blob: fa096f9c824d7d654f65cef2193c38c57fd6d305 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001The CIFS VFS support for Linux supports many advanced network filesystem
Uwe Kleine-König1b3c3712007-02-17 19:23:03 +01002features such as hierarchical dfs like namespace, hardlinks, locking and more.
Linus Torvalds1da177e2005-04-16 15:20:36 -07003It was designed to comply with the SNIA CIFS Technical Reference (which
4supersedes the 1992 X/Open SMB Standard) as well as to perform best practice
5practical interoperability with Windows 2000, Windows XP, Samba and equivalent
6servers.
7
8For questions or bug reports please contact:
9 sfrench@samba.org (sfrench@us.ibm.com)
10
11Build instructions:
12==================
13For Linux 2.4:
141) Get the kernel source (e.g.from http://www.kernel.org)
15and download the cifs vfs source (see the project page
16at http://us1.samba.org/samba/Linux_CIFS_client.html)
17and change directory into the top of the kernel directory
18then patch the kernel (e.g. "patch -p1 < cifs_24.patch")
19to add the cifs vfs to your kernel configure options if
20it has not already been added (e.g. current SuSE and UL
21users do not need to apply the cifs_24.patch since the cifs vfs is
22already in the kernel configure menu) and then
23mkdir linux/fs/cifs and then copy the current cifs vfs files from
24the cifs download to your kernel build directory e.g.
25
26 cp <cifs_download_dir>/fs/cifs/* to <kernel_download_dir>/fs/cifs
27
282) make menuconfig (or make xconfig)
293) select cifs from within the network filesystem choices
304) save and exit
315) make dep
326) make modules (or "make" if CIFS VFS not to be built as a module)
33
34For Linux 2.6:
Adrian Bunkdfc1e142005-05-05 16:15:51 -0700351) Download the kernel (e.g. from http://www.kernel.org)
36and change directory into the top of the kernel directory tree
37(e.g. /usr/src/linux-2.5.73)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700382) make menuconfig (or make xconfig)
393) select cifs from within the network filesystem choices
404) save and exit
415) make
42
43
44Installation instructions:
45=========================
46If you have built the CIFS vfs as module (successfully) simply
47type "make modules_install" (or if you prefer, manually copy the file to
48the modules directory e.g. /lib/modules/2.4.10-4GB/kernel/fs/cifs/cifs.o).
49
50If you have built the CIFS vfs into the kernel itself, follow the instructions
51for your distribution on how to install a new kernel (usually you
52would simply type "make install").
53
54If you do not have the utility mount.cifs (in the Samba 3.0 source tree and on
55the CIFS VFS web site) copy it to the same directory in which mount.smbfs and
56similar files reside (usually /sbin). Although the helper software is not
57required, mount.cifs is recommended. Eventually the Samba 3.0 utility program
58"net" may also be helpful since it may someday provide easier mount syntax for
59users who are used to Windows e.g. net use <mount point> <UNC name or cifs URL>
60Note that running the Winbind pam/nss module (logon service) on all of your
61Linux clients is useful in mapping Uids and Gids consistently across the
62domain to the proper network user. The mount.cifs mount helper can be
63trivially built from Samba 3.0 or later source e.g. by executing:
64
65 gcc samba/source/client/mount.cifs.c -o mount.cifs
66
67If cifs is built as a module, then the size and number of network buffers
68and maximum number of simultaneous requests to one server can be configured.
69Changing these from their defaults is not recommended. By executing modinfo
70 modinfo kernel/fs/cifs/cifs.ko
71on kernel/fs/cifs/cifs.ko the list of configuration changes that can be made
72at module initialization time (by running insmod cifs.ko) can be seen.
73
74Allowing User Mounts
75====================
76To permit users to mount and unmount over directories they own is possible
77with the cifs vfs. A way to enable such mounting is to mark the mount.cifs
Steve French099a58f2005-04-28 22:41:07 -070078utility as suid (e.g. "chmod +s /sbin/mount.cifs). To enable users to
Linus Torvalds1da177e2005-04-16 15:20:36 -070079umount shares they mount requires
801) mount.cifs version 1.4 or later
812) an entry for the share in /etc/fstab indicating that a user may
82unmount it e.g.
83//server/usersharename /mnt/username cifs user 0 0
84
85Note that when the mount.cifs utility is run suid (allowing user mounts),
86in order to reduce risks, the "nosuid" mount flag is passed in on mount to
87disallow execution of an suid program mounted on the remote target.
88When mount is executed as root, nosuid is not passed in by default,
89and execution of suid programs on the remote target would be enabled
90by default. This can be changed, as with nfs and other filesystems,
91by simply specifying "nosuid" among the mount options. For user mounts
92though to be able to pass the suid flag to mount requires rebuilding
93mount.cifs with the following flag:
94
95 gcc samba/source/client/mount.cifs.c -DCIFS_ALLOW_USR_SUID -o mount.cifs
96
97There is a corresponding manual page for cifs mounting in the Samba 3.0 and
98later source tree in docs/manpages/mount.cifs.8
99
Steve French099a58f2005-04-28 22:41:07 -0700100Allowing User Unmounts
101======================
102To permit users to ummount directories that they have user mounted (see above),
103the utility umount.cifs may be used. It may be invoked directly, or if
Steve French0cb766a2005-04-28 22:41:11 -0700104umount.cifs is placed in /sbin, umount can invoke the cifs umount helper
Steve French099a58f2005-04-28 22:41:07 -0700105(at least for most versions of the umount utility) for umount of cifs
Steve French0cb766a2005-04-28 22:41:11 -0700106mounts, unless umount is invoked with -i (which will avoid invoking a umount
107helper). As with mount.cifs, to enable user unmounts umount.cifs must be marked
108as suid (e.g. "chmod +s /sbin/umount.cifs") or equivalent (some distributions
109allow adding entries to a file to the /etc/permissions file to achieve the
110equivalent suid effect). For this utility to succeed the target path
111must be a cifs mount, and the uid of the current user must match the uid
112of the user who mounted the resource.
Steve French099a58f2005-04-28 22:41:07 -0700113
114Also note that the customary way of allowing user mounts and unmounts is
115(instead of using mount.cifs and unmount.cifs as suid) to add a line
116to the file /etc/fstab for each //server/share you wish to mount, but
117this can become unwieldy when potential mount targets include many
118or unpredictable UNC names.
119
Linus Torvalds1da177e2005-04-16 15:20:36 -0700120Samba Considerations
121====================
122To get the maximum benefit from the CIFS VFS, we recommend using a server that
123supports the SNIA CIFS Unix Extensions standard (e.g. Samba 2.2.5 or later or
124Samba 3.0) but the CIFS vfs works fine with a wide variety of CIFS servers.
125Note that uid, gid and file permissions will display default values if you do
126not have a server that supports the Unix extensions for CIFS (such as Samba
1272.2.5 or later). To enable the Unix CIFS Extensions in the Samba server, add
128the line:
129
130 unix extensions = yes
131
132to your smb.conf file on the server. Note that the following smb.conf settings
133are also useful (on the Samba server) when the majority of clients are Unix or
134Linux:
135
136 case sensitive = yes
137 delete readonly = yes
138 ea support = yes
139
140Note that server ea support is required for supporting xattrs from the Linux
141cifs client, and that EA support is present in later versions of Samba (e.g.
1423.0.6 and later (also EA support works in all versions of Windows, at least to
143shares on NTFS filesystems). Extended Attribute (xattr) support is an optional
144feature of most Linux filesystems which may require enabling via
145make menuconfig. Client support for extended attributes (user xattr) can be
146disabled on a per-mount basis by specifying "nouser_xattr" on mount.
147
148The CIFS client can get and set POSIX ACLs (getfacl, setfacl) to Samba servers
149version 3.10 and later. Setting POSIX ACLs requires enabling both XATTR and
150then POSIX support in the CIFS configuration options when building the cifs
151module. POSIX ACL support can be disabled on a per mount basic by specifying
152"noacl" on mount.
153
154Some administrators may want to change Samba's smb.conf "map archive" and
155"create mask" parameters from the default. Unless the create mask is changed
156newly created files can end up with an unnecessarily restrictive default mode,
157which may not be what you want, although if the CIFS Unix extensions are
158enabled on the server and client, subsequent setattr calls (e.g. chmod) can
159fix the mode. Note that creating special devices (mknod) remotely
160may require specifying a mkdev function to Samba if you are not using
161Samba 3.0.6 or later. For more information on these see the manual pages
162("man smb.conf") on the Samba server system. Note that the cifs vfs,
163unlike the smbfs vfs, does not read the smb.conf on the client system
164(the few optional settings are passed in on mount via -o parameters instead).
