[pull request][net-next 00/15] mlx5 updates 2023-05-19

[pull request][net-next 00/15] mlx5 updates 2023-05-19

[Saeed Mahameed]

From: Saeed Mahameed <saeedm@nvidia.com>

This series provides misc updates to mlx5 driver.

For more information please see tag log below.

Please pull and let me know if there is any problem.

Thanks,
Saeed.


The following changes since commit 20d5e0ef252a151ea6585cfccf32def81a624666:

  net: arc: Make arc_emac_remove() return void (2023-05-19 13:33:28 +0100)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-updates-2023-05-19

for you to fetch changes up to f5d87b47a1d9dc14c048c84935397d97833ac706:

  net/mlx5e: E-Switch, Initialize E-Switch for eswitch manager (2023-05-19 10:50:31 -0700)

----------------------------------------------------------------
mlx5-updates-2023-05-19

mlx5 misc changes and code clean up:

The following series contains general changes for improving
E-Switch driver behavior.

1) improving condition checking
2) Code clean up
3) Using metadata matching on send-to-vport rules.
4) Using RoCE v2 instead of v1 for loopback rules.

----------------------------------------------------------------
Roi Dayan (15):
      net/mlx5: Remove redundant esw multiport validate function
      net/mlx5: E-Switch, Remove redundant check
      net/mlx5e: E-Switch, Remove flow_source check for metadata matching
      net/mlx5e: Remove redundant __func__ arg from fs_err() calls
      net/mlx5e: E-Switch, Update when to set other vport context
      net/mlx5e: E-Switch, Allow get vport api if esw exists
      net/mlx5e: E-Switch, Use metadata for vport matching in send-to-vport rules
      net/mlx5: Remove redundant vport_group_manager cap check
      net/mlx5e: E-Switch, Check device is PF when stopping esw offloads
      net/mlx5e: E-Switch: move debug print of adding mac to correct place
      net/mlx5e: E-Switch, Add a check that log_max_l2_table is valid
      net/mlx5: E-Switch, Use RoCE version 2 for loopback traffic
      net/mlx5: E-Switch, Use metadata matching for RoCE loopback rule
      net/mlx5: devlink, Only show PF related devlink warning when needed
      net/mlx5e: E-Switch, Initialize E-Switch for eswitch manager

 drivers/net/ethernet/mellanox/mlx5/core/devlink.c  |  26 +-----
 drivers/net/ethernet/mellanox/mlx5/core/en_fs.c    |  12 +--
 .../ethernet/mellanox/mlx5/core/esw/acl/helper.c   |   3 +-
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c  |  22 +++--
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.h  |   8 ++
 .../ethernet/mellanox/mlx5/core/eswitch_offloads.c | 103 ++++++++++++++-------
 drivers/net/ethernet/mellanox/mlx5/core/fw.c       |   5 +-
 drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c |   2 +-
 drivers/net/ethernet/mellanox/mlx5/core/rdma.c     |  24 ++---
 drivers/net/ethernet/mellanox/mlx5/core/sriov.c    |   4 -
 drivers/net/ethernet/mellanox/mlx5/core/vport.c    |   3 +-
 11 files changed, 108 insertions(+), 104 deletions(-)

[net-next 01/15] net/mlx5: Remove redundant esw multiport validate function

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

The function didn't validate the value and doesn't require value
validation as it will always be valid true or false values.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/devlink.c | 23 +------------------
 1 file changed, 1 insertion(+), 22 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 4b607785d694..0e07971e024a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -464,27 +464,6 @@ static int mlx5_devlink_esw_multiport_get(struct devlink *devlink, u32 id,
 	ctx->val.vbool = mlx5_lag_is_mpesw(dev);
 	return 0;
 }
-
-static int mlx5_devlink_esw_multiport_validate(struct devlink *devlink, u32 id,
-					       union devlink_param_value val,
-					       struct netlink_ext_ack *extack)
-{
-	struct mlx5_core_dev *dev = devlink_priv(devlink);
-
-	if (!MLX5_ESWITCH_MANAGER(dev)) {
-		NL_SET_ERR_MSG_MOD(extack, "E-Switch is unsupported");
-		return -EOPNOTSUPP;
-	}
-
-	if (mlx5_eswitch_mode(dev) != MLX5_ESWITCH_OFFLOADS) {
-		NL_SET_ERR_MSG_MOD(extack,
-				   "E-Switch must be in switchdev mode");
-		return -EBUSY;
-	}
-
-	return 0;
-}
-
 #endif
 
 static int mlx5_devlink_eq_depth_validate(struct devlink *devlink, u32 id,
@@ -563,7 +542,7 @@ static const struct devlink_param mlx5_devlink_params[] = {
 			     BIT(DEVLINK_PARAM_CMODE_RUNTIME),
 			     mlx5_devlink_esw_multiport_get,
 			     mlx5_devlink_esw_multiport_set,
-			     mlx5_devlink_esw_multiport_validate),
+			     NULL),
 #endif
 	DEVLINK_PARAM_GENERIC(IO_EQ_SIZE, BIT(DEVLINK_PARAM_CMODE_DRIVERINIT),
 			      NULL, NULL, mlx5_devlink_eq_depth_validate),
-- 
2.40.1

