2006-06-01 02:40:51 +04:00
/*
* fs / cifs / sess . c
*
* SMB / CIFS session setup handling routines
*
2009-04-30 21:45:10 +04:00
* Copyright ( c ) International Business Machines Corp . , 2006 , 2009
2006-06-01 02:40:51 +04:00
* Author ( s ) : Steve French ( sfrench @ us . ibm . com )
*
* This library is free software ; you can redistribute it and / or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation ; either version 2.1 of the License , or
* ( at your option ) any later version .
*
* This library 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 Lesser General Public License for more details .
*
* You should have received a copy of the GNU Lesser General Public License
* along with this library ; if not , write to the Free Software
* Foundation , Inc . , 59 Temple Place , Suite 330 , Boston , MA 02111 - 1307 USA
*/
# include "cifspdu.h"
# include "cifsglob.h"
# include "cifsproto.h"
# include "cifs_unicode.h"
# include "cifs_debug.h"
# include "ntlmssp.h"
# include "nterr.h"
2006-06-01 09:09:10 +04:00
# include <linux/utsname.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 11:04:11 +03:00
# include <linux/slab.h>
2007-11-17 02:37:35 +03:00
# include "cifs_spnego.h"
2006-06-01 02:40:51 +04:00
2010-04-24 15:57:47 +04:00
/*
* Checks if this is the first smb session to be reconnected after
* the socket has been reestablished ( so we know whether to use vc 0 ) .
* Called while holding the cifs_tcp_ses_lock , so do not block
*/
2011-05-27 08:34:02 +04:00
static bool is_first_ses_reconnect ( struct cifs_ses * ses )
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
{
struct list_head * tmp ;
2011-05-27 08:34:02 +04:00
struct cifs_ses * tmp_ses ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
list_for_each ( tmp , & ses - > server - > smb_ses_list ) {
2011-05-27 08:34:02 +04:00
tmp_ses = list_entry ( tmp , struct cifs_ses ,
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
smb_ses_list ) ;
if ( tmp_ses - > need_reconnect = = false )
return false ;
}
/* could not find a session that was already connected,
this must be the first one we are reconnecting */
return true ;
}
/*
* vc number 0 is treated specially by some servers , and should be the
* first one we request . After that we can use vcnumbers up to maxvcs ,
* one for each smb session ( some Windows versions set maxvcs incorrectly
* so maxvc = 1 can be ignored ) . If we have too many vcs , we can reuse
* any vc but zero ( some servers reset the connection on vcnum zero )
*
*/
2011-05-27 08:34:02 +04:00
static __le16 get_next_vcnum ( struct cifs_ses * ses )
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
{
__u16 vcnum = 0 ;
struct list_head * tmp ;
2011-05-27 08:34:02 +04:00
struct cifs_ses * tmp_ses ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
__u16 max_vcs = ses - > server - > max_vcs ;
__u16 i ;
int free_vc_found = 0 ;
/* Quoting the MS-SMB specification: "Windows-based SMB servers set this
field to one but do not enforce this limit , which allows an SMB client
to establish more virtual circuits than allowed by this value . . . but
other server implementations can enforce this limit . " */
if ( max_vcs < 2 )
max_vcs = 0xFFFF ;
2010-10-18 21:59:37 +04:00
spin_lock ( & cifs_tcp_ses_lock ) ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
if ( ( ses - > need_reconnect ) & & is_first_ses_reconnect ( ses ) )
goto get_vc_num_exit ; /* vcnum will be zero */
for ( i = ses - > server - > srv_count - 1 ; i < max_vcs ; i + + ) {
if ( i = = 0 ) /* this is the only connection, use vc 0 */
break ;
free_vc_found = 1 ;
list_for_each ( tmp , & ses - > server - > smb_ses_list ) {
2011-05-27 08:34:02 +04:00
tmp_ses = list_entry ( tmp , struct cifs_ses ,
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
smb_ses_list ) ;
if ( tmp_ses - > vcnum = = i ) {
free_vc_found = 0 ;
break ; /* found duplicate, try next vcnum */
}
}
if ( free_vc_found )
break ; /* we found a vcnumber that will work - use it */
}
if ( i = = 0 )
vcnum = 0 ; /* for most common case, ie if one smb session, use
vc zero . Also for case when no free vcnum , zero
is safest to send ( some clients only send zero ) */
else if ( free_vc_found = = 0 )
vcnum = 1 ; /* we can not reuse vc=0 safely, since some servers
reset all uids on that , but 1 is ok . */
else
vcnum = i ;
ses - > vcnum = vcnum ;
get_vc_num_exit :
2010-10-18 21:59:37 +04:00
spin_unlock ( & cifs_tcp_ses_lock ) ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
2009-05-01 20:21:04 +04:00
return cpu_to_le16 ( vcnum ) ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
}
2011-05-27 08:34:02 +04:00
static __u32 cifs_ssetup_hdr ( struct cifs_ses * ses , SESSION_SETUP_ANDX * pSMB )
2006-06-01 02:40:51 +04:00
{
__u32 capabilities = 0 ;
/* init fields common to all four types of SessSetup */
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
/* Note that offsets for first seven fields in req struct are same */
/* in CIFS Specs so does not matter which of 3 forms of struct */
/* that we use in next few lines */
/* Note that header is initialized to zero in header_assemble */
2006-06-01 02:40:51 +04:00
pSMB - > req . AndXCommand = 0xFF ;
2011-10-11 14:41:32 +04:00
pSMB - > req . MaxBufferSize = cpu_to_le16 ( min_t ( u32 ,
CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4 ,
USHRT_MAX ) ) ;
2006-06-01 02:40:51 +04:00
pSMB - > req . MaxMpxCount = cpu_to_le16 ( ses - > server - > maxReq ) ;
[CIFS] Fix multiuser mounts so server does not invalidate earlier security contexts
When two different users mount the same Windows 2003 Server share using CIFS,
the first session mounted can be invalidated. Some servers invalidate the first
smb session when a second similar user (e.g. two users who get mapped by server to "guest")
authenticates an smb session from the same client.
By making sure that we set the 2nd and subsequent vc numbers to nonzero values,
this ensures that we will not have this problem.
Fixes Samba bug 6004, problem description follows:
How to reproduce:
- configure an "open share" (full permissions to Guest user) on Windows 2003
Server (I couldn't reproduce the problem with Samba server or Windows older
than 2003)
- mount the share twice with different users who will be authenticated as guest.
noacl,noperm,user=john,dir_mode=0700,domain=DOMAIN,rw
noacl,noperm,user=jeff,dir_mode=0700,domain=DOMAIN,rw
Result:
- just the mount point mounted last is accessible:
Signed-off-by: Steve French <sfrench@us.ibm.com>
2009-02-20 08:43:09 +03:00
pSMB - > req . VcNumber = get_next_vcnum ( ses ) ;
2006-06-01 02:40:51 +04:00
/* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */
2007-07-07 23:25:05 +04:00
/* BB verify whether signing required on neg or just on auth frame
2006-06-01 02:40:51 +04:00
( and NTLM case ) */
capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS |
CAP_LARGE_WRITE_X | CAP_LARGE_READ_X ;
2011-05-27 08:34:02 +04:00
if ( ses - > server - > sec_mode &
2007-07-07 23:25:05 +04:00
( SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED ) )
2006-06-01 02:40:51 +04:00
pSMB - > req . hdr . Flags2 | = SMBFLG2_SECURITY_SIGNATURE ;
if ( ses - > capabilities & CAP_UNICODE ) {
pSMB - > req . hdr . Flags2 | = SMBFLG2_UNICODE ;
capabilities | = CAP_UNICODE ;
}
if ( ses - > capabilities & CAP_STATUS32 ) {
pSMB - > req . hdr . Flags2 | = SMBFLG2_ERR_STATUS ;
capabilities | = CAP_STATUS32 ;
}
if ( ses - > capabilities & CAP_DFS ) {
pSMB - > req . hdr . Flags2 | = SMBFLG2_DFS ;
capabilities | = CAP_DFS ;
}
2007-08-31 02:09:15 +04:00
if ( ses - > capabilities & CAP_UNIX )
2006-06-01 02:40:51 +04:00
capabilities | = CAP_UNIX ;
return capabilities ;
}
2007-10-16 21:32:19 +04:00
static void
unicode_oslm_strings ( char * * pbcc_area , const struct nls_table * nls_cp )
{
char * bcc_ptr = * pbcc_area ;
int bytes_ret = 0 ;
/* Copy OS version */
2012-01-19 08:32:33 +04:00
bytes_ret = cifs_strtoUTF16 ( ( __le16 * ) bcc_ptr , " Linux version " , 32 ,
nls_cp ) ;
2007-10-16 21:32:19 +04:00
bcc_ptr + = 2 * bytes_ret ;
2012-01-19 08:32:33 +04:00
bytes_ret = cifs_strtoUTF16 ( ( __le16 * ) bcc_ptr , init_utsname ( ) - > release ,
32 , nls_cp ) ;
2007-10-16 21:32:19 +04:00
bcc_ptr + = 2 * bytes_ret ;
bcc_ptr + = 2 ; /* trailing null */
2012-01-19 08:32:33 +04:00
bytes_ret = cifs_strtoUTF16 ( ( __le16 * ) bcc_ptr , CIFS_NETWORK_OPSYS ,
32 , nls_cp ) ;
2007-10-16 21:32:19 +04:00
bcc_ptr + = 2 * bytes_ret ;
bcc_ptr + = 2 ; /* trailing null */
* pbcc_area = bcc_ptr ;
}
2011-05-27 08:34:02 +04:00
static void unicode_domain_string ( char * * pbcc_area , struct cifs_ses * ses ,
2007-10-16 21:32:19 +04:00
const struct nls_table * nls_cp )
{
char * bcc_ptr = * pbcc_area ;
int bytes_ret = 0 ;
/* copy domain */
if ( ses - > domainName = = NULL ) {
/* Sending null domain better than using a bogus domain name (as
we did briefly in 2.6 .18 ) since server will use its default */
* bcc_ptr = 0 ;
* ( bcc_ptr + 1 ) = 0 ;
bytes_ret = 0 ;
} else
2012-01-19 08:32:33 +04:00
bytes_ret = cifs_strtoUTF16 ( ( __le16 * ) bcc_ptr , ses - > domainName ,
256 , nls_cp ) ;
2007-10-16 21:32:19 +04:00
bcc_ptr + = 2 * bytes_ret ;
bcc_ptr + = 2 ; /* account for null terminator */
* pbcc_area = bcc_ptr ;
}
2011-05-27 08:34:02 +04:00
static void unicode_ssetup_strings ( char * * pbcc_area , struct cifs_ses * ses ,
2007-07-07 23:25:05 +04:00
const struct nls_table * nls_cp )
2006-06-01 02:40:51 +04:00
{
2007-07-07 23:25:05 +04:00
char * bcc_ptr = * pbcc_area ;
2006-06-01 02:40:51 +04:00
int bytes_ret = 0 ;
/* BB FIXME add check that strings total less
than 335 or will need to send them as arrays */
2006-06-27 23:50:57 +04:00
/* unicode strings, must be word aligned before the call */
/* if ((long) bcc_ptr % 2) {
2006-06-01 02:40:51 +04:00
* bcc_ptr = 0 ;
bcc_ptr + + ;
2006-06-27 23:50:57 +04:00
} */
2006-06-01 02:40:51 +04:00
/* copy user */
2011-02-25 10:11:56 +03:00
if ( ses - > user_name = = NULL ) {
2006-11-09 02:10:46 +03:00
/* null user mount */
* bcc_ptr = 0 ;
* ( bcc_ptr + 1 ) = 0 ;
2010-02-06 10:08:53 +03:00
} else {
2012-01-19 08:32:33 +04:00
bytes_ret = cifs_strtoUTF16 ( ( __le16 * ) bcc_ptr , ses - > user_name ,
MAX_USERNAME_SIZE , nls_cp ) ;
2006-06-01 02:40:51 +04:00
}
bcc_ptr + = 2 * bytes_ret ;
bcc_ptr + = 2 ; /* account for null termination */
2007-10-16 21:32:19 +04:00
unicode_domain_string ( & bcc_ptr , ses , nls_cp ) ;
unicode_oslm_strings ( & bcc_ptr , nls_cp ) ;
2006-06-01 02:40:51 +04:00
* pbcc_area = bcc_ptr ;
}
2011-05-27 08:34:02 +04:00
static void ascii_ssetup_strings ( char * * pbcc_area , struct cifs_ses * ses ,
2007-07-07 23:25:05 +04:00
const struct nls_table * nls_cp )
2006-06-01 02:40:51 +04:00
{
2007-07-07 23:25:05 +04:00
char * bcc_ptr = * pbcc_area ;
2006-06-01 02:40:51 +04:00
/* copy user */
/* BB what about null user mounts - check that we do this BB */
2007-07-07 23:25:05 +04:00
/* copy user */
2012-02-03 01:28:28 +04:00
if ( ses - > user_name ! = NULL ) {
2011-02-25 10:11:56 +03:00
strncpy ( bcc_ptr , ses - > user_name , MAX_USERNAME_SIZE ) ;
2012-02-03 01:28:28 +04:00
bcc_ptr + = strnlen ( ses - > user_name , MAX_USERNAME_SIZE ) ;
}
2011-02-25 10:11:56 +03:00
/* else null user mount */
2006-06-01 02:40:51 +04:00
* bcc_ptr = 0 ;
2007-07-07 23:25:05 +04:00
bcc_ptr + + ; /* account for null termination */
2006-06-01 02:40:51 +04:00
2007-07-07 23:25:05 +04:00
/* copy domain */
if ( ses - > domainName ! = NULL ) {
strncpy ( bcc_ptr , ses - > domainName , 256 ) ;
2006-06-01 02:40:51 +04:00
bcc_ptr + = strnlen ( ses - > domainName , 256 ) ;
2007-07-07 23:25:05 +04:00
} /* else we will send a null domain name
2006-11-09 02:10:46 +03:00
so the server will default to its own domain */
2006-06-01 02:40:51 +04:00
* bcc_ptr = 0 ;
bcc_ptr + + ;
/* BB check for overflow here */
strcpy ( bcc_ptr , " Linux version " ) ;
bcc_ptr + = strlen ( " Linux version " ) ;
2006-10-02 13:18:13 +04:00
strcpy ( bcc_ptr , init_utsname ( ) - > release ) ;
bcc_ptr + = strlen ( init_utsname ( ) - > release ) + 1 ;
2006-06-01 02:40:51 +04:00
strcpy ( bcc_ptr , CIFS_NETWORK_OPSYS ) ;
bcc_ptr + = strlen ( CIFS_NETWORK_OPSYS ) + 1 ;
2007-07-07 23:25:05 +04:00
