Commit da6b0bb0 authored by Maor Gottlieb's avatar Maor Gottlieb Committed by Saeed Mahameed

net/mlx5: Lag, use steering to select the affinity port in LAG

Use the steering based solution for select the affinity port
when the LAG mode is based on hash policy and the device support
in port selection flow table.
Signed-off-by: default avatarMaor Gottlieb <maorg@nvidia.com>
Reviewed-by: default avatarMark Bloch <mbloch@nvidia.com>
Signed-off-by: default avatarSaeed Mahameed <saeedm@nvidia.com>
parent b7267869
...@@ -47,16 +47,21 @@ ...@@ -47,16 +47,21 @@
static DEFINE_SPINLOCK(lag_lock); static DEFINE_SPINLOCK(lag_lock);
static int mlx5_cmd_create_lag(struct mlx5_core_dev *dev, u8 remap_port1, static int mlx5_cmd_create_lag(struct mlx5_core_dev *dev, u8 remap_port1,
u8 remap_port2, bool shared_fdb) u8 remap_port2, bool shared_fdb, u8 flags)
{ {
u32 in[MLX5_ST_SZ_DW(create_lag_in)] = {}; u32 in[MLX5_ST_SZ_DW(create_lag_in)] = {};
void *lag_ctx = MLX5_ADDR_OF(create_lag_in, in, ctx); void *lag_ctx = MLX5_ADDR_OF(create_lag_in, in, ctx);
MLX5_SET(create_lag_in, in, opcode, MLX5_CMD_OP_CREATE_LAG); MLX5_SET(create_lag_in, in, opcode, MLX5_CMD_OP_CREATE_LAG);
MLX5_SET(lagc, lag_ctx, tx_remap_affinity_1, remap_port1);
MLX5_SET(lagc, lag_ctx, tx_remap_affinity_2, remap_port2);
MLX5_SET(lagc, lag_ctx, fdb_selection_mode, shared_fdb); MLX5_SET(lagc, lag_ctx, fdb_selection_mode, shared_fdb);
if (!(flags & MLX5_LAG_FLAG_HASH_BASED)) {
MLX5_SET(lagc, lag_ctx, tx_remap_affinity_1, remap_port1);
MLX5_SET(lagc, lag_ctx, tx_remap_affinity_2, remap_port2);
} else {
MLX5_SET(lagc, lag_ctx, port_select_mode,
MLX5_LAG_PORT_SELECT_MODE_PORT_SELECT_FT);
}
return mlx5_cmd_exec_in(dev, create_lag, in); return mlx5_cmd_exec_in(dev, create_lag, in);
} }
...@@ -199,6 +204,15 @@ static void mlx5_infer_tx_affinity_mapping(struct lag_tracker *tracker, ...@@ -199,6 +204,15 @@ static void mlx5_infer_tx_affinity_mapping(struct lag_tracker *tracker,
*port1 = 2; *port1 = 2;
} }
static int _mlx5_modify_lag(struct mlx5_lag *ldev, u8 v2p_port1, u8 v2p_port2)
{
struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev;
if (ldev->flags & MLX5_LAG_FLAG_HASH_BASED)
return mlx5_lag_port_sel_modify(ldev, v2p_port1, v2p_port2);
return mlx5_cmd_modify_lag(dev0, v2p_port1, v2p_port2);
}
void mlx5_modify_lag(struct mlx5_lag *ldev, void mlx5_modify_lag(struct mlx5_lag *ldev,
struct lag_tracker *tracker) struct lag_tracker *tracker)
{ {
...@@ -211,39 +225,56 @@ void mlx5_modify_lag(struct mlx5_lag *ldev, ...@@ -211,39 +225,56 @@ void mlx5_modify_lag(struct mlx5_lag *ldev,
if (v2p_port1 != ldev->v2p_map[MLX5_LAG_P1] || if (v2p_port1 != ldev->v2p_map[MLX5_LAG_P1] ||
v2p_port2 != ldev->v2p_map[MLX5_LAG_P2]) { v2p_port2 != ldev->v2p_map[MLX5_LAG_P2]) {
err = _mlx5_modify_lag(ldev, v2p_port1, v2p_port2);
if (err) {
mlx5_core_err(dev0,
"Failed to modify LAG (%d)\n",
err);
return;
}
ldev->v2p_map[MLX5_LAG_P1] = v2p_port1; ldev->v2p_map[MLX5_LAG_P1] = v2p_port1;
ldev->v2p_map[MLX5_LAG_P2] = v2p_port2; ldev->v2p_map[MLX5_LAG_P2] = v2p_port2;
mlx5_core_info(dev0, "modify lag map port 1:%d port 2:%d", mlx5_core_info(dev0, "modify lag map port 1:%d port 2:%d",
ldev->v2p_map[MLX5_LAG_P1], ldev->v2p_map[MLX5_LAG_P1],
ldev->v2p_map[MLX5_LAG_P2]); ldev->v2p_map[MLX5_LAG_P2]);
err = mlx5_cmd_modify_lag(dev0, v2p_port1, v2p_port2);
if (err)
mlx5_core_err(dev0,
"Failed to modify LAG (%d)\n",
err);
} }
} }
static void mlx5_lag_set_port_sel_mode(struct mlx5_lag *ldev,
struct lag_tracker *tracker, u8 *flags)
{
bool roce_lag = !!(*flags & MLX5_LAG_FLAG_ROCE);
struct lag_func *dev0 = &ldev->pf[MLX5_LAG_P1];
if (roce_lag ||
!MLX5_CAP_PORT_SELECTION(dev0->dev, port_select_flow_table) ||
tracker->tx_type != NETDEV_LAG_TX_TYPE_HASH)
return;
*flags |= MLX5_LAG_FLAG_HASH_BASED;
}
static char *get_str_port_sel_mode(u8 flags)
{
if (flags & MLX5_LAG_FLAG_HASH_BASED)
return "hash";
return "queue_affinity";
}
static int mlx5_create_lag(struct mlx5_lag *ldev, static int mlx5_create_lag(struct mlx5_lag *ldev,
struct lag_tracker *tracker, struct lag_tracker *tracker,
bool shared_fdb) bool shared_fdb, u8 flags)
{ {
struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev; struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev;
struct mlx5_core_dev *dev1 = ldev->pf[MLX5_LAG_P2].dev; struct mlx5_core_dev *dev1 = ldev->pf[MLX5_LAG_P2].dev;
u32 in[MLX5_ST_SZ_DW(destroy_lag_in)] = {}; u32 in[MLX5_ST_SZ_DW(destroy_lag_in)] = {};
int err; int err;
mlx5_infer_tx_affinity_mapping(tracker, &ldev->v2p_map[MLX5_LAG_P1], mlx5_core_info(dev0, "lag map port 1:%d port 2:%d shared_fdb:%d mode:%s",
&ldev->v2p_map[MLX5_LAG_P2]);
mlx5_core_info(dev0, "lag map port 1:%d port 2:%d shared_fdb:%d",
ldev->v2p_map[MLX5_LAG_P1], ldev->v2p_map[MLX5_LAG_P2], ldev->v2p_map[MLX5_LAG_P1], ldev->v2p_map[MLX5_LAG_P2],
shared_fdb); shared_fdb, get_str_port_sel_mode(flags));
err = mlx5_cmd_create_lag(dev0, ldev->v2p_map[MLX5_LAG_P1], err = mlx5_cmd_create_lag(dev0, ldev->v2p_map[MLX5_LAG_P1],
ldev->v2p_map[MLX5_LAG_P2], shared_fdb); ldev->v2p_map[MLX5_LAG_P2], shared_fdb, flags);
if (err) { if (err) {
mlx5_core_err(dev0, mlx5_core_err(dev0,
"Failed to create LAG (%d)\n", "Failed to create LAG (%d)\n",
...@@ -279,16 +310,32 @@ int mlx5_activate_lag(struct mlx5_lag *ldev, ...@@ -279,16 +310,32 @@ int mlx5_activate_lag(struct mlx5_lag *ldev,
struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev; struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev;
int err; int err;
err = mlx5_create_lag(ldev, tracker, shared_fdb); mlx5_infer_tx_affinity_mapping(tracker, &ldev->v2p_map[MLX5_LAG_P1],
&ldev->v2p_map[MLX5_LAG_P2]);
mlx5_lag_set_port_sel_mode(ldev, tracker, &flags);
if (flags & MLX5_LAG_FLAG_HASH_BASED) {
err = mlx5_lag_port_sel_create(ldev, tracker->hash_type,
ldev->v2p_map[MLX5_LAG_P1],
ldev->v2p_map[MLX5_LAG_P2]);
if (err) {
mlx5_core_err(dev0,
"Failed to create LAG port selection(%d)\n",
err);
return err;
}
}
err = mlx5_create_lag(ldev, tracker, shared_fdb, flags);
if (err) { if (err) {
if (roce_lag) { if (flags & MLX5_LAG_FLAG_HASH_BASED)
mlx5_lag_port_sel_destroy(ldev);
if (roce_lag)
mlx5_core_err(dev0, mlx5_core_err(dev0,
"Failed to activate RoCE LAG\n"); "Failed to activate RoCE LAG\n");
} else { else
mlx5_core_err(dev0, mlx5_core_err(dev0,
"Failed to activate VF LAG\n" "Failed to activate VF LAG\n"
"Make sure all VFs are unbound prior to VF LAG activation or deactivation\n"); "Make sure all VFs are unbound prior to VF LAG activation or deactivation\n");
}
return err; return err;
} }
...@@ -302,6 +349,7 @@ static int mlx5_deactivate_lag(struct mlx5_lag *ldev) ...@@ -302,6 +349,7 @@ static int mlx5_deactivate_lag(struct mlx5_lag *ldev)
struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev; struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev;
u32 in[MLX5_ST_SZ_DW(destroy_lag_in)] = {}; u32 in[MLX5_ST_SZ_DW(destroy_lag_in)] = {};
bool roce_lag = __mlx5_lag_is_roce(ldev); bool roce_lag = __mlx5_lag_is_roce(ldev);
u8 flags = ldev->flags;
int err; int err;
ldev->flags &= ~MLX5_LAG_MODE_FLAGS; ldev->flags &= ~MLX5_LAG_MODE_FLAGS;
...@@ -324,6 +372,8 @@ static int mlx5_deactivate_lag(struct mlx5_lag *ldev) ...@@ -324,6 +372,8 @@ static int mlx5_deactivate_lag(struct mlx5_lag *ldev)
"Failed to deactivate VF LAG; driver restart required\n" "Failed to deactivate VF LAG; driver restart required\n"
"Make sure all VFs are unbound prior to VF LAG activation or deactivation\n"); "Make sure all VFs are unbound prior to VF LAG activation or deactivation\n");
} }
} else if (flags & MLX5_LAG_FLAG_HASH_BASED) {
mlx5_lag_port_sel_destroy(ldev);
} }
return err; return err;
......
...@@ -18,10 +18,12 @@ enum { ...@@ -18,10 +18,12 @@ enum {
MLX5_LAG_FLAG_SRIOV = 1 << 1, MLX5_LAG_FLAG_SRIOV = 1 << 1,
MLX5_LAG_FLAG_MULTIPATH = 1 << 2, MLX5_LAG_FLAG_MULTIPATH = 1 << 2,
MLX5_LAG_FLAG_READY = 1 << 3, MLX5_LAG_FLAG_READY = 1 << 3,
MLX5_LAG_FLAG_HASH_BASED = 1 << 4,
}; };
#define MLX5_LAG_MODE_FLAGS (MLX5_LAG_FLAG_ROCE | MLX5_LAG_FLAG_SRIOV |\ #define MLX5_LAG_MODE_FLAGS (MLX5_LAG_FLAG_ROCE | MLX5_LAG_FLAG_SRIOV |\
MLX5_LAG_FLAG_MULTIPATH) MLX5_LAG_FLAG_MULTIPATH | \
MLX5_LAG_FLAG_HASH_BASED)
struct lag_func { struct lag_func {
struct mlx5_core_dev *dev; struct mlx5_core_dev *dev;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment