	Hmmn.. and that doesn't **seem** to match up with
	Kerberos, either.   I wonder if it's going to change
	in NT 5?

--dave
--
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From Dirk.DeWachter at rug.ac.be  Thu Sep  3 17:12:20 1998
From: Dirk.DeWachter at rug.ac.be (Dirk De Wachter)
Date: Tue Dec  2 03:02:08 2003
Subject: OTHER: I want to write a front end (PR#9501) (fwd)
In-Reply-To: <199809030447.XAA05210@unet.unet.umn.edu>
Message-ID: <199809031513.RAA07345@navier.rug.ac.be>

> Cory,
> 
> A visual representation of the browse list would be a nice thing to
> have, but I'm not sure what you would do with such a list unless you
> had an SMB file system implementation, eg. SMBFS.  Perhaps your
> product should be an add-on to SMBFS instead of Samba.
> 
> Chris -)-----
> 
Dear Chris & Cory,

If a command can be linked to this graphical list, one may e.g.
fire up smbclient to the specific share or mount it with Sharity.
After all, it is samba that contains the tools to query the browse 
list (and as far as I know, not SMBFS).

Kind regards,

Dirk De Wachter

--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--
  Dirk De Wachter, MScEE, MScBME, PhD       http://allserv.rug.ac.be/~ddwachte
  scientific researcher, systems administrator mailto:Dirk.DeWachter@rug.ac.be
  Hydraulics Laboratory, Ibitech, University Gent         voice:+32 9 264 3281
  St.-Pietersnieuwstraat 41, B-9000 Gent Belgium          faxto:+32 9 264 3595
--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--
From davecb at Canada.Sun.COM  Thu Sep  3 15:38:54 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:08 2003
Subject: Security model in samba-2
References: <Pine.NEB.3.96.980903134311.4476Q-100000@cb1-gw.cb1.com> <35EEABD2.FFDF8850@canada.sun.com>
Message-ID: <35EEB80E.34253E97@canada.sun.com>

David Collier-Brown wrote:
>         And it doesn't even map into the old
>         ``orange book'' (military) concepts of secuity
>         categories and levels within them...

	I should comment that Samba's security model
	``fits inside'' an orange book categoey and level.

	A colleague and I described samba security as three-level,
	but not in the ``secret" -vs- "top secret" sense
	of levels: more like
		1) host level -- what machine can connect
		2) Service-level -- per-share limitations
		3) user-level -- limitations on particular users.
	all sitting on top of (inside of) normal Unix 
	security.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From william at hae.com  Thu Sep  3 16:33:34 1998
From: william at hae.com (William Stuart)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.OSF.3.96.980903075559.2969A-100000@barney.spc.cc.tx.us>
Message-ID: <Pine.BSF.4.02A.9809030927390.24456-100000@shell3.ba.best.com>

Tim--

In regards to your comment about Microsoft and SP4 breaking SAMBA, do you
think it true or were you just spouting?

If it is true, the DOJ might be interested in hearing it.  They are adding
the breaking of other, competing software systems to their allegations,
but their latest incident of this was DR DOS,  5 or 6 years ago.

Just a thought.

---
William Stuart  (william@hae.com)
"Don't rush me sonny.  You rush a miracle man you get rotten miracles."
                     --Miracle Max, "The Princess Bride"

On Thu, 3 Sep 1998, Tim Winders wrote:

> Date: Thu, 3 Sep 1998 23:01:48 +1000
> From: Tim Winders <twinders@SPC.cc.tx.us>
> To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> Subject: Re: NT4 sp4, anyone working on it?
> 
> On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:
> 
> > On Wed, 2 Sep 1998, Tim Winders wrote:
> > 
> > > Luke -
> > > 
> > > Here are two gzipped capture files printed out with the all levels as
> > 
> > tim,
> > 
> > the captures show that we are replying to the "Bind Ack" with the
> > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> > two hoots about this (and, to be honest, i don't think we do either, in
> > the bind ack response).
> > 
> > however, it looks like someone's being giving the nt dce/rpc code the "nit
> > comb" treatment...
> 
> Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> Now, is this something "easy" to fix?
> 
> === Tim
> 
> ---------------------------------------------------------------------
> |  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
> |  Network Administrator         |  Phone:  806-894-9611 x 2369     |
> |  South Plains College          |  Fax:    806-897-4711            |
> ---------------------------------------------------------------------
> 
> 
> 

From twinders at SPC.cc.tx.us  Thu Sep  3 17:03:53 1998
From: twinders at SPC.cc.tx.us (Tim Winders)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.BSF.4.02A.9809030927390.24456-100000@shell3.ba.best.com>
Message-ID: <Pine.OSF.3.96.980903120218.6350A-100000@barney.spc.cc.tx.us>

You *NEVER* know with Microsoft and I doubt we could *PROVE* it.  I really
was just spouting, but I believe MS intentially breaks 3rd party software
and claims it was for other reasons.  Take the Win98 upgrade and DLL files
as an example...

On Thu, 3 Sep 1998, William Stuart wrote:

> In regards to your comment about Microsoft and SP4 breaking SAMBA, do you
> think it true or were you just spouting?
> 
> If it is true, the DOJ might be interested in hearing it.  They are adding
> the breaking of other, competing software systems to their allegations,
> but their latest incident of this was DR DOS,  5 or 6 years ago.
> 
> Just a thought.


> On Thu, 3 Sep 1998, Tim Winders wrote:
> 
> > Date: Thu, 3 Sep 1998 23:01:48 +1000
> > From: Tim Winders <twinders@SPC.cc.tx.us>
> > To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> > Subject: Re: NT4 sp4, anyone working on it?
> > 
> > On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:
> > 
> > > On Wed, 2 Sep 1998, Tim Winders wrote:
> > > 
> > > > Luke -
> > > > 
> > > > Here are two gzipped capture files printed out with the all levels as
> > > 
> > > tim,
> > > 
> > > the captures show that we are replying to the "Bind Ack" with the
> > > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> > > two hoots about this (and, to be honest, i don't think we do either, in
> > > the bind ack response).
> > > 
> > > however, it looks like someone's being giving the nt dce/rpc code the "nit
> > > comb" treatment...
> > 
> > Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> > Now, is this something "easy" to fix?
> > 
> > === Tim
> > 
> > ---------------------------------------------------------------------
> > |  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
> > |  Network Administrator         |  Phone:  806-894-9611 x 2369     |
> > |  South Plains College          |  Fax:    806-897-4711            |
> > ---------------------------------------------------------------------
> > 
> > 
> > 
> 
> 

=== Tim

---------------------------------------------------------------------
|  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
|  Network Administrator         |  Phone:  806-894-9611 x 2369     |
|  South Plains College          |  Fax:    806-897-4711            |
---------------------------------------------------------------------


From davecb at Canada.Sun.COM  Thu Sep  3 18:26:03 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
References: <Pine.BSF.4.02A.9809030927390.24456-100000@shell3.ba.best.com>
Message-ID: <35EEDF3B.A4579355@canada.sun.com>

William Stuart wrote:
> In regards to your comment about Microsoft and SP4 breaking SAMBA,
...
> If it is true, the DOJ might be interested in hearing it.  They are adding
> the breaking of other, competing software systems to their allegations,
> but their latest incident of this was DR DOS,  5 or 6 years ago.

	All to easy to do this sort of thing while
	fixing bugs....

--dave (who very much doesn't like Microsoft) c-b
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From Jean-Francois.Micouleau at utc.fr  Fri Sep  4 10:40:05 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:09 2003
Subject: Share names in uppercase
Message-ID: <Pine.OSF.4.02A.9809041228100.6849-100000@kappa.utc.fr>


in rpc_server/srv_srvsvc.c: make_srv_share_1_info there is a
strupper(lp_servicename()) making all the share names uppercased.

I checked against an NT server which doesn't do it. So why ???

There is also a cosmetic thing in swat. In the advanced view of the share,
the printers options are available.

	Jean Francois

-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------

From lkcl at switchboard.net  Fri Sep  4 15:57:54 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: OTHER: I want to write a front end (PR#9501) (fwd)
In-Reply-To: <199809031513.RAA07345@navier.rug.ac.be>
Message-ID: <Pine.NEB.3.96.980904165728.9047G-100000@cb1-gw.cb1.com>

smb2www, a perl script wrapper around smbclient.  available under the gpl.

On Fri, 4 Sep 1998, Dirk De Wachter wrote:

> > Cory,
> > 
> > A visual representation of the browse list would be a nice thing to
> > have, but I'm not sure what you would do with such a list unless you
> > had an SMB file system implementation, eg. SMBFS.  Perhaps your
> > product should be an add-on to SMBFS instead of Samba.
> > 
> > Chris -)-----
> > 
> Dear Chris & Cory,
> 
> If a command can be linked to this graphical list, one may e.g.
> fire up smbclient to the specific share or mount it with Sharity.
> After all, it is samba that contains the tools to query the browse 
> list (and as far as I know, not SMBFS).
> 
> Kind regards,
> 
> Dirk De Wachter
> 
> --*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--
>   Dirk De Wachter, MScEE, MScBME, PhD       http://allserv.rug.ac.be/~ddwachte
>   scientific researcher, systems administrator mailto:Dirk.DeWachter@rug.ac.be
>   Hydraulics Laboratory, Ibitech, University Gent         voice:+32 9 264 3281
>   St.-Pietersnieuwstraat 41, B-9000 Gent Belgium          faxto:+32 9 264 3595
> --~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Fri Sep  4 15:59:40 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Security model in samba-2
In-Reply-To: <35EEB80E.34253E97@canada.sun.com>
Message-ID: <Pine.NEB.3.96.980904165817.9047H-100000@cb1-gw.cb1.com>

> 	I should comment that Samba's security model
> 	``fits inside'' an orange book categoey and level.
> 
> 	A colleague and I described samba security as three-level,
> 	but not in the ``secret" -vs- "top secret" sense
> 	of levels: more like
> 		1) host level -- what machine can connect
> 		2) Service-level -- per-share limitations
> 		3) user-level -- limitations on particular users.
> 	all sitting on top of (inside of) normal Unix 
> 	security.

because you can include files inline using any %sub macro, you can do
groups, caller's name, calling name etc etc.  lots.

From jallison at cthulhu.engr.sgi.com  Fri Sep  4 16:05:24 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:09 2003
Subject: Share names in uppercase
References: <Pine.OSF.4.02A.9809041228100.6849-100000@kappa.utc.fr>
Message-ID: <35F00FC4.3896B421@engr.sgi.com>

Jean-Francois Micouleau wrote:
> 
> in rpc_server/srv_srvsvc.c: make_srv_share_1_info there is a
> strupper(lp_servicename()) making all the share names uppercased.
> 
> I checked against an NT server which doesn't do it. So why ???
> 
> There is also a cosmetic thing in swat. In the advanced view of the share,
> the printers options are available.

Thanks for spotting that - it had irritated me for a while
but I always thought it was an artifact of doing share
names over RPC with non-unicode.

I'm glad you pointed out that *we* were doing it :-).

It's *gone* (from the master sources :-).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From lkcl at switchboard.net  Fri Sep  4 16:14:59 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.BSF.4.02A.9809030927390.24456-100000@shell3.ba.best.com>
Message-ID: <Pine.NEB.3.96.980904171224.9047L-100000@cb1-gw.cb1.com>

william, tim.

microsoft have made their code "more robust".  i can only presume that
they were ignoring a vital length field (the fragment length in the bind
ack response) in _pre_ NT4 SP4.  i suspect that there is therefore a
buffer overflow security hole there, somewhere...

On Fri, 4 Sep 1998, William Stuart wrote:

> Tim--
> 
> In regards to your comment about Microsoft and SP4 breaking SAMBA, do you
> think it true or were you just spouting?
> 
> If it is true, the DOJ might be interested in hearing it.  They are adding
> the breaking of other, competing software systems to their allegations,
> but their latest incident of this was DR DOS,  5 or 6 years ago.
> 
> Just a thought.
> 
> ---
> William Stuart  (william@hae.com)
> "Don't rush me sonny.  You rush a miracle man you get rotten miracles."
>                      --Miracle Max, "The Princess Bride"
> 
> On Thu, 3 Sep 1998, Tim Winders wrote:
> 
> > Date: Thu, 3 Sep 1998 23:01:48 +1000
> > From: Tim Winders <twinders@SPC.cc.tx.us>
> > To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> > Subject: Re: NT4 sp4, anyone working on it?
> > 
> > On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:
> > 
> > > On Wed, 2 Sep 1998, Tim Winders wrote:
> > > 
> > > > Luke -
> > > > 
> > > > Here are two gzipped capture files printed out with the all levels as
> > > 
> > > tim,
> > > 
> > > the captures show that we are replying to the "Bind Ack" with the
> > > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> > > two hoots about this (and, to be honest, i don't think we do either, in
> > > the bind ack response).
> > > 
> > > however, it looks like someone's being giving the nt dce/rpc code the "nit
> > > comb" treatment...
> > 
> > Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> > Now, is this something "easy" to fix?
> > 
> > === Tim
> > 
> > ---------------------------------------------------------------------
> > |  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
> > |  Network Administrator         |  Phone:  806-894-9611 x 2369     |
> > |  South Plains College          |  Fax:    806-897-4711            |
> > ---------------------------------------------------------------------
> > 
> > 
> > 
> 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Fri Sep  4 16:33:14 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Reg Developiing a file system in solaris
In-Reply-To: <9809041508.AA13467@tagore.wipinfo.soft.net>
Message-ID: <Pine.NEB.3.96.980904173009.9047R-100000@cb1-gw.cb1.com>

dorairaj,

there is someone i can't recall who (i haven't spoken to him for a few
months), but they are on the samba digest lists, who is writing a "VFS" -
virtual file system for solaris.  his aim is to write a redirector. 
however, as he is at a university and has access to source, he is under
some sort of license by sun not to tell all... 

if you know who you are, can you please contact dorairaj?  thank you!


On Fri, 4 Sep 1998, Dorairaj V wrote:

> Hi ,
>   I want to develop  a CIFS file system on Solaris .
>  Is anyone aware of any documentation available to develop a file
> system in Solaris .
> Any reply in this regard is highly appreciated
> 
> Cheeers
> 
> Dorai
> 
> 
> 
> 
> 
> 
> 
> 
> *************************************************
>          OFFICE              *         RESIDENCE
>  *************************************************
>   Dorairaj V                 * Dorairaj V
>   (Senior Engineer -Software *  273-A Basavanpura Road
>   CEC-STAR Group , NDF  )    *  Behind Police Qtrs
>  Wipro Global R& D           *  Krishnarajapuram
>  30 Mission Road             *  Bangalore -36
>  1'st main S.R.Nagar         *
>  Bangalore -25               *  Ph : 91-80-8518552
> *************************************************************
> Voice : +91+80  2241730/28/33/35/68  ,+91+80 2244867/68/69/70   extn 3404
> Cyber :dorai@wipinfo.soft.net
> ************************************************************
> ~
> 
> ----------------------------------------------------------------
> Users Guide http://www.microsoft.com/workshop/essentials/mail.asp
> contains important info including how to unsubscribe.  Save time, search
> the archives at http://discuss.microsoft.com/archives/index.html
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>


From lkcl at switchboard.net  Fri Sep  4 16:40:02 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Share names in uppercase
In-Reply-To: <Pine.OSF.4.02A.9809041228100.6849-100000@kappa.utc.fr>
Message-ID: <Pine.NEB.3.96.980904173549.9047T-100000@cb1-gw.cb1.com>

On Fri, 4 Sep 1998, Jean-Francois Micouleau wrote:

> 
> in rpc_server/srv_srvsvc.c: make_srv_share_1_info there is a
> strupper(lp_servicename()) making all the share names uppercased.
> 
> I checked against an NT server which doesn't do it. So why ???

my mistake.


From davecb at Canada.Sun.COM  Fri Sep  4 16:49:22 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:09 2003
Subject: Security model in samba-2
References: <Pine.NEB.3.96.980904165817.9047H-100000@cb1-gw.cb1.com>
Message-ID: <35F01A12.38D00002@canada.sun.com>

I wrote:
> >       A colleague and I described samba security as three-level,
> >       but not in the ``secret" -vs- "top secret" sense
> >       of levels: more like
> >               1) host level -- what machine can connect
> >               2) Service-level -- per-share limitations
> >               3) user-level -- limitations on particular users.
> >       all sitting on top of (inside of) normal Unix
> >       security.
> 
Luke Kenneth Casson Leighton wrote:
> because you can include files inline using any %sub macro, you can do
> groups, caller's name, calling name etc etc.  lots.

	Quite! You can expand the per-service
	level into any combination of restrictions
	that can be expressed in  %-macros.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From twinders at SPC.cc.tx.us  Fri Sep  4 17:50:37 1998
From: twinders at SPC.cc.tx.us (Tim Winders)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.NEB.3.96.980904171224.9047L-100000@cb1-gw.cb1.com>
Message-ID: <Pine.OSF.3.96.980904124930.15500B-100000@barney.spc.cc.tx.us>

Thanks Luke.  I was contacted by a reporter from ZD News today questioning
me on this information!  She asked me to keep her up to date on the
"developments".

=== Tim

On Sat, 5 Sep 1998, Luke Kenneth Casson Leighton wrote:

> william, tim.
> 
> microsoft have made their code "more robust".  i can only presume that
> they were ignoring a vital length field (the fragment length in the bind
> ack response) in _pre_ NT4 SP4.  i suspect that there is therefore a
> buffer overflow security hole there, somewhere...
> 
> On Fri, 4 Sep 1998, William Stuart wrote:
> 
> > Tim--
> > 
> > In regards to your comment about Microsoft and SP4 breaking SAMBA, do you
> > think it true or were you just spouting?
> > 
> > If it is true, the DOJ might be interested in hearing it.  They are adding
> > the breaking of other, competing software systems to their allegations,
> > but their latest incident of this was DR DOS,  5 or 6 years ago.
> > 
> > Just a thought.
> > 
> > ---
> > William Stuart  (william@hae.com)
> > "Don't rush me sonny.  You rush a miracle man you get rotten miracles."
> >                      --Miracle Max, "The Princess Bride"
> > 
> > On Thu, 3 Sep 1998, Tim Winders wrote:
> > 
> > > Date: Thu, 3 Sep 1998 23:01:48 +1000
> > > From: Tim Winders <twinders@SPC.cc.tx.us>
> > > To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> > > Subject: Re: NT4 sp4, anyone working on it?
> > > 
> > > On Thu, 3 Sep 1998, Luke Kenneth Casson Leighton wrote:
> > > 
> > > > On Wed, 2 Sep 1998, Tim Winders wrote:
> > > > 
> > > > > Luke -
> > > > > 
> > > > > Here are two gzipped capture files printed out with the all levels as
> > > > 
> > > > tim,
> > > > 
> > > > the captures show that we are replying to the "Bind Ack" with the
> > > > "fragmentation length" 16 bytes short.  obviously, pre-ntsp4 didn't care
> > > > two hoots about this (and, to be honest, i don't think we do either, in
> > > > the bind ack response).
> > > > 
> > > > however, it looks like someone's being giving the nt dce/rpc code the "nit
> > > > comb" treatment...
> > > 
> > > Hurrah!  Obviously a step by MS to try to stop people from using Samba!
> > > Now, is this something "easy" to fix?
> > > 
> > > === Tim
> > > 
> > > ---------------------------------------------------------------------
> > > |  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
> > > |  Network Administrator         |  Phone:  806-894-9611 x 2369     |
> > > |  South Plains College          |  Fax:    806-897-4711            |
> > > ---------------------------------------------------------------------
> > > 
> > > 
> > > 
> > 
> > 
> 
> <a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
> <a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
> <a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>
> 
> 

=== Tim

---------------------------------------------------------------------
|  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
|  Network Administrator         |  Phone:  806-894-9611 x 2369     |
|  South Plains College          |  Fax:    806-897-4711            |
---------------------------------------------------------------------


From jallison at cthulhu.engr.sgi.com  Fri Sep  4 19:03:43 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
References: <Pine.OSF.3.96.980904124930.15500B-100000@barney.spc.cc.tx.us>
Message-ID: <35F0398F.AAD5CD0B@engr.sgi.com>

Tim Winders wrote:
> 
> Thanks Luke.  I was contacted by a reporter from ZD News today questioning
> me on this information!  She asked me to keep her up to date on the
> "developments".
> 
> === Tim
> 

Tim - have you tried SP4 with Samba 1.9.18p10 ?
Does that still work ok ? I was presuming it
was just our alpha domain code that was causing problems
not the stable version.

We don't have SP4 here yet so I can't test it - as
soon as we do I'll fix whatever problems I find.

Cheers,

	Jeremy.

PS. I'm also talking to ZD News to clarify the problem.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From william at hae.com  Fri Sep  4 19:29:11 1998
From: william at hae.com (William Stuart)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.OSF.3.96.980904124930.15500B-100000@barney.spc.cc.tx.us>
Message-ID: <Pine.BSF.4.02A.9809041227460.16632-100000@shell3.ba.best.com>

Looks like we have a few reporters on the list.

Hmmm.  Better watch what we say.

---
William Stuart  (william@hae.com)
"Don't rush me sonny.  You rush a miracle man you get rotten miracles."
                     --Miracle Max, "The Princess Bride"

On Sat, 5 Sep 1998, Tim Winders wrote:

> Date: Sat, 5 Sep 1998 03:52:56 +1000
> From: Tim Winders <twinders@SPC.cc.tx.us>
> To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> Subject: Re: NT4 sp4, anyone working on it?
> 
> Thanks Luke.  I was contacted by a reporter from ZD News today questioning
> me on this information!  She asked me to keep her up to date on the
> "developments".
> 
> === Tim
> 

From sambaman at norse.myths.com  Sat Sep  5 02:25:33 1998
From: sambaman at norse.myths.com (Red Hat Linux User)
Date: Tue Dec  2 03:02:09 2003
Subject: subscribe
Message-ID: <199809050225.WAA30614@norse.myths.com>

subscribe
From twinders at SPC.cc.tx.us  Sat Sep  5 03:34:59 1998
From: twinders at SPC.cc.tx.us (Tim Winders)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.BSF.4.02A.9809041227460.16632-100000@shell3.ba.best.com>
Message-ID: <Pine.OSF.3.96.980904223424.9818B-100000@barney.spc.cc.tx.us>

It took me by *BIG* surprise when I got a PHONE CALL today.  Wow, I was
dumb-fuddled...

On Fri, 4 Sep 1998, William Stuart wrote:

> Looks like we have a few reporters on the list.
> 
> Hmmm.  Better watch what we say.
> 
> ---
> William Stuart  (william@hae.com)
> "Don't rush me sonny.  You rush a miracle man you get rotten miracles."
>                      --Miracle Max, "The Princess Bride"
> 
> On Sat, 5 Sep 1998, Tim Winders wrote:
> 
> > Date: Sat, 5 Sep 1998 03:52:56 +1000
> > From: Tim Winders <twinders@SPC.cc.tx.us>
> > To: Multiple recipients of list <samba-technical@samba.anu.edu.au>
> > Subject: Re: NT4 sp4, anyone working on it?
> > 
> > Thanks Luke.  I was contacted by a reporter from ZD News today questioning
> > me on this information!  She asked me to keep her up to date on the
> > "developments".
> > 
> > === Tim
> > 
> 
> 

=== Tim

---------------------------------------------------------------------
|  Tim Winders, CNE, MCSE        |  Email:  TWinders@SPC.cc.tx.us   |
|  Network Administrator         |  Phone:  806-894-9611 x 2369     |
|  South Plains College          |  Fax:    806-897-4711            |
---------------------------------------------------------------------


From tridge at samba.anu.edu.au  Sun Sep  6 12:49:32 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:09 2003
Subject: icons problem 
Message-ID: <19980906124936Z12606479-25139+5750@samba.anu.edu.au>

A number of people have reported that application icons don't appear
for files on Samba servers when running with security=server. This has
been a problem for a long time, but as we haven't been able to
reproduce it we haven't been able to fix it.

This is just to say that I've now reproduced the problem, and I've
caught a sniff of it with tcpdump. I haven't worked out what is wrong
yet, but hopefully I will soon.

Reproducing that has made me very happy - this problem has annoyed me
for a long time!
From tridge at samba.anu.edu.au  Sun Sep  6 13:49:34 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:09 2003
Subject: icons problem - timing problem?
Message-ID: <19980906134947Z12606814-20449+5658@samba.anu.edu.au>

I've narrowed it down a bit. It seems to be a timing problem. A Win95
client will not attempt to display icons if the server is "slow". The
critical thing appear to be how long the server takes to respond to
the negprot and session setup requests.

When using security=server Samba responds slowly to these requests
because it needs to forward them to the password server. this is
exacerbated by Smaba sending a deliberately incorrect initial password
to the password server (to test for the security bug in NT where it
grants all session requests, regardless of passwords). NT responds
very slowly to an incorrect password (maybe a deliberate feature to
try to prevent password cracking?).

I don't have a good solution yet, but at least I think I know what is
going on. I've also managed to reproduce the "no icons" problem with
user level security by adding some sleep() calls in the negprot and
session setup code.

From lkcl at switchboard.net  Sun Sep  6 16:06:18 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Share names in uppercase
In-Reply-To: <35F00FC4.3896B421@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980906170538.13193A-100000@cb1-gw.cb1.com>

naah - i just messed it up by uppercasing the share name instead of the
workstation name.

On Sat, 5 Sep 1998, Jeremy Allison wrote:

> Thanks for spotting that - it had irritated me for a while
> but I always thought it was an artifact of doing share
> names over RPC with non-unicode.


From lkcl at switchboard.net  Sun Sep  6 19:07:38 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.BSF.4.02A.9809041227460.16632-100000@shell3.ba.best.com>
Message-ID: <Pine.NEB.3.96.980906200504.13425D-100000@cb1-gw.cb1.com>

On Sat, 5 Sep 1998, William Stuart wrote:

> Looks like we have a few reporters on the list.
> 
> Hmmm.  Better watch what we say.

nuts!  hazel nuts!  pigs and flying saucers!  (my friend reading this is
telling me to stop sniffing glue...)

i know what this list is for: technical issues related to the development
of samba.  there are a number of reasons why i discuss everything i do,
publicly.  one of these is so that other people can choose either to keep
informed, or to help out.

luke

From borsenkow.msk at sni.de  Mon Sep  7 15:01:27 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:09 2003
Subject: Security model in samba-2
In-Reply-To: <Pine.NEB.3.96.980904165817.9047H-100000@cb1-gw.cb1.com>
Message-ID: <000901bdda70$6383db60$21c9ca95@ao13.mow.sni.de>



> -----Original Message-----
> From: samba-technical@samba.anu.edu.au
> [mailto:samba-technical@samba.anu.edu.au]On Behalf Of Luke Kenneth
> Casson Leighton
> Sent: Friday, September 04, 1998 8:05 PM
> To: Multiple recipients of list
> Subject: Re: Security model in samba-2
> 
> 
> > 	I should comment that Samba's security model
> > 	``fits inside'' an orange book categoey and level.
> > 
> > 	A colleague and I described samba security as three-level,
> > 	but not in the ``secret" -vs- "top secret" sense
> > 	of levels: more like
> > 		1) host level -- what machine can connect
> > 		2) Service-level -- per-share limitations
> > 		3) user-level -- limitations on particular users.
> > 	all sitting on top of (inside of) normal Unix 
> > 	security.
> 
> because you can include files inline using any %sub macro, you can do
> groups, caller's name, calling name etc etc.  lots.
> 
> 


Unfortunately, I cannot access NT domain information now (or am I wrong?) There is no way to say:

	valid domains = ...
	valid domain groups = DOM1\gr1, DOM2\gr2 ...
	valid domain users = DOM1\user1 ...

or like.

I repeat: in case of SAMBA particicpating in NT domain, the user name *only* is not enough. It simply does not provide enough authentication in case of trust relationships between domains. It is possible to emulate it, but it means, SAMBA server has to have complete knowledge of all possible domains/users/groups/hosts in any other domain.

/Andrej

From Dirk.DeWachter at rug.ac.be  Mon Sep  7 18:12:09 1998
From: Dirk.DeWachter at rug.ac.be (Dirk De Wachter)
Date: Tue Dec  2 03:02:09 2003
Subject: icons problem 
In-Reply-To: <19980906124936Z12606479-25139+5750@samba.anu.edu.au>
Message-ID: <199809071612.SAA17410@navier.rug.ac.be>

> A number of people have reported that application icons don't appear
> for files on Samba servers when running with security=server. This
> has been a problem for a long time, but as we haven't been able to
> reproduce it we haven't been able to fix it.
> 
> This is just to say that I've now reproduced the problem, and I've
> caught a sniff of it with tcpdump. I haven't worked out what is
> wrong yet, but hopefully I will soon.
> 
> Reproducing that has made me very happy - this problem has annoyed
> me for a long time!
> 
I always thought this had to do with some timeouts. If the server 
would respond a little bit faster (and really I have no idea wat 
actions are involved to get the icons) the icons would appear on the 
screen. There is however a difference between NT and Win95 (or is 
this just because my NT machines are more powerfull in terms of 
networking?)
Anyway, Microsoft gives away the main Office 97 icons to be installed 
locally. If MS Office builds its screenbar, it has to query the icons 
of the different applications. If the server tends to respond too 
slow, the icons cannot be shown. One can solve this to use locally 
installed icons for the shortcuts. This works perfectly (and also for 
the Start-icons).
Perhaps this "slow" response is the same thing that we see when 
connecting to a Samba server to get NT-profiles?

Kind regards,

Dirk

--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--*--
  Dirk De Wachter, MScEE, MScBME, PhD       http://allserv.rug.ac.be/~ddwachte
  scientific researcher, systems administrator mailto:Dirk.DeWachter@rug.ac.be
  Hydraulics Laboratory, Ibitech, University Gent         voice:+32 9 264 3281
  St.-Pietersnieuwstraat 41, B-9000 Gent Belgium          faxto:+32 9 264 3595
--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--~--
From lkcl at switchboard.net  Mon Sep  7 16:56:51 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: NT4 sp4, anyone working on it?
In-Reply-To: <Pine.OSF.3.96.980906143654.30674B-100000@barney.spc.cc.tx.us>
Message-ID: <Pine.NEB.3.96.980907175552.15494B-100000@cb1-gw.cb1.com>

ok, can people check out smbd/ipc.c and try it?  i've added 0x10 to the
fragment length of the header on the "bind acknowledgment" response.

From lkcl at switchboard.net  Mon Sep  7 17:10:16 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Security model in samba-2
In-Reply-To: <000901bdda70$6383db60$21c9ca95@ao13.mow.sni.de>
Message-ID: <Pine.NEB.3.96.980907180517.15494G-100000@cb1-gw.cb1.com>

On Mon, 7 Sep 1998, Andrej Borsenkow wrote:

> Unfortunately, I cannot access NT domain information now (or am I wrong?) There is no way to say:
> 
> 	valid domains = ...
> 	valid domain groups = DOM1\gr1, DOM2\gr2 ...
> 	valid domain users = DOM1\user1 ...
> 
> or like.

the only work that has been done, so far, is to make samba either a
stand-alone PDC, or to make samba join a single domain.  no inter-domain
trust relationships; no BDCs.
 
> I repeat: in case of SAMBA particicpating in NT domain, the user name *only* is not enough. It simply does not provide enough authentication in case of trust relationships between domains. It is possible to emulate it, but it means, SAMBA server has to have complete knowledge of all possible domains/users/groups/hosts in any other domain.

[andrej, can you please break up your lines to a max of 70 chars per
line?  when "including" one of your messages, an entire paragraph comes
out as a single line, as demonstrated above.  thanks!]

so, at the moment, the NT "domain" model just doesn't apply.  across all
trusted domains, the user names must be unique.

hm.  what about modifying the "map username =" so that it takes the
format:
\\DOMAIN\ntusername unixusername

luke

From lkcl at switchboard.net  Mon Sep  7 17:10:57 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: icons problem 
In-Reply-To: <199809071612.SAA17410@navier.rug.ac.be>
Message-ID: <Pine.NEB.3.96.980907181045.15494H-100000@cb1-gw.cb1.com>

> Perhaps this "slow" response is the same thing that we see when 
> connecting to a Samba server to get NT-profiles?

yes.

From tridge at samba.anu.edu.au  Mon Sep  7 22:29:00 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:09 2003
Subject: icons problem
In-Reply-To: <199809071612.SAA17410@navier.rug.ac.be>
	(Dirk.DeWachter@rug.ac.be)
References: <199809071612.SAA17410@navier.rug.ac.be>
Message-ID: <19980907222911Z12669357-7510+5972@samba.anu.edu.au>

> Anyway, Microsoft gives away the main Office 97 icons to be installed 
> locally.

yes, I saw that in the knowledgebase. The particular case I was
looking at was general .exe files though, not just the office ones.

> Perhaps this "slow" response is the same thing that we see when 
> connecting to a Samba server to get NT-profiles?

that's very possible. If I get time I might try to reproduce the "slow
server" profiles message on NT by adding some sleep() calls, and try
to quantify the timing threshold.

From Jean-Francois.Micouleau at utc.fr  Tue Sep  8 20:40:38 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
Message-ID: <Pine.OSF.4.02A.9809082224180.31943-100000@kappa.utc.fr>



Is there a way to disable/enable the auto-(re)loading of smb.conf while an
smbd is running ? Something like

disable_auto_reload() and
enable_auto_reload()

I need that in the spoolss code to prevent a risk of clash while a printer
is open.

	Jean Francois

-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------


From cartegw at Eng.Auburn.EDU  Tue Sep  8 22:24:25 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
Message-ID: <35F5AE99.CC1CBE7A@eng.auburn.edu>

Greetings again...

Not sure what to think about this one...

I'm running the 2.0 prealpha code downloaded via CVS on August 26
running on Solaris 2.5.1(sparc)

Here's the log message

------------------------------------------------------
[1998/09/08 16:30:24, 0]
nmbd/nmbd_become_dmb.c:become_domain_master_query_fail(360)
  become_domain_master_query_fail: Error 0 returned when querying 
  WINS server for name ENG-NT<1b>.
------------------------------------------------------

Also seeing

------------------------------------------------------
find_response_record: response packet id 6316 received with no matching
record.
find_subnet_for_nmb_packet: response record not found for response id
6316
------------------------------------------------------

packet id <6316> appears to be the request that is timing out.


I've done some debugging and it seems to be timeing out when trying to 
register the DOMAINNAME<1b> record with the 1.9.18p4 WINS server.

The catch is that it will only fail to register this name if the WINS
server is on the same subnet.  If the WINS erver is on another subnet,
everything goes as planned. Does this sound familar to anyone?  Not 
sure I quite understand the behavior here.

I gone through the code and looked at backtraces of the call stack 
during execution and can't seem to get why it's doing this.  

Could someone enlightne me a little?  I can provide as much or 
as little information as you want.  I'll keep looking in the 
meantime.



Thanks again,
j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From tridge at samba.anu.edu.au  Wed Sep  9 02:28:32 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <Pine.OSF.4.02A.9809082224180.31943-100000@kappa.utc.fr> (message
	from Jean-Francois Micouleau on Wed, 9 Sep 1998 06:45:32 +1000)
References: <Pine.OSF.4.02A.9809082224180.31943-100000@kappa.utc.fr>
Message-ID: <19980909022843Z12669336-20449+6168@samba.anu.edu.au>

> Is there a way to disable/enable the auto-(re)loading of smb.conf while an
> smbd is running ? Something like
> 
> disable_auto_reload() and
> enable_auto_reload()

there isn't a way currently.
 
> I need that in the spoolss code to prevent a risk of clash while a printer
> is open.

can you explain this a little more? What sort of clash do you mean?

I've actually been tempted to change smbd and nmbd to only reload on
HUP and when a new connection is established (ie. when the daemon
forks). What do other people think about this? I originally put in the
auto-reload code because I thought it would be a convenience but I
think it may actually be counter-productive. It certainly doesn't
follow the principle of least surprise - most Unix daemons only reload
on HUP.


From sambaman at norse.myths.com  Wed Sep  9 02:56:39 1998
From: sambaman at norse.myths.com (David P. Murphy)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
Message-ID: <199809090256.WAA00333@norse.myths.com>

>I've actually been tempted to change smbd and nmbd to only reload on
>HUP and when a new connection is established (ie. when the daemon
>forks). What do other people think about this? I originally put in the
>auto-reload code because I thought it would be a convenience but I
>think it may actually be counter-productive. It certainly doesn't
>follow the principle of least surprise - most Unix daemons only reload
>on HUP.

which is exactly what happened to me the first time i modified the
config file . . . almost did a 'killall -HUP smbd' out of habit,
but read the man page first "just to confirm" and was quite
surprised to not find the capability.

i would certainly agree that reloading only upon SIGHUP would be
the Right Thing to do.

ok
dpm
From Jean-Francois.Micouleau at utc.fr  Wed Sep  9 09:20:30 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <19980909022843Z12669336-20449+6168@samba.anu.edu.au>
Message-ID: <Pine.OSF.4.02A.9809091110070.27769-100000@kappa.utc.fr>

On Wed, 9 Sep 1998, Andrew Tridgell wrote:

> can you explain this a little more? What sort of clash do you mean?

In a struct I store the handle (20 bytes) and the corresponding
lp_servicename(). I need to do so as all the spoolss functions are
referencing an handle.

Now imagine, a user open a printer (to check the spooled jobs for
instance) and the samba admin decides to delete this share, and the
smb.conf file is reloaded.

As I'm lazy, I don't want to add code everywhere to check if the share
name stored in the handle struct still exist, I prefer to disable
auto-reloading of smb.conf until all printers are open by a given smbd
process.


> I've actually been tempted to change smbd and nmbd to only reload on
> HUP and when a new connection is established (ie. when the daemon
> forks). What do other people think about this?

Yes. It's more Unixlly correct. Jeremy will like it :-)

> I originally put in the
> auto-reload code because I thought it would be a convenience but I
> think it may actually be counter-productive. It certainly doesn't
> follow the principle of least surprise - most Unix daemons only reload
> on HUP.

	J.F.

-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------

From tridge at samba.anu.edu.au  Wed Sep  9 10:08:26 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <Pine.OSF.4.02A.9809091110070.27769-100000@kappa.utc.fr> (message
	from Jean-Francois Micouleau on Wed, 9 Sep 1998 11:20:30 +0200 (MET
	DST))
References: <Pine.OSF.4.02A.9809091110070.27769-100000@kappa.utc.fr>
Message-ID: <19980909100830Z12669313-7009+6238@samba.anu.edu.au>

> In a struct I store the handle (20 bytes) and the corresponding
> lp_servicename(). I need to do so as all the spoolss functions are
> referencing an handle.

ok
 
> Now imagine, a user open a printer (to check the spooled jobs for
> instance) and the samba admin decides to delete this share, and the
> smb.conf file is reloaded.

rare, but possible
 
> As I'm lazy, I don't want to add code everywhere to check if the share
> name stored in the handle struct still exist, I prefer to disable
> auto-reloading of smb.conf until all printers are open by a given smbd
> process.

no, don't do that. What if the client doesn't open all printers? Then
you'd leave the smb.conf loading permanently disabled.

I think the correct solution is to either add checks to see if the
share name is valid (it can't be that hard!) or just not ignore the
checks and bear with a possible segv if the reload happens when the
share is deleted.

there is also a 3rd alternative. Samba doesn't allow a reload to
delete a share that is currently in use. (see the call to
lp_killunused() in server.c). You could add a "inuse" flag to a share
and make conn_snum_used() check that flag. Then add lp_set_used() and
lp_unset_used() calls.

From davecb at Canada.Sun.COM  Wed Sep  9 12:27:42 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
References: <199809090256.WAA00333@norse.myths.com>
Message-ID: <35F6743E.79111F2E@canada.sun.com>

Tridge quoth:
> >I've actually been tempted to change smbd and nmbd to only reload on
> >HUP and when a new connection is established (ie. when the daemon
> >forks). What do other people think about this? 

	Reloading when a service is created will still
	trigger the (admittedly unusual) case the initial
	commentator described.

	Utterly-least-surprise for a non-inetd daemon is 
	ONLY on hup.  For an inetd deamon, utterly would 
	increase to on daemon-creation and HUP [but see 1 
	below]

	Conversely, right now it's ``whenever the config file
	changes''.  This is surprising, but it's not a bad thing.
	
	I see only two dangers in the whole system:
	1) Different states of different daemons.
	   An inetd-started daemon, unless it polls for
	   updates, could easily end up running from a
	   different config file than the next one started.
	   This argues against dropping the poll.
	2) Unexpected change during production operation.
	   If the daemons all reread changed files, some things
	   could blow up right under users.  This is the
	   general case of the printer problem mentioned.
	   This argues (on first glance) against any updating.

	I'm a former security person: B2 systems react 
	**immediately** to any change in someone's permissions.
	This means the security officer (sysadmin) has to be careful.

	It's not an onerous requirement.  I run production
	and test systems, and test on the test system, where the
	immediate change is A Real Good Thing.  I just have
	(well, had) to be careful when changing the production
	system, because I know the change will happen **now**.

	In summary:  I think the file should be checked **at least**
	on HUP or new connection.  I like the current model,
	as it approaches `whenever the config file changes''.
	They only reason I'd back off from that is performance.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From Jean-Francois.Micouleau at utc.fr  Wed Sep  9 14:26:04 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <19980909100830Z12669313-7009+6238@samba.anu.edu.au>
Message-ID: <Pine.OSF.4.02A.9809091619580.17957-100000@kappa.utc.fr>

On Wed, 9 Sep 1998, Andrew Tridgell wrote:

> > auto-reloading of smb.conf until all printers are open by a given smbd
> > process.

Oops, my fault. I mean: until all opened printers are closed.

> no, don't do that. What if the client doesn't open all printers? Then
> you'd leave the smb.conf loading permanently disabled.
> 
> I think the correct solution is to either add checks to see if the
> share name is valid (it can't be that hard!) or just not ignore the
> checks and bear with a possible segv if the reload happens when the
> share is deleted.
> 
> there is also a 3rd alternative. Samba doesn't allow a reload to
> delete a share that is currently in use. (see the call to
> lp_killunused() in server.c). You could add a "inuse" flag to a share
> and make conn_snum_used() check that flag. Then add lp_set_used() and
> lp_unset_used() calls.

I like that.

Before coding, let's state on the auto-reloading. Do we want to keep it or
move to the kill -HUP unix habit ?

-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------

From green at UMDNJ.EDU  Wed Sep  9 14:34:10 1998
From: green at UMDNJ.EDU (green@UMDNJ.EDU)
Date: Tue Dec  2 03:02:09 2003
Subject: samba-2.0.0-alpha1 portability issues
In-Reply-To: <35EC1D2F.D00908DF@canada.sun.com>
Message-ID: <Pine.HPX.4.02.9809012321160.3051-100000@rwja.UMDNJ.EDU>

This isn't really a response to David's posting, but I'm not sure where
else to report this, so...

Configuring and making this alpha on an HPUX 10.20 box (a 9000/715), I
had a couple of notes:

- the resulting Makefile doesn't have an important flag for the HP ANSI
compiler;  the initial value for CFLAGS is only "-g", but should 
include "-Ae" to enable the ANSI features (this is equivalent to using
"-Aa -D_HPUX_SOURCE").  As I found later, this should be set in the
environment, but anyone else on an HP box should be so warned...

- there was a lone warning during the make:
Compiling lib/netmask.c
cc: "lib/netmask.c", line 317: warning 618: Declaration of "in_addr" not
visible outside this function prototype scope.

Otherwise, everything seems to make okay.  So far.

- when starting, both daemons start and then poop out with similar log
messages, this one for smbd:

[1998/08/31 18:32:31, 1] smbd/server.c:(630)
  smbd version 2.0.0-prealpha started.
  Copyright Andrew Tridgell 1992-1997
[1998/08/31 18:32:32, 0] lib/pidfile.c:(82)
  ERROR: smbd : fcntl lock of file /products/samba/var/locks/smbd.pid
failed. Error was No such file or directory

However, both smbd.pid and nmbd.pid *do* exist.  They are both
zero-sized files, but with different permissions:
-rw-------   1 root       sys              0 Aug 31 17:09 nmbd.pid
-rw-r--r--   1 root       root             0 Aug 31 17:09 smbd.pid

Both the nmbd and smbd errors are for the same line in pidfile.c.

For the last few releases, Samba has be a nobrainer to make and install;
I guess I just got spoiled.  Since neither daemon will run, I can't give
any more diagnostics.  <sigh>

c
-- 
Clifford Green               Internet -  green@umdnj.edu
Academic Computing Services     voice -     732-235-5250
UMDNJ-IST                         fax -     732-235-5252
Avoid loud & aggressive persons, they are vexatious to the spirit.




From cartegw at Eng.Auburn.EDU  Wed Sep  9 14:42:15 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
References: <35F5AE99.CC1CBE7A@eng.auburn.edu>
Message-ID: <35F693C7.A20EE6A4@eng.auburn.edu>

Gerald W. Carter wrote:
> 
> I've done some debugging and it seems to be timeing out when trying to
> register the DOMAINNAME<1b> record with the 1.9.18p4 WINS server.
> 
> The catch is that it will only fail to register this name if the WINS
> server is on the same subnet.  If the WINS erver is on another subnet,
> everything goes as planned. Does this sound familar to anyone?  Not
> sure I quite understand the behavior here.

Some more news.  I think I was wrong in the above statements.  Here 
is a little more information.  These log entries were taken from the
19.18p4 WINS server

-------PDC that registers ENG-NT<1b> ok--------------------------
09/09/1998 08:46:05 received a packet of len 50 from (131.204.31.43)
port 137
nmb packet from aaa.bbb.ccc.ddd(137) header: id=29426 opcode=Query(0)
response=No
    header: flags: bcast=No rec_avail=No rec_des=Yes trunc=No auth=No
    header: rcode=0 qdcount=1 ancount=0 nscount=0 arcount=0
    question: q_name=ENG-NT<1b> q_type=32 q_class=1
 
nmb packet from aaa.bbb.ccc.ddd(137) header: id=29426 opcode=Query(0)
response=Yes
    header: flags: bcast=No rec_avail=Yes rec_des=Yes trunc=No auth=Yes
    header: rcode=3 qdcount=0 ancount=1 nscount=0 arcount=0
    answers: nmb_name=ENG-NT<1b> rr_type=32 rr_class=1 ttl=0
------------------------------------------------------------------

-------PDC that does not register ENG-NT<1b> ok-------------------
nmb packet from aaa.bbb.xxx.yyy(137) header: id=27254 opcode=Query(0)
response=No
    header: flags: bcast=No rec_avail=No rec_des=Yes trunc=No auth=No
    header: rcode=0 qdcount=1 ancount=0 nscount=0 arcount=0
    question: q_name=ENG-NT<1b> q_type=32 q_class=1
 
nmb packet from aaa.bbb.xxx.yyy(137) header: id=27254 opcode=Query(0)
response=Yes
    header: flags: bcast=No rec_avail=Yes rec_des=Yes trunc=No auth=Yes
    header: rcode=3 qdcount=0 ancount=1 nscount=0 arcount=0
    answers: nmb_name=ENG-NT<1b> rr_type=32 rr_class=1 ttl=0
------------------------------------------------------------------

As you can see the rcode = NAM_ERR in both responses ( which is 
correct ).  However he gdb backtrace on the 2.0.0-prealpha
( the one that won't register ) looks like

--------------------------------------------------------------------
Breakpoint 1, become_domain_master_query_fail (subrec=0xbfdf0,
rrec=0xbde90,
    question_name=0xc7644, fail_code=0) at nmbd/nmbd_become_dmb.c:248
248       if((subrec == unicast_subnet) && (fail_code != NAM_ERR))
(gdb) bt
#0  become_domain_master_query_fail (subrec=0xbfdf0, rrec=0xbde90,
question_name=0xc7644,
    fail_code=0) at nmbd/nmbd_become_dmb.c:248
#1  0x28340 in query_name_timeout_response (subrec=0xbfdf0,
rrec=0xbde90)
    at nmbd/nmbd_namequery.c:137
#2  0x2edb8 in retransmit_or_expire_response_records (t=905304387) at
nmbd/nmbd_packets.c:1639
#3  0x1a9c4 in process () at nmbd/nmbd.c:387
#4  0x1bd64 in main (argc=6, argv=0xeffff654) at nmbd/nmbd.c:763
--------------------------------------------------------------------

Here the fail_code has been set to '0'.  I'm still trying to trace 
the timeout which BTW the way occurs when the WINS server is 
lightly loaded as well.

I'm tracing things through on the 2.0.0 code.  The 18p4 WINS server 
code appears to be working correctly.  Does this sound like the right?

I am noticing from the netmon trace that the 2.0.0 server that won't 
register the <1b> record, registers all else by broadcast rather than 
unicast directly to the WINS server.  I have two servers running on 
each of two Sun's.  one server is on a real interface and the other is
on a virtual interface.  That's a total of four servers on on two 
machines ( just to clarify ).

The 2.0.0 PDC is on a virtual interface on one machine and the 18p4
WINS server is on the real interface on another machine.  All this is 
setup using the "interfaces" and "socket address" parameters.

Will keep looking but if anyone has any suggestions for debugging the
current situation, let me know.




Thanks again,
j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From cartegw at Eng.Auburn.EDU  Wed Sep  9 16:02:56 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
References: <35F5AE99.CC1CBE7A@eng.auburn.edu> <35F693C7.A20EE6A4@eng.auburn.edu>
Message-ID: <35F6A6B0.F14BA0F5@eng.auburn.edu>

Gerald W. Carter wrote:
> 
> I am noticing from the netmon trace that the 2.0.0 server that won't
> register the <1b> record, registers all else by broadcast rather than
> unicast directly to the WINS server. 

Hmmm...the things you learn with netmon and gdb :)

I was mistaken.  the problem PDC does not recieve an answer to the
DOMAINNAME<1b> query.  

-----------------------------
696       if(initiate_name_query_packet( p ) == False)
(gdb) print *p
$8 = {next = 0x0, prev = 0x0, locked = 0, ip = {S_un = {S_un_b = {s_b1 =
131 
      '\203', s_b2 = 204 '?', s_b3 = 12 '\f', s_b4 = 43 '+'}, S_un_w = 
     {s_w1 = 33740, s_w2 = 3115}, S_addr = 2211187755}}, port = 137, fd
= 5,
     timestamp = 905356469, packet_type = NMB_PACKET, packet = {nmb = 
     {header = {name_trn_id = 3682, opcode = 0, response = 0, nm_flags = 
     {bcast = 0, recursion_available = 0, recursion_desired = 0, trunc =
0,
     authoritative = 0}, rcode = 0, qdcount = 1, ancount = 0, nscount =
0,
     arcount = 0}, question = {question_name = {
     name = "ENG-NT\000\000\000\000\000\000\000\000\000\000", 
     scope = '\000' <repeats 63 times>, name_type = 27}, question_type =
32, 
     question_class = 1}, answers = 0x0, nsrecs = 0x0, additional =
0x0}, 
     dgram = {header = { msg_type = 3682, flags = {node_type = B_NODE, 
     first = 0, more = 0}, dgm_id = 0, source_ip = {S_un = {S_un_b = 
     {s_b1 = 0 '\000', s_b2 = 0 '\000', s_b3 = 0 '\000', s_b4 = 0
'\000'},
     S_un_w = {s_w1 = 0, s_w2 = 0}, S_addr = 0}}, source_port = 0, 
     dgm_length = 0, packet_offset = 0}, source_name = {
     name = "\000\000\000\001", '\000' <repeats 12 times>, "E", 
     scope = "NG-NT", '\000' <repeats 58 times>, name_type = 0},
dest_name = {
     name = '\000' <repeats 15 times>, "\e", 
     scope = "\000\000 \000\000\000\001", '\000' <repeats 56 times>, 
     name_type = 0}, datasize = 0, data = '\000' <repeats 575 times>}}}
-------------------------------

This is a general question.  This is a queued netbios packet to be sent 
from the problem PDC to the WINS server.  Why is the bcast flag set to 
0 but the node_type = B_NODE?



Thanks again,
j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From lkcl at switchboard.net  Wed Sep  9 16:27:43 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <19980909022843Z12669336-20449+6168@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980909172548.23463E-100000@cb1-gw.cb1.com>

any shares currently open don't get reloaded / modified: no capability
exists to deal with this situation.  i quite like the auto-load thing, but
then again, i know how it works and its limitations.

> I've actually been tempted to change smbd and nmbd to only reload on
> HUP and when a new connection is established (ie. when the daemon
> forks). What do other people think about this? I originally put in the

From jallison at cthulhu.engr.sgi.com  Wed Sep  9 16:33:36 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
References: <Pine.OSF.4.02A.9809082224180.31943-100000@kappa.utc.fr> <19980909022843Z12669336-20449+6168@samba.anu.edu.au>
Message-ID: <35F6ADE0.7D800936@engr.sgi.com>

Andrew Tridgell wrote:

> 
> I've actually been tempted to change smbd and nmbd to only reload on
> HUP and when a new connection is established (ie. when the daemon
> forks). What do other people think about this? I originally put in the
> auto-reload code because I thought it would be a convenience but I
> think it may actually be counter-productive. It certainly doesn't
> follow the principle of least surprise - most Unix daemons only reload
> on HUP.

That would work for me. I've seen people get very confused
when some things are changed when smb.conf is updated and some
things aren't.

But do it *soon* (before Samba-2 gets many more alpha's).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From davecb at Canada.Sun.COM  Wed Sep  9 16:43:16 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
References: <Pine.NEB.3.96.980909172548.23463E-100000@cb1-gw.cb1.com>
Message-ID: <35F6B024.482533A8@canada.sun.com>

Luke Kenneth Casson Leighton wrote:
> 
> any shares currently open don't get reloaded / modified: no capability
> exists to deal with this situation. 

	Good!  That eliminates one of my cases from
	consideration (and you can still put someone
	the hosts deny list and kill his server to
	blow him away)

	 i quite like the auto-load thing, but
> then again, i know how it works and its limitations.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From lkcl at switchboard.net  Wed Sep  9 16:46:25 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
In-Reply-To: <35F6A6B0.F14BA0F5@eng.auburn.edu>
Message-ID: <Pine.NEB.3.96.980909174443.23463K-100000@cb1-gw.cb1.com>

> This is a general question.  This is a queued netbios packet to be sent 
> from the problem PDC to the WINS server.  Why is the bcast flag set to 
> 0 but the node_type = B_NODE?

it's because the query is specifically sent to that machine.  only that
machine is expected to respond.

remember that NetBIOS itself used to be a transport, so you can set "bcast
= True" which is equivalent to sending to x.x.255.255 or x.x.x.255 or
whatever, and "bcast = false" which is equivalent to sending to x.x.x.x

luke

From lkcl at switchboard.net  Wed Sep  9 16:50:10 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: auto-loading of smb.conf
In-Reply-To: <35F6ADE0.7D800936@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980909174746.23463N-100000@cb1-gw.cb1.com>

> That would work for me. I've seen people get very confused
> when some things are changed when smb.conf is updated and some
> things aren't.

wouldn't make any difference... *unless* a kill -HUP is used, say, to
cause the server to *drop* all open shares, forcing the clients to
reconnect.

this would have the effect of making all shares the same when smb.conf is
updated, but is likely to have nasty side-effects.  but SMB is supposed to
be resilient to that kind of thing, innit? :)

From cartegw at Eng.Auburn.EDU  Wed Sep  9 17:09:00 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
References: <Pine.NEB.3.96.980909174443.23463K-100000@cb1-gw.cb1.com>
Message-ID: <35F6B62C.CA0A2225@eng.auburn.edu>

Luke Kenneth Casson Leighton wrote:
> 
> > This is a general question.  This is a queued netbios packet to be sent
> > from the problem PDC to the WINS server.  Why is the bcast flag set to
> > 0 but the node_type = B_NODE?
> 
> it's because the query is specifically sent to that machine.  only that
> machine is expected to respond.
> 
> remember that NetBIOS itself used to be a transport, so you can set "bcast
> = True" which is equivalent to sending to x.x.255.255 or x.x.x.255 or
> whatever, and "bcast = false" which is equivalent to sending to x.x.x.x

Right.  I got that.  Question was directed mot to the B_NODE flag.
If I have set the 'wins server' paramet in smb.conf the samba_nb_type 
should be set to NB_MFLAG arther than NB_BFLAG which is what is showing 
up in the outgoing packet

void set_samba_nb_type(void)
{
  if( lp_wins_support() || (*lp_wins_server()) )
    samba_nb_type = NB_MFLAG;       /* samba is a 'hybrid' node type. */
  else
    samba_nb_type = NB_BFLAG;       /* samba is broadcast-only node
type. */
} /* set_samba_nb_type */



j- 
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From Jean-Francois.Micouleau at utc.fr  Wed Sep  9 18:02:40 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:09 2003
Subject: Autocad R14 bug with Samba
In-Reply-To: <35F6ADE0.7D800936@engr.sgi.com>
Message-ID: <Pine.OSF.4.02A.9809091959240.12084-100000@kappa.utc.fr>


I've got a netmon trace of the autocad R14 bug when you try to open a file
and the reply is "The file is in read-only mode, blah, blah,..."

Anybody interrested ???

Where should I upload it ?

-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------

From lkcl at switchboard.net  Wed Sep  9 18:26:18 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
In-Reply-To: <35F6B62C.CA0A2225@eng.auburn.edu>
Message-ID: <Pine.NEB.3.96.980909192438.24096B-100000@cb1-gw.cb1.com>

> > > This is a general question.  This is a queued netbios packet to be sent
> > > from the problem PDC to the WINS server.  Why is the bcast flag set to
> > > 0 but the node_type = B_NODE?

> 
> Right.  I got that.  Question was directed mot to the B_NODE flag.
> If I have set the 'wins server' paramet in smb.conf the samba_nb_type 
> should be set to NB_MFLAG arther than NB_BFLAG which is what is showing 
> up in the outgoing packet

> void set_samba_nb_type(void)
> {
>   if( lp_wins_support() || (*lp_wins_server()) )
>     samba_nb_type = NB_MFLAG;       /* samba is a 'hybrid' node type. */
>   else
>     samba_nb_type = NB_BFLAG;       /* samba is broadcast-only node
> type. */
> } /* set_samba_nb_type */

names are not just registered as "mixed" or "hybrid": they must also be
registered independently on each netbios-broadcast subnet.  as type "B".

luke

From jallison at cthulhu.engr.sgi.com  Wed Sep  9 18:40:20 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:09 2003
Subject: Autocad R14 bug with Samba
References: <Pine.OSF.4.02A.9809091959240.12084-100000@kappa.utc.fr>
Message-ID: <35F6CB94.45CA3044@engr.sgi.com>

Jean-Francois Micouleau wrote:
> 
> I've got a netmon trace of the autocad R14 bug when you try to open a file
> and the reply is "The file is in read-only mode, blah, blah,..."
> 
> Anybody interrested ???
> 
> Where should I upload it ?
> 

Can you reproduce this at will ? Is the netmon trace
with Samba-2.0 alpha ?

If so please upload it somewher onto samba.anu.edu.au
and let me know where. I've been wanting to kill this
one for *years* !

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From cartegw at Eng.Auburn.EDU  Wed Sep  9 18:46:43 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:09 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
In-Reply-To: <Pine.NEB.3.96.980909192438.24096B-100000@cb1-gw.cb1.com>
Message-ID: <Pine.SOL.3.96.980909133823.20087A-100000@orwell.eng.auburn.edu>

On Wed, 9 Sep 1998, Luke Kenneth Casson Leighton wrote:

> names are not just registered as "mixed" or "hybrid": they must also be
> registered independently on each netbios-broadcast subnet.  as type "B".

I understand why the B_NODE flag is there.  The packet union in
packet_struct { nmb_packet, dgram_packet }.  Both were showing up in the
gdb printout.  Since B_NODE is 0 it was showning up in the dgram_struct
which didn't really matter since the packet_type was NMB_PACKET.

At least I understand that but it looks like it is back to the drawing
board on the original problem.




Thanks again,
j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )


From cartegw at Eng.Auburn.EDU  Wed Sep  9 19:27:26 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:10 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
In-Reply-To: <Pine.SOL.3.96.980909133823.20087A-100000@orwell.eng.auburn.edu>
Message-ID: <Pine.SOL.3.96.980909142247.20094B-100000@orwell.eng.auburn.edu>

I found it!!! :)

I'm still not convienced that somethings wierd in the code, but for now I
have a workaround.

The problem was that the repsonse from the 18p4 WINS server were going to
the real interface on the second machine rather the virtual interface that
the PDC was sitting on.  Only for the <1b> record though.  the rest worked
fine.

The workaround....

- start the PDC on the virtual interface before starting the 18p7 server
on the real interface.



I am going to keep looking through the code and see why things act
differently for the <1b> name.  Will report back later.  Right now I have
a few NT boxes to put back in the domain.



Catch ya later,
j- 
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )

From tridge at samba.anu.edu.au  Wed Sep  9 23:20:22 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:10 2003
Subject: auto-loading of smb.conf
In-Reply-To: <Pine.NEB.3.96.980909172548.23463E-100000@cb1-gw.cb1.com>
	(message from Luke Kenneth Casson Leighton on Wed, 9 Sep 1998 17:27:43
	+0100 (BST))
References: <Pine.NEB.3.96.980909172548.23463E-100000@cb1-gw.cb1.com>
Message-ID: <19980909232032Z12670119-7510+6556@samba.anu.edu.au>

> any shares currently open don't get reloaded / modified: 

they can get modified, they just can't get deleted.

From tridge at samba.anu.edu.au  Wed Sep  9 23:30:17 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:10 2003
Subject: Autocad R14 bug with Samba
In-Reply-To: <Pine.OSF.4.02A.9809091959240.12084-100000@kappa.utc.fr> (message
	from Jean-Francois Micouleau on Thu, 10 Sep 1998 04:06:07 +1000)
References: <Pine.OSF.4.02A.9809091959240.12084-100000@kappa.utc.fr>
Message-ID: <19980909233025Z12669487-25139+6492@samba.anu.edu.au>

> I've got a netmon trace of the autocad R14 bug when you try to open a file
> and the reply is "The file is in read-only mode, blah, blah,..."
> 
> Anybody interrested ???

yes!
 
> Where should I upload it ?

ftp://samba.anu.edu.au/pub/upload/

then let me know what its called 

     Cheers, Tridge
From tridge at samba.anu.edu.au  Thu Sep 10 00:48:31 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:10 2003
Subject: lstat() and fstat()
Message-ID: <19980910004832Z12669903-7510+6579@samba.anu.edu.au>

Jeremy,

I just noticed something that could explain a couple of strange oplock
problems (though not all I suspect). We use a mixture of lstat() and
fstat() in Samba. We also use the inode/device to control oplocks. The
problem is that in the case of a symlink lstat() will return the
inode/device of the link and fstat() will return the inode/device of
the file that the link points to. So if we compare the dev/inode
during a oplock break we will get a mismatch.

I haven't fixed this, I just thought I should mention it before I
forget.
From lkcl at switchboard.net  Thu Sep 10 14:34:42 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
In-Reply-To: <Pine.SOL.3.96.980909133823.20087A-100000@orwell.eng.auburn.edu>
Message-ID: <Pine.NEB.3.96.980910153403.26762B-100000@cb1-gw.cb1.com>

On Wed, 9 Sep 1998, Gerald W. Carter wrote:

> On Wed, 9 Sep 1998, Luke Kenneth Casson Leighton wrote:
> 
> > names are not just registered as "mixed" or "hybrid": they must also be
> > registered independently on each netbios-broadcast subnet.  as type "B".
> 
> I understand why the B_NODE flag is there.  The packet union in
> packet_struct { nmb_packet, dgram_packet }.  Both were showing up in the
> gdb printout.  Since B_NODE is 0 it was showning up in the dgram_struct
> which didn't really matter since the packet_type was NMB_PACKET.

sorry?  B_NODE flag has nothing to do with whether the packet is nmb (port
137) or dgram (port 138).

From lkcl at switchboard.net  Thu Sep 10 14:36:05 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: NT4 sp4, anyone working on it?
Message-ID: <Pine.NEB.3.96.980910153552.26762D-100000@cb1-gw.cb1.com>

ok, the fix to get NT SP4 to talk to samba is really simple, and has been
confirmed as working, by tim winders (ta, tim!).  so, it looks like NT SP4
code is a little more robust (hooray).

in ipc.c, search for make_rpc_hdr:

    /***/
	/*** then do the header, now we know the length ***/
    /***/

	make_rpc_hdr(&p->hdr, RPC_BINDACK, RPC_FLG_FIRST | RPC_FLG_LAST,
				 p->hdr.call_id,
	             p->rdata.offset + p->rauth.offset + 0x10,
	                                               ^^^^^^
	             p->rauth.offset);

	smb_io_rpc_hdr("", &p->hdr, &p->rhdr, 0);
	mem_realloc_data(p->rhdr.data, p->rdata.offset);



From lkcl at switchboard.net  Thu Sep 10 14:47:31 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: lstat() and fstat()
In-Reply-To: <19980910004832Z12669903-7510+6579@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980910154428.26762H-100000@cb1-gw.cb1.com>

> problems (though not all I suspect). We use a mixture of lstat() and
> fstat() in Samba. We also use the inode/device to control oplocks. The

albert chin-a-young wanted to know if it's possible to call "lstat" when
doing directory checks and anything unimportant like that (because it
doesn't trigger an automount storm), and doing "fstat" when doing file
checks.

the problem is that he has over 150 automount directories /
subdirectories, and of course idiot-explorer.exe goes and browses the
whole lot recursively, causing them to all be automounted.

the ideal solution is to use DFS-referrals, but we haven't quite got that
in, yet.  but nigel williams has written the low-level code to do them!

luke

From cartegw at Eng.Auburn.EDU  Thu Sep 10 15:04:29 1998
From: cartegw at Eng.Auburn.EDU (Gerald Carter)
Date: Tue Dec  2 03:02:10 2003
Subject: Error registering DOMAINNAME<1b> with 1.9.18p4 WINS server
References: <Pine.NEB.3.96.980910153403.26762B-100000@cb1-gw.cb1.com>
Message-ID: <35F7EA7D.AA64FF4E@eng.auburn.edu>

Luke Kenneth Casson Leighton wrote:
> 
> sorry?  B_NODE flag has nothing to do with whether the packet is nmb 
> (port 137) or dgram (port 138).

Right but the gdb output was showing both union members (nmb & dgram)
for the packet field in the packet_struct

struct packet_struct
{
  struct packet_struct *next;
  struct packet_struct *prev;
  BOOL locked;
  struct in_addr ip;
  int port;
  int fd;
  time_t timestamp;
  enum packet_type packet_type;
  union {
    struct nmb_packet nmb;
    struct dgram_packet dgram;
  } packet;
    ^^^^^^
};


Becasue of the was the byte was laid out in memory, 

   **if** (the packet would have been of type DGRAM_PACKET) then

the type would have been B_NODE ( which is 0x0 ).  Since the 
packet was of type NMB_PACKET, then this was irrelavent.

Make sense?  Really was just that I wasn't paying attention to 
what I was reading.  Sorry :)




Thanks again,
j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From m.chapman at student.unsw.edu.au  Fri Sep 11 08:26:54 1998
From: m.chapman at student.unsw.edu.au (Matthew Chapman)
Date: Tue Dec  2 03:02:10 2003
Subject: SAMBA Passwords and multiple (trusted) NT domains
References: <6t9bvo$sv0@transfer.stratus.com>
Message-ID: <35F8DECE.C0942505@student.unsw.edu.au>

> Let's say we have two different NT domains - "A" and B". Both of
> these NT domains "trust" each other so that users can access
> resources on each without a problem.
>
> Now lets say we have a SAMBA server that is configured to use
> NT Server "A" for password authentication as follows:
>
>   security = server
>   password server = name of NT server for domain "A"
>
> People from NT domain "A" can use this SAMBA server without
> any problem as you would expect. Their passwords are verified
> by the server and they exist in the NT domain "A".
>
> BUT...
>
> People from domain "B" cannot access the SAMBA server. Although
> the NT server for domain "A" knows about domain "B" where they
> exist (and they're able to browse resources in domain B when using
> NT), apparently the password verification mechanism used by SAMBA
> can't tell NT "go to domain "B" for this user".
>
> Any thoughts or insights on how to make this work (if indeed it's even
> possible) would be greatly appreciated. My fallback plan is to run another
> SAMBA server pointing to domain "B" for these folks.

I'm thinking that in theory this should work. Samba should pass both domain
and username information to the password server, which then authenticates
against the trusted domain.

However I think Samba either doesn't send the correct domain information or
the password server ignores it (this is not entirely unlikely; as far as I
know Win95 and WfWg don't let you specify an alternate domain so Microsoft
could have taken a shortcut in their LanManager code)... any comments ppl?

    Matt


--
Matt Chapman
E-mail: mattyc@cyberdude.com




From oliva at dcc.unicamp.br  Sat Sep 12 08:52:17 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:10 2003
Subject: building on a separate tree
Message-ID: <orzpc5tuoe.fsf@tiete.dcc.unicamp.br>

Hi there!

I've recently checked out a copy of SAMBA with anonymous CVS, and I
found some trouble building it in a separate tree.  I have also found
some errors and inconsistencies in configure.in and Makefile.in, which
I have cleaned up.

The patch that fixes all these problems is attached.  I have ommitted
configure, since it can be easily re-created from configure.in.  I
have created four new files: one is named stamp-h.in, and three other
three, dummy.in.  Patches that create them are at the end of the patch
file, but they won't be added to the CVS repository automatically by
applying the patch (as if you didn't know! :-) :-)

stamp-h.in is part of a technique used in automake for automagically
updating files that are automatically generated from configure.in.  I
have modified Makefile.in so that it implements this technique.  Now,
when you modify configure.in, configure is rebuild and rerun, which
creates a new config.status.  Then, this new config.status is used to
rebuild the Makefile itself.  However, it is not working well for
include/config.h, because no files depend on it.  Setting
INCLUDES=include/config.h did not work, I don't understand why.  I
have tried to arrange that Makefile depends on include/config.h, but
this has some strange implications for GNU make.  Anyway, it is an
alternative if we can't find anything else.

The dummy.in files are for the directories that would not be created
in the build tree, because a dummy file would not be created in them
by the configure script.  I have also modified the existing dummy.in
files so that they contain at least one byte: some platforms don't
like empty files :-(


-Iinclude had to be included in list of preprocessor flags, otherwise
include/config.h, that does not live in the source tree, would not be
found.

Another problem was that installbin.sh and the other scripts would not
create the root of the installation tree ($prefix), only their
subdirectories.  I have fixed this problem by using the install-sh
script to create all needed directories; it will create parent
directories too, if needed.  However, I haven't gone to the trouble of
removing the now-obsolete parts of the scripts, that used to be
responsible for creating directories.

Note that installdirs will sometimes creates some unneeded
directories.  It might have been better to arrange that each install
target calls install-sh with the list of directories is uses by
itself.

I have modified the `clean' and `realclean' targets, so that `clean'
will only clean up intermediate build files, `realclean' will remove
the executables too, and `distclean' (new target) will remove
configure-generated files.


In acconfig.h, I have added a typedef whose lack autoheader was
complaining about.  From the patch to config.h.in, it looks like
someone has modified config.h.in instead of acconfig.h.


In configure.in, I have reduced the number of autoconf warnings about
cross-compilation due to the use of AC_TRY_RUN, and added a fatal
error to prevent SAMBA from being cross-compiled unless the results of
all tests are provided manually, by editing config.cache or defining
the corresponding environment variables.

I have modified all tests that would explicitly print `checking ...'
with calls to AC_CACHE_CHECK, that will not print checking messages
and results if configure --quiet is run.

I have created cache variables where appropriate, so that
reconfiguration is now much faster.

The code for auto-detecting crypt, pam_authenticate and connect would
write incorrect information in the cache, causing reconfigurations to
not include the appropriate libraries in the LIBS variable.  Now it
does the right thing.

The source files for some tests would not be found when
builddir!=srcdir; now they will.


In lib/netmask.c, I have removed a warning message that would be
printed if struct in_addr were not defined in any system header file:
gcc would complain because it would be declared inside an argument
list.  I have provided a forward declaration.


installcp.sh would not find the codepage_def.* files when
builddir!=srcdir; I have modified it so that it accepts the srcdir as
its first argument.  I found that safer than tweaking with $0 within
the script.


Since the patch ended up quite large, I have compressed it with bzip2.
I hope this helps.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil
-------------- next part --------------
A non-text attachment was scrubbed...
Name: build-tree.diff.bz2
Type: application/octet-stream
Size: 6553 bytes
Desc: not available
Url : http://lists.samba.org/archive/samba-technical/attachments/19980912/90b687cd/build-tree.diff.obj
From oliva at dcc.unicamp.br  Sat Sep 12 14:01:50 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and PASSWD_FD
In-Reply-To: Jeremy Allison's message of "Fri, 11 Sep 1998 09:57:58 -0700"
References: <19980908232913Z12669591-7510+6216@samba.anu.edu.au> <35F95696.475288E1@engr.sgi.com>
Message-ID: <or3e9xtgch.fsf_-_@tiete.dcc.unicamp.br>

Jeremy Allison <jallison@cthulhu.engr.sgi.com> writes:

>> 1) create a new command, say, du, that will traverse the directory
>> tree without printing anything but the total bytes. (preferred option)

> Yep - that sounds good to me.

The attached patch implements the interactive command `du' for
smbclient.  It behaves *exactly* as dir, except that it won't print
the directories and filenames as it traverses them.

I have also implemented the `n' switch for the `tar' interactive
command, as well as for the command-line variant of it.  I have chosen
`n' because several Unix commands use the -n switch as a `do-nothing'
(AKA `dry-run') request.  BTW, dry_run will also be implicitly enabled
when creating a tar-file to /dev/null.  The speed-up is wonderful! :-)

Anyway, I still feel annoyed about the way smbclient handles
interactive tar invocations writing to stdout or reading from stdin.
One of the problems is that it inconditionally closes tarhandle, which
may cause it to close stdout and/or stdin when it should be requesting
for more commands.  The second problem is the fact that messages are
printed to stdout by default, even if a tar-file is being created to
it.  IMO, if a tar-file is to be written to stdout, nothing else
should, -E should be implicitly assumed.  I'm pretty sure I've already
submitted a patch that fixed this problem, quite a long time ago, but
I no longer have a copy of it, because someone had told me the problem
was fixed already, and I believed that.  I may write it up again, if
you'd agree on taking one more look at it. :-)

>> How about support for PASSWD_FILE and/or PASSWD_FD?

> Yes, a du and password file patch would be very welcome
> and integrated into Samba 2.0 asap !

With the attached patch, command line overrides PASSWD_FD that
overrides PASSWD_FILE that overrides PASSWD.  PASSWD_FD expects a file
descriptor number, and PASSWD_FILE, a file name.  In the second case,
the file will be opened for reading.  In either case, at most 127
bytes will be read (just like get_pass does), and anything after the
first occurrence of a line-feed or a NUL (if they occur at all) is
ignored.  Empty passwords are reported as such; read errors are
reported too.  After the password is read, in the PASSWD_FILE case,
the opened file is closed.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil
-------------- next part --------------
A non-text attachment was scrubbed...
Name: clitar.diff
Type: application/octet-stream
Size: 8092 bytes
Desc: not available
Url : http://lists.samba.org/archive/samba-technical/attachments/19980912/cf628905/clitar.obj
From Jean-Francois.Micouleau at utc.fr  Sat Sep 12 21:28:12 1998
From: Jean-Francois.Micouleau at utc.fr (Jean-Francois Micouleau)
Date: Tue Dec  2 03:02:10 2003
Subject: profile directory
In-Reply-To: <3.0.5.32.19980912142000.008f9500@pophost.eng.auburn.edu>
Message-ID: <Pine.OSF.4.02A.9809122316180.3776-100000@kappa.utc.fr>

On Sun, 13 Sep 1998, Gerald Carter wrote:

> At 03:54 AM 9/13/98 +1000, Yaroslav Halchinsky wrote:
> >The first is when new user logs in a NT workstation samba creates
> >not the *directory* `profile' but *file* `profile' and in case 
> >profiles do not work at all. Solution - create directory `profile' 
> >first then log in - it works fine.
> 
> Does anyone know if 
> the profile problem mentioned is a bug in the NT client or the 
> Samba code?


I took a netmon trace of an openning session for a user who doesn't
already have a profile. The NT client is sending an NT Transact NT create
call with the flags value set to 0. 

If I read correctly the CIFS spec, the flags value should be 8 here.

I don't have an NT server to check against, so I'm curious of what an NTS
is doing in this case.

	J.F.


-----------------------------------------------------------
Pinky: "What are we going to do tonight, Brain?"
Brain: "The same thing we do every night, Pinky :
	try to install Windows NT !"
-----------------------------------------------------------

From sharpe at ns.aus.com  Sun Sep 13 08:51:40 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and
  PASSWD_FD
In-Reply-To: <or3e9xtgch.fsf_-_@tiete.dcc.unicamp.br>
References: <Jeremy Allison's message of "Fri, 11 Sep 1998 09:57:58 -0700">
 <19980908232913Z12669591-7510+6216@samba.anu.edu.au>
 <35F95696.475288E1@engr.sgi.com>
Message-ID: <3.0.5.32.19980913175140.00796b50@mail.adelaide.on.net>

Hi Alexandre and Jeremy,

At 03:06 AM 9/13/98 +1000, you wrote:
>Jeremy Allison <jallison@cthulhu.engr.sgi.com> writes:
>
>>> 1) create a new command, say, du, that will traverse the directory
>>> tree without printing anything but the total bytes. (preferred option)>
Yep - that sounds good to me.
>
>The attached patch implements the interactive command `du' for
>smbclient.  It behaves *exactly* as dir, except that it won't print
>the directories and filenames as it traverses them.I have also implemented
the `n' switch for the `tar' interactive
>command, as well as for the command-line variant of it.  I have chosen
>`n' because several Unix commands use the -n switch as a `do-nothing'
>(AKA `dry-run') request.  BTW, dry_run will also be implicitly enabled
>when creating a tar-file to /dev/null.  The speed-up is wonderful! :-)
>
>Anyway, I still feel annoyed about the way smbclient handles
>interactive tar invocations writing to stdout or reading from stdin.
>One of the problems is that it inconditionally closes tarhandle, which
>may cause it to close stdout and/or stdin when it should be requesting
>for more commands.  The second problem is the fact that messages are
>printed to stdout by default, even if a tar-file is being created to
>it.  IMO, if a tar-file is to be written to stdout, nothing else
>should, -E should be implicitly assumed.  I'm pretty sure I've already
>submitted a patch that fixed this problem, quite a long time ago, but
>I no longer have a copy of it, because someone had told me the problem
>was fixed already, and I believed that.  I may write it up again, if
>you'd agree on taking one more look at it. :-)

The problem with smbtar/clitar is that I introduced a bug when adding other
functionality to it, for which I apologise. I have seen several patches,
including one with Amanda, none of which fix the real problem, which was/is
a stupid printf in the code :-(

I have sent a correct patch to several people, and will make sure that the
correct patch gets intp the 2.0 source tree, but nothing is likely to get
into 1.9.18 now as we are no longer working on that three anymore I think.

All the patched you include below will be looked at when by me when
client.c has been restructured for 2.0.  This will probably be done by me
next week.

>> How about support for PASSWD_FILE and/or PASSWD_FD?
>
>> Yes, a du and password file patch would be very welcome
>> and integrated into Samba 2.0 asap !With the attached patch, command line 
>>overrides PASSWD_FD that
>overrides PASSWD_FILE that overrides PASSWD.  PASSWD_FD expects a file
>descriptor number, and PASSWD_FILE, a file name.  In the second case,
>the file will be opened for reading.  In either case, at most 127
>bytes will be read (just like get_pass does), and anything after the
>first occurrence of a line-feed or a NUL (if they occur at all) is
>ignored.  Empty passwords are reported as such; read errors are
>reported too.  After the password is read, in the PASSWD_FILE case,
>the opened file is closed.
>
>-- 
>Alexandre Oliva
>mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
>http://www.dcc.unicamp.br/~oliva
>Universidade Estadual de Campinas, SP,
Brasil--Multipart_Sat_Sep_12_14:01:50_1998-1
>Content-Type: application/octet-stream; type=patch
>Content-Disposition: attachment; filename="clitar.diff"
>Content-Transfer-Encoding: 7bit
>
>Attachment Converted: "C:\docs\Eudora-recvd\clitar.diff"
>--Multipart_Sat_Sep_12_14:01:50_1998-1--
>
>

Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From oliva at dcc.unicamp.br  Sun Sep 13 19:55:22 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: Richard Sharpe's message of "Sun, 13 Sep 1998 17:51:40 +0900"
References: <Jeremy Allison's message of "Fri, 11 Sep 1998 09:57:58 -0700"> <19980908232913Z12669591-7510+6216@samba.anu.edu.au> <35F95696.475288E1@engr.sgi.com> <3.0.5.32.19980913175140.00796b50@mail.adelaide.on.net>
Message-ID: <or3e9vr5b9.fsf@tiete.dcc.unicamp.br>

Richard Sharpe <sharpe@ns.aus.com> writes:

>> The second problem is the fact that messages are printed to stdout
>> by default, even if a tar-file is being created to it.  IMO, if a
>> tar-file is to be written to stdout, nothing else should, -E should
>> be implicitly assumed.

> The problem with smbtar/clitar is that I introduced a bug when
> adding other functionality to it, for which I apologise. I have seen
> several patches, including one with Amanda, none of which fix the
> real problem, which was/is a stupid printf in the code :-(

I'm not talking about this particular printf; I'm talking about output 
in general.  If a tar-file is written to stdout, I wouldn't like any
debugging output to get written to stdout too; so -E should be
implicitly assumed in the case of writing a tar-file to stdout.

Sorry if I was not clear enough in the previous message.

I'm not sure this has been fixed in the current sources; my local
build based on yesterday's CVS tree won't connect to SMB servers I've
got access to, not even one running SAMBA 1.9.18p10 :-(

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From lkcl at switchboard.net  Mon Sep 14 14:12:59 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: <or3e9vr5b9.fsf@tiete.dcc.unicamp.br>
Message-ID: <Pine.NEB.3.96.980914151045.14920D-100000@cb1-gw.cb1.com>

> I'm not talking about this particular printf; I'm talking about output 
> in general.  If a tar-file is written to stdout, I wouldn't like any
> debugging output to get written to stdout too; so -E should be
> implicitly assumed in the case of writing a tar-file to stdout.

debug output should go to stderr, with an option to select a file on the
command-line.  interactive commands, and tar output, should go to stdout.

> I'm not sure this has been fixed in the current sources; my local
> build based on yesterday's CVS tree won't connect to SMB servers I've
> got access to, not even one running SAMBA 1.9.18p10 :-(

alexandre, please can you send a message to samba-bugs, with a tcpdump or
netmon trace attached?  thanks!

From admin at intergrafix.net  Mon Sep 14 15:40:53 1998
From: admin at intergrafix.net (System Administrator)
Date: Tue Dec  2 03:02:10 2003
Subject: user mapping and permission problem
Message-ID: <Pine.LNX.4.03.9809141140300.3483-100000@athena.intergrafix.net>


Hi,

I have some web directories that our designers share. 
The directories and files underneath are chowne as <user>.web
where <user> is the <user> the webserver is run as with read and execute
permission and 'web' is the group that the web designers are in with
read,write,and execute permission.

My problem is, when they transfer files over, the owner changes from
<user> to the web designer's user name, which then makes the files they
transfered inaccessible by the web server.

If i use the 'force user = <user>' directive it gives me access denied
because <user> doesn't have write permissions (too much of a security risk
to give the webserver user write permissions). Is there any way to let my
designers transfer files to these directories and have it keep the same
owner of the files?

I'm running samba 1.9.18p4 on Caldera OpenLinux with kernel 2.0.35

Thanx,

-Tony
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.
Anthony J. Biacco                           Network Administrator/Engineer
admin@intergrafix.net                        Intergrafix Internet Services

          "The best way to predict the future, is to invent it."
http://cygnus.ncohafmuta.com                    http://www.intergrafix.net
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.


From davecb at Canada.Sun.COM  Mon Sep 14 19:52:20 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.NEB.3.96.980826145113.8209F-100000@cb1-gw.cb1.com> <35E42545.4B27F08A@canada.sun.com>
Message-ID: <35FD73F4.3500856C@canada.sun.com>

Once upon a time David Collier-Brown wrote:
|				I do expect to see
| significant additions from the NTDOMAIN project's work, mind you!

	Who's a good person to predict and/or discuss
	the organization of the ``domain'' options?

	I need to finish a write-up in the next few
	days, so that it can make technical review.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From mhw at wittsend.com  Tue Sep 15 03:34:41 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:10 2003
Subject: smbmount et al...
In-Reply-To: <19980902035029Z12669168-7009+4496@samba.anu.edu.au> from Andrew Tridgell at "Sep 2, 1998  1:51:47 pm"
Message-ID: <199809150334.XAA31988@alcove.wittsend.com>

Yo!  Tridge...

Andrew Tridgell enscribed thusly:
> > What's the magic incantation or did something get missed in the conversion?

> it got missed :)

	Kinda figured that...

> We're still trying to work out if we want to include smbmount in Samba
> 2.0. The problem with including it is that we don't have anyone who
> can look after it (handle bug reports etc) and Jeremy and I don't use
> it at all. It might have been a mistake to include it in Samba in the
> first place as it is so Linux specific.

	Ok...  I waited until I had Luke safely back in my office and kinda
talked things over with him about this.

	Since you and Jeremy don't use it, but I do, and I've posted fixes
in the past, I guess I'm a prime candidate.  Judging from the number of
requests I've gotten for my smbmount version switching script, I would
say there are a number of others using it (or trying to) as well.  I've
also been fielding complaints / help requests on the linux lists for some
time now.

	I'm also one of the primary bitchers about the change in syntax
from the old smbfs version of smbmount to the new samba version of smbmount.
I'm not going to go down the road of why it got where it is or why it should
or shouldn't have.  We have users that are heavily invested in it and we
have some responsibility to the sins of the past...  So be it.

	I've dug into it and gotten smbmount, smbmnt, and smbumount to
compile under the current CVS tree under both libc (RedHat 4.x) and glibc
(RedHat 5.x) systems.  The patch against the current CVS tree is attached
below.  It only affects Makefile.in, configure.in, and the three smb*.c
files under the client directory.  That SHOULD be safe enough to plug
into CVS without screwing anything else up.  While the smbmount stuff
only applies to Linux, I did not add Linux checks or validation to the
autoconf stuff.  Your shot to call if you think it needs that level of
idiot proofing.  I'll do it if you ask...

	Luke thinks we should just let things "float" and when people want
things fixed, someone will "pony up to the bar" and fix them.  I tend to
agree with him and I really don't want to "own" this particular "tar baby".
However, if that's what it takes to keep this thing rolling, then so be it.
I'll take it and deal with it one way or the other, if that will make life
easier on you and Jeremy.

	I knew Luke was working with some smbclient stuff and I know the
smbmount is some perversion of the smbclient.  If Luke and I can coordinate,
we should be able to keep each other from commiting too many heinious sins.
We might even get some real work done!  :-)  I'll help him out, best I can
and he can keep me from gumming up the works by thrashing around in a
setup I haven't worked in before.  If that's ok with you, that is (and if
you aren't overly worried about the two of us together creating new
synergistic random acts of terrorism)...

	Here is my current patch.  Let me know what you want to do with
it and what you would like me to do.  Yes, I'm volunteering, for what
it's worth.  I just want to keep this thing compiling and working until
if and when anything else better comes along.  Hell...  Might as well
make it something to be happy about...

	BTW...  Just to make sure I maximize confusion all the way around...

	mhw@iss.net == mhw@wittsend.com

	My personal home domain is WittsEnd.com and mhw@wittsend.com is my
home address.  My office persona is mhw@iss.net.  I may send E-Mail from
either or both.  I check both continuously through the day whether I'm
telecommuting from home (3 days a week) or in the office (currently Monday
and Friday).  I know I sent some E-Mail to you and Jeremy, last time Luke
was here, from my iss.net account.  This is just letting you know that I
be he...  :-) ;-> :-/

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!

===== Begin attachment ===== samba.smbmount.diff =====
diff -urN samba/source/Makefile.in samba.smbmount/source/Makefile.in
--- samba/source/Makefile.in	Mon Sep 14 22:57:50 1998
+++ samba.smbmount/source/Makefile.in	Mon Sep 14 18:21:44 1998
@@ -75,7 +75,8 @@
 SPROGS = bin/smbd bin/nmbd bin/swat
 PROGS1 = bin/smbclient bin/testparm bin/testprns bin/smbrun bin/smbstatus 
 PROGS2 = bin/smbpasswd bin/make_smbcodepage
-PROGS = $(PROGS1) $(PROGS2) bin/nmblookup bin/make_printerdef 
+MPROGS = @MPROGS@
+PROGS = $(PROGS1) $(PROGS2) $(MPROGS) bin/nmblookup bin/make_printerdef 
 
 SCRIPTS = script/smbtar script/addtosmbpass
 
@@ -180,6 +181,15 @@
 CLIENT_OBJ = client/client.o client/clientutil.o client/clitar.o \
              $(PARAM_OBJ) $(LIBSMB_OBJ) $(UBIQX_OBJ) $(LIB_OBJ)
 
+MOUNT_OBJ = client/smbmount.o client/clientutil.o \
+             $(PARAM_OBJ) $(LIBSMB_OBJ) $(UBIQX_OBJ) $(LIB_OBJ)
+
+MNT_OBJ = client/smbmnt.o \
+             $(PARAM_OBJ) $(LIBSMB_OBJ) $(UBIQX_OBJ) $(LIB_OBJ)
+
+UMOUNT_OBJ = client/smbumount.o \
+             $(PARAM_OBJ) $(LIBSMB_OBJ) $(UBIQX_OBJ) $(LIB_OBJ)
+
 NMBLOOKUP_OBJ = utils/nmblookup.o $(PARAM_OBJ) $(UBIQX_OBJ) \
                 $(LIBSMB_OBJ) $(LIB_OBJ)
 
@@ -224,6 +234,18 @@
 bin/smbclient: $(CLIENT_OBJ) 
 	@echo Linking $@
 	@$(CC) $(FLAGS) -o $@ $(CLIENT_OBJ) $(LIBS)
+
+bin/smbmount: $(MOUNT_OBJ) 
+	@echo Linking $@
+	@$(CC) $(FLAGS) -o $@ $(MOUNT_OBJ) $(LIBS)
+
+bin/smbmnt: $(MNT_OBJ) 
+	@echo Linking $@
+	@$(CC) $(FLAGS) -o $@ $(MNT_OBJ) $(LIBS)
+
+bin/smbumount: $(UMOUNT_OBJ) 
+	@echo Linking $@
+	@$(CC) $(FLAGS) -o $@ $(UMOUNT_OBJ) $(LIBS)
 
 bin/testparm: $(TESTPARM_OBJ)
 	@echo Linking $@
diff -urN samba/source/client/smbmnt.c samba.smbmount/source/client/smbmnt.c
--- samba/source/client/smbmnt.c	Mon Sep 14 22:57:55 1998
+++ samba.smbmount/source/client/smbmnt.c	Mon Sep 14 23:04:33 1998
@@ -5,33 +5,22 @@
  *
  */
 
-#include <stdio.h>
-#include <string.h>
-#include <signal.h>
-#include <pwd.h>
-#include <grp.h>
-#include <sys/socket.h>
-#include <sys/param.h>
-#include <netinet/in.h>
-#include <netdb.h>
-#include <sys/stat.h>
-#include <sys/types.h>
-/* #include <sys/wait.h> */  /* generates a warning here */
-extern pid_t waitpid(pid_t, int *, int);
-#include <sys/errno.h>
-#include <unistd.h>
-#include <fcntl.h>
-#include <errno.h>
-#include <ctype.h>
-#include <stdlib.h>
-#include <sys/mount.h>
+#include "includes.h"
+
 #include <mntent.h>
 
-#include <linux/fs.h>
+#include <asm/types.h>
 #include <linux/smb.h>
 #include <linux/smb_mount.h>
-
 #include <asm/unistd.h>
+
+#ifndef	MS_MGC_VAL
+/* This may look strange but MS_MGC_VAL is what we are looking for and
+	is what we need from <linux/fs.h> under libc systems and is
+	provided in standard includes on glibc systems.  So...  We
+	switch on what we need...  */
+#include <linux/fs.h>
+#endif
 
 static char *progname;
 
diff -urN samba/source/client/smbmount.c samba.smbmount/source/client/smbmount.c
--- samba/source/client/smbmount.c	Mon Sep 14 22:57:57 1998
+++ samba.smbmount/source/client/smbmount.c	Mon Sep 14 18:22:22 1998
@@ -28,6 +28,8 @@
 #endif
 
 #include "includes.h"
+
+#include <asm/types.h>
 #include <linux/smb_fs.h>
 static struct smb_conn_opt conn_options;
 
diff -urN samba/source/client/smbumount.c samba.smbmount/source/client/smbumount.c
--- samba/source/client/smbumount.c	Mon Sep 14 22:57:58 1998
+++ samba.smbmount/source/client/smbumount.c	Mon Sep 14 18:22:22 1998
@@ -5,30 +5,11 @@
  *
  */
 
-#include <stdio.h>
-#include <string.h>
-#include <signal.h>
-#include <pwd.h>
-#include <grp.h>
-#include <sys/socket.h>
-#include <sys/param.h>
-#include <netinet/in.h>
-#include <netdb.h>
-#include <sys/stat.h>
-#include <sys/types.h>
-/* #include <sys/wait.h> */  /* generates a warning here */
-extern pid_t waitpid(pid_t, int *, int);
-#include <sys/errno.h>
-#include <unistd.h>
-#include <fcntl.h>
-#include <errno.h>
-#include <ctype.h>
-#include <stdlib.h>
-#include <sys/mount.h>
+#include "includes.h"
+
 #include <mntent.h>
 
-#include <sys/ioctl.h>
-#include <linux/fs.h>
+#include <asm/types.h>
 #include <linux/smb.h>
 #include <linux/smb_mount.h>
 #include <linux/smb_fs.h>
diff -urN samba/source/configure.in samba.smbmount/source/configure.in
--- samba/source/configure.in	Mon Sep 14 22:57:54 1998
+++ samba.smbmount/source/configure.in	Mon Sep 14 18:18:25 1998
@@ -6,6 +6,7 @@
 AC_PROG_CC
 AC_PROG_INSTALL
 AC_SUBST(SHELL)
+AC_SUBST(MPROGS)
 AC_PROG_AWK
 
 AC_CANONICAL_SYSTEM
@@ -416,6 +417,27 @@
     ;;
   esac ],
   AC_MSG_RESULT(no)
+)
+
+#################################################
+# check for smbmount support
+AC_MSG_CHECKING(whether to use SMBMOUNT)
+AC_ARG_WITH(smbmount,
+[  --with-smbmount     Include SMBMOUNT (Linux only) support
+  --without-smbmount  Don't include SMBMOUNT support (default)],
+[ case "$withval" in
+  yes)
+    AC_MSG_RESULT(yes)
+    AC_DEFINE(WITH_SMBMOUNT)
+    MPROGS="bin/smbmount bin/smbmnt bin/smbumount"
+    ;;
+  *)
+    AC_MSG_RESULT(no)
+    MPROGS=
+    ;;
+  esac ],
+  AC_MSG_RESULT(no)
+  MPROGS=
 )
 
 #################################################
===== End attachment ===== samba.smbmount.diff =====
From davecb at Canada.Sun.COM  Tue Sep 15 12:09:07 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.NEB.3.96.980826145113.8209F-100000@cb1-gw.cb1.com>
	 <35E42545.4B27F08A@canada.sun.com> <3.0.5.32.19980914231903.0091e7c0@pophost.eng.auburn.edu>
Message-ID: <35FE58E3.9E910EBD@canada.sun.com>

Gerald Carter wrote:
> Not sure I can help, but what do you need to know?
> All my documentation on the NTDOM stuff is linked
> off the main samba site under the "NT Domains FAQ"

	I've read it, and it diluted my ignorance most 
	wonderfully, but I'm primarily looking for a bit 
	of information about the semantics of the new options.

These have been discussed to some extent
      domain controller = 
      domain sid = 
      machine password = 
	and
      security = domain

These haven't.
      domain groups = 
      domain admin group = 
      domain guest group = 
      domain admin users = 
      domain guest users = 

      groupname map = 

Can anyone tell me what the basic intention of these are?
They're new, and not overly documented: they were
mentioned here once as a future work item.

A quick look at the sources shows the domain admin group
lp_domain_admin_group, is used as a user list in api_net_sam_logon
samr_reply_query_usergroups and get_domain_user_groups,
which means its some sort of unix-like group, but what did the
authors **mean*** by it?
	0) a group, just like a unix group
	1) an nt group, distinct from a unix group
	2) a logical merging of the two concepts
	or
	3) an illogical merging of the two (;-))
the existance of groupname map implies any of 0-2.

To put it in the shortest possible terms, I'm
asking the ``why'' question, because the code answers
the ``what'' question.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From cartegw at Eng.Auburn.EDU  Tue Sep 15 12:53:40 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
In-Reply-To: <35FE58E3.9E910EBD@canada.sun.com>
Message-ID: <Pine.SOL.3.96.980915074409.4152B-100000@orwell.eng.auburn.edu>

On Tue, 15 Sep 1998, David Collier-Brown wrote:

> Gerald Carter wrote:
> > Not sure I can help, but what do you need to know?
> > All my documentation on the NTDOM stuff is linked
> > off the main samba site under the "NT Domains FAQ"
> 
> 	I've read it, and it diluted my ignorance most 
> 	wonderfully, but I'm primarily looking for a bit 
> 	of information about the semantics of the new options.
> 
> These have been discussed to some extent
>       domain controller = 

Obselete I think

>       domain sid =

Has been replaced.  The MACHINE SID is generated randomly an now stored in
private/MACHINE.SID (this value is obtained from the value of "domain sid
=" if it exists but other than that the parameter is ignored once
MACHINE.SID is created.

>       machine password = 
> 	and
>       security = domain

Instructions for adding to a domain are in the FAQ.  Once Jeremy is
through with it, it will not be necessary to have matching unix accounts
on the Samba box (or use the username map option to map NT RID's to unix
uids )
 
> These haven't.
>       domain groups = 

List of group RID's to add to the user's info token upon login (i think )
havne't played much with this one.  Haven't found much use yet ( although
combined with the inlude directive, could provide / restrict access to
certain things like dial-up access, etc... )  Will really come in handy
once the RID <-> uid mapping and lsarLookupNames is done.

>       domain admin group = 

Add's the well known group DOMAIN ADMIN group to the user info token which
is sent back during  login.  Takes a list of usernames as a value.

This provides a simple method of creating domain administrative accounts.

>       domain guest group =

Same as domain admin group except it uses the well know GUEST RID.

>       domain admin users = 
>       domain guest users = 

Obselete and soon to be removed.

>       groupname map = 

Works like "username map" except maps NT groups to unix groups ( not in
there yet, is it? )  The idea is to allow mappings such as wheel <->
"Domain Admins", etc...

> Can anyone tell me what the basic intention of these are?
> They're new, and not overly documented: they were
> mentioned here once as a future work item.

Hope that helps.

> A quick look at the sources shows the domain admin group
> lp_domain_admin_group, is used as a user list in api_net_sam_logon
> samr_reply_query_usergroups and get_domain_user_groups,
> which means its some sort of unix-like group, but what did the
> authors **mean*** by it?
> 	0) a group, just like a unix group
> 	1) an nt group, distinct from a unix group
> 	2) a logical merging of the two concepts
> 	or
> 	3) an illogical merging of the two (;-))
> the existance of groupname map implies any of 0-2.
> 
> To put it in the shortest possible terms, I'm
> asking the ``why'' question, because the code answers
> the ``what'' question.



Sorry if I have repeated previous information.  Hope this helps,
j-

________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )

From davecb at Canada.Sun.COM  Tue Sep 15 12:56:35 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.NEB.3.96.980826145113.8209F-100000@cb1-gw.cb1.com>
		 <35E42545.4B27F08A@canada.sun.com> <3.0.5.32.19980914231903.0091e7c0@pophost.eng.auburn.edu> <35FE58E3.9E910EBD@canada.sun.com>
Message-ID: <35FE6403.355916C0@canada.sun.com>

David Collier-Brown wrote:
> To put it in the shortest possible terms, I'm
> asking the ``why'' question, because the code answers
> the ``what'' question.

	Just so people know, I've been through the
	wishlist and Luke's plans documents.

	I know little, not nothing (:-))

--dave
From borsenkow.msk at sni.de  Tue Sep 15 12:58:36 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:10 2003
Subject: samba-2.0-alpha4 with NT5beta2
Message-ID: <003c01bde0a8$8d6b50f0$21c9ca95@ao13.mowp.siemens.ru>

We have one guy with NT5.0beta2 (may be, higher). He cannot use samba. The current version is samba-2.0-alpha4. Here is syslog for it. Has anybody tried it already? Is it NT5 problem? In any case, would be nice if it samba supported it.

greetings

/Andrej

Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   abtest (149.202.201.248) connect to service source as user nobody (uid=104, gid=102) (pid 8773)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/nttrans.c:(333)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/nttrans.c:(333)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 100080
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 1] smbd/service.c:(482)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   abtest (149.202.201.248) connect to service source as user nobody (uid=104, gid=102) (pid 8773)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 1] smbd/service.c:(506)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   abtest (149.202.201.248) closed connection to service source
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/nttrans.c:(333)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/nttrans.c:(333)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/open.c:(983)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   open_directory: unable to stat name = *. Error was No such file or directory
Sep 15 14:51:56 itsrm2 unix: smbd[8773]: [1998/09/15 14:51:56, 0] smbd/open.c:(983)
Sep 15 14:51:56 itsrm2 unix: smbd[8773]:   open_directory: unable to stat name = *. Error was No such file or directory
Sep 15 14:52:05 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:05, 0] smbd/nttrans.c:(333)
Sep 15 14:52:05 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:52:05 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:05, 0] smbd/nttrans.c:(333)
Sep 15 14:52:05 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:52:05 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:05, 0] smbd/open.c:(983)
Sep 15 14:52:05 itsrm2 unix: smbd[8773]:   open_directory: unable to stat name = *. Error was No such file or directory
Sep 15 14:52:49 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:49, 0] smbd/nttrans.c:(333)
Sep 15 14:52:49 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:52:49 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:49, 0] smbd/nttrans.c:(333)
Sep 15 14:52:49 itsrm2 unix: smbd[8773]:   map_share_mode: Incorrect value for desired_access = 80
Sep 15 14:52:49 itsrm2 unix: smbd[8773]: [1998/09/15 14:52:49, 0] smbd/open.c:(983)
Sep 15 14:52:49 itsrm2 unix: smbd[8773]:   open_directory: unable to stat name = *. Error was No such file or directory

From mhw at wittsend.com  Tue Sep 15 13:38:21 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:10 2003
Subject: smbmount et al...
In-Reply-To: <199809150334.XAA31988@alcove.wittsend.com> from "Michael H. Warfield" at "Sep 15, 1998  1:35:36 pm"
Message-ID: <199809151338.JAA00719@alcove.wittsend.com>

Michael H. Warfield enscribed thusly:
> Yo!  Tridge...

	Ooopppsss...

	Didn't mean for this whole thing to go back to samba-technical.

	Deepest apologies to anyone who objects to getting slapped with
a patch on a mailing list.

	Sigh...  That's what I get for typing replies late at night.

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From davecb at Canada.Sun.COM  Tue Sep 15 14:11:34 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.SOL.3.96.980915074409.4152B-100000@orwell.eng.auburn.edu>
Message-ID: <35FE7596.DE70ABAD@canada.sun.com>

Gerald W. Carter wrote:
>  Hope this helps,


	Sure does!  Raised more questions (;-))

	In security=domain, the need for a Unix password
	file is going to disappear (soon, not instantly)

	Is the magic [homes] share going to remain magic?
	I'd assume so, but it's non-obvious what uid
	is going to be invented for the unix user, and
	I assume you'll need to put in an explicit 
	path =  /some/place/%u or the like into [homes].


--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From cartegw at Eng.Auburn.EDU  Tue Sep 15 14:31:18 1998
From: cartegw at Eng.Auburn.EDU (Gerald Carter)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.SOL.3.96.980915074409.4152B-100000@orwell.eng.auburn.edu> <35FE7596.DE70ABAD@canada.sun.com>
Message-ID: <35FE7A36.60305F82@eng.auburn.edu>

David Collier-Brown wrote:
> 
>         In security=domain, the need for a Unix password
>         file is going to disappear (soon, not instantly)

Yes.  Jeremy's "appliance mode"
 
>         Is the magic [homes] share going to remain magic?
>         I'd assume so, but it's non-obvious what uid
>         is going to be invented for the unix user, and
>         I assume you'll need to put in an explicit
>         path =  /some/place/%u or the like into [homes].

I use this solution in labs where I don't actually use the real home
directory from the NIS maps.  Rather I use a local disk area such as
/export/home/%U



j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From davecb at Canada.Sun.COM  Tue Sep 15 14:36:31 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: An implication of domains...
Message-ID: <35FE7B6F.DF8DD1CB@canada.sun.com>

A side point just surfaced in the NT domains discussion.

  Right now, Samba can be a PDC using nt authentication and
a private password file, or a workgroup server using
unix authentication and /etc/passwd.

  Can Samba be a PDC while still taking most of its information
from a Unix authentication domain?  After all, the
password is only one of 7 fields we use:
	user name
	password
	user full name (nee GCOS)
	uid
	primary gid
	other gids (from /etc/groups)
	home directory

  I can see Unix bigots like myself being prepared to live
with the travails of password synchronization in order to
have an ``NT Gateway server'' in the form of Samba, and
be able to administer our authentication domain for everyone
using yp/nis, nis+, kerberos/hesiod or blue pages.
(In fact, that's what some of our sites are doing using
just security=user and yellow pages.)

  Note that is the value-add that commercial products
have, but at the cost of locking you in to their way of
doing things.

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From jallison at cthulhu.engr.sgi.com  Tue Sep 15 16:52:06 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:10 2003
Subject: An implication of domains...
References: <35FE7B6F.DF8DD1CB@canada.sun.com>
Message-ID: <35FE9B36.215440E3@engr.sgi.com>

David Collier-Brown wrote:

> 
>   Can Samba be a PDC while still taking most of its information
> from a Unix authentication domain?  After all, the
> password is only one of 7 fields we use:
>         user name
>         password
>         user full name (nee GCOS)
>         uid
>         primary gid
>         other gids (from /etc/groups)
>         home directory
> 
>   I can see Unix bigots like myself being prepared to live
> with the travails of password synchronization in order to
> have an ``NT Gateway server'' in the form of Samba, and
> be able to administer our authentication domain for everyone
> using yp/nis, nis+, kerberos/hesiod or blue pages.
> (In fact, that's what some of our sites are doing using
> just security=user and yellow pages.)
> 

Well I was considering migrating the info in a smbpasswd file
into a gdbm database to give us full functionality w.r.t 
PDC account parameters.

Things will be much better with NT5.x as we can use
kerberos for all authentication and the password sync
problems go away....... sometime next century :-).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From borsenkow.msk at sni.de  Tue Sep 15 17:04:16 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:10 2003
Subject: DOM\user supported?
Message-ID: <000401bde0ca$dee7f100$21c9ca95@ao13.mowp.siemens.ru>

samba-2alpha4 is part of domain; the securioty is set to be "domain". We have a WinNT server, which is not part of our domain. Let's say, domain is DOMAIN, other server is TEST and user is ehhh... user :)

If user on this server tries to connect to SAMBA, it is first rejected with correct message, that PDC of domain DOMAIN cannot authorise user "user" from domain TEST of other server. But if he tries to now "connect as" DOMAIN\user (this account *does* exists on our domain), SAMBA reports exactly the same!! That is, DOMAIN part seems to be silently ignored?

Who interprets DOMAIN part in DOMAIN\user? Client or server? That is, is it client who sends wrong information or SAMBA just does not know, what to do with this form?

thank you

From davecb at Canada.Sun.COM  Tue Sep 15 17:15:32 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: An implication of domains...
References: <35FE7B6F.DF8DD1CB@canada.sun.com> <35FE9B36.215440E3@engr.sgi.com>
Message-ID: <35FEA0B4.42044EA9@canada.sun.com>

Jeremy Allison wrote:
> Things will be much better with NT5.x as we can use
> kerberos for all authentication and the password sync
> problems go away....... sometime next century :-).

	Alas, they haven't done Hesiod,
	so all the **other** stuff that lives in 
	/etc/passwd has to move to LDAP/x500.

	Which is probably going to take another
	century.  Feh. I'll stick to blue pages(:-))

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From lkcl at switchboard.net  Tue Sep 15 18:30:47 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
In-Reply-To: <35FE58E3.9E910EBD@canada.sun.com>
Message-ID: <Pine.NEB.3.96.980915192305.18826I-100000@cb1-gw.cb1.com>

> These have been discussed to some extent
>       domain controller = 

redundant, being removed.

>       domain sid = 

redundant, being removed.

>       machine password = 

clueless.  presumably this is the initial "workstation trust account"
password.  it should not be called "machine password".  see ACB_WKSTRUST.

> 	and
>       security = domain

this makes samba join an nt domain, just like an nt workstation.  it then
uses "dce/rpc Network LsaSamLogon" passthrough authentication instead of
the stupid "CIFS" passthrough authentication.

> These haven't.
>       domain groups = 
>       domain admin group = 
>       domain guest group = 
>       domain admin users = 
>       domain guest users = 

these are to be replaced with a better system.

>       groupname map = 

this is part of the better system, and will be equivalent to "map
username" but for NT<->unix groups instead of NT<->unix users.

> Can anyone tell me what the basic intention of these are?

the intention is to provide mapping for unix groups to NT "local groups",
a.k.a NT "aliases", and to provide mapping for unix groups to NT "domain
groups".

yes, these two _are_ different, and it is important to distinguish between
them at the NT/Samba level.  even if unix really knows nothing about the
difference [between aliases and groups], at least it can be made to look
like it does :)

luke

From davecb at Canada.Sun.COM  Tue Sep 15 18:43:39 1998
From: davecb at Canada.Sun.COM (David Collier-Brown)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
References: <Pine.NEB.3.96.980915192305.18826I-100000@cb1-gw.cb1.com>
Message-ID: <35FEB55B.3F5E716A@canada.sun.com>

Luke Kenneth Casson Leighton wrote:
> >       domain admin users =
> >       domain guest users =
> 
> these are to be replaced with a better system.
> 
> >       groupname map =
> 
> this is part of the better system, and will be equivalent to "map
> username" but for NT<->unix groups instead of NT<->unix users.

	Sounds good: I had actually started documenting 
	domain groups, domain admin group and domain guest group
	as if they were the basic functionality and
	groupname map as the add-on.

	Being both lazy and cautious, I only wrote
	two paragraphs (;-))  Easy to fix.

	Can you say a few words about the new system
	so I can turn it into a overview (if perhaps
	of vacuous generalities)


> the intention is to provide mapping for unix groups to NT "local groups",
> a.k.a NT "aliases", and to provide mapping for unix groups to NT "domain
> groups".
> 
> yes, these two _are_ different, and it is important to distinguish between
> them at the NT/Samba level.  even if unix really knows nothing about the
> difference [between aliases and groups], at least it can be made to look
> like it does :)

	Ok, I'll draft a description of that
	and get back to you...

--dave
-- 
David Collier-Brown,  | Cherish your enemies.  They're harder to
185 Ellerslie Ave.,   | come by than friends and more motivated.
Willowdale, Ontario   | davecb@canada.sun.com, hobbes.ss.org
N2M 1Y3. 416-223-8968 | http://java.science.yorku.ca/~davecb
From lkcl at switchboard.net  Tue Sep 15 18:40:48 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
In-Reply-To: <35FE7596.DE70ABAD@canada.sun.com>
Message-ID: <Pine.NEB.3.96.980915193832.18826L-100000@cb1-gw.cb1.com>

On Wed, 16 Sep 1998, David Collier-Brown wrote:

> Gerald W. Carter wrote:
> >  Hope this helps,
> 
> 
> 	Sure does!  Raised more questions (;-))
> 
> 	In security=domain, the need for a Unix password
> 	file is going to disappear (soon, not instantly)

true _only_ on a "black-box" system where you do not need any logins other
than root.
 
> 	Is the magic [homes] share going to remain magic?

i see no reason why not.

> 	I'd assume so, but it's non-obvious what uid
> 	is going to be invented for the unix user, and

with the "black-box" option, not only will it not be obvious what unix uid
will be used, but it will also be irrelevant what unix uid is used.  using
samba, it will look like an NT box.  who cares if it's a unix box
underneath!

From lkcl at switchboard.net  Tue Sep 15 18:44:59 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: An implication of domains...
In-Reply-To: <35FE7B6F.DF8DD1CB@canada.sun.com>
Message-ID: <Pine.NEB.3.96.980915194332.18826M-100000@cb1-gw.cb1.com>

>   Can Samba be a PDC while still taking most of its information
> from a Unix authentication domain?  After all, the

of course.  but remember that you will still need an NT and LM 16 byte
hash, which is obtained via the "password database api", which currently
only has private/smbpasswd working yet.

which reminds me.  the smbpasswd command should be modified to use the
password database api, _not_ writining direct to private/smbpasswd.

luke

From lkcl at switchboard.net  Tue Sep 15 18:49:05 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: DOM\user supported?
In-Reply-To: <000401bde0ca$dee7f100$21c9ca95@ao13.mowp.siemens.ru>
Message-ID: <Pine.NEB.3.96.980915194754.18826O-100000@cb1-gw.cb1.com>

andrej,

does the account "user" exist on the samba unix server?  it is a
requirement of the "security = domain" system that a local account with
the same name as the nt domain exists on the samba server.

luke

On Wed, 16 Sep 1998, Andrej Borsenkow wrote:

> samba-2alpha4 is part of domain; the securioty is set to be "domain". We have a WinNT server, which is not part of our domain. Let's say, domain is DOMAIN, other server is TEST and user is ehhh... user :)
> 
> If user on this server tries to connect to SAMBA, it is first rejected with correct message, that PDC of domain DOMAIN cannot authorise user "user" from domain TEST of other server. But if he tries to now "connect as" DOMAIN\user (this account *does* exists on our domain), SAMBA reports exactly the same!! That is, DOMAIN part seems to be silently ignored?
> 
> Who interprets DOMAIN part in DOMAIN\user? Client or server? That is, is it client who sends wrong information or SAMBA just does not know, what to do with this form?
> 
> thank you
> 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From jallison at cthulhu.engr.sgi.com  Tue Sep 15 18:54:00 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:10 2003
Subject: DOM\user supported?
References: <000401bde0ca$dee7f100$21c9ca95@ao13.mowp.siemens.ru>
Message-ID: <35FEB7C8.DD1A150E@engr.sgi.com>

Andrej Borsenkow wrote:
> 
> samba-2alpha4 is part of domain; the securioty is set to be "domain". We have a WinNT server, which is not part of our domain. Let's say, domain is DOMAIN, other server is TEST and user is ehhh... user :)
> 
> If user on this server tries to connect to SAMBA, it is first rejected with correct message, that PDC of domain DOMAIN cannot authorise user "user" from domain TEST of other server. But if he tries to now "connect as" DOMAIN\user (this account *does* exists on our domain), SAMBA reports exactly the same!! That is, DOMAIN part seems to be silently ignored?
> 
> Who interprets DOMAIN part in DOMAIN\user? Client or server? That is, is it client who sends wrong information or SAMBA just does not know, what to do with this form?

The smbd just sends both username and Domain to the PDC
for authentication. Currently, once a "user is authorised"
return is received back Samba drops the Domain portion and
then just uses the user part as normal.

Get a high level debug log to see what's happening.

Cheers,

	Jeremy Allison,
	Samba Team.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From lkcl at switchboard.net  Tue Sep 15 18:59:07 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Changes in Samba 2.0
In-Reply-To: <35FEB55B.3F5E716A@canada.sun.com>
Message-ID: <Pine.NEB.3.96.980915195417.18826T-100000@cb1-gw.cb1.com>

> 	Sounds good: I had actually started documenting 
> 	domain groups, domain admin group and domain guest group
> 	as if they were the basic functionality and
> 	groupname map as the add-on.
> 
> 	Being both lazy and cautious, I only wrote
> 	two paragraphs (;-))  Easy to fix.

good, because they won't exist!
 
> 	Can you say a few words about the new system
> 	so I can turn it into a overview (if perhaps
> 	of vacuous generalities)

eek, not really.  basically, that a system will exist to provide full
"User Manager for Domains" capability, providing a mapping between NT
users, groups and aliases to unix users and groups.  where unix is lacking
in functionality and basic concepts that NT supports, samba will fill in
the missing gaps.  with respect to users and groups (for USRMGR.EXE
support) this will be provided by a "password database API" and a
"group/aliases database API".


From Volker.Lendecke at SerNet.DE  Wed Sep 16 07:20:14 1998
From: Volker.Lendecke at SerNet.DE (Volker.Lendecke@SerNet.DE)
Date: Tue Dec  2 03:02:10 2003
Subject: smbmount et al...
In-Reply-To: <199809150334.XAA31988@alcove.wittsend.com> (mhw@wittsend.com)
Message-ID: <E0zJBss-00003I-00@delphin.SerNet.DE>

-----BEGIN PGP SIGNED MESSAGE-----

Hello!

> 	I'm also one of the primary bitchers about the change in syntax
> from the old smbfs version of smbmount to the new samba version of smbmount.
> I'm not going to go down the road of why it got where it is or why it should
> or shouldn't have.  We have users that are heavily invested in it and we
> have some responsibility to the sins of the past...  So be it.

Is it only the syntax that annoys you or is it also the kernel-level
interface? I must admit that the syntax is really ugly, but it was the
fastest way to get something use the kernel interface. IMO this kernel
interface is necessary, as I did not want all the fancy stuff like
netbios name lookup or encrypted passwords in the kernel. The syntax
could be improved, making the 2.1 stuff look very similar to the 2.0
smbmount and still use smbclient to do the authentication.

Volker

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3i
Charset: noconv

iQCVAwUBNf9moz/9BWnmOc5FAQGG6AP/ag7SYN8D1MEVpSjvnR3r1JsgoI1z2R/0
cC/gN5+thU+aBISRDs6fcinHFFgy0sJ3W6tiCo5H2Q14J7qtwySxSJiJDVew7qWy
I+40azLKwkg4vaNgTeUrvSuPC2iAPieyFcy8sNlQa2znJULnzUvJZ+EM1yX+lruG
DEyJ+dSddPQ=
=UHE2
-----END PGP SIGNATURE-----
From china at pprd.abbott.com  Wed Sep 16 14:39:36 1998
From: china at pprd.abbott.com (Albert Chin-A-Young)
Date: Tue Dec  2 03:02:10 2003
Subject: Multiple Samba servers
Message-ID: <Pine.GSO.3.96.980916092206.11749A-100000@contac.pprd.abbott.com>

We have two Samba servers set up here acting as live failover for each
other. Users map to a DNS name (ufs) which looks like:
	ufs.pprd.abbott.com.    0       A       10.250.95.200
	ufs.pprd.abbott.com.    0       A       10.250.95.199

This works just dandy when one of the machines falls off the net.
However, it presents a problem for locking. Would it be worthwhile to
add functionality to Samba to advertise locking information to other
Samba servers?

