The btrfs allocator uses list_for_each to walk the available block
groups when searching for free blocks. It starts off with a hint
to help find the best block group for a given allocation.
The hint is resolved into a block group, but we don't properly check
to make sure the block group we find isn't in the middle of being
freed due to filesystem shrinking or balancing. If it is being
freed, the list pointers in it are bogus and can't be trusted. But,
the code happily goes along and uses them in the list_for_each loop,
leading to all kinds of fun.
The fix used here is to check to make sure the block group we find really
is on the list before we use it. list_del_init is used when removing
it from the list, so we can do a proper check.
The allocation clustering code has a similar bug where it will trust
the block group in the current free space cluster. If our allocation
flags have changed (going from single spindle dup to raid1 for example)
because the drives in the FS have changed, we're not allowed to use
the old block group any more.
The fix used here is to check the current cluster against the
current allocation flags.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The notreelog and flushoncommit mount options were being printed slightly
differently.
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
In Li Zefan's commit dae7b665cf,
a combination call of kmalloc() and copy_from_user() is replaced by
memdup_user(). So btrfs_ioctl_resize() doesn't use GFP_NOFS any more.
Signed-off-by: Li Hong <lihong.hi@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
These debugging WARN_ONs make too much console noise during regular
IO failures. An IO failure will still generate a number of messages
as we verify checksums etc, but these two are not needed.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When a btrfs metadata read fails, the first thing we try to do is find
a good copy on another mirror of the block. If this fails, read_tree_block()
ends up returning a buffer that isn't up to date.
The btrfs btree reading code was reworked to drop locks and repeat
the search when IO was done, but the changes didn't add a check for failed
reads. The end result was looping forever on buffers that were never
going to become up to date.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This flag is used to decide when we need to send a given file through
the ordered code to make sure it is fully written before a transaction
commits. It was not being properly set to zero when the inode was
being setup.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This changes btrfs_read_locked_inode() to peek ahead in the btree for acl items.
If it is certain a given inode has no acls, it will set the in memory acl
fields to null to avoid acl lookups completely.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Linus noticed the btrfs code to cache acls wasn't properly caching
a NULL acl when the inode didn't have any acls. This meant the common
case of no acls resulted in expensive btree searches every time the
kernel checked permissions (which is quite often).
This is a modified version of Linus' original patch:
Properly set initial acl fields to BTRFS_ACL_NOT_CACHED in the inode.
This forces an acl lookup when permission checks are done.
Fix btrfs_get_acl to avoid lookups and locking when the inode acls fields
are set to null.
Fix btrfs_get_acl to use the right return value from __btrfs_getxattr
when deciding to cache a NULL acl. It was storing a NULL acl when
__btrfs_getxattr return -ENOENT, but __btrfs_getxattr was actually returning
-ENODATA for this case.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Just happened to notice a bunch of %llu vs u64 warnings. Here's a patch
to cast them all.
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
A small warning popped up on ia64 because inode-map.c was comparing a
u64 object id with the ULL FIRST_FREE_OBJECTID. My first thought was
that all the OBJECTID constants should contain the u64 cast because
btrfs code deals entirely in u64s. But then I saw how large that was,
and figured I'd just fix the max() call.
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs has printks for various IO errors, including bad checksums and
mismatches between what we expect the block headers to contain and what
we actually find on the disk.
Longer term we need a real reporting mechanism for this, but for now
printk is going to have to do.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Previously, we updated a device's size prior to attempting a shrink
operation. This patch moves the device resizing logic to only happen if
the shrink completes successfully. In the process, it introduces a new
field to btrfs_device -- disk_total_bytes -- to track the on-disk size.
Signed-off-by: Chris Ball <cjb@laptop.org>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
After a transaction commit, the old root of the subvol btrees are sent through
snapshot removal. This is what actually frees up any blocks replaced by
COW, and anything the old blocks pointed to.
Snapshot deletion will pause when a transaction commit has started, which
helps to avoid a huge amount of delayed reference count updates piling up
as the transaction is trying to close.
But, this pause happens after the snapshot deletion process has asked other
procs on the system to throttle back a bit so that it can make progress.
We don't want to throttle everyone while we're waiting for the transaction
commit, it leads to deadlocks in the user transaction ioctls used by Ceph
and makes things slower in general.
This patch changes things to avoid the throttling while we sleep.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The btrfs fallocate call takes an extent lock on the entire range
being fallocated, and then runs through insert_reserved_extent on each
extent as they are allocated.
The problem with this is that btrfs_drop_extents may decide to try
and take the same extent lock fallocate was already holding. The solution
used here is to push down knowledge of the range that is already locked
going into btrfs_drop_extents.
It turns out that at least one other caller had the same bug.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Currently the extent_map code is only for btrfs so don't export it's
symbols.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Get rid of the hacks for building out of tree, and always use += for
assigning to the object lists.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch makes the chunk allocator keep a good ratio of metadata vs data
block groups. By default for every 8 data block groups, we'll allocate 1
metadata chunk, or about 12% of the disk will be allocated for metadata. This
can be changed by specifying the metadata_ratio mount option.
