2004-06-19 12:15:41 +04:00
/*
Unix SMB / CIFS implementation .
RFC2478 Compliant SPNEGO implementation
2004-06-20 04:58:09 +04:00
Copyright ( C ) Jim McDonough < jmcd @ us . ibm . com > 2003
Copyright ( C ) Andrew Bartlett < abartlet @ samba . org > 2004
2004-06-19 12:15:41 +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
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"
# undef DBGC_CLASS
# define DBGC_CLASS DBGC_AUTH
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 13:40:10 +04:00
enum spnego_state_position {
SPNEGO_SERVER_START ,
SPNEGO_CLIENT_GET_MECHS ,
SPNEGO_CLIENT_SEND_MECHS ,
SPNEGO_TARG ,
SPNEGO_FALLBACK ,
SPNEGO_DONE
} ;
struct spnego_state {
TALLOC_CTX * mem_ctx ;
uint_t ref_count ;
enum spnego_message_type expected_packet ;
enum spnego_message_type state_position ;
2004-07-06 04:09:10 +04:00
negResult_t result ;
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 13:40:10 +04:00
struct gensec_security * sub_sec_security ;
} ;
static NTSTATUS gensec_spnego_client_start ( struct gensec_security * gensec_security )
2004-06-19 12:15:41 +04:00
{
2004-06-20 04:58:09 +04:00
struct spnego_state * spnego_state ;
TALLOC_CTX * mem_ctx = talloc_init ( " gensec_spengo_client_start " ) ;
if ( ! mem_ctx ) {
return NT_STATUS_NO_MEMORY ;
}
spnego_state = talloc_p ( mem_ctx , struct spnego_state ) ;
if ( ! spnego_state ) {
return NT_STATUS_NO_MEMORY ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
spnego_state - > expected_packet = SPNEGO_NEG_TOKEN_INIT ;
spnego_state - > state_position = SPNEGO_CLIENT_GET_MECHS ;
spnego_state - > result = SPNEGO_ACCEPT_INCOMPLETE ;
spnego_state - > mem_ctx = mem_ctx ;
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 13:40:10 +04:00
spnego_state - > sub_sec_security = NULL ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
gensec_security - > private_data = spnego_state ;
return NT_STATUS_OK ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
/*
wrappers for the spnego_ * ( ) 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 13:40:10 +04:00
static NTSTATUS gensec_spnego_unseal_packet ( struct gensec_security * gensec_security ,
TALLOC_CTX * mem_ctx ,
uint8_t * data , size_t length , DATA_BLOB * sig )
2004-06-19 12:15:41 +04:00
{
2004-06-20 04:58:09 +04:00
struct spnego_state * spnego_state = gensec_security - > private_data ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
if ( spnego_state - > state_position ! = SPNEGO_DONE
& & spnego_state - > state_position ! = SPNEGO_FALLBACK ) {
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
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 13:40:10 +04:00
return gensec_unseal_packet ( spnego_state - > sub_sec_security ,
mem_ctx , data , length , sig ) ;
2004-06-20 04:58:09 +04:00
}
2004-06-19 12:15:41 +04:00
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 13:40:10 +04:00
static NTSTATUS gensec_spnego_check_packet ( struct gensec_security * gensec_security ,
2004-06-20 04:58:09 +04:00
TALLOC_CTX * mem_ctx ,
const uint8_t * data , size_t length ,
const DATA_BLOB * sig )
{
struct spnego_state * spnego_state = gensec_security - > private_data ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
return NT_STATUS_NOT_IMPLEMENTED ;
if ( spnego_state - > state_position ! = SPNEGO_DONE
& & spnego_state - > state_position ! = SPNEGO_FALLBACK ) {
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
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 13:40:10 +04:00
return gensec_check_packet ( spnego_state - > sub_sec_security ,
mem_ctx , data , length , sig ) ;
2004-06-20 04:58:09 +04:00
}
2004-06-19 12:15:41 +04:00
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 13:40:10 +04:00
static NTSTATUS gensec_spnego_seal_packet ( struct gensec_security * gensec_security ,
2004-06-20 04:58:09 +04:00
TALLOC_CTX * mem_ctx ,
uint8_t * data , size_t length ,
DATA_BLOB * sig )
{
struct spnego_state * spnego_state = gensec_security - > private_data ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
return NT_STATUS_NOT_IMPLEMENTED ;
if ( spnego_state - > state_position ! = SPNEGO_DONE
& & spnego_state - > state_position ! = SPNEGO_FALLBACK ) {
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
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 13:40:10 +04:00
return gensec_seal_packet ( spnego_state - > sub_sec_security ,
mem_ctx , data , length , sig ) ;
2004-06-20 04:58:09 +04:00
}
2004-06-19 12:15:41 +04:00
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 13:40:10 +04:00
static NTSTATUS gensec_spnego_sign_packet ( struct gensec_security * gensec_security ,
2004-06-20 04:58:09 +04:00
TALLOC_CTX * mem_ctx ,
const uint8_t * data , size_t length ,
DATA_BLOB * sig )
{
struct spnego_state * spnego_state = gensec_security - > private_data ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
if ( spnego_state - > state_position ! = SPNEGO_DONE
& & spnego_state - > state_position ! = SPNEGO_FALLBACK ) {
return NT_STATUS_INVALID_PARAMETER ;
}
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 13:40:10 +04:00
return gensec_sign_packet ( spnego_state - > sub_sec_security ,
mem_ctx , data , length , sig ) ;
2004-06-19 12:15:41 +04:00
}
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 13:40:10 +04:00
static NTSTATUS gensec_spnego_session_key ( struct gensec_security * gensec_security ,
2004-06-20 04:58:09 +04:00
DATA_BLOB * session_key )
2004-06-19 12:15:41 +04:00
{
2004-06-20 04:58:09 +04:00
struct spnego_state * spnego_state = gensec_security - > private_data ;
if ( spnego_state - > state_position ! = SPNEGO_DONE
& & spnego_state - > state_position ! = SPNEGO_FALLBACK ) {
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
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 13:40:10 +04:00
return gensec_session_key ( spnego_state - > sub_sec_security ,
session_key ) ;
2004-06-19 12:15:41 +04:00
}
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 13:40:10 +04:00
static NTSTATUS gensec_spnego_update ( struct gensec_security * gensec_security , TALLOC_CTX * out_mem_ctx ,
2004-06-20 04:58:09 +04:00
const DATA_BLOB in , DATA_BLOB * out )
2004-06-19 12:15:41 +04:00
{
2004-06-20 04:58:09 +04:00
struct spnego_state * spnego_state = gensec_security - > private_data ;
DATA_BLOB null_data_blob = data_blob ( NULL , 0 ) ;
DATA_BLOB unwrapped_out ;
struct spnego_data spnego_out ;
struct spnego_data spnego ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
ssize_t len ;
2004-06-19 12:15:41 +04:00
2004-06-20 04:58:09 +04:00
if ( ! out_mem_ctx ) {
out_mem_ctx = spnego_state - > mem_ctx ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
if ( spnego_state - > state_position = = SPNEGO_FALLBACK ) {
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 13:40:10 +04:00
return gensec_update ( spnego_state - > sub_sec_security ,
out_mem_ctx , in , out ) ;
2004-06-19 12:15:41 +04:00
}
2004-07-06 04:09:10 +04:00
len = spnego_read_data ( in , & spnego ) ;
if ( len = = - 1 & & spnego_state - > state_position = = SPNEGO_SERVER_START ) {
2004-06-20 04:58:09 +04:00
int i ;
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 13:40:10 +04:00
int num_ops ;
const struct gensec_security_ops * * all_ops = gensec_security_all ( & num_ops ) ;
for ( i = 0 ; i < num_ops ; i + + ) {
2004-06-20 04:58:09 +04:00
NTSTATUS 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 13:40:10 +04:00
if ( ! all_ops [ i ] - > oid ) {
2004-06-20 04:58:09 +04:00
continue ;
}
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 13:40:10 +04:00
nt_status = gensec_server_start ( & spnego_state - > sub_sec_security ) ;
2004-06-20 04:58:09 +04:00
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 13:40:10 +04:00
return nt_status ;
}
nt_status = gensec_start_mech_by_oid ( spnego_state - > sub_sec_security ,
all_ops [ i ] - > oid ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
gensec_end ( & spnego_state - > sub_sec_security ) ;
2004-06-20 04:58:09 +04:00
continue ;
}
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 13:40:10 +04:00
nt_status = gensec_update ( spnego_state - > sub_sec_security ,
out_mem_ctx , in , out ) ;
2004-06-20 04:58:09 +04:00
if ( NT_STATUS_EQUAL ( nt_status , NT_STATUS_MORE_PROCESSING_REQUIRED ) ) {
spnego_state - > state_position = SPNEGO_FALLBACK ;
return 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 13:40:10 +04:00
gensec_end ( & spnego_state - > sub_sec_security ) ;
2004-06-19 12:15:41 +04:00
}
2004-06-20 04:58:09 +04:00
DEBUG ( 1 , ( " Failed to parse SPENGO request \n " ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
2004-07-06 04:09:10 +04:00
} else {
2004-06-20 04:58:09 +04:00
2004-07-06 04:09:10 +04:00
if ( spnego . type ! = spnego_state - > expected_packet ) {
spnego_free_data ( & spnego ) ;
DEBUG ( 1 , ( " Invalid SPENGO request: %d, expected %d \n " , spnego . type ,
spnego_state - > expected_packet ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
}
2004-06-19 12:15:41 +04:00
2004-07-06 04:09:10 +04:00
if ( spnego_state - > state_position = = SPNEGO_CLIENT_GET_MECHS ) {
2004-06-19 12:15:41 +04:00
2004-07-06 04:09:10 +04:00
/* The server offers a list of mechanisms */
char * * mechType = spnego . negTokenInit . mechTypes ;
char * my_mechs [ ] = { NULL , NULL } ;
int i ;
NTSTATUS nt_status ;
for ( i = 0 ; mechType [ i ] ; i + + ) {
nt_status = gensec_client_start ( & spnego_state - > sub_sec_security ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
break ;
}
nt_status = gensec_start_mech_by_oid ( spnego_state - > sub_sec_security ,
mechType [ i ] ) ;
if ( ! NT_STATUS_IS_OK ( nt_status ) ) {
gensec_end ( & spnego_state - > sub_sec_security ) ;
continue ;
}
if ( i = = 0 ) {
nt_status = gensec_update ( spnego_state - > sub_sec_security ,
out_mem_ctx ,
spnego . negTokenInit . mechToken ,
& unwrapped_out ) ;
} else {
/* only get the helping start blob for the first OID */
nt_status = gensec_update ( spnego_state - > sub_sec_security ,
out_mem_ctx ,
null_data_blob ,
& unwrapped_out ) ;
}
if ( ! NT_STATUS_EQUAL ( nt_status , NT_STATUS_MORE_PROCESSING_REQUIRED ) ) {
DEBUG ( 1 , ( " SPENGO(%s) NEG_TOKEN_INIT failed: %s \n " ,
spnego_state - > sub_sec_security - > ops - > name , nt_errstr ( nt_status ) ) ) ;
gensec_end ( & spnego_state - > sub_sec_security ) ;
} else {
break ;
}
}
if ( ! mechType [ i ] ) {
DEBUG ( 1 , ( " SPENGO: Could not find a suitable mechtype in NEG_TOKEN_INIT \n " ) ) ;
}
2004-06-19 12:15:41 +04:00
2004-07-06 04:09:10 +04:00
spnego_free_data ( & spnego ) ;
if ( ! NT_STATUS_EQUAL ( nt_status , NT_STATUS_MORE_PROCESSING_REQUIRED ) ) {
return nt_status ;
}
/* compose reply */
my_mechs [ 0 ] = spnego_state - > sub_sec_security - > ops - > oid ;
2004-06-20 04:58:09 +04:00
2004-07-06 04:09:10 +04:00
spnego_out . type = SPNEGO_NEG_TOKEN_INIT ;
spnego_out . negTokenInit . mechTypes = my_mechs ;
spnego_out . negTokenInit . reqFlags = 0 ;
spnego_out . negTokenInit . mechListMIC = null_data_blob ;
spnego_out . negTokenInit . mechToken = unwrapped_out ;
if ( spnego_write_data ( out_mem_ctx , out , & spnego_out ) = = - 1 ) {
DEBUG ( 1 , ( " Failed to write SPENGO reply to NEG_TOKEN_INIT \n " ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
2004-06-20 04:58:09 +04:00
}
2004-07-06 04:09:10 +04:00
/* set next state */
spnego_state - > expected_packet = SPNEGO_NEG_TOKEN_TARG ;
spnego_state - > state_position = SPNEGO_TARG ;
return nt_status ;
