1 The CIFS VFS support for Linux supports many advanced network filesystem
2 features such as heirarchical dfs like namespace, hardlinks, locking and more.
3 It was designed to comply with the SNIA CIFS Technical Reference (which
4 supersedes the 1992 X/Open SMB Standard) as well as to perform best practice
5 practical interoperability with Windows 2000, Windows XP, Samba and equivalent
8 For questions or bug reports please contact:
9 sfrench@samba.org (sfrench@us.ibm.com)
14 1) Get the kernel source (e.g.from http://www.kernel.org)
15 and download the cifs vfs source (see the project page
16 at http://us1.samba.org/samba/Linux_CIFS_client.html)
17 and change directory into the top of the kernel directory
18 then patch the kernel (e.g. "patch -p1 < cifs_24.patch")
19 to add the cifs vfs to your kernel configure options if
20 it has not already been added (e.g. current SuSE and UL
21 users do not need to apply the cifs_24.patch since the cifs vfs is
22 already in the kernel configure menu) and then
23 mkdir linux/fs/cifs and then copy the current cifs vfs files from
24 the cifs download to your kernel build directory e.g.
26 cp <cifs_download_dir>/fs/cifs/* to <kernel_download_dir>/fs/cifs
28 2) make menuconfig (or make xconfig)
29 3) select cifs from within the network filesystem choices
32 6) make modules (or "make" if CIFS VFS not to be built as a module)
35 1) Download the kernel (e.g. from http://www.kernel.org)
36 and change directory into the top of the kernel directory tree
37 (e.g. /usr/src/linux-2.5.73)
38 2) make menuconfig (or make xconfig)
39 3) select cifs from within the network filesystem choices
44 Installation instructions:
45 =========================
46 If you have built the CIFS vfs as module (successfully) simply
47 type "make modules_install" (or if you prefer, manually copy the file to
48 the modules directory e.g. /lib/modules/2.4.10-4GB/kernel/fs/cifs/cifs.o).
50 If you have built the CIFS vfs into the kernel itself, follow the instructions
51 for your distribution on how to install a new kernel (usually you
52 would simply type "make install").
54 If you do not have the utility mount.cifs (in the Samba 3.0 source tree and on
55 the CIFS VFS web site) copy it to the same directory in which mount.smbfs and
56 similar files reside (usually /sbin). Although the helper software is not
57 required, 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
59 users who are used to Windows e.g. net use <mount point> <UNC name or cifs URL>
60 Note that running the Winbind pam/nss module (logon service) on all of your
61 Linux clients is useful in mapping Uids and Gids consistently across the
62 domain to the proper network user. The mount.cifs mount helper can be
63 trivially built from Samba 3.0 or later source e.g. by executing:
65 gcc samba/source/client/mount.cifs.c -o mount.cifs
67 If cifs is built as a module, then the size and number of network buffers
68 and maximum number of simultaneous requests to one server can be configured.
69 Changing these from their defaults is not recommended. By executing modinfo
70 modinfo kernel/fs/cifs/cifs.ko
71 on kernel/fs/cifs/cifs.ko the list of configuration changes that can be made
72 at module initialization time (by running insmod cifs.ko) can be seen.
76 To permit users to mount and unmount over directories they own is possible
77 with the cifs vfs. A way to enable such mounting is to mark the mount.cifs
78 utility as suid (e.g. "chmod +s /sbin/mount.cifs). To enable users to
79 umount shares they mount requires
80 1) mount.cifs version 1.4 or later
81 2) an entry for the share in /etc/fstab indicating that a user may
83 //server/usersharename /mnt/username cifs user 0 0
85 Note that when the mount.cifs utility is run suid (allowing user mounts),
86 in order to reduce risks, the "nosuid" mount flag is passed in on mount to
87 disallow execution of an suid program mounted on the remote target.
88 When mount is executed as root, nosuid is not passed in by default,
89 and execution of suid programs on the remote target would be enabled
90 by default. This can be changed, as with nfs and other filesystems,
91 by simply specifying "nosuid" among the mount options. For user mounts
92 though to be able to pass the suid flag to mount requires rebuilding
93 mount.cifs with the following flag:
95 gcc samba/source/client/mount.cifs.c -DCIFS_ALLOW_USR_SUID -o mount.cifs
97 There is a corresponding manual page for cifs mounting in the Samba 3.0 and
98 later source tree in docs/manpages/mount.cifs.8
100 Allowing User Unmounts
101 ======================
102 To permit users to ummount directories that they have user mounted (see above),
103 the utility umount.cifs may be used. It may be invoked directly, or if
104 umount.cifs is placed in /sbin, umount can invoke the cifs umount helper
105 (at least for most versions of the umount utility) for umount of cifs
106 mounts, unless umount is invoked with -i (which will avoid invoking a umount
107 helper). As with mount.cifs, to enable user unmounts umount.cifs must be marked
108 as suid (e.g. "chmod +s /sbin/umount.cifs") or equivalent (some distributions
109 allow adding entries to a file to the /etc/permissions file to achieve the
110 equivalent suid effect). For this utility to succeed the target path
111 must be a cifs mount, and the uid of the current user must match the uid
112 of the user who mounted the resource.
114 Also 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
116 to the file /etc/fstab for each //server/share you wish to mount, but
117 this can become unwieldy when potential mount targets include many
118 or unpredictable UNC names.
122 To get the maximum benefit from the CIFS VFS, we recommend using a server that
123 supports the SNIA CIFS Unix Extensions standard (e.g. Samba 2.2.5 or later or
124 Samba 3.0) but the CIFS vfs works fine with a wide variety of CIFS servers.
125 Note that uid, gid and file permissions will display default values if you do
126 not have a server that supports the Unix extensions for CIFS (such as Samba
127 2.2.5 or later). To enable the Unix CIFS Extensions in the Samba server, add
130 unix extensions = yes
132 to your smb.conf file on the server. Note that the following smb.conf settings
133 are also useful (on the Samba server) when the majority of clients are Unix or
137 delete readonly = yes
140 Note that server ea support is required for supporting xattrs from the Linux
141 cifs client, and that EA support is present in later versions of Samba (e.g.
142 3.0.6 and later (also EA support works in all versions of Windows, at least to
143 shares on NTFS filesystems). Extended Attribute (xattr) support is an optional
144 feature of most Linux filesystems which may require enabling via
145 make menuconfig. Client support for extended attributes (user xattr) can be
146 disabled on a per-mount basis by specifying "nouser_xattr" on mount.
148 The CIFS client can get and set POSIX ACLs (getfacl, setfacl) to Samba servers
149 version 3.10 and later. Setting POSIX ACLs requires enabling both XATTR and
150 then POSIX support in the CIFS configuration options when building the cifs
151 module. POSIX ACL support can be disabled on a per mount basic by specifying
154 Some 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
156 newly created files can end up with an unnecessarily restrictive default mode,
157 which may not be what you want, although if the CIFS Unix extensions are
158 enabled on the server and client, subsequent setattr calls (e.g. chmod) can
159 fix the mode. Note that creating special devices (mknod) remotely
160 may require specifying a mkdev function to Samba if you are not using
161 Samba 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,
163 unlike 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).
165 Note that Samba 2.2.7 or later includes a fix that allows the CIFS VFS to delete
166 open files (required for strict POSIX compliance). Windows Servers already
167 supported this feature. Samba server does not allow symlinks that refer to files
168 outside of the share, so in Samba versions prior to 3.0.6, most symlinks to
169 files with absolute paths (ie beginning with slash) such as:
171 would be forbidden. Samba 3.0.6 server or later includes the ability to create
172 such symlinks safely by converting unsafe symlinks (ie symlinks to server
173 files that are outside of the share) to a samba specific format on the server
174 that is ignored by local server applications and non-cifs clients and that will
175 not be traversed by the Samba server). This is opaque to the Linux client
176 application using the cifs vfs. Absolute symlinks will work to Samba 3.0.5 or
177 later, but only for remote clients using the CIFS Unix extensions, and will
178 be invisbile to Windows clients and typically will not affect local
179 applications running on the same server as Samba.
183 Once 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
187 mount -t cifs //9.53.216.11/e$ /mnt -o user=myname,pass=mypassword
189 Before -o the option -v may be specified to make the mount.cifs
190 mount helper display the mount steps more verbosely.
191 After -o the following commonly used cifs vfs specific options
198 Other cifs mount options are described below. Use of TCP names (in addition to
199 ip addresses) is available if the mount helper (mount.cifs) is installed. If
200 you do not trust the server to which are mounted, or if you do not have
201 cifs signing enabled (and the physical network is insecure), consider use
202 of the standard mount options "noexec" and "nosuid" to reduce the risk of
203 running an altered binary on your local system (downloaded from a hostile server
204 or altered by a hostile router).
206 Although mounting using format corresponding to the CIFS URL specification is
207 not possible in mount.cifs yet, it is possible to use an alternate format
208 for the server and sharename (which is somewhat similar to NFS style mount
209 syntax) 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
212 When using the mount helper mount.cifs, passwords may be specified via alternate
213 mechanisms, instead of specifying it after -o using the normal "pass=" syntax
215 1) By including it in a credential file. Specify credentials=filename as one
216 of the mount options. Credential files contain two lines
218 password=your_password
219 2) By specifying the password in the PASSWD environment variable (similarly
220 the user name can be taken from the USER environment variable).
221 3) By specifying the password in a file by name via PASSWD_FILE
222 4) By specifying the password in a file by file descriptor via PASSWD_FD
224 If no password is provided, mount.cifs will prompt for password entry
228 Servers must support the NTLM SMB dialect (which is the most recent, supported
229 by Samba and Windows NT version 4, 2000 and XP and many other SMB/CIFS servers)
230 Servers must support either "pure-TCP" (port 445 TCP/IP CIFS connections) or RFC
231 1001/1002 support for "Netbios-Over-TCP/IP." Neither of these is likely to be a
232 problem as most servers support this. IPv6 support is planned for the future,
233 and is almost complete.
235 Valid filenames differ between Windows and Linux. Windows typically restricts
236 filenames which contain certain reserved characters (e.g.the character :
237 which is used to delimit the beginning of a stream name by Windows), while
238 Linux allows a slightly wider set of valid characters in filenames. Windows
239 servers can remap such characters when an explicit mapping is specified in
240 the Server's registry. Samba starting with version 3.10 will allow such
241 filenames (ie those which contain valid Linux characters, which normally
242 would be forbidden for Windows/CIFS semantics) as long as the server is
243 configured for Unix Extensions (and the client has not disabled
244 /proc/fs/cifs/LinuxExtensionsEnabled).
247 CIFS VFS Mount Options
248 ======================
249 A partial list of the supported mount options follows:
250 user The user name to use when trying to establish
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
258 domain Set the SMB/CIFS workgroup name prepended to the
259 username during CIFS session establishment
260 uid If CIFS Unix extensions are not supported by the server
261 this overrides the default uid for inodes. For mounts to
262 servers which do support the CIFS Unix extensions, such
263 as a properly configured Samba server, the server provides
264 the uid, gid and mode. For servers which do not support
265 the Unix extensions, the default uid (and gid) returned on
266 lookup of existing files is the uid (gid) of the person
267 who executed the mount (root, except when mount.cifs
268 is configured setuid for user mounts) unless the "uid="
269 (gid) mount option is specified. For the uid (gid) of newly
270 created files and directories, ie files created since
271 the last mount of the server share, the expected uid
272 (gid) is cached as as long as the inode remains in
273 memory on the client. Also note that permission
274 checks (authorization checks) on accesses to a file occur
275 at the server, but there are cases in which an administrator
276 may want to restrict at the client as well. For those
277 servers which do not report a uid/gid owner
278 (such as Windows), permissions can also be checked at the
279 client, and a crude form of client side permission checking
280 can be enabled by specifying file_mode and dir_mode on
281 the client. Note that the mount.cifs helper must be
282 at version 1.10 or higher to support specifying the uid
283 (or gid) in non-numberic form.
284 gid If CIFS Unix extensions are not supported by the server
285 this overrides the default gid for inodes.
286 file_mode If CIFS Unix extensions are not supported by the server
287 this overrides the default mode for file inodes.
288 dir_mode If CIFS Unix extensions are not supported by the server
289 this overrides the default mode for directory inodes.
290 port attempt to contact the server on this tcp port, before
291 trying the usual ports (port 445, then 139).
292 iocharset Codepage used to convert local path names to and from
293 Unicode. Unicode is used by default for network path
294 names if the server supports it. If iocharset is
295 not specified then the nls_default specified
296 during the local client kernel build will be used.
297 If server does not support Unicode, this parameter is
299 rsize default read size (usually 16K)
300 wsize default write size (usually 16K, 32K is often better over GigE)
301 maximum wsize currently allowed by CIFS is 57344 (14 4096 byte
303 rw mount the network share read-write (note that the
304 server may still consider the share read-only)
305 ro mount network share read-only
306 version used to distinguish different versions of the
307 mount helper utility (not typically needed)
308 sep if first mount option (after the -o), overrides
309 the comma as the separator between the mount
311 -o user=myname,password=mypassword,domain=mydom
312 could be passed instead with period as the separator by
313 -o sep=.user=myname.password=mypassword.domain=mydom
314 this might be useful when comma is contained within username
315 or password or domain. This option is less important
316 when the cifs mount helper cifs.mount (version 1.1 or later)
318 nosuid Do not allow remote executables with the suid bit
319 program to be executed. This is only meaningful for mounts
320 to servers such as Samba which support the CIFS Unix Extensions.
321 If you do not trust the servers in your network (your mount
322 targets) it is recommended that you specify this option for
324 exec Permit execution of binaries on the mount.
325 noexec Do not permit execution of binaries on the mount.
326 dev Recognize block devices on the remote mount.
327 nodev Do not recognize devices on the remote mount.
328 suid Allow remote files on this mountpoint with suid enabled to
329 be executed (default for mounts when executed as root,
330 nosuid is default for user mounts).
331 credentials Although ignored by the cifs kernel component, it is used by
332 the mount helper, mount.cifs. When mount.cifs is installed it
333 opens and reads the credential file specified in order
334 to obtain the userid and password arguments which are passed to
336 guest Although ignored by the kernel component, the mount.cifs
337 mount helper will not prompt the user for a password
338 if guest is specified on the mount options. If no
339 password is specified a null password will be used.
340 perm Client does permission checks (vfs_permission check of uid
341 and gid of the file against the mode and desired operation),
342 Note that this is in addition to the normal ACL check on the
343 target machine done by the server software.
344 Client permission checking is enabled by default.
345 noperm Client does not do permission checks. This can expose
346 files on this mount to access by other users on the local
347 client system. It is typically only needed when the server
348 supports the CIFS Unix Extensions but the UIDs/GIDs on the
349 client and server system do not match closely enough to allow
350 access by the user doing the mount, but it may be useful with
351 non CIFS Unix Extension mounts for cases in which the default
352 mode is specified on the mount but is not to be enforced on the
353 client (e.g. perhaps when MultiUserMount is enabled)
354 Note that this does not affect the normal ACL check on the
355 target machine done by the server software (of the server
356 ACL against the user name provided at mount time).
357 serverino Use servers inode numbers instead of generating automatically
358 incrementing inode numbers on the client. Although this will
359 make it easier to spot hardlinked files (as they will have
360 the same inode numbers) and inode numbers may be persistent,
361 note that the server does not guarantee that the inode numbers
362 are unique if multiple server side mounts are exported under a
363 single share (since inode numbers on the servers might not
364 be unique if multiple filesystems are mounted under the same
365 shared higher level directory). Note that this requires that
366 the server support the CIFS Unix Extensions as other servers
367 do not return a unique IndexNumber on SMB FindFirst (most
368 servers return zero as the IndexNumber). Parameter has no
369 effect to Windows servers and others which do not support the
370 CIFS Unix Extensions.
371 noserverino Client generates inode numbers (rather than using the actual one
372 from the server) by default.
