From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jason Gunthorpe Subject: Re: [PATCH mlx5-next 08/10] net/mlx5: Support extended destination format in flow steering command Date: Mon, 10 Dec 2018 16:16:20 +0000 Message-ID: <20181210161614.GG3459@mellanox.com> References: <20181210030442.7543-1-saeedm@mellanox.com> <20181210030442.7543-9-saeedm@mellanox.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <20181210030442.7543-9-saeedm@mellanox.com> Content-Language: en-US Content-ID: <9F6E58AC58DF0D438B98ECFD51BB774B@eurprd05.prod.outlook.com> Sender: netdev-owner@vger.kernel.org To: Saeed Mahameed Cc: Leon Romanovsky , "netdev@vger.kernel.org" , "linux-rdma@vger.kernel.org" , Eli Britstein List-Id: linux-rdma@vger.kernel.org On Sun, Dec 09, 2018 at 07:04:40PM -0800, Saeed Mahameed wrote: > From: Eli Britstein >=20 > Update the flow steering command formatting according to the extended > destination API. > Note that the FW dictates that multi destination FTEs that involve at > least one encap must use the extended destination format, while single > destination ones must use the legacy format. > Using extended destination format requires FW support. Check for its > capabilities and return error if not supported. >=20 > Signed-off-by: Eli Britstein > Reviewed-by: Or Gerlitz > Reviewed-by: Oz Shlomo > Signed-off-by: Saeed Mahameed > .../net/ethernet/mellanox/mlx5/core/fs_cmd.c | 80 +++++++++++++++++-- > include/linux/mlx5/fs.h | 2 + > 2 files changed, 75 insertions(+), 7 deletions(-) >=20 > diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/n= et/ethernet/mellanox/mlx5/core/fs_cmd.c > index dda63dedaa49..a18fbea7a931 100644 > +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c > @@ -308,22 +308,68 @@ static int mlx5_cmd_destroy_flow_group(struct mlx5_= core_dev *dev, > return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); > } > =20 > +static int mlx5_set_extended_dest(struct mlx5_core_dev *dev, > + struct fs_fte *fte, bool *extended_dest) > +{ > + int fw_log_max_fdb_encap_uplink =3D > + MLX5_CAP_ESW(dev, log_max_fdb_encap_uplink); > + int num_fwd_destinations =3D 0; > + struct mlx5_flow_rule *dst; > + int num_encap =3D 0; > + > + *extended_dest =3D false; > + if (!(fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST)) > + return 0; > + > + list_for_each_entry(dst, &fte->node.children, node.list) { > + if (dst->dest_attr.type =3D=3D MLX5_FLOW_DESTINATION_TYPE_COUNTER) > + continue; > + if (dst->dest_attr.type =3D=3D MLX5_FLOW_DESTINATION_TYPE_VPORT && > + dst->dest_attr.vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID) > + num_encap++; > + num_fwd_destinations++; > + } > + if (num_fwd_destinations > 1 && num_encap > 0) > + *extended_dest =3D true; > + > + if (*extended_dest && !fw_log_max_fdb_encap_uplink) { > + pr_warn("FW does not support extended destination"); > + return -EOPNOTSUPP; > + } > + if (num_encap > (1 << fw_log_max_fdb_encap_uplink)) { > + pr_warn("FW does not support more than %d encaps", > + 1 << fw_log_max_fdb_encap_uplink); Drivers should avoid using pr_ if they have an struct device, in this case &dev->pdev->dev .. This is probably a comment for the entire core code directory I suppose? Jason