Merge tag 'mlx5-updates-2021-11-16' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2021-11-16

Updates for mlx5 driver:

1) Support ethtool cq mode
2) Static allocation of mod header object for the common case
3) TC support for when local and remote VTEPs are in the same
4) Create E-Switch QoS objects on demand to save on resources
5) Minor code improvements
====================

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

23 files changed, 432 insertions, 321 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index f0ac6b0d9653..48b12ee44b8d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -1148,9 +1148,12 @@ void mlx5e_ethtool_get_channels(struct mlx5e_priv *priv,
 int mlx5e_ethtool_set_channels(struct mlx5e_priv *priv,
 			       struct ethtool_channels *ch);
 int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv,
-			       struct ethtool_coalesce *coal);
+			       struct ethtool_coalesce *coal,
+			       struct kernel_ethtool_coalesce *kernel_coal);
 int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv,
-			       struct ethtool_coalesce *coal);
+			       struct ethtool_coalesce *coal,
+			       struct kernel_ethtool_coalesce *kernel_coal,
+			       struct netlink_ext_ack *extack);
 int mlx5e_ethtool_get_link_ksettings(struct mlx5e_priv *priv,
 				     struct ethtool_link_ksettings *link_ksettings);
 int mlx5e_ethtool_set_link_ksettings(struct mlx5e_priv *priv,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c
index 7edde4d536fd..17325c5d6516 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.c
@@ -155,3 +155,61 @@ struct mlx5_modify_hdr *mlx5e_mod_hdr_get(struct mlx5e_mod_hdr_handle *mh)
 	return mh->modify_hdr;
 }
 
+char *
+mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace,
+		    struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts)
+{
+	int new_num_actions, max_hw_actions;
+	size_t new_sz, old_sz;
+	void *ret;
+
+	if (mod_hdr_acts->num_actions < mod_hdr_acts->max_actions)
+		goto out;
+
+	max_hw_actions = mlx5e_mod_hdr_max_actions(mdev, namespace);
+	new_num_actions = min(max_hw_actions,
+			      mod_hdr_acts->actions ?
+			      mod_hdr_acts->max_actions * 2 : 1);
+	if (mod_hdr_acts->max_actions == new_num_actions)
+		return ERR_PTR(-ENOSPC);
+
+	new_sz = MLX5_MH_ACT_SZ * new_num_actions;
+	old_sz = mod_hdr_acts->max_actions * MLX5_MH_ACT_SZ;
+
+	if (mod_hdr_acts->is_static) {
+		ret = kzalloc(new_sz, GFP_KERNEL);
+		if (ret) {
+			memcpy(ret, mod_hdr_acts->actions, old_sz);
+			mod_hdr_acts->is_static = false;
+		}
+	} else {
+		ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL);
+		if (ret)
+			memset(ret + old_sz, 0, new_sz - old_sz);
+	}
+	if (!ret)
+		return ERR_PTR(-ENOMEM);
+
+	mod_hdr_acts->actions = ret;
+	mod_hdr_acts->max_actions = new_num_actions;
+
+out:
+	return mod_hdr_acts->actions + (mod_hdr_acts->num_actions * MLX5_MH_ACT_SZ);
+}
+
+void
+mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts)
+{
+	if (!mod_hdr_acts->is_static)
+		kfree(mod_hdr_acts->actions);
+
+	mod_hdr_acts->actions = NULL;
+	mod_hdr_acts->num_actions = 0;
+	mod_hdr_acts->max_actions = 0;
+}
+
+char *
+mlx5e_mod_hdr_get_item(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts, int pos)
+{
+	return mod_hdr_acts->actions + (pos * MLX5_MH_ACT_SZ);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h
index 33b23d8f9182..b8dac418d0a5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/mod_hdr.h
@@ -7,14 +7,32 @@
 #include <linux/hashtable.h>
 #include <linux/mlx5/fs.h>
 
+#define MLX5_MH_ACT_SZ MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto)
+
 struct mlx5e_mod_hdr_handle;
 
 struct mlx5e_tc_mod_hdr_acts {
 	int num_actions;
 	int max_actions;
+	bool is_static;
 	void *actions;
 };
 
+#define DECLARE_MOD_HDR_ACTS_ACTIONS(name, len) \
+	u8 name[len][MLX5_MH_ACT_SZ] = {}
+
+#define DECLARE_MOD_HDR_ACTS(name, acts_arr) \
+	struct mlx5e_tc_mod_hdr_acts name = { \
+		.max_actions = ARRAY_SIZE(acts_arr), \
+		.is_static = true, \
+		.actions = acts_arr, \
+	}
+
+char *mlx5e_mod_hdr_alloc(struct mlx5_core_dev *mdev, int namespace,
+			  struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts);
+void mlx5e_mod_hdr_dealloc(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts);
+char *mlx5e_mod_hdr_get_item(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts, int pos);
+
 struct mlx5e_mod_hdr_handle *
 mlx5e_mod_hdr_attach(struct mlx5_core_dev *mdev,
 		     struct mod_hdr_tbl *tbl,
@@ -28,4 +46,12 @@ struct mlx5_modify_hdr *mlx5e_mod_hdr_get(struct mlx5e_mod_hdr_handle *mh);
 void mlx5e_mod_hdr_tbl_init(struct mod_hdr_tbl *tbl);
 void mlx5e_mod_hdr_tbl_destroy(struct mod_hdr_tbl *tbl);
 
+static inline int mlx5e_mod_hdr_max_actions(struct mlx5_core_dev *mdev, int namespace)
+{
+	if (namespace == MLX5_FLOW_NAMESPACE_FDB) /* FDB offloading */
+		return MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, max_modify_header_actions);
+	else /* namespace is MLX5_FLOW_NAMESPACE_KERNEL - NIC offloading */
+		return MLX5_CAP_FLOWTABLE_NIC_RX(mdev, max_modify_header_actions);
+}
+
 #endif /* __MLX5E_EN_MOD_HDR_H__ */
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
index df6888c4793c..ff4b4f8a5a9d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/sample.c
@@ -5,6 +5,7 @@
 #include <net/psample.h>
 #include "en/mapping.h"
 #include "en/tc/post_act.h"
+#include "en/mod_hdr.h"
 #include "sample.h"
 #include "eswitch.h"
 #include "en_tc.h"
@@ -255,12 +256,12 @@ sample_modify_hdr_get(struct mlx5_core_dev *mdev, u32 obj_id,
 		goto err_modify_hdr;
 	}
 
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 	return modify_hdr;
 
 err_modify_hdr:
 err_post_act:
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 err_set_regc0:
 	return ERR_PTR(err);
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
index c1c6e74c79c4..9a31f45a9d9b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c
@@ -36,6 +36,12 @@
 #define MLX5_CT_LABELS_BITS (mlx5e_tc_attr_to_reg_mappings[LABELS_TO_REG].mlen)
 #define MLX5_CT_LABELS_MASK GENMASK(MLX5_CT_LABELS_BITS - 1, 0)
 
+/* Statically allocate modify actions for
+ * ipv6 and port nat (5) + tuple fields (4) + nic mode zone restore (1) = 10.
+ * This will be increased dynamically if needed (for the ipv6 snat + dnat).
+ */
+#define MLX5_CT_MIN_MOD_ACTS 10
+
 #define ct_dbg(fmt, args...)\
 	netdev_dbg(ct_priv->netdev, "ct_debug: " fmt "\n", ##args)
 
@@ -609,22 +615,15 @@ mlx5_tc_ct_entry_create_nat(struct mlx5_tc_ct_priv *ct_priv,
 	struct flow_action *flow_action = &flow_rule->action;
 	struct mlx5_core_dev *mdev = ct_priv->dev;
 	struct flow_action_entry *act;
-	size_t action_size;
 	char *modact;
 	int err, i;
 
-	action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto);
-
 	flow_action_for_each(i, act, flow_action) {
 		switch (act->id) {
 		case FLOW_ACTION_MANGLE: {
-			err = alloc_mod_hdr_actions(mdev, ct_priv->ns_type,
-						    mod_acts);
-			if (err)
-				return err;
-
-			modact = mod_acts->actions +
-				 mod_acts->num_actions * action_size;
+			modact = mlx5e_mod_hdr_alloc(mdev, ct_priv->ns_type, mod_acts);
+			if (IS_ERR(modact))
+				return PTR_ERR(modact);
 
 			err = mlx5_tc_ct_parse_mangle_to_mod_act(act, modact);
 			if (err)
@@ -652,7 +651,8 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv,
 				struct mlx5e_mod_hdr_handle **mh,
 				u8 zone_restore_id, bool nat)
 {
-	struct mlx5e_tc_mod_hdr_acts mod_acts = {};
+	DECLARE_MOD_HDR_ACTS_ACTIONS(actions_arr, MLX5_CT_MIN_MOD_ACTS);
+	DECLARE_MOD_HDR_ACTS(mod_acts, actions_arr);
 	struct flow_action_entry *meta;
 	u16 ct_state = 0;
 	int err;
@@ -706,11 +706,11 @@ mlx5_tc_ct_entry_create_mod_hdr(struct mlx5_tc_ct_priv *ct_priv,
 		attr->modify_hdr = mlx5e_mod_hdr_get(*mh);
 	}
 
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 	return 0;
 
 err_mapping:
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 	mlx5_put_label_mapping(ct_priv, attr->ct_attr.ct_labels_id);
 	return err;
 }
@@ -907,12 +907,9 @@ mlx5_tc_ct_shared_counter_get(struct mlx5_tc_ct_priv *ct_priv,
 	struct mlx5_ct_tuple rev_tuple = entry->tuple;
 	struct mlx5_ct_counter *shared_counter;
 	struct mlx5_ct_entry *rev_entry;
-	__be16 tmp_port;
 
 	/* get the reversed tuple */
-	tmp_port = rev_tuple.port.src;
-	rev_tuple.port.src = rev_tuple.port.dst;
-	rev_tuple.port.dst = tmp_port;
+	swap(rev_tuple.port.src, rev_tuple.port.dst);
 
 	if (rev_tuple.addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
 		__be32 tmp_addr = rev_tuple.ip.src_v4;
@@ -1445,7 +1442,7 @@ static int tc_ct_pre_ct_add_rules(struct mlx5_ct_ft *ct_ft,
 	}
 	pre_ct->miss_rule = rule;
 
-	dealloc_mod_hdr_actions(&pre_mod_acts);
+	mlx5e_mod_hdr_dealloc(&pre_mod_acts);
 	kvfree(spec);
 	return 0;
 
@@ -1454,7 +1451,7 @@ err_miss_rule:
 err_flow_rule:
 	mlx5_modify_header_dealloc(dev, pre_ct->modify_hdr);
 err_mapping:
-	dealloc_mod_hdr_actions(&pre_mod_acts);
+	mlx5e_mod_hdr_dealloc(&pre_mod_acts);
 	kvfree(spec);
 	return err;
 }
@@ -1850,14 +1847,14 @@ __mlx5_tc_ct_flow_offload(struct mlx5_tc_ct_priv *ct_priv,
 	}
 
 	attr->ct_attr.ct_flow = ct_flow;
-	dealloc_mod_hdr_actions(&pre_mod_acts);
+	mlx5e_mod_hdr_dealloc(&pre_mod_acts);
 
 	return ct_flow->pre_ct_rule;
 
 err_insert_orig:
 	mlx5_modify_header_dealloc(priv->mdev, pre_ct_attr->modify_hdr);
 err_mapping:
-	dealloc_mod_hdr_actions(&pre_mod_acts);
+	mlx5e_mod_hdr_dealloc(&pre_mod_acts);
 	mlx5_chains_put_chain_mapping(ct_priv->chains, ct_flow->chain_mapping);
 err_get_chain:
 	kfree(ct_flow->pre_ct_attr);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
index a5e450973225..33815246fead 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c
@@ -103,7 +103,7 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv,
 }
 
 static int mlx5e_route_lookup_ipv4_get(struct mlx5e_priv *priv,
-				       struct net_device *mirred_dev,
+				       struct net_device *dev,
 				       struct mlx5e_tc_tun_route_attr *attr)
 {
 	struct net_device *route_dev;
@@ -122,13 +122,13 @@ static int mlx5e_route_lookup_ipv4_get(struct mlx5e_priv *priv,
 		uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH);
 		attr->fl.fl4.flowi4_oif = uplink_dev->ifindex;
 	} else {
-		struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(mirred_dev);
+		struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(dev);
 
 		if (tunnel && tunnel->get_remote_ifindex)
-			attr->fl.fl4.flowi4_oif = tunnel->get_remote_ifindex(mirred_dev);
+			attr->fl.fl4.flowi4_oif = tunnel->get_remote_ifindex(dev);
 	}
 
-	rt = ip_route_output_key(dev_net(mirred_dev), &attr->fl.fl4);
+	rt = ip_route_output_key(dev_net(dev), &attr->fl.fl4);
 	if (IS_ERR(rt))
 		return PTR_ERR(rt);
 
@@ -440,10 +440,10 @@ release_neigh:
 
 #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6)
 static int mlx5e_route_lookup_ipv6_get(struct mlx5e_priv *priv,
-				       struct net_device *mirred_dev,
+				       struct net_device *dev,
 				       struct mlx5e_tc_tun_route_attr *attr)
 {
-	struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(mirred_dev);
+	struct mlx5e_tc_tunnel *tunnel = mlx5e_get_tc_tun(dev);
 	struct net_device *route_dev;
 	struct net_device *out_dev;
 	struct dst_entry *dst;
@@ -451,8 +451,8 @@ static int mlx5e_route_lookup_ipv6_get(struct mlx5e_priv *priv,
 	int ret;
 
 	if (tunnel && tunnel->get_remote_ifindex)
-		attr->fl.fl6.flowi6_oif = tunnel->get_remote_ifindex(mirred_dev);
-	dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(mirred_dev), NULL, &attr->fl.fl6,
+		attr->fl.fl6.flowi6_oif = tunnel->get_remote_ifindex(dev);
+	dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(dev), NULL, &attr->fl.fl6,
 					      NULL);
 	if (IS_ERR(dst))
 		return PTR_ERR(dst);
@@ -708,7 +708,8 @@ release_neigh:
 
 int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv,
 			      struct mlx5_flow_spec *spec,
-			      struct mlx5_flow_attr *flow_attr)
+			      struct mlx5_flow_attr *flow_attr,
+			      struct net_device *filter_dev)
 {
 	struct mlx5_esw_flow_attr *esw_attr = flow_attr->esw_attr;
 	struct mlx5e_tc_int_port *int_port;
@@ -720,14 +721,14 @@ int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv,
 		/* Addresses are swapped for decap */
 		attr.fl.fl4.saddr = esw_attr->rx_tun_attr->dst_ip.v4;
 		attr.fl.fl4.daddr = esw_attr->rx_tun_attr->src_ip.v4;
-		err = mlx5e_route_lookup_ipv4_get(priv, priv->netdev, &attr);
+		err = mlx5e_route_lookup_ipv4_get(priv, filter_dev, &attr);
 	}
 #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6)
 	else if (flow_attr->tun_ip_version == 6) {
 		/* Addresses are swapped for decap */
 		attr.fl.fl6.saddr = esw_attr->rx_tun_attr->dst_ip.v6;
 		attr.fl.fl6.daddr = esw_attr->rx_tun_attr->src_ip.v6;
-		err = mlx5e_route_lookup_ipv6_get(priv, priv->netdev, &attr);
+		err = mlx5e_route_lookup_ipv6_get(priv, filter_dev, &attr);
 	}
 #endif
 	else
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h
index aa092eaeaec3..b38f693bbb52 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.h
@@ -94,7 +94,8 @@ mlx5e_tc_tun_update_header_ipv6(struct mlx5e_priv *priv,
 #endif
 int mlx5e_tc_tun_route_lookup(struct mlx5e_priv *priv,
 			      struct mlx5_flow_spec *spec,
-			      struct mlx5_flow_attr *attr);
+			      struct mlx5_flow_attr *attr,
+			      struct net_device *filter_dev);
 
 bool mlx5e_tc_tun_device_to_offload(struct mlx5e_priv *priv,
 				    struct net_device *netdev);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
index 660cca73c36c..de16bbc08679 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun_encap.c
@@ -1153,7 +1153,7 @@ int mlx5e_attach_decap_route(struct mlx5e_priv *priv,
 
 	tbl_time_before = mlx5e_route_tbl_get_last_update(priv);
 	tbl_time_after = tbl_time_before;
-	err = mlx5e_tc_tun_route_lookup(priv, &parse_attr->spec, attr);
+	err = mlx5e_tc_tun_route_lookup(priv, &parse_attr->spec, attr, parse_attr->filter_dev);
 	if (err || !esw_attr->rx_tun_attr->decap_vport)
 		goto out;
 
