Skip to content

E
etercifs
Commit 3c073451 authored by Konstantin Baev's avatar Konstantin Baev
Browse files
Options

add docs (README.ETER is empty yet)

parent c280cf3b
Show whitespace changes
Inline Side-by-side
Showing with 1829 additions and 1 deletion
etercifs.spec
View file @ 3c073451
......@@ -21,7 +21,7 @@
Name: etercifs
Version: 4.0.1
Release: alt1
Release: alt2
Summary: Advanced Common Internet File System for Linux with Etersoft extension
......@@ -109,8 +109,13 @@ cp %SOURCE27 %buildroot/%etercifs_src/%src_package_name-2.6.27-%src_2_6_27_versi
%_datadir/%name
%_initdir/%name
%_initdir/%name.outformat
%doc README.ETER AUTHORS CHANGES README TODO
%changelog
* Wed Dec 17 2008 Konstantin Baev <kipruss@altlinux.org> 4.0.1-alt2
- minor design changes in sources code
- add docs
* Tue Dec 16 2008 Konstantin Baev <kipruss@altlinux.org> 4.0.1-alt1
- update all sources: add code, that fixing bug Eter#2929
- update sources/2.6.27 (up to 2.6.27.9)
......
etercifs/AUTHORS 0 → 100644
View file @ 3c073451
Original Author
===============
Steve French (sfrench@samba.org)
The author wishes to express his appreciation and thanks to:
Andrew Tridgell (Samba team) for his early suggestions about smb/cifs VFS
improvements. Thanks to IBM for allowing me time and test resources to pursue
this project, to Jim McDonough from IBM (and the Samba Team) for his help, to
the IBM Linux JFS team for explaining many esoteric Linux filesystem features.
Jeremy Allison of the Samba team has done invaluable work in adding the server
side of the original CIFS Unix extensions and reviewing and implementing
portions of the newer CIFS POSIX extensions into the Samba 3 file server. Thank
Dave Boutcher of IBM Rochester (author of the OS/400 smb/cifs filesystem client)
for proving years ago that very good smb/cifs clients could be done on Unix-like
operating systems. Volker Lendecke, Andrew Tridgell, Urban Widmark, John
Newbigin and others for their work on the Linux smbfs module. Thanks to
the other members of the Storage Network Industry Association CIFS Technical
Workgroup for their work specifying this highly complex protocol and finally
thanks to the Samba team for their technical advice and encouragement.
Patch Contributors
------------------
Zwane Mwaikambo
Andi Kleen
Amrut Joshi
Shobhit Dayal
Sergey Vlasov
Richard Hughes
Yury Umanets
Mark Hamzy (for some of the early cifs IPv6 work)
Domen Puncer
Jesper Juhl (in particular for lots of whitespace/formatting cleanup)
Vince Negri and Dave Stahl (for finding an important caching bug)
Adrian Bunk (kcalloc cleanups)
Miklos Szeredi
Kazeon team for various fixes especially for 2.4 version.
Asser Ferno (Change Notify support)
Shaggy (Dave Kleikamp) for inumerable small fs suggestions and some good cleanup
Igor Mammedov (DFS support)
Test case and Bug Report contributors
-------------------------------------
Thanks to those in the community who have submitted detailed bug reports
and debug of problems they have found: Jochen Dolze, David Blaine,
Rene Scharfe, Martin Josefsson, Alexander Wild, Anthony Liguori,
Lars Muller, Urban Widmark, Massimiliano Ferrero, Howard Owen,
Olaf Kirch, Kieron Briggs, Nick Millington and others. Also special
mention to the Stanford Checker (SWAT) which pointed out many minor
bugs in error paths. Valuable suggestions also have come from Al Viro
and Dave Miller.
And thanks to the IBM LTC and Power test teams and SuSE testers for
finding multiple bugs during excellent stress test runs.
etercifs/CHANGES 0 → 100644
View file @ 3c073451
Version 1.54
------------
Fix premature write failure on congested networks (we would give up
on EAGAIN from the socket too quickly on large writes).
Cifs_mkdir and cifs_create now respect the setgid bit on parent dir.
Fix endian problems in acl (mode from/to cifs acl) on bigendian
architectures. Fix problems with preserving timestamps on copying open
files (e.g. "cp -a") to Windows servers. For mkdir and create honor setgid bit
on parent directory when server supports Unix Extensions but not POSIX
create. Update cifs.upcall version to handle new Kerberos sec flags
(this requires update of cifs.upcall program from Samba). Fix memory leak
on dns_upcall (resolving DFS referralls). Fix plain text password
authentication (requires setting SecurityFlags to 0x30030 to enable
lanman and plain text though). Fix writes to be at correct offset when
file is open with O_APPEND and file is on a directio (forcediretio) mount.
Version 1.53
------------
DFS support added (Microsoft Distributed File System client support needed
for referrals which enable a hierarchical name space among servers).
Disable temporary caching of mode bits to servers which do not support
storing of mode (e.g. Windows servers, when client mounts without cifsacl
mount option) and add new "dynperm" mount option to enable temporary caching
of mode (enable old behavior). Fix hang on mount caused when server crashes
tcp session during negotiate protocol.
Version 1.52
------------
Fix oops on second mount to server when null auth is used.
Enable experimental Kerberos support. Return writebehind errors on flush
and sync so that events like out of disk space get reported properly on
cached files. Fix setxattr failure to certain Samba versions. Fix mount
of second share to disconnected server session (autoreconnect on this).
Add ability to modify cifs acls for handling chmod (when mounted with
cifsacl flag). Fix prefixpath path separator so we can handle mounts
with prefixpaths longer than one directory (one path component) when
mounted to Windows servers. Fix slow file open when cifsacl
enabled. Fix memory leak in FindNext when the SMB call returns -EBADF.
Version 1.51
------------
Fix memory leak in statfs when mounted to very old servers (e.g.
Windows 9x). Add new feature "POSIX open" which allows servers
which support the current POSIX Extensions to provide better semantics
(e.g. delete for open files opened with posix open). Take into
account umask on posix mkdir not just older style mkdir. Add
ability to mount to IPC$ share (which allows CIFS named pipes to be
opened, read and written as if they were files). When 1st tree
connect fails (e.g. due to signing negotiation failure) fix
leak that causes cifsd not to stop and rmmod to fail to cleanup
cifs_request_buffers pool. Fix problem with POSIX Open/Mkdir on
bigendian architectures. Fix possible memory corruption when
EAGAIN returned on kern_recvmsg. Return better error if server
requires packet signing but client has disabled it. When mounted
with cifsacl mount option - mode bits are approximated based
on the contents of the ACL of the file or directory. When cifs
mount helper is missing convert make sure that UNC name
has backslash (not forward slash) between ip address of server
and the share name.
Version 1.50
------------
Fix NTLMv2 signing. NFS server mounted over cifs works (if cifs mount is
done with "serverino" mount option). Add support for POSIX Unlink
(helps with certain sharing violation cases when server such as
Samba supports newer POSIX CIFS Protocol Extensions). Add "nounix"
mount option to allow disabling the CIFS Unix Extensions for just
that mount. Fix hang on spinlock in find_writable_file (race when
reopening file after session crash). Byte range unlock request to
windows server could unlock more bytes (on server copy of file)
than intended if start of unlock request is well before start of
a previous byte range lock that we issued.
Version 1.49
------------
IPv6 support. Enable ipv6 addresses to be passed on mount (put the ipv6
address after the "ip=" mount option, at least until mount.cifs is fixed to
handle DNS host to ipv6 name translation). Accept override of uid or gid
on mount even when Unix Extensions are negotiated (it used to be ignored
when Unix Extensions were ignored). This allows users to override the
default uid and gid for files when they are certain that the uids or
gids on the server do not match those of the client. Make "sec=none"
mount override username (so that null user connection is attempted)
to match what documentation said. Support for very large reads, over 127K,
available to some newer servers (such as Samba 3.0.26 and later but
note that it also requires setting CIFSMaxBufSize at module install
time to a larger value which may hurt performance in some cases).
Make sign option force signing (or fail if server does not support it).
Version 1.48
------------
Fix mtime bouncing around from local idea of last write times to remote time.
Fix hang (in i_size_read) when simultaneous size update of same remote file
on smp system corrupts sequence number. Do not reread unnecessarily partial page
(which we are about to overwrite anyway) when writing out file opened rw.
When DOS attribute of file on non-Unix server's file changes on the server side
from read-only back to read-write, reflect this change in default file mode
(we had been leaving a file's mode read-only until the inode were reloaded).
Allow setting of attribute back to ATTR_NORMAL (removing readonly dos attribute
when archive dos attribute not set and we are changing mode back to writeable
on server which does not support the Unix Extensions). Remove read only dos
attribute on chmod when adding any write permission (ie on any of
user/group/other (not all of user/group/other ie 0222) when
mounted to windows. Add support for POSIX MkDir (slight performance
enhancement and eliminates the network race between the mkdir and set
path info of the mode).
Version 1.47
------------
Fix oops in list_del during mount caused by unaligned string.
Fix file corruption which could occur on some large file
copies caused by writepages page i/o completion bug.
