summaryrefslogtreecommitdiff
path: root/drivers/infiniband/hw/mlx4/qp.c
diff options
context:
space:
mode:
authorJack Morgenstein <jackm@dev.mellanox.co.il>2009-05-07 21:35:13 -0700
committerRoland Dreier <rolandd@cisco.com>2009-05-07 21:35:13 -0700
commit2b6b7d4be487bada8c727df829c25068c7b5e5a3 (patch)
treed1fa04daa97cafe89f04b9f16c8a58925f364203 /drivers/infiniband/hw/mlx4/qp.c
parent56a50adda49b2020156616c4eb15353e0f9ad7de (diff)
IB/mlx4: Don't overwrite fast registration page list when posting work request
The low-level mlx4 driver modified the page-list addresses for fast register work requests post send to big-endian, and set a "present" bit. This caused problems later when the consumer attempted to unmap the pages using the page-list (using the list addresses which were assumed to be still in CPU-endian order). Fix the mlx4 driver to allocate two buffers and use a private buffer for the hardware-format bus addresses. This patch fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1571>, an NFS/RDMA server crash. The cause of the crash was found by Vu Pham of Mellanox. The fix is along the lines suggested by Steve Wise in comment #21 in bug 1571. Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
Diffstat (limited to 'drivers/infiniband/hw/mlx4/qp.c')
-rw-r--r--drivers/infiniband/hw/mlx4/qp.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index f385a24d31d2..20724aee76f4 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1365,7 +1365,7 @@ static void set_fmr_seg(struct mlx4_wqe_fmr_seg *fseg, struct ib_send_wr *wr)
int i;
for (i = 0; i < wr->wr.fast_reg.page_list_len; ++i)
- wr->wr.fast_reg.page_list->page_list[i] =
+ mfrpl->mapped_page_list[i] =
cpu_to_be64(wr->wr.fast_reg.page_list->page_list[i] |
MLX4_MTT_FLAG_PRESENT);