RDMA/hns: Fix the bug with rq sge

BugLink: https://bugs.launchpad.net/bugs/1770974
When received multiply rq sge, it should tag the
invalid lkey for the last non-zero length sge
when have some sges' length are zero. This patch
fixes it.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
(cherry picked from commit 778cc5a8b75eee62d330059a2655b515cda43278)
Signed-off-by: dann frazier <dann.frazier@canonical.com>
Acked-by: Kleber Sacilotto de Souza <kleber.souza@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Khalid Elmously <khalid.elmously@canonical.com>

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 82afc5b47a2acfff58e916e14de99f96b5fa88c9..a46c15d3f6a54096f3064923fdd05045f6570253 100644 (file)
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -547,8 +547,8 @@ static int hns_roce_v2_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
                }
 
                if (i < hr_qp->rq.max_gs) {
-                       dseg[i].lkey = cpu_to_le32(HNS_ROCE_INVALID_LKEY);
-                       dseg[i].addr = 0;
+                       dseg->lkey = cpu_to_le32(HNS_ROCE_INVALID_LKEY);
+                       dseg->addr = 0;
                }
 
                /* rq support inline data */