165Note that Samba 2.2.7 or later includes a fix that allows the CIFS VFS to delete
166open files (required for strict POSIX compliance). Windows Servers already
167supported this feature. Samba server does not allow symlinks that refer to files
168outside of the share, so in Samba versions prior to 3.0.6, most symlinks to
169files with absolute paths (ie beginning with slash) such as:
170 ln -s /mnt/foo bar
171would be forbidden. Samba 3.0.6 server or later includes the ability to create
172such symlinks safely by converting unsafe symlinks (ie symlinks to server
173files that are outside of the share) to a samba specific format on the server
174that is ignored by local server applications and non-cifs clients and that will
175not be traversed by the Samba server). This is opaque to the Linux client
176application using the cifs vfs. Absolute symlinks will work to Samba 3.0.5 or
177later, but only for remote clients using the CIFS Unix extensions, and will
178be invisbile to Windows clients and typically will not affect local
179applications running on the same server as Samba.
180
181Use instructions:
182================
183Once the CIFS VFS support is built into the kernel or installed as a module
184(cifs.o), you can use mount syntax like the following to access Samba or Windows
185servers:
186
187 mount -t cifs //9.53.216.11/e$ /mnt -o user=myname,pass=mypassword
188
189Before -o the option -v may be specified to make the mount.cifs
190mount helper display the mount steps more verbosely.
191After -o the following commonly used cifs vfs specific options
192are supported:
193
194 user=<username>
195 pass=<password>
196 domain=<domain name>
197
198Other cifs mount options are described below. Use of TCP names (in addition to
199ip addresses) is available if the mount helper (mount.cifs) is installed. If
200you do not trust the server to which are mounted, or if you do not have
201cifs signing enabled (and the physical network is insecure), consider use
202of the standard mount options "noexec" and "nosuid" to reduce the risk of
203running an altered binary on your local system (downloaded from a hostile server
204or altered by a hostile router).
205
206Although mounting using format corresponding to the CIFS URL specification is
207not possible in mount.cifs yet, it is possible to use an alternate format
208for the server and sharename (which is somewhat similar to NFS style mount
209syntax) instead of the more widely used UNC format (i.e. \\server\share):
210 mount -t cifs tcp_name_of_server:share_name /mnt -o user=myname,pass=mypasswd
211
212When using the mount helper mount.cifs, passwords may be specified via alternate
213mechanisms, instead of specifying it after -o using the normal "pass=" syntax
214on the command line:
2151) By including it in a credential file. Specify credentials=filename as one
216of the mount options. Credential files contain two lines
217 username=someuser
218 password=your_password
2192) By specifying the password in the PASSWD environment variable (similarly
220the user name can be taken from the USER environment variable).
2213) By specifying the password in a file by name via PASSWD_FILE
2224) By specifying the password in a file by file descriptor via PASSWD_FD
223
224If no password is provided, mount.cifs will prompt for password entry
225
226Restrictions
227============
228Servers must support the NTLM SMB dialect (which is the most recent, supported
229by Samba and Windows NT version 4, 2000 and XP and many other SMB/CIFS servers)
230Servers must support either "pure-TCP" (port 445 TCP/IP CIFS connections) or RFC
2311001/1002 support for "Netbios-Over-TCP/IP." Neither of these is likely to be a
232problem as most servers support this. IPv6 support is planned for the future,
233and is almost complete.
234
235Valid filenames differ between Windows and Linux. Windows typically restricts
236filenames which contain certain reserved characters (e.g.the character :
237which is used to delimit the beginning of a stream name by Windows), while
238Linux allows a slightly wider set of valid characters in filenames. Windows
239servers can remap such characters when an explicit mapping is specified in
240the Server's registry. Samba starting with version 3.10 will allow such
241filenames (ie those which contain valid Linux characters, which normally
242would be forbidden for Windows/CIFS semantics) as long as the server is
243configured for Unix Extensions (and the client has not disabled
244/proc/fs/cifs/LinuxExtensionsEnabled).
