The sectorsize member of btrfs_block_group_cache is unused. So remove it, this
reduces the number of holes in the struct.

With patch:
/* size: 856, cachelines: 14, members: 40 */
/* sum members: 837, holes: 4, sum holes: 19 */
/* bit holes: 1, sum bit holes: 29 bits */
/* last cacheline: 24 bytes */

Without patch:
/* size: 864, cachelines: 14, members: 41 */
/* sum members: 841, holes: 5, sum holes: 23 */
/* bit holes: 1, sum bit holes: 29 bits */
/* last cacheline: 32 bytes */
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If a lot of metadata is reserved for outstanding delayed allocations, we
rely on shrink_delalloc() to reclaim metadata space in order to fulfill
reservation tickets. However, shrink_delalloc() has a shortcut where if
it determines that space can be overcommitted, it will stop early. This
made sense before the ticketed enospc system, but now it means that
shrink_delalloc() will often not reclaim enough space to fulfill any
tickets, leading to an early ENOSPC. (Reservation tickets don't care
about being able to overcommit, they need every byte accounted for.)

Fix it by getting rid of the shortcut so that shrink_delalloc() reclaims
all of the metadata it is supposed to. This fixes early ENOSPCs we were
seeing when doing a btrfs receive to populate a new filesystem, as well
as early ENOSPCs Christoph saw when doing a big cp -r onto Btrfs.

Fixes: 957780eb ("Btrfs: introduce ticketed enospc infrastructure")
Tested-by: NChristoph Anton Mitterer <mail@christoph.anton.mitterer.name>
Cc: stable@vger.kernel.org
Reviewed-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If we have a block group that is all of the following:
1) uncached in memory
2) is read-only
3) has a disk cache state that indicates we need to recreate the cache

AND the file system has enough free space fragmentation such that the
request for an extent of a given size can't be honored;

AND have a single CPU core;

AND it's the block group with the highest starting offset such that
there are no opportunities (like reading from disk) for the loop to
yield the CPU;

We can end up with a lockup.

The root cause is simple.  Once we're in the position that we've read in
all of the other block groups directly and none of those block groups
can honor the request, there are no more opportunities to sleep.  We end
up trying to start a caching thread which never gets run if we only have
one core.  This *should* present as a hung task waiting on the caching
thread to make some progress, but it doesn't.  Instead, it degrades into
a busy loop because of the placement of the read-only check.

During the first pass through the loop, block_group->cached will be set
to BTRFS_CACHE_STARTED and have_caching_bg will be set.  Then we hit the
read-only check and short circuit the loop.  We're not yet in
LOOP_CACHING_WAIT, so we skip that loop back before going through the
loop again for other raid groups.

Then we move to LOOP_CACHING_WAIT state.

During the this pass through the loop, ->cached will still be
BTRFS_CACHE_STARTED, which means it's not cached, so we'll enter
cache_block_group, do a lot of nothing, and return, and also set
have_caching_bg again.  Then we hit the read-only check and short circuit
the loop.  The same thing happens as before except now we DO trigger
the LOOP_CACHING_WAIT && have_caching_bg check and loop back up to the
top.  We do this forever.

There are two fixes in this patch since they address the same underlying
bug.

The first is to add a cond_resched to the end of the loop to ensure
that the caching thread always has an opportunity to run.  This will
fix the soft lockup issue, but find_free_extent will still loop doing
nothing until the thread has completed.

The second is to move the read-only check to the top of the loop.  We're
never going to return an allocation within a read-only block group so
we may as well skip it early.  The check for ->cached == BTRFS_CACHE_ERROR
would cause the same problem except that BTRFS_CACHE_ERROR is considered
a "done" state and we won't re-set have_caching_bg again.

Many thanks to Stephan Kulow <coolo@suse.de> for his excellent help in
the testing process.
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Dave Jones hit a WARN_ON(nr < 0) in btrfs_wait_ordered_roots() with
v4.12-rc6.  This was because commit 70e7af24 made it possible for
calc_reclaim_items_nr() to return a negative number.  It's not really a
bug in that commit, it just didn't go far enough down the stack to find
all the possible 64->32 bit overflows.

