Commit ddbb5ddc authored by Rongwei Liu's avatar Rongwei Liu Committed by Jakub Kicinski
Browse files

net/mlx5: LAG, Refactor lag logic 



Wrap the lag pf access into two new macros:
1. ldev_for_each()
2. ldev_for_each_reverse()
The maximum number of lag ports and the index to `natvie_port_num`
mapping will be handled by the two new macros.
Users shouldn't use the for loop anymore.

Signed-off-by: default avatarRongwei Liu <rongweil@nvidia.com>
Reviewed-by: default avatarSaeed Mahameed <saeedm@nvidia.com>
Reviewed-by: default avatarMark Bloch <mbloch@nvidia.com>
Signed-off-by: default avatarTariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/20241219175841.1094544-2-tariqt@nvidia.com


Signed-off-by: default avatarJakub Kicinski <kuba@kernel.org>
parent 8d94a744

drivers/net/ethernet/mellanox/mlx5/core/lag/debugfs.c

+5 −8
Original line number Diff line number Diff line
@@ -105,20 +105,20 @@ static int mapping_show(struct seq_file *file, void *priv) 
	struct mlx5_lag *ldev;

	bool hash = false;

	bool lag_active;

	int i, idx = 0;

	int num_ports;

	int i;



	ldev = mlx5_lag_dev(dev);

	mutex_lock(&ldev->lock);

	lag_active = __mlx5_lag_is_active(ldev);

	if (lag_active) {

		if (test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &ldev->mode_flags)) {

			mlx5_infer_tx_enabled(&ldev->tracker, ldev->ports, ports,

			mlx5_infer_tx_enabled(&ldev->tracker, ldev, ports,

					      &num_ports);

			hash = true;

		} else {

			for (i = 0; i < ldev->ports; i++)

				ports[i] = ldev->v2p_map[i];

			mlx5_ldev_for_each(i, 0, ldev)

				ports[idx++] = ldev->v2p_map[i];

			num_ports = ldev->ports;

		}

	}
@@ -144,11 +144,8 @@ static int members_show(struct seq_file *file, void *priv) 


	ldev = mlx5_lag_dev(dev);

	mutex_lock(&ldev->lock);

	for (i = 0; i < ldev->ports; i++) {

		if (!ldev->pf[i].dev)

			continue;

	mlx5_ldev_for_each(i, 0, ldev)

		seq_printf(file, "%s\n", dev_name(ldev->pf[i].dev->device));

	}

	mutex_unlock(&ldev->lock);



	return 0;

drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c

+91 −90
Original line number Diff line number Diff line
@@ -43,10 +43,6 @@ 
#include "mp.h"

#include "mpesw.h"



enum {

	MLX5_LAG_EGRESS_PORT_1 = 1,

	MLX5_LAG_EGRESS_PORT_2,

};



/* General purpose, use for short periods of time.

 * Beware of lock dependencies (preferably, no locks should be acquired
@@ -72,7 +68,7 @@ static u8 lag_active_port_bits(struct mlx5_lag *ldev) 
	int num_enabled;

	int idx;



	mlx5_infer_tx_enabled(&ldev->tracker, ldev->ports, enabled_ports,

	mlx5_infer_tx_enabled(&ldev->tracker, ldev, enabled_ports,

			      &num_enabled);

	for (idx = 0; idx < num_enabled; idx++)

		active_port |= BIT_MASK(enabled_ports[idx]);
@@ -113,7 +109,7 @@ static int mlx5_cmd_create_lag(struct mlx5_core_dev *dev, u8 *ports, int mode, 
	return mlx5_cmd_exec_in(dev, create_lag, in);

}



static int mlx5_cmd_modify_lag(struct mlx5_core_dev *dev, u8 num_ports,

static int mlx5_cmd_modify_lag(struct mlx5_core_dev *dev, struct mlx5_lag *ldev,

			       u8 *ports)

{

	u32 in[MLX5_ST_SZ_DW(modify_lag_in)] = {};
@@ -148,33 +144,31 @@ int mlx5_cmd_destroy_vport_lag(struct mlx5_core_dev *dev) 
}

