This adds support for nocow mode, where we do writes in-place when
possible. Patch components:
- New boolean filesystem and inode option, nocow: note that when nocow
is enabled, data checksumming and compression are implicitly disabled
- To prevent in-place writes from racing with data moves
(data_update.c) or bucket reuse (i.e. a bucket being reused and
re-allocated while a nocow write is in flight, we have a new locking
mechanism.
Buckets can be locked for either data update or data move, using a
fixed size hash table of two_state_shared locks. We don't have any
chaining, meaning updates and moves to different buckets that hash to
the same lock will wait unnecessarily - we'll want to watch for this
becoming an issue.
- The allocator path also needs to check for in-place writes in flight
to a given bucket before giving it out: thus we add another counter
to bucket_alloc_state so we can track this.
- Fsync now may need to issue cache flushes to block devices instead of
flushing the journal. We add a device bitmask to bch_inode_info,
ei_devs_need_flush, which tracks devices that need to have flushes
issued - note that this will lead to unnecessary flushes when other
codepaths have already issued flushes, we may want to replace this with
a sequence number.
- New nocow write path: look up extents, and if they're writable write
to them - otherwise fall back to the normal COW write path.
XXX: switch to sequence numbers instead of bitmask for devs needing
journal flush
XXX: ei_quota_lock being a mutex means bch2_nocow_write_done() needs to
run in process context - see if we can improve this
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The data update path requires special support for unwritten extents - we
still need to be able to move them, but there's no need to read or write
anything.
This patch adds a new error code to tell bch2_move_extent() that we're
short circuiting the read, and adds bch2_update_unwritten_extent() to
create a reservation then call __bch2_data_update_index_update().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
- bch2_extent_merge checks unwritten bit
- read path returns 0s for unwritten extents without actually reading
- reflink path skips over unwritten extents
- bch2_bkey_ptrs_invalid() checks for extents with both written and
unwritten extents, and non-normal extents (stripes, btree ptrs) with
unwritten ptrs
- fiemap checks for unwritten extents and returns
FIEMAP_EXTENT_UNWRITTEN
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In the io path, when we do the extent update we also have to update the
inode - for i_size and i_sectors updates, as well as for bi_journal_seq
for fsync.
This factors that out into a new helper which will be used in the new
nocow mode, in the unwritten extent conversion path.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This factors out part of __bchfs_fallocate() in fs-io.c into an new,
lower level io.c helper, which creates a single extent reservation.
This is prep work for nocow support - the new helper will shortly gain
the ability to create unwritten extents.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This takes advantage of the new inode type to skip the expensive
pack/unpack when inode updates are required in the extent update path.
Additionally, we now skip the inode update entirely when i_sectors and
i_size aren't changing.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The extend update path had an optimization to avoid updating the inode
if we knew we were definitely not extending the file. But now that we're
updating inodes on every extent update - for fsync - that code can be
deleted.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Move bi_size and bi_sectors into the non-varint portion of the inode, so
that the write path can update them without going through the relatively
expensive unpack/pack operations.
Other changes:
- Add a field for the offset of the varint section, so we can add new
non-varint fields without needing a new inode type, like alloc_v3
- Move bi_mode into the flags field, so that the varint section can be
u64 aligned
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bch2_gc may require snapshots to be started - the repair path when
checking the reflink btree may do updates to the extents btree.
This moves bch2_fs_initialize_subvolumes() and bch2_fs_snapshots_start()
to before bch2_gc() - since we haven't gone RW yet, the updates in
bch2_fs_initialize_subvolumes() are done via the journal replay keys
list, so it's fine to do this before bch2_gc().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This factors out a new helper from bch2_dev_freespace_init(),
bch2_get_key_or_hole(), and uses it in bch2_check_alloc_info(): we're
now able to process holes in the alloc btree as ranges, instead of one
bucket at a time.
This will improve fsck performance on new filesystems, or filesystems
where not every bucket has been used yet.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This makes bch2_dev_freespace_init() much faster: instead of processing
every bucket on the device one at a time, we handle ranges of missing
keys all at once: the freespace btree is an extents style btree, so we
only have to insert one freespace key for every range of missing keys
in the alloc btree.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The btree key cache mainly helps with lock contention, at the cost of
additional memory overhead. During some fsck passes the memory overhead
really matters, but fsck is single threaded so lock contention is an
issue - so skipping the key cache during fsck will help with
performance.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Similer to the previous patch for check_backpointers_to_extents(), if
the alloc + backpointers btrees do not fit in ram we need to run into
multiple passes.
The counting of btree nodes that fit in memory is different here,
because we have to walk the alloc and backpointers btrees at the same
time, since a backpointer could reside in either of them and we don't
know which without checking both.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
When the extents + reflink btrees don't fit into memory this fsck pass
becomes _much_ slower, since we're doing random lookups.
This patch changes this pass to check how much of the relevant btrees
will fit into memory, and run in multiple passes if needed.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
With the new backpointer based copygc we don't need an explicit copygc
reserve, we're always evacuating buckets one at a time - so this is no
longer needed, and in fact removing it fixes a deadlock in
bch2_dev_allocator_remove().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This is only a start to updating erasure coding for backpointers - it's
still not working yet. The subsequent patch will delete our old in
memory backpointers for copygc, and this fixes a spurious EPERM
bug/error message.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Previously, copygc needed to walk the entire extents & reflink btrees to
find extents that needed to be moved.
Now that we have backpointers, this patch implements
bch2_evacuate_bucket() in the move code, which copygc now uses for
evacuating mostly empty buckets.
Also, thanks to the new backpointers code, copygc can now move btree
nodes.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This patch adds backpointers: we now have a reverse index from device
and offset on that device (specifically, offset within a bucket) back to
btree nodes and (non cached) data extents.
The first 40 backpointers within a bucket are stored in the alloc key;
after that backpointers spill over to the next backpointers btree. This
is to help avoid performance regressions from additional btree updates
on large streaming workloads.
This patch adds all the code for creating, checking and repairing
backpointers. The next patch in the series is going to use backpointers
for copygc - finally getting rid of the need to scan all extents to do
copygc.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds a new method of doing btree updates - a straight write buffer,
implemented as a flat fixed size array.
This is only useful when we don't need to read from the btree in order
to do the update, and when reading is infrequent - perfect for the LRU
btree.
This will make LRU btree updates fast enough that we'll be able to use
it for persistently indexing buckets by fragmentation, which will be a
massive boost to copygc performance.
Changes:
- A new btree_insert_type enum, for btree_insert_entries. Specifies
btree, btree key cache, or btree write buffer.
- bch2_trans_update_buffered(): updates via the btree write buffer
don't need a btree path, so we need a new update path.
- Transaction commit path changes:
The update to the btree write buffer both mutates global, and can
fail if there isn't currently room. Therefore we do all write buffer
updates in the transaction all at once, and also if it fails we have
to revert filesystem usage counter changes.
If there isn't room we flush the write buffer in the transaction
commit error path and retry.
- A new persistent option, for specifying the number of entries in the
write buffer.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
It's possible to do btree updates before going RW by adding them to the
list of updates for journal replay to do, but this is limited by what
fits in RAM. This patch switches the second alloc info phase to run
after going RW - btree_gc has already ensured the alloc btree itself is
correct - and tweaks the allocation path to deal with the potential
small inconsistencies.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In the distant past, it wasn't possible to start copygc until after
journal replay had finished. Now, the btree iterator code overlays keys
from the journal, so there's no reason not to start it earlier - and it
solves a rare deadlock.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Recursive transaction commits are occasionally necessary - in
particular, for the upcoming btree write buffer's flush path.
This avoids bugs due to trans->flags being accidentally mutated
mid-commit, which can cause c->writes refcount leaks.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
When we unlock in order to submit IO, the next relock event is likely to
fail if submit_bio() blocked - we shouldn't those events in our _fail
stats, since those are expected events and shouldn't cause test
failures.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds a debug mode where we split up the c->writes refcount into
distinct refcounts for every codepath that takes a reference, and adds
sysfs code to print the value of each ref.
This will make it easier to debug shutdown hangs due to refcount leaks.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The btree_node_write_blocked bit was a later addition to this code,
it only mirrors the state of the b->write_blocked list (empty or
nonempty) - unfortunately, when it was added it wasn't correctly kept in
sync - oops.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We need to call bch2_trans_update_max_paths() before marking the new
path as allocated, since we're not initializing it yet.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
It's important that in BTREE_ITER_FILTER_SNAPSHOTS mode we always use
peek_upto() and provide an end for the interval we're searching for -
otherwise, when we hit the end of the inode the next inode be in a
different subvolume and not have any keys in the current snapshot, and
we'd iterate over arbitrarily many keys before returning one.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a (harmless) lockdep splat, due to a lock order violation in
the key cache exit path.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds a threshold for the maximum spin time, similar to the rwsem
code, and a flag to the lock itself indicating when we've spun too long
so other threads also refrain from spinning.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This uses the new _ip() interface to six locks and hooks it up to
btree_path->ip_allocated, when available.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds _ip variations of the various lock functions that allow an IP
to be passed in, which is used by lockstat.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This replaces various BUG_ON() assertions with panics that tell us where
the restart was done and the restart type.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This ensures that failure to read a btree node error is treated as a
topology error, and returns the correct error so that the topology
repair pass will be run.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
On hash collision when we have to check for duplicates or incorrect
hash value, we weren't specifying a snapshot ID to iterate with.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Easy workaround for a lockdep splat - and since bch2_prt_backtrace() is
only used in debug code this is fine.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This improves io_opts() and makes it a non-inline function - it's big
enough that it probably shouldn't be.
Also, bch_io_opts no longer needs fields for whether options are
defined, so we can slim it down a bit.
We'd like to stop passing around the full bch_io_opts, but that'll be
tricky because of bch2_rebalance_add_key().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Provide inline versions of some allocation functions
- bch2_alloc_sectors_done_inlined()
- bch2_alloc_sectors_append_ptrs_inlined()
and use them in the core IO path.
Also, inline bch2_extent_update_i_size_sectors() and
bch2_bkey_append_ptr().
In the core write path, function call overhead matters - every function
call is a jump to a new location and a potential cache miss.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This separates out the slowpath into a separate function, and inlines
bch2_alloc_v4_mut into bch2_trans_start_alloc_update(), the main place
it's called.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
