linux/fs/cifs/sess.c

1754 lines
47 KiB
C
Raw Normal View History

/*
* fs/cifs/sess.c
*
* SMB/CIFS session setup handling routines
*
* Copyright (c) International Business Machines Corp., 2006, 2009
* 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"
#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 16:04:11 +08:00
#include <linux/slab.h>
#include "cifs_spnego.h"
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
#include "smb2proto.h"
#include "fs_context.h"
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
static int
cifs_ses_add_channel(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses,
struct cifs_server_iface *iface);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
bool
is_server_using_iface(struct TCP_Server_Info *server,
struct cifs_server_iface *iface)
{
struct sockaddr_in *i4 = (struct sockaddr_in *)&iface->sockaddr;
struct sockaddr_in6 *i6 = (struct sockaddr_in6 *)&iface->sockaddr;
struct sockaddr_in *s4 = (struct sockaddr_in *)&server->dstaddr;
struct sockaddr_in6 *s6 = (struct sockaddr_in6 *)&server->dstaddr;
if (server->dstaddr.ss_family != iface->sockaddr.ss_family)
return false;
if (server->dstaddr.ss_family == AF_INET) {
if (s4->sin_addr.s_addr != i4->sin_addr.s_addr)
return false;
} else if (server->dstaddr.ss_family == AF_INET6) {
if (memcmp(&s6->sin6_addr, &i6->sin6_addr,
sizeof(i6->sin6_addr)) != 0)
return false;
} else {
/* unknown family.. */
return false;
}
return true;
}
bool is_ses_using_iface(struct cifs_ses *ses, struct cifs_server_iface *iface)
{
int i;
for (i = 0; i < ses->chan_count; i++) {
if (is_server_using_iface(ses->chans[i].server, iface))
return true;
}
return false;
}
/* returns number of channels added */
int cifs_try_adding_channels(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses)
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
{
int old_chan_count = ses->chan_count;
int left = ses->chan_max - ses->chan_count;
int i = 0;
int rc = 0;
int tries = 0;
struct cifs_server_iface *ifaces = NULL;
size_t iface_count;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
if (left <= 0) {
cifs_dbg(FYI,
"ses already at max_channels (%zu), nothing to open\n",
ses->chan_max);
return 0;
}
if (ses->server->dialect < SMB30_PROT_ID) {
cifs_dbg(VFS, "multichannel is not supported on this protocol version, use 3.0 or above\n");
return 0;
}
/*
* Make a copy of the iface list at the time and use that
* instead so as to not hold the iface spinlock for opening
* channels
*/
spin_lock(&ses->iface_lock);
iface_count = ses->iface_count;
if (iface_count <= 0) {
spin_unlock(&ses->iface_lock);
cifs_dbg(VFS, "no iface list available to open channels\n");
return 0;
}
ifaces = kmemdup(ses->iface_list, iface_count*sizeof(*ifaces),
GFP_ATOMIC);
if (!ifaces) {
spin_unlock(&ses->iface_lock);
return 0;
}
spin_unlock(&ses->iface_lock);
/*
* Keep connecting to same, fastest, iface for all channels as
* long as its RSS. Try next fastest one if not RSS or channel
* creation fails.
*/
while (left > 0) {
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
struct cifs_server_iface *iface;
tries++;
if (tries > 3*ses->chan_max) {
cifs_dbg(FYI, "too many channel open attempts (%d channels left to open)\n",
left);
break;
}
iface = &ifaces[i];
if (is_ses_using_iface(ses, iface) && !iface->rss_capable) {
i = (i+1) % iface_count;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
continue;
}
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
rc = cifs_ses_add_channel(cifs_sb, ses, iface);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
if (rc) {
cifs_dbg(FYI, "failed to open extra channel on iface#%d rc=%d\n",
i, rc);
i = (i+1) % iface_count;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
continue;
}
cifs_dbg(FYI, "successfully opened new channel on iface#%d\n",
i);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
left--;
}
kfree(ifaces);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
return ses->chan_count - old_chan_count;
}
/*
* If server is a channel of ses, return the corresponding enclosing
* cifs_chan otherwise return NULL.
*/
struct cifs_chan *
cifs_ses_find_chan(struct cifs_ses *ses, struct TCP_Server_Info *server)
{
int i;
for (i = 0; i < ses->chan_count; i++) {
if (ses->chans[i].server == server)
return &ses->chans[i];
}
return NULL;
}
static int
cifs_ses_add_channel(struct cifs_sb_info *cifs_sb, struct cifs_ses *ses,
struct cifs_server_iface *iface)
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
{
struct cifs_chan *chan;
struct smb3_fs_context ctx = {NULL};
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
static const char unc_fmt[] = "\\%s\\foo";
char unc[sizeof(unc_fmt)+SERVER_NAME_LEN_WITH_NULL] = {0};
struct sockaddr_in *ipv4 = (struct sockaddr_in *)&iface->sockaddr;
struct sockaddr_in6 *ipv6 = (struct sockaddr_in6 *)&iface->sockaddr;
int rc;
unsigned int xid = get_xid();
if (iface->sockaddr.ss_family == AF_INET)
cifs_dbg(FYI, "adding channel to ses %p (speed:%zu bps rdma:%s ip:%pI4)\n",
ses, iface->speed, iface->rdma_capable ? "yes" : "no",
&ipv4->sin_addr);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
else
cifs_dbg(FYI, "adding channel to ses %p (speed:%zu bps rdma:%s ip:%pI4)\n",
ses, iface->speed, iface->rdma_capable ? "yes" : "no",
&ipv6->sin6_addr);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/*
* Setup a ctx with mostly the same info as the existing
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
* session and overwrite it with the requested iface data.
*
* We need to setup at least the fields used for negprot and
* sesssetup.
*
* We only need the ctx here, so we can reuse memory from
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
* the session and server without caring about memory
* management.
*/
/* Always make new connection for now (TODO?) */
ctx.nosharesock = true;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/* Auth */
ctx.domainauto = ses->domainAuto;
ctx.domainname = ses->domainName;
ctx.username = ses->user_name;
ctx.password = ses->password;
ctx.sectype = ses->sectype;
ctx.sign = ses->sign;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/* UNC and paths */
/* XXX: Use ses->server->hostname? */
sprintf(unc, unc_fmt, ses->ip_addr);
ctx.UNC = unc;
ctx.prepath = "";
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/* Reuse same version as master connection */
ctx.vals = ses->server->vals;
ctx.ops = ses->server->ops;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
ctx.noblocksnd = ses->server->noblocksnd;
ctx.noautotune = ses->server->noautotune;
ctx.sockopt_tcp_nodelay = ses->server->tcp_nodelay;
ctx.echo_interval = ses->server->echo_interval / HZ;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/*
* This will be used for encoding/decoding user/domain/pw
* during sess setup auth.
