net/mlx5e: Add recovery flow for tx devlink health reporter for unhealthy PTP SQ
authorRahul Rameshbabu <rrameshbabu@nvidia.com>
Wed, 9 Aug 2023 04:10:21 +0000 (21:10 -0700)
committerSaeed Mahameed <saeedm@nvidia.com>
Mon, 14 Aug 2023 21:40:20 +0000 (14:40 -0700)
commit53b836a44db4259b94ffcfff321fb3d63f976b76
treea4ecb2462bb2fd669c0ceb0ab4af4a0ab2270882
parent3178308ad4ca38955cad684d235153d4939f1fcd
net/mlx5e: Add recovery flow for tx devlink health reporter for unhealthy PTP SQ

A new check for the tx devlink health reporter is introduced for
determining when the PTP port timestamping SQ is considered unhealthy. If
there are enough CQEs considered never to be delivered, the space that can
be utilized on the SQ decreases significantly, impacting performance and
usability of the SQ. The health reporter is triggered when the number of
likely never delivered port timestamping CQEs that utilize the space of the
PTP SQ is greater than 93.75% of the total capacity of the SQ. A devlink
health reporter recover method is also provided for this specific TX error
context that restarts the PTP SQ.

Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Documentation/networking/devlink/mlx5.rst
drivers/net/ethernet/mellanox/mlx5/core/en/health.h
drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c
drivers/net/ethernet/mellanox/mlx5/core/en/ptp.h
drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c