This switches calc_reclaim_items_nr() to return a u64 and changes everyone
that uses the results of that math to u64 as well.
Reported-by: NDave Jones <davej@codemonkey.org.uk>
Fixes: 70e7af24 ("Btrfs: fix delalloc accounting leak caused by u32 overflow")
Signed-off-by: NChris Mason <clm@fb.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

[BUG]
For the following case, btrfs can underflow qgroup reserved space
at an error path:
(Page size 4K, function name without "btrfs_" prefix)

         Task A                  |             Task B
----------------------------------------------------------------------
Buffered_write [0, 2K)           |
|- check_data_free_space()       |
|  |- qgroup_reserve_data()      |
|     Range aligned to page      |
|     range [0, 4K)          <<< |
|     4K bytes reserved      <<< |
|- copy pages to page cache      |
                                 | Buffered_write [2K, 4K)
                                 | |- check_data_free_space()
                                 | |  |- qgroup_reserved_data()
                                 | |     Range alinged to page
                                 | |     range [0, 4K)
                                 | |     Already reserved by A <<<
                                 | |     0 bytes reserved      <<<
                                 | |- delalloc_reserve_metadata()
                                 | |  And it *FAILED* (Maybe EQUOTA)
                                 | |- free_reserved_data_space()
                                      |- qgroup_free_data()
                                         Range aligned to page range
                                         [0, 4K)
                                         Freeing 4K
(Special thanks to Chandan for the detailed report and analyse)

[CAUSE]
Above Task B is freeing reserved data range [0, 4K) which is actually
reserved by Task A.

And at writeback time, page dirty by Task A will go through writeback
routine, which will free 4K reserved data space at file extent insert
time, causing the qgroup underflow.

[FIX]
For btrfs_qgroup_free_data(), add @reserved parameter to only free
data ranges reserved by previous btrfs_qgroup_reserve_data().
So in above case, Task B will try to free 0 byte, so no underflow.
Reported-by: NChandan Rajendra <chandan@linux.vnet.ibm.com>
Signed-off-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
Reviewed-by: NChandan Rajendra <chandan@linux.vnet.ibm.com>
Tested-by: NChandan Rajendra <chandan@linux.vnet.ibm.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Introduce a new parameter, struct extent_changeset for
btrfs_qgroup_reserved_data() and its callers.

Such extent_changeset was used in btrfs_qgroup_reserve_data() to record
which range it reserved in current reserve, so it can free it in error
paths.

The reason we need to export it to callers is, at buffered write error
path, without knowing what exactly which range we reserved in current
allocation, we can free space which is not reserved by us.

This will lead to qgroup reserved space underflow.
Reviewed-by: NChandan Rajendra <chandan@linux.vnet.ibm.com>
Signed-off-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

btrfs_qgroup_release/free_data() only returns 0 or a negative error
number (ENOMEM is the only possible error).

This is normally good enough, but sometimes we need the exact byte
count it freed/released.

Change it to return actually released/freed bytenr number instead of 0
for success.
And slightly modify related extent_changeset structure, since in btrfs
one no-hole data extent won't be larger than 128M, so "unsigned int"
is large enough for the use case.
Signed-off-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The total_bytes_pinned counter is completely broken when accounting
delayed refs:

- If two drops for the same extent are merged, we will decrement
  total_bytes_pinned twice but only increment it once.
- If an add is merged into a drop or vice versa, we will decrement the
  total_bytes_pinned counter but never increment it.
- If multiple references to an extent are dropped, we will account it
  multiple times, potentially vastly over-estimating the number of bytes
  that will be freed by a commit and doing unnecessary work when we're
  close to ENOSPC.

The last issue is relatively minor, but the first two make the
total_bytes_pinned counter leak or underflow very often. These
accounting issues were introduced in b150a4f1 ("Btrfs: use a percpu
to keep track of possibly pinned bytes"), but they were papered over by
zeroing out the counter on every commit until d288db5d ("Btrfs: fix
race of using total_bytes_pinned").

We need to make sure that an extent is accounted as pinned exactly once
if and only if we will drop references to it when when the transaction
is committed. Ideally we would only add to total_bytes_pinned when the
*last* reference is dropped, but this information isn't readily
available for data extents. Again, this over-estimation can lead to
extra commits when we're close to ENOSPC, but it's not as bad as before.

The fix implemented here is to increment total_bytes_pinned when the
total refmod count for an extent goes negative and decrement it if the
refmod count goes back to non-negative or after we've run all of the
delayed refs for that extent.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

We need this to decide when to account pinned bytes.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Currently, we only increment total_bytes_pinned in
btrfs_free_tree_block() when dropping the last reference on the block.
However, when the delayed ref is run later, we will decrement
total_bytes_pinned regardless of whether it was the last reference or
not. This causes the counter to underflow when the reference we dropped
was not the last reference. Fix it by incrementing the counter
unconditionally, which is what btrfs_free_extent() does. This makes
total_bytes_pinned an overestimate when references to shared extents are
dropped, but in the worst case this will just make us try to commit the
transaction to try to free up space and find we didn't free enough.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The extents marked in pin_down_extent() will be unpinned later in
unpin_extent_range(), which decrements total_bytes_pinned.
pin_down_extent() must increment the counter to avoid underflowing it.
Also adjust btrfs_free_tree_block() to avoid accounting for the same
extent twice.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The value of flags is one of DATA/METADATA/SYSTEM, they must exist at
when add_pinned_bytes is called.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
[ added changelog ]
Signed-off-by: NDavid Sterba <dsterba@suse.com>

There are a few places where we pass in a negative num_bytes, so make it
signed for clarity. Also move it up in the file since later patches will
need it there.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Following the factoring out of the creation code udpate_space_info can
only be called for already-existing space_info structs. As such it
cannot fail.  Remove superfluous error handling and make the function
return void.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Currently the struct space_info creation code is intermixed in the
udpate_space_info function. There are well-defined points at which the
we actually want to create brand-new space_info structs (e.g. during
mount of the filesystem as well as sometimes when adding/initialising
new chunks). In such cases update_space_info is called with 0 as the
bytes parameter. All of this makes for spaghetti code.

Fix it by factoring out the creation code in a separate
create_space_info structure. This also allows to simplify the internals.
Also remove BUG_ON from do_alloc_chunk since the callers handle errors.
Furthermore it will make the update_space_info function not fail,
allowing us to remove error handling in callers. This will come in a
follow up patch.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

We commit transaction in order to reclaim space from pinned bytes because
it could process delayed refs, and in may_commit_transaction(), we check
first if pinned bytes are enough for the required space, we then check if
that plus bytes reserved for delayed insert are enough for the required
space.

This changes the code to the above logic.

Fixes: b150a4f1 ("Btrfs: use a percpu to keep track of possibly pinned bytes")
Tested-by: NNikolay Borisov <nborisov@suse.com>
Reported-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

can_overcommit using the root to determine the allocation profile
is the only use of a root in the call graph below reserve_metadata_bytes.

It turns out that we only need to know whether the allocation is for
the chunk root or not -- and we can pass that around as a bool instead.

This allows us to pull root usage out of the reservation path all the
way up to reserve_metadata_bytes itself, which uses it only to compare
against fs_info->chunk_root to set the bool.  In turn, this eliminates
a bunch of races where we use a particular root too early in the mount
process.
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

There are two places where we don't already know what kind of alloc
profile we need before calling btrfs_get_alloc_profile, but we need
access to a root everywhere we call it.

This patch adds helpers for btrfs_{data,metadata,system}_alloc_profile()
and relegates btrfs_system_alloc_profile to a static for use in those
two cases.  The next patch will eliminate one of those.
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The ->free_chunk_space variable is used to track the unallocated space
and access to it is protected by a spinlock, which is not used for
anything else.  Make the code a bit self-explanatory by switching the
variable to an atomic64_t type and kill the spinlock.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
[ not a performance critical code, use of atomic type is ok ]
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If we have to recover relocation during mount, we'll ultimately have to
evict the orphan inode.  That goes through the reservation dance, where
priority_reclaim_metadata_space and flush_space expect fs_info->fs_root
to be valid.  That's the next thing to be set up during mount, so we
crash, almost always in flush_space trying to join the transaction
but priority_reclaim_metadata_space is possible as well.  This call
path has been problematic in the past WRT whether ->fs_root is valid
yet.  Commit 957780eb (Btrfs: introduce ticketed enospc
infrastructure) added new users that are called in the direct path
instead of the async path that had already been worked around.

The thing is that we don't actually need the fs_root, specifically, for
anything.  We either use it to determine whether the root is the
chunk_root for use in choosing an allocation profile or as a root to pass
btrfs_join_transaction before immediately committing it.  Anything that
isn't the chunk root works in the former case and any root works in
the latter.