*/
ctx.local_nls = cifs_sb->local_nls;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/* Use RDMA if possible */
ctx.rdma = iface->rdma_capable;
memcpy(&ctx.dstaddr, &iface->sockaddr, sizeof(struct sockaddr_storage));
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/* reuse master con client guid */
memcpy(&ctx.client_guid, ses->server->client_guid,
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
SMB2_CLIENT_GUID_SIZE);
ctx.use_client_guid = true;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
mutex_lock(&ses->session_mutex);
chan = ses->binding_chan = &ses->chans[ses->chan_count];
chan->server = cifs_get_tcp_session(&ctx);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
if (IS_ERR(chan->server)) {
rc = PTR_ERR(chan->server);
chan->server = NULL;
goto out;
}
spin_lock(&cifs_tcp_ses_lock);
chan->server->is_channel = true;
spin_unlock(&cifs_tcp_ses_lock);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
/*
* We need to allocate the server crypto now as we will need
* to sign packets before we generate the channel signing key
* (we sign with the session key)
*/
rc = smb311_crypto_shash_allocate(chan->server);
if (rc) {
cifs_dbg(VFS, "%s: crypto alloc failed\n", __func__);
goto out;
}
ses->binding = true;
rc = cifs_negotiate_protocol(xid, ses);
if (rc)
goto out;
rc = cifs_setup_session(xid, ses, cifs_sb->local_nls);
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
if (rc)
goto out;
/* success, put it on the list
* XXX: sharing ses between 2 tcp servers is not possible, the
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
* way "internal" linked lists works in linux makes element
* only able to belong to one list
*
* the binding session is already established so the rest of
* the code should be able to look it up, no need to add the
* ses to the new server.
*/
ses->chan_count++;
atomic_set(&ses->chan_seq, 0);
out:
ses->binding = false;
ses->binding_chan = NULL;
cifs: try opening channels after mounting After doing mount() successfully we call cifs_try_adding_channels() which will open as many channels as it can. Channels are closed when the master session is closed. The master connection becomes the first channel. ,-------------> global cifs_tcp_ses_list <-------------------------. | | '- TCP_Server_Info <--> TCP_Server_Info <--> TCP_Server_Info <-' (master con) (chan#1 con) (chan#2 con) | ^ ^ ^ v '--------------------|--------------------' cifs_ses | - chan_count = 3 | - chans[] ---------------------' - smb3signingkey[] (master signing key) Note how channel connections don't have sessions. That's because cifs_ses can only be part of one linked list (list_head are internal to the elements). For signing keys, each channel has its own signing key which must be used only after the channel has been bound. While it's binding it must use the master session signing key. For encryption keys, since channel connections do not have sessions attached we must now find matching session by looping over all sessions in smb2_get_enc_key(). Each channel is opened like a regular server connection but at the session setup request step it must set the SMB2_SESSION_REQ_FLAG_BINDING flag and use the session id to bind to. Finally, while sending in compound_send_recv() for requests that aren't negprot, ses-setup or binding related, use a channel by cycling through the available ones (round-robin). Signed-off-by: Aurelien Aptel <aaptel@suse.com> Signed-off-by: Steve French <stfrench@microsoft.com>
2019-09-20 12:31:10 +08:00
mutex_unlock(&ses->session_mutex);
if (rc && chan->server)
cifs_put_tcp_session(chan->server, 0);
return rc;
}
static __u32 cifs_ssetup_hdr(struct cifs_ses *ses, SESSION_SETUP_ANDX *pSMB)
{
__u32 capabilities = 0;
/* init fields common to all four types of SessSetup */
/* 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 */
pSMB->req.AndXCommand = 0xFF;
pSMB->req.MaxBufferSize = cpu_to_le16(min_t(u32,
CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4,
USHRT_MAX));
pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq);
pSMB->req.VcNumber = cpu_to_le16(1);
/* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */
/* BB verify whether signing required on neg or just on auth frame
(and NTLM case) */
capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS |
CAP_LARGE_WRITE_X | CAP_LARGE_READ_X;
if (ses->server->sign)
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;
}
if (ses->capabilities & CAP_UNIX)
capabilities |= CAP_UNIX;
return capabilities;
}
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 */
bytes_ret = cifs_strtoUTF16((__le16 *)bcc_ptr, "Linux version ", 32,
nls_cp);
bcc_ptr += 2 * bytes_ret;
bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, init_utsname()->release,
32, nls_cp);
bcc_ptr += 2 * bytes_ret;
bcc_ptr += 2; /* trailing null */
bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, CIFS_NETWORK_OPSYS,
32, nls_cp);
bcc_ptr += 2 * bytes_ret;
bcc_ptr += 2; /* trailing null */
*pbcc_area = bcc_ptr;
}
static void unicode_domain_string(char **pbcc_area, struct cifs_ses *ses,
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
bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, ses->domainName,
CIFS_MAX_DOMAINNAME_LEN, nls_cp);
bcc_ptr += 2 * bytes_ret;
bcc_ptr += 2; /* account for null terminator */
*pbcc_area = bcc_ptr;
}
static void unicode_ssetup_strings(char **pbcc_area, struct cifs_ses *ses,
const struct nls_table *nls_cp)
{
char *bcc_ptr = *pbcc_area;
int bytes_ret = 0;
/* BB FIXME add check that strings total less
than 335 or will need to send them as arrays */
/* unicode strings, must be word aligned before the call */
/* if ((long) bcc_ptr % 2) {
*bcc_ptr = 0;
bcc_ptr++;
} */
/* copy user */
