2003-01-15 15:52:38 +03:00
/*
Unix SMB / Netbios implementation .
Version 3.0
handle NLTMSSP , server side
Copyright ( C ) Andrew Tridgell 2001
Copyright ( C ) Andrew Bartlett 2001 - 2003
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"
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
static NTSTATUS ntlmssp_client_initial ( struct ntlmssp_state * ntlmssp_state ,
DATA_BLOB reply , DATA_BLOB * next_request ) ;
static NTSTATUS ntlmssp_server_negotiate ( struct ntlmssp_state * ntlmssp_state ,
const DATA_BLOB in , DATA_BLOB * out ) ;
static NTSTATUS ntlmssp_client_challenge ( struct ntlmssp_state * ntlmssp_state ,
const DATA_BLOB reply , DATA_BLOB * next_request ) ;
static NTSTATUS ntlmssp_server_auth ( struct ntlmssp_state * ntlmssp_state ,
const DATA_BLOB request , DATA_BLOB * reply ) ;
/**
* Callbacks for NTLMSSP - for both client and server operating modes
*
*/
static const struct ntlmssp_callbacks {
enum NTLMSSP_ROLE role ;
enum NTLM_MESSAGE_TYPE ntlmssp_command ;
NTSTATUS ( * fn ) ( struct ntlmssp_state * ntlmssp_state ,
DATA_BLOB in , DATA_BLOB * out ) ;
} ntlmssp_callbacks [ ] = {
{ NTLMSSP_CLIENT , NTLMSSP_INITIAL , ntlmssp_client_initial } ,
{ NTLMSSP_SERVER , NTLMSSP_NEGOTIATE , ntlmssp_server_negotiate } ,
{ NTLMSSP_CLIENT , NTLMSSP_CHALLENGE , ntlmssp_client_challenge } ,
{ NTLMSSP_SERVER , NTLMSSP_AUTH , ntlmssp_server_auth } ,
{ NTLMSSP_CLIENT , NTLMSSP_UNKNOWN , NULL } ,
{ NTLMSSP_SERVER , NTLMSSP_UNKNOWN , NULL }
} ;
2003-01-15 15:52:38 +03:00
/**
2003-02-24 05:55:00 +03:00
* Print out the NTLMSSP flags for debugging
* @ param neg_flags The flags from the packet
*/
void debug_ntlmssp_flags ( uint32 neg_flags )
{
DEBUG ( 3 , ( " Got NTLMSSP neg_flags=0x%08x \n " , neg_flags ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_UNICODE )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_UNICODE \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_OEM )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_OEM \n " ) ) ;
if ( neg_flags & NTLMSSP_REQUEST_TARGET )
DEBUGADD ( 4 , ( " NTLMSSP_REQUEST_TARGET \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_SIGN )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_SIGN \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_SEAL )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_SEAL \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_LM_KEY )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_LM_KEY \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_NETWARE )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_NETWARE \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_NTLM )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_NTLM \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_DOMAIN_SUPPLIED )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_DOMAIN_SUPPLIED \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_WORKSTATION_SUPPLIED )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_WORKSTATION_SUPPLIED \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_THIS_IS_LOCAL_CALL )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_THIS_IS_LOCAL_CALL \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_ALWAYS_SIGN )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_ALWAYS_SIGN \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_NTLM2 )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_NTLM2 \n " ) ) ;
if ( neg_flags & NTLMSSP_CHAL_TARGET_INFO )
DEBUGADD ( 4 , ( " NTLMSSP_CHAL_TARGET_INFO \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_128 )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_128 \n " ) ) ;
if ( neg_flags & NTLMSSP_NEGOTIATE_KEY_EXCH )
DEBUGADD ( 4 , ( " NTLMSSP_NEGOTIATE_KEY_EXCH \n " ) ) ;
}
/**
* Default challenge generation code .
2003-01-15 15:52:38 +03:00
*
*/
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
static const uint8 * get_challenge ( const struct ntlmssp_state * ntlmssp_state )
2003-01-15 15:52:38 +03:00
{
static uchar chal [ 8 ] ;
generate_random_buffer ( chal , sizeof ( chal ) , False ) ;
return chal ;
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/**
* Default ' we can set the challenge to anything we like ' implementation
*
*/
static BOOL may_set_challenge ( const struct ntlmssp_state * ntlmssp_state )
{
return True ;
}
/**
* Default ' we can set the challenge to anything we like ' implementation
*
* Does not actually do anything , as the value is always in the structure anyway .
*
*/
static NTSTATUS set_challenge ( struct ntlmssp_state * ntlmssp_state , DATA_BLOB * challenge )
{
SMB_ASSERT ( challenge - > length = = 8 ) ;
return NT_STATUS_OK ;
}
/**
* Set a username on an NTLMSSP context - ensures it is talloc ( ) ed
*
*/
NTSTATUS ntlmssp_set_username ( NTLMSSP_STATE * ntlmssp_state , const char * user )
{
ntlmssp_state - > user = talloc_strdup ( ntlmssp_state - > mem_ctx , user ) ;
if ( ! ntlmssp_state - > user ) {
return NT_STATUS_NO_MEMORY ;
}
return NT_STATUS_OK ;
}
/**
* Set a password on an NTLMSSP context - ensures it is talloc ( ) ed
*
*/
NTSTATUS ntlmssp_set_password ( NTLMSSP_STATE * ntlmssp_state , const char * password )
{
if ( ! password ) {
ntlmssp_state - > password = NULL ;
} else {
ntlmssp_state - > password = talloc_strdup ( ntlmssp_state - > mem_ctx , password ) ;
if ( ! ntlmssp_state - > password ) {
return NT_STATUS_NO_MEMORY ;
}
}
return NT_STATUS_OK ;
}
/**
* Set a domain on an NTLMSSP context - ensures it is talloc ( ) ed
*
*/
NTSTATUS ntlmssp_set_domain ( NTLMSSP_STATE * ntlmssp_state , const char * domain )
{
ntlmssp_state - > domain = talloc_strdup ( ntlmssp_state - > mem_ctx , domain ) ;
if ( ! ntlmssp_state - > domain ) {
return NT_STATUS_NO_MEMORY ;
}
return NT_STATUS_OK ;
}
/**
* Set a workstation on an NTLMSSP context - ensures it is talloc ( ) ed
*
*/
NTSTATUS ntlmssp_set_workstation ( NTLMSSP_STATE * ntlmssp_state , const char * workstation )
{
ntlmssp_state - > workstation = talloc_strdup ( ntlmssp_state - > mem_ctx , workstation ) ;
if ( ! ntlmssp_state - > domain ) {
return NT_STATUS_NO_MEMORY ;
}
return NT_STATUS_OK ;
}
/**
* Store a DATA_BLOB containing an NTLMSSP response , for use later .