* pbcc_area = bcc_ptr ;
2006-06-01 02:40:51 +04:00
}
2009-04-30 15:16:21 +04:00
static void
2011-05-27 08:34:02 +04:00
decode_unicode_ssetup ( char * * pbcc_area , int bleft , struct cifs_ses * ses ,
2009-04-30 15:16:21 +04:00
const struct nls_table * nls_cp )
2006-06-01 02:40:51 +04:00
{
2009-04-30 15:16:21 +04:00
int len ;
2007-07-07 23:25:05 +04:00
char * data = * pbcc_area ;
2006-06-01 02:40:51 +04:00
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " bleft %d " , bleft ) ;
2006-06-01 02:40:51 +04:00
2007-08-31 02:09:15 +04:00
kfree ( ses - > serverOS ) ;
2012-01-19 08:32:33 +04:00
ses - > serverOS = cifs_strndup_from_utf16 ( data , bleft , true , nls_cp ) ;
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " serverOS=%s " , ses - > serverOS ) ;
2009-04-30 15:16:21 +04:00
len = ( UniStrnlen ( ( wchar_t * ) data , bleft / 2 ) * 2 ) + 2 ;
data + = len ;
bleft - = len ;
if ( bleft < = 0 )
return ;
2006-06-01 02:40:51 +04:00
2007-08-31 02:09:15 +04:00
kfree ( ses - > serverNOS ) ;
2012-01-19 08:32:33 +04:00
ses - > serverNOS = cifs_strndup_from_utf16 ( data , bleft , true , nls_cp ) ;
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " serverNOS=%s " , ses - > serverNOS ) ;
2009-04-30 15:16:21 +04:00
len = ( UniStrnlen ( ( wchar_t * ) data , bleft / 2 ) * 2 ) + 2 ;
data + = len ;
bleft - = len ;
if ( bleft < = 0 )
return ;
2007-07-07 23:25:05 +04:00
2007-08-31 02:09:15 +04:00
kfree ( ses - > serverDomain ) ;
2012-01-19 08:32:33 +04:00
ses - > serverDomain = cifs_strndup_from_utf16 ( data , bleft , true , nls_cp ) ;
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " serverDomain=%s " , ses - > serverDomain ) ;
2007-07-07 23:25:05 +04:00
2009-04-30 15:16:21 +04:00
return ;
2006-06-01 02:40:51 +04:00
}
2011-01-20 21:36:51 +03:00
static int decode_ascii_ssetup ( char * * pbcc_area , __u16 bleft ,
2011-05-27 08:34:02 +04:00
struct cifs_ses * ses ,
2007-07-07 23:25:05 +04:00
const struct nls_table * nls_cp )
2006-06-01 02:40:51 +04:00
{
int rc = 0 ;
int len ;
2007-07-07 23:25:05 +04:00
char * bcc_ptr = * pbcc_area ;
2006-06-01 02:40:51 +04:00
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " decode sessetup ascii. bleft %d " , bleft ) ;
2007-07-13 04:33:32 +04:00
2006-06-01 02:40:51 +04:00
len = strnlen ( bcc_ptr , bleft ) ;
2007-07-07 23:25:05 +04:00
if ( len > = bleft )
2006-06-01 02:40:51 +04:00
return rc ;
2007-07-13 04:33:32 +04:00
2007-08-31 02:09:15 +04:00
kfree ( ses - > serverOS ) ;
2006-06-01 02:40:51 +04:00
ses - > serverOS = kzalloc ( len + 1 , GFP_KERNEL ) ;
2007-07-07 23:25:05 +04:00
if ( ses - > serverOS )
2006-06-01 02:40:51 +04:00
strncpy ( ses - > serverOS , bcc_ptr , len ) ;
2007-07-07 23:25:05 +04:00
if ( strncmp ( ses - > serverOS , " OS/2 " , 4 ) = = 0 ) {
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " OS/2 server " ) ;
2006-09-30 08:13:17 +04:00
ses - > flags | = CIFS_SES_OS2 ;
}
2006-06-01 02:40:51 +04:00
bcc_ptr + = len + 1 ;
bleft - = len + 1 ;
len = strnlen ( bcc_ptr , bleft ) ;
2007-07-07 23:25:05 +04:00
if ( len > = bleft )
2006-06-01 02:40:51 +04:00
return rc ;
2007-08-31 02:09:15 +04:00
kfree ( ses - > serverNOS ) ;
2006-06-01 02:40:51 +04:00
ses - > serverNOS = kzalloc ( len + 1 , GFP_KERNEL ) ;
2007-07-07 23:25:05 +04:00
if ( ses - > serverNOS )
2006-06-01 02:40:51 +04:00
strncpy ( ses - > serverNOS , bcc_ptr , len ) ;
bcc_ptr + = len + 1 ;
bleft - = len + 1 ;
2007-07-07 23:25:05 +04:00
len = strnlen ( bcc_ptr , bleft ) ;
if ( len > bleft )
return rc ;
2006-06-01 02:40:51 +04:00
2006-09-30 08:13:17 +04:00
/* No domain field in LANMAN case. Domain is
returned by old servers in the SMB negprot response */
/* BB For newer servers which do not support Unicode,
but thus do return domain here we could add parsing
for it later , but it is not very important */
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " ascii: bytes left %d " , bleft ) ;
2006-06-01 02:40:51 +04:00
return rc ;
}
2009-05-04 12:37:12 +04:00
static int decode_ntlmssp_challenge ( char * bcc_ptr , int blob_len ,
2011-05-27 08:34:02 +04:00
struct cifs_ses * ses )
2009-05-04 12:37:12 +04:00
{
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
unsigned int tioffset ; /* challenge message target info area */
unsigned int tilen ; /* challenge message target info area length */
2009-05-04 12:37:12 +04:00
CHALLENGE_MESSAGE * pblob = ( CHALLENGE_MESSAGE * ) bcc_ptr ;
if ( blob_len < sizeof ( CHALLENGE_MESSAGE ) ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " challenge blob len %d too small " , blob_len ) ;
2009-05-04 12:37:12 +04:00
return - EINVAL ;
}
if ( memcmp ( pblob - > Signature , " NTLMSSP " , 8 ) ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " blob signature incorrect %s " , pblob - > Signature ) ;
2009-05-04 12:37:12 +04:00
return - EINVAL ;
}
if ( pblob - > MessageType ! = NtLmChallenge ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " Incorrect message type %d " , pblob - > MessageType ) ;
2009-05-04 12:37:12 +04:00
return - EINVAL ;
}
2010-10-28 18:53:07 +04:00
memcpy ( ses - > ntlmssp - > cryptkey , pblob - > Challenge , CIFS_CRYPTO_KEY_SIZE ) ;
2009-05-04 12:37:12 +04:00
/* BB we could decode pblob->NegotiateFlags; some may be useful */
/* In particular we can examine sign flags */
/* BB spec says that if AvId field of MsvAvTimestamp is populated then
we must set the MIC field of the AUTHENTICATE_MESSAGE */
2010-10-28 18:53:07 +04:00
ses - > ntlmssp - > server_flags = le32_to_cpu ( pblob - > NegotiateFlags ) ;
2011-03-13 08:08:25 +03:00
tioffset = le32_to_cpu ( pblob - > TargetInfoArray . BufferOffset ) ;
tilen = le16_to_cpu ( pblob - > TargetInfoArray . Length ) ;
2012-01-31 12:52:01 +04:00
if ( tioffset > blob_len | | tioffset + tilen > blob_len ) {
cERROR ( 1 , " tioffset + tilen too high %u + %u " , tioffset , tilen ) ;
return - EINVAL ;
}
2010-10-28 18:53:07 +04:00