245
246
247CIFS VFS Mount Options
248======================
249A partial list of the supported mount options follows:
250 user The user name to use when trying to establish
251 the CIFS session.
252 password The user password. If the mount helper is
253 installed, the user will be prompted for password
254 if it is not supplied.
255 ip The ip address of the target server
256 unc The target server Universal Network Name (export) to
257 mount.
258 domain Set the SMB/CIFS workgroup name prepended to the
259 username during CIFS session establishment
Steve French4523cc32007-04-30 20:13:06 +0000260 uid Set the default uid for inodes. For mounts to servers
261 which do support the CIFS Unix extensions, such as a
262 properly configured Samba server, the server provides
263 the uid, gid and mode so this parameter should not be
264 specified unless the server and clients uid and gid
265 numbering differ. If the server and client are in the
266 same domain (e.g. running winbind or nss_ldap) and
267 the server supports the Unix Extensions then the uid
268 and gid can be retrieved from the server (and uid
269 and gid would not have to be specifed on the mount.
270 For servers which do not support the CIFS Unix
271 extensions, the default uid (and gid) returned on lookup
272 of existing files will be the uid (gid) of the person
Linus Torvalds1da177e2005-04-16 15:20:36 -0700273 who executed the mount (root, except when mount.cifs
274 is configured setuid for user mounts) unless the "uid="
275 (gid) mount option is specified. For the uid (gid) of newly
276 created files and directories, ie files created since
277 the last mount of the server share, the expected uid
Matt LaPlantecab00892006-10-03 22:36:44 +0200278 (gid) is cached as long as the inode remains in
Linus Torvalds1da177e2005-04-16 15:20:36 -0700279 memory on the client. Also note that permission
280 checks (authorization checks) on accesses to a file occur
281 at the server, but there are cases in which an administrator
282 may want to restrict at the client as well. For those
283 servers which do not report a uid/gid owner
284 (such as Windows), permissions can also be checked at the
285 client, and a crude form of client side permission checking
286 can be enabled by specifying file_mode and dir_mode on
Steve French6473a552005-11-29 20:20:10 -0800287 the client. Note that the mount.cifs helper must be
288 at version 1.10 or higher to support specifying the uid
289 (or gid) in non-numberic form.
Steve French4523cc32007-04-30 20:13:06 +0000290 gid Set the default gid for inodes (similar to above).
Linus Torvalds1da177e2005-04-16 15:20:36 -0700291 file_mode If CIFS Unix extensions are not supported by the server
292 this overrides the default mode for file inodes.
293 dir_mode If CIFS Unix extensions are not supported by the server
294 this overrides the default mode for directory inodes.
295 port attempt to contact the server on this tcp port, before
296 trying the usual ports (port 445, then 139).
297 iocharset Codepage used to convert local path names to and from
298 Unicode. Unicode is used by default for network path
299 names if the server supports it. If iocharset is
300 not specified then the nls_default specified
301 during the local client kernel build will be used.
302 If server does not support Unicode, this parameter is
303 unused.
Steve French75865f8c2007-06-24 18:30:48 +0000304 rsize default read size (usually 16K). The client currently
305 can not use rsize larger than CIFSMaxBufSize. CIFSMaxBufSize
306 defaults to 16K and may be changed (from 8K to the maximum
307 kmalloc size allowed by your kernel) at module install time
308 for cifs.ko. Setting CIFSMaxBufSize to a very large value
309 will cause cifs to use more memory and may reduce performance
310 in some cases. To use rsize greater than 127K (the original
311 cifs protocol maximum) also requires that the server support
312 a new Unix Capability flag (for very large read) which some
313 newer servers (e.g. Samba 3.0.26 or later) do. rsize can be
314 set from a minimum of 2048 to a maximum of 130048 (127K or
315 CIFSMaxBufSize, whichever is smaller)
316 wsize default write size (default 57344)
317 maximum wsize currently allowed by CIFS is 57344 (fourteen
318 4096 byte pages)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700319 rw mount the network share read-write (note that the
320 server may still consider the share read-only)
321 ro mount network share read-only
322 version used to distinguish different versions of the
323 mount helper utility (not typically needed)
324 sep if first mount option (after the -o), overrides
325 the comma as the separator between the mount
326 parms. e.g.
