Merge tag 'mlx5-updates-2021-10-15' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5-updates-2021-10-15

1) From Rongwei Liu:

Use system_image_guid and native_port_num when bonding.
Don't relay on PCIe ids anymore.

With some specific NIC, the physical devices may have PCIe IDs like
0001:01:00.0/1 and 0002:02:00.0/1. All of these devices should have
the same system_image_guid and device index can be queried from
native_port_num.

For matching sibling devices/port of the same HCA, compare the HCA
GUID reported on each device rather than just assuming PCIe ids have
similar attributes.

2) From Amir Tzin: Use HCA defined Timouts

Replace hard coded timeouts with values stored by firmware in default
timeouts register (DTOR). Timeouts are read during driver load. If DTOR
is not supported by firmware then fallback to hard coded defaults
instead.

3) From Shay Drory: Disable roce at HCA level
Disable RoCE in Firmware when devlink roce parameter is set to off.

4) A small set of trivial cleanups
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/mellanox/mlx5/core/Makefile

@@ -17,7 +17,7 @@ mlx5_core-y :=	main.o cmd.o debugfs.o fw.o eq.o uar.o pagealloc.o \
 		fs_counters.o fs_ft_pool.o rl.o lag.o dev.o events.o wq.o lib/gid.o \
 		lib/devcom.o lib/pci_vsc.o lib/dm.o lib/fs_ttc.o diag/fs_tracepoint.o \
 		diag/fw_tracer.o diag/crdump.o devlink.o diag/rsc_dump.o \
-		fw_reset.o qos.o
+		fw_reset.o qos.o lib/tout.o
 
 #
 # Netdev basic

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

@@ -45,6 +45,7 @@
 
 #include "mlx5_core.h"
 #include "lib/eq.h"
+#include "lib/tout.h"
 
 enum {
 	CMD_IF_REV = 5,
@@ -225,9 +226,13 @@ static void set_signature(struct mlx5_cmd_work_ent *ent, int csum)
 
 static void poll_timeout(struct mlx5_cmd_work_ent *ent)
 {
-	unsigned long poll_end = jiffies + msecs_to_jiffies(MLX5_CMD_TIMEOUT_MSEC + 1000);
+	struct mlx5_core_dev *dev = container_of(ent->cmd, struct mlx5_core_dev, cmd);
+	u64 cmd_to_ms = mlx5_tout_ms(dev, CMD);
+	unsigned long poll_end;
 	u8 own;
 
+	poll_end = jiffies + msecs_to_jiffies(cmd_to_ms + 1000);
+
 	do {
 		own = READ_ONCE(ent->lay->status_own);
 		if (!(own & CMD_OWNER_HW)) {
@@ -925,15 +930,18 @@ static void cmd_work_handler(struct work_struct *work)
 {
 	struct mlx5_cmd_work_ent *ent = container_of(work, struct mlx5_cmd_work_ent, work);
 	struct mlx5_cmd *cmd = ent->cmd;
-	struct mlx5_core_dev *dev = container_of(cmd, struct mlx5_core_dev, cmd);
-	unsigned long cb_timeout = msecs_to_jiffies(MLX5_CMD_TIMEOUT_MSEC);
+	bool poll_cmd = ent->polling;
 	struct mlx5_cmd_layout *lay;
+	struct mlx5_core_dev *dev;
+	unsigned long cb_timeout;
 	struct semaphore *sem;
 	unsigned long flags;
-	bool poll_cmd = ent->polling;
 	int alloc_ret;
 	int cmd_mode;
 
+	dev = container_of(cmd, struct mlx5_core_dev, cmd);
+	cb_timeout = msecs_to_jiffies(mlx5_tout_ms(dev, CMD));
+
 	complete(&ent->handling);
 	sem = ent->page_queue ? &cmd->pages_sem : &cmd->sem;
 	down(sem);
@@ -1073,7 +1081,7 @@ static void wait_func_handle_exec_timeout(struct mlx5_core_dev *dev,
 
 static int wait_func(struct mlx5_core_dev *dev, struct mlx5_cmd_work_ent *ent)
 {
-	unsigned long timeout = msecs_to_jiffies(MLX5_CMD_TIMEOUT_MSEC);
+	unsigned long timeout = msecs_to_jiffies(mlx5_tout_ms(dev, CMD));
 	struct mlx5_cmd *cmd = &dev->cmd;
 	int err;
 

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

@@ -33,6 +33,7 @@
 #include <linux/mlx5/driver.h>
 #include <linux/mlx5/eswitch.h>
 #include <linux/mlx5/mlx5_ifc_vdpa.h>
+#include <linux/mlx5/vport.h>
 #include "mlx5_core.h"
 
 /* intf dev list mutex */
@@ -537,6 +538,16 @@ int mlx5_rescan_drivers_locked(struct mlx5_core_dev *dev)
 	return add_drivers(dev);
 }
 
+static bool mlx5_same_hw_devs(struct mlx5_core_dev *dev, struct mlx5_core_dev *peer_dev)
+{
+	u64 fsystem_guid, psystem_guid;
+
+	fsystem_guid = mlx5_query_nic_system_image_guid(dev);
+	psystem_guid = mlx5_query_nic_system_image_guid(peer_dev);
+
+	return (fsystem_guid && psystem_guid && fsystem_guid == psystem_guid);
+}
+
 static u32 mlx5_gen_pci_id(const struct mlx5_core_dev *dev)
 {
 	return (u32)((pci_domain_nr(dev->pdev->bus) << 16) |
@@ -556,7 +567,8 @@ static int next_phys_dev(struct device *dev, const void *data)
 	if (mdev == curr)
 		return 0;
 
-	if (mlx5_gen_pci_id(mdev) != mlx5_gen_pci_id(curr))
+	if (!mlx5_same_hw_devs(mdev, (struct mlx5_core_dev *)curr) &&
+	    mlx5_gen_pci_id(mdev) != mlx5_gen_pci_id(curr))
 		return 0;
 
 	return 1;

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

@@ -454,7 +454,8 @@ static int mlx5_devlink_enable_roce_validate(struct devlink *devlink, u32 id,
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
 	bool new_state = val.vbool;
 
