Commit b5dcbf47 authored by Hal Rosenstock's avatar Hal Rosenstock Committed by Roland Dreier

[PATCH] IB: RMPP fixes

- Fix payload length of middle RMPP sent segments. Middle payload
  lengths should be 0 on the send side.

  (This is perhaps a compliance and should not be an interop issue as
  middle payload lengths are supposed to be ignored on receive).

- Fix length in first segment of multipacket sends

  (This is a compliance issue but does not affect at least OpenIB to
  OpenIB RMPP transfers).
Signed-off-by: default avatarHal Rosenstock <halr@voltaire.com>
Signed-off-by: default avatarRoland Dreier <rolandd@cisco.com>
parent 30a7e8ef
......@@ -593,7 +593,8 @@ static int send_next_seg(struct ib_mad_send_wr_private *mad_send_wr)
rmpp_mad->rmpp_hdr.paylen_newwin =
cpu_to_be32(mad_send_wr->total_seg *
(sizeof(struct ib_rmpp_mad) -
offsetof(struct ib_rmpp_mad, data)));
offsetof(struct ib_rmpp_mad, data)) -
mad_send_wr->pad);
mad_send_wr->sg_list[0].length = sizeof(struct ib_rmpp_mad);
} else {
mad_send_wr->send_wr.num_sge = 2;
......@@ -602,6 +603,7 @@ static int send_next_seg(struct ib_mad_send_wr_private *mad_send_wr)
mad_send_wr->sg_list[1].length = sizeof(struct ib_rmpp_mad) -
mad_send_wr->data_offset;
mad_send_wr->sg_list[1].lkey = mad_send_wr->sg_list[0].lkey;
rmpp_mad->rmpp_hdr.paylen_newwin = 0;
}
if (mad_send_wr->seg_num == mad_send_wr->total_seg) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment