From a13ccb04af4aa6632c11d59ddf6555aa80ffb139 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Mon, 23 Feb 2015 19:50:47 +0100
Subject: s390/scm_block: fix off by one during cluster reservation
We increase the msb_count after we're finished building the request.
That way we can always access the current request via
scmrq->request[msb_count] . But once the request is started we need
to make sure that the array index stays below msb_count.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
drivers/s390/block/scm_blk_cluster.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
(limited to 'drivers/s390')
diff --git a/drivers/s390/block/scm_blk_cluster.c b/drivers/s390/block/scm_blk_cluster.c
index 09db45296eed..7497ddde2dd6 100644
--- a/drivers/s390/block/scm_blk_cluster.c
+++ b/drivers/s390/block/scm_blk_cluster.c
@@ -92,7 +92,7 @@ bool scm_reserve_cluster(struct scm_request *scmrq)
add = 0;
continue;
}
- for (pos = 0; pos <= iter->aob->request.msb_count; pos++) {
+ for (pos = 0; pos < iter->aob->request.msb_count; pos++) {
if (clusters_intersect(req, iter->request[pos]) &&
(rq_data_dir(req) == WRITE ||
rq_data_dir(iter->request[pos]) == WRITE)) {
--
cgit v1.2.1