327 -o user=myname,password=mypassword,domain=mydom
328 could be passed instead with period as the separator by
329 -o sep=.user=myname.password=mypassword.domain=mydom
330 this might be useful when comma is contained within username
331 or password or domain. This option is less important
332 when the cifs mount helper cifs.mount (version 1.1 or later)
333 is used.
334 nosuid Do not allow remote executables with the suid bit
335 program to be executed. This is only meaningful for mounts
336 to servers such as Samba which support the CIFS Unix Extensions.
337 If you do not trust the servers in your network (your mount
338 targets) it is recommended that you specify this option for
339 greater security.
340 exec Permit execution of binaries on the mount.
341 noexec Do not permit execution of binaries on the mount.
342 dev Recognize block devices on the remote mount.
343 nodev Do not recognize devices on the remote mount.
344 suid Allow remote files on this mountpoint with suid enabled to
345 be executed (default for mounts when executed as root,
346 nosuid is default for user mounts).
347 credentials Although ignored by the cifs kernel component, it is used by
348 the mount helper, mount.cifs. When mount.cifs is installed it
349 opens and reads the credential file specified in order
350 to obtain the userid and password arguments which are passed to
351 the cifs vfs.
352 guest Although ignored by the kernel component, the mount.cifs
353 mount helper will not prompt the user for a password
354 if guest is specified on the mount options. If no
355 password is specified a null password will be used.
356 perm Client does permission checks (vfs_permission check of uid
357 and gid of the file against the mode and desired operation),
358 Note that this is in addition to the normal ACL check on the
359 target machine done by the server software.
360 Client permission checking is enabled by default.
361 noperm Client does not do permission checks. This can expose
362 files on this mount to access by other users on the local
363 client system. It is typically only needed when the server
364 supports the CIFS Unix Extensions but the UIDs/GIDs on the
365 client and server system do not match closely enough to allow
Steve French6473a552005-11-29 20:20:10 -0800366 access by the user doing the mount, but it may be useful with
367 non CIFS Unix Extension mounts for cases in which the default
368 mode is specified on the mount but is not to be enforced on the
369 client (e.g. perhaps when MultiUserMount is enabled)
Linus Torvalds1da177e2005-04-16 15:20:36 -0700370 Note that this does not affect the normal ACL check on the
371 target machine done by the server software (of the server
372 ACL against the user name provided at mount time).
Steve French7521a3c2007-07-11 18:30:34 +0000373 serverino Use server's inode numbers instead of generating automatically
Linus Torvalds1da177e2005-04-16 15:20:36 -0700374 incrementing inode numbers on the client. Although this will
375 make it easier to spot hardlinked files (as they will have
376 the same inode numbers) and inode numbers may be persistent,
377 note that the server does not guarantee that the inode numbers
378 are unique if multiple server side mounts are exported under a
379 single share (since inode numbers on the servers might not
380 be unique if multiple filesystems are mounted under the same
Steve French7521a3c2007-07-11 18:30:34 +0000381 shared higher level directory). Note that some older
382 (e.g. pre-Windows 2000) do not support returning UniqueIDs
383 or the CIFS Unix Extensions equivalent and for those
384 this mount option will have no effect. Exporting cifs mounts
385 under nfsd requires this mount option on the cifs mount.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700386 noserverino Client generates inode numbers (rather than using the actual one
387 from the server) by default.
388 setuids If the CIFS Unix extensions are negotiated with the server
389 the client will attempt to set the effective uid and gid of
390 the local process on newly created files, directories, and
Steve French6473a552005-11-29 20:20:10 -0800391 devices (create, mkdir, mknod). If the CIFS Unix Extensions
392 are not negotiated, for newly created files and directories
Matt LaPlantecab00892006-10-03 22:36:44 +0200393 instead of using the default uid and gid specified on
Steve French6473a552005-11-29 20:20:10 -0800394 the mount, cache the new file's uid and gid locally which means
395 that the uid for the file can change when the inode is
396 reloaded (or the user remounts the share).
Linus Torvalds1da177e2005-04-16 15:20:36 -0700397 nosetuids The client will not attempt to set the uid and gid on
398 on newly created files, directories, and devices (create,
399 mkdir, mknod) which will result in the server setting the
400 uid and gid to the default (usually the server uid of the
Steve French67594fe2005-05-17 13:04:49 -0500401 user who mounted the share). Letting the server (rather than
Steve French6473a552005-11-29 20:20:10 -0800402 the client) set the uid and gid is the default. If the CIFS
403 Unix Extensions are not negotiated then the uid and gid for
404 new files will appear to be the uid (gid) of the mounter or the
405 uid (gid) parameter specified on the mount.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700406 netbiosname When mounting to servers via port 139, specifies the RFC1001
407 source name to use to represent the client netbios machine
408 name when doing the RFC1001 netbios session initialize.
409 direct Do not do inode data caching on files opened on this mount.
410 This precludes mmaping files on this mount. In some cases
411 with fast networks and little or no caching benefits on the
412 client (e.g. when the application is doing large sequential
413 reads bigger than page size without rereading the same data)
414 this can provide better performance than the default
Steve French67594fe2005-05-17 13:04:49 -0500415 behavior which caches reads (readahead) and writes
Linus Torvalds1da177e2005-04-16 15:20:36 -0700416 (writebehind) through the local Linux client pagecache
417 if oplock (caching token) is granted and held. Note that
418 direct allows write operations larger than page size
419 to be sent to the server.
420 acl Allow setfacl and getfacl to manage posix ACLs if server
421 supports them. (default)
422 noacl Do not allow setfacl and getfacl calls on this mount
423 user_xattr Allow getting and setting user xattrs as OS/2 EAs (extended
424 attributes) to the server (default) e.g. via setfattr
425 and getfattr utilities.
Steve Frenchea4c07d2006-08-16 19:44:25 +0000426 nouser_xattr Do not allow getfattr/setfattr to get/set/list xattrs
Steve French737b7582005-04-28 22:41:06 -0700427 mapchars Translate six of the seven reserved characters (not backslash)
428 *?<>|:
Steve French6a0b4822005-04-28 22:41:05 -0700429 to the remap range (above 0xF000), which also
430 allows the CIFS client to recognize files created with
431 such characters by Windows's POSIX emulation. This can
432 also be useful when mounting to most versions of Samba
433 (which also forbids creating and opening files
434 whose names contain any of these seven characters).
435 This has no effect if the server does not support
436 Unicode on the wire.
437 nomapchars Do not translate any of these seven characters (default).
Steve Frenchc46fa8a2005-08-18 20:49:57 -0700438 nocase Request case insensitive path name matching (case
439 sensitive is the default if the server suports it).
Steve French82940a42006-03-02 03:24:57 +0000440 posixpaths If CIFS Unix extensions are supported, attempt to
441 negotiate posix path name support which allows certain
442 characters forbidden in typical CIFS filenames, without
443 requiring remapping. (default)
444 noposixpaths If CIFS Unix extensions are supported, do not request
445 posix path name support (this may cause servers to
446 reject creatingfile with certain reserved characters).
Steve Frencha403a0a2007-07-26 15:54:16 +0000447 nounix Disable the CIFS Unix Extensions for this mount (tree
448 connection). This is rarely needed, but it may be useful
449 in order to turn off multiple settings all at once (ie
450 posix acls, posix locks, posix paths, symlink support
451 and retrieving uids/gids/mode from the server) or to
452 work around a bug in server which implement the Unix
453 Extensions.
Steve Frenchc46fa8a2005-08-18 20:49:57 -0700454 nobrl Do not send byte range lock requests to the server.
455 This is necessary for certain applications that break
456 with cifs style mandatory byte range locks (and most
457 cifs servers do not yet support requesting advisory
458 byte range locks).
Steve French0cb766a2005-04-28 22:41:11 -0700459 remount remount the share (often used to change from ro to rw mounts
460 or vice versa)
Steve French6473a552005-11-29 20:20:10 -0800461 sfu When the CIFS Unix Extensions are not negotiated, attempt to
462 create device files and fifos in a format compatible with
463 Services for Unix (SFU). In addition retrieve bits 10-12
464 of the mode via the SETFILEBITS extended attribute (as
Matt LaPlantecab00892006-10-03 22:36:44 +0200465 SFU does). In the future the bottom 9 bits of the
Steve French6473a552005-11-29 20:20:10 -0800466 mode also will be emulated using queries of the security
467 descriptor (ACL).