if ( tilen ) {
ses - > auth_key . response = kmalloc ( tilen , GFP_KERNEL ) ;
if ( ! ses - > auth_key . response ) {
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
cERROR ( 1 , " Challenge target info allocation failure " ) ;
return - ENOMEM ;
}
2010-10-28 18:53:07 +04:00
memcpy ( ses - > auth_key . response , bcc_ptr + tioffset , tilen ) ;
ses - > auth_key . len = tilen ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
}
2009-05-04 12:37:12 +04:00
return 0 ;
}
/* BB Move to ntlmssp.c eventually */
/* We do not malloc the blob, it is passed in pbuffer, because
it is fixed size , and small , making this approach cleaner */
static void build_ntlmssp_negotiate_blob ( unsigned char * pbuffer ,
2011-05-27 08:34:02 +04:00
struct cifs_ses * ses )
2009-05-04 12:37:12 +04:00
{
NEGOTIATE_MESSAGE * sec_blob = ( NEGOTIATE_MESSAGE * ) pbuffer ;
__u32 flags ;
2010-12-11 23:19:22 +03:00
memset ( pbuffer , 0 , sizeof ( NEGOTIATE_MESSAGE ) ) ;
2009-05-04 12:37:12 +04:00
memcpy ( sec_blob - > Signature , NTLMSSP_SIGNATURE , 8 ) ;
sec_blob - > MessageType = NtLmNegotiate ;
/* BB is NTLMV2 session security format easier to use here? */
flags = NTLMSSP_NEGOTIATE_56 | NTLMSSP_REQUEST_TARGET |
NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE |
2010-12-11 23:19:22 +03:00
NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC ;
2011-05-27 08:34:02 +04:00
if ( ses - > server - > sec_mode &
2010-10-21 23:25:08 +04:00
( SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED ) ) {
2010-09-09 01:09:27 +04:00
flags | = NTLMSSP_NEGOTIATE_SIGN ;
2010-10-21 23:25:08 +04:00
if ( ! ses - > server - > session_estab )
2011-07-10 15:55:32 +04:00
flags | = NTLMSSP_NEGOTIATE_KEY_XCH ;
2010-10-21 23:25:08 +04:00
}
2009-05-04 12:37:12 +04:00
2010-12-11 23:19:22 +03:00
sec_blob - > NegotiateFlags = cpu_to_le32 ( flags ) ;
2009-05-04 12:37:12 +04:00
sec_blob - > WorkstationName . BufferOffset = 0 ;
sec_blob - > WorkstationName . Length = 0 ;
sec_blob - > WorkstationName . MaximumLength = 0 ;
/* Domain name is sent on the Challenge not Negotiate NTLMSSP request */
sec_blob - > DomainName . BufferOffset = 0 ;
sec_blob - > DomainName . Length = 0 ;
sec_blob - > DomainName . MaximumLength = 0 ;
}
/* We do not malloc the blob, it is passed in pbuffer, because its
maximum possible size is fixed and small , making this approach cleaner .
This function returns the length of the data in the blob */
static int build_ntlmssp_auth_blob ( unsigned char * pbuffer ,
2010-10-19 20:47:52 +04:00
u16 * buflen ,
2011-05-27 08:34:02 +04:00
struct cifs_ses * ses ,
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
const struct nls_table * nls_cp )
2009-05-04 12:37:12 +04:00
{
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
int rc ;
2009-05-04 12:37:12 +04:00
AUTHENTICATE_MESSAGE * sec_blob = ( AUTHENTICATE_MESSAGE * ) pbuffer ;
__u32 flags ;
unsigned char * tmp ;
memcpy ( sec_blob - > Signature , NTLMSSP_SIGNATURE , 8 ) ;
sec_blob - > MessageType = NtLmAuthenticate ;
flags = NTLMSSP_NEGOTIATE_56 |
NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_TARGET_INFO |
NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE |
2010-12-11 23:19:22 +03:00
NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC ;
2011-05-27 08:34:02 +04:00
if ( ses - > server - > sec_mode &
2011-07-10 15:55:32 +04:00
( SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED ) ) {
2009-05-04 12:37:12 +04:00
flags | = NTLMSSP_NEGOTIATE_SIGN ;
2011-07-10 15:55:32 +04:00
if ( ! ses - > server - > session_estab )
flags | = NTLMSSP_NEGOTIATE_KEY_XCH ;
}
2009-05-04 12:37:12 +04:00
tmp = pbuffer + sizeof ( AUTHENTICATE_MESSAGE ) ;
2010-12-11 23:19:22 +03:00
sec_blob - > NegotiateFlags = cpu_to_le32 ( flags ) ;
2009-05-04 12:37:12 +04:00
sec_blob - > LmChallengeResponse . BufferOffset =
cpu_to_le32 ( sizeof ( AUTHENTICATE_MESSAGE ) ) ;
sec_blob - > LmChallengeResponse . Length = 0 ;
sec_blob - > LmChallengeResponse . MaximumLength = 0 ;
2010-09-09 01:10:58 +04:00
sec_blob - > NtChallengeResponse . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
2010-10-21 15:42:55 +04:00
rc = setup_ntlmv2_rsp ( ses , nls_cp ) ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
if ( rc ) {
cERROR ( 1 , " Error %d during NTLMSSP authentication " , rc ) ;
goto setup_ntlmv2_ret ;
}
2010-10-21 15:42:55 +04:00
memcpy ( tmp , ses - > auth_key . response + CIFS_SESS_KEY_SIZE ,
ses - > auth_key . len - CIFS_SESS_KEY_SIZE ) ;
tmp + = ses - > auth_key . len - CIFS_SESS_KEY_SIZE ;
2010-09-09 01:10:58 +04:00
2010-10-21 15:42:55 +04:00
sec_blob - > NtChallengeResponse . Length =
cpu_to_le16 ( ses - > auth_key . len - CIFS_SESS_KEY_SIZE ) ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
sec_blob - > NtChallengeResponse . MaximumLength =
2010-10-21 15:42:55 +04:00
cpu_to_le16 ( ses - > auth_key . len - CIFS_SESS_KEY_SIZE ) ;
2009-05-04 12:37:12 +04:00
if ( ses - > domainName = = NULL ) {
sec_blob - > DomainName . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > DomainName . Length = 0 ;
sec_blob - > DomainName . MaximumLength = 0 ;
tmp + = 2 ;
} else {
int len ;
2012-01-19 08:32:33 +04:00
len = cifs_strtoUTF16 ( ( __le16 * ) tmp , ses - > domainName ,
MAX_USERNAME_SIZE , nls_cp ) ;
2009-05-04 12:37:12 +04:00
len * = 2 ; /* unicode is 2 bytes each */
sec_blob - > DomainName . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > DomainName . Length = cpu_to_le16 ( len ) ;
sec_blob - > DomainName . MaximumLength = cpu_to_le16 ( len ) ;
tmp + = len ;
}
2011-02-25 10:11:56 +03:00
if ( ses - > user_name = = NULL ) {
2009-05-04 12:37:12 +04:00
sec_blob - > UserName . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > UserName . Length = 0 ;
sec_blob - > UserName . MaximumLength = 0 ;
tmp + = 2 ;
} else {
int len ;
2012-01-19 08:32:33 +04:00
len = cifs_strtoUTF16 ( ( __le16 * ) tmp , ses - > user_name ,