if (ses->user_name == NULL) {
/* null user mount */
*bcc_ptr = 0;
*(bcc_ptr+1) = 0;
} else {
bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, ses->user_name,
CIFS_MAX_USERNAME_LEN, nls_cp);
}
bcc_ptr += 2 * bytes_ret;
bcc_ptr += 2; /* account for null termination */
unicode_domain_string(&bcc_ptr, ses, nls_cp);
unicode_oslm_strings(&bcc_ptr, nls_cp);
*pbcc_area = bcc_ptr;
}
static void ascii_ssetup_strings(char **pbcc_area, struct cifs_ses *ses,
const struct nls_table *nls_cp)
{
char *bcc_ptr = *pbcc_area;
int len;
/* copy user */
/* BB what about null user mounts - check that we do this BB */
/* copy user */
if (ses->user_name != NULL) {
len = strscpy(bcc_ptr, ses->user_name, CIFS_MAX_USERNAME_LEN);
if (WARN_ON_ONCE(len < 0))
len = CIFS_MAX_USERNAME_LEN - 1;
bcc_ptr += len;
}
/* else null user mount */
*bcc_ptr = 0;
bcc_ptr++; /* account for null termination */
/* copy domain */
if (ses->domainName != NULL) {
len = strscpy(bcc_ptr, ses->domainName, CIFS_MAX_DOMAINNAME_LEN);
if (WARN_ON_ONCE(len < 0))
len = CIFS_MAX_DOMAINNAME_LEN - 1;
bcc_ptr += len;
} /* else we will send a null domain name
so the server will default to its own domain */
*bcc_ptr = 0;
bcc_ptr++;
/* BB check for overflow here */
strcpy(bcc_ptr, "Linux version ");
bcc_ptr += strlen("Linux version ");
strcpy(bcc_ptr, init_utsname()->release);
bcc_ptr += strlen(init_utsname()->release) + 1;
strcpy(bcc_ptr, CIFS_NETWORK_OPSYS);
bcc_ptr += strlen(CIFS_NETWORK_OPSYS) + 1;
*pbcc_area = bcc_ptr;
}
static void
decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifs_ses *ses,
const struct nls_table *nls_cp)
{
int len;
char *data = *pbcc_area;
cifs_dbg(FYI, "bleft %d\n", bleft);
kfree(ses->serverOS);
ses->serverOS = cifs_strndup_from_utf16(data, bleft, true, nls_cp);
cifs_dbg(FYI, "serverOS=%s\n", ses->serverOS);
len = (UniStrnlen((wchar_t *) data, bleft / 2) * 2) + 2;
data += len;
bleft -= len;
if (bleft <= 0)
return;
kfree(ses->serverNOS);
ses->serverNOS = cifs_strndup_from_utf16(data, bleft, true, nls_cp);
cifs_dbg(FYI, "serverNOS=%s\n", ses->serverNOS);
len = (UniStrnlen((wchar_t *) data, bleft / 2) * 2) + 2;
data += len;
bleft -= len;
if (bleft <= 0)
return;
kfree(ses->serverDomain);
ses->serverDomain = cifs_strndup_from_utf16(data, bleft, true, nls_cp);
cifs_dbg(FYI, "serverDomain=%s\n", ses->serverDomain);
return;
}
static void decode_ascii_ssetup(char **pbcc_area, __u16 bleft,
struct cifs_ses *ses,
const struct nls_table *nls_cp)
{
int len;
char *bcc_ptr = *pbcc_area;
cifs_dbg(FYI, "decode sessetup ascii. bleft %d\n", bleft);
len = strnlen(bcc_ptr, bleft);
if (len >= bleft)
return;
kfree(ses->serverOS);
ses->serverOS = kmalloc(len + 1, GFP_KERNEL);
if (ses->serverOS) {
memcpy(ses->serverOS, bcc_ptr, len);
ses->serverOS[len] = 0;
if (strncmp(ses->serverOS, "OS/2", 4) == 0)
cifs_dbg(FYI, "OS/2 server\n");
}
bcc_ptr += len + 1;
bleft -= len + 1;
len = strnlen(bcc_ptr, bleft);
if (len >= bleft)
return;
kfree(ses->serverNOS);
ses->serverNOS = kmalloc(len + 1, GFP_KERNEL);
if (ses->serverNOS) {
memcpy(ses->serverNOS, bcc_ptr, len);
ses->serverNOS[len] = 0;
}
bcc_ptr += len + 1;
bleft -= len + 1;
len = strnlen(bcc_ptr, bleft);
if (len > bleft)
return;
/* 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 */
cifs_dbg(FYI, "ascii: bytes left %d\n", bleft);
}
int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len,
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 11:02:18 +08:00
unsigned int tioffset; /* challenge message target info area */
unsigned int tilen; /* challenge message target info area length */
CHALLENGE_MESSAGE *pblob = (CHALLENGE_MESSAGE *)bcc_ptr;
if (blob_len < sizeof(CHALLENGE_MESSAGE)) {
cifs_dbg(VFS, "challenge blob len %d too small\n", blob_len);
return -EINVAL;
}
if (memcmp(pblob->Signature, "NTLMSSP", 8)) {
cifs_dbg(VFS, "blob signature incorrect %s\n",
pblob->Signature);
return -EINVAL;
}
if (pblob->MessageType != NtLmChallenge) {
cifs_dbg(VFS, "Incorrect message type %d\n",
pblob->MessageType);
return -EINVAL;
}
memcpy(ses->ntlmssp->cryptkey, pblob->Challenge, CIFS_CRYPTO_KEY_SIZE);
/* 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 */
ses->ntlmssp->server_flags = le32_to_cpu(pblob->NegotiateFlags);
tioffset = le32_to_cpu(pblob->TargetInfoArray.BufferOffset);
tilen = le16_to_cpu(pblob->TargetInfoArray.Length);
if (tioffset > blob_len || tioffset + tilen > blob_len) {
cifs_dbg(VFS, "tioffset + tilen too high %u + %u\n",
tioffset, tilen);
return -EINVAL;
}
if (tilen) {
ses->auth_key.response = kmemdup(bcc_ptr + tioffset, tilen,
GFP_KERNEL);
if (!ses->auth_key.response) {
cifs_dbg(VFS, "Challenge target info alloc failure\n");
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 11:02:18 +08:00
return -ENOMEM;
}
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 11:02:18 +08:00
}
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 */
void build_ntlmssp_negotiate_blob(unsigned char *pbuffer,
struct cifs_ses *ses)
{
struct TCP_Server_Info *server = cifs_ses_server(ses);
NEGOTIATE_MESSAGE *sec_blob = (NEGOTIATE_MESSAGE *)pbuffer;
__u32 flags;
memset(pbuffer, 0, sizeof(NEGOTIATE_MESSAGE));
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 |
NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC |
NTLMSSP_NEGOTIATE_SEAL;
if (server->sign)
flags |= NTLMSSP_NEGOTIATE_SIGN;
if (!server->session_estab || ses->ntlmssp->sesskey_per_smbsess)
flags |= NTLMSSP_NEGOTIATE_KEY_XCH;
sec_blob->NegotiateFlags = cpu_to_le32(flags);
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;
}
static int size_of_ntlmssp_blob(struct cifs_ses *ses)
{
int sz = sizeof(AUTHENTICATE_MESSAGE) + ses->auth_key.len
- CIFS_SESS_KEY_SIZE + CIFS_CPHTXT_SIZE + 2;
if (ses->domainName)
sz += 2 * strnlen(ses->domainName, CIFS_MAX_DOMAINNAME_LEN);
