This function is the entry to the extent allocator and as such has
quite a number of parameters. Some of those have subtle effects on the
allocation algorithm. Document the parameters.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NAnand Jain <anand.jain@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

As with every function which deals with modifying the btree
btrfs_uuid_tree_rem can fail for any number of reasons (ie. EIO/ENOMEM).
Handle return error value from this function gracefully by aborting the
transaction.

Fixes: dd5f9615 ("Btrfs: maintain subvolume items in the UUID tree")
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The kernel would like to have all stack VLA usage removed[1].
Unfortunately using an integer constant variable as the size of an
array is still considered a VLA. Instead let's use directly sizeof(var)
which removes the VLA usage. Use the occasion to remove csum_size
altogether and use sizeof() also for the size passed to memcmp

[1]: https://lkml.org/lkml/2018/3/7/621Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

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

Remove parameters not used by any of the callbacks.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Remove parameters not used by any of the callbacks.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The callbacks make use of different parameters that are passed to the
other type unnecessarily. This patch adds separate types for each and
the unused parameters will be removed.

The type extent_submit_bio_hook_t keeps all parameters and can be used
where the start/done types are not appropriate.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

A useless wrapper around tree_mod_log_insert_root that hides missing
error handling. Move it to the callers.
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

A trivial wrapper that can be simply opencoded and makes the GFP
allocation request more visible. The error handling is now moved to the
callers.
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The wrapper is effectively an alias for tree_mod_log_insert_move but
also hides the missing error handling. To make that more visible, lift
the BUG_ON to the callers.
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The wrappers are trivial and do not bring any extra value on top of the
plain locking primitives.
Reviewed-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The tree_mod_move is not used anywhere and can be embedded as anonymous
structure.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

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

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

It's provided by the extent_buffer.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

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

The merge call was factored out to a separate helper but it's a trivial
one and arguably we can opencode it and cache the value.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The value of page_end is only stored to end, no other use.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

All callers pass a valid pointer so we can drop the redundant checks.
The call to submit_one_bio never happend and can be removed.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

In case of raid56, writes and rebuilds always take BTRFS_STRIPE_LEN(64K)
as unit, however, scrub_extent() sets blocksize as unit, so rebuild
process may be triggered on every block on a same stripe.

A typical example would be that when we're replacing a disappeared disk,
all reads on the disks get -EIO, every block (size is 4K if blocksize is
4K) would go thru these,

scrub_handle_errored_block
  scrub_recheck_block # re-read pages one by one
  scrub_recheck_block # rebuild by calling raid56_parity_recover()
                        page by page

Although with raid56 stripe cache most of reads during rebuild can be
avoided, the parity recover calculation(xor or raid6 algorithms) needs to
be done $(BTRFS_STRIPE_LEN / blocksize) times.

This makes it smarter by doing raid56 scrub/replace on stripe length.
Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Sort mount options by the primary name, followed by the 'no-'
counterpart if it exists. Group the deprecated and debugging options.
Enum and token defintions are synced.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Btrfs has two mount options for SSD optimizations: ssd and ssd_spread.
Presently there is an option to disable all SSD optimizations, but there
isn't an option to disable just ssd_spread.

This patch adds a mount option nossd_spread that disables ssd_spread
only.
Reviewed-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NHoward McLauchlan <hmclauchlan@fb.com>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Since userspace transaction have been removed we no longer have use
for this field so delete it.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Now that the userspace transaction ioctls have been removed,
TRANS_USERSPACE is no longer used hence we can remove it.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Now that the userspace transaction IOCTL have been removed, this member
is no longer used so just remove it
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Commit 3558d4f8 ("btrfs: Deprecate userspace transaction ioctls")
marked the beginning of the end of userspace transaction. This commit
finishes the job! There are no known users and ceph does not use the
ioctl anymore.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Acked-by: NSage Weil <sage@redhat.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

btrfs: qgroup: Fix root item corruption when multiple same source snapshots are created with quota enabled

When multiple pending snapshots referring to the same source subvolume
are executed, enabled quota will cause root item corruption, where root
items are using old bytenr (no backref in extent tree).

This can be triggered by fstests btrfs/152.

The cause is when source subvolume is still dirty, extra commit
(simplied transaction commit) of qgroup_account_snapshot() can skip
dirty roots not recorded in current transaction, making root item of
source subvolume not updated.

Fix it by forcing recording source subvolume in current transaction
before qgroup sub-transaction commit.
Reported-by: NJustin Maggard <jmaggard@netgear.com>
Signed-off-by: NQu Wenruo <wqu@suse.com>
Reviewed-by: NFilipe Manana <fdmanana@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

When performing an unlock on an extent buffer we'd like to order the
decrement of extent_buffer::blocking_writers with waking up any
waiters. In such situations it's sufficient to use smp_mb__after_atomic
rather than the heavy smp_mb. On architectures where atomic operations
are fully ordered (such as x86 or s390) unconditionally executing
a heavyweight smp_mb instruction causes a severe hit to performance
while bringin no improvements in terms of correctness.

The better thing is to use the appropriate smp_mb__after_atomic routine
which will do the correct thing (invoke a full smp_mb or in the case
of ordered atomics insert a compiler barrier). Put another way,
an RMW atomic op + smp_load__after_atomic equals, in terms of
semantics, to a full smp_mb. This ensures that none of the problems
described in the accompanying comment of waitqueue_active occur.
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>

Some functions can filter metadata by the generation. Add a define that
will annotate such arguments.
Signed-off-by: NAnand Jain <anand.jain@oracle.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
[ update changelog ]
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The called function name is self explanatory.
Signed-off-by: NDavid Sterba <dsterba@suse.com>

The current implementation of btrfs_page_exists_in_range() gives the
wrong answer if the workingset code has stored a shadow entry in the
page cache.  The filemap_range_has_page() function does not have this
problem, and it's shared code, so use it instead.
eigned-off-by: NMatthew Wilcox <mawilcox@microsoft.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

In the last step of scrub_handle_error_block, we try to combine good
copies on all possible mirrors, this works fine for raid1 and raid10,
but not for raid56 as it's doing parity rebuild.

If parity rebuild doesn't get back with correct data which matches its
checksum, in case of replace we'd rather write what is stored in the
source device than the data calculuated from parity.
Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

async_missing_raid56() is identical to async_read_rebuild().
Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

Previously, btrfs_inode_by_name() returned 0 which left caller to check
objectid of location even location if the type was invalid.

Let btrfs_inode_by_name() return -EUCLEAN if a corrupted location of a
dir entry is found.  Removal of label out_err also simplifies the
function.
Signed-off-by: NSu Yue <suy.fnst@cn.fujitsu.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
[ drop unlikely ]
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This function btrfs_close_extra_devices() is about freeing
extra devids which once it may have belonged to this filesystem.
So rename it and add the comment. The _devid suffix is
appropriate as this function won't handle devices which are
outside of the filesytem being mounted.
Signed-off-by: NAnand Jain <anand.jain@oracle.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>

This argument is always set to the root of the inode, which is also
passed. So let's get a reference inside the function and simplify
the arg list.
Signed-off-by: NNikolay Borisov <nborisov@suse.com>
Reviewed-by: NDavid Sterba <dsterba@suse.com>
Signed-off-by: NDavid Sterba <dsterba@suse.com>