A simple fix is to use a root we know will always be there: the
extent_root.

Cc: <stable@vger.kernel.org> # v4.8+
Fixes: 957780eb (Btrfs: introduce ticketed enospc infrastructure)
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If we fail to add the space_info kobject, we'll leak the memory
for the percpu counter.

Fixes: 6ab0a202 (btrfs: publish allocation data in sysfs)
Cc: <stable@vger.kernel.org> # v3.14+
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Unlike mirror based profiles, RAID5/6 recovery needs to read out the
whole full stripe.

And if we don't do proper protection, it can easily cause race condition.

Introduce 2 new functions: lock_full_stripe() and unlock_full_stripe()
for RAID5/6.
Which store a rb_tree of mutexes for full stripes, so scrub callers can
use them to lock a full stripe to avoid race.
Signed-off-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
[ minor comment adjustments ]
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The members have been effectively unused since "Btrfs: rework qgroup
accounting" (fcebe456), there's no substitute for
assert_qgroups_uptodate so it's removed as well.
Reviewed-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

We also don't bother to flush free space cache while with free space
tree.

Cc: David Sterba <dsterba@suse.cz>
Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

refcount_t type and corresponding API should be
used instead of atomic_t when the variable is used as
a reference counter. This allows to avoid accidental
refcounter overflows that might lead to use-after-free
situations.
Signed-off-by: NElena Reshetova <elena.reshetova@intel.com>
Signed-off-by: NHans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NDavid Windsor <dwindsor@gmail.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

refcount_t type and corresponding API should be
used instead of atomic_t when the variable is used as
a reference counter. This allows to avoid accidental
refcounter overflows that might lead to use-after-free
situations.
Signed-off-by: NElena Reshetova <elena.reshetova@intel.com>
Signed-off-by: NHans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NDavid Windsor <dwindsor@gmail.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

refcount_t type and corresponding API should be
used instead of atomic_t when the variable is used as
a reference counter. This allows to avoid accidental
refcounter overflows that might lead to use-after-free
situations.
Signed-off-by: NElena Reshetova <elena.reshetova@intel.com>
Signed-off-by: NHans Liljestrand <ishkamiel@gmail.com>
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NDavid Windsor <dwindsor@gmail.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Instead of including the full <linux/signal.h>, we are going to include the
types-only <linux/signal_types.h> header in <linux/sched.h>, to further
decouple the scheduler header from the signal headers.

This means that various files which relied on the full <linux/signal.h> need
to be updated to gain an explicit dependency on it.

Update the code that relies on sched.h's inclusion of the <linux/signal.h> header.
Acked-by: NLinus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

At close_ctree() we free the block groups and then only after we wait for
any running worker kthreads to finish and shutdown the workqueues. This
behaviour is racy and it triggers an assertion failure when freeing block
groups because while we are doing it we can have for example a block group
caching kthread running, and in that case the block group's reference
count can still be greater than 1 by the time we assert its reference count
is 1, leading to an assertion failure:

