net/mlx5: DR, Add support for isolate_vl_tc QP
authorYevgeny Kliteynik <kliteyn@nvidia.com>
Mon, 2 Nov 2020 23:31:53 +0000 (01:31 +0200)
committerSaeed Mahameed <saeedm@nvidia.com>
Tue, 20 Apr 2021 03:17:46 +0000 (20:17 -0700)
commitaeacb52a8de7046be5399ba311f49bce96e1b269
tree3914c07151a087b4c539557bded653f0ddeb9725
parent7304d603a57a1edecfecfbcc26f85edcda4cae81
net/mlx5: DR, Add support for isolate_vl_tc QP

When using SW steering, rule insertion rate depends on the RDMA RC QP
performance used for writing to the ICM. During stress this QP is competing
on the HW resources with all the other QPs that are used to send data.
To protect SW steering QP's performance in such cases, we set this QP to
use isolated VL. The VL number is reserved by FW and is not exposed to the
driver.
Support for this QP on isolated VL exists only when both force-loopback and
isolate_vl_tc capabilities are set.

Signed-off-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c
drivers/net/ethernet/mellanox/mlx5/core/steering/dr_send.c
drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h
include/linux/mlx5/mlx5_ifc.h