-	if (new_state && !MLX5_CAP_GEN(dev, roce)) {
+	if (new_state && !MLX5_CAP_GEN(dev, roce) &&
+	    !MLX5_CAP_GEN(dev, roce_rw_supported)) {
 		NL_SET_ERR_MSG_MOD(extack, "Device doesn't support RoCE");
 		return -EOPNOTSUPP;
 	}

drivers/net/ethernet/mellanox/mlx5/core/en/devlink.c

@@ -24,7 +24,7 @@ int mlx5e_devlink_port_register(struct mlx5e_priv *priv)
 
 	if (mlx5_core_is_pf(priv->mdev)) {
 		attrs.flavour = DEVLINK_PORT_FLAVOUR_PHYSICAL;
-		attrs.phys.port_number = PCI_FUNC(priv->mdev->pdev->devfn);
+		attrs.phys.port_number = mlx5_get_dev_index(priv->mdev);
 		if (MLX5_ESWITCH_MANAGER(priv->mdev)) {
 			mlx5e_devlink_get_port_parent_id(priv->mdev, &ppid);
 			memcpy(attrs.switch_id.id, ppid.id, ppid.id_len);

drivers/net/ethernet/mellanox/mlx5/core/en/fs.h

@@ -125,15 +125,15 @@ struct mlx5e_ethtool_steering {
 
 void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv);
 void mlx5e_ethtool_cleanup_steering(struct mlx5e_priv *priv);
-int mlx5e_ethtool_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd);
-int mlx5e_ethtool_get_rxnfc(struct net_device *dev,
+int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd);
+int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv,
 			    struct ethtool_rxnfc *info, u32 *rule_locs);
 #else
 static inline void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv)    { }
 static inline void mlx5e_ethtool_cleanup_steering(struct mlx5e_priv *priv) { }
-static inline int mlx5e_ethtool_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
+static inline int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd)
 { return -EOPNOTSUPP; }
-static inline int mlx5e_ethtool_get_rxnfc(struct net_device *dev,
+static inline int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv,
 					  struct ethtool_rxnfc *info, u32 *rule_locs)
 { return -EOPNOTSUPP; }
 #endif /* CONFIG_MLX5_EN_RXNFC */

drivers/net/ethernet/mellanox/mlx5/core/en/health.h

@@ -32,7 +32,6 @@ void mlx5e_reporter_rq_cqe_err(struct mlx5e_rq *rq);
 void mlx5e_reporter_rx_timeout(struct mlx5e_rq *rq);
 
 #define MLX5E_REPORTER_PER_Q_MAX_LEN 256
-#define MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC 2000
 
 struct mlx5e_err_ctx {
 	int (*recover)(void *ctx);

drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c

@@ -6,6 +6,7 @@
 #include "txrx.h"
 #include "devlink.h"
 #include "ptp.h"
+#include "lib/tout.h"
 
 static int mlx5e_query_rq_state(struct mlx5_core_dev *dev, u32 rqn, u8 *state)
 {
@@ -32,8 +33,10 @@ static int mlx5e_query_rq_state(struct mlx5_core_dev *dev, u32 rqn, u8 *state)
 
 static int mlx5e_wait_for_icosq_flush(struct mlx5e_icosq *icosq)
 {
-	unsigned long exp_time = jiffies +
-				 msecs_to_jiffies(MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC);
+	struct mlx5_core_dev *dev = icosq->channel->mdev;
+	unsigned long exp_time;
+
+	exp_time = jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, FLUSH_ON_ERROR));
 
 	while (time_before(jiffies, exp_time)) {
 		if (icosq->cc == icosq->pc)

drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c

@@ -4,11 +4,14 @@
 #include "health.h"
 #include "en/ptp.h"
 #include "en/devlink.h"
+#include "lib/tout.h"
 
 static int mlx5e_wait_for_sq_flush(struct mlx5e_txqsq *sq)
 {
-	unsigned long exp_time = jiffies +
-				 msecs_to_jiffies(MLX5E_REPORTER_FLUSH_TIMEOUT_MSEC);
+	struct mlx5_core_dev *dev = sq->mdev;
+	unsigned long exp_time;
+
+	exp_time = jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, FLUSH_ON_ERROR));
 
 	while (time_before(jiffies, exp_time)) {
 		if (sq->cc == sq->pc)

drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c

@@ -2141,6 +2141,7 @@ mlx5_tc_ct_init(struct mlx5e_priv *priv, struct mlx5_fs_chains *chains,
 err_ct_tuples_ht:
 	rhashtable_destroy(&ct_priv->zone_ht);
 err_ct_zone_ht:
+	mlx5_chains_destroy_global_table(chains, ct_priv->ct_nat);
 err_ct_nat_tbl:
 	mlx5_chains_destroy_global_table(chains, ct_priv->ct);
 err_ct_tbl:

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

@@ -2137,12 +2137,14 @@ int mlx5e_get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
 		return 0;
 	}
 
-	return mlx5e_ethtool_get_rxnfc(dev, info, rule_locs);
+	return mlx5e_ethtool_get_rxnfc(priv, info, rule_locs);
 }
 
 int mlx5e_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
 {
-	return mlx5e_ethtool_set_rxnfc(dev, cmd);
+	struct mlx5e_priv *priv = netdev_priv(dev);
+
+	return mlx5e_ethtool_set_rxnfc(priv, cmd);
 }
 
 static int query_port_status_opcode(struct mlx5_core_dev *mdev, u32 *status_opcode)

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

@@ -937,9 +937,8 @@ static int mlx5e_get_rss_hash_opt(struct mlx5e_priv *priv,
 	return 0;
 }
 
-int mlx5e_ethtool_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
+int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd)
 {
-	struct mlx5e_priv *priv = netdev_priv(dev);
 	int err = 0;
 
 	switch (cmd->cmd) {
@@ -960,10 +959,9 @@ int mlx5e_ethtool_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
 	return err;
 }
 
-int mlx5e_ethtool_get_rxnfc(struct net_device *dev,
+int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv,
 			    struct ethtool_rxnfc *info, u32 *rule_locs)
 {
-	struct mlx5e_priv *priv = netdev_priv(dev);
 	int err = 0;
 
 	switch (info->cmd) {

drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c

@@ -677,6 +677,10 @@ mlx5_esw_bridge_egress_flow_create(u16 vport_num, u16 esw_owner_vhca_id, const u
 	if (!rule_spec)
 		return ERR_PTR(-ENOMEM);
 
+	if (MLX5_CAP_ESW_FLOWTABLE(bridge->br_offloads->esw->dev, flow_source) &&
+	    vport_num == MLX5_VPORT_UPLINK)
+		rule_spec->flow_context.flow_source =
+			MLX5_FLOW_CONTEXT_FLOW_SOURCE_LOCAL_VPORT;
 	rule_spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
 
 	dmac_v = MLX5_ADDR_OF(fte_match_param, rule_spec->match_value,

drivers/net/ethernet/mellanox/mlx5/core/esw/devlink_port.c

@@ -36,7 +36,7 @@ static struct devlink_port *mlx5_esw_dl_port_alloc(struct mlx5_eswitch *esw, u16
 		return NULL;
 
 	mlx5_esw_get_port_parent_id(dev, &ppid);
-	pfnum = PCI_FUNC(dev->pdev->devfn);
+	pfnum = mlx5_get_dev_index(dev);
 	external = mlx5_core_is_ecpf_esw_manager(dev);
 	if (external)
 		controller_num = dev->priv.eswitch->offloads.host_number + 1;
@@ -149,7 +149,7 @@ int mlx5_esw_devlink_sf_port_register(struct mlx5_eswitch *esw, struct devlink_p
 	if (IS_ERR(vport))
 		return PTR_ERR(vport);
 
-	pfnum = PCI_FUNC(dev->pdev->devfn);
+	pfnum = mlx5_get_dev_index(dev);
 	mlx5_esw_get_port_parent_id(dev, &ppid);
 	memcpy(dl_port->attrs.switch_id.id, &ppid.id[0], ppid.id_len);
 	dl_port->attrs.switch_id.id_len = ppid.id_len;

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

@@ -2798,7 +2798,7 @@ u32 mlx5_esw_match_metadata_alloc(struct mlx5_eswitch *esw)
 	int id;
 
 	/* Only 4 bits of pf_num */
-	pf_num = PCI_FUNC(esw->dev->pdev->devfn);
+	pf_num = mlx5_get_dev_index(esw->dev);
 	if (pf_num > max_pf_num)
 		return 0;
 

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

@@ -35,6 +35,7 @@
 #include <linux/module.h>
 #include "mlx5_core.h"
 #include "../../mlxfw/mlxfw.h"
+#include "lib/tout.h"
 #include "accel/tls.h"
 
 enum {
@@ -317,10 +318,9 @@ int mlx5_cmd_force_teardown_hca(struct mlx5_core_dev *dev)
 	return 0;
 }
 
-#define MLX5_FAST_TEARDOWN_WAIT_MS   3000
 int mlx5_cmd_fast_teardown_hca(struct mlx5_core_dev *dev)
 {
-	unsigned long end, delay_ms = MLX5_FAST_TEARDOWN_WAIT_MS;
+	unsigned long end, delay_ms = mlx5_tout_ms(dev, TEARDOWN);
 	u32 out[MLX5_ST_SZ_DW(teardown_hca_out)] = {};
 	u32 in[MLX5_ST_SZ_DW(teardown_hca_in)] = {};
 	int state;
@@ -618,17 +618,18 @@ static void mlx5_fsm_release(struct mlxfw_dev *mlxfw_dev, u32 fwhandle)
 			 fwhandle, 0);
 }
 
-#define MLX5_FSM_REACTIVATE_TOUT 5000 /* msecs */
 static int mlx5_fsm_reactivate(struct mlxfw_dev *mlxfw_dev, u8 *status)
 {
-	unsigned long exp_time = jiffies + msecs_to_jiffies(MLX5_FSM_REACTIVATE_TOUT);
 	struct mlx5_mlxfw_dev *mlx5_mlxfw_dev =
 		container_of(mlxfw_dev, struct mlx5_mlxfw_dev, mlxfw_dev);
 	struct mlx5_core_dev *dev = mlx5_mlxfw_dev->mlx5_core_dev;
 	u32 out[MLX5_ST_SZ_DW(mirc_reg)];
 	u32 in[MLX5_ST_SZ_DW(mirc_reg)];
+	unsigned long exp_time;
 	int err;
 
+	exp_time = jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, FSM_REACTIVATE));
+
 	if (!MLX5_CAP_MCAM_REG2(dev, mirc))
 		return -EOPNOTSUPP;
 

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

@@ -3,6 +3,7 @@
 
 #include "fw_reset.h"
 #include "diag/fw_tracer.h"
+#include "lib/tout.h"
 
 enum {
 	MLX5_FW_RESET_FLAGS_RESET_REQUESTED,
@@ -228,8 +229,6 @@ static void mlx5_sync_reset_request_event(struct work_struct *work)
 		mlx5_core_warn(dev, "PCI Sync FW Update Reset Ack. Device reset is expected.\n");
 }
 
-#define MLX5_PCI_LINK_UP_TIMEOUT 2000
-
 static int mlx5_pci_link_toggle(struct mlx5_core_dev *dev)
 {
 	struct pci_bus *bridge_bus = dev->pdev->bus;
@@ -286,7 +285,7 @@ static int mlx5_pci_link_toggle(struct mlx5_core_dev *dev)
 		goto restore;
 	}
 
-	timeout = jiffies + msecs_to_jiffies(MLX5_PCI_LINK_UP_TIMEOUT);
+	timeout = jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, PCI_TOGGLE));
 	do {
 		err = pci_read_config_word(bridge, cap + PCI_EXP_LNKSTA, &reg16);
 		if (err)
@@ -299,8 +298,8 @@ static int mlx5_pci_link_toggle(struct mlx5_core_dev *dev)
 	if (reg16 & PCI_EXP_LNKSTA_DLLLA) {
 		mlx5_core_info(dev, "PCI Link up\n");
 	} else {
-		mlx5_core_err(dev, "PCI link not ready (0x%04x) after %d ms\n",
-			      reg16, MLX5_PCI_LINK_UP_TIMEOUT);
+		mlx5_core_err(dev, "PCI link not ready (0x%04x) after %llu ms\n",
+			      reg16, mlx5_tout_ms(dev, PCI_TOGGLE));
 		err = -ETIMEDOUT;
 	}
 
@@ -395,16 +394,15 @@ static int fw_reset_event_notifier(struct notifier_block *nb, unsigned long acti
 	return NOTIFY_OK;
 }
 
-#define MLX5_FW_RESET_TIMEOUT_MSEC 5000
 int mlx5_fw_reset_wait_reset_done(struct mlx5_core_dev *dev)
 {
-	unsigned long timeout = msecs_to_jiffies(MLX5_FW_RESET_TIMEOUT_MSEC);
+	unsigned long timeout = msecs_to_jiffies(mlx5_tout_ms(dev, PCI_SYNC_UPDATE));
 	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
 	int err;
 
 	if (!wait_for_completion_timeout(&fw_reset->done, timeout)) {
-		mlx5_core_warn(dev, "FW sync reset timeout after %d seconds\n",
-			       MLX5_FW_RESET_TIMEOUT_MSEC / 1000);
+		mlx5_core_warn(dev, "FW sync reset timeout after %llu seconds\n",
+			       mlx5_tout_ms(dev, PCI_SYNC_UPDATE) / 1000);
 		err = -ETIMEDOUT;
 		goto out;
 	}

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

@@ -40,10 +40,10 @@
 #include "lib/eq.h"
 #include "lib/mlx5.h"
 #include "lib/pci_vsc.h"
+#include "lib/tout.h"
 #include "diag/fw_tracer.h"
 
 enum {
-	MLX5_HEALTH_POLL_INTERVAL	= 2 * HZ,
 	MAX_MISSES			= 3,
 };
 
@@ -219,11 +219,9 @@ void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force)
 	mutex_unlock(&dev->intf_state_mutex);
 }
 
-#define MLX5_CRDUMP_WAIT_MS	60000
-#define MLX5_FW_RESET_WAIT_MS	1000
 void mlx5_error_sw_reset(struct mlx5_core_dev *dev)
 {
-	unsigned long end, delay_ms = MLX5_FW_RESET_WAIT_MS;
+	unsigned long end, delay_ms = mlx5_tout_ms(dev, PCI_TOGGLE);
 	int lock = -EBUSY;
 
 	mutex_lock(&dev->intf_state_mutex);
@@ -237,7 +235,7 @@ void mlx5_error_sw_reset(struct mlx5_core_dev *dev)
 		lock = lock_sem_sw_reset(dev, true);
 
 		if (lock == -EBUSY) {
-			delay_ms = MLX5_CRDUMP_WAIT_MS;
+			delay_ms = mlx5_tout_ms(dev, FULL_CRDUMP);
 			goto recover_from_sw_reset;
 		}
 		/* Execute SW reset */
@@ -307,13 +305,11 @@ static void mlx5_handle_bad_state(struct mlx5_core_dev *dev)
 	mlx5_disable_device(dev);
 }
 
