write_lock(&current->fs->lock) guarantees we can't wrongly miss
LSM_UNSAFE_SHARE, this is what we care about. Use rcu_read_lock()
instead of ->siglock to iterate over the sub-threads. We must see
all CLONE_THREAD|CLONE_FS threads which didn't pass exit_fs(), it
takes fs->lock too.

With or without this patch we can miss the freshly cloned thread
and set LSM_UNSAFE_SHARE, we don't care.
Signed-off-by: NOleg Nesterov <oleg@redhat.com>
Acked-by: NRoland McGrath <roland@redhat.com>
[ Fixed lock/unlock typo  - Hugh ]
Acked-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

If do_execve() fails after check_unsafe_exec(), it clears fs->in_exec
unconditionally. This is wrong if we race with our sub-thread which
also does do_execve:

	Two threads T1 and T2 and another process P, all share the same
	->fs.

	T1 starts do_execve(BAD_FILE). It calls check_unsafe_exec(), since
	->fs is shared, we set LSM_UNSAFE but not ->in_exec.

	P exits and decrements fs->users.

	T2 starts do_execve(), calls check_unsafe_exec(), now ->fs is not
	shared, we set fs->in_exec.

	T1 continues, open_exec(BAD_FILE) fails, we clear ->in_exec and
	return to the user-space.

	T1 does clone(CLONE_FS /* without CLONE_THREAD */).

	T2 continues without LSM_UNSAFE_SHARE while ->fs is shared with
	another process.

Change check_unsafe_exec() to return res = 1 if we set ->in_exec, and change
do_execve() to clear ->in_exec depending on res.

When do_execve() suceeds, it is safe to clear ->in_exec unconditionally.
It can be set only if we don't share ->fs with another process, and since
we already killed all sub-threads either ->in_exec == 0 or we are the
only user of this ->fs.

Also, we do not need fs->lock to clear fs->in_exec.
Signed-off-by: NOleg Nesterov <oleg@redhat.com>
Acked-by: NRoland McGrath <roland@redhat.com>
Acked-by: NHugh Dickins <hugh@veritas.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The cpu idle field in the output of /proc/stat is too small for cpus
that have been idle for more than a tick. Add the architecture hook
arch_idle_time that allows to add the not accounted idle time of a
sleeping cpu without waking the cpu.

The s390 implementation of arch_idle_time uses the already existing
s390_idle_data per_cpu variable to find the sleep time of a neighboring
idle cpu.
Signed-off-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>

GFS2 has a goal block associated with each inode indicating the
search start position for future block allocations (in fact there
are two, but thats for backward compatibility with GFS1 as they
are set to identical locations in GFS2).

In some circumstances, depending on the ordering of updates to
the inode it was possible for the goal block settings to not
be updated on disk. This patch ensures that the goal block will
always get updated, thus reducing the potential for searching
the same (already allocated) blocks again when looking for free
space during block allocation.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

The new bitfit algorithm was counting from the wrong end of
64 bit words in the bitfield. This fixes it by using __ffs64
instead of fls64
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

If the Orlov allocator is having trouble finding an appropriate block
group, the fallback code could loop forever, causing a soft lockup
warning in find_group_orlov():

BUG: soft lockup - CPU#0 stuck for 61s! [cp:11728]
     ...
Pid: 11728, comm: cp Not tainted (2.6.30-rc1-dirty #77) Lenovo          
EIP: 0060:[<c021650e>] EFLAGS: 00000246 CPU: 0
EIP is at ext4_get_group_desc+0x54/0x9d
    ...
Call Trace:
 [<c0218021>] find_group_orlov+0x2ee/0x334
 [<c0120a5f>] ? sched_clock+0x8/0xb
 [<c02188e3>] ext4_new_inode+0x2cf/0xb1a
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>

Instead of just checking that the extent block number is greater or
equal than s_first_data_block, make sure it it is not pointing into
the block group descriptors, since that is clearly wrong.  This helps
prevent filesystem from getting very badly corrupted in case an extent
block is corrupted.
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>

When using filename encryption with eCryptfs, the value of the symlink
in the lower filesystem is encrypted and stored as a Tag 70 packet.
This results in a longer symlink target than if the target value wasn't
encrypted.

Users were reporting these messages in their syslog:

[ 45.653441] ecryptfs_parse_tag_70_packet: max_packet_size is [56]; real
packet size is [51]
[ 45.653444] ecryptfs_decode_and_decrypt_filename: Could not parse tag
70 packet from filename; copying through filename as-is

This was due to bufsiz, one the arguments in readlink(), being used to
when allocating the buffer passed to the lower inode's readlink().
That symlink target may be very large, but when decoded and decrypted,
could end up being smaller than bufsize.

