Merge branch 'mlx5-bulk-flow-stats-sriov-tc-offloads'

Saeed Mahameed says:

====================
Mellanox 100G mlx5 Bulk flow statistics and SRIOV TC offloads

This series from Amir and Or deals with two enhancements for the mlx5 TC offloads.

The 1st two patches add bulk reading of flow counters. Few bulk counter queries are
used instead of issuing thousands firmware commands per second to get statistics of all
flows set to HW.

The next patches add TC based SRIOV offloading to mlx5, as a follow up for the e-switch
offloads mode and the VF representors. When the e-switch is set to the (new) "offloads"
mode, we can now offload TC/flower drop and forward rules, the forward action we offload
is TC mirred/redirect.

The above is done by the VF representor netdevices exporting the setup_tc ndo where from
there we're re-using and enhancing the existing mlx5 TC offloads sub-module which now
works for both the NIC and the SRIOV cases.

The series is applied on top b38a75d2d324 ('mlxsw: core: Trace EMAD messages')
and it has no merge issues with the on-going net submission ('mlx5 tx timeout watchdog fixes')

V2:
    - Fixed compilation warning.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 70 insertions, 7 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 1842dfb4636b..a357e8eeeed8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -38,6 +38,54 @@
 #include "mlx5_core.h"
 #include "eswitch.h"
 
+enum {
+	FDB_FAST_PATH = 0,
+	FDB_SLOW_PATH
+};
+
+struct mlx5_flow_rule *
+mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw,
+				struct mlx5_flow_spec *spec,
+				u32 action, u32 src_vport, u32 dst_vport)
+{
+	struct mlx5_flow_destination dest = { 0 };
+	struct mlx5_fc *counter = NULL;
+	struct mlx5_flow_rule *rule;
+	void *misc;
+
+	if (esw->mode != SRIOV_OFFLOADS)
+		return ERR_PTR(-EOPNOTSUPP);
+
+	if (action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) {
+		dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
+		dest.vport_num = dst_vport;
+		action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
+	} else if (action & MLX5_FLOW_CONTEXT_ACTION_COUNT) {
+		counter = mlx5_fc_create(esw->dev, true);
+		if (IS_ERR(counter))
+			return ERR_CAST(counter);
+		dest.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER;
+		dest.counter = counter;
+	}
+
+	misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters);
+	MLX5_SET(fte_match_set_misc, misc, source_port, src_vport);
+
+	misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters);
+	MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
+
+	spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS |
+				      MLX5_MATCH_MISC_PARAMETERS;
+
+	rule = mlx5_add_flow_rule((struct mlx5_flow_table *)esw->fdb_table.fdb,
+				  spec, action, 0, &dest);
+
+	if (IS_ERR(rule))
+		mlx5_fc_destroy(esw->dev, counter);
+
+	return rule;
+}
+
 static struct mlx5_flow_rule *
 mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, int vport, u32 sqn)
 {
@@ -149,7 +197,7 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw)
 	dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
 	dest.vport_num = 0;
 
-	flow_rule = mlx5_add_flow_rule(esw->fdb_table.fdb, spec,
+	flow_rule = mlx5_add_flow_rule(esw->fdb_table.offloads.fdb, spec,
 				       MLX5_FLOW_CONTEXT_ACTION_FWD_DEST,
 				       0, &dest);
 	if (IS_ERR(flow_rule)) {
@@ -165,6 +213,8 @@ out:
 }
 
 #define MAX_PF_SQ 256
+#define ESW_OFFLOADS_NUM_ENTRIES (1 << 13) /* 8K */
+#define ESW_OFFLOADS_NUM_GROUPS  4
 
 static int esw_create_offloads_fdb_table(struct mlx5_eswitch *esw, int nvports)
 {
@@ -190,15 +240,25 @@ static int esw_create_offloads_fdb_table(struct mlx5_eswitch *esw, int nvports)
 	esw_debug(dev, "Create offloads FDB table, log_max_size(%d)\n",
 		  MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size));
 
-	table_size = nvports + MAX_PF_SQ + 1;
-	fdb = mlx5_create_flow_table(root_ns, 0, table_size, 0);
+	fdb = mlx5_create_auto_grouped_flow_table(root_ns, FDB_FAST_PATH,
+						  ESW_OFFLOADS_NUM_ENTRIES,
+						  ESW_OFFLOADS_NUM_GROUPS, 0);
 	if (IS_ERR(fdb)) {
 		err = PTR_ERR(fdb);
-		esw_warn(dev, "Failed to create FDB Table err %d\n", err);
-		goto fdb_err;
+		esw_warn(dev, "Failed to create Fast path FDB Table err %d\n", err);
+		goto fast_fdb_err;
 	}
 	esw->fdb_table.fdb = fdb;
 
+	table_size = nvports + MAX_PF_SQ + 1;
+	fdb = mlx5_create_flow_table(root_ns, FDB_SLOW_PATH, table_size, 0);
+	if (IS_ERR(fdb)) {
+		err = PTR_ERR(fdb);
+		esw_warn(dev, "Failed to create slow path FDB Table err %d\n", err);
+		goto slow_fdb_err;
+	}
+	esw->fdb_table.offloads.fdb = fdb;
+
 	/* create send-to-vport group */
 	memset(flow_group_in, 0, inlen);
 	MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable,
@@ -247,8 +307,10 @@ miss_rule_err:
 miss_err:
 	mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_grp);
 send_vport_err:
-	mlx5_destroy_flow_table(fdb);
-fdb_err:
+	mlx5_destroy_flow_table(esw->fdb_table.offloads.fdb);
+slow_fdb_err:
+	mlx5_destroy_flow_table(esw->fdb_table.fdb);
+fast_fdb_err:
 ns_err:
 	kvfree(flow_group_in);
 	return err;
@@ -264,6 +326,7 @@ static void esw_destroy_offloads_fdb_table(struct mlx5_eswitch *esw)
 	mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_grp);
 	mlx5_destroy_flow_group(esw->fdb_table.offloads.miss_grp);
 
+	mlx5_destroy_flow_table(esw->fdb_table.offloads.fdb);
 	mlx5_destroy_flow_table(esw->fdb_table.fdb);
 }