EXPORT_SYMBOL(mlx5_cmd_destroy_vport_lag);



static void mlx5_infer_tx_disabled(struct lag_tracker *tracker, u8 num_ports,

static void mlx5_infer_tx_disabled(struct lag_tracker *tracker, struct mlx5_lag *ldev,

				   u8 *ports, int *num_disabled)

{

	int i;



	*num_disabled = 0;

	for (i = 0; i < num_ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev)

		if (!tracker->netdev_state[i].tx_enabled ||

		    !tracker->netdev_state[i].link_up)

			ports[(*num_disabled)++] = i;

}

}



void mlx5_infer_tx_enabled(struct lag_tracker *tracker, u8 num_ports,

void mlx5_infer_tx_enabled(struct lag_tracker *tracker, struct mlx5_lag *ldev,

			   u8 *ports, int *num_enabled)

{

	int i;



	*num_enabled = 0;

	for (i = 0; i < num_ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev)

		if (tracker->netdev_state[i].tx_enabled &&

		    tracker->netdev_state[i].link_up)

			ports[(*num_enabled)++] = i;

	}



	if (*num_enabled == 0)

		mlx5_infer_tx_disabled(tracker, num_ports, ports, num_enabled);

		mlx5_infer_tx_disabled(tracker, ldev, ports, num_enabled);

}



static void mlx5_lag_print_mapping(struct mlx5_core_dev *dev,
@@ -192,7 +186,7 @@ static void mlx5_lag_print_mapping(struct mlx5_core_dev *dev, 
	int j;



	if (test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &flags)) {

		mlx5_infer_tx_enabled(tracker, ldev->ports, enabled_ports,

		mlx5_infer_tx_enabled(tracker, ldev, enabled_ports,

				      &num_enabled);

		for (i = 0; i < num_enabled; i++) {

			err = scnprintf(buf + written, 4, "%d, ", enabled_ports[i] + 1);
@@ -203,7 +197,7 @@ static void mlx5_lag_print_mapping(struct mlx5_core_dev *dev, 
		buf[written - 2] = 0;

		mlx5_core_info(dev, "lag map active ports: %s\n", buf);

	} else {

		for (i = 0; i < ldev->ports; i++) {

		mlx5_ldev_for_each(i, 0, ldev) {

			for (j  = 0; j < ldev->buckets; j++) {

				idx = i * ldev->buckets + j;

				err = scnprintf(buf + written, 10,
@@ -286,7 +280,7 @@ int mlx5_lag_dev_get_netdev_idx(struct mlx5_lag *ldev, 
{

	int i;



	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		if (ldev->pf[i].netdev == ndev)

			return i;
@@ -310,7 +304,7 @@ static bool __mlx5_lag_is_sriov(struct mlx5_lag *ldev) 
 * with mapping that points to active ports.

 */

static void mlx5_infer_tx_affinity_mapping(struct lag_tracker *tracker,

					   u8 num_ports,

					   struct mlx5_lag *ldev,

					   u8 buckets,

					   u8 *ports)

{
@@ -323,7 +317,7 @@ static void mlx5_infer_tx_affinity_mapping(struct lag_tracker *tracker, 
	int i;

	int j;



	for (i = 0; i < num_ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		if (tracker->netdev_state[i].tx_enabled &&

		    tracker->netdev_state[i].link_up)

			enabled[enabled_ports_num++] = i;
@@ -334,15 +328,16 @@ static void mlx5_infer_tx_affinity_mapping(struct lag_tracker *tracker, 
	/* Use native mapping by default where each port's buckets

	 * point the native port: 1 1 1 .. 1 2 2 2 ... 2 3 3 3 ... 3 etc

	 */

	for (i = 0; i < num_ports; i++)

	mlx5_ldev_for_each(i, 0, ldev) {

		for (j = 0; j < buckets; j++) {

			idx = i * buckets + j;

			ports[idx] = MLX5_LAG_EGRESS_PORT_1 + i;

			ports[idx] = i + 1;

		}

	}



	/* If all ports are disabled/enabled keep native mapping */

	if (enabled_ports_num == num_ports ||

	    disabled_ports_num == num_ports)

	if (enabled_ports_num == ldev->ports ||

	    disabled_ports_num == ldev->ports)

		return;



	/* Go over the disabled ports and for each assign a random active port */
@@ -358,7 +353,7 @@ static bool mlx5_lag_has_drop_rule(struct mlx5_lag *ldev) 
{

	int i;



	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		if (ldev->pf[i].has_drop)

			return true;

	return false;
@@ -368,7 +363,7 @@ static void mlx5_lag_drop_rule_cleanup(struct mlx5_lag *ldev) 
{

	int i;



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		if (!ldev->pf[i].has_drop)

			continue;
@@ -396,7 +391,7 @@ static void mlx5_lag_drop_rule_setup(struct mlx5_lag *ldev, 
	if (!ldev->tracker.has_inactive)

		return;



	mlx5_infer_tx_disabled(tracker, ldev->ports, disabled_ports, &num_disabled);

	mlx5_infer_tx_disabled(tracker, ldev, disabled_ports, &num_disabled);



	for (i = 0; i < num_disabled; i++) {

		disabled_index = disabled_ports[i];
@@ -442,7 +437,7 @@ static int _mlx5_modify_lag(struct mlx5_lag *ldev, u8 *ports) 


		return mlx5_cmd_modify_active_port(dev0, active_ports);

	}

	return mlx5_cmd_modify_lag(dev0, ldev->ports, ports);

	return mlx5_cmd_modify_lag(dev0, ldev, ports);

}



static struct net_device *mlx5_lag_active_backup_get_netdev(struct mlx5_core_dev *dev)
@@ -458,7 +453,7 @@ static struct net_device *mlx5_lag_active_backup_get_netdev(struct mlx5_core_dev 
	if (!ldev)

		goto unlock;



	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		if (ldev->tracker.netdev_state[i].tx_enabled)

			ndev = ldev->pf[i].netdev;

	if (!ndev)
@@ -483,9 +478,9 @@ void mlx5_modify_lag(struct mlx5_lag *ldev, 
	int i;

	int j;



	mlx5_infer_tx_affinity_mapping(tracker, ldev->ports, ldev->buckets, ports);

	mlx5_infer_tx_affinity_mapping(tracker, ldev, ldev->buckets, ports);



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		for (j = 0; j < ldev->buckets; j++) {

			idx = i * ldev->buckets + j;

			if (ports[idx] == ldev->v2p_map[idx])
@@ -596,9 +591,9 @@ static int mlx5_lag_create_single_fdb(struct mlx5_lag *ldev) 
	struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev;

	struct mlx5_eswitch *master_esw = dev0->priv.eswitch;

	int err;

	int i;

	int i, j;



	for (i = MLX5_LAG_P1 + 1; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 1, ldev) {

		struct mlx5_eswitch *slave_esw = ldev->pf[i].dev->priv.eswitch;



		err = mlx5_eswitch_offloads_single_fdb_add_one(master_esw,
@@ -608,9 +603,9 @@ static int mlx5_lag_create_single_fdb(struct mlx5_lag *ldev) 
	}

	return 0;

err:

	for (; i > MLX5_LAG_P1; i--)

	mlx5_ldev_for_each_reverse(j, i, 1, ldev)

		mlx5_eswitch_offloads_single_fdb_del_one(master_esw,

							 ldev->pf[i].dev->priv.eswitch);

							 ldev->pf[j].dev->priv.eswitch);

	return err;

}
@@ -671,7 +666,7 @@ int mlx5_activate_lag(struct mlx5_lag *ldev, 
		return err;



	if (mode != MLX5_LAG_MODE_MPESW) {

		mlx5_infer_tx_affinity_mapping(tracker, ldev->ports, ldev->buckets, ldev->v2p_map);

		mlx5_infer_tx_affinity_mapping(tracker, ldev, ldev->buckets, ldev->v2p_map);

		if (test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &flags)) {

			err = mlx5_lag_port_sel_create(ldev, tracker->hash_type,

						       ldev->v2p_map);
@@ -722,7 +717,7 @@ int mlx5_deactivate_lag(struct mlx5_lag *ldev) 
	mlx5_lag_mp_reset(ldev);



	if (test_bit(MLX5_LAG_MODE_FLAG_SHARED_FDB, &flags)) {

		for (i = MLX5_LAG_P1 + 1; i < ldev->ports; i++)

		mlx5_ldev_for_each(i, 1, ldev)

			mlx5_eswitch_offloads_single_fdb_del_one(master_esw,

								 ldev->pf[i].dev->priv.eswitch);

		clear_bit(MLX5_LAG_MODE_FLAG_SHARED_FDB, &flags);
@@ -766,7 +761,7 @@ bool mlx5_lag_check_prereq(struct mlx5_lag *ldev) 
			return false;



#ifdef CONFIG_MLX5_ESWITCH

	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		dev = ldev->pf[i].dev;

		if (mlx5_eswitch_num_vfs(dev->priv.eswitch) && !is_mdev_switchdev_mode(dev))

			return false;
@@ -774,17 +769,17 @@ bool mlx5_lag_check_prereq(struct mlx5_lag *ldev) 


	dev = ldev->pf[MLX5_LAG_P1].dev;

	mode = mlx5_eswitch_mode(dev);

	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		if (mlx5_eswitch_mode(ldev->pf[i].dev) != mode)

			return false;



#else

	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		if (mlx5_sriov_is_enabled(ldev->pf[i].dev))

			return false;

#endif

	roce_support = mlx5_get_roce_state(ldev->pf[MLX5_LAG_P1].dev);

	for (i = 1; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, MLX5_LAG_P2, ldev)

		if (mlx5_get_roce_state(ldev->pf[i].dev) != roce_support)

			return false;
@@ -795,10 +790,7 @@ void mlx5_lag_add_devices(struct mlx5_lag *ldev) 
{

	int i;



	for (i = 0; i < ldev->ports; i++) {

		if (!ldev->pf[i].dev)

			continue;



	mlx5_ldev_for_each(i, 0, ldev) {

		if (ldev->pf[i].dev->priv.flags &

		    MLX5_PRIV_FLAGS_DISABLE_ALL_ADEV)

			continue;
@@ -812,10 +804,7 @@ void mlx5_lag_remove_devices(struct mlx5_lag *ldev) 
{

	int i;



	for (i = 0; i < ldev->ports; i++) {

		if (!ldev->pf[i].dev)

			continue;



	mlx5_ldev_for_each(i, 0, ldev) {

		if (ldev->pf[i].dev->priv.flags &

		    MLX5_PRIV_FLAGS_DISABLE_ALL_ADEV)

			continue;
@@ -842,7 +831,7 @@ void mlx5_disable_lag(struct mlx5_lag *ldev) 
			dev0->priv.flags |= MLX5_PRIV_FLAGS_DISABLE_IB_ADEV;

			mlx5_rescan_drivers_locked(dev0);

		}

		for (i = 1; i < ldev->ports; i++)

		mlx5_ldev_for_each(i, MLX5_LAG_P2, ldev)

			mlx5_nic_vport_disable_roce(ldev->pf[i].dev);

	}
@@ -854,7 +843,7 @@ void mlx5_disable_lag(struct mlx5_lag *ldev) 
		mlx5_lag_add_devices(ldev);



	if (shared_fdb)

		for (i = 0; i < ldev->ports; i++)

		mlx5_ldev_for_each(i, 0, ldev)

			if (!(ldev->pf[i].dev->priv.flags & MLX5_PRIV_FLAGS_DISABLE_ALL_ADEV))

				mlx5_eswitch_reload_ib_reps(ldev->pf[i].dev->priv.eswitch);

}
@@ -864,7 +853,7 @@ static bool mlx5_shared_fdb_supported(struct mlx5_lag *ldev) 
	struct mlx5_core_dev *dev;

	int i;



	for (i = MLX5_LAG_P1 + 1; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, MLX5_LAG_P1 + 1, ldev) {

		dev = ldev->pf[i].dev;

		if (is_mdev_switchdev_mode(dev) &&

		    mlx5_eswitch_vport_match_metadata_enabled(dev->priv.eswitch) &&
@@ -892,11 +881,11 @@ static bool mlx5_lag_is_roce_lag(struct mlx5_lag *ldev) 
	bool roce_lag = true;

	int i;



