2003-07-01 21:51:52 +04:00
/*
2002-01-30 09:08:46 +03:00
Unix SMB / CIFS implementation .
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
Authentication utility functions
Copyright ( C ) Andrew Tridgell 1992 - 1998
Copyright ( C ) Andrew Bartlett 2001
2001-11-04 02:34:24 +03:00
Copyright ( C ) Jeremy Allison 2000 - 2001
2002-09-25 19:19:00 +04:00
Copyright ( C ) Rafal Szczesniak 2002
2006-02-04 01:19:41 +03:00
Copyright ( C ) Volker Lendecke 2006
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
This program is free software ; you can redistribute it and / or modify
it under the terms of the GNU General Public License as published by
2007-07-09 23:25:36 +04:00
the Free Software Foundation ; either version 3 of the License , or
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
( at your option ) any later version .
This program is distributed in the hope that it will be useful ,
but WITHOUT ANY WARRANTY ; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE . See the
GNU General Public License for more details .
You should have received a copy of the GNU General Public License
2007-07-10 04:52:41 +04:00
along with this program . If not , see < http : //www.gnu.org/licenses/>.
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
*/
2001-10-02 08:29:50 +04:00
# include "includes.h"
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
2002-07-15 14:35:28 +04:00
# undef DBGC_CLASS
# define DBGC_CLASS DBGC_AUTH
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
/****************************************************************************
Create a UNIX user on demand .
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2003-05-26 03:56:41 +04:00
static int smb_create_user ( const char * domain , const char * unix_username , const char * homedir )
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
{
2007-11-14 21:37:18 +03:00
TALLOC_CTX * ctx = talloc_tos ( ) ;
char * add_script ;
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
int ret ;
2007-11-14 21:37:18 +03:00
add_script = talloc_strdup ( ctx , lp_adduser_script ( ) ) ;
if ( ! add_script | | ! * add_script ) {
2001-11-04 02:34:24 +03:00
return - 1 ;
2007-11-14 21:37:18 +03:00
}
add_script = talloc_all_string_sub ( ctx ,
add_script ,
" %u " ,
unix_username ) ;
if ( ! add_script ) {
return - 1 ;
}
if ( domain ) {
add_script = talloc_all_string_sub ( ctx ,
add_script ,
" %D " ,
domain ) ;
if ( ! add_script ) {
return - 1 ;
}
}
if ( homedir ) {
add_script = talloc_all_string_sub ( ctx ,
add_script ,
" %H " ,
homedir ) ;
if ( ! add_script ) {
return - 1 ;
}
}
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
ret = smbrun ( add_script , NULL ) ;
2005-03-03 19:52:44 +03:00
flush_pwnam_cache ( ) ;
2007-11-14 21:37:18 +03:00
DEBUG ( ret ? 0 : 3 ,
( " smb_create_user: Running the command `%s' gave %d \n " ,
add_script , ret ) ) ;
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
return ret ;
}
2001-10-31 13:46:25 +03:00
/****************************************************************************
Create an auth_usersupplied_data structure
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-11-14 21:37:18 +03:00
static NTSTATUS make_user_info ( auth_usersupplied_info * * user_info ,
const char * smb_name ,
2002-09-25 19:19:00 +04:00
const char * internal_username ,
2007-11-14 21:37:18 +03:00
const char * client_domain ,
2002-09-25 19:19:00 +04:00
const char * domain ,
2007-11-14 21:37:18 +03:00
const char * wksta_name ,
2004-04-03 19:41:32 +04:00
DATA_BLOB * lm_pwd , DATA_BLOB * nt_pwd ,
DATA_BLOB * lm_interactive_pwd , DATA_BLOB * nt_interactive_pwd ,
2007-11-14 21:37:18 +03:00
DATA_BLOB * plaintext ,
2007-10-19 04:40:25 +04:00
bool encrypted )
2001-10-31 13:46:25 +03:00
{
DEBUG ( 5 , ( " attempting to make a user_info for %s (%s) \n " , internal_username , smb_name ) ) ;
2004-12-07 21:25:53 +03:00
* user_info = SMB_MALLOC_P ( auth_usersupplied_info ) ;
2006-03-07 23:14:47 +03:00
if ( * user_info = = NULL ) {
2003-07-25 08:24:40 +04:00
DEBUG ( 0 , ( " malloc failed for user_info (size %lu) \n " , ( unsigned long ) sizeof ( * user_info ) ) ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
ZERO_STRUCTP ( * user_info ) ;
2002-03-01 04:24:30 +03:00
DEBUG ( 5 , ( " making strings for %s's user_info struct \n " , internal_username ) ) ;
2001-10-31 13:46:25 +03:00
2006-02-04 01:19:41 +03:00
( * user_info ) - > smb_name = SMB_STRDUP ( smb_name ) ;
if ( ( * user_info ) - > smb_name = = NULL ) {
2001-10-31 13:46:25 +03:00
free_user_info ( user_info ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
( * user_info ) - > internal_username = SMB_STRDUP ( internal_username ) ;
if ( ( * user_info ) - > internal_username = = NULL ) {
2001-10-31 13:46:25 +03:00
free_user_info ( user_info ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
( * user_info ) - > domain = SMB_STRDUP ( domain ) ;
if ( ( * user_info ) - > domain = = NULL ) {
2001-10-31 13:46:25 +03:00
free_user_info ( user_info ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
( * user_info ) - > client_domain = SMB_STRDUP ( client_domain ) ;
if ( ( * user_info ) - > client_domain = = NULL ) {
2001-10-31 13:46:25 +03:00
free_user_info ( user_info ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
( * user_info ) - > wksta_name = SMB_STRDUP ( wksta_name ) ;
if ( ( * user_info ) - > wksta_name = = NULL ) {
2001-10-31 13:46:25 +03:00
free_user_info ( user_info ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_MEMORY ;
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
}
2001-10-31 13:46:25 +03:00
2002-07-15 14:35:28 +04:00
DEBUG ( 5 , ( " making blobs for %s's user_info struct \n " , internal_username ) ) ;
2001-10-31 13:46:25 +03:00
2004-04-03 19:41:32 +04:00
if ( lm_pwd )
( * user_info ) - > lm_resp = data_blob ( lm_pwd - > data , lm_pwd - > length ) ;
if ( nt_pwd )
( * user_info ) - > nt_resp = data_blob ( nt_pwd - > data , nt_pwd - > length ) ;
if ( lm_interactive_pwd )
( * user_info ) - > lm_interactive_pwd = data_blob ( lm_interactive_pwd - > data , lm_interactive_pwd - > length ) ;
if ( nt_interactive_pwd )
( * user_info ) - > nt_interactive_pwd = data_blob ( nt_interactive_pwd - > data , nt_interactive_pwd - > length ) ;
if ( plaintext )
( * user_info ) - > plaintext_password = data_blob ( plaintext - > data , plaintext - > length ) ;
2001-10-31 13:46:25 +03:00
( * user_info ) - > encrypted = encrypted ;
2005-11-08 09:19:34 +03:00
( * user_info ) - > logon_parameters = 0 ;
2001-10-31 13:46:25 +03:00
DEBUG ( 10 , ( " made an %sencrypted user_info for %s (%s) \n " , encrypted ? " " : " un " , internal_username , smb_name ) ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_OK ;
This is my 'Authentication Rewrite' version 1.01, mostly as submitted to
samba-technical a few weeks ago.
The idea here is to standardize the checking of user names and passwords,
thereby ensuring that all authtentications pass the same standards. The
interface currently implemented in as
nt_status = check_password(user_info, server_info)
where user_info contains (mostly) the authentication data, and server_info
contains things like the user-id they got, and their resolved user name.
The current ugliness with the way the structures are created will be killed
the next revision, when they will be created and malloced by creator functions.
This patch also includes the first implementation of NTLMv2 in HEAD, but which
needs some more testing. We also add a hack to allow plaintext passwords to be
compared with smbpasswd, not the system password database.
Finally, this patch probably reintroduces the PAM accounts bug we had in
2.2.0, I'll fix that once this hits the tree. (I've just finished testing
it on a wide variety of platforms, so I want to get this patch in).
(This used to be commit b30b6202f31d339b48d51c0d38174cafd1cfcd42)
2001-08-03 17:09:23 +04:00
}
2001-10-31 13:46:25 +03:00
/****************************************************************************
Create an auth_usersupplied_data structure after appropriate mapping .
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2002-09-25 19:19:00 +04:00
NTSTATUS make_user_info_map ( auth_usersupplied_info * * user_info ,
const char * smb_name ,
const char * client_domain ,
const char * wksta_name ,
2004-04-03 19:41:32 +04:00
DATA_BLOB * lm_pwd , DATA_BLOB * nt_pwd ,
DATA_BLOB * lm_interactive_pwd , DATA_BLOB * nt_interactive_pwd ,
DATA_BLOB * plaintext ,
2007-10-19 04:40:25 +04:00
bool encrypted )
2001-10-31 13:46:25 +03:00
{
2001-11-11 14:34:46 +03:00
const char * domain ;
2006-05-06 23:24:35 +04:00
NTSTATUS result ;
2007-10-19 04:40:25 +04:00
bool was_mapped ;
2001-10-31 13:46:25 +03:00
fstring internal_username ;
fstrcpy ( internal_username , smb_name ) ;
2006-05-06 23:24:35 +04:00
was_mapped = map_username ( internal_username ) ;
2002-09-25 19:19:00 +04:00
2002-03-23 12:01:30 +03:00
DEBUG ( 5 , ( " make_user_info_map: Mapping user [%s] \\ [%s] from workstation [%s] \n " ,
client_domain , smb_name , wksta_name ) ) ;
2001-10-31 13:46:25 +03:00
2003-06-28 12:29:42 +04:00
/* don't allow "" as a domain, fixes a Win9X bug
2003-08-08 09:11:11 +04:00
where it doens ' t supply a domain for logon script
' net use ' commands . */
2002-01-18 11:12:10 +03:00
2003-06-28 12:29:42 +04:00
if ( * client_domain )
2002-01-18 11:12:10 +03:00
domain = client_domain ;
2003-06-28 12:29:42 +04:00
else
domain = lp_workgroup ( ) ;
2002-01-18 11:12:10 +03:00
2003-06-28 12:29:42 +04:00
/* do what win2k does. Always map unknown domains to our own
2003-07-01 08:11:42 +04:00
and let the " passdb backend " handle unknown users . */
2003-04-24 15:56:09 +04:00
2003-08-08 09:11:11 +04:00
if ( ! is_trusted_domain ( domain ) & & ! strequal ( domain , get_global_sam_name ( ) ) )
2006-02-21 17:34:11 +03:00
domain = my_sam_name ( ) ;
2003-06-28 12:29:42 +04:00
/* we know that it is a trusted domain (and we are allowing them) or it is our domain */
2001-10-31 13:46:25 +03:00
2006-05-06 23:24:35 +04:00
result = make_user_info ( user_info , smb_name , internal_username ,
2004-04-03 19:41:32 +04:00
client_domain , domain , wksta_name ,
lm_pwd , nt_pwd ,
lm_interactive_pwd , nt_interactive_pwd ,
plaintext , encrypted ) ;
2006-05-06 23:24:35 +04:00
if ( NT_STATUS_IS_OK ( result ) ) {
( * user_info ) - > was_mapped = was_mapped ;
}
return result ;
2001-10-31 13:46:25 +03:00
}
/****************************************************************************
Create an auth_usersupplied_data , making the DATA_BLOBs here .
