When we did stress test for the space relocation, the deadlock happened.
By debugging, We found it was caused by the carelessness that we forgot
to unlock the read lock of the extent buffers in btrfs_orphan_cleanup()
before we end the transaction handle, so the transaction commit task waited
the task, which called btrfs_orphan_cleanup(), to unlock the extent buffer,
but that task waited the commit task to end the transaction commit, and
the deadlock happened. Fix it.
Signed-ff-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

I-node cache forgets to reserve the space when writing out it. And when
we do some stress test, such as synctest, it will trigger WARN_ON() in
use_block_rsv().

WARNING: at fs/btrfs/extent-tree.c:5718 btrfs_alloc_free_block+0xbf/0x281 [btrfs]()
...
Call Trace:
 [<ffffffff8104df86>] warn_slowpath_common+0x80/0x98
 [<ffffffff8104dfb3>] warn_slowpath_null+0x15/0x17
 [<ffffffffa0369c60>] btrfs_alloc_free_block+0xbf/0x281 [btrfs]
 [<ffffffff810cbcb8>] ? __set_page_dirty_nobuffers+0xfe/0x108
 [<ffffffffa035c040>] __btrfs_cow_block+0x118/0x3b5 [btrfs]
 [<ffffffffa035c7ba>] btrfs_cow_block+0x103/0x14e [btrfs]
 [<ffffffffa035e4c4>] btrfs_search_slot+0x249/0x6a4 [btrfs]
 [<ffffffffa036d086>] btrfs_lookup_inode+0x2a/0x8a [btrfs]
 [<ffffffffa03788b7>] btrfs_update_inode+0xaa/0x141 [btrfs]
 [<ffffffffa036d7ec>] btrfs_save_ino_cache+0xea/0x202 [btrfs]
 [<ffffffffa03a761e>] ? btrfs_update_reloc_root+0x17e/0x197 [btrfs]
 [<ffffffffa0373867>] commit_fs_roots+0xaa/0x158 [btrfs]
 [<ffffffffa03746a6>] btrfs_commit_transaction+0x405/0x731 [btrfs]
 [<ffffffff810690df>] ? wake_up_bit+0x25/0x25
 [<ffffffffa039d652>] ? btrfs_log_dentry_safe+0x43/0x51 [btrfs]
 [<ffffffffa0381c5f>] btrfs_sync_file+0x16a/0x198 [btrfs]
 [<ffffffff81122806>] ? mntput+0x21/0x23
 [<ffffffff8112d150>] vfs_fsync_range+0x18/0x21
 [<ffffffff8112d170>] vfs_fsync+0x17/0x19
 [<ffffffff8112d316>] do_fsync+0x29/0x3e
 [<ffffffff8112d348>] sys_fsync+0xb/0xf
 [<ffffffff81468352>] system_call_fastpath+0x16/0x1b

Sometimes it causes BUG_ON() in the reservation code of the delayed inode
is triggered.

So we must reserve enough space for inode cache.

Note: If we can not reserve the enough space for inode cache, we will
give up writing out it.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

btrfs_previous_item() just search the b+ tree, do not COW the nodes or leaves,
if we modify the result of it, the meta-data will be broken. fix it.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Josef sent along an incremental to the inode reservation
code to make sure we try and fall back to directly updating
the inode item if things go horribly wrong.

This reworks that patch slightly, adding a fallback function
that will always try to update the inode item directly without
going through the delayed_inode code.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Commits 6c41761f and 45ea6095 introduced the possibility of NULL pointer
dereference on error paths, also we would leave all devices busy and
leak fs_info with all sub-structures on error when trying to mount an
already mounted fs to a different directory.

Fix this by doing all allocations before trying to open any of the
devices, adjust error path for mount-already-mounted-fs case.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Fix a bug introduced by 7e662854 where we would leave devices busy on
certain error paths in open_ctree().  fs_info is guaranteed to be
non-NULL now so it's safe to dereference it on all error paths.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Fix bugs introduced by 6c41761f.  Firstly, after failing to allocate any
of the tree roots (first 'goto fail' in open_ctree()) we would
dereference a NULL fs_info pointer in free_fs_info().  Secondly, after
failures from init_srcu_struct(), setup_bdi() and new_inode() we would
leak all earlier allocated roots: fs_info fields haven't been
initialized yet so free_fs_info() is rendered useless.

