block: Always check length of all iov entries in blk_rq_map_user_iov()
Ben Hutchings [Sun, 13 Nov 2011 18:58:09 +0000 (19:58 +0100)]
commit 6b76106d8ef31111d6fc469564b83b5f5542794f upstream.

Even after commit 5478755616ae2ef1ce144dded589b62b2a50d575
("block: check for proper length of iov entries earlier ...")
we still won't check for zero-length entries after an unaligned
entry.  Remove the break-statement, so all entries are checked.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

block/blk-map.c

index e663ac2..164cd00 100644 (file)
@@ -204,10 +204,11 @@ int blk_rq_map_user_iov(struct request_queue *q, struct request *rq,
                if (!iov[i].iov_len)
                        return -EINVAL;
 
-               if (uaddr & queue_dma_alignment(q)) {
+               /*
+                * Keep going so we check length of all segments
+                */
+               if (uaddr & queue_dma_alignment(q))
                        unaligned = 1;
-                       break;
-               }
        }
 
        if (unaligned || (q->dma_pad_mask & len) || map_data)