20 years agoRe-add bail on failure.
Andrew Bartlett [Thu, 3 Jan 2002 06:10:25 +0000 (06:10 +0000)]
Re-add bail on failure.

20 years agoFix talloc msg registration
Martin Pool [Thu, 3 Jan 2002 05:57:41 +0000 (05:57 +0000)]
Fix talloc msg registration

20 years agoPut a name on lp_talloc pool
Martin Pool [Thu, 3 Jan 2002 05:57:21 +0000 (05:57 +0000)]
Put a name on lp_talloc pool

20 years agoMake talloctort call talloc_describe_all before exiting
Martin Pool [Thu, 3 Jan 2002 05:57:02 +0000 (05:57 +0000)]
Make talloctort call talloc_describe_all before exiting

20 years agoCope nicely with pools with no name.
Martin Pool [Thu, 3 Jan 2002 05:52:27 +0000 (05:52 +0000)]
Cope nicely with pools with no name.

20 years agofixed ERRMAPEXTRACT torture to work with win2k
Andrew Tridgell [Thu, 3 Jan 2002 05:48:49 +0000 (05:48 +0000)]
fixed ERRMAPEXTRACT torture to work with win2k

20 years agosmbd children should also register for POOL_USAGE
Martin Pool [Thu, 3 Jan 2002 05:41:13 +0000 (05:41 +0000)]
smbd children should also register for POOL_USAGE

20 years agoCleanup format.
Martin Pool [Thu, 3 Jan 2002 05:36:21 +0000 (05:36 +0000)]
Cleanup format.

20 years agotalloc accounting now includes number of chunks and bytes allocated.
Martin Pool [Thu, 3 Jan 2002 05:34:33 +0000 (05:34 +0000)]
talloc accounting now includes number of chunks and bytes allocated.

20 years agoSend back talloc allocation information when POOL_USAGE is received.
Martin Pool [Thu, 3 Jan 2002 05:26:57 +0000 (05:26 +0000)]
Send back talloc allocation information when POOL_USAGE is received.

20 years agoAdd talloc_asprintf_append, which grows an existing string buffer to
Martin Pool [Thu, 3 Jan 2002 05:25:30 +0000 (05:25 +0000)]
Add talloc_asprintf_append, which grows an existing string buffer to
contain new print-formatted information.  (Also
talloc_vasprintf_append.)  Idea borrowed from glib.

20 years agosmbd now sends a reply to MSG_REQ_POOL_USAGE, though it's pretty
Martin Pool [Thu, 3 Jan 2002 04:22:55 +0000 (04:22 +0000)]
smbd now sends a reply to MSG_REQ_POOL_USAGE, though it's pretty
boring so far.

20 years agosmbd should register to receive MSG_REQ_POOL_USAGE. Response should
Martin Pool [Thu, 3 Jan 2002 03:53:41 +0000 (03:53 +0000)]
smbd should register to receive MSG_REQ_POOL_USAGE.  Response should
be information about memory usage, but this is not done yet.

20 years agoFixed crash bug in display_print_info_2()
Tim Potter [Thu, 3 Jan 2002 03:51:13 +0000 (03:51 +0000)]
Fixed crash bug in display_print_info_2()

20 years agoIgnore dmalloc.log.
Martin Pool [Thu, 3 Jan 2002 03:47:09 +0000 (03:47 +0000)]
Ignore dmalloc.log.

20 years agoIf we receive a message that is not handled then emit a gentle warning
Martin Pool [Thu, 3 Jan 2002 03:42:33 +0000 (03:42 +0000)]
If we receive a message that is not handled then emit a gentle warning
to the log.

20 years agodebug output on receipt of message.
Martin Pool [Thu, 3 Jan 2002 03:35:02 +0000 (03:35 +0000)]
debug output on receipt of message.

20 years agoOops, rerun autoconf after last change.
Martin Pool [Thu, 3 Jan 2002 03:14:51 +0000 (03:14 +0000)]
Oops, rerun autoconf after last change.

20 years agoWelcome to preprocessor hell. Had to put a #ifndef around SAFE_FREE to
Tim Potter [Thu, 3 Jan 2002 03:07:51 +0000 (03:07 +0000)]
Welcome to preprocessor hell.  Had to put a #ifndef around SAFE_FREE to
stop smb_macros.h and tdb.h from fighting with each other.

I tried to rearrange the #include file order but that breaks other stuff.

20 years agoAdd instructions on using dmalloc.
Martin Pool [Thu, 3 Jan 2002 03:04:35 +0000 (03:04 +0000)]
Add instructions on using dmalloc.

20 years agoFixed crash bug in display_print_info_1()
Tim Potter [Thu, 3 Jan 2002 02:48:23 +0000 (02:48 +0000)]
Fixed crash bug in display_print_info_1()

20 years agoInitialise result on success in cli_spoolss_enum_printers() so we don't get
Tim Potter [Thu, 3 Jan 2002 02:47:50 +0000 (02:47 +0000)]
Initialise result on success in cli_spoolss_enum_printers() so we don't get
stuck in an infinite loop.

20 years agoThe werror<->nt status code map has changed and has broken all the spoolss
Tim Potter [Thu, 3 Jan 2002 02:36:08 +0000 (02:36 +0000)]
The werror<->nt status code map has changed and has broken all the spoolss
commands in rpcclient.  Replacing ERROR_INSUFFICIENT_BUFFER with

I always thought the caller (i.e cmd_spoolss.c) should take care of the
whole requested/needed buffer size thingy though...

20 years agoRemoved unused #define
Tim Potter [Thu, 3 Jan 2002 02:28:51 +0000 (02:28 +0000)]
Removed unused #define

20 years agodebug statement fixups.
Jeremy Allison [Wed, 2 Jan 2002 23:28:55 +0000 (23:28 +0000)]
debug statement fixups.
Merge SAFE_FREE fix in tdb from 2.2, and IRIX fix.

20 years agoremoved stuff I wasn't using
Gerald Carter [Wed, 2 Jan 2002 23:11:24 +0000 (23:11 +0000)]
removed stuff I wasn't using

20 years agosync up ldap defaults with 2.2
Gerald Carter [Wed, 2 Jan 2002 22:44:15 +0000 (22:44 +0000)]
sync up ldap defaults with 2.2

20 years agoooops./ forgot one
Gerald Carter [Wed, 2 Jan 2002 22:22:22 +0000 (22:22 +0000)]
ooops./  forgot one

20 years agofixing compile error
Gerald Carter [Wed, 2 Jan 2002 22:02:11 +0000 (22:02 +0000)]
fixing compile error

20 years agoWe should be SHLD not LINK when creating shared libraries.
Jeremy Allison [Wed, 2 Jan 2002 21:54:38 +0000 (21:54 +0000)]
We should be SHLD not LINK when creating shared libraries.
Tidyup in passdb.c

20 years agoWe go to a lot of effort to avoid strcpy() in Samba, but its not much use if
Andrew Bartlett [Wed, 2 Jan 2002 09:02:28 +0000 (09:02 +0000)]
We go to a lot of effort to avoid strcpy() in Samba, but its not much use if
people start doing using strlen(src) in strncpy() :-)

This occured when the talloc based passdb was reverted in favor of a pstring
based one.  In the talloc version this was fine, becouse the buffer was
strlen(src) bytes long.  This no longer applies...

This doesn't seem to have broken anything, but I'm chasing up an oddity with NT
password changing from the last merge.

Test & MERGE for 2.2.3

Andrew Bartlett

20 years agoAdd prs_dump_before to dump everything from the start of the prs
Martin Pool [Wed, 2 Jan 2002 07:48:07 +0000 (07:48 +0000)]
Add prs_dump_before to dump everything from the start of the prs
buffer up to the current position, and use this to dump pipe buffers
just before parsing.

20 years agoActually enforce the passdb API.
Andrew Bartlett [Wed, 2 Jan 2002 07:41:54 +0000 (07:41 +0000)]
Actually enforce the passdb API.