373 setuids If the CIFS Unix extensions are negotiated with the server
374 the client will attempt to set the effective uid and gid of
375 the local process on newly created files, directories, and
376 devices (create, mkdir, mknod). If the CIFS Unix Extensions
377 are not negotiated, for newly created files and directories
378 instead of using the default uid and gid specified on the
379 the mount, cache the new file's uid and gid locally which means
380 that the uid for the file can change when the inode is
381 reloaded (or the user remounts the share).
382 nosetuids The client will not attempt to set the uid and gid on
383 on newly created files, directories, and devices (create,
384 mkdir, mknod) which will result in the server setting the
385 uid and gid to the default (usually the server uid of the
386 user who mounted the share). Letting the server (rather than
387 the client) set the uid and gid is the default. If the CIFS
388 Unix Extensions are not negotiated then the uid and gid for
389 new files will appear to be the uid (gid) of the mounter or the
390 uid (gid) parameter specified on the mount.
391 netbiosname When mounting to servers via port 139, specifies the RFC1001
392 source name to use to represent the client netbios machine
393 name when doing the RFC1001 netbios session initialize.
394 direct Do not do inode data caching on files opened on this mount.
395 This precludes mmaping files on this mount. In some cases
396 with fast networks and little or no caching benefits on the
397 client (e.g. when the application is doing large sequential
398 reads bigger than page size without rereading the same data)
399 this can provide better performance than the default
400 behavior which caches reads (readahead) and writes
401 (writebehind) through the local Linux client pagecache
402 if oplock (caching token) is granted and held. Note that
403 direct allows write operations larger than page size
404 to be sent to the server.
405 acl Allow setfacl and getfacl to manage posix ACLs if server
406 supports them. (default)
407 noacl Do not allow setfacl and getfacl calls on this mount
408 user_xattr Allow getting and setting user xattrs as OS/2 EAs (extended
409 attributes) to the server (default) e.g. via setfattr
410 and getfattr utilities.
411 nouser_xattr Do not allow getfattr/setfattr to get/set xattrs
412 mapchars Translate six of the seven reserved characters (not backslash)
414 to the remap range (above 0xF000), which also
415 allows the CIFS client to recognize files created with
416 such characters by Windows's POSIX emulation. This can
417 also be useful when mounting to most versions of Samba
418 (which also forbids creating and opening files
419 whose names contain any of these seven characters).
420 This has no effect if the server does not support
422 nomapchars Do not translate any of these seven characters (default).
423 nocase Request case insensitive path name matching (case
424 sensitive is the default if the server suports it).
425 nobrl Do not send byte range lock requests to the server.
426 This is necessary for certain applications that break
427 with cifs style mandatory byte range locks (and most
428 cifs servers do not yet support requesting advisory
430 remount remount the share (often used to change from ro to rw mounts
432 sfu When the CIFS Unix Extensions are not negotiated, attempt to
433 create device files and fifos in a format compatible with
434 Services for Unix (SFU). In addition retrieve bits 10-12
435 of the mode via the SETFILEBITS extended attribute (as
436 SFU does). In the future the bottom 9 bits of the mode
437 mode also will be emulated using queries of the security
439 sec Security mode. Allowed values are:
440 none attempt to connection as a null user (no name)
441 krb5 Use Kerberos version 5 authentication
442 krb5i Use Kerberos authentication and packet signing
443 ntlm Use NTLM password hashing (default)
444 ntlmi Use NTLM password hashing with signing (if
445 /proc/fs/cifs/PacketSigningEnabled on or if
446 server requires signing also can be the default)
447 ntlmv2 Use NTLMv2 password hashing
448 ntlmv2i Use NTLMv2 password hashing with packet signing
450 The mount.cifs mount helper also accepts a few mount options before -o
453 -S take password from stdin (equivalent to setting the environment
454 variable "PASSWD_FD=0"
455 -V print mount.cifs version
456 -? display simple usage information
458 With recent 2.6 kernel versions of modutils, the version of the cifs kernel
459 module can be displayed via modinfo.
461 Misc /proc/fs/cifs Flags and Debug Info
462 =======================================
463 Informational pseudo-files:
464 DebugData Displays information about active CIFS sessions
465 and shares, as well as the cifs.ko version.