Fix this by initializing fs_info pointer and fs_info fields before any
allocations happen.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

btrfs_parse_early_options() can fail due to error while scanning devices
(-o device= option), but still strdup() subvol_name string:

mount -o subvol=SUBV,device=BAD_DEVICE <dev> <mnt>

So free subvol_name string on error.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

Don't leak subvol_name string in case multiple subvol= options are
given.  "The lastest option is effective" behavior (consistent with
subvolid= and subvolrootid= options) is preserved.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

People have been reporting ENOSPC crashes in finish_ordered_io.  This is because
we try to steal from the delalloc block rsv to satisfy a reservation to update
the inode.  The problem with this is we don't explicitly save space for updating
the inode when doing delalloc.  This is kind of a problem and we've gotten away
with this because way back when we just stole from the delalloc reserve without
any questions, and this worked out fine because generally speaking the leaf had
been modified either by the mtime update when we did the original write or
because we just updated the leaf when we inserted the file extent item, only on
rare occasions had the leaf not actually been modified, and that was still ok
because we'd just use a block or two out of the over-reservation that is
delalloc.

Then came the delayed inode stuff.  This is amazing, except it wants a full
reservation for updating the inode since it may do it at some point down the
road after we've written the blocks and we have to recow everything again.  This
worked out because the delayed inode stuff just stole from the global reserve,
that is until recently when I changed that because it caused other problems.

So here we are, we're doing everything right and being screwed for it.  So take
an extra reservation for the inode at delalloc reservation time and carry it
through the life of the delalloc reservation.  If we need it we can steal it in
the delayed inode stuff.  If we have already stolen it try and do a normal
metadata reservation.  If that fails try to steal from the delalloc reservation.
If _that_ fails we'll get a WARN_ON() so I can start thinking of a better way to
solve this and in the meantime we'll steal from the global reserve.

With this patch I ran xfstests 13 in a loop for a couple of hours and didn't see
any problems.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

If we fail to reserve space in the transaction during truncate, we can
error out with a NULL trans handle.  The cleanup code needs an extra
check to make sure we aren't trying to use the bad handle.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

On error path 'tree_root' is treed in 'free_fs_info()'.
No need to free it explicitely. Noticed by SLUB in debug mode:

Complete reproducer under usermode linux (discovered on real
machine):

    bdev=/dev/ubda
    btr_root=/btr
    /mkfs.btrfs $bdev
    mount $bdev $btr_root
    mkdir $btr_root/subvols/
    cd $btr_root/subvols/
    /btrfs su cr foo
    /btrfs su cr bar
    mount $bdev -osubvol=subvols/foo $btr_root/subvols/bar
    umount $btr_root/subvols/bar

which gives

device fsid 4d55aa28-45b1-474b-b4ec-da912322195e devid 1 transid 7 /dev/ubda
=============================================================================
BUG kmalloc-2048: Object already free
-----------------------------------------------------------------------------

INFO: Allocated in btrfs_mount+0x389/0x7f0 age=0 cpu=0 pid=277
INFO: Freed in btrfs_mount+0x51c/0x7f0 age=0 cpu=0 pid=277
INFO: Slab 0x0000000062886200 objects=15 used=9 fp=0x0000000070b4d2d0 flags=0x4081
INFO: Object 0x0000000070b4d2d0 @offset=21200 fp=0x0000000070b4a968
...
Call Trace:
70b31948:  [<6008c522>] print_trailer+0xe2/0x130
70b31978:  [<6008c5aa>] object_err+0x3a/0x50
70b319a8:  [<6008e242>] free_debug_processing+0x142/0x2a0
70b319e0:  [<600ebf6f>] btrfs_mount+0x55f/0x7f0
70b319f8:  [<6008e5c1>] __slab_free+0x221/0x2d0
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
Cc: Arne Jansen <sensille@gmx.net>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

During log replay, can commit the transaction before the fs_root
pointers are setup, so we have to make sure they are not null before
trying to use them.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

