Add other eswitch support

[PATCH rdma-next 0/7] Add other eswitch support

Posted by Edward Srouji 3 months, 1 week ago

From: Patrisious Haddad <phaddad@nvidia.com>

From Patrisious:
When the device in switchdev mode, the RDMA device manages all the
vports which belong to its representors, which can lead to a situation
where the PF that is used to manage the RDMA device isn't the native PF
of some of the vports it manages.

Add infrastructure to allow the master PF to manage all the hardware
resources for the vports under its management.
Whereas currently the only such resource is RDMA TRANSPORT steering
domains.

That is done by adding new FW argument other_eswitch which is passed by
the driver to the FW to allow the master PF to properly manage vports
belonging to other native PF.

---
Patrisious Haddad (7):
      net/mlx5: Add OTHER_ESWITCH HW capabilities
      net/mlx5: fs, Add other_eswitch support for steering tables
      net/mlx5: fs, set non default device per namespace
      RDMA/mlx5: Change default device for LAG slaves in RDMA TRANSPORT namespaces
      RDMA/mlx5: Add other_eswitch support for devx destruction
      RDMA/mlx5: Refactor _get_prio() function
      RDMA/mlx5: Add other eswitch support to userspace tables

 drivers/infiniband/hw/mlx5/devx.c                 | 14 +++++
 drivers/infiniband/hw/mlx5/fs.c                   | 65 ++++++++++++--------
 drivers/infiniband/hw/mlx5/ib_rep.c               | 74 ++++++++++++++++++++++-
 drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c  | 31 ++++++++++
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 74 ++++++++++++++++++++---
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.h | 19 +-----
 include/linux/mlx5/fs.h                           | 24 ++++++++
 include/linux/mlx5/mlx5_ifc.h                     | 47 +++++++++-----
 8 files changed, 277 insertions(+), 71 deletions(-)
---
base-commit: be180c847a6db6646d7bb4740a1d73f6f67d1030
change-id: 20251029-support-other-eswitch-0ed3232ce04e

Best regards,
-- 
Edward Srouji <edwards@nvidia.com>

Re: [PATCH rdma-next 0/7] Add other eswitch support

Posted by Leon Romanovsky 3 months ago

On Wed, Oct 29, 2025 at 05:42:52PM +0200, Edward Srouji wrote:
> From: Patrisious Haddad <phaddad@nvidia.com>
> 
> From Patrisious:
> When the device in switchdev mode, the RDMA device manages all the
> vports which belong to its representors, which can lead to a situation
> where the PF that is used to manage the RDMA device isn't the native PF
> of some of the vports it manages.
> 
> Add infrastructure to allow the master PF to manage all the hardware
> resources for the vports under its management.
> Whereas currently the only such resource is RDMA TRANSPORT steering
> domains.
> 
> That is done by adding new FW argument other_eswitch which is passed by
> the driver to the FW to allow the master PF to properly manage vports
> belonging to other native PF.
> 
> ---
> Patrisious Haddad (7):
>       net/mlx5: Add OTHER_ESWITCH HW capabilities
>       net/mlx5: fs, Add other_eswitch support for steering tables
>       net/mlx5: fs, set non default device per namespace
>       RDMA/mlx5: Change default device for LAG slaves in RDMA TRANSPORT namespaces
>       RDMA/mlx5: Add other_eswitch support for devx destruction
>       RDMA/mlx5: Refactor _get_prio() function
>       RDMA/mlx5: Add other eswitch support to userspace tables
> 

Thanks, applied.