* This copies the data blob
*/
NTSTATUS ntlmssp_store_response ( NTLMSSP_STATE * ntlmssp_state ,
DATA_BLOB response )
{
ntlmssp_state - > stored_response = data_blob_talloc ( ntlmssp_state - > mem_ctx ,
response . data , response . length ) ;
return NT_STATUS_OK ;
}
/**
* Next state function for the NTLMSSP state machine
*
* @ param ntlmssp_state NTLMSSP State
* @ param in The packet in from the NTLMSSP partner , as a DATA_BLOB
* @ param out The reply , as an allocated DATA_BLOB , caller to free .
* @ return Errors , NT_STATUS_MORE_PROCESSING_REQUIRED or NT_STATUS_OK .
*/
NTSTATUS ntlmssp_update ( NTLMSSP_STATE * ntlmssp_state ,
const DATA_BLOB in , DATA_BLOB * out )
{
DATA_BLOB input ;
uint32 ntlmssp_command ;
int i ;
* out = data_blob ( NULL , 0 ) ;
if ( ! in . length & & ntlmssp_state - > stored_response . length ) {
input = ntlmssp_state - > stored_response ;
/* we only want to read the stored response once - overwrite it */
ntlmssp_state - > stored_response = data_blob ( NULL , 0 ) ;
} else {
input = in ;
}
if ( ! input . length ) {
switch ( ntlmssp_state - > role ) {
case NTLMSSP_CLIENT :
ntlmssp_command = NTLMSSP_INITIAL ;
break ;
case NTLMSSP_SERVER :
/* 'datagram' mode - no neg packet */
ntlmssp_command = NTLMSSP_NEGOTIATE ;
break ;
}
} else {
if ( ! msrpc_parse ( & input , " Cd " ,
" NTLMSSP " ,
& ntlmssp_command ) ) {
DEBUG ( 1 , ( " Failed to parse NTLMSSP packet, could not extract NTLMSSP command \n " ) ) ;
dump_data ( 2 , ( const char * ) input . data , input . length ) ;
return NT_STATUS_INVALID_PARAMETER ;
}
}
if ( ntlmssp_command ! = ntlmssp_state - > expected_state ) {
DEBUG ( 1 , ( " got NTLMSSP command %u, expected %u \n " , ntlmssp_command , ntlmssp_state - > expected_state ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
}
for ( i = 0 ; ntlmssp_callbacks [ i ] . fn ; i + + ) {
if ( ntlmssp_callbacks [ i ] . role = = ntlmssp_state - > role
& & ntlmssp_callbacks [ i ] . ntlmssp_command = = ntlmssp_command ) {
return ntlmssp_callbacks [ i ] . fn ( ntlmssp_state , input , out ) ;
}
}
DEBUG ( 1 , ( " failed to find NTLMSSP callback for NTLMSSP mode %u, command %u \n " ,
ntlmssp_state - > role , ntlmssp_command ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
}
/**
* End an NTLMSSP state machine
*
* @ param ntlmssp_state NTLMSSP State , free ( ) ed by this function
*/
void ntlmssp_end ( NTLMSSP_STATE * * ntlmssp_state )
{
TALLOC_CTX * mem_ctx = ( * ntlmssp_state ) - > mem_ctx ;
( * ntlmssp_state ) - > ref_count - - ;
if ( ( * ntlmssp_state ) - > ref_count = = 0 ) {
data_blob_free ( & ( * ntlmssp_state ) - > chal ) ;
data_blob_free ( & ( * ntlmssp_state ) - > lm_resp ) ;
data_blob_free ( & ( * ntlmssp_state ) - > nt_resp ) ;
talloc_destroy ( mem_ctx ) ;
}
* ntlmssp_state = NULL ;
return ;
}
2003-02-24 05:55:00 +03:00
/**
* Determine correct target name flags for reply , given server role
2003-03-18 14:53:20 +03:00
* and negotiated flags
2003-02-24 05:55:00 +03:00
*
* @ param ntlmssp_state NTLMSSP State
* @ param neg_flags The flags from the packet
* @ param chal_flags The flags to be set in the reply packet
* @ return The ' target name ' string .
*/
2003-01-15 15:52:38 +03:00
2003-02-24 05:55:00 +03:00
static const char * ntlmssp_target_name ( struct ntlmssp_state * ntlmssp_state ,
2003-01-15 15:52:38 +03:00
uint32 neg_flags , uint32 * chal_flags )
{
if ( neg_flags & NTLMSSP_REQUEST_TARGET ) {
* chal_flags | = NTLMSSP_CHAL_TARGET_INFO ;
* chal_flags | = NTLMSSP_REQUEST_TARGET ;
2003-01-16 07:40:07 +03:00
if ( ntlmssp_state - > server_role = = ROLE_STANDALONE ) {
2003-01-15 15:52:38 +03:00
* chal_flags | = NTLMSSP_TARGET_TYPE_SERVER ;
return ntlmssp_state - > get_global_myname ( ) ;
} else {
* chal_flags | = NTLMSSP_TARGET_TYPE_DOMAIN ;
return ntlmssp_state - > get_domain ( ) ;
} ;
} else {
return " " ;
}
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
static void ntlmssp_handle_neg_flags ( struct ntlmssp_state * ntlmssp_state ,
uint32 neg_flags , BOOL allow_lm ) {
if ( neg_flags & NTLMSSP_NEGOTIATE_UNICODE ) {
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_UNICODE ;
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_OEM ;
ntlmssp_state - > unicode = True ;
} else {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_UNICODE ;
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_OEM ;
ntlmssp_state - > unicode = False ;
}
if ( neg_flags & NTLMSSP_NEGOTIATE_LM_KEY & & allow_lm ) {
/* other end forcing us to use LM */
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_LM_KEY ;
ntlmssp_state - > use_ntlmv2 = False ;
} else {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_LM_KEY ;
}
if ( ! ( neg_flags & NTLMSSP_NEGOTIATE_NTLM2 ) ) {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_NTLM2 ;
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_LM_KEY ;
}
if ( ! ( neg_flags & NTLMSSP_NEGOTIATE_128 ) ) {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_128 ;
}
if ( ! ( neg_flags & NTLMSSP_NEGOTIATE_KEY_EXCH ) ) {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_KEY_EXCH ;
}
if ( ( neg_flags & NTLMSSP_REQUEST_TARGET ) ) {
ntlmssp_state - > neg_flags | = NTLMSSP_REQUEST_TARGET ;
}
}
2003-02-24 05:55:00 +03:00
/**
* Next state function for the Negotiate packet
*
* @ param ntlmssp_state NTLMSSP State
* @ param request The request , as a DATA_BLOB
* @ param request The reply , as an allocated DATA_BLOB , caller to free .