Re: [net-next 01/15] net/mlx5: Remove redundant esw multiport validate function

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (main)
by Saeed Mahameed <saeedm@nvidia.com>:

On Fri, 19 May 2023 10:55:43 -0700 you wrote:
> From: Roi Dayan <roid@nvidia.com>
> 
> The function didn't validate the value and doesn't require value
> validation as it will always be valid true or false values.
> 
> Signed-off-by: Roi Dayan <roid@nvidia.com>
> Reviewed-by: Maor Dickman <maord@nvidia.com>
> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
> 
> [...]

Here is the summary with links:
  - [net-next,01/15] net/mlx5: Remove redundant esw multiport validate function
    https://git.kernel.org/netdev/net-next/c/c511822fe2c9
  - [net-next,02/15] net/mlx5: E-Switch, Remove redundant check
    https://git.kernel.org/netdev/net-next/c/2abe501751ed
  - [net-next,03/15] net/mlx5e: E-Switch, Remove flow_source check for metadata matching
    https://git.kernel.org/netdev/net-next/c/edab80b89337
  - [net-next,04/15] net/mlx5e: Remove redundant __func__ arg from fs_err() calls
    https://git.kernel.org/netdev/net-next/c/806815bf3c1d
  - [net-next,05/15] net/mlx5e: E-Switch, Update when to set other vport context
    https://git.kernel.org/netdev/net-next/c/c97c9fe48ae3
  - [net-next,06/15] net/mlx5e: E-Switch, Allow get vport api if esw exists
    https://git.kernel.org/netdev/net-next/c/99db5669f663
  - [net-next,07/15] net/mlx5e: E-Switch, Use metadata for vport matching in send-to-vport rules
    https://git.kernel.org/netdev/net-next/c/29bcb6e4fe70
  - [net-next,08/15] net/mlx5: Remove redundant vport_group_manager cap check
    https://git.kernel.org/netdev/net-next/c/6cb9318a2534
  - [net-next,09/15] net/mlx5e: E-Switch, Check device is PF when stopping esw offloads
    https://git.kernel.org/netdev/net-next/c/bea416c7e970
  - [net-next,10/15] net/mlx5e: E-Switch: move debug print of adding mac to correct place
    https://git.kernel.org/netdev/net-next/c/292243d13b18
  - [net-next,11/15] net/mlx5e: E-Switch, Add a check that log_max_l2_table is valid
    https://git.kernel.org/netdev/net-next/c/3d7c5f78b8ce
  - [net-next,12/15] net/mlx5: E-Switch, Use RoCE version 2 for loopback traffic
    https://git.kernel.org/netdev/net-next/c/c24246d07a94
  - [net-next,13/15] net/mlx5: E-Switch, Use metadata matching for RoCE loopback rule
    https://git.kernel.org/netdev/net-next/c/7eb197fd83a3
  - [net-next,14/15] net/mlx5: devlink, Only show PF related devlink warning when needed
    https://git.kernel.org/netdev/net-next/c/0279b5454c0e
  - [net-next,15/15] net/mlx5e: E-Switch, Initialize E-Switch for eswitch manager
    https://git.kernel.org/netdev/net-next/c/f5d87b47a1d9

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html

[net-next 02/15] net/mlx5: E-Switch, Remove redundant check

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

The call to mlx5_eswitch_enable() also does the same check
and if E-Switch not supported it returns 0 without any change.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/sriov.c | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
index 20d7662c10fb..f07d00929162 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
@@ -74,9 +74,6 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs)
 	struct mlx5_core_sriov *sriov = &dev->priv.sriov;
 	int err, vf, num_msix_count;
 
-	if (!MLX5_ESWITCH_MANAGER(dev))
-		goto enable_vfs_hca;
-
 	err = mlx5_eswitch_enable(dev->priv.eswitch, num_vfs);
 	if (err) {
 		mlx5_core_warn(dev,
@@ -84,7 +81,6 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs)
 		return err;
 	}
 
-enable_vfs_hca:
 	num_msix_count = mlx5_get_default_msix_vec_count(dev, num_vfs);
 	for (vf = 0; vf < num_vfs; vf++) {
 		/* Notify the VF before its enablement to let it set
-- 
2.40.1

[net-next 03/15] net/mlx5e: E-Switch, Remove flow_source check for metadata matching

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

There is no reason to check for flow_source cap to allow metadata
matching. When flow_source match is being used the flow_source cap
is being checked.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 69215ffb9999..ecd12a0c6f07 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -2827,9 +2827,6 @@ bool mlx5_esw_vport_match_metadata_supported(const struct mlx5_eswitch *esw)
 	      MLX5_FDB_TO_VPORT_REG_C_0))
 		return false;
 
-	if (!MLX5_CAP_ESW_FLOWTABLE(esw->dev, flow_source))
-		return false;
-
 	return true;
 }
 
-- 
2.40.1

[net-next 04/15] net/mlx5e: Remove redundant __func__ arg from fs_err() calls

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

fs_err() already logs the function name. remote the arg so the function
name will not be logged twice.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index 33bfe4d7338b..934b0d5ce1b3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -283,7 +283,7 @@ static int __mlx5e_add_vlan_rule(struct mlx5e_flow_steering *fs,
 	if (IS_ERR(*rule_p)) {
 		err = PTR_ERR(*rule_p);
 		*rule_p = NULL;
-		fs_err(fs, "%s: add rule failed\n", __func__);
+		fs_err(fs, "add rule failed\n");
 	}
 
 	return err;
@@ -395,8 +395,7 @@ int mlx5e_add_vlan_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num
 	if (IS_ERR(rule)) {
 		err = PTR_ERR(rule);
 		fs->vlan->trap_rule = NULL;
-		fs_err(fs, "%s: add VLAN trap rule failed, err %d\n",
-		       __func__, err);
+		fs_err(fs, "add VLAN trap rule failed, err %d\n", err);
 		return err;
 	}
 	fs->vlan->trap_rule = rule;
@@ -421,8 +420,7 @@ int mlx5e_add_mac_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num)
 	if (IS_ERR(rule)) {
 		err = PTR_ERR(rule);
 		fs->l2.trap_rule = NULL;
-		fs_err(fs, "%s: add MAC trap rule failed, err %d\n",
-		       __func__, err);
+		fs_err(fs, "add MAC trap rule failed, err %d\n", err);
 		return err;
 	}
 	fs->l2.trap_rule = rule;
@@ -763,7 +761,7 @@ static int mlx5e_add_promisc_rule(struct mlx5e_flow_steering *fs)
 	if (IS_ERR(*rule_p)) {
 		err = PTR_ERR(*rule_p);
 		*rule_p = NULL;
-		fs_err(fs, "%s: add promiscuous rule failed\n", __func__);
+		fs_err(fs, "add promiscuous rule failed\n");
 	}
 	kvfree(spec);
 	return err;
@@ -995,7 +993,7 @@ static int mlx5e_add_l2_flow_rule(struct mlx5e_flow_steering *fs,
 
 	ai->rule = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1);
 	if (IS_ERR(ai->rule)) {
-		fs_err(fs, "%s: add l2 rule(mac:%pM) failed\n", __func__, mv_dmac);
+		fs_err(fs, "add l2 rule(mac:%pM) failed\n", mv_dmac);
 		err = PTR_ERR(ai->rule);
 		ai->rule = NULL;
 	}
-- 
2.40.1

[net-next 05/15] net/mlx5e: E-Switch, Update when to set other vport context

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Other vport context should be set if vport number is not 0.
In case of ECPF, vport 0 represents the host PF representor so also
need to set other vport context.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/esw/acl/helper.c | 3 ++-
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c        | 3 ++-
 drivers/net/ethernet/mellanox/mlx5/core/vport.c          | 3 ++-
 3 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/helper.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/helper.c
index 45b839116212..d599e50af346 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/helper.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/acl/helper.c
@@ -35,7 +35,8 @@ esw_acl_table_create(struct mlx5_eswitch *esw, struct mlx5_vport *vport, int ns,
 	}
 
 	ft_attr.max_fte = size;
-	ft_attr.flags = MLX5_FLOW_TABLE_OTHER_VPORT;
+	if (vport_num || mlx5_core_is_ecpf(esw->dev))
+		ft_attr.flags = MLX5_FLOW_TABLE_OTHER_VPORT;
 	acl = mlx5_create_vport_flow_table(root_ns, &ft_attr, vport_num);
 	if (IS_ERR(acl)) {
 		err = PTR_ERR(acl);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 901c53751b0a..bf97a593d1d4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -113,7 +113,8 @@ static int arm_vport_context_events_cmd(struct mlx5_core_dev *dev, u16 vport,
 		 opcode, MLX5_CMD_OP_MODIFY_NIC_VPORT_CONTEXT);
 	MLX5_SET(modify_nic_vport_context_in, in, field_select.change_event, 1);
 	MLX5_SET(modify_nic_vport_context_in, in, vport_number, vport);
-	MLX5_SET(modify_nic_vport_context_in, in, other_vport, 1);
+	if (vport || mlx5_core_is_ecpf(dev))
+		MLX5_SET(modify_nic_vport_context_in, in, other_vport, 1);
 	nic_vport_ctx = MLX5_ADDR_OF(modify_nic_vport_context_in,
 				     in, nic_vport_context);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/vport.c b/drivers/net/ethernet/mellanox/mlx5/core/vport.c
index ba7e3df22413..bc66b078a8a1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/vport.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/vport.c
@@ -288,7 +288,8 @@ int mlx5_query_nic_vport_mac_list(struct mlx5_core_dev *dev,
 		 MLX5_CMD_OP_QUERY_NIC_VPORT_CONTEXT);
 	MLX5_SET(query_nic_vport_context_in, in, allowed_list_type, list_type);
 	MLX5_SET(query_nic_vport_context_in, in, vport_number, vport);
-	MLX5_SET(query_nic_vport_context_in, in, other_vport, 1);
+	if (vport || mlx5_core_is_ecpf(dev))
+		MLX5_SET(query_nic_vport_context_in, in, other_vport, 1);
 
 	err = mlx5_cmd_exec(dev, in, sizeof(in), out, out_sz);
 	if (err)
-- 
2.40.1

[net-next 06/15] net/mlx5e: E-Switch, Allow get vport api if esw exists

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

We could have an esw manager device which is not a vport group manager.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index bf97a593d1d4..692cea3a6383 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -92,7 +92,7 @@ mlx5_eswitch_get_vport(struct mlx5_eswitch *esw, u16 vport_num)
 {
 	struct mlx5_vport *vport;
 
-	if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager))
+	if (!esw)
 		return ERR_PTR(-EPERM);
 
 	vport = xa_load(&esw->vports, vport_num);
-- 
2.40.1

[net-next 07/15] net/mlx5e: E-Switch, Use metadata for vport matching in send-to-vport rules

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Like other rules use metadata matching if supported instead of
source_port.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../mellanox/mlx5/core/eswitch_offloads.c     | 62 +++++++++++++------
 1 file changed, 43 insertions(+), 19 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index ecd12a0c6f07..66a522d08be1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -838,6 +838,7 @@ mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *on_esw,
 	struct mlx5_flow_handle *flow_rule;
 	struct mlx5_flow_spec *spec;
 	void *misc;
+	u16 vport;
 
 	spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
 	if (!spec) {
@@ -847,20 +848,43 @@ mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *on_esw,
 
 	misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters);
 	MLX5_SET(fte_match_set_misc, misc, source_sqn, sqn);
-	/* source vport is the esw manager */
-	MLX5_SET(fte_match_set_misc, misc, source_port, from_esw->manager_vport);
-	if (MLX5_CAP_ESW(on_esw->dev, merged_eswitch))
-		MLX5_SET(fte_match_set_misc, misc, source_eswitch_owner_vhca_id,
-			 MLX5_CAP_GEN(from_esw->dev, vhca_id));
 
 	misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters);
 	MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_sqn);
-	MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
-	if (MLX5_CAP_ESW(on_esw->dev, merged_eswitch))
-		MLX5_SET_TO_ONES(fte_match_set_misc, misc,
-				 source_eswitch_owner_vhca_id);
 
 	spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
+
+	/* source vport is the esw manager */
+	vport = from_esw->manager_vport;
+
+	if (mlx5_eswitch_vport_match_metadata_enabled(on_esw)) {
+		misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters_2);
+		MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0,
+			 mlx5_eswitch_get_vport_metadata_for_match(from_esw, vport));
+
+		misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters_2);
+		MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0,
+			 mlx5_eswitch_get_vport_metadata_mask());
+
+		spec->match_criteria_enable |= MLX5_MATCH_MISC_PARAMETERS_2;
+	} else {
+		misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters);
+		MLX5_SET(fte_match_set_misc, misc, source_port, vport);
+
+		if (MLX5_CAP_ESW(on_esw->dev, merged_eswitch))
+			MLX5_SET(fte_match_set_misc, misc, source_eswitch_owner_vhca_id,
+				 MLX5_CAP_GEN(from_esw->dev, vhca_id));
+
+		misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters);
+		MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
+
+		if (MLX5_CAP_ESW(on_esw->dev, merged_eswitch))
+			MLX5_SET_TO_ONES(fte_match_set_misc, misc,
+					 source_eswitch_owner_vhca_id);
+
+		spec->match_criteria_enable |= MLX5_MATCH_MISC_PARAMETERS;
+	}
+
 	dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT;
 	dest.vport.num = rep->vport;
 	dest.vport.vhca_id = MLX5_CAP_GEN(rep->esw->dev, vhca_id);
@@ -1270,7 +1294,8 @@ esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag)
 #define MAX_SQ_NVPORTS 32
 
