RDMA/hns: Fix wrong PBL offset when VA is not aligned to PAGE_SIZE

ROCE uses "VA % buf_page_size" to caclulate the offset in the PBL's first
page, the actual PA corresponding to the MR's VA is equal to MR's PA plus
this offset. The first PA in PBL has already been aligned to PAGE_SIZE
after calling ib_umem_get(), but the MR's VA may not. If the buf_page_size
is smaller than the PAGE_SIZE, this will lead the HW to access the wrong
memory because the offset is smaller than expected.

Fixes: 9b2cf76c9f05 ("RDMA/hns: Optimize PBL buffer allocation process")
Link: https://lore.kernel.org/r/1594726935-45666-1-git-send-email-liweihang@huawei.com
Signed-off-by: Xi Wang <wangxi11@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index 0e71ebee9e5202d7d2e09fa5b2682b1e2185e547..6b226a5eb7db6fbc808c6420d81675932f437107 100644 (file)
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -120,7 +120,7 @@ static int alloc_mr_pbl(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr,
 
        mr->pbl_hop_num = is_fast ? 1 : hr_dev->caps.pbl_hop_num;
        buf_attr.page_shift = is_fast ? PAGE_SHIFT :
-                             hr_dev->caps.pbl_buf_pg_sz + HNS_HW_PAGE_SHIFT;
+                             hr_dev->caps.pbl_buf_pg_sz + PAGE_SHIFT;
        buf_attr.region[0].size = length;
        buf_attr.region[0].hopnum = mr->pbl_hop_num;
        buf_attr.region_count = 1;