MAX_USERNAME_SIZE , nls_cp ) ;
2009-05-04 12:37:12 +04:00
len * = 2 ; /* unicode is 2 bytes each */
sec_blob - > UserName . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > UserName . Length = cpu_to_le16 ( len ) ;
sec_blob - > UserName . MaximumLength = cpu_to_le16 ( len ) ;
tmp + = len ;
}
sec_blob - > WorkstationName . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > WorkstationName . Length = 0 ;
sec_blob - > WorkstationName . MaximumLength = 0 ;
tmp + = 2 ;
2010-12-11 23:19:22 +03:00
if ( ( ( ses - > ntlmssp - > server_flags & NTLMSSP_NEGOTIATE_KEY_XCH ) | |
( ses - > ntlmssp - > server_flags & NTLMSSP_NEGOTIATE_EXTENDED_SEC ) )
& & ! calc_seckey ( ses ) ) {
2010-10-28 18:53:07 +04:00
memcpy ( tmp , ses - > ntlmssp - > ciphertext , CIFS_CPHTXT_SIZE ) ;
2010-10-21 23:25:08 +04:00
sec_blob - > SessionKey . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > SessionKey . Length = cpu_to_le16 ( CIFS_CPHTXT_SIZE ) ;
sec_blob - > SessionKey . MaximumLength =
cpu_to_le16 ( CIFS_CPHTXT_SIZE ) ;
tmp + = CIFS_CPHTXT_SIZE ;
} else {
sec_blob - > SessionKey . BufferOffset = cpu_to_le32 ( tmp - pbuffer ) ;
sec_blob - > SessionKey . Length = 0 ;
sec_blob - > SessionKey . MaximumLength = 0 ;
}
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
setup_ntlmv2_ret :
2010-10-19 20:47:52 +04:00
* buflen = tmp - pbuffer ;
return rc ;
2009-05-04 12:37:12 +04:00
}
2007-07-07 23:25:05 +04:00
int
2011-05-27 08:34:02 +04:00
CIFS_SessSetup ( unsigned int xid , struct cifs_ses * ses ,
2010-04-24 15:57:47 +04:00
const struct nls_table * nls_cp )
2006-06-01 02:40:51 +04:00
{
int rc = 0 ;
int wct ;
struct smb_hdr * smb_buf ;
char * bcc_ptr ;
2006-06-27 10:28:30 +04:00
char * str_area ;
2006-06-01 02:40:51 +04:00
SESSION_SETUP_ANDX * pSMB ;
__u32 capabilities ;
2011-01-20 21:36:51 +03:00
__u16 count ;
2007-11-17 02:37:35 +03:00
int resp_buf_type ;
struct kvec iov [ 3 ] ;
2006-06-01 02:40:51 +04:00
enum securityEnum type ;
2011-01-20 21:36:51 +03:00
__u16 action , bytes_remaining ;
2007-11-17 02:37:35 +03:00
struct key * spnego_key = NULL ;
2009-05-04 12:37:12 +04:00
__le32 phase = NtLmNegotiate ; /* NTLMSSP, if needed, is multistage */
2010-10-19 20:47:52 +04:00
u16 blob_len ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
char * ntlmsspblob = NULL ;
2006-06-04 09:53:15 +04:00
2007-07-07 23:25:05 +04:00
if ( ses = = NULL )
2006-06-01 02:40:51 +04:00
return - EINVAL ;
type = ses - > server - > secType ;
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " sess setup type %d " , type ) ;
2010-10-28 18:53:07 +04:00
if ( type = = RawNTLMSSP ) {
/* if memory allocation is successful, caller of this function
* frees it .
*/
ses - > ntlmssp = kmalloc ( sizeof ( struct ntlmssp_auth ) , GFP_KERNEL ) ;
if ( ! ses - > ntlmssp )
return - ENOMEM ;
}
2009-05-04 12:37:12 +04:00
ssetup_ntlmssp_authenticate :
if ( phase = = NtLmChallenge )
phase = NtLmAuthenticate ; /* if ntlmssp, now final phase */
2007-07-07 23:25:05 +04:00
if ( type = = LANMAN ) {
2006-06-01 02:40:51 +04:00
# ifndef CONFIG_CIFS_WEAK_PW_HASH
/* LANMAN and plaintext are less secure and off by default.
So we make this explicitly be turned on in kconfig ( in the
build ) and turned on at runtime ( changed from the default )
in proc / fs / cifs or via mount parm . Unfortunately this is
needed for old Win ( e . g . Win95 ) , some obscure NAS and OS / 2 */
return - EOPNOTSUPP ;
# endif
wct = 10 ; /* lanman 2 style sessionsetup */
2007-07-07 23:25:05 +04:00
} else if ( ( type = = NTLM ) | | ( type = = NTLMv2 ) ) {
2006-06-05 02:21:07 +04:00
/* For NTLMv2 failures eventually may need to retry NTLM */
2006-06-01 02:40:51 +04:00
wct = 13 ; /* old style NTLM sessionsetup */
2007-07-07 23:25:05 +04:00
} else /* same size: negotiate or auth, NTLMSSP or extended security */
2006-06-01 02:40:51 +04:00
wct = 12 ;
rc = small_smb_init_no_tc ( SMB_COM_SESSION_SETUP_ANDX , wct , ses ,
( void * * ) & smb_buf ) ;
2007-07-07 23:25:05 +04:00
if ( rc )
2006-06-01 02:40:51 +04:00
return rc ;
pSMB = ( SESSION_SETUP_ANDX * ) smb_buf ;
capabilities = cifs_ssetup_hdr ( ses , pSMB ) ;
2006-06-27 10:28:30 +04:00
2007-11-17 02:37:35 +03:00
/* we will send the SMB in three pieces:
a fixed length beginning part , an optional
SPNEGO blob ( which can be zero length ) , and a
last part which will include the strings
and rest of bcc area . This allows us to avoid
a large buffer 17 K allocation */
2007-07-07 23:25:05 +04:00
iov [ 0 ] . iov_base = ( char * ) pSMB ;
2011-04-29 09:40:20 +04:00
iov [ 0 ] . iov_len = be32_to_cpu ( smb_buf - > smb_buf_length ) + 4 ;
2006-06-27 10:28:30 +04:00
2007-11-17 02:37:35 +03:00
/* setting this here allows the code at the end of the function
to free the request buffer if there ' s an error */
resp_buf_type = CIFS_SMALL_BUFFER ;
2006-06-27 10:28:30 +04:00
/* 2000 big enough to fit max user, domain, NOS name etc. */
str_area = kmalloc ( 2000 , GFP_KERNEL ) ;
2007-08-18 04:15:20 +04:00
if ( str_area = = NULL ) {
2007-11-17 02:37:35 +03:00
rc = - ENOMEM ;
goto ssetup_exit ;
2007-08-18 04:15:20 +04:00
}
2006-06-27 10:28:30 +04:00
bcc_ptr = str_area ;
2006-06-01 02:40:51 +04:00
2006-09-30 08:13:17 +04:00
ses - > flags & = ~ CIFS_SES_LANMAN ;
2007-11-17 02:37:35 +03:00
iov [ 1 ] . iov_base = NULL ;
iov [ 1 ] . iov_len = 0 ;
2007-07-07 23:25:05 +04:00
if ( type = = LANMAN ) {
2006-06-01 02:40:51 +04:00
# ifdef CONFIG_CIFS_WEAK_PW_HASH
2011-02-17 23:38:31 +03:00
char lnm_session_key [ CIFS_AUTH_RESP_SIZE ] ;
2006-06-01 02:40:51 +04:00
2008-08-28 19:32:22 +04:00
pSMB - > req . hdr . Flags2 & = ~ SMBFLG2_UNICODE ;
2006-06-01 02:40:51 +04:00
/* no capabilities flags in old lanman negotiation */
2011-02-17 23:38:31 +03:00
pSMB - > old_req . PasswordLength = cpu_to_le16 ( CIFS_AUTH_RESP_SIZE ) ;
2006-06-01 02:40:51 +04:00
2010-10-28 00:20:36 +04:00
/* Calculate hash with password and copy into bcc_ptr.