To fix this, the buffer passed to the lower inode's readlink() will
always be PATH_MAX in size when filename encryption is enabled.  Any
necessary truncation occurs after the decoding and decrypting.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

This patch locks the lower directory inode's i_mutex before calling
lookup_one_len() to find the appropriate dentry in the lower filesystem.
This bug was found thanks to the warning set in commit 2f9092e1.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

A feature was added to the eCryptfs umount helper to automatically
unlink the keys used for an eCryptfs mount from the kernel keyring upon
umount.  This patch keeps the unrecognized mount option warnings for
ecryptfs_unlink_sigs out of the logs.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

ecryptfs_passthrough is a mount option that allows eCryptfs to allow
data to be written to non-eCryptfs files in the lower filesystem.  The
passthrough option was causing data corruption due to it not always
being treated as a non-eCryptfs file.

The first 8 bytes of an eCryptfs file contains the decrypted file size.
This value was being written to the non-eCryptfs files, too.  Also,
extra 0x00 characters were being written to make the file size a
multiple of PAGE_CACHE_SIZE.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

The filename encryption key signature is not properly displayed in
/proc/mounts.  The "ecryptfs_sig=" mount option name is displayed for
all global authentication tokens, included those for filename keys.

This patch checks the global authentication token flags to determine if
the key is a FEKEK or FNEK and prints the appropriate mount option name
before the signature.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

If data is NULL, msg_ctx->msg is set to NULL and then dereferenced
afterwards.  ecryptfs_send_raw_message() is the only place that
ecryptfs_send_miscdev() is called with data being NULL, but the only
caller of that function (ecryptfs_process_helo()) is never called.  In
short, there is currently no way to trigger the NULL pointer
dereference.

This patch removes the two unused functions and modifies
ecryptfs_send_miscdev() to remove the NULL dereferences.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Copies the lower inode attributes to the upper inode before passing the
upper inode to d_instantiate().  This is important for
security_d_instantiate().

The problem was discovered by a user seeing SELinux denials like so:

type=AVC msg=audit(1236812817.898:47): avc:  denied  { 0x100000 } for
pid=3584 comm="httpd" name="testdir" dev=ecryptfs ino=943872
scontext=root:system_r:httpd_t:s0
tcontext=root:object_r:httpd_sys_content_t:s0 tclass=file

Notice target class is file while testdir is really a directory,
confusing the permission translation (0x100000) due to the wrong i_mode.
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Impact: remove possible deadlock condition

There is no reason to use mempool backed allocation for map functions.
Also, because kern mapping is used inside LLDs (e.g. for EH), using
mempool backed allocation can lead to deadlock under extreme
conditions (mempool already consumed by the time a request reached EH
and requests are blocked on EH).

Switch copy/map functions to bio_kmalloc().
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

Impact: fix bio_kmalloc() and its destruction path

bio_kmalloc() was broken in two ways.

* bvec_alloc_bs() first allocates bvec using kmalloc() and then
  ignores it and allocates again like non-kmalloc bvecs.

* bio_kmalloc_destructor() didn't check for and free bio integrity
  data.

This patch fixes the above problems.  kmalloc patch is separated out
from bio_alloc_bioset() and allocates the requested number of bvecs as
inline bvecs.

* bio_alloc_bioset() no longer takes NULL @bs.  None other than
  bio_kmalloc() used it and outside users can't know how it was
  allocated anyway.

* Define and use BIO_POOL_NONE so that pool index check in
  bvec_free_bs() triggers if inline or kmalloc allocated bvec gets
  there.

* Relocate destructors on top of each allocation function so that how
  they're used is more clear.

Jens Axboe suggested allocating bvecs inline.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

This fixes the following BUG:

  # mount -o size=MM -t hugetlbfs none /huge
  hugetlbfs: Bad value 'MM' for mount option 'size=MM'
  ------------[ cut here ]------------
  kernel BUG at fs/super.c:996!

Due to

	BUG_ON(!mnt->mnt_sb);

in vfs_kern_mount().

Also, remove unused #include <linux/quotaops.h>

Cc: William Irwin <wli@holomorphy.com>
Cc: <stable@kernel.org>
Signed-off-by: NAkinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Btrfs fallocate was incorrectly starting a transaction with a lock held
on the extent_io tree for the file, which could deadlock.  Strictly
speaking it was using join_transaction which would be safe, but it is better
to move the transaction outside of the lock.

When preallocated extents are overwritten, btrfs_mark_buffer_dirty was
being called on an unlocked buffer.  This was triggering an assertion and
oops because the lock is supposed to be held.

The bug was calling btrfs_mark_buffer_dirty on a leaf after btrfs_del_item had
been run.  btrfs_del_item takes care of dirtying things, so the solution is a
to skip the btrfs_mark_buffer_dirty call in this case.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Commit ae46141f (NFSv3: Fix posix ACL code)
introduces a bug in the calculation of the XDR header iovec. In the case
where we are inlining the acls, we need to adjust the length of the iovec
req->rq_svec, in addition to adjusting the total buffer length.
Tested-by: NLeonardo Chiquitto <leonardo.lists@gmail.com>
Tested-by: NSuresh Jayaraman <sjayaraman@suse.de>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

"int get_filesystem_list(char * buf)" is called by only
"static void __init get_fs_names(char *page)".
We can mark get_filesystem_list() as "__init".
Signed-off-by: NTetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

There's really no reason to keep vfs_stat_fd and vfs_lstat_fd with
Oleg's vfs_fstatat.  Use vfs_fstatat for the few cases having the
directory fd, and switch all others to vfs_stat / vfs_lstat.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This is a version incorporating Christoph's suggestion.

Separate out common *fstatat functionality into a single function
instead of duplicating it all over the code.
Signed-off-by: NOleg Drokin <green@linuxhacker.ru>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user().
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user()
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user()
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user().
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user().

Note this changes some GFP_NOFS to GFP_KERNEL, since copy_from_user() may
cause pagefault, it's pointless to pass GFP_NOFS to kmalloc().
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Remove open-coded memdup_user()
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

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

Commit 14f7dd63 ("Copy XFS readdir hack into nfsd code") introduced a
bug to generic code which had been extant for a long time in the XFS
version -- it started to call through into lookup_one_len() and hence
into the file systems' ->lookup() methods without i_mutex held on the
directory.

This patch fixes it by locking the directory's i_mutex again before
calling the filldir functions. The original deadlocks which commit
14f7dd63 was designed to avoid are still avoided, because they were due
to fs-internal locking, not i_mutex.

While we're at it, fix the return type of nfsd_buffered_readdir() which
should be a __be32 not an int -- it's an NFS errno, not a Linux errno.
And return nfserrno(-ENOMEM) when allocation fails, not just -ENOMEM.
Sparse would have caught that, if it wasn't so busy bitching about
__cold__.

Commit 05f4f678 ("nfsd4: don't do lookup within readdir in recovery
code") introduced a similar problem with calling lookup_one_len()
without i_mutex, which this patch also addresses. To fix that, it was
necessary to fix the called functions so that they expect i_mutex to be
held; that part was done by J. Bruce Fields.
Signed-off-by: NDavid Woodhouse <David.Woodhouse@intel.com>
Umm-I-can-live-with-that-by: NAl Viro <viro@zeniv.linux.org.uk>
Reported-by: NJ. R. Okajima <hooanon05@yahoo.co.jp>
Tested-by: NJ. Bruce Fields <bfields@citi.umich.edu>
LKML-Reference: <8036.1237474444@jrobl>
Cc: stable@kernel.org
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

In file included from fs/compat_ioctl.c:61:
include/linux/loop.h:59: error: field 'lo_bio_list' has incomplete type
Signed-off-by: NAlexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

mnt should remain the same for all iterations through the list;
as it is, if we have a busy mount, mnt follows into it and isn't
restored for the next iteration.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

is_under() will DTRT anyway.  And yes, is_subdir() behaviour
is intentional.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

AFAICS, we have a subtle bug there: if we have crossed mountpoint
*and* it got mount --move'd away, we'll be holding only one
reference to fs containing dentry - exp->ex_path.mnt.  IOW, we
ought to dput() before exp_put().
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

We shouldn't just touch the namespace of current process
Caught-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Missing conversion from kernel to userland dev_t; this sucker
breaks as soon as we get sufficiently many autofs mounts for
new_encode_dev(s_dev) != s_dev.

Note: this is the minimal fix.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

reada_for_balance was using the wrong index into the path node array,
so it wasn't reading the right blocks.  We never directly used the
results of the read done by this function because the btree search is
started over at the end.

This fixes reada_for_balance to reada in the correct node and to
avoid searching past the last slot in the node.  It also makes sure to
hold the parent lock while we are finding the nodes to read.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

The extent_io writepage call updates the writepage index in the inode
as it makes progress.  But, it was doing the update after unlocking the page,
which isn't legal because page->mapping can't be trusted once the page
is unlocked.

This lead to an oops, especially common with compression turned on.  The
fix here is to update the writeback index before unlocking the page.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Btrfs is using WRITE_SYNC_PLUG to send down synchronous IOs with a
higher priority.  But, the checksumming helper threads prevent it
from being fully effective.

There are two problems.  First, a big queue of pending checksumming
will delay the synchronous IO behind other lower priority writes.  Second,
the checksumming uses an ordered async work queue.  The ordering makes sure
that IOs are sent to the block layer in the same order they are sent
to the checksumming threads.  Usually this gives us less seeky IO.

But, when we start mixing IO priorities, the lower priority IO can delay
the higher priority IO.

This patch solves both problems by adding a high priority list to the async
helper threads, and a new btrfs_set_work_high_prio(), which is used
to make put a new async work item onto the higher priority list.

The ordering is still done on high priority IO, but all of the high
priority bios are ordered separately from the low priority bios.  This
ordering is purely an IO optimization, it is not involved in data
or metadata integrity.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Part of reducing fsync/O_SYNC/O_DIRECT latencies is using WRITE_SYNC for
writes we plan on waiting on in the near future.  This patch
mirrors recent changes in other filesystems and the generic code to
use WRITE_SYNC when WB_SYNC_ALL is passed and to use WRITE_SYNC for
other latency critical writes.

Btrfs uses async worker threads for checksumming before the write is done,
and then again to actually submit the bios.  The bio submission code just
runs a per-device list of bios that need to be sent down the pipe.

This list is split into low priority and high priority lists so the
WRITE_SYNC IO happens first.
Signed-off-by: NChris Mason <chris.mason@oracle.com>