This is simply the number of data block groups that have to be allocated to
force a metadata chunk allocation. By making sure we allocate metadata chunks
more often, we are less likely to get into situations where the whole disk
has been allocated as data block groups.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs fallocate was incorrectly starting a transaction with a lock held
on the extent_io tree for the file, which could deadlock. Strictly
speaking it was using join_transaction which would be safe, but it is better
to move the transaction outside of the lock.
When preallocated extents are overwritten, btrfs_mark_buffer_dirty was
being called on an unlocked buffer. This was triggering an assertion and
oops because the lock is supposed to be held.
The bug was calling btrfs_mark_buffer_dirty on a leaf after btrfs_del_item had
been run. btrfs_del_item takes care of dirtying things, so the solution is a
to skip the btrfs_mark_buffer_dirty call in this case.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Remove open-coded memdup_user().
Note this changes some GFP_NOFS to GFP_KERNEL, since copy_from_user() may
cause pagefault, it's pointless to pass GFP_NOFS to kmalloc().
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
reada_for_balance was using the wrong index into the path node array,
so it wasn't reading the right blocks. We never directly used the
results of the read done by this function because the btree search is
started over at the end.
This fixes reada_for_balance to reada in the correct node and to
avoid searching past the last slot in the node. It also makes sure to
hold the parent lock while we are finding the nodes to read.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The extent_io writepage call updates the writepage index in the inode
as it makes progress. But, it was doing the update after unlocking the page,
which isn't legal because page->mapping can't be trusted once the page
is unlocked.
This lead to an oops, especially common with compression turned on. The
fix here is to update the writeback index before unlocking the page.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs is using WRITE_SYNC_PLUG to send down synchronous IOs with a
higher priority. But, the checksumming helper threads prevent it
from being fully effective.
There are two problems. First, a big queue of pending checksumming
will delay the synchronous IO behind other lower priority writes. Second,
the checksumming uses an ordered async work queue. The ordering makes sure
that IOs are sent to the block layer in the same order they are sent
to the checksumming threads. Usually this gives us less seeky IO.
But, when we start mixing IO priorities, the lower priority IO can delay
the higher priority IO.
This patch solves both problems by adding a high priority list to the async
helper threads, and a new btrfs_set_work_high_prio(), which is used
to make put a new async work item onto the higher priority list.
The ordering is still done on high priority IO, but all of the high
priority bios are ordered separately from the low priority bios. This
ordering is purely an IO optimization, it is not involved in data
or metadata integrity.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Part of reducing fsync/O_SYNC/O_DIRECT latencies is using WRITE_SYNC for
writes we plan on waiting on in the near future. This patch
mirrors recent changes in other filesystems and the generic code to
use WRITE_SYNC when WB_SYNC_ALL is passed and to use WRITE_SYNC for
other latency critical writes.
Btrfs uses async worker threads for checksumming before the write is done,
and then again to actually submit the bios. The bio submission code just
runs a per-device list of bios that need to be sent down the pipe.
This list is split into low priority and high priority lists so the
WRITE_SYNC IO happens first.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Remove an unneeded return statement and conditional
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
We get this on 32 builds:
fs/built-in.o: In function `extent_fiemap':
(.text+0x1019f2): undefined reference to `__ucmpdi2'
Happens because of a switch statement with a 64 bit argument.
Convert this to an if statement to fix this.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_new_inode doesn't call iput to free the inode
when it fails.
Signed-off-by: Shen Feng <shen@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Need to check kthread_should_stop after schedule_timeout() before calling
schedule(). This causes threads to sleep with potentially no one to wake them
up causing mount(2) to hang in btrfs_stop_workers waiting for threads to stop.
Signed-off-by: Amit Gud <gud@ksu.edu>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The 'flushoncommit' mount option forces any data dirtied by a write in a
prior transaction to commit as part of the current commit. This makes
the committed state a fully consistent view of the file system from the
application's perspective (i.e., it includes all completed file system
operations). This was previously the behavior only when a snapshot is
created.
This is used by Ceph to ensure that completed writes make it to the
platter along with the metadata operations they are bound to (by
BTRFS_IOC_TRANS_{START,END}).
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Add a 'notreelog' mount option to disable the tree log (used by fsync,
O_SYNC writes). This is much slower, but the tree logging produces
inconsistent views into the FS for ceph.
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs options can change at times other than mount, yet /proc/mounts shows the
options string used when the fs was mounted (an example would be when btrfs
determines that barriers aren't useful and turns them off.) This patch
instead outputs the actual options in use by btrfs.
Signed-off-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Because btrfs is copy-on-write, we end up picking new locations for
blocks very often. This makes it fairly difficult to maintain perfect
read patterns over time, but we can at least do some optimizations
for writes.
This is done today by remembering the last place we allocated and
trying to find a free space hole big enough to hold more than just one
allocation. The end result is that we tend to write sequentially to
the drive.
This happens all the time for metadata and it happens for data
when mounted -o ssd. But, the way we record it is fairly racey
and it tends to fragment the free space over time because we are trying
to allocate fairly large areas at once.
This commit gets rid of the races by adding a free space cluster object
with dedicated locking to make sure that only one process at a time
is out replacing the cluster.
The free space fragmentation is somewhat solved by allowing a cluster
to be comprised of smaller free space extents. This part definitely
adds some CPU time to the cluster allocations, but it allows the allocator
to consume the small holes left behind by cow.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_next_leaf was using blocking locks when it could have been using
faster spinning ones instead. This adds a few extra checks around
the pieces that block and switches over to spinning locks.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch removes the pinned_mutex. The extent io map has an internal tree
lock that protects the tree itself, and since we only copy the extent io map
when we are committing the transaction we don't need it there. We also don't
need it when caching the block group since searching through the tree is also
protected by the internal map spin lock.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
This patch removes the block group alloc mutex used to protect the free space
tree for allocations and replaces it with a spin lock which is used only to
protect the free space rb tree. This means we only take the lock when we are
directly manipulating the tree, which makes us a touch faster with
multi-threaded workloads.
This patch also gets rid of btrfs_find_free_space and replaces it with
btrfs_find_space_for_alloc, which takes the number of bytes you want to
allocate, and empty_size, which is used to indicate how much free space should
be at the end of the allocation.
It will return an offset for the allocator to use. If we don't end up using it
we _must_ call btrfs_add_free_space to put it back. This is the tradeoff to
kill the alloc_mutex, since we need to make sure nobody else comes along and
takes our space.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
I've replaced the strange looping constructs with a list_for_each_entry on
space_info->block_groups. If we have a hint we just jump into the loop with
the block group and start looking for space. If we don't find anything we
start at the beginning and start looking. We never come out of the loop with a
ref on the block_group _unless_ we found space to use, then we drop it after we
set the trans block_group.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
This patch cleans up the free space cache code a bit. It better documents the
idiosyncrasies of tree_search_offset and makes the code make a bit more sense.
I took out the info allocation at the start of __btrfs_add_free_space and put it
where it makes more sense. This was left over cruft from when alloc_mutex
existed. Also all of the re-searches we do to make sure we inserted properly.
Signed-off-by: Josef Bacik <jbacik@redhat.com>
Btrfs pages being written get set to writeback, and then may go through
a number of steps before they hit the block layer. This includes compression,
checksumming and async bio submission.
The end result is that someone who writes a page and then does
wait_on_page_writeback is likely to unplug the queue before the bio they
cared about got there.
We could fix this by marking bios sync, or by doing more frequent unplugs,
but this commit just changes the async bio submission code to unplug
after it has processed all the bios for a device. The async bio submission
does a fair job of collection bios, so this shouldn't be a huge problem
for reducing merging at the elevator.
For streaming O_DIRECT writes on a 5 drive array, it boosts performance
from 386MB/s to 460MB/s.
Thanks to Hisashi Hifumi for helping with this work.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Btrfs uses async helper threads to submit write bios so the checksumming
helper threads don't block on the disk.
The submit bio threads may process bios for more than one block device,
so when they find one device congested they try to move on to other
devices instead of blocking in get_request_wait for one device.
This does a pretty good job of keeping multiple devices busy, but the
congested flag has a number of problems. A congested device may still
give you a request, and other procs that aren't backing off the congested
device may starve you out.
This commit uses the io_context stored in current to decide if our process
has been made a batching process by the block layer. If so, it keeps
sending IO down for at least one batch. This helps make sure we do
a good amount of work each time we visit a bdev, and avoids large IO
stalls in multi-device workloads.
It's also very ugly. A better solution is in the works with Jens Axboe.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
page_mkwrite is called with neither the page lock nor the ptl held. This
means a page can be concurrently truncated or invalidated out from
underneath it. Callers are supposed to prevent truncate races themselves,
however previously the only thing they can do in case they hit one is to
raise a SIGBUS. A sigbus is wrong for the case that the page has been
invalidated or truncated within i_size (eg. hole punched). Callers may
also have to perform memory allocations in this path, where again, SIGBUS
would be wrong.
The previous patch ("mm: page_mkwrite change prototype to match fault")
made it possible to properly specify errors. Convert the generic buffer.c
code and btrfs to return sane error values (in the case of page removed
from pagecache, VM_FAULT_NOPAGE will cause the fault handler to exit
without doing anything, and the fault will be retried properly).
This fixes core code, and converts btrfs as a template/example. All other
filesystems defining their own page_mkwrite should be fixed in a similar
manner.
Acked-by: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Chris Mason