The cache alias problem will happen if the changes of user shared mapping
is not flushed before copying, then user and kernel mapping may be mapped
into two different cache line, it is impossible to guarantee the coherence
after iov_iter_copy_from_user_atomic.  So the right steps should be:

	flush_dcache_page(page);
	kmap_atomic(page);
	write to page;
	kunmap_atomic(page);
	flush_dcache_page(page);

More precisely, we might create two new APIs flush_dcache_user_page and
flush_dcache_kern_page to replace the two flush_dcache_page accordingly.

Here is a snippet tested on omap2430 with VIPT cache, and I think it is
not ARM-specific:

	int val = 0x11111111;
	fd = open("abc", O_RDWR);
	addr = mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);
	*(addr+0) = 0x44444444;
	tmp = *(addr+0);
	*(addr+1) = 0x77777777;
	write(fd, &val, sizeof(int));
	close(fd);

The results are not always 0x11111111 0x77777777 at the beginning as expected.  Sometimes we see 0x44444444 0x77777777.
Signed-off-by: NAnfei <anfei.zhou@gmail.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: <linux-arch@vger.kernel.org>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Commit 221af7f8 ("Split 'flush_old_exec' into two functions") split
the function at the point of no return - ie right where there were no
more error cases to check.  That made sense from a technical standpoint,
but when we then also combined it with the actual personality setting
going in between flush_old_exec() and setup_new_exec(), it needs to be a
bit more careful.

In particular, we need to make sure that we really flush the old
personality bits in the 'flush' stage, rather than later in the 'setup'
stage, since otherwise we might be flushing the _new_ personality state
that we're just setting up.

So this moves the flags and personality flushing (and 'flush_thread()',
which is the arch-specific function that generally resets lazy FP state
etc) of the old process into flush_old_exec(), so that it doesn't affect
any state that execve() is setting up for the new process environment.

This was reported by Michal Simek as breaking his Microblaze qemu
environment.
Reported-and-tested-by: NMichal Simek <michal.simek@petalogix.com>
Cc: Peter Anvin <hpa@zytor.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This is called under a glock, so its a good plan to use GFP_NOFS
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

The do_div() call needs to remain.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

ince gfs2 writes the rindex file a block at a time, and releases the
exclusive lock after each block, it is possible that another process
will grab the lock in the middle of the write.  Since rindex entries are
not an even divisor of blocks, that other process may see partial
entries.  On grows, this is fine.  The process can simply ignore the the
partial entires. Previously, the code withdrew when it saw partial
entries. Now it simply ignores them.
Signed-off-by: NBenjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

This fixes incorrect usage of nilfs_segctor_confirm() test function in
nilfs_segctor_destroy(); nilfs_segctor_confirm() returns zero if the
filesystem is not clean, so its use in nilfs_segctor_destroy() needs
inversion.
Signed-off-by: NRyusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>

Fix two bugs in the bio integrity code:

 use_bip_pool() always returns 0 because it checks against the wrong limit,
 causing the mempool to be used only when regular allocation fails.

 When the mempool is used as a fallback we don't free the data properly.
Signed-Off-By: NChuck Ebbert <cebbert@redhat.com>
Acked-by: NMartin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

'flush_old_exec()' is the point of no return when doing an execve(), and
it is pretty badly misnamed.  It doesn't just flush the old executable
environment, it also starts up the new one.

Which is very inconvenient for things like setting up the new
personality, because we want the new personality to affect the starting
of the new environment, but at the same time we do _not_ want the new
personality to take effect if flushing the old one fails.

As a result, the x86-64 '32-bit' personality is actually done using this
insane "I'm going to change the ABI, but I haven't done it yet" bit
(TIF_ABI_PENDING), with SET_PERSONALITY() not actually setting the
personality, but just the "pending" bit, so that "flush_thread()" can do
the actual personality magic.

This patch in no way changes any of that insanity, but it does split the
'flush_old_exec()' function up into a preparatory part that can fail
(still called flush_old_exec()), and a new part that will actually set
up the new exec environment (setup_new_exec()).  All callers are changed
to trivially comply with the new world order.
Signed-off-by: NH. Peter Anvin <hpa@zytor.com>
Cc: stable@kernel.org
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

If you have a disk failure in RAID1 and then add a new disk to the
array, and then try to remove the missing volume, it will fail.  The
reason is the sanity check only looks at the total number of rw devices,
which is just 2 because we have 2 good disks and 1 bad one.  Instead
check the total number of devices in the array to make sure we can
actually remove the device.  Tested this with a failed disk setup and
with this test we can now run

btrfs-vol -r missing /mount/point

and it works fine.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Hit this problem while testing RAID1 failure stuff.  open_bdev_exclusive
returns ERR_PTR(), not NULL.  So change the return value properly.  This
is important if you accidently specify a device that doesn't exist when
trying to add a new device to an array, you will panic the box
dereferencing bdev.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

If a RAID setup has chunks that span multiple disks, and one of those
disks has failed, btrfs_chunk_readonly will return 1 since one of the
disks in that chunk's stripes is dead and therefore not writeable.  So
instead if we are in degraded mode, return 0 so we can go ahead and
allocate stuff.  Without this patch all of the block groups in a RAID1
setup will end up read-only, which will mean we can't add new disks to
the array since we won't be able to make allocations.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

This patch revert's commit

6c090a11

Since it introduces this problem where we can run orphan cleanup on a
volume that can have orphan entries re-added.  Instead of my original
fix, Yan Zheng pointed out that we can just revert my original fix and
then run the orphan cleanup in open_ctree after we look up the fs_root.
I have tested this with all the tests that gave me problems and this
patch fixes both problems.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

In btrfs_init_acl() cloned acl is not released
Signed-off-by: NYang Hongyang <yanghy@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

commit f2bc9dd07e3424c4ec5f3949961fe053d47bc825
Author: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Date:   Wed Jan 20 12:57:53 2010 +0530

    Btrfs: Use correct values when updating inode i_size on fallocate

    Even though we allocate more, we should be updating inode i_size
    as per the arguments passed
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

This patch removes tree_search() in extent_map.c because it is not called by
anything.
Signed-off-by: NMiao Xie <miaox@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

The default btrfs mount -o compress mode will quickly back off
compressing a file if it notices that compression does not reduce the
size of the data being written.  This can save considerable CPU because
all future writes to the file go through uncompressed.

But some files are both very large and have mixed data stored in
them.  In that case, we want to add the ability to always try
compressing data before writing it.

This commit adds mount -o compress-force.  A later commit will add
a new inode flag that does the same thing.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

We have to properly decrease bi_size in order to merge_bvec_fn return
right result.  Otherwise this result in false merge rejects for two
absolutely valid bio_vecs.  This may cause significant performance
penalty for example fs_block_size == 1k and block device is raid0 with
small chunk_size = 8k. Then it is impossible to merge 7-th fs-block in
to bio which already has 6 fs-blocks.

Cc: <stable@kernel.org>
Signed-off-by: NDmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

Vmalloc is called to allocate journal->j_cnode_free_list but
we hold the reiserfs lock at this time, which raises a
{RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} lock inversion.

Just drop the reiserfs lock at this time, as it's not even
needed but kept for paranoid reasons.

This fixes:

[ INFO: inconsistent lock state ]
2.6.33-rc5 #1
---------------------------------
inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage.
kswapd0/313 [HC0[0]:SC0[0]:HE1:SE1] takes:
 (&REISERFS_SB(s)->lock){+.+.?.}, at: [<c11118c8>]
reiserfs_write_lock_once+0x28/0x50
{RECLAIM_FS-ON-W} state was registered at:
  [<c104ee32>] mark_held_locks+0x62/0x90
  [<c104eefa>] lockdep_trace_alloc+0x9a/0xc0
  [<c108f7b6>] kmem_cache_alloc+0x26/0xf0
  [<c108621c>] __get_vm_area_node+0x6c/0xf0
  [<c108690e>] __vmalloc_node+0x7e/0xa0
  [<c1086aab>] vmalloc+0x2b/0x30
  [<c110e1fb>] journal_init+0x6cb/0xa10
  [<c10f90a2>] reiserfs_fill_super+0x342/0xb80
  [<c1095665>] get_sb_bdev+0x145/0x180
  [<c10f68e1>] get_super_block+0x21/0x30
  [<c1094520>] vfs_kern_mount+0x40/0xd0
  [<c1094609>] do_kern_mount+0x39/0xd0
  [<c10aaa97>] do_mount+0x2c7/0x6d0
  [<c10aaf06>] sys_mount+0x66/0xa0
  [<c16198a7>] mount_block_root+0xc4/0x245
  [<c1619a81>] mount_root+0x59/0x5f
  [<c1619b98>] prepare_namespace+0x111/0x14b
  [<c1619269>] kernel_init+0xcf/0xdb
  [<c100303a>] kernel_thread_helper+0x6/0x1c
irq event stamp: 63236801
hardirqs last  enabled at (63236801): [<c134e7fa>]
__mutex_unlock_slowpath+0x9a/0x120
hardirqs last disabled at (63236800): [<c134e799>]
__mutex_unlock_slowpath+0x39/0x120
softirqs last  enabled at (63218800): [<c102f451>] __do_softirq+0xc1/0x110
softirqs last disabled at (63218789): [<c102f4ed>] do_softirq+0x4d/0x60

other info that might help us debug this:
2 locks held by kswapd0/313:
 #0:  (shrinker_rwsem){++++..}, at: [<c1074bb4>] shrink_slab+0x24/0x170
 #1:  (&type->s_umount_key#19){++++..}, at: [<c10a2edd>]
shrink_dcache_memory+0xfd/0x1a0

stack backtrace:
Pid: 313, comm: kswapd0 Not tainted 2.6.33-rc5 #1
Call Trace:
 [<c134db2c>] ? printk+0x18/0x1c
 [<c104e7ef>] print_usage_bug+0x15f/0x1a0
 [<c104ebcf>] mark_lock+0x39f/0x5a0
 [<c104d66b>] ? trace_hardirqs_off+0xb/0x10
 [<c1052c50>] ? check_usage_forwards+0x0/0xf0
 [<c1050c24>] __lock_acquire+0x214/0xa70
 [<c10438c5>] ? sched_clock_cpu+0x95/0x110
 [<c10514fa>] lock_acquire+0x7a/0xa0
 [<c11118c8>] ? reiserfs_write_lock_once+0x28/0x50
 [<c134f03f>] mutex_lock_nested+0x5f/0x2b0
 [<c11118c8>] ? reiserfs_write_lock_once+0x28/0x50
 [<c11118c8>] ? reiserfs_write_lock_once+0x28/0x50
 [<c11118c8>] reiserfs_write_lock_once+0x28/0x50
 [<c10f05b0>] reiserfs_delete_inode+0x50/0x140
 [<c10a653f>] ? generic_delete_inode+0x5f/0x150
 [<c10f0560>] ? reiserfs_delete_inode+0x0/0x140
 [<c10a657c>] generic_delete_inode+0x9c/0x150
 [<c10a666d>] generic_drop_inode+0x3d/0x60
 [<c10a5597>] iput+0x47/0x50
 [<c10a2a4f>] dentry_iput+0x6f/0xf0
 [<c10a2af4>] d_kill+0x24/0x50
 [<c10a2d3d>] __shrink_dcache_sb+0x21d/0x2b0
 [<c10a2f0f>] shrink_dcache_memory+0x12f/0x1a0
 [<c1074c9e>] shrink_slab+0x10e/0x170
 [<c1075177>] kswapd+0x477/0x6a0
 [<c1072d10>] ? isolate_pages_global+0x0/0x1b0
 [<c103e160>] ? autoremove_wake_function+0x0/0x40
 [<c1074d00>] ? kswapd+0x0/0x6a0
 [<c103de6c>] kthread+0x6c/0x80
 [<c103de00>] ? kthread+0x0/0x80
 [<c100303a>] kernel_thread_helper+0x6/0x1c
Reported-by: NAlexander Beregalov <a.beregalov@gmail.com>
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: Christian Kujau <lists@nerdbynature.de>
Cc: Chris Mason <chris.mason@oracle.com>

if 9P ->get_sb() fails late (at root inode or root dentry
allocation), we'll hit its ->kill_sb() with NULL ->s_root
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

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

if we'd just got success from it, vfsmount won't be NULL
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

double iput(), leaks...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Error handling in that sucker got broken back in 2003.  If function
returns 0 on failure, it's not nice to add return -EINVAL into it.
Adding return 1 on other failure exits is also not a good thing (and
yes, original success exits with 1 and some of failure exits with 0
are still there; so's the original logics in callers).
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

A couple of fields in affs_sb_info is used in follow_link() and
symlink() for handling AFFS "absolute" symlinks.  Need locking
against affs_remount() updates.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

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

Commit 70362511 exposed that f_modown()
should call write_lock_irqsave instead of just write_lock_irq so that
because a caller could have a spinlock held and it would not be good to
renable interrupts.

Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Tavis Ormandy <taviso@google.com>
Cc: stable <stable@kernel.org>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

KVM needs a wait to atomically remove themselves from the eventfd ->poll()
wait queue head, in order to handle correctly their IRQfd deassign
operation.

This patch introduces such API, plus a way to read an eventfd from its
context.
Signed-off-by: NDavide Libenzi <davidel@xmailserver.org>
Signed-off-by: NAvi Kivity <avi@redhat.com>

After the commit fb07a5f8 ("compat_ioctl: remove all VT ioctl
handling"), I got this error message on 64-bit mips kernel with 32-bit
busybox userland:

ioctl32(init:1): Unknown cmd fd(0) cmd(00005600){t:'V';sz:0} arg(7fd76480) on /dev/console

The cmd 5600 is VT_OPENQRY.  The busybox's init issues this ioctl to
know vt-console or serial-console.  If the console was serial console,
VT ioctls are not handled by the serial driver.

And by quick search, I found some programs using VT_GETMODE to check
vt-console is available or not.
Signed-off-by: NAtsushi Nemoto <anemo@mba.ocn.ne.jp>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The "full_alg_name" variable is used on a couple error paths, so we
shouldn't free it until the end.
Signed-off-by: NDan Carpenter <error27@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

The variable lower_dentry is initialized twice to the same (side effect-free)
expression.  Drop one initialization.

A simplified version of the semantic match that finds this problem is:
(http://coccinelle.lip6.fr/)

// <smpl>
@forall@
idexpression *x;
identifier f!=ERR_PTR;
@@

x = f(...)
... when != x
(
x = f(...,<+...x...+>,...)
|
* x = f(...)
)
// </smpl>
Signed-off-by: NJulia Lawall <julia@diku.dk>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Signed-off-by: NErez Zadok <ezk@cs.sunysb.edu>
Acked-by: NDustin Kirkland <kirkland@canonical.com>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

ecryptfs_interpose checks if one of the flags passed is
ECRYPTFS_INTERPOSE_FLAG_D_ADD, defined as 0x00000001 in ecryptfs_kernel.h.
But the only user of ecryptfs_interpose to pass a non-zero flag to it, has
hard-coded the value as "1". This could spell trouble if any of these values
changes in the future.
Signed-off-by: NErez Zadok <ezk@cs.sunysb.edu>
Cc: Dustin Kirkland <kirkland@canonical.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Unnecessary because it would unhash perfectly valid dentries, causing them
to have to be re-looked up the next time they're needed, which presumably is
right after.
Signed-off-by: NAseem Rastogi <arastogi@cs.sunysb.edu>
Signed-off-by: NShrikar archak <shrikar84@gmail.com>
Signed-off-by: NErez Zadok <ezk@cs.sunysb.edu>
Cc: Saumitra Bhanage <sbhanage@cs.sunysb.edu>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Signed-off-by: NErez Zadok <ezk@cs.sunysb.edu>
Cc: Dustin Kirkland <kirkland@canonical.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Ecryptfs_open dereferences a pointer to the private lower file (the one
stored in the ecryptfs inode), without checking if the pointer is NULL.
Right afterward, it initializes that pointer if it is NULL.  Swap order of
statements to first initialize.  Bug discovered by Duckjin Kang.
Signed-off-by: NDuckjin Kang <fromdj2k@gmail.com>
Signed-off-by: NErez Zadok <ezk@cs.sunysb.edu>
Cc: Dustin Kirkland <kirkland@canonical.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: <stable@kernel.org>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

