btrfs: abort transaction when sibling keys check fails for leaves
authorFilipe Manana <fdmanana@suse.com>
Wed, 26 Apr 2023 10:51:35 +0000 (11:51 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 9 Jun 2023 08:34:07 +0000 (10:34 +0200)
commit0433baa89318728b5e451a67b8d16f1d21f5e2dd
tree943b421c558f722d0f5ede88f95154a87df8a33e
parente2d161c539a4e1d48e4575d9fbcbcd09762899d6
btrfs: abort transaction when sibling keys check fails for leaves

[ Upstream commit 9ae5afd02a03d4e22a17a9609b19400b77c36273 ]

If the sibling keys check fails before we move keys from one sibling
leaf to another, we are not aborting the transaction - we leave that to
some higher level caller of btrfs_search_slot() (or anything else that
uses it to insert items into a b+tree).

This means that the transaction abort will provide a stack trace that
omits the b+tree modification call chain. So change this to immediately
abort the transaction and therefore get a more useful stack trace that
shows us the call chain in the bt+tree modification code.

It's also important to immediately abort the transaction just in case
some higher level caller is not doing it, as this indicates a very
serious corruption and we should stop the possibility of doing further
damage.

Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/btrfs/ctree.c