On Mon, Mar 13, 2017 at 02:41:10PM +0800, Yanjun Zhu wrote:
>
>
> On 2017/3/13 14:32, Leon Romanovsky wrote:
> > On Mon, Mar 13, 2017 at 01:43:45AM -0400, Zhu Yanjun wrote:
> > > When rdma_accept fails, rdma_reject is called in it. As such, it is
> > > not necessary to execute rdma_reject again.
> > It is not always correct, according to the code, rdma_accept can fail
> > and will return EINVAL and skip calling to rdma_reject.
> Sure. When -EINVAL is returned, By this function cma_comp(id_priv,
> RDMA_CM_CONNECT), the connection
> is lost. As such, it does not matter whether sending rdma_reject or not.
Thanks for the explanation..
>
> Zhu Yanjun
> >
> > 3725 int rdma_accept(struct rdma_cm_id *id, struct rdma_conn_param *conn_param)
> > 3726 {
> > 3727 struct rdma_id_private *id_priv;
> > 3728 int ret;
> > 3729
> > 3730 id_priv = container_of(id, struct rdma_id_private, id);
> > 3731
> > 3732 id_priv->owner = task_pid_nr(current);
> > 3733
> > 3734 if (!cma_comp(id_priv, RDMA_CM_CONNECT))
> > 3735 return -EINVAL;
> > 3736
> >
> >
> > > Cc: Joe Jin <joe.jin@xxxxxxxxxx>
> > > Cc: Junxiao Bi <junxiao.bi@xxxxxxxxxx>
> > > Acked-by: Santosh Shilimkar <santosh.shilimkar@xxxxxxxxxx>
> > > Signed-off-by: Zhu Yanjun <yanjun.zhu@xxxxxxxxxx>
> > > ---
> > > net/rds/ib_cm.c | 5 ++---
> > > 1 file changed, 2 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c
> > > index ce3775a..4b9405c 100644
> > > --- a/net/rds/ib_cm.c
> > > +++ b/net/rds/ib_cm.c
> > > @@ -677,9 +677,8 @@ int rds_ib_cm_handle_connect(struct rdma_cm_id *cm_id,
> > > event->param.conn.initiator_depth);
> > >
> > > /* rdma_accept() calls rdma_reject() internally if it fails */
> > > - err = rdma_accept(cm_id, &conn_param);
> > > - if (err)
> > > - rds_ib_conn_error(conn, "rdma_accept failed (%d)\n", err);
> > > + if (rdma_accept(cm_id, &conn_param))
> > > + rds_ib_conn_error(conn, "rdma_accept failed\n");
> > >
> > > out:
> > > if (conn)
> > > --
> > > 2.7.4
> > >
>
Attachment:
signature.asc
Description: PGP signature