Adrian reported that mkfontscale didn't work inside of eCryptfs mounts.
Strace revealed the following:

open("./", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY|O_CLOEXEC) = 3
fcntl64(3, F_GETFD) = 0x1 (flags FD_CLOEXEC)
open("./fonts.scale", O_WRONLY|O_CREAT|O_TRUNC, 0666) = 4
getdents(3, /* 80 entries */, 32768) = 2304
open("./.", O_RDONLY) = 5
fcntl64(5, F_SETFD, FD_CLOEXEC) = 0
fstat64(5, {st_mode=S_IFDIR|0755, st_size=16384, ...}) = 0
mmap2(NULL, 16384, PROT_READ, MAP_PRIVATE, 5, 0) = 0xb7fcf000
close(5) = 0
--- SIGBUS (Bus error) @ 0 (0) ---
+++ killed by SIGBUS +++

The mmap2() on a directory was successful, resulting in a SIGBUS
signal later.  This patch removes mmap() from the list of possible
ecryptfs_dir_fops so that mmap() isn't possible on eCryptfs directory
files.

https://bugs.launchpad.net/ecryptfs/+bug/400443Reported-by: NAdrian C. <anrxc@sysphere.org>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

The i_blocks field of an eCryptfs inode cannot be trusted, but
generic_fillattr() uses it to instantiate the blocks field of a stat()
syscall when a filesystem doesn't implement its own getattr().  Users
have noticed that the output of du is incorrect on newly created files.

This patch creates ecryptfs_getattr() which calls into the lower
filesystem's getattr() so that eCryptfs can use its kstat.blocks value
after calling generic_fillattr().  It is important to note that the
block count includes the eCryptfs metadata stored in the beginning of
the lower file plus any padding used to fill an extent before
encryption.

https://bugs.launchpad.net/ecryptfs/+bug/390833Reported-by: NDominic Sacré <dominic.sacre@gmx.de>
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

When truncating inodes in the lower filesystem, eCryptfs directly
invoked vmtruncate(). As Christoph Hellwig pointed out, vmtruncate() is
a filesystem helper function, but filesystems may need to do more than
just a call to vmtruncate().

This patch moves the lower inode truncation out of ecryptfs_truncate()
and renames the function to truncate_upper().  truncate_upper() updates
an iattr for the lower inode to indicate if the lower inode needs to be
truncated upon return.  ecryptfs_setattr() then calls notify_change(),
using the updated iattr for the lower inode, to complete the truncation.

For eCryptfs functions needing to truncate, ecryptfs_truncate() is
reintroduced as a simple way to truncate the upper inode to a specified
size and then truncate the lower inode accordingly.

https://bugs.launchpad.net/bugs/451368Reported-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NDustin Kirkland <kirkland@canonical.com>
Cc: ecryptfs-devel@lists.launchpad.net
Cc: linux-fsdevel@vger.kernel.org
Signed-off-by: NTyler Hicks <tyhicks@linux.vnet.ibm.com>

fs/bio.c:81:33: warning: symbol 'bslab' shadows an earlier one
fs/bio.c:74:25: originally declared here
Signed-off-by: NThiago Farina <tfransosi@gmail.com>
Signed-off-by: NJens Axboe <jens.axboe@oracle.com>

We can race with the unmount of an fs and the stopping of a kthread where we
will free the block group before we're done using it.  The reason for this is
because we do not hold a reference on the block group while its caching, since
the allocator drops its reference once it exits or moves on to the next block
group.  This patch fixes the problem by taking a reference to the block group
before we start caching and dropping it when we're done to make sure all
accesses to the block group are safe.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>