Thou shalt not reference SAM_ACCOUNT members directly - always use

This is achived by making the whole of SAM_ACCOUNT have a .private member,
where the real members live.  This caught a pile of examples, and these have
beeen fixed.

The pdb_get..() functions are 'const' (have been for some time) and this
required a few small changes to constify other functions.

I've also added some debugs to the pdb get and set, they can be removed if

I've rewritten the copy_id2x_to_sam_pass() functions to use the new passdb
interface, but I need the flags info to do it properly.

The pdb_free_sam() funciton now blanks out the LM and NT hashes, and as such
I have removed many extra 'samr_clear_sam_passwd(smbpass)' calls as a result.

Finally, any and all testing is always appriciated - but the basics seem to

Andrew Bartlett

20 years agoMake user enumeration work for w2k and win9x.
Jeremy Allison [Wed, 2 Jan 2002 07:27:33 +0000 (07:27 +0000)]
Make user enumeration work for w2k and win9x.
This means different return sizes depending on client bugs. This sucks :-(.

20 years agoAnother touch of 'const'
Andrew Bartlett [Wed, 2 Jan 2002 06:55:21 +0000 (06:55 +0000)]
Another touch of 'const'

20 years agoDoc.
Martin Pool [Wed, 2 Jan 2002 06:29:56 +0000 (06:29 +0000)]

20 years agoAlso capture received data
Martin Pool [Wed, 2 Jan 2002 06:08:02 +0000 (06:08 +0000)]
Also capture received data

20 years agostruct cli_state remembers the pipe name that it's talking to, if any,
Martin Pool [Wed, 2 Jan 2002 05:39:49 +0000 (05:39 +0000)]
struct cli_state remembers the pipe name that it's talking to, if any,
so that we can print it in later debug messages.

Call prs_dump to dump out requests sent by the client at sufficiently
high debug levels.

20 years agostruct cli_state remembers the pipe name that it's talking to, if any,
Martin Pool [Wed, 2 Jan 2002 05:38:09 +0000 (05:38 +0000)]
struct cli_state remembers the pipe name that it's talking to, if any,
so that we can print it in later debug messages.

20 years agoMinor update to make the output 'real C'. (The output is intended to be a C
Andrew Bartlett [Wed, 2 Jan 2002 03:54:40 +0000 (03:54 +0000)]
Minor update to make the output 'real C'.  (The output is intended to be a C
format table for inclusion back into Samba).

Andrew Bartlett

20 years agoRemoved reference to rpc_client/cli_netlogon.o from dependencies for
Tim Potter [Wed, 2 Jan 2002 00:52:30 +0000 (00:52 +0000)]
Removed reference to rpc_client/cli_netlogon.o from dependencies for
samsync.  The libsmb version should do fine.

20 years agothe pdbedit tool manpage
Simo Sorce [Wed, 2 Jan 2002 00:40:47 +0000 (00:40 +0000)]
the pdbedit tool manpage

20 years agotry to handle broken const in headers on cray unicos
Andrew Tridgell [Tue, 1 Jan 2002 06:55:33 +0000 (06:55 +0000)]
try to handle broken const in headers on cray unicos

20 years agoThis brings the NT->DOS error mapping into better line with what NT does.
Andrew Bartlett [Tue, 1 Jan 2002 06:16:43 +0000 (06:16 +0000)]
This brings the NT->DOS error mapping into better line with what NT does.

I'll post the changes to the actual map to the list for comment, but this fixes
the 'unknown' case.

Andrew Bartlett

20 years agoAdd a function to convert 'NT_STATUS...' strings back into their actual error
Andrew Bartlett [Tue, 1 Jan 2002 05:52:02 +0000 (05:52 +0000)]
Add a function to convert 'NT_STATUS...' strings back into their actual error

Andrew Bartlett

20 years agoAllow usernames in the form of 'NT_STATUS_....' to map to that as the error
Andrew Bartlett [Tue, 1 Jan 2002 05:51:03 +0000 (05:51 +0000)]
Allow usernames in the form of 'NT_STATUS_....' to map to that as the error
when using the 'name_to_ntstatus' auth module.

This could be useful in testing.

Andrew Bartlett

20 years agoNow that winbind doesn't rely on this, we may as well remove it...
Andrew Bartlett [Tue, 1 Jan 2002 05:49:27 +0000 (05:49 +0000)]
Now that winbind doesn't rely on this, we may as well remove it...

Andrew Bartlett

20 years agoFurther rpc_client removal, this time from winbindd.
Andrew Bartlett [Tue, 1 Jan 2002 04:50:45 +0000 (04:50 +0000)]
Further rpc_client removal, this time from winbindd.

Also removed the dependency on auth_util.o, which makes things nicer.

makefile variable - becouse Samba dependencies are starting to be sane again!

Andrew Bartlett

20 years agorenamed ans1.h to asn_1.h to prevent conflict caused by krb5 headers
Andrew Tridgell [Tue, 1 Jan 2002 04:05:10 +0000 (04:05 +0000)]
renamed ans1.h to asn_1.h to prevent conflict caused by krb5 headers
on some platforms using "" instead of <> in include statements

20 years agoA farily large commit:
Andrew Bartlett [Tue, 1 Jan 2002 03:10:32 +0000 (03:10 +0000)]
A farily large commit:

 - Move rpc_client/cli_trust.c to smbd/change_trust_pw.c
  - It hasn't been used by anything else since smbpasswd lost its -j

 - Add a TALLOC_CTX to the auth subsytem.  These are only valid for the length
   of the calls to the individual modules, if you want a longer context hide it
   in your private data.

   Similarly, all returns (like the server_info) should still be malloced.

 - Move the 'ntdomain' module (security=domain in oldspeak) over to use the new
   libsmb domain logon code.  Also rework much of the code to use some better
   helper functions for the connection - getting us much better error returns
   (the new code is NTSTATUS).

   The only remaining thing to do is to figure out if tpot's 0xdead 0xbeef for
   the LUID feilds is sufficient, or if we should do random LUIDs as per the old

   Similarly, I'll move winbind over to this when I get a chance.

This leaves the SPOOLSS code and some cli_pipe code as the only stuff still in
rpc_client, at least as far as smbd is concerned.

While I've given this a basic rundown, any testing is as always appriciated.

Andrew Bartlett

20 years agoAdd a specialised version of tpot's libsmb samlogon code for use with
Andrew Bartlett [Tue, 1 Jan 2002 02:34:29 +0000 (02:34 +0000)]
Add a specialised version of tpot's libsmb samlogon code for use with
presupplied challange-response pairs, and only using the 'network' version.

This will be used to move the auth subsystem over to a libsmb (rather than
rpc_client) base.

Andrew Bartlett

20 years agoFix up C99 comment.
Andrew Bartlett [Tue, 1 Jan 2002 02:31:32 +0000 (02:31 +0000)]
Fix up C99 comment.
// -> /* */

20 years agoNamed constants are always much better than magic numbers...
Andrew Bartlett [Tue, 1 Jan 2002 00:18:47 +0000 (00:18 +0000)]
Named constants are always much better than magic numbers...
(I defined the constants when I put in 'net rpc shutdown')

Andrew Bartlett

20 years agolast_enum not used.
Jeremy Allison [Mon, 31 Dec 2001 23:24:48 +0000 (23:24 +0000)]
last_enum not used.

20 years agoFixed enumeration of a large (<1500) users from a Samba PDC to a W2k
Jeremy Allison [Mon, 31 Dec 2001 22:37:51 +0000 (22:37 +0000)]
Fixed enumeration of a large (<1500) users from a Samba PDC to a W2k
member server. Firstly, use the same max enumeration size (0x400) as W2K
uses, otherwise W2K won't ask for any more. Secondly, if a enumeration
request with a non-zero offset comes in on a handle that hasn't started
an enumeration, don't bitch about it (return NT_STATUS_UNSUCCESSFUL),
just load the db on that handle and return at that offset.

20 years agoadded a simple test to see whether building shared libraries actually
Andrew Tridgell [Mon, 31 Dec 2001 22:10:24 +0000 (22:10 +0000)]
added a simple test to see whether building shared libraries actually

20 years agoops, froget to set the values.
Simo Sorce [Mon, 31 Dec 2001 16:01:04 +0000 (16:01 +0000)]
ops, froget to set the values.
jerry, can you look at theis where we use standard_sub_advanced() fns?
I think this structure should be backported to 2.2 because we do not know if the SAM_ACCOUNT strings have enough space to contain the "substituted" string.
(Yes, just now we know they are pstrings, but we may change them into alloced one, I'm a strong suported of alloced strings as 1024 bytes are not always enough and are often too much)

20 years agoconst religion
Simo Sorce [Mon, 31 Dec 2001 15:48:32 +0000 (15:48 +0000)]
const religion

20 years agoport mods from 2.2
Simo Sorce [Mon, 31 Dec 2001 15:48:03 +0000 (15:48 +0000)]
port mods from 2.2

20 years agoreverted to 1.24 and manually merged in changes from 2.2
Gerald Carter [Mon, 31 Dec 2001 14:39:26 +0000 (14:39 +0000)]
reverted to 1.24 and manually merged in changes from 2.2

20 years agocope with systems that don't have full gssapi libs
Andrew Tridgell [Mon, 31 Dec 2001 14:31:13 +0000 (14:31 +0000)]
cope with systems that don't have full gssapi libs

20 years agomore irix -64 portability fixes
Andrew Tridgell [Mon, 31 Dec 2001 14:05:22 +0000 (14:05 +0000)]
more irix -64 portability fixes

20 years ago- portablitity fixes for cc -64 on irix
Andrew Tridgell [Mon, 31 Dec 2001 13:46:26 +0000 (13:46 +0000)]
- portablitity fixes for cc -64 on irix
- fixed gid* bug in rpc_server

20 years agoEnsure the output cli can't have spurious values if the connection fails...
Andrew Bartlett [Mon, 31 Dec 2001 13:06:10 +0000 (13:06 +0000)]
Ensure the output cli can't have spurious values if the connection fails...

20 years agoAdd 'net rpc shutdown' and 'net rpc abortshutdown'.
Andrew Bartlett [Mon, 31 Dec 2001 13:00:59 +0000 (13:00 +0000)]
Add 'net rpc shutdown' and 'net rpc abortshutdown'.

These two little features are very useful, but the passing of options about
needs some serious work.  The popt stuff in the shutdown code is #ifdef'ed out
until the main popt loop can be convinced not to chew on the options :-(

Andrew Bartlett

20 years agoFix up the comment in the copyright header
Andrew Bartlett [Mon, 31 Dec 2001 12:50:44 +0000 (12:50 +0000)]
Fix up the comment in the copyright header

20 years agoGive the main loop talloc context a name (using mbp's new talloc naming
Andrew Bartlett [Mon, 31 Dec 2001 12:41:24 +0000 (12:41 +0000)]
Give the main loop talloc context a name (using mbp's new talloc naming

Andrew Bartlett

20 years agoMake --with-tdbsam compile again, given the new 'am I setting a default' flags.
Andrew Bartlett [Mon, 31 Dec 2001 11:25:48 +0000 (11:25 +0000)]
Make --with-tdbsam compile again, given the new 'am I setting a default' flags.

Andrew Bartlett

20 years agoAllow this to build without LDAP, as per the example below it.
Andrew Bartlett [Mon, 31 Dec 2001 11:14:38 +0000 (11:14 +0000)]
Allow this to build without LDAP, as per the example below it.

Andrew Bartlett

20 years agofix compile error
Gerald Carter [Mon, 31 Dec 2001 02:04:08 +0000 (02:04 +0000)]
fix compile error

20 years agoremoved extern samlogon_user
Gerald Carter [Mon, 31 Dec 2001 01:56:05 +0000 (01:56 +0000)]
removed extern samlogon_user

20 years agoremove samlogon_user
Gerald Carter [Mon, 31 Dec 2001 00:09:15 +0000 (00:09 +0000)]
remove samlogon_user

20 years agosome merges from 2.2. Still need to merge in changes from pdb_tdb.c
Gerald Carter [Mon, 31 Dec 2001 00:06:51 +0000 (00:06 +0000)]
some merges from 2.2.  Still need to merge in changes from pdb_tdb.c
but it will take more time as I don't want to loose any fixes that
are only in HEAD.

20 years agofreeing the wrong pointer, sorry my mistake.
Simo Sorce [Sun, 30 Dec 2001 22:55:04 +0000 (22:55 +0000)]
freeing the wrong pointer, sorry my mistake.

20 years agoUnicode byte ording fix from Benjamin Kuit <>
Tim Potter [Sun, 30 Dec 2001 21:25:46 +0000 (21:25 +0000)]
Unicode byte ording fix from Benjamin Kuit <>

20 years agoutil_sid.c - respect a const variabile (addedd strdup)
Simo Sorce [Sun, 30 Dec 2001 19:21:25 +0000 (19:21 +0000)]
util_sid.c - respect a const variabile (addedd strdup)
cli_reg.c  - indentation
pdb_ldap.c - some checks on init fns parameters
pdb_tdb.c  - some checks on init fns parameters + make sure we close the db on failure

20 years agoAdd a pile of doxygen style comments to various parts of Samba. Many of these
Andrew Bartlett [Sun, 30 Dec 2001 10:54:58 +0000 (10:54 +0000)]
Add a pile of doxygen style comments to various parts of Samba.  Many of these
probably will never actually be genearted, but I like the style in any case.

Also fix a segfault in 'net rpc' when the login failed and a small memory leak
on failure in the auth_info.c code.

Andrew Bartlett

20 years agoFinally remove these files, which moved (by cvs backend magic) to source/auth
Andrew Bartlett [Sun, 30 Dec 2001 10:50:46 +0000 (10:50 +0000)]
Finally remove these files, which moved (by cvs backend magic) to source/auth
at the start of the month.

20 years agoFinish idra's cleanup of the RPC remote shutdown code.
Andrew Bartlett [Sun, 30 Dec 2001 10:19:56 +0000 (10:19 +0000)]
Finish idra's cleanup of the RPC remote shutdown code.

20 years agoremove unneaded memory leaking code.
Simo Sorce [Sun, 30 Dec 2001 08:37:34 +0000 (08:37 +0000)]
remove unneaded memory leaking code.

20 years agoFix the compile on systems without a full kerberos kit.
Andrew Bartlett [Sun, 30 Dec 2001 06:20:23 +0000 (06:20 +0000)]
Fix the compile on systems without a full kerberos kit.

Andrew Bartlett

20 years agoMake Samba compile on RH 6.2 again.
Andrew Bartlett [Sun, 30 Dec 2001 05:59:43 +0000 (05:59 +0000)]
Make Samba compile on RH 6.2 again.

We now include the libber.h file if required, but currently we just don't use
ldap.  (I'll chase this up).

In the meantime, I've moved the ads_status code about, its now in its own file,
and has a couple of #ifdefs to allow smbd to link - becouse the lack of LDAP
caused HAVE_ADS to be undefined. (I hope its not too ugly).

Andrew Bartlett

20 years agoWhen running interactive we want to set our own process group for
Jeremy Allison [Sun, 30 Dec 2001 01:46:38 +0000 (01:46 +0000)]
When running interactive we want to set our own process group for
signal management.

20 years agopdb_getsampwnuid() merge from 2.2
Gerald Carter [Sun, 30 Dec 2001 00:03:47 +0000 (00:03 +0000)]
pdb_getsampwnuid() merge from 2.2

20 years agoDocumented -i modes for nmbd/smbd.
Jeremy Allison [Sat, 29 Dec 2001 21:15:40 +0000 (21:15 +0000)]
Documented -i modes for nmbd/smbd.

20 years agoFixup -i interactive modes.
Jeremy Allison [Sat, 29 Dec 2001 21:13:53 +0000 (21:13 +0000)]
Fixup -i interactive modes.

20 years agoRemoved extra lp_adduser() call.
Jeremy Allison [Sat, 29 Dec 2001 20:29:43 +0000 (20:29 +0000)]
Removed extra lp_adduser() call.
Fixed up error returns in get_correct_cversion().

20 years agofile Samba-LDAP-HOWTO.sgml was initially added on branch SAMBA_2_2.
Gerald Carter [Sat, 29 Dec 2001 16:20:49 +0000 (16:20 +0000)]
file Samba-LDAP-HOWTO.sgml was initially added on branch SAMBA_2_2.

20 years agopassing NULL to print_queue_status()'s status parameter could end in a segfault
Simo Sorce [Sat, 29 Dec 2001 09:52:24 +0000 (09:52 +0000)]
passing NULL to print_queue_status()'s status parameter could end in a segfault

20 years agoremoved unused files
Gerald Carter [Fri, 28 Dec 2001 02:28:39 +0000 (02:28 +0000)]
removed unused files

20 years agomerge from 2.2
Gerald Carter [Fri, 28 Dec 2001 02:03:36 +0000 (02:03 +0000)]
merge from 2.2

20 years agomoving SAM_ACCOUNT to include a bit field for initialized
Gerald Carter [Thu, 27 Dec 2001 06:38:04 +0000 (06:38 +0000)]
moving SAM_ACCOUNT to include a bit field for initialized
members (such as uid and gid).  This way we will be able to
keep ourselves from writing out default smb.conf settings when
the admin doesn't want to,  That part is not done yet.

Tested compiles with ldap/tdb/smbpasswd.  Tested connection with smbpasswd

oh...and smbpasswd doesn'y automatically expire accounts after 21 days
from the last password change either now.  Just ifdef'd out that code
in build_sam_account().

Will merge updates into 2.2 as they are necessary.


20 years agosync with 2.2
Gerald Carter [Wed, 26 Dec 2001 05:35:40 +0000 (05:35 +0000)]
sync with 2.2

20 years agomerge Jeremy's gcc3 warning fix from 2.2
Herb Lewis [Sat, 22 Dec 2001 01:15:58 +0000 (01:15 +0000)]
merge Jeremy's gcc3 warning fix from 2.2

20 years agosync up packaging scripts
Herb Lewis [Sat, 22 Dec 2001 00:54:10 +0000 (00:54 +0000)]
sync up packaging scripts

20 years agomerge IRIX winbind support from Samba 2.2 branch
Herb Lewis [Sat, 22 Dec 2001 00:51:32 +0000 (00:51 +0000)]
merge IRIX winbind support from Samba 2.2 branch

20 years agoRemoved unused variables.
Jeremy Allison [Fri, 21 Dec 2001 23:28:04 +0000 (23:28 +0000)]
Removed unused variables.

20 years agoThe jf_get_sampasswd_entries() is not used anymore.
Jeremy Allison [Fri, 21 Dec 2001 23:25:32 +0000 (23:25 +0000)]
The jf_get_sampasswd_entries() is not used anymore.

20 years agoadded a copyright notice.
Jean-François Micouleau [Fri, 21 Dec 2001 22:54:53 +0000 (22:54 +0000)]
added a copyright notice.


20 years agodoesn't keep track of the struct's size we enumerate. W2K doesn't
Jean-François Micouleau [Fri, 21 Dec 2001 22:34:49 +0000 (22:34 +0000)]
doesn't keep track of the struct's size we enumerate. W2K doesn't
calculate them and always reply a size of 32 bytes whereas NT4 did the
maths. Anyway, it looks like the clients don't complain.

in query_dom_info() at level 2, return the real number of users and
groups. That's the fix to the W95/98 userlist bug !

as W95/98 does a query_dom_info(2) followed by a query_disp_info(4) on
the SAME context handle (err we call it an lsa policy handle ! plain
wrong name), I was tempted to keep the snapshoot in memory, to prevent
2 full user db enumerations in a row and just have one shared. But if some
client does the 2 calls on two different handles, we would have 2 copies
in memory not free'ed before the samr_close().

We still have too many fixed constant and too many magic values in that
code. And btw, I really hates how the sequence number is generated !