2001-11-22 00:10:13 +03:00
Decrypt and encrypt the passwords .
2001-10-31 13:46:25 +03:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-10-19 04:40:25 +04:00
bool make_user_info_netlogon_network ( auth_usersupplied_info * * user_info ,
2002-01-20 11:58:21 +03:00
const char * smb_name ,
const char * client_domain ,
const char * wksta_name ,
2005-11-08 09:19:34 +03:00
uint32 logon_parameters ,
2006-02-04 01:19:41 +03:00
const uchar * lm_network_pwd ,
int lm_pwd_len ,
const uchar * nt_network_pwd ,
int nt_pwd_len )
2001-10-31 13:46:25 +03:00
{
2007-10-19 04:40:25 +04:00
bool ret ;
2006-02-04 01:19:41 +03:00
NTSTATUS status ;
2001-10-31 13:46:25 +03:00
DATA_BLOB lm_blob = data_blob ( lm_network_pwd , lm_pwd_len ) ;
DATA_BLOB nt_blob = data_blob ( nt_network_pwd , nt_pwd_len ) ;
2001-11-01 08:02:41 +03:00
2006-02-04 01:19:41 +03:00
status = make_user_info_map ( user_info ,
smb_name , client_domain ,
wksta_name ,
lm_pwd_len ? & lm_blob : NULL ,
nt_pwd_len ? & nt_blob : NULL ,
NULL , NULL , NULL ,
True ) ;
2005-11-08 09:19:34 +03:00
2006-02-04 01:19:41 +03:00
if ( NT_STATUS_IS_OK ( status ) ) {
2005-11-08 09:19:34 +03:00
( * user_info ) - > logon_parameters = logon_parameters ;
}
2006-02-04 01:19:41 +03:00
ret = NT_STATUS_IS_OK ( status ) ? True : False ;
2005-11-08 09:19:34 +03:00
2001-10-31 13:46:25 +03:00
data_blob_free ( & lm_blob ) ;
data_blob_free ( & nt_blob ) ;
return ret ;
}
/****************************************************************************
Create an auth_usersupplied_data , making the DATA_BLOBs here .
2001-11-22 00:10:13 +03:00
Decrypt and encrypt the passwords .
2001-10-31 13:46:25 +03:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-10-19 04:40:25 +04:00
bool make_user_info_netlogon_interactive ( auth_usersupplied_info * * user_info ,
2002-01-05 07:55:41 +03:00
const char * smb_name ,
const char * client_domain ,
const char * wksta_name ,
2005-11-08 09:19:34 +03:00
uint32 logon_parameters ,
2002-01-05 07:55:41 +03:00
const uchar chal [ 8 ] ,
const uchar lm_interactive_pwd [ 16 ] ,
const uchar nt_interactive_pwd [ 16 ] ,
const uchar * dc_sess_key )
2001-10-31 13:46:25 +03:00
{
2007-03-28 17:26:43 +04:00
unsigned char lm_pwd [ 16 ] ;
unsigned char nt_pwd [ 16 ] ;
2001-10-31 13:46:25 +03:00
unsigned char local_lm_response [ 24 ] ;
unsigned char local_nt_response [ 24 ] ;
unsigned char key [ 16 ] ;
2001-11-01 08:02:41 +03:00
ZERO_STRUCT ( key ) ;
2001-10-31 13:46:25 +03:00
memcpy ( key , dc_sess_key , 8 ) ;
2006-02-04 01:19:41 +03:00
if ( lm_interactive_pwd )
memcpy ( lm_pwd , lm_interactive_pwd , sizeof ( lm_pwd ) ) ;
if ( nt_interactive_pwd )
memcpy ( nt_pwd , nt_interactive_pwd , sizeof ( nt_pwd ) ) ;
2001-10-31 13:46:25 +03:00
# ifdef DEBUG_PASSWORD
DEBUG ( 100 , ( " key: " ) ) ;
2007-03-28 17:34:59 +04:00
dump_data ( 100 , key , sizeof ( key ) ) ;
2001-10-31 13:46:25 +03:00
DEBUG ( 100 , ( " lm owf password: " ) ) ;
2001-11-01 08:02:41 +03:00
dump_data ( 100 , lm_pwd , sizeof ( lm_pwd ) ) ;
2001-10-31 13:46:25 +03:00
DEBUG ( 100 , ( " nt owf password: " ) ) ;
2001-11-01 08:02:41 +03:00
dump_data ( 100 , nt_pwd , sizeof ( nt_pwd ) ) ;
2001-10-31 13:46:25 +03:00
# endif
2004-04-03 19:41:32 +04:00
if ( lm_interactive_pwd )
2007-03-28 17:26:43 +04:00
SamOEMhash ( lm_pwd , key , sizeof ( lm_pwd ) ) ;
2004-04-03 19:41:32 +04:00
if ( nt_interactive_pwd )
2007-03-28 17:26:43 +04:00
SamOEMhash ( nt_pwd , key , sizeof ( nt_pwd ) ) ;
2001-10-31 13:46:25 +03:00
# ifdef DEBUG_PASSWORD
DEBUG ( 100 , ( " decrypt of lm owf password: " ) ) ;
2001-11-01 08:02:41 +03:00
dump_data ( 100 , lm_pwd , sizeof ( lm_pwd ) ) ;
2001-10-31 13:46:25 +03:00
DEBUG ( 100 , ( " decrypt of nt owf password: " ) ) ;
2001-11-01 08:02:41 +03:00
dump_data ( 100 , nt_pwd , sizeof ( nt_pwd ) ) ;
2001-10-31 13:46:25 +03:00
# endif
2004-04-03 19:41:32 +04:00
if ( lm_interactive_pwd )
2007-03-28 17:26:43 +04:00
SMBOWFencrypt ( lm_pwd , chal ,
2006-02-04 01:19:41 +03:00
local_lm_response ) ;
2004-04-03 19:41:32 +04:00
if ( nt_interactive_pwd )
2007-03-28 17:26:43 +04:00
SMBOWFencrypt ( nt_pwd , chal ,
2006-02-04 01:19:41 +03:00
local_nt_response ) ;
2001-10-31 13:46:25 +03:00
2002-03-25 02:25:05 +03:00
/* Password info paranoia */
2001-10-31 13:46:25 +03:00
ZERO_STRUCT ( key ) ;
{
2007-10-19 04:40:25 +04:00
bool ret ;
2002-09-25 19:19:00 +04:00
NTSTATUS nt_status ;
2004-04-03 19:41:32 +04:00
DATA_BLOB local_lm_blob ;
DATA_BLOB local_nt_blob ;
2001-10-31 13:46:25 +03:00
2004-04-03 19:41:32 +04:00
DATA_BLOB lm_interactive_blob ;
DATA_BLOB nt_interactive_blob ;
if ( lm_interactive_pwd ) {
2006-02-04 01:19:41 +03:00
local_lm_blob = data_blob ( local_lm_response ,
sizeof ( local_lm_response ) ) ;
lm_interactive_blob = data_blob ( lm_pwd ,
sizeof ( lm_pwd ) ) ;
2004-04-03 19:41:32 +04:00
ZERO_STRUCT ( lm_pwd ) ;
}
if ( nt_interactive_pwd ) {
2006-02-04 01:19:41 +03:00
local_nt_blob = data_blob ( local_nt_response ,
sizeof ( local_nt_response ) ) ;
nt_interactive_blob = data_blob ( nt_pwd ,
sizeof ( nt_pwd ) ) ;
2004-04-03 19:41:32 +04:00
ZERO_STRUCT ( nt_pwd ) ;
}
2001-11-01 08:02:41 +03:00
2006-02-04 01:19:41 +03:00
nt_status = make_user_info_map (
user_info ,
smb_name , client_domain , wksta_name ,
lm_interactive_pwd ? & local_lm_blob : NULL ,
nt_interactive_pwd ? & local_nt_blob : NULL ,
lm_interactive_pwd ? & lm_interactive_blob : NULL ,
nt_interactive_pwd ? & nt_interactive_blob : NULL ,
NULL , True ) ;
2004-04-03 19:41:32 +04:00
2005-11-08 09:19:34 +03:00
if ( NT_STATUS_IS_OK ( nt_status ) ) {
( * user_info ) - > logon_parameters = logon_parameters ;
}
2002-09-25 19:19:00 +04:00
ret = NT_STATUS_IS_OK ( nt_status ) ? True : False ;
2001-10-31 13:46:25 +03:00
data_blob_free ( & local_lm_blob ) ;
data_blob_free ( & local_nt_blob ) ;
2004-04-03 19:41:32 +04:00
data_blob_free ( & lm_interactive_blob ) ;
data_blob_free ( & nt_interactive_blob ) ;
2001-10-31 13:46:25 +03:00
return ret ;
}
}
/****************************************************************************
Create an auth_usersupplied_data structure
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-10-19 04:40:25 +04:00
bool make_user_info_for_reply ( auth_usersupplied_info * * user_info ,
2002-01-20 11:58:21 +03:00
const char * smb_name ,
const char * client_domain ,
2002-01-05 07:55:41 +03:00
const uint8 chal [ 8 ] ,
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
DATA_BLOB plaintext_password )
2001-10-31 13:46:25 +03:00
{
DATA_BLOB local_lm_blob ;
DATA_BLOB local_nt_blob ;
2002-09-25 19:19:00 +04:00
NTSTATUS ret = NT_STATUS_UNSUCCESSFUL ;
2001-10-31 13:46:25 +03:00
/*
* Not encrypted - do so .
*/
2006-02-04 01:19:41 +03:00
DEBUG ( 5 , ( " make_user_info_for_reply: User passwords not in encrypted "
" format. \n " ) ) ;
2001-10-31 13:46:25 +03:00
if ( plaintext_password . data ) {
unsigned char local_lm_response [ 24 ] ;
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
# ifdef DEBUG_PASSWORD
2006-02-04 01:19:41 +03:00
DEBUG ( 10 , ( " Unencrypted password (len %d): \n " ,
( int ) plaintext_password . length ) ) ;
2007-03-28 17:34:59 +04:00
dump_data ( 100 , plaintext_password . data ,
2006-02-04 01:19:41 +03:00
plaintext_password . length ) ;
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
# endif
2001-10-31 13:46:25 +03:00
2006-02-04 01:19:41 +03:00
SMBencrypt ( ( const char * ) plaintext_password . data ,
( const uchar * ) chal , local_lm_response ) ;
2001-10-31 13:46:25 +03:00
local_lm_blob = data_blob ( local_lm_response , 24 ) ;
2001-11-22 00:10:13 +03:00
/* We can't do an NT hash here, as the password needs to be
case insensitive */
2007-05-14 16:16:20 +04:00
local_nt_blob = data_blob_null ;
2001-10-31 13:46:25 +03:00
} else {
2007-05-14 16:16:20 +04:00
local_lm_blob = data_blob_null ;
local_nt_blob = data_blob_null ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
ret = make_user_info_map (
user_info , smb_name , client_domain ,
get_remote_machine_name ( ) ,
local_lm_blob . data ? & local_lm_blob : NULL ,
local_nt_blob . data ? & local_nt_blob : NULL ,
NULL , NULL ,
plaintext_password . data ? & plaintext_password : NULL ,
False ) ;
2001-10-31 13:46:25 +03:00
data_blob_free ( & local_lm_blob ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_IS_OK ( ret ) ? True : False ;
2001-10-31 13:46:25 +03:00
}
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
/****************************************************************************
Create an auth_usersupplied_data structure
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2002-09-25 19:19:00 +04:00
NTSTATUS make_user_info_for_reply_enc ( auth_usersupplied_info * * user_info ,
const char * smb_name ,
const char * client_domain ,
DATA_BLOB lm_resp , DATA_BLOB nt_resp )
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
{
return make_user_info_map ( user_info , smb_name ,
2004-04-03 19:41:32 +04:00
client_domain ,
get_remote_machine_name ( ) ,
lm_resp . data ? & lm_resp : NULL ,
nt_resp . data ? & nt_resp : NULL ,
NULL , NULL , NULL ,
True ) ;
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
}
2001-11-09 01:19:01 +03:00
/****************************************************************************
Create a guest user_info blob , for anonymous authenticaion .
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-10-19 04:40:25 +04:00
bool make_user_info_guest ( auth_usersupplied_info * * user_info )
2001-11-09 01:19:01 +03:00
{
2002-09-25 19:19:00 +04:00
NTSTATUS nt_status ;
2001-11-09 01:19:01 +03:00
2002-09-25 19:19:00 +04:00
nt_status = make_user_info ( user_info ,
2004-04-03 19:41:32 +04:00
" " , " " ,
" " , " " ,
" " ,
NULL , NULL ,
NULL , NULL ,
NULL ,
True ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_IS_OK ( nt_status ) ? True : False ;
}
2006-08-29 23:14:25 +04:00
static int server_info_dtor ( auth_serversupplied_info * server_info )
2006-02-04 01:19:41 +03:00
{
2006-11-18 20:05:50 +03:00
TALLOC_FREE ( server_info - > sam_account ) ;
2006-02-04 01:19:41 +03:00
ZERO_STRUCTP ( server_info ) ;
return 0 ;
}
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
/***************************************************************************
2006-02-20 20:59:58 +03:00
Make a server_info struct . Free with TALLOC_FREE ( ) .
2006-02-04 01:19:41 +03:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
static auth_serversupplied_info * make_server_info ( TALLOC_CTX * mem_ctx )
{
struct auth_serversupplied_info * result ;
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
result = TALLOC_ZERO_P ( mem_ctx , auth_serversupplied_info ) ;
if ( result = = NULL ) {
DEBUG ( 0 , ( " talloc failed \n " ) ) ;
return NULL ;
}
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
talloc_set_destructor ( result , server_info_dtor ) ;
/* Initialise the uid and gid values to something non-zero
which may save us from giving away root access if there
is a bug in allocating these fields . */
result - > uid = - 1 ;
result - > gid = - 1 ;
return result ;
}
2005-12-11 07:39:33 +03:00
2006-02-04 01:19:41 +03:00
/***************************************************************************
2006-02-20 23:09:36 +03:00
Make ( and fill ) a user_info struct from a struct samu
2006-02-04 01:19:41 +03:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
NTSTATUS make_server_info_sam ( auth_serversupplied_info * * server_info ,
2006-02-20 23:09:36 +03:00
struct samu * sampass )
2006-02-04 01:19:41 +03:00
{
NTSTATUS status ;
struct passwd * pwd ;
gid_t * gids ;
auth_serversupplied_info * result ;
2006-08-22 20:01:24 +04:00
int i ;
size_t num_gids ;
DOM_SID unix_group_sid ;
2006-02-04 01:19:41 +03:00
2006-02-12 00:27:08 +03:00
if ( ! ( result = make_server_info ( NULL ) ) ) {
2006-02-04 01:19:41 +03:00
return NT_STATUS_NO_MEMORY ;
2005-12-11 07:39:33 +03:00
}
2007-04-02 07:46:13 +04:00
if ( ! ( pwd = getpwnam_alloc ( result , pdb_get_username ( sampass ) ) ) ) {
DEBUG ( 1 , ( " User %s in passdb, but getpwnam() fails! \n " ,
pdb_get_username ( sampass ) ) ) ;
2007-04-02 09:53:34 +04:00
TALLOC_FREE ( result ) ;
2007-04-02 07:46:13 +04:00
return NT_STATUS_NO_SUCH_USER ;
}
2006-02-04 01:19:41 +03:00
result - > sam_account = sampass ;
2007-04-02 07:46:13 +04:00
/* Ensure thaat the sampass will be freed with the result */
talloc_steal ( result , sampass ) ;
result - > unix_name = pwd - > pw_name ;
/* Ensure that we keep pwd->pw_name, because we will free pwd below */
talloc_steal ( result , pwd - > pw_name ) ;
2006-02-04 01:19:41 +03:00
result - > gid = pwd - > pw_gid ;
result - > uid = pwd - > pw_uid ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( pwd ) ;
2005-12-11 07:39:33 +03:00
2006-02-04 01:19:41 +03:00
status = pdb_enum_group_memberships ( result , sampass ,
& result - > sids , & gids ,
& result - > num_sids ) ;
2005-12-11 07:39:33 +03:00
2006-02-04 01:19:41 +03:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
DEBUG ( 10 , ( " pdb_enum_group_memberships failed: %s \n " ,
nt_errstr ( status ) ) ) ;
result - > sam_account = NULL ; /* Don't free on error exit. */
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( result ) ;
2006-02-04 01:19:41 +03:00
return status ;
2002-09-25 19:19:00 +04:00
}
2006-08-22 20:01:24 +04:00
/* Add the "Unix Group" SID for each gid to catch mapped groups
and their Unix equivalent . This is to solve the backwards
compatibility problem of ' valid users = + ntadmin ' where
ntadmin has been paired with " Domain Admins " in the group
mapping table . Otherwise smb . conf would need to be changed
to ' valid user = " Domain Admins " ' . - - jerry */
num_gids = result - > num_sids ;
for ( i = 0 ; i < num_gids ; i + + ) {
if ( ! gid_to_unix_groups_sid ( gids [ i ] , & unix_group_sid ) ) {
DEBUG ( 1 , ( " make_server_info_sam: Failed to create SID "
" for gid %d! \n " , gids [ i ] ) ) ;
continue ;
}
2006-12-09 05:58:18 +03:00
if ( ! add_sid_to_array_unique ( result , & unix_group_sid ,
& result - > sids , & result - > num_sids ) ) {
result - > sam_account = NULL ; /* Don't free on error exit. */
TALLOC_FREE ( result ) ;
return NT_STATUS_NO_MEMORY ;
}
2006-08-22 20:01:24 +04:00
}
2005-01-13 21:20:37 +03:00
2006-02-04 01:19:41 +03:00
/* For now we throw away the gids and convert via sid_to_gid
* later . This needs fixing , but I ' d like to get the code straight and
* simple first . */
2006-08-22 20:01:24 +04:00
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( gids ) ;
2005-01-13 21:20:37 +03:00
2006-02-04 01:19:41 +03:00
DEBUG ( 5 , ( " make_server_info_sam: made server info for user %s -> %s \n " ,
pdb_get_username ( sampass ) , result - > unix_name ) ) ;
2005-02-11 13:32:46 +03:00
2006-02-04 01:19:41 +03:00
* server_info = result ;
2005-12-18 21:06:15 +03:00
2006-02-04 01:19:41 +03:00
return NT_STATUS_OK ;
}
static NTSTATUS log_nt_token ( TALLOC_CTX * tmp_ctx , NT_USER_TOKEN * token )
{
char * command ;
char * group_sidstr ;
size_t i ;
if ( ( lp_log_nt_token_command ( ) = = NULL ) | |
( strlen ( lp_log_nt_token_command ( ) ) = = 0 ) ) {
return NT_STATUS_OK ;
2005-02-11 13:32:46 +03:00
}
2006-02-04 01:19:41 +03:00
group_sidstr = talloc_strdup ( tmp_ctx , " " ) ;
for ( i = 1 ; i < token - > num_sids ; i + + ) {
group_sidstr = talloc_asprintf (
tmp_ctx , " %s %s " , group_sidstr ,
2007-12-15 23:10:58 +03:00
sid_string_talloc ( tmp_ctx , & token - > user_sids [ i ] ) ) ;
2006-02-04 01:19:41 +03:00
}
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
command = talloc_string_sub (
tmp_ctx , lp_log_nt_token_command ( ) ,
2007-12-15 23:10:58 +03:00
" %s " , sid_string_talloc ( tmp_ctx , & token - > user_sids [ 0 ] ) ) ;
2006-02-04 01:19:41 +03:00
command = talloc_string_sub ( tmp_ctx , command , " %t " , group_sidstr ) ;
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
if ( command = = NULL ) {
return NT_STATUS_NO_MEMORY ;
}
DEBUG ( 8 , ( " running command: [%s] \n " , command ) ) ;
if ( smbrun ( command , NULL ) ! = 0 ) {
DEBUG ( 0 , ( " Could not log NT token \n " ) ) ;
return NT_STATUS_ACCESS_DENIED ;
}
return NT_STATUS_OK ;
}
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
/*
* Create the token to use from server_info - > sam_account and
* server_info - > sids ( the info3 / sam groups ) . Find the unix gids .
*/
2005-09-30 21:13:37 +04:00
2006-02-04 01:19:41 +03:00
NTSTATUS create_local_token ( auth_serversupplied_info * server_info )
2005-09-30 21:13:37 +04:00
{
2006-02-04 01:19:41 +03:00
TALLOC_CTX * mem_ctx ;
NTSTATUS status ;
size_t i ;
2005-09-30 21:13:37 +04:00
2006-02-04 01:19:41 +03:00
mem_ctx = talloc_new ( NULL ) ;
if ( mem_ctx = = NULL ) {
DEBUG ( 0 , ( " talloc_new failed \n " ) ) ;
return NT_STATUS_NO_MEMORY ;
2005-09-30 21:13:37 +04:00
}
2006-02-04 01:19:41 +03:00
2006-07-13 20:28:38 +04:00
/*
* If winbind is not around , we can not make much use of the SIDs the
* domain controller provided us with . Likewise if the user name was
* mapped to some local unix user .
*/
2006-08-04 04:00:15 +04:00
if ( ( ( lp_server_role ( ) = = ROLE_DOMAIN_MEMBER ) & & ! winbind_ping ( ) ) | |
( server_info - > was_mapped ) ) {
2006-05-06 23:24:35 +04:00
status = create_token_from_username ( server_info ,
server_info - > unix_name ,
server_info - > guest ,
& server_info - > uid ,
& server_info - > gid ,
& server_info - > unix_name ,
& server_info - > ptok ) ;
2006-07-13 19:37:58 +04:00
2006-05-06 23:24:35 +04:00
} else {
2006-07-13 19:37:58 +04:00
server_info - > ptok = create_local_nt_token (
2006-05-06 23:24:35 +04:00
server_info ,
pdb_get_user_sid ( server_info - > sam_account ) ,
server_info - > guest ,
server_info - > num_sids , server_info - > sids ) ;
2006-07-13 19:37:58 +04:00
status = server_info - > ptok ?
NT_STATUS_OK : NT_STATUS_NO_SUCH_USER ;
2006-03-08 18:18:14 +03:00
}
2006-07-13 19:03:46 +04:00
2006-07-13 19:37:58 +04:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
TALLOC_FREE ( mem_ctx ) ;
return status ;
}
2006-02-04 01:19:41 +03:00
/* Convert the SIDs to gids. */
server_info - > n_groups = 0 ;
server_info - > groups = NULL ;
/* Start at index 1, where the groups start. */
2007-04-20 02:26:09 +04:00
for ( i = 1 ; i < server_info - > ptok - > num_sids ; i + + ) {
gid_t gid ;
DOM_SID * sid = & server_info - > ptok - > user_sids [ i ] ;
2006-02-04 01:19:41 +03:00
2007-04-20 02:26:09 +04:00
if ( ! sid_to_gid ( sid , & gid ) ) {
DEBUG ( 10 , ( " Could not convert SID %s to gid, "
" ignoring it \n " , sid_string_static ( sid ) ) ) ;
continue ;
2006-12-09 05:58:18 +03:00
}
2007-04-20 02:26:09 +04:00
add_gid_to_array_unique ( server_info , gid , & server_info - > groups ,
& server_info - > n_groups ) ;
2006-02-04 01:19:41 +03:00
}
2007-04-20 02:26:09 +04:00
2006-02-04 01:19:41 +03:00
debug_nt_user_token ( DBGC_AUTH , 10 , server_info - > ptok ) ;
status = log_nt_token ( mem_ctx , server_info - > ptok ) ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( mem_ctx ) ;
2006-02-04 01:19:41 +03:00
return status ;
2005-09-30 21:13:37 +04:00
}
2006-02-04 01:19:41 +03:00
/*
* Create an artificial NT token given just a username . ( Initially indended
* for force user )
2002-09-25 19:19:00 +04:00
*
2006-02-04 01:19:41 +03:00
* We go through lookup_name ( ) to avoid problems we had with ' winbind use
* default domain ' .
2003-10-07 20:34:23 +04:00
*
2006-02-04 01:19:41 +03:00
* We have 3 cases :
*
* unmapped unix users : Go directly to nss to find the user ' s group .
*
* A passdb user : The list of groups is provided by pdb_enum_group_memberships .
*
* If the user is provided by winbind , the primary gid is set to " domain
* users " of the user's domain. For an explanation why this is necessary, see
* the thread starting at
* http : //lists.samba.org/archive/samba-technical/2006-January/044803.html.
*/
NTSTATUS create_token_from_username ( TALLOC_CTX * mem_ctx , const char * username ,
2007-10-19 04:40:25 +04:00
bool is_guest ,
2006-02-04 01:19:41 +03:00
uid_t * uid , gid_t * gid ,
char * * found_username ,
struct nt_user_token * * token )
2002-09-25 19:19:00 +04:00
{
2006-02-04 01:19:41 +03:00
NTSTATUS result = NT_STATUS_NO_SUCH_USER ;
TALLOC_CTX * tmp_ctx ;
DOM_SID user_sid ;
2006-09-08 18:28:06 +04:00
enum lsa_SidType type ;
2006-02-04 01:19:41 +03:00
gid_t * gids ;
DOM_SID * group_sids ;
2006-08-23 06:45:45 +04:00
DOM_SID unix_group_sid ;
2006-02-04 01:19:41 +03:00
size_t num_group_sids ;
2006-08-23 06:45:45 +04:00
size_t num_gids ;
size_t i ;
2003-02-24 05:35:54 +03:00
2006-02-04 01:19:41 +03:00
tmp_ctx = talloc_new ( NULL ) ;
if ( tmp_ctx = = NULL ) {
DEBUG ( 0 , ( " talloc_new failed \n " ) ) ;
return NT_STATUS_NO_MEMORY ;
}
2002-09-25 19:19:00 +04:00
2006-08-05 00:35:52 +04:00
if ( ! lookup_name_smbconf ( tmp_ctx , username , LOOKUP_NAME_ALL ,
2006-02-04 01:19:41 +03:00
NULL , NULL , & user_sid , & type ) ) {
2006-08-05 00:35:52 +04:00
DEBUG ( 1 , ( " lookup_name_smbconf for %s failed \n " , username ) ) ;
2006-02-04 01:19:41 +03:00
goto done ;
}
2003-08-10 03:12:35 +04:00
2006-02-04 01:19:41 +03:00
if ( type ! = SID_NAME_USER ) {
DEBUG ( 1 , ( " %s is a %s, not a user \n " , username ,
sid_type_lookup ( type ) ) ) ;
goto done ;
2002-09-25 19:19:00 +04:00
}
2004-11-12 18:49:47 +03:00
2006-02-04 01:19:41 +03:00
if ( ! sid_to_uid ( & user_sid , uid ) ) {
DEBUG ( 1 , ( " sid_to_uid for %s (%s) failed \n " ,
username , sid_string_static ( & user_sid ) ) ) ;
goto done ;
}
2003-08-10 03:12:35 +04:00
2006-07-07 22:22:26 +04:00
if ( sid_check_is_in_our_domain ( & user_sid ) ) {
2007-10-19 04:40:25 +04:00
bool ret ;
2003-07-08 09:37:13 +04:00
2006-02-04 01:19:41 +03:00
/* This is a passdb user, so ask passdb */
2003-07-08 09:37:13 +04:00
2006-02-20 23:09:36 +03:00
struct samu * sam_acct = NULL ;
2001-10-31 13:46:25 +03:00
2006-02-21 17:34:11 +03:00
if ( ! ( sam_acct = samu_new ( tmp_ctx ) ) ) {
2006-06-15 05:54:09 +04:00
result = NT_STATUS_NO_MEMORY ;
2006-02-04 01:19:41 +03:00
goto done ;
}
2001-11-11 14:34:46 +03:00
2007-10-12 15:20:07 +04:00
become_root ( ) ;
ret = pdb_getsampwsid ( sam_acct , & user_sid ) ;
unbecome_root ( ) ;
if ( ! ret ) {
2006-02-04 01:19:41 +03:00
DEBUG ( 1 , ( " pdb_getsampwsid(%s) for user %s failed \n " ,
sid_string_static ( & user_sid ) , username ) ) ;
2006-07-07 22:53:19 +04:00
DEBUGADD ( 1 , ( " Fall back to unix user %s \n " , username ) ) ;
goto unix_user ;
2006-02-04 01:19:41 +03:00
}
2002-09-25 19:19:00 +04:00
2006-12-10 08:23:47 +03:00
result = pdb_enum_group_memberships ( tmp_ctx , sam_acct ,
2006-02-04 01:19:41 +03:00
& group_sids , & gids ,
& num_group_sids ) ;
if ( ! NT_STATUS_IS_OK ( result ) ) {
DEBUG ( 10 , ( " enum_group_memberships failed for %s \n " ,
username ) ) ;
2006-08-04 03:44:07 +04:00
DEBUGADD ( 1 , ( " Fall back to unix user %s \n " , username ) ) ;
goto unix_user ;
2006-02-04 01:19:41 +03:00
}
2001-10-31 13:46:25 +03:00
2006-08-04 04:00:15 +04:00
/* see the smb_panic() in pdb_default_enum_group_memberships */
SMB_ASSERT ( num_group_sids > 0 ) ;
* gid = gids [ 0 ] ;
2006-12-10 08:23:47 +03:00
/* Ensure we're returning the found_username on the right context. */
2006-02-04 01:19:41 +03:00
* found_username = talloc_strdup ( mem_ctx ,
pdb_get_username ( sam_acct ) ) ;
2003-06-12 18:24:15 +04:00
2006-07-07 22:22:26 +04:00
} else if ( sid_check_is_in_unix_users ( & user_sid ) ) {
/* This is a unix user not in passdb. We need to ask nss
* directly , without consulting passdb */
struct passwd * pass ;
2006-07-07 22:53:19 +04:00
/*
* This goto target is used as a fallback for the passdb
* case . The concrete bug report is when passdb gave us an
* unmapped gid .
*/
unix_user :
uid_to_unix_users_sid ( * uid , & user_sid ) ;
2006-07-07 22:22:26 +04:00
pass = getpwuid_alloc ( tmp_ctx , * uid ) ;
if ( pass = = NULL ) {
DEBUG ( 1 , ( " getpwuid(%d) for user %s failed \n " ,
* uid , username ) ) ;
goto done ;
}
if ( ! getgroups_unix_user ( tmp_ctx , username , pass - > pw_gid ,
& gids , & num_group_sids ) ) {
DEBUG ( 1 , ( " getgroups_unix_user for user %s failed \n " ,
username ) ) ;
goto done ;
}
2007-04-30 06:39:34 +04:00
if ( num_group_sids ) {
group_sids = TALLOC_ARRAY ( tmp_ctx , DOM_SID , num_group_sids ) ;
if ( group_sids = = NULL ) {
DEBUG ( 1 , ( " TALLOC_ARRAY failed \n " ) ) ;
result = NT_STATUS_NO_MEMORY ;
goto done ;
}
} else {
group_sids = NULL ;
2006-07-07 22:22:26 +04:00
}
for ( i = 0 ; i < num_group_sids ; i + + ) {
gid_to_sid ( & group_sids [ i ] , gids [ i ] ) ;
}
2006-08-04 04:00:15 +04:00
/* In getgroups_unix_user we always set the primary gid */
SMB_ASSERT ( num_group_sids > 0 ) ;
* gid = gids [ 0 ] ;
2006-07-07 22:22:26 +04:00
2006-12-10 08:23:47 +03:00
/* Ensure we're returning the found_username on the right context. */
* found_username = talloc_strdup ( mem_ctx , pass - > pw_name ) ;
2006-02-04 01:19:41 +03:00
} else {
2003-06-12 18:24:15 +04:00
2006-02-04 01:19:41 +03:00
/* This user is from winbind, force the primary gid to the
* user ' s " domain users " group . Under certain circumstances
* ( user comes from NT4 ) , this might be a loss of
* information . But we can not rely on winbind getting the
* correct info . AD might prohibit winbind looking up that
* information . */
2003-06-12 18:24:15 +04:00
2006-02-04 01:19:41 +03:00
uint32 dummy ;
2003-06-12 18:24:15 +04:00
2006-12-09 05:58:18 +03:00
num_group_sids = 1 ;
2007-04-28 03:18:41 +04:00
group_sids = TALLOC_ARRAY ( tmp_ctx , DOM_SID , num_group_sids ) ;
2006-12-09 05:58:18 +03:00
if ( group_sids = = NULL ) {
2007-04-28 03:18:41 +04:00
DEBUG ( 1 , ( " TALLOC_ARRAY failed \n " ) ) ;
2006-12-09 05:58:18 +03:00
result = NT_STATUS_NO_MEMORY ;
goto done ;
}
sid_copy ( & group_sids [ 0 ] , & user_sid ) ;
sid_split_rid ( & group_sids [ 0 ] , & dummy ) ;
sid_append_rid ( & group_sids [ 0 ] , DOMAIN_GROUP_RID_USERS ) ;
2003-06-12 18:24:15 +04:00
2006-12-09 05:58:18 +03:00
if ( ! sid_to_gid ( & group_sids [ 0 ] , gid ) ) {
2006-02-04 01:19:41 +03:00
DEBUG ( 1 , ( " sid_to_gid(%s) failed \n " ,
2006-12-09 05:58:18 +03:00
sid_string_static ( & group_sids [ 0 ] ) ) ) ;
2006-02-04 01:19:41 +03:00
goto done ;
}
2003-06-12 18:24:15 +04:00
2006-09-30 01:26:33 +04:00
gids = gid ;
2003-06-12 18:24:15 +04:00
2006-12-10 08:23:47 +03:00
/* Ensure we're returning the found_username on the right context. */
2006-02-04 01:19:41 +03:00
* found_username = talloc_strdup ( mem_ctx , username ) ;
2003-07-03 18:36:42 +04:00
}
2003-06-12 18:24:15 +04:00
2006-08-23 06:45:45 +04:00
/* Add the "Unix Group" SID for each gid to catch mapped groups
and their Unix equivalent . This is to solve the backwards
compatibility problem of ' valid users = + ntadmin ' where
ntadmin has been paired with " Domain Admins " in the group
mapping table . Otherwise smb . conf would need to be changed
to ' valid user = " Domain Admins " ' . - - jerry */
num_gids = num_group_sids ;
for ( i = 0 ; i < num_gids ; i + + ) {
gid_t high , low ;
/* don't pickup anything managed by Winbind */
if ( lp_idmap_gid ( & low , & high ) & & ( gids [ i ] > = low ) & & ( gids [ i ] < = high ) )
continue ;
if ( ! gid_to_unix_groups_sid ( gids [ i ] , & unix_group_sid ) ) {
DEBUG ( 1 , ( " create_token_from_username: Failed to create SID "
" for gid %d! \n " , gids [ i ] ) ) ;
continue ;
}
2006-12-10 08:23:47 +03:00
if ( ! add_sid_to_array_unique ( tmp_ctx , & unix_group_sid ,
2006-12-09 05:58:18 +03:00
& group_sids , & num_group_sids ) ) {
result = NT_STATUS_NO_MEMORY ;
goto done ;
}
2006-08-23 06:45:45 +04:00
}
2006-12-10 08:23:47 +03:00
/* Ensure we're creating the nt_token on the right context. */
2006-08-04 04:00:15 +04:00
* token = create_local_nt_token ( mem_ctx , & user_sid ,
2006-02-04 01:19:41 +03:00
is_guest , num_group_sids , group_sids ) ;
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
if ( ( * token = = NULL ) | | ( * found_username = = NULL ) ) {
result = NT_STATUS_NO_MEMORY ;
goto done ;
}
result = NT_STATUS_OK ;
done :
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( tmp_ctx ) ;
2006-02-04 01:19:41 +03:00
return result ;
2001-10-31 13:46:25 +03:00
}
2006-02-13 20:08:25 +03:00
/***************************************************************************
Build upon create_token_from_username :
Expensive helper function to figure out whether a user given its name is
member of a particular group .
