If prev_badblocks() returns '-1', it means no valid badblocks record
before the checking range. It doesn't make sense to check whether
the input checking range is overlapped with the non-existed invalid
front range.
This patch checkes whether 'prev >= 0' is true before calling
overlap_front(), to void such invalid operations.
Fixes: 3ea3354cb9f0 ("badblocks: improve badblocks_check() for multiple ranges handling")
Reported-and-tested-by: Ira Weiny <ira.weiny@xxxxxxxxx>
Signed-off-by: Coly Li <colyli@xxxxxxx>
Link: https://lore.kernel.org/nvdimm/3035e75a-9be0-4bc3-8d4a-6e52c207f277@xxxxxxxxxxxxx/
Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
Cc: Geliang Tang <geliang.tang@xxxxxxxx>
Cc: Hannes Reinecke <hare@xxxxxxx>
Cc: Jens Axboe <axboe@xxxxxxxxx>
Cc: NeilBrown <neilb@xxxxxxx>
Cc: Vishal L Verma <vishal.l.verma@xxxxxxxxx>
Cc: Xiao Ni <xni@xxxxxxxxxx>
---
block/badblocks.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/block/badblocks.c b/block/badblocks.c
index fc92d4e18aa3..db4ec8b9b2a8 100644
--- a/block/badblocks.c
+++ b/block/badblocks.c
@@ -1312,12 +1312,14 @@ static int _badblocks_check(struct badblocks *bb, sector_t s, int sectors,
prev = prev_badblocks(bb, &bad, hint);
/* start after all badblocks */
- if ((prev + 1) >= bb->count && !overlap_front(bb, prev, &bad)) {
+ if ((prev >= 0) &&
+ ((prev + 1) >= bb->count) && !overlap_front(bb, prev, &bad)) {
len = sectors;
goto update_sectors;
}
- if (overlap_front(bb, prev, &bad)) {
+ /* Overlapped with front badblocks record */
+ if ((prev >= 0) && overlap_front(bb, prev, &bad)) {
if (BB_ACK(p[prev]))
acked_badblocks++;
else
--
2.35.3
