bbaumbach/samba.git
24 years agoThis commit was manufactured by cvs2svn to create tag samba-misc-tags/tng/release-alpha-1-0
cvs2svn Import User [Fri, 17 Mar 2000 19:32:14 +0000 (19:32 +0000)]
This commit was manufactured by cvs2svn to create tag
'release-alpha-1-0'.

24 years agopreparing for release of alpha.1.0
Samba Release Account [Fri, 17 Mar 2000 19:32:13 +0000 (19:32 +0000)]
preparing for release of alpha.1.0

24 years agook, this one's weird. firstly, sam_set_userinfo doesn't align the
Luke Leighton [Fri, 17 Mar 2000 19:15:50 +0000 (19:15 +0000)]
ok, this one's weird.  firstly, sam_set_userinfo doesn't align the
2nd info level at the container, but sam_set_userinfo2 *does*.

secondly, it _looks_ like there are two revisions for SAM database
entries, where some unknown trash is _not_ placed in the structure
(at the end: unknown_6 and padding4) if the revision number (unknown_5)
is zero.

have to keep an eye on this one over the next year or so, to make sure
all cases are caught.

third, it _looks_ like if you do a password set that the account disabled
acb_info bit is automatically unset, but i cannot confirm that right
now, i'll just have to wing-it (in the next commit)

24 years agowhen a GETDC datagram is received, if the dgram type is UNIQUE instead of
Luke Leighton [Fri, 17 Mar 2000 18:30:10 +0000 (18:30 +0000)]
when a GETDC datagram is received, if the dgram type is UNIQUE instead of
GROUP, it is processed *differently*, there is a domain-name string in it,
so i skip 22 bytes and _then_ read the version number etc.

this must go in 2.0.x and cvs main (oh, and i'm not going to do it, i'm
not "trusted" enough to do the job "correctly").

this will be the only notification that this is needed in 2.0.x and cvs
main branches.

24 years agoFrom Elrond@Wunder-Nett.org Sat Mar 18 04:49:11 2000
Luke Leighton [Fri, 17 Mar 2000 18:03:47 +0000 (18:03 +0000)]
From Elrond@Wunder-Nett.org Sat Mar 18 04:49:11 2000
Date: Fri, 17 Mar 2000 16:02:28 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: Re: SVC_UNKNOWN_3

On Fri, Mar 17, 2000 at 08:27:49AM +1100, Luke Kenneth Casson Leighton wrote:
> it's probably a delete service :-) :-)
>
> check your box!!!
[...]

Hi Luke,

You pass in a Service-Control-Manager-Handle, not a
service-handle, so it can't be service-delete. It could of
course be scman-del... which would be quite crazy.
(includind the fact, that this is done by srvmgr, which
hasn't any means to add/del a service...)

But it was all my fault. I copied cli_svc.c:svc_close()
over to svc_unknown_3... and guess what... I changed
everything except the Opcode...

After that I was more successful: You pass in a scman_hnd,
and get back some other handle...I don't know yet, what it
is good for... I added an appropiate server-side
implementation and srvmgr.exe got a bit further, but now it
calls opcode 11 on a service-handle. Maybe more soon.
(I'm pressing the "Startup"-button in Services in srvmgr)

The patch contains:
- complete server and client-side implementation of
  SVC_UNKNOWN_3 (yet no idea, what it is, but we have it)
- --with-privatedir=... from HEAD is now completely in TNG
- merged some things from HEAD into lib/time.c
  PLEASE DON'T run indent on lib/time.c !
[lkcl: okie]
- dito for param/loadparm.c
  PLEASE DON'T run indent on param/loadparm.c !
[lkcl: okie]

    Elrond

24 years agoindent rules suggested by elrond. comments added into nmbd packet tracing.
Luke Leighton [Fri, 17 Mar 2000 17:49:47 +0000 (17:49 +0000)]
indent rules suggested by elrond.  comments added into nmbd packet tracing.

24 years agomaking sure that req chal would not be affected by having prs_align()
Luke Leighton [Fri, 17 Mar 2000 17:48:19 +0000 (17:48 +0000)]
making sure that req chal would not be affected by having prs_align()
added to *inside* smb_io_unistr2().  this is the _only_ place where
alignment rules are broken.

