Looks like this bug has been here since these write counts were
introduced, not sure why it was just noticed now.

Thanks also to Jan Kara for pointing out the problem.

Cc: stable@vger.kernel.org
Reported-by: NMatthew Rahtz <mrahtz@rapitasystems.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

There is a regression in
	208d0acc 2014-01-07 nfsd4: break only delegations when appropriate
which deletes an nfserrno() call in nfsd_setattr() (by accident,
probably), and NFSD becomes ignoring an error from VFS.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

If a file is unlinked or renamed between the time when we do the local
open and the time when we get the delegation, then we will return to the
client indicating that it holds a delegation even though the file no
longer exists under the name it was open under.

But a client performing an open-by-name, when it is returned a
delegation, must be able to assume that the file is still linked at the
name it was opened under.

So, hold the parent i_mutex for longer to prevent concurrent renames or
unlinks.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Remove the boilerplate code to marshall and unmarhall ACL objects into
xattrs and operate on the posix_acl objects directly.  Also move all
the ACL handling code into nfs?acl.c where it belongs.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

As a temporary fix, nfsd was breaking all leases on unlink, link,
rename, and setattr.

Now that we can distinguish between leases and delegations, we can be
nicer and break only the delegations, and not bother lease-holders with
operations they don't care about.

And we get to delete some code while we're at it.

Note that in the presence of delegations the vfs calls here all return
-EWOULDBLOCK instead of blocking, so nfsd threads will not get stuck
waiting for delegation returns.
Acked-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

There is an inconsistency in the handling of SUID/SGID file
bits after chown() between NFS and other local file systems.

Local file systems (for example, ext3, ext4, xfs, btrfs) revoke
SUID/SGID bits after chown() on a regular file even if
the owner/group of the file has not been changed:

~# touch file; chmod ug+s file; chmod u+x file
~# ls -l file
-rwsr-Sr-- 1 root root 0 Dec  6 04:49 file
~# chown root file; ls -l file
-rwxr-Sr-- 1 root root 0 Dec  6 04:49 file

but NFS doesn't do that:

~# touch file; chmod ug+s file; chmod u+x file
~# ls -l file
-rwsr-Sr-- 1 root root 0 Dec  6 04:49 file
~# chown root file; ls -l file
-rwsr-Sr-- 1 root root 0 Dec  6 04:49 file

NFS does that only if the owner/group has been changed:

~# touch file; chmod ug+s file; chmod u+x file
~# ls -l file
-rwsr-Sr-- 1 root root 0 Dec  6 05:02 file
~# chown bin file; ls -l file
-rwxr-Sr-- 1 bin root 0 Dec  6 05:02 file

See: http://pubs.opengroup.org/onlinepubs/9699919799/functions/chown.html

 "If the specified file is a regular file, one or more of
 the S_IXUSR, S_IXGRP, or S_IXOTH bits of the file mode are set,
 and the process has appropriate privileges, it is
 implementation-defined whether the set-user-ID and set-group-ID
 bits are altered."

So both variants are acceptable by POSIX.

This patch makes NFS to behave like local file systems.
Signed-off-by: NStanislav Kholmanskikh <stanislav.kholmanskikh@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Use a straight goto error label style in nfsd_setattr to make sure
we always do the put_write_access call after we got it earlier.

Note that the we have been failing to do that in the case
nfsd_break_lease() returns an error, a bug introduced into 2.6.38 with
6a76bebe "nfsd4: break lease on nfsd
setattr".
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Split out two helpers to make the code more readable and easier to verify
for correctness.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

NFSv4 uses leases to guarantee that clients can cache metadata as well
as data.

Cc: Mikulas Patocka <mikulas@artax.karlin.mff.cuni.cz>
Cc: David Howells <dhowells@redhat.com>
Cc: Tyler Hicks <tyhicks@canonical.com>
Cc: Dustin Kirkland <dustin.kirkland@gazzang.com>
Acked-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Cc: Tyler Hicks <tyhicks@canonical.com>
Cc: Dustin Kirkland <dustin.kirkland@gazzang.com>
Acked-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Cc: David Howells <dhowells@redhat.com>
Acked-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

We need to break delegations on any operation that changes the set of
links pointing to an inode.  Start with unlink.

Such operations also hold the i_mutex on a parent directory.  Breaking a
delegation may require waiting for a timeout (by default 90 seconds) in
the case of a unresponsive NFS client.  To avoid blocking all directory
operations, we therefore drop locks before waiting for the delegation.
The logic then looks like:

	acquire locks
	...
	test for delegation; if found:
		take reference on inode
		release locks
		wait for delegation break
		drop reference on inode
		retry

It is possible this could never terminate.  (Even if we take precautions
to prevent another delegation being acquired on the same inode, we could
get a different inode on each retry.)  But this seems very unlikely.

The initial test for a delegation happens after the lock on the target
inode is acquired, but the directory inode may have been acquired
further up the call stack.  We therefore add a "struct inode **"
argument to any intervening functions, which we use to pass the inode
back up to the caller in the case it needs a delegation synchronously
broken.

Cc: David Howells <dhowells@redhat.com>
Cc: Tyler Hicks <tyhicks@canonical.com>
Cc: Dustin Kirkland <dustin.kirkland@gazzang.com>
Acked-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The following call chain:
------------------------------------------------------------
nfs4_get_vfs_file
- nfsd_open
  - dentry_open
    - do_dentry_open
      - __get_file_write_access
        - get_write_access
          - return atomic_inc_unless_negative(&inode->i_writecount) ? 0 : -ETXTBSY;
------------------------------------------------------------

can result in the following state:
------------------------------------------------------------
struct nfs4_file {
...
  fi_fds = {0xffff880c1fa65c80, 0xffffffffffffffe6, 0x0},
  fi_access = {{
      counter = 0x1
    }, {
      counter = 0x0
    }},
...
------------------------------------------------------------

1) First time around, in nfs4_get_vfs_file() fp->fi_fds[O_WRONLY] is
NULL, hence nfsd_open() is called where we get status set to an error
and fp->fi_fds[O_WRONLY] to -ETXTBSY. Thus we do not reach
nfs4_file_get_access() and fi_access[O_WRONLY] is not incremented.

2) Second time around, in nfs4_get_vfs_file() fp->fi_fds[O_WRONLY] is
NOT NULL (-ETXTBSY), so nfsd_open() is NOT called, but
nfs4_file_get_access() IS called and fi_access[O_WRONLY] is incremented.
Thus we leave a landmine in the form of the nfs4_file data structure in
an incorrect state.

3) Eventually, when __nfs4_file_put_access() is called it finds
fi_access[O_WRONLY] being non-zero, it decrements it and calls
nfs4_file_put_fd() which tries to fput -ETXTBSY.
------------------------------------------------------------
...
     [exception RIP: fput+0x9]
     RIP: ffffffff81177fa9  RSP: ffff88062e365c90  RFLAGS: 00010282
     RAX: ffff880c2b3d99cc  RBX: ffff880c2b3d9978  RCX: 0000000000000002
     RDX: dead000000100101  RSI: 0000000000000001  RDI: ffffffffffffffe6
     RBP: ffff88062e365c90   R8: ffff88041fe797d8   R9: ffff88062e365d58
     R10: 0000000000000008  R11: 0000000000000000  R12: 0000000000000001
     R13: 0000000000000007  R14: 0000000000000000  R15: 0000000000000000
     ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
  #9 [ffff88062e365c98] __nfs4_file_put_access at ffffffffa0562334 [nfsd]
 #10 [ffff88062e365cc8] nfs4_file_put_access at ffffffffa05623ab [nfsd]
 #11 [ffff88062e365ce8] free_generic_stateid at ffffffffa056634d [nfsd]
 #12 [ffff88062e365d18] release_open_stateid at ffffffffa0566e4b [nfsd]
 #13 [ffff88062e365d38] nfsd4_close at ffffffffa0567401 [nfsd]
 #14 [ffff88062e365d88] nfsd4_proc_compound at ffffffffa0557f28 [nfsd]
 #15 [ffff88062e365dd8] nfsd_dispatch at ffffffffa054543e [nfsd]
 #16 [ffff88062e365e18] svc_process_common at ffffffffa04ba5a4 [sunrpc]
 #17 [ffff88062e365e98] svc_process at ffffffffa04babe0 [sunrpc]
 #18 [ffff88062e365eb8] nfsd at ffffffffa0545b62 [nfsd]
 #19 [ffff88062e365ee8] kthread at ffffffff81090886
 #20 [ffff88062e365f48] kernel_thread at ffffffff8100c14a
------------------------------------------------------------

Cc: stable@vger.kernel.org
Signed-off-by: NHarshula Jayasuriya <harshula@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

iterate_dir(): new helper, replacing vfs_readdir().

struct dir_context: contains the readdir callback (and will get more stuff
in it), embedded into whatever data that callback wants to deal with;
eventually, we'll be passing it to ->readdir() replacement instead of
(data,filldir) pair.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Implement labeled NFS on the server: encoding and decoding, and writing
and reading, of file labels.

