Kent Overstreet [Fri, 20 Dec 2024 09:46:00 +0000 (04:46 -0500)]
bcachefs: Fixes for snapshot_tree.master_subvol
Ensure that snapshot_tree.master_subvol is cleared when we delete the
master subvolume in a tree of snapshots, and allow for snapshot trees
that don't have a master subvolume in fsck.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 21 Dec 2024 04:56:42 +0000 (23:56 -0500)]
bcachefs: Don't rely on snapshot_tree.master_subvol for reattaching
Previously, fsck used the snapshot tree's master subvol for finding the
root inode number - but the master subvol might have been deleting, and
setting a new one should be a user operation; meaning we can't rely on
it existing.
Fortunately, for finding the root inode number in a tree of snapshots,
finding any associated subvolume works.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 20 Dec 2024 10:20:01 +0000 (05:20 -0500)]
bcachefs: bch2_kvmalloc()
Add a version of kvmalloc() that doesn't have the INT_MAX limit; large
filesystems do hit this.
We'll want to get rid of the in-memory bucket gens array, but we're not
there quite yet.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 16 Dec 2024 21:41:25 +0000 (16:41 -0500)]
bcachefs: Fix assert for online fsck
We can't check if we're racing with fsck ending until mark_lock is held.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 16 Dec 2024 18:58:02 +0000 (13:58 -0500)]
bcachefs: Handle -BCH_ERR_need_mark_replicas in gc
Locking considerations (possibly no longer relevant?) mean that when an
accounting update needs a new superblock replicas entry to be created,
it's deferred to the transaction commit error path.
But accounting updates for gc/fcsk aren't done from the transaction
commit path - so we need to handle
-BCH_ERR_btree_insert_need_mark_replicas locally.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 8 Jun 2024 21:01:31 +0000 (17:01 -0400)]
bcachefs: Write lock btree node in key cache fills
this addresses a key cache coherency bug
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 15 Dec 2024 07:24:30 +0000 (02:24 -0500)]
bcachefs: kill __bch2_btree_iter_flags()
bch2_btree_iter_flags() now takes a level parameter; this fixes a bug
where using a node iterator on a leaf wouldn't set
BTREE_ITER_with_key_cache, leading to fun cache coherency bugs.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 15 Dec 2024 07:03:11 +0000 (02:03 -0500)]
bcachefs: Drop redundant "read error" call from btree_gc
The btree node read error path already calls topology error, so this is
entirely redundant, and we're not specific enough about our error codes
- this was triggering for bucket_ref_update() errors.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 15 Dec 2024 06:52:54 +0000 (01:52 -0500)]
bcachefs: Drop racy warning
Checking for writing past i_size after unlocking the folio and clearing
the dirty bit is racy, and we already check it at the start.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 05:55:48 +0000 (00:55 -0500)]
bcachefs: better check_bp_exists() error message
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Hongbo Li [Tue, 12 Nov 2024 08:15:47 +0000 (16:15 +0800)]
bcachefs: add counter_flags for counters
In bcachefs, io_read and io_write counter record the amount
of data which has been read and written. They increase in
unit of sector, so to display correctly, they need to be
shifted to the left by the size of a sector. Other counters
like io_move, move_extent_{read, write, finish} also have
this problem.
In order to support different unit, we add extra column to
mark the counter type by using TYPE_COUNTER and TYPE_SECTORS
in BCH_PERSISTENT_COUNTERS().
Fixes:
1c6fdbd8f246 ("bcachefs: Initial commit")
Signed-off-by: Hongbo Li <lihongbo22@huawei.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 10 Dec 2024 19:19:30 +0000 (14:19 -0500)]
bcachefs: bcachefs_metadata_version_autofix_errors
It's time to make self healing the default: change the error action for
old filesystems to fix_safe, matching the default for current
filesystems.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 2 Dec 2024 02:44:38 +0000 (21:44 -0500)]
bcachefs: bcachefs_metadata_version_persistent_inode_cursors
Persistent cursors for inode allocation.
A free inodes btree would add substantial overhead to inode allocation
and freeing - a "next num to allocate" cursor is always going to be
faster.
We just need it to be persistent, to avoid scanning the inodes btree
from the start on startup.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 3 Aug 2023 00:27:38 +0000 (20:27 -0400)]
bcachefs: bcachefs_metadata_version_inode_depth
This adds a new inode field, bi_depth, for directory inodes: this allows
us to make the check_directory_structure pass much more efficient.
Currently, to ensure the filesystem is fully connect and has no loops,
for every directory we follow backpointers until we find the root. But
by adding a depth counter, it sufficies to only check the parent of each
directory, and check that the parent's bi_depth is smaller.
(fsck doesn't require that bi_depth = parent->bi_depth + 1; if a rename
causes bi_depth off, but the chain to the root is still strictly
decreasing, then the algorithm still works and there's no need for fsck
to fixup the bi_depth fields).
We've already checked backpointers, so we know that every directory
(excluding the root)has a valid parent: if bi_depth is always
decreasing, every chain must terminate, and terminate at the root
directory.
bi_depth will not necessarily be correct when fsck runs, due to
directory renames - we can't change bi_depth on every child directory
when renaming a directory. That's ok; fsck will silently fix the
bi_depth field as needed, and future fsck runs will be much faster.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 20 Oct 2024 06:12:21 +0000 (02:12 -0400)]
bcachefs: Option changes now get propagated to reflinked data
Now that bch2_move_get_io_opts() re-propagates changed inode io options
to bch_extent_rebalance, we can properly suport changing IO path options
for reflinked data.
Changing a per-file IO path option, either via the xattr interface or
via the BCHFS_IOC_REINHERIT_ATTRS ioctl, will now trigger a scan (the
inode number is marked as needing a scan, via
bch2_set_rebalance_needs_scan()), and rebalance will use
bch2_move_data(), which will walk the inode number and pick up the new
options.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 7 Nov 2024 04:16:24 +0000 (23:16 -0500)]
bcachefs: bcachefs_metadata_version_reflink_p_may_update_opts
Previously, io path option changes on a file would be picked up
automatically and applied to existing data - but not for reflinked data,
as we had no way of doing this safely. A user may have had permission to
copy (and reflink) a given file, but not write to it, and if so they
shouldn't be allowed to change e.g. nr_replicas or other options.
This uses the incompat feature mechanism in the previous patch to add a
new incompatible flag to bch_reflink_p, indicating whether a given
reflink pointer may propagate io path option changes back to the
indirect extent.
In this initial patch we're only setting it for the source extents.
We'd like to set it for the destination in a reflink copy, when the user
has write access to the source, but that requires mnt_idmap which is not
curretly plumbed up to remap_file_range.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 12 Nov 2024 02:50:29 +0000 (21:50 -0500)]
bcachefs: BCH_SB_VERSION_INCOMPAT
We've been getting away from feature bits: they don't have any kind of
ordering, and thus it's possible for people to enable weird combinations
of features that were never tested or intended to be run.
Much better to just give every new feature, compatible or incompatible,
a version number.
Additionally, we probably won't ever rev the major version number: major
version numbers represent incompatible versions, but that doesn't really
fit with how we actually roll out incompatible features - we need a
better way of rolling out incompatible features.
So, this patch adds two new superblock fields:
- BCH_SB_VERSION_INCOMPAT
- BCH_SB_VERSION_INCOMPAT_ALLOWED
BCH_SB_VERSION_INCOMPAT_ALLOWED indicates that incompatible features up
to version number x are allowed to be used without user prompting, but
it does not by itself deny old versions from mounting.
BCH_SB_VERSION_INCOMPAT does deny old versions from mounting, and must
be <= BCH_SB_VERSION_INCOMPAT_ALLOWED.
BCH_SB_VERSION_INCOMPAT will only be set when a codepath attempts to use
an incompatible feature, so as to not unnecessarily break compatibility
with old versions.
bch2_request_incompat_feature() is the new interface to check if an
incompatible feature may be used.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 15 Nov 2024 01:47:32 +0000 (20:47 -0500)]
bcachefs: Only run check_backpointers_to_extents in debug mode
The backpointers passes, check_backpointers_to_extents() and
check_extents_to_backpointers() are the most expensive fsck passes.
Now that we're running the same check and repair code when using a
backpointer at runtime (via bch2_backpointer_get_key()) that fsck does,
there's no reason fsck needs to - except to verify that the filesystem
really has no errors in debug mode.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 10 Dec 2024 19:04:39 +0000 (14:04 -0500)]
bcachefs: better backpointer_target_not_found() error message
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 12 Nov 2024 08:46:31 +0000 (03:46 -0500)]
bcachefs: bch2_backpointer_get_key() now repairs dangling backpointers
Continuing on with the self healing theme, we should be running any
check and repair code at runtime that we can - instead of declaring the
filesystemt inconsistent.
This will also let us skip running the backpointers -> extents fsck pass
except in debug mode.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 15 Nov 2024 21:31:54 +0000 (16:31 -0500)]
bcachefs: check_extents_to_backpointers() now only checks buckets with mismatches
Instead of walking every extent and every backpointer it points to,
first sum up backpointers in each bucket and check for mismatches, and
only look for missing backpointers if mismatches were detected, and only
check extents in those buckets.
This is a major fsck scalability improvement, since the two backpointers
passes (backpointers -> extents and extents -> backpointers) are the
most expensive fsck passes by far.
Additionally, to speed up the upgrade for backpointer bucket gens, or in
situations when we have to rebuild alloc info, add a special case for
when no backpointers are found in a bucket - don't check each individual
backpointer (in particular, avoiding the write buffer flushes), just
recreate them.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 15 Nov 2024 03:13:29 +0000 (22:13 -0500)]
bcachefs: Add write buffer flush param to backpointer_get_key()
In an upcoming patch bch2_backpointer_get_key() will be repairing when
it finds a dangling backpointer; it will need to flush the btree write
buffer before it can definitively say there's an error.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 17 Nov 2024 23:37:41 +0000 (18:37 -0500)]
bcachefs: kill __bch2_extent_ptr_to_bp()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 18 Nov 2024 04:58:21 +0000 (23:58 -0500)]
bcachefs: bch2_extent_ptr_to_bp() no longer depends on device
bch_backpointer no longer contains the bucket_offset field, it's just a
direct LBA mapping (with low bits to account for compressed extent
splitting), so we don't need to refer to the device to construct it
anymore.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 22:41:43 +0000 (17:41 -0500)]
bcachefs: bcachefs_metadata_version_disk_accounting_big_endian
Fix sort order for disk accounting keys, in order to fix a regression on
mount times.
The typetag is now the most significant byte of the key, meaning disk
accounting keys of the same type now sort together.
This lets us skip over disk accounting keys that aren't mirrored in
memory when reading accounting at startup, instead of having them
interleaved with other counter types.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 17 Nov 2024 04:53:07 +0000 (23:53 -0500)]
bcachefs: bcachefs_metadata_version_backpointer_bucket_gen
New on disk format version: backpointers new include the generation
number of the bucket they refer to, and the obsolete bucket_offset field
(no longer needed because we no longer store backpointers in alloc keys)
is gone.
This is an expensive forced upgrade - hopefully the last; we have to run
the extents_to_backpointers recovery pass to regenerate backpointers.
It's a forced incompatible upgrade because the alternative would've been
permamently making backpointers bigger, and as one of the biggest btrees
(along with the extents btree) that's not an ideal option.
It's worth it though, because this allows us to make the
check_extents_to_backpointers pass drastically cheaper: an upcoming
patch changes it to sum up backpointers in a bucket and check the sum
against the sector counts for that bucket, only looking for missing
backpointers if they don't match (and then only for specific buckets).
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 13 Dec 2024 10:58:34 +0000 (05:58 -0500)]
bcachefs: bch2_btree_path_peek_slot() doesn't return errors
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 13 Dec 2024 10:43:00 +0000 (05:43 -0500)]
bcachefs: trace_key_cache_fill
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 09:00:40 +0000 (04:00 -0500)]
bcachefs: Log message in journal for snapshot deletion
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 05:44:28 +0000 (00:44 -0500)]
bcachefs: bch2_trans_log_msg()
Export a helper for logging to the journal when we're already in a
transaction context.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 09:03:32 +0000 (04:03 -0500)]
bcachefs: Kill snapshot_t->equiv
Now entirely dead code.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 08:03:58 +0000 (03:03 -0500)]
bcachefs: Snapshot deletion no longer uses snapshot_t->equiv
Switch to generating a private list of interior nodes to delete, instead
of using the equivalence class in the global data structure.
This eliminates possible races with snapshot creation, and is much
cleaner - it'll let us delete a lot of janky code for calculating and
maintaining the equivalence classes.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 07:41:37 +0000 (02:41 -0500)]
bcachefs: Kill equiv_seen arg to delete_dead_snapshots_process_key()
When deleting dead snapshots, we move keys from redundant interior
snapshot nodes to child nodes - unless there's already a key, in which
case the ancestor key is deleted.
Previously, we tracked via equiv_seen whether the child snapshot had a
key, but this was tricky w.r.t. transaction restarts, and not
transactionally safe w.r.t. updates in the child snapshot.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 07:27:52 +0000 (02:27 -0500)]
bcachefs: Don't run overwrite triggers before insert
This breaks when the trigger is inserting updates for the same btree, as
the inode trigger now does.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 07:32:32 +0000 (02:32 -0500)]
bcachefs: alloc_data_type_set() happens in alloc trigger
Originally, we ran insert triggers before overwrite so that if an extent
was being moved (by fallocate insert/collapse range), the bucket sector
count wouldn't hit 0 partway through, and so we don't trigger state
changes caused by that too soon.
But this is better solved by just moving the data type change to the
alloc trigger itself, where it's already called.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 13 Dec 2024 10:29:27 +0000 (05:29 -0500)]
bcachefs: Fix key cache + BTREE_ITER_all_snapshots
Normally, whitouts (KEY_TYPE_whitout) are filtered from btree lookups,
since they exist only to represent deletions of keys in ancestor
snapshots - except, they should not be filtered in
BTREE_ITER_all_snapshots mode, so that e.g. snapshot deletion can clean
them up.
This means that that the key cache has to store whiteouts, and key cache
fills cannot filter them.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 07:26:15 +0000 (02:26 -0500)]
bcachefs: Fix btree_trans_peek_key_cache() BTREE_ITER_all_snapshots
In BTREE_ITER_all_snapshots mode, we're required to only return keys
where the snapshot field matches the iterator position -
BTREE_ITER_filter_snapshots requires pulling keys into the key cache
from ancestor snapshots, so we have to check for that.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 13 Dec 2024 11:02:24 +0000 (06:02 -0500)]
bcachefs: tidy btree_trans_peek_journal()
Change to match bch2_btree_trans_peek_updates() calling convention.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 12 Dec 2024 08:38:14 +0000 (03:38 -0500)]
bcachefs: tidy up __bch2_btree_iter_peek()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 02:10:27 +0000 (21:10 -0500)]
bcachefs: check_indirect_extents can run online
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 10 Dec 2024 18:23:47 +0000 (13:23 -0500)]
bcachefs: Refactor c->opts.reconstruct_alloc
Now handled in one place.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Nathan Chancellor [Tue, 10 Dec 2024 18:12:07 +0000 (11:12 -0700)]
bcachefs: Add empty statement between label and declaration in check_inode_hash_info_matches_root()
Clang 18 and newer warns (or errors with CONFIG_WERROR=y):
fs/bcachefs/str_hash.c:164:2: error: label followed by a declaration is a C23 extension [-Werror,-Wc23-extensions]
164 | struct bch_inode_unpacked inode;
| ^
In Clang 17 and prior, this is an unconditional hard error:
fs/bcachefs/str_hash.c:164:2: error: expected expression
164 | struct bch_inode_unpacked inode;
| ^
fs/bcachefs/str_hash.c:165:30: error: use of undeclared identifier 'inode'
165 | ret = bch2_inode_unpack(k, &inode);
| ^
fs/bcachefs/str_hash.c:169:55: error: use of undeclared identifier 'inode'
169 | struct bch_hash_info hash2 = bch2_hash_info_init(c, &inode);
| ^
fs/bcachefs/str_hash.c:171:40: error: use of undeclared identifier 'inode'
171 | ret = repair_inode_hash_info(trans, &inode);
| ^
Add an empty statement between the label and the declaration to fix the
warning/error without disturbing the code too much.
Fixes:
2519d3b0d656 ("bcachefs: bch2_str_hash_check_key() now checks inode hash info")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/
202412092339.QB7hffGC-lkp@intel.com/
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 10 Dec 2024 15:29:12 +0000 (10:29 -0500)]
bcachefs: trace_write_buffer_maybe_flush
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 06:31:43 +0000 (01:31 -0500)]
bcachefs: bch2_snapshot_exists()
bch2_snapshot_equiv() is going away; convert users that just wanted to
know if the snapshot exists to something better
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 03:30:19 +0000 (22:30 -0500)]
bcachefs: bch2_check_key_has_snapshot() prints btree id
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 02:47:34 +0000 (21:47 -0500)]
bcachefs: bch2_str_hash_check_key() now checks inode hash info
Versions of the same inode in different snapshots must have the same
hash info; this is critical for lookups to work correctly.
We're going to be running the str_hash checks online, at readdir or
xattr list time, so we now need str_hash_check_key() to check for inode
hash seed mismatches, since it won't be run right after check_inodes().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 03:00:36 +0000 (22:00 -0500)]
bcachefs: Don't BUG_ON() inode unpack error
Bkey validation checks that inodes are well-formed and unpack
successfully, so an unpack error should always indicate memory
corruption or some other kind of hardware bug - but these are still
errors we can recover from.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 02:42:49 +0000 (21:42 -0500)]
bcachefs: Use proper errcodes for inode unpack errors
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 01:55:03 +0000 (20:55 -0500)]
bcachefs: kill sysfs internal/accounting
Since we added per-inode counters there's now far too many counters to
show in one shot - if we want this in the future, it'll have to be in
debugfs.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 8 Dec 2024 09:11:21 +0000 (04:11 -0500)]
bcachefs: Kill unnecessary mark_lock usage
We can't hold mark_lock while calling fsck_err() - that's a deadlock,
mark_lock is meant to be a leaf node lock.
It's also unnecessary for gc_bucket() and bucket_gen(); rcu suffices
since the bucket_gens array describes its size, and we can't race with
device removal or resize during gc/fsck since that takes state lock.
Reported-by: syzbot+38641fcbda1aaffefdd4@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 9 Dec 2024 11:00:33 +0000 (06:00 -0500)]
bcachefs: Don't start rewriting btree nodes until after journal replay
This fixes a deadlock during journal replay when btree node read errors
kick off a ton of rewrites: we don't want them competing with journal
replay.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 04:15:05 +0000 (23:15 -0500)]
bcachefs: Fix reuse of bucket before journal flush on multiple empty -> nonempty transition
For each bucket we track when the bucket became nonempty and when it
became empty again: if we can ensure that there will be no journal
flushes in the range [nonempty, empty) (possibly because they occured at
the same journal sequence number), then it's safe to reuse the bucket
without waiting for a journal commit.
This is a major performance optimization for erasure coding, where
writes are initially replicated, but the extra replicas are quickly
dropped: if those buckets are reused and overwritten without issuing a
cache flush to the underlying device, then they only cost bus bandwidth.
But there's a tricky corner case when there's multiple empty -> nonempty
-> empty transitions in quick succession, i.e. when data is getting
overwritten immediately as it's being written.
If this happens and the previous empty transition hasn't been flushed,
we need to continue tracking the previous nonempty transition - not
start a new one.
Fixing this means we now need to track both the nonempty and empty
transitions in bch_alloc_v4.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 8 Dec 2024 05:28:16 +0000 (00:28 -0500)]
bcachefs: bch2_journal_noflush_seq() now takes [start, end)
Harder to screw up if we're explicit about the range, and more correct
as journal reservations can be outstanding on multiple journal entries
simultaneously.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 8 Dec 2024 01:43:07 +0000 (20:43 -0500)]
bcachefs: Set bucket needs discard, inc gen on empty -> nonempty transition
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 17:35:43 +0000 (12:35 -0500)]
bcachefs: Don't add unknown accounting types to eytzinger tree
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 8 Dec 2024 02:36:15 +0000 (21:36 -0500)]
bcachefs: Plumb bkey_validate_context to journal_entry_validate
This lets us print the exact location in the journal if it was found in
the journal, or correctly print if it was found in the superblock.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 00:23:22 +0000 (19:23 -0500)]
bcachefs: Use a heap for handling overwrites in btree node scan
Fix an O(n^2) issue when we find many overlapping (overwritten) btree
nodes - especially when one node overwrites many smaller nodes.
This was discovered to be an issue with the bcachefs
merge_torture_flakey test - if we had a large btree that was then
emptied, the number of difficult overwrites can be unbounded.
Cc: Kuan-Wei Chiu <visitorckw@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 00:16:02 +0000 (19:16 -0500)]
lib min_heap: Switch to size_t
size_t is the correct type for a count of objects that can fit in
memory: this also means heaps now have the same memory layout as darrays
(fs/bcachefs/darray.h), and darrays can be used as heaps.
Cc: Kuan-Wei Chiu <visitorckw@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Coly Li <colyli@suse.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 03:37:42 +0000 (22:37 -0500)]
bcachefs: Minor bucket alloc optimization
Check open buckets and buckets waiting for journal commit before doing
other expensive lookups.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 00:49:46 +0000 (19:49 -0500)]
bcachefs: Mark more errors autofix
tested repairing from a bug uncovered by the merge_torture_flakey test
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 7 Dec 2024 01:11:16 +0000 (20:11 -0500)]
bcachefs: fix bch2_btree_node_header_to_text() format string
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 17:35:17 +0000 (12:35 -0500)]
bcachefs: Journal space calculations should skip durability=0 devices
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 04:36:33 +0000 (23:36 -0500)]
bcachefs: factor out str_hash.c
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 04:40:26 +0000 (23:40 -0500)]
bcachefs: kill flags param to bch2_subvolume_get()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 01:43:01 +0000 (20:43 -0500)]
bcachefs: Don't call bch2_btree_interior_update_will_free_node() until after update succeeds
Originally, btree splits always succeeded once we got to the point of
recursing to the btree_insert_node() call.
But that changed when we switched to not taking intent locks all the way
up to the root, and that introduced a bug, because
bch2_btree_interior_update_will_free_node() cancels paending writes and
reparents a node that's going to be made visible on disk by another
btree update to the current btree update.
This was discovered in recent backpointers work, because
bch2_btree_interior_update_will_free_node() also clears the
will_make_reachable flag, causing backpointer target lookup to
spuriously thing it had found a dangling backpointer (when the
backpointer just hadn't been created yet by
btree_update_nodes_written()).
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 00:46:35 +0000 (19:46 -0500)]
bcachefs: Make sure __bch2_run_explicit_recovery_pass() signals to rewind
We should always signal to rewind if the requested pass hasn't been run,
even if called multiple times.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 00:41:38 +0000 (19:41 -0500)]
bcachefs: Call bch2_btree_lost_data() on btree read error
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 23:14:14 +0000 (18:14 -0500)]
bcachefs: Journal write path refactoring, debug improvements
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 5 Dec 2024 00:21:22 +0000 (19:21 -0500)]
bcachefs: dev_alloc_list.devs -> dev_alloc_list.data
This lets us use darray macros on dev_alloc_list (and it will become a
darray eventually, when we increase the maximum number of devices).
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 23:16:25 +0000 (18:16 -0500)]
bcachefs: Fix failure to allocate journal write on discard retry
When allocating a journal write fails, then retries after doing
discards, we were failing to count already allocated replicas.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 22:53:38 +0000 (17:53 -0500)]
bcachefs: BCH_ERR_insufficient_journal_devices
kill another standard error code use
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 22:48:06 +0000 (17:48 -0500)]
bcachefs: Silence "unable to allocate journal write" if we're already RO
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 22:44:25 +0000 (17:44 -0500)]
bcachefs: trace_accounting_mem_insert
Add a tracepoint for inserting new accounting entries: we're seeing odd
spinning behaviour in accounting read.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 06:19:28 +0000 (01:19 -0500)]
bcachefs: Advance to next bp on BCH_ERR_backpointer_to_overwritten_btree_node
Don't spin.
Fixes:
de95cc201a97 ("bcachefs: Kill bch2_get_next_backpointer()")
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 03:03:18 +0000 (22:03 -0500)]
bcachefs: Simplify disk accounting validate late
The validate late path was iterating over accounting entries in
eytzinger order, which is unnecessarily tricky when we may have to
remove entries.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 2 Dec 2024 02:35:11 +0000 (21:35 -0500)]
bcachefs: logged ops only use inum 0 of logged ops btree
we wish to use the logged ops btree for other items that aren't strictly
logged ops: cursors for inode allocation
There's no reason to create another cached btree for inode allocator
cursors - so reserve different parts of the keyspace for different
purposes.
Older versions will ignore or delete the cursors.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 4 Dec 2024 02:22:26 +0000 (21:22 -0500)]
bcachefs: rcu_pending now works in userspace
Introduce a typedef to handle the difference between unsigned
long/struct urcu_gp_poll_state.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Geert Uytterhoeven [Tue, 3 Dec 2024 16:40:10 +0000 (17:40 +0100)]
bcachefs: BCACHEFS_PATH_TRACEPOINTS should depend on TRACING
When tracing is disabled, there is no point in asking the user about
enabling extra btree_path tracepoints in bcachefs.
Fixes:
32ed4a620c5405be ("bcachefs: Btree path tracepoints")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 3 Dec 2024 04:36:38 +0000 (23:36 -0500)]
bcachefs: Fix allocating too big journal entry
The "journal space available" calculations didn't take into account
mismatched bucket sizes; we need to take the minimum space available out
of our devices.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 1 Dec 2024 21:39:54 +0000 (16:39 -0500)]
bcachefs: Improve "unable to allocate journal write" message
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 1 Dec 2024 04:27:45 +0000 (23:27 -0500)]
bcachefs: fix bch2_journal_key_insert_take() seq
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 23:53:26 +0000 (18:53 -0500)]
bcachefs: bch2_async_btree_node_rewrites_flush()
Add a method to flush btree node rewrites at the end of recovery, to
ensure that corrected errors are persisted.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 23:17:00 +0000 (18:17 -0500)]
bcachefs: If we did repair on a btree node, make sure we rewrite it
Ensure that "invalid bkey" repair gets persisted, so that it doesn't
repeatedly spam the logs.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 23:20:42 +0000 (18:20 -0500)]
bcachefs: bkey_fsck_err now respects errors_silent
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 30 Nov 2024 00:13:54 +0000 (19:13 -0500)]
bcachefs: list_pop_entry()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 14 Nov 2024 04:08:57 +0000 (23:08 -0500)]
bcachefs: Convert write path errors to inum_to_path()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 28 Sep 2024 19:40:49 +0000 (15:40 -0400)]
bcachefs: bch2_inum_to_path()
Add a function for walking backpointers to find a path from a given
inode number, and convert various error messages to use it.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 30 Nov 2024 02:12:47 +0000 (21:12 -0500)]
bcachefs: Fix fsck.c build in userspace
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Yang Li [Fri, 29 Nov 2024 06:38:27 +0000 (14:38 +0800)]
bcachefs: Add missing parameter description to bch2_bucket_alloc_trans()
The function bch2_bucket_alloc_trans() lacked a description for the
nowait parameter in its documentation comment block. This patch adds the
missing description to ensure all parameters are properly documented.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=12179
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 00:30:23 +0000 (19:30 -0500)]
bcachefs: Don't recurse in check_discard_freespace_key
When calling check_discard_freeespace_key from the allocator, we can't
repair without recursing - run it asynchronously instead.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 29 Nov 2024 00:02:18 +0000 (19:02 -0500)]
bcachefs: Check for extent crc uncompressed/compressed size mismatch
When not compressed, these must be equal - this fixes an assertion pop
in bch2_rechecksum_bio().
Reported-by: syzbot+50d3544c9b8db9c99fd2@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 23:05:06 +0000 (18:05 -0500)]
bcachefs: bch2_trans_relock() is trylock for lockdep
fix some spurious lockdep splats
Reported-by: syzbot+e088be3c2d5c05aaac35@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 22:57:55 +0000 (17:57 -0500)]
bcachefs: cryptographic MACs on superblock are not (yet?) supported
We should add support for cryptographic macs on the superblock - and it
won't be hard, but it'll need an incompatible feature bit (and we have a
new incompatible feature versioning scheme coming).
For now, just add a guard to avoid a dull ptr deref in gen_poly_key().
Reported-by: syzbot+dd3d9835055dacb66f35@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 22:48:20 +0000 (17:48 -0500)]
bcachefs: Check for inode journal seq in the future
More check and repair code: this fixes a warning in
bch2_journal_flush_seq_async()
Reported-by: syzbot+d119b445ec739e7f3068@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 21:59:40 +0000 (16:59 -0500)]
bcachefs: Check for bucket journal seq in the future
This fixes an assertion pop in bch2_journal_noflush_seq() - log the
error to the superblock and continue instead.
Reported-by: syzbot+85700120f75fc10d4e18@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 21:25:41 +0000 (16:25 -0500)]
bcachefs: do_fsck_ask_yn()
__bch2_fsck_err() is huge, and badly needs more refactoring
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 21:14:06 +0000 (16:14 -0500)]
bcachefs: Don't error out when logging fsck error
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 21:09:15 +0000 (16:09 -0500)]
bcachefs: mark more errors AUTOFIX
mark errors as autofix where syzbot has hit the repair paths
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 21:09:04 +0000 (16:09 -0500)]
bcachefs: add missing printbuf_reset()
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 28 Nov 2024 20:10:24 +0000 (15:10 -0500)]
bcachefs: Fix journal_iter list corruption
Fix exiting an iterator that wasn't initialized.
Reported-by: syzbot+2f7c2225ed8a5cb24af1@syzkaller.appspotmail.com
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>