* @ return Errors or MORE_PROCESSING_REQUIRED if a reply is sent .
*/
static NTSTATUS ntlmssp_server_negotiate ( struct ntlmssp_state * ntlmssp_state ,
const DATA_BLOB request , DATA_BLOB * reply )
2003-01-15 15:52:38 +03:00
{
DATA_BLOB struct_blob ;
fstring dnsname , dnsdomname ;
2003-03-24 12:54:13 +03:00
uint32 neg_flags = 0 ;
uint32 ntlmssp_command , chal_flags ;
2003-01-15 15:52:38 +03:00
char * cliname = NULL , * domname = NULL ;
const uint8 * cryptkey ;
const char * target_name ;
/* parse the NTLMSSP packet */
#if 0
file_save ( " ntlmssp_negotiate.dat " , request . data , request . length ) ;
# endif
2003-03-24 12:54:13 +03:00
if ( request . length ) {
if ( ! msrpc_parse ( & request , " CddAA " ,
" NTLMSSP " ,
& ntlmssp_command ,
& neg_flags ,
& cliname ,
& domname ) ) {
DEBUG ( 1 , ( " ntlmssp_server_negotiate: failed to parse NTLMSSP: \n " ) ) ;
2003-08-15 05:42:30 +04:00
dump_data ( 2 , ( const char * ) request . data , request . length ) ;
2003-03-24 12:54:13 +03:00
return NT_STATUS_INVALID_PARAMETER ;
}
SAFE_FREE ( cliname ) ;
SAFE_FREE ( domname ) ;
debug_ntlmssp_flags ( neg_flags ) ;
2003-01-15 15:52:38 +03:00
}
2003-03-24 12:54:13 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
ntlmssp_handle_neg_flags ( ntlmssp_state , neg_flags , lp_lanman_auth ( ) ) ;
2003-01-15 15:52:38 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
chal_flags = ntlmssp_state - > neg_flags ;
2003-01-15 15:52:38 +03:00
target_name = ntlmssp_target_name ( ntlmssp_state ,
neg_flags , & chal_flags ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
if ( target_name = = NULL )
2003-08-12 01:00:13 +04:00
return NT_STATUS_INVALID_PARAMETER ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* Ask our caller what challenge they would like in the packet */
cryptkey = ntlmssp_state - > get_challenge ( ntlmssp_state ) ;
/* Check if we may set the challenge */
if ( ! ntlmssp_state - > may_set_challenge ( ntlmssp_state ) ) {
ntlmssp_state - > neg_flags & = ~ NTLMSSP_NEGOTIATE_NTLM2 ;
}
ntlmssp_state - > chal = data_blob_talloc ( ntlmssp_state - > mem_ctx , cryptkey , 8 ) ;
ntlmssp_state - > internal_chal = data_blob_talloc ( ntlmssp_state - > mem_ctx , cryptkey , 8 ) ;
2003-01-15 15:52:38 +03:00
/* This should be a 'netbios domain -> DNS domain' mapping */
dnsdomname [ 0 ] = ' \0 ' ;
2003-08-12 00:51:47 +04:00
get_mydomname ( dnsdomname ) ;
2003-07-03 23:11:31 +04:00
strlower_m ( dnsdomname ) ;
2003-01-15 15:52:38 +03:00
dnsname [ 0 ] = ' \0 ' ;
2003-08-12 00:51:47 +04:00
get_myfullname ( dnsname ) ;
2003-07-03 23:11:31 +04:00
strlower_m ( dnsname ) ;
2003-01-15 15:52:38 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* This creates the 'blob' of names that appears at the end of the packet */
2003-01-15 15:52:38 +03:00
if ( chal_flags & NTLMSSP_CHAL_TARGET_INFO )
{
const char * target_name_dns = " " ;
if ( chal_flags | = NTLMSSP_TARGET_TYPE_DOMAIN ) {
target_name_dns = dnsdomname ;
} else if ( chal_flags | = NTLMSSP_TARGET_TYPE_SERVER ) {
target_name_dns = dnsname ;
}
msrpc_gen ( & struct_blob , " aaaaa " ,
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
NTLMSSP_NAME_TYPE_DOMAIN , target_name ,
NTLMSSP_NAME_TYPE_SERVER , ntlmssp_state - > get_global_myname ( ) ,
NTLMSSP_NAME_TYPE_DOMAIN_DNS , dnsdomname ,
NTLMSSP_NAME_TYPE_SERVER_DNS , dnsname ,
0 , " " ) ;
2003-01-15 15:52:38 +03:00
} else {
struct_blob = data_blob ( NULL , 0 ) ;
}
{
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* Marshel the packet in the right format, be it unicode or ASCII */
2003-01-15 15:52:38 +03:00
const char * gen_string ;
if ( ntlmssp_state - > unicode ) {
gen_string = " CdUdbddB " ;
} else {
gen_string = " CdAdbddB " ;
}
msrpc_gen ( reply , gen_string ,
" NTLMSSP " ,
NTLMSSP_CHALLENGE ,
target_name ,
chal_flags ,
cryptkey , 8 ,
0 , 0 ,
struct_blob . data , struct_blob . length ) ;
}
data_blob_free ( & struct_blob ) ;
2003-02-24 05:55:00 +03:00
ntlmssp_state - > expected_state = NTLMSSP_AUTH ;
2003-01-15 15:52:38 +03:00
return NT_STATUS_MORE_PROCESSING_REQUIRED ;
}
2003-02-24 05:55:00 +03:00
/**
* Next state function for the Authenticate packet
*
* @ param ntlmssp_state NTLMSSP State
* @ param request The request , as a DATA_BLOB
* @ param request The reply , as an allocated DATA_BLOB , caller to free .