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2006-12-10 08:23:47 +03:00
2007-10-19 04:40:25 +04:00
bool user_in_group_sid ( const char * username , const DOM_SID * group_sid )
2006-02-13 20:08:25 +03:00
{
NTSTATUS status ;
uid_t uid ;
gid_t gid ;
char * found_username ;
struct nt_user_token * token ;
2007-10-19 04:40:25 +04:00
bool result ;
2006-02-13 20:08:25 +03:00
TALLOC_CTX * mem_ctx ;
mem_ctx = talloc_new ( NULL ) ;
if ( mem_ctx = = NULL ) {
DEBUG ( 0 , ( " talloc_new failed \n " ) ) ;
return False ;
}
status = create_token_from_username ( mem_ctx , username , False ,
& uid , & gid , & found_username ,
& token ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
DEBUG ( 10 , ( " could not create token for %s \n " , username ) ) ;
return False ;
}
result = nt_token_check_sid ( group_sid , token ) ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( mem_ctx ) ;
2006-02-13 20:08:25 +03:00
return result ;
}
2007-10-19 04:40:25 +04:00
bool user_in_group ( const char * username , const char * groupname )
2006-02-13 20:08:25 +03:00
{
TALLOC_CTX * mem_ctx ;
DOM_SID group_sid ;
2007-10-19 04:40:25 +04:00
bool ret ;
2006-02-13 20:08:25 +03:00
mem_ctx = talloc_new ( NULL ) ;
if ( mem_ctx = = NULL ) {
DEBUG ( 0 , ( " talloc_new failed \n " ) ) ;
return False ;
}
ret = lookup_name ( mem_ctx , groupname , LOOKUP_NAME_ALL ,
NULL , NULL , & group_sid , NULL ) ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( mem_ctx ) ;
2006-02-13 20:08:25 +03:00
if ( ! ret ) {
2006-05-18 06:05:53 +04:00
DEBUG ( 10 , ( " lookup_name for (%s) failed. \n " , groupname ) ) ;
2006-02-13 20:08:25 +03:00
return False ;
}
return user_in_group_sid ( username , & group_sid ) ;
}
2001-11-11 14:34:46 +03:00
/***************************************************************************
Make ( and fill ) a user_info struct from a ' struct passwd ' by conversion
2006-02-20 23:09:36 +03:00
to a struct samu
2001-11-11 14:34:46 +03:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2004-03-16 19:41:54 +03:00
NTSTATUS make_server_info_pw ( auth_serversupplied_info * * server_info ,
char * unix_username ,
struct passwd * pwd )
2001-10-31 13:46:25 +03:00
{
2006-02-04 01:19:41 +03:00
NTSTATUS status ;
2006-02-20 23:09:36 +03:00
struct samu * sampass = NULL ;
2006-02-04 01:19:41 +03:00
gid_t * gids ;
2007-06-17 23:23:32 +04:00
char * qualified_name = NULL ;
TALLOC_CTX * mem_ctx = NULL ;
DOM_SID u_sid ;
enum lsa_SidType type ;
2006-02-04 01:19:41 +03:00
auth_serversupplied_info * result ;
2006-02-21 17:34:11 +03:00
if ( ! ( sampass = samu_new ( NULL ) ) ) {
return NT_STATUS_NO_MEMORY ;
}
status = samu_set_unix ( sampass , pwd ) ;
2006-02-04 01:19:41 +03:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
result = make_server_info ( NULL ) ;
2006-06-12 15:03:49 +04:00
if ( result = = NULL ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sampass ) ;
2006-06-12 15:03:49 +04:00
return NT_STATUS_NO_MEMORY ;
2003-06-12 18:24:15 +04:00
}
2006-02-04 01:19:41 +03:00
result - > sam_account = sampass ;
result - > unix_name = talloc_strdup ( result , unix_username ) ;
result - > uid = pwd - > pw_uid ;
result - > gid = pwd - > pw_gid ;
2003-06-12 18:24:15 +04:00
2006-02-04 01:19:41 +03:00
status = pdb_enum_group_memberships ( result , sampass ,
& result - > sids , & gids ,
& result - > num_sids ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
DEBUG ( 10 , ( " pdb_enum_group_memberships failed: %s \n " ,
nt_errstr ( status ) ) ) ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( result ) ;
2006-02-04 01:19:41 +03:00
return status ;
2003-06-12 18:24:15 +04:00
}
2007-06-17 23:23:32 +04:00
/*
* The SID returned in server_info - > sam_account is based
* on our SAM sid even though for a pure UNIX account this should
* not be the case as it doesn ' t really exist in the SAM db .
* This causes lookups on " [in]valid users " to fail as they
* will lookup this name as a " Unix User " SID to check against
* the user token . Fix this by adding the " Unix User " \ unix_username
* SID to the sid array . The correct fix should probably be
* changing the server_info - > sam_account user SID to be a
* S - 1 - 22 Unix SID , but this might break old configs where
* plaintext passwords were used with no SAM backend .
*/
mem_ctx = talloc_init ( " make_server_info_pw_tmp " ) ;
if ( ! mem_ctx ) {
TALLOC_FREE ( result ) ;
return NT_STATUS_NO_MEMORY ;
}
qualified_name = talloc_asprintf ( mem_ctx , " %s \\ %s " ,
unix_users_domain_name ( ) ,
unix_username ) ;
if ( ! qualified_name ) {
TALLOC_FREE ( result ) ;
TALLOC_FREE ( mem_ctx ) ;
return NT_STATUS_NO_MEMORY ;
}
if ( ! lookup_name ( mem_ctx , qualified_name , LOOKUP_NAME_ALL ,
NULL , NULL ,
& u_sid , & type ) ) {
TALLOC_FREE ( result ) ;
TALLOC_FREE ( mem_ctx ) ;
return NT_STATUS_NO_SUCH_USER ;
}
TALLOC_FREE ( mem_ctx ) ;
if ( type ! = SID_NAME_USER ) {
TALLOC_FREE ( result ) ;
return NT_STATUS_NO_SUCH_USER ;
}
if ( ! add_sid_to_array_unique ( result , & u_sid ,
& result - > sids ,
& result - > num_sids ) ) {
TALLOC_FREE ( result ) ;
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
/* For now we throw away the gids and convert via sid_to_gid
* later . This needs fixing , but I ' d like to get the code straight and
* simple first . */
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( gids ) ;
2003-07-03 18:36:42 +04:00
2006-02-04 01:19:41 +03:00
* server_info = result ;
return NT_STATUS_OK ;
2001-10-31 13:46:25 +03:00
}
2001-11-11 14:34:46 +03:00
/***************************************************************************
2002-09-25 19:19:00 +04:00
Make ( and fill ) a user_info struct for a guest login .
2006-02-04 01:19:41 +03:00
This * must * succeed for smbd to start . If there is no mapping entry for
the guest gid , then create one .
2006-06-11 20:13:41 +04:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2001-11-11 14:34:46 +03:00
2004-11-06 01:53:35 +03:00
static NTSTATUS make_new_server_info_guest ( auth_serversupplied_info * * server_info )
2001-10-31 13:46:25 +03:00
{
2006-02-04 01:19:41 +03:00
NTSTATUS status ;
2006-02-20 23:09:36 +03:00
struct samu * sampass = NULL ;
2002-09-25 19:19:00 +04:00
DOM_SID guest_sid ;
2007-10-19 04:40:25 +04:00
bool ret ;
2006-09-04 13:51:47 +04:00
static const char zeros [ 16 ] = { 0 , } ;
2002-09-25 19:19:00 +04:00
2006-02-21 17:34:11 +03:00
if ( ! ( sampass = samu_new ( NULL ) ) ) {
return NT_STATUS_NO_MEMORY ;
2001-10-31 13:46:25 +03:00
}
2002-09-25 19:19:00 +04:00
sid_copy ( & guest_sid , get_global_sam_sid ( ) ) ;
sid_append_rid ( & guest_sid , DOMAIN_USER_RID_GUEST ) ;
2001-11-04 02:34:24 +03:00
2002-11-16 00:43:57 +03:00
become_root ( ) ;
2006-02-04 01:19:41 +03:00
ret = pdb_getsampwsid ( sampass , & guest_sid ) ;
unbecome_root ( ) ;
if ( ! ret ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sampass ) ;
2002-09-25 19:19:00 +04:00
return NT_STATUS_NO_SUCH_USER ;
2001-10-31 13:46:25 +03:00
}
2006-02-04 01:19:41 +03:00
status = make_server_info_sam ( server_info , sampass ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sampass ) ;
2006-02-12 00:27:08 +03:00
return status ;
2003-02-24 05:35:54 +03:00
}
2006-02-04 01:19:41 +03:00
( * server_info ) - > guest = True ;
2002-09-25 19:19:00 +04:00
2006-02-04 01:19:41 +03:00
status = create_local_token ( * server_info ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
DEBUG ( 10 , ( " create_local_token failed: %s \n " ,
nt_errstr ( status ) ) ) ;
return status ;
}
/* annoying, but the Guest really does have a session key, and it is
all zeros ! */
( * server_info ) - > user_session_key = data_blob ( zeros , sizeof ( zeros ) ) ;
( * server_info ) - > lm_session_key = data_blob ( zeros , sizeof ( zeros ) ) ;
return NT_STATUS_OK ;
2001-10-31 13:46:25 +03:00
}
2004-11-06 01:53:35 +03:00
static auth_serversupplied_info * copy_serverinfo ( auth_serversupplied_info * src )
{
auth_serversupplied_info * dst ;
2006-02-04 01:19:41 +03:00
dst = make_server_info ( NULL ) ;
if ( dst = = NULL ) {
2004-11-06 01:53:35 +03:00
return NULL ;
2006-02-04 01:19:41 +03:00
}
2004-11-06 01:53:35 +03:00
dst - > guest = src - > guest ;
dst - > uid = src - > uid ;
dst - > gid = src - > gid ;
dst - > n_groups = src - > n_groups ;
2006-06-15 05:54:09 +04:00
if ( src - > n_groups ! = 0 ) {
2007-04-28 03:18:41 +04:00
dst - > groups = ( gid_t * ) TALLOC_MEMDUP (
2006-08-04 16:15:53 +04:00
dst , src - > groups , sizeof ( gid_t ) * dst - > n_groups ) ;
2006-06-15 05:54:09 +04:00
} else {
2004-11-06 01:53:35 +03:00
dst - > groups = NULL ;
2006-06-15 05:54:09 +04:00
}
if ( src - > ptok ) {
dst - > ptok = dup_nt_token ( dst , src - > ptok ) ;
if ( ! dst - > ptok ) {
TALLOC_FREE ( dst ) ;
return NULL ;
}
}
2006-02-25 00:36:40 +03:00
dst - > user_session_key = data_blob_talloc ( dst , src - > user_session_key . data ,
2006-06-15 05:54:09 +04:00
src - > user_session_key . length ) ;
2006-02-25 00:36:40 +03:00
dst - > lm_session_key = data_blob_talloc ( dst , src - > lm_session_key . data ,
2006-06-15 05:54:09 +04:00
src - > lm_session_key . length ) ;
dst - > sam_account = samu_new ( NULL ) ;
if ( ! dst - > sam_account ) {
TALLOC_FREE ( dst ) ;
return NULL ;
}
if ( ! pdb_copy_sam_account ( dst - > sam_account , src - > sam_account ) ) {
TALLOC_FREE ( dst ) ;
return NULL ;
}
2006-02-25 00:36:40 +03:00
2004-11-06 01:53:35 +03:00
dst - > pam_handle = NULL ;
2006-02-04 01:19:41 +03:00
dst - > unix_name = talloc_strdup ( dst , src - > unix_name ) ;
2006-06-15 05:54:09 +04:00
if ( ! dst - > unix_name ) {
TALLOC_FREE ( dst ) ;
return NULL ;
}
2004-11-06 01:53:35 +03:00
return dst ;
}
static auth_serversupplied_info * guest_info = NULL ;
2007-10-19 04:40:25 +04:00
bool init_guest_info ( void )
2004-11-06 01:53:35 +03:00
{
if ( guest_info ! = NULL )
return True ;
return NT_STATUS_IS_OK ( make_new_server_info_guest ( & guest_info ) ) ;
}
NTSTATUS make_server_info_guest ( auth_serversupplied_info * * server_info )
{
* server_info = copy_serverinfo ( guest_info ) ;
return ( * server_info ! = NULL ) ? NT_STATUS_OK : NT_STATUS_NO_MEMORY ;
}
2007-10-19 04:40:25 +04:00
bool copy_current_user ( struct current_user * dst , struct current_user * src )
2006-07-11 22:01:26 +04:00
{
gid_t * groups ;
NT_USER_TOKEN * nt_token ;
2006-08-04 16:15:53 +04:00
groups = ( gid_t * ) memdup ( src - > ut . groups ,
sizeof ( gid_t ) * src - > ut . ngroups ) ;
2006-07-11 22:01:26 +04:00
if ( ( src - > ut . ngroups ! = 0 ) & & ( groups = = NULL ) ) {
return False ;
}
nt_token = dup_nt_token ( NULL , src - > nt_user_token ) ;
if ( nt_token = = NULL ) {
SAFE_FREE ( groups ) ;
return False ;
}
dst - > conn = src - > conn ;
dst - > vuid = src - > vuid ;
dst - > ut . uid = src - > ut . uid ;
dst - > ut . gid = src - > ut . gid ;
dst - > ut . ngroups = src - > ut . ngroups ;
dst - > ut . groups = groups ;
dst - > nt_user_token = nt_token ;
return True ;
}
2007-10-19 04:40:25 +04:00
bool set_current_user_guest ( struct current_user * dst )
2006-07-11 22:01:26 +04:00
{
gid_t * groups ;
NT_USER_TOKEN * nt_token ;
2006-08-04 16:15:53 +04:00
groups = ( gid_t * ) memdup ( guest_info - > groups ,
sizeof ( gid_t ) * guest_info - > n_groups ) ;
2006-07-11 22:01:26 +04:00
if ( groups = = NULL ) {
return False ;
}
nt_token = dup_nt_token ( NULL , guest_info - > ptok ) ;
if ( nt_token = = NULL ) {
SAFE_FREE ( groups ) ;
return False ;
}
TALLOC_FREE ( dst - > nt_user_token ) ;
SAFE_FREE ( dst - > ut . groups ) ;
/* dst->conn is never really dereferenced, it's only tested for
* equality in uid . c */
dst - > conn = NULL ;
dst - > vuid = UID_FIELD_INVALID ;
dst - > ut . uid = guest_info - > uid ;
dst - > ut . gid = guest_info - > gid ;
dst - > ut . ngroups = guest_info - > n_groups ;
dst - > ut . groups = groups ;
dst - > nt_user_token = nt_token ;
return True ;
}
2003-06-12 11:21:26 +04:00
/***************************************************************************
Purely internal function for make_server_info_info3
Fill the sam account from getpwnam
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2003-07-03 18:36:42 +04:00
static NTSTATUS fill_sam_account ( TALLOC_CTX * mem_ctx ,
const char * domain ,
2003-06-11 20:36:04 +04:00
const char * username ,
2003-07-03 18:36:42 +04:00
char * * found_username ,
2003-06-12 18:24:15 +04:00
uid_t * uid , gid_t * gid ,
2006-05-06 23:24:35 +04:00
struct samu * account ,
2007-10-19 04:40:25 +04:00
bool * username_was_mapped )
2003-06-11 20:36:04 +04:00
{
2005-06-07 21:52:19 +04:00
NTSTATUS nt_status ;
2004-10-23 00:15:24 +04:00
fstring dom_user , lower_username ;
2004-03-16 19:41:54 +03:00
fstring real_username ;
2003-06-11 20:36:04 +04:00
struct passwd * passwd ;
2004-10-23 00:15:24 +04:00
fstrcpy ( lower_username , username ) ;
strlower_m ( lower_username ) ;
fstr_sprintf ( dom_user , " %s%c%s " , domain , * lp_winbind_separator ( ) ,
lower_username ) ;
2003-06-11 20:36:04 +04:00
2006-03-20 13:55:48 +03:00
/* Get the passwd struct. Try to create the account is necessary. */
2003-06-11 20:36:04 +04:00
2006-05-06 23:24:35 +04:00
* username_was_mapped = map_username ( dom_user ) ;
2004-10-23 00:15:24 +04:00
2006-02-04 01:19:41 +03:00
if ( ! ( passwd = smb_getpwnam ( NULL , dom_user , real_username , True ) ) )
2003-06-11 20:36:04 +04:00
return NT_STATUS_NO_SUCH_USER ;
2003-06-12 18:24:15 +04:00
* uid = passwd - > pw_uid ;
* gid = passwd - > pw_gid ;
2004-03-16 19:41:54 +03:00
/* This is pointless -- there is no suport for differing
2003-10-07 20:34:23 +04:00
unix and windows names . Make sure to always store the
2003-10-20 20:49:45 +04:00
one we actually looked up and succeeded . Have I mentioned
2003-10-07 20:34:23 +04:00
why I hate the ' winbind use default domain ' parameter ?
- - jerry */
2004-03-16 19:41:54 +03:00
* found_username = talloc_strdup ( mem_ctx , real_username ) ;
2003-12-19 03:33:09 +03:00
2006-02-21 17:34:11 +03:00
DEBUG ( 5 , ( " fill_sam_account: located username was [%s] \n " , * found_username ) ) ;
2003-07-03 18:36:42 +04:00
2006-02-21 17:34:11 +03:00
nt_status = samu_set_unix ( account , passwd ) ;
2006-02-20 20:59:58 +03:00
TALLOC_FREE ( passwd ) ;
2006-02-21 17:34:11 +03:00
2005-06-07 21:52:19 +04:00
return nt_status ;
2003-06-11 20:36:04 +04:00
}
2003-10-20 20:49:45 +04:00
/****************************************************************************
2003-11-06 20:28:44 +03:00
Wrapper to allow the getpwnam ( ) call to strip the domain name and
try again in case a local UNIX user is already there . Also run through
the username if we fallback to the username only .
2003-10-20 20:49:45 +04:00
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2006-02-04 01:19:41 +03:00
struct passwd * smb_getpwnam ( TALLOC_CTX * mem_ctx , char * domuser ,
2007-10-19 04:40:25 +04:00
fstring save_username , bool create )
2003-10-20 20:49:45 +04:00
{
2003-12-06 05:34:02 +03:00
struct passwd * pw = NULL ;
2003-10-20 20:49:45 +04:00
char * p ;
2004-07-07 01:43:12 +04:00
fstring username ;
2004-03-16 19:41:54 +03:00
/* we only save a copy of the username it has been mangled
by winbindd use default domain */
save_username [ 0 ] = ' \0 ' ;
2004-07-07 01:43:12 +04:00
/* don't call map_username() here since it has to be done higher
up the stack so we don ' t call it mutliple times */
fstrcpy ( username , domuser ) ;
2004-03-16 19:41:54 +03:00
2004-07-07 01:43:12 +04:00
p = strchr_m ( username , * lp_winbind_separator ( ) ) ;
2004-03-16 19:41:54 +03:00
/* code for a DOMAIN\user string */
if ( p ) {
2004-07-07 01:43:12 +04:00
fstring strip_username ;
2006-02-04 01:19:41 +03:00
pw = Get_Pwnam_alloc ( mem_ctx , domuser ) ;
2004-03-16 19:41:54 +03:00
if ( pw ) {
/* make sure we get the case of the username correct */
/* work around 'winbind use default domain = yes' */
if ( ! strchr_m ( pw - > pw_name , * lp_winbind_separator ( ) ) ) {
char * domain ;
2004-07-07 01:43:12 +04:00
/* split the domain and username into 2 strings */
2004-03-16 19:41:54 +03:00
* p = ' \0 ' ;
2004-07-07 01:43:12 +04:00
domain = username ;
2004-03-16 19:41:54 +03:00
fstr_sprintf ( save_username , " %s%c%s " , domain , * lp_winbind_separator ( ) , pw - > pw_name ) ;
}
else
fstrcpy ( save_username , pw - > pw_name ) ;
2003-10-20 20:49:45 +04:00
2004-03-16 19:41:54 +03:00
/* whew -- done! */
return pw ;
}
2003-10-20 20:49:45 +04:00
2004-03-16 19:41:54 +03:00
/* setup for lookup of just the username */
2004-07-07 01:43:12 +04:00
/* remember that p and username are overlapping memory */
2003-10-20 20:49:45 +04:00
2004-03-16 23:28:47 +03:00
p + + ;
fstrcpy ( strip_username , p ) ;
2004-07-07 01:43:12 +04:00
fstrcpy ( username , strip_username ) ;
2004-03-16 19:41:54 +03:00
}
/* just lookup a plain username */
2006-02-04 01:19:41 +03:00
pw = Get_Pwnam_alloc ( mem_ctx , username ) ;
2004-03-16 19:41:54 +03:00
2006-04-14 23:36:36 +04:00
/* Create local user if requested but only if winbindd
is not running . We need to protect against cases
where winbindd is failing and then prematurely
creating users in / etc / passwd */
2004-03-16 19:41:54 +03:00
2006-04-14 23:36:36 +04:00
if ( ! pw & & create & & ! winbind_ping ( ) ) {
2004-03-16 19:41:54 +03:00
/* Don't add a machine account. */
2004-07-07 01:43:12 +04:00
if ( username [ strlen ( username ) - 1 ] = = ' $ ' )
2004-03-16 19:41:54 +03:00
return NULL ;
2003-10-20 20:49:45 +04:00
2005-05-31 06:23:47 +04:00
smb_create_user ( NULL , username , NULL ) ;
2006-02-04 01:19:41 +03:00
pw = Get_Pwnam_alloc ( mem_ctx , username ) ;
2003-11-06 20:28:44 +03:00
}
2004-03-16 19:41:54 +03:00
/* one last check for a valid passwd struct */
if ( pw )
fstrcpy ( save_username , pw - > pw_name ) ;
2003-10-20 20:49:45 +04:00
2003-12-06 05:34:02 +03:00
return pw ;
2003-10-20 20:49:45 +04:00
}
2002-07-15 14:35:28 +04:00
/***************************************************************************
Make a server_info struct from the info3 returned by a domain logon
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
NTSTATUS make_server_info_info3 ( TALLOC_CTX * mem_ctx ,
const char * sent_nt_username ,
const char * domain ,
auth_serversupplied_info * * server_info ,
NET_USER_INFO_3 * info3 )
{
2006-09-04 13:51:47 +04:00
static const char zeros [ 16 ] = { 0 , } ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
2002-07-15 14:35:28 +04:00
NTSTATUS nt_status = NT_STATUS_OK ;
2003-07-03 18:36:42 +04:00
char * found_username ;
2002-07-15 14:35:28 +04:00
const char * nt_domain ;
const char * nt_username ;
2006-02-20 23:09:36 +03:00
struct samu * sam_account = NULL ;
2002-07-15 14:35:28 +04:00
DOM_SID user_sid ;
DOM_SID group_sid ;
2007-10-19 04:40:25 +04:00
bool username_was_mapped ;
2002-07-15 14:35:28 +04:00
2003-06-12 18:24:15 +04:00
uid_t uid ;
gid_t gid ;
2002-07-15 14:35:28 +04:00
2006-02-04 01:19:41 +03:00
auth_serversupplied_info * result ;
2002-07-15 14:35:28 +04:00
/*
Here is where we should check the list of
trusted domains , and verify that the SID
matches .
*/
sid_copy ( & user_sid , & info3 - > dom_sid . sid ) ;
if ( ! sid_append_rid ( & user_sid , info3 - > user_rid ) ) {
return NT_STATUS_INVALID_PARAMETER ;
}
sid_copy ( & group_sid , & info3 - > dom_sid . sid ) ;
if ( ! sid_append_rid ( & group_sid , info3 - > group_rid ) ) {
return NT_STATUS_INVALID_PARAMETER ;
}
2007-11-28 01:35:30 +03:00
if ( ! ( nt_username = unistr2_to_ascii_talloc ( mem_ctx , & ( info3 - > uni_user_name ) ) ) ) {
2006-02-04 01:19:41 +03:00
/* If the server didn't give us one, just use the one we sent
* them */
2002-07-15 14:35:28 +04:00
nt_username = sent_nt_username ;
}
2007-11-28 01:35:30 +03:00
if ( ! ( nt_domain = unistr2_to_ascii_talloc ( mem_ctx , & ( info3 - > uni_logon_dom ) ) ) ) {
2006-02-04 01:19:41 +03:00
/* If the server didn't give us one, just use the one we sent
* them */
2003-12-31 03:31:43 +03:00
nt_domain = domain ;
2002-07-15 14:35:28 +04:00
}
2003-07-07 09:11:10 +04:00
2003-07-11 21:50:59 +04:00
/* try to fill the SAM account.. If getpwnam() fails, then try the
2004-10-23 00:15:24 +04:00
add user script ( 2.2 . x behavior ) .
We use the _unmapped_ username here in an attempt to provide
consistent username mapping behavior between kerberos and NTLM [ SSP ]
2006-02-04 01:19:41 +03:00
authentication in domain mode security . I . E . Username mapping
should be applied to the fully qualified username
( e . g . DOMAIN \ user ) and not just the login name . Yes this means we
called map_username ( ) unnecessarily in make_user_info_map ( ) but
that is how the current code is designed . Making the change here
is the least disruptive place . - - jerry */
2003-07-07 09:11:10 +04:00
2006-02-21 17:34:11 +03:00
if ( ! ( sam_account = samu_new ( NULL ) ) ) {
return NT_STATUS_NO_MEMORY ;
}
2006-05-02 16:13:23 +04:00
/* this call will try to create the user if necessary */
2004-10-23 00:15:24 +04:00
nt_status = fill_sam_account ( mem_ctx , nt_domain , sent_nt_username ,
2006-05-06 23:24:35 +04:00
& found_username , & uid , & gid , sam_account ,
& username_was_mapped ) ;
2003-07-07 09:11:10 +04:00
2002-07-15 14:35:28 +04:00
2005-06-09 22:45:56 +04:00
/* if we still don't have a valid unix account check for
2006-02-27 05:44:41 +03:00
' map to guest = bad uid ' */
2005-06-09 22:45:56 +04:00
2002-07-15 14:35:28 +04:00
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
2006-02-21 17:34:11 +03:00
TALLOC_FREE ( sam_account ) ;
2005-06-09 22:45:56 +04:00
if ( lp_map_to_guest ( ) = = MAP_TO_GUEST_ON_BAD_UID ) {
make_server_info_guest ( server_info ) ;
return NT_STATUS_OK ;
}
2002-07-15 14:35:28 +04:00
return nt_status ;
}
2003-07-03 18:36:42 +04:00
if ( ! pdb_set_nt_username ( sam_account , nt_username , PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2003-07-03 18:36:42 +04:00
return NT_STATUS_NO_MEMORY ;
2002-07-15 14:35:28 +04:00
}
2003-07-03 18:36:42 +04:00
if ( ! pdb_set_username ( sam_account , nt_username , PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2002-11-02 06:47:48 +03:00
if ( ! pdb_set_domain ( sam_account , nt_domain , PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2003-07-03 18:36:42 +04:00
if ( ! pdb_set_user_sid ( sam_account , & user_sid , PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2003-07-03 18:36:42 +04:00
return NT_STATUS_UNSUCCESSFUL ;
}
if ( ! pdb_set_group_sid ( sam_account , & group_sid , PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2003-07-03 18:36:42 +04:00
return NT_STATUS_UNSUCCESSFUL ;
}
2006-02-04 01:19:41 +03:00
if ( ! pdb_set_fullname ( sam_account ,
unistr2_static ( & ( info3 - > uni_full_name ) ) ,
2003-07-03 18:36:42 +04:00
PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
if ( ! pdb_set_logon_script ( sam_account ,
unistr2_static ( & ( info3 - > uni_logon_script ) ) ,
PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
if ( ! pdb_set_profile_path ( sam_account ,
unistr2_static ( & ( info3 - > uni_profile_path ) ) ,
PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
if ( ! pdb_set_homedir ( sam_account ,
unistr2_static ( & ( info3 - > uni_home_dir ) ) ,
PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
if ( ! pdb_set_dir_drive ( sam_account ,
unistr2_static ( & ( info3 - > uni_dir_drive ) ) ,
PDB_CHANGED ) ) {
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2002-07-15 14:35:28 +04:00
return NT_STATUS_NO_MEMORY ;
2006-02-27 13:32:45 +03:00
}
if ( ! pdb_set_acct_ctrl ( sam_account , info3 - > acct_flags , PDB_CHANGED ) ) {
TALLOC_FREE ( sam_account ) ;
return NT_STATUS_NO_MEMORY ;
2002-07-15 14:35:28 +04:00
}
2007-05-12 23:53:47 +04:00
if ( ! pdb_set_pass_last_set_time (
sam_account ,
nt_time_to_unix ( info3 - > pass_last_set_time ) ,
PDB_CHANGED ) ) {
TALLOC_FREE ( sam_account ) ;
return NT_STATUS_NO_MEMORY ;
}
if ( ! pdb_set_pass_can_change_time (
sam_account ,
nt_time_to_unix ( info3 - > pass_can_change_time ) ,
PDB_CHANGED ) ) {
TALLOC_FREE ( sam_account ) ;
return NT_STATUS_NO_MEMORY ;
}
if ( ! pdb_set_pass_must_change_time (
sam_account ,
nt_time_to_unix ( info3 - > pass_must_change_time ) ,
PDB_CHANGED ) ) {
TALLOC_FREE ( sam_account ) ;
return NT_STATUS_NO_MEMORY ;
}
2006-02-04 01:19:41 +03:00
result = make_server_info ( NULL ) ;
if ( result = = NULL ) {
2002-09-25 19:19:00 +04:00
DEBUG ( 4 , ( " make_server_info failed! \n " ) ) ;
2006-02-20 23:09:36 +03:00
TALLOC_FREE ( sam_account ) ;
2006-02-04 01:19:41 +03:00
return NT_STATUS_NO_MEMORY ;
2002-07-15 14:35:28 +04:00
}
Here's the code to make winbindd work on a Samba DC
to handle domain trusts. Jeremy and I talked about this
and it's going in as working code. It keeps winbind clean
and solves the trust problem with minimal changes.
To summarize, there are 2 basic cases where the deadlock would
occur. (1) lookuping up secondary groups for a user, and
(2) get[gr|pw]nam() calls that fall through the NSS layer because
they don't exist anywhere.
o To handle case #1, we bypass winbindd in sys_getgrouplist() unless
the username includes the 'winbind separator'.
o Case #2 is handled by adding checks in winbindd to return failure
if we are a DC and the domain matches our own.
This code has been tested using basic share connections, domain
logons, and with pam_winbind (both with and without 'winbind
use default domain'). The 'trustdomain' auth module should work
as well if an admin wants to manually create UNIX users for
acounts in the trusted domains.