24 years agoafsticket defines
Luke Leighton [Fri, 17 Mar 2000 17:36:40 +0000 (17:36 +0000)]
afsticket defines

24 years agoFrom Elrond@Wunder-Nett.org Fri Mar 17 08:34:21 2000
Luke Leighton [Thu, 16 Mar 2000 22:06:53 +0000 (22:06 +0000)]
From Elrond@Wunder-Nett.org Fri Mar 17 08:34:21 2000
Date: Thu, 16 Mar 2000 16:29:53 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: SVC_UNKNOWN_3

Hi Luke,

I played a bit with srvmgr around and it did a opnum=3 on
svcctl, so I tried to decode the request. For me it looked
lit it only consists of a sc_man_hnd. So I added some
parsing code and an "svcunk3" to rpcclient, so I could ask
the NT-box for the answer. The answer were 20 NUL-bytes. I
guess, this is a closed handle and a status. Also the
following svc_close(sc_man_hnd) seemed to fail somehow (I
didn't realy see the fault-pdu or so... but I did not see
an svc_io_r_close_hnd either.)

[lkcl: i suspect that this is a "delete service" call,
       which is why the close fails because the handle
       is alread closed!]

All this looks to me like opnum=3 is something like
SVC_CLOSE2 or so.

Could you apply the patch and verify that?

I also did the following:
- modified configure.in, so it is more like HEAD (still lot
  to do)
- fixed some issues in groupdb/builtinunix.c:
  When I created aliases with domain alias map, all these
  aliases also existed in the Builtin domain, so usrmgr
  showed them all doubled. (might be interesting for
  HEAD...)
- made global_sid_S_1_1_0 static, one should use
  global_sid_everyone instead.
- continued my share_get_info-effort by writing *_free_*
  for all the fun. Yet have to use them.

    Elrond

24 years agoran indent on afsticket.c because its format was confusing make proto.
Luke Leighton [Thu, 16 Mar 2000 21:48:59 +0000 (21:48 +0000)]
ran indent on afsticket.c because its format was confusing make proto.

24 years agopatch from johan hedin <johanh@fusion.kth.se> for AFS ticket support.
Luke Leighton [Thu, 16 Mar 2000 21:38:20 +0000 (21:38 +0000)]
patch from johan hedin <johanh@fusion.kth.se> for AFS ticket support.

24 years agofixed word-order issue in password set and password change.
Luke Leighton [Wed, 15 Mar 2000 23:47:51 +0000 (23:47 +0000)]
fixed word-order issue in password set and password change.

24 years agopreparing for release of alpha.0.16
Samba Release Account [Wed, 15 Mar 2000 23:34:43 +0000 (23:34 +0000)]
preparing for release of alpha.0.16

24 years agogratuitous indent on files modified in the last two days.
Luke Leighton [Wed, 15 Mar 2000 23:28:14 +0000 (23:28 +0000)]
gratuitous indent on files modified in the last two days.

24 years ago1) SAM_USER_INFO_12 what i thought was acb_info isn't. this may be
Luke Leighton [Wed, 15 Mar 2000 02:08:59 +0000 (02:08 +0000)]
1) SAM_USER_INFO_12 what i thought was acb_info isn't.  this may be
responsible for some of the password set failure problems: treating
the wrong field as an acb_info would result in rejection of password
sets.

2) in attempting to track down server-password-set problems i noticed
that the "security = domain" mode sets the password incorrectly.

24 years agosneaking suspicion that the ProductOptions key is being reported the wrong
Luke Leighton [Tue, 14 Mar 2000 23:48:21 +0000 (23:48 +0000)]
sneaking suspicion that the ProductOptions key is being reported the wrong
way round (BDC instead of PDC, PDC instead of BDC).

24 years agopatch from michael breuer for irix.
Luke Leighton [Tue, 14 Mar 2000 21:19:44 +0000 (21:19 +0000)]
patch from michael breuer for irix.

24 years agofinally decided to increase debug log level for "socket connect failed"
Luke Leighton [Tue, 14 Mar 2000 21:11:06 +0000 (21:11 +0000)]
finally decided to increase debug log level for "socket connect failed"
message from level 0 to level 1.

24 years agoFrom Elrond@Wunder-Nett.org Wed Mar 15 06:14:36 2000
Luke Leighton [Tue, 14 Mar 2000 20:53:28 +0000 (20:53 +0000)]
From Elrond@Wunder-Nett.org Wed Mar 15 06:14:36 2000
Date: Tue, 14 Mar 2000 19:23:04 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: Re: _lsa_lookup_names

On Tue, Mar 14, 2000 at 08:20:26AM +1100, Luke Kenneth Casson Leighton wrote:
> elrond, this function isn't used, yet.

I did the conversion this morning and the diff is
appended.

[lkcl: thanks elrond]

I also added some more handle-names. Hope someone else
likes them. ;)

[lkcl: me! me! :)]

    Elrond

24 years agoAnother #ifndef #define #endif bug.
Tim Potter [Tue, 14 Mar 2000 03:22:29 +0000 (03:22 +0000)]
Another #ifndef #define #endif bug.

24 years agoSpotted error in #ifndef _XXX #define _XXX #endif thing.
Tim Potter [Tue, 14 Mar 2000 03:18:33 +0000 (03:18 +0000)]
Spotted error in #ifndef _XXX #define _XXX #endif thing.

24 years agopreparing for release of alpha.0.15
Samba Release Account [Mon, 13 Mar 2000 20:51:34 +0000 (20:51 +0000)]
preparing for release of alpha.0.15

24 years agoFrom Elrond@Wunder-Nett.org Tue Mar 14 06:07:16 2000
Luke Leighton [Mon, 13 Mar 2000 20:21:58 +0000 (20:21 +0000)]
From Elrond@Wunder-Nett.org Tue Mar 14 06:07:16 2000
Date: Mon, 13 Mar 2000 18:26:00 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: more cleanup

- removed most of the old --with-lib*-stuff, we're using
  libtool for this
- srv_samr_passdb.c: Added some handle-names for debugging
  (helped me somehow to track netlogond-probs)
- made a whole bunch of functions in display_srv.c and
  display_sam.c
- Added a bunch of error-checking in lsarpcd, it looks like
  it tries to check/change the trust-accoutn password as a
  normal user (or even guest)

[lkcl: checks can't hurt: _lsa_open_secret is called, direct, not
       client-side marshalling wrapper, lsa_open_secret, etc]

    Elrond

24 years agoFrom Elrond@Wunder-Nett.org Tue Mar 14 06:01:05 2000
Luke Leighton [Mon, 13 Mar 2000 19:09:13 +0000 (19:09 +0000)]
From Elrond@Wunder-Nett.org Tue Mar 14 06:01:05 2000
Date: Mon, 13 Mar 2000 18:16:51 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: netlogond: crash and ggroup membership

Hi Luke,

netlogond crashes, when my workstation tries to change its
trust-account password. I tracked this to some point and
fixed those points. Since my workstation only tries to
change the pw every 15 minutes, testing is quite
timeconsuming and I didn't yet test my last fix, that I did
in this area.

After some funny errors (most notable: root is in Domain
Admins, but I couldn't change any thing in the
network-settings), I noticed, that netlogond didn't
transmit _any_ group memberships, when I logged into the
workstation, I found this and fixed it.

Could you please take a look at the following functions:
- direct_samr_userinfo
- _net_srv_pwset

I modified both, but I'm not too sure, I did it the right
way.

[lkcl: looks good, elrond.  explains why groups didn't get set up!]

24 years agoAdded standard shell command thing using '!' as first character.
Tim Potter [Mon, 13 Mar 2000 01:32:48 +0000 (01:32 +0000)]
Added standard shell command thing using '!' as first character.

24 years agoFrom Elrond@Wunder-Nett.org Mon Mar 13 10:35:03 2000
Luke Leighton [Mon, 13 Mar 2000 00:01:04 +0000 (00:01 +0000)]
From Elrond@Wunder-Nett.org Mon Mar 13 10:35:03 2000
Date: Sat, 11 Mar 2000 17:57:54 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: more cleanup, global_sid_builtin

Hi Luke,

- more cleanup on parse_samr.c
- micro-merge from HEAD for Makefile
- use smb_io_rpc_uuid-return-value in smb_io_pol_hnd
- replaced nearly all &global_sid_S_1_5_20 by
  global_sid_builtin (inspired from a post by tim, he
  didn't apply this patch)
- closing an lsa-handle in rpcclient->createuser only, if
  also opened

24 years agopreparing for release of alpha.0.14
Samba Release Account [Sat, 11 Mar 2000 01:01:17 +0000 (01:01 +0000)]
preparing for release of alpha.0.14

24 years agoremoved check_vuser_ok() as it restricts things too much, unexpectedly.
Luke Leighton [Sat, 11 Mar 2000 00:58:12 +0000 (00:58 +0000)]
removed check_vuser_ok() as it restricts things too much, unexpectedly.
"invalid users = root" will stop any msrpc daemons working, including
being able to add users to the sam database (!!!! not good!)