-- 
albert chin (china@pprd.abbott.com)

From lkcl at switchboard.net  Wed Sep 16 22:33:03 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: smbmount et al...
In-Reply-To: <E0zJBss-00003I-00@delphin.SerNet.DE>
Message-ID: <Pine.NEB.3.96.980916233017.22799O-100000@cb1-gw.cb1.com>

> Is it only the syntax that annoys you or is it also the kernel-level
> interface? 

well, mike has to rename smbmount to something different, and write a
shell script named "smbmount" that checks the version of the host OS (2.0
or 2.1, whatever), which then calls the renamed-smbmount command as
appropriate. 

all this because "automount" scripts are broken because of the different
syntax.

From oliva at dcc.unicamp.br  Thu Sep 17 00:02:11 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: Luke Kenneth Casson Leighton's message of "Mon, 14 Sep 1998 15:12:59 +0100 (BST)"
References: <Pine.NEB.3.96.980914151045.14920D-100000@cb1-gw.cb1.com>
Message-ID: <orn27z8mrw.fsf@tiete.dcc.unicamp.br>

Luke Kenneth Casson Leighton <lkcl@switchboard.net> writes:

>> I'm not sure this has been fixed in the current sources; my local
>> build based on yesterday's CVS tree won't connect to SMB servers I've
>> got access to, not even one running SAMBA 1.9.18p10 :-(

> alexandre, please can you send a message to samba-bugs, with a tcpdump or
> netmon trace attached?  thanks!

Sure.  I've investigated the problem: it will not occur on my x86
(RedHat Linux 5.1) builds, only on the sparc (Solaris 2.[56]) builds.
Furthermore, I'm only unable to connect to the server if it demands
encryption.  This seems to indicate some byte-ordering problem in the
encryption code.

root@emilia:~[139]# tcpdump -i eth1 -x host sec-server.dcc.unicamp.br
tcpdump: listening on eth1

First, I tried to connect from host saci (143.106.24.137), running
Solaris 2.6, to sec-server, running MSWNT4 (ugh! :-)

01:01:04.597451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: S 3669535075:3669535075(0) win 8760 <mss 1460> (DF)
                         4500 002c 2d8d 4000 fe06 0795 8f6a 1889
                         8f6a 104c 80f7 008b dab8 a963 0000 0000
                         6002 2238 28a6 0000 0204 05b4 3600
01:01:04.597451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: S 1865357290:1865357290(0) ack 3669535076 win 8760 <mss 1460> (DF)
                         4500 002c df15 4000 7c06 d80c 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 17ea dab8 a964
                         6012 2238 a17b 0000 0204 05b4
01:01:04.597451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 1 win 8760 (DF)
                         4500 0028 2d8e 4000 fe06 0798 8f6a 1889
                         8f6a 104c 80f7 008b dab8 a964 6f2f 17eb
                         5010 2238 b938 0000 8f6a 1889 008b
01:01:04.857451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: P 1:77(76) ack 1 win 8760 (DF)
                         4500 0074 2d8f 4000 fe06 074b 8f6a 1889
                         8f6a 104c 80f7 008b dab8 a964 6f2f 17eb
                         5018 2238 9015 0000 8100 0048 2046 4445
                         4645 4443 4e46
01:01:04.857451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: P 1:5(4) ack 77 win 8684 (DF)
                         4500 002c ec15 4000 7c06 cb0c 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 17eb dab8 a9b0
                         5018 21ec 372c 0000 8200 0000
01:01:04.857451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 5 win 8760 (DF)
                         4500 0028 2d90 4000 fe06 0796 8f6a 1889
                         8f6a 104c 80f7 008b dab8 a9b0 6f2f 17ef
                         5010 2238 b8e8 0000 0000 020d 712a
01:01:04.857451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: P 77:245(168) ack 5 win 8760 (DF)
                         4500 00d0 2d91 4000 fe06 06ed 8f6a 1889
                         8f6a 104c 80f7 008b dab8 a9b0 6f2f 17ef
                         5018 2238 cf2f 0000 0000 00a4 ff53 4d42
                         7200 0000 0000
01:01:04.867451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: P 5:114(109) ack 245 win 8516 (DF)
                         4500 0095 ed15 4000 7c06 c9a3 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 17ef dab8 aa58
                         5018 2144 ebb9 0000 0000 0069 ff53 4d42
                         7200 0000 0080
01:01:04.917451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 114 win 8760 (DF)
                         4500 0028 2d92 4000 fe06 0794 8f6a 1889
                         8f6a 104c 80f7 008b dab8 aa58 6f2f 185c
                         5010 2238 b7d3 0000 0000 0000 6056
01:01:07.407451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: P 245:386(141) ack 114 win 8760 (DF)
                         4500 00b5 2d93 4000 fe06 0706 8f6a 1889
                         8f6a 104c 80f7 008b dab8 aa58 6f2f 185c
                         5018 2238 f134 0000 0000 0089 ff53 4d42
                         7300 0000 0008
01:01:07.587451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: . ack 386 win 8375 (DF)
                         4500 0028 8016 4000 7c06 3710 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 185c dab8 aae5
                         5010 20b7 b8c7 0000
01:01:13.097451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: P 114:153(39) ack 386 win 8375 (DF)
                         4500 004f 8517 4000 7c06 31e8 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 185c dab8 aae5
                         5018 20b7 6539 0000 0000 0023 ff53 4d42
                         7301 0005 0088
01:01:13.107451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: F 386:386(0) ack 153 win 8760 (DF)
                         4500 0028 2d94 4000 fe06 0792 8f6a 1889
                         8f6a 104c 80f7 008b dab8 aae5 6f2f 1883
                         5011 2238 b71e 0000 2e44 4343 2e55
01:01:13.117451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.137.33015: F 153:153(0) ack 387 win 8375 (DF)
                         4500 0028 8617 4000 7c06 310f 8f6a 104c
                         8f6a 1889 008b 80f7 6f2f 1883 dab8 aae6
                         5011 20b7 b89e 0000
01:01:13.117451 143.106.24.137.33015 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 154 win 8760 (DF)
                         4500 0028 2d95 4000 fe06 0791 8f6a 1889
                         8f6a 104c 80f7 008b dab8 aae6 6f2f 1884
                         5010 2238 b71d 0000 8f6a 1889 008b

smbclient said:

Session setup failed for username=backup myname=SACI destname=SEC-SERVER.DCC.UNICAMP.BR   ERRDOS - ERRnoaccess (Access denied.)
You might find the -U, -W or -n options useful
Sometimes you have to use `-n USERNAME' (particularly with OS/2)
Some servers also insist on uppercase-only passwords

Then I tried from barnabe (143.106.24.132), running RedHat Linux 5.1,
and succeeded:

01:01:28.647451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: S 188339082:188339082(0) win 512 <mss 1460>
                         4500 002c 4d00 0000 3f06 e727 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d38a 0000 0000
                         6002 0200 66db 0000 0204 05b4 2d53
01:01:28.657451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: S 1865381355:1865381355(0) ack 188339083 win 8760 <mss 1460> (DF)
                         4500 002c 581b 4000 7c06 5f0c 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 75eb 0b39 d38b
                         6012 2238 6177 0000 0204 05b4
01:01:28.657451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 1 win 32120 (DF)
                         4500 0028 4d01 4000 3f06 a72a 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d38b 6f2f 75ec
                         5010 7d78 1df4 0000 0204 05b4 2d53
01:01:28.907451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: P 1:77(76) ack 1 win 32120 (DF)
                         4500 0074 4d02 4000 3f06 a6dd 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d38b 6f2f 75ec
                         5018 7d78 e9ca 0000 8100 0048 2046 4445
                         4645 4443 4e46
01:01:28.917451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: P 1:5(4) ack 77 win 8684 (DF)
                         4500 002c 7a1b 4000 7c06 3d0c 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 75ec 0b39 d3d7
                         5018 21ec f727 0000 8200 0000
01:01:28.917451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: P 77:245(168) ack 5 win 32120 (DF)
                         4500 00d0 4d03 4000 3f06 a680 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d3d7 6f2f 75f0
                         5018 7d78 7162 0000 0000 00a4 ff53 4d42
                         7200 0000 0000
01:01:28.917451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: P 5:114(109) ack 245 win 8516 (DF)
                         4500 0095 7b1b 4000 7c06 3ba3 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 75f0 0b39 d47f
                         5018 2144 297b 0000 0000 0069 ff53 4d42
                         7200 0000 0080
01:01:28.937451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 114 win 32120 (DF)
                         4500 0028 4d06 4000 3f06 a725 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d47f 6f2f 765d
                         5010 7d78 1c8f 0000 0000 0059 e897
01:01:31.177451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: P 245:386(141) ack 114 win 32120 (DF)
                         4500 00b5 4d3c 4000 3f06 a662 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d47f 6f2f 765d
                         5018 7d78 0854 0000 0000 0089 ff53 4d42
                         7300 0000 0008
01:01:31.327451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: . ack 386 win 8375 (DF)
                         4500 0028 9c1c 4000 7c06 1b0f 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 765d 0b39 d50c
                         5010 20b7 78c3 0000
01:01:31.377451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: P 114:207(93) ack 386 win 8375 (DF)
                         4500 0085 9d1c 4000 7c06 19b2 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 765d 0b39 d50c
                         5018 20b7 c02b 0000 0000 0059 ff53 4d42
                         7300 0000 0088
01:01:31.397451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 207 win 32120 (DF)
                         4500 0028 4d45 4000 3f06 a6e6 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d50c 6f2f 76ba
                         5010 7d78 1ba5 0000 0000 00d9 70f3
01:01:31.407451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: P 386:468(82) ack 207 win 32120 (DF)
                         4500 007a 4d47 4000 3f06 a692 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d50c 6f2f 76ba
                         5018 7d78 3494 0000 0000 004e ff53 4d42
                         7500 0000 0008
01:01:31.527451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: . ack 468 win 8293 (DF)
                         4500 0028 b31c 4000 7c06 040f 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 76ba 0b39 d55e
                         5010 2065 7866 0000
01:01:31.917451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: P 207:260(53) ack 468 win 8293 (DF)
                         4500 005d f11c 4000 7c06 c5d9 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 76ba 0b39 d55e
                         5018 2065 1b19 0000 0000 0031 ff53 4d42
                         7500 0000 0088
01:01:31.937451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 260 win 32120 (DF)
                         4500 0028 4d51 4000 3f06 a6da 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d55e 6f2f 76ef
                         5010 7d78 1b1e 0000 0000 004d 3e4b
01:01:33.117451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: P 468:507(39) ack 260 win 32120 (DF)
                         4500 004f 4d68 4000 3f06 a69c 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d55e 6f2f 76ef
                         5018 7d78 958e 0000 0000 0023 ff53 4d42
                         7100 0000 0008
01:01:33.137451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: P 260:299(39) ack 507 win 8254 (DF)
                         4500 004f 5e1d 4000 7c06 58e7 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 76ef 0b39 d585
                         5018 203e f221 0000 0000 0023 ff53 4d42
                         7100 0000 0088
01:01:33.137451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: F 507:507(0) ack 299 win 32120
                         4500 0028 4d6a 0000 3f06 e6c1 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d585 6f2f 7716
                         5011 7d78 1acf 0000 0000 0023 ff53
01:01:33.167451 sec-server.dcc.unicamp.br.netbios-ssn > 143.106.24.132.2136: F 299:299(0) ack 508 win 8254 (DF)
                         4500 0028 621d 4000 7c06 550e 8f6a 104c
                         8f6a 1884 008b 0858 6f2f 7716 0b39 d586
                         5011 203e 7808 0000
01:01:33.167451 143.106.24.132.2136 > sec-server.dcc.unicamp.br.netbios-ssn: . ack 300 win 32120 (DF)
                         4500 0028 4d6e 4000 3f06 a6bd 8f6a 1884
                         8f6a 104c 0858 008b 0b39 d586 6f2f 7717
                         5010 7d78 1ace 0000 0000 0055 abe7


Both builds were based on the current CVS tree.  The compiler is egcs
1.1, if that matters.

Good luck! :-)

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From tridge at samba.anu.edu.au  Thu Sep 17 08:40:19 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:10 2003
Subject: fork bomb fixed
Message-ID: <19980917084033Z12609332-20449+8333@samba.anu.edu.au>

I've fixed the fork bomb reported on samba-ntdom. It was a bug in some
code I added recently to do background browse synchronisation.

I'm released a new 2.0alpha (alpha5) as this bug can easily bring
machines down. 

Sorry for the slow response on this. I wasn't on the samba-ntdom list
for some reason and didn't know about the problem till Jeremy told me
about it this morning.
From lkcl at switchboard.net  Thu Sep 17 16:11:08 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:10 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: <orn27z8mrw.fsf@tiete.dcc.unicamp.br>
Message-ID: <Pine.NEB.3.96.980917170903.24970I-100000@cb1-gw.cb1.com>

> Sure.  I've investigated the problem: it will not occur on my x86
> (RedHat Linux 5.1) builds, only on the sparc (Solaris 2.[56]) builds.
> Furthermore, I'm only unable to connect to the server if it demands
> encryption.  This seems to indicate some byte-ordering problem in the
> encryption code.

ah.
 
> root@emilia:~[139]# tcpdump -i eth1 -x host sec-server.dcc.unicamp.br

... -n -s 1500 options are needed, here.  and a version of tcpdump that
decodes netbios packets also needed.  raw data's hard to stare at unless
you're on a roll and do it all the time...

From sharpe at ns.aus.com  Thu Sep 17 22:14:18 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: Fixing client.c
Message-ID: <3.0.5.32.19980918071418.0085bc70@mail.adelaide.on.net>

Hi,

I have been working on rewriting parts of clitar.c to make it easier to
understand and modify, and now want to fix client.c as we agreed a while ago.

That is, merge in the stuff that Luke did in the NTDOM branch some time ago.

Can anyone tell me how to pull down the relevant stuff so I can merge it in?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Fri Sep 18 04:09:57 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: option ordering
In-Reply-To: <19980917230705Z12669565-20449+8538@samba.anu.edu.au>
	(jra@samba.anu.edu.au)
References: <19980917230705Z12669565-20449+8538@samba.anu.edu.au>
Message-ID: <19980918041005Z12664238-25139+8798@samba.anu.edu.au>

> param/loadparm.c: Moved "blocking locks" into locking section. Alphabetised
>                   locking options. Question - shuld we do this for all options ?

maybe. The only thing it affects is the order they appear in
SWAT. Certainly the ordering in SWAT could be improved, the question
is whether alphabetical ordering is best.

From tridge at samba.anu.edu.au  Fri Sep 18 04:24:39 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: Fixing client.c
In-Reply-To: <3.0.5.32.19980918071418.0085bc70@mail.adelaide.on.net> (message
	from Richard Sharpe on Fri, 18 Sep 1998 07:49:18 +1000)
References: <3.0.5.32.19980918071418.0085bc70@mail.adelaide.on.net>
Message-ID: <19980918042452Z12670293-25139+8799@samba.anu.edu.au>

> I have been working on rewriting parts of clitar.c to make it easier to
> understand and modify, and now want to fix client.c as we agreed a while ago.

thanks!

there are two things to be done:

1) convert client.c to use clientgen.c and allow us to delete
   clientutil.c. This will certainly require that clientgen.c be
   expanded a bit because it doesn't have all the functions needed. 

2) add NT pipe calls that Luke added.

When doing (1) you will come across some weird bits of code that try
to optimise the use of chained open/read/close and readraw/readX for
lots of different situations. This stuff made smbclient fast but
ugly. I suggest you instead make the code really simple (just calling
a cli_read() and cli_write() function) and we can worry about client
optimisation later. 

Eventually we should add cli_open_read() and cli_read_close()
functions which will get the speed back, but it just isn't important
enough to do now and will make your job much harder.

oh, and don't try to do (1) and (2) at the same time. Leave a few days
between them for things to settle down :)
 
> That is, merge in the stuff that Luke did in the NTDOM branch some time ago.

It is more than a simple "merge" because Luke modified clientgen.c
quite a lot. That makes a hack-and-slash merge very difficult. You'll
need to instead work out what Lukes code did and take it across bit by
bit. It is probably a few days work.

> Can anyone tell me how to pull down the relevant stuff so I can merge it in?

you can get Lukes code like this:

cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot login
Password: cvs
cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot co -r BRANCH_NTDOM sambaold


From oliva at dcc.unicamp.br  Fri Sep 18 01:42:57 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:11 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: Luke Kenneth Casson Leighton's message of "Thu, 17 Sep 1998 17:11:08 +0100 (BST)"
References: <Pine.NEB.3.96.980917170903.24970I-100000@cb1-gw.cb1.com>
Message-ID: <orvhmm2fqm.fsf@tiete.dcc.unicamp.br>

I wrote:

>> Sure.  I've investigated the problem: it will not occur on my x86
>> (RedHat Linux 5.1) builds, only on the sparc (Solaris 2.[56]) builds.
>> Furthermore, I'm only unable to connect to the server if it demands
>> encryption.  This seems to indicate some byte-ordering problem in the
>> encryption code.

Luke Kenneth Casson Leighton <lkcl@switchboard.net> writes:

> ... -n -s 1500 options are needed, here.  and a version of tcpdump that
> decodes netbios packets also needed.  raw data's hard to stare at unless
> you're on a roll and do it all the time...

How about this?  143.106.23.129 is the WNT4 server; 143.106.24.138 is
a Solaris 2.6 client, and 143.106.24.132 is a RedHat Linux 5.1 x86.


02:38:58.207451 143.106.24.138.32809 > 143.106.23.129.139: S 279378652:279378652(0) win 8760 <mss 1460> (DF)
02:38:58.217451 143.106.23.129.139 > 143.106.24.138.32809: S 1957631074:1957631074(0) ack 279378653 win 8760 <mss 1460> (DF)
02:38:58.217451 143.106.24.138.32809 > 143.106.23.129.139: . ack 1 win 8760 (DF)
02:38:58.467451 143.106.24.138.32809 > 143.106.23.129.139: P 1:77(76) ack 1 win 8760
>>> NBT Packet
NBT Session Request
Flags=0x81000048
Destination=SEC-SERVER      NameType=0x20 (Server)
Source=RABICO          NameType=0x00 (Workstation)
Data: (4 bytes)
[000] 00 00 00 00                                       .... 

 (DF)
02:38:58.477451 143.106.23.129.139 > 143.106.24.138.32809: P 1:5(4) ack 77 win 8684
>>> NBT Packet
NBT Session Granted
Flags=0x82000000

 (DF)
02:38:58.477451 143.106.24.138.32809 > 143.106.23.129.139: . ack 5 win 8760 (DF)
02:38:58.477451 143.106.24.138.32809 > 143.106.23.129.139: P 77:245(168) ack 5 win 8760
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=164

SMB PACKET: SMBnegprot (REQUEST)
SMB Command   =  0x72
Error class   =  0x0
Error code    =  0
Flags1        =  0x0
Flags2        =  0x0
Tree ID       =  0
Proc ID       =  399
UID           =  404
MID           =  499
Word Count    =  0
Dialect=PC NETWORK PROGRAM 1.0
Dialect=MICROSOFT NETWORKS 1.03
Dialect=MICROSOFT NETWORKS 3.0
Dialect=LANMAN1.0
Dialect=LM1.2X002
Dialect=Samba
Dialect=NT LM 0.12
Dialect=NT LANMAN 1.0


 (DF)
02:38:58.477451 143.106.23.129.139 > 143.106.24.138.32809: P 5:114(109) ack 245 win 8516
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=105

SMB PACKET: SMBnegprot (REPLY)
SMB Command   =  0x72
Error class   =  0x0
Error code    =  0
Flags1        =  0x80
Flags2        =  0x0
Tree ID       =  0
Proc ID       =  399
UID           =  404
MID           =  499
Word Count    =  17
NT1 Protocol
DialectIndex=6
SecMode=0x3
MaxMux=50
NumVcs=1
MaxBuffer=4356
RawSize=65536
SessionKey=0x0
Capabilities=0x43FD
ServerTime=Fri Sep 18 01:40:35 1998
TimeZone=180
CryptKey=Data: (1 bytes)
[000] 08                                                . 
[000] CA 17 1E 88 76 77 ED 44  53 00 45 00 43 00 52 00  ....vw.D S.E.C.R.
[010] 45 00 54 00 41 00 52 00  49 00 41 00 2D 00 49 00  E.T.A.R. I.A.-.I.
[020] 43 00 00 00                                       C... 


 (DF)
02:38:58.517451 143.106.24.138.32809 > 143.106.23.129.139: . ack 114 win 8760 (DF)
02:39:01.927451 143.106.24.138.32809 > 143.106.23.129.139: P 245:386(141) ack 114 win 8760
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=137

SMB PACKET: SMBsesssetupX (REQUEST)
SMB Command   =  0x73
Error class   =  0x0
Error code    =  0
Flags1        =  0x8
Flags2        =  0x1
Tree ID       =  0
Proc ID       =  399
UID           =  404
MID           =  499
Word Count    =  13
Com2=0xFF
Res1=0x0
Off2=0
MaxBuffer=65535
MaxMpx=2
VcNumber=399
SessionKey=0x0
CaseInsensitivePasswordLength=24
CaseSensitivePasswordLength=24
Res=0x0
Capabilities=0x0
Pass1&Pass2&Account&Domain&OS&LanMan=
[000] 73 CA 3D 5E 28 4E B4 A9  13 B3 C3 DC 3F D4 AA C8  s.=^(N.. ....?...
[010] E5 2A B4 FE FA 0D 10 00  10 F6 2B CC CB 92 26 A6  .*...... ..+...&.
[020] 4E DB AF EA 78 A1 99 82  B0 0A 56 6A 3D 9B F0 7A  N...x... ..Vj=..z
[030] 62 61 63 6B 75 70 00 57  4F 52 4B 47 52 4F 55 50  backup.W ORKGROUP
[040] 00 55 6E 69 78 00 53 61  6D 62 61 00              .Unix.Sa mba.


 (DF)
02:39:02.097451 143.106.23.129.139 > 143.106.24.138.32809: . ack 386 win 8375 (DF)
02:39:04.977451 143.106.23.129.139 > 143.106.24.138.32809: P 114:153(39) ack 386 win 8375
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=35

SMB PACKET: SMBsesssetupX (REPLY)
SMB Command   =  0x73
Error class   =  0x1
Error code    =  5
Flags1        =  0x88
Flags2        =  0x1
Tree ID       =  0
Proc ID       =  399
UID           =  404
MID           =  499
Word Count    =  0
SMBError = ERRDOS - ERRnoaccess (Access denied.)


 (DF)
02:39:04.977451 143.106.24.138.32809 > 143.106.23.129.139: F 386:386(0) ack 153 win 8760 (DF)
02:39:04.977451 143.106.23.129.139 > 143.106.24.138.32809: F 153:153(0) ack 387 win 8375 (DF)
02:39:04.977451 143.106.24.138.32809 > 143.106.23.129.139: . ack 154 win 8760 (DF)
02:39:20.657451 143.106.24.132.2178 > 143.106.16.76.139: S 3635411348:3635411348(0) win 512 <mss 1460>
02:39:20.657451 143.106.16.76.139 > 143.106.24.132.2178: S 1957653516:1957653516(0) ack 3635411349 win 8760 <mss 1460> (DF)
02:39:20.657451 143.106.24.132.2178 > 143.106.16.76.139: . ack 1 win 32120 (DF)
02:39:20.917451 143.106.24.132.2178 > 143.106.16.76.139: P 1:77(76) ack 1 win 32120
>>> NBT Packet
NBT Session Request
Flags=0x81000048
Destination=SEC-SERVER      NameType=0x20 (Server)
Source=BARNABE         NameType=0x00 (Workstation)
Data: (4 bytes)
[000] 00 00 00 00                                       .... 

 (DF)
02:39:20.917451 143.106.16.76.139 > 143.106.24.132.2178: P 1:5(4) ack 77 win 8684
>>> NBT Packet
NBT Session Granted
Flags=0x82000000

 (DF)
02:39:20.917451 143.106.24.132.2178 > 143.106.16.76.139: P 77:245(168) ack 5 win 32120
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=164

SMB PACKET: SMBnegprot (REQUEST)
SMB Command   =  0x72
Error class   =  0x0
Error code    =  0
Flags1        =  0x0
Flags2        =  0x0
Tree ID       =  0
Proc ID       =  23115
UID           =  404
MID           =  23215
Word Count    =  0
Dialect=PC NETWORK PROGRAM 1.0
Dialect=MICROSOFT NETWORKS 1.03
Dialect=MICROSOFT NETWORKS 3.0
Dialect=LANMAN1.0
Dialect=LM1.2X002
Dialect=Samba
Dialect=NT LM 0.12
Dialect=NT LANMAN 1.0


 (DF)
02:39:20.927451 143.106.16.76.139 > 143.106.24.132.2178: P 5:114(109) ack 245 win 8516
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=105

SMB PACKET: SMBnegprot (REPLY)
SMB Command   =  0x72
Error class   =  0x0
Error code    =  0
Flags1        =  0x80
Flags2        =  0x0
Tree ID       =  0
Proc ID       =  23115
UID           =  404
MID           =  23215
Word Count    =  17
NT1 Protocol
DialectIndex=6
SecMode=0x3
MaxMux=50
NumVcs=1
MaxBuffer=4356
RawSize=65536
SessionKey=0x0
Capabilities=0x43FD
ServerTime=Fri Sep 18 01:40:57 1998
TimeZone=180
CryptKey=Data: (1 bytes)
[000] 08                                                . 
[000] AB 16 56 23 ED E0 2C 32  53 00 45 00 43 00 52 00  ..V#..,2 S.E.C.R.
[010] 45 00 54 00 41 00 52 00  49 00 41 00 2D 00 49 00  E.T.A.R. I.A.-.I.
[020] 43 00 00 00                                       C... 


 (DF)
02:39:20.937451 143.106.24.132.2178 > 143.106.16.76.139: . ack 114 win 32120 (DF)
02:39:24.387451 143.106.24.132.2178 > 143.106.16.76.139: P 245:386(141) ack 114 win 32120
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=137

SMB PACKET: SMBsesssetupX (REQUEST)
SMB Command   =  0x73
Error class   =  0x0
Error code    =  0
Flags1        =  0x8
Flags2        =  0x1
Tree ID       =  0
Proc ID       =  23115
UID           =  404
MID           =  23215
Word Count    =  13
Com2=0xFF
Res1=0x0
Off2=0
MaxBuffer=65535
MaxMpx=2
VcNumber=23115
SessionKey=0x0
CaseInsensitivePasswordLength=24
CaseSensitivePasswordLength=24
Res=0x0
Capabilities=0x0
Pass1&Pass2&Account&Domain&OS&LanMan=
[000] 1F 2C F0 C0 AB B4 75 3A  A6 FD 73 E1 EE CE C2 81  .,....u: ..s.....
[010] 17 27 EF 04 AD 7B F6 31  19 D6 DE FC 75 87 0E B4  .'...{.1 ....u...
[020] CC FF 7E 84 E1 B1 3A 9D  A5 E7 3F F5 D6 76 E2 C0  ..~...:. ..?..v..
[030] 62 61 63 6B 75 70 00 57  4F 52 4B 47 52 4F 55 50  backup.W ORKGROUP
[040] 00 55 6E 69 78 00 53 61  6D 62 61 00              .Unix.Sa mba.


 (DF)
02:39:26.547451 143.106.24.132.2178 > 143.106.16.76.139: P 468:507(39) ack 260 win 32120
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=35

SMB PACKET: SMBtdis (REQUEST)
SMB Command   =  0x71
Error class   =  0x0
Error code    =  0
Flags1        =  0x8
Flags2        =  0x1
Tree ID       =  30726
Proc ID       =  23115
UID           =  45056
MID           =  23215
Word Count    =  0
smb_bcc=0


 (DF)
02:39:26.547451 143.106.16.76.139 > 143.106.24.132.2178: P 260:299(39) ack 507 win 8254
>>> NBT Packet
NBT Session Packet
Flags=0x0
Length=35

SMB PACKET: SMBtdis (REPLY)
SMB Command   =  0x71
Error class   =  0x0
Error code    =  0
Flags1        =  0x88
Flags2        =  0x1
Tree ID       =  30726
Proc ID       =  23115
UID           =  45056
MID           =  23215
Word Count    =  0
smb_bcc=0


 (DF)
02:39:26.547451 143.106.24.132.2178 > 143.106.16.76.139: F 507:507(0) ack 299 win 32120
02:39:26.557451 143.106.16.76.139 > 143.106.24.132.2178: F 299:299(0) ack 508 win 8254 (DF)
02:39:26.557451 143.106.24.132.2178 > 143.106.16.76.139: . ack 300 win 32120 (DF)

29 packets received by filter
0 packets dropped by kernel


-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From sharpe at ns.aus.com  Fri Sep 18 05:33:21 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: clitar changes and du patches
Message-ID: <3.0.5.32.19980918143321.0086f760@mail.adelaide.on.net>

Hi folks, 

my clitar changes to clean up the restore code are coming along fine. The
code is significantly cleaner and works somewhat at the moment.

I want to start releasing things, but guard against it being used by anyone
other that testers at first.

I am thinking of adding #IFDEF RESTORE_CLEANUP around my new code and
adding a -with_RESTORE_CLEANUP to autoconf ...

I am also looking at adding Alexandre Oliva's du patches.

Anyone have any comments?


Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Fri Sep 18 05:24:33 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: clitar changes and du patches
In-Reply-To: <3.0.5.32.19980918143321.0086f760@mail.adelaide.on.net> (message
	from Richard Sharpe on Fri, 18 Sep 1998 15:07:56 +1000)
References: <3.0.5.32.19980918143321.0086f760@mail.adelaide.on.net>
Message-ID: <19980918052444Z12664217-7510+8777@samba.anu.edu.au>

> my clitar changes to clean up the restore code are coming along fine. The
> code is significantly cleaner and works somewhat at the moment.

excellent.
 
> I want to start releasing things, but guard against it being used by anyone
> other that testers at first.

that's what the alpha branches are for

it can't be worse than my stuffup with the background wins database
save - I crashed lots of peoples machines! The number of nmbd
processes doubled every 5 minutes :)
 
> I am thinking of adding #IFDEF RESTORE_CLEANUP around my new code and
> adding a -with_RESTORE_CLEANUP to autoconf ...

if the new code is cleaner then just delete the old code. Don't be
timid! 
 
> I am also looking at adding Alexandre Oliva's du patches.

great.
 
> Anyone have any comments?

I just thought of another client issue you might like to think
about. Several people have complained that the client is unreliable,
that when taking backups (particularly of Win95 and WfWG machines) it
sometimes gets errors and skips files. I believe this is caused by
resource starvation in the MS servers. To work round this we need to
catch these errors and when they happen sleep for a moment before
retrying the operation.

From lkcl at switchboard.net  Fri Sep 18 14:20:16 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: Fixing client.c
In-Reply-To: <3.0.5.32.19980918071418.0085bc70@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980918151211.24970B-100000@cb1-gw.cb1.com>

cvs -t co sambaold -r BRANCH_NTDOM.  richard, i think you/we would be
better off looking at what i did and doing the same thing, rather than
grabbing BRANCH_NTDOM.

first off: remove clientutil.c

second: global/search/replace all references to the silly global variables
that client.c and clitar.c use in clientutil.c, replacing them with
clientgen.c structure member variables instead.

third (optional): put all other silly global variables that client.c and
clitar.c use for their own purposes into a structure.

fourth (optional): pass the clientgen.c structure and the structure in (3)
above as _arguments_ to all the client.c and clitar.c functions instead of
them being global member variables.

then, putting the BRANCH_NTDOM code in will be a hell of a lot easier.

luke

On Fri, 18 Sep 1998, Richard Sharpe wrote:

> Hi,
> 
> I have been working on rewriting parts of clitar.c to make it easier to
> understand and modify, and now want to fix client.c as we agreed a while ago.
> 
> That is, merge in the stuff that Luke did in the NTDOM branch some time ago.
> 
> Can anyone tell me how to pull down the relevant stuff so I can merge it in?
> 
> 
> 
> Regards
> -------
> Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
> NS Computer Software and Services P/L, 
> Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
> Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Fri Sep 18 14:29:53 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: <orvhmm2fqm.fsf@tiete.dcc.unicamp.br>
Message-ID: <Pine.NEB.3.96.980918152537.24970C-100000@cb1-gw.cb1.com>

ok, nothing wrong with that: you may be right about the byte alignment. 
ok, you written c code before?  try taking the lm and nt password hashes
(16 bytes) and the challenge you see at the start of the Negprot response. 
the challenge is the 8 bytes after the "08" you see on a line of its own,
and is followed by a unicode string "secretatiat-02".

call E_P24 on them (smbdes.c)

or, call SMBNTencrypt and SMBencrypt on your clear-text password and the 8
byte challenge (smbencrypt.c i think).

try compiling this on both the linux box and the solaris one.  if you get
different results, then it's byte alignment.

From jarubio at ultra.citi.com.mx  Fri Sep 18 14:52:18 1998
From: jarubio at ultra.citi.com.mx (Javier Rubio)
Date: Tue Dec  2 03:02:11 2003
Subject: Solaris 2.5.1 reboot problem since samba installation
Message-ID: <360273A2.11EB3989@ultra.citi.com.mx>

I have a sparc ultra 1, with solaris 2.5.1 (SunOs 5.1) , and samba
1.9.17, but, since samba was installed, i have problems when i turn off
or reboot me sparc, the /sbin directory is changed, execution
permissions
are removed, and the sparc no boot.

What i can do?

|-------------------------------------------------------
| Javier Rubio Condelle    -=-  Ingeniero de Integracion
| CITI S.A. de C.V.       -=-        Area de Integracion
| E-mail: jarubio@citi.com.mx - Tel: (8) 52 3572267 x 22
| Beeper: 1511111 Clave: 5115624   -=-        << CITI >>
| BeeperMail: beep_jarubio@citi.com.mx - Message=Subject
|----------------------------------------------------JAR


From jallison at cthulhu.engr.sgi.com  Fri Sep 18 17:45:54 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: CVS update: samba/source/client
References: <19980918124756Z12666551-7510+8896@samba.anu.edu.au>
Message-ID: <36029C52.4AE27FEE@engr.sgi.com>

Hi all,

I'm afraid someone broke the build. We've already
had one complaint on samba-dom about it.

Here's a (gentle :-) pointer to using autoconf :-).

You can't just put things like (from include/includes.h)

---------cut here----------
/* This is the naughty bit.  Autoconf should declare these symbols if
   it finds that GNU Readline is installed.  */

#define HAVE_LIBREADLINE
#define HAVE_READLINE_READLINE_H
#define HAVE_READLINE_HISTORY_H
---------end cut--------------

Into the code without writing the autoconf tests,
especially when you turn the code *on* by default :-).

The build now only works on Linux and other GNU
systems with readline natively installed.

To fix this (as most of the correct tests are
in the code) I added the following to configure.in
to test for the existance of the readline header
files:

AC_CHECK_HEADERS(readline.h history.h readline/readline.h)
AC_CHECK_HEADERS(readline/history.h)

Then to back this up added into include/config.h.in the
lines :

/* Define if you have the <history.h> header file.  */
#undef HAVE_HISTORY_H
/* Define if you have the <readline.h> header file.  */
#undef HAVE_READLINE_H
/* Define if you have the <readline/history.h> header file.  */
#undef HAVE_READLINE_HISTORY_H
/* Define if you have the <readline/readline.h> header file.  */
#undef HAVE_READLINE_READLINE_H

This now means the correct defines are detected at configure
time.

Finally, to check for the existance of the readline 
library, knowing that the function call readline()
exists within it, I added to configure.in :

###############################################
# test for where we get readline() from
if test "$ac_cv_header_readline_h" = "yes" ||
   test "$ac_cv_header_readline_readline_h" = "yes"; then
  AC_CHECK_LIB(readline,readline)
  AC_DEFINE(HAVE_LIBREADLINE)
fi

Hope this helps other autoconf users in Samba.

Cheers,

	Jeremy.




-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From oliva at dcc.unicamp.br  Fri Sep 18 18:00:46 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:11 2003
Subject: CVS update: samba/source/client
In-Reply-To: Jeremy Allison's message of "Sat, 19 Sep 1998 03:58:00 +1000"
References: <19980918124756Z12666551-7510+8896@samba.anu.edu.au> <36029C52.4AE27FEE@engr.sgi.com>
Message-ID: <orzpbxqgox.fsf@araguaia.dcc.unicamp.br>

Jeremy Allison <jallison@cthulhu.engr.sgi.com> writes:

> Then to back this up added into include/config.h.in the
> lines :

Actually, adding anything to include/config.h.in is just a waste of
time, since it is automatically generated from acconfig.h by
autoheader. :-)

> Hope this helps other autoconf users in Samba.

BTW, the patch I've recently posted to this list contains a lot of
fixes of autoconf-related problems, so it would be nice to have it
installed soon, at least to help prevent further addition of incorrect
code to configure.in :-)

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From sharpe at ns.aus.com  Fri Sep 18 11:47:32 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: nmblookup for names on samba server don't work
Message-ID: <3.0.5.32.19980918204732.0085be30@mail.adelaide.on.net>

Hi,

we are having a problem at NCI.

PCs are registered in WINS.dat, I can see themas, say:

RJSPC1        <00>

but if I do an nmblookup rjspc1 I get back a status saying that the name
could not be found.  Can anyone say why that is the case?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Fri Sep 18 23:13:10 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: Problems making samba 2.0.0 under Linux
Message-ID: <3.0.5.32.19980919081310.008601a0@mail.adelaide.on.net>

Hi,

I am having the following problem under Linux 5.0 with GCC with Samba 2.0.0

>Using FLAGS = -g -O2 -I./include -I./ubiqx
>-DSMBLOGFILE="/usr/local/samba/var/log.smb"
>-DNMBLOGFILE="/usr/local/samba/var/log.nmb"
>-DCONFIGFILE="/usr/local/samba/lib/smb.conf"
>-DLMHOSTSFILE="/usr/local/samba/lib/lmhosts"   
>-DSWATDIR="/usr/local/samba/swat" -DSBINDIR="/usr/local/samba/bin"
>-DLOCKDIR="/usr/local/samba/var/locks" -DSMBRUN="/usr/local/samba/bin/smbrun"
>-DCODEPAGEDIR="/usr/local/samba/lib/codepages"
>-DDRIVERFILE="/usr/local/samba/lib/printers.def" -DHAVE_INCLUDES_H
>-DSMB_PASSWD="/usr/local/samba/bin/smbpasswd"
>-DSMB_PASSWD_FILE="/usr/local/samba/private/smbpasswd"
>Using LIBS = -lpam -ldl -lcrypt 
>Compiling printing/printing.c
>printing/printing.c: In function `parse_lpq_hpux':
>printing/printing.c:580: parse error before character constant
>make: *** [printing/printing.o] Error 1

Any ideas ... I have looked at it briefly ...but can't see the solution yet.

Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From jallison at cthulhu.engr.sgi.com  Fri Sep 18 22:57:30 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: nmblookup for names on samba server don't work
References: <3.0.5.32.19980918204732.0085be30@mail.adelaide.on.net>
Message-ID: <3602E55A.64F4EAB5@engr.sgi.com>

Richard Sharpe wrote:
> 
> Hi,
> 
> we are having a problem at NCI.
> 
> PCs are registered in WINS.dat, I can see themas, say:
> 
> RJSPC1        <00>
> 
> but if I do an nmblookup rjspc1 I get back a status saying that the name
> could not be found.  Can anyone say why that is the case?
> 

By default nmblookup is looking for a <0x20> name.
Change the type it's looking for by doing :

nmblookup rjspc1#0

And it should find it.

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From tridge at samba.anu.edu.au  Fri Sep 18 22:59:51 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: nmblookup for names on samba server don't work
In-Reply-To: <3.0.5.32.19980918204732.0085be30@mail.adelaide.on.net> (message
	from Richard Sharpe on Sat, 19 Sep 1998 08:46:25 +1000)
References: <3.0.5.32.19980918204732.0085be30@mail.adelaide.on.net>
Message-ID: <19980918225954Z12617671-7510+9046@samba.anu.edu.au>

> PCs are registered in WINS.dat, I can see themas, say:
> 
> RJSPC1        <00>
> 
> but if I do an nmblookup rjspc1 I get back a status saying that the name
> could not be found.  Can anyone say why that is the case?

do you have "WINS proxy" enabled? Are you doing a WINS query?

you need to either do a WINS (as opposed to bcast) query or enable
WINS proxy support. 

To do a WINS query use:

nmblookup -R -U server name

where server is your WINS server and name is the name you want to
query.

Samba used to always answer such queries, even if malformed. Jeremy
fixed it to do the "right thing" in his rewrite, but that means you
need to craft queries more carefully.
From jallison at cthulhu.engr.sgi.com  Fri Sep 18 23:15:04 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: Problems making samba 2.0.0 under Linux
References: <3.0.5.32.19980919081310.008601a0@mail.adelaide.on.net>
Message-ID: <3602E978.67284165@engr.sgi.com>

Richard Sharpe wrote:
> 
> >printing/printing.c: In function `parse_lpq_hpux':
> >printing/printing.c:580: parse error before character constant
> >make: *** [printing/printing.o] Error 1
> 
> Any ideas ... I have looked at it briefly ...but can't see the solution yet.

Re-check out via CVS. I have fixed this (was a variable
name of TAB - I changed it to htab as TAB can be
defined in one of the system header files on Linux).

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Sat Sep 19 01:25:25 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: Samba-2 on AIX
Message-ID: <36030805.FE23D689@engr.sgi.com>

Ok - I just tested that Samba-2 will compile on our AIX
box (samba2) - and if you do a :

CFLAGS="-D_LARGE_FILES -O" ./configure

it even does the 64 bit file support !

However, an annoyance is that the arguments
to things like accept() on AIX have a 3rd
argument of size_t *, rather than int *.

As our current code uses int * then we get
a few warnings from AIX. I believe POSIX is
going towards a socklen_t data type to fix these
problems but there doesn't seem to be a good
way to autoconf this fix as even if the args
are not correct the compiler just gives a
warning, not an error.

Any good ideas on fixing this in a portable
way ?

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From sharpe at ns.aus.com  Sat Sep 19 11:44:25 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: timing out names in a WINS server
Message-ID: <3.0.5.32.19980919204425.00874390@mail.adelaide.on.net>

Hi,

I have noticed that WINS names hang around in WINS.DAT for a very long time.

Is there any way to time them out?  If the same name is registered later,
an earlier incarnation of the name should be thrown out of WINS should it not?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 19 11:45:47 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: Unknown server error 1311
Message-ID: <3.0.5.32.19980919204547.0085aa40@mail.adelaide.on.net>

Hi folks,

I was trying to do 'smbclient -L pcname -I ipaddr'

where PC name is my laptop running Win 95 that was dialed in to a Linux
machine using WINS, and ipaddr was the IP address assigned to the PC.

I got the following:

SMB: R negotiate - Server Error, (1311) Unknown
    SMB: SMB Status = Server Error, (1311) Unknown
        SMB: Error class = Server Error
        SMB: Server error code = 0x051F
    SMB: Header: PID = 0x7E15 TID = 0x0000 MID = 0x7E79 UID = 0x0000
        SMB: Tree ID      (TID) = 0 (0x0)
        SMB: Process ID   (PID) = 32277 (0x7E15)
        SMB: User ID      (UID) = 0 (0x0)
        SMB: Multiplex ID (MID) = 32377 (0x7E79)
        SMB: Flags Summary = 128 (0x80)
            SMB: .......0 = Lock & Read and Write & Unlock not supported
            SMB: ......0. = Send No Ack not supported
            SMB: ....0... = Using case sensitive pathnames
            SMB: ...0.... = No canonicalized pathnames
            SMB: ..0..... = No Opportunistic lock
            SMB: .0...... = No Change Notify
            SMB: 1....... = Server response
        SMB: flags2 Summary = 0 (0x0)
            SMB: ...............0 = Understands only DOS 8.3 filenames
            SMB: ..............0. = Does not understand extended attributes
            SMB: ...0............ = No DFS capabilities
            SMB: ..0............. = No paging of IO
            SMB: .0.............. = Using SMB status codes
            SMB: 0............... = Using ASCII strings
    SMB: Command = C negotiate
        SMB: Word count = 0

Does anyone know what it means?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 19 15:17:25 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: Win95 TCP/IP and SMB
Message-ID: <3.0.5.32.19980920001725.0085faa0@mail.adelaide.on.net>

Hmmm,

if you write more than max_xmit on an SMB session, Win95 freezes the
connection and the client times out and aborts ... TCP stays alive, and
client send a tree disconnect ... wierd.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 19 22:23:47 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: Serious bug in the clitar code uploaded into 2.0.0
Message-ID: <3.0.5.32.19980920072347.0085d210@mail.adelaide.on.net>

Hi guys,

I have a serious bug in the code uploaded into 2.0.0. It has to do with
Win95 and other OSes that use a MAX_XMIT with odd values that are not a
multiple of TBLOCK (= 512).  Win 95 uses 2920.

I expect to fix the problem over the next few days, and verify that clitar
works with at least Samba on Linux, Win 95 and Win NT.

Please don't release until I verify that it is working.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 19 22:51:32 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS, dialup, and netlogon problems
Message-ID: <3.0.5.32.19980920075132.0085e480@mail.adelaide.on.net>

Hi,

I am having problems at NCI with respect to WINS.

1. We are having a lot of problems with dial-in users not being logged in.
After testing, I am pretty certain that this has to do with the user's PC's
entries being in wins.dat multiple times. That is:

1. The user's PC logs in one night and they get:

    IP address =172.16.10.120

   so they have a PC#00, PC#03 and PC#20 all saying 172.16.10.120

   Next night they log in, because a different modem is used, they get:

    IP address = 172.16.10.121

   so they have a PC#00, PC#03, and PC#20 all saying 172.16.10.121 as well
   as the entries above!

   We have 1.9.18p10 and I have seen the above.

2.  We are doing things with smbclient against the above dial-in PCs, but
smbclient will not use WINS, and if you put WINS SERVER = <server ip
address> in the smb.conf pointing to the machine which runs samba (which is
the same machine that we do smbclient on) and have WINS SUPPORT = on, since
we are a WINS server, nmbd refuses to run.

How are we to get smbclient to do WINS lookups?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 19 23:40:59 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS.DAT and multiple registrations of the same name
Message-ID: <3.0.5.32.19980920084059.0085d5a0@mail.adelaide.on.net>

Hi Folks, 

here is the results of nmblookup -R -U rjspc1 and the wins.dat from my
favourite Samba 1.9.18p10 site!

Sending queries to 172.30.0.33
172.30.0.160 rjspc1<00>
172.30.0.163 rjspc1<00>
Sending queries to 172.30.0.33
172.30.0.160 rjspc1<20>
172.30.0.163 rjspc1<20>
"NCIADL2#20" 906502422 172.30.0.33 46R
"NCIADL2#03" 906502422 172.30.0.33 46R
"NCIADL2#00" 906502422 172.30.0.33 46R
"GOLIATH#20" 906502422 172.30.0.33 44R
"GOLIATH#03" 906502422 172.30.0.33 44R
"GOLIATH#00" 906502422 172.30.0.33 44R
"NCINET#00" 906502455 255.255.255.255 c4R
"NCINET#1e" 906502455 255.255.255.255 c4R
"NCINET#1c" 906502422 172.30.0.33 c4R
"NCINET#1b" 906502422 172.30.0.33 44R
"RJSPC1#03" 906504820 172.30.0.163 172.30.0.160  4R
"RJSPC1#00" 906504824 172.30.0.163 172.30.0.160  4R
"WORKGROUPNAME#00" 906504838 255.255.255.255 84R
"RICHARDS#03" 906502264 172.30.0.163  4R
"RJSPC1#20" 906504815 172.30.0.163 172.30.0.160  4R
"WORKGROUPNAME#1e" 906504834 255.255.255.255 84R

The names RJSPC1#xx have multiple addresses against them.  This seems to
suggest that Samba will allow the same name to be registered more than once!

What is the go here?


Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Sun Sep 20 05:50:18 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: Samba-2 on AIX
In-Reply-To: <36030805.FE23D689@engr.sgi.com> (message from Jeremy Allison on
	Sat, 19 Sep 1998 11:33:14 +1000)
References: <36030805.FE23D689@engr.sgi.com>
Message-ID: <19980920055026Z12670232-7510+9498@samba.anu.edu.au>

> However, an annoyance is that the arguments
> to things like accept() on AIX have a 3rd
> argument of size_t *, rather than int *.

yea, I noticed that and decided to ignore it. It doesn't actually
cause any problems.

> Any good ideas on fixing this in a portable
> way ?

well, one right way to fix it is to have a ACCEPT_CAST define
somewhere. The difficult thing is auto-detecting what the correct
value should be.

From tridge at samba.anu.edu.au  Sun Sep 20 06:10:03 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: Serious bug in the clitar code uploaded into 2.0.0
In-Reply-To: <3.0.5.32.19980920072347.0085d210@mail.adelaide.on.net> (message
	from Richard Sharpe on Sun, 20 Sep 1998 07:58:28 +1000)
References: <3.0.5.32.19980920072347.0085d210@mail.adelaide.on.net>
Message-ID: <19980920061017Z12669905-7510+9501@samba.anu.edu.au>

> Please don't release until I verify that it is working.

alpha release are happening automatically every Tuesday at 5pm
Australian time.

Don't worry though, these alphas are meant to be for the brave only!
From oliva at dcc.unicamp.br  Sun Sep 20 10:03:25 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:11 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: Luke Kenneth Casson Leighton's message of "Fri, 18 Sep 1998 15:29:53 +0100 (BST)"
References: <Pine.NEB.3.96.980918152537.24970C-100000@cb1-gw.cb1.com>
Message-ID: <or1zp7aqci.fsf@araguaia.dcc.unicamp.br>

Luke Kenneth Casson Leighton <lkcl@switchboard.net> writes:

> ok, nothing wrong with that: you may be right about the byte
> alignment.

I was not, just false alarm.  Solaris' getpass() will return a string
with at most 8 characters, and my password had 9.  It's amazing that I
hadn't noticed that before...  I guess that's why SAMBA 1.9.18p10
would replace getpass on Solaris...

Unfortunately, I can't think of any way to detect, at configure time,
whether getpass() should be replaced or not, except using tools such
as expect or some other kind of pty controller :-(

Perhaps we should #define REPLACE_GETPASS by default, until we find a
work-around.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From tridge at samba.anu.edu.au  Sun Sep 20 15:30:08 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: <or1zp7aqci.fsf@araguaia.dcc.unicamp.br> (message from Alexandre
	Oliva on Sun, 20 Sep 1998 23:10:06 +1000)
References: <Pine.NEB.3.96.980918152537.24970C-100000@cb1-gw.cb1.com> <or1zp7aqci.fsf@araguaia.dcc.unicamp.br>
Message-ID: <19980920153016Z12670506-25139+9557@samba.anu.edu.au>

> Perhaps we should #define REPLACE_GETPASS by default, until we find a
> work-around.

only on systems that have the necessary headers for our replacement
getpass code to work. 

in fact, maybe should just have a configure test to see if the 
getpass code compiles, and if it does then use it, otherwise use the
system one.

From landrum at lightlink.com  Mon Sep 21 02:38:56 1998
From: landrum at lightlink.com (Alfred Landrum)
Date: Tue Dec  2 03:02:11 2003
Subject: Porting Samba to BeOS and other non-unix platforms
Message-ID: <Pine.SUN.3.95.980920223345.5627E-100000@adore.lightlink.com>

Hello all,

I'd like to talk to anyone who is porting Samba to a non-unix platform.  I
would like to port Samba to the BeOS, and i thought that people porting to
other various platforms might could get together and help each other out. 

I spoke with Tridge about a month ago (been busy), who said: 

>ok, but when you make changes please make sure you don't do:
>
>#ifdef BEOS 
>do something 
>#else 
>do something else 
>#endif 
> 
>if you do that throughout the code then I'll reject your 
>patch. Instead you will need to add abstraction functions in Samba so 
>that the main Samba code calls general functions (perhaps in 
>lib/system.c or os/beos.c). Then those functions can do BEOS specific 
>stuff on BEOS.  
> 
>I know this is more work for you, but it really has to be done that 
>way. Otherwise we end up with code that is just a nest of #ifdefs for 
>each weird platform that comes along. 

Note: I had just spent two days placing aproximately one billion "#ifdef
BEOS"'s into 1.9.18p7. :)  So, a (quick, bad) port of Samba does exist for
Be, if anyone is interested. 

So, I thought anyone doing ports to the weird platforms could get together
and abstract out the parts we need in some coherent form. If you are,
contact me and we can discuss our respective OS's weirdness. 

Thanks,
Alfred Landrum

FYI:  I interned for Be this summer, and wrote a cifs file system for
them.  If anyone here has interests in Be, Rel4 (due out in Nov/Dec)
should include this. 

PS: Samba rocks. Thank you all.



From mhaigh at village.vut.edu.au  Mon Sep 21 03:35:45 1998
From: mhaigh at village.vut.edu.au (Mick Haigh)
Date: Tue Dec  2 03:02:11 2003
Subject: Samba not printing
Message-ID: <3605C991.1F4337B0@village.vut.edu.au>

Hia guys.

The last few times I've checked out the alpha code I have been unable to
print from a Windows 95 machine.  Printing still works fine from an
NT4.0 box.

A look through a level 5 log file didn't show up anything that (as a
newbie to the code) I thought looked out of place.  If any of you would
like a copy of a log file (at whatever level) to have a look at, give me
a yell.

Thanks.

From Volker.Lendecke at SerNet.DE  Mon Sep 21 06:57:51 1998
From: Volker.Lendecke at SerNet.DE (Volker.Lendecke@SerNet.DE)
Date: Tue Dec  2 03:02:11 2003
Subject: clitar changes and du patches
In-Reply-To: <19980918052444Z12664217-7510+8777@samba.anu.edu.au> (message
	from Andrew Tridgell on Fri, 18 Sep 1998 15:26:27 +1000)
Message-ID: <E0zKzux-000039-00@delphin.SerNet.DE>

-----BEGIN PGP SIGNED MESSAGE-----

> I just thought of another client issue you might like to think
> about. Several people have complained that the client is unreliable,
> that when taking backups (particularly of Win95 and WfWG machines) it
> sometimes gets errors and skips files. I believe this is caused by
> resource starvation in the MS servers. To work round this we need to
> catch these errors and when they happen sleep for a moment before
> retrying the operation.

Is it possible that this is exactly the same problem as the one I
tried to solve with the smbfs WIN95_WORKAROUND (or so?). I thought
that smbclient does not have that problem as it uses a buffer size of
512 entries.

Volker

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3i
Charset: noconv

iQCVAwUBNgX46D/9BWnmOc5FAQEW7AP+NM8QmeaVu/g74SeOc0c3RYc0rhuc4k+v
PoB1t62qSqawsDJZb3u52vfG6FpZaPiZGnsTy/FOKooGDRmnX5Yx2o/r5ZPMfjFw
p56YtR5ev6kPJbI3mOVYmLEEZedFaWuWEkoH2kCY/MM4WR2/Xy+oIy7llGKPZic9
UkNn8uxG9Cc=
=Ccte
-----END PGP SIGNATURE-----
From oliva at dcc.unicamp.br  Mon Sep 21 08:33:15 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:11 2003
Subject: Patch for smbclient command du, tar cn, PASSWD_FILE and   PASSWD_FD
In-Reply-To: Andrew Tridgell's message of "Mon, 21 Sep 1998 01:30:08 +1000"
References: <Pine.NEB.3.96.980918152537.24970C-100000@cb1-gw.cb1.com> <or1zp7aqci.fsf@araguaia.dcc.unicamp.br> <19980920153016Z12670506-25139+9557@samba.anu.edu.au>
Message-ID: <or1zp58zus.fsf@araguaia.dcc.unicamp.br>

Andrew Tridgell <tridge@samba.anu.edu.au> writes:

> in fact, maybe should just have a configure test to see if the 
> getpass code compiles, and if it does then use it, otherwise use the
> system one.

Yup, committed.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From lkcl at switchboard.net  Mon Sep 21 15:48:15 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: Samba-2 on AIX
In-Reply-To: <36030805.FE23D689@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980921164732.5949H-100000@cb1-gw.cb1.com>

> As our current code uses int * then we get
> a few warnings from AIX. I believe POSIX is
> going towards a socklen_t data type to fix these
> problems but there doesn't seem to be a good
> way to autoconf this fix as even if the args
> are not correct the compiler just gives a
> warning, not an error.

you could compile the autoconf test with an options that forces warnings
to be treated as errors.  possible?

From lkcl at switchboard.net  Mon Sep 21 15:53:53 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: timing out names in a WINS server
In-Reply-To: <3.0.5.32.19980919204425.00874390@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980921165232.5949L-100000@cb1-gw.cb1.com>

On Sat, 19 Sep 1998, Richard Sharpe wrote:

> Hi,
> 
> I have noticed that WINS names hang around in WINS.DAT for a very long time.

"expired" names should stay there forever, for informational purposes.
the expiry time should be whatever the client requests but the server
limits (typically an hour?)

From lkcl at switchboard.net  Mon Sep 21 16:01:57 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS, dialup, and netlogon problems
In-Reply-To: <3.0.5.32.19980920075132.0085e480@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980921165603.5949N-100000@cb1-gw.cb1.com>

On Sun, 20 Sep 1998, Richard Sharpe wrote:

> Hi,
> 
> I am having problems at NCI with respect to WINS.
> 
> 1. We are having a lot of problems with dial-in users not being logged in.
> After testing, I am pretty certain that this has to do with the user's PC's
> entries being in wins.dat multiple times. That is:
> 
> 1. The user's PC logs in one night and they get:
> 
>     IP address =172.16.10.120
> 
>    so they have a PC#00, PC#03 and PC#20 all saying 172.16.10.120
> 
>    Next night they log in, because a different modem is used, they get:
> 
>     IP address = 172.16.10.121
> 
>    so they have a PC#00, PC#03, and PC#20 all saying 172.16.10.121 as well
>    as the entries above!

that should be absolutely fine, _as long as_ the "expired" entries in
wins.dat are treated as such, and deleted once a new ip address comes in
for the name.

code something like this:

lookup_name(PC#00)
if (exists())
{
	if (!expired(PC#0))
	{
		send_WACK(); /* to querier: tells them to wait */
		query_current_owner(); /* see if they still want the name*/
		if (still_owns_name())
		{
			reject_name(PC#0);
		}
		else
		{
			delete_old_owner();
			register_name(PC#0);
		}
	}
	else
	{
		delete_expired_name(PC#0);
		register_name(PC#0);
	}
}
else
{
	register_name(PC#0);
}

From jallison at cthulhu.engr.sgi.com  Mon Sep 21 16:32:37 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS, dialup, and netlogon problems
References: <3.0.5.32.19980920075132.0085e480@mail.adelaide.on.net>
Message-ID: <36067FA5.E1543207@engr.sgi.com>

Richard Sharpe wrote:

> 1. The user's PC logs in one night and they get:
> 
>     IP address =172.16.10.120
> 
>    so they have a PC#00, PC#03 and PC#20 all saying 172.16.10.120
> 
>    Next night they log in, because a different modem is used, they get:
> 
>     IP address = 172.16.10.121
> 
>    so they have a PC#00, PC#03, and PC#20 all saying 172.16.10.121 as well
>    as the entries above!
> 
>    We have 1.9.18p10 and I have seen the above.
> 

Well the nmbd in 1.9.18p10 will register a name multiple-times
for the same IP address, but only if the second registration
is a multi-homed registration.

Are these machines coming in over a dialup line ? If so this
dialup IP registration may very well be a multi-homed 
registration.

> 2.  We are doing things with smbclient against the above dial-in PCs, but
> smbclient will not use WINS, and if you put WINS SERVER = <server ip
> address> in the smb.conf pointing to the machine which runs samba (which is
> the same machine that we do smbclient on) and have WINS SUPPORT = on, since
> we are a WINS server, nmbd refuses to run.
> 
> How are we to get smbclient to do WINS lookups?

Doesn't smbclient use the host lookup switchtable 
function ? In which case it should use either
gethostbyname(), WINS or broadcast - depending
on how you have the "name resolve order" set.

Jeremy.


-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Mon Sep 21 16:34:01 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS.DAT and multiple registrations of the same name
References: <3.0.5.32.19980920084059.0085d5a0@mail.adelaide.on.net>
Message-ID: <36067FF9.695EE032@engr.sgi.com>

Richard Sharpe wrote:

> The names RJSPC1#xx have multiple addresses against them.  This seems to
> suggest that Samba will allow the same name to be registered more than once!
> 
> What is the go here?

That's perfectly correct - so long as the machine with
the multiple addresses has one IP interface for each
address.

That's how multi-homed machines use NetBIOS.

Cheers,

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Mon Sep 21 16:50:09 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: Samba-2 on AIX
References: <36030805.FE23D689@engr.sgi.com> <19980920055026Z12670232-7510+9498@samba.anu.edu.au>
Message-ID: <360683C1.167CAD2D@engr.sgi.com>

Andrew Tridgell wrote:
> 
> > However, an annoyance is that the arguments
> > to things like accept() on AIX have a 3rd
> > argument of size_t *, rather than int *.
> 
> yea, I noticed that and decided to ignore it. It doesn't actually
> cause any problems.
> 
> > Any good ideas on fixing this in a portable
> > way ?
> 
> well, one right way to fix it is to have a ACCEPT_CAST define
> somewhere. The difficult thing is auto-detecting what the correct
> value should be.

Well we should just define these types to be
socklen_t (see the new edition of Steven's
UNIX Network Programming - you can get it
here when you come over if you haven't already :-)
and allow autoconf to try and find it in the header
files.

If it fails (as it will on most systems) we need
a way to get autoconf to look for the type of
the last argument of the prototype for accept()
in the system header files and define that as
the type of socklen_t.

Any good way of doing this ?

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Mon Sep 21 17:32:41 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: CVS update: samba/source/smbd
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au>
Message-ID: <36068DB9.16F0096A@engr.sgi.com>

tridge@samba.anu.edu.au wrote:
> 
> The last change was because EB wanted the right volume label. The code
> we had used a non unicode volume label but tests with W95->NT4 showed
> that it has to be unicode. There was a note in the code from Jeremy
> saying that he thought it should _not_ be unicode. Jeremy, can you
> explain why? It certainly didn't work as non-unicode (the client
> displays a garbage volume label) and when I fixed it to use unicode it
> all worked from Win95.

Yep - the reason is that NT Server 4.0 SP3 under these circumstances
insists on the label being ascii. Even if you set the "this packet
is unicode" bit it still treats the label as ascii.

This is a nasty problem -  it seems that Win95 insists on
UNICODE and ignores the "this is ascii" bit, and NT4.0 SP3
insists on ascii, and ignores the "this is unicode" bit.

Hmmmmmm. Client specific code maybe ? (Gag.....)

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From lkcl at switchboard.net  Mon Sep 21 21:08:19 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS.DAT and multiple registrations of the same name
In-Reply-To: <36067FF9.695EE032@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980921220613.6876B-100000@cb1-gw.cb1.com>

On Tue, 22 Sep 1998, Jeremy Allison wrote:

> Richard Sharpe wrote:
> 
> > The names RJSPC1#xx have multiple addresses against them.  This seems to
> > suggest that Samba will allow the same name to be registered more than once!
> > 
> > What is the go here?
> 
> That's perfectly correct - so long as the machine with
> the multiple addresses has one IP interface for each
> address.

if it's an ordinary registration (non-multi-homed) i think what's
happening here is that we are not removing the "expired"  entries from
wins.dat when the new (non-multi-homed) registration comes in. 

> That's how multi-homed machines use NetBIOS.

surely, only if they send a "register multi-homed" packet which microsoft
had to add as a non-rfc1001/2 job?

From lkcl at switchboard.net  Mon Sep 21 21:22:05 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:11 2003
Subject: map "groupname, aliasname, username" issues
In-Reply-To: <3606A05F.9491D686@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980921220943.6876D-100000@cb1-gw.cb1.com>

On Mon, 21 Sep 1998, Jeremy Allison wrote:

> Luke Kenneth Casson Leighton wrote:
> > 
> > instead there will be:
> > 
> > map username
> > groupname map
> 
> We haven't had the argument about this one yet.

hee hee
 
> > alaisname map
> 
> Convince me we need it :-).

sure.

1) you want to give someone local admin rights, but not domain admin
rights, how do you do that?  with only "map groupname", you can't.  you
need "map aliasname".

2) complicated-explanation-that-john-t-gave-me-one-day but i'll try and
see if i can get it right.  it may only apply to NT, as you can always do
unix file permission manipulation.

let's say that you want to move some files from one domain to another, and
you are going to shut down the first domain and bring back a second
domain.  this presents a problem: your SIDs are going to become invalid
(as they are in the first domain). 

so, what you do is you create a "local group", and you make all the
"domain users" and "domain groups" of the about-to-be-retired domain a
_member_ of that local group. 

you then add access rights of this "local group" to all the files you're
going to move to the new domain, prior to taking down the old domain.  

if we are going to be a PDC, then we need to be able to allow users the
right to use "local groups", a.k.a "aliases", and add such permissions to
files.  which means that administrators need to be able to create them,
which i envisage should be provided by something identical in
functionality to "map groupname", and i suggest that we call it "map
aliasname".

luke


From jallison at cthulhu.engr.sgi.com  Mon Sep 21 22:32:27 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: WINS.DAT and multiple registrations of the same name
References: <Pine.NEB.3.96.980921220613.6876B-100000@cb1-gw.cb1.com>
Message-ID: <3606D3FB.988B6E89@engr.sgi.com>

Luke Kenneth Casson Leighton wrote:
> 
> if it's an ordinary registration (non-multi-homed) i think what's
> happening here is that we are not removing the "expired"  entries from
> wins.dat when the new (nowins_process_name_registration_requestn-multi-homed) registration comes in.
> 

Well the code flow in nmbd is :

if( registration request is for name that exists AND ip of new
    reg != IP in WINS.dat ) {

  Send_wack_packet to requestor :

  Send query name to original IP in WINS.DAT

  if( original IP owner still replied and still wants
      the name )

     send registration reject packet to new requestor.

  else

     remove old name from WINS.DAT

     add new name/IP address pair.

}

It's all in nmbd/nmbd_winsserver in wins_process_name_registration_request()
and wins_register_query_fail().

wins_register_query_fail() is the function that removes the
original name in the code section that goes :

---------------------------------cut here---------------------------

  /*
   * We want to just add the name, as we now know the original owner
   * didn't want it. But we can't just do that as an arbitary
   * amount of time may have taken place between the name query
   * request and this timeout/error response. So we check that
   * the name still exists and is in the same state - if so
   * we remove it and call wins_process_name_registration_request()
   * as we know it will do the right thing now.
   */

  namerec = find_name_on_subnet(subrec, question_name, FIND_ANY_NAME);

  if( (namerec != NULL)
   && (namerec->data.source == REGISTER_NAME)
   && ip_equal(rrec->packet->ip, *namerec->data.ip) )
  {
    remove_name_from_namelist( subrec, namerec);
    namerec = NULL;
  }

-----------------------------end cut----------------------------------


Now if richard can instrument this section of code and see
why it is not removing the name from the namelist then
this will tell us what may be wrong in his case.

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Mon Sep 21 22:33:57 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: map "groupname, aliasname, username" issues
References: <Pine.NEB.3.96.980921220943.6876D-100000@cb1-gw.cb1.com>
Message-ID: <3606D455.4B74A63A@engr.sgi.com>

Luke Kenneth Casson Leighton wrote:
> 
> 2) complicated-explanation-that-john-t-gave-me-one-day but i'll try and
> see if i can get it right.  it may only apply to NT, as you can always do
> unix file permission manipulation.
> 

....complex explaination snipped.....

I need to think about this a lot. There may be
a way to achieve the same with UNIX groups on
samba servers within a domain.

Let me get back to you on this one.....

Jeremy.


-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From tridge at samba.anu.edu.au  Tue Sep 22 00:07:06 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: QFSINFO/258
In-Reply-To: <36068DB9.16F0096A@engr.sgi.com> (message from Jeremy Allison on
	Mon, 21 Sep 1998 10:32:41 -0700)
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au> <36068DB9.16F0096A@engr.sgi.com>
Message-ID: <19980922000719Z12670122-25139+9974@samba.anu.edu.au>

> Yep - the reason is that NT Server 4.0 SP3 under these circumstances
> insists on the label being ascii. Even if you set the "this packet
> is unicode" bit it still treats the label as ascii.

is it SP3 specific? Did they really change that in a service pack??

> 
> This is a nasty problem -  it seems that Win95 insists on
> UNICODE and ignores the "this is ascii" bit, and NT4.0 SP3
> insists on ascii, and ignores the "this is unicode" bit.

nasty. How do you get NT4 to generate one of these requests?

and the really weird thing is that NT4 sends it as unicode.

From tridge at samba.anu.edu.au  Tue Sep 22 00:10:17 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: QFSINFO/258
In-Reply-To: <Pine.NEB.3.96.980921220901.6876C-100000@cb1-gw.cb1.com> (message
	from Luke Kenneth Casson Leighton on Tue, 22 Sep 1998 07:22:16 +1000)
References: <Pine.NEB.3.96.980921220901.6876C-100000@cb1-gw.cb1.com>
Message-ID: <19980922001023Z12670680-25139+10011@samba.anu.edu.au>

> can we just throw in ascii->unicode conversions all over the place, and
> have done with it?

if NT4 is assuming ascii then converting to unicode won't help this
problem!
From jallison at cthulhu.engr.sgi.com  Tue Sep 22 01:10:43 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:11 2003
Subject: QFSINFO/258
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au> <36068DB9.16F0096A@engr.sgi.com> <19980922000719Z12670122-25139+9974@samba.anu.edu.au>
Message-ID: <3606F913.39827984@engr.sgi.com>

Andrew Tridgell wrote:
> 
> > Yep - the reason is that NT Server 4.0 SP3 under these circumstances
> > insists on the label being ascii. Even if you set the "this packet
> > is unicode" bit it still treats the label as ascii.
> 
> is it SP3 specific? Did they really change that in a service pack??
> 

Nope - I don't think it's a SP3 specific thing. It's an anoying
bug that's been around a while.

> nasty. How do you get NT4 to generate one of these requests?
> 

Dead easy. Mount a Samba drive, right click on the
drive icon and bingo - you'll see the first letter
of the share name as the label info (as the first
UNICODE null is treated as ascii end-of-string).

Set it to send as ascii and NT can see the whole
label name.

> and the really weird thing is that NT4 sends it as unicode.

Well NT negotiates UNICODE so it would between MS 
clients. The thing to see is if the problem happens
looking at a Win95 server from NT, as that negotiates
the same non-unicode that Samba does.

Cheers,

Jeremy.
-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From sharpe at ns.aus.com  Sun Sep 20 13:49:00 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:11 2003
Subject: smbclient and WINS lookups
Message-ID: <3.0.5.32.19980920224900.00793c00@mail.adelaide.on.net>

Hi,

there seems to be a problem with smbclient in that it appears it cannot do
a WINS lookup if you are trying to use smbclient on the same machine as a
SAMBA server, especially if the SAMBA server is offering WINS capabilities.

Can anyone comment on this, as I need to use WINS with smbclient on the
same machine as the SAMBA and WINS server ...



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Tue Sep 22 12:00:20 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:11 2003
Subject: QFSINFO/258
In-Reply-To: <3606F913.39827984@engr.sgi.com> (message from Jeremy Allison on
	Mon, 21 Sep 1998 18:10:43 -0700)
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au> <36068DB9.16F0096A@engr.sgi.com> <19980922000719Z12670122-25139+9974@samba.anu.edu.au> <3606F913.39827984@engr.sgi.com>
Message-ID: <19980922120030Z12617670-7510+10257@samba.anu.edu.au>

> Nope - I don't think it's a SP3 specific thing. It's an anoying
> bug that's been around a while.

then I'm confused

> Dead easy. Mount a Samba drive, right click on the
> drive icon and bingo - you'll see the first letter
> of the share name as the label info (as the first
> UNICODE null is treated as ascii end-of-string).

did that, and you get "remote server doesn't support requests".

> Set it to send as ascii and NT can see the whole
> label name.

nope, it doesn't even use QSFINFO/258
 
> > and the really weird thing is that NT4 sends it as unicode.
> 
> Well NT negotiates UNICODE so it would between MS 
> clients. The thing to see is if the problem happens
> looking at a Win95 server from NT, as that negotiates
> the same non-unicode that Samba does.

that's exactly what my patch fixed. I was testing from Win95 -> NT4
and NT4 gave back unicode. I changed Samba to do the same.

Are you sure you've seen NT4 generate non-unicode for this request?

Can you try it again? I can't believe it is a US/non-US thing ...

From sharpe at ns.aus.com  Tue Sep 22 12:29:06 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: WINS, dialup, and netlogon problems
In-Reply-To: <36067FA5.E1543207@engr.sgi.com>
References: <3.0.5.32.19980920075132.0085e480@mail.adelaide.on.net>
Message-ID: <3.0.5.32.19980922212906.00882280@mail.adelaide.on.net>

At 09:32 AM 9/21/98 -0700, Jeremy Allison wrote:
>Richard Sharpe wrote:
>
>> 1. The user's PC logs in one night and they get:
>> 
>>     IP address =172.16.10.120
>> 
>>    so they have a PC#00, PC#03 and PC#20 all saying 172.16.10.120
>> 
>>    Next night they log in, because a different modem is used, they get:
>> 
>>     IP address = 172.16.10.121
>> 
>>    so they have a PC#00, PC#03, and PC#20 all saying 172.16.10.121 as well
>>    as the entries above!
>> 
>>    We have 1.9.18p10 and I have seen the above.
>> 
>
>Well the nmbd in 1.9.18p10 will register a name multiple-times
>for the same IP address, but only if the second registration
>is a multi-homed registration.

Except that I think I am seeing the same name registered with different IP
addresses.

>Are these machines coming in over a dialup line ? If so this
>dialup IP registration may very well be a multi-homed 
>registration.

Same machine calling in at different times and getting different modems
that map to different IP addresses.

>> 2.  We are doing things with smbclient against the above dial-in PCs, but
>> smbclient will not use WINS, and if you put WINS SERVER = <server ip
>> address> in the smb.conf pointing to the machine which runs samba (which is
>> the same machine that we do smbclient on) and have WINS SUPPORT = on, since
>> we are a WINS server, nmbd refuses to run.
>> 
>> How are we to get smbclient to do WINS lookups?
>
>Doesn't smbclient use the host lookup switchtable 
>function ? In which case it should use either
>gethostbyname(), WINS or broadcast - depending
>on how you have the "name resolve order" set.

smbclient gets its WINS server address from the smb.conf file.  Since the
server is already a WINS server, we cannot have a WINS SERVER = line in the
smb.conf file as well. 

I have to look into a command line way of specifying the WINS server.

>Jeremy.
>


Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From Volker.Lendecke at SerNet.DE  Tue Sep 22 12:11:13 1998
From: Volker.Lendecke at SerNet.DE (Volker.Lendecke@SerNet.DE)
Date: Tue Dec  2 03:02:12 2003
Subject: clitar changes and du patches
In-Reply-To: <3.0.5.32.19980922212056.0085fca0@mail.adelaide.on.net> (message
	from Richard Sharpe on Tue, 22 Sep 1998 21:20:56 +0900)
Message-ID: <E0zLRHl-0000Jo-00@delphin.SerNet.DE>

-----BEGIN PGP SIGNED MESSAGE-----

> >Is it possible that this is exactly the same problem as the one I
> >tried to solve with the smbfs WIN95_WORKAROUND (or so?). I thought
> >that smbclient does not have that problem as it uses a buffer size of
> >512 entries.
> 
> smbclient uses a TAR block size of 512 bytes, however, it writes to the
> server with blocks of size MAX_XMIT ...

I meant the following '512' from file source/client/client.c

/****************************************************************************
  do a directory listing, calling fn on each file found. Use the TRANSACT2
  call for long filenames
  ****************************************************************************/
static int do_long_dir(char *inbuf,char *outbuf,char *Mask,int attribute,void (*fn)(file_info *),BOOL recurse_dir, BOOL dirstoo)
{
  int max_matches = 512;
                    ^^^

  int info_level = Protocol<PROTOCOL_NT1?1:260; /* NT uses 260, OS/2 uses 2. Both accept 1. */
  char *p;

Volker

-----BEGIN PGP SIGNATURE-----
Version: 2.6.3i
Charset: noconv

iQCVAwUBNgeT2T/9BWnmOc5FAQHDGwP+PQat59kDAO39S4f7ELQruWev0b3zQFOu
7UxAFyJD8rTgEwNuWggi/PmZpW95nFMVEgQ2LVRAjjskS8Du0NgjvRztd5v3BVoc
X3zSzuYTM9MjF1bm7xIdyfuAu70DzMaYrukgXaP3p1mt+CLvgTWXCeCF2WIEI8Rd
u+/rgKFGpzI=
=c46M
-----END PGP SIGNATURE-----
From lkcl at switchboard.net  Tue Sep 22 13:11:13 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: QFSINFO/258
In-Reply-To: <19980922001023Z12670680-25139+10011@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980922141027.8514B-100000@cb1-gw.cb1.com>

ok, what i mean is, if we add support for "CAP_UNICODE" in the SMB
negotiation, does this problem go away?

On Tue, 22 Sep 1998, Andrew Tridgell wrote:

> > can we just throw in ascii->unicode conversions all over the place, and
> > have done with it?
> 
> if NT4 is assuming ascii then converting to unicode won't help this
> problem!
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From tridge at samba.anu.edu.au  Tue Sep 22 13:30:21 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: QFSINFO/258
In-Reply-To: <Pine.NEB.3.96.980922141027.8514B-100000@cb1-gw.cb1.com> (message
	from Luke Kenneth Casson Leighton on Tue, 22 Sep 1998 14:11:13 +0100
	(BST))
References: <Pine.NEB.3.96.980922141027.8514B-100000@cb1-gw.cb1.com>
Message-ID: <19980922133034Z12669365-25139+10252@samba.anu.edu.au>

> ok, what i mean is, if we add support for "CAP_UNICODE" in the SMB
> negotiation, does this problem go away?

not this particular problem, but others might!

From lkcl at switchboard.net  Tue Sep 22 14:02:24 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: QFSINFO/258
In-Reply-To: <19980922120030Z12617670-7510+10257@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980922141757.8514G-100000@cb1-gw.cb1.com>

On Tue, 22 Sep 1998, Andrew Tridgell wrote:

> > Nope - I don't think it's a SP3 specific thing. It's an anoying
> > bug that's been around a while.
> 
> then I'm confused
> 
> > Dead easy. Mount a Samba drive, right click on the
> > drive icon and bingo - you'll see the first letter
> > of the share name as the label info (as the first
> > UNICODE null is treated as ascii end-of-string).
> 
> did that, and you get "remote server doesn't support requests".

oops!  a 1.9.18p10 or other server that doesn't support dce/rpc calls is
needed, methinks. 

From lkcl at switchboard.net  Tue Sep 22 14:06:48 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: WINS, dialup, and netlogon problems
In-Reply-To: <3.0.5.32.19980922212906.00882280@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980922150313.8514H-100000@cb1-gw.cb1.com>

> smbclient gets its WINS server address from the smb.conf file.  Since the
> server is already a WINS server, we cannot have a WINS SERVER = line in the
> smb.conf file as well. 

*ah* - well spotted, richard.  see, the problem comes in that because we
use smb.conf for multiple "processing entities", some of which may be in
the same process and some may not, we get confusion.

nmbd contains both client and server capabilities, with respect to WINS.

strictly speaking, we _should_ require that people specify "wins server =
yes" and also specify "wins server = 127.0.0.1" or "wins server =
your.own.ip.address".

> I have to look into a command line way of specifying the WINS server.

that's one solution.  another is to say "if wins server = yes" then use
127.0.0.1 for wins server ip address.

