Reduce some branches and memory accesses in dcache lookup by adding dentry
flags to indicate common d_ops are set, rather than having to check them.
This saves a pointer memory access (dentry->d_op) in common path lookup
situations, and saves another pointer load and branch in cases where we
have d_op but not the particular operation.

Patched with:

git grep -E '[.>]([[:space:]])*d_op([[:space:]])*=' | xargs sed -e 's/\([^\t ]*\)->d_op = \(.*\);/d_set_d_op(\1, \2);/' -e 's/\([^\t ]*\)\.d_op = \(.*\);/d_set_d_op(\&\1, \2);/' -i
Signed-off-by: NNick Piggin <npiggin@kernel.dk>

RCU free the struct inode. This will allow:

- Subsequent store-free path walking patch. The inode must be consulted for
  permissions when walking, so an RCU inode reference is a must.
- sb_inode_list_lock to be moved inside i_lock because sb list walkers who want
  to take i_lock no longer need to take sb_inode_list_lock to walk the list in
  the first place. This will simplify and optimize locking.
- Could remove some nested trylock loops in dcache code
- Could potentially simplify things a bit in VM land. Do not need to take the
  page lock to follow page->mapping.

The downsides of this is the performance cost of using RCU. In a simple
creat/unlink microbenchmark, performance drops by about 10% due to inability to
reuse cache-hot slab objects. As iterations increase and RCU freeing starts
kicking over, this increases to about 20%.

In cases where inode lifetimes are longer (ie. many inodes may be allocated
during the average life span of a single inode), a lot of this cache reuse is
not applicable, so the regression caused by this patch is smaller.

The cache-hot regression could largely be avoided by using SLAB_DESTROY_BY_RCU,
however this adds some complexity to list walking and store-free path walking,
so I prefer to implement this at a later date, if it is shown to be a win in
real situations. I haven't found a regression in any non-micro benchmark so I
doubt it will be a problem.
Signed-off-by: NNick Piggin <npiggin@kernel.dk>

Change d_delete from a dentry deletion notification to a dentry caching
advise, more like ->drop_inode. Require it to be constant and idempotent,
and not take d_lock. This is how all existing filesystems use the callback
anyway.

This makes fine grained dentry locking of dput and dentry lru scanning
much simpler.
Signed-off-by: NNick Piggin <npiggin@kernel.dk>

Buggered-in: 76dda93c ("Btrfs: add snapshot/subvolume destroy
ioctl")
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Acked-by: NChris Mason <chris.mason@oracle.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The extent allocator has code that allows us to fill
allocations from any available block group, even if it doesn't
match the raid level we've requested.

This was put in because adding a new drive to a filesystem
made with the default mkfs options actually upgrades the metadata from
single spindle dup to full RAID1.

But, the code also allows us to allocate from a raid0 chunk when we
really want a raid1 or raid10 chunk.  This can cause big trouble because
mkfs creates a small (4MB) raid0 chunk for data and metadata which then
goes unused for raid1/raid10 installs.

The allocator will happily wander in and allocate from that chunk when
things get tight, which is not correct.

The fix here is to make sure that we provide duplication when the
caller has asked for it.  It does all the dups to be any raid level,
which preserves the dup->raid1 upgrade abilities.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

When we mount in RAID degraded mode without adding a new device to
replace the failed one, we can end up using the wrong RAID flags for
allocations.

This results in strange combinations of block groups (raid1 in a raid10
filesystem) and corruptions when we try to allocate blocks from single
spindle chunks on drives that are actually missing.

The first device has two small 4MB chunks in it that mkfs creates and
these are usually unused in a raid1 or raid10 setup.  But, in -o degraded,
the allocator will fall back to these because the mask of desired raid groups
isn't correct.

The fix here is to count the missing devices as we build up the list
of devices in the system.  This count is used when picking the
raid level to make sure we continue using the same levels that were
in place before we lost a drive.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

If we just get a plain IO error when we read tree roots, the code
wasn't properly sending that error up the chain.  This allowed mounts to
continue when they should failed, and allowed operations
on partially setup root structs.  The end result was usually oopsen
on spinlocks that hadn't been spun up correctly.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

... regarding an unused function when !MIGRATION, and regarding a
printk() format string vs argument mismatch.
Signed-off-by: NJan Beulich <jbeulich@novell.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

If we had reserved some bytes in struct btrfs_ioctl_vol_args, we
wouldn't have to create a new structure for async snapshot creation.

Here we convert async snapshot ioctl to use a more generic ABI, as
we'll add more ioctls for snapshots/subvolumes in the future, readonly
snapshots for example.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

This problem is found in meego testing:
http://bugs.meego.com/show_bug.cgi?id=6672
A file in btrfs is mmaped and the mmaped buffer is passed to pwrite to write to the same page
of the same file. In btrfs_file_aio_write(), the pages is locked by prepare_pages(). So when
btrfs_copy_from_user() is called, page fault happens and the same page needs to be locked again
in filemap_fault(). The fix is to move iov_iter_fault_in_readable() before prepage_pages() to make page
fault happen before pages are locked. And also disable page fault in critical region in
btrfs_copy_from_user().

Reviewed-by: Yan, Zheng<zheng.z.yan@intel.com>
Signed-off-by: NZhong, Xin <xin.zhong@intel.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

We should drop dentry before deactivating the superblock, otherwise
we can hit this bug:

BUG: Dentry f349a690{i=100,n=/} still in use (1) [unmount of btrfs loop1]
...

Steps to reproduce the bug:

  # mount /dev/loop1 /mnt
  # mkdir save
  # btrfs subvolume snapshot /mnt save/snap1
  # umount /mnt
  # mount -o subvol=save/snap1 /dev/loop1 /mnt
  (crash)
Reported-by: NMichael Niederle <mniederle@gmx.at>
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

We were incorrectly taking the async path even for the sync ioctls by
passing in &transid unconditionally.

There's ample room for further cleanup here, but this keeps the fix simple.
Signed-off-by: NSage Weil <sage@newdream.net>
Reviewed-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

"start + num_bytes >= actual_end" can happen when compressed page writeback races
with file truncation. In that case we need unlock and release pages past the end
of file.
Signed-off-by: NYan, Zheng <zheng.z.yan@intel.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Not being able to delete an orphan item isn't a horrible thing.  The worst that
happens is the next time around we try and do the orphan cleanup and we can't
find the referenced object and just delete the item and move on.
Signed-off-by: NJosef Bacik <josef@redhat.com>

If the orphan item doesn't exist, we return 1, which doesn't make any sense to
the callers.  Instead return -ENOENT if we didn't find the item.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>

Since the fast caching uses normal tree locking, we can possibly deadlock if we
get to the caching via a btrfs_search_slot() on the tree_root.  So just check to
see if the root we are on is the tree root, and just don't do the fast caching.
Reported-by: NSage Weil <sage@newdream.net>
Signed-off-by: NJosef Bacik <josef@redhat.com>

Currently if the space cache inode generation number doesn't match the
generation number in the space cache header we will just fail to load the space
cache, but we won't mark the space cache as an error, so we'll keep getting that
error each time somebody tries to cache that block group until we actually clear
the thing.  Fix this by marking the space cache as having an error so we only
get the message once.  This patch also makes it so that we don't try and setup
space cache for a block group that isn't cached, since we won't be able to write
it out anyway.  None of these problems are actual problems, they are just
annoying and sub-optimal.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>

This fixes a bug where we use dip after we have freed it.  Instead just use the
file_offset that was passed to the function.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>

Fixes compile error
Signed-off-by: NChris Mason <chris.mason@oracle.com>

The new DIO bio splitting code has problems when the bio
spans more than one ordered extent.  This will happen as the
generic DIO code merges our get_blocks calls together into
a bigger single bio.

This fixes things by walking forward in the ordered extent
code finding all the overlapping ordered extents and completing them
all at once.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

There is a problem with how we use sget, it searches through the list of supers
attached to the fs_type looking for a super with the same fs_devices as what
we're trying to mount.  This depends on sb->s_fs_info being filled, but we don't
fill that in until we get to btrfs_fill_super, so we could hit supers on the
fs_type super list that have a null s_fs_info.  In order to fix that we need to
go ahead and setup a blank root with a blank fs_info to hold fs_devices, that
way our test will work out right and then we can set s_fs_info in
btrfs_set_super, and then open_ctree will simply use our pre-allocated root and
fs_info when setting everything up.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

There are two big problems currently with FIEMAP

1) We return extents for holes.  This isn't supposed to happen, we just don't
return extents for holes and then userspace interprets the lack of an extent as
a hole.

2) We sometimes don't set FIEMAP_EXTENT_LAST properly.  This is because we wait
to see a EXTENT_FLAG_VACANCY flag on the em, but this won't happen if say we ask
fiemap to map up to the last extent in a file, and there is nothing but holes up
to the i_size.  To fix this we need to lookup the last extent in this file and
save the logical offset, so if we happen to try and map that extent we can be
sure to set FIEMAP_EXTENT_LAST.

With this patch we now pass xfstest 225, which we never have before.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

When mounting a btrfs file system btrfs_test_super() may attempt to
use sb->s_fs_info, the btrfs root, of a super block that is going away
and that has had the btrfs root set to NULL in its ->put_super(). But
if the super block is going away it cannot be an existing super block
so we can return false in this case.
Signed-off-by: NIan Kent <raven@themaw.net>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Currently we fail xfstest 236 because we're not updating the inode ctime on
link.  This is a simple fix, and makes it so we pass 236 now.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

We have been failing xfstest 228 forever, because we don't check to make sure
the new inode size is acceptable as far as RLIMIT is concerned.  Just check to
make sure it's ok to create a inode with this new size and error out if not.
With this patch we now pass 228.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

There is a typo in __btrfs_prealloc_file_range() where we set the i_size to
actual_len/cur_offset, and then just set it to cur_offset again, and do the same
with btrfs_ordered_update_i_size().  This fixes it back to keeping i_size in a
local variable and then updating i_size properly.  Tested this with

xfs_io -F -f -c "falloc 0 1" -c "pwrite 0 1" foo

stat'ing foo gives us a size of 1 instead of 4096 like it was.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

If we fail to find a pointer in the radix tree, don't try
to deref the NULL one we do have.
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Everybody who calls btrfs_add_nondir just passes in the dentry of the new file
and then dereference dentry->d_parent->d_inode, but everybody who calls
btrfs_add_nondir() are already passed the parent's inode.  So instead of
dereferencing dentry->d_parent, just make btrfs_add_nondir take the dir inode as
an argument and pass that along so we don't have to worry about d_parent.
Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Since we walk up the path logging all of the parts of the inode's path, we need
to hold i_mutex to make sure that the inode is not renamed while we're logging
everything.  btrfs_log_dentry_safe does dget_parent and all of that jazz, but we
may get unexpected results if the rename changes the inode's location while
we're higher up the path logging those dentries, so do this for safety reasons.
Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

There are lots of places where we do dentry->d_parent->d_inode without holding
the dentry->d_lock.  This could cause problems with rename.  So instead we need
to use dget_parent() and hold the reference to the parent as long as we are
going to use it's inode and then dput it at the end.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Cc: raven@themaw.net
Signed-off-by: NChris Mason <chris.mason@oracle.com>

When creating new inodes we don't setup inode->i_generation.  So if we generate
an fh with a newly created inode we save the generation of 0, but if we flush
the inode to disk and have to read it back when getting the inode on the server
we'll have the right i_generation, so gens wont match and we get ESTALE.  This
patch properly sets inode->i_generation when we create the new inode and now I'm
no longer getting ESTALE.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

People kept reporting NFS issues, specifically getting ESTALE alot.  I figured
out how to reproduce the problem

SERVER
mkfs.btrfs /dev/sda1
mount /dev/sda1 /mnt/btrfs-test
<add /mnt/btrfs-test to /etc/exports>
btrfs subvol create /mnt/btrfs-test/foo
service nfs start

CLIENT
mount server:/mnt/btrfs /mnt/test
cd /mnt/test/foo
ls

SERVER
echo 3 > /proc/sys/vm/drop_caches

CLIENT
ls			<-- get an ESTALE here

This is because the standard way to lookup a name in nfsd is to use readdir, and
what it does is do a readdir on the parent directory looking for the inode of
the child.  So in this case the parent being / and the child being foo.  Well
subvols all have the same inode number, so doing a readdir of / looking for
inode 256 will return '.', which obviously doesn't match foo.  So instead we
need to have our own .get_name so that we can find the right name.

Our .get_name will either lookup the inode backref or the root backref,
whichever we're looking for, and return the name we find.  Running the above
reproducer with this patch results in everything acting the way its supposed to.
Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Fixes these sparse warnings:
fs/btrfs/ctree.h:811:17: error: dubious one-bit signed bitfield
fs/btrfs/ctree.h:812:20: error: dubious one-bit signed bitfield
fs/btrfs/ctree.h:813:19: error: dubious one-bit signed bitfield
Signed-off-by: NMariusz Kozlowski <mk@lab.zgora.pl>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Symlinks and files of other types show different device numbers, though
they are on the same partition:

 $ touch tmp; ln -s tmp tmp2; stat tmp tmp2
   File: `tmp'
   Size: 0         	Blocks: 0          IO Block: 4096   regular empty file
 Device: 15h/21d	Inode: 984027      Links: 1
 --- snip ---
   File: `tmp2' -> `tmp'
   Size: 3         	Blocks: 0          IO Block: 4096   symbolic link
 Device: 13h/19d	Inode: 984028      Links: 1
Reported-by: NToke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

Set src_offset = 0, src_length = 20K, dest_offset = 20K. And the
original filesize of the dest file 'file2' is 30K:

  # ls -l /mnt/file2
  -rw-r--r-- 1 root root 30720 Nov 18 16:42 /mnt/file2

Now clone file1 to file2, the dest file should be 40K, but it
still shows 30K:

  # ls -l /mnt/file2
  -rw-r--r-- 1 root root 30720 Nov 18 16:42 /mnt/file2
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

We've done the check for src_offset and src_length, and We should
also check dest_offset, otherwise we'll corrupt the destination
file:

  (After cloning file1 to file2 with unaligned dest_offset)
  # cat /mnt/file2
  cat: /mnt/file2: Input/output error
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

When I added the clear_cache option I screwed up and took the break out of
the space_cache case statement, so whenever you mount with space_cache you also
get clear_cache, which does you no good if you say set space_cache in fstab so
it always gets set.  This patch adds the break back in properly.
Signed-off-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

'unused' calculated with wrong sign in reserve_metadata_bytes().
This might have lead to unwanted over-reservations.
Signed-off-by: NArne Jansen <sensille@gmx.net>
Reviewed-by: NJosef Bacik <josef@redhat.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

btrfs paniced when we write >64KB data by direct IO at one time.

Reproduce steps:
 # mkfs.btrfs /dev/sda5 /dev/sda6
 # mount /dev/sda5 /mnt
 # dd if=/dev/zero of=/mnt/tmpfile bs=100K count=1 oflag=direct

Then btrfs paniced:
mapping failed logical 1103155200 bio len 69632 len 12288
------------[ cut here ]------------
kernel BUG at fs/btrfs/volumes.c:3010!
[SNIP]
Pid: 1992, comm: btrfs-worker-0 Not tainted 2.6.37-rc1 #1 D2399/PRIMERGY
RIP: 0010:[<ffffffffa03d1462>]  [<ffffffffa03d1462>] btrfs_map_bio+0x202/0x210 [btrfs]
[SNIP]
Call Trace:
 [<ffffffffa03ab3eb>] __btrfs_submit_bio_done+0x1b/0x20 [btrfs]
 [<ffffffffa03a35ff>] run_one_async_done+0x9f/0xb0 [btrfs]
 [<ffffffffa03d3d20>] run_ordered_completions+0x80/0xc0 [btrfs]
 [<ffffffffa03d45a4>] worker_loop+0x154/0x5f0 [btrfs]
 [<ffffffffa03d4450>] ? worker_loop+0x0/0x5f0 [btrfs]
 [<ffffffffa03d4450>] ? worker_loop+0x0/0x5f0 [btrfs]
 [<ffffffff81083216>] kthread+0x96/0xa0
 [<ffffffff8100cec4>] kernel_thread_helper+0x4/0x10
 [<ffffffff81083180>] ? kthread+0x0/0xa0
 [<ffffffff8100cec0>] ? kernel_thread_helper+0x0/0x10

We fix this problem by splitting bios when we submit bios.
Reported-by: NTsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: NMiao Xie <miaox@cn.fujitsu.com>
Tested-by: NTsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>

extent_bio_alloc() and compressed_bio_alloc() are similar, cleanup
similar source code.
Signed-off-by: NMiao Xie <miaox@cn.fujitsu.com>
Signed-off-by: NChris Mason <chris.mason@oracle.com>