We need to allow filesystems with metadata from newer versions to be
mountable and usable by older versions.
This patch enables us to roll out new btrees without a new major version
number; we can now handle btree roots for unknown btree types.
The unknown btree roots will be retained, and fsck (including
backpointers) will check them, the same as other btree types.
We add a dynamic array for the extra, unknown btree roots, in addition
to the fixed size btree root array, and add new helpers for looking up
btree roots.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
A crash immediately after device removal can result in an
unmountable filesystem due to recovery failure. The following
command reliably reproduces on a multi-device fs:
bcachefs device remove <dev> && xfs_io -xc shutdown <mnt>
The post-crash mount fails with an error similar to the following,
reported by fsck:
invalid journal entry dev_usage at offset 7994/8034 seq 12: bad dev, fixing
This refers to a device usage entry in the journal that refers to
the index of the just removed device. Recovery considers this an
invalid entry and fails to proceed.
Device usage entries are added to journal buffer writes via
bch_journal_write() -> bch2_journal_super_entries_add_common(),
which means any journal buffer write has content that refers to
member devices at the time of the journal write.
The device remove sequence already removes metadata references to
the device being removed. It then flushes any pins that refer to the
device, clears replica entries, removes the in-memory device object
and lastly updates the superblock to reflect that the device is no
longer present. The problem is that any journal writes that occur
during this sequence will include a dev usage entry so long as the
device is present. To avoid this problem, we can flush the journal
once more after the device entry is removed from the in-core
structures, but before the superblock is updated to fully remove the
device on-disk.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a null ptr deref in bch2_free_pending_node_rewrites() when
the list head wasn't initialized.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
As with previous conversions, replace -ENOENT uses with more informative
private error codes.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
If a device has keys in the bucket_gens btree associated with its
buckets and is removed from a bcachefs volume, fsck will complain
about the presence of keys associated with an invalid device index.
A repair removes the associated keys and restores correctness.
Update bch2_dev_remove_alloc() to remove device related keys at
device removal time to avoid the problem.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Like in the recovery, and device add, we have to check if devices don't
have the freespace btree initialized - this was missed in the device hot
add path.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
A workqueue resource deadlock has been observed when running fsck
on a filesystem with a full/stuck journal. fsck is not currently
able to repair the fs due to fairly rapid emergency shutdown, but
rather than exit gracefully the fsck process hangs during the
shutdown sequence. Fortunately this is easily recoverable from
userspace, but the root cause involves code shared between the
kernel and userspace and so should be addressed.
The deadlock scenario involves the main task in the bch2_fs_stop()
-> bch2_fs_read_only() path waiting on write references to drain
with the fs state lock held. A bch2_read_only_work() workqueue task
is scheduled on the system_long_wq, blocked on the state lock.
Finally, various other write ref holding workqueue tasks are
scheduled to run on the same workqueue and must complete in order to
release references that the initial task is waiting on.
To avoid this problem, we can split the dependent workqueue tasks
across different workqueues. It's a bit of a waste to create a
dedicated wq for the read-only worker, but there are several tasks
throughout the fs that follow the pattern of acquiring a write
reference and then scheduling to the system wq. Use a local wq
for such tasks to break the subtle dependency between these and the
read-only worker.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a bug in bch2_evict_subvolume_inodes(): d_mark_dontcache()
doesn't handle the case where i_count is already 0, we need to grab and
put the inode in order for it to be dropped.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This implements a new shutdown path for erasure coding, which is needed
for the upcoming BCH_WRITE_WAIT_FOR_EC write path.
The process is:
- Cancel new stripes being built up
- Close out/cancel open buckets on write points or the partial list
that are for stripes
- Shutdown rebalance/copygc
- Then wait for in flight new stripes to finish
With BCH_WRITE_WAIT_FOR_EC, move ops will be waiting on stripes to fill
up before they complete; the new ec shutdown path is needed for shutting
down copygc/rebalance without deadlocking.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This also adds bch2_write_op_to_text(): now we can see outstand moves,
useful for debugging shutdown with the upcoming BCH_WRITE_WAIT_FOR_EC
and likely for other things in the future.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds private error codes for most (but not all) of our ENOMEM uses,
which makes it easier to track down assorted allocation failures.
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 improves the nocow lock table so that hash table entries have
multiple locks, and locks specify which bucket they're for - i.e. we can
now resolve hash collisions.
This is important because the allocator has to skip buckets that are
locked in the nocow lock table, and previously hash collisions would
cause it to spuriously skip unlocked buckets.
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 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>
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>
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>
We shouldn't be overloading standard error codes now that we have
provisions for bcachefs-specific errorcodes: this patch converts super.c
and super-io.c to per error site errcodes, with a bit of cleanup.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
On startup, we need to ensure the first journal entry written is a flush
write: after a clean shutdown we generally don't read the journal, which
means we might be overwriting whatever was there previously, and there
must always be at least one flush entry in the journal or recovery will
fail.
Found by fstests generic/388.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
checkpatch.pl gives lots of warnings that we don't want - suggested
ignore list:
ASSIGN_IN_IF
UNSPECIFIED_INT - bcachefs coding style prefers single token type names
NEW_TYPEDEFS - typedefs are occasionally good
FUNCTION_ARGUMENTS - we prefer to look at functions in .c files
(hopefully with docbook documentation), not .h
file prototypes
MULTISTATEMENT_MACRO_USE_DO_WHILE
- we have _many_ x-macros and other macros where
we can't do this
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We need counters to be initialized before initializing shrinkers - the
shrinker callbacks will update those counters. This fixes a segfault in
userspace.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Continuing the saga of introducing private dedicated error codes for
each error path, this patch converts ENOSPC to error codes that are
subtypes of ENOSPC. We've recently had a test failure where we got
-ENOSPC where we shouldn't have, and didn't have enough information to
tell where it came from, so this patch will solve that problem.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Device labels are represented as pointers in the member info section: we
need to get and then set the label for it to be kept correctly.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Like bch2_do_discards(), we should check if this needs to be done when
going rw.
Also, add some sysfs code for debugging bucket invalidation.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This converts bcachefs to the modern printbuf interface/implementation,
synced with the version to be submitted upstream.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds a new superblock field for persisting counters
and adds a sysfs interface in counters/ exposing these counters.
The superblock field is ignored by older versions letting us avoid
an on disk version bump.
Each sysfs file outputs a counter that tracks since filesystem
creation and a counter for the current mount session.
Signed-off-by: Daniel Hill <daniel@gluo.nz>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We're seeing occasional firings of the assertion in the key cache
shutdown code that nr_dirty == 0, which means we must sometimes be doing
transaction commits after we've gone read only.
Cleanups & changes:
- BCH_FS_ALLOC_CLEAN renamed to BCH_FS_CLEAN_SHUTDOWN
- new helper bch2_btree_interior_updates_flush(), which returns true if
it had to wait
- bch2_btree_flush_writes() now also returns true if there were btree
writes in flight
- __bch2_fs_read_only now checks if btree writes were in flight in the
shutdown loop: btree write completion does a transaction update, to
update the pointer in the parent node
- assert that !BCH_FS_CLEAN_SHUTDOWN in __bch2_trans_commit
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
- We weren't clearing the LRU btree
- bch2_alloc_read() runs before bch2_check_alloc_key() deletes alloc
keys for devices/buckets that don't exists, so it needs to check for
that
- bch2_check_lrus() needs to check that buckets exists
- improve some error messages
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
We need to ensure that work structs in bch_fs always get initialized -
otherwise an error in filesystem initialization can pop a warning in the
workqueue code when we try to cancel a work struct that wasn't
initialized.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Previously, the journal read path used a linked list for storing the
journal entries we read from disk. But there's been a bug that's been
causing journal_flush_delay to incorrectly be set to 0, leading to far
more journal entries than is normal being written out, which then means
filesystems are no longer able to start due to the O(n^2) behaviour of
inserting into/searching that linked list.
Fix this by switching to a radix tree.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Previously, we were missing accounting for buckets in need_gc_gens and
need_discard states. This matters because buckets in those states need
other btree operations done before they can be used, so they can't be
conuted when checking current number of free buckets against the
allocation watermark.
Also, we weren't directly counting free buckets at all. Now, data type 0
== BCH_DATA_free, and free buckets are counted; this means we can get
rid of the separate (poorly defined) count of unavailable buckets.
This is a new on disk format version, with upgrade and fsck required for
the accounting changes.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
- Move checks for whether the device & bucket are valid from the
.key_invalid method to bch2_check_alloc_key(). This is because
.key_invalid() is called on keys that may no longer exist (post
journal replay), which is a problem when removing/resizing devices.
- We weren't checking the need_discard btree to ensure that every set
bucket has a corresponding alloc key. This refactors the code for
checking the freespace btree, so that it now checks both.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
In the old allocator code, buckets would be discarded just prior to
being used - this made sense in bcache where we were discarding buckets
just after invalidating the cached data they contain, but in a
filesystem where we typically have more free space we want to be
discarding buckets when they become empty.
This patch implements the new behaviour - it checks the need_discard
btree for buckets awaiting discards, and then clears the appropriate
bit in the alloc btree, which moves the buckets to the freespace btree.
Additionally, discards are now enabled by default.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now that we have new persistent data structures for the allocator, this
patch converts the allocator to use them.
Now, foreground bucket allocation uses the freespace btree to find
buckets to allocate, instead of popping buckets off the freelist.
The background allocator threads are no longer needed and are deleted,
as well as the allocator freelists. Now we only need background tasks
for invalidating buckets containing cached data (when we are low on
empty buckets), and for issuing discards.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds two new btrees for the upcoming allocator rewrite: an extents
btree of free buckets, and a btree for buckets awaiting discards.
We also add a new trigger for alloc keys to keep the new btrees up to
date, and a compatibility path to initialize them on existing
filesystems.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In the write path, after the write to the block device(s) complete we
have to punt to process context to do the btree update.
Instead of using the work item embedded in op->cl, this patch switches
to a per write-point work item. This helps with two different issues:
- lock contention: btree updates to the same writepoint will (usually)
be updating the same alloc keys
- context switch overhead: when we're bottlenecked on btree updates,
having a thread (running out of a work item) checking the write point
for completed ops is cheaper than queueing up a new work item and
waking up a kworker.
In an arbitrary benchmark, 4k random writes with fio running inside a
VM, this patch resulted in a 10% improvement in total iops.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now we've got strings for metadata versions - this changes
bch2_sb_to_text() and our mount log message to use it.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Options no longer have to be manually added to bch2_sb_to_text() - it
now uses the master list of options in opts.h. Also, improve some of the
formatting by converting it to tabstops.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We were accidentally using default mount options and overwriting the
discard flag.
Signed-off-by: Daniel Hill <daniel@gluo.nz>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This patch changes printbufs dynamically allocate and reallocate a
buffer as needed. Stack usage has become a bit of a problem, and a major
cause of that has been static size string buffers on the stack.
The most involved part of this refactoring is that printbufs must now be
exited with printbuf_exit().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This patch was originally to work around the journal geting stuck in
nochanges mode - but that was just a hack, we needed to fix the actual
bug. It should be fixed now, so revert it.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
bch2_dev_lookup() is used from the extended attribute set methods, for
setting the target options, where we're already holding an inode lock -
it turns out pathname lookups also take inode locks, so that was
susceptible to deadlocks.
Fortunately we already stash the device name in ca->name. This does
change user-visible behaviour though: instead of specifying e.g.
/dev/sda1, user must now specify sda1.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Like the previous patches, this converts bch2_gc_gens() to use the alloc
btree directly, and private arrays of generation numbers for its own
recalculation of oldest_gen.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Add a new helper that returns true if the given btree ID uses the btree
key cache. This enables some new cleanups, since the helper can check
the options for whether caching is enabled on a given btree.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Implement a hash table, using cuckoo hashing, for empty buckets that are
waiting on a journal commit before they can be reused.
This replaces the journal_seq field of bucket_mark, and is part of
eventually getting rid of the in memory bucket array.
We may need to make bch2_bucket_needs_journal_commit() lockless, pending
profiling and testing.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This reverts commit f95b61228efd04c9c158123da5827c96e9773b29.
It turns out, we're seeing filesystems in the wild end up with
blacklisted btree node bsets - this should not be happening, and until
we understand why and fix it we need to keep this code around.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
- Add a shim uuid_unparse_lower() in the kernel, since %pU doesn't work
in userspace
- We don't need to print the bcachefs: or the filesystem name prefix in
userspace
- Improve a few error messages
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This patch converts bch2_sb_validate() and the .validate methods for the
various superblock sections to take printbuf, to which they can print
detailed error messages, including printing the entire section that was
invalid.
This is a great improvement over the previous situation, where we could
only return static strings that didn't have precise information about
what was wrong.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Add a field to bch_dev for the dev_t of the underlying block device -
this fixes a null ptr deref in tracepoints.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This converts the error messages in the device add to a better style,
and adds some missing ones.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Since metadata version bcachefs_metadata_version_btree_ptr_sectors_written,
we haven't needed the journal seq blacklist mechanism for ignoring
blacklisted btree node writes - we now only need it for ignoring journal
entries that were written after the newest flush journal entry, and then
we only need to keep those blacklist entries around until journal replay
is finished.
That means we can delete the code for scanning btree nodes to GC
journal_seq_blacklist entries.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Now that bch2_bucket_alloc_new_fs() isn't looking at bucket marks to
decide what buckets are eligible to allocate, we can clean up the
filesystem initialization and device add paths. Previously, we had to
use ancient code to mark superblock/journal buckets in the in memory
bucket marks as we allocated them, and then zero that out and re-do that
marking using the newer transational bucket mark paths. Now, we can
simply delete the in-memory bucket marking.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This changes bch2_bucket_alloc_new_fs() to a simple bump allocator that
doesn't need to use the in memory bucket array, part of a larger patch
series to entirely get rid of the in memory bucket array, except for
gc/fsck.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This adds flags for options that must be a power of two (block size and
btree node size), and options that are stored in the superblock as a
power of two (encoded extent max).
Also: options are now stored in memory in the same units they're
displayed in (bytes): we now convert when getting and setting from the
superblock.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This adds more latency/event measurements and breaks some apart into
more events. Journal writes are broken apart into flush writes and
noflush writes, btree compactions are broken out from btree splits,
btree mergers are added, as well as btree_interior_updates - foreground
and total.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This converts journal_write_delay, journal_flush_disabled, and
journal_reclaim_delay to normal filesystems options, and also adds them
to the superblock.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
It used to be that error reporting in the startup path was done by
returning strings describing the error, but that turned out to be a
rather silly idea - if there's something we can describe about the
error, just print it right away.
This converts a good chunk of code to returning error codes, as is more
typical style.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
bch2_fs_encryption_init() correctly passes back -ENOKEY from request_key()
when no unlock key is found, or -EINVAL if superblock decryption fails
because of an invalid key. However, these get absorbed into a generic NULL
return from bch2_fs_alloc() and later returned to user space as -ENOMEM,
leading to a misleading error from mount(1):
mount(2) system call failed: Out of memory.
Return explicit error pointers out of bch2_fs_alloc() and handle them in
both callers, so the user instead sees
mount(2) system call failed: Required key not available.
when attempting to mount a filesystem which is still locked.
Signed-off-by: Chris Webb <chris@arachsys.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This tweaks the journal code to always act as if there's space available
in nochanges mode, when we're not going to be doing any writes. This
helps in recovering filesystems that won't mount because they need
journal replay and the journal has gotten stuck.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Back when we relied on the journal sequence number blacklist machinery
for consistency between btree and the journal, we needed to ensure a new
journal entry was written before any btree writes were done. But, this
had the side effect of consuming some space in the journal prior to
doing journal replay - which could lead to a very wedged filesystem,
since we don't yet have a way to grow the journal prior to going RW.
Fortunately, the journal sequence number blacklist machinery isn't
needed anymore, as btree node pointers now record the numer of sectors
currently written to that node - that code should all be ripped out.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This is a hacky but effective fix to device usage stats for superblock
and journal being wrong on a newly added device (following the comment
that already told us how it needed to be done!)
Reported-by: Chris Webb <chris@arachsys.com>
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
It was missing a lockrestart_do(), to call bch2_trans_begin() and also
handle transaction restarts.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This patch adds subvolume.c - support for the subvolumes and snapshots
btrees and related data types and on disk data structures. The next
patches will start hooking up this new code to existing code.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
This splits btree_iter into two components: btree_iter is now the
externally visible componont, and it points to a btree_path which is now
reference counted.
This means we no longer have to clone iterators up front if they might
be mutated - btree_path can be shared by multiple iterators, and cloned
if an iterator would mutate a shared btree_path. This will help us use
iterators more efficiently, as well as slimming down the main long lived
state in btree_trans, and significantly cleans up the logic for iterator
lifetimes.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds progress stats to sysfs for copygc, rebalance, recovery, and the
cmd_job ioctls.
Signed-off-by: Brett Holman <bholman.devel@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This closes a significant hole (and last known hole) in our ability to
verify metadata. Previously, since btree nodes are log structured, we
couldn't detect lost btree writes that weren't the first write to a
given node. Additionally, this seems to have lead to some significant
metadata corruption on multi device filesystems with metadata
replication: since a write may have made it to one device and not
another, if we read that btree node back from the replica that did have
that write and started appending after that point, the other replica
would have a gap in the bset entries and reading from that replica
wouldn't find the rest of the bsets.
But, since updates to interior btree nodes are now journalled, we can
close this hole by updating pointers to btree nodes after every write
with the currently written number of sectors, without negatively
affecting performance. This means we will always detect lost or corrupt
metadata - it also means that our btree is now a curious hybrid of COW
and non COW btrees, with all the benefits of both (excluding
complexity).
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
bcachefs-tools recently started putting a backup superblock at the end
of the device. This causes a problem if the bucket size doesn't divide
the device size - but we can fix it by just skipping marking that part.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
We can't use btree_update_wq becuase btree updates may be waiting on
btree writes to complete.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Also, clean up workqueue usage - we shouldn't be using system
workqueues, pretty much everything we do needs to be on our own
WQ_MEM_RECLAIM workqueues.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Writeback throttling is a kernel config option and not always enabled.
When it's not enabled we need a fallback, to avoid unbounded memory
pinning and work item backlogs.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Ensure that the block device pointer in a superblock handle is not
null before dereferencing it in bch2_dev_to_fs. The block device pointer
may be null when mounting a new bcachefs filesystem given another mounted
bcachefs filesystem exists that has at least one device that is offline.
Signed-off-by: Dan Robertson <dan@dlrobertson.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We really need debug mode assertions that ca->ref and ca->io_ref are
used correctly.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This splits out btree topology repair into a separate pass, and makes
some improvements:
- When we have to pick which of two overlapping nodes to drop keys
from, we use the btree node header sequence number to preserve the
newer node
- the gc code has been changed so that it doesn't bail out if we're
continuing/ignoring on fsck error - this way the dump tool can skip
running the repair pass but still walk all reachable metadata
- add a new superblock flag indicating when a filesystem is known to
have btree topology issues, and the topology repair pass should be
run
- changing the start/end of a node might mean keys in that node have to
be deleted: this patch handles that better by splitting it out into a
separate function and running it explicitly in the topology repair
code, previously those keys were only being dropped when the btree
node was read in.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In userspace, we don't really have a well defined PAGE_SIZE and shouln't
be relying on it. This is some more incremental work to remove
references to it.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This patch standardizes all the enums that have associated string tables
(probably more enums should have string tables).
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Especially in userspace, we sometime run into resource exhaustion issues
with starting up threads after mark and sweep/fsck.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a regression from the patch
bcachefs: Fix copygc dying on startup
In general only the allocator thread itself should be updating
ca->allocator_state, the thread waking up the allocator setting it is an
ugly hack only needed to avoid racing with the copygc threads when we're
first starting up.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We have a separate mechanism for ratelimiting copygc now - the pd
controller has only been causing problems.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We're seeing failures to mount because of a failure to start the
allocator threads, which currently happens fairly late in the mount
process, after walking all metadata, and kthread_create() fails if
something has tried to kill the mount process, which is probably not
what we want.
This patch avoids this issue by creating, but not starting, the
allocator threads when we preallocate all of our other in memory data
structures.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
When the replicas mechanism was added, for tracking data by which drives
it's replicated on, the check for whether we have sufficient devices was
never updated to make use of it. This patch finally does that.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes some arithmetic bugs in "bcachefs: Journal updates to dev
usage" - additionally, it cleans things up by switching everything that
goes in every journal entry to the journal_entry_res mechanism.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This eliminates the need to scan every bucket to regenerate dev_usage at
mount time.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Originally, bcachefs - going back to bcache - stored, for each bucket, a
16 bit counter corresponding to how long it had been since the bucket
was read from. But, this required periodically rescaling counters on
every bucket to avoid wraparound. That wasn't an issue in bcache, where
we'd perodically rewrite the per bucket metadata all at once, but in
bcachefs we're trying to avoid having to walk every single bucket.
This patch switches to persisting 64 bit io clocks, corresponding to the
64 bit bucket timestaps introduced in the previous patch with
KEY_TYPE_alloc_v2.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Some errors may need to be fixed in order for GC to successfully run -
walk and mark all metadata. But we can't start the allocators and do
normal btree updates until after GC has completed, and allocation
information is known to be consistent, so we need a different method of
doing btree updates.
Fortunately, we already have code for walking the btree while overlaying
keys from the journal to be replayed. This patch adds an update path
that adds keys to the list of keys to be replayed by journal replay, and
also fixes up iterators.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
More work towards getting rid of the in memory struct bucket: this path
adds code for marking superblock and journal buckets via the btree, and
uses it in the device add and journal resize paths.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This is to make it more amenable for serialization.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
There was a race: btree node writes drop their reference on journal pins
before clearing the btree_node_write_in_flight flag.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Various filesystem usage counters are kept in percpu counters, with one
set per in flight journal buffer. Right now all the code that deals with
it assumes that there's only two buffers/sets of counters, but the
number of journal bufs is getting increased to 4 in the next patch - so
refactor that code to not assume a constant.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We allocate a lot of these, and we're seeing sporading OOMs - this will
help with tracking those down.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Previous varint implementation used by the inode code was not nearly as
fast as it could have been; partly because it was attempting to encode
integers up to 96 bits (for timestamps) but this meant that encoding and
decoding the length required a table lookup.
Instead, we'll just encode timestamps greater than 64 bits as two
separate varints; this will make decoding/encoding of inodes
significantly faster overall.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Allocating our array of btree iters is a big enough allocation that it
hits the buddy allocator, and we're seeing lots of lock contention.
Sticking a single element buffer in front of it should help.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This shards new inodes into different btree nodes by using the processor
ID for the high bits of the new inode number. Much faster than the
previous inode create optimization - this also helps with sharding in
the other btrees that index by inode number.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
With the btree key cache code, journal reclaim now has a lot more work
to do. It could be the case that after journal reclaim has finished one
iteration there's already more work to do, so put it in a loop to check
for that.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
On workloads that do a lot of multithreaded creates all at once, lock
contention on the inodes btree turns out to still be an issue.
This patch adds a small buffer of inode numbers that are known to be
free, so that we can avoid touching the btree on every create. Also,
this changes inode creates to update via the btree key cache for the
initial create.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a bug where the clock hands in the journal and superblock
didn't match, because we were still incrementing the read clock hand
while read-only.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now that we've got transactional alloc info updates (and have for
awhile), we don't need to write it out on shutdown, and we don't need to
write it out on startup except when GC found errors - this is a big
improvement to mount/unmount performance.
This patch also fixes a few bugs where we weren't writing out alloc
info (on new filesystems, and new devices) and should have been.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The copygc threads errors out and makes the filesystem go RO if it ever
tries to run and discovers it has no reserve allocated - which is a
problem if it races with the allocator thread and its reserve hasn't
been filled yet.
The allocator thread doesn't start filling the copygc reserve until
after BCH_FS_STARTED has been set, so make sure to wake up the allocator
threads after setting that and before starting copygc.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
At some point bch2_fs_alloc() was changed to always call bch2_fs_free()
in the error path, which means we need c->cl to always be initialized.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
There was a long standing race in the mount/unmount code - the VFS
intends for mount/unmount synchronizatino to be handled by the list of
superblocks, but we were still holding devices open after tearing down
our superblock in the unmount path.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Also - make sure to show the devices we actually have open in /proc
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Previously, we would start doing btree updates before writing the first
journal entry; if this was after an unclean shutdown, this could cause
those btree updates to not be blacklisted.
Also, move some code to headers for userspace debug tools.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Per device copygc threads don't move data to different devices and they
make fragmentation works - they don't make much sense anymore.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This is mainly to solve a lock ordering issue, and also simplifies the
code a bit.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Soon we'll be able to modify existing stripes - replacing empty blocks
with new blocks and new p/q blocks. This patch updates the trigger code
to handle pointers changing in an existing stripe; also, it
significantly improves how the stripes heap works, which means we can
get rid of the stripe creation/deletion lock.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This introduces a new kind of btree iterator, cached iterators, which
point to keys cached in a hash table. The cache also acts as a write
cache - in the update path, we journal the update but defer updating the
btree until the cached entry is flushed by journal reclaim.
Cache coherency is for now up to the users to handle, which isn't ideal
but should be good enough for now.
These new iterators will be used for updating inodes and alloc info (the
alloc and stripes btrees).
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
It's not needed anymore since we can now write to buckets before
updating the alloc btree.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now that interior btree updates are fully transactional, we don't need
to write out alloc info in a loop. However, interior btree updates do
put more things in the journal, so we still need a loop in the RO
sequence.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We now update the alloc info (bucket sector counts) atomically with
journalling the update to the interior btree nodes, and we also set new
btree roots atomically with the journalled part of the btree update.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
When updates to interior nodes started being journalled, that meant that
after an unclean shutdown, until journal replay is done we can't walk
the btree without overlaying the updates from the journal.
The initial btree gc was changed to walk the btree overlaying keys from
the journal - but bch2_alloc_read() and bch2_stripes_read() were missed.
Major whoops...
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Previously, there was at least one error path where we could mark the
filesystem clean when we hadn't sucessfully written out alloc info.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
- Always pass BTREE_INSERT_USE_RESERVE when writing alloc btree keys
- Don't strand buckest on the copygc freelist until after recovery is
done and we're starting copygc.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This will be used by the userspace debug tools.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This will help for debugging hangs during unmount
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Prep work for journalling updates to interior nodes - enforcing ordering
will greatly simplify those changes.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The device remove test was sporadically failing, because we hadn't
finished dropping btree sector counts for the device when
bch2_replicas_gc2() was called - mainly due to in flight journal writes.
We don't yet have a good mechanism for flushing the counts that
correspend to open journal entries yet.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
BTREE_INSERT_LAZY_RW was added for this since this code was written; use
it instead.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Long overdue cleanup - this converts btree_node_iter_large uses to
sort_iter.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This implements code for storing small bkeys on the stack and allocating
out of a mempool if they're too big.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
We weren't checking for errors when trying to delet stripes, which meant
ec_stripe_delete_work() would spin trying to delete the same stripe over
and over.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
mark_lock is a frequently taken lock, and there's also potential for
deadlocks since currently bch2_clear_page_bits which is called from
memory reclaim has to take it to drop disk reservations.
The disk reservation get path takes it when it recalculates the number
of sectors known to be available, but it's not really needed for
consistency. We just want to make sure we only have one thread updating
the sectors_available count, which we can do with a dedicated mutex.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Now, we store blacklisted journal sequence numbers in the superblock,
not the journal: this helps to greatly simplify the code, and more
importantly it's now implemented in a way that doesn't require all btree
nodes to be visited before starting the journal - instead, we
unconditionally blacklist the next 4 journal sequence numbers after an
unclean shutdown.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In flight btree updates could update alloc info until they're flushed -
so we have to try writing again after they've been flushed.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
- Does not persist alloc info for stripes yet
- Also does not yet include filesystem block/sector counts yet, from
struct fs_usage
- Not made use of just yet
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
journal reclaim writes btree nodes, which can end up waiting for in
flight btree writes to complete, and btree write completions run out of
workqueues - so we can't run out of the same workqueue or we risk
deadlock
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
this lets us get rid of a lot of extra switch statements - in a lot of
places we dispatch on the btree node type, and then the key type, so
this is a nice cleanup across a lot of code.
Also improve the on disk format versioning stuff.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This means we can now use gc to verify the allocation information -
important for testing persistant alloc info
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This way the last mount time is actually meaningful instead of just being
various times from 1970 (which happens with the monotonic clock).
Also, roundup_pow_of_two() is undefined when passed in 0, so check before
calling it.
Signed-off-by: Tim Schlueter <schlueter.tim@linux.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Due to compression, the different replicas of a replicated extent don't
necessarily have to take up the same amount of space - so replicated
data sector counts shouldn't be stored divided by the number of
replicas.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Initially forked from drivers/md/bcache, bcachefs is a new copy-on-write
filesystem with every feature you could possibly want.
Website: https://bcachefs.org
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