 static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
-					   u32 *flow_group_in)
+					   u32 *flow_group_in,
+					   int match_params)
 {
 	void *match_criteria = MLX5_ADDR_OF(create_flow_group_in,
 					    flow_group_in,
@@ -1279,7 +1304,7 @@ static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
 	if (mlx5_eswitch_vport_match_metadata_enabled(esw)) {
 		MLX5_SET(create_flow_group_in, flow_group_in,
 			 match_criteria_enable,
-			 MLX5_MATCH_MISC_PARAMETERS_2);
+			 MLX5_MATCH_MISC_PARAMETERS_2 | match_params);
 
 		MLX5_SET(fte_match_param, match_criteria,
 			 misc_parameters_2.metadata_reg_c_0,
@@ -1287,7 +1312,7 @@ static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
 	} else {
 		MLX5_SET(create_flow_group_in, flow_group_in,
 			 match_criteria_enable,
-			 MLX5_MATCH_MISC_PARAMETERS);
+			 MLX5_MATCH_MISC_PARAMETERS | match_params);
 
 		MLX5_SET_TO_ONES(fte_match_param, match_criteria,
 				 misc_parameters.source_port);
@@ -1463,14 +1488,13 @@ esw_create_send_to_vport_group(struct mlx5_eswitch *esw,
 
 	memset(flow_group_in, 0, inlen);
 
-	MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable,
-		 MLX5_MATCH_MISC_PARAMETERS);
+	esw_set_flow_group_source_port(esw, flow_group_in, MLX5_MATCH_MISC_PARAMETERS);
 
 	match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria);
-
 	MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_sqn);
-	MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_port);
-	if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) {
+
+	if (!mlx5_eswitch_vport_match_metadata_enabled(esw) &&
+	    MLX5_CAP_ESW(esw->dev, merged_eswitch)) {
 		MLX5_SET_TO_ONES(fte_match_param, match_criteria,
 				 misc_parameters.source_eswitch_owner_vhca_id);
 		MLX5_SET(create_flow_group_in, flow_group_in,
@@ -1558,7 +1582,7 @@ esw_create_peer_esw_miss_group(struct mlx5_eswitch *esw,
 
 	memset(flow_group_in, 0, inlen);
 
-	esw_set_flow_group_source_port(esw, flow_group_in);
+	esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	if (!mlx5_eswitch_vport_match_metadata_enabled(esw)) {
 		match_criteria = MLX5_ADDR_OF(create_flow_group_in,
@@ -1845,7 +1869,7 @@ static int esw_create_vport_rx_group(struct mlx5_eswitch *esw)
 		return -ENOMEM;
 
 	/* create vport rx group */
-	esw_set_flow_group_source_port(esw, flow_group_in);
+	esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0);
 	MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, nvports - 1);
-- 
2.40.1

[net-next 08/15] net/mlx5: Remove redundant vport_group_manager cap check

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

It's enough to check for esw_manager cap for get the esw flow table
caps.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fw.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
index 7bb7be01225a..fb2035a5ec99 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
@@ -196,14 +196,11 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev)
 			return err;
 	}
 
-	if (MLX5_CAP_GEN(dev, vport_group_manager) &&
-	    MLX5_ESWITCH_MANAGER(dev)) {
+	if (MLX5_ESWITCH_MANAGER(dev)) {
 		err = mlx5_core_get_caps(dev, MLX5_CAP_ESWITCH_FLOW_TABLE);
 		if (err)
 			return err;
-	}
 
-	if (MLX5_ESWITCH_MANAGER(dev)) {
 		err = mlx5_core_get_caps(dev, MLX5_CAP_ESWITCH);
 		if (err)
 			return err;
-- 
2.40.1

[net-next 09/15] net/mlx5e: E-Switch, Check device is PF when stopping esw offloads

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Checking sriov is done on the pci device so it can return true
on other devices like SF but nothing should be done in this case.
Add a check that the device is PF.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 66a522d08be1..44b5d1359155 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -3301,7 +3301,7 @@ static int esw_offloads_stop(struct mlx5_eswitch *esw,
 	/* If changing from switchdev to legacy mode without sriov enabled,
 	 * no need to create legacy fdb.
 	 */
-	if (!mlx5_sriov_is_enabled(esw->dev))
+	if (!mlx5_core_is_pf(esw->dev) || !mlx5_sriov_is_enabled(esw->dev))
 		return 0;
 
 	err = mlx5_eswitch_enable_locked(esw, MLX5_ESWITCH_IGNORE_NUM_VFS);
-- 
2.40.1

[net-next 10/15] net/mlx5e: E-Switch: move debug print of adding mac to correct place

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Move the debug print inside the if clause that actually does the change.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 692cea3a6383..3e3963424285 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -310,11 +310,12 @@ static int esw_add_uc_addr(struct mlx5_eswitch *esw, struct vport_addr *vaddr)
 
 fdb_add:
 	/* SRIOV is enabled: Forward UC MAC to vport */
-	if (esw->fdb_table.legacy.fdb && esw->mode == MLX5_ESWITCH_LEGACY)
+	if (esw->fdb_table.legacy.fdb && esw->mode == MLX5_ESWITCH_LEGACY) {
 		vaddr->flow_rule = esw_fdb_set_vport_rule(esw, mac, vport);
 
-	esw_debug(esw->dev, "\tADDED UC MAC: vport[%d] %pM fr(%p)\n",
-		  vport, mac, vaddr->flow_rule);
+		esw_debug(esw->dev, "\tADDED UC MAC: vport[%d] %pM fr(%p)\n",
+			  vport, mac, vaddr->flow_rule);
+	}
 
 	return 0;
 }
-- 
2.40.1

[net-next 11/15] net/mlx5e: E-Switch, Add a check that log_max_l2_table is valid

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

If log_max_l2_table is 0 there is no really room for one L2 address.
and should be treated as not supported.
Do the check in MPFS init and for vport context events which
both used to update L2 address.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c  | 6 +++++-
 drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c | 2 +-
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 3e3963424285..9b71819e049a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -712,6 +712,9 @@ void esw_vport_change_handle_locked(struct mlx5_vport *vport)
 	struct mlx5_eswitch *esw = dev->priv.eswitch;
 	u8 mac[ETH_ALEN];
 
+	if (!MLX5_CAP_GEN(dev, log_max_l2_table))
+		return;
+
 	mlx5_query_nic_vport_mac_address(dev, vport->vport, true, mac);
 	esw_debug(dev, "vport[%d] Context Changed: perm mac: %pM\n",
 		  vport->vport, mac);
@@ -948,7 +951,8 @@ void mlx5_esw_vport_disable(struct mlx5_eswitch *esw, u16 vport_num)
 	vport->enabled = false;
 
 	/* Disable events from this vport */
-	arm_vport_context_events_cmd(esw->dev, vport->vport, 0);
+	if (MLX5_CAP_GEN(esw->dev, log_max_l2_table))
+		arm_vport_context_events_cmd(esw->dev, vport->vport, 0);
 
 	if (!mlx5_esw_is_manager_vport(esw, vport->vport) &&
 	    MLX5_CAP_GEN(esw->dev, vhca_resource_manager))
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c
index 8ff16318e32d..4450091e181a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/mpfs.c
@@ -99,7 +99,7 @@ int mlx5_mpfs_init(struct mlx5_core_dev *dev)
 	int l2table_size = 1 << MLX5_CAP_GEN(dev, log_max_l2_table);
 	struct mlx5_mpfs *mpfs;
 
-	if (!MLX5_ESWITCH_MANAGER(dev))
+	if (!MLX5_ESWITCH_MANAGER(dev) || l2table_size == 1)
 		return 0;
 
 	mpfs = kzalloc(sizeof(*mpfs), GFP_KERNEL);
-- 
2.40.1

[net-next 12/15] net/mlx5: E-Switch, Use RoCE version 2 for loopback traffic

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Could be port initializing eswitch doesn't support RoCE version 1
but all ports should support RoCE version 2.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/rdma.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/rdma.c b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
index 540cf05f6373..15bb562b3846 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
@@ -115,7 +115,7 @@ static int mlx5_rdma_enable_roce_steering(struct mlx5_core_dev *dev)
 
 static void mlx5_rdma_del_roce_addr(struct mlx5_core_dev *dev)
 {
-	mlx5_core_roce_gid_set(dev, 0, 0, 0,
+	mlx5_core_roce_gid_set(dev, 0, MLX5_ROCE_VERSION_2, 0,
 			       NULL, NULL, false, 0, 1);
 }
 
@@ -135,7 +135,7 @@ static int mlx5_rdma_add_roce_addr(struct mlx5_core_dev *dev)
 
 	mlx5_rdma_make_default_gid(dev, &gid);
 	return mlx5_core_roce_gid_set(dev, 0,
-				      MLX5_ROCE_VERSION_1,
+				      MLX5_ROCE_VERSION_2,
 				      0, gid.raw, mac,
 				      false, 0, 1);
 }
-- 
2.40.1

[net-next 13/15] net/mlx5: E-Switch, Use metadata matching for RoCE loopback rule

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Use metadata matching for RoCE loopback rule if device is configured
to use metadata for source port matching.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 .../net/ethernet/mellanox/mlx5/core/eswitch.h |  8 ++++
 .../mellanox/mlx5/core/eswitch_offloads.c     | 46 +++++++++++--------
 .../net/ethernet/mellanox/mlx5/core/rdma.c    | 20 ++------
 3 files changed, 40 insertions(+), 34 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index 1a042c981713..6c5c05c414fb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -683,6 +683,14 @@ mlx5_esw_vporttbl_put(struct mlx5_eswitch *esw, struct mlx5_vport_tbl_attr *attr
 struct mlx5_flow_handle *
 esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag);
 
+void mlx5_esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
+					 u32 *flow_group_in,
+					 int match_params);
+
+void mlx5_esw_set_spec_source_port(struct mlx5_eswitch *esw,
+				   u16 vport,
+				   struct mlx5_flow_spec *spec);
+
 int esw_offloads_load_rep(struct mlx5_eswitch *esw, u16 vport_num);
 void esw_offloads_unload_rep(struct mlx5_eswitch *esw, u16 vport_num);
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 44b5d1359155..76db72f339d6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -1293,9 +1293,10 @@ esw_add_restore_rule(struct mlx5_eswitch *esw, u32 tag)
 #define MAX_PF_SQ 256
 #define MAX_SQ_NVPORTS 32
 
-static void esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
-					   u32 *flow_group_in,
-					   int match_params)
+void
+mlx5_esw_set_flow_group_source_port(struct mlx5_eswitch *esw,
+				    u32 *flow_group_in,
+				    int match_params)
 {
 	void *match_criteria = MLX5_ADDR_OF(create_flow_group_in,
 					    flow_group_in,
@@ -1488,7 +1489,7 @@ esw_create_send_to_vport_group(struct mlx5_eswitch *esw,
 
 	memset(flow_group_in, 0, inlen);
 
-	esw_set_flow_group_source_port(esw, flow_group_in, MLX5_MATCH_MISC_PARAMETERS);
+	mlx5_esw_set_flow_group_source_port(esw, flow_group_in, MLX5_MATCH_MISC_PARAMETERS);
 
 	match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria);
 	MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_sqn);
@@ -1582,7 +1583,7 @@ esw_create_peer_esw_miss_group(struct mlx5_eswitch *esw,
 
 	memset(flow_group_in, 0, inlen);
 
-	esw_set_flow_group_source_port(esw, flow_group_in, 0);
+	mlx5_esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	if (!mlx5_eswitch_vport_match_metadata_enabled(esw)) {
 		match_criteria = MLX5_ADDR_OF(create_flow_group_in,
@@ -1869,7 +1870,7 @@ static int esw_create_vport_rx_group(struct mlx5_eswitch *esw)
 		return -ENOMEM;
 
 	/* create vport rx group */
-	esw_set_flow_group_source_port(esw, flow_group_in, 0);
+	mlx5_esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0);
 	MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, nvports - 1);
@@ -1939,21 +1940,13 @@ static void esw_destroy_vport_rx_drop_group(struct mlx5_eswitch *esw)
 		mlx5_destroy_flow_group(esw->offloads.vport_rx_drop_group);
 }
 
-struct mlx5_flow_handle *
-mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport,
-				  struct mlx5_flow_destination *dest)
+void
+mlx5_esw_set_spec_source_port(struct mlx5_eswitch *esw,
+			      u16 vport,
+			      struct mlx5_flow_spec *spec)
 {
-	struct mlx5_flow_act flow_act = {0};
-	struct mlx5_flow_handle *flow_rule;
-	struct mlx5_flow_spec *spec;
 	void *misc;
 
-	spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
-	if (!spec) {
-		flow_rule = ERR_PTR(-ENOMEM);
-		goto out;
-	}
-
 	if (mlx5_eswitch_vport_match_metadata_enabled(esw)) {
 		misc = MLX5_ADDR_OF(fte_match_param, spec->match_value, misc_parameters_2);
 		MLX5_SET(fte_match_set_misc2, misc, metadata_reg_c_0,
@@ -1973,6 +1966,23 @@ mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport,
 
 		spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
 	}
+}
+
+struct mlx5_flow_handle *
+mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport,
+				  struct mlx5_flow_destination *dest)
+{
+	struct mlx5_flow_act flow_act = {0};
+	struct mlx5_flow_handle *flow_rule;
+	struct mlx5_flow_spec *spec;
+
+	spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+	if (!spec) {
+		flow_rule = ERR_PTR(-ENOMEM);
+		goto out;
+	}
+
+	mlx5_esw_set_spec_source_port(esw, vport, spec);
 
 	flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
 	flow_rule = mlx5_add_flow_rules(esw->offloads.ft_offloads, spec,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/rdma.c b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
index 15bb562b3846..a42f6cd99b74 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/rdma.c
@@ -30,9 +30,8 @@ static int mlx5_rdma_enable_roce_steering(struct mlx5_core_dev *dev)
 	struct mlx5_flow_spec *spec;
 	struct mlx5_flow_table *ft;
 	struct mlx5_flow_group *fg;
-	void *match_criteria;
+	struct mlx5_eswitch *esw;
 	u32 *flow_group_in;
-	void *misc;
 	int err;
 
 	if (!(MLX5_CAP_FLOWTABLE_RDMA_RX(dev, ft_support) &&
@@ -63,12 +62,8 @@ static int mlx5_rdma_enable_roce_steering(struct mlx5_core_dev *dev)
 		goto free;
 	}
 
-	MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable,
-		 MLX5_MATCH_MISC_PARAMETERS);
-	match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in,
-				      match_criteria);
-	MLX5_SET_TO_ONES(fte_match_param, match_criteria,
-			 misc_parameters.source_port);
+	esw = dev->priv.eswitch;
+	mlx5_esw_set_flow_group_source_port(esw, flow_group_in, 0);
 
 	fg = mlx5_create_flow_group(ft, flow_group_in);
 	if (IS_ERR(fg)) {
@@ -77,14 +72,7 @@ static int mlx5_rdma_enable_roce_steering(struct mlx5_core_dev *dev)
 		goto destroy_flow_table;
 	}
 
-	spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
-	misc = MLX5_ADDR_OF(fte_match_param, spec->match_value,
-			    misc_parameters);
-	MLX5_SET(fte_match_set_misc, misc, source_port,
-		 dev->priv.eswitch->manager_vport);
-	misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria,
-			    misc_parameters);
-	MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
+	mlx5_esw_set_spec_source_port(esw, esw->manager_vport, spec);
 
 	flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW;
 	flow_rule = mlx5_add_flow_rules(ft, spec, &flow_act, NULL, 0);
-- 
2.40.1

[net-next 14/15] net/mlx5: devlink, Only show PF related devlink warning when needed

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Limit the PF related warning to show if device is actually a PF.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/devlink.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 0e07971e024a..bfaec67abf0d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -162,9 +162,8 @@ static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 		return -EOPNOTSUPP;
 	}
 
-	if (pci_num_vf(pdev)) {
+	if (mlx5_core_is_pf(dev) && pci_num_vf(pdev))
 		NL_SET_ERR_MSG_MOD(extack, "reload while VFs are present is unfavorable");
-	}
 
 	switch (action) {
 	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
-- 
2.40.1

[net-next 15/15] net/mlx5e: E-Switch, Initialize E-Switch for eswitch manager

[Saeed Mahameed]

From: Roi Dayan <roid@nvidia.com>

Initialize eswitch instance for a function which is eswitch manager
but not a vport group manager.

Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index 9b71819e049a..31956cd9d1bb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -1622,7 +1622,7 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev)
 	struct mlx5_eswitch *esw;
 	int err;
 
-	if (!MLX5_VPORT_MANAGER(dev))
+	if (!MLX5_VPORT_MANAGER(dev) && !MLX5_ESWITCH_MANAGER(dev))
 		return 0;
 
 	esw = kzalloc(sizeof(*esw), GFP_KERNEL);
@@ -1692,7 +1692,7 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev)
 
 void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw)
 {
-	if (!esw || !MLX5_VPORT_MANAGER(esw->dev))
+	if (!esw)
 		return;
 
 	esw_info(esw->dev, "cleanup\n");
-- 
2.40.1

Re: [pull request][net-next 00/15] mlx5 updates 2023-05-19

[Jakub Kicinski]

On Fri, 19 May 2023 10:55:42 -0700 Saeed Mahameed wrote:
> mlx5 misc changes and code clean up:
> 
> The following series contains general changes for improving
> E-Switch driver behavior.
> 
> 1) improving condition checking
> 2) Code clean up
> 3) Using metadata matching on send-to-vport rules.
> 4) Using RoCE v2 instead of v1 for loopback rules.

Acked-by: Jakub Kicinski <kuba@kernel.org>

FWIW..