From crh at NTS.Umn.EDU  Tue Sep 22 14:28:18 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:12 2003
Subject: Porting Samba to BeOS and other non-unix platforms
In-Reply-To: <Pine.SUN.3.95.980920223345.5627E-100000@adore.lightlink.com> from "Alfred Landrum" at Sep 22, 98 01:31:37 am
Message-ID: <199809221428.JAA26654@unet.unet.umn.edu>

Alfred,

Thanks for your interest.  You might take a look at the Amiga port of
Samba.  There is a link to the site on the Samba web page.  On the Amiga
platform the port is partially handled by the existance of a free,
third-party, unix emulation library.  Unfortunately, this library is not
complete, so there are still problems.

As Andrew outlined, the goal would be to isolate functions that must be
re-implemented on different platforms.  The Samba code might then call a
stub function which would call the appropriate function for the target
system.  The first step is to identify the functions that would need such 
treatment.

Chris -)-----

PS.  Where are you located and is there any chance that you'll be going 
     to the CIFS conference next month?

> Hello all,
> 
> I'd like to talk to anyone who is porting Samba to a non-unix platform.  I
> would like to port Samba to the BeOS, and i thought that people porting to
> other various platforms might could get together and help each other out. 
> 
> I spoke with Tridge about a month ago (been busy), who said: 
> 
> >ok, but when you make changes please make sure you don't do:
> >
> >#ifdef BEOS 
> >do something 
> >#else 
> >do something else 
> >#endif 
> > 
> >if you do that throughout the code then I'll reject your 
> >patch. Instead you will need to add abstraction functions in Samba so 
> >that the main Samba code calls general functions (perhaps in 
> >lib/system.c or os/beos.c). Then those functions can do BEOS specific 
> >stuff on BEOS.  
> > 
> >I know this is more work for you, but it really has to be done that 
> >way. Otherwise we end up with code that is just a nest of #ifdefs for 
> >each weird platform that comes along. 
> 
> Note: I had just spent two days placing aproximately one billion "#ifdef
> BEOS"'s into 1.9.18p7. :)  So, a (quick, bad) port of Samba does exist for
> Be, if anyone is interested. 
> 
> So, I thought anyone doing ports to the weird platforms could get together
> and abstract out the parts we need in some coherent form. If you are,
> contact me and we can discuss our respective OS's weirdness. 
> 
> Thanks,
> Alfred Landrum
> 
> FYI:  I interned for Be this summer, and wrote a cifs file system for
> them.  If anyone here has interests in Be, Rel4 (due out in Nov/Dec)
> should include this. 
> 
> PS: Samba rocks. Thank you all.
> 
> 
> 


-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From crh at NTS.Umn.EDU  Tue Sep 22 14:56:03 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:12 2003
Subject: Open LDAP announcement (fwd)
Message-ID: <199809221456.JAA00263@unet.unet.umn.edu>

FYI...

Forwarded message:
> Date: Wed, 26 Aug 1998 11:50:53 -0700
> To: ldap@umich.edu
> From: OpenLDAP <Info@OpenLDAP.Org>
> Subject: [ldap] Announcing OpenLDAP 1.0, an open source LDAP distribution
> Mime-Version: 1.0
> 
> The OpenLDAP Project is pleased to announce the immediate availability
> of OpenLDAP 1.0.  OpenLDAP is a complete Open Source suite of client and
> server applications derived from University of Michigan LDAP v3.3.
> Enhancements include advanced security features, updated platform support,
> and numerous bug fixes.
> 
> This software is available for download:
> 	http://www.OpenLDAP.org/download.html
> or AnonCVS:
> 	http://www.OpenLDAP.org/repo.html
> 
> This is the first release from The OpenLDAP Project.  Volunteers are
> needed.  Please help us develop the next generation LDAP suite.  For
> more information, check out our web site at:
> 	http://www.OpenLDAP.org/
> 
> The OpenLDAP Project is coordinated by the OpenLDAP Foundation, a
> non-profit corporation.  The OpenLDAP Foundation was established to
> promote the development of open source LDAP software and information.
> You can read more about the foundation at:
> 	http://www.OpenLDAP.org/foundation.html
> 
> Enjoy!
> 	Kurt D. Zeilenga <Kurt@OpenLDAP.Org>
> 	Cofounder/Chief Architect
> 	The OpenLDAP Project

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From trep at dem.qc.ca  Tue Sep 22 15:41:22 1998
From: trep at dem.qc.ca (Pierre-Jules Tremblay)
Date: Tue Dec  2 03:02:12 2003
Subject: Problem with ksh UWIN 1.6 and samba
Message-ID: <199809221541.LAA14670@ursula.dem.qc.ca>

A non-text attachment was scrubbed...
Name: not available
Type: text
Size: 2374 bytes
Desc: not available
Url : http://lists.samba.org/archive/samba-technical/attachments/19980922/f56721af/attachment.bat
From admin at intergrafix.net  Tue Sep 22 16:00:48 1998
From: admin at intergrafix.net (System Administrator)
Date: Tue Dec  2 03:02:12 2003
Subject: mapping NT drives to samba shares question
Message-ID: <Pine.LNX.4.03.9809221145230.6015-100000@athena.intergrafix.net>

I still cant get on the samba@ mailing list so I'm posting here.

Running samba 1.9.18p4 on linux. Let's say the machine is called "test"

I have shares in my smb.conf configured as follows. (example)

[excelldent]  
   force group = excell
   comment = Web stuff
   path = /servers/www/sites/excelldent.com
   valid users = webtech
   public = no
   writable = yes
   printable = no
   force create mode = 750
   create mask = 0750
   directory mask = 0750
;   fake oplocks = yes 
   short preserve case = yes
   preserve case = yes

this works all fine and dandy, shows up as a share under the computer
"test" in Network Neighborhood. and I can do this with the rest of my
shares/web sites

my problem is that i can map a drive to the root machine. for someone to
copy a file to say excelldent (from above) they have to map a drive to
each share.

Is there a map to map a drive letter to the root machine. Or even make a
dummy share that i can map a drive to, and put all my real shares
underneath that dummy share.

Thanx,

-Tony
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.
Anthony J. Biacco                           Network Administrator/Engineer
admin@intergrafix.net                        Intergrafix Internet Services

          "The best way to predict the future, is to invent it."
http://cygnus.ncohafmuta.com                    http://www.intergrafix.net
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.

From admin at intergrafix.net  Tue Sep 22 16:24:05 1998
From: admin at intergrafix.net (System Administrator)
Date: Tue Dec  2 03:02:12 2003
Subject: mapping NT drives to samba shares question
In-Reply-To: <Pine.LNX.4.03.9809221145230.6015-100000@athena.intergrafix.net>
Message-ID: <Pine.LNX.4.03.9809221223330.7456-100000@athena.intergrafix.net>

On Wed, 23 Sep 1998, System Administrator wrote:

> my problem is that i can map a drive to the root machine. for someone to
> copy a file to say excelldent (from above) they have to map a drive to
> each share.
> 

that's CAN'T map a drive..sorry

-Tony
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.
Anthony J. Biacco                           Network Administrator/Engineer
admin@intergrafix.net                        Intergrafix Internet Services

          "The best way to predict the future, is to invent it."
http://cygnus.ncohafmuta.com                    http://www.intergrafix.net
.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-._.-.

From jallison at cthulhu.engr.sgi.com  Tue Sep 22 16:40:55 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:12 2003
Subject: QFSINFO/258
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au> <36068DB9.16F0096A@engr.sgi.com> <19980922000719Z12670122-25139+9974@samba.anu.edu.au> <3606F913.39827984@engr.sgi.com> <19980922120030Z12617670-7510+10257@samba.anu.edu.au>
Message-ID: <3607D317.30CCA971@engr.sgi.com>

Andrew Tridgell wrote:

> > Dead easy. Mount a Samba drive, right click on the
> > drive icon and bingo - you'll see the first letter
> > of the share name as the label info (as the first
> > UNICODE null is treated as ascii end-of-string).
> 
> did that, and you get "remote server doesn't support requests".
> 

Ah - sorry. I didn't explain well enough.

Map a drive letter to a Samba share.

Double click on "My Computer".

Right click on the "network drive" icon for the
drive letter you just mapped and you get a menu with 
top entry "Open"... bottom entry "Properties".

Left click on properties, you will get the
drive information screen, called "General"
with a field called "Label". The contents of
this field should be the volume label - but
with a unicode return to an NT client that
didn't negotiate unicode with the server it is
just the first character.

> > Well NT negotiates UNICODE so it would between MS
> > clients. The thing to see is if the problem happens
> > looking at a Win95 server from NT, as that negotiates
> > the same non-unicode that Samba does.
> 
> that's exactly what my patch fixed. I was testing from Win95 -> NT4
> and NT4 gave back unicode. I changed Samba to do the same.
> 

The important thing to test is Nt4 client -> Win95 server.
I only have easy access to NT here so it's hard for me
to test Win95 things.

> Are you sure you've seen NT4 generate non-unicode for this request?
> 

No - NT probably does generate unicode for this request.
The problem is what NT *generates* is irrelevent. It's
what NT will *accept* as a client when the server is operating
in non-unicode mode that's important. And NT wants ascii
in this field (and Win95 doesn't :-).

In the CIFS6 spec the field is marked as either ascii
or unicode, depending on the negotiated options. In
the earlier spec the field is marked as "
unicode only".

My guess is the earlier spec is based on the Win95 code
review, the later spec on the WinNT code review. It really
seems to be a behaviour that needs to be different depending
on the client (much as I *hate* these things).

Now what was that good way you decided of determining
if a client was Win95 or NT again.... ? :-) :-).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From lkcl at switchboard.net  Tue Sep 22 20:02:38 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: max buffer sizes
Message-ID: <Pine.NEB.3.96.980922210043.8514i-100000@cb1-gw.cb1.com>

does "Max Buffer Size", negotiated by the SMBnegprot response and the
SMBsesssetupX request, set the limit on SMBreadX and SMBwriteX calls?

or is it "Max Raw Size"?

thanks in advance,

luke

From oliva at dcc.unicamp.br  Tue Sep 22 20:44:44 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:12 2003
Subject: Samba-2 on AIX
In-Reply-To: Jeremy Allison's message of "Tue, 22 Sep 1998 03:11:54 +1000"
References: <36030805.FE23D689@engr.sgi.com> <19980920055026Z12670232-7510+9498@samba.anu.edu.au> <360683C1.167CAD2D@engr.sgi.com>
Message-ID: <oriuif67bn.fsf@araguaia.dcc.unicamp.br>

Jeremy Allison <jallison@cthulhu.engr.sgi.com> writes:

> If it fails (as it will on most systems) we need
> a way to get autoconf to look for the type of
> the last argument of the prototype for accept()
> in the system header files and define that as
> the type of socklen_t.

> Any good way of doing this ?

Nope.  In fact, accept() is a macro on some systems, which opens a can
of worms.  I'd better live with the warnings :-(

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From tridge at samba.anu.edu.au  Wed Sep 23 01:10:11 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: QFSINFO/258
In-Reply-To: <3607D317.30CCA971@engr.sgi.com> (message from Jeremy Allison on
	Tue, 22 Sep 1998 09:40:55 -0700)
References: <19980920154822Z12670510-20449+9432@samba.anu.edu.au> <36068DB9.16F0096A@engr.sgi.com> <19980922000719Z12670122-25139+9974@samba.anu.edu.au> <3606F913.39827984@engr.sgi.com> <19980922120030Z12617670-7510+10257@samba.anu.edu.au> <3607D317.30CCA971@engr.sgi.com>
Message-ID: <19980923011019Z12642322-7510+10454@samba.anu.edu.au>

ok, I've got my NT box back together now (it is hard to test stuff
with no video card) and have seen the NT client behaviour. I'll see if
I can come up with an improved client type detector that allows us to
reliably modify our behaviour according to what client it is.
From mhaigh at village.vut.edu.au  Wed Sep 23 02:22:06 1998
From: mhaigh at village.vut.edu.au (Mick Haigh)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
Message-ID: <36085B4E.3173CCF1@village.vut.edu.au>

Something similar to this has been posted to the samba group a few
times, but I didn't see any replys.  Anyway - the last few revisions of
the code that I've checked out (including one from 12:11pm Sept 23 EST)
have broken printing from Win95 machines.

NT prints fine but when I try to print from a 95 box I get

There was an error writing to \\village\laser for printer (HP LaserJet
4M Plus):
There was a problem printing to the printer due to an unknown system
error.
Restart Windows, and the try printer again.
This printer will be set to work offline.
To save your print job in the local printer queue, click OK.

A log level of 10 shows nothing obviously untoward (to someone not
exactly familiar with the code).

Any ideas??

Mick

From tridge at samba.anu.edu.au  Wed Sep 23 02:50:22 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
In-Reply-To: <36085B4E.3173CCF1@village.vut.edu.au> (message from Mick Haigh
	on Wed, 23 Sep 1998 12:19:18 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au>
Message-ID: <19980923025025Z12639646-7009+10376@samba.anu.edu.au>

> Something similar to this has been posted to the samba group a few
> times, but I didn't see any replys.  Anyway - the last few revisions of
> the code that I've checked out (including one from 12:11pm Sept 23 EST)
> have broken printing from Win95 machines.

It's not as simple as that. Printing works fine from Win95 for me (my
wife uses it all the time).

> There was an error writing to \\village\laser for printer (HP LaserJet
> 4M Plus):
> There was a problem printing to the printer due to an unknown system
> error.
> Restart Windows, and the try printer again.

you may have to do a comparitive sniff to work out what's
wrong. Download a earlier version (one that works) then get a sniff or
log of that printing. Then use the current version and get another
log. Then look at them to find the first point at which they diverge.
Then stare at the surrounding packets till you find what the
difference is in how Samba reponded.

Alternatively (if you don't like staring at sniffs) use the -D option
to CVS to pull out different versions till you find the exact code
change that caused the problem. For example:

cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot co -D "9/1/98" samba/source
... compile and test in samba/source ... note result ...
rm -rf samba/source
cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot co -D "9/20/98" samba/source
... compile and test in samba/source ... note result ...

then use a bisection search on the date to find the day (and even
hour/minute) that the code was broken. You'll probably find it in a
few tests. Then use diff to see exactly what bit of code caused the
problem.

The advantage of this method is that it is almost completely
automated. You don't need to know anything about how the code works to
find the bug!

From mhaigh at village.vut.edu.au  Wed Sep 23 06:45:08 1998
From: mhaigh at village.vut.edu.au (Mick Haigh)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au>
Message-ID: <360898F4.8A8BFA93@village.vut.edu.au>

Andrew Tridgell wrote:

> > Something similar to this has been posted to the samba group a few
> > times, but I didn't see any replys.  Anyway - the last few revisions of
> > the code that I've checked out (including one from 12:11pm Sept 23 EST)
> > have broken printing from Win95 machines.
>
> It's not as simple as that. Printing works fine from Win95 for me (my
> wife uses it all the time).
>
> > There was an error writing to \\village\laser for printer (HP LaserJet
> > 4M Plus):
> > There was a problem printing to the printer due to an unknown system
> > error.
> > Restart Windows, and the try printer again.
>
> you may have to do a comparitive sniff to work out what's
> wrong. Download a earlier version (one that works) then get a sniff or
> log of that printing. Then use the current version and get another
> log. Then look at them to find the first point at which they diverge.
> Then stare at the surrounding packets till you find what the
> difference is in how Samba reponded.
>
> Alternatively (if you don't like staring at sniffs) use the -D option
> to CVS to pull out different versions till you find the exact code
> change that caused the problem. For example:
>
> cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot co -D "9/1/98" samba/source
> .. compile and test in samba/source ... note result ...
> rm -rf samba/source
> cvs -d :pserver:cvs@samba.anu.edu.au:/cvsroot co -D "9/20/98" samba/source
> .. compile and test in samba/source ... note result ...

Ok - I've done this for the time being.  I couldn't (easily) get the source
from 9/19/98 to build, however the source from th 18th worked fine, but the
source from the 20th didn't.

I'll have a bit more of a look tomorrow.

Cheers.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: mhaigh.vcf
Type: text/x-vcard
Size: 201 bytes
Desc: Card for Mick Haigh
Url : http://lists.samba.org/archive/samba-technical/attachments/19980923/a3665fcc/mhaigh.vcf
From tridge at samba.anu.edu.au  Wed Sep 23 08:00:14 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
In-Reply-To: <360898F4.8A8BFA93@village.vut.edu.au> (message from Mick Haigh
	on Wed, 23 Sep 1998 16:41:00 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au>
Message-ID: <19980923080019Z12587999-20449+10428@samba.anu.edu.au>

> Ok - I've done this for the time being.  I couldn't (easily) get the source
> from 9/19/98 to build, however the source from th 18th worked fine, but the
> source from the 20th didn't.

unfortunately there were quite a few changes from the 18th to the
20th. 

This is the diffstat info:

 client.c               |  254 +++++++++++++++++++++--
 clitar.c               |  287 +++++++++++++++++++++-----
 config.h.in            |   23 ++
 configure              |  540 ++++++++++++++++++++++++++++++-------------------
 configure.in           |   32 ++
 dfree.c                |   44 +--
 fileio.c               |   22 -
 filename.c             |   16 +
 includes.h             |   34 +++
 ipc.c                  |   24 --
 loadparm.c             |   30 ++
 negprot.c              |   17 -
 nttrans.c              |   44 +--
 open.c                 |   85 ++++---
 oplock.c               |   60 +++++
 proto.h                |   22 +
 reply.c                |   38 ---
 smb.h                  |   18 +
 smbpass.c              |   14 -
 system.c               |   32 ++
 trans2.c               |   75 ++----
 util.c                 |   33 --

you can ignore the client changes. 

To try to narrow it own further I'd take a 20th source tree and
convert it file by file back to a 18th tree, testing after each
file. I think that will be possible without too much hacking. (going
the other way would be very tricky!).

you could also test after every few files, which would save you some
testing and allow you to narrow it down to 1 file in just a few
tests. (if you are lucky).

In case you don't know, you can get a complete diff from the 18th to
the 20th by doing:

    cvs diff -u -D "9/18/98" -D "9/20/98" .

in the tree you want to look at (usually in a current
tree). Unfortunately this produces 4696 lines of diff, which is why we
need to narrow it down a bit more.

Cheers, Tridge

PS: Thanks for taking the time to help us track this down!
From lkcl at switchboard.net  Wed Sep 23 21:13:10 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: NT Server Manager..
In-Reply-To: <Pine.NEB.3.96.980923160118.8514A-100000@cb1-gw.cb1.com>
Message-ID: <Pine.NEB.3.96.980923221032.18607B-100000@cb1-gw.cb1.com>

On Thu, 24 Sep 1998, Luke Kenneth Casson Leighton wrote:

> >  Its amazing the responses i have gotten from my original 
> > posting. Many keep saying you cant use NT Server Manager 
> > and most seem to think i am refering to "NT User Manager" 
> > well the fact is that in 1.19 you could use both of them 
> 
> you are absolutely correct.  _why_ doesn't it work???  not good enough!
> i'm on it, ok?

ok.

\PIPE\lsarpc is opened simultaneously as \PIPE\samr.

samr is opened; calls to samr are made; lsarpc is opened; lsarpc calls are
made. 

then, a samr call is made, and samba responds "INVALID_PIPE_HANDLE".
oops.

the client responds by closing samr, and later it closes lsarpc.

that's the techie symptoms; now to track down the problem (corruption of
the pipe handles?)

From lkcl at switchboard.net  Wed Sep 23 21:39:14 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: NT Server Manager..
In-Reply-To: <Pine.NEB.3.96.980923221032.18607B-100000@cb1-gw.cb1.com>
Message-ID: <Pine.NEB.3.96.980923223811.18607C-100000@cb1-gw.cb1.com>

On Thu, 24 Sep 1998, Luke Kenneth Casson Leighton wrote:

> On Thu, 24 Sep 1998, Luke Kenneth Casson Leighton wrote:
> 
> > >  Its amazing the responses i have gotten from my original 
> > > posting. Many keep saying you cant use NT Server Manager 
> > > and most seem to think i am refering to "NT User Manager" 
> > > well the fact is that in 1.19 you could use both of them 
> > 
> > you are absolutely correct.  _why_ doesn't it work???  not good enough!
> > i'm on it, ok?
> 
> ok.
> 
> \PIPE\lsarpc is opened simultaneously as \PIPE\samr.
> 
> samr is opened; calls to samr are made; lsarpc is opened; lsarpc calls are
> made. 
> 
> then, a samr call is made, and samba responds "INVALID_PIPE_HANDLE".
> oops.
> 
> the client responds by closing samr, and later it closes lsarpc.
> 
> that's the techie symptoms; now to track down the problem (corruption of
> the pipe handles?)

got it :-)

in the pipe handle code, ZERO_STRUCT(p) needed to come _before_
DLIST_ADD(Pipes, p).


From sharpe at ns.aus.com  Thu Sep 24 05:01:30 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: SMBcreate under Samba lower-cases names, it seems
Message-ID: <3.0.5.32.19980924140130.00888100@mail.adelaide.on.net>

During testing of my recent changes to clitar, I tested against Samba.

I used smbclient to create a tar file, and then restored it to a Samba server.

All uppercase characters in file names were case folded to lowercase
characters when the file was created with SMBcreat calls.

Is this the correct behaviour?  I would have hoped that the case would be
preserved on an OS that could do that.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Thu Sep 24 04:59:59 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: SMBcreate under Samba lower-cases names, it seems
In-Reply-To: <3.0.5.32.19980924140130.00888100@mail.adelaide.on.net> (message
	from Richard Sharpe on Thu, 24 Sep 1998 14:35:09 +1000)
References: <3.0.5.32.19980924140130.00888100@mail.adelaide.on.net>
Message-ID: <19980924050009Z12670678-25139+10802@samba.anu.edu.au>

> All uppercase characters in file names were case folded to lowercase
> characters when the file was created with SMBcreat calls.
> 
> Is this the correct behaviour?  I would have hoped that the case would be
> preserved on an OS that could do that.

it is correct only if you don't have the "preserve case" option
set. For Samba 2.0 the default is to preserve case. In 1.9.18 the
defaults were different.

I just tested this and 2.0 does seem to do things right, so I think
your smb.conf might be the problem.
From jallison at cthulhu.engr.sgi.com  Fri Sep 18 21:29:38 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:12 2003
Subject: CVS update: samba/source/client
References: <19980918124756Z12666551-7510+8896@samba.anu.edu.au> <36029C52.4AE27FEE@engr.sgi.com> <orzpbxqgox.fsf@araguaia.dcc.unicamp.br>
Message-ID: <3602D0C2.ED6F3C00@engr.sgi.com>

Alexandre Oliva wrote:

> Actually, adding anything to include/config.h.in is just a waste of
> time, since it is automatically generated from acconfig.h by
> autoheader. :-)
> 

Thanks for that - I missed that (I've been adding to config.h.in
in error :-). Guess I should read the comments :-).

I'll go and fix this properly now :-).

> BTW, the patch I've recently posted to this list contains a lot of
> fixes of autoconf-related problems, so it would be nice to have it
> installed soon, at least to help prevent further addition of incorrect
> code to configure.in :-)

I'm not sure I've seen that patch. When did you post it
and was it to samba-tech ?

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From oliva at dcc.unicamp.br  Thu Sep 24 03:54:26 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:12 2003
Subject: CVS update: samba/source/client
In-Reply-To: Jeremy Allison's message of "Fri, 18 Sep 1998 14:29:38 -0700"
References: <19980918124756Z12666551-7510+8896@samba.anu.edu.au> <36029C52.4AE27FEE@engr.sgi.com> <orzpbxqgox.fsf@araguaia.dcc.unicamp.br> <3602D0C2.ED6F3C00@engr.sgi.com>
Message-ID: <or4stym259.fsf@araguaia.dcc.unicamp.br>

Jeremy Allison <jallison@cthulhu.engr.sgi.com> writes:

>> BTW, the patch I've recently posted to this list contains a lot of
>> fixes of autoconf-related problems

> I'm not sure I've seen that patch.

Never mind, I've installed it myself.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From Craig.Bridge at seagatesoftware.com  Thu Sep 24 10:19:37 1998
From: Craig.Bridge at seagatesoftware.com (Craig Bridge)
Date: Tue Dec  2 03:02:12 2003
Subject: toupper case grief with Windows NT5 Beta 2 and Samba1.9.18p10-51.
Message-ID: <D0633F10124BD211B9C90000947B08F901F15D@lmoxch02.smg.seagatesoftware.com>

I'm having toupper case grief with Windows NT 5 Beta 2 created files on
Red Hat Samba 1.9.18p10-51.1 that I don't have with Windows NT 4.

I'm willing to get down and dirty in the code and have the resources and
general background to do so. I'd just like some _BRIEF_ guidance from
the SAMBA development team on where to set key breakpoints, any
sniffer tools, debug output, and what to look for on the SAMBA side.

Direct resposes to <cbridge@seagatesoftware.com>, or pointing me to
another forum, etc may be more appropriate, I'm new here... but bring
considerable NT kernel mode debugger experience.

NT 4 is cool with SAMBA shares with smb.conf parameters:
   preserve case = yes
   short preserve case = yes
   default case = lower

I was fairly succesful with:
   case sensitive = yes
and created "junk", "JUNK", and "JuNk" files that explorer shows as
"junk", "Junk" (explorer convention for display of all uppercase names,
and "JuNk" as expected).  However, NT 4's matcher doesn't always get
the correct match the first time depending on what is in the directory
cache.  This is something that the NFS on NT solutions solved!
I'll investigate futher if time permits.

Note: with "case sensitive = no", you can't create the 3 varients from
NT
on the SAMBA share, so it doesn't come close to an NFS solution in
exposing a full Unix file system name space.

NT 5 Beta 2 touppers everything with either of the above configurations;
however, NT 5 and NT 4 play together without this grief using SMB.

I'd _REALLY_ like to figure out why NT 4 and NT 5 Beta 2 aren't
handling SAMBA the same. I don't want NT 5 to force me to live with:
  preserve case = no
  short preserve case = no
  default case = lower

Thanks,
-Craig


From sharpe at ns.aus.com  Thu Sep 24 14:22:38 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: clitar restore rewrite 
Message-ID: <3.0.5.32.19980924232238.008666d0@mail.adelaide.on.net>

Hi Folks,

I have been rewriting the restore side of clitar to make it more readable
and hopefully, easier to maintain.

I have done a lot of testing of this code but need more testing from people.

Here is what I have done:

- restored files to a samba-based share and then did a diff agains the
originals to find no changes

- restored files to a Win95 machine to ensure that file sizes and dates are
correct.

Win95 objects to a change date request, but seems to change the date as
requested anyway.  I will have to use netmon to check this out.

What I have not done yet is a diff of the files under Win95 to check if
they are the same. Is there a diff command?

Also, I would like more people to check the restore function to ensure that
it works correctly.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Thu Sep 24 15:39:50 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: toupper case grief with Windows NT5 Beta 2 and Samba1.9.18p10-51.
In-Reply-To: 
	<D0633F10124BD211B9C90000947B08F901F15D@lmoxch02.smg.seagatesoftware.com>
	(message from Craig Bridge on Thu, 24 Sep 1998 20:23:21 +1000)
References: <D0633F10124BD211B9C90000947B08F901F15D@lmoxch02.smg.seagatesoftware.com>
Message-ID: <19980924153956Z12670774-7009+10852@samba.anu.edu.au>

> I was fairly succesful with:
>    case sensitive = yes

that is normally not a good idea, unless you know that all your
applications handle a case sensitive filesystem. (case sensitivity is
largely a application issue, not a OS issue. many apps change case
arbitrarily) 

> and created "junk", "JUNK", and "JuNk" files that explorer shows as
> "junk", "Junk" (explorer convention for display of all uppercase names,
> and "JuNk" as expected).  However, NT 4's matcher doesn't always get
> the correct match the first time depending on what is in the directory
> cache.  This is something that the NFS on NT solutions solved!

first off, find out if what you are trying to do works with a NT->NT
connection over SMB. If it does then Samba can probably be made to do
the same thing. If it doesn't then it is quite likely a client issue
(and thus something we can't do anything about).

> Note: with "case sensitive = no", you can't create the 3 varients
> from NT on the SAMBA share, so it doesn't come close to an NFS
> solution in exposing a full Unix file system name space.

the reason we default to "case sensitive = no" is that many apps can't
handle it otherwise. If you look witha sniffer at what common apps
(such as ms office) do you will find that they change the case of
files before sending the names back to the server. That makes it
impossible for it to work right with many apps no matter what Samba
does.
 
> NT 5 Beta 2 touppers everything with either of the above configurations;
> however, NT 5 and NT 4 play together without this grief using SMB.

that is interesting. If NT->NT over SMB does what you want then we can
probably make Samba do the same. You'll need to start using a SMB
sniffer (such as tcpdump-smb) to work out what is different about how
Samba and NT handle things. 

Note that the sniffs will be quite different (because NT will use
unicode) but I doubt this is the cause of the problem. unicode and
case sensitivity should be orthogonal issues.

Hmmm, I wonder if any apps look at the file system type (the fstype
string) to determine case behaviour? Might be worth trying with
FSTYPE_STRING set to "NTFS".

Cheers, Tridge
From tridge at samba.anu.edu.au  Fri Sep 25 03:49:37 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
In-Reply-To: <360AEA1E.C2E7CAE8@village.vut.edu.au> (message from Mick Haigh
	on Fri, 25 Sep 1998 10:55:58 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au>
Message-ID: <19980925034944Z12616741-25139+11120@samba.anu.edu.au>

I think I might have foun the problem. In reply_getatr() we have this
comment:

  unix_convert(fname,conn,0,&bad_path,&sbuf);

  /* dos smetimes asks for a stat of "" - it returns a "hidden directory"
     under WfWg - weird! */
  if (! (*fname))
    {
      mode = aHIDDEN | aDIR;
      if (!CAN_WRITE(conn)) mode |= aRONLY;
      size = 0;
      mtime = 0;
      ok = True;
    }
  else {
  .....
}

the new unix_convert() code will break that as it will never return a
null filename.

Jeremy, how do you want to fix this? Change unix_convert() to allow
null filenames?

From tridge at samba.anu.edu.au  Fri Sep 25 03:49:37 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
In-Reply-To: <360AEA1E.C2E7CAE8@village.vut.edu.au> (message from Mick Haigh
	on Fri, 25 Sep 1998 10:55:58 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au>
Message-ID: <19980925034944Z12597158-7009+11097@samba.anu.edu.au>

> Ok - at this stage I've got it narrowed down to a change in either
> fileio.c or filename.c.  I'll have a more in depth look at the
> changes this afternoon.

the only fileio.c changed was a cleanup of the use of mmap(), which
doesn't affect printing and is only used if you compiled with
--with-mmap anyway. So it must be filename.c

In filename.c the diffs were:


    --- samba/source/smbd/filename.c        1998/09/10 00:35:09     1.13
    +++ /cvsroot/samba/source/smbd/filename.c       1998/09/19 03:34:12     1.14
    @@ -150,7 +150,8 @@
       /*
	* Don't cache trivial valid directory entries.
	*/
    -  if((strcmp(full_orig_name, ".") == 0) || (strcmp(full_orig_name, "..") == 0))
    +  if((*full_orig_name == '\0') || (strcmp(full_orig_name, ".") == 0) ||
    +     (strcmp(full_orig_name, "..") == 0))
	 return;

       /*
    @@ -245,7 +246,7 @@
       /*
	* Don't lookup trivial valid directory entries.
	*/
    -  if((strcmp(name, ".") == 0) || (strcmp(name, "..") == 0)) {
    +  if((*name == '\0') || (strcmp(name, ".") == 0) || (strcmp(name, "..") == 0)) {
	 global_stat_cache_misses++;
	 return False;
       }
    @@ -364,6 +365,17 @@
	*/

       trim_string(name,"/","/");
    +
    +  /*
    +   * If we trimmed down to a single '\0' character
    +   * then we should use the "." directory to avoid
    +   * searching the cache.
    +   */
    +
    +  if(!*name) {
    +    name[0] = '.';
    +    name[1] = '\0';
    +  }

       /*
	* Ensure saved_last_component is valid even if file exists.


These changes were made by Jeremy as part of his optimisation
efforts. I suspect that the printing stuff somehow relies on the old
behaviour of this code.
    
The changes are to three functions:
    
1) stat_cache_add()
1) stat_cache_lookup()
2) unix_convert()
    
The bug is almost certainly in the unix_convert() change. Can you
please confirm this by setting "stat cache = no" in smb.conf? That
will disable the other two functions.

If it is unix_convert() then please try removing these lines:

      if(!*name) {
        name[0] = '.';
        name[1] = '\0';
      }

and see if it fixes it. If it does then we just need to work out which
call to this is the problem ...
    
From sharpe at ns.aus.com  Fri Sep 25 04:21:34 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: Heaps of errors in the current 2.0.0 code :-(
Message-ID: <3.0.5.32.19980925132134.0087f9a0@mail.adelaide.on.net>

Hi,

someone has checked in some code that seems to give me heaps of errors now,
like:

>smbd/reply.c: In function `reply_lockingX':
>smbd/reply.c:3547: warning: `count' might be used uninitialized in this
function
>smbd/reply.c:3547: warning: `offset' might be used uninitialized in this 
>function



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Fri Sep 25 03:59:40 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
In-Reply-To: <360AEA1E.C2E7CAE8@village.vut.edu.au> (message from Mick Haigh
	on Fri, 25 Sep 1998 10:55:58 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au>
Message-ID: <19980925035940Z12597158-7510+11159@samba.anu.edu.au>

Mick,

I've commited a change that might fix things (assuming we have
tracked it down to the right code). Can you let me kow if it works
now?
From sharpe at ns.aus.com  Fri Sep 25 04:28:14 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: windiff shows no differences with files restored by clitar.c
Message-ID: <3.0.5.32.19980925132814.0088d530@mail.adelaide.on.net>

Hi,

Well, with the help of a couple of people I discovered windiff and used it
to check out restored files.

So far, I have found no differences ...

looks like the new clitar code can be declared to be working ...

Now to document a bunch of changes that have not yet been documented.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From borsenkow.msk at sni.de  Fri Sep 25 07:09:24 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:12 2003
Subject: SAMBA cannot authenticate against NT4.0SP4-rc1.99
Message-ID: <000f01bde853$6cd749e0$21c9ca95@ao13.mowp.siemens.ru>

For some reasons our guys who are in charge of NT server installed
SP4-rc1.99 on WinNT4 servers. With effect, that now samba-2.0.0-alpha7
running with security = domain cannot authenticate itself any more in
domain. I tried to remove it from domain and add back - all results in the
same error message:

domain_client_validate: unable to open the domain client session to machine
ITS_SERVER. Error was : code 0.

I attach zip with both raw cap file and pretty printout. It is between
its_app (BDC) and itsrm2 (SAMBA) doing

smbpasswd -j ITS -r ITS_APP

The same happens with PDC; I just happen to have netmon on BDC :))

Or, yes, machine sid is

S-1-5-21-4291449084-786989410-2010929659

/Andrej
-------------- next part --------------
A non-text attachment was scrubbed...
Name: nt4sp4.zip
Type: application/x-zip-compressed
Size: 17130 bytes
Desc: not available
Url : http://lists.samba.org/archive/samba-technical/attachments/19980925/360f3e8f/nt4sp4.bin
From borsenkow.msk at sni.de  Fri Sep 25 07:24:16 1998
From: borsenkow.msk at sni.de (Andrej Borsenkow)
Date: Tue Dec  2 03:02:12 2003
Subject: SAMBA cannot authenticate against NT4.0SP4-rc1.99
In-Reply-To: <000f01bde853$6cd749e0$21c9ca95@ao13.mowp.siemens.ru>
Message-ID: <001301bde855$80b0da60$21c9ca95@ao13.mowp.siemens.ru>

Just to add, that security = server works fine with server set to PDC. It is
just security = domain that is broken.

/Andrej


From lkcl at switchboard.net  Fri Sep 25 15:00:52 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: Heaps of errors in the current 2.0.0 code :-(
In-Reply-To: <3.0.5.32.19980925132134.0087f9a0@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980925155833.1678K-100000@cb1-gw.cb1.com>

it's probably caused by the #ifdef LARGE_SMB_OFF_T, and unless
large_file_format is always false, then yes: the warning you see below is
absolutely correct!

On Fri, 25 Sep 1998, Richard Sharpe wrote:

> Hi,
> 
> someone has checked in some code that seems to give me heaps of errors now,
> like:
> 
> >smbd/reply.c: In function `reply_lockingX':
> >smbd/reply.c:3547: warning: `count' might be used uninitialized in this
> function
> >smbd/reply.c:3547: warning: `offset' might be used uninitialized in this 
> >function
> 
> 
> 
> Regards
> -------
> Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
> NS Computer Software and Services P/L, 
> Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
> Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From jallison at cthulhu.engr.sgi.com  Fri Sep 25 16:48:07 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:12 2003
Subject: printing from Win95
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au> <19980925034944Z12616741-25139+11120@samba.anu.edu.au>
Message-ID: <360BC947.2DFE0228@engr.sgi.com>

Andrew Tridgell wrote:

> the new unix_convert() code will break that as it will never return a
> null filename.
> 
> Jeremy, how do you want to fix this? Change unix_convert() to allow
> null filenames?

Actually the code I added was simply a "paranioa" check.
The code actually already worked ok without it.

Of course I never realized something actually *depended*
on unix_convert() returning a pointer to a null byte :-).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From lkcl at switchboard.net  Fri Sep 25 19:32:32 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:12 2003
Subject: rpcclient
Message-ID: <Pine.NEB.3.96.980925203032.1678c-100000@cb1-gw.cb1.com>

to get round the problems of merging smbclient dce/rpc code (from
BRANCH_NTDOM) into the current cvs tree, i've instead created a
"rpcclient" command.

its first duty was to do a "LsaQueryInfoPolicy" at levels 3 & 5, with its
only command (lsaquery) which it passed with flying colours.

hooray.

From sharpe at ns.aus.com  Sat Sep 26 00:47:01 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: Still problems with compiling samba with Jeremy's flags
Message-ID: <3.0.5.32.19980926094701.00866a40@mail.adelaide.on.net>

Hi,

because I did a make clean and then a make with Jeremy's flags, I have
found a number of places where there are still problems.

I have fixed all the cases where variables like count and offset are
declared but not initialized in the declaration.

However, I also found one place (passdb/password_check.c) where gcc
complained about crypt having an implicit declaration.

I don't seem able to fix this ... have tried #define _XOPEN_SOURCE but that
does not work.



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 26 01:00:14 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: Source tree currently broken?
Message-ID: <3.0.5.32.19980926100014.0087c530@mail.adelaide.on.net>

Hmmm,

can't link the source tree as someone has introduced smblib/pwd_cache.c and
forgot to tell Makefile.in about it?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From sharpe at ns.aus.com  Sat Sep 26 01:14:28 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: Source tree currently broken?
In-Reply-To: <3.0.5.32.19980926100014.0087c530@mail.adelaide.on.net>
Message-ID: <3.0.5.32.19980926101428.0087c520@mail.adelaide.on.net>

At 10:37 AM 9/26/98 +1000, I wrote:
>Hmmm,
>
>can't link the source tree as someone has introduced smblib/pwd_cache.c and
>forgot to tell Makefile.in about it?

Spoke too soon ... Picked up the latest updates and reran configure :-(



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Sat Sep 26 01:10:34 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:12 2003
Subject: Still problems with compiling samba with Jeremy's flags
In-Reply-To: <3.0.5.32.19980926094701.00866a40@mail.adelaide.on.net> (message
	from Richard Sharpe on Sat, 26 Sep 1998 10:25:10 +1000)
References: <3.0.5.32.19980926094701.00866a40@mail.adelaide.on.net>
Message-ID: <19980926011044Z12670712-25139+11437@samba.anu.edu.au>

> However, I also found one place (passdb/password_check.c) where gcc
> complained about crypt having an implicit declaration.

I've seen that and I think that it is a bug in the glibc headers. I
haven't bothered to put in a workaround yet.


From jallison at cthulhu.engr.sgi.com  Sat Sep 26 01:17:59 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:12 2003
Subject: Still problems with compiling samba with Jeremy's flags
References: <3.0.5.32.19980926094701.00866a40@mail.adelaide.on.net>
Message-ID: <360C40C7.424328F0@engr.sgi.com>

Richard Sharpe wrote:

> I have fixed all the cases where variables like count and offset are
> declared but not initialized in the declaration.
> 

Ah - I see what that is now.

I am compiling on a system with 64 bit file support.
Now the code path taken through the locking code
could lead to uninitialized variables being used
if the client sends a large file locking request
but the server doesn't support 64 bit files.

However the bit that the server uses to tell
the client that 64 bit requests are ok is only set
if the server supports large files.

I need to re-structure the code slightly to
remove this problem ultimately, but setting
the count & offset to zero is probably good enough
(as it really would be undefined behavior if the server told
the client it couldn't support large file requests and
the client sent them anyway :-).

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From sharpe at ns.aus.com  Sat Sep 26 02:08:59 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:12 2003
Subject: Still build problems with Samba
Message-ID: <3.0.5.32.19980926110859.00867220@mail.adelaide.on.net>

Hi,

here is the latest problem

>make: *** No rule to make target `rpcclient/rpcclient.o', needed by 
>`bin/rpcclient'.  Stop.


Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From jallison at cthulhu.engr.sgi.com  Sat Sep 26 02:39:39 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:12 2003
Subject: Still build problems with Samba
References: <3.0.5.32.19980926110859.00867220@mail.adelaide.on.net>
Message-ID: <360C53EB.F59C941D@engr.sgi.com>

Richard Sharpe wrote:
> 
> Hi,
> 
> here is the latest problem
> 
> >make: *** No rule to make target `rpcclient/rpcclient.o', needed by
> >`bin/rpcclient'.  Stop.

CVS checkout again. It builds ok here (and Andrew also
reports it will build).

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From james at cows.ml.org  Sat Sep 26 04:00:19 1998
From: james at cows.ml.org (James Willard)
Date: Tue Dec  2 03:02:13 2003
Subject: WINS/DC Utilities
Message-ID: <199809260400.AAA12489@cows.ml.org>


   I have not been able to find any kind of WINS or Domain Controller test
utilities for Linux. I am wanting to be able to see exactly what is going on
in the network such as finding the PDC or viewing a WINS database. Can anyone
suggest utilities that would help me with this?

James Willard
james@cows.ml.org

-- 
From mhaigh at village.vut.edu.au  Sat Sep 26 08:34:37 1998
From: mhaigh at village.vut.edu.au (Mick Haigh)
Date: Tue Dec  2 03:02:13 2003
Subject: printing from Win95
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au> <19980925035940Z12597158-7510+11159@samba.anu.edu.au>
Message-ID: <360CA71C.9FB61357@village.vut.edu.au>

Andrew Tridgell wrote:

> Mick,
>
> I've commited a change that might fix things (assuming we have
> tracked it down to the right code). Can you let me kow if it works
> now?

Sorry I took so long to reply.  Had a major equipment failure and 650
students jumping down my throat :(

Anyway - seems to work like a dream.  Just a question - why would you
(your wife??) still have been able to print??

Mick

From tridge at samba.anu.edu.au  Sat Sep 26 10:00:00 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: printing from Win95
In-Reply-To: <360CA71C.9FB61357@village.vut.edu.au> (message from Mick Haigh
	on Sat, 26 Sep 1998 18:34:37 +1000)
References: <36085B4E.3173CCF1@village.vut.edu.au> <19980923025025Z12639646-7009+10376@samba.anu.edu.au> <360898F4.8A8BFA93@village.vut.edu.au> <19980923080019Z12587999-20449+10428@samba.anu.edu.au> <360AEA1E.C2E7CAE8@village.vut.edu.au> <19980925035940Z12597158-7510+11159@samba.anu.edu.au> <360CA71C.9FB61357@village.vut.edu.au>
Message-ID: <19980926100006Z12670695-20449+11459@samba.anu.edu.au>

> Anyway - seems to work like a dream.  Just a question - why would you
> (your wife??) still have been able to print??

The problem wasn't actually a printing problem, it was that Samba no
longer responded to the (rather unusual) stat("") with "hidden
directory", which is a myterious requirement that we discovered years
ago.

Obviously Sue's machine either didn't make that request for some
reason or didn't mind when it failed. It's probably some side effect
of how she was connected to my box (win95 domain logon etc).



From sharpe at ns.aus.com  Mon Sep 28 04:37:43 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:13 2003
Subject: How to get lots of compile toime checks now?
Message-ID: <3.0.5.32.19980928133743.0089ada0@mail.adelaide.on.net>

Hi,

with the latest changes to Samba 2.0.0, I no longer seem able to pass
things like -Wall -Wstrict-prototypes etc to the compiles by editing the
Makefile that is generated by configure.

How do I accomplish this?



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From tridge at samba.anu.edu.au  Mon Sep 28 04:40:17 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: How to get lots of compile toime checks now?
In-Reply-To: <3.0.5.32.19980928133743.0089ada0@mail.adelaide.on.net> (message
	from Richard Sharpe on Mon, 28 Sep 1998 14:11:24 +1000)
References: <3.0.5.32.19980928133743.0089ada0@mail.adelaide.on.net>
Message-ID: <19980928044028Z12616741-7009+11911@samba.anu.edu.au>

> with the latest changes to Samba 2.0.0, I no longer seem able to pass
> things like -Wall -Wstrict-prototypes etc to the compiles by editing the
> Makefile that is generated by configure.
> 
> How do I accomplish this?

I have a script called "config" that contains the following:

#!/bin/sh
export CFLAGS="-g -Wall -Wshadow -Wstrict-prototypes -Wpointer-arith -Wcast-qual"
./configure
From sharpe at ns.aus.com  Mon Sep 28 06:02:49 1998
From: sharpe at ns.aus.com (Richard Sharpe)
Date: Tue Dec  2 03:02:13 2003
Subject: Remaining compile problems in Samba 2.0.0
Message-ID: <3.0.5.32.19980928150249.00894bd0@mail.adelaide.on.net>

Hi,

these are the remaining compilation warning problems in Samba 2.0.0 when I
compile with the following options:

>Using FLAGS = -g -O2 -Wall -Wshadow -Wstrict-prototypes -Wpointer-arith -
>Compiling passdb/pass_check.c
>passdb/pass_check.c: In function `password_check':
>passdb/pass_check.c:739: warning: implicit declaration of function `crypt'
>Compiling client/client.c
>client/client.c: In function `process':
>client/client.c:3655: warning: declaration of `prompt' shadows global
declaration



Regards
-------
Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
NS Computer Software and Services P/L, 
Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
From cartegw at Eng.Auburn.EDU  Sun Sep 27 21:32:40 1998
From: cartegw at Eng.Auburn.EDU (Gerald Carter)
Date: Tue Dec  2 03:02:13 2003
Subject: WINS/DC Utilities
References: <199809260400.AAA12489@cows.ml.org>
Message-ID: <360EAEF8.1E7A084@eng.auburn.edu>

James Willard wrote:
> 
> I have not been able to find any kind of WINS 
> or Domain Controller test utilities for Linux. I 
> am wanting to be able to see exactly what is going 
> on in the network such as finding the PDC or viewing 
> a WINS database. Can anyone suggest utilities that 
> would help me with this?

MS's Network Monitor is very good for viewing traffic on the
wire.  

For WINS I am guessing you are talking about an NT WINS
server.  No idea here unless you right some programs to query
the JET database format.

Hope this helps,
j- 
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )


From cartegw at Eng.Auburn.EDU  Mon Sep 28 13:47:33 1998
From: cartegw at Eng.Auburn.EDU (Gerald Carter)
Date: Tue Dec  2 03:02:13 2003
Subject: LFN diaply in DOS window on NT
Message-ID: <360F9375.3AE70FAF@eng.auburn.edu>

Andrew,

I went back and tried to do something like

	type buildprofile

from a DOS windows under NT against CVS coded dated Aug 31 as 
well as against 1.9.18p7 and it worked.  I checked the FS type 
and it reported 'Samba'.

Just noting this in response to the recent change from FS type 
Samba to NTFS.  Not sure if it really matters.  



j- 
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )
From tridge at samba.anu.edu.au  Mon Sep 28 14:00:15 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: LFN diaply in DOS window on NT
In-Reply-To: <360F9375.3AE70FAF@eng.auburn.edu> (message from Gerald Carter on
	Mon, 28 Sep 1998 23:48:51 +1000)
References: <360F9375.3AE70FAF@eng.auburn.edu>
Message-ID: <19980928140015Z12597176-20449+12000@samba.anu.edu.au>

> I went back and tried to do something like
> 
> 	type buildprofile
> 
> from a DOS windows under NT against CVS coded dated Aug 31 as 
> well as against 1.9.18p7 and it worked.  I checked the FS type 
> and it reported 'Samba'.

yes. The difference I noted (that prompted me to make the change) was
in directory listings. My wifes Win95 box didn't show long names in a
"dir" without the fstype being set to NTFS or HPFS.

From lkcl at switchboard.net  Mon Sep 28 14:04:31 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: Source tree currently broken?
In-Reply-To: <3.0.5.32.19980926100014.0087c530@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980928150344.8771E-100000@cb1-gw.cb1.com>

que??  no, it's there, honest.

On Sat, 26 Sep 1998, Richard Sharpe wrote:

> Hmmm,
> 
> can't link the source tree as someone has introduced smblib/pwd_cache.c and
> forgot to tell Makefile.in about it?
> 
> 
> 
> Regards
> -------
> Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
> NS Computer Software and Services P/L, 
> Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
> Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Mon Sep 28 14:15:17 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: Still build problems with Samba
In-Reply-To: <3.0.5.32.19980926110859.00867220@mail.adelaide.on.net>
Message-ID: <Pine.NEB.3.96.980928151509.8771H-100000@cb1-gw.cb1.com>

it's there, i promise!

On Sat, 26 Sep 1998, Richard Sharpe wrote:

> Hi,
> 
> here is the latest problem
> 
> >make: *** No rule to make target `rpcclient/rpcclient.o', needed by 
> >`bin/rpcclient'.  Stop.
> 
> 
> Regards
> -------
> Richard Sharpe, sharpe@ns.aus.com, NIC-Handle:RJS96
> NS Computer Software and Services P/L, 
> Ph: +61-8-8281-0063, FAX: +61-8-8250-2080, 
> Samba, Linux, Apache, Digital UNIX, AIX, Netscape, Stronghold, C, ...
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Mon Sep 28 14:20:59 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: WINS/DC Utilities
In-Reply-To: <199809260400.AAA12489@cows.ml.org>
Message-ID: <Pine.NEB.3.96.980928152029.8771J-100000@cb1-gw.cb1.com>

well i just created a new command called "rpcclient".  and you can use
nmblookup for wins.

On Sat, 26 Sep 1998, James Willard wrote:

> 
>    I have not been able to find any kind of WINS or Domain Controller test
> utilities for Linux. I am wanting to be able to see exactly what is going on
> in the network such as finding the PDC or viewing a WINS database. Can anyone
> suggest utilities that would help me with this?
> 
> James Willard
> james@cows.ml.org
> 
> -- 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From mhw at wittsend.com  Mon Sep 28 14:57:51 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <E0zJBss-00003I-00@delphin.SerNet.DE> from "Volker.Lendecke@SerNet.DE" at "Sep 16, 1998  5:22: 6 pm"
Message-ID: <199809281457.KAA15937@alcove.wittsend.com>

Hello!

	My original reply to this message appears to have been eaten by
a wayward IPv6 enabled version of sendmail I was putzing with at the time...
Since my status with the Samba team is changing and it looks like I'm going
to end up more involved with this...  Here we go again...

	I have mention of what appears to be an smbfs kernel module problem
later on in this message, so I'm also including the kernel developer list
in this.  Volker may fix the kernel module himself, or he may point me toward
it, or someone else may take on the task...

Volker.Lendecke@SerNet.DE enscribed thusly:

> Hello!

> > 	I'm also one of the primary bitchers about the change in syntax
> > from the old smbfs version of smbmount to the new samba version of smbmount.
> > I'm not going to go down the road of why it got where it is or why it should
> > or shouldn't have.  We have users that are heavily invested in it and we
> > have some responsibility to the sins of the past...  So be it.

> Is it only the syntax that annoys you or is it also the kernel-level
> interface? I must admit that the syntax is really ugly, but it was the
> fastest way to get something use the kernel interface. IMO this kernel
> interface is necessary, as I did not want all the fancy stuff like
> netbios name lookup or encrypted passwords in the kernel. The syntax
> could be improved, making the 2.1 stuff look very similar to the 2.0
> smbmount and still use smbclient to do the authentication.

	It's primarily the syntax change.  The syntax change broke the
bejesus out of autofs, and any other scripts that thought they understood
smbmount, and to call the fix "less than pretty" would be paying it high
praise.  One optional parameter became positional while several optional
parameters got combined with one positional parameter and subsumed into a
mount command string single parameter!

	This is compounded by the problems introduced by many of us who
switch back and forth between 2.0 and 2.1 kernels for testing.  Due to
the change in the kernel interface, we are stuck with two binaries anyways.
But we also end up with two incompatible command syntaxes to deal with as
well.  So, now, what should have been understood and buried in one command
(even if it resolves to two binaries) requires anyone and anything which
calls it to also understand and switch syntax based on the conditions that
would switch the binaries.  It would have been nice to support backward
compatibility in the new program against the old interface, but that's
a minor point compared to the frontend problems.  I understand the need
for the kernel interface change, agree with it, and can live with that.

	It gets really entertaining to look at some of the headstands that
some people are doing to get some of this to work!  Someone else told
me that they had a "mount.smb" script that they used with mount.  If
you call mount with a file system type and it finds a "mount.type"
executable, it execs that executable.  His script would then look up user
names and passwords and anything else it needed and call smbmount.  Well...
He also needed to workaround the byzantine syntax gyrations depending on
kernel version.  So he calls my smbmount script which, in turn, calls
smbmount.samba, which, in turn, calls mount!  Here's how that one plays out:

	autofs -> mount -t smb -> mount.smb -> smbmount -> mount.samba -> mount

	Ultimately, mount ends up calling mount...

	That's scary!


	Talking with Luke in my office, I understand some of the history and
the why's and wherefore's behind putting smbmount, et al, into samba.  I
also understand that Tridge and Jeremy, et al, are rightfully frustrated by
a package that they don't use and currently can't support.  Hopefully, I'm
going to aid in easing that problem instead of compounding it by piling
on more complaints.

	I looked into fixing autofs to understand the new syntax.  I
thought "how hard could it be?"  Yeah right.  My thought was to first get
autofs to understand the new syntax and then get it to go "bi" and switch
depending on kernel version.  The current autofs plugs in the two position
parameters (the resource and the mount point) and then proceeds to process
the remaining parameters into options.  As I dug deeper into what was going
to be required to rearrange those parameters, plus synthesize the mount
command string with the mount point PLUS optional parameters, it just kept
getting uglier and uglier.  It was going to need a complete rewrite of
a couple of core parsing routines and then I realized that to "switch on
the fly" based on kernel version, I was going to have to have both the
old set of routines and the new set of routines, rather than even attempt
to get a single routine to parse out and synthesize the parameter list.
Then, on top of that, it was going to have to know how to find each of
those two unique binaries...  Sigh...

	I finally threw up my hands and wrote an smbmount wrapper script
that takes the place of smbmount, and checks the kernel version.  It then
calls smbmount.smbfs if the kernel version is 2.0.x, or munges the parameters
into the new syntax and calls smbmount.samba if the kernel version is 2.1.x
(anyone still using 2.1.x prior to 2.1.70 gets no sympathy...)  I also
found that I had to do some "magic" with the smbmount.samba output and
input, since the samba smbmount seems to hang if it is piped to syslog.  My
hacks for that then seem to have broken interactive use of smbmount (password
prompting got screwed).  Just found that one two days ago...  (Grumble)
I've got some other patches to the smbmount script that have been
contributed by others that I haven't integrated in as yet.

	The script is available at:

	http://www.wittsend.com/smbmount.html

	With the 2.2 kernels on the horizon, backward compatibility may
become less and less an issue, but it would be nice to have the new smbmount
program recognize older kernel version and fork off the old smbmount as
needed.  That would eliminate my script and simplify things immensely
(especially installation and updates in a mix environment).  If I can
munge around the parameter parsing to understand the old syntax, that would
solve a lot of problems.  The new syntax has been in use long enough, at
this point, that it can't be just abandoned either (or I'm doing exactly
what I'm bitching about to everyone else).  Yer damned if you do and
damn'ed if you don't...  Input and feedback on what direction to take
here would be vastly appreciated!

	Right now I have several things that I'm looking at.  I've already
started going through RECENT messages to samba-bugs regarding smbfs and
have responded to one or two.  Most have to do with version breaks and
compiling under different distributions.  I'm not sure I'm going to wade
in and answer everything going back 6 months or more, but I'm trying to
get to things only a month or two old.  I have at least one smbfs kernel
module problem on my list so I'm also forwarding this message to the
linux-kernel development list.


	Here is what is currently on my radar scope:

	Samba smbmount/smbmnt/smbumount problems:

	1) Smbmount (et al) is not currently supported by the samba 2.0
autoconfigure.

	Status: Fixed as soon as I can get the changes committed in.


	2) Smbmount does not compile under RedHat 5.1 (or, presumably other
glibc2 systems).

	Status: Fixed as soon as I can get the changes committed in.


	3) Smbmount goes "daemon" premature causing autofs and other scripts
to return before the mount has succeeded.

	Symptoms:  If you do an "ls" of an automounted directory that is not
currently mounted, the first attempt will sometimes come back with an empty
directory.  Subsequent access shows the directory properly mounted and can
be listed.  If you "cd" into a root directory of an automounted partition,
the cd will succeed and a simple "ls" will work but an "ls -l" will return
"no such file or directory" for every filename in the directory.  Syslog
entries show multiple mount attempts every time autofs attempts to mount
an smbfs mountpoint.

	What has happened is that smbmount forks a copy of itself, which
becomes the smbfs mount daemon, before the mount is completed.  The exiting
parent causes autofs to continue as if the mount were finished.  Subsequent
accesses may result in additional attempts to mount the pending mount.
The subsequent attempts eventually result in a "mount point busy" error.
In the resulting timing windows, some commands may complete, with erronious
results.  An "ls" of a directory may terminate, complete, having gotten no
data.  If you cd to a directory, you may have cd'ed to the directory before
the share is mounted on top of the directory you just cd'ed into.

	This isn't a problem when running smbmount manually, since the
mount completes long before the user can enter in the next command, but
the timing windows when this is being done from other programs is just
horrendous.

	Status: Pending.  I think this should be relatively easy to shoot.


	4) Smbmount hangs when stdout/stdin/stderr is piped to syslog.

	I haven't delved into this one very deeply as yet, but it is a royal
pain in the fanny with regard to autofs.  My wild ass guess is that it has
something to do with the password prompting routines...

	Status: Pending.  I have no real clue as yet...


	Smbfs kernel module (2.1.122):

	1) Date stamps bear little resemblance to reality.

	Here is an example...

	Three systems:

	Chaos - Linux 2.1.122
	Amber - Linux 2.0.35
	Phil  - Windows NT 4.0 SP3

	\\phil\public is mounted on /mnt/phil/public on both Chaos and Amber.

	On amber "ls -l /mnt/public" date stamps look reasonable:

drwxr-xr-x   1 root     root          512 Aug 27 16:14 t_elvis
drwxr-xr-x   1 root     root          512 Sep 10 12:06 tellison
drwxr-xr-x   1 root     root          512 Sep 15 16:24 telljohann
-rwxr-xr-x   1 root     root        83968 Jun 11 13:14 th.doc
drwxr-xr-x   1 root     root          512 Mar 11  1998 tmbackup
drwxr-xr-x   1 root     root          512 Jul 14 16:55 to_ftp
drwxr-xr-x   1 root     root          512 Aug 19 12:06 tpage
drwxr-xr-x   1 root     root          512 Aug  7 17:01 tsavino
drwxr-xr-x   1 root     root          512 May 18 16:40 unix_for_nt
-rwxr-xr-x   1 root     root        36352 May  9  1996 view_iss.xls
drwxr-xr-x   1 root     root          512 Jan 27  1998 vulndb
drwxr-xr-x   1 root     root          512 May 29 12:17 web_pages_jl
drwxr-xr-x   1 root     root          512 Aug 25 18:21 wingate
-rwxr-xr-x   1 root     root       709120 Oct 21  1997 winzip95.exe
drwxr-xr-x   1 root     root          512 Jul 28 09:55 winzipnt
-rwxr-xr-x   1 root     root       637728 Jan  6  1934 wrd6ex32.exe
-rwxr-xr-x   1 root     root       439904 Jun 17 22:18 xlock
drwxr-xr-x   1 root     root          512 Apr 21 11:32 xtian

	Same listing on chaos shows date/time stamps all over the map...

drwxr-xr-x   1 root     root          512 Oct 12  1954 t_elvis
drwxr-xr-x   1 root     root          512 Apr 30  1920 tellison
drwxr-xr-x   1 root     root          512 Aug 21  2026 telljohann
-rwxr-xr-x   1 root     root        83968 Nov  8  1971 th.doc
drwxr-xr-x   1 root     root          512 Feb  4  1981 tmbackup
drwxr-xr-x   1 root     root          512 Nov  5  2016 to_ftp
drwxr-xr-x   1 root     root          512 Jan 14  1969 tpage
drwxr-xr-x   1 root     root          512 Jun 25  2001 tsavino
drwxr-xr-x   1 root     root          512 Nov  7  1944 unix_for_nt
-rwxr-xr-x   1 root     root        36352 Oct 10  1929 view_iss.xls
drwxr-xr-x   1 root     root          512 Nov 20  2024 vulndb
drwxr-xr-x   1 root     root          512 Mar 17  1960 web_pages_jl
drwxr-xr-x   1 root     root          512 Mar 17  2015 wingate
-rwxr-xr-x   1 root     root       709120 Jun 14  2023 winzip95.exe
drwxr-xr-x   1 root     root          512 Oct 31  1960 winzipnt
-rwxr-xr-x   1 root     root       637728 Mar 18  1944 wrd6ex32.exe
-rwxr-xr-x   1 root     root       439904 Jun  5  1962 xlock
drwxr-xr-x   1 root     root          512 Feb 23  2016 xtian

	A directory listing from smbclient agrees with the amber (2.0.35)
date stamps and not with the chaos (2.1.122) date stamps.


	I guess this is an invitation to "open season" but...  Have I missed
anything?  Are there other issues that need investigating?

	I DON'T want to end up looking like a bull in a china shop, so I will
need feedback from other users of this animal...

	I'm currently waiting on Tridge before I can commit the changes I've
already got.  ITMT, I'm continuing to look over the smbfs messages up on
the samba-bugs site...

	IMHO...  We are in a feature freeze in the 2.1 kernels and building
toward a 2.2 release.  Stabilizing smbfs and fixing bugs fits right in there.
The smbmount stuff I'm working on is in the Samba 2.0 alpha line.  I don't
know if we are going to need to retrofit things back into the "released"
Samba series or not.  Smbmount will not compile on RedHat 5.x systems
currently.  We may yet have a beta Samba 2.0 out before Linux 2.2 hits
the street anyways.  It's not a jab at the kernel guys but a reflection
on the complexity they are dealing with.  Linus was talking about a 2.1
freeze last year at the Atlanta Linux Showcase and we've finally just
gotten there.  Where some of this fits in has yet to be clarified.

> Volker


	Regards,
	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From james at cows.ml.org  Mon Sep 28 15:04:35 1998
From: james at cows.ml.org (James Willard)
Date: Tue Dec  2 03:02:13 2003
Subject: WINS/DC Utilities
In-Reply-To: <Pine.NEB.3.96.980928152029.8771J-100000@cb1-gw.cb1.com> from "Luke Kenneth Casson Leighton" at Sep 28, 98 03:20:59 pm
Message-ID: <199809281504.LAA08344@cows.ml.org>

I am referring to a utility something like WINSDMP and WINSCHK from the NT
resource kit. It allows you to view the WINS database, flush entries,
initiate an integrity check, etc.

Does samba have the capability to be administrated by the NT Server Manager
yet in the NTDOM part of the code?

Thanks,

James Willard
james@cows.ml.org

> 
> well i just created a new command called "rpcclient".  and you can use
> nmblookup for wins.
> 
> On Sat, 26 Sep 1998, James Willard wrote:
> 
> > 
> >    I have not been able to find any kind of WINS or Domain Controller test
> > utilities for Linux. I am wanting to be able to see exactly what is going on
> > in the network such as finding the PDC or viewing a WINS database. Can anyone
> > suggest utilities that would help me with this?
> > 
> > James Willard
> > james@cows.ml.org
> > 
> > -- 
> > 
> 
> <a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
> <a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
> <a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>
> 


-- 
From lkcl at switchboard.net  Mon Sep 28 15:01:46 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: WINS/DC Utilities
In-Reply-To: <199809281504.LAA08344@cows.ml.org>
Message-ID: <Pine.NEB.3.96.980928160111.8771X-100000@cb1-gw.cb1.com>

On Mon, 28 Sep 1998, James Willard wrote:

> I am referring to a utility something like WINSDMP and WINSCHK from the NT
> resource kit. It allows you to view the WINS database, flush entries,
> initiate an integrity check, etc.

noy yet - edit wins.dat manually.
 
> Does samba have the capability to be administrated by the NT Server Manager
> yet in the NTDOM part of the code?

read-only, yes.
 
> Thanks,
> 
> James Willard
> james@cows.ml.org
> 
> > 
> > well i just created a new command called "rpcclient".  and you can use
> > nmblookup for wins.
> > 
> > On Sat, 26 Sep 1998, James Willard wrote:
> > 
> > > 
> > >    I have not been able to find any kind of WINS or Domain Controller test
> > > utilities for Linux. I am wanting to be able to see exactly what is going on
> > > in the network such as finding the PDC or viewing a WINS database. Can anyone
> > > suggest utilities that would help me with this?
> > > 
> > > James Willard
> > > james@cows.ml.org
> > > 
> > > -- 
> > > 
> > 
> > <a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
> > <a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
> > <a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>
> > 
> 
> 
> -- 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From whawes at transmeta.com  Mon Sep 28 15:38:21 1998
From: whawes at transmeta.com (Bill Hawes)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
References: <199809281457.KAA15937@alcove.wittsend.com>
Message-ID: <360FAD6C.2BCB808C@transmeta.com>

Michael H. Warfield wrote:

>         Smbfs kernel module (2.1.122):
>
>         1) Date stamps bear little resemblance to reality.
>
>         Here is an example...
>
>         Three systems:
>
>         Chaos - Linux 2.1.122
>         Amber - Linux 2.0.35
>         Phil  - Windows NT 4.0 SP3
>
>         \\phil\public is mounted on /mnt/phil/public on both Chaos and Amber.
>
>         On amber "ls -l /mnt/public" date stamps look reasonable:
>
> drwxr-xr-x   1 root     root          512 Aug 27 16:14 t_elvis
> drwxr-xr-x   1 root     root          512 Sep 10 12:06 tellison
> drwxr-xr-x   1 root     root          512 Sep 15 16:24 telljohann

Hi Michael,

I intend to maintain the kernel smbfs code, so please direct any kernel-related bug
reports to the linux-kernel list.

With regards to the timestamp problem, I would suspect that you've mounted the
volume with incorrect bug-workaround flags. There are differences in the way
timestamps are handled in Win95 vs NT smb servers, so if smbfs thinks it's talking
to a Win95 system, timestamps will probably be wrong.

If you need to support both NT and Win 95 systems using the same kernel modules,
you need to _not_ specify the Win 95 kernel config option, but instead use the
mount time flags.

Regards,
Bill


From mhw at wittsend.com  Mon Sep 28 15:51:02 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <360FAD6C.2BCB808C@transmeta.com> from Bill Hawes at "Sep 28, 1998  8:38:21 am"
Message-ID: <199809281551.LAA16147@alcove.wittsend.com>

Bill Hawes enscribed thusly:
> Michael H. Warfield wrote:

> >         Smbfs kernel module (2.1.122):

> >         1) Date stamps bear little resemblance to reality.

> >         Here is an example...

> >         Three systems:

> >         Chaos - Linux 2.1.122
> >         Amber - Linux 2.0.35
> >         Phil  - Windows NT 4.0 SP3

> >         \\phil\public is mounted on /mnt/phil/public on both Chaos and Amber.

> >         On amber "ls -l /mnt/public" date stamps look reasonable:

> > drwxr-xr-x   1 root     root          512 Aug 27 16:14 t_elvis
> > drwxr-xr-x   1 root     root          512 Sep 10 12:06 tellison
> > drwxr-xr-x   1 root     root          512 Sep 15 16:24 telljohann

> Hi Michael,

> I intend to maintain the kernel smbfs code, so please direct any kernel-related bug
> reports to the linux-kernel list.

	That's why I added it...  :-)

> With regards to the timestamp problem, I would suspect that you've mounted the
> volume with incorrect bug-workaround flags. There are differences in the way
> timestamps are handled in Win95 vs NT smb servers, so if smbfs thinks it's talking
> to a Win95 system, timestamps will probably be wrong.

	Oh great...

> If you need to support both NT and Win 95 systems using the same kernel modules,
> you need to _not_ specify the Win 95 kernel config option, but instead use the
> mount time flags.

	Ok...  That fits because, yes I did specify the Win95 bug workaround
option.  You say to use the mount time flags.  Where does that fit in the
smbmount syntax?  I noticed that the smbmount doc's are, shall we say, a bit
weak.  Is this an additional option to the -c "mount /mountpoint ...." string
or is it something that smbmount should be parsing (I would expect the later
but would not be surprised by the former).  I don't find anything in
particular in the smbmount command line parsing that would point a finger
at what it is suppose to be.  The docs on the mount command basically
say it's in smbmount's domain.  If it's suppose to be in smbmount that
either means I'm overlooking it (real easy) or it's another one on my list
to do...

> Regards,
> Bill

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From whawes at transmeta.com  Mon Sep 28 16:10:11 1998
From: whawes at transmeta.com (Bill Hawes)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
References: <199809281551.LAA16147@alcove.wittsend.com>
Message-ID: <360FB4E3.93FC9F01@transmeta.com>

Michael H. Warfield wrote:

> > If you need to support both NT and Win 95 systems using the same kernel modules,
> > you need to _not_ specify the Win 95 kernel config option, but instead use the
> > mount time flags.
>
>         Ok...  That fits because, yes I did specify the Win95 bug workaround
> option.  You say to use the mount time flags.  Where does that fit in the
> smbmount syntax?  I noticed that the smbmount doc's are, shall we say, a bit
> weak.  Is this an additional option to the -c "mount /mountpoint ...." string
> or is it something that smbmount should be parsing (I would expect the later
> but would not be surprised by the former).  I don't find anything in
> particular in the smbmount command line parsing that would point a finger
> at what it is suppose to be.  The docs on the mount command basically
> say it's in smbmount's domain.  If it's suppose to be in smbmount that
> either means I'm overlooking it (real easy) or it's another one on my list
> to do...

Hi Mike,

The smbmnt man page describes the use of the -f argument to specifiy bug workarounds,
though probably it should just pull in the information from the smbfs.txt file rather
than referring to it. But at the time the kernel support was changing rapidly, so it
seemed better to do it this way.

The basic problem is that smbfs really needs to know which flavor of server it's talking
to, as the smb protocol level isn't sufficient. For example, for one of the smb messages
Win 95 swaps the order of timestamp fields, contrary to SMB docs _and_ Win NT practice.

Regards,
Bill



From lkcl at switchboard.net  Mon Sep 28 18:39:46 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <360FB4E3.93FC9F01@transmeta.com>
Message-ID: <Pine.NEB.3.96.980928193814.8771k-100000@cb1-gw.cb1.com>

> The basic problem is that smbfs really needs to know which flavor of server it's talking
> to, as the smb protocol level isn't sufficient. For example, for one of the smb messages
> Win 95 swaps the order of timestamp fields, contrary to SMB docs _and_ Win NT practice.

nt does this by doing a "NetServerEnum" dce/rpc call
win95 does this by issuing an SMBopenX on a dce/rpc pipe "\PIPE\lsass" and
then closing it, if the open succeeded.

... identifies whether the server is nt or not-nt.

From oliva at dcc.unicamp.br  Mon Sep 28 22:39:37 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:13 2003
Subject: Still build problems with Samba
In-Reply-To: Richard Sharpe's message of "Sat, 26 Sep 1998 11:50:30 +1000"
References: <3.0.5.32.19980926110859.00867220@mail.adelaide.on.net>
Message-ID: <orn27jlv4m.fsf@araguaia.dcc.unicamp.br>

Richard Sharpe <sharpe@ns.aus.com> writes:

> here is the latest problem

>> make: *** No rule to make target `rpcclient/rpcclient.o', needed by 
>> `bin/rpcclient'.  Stop.

cvs up -d

rcpclient seems to be a new directory

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From tridge at samba.anu.edu.au  Mon Sep 28 23:30:07 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <360FAD6C.2BCB808C@transmeta.com> (message from Bill Hawes on
	Tue, 29 Sep 1998 02:15:58 +1000)
References: <199809281457.KAA15937@alcove.wittsend.com> <360FAD6C.2BCB808C@transmeta.com>
Message-ID: <19980928233015Z12668397-7009+12138@samba.anu.edu.au>

> With regards to the timestamp problem, I would suspect that you've
> mounted the volume with incorrect bug-workaround flags. There are
> differences in the way timestamps are handled in Win95 vs NT smb
> servers, so if smbfs thinks it's talking to a Win95 system,
> timestamps will probably be wrong.

you might like to change smbfs to look at the capabilities bits from
the negprot response:

if (capabilities & CAP_NT_SMBS) {
  use_nt_code();
} else {
  use_win95_code();
}

or should smbmount do this and pass a flag to the kernel?

From tridge at samba.anu.edu.au  Tue Sep 29 01:10:05 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <199809281457.KAA15937@alcove.wittsend.com> (mhw@wittsend.com)
References: <199809281457.KAA15937@alcove.wittsend.com>
Message-ID: <19980929011007Z12645666-7510+12292@samba.anu.edu.au>

> The smbmount stuff I'm working on is in the Samba 2.0 alpha line.  I don't
> know if we are going to need to retrofit things back into the "released"
> Samba series or not.

no, we won't do that. 1.9.18 is dead.

Cheers, Tridge
From mhw at wittsend.com  Tue Sep 29 02:57:07 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <19980928233015Z12668397-7009+12138@samba.anu.edu.au> from Andrew Tridgell at "Sep 29, 1998  9:35:55 am"
Message-ID: <199809290257.WAA07841@alcove.wittsend.com>

Andrew Tridgell enscribed thusly:
> > With regards to the timestamp problem, I would suspect that you've
> > mounted the volume with incorrect bug-workaround flags. There are
> > differences in the way timestamps are handled in Win95 vs NT smb
> > servers, so if smbfs thinks it's talking to a Win95 system,
> > timestamps will probably be wrong.

> you might like to change smbfs to look at the capabilities bits from
> the negprot response:

> if (capabilities & CAP_NT_SMBS) {
>   use_nt_code();
> } else {
>   use_win95_code();
> }

> or should smbmount do this and pass a flag to the kernel?

	I think the later...  Luke and I and Dave LeBlanc (my evil NT
twin at Internet Security Systems) were discussing this today.  Dave
thinks that doing a "GetServerInfo" at the appropriate level should
give the information we are looking for.  There are bits which will
identify WfWg and W95.  We can apply appropriate discrimination at
the time the connection is established.  Luke had a different approach.
He suggested opening an RPC pipe and, if that failed, then applying the
bug work-around.  That would ALSO be at connection time.  Downside to Luke's
approach would be if someone added that capability to Win95 (YUCK).  This is,
of course, all dependent on if smbmount is not currently doing the correct
thing ANYWAYS...  I actually DO NOT KNOW - YET.  I do know that disabling the
Win95 bug work-around compile option DID fix the problem I was seeing (thank
you, Bill).  Now, I need to dig into the guts and see if smbmount (or smbmnt)
is doing the right thing based on the protocols...  As I told Bill...  Looks
like one more on my plate after all...

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From mhw at wittsend.com  Tue Sep 29 03:07:19 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <19980929011007Z12645666-7510+12292@samba.anu.edu.au> from Andrew Tridgell at "Sep 29, 1998 11:12:55 am"
Message-ID: <199809290307.XAA07973@alcove.wittsend.com>

Andrew Tridgell enscribed thusly:
> > The smbmount stuff I'm working on is in the Samba 2.0 alpha line.  I don't
> > know if we are going to need to retrofit things back into the "released"
> > Samba series or not.

> no, we won't do that. 1.9.18 is dead.

	So be it...  No further consideration given.

> Cheers, Tridge

	Thanks!

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From tridge at samba.anu.edu.au  Tue Sep 29 03:40:02 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <199809290257.WAA07841@alcove.wittsend.com> (mhw@wittsend.com)
References: <199809290257.WAA07841@alcove.wittsend.com>
Message-ID: <19980929034006Z12643063-25139+12275@samba.anu.edu.au>

> 	I think the later...  Luke and I and Dave LeBlanc (my evil NT
> twin at Internet Security Systems) were discussing this today.  Dave
> thinks that doing a "GetServerInfo" at the appropriate level should
> give the information we are looking for.

the advatnage of using the capabilities bits in the negprot is that no
new calls are needed. You're already getting the negprot response, you
just need to look at the CAP_NT_SMBS bit (or store the capabilities
for future reference)

From herb at chomps.engr.sgi.com  Tue Sep 29 04:06:22 1998
From: herb at chomps.engr.sgi.com (Herb Lewis)
Date: Tue Dec  2 03:02:13 2003
Subject: .cvsignore
Message-ID: <36105CBE.2AA7981B@chomps.engr.sgi.com>

All the .cvsignore files have an entry for "dummy" but the new
configure creates ".dummy" files in all directories. One of the
2 needs to be changed so you don't keep getting extra files found
during cvs update. Which should it be "dummy" or ".dummy" and who
is going to make the change?

-- 
======================================================================
Herb Lewis                                   Silicon Graphics 
Technical Marketing                          2011 N Shoreline Blvd
Network Systems Division                     Mountain View, CA  94043  
herb@sgi.com                                 Tel: 650-933-2177
http://www.sgi.com                           Fax: 650-932-2177          
======================================================================
From oliva at dcc.unicamp.br  Tue Sep 29 04:16:33 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:13 2003
Subject: .cvsignore
In-Reply-To: Herb Lewis's message of "Tue, 29 Sep 1998 14:10:44 +1000"
References: <36105CBE.2AA7981B@chomps.engr.sgi.com>
Message-ID: <orpvcfh7tq.fsf@araguaia.dcc.unicamp.br>

Herb Lewis <herb@chomps.engr.sgi.com> writes:

> Which should it be "dummy" or ".dummy" and who is going to make the
> change?

None of them, IMHO.  Since I joined the team, I've been trying to
figure out why these files are needed.

It seems to me that they were only created as an ugly mechanism to
ensure that subdirectories are properly created in the build tree.
I'm sure I can figure out a better way to do that, that does not
require these files.

Can anyone think of any reason for not removing them once and for all?

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From tridge at samba.anu.edu.au  Tue Sep 29 08:51:25 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: .cvsignore
In-Reply-To: <orpvcfh7tq.fsf@araguaia.dcc.unicamp.br> (message from Alexandre
	Oliva on Tue, 29 Sep 1998 14:27:10 +1000)
References: <36105CBE.2AA7981B@chomps.engr.sgi.com> <orpvcfh7tq.fsf@araguaia.dcc.unicamp.br>
Message-ID: <19980929085131Z12669351-25139+12316@samba.anu.edu.au>

> Can anyone think of any reason for not removing them once and for all?

if you've got a better method then please delete them! It was just a
hack :)
From herb at chomps.engr.sgi.com  Tue Sep 29 13:47:56 1998
From: herb at chomps.engr.sgi.com (Herb Lewis)
Date: Tue Dec  2 03:02:13 2003
Subject: .cvsignore
References: <36105CBE.2AA7981B@chomps.engr.sgi.com> <orpvcfh7tq.fsf@araguaia.dcc.unicamp.br>
Message-ID: <3610E50C.7C6B9C97@chomps.engr.sgi.com>

Alexandre Oliva wrote:
> 
> Herb Lewis <herb@chomps.engr.sgi.com> writes:
> 
> > Which should it be "dummy" or ".dummy" and who is going to make the
> > change?
> 
> None of them, IMHO.  Since I joined the team, I've been trying to
> figure out why these files are needed.
> 
> It seems to me that they were only created as an ugly mechanism to
> ensure that subdirectories are properly created in the build tree.
> I'm sure I can figure out a better way to do that, that does not
> require these files.
> 
> Can anyone think of any reason for not removing them once and for all?
> 

While you are at it, what about include/stamp-h? It should be added
to .cvsignore (like config.h which is also created) or find a way to
do away with it also.