* Encryption Key ( stored as in cryptkey ) gets used if the
* security mode bit in Negottiate Protocol response states
* to use challenge / response method ( i . e . Password bit is 1 ) .
*/
2011-04-19 22:23:31 +04:00
rc = calc_lanman_hash ( ses - > password , ses - > server - > cryptkey ,
2011-05-27 08:34:02 +04:00
ses - > server - > sec_mode & SECMODE_PW_ENCRYPT ?
2008-12-06 04:41:21 +03:00
true : false , lnm_session_key ) ;
2007-07-07 23:25:05 +04:00
ses - > flags | = CIFS_SES_LANMAN ;
2011-02-17 23:38:31 +03:00
memcpy ( bcc_ptr , ( char * ) lnm_session_key , CIFS_AUTH_RESP_SIZE ) ;
bcc_ptr + = CIFS_AUTH_RESP_SIZE ;
2006-06-01 02:40:51 +04:00
/* can not sign if LANMAN negotiated so no need
to calculate signing key ? but what if server
changed to do higher than lanman dialect and
we reconnected would we ever calc signing_key ? */
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " Negotiating LANMAN setting up strings " ) ;
2006-06-01 02:40:51 +04:00
/* Unicode not allowed for LANMAN dialects */
ascii_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
2007-07-07 23:25:05 +04:00
# endif
2006-06-01 02:40:51 +04:00
} else if ( type = = NTLM ) {
pSMB - > req_no_secext . Capabilities = cpu_to_le32 ( capabilities ) ;
pSMB - > req_no_secext . CaseInsensitivePasswordLength =
2010-10-21 15:42:55 +04:00
cpu_to_le16 ( CIFS_AUTH_RESP_SIZE ) ;
2006-06-01 02:40:51 +04:00
pSMB - > req_no_secext . CaseSensitivePasswordLength =
2010-10-21 15:42:55 +04:00
cpu_to_le16 ( CIFS_AUTH_RESP_SIZE ) ;
/* calculate ntlm response and session key */
2011-10-20 22:21:59 +04:00
rc = setup_ntlm_response ( ses , nls_cp ) ;
2010-10-21 15:42:55 +04:00
if ( rc ) {
cERROR ( 1 , " Error %d during NTLM authentication " , rc ) ;
goto ssetup_exit ;
}
2007-07-13 04:33:32 +04:00
2010-10-21 15:42:55 +04:00
/* copy ntlm response */
memcpy ( bcc_ptr , ses - > auth_key . response + CIFS_SESS_KEY_SIZE ,
CIFS_AUTH_RESP_SIZE ) ;
bcc_ptr + = CIFS_AUTH_RESP_SIZE ;
memcpy ( bcc_ptr , ses - > auth_key . response + CIFS_SESS_KEY_SIZE ,
CIFS_AUTH_RESP_SIZE ) ;
bcc_ptr + = CIFS_AUTH_RESP_SIZE ;
2006-06-01 02:40:51 +04:00
2007-07-07 23:25:05 +04:00
if ( ses - > capabilities & CAP_UNICODE ) {
2006-06-27 23:50:57 +04:00
/* unicode strings must be word aligned */
if ( iov [ 0 ] . iov_len % 2 ) {
* bcc_ptr = 0 ;
2007-07-07 23:25:05 +04:00
bcc_ptr + + ;
}
2006-06-01 23:20:10 +04:00
unicode_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
2006-06-27 23:50:57 +04:00
} else
2006-06-01 23:20:10 +04:00
ascii_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
} else if ( type = = NTLMv2 ) {
pSMB - > req_no_secext . Capabilities = cpu_to_le32 ( capabilities ) ;
/* LM2 password would be here if we supported it */
pSMB - > req_no_secext . CaseInsensitivePasswordLength = 0 ;
2010-10-21 15:42:55 +04:00
/* calculate nlmv2 response and session key */
rc = setup_ntlmv2_rsp ( ses , nls_cp ) ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
if ( rc ) {
cERROR ( 1 , " Error %d during NTLMv2 authentication " , rc ) ;
goto ssetup_exit ;
}
2010-10-21 15:42:55 +04:00
memcpy ( bcc_ptr , ses - > auth_key . response + CIFS_SESS_KEY_SIZE ,
ses - > auth_key . len - CIFS_SESS_KEY_SIZE ) ;
bcc_ptr + = ses - > auth_key . len - CIFS_SESS_KEY_SIZE ;
2010-10-05 04:56:13 +04:00
/* set case sensitive password length after tilen may get
* assigned , tilen is 0 otherwise .
