diff options
| author | Kent Overstreet <kent.overstreet@gmail.com> | 2022-03-06 21:17:43 -0500 |
|---|---|---|
| committer | Kent Overstreet <kent.overstreet@gmail.com> | 2022-03-12 20:14:13 -0500 |
| commit | 354d8bc719546419d8a051f100a041df45481b5c (patch) | |
| tree | 506603b75b611d0f56b768a782bd113b3e1ad4ed | |
| parent | 0094a8e02292d24ab6e2a5cbf5afc31902b57e87 (diff) | |
bcachefs: Fix lock ordering under traverse_all()
traverse_all() traverses btree paths in sorted order, so it should never
see transaction restarts due to lock ordering violations. But some code
in __bch2_btree_path_upgrade(), while necessary when not running under
traverse_all(), was causing some confusing lock ordering violations -
disabling this code under traverse_all() will let us put in some more
assertions.
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
| -rw-r--r-- | fs/bcachefs/btree_iter.c | 17 |
1 files changed, 9 insertions, 8 deletions
diff --git a/fs/bcachefs/btree_iter.c b/fs/bcachefs/btree_iter.c index 847ea7244121..23923503dc6e 100644 --- a/fs/bcachefs/btree_iter.c +++ b/fs/bcachefs/btree_iter.c @@ -486,14 +486,15 @@ bool __bch2_btree_path_upgrade(struct btree_trans *trans, * before interior nodes - now that's handled by * bch2_btree_path_traverse_all(). */ - trans_for_each_path(trans, linked) - if (linked != path && - linked->cached == path->cached && - linked->btree_id == path->btree_id && - linked->locks_want < new_locks_want) { - linked->locks_want = new_locks_want; - btree_path_get_locks(trans, linked, true); - } + if (!path->cached && !trans->in_traverse_all) + trans_for_each_path(trans, linked) + if (linked != path && + linked->cached == path->cached && + linked->btree_id == path->btree_id && + linked->locks_want < new_locks_want) { + linked->locks_want = new_locks_want; + btree_path_get_locks(trans, linked, true); + } return false; } |