24 years agotim, become_xxx shouldn't be called in winbindd.
Luke Leighton [Sat, 11 Mar 2000 00:37:36 +0000 (00:37 +0000)]
tim, become_xxx shouldn't be called in winbindd.

24 years agopreparing for release of alpha.0.13
Samba Release Account [Fri, 10 Mar 2000 21:59:21 +0000 (21:59 +0000)]
preparing for release of alpha.0.13

24 years agoalways check that crypt may return NULL
Luke Leighton [Fri, 10 Mar 2000 21:54:22 +0000 (21:54 +0000)]
always check that crypt may return NULL

24 years agoadded debug reporting that tells you if "valid users" succeeds or fails,
Luke Leighton [Fri, 10 Mar 2000 21:43:20 +0000 (21:43 +0000)]
added debug reporting that tells you if "valid users" succeeds or fails,
on msrpc daemons.

24 years ago- spoolssd linked against shared libubiqx, not static.
Luke Leighton [Fri, 10 Mar 2000 19:53:33 +0000 (19:53 +0000)]
- spoolssd linked against shared libubiqx, not static.
- finaly fixed %U (at least for me), it was one line in
  passdb/sampass.c
- started to merge some configure things from HEAD (not
  that we currently realy need it)
- added some debugging to lsarpcd (policy-handles), so
  maybe we'll get those mysterious crashes with ldap;
  made some functions static
- prs_set_offset() (from J.F.) didn't prs_grow enough. I
  don't know, if he needs prs_set_offset to actually grow
  any buffer (for marshalling), but maybe it helps
- parse_samr.c cleanup:
  + rewrote most of the memory-things to use g_new
  + removed useless prs_align after smb_io_pol_hnd()
  + converted more memcpy(&q->pol, pol, sizeof()) to
    q->pol = *pol;
  this all made parse_samr.c about 100 lines smaller.

24 years agopatch from elrond for configure tests.
Luke Leighton [Fri, 10 Mar 2000 18:15:46 +0000 (18:15 +0000)]
patch from elrond for configure tests.

24 years agoinitialiase USER_INFO_CTR to null
Luke Leighton [Fri, 10 Mar 2000 02:10:43 +0000 (02:10 +0000)]
initialiase USER_INFO_CTR to null

24 years agouse write_socket() not a direct write()
Luke Leighton [Fri, 10 Mar 2000 02:01:40 +0000 (02:01 +0000)]
use write_socket() not a direct write()

24 years agopatch from elrond: configure cleanups
Luke Leighton [Thu, 9 Mar 2000 20:00:37 +0000 (20:00 +0000)]
patch from elrond: configure cleanups

24 years agologin validation level 2 and 3 negotiated from netr_auth2 neg_flags.
Luke Leighton [Thu, 9 Mar 2000 04:30:57 +0000 (04:30 +0000)]
login validation level 2 and 3 negotiated from netr_auth2 neg_flags.

24 years agopreparing for release of alpha.0.12
Samba Release Account [Wed, 8 Mar 2000 23:34:02 +0000 (23:34 +0000)]
preparing for release of alpha.0.12

24 years agojames f. hranicky pointed out that a standalone workstation couldn't
Luke Leighton [Wed, 8 Mar 2000 22:50:33 +0000 (22:50 +0000)]
james f. hranicky pointed out that a standalone workstation couldn't
do a net use \\samba-tng\share /user:tnguser

you had to do net use \\samba-tng\share /user:samba-tngdomainname\tnguser

i solved this by putting "if domain name is unrecognised, use local SAM"
where domain name, by default on standalone workstations, is the name
of the workstation itself.

24 years agoindent defaults
Luke Leighton [Wed, 8 Mar 2000 22:08:33 +0000 (22:08 +0000)]
indent defaults

24 years agoran indent.
Luke Leighton [Wed, 8 Mar 2000 22:07:42 +0000 (22:07 +0000)]
ran indent.

24 years ago#include missing in srv_spoolss.c
Luke Leighton [Wed, 8 Mar 2000 21:01:42 +0000 (21:01 +0000)]
#include missing in srv_spoolss.c

24 years agofound two bugs in lsalookup* - sid_name_use type is uint16 - by bouncing
Luke Leighton [Wed, 8 Mar 2000 20:49:08 +0000 (20:49 +0000)]
found two bugs in lsalookup* - sid_name_use type is uint16 - by bouncing
off sun's cascade box.  HOORAY!

good suggestion, elrond.

