| Message ID | 1433770345-17354-1-git-send-email-monis@mellanox.com (mailing list archive) |
|---|---|
| State | Rejected |
| Headers | show |
On Mon, Jun 08, 2015 at 04:32:25PM +0300, Moni Shoua wrote: > The flag RDMA_CORE_CAP_IB_SA is set when link layer is InfiniBand but > also when link layer is Ethernet. Why? Ethernet does not have an SA nor need one. > > Therefore, link layer is not implied by > this flag. Of course not. This flag is explicitly saying this port needs to use the IB SA to resolve the route information. cma_resolve_ib_route finds the route information by issuing an SA query. Therefore it should only be called on ports which require an SA. Was this change inspired by a bug being found? If so it seems the underlying device needs to be fixed to not set the RDMA_CORE_CAP_IB_SA flag when the link layer is ethernet. Ira > > --- > drivers/infiniband/core/cma.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c > index 3b943b7..f96d6fd 100644 > --- a/drivers/infiniband/core/cma.c > +++ b/drivers/infiniband/core/cma.c > @@ -1993,7 +1993,7 @@ int rdma_resolve_route(struct rdma_cm_id *id, int timeout_ms) > return -EINVAL; > > atomic_inc(&id_priv->refcount); > - if (rdma_cap_ib_sa(id->device, id->port_num)) > + if (rdma_protocol_ib(id->device, id->port_num)) > ret = cma_resolve_ib_route(id_priv, timeout_ms); > else if (rdma_protocol_roce(id->device, id->port_num)) > ret = cma_resolve_iboe_route(id_priv); > -- > 1.7.1 > -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> The flag RDMA_CORE_CAP_IB_SA is set when link layer is InfiniBand but > also when link layer is Ethernet. Therefore, link layer is not implied by This flag should NOT be set for Ethernet. What is setting this? > this flag. > > --- > drivers/infiniband/core/cma.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c > index 3b943b7..f96d6fd 100644 > --- a/drivers/infiniband/core/cma.c > +++ b/drivers/infiniband/core/cma.c > @@ -1993,7 +1993,7 @@ int rdma_resolve_route(struct rdma_cm_id *id, int > timeout_ms) > return -EINVAL; > > atomic_inc(&id_priv->refcount); > - if (rdma_cap_ib_sa(id->device, id->port_num)) > + if (rdma_protocol_ib(id->device, id->port_num)) This change looks wrong. The check here has nothing to do with the link layer. It is stating that route resolution must be done using an SA. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 3b943b7..f96d6fd 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1993,7 +1993,7 @@ int rdma_resolve_route(struct rdma_cm_id *id, int timeout_ms) return -EINVAL; atomic_inc(&id_priv->refcount); - if (rdma_cap_ib_sa(id->device, id->port_num)) + if (rdma_protocol_ib(id->device, id->port_num)) ret = cma_resolve_ib_route(id_priv, timeout_ms); else if (rdma_protocol_roce(id->device, id->port_num)) ret = cma_resolve_iboe_route(id_priv);