} else if ( spnego_state - > state_position = = SPNEGO_TARG ) {
NTSTATUS nt_status ;
if ( spnego . negTokenTarg . negResult = = SPNEGO_REJECT ) {
return NT_STATUS_ACCESS_DENIED ;
2004-06-20 04:58:09 +04:00
}
2004-06-19 12:15:41 +04:00
2004-07-06 04:09:10 +04:00
if ( spnego . negTokenTarg . responseToken . length ) {
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 13:40:10 +04:00
nt_status = gensec_update ( spnego_state - > sub_sec_security ,
out_mem_ctx ,
2004-07-06 04:09:10 +04:00
spnego . negTokenTarg . responseToken ,
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 13:40:10 +04:00
& unwrapped_out ) ;
2004-06-20 04:58:09 +04:00
} else {
2004-07-06 04:09:10 +04:00
unwrapped_out = data_blob ( NULL , 0 ) ;
nt_status = NT_STATUS_OK ;
2004-06-20 04:58:09 +04:00
}
2004-07-06 04:09:10 +04:00
if ( NT_STATUS_IS_OK ( nt_status )
& & ( spnego . negTokenTarg . negResult ! = SPNEGO_ACCEPT_COMPLETED ) ) {
nt_status = NT_STATUS_INVALID_PARAMETER ;
2004-06-20 04:58:09 +04:00
}
2004-06-19 12:15:41 +04:00
2004-07-06 04:09:10 +04:00
spnego_state - > result = spnego . negTokenTarg . negResult ;
spnego_free_data ( & spnego ) ;
2004-06-20 04:58:09 +04:00
2004-07-06 04:09:10 +04:00
if ( NT_STATUS_EQUAL ( nt_status , NT_STATUS_MORE_PROCESSING_REQUIRED ) ) {
/* compose reply */
spnego_out . type = SPNEGO_NEG_TOKEN_TARG ;
spnego_out . negTokenTarg . negResult = SPNEGO_ACCEPT_INCOMPLETE ;
spnego_out . negTokenTarg . supportedMech
= spnego_state - > sub_sec_security - > ops - > oid ;
;
spnego_out . negTokenTarg . responseToken = unwrapped_out ;
spnego_out . negTokenTarg . mechListMIC = null_data_blob ;
2004-06-23 19:32:44 +04:00
if ( spnego_write_data ( out_mem_ctx , out , & spnego_out ) = = - 1 ) {
2004-06-20 04:58:09 +04:00
DEBUG ( 1 , ( " Failed to write SPENGO reply to NEG_TOKEN_TARG \n " ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
2004-07-06 04:09:10 +04:00
spnego_state - > state_position = SPNEGO_TARG ;
} else if ( NT_STATUS_IS_OK ( nt_status ) ) {
spnego_state - > state_position = SPNEGO_DONE ;
} else {
DEBUG ( 1 , ( " SPENGO(%s) login failed: %s \n " ,
spnego_state - > sub_sec_security - > ops - > name ,
nt_errstr ( nt_status ) ) ) ;
return nt_status ;
}
2004-07-06 03:26:07 +04:00
return nt_status ;
2004-07-06 04:09:10 +04:00
} else {
spnego_free_data ( & spnego ) ;
DEBUG ( 1 , ( " Invalid SPENGO request: %d \n " , spnego . type ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
2004-06-19 12:15:41 +04:00
}
}
}
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 13:40:10 +04:00
static void gensec_spnego_end ( struct gensec_security * gensec_security )
2004-06-20 04:58:09 +04:00
{
struct spnego_state * spnego_state = gensec_security - > private_data ;
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 13:40:10 +04:00
if ( spnego_state - > sub_sec_security ) {
gensec_end ( & spnego_state - > sub_sec_security ) ;
}
2004-06-20 04:58:09 +04:00
talloc_destroy ( spnego_state - > mem_ctx ) ;
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 13:40:10 +04:00
static const struct gensec_security_ops gensec_spnego_security_ops = {
. name = " spnego " ,
. sasl_name = " GSS-SPNEGO " ,
. oid = OID_SPNEGO ,
. client_start = gensec_spnego_client_start ,
. update = gensec_spnego_update ,
. seal_packet = gensec_spnego_seal_packet ,
. sign_packet = gensec_spnego_sign_packet ,
. check_packet = gensec_spnego_check_packet ,
. unseal_packet = gensec_spnego_unseal_packet ,
. session_key = gensec_spnego_session_key ,
. end = gensec_spnego_end
} ;
NTSTATUS gensec_spengo_init ( void )
{
NTSTATUS ret ;
ret = register_backend ( " gensec " , & gensec_spnego_security_ops ) ;
if ( ! NT_STATUS_IS_OK ( ret ) ) {
DEBUG ( 0 , ( " Failed to register '%s' gensec backend! \n " ,
gensec_spnego_security_ops . name ) ) ;
return ret ;
}
return ret ;
}