* @ return Errors or NT_STATUS_OK .
*/
static NTSTATUS ntlmssp_server_auth ( struct ntlmssp_state * ntlmssp_state ,
const DATA_BLOB request , DATA_BLOB * reply )
2003-01-15 15:52:38 +03:00
{
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
DATA_BLOB encrypted_session_key = data_blob ( NULL , 0 ) ;
DATA_BLOB nt_session_key = data_blob ( NULL , 0 ) ;
DATA_BLOB lm_session_key = data_blob ( NULL , 0 ) ;
DATA_BLOB session_key = data_blob ( NULL , 0 ) ;
uint32 ntlmssp_command , auth_flags ;
2003-01-15 15:52:38 +03:00
NTSTATUS nt_status ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* used by NTLM2 */
BOOL doing_ntlm2 = False ;
uchar session_nonce [ 16 ] ;
uchar session_nonce_hash [ 16 ] ;
2003-01-15 15:52:38 +03:00
const char * parse_string ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
char * domain = NULL ;
char * user = NULL ;
char * workstation = NULL ;
2003-01-15 15:52:38 +03:00
/* parse the NTLMSSP packet */
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
* reply = data_blob ( NULL , 0 ) ;
2003-01-15 15:52:38 +03:00
#if 0
file_save ( " ntlmssp_auth.dat " , request . data , request . length ) ;
# endif
if ( ntlmssp_state - > unicode ) {
parse_string = " CdBBUUUBd " ;
} else {
parse_string = " CdBBAAABd " ;
}
data_blob_free ( & ntlmssp_state - > lm_resp ) ;
data_blob_free ( & ntlmssp_state - > nt_resp ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
ntlmssp_state - > user = NULL ;
ntlmssp_state - > domain = NULL ;
ntlmssp_state - > workstation = NULL ;
2003-01-15 15:52:38 +03:00
/* now the NTLMSSP encoded auth hashes */
if ( ! msrpc_parse ( & request , parse_string ,
" NTLMSSP " ,
& ntlmssp_command ,
& ntlmssp_state - > lm_resp ,
& ntlmssp_state - > nt_resp ,
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
& domain ,
& user ,
& workstation ,
& encrypted_session_key ,
& auth_flags ) ) {
2003-03-24 12:54:13 +03:00
DEBUG ( 1 , ( " ntlmssp_server_auth: failed to parse NTLMSSP: \n " ) ) ;
2003-08-15 05:42:30 +04:00
dump_data ( 2 , ( const char * ) request . data , request . length ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
data_blob_free ( & encrypted_session_key ) ;
auth_flags = 0 ;
/* Try again with a shorter string (Win9X truncates this packet) */
if ( ntlmssp_state - > unicode ) {
parse_string = " CdBBUUU " ;
} else {
parse_string = " CdBBAAA " ;
}
/* now the NTLMSSP encoded auth hashes */
if ( ! msrpc_parse ( & request , parse_string ,
" NTLMSSP " ,
& ntlmssp_command ,
& ntlmssp_state - > lm_resp ,
& ntlmssp_state - > nt_resp ,
& domain ,
& user ,
& workstation ) ) {
DEBUG ( 1 , ( " ntlmssp_server_auth: failed to parse NTLMSSP: \n " ) ) ;
dump_data ( 2 , ( const char * ) request . data , request . length ) ;
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
return NT_STATUS_INVALID_PARAMETER ;
}
2003-01-15 15:52:38 +03:00
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_set_domain ( ntlmssp_state , domain ) ) ) {
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
data_blob_free ( & encrypted_session_key ) ;
return nt_status ;
}
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_set_username ( ntlmssp_state , user ) ) ) {
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
data_blob_free ( & encrypted_session_key ) ;
return nt_status ;
}
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_set_workstation ( ntlmssp_state , workstation ) ) ) {
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
data_blob_free ( & encrypted_session_key ) ;
return nt_status ;
}
SAFE_FREE ( domain ) ;
SAFE_FREE ( user ) ;
SAFE_FREE ( workstation ) ;
2003-07-25 08:24:40 +04:00
DEBUG ( 3 , ( " Got user=[%s] domain=[%s] workstation=[%s] len1=%lu len2=%lu \n " ,
ntlmssp_state - > user , ntlmssp_state - > domain , ntlmssp_state - > workstation , ( unsigned long ) ntlmssp_state - > lm_resp . length , ( unsigned long ) ntlmssp_state - > nt_resp . length ) ) ;
2003-01-15 15:52:38 +03:00
#if 0
file_save ( " nthash1.dat " , & ntlmssp_state - > nt_resp . data , & ntlmssp_state - > nt_resp . length ) ;
file_save ( " lmhash1.dat " , & ntlmssp_state - > lm_resp . data , & ntlmssp_state - > lm_resp . length ) ;
# endif
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
/* NTLM2 uses a 'challenge' that is made of up both the server challenge, and a
client challenge
2003-01-15 15:52:38 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
However , the NTLM2 flag may still be set for the real NTLMv2 logins , be careful .