	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		roce_lag = roce_lag && !mlx5_sriov_is_enabled(ldev->pf[i].dev);



#ifdef CONFIG_MLX5_ESWITCH

	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		roce_lag = roce_lag && is_mdev_legacy_mode(ldev->pf[i].dev);

#endif
@@ -956,7 +945,7 @@ static void mlx5_do_bond(struct mlx5_lag *ldev) 
		} else if (roce_lag) {

			dev0->priv.flags &= ~MLX5_PRIV_FLAGS_DISABLE_IB_ADEV;

			mlx5_rescan_drivers_locked(dev0);

			for (i = 1; i < ldev->ports; i++) {

			mlx5_ldev_for_each(i, MLX5_LAG_P2, ldev) {

				if (mlx5_get_roce_state(ldev->pf[i].dev))

					mlx5_nic_vport_enable_roce(ldev->pf[i].dev);

			}
@@ -966,7 +955,7 @@ static void mlx5_do_bond(struct mlx5_lag *ldev) 
			dev0->priv.flags &= ~MLX5_PRIV_FLAGS_DISABLE_IB_ADEV;

			mlx5_rescan_drivers_locked(dev0);



			for (i = 0; i < ldev->ports; i++) {

			mlx5_ldev_for_each(i, 0, ldev) {

				err = mlx5_eswitch_reload_ib_reps(ldev->pf[i].dev->priv.eswitch);

				if (err)

					break;
@@ -977,7 +966,7 @@ static void mlx5_do_bond(struct mlx5_lag *ldev) 
				mlx5_rescan_drivers_locked(dev0);

				mlx5_deactivate_lag(ldev);

				mlx5_lag_add_devices(ldev);

				for (i = 0; i < ldev->ports; i++)

				mlx5_ldev_for_each(i, 0, ldev)

					mlx5_eswitch_reload_ib_reps(ldev->pf[i].dev->priv.eswitch);

				mlx5_core_err(dev0, "Failed to enable lag\n");

				return;
@@ -1010,12 +999,9 @@ struct mlx5_devcom_comp_dev *mlx5_lag_get_devcom_comp(struct mlx5_lag *ldev) 
	int i;



	mutex_lock(&ldev->lock);

	for (i = 0; i < ldev->ports; i++) {

		if (ldev->pf[i].dev) {

	i = mlx5_get_next_ldev_func(ldev, 0);

	if (i < MLX5_MAX_PORTS)

		devcom = ldev->pf[i].dev->priv.hca_devcom_comp;

			break;

		}

	}

	mutex_unlock(&ldev->lock);

	return devcom;

}
@@ -1068,7 +1054,7 @@ static int mlx5_handle_changeupper_event(struct mlx5_lag *ldev, 
	u8 bond_status = 0;

	int num_slaves = 0;

	int changed = 0;

	int idx;

	int i, idx = -1;



	if (!netif_is_lag_master(upper))

		return 0;
@@ -1083,8 +1069,13 @@ static int mlx5_handle_changeupper_event(struct mlx5_lag *ldev, 
	 */

	rcu_read_lock();

	for_each_netdev_in_bond_rcu(upper, ndev_tmp) {

		idx = mlx5_lag_dev_get_netdev_idx(ldev, ndev_tmp);

		if (idx >= 0) {

		mlx5_ldev_for_each(i, 0, ldev) {

			if (ldev->pf[i].netdev == ndev_tmp) {

				idx++;

				break;