else
sz += 2;
if (ses->user_name)
sz += 2 * strnlen(ses->user_name, CIFS_MAX_USERNAME_LEN);
else
sz += 2;
return sz;
}
int build_ntlmssp_auth_blob(unsigned char **pbuffer,
u16 *buflen,
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 11:02:18 +08:00
const struct nls_table *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 11:02:18 +08:00
int rc;
AUTHENTICATE_MESSAGE *sec_blob;
__u32 flags;
unsigned char *tmp;
rc = setup_ntlmv2_rsp(ses, nls_cp);
if (rc) {
cifs_dbg(VFS, "Error %d during NTLMSSP authentication\n", rc);
*buflen = 0;
goto setup_ntlmv2_ret;
}
*pbuffer = kmalloc(size_of_ntlmssp_blob(ses), GFP_KERNEL);
if (!*pbuffer) {
rc = -ENOMEM;
cifs_dbg(VFS, "Error %d during NTLMSSP allocation\n", rc);
*buflen = 0;
goto setup_ntlmv2_ret;
}
sec_blob = (AUTHENTICATE_MESSAGE *)*pbuffer;
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 |
NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC |
NTLMSSP_NEGOTIATE_SEAL;
if (ses->server->sign)
flags |= NTLMSSP_NEGOTIATE_SIGN;
if (!ses->server->session_estab || ses->ntlmssp->sesskey_per_smbsess)
flags |= NTLMSSP_NEGOTIATE_KEY_XCH;
tmp = *pbuffer + sizeof(AUTHENTICATE_MESSAGE);
sec_blob->NegotiateFlags = cpu_to_le32(flags);
sec_blob->LmChallengeResponse.BufferOffset =
cpu_to_le32(sizeof(AUTHENTICATE_MESSAGE));
sec_blob->LmChallengeResponse.Length = 0;
sec_blob->LmChallengeResponse.MaximumLength = 0;
sec_blob->NtChallengeResponse.BufferOffset =
cpu_to_le32(tmp - *pbuffer);
if (ses->user_name != NULL) {
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;
sec_blob->NtChallengeResponse.Length =
cpu_to_le16(ses->auth_key.len - CIFS_SESS_KEY_SIZE);
sec_blob->NtChallengeResponse.MaximumLength =
cpu_to_le16(ses->auth_key.len - CIFS_SESS_KEY_SIZE);
} else {
/*
* don't send an NT Response for anonymous access
*/
sec_blob->NtChallengeResponse.Length = 0;
sec_blob->NtChallengeResponse.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 11:02:18 +08:00
}
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;
len = cifs_strtoUTF16((__le16 *)tmp, ses->domainName,
CIFS_MAX_DOMAINNAME_LEN, nls_cp);
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;
}
if (ses->user_name == NULL) {
sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - *pbuffer);
sec_blob->UserName.Length = 0;
sec_blob->UserName.MaximumLength = 0;
tmp += 2;
} else {
int len;
len = cifs_strtoUTF16((__le16 *)tmp, ses->user_name,
CIFS_MAX_USERNAME_LEN, nls_cp);
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;
if (((ses->ntlmssp->server_flags & NTLMSSP_NEGOTIATE_KEY_XCH) ||
(ses->ntlmssp->server_flags & NTLMSSP_NEGOTIATE_EXTENDED_SEC))
&& !calc_seckey(ses)) {
memcpy(tmp, ses->ntlmssp->ciphertext, CIFS_CPHTXT_SIZE);
sec_blob->SessionKey.BufferOffset = cpu_to_le32(tmp - *pbuffer);
NTLM auth and sign - Define crypto hash functions and create and send keys needed for key exchange Mark dependency on crypto modules in Kconfig. Defining per structures sdesc and cifs_secmech which are used to store crypto hash functions and contexts. They are stored per smb connection and used for all auth mechs to genereate hash values and signatures. Allocate crypto hashing functions, security descriptiors, and respective contexts when a smb/tcp connection is established. Release them when a tcp/smb connection is taken down. md5 and hmac-md5 are two crypto hashing functions that are used throught the life of an smb/tcp connection by various functions that calcualte signagure and ntlmv2 hash, HMAC etc. structure ntlmssp_auth is defined as per smb connection. ntlmssp_auth holds ciphertext which is genereated by rc4/arc4 encryption of secondary key, a nonce using ntlmv2 session key and sent in the session key field of the type 3 message sent by the client during ntlmssp negotiation/exchange A key is exchanged with the server if client indicates so in flags in type 1 messsage and server agrees in flag in type 2 message of ntlmssp negotiation. If both client and agree, a key sent by client in type 3 message of ntlmssp negotiation in the session key field. The key is a ciphertext generated off of secondary key, a nonce, using ntlmv2 hash via rc4/arc4. Signing works for ntlmssp in this patch. The sequence number within the server structure needs to be zero until session is established i.e. till type 3 packet of ntlmssp exchange of a to be very first smb session on that smb connection is sent. Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-10-22 03:25:08 +08:00
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);
NTLM auth and sign - Define crypto hash functions and create and send keys needed for key exchange Mark dependency on crypto modules in Kconfig. Defining per structures sdesc and cifs_secmech which are used to store crypto hash functions and contexts. They are stored per smb connection and used for all auth mechs to genereate hash values and signatures. Allocate crypto hashing functions, security descriptiors, and respective contexts when a smb/tcp connection is established. Release them when a tcp/smb connection is taken down. md5 and hmac-md5 are two crypto hashing functions that are used throught the life of an smb/tcp connection by various functions that calcualte signagure and ntlmv2 hash, HMAC etc. structure ntlmssp_auth is defined as per smb connection. ntlmssp_auth holds ciphertext which is genereated by rc4/arc4 encryption of secondary key, a nonce using ntlmv2 session key and sent in the session key field of the type 3 message sent by the client during ntlmssp negotiation/exchange A key is exchanged with the server if client indicates so in flags in type 1 messsage and server agrees in flag in type 2 message of ntlmssp negotiation. If both client and agree, a key sent by client in type 3 message of ntlmssp negotiation in the session key field. The key is a ciphertext generated off of secondary key, a nonce, using ntlmv2 hash via rc4/arc4. Signing works for ntlmssp in this patch. The sequence number within the server structure needs to be zero until session is established i.e. till type 3 packet of ntlmssp exchange of a to be very first smb session on that smb connection is sent. Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2010-10-22 03:25:08 +08:00