@@ -1474,7 +1474,7 @@ static void mlx5e_reoffload_decap(struct mlx5e_priv *priv,
 
 		parse_attr = attr->parse_attr;
 		spec = &parse_attr->spec;
-		err = mlx5e_tc_tun_route_lookup(priv, spec, attr);
+		err = mlx5e_tc_tun_route_lookup(priv, spec, attr, parse_attr->filter_dev);
 		if (err) {
 			mlx5_core_warn(priv->mdev, "Failed to lookup route for flow, %d\n",
 				       err);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index c2ea5fad48dd..45bdfcb3dcc7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -511,7 +511,8 @@ static int mlx5e_set_channels(struct net_device *dev,
 }
 
 int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv,
-			       struct ethtool_coalesce *coal)
+			       struct ethtool_coalesce *coal,
+			       struct kernel_ethtool_coalesce *kernel_coal)
 {
 	struct dim_cq_moder *rx_moder, *tx_moder;
 
@@ -528,6 +529,11 @@ int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv,
 	coal->tx_max_coalesced_frames	= tx_moder->pkts;
 	coal->use_adaptive_tx_coalesce	= priv->channels.params.tx_dim_enabled;
 
+	kernel_coal->use_cqe_mode_rx =
+		MLX5E_GET_PFLAG(&priv->channels.params, MLX5E_PFLAG_RX_CQE_BASED_MODER);
+	kernel_coal->use_cqe_mode_tx =
+		MLX5E_GET_PFLAG(&priv->channels.params, MLX5E_PFLAG_TX_CQE_BASED_MODER);
+
 	return 0;
 }
 
@@ -538,7 +544,7 @@ static int mlx5e_get_coalesce(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 
-	return mlx5e_ethtool_get_coalesce(priv, coal);
+	return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal);
 }
 
 #define MLX5E_MAX_COAL_TIME		MLX5_MAX_CQ_PERIOD
@@ -578,14 +584,26 @@ mlx5e_set_priv_channels_rx_coalesce(struct mlx5e_priv *priv, struct ethtool_coal
 	}
 }
 
+/* convert a boolean value of cq_mode to mlx5 period mode
+ * true  : MLX5_CQ_PERIOD_MODE_START_FROM_CQE
+ * false : MLX5_CQ_PERIOD_MODE_START_FROM_EQE
+ */
+static int cqe_mode_to_period_mode(bool val)
+{
+	return val ? MLX5_CQ_PERIOD_MODE_START_FROM_CQE : MLX5_CQ_PERIOD_MODE_START_FROM_EQE;
+}
+
 int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv,
-			       struct ethtool_coalesce *coal)
+			       struct ethtool_coalesce *coal,
+			       struct kernel_ethtool_coalesce *kernel_coal,
+			       struct netlink_ext_ack *extack)
 {
 	struct dim_cq_moder *rx_moder, *tx_moder;
 	struct mlx5_core_dev *mdev = priv->mdev;
 	struct mlx5e_params new_params;
 	bool reset_rx, reset_tx;
 	bool reset = true;
+	u8 cq_period_mode;
 	int err = 0;
 
 	if (!MLX5_CAP_GEN(mdev, cq_moderation))
@@ -605,6 +623,12 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv,
 		return -ERANGE;
 	}
 
+	if ((kernel_coal->use_cqe_mode_rx || kernel_coal->use_cqe_mode_tx) &&
+	    !MLX5_CAP_GEN(priv->mdev, cq_period_start_from_cqe)) {
+		NL_SET_ERR_MSG_MOD(extack, "cqe_mode_rx/tx is not supported on this device");
+		return -EOPNOTSUPP;
+	}
+
 	mutex_lock(&priv->state_lock);
 	new_params = priv->channels.params;
 
@@ -621,6 +645,18 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv,
 	reset_rx = !!coal->use_adaptive_rx_coalesce != priv->channels.params.rx_dim_enabled;
 	reset_tx = !!coal->use_adaptive_tx_coalesce != priv->channels.params.tx_dim_enabled;
 
+	cq_period_mode = cqe_mode_to_period_mode(kernel_coal->use_cqe_mode_rx);
+	if (cq_period_mode != rx_moder->cq_period_mode) {
+		mlx5e_set_rx_cq_mode_params(&new_params, cq_period_mode);
+		reset_rx = true;
+	}
+
+	cq_period_mode = cqe_mode_to_period_mode(kernel_coal->use_cqe_mode_tx);
+	if (cq_period_mode != tx_moder->cq_period_mode) {
+		mlx5e_set_tx_cq_mode_params(&new_params, cq_period_mode);
+		reset_tx = true;
+	}
+
 	if (reset_rx) {
 		u8 mode = MLX5E_GET_PFLAG(&new_params,
 					  MLX5E_PFLAG_RX_CQE_BASED_MODER);
@@ -656,9 +692,9 @@ static int mlx5e_set_coalesce(struct net_device *netdev,
 			      struct kernel_ethtool_coalesce *kernel_coal,
 			      struct netlink_ext_ack *extack)
 {
-	struct mlx5e_priv *priv    = netdev_priv(netdev);
+	struct mlx5e_priv *priv = netdev_priv(netdev);
 
-	return mlx5e_ethtool_set_coalesce(priv, coal);
+	return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack);
 }
 
 static void ptys2ethtool_supported_link(struct mlx5_core_dev *mdev,
@@ -2358,7 +2394,8 @@ static void mlx5e_get_rmon_stats(struct net_device *netdev,
 const struct ethtool_ops mlx5e_ethtool_ops = {
 	.supported_coalesce_params = ETHTOOL_COALESCE_USECS |
 				     ETHTOOL_COALESCE_MAX_FRAMES |
-				     ETHTOOL_COALESCE_USE_ADAPTIVE,
+				     ETHTOOL_COALESCE_USE_ADAPTIVE |
+				     ETHTOOL_COALESCE_USE_CQE,
 	.get_drvinfo       = mlx5e_get_drvinfo,
 	.get_link          = ethtool_op_get_link,
 	.get_link_ext_state  = mlx5e_get_link_ext_state,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index e58a9ec42553..8c81aeba07db 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -258,7 +258,7 @@ static int mlx5e_rep_get_coalesce(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 
-	return mlx5e_ethtool_get_coalesce(priv, coal);
+	return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal);
 }
 
 static int mlx5e_rep_set_coalesce(struct net_device *netdev,
@@ -268,7 +268,7 @@ static int mlx5e_rep_set_coalesce(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 
-	return mlx5e_ethtool_set_coalesce(priv, coal);
+	return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack);
 }
 
 static u32 mlx5e_rep_get_rxfh_key_size(struct net_device *netdev)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
index 2a9bfc3ffa2e..3c91a11e27ad 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c
@@ -2076,7 +2076,7 @@ static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(ptp)
 
 	for (i = 0; i < NUM_PTP_CH_STATS; i++)
 		sprintf(data + (idx++) * ETH_GSTRING_LEN,
-			ptp_ch_stats_desc[i].format);
+			"%s", ptp_ch_stats_desc[i].format);
 
 	if (priv->tx_ptp_opened) {
 		for (tc = 0; tc < priv->max_opened_tc; tc++)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 835caa1c7b74..55e384abd364 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -71,7 +71,6 @@
 #include "lag/mp.h"
 
 #define nic_chains(priv) ((priv)->fs.tc.chains)
-#define MLX5_MH_ACT_SZ MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto)
 
 #define MLX5E_TC_TABLE_NUM_GROUPS 4
 #define MLX5E_TC_TABLE_MAX_GROUP_SIZE BIT(18)
@@ -209,12 +208,9 @@ mlx5e_tc_match_to_reg_set_and_get_id(struct mlx5_core_dev *mdev,
 	char *modact;
 	int err;
 
-	err = alloc_mod_hdr_actions(mdev, ns, mod_hdr_acts);
-	if (err)
-		return err;
-
-	modact = mod_hdr_acts->actions +
-		 (mod_hdr_acts->num_actions * MLX5_MH_ACT_SZ);
+	modact = mlx5e_mod_hdr_alloc(mdev, ns, mod_hdr_acts);
+	if (IS_ERR(modact))
+		return PTR_ERR(modact);
 
 	/* Firmware has 5bit length field and 0 means 32bits */
 	if (mlen == 32)
@@ -333,7 +329,7 @@ void mlx5e_tc_match_to_reg_mod_hdr_change(struct mlx5_core_dev *mdev,
 	int mlen = mlx5e_tc_attr_to_reg_mappings[type].mlen;
 	char *modact;
 
-	modact = mod_hdr_acts->actions + (act_id * MLX5_MH_ACT_SZ);
+	modact = mlx5e_mod_hdr_get_item(mod_hdr_acts, act_id);
 
 	/* Firmware has 5bit length field and 0 means 32bits */
 	if (mlen == 32)
@@ -1076,7 +1072,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv,
 
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) {
 		err = mlx5e_attach_mod_hdr(priv, flow, parse_attr);
-		dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts);
+		mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts);
 		if (err)
 			return err;
 	}
@@ -1132,16 +1128,16 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv,
 	}
 	mutex_unlock(&priv->fs.tc.t_lock);
 
-	kvfree(attr->parse_attr);
-
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR)
 		mlx5e_detach_mod_hdr(priv, flow);
 
-	mlx5_fc_destroy(priv->mdev, attr->counter);
+	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT)
+		mlx5_fc_destroy(priv->mdev, attr->counter);
 
 	if (flow_flag_test(flow, HAIRPIN))
 		mlx5e_hairpin_flow_del(priv, flow);
 
+	kvfree(attr->parse_attr);
 	kfree(flow->attr);
 }
 
@@ -1623,15 +1619,12 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	mlx5_tc_ct_match_del(get_ct_priv(priv), &flow->attr->ct_attr);
 
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) {
-		dealloc_mod_hdr_actions(&attr->parse_attr->mod_hdr_acts);
+		mlx5e_mod_hdr_dealloc(&attr->parse_attr->mod_hdr_acts);
 		if (vf_tun && attr->modify_hdr)
 			mlx5_modify_header_dealloc(priv->mdev, attr->modify_hdr);
 		else
 			mlx5e_detach_mod_hdr(priv, flow);
 	}
-	kfree(attr->sample_attr);
-	kvfree(attr->parse_attr);
-	kvfree(attr->esw_attr->rx_tun_attr);
 
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT)
 		mlx5_fc_destroy(esw_attr->counter_dev, attr->counter);
@@ -1645,6 +1638,9 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 	if (flow_flag_test(flow, L3_TO_L2_DECAP))
 		mlx5e_detach_decap(priv, flow);
 
+	kfree(attr->sample_attr);
+	kvfree(attr->esw_attr->rx_tun_attr);
+	kvfree(attr->parse_attr);
 	kfree(flow->attr);
 }
 
@@ -2766,13 +2762,12 @@ static int offload_pedit_fields(struct mlx5e_priv *priv,
 				struct netlink_ext_ack *extack)
 {
 	struct pedit_headers *set_masks, *add_masks, *set_vals, *add_vals;
-	int i, action_size, first, last, next_z;
 	void *headers_c, *headers_v, *action, *vals_p;
 	u32 *s_masks_p, *a_masks_p, s_mask, a_mask;
 	struct mlx5e_tc_mod_hdr_acts *mod_acts;
-	struct mlx5_fields *f;
 	unsigned long mask, field_mask;
-	int err;
+	int i, first, last, next_z;
+	struct mlx5_fields *f;
 	u8 cmd;
 
 	mod_acts = &parse_attr->mod_hdr_acts;
@@ -2784,8 +2779,6 @@ static int offload_pedit_fields(struct mlx5e_priv *priv,
 	set_vals = &hdrs[0].vals;
 	add_vals = &hdrs[1].vals;
 
-	action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto);
-
 	for (i = 0; i < ARRAY_SIZE(fields); i++) {
 		bool skip;
 
@@ -2853,18 +2846,16 @@ static int offload_pedit_fields(struct mlx5e_priv *priv,
 			return -EOPNOTSUPP;
 		}
 
-		err = alloc_mod_hdr_actions(priv->mdev, namespace, mod_acts);
-		if (err) {
+		action = mlx5e_mod_hdr_alloc(priv->mdev, namespace, mod_acts);
+		if (IS_ERR(action)) {
 			NL_SET_ERR_MSG_MOD(extack,
 					   "too many pedit actions, can't offload");
 			mlx5_core_warn(priv->mdev,
 				       "mlx5: parsed %d pedit actions, can't do more\n",
 				       mod_acts->num_actions);
-			return err;
+			return PTR_ERR(action);
 		}
 
-		action = mod_acts->actions +
-			 (mod_acts->num_actions * action_size);
 		MLX5_SET(set_action_in, action, action_type, cmd);
 		MLX5_SET(set_action_in, action, field, f->field);
 
@@ -2894,57 +2885,6 @@ static int offload_pedit_fields(struct mlx5e_priv *priv,
 	return 0;
 }
 
-static int mlx5e_flow_namespace_max_modify_action(struct mlx5_core_dev *mdev,
-						  int namespace)
-{
-	if (namespace == MLX5_FLOW_NAMESPACE_FDB) /* FDB offloading */
-		return MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, max_modify_header_actions);
-	else /* namespace is MLX5_FLOW_NAMESPACE_KERNEL - NIC offloading */
-		return MLX5_CAP_FLOWTABLE_NIC_RX(mdev, max_modify_header_actions);
-}
-
-int alloc_mod_hdr_actions(struct mlx5_core_dev *mdev,
-			  int namespace,
-			  struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts)
-{
-	int action_size, new_num_actions, max_hw_actions;
-	size_t new_sz, old_sz;
-	void *ret;
-
-	if (mod_hdr_acts->num_actions < mod_hdr_acts->max_actions)
-		return 0;
-
-	action_size = MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto);
-
-	max_hw_actions = mlx5e_flow_namespace_max_modify_action(mdev,
-								namespace);
-	new_num_actions = min(max_hw_actions,
-			      mod_hdr_acts->actions ?
-			      mod_hdr_acts->max_actions * 2 : 1);
-	if (mod_hdr_acts->max_actions == new_num_actions)
-		return -ENOSPC;
-
-	new_sz = action_size * new_num_actions;
-	old_sz = mod_hdr_acts->max_actions * action_size;
-	ret = krealloc(mod_hdr_acts->actions, new_sz, GFP_KERNEL);
-	if (!ret)
-		return -ENOMEM;
-
-	memset(ret + old_sz, 0, new_sz - old_sz);
-	mod_hdr_acts->actions = ret;
-	mod_hdr_acts->max_actions = new_num_actions;
-
-	return 0;
-}
-
-void dealloc_mod_hdr_actions(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts)
-{
-	kfree(mod_hdr_acts->actions);
-	mod_hdr_acts->actions = NULL;
-	mod_hdr_acts->num_actions = 0;
-	mod_hdr_acts->max_actions = 0;
-}
-
 static const struct pedit_headers zero_masks = {};
 
 static int
@@ -2967,7 +2907,7 @@ parse_pedit_to_modify_hdr(struct mlx5e_priv *priv,
 		goto out_err;
 	}
 
-	if (!mlx5e_flow_namespace_max_modify_action(priv->mdev, namespace)) {
+	if (!mlx5e_mod_hdr_max_actions(priv->mdev, namespace)) {
 		NL_SET_ERR_MSG_MOD(extack,
 				   "The pedit offload action is not supported");
 		goto out_err;
@@ -3060,7 +3000,7 @@ static int alloc_tc_pedit_action(struct mlx5e_priv *priv, int namespace,
 	return 0;
 
 out_dealloc_parsed_actions:
-	dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts);
+	mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts);
 	return err;
 }
 