*/
if ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_NTLM2 ) {
if ( ntlmssp_state - > nt_resp . length = = 24 & & ntlmssp_state - > lm_resp . length = = 24 ) {
struct MD5Context md5_session_nonce_ctx ;
SMB_ASSERT ( ntlmssp_state - > internal_chal . data & & ntlmssp_state - > internal_chal . length = = 8 ) ;
doing_ntlm2 = True ;
memcpy ( session_nonce , ntlmssp_state - > internal_chal . data , 8 ) ;
memcpy ( & session_nonce [ 8 ] , ntlmssp_state - > lm_resp . data , 8 ) ;
MD5Init ( & md5_session_nonce_ctx ) ;
MD5Update ( & md5_session_nonce_ctx , session_nonce , 16 ) ;
MD5Final ( session_nonce_hash , & md5_session_nonce_ctx ) ;
ntlmssp_state - > chal = data_blob_talloc ( ntlmssp_state - > mem_ctx , session_nonce_hash , 8 ) ;
/* LM response is no longer useful */
data_blob_free ( & ntlmssp_state - > lm_resp ) ;
/* We changed the effective challenge - set it */
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_state - > set_challenge ( ntlmssp_state , & ntlmssp_state - > chal ) ) ) {
data_blob_free ( & encrypted_session_key ) ;
return nt_status ;
}
}
}
/* Finally, actually ask if the password is OK */
if ( ! NT_STATUS_IS_OK ( nt_status = ntlmssp_state - > check_password ( ntlmssp_state , & nt_session_key , & lm_session_key ) ) ) {
data_blob_free ( & encrypted_session_key ) ;
return nt_status ;
}
dump_data_pw ( " NT session key: \n " , nt_session_key . data , nt_session_key . length ) ;
dump_data_pw ( " LM first-8: \n " , lm_session_key . data , lm_session_key . length ) ;
/* Handle the different session key derivation for NTLM2 */
if ( doing_ntlm2 ) {
if ( nt_session_key . data & & nt_session_key . length = = 16 ) {
session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 16 ) ;
hmac_md5 ( nt_session_key . data , session_nonce ,
sizeof ( session_nonce ) , session_key . data ) ;
dump_data_pw ( " NTLM2 session key: \n " , session_key . data , session_key . length ) ;
}
} else if ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_LM_KEY ) {
if ( lm_session_key . data & & lm_session_key . length > = 8 & &
ntlmssp_state - > lm_resp . data & & ntlmssp_state - > lm_resp . length = = 24 ) {
session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 16 ) ;
SMBsesskeygen_lmv1 ( lm_session_key . data , ntlmssp_state - > lm_resp . data ,
session_key . data ) ;
dump_data_pw ( " LM session key: \n " , session_key . data , session_key . length ) ;
}
} else if ( nt_session_key . data ) {
session_key = nt_session_key ;
dump_data_pw ( " unmodified session key: \n " , session_key . data , session_key . length ) ;
}
/* With KEY_EXCH, the client supplies the proposed session key,
but encrypts it with the long - term key */
if ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_KEY_EXCH ) {
if ( ! encrypted_session_key . data | | encrypted_session_key . length ! = 16 ) {
data_blob_free ( & encrypted_session_key ) ;
DEBUG ( 1 , ( " Client-supplied KEY_EXCH session key was of invalid length (%u)! \n " ,
encrypted_session_key . length ) ) ;
return NT_STATUS_INVALID_PARAMETER ;
} else if ( ! session_key . data | | session_key . length ! = 16 ) {
DEBUG ( 5 , ( " server session key is invalid (len == %u), cannot do KEY_EXCH! \n " ,
session_key . length ) ) ;
} else {
dump_data_pw ( " KEY_EXCH session key (enc): \n " , encrypted_session_key . data , encrypted_session_key . length ) ;
SamOEMhash ( encrypted_session_key . data ,
session_key . data ,
encrypted_session_key . length ) ;
ntlmssp_state - > session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx ,
encrypted_session_key . data ,
encrypted_session_key . length ) ;
dump_data_pw ( " KEY_EXCH session key: \n " , session_key . data , session_key . length ) ;
}
} else {
ntlmssp_state - > session_key = session_key ;
}
data_blob_free ( & encrypted_session_key ) ;
/* allow arbitarily many authentications */
ntlmssp_state - > expected_state = NTLMSSP_AUTH ;
2003-01-15 15:52:38 +03:00
return nt_status ;
}
2003-02-24 05:55:00 +03:00
/**
* Create an NTLMSSP state machine
*
2003-03-18 14:53:20 +03:00
* @ param ntlmssp_state NTLMSSP State , allocated by this function
2003-02-24 05:55:00 +03:00
*/
NTSTATUS ntlmssp_server_start ( NTLMSSP_STATE * * ntlmssp_state )
{
TALLOC_CTX * mem_ctx ;
mem_ctx = talloc_init ( " NTLMSSP context " ) ;
* ntlmssp_state = talloc_zero ( mem_ctx , sizeof ( * * ntlmssp_state ) ) ;
if ( ! * ntlmssp_state ) {
DEBUG ( 0 , ( " ntlmssp_server_start: talloc failed! \n " ) ) ;
talloc_destroy ( mem_ctx ) ;
return NT_STATUS_NO_MEMORY ;
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > role = NTLMSSP_SERVER ;
2003-02-24 05:55:00 +03:00
( * ntlmssp_state ) - > mem_ctx = mem_ctx ;
( * ntlmssp_state ) - > get_challenge = get_challenge ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > set_challenge = set_challenge ;
( * ntlmssp_state ) - > may_set_challenge = may_set_challenge ;
2003-02-24 05:55:00 +03:00
( * ntlmssp_state ) - > get_global_myname = global_myname ;
( * ntlmssp_state ) - > get_domain = lp_workgroup ;
( * ntlmssp_state ) - > server_role = ROLE_DOMAIN_MEMBER ; /* a good default */
( * ntlmssp_state ) - > expected_state = NTLMSSP_NEGOTIATE ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > ref_count = 1 ;
2003-02-24 05:55:00 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > neg_flags =
NTLMSSP_NEGOTIATE_128 |
NTLMSSP_NEGOTIATE_NTLM |
NTLMSSP_NEGOTIATE_NTLM2 |
NTLMSSP_NEGOTIATE_KEY_EXCH |
NTLMSSP_NEGOTIATE_SIGN ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_OK ;
}
/*********************************************************************
Client side NTLMSSP
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
/**
* Next state function for the Initial packet
*
* @ param ntlmssp_state NTLMSSP State
* @ param request The request , as a DATA_BLOB . reply . data must be NULL
* @ param request The reply , as an allocated DATA_BLOB , caller to free .
* @ return Errors or NT_STATUS_OK .