Steve French750d1152006-06-27 06:28:30 +0000468 sign Must use packet signing (helps avoid unwanted data modification
469 by intermediate systems in the route). Note that signing
470 does not work with lanman or plaintext authentication.
471 sec Security mode. Allowed values are:
Steve Frenchbf820672005-12-01 22:32:42 -0800472 none attempt to connection as a null user (no name)
473 krb5 Use Kerberos version 5 authentication
474 krb5i Use Kerberos authentication and packet signing
475 ntlm Use NTLM password hashing (default)
476 ntlmi Use NTLM password hashing with signing (if
477 /proc/fs/cifs/PacketSigningEnabled on or if
478 server requires signing also can be the default)
479 ntlmv2 Use NTLMv2 password hashing
480 ntlmv2i Use NTLMv2 password hashing with packet signing
Steve French189acaa2006-06-23 02:33:48 +0000481 lanman (if configured in kernel config) use older
482 lanman hash
Steve Frenchbf820672005-12-01 22:32:42 -0800483
Linus Torvalds1da177e2005-04-16 15:20:36 -0700484The mount.cifs mount helper also accepts a few mount options before -o
485including:
486
487 -S take password from stdin (equivalent to setting the environment
488 variable "PASSWD_FD=0"
489 -V print mount.cifs version
490 -? display simple usage information
491
Jeff Layton8426c392007-05-05 03:27:49 +0000492With most 2.6 kernel versions of modutils, the version of the cifs kernel
Linus Torvalds1da177e2005-04-16 15:20:36 -0700493module can be displayed via modinfo.
494
495Misc /proc/fs/cifs Flags and Debug Info
496=======================================
497Informational pseudo-files:
498DebugData Displays information about active CIFS sessions
Steve French09d1db52005-04-28 22:41:08 -0700499 and shares, as well as the cifs.ko version.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700500Stats Lists summary resource usage information as well as per
501 share statistics, if CONFIG_CIFS_STATS in enabled
502 in the kernel configuration.
503
504Configuration pseudo-files:
505MultiuserMount If set to one, more than one CIFS session to
506 the same server ip address can be established
507 if more than one uid accesses the same mount
508 point and if the uids user/password mapping
509 information is available. (default is 0)
510PacketSigningEnabled If set to one, cifs packet signing is enabled
511 and will be used if the server requires
512 it. If set to two, cifs packet signing is
513 required even if the server considers packet
514 signing optional. (default 1)
Steve French254e55e2006-06-04 05:53:15 +0000515SecurityFlags Flags which control security negotiation and
516 also packet signing. Authentication (may/must)
517 flags (e.g. for NTLM and/or NTLMv2) may be combined with
518 the signing flags. Specifying two different password
519 hashing mechanisms (as "must use") on the other hand
520 does not make much sense. Default flags are
521 0x07007
522 (NTLM, NTLMv2 and packet signing allowed). Maximum
523 allowable flags if you want to allow mounts to servers
524 using weaker password hashes is 0x37037 (lanman,
525 plaintext, ntlm, ntlmv2, signing allowed):
526
527 may use packet signing 0x00001
528 must use packet signing 0x01001
529 may use NTLM (most common password hash) 0x00002
530 must use NTLM 0x02002
531 may use NTLMv2 0x00004
532 must use NTLMv2 0x04004
533 may use Kerberos security (not implemented yet) 0x00008
534 must use Kerberos (not implemented yet) 0x08008
535 may use lanman (weak) password hash 0x00010
536 must use lanman password hash 0x10010
537 may use plaintext passwords 0x00020
538 must use plaintext passwords 0x20020
539 (reserved for future packet encryption) 0x00040
540
Jeff Layton8426c392007-05-05 03:27:49 +0000541cifsFYI If set to non-zero value, additional debug information
542 will be logged to the system error log. This field
543 contains three flags controlling different classes of
544 debugging entries. The maximum value it can be set
545 to is 7 which enables all debugging points (default 0).