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 11:02:18 +08:00
*buflen = tmp - *pbuffer;
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 11:02:18 +08:00
setup_ntlmv2_ret:
return rc;
}
enum securityEnum
cifs_select_sectype(struct TCP_Server_Info *server, enum securityEnum requested)
{
switch (server->negflavor) {
case CIFS_NEGFLAVOR_EXTENDED:
switch (requested) {
case Kerberos:
case RawNTLMSSP:
return requested;
case Unspecified:
if (server->sec_ntlmssp &&
(global_secflags & CIFSSEC_MAY_NTLMSSP))
return RawNTLMSSP;
if ((server->sec_kerberos || server->sec_mskerberos) &&
(global_secflags & CIFSSEC_MAY_KRB5))
return Kerberos;
fallthrough;
default:
return Unspecified;
}
case CIFS_NEGFLAVOR_UNENCAP:
switch (requested) {
case NTLM:
case NTLMv2:
return requested;
case Unspecified:
if (global_secflags & CIFSSEC_MAY_NTLMV2)
return NTLMv2;
if (global_secflags & CIFSSEC_MAY_NTLM)
return NTLM;
break;
default:
break;
}
fallthrough; /* to attempt LANMAN authentication next */
case CIFS_NEGFLAVOR_LANMAN:
switch (requested) {
case LANMAN:
return requested;
case Unspecified:
if (global_secflags & CIFSSEC_MAY_LANMAN)
return LANMAN;
fallthrough;
default:
return Unspecified;
}
default:
return Unspecified;
}
}
struct sess_data {
unsigned int xid;
struct cifs_ses *ses;
struct nls_table *nls_cp;
void (*func)(struct sess_data *);
int result;
/* 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 17K allocation
*/
int buf0_type;
struct kvec iov[3];
};
static int
sess_alloc_buffer(struct sess_data *sess_data, int wct)
{
int rc;
struct cifs_ses *ses = sess_data->ses;
struct smb_hdr *smb_buf;
rc = small_smb_init_no_tc(SMB_COM_SESSION_SETUP_ANDX, wct, ses,
(void **)&smb_buf);
if (rc)
return rc;
sess_data->iov[0].iov_base = (char *)smb_buf;
sess_data->iov[0].iov_len = be32_to_cpu(smb_buf->smb_buf_length) + 4;
/*
* This variable will be used to clear the buffer
* allocated above in case of any error in the calling function.
*/
sess_data->buf0_type = CIFS_SMALL_BUFFER;
/* 2000 big enough to fit max user, domain, NOS name etc. */
sess_data->iov[2].iov_base = kmalloc(2000, GFP_KERNEL);
if (!sess_data->iov[2].iov_base) {
rc = -ENOMEM;
goto out_free_smb_buf;
}
return 0;
out_free_smb_buf:
kfree(smb_buf);
sess_data->iov[0].iov_base = NULL;
sess_data->iov[0].iov_len = 0;
sess_data->buf0_type = CIFS_NO_BUFFER;
return rc;
}
static void
sess_free_buffer(struct sess_data *sess_data)
{
free_rsp_buf(sess_data->buf0_type, sess_data->iov[0].iov_base);
sess_data->buf0_type = CIFS_NO_BUFFER;
kfree(sess_data->iov[2].iov_base);
}
static int
sess_establish_session(struct sess_data *sess_data)
{
struct cifs_ses *ses = sess_data->ses;
mutex_lock(&ses->server->srv_mutex);
if (!ses->server->session_estab) {
if (ses->server->sign) {
ses->server->session_key.response =
kmemdup(ses->auth_key.response,
ses->auth_key.len, GFP_KERNEL);
if (!ses->server->session_key.response) {
mutex_unlock(&ses->server->srv_mutex);
return -ENOMEM;
}
ses->server->session_key.len =
ses->auth_key.len;
}
ses->server->sequence_number = 0x2;
ses->server->session_estab = true;
}
mutex_unlock(&ses->server->srv_mutex);
cifs_dbg(FYI, "CIFS session established successfully\n");
spin_lock(&GlobalMid_Lock);
ses->status = CifsGood;
ses->need_reconnect = false;
spin_unlock(&GlobalMid_Lock);
return 0;
}
static int
sess_sendreceive(struct sess_data *sess_data)
{
int rc;
struct smb_hdr *smb_buf = (struct smb_hdr *) sess_data->iov[0].iov_base;
__u16 count;
struct kvec rsp_iov = { NULL, 0 };
count = sess_data->iov[1].iov_len + sess_data->iov[2].iov_len;
be32_add_cpu(&smb_buf->smb_buf_length, count);
put_bcc(count, smb_buf);
rc = SendReceive2(sess_data->xid, sess_data->ses,
sess_data->iov, 3 /* num_iovecs */,
&sess_data->buf0_type,
CIFS_LOG_ERROR, &rsp_iov);
cifs_small_buf_release(sess_data->iov[0].iov_base);
memcpy(&sess_data->iov[0], &rsp_iov, sizeof(struct kvec));
return rc;
}
/*
* 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
*/
#ifdef CONFIG_CIFS_WEAK_PW_HASH
static void
sess_auth_lanman(struct sess_data *sess_data)
{
int rc = 0;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
char *bcc_ptr;
struct cifs_ses *ses = sess_data->ses;
char lnm_session_key[CIFS_AUTH_RESP_SIZE];
__u16 bytes_remaining;
/* lanman 2 style sessionsetup */
/* wct = 10 */
rc = sess_alloc_buffer(sess_data, 10);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
bcc_ptr = sess_data->iov[2].iov_base;
(void)cifs_ssetup_hdr(ses, pSMB);
pSMB->req.hdr.Flags2 &= ~SMBFLG2_UNICODE;
if (ses->user_name != NULL) {
/* no capabilities flags in old lanman negotiation */
pSMB->old_req.PasswordLength = cpu_to_le16(CIFS_AUTH_RESP_SIZE);
/* Calculate hash with password and copy into bcc_ptr.
* Encryption Key (stored as in cryptkey) gets used if the
* security mode bit in Negotiate Protocol response states
* to use challenge/response method (i.e. Password bit is 1).
*/
rc = calc_lanman_hash(ses->password, ses->server->cryptkey,
ses->server->sec_mode & SECMODE_PW_ENCRYPT ?
true : false, lnm_session_key);
if (rc)
goto out;
memcpy(bcc_ptr, (char *)lnm_session_key, CIFS_AUTH_RESP_SIZE);
bcc_ptr += CIFS_AUTH_RESP_SIZE;
} else {
pSMB->old_req.PasswordLength = 0;
}
/*
* 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?