*/
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
static NTSTATUS ntlmssp_client_initial ( struct ntlmssp_state * ntlmssp_state ,
2003-02-24 05:55:00 +03:00
DATA_BLOB reply , DATA_BLOB * next_request )
{
if ( ntlmssp_state - > unicode ) {
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_UNICODE ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
} else {
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_OEM ;
2003-02-24 05:55:00 +03:00
}
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
if ( ntlmssp_state - > use_ntlmv2 ) {
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_NTLM2 ;
}
2003-08-12 09:53:43 +04:00
ntlmssp_state - > neg_flags | = NTLMSSP_NEGOTIATE_NTLM2 ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
2003-02-24 05:55:00 +03:00
/* generate the ntlmssp negotiate packet */
msrpc_gen ( next_request , " CddAA " ,
" NTLMSSP " ,
NTLMSSP_NEGOTIATE ,
ntlmssp_state - > neg_flags ,
ntlmssp_state - > get_domain ( ) ,
ntlmssp_state - > get_global_myname ( ) ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
ntlmssp_state - > expected_state = NTLMSSP_CHALLENGE ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_MORE_PROCESSING_REQUIRED ;
}
/**
* Next state function for the Challenge Packet . Generate an auth packet .
*
* @ param ntlmssp_state NTLMSSP State
* @ param request The request , as a DATA_BLOB . reply . data must be NULL
* @ param request The reply , as an allocated DATA_BLOB , caller to free .
* @ return Errors or NT_STATUS_OK .
*/
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
static NTSTATUS ntlmssp_client_challenge ( struct ntlmssp_state * ntlmssp_state ,
2003-02-24 05:55:00 +03:00
const DATA_BLOB reply , DATA_BLOB * next_request )
{
uint32 chal_flags , ntlmssp_command , unkn1 , unkn2 ;
DATA_BLOB server_domain_blob ;
DATA_BLOB challenge_blob ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
DATA_BLOB struct_blob = data_blob ( NULL , 0 ) ;
2003-02-24 05:55:00 +03:00
char * server_domain ;
const char * chal_parse_string ;
const char * auth_gen_string ;
DATA_BLOB lm_response = data_blob ( NULL , 0 ) ;
DATA_BLOB nt_response = data_blob ( NULL , 0 ) ;
DATA_BLOB session_key = data_blob ( NULL , 0 ) ;
2003-09-06 23:23:24 +04:00
DATA_BLOB encrypted_session_key = data_blob ( NULL , 0 ) ;
2003-02-24 05:55:00 +03:00
if ( ! msrpc_parse ( & reply , " CdBd " ,
" NTLMSSP " ,
& ntlmssp_command ,
& server_domain_blob ,
& chal_flags ) ) {
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
DEBUG ( 1 , ( " Failed to parse the NTLMSSP Challenge: (#1) \n " ) ) ;
2003-08-15 05:42:30 +04:00
dump_data ( 2 , ( const char * ) reply . data , reply . length ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
2003-02-24 05:55:00 +03:00
return NT_STATUS_INVALID_PARAMETER ;
}
data_blob_free ( & server_domain_blob ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
DEBUG ( 3 , ( " Got challenge flags: \n " ) ) ;
debug_ntlmssp_flags ( chal_flags ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
ntlmssp_handle_neg_flags ( ntlmssp_state , chal_flags , lp_client_lanman_auth ( ) ) ;
if ( ntlmssp_state - > unicode ) {
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
if ( chal_flags & NTLMSSP_CHAL_TARGET_INFO ) {
chal_parse_string = " CdUdbddB " ;
} else {
chal_parse_string = " CdUdbdd " ;
}
2003-02-24 05:55:00 +03:00
auth_gen_string = " CdBBUUUBd " ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
} else {
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
if ( chal_flags & NTLMSSP_CHAL_TARGET_INFO ) {
chal_parse_string = " CdAdbddB " ;
} else {
chal_parse_string = " CdAdbdd " ;
}
2003-09-06 23:23:24 +04:00
2003-02-24 05:55:00 +03:00
auth_gen_string = " CdBBAAABd " ;
2003-09-06 23:23:24 +04:00
}
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
DEBUG ( 3 , ( " NTLMSSP: Set final flags: \n " ) ) ;
debug_ntlmssp_flags ( ntlmssp_state - > neg_flags ) ;
2003-02-24 05:55:00 +03:00
if ( ! msrpc_parse ( & reply , chal_parse_string ,
" NTLMSSP " ,
& ntlmssp_command ,
& server_domain ,
& chal_flags ,
& challenge_blob , 8 ,
& unkn1 , & unkn2 ,
& struct_blob ) ) {
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
DEBUG ( 1 , ( " Failed to parse the NTLMSSP Challenge: (#2) \n " ) ) ;
2003-08-15 05:42:30 +04:00
dump_data ( 2 , ( const char * ) reply . data , reply . length ) ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_INVALID_PARAMETER ;
}
2003-08-11 00:18:05 +04:00
ntlmssp_state - > server_domain = talloc_strdup ( ntlmssp_state - > mem_ctx ,
server_domain ) ;
2003-02-24 05:55:00 +03:00
SAFE_FREE ( server_domain ) ;
if ( challenge_blob . length ! = 8 ) {
2003-05-09 18:42:20 +04:00
data_blob_free ( & struct_blob ) ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_INVALID_PARAMETER ;
}
2003-08-14 05:08:00 +04:00
if ( ! ntlmssp_state - > password ) {
/* do nothing - blobs are zero length */
} else if ( ntlmssp_state - > use_ntlmv2 ) {
2003-02-24 05:55:00 +03:00
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
if ( ! struct_blob . length ) {
/* be lazy, match win2k - we can't do NTLMv2 without it */
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
DEBUG ( 1 , ( " Server did not provide 'target information', required for NTLMv2 \n " ) ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
return NT_STATUS_INVALID_PARAMETER ;
}
2003-02-24 05:55:00 +03:00
/* TODO: if the remote server is standalone, then we should replace 'domain'
with the server name as supplied above */
if ( ! SMBNTLMv2encrypt ( ntlmssp_state - > user ,
ntlmssp_state - > domain ,
2003-05-09 18:42:20 +04:00
ntlmssp_state - > password , & challenge_blob ,
& struct_blob ,
& lm_response , & nt_response , & session_key ) ) {
2003-02-24 05:55:00 +03:00
data_blob_free ( & challenge_blob ) ;
2003-05-09 18:42:20 +04:00
data_blob_free ( & struct_blob ) ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_NO_MEMORY ;
}
2003-08-12 09:53:43 +04:00
} else if ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_NTLM2 ) {
struct MD5Context md5_session_nonce_ctx ;
uchar nt_hash [ 16 ] ;
uchar session_nonce [ 16 ] ;
2003-09-06 23:23:24 +04:00
uchar session_nonce_hash [ 16 ] ;
uchar nt_session_key [ 16 ] ;
2003-08-12 09:53:43 +04:00
E_md4hash ( ntlmssp_state - > password , nt_hash ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
lm_response = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 24 ) ;
2003-08-12 09:53:43 +04:00
generate_random_buffer ( lm_response . data , 8 , False ) ;
memset ( lm_response . data + 8 , 0 , 16 ) ;
2003-09-06 23:23:24 +04:00
memcpy ( session_nonce , challenge_blob . data , 8 ) ;
memcpy ( & session_nonce [ 8 ] , lm_response . data , 8 ) ;
2003-08-12 09:53:43 +04:00
MD5Init ( & md5_session_nonce_ctx ) ;
MD5Update ( & md5_session_nonce_ctx , challenge_blob . data , 8 ) ;
MD5Update ( & md5_session_nonce_ctx , lm_response . data , 8 ) ;
2003-09-06 23:23:24 +04:00
MD5Final ( session_nonce_hash , & md5_session_nonce_ctx ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
DEBUG ( 5 , ( " NTLMSSP challenge set by NTLM2 \n " ) ) ;
DEBUG ( 5 , ( " challenge is: \n " ) ) ;
dump_data ( 5 , session_nonce_hash , 8 ) ;
2003-08-12 09:53:43 +04:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
nt_response = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 24 ) ;
2003-08-12 09:53:43 +04:00
SMBNTencrypt ( ntlmssp_state - > password ,
2003-09-06 23:23:24 +04:00
session_nonce_hash ,
2003-08-12 09:53:43 +04:00
nt_response . data ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 16 ) ;
2003-09-06 23:23:24 +04:00
SMBsesskeygen_ntv1 ( nt_hash , NULL , nt_session_key ) ;
hmac_md5 ( nt_session_key , session_nonce , sizeof ( session_nonce ) , session_key . data ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
dump_data_pw ( " NTLM2 session key: \n " , session_key . data , session_key . length ) ;
2003-02-24 05:55:00 +03:00
} else {
2003-08-12 09:53:43 +04:00
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
uchar lm_hash [ 16 ] ;
2003-02-24 05:55:00 +03:00
uchar nt_hash [ 16 ] ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
E_deshash ( ntlmssp_state - > password , lm_hash ) ;
2003-02-24 05:55:00 +03:00
E_md4hash ( ntlmssp_state - > password , nt_hash ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
/* lanman auth is insecure, it may be disabled */
2003-02-24 05:55:00 +03:00
if ( lp_client_lanman_auth ( ) ) {
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
lm_response = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 24 ) ;
2003-02-24 05:55:00 +03:00
SMBencrypt ( ntlmssp_state - > password , challenge_blob . data ,
lm_response . data ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
}
2003-02-24 05:55:00 +03:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
nt_response = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 24 ) ;
2003-02-24 05:55:00 +03:00
SMBNTencrypt ( ntlmssp_state - > password , challenge_blob . data ,
nt_response . data ) ;
2003-08-12 09:53:43 +04:00
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx , NULL , 16 ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
if ( ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_LM_KEY )
2003-08-12 09:53:43 +04:00
& & lp_client_lanman_auth ( ) ) {
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
SMBsesskeygen_lmv1 ( lm_hash , lm_response . data ,
session_key . data ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
dump_data_pw ( " LM session key \n " , session_key . data , session_key . length ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
} else {
SMBsesskeygen_ntv1 ( nt_hash , NULL , session_key . data ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
dump_data_pw ( " NT session key: \n " , session_key . data , session_key . length ) ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
}
2003-02-24 05:55:00 +03:00
}
2003-05-09 18:42:20 +04:00
data_blob_free ( & struct_blob ) ;
2003-02-24 05:55:00 +03:00
2003-09-06 23:23:24 +04:00
/* Key exchange encryptes a new client-generated session key with
the password - derived key */
if ( ntlmssp_state - > neg_flags & NTLMSSP_NEGOTIATE_KEY_EXCH ) {
uint8 client_session_key [ 16 ] ;
generate_random_buffer ( client_session_key , sizeof ( client_session_key ) , False ) ;
encrypted_session_key = data_blob ( client_session_key , sizeof ( client_session_key ) ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
dump_data_pw ( " KEY_EXCH session key: \n " , encrypted_session_key . data , encrypted_session_key . length ) ;
2003-09-06 23:23:24 +04:00
SamOEMhash ( encrypted_session_key . data , session_key . data , encrypted_session_key . length ) ;
data_blob_free ( & session_key ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
session_key = data_blob_talloc ( ntlmssp_state - > mem_ctx , client_session_key , sizeof ( client_session_key ) ) ;
dump_data_pw ( " KEY_EXCH session key (enc): \n " , encrypted_session_key . data , encrypted_session_key . length ) ;
2003-09-06 23:23:24 +04:00
}
2003-02-24 05:55:00 +03:00
/* this generates the actual auth packet */
if ( ! msrpc_gen ( next_request , auth_gen_string ,
" NTLMSSP " ,
NTLMSSP_AUTH ,
lm_response . data , lm_response . length ,
nt_response . data , nt_response . length ,
ntlmssp_state - > domain ,
ntlmssp_state - > user ,
ntlmssp_state - > get_global_myname ( ) ,
2003-09-06 23:23:24 +04:00
encrypted_session_key . data , encrypted_session_key . length ,
2003-02-24 05:55:00 +03:00
ntlmssp_state - > neg_flags ) ) {
return NT_STATUS_NO_MEMORY ;
}
2003-09-06 23:23:24 +04:00
data_blob_free ( & encrypted_session_key ) ;
2003-03-18 14:53:20 +03:00
data_blob_free ( & ntlmssp_state - > chal ) ;
2003-02-24 05:55:00 +03:00
2003-03-18 14:53:20 +03:00
ntlmssp_state - > chal = challenge_blob ;
ntlmssp_state - > lm_resp = lm_response ;
ntlmssp_state - > nt_resp = nt_response ;
2003-02-24 05:55:00 +03:00
ntlmssp_state - > session_key = session_key ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
ntlmssp_state - > expected_state = NTLMSSP_UNKNOWN ;
2003-02-24 05:55:00 +03:00
return NT_STATUS_MORE_PROCESSING_REQUIRED ;
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
NTSTATUS ntlmssp_client_start ( NTLMSSP_STATE * * ntlmssp_state )
2003-02-24 05:55:00 +03:00
{
TALLOC_CTX * mem_ctx ;
mem_ctx = talloc_init ( " NTLMSSP Client context " ) ;
* ntlmssp_state = talloc_zero ( mem_ctx , sizeof ( * * ntlmssp_state ) ) ;
if ( ! * ntlmssp_state ) {
DEBUG ( 0 , ( " ntlmssp_server_start: talloc failed! \n " ) ) ;
talloc_destroy ( mem_ctx ) ;
return NT_STATUS_NO_MEMORY ;
}
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > role = NTLMSSP_CLIENT ;
2003-02-24 05:55:00 +03:00
( * ntlmssp_state ) - > mem_ctx = mem_ctx ;
( * ntlmssp_state ) - > get_global_myname = global_myname ;
( * ntlmssp_state ) - > get_domain = lp_workgroup ;
( * ntlmssp_state ) - > unicode = True ;
Jeremy requested that I get my NTLMSSP patch into CVS. He didn't request
the schannel code, but I've included that anyway. :-)
This patch revives the client-side NTLMSSP support for RPC named pipes
in Samba, and cleans up the client and server schannel code. The use of the
new code is enabled by the 'sign', 'seal' and 'schannel' commands in
rpcclient.
The aim was to prove that our separate NTLMSSP client library actually
implements NTLMSSP signing and sealing as per Microsoft's NTLMv1 implementation,
in the hope that knowing this will assist us in correctly implementing
NTLMSSP signing for SMB packets. (Still not yet functional)
This patch replaces the NTLMSSP implementation in rpc_client/cli_pipe.c with
calls to libsmb/ntlmssp.c. In the process, we have gained the ability to
use the more secure NT password, and the ability to sign-only, instead of
having to seal the pipe connection. (Previously we were limited to sealing,
and could only use the LM-password derived key).
Our new client-side NTLMSSP code also needed alteration to cope with our
comparatively simple server-side implementation. A future step is to replace
it with calls to the same NTLMSSP library.
Also included in this patch is the schannel 'sign only' patch I submitted to
the team earlier. While not enabled (and not functional, at this stage) the
work in this patch makes the code paths *much* easier to follow. I have also
included similar hooks in rpccleint to allow the use of schannel on *any* pipe.
rpcclient now defaults to not using schannel (or any other extra per-pipe
authenticiation) for any connection. The 'schannel' command enables schannel
for all pipes until disabled.
This code is also much more secure than the previous code, as changes to our
cli_pipe routines ensure that the authentication footer cannot be removed
by an attacker, and more error states are correctly handled.
(The same needs to be done to our server)
Andrew Bartlett
(This used to be commit 5472ddc9eaf4e79c5b2e1c8ee8c7f190dc285f19)
2003-07-14 12:46:32 +04:00
( * ntlmssp_state ) - > use_ntlmv2 = lp_client_ntlmv2_auth ( ) ;
Changes all over the shop, but all towards:
- NTLM2 support in the server
- KEY_EXCH support in the server
- variable length session keys.
In detail:
- NTLM2 is an extension of NTLMv1, that is compatible with existing
domain controllers (unlike NTLMv2, which requires a DC upgrade).
* This is known as 'NTLMv2 session security' *
(This is not yet implemented on the RPC pipes however, so there may
well still be issues for PDC setups, particuarly around password
changes. We do not fully understand the sign/seal implications of
NTLM2 on RPC pipes.)
This requires modifications to our authentication subsystem, as we
must handle the 'challege' input into the challenge-response algorithm
being changed. This also needs to be turned off for
'security=server', which does not support this.
- KEY_EXCH is another 'security' mechanism, whereby the session key
actually used by the server is sent by the client, rather than being
the shared-secret directly or indirectly.
- As both these methods change the session key, the auth subsystem
needed to be changed, to 'override' session keys provided by the
backend.
- There has also been a major overhaul of the NTLMSSP subsystem, to merge the 'client' and 'server' functions, so they both operate on a single structure. This should help the SPNEGO implementation.
- The 'names blob' in NTLMSSP is always in unicode - never in ascii.
Don't make an ascii version ever.
- The other big change is to allow variable length session keys. We
have always assumed that session keys are 16 bytes long - and padded
to this length if shorter. However, Kerberos session keys are 8 bytes
long, when the krb5 login uses DES.
* This fix allows SMB signging on machines not yet running MIT KRB5 1.3.1. *
- Add better DEBUG() messages to ntlm_auth, warning administrators of
misconfigurations that prevent access to the privileged pipe. This
should help reduce some of the 'it just doesn't work' issues.
- Fix data_blob_talloc() to behave the same way data_blob() does when
passed a NULL data pointer. (just allocate)
REMEMBER to make clean after this commit - I have changed plenty of data structures...
(This used to be commit f3bbc87b0dac63426cda6fac7a295d3aad810ecc)
2003-11-22 16:19:38 +03:00
( * ntlmssp_state ) - > expected_state = NTLMSSP_INITIAL ;
( * ntlmssp_state ) - > ref_count = 1 ;
2003-02-24 05:55:00 +03:00
( * ntlmssp_state ) - > neg_flags =
2003-03-18 14:53:20 +03:00
NTLMSSP_NEGOTIATE_128 |
2003-02-24 05:55:00 +03:00
NTLMSSP_NEGOTIATE_NTLM |
2003-09-06 23:23:24 +04:00
NTLMSSP_NEGOTIATE_NTLM2 |
NTLMSSP_NEGOTIATE_KEY_EXCH |
2003-09-16 05:27:46 +04:00
/*
* 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 .
* */
NTLMSSP_NEGOTIATE_SIGN |
2003-02-24 05:55:00 +03:00
NTLMSSP_REQUEST_TARGET ;
return NT_STATUS_OK ;
}