24 years agoattempting to get trust relationship setup going again.
Luke Leighton [Wed, 8 Mar 2000 09:24:01 +0000 (09:24 +0000)]
attempting to get trust relationship setup going again.

24 years agopreparing for release of alpha.0.11
Samba Release Account [Wed, 8 Mar 2000 08:09:10 +0000 (08:09 +0000)]
preparing for release of alpha.0.11

24 years ago- acb_info is a 32-bit field in samr stuff (oops!!) sun cascade with
Luke Leighton [Wed, 8 Mar 2000 07:59:55 +0000 (07:59 +0000)]
- acb_info is a 32-bit field in samr stuff (oops!!)  sun cascade with
a different byte order from nt is VERY useful, it can be used to
identify the size of structure members because the byte order will
be swapped, in comparative traces.

- auth_resp in NETLOGON is a parameter to net_r_sam_logon, NOT part of
 the NET_USER_INFO struct.

- domain_client_validate was grabbing the secret $MACHINE.ACC, then
  the unicode string needed to be byte-swapped before being passwd
  to nt_owf_genW.

24 years ago1) logon hours _is_ a 32-bit not a 16-bit. oops
Luke Leighton [Wed, 8 Mar 2000 05:24:12 +0000 (05:24 +0000)]
1) logon hours _is_ a 32-bit not a 16-bit.  oops

2) logon to local workstation when it is a member of a domain is also
   allowed, by making domain_client_validate accept our own localservername
   as the domain name (a la MYSERVERNAME\user)

24 years agoyet another stage in net user info level 2. server-side and client-side
Luke Leighton [Wed, 8 Mar 2000 01:12:38 +0000 (01:12 +0000)]
yet another stage in net user info level 2.  server-side and client-side
support but no neg_flags used, yet.

24 years agonext stage of having a NET_USER_INFO_CTR
Luke Leighton [Tue, 7 Mar 2000 23:54:55 +0000 (23:54 +0000)]
next stage of having a NET_USER_INFO_CTR

24 years agocreated a NET_USER_INFO_2.
Luke Leighton [Tue, 7 Mar 2000 23:07:04 +0000 (23:07 +0000)]
created a NET_USER_INFO_2.

24 years agopreparing for release of alpha.0.10
Samba Release Account [Tue, 7 Mar 2000 20:15:38 +0000 (20:15 +0000)]
preparing for release of alpha.0.10

24 years agojean-francois' tng patch, updated. cleaned up some of the
Luke Leighton [Tue, 7 Mar 2000 20:07:00 +0000 (20:07 +0000)]
jean-francois' tng patch, updated.  cleaned up some of the
global / local variable clashes (found with ./configure.developer)

24 years agomore of those wonderful connection-related issues, found by rpctorture
Luke Leighton [Tue, 7 Mar 2000 19:26:16 +0000 (19:26 +0000)]
more of those wonderful connection-related issues, found by rpctorture

24 years agopatch from jf who was having difficulty doing cvs commits on TNG.
Luke Leighton [Tue, 7 Mar 2000 18:29:17 +0000 (18:29 +0000)]
patch from jf who was having difficulty doing cvs commits on TNG.

24 years agomake proto
Luke Leighton [Tue, 7 Mar 2000 18:24:33 +0000 (18:24 +0000)]
make proto

24 years agoFrom Elrond@Wunder-Nett.org Wed Mar 8 05:16:00 2000
Luke Leighton [Tue, 7 Mar 2000 18:18:44 +0000 (18:18 +0000)]
From Elrond@Wunder-Nett.org Wed Mar  8 05:16:00 2000
Date: Tue, 7 Mar 2000 18:03:39 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: cleanup: sam_enum_domains (passdb) and share info 502

Hi Luke,

- munged around with share_info_502, cause I want to
  integrate ShareGetInfo and ShareEnum more, so both
  support the same info-levels and there isn't so much
  doubled code. (note: I want that, it's not currently
  anywhere near there.)
- moved enumdomains from lib/sids.c to
  samrd/srv_samr_passdb.c, it's only used there.

24 years agoPOL_HND_SIZE not defined any more
Luke Leighton [Tue, 7 Mar 2000 18:12:42 +0000 (18:12 +0000)]
POL_HND_SIZE not defined any more

24 years agoset user info 2 - 0x12 sorted out, client and server side. i hope.
Luke Leighton [Tue, 7 Mar 2000 05:17:04 +0000 (05:17 +0000)]
set user info 2 - 0x12 sorted out, client and server side.  i hope.

24 years agoattempting to get samr_set_userinfo2 info level 0x12 working.
Luke Leighton [Tue, 7 Mar 2000 02:57:08 +0000 (02:57 +0000)]
attempting to get samr_set_userinfo2 info level 0x12 working.

24 years agosolving issue of cascade joining domain. it's decided that the
Luke Leighton [Tue, 7 Mar 2000 00:16:37 +0000 (00:16 +0000)]
solving issue of cascade joining domain.  it's decided that the
password set is to be an info level 0x12, due to get_usrdom_pwinfo.  eh???

24 years agoinfo level 23 password change also has alignment issues
Luke Leighton [Mon, 6 Mar 2000 23:30:03 +0000 (23:30 +0000)]
info level 23 password change also has alignment issues

24 years agoset_userinfo at level 23 was wrong.
Luke Leighton [Mon, 6 Mar 2000 23:25:18 +0000 (23:25 +0000)]
set_userinfo at level 23 was wrong.

24 years agooi! who changed set_tdbsam to set_tdbsid without also changing all
Luke Leighton [Mon, 6 Mar 2000 20:01:47 +0000 (20:01 +0000)]
oi!  who changed set_tdbsam to set_tdbsid without also changing all
occurrences of get_tdbsam to get_tdbsid!!!!

who added global_sid_S_1_1 to tdb_samr_connect?

24 years agopatch from elrond for libtool hack
Luke Leighton [Mon, 6 Mar 2000 18:23:36 +0000 (18:23 +0000)]
patch from elrond for libtool hack

24 years agopreparing for release of alpha.0.9
Samba Release Account [Mon, 6 Mar 2000 10:40:35 +0000 (10:40 +0000)]
preparing for release of alpha.0.9

24 years agofascinating: may be the answer to some of the login problems: byte ordering
Luke Leighton [Mon, 6 Mar 2000 10:33:12 +0000 (10:33 +0000)]
fascinating: may be the answer to some of the login problems: byte ordering
in passwords.  AAGH!

24 years agointeresting. had to add code to stop cascade's SMBtconX encrypted password
Luke Leighton [Mon, 6 Mar 2000 09:34:19 +0000 (09:34 +0000)]
interesting.  had to add code to stop cascade's SMBtconX encrypted password
from being used, i have a) no idea if it's valid b) no idea how to decode
it!

24 years agohad to create a table referencing the incoming DCE/RPC context_id to an
Luke Leighton [Mon, 6 Mar 2000 08:16:09 +0000 (08:16 +0000)]
had to create a table referencing the incoming DCE/RPC context_id to an
internal context id.

also fixing issues where i had earlier removed POL_HND_SIZE #define.

24 years agook, got bigendian byte order requests working.
Luke Leighton [Mon, 6 Mar 2000 07:23:36 +0000 (07:23 +0000)]
ok, got bigendian byte order requests working.

24 years agoturns out that POLICY_HND actually contains an RPC_UUID. only found this
Luke Leighton [Mon, 6 Mar 2000 05:13:26 +0000 (05:13 +0000)]
turns out that POLICY_HND actually contains an RPC_UUID.  only found this
out by receiving little-endian DCE/RPC packets and sending big-endian
DCE/RPC packets, of course the POLICY_HND was wrong...

24 years agofixing careful alignment IVAL const issues
Luke Leighton [Mon, 6 Mar 2000 04:09:20 +0000 (04:09 +0000)]
fixing careful alignment IVAL const issues

24 years agofirst connectathon 2000 tests produces results!
Luke Leighton [Mon, 6 Mar 2000 03:25:59 +0000 (03:25 +0000)]
first connectathon 2000 tests produces results!

cascade shows that pipe names are case-insensitive.  they send all uppercase.

24 years agocareful alignment needs const versions of SVAL, IVAL etc.
Luke Leighton [Mon, 6 Mar 2000 03:18:20 +0000 (03:18 +0000)]
careful alignment needs const versions of SVAL, IVAL etc.

24 years agopreparing for release of alpha.0.8
Samba Release Account [Fri, 3 Mar 2000 20:12:15 +0000 (20:12 +0000)]
preparing for release of alpha.0.8

24 years agounbecome_root()( fix - only actually do an unbecome_root
Luke Leighton [Fri, 3 Mar 2000 20:09:04 +0000 (20:09 +0000)]
unbecome_root()( fix - only actually do an unbecome_root
when root depth reaches zero again.