While we're allocating ram for a new transaction, we drop our spinlock.
When we get the lock back, we do check to see if a transaction started
while we slept, but we don't check to make sure it isn't blocked
because a commit has already started.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

In case we were able to map less than we wanted (length < PAGE_SIZE
clause is true) btrfs_bio is still allocated and we have to free it.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Signed-off-by: Chris Mason <chris.mason@oracle.com>

Signed-off-by: Chris Mason <chris.mason@oracle.com>

If we don't stop them, they linger around corrupting
memory by using pointers to freed things.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

The scrub readahead branch brought in a new error handling hook,
but it was leaking extent_buffer references.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

The new ioctls to follow backrefs are not clean for 32/64 bit
compat.  This reworks them for u64s everywhere.  They are brand new, so
there are no problems with changing the interface now.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Conflicts:
	fs/btrfs/Makefile
	fs/btrfs/extent_io.c
	fs/btrfs/extent_io.h
	fs/btrfs/scrub.c
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Conflicts:
	fs/btrfs/ctree.h
Signed-off-by: Chris Mason <chris.mason@oracle.com>

We all keep getting those stupid warnings from use_block_rsv when running
stress.sh, and it's because the delayed insertion stuff is being stupid.  It's
not the delayed insertion stuffs fault, it's all just stupid.  When marking an
inode dirty for oh say updating the time on it, we just do a
btrfs_join_transaction, which doesn't reserve any space.  This is stupid because
we're going to have to have space reserve to make this change, but we do it
because it's fast because chances are we're going to call it over and over again
and it doesn't matter.  Well thanks to the delayed insertion stuff this is
mostly the case, so we do actually need to make this reservation.  So if
trans->bytes_reserved is 0 then try to do a normal reservation.  If not return
ENOSPC which will make the btrfs_dirty_inode start a proper transaction which
will let it do the whole ENOSPC dance and reserve enough space for the delayed
insertion to steal the reservation from the transaction.

The other stupid thing we do is not reserve space for the inode when writing to
the thing.  Usually this is ok since we have to update the time so we'd have
already done all this work before we get to the endio stuff, so it doesn't
matter.  But this is stupid because we could write the data after the
transaction commits where we changed the mtime of the inode so we have to cow
all the way down to the inode anyway.  This used to be masked by the delalloc
reservation stuff, but because we delay the update it doesn't get masked in this
case.  So again the delayed insertion stuff bites us in the ass.  So if our
trans->block_rsv is delalloc, just steal the reservation from the delalloc
reserve.  Hopefully this won't bite us in the ass, but I've said that before.

With this patch stress.sh no longer spits out those stupid warnings (famous last
words).  Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Failure testing was tripping up over stale PageError bits in
metadata pages.  If we have an io error on a block, and later on
end up reusing it, nobody ever clears PageError on those pages.

During commit, we'll find PageError and think we had trouble writing
the block, which will lead to aborts and other problems.

This changes clean_tree_block and the btrfs writepage code to
clear the PageError bit.  In both cases we're either completely
done with the page or the page has good stuff and the error bit
is no longer valid.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

Because of the overcommit stuff I had to make it so that we committed the
transaction all the time in reserve_metadata_bytes in case we had overcommitted
because of delayed items.  This was because previously we had no way of knowing
how much space was reserved for delayed items.  Now that we have the
delayed_block_rsv we can check it to see if committing the transaction would get
us anywhere.  This patch breaks out the committing logic into a helper function
that will check to see if committing the transaction would free enough space for
us to get anything done.  With this patch xfstests 83 goes from taking 445
seconds to taking 28 seconds on my box.  Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

I've been hitting warnings in use_block_rsv when running the delayed insertion
stuff.  It's because we will readjust global block rsv based on what is in use,
which means we could end up discarding reservations that are for the delayed
insertion stuff.  So instead create a seperate block rsv for the delayed
insertion stuff.  This will also make it easier to debug problems with the
delayed insertion reservations since we will know that only the delayed
insertion code touches this block_rsv.  Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

This takes some of the free space in the btrfs super block
to record information about most of the roots in the last four
commits.

