net/mlx5: SF, Use all available cpu for setting cpu affinity

Browse source

Currently all SFs are using the same CPUs. Spreading SF over CPUs, in
round-robin manner, in order to achieve better distribution of the SFs
over available CPUs.

Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

...

This commit is contained in:

Shay Drory 2021-11-23 12:50:19 +02:00 • committed by Saeed Mahameed

parent 79b60ca83b

commit 061f5b2358

5 changed files with 201 additions and 34 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

11

drivers/net/ethernet/mellanox/mlx5/core/eq.c

View file

139

drivers/net/ethernet/mellanox/mlx5/core/irq_affinity.c

View file

13

drivers/net/ethernet/mellanox/mlx5/core/mlx5_irq.h

View file

64

drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c

View file

8

drivers/net/ethernet/mellanox/mlx5/core/pci_irq.h

View file