net/mlx5: fix tunnel rules validation on VF representor

MLX5 PMD implicitly adds vxlan_decap flow action to tunnel offload
match type rules. However, VXLAN decap action on VF representors is
not supported on MLX5 PMD hardware.

The patch rejects attempt to create tunnel offload flow rules on VF
representor.

Refer commit 9c4971e5231d ("net/mlx5: update VLAN and encap actions validation")

Fixes: 4ec6360de37d ("net/mlx5: implement tunnel offload")
Cc: stable@dpdk.org
Signed-off-by: Gregory Etelson <getelson@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>

diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 7c2b389c630e4daebb10c936d9bf57f61aa95b25..4f638e24ad171eba9cfc387639a6c40b67f58ebf 100644 (file)
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -5280,6 +5280,11 @@ flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
        } else {
                tunnel = NULL;
        }
+       if (tunnel && priv->representor)
+               return rte_flow_error_set(error, ENOTSUP,
+                                         RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL,
+                                         "decap not supported "
+                                         "for VF representor");
        grp_info.std_tbl_fix = tunnel_use_standard_attr_group_translate
                                (dev, tunnel, attr, items, actions);
        ret = flow_dv_validate_attributes(dev, tunnel, attr, &grp_info, error);