net/mlx5: E-Switch, Allow transition to offloads mode for ECPF

Currently, the e-switch driver requires going to legacy mode before
changing to the offloads mode. This makes sense for regular case as
the legacy mode is done by creating VFs.

However, it's problematic when ECPF is the eswitch manager. In such
case, ECPF will control the vports on peer host including the peer
PF and VFs. But ECPF doesn't need and shall not create VFs as the
VFs are created in the peer PF host.

Grant ECPF the ability to change from none to the offloads mode. Note
that currently the only way to go back to none mode is by unloading
the ECPF driver.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
This commit is contained in:
Bodong Wang 2018-12-20 17:28:06 -06:00 committed by Saeed Mahameed
parent a3888f33db
commit c96692fb8f

View File

@ -1250,7 +1250,8 @@ static int esw_offloads_start(struct mlx5_eswitch *esw,
{
int err, err1, num_vfs = esw->dev->priv.sriov.num_vfs;
if (esw->mode != SRIOV_LEGACY) {
if (esw->mode != SRIOV_LEGACY &&
!mlx5_core_is_ecpf_esw_manager(esw->dev)) {
NL_SET_ERR_MSG_MOD(extack,
"Can't set offloads mode, SRIOV legacy not enabled");
return -EINVAL;
@ -1846,7 +1847,8 @@ static int mlx5_devlink_eswitch_check(struct devlink *devlink)
if(!MLX5_ESWITCH_MANAGER(dev))
return -EPERM;
if (dev->priv.eswitch->mode == SRIOV_NONE)
if (dev->priv.eswitch->mode == SRIOV_NONE &&
!mlx5_core_is_ecpf_esw_manager(dev))
return -EOPNOTSUPP;
return 0;