Other misc fixes:
* we need to fix check_ntlm_password() to be able to determine
if an auth module is authoritative over a user (NT_STATUS_WRONG_PASSWORD,
etc...). I worked around my specific situation, but this needs to be
fixed. the winbindd auth module was causing delays.
* fix named server mutex deadlock between trust domain auth module
and winbindd looking up a uid
* make sure SAM_ACCOUNT gets stored in the server_info struct for the
_net_sam_logon() reply.
Configuration details:
The recommended method for supporting trusts is to use winbind.
The gets us around some of the server mutex issues as well.
* set 'files winbind' for passwd: and group: in /etc/nsswitch.conf
* create domain trusts like normal
* join winbind on the pdc to the Samba domain using 'net rpc join'
* add normal parameters to smb.conf for winbind
* set 'auth method = guest sam winbind'
* start smbd, nmbd, & winbindd
Problems that remain:
* join a Windows 2k/XP box to a Samba domain.
* create a 2-way trust between the Samba domain
and an NT domain
* logon to the windows client as a user from theh trusted
domain
* try to browse server in the trusted domain (or other
workstations). an NT client seems to work ok, but 2k
and XP either prompt for passwords or fail with errors.
apparanently this never got tested since no one has ever been
able to logon as a trusted user to a Samba domain from a Windows
client.
(This used to be commit f804b590f9dbf1f0147c06a0a2f12e221ae6fc3b)
2003-06-29 07:39:50 +04:00
/* save this here to _net_sam_logon() doesn't fail (it assumes a
2006-02-20 23:09:36 +03:00
valid struct samu ) */
Here's the code to make winbindd work on a Samba DC
to handle domain trusts. Jeremy and I talked about this
and it's going in as working code. It keeps winbind clean
and solves the trust problem with minimal changes.
To summarize, there are 2 basic cases where the deadlock would
occur. (1) lookuping up secondary groups for a user, and
(2) get[gr|pw]nam() calls that fall through the NSS layer because
they don't exist anywhere.
o To handle case #1, we bypass winbindd in sys_getgrouplist() unless
the username includes the 'winbind separator'.
o Case #2 is handled by adding checks in winbindd to return failure
if we are a DC and the domain matches our own.
This code has been tested using basic share connections, domain
logons, and with pam_winbind (both with and without 'winbind
use default domain'). The 'trustdomain' auth module should work
as well if an admin wants to manually create UNIX users for
acounts in the trusted domains.
Other misc fixes:
* we need to fix check_ntlm_password() to be able to determine
if an auth module is authoritative over a user (NT_STATUS_WRONG_PASSWORD,
etc...). I worked around my specific situation, but this needs to be
fixed. the winbindd auth module was causing delays.
* fix named server mutex deadlock between trust domain auth module
and winbindd looking up a uid
* make sure SAM_ACCOUNT gets stored in the server_info struct for the
_net_sam_logon() reply.
Configuration details:
The recommended method for supporting trusts is to use winbind.
The gets us around some of the server mutex issues as well.
* set 'files winbind' for passwd: and group: in /etc/nsswitch.conf
* create domain trusts like normal
* join winbind on the pdc to the Samba domain using 'net rpc join'
* add normal parameters to smb.conf for winbind
* set 'auth method = guest sam winbind'
* start smbd, nmbd, & winbindd
Problems that remain:
* join a Windows 2k/XP box to a Samba domain.
* create a 2-way trust between the Samba domain
and an NT domain
* logon to the windows client as a user from theh trusted
domain
* try to browse server in the trusted domain (or other
workstations). an NT client seems to work ok, but 2k
and XP either prompt for passwords or fail with errors.
apparanently this never got tested since no one has ever been
able to logon as a trusted user to a Samba domain from a Windows
client.
(This used to be commit f804b590f9dbf1f0147c06a0a2f12e221ae6fc3b)
2003-06-29 07:39:50 +04:00
2006-02-04 01:19:41 +03:00
result - > sam_account = sam_account ;
result - > unix_name = talloc_strdup ( result , found_username ) ;
2003-07-03 18:36:42 +04:00
2003-06-12 18:24:15 +04:00
/* Fill in the unix info we found on the way */
2006-02-04 01:19:41 +03:00
result - > uid = uid ;
result - > gid = gid ;
2002-09-25 19:19:00 +04:00
/* Create a 'combined' list of all SIDs we might want in the SD */
2006-02-04 01:19:41 +03:00
result - > num_sids = 0 ;
result - > sids = NULL ;
2002-09-25 19:19:00 +04:00
2007-07-17 15:47:17 +04:00
nt_status = sid_array_from_info3 ( result , info3 ,
2006-12-09 05:58:18 +03:00
& result - > sids ,
2007-07-17 15:47:17 +04:00
& result - > num_sids ,
False ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
TALLOC_FREE ( result ) ;
return nt_status ;
2002-09-25 19:19:00 +04:00
}
2007-11-28 01:35:30 +03:00
result - > login_server = unistr2_to_ascii_talloc ( result ,
2006-02-04 01:19:41 +03:00
& ( info3 - > uni_logon_srv ) ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* ensure we are never given NULL session keys */
2002-09-25 19:19:00 +04:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
if ( memcmp ( info3 - > user_sess_key , zeros , sizeof ( zeros ) ) = = 0 ) {
2007-05-14 16:16:20 +04:00
result - > user_session_key = data_blob_null ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
} else {
2006-02-04 01:19:41 +03:00
result - > user_session_key = data_blob_talloc (
result , info3 - > user_sess_key ,
sizeof ( info3 - > user_sess_key ) ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
}
2003-03-24 12:54:13 +03:00
2004-12-20 14:36:39 +03:00
if ( memcmp ( info3 - > lm_sess_key , zeros , 8 ) = = 0 ) {
2007-05-14 16:16:20 +04:00
result - > lm_session_key = data_blob_null ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
} else {
2006-02-04 01:19:41 +03:00
result - > lm_session_key = data_blob_talloc (
result , info3 - > lm_sess_key ,
sizeof ( info3 - > lm_sess_key ) ) ;
}
2006-05-06 23:24:35 +04:00
result - > was_mapped = username_was_mapped ;
2006-02-04 01:19:41 +03:00
* server_info = result ;
2004-12-20 14:36:39 +03:00
2002-07-15 14:35:28 +04:00
return NT_STATUS_OK ;
}
2002-09-25 19:19:00 +04:00
/***************************************************************************
Free a user_info struct
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
void free_user_info ( auth_usersupplied_info * * user_info )
{
DEBUG ( 5 , ( " attempting to free (and zero) a user_info structure \n " ) ) ;
if ( * user_info ! = NULL ) {
2006-02-04 01:19:41 +03:00
if ( ( * user_info ) - > smb_name ) {
DEBUG ( 10 , ( " structure was created for %s \n " ,
( * user_info ) - > smb_name ) ) ;
2002-09-25 19:19:00 +04:00
}
2006-02-04 01:19:41 +03:00
SAFE_FREE ( ( * user_info ) - > smb_name ) ;
SAFE_FREE ( ( * user_info ) - > internal_username ) ;
SAFE_FREE ( ( * user_info ) - > client_domain ) ;
SAFE_FREE ( ( * user_info ) - > domain ) ;
SAFE_FREE ( ( * user_info ) - > wksta_name ) ;
2002-09-25 19:19:00 +04:00
data_blob_free ( & ( * user_info ) - > lm_resp ) ;
data_blob_free ( & ( * user_info ) - > nt_resp ) ;
2004-04-03 19:41:32 +04:00
data_blob_clear_free ( & ( * user_info ) - > lm_interactive_pwd ) ;
data_blob_clear_free ( & ( * user_info ) - > nt_interactive_pwd ) ;
2002-09-25 19:19:00 +04:00
data_blob_clear_free ( & ( * user_info ) - > plaintext_password ) ;
ZERO_STRUCT ( * * user_info ) ;
}
SAFE_FREE ( * user_info ) ;
}
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
/***************************************************************************
Make an auth_methods struct
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2007-10-19 04:40:25 +04:00
bool make_auth_methods ( struct auth_context * auth_context , auth_methods * * auth_method )
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
{
2002-01-05 07:55:41 +03:00
if ( ! auth_context ) {
2006-02-04 01:19:41 +03:00
smb_panic ( " no auth_context supplied to "
" make_auth_methods()! \n " ) ;
2002-01-05 07:55:41 +03:00
}
if ( ! auth_method ) {
2006-02-04 01:19:41 +03:00
smb_panic ( " make_auth_methods: pointer to auth_method pointer "
" is NULL! \n " ) ;
2002-01-05 07:55:41 +03:00
}
2004-12-07 21:25:53 +03:00
* auth_method = TALLOC_P ( auth_context - > mem_ctx , auth_methods ) ;
This is another rather major change to the samba authenticaion
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a9757739711dbb82035a4dfe6b40c99)
2001-11-24 15:12:38 +03:00
if ( ! * auth_method ) {
DEBUG ( 0 , ( " make_auth_method: malloc failed! \n " ) ) ;
return False ;
}
ZERO_STRUCTP ( * auth_method ) ;
return True ;
}
2003-07-01 21:51:52 +04:00
/**
* Verify whether or not given domain is trusted .
*
* @ param domain_name name of the domain to be verified
* @ return true if domain is one of the trusted once or
* false if otherwise
* */
2007-10-19 04:40:25 +04:00
bool is_trusted_domain ( const char * dom_name )
2003-07-01 21:51:52 +04:00
{
DOM_SID trustdom_sid ;
2007-10-19 04:40:25 +04:00
bool ret ;
2003-07-01 21:51:52 +04:00
2003-07-08 21:19:37 +04:00
/* no trusted domains for a standalone server */
if ( lp_server_role ( ) = = ROLE_STANDALONE )
return False ;
2003-07-01 21:51:52 +04:00
/* if we are a DC, then check for a direct trust relationships */
2004-06-03 22:00:22 +04:00
if ( IS_DC ) {
2003-07-01 21:51:52 +04:00
become_root ( ) ;
2006-02-04 01:19:41 +03:00
DEBUG ( 5 , ( " is_trusted_domain: Checking for domain trust with "
" [%s] \n " , dom_name ) ) ;
2007-01-16 11:17:26 +03:00
ret = pdb_get_trusteddom_pw ( dom_name , NULL , NULL , NULL ) ;
2003-07-01 21:51:52 +04:00
unbecome_root ( ) ;
if ( ret )
return True ;
}
else {
2005-06-03 19:42:03 +04:00
NSS_STATUS result ;
2003-07-01 21:51:52 +04:00
2005-06-03 19:42:03 +04:00
/* If winbind is around, ask it */
result = wb_is_trusted_domain ( dom_name ) ;
if ( result = = NSS_STATUS_SUCCESS ) {
return True ;
}
if ( result = = NSS_STATUS_NOTFOUND ) {
/* winbind could not find the domain */
return False ;
}
/* The only other possible result is that winbind is not up
and running . We need to update the trustdom_cache
ourselves */
update_trustdom_cache ( ) ;
2003-07-01 21:51:52 +04:00
}
/* now the trustdom cache should be available a DC could still
* have a transitive trust so fall back to the cache of trusted
* domains ( like a domain member would use */
if ( trustdom_cache_fetch ( dom_name , & trustdom_sid ) ) {
return True ;
}
return False ;
}
2002-01-05 07:55:41 +03:00