*/
pSMB - > req_no_secext . CaseSensitivePasswordLength =
2010-10-27 03:10:24 +04:00
cpu_to_le16 ( ses - > auth_key . len - CIFS_SESS_KEY_SIZE ) ;
2010-10-05 04:56:13 +04:00
2007-07-07 23:25:05 +04:00
if ( ses - > capabilities & CAP_UNICODE ) {
if ( iov [ 0 ] . iov_len % 2 ) {
2006-06-27 23:50:57 +04:00
* bcc_ptr = 0 ;
2007-08-31 02:09:15 +04:00
bcc_ptr + + ;
}
2006-06-01 02:40:51 +04:00
unicode_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
2006-06-27 23:50:57 +04:00
} else
2006-06-01 02:40:51 +04:00
ascii_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
2010-04-24 15:57:49 +04:00
} else if ( type = = Kerberos ) {
2007-11-17 02:37:35 +03:00
# ifdef CONFIG_CIFS_UPCALL
struct cifs_spnego_msg * msg ;
2010-10-21 15:42:55 +04:00
2007-11-17 02:37:35 +03:00
spnego_key = cifs_get_spnego_key ( ses ) ;
if ( IS_ERR ( spnego_key ) ) {
rc = PTR_ERR ( spnego_key ) ;
spnego_key = NULL ;
goto ssetup_exit ;
}
msg = spnego_key - > payload . data ;
2008-08-26 04:37:14 +04:00
/* check version field to make sure that cifs.upcall is
sending us a response in an expected form */
if ( msg - > version ! = CIFS_SPNEGO_UPCALL_VERSION ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " incorrect version of cifs.upcall (expected "
2008-08-26 04:37:14 +04:00
" %d but got %d) " ,
2010-04-21 07:50:45 +04:00
CIFS_SPNEGO_UPCALL_VERSION , msg - > version ) ;
2008-08-26 04:37:14 +04:00
rc = - EKEYREJECTED ;
goto ssetup_exit ;
}
2010-10-21 15:42:55 +04:00
ses - > auth_key . response = kmalloc ( msg - > sesskey_len , GFP_KERNEL ) ;
if ( ! ses - > auth_key . response ) {
cERROR ( 1 , " Kerberos can't allocate (%u bytes) memory " ,
msg - > sesskey_len ) ;
rc = - ENOMEM ;
2007-11-17 02:37:35 +03:00
goto ssetup_exit ;
}
2010-10-21 15:42:55 +04:00
memcpy ( ses - > auth_key . response , msg - > data , msg - > sesskey_len ) ;
2010-10-14 03:15:00 +04:00
ses - > auth_key . len = msg - > sesskey_len ;
2010-10-21 15:42:55 +04:00
2006-06-01 02:40:51 +04:00
pSMB - > req . hdr . Flags2 | = SMBFLG2_EXT_SEC ;
capabilities | = CAP_EXTENDED_SECURITY ;
pSMB - > req . Capabilities = cpu_to_le32 ( capabilities ) ;
2007-11-17 02:37:35 +03:00
iov [ 1 ] . iov_base = msg - > data + msg - > sesskey_len ;
iov [ 1 ] . iov_len = msg - > secblob_len ;
pSMB - > req . SecurityBlobLength = cpu_to_le16 ( iov [ 1 ] . iov_len ) ;
if ( ses - > capabilities & CAP_UNICODE ) {
/* unicode strings must be word aligned */
2007-12-31 07:56:21 +03:00
if ( ( iov [ 0 ] . iov_len + iov [ 1 ] . iov_len ) % 2 ) {
2007-11-17 02:37:35 +03:00
* bcc_ptr = 0 ;
bcc_ptr + + ;
}
unicode_oslm_strings ( & bcc_ptr , nls_cp ) ;
unicode_domain_string ( & bcc_ptr , ses , nls_cp ) ;
} else
/* BB: is this right? */
ascii_ssetup_strings ( & bcc_ptr , ses , nls_cp ) ;
# else /* ! CONFIG_CIFS_UPCALL */
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " Kerberos negotiated but upcall support disabled! " ) ;
2007-11-17 02:37:35 +03:00
rc = - ENOSYS ;
goto ssetup_exit ;
# endif /* CONFIG_CIFS_UPCALL */
2011-01-07 19:30:28 +03:00
} else if ( type = = RawNTLMSSP ) {
if ( ( pSMB - > req . hdr . Flags2 & SMBFLG2_UNICODE ) = = 0 ) {
cERROR ( 1 , " NTLMSSP requires Unicode support " ) ;
rc = - ENOSYS ;
goto ssetup_exit ;
}
cFYI ( 1 , " ntlmssp session setup phase %d " , phase ) ;
pSMB - > req . hdr . Flags2 | = SMBFLG2_EXT_SEC ;
capabilities | = CAP_EXTENDED_SECURITY ;
pSMB - > req . Capabilities | = cpu_to_le32 ( capabilities ) ;
switch ( phase ) {
case NtLmNegotiate :
build_ntlmssp_negotiate_blob (
pSMB - > req . SecurityBlob , ses ) ;
iov [ 1 ] . iov_len = sizeof ( NEGOTIATE_MESSAGE ) ;
iov [ 1 ] . iov_base = pSMB - > req . SecurityBlob ;
pSMB - > req . SecurityBlobLength =
cpu_to_le16 ( sizeof ( NEGOTIATE_MESSAGE ) ) ;
break ;
case NtLmAuthenticate :
/*
* 5 is an empirical value , large enough to hold
* authenticate message plus max 10 of av paris ,
* domain , user , workstation names , flags , etc .
*/
ntlmsspblob = kzalloc (
5 * sizeof ( struct _AUTHENTICATE_MESSAGE ) ,
GFP_KERNEL ) ;
if ( ! ntlmsspblob ) {
cERROR ( 1 , " Can't allocate NTLMSSP blob " ) ;
rc = - ENOMEM ;
2009-05-04 12:37:12 +04:00
goto ssetup_exit ;
}
2011-01-07 19:30:28 +03:00
rc = build_ntlmssp_auth_blob ( ntlmsspblob ,
& blob_len , ses , nls_cp ) ;
if ( rc )
2009-05-04 12:37:12 +04:00
goto ssetup_exit ;
2011-01-07 19:30:28 +03:00
iov [ 1 ] . iov_len = blob_len ;
iov [ 1 ] . iov_base = ntlmsspblob ;
pSMB - > req . SecurityBlobLength = cpu_to_le16 ( blob_len ) ;
/*
* Make sure that we tell the server that we are using
* the uid that it just gave us back on the response
* ( challenge )
*/
smb_buf - > Uid = ses - > Suid ;
break ;
default :
cERROR ( 1 , " invalid phase %d " , phase ) ;
2009-05-04 12:37:12 +04:00
rc = - ENOSYS ;
goto ssetup_exit ;
}
2011-01-07 19:30:28 +03:00
/* unicode strings must be word aligned */
if ( ( iov [ 0 ] . iov_len + iov [ 1 ] . iov_len ) % 2 ) {
* bcc_ptr = 0 ;
bcc_ptr + + ;
}
unicode_oslm_strings ( & bcc_ptr , nls_cp ) ;
} else {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " secType %d not supported! " , type ) ;
2007-11-17 02:37:35 +03:00
rc = - ENOSYS ;
goto ssetup_exit ;
2006-06-01 02:40:51 +04:00
}
2007-11-17 02:37:35 +03:00
iov [ 2 ] . iov_base = str_area ;
iov [ 2 ] . iov_len = ( long ) bcc_ptr - ( long ) str_area ;
count = iov [ 1 ] . iov_len + iov [ 2 ] . iov_len ;
2011-04-29 09:40:20 +04:00
smb_buf - > smb_buf_length =
cpu_to_be32 ( be32_to_cpu ( smb_buf - > smb_buf_length ) + count ) ;
2006-06-01 02:40:51 +04:00
2011-05-04 16:05:26 +04:00
put_bcc ( count , smb_buf ) ;
2006-06-01 02:40:51 +04:00
2007-11-17 02:37:35 +03:00
rc = SendReceive2 ( xid , ses , iov , 3 /* num_iovecs */ , & resp_buf_type ,
2011-01-11 15:24:23 +03:00
CIFS_LOG_ERROR ) ;
2006-06-01 02:40:51 +04:00
/* SMB request buf freed in SendReceive2 */
pSMB = ( SESSION_SETUP_ANDX * ) iov [ 0 ] . iov_base ;
smb_buf = ( struct smb_hdr * ) iov [ 0 ] . iov_base ;
2009-05-04 12:37:12 +04:00