546 Some debugging statements are not compiled into the
547 cifs kernel unless CONFIG_CIFS_DEBUG2 is enabled in the
548 kernel configuration. cifsFYI may be set to one or
549 nore of the following flags (7 sets them all):
550
551 log cifs informational messages 0x01
552 log return codes from cifs entry points 0x02
Steve French0ec54aa2007-05-05 22:08:06 +0000553 log slow responses (ie which take longer than 1 second)
554 CONFIG_CIFS_STATS2 must be enabled in .config 0x04
Jeff Layton8426c392007-05-05 03:27:49 +0000555
556
Linus Torvalds1da177e2005-04-16 15:20:36 -0700557traceSMB If set to one, debug information is logged to the
558 system error log with the start of smb requests
559 and responses (default 0)
560LookupCacheEnable If set to one, inode information is kept cached
561 for one second improving performance of lookups
562 (default 1)
563OplockEnabled If set to one, safe distributed caching enabled.
564 (default 1)
565LinuxExtensionsEnabled If set to one then the client will attempt to
566 use the CIFS "UNIX" extensions which are optional
567 protocol enhancements that allow CIFS servers
568 to return accurate UID/GID information as well
569 as support symbolic links. If you use servers
570 such as Samba that support the CIFS Unix
571 extensions but do not want to use symbolic link
572 support and want to map the uid and gid fields
573 to values supplied at mount (rather than the
574 actual values, then set this to zero. (default 1)
Steve French60808232006-04-22 15:53:05 +0000575Experimental When set to 1 used to enable certain experimental
576 features (currently enables multipage writes
577 when signing is enabled, the multipage write
578 performance enhancement was disabled when
579 signing turned on in case buffer was modified
580 just before it was sent, also this flag will
581 be used to use the new experimental sessionsetup
582 code).
Linus Torvalds1da177e2005-04-16 15:20:36 -0700583
584These experimental features and tracing can be enabled by changing flags in
585/proc/fs/cifs (after the cifs module has been installed or built into the
586kernel, e.g. insmod cifs). To enable a feature set it to 1 e.g. to enable
587tracing to the kernel message log type:
588
Steve French1047abc2005-10-11 19:58:06 -0700589 echo 7 > /proc/fs/cifs/cifsFYI
Linus Torvalds1da177e2005-04-16 15:20:36 -0700590
Steve French1047abc2005-10-11 19:58:06 -0700591cifsFYI functions as a bit mask. Setting it to 1 enables additional kernel
592logging of various informational messages. 2 enables logging of non-zero
593SMB return codes while 4 enables logging of requests that take longer
594than one second to complete (except for byte range lock requests).
595Setting it to 4 requires defining CONFIG_CIFS_STATS2 manually in the
596source code (typically by setting it in the beginning of cifsglob.h),
597and setting it to seven enables all three. Finally, tracing
598the start of smb requests and responses can be enabled via:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700599
600 echo 1 > /proc/fs/cifs/traceSMB
601
Steve French75865f8c2007-06-24 18:30:48 +0000602Two other experimental features are under development. To test these
603requires enabling CONFIG_CIFS_EXPERIMENTAL
Linus Torvalds1da177e2005-04-16 15:20:36 -0700604
Steve French75865f8c2007-06-24 18:30:48 +0000605 ipv6 enablement
Linus Torvalds1da177e2005-04-16 15:20:36 -0700606
607 DNOTIFY fcntl: needed for support of directory change
608 notification and perhaps later for file leases)
609
610Per share (per client mount) statistics are available in /proc/fs/cifs/Stats
611if the kernel was configured with cifs statistics enabled. The statistics
612represent the number of successful (ie non-zero return code from the server)
613SMB responses to some of the more common commands (open, delete, mkdir etc.).
614Also recorded is the total bytes read and bytes written to the server for
615that share. Note that due to client caching effects this can be less than the
616number of bytes read and written by the application running on the client.
617The statistics for the number of total SMBs and oplock breaks are different in
618that they represent all for that share, not just those for which the server
619returned success.
620
621Also note that "cat /proc/fs/cifs/DebugData" will display information about
622the active sessions and the shares that are mounted. Note: NTLMv2 enablement
Steve French09d1db52005-04-28 22:41:08 -0700623will not work since its implementation is not quite complete yet. Do not alter
624the ExtendedSecurity configuration value unless you are doing specific testing.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700625Enabling extended security works to Windows 2000 Workstations and XP but not to
626Windows 2000 server or Samba since it does not usually send "raw NTLMSSP"
627(instead it sends NTLMSSP encapsulated in SPNEGO/GSSAPI, which support is not
628complete in the CIFS VFS yet).