			}

		}

		if (i < MLX5_MAX_PORTS) {

			slave = bond_slave_get_rcu(ndev_tmp);

			if (slave)

				has_inactive |= bond_is_slave_inactive(slave);
@@ -1240,9 +1231,6 @@ static void mlx5_ldev_add_netdev(struct mlx5_lag *ldev, 
	unsigned int fn = mlx5_get_dev_index(dev);

	unsigned long flags;



	if (fn >= ldev->ports)

		return;



	spin_lock_irqsave(&lag_lock, flags);

	ldev->pf[fn].netdev = netdev;

	ldev->tracker.netdev_state[fn].link_up = 0;
@@ -1257,7 +1245,7 @@ static void mlx5_ldev_remove_netdev(struct mlx5_lag *ldev, 
	int i;



	spin_lock_irqsave(&lag_lock, flags);

	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		if (ldev->pf[i].netdev == netdev) {

			ldev->pf[i].netdev = NULL;

			break;
@@ -1271,9 +1259,6 @@ static void mlx5_ldev_add_mdev(struct mlx5_lag *ldev, 
{

	unsigned int fn = mlx5_get_dev_index(dev);



	if (fn >= ldev->ports)

		return;



	ldev->pf[fn].dev = dev;

	dev->priv.lag = ldev;

}
@@ -1281,16 +1266,13 @@ static void mlx5_ldev_add_mdev(struct mlx5_lag *ldev, 
static void mlx5_ldev_remove_mdev(struct mlx5_lag *ldev,

				  struct mlx5_core_dev *dev)

{

	int i;



	for (i = 0; i < ldev->ports; i++)

		if (ldev->pf[i].dev == dev)

			break;

	int fn;



	if (i == ldev->ports)

	fn = mlx5_get_dev_index(dev);

	if (ldev->pf[fn].dev != dev)

		return;



	ldev->pf[i].dev = NULL;

	ldev->pf[fn].dev = NULL;

	dev->priv.lag = NULL;

}
@@ -1406,7 +1388,6 @@ void mlx5_lag_add_netdev(struct mlx5_core_dev *dev, 


	mutex_lock(&ldev->lock);

	mlx5_ldev_add_netdev(ldev, dev, netdev);



	for (i = 0; i < ldev->ports; i++)

		if (!ldev->pf[i].netdev)

			break;
@@ -1417,6 +1398,26 @@ void mlx5_lag_add_netdev(struct mlx5_core_dev *dev, 
	mlx5_queue_bond_work(ldev, 0);

}



int mlx5_get_pre_ldev_func(struct mlx5_lag *ldev, int start_idx, int end_idx)

{

	int i;



	for (i = start_idx; i >= end_idx; i--)

		if (ldev->pf[i].dev)

			return i;

	return -1;

}



int mlx5_get_next_ldev_func(struct mlx5_lag *ldev, int start_idx)

{

	int i;



	for (i = start_idx; i < MLX5_MAX_PORTS; i++)

		if (ldev->pf[i].dev)

			return i;

	return MLX5_MAX_PORTS;

}



bool mlx5_lag_is_roce(struct mlx5_core_dev *dev)

{

	struct mlx5_lag *ldev;
@@ -1467,7 +1468,7 @@ bool mlx5_lag_is_master(struct mlx5_core_dev *dev) 
{

	struct mlx5_lag *ldev;

	unsigned long flags;

	bool res;

	bool res = false;



	spin_lock_irqsave(&lag_lock, flags);

	ldev = mlx5_lag_dev(dev);
@@ -1555,7 +1556,7 @@ u8 mlx5_lag_get_slave_port(struct mlx5_core_dev *dev, 
	if (!(ldev && __mlx5_lag_is_roce(ldev)))

		goto unlock;



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		if (ldev->pf[i].netdev == slave) {

			port = i;

			break;
@@ -1594,13 +1595,13 @@ struct mlx5_core_dev *mlx5_lag_get_next_peer_mdev(struct mlx5_core_dev *dev, int 
	if (!ldev)

		goto unlock;



	if (*i == ldev->ports)

	if (*i == MLX5_MAX_PORTS)

		goto unlock;

	for (idx = *i; idx < ldev->ports; idx++)

	mlx5_ldev_for_each(idx, *i, ldev)

		if (ldev->pf[idx].dev != dev)

			break;



	if (idx == ldev->ports) {

	if (idx == MLX5_MAX_PORTS) {

		*i = idx;

		goto unlock;

	}
@@ -1621,10 +1622,10 @@ int mlx5_lag_query_cong_counters(struct mlx5_core_dev *dev, 
{

	int outlen = MLX5_ST_SZ_BYTES(query_cong_statistics_out);

	struct mlx5_core_dev **mdev;

	int ret = 0, i, j, idx = 0;

	struct mlx5_lag *ldev;

	unsigned long flags;

	int num_ports;

	int ret, i, j;

	void *out;



	out = kvzalloc(outlen, GFP_KERNEL);
@@ -1643,8 +1644,8 @@ int mlx5_lag_query_cong_counters(struct mlx5_core_dev *dev, 
	ldev = mlx5_lag_dev(dev);

	if (ldev && __mlx5_lag_is_active(ldev)) {

		num_ports = ldev->ports;

		for (i = 0; i < ldev->ports; i++)

			mdev[i] = ldev->pf[i].dev;

		mlx5_ldev_for_each(i, 0, ldev)

			mdev[idx++] = ldev->pf[i].dev;

	} else {

		num_ports = 1;

		mdev[MLX5_LAG_P1] = dev;

drivers/net/ethernet/mellanox/mlx5/core/lag/lag.h

+13 −1
Original line number Diff line number Diff line
@@ -103,7 +103,7 @@ int mlx5_lag_dev_get_netdev_idx(struct mlx5_lag *ldev, 
				struct net_device *ndev);



char *mlx5_get_str_port_sel_mode(enum mlx5_lag_mode mode, unsigned long flags);

void mlx5_infer_tx_enabled(struct lag_tracker *tracker, u8 num_ports,

void mlx5_infer_tx_enabled(struct lag_tracker *tracker, struct mlx5_lag *ldev,

			   u8 *ports, int *num_enabled);



void mlx5_ldev_add_debugfs(struct mlx5_core_dev *dev);
@@ -119,9 +119,21 @@ static inline bool mlx5_lag_is_supported(struct mlx5_core_dev *dev) 
	if (!MLX5_CAP_GEN(dev, vport_group_manager) ||

	    !MLX5_CAP_GEN(dev, lag_master) ||

	    MLX5_CAP_GEN(dev, num_lag_ports) < 2 ||

	    mlx5_get_dev_index(dev) >= MLX5_MAX_PORTS ||

	    MLX5_CAP_GEN(dev, num_lag_ports) > MLX5_MAX_PORTS)

		return false;

	return true;

}



#define mlx5_ldev_for_each(i, start_index, ldev) \

	for (int tmp = start_index; tmp = mlx5_get_next_ldev_func(ldev, tmp), \

	     i = tmp, tmp < MLX5_MAX_PORTS; tmp++)



#define mlx5_ldev_for_each_reverse(i, start_index, end_index, ldev)      \

	for (int tmp = start_index, tmp1 = end_index; \

	     tmp = mlx5_get_pre_ldev_func(ldev, tmp, tmp1), \

	     i = tmp, tmp >= tmp1; tmp--)



int mlx5_get_pre_ldev_func(struct mlx5_lag *ldev, int start_idx, int end_idx);

int mlx5_get_next_ldev_func(struct mlx5_lag *ldev, int start_idx);

#endif /* __MLX5_LAG_H__ */

drivers/net/ethernet/mellanox/mlx5/core/lag/mp.c

+16 −8
Original line number Diff line number Diff line
@@ -153,6 +153,7 @@ static void mlx5_lag_fib_route_event(struct mlx5_lag *ldev, unsigned long event, 
	struct net_device *nh_dev0, *nh_dev1;

	struct fib_info *fi = fen_info->fi;

	struct lag_mp *mp = &ldev->lag_mp;

	int i, dev_idx = 0;



	/* Handle delete event */

	if (event == FIB_EVENT_ENTRY_DEL) {
@@ -186,10 +187,12 @@ static void mlx5_lag_fib_route_event(struct mlx5_lag *ldev, unsigned long event, 


	if (!nh_dev1) {

		if (__mlx5_lag_is_active(ldev)) {

			int i = mlx5_lag_dev_get_netdev_idx(ldev, nh_dev0);



			i++;

			mlx5_lag_set_port_affinity(ldev, i);

			mlx5_ldev_for_each(i, 0, ldev) {

				dev_idx++;

				if (ldev->pf[i].netdev == nh_dev0)

					break;

			}

			mlx5_lag_set_port_affinity(ldev, dev_idx);

			mlx5_lag_fib_set(mp, fi, fen_info->dst, fen_info->dst_len);

		}
@@ -214,6 +217,7 @@ static void mlx5_lag_fib_nexthop_event(struct mlx5_lag *ldev, 
				       struct fib_info *fi)

{

	struct lag_mp *mp = &ldev->lag_mp;

	int i, dev_idx = 0;



	/* Check the nh event is related to the route */

	if (!mp->fib.mfi || mp->fib.mfi != fi)
@@ -221,11 +225,15 @@ static void mlx5_lag_fib_nexthop_event(struct mlx5_lag *ldev, 


	/* nh added/removed */

	if (event == FIB_EVENT_NH_DEL) {

		int i = mlx5_lag_dev_get_netdev_idx(ldev, fib_nh->fib_nh_dev);

		mlx5_ldev_for_each(i, 0, ldev) {

			if (ldev->pf[i].netdev == fib_nh->fib_nh_dev)

				break;

			dev_idx++;

		}



		if (i >= 0) {

			i = (i + 1) % 2 + 1; /* peer port */

			mlx5_lag_set_port_affinity(ldev, i);

		if (dev_idx >= 0) {

			dev_idx = (dev_idx + 1) % 2 + 1; /* peer port */

			mlx5_lag_set_port_affinity(ldev, dev_idx);

		}

	} else if (event == FIB_EVENT_NH_ADD &&

		   fib_info_num_path(fi) == 2) {

drivers/net/ethernet/mellanox/mlx5/core/lag/mpesw.c

+5 −5
Original line number Diff line number Diff line
@@ -15,7 +15,7 @@ static void mlx5_mpesw_metadata_cleanup(struct mlx5_lag *ldev) 
	u32 pf_metadata;

	int i;



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		dev = ldev->pf[i].dev;

		esw = dev->priv.eswitch;

		pf_metadata = ldev->lag_mpesw.pf_metadata[i];
@@ -36,7 +36,7 @@ static int mlx5_mpesw_metadata_set(struct mlx5_lag *ldev) 
	u32 pf_metadata;

	int i, err;



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		dev = ldev->pf[i].dev;

		esw = dev->priv.eswitch;

		pf_metadata = mlx5_esw_match_metadata_alloc(esw);
@@ -52,7 +52,7 @@ static int mlx5_mpesw_metadata_set(struct mlx5_lag *ldev) 
			goto err_metadata;

	}



	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		dev = ldev->pf[i].dev;

		mlx5_notifier_call_chain(dev->priv.events, MLX5_DEV_EVENT_MULTIPORT_ESW,

					 (void *)0);
@@ -98,7 +98,7 @@ static int enable_mpesw(struct mlx5_lag *ldev) 


	dev0->priv.flags &= ~MLX5_PRIV_FLAGS_DISABLE_IB_ADEV;

	mlx5_rescan_drivers_locked(dev0);

	for (i = 0; i < ldev->ports; i++) {

	mlx5_ldev_for_each(i, 0, ldev) {

		err = mlx5_eswitch_reload_ib_reps(ldev->pf[i].dev->priv.eswitch);

		if (err)

			goto err_rescan_drivers;
@@ -112,7 +112,7 @@ static int enable_mpesw(struct mlx5_lag *ldev) 
	mlx5_deactivate_lag(ldev);

err_add_devices:

	mlx5_lag_add_devices(ldev);

	for (i = 0; i < ldev->ports; i++)

	mlx5_ldev_for_each(i, 0, ldev)

		mlx5_eswitch_reload_ib_reps(ldev->pf[i].dev->priv.eswitch);

	mlx5_mpesw_metadata_cleanup(ldev);

	return err;