It also adds a -o recovery to use the root history log when
we're not able to read the tree of tree roots, the extent
tree root, the device tree root or the csum root.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

fs_info has now ~9kb, more than fits into one page. This will cause
mount failure when memory is too fragmented. Top space consumers are
super block structures super_copy and super_for_commit, ~2.8kb each.
Allocate them dynamically. fs_info will be ~3.5kb. (measured on x86_64)

Add a wrapper for freeing fs_info and all of it's dynamically allocated
members.
Signed-off-by: David Sterba <dsterba@suse.cz>

We no longer use the orphan block rsv for holding the reservation for truncating
the inode, so instead use the global block rsv and check to make sure it has
enough space for us to truncate the space.  Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

I fixed a problem where we weren't reserving space for an orphan item when we
had to fallback to using the global reserve for an unlink, but I introduced
another problem.  I was migrating the bytes from the transaction reserve to the
global reserve and then releasing from the global reserve in
btrfs_end_transaction().  The problem with this is that a migrate will jack up
the size for the destination, but leave the size alone for the source, with the
idea that you can do a release normally on the source and it all washes out, and
then you can do a release again on the destination and it works out right.  My
way was skipping the release on the trans_block_rsv which still had the jacked
up size from our original reservation.  So instead release manually from the
global reserve if this transaction was using it, and then set the
trans->block_rsv back to the trans_block_rsv so that btrfs_end_transaction
cleans everything up properly.  With this patch xfstest 83 doesn't emit warnings
about leaking space.  Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

write_cache_pages tries to build up a large bio to stuff down the pipe.
But if it needs to wait for a page lock, it needs to make sure and send
down any pending writes so we don't deadlock with anyone who has the
page lock and is waiting for writeback of things inside the bio.

Dave Sterba triggered this as a deadlock between the autodefrag code and
the extent write_cache_pages
Signed-off-by: Chris Mason <chris.mason@oracle.com>

The tree log had two important bugs that could cause corruptions after a
crash.  Sometimes we were allowing tree log blocks to be reused after
the tree log was committed but before the transaction commit was done.

This allowed a future metadata write to overwrite the tree log data.  It
is fixed by adding a new variant of freeing reserved extents that always
pins them.  Credit goes to Stefan Behrens and Arne Jansen for many many
hours spent tracking this bug down.

During tree log replay, we do a pass through the tree log and pin all
the extents we find.  This makes sure the replay code won't go in and
use any of those blocks for new allocations during replay.  The problem
is the free space cache isn't honoring these pinned extents.  So the
allocator can end up handing them out, leading to all kinds of problems
during replay.

The fix here is to force any free space cache to load while we pin the
extents, and then to make sure we remove the pinned extents from the
free space rbtree.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Reported-by: Stefan Behrens <sbehrens@giantdisaster.de>

btrfs_remove_free_space needs to make sure to set ret back to a
valid return value after setting it to EAGAIN, otherwise we return
it to the callers.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

When we're doing log commits, we try to wait for more writers to come in
and make the commit bigger.  This helps improve performance on rotating
disks, but on SSDs it adds latencies.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

The WARN_ON under some circumstances heavily polute log and slow down
the machine. This is just a safety, as the warning should be fixed by
another patch, nevertheless, it still pops up during testing.
Signed-off-by: David Sterba <dsterba@suse.cz>

There's a missing test whether the path passed to subvol=path option
during mount is a real subvolume, allowing any directory located in
default subovlume to be passed and accepted for mount.

(current btrfs progs prevent this early)
$ btrfs subvol snapshot . p1-snap
ERROR: '.' is not a subvolume

(with "is subvolume?" test bypassed)
$ btrfs subvol snapshot . p1-snap
Create a snapshot of '.' in './p1-snap'

$ btrfs subvol list -p .
ID 258 parent 5 top level 5 path subvol
ID 259 parent 5 top level 5 path subvol1
ID 260 parent 5 top level 5 path default-subvol1
ID 262 parent 5 top level 5 path p1/p1-snapshot
ID 263 parent 259 top level 5 path subvol1/subvol1-snap

The problem I see is that this makes a false impression of snapshotting the
given subvolume but in fact snapshots the default one: a user expects outcome
like ID 263 but in fact gets ID 262 .

This patch makes mount fail with EINVAL with a message in syslog.
Signed-off-by: David Sterba <dsterba@suse.cz>