2004-06-20 00:58:09 +00:00
/*
Unix SMB / CIFS implementation .
dcerpc authentication operations
Copyright ( C ) Andrew Tridgell 2003
Copyright ( C ) Andrew Bartlett < abartlet @ samba . org > 2004
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
the Free Software Foundation ; either version 2 of the License , or
( 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
along with this program ; if not , write to the Free Software
Foundation , Inc . , 675 Mass Ave , Cambridge , MA 0213 9 , USA .
*/
# include "includes.h"
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state {
struct auth_context * auth_context ;
struct auth_serversupplied_info * server_info ;
struct ntlmssp_state * ntlmssp_state ;
} ;
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
/**
* Return the challenge as determined by the authentication subsystem
* @ return an 8 byte random challenge
*/
static const uint8_t * auth_ntlmssp_get_challenge ( const struct ntlmssp_state * ntlmssp_state )
{
struct gensec_ntlmssp_state * gensec_ntlmssp_state = ntlmssp_state - > auth_context ;
return gensec_ntlmssp_state - > auth_context - > get_ntlm_challenge ( gensec_ntlmssp_state - > auth_context ) ;
}
/**
* Some authentication methods ' fix ' the challenge , so we may not be able to set it
*
* @ return If the effective challenge used by the auth subsystem may be modified
*/
static BOOL auth_ntlmssp_may_set_challenge ( const struct ntlmssp_state * ntlmssp_state )
{
struct gensec_ntlmssp_state * gensec_ntlmssp_state = ntlmssp_state - > auth_context ;
return gensec_ntlmssp_state - > auth_context - > challenge_may_be_modified ;
}
/**
* NTLM2 authentication modifies the effective challenge ,
* @ param challenge The new challenge value
*/
static NTSTATUS auth_ntlmssp_set_challenge ( struct ntlmssp_state * ntlmssp_state , DATA_BLOB * challenge )
{
struct gensec_ntlmssp_state * gensec_ntlmssp_state = ntlmssp_state - > auth_context ;
struct auth_context * auth_context = gensec_ntlmssp_state - > auth_context ;
SMB_ASSERT ( challenge - > length = = 8 ) ;
2004-09-26 03:50:24 +00:00
auth_context - > challenge = data_blob_talloc ( auth_context ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
challenge - > data , challenge - > length ) ;
auth_context - > challenge_set_by = " NTLMSSP callback (NTLM2) " ;
DEBUG ( 5 , ( " auth_context challenge set by %s \n " , auth_context - > challenge_set_by ) ) ;
DEBUG ( 5 , ( " challenge is: \n " ) ) ;
dump_data ( 5 , ( const char * ) auth_context - > challenge . data , auth_context - > challenge . length ) ;
return NT_STATUS_OK ;
}
/**
* Check the password on an NTLMSSP login .
*
* Return the session keys used on the connection .
*/
static NTSTATUS auth_ntlmssp_check_password ( struct ntlmssp_state * ntlmssp_state , DATA_BLOB * user_session_key , DATA_BLOB * lm_session_key )
{
struct gensec_ntlmssp_state * gensec_ntlmssp_state = ntlmssp_state - > auth_context ;
struct auth_usersupplied_info * user_info = NULL ;
NTSTATUS nt_status ;
#if 0
/* the client has given us its machine name (which we otherwise would not get on port 445).
we need to possibly reload smb . conf if smb . conf includes depend on the machine name */
set_remote_machine_name ( gensec_ntlmssp_state - > ntlmssp_state - > workstation , True ) ;
/* setup the string used by %U */
/* sub_set_smb_name checks for weird internally */
sub_set_smb_name ( gensec_ntlmssp_state - > ntlmssp_state - > user ) ;
reload_services ( True ) ;
# endif
2004-09-26 23:39:18 +00:00
nt_status = make_user_info_map ( ntlmssp_state ,
& user_info ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state - > ntlmssp_state - > user ,
gensec_ntlmssp_state - > ntlmssp_state - > domain ,
gensec_ntlmssp_state - > ntlmssp_state - > workstation ,
gensec_ntlmssp_state - > ntlmssp_state - > lm_resp . data ? & gensec_ntlmssp_state - > ntlmssp_state - > lm_resp : NULL ,
gensec_ntlmssp_state - > ntlmssp_state - > nt_resp . data ? & gensec_ntlmssp_state - > ntlmssp_state - > nt_resp : NULL ,
NULL , NULL , NULL ,
True ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
return nt_status ;
}
nt_status = gensec_ntlmssp_state - > auth_context - > check_ntlm_password ( gensec_ntlmssp_state - > auth_context ,
user_info , & gensec_ntlmssp_state - > server_info ) ;
free_user_info ( & user_info ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
return nt_status ;
}
if ( gensec_ntlmssp_state - > server_info - > user_session_key . length ) {
DEBUG ( 10 , ( " Got NT session key of length %u \n " , gensec_ntlmssp_state - > server_info - > user_session_key . length ) ) ;
2004-09-26 02:16:25 +00:00
* user_session_key = data_blob_talloc ( ntlmssp_state ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state - > server_info - > user_session_key . data ,
gensec_ntlmssp_state - > server_info - > user_session_key . length ) ;
}
if ( gensec_ntlmssp_state - > server_info - > lm_session_key . length ) {
DEBUG ( 10 , ( " Got LM session key of length %u \n " , gensec_ntlmssp_state - > server_info - > lm_session_key . length ) ) ;
2004-09-26 02:16:25 +00:00
* lm_session_key = data_blob_talloc ( ntlmssp_state ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state - > server_info - > lm_session_key . data ,
gensec_ntlmssp_state - > server_info - > lm_session_key . length ) ;
}
return nt_status ;
}
static NTSTATUS gensec_ntlmssp_start ( struct gensec_security * gensec_security )
{
struct gensec_ntlmssp_state * gensec_ntlmssp_state ;
2004-09-26 02:16:25 +00:00
gensec_ntlmssp_state = talloc_p ( gensec_security , struct gensec_ntlmssp_state ) ;
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
if ( ! gensec_ntlmssp_state ) {
return NT_STATUS_NO_MEMORY ;
}
gensec_ntlmssp_state - > ntlmssp_state = NULL ;
gensec_ntlmssp_state - > auth_context = NULL ;
gensec_ntlmssp_state - > server_info = NULL ;
gensec_security - > private_data = gensec_ntlmssp_state ;
return NT_STATUS_OK ;
}
static NTSTATUS gensec_ntlmssp_server_start ( struct gensec_security * gensec_security )
{
NTSTATUS nt_status ;
NTSTATUS status ;
struct ntlmssp_state * ntlmssp_state ;
struct gensec_ntlmssp_state * gensec_ntlmssp_state ;
status = gensec_ntlmssp_start ( gensec_security ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
gensec_ntlmssp_state = gensec_security - > private_data ;
2004-09-26 02:16:25 +00:00
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_server_start ( gensec_security ,
& gensec_ntlmssp_state - > ntlmssp_state ) ) ) {
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
return nt_status ;
}
2004-08-25 02:25:20 +00:00
if ( gensec_security - > want_features & GENSEC_WANT_SIGN ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_SIGN ;
}
if ( gensec_security - > want_features & GENSEC_WANT_SEAL ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_SEAL ;
}
2004-08-29 11:28:31 +00:00
/* IF we are not doing Signing or Sealing, we can actually do
* NTLM2 . When we crack the crypto puzzle , then we can enable
* this always , in the constant flags */
if ( ! ( gensec_security - > want_features & GENSEC_WANT_SIGN ) & & ! ( gensec_security - > want_features & GENSEC_WANT_SEAL ) ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_NTLM2 ;
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
ntlmssp_state = gensec_ntlmssp_state - > ntlmssp_state ;
2004-09-26 03:50:24 +00:00
nt_status = make_auth_context_subsystem ( gensec_security , & gensec_ntlmssp_state - > auth_context ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
return nt_status ;
}
ntlmssp_state - > auth_context = gensec_ntlmssp_state ;
ntlmssp_state - > get_challenge = auth_ntlmssp_get_challenge ;
ntlmssp_state - > may_set_challenge = auth_ntlmssp_may_set_challenge ;
ntlmssp_state - > set_challenge = auth_ntlmssp_set_challenge ;
ntlmssp_state - > check_password = auth_ntlmssp_check_password ;
ntlmssp_state - > server_role = lp_server_role ( ) ;
return NT_STATUS_OK ;
}
static NTSTATUS gensec_ntlmssp_client_start ( struct gensec_security * gensec_security )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state ;
char * password = NULL ;
2004-06-20 00:58:09 +00:00
NTSTATUS status ;
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
status = gensec_ntlmssp_start ( gensec_security ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state = gensec_security - > private_data ;
2004-09-26 02:16:25 +00:00
status = ntlmssp_client_start ( gensec_security ,
& gensec_ntlmssp_state - > ntlmssp_state ) ;
2004-06-20 00:58:09 +00:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
2004-08-25 02:25:20 +00:00
if ( gensec_security - > want_features & GENSEC_WANT_SESSION_KEY ) {
/*
* We need to set this to allow a later SetPassword
* via the SAMR pipe to succeed . Strange . . . . We could
* also add NTLMSSP_NEGOTIATE_SEAL here . JRA .
*
* Without this , Windows will not create the master key
* that it thinks is only used for NTLMSSP signing and
* sealing . ( It is actually pulled out and used directly )
*/
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_SIGN ;
}
if ( gensec_security - > want_features & GENSEC_WANT_SIGN ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_SIGN ;
}
if ( gensec_security - > want_features & GENSEC_WANT_SEAL ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_SEAL ;
}
2004-08-29 11:28:31 +00:00
/* IF we are not doing Signing or Sealing, we can actually do
* NTLM2 . When we crack the crypto puzzle , then we can enable
* this always , in the constant flags */
if ( ! ( gensec_security - > want_features & GENSEC_WANT_SIGN ) & & ! ( gensec_security - > want_features & GENSEC_WANT_SEAL ) ) {
gensec_ntlmssp_state - > ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_NTLM2 ;
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
status = ntlmssp_set_domain ( gensec_ntlmssp_state - > ntlmssp_state ,
gensec_security - > user . domain ) ;
2004-06-20 00:58:09 +00:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
status = ntlmssp_set_username ( gensec_ntlmssp_state - > ntlmssp_state ,
gensec_security - > user . name ) ;
2004-06-20 00:58:09 +00:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
2004-09-26 02:16:25 +00:00
status = gensec_get_password ( gensec_security , gensec_ntlmssp_state , & password ) ;
2004-06-20 00:58:09 +00:00
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
if ( password ) {
status = ntlmssp_set_password ( gensec_ntlmssp_state - > ntlmssp_state ,
password ) ;
if ( ! NT_STATUS_IS_OK ( status ) ) {
return status ;
}
}