[19041.198004] assertion failed: atomic_read(&block_group->count) == 1, file: fs/btrfs/extent-tree.c, line: 9799
[19041.200584] ------------[ cut here ]------------
[19041.201692] kernel BUG at fs/btrfs/ctree.h:3418!
[19041.202830] invalid opcode: 0000 [#1] PREEMPT SMP
[19041.203929] Modules linked in: btrfs xor raid6_pq dm_flakey dm_mod crc32c_generic ppdev sg psmouse acpi_cpufreq pcspkr parport_pc evdev tpm_tis parport tpm_tis_core i2c_piix4 i2c_core tpm serio_raw processor button loop autofs4 ext4 crc16 jbd2 mbcache sr_mod cdrom sd_mod ata_generic virtio_scsi ata_piix virtio_pci libata virtio_ring virtio e1000 scsi_mod floppy [last unloaded: btrfs]
[19041.208082] CPU: 6 PID: 29051 Comm: umount Not tainted 4.9.0-rc7-btrfs-next-36+ #1
[19041.208082] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.9.1-0-gb3ef39f-prebuilt.qemu-project.org 04/01/2014
[19041.208082] task: ffff88015f028980 task.stack: ffffc9000ad34000
[19041.208082] RIP: 0010:[<ffffffffa03e319e>]  [<ffffffffa03e319e>] assfail.constprop.41+0x1c/0x1e [btrfs]
[19041.208082] RSP: 0018:ffffc9000ad37d60  EFLAGS: 00010286
[19041.208082] RAX: 0000000000000061 RBX: ffff88015ecb4000 RCX: 0000000000000001
[19041.208082] RDX: ffff88023f392fb8 RSI: ffffffff817ef7ba RDI: 00000000ffffffff
[19041.208082] RBP: ffffc9000ad37d60 R08: 0000000000000001 R09: 0000000000000000
[19041.208082] R10: ffffc9000ad37cb0 R11: ffffffff82f2b66d R12: ffff88023431d170
[19041.208082] R13: ffff88015ecb40c0 R14: ffff88023431d000 R15: ffff88015ecb4100
[19041.208082] FS:  00007f44f3d42840(0000) GS:ffff88023f380000(0000) knlGS:0000000000000000
[19041.208082] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[19041.208082] CR2: 00007f65d623b000 CR3: 00000002166f2000 CR4: 00000000000006e0
[19041.208082] Stack:
[19041.208082]  ffffc9000ad37d98 ffffffffa035989f ffff88015ecb4000 ffff88015ecb5630
[19041.208082]  ffff88014f6be000 0000000000000000 00007ffcf0ba6a10 ffffc9000ad37df8
[19041.208082]  ffffffffa0368cd4 ffff88014e9658e0 ffffc9000ad37e08 ffffffff811a634d
[19041.208082] Call Trace:
[19041.208082]  [<ffffffffa035989f>] btrfs_free_block_groups+0x17f/0x392 [btrfs]
[19041.208082]  [<ffffffffa0368cd4>] close_ctree+0x1c5/0x2e1 [btrfs]
[19041.208082]  [<ffffffff811a634d>] ? evict_inodes+0x132/0x141
[19041.208082]  [<ffffffffa034356d>] btrfs_put_super+0x15/0x17 [btrfs]
[19041.208082]  [<ffffffff8118fc32>] generic_shutdown_super+0x6a/0xeb
[19041.208082]  [<ffffffff8119004f>] kill_anon_super+0x12/0x1c
[19041.208082]  [<ffffffffa0343370>] btrfs_kill_super+0x16/0x21 [btrfs]
[19041.208082]  [<ffffffff8118fad1>] deactivate_locked_super+0x3b/0x68
[19041.208082]  [<ffffffff8118fb34>] deactivate_super+0x36/0x39
[19041.208082]  [<ffffffff811a9946>] cleanup_mnt+0x58/0x76
[19041.208082]  [<ffffffff811a99a2>] __cleanup_mnt+0x12/0x14
[19041.208082]  [<ffffffff81071573>] task_work_run+0x6f/0x95
[19041.208082]  [<ffffffff81001897>] prepare_exit_to_usermode+0xa3/0xc1
[19041.208082]  [<ffffffff81001a23>] syscall_return_slowpath+0x16e/0x1d2
[19041.208082]  [<ffffffff814c607d>] entry_SYSCALL_64_fastpath+0xab/0xad
[19041.208082] Code: c7 ae a0 3e a0 48 89 e5 e8 4e 74 d4 e0 0f 0b 55 89 f1 48 c7 c2 0b a4 3e a0 48 89 fe 48 c7 c7 a4 a6 3e a0 48 89 e5 e8 30 74 d4 e0 <0f> 0b 55 31 d2 48 89 e5 e8 d5 b9 f7 ff 5d c3 48 63 f6 55 31 c9
[19041.208082] RIP  [<ffffffffa03e319e>] assfail.constprop.41+0x1c/0x1e [btrfs]
[19041.208082]  RSP <ffffc9000ad37d60>
[19041.279264] ---[ end trace 23330586f16f064d ]---

This started happening as of kernel 4.8, since commit f3bca802
("Btrfs: add ASSERT for block group's memory leak") introduced these
assertions.

So fix this by freeing the block groups only after waiting for all
worker kthreads to complete and shutdown the workqueues.
Signed-off-by: NFilipe Manana <fdmanana@suse.com>
Reviewed-by: NLiu Bo <bo.li.liu@oracle.com>

The free space cache APIs accept a root but always use the tree root.

Also, btrfs_truncate_free_space_cache accepts a root AND an inode but
the inode always points to the root anyway, so let's just pass the inode.
Signed-off-by: NJeff Mahoney <jeffm@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>