@@ -3484,12 +3424,12 @@ actions_prepare_mod_hdr_actions(struct mlx5e_priv *priv,
 	if (err)
 		return err;
 
-	/* In case all pedit actions are skipped, remove the MOD_HDR flag. */
 	if (parse_attr->mod_hdr_acts.num_actions > 0)
 		return 0;
 
+	/* In case all pedit actions are skipped, remove the MOD_HDR flag. */
 	attr->action &= ~MLX5_FLOW_CONTEXT_ACTION_MOD_HDR;
-	dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts);
+	mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts);
 
 	if (ns_type != MLX5_FLOW_NAMESPACE_FDB)
 		return 0;
@@ -4708,7 +4648,7 @@ mlx5e_add_nic_flow(struct mlx5e_priv *priv,
 
 err_free:
 	flow_flag_set(flow, FAILED);
-	dealloc_mod_hdr_actions(&parse_attr->mod_hdr_acts);
+	mlx5e_mod_hdr_dealloc(&parse_attr->mod_hdr_acts);
 	mlx5e_flow_put(priv, flow);
 out:
 	return err;
@@ -5008,14 +4948,8 @@ static int scan_tc_matchall_fdb_actions(struct mlx5e_priv *priv,
 int mlx5e_tc_configure_matchall(struct mlx5e_priv *priv,
 				struct tc_cls_matchall_offload *ma)
 {
-	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
 	struct netlink_ext_ack *extack = ma->common.extack;
 
-	if (!mlx5_esw_qos_enabled(esw)) {
-		NL_SET_ERR_MSG_MOD(extack, "QoS is not supported on this device");
-		return -EOPNOTSUPP;
-	}
-
 	if (ma->common.prio != 1) {
 		NL_SET_ERR_MSG_MOD(extack, "only priority 1 is supported");
 		return -EINVAL;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
index fdb222793027..eb042f0f5a41 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h
@@ -247,11 +247,6 @@ int mlx5e_tc_add_flow_mod_hdr(struct mlx5e_priv *priv,
 			      struct mlx5e_tc_flow_parse_attr *parse_attr,
 			      struct mlx5e_tc_flow *flow);
 
-int alloc_mod_hdr_actions(struct mlx5_core_dev *mdev,
-			  int namespace,
-			  struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts);
-void dealloc_mod_hdr_actions(struct mlx5e_tc_mod_hdr_acts *mod_hdr_acts);
-
 struct mlx5e_tc_flow;
 u32 mlx5e_tc_get_flow_tun_id(struct mlx5e_tc_flow *flow);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c
index 425c91814b34..c275fe028b6d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c
@@ -14,6 +14,7 @@
 #include "fs_core.h"
 #include "esw/indir_table.h"
 #include "lib/fs_chains.h"
+#include "en/mod_hdr.h"
 
 #define MLX5_ESW_INDIR_TABLE_SIZE 128
 #define MLX5_ESW_INDIR_TABLE_RECIRC_IDX_MAX (MLX5_ESW_INDIR_TABLE_SIZE - 2)
@@ -226,7 +227,7 @@ static int mlx5_esw_indir_table_rule_get(struct mlx5_eswitch *esw,
 		goto err_handle;
 	}
 
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 	rule->handle = handle;
 	rule->vni = esw_attr->rx_tun_attr->vni;
 	rule->mh = flow_act.modify_hdr;
@@ -243,7 +244,7 @@ err_table:
 	mlx5_modify_header_dealloc(esw->dev, flow_act.modify_hdr);
 err_mod_hdr_alloc:
 err_mod_hdr_regc1:
-	dealloc_mod_hdr_actions(&mod_acts);
+	mlx5e_mod_hdr_dealloc(&mod_acts);
 err_mod_hdr_regc0:
 err_ethertype:
 	kfree(rule);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c
index df277a6cddc0..2b52f7c09152 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c
@@ -522,9 +522,7 @@ int mlx5_eswitch_set_vport_rate(struct mlx5_eswitch *esw, u16 vport,
 		return PTR_ERR(evport);
 
 	mutex_lock(&esw->state_lock);
-	err = mlx5_esw_qos_set_vport_min_rate(esw, evport, min_rate, NULL);
-	if (!err)
-		err = mlx5_esw_qos_set_vport_max_rate(esw, evport, max_rate, NULL);
+	err = mlx5_esw_qos_set_vport_rate(esw, evport, max_rate, min_rate);
 	mutex_unlock(&esw->state_lock);
 	return err;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c
index c6cc67cb4f6a..ff0a07a91992 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c
@@ -204,10 +204,8 @@ static int esw_qos_normalize_groups_min_rate(struct mlx5_eswitch *esw, u32 divid
 	return 0;
 }
 
-int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw,
-				    struct mlx5_vport *evport,
-				    u32 min_rate,
-				    struct netlink_ext_ack *extack)
+static int esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport,
+				      u32 min_rate, struct netlink_ext_ack *extack)
 {
 	u32 fw_max_bw_share, previous_min_rate;
 	bool min_rate_supported;
@@ -231,10 +229,8 @@ int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw,
 	return err;
 }
 
-int mlx5_esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw,
-				    struct mlx5_vport *evport,
-				    u32 max_rate,
-				    struct netlink_ext_ack *extack)
+static int esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport,
+				      u32 max_rate, struct netlink_ext_ack *extack)
 {
 	u32 act_max_rate = max_rate;
 	bool max_rate_supported;
@@ -432,16 +428,13 @@ static int esw_qos_vport_update_group(struct mlx5_eswitch *esw,
 }
 
 static struct mlx5_esw_rate_group *
-esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack)
+__esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack)
 {
 	u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {};
 	struct mlx5_esw_rate_group *group;
 	u32 divider;
 	int err;
 
-	if (!MLX5_CAP_QOS(esw->dev, log_esw_max_sched_depth))
-		return ERR_PTR(-EOPNOTSUPP);
-
 	group = kzalloc(sizeof(*group), GFP_KERNEL);
 	if (!group)
 		return ERR_PTR(-ENOMEM);
@@ -482,9 +475,32 @@ err_sched_elem:
 	return ERR_PTR(err);
 }
 
-static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw,
-				      struct mlx5_esw_rate_group *group,
-				      struct netlink_ext_ack *extack)
+static int esw_qos_get(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack);
+static void esw_qos_put(struct mlx5_eswitch *esw);
+
+static struct mlx5_esw_rate_group *
+esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack)
+{
+	struct mlx5_esw_rate_group *group;
+	int err;
+
+	if (!MLX5_CAP_QOS(esw->dev, log_esw_max_sched_depth))
+		return ERR_PTR(-EOPNOTSUPP);
+
+	err = esw_qos_get(esw, extack);
+	if (err)
+		return ERR_PTR(err);
+
+	group = __esw_qos_create_rate_group(esw, extack);
+	if (IS_ERR(group))
+		esw_qos_put(esw);
+
+	return group;
+}
+
+static int __esw_qos_destroy_rate_group(struct mlx5_eswitch *esw,
+					struct mlx5_esw_rate_group *group,
+					struct netlink_ext_ack *extack)
 {
 	u32 divider;
 	int err;
@@ -503,7 +519,21 @@ static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw,
 		NL_SET_ERR_MSG_MOD(extack, "E-Switch destroy TSAR_ID failed");
 
 	trace_mlx5_esw_group_qos_destroy(esw->dev, group, group->tsar_ix);
+
 	kfree(group);
+
+	return err;
+}
+
+static int esw_qos_destroy_rate_group(struct mlx5_eswitch *esw,
+				      struct mlx5_esw_rate_group *group,
+				      struct netlink_ext_ack *extack)
+{
+	int err;
+
+	err = __esw_qos_destroy_rate_group(esw, group, extack);
+	esw_qos_put(esw);
+
 	return err;
 }
 
@@ -526,7 +556,7 @@ static bool esw_qos_element_type_supported(struct mlx5_core_dev *dev, int type)
 	return false;
 }
 
-void mlx5_esw_qos_create(struct mlx5_eswitch *esw)
+static int esw_qos_create(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack)
 {
 	u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {};
 	struct mlx5_core_dev *dev = esw->dev;
@@ -534,14 +564,10 @@ void mlx5_esw_qos_create(struct mlx5_eswitch *esw)
 	int err;
 
 	if (!MLX5_CAP_GEN(dev, qos) || !MLX5_CAP_QOS(dev, esw_scheduling))
-		return;
+		return -EOPNOTSUPP;
 
 	if (!esw_qos_element_type_supported(dev, SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR))
-		return;
-
-	mutex_lock(&esw->state_lock);
-	if (esw->qos.enabled)
-		goto unlock;
+		return -EOPNOTSUPP;
 
 	MLX5_SET(scheduling_context, tsar_ctx, element_type,
 		 SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR);
@@ -555,75 +581,93 @@ void mlx5_esw_qos_create(struct mlx5_eswitch *esw)
 						 &esw->qos.root_tsar_ix);
 	if (err) {
 		esw_warn(dev, "E-Switch create root TSAR failed (%d)\n", err);
-		goto unlock;
+		return err;
 	}
 
 	INIT_LIST_HEAD(&esw->qos.groups);
 	if (MLX5_CAP_QOS(dev, log_esw_max_sched_depth)) {
-		esw->qos.group0 = esw_qos_create_rate_group(esw, NULL);
+		esw->qos.group0 = __esw_qos_create_rate_group(esw, extack);
 		if (IS_ERR(esw->qos.group0)) {
 			esw_warn(dev, "E-Switch create rate group 0 failed (%ld)\n",
 				 PTR_ERR(esw->qos.group0));
 			goto err_group0;
 		}
 	}
-	esw->qos.enabled = true;
-unlock:
-	mutex_unlock(&esw->state_lock);
-	return;
+	refcount_set(&esw->qos.refcnt, 1);
+
+	return 0;
 
 err_group0:
-	err = mlx5_destroy_scheduling_element_cmd(esw->dev,
-						  SCHEDULING_HIERARCHY_E_SWITCH,
-						  esw->qos.root_tsar_ix);
-	if (err)
-		esw_warn(esw->dev, "E-Switch destroy root TSAR failed (%d)\n", err);
-	mutex_unlock(&esw->state_lock);
+	if (mlx5_destroy_scheduling_element_cmd(esw->dev, SCHEDULING_HIERARCHY_E_SWITCH,
+						esw->qos.root_tsar_ix))
+		esw_warn(esw->dev, "E-Switch destroy root TSAR failed.\n");
+
+	return err;
 }
 
-void mlx5_esw_qos_destroy(struct mlx5_eswitch *esw)
+static void esw_qos_destroy(struct mlx5_eswitch *esw)
 {
-	struct devlink *devlink = priv_to_devlink(esw->dev);
 	int err;
 
-	devlink_rate_nodes_destroy(devlink);
-	mutex_lock(&esw->state_lock);
-	if (!esw->qos.enabled)
-		goto unlock;
-
 	if (esw->qos.group0)
-		esw_qos_destroy_rate_group(esw, esw->qos.group0, NULL);
+		__esw_qos_destroy_rate_group(esw, esw->qos.group0, NULL);
 
 	err = mlx5_destroy_scheduling_element_cmd(esw->dev,
 						  SCHEDULING_HIERARCHY_E_SWITCH,
 						  esw->qos.root_tsar_ix);
 	if (err)
 		esw_warn(esw->dev, "E-Switch destroy root TSAR failed (%d)\n", err);
+}
 
-	esw->qos.enabled = false;
-unlock:
-	mutex_unlock(&esw->state_lock);
+static int esw_qos_get(struct mlx5_eswitch *esw, struct netlink_ext_ack *extack)
+{
+	int err = 0;
+
+	lockdep_assert_held(&esw->state_lock);
+
+	if (!refcount_inc_not_zero(&esw->qos.refcnt)) {
+		/* esw_qos_create() set refcount to 1 only on success.
+		 * No need to decrement on failure.
+		 */
+		err = esw_qos_create(esw, extack);
+	}
+
+	return err;
 }
 
-int mlx5_esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport,
-			      u32 max_rate, u32 bw_share)
+static void esw_qos_put(struct mlx5_eswitch *esw)
+{
+	lockdep_assert_held(&esw->state_lock);
+	if (refcount_dec_and_test(&esw->qos.refcnt))
+		esw_qos_destroy(esw);
+}
+
+static int esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport,
+				u32 max_rate, u32 bw_share, struct netlink_ext_ack *extack)
 {
 	int err;
 
 	lockdep_assert_held(&esw->state_lock);
-	if (!esw->qos.enabled)
+	if (vport->qos.enabled)
 		return 0;
 
-	if (vport->qos.enabled)
-		return -EEXIST;
+	err = esw_qos_get(esw, extack);
+	if (err)
+		return err;
 
 	vport->qos.group = esw->qos.group0;
 
 	err = esw_qos_vport_create_sched_element(esw, vport, max_rate, bw_share);
-	if (!err) {
-		vport->qos.enabled = true;
-		trace_mlx5_esw_vport_qos_create(vport, bw_share, max_rate);
-	}
+	if (err)
+		goto err_out;
+
+	vport->qos.enabled = true;
+	trace_mlx5_esw_vport_qos_create(vport, bw_share, max_rate);
+
+	return 0;
+
+err_out:
+	esw_qos_put(esw);
 
 	return err;
 }
@@ -633,7 +677,7 @@ void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vpo
 	int err;
 
 	lockdep_assert_held(&esw->state_lock);
-	if (!esw->qos.enabled || !vport->qos.enabled)
+	if (!vport->qos.enabled)
 		return;
 	WARN(vport->qos.group && vport->qos.group != esw->qos.group0,
 	     "Disabling QoS on port before detaching it from group");
@@ -645,8 +689,27 @@ void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vpo
 		esw_warn(esw->dev, "E-Switch destroy TSAR vport element failed (vport=%d,err=%d)\n",
 			 vport->vport, err);
 
-	vport->qos.enabled = false;
+	memset(&vport->qos, 0, sizeof(vport->qos));
 	trace_mlx5_esw_vport_qos_destroy(vport);
+
+	esw_qos_put(esw);
+}
+
+int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *vport,
+				u32 min_rate, u32 max_rate)
+{
+	int err;
+
+	lockdep_assert_held(&esw->state_lock);
+	err = esw_qos_vport_enable(esw, vport, 0, 0, NULL);
+	if (err)
+		return err;
+
+	err = esw_qos_set_vport_min_rate(esw, vport, min_rate, NULL);
+	if (!err)
+		err = esw_qos_set_vport_max_rate(esw, vport, max_rate, NULL);
+
+	return err;
 }
 
 int mlx5_esw_qos_modify_vport_rate(struct mlx5_eswitch *esw, u16 vport_num, u32 rate_mbps)
@@ -654,22 +717,29 @@ int mlx5_esw_qos_modify_vport_rate(struct mlx5_eswitch *esw, u16 vport_num, u32
 	u32 ctx[MLX5_ST_SZ_DW(scheduling_context)] = {};
 	struct mlx5_vport *vport;
 	u32 bitmask;
+	int err;
 
 	vport = mlx5_eswitch_get_vport(esw, vport_num);
 	if (IS_ERR(vport))
 		return PTR_ERR(vport);
 
-	if (!vport->qos.enabled)
-		return -EOPNOTSUPP;
-
-	MLX5_SET(scheduling_context, ctx, max_average_bw, rate_mbps);
-	bitmask = MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_MAX_AVERAGE_BW;
+	mutex_lock(&esw->state_lock);
+	if (!vport->qos.enabled) {
+		/* Eswitch QoS wasn't enabled yet. Enable it and vport QoS. */
+		err = esw_qos_vport_enable(esw, vport, rate_mbps, vport->qos.bw_share, NULL);
+	} else {
+		MLX5_SET(scheduling_context, ctx, max_average_bw, rate_mbps);
+
+		bitmask = MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_MAX_AVERAGE_BW;
+		err = mlx5_modify_scheduling_element_cmd(esw->dev,
+							 SCHEDULING_HIERARCHY_E_SWITCH,
+							 ctx,
+							 vport->qos.esw_tsar_ix,
+							 bitmask);
+	}
+	mutex_unlock(&esw->state_lock);
 
-	return mlx5_modify_scheduling_element_cmd(esw->dev,
-						  SCHEDULING_HIERARCHY_E_SWITCH,
-						  ctx,
-						  vport->qos.esw_tsar_ix,
-						  bitmask);
+	return err;
 }
 
 #define MLX5_LINKSPEED_UNIT 125000 /* 1Mbps in Bps */
@@ -728,7 +798,12 @@ int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void
 		return err;
 
 	mutex_lock(&esw->state_lock);
-	err = mlx5_esw_qos_set_vport_min_rate(esw, vport, tx_share, extack);
+	err = esw_qos_vport_enable(esw, vport, 0, 0, extack);
+	if (err)
+		goto unlock;
+
+	err = esw_qos_set_vport_min_rate(esw, vport, tx_share, extack);
+unlock:
 	mutex_unlock(&esw->state_lock);
 	return err;
 }
@@ -749,7 +824,12 @@ int mlx5_esw_devlink_rate_leaf_tx_max_set(struct devlink_rate *rate_leaf, void *
 		return err;
 
 	mutex_lock(&esw->state_lock);
-	err = mlx5_esw_qos_set_vport_max_rate(esw, vport, tx_max, extack);
+	err = esw_qos_vport_enable(esw, vport, 0, 0, extack);
+	if (err)
+		goto unlock;
+
+	err = esw_qos_set_vport_max_rate(esw, vport, tx_max, extack);
+unlock:
 	mutex_unlock(&esw->state_lock);
 	return err;
 }
@@ -846,7 +926,9 @@ int mlx5_esw_qos_vport_update_group(struct mlx5_eswitch *esw,
 	int err;
 
 	mutex_lock(&esw->state_lock);
-	err = esw_qos_vport_update_group(esw, vport, group, extack);
+	err = esw_qos_vport_enable(esw, vport, 0, 0, extack);
+	if (!err)
+		err = esw_qos_vport_update_group(esw, vport, group, extack);
 	mutex_unlock(&esw->state_lock);
 	return err;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h
index 28451abe2d2f..0141e9d52037 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.h
@@ -6,18 +6,8 @@
 
 #ifdef CONFIG_MLX5_ESWITCH
 
-int mlx5_esw_qos_set_vport_min_rate(struct mlx5_eswitch *esw,
-				    struct mlx5_vport *evport,
-				    u32 min_rate,
-				    struct netlink_ext_ack *extack);
-int mlx5_esw_qos_set_vport_max_rate(struct mlx5_eswitch *esw,
-				    struct mlx5_vport *evport,
-				    u32 max_rate,
-				    struct netlink_ext_ack *extack);
-void mlx5_esw_qos_create(struct mlx5_eswitch *esw);
-void mlx5_esw_qos_destroy(struct mlx5_eswitch *esw);
-int mlx5_esw_qos_vport_enable(struct mlx5_eswitch *esw, struct mlx5_vport *vport,
-			      u32 max_rate, u32 bw_share);
+int mlx5_esw_qos_set_vport_rate(struct mlx5_eswitch *esw, struct mlx5_vport *evport,
+				u32 max_rate, u32 min_rate);
 void mlx5_esw_qos_vport_disable(struct mlx5_eswitch *esw, struct mlx5_vport *vport);
 
 int mlx5_esw_devlink_rate_leaf_tx_share_set(struct devlink_rate *rate_leaf, void *priv,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index ec136b499204..46532dd42b43 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -781,9 +781,6 @@ static int esw_vport_setup(struct mlx5_eswitch *esw, struct mlx5_vport *vport)
 	if (err)
 		return err;
 
-	/* Attach vport to the eswitch rate limiter */
-	mlx5_esw_qos_vport_enable(esw, vport, vport->qos.max_rate, vport->qos.bw_share);
-
 	if (mlx5_esw_is_manager_vport(esw, vport_num))
 		return 0;
 
@@ -1260,8 +1257,6 @@ int mlx5_eswitch_enable_locked(struct mlx5_eswitch *esw, int mode, int num_vfs)
 
 	mlx5_eswitch_update_num_of_vfs(esw, num_vfs);
 
-	mlx5_esw_qos_create(esw);
-
 	esw->mode = mode;
 
 	if (mode == MLX5_ESWITCH_LEGACY) {
@@ -1290,7 +1285,6 @@ abort:
 	if (mode == MLX5_ESWITCH_OFFLOADS)
 		mlx5_rescan_drivers(esw->dev);
 
-	mlx5_esw_qos_destroy(esw);
 	mlx5_esw_acls_ns_cleanup(esw);
 	return err;
 }
@@ -1330,6 +1324,7 @@ int mlx5_eswitch_enable(struct mlx5_eswitch *esw, int num_vfs)
 
 void mlx5_eswitch_disable_locked(struct mlx5_eswitch *esw, bool clear_vf)
 {
+	struct devlink *devlink = priv_to_devlink(esw->dev);
 	int old_mode;
 
 	lockdep_assert_held_write(&esw->mode_lock);
@@ -1359,7 +1354,8 @@ void mlx5_eswitch_disable_locked(struct mlx5_eswitch *esw, bool clear_vf)
 	if (old_mode == MLX5_ESWITCH_OFFLOADS)
 		mlx5_rescan_drivers(esw->dev);
 
-	mlx5_esw_qos_destroy(esw);
+	devlink_rate_nodes_destroy(devlink);
+
 	mlx5_esw_acls_ns_cleanup(esw);
 
 	if (clear_vf)
@@ -1568,6 +1564,7 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev)
 	lockdep_register_key(&esw->mode_lock_key);
 	init_rwsem(&esw->mode_lock);
 	lockdep_set_class(&esw->mode_lock, &esw->mode_lock_key);
+	refcount_set(&esw->qos.refcnt, 0);
 
 	esw->enabled_vports = 0;
 	esw->mode = MLX5_ESWITCH_NONE;
@@ -1601,6 +1598,7 @@ void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw)
 
 	esw->dev->priv.eswitch = NULL;
 	destroy_workqueue(esw->work_queue);
+	WARN_ON(refcount_read(&esw->qos.refcnt));
 	lockdep_unregister_key(&esw->mode_lock_key);
 	mutex_destroy(&esw->state_lock);
 	WARN_ON(!xa_empty(&esw->offloads.vhca_map));
@@ -1690,82 +1688,6 @@ bool mlx5_esw_is_sf_vport(struct mlx5_eswitch *esw, u16 vport_num)
 	return mlx5_esw_check_port_type(esw, vport_num, MLX5_ESW_VPT_SF);
 }
 
-static bool
-is_port_function_supported(struct mlx5_eswitch *esw, u16 vport_num)
-{
-	return vport_num == MLX5_VPORT_PF ||
-	       mlx5_eswitch_is_vf_vport(esw, vport_num) ||
-	       mlx5_esw_is_sf_vport(esw, vport_num);
-}
-
-int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port,
-					   u8 *hw_addr, int *hw_addr_len,
-					   struct netlink_ext_ack *extack)
-{
-	struct mlx5_eswitch *esw;
-	struct mlx5_vport *vport;
-	int err = -EOPNOTSUPP;
-	u16 vport_num;
-
-	esw = mlx5_devlink_eswitch_get(port->devlink);
-	if (IS_ERR(esw))
-		return PTR_ERR(esw);
-
-	vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index);
-	if (!is_port_function_supported(esw, vport_num))
-		return -EOPNOTSUPP;
-
-	vport = mlx5_eswitch_get_vport(esw, vport_num);
-	if (IS_ERR(vport)) {
-		NL_SET_ERR_MSG_MOD(extack, "Invalid port");
-		return PTR_ERR(vport);
-	}
-
-	mutex_lock(&esw->state_lock);
-	if (vport->enabled) {
-		ether_addr_copy(hw_addr, vport->info.mac);
-		*hw_addr_len = ETH_ALEN;
-		err = 0;
-	}
-	mutex_unlock(&esw->state_lock);
-	return err;
-}
-
-int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port,
-					   const u8 *hw_addr, int hw_addr_len,
-					   struct netlink_ext_ack *extack)
-{
-	struct mlx5_eswitch *esw;
-	struct mlx5_vport *vport;
-	int err = -EOPNOTSUPP;
-	u16 vport_num;
-
-	esw = mlx5_devlink_eswitch_get(port->devlink);
-	if (IS_ERR(esw)) {
-		NL_SET_ERR_MSG_MOD(extack, "Eswitch doesn't support set hw_addr");
-		return PTR_ERR(esw);
-	}
-
-	vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index);
-	if (!is_port_function_supported(esw, vport_num)) {
-		NL_SET_ERR_MSG_MOD(extack, "Port doesn't support set hw_addr");
-		return -EINVAL;
-	}
-	vport = mlx5_eswitch_get_vport(esw, vport_num);
-	if (IS_ERR(vport)) {
-		NL_SET_ERR_MSG_MOD(extack, "Invalid port");
-		return PTR_ERR(vport);
-	}
-
-	mutex_lock(&esw->state_lock);
-	if (vport->enabled)
-		err = mlx5_esw_set_vport_mac_locked(esw, vport, hw_addr);
-	else
-		NL_SET_ERR_MSG_MOD(extack, "Eswitch vport is disabled");
-	mutex_unlock(&esw->state_lock);
-	return err;
-}
-
 int mlx5_eswitch_set_vport_state(struct mlx5_eswitch *esw,
 				 u16 vport, int link_state)
 {
@@ -1822,8 +1744,10 @@ int mlx5_eswitch_get_vport_config(struct mlx5_eswitch *esw,
 	ivi->qos = evport->info.qos;
 	ivi->spoofchk = evport->info.spoofchk;
 	ivi->trusted = evport->info.trusted;
-	ivi->min_tx_rate = evport->qos.min_rate;
-	ivi->max_tx_rate = evport->qos.max_rate;
+	if (evport->qos.enabled) {
+		ivi->min_tx_rate = evport->qos.min_rate;
+		ivi->max_tx_rate = evport->qos.max_rate;
+	}
 	mutex_unlock(&esw->state_lock);
 
 	return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index 42f8ee2e5d9f..513f741d16c7 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -308,10 +308,14 @@ struct mlx5_eswitch {
 	atomic64_t user_count;
 
 	struct {
-		bool            enabled;
 		u32             root_tsar_ix;
 		struct mlx5_esw_rate_group *group0;
 		struct list_head groups; /* Protected by esw->state_lock */
+
+		/* Protected by esw->state_lock.
+		 * Initially 0, meaning no QoS users and QoS is disabled.
+		 */
+		refcount_t refcnt;
 	} qos;
 
 	struct mlx5_esw_bridge_offloads *br_offloads;
@@ -516,11 +520,6 @@ int mlx5_eswitch_del_vlan_action(struct mlx5_eswitch *esw,
 int __mlx5_eswitch_set_vport_vlan(struct mlx5_eswitch *esw,
 				  u16 vport, u16 vlan, u8 qos, u8 set_flags);
 
-static inline bool mlx5_esw_qos_enabled(struct mlx5_eswitch *esw)
-{
-	return esw->qos.enabled;
-}
-
 static inline bool mlx5_eswitch_vlan_actions_supported(struct mlx5_core_dev *dev,
 						       u8 vlan_depth)
 {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index f4eaa5893886..4bd502ae82b6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -3862,3 +3862,62 @@ u32 mlx5_eswitch_get_vport_metadata_for_set(struct mlx5_eswitch *esw,
 	return vport->metadata;
 }
 EXPORT_SYMBOL(mlx5_eswitch_get_vport_metadata_for_set);
+
+static bool
+is_port_function_supported(struct mlx5_eswitch *esw, u16 vport_num)
+{
+	return vport_num == MLX5_VPORT_PF ||
+	       mlx5_eswitch_is_vf_vport(esw, vport_num) ||
+	       mlx5_esw_is_sf_vport(esw, vport_num);
+}
+
+int mlx5_devlink_port_function_hw_addr_get(struct devlink_port *port,
+					   u8 *hw_addr, int *hw_addr_len,
+					   struct netlink_ext_ack *extack)
+{
+	struct mlx5_eswitch *esw;
+	struct mlx5_vport *vport;
+	u16 vport_num;
+
+	esw = mlx5_devlink_eswitch_get(port->devlink);
+	if (IS_ERR(esw))
+		return PTR_ERR(esw);
+
+	vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index);
+	if (!is_port_function_supported(esw, vport_num))
+		return -EOPNOTSUPP;
+
+	vport = mlx5_eswitch_get_vport(esw, vport_num);
+	if (IS_ERR(vport)) {
+		NL_SET_ERR_MSG_MOD(extack, "Invalid port");
+		return PTR_ERR(vport);
+	}
+
+	mutex_lock(&esw->state_lock);
+	ether_addr_copy(hw_addr, vport->info.mac);
+	*hw_addr_len = ETH_ALEN;
+	mutex_unlock(&esw->state_lock);
+	return 0;
+}
+
+int mlx5_devlink_port_function_hw_addr_set(struct devlink_port *port,
+					   const u8 *hw_addr, int hw_addr_len,
+					   struct netlink_ext_ack *extack)
+{
+	struct mlx5_eswitch *esw;
+	u16 vport_num;
+
+	esw = mlx5_devlink_eswitch_get(port->devlink);
+	if (IS_ERR(esw)) {
+		NL_SET_ERR_MSG_MOD(extack, "Eswitch doesn't support set hw_addr");
+		return PTR_ERR(esw);
+	}
+
+	vport_num = mlx5_esw_devlink_port_index_to_vport_num(port->index);
+	if (!is_port_function_supported(esw, vport_num)) {
+		NL_SET_ERR_MSG_MOD(extack, "Port doesn't support set hw_addr");
+		return -EINVAL;
+	}
+
+	return mlx5_eswitch_set_vport_mac(esw, vport_num, hw_addr);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c
index 64f1abc4dc36..75121bc1eaa5 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c
@@ -420,6 +420,11 @@ static void print_health_info(struct mlx5_core_dev *dev)
 	if (!ioread8(&h->synd))
 		return;
 
+	if (ioread32be(&h->fw_ver) == 0xFFFFFFFF) {
+		mlx5_log(dev, LOGLEVEL_ERR, "PCI slot is unavailable\n");
+		return;
+	}
+
 	rfr_severity = ioread8(&h->rfr_severity);
 	severity  = mlx5_health_get_severity(rfr_severity);
 	mlx5_log(dev, severity, "Health issue observed, %s, severity(%d) %s:\n",
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c
index 962d41418ce7..f23e33ac9c6b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c
@@ -105,7 +105,7 @@ static int mlx5i_set_coalesce(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = mlx5i_epriv(netdev);
 
-	return mlx5e_ethtool_set_coalesce(priv, coal);
+	return mlx5e_ethtool_set_coalesce(priv, coal, kernel_coal, extack);
 }
 
 static int mlx5i_get_coalesce(struct net_device *netdev,
@@ -115,7 +115,7 @@ static int mlx5i_get_coalesce(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = mlx5i_epriv(netdev);
 
-	return mlx5e_ethtool_get_coalesce(priv, coal);
+	return mlx5e_ethtool_get_coalesce(priv, coal, kernel_coal);
 }
 
 static int mlx5i_get_ts_info(struct net_device *netdev,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
index 830444f927d4..19bf2b66707d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
@@ -479,7 +479,7 @@ irq_pool_alloc(struct mlx5_core_dev *dev, int start, int size, char *name,
 	pool->xa_num_irqs.max = start + size - 1;
 	if (name)
 		snprintf(pool->name, MLX5_MAX_IRQ_NAME - MLX5_MAX_IRQ_IDX_CHARS,
-			 name);
+			 "%s", name);
 	pool->min_threshold = min_threshold * MLX5_EQ_REFS_PER_IRQ;
 	pool->max_threshold = max_threshold * MLX5_EQ_REFS_PER_IRQ;
 	mlx5_core_dbg(dev, "pool->name = %s, pool->size = %d, pool->start = %d",