diff options
author | Shun Hao <shunh@nvidia.com> | 2022-01-17 15:01:12 +0300 |
---|---|---|
committer | Saeed Mahameed <saeedm@nvidia.com> | 2022-03-10 00:33:03 +0300 |
commit | aa818fbf8f36e8c6f3e608ea960567906c2d6112 (patch) | |
tree | f457f09cfd35a2abf73582ff16ba9e0fbf92e34f /drivers | |
parent | 32071187e9fb18da62f5be569bd2ea0d7a981ee8 (diff) | |
download | linux-aa818fbf8f36e8c6f3e608ea960567906c2d6112.tar.xz |
net/mlx5: DR, Align mlx5dv_dr API vport action with FW behavior
This aligns the behavior with FW when creating an FDB rule with wire
vport destination but no source port matching. Until now such rules
would fail on internal DR RX rule creation since the source and
destination are the wire vport.
The new behavior is the same as done on FW steering, if destination is
wire, we will create both TX and RX rules, but the RX packet coming from
wire will be dropped due to loopback not supported.
Signed-off-by: Shun Hao <shunh@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Diffstat (limited to 'drivers')
-rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c | 12 |
1 files changed, 3 insertions, 9 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c index c61a5e83c78c..743422acc3d8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c @@ -669,15 +669,9 @@ int mlx5dr_actions_build_ste_arr(struct mlx5dr_matcher *matcher, case DR_ACTION_TYP_VPORT: attr.hit_gvmi = action->vport->caps->vhca_gvmi; dest_action = action; - if (rx_rule) { - if (action->vport->caps->num == MLX5_VPORT_UPLINK) { - mlx5dr_dbg(dmn, "Device doesn't support Loopback on WIRE vport\n"); - return -EOPNOTSUPP; - } - attr.final_icm_addr = action->vport->caps->icm_address_rx; - } else { - attr.final_icm_addr = action->vport->caps->icm_address_tx; - } + attr.final_icm_addr = rx_rule ? + action->vport->caps->icm_address_rx : + action->vport->caps->icm_address_tx; break; case DR_ACTION_TYP_POP_VLAN: if (!rx_rule && !(dmn->ste_ctx->actions_caps & |