if ( ( type = = RawNTLMSSP ) & & ( smb_buf - > Status . CifsError = =
cpu_to_le32 ( NT_STATUS_MORE_PROCESSING_REQUIRED ) ) ) {
if ( phase ! = NtLmNegotiate ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " Unexpected more processing error " ) ;
2009-05-04 12:37:12 +04:00
goto ssetup_exit ;
}
/* NTLMSSP Negotiate sent now processing challenge (response) */
phase = NtLmChallenge ; /* process ntlmssp challenge */
rc = 0 ; /* MORE_PROC rc is not an error here, but expected */
}
if ( rc )
goto ssetup_exit ;
2007-07-07 23:25:05 +04:00
if ( ( smb_buf - > WordCount ! = 3 ) & & ( smb_buf - > WordCount ! = 4 ) ) {
2006-06-01 02:40:51 +04:00
rc = - EIO ;
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " bad word count %d " , smb_buf - > WordCount ) ;
2006-06-01 02:40:51 +04:00
goto ssetup_exit ;
}
action = le16_to_cpu ( pSMB - > resp . Action ) ;
if ( action & GUEST_LOGIN )
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " Guest login " ) ; /* BB mark SesInfo struct? */
2006-06-01 02:40:51 +04:00
ses - > Suid = smb_buf - > Uid ; /* UID left in wire format (le) */
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " UID = %d " , ses - > Suid ) ;
2006-06-01 02:40:51 +04:00
/* response can have either 3 or 4 word count - Samba sends 3 */
/* and lanman response is 3 */
2011-01-20 21:36:51 +03:00
bytes_remaining = get_bcc ( smb_buf ) ;
2006-06-01 02:40:51 +04:00
bcc_ptr = pByteArea ( smb_buf ) ;
2007-07-07 23:25:05 +04:00
if ( smb_buf - > WordCount = = 4 ) {
2006-06-01 02:40:51 +04:00
blob_len = le16_to_cpu ( pSMB - > resp . SecurityBlobLength ) ;
2007-07-07 23:25:05 +04:00
if ( blob_len > bytes_remaining ) {
2010-04-21 07:50:45 +04:00
cERROR ( 1 , " bad security blob length %d " , blob_len ) ;
2006-06-01 02:40:51 +04:00
rc = - EINVAL ;
goto ssetup_exit ;
}
2009-05-04 12:37:12 +04:00
if ( phase = = NtLmChallenge ) {
rc = decode_ntlmssp_challenge ( bcc_ptr , blob_len , ses ) ;
/* now goto beginning for ntlmssp authenticate phase */
if ( rc )
goto ssetup_exit ;
}
bcc_ptr + = blob_len ;
2006-06-01 02:40:51 +04:00
bytes_remaining - = blob_len ;
2007-07-07 23:25:05 +04:00
}
2006-06-01 02:40:51 +04:00
/* BB check if Unicode and decode strings */
2011-04-27 21:25:51 +04:00
if ( bytes_remaining = = 0 ) {
/* no string area to decode, do nothing */
} else if ( smb_buf - > Flags2 & SMBFLG2_UNICODE ) {
2009-04-14 19:00:53 +04:00
/* unicode string area must be word-aligned */
if ( ( ( unsigned long ) bcc_ptr - ( unsigned long ) smb_buf ) % 2 ) {
+ + bcc_ptr ;
- - bytes_remaining ;
}
2009-04-30 15:16:21 +04:00
decode_unicode_ssetup ( & bcc_ptr , bytes_remaining , ses , nls_cp ) ;
2009-04-14 19:00:53 +04:00
} else {
2007-07-17 21:34:02 +04:00
rc = decode_ascii_ssetup ( & bcc_ptr , bytes_remaining ,
ses , nls_cp ) ;
2009-04-14 19:00:53 +04:00
}
2007-07-13 04:33:32 +04:00
2006-06-01 02:40:51 +04:00
ssetup_exit :
2008-09-24 19:32:59 +04:00
if ( spnego_key ) {
key_revoke ( spnego_key ) ;
2007-11-17 02:37:35 +03:00
key_put ( spnego_key ) ;
2008-09-24 19:32:59 +04:00
}
2006-06-27 10:28:30 +04:00
kfree ( str_area ) ;
cifs NTLMv2/NTLMSSP ntlmv2 within ntlmssp autentication code
Attribue Value (AV) pairs or Target Info (TI) pairs are part of
ntlmv2 authentication.
Structure ntlmv2_resp had only definition for two av pairs.
So removed it, and now allocation of av pairs is dynamic.
For servers like Windows 7/2008, av pairs sent by server in
challege packet (type 2 in the ntlmssp exchange/negotiation) can
vary.
Server sends them during ntlmssp negotiation. So when ntlmssp is used
as an authentication mechanism, type 2 challenge packet from server
has this information. Pluck it and use the entire blob for
authenticaiton purpose. If user has not specified, extract
(netbios) domain name from the av pairs which is used to calculate
ntlmv2 hash. Servers like Windows 7 are particular about the AV pair
blob.
Servers like Windows 2003, are not very strict about the contents
of av pair blob used during ntlmv2 authentication.
So when security mechanism such as ntlmv2 is used (not ntlmv2 in ntlmssp),
there is no negotiation and so genereate a minimal blob that gets
used in ntlmv2 authentication as well as gets sent.
Fields tilen and tilbob are session specific. AV pair values are defined.
To calculate ntlmv2 response we need ti/av pair blob.
For sec mech like ntlmssp, the blob is plucked from type 2 response from
the server. From this blob, netbios name of the domain is retrieved,
if user has not already provided, to be included in the Target String
as part of ntlmv2 hash calculations.
For sec mech like ntlmv2, create a minimal, two av pair blob.
The allocated blob is freed in case of error. In case there is no error,
this blob is used in calculating ntlmv2 response (in CalcNTLMv2_response)
and is also copied on the response to the server, and then freed.
The type 3 ntlmssp response is prepared on a buffer,
5 * sizeof of struct _AUTHENTICATE_MESSAGE, an empirical value large
enough to hold _AUTHENTICATE_MESSAGE plus a blob with max possible
10 values as part of ntlmv2 response and lmv2 keys and domain, user,
workstation names etc.
Also, kerberos gets selected as a default mechanism if server supports it,
over the other security mechanisms.
Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-09-19 07:02:18 +04:00
kfree ( ntlmsspblob ) ;
ntlmsspblob = NULL ;
2007-07-07 23:25:05 +04:00
if ( resp_buf_type = = CIFS_SMALL_BUFFER ) {
2010-04-21 07:50:45 +04:00
cFYI ( 1 , " ssetup freeing small buf %p " , iov [ 0 ] . iov_base ) ;
2006-06-01 02:40:51 +04:00
cifs_small_buf_release ( iov [ 0 ] . iov_base ) ;
2007-07-07 23:25:05 +04:00
} else if ( resp_buf_type = = CIFS_LARGE_BUFFER )
2006-06-01 02:40:51 +04:00
cifs_buf_release ( iov [ 0 ] . iov_base ) ;
2009-05-04 12:37:12 +04:00
/* if ntlmssp, and negotiate succeeded, proceed to authenticate phase */
if ( ( phase = = NtLmChallenge ) & & ( rc = = 0 ) )
goto ssetup_ntlmssp_authenticate ;
2006-06-01 02:40:51 +04:00
return rc ;
}