Add a tracepoint for downcasting private errors to standard errors, so
they can be recovered even when not logged; also, add some
documentation.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs btree nodes are big - typically 256k - and btree roots are
pinned in memory. As we're now up to 18 btrees, we now have significant
memory overhead in mostly empty btree roots.
And in the future we're going to start enforcing that certain btree node
boundaries exist, to solve lock contention issues - analagous to XFS's
AGIs.
Thus, we need to start allocating smaller btree node buffers when we
can. This patch changes code that refers to the filesystem constant
c->opts.btree_node_size to refer to the btree node buffer size -
btree_buf_bytes() - where appropriate.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Factor out bch2_journal_bufs_to_text(), and use it in the
journal_entry_full() tracepoint; when we can't get a journal reservation
we need to know the outstanding journal entry sizes to know if the
problem is due to excessive flushing.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
These were for extra info in tracepoints for debugging a specialized
issue - we do not want to bloat btree_path for this, at least in release
builds.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In the CI, we're seeing tests failing due to excessive would_deadlock
transaction restarts - the tracepoint now includes the lock cycle that
occured.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
- add a tracepoint for write_buffer_flush_sync; this is expensive
- fix the write_buffer_flush_slowpath tracepoint
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Add a tracepoint for rebalance, printing out
- the target option
- the compression option
- the key being rebalanced
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Renamed from trace_move_extent_alloc_mem_fail, because there are other
reasons we colud fail (disk space allocation failure).
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This deletes the complicated and somewhat expensive journal
pre-reservation machinery in favor of just using journal watermarks:
when the journal is more than half full, we run journal reclaim more
aggressively, and when the journal is more than 3/4s full we only allow
journal reclaim to get new journal reservations.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We should only be downgrading locks on success - otherwise, our
transaction restarts won't be getting the correct locks and we'll
livelock.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
data_progress_list is gone - it was redundant with moving_context_list
The upcoming rebalance rewrite is going to have it using two different
move_stats objects with the same moving_context, depending on whether
it's scanning or using the rebalance_work btree - this patch plumbs
stats around a bit differently so that will work.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Since we can run with unknown btree IDs, we can't directly index btree
IDs into fixed size arrays.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In https://github.com/koverstreet/bcachefs/issues/450, we're seeing
unexplained btree_path_relock_fail events - according to the information
currently in the tracepoint, it appears the relock should be succeeding.
This adds lock counts to the tracepoint to help track it down.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Using drop_locks_do() ensures that every unlock() is paired with a
relock(), with proper error checking.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
As suggested by Linus, this drops the six_lock_state union in favor of
raw bitmasks.
On the one hand, bitfields give more type-level structure to the code.
However, a significant amount of the code was working with
six_lock_state as a u64/atomic64_t, and the conversions from the
bitfields to the u64 were deemed a bit too out-there.
More significantly, because bitfield order is poorly defined (#ifdef
__LITTLE_ENDIAN_BITFIELD can be used, but is gross), incrementing the
sequence number would overflow into the rest of the bitfield if the
compiler didn't put the sequence number at the high end of the word.
The new code is a bit saner when we're on an architecture without real
atomic64_t support - all accesses to lock->state now go through
atomic64_*() operations.
On architectures with real atomic64_t support, we additionally use
atomic bit ops for setting/clearing individual bits.
Text size: 7467 bytes -> 4649 bytes - compilers still suck at
bitfields.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Move path tracepoints now include the key being moved. Also, add new
tracepoints for the start of move_extent, and evacuate_bucket.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This greatly expands the move_extent_fail tracepoint - now it includes
all the information we have available, including exactly why the extent
wasn't updated.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Seeing occasional test failures where we get stuck in a livelock that
involves this event - this will help track it down.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now, any open_bucket can go on the partial list: allocating from the
partial list has been moved to its own dedicated function,
open_bucket_add_bucets() -> bucket_alloc_set_partial().
In particular, this means that erasure coded buckets can safely go on
the partial list; the new location works with the "allocate an ec bucket
first, then the rest" logic.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Print bucket in dev:bucket notation, to be consistent with how we refer
to buckets elsewhere.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now that we have much more efficient updates to the LRU btree, this
patch adds a new LRU that indexes buckets by fragmentation.
This means copygc no longer has to scan every bucket to find buckets
that need to be evacuated.
Changes:
- A new field in bch_alloc_v4, fragmentation_lru - this corresponds to
the bucket's position in the fragmentation LRU. We add a new field
for this instead of calculating it as needed because we may make the
fragmentation LRU optional; this field indicates whether a bucket is
on the fragmentation LRU.
Also, zoned devices will introduce variable bucket sizes; explicitly
recording the LRU position will be safer for them.
- A new copygc path for using the fragmentation LRU instead of
scanning every bucket and building up an in-memory heap.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
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>
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 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>
This refactoring puts our various allocation path counters into a
dedicated struct - the upcoming nocow patch is going to add another
counter.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
It now includes more info - whether the bucket was for metadata or data
- and also call it in the same place as the bucket_alloc_fail
tracepoint.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We've outgrown our own deadlock avoidance strategy.
The btree iterator API provides an interface where the user doesn't need
to concern themselves with lock ordering - different btree iterators can
be traversed in any order. Without special care, this will lead to
deadlocks.
Our previous strategy was to define a lock ordering internally, and
whenever we attempt to take a lock and trylock() fails, we'd check if
the current btree transaction is holding any locks that cause a lock
ordering violation. If so, we'd issue a transaction restart, and then
bch2_trans_begin() would re-traverse all previously used iterators, but
in the correct order.
That approach had some issues, though.
- Sometimes we'd issue transaction restarts unnecessarily, when no
deadlock would have actually occured. Lock ordering restarts have
become our primary cause of transaction restarts, on some workloads
totally 20% of actual transaction commits.
- To avoid deadlock or livelock, we'd often have to take intent locks
when we only wanted a read lock: with the lock ordering approach, it
is actually illegal to hold _any_ read lock while blocking on an intent
lock, and this has been causing us unnecessary lock contention.
- It was getting fragile - the various lock ordering rules are not
trivial, and we'd been seeing occasional livelock issues related to
this machinery.
So, since bcachefs is already a relational database masquerading as a
filesystem, we're stealing the next traditional database technique and
switching to a cycle detector for avoiding deadlocks.
When we block taking a btree lock, after adding ourself to the waitlist
but before sleeping, we do a DFS of btree transactions waiting on other
btree transactions, starting with the current transaction and walking
our held locks, and transactions blocking on our held locks.
If we find a cycle, we emit a transaction restart. Occasionally (e.g.
the btree split path) we can not allow the lock() operation to fail, so
if necessary we'll tell another transaction that it has to fail.
Result: trans_restart_would_deadlock events are reduced by a factor of
10 to 100, and we'll be able to delete a whole bunch of grotty, fragile
code.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Centralizing the transaction restart/tracepoint in
bch2_btree_path_upgrade() lets us improve the tracepoint - now it emits
old and new locks_want.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Also, do some reorganizing/renaming, convert atomic counters in bch_fs
to persistent counters, and add a few missing counters.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
It now prints the error name when the btree node is an error pointer;
also, don't trace failures when the the btree node is
BCH_ERR_no_btree_node_up.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six_lock_count now counts up whether a write lock held, and this patch
now also correctly counts six_lock->intent_lock_recurse.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Our types are exported to the tracepoint code, so it's not necessary to
break things out individually when passing them to tracepoints - we can
also call other functions from TP_fast_assign().
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
- use strlcpy(), not strncpy()
- add tracepoints for btree_path alloc and free
- give the tracepoint for key cache upgrade fail a proper name
- add a tracepoint for btree_node_upgrade_fail
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
In CONFIG_BCACHEFS_DEBUG mode, we'll now randomly issue transaction
restarts - with a decaying probability based on the number of restarts
we've already had, to ensure that transactions eventually make forward
progress. This should help shake out some bugs.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