Enabled with CONFIG_NFSD_V4_SECURITY_LABEL.
Signed-off-by: NMatthew N. Dodd <Matthew.Dodd@sparta.com>
Signed-off-by: NMiguel Rodel Felipe <Rodel_FM@dsi.a-star.edu.sg>
Signed-off-by: NPhua Eu Gene <PHUA_Eu_Gene@dsi.a-star.edu.sg>
Signed-off-by: NKhin Mi Mi Aung <Mi_Mi_AUNG@dsi.a-star.edu.sg>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

We again check for the EXDEV a little later on, so the first check is
redundant.  This check is also slightly racier, since a badly timed
eviction from the export cache could leave us with the two fh_export
pointers pointing to two different cache entries which each refer to the
same underlying export.

It's better to compare vfsmounts as the later check does, but that
leaves a minor security hole in the case where the two exports refer to
two different directories especially if (for example) they have
different root-squashing options.

So, compare ex_path.dentry too.
Reported-by: NJoe Habermann <joe.habermann@gmail.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

vfs_writev() updates the offset argument - but the code then passes the
offset to vfs_fsync_range(). Since offset now points to the offset after
what was just written, this is probably not what was intended

Introduced by face1502 "nfsd: use
vfs_fsync_range(), not O_SYNC, for stable writes".
Signed-off-by: NKent Overstreet <koverstreet@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: stable@vger.kernel.org
Reviewed-by: NZach Brown <zab@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Use uid_eq(uid, GLOBAL_ROOT_UID) instead of !uid.
Use gid_eq(gid, GLOBAL_ROOT_GID) instead of !gid.
Use uid_eq(uid, INVALID_UID) instead of uid == -1
Use gid_eq(uid, INVALID_GID) instead of gid == -1
Use uid = GLOBAL_ROOT_UID instead of uid = 0;
Use gid = GLOBAL_ROOT_GID instead of gid = 0;
Use !uid_eq(uid1, uid2) instead of uid1 != uid2.
Use !gid_eq(gid1, gid2) instead of gid1 != gid2.
Use uid_eq(uid1, uid2) instead of uid1 == uid2.

Cc: "J. Bruce Fields" <bfields@fieldses.org>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>

This reverts commit 79f77bf9.

This is obviously wrong, and I have no idea how I missed seeing the
warning in testing: I must just not have looked at the right logs.  The
caller bumps rq_resused/rq_next_page, so it will always be hit on a
large enough read.
Reported-by: NDave Jones <davej@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

It may be a matter of personal taste, but I find this makes the code
clearer.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

As far as I can tell this shouldn't currently happen--or if it does,
something is wrong and data is going to be corrupted.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

With NFSv4, if we create a file then open it we explicit avoid checking
the permissions on the file during the open because the fact that we
created it ensures we should be allow to open it (the create and the
open should appear to be a single operation).

However if the reply to an EXCLUSIVE create gets lots and the client
resends the create, the current code will perform the permission check -
because it doesn't realise that it did the open already..

This patch should fix this.

Note that I haven't actually seen this cause a problem.  I was just
looking at the code trying to figure out a different EXCLUSIVE open
related issue, and this looked wrong.

(Fix confirmed with pynfs 4.0 test OPEN4--bfields)

Cc: stable@kernel.org
Signed-off-by: NNeilBrown <neilb@suse.de>
[bfields: use OWNER_OVERRIDE and update for 4.1]
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

NFSv4 shares the same struct file across multiple writes.  (And we'd
like NFSv2 and NFSv3 to do that as well some day.)

So setting O_SYNC on the struct file as a way to request a synchronous
write doesn't work.

Instead, do a vfs_fsync_range() in that case.
Reported-by: NPeter Staubach <pstaubach@exagrid.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

I don't really see how you could claim to support nfsd and not support
fsync somehow.

And in practice a quick look through the exportable filesystems suggests
the only ones without an ->fsync are read-only (efs, isofs, squashfs) or
in-memory (shmem).

Also, performing a write and then returning an error if the sync fails
(as we would do here in the wgather case) seems unhelpful to clients.

Also remove an incorrect comment.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

 - Pass the user namespace the uid and gid values in the xattr are stored
   in into posix_acl_from_xattr.

 - Pass the user namespace kuid and kgid values should be converted into
   when storing uid and gid values in an xattr in posix_acl_to_xattr.

- Modify all callers of posix_acl_from_xattr and posix_acl_to_xattr to
  pass in &init_user_ns.

In the short term this change is not strictly needed but it makes the
code clearer.  In the longer term this change is necessary to be able to
mount filesystems outside of the initial user namespace that natively
store posix acls in the linux xattr format.

Cc: Theodore Tso <tytso@mit.edu>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Jan Kara <jack@suse.cz>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>

As we already do in readv, writev.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

When mnt_want_write() starts to handle freezing it will get a full lock
semantics requiring proper lock ordering. So push mnt_want_write() call
consistently outside of i_mutex.

CC: linux-nfs@vger.kernel.org
CC: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

all we want is a boolean flag, same as the method gets now
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

We normally allow the owner of a file to override permissions checks on
IO operations, since:
	- the client will take responsibility for doing an access check
	  on open;
	- the permission checks offer no protection against malicious
	  clients--if they can authenticate as the file's owner then
	  they can always just change its permissions;
	- checking permission on each IO operation breaks the usual
	  posix rule that permission is checked only on open.

However, we've never allowed the owner to override permissions on
readdir operations, even though the above logic would also apply to
directories.  I've never heard of this causing a problem, probably
because a) simultaneously opening and creating a directory (with
restricted mode) isn't possible, and b) opening a directory, then
chmod'ing it, is rare.

Our disallowal of owner-override on directories appears to be an
accident, though--the readdir itself succeeds, and then we fail just
because lookup_one_len() calls in our filldir methods fail.

I'm not sure what the easiest fix for that would be.  For now, just make
this behavior obvious by denying the override right at the start.

This also fixes some odd v4 behavior: with the rdattr_error attribute
requested, it would perform the readdir but return an ACCES error with
each entry.
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

They're equivalent, but SEEK_SET is more informative...

Cc: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Allow a v3 unchecked open of a non-regular file succeed as if it were a
lookup; typically a client in such a case will want to fall back on a
local open, so succeeding and giving it the filehandle is more useful
than failing with nfserr_exist, which makes it appear that nothing at
all exists by that name.

Similarly for v4, on an open-create, return the same errors we would on
an attempt to open a non-regular file, instead of returning
nfserr_exist.

This fixes a problem found doing a v4 open of a symlink with
O_RDONLY|O_CREAT, which resulted in the current client returning EEXIST.

Thanks also to Trond for analysis.

Cc: stable@kernel.org
Reported-by: NOrion Poplawski <orion@cora.nwra.com>
Tested-by: NOrion Poplawski <orion@cora.nwra.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Use 32-bit or 64-bit llseek() hashes for directory offsets depending on
the NFS version. NFSv2 gets 32-bit hashes only.

NOTE: This patch got rather complex as Christoph asked to set the
filp->f_mode flag in the open call or immediatly after dentry_open()
in nfsd_open() to avoid races.
Personally I still do not see a reason for that and in my opinion
FMODE_32BITHASH/FMODE_64BITHASH flags could be set nfsd_readdir(), as it
follows directly after nfsd_open() without a chance of races.
Signed-off-by: NBernd Schubert <bernd.schubert@itwm.fraunhofer.de>
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>
Acked-by: J. Bruce Fields<bfields@redhat.com>

Just rename this variable, as the next patch will add a flag and
'access' as variable name would not be correct any more.
Signed-off-by: NBernd Schubert <bernd.schubert@itwm.fraunhofer.de>
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>
Acked-by: J. Bruce Fields<bfields@redhat.com>

As of fedfs-utils-0.8.0, user space stores all NFS junction
information in a single extended attribute: "trusted.junction.nfs".

Both FedFS and NFS basic junctions are stored in this one attribute,
and the intention is that all future forms of NFS junction metadata
will be stored in this attribute.  Other protocols may use a different
extended attribute.

Thus NFSD needs to look only for that one extended attribute.  The
"trusted.junction.type" xattr is deprecated.  fedfs-utils-0.8.0 will
continue to attach a "trusted.junction.type" xattr to junctions, but
future fedfs-utils releases may no longer do that.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NJ. Bruce Fields <bfields@redhat.com>