-- 
======================================================================
Herb Lewis                                   Silicon Graphics 
Technical Marketing                          2011 N Shoreline Blvd
Network Systems Division                     Mountain View, CA  94043  
herb@sgi.com                                 Tel: 650-933-2177
http://www.sgi.com                           Fax: 650-932-2177          
======================================================================
From mhw at wittsend.com  Tue Sep 29 14:29:56 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:13 2003
Subject: nmbd dying...
Message-ID: <199809291429.KAA07462@alcove.wittsend.com>

Hey gang...

	I've been noticing for some time that I've been having a problem
with nmbd and the latest snapshots.  In particular, I've noticed that
after a while, nmbd disappears.  More specifically, on the system I have
running as a WINS server, the parent nmbd disappears.  On the systems
that are NOT the WINS server, the only nmbd process disappears.  One
system is a RedHat 5.1 system running Linux 2.1.123 and another is a
RedHat 4.2 system running Linux 2.0.35.  Both exhibit the same problem
and the problem was still there with todays snapshot.

	I think I saw a couple of other messages (now deleted) regarding
something that might have been similar on this list a few days ago...  I
think one was from Luke and I'm going to check with him when I get into
the office...

	I decided to take a closer peek when I noticed the nmbd processes
on two systems disappeared simultaneously just a few minutes ago.

	I restarted the samba subsystems and then ran gdb on the nmbd binary.
I then attached to the main nmbd process and started to play around expecting
it to eventually DO SOMETHING.  Well, it did, and not at all what I expected.
As soon as I hit "Ctrl-Alt-Delete" on my Windows NT box, nmbd segfaulted.

	I can reproduce it every time.  Get nmbd up and running and then
go to log into the workstation and the nmbd processes are toast.  I've
gotten two gdb traces, both the same.

	The workstation is running Windows NT 4.0 SP3 and the last login was
to the WittsEnd domain for which Samba on alcove.wittsend.com is the PDC.

	The gdb trace is from Alcove...

	Here is the trace from gdb:

(gdb) attach 7419
Attaching to program `/mnt2/src/samba/source/bin/nmbd', process 7419
Reading symbols from /usr/lib/libreadline.so.3...done.
Reading symbols from /lib/libdl.so.2...done.
Reading symbols from /lib/libcrypt.so.1...done.
Reading symbols from /lib/libpam.so.0...done.
Reading symbols from /lib/libc.so.6...done.
Reading symbols from /lib/libtermcap.so.2...done.
Reading symbols from /lib/ld-linux.so.2...done.
Reading symbols from /lib/libnss_files.so.1...done.
Reading symbols from /lib/libnss_nis.so.1...done.
Reading symbols from /lib/libnsl.so.1...done.
Reading symbols from /lib/libnss_dns.so.1...done.
Reading symbols from /lib/libresolv.so.2...done.
0x400da83e in __select ()
(gdb) continue
Continuing.

Program received signal SIGSEGV, Segmentation fault.
process_logon_packet (p=0x80abf88, buf=0x80ac0dc "\022", len=104, 
    mailslot=0x807e868 "\\MAILSLOT\\NET\\NTLOGON")
    at nmbd/nmbd_processlogon.c:174
174           ntversion = IVAL(q, 0); q += 4;
(gdb) where
#0  process_logon_packet (p=0x80abf88, buf=0x80ac0dc "\022", len=104, 
    mailslot=0x807e868 "\\MAILSLOT\\NET\\NTLOGON")
    at nmbd/nmbd_processlogon.c:174
#1  0x805485c in process_dgram (p=0x80abf88) at nmbd/nmbd_packets.c:1277
#2  0x8054f9a in run_packet_queue () at nmbd/nmbd_packets.c:1588
#3  0x804acbc in process () at nmbd/nmbd.c:264
#4  0x804b868 in main (argc=2, argv=0xbffff914) at nmbd/nmbd.c:763
(gdb) 

	Anything else I can get'cha?

	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From lkcl at switchboard.net  Tue Sep 29 15:15:19 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
In-Reply-To: <199809290257.WAA07841@alcove.wittsend.com>
Message-ID: <Pine.NEB.3.96.980929161235.8771y-100000@cb1-gw.cb1.com>

On Tue, 29 Sep 1998, Michael H. Warfield wrote:

> Andrew Tridgell enscribed thusly:
> > > With regards to the timestamp problem, I would suspect that you've
> > > mounted the volume with incorrect bug-workaround flags. There are
> > > differences in the way timestamps are handled in Win95 vs NT smb
> > > servers, so if smbfs thinks it's talking to a Win95 system,
> > > timestamps will probably be wrong.
> 
> > you might like to change smbfs to look at the capabilities bits from
> > the negprot response:
> 
> > if (capabilities & CAP_NT_SMBS) {
> >   use_nt_code();
> > } else {
> >   use_win95_code();
> > }
> 
> > or should smbmount do this and pass a flag to the kernel?
> 
> 	I think the later...  Luke and I and Dave LeBlanc (my evil NT
> twin at Internet Security Systems) were discussing this today.  Dave
> thinks that doing a "GetServerInfo" at the appropriate level should
> give the information we are looking for.  There are bits which will
> identify WfWg and W95.

the call is a dce/rpc "NetServerGetInfo" call.

>  We can apply appropriate discrimination at
> the time the connection is established.  Luke had a different approach.
> He suggested opening an RPC pipe and, if that failed, then applying the
> bug work-around.  That would ALSO be at connection time.  Downside to Luke's
> approach would be if someone added that capability to Win95 (YUCK).

no, that's also ok.  the pipe opening is part of doing the
"NetServerGetInfo" call.  the thing is that this is what i have observed
ms systems doing: just opening "\PIPE\srvsvc" or "\PIPE\lsass" not sure
which and then closing it.

heck, if microsoft want to add dce/rpc code to win98, good luck to them:
they'll have to find the solution to this problem and then we just watch
on the wire what they did and follow suite.

From whawes at transmeta.com  Tue Sep 29 15:25:02 1998
From: whawes at transmeta.com (Bill Hawes)
Date: Tue Dec  2 03:02:13 2003
Subject: smbmount and smbfs (was Re: smbmount et al...)
References: <199809290257.WAA07841@alcove.wittsend.com>
Message-ID: <3610FBCE.113AE4A4@transmeta.com>

Michael H. Warfield wrote:

> Andrew Tridgell enscribed thusly:
>
> > you might like to change smbfs to look at the capabilities bits from
> > the negprot response:
>
> > if (capabilities & CAP_NT_SMBS) {
> >   use_nt_code();
> > } else {
> >   use_win95_code();
> > }
>
> > or should smbmount do this and pass a flag to the kernel?
>
>         I think the later...  Luke and I and Dave LeBlanc (my evil NT
> twin at Internet Security Systems) were discussing this today.  Dave
> thinks that doing a "GetServerInfo" at the appropriate level should
> give the information we are looking for.  There are bits which will
> identify WfWg and W95.  We can apply appropriate discrimination at
> the time the connection is established.

It sounds reasonable to me to let smbmount make a determination of the
appropriate bug-fix flags based on its protocol negotiation. Right now the only
flag that's needed is the w95 workaround -- WfWg should work OK based on its
protocol level, and NT will work provided the w95 workarounds aren't turned on.

Hopefully the situation won't change with w98 or future NT releases, but if they
do we'll just have to update the utilities and kernel as needed.

To implement auto-selection of bug-fixes, perhaps smbmount can pass a "server
type" argument to smbmnt, and smbmnt can fold that into the appropriate -f
argument.

Regards,
Bill


From crh at NTS.Umn.EDU  Tue Sep 29 16:59:51 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:13 2003
Subject: Assigned numbers.
Message-ID: <199809291659.LAA08694@unet.unet.umn.edu>

I've been poking around a bit and have a few questions.

  Port 42:   It appears as though this is being used for WINS replication. 
             Any truth to the rumor?

  name             42/tcp    Host Name Server
  name             42/udp    Host Name Server
  nameserver       42/tcp    Host Name Server
  nameserver       42/udp    Host Name Server

  Port 135:  This is listed as a DCE/RPC termination port.  Is this being
             used by the NT DCE/RPC stuff at all?

  epmap           135/tcp    DCE endpoint resolution
  epmap           135/udp    DCE endpoint resolution
  #                          Joe Pato <pato@apollo.hp.com>

  Ports 137, These are assigned for both tcp & udp.  It appears, however,
        138, that they are not used that way.  In nmbd.c:open_sockets(),
      & 139: both the NetBIOS Name Server port and the Datagram port are
             opened as SOCK_DGRAM.  RFC1002 talks about using both udp
             and tcp for name resolution.  Is this done?

             The RFC says (in section 4.2.1.1) that the truncation flag
             is:
                      "Set if this message was truncated because the
                       datagram carrying it would be greater than
                       576 bytes in length.  Use TCP to get the
                       information from the NetBIOS Name Server."

Going through the latest port assigments at 
  http://www.isi.edu/in-notes/iana/assignments/port-numbers
I notice several of entries by Microsoft.  I don't know how many of these
are new.  Comments?

Oh, BTW...  In nameserv.h it is written:  

#define MAX_DGRAM_SIZE (576) /* tcp/ip datagram limit is 576 bytes */

Shouldn't we say "udp/ip"?

Yours pedantically,

Chris -)-----

--
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From oliva at dcc.unicamp.br  Tue Sep 29 17:18:21 1998
From: oliva at dcc.unicamp.br (Alexandre Oliva)
Date: Tue Dec  2 03:02:13 2003
Subject: .cvsignore
In-Reply-To: Herb Lewis's message of "Tue, 29 Sep 1998 06:47:56 -0700"
References: <36105CBE.2AA7981B@chomps.engr.sgi.com> <orpvcfh7tq.fsf@araguaia.dcc.unicamp.br> <3610E50C.7C6B9C97@chomps.engr.sgi.com>
Message-ID: <orbtnyhm76.fsf@araguaia.dcc.unicamp.br>

Herb Lewis <herb@chomps.engr.sgi.com> writes:

> what about include/stamp-h? It should be added
> to .cvsignore

Done.  Please complain again if I introduce new files and forget to
add them to .cvsignore.  Since I always build in a separate tree, I'll 
hardly notice unless someone reminds me of that.

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil

From ivey at realminfo.com  Tue Sep 29 19:03:54 1998
From: ivey at realminfo.com (Michael D. Ivey)
Date: Tue Dec  2 03:02:13 2003
Subject: latest code fails with NTserver, works with 95
Message-ID: <19980929150354.A10741@bagheera.realminfo.com>

In my never-ending quest to get everything working, I
now can't get my 2 NT servers into the domain.  The 95
clients were having trouble at first, but now they seem
ok.  NT boxes say "No domain controller available" and
fail miserably.  'net view' doesn't show the server, even.

Any suggestions?  Also, please tell me how to update the
smb.conf without kicking everyone off...I can't get a
kill -HUP to do it.

Thanks.

/mdi

-- 
Michael D. Ivey  -  Director of Emerging Technologies
ivey@realminfo.com    http://www.realminfo.com/~ivey/
From jallison at cthulhu.engr.sgi.com  Tue Sep 29 19:22:51 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:13 2003
Subject: Samba AutoCAD 14 problem - info (maybe a solution....)
Message-ID: <3611338B.20ED5CB5@engr.sgi.com>

"Kelly Uehling" <kellyu@mwcad.com> writes in comp.protocols.smb:

>This has been an ongoing problem with Unix SMB file servers. While perusing
>the autodesk.autocad.network newsgroup, I came across someone else having
>the same problem with a Syntax 5.3 server. Syntax's response was as follows:
>
>    By Syntax Technical Support (Support) on Thursday, September 3, 1998 -
>03:53 pm:
>
>    The problem:
>    When using AutoCad 14 on an NT workstation w/Service Pack 3, accessing a
>file from a TAS NB realm volume, the file will be opened read-only
>occasionally.
>
>    Background:
>    When we investigated this problem using an NT workstation without SP3
>the problem was fixed by reducing the mux value to 1. When we did this the
>NT did not try to do named pipe opens for srvsvc or SMBtrans API's. Now with
>SP3 the NT will do these calls even with mux set to 1.
>
>    The problem has been narrowed to the SMBtrans API call comming between
>two SMB's, SMBgetattr filename.dwk and the SMBopenX filename.dwg. The .dwk
>file does not exist and this should not cause a problem. It is interesting
>to note that the r/w open of the .dwg file never fails and even where
>AutoCad is saying the file is read-only the file is being opened
>successfully r/w.
>
>    It is also interesting that AutoCad is quite flexible in the way it does
>things. No two traces are exactly the same even though the same actions are
>done each time. The traces are basically the same but actions can come in
>different orders which seems to point to multi-threading of the application.
>
>    It is also interesting that AutoCad is quite flexible in the way it does
>things. No two traces are exactly the same even though the same actions are
>done each time. The traces are basically the same but actions can come in
>different orders which seems to point to multi-threading of the application.
>
>    It is suspected that the API transaction is somehow causing AutoCad to
>pick up the failure of the getattr SMB for filename.dwk as a failure for
>filename.dwg.
>
>    AutoDesk has an ackowleged problem with NT using a Novell 32 bit client
>that is exactly like this. Although this is a different realm it may be the
>same problem to the application.
>
>Maybe this will help track the problem down in Samba. Maybe it won't.

Kelly - you're a *wonderful* person !

This looks *exactly* like the issue I see in the Samba
logs - the getattr failure for <filename>.dwk.

The problem is that if you set maxmux to 1 in
Samba then NT stops being able to load user
profiles from a Samba share (that's why we upped it 
to 50, as NT uses). If you're not hung for one problem 
you're hung for another.... :-).

So what it *actually* looks like is a bug
in the Microsoft redirector that is compounded
by the way AutoCAD 14 checks for read only.

The way to check this is for Samba users having
this problem to set in the [global] section of
their smb.conf the parameter :

max mux = 1

and re-start Samba. If this information is
correct the AutoCAD problem will not then
occur (although you will not be able to
store NT user profiles on that Samba server
due to the *other* bug in the NT redirector :-).

My guess is that this doesn't happen on NT
as theirs is a multi-threaded implementation
and just happens to process the SMB requests
that AutoCAD sends in a different order - if
they processed them fist-come-first served, as
Samba does then they'd get the same problem with
AutoCAD !

Well I never thought I'd say so, but if this
fix works then "Thanks a *BUNCH*, Syntax :-)".

Cheers,

        Jeremy Allison,
        Samba Team.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From ivey at realminfo.com  Tue Sep 29 19:53:51 1998
From: ivey at realminfo.com (Michael D. Ivey)
Date: Tue Dec  2 03:02:13 2003
Subject: more info on the nmbd problem
Message-ID: <19980929155351.A11001@bagheera.realminfo.com>

> Michael H. Warfield (mhw@wittsend.com)
> Tue, 29 Sep 1998 10:29:56 -0400 (EDT) 
> 
>         I've been noticing for some time that I've been having a problem 
> with nmbd and the latest snapshots. In particular, I've noticed that 
> after a while, nmbd disappears. More specifically, on the system I have 

I'm having the same problem...here's my trace:

[root@alice source]# nmbd -D; ps axf |grep nmb
 9816  ?  R    0:00 nmbd -D 
 9818  ?  R    0:00  \_ nmbd -D 
[root@alice source]# gdb
GNU gdb 4.17
Copyright 1998 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-redhat-linux".
(gdb) attach 9816
Attaching to process 9816
0x400d983e in ?? ()
(gdb) continue 
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0x8055eb2 in ?? ()
(gdb) where
#0  0x8055eb2 in ?? ()
#1  0x80548bc in ?? ()
#2  0x8054ffa in ?? ()
#3  0x804ad1c in ?? ()
#4  0x804b8c8 in ?? ()
(gdb) 


I suspect this is what's keeping my NT servers from being able to login?

My list subscription hasn't gone through yet, but I'm reading in the
archive...any help would be appreciated.

/mdi

-- 
Michael D. Ivey  -  Director of Emerging Technologies
ivey@realminfo.com    http://www.realminfo.com/~ivey/
From lkcl at switchboard.net  Tue Sep 29 21:22:09 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:13 2003
Subject: dynamic-allocated string array "holder"
Message-ID: <Pine.NEB.3.96.980929222058.8771c-100000@cb1-gw.cb1.com>

do we have such code to do this, modularised or otherwise?

From tridge at samba.anu.edu.au  Tue Sep 29 23:40:06 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:13 2003
Subject: Assigned numbers.
In-Reply-To: <199809291659.LAA08694@unet.unet.umn.edu> (crh@NTS.Umn.EDU)
References: <199809291659.LAA08694@unet.unet.umn.edu>
Message-ID: <19980929234006Z12670182-7510+12623@samba.anu.edu.au>

>   Port 42:   It appears as though this is being used for WINS replication. 
>              Any truth to the rumor?
> 
>   name             42/tcp    Host Name Server
>   name             42/udp    Host Name Server
>   nameserver       42/tcp    Host Name Server
>   nameserver       42/udp    Host Name Server

that's a new one for me.
 
>   Port 135:  This is listed as a DCE/RPC termination port.  Is this being
>              used by the NT DCE/RPC stuff at all?
> 
>   epmap           135/tcp    DCE endpoint resolution
>   epmap           135/udp    DCE endpoint resolution

it is used, but I don't know exactly what for. I've seen packets to it
though.

>   Ports 137, These are assigned for both tcp & udp.  It appears, however,

It seems that the IANA always does that. When I asked for a port for
rsync they gave me UDP as well as TCP even though I only asked for
TCP.

>                       "Set if this message was truncated because the
>                        datagram carrying it would be greater than
>                        576 bytes in length.  Use TCP to get the
>                        information from the NetBIOS Name Server."

I don't believe the MS implementations do this. Anyone see it?

From tridge at samba.anu.edu.au  Wed Sep 30 01:00:09 1998
From: tridge at samba.anu.edu.au (Andrew Tridgell)
Date: Tue Dec  2 03:02:14 2003
Subject: dynamic-allocated string array "holder"
In-Reply-To: <Pine.NEB.3.96.980929222058.8771c-100000@cb1-gw.cb1.com> (message
	from Luke Kenneth Casson Leighton on Wed, 30 Sep 1998 07:35:33 +1000)
References: <Pine.NEB.3.96.980929222058.8771c-100000@cb1-gw.cb1.com>
Message-ID: <19980930010009Z12669225-20449+12480@samba.anu.edu.au>

depends, what is a "string array holder" ??

the only dynamic string support we have is:

string_init()
string_set()
string_free()

it's used in a few places, particularly the loadparm code.

If you want arrays of strings then maybe build functions for them on
top of the above functions?

From jallison at cthulhu.engr.sgi.com  Tue Sep 29 17:24:22 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:14 2003
Subject: nmbd dying...
References: <199809291429.KAA07462@alcove.wittsend.com>
Message-ID: <361117C6.5B14296C@engr.sgi.com>

Michael H. Warfield wrote:
> 
>         I can reproduce it every time.  Get nmbd up and running and then
> go to log into the workstation and the nmbd processes are toast.  I've
> gotten two gdb traces, both the same.

peloy@ven.ra.rockwell.com wrote:

> a CVS update done today Sep. 29 left my nmbd in an unstable state. Whn
> I try to log into the Samba PDC nmbd dies:
> 
> 
> [1998/09/29 09:52:12, 1] nmbd/nmbd_processlogon.c:process_logon_packet(69)
>   process_logon_packet: Logon from 130.151.17.156: code = 12
> [1998/09/29 09:52:12, 0] lib/fault.c:fault_report(40)

Arggg. My fault sorry.

Dead code elimination. I removed a line such as

allowableaccount = IVAL(q, 0); q += 4;

as the variable "allowableaccount" was unused.
I missed the bloody "q += 4;".

This is why multiple statements on one line are
a *BUG*. I will eventualy remove *ALL* these (luke
you have been warned :-).

Please CVS update and it should be fixed. Please
accept my apologies.

If it isn't (and you get another core dump in
the nmbd logon code) - sending the gdb backtrace
allowed me to find it instantly so please do
that again - it really helped, thanks.

Cheers,

Jeremy.


-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From mhw at wittsend.com  Wed Sep 30 13:30:31 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
In-Reply-To: <19980929234006Z12670182-7510+12623@samba.anu.edu.au> from Andrew Tridgell at "Sep 30, 1998  9:41:51 am"
Message-ID: <199809301330.JAA22457@alcove.wittsend.com>

Andrew Tridgell enscribed thusly:
> >   Port 42:   It appears as though this is being used for WINS replication. 
> >              Any truth to the rumor?

> >   name             42/tcp    Host Name Server
> >   name             42/udp    Host Name Server
> >   nameserver       42/tcp    Host Name Server
> >   nameserver       42/udp    Host Name Server

	Aaaauuuugggghhhhh!!!!  It arises from the dead!

> that's a new one for me.

	No...  Not a new one at all.  An old one.  An evil one.  "The star
stones must have come unstuck" one (for H.P. Lovecraft fans)...  :-)

	The "nameserver" or "hostname" service is an OLD service that has
been obsolete for longer than I've been involved with TCP/IP (almost 15
years now?  Gees, I'm getting old...)  It's listed back in several RFC's
that actually have only 3 digits!  (start at 961 and work your way through
the "obsoleted by" list until you end up at 1011).  It has it's roots in
IEN 116 which described what amounted to a "what's in your host file" type
protocol.

	This is from RFC 1011...  I think it says it best.  It describes
what this thing was, what its problems were and the fact that it was
abandoned in favor of the DNS.  RFC 1011 is dated 1987...  Nuff said.

]    Host Name Server Protocol  ----------------------------- (NAMESERVER)
] 
]       STATUS:  Experimental
] 
]       SPECIFICATION:  IEN 116 (in DPH)
] 
]       COMMENTS:
] 
]          Provides machine oriented procedure for translating a host name
]          to an Internet Address.
] 
]          This specification has significant problems:  1) The name
]          syntax is out of date.  2) The protocol details are ambiguous,
]          in particular, the length octet either does or doesn't include
]          itself and the op code.  3) The extensions are not supported by
]          any known implementation.
] 
]          This protocol is now abandoned in favor of the DOMAIN protocol.
]          Further implementations of this protocol are not advised.
] 
]          Please discuss any plans for implementation or use of this
]          protocol with the contact.

	It's unfortunate that once a port is allocated, even to a protocol
which never had a snowball's chance in hell, it becomes difficult (or
nearly impossible) to call it back or retire it.  This one has been around
long enough to be called ancient.  We have a recent candidate in the SMTPS
(SMTP over SSL) which was allocated and then almost immediately withdrawn.
Not fast enough before Microsoft decided to stick it into exchange and
outlook, though...

	Unfortunately, IANA's assigned numbers list does not have a column
or an annotation that says "This number is assigned but this is a bad
thing - don't do this".

> >   Port 135:  This is listed as a DCE/RPC termination port.  Is this being
> >              used by the NT DCE/RPC stuff at all?

> >   epmap           135/tcp    DCE endpoint resolution
> >   epmap           135/udp    DCE endpoint resolution

> it is used, but I don't know exactly what for. I've seen packets to it
> though.

> >   Ports 137, These are assigned for both tcp & udp.  It appears, however,

> It seems that the IANA always does that. When I asked for a port for
> rsync they gave me UDP as well as TCP even though I only asked for
> TCP.

	That's now official policy buried on their web pages (or maybe it's
in the application for a port assignment) somewhere.  I vaguely remember
some confusion or some problems many years ago when two different services
got allocated the same port number, one on UDP and one on TCP.

> >                       "Set if this message was truncated because the
> >                        datagram carrying it would be greater than
> >                        576 bytes in length.  Use TCP to get the
> >                        information from the NetBIOS Name Server."

> I don't believe the MS implementations do this. Anyone see it?


	Mike
-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From cartegw at Eng.Auburn.EDU  Wed Sep 30 12:55:52 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:14 2003
Subject: more info on the nmbd problem
In-Reply-To: <19980929155351.A11001@bagheera.realminfo.com>
Message-ID: <Pine.SOL.3.96.980930075453.8210B-100000@orwell.eng.auburn.edu>

On Wed, 30 Sep 1998, Michael D. Ivey wrote:

> > Michael H. Warfield (mhw@wittsend.com)
> > Tue, 29 Sep 1998 10:29:56 -0400 (EDT) 
> > 
> >         I've been noticing for some time that I've been having a problem 
> > with nmbd and the latest snapshots. In particular, I've noticed that 
> > after a while, nmbd disappears. More specifically, on the system I have 
> 
> I'm having the same problem...here's my trace:
> 
> [root@alice source]# nmbd -D; ps axf |grep nmb
>  9816  ?  R    0:00 nmbd -D 
>  9818  ?  R    0:00  \_ nmbd -D 
> [root@alice source]# gdb
> GNU gdb 4.17
> Copyright 1998 Free Software Foundation, Inc.
> GDB is free software, covered by the GNU General Public License, and you are
> welcome to change it and/or distribute copies of it under certain conditions.
> Type "show copying" to see the conditions.
> There is absolutely no warranty for GDB.  Type "show warranty" for details.
> This GDB was configured as "i386-redhat-linux".
> (gdb) attach 9816
> Attaching to process 9816
> 0x400d983e in ?? ()
> (gdb) continue 
> Continuing.
> 
> Program received signal SIGSEGV, Segmentation fault.
> 0x8055eb2 in ?? ()
> (gdb) where
> #0  0x8055eb2 in ?? ()
> #1  0x80548bc in ?? ()
> #2  0x8054ffa in ?? ()
> #3  0x804ad1c in ?? ()
> #4  0x804b8c8 in ?? ()
> (gdb) 
> 

I think someone else send a backtrace from gdb as well, but just in
case...

Could you recompile gdb with the -g flag (for debugging support) and then
perform the stpes you just did.  See what happens.





j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )

From cartegw at Eng.Auburn.EDU  Wed Sep 30 12:51:26 1998
From: cartegw at Eng.Auburn.EDU (Gerald W. Carter)
Date: Tue Dec  2 03:02:14 2003
Subject: latest code fails with NTserver, works with 95
In-Reply-To: <19980929150354.A10741@bagheera.realminfo.com>
Message-ID: <Pine.SOL.3.96.980930074527.8210A-100000@orwell.eng.auburn.edu>

On Wed, 30 Sep 1998, Michael D. Ivey wrote:

> In my never-ending quest to get everything working, I
> now can't get my 2 NT servers into the domain.  The 95
> clients were having trouble at first, but now they seem
> ok.  NT boxes say "No domain controller available" and
> fail miserably.  'net view' doesn't show the server, even.
> 
> Any suggestions?  Also, please tell me how to update the
> smb.conf without kicking everyone off...I can't get a
> kill -HUP to do it.
> 

smbd will reread the configuration periodically.  Don't know exactly how
often.





j-
________________________________________________________________________
                            Gerald ( Jerry ) Carter	
Engineering Network Services                           Auburn University 
jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

       "...a hundred billion castaways looking for a home."
                                  - Sting "Message in a Bottle" ( 1979 )

From lkcl at switchboard.net  Wed Sep 30 14:34:58 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
In-Reply-To: <199809291659.LAA08694@unet.unet.umn.edu>
Message-ID: <Pine.NEB.3.96.980930153327.23009A-100000@cb1-gw.cb1.com>

On Wed, 30 Sep 1998, Christopher R. Hertel wrote:

> I've been poking around a bit and have a few questions.
> 
>   Port 42:   It appears as though this is being used for WINS replication. 
>              Any truth to the rumor?

don't know.
 
>   name             42/tcp    Host Name Server
>   name             42/udp    Host Name Server
>   nameserver       42/tcp    Host Name Server
>   nameserver       42/udp    Host Name Server
> 
>   Port 135:  This is listed as a DCE/RPC termination port.  Is this being
>              used by the NT DCE/RPC stuff at all?

it's the standard dce/rpc negotiation port.  yes, i recall, when running
winsmgr.exe 2 years ago that traffic occurred on this port, and then some
other traffic occurred on a high port number.

From lkcl at switchboard.net  Wed Sep 30 14:51:39 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
In-Reply-To: <19980929234006Z12670182-7510+12623@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980930154934.23009G-100000@cb1-gw.cb1.com>

> >                       "Set if this message was truncated because the
> >                        datagram carrying it would be greater than
> >                        576 bytes in length.  Use TCP to get the
> >                        information from the NetBIOS Name Server."
> 
> I don't believe the MS implementations do this. Anyone see it?

never seen it: microsoft's WINSMGR.HLP documentation states that they
deliberately limit the number of dynamic entries for multi-homed and
internet-group<1c> names to 28: this is the maximum that will fit in a 576
byte UDP packet.

they do not limit the number of _static_ internet-group<1c> names that can
be manually added into a ms-WINS database.

From lkcl at switchboard.net  Wed Sep 30 14:58:14 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:14 2003
Subject: dynamic-allocated string array "holder"
In-Reply-To: <19980930010009Z12669225-20449+12480@samba.anu.edu.au>
Message-ID: <Pine.NEB.3.96.980930155714.23009I-100000@cb1-gw.cb1.com>

On Wed, 30 Sep 1998, Andrew Tridgell wrote:

> depends, what is a "string array holder" ??

similar to what you have there:

string_array_init(struct string **str, int size)
stringarray_free (struct string **str);

would do the trick.
 
> the only dynamic string support we have is:
> 
> string_init()
> string_set()
> string_free()
> 
> it's used in a few places, particularly the loadparm code.
> 
> If you want arrays of strings then maybe build functions for them on
> top of the above functions?
> 
> 

<a href="mailto:lkcl@samba.anu.edu.au" > Luke Kenneth Casson Leighton  </a>
<a href="http://mailhost.cb1.com/~lkcl"> Samba and Network Development </a>
<a href="http://www.samba.co.uk"       > Samba and Network Consultancy </a>

From lkcl at switchboard.net  Wed Sep 30 15:10:15 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:14 2003
Subject: nmbd dying...
In-Reply-To: <361117C6.5B14296C@engr.sgi.com>
Message-ID: <Pine.NEB.3.96.980930160603.23009L-100000@cb1-gw.cb1.com>

> Dead code elimination. I removed a line such as
> 
> allowableaccount = IVAL(q, 0); q += 4;
> 
> as the variable "allowableaccount" was unused.
> I missed the bloody "q += 4;".
> 
> This is why multiple statements on one line are
> a *BUG*. I will eventualy remove *ALL* these (luke
> you have been warned :-).

oh, come on!  i don't do that!  it breaks my coding standards.  the only
curcumstances i will do something like that is:

if (some_bit_test(0)) set_something0;
if (some_bit_test(1)) set_something1;
if (some_bit_test(2)) set_something2;
...

those q += 4 lines are highly prevalent in ipc.c packet-parsing code.

also, for "informational" purposes, i would recommend that
allowableaccount is put back in, and printed out with a "DEBUG" statement.
if we re-examine the code and find we need to add a feature which uses
allowableaccount (whatever it is) then having this variable removed will
make that job potentially difficult to understand.

From lkcl at switchboard.net  Wed Sep 30 16:44:02 1998
From: lkcl at switchboard.net (Luke Kenneth Casson Leighton)
Date: Tue Dec  2 03:02:14 2003
Subject: warning in smbd/password.c: line 167
Message-ID: <Pine.NEB.3.96.980930174308.23438A-100000@cb1-gw.cb1.com>

uid_t and gid_t appear to be unsigned: therefore a test for < 0 gives a
warning.

From crh at NTS.Umn.EDU  Wed Sep 30 17:45:00 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
In-Reply-To: <199809301330.JAA22457@alcove.wittsend.com> from "Michael H. Warfield" at Sep 30, 98 11:32:09 pm
Message-ID: <199809301745.MAA01267@unet.unet.umn.edu>

> 	Aaaauuuugggghhhhh!!!!  It arises from the dead!

The suggestion was that Microsoft has stolen the body to use for its own 
dastardly purposes.  That is, since no one actually uses 42 (except in 
advanced infinite improbability algorithms), M$ might be using it for 
WINS replication.

I have *no* evidence that this is actually being done.  I have been too
burried in other floob to put boxes on a wire and see.  It was suggested
by one of our security folk, and I was just wondering if anyone knew. 

Sorry to have raised a panic.  It's safe to put away the silver bullets 
now.  (Well, maybe not...)

Chris -)-----

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From mhw at wittsend.com  Wed Sep 30 18:09:21 1998
From: mhw at wittsend.com (Michael H. Warfield)
Date: Tue Dec  2 03:02:14 2003
Subject: more info on the nmbd problem
In-Reply-To: <Pine.SOL.3.96.980930075453.8210B-100000@orwell.eng.auburn.edu> from "Gerald W. Carter" at "Oct 1, 1998  0: 1:53 am"
Message-ID: <199809301809.OAA28364@alcove.wittsend.com>

Gerald W. Carter enscribed thusly:
> On Wed, 30 Sep 1998, Michael D. Ivey wrote:

> > > Michael H. Warfield (mhw@wittsend.com)
> > > Tue, 29 Sep 1998 10:29:56 -0400 (EDT) 

> > >         I've been noticing for some time that I've been having a problem 
> > > with nmbd and the latest snapshots. In particular, I've noticed that 
> > > after a while, nmbd disappears. More specifically, on the system I have 

> > I'm having the same problem...here's my trace:

> > [root@alice source]# nmbd -D; ps axf |grep nmb
> >  9816  ?  R    0:00 nmbd -D 
> >  9818  ?  R    0:00  \_ nmbd -D 
> > [root@alice source]# gdb
> > GNU gdb 4.17
> > Copyright 1998 Free Software Foundation, Inc.
> > GDB is free software, covered by the GNU General Public License, and you are
> > welcome to change it and/or distribute copies of it under certain conditions.
> > Type "show copying" to see the conditions.
> > There is absolutely no warranty for GDB.  Type "show warranty" for details.
> > This GDB was configured as "i386-redhat-linux".
> > (gdb) attach 9816
> > Attaching to process 9816
> > 0x400d983e in ?? ()
> > (gdb) continue 
> > Continuing.

> > Program received signal SIGSEGV, Segmentation fault.
> > 0x8055eb2 in ?? ()
> > (gdb) where
> > #0  0x8055eb2 in ?? ()
> > #1  0x80548bc in ?? ()
> > #2  0x8054ffa in ?? ()
> > #3  0x804ad1c in ?? ()
> > #4  0x804b8c8 in ?? ()
> > (gdb) 


> I think someone else send a backtrace from gdb as well, but just in
> case...

	Yeah...  I was the one with the other backtrace.  Looks like
Jeremy found the problem and shot it.  Check out the latest CVS source
and try again.  It's now fixed for me.

> Could you recompile gdb with the -g flag (for debugging support) and then
> perform the stpes you just did.  See what happens.




> j-
> ________________________________________________________________________
>                             Gerald ( Jerry ) Carter	
> Engineering Network Services                           Auburn University 
> jerry@eng.auburn.edu             http://www.eng.auburn.edu/users/cartegw

>        "...a hundred billion castaways looking for a home."
>                                   - Sting "Message in a Bottle" ( 1979 )


-- 
 Michael H. Warfield    |  (770) 985-6132   |  mhw@WittsEnd.com
  (The Mad Wizard)      |  (770) 925-8248   |  http://www.wittsend.com/mhw/
  NIC whois:  MHW9      |  An optimist believes we live in the best of all
 PGP Key: 0xDF1DD471    |  possible worlds.  A pessimist is sure of it!
From jallison at cthulhu.engr.sgi.com  Wed Sep 30 17:21:05 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:14 2003
Subject: nmbd dying...
References: <Pine.NEB.3.96.980930160603.23009L-100000@cb1-gw.cb1.com>
Message-ID: <36126881.6C64D3A5@engr.sgi.com>

Luke Kenneth Casson Leighton wrote:

> 
> oh, come on!  i don't do that!  it breaks my coding standards. 

Yeah well the coding standards have a bug then :-).

Please - *no* multiline statements. They are a *nasty* bug
waiting to happen (IMHO).

> ...
> 
> those q += 4 lines are highly prevalent in ipc.c packet-parsing code.
> 

I know - and I'm splitting them into 2 lines wherever I find them.

> also, for "informational" purposes, i would recommend that
> allowableaccount is put back in, and printed out with a "DEBUG" statement.
> if we re-examine the code and find we need to add a feature which uses
> allowableaccount (whatever it is) then having this variable removed will
> make that job potentially difficult to understand.

True - time permitting I may add the code back in
bracketed with #ifdef 0's.

Cheers,

	Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From jallison at cthulhu.engr.sgi.com  Wed Sep 30 20:46:05 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
References: <199809301745.MAA01267@unet.unet.umn.edu>
Message-ID: <3612988D.E6A5460A@engr.sgi.com>

Christopher R. Hertel wrote:

> The suggestion was that Microsoft has stolen the body to use for its own
> dastardly purposes.  That is, since no one actually uses 42 (except in
> advanced infinite improbability algorithms), M$ might be using it for
> WINS replication.
> 

Nah - not a chance. They'll be using an undocumented
set of DCE/RPC calls. I know them. I know how they
think :-).

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------
From crh at NTS.Umn.EDU  Wed Sep 30 20:59:52 1998
From: crh at NTS.Umn.EDU (Christopher R. Hertel)
Date: Tue Dec  2 03:02:14 2003
Subject: Assigned numbers.
In-Reply-To: <199809301330.JAA22457@alcove.wittsend.com> from "Michael H. Warfield" at Sep 30, 98 11:32:09 pm
Message-ID: <199809302059.PAA29029@unet.unet.umn.edu>

FYI on port 42...

We don't block it at our border (but probably should).  Looking through
the last 21 days of packet-drop logs (packets which hit the floor because
they didn't pass the filter) I find *one* instance of an outside machine
attempting to connect from port 42 to port 137/udp.  There is only one
other reference to port 42 and it's not interesting (someone talking to
themselves). 

If I get a chance, I'll look at flows.

Chris -)-----

-- 
Christopher R. Hertel -)-----                   University of Minnesota
crh@nts.umn.edu              Networking and Telecommunications Services
From jallison at cthulhu.engr.sgi.com  Wed Sep 30 19:28:31 1998
From: jallison at cthulhu.engr.sgi.com (Jeremy Allison)
Date: Tue Dec  2 03:02:14 2003
Subject: Strange problem writing files (PR#10099)
References: <19980930132121Z12670658-7009+12686@samba.anu.edu.au>
Message-ID: <3612865F.6019B4DA@engr.sgi.com>

anders.blomdell@control.lth.se wrote:
> 
> One of our NT programs (a version of latex) has problem opening files with names like '.\dir\file' if 'protocol = NT1' in smb.conf. If 'protocol = LANMAN2', everything works OK. This is what samba gets in the different cases (the problem seems to be that NT1 does not get the full path):
> 

Anders,

	I think I've fixed this. Could you please do a CVS
update and try again.

I'm CC:ing this to Samba-Tech and Samba-NTDOM as people
on these lists will also be interested in this fix.

As Samba now allows ChangeNotify it allows NT clients
to open directories. The NTCreateX call can take a
directory FID as well as a filename. If there is a
valid directory FID then the filename is supposed to
be parsed *relative to that directory*.

When I originally implemented NT SMB's I didn't allow
directory opens and so this wasn't an issue. Now I do,
but I forgot to implement the directory-relative opens.

I've now implemented these directory-relative opens
but I cannot get NT to generate such an open to check
if my code is correct.

If you could re-run your test case and confirm that
this fixes the promblem I'd be really grateful. This
also may be the key to the problem someone was having
with execing unzip from a program also....

Cheers,

	Jeremy.


-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------