466 Stats Lists summary resource usage information as well as per
467 share statistics, if CONFIG_CIFS_STATS in enabled
468 in the kernel configuration.
470 Configuration pseudo-files:
471 MultiuserMount If set to one, more than one CIFS session to
472 the same server ip address can be established
473 if more than one uid accesses the same mount
474 point and if the uids user/password mapping
475 information is available. (default is 0)
476 PacketSigningEnabled If set to one, cifs packet signing is enabled
477 and will be used if the server requires
478 it. If set to two, cifs packet signing is
479 required even if the server considers packet
480 signing optional. (default 1)
481 cifsFYI If set to one, additional debug information is
482 logged to the system error log. (default 0)
483 ExtendedSecurity If set to one, SPNEGO session establishment
484 is allowed which enables more advanced
485 secure CIFS session establishment (default 0)
486 NTLMV2Enabled If set to one, more secure password hashes
487 are used when the server supports them and
488 when kerberos is not negotiated (default 0)
489 traceSMB If set to one, debug information is logged to the
490 system error log with the start of smb requests
491 and responses (default 0)
492 LookupCacheEnable If set to one, inode information is kept cached
493 for one second improving performance of lookups
495 OplockEnabled If set to one, safe distributed caching enabled.
497 LinuxExtensionsEnabled If set to one then the client will attempt to
498 use the CIFS "UNIX" extensions which are optional
499 protocol enhancements that allow CIFS servers
500 to return accurate UID/GID information as well
501 as support symbolic links. If you use servers
502 such as Samba that support the CIFS Unix
503 extensions but do not want to use symbolic link
504 support and want to map the uid and gid fields
505 to values supplied at mount (rather than the
506 actual values, then set this to zero. (default 1)
508 These experimental features and tracing can be enabled by changing flags in
509 /proc/fs/cifs (after the cifs module has been installed or built into the
510 kernel, e.g. insmod cifs). To enable a feature set it to 1 e.g. to enable
511 tracing to the kernel message log type:
513 echo 7 > /proc/fs/cifs/cifsFYI
515 cifsFYI functions as a bit mask. Setting it to 1 enables additional kernel
516 logging of various informational messages. 2 enables logging of non-zero
517 SMB return codes while 4 enables logging of requests that take longer
518 than one second to complete (except for byte range lock requests).
519 Setting it to 4 requires defining CONFIG_CIFS_STATS2 manually in the
520 source code (typically by setting it in the beginning of cifsglob.h),
521 and setting it to seven enables all three. Finally, tracing
522 the start of smb requests and responses can be enabled via:
524 echo 1 > /proc/fs/cifs/traceSMB
526 Two other experimental features are under development and to test
527 require enabling CONFIG_CIFS_EXPERIMENTAL
529 More efficient write operations
531 DNOTIFY fcntl: needed for support of directory change
532 notification and perhaps later for file leases)
534 Per share (per client mount) statistics are available in /proc/fs/cifs/Stats
535 if the kernel was configured with cifs statistics enabled. The statistics
536 represent the number of successful (ie non-zero return code from the server)
537 SMB responses to some of the more common commands (open, delete, mkdir etc.).
538 Also recorded is the total bytes read and bytes written to the server for
539 that share. Note that due to client caching effects this can be less than the
540 number of bytes read and written by the application running on the client.
541 The statistics for the number of total SMBs and oplock breaks are different in
542 that they represent all for that share, not just those for which the server
545 Also note that "cat /proc/fs/cifs/DebugData" will display information about
546 the active sessions and the shares that are mounted. Note: NTLMv2 enablement
547 will not work since its implementation is not quite complete yet. Do not alter
548 the ExtendedSecurity configuration value unless you are doing specific testing.
549 Enabling extended security works to Windows 2000 Workstations and XP but not to
550 Windows 2000 server or Samba since it does not usually send "raw NTLMSSP"
551 (instead it sends NTLMSSP encapsulated in SPNEGO/GSSAPI, which support is not
552 complete in the CIFS VFS yet).
projects / linux-2.6 / blob