-/* How much time to wait until health resetting the driver (in msecs) */
-#define MLX5_RECOVERY_WAIT_MSECS 60000
 int mlx5_health_wait_pci_up(struct mlx5_core_dev *dev)
 {
 	unsigned long end;
 
-	end = jiffies + msecs_to_jiffies(MLX5_RECOVERY_WAIT_MSECS);
+	end = jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, FW_RESET));
 	while (sensor_pci_not_working(dev)) {
 		if (time_after(jiffies, end))
 			return -ETIMEDOUT;
@@ -674,13 +670,13 @@ static void mlx5_fw_reporters_destroy(struct mlx5_core_dev *dev)
 		devlink_health_reporter_destroy(health->fw_fatal_reporter);
 }
 
-static unsigned long get_next_poll_jiffies(void)
+static unsigned long get_next_poll_jiffies(struct mlx5_core_dev *dev)
 {
 	unsigned long next;
 
 	get_random_bytes(&next, sizeof(next));
 	next %= HZ;
-	next += jiffies + MLX5_HEALTH_POLL_INTERVAL;
+	next += jiffies + msecs_to_jiffies(mlx5_tout_ms(dev, HEALTH_POLL_INTERVAL));
 
 	return next;
 }
@@ -740,11 +736,12 @@ static void poll_health(struct timer_list *t)
 		queue_work(health->wq, &health->report_work);
 
 out:
-	mod_timer(&health->timer, get_next_poll_jiffies());
+	mod_timer(&health->timer, get_next_poll_jiffies(dev));
 }
 
 void mlx5_start_health_poll(struct mlx5_core_dev *dev)
 {
+	u64 poll_interval_ms =  mlx5_tout_ms(dev, HEALTH_POLL_INTERVAL);
 	struct mlx5_core_health *health = &dev->priv.health;
 
 	timer_setup(&health->timer, poll_health, 0);
@@ -753,7 +750,7 @@ void mlx5_start_health_poll(struct mlx5_core_dev *dev)
 	health->health = &dev->iseg->health;
 	health->health_counter = &dev->iseg->health_counter;
 
-	health->timer.expires = round_jiffies(jiffies + MLX5_HEALTH_POLL_INTERVAL);
+	health->timer.expires = jiffies + msecs_to_jiffies(poll_interval_ms);
 	add_timer(&health->timer);
 }
 

drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c

@@ -33,6 +33,11 @@
 #include "en.h"
 #include "ipoib.h"
 
+static u32 mlx5i_flow_type_mask(u32 flow_type)
+{
+	return flow_type & ~(FLOW_EXT | FLOW_MAC_EXT | FLOW_RSS);
+}
+
 static void mlx5i_get_drvinfo(struct net_device *dev,
 			      struct ethtool_drvinfo *drvinfo)
 {
@@ -217,6 +222,27 @@ static int mlx5i_get_link_ksettings(struct net_device *netdev,
 	return 0;
 }
 
+#ifdef CONFIG_MLX5_EN_RXNFC
+static int mlx5i_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
+{
+	struct mlx5e_priv *priv = mlx5i_epriv(dev);
+	struct ethtool_rx_flow_spec *fs = &cmd->fs;
+
+	if (mlx5i_flow_type_mask(fs->flow_type) == ETHER_FLOW)
+		return -EINVAL;
+
+	return mlx5e_ethtool_set_rxnfc(priv, cmd);
+}
+
+static int mlx5i_get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
+			   u32 *rule_locs)
+{
+	struct mlx5e_priv *priv = mlx5i_epriv(dev);
+
+	return mlx5e_ethtool_get_rxnfc(priv, info, rule_locs);
+}
+#endif
+
 const struct ethtool_ops mlx5i_ethtool_ops = {
 	.supported_coalesce_params = ETHTOOL_COALESCE_USECS |
 				     ETHTOOL_COALESCE_MAX_FRAMES |
@@ -233,6 +259,10 @@ const struct ethtool_ops mlx5i_ethtool_ops = {
 	.get_coalesce       = mlx5i_get_coalesce,
 	.set_coalesce       = mlx5i_set_coalesce,
 	.get_ts_info        = mlx5i_get_ts_info,
+#ifdef CONFIG_MLX5_EN_RXNFC
+	.get_rxnfc          = mlx5i_get_rxnfc,
+	.set_rxnfc          = mlx5i_set_rxnfc,
+#endif
 	.get_link_ksettings = mlx5i_get_link_ksettings,
 	.get_link           = ethtool_op_get_link,
 };

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

@@ -336,6 +336,8 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
 		goto err_destroy_arfs_tables;
 	}
 
+	mlx5e_ethtool_init_steering(priv);
+
 	return 0;
 
 err_destroy_arfs_tables:
@@ -348,6 +350,7 @@ static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv)
 {
 	mlx5e_destroy_ttc_table(priv);
 	mlx5e_arfs_destroy_tables(priv);
+	mlx5e_ethtool_cleanup_steering(priv);
 }
 
 static int mlx5i_init_rx(struct mlx5e_priv *priv)

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

@@ -688,7 +688,7 @@ static void mlx5_ldev_add_netdev(struct mlx5_lag *ldev,
 				 struct mlx5_core_dev *dev,
 				 struct net_device *netdev)
 {
-	unsigned int fn = PCI_FUNC(dev->pdev->devfn);
+	unsigned int fn = mlx5_get_dev_index(dev);
 
 	if (fn >= MLX5_MAX_PORTS)
 		return;
@@ -718,7 +718,7 @@ static void mlx5_ldev_remove_netdev(struct mlx5_lag *ldev,
 static void mlx5_ldev_add_mdev(struct mlx5_lag *ldev,
 			       struct mlx5_core_dev *dev)
 {
-	unsigned int fn = PCI_FUNC(dev->pdev->devfn);
+	unsigned int fn = mlx5_get_dev_index(dev);
 
 	if (fn >= MLX5_MAX_PORTS)
 		return;

drivers/net/ethernet/mellanox/mlx5/core/lib/tout.c

+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/* Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */
+
+#include <linux/mlx5/driver.h>
+#include "lib/tout.h"
+
+struct mlx5_timeouts {
+	u64 to[MAX_TIMEOUT_TYPES];
+};
+
+static const u32 tout_def_sw_val[MAX_TIMEOUT_TYPES] = {
+	[MLX5_TO_FW_PRE_INIT_TIMEOUT_MS] = 120000,
+	[MLX5_TO_FW_PRE_INIT_WARN_MESSAGE_INTERVAL_MS] = 20000,
+	[MLX5_TO_FW_PRE_INIT_WAIT_MS] = 2,
+	[MLX5_TO_FW_INIT_MS] = 2000,
+	[MLX5_TO_CMD_MS] = 60000,
+	[MLX5_TO_PCI_TOGGLE_MS] =  2000,
+	[MLX5_TO_HEALTH_POLL_INTERVAL_MS] =  2000,
+	[MLX5_TO_FULL_CRDUMP_MS] = 60000,
+	[MLX5_TO_FW_RESET_MS] = 60000,
+	[MLX5_TO_FLUSH_ON_ERROR_MS] = 2000,
+	[MLX5_TO_PCI_SYNC_UPDATE_MS] = 5000,
+	[MLX5_TO_TEARDOWN_MS] = 3000,
+	[MLX5_TO_FSM_REACTIVATE_MS] = 5000,
+	[MLX5_TO_RECLAIM_PAGES_MS] = 5000,
+	[MLX5_TO_RECLAIM_VFS_PAGES_MS] = 120000
+};
+
+static void tout_set(struct mlx5_core_dev *dev, u64 val, enum mlx5_timeouts_types type)
+{
+	dev->timeouts->to[type] = val;
+}
+
+static void tout_set_def_val(struct mlx5_core_dev *dev)
+{
+	int i;
+
+	for (i = MLX5_TO_FW_PRE_INIT_TIMEOUT_MS; i < MAX_TIMEOUT_TYPES; i++)
+		tout_set(dev, tout_def_sw_val[i], i);
+}
+
+int mlx5_tout_init(struct mlx5_core_dev *dev)
+{
+	dev->timeouts = kmalloc(sizeof(*dev->timeouts), GFP_KERNEL);
+	if (!dev->timeouts)
+		return -ENOMEM;
+
+	tout_set_def_val(dev);
+	return 0;
+}
+
+void mlx5_tout_cleanup(struct mlx5_core_dev *dev)
+{
+	kfree(dev->timeouts);
+}
+
+/* Time register consists of two fields to_multiplier(time out multiplier)
+ * and to_value(time out value). to_value is the quantity of the time units and
+ * to_multiplier is the type and should be one off these four values.
+ * 0x0: millisecond
+ * 0x1: seconds
+ * 0x2: minutes
+ * 0x3: hours
+ * this function converts the time stored in the two register fields into
+ * millisecond.
+ */
+static u64 tout_convert_reg_field_to_ms(u32 to_mul, u32 to_val)
+{
+	u64 msec = to_val;
+
+	to_mul &= 0x3;
+	/* convert hours/minutes/seconds to miliseconds */
+	if (to_mul)
+		msec *= 1000 * int_pow(60, to_mul - 1);
+
+	return msec;
+}
+
+static u64 tout_convert_iseg_to_ms(u32 iseg_to)
+{
+	return tout_convert_reg_field_to_ms(iseg_to >> 29, iseg_to & 0xfffff);
+}
+
+static bool tout_is_supported(struct mlx5_core_dev *dev)
+{
+	return !!ioread32be(&dev->iseg->cmd_q_init_to);
+}
+
+void mlx5_tout_query_iseg(struct mlx5_core_dev *dev)
+{
+	u32 to;
+
+	if (!tout_is_supported(dev))
+		return;
+
+	to = ioread32be(&dev->iseg->cmd_q_init_to);
+	tout_set(dev, tout_convert_iseg_to_ms(to), MLX5_TO_FW_INIT_MS);
+
+	to = ioread32be(&dev->iseg->cmd_exec_to);
+	tout_set(dev, tout_convert_iseg_to_ms(to), MLX5_TO_CMD_MS);
+}
+
+u64 _mlx5_tout_ms(struct mlx5_core_dev *dev, enum mlx5_timeouts_types type)
+{
+	return dev->timeouts->to[type];
+}
+
+#define MLX5_TIMEOUT_QUERY(fld, reg_out) \
+	({ \
+	struct mlx5_ifc_default_timeout_bits *time_field; \
+	u32 to_multi, to_value; \
+	u64 to_val_ms; \
+	\
+	time_field = MLX5_ADDR_OF(dtor_reg, reg_out, fld); \
+	to_multi = MLX5_GET(default_timeout, time_field, to_multiplier); \
+	to_value = MLX5_GET(default_timeout, time_field, to_value); \
+	to_val_ms = tout_convert_reg_field_to_ms(to_multi, to_value); \
+	to_val_ms; \
+	})
+
+#define MLX5_TIMEOUT_FILL(fld, reg_out, dev, to_type, to_extra) \
+	({ \
+	u64 fw_to = MLX5_TIMEOUT_QUERY(fld, reg_out); \
+	tout_set(dev, fw_to + (to_extra), to_type); \
+	fw_to; \
+	})
+
+static int tout_query_dtor(struct mlx5_core_dev *dev)
+{
+	u64 pcie_toggle_to_val, tear_down_to_val;
+	u32 out[MLX5_ST_SZ_DW(dtor_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(dtor_reg)] = {};
+	int err;
+
+	err = mlx5_core_access_reg(dev, in, sizeof(in), out, sizeof(out), MLX5_REG_DTOR, 0, 0);
+	if (err)
+		return err;
+
+	pcie_toggle_to_val = MLX5_TIMEOUT_FILL(pcie_toggle_to, out, dev, MLX5_TO_PCI_TOGGLE_MS, 0);
+	MLX5_TIMEOUT_FILL(fw_reset_to, out, dev, MLX5_TO_FW_RESET_MS, pcie_toggle_to_val);
+
+	tear_down_to_val = MLX5_TIMEOUT_FILL(tear_down_to, out, dev, MLX5_TO_TEARDOWN_MS, 0);
+	MLX5_TIMEOUT_FILL(pci_sync_update_to, out, dev, MLX5_TO_PCI_SYNC_UPDATE_MS,
+			  tear_down_to_val);
+
+	MLX5_TIMEOUT_FILL(health_poll_to, out, dev, MLX5_TO_HEALTH_POLL_INTERVAL_MS, 0);
+	MLX5_TIMEOUT_FILL(full_crdump_to, out, dev, MLX5_TO_FULL_CRDUMP_MS, 0);
+	MLX5_TIMEOUT_FILL(flush_on_err_to, out, dev, MLX5_TO_FLUSH_ON_ERROR_MS, 0);
+	MLX5_TIMEOUT_FILL(fsm_reactivate_to, out, dev, MLX5_TO_FSM_REACTIVATE_MS, 0);
+	MLX5_TIMEOUT_FILL(reclaim_pages_to, out, dev, MLX5_TO_RECLAIM_PAGES_MS, 0);
+	MLX5_TIMEOUT_FILL(reclaim_vfs_pages_to, out, dev, MLX5_TO_RECLAIM_VFS_PAGES_MS, 0);
+
+	return 0;
+}
+
+int mlx5_tout_query_dtor(struct mlx5_core_dev *dev)
+{
+	if (tout_is_supported(dev))
+		return tout_query_dtor(dev);
+
+	return 0;
+}

drivers/net/ethernet/mellanox/mlx5/core/lib/tout.h

+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */
+
+#ifndef MLX5_TIMEOUTS_H
+#define MLX5_TIMEOUTS_H
+
+enum mlx5_timeouts_types {
+	/* pre init timeouts (not read from FW) */
+	MLX5_TO_FW_PRE_INIT_TIMEOUT_MS,
+	MLX5_TO_FW_PRE_INIT_WARN_MESSAGE_INTERVAL_MS,
+	MLX5_TO_FW_PRE_INIT_WAIT_MS,
+
+	/* init segment timeouts */
+	MLX5_TO_FW_INIT_MS,
+	MLX5_TO_CMD_MS,
+
+	/* DTOR timeouts */
+	MLX5_TO_PCI_TOGGLE_MS,
+	MLX5_TO_HEALTH_POLL_INTERVAL_MS,
+	MLX5_TO_FULL_CRDUMP_MS,
+	MLX5_TO_FW_RESET_MS,
+	MLX5_TO_FLUSH_ON_ERROR_MS,
+	MLX5_TO_PCI_SYNC_UPDATE_MS,
+	MLX5_TO_TEARDOWN_MS,
+	MLX5_TO_FSM_REACTIVATE_MS,
+	MLX5_TO_RECLAIM_PAGES_MS,
+	MLX5_TO_RECLAIM_VFS_PAGES_MS,
+
+	MAX_TIMEOUT_TYPES
+};
+
+struct mlx5_core_dev;
+int mlx5_tout_init(struct mlx5_core_dev *dev);
+void mlx5_tout_cleanup(struct mlx5_core_dev *dev);
+void mlx5_tout_query_iseg(struct mlx5_core_dev *dev);
+int mlx5_tout_query_dtor(struct mlx5_core_dev *dev);
+u64 _mlx5_tout_ms(struct mlx5_core_dev *dev, enum mlx5_timeouts_types type);
+
+#define mlx5_tout_ms(dev, type) _mlx5_tout_ms(dev, MLX5_TO_##type##_MS)
+
+# endif /* MLX5_TIMEOUTS_H */

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

@@ -60,6 +60,7 @@
 #include "devlink.h"
 #include "fw_reset.h"
 #include "lib/mlx5.h"
+#include "lib/tout.h"
 #include "fpga/core.h"
 #include "fpga/ipsec.h"
 #include "accel/ipsec.h"
@@ -176,11 +177,6 @@ static struct mlx5_profile profile[] = {
 	},
 };
 
-#define FW_INIT_TIMEOUT_MILI		2000
-#define FW_INIT_WAIT_MS			2
-#define FW_PRE_INIT_TIMEOUT_MILI	120000
-#define FW_INIT_WARN_MESSAGE_INTERVAL	20000
-
 static int fw_initializing(struct mlx5_core_dev *dev)
 {
 	return ioread32be(&dev->iseg->initializing) >> 31;
@@ -193,8 +189,6 @@ static int wait_fw_init(struct mlx5_core_dev *dev, u32 max_wait_mili,
 	unsigned long end = jiffies + msecs_to_jiffies(max_wait_mili);
 	int err = 0;
 
-	BUILD_BUG_ON(FW_PRE_INIT_TIMEOUT_MILI < FW_INIT_WARN_MESSAGE_INTERVAL);
-
 	while (fw_initializing(dev)) {
 		if (time_after(jiffies, end)) {
 			err = -EBUSY;
@@ -205,7 +199,7 @@ static int wait_fw_init(struct mlx5_core_dev *dev, u32 max_wait_mili,
 				       jiffies_to_msecs(end - warn) / 1000);
 			warn = jiffies + msecs_to_jiffies(warn_time_mili);
 		}
-		msleep(FW_INIT_WAIT_MS);
+		msleep(mlx5_tout_ms(dev, FW_PRE_INIT_WAIT));
 	}
 
 	return err;
@@ -564,15 +558,38 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx)
 		MLX5_SET(cmd_hca_cap, set_hca_cap, num_total_dynamic_vf_msix,
 			 MLX5_CAP_GEN_MAX(dev, num_total_dynamic_vf_msix));
 
+	if (MLX5_CAP_GEN(dev, roce_rw_supported))
+		MLX5_SET(cmd_hca_cap, set_hca_cap, roce, mlx5_is_roce_init_enabled(dev));
+
 	return set_caps(dev, set_ctx, MLX5_SET_HCA_CAP_OP_MOD_GENERAL_DEVICE);
 }
 
+/* Cached MLX5_CAP_GEN(dev, roce) can be out of sync this early in the
+ * boot process.
+ * In case RoCE cap is writable in FW and user/devlink requested to change the
+ * cap, we are yet to query the final state of the above cap.
+ * Hence, the need for this function.
+ *
+ * Returns
+ * True:
+ * 1) RoCE cap is read only in FW and already disabled
+ * OR:
+ * 2) RoCE cap is writable in FW and user/devlink requested it off.
+ *
+ * In any other case, return False.
+ */
+static bool is_roce_fw_disabled(struct mlx5_core_dev *dev)
+{
+	return (MLX5_CAP_GEN(dev, roce_rw_supported) && !mlx5_is_roce_init_enabled(dev)) ||
+		(!MLX5_CAP_GEN(dev, roce_rw_supported) && !MLX5_CAP_GEN(dev, roce));
+}
+
 static int handle_hca_cap_roce(struct mlx5_core_dev *dev, void *set_ctx)
 {
 	void *set_hca_cap;
 	int err;
 
-	if (!MLX5_CAP_GEN(dev, roce))
+	if (is_roce_fw_disabled(dev))
 		return 0;
 
 	err = mlx5_core_get_caps(dev, MLX5_CAP_ROCE);
@@ -975,25 +992,34 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot)
 	if (mlx5_core_is_pf(dev))
 		pcie_print_link_status(dev->pdev);
 
+	err = mlx5_tout_init(dev);
+	if (err) {
+		mlx5_core_err(dev, "Failed initializing timeouts, aborting\n");
+		return err;
+	}
+
 	/* wait for firmware to accept initialization segments configurations
 	 */
-	err = wait_fw_init(dev, FW_PRE_INIT_TIMEOUT_MILI, FW_INIT_WARN_MESSAGE_INTERVAL);
+	err = wait_fw_init(dev, mlx5_tout_ms(dev, FW_PRE_INIT_TIMEOUT),
+			   mlx5_tout_ms(dev, FW_PRE_INIT_WARN_MESSAGE_INTERVAL));
 	if (err) {
-		mlx5_core_err(dev, "Firmware over %d MS in pre-initializing state, aborting\n",
-			      FW_PRE_INIT_TIMEOUT_MILI);
-		return err;
+		mlx5_core_err(dev, "Firmware over %llu MS in pre-initializing state, aborting\n",
+			      mlx5_tout_ms(dev, FW_PRE_INIT_TIMEOUT));
+		goto err_tout_cleanup;
 	}
 
 	err = mlx5_cmd_init(dev);
 	if (err) {
 		mlx5_core_err(dev, "Failed initializing command interface, aborting\n");
-		return err;
+		goto err_tout_cleanup;
 	}
 
-	err = wait_fw_init(dev, FW_INIT_TIMEOUT_MILI, 0);
+	mlx5_tout_query_iseg(dev);
+
+	err = wait_fw_init(dev, mlx5_tout_ms(dev, FW_INIT), 0);
 	if (err) {
-		mlx5_core_err(dev, "Firmware over %d MS in initializing state, aborting\n",
-			      FW_INIT_TIMEOUT_MILI);
+		mlx5_core_err(dev, "Firmware over %llu MS in initializing state, aborting\n",
+			      mlx5_tout_ms(dev, FW_INIT));
 		goto err_cmd_cleanup;
 	}
 
@@ -1017,6 +1043,12 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot)
 		goto err_disable_hca;
 	}
 
+	err = mlx5_tout_query_dtor(dev);
+	if (err) {
+		mlx5_core_err(dev, "failed to read dtor\n");
+		goto reclaim_boot_pages;
+	}
+
 	err = set_hca_ctrl(dev);
 	if (err) {
 		mlx5_core_err(dev, "set_hca_ctrl failed\n");
@@ -1062,6 +1094,8 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot)
 err_cmd_cleanup:
 	mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
 	mlx5_cmd_cleanup(dev);
+err_tout_cleanup:
+	mlx5_tout_cleanup(dev);
 
 	return err;
 }
@@ -1080,6 +1114,7 @@ static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot)
 	mlx5_core_disable_hca(dev, 0);
 	mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN);
 	mlx5_cmd_cleanup(dev);
+	mlx5_tout_cleanup(dev);
 
 	return 0;
 }

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

@@ -38,6 +38,7 @@
 #include <linux/xarray.h>
 #include "mlx5_core.h"
 #include "lib/eq.h"
+#include "lib/tout.h"
 
 enum {
 	MLX5_PAGES_CANT_GIVE	= 0,
@@ -64,11 +65,6 @@ struct fw_page {
 	unsigned int free_count;
 };
 
-enum {
-	MAX_RECLAIM_TIME_MSECS	= 5000,
-	MAX_RECLAIM_VFS_PAGES_TIME_MSECS = 2 * 1000 * 60,
-};
-
 enum {
 	MLX5_MAX_RECLAIM_TIME_MILI	= 5000,
 	MLX5_NUM_4K_IN_PAGE		= PAGE_SIZE / MLX5_ADAPTER_PAGE_SIZE,
@@ -641,7 +637,8 @@ static int optimal_reclaimed_pages(void)
 static int mlx5_reclaim_root_pages(struct mlx5_core_dev *dev,
 				   struct rb_root *root, u16 func_id)
 {
-	unsigned long end = jiffies + msecs_to_jiffies(MAX_RECLAIM_TIME_MSECS);
+	u64 recl_pages_to_jiffies = msecs_to_jiffies(mlx5_tout_ms(dev, RECLAIM_PAGES));
+	unsigned long end = jiffies + recl_pages_to_jiffies;
 
 	while (!RB_EMPTY_ROOT(root)) {
 		int nclaimed;
@@ -656,7 +653,7 @@ static int mlx5_reclaim_root_pages(struct mlx5_core_dev *dev,
 		}
 
 		if (nclaimed)
-			end = jiffies + msecs_to_jiffies(MAX_RECLAIM_TIME_MSECS);
+			end = jiffies + recl_pages_to_jiffies;
 
 		if (time_after(jiffies, end)) {
 			mlx5_core_warn(dev, "FW did not return all pages. giving up...\n");
@@ -727,7 +724,8 @@ void mlx5_pagealloc_stop(struct mlx5_core_dev *dev)
 
 int mlx5_wait_for_pages(struct mlx5_core_dev *dev, int *pages)
 {
-	unsigned long end = jiffies + msecs_to_jiffies(MAX_RECLAIM_VFS_PAGES_TIME_MSECS);
+	u64 recl_vf_pages_to_jiffies = msecs_to_jiffies(mlx5_tout_ms(dev, RECLAIM_VFS_PAGES));
+	unsigned long end = jiffies + recl_vf_pages_to_jiffies;
 	int prev_pages = *pages;
 
 	/* In case of internal error we will free the pages manually later */
@@ -743,7 +741,7 @@ int mlx5_wait_for_pages(struct mlx5_core_dev *dev, int *pages)
 			return -ETIMEDOUT;
 		}
 		if (*pages < prev_pages) {
-			end = jiffies + msecs_to_jiffies(MAX_RECLAIM_VFS_PAGES_TIME_MSECS);
+			end = jiffies + recl_vf_pages_to_jiffies;
 			prev_pages = *pages;
 		}
 		msleep(50);

drivers/net/ethernet/mellanox/mlx5/core/sf/devlink.c

@@ -323,7 +323,7 @@ mlx5_sf_new_check_attr(struct mlx5_core_dev *dev, const struct devlink_port_new_
 		NL_SET_ERR_MSG_MOD(extack, "External controller is unsupported");
 		return -EOPNOTSUPP;
 	}
-	if (new_attr->pfnum != PCI_FUNC(dev->pdev->devfn)) {
+	if (new_attr->pfnum != mlx5_get_dev_index(dev)) {
 		NL_SET_ERR_MSG_MOD(extack, "Invalid pfnum supplied");
 		return -EOPNOTSUPP;
 	}

drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c

@@ -854,6 +854,7 @@ mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn,
 	struct mlx5dr_action *action;
 	bool reformat_req = false;
 	u32 num_of_ref = 0;
+	u32 ref_act_cnt;
 	int ret;
 	int i;
 
@@ -862,11 +863,14 @@ mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn,
 		return NULL;
 	}
 
-	hw_dests = kzalloc(sizeof(*hw_dests) * num_of_dests, GFP_KERNEL);
+	hw_dests = kcalloc(num_of_dests, sizeof(*hw_dests), GFP_KERNEL);
 	if (!hw_dests)
 		return NULL;
 
-	ref_actions = kzalloc(sizeof(*ref_actions) * num_of_dests * 2, GFP_KERNEL);
+	if (unlikely(check_mul_overflow(num_of_dests, 2u, &ref_act_cnt)))
+		goto free_hw_dests;
+
+	ref_actions = kcalloc(ref_act_cnt, sizeof(*ref_actions), GFP_KERNEL);
 	if (!ref_actions)
 		goto free_hw_dests;
 

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

@@ -421,19 +421,21 @@ int mlx5_query_nic_vport_system_image_guid(struct mlx5_core_dev *mdev,
 {
 	u32 *out;
 	int outlen = MLX5_ST_SZ_BYTES(query_nic_vport_context_out);
+	int err;
 
 	out = kvzalloc(outlen, GFP_KERNEL);
 	if (!out)
 		return -ENOMEM;
 
-	mlx5_query_nic_vport_context(mdev, 0, out);
+	err = mlx5_query_nic_vport_context(mdev, 0, out);
+	if (err)
+		goto out;
 