Seek to SEEK_END forces check for update of file size for non-cached
files. Allow file size to be updated on remote extend of locally open,
non-cached file. Fix reconnect to newer Samba servers (or other servers
which support the CIFS Unix/POSIX extensions) so that we again tell the
server the Unix/POSIX cifs capabilities which we support (SetFSInfo).
Add experimental support for new POSIX Open/Mkdir (which returns
stat information on the open, and allows setting the mode).
Version 1.46
------------
Support deep tree mounts. Better support OS/2, Win9x (DOS) time stamps.
Allow null user to be specified on mount ("username="). Do not return
EINVAL on readdir when filldir fails due to overwritten blocksize
(fixes FC problem). Return error in rename 2nd attempt retry (ie report
if rename by handle also fails, after rename by path fails, we were
not reporting whether the retry worked or not). Fix NTLMv2 to
work to Windows servers (mount with option "sec=ntlmv2").
Version 1.45
------------
Do not time out lockw calls when using posix extensions. Do not
time out requests if server still responding reasonably fast
on requests on other threads. Improve POSIX locking emulation,
(lock cancel now works, and unlock of merged range works even
to Windows servers now). Fix oops on mount to lanman servers
(win9x, os/2 etc.) when null password. Do not send listxattr
(SMB to query all EAs) if nouser_xattr specified. Fix SE Linux
problem (instantiate inodes/dentries in right order for readdir).
Version 1.44
------------
Rewritten sessionsetup support, including support for legacy SMB
session setup needed for OS/2 and older servers such as Windows 95 and 98.
Fix oops on ls to OS/2 servers. Add support for level 1 FindFirst
so we can do search (ls etc.) to OS/2. Do not send NTCreateX
or recent levels of FindFirst unless server says it supports NT SMBs
(instead use legacy equivalents from LANMAN dialect). Fix to allow
NTLMv2 authentication support (now can use stronger password hashing
on mount if corresponding /proc/fs/cifs/SecurityFlags is set (0x4004).
Allow override of global cifs security flags on mount via "sec=" option(s).
Version 1.43
------------
POSIX locking to servers which support CIFS POSIX Extensions
(disabled by default controlled by proc/fs/cifs/Experimental).
Handle conversion of long share names (especially Asian languages)
to Unicode during mount. Fix memory leak in sess struct on reconnect.
Fix rare oops after acpi suspend. Fix O_TRUNC opens to overwrite on
cifs open which helps rare case when setpathinfo fails or server does
not support it.
Version 1.42
------------
Fix slow oplock break when mounted to different servers at the same time and
the tids match and we try to find matching fid on wrong server. Fix read
looping when signing required by server (2.6.16 kernel only). Fix readdir
vs. rename race which could cause each to hang. Return . and .. even
if server does not. Allow searches to skip first three entries and
begin at any location. Fix oops in find_writeable_file.
Version 1.41
------------
Fix NTLMv2 security (can be enabled in /proc/fs/cifs) so customers can
configure stronger authentication. Fix sfu symlinks so they can
be followed (not just recognized). Fix wraparound of bcc on
read responses when buffer size over 64K and also fix wrap of
max smb buffer size when CIFSMaxBufSize over 64K. Fix oops in
cifs_user_read and cifs_readpages (when EAGAIN on send of smb
on socket is returned over and over). Add POSIX (advisory) byte range
locking support (requires server with newest CIFS UNIX Extensions
to the protocol implemented). Slow down negprot slightly in port 139
RFC1001 case to give session_init time on buggy servers.
Version 1.40
------------
Use fsuid (fsgid) more consistently instead of uid (gid). Improve performance
of readpages by eliminating one extra memcpy. Allow update of file size
from remote server even if file is open for write as long as mount is
directio. Recognize share mode security and send NTLM encrypted password
on tree connect if share mode negotiated.
Version 1.39
------------
Defer close of a file handle slightly if pending writes depend on that handle
(this reduces the EBADF bad file handle errors that can be logged under heavy
stress on writes). Modify cifs Kconfig options to expose CONFIG_CIFS_STATS2
Fix SFU style symlinks and mknod needed for servers which do not support the
CIFS Unix Extensions. Fix setfacl/getfacl on bigendian. Timeout negative
dentries so files that the client sees as deleted but that later get created
on the server will be recognized. Add client side permission check on setattr.
Timeout stuck requests better (where server has never responded or sent corrupt
responses)
Version 1.38
------------
Fix tcp socket retransmission timeouts (e.g. on ENOSPACE from the socket)
to be smaller at first (but increasing) so large write performance performance
over GigE is better. Do not hang thread on illegal byte range lock response
from Windows (Windows can send an RFC1001 size which does not match smb size) by
allowing an SMBs TCP length to be up to a few bytes longer than it should be.
wsize and rsize can now be larger than negotiated buffer size if server
supports large readx/writex, even when directio mount flag not specified.
Write size will in many cases now be 16K instead of 4K which greatly helps
file copy performance on lightly loaded networks. Fix oops in dnotify
when experimental config flag enabled. Make cifsFYI more granular.
Version 1.37
------------
Fix readdir caching when unlink removes file in current search buffer,
and this is followed by a rewind search to just before the deleted entry.
Do not attempt to set ctime unless atime and/or mtime change requested
(most servers throw it away anyway). Fix length check of received smbs
to be more accurate. Fix big endian problem with mapchars mount option,
and with a field returned by statfs.
Version 1.36
------------
Add support for mounting to older pre-CIFS servers such as Windows9x and ME.
For these older servers, add option for passing netbios name of server in
on mount (servernetbiosname). Add suspend support for power management, to
avoid cifsd thread preventing software suspend from working.
Add mount option for disabling the default behavior of sending byte range lock
requests to the server (necessary for certain applications which break with
mandatory lock behavior such as Evolution), and also mount option for
requesting case insensitive matching for path based requests (requesting
case sensitive is the default).
Version 1.35
------------
Add writepage performance improvements. Fix path name conversions
for long filenames on mounts which were done with "mapchars" mount option
specified. Ensure multiplex ids do not collide. Fix case in which
rmmod can oops if done soon after last unmount. Fix truncated
search (readdir) output when resume filename was a long filename.
Fix filename conversion when mapchars mount option was specified and
filename was a long filename.
Version 1.34
------------
Fix error mapping of the TOO_MANY_LINKS (hardlinks) case.
Do not oops if root user kills cifs oplock kernel thread or
kills the cifsd thread (NB: killing the cifs kernel threads is not
recommended, unmount and rmmod cifs will kill them when they are
no longer needed). Fix readdir to ASCII servers (ie older servers
which do not support Unicode) and also require asterisk.
Fix out of memory case in which data could be written one page
off in the page cache.
Version 1.33
------------
Fix caching problem, in which readdir of directory containing a file
which was cached could cause the file's time stamp to be updated
without invalidating the readahead data (so we could get stale
file data on the client for that file even as the server copy changed).
Cleanup response processing so cifsd can not loop when abnormally
terminated.
Version 1.32
------------
Fix oops in ls when Transact2 FindFirst (or FindNext) returns more than one
transact response for an SMB request and search entry split across two frames.
Add support for lsattr (getting ext2/ext3/reiserfs attr flags from the server)
as new protocol extensions. Do not send Get/Set calls for POSIX ACLs
unless server explicitly claims to support them in CIFS Unix extensions
POSIX ACL capability bit. Fix packet signing when multiuser mounting with
different users from the same client to the same server. Fix oops in
cifs_close. Add mount option for remapping reserved characters in
filenames (also allow recognizing files with created by SFU which have any
of these seven reserved characters, except backslash, to be recognized).
Fix invalid transact2 message (we were sometimes trying to interpret
oplock breaks as SMB responses). Add ioctl for checking that the
current uid matches the uid of the mounter (needed by umount.cifs).
Reduce the number of large buffer allocations in cifs response processing
(significantly reduces memory pressure under heavy stress with multiple
processes accessing the same server at the same time).
Version 1.31
------------
Fix updates of DOS attributes and time fields so that files on NT4 servers
do not get marked delete on close. Display sizes of cifs buffer pools in
cifs stats. Fix oops in unmount when cifsd thread being killed by
shutdown. Add generic readv/writev and aio support. Report inode numbers
consistently in readdir and lookup (when serverino mount option is
specified use the inode number that the server reports - for both lookup
and readdir, otherwise by default the locally generated inode number is used
for inodes created in either path since servers are not always able to
provide unique inode numbers when exporting multiple volumes from under one
sharename).
Version 1.30
------------
Allow new nouser_xattr mount parm to disable xattr support for user namespace.
Do not flag user_xattr mount parm in dmesg. Retry failures setting file time
(mostly affects NT4 servers) by retry with handle based network operation.
Add new POSIX Query FS Info for returning statfs info more accurately.
Handle passwords with multiple commas in them.
Version 1.29
------------
Fix default mode in sysfs of cifs module parms. Remove old readdir routine.
Fix capabilities flags for large readx so as to allow reads larger than 64K.
Version 1.28
------------
Add module init parm for large SMB buffer size (to allow it to be changed
from its default of 16K) which is especially useful for large file copy
when mounting with the directio mount option. Fix oops after
returning from mount when experimental ExtendedSecurity enabled and
SpnegoNegotiated returning invalid error. Fix case to retry better when
peek returns from 1 to 3 bytes on socket which should have more data.