*/
cifs_dbg(FYI, "Negotiating LANMAN setting up strings\n");
/* Unicode not allowed for LANMAN dialects */
ascii_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
sess_data->iov[2].iov_len = (long) bcc_ptr -
(long) sess_data->iov[2].iov_base;
rc = sess_sendreceive(sess_data);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
/* lanman response has a word count of 3 */
if (smb_buf->WordCount != 3) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out;
}
if (le16_to_cpu(pSMB->resp.Action) & GUEST_LOGIN)
cifs_dbg(FYI, "Guest login\n"); /* BB mark SesInfo struct? */
ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */
cifs_dbg(FYI, "UID = %llu\n", ses->Suid);
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
/* BB check if Unicode and decode strings */
if (bytes_remaining == 0) {
/* no string area to decode, do nothing */
} else if (smb_buf->Flags2 & SMBFLG2_UNICODE) {
/* unicode string area must be word-aligned */
if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) {
++bcc_ptr;
--bytes_remaining;
}
decode_unicode_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
} else {
decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
}
rc = sess_establish_session(sess_data);
out:
sess_data->result = rc;
sess_data->func = NULL;
sess_free_buffer(sess_data);
}
#endif
static void
sess_auth_ntlm(struct sess_data *sess_data)
{
int rc = 0;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
char *bcc_ptr;
struct cifs_ses *ses = sess_data->ses;
__u32 capabilities;
__u16 bytes_remaining;
/* old style NTLM sessionsetup */
/* wct = 13 */
rc = sess_alloc_buffer(sess_data, 13);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
bcc_ptr = sess_data->iov[2].iov_base;
capabilities = cifs_ssetup_hdr(ses, pSMB);
pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities);
if (ses->user_name != NULL) {
pSMB->req_no_secext.CaseInsensitivePasswordLength =
cpu_to_le16(CIFS_AUTH_RESP_SIZE);
pSMB->req_no_secext.CaseSensitivePasswordLength =
cpu_to_le16(CIFS_AUTH_RESP_SIZE);
/* calculate ntlm response and session key */
rc = setup_ntlm_response(ses, sess_data->nls_cp);
if (rc) {
cifs_dbg(VFS, "Error %d during NTLM authentication\n",
rc);
goto out;
}
/* 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;
} else {
pSMB->req_no_secext.CaseInsensitivePasswordLength = 0;
pSMB->req_no_secext.CaseSensitivePasswordLength = 0;
}
if (ses->capabilities & CAP_UNICODE) {
/* unicode strings must be word aligned */
if (sess_data->iov[0].iov_len % 2) {
*bcc_ptr = 0;
bcc_ptr++;
}
unicode_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
} else {
ascii_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
}
sess_data->iov[2].iov_len = (long) bcc_ptr -
(long) sess_data->iov[2].iov_base;
rc = sess_sendreceive(sess_data);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
if (smb_buf->WordCount != 3) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out;
}
if (le16_to_cpu(pSMB->resp.Action) & GUEST_LOGIN)
cifs_dbg(FYI, "Guest login\n"); /* BB mark SesInfo struct? */
ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */
cifs_dbg(FYI, "UID = %llu\n", ses->Suid);
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
/* BB check if Unicode and decode strings */
if (bytes_remaining == 0) {
/* no string area to decode, do nothing */
} else if (smb_buf->Flags2 & SMBFLG2_UNICODE) {
/* unicode string area must be word-aligned */
if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) {
++bcc_ptr;
--bytes_remaining;
}
decode_unicode_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
} else {
decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
}
rc = sess_establish_session(sess_data);
out:
sess_data->result = rc;
sess_data->func = NULL;
sess_free_buffer(sess_data);
kfree(ses->auth_key.response);
ses->auth_key.response = NULL;
}
static void
sess_auth_ntlmv2(struct sess_data *sess_data)
{
int rc = 0;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
char *bcc_ptr;
struct cifs_ses *ses = sess_data->ses;
__u32 capabilities;
__u16 bytes_remaining;
/* old style NTLM sessionsetup */
/* wct = 13 */
rc = sess_alloc_buffer(sess_data, 13);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
bcc_ptr = sess_data->iov[2].iov_base;
capabilities = cifs_ssetup_hdr(ses, pSMB);
pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities);
/* LM2 password would be here if we supported it */
pSMB->req_no_secext.CaseInsensitivePasswordLength = 0;
if (ses->user_name != NULL) {
/* calculate nlmv2 response and session key */
rc = setup_ntlmv2_rsp(ses, sess_data->nls_cp);
if (rc) {
cifs_dbg(VFS, "Error %d during NTLMv2 authentication\n", rc);
goto out;
}
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;
/* set case sensitive password length after tilen may get
* assigned, tilen is 0 otherwise.
*/
pSMB->req_no_secext.CaseSensitivePasswordLength =
cpu_to_le16(ses->auth_key.len - CIFS_SESS_KEY_SIZE);
} else {
pSMB->req_no_secext.CaseSensitivePasswordLength = 0;
}
if (ses->capabilities & CAP_UNICODE) {
if (sess_data->iov[0].iov_len % 2) {
*bcc_ptr = 0;
bcc_ptr++;
}
unicode_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
} else {
ascii_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
}
sess_data->iov[2].iov_len = (long) bcc_ptr -
(long) sess_data->iov[2].iov_base;
rc = sess_sendreceive(sess_data);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
if (smb_buf->WordCount != 3) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out;
}
if (le16_to_cpu(pSMB->resp.Action) & GUEST_LOGIN)
cifs_dbg(FYI, "Guest login\n"); /* BB mark SesInfo struct? */
ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */
cifs_dbg(FYI, "UID = %llu\n", ses->Suid);
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
/* BB check if Unicode and decode strings */
if (bytes_remaining == 0) {
/* no string area to decode, do nothing */
} else if (smb_buf->Flags2 & SMBFLG2_UNICODE) {
/* unicode string area must be word-aligned */
if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) {
++bcc_ptr;
--bytes_remaining;
}
decode_unicode_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
} else {
decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
}
rc = sess_establish_session(sess_data);
out:
sess_data->result = rc;
sess_data->func = NULL;
sess_free_buffer(sess_data);
kfree(ses->auth_key.response);
ses->auth_key.response = NULL;
}
#ifdef CONFIG_CIFS_UPCALL
static void
sess_auth_kerberos(struct sess_data *sess_data)
{
int rc = 0;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
char *bcc_ptr;
struct cifs_ses *ses = sess_data->ses;
__u32 capabilities;
__u16 bytes_remaining;
struct key *spnego_key = NULL;
struct cifs_spnego_msg *msg;
u16 blob_len;
/* extended security */
/* wct = 12 */
rc = sess_alloc_buffer(sess_data, 12);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
bcc_ptr = sess_data->iov[2].iov_base;
capabilities = cifs_ssetup_hdr(ses, pSMB);
spnego_key = cifs_get_spnego_key(ses);
if (IS_ERR(spnego_key)) {
rc = PTR_ERR(spnego_key);
spnego_key = NULL;
goto out;
}
msg = spnego_key->payload.data[0];
/*
* 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) {
cifs_dbg(VFS, "incorrect version of cifs.upcall (expected %d but got %d)\n",
CIFS_SPNEGO_UPCALL_VERSION, msg->version);
rc = -EKEYREJECTED;
goto out_put_spnego_key;
}
ses->auth_key.response = kmemdup(msg->data, msg->sesskey_len,
GFP_KERNEL);
if (!ses->auth_key.response) {
cifs_dbg(VFS, "Kerberos can't allocate (%u bytes) memory\n",
msg->sesskey_len);
rc = -ENOMEM;
goto out_put_spnego_key;
}
ses->auth_key.len = msg->sesskey_len;
pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC;
capabilities |= CAP_EXTENDED_SECURITY;
pSMB->req.Capabilities = cpu_to_le32(capabilities);
sess_data->iov[1].iov_base = msg->data + msg->sesskey_len;
sess_data->iov[1].iov_len = msg->secblob_len;
pSMB->req.SecurityBlobLength = cpu_to_le16(sess_data->iov[1].iov_len);
if (ses->capabilities & CAP_UNICODE) {
/* unicode strings must be word aligned */
if ((sess_data->iov[0].iov_len
+ sess_data->iov[1].iov_len) % 2) {
*bcc_ptr = 0;
bcc_ptr++;
}
unicode_oslm_strings(&bcc_ptr, sess_data->nls_cp);
unicode_domain_string(&bcc_ptr, ses, sess_data->nls_cp);
} else {
/* BB: is this right? */
ascii_ssetup_strings(&bcc_ptr, ses, sess_data->nls_cp);
}
sess_data->iov[2].iov_len = (long) bcc_ptr -
(long) sess_data->iov[2].iov_base;
rc = sess_sendreceive(sess_data);
if (rc)
goto out_put_spnego_key;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
if (smb_buf->WordCount != 4) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out_put_spnego_key;
}
if (le16_to_cpu(pSMB->resp.Action) & GUEST_LOGIN)
cifs_dbg(FYI, "Guest login\n"); /* BB mark SesInfo struct? */
ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */
cifs_dbg(FYI, "UID = %llu\n", ses->Suid);
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
blob_len = le16_to_cpu(pSMB->resp.SecurityBlobLength);
if (blob_len > bytes_remaining) {
cifs_dbg(VFS, "bad security blob length %d\n",
blob_len);
rc = -EINVAL;
goto out_put_spnego_key;
}
bcc_ptr += blob_len;
bytes_remaining -= blob_len;
/* BB check if Unicode and decode strings */
if (bytes_remaining == 0) {
/* no string area to decode, do nothing */
} else if (smb_buf->Flags2 & SMBFLG2_UNICODE) {
/* unicode string area must be word-aligned */
if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) {
++bcc_ptr;
--bytes_remaining;
}
decode_unicode_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
} else {
decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
}
rc = sess_establish_session(sess_data);
out_put_spnego_key:
key_invalidate(spnego_key);
key_put(spnego_key);
out:
sess_data->result = rc;
sess_data->func = NULL;
sess_free_buffer(sess_data);
kfree(ses->auth_key.response);
ses->auth_key.response = NULL;
}
#endif /* ! CONFIG_CIFS_UPCALL */
/*
* The required kvec buffers have to be allocated before calling this
* function.
*/
static int
_sess_auth_rawntlmssp_assemble_req(struct sess_data *sess_data)
{
SESSION_SETUP_ANDX *pSMB;
struct cifs_ses *ses = sess_data->ses;
__u32 capabilities;
char *bcc_ptr;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
capabilities = cifs_ssetup_hdr(ses, pSMB);
if ((pSMB->req.hdr.Flags2 & SMBFLG2_UNICODE) == 0) {
cifs_dbg(VFS, "NTLMSSP requires Unicode support\n");
return -ENOSYS;
}
pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC;
capabilities |= CAP_EXTENDED_SECURITY;
pSMB->req.Capabilities |= cpu_to_le32(capabilities);
bcc_ptr = sess_data->iov[2].iov_base;
/* unicode strings must be word aligned */
if ((sess_data->iov[0].iov_len + sess_data->iov[1].iov_len) % 2) {
*bcc_ptr = 0;
bcc_ptr++;
}
unicode_oslm_strings(&bcc_ptr, sess_data->nls_cp);
sess_data->iov[2].iov_len = (long) bcc_ptr -
(long) sess_data->iov[2].iov_base;
return 0;
}
static void
sess_auth_rawntlmssp_authenticate(struct sess_data *sess_data);
static void
sess_auth_rawntlmssp_negotiate(struct sess_data *sess_data)
{
int rc;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
struct cifs_ses *ses = sess_data->ses;
__u16 bytes_remaining;
char *bcc_ptr;
u16 blob_len;
cifs_dbg(FYI, "rawntlmssp session setup negotiate phase\n");
/*
* if memory allocation is successful, caller of this function
* frees it.
