btrfs: fix search key advancing condition
authorNaohiro Aota <naota@elisp.net>
Tue, 30 Jun 2015 02:25:43 +0000 (11:25 +0900)
committerChris Mason <clm@fb.com>
Sun, 9 Aug 2015 13:17:02 +0000 (06:17 -0700)
commitdd81d459a37d73cfa39896bd070e7b92e66e3628
tree993856780c7913312bbe2676ffa721ab11b1091e
parentd6589101b67a55107652050dfbf414403a93e351
btrfs: fix search key advancing condition

The search key advancing condition used in copy_to_sk() is loose. It can
advance the key even if it reaches sk->max_*: e.g. when the max key = (512,
1024, -1) and the current key = (512, 1025, 10), it increments the
offset by 1, continues hopeless search from (512, 1025, 11). This issue
make ioctl() to take unexpectedly long time scanning all the leaf a blocks
one by one.

This commit fix the problem using standard way of key comparison:
btrfs_comp_cpu_keys()

Signed-off-by: Naohiro Aota <naota@elisp.net>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
fs/btrfs/ioctl.c