Fixed path based calls (such as cifs lookup) to handle path names
longer than 530 (now can handle PATH_MAX). Fix pass through authentication
from Samba server to DC (Samba required dummy LM password).
Version 1.27
------------
Turn off DNOTIFY (directory change notification support) by default
(unless built with the experimental flag) to fix hang with KDE
file browser. Fix DNOTIFY flag mappings. Fix hang (in wait_event
waiting on an SMB response) in SendReceive when session dies but
reconnects quickly from another task. Add module init parms for
minimum number of large and small network buffers in the buffer pools,
and for the maximum number of simultaneous requests.
Version 1.26
------------
Add setfacl support to allow setting of ACLs remotely to Samba 3.10 and later
and other POSIX CIFS compliant servers. Fix error mapping for getfacl
to EOPNOTSUPP when server does not support posix acls on the wire. Fix
improperly zeroed buffer in CIFS Unix extensions set times call.
Version 1.25
------------
Fix internationalization problem in cifs readdir with filenames that map to
longer UTF-8 strings than the string on the wire was in Unicode. Add workaround
for readdir to netapp servers. Fix search rewind (seek into readdir to return
non-consecutive entries). Do not do readdir when server negotiates
buffer size to small to fit filename. Add support for reading POSIX ACLs from
the server (add also acl and noacl mount options).
Version 1.24
------------
Optionally allow using server side inode numbers, rather than client generated
ones by specifying mount option "serverino" - this is required for some apps
to work which double check hardlinked files and have persistent inode numbers.
Version 1.23
------------
Multiple bigendian fixes. On little endian systems (for reconnect after
network failure) fix tcp session reconnect code so we do not try first
to reconnect on reverse of port 445. Treat reparse points (NTFS junctions)
as directories rather than symlinks because we can do follow link on them.
Version 1.22
------------
Add config option to enable XATTR (extended attribute) support, mapping
xattr names in the "user." namespace space to SMB/CIFS EAs. Lots of
minor fixes pointed out by the Stanford SWAT checker (mostly missing
or out of order NULL pointer checks in little used error paths).
Version 1.21
------------
Add new mount parm to control whether mode check (generic_permission) is done
on the client. If Unix extensions are enabled and the uids on the client
and server do not match, client permission checks are meaningless on
server uids that do not exist on the client (this does not affect the
normal ACL check which occurs on the server). Fix default uid
on mknod to match create and mkdir. Add optional mount parm to allow
override of the default uid behavior (in which the server sets the uid
and gid of newly created files). Normally for network filesystem mounts
user want the server to set the uid/gid on newly created files (rather than
using uid of the client processes you would in a local filesystem).
Version 1.20
------------
Make transaction counts more consistent. Merge /proc/fs/cifs/SimultaneousOps
info into /proc/fs/cifs/DebugData. Fix oops in rare oops in readdir
(in build_wildcard_path_from_dentry). Fix mknod to pass type field
(block/char/fifo) properly. Remove spurious mount warning log entry when
credentials passed as mount argument. Set major/minor device number in
inode for block and char devices when unix extensions enabled.
Version 1.19
------------
Fix /proc/fs/cifs/Stats and DebugData display to handle larger
amounts of return data. Properly limit requests to MAX_REQ (50
is the usual maximum active multiplex SMB/CIFS requests per server).
Do not kill cifsd (and thus hurt the other SMB session) when more than one
session to the same server (but with different userids) exists and one
of the two user's smb sessions is being removed while leaving the other.
Do not loop reconnecting in cifsd demultiplex thread when admin
kills the thread without going through unmount.
Version 1.18
------------
Do not rename hardlinked files (since that should be a noop). Flush
cached write behind data when reopening a file after session abend,
except when already in write. Grab per socket sem during reconnect
to avoid oops in sendmsg if overlapping with reconnect. Do not
reset cached inode file size on readdir for files open for write on
client.
Version 1.17
------------
Update number of blocks in file so du command is happier (in Linux a fake
blocksize of 512 is required for calculating number of blocks in inode).
Fix prepare write of partial pages to read in data from server if possible.
Fix race on tcpStatus field between unmount and reconnection code, causing
cifsd process sometimes to hang around forever. Improve out of memory
checks in cifs_filldir
Version 1.16
------------
Fix incorrect file size in file handle based setattr on big endian hardware.
Fix oops in build_path_from_dentry when out of memory. Add checks for invalid
and closing file structs in writepage/partialpagewrite. Add statistics
for each mounted share (new menuconfig option). Fix endianness problem in
volume information displayed in /proc/fs/cifs/DebugData (only affects
affects big endian architectures). Prevent renames while constructing
path names for open, mkdir and rmdir.
Version 1.15
------------
Change to mempools for alloc smb request buffers and multiplex structs
to better handle low memory problems (and potential deadlocks).
Version 1.14
------------
Fix incomplete listings of large directories on Samba servers when Unix
extensions enabled. Fix oops when smb_buffer can not be allocated. Fix
rename deadlock when writing out dirty pages at same time.
Version 1.13
------------
Fix open of files in which O_CREATE can cause the mode to change in
some cases. Fix case in which retry of write overlaps file close.
Fix PPC64 build error. Reduce excessive stack usage in smb password
hashing. Fix overwrite of Linux user's view of file mode to Windows servers.
Version 1.12
------------
Fixes for large file copy, signal handling, socket retry, buffer
allocation and low memory situations.
Version 1.11
------------
Better port 139 support to Windows servers (RFC1001/RFC1002 Session_Initialize)
also now allowing support for specifying client netbiosname. NT4 support added.
Version 1.10
------------
Fix reconnection (and certain failed mounts) to properly wake up the
blocked users thread so it does not seem hung (in some cases was blocked
until the cifs receive timeout expired). Fix spurious error logging
to kernel log when application with open network files killed.
Version 1.09
------------
Fix /proc/fs module unload warning message (that could be logged
to the kernel log). Fix intermittent failure in connectathon
test7 (hardlink count not immediately refreshed in case in which
inode metadata can be incorrectly kept cached when time near zero)
Version 1.08
------------
Allow file_mode and dir_mode (specified at mount time) to be enforced
locally (the server already enforced its own ACLs too) for servers
that do not report the correct mode (do not support the
CIFS Unix Extensions).
Version 1.07
------------
Fix some small memory leaks in some unmount error paths. Fix major leak
of cache pages in readpages causing multiple read oriented stress
testcases (including fsx, and even large file copy) to fail over time.
Version 1.06
------------
Send NTCreateX with ATTR_POSIX if Linux/Unix extensions negotiated with server.
This allows files that differ only in case and improves performance of file
creation and file open to such servers. Fix semaphore conflict which causes
slow delete of open file to Samba (which unfortunately can cause an oplock
break to self while vfs_unlink held i_sem) which can hang for 20 seconds.
Version 1.05
------------
fixes to cifs_readpages for fsx test case
Version 1.04
------------
Fix caching data integrity bug when extending file size especially when no
oplock on file. Fix spurious logging of valid already parsed mount options
that are parsed outside of the cifs vfs such as nosuid.
Version 1.03
------------
Connect to server when port number override not specified, and tcp port
unitialized. Reset search to restart at correct file when kernel routine
filldir returns error during large directory searches (readdir).
Version 1.02
------------
Fix caching problem when files opened by multiple clients in which
page cache could contain stale data, and write through did
not occur often enough while file was still open when read ahead
(read oplock) not allowed. Treat "sep=" when first mount option
as an override of comma as the default separator between mount
options.
Version 1.01
------------
Allow passwords longer than 16 bytes. Allow null password string.
Version 1.00
------------
Gracefully clean up failed mounts when attempting to mount to servers such as
Windows 98 that terminate tcp sessions during protocol negotiation. Handle
embedded commas in mount parsing of passwords.
Version 0.99
------------
Invalidate local inode cached pages on oplock break and when last file
instance is closed so that the client does not continue using stale local
copy rather than later modified server copy of file. Do not reconnect
when server drops the tcp session prematurely before negotiate
protocol response. Fix oops in reopen_file when dentry freed. Allow
the support for CIFS Unix Extensions to be disabled via proc interface.
Version 0.98
------------
Fix hang in commit_write during reconnection of open files under heavy load.
Fix unload_nls oops in a mount failure path. Serialize writes to same socket
which also fixes any possible races when cifs signatures are enabled in SMBs
being sent out of signature sequence number order.
Version 0.97
------------
Fix byte range locking bug (endian problem) causing bad offset and
length.
Version 0.96
------------
Fix oops (in send_sig) caused by CIFS unmount code trying to
wake up the demultiplex thread after it had exited. Do not log
error on harmless oplock release of closed handle.
Version 0.95
------------
Fix unsafe global variable usage and password hash failure on gcc 3.3.1
Fix problem reconnecting secondary mounts to same server after session
failure. Fix invalid dentry - race in mkdir when directory gets created
by another client between the lookup and mkdir.
Version 0.94
------------
Fix to list processing in reopen_files. Fix reconnection when server hung
but tcpip session still alive. Set proper timeout on socket read.
Version 0.93
------------
Add missing mount options including iocharset. SMP fixes in write and open.
Fix errors in reconnecting after TCP session failure. Fix module unloading
of default nls codepage
Version 0.92
------------
Active smb transactions should never go negative (fix double FreeXid). Fix
list processing in file routines. Check return code on kmalloc in open.
Fix spinlock usage for SMP.
Version 0.91
------------
Fix oops in reopen_files when invalid dentry. drop dentry on server rename
and on revalidate errors. Fix cases where pid is now tgid. Fix return code
on create hard link when server does not support them.
Version 0.90
------------
Fix scheduling while atomic error in getting inode info on newly created file.
Fix truncate of existing files opened with O_CREAT but not O_TRUNC set.
Version 0.89
------------
Fix oops on write to dead tcp session. Remove error log write for case when file open
O_CREAT but not O_EXCL
Version 0.88
------------
Fix non-POSIX behavior on rename of open file and delete of open file by taking
advantage of trans2 SetFileInfo rename facility if available on target server.
Retry on ENOSPC and EAGAIN socket errors.
Version 0.87
------------
Fix oops on big endian readdir. Set blksize to be even power of two (2**blkbits) to fix
allocation size miscalculation. After oplock token lost do not read through
cache.
Version 0.86
------------
Fix oops on empty file readahead. Fix for file size handling for locally cached files.
Version 0.85
------------
Fix oops in mkdir when server fails to return inode info. Fix oops in reopen_files
during auto reconnection to server after server recovered from failure.
Version 0.84
------------
Finish support for Linux 2.5 open/create changes, which removes the
redundant NTCreate/QPathInfo/close that was sent during file create.
Enable oplock by default. Enable packet signing by default (needed to
access many recent Windows servers)
Version 0.83
------------
Fix oops when mounting to long server names caused by inverted parms to kmalloc.
Fix MultiuserMount (/proc/fs/cifs configuration setting) so that when enabled
we will choose a cifs user session (smb uid) that better matches the local
uid if a) the mount uid does not match the current uid and b) we have another
session to the same server (ip address) for a different mount which
matches the current local uid.
Version 0.82
------------
Add support for mknod of block or character devices. Fix oplock
code (distributed caching) to properly send response to oplock
break from server.
Version 0.81
------------
Finish up CIFS packet digital signing for the default
NTLM security case. This should help Windows 2003
network interoperability since it is common for
packet signing to be required now. Fix statfs (stat -f)
which recently started returning errors due to
invalid value (-1 instead of 0) being set in the
struct kstatfs f_ffiles field.
Version 0.80
-----------
Fix oops on stopping oplock thread when removing cifs when
built as module.
Version 0.79
------------
Fix mount options for ro (readonly), uid, gid and file and directory mode.
Version 0.78
------------
Fix errors displayed on failed mounts to be more understandable.
Fixed various incorrect or misleading smb to posix error code mappings.
Version 0.77
------------
Fix display of NTFS DFS junctions to display as symlinks.
They are the network equivalent. Fix oops in
cifs_partialpagewrite caused by missing spinlock protection
of openfile linked list. Allow writebehind caching errors to
be returned to the application at file close.
Version 0.76
------------
Clean up options displayed in /proc/mounts by show_options to
be more consistent with other filesystems.
Version 0.75
------------
Fix delete of readonly file to Windows servers. Reflect
presence or absence of read only dos attribute in mode
bits for servers that do not support CIFS Unix extensions.
Fix shortened results on readdir of large directories to
servers supporting CIFS Unix extensions (caused by
incorrect resume key).
Version 0.74
------------
Fix truncate bug (set file size) that could cause hangs e.g. running fsx
Version 0.73
------------
unload nls if mount fails.
Version 0.72
------------
Add resume key support to search (readdir) code to workaround
Windows bug. Add /proc/fs/cifs/LookupCacheEnable which
allows disabling caching of attribute information for
lookups.
Version 0.71
------------
Add more oplock handling (distributed caching code). Remove
dead code. Remove excessive stack space utilization from
symlink routines.
Version 0.70
------------
Fix oops in get dfs referral (triggered when null path sent in to
mount). Add support for overriding rsize at mount time.
Version 0.69
------------
Fix buffer overrun in readdir which caused intermittent kernel oopses.
Fix writepage code to release kmap on write data. Allow "-ip=" new
mount option to be passed in on parameter distinct from the first part
(server name portion of) the UNC name. Allow override of the
tcp port of the target server via new mount option "-port="
Version 0.68
------------
Fix search handle leak on rewind. Fix setuid and gid so that they are
reflected in the local inode immediately. Cleanup of whitespace
to make 2.4 and 2.5 versions more consistent.
Version 0.67
------------
Fix signal sending so that captive thread (cifsd) exits on umount
(which was causing the warning in kmem_cache_free of the request buffers
at rmmod time). This had broken as a sideeffect of the recent global
kernel change to daemonize. Fix memory leak in readdir code which
showed up in "ls -R" (and applications that did search rewinding).
Version 0.66
------------
Reconnect tids and fids after session reconnection (still do not
reconnect byte range locks though). Fix problem caching
lookup information for directory inodes, improving performance,
especially in deep directory trees. Fix various build warnings.
Version 0.65
------------
Finish fixes to commit write for caching/readahead consistency. fsx
now works to Samba servers. Fix oops caused when readahead
was interrupted by a signal.
Version 0.64
------------
Fix data corruption (in partial page after truncate) that caused fsx to
fail to Windows servers. Cleaned up some extraneous error logging in
common error paths. Add generic sendfile support.
Version 0.63
------------
Fix memory leak in AllocMidQEntry.
Finish reconnection logic, so connection with server can be dropped
(or server rebooted) and the cifs client will reconnect.
Version 0.62
------------
Fix temporary socket leak when bad userid or password specified
(or other SMBSessSetup failure). Increase maximum buffer size to slightly
over 16K to allow negotiation of up to Samba and Windows server default read
sizes. Add support for readpages
Version 0.61
------------
Fix oops when username not passed in on mount. Extensive fixes and improvements
to error logging (strip redundant newlines, change debug macros to ensure newline
passed in and to be more consistent). Fix writepage wrong file handle problem,
a readonly file handle could be incorrectly used to attempt to write out
file updates through the page cache to multiply open files. This could cause
the iozone benchmark to fail on the fwrite test. Fix bug mounting two different
shares to the same Windows server when using different usernames
(doing this to Samba servers worked but Windows was rejecting it) - now it is
possible to use different userids when connecting to the same server from a
Linux client. Fix oops when treeDisconnect called during unmount on
previously freed socket.
Version 0.60
------------
Fix oops in readpages caused by not setting address space operations in inode in
rare code path.
Version 0.59
------------
Includes support for deleting of open files and renaming over existing files (per POSIX
requirement). Add readlink support for Windows junction points (directory symlinks).
Version 0.58
------------
Changed read and write to go through pagecache. Added additional address space operations.
Memory mapped operations now working.
Version 0.57
------------
Added writepage code for additional memory mapping support. Fixed leak in xids causing
the simultaneous operations counter (/proc/fs/cifs/SimultaneousOps) to increase on
every stat call. Additional formatting cleanup.
Version 0.56
------------
Fix bigendian bug in order of time conversion. Merge 2.5 to 2.4 version. Formatting cleanup.
Version 0.55
------------
Fixes from Zwane Mwaikambo for adding missing return code checking in a few places.
Also included a modified version of his fix to protect global list manipulation of
the smb session and tree connection and mid related global variables.
Version 0.54
------------
Fix problem with captive thread hanging around at unmount time. Adjust to 2.5.42-pre
changes to superblock layout. Remove wasteful allocation of smb buffers (now the send
buffer is reused for responses). Add more oplock handling. Additional minor cleanup.
Version 0.53
------------
More stylistic updates to better match kernel style. Add additional statistics
for filesystem which can be viewed via /proc/fs/cifs. Add more pieces of NTLMv2
and CIFS Packet Signing enablement.
Version 0.52
------------
Replace call to sleep_on with safer wait_on_event.
Make stylistic changes to better match kernel style recommendations.
Remove most typedef usage (except for the PDUs themselves).
Version 0.51
------------
Update mount so the -unc mount option is no longer required (the ip address can be specified
in a UNC style device name. Implementation of readpage/writepage started.
Version 0.50
------------
Fix intermittent problem with incorrect smb header checking on badly
fragmented tcp responses
Version 0.49
------------
Fixes to setting of allocation size and file size.
Version 0.48
------------
Various 2.5.38 fixes. Now works on 2.5.38
Version 0.47
------------
Prepare for 2.5 kernel merge. Remove ifdefs.
Version 0.46
------------
Socket buffer management fixes. Fix dual free.
Version 0.45
------------
Various big endian fixes for hardlinks and symlinks and also for dfs.
Version 0.44
------------
Various big endian fixes for servers with Unix extensions such as Samba
Version 0.43
------------
Various FindNext fixes for incorrect filenames on large directory searches on big endian
clients. basic posix file i/o tests now work on big endian machines, not just le
Version 0.42
------------
SessionSetup and NegotiateProtocol now work from Big Endian machines.
Various Big Endian fixes found during testing on the Linux on 390. Various fixes for compatibility with older
versions of 2.4 kernel (now builds and works again on kernels at least as early as 2.4.7).
Version 0.41
------------
Various minor fixes for Connectathon Posix "basic" file i/o test suite. Directory caching fixed so hardlinked
files now return the correct number of links on fstat as they are repeatedly linked and unlinked.
Version 0.40
------------
Implemented "Raw" (i.e. not encapsulated in SPNEGO) NTLMSSP (i.e. the Security Provider Interface used to negotiate
session advanced session authentication). Raw NTLMSSP is preferred by Windows 2000 Professional and Windows XP.
Began implementing support for SPNEGO encapsulation of NTLMSSP based session authentication blobs
(which is the mechanism preferred by Windows 2000 server in the absence of Kerberos).
Version 0.38
------------
Introduced optional mount helper utility mount.cifs and made coreq changes to cifs vfs to enable
it. Fixed a few bugs in the DFS code (e.g. bcc two bytes too short and incorrect uid in PDU).
Version 0.37
------------
Rewrote much of connection and mount/unmount logic to handle bugs with
multiple uses to same share, multiple users to same server etc.
Version 0.36
------------
Fixed major problem with dentry corruption (missing call to dput)
Version 0.35
------------
Rewrite of readdir code to fix bug. Various fixes for bigendian machines.
Begin adding oplock support. Multiusermount and oplockEnabled flags added to /proc/fs/cifs
although corresponding function not fully implemented in the vfs yet
Version 0.34
------------
Fixed dentry caching bug, misc. cleanup
Version 0.33
------------
Fixed 2.5 support to handle build and configure changes as well as misc. 2.5 changes. Now can build
on current 2.5 beta version (2.5.24) of the Linux kernel as well as on 2.4 Linux kernels.
Support for STATUS codes (newer 32 bit NT error codes) added. DFS support begun to be added.
Version 0.32
------------
Unix extensions (symlink, readlink, hardlink, chmod and some chgrp and chown) implemented
and tested against Samba 2.2.5
Version 0.31
------------
1) Fixed lockrange to be correct (it was one byte too short)
2) Fixed GETLK (i.e. the fcntl call to test a range of bytes in a file to see if locked) to correctly
show range as locked when there is a conflict with an existing lock.
3) default file perms are now 2767 (indicating support for mandatory locks) instead of 777 for directories
in most cases. Eventually will offer optional ability to query server for the correct perms.
3) Fixed eventual trap when mounting twice to different shares on the same server when the first succeeded
but the second one was invalid and failed (the second one was incorrectly disconnecting the tcp and smb
session)
4) Fixed error logging of valid mount options
5) Removed logging of password field.
6) Moved negotiate, treeDisconnect and uloggoffX (only tConx and SessSetup remain in connect.c) to cifssmb.c
and cleaned them up and made them more consistent with other cifs functions.
7) Server support for Unix extensions is now fully detected and FindFirst is implemented both ways
(with or without Unix extensions) but FindNext and QueryPathInfo with the Unix extensions are not completed,
nor is the symlink support using the Unix extensions
8) Started adding the readlink and follow_link code
Version 0.3
-----------
Initial drop
etercifs/README 0 → 100644
View file @ 3c073451
The CIFS VFS support for Linux supports many advanced network filesystem
features such as hierarchical dfs like namespace, hardlinks, locking and more.
It was designed to comply with the SNIA CIFS Technical Reference (which
supersedes the 1992 X/Open SMB Standard) as well as to perform best practice
practical interoperability with Windows 2000, Windows XP, Samba and equivalent
servers. This code was developed in participation with the Protocol Freedom
Information Foundation.
Please see
http://protocolfreedom.org/ and
http://samba.org/samba/PFIF/
for more details.
For questions or bug reports please contact:
sfrench@samba.org (sfrench@us.ibm.com)
Build instructions:
==================
For Linux 2.4:
1) Get the kernel source (e.g.from http://www.kernel.org)
and download the cifs vfs source (see the project page
at http://us1.samba.org/samba/Linux_CIFS_client.html)
and change directory into the top of the kernel directory
then patch the kernel (e.g. "patch -p1 < cifs_24.patch")
to add the cifs vfs to your kernel configure options if
it has not already been added (e.g. current SuSE and UL
users do not need to apply the cifs_24.patch since the cifs vfs is
already in the kernel configure menu) and then
mkdir linux/fs/cifs and then copy the current cifs vfs files from
the cifs download to your kernel build directory e.g.
cp <cifs_download_dir>/fs/cifs/* to <kernel_download_dir>/fs/cifs
2) make menuconfig (or make xconfig)
3) select cifs from within the network filesystem choices
4) save and exit
5) make dep
6) make modules (or "make" if CIFS VFS not to be built as a module)
For Linux 2.6:
1) Download the kernel (e.g. from http://www.kernel.org)
and change directory into the top of the kernel directory tree
(e.g. /usr/src/linux-2.5.73)
2) make menuconfig (or make xconfig)
3) select cifs from within the network filesystem choices
4) save and exit
5) make
Installation instructions:
=========================
If you have built the CIFS vfs as module (successfully) simply
type "make modules_install" (or if you prefer, manually copy the file to
the modules directory e.g. /lib/modules/2.4.10-4GB/kernel/fs/cifs/cifs.o).
If you have built the CIFS vfs into the kernel itself, follow the instructions
for your distribution on how to install a new kernel (usually you
would simply type "make install").
If you do not have the utility mount.cifs (in the Samba 3.0 source tree and on
the CIFS VFS web site) copy it to the same directory in which mount.smbfs and
similar files reside (usually /sbin). Although the helper software is not
required, mount.cifs is recommended. Eventually the Samba 3.0 utility program
"net" may also be helpful since it may someday provide easier mount syntax for
users who are used to Windows e.g.
net use <mount point> <UNC name or cifs URL>
Note that running the Winbind pam/nss module (logon service) on all of your
Linux clients is useful in mapping Uids and Gids consistently across the
domain to the proper network user. The mount.cifs mount helper can be
trivially built from Samba 3.0 or later source e.g. by executing:
gcc samba/source/client/mount.cifs.c -o mount.cifs
If cifs is built as a module, then the size and number of network buffers
and maximum number of simultaneous requests to one server can be configured.
Changing these from their defaults is not recommended. By executing modinfo
modinfo kernel/fs/cifs/cifs.ko
on kernel/fs/cifs/cifs.ko the list of configuration changes that can be made
at module initialization time (by running insmod cifs.ko) can be seen.
Allowing User Mounts
====================
To permit users to mount and unmount over directories they own is possible
with the cifs vfs. A way to enable such mounting is to mark the mount.cifs
utility as suid (e.g. "chmod +s /sbin/mount.cifs). To enable users to
umount shares they mount requires
1) mount.cifs version 1.4 or later
2) an entry for the share in /etc/fstab indicating that a user may
unmount it e.g.
//server/usersharename /mnt/username cifs user 0 0
Note that when the mount.cifs utility is run suid (allowing user mounts),
in order to reduce risks, the "nosuid" mount flag is passed in on mount to
disallow execution of an suid program mounted on the remote target.
When mount is executed as root, nosuid is not passed in by default,
and execution of suid programs on the remote target would be enabled
by default. This can be changed, as with nfs and other filesystems,
by simply specifying "nosuid" among the mount options. For user mounts
though to be able to pass the suid flag to mount requires rebuilding
mount.cifs with the following flag:
gcc samba/source/client/mount.cifs.c -DCIFS_ALLOW_USR_SUID -o mount.cifs
There is a corresponding manual page for cifs mounting in the Samba 3.0 and
later source tree in docs/manpages/mount.cifs.8
Allowing User Unmounts
======================
To permit users to ummount directories that they have user mounted (see above),
the utility umount.cifs may be used. It may be invoked directly, or if
umount.cifs is placed in /sbin, umount can invoke the cifs umount helper
(at least for most versions of the umount utility) for umount of cifs
mounts, unless umount is invoked with -i (which will avoid invoking a umount
helper). As with mount.cifs, to enable user unmounts umount.cifs must be marked
as suid (e.g. "chmod +s /sbin/umount.cifs") or equivalent (some distributions
allow adding entries to a file to the /etc/permissions file to achieve the
equivalent suid effect). For this utility to succeed the target path
must be a cifs mount, and the uid of the current user must match the uid
of the user who mounted the resource.
Also note that the customary way of allowing user mounts and unmounts is
(instead of using mount.cifs and unmount.cifs as suid) to add a line
to the file /etc/fstab for each //server/share you wish to mount, but
this can become unwieldy when potential mount targets include many
or unpredictable UNC names.
Samba Considerations
====================
To get the maximum benefit from the CIFS VFS, we recommend using a server that
supports the SNIA CIFS Unix Extensions standard (e.g. Samba 2.2.5 or later or
Samba 3.0) but the CIFS vfs works fine with a wide variety of CIFS servers.
Note that uid, gid and file permissions will display default values if you do
not have a server that supports the Unix extensions for CIFS (such as Samba
2.2.5 or later). To enable the Unix CIFS Extensions in the Samba server, add
the line:
unix extensions = yes
to your smb.conf file on the server. Note that the following smb.conf settings
are also useful (on the Samba server) when the majority of clients are Unix or
Linux:
case sensitive = yes
delete readonly = yes
ea support = yes
Note that server ea support is required for supporting xattrs from the Linux
cifs client, and that EA support is present in later versions of Samba (e.g.
3.0.6 and later (also EA support works in all versions of Windows, at least to
shares on NTFS filesystems). Extended Attribute (xattr) support is an optional
feature of most Linux filesystems which may require enabling via
make menuconfig. Client support for extended attributes (user xattr) can be
disabled on a per-mount basis by specifying "nouser_xattr" on mount.
The CIFS client can get and set POSIX ACLs (getfacl, setfacl) to Samba servers
version 3.10 and later. Setting POSIX ACLs requires enabling both XATTR and
then POSIX support in the CIFS configuration options when building the cifs
module. POSIX ACL support can be disabled on a per mount basic by specifying
"noacl" on mount.
Some administrators may want to change Samba's smb.conf "map archive" and
"create mask" parameters from the default. Unless the create mask is changed
newly created files can end up with an unnecessarily restrictive default mode,
which may not be what you want, although if the CIFS Unix extensions are
enabled on the server and client, subsequent setattr calls (e.g. chmod) can
fix the mode. Note that creating special devices (mknod) remotely
may require specifying a mkdev function to Samba if you are not using
Samba 3.0.6 or later. For more information on these see the manual pages
("man smb.conf") on the Samba server system. Note that the cifs vfs,
unlike the smbfs vfs, does not read the smb.conf on the client system
(the few optional settings are passed in on mount via -o parameters instead).
Note that Samba 2.2.7 or later includes a fix that allows the CIFS VFS to delete
open files (required for strict POSIX compliance). Windows Servers already
supported this feature. Samba server does not allow symlinks that refer to files
outside of the share, so in Samba versions prior to 3.0.6, most symlinks to
files with absolute paths (ie beginning with slash) such as:
ln -s /mnt/foo bar
would be forbidden. Samba 3.0.6 server or later includes the ability to create
such symlinks safely by converting unsafe symlinks (ie symlinks to server
files that are outside of the share) to a samba specific format on the server
that is ignored by local server applications and non-cifs clients and that will
not be traversed by the Samba server). This is opaque to the Linux client
application using the cifs vfs. Absolute symlinks will work to Samba 3.0.5 or
later, but only for remote clients using the CIFS Unix extensions, and will
be invisbile to Windows clients and typically will not affect local
applications running on the same server as Samba.
Use instructions:
================
Once the CIFS VFS support is built into the kernel or installed as a module
(cifs.o), you can use mount syntax like the following to access Samba or Windows
servers:
mount -t cifs //9.53.216.11/e$ /mnt -o user=myname,pass=mypassword
Before -o the option -v may be specified to make the mount.cifs
mount helper display the mount steps more verbosely.
After -o the following commonly used cifs vfs specific options
are supported:
user=<username>
pass=<password>
domain=<domain name>
Other cifs mount options are described below. Use of TCP names (in addition to
ip addresses) is available if the mount helper (mount.cifs) is installed. If
you do not trust the server to which are mounted, or if you do not have
cifs signing enabled (and the physical network is insecure), consider use
of the standard mount options "noexec" and "nosuid" to reduce the risk of
running an altered binary on your local system (downloaded from a hostile server
or altered by a hostile router).
Although mounting using format corresponding to the CIFS URL specification is
not possible in mount.cifs yet, it is possible to use an alternate format
for the server and sharename (which is somewhat similar to NFS style mount
syntax) instead of the more widely used UNC format (i.e. \\server\share):
mount -t cifs tcp_name_of_server:share_name /mnt -o user=myname,pass=mypasswd
When using the mount helper mount.cifs, passwords may be specified via alternate
mechanisms, instead of specifying it after -o using the normal "pass=" syntax
on the command line:
1) By including it in a credential file. Specify credentials=filename as one
of the mount options. Credential files contain two lines
username=someuser
password=your_password
2) By specifying the password in the PASSWD environment variable (similarly
the user name can be taken from the USER environment variable).
3) By specifying the password in a file by name via PASSWD_FILE
4) By specifying the password in a file by file descriptor via PASSWD_FD
If no password is provided, mount.cifs will prompt for password entry
Restrictions
============
Servers must support either "pure-TCP" (port 445 TCP/IP CIFS connections) or RFC
1001/1002 support for "Netbios-Over-TCP/IP." This is not likely to be a
problem as most servers support this.
Valid filenames differ between Windows and Linux. Windows typically restricts
filenames which contain certain reserved characters (e.g.the character :
which is used to delimit the beginning of a stream name by Windows), while
Linux allows a slightly wider set of valid characters in filenames. Windows
servers can remap such characters when an explicit mapping is specified in
the Server's registry. Samba starting with version 3.10 will allow such
filenames (ie those which contain valid Linux characters, which normally
would be forbidden for Windows/CIFS semantics) as long as the server is
configured for Unix Extensions (and the client has not disabled
/proc/fs/cifs/LinuxExtensionsEnabled).
CIFS VFS Mount Options
======================
A partial list of the supported mount options follows:
user The user name to use when trying to establish
the CIFS session.
password The user password. If the mount helper is
installed, the user will be prompted for password
if not supplied.
ip The ip address of the target server
unc The target server Universal Network Name (export) to
mount.
domain Set the SMB/CIFS workgroup name prepended to the
username during CIFS session establishment
uid Set the default uid for inodes. For mounts to servers
which do support the CIFS Unix extensions, such as a
properly configured Samba server, the server provides
the uid, gid and mode so this parameter should not be
specified unless the server and clients uid and gid
numbering differ. If the server and client are in the
same domain (e.g. running winbind or nss_ldap) and
the server supports the Unix Extensions then the uid
and gid can be retrieved from the server (and uid
and gid would not have to be specifed on the mount.
For servers which do not support the CIFS Unix
extensions, the default uid (and gid) returned on lookup
of existing files will be the uid (gid) of the person
who executed the mount (root, except when mount.cifs
is configured setuid for user mounts) unless the "uid="
(gid) mount option is specified. For the uid (gid) of newly
created files and directories, ie files created since
the last mount of the server share, the expected uid
(gid) is cached as long as the inode remains in
memory on the client. Also note that permission
checks (authorization checks) on accesses to a file occur
at the server, but there are cases in which an administrator
may want to restrict at the client as well. For those
servers which do not report a uid/gid owner
(such as Windows), permissions can also be checked at the
client, and a crude form of client side permission checking
can be enabled by specifying file_mode and dir_mode on
the client. Note that the mount.cifs helper must be
at version 1.10 or higher to support specifying the uid
(or gid) in non-numeric form.
gid Set the default gid for inodes (similar to above).
file_mode If CIFS Unix extensions are not supported by the server
this overrides the default mode for file inodes.
dir_mode If CIFS Unix extensions are not supported by the server
this overrides the default mode for directory inodes.
port attempt to contact the server on this tcp port, before
trying the usual ports (port 445, then 139).
iocharset Codepage used to convert local path names to and from
Unicode. Unicode is used by default for network path
names if the server supports it. If iocharset is
not specified then the nls_default specified
during the local client kernel build will be used.
If server does not support Unicode, this parameter is
unused.
rsize default read size (usually 16K). The client currently
can not use rsize larger than CIFSMaxBufSize. CIFSMaxBufSize
defaults to 16K and may be changed (from 8K to the maximum
kmalloc size allowed by your kernel) at module install time
for cifs.ko. Setting CIFSMaxBufSize to a very large value
will cause cifs to use more memory and may reduce performance
in some cases. To use rsize greater than 127K (the original
cifs protocol maximum) also requires that the server support
a new Unix Capability flag (for very large read) which some
newer servers (e.g. Samba 3.0.26 or later) do. rsize can be
set from a minimum of 2048 to a maximum of 130048 (127K or
CIFSMaxBufSize, whichever is smaller)
wsize default write size (default 57344)
maximum wsize currently allowed by CIFS is 57344 (fourteen
4096 byte pages)
rw mount the network share read-write (note that the
server may still consider the share read-only)
ro mount network share read-only
version used to distinguish different versions of the
mount helper utility (not typically needed)
sep if first mount option (after the -o), overrides
the comma as the separator between the mount
parms. e.g.
-o user=myname,password=mypassword,domain=mydom
could be passed instead with period as the separator by
-o sep=.user=myname.password=mypassword.domain=mydom
this might be useful when comma is contained within username
or password or domain. This option is less important
when the cifs mount helper cifs.mount (version 1.1 or later)
is used.
nosuid Do not allow remote executables with the suid bit
program to be executed. This is only meaningful for mounts
to servers such as Samba which support the CIFS Unix Extensions.
If you do not trust the servers in your network (your mount
targets) it is recommended that you specify this option for
greater security.
exec Permit execution of binaries on the mount.
noexec Do not permit execution of binaries on the mount.
dev Recognize block devices on the remote mount.
nodev Do not recognize devices on the remote mount.
suid Allow remote files on this mountpoint with suid enabled to
be executed (default for mounts when executed as root,
nosuid is default for user mounts).
credentials Although ignored by the cifs kernel component, it is used by
the mount helper, mount.cifs. When mount.cifs is installed it
opens and reads the credential file specified in order
to obtain the userid and password arguments which are passed to
the cifs vfs.
guest Although ignored by the kernel component, the mount.cifs
mount helper will not prompt the user for a password
if guest is specified on the mount options. If no
password is specified a null password will be used.
perm Client does permission checks (vfs_permission check of uid
and gid of the file against the mode and desired operation),
Note that this is in addition to the normal ACL check on the
target machine done by the server software.
Client permission checking is enabled by default.
noperm Client does not do permission checks. This can expose
files on this mount to access by other users on the local
client system. It is typically only needed when the server
supports the CIFS Unix Extensions but the UIDs/GIDs on the
client and server system do not match closely enough to allow
access by the user doing the mount, but it may be useful with
non CIFS Unix Extension mounts for cases in which the default
mode is specified on the mount but is not to be enforced on the
client (e.g. perhaps when MultiUserMount is enabled)
Note that this does not affect the normal ACL check on the
target machine done by the server software (of the server
ACL against the user name provided at mount time).
serverino Use server's inode numbers instead of generating automatically
incrementing inode numbers on the client. Although this will
make it easier to spot hardlinked files (as they will have
the same inode numbers) and inode numbers may be persistent,
note that the server does not guarantee that the inode numbers
are unique if multiple server side mounts are exported under a
single share (since inode numbers on the servers might not
be unique if multiple filesystems are mounted under the same
shared higher level directory). Note that some older
(e.g. pre-Windows 2000) do not support returning UniqueIDs
or the CIFS Unix Extensions equivalent and for those
this mount option will have no effect. Exporting cifs mounts
under nfsd requires this mount option on the cifs mount.
noserverino Client generates inode numbers (rather than using the actual one
from the server) by default.
setuids If the CIFS Unix extensions are negotiated with the server
the client will attempt to set the effective uid and gid of
the local process on newly created files, directories, and
devices (create, mkdir, mknod). If the CIFS Unix Extensions
are not negotiated, for newly created files and directories
instead of using the default uid and gid specified on
the mount, cache the new file's uid and gid locally which means
that the uid for the file can change when the inode is
reloaded (or the user remounts the share).
nosetuids The client will not attempt to set the uid and gid on
on newly created files, directories, and devices (create,
mkdir, mknod) which will result in the server setting the
uid and gid to the default (usually the server uid of the
user who mounted the share). Letting the server (rather than
the client) set the uid and gid is the default. If the CIFS
Unix Extensions are not negotiated then the uid and gid for
new files will appear to be the uid (gid) of the mounter or the
uid (gid) parameter specified on the mount.
netbiosname When mounting to servers via port 139, specifies the RFC1001
source name to use to represent the client netbios machine
name when doing the RFC1001 netbios session initialize.
direct Do not do inode data caching on files opened on this mount.
This precludes mmaping files on this mount. In some cases
with fast networks and little or no caching benefits on the
client (e.g. when the application is doing large sequential
reads bigger than page size without rereading the same data)
this can provide better performance than the default
behavior which caches reads (readahead) and writes
(writebehind) through the local Linux client pagecache
if oplock (caching token) is granted and held. Note that
direct allows write operations larger than page size
to be sent to the server.
acl Allow setfacl and getfacl to manage posix ACLs if server
supports them. (default)
noacl Do not allow setfacl and getfacl calls on this mount
user_xattr Allow getting and setting user xattrs (those attributes whose
name begins with "user." or "os2.") as OS/2 EAs (extended
attributes) to the server. This allows support of the
setfattr and getfattr utilities. (default)
nouser_xattr Do not allow getfattr/setfattr to get/set/list xattrs
mapchars Translate six of the seven reserved characters (not backslash)
*?<>|:
to the remap range (above 0xF000), which also
allows the CIFS client to recognize files created with
such characters by Windows's POSIX emulation. This can
also be useful when mounting to most versions of Samba
(which also forbids creating and opening files
whose names contain any of these seven characters).
This has no effect if the server does not support
Unicode on the wire.
nomapchars Do not translate any of these seven characters (default).
nocase Request case insensitive path name matching (case
sensitive is the default if the server suports it).
(mount option "ignorecase" is identical to "nocase")
posixpaths If CIFS Unix extensions are supported, attempt to
negotiate posix path name support which allows certain
characters forbidden in typical CIFS filenames, without
requiring remapping. (default)
noposixpaths If CIFS Unix extensions are supported, do not request
posix path name support (this may cause servers to
reject creatingfile with certain reserved characters).
nounix Disable the CIFS Unix Extensions for this mount (tree
connection). This is rarely needed, but it may be useful
in order to turn off multiple settings all at once (ie
posix acls, posix locks, posix paths, symlink support
and retrieving uids/gids/mode from the server) or to
work around a bug in server which implement the Unix
Extensions.
nobrl Do not send byte range lock requests to the server.
This is necessary for certain applications that break
with cifs style mandatory byte range locks (and most
cifs servers do not yet support requesting advisory
byte range locks).
remount remount the share (often used to change from ro to rw mounts
or vice versa)
cifsacl Report mode bits (e.g. on stat) based on the Windows ACL for
the file. (EXPERIMENTAL)
servern Specify the server 's netbios name (RFC1001 name) to use
when attempting to setup a session to the server.
This is needed for mounting to some older servers (such
as OS/2 or Windows 98 and Windows ME) since they do not
support a default server name. A server name can be up
to 15 characters long and is usually uppercased.
sfu When the CIFS Unix Extensions are not negotiated, attempt to
create device files and fifos in a format compatible with
Services for Unix (SFU). In addition retrieve bits 10-12
of the mode via the SETFILEBITS extended attribute (as
SFU does). In the future the bottom 9 bits of the
mode also will be emulated using queries of the security
descriptor (ACL).
sign Must use packet signing (helps avoid unwanted data modification
by intermediate systems in the route). Note that signing
does not work with lanman or plaintext authentication.
seal Must seal (encrypt) all data on this mounted share before
sending on the network. Requires support for Unix Extensions.
Note that this differs from the sign mount option in that it
causes encryption of data sent over this mounted share but other
shares mounted to the same server are unaffected.
sec Security mode. Allowed values are:
none attempt to connection as a null user (no name)
krb5 Use Kerberos version 5 authentication
krb5i Use Kerberos authentication and packet signing
ntlm Use NTLM password hashing (default)
ntlmi Use NTLM password hashing with signing (if
/proc/fs/cifs/PacketSigningEnabled on or if
server requires signing also can be the default)
ntlmv2 Use NTLMv2 password hashing
ntlmv2i Use NTLMv2 password hashing with packet signing
lanman (if configured in kernel config) use older
lanman hash
hard Retry file operations if server is not responding
soft Limit retries to unresponsive servers (usually only
one retry) before returning an error. (default)
The mount.cifs mount helper also accepts a few mount options before -o
including:
-S take password from stdin (equivalent to setting the environment
variable "PASSWD_FD=0"
-V print mount.cifs version
-? display simple usage information
With most 2.6 kernel versions of modutils, the version of the cifs kernel
module can be displayed via modinfo.
Misc /proc/fs/cifs Flags and Debug Info
=======================================
Informational pseudo-files:
DebugData Displays information about active CIFS sessions
and shares, as well as the cifs.ko version.
Stats Lists summary resource usage information as well as per
share statistics, if CONFIG_CIFS_STATS in enabled
in the kernel configuration.
Configuration pseudo-files:
MultiuserMount If set to one, more than one CIFS session to
the same server ip address can be established
if more than one uid accesses the same mount
point and if the uids user/password mapping
information is available. (default is 0)
PacketSigningEnabled If set to one, cifs packet signing is enabled
and will be used if the server requires
it. If set to two, cifs packet signing is
required even if the server considers packet
signing optional. (default 1)
SecurityFlags Flags which control security negotiation and
also packet signing. Authentication (may/must)
flags (e.g. for NTLM and/or NTLMv2) may be combined with
the signing flags. Specifying two different password
hashing mechanisms (as "must use") on the other hand
does not make much sense. Default flags are
0x07007
(NTLM, NTLMv2 and packet signing allowed). The maximum
allowable flags if you want to allow mounts to servers
using weaker password hashes is 0x37037 (lanman,
plaintext, ntlm, ntlmv2, signing allowed). Some
SecurityFlags require the corresponding menuconfig
options to be enabled (lanman and plaintext require
CONFIG_CIFS_WEAK_PW_HASH for example). Enabling
plaintext authentication currently requires also
enabling lanman authentication in the security flags
because the cifs module only supports sending
laintext passwords using the older lanman dialect
form of the session setup SMB. (e.g. for authentication
using plain text passwords, set the SecurityFlags
to 0x30030):
may use packet signing 0x00001
must use packet signing 0x01001
may use NTLM (most common password hash) 0x00002
must use NTLM 0x02002
may use NTLMv2 0x00004
must use NTLMv2 0x04004
may use Kerberos security 0x00008
must use Kerberos 0x08008
may use lanman (weak) password hash 0x00010
must use lanman password hash 0x10010
may use plaintext passwords 0x00020
must use plaintext passwords 0x20020
(reserved for future packet encryption) 0x00040
cifsFYI If set to non-zero value, additional debug information
will be logged to the system error log. This field
contains three flags controlling different classes of
debugging entries. The maximum value it can be set
to is 7 which enables all debugging points (default 0).
Some debugging statements are not compiled into the
cifs kernel unless CONFIG_CIFS_DEBUG2 is enabled in the
kernel configuration. cifsFYI may be set to one or
nore of the following flags (7 sets them all):
log cifs informational messages 0x01
log return codes from cifs entry points 0x02
log slow responses (ie which take longer than 1 second)
CONFIG_CIFS_STATS2 must be enabled in .config 0x04
traceSMB If set to one, debug information is logged to the
system error log with the start of smb requests
and responses (default 0)
LookupCacheEnable If set to one, inode information is kept cached
for one second improving performance of lookups
(default 1)
OplockEnabled If set to one, safe distributed caching enabled.
(default 1)
LinuxExtensionsEnabled If set to one then the client will attempt to
use the CIFS "UNIX" extensions which are optional
protocol enhancements that allow CIFS servers
to return accurate UID/GID information as well
as support symbolic links. If you use servers
such as Samba that support the CIFS Unix
extensions but do not want to use symbolic link
support and want to map the uid and gid fields
to values supplied at mount (rather than the
actual values, then set this to zero. (default 1)
Experimental When set to 1 used to enable certain experimental
features (currently enables multipage writes
when signing is enabled, the multipage write
performance enhancement was disabled when
signing turned on in case buffer was modified
just before it was sent, also this flag will
be used to use the new experimental directory change
notification code).
These experimental features and tracing can be enabled by changing flags in
/proc/fs/cifs (after the cifs module has been installed or built into the
kernel, e.g. insmod cifs). To enable a feature set it to 1 e.g. to enable
tracing to the kernel message log type:
echo 7 > /proc/fs/cifs/cifsFYI
cifsFYI functions as a bit mask. Setting it to 1 enables additional kernel
logging of various informational messages. 2 enables logging of non-zero
SMB return codes while 4 enables logging of requests that take longer
than one second to complete (except for byte range lock requests).
Setting it to 4 requires defining CONFIG_CIFS_STATS2 manually in the
source code (typically by setting it in the beginning of cifsglob.h),
and setting it to seven enables all three. Finally, tracing
the start of smb requests and responses can be enabled via:
echo 1 > /proc/fs/cifs/traceSMB
Two other experimental features are under development. To test these
requires enabling CONFIG_CIFS_EXPERIMENTAL
cifsacl support needed to retrieve approximated mode bits based on
the contents on the CIFS ACL.
DNOTIFY fcntl: needed for support of directory change
notification and perhaps later for file leases)
Per share (per client mount) statistics are available in /proc/fs/cifs/Stats
if the kernel was configured with cifs statistics enabled. The statistics
represent the number of successful (ie non-zero return code from the server)
SMB responses to some of the more common commands (open, delete, mkdir etc.).
Also recorded is the total bytes read and bytes written to the server for
that share. Note that due to client caching effects this can be less than the
number of bytes read and written by the application running on the client.
The statistics for the number of total SMBs and oplock breaks are different in
that they represent all for that share, not just those for which the server
returned success.
Also note that "cat /proc/fs/cifs/DebugData" will display information about
the active sessions and the shares that are mounted.
Enabling Kerberos (extended security) works but requires version 1.2 or later
of the helper program cifs.upcall to be present and to be configured in the
/etc/request-key.conf file. The cifs.upcall helper program is from the Samba
project(http://www.samba.org). NTLM and NTLMv2 and LANMAN support do not
require this helper. Note that NTLMv2 security (which does not require the
cifs.upcall helper program), instead of using Kerberos, is sufficient for
some use cases.
Enabling DFS support (used to access shares transparently in an MS-DFS
global name space) requires that CONFIG_CIFS_EXPERIMENTAL be enabled. In
addition, DFS support for target shares which are specified as UNC
names which begin with host names (rather than IP addresses) requires
a user space helper (such as cifs.upcall) to be present in order to
translate host names to ip address, and the user space helper must also
be configured in the file /etc/request-key.conf
To use cifs Kerberos and DFS support, the Linux keyutils package should be
installed and something like the following lines should be added to the
/etc/request-key.conf file:
create cifs.spnego * * /usr/local/sbin/cifs.upcall %k
create dns_resolver * * /usr/local/sbin/cifs.upcall %k
etercifs/TODO 0 → 100644
View file @ 3c073451
Version 1.53 May 20, 2008
A Partial List of Missing Features
==================================
Contributions are welcome. There are plenty of opportunities
for visible, important contributions to this module. Here
is a partial list of the known problems and missing features:
a) Support for SecurityDescriptors(Windows/CIFS ACLs) for chmod/chgrp/chown
so that these operations can be supported to Windows servers
b) Mapping POSIX ACLs (and eventually NFSv4 ACLs) to CIFS
SecurityDescriptors
c) Better pam/winbind integration (e.g. to handle uid mapping
better)
d) Cleanup now unneeded SessSetup code in
fs/cifs/connect.c and add back in NTLMSSP code if any servers
need it
e) fix NTLMv2 signing when two mounts with different users to same
server.
f) Directory entry caching relies on a 1 second timer, rather than
using FindNotify or equivalent. - (started)
g) quota support (needs minor kernel change since quota calls
to make it to network filesystems or deviceless filesystems)
h) investigate sync behavior (including syncpage) and check
for proper behavior of intr/nointr
i) improve support for very old servers (OS/2 and Win9x for example)
Including support for changing the time remotely (utimes command).
j) hook lower into the sockets api (as NFS/SunRPC does) to avoid the
extra copy in/out of the socket buffers in some cases.
k) Better optimize open (and pathbased setfilesize) to reduce the
oplock breaks coming from windows srv. Piggyback identical file
opens on top of each other by incrementing reference count rather
than resending (helps reduce server resource utilization and avoid
spurious oplock breaks).
l) Improve performance of readpages by sending more than one read
at a time when 8 pages or more are requested. In conjuntion
add support for async_cifs_readpages.
m) Add support for storing symlink info to Windows servers
in the Extended Attribute format their SFU clients would recognize.
n) Finish fcntl D_NOTIFY support so kde and gnome file list windows
will autorefresh (partially complete by Asser). Needs minor kernel
vfs change to support removing D_NOTIFY on a file.
o) Add GUI tool to configure /proc/fs/cifs settings and for display of
the CIFS statistics (started)
p) implement support for security and trusted categories of xattrs
(requires minor protocol extension) to enable better support for SELINUX
q) Implement O_DIRECT flag on open (already supported on mount)
r) Create UID mapping facility so server UIDs can be mapped on a per
mount or a per server basis to client UIDs or nobody if no mapping
exists. This is helpful when Unix extensions are negotiated to
allow better permission checking when UIDs differ on the server
and client. Add new protocol request to the CIFS protocol
standard for asking the server for the corresponding name of a
particular uid.
s) Add support for CIFS Unix and also the newer POSIX extensions to the
server side for Samba 4.
t) In support for OS/2 (LANMAN 1.2 and LANMAN2.1 based SMB servers)
need to add ability to set time to server (utimes command)
u) DOS attrs - returned as pseudo-xattr in Samba format (check VFAT and NTFS for this too)
v) mount check for unmatched uids
w) Add support for new vfs entry points for setlease and fallocate
x) Fix Samba 3 server to handle Linux kernel aio so dbench with lots of
processes can proceed better in parallel (on the server)
y) Fix Samba 3 to handle reads/writes over 127K (and remove the cifs mount
restriction of wsize max being 127K)
KNOWN BUGS (updated April 24, 2007)
====================================
See http://bugzilla.samba.org - search on product "CifsVFS" for
current bug list.
1) existing symbolic links (Windows reparse points) are recognized but
can not be created remotely. They are implemented for Samba and those that
support the CIFS Unix extensions, although earlier versions of Samba
overly restrict the pathnames.
2) follow_link and readdir code does not follow dfs junctions
but recognizes them
3) create of new files to FAT partitions on Windows servers can
succeed but still return access denied (appears to be Windows
server not cifs client problem) and has not been reproduced recently.
NTFS partitions do not have this problem.
4) Unix/POSIX capabilities are reset after reconnection, and affect
a few fields in the tree connection but we do do not know which
superblocks to apply these changes to. We should probably walk
the list of superblocks to set these. Also need to check the
flags on the second mount to the same share, and see if we
can do the same trick that NFS does to remount duplicate shares.
Misc testing to do
==================
1) check out max path names and max path name components against various server
types. Try nested symlinks (8 deep). Return max path name in stat -f information
2) Modify file portion of ltp so it can run against a mounted network
share and run it against cifs vfs in automated fashion.
3) Additional performance testing and optimization using iozone and similar -
there are some easy changes that can be done to parallelize sequential writes,
and when signing is disabled to request larger read sizes (larger than
negotiated size) and send larger write sizes to modern servers.
4) More exhaustively test against less common servers. More testing
against Windows 9x, Windows ME servers.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment