Kent Overstreet [Thu, 28 Dec 2023 04:19:09 +0000 (23:19 -0500)]
bcachefs: trans_mark now takes bkey_s
Prep work for disk space accounting rewrite: we're going to want to use
a single callback for both of our current triggers, so we need to change
them to have the same type signature first.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 28 Jun 2023 01:02:27 +0000 (21:02 -0400)]
bcachefs: bch_member->seq
Add new fields for split brain detection:
- bch_member->seq, which tracks the sequence number of the last superblock
write that happened to each member device
- bch_sb->write_time, which tracks the time of the last superblock write,
to allow detection of when two members have diverged but had the same
number of superblock writes.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 23 Dec 2023 22:50:29 +0000 (17:50 -0500)]
bcachefs: Fix nochanges/read_only interaction
nochanges means "we cannot issue writes at all"; it's possible to go
into a pseudo read-write mode where we pin dirty metadata in memory,
which is used for fsck in dry run mode and doing journal replay on a
read only mount, but we do not want to allow an actual read-write mount
in nochanges mode.
But we do always want to allow early read-write, during recovery - this
patch clarifies that.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 15 Dec 2023 19:13:48 +0000 (14:13 -0500)]
bcachefs: Fix reattach_inode() for snapshots
reattach_inode() was broken w.r.t. snapshots - we'd lookup the subvolume
to look up lost+found, but if we're in an interior node snapshot that
didn't make any sense.
Instead, this adds a dirent path for creating in a specific snapshot,
skipping the subvolume; and we also make sure to create lost+found in
the root snapshot, to avoid conflicts with lost+found being created in
overlapping snapshots.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 11 Dec 2023 00:26:30 +0000 (19:26 -0500)]
bcachefs: growable btree_paths
XXX: we're allocating memory with btree locks held - bad
We need to plumb through an error path so we can do
allocate_dropping_locks() - but we're merging this now because it fixes
a transaction path overflow caused by indirect extent fragmentation, and
the resize path is rare.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Since the btree_paths array is now about to become growable, we have to
be careful not to refer to paths by pointer across contexts where they
may be reallocated.
This fixes the remaining btree_interior_update() paths - split and
merge.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
- Some tweaks to greatly reduce locking overhead for the list of btree
transactions, so that it can always be enabled: leave btree_trans
objects on the list when they're on the percpu single item freelist,
and only check for duplicates in the same process when
CONFIG_BCACHEFS_DEBUG is enabled
- don't zero out the full btree_trans() unless we allocated it from
the mempool
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Wed, 13 Dec 2023 01:08:29 +0000 (20:08 -0500)]
bcachefs: rcu protect trans->paths
Upcoming patches are going to be changing trans->paths to a
reallocatable buffer. We need to guard against use after free when it's
used by other threads; this introduces RCU protection to those paths and
changes them to check for trans->paths == NULL
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Randy Dunlap [Wed, 20 Dec 2023 06:44:09 +0000 (22:44 -0800)]
bcachefs: mean and variance: fix kernel-doc for function params
Add missing function parameter descriptions in mean_and_variance.c.
The also eliminates the "Excess function parameter" warnings.
Prevents these kernel-doc warnings:
mean_and_variance.c:67: warning: Function parameter or member 's' not described in 'mean_and_variance_get_mean'
mean_and_variance.c:78: warning: Function parameter or member 's1' not described in 'mean_and_variance_get_variance'
mean_and_variance.c:94: warning: Function parameter or member 's' not described in 'mean_and_variance_get_stddev'
mean_and_variance.c:108: warning: Function parameter or member 's' not described in 'mean_and_variance_weighted_update'
mean_and_variance.c:108: warning: Function parameter or member 'x' not described in 'mean_and_variance_weighted_update'
mean_and_variance.c:108: warning: Excess function parameter 's1' description in 'mean_and_variance_weighted_update'
mean_and_variance.c:108: warning: Excess function parameter 's2' description in 'mean_and_variance_weighted_update'
mean_and_variance.c:134: warning: Function parameter or member 's' not described in 'mean_and_variance_weighted_get_mean'
mean_and_variance.c:143: warning: Function parameter or member 's' not described in 'mean_and_variance_weighted_get_variance'
mean_and_variance.c:153: warning: Function parameter or member 's' not described in 'mean_and_variance_weighted_get_stddev'
Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Brian Foster <bfoster@redhat.com> Cc: linux-bcachefs@vger.kernel.org Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 23 Dec 2023 02:58:43 +0000 (21:58 -0500)]
bcachefs: check for failure to downgrade
With the upcoming member seq patch, it's now critical that we don't ever
write to a superblock that hasn't been version downgraded - failure to
update member seq fields will cause split brain detection to fire
erroniously.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Brian Foster [Tue, 19 Dec 2023 14:02:15 +0000 (09:02 -0500)]
bcachefs: clean up some dead fallocate code
The have_reservation local variable in bch2_extent_fallocate() is
initialized to false and set to true further down in the function.
Between this two points, one branch of code checks for negative
value and one for positive, and nothing ever checks the variable
after it is set to true. Clean up some of the unnecessary logic and
code.
Signed-off-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Tue, 19 Dec 2023 23:08:19 +0000 (18:08 -0500)]
bcachefs: Make sure allocation failure errors are logged
The previous patch fixed a bug in allocation path error handling, and it
would've been noticed sooner had it been logged properly.