gensec_security - > private_data = gensec_ntlmssp_state ;
2004-06-20 00:58:09 +00:00
return status ;
}
/*
wrappers for the ntlmssp_ * ( ) functions
*/
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static NTSTATUS gensec_ntlmssp_unseal_packet ( struct gensec_security * gensec_security ,
2004-09-11 15:11:36 +00:00
TALLOC_CTX * mem_ctx ,
uint8_t * data , size_t length ,
const uint8_t * whole_pdu , size_t pdu_length ,
DATA_BLOB * sig )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
2004-09-11 15:11:36 +00:00
return ntlmssp_unseal_packet ( gensec_ntlmssp_state - > ntlmssp_state , mem_ctx , data , length , whole_pdu , pdu_length , sig ) ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static NTSTATUS gensec_ntlmssp_check_packet ( struct gensec_security * gensec_security ,
2004-09-11 15:11:36 +00:00
TALLOC_CTX * mem_ctx ,
const uint8_t * data , size_t length ,
const uint8_t * whole_pdu , size_t pdu_length ,
const DATA_BLOB * sig )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
2004-09-11 15:11:36 +00:00
return ntlmssp_check_packet ( gensec_ntlmssp_state - > ntlmssp_state , mem_ctx , data , length , whole_pdu , pdu_length , sig ) ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static NTSTATUS gensec_ntlmssp_seal_packet ( struct gensec_security * gensec_security ,
2004-09-11 15:11:36 +00:00
TALLOC_CTX * mem_ctx ,
uint8_t * data , size_t length ,
const uint8_t * whole_pdu , size_t pdu_length ,
DATA_BLOB * sig )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
2004-09-11 15:11:36 +00:00
return ntlmssp_seal_packet ( gensec_ntlmssp_state - > ntlmssp_state , mem_ctx , data , length , whole_pdu , pdu_length , sig ) ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static NTSTATUS gensec_ntlmssp_sign_packet ( struct gensec_security * gensec_security ,
2004-09-11 15:11:36 +00:00
TALLOC_CTX * mem_ctx ,
const uint8_t * data , size_t length ,
const uint8_t * whole_pdu , size_t pdu_length ,
DATA_BLOB * sig )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
2004-09-11 15:11:36 +00:00
return ntlmssp_sign_packet ( gensec_ntlmssp_state - > ntlmssp_state , mem_ctx , data , length , whole_pdu , pdu_length , sig ) ;
}
static size_t gensec_ntlmssp_sig_size ( struct gensec_security * gensec_security )
{
return NTLMSSP_SIG_SIZE ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static NTSTATUS gensec_ntlmssp_session_key ( struct gensec_security * gensec_security ,
2004-07-11 10:38:31 +00:00
DATA_BLOB * session_key )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
return ntlmssp_session_key ( gensec_ntlmssp_state - > ntlmssp_state , session_key ) ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
/**
* Next state function for the wrapped NTLMSSP state machine
*
2004-07-11 10:38:31 +00:00
* @ param gensec_security GENSEC state , initialised to NTLMSSP
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
* @ param out_mem_ctx The TALLOC_CTX for * out to be allocated on
* @ param in The request , as a DATA_BLOB
* @ param out The reply , as an talloc ( ) ed DATA_BLOB , on * out_mem_ctx
* @ return Error , MORE_PROCESSING_REQUIRED if a reply is sent ,
* or NT_STATUS_OK if the user is authenticated .
*/
static NTSTATUS gensec_ntlmssp_update ( struct gensec_security * gensec_security , TALLOC_CTX * out_mem_ctx ,
2004-07-11 10:38:31 +00:00
const DATA_BLOB in , DATA_BLOB * out )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
return ntlmssp_update ( gensec_ntlmssp_state - > ntlmssp_state , out_mem_ctx , in , out ) ;
}
/**
* Return the credentials of a logged on user , including session keys
* etc .
*
* Only valid after a successful authentication
*
* May only be called once per authentication .
*
*/
static NTSTATUS gensec_ntlmssp_session_info ( struct gensec_security * gensec_security ,
2004-10-25 04:25:29 +00:00
struct auth_session_info * * session_info )
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
{
NTSTATUS nt_status ;
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-10-25 04:25:29 +00:00
nt_status = make_session_info ( gensec_ntlmssp_state , gensec_ntlmssp_state - > server_info , session_info ) ;
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
return nt_status ;
}
2004-09-26 01:43:05 +00:00
( * session_info ) - > session_key = data_blob_talloc ( * session_info ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state - > ntlmssp_state - > session_key . data ,
gensec_ntlmssp_state - > ntlmssp_state - > session_key . length ) ;
2004-09-26 01:43:05 +00:00
( * session_info ) - > workstation = talloc_strdup ( * session_info ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
gensec_ntlmssp_state - > ntlmssp_state - > workstation ) ;
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
return NT_STATUS_OK ;
2004-06-20 00:58:09 +00:00
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static void gensec_ntlmssp_end ( struct gensec_security * gensec_security )
2004-06-20 00:58:09 +00:00
{
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
struct gensec_ntlmssp_state * gensec_ntlmssp_state = gensec_security - > private_data ;
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
if ( gensec_ntlmssp_state - > ntlmssp_state ) {
ntlmssp_end ( & gensec_ntlmssp_state - > ntlmssp_state ) ;
}
2004-06-20 00:58:09 +00:00
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
if ( gensec_ntlmssp_state - > auth_context ) {
free_auth_context ( & gensec_ntlmssp_state - > auth_context ) ;
}
if ( gensec_ntlmssp_state - > server_info ) {
free_server_info ( & gensec_ntlmssp_state - > server_info ) ;
}
2004-09-26 02:16:25 +00:00
talloc_free ( gensec_ntlmssp_state ) ;
2004-06-20 00:58:09 +00:00
gensec_security - > private_data = NULL ;
}
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
static const struct gensec_security_ops gensec_ntlmssp_security_ops = {
. name = " ntlmssp " ,
. sasl_name = " NTLM " ,
. auth_type = DCERPC_AUTH_TYPE_NTLMSSP ,
. oid = OID_NTLMSSP ,
. client_start = gensec_ntlmssp_client_start ,
. server_start = gensec_ntlmssp_server_start ,
. update = gensec_ntlmssp_update ,
. seal_packet = gensec_ntlmssp_seal_packet ,
2004-09-11 15:11:36 +00:00
. sig_size = gensec_ntlmssp_sig_size ,
r1294: A nice, large, commit...
This implements gensec for Samba's server side, and brings gensec up
to the standards of a full subsystem.
This means that use of the subsystem is by gensec_* functions, not
function pointers in structures (this is internal). This causes
changes in all the existing gensec users.
Our RPC server no longer contains it's own generalised security
scheme, and now calls gensec directly.
Gensec has also taken over the role of auth/auth_ntlmssp.c
An important part of gensec, is the output of the 'session_info'
struct. This is now reference counted, so that we can correctly free
it when a pipe is closed, no matter if it was inherited, or created by
per-pipe authentication.
The schannel code is reworked, to be in the same file for client and
server.
ntlm_auth is reworked to use gensec.
The major problem with this code is the way it relies on subsystem
auto-initialisation. The primary reason for this commit now.is to
allow these problems to be looked at, and fixed.
There are problems with the new code:
- I've tested it with smbtorture, but currently don't have VMware and
valgrind working (this I'll fix soon).
- The SPNEGO code is client-only at this point.
- We still do not do kerberos.
Andrew Bartlett
(This used to be commit 07fd885fd488fd1051eacc905a2d4962f8a018ec)
2004-06-29 09:40:10 +00:00
. sign_packet = gensec_ntlmssp_sign_packet ,
. check_packet = gensec_ntlmssp_check_packet ,
. unseal_packet = gensec_ntlmssp_unseal_packet ,
. session_key = gensec_ntlmssp_session_key ,
. session_info = gensec_ntlmssp_session_info ,
. end = gensec_ntlmssp_end
} ;
NTSTATUS gensec_ntlmssp_init ( void )
{
NTSTATUS ret ;
ret = register_backend ( " gensec " , & gensec_ntlmssp_security_ops ) ;
if ( ! NT_STATUS_IS_OK ( ret ) ) {
DEBUG ( 0 , ( " Failed to register '%s' gensec backend! \n " ,
gensec_ntlmssp_security_ops . name ) ) ;
return ret ;
}
/* ugly cludge, but we need the auth subsystem for this to work */
auth_init ( ) ;
return ret ;
}