24 years agoCheck for null tdb context in tdb_exists()
Tim Potter [Wed, 1 Mar 2000 01:11:09 +0000 (01:11 +0000)]
Check for null tdb context in tdb_exists()

24 years agoMerge of changes from HEAD branch.
Tim Potter [Tue, 29 Feb 2000 22:52:28 +0000 (22:52 +0000)]
Merge of changes from HEAD branch.

24 years agoallow nested become_root() calls. this is a hack.
Luke Leighton [Tue, 29 Feb 2000 18:04:51 +0000 (18:04 +0000)]
allow nested become_root() calls.  this is a hack.

24 years agoDCE/RPC header - pack type tests
Luke Leighton [Mon, 28 Feb 2000 20:07:44 +0000 (20:07 +0000)]
DCE/RPC header - pack type tests

24 years agoFrom Elrond@Wunder-Nett.org Tue Feb 29 05:53:10 2000
Luke Leighton [Mon, 28 Feb 2000 19:41:09 +0000 (19:41 +0000)]
From Elrond@Wunder-Nett.org Tue Feb 29 05:53:10 2000
Date: Mon, 28 Feb 2000 17:29:48 +0100
From: Elrond <Elrond@Wunder-Nett.org>
To: Luke Kenneth Casson Leighton <lkcl@samba.org>
Subject: _samr_query_sec_obj

Hi Luke,

- moved the SD-generation code out of
  samrd/srv_samr_sam_tdb.c:_samr_query_sec_obj into its own
  samrd/samr_util.c
- updated the SD to have the same order and include
  Builtin\Account Operators, as seen on an nt4sp4 ws.
- used this function in samrpassd (which was broken here
  and usrmgr didn't work) and samrtdbd (for consistency)
  (I can't compile-check samrnt5ldap, so I will send Luke
  Howard a patch to inspect.)
- clean-up: Mainly policy-handles... but also other stuff
- debugging and type-checking of data associated with
  POLICY_HND in samrtdbd
  (I suspect, there's a problem in sam_query_sec_obj, see
  the new WARNING, that appears.)
- rpcclient/samedit: enumaliases has new option:
  [-d Domain], so I could do "enumaliases -d Builtin"

After this patch my usrmgr worked better, but now it looks
like it is thinking, it talks to a WS instead of a server.
I can't see the domain-groups, only aliases, and in the
user-properties, there are only 3 buttons for "Groups",
"Profiles" and "RAS", the others are missing... I tried a
long time to find out about it... I failed, so could you
have a look at this?

[lkcl: this because winregd is returning "i am a wksta"]

Okay, and another thing:

You seemed to look for the bit on the user-objects, that
allows password-changing: I think, it must be 0x0004,
cause:
Everyone has: 0x2035b
Normal user : 0x20044

0x20040 is included in the bits, everyone has, 0x0004 is
left over....

Just some thoughts...

    Elrond

24 years agoreversed irix include standard.h patch because it breaks other irix
Luke Leighton [Mon, 28 Feb 2000 15:56:46 +0000 (15:56 +0000)]
reversed irix include standard.h patch because it breaks other irix
configures.

24 years agopreparing for release of alpha.0.7
Samba Release Account [Mon, 28 Feb 2000 01:00:35 +0000 (01:00 +0000)]
preparing for release of alpha.0.7

24 years agoThis commit was manufactured by cvs2svn to create branch 'SAMBA_TNG'.
cvs2svn Import User [Mon, 28 Feb 2000 00:51:55 +0000 (00:51 +0000)]
This commit was manufactured by cvs2svn to create branch 'SAMBA_TNG'.

24 years agoMakefile for tdb directory.
Tim Potter [Mon, 28 Feb 2000 00:51:53 +0000 (00:51 +0000)]
Makefile for tdb directory.

24 years agoOops - put back filename argument to tdb_open().
Tim Potter [Mon, 28 Feb 2000 00:38:45 +0000 (00:38 +0000)]
Oops - put back filename argument to tdb_open().

24 years agoModifications to tdb_traverse() arguments to remove compile warnings.
Tim Potter [Mon, 28 Feb 2000 00:37:13 +0000 (00:37 +0000)]
Modifications to tdb_traverse() arguments to remove compile warnings.

24 years agoAllow tdb databases to be created "in memory" without having a file
Tim Potter [Mon, 28 Feb 2000 00:25:37 +0000 (00:25 +0000)]
Allow tdb databases to be created "in memory" without having a file
associated with them.  tdb can then be used as a hash table data
structure.

24 years agoAdded tdbtool to .cvsignore list.
Tim Potter [Mon, 28 Feb 2000 00:22:45 +0000 (00:22 +0000)]
Added tdbtool to .cvsignore list.

24 years agoadding pack_type data representation. tested as-was (i.e. not against
Luke Leighton [Mon, 28 Feb 2000 00:00:53 +0000 (00:00 +0000)]
adding pack_type data representation.  tested as-was (i.e. not against
a non-intel-byte-order m/c).

24 years agoadded bind nack to server-side.
Luke Leighton [Sun, 27 Feb 2000 19:12:07 +0000 (19:12 +0000)]
added bind nack to server-side.

jeremy, the majority of "negative" responses are actually "fault" pdus.

the only circumstances in which a bind nack is returned is if there is
something wrong with a bind request.  e.g the NTLMSSP auth bind-request
(negotiate) stage, which contains the client hostname and client domain,
contains an unrecognised name such as NULL.

e.g the NETSEC (netlogon secure channel) auth bind-request stage which
again happens (coincidentally) to contain the client hostname and client
domain, contains a hostname/domainname tuple for which no NetrReqChal+
NetrAuth2 with flags 0x400001ff has just previously been done.

i.e: NetrReqChal("\\myserver", "mydomain", ...);
NetrAuth2("\\myserver", mydomain, 0x400001ff, ...);
[now do netsec]
NetSecBindRequest("\\somestupidservername", "totalgarbagedomainname")
this should be rejected with a bind NACK.

24 years agotrying to track down why setgroups() fails.
Luke Leighton [Sun, 27 Feb 2000 18:11:29 +0000 (18:11 +0000)]
trying to track down why setgroups() fails.

24 years agotesting password change showed up some bugs in NTLMSSP calcs, plus one in
Luke Leighton [Sun, 27 Feb 2000 17:52:19 +0000 (17:52 +0000)]
testing password change showed up some bugs in NTLMSSP calcs, plus one in
cli_samr.c.  from the NetrSamLogon the first 8 bytes of the
LM# are received, i forgot to pass this over.

24 years agoadded enumprintprocessordatatypes
Jean-François Micouleau [Sat, 26 Feb 2000 23:01:02 +0000 (23:01 +0000)]
added enumprintprocessordatatypes
now NT is happy and the "always send data in RAW mode" is checked

J.F.

24 years agorewrote enumprinterdata. still a bug in it but reproducing it hard and
Jean-François Micouleau [Sat, 26 Feb 2000 22:22:24 +0000 (22:22 +0000)]
rewrote enumprinterdata. still a bug in it but reproducing it hard and
borring.

I need a client test program urgently!!!

rewrote setprinter, doesn't coredump anymore, and no memleak.

J.F.

24 years agoclient/client.c:
Jeremy Allison [Fri, 25 Feb 2000 22:25:25 +0000 (22:25 +0000)]
client/client.c:
libsmb/clientgen.c: Fixes for Win2k smbclient browsing.
Other fixes implement smbpasswd -x user to delete users. Also allows swat
to do the same.
Jeremy.

24 years agopatch from michael breuer for irix TCP_NODELAY include
Luke Leighton [Fri, 25 Feb 2000 21:07:31 +0000 (21:07 +0000)]
patch from michael breuer for irix TCP_NODELAY include

24 years agoconst issues
Luke Leighton [Fri, 25 Feb 2000 05:47:45 +0000 (05:47 +0000)]
const issues

24 years agoremove unused var
Luke Leighton [Fri, 25 Feb 2000 05:44:46 +0000 (05:44 +0000)]
remove unused var

24 years agotim, after speaking with andrew he noticed that you were calling become_root()
Luke Leighton [Fri, 25 Feb 2000 05:43:49 +0000 (05:43 +0000)]
tim, after speaking with andrew he noticed that you were calling become_root()
and unbecome_root().  client-side code CANNOT call these functions, it must
only call the *stub* versions of these.

the only circumstances under which it is ok for client-side code to call
become_root() and unbecome_root() is if you are actually calling client-side
code from inside a root process to the msrpc loop-back interface.

this is because you cannot _access_ the msrpc loop-back interface unless you
are root (use a server name of "\\." instead of "\\someservername").