Generally speaking, errors that shouldn't happen in normal operation and
are being returned up the stack should be logged: the write path was
already logging IO errors, but non IO errors were missed.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: Replace zero-length array with flex-array member and use __counted_by
Fake flexible arrays (zero-length and one-element arrays) are
deprecated, and should be replaced by flexible-array members.
So, replace zero-length array with a flexible-array member in
`struct bch_ioctl_fsck_offline`.
Also annotate array `devs` with `__counted_by()` to prepare for the
coming implementation by GCC and Clang of the `__counted_by` attribute.
Flexible array members annotated with `__counted_by` can have their
accesses bounds-checked at run-time via `CONFIG_UBSAN_BOUNDS` (for
array indexing) and `CONFIG_FORTIFY_SOURCE` (for strcpy/memcpy-family
functions).
This fixes the following -Warray-bounds warnings:
fs/bcachefs/chardev.c: In function 'bch2_ioctl_fsck_offline':
fs/bcachefs/chardev.c:363:34: warning: array subscript 0 is outside array bounds of '__u64[0]' {aka 'long long unsigned int[]'} [-Warray-bounds=]
363 | if (copy_from_user(devs, &user_arg->devs[0], sizeof(user_arg->devs[0]) * arg.nr_devs)) {
| ^~~~~~~~~~~~~~~~~~
In file included from fs/bcachefs/chardev.c:5:
fs/bcachefs/bcachefs_ioctl.h:400:33: note: while referencing 'devs'
400 | __u64 devs[0];
This results in no differences in binary output.
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Mon, 11 Dec 2023 07:13:33 +0000 (02:13 -0500)]
bcachefs: Drop journal entry compaction
Previously, we dropped empty journal entries and coalesced entries that
could be - but it's not worth the overhead; we very rarely leave unused
journal entries after getting a journal reservation.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sat, 4 Nov 2023 04:06:56 +0000 (00:06 -0400)]
bcachefs: Inline btree write buffer sort
The sort in the btree write buffer flush path is a very hot path, and
it's particularly performance sensitive since it's single threaded and
can block every other thread on a multithreaded write workload.
It's well worth doing a sort with inlined cmp and swap functions.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Thu, 2 Nov 2023 22:57:19 +0000 (18:57 -0400)]
bcachefs: btree write buffer now slurps keys from journal
Previosuly, the transaction commit path would have to add keys to the
btree write buffer as a separate operation, requiring additional global
synchronization.
This patch introduces a new journal entry type, which indicates that the
keys need to be copied into the btree write buffer prior to being
written out. We switch the journal entry type back to
JSET_ENTRY_btree_keys prior to write, so this is not an on disk format
change.
Flushing the btree write buffer may require pulling keys out of journal
entries yet to be written, and quiescing outstanding journal
reservations; we previously added journal->buf_lock for synchronization
with the journal write path.
We also can't put strict bounds on the number of keys in the journal
destined for the write buffer, which means we might overflow the size of
the preallocated buffer and have to reallocate - this introduces a
potentially fatal memory allocation failure. This is something we'll
have to watch for, if it becomes an issue in practice we can do
additional mitigation.
The transaction commit path no longer has to explicitly check if the
write buffer is full and wait on flushing; this is another performance
optimization. Instead, when the btree write buffer is close to full we
change the journal watermark, so that only reservations for journal
reclaim are allowed.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 10 Dec 2023 22:44:04 +0000 (17:44 -0500)]
bcachefs: bch2_trans_node_add no longer uses trans_for_each_path()
In the future we'll be making trans->paths resizable and potentially
having _many_ more paths (for fsck); we need to start fixing algorithms
that walk each path in a transaction where possible.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Sun, 10 Dec 2023 21:48:22 +0000 (16:48 -0500)]
bcachefs: Improve trans->extra_journal_entries
Instead of using a darray, we now allocate journal entries for the
transaction commit path with our normal bump allocator - with an inlined
fastpath, and using btree_transaction_stats to remember how much to
initially allocate so as to avoid transaction restarts.
This is prep work for converting write buffer updates to use this
mechanism.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Randy Dunlap [Sun, 10 Dec 2023 06:06:44 +0000 (22:06 -0800)]
bcachefs: six lock: fix typos
Fix a few typos in the six.h header file.
Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Brian Foster <bfoster@redhat.com> Cc: linux-bcachefs@vger.kernel.org Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 8 Dec 2023 04:28:26 +0000 (23:28 -0500)]
bcachefs: Kill for_each_btree_key()
for_each_btree_key() handles transaction restarts, like
for_each_btree_key2(), but only calls bch2_trans_begin() after a
transaction restart - for_each_btree_key2() wraps every loop iteration
in a transaction.
The for_each_btree_key() behaviour is problematic when it leads to
holding the SRCU lock that prevents key cache reclaim for an unbounded
amount of time - there's no real need to keep it around.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Kent Overstreet [Fri, 8 Dec 2023 04:50:38 +0000 (23:50 -0500)]
bcachefs: Fix bch2_read_btree()
In the debugfs code, we had an incorrect use of drop_locks_do(); on
transaction restart we don't want to restart the current loop iteration,
since we've already emitted the current key to the buffer for userspace.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
projects / mirror_ubuntu-kernels.git / log