 	*system_image_guid = MLX5_GET64(query_nic_vport_context_out, out,
 					nic_vport_context.system_image_guid);
-
+out:
 	kvfree(out);
-
-	return 0;
+	return err;
 }
 EXPORT_SYMBOL_GPL(mlx5_query_nic_vport_system_image_guid);
 
@@ -1133,19 +1135,20 @@ EXPORT_SYMBOL_GPL(mlx5_nic_vport_unaffiliate_multiport);
 u64 mlx5_query_nic_system_image_guid(struct mlx5_core_dev *mdev)
 {
 	int port_type_cap = MLX5_CAP_GEN(mdev, port_type);
-	u64 tmp = 0;
+	u64 tmp;
+	int err;
 
 	if (mdev->sys_image_guid)
 		return mdev->sys_image_guid;
 
 	if (port_type_cap == MLX5_CAP_PORT_TYPE_ETH)
-		mlx5_query_nic_vport_system_image_guid(mdev, &tmp);
+		err = mlx5_query_nic_vport_system_image_guid(mdev, &tmp);
 	else
-		mlx5_query_hca_vport_system_image_guid(mdev, &tmp);
+		err = mlx5_query_hca_vport_system_image_guid(mdev, &tmp);
 
-	mdev->sys_image_guid = tmp;
+	mdev->sys_image_guid = err ? 0 : tmp;
 
-	return tmp;
+	return mdev->sys_image_guid;
 }
 EXPORT_SYMBOL_GPL(mlx5_query_nic_system_image_guid);
 

include/linux/mlx5/device.h

@@ -577,7 +577,9 @@ struct mlx5_init_seg {
 	__be32			rsvd1[120];
 	__be32			initializing;
 	struct health_buffer	health;
-	__be32			rsvd2[880];
+	__be32			rsvd2[878];
+	__be32			cmd_exec_to;
+	__be32			cmd_q_init_to;
 	__be32			internal_timer_h;
 	__be32			internal_timer_l;
 	__be32			rsvd3[2];

include/linux/mlx5/driver.h

@@ -66,10 +66,6 @@ enum {
 };
 
 enum {
-	/* one minute for the sake of bringup. Generally, commands must always
-	 * complete and we may need to increase this timeout value
-	 */
-	MLX5_CMD_TIMEOUT_MSEC	= 60 * 1000,
 	MLX5_CMD_WQ_MAX_NAME	= 32,
 };
 
@@ -156,6 +152,7 @@ enum {
 	MLX5_REG_MIRC		 = 0x9162,
 	MLX5_REG_SBCAM		 = 0xB01F,
 	MLX5_REG_RESOURCE_DUMP   = 0xC000,
+	MLX5_REG_DTOR            = 0xC00E,
 };
 
 enum mlx5_qpts_trust_state {
@@ -754,6 +751,7 @@ struct mlx5_core_dev {
 		u32 qcam[MLX5_ST_SZ_DW(qcam_reg)];
 		u8  embedded_cpu;
 	} caps;
+	struct mlx5_timeouts	*timeouts;
 	u64			sys_image_guid;
 	phys_addr_t		iseg_base;
 	struct mlx5_init_seg __iomem *iseg;
@@ -1245,6 +1243,16 @@ static inline int mlx5_core_native_port_num(struct mlx5_core_dev *dev)
 	return MLX5_CAP_GEN(dev, native_port_num);
 }
 
+static inline int mlx5_get_dev_index(struct mlx5_core_dev *dev)
+{
+	int idx = MLX5_CAP_GEN(dev, native_port_num);
+
+	if (idx >= 1 && idx <= MLX5_MAX_PORTS)
+		return idx - 1;
+	else
+		return PCI_FUNC(dev->pdev->devfn);
+}
+
 enum {
 	MLX5_TRIGGERED_CMD_COMP = (u64)1 << 32,
 };
@@ -1253,11 +1261,12 @@ static inline bool mlx5_is_roce_init_enabled(struct mlx5_core_dev *dev)
 {
 	struct devlink *devlink = priv_to_devlink(dev);
 	union devlink_param_value val;
+	int err;
 
-	devlink_param_driverinit_value_get(devlink,
+	err = devlink_param_driverinit_value_get(devlink,
 						 DEVLINK_PARAM_GENERIC_ID_ENABLE_ROCE,
 						 &val);
-	return val.vbool;
+	return err ? MLX5_CAP_GEN(dev, roce) : val.vbool;
 }
 
 #endif /* MLX5_DRIVER_H */

include/linux/mlx5/mlx5_ifc.h

@@ -1306,7 +1306,8 @@ struct mlx5_ifc_cmd_hca_cap_bits {
 	u8         vhca_resource_manager[0x1];
 
 	u8         hca_cap_2[0x1];
-	u8         reserved_at_21[0x2];
+	u8         reserved_at_21[0x1];
+	u8         dtor[0x1];
 	u8         event_on_vhca_state_teardown_request[0x1];
 	u8         event_on_vhca_state_in_use[0x1];
 	u8         event_on_vhca_state_active[0x1];
@@ -1587,7 +1588,8 @@ struct mlx5_ifc_cmd_hca_cap_bits {
 	u8         log_max_tis_per_sq[0x5];
 
 	u8         ext_stride_num_range[0x1];
-	u8         reserved_at_3a1[0x2];
+	u8         roce_rw_supported[0x1];
+	u8         reserved_at_3a2[0x1];
 	u8         log_max_stride_sz_rq[0x5];
 	u8         reserved_at_3a8[0x3];
 	u8         log_min_stride_sz_rq[0x5];
@@ -2807,6 +2809,40 @@ struct mlx5_ifc_dropped_packet_logged_bits {
 	u8         reserved_at_0[0xe0];
 };
 
+struct mlx5_ifc_default_timeout_bits {
+	u8         to_multiplier[0x3];
+	u8         reserved_at_3[0x9];
+	u8         to_value[0x14];
+};
+
+struct mlx5_ifc_dtor_reg_bits {
+	u8         reserved_at_0[0x20];
+
+	struct mlx5_ifc_default_timeout_bits pcie_toggle_to;
+
+	u8         reserved_at_40[0x60];
+
+	struct mlx5_ifc_default_timeout_bits health_poll_to;
+
+	struct mlx5_ifc_default_timeout_bits full_crdump_to;
+
+	struct mlx5_ifc_default_timeout_bits fw_reset_to;
+
+	struct mlx5_ifc_default_timeout_bits flush_on_err_to;
+
+	struct mlx5_ifc_default_timeout_bits pci_sync_update_to;
+
+	struct mlx5_ifc_default_timeout_bits tear_down_to;
+
+	struct mlx5_ifc_default_timeout_bits fsm_reactivate_to;
+
+	struct mlx5_ifc_default_timeout_bits reclaim_pages_to;
+
+	struct mlx5_ifc_default_timeout_bits reclaim_vfs_pages_to;
+
+	u8         reserved_at_1c0[0x40];
+};
+
 enum {
 	MLX5_CQ_ERROR_SYNDROME_CQ_OVERRUN                 = 0x1,
 	MLX5_CQ_ERROR_SYNDROME_CQ_ACCESS_VIOLATION_ERROR  = 0x2,