A crafted btrfs image with incorrect chunk<->block group mapping will
trigger a lot of unexpected things as the mapping is essential.

Although the problem can be caught by block group item checker
added in "btrfs: tree-checker: Verify block_group_item", it's still not
sufficient.  A sufficiently valid block group item can pass the check
added by the mentioned patch but could fail to match the existing chunk.

This patch will add extra block group -> chunk mapping check, to ensure
we have a completely matching (start, len, flags) chunk for each block
group at mount time.

Here we reuse the original helper find_first_block_group(), which is
already doing the basic bg -> chunk checks, adding further checks of the
start/len and type flags.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=199837Reported-by: NXu Wen <wen.xu@gatech.edu>
Signed-off-by: NQu Wenruo <wqu@suse.com>
Reviewed-by: NSu Yue <suy.fnst@cn.fujitsu.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

There is no user of this function anymore.

This was forgotten to be removed in commit a575ceeb
("Btrfs: get rid of unused orphan infrastructure").
Signed-off-by: NMisono Tomohiro <misono.tomohiro@jp.fujitsu.com>
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed transaction handle.
Signed-off-by: NLu Fengqi <lufq.fnst@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Leftover after fix e339a6b0 ("Btrfs: __btrfs_mod_ref should always
use no_quota"), that removed it from the function calls but not the
structure.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If we're trying to make a data reservation and we have to allocate a
data chunk we could leak ret == 1, as do_chunk_alloc() will return 1 if
it allocated a chunk.  Since the end of the function is the success path
just return 0.

CC: stable@vger.kernel.org # 4.4+
Signed-off-by: NJosef Bacik <josef@toxicpanda.com>
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed transaction handle.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NLu Fengqi <lufq.fnst@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

In find_free_extent() under checks: label, we have the following code:

		search_start = ALIGN(offset, fs_info->stripesize);
		/* move on to the next group */
		if (search_start + num_bytes >
		    block_group->key.objectid + block_group->key.offset) {
			btrfs_add_free_space(block_group, offset, num_bytes);
			goto loop;
		}
		if (offset < search_start)
			btrfs_add_free_space(block_group, offset,
					     search_start - offset);
		BUG_ON(offset > search_start);

However ALIGN() is rounding up, thus @search_start >= @offset and that
BUG_ON() will never be triggered.
Signed-off-by: NQu Wenruo <wqu@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

There are many places that open code the duplicity factor of the block
group profiles, create a common helper. This can be easily extended for
more copies.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The fs_info can be fetched from the transaction handle directly.
Signed-off-by: NLu Fengqi <lufq.fnst@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be fetched from the transaction handle.
Signed-off-by: NLu Fengqi <lufq.fnst@cn.fujitsu.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Introduce a small helper, btrfs_mark_bg_unused(), to acquire locks and
add a block group to unused_bgs list.

No functional modification, and only 3 callers are involved.
Signed-off-by: NQu Wenruo <wqu@suse.com>
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

do_chunk_alloc implements logic to detect whether there is currently
pending chunk allocation (by means of space_info->chunk_alloc being
set) and if so it loops around to the 'again' label. Additionally,
based on the state of the space_info (e.g. whether it's full or not)
and the return value of should_alloc_chunk() it decides whether this
is a "hard" error (ENOSPC) or we can just return 0.

This patch refactors all of this:

1. Put order to the scattered ifs handling the various cases in an
easy-to-read if {} else if{} branches. This makes clear the various
cases we are interested in handling.

2. Call should_alloc_chunk only once and use the result in the
if/else if constructs. All of this is done under space_info->lock, so
even before multiple calls of should_alloc_chunk were unnecessary.

3. Rewrite the "do {} while()" loop currently implemented via label
into an explicit loop construct.

4. Move the mutex locking for the case where the caller is the one doing
the allocation. For the case where the caller needs to wait a concurrent
allocation, introduce a pair of mutex_lock/mutex_unlock to act as a
barrier and reword the comment.

5. Switch local vars to bool type where pertinent.

All in all this shouldn't introduce any functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

In commit b150a4f1 ("Btrfs: use a percpu to keep track of possibly
pinned bytes") we use total_bytes_pinned to track how many bytes we are
going to free in this transaction. When we are close to ENOSPC, we check it
and know if we can make the allocation by commit the current transaction.
For every data/metadata extent we are going to free, we add
total_bytes_pinned in btrfs_free_extent() and btrfs_free_tree_block(), and
release it in unpin_extent_range() when we finish the transaction. So this
is a variable we frequently update but rarely read - just the suitable
use of percpu_counter. But in previous commit we update total_bytes_pinned
by default 32 batch size, making every update essentially a spin lock
protected update. Since every spin lock/unlock operation involves syncing
a globally used variable and some kind of barrier in a SMP system, this is
more expensive than using total_bytes_pinned as a simple atomic64_t.

So fix this by using a customized batch size. Since we only read
total_bytes_pinned when we are close to ENOSPC and fail to allocate new
chunk, we can use a really large batch size and have nearly no penalty
in most cases.

[Test]
We tested the patch on a 4-cores x86 machine:

1. fallocate a 16GiB size test file
2. take snapshot (so all following writes will be COW)
3. run a 180 sec, 4 jobs, 4K random write fio on test file

We also added a temporary lockdep class on percpu_counter's spin lock
used by total_bytes_pinned to track it by lock_stat.

[Results]
unpatched:
lock_stat version 0.4
-----------------------------------------------------------------------
                              class name    con-bounces    contentions
waittime-min   waittime-max waittime-total   waittime-avg    acq-bounces
acquisitions   holdtime-min   holdtime-max holdtime-total   holdtime-avg

               total_bytes_pinned_percpu:            82             82
        0.21           0.61          29.46           0.36         298340
      635973           0.09          11.01      173476.25           0.27

patched:
lock_stat version 0.4
-----------------------------------------------------------------------
                              class name    con-bounces    contentions
waittime-min   waittime-max waittime-total   waittime-avg    acq-bounces
acquisitions   holdtime-min   holdtime-max holdtime-total   holdtime-avg

               total_bytes_pinned_percpu:             1              1
        0.62           0.62           0.62           0.62          13601
       31542           0.14           9.61       11016.90           0.35

[Analysis]
Since the spin lock only protects a single in-memory variable, the
contentions (number of lock acquisitions that had to wait) in both
unpatched and patched version are low. But when we see acquisitions and
acq-bounces, we get much lower counts in patched version. Here the most
important metric is acq-bounces. It means how many times the lock gets
transferred between different cpus, so the patch can really reduce
cacheline bouncing of spin lock (also the global counter of percpu_counter)
in a SMP system.

Fixes: b150a4f1 ("Btrfs: use a percpu to keep track of possibly pinned bytes")
Signed-off-by: NEthan Lien <ethanlien@synology.com>
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Functions that get btrfs inode can simply reach the fs_info by
dereferencing the root and this looks a bit more straightforward
compared to the btrfs_sb(...) indirection.

If the transaction handle is available and not NULL it's used instead.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The v0 extent type checks are the right case for the unlikely
annotations as we don't expect to ever see them, so let's give the
compiler some hint.
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Following the removal of the v0 handling code let's be courteous and
print an error message when such extents are handled. In the cases
where we have a transaction just abort it, otherwise just call
btrfs_handle_fs_error. Both cases result in the FS being re-mounted RO.

In case the error handling would be too intrusive, leave the BUG_ON in
place, like extent_data_ref_count, other proper handling would catch
that earlier.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The v0 compat code was introduced in commit 5d4f98a2
("Btrfs: Mixed back reference  (FORWARD ROLLING FORMAT CHANGE)") 9
years ago, which was merged in 2.6.31. This means that the code is
there to support filesystems which are _VERY_ old and if you are using
btrfs on such an old kernel, you have much bigger problems. This coupled
with the fact that no one is likely testing/maintining this code likely
means it has bugs lurking. All things considered I think 43 kernel
releases later it's high time this remnant of the past got removed.

This patch removes all code wrapped in #ifdefs but leaves the BUG_ONs in case
we have a v0 with no support intact as a sort of safety-net.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

If type of extent_inline_ref found is not expected, filesystem may have
been corrupted, should return EUCLEAN instead of EINVAL.
Signed-off-by: NSu Yue <suy.fnst@cn.fujitsu.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

[BUG]
Under certain KVM load and LTP tests, it is possible to hit the
following calltrace if quota is enabled:

BTRFS critical (device vda2): unable to find logical 8820195328 length 4096
BTRFS critical (device vda2): unable to find logical 8820195328 length 4096

WARNING: CPU: 0 PID: 49 at ../block/blk-core.c:172 blk_status_to_errno+0x1a/0x30
CPU: 0 PID: 49 Comm: kworker/u2:1 Not tainted 4.12.14-15-default #1 SLE15 (unreleased)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.0.0-prebuilt.qemu-project.org 04/01/2014
Workqueue: btrfs-endio-write btrfs_endio_write_helper [btrfs]
task: ffff9f827b340bc0 task.stack: ffffb4f8c0304000
RIP: 0010:blk_status_to_errno+0x1a/0x30
Call Trace:
 submit_extent_page+0x191/0x270 [btrfs]
 ? btrfs_create_repair_bio+0x130/0x130 [btrfs]
 __do_readpage+0x2d2/0x810 [btrfs]
 ? btrfs_create_repair_bio+0x130/0x130 [btrfs]
 ? run_one_async_done+0xc0/0xc0 [btrfs]
 __extent_read_full_page+0xe7/0x100 [btrfs]
 ? run_one_async_done+0xc0/0xc0 [btrfs]
 read_extent_buffer_pages+0x1ab/0x2d0 [btrfs]
 ? run_one_async_done+0xc0/0xc0 [btrfs]
 btree_read_extent_buffer_pages+0x94/0xf0 [btrfs]
 read_tree_block+0x31/0x60 [btrfs]
 read_block_for_search.isra.35+0xf0/0x2e0 [btrfs]
 btrfs_search_slot+0x46b/0xa00 [btrfs]
 ? kmem_cache_alloc+0x1a8/0x510
 ? btrfs_get_token_32+0x5b/0x120 [btrfs]
 find_parent_nodes+0x11d/0xeb0 [btrfs]
 ? leaf_space_used+0xb8/0xd0 [btrfs]
 ? btrfs_leaf_free_space+0x49/0x90 [btrfs]
 ? btrfs_find_all_roots_safe+0x93/0x100 [btrfs]
 btrfs_find_all_roots_safe+0x93/0x100 [btrfs]
 btrfs_find_all_roots+0x45/0x60 [btrfs]
 btrfs_qgroup_trace_extent_post+0x20/0x40 [btrfs]
 btrfs_add_delayed_data_ref+0x1a3/0x1d0 [btrfs]
 btrfs_alloc_reserved_file_extent+0x38/0x40 [btrfs]
 insert_reserved_file_extent.constprop.71+0x289/0x2e0 [btrfs]
 btrfs_finish_ordered_io+0x2f4/0x7f0 [btrfs]
 ? pick_next_task_fair+0x2cd/0x530
 ? __switch_to+0x92/0x4b0
 btrfs_worker_helper+0x81/0x300 [btrfs]
 process_one_work+0x1da/0x3f0
 worker_thread+0x2b/0x3f0
 ? process_one_work+0x3f0/0x3f0
 kthread+0x11a/0x130
 ? kthread_create_on_node+0x40/0x40
 ret_from_fork+0x35/0x40

BTRFS critical (device vda2): unable to find logical 8820195328 length 16384
BTRFS: error (device vda2) in btrfs_finish_ordered_io:3023: errno=-5 IO failure
BTRFS info (device vda2): forced readonly
BTRFS error (device vda2): pending csums is 2887680

[CAUSE]
It's caused by race with block group auto removal:

- There is a meta block group X, which has only one tree block
  The tree block belongs to fs tree 257.
- In current transaction, some operation modified fs tree 257
  The tree block gets COWed, so the block group X is empty, and marked
  as unused, queued to be deleted.
- Some workload (like fsync) wakes up cleaner_kthread()
  Which will call btrfs_delete_unused_bgs() to remove unused block
  groups.
  So block group X along its chunk map get removed.
- Some delalloc work finished for fs tree 257
  Quota needs to get the original reference of the extent, which will
  read tree blocks of commit root of 257.
  Then since the chunk map gets removed, the above warning gets
  triggered.

[FIX]
Just let btrfs_delete_unused_bgs() skip block group which still has
pinned bytes.

However there is a minor side effect: currently we only queue empty
blocks at update_block_group(), and such empty block group with pinned
bytes won't go through update_block_group() again, such block group
won't be removed, until it gets new extent allocated and removed.
Signed-off-by: NQu Wenruo <wqu@suse.com>
Reviewed-by: NFilipe Manana <fdmanana@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

When a new extent buffer is allocated there are a few mandatory fields
which need to be set in order for the buffer to be sane: level,
generation, bytenr, backref_rev, owner and FSID/UUID. Currently this
is open coded in the callers of btrfs_alloc_tree_block, meaning it's
fairly high in the abstraction hierarchy of operations. This patch
solves this by simply moving this init code in btrfs_init_new_buffer,
since this is the function which initializes a newly allocated
extent buffer. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed transaction handle.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed bg cache.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from trans since the function is always called
within a valid transaction.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced directly from the transaction handle since it's
always valid.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed transaction handle, since it's
always valid.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed transaction handle, since it's
always valid.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed block group.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from the passed block group.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from trans since the function is always called
within a transaction.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can be referenced from trans since the function is always called
within a transaction.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function is always called with a valid transaction handle from
where fs_info can be referenced. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It can always be referneced from the passed transaction handle since
it's always valid. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

fs_info can be refenreced from the transaction handle, since it's always
valid. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The argument is no longer used so remove it.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function is always called with a valid transaction handle so
fs_info can be referenced from there. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function is always called with a valid transaction from where
fs_info can be referenced. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function already takes a transaction which holds a reference to
the fs_info struct. Use that reference and remove the extra arg. No
functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

fs_info can be referenced from the transaction handle, which is always
valid. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function is always called with a valid transaction handle so we
can reference the fs_info from there. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function is always called with a valid transaction handle from
where we can reference fs_info. No functional changes.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NQu Wenruo <wqu@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>