*/
ses->ntlmssp = kmalloc(sizeof(struct ntlmssp_auth), GFP_KERNEL);
if (!ses->ntlmssp) {
rc = -ENOMEM;
goto out;
}
ses->ntlmssp->sesskey_per_smbsess = false;
/* wct = 12 */
rc = sess_alloc_buffer(sess_data, 12);
if (rc)
goto out;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
/* Build security blob before we assemble the request */
build_ntlmssp_negotiate_blob(pSMB->req.SecurityBlob, ses);
sess_data->iov[1].iov_len = sizeof(NEGOTIATE_MESSAGE);
sess_data->iov[1].iov_base = pSMB->req.SecurityBlob;
pSMB->req.SecurityBlobLength = cpu_to_le16(sizeof(NEGOTIATE_MESSAGE));
rc = _sess_auth_rawntlmssp_assemble_req(sess_data);
if (rc)
goto out;
rc = sess_sendreceive(sess_data);
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
/* If true, rc here is expected and not an error */
if (sess_data->buf0_type != CIFS_NO_BUFFER &&
smb_buf->Status.CifsError ==
cpu_to_le32(NT_STATUS_MORE_PROCESSING_REQUIRED))
rc = 0;
if (rc)
goto out;
cifs_dbg(FYI, "rawntlmssp session setup challenge phase\n");
if (smb_buf->WordCount != 4) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out;
}
ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */
cifs_dbg(FYI, "UID = %llu\n", ses->Suid);
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
blob_len = le16_to_cpu(pSMB->resp.SecurityBlobLength);
if (blob_len > bytes_remaining) {
cifs_dbg(VFS, "bad security blob length %d\n",
blob_len);
rc = -EINVAL;
goto out;
}
rc = decode_ntlmssp_challenge(bcc_ptr, blob_len, ses);
out:
sess_free_buffer(sess_data);
if (!rc) {
sess_data->func = sess_auth_rawntlmssp_authenticate;
return;
}
/* Else error. Cleanup */
kfree(ses->auth_key.response);
ses->auth_key.response = NULL;
kfree(ses->ntlmssp);
ses->ntlmssp = NULL;
sess_data->func = NULL;
sess_data->result = rc;
}
static void
sess_auth_rawntlmssp_authenticate(struct sess_data *sess_data)
{
int rc;
struct smb_hdr *smb_buf;
SESSION_SETUP_ANDX *pSMB;
struct cifs_ses *ses = sess_data->ses;
__u16 bytes_remaining;
char *bcc_ptr;
unsigned char *ntlmsspblob = NULL;
u16 blob_len;
cifs_dbg(FYI, "rawntlmssp session setup authenticate phase\n");
/* wct = 12 */
rc = sess_alloc_buffer(sess_data, 12);
if (rc)
goto out;
/* Build security blob before we assemble the request */
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)pSMB;
rc = build_ntlmssp_auth_blob(&ntlmsspblob,
&blob_len, ses, sess_data->nls_cp);
if (rc)
goto out_free_ntlmsspblob;
sess_data->iov[1].iov_len = blob_len;
sess_data->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;
rc = _sess_auth_rawntlmssp_assemble_req(sess_data);
if (rc)
goto out_free_ntlmsspblob;
rc = sess_sendreceive(sess_data);
if (rc)
goto out_free_ntlmsspblob;
pSMB = (SESSION_SETUP_ANDX *)sess_data->iov[0].iov_base;
smb_buf = (struct smb_hdr *)sess_data->iov[0].iov_base;
if (smb_buf->WordCount != 4) {
rc = -EIO;
cifs_dbg(VFS, "bad word count %d\n", smb_buf->WordCount);
goto out_free_ntlmsspblob;
}
if (le16_to_cpu(pSMB->resp.Action) & GUEST_LOGIN)
cifs_dbg(FYI, "Guest login\n"); /* BB mark SesInfo struct? */
if (ses->Suid != smb_buf->Uid) {
ses->Suid = smb_buf->Uid;
cifs_dbg(FYI, "UID changed! new UID = %llu\n", ses->Suid);
}
bytes_remaining = get_bcc(smb_buf);
bcc_ptr = pByteArea(smb_buf);
blob_len = le16_to_cpu(pSMB->resp.SecurityBlobLength);
if (blob_len > bytes_remaining) {
cifs_dbg(VFS, "bad security blob length %d\n",
blob_len);
rc = -EINVAL;
goto out_free_ntlmsspblob;
}
bcc_ptr += blob_len;
bytes_remaining -= blob_len;
/* BB check if Unicode and decode strings */
if (bytes_remaining == 0) {
/* no string area to decode, do nothing */
} else if (smb_buf->Flags2 & SMBFLG2_UNICODE) {
/* unicode string area must be word-aligned */
if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) {
++bcc_ptr;
--bytes_remaining;
}
decode_unicode_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
} else {
decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,
sess_data->nls_cp);
}
out_free_ntlmsspblob:
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 11:02:18 +08:00
kfree(ntlmsspblob);
out:
sess_free_buffer(sess_data);
if (!rc)
rc = sess_establish_session(sess_data);
/* Cleanup */
kfree(ses->auth_key.response);
ses->auth_key.response = NULL;
kfree(ses->ntlmssp);
ses->ntlmssp = NULL;
sess_data->func = NULL;
sess_data->result = rc;
}
static int select_sec(struct cifs_ses *ses, struct sess_data *sess_data)
{
int type;
type = cifs_select_sectype(ses->server, ses->sectype);
cifs_dbg(FYI, "sess setup type %d\n", type);
if (type == Unspecified) {
cifs_dbg(VFS, "Unable to select appropriate authentication method!\n");
return -EINVAL;
}
switch (type) {
case LANMAN:
/* 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 */
#ifdef CONFIG_CIFS_WEAK_PW_HASH
sess_data->func = sess_auth_lanman;
break;
#else
return -EOPNOTSUPP;
#endif
case NTLM:
sess_data->func = sess_auth_ntlm;
break;
case NTLMv2:
sess_data->func = sess_auth_ntlmv2;
break;
case Kerberos:
#ifdef CONFIG_CIFS_UPCALL
sess_data->func = sess_auth_kerberos;
break;
#else
cifs_dbg(VFS, "Kerberos negotiated but upcall support disabled!\n");
return -ENOSYS;
#endif /* CONFIG_CIFS_UPCALL */
case RawNTLMSSP:
sess_data->func = sess_auth_rawntlmssp_negotiate;
break;
default:
cifs_dbg(VFS, "secType %d not supported!\n", type);
return -ENOSYS;
}
return 0;
}
int CIFS_SessSetup(const unsigned int xid, struct cifs_ses *ses,
const struct nls_table *nls_cp)
{
int rc = 0;
struct sess_data *sess_data;
if (ses == NULL) {
WARN(1, "%s: ses == NULL!", __func__);
return -EINVAL;
}
sess_data = kzalloc(sizeof(struct sess_data), GFP_KERNEL);
if (!sess_data)
return -ENOMEM;
rc = select_sec(ses, sess_data);
if (rc)
goto out;
sess_data->xid = xid;
sess_data->ses = ses;
sess_data->buf0_type = CIFS_NO_BUFFER;
sess_data->nls_cp = (struct nls_table *) nls_cp;
while (sess_data->func)
sess_data->func(sess_data);
/* Store result before we free sess_data */
rc = sess_data->result;
out:
kfree(sess_data);
return rc;
}