diff options
| author | Kent Overstreet <kent.overstreet@linux.dev> | 2023-08-06 16:04:37 +0200 |
|---|---|---|
| committer | Kent Overstreet <kent.overstreet@linux.dev> | 2023-10-22 23:10:09 +0200 |
| commit | e08e63e44e65761e6eb53fcd568d194f607daa61 (patch) | |
| tree | 9128db786cae25b4eddfaccff21a3625885ec946 /fs/bcachefs/btree_io.c | |
| parent | bcachefs: kill EBUG_ON() redefinition in bkey.c (diff) | |
| download | linux-e08e63e44e65761e6eb53fcd568d194f607daa61.tar.xz linux-e08e63e44e65761e6eb53fcd568d194f607daa61.zip | |
bcachefs: BCH_COMPAT_bformat_overflow_done no longer required
Awhile back, we changed bkey_format generation to ensure that the packed
representation could never represent fields larger than the unpacked
representation.
This was to ensure that bkey_packed_successor() always gave a sensible
result, but in the current code bkey_packed_successor() is only used in
a debug assertion - not for anything important.
This kills the requirement that we've gotten rid of those weird bkey
formats, and instead changes the assertion to check if we're dealing
with an old weird bkey format.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Diffstat (limited to 'fs/bcachefs/btree_io.c')
| -rw-r--r-- | fs/bcachefs/btree_io.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/fs/bcachefs/btree_io.c b/fs/bcachefs/btree_io.c index a3da5b4bcd21..cba3c081b1d0 100644 --- a/fs/bcachefs/btree_io.c +++ b/fs/bcachefs/btree_io.c @@ -779,7 +779,7 @@ static int validate_bset(struct bch_fs *c, struct bch_dev *ca, compat_btree_node(b->c.level, b->c.btree_id, version, BSET_BIG_ENDIAN(i), write, bn); - btree_err_on(bch2_bkey_format_validate(&bn->format, &buf1), + btree_err_on(bch2_bkey_format_invalid(c, &bn->format, write, &buf1), -BCH_ERR_btree_node_read_err_bad_node, c, ca, b, i, "invalid bkey format: %s\n %s", buf1.buf, (printbuf_reset(&buf2), |