Btree nodes are log structured; thus, we need to emit whiteouts when
we're deleting a key that's been written out to disk.
k->needs_whiteout tracks whether a key will need a whiteout when it's
deleted, and this requires some careful handling; e.g. the key we're
deleting may not have been written out to disk, but it may have
overwritten a key that was - thus we need to carry this flag around on
overwrites.
Invariants:
There may be multiple key for the same position in a given node (because
of overwrites), but only one of them will be a live (non deleted) key,
and only one key for a given position will have the needs_whiteout flag
set.
Additionally, we don't want to carry around whiteouts that need to be
written in the main searchable part of a btree node - btree_iter_peek()
will have to skip past them, and this can lead to an O(n^2) issues when
doing sequential deletions (e.g. inode rm/truncate). So there's a
separate region in the btree node buffer for unwritten whiteouts; these
are merge sorted with the rest of the keys we're writing in the btree
node write path.
The unwritten whiteouts was a later optimization that bch2_sort_keys()
didn't take into account; the unwritten whiteouts area means that we
never have deleted keys with needs_whiteout set in the main searchable
part of a btree node.
That means we can simplify and optimize some sort paths, and eliminate
an assertion that syzbot found:
- Unless we're in the btree node write path, it's always ok to drop
whiteouts when sorting
- When sorting for a btree node write, we drop the whiteout if it's not
from the unwritten whiteouts area, or if it's overwritten by a real
key at the same position.
This completely eliminates some tricky logic for propagating the
needs_whiteout flag: syzbot was able to hit the assertion that checked
that there shouldn't be more than one key at the same pos with
needs_whiteout set, likely due to a combination of flipping on
needs_whiteout on all written keys (they need whiteouts if overwritten),
combined with not always dropping unneeded whiteouts, and the tricky
logic in the sort path for preserving needs_whiteout that wasn't really
needed.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Small performance optimization; an open coded loop is better than rep ;
movsq for small copies.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Rust bindgen doesn't cope well with anonymous structs and unions. This
patch drops the fancy anonymous structs & unions in bkey_i that let us
use the same helpers for bkey_i and bkey_packed; since bkey_packed is an
internal type that's never exposed to outside code, it's only a minor
inconvenienc.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This improves the bkey_format calculation when splitting btree nodes.
Previously, we'd use a format calculated for the original node for the
lower of the two new nodes.
This was particularly bad on sequential insertions, where we iteratively
split the last btree node, whos format has to include KEY_MAX.
Now, we calculate formats precisely for the keys the two new nodes will
contain. This also should make splitting a bit more efficient, since
we're only copying keys once (from the original node to the new node,
instead of new node, replacement node, then upper split).
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This adds an inlined version of bch2_bkey_cmp_packed(), and uses it in
bch2_sort_keys(), where it's part of the inner loop.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
The main function of bch2_sort_repack_merge() was to call .key_normalize
on every key, which drops stale (cached) pointers - it hasn't actually
merged extents in quite some time.
But bch2_gc_gens() now works on individual keys - we used to gc old gens
by rewriting entire btree nodes. With that gone, there's no need for
internal btree code to be calling .key_normalize anymore.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Bkey noops were introduced to deal with trimming inline data extents in
place in the btree: if the u64s field of a bkey was 0, that u64 was a
noop and we'd start looking for the next bkey immediately after it.
But extent handling has been lifted above the btree - we no longer
modify existing extents in place in the btree, and the compatibilty code
for old style extent btree nodes is gone, so we can completely drop this
code.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
KEY_TYPE_discard used to be used for extent whiteouts, but when handling
over overlapping extents was lifted above the core btree code it became
unused. This patch updates various code to reflect that.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs has been aggressively migrating filesystems and btree nodes to
the new format for quite some time - this shouldn't affect anyone
anymore, and lets us delete a _lot_ of code. Also, it frees up
KEY_TYPE_discard for a new whiteout key type for snapshots.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
With various newer key types - stripe keys, inline data extents - the
old approach of calculating the maximum size of the value is becoming
more and more error prone. Better to switch to bkey_on_stack, which can
dynamically allocate if necessary to handle any size bkey.
In particular we also want to get rid of BKEY_EXTENT_VAL_U64s_MAX.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This only did anything in two places, and those can just be replaced
wiht bkey_cmp_left_packed()).
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Extent merging is currently broken, and will be reimplemented
differently soon - right now it only happens when btree nodes are being
compacted, which makes it difficult to test.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Recently the extent update path started emmiting 0 size whiteouts on
extent overwrite, as part of transitioning to moving extent handling
out of the core btree code.
Unfortunately, this broke the old code path that handles overlapping
extents when reading in btree nodes - it relies on sorting incomming
extents by start position, but the 0 size whiteouts broke that ordering.
Skipping over them before the main algorithm sees them fixes this.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This fixes a nasty memory corruption with other threads that are still
reading the btree node being compacted.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Previously, partial overwrites of existing extents were handled
implicitly by the btree code; when reading in a btree node, we'd do a
mergesort of the different bsets and detect and fix partially
overlapping extents during that mergesort.
That approach won't work with snapshots: this changes extents to work
like regular keys as far as the btree code is concerned, where a 0 size
KEY_TYPE_deleted whiteout will completely overwrite an existing extent.
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>
The whiteout compaction path - as opposed to just dropping whiteouts -
is now only needed for extents, and soon will only be needed for extent
btree nodes in the old format.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
This changes bch2_cut_front and bch2_cut_back so that they're able to
shorten the size of the value, and it also changes the extent update
path to update the accounting in the btree node when this happens.
When the size of the value is shortened, they zero out the space that's
no longer used, so it's interpreted as noops (as implemented in the last
patch).
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
For upcoming inline data extents, we're going to need to be able to
shorten the value of existing bkeys in the btree - and to make that work
we're going to be able to need to pad out the space the value previously
took up with something.
This patch changes the various code that iterates over bkeys to handle
k->u64s == 0 as meaning "skip the next 8 bytes".
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>
Prep work for extents with inline data
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bch2_bkey_normalize() modifies the value, and we were modifying the
original value in the src btree node - but, we're called without a write
lock held on the src node. Oops...
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>
