[pull request][net-next 00/16] Mellanox, mlx5e updates 2018-05-14

[pull request][net-next 00/16] Mellanox, mlx5e updates 2018-05-14

[Saeed Mahameed]

Hi Dave,

The following series contains misc updates for mlx5e netdevice driver.
For more information please see tag log below.

Please pull and let me know if there's any problem.

Thanks,
Saeed.

---

The following changes since commit 7cfac881660ac7b7229950e0a942201be63c15d1:

  cxgb4: do not fail vf instatiation in slave mode (2018-05-14 16:42:09 -0400)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5e-updates-2018-05-14

for you to fetch changes up to 0e5c04f6b52b248cfdb7e791fd5702f12270df7b:

  net/mlx5e: Remove MLX5E_TEST_BIT macro (2018-05-14 15:10:21 -0700)

----------------------------------------------------------------
mlx5e-updates-2018-05-14

Misc update for mlx5e netdevice driver

>From Gal Pressman:
 - Remove MLX5E_TEST_BIT macros and use test_bit instead
 - Use __set_bit when possible

>From Eran Ben Elisha:
  - Improve debug print on initial RX posting timeout

>From Or Gerlitz:
 - Support offloaded TC flows with no matches on headers
 - mlx5e TC cleanups

Trivial cleanups From Roi, Tariq and Saeed:
  - Use bool as return type for mlx5e_xdp_handle
  - Use u8 instead of int for LRO number of segments
  - Skip redundant checks when providing NUD lastuse feedback
  - Remove redundant vport context vlan update

----------------------------------------------------------------
Eran Ben Elisha (1):
      net/mlx5e: Report all channels with min RX WQEs timeout

Gal Pressman (3):
      net/mlx5e: Use __set_bit for adaptive-moderation bit in RQ state
      net/mlx5e: Use test bit in en accel xmit flow
      net/mlx5e: Remove MLX5E_TEST_BIT macro

Or Gerlitz (8):
      net/mlx5e: Remove double defined DMAC header re-write element
      net/mlx5e: Clean static checker complaints on TC offload and VF reps code
      net/mlx5e: Avoid redundant zeroing of offloaded TC flow attributes
      net/mlx5e: Return success when TC offloaded fdb actions parsed ok
      net/mlx5e: Use local actions var while processing offloaded TC flow actions
      net/mlx5e: Properly order min inline mode setup while parsing TC matches
      net/mlx5e: Get the required HW match level while parsing TC flow matches
      net/mlx5e: Support offloaded TC flows with no matches on headers

Roi Dayan (1):
      net/mlx5e: Skip redundant checks when providing NUD lastuse feedback

Saeed Mahameed (1):
      net/mlx5e: Remove redundant vport context vlan update

Tariq Toukan (2):
      net/mlx5e: Use u8 instead of int for LRO number of segments
      net/mlx5e: Use bool as return type for mlx5e_xdp_handle

 drivers/net/ethernet/mellanox/mlx5/core/en.h       |   2 -
 .../mellanox/mlx5/core/en_accel/en_accel.h         |   4 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_fs.c    |   1 -
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c  |  25 ++-
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c   |   8 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c    |  19 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c    | 224 ++++++++++++---------
 drivers/net/ethernet/mellanox/mlx5/core/en_tx.c    |   2 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c  |   4 +-
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.h  |   8 +
 .../ethernet/mellanox/mlx5/core/eswitch_offloads.c |   8 +-
 11 files changed, 173 insertions(+), 132 deletions(-)

[net-next 01/16] net/mlx5e: Remove redundant vport context vlan update

[Saeed Mahameed]

In delete vlan flow an extra call to mlx5e_vport_context_update_vlans
was added by mistake, remove it.

Fixes: 86d722ad2c3b ("net/mlx5: Use flow steering infrastructure for mlx5_en")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Gal Pressman <galp@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index f64dda2bed31..76cc10e44080 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -277,7 +277,6 @@ static void mlx5e_del_vlan_rule(struct mlx5e_priv *priv,
 		}
 		break;
 	case MLX5E_VLAN_RULE_TYPE_MATCH_CTAG_VID:
-		mlx5e_vport_context_update_vlans(priv);
 		if (priv->fs.vlan.active_cvlans_rule[vid]) {
 			mlx5_del_flow_rules(priv->fs.vlan.active_cvlans_rule[vid]);
 			priv->fs.vlan.active_cvlans_rule[vid] = NULL;
-- 
2.17.0

[net-next 02/16] net/mlx5e: Skip redundant checks when providing NUD lastuse feedback

[Saeed Mahameed]

From: Roi Dayan <roid@mellanox.com>

It's redundant to continue the loop if we found one flow whose lastuse value
being newer than the last one we reported, since this is enough for us to
trigger a NUD update (neigh_event_send()).

Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index b94276db3ce9..9b0e3ccffa66 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -982,6 +982,8 @@ void mlx5e_tc_update_neigh_used_value(struct mlx5e_neigh_hash_entry *nhe)
 				}
 			}
 		}
+		if (neigh_used)
+			break;
 	}
 
 	if (neigh_used) {
-- 
2.17.0

[net-next 03/16] net/mlx5e: Use u8 instead of int for LRO number of segments

[Saeed Mahameed]

From: Tariq Toukan <tariqt@mellanox.com>

Range of LRO number of segments fits in u8.
Also, bring initialization and declaration together to
save code.

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index 7bbf0db27a01..e10422b845ca 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -681,11 +681,10 @@ static inline void mlx5e_build_rx_skb(struct mlx5_cqe64 *cqe,
 				      struct mlx5e_rq *rq,
 				      struct sk_buff *skb)
 {
+	u8 lro_num_seg = be32_to_cpu(cqe->srqn) >> 24;
 	struct net_device *netdev = rq->netdev;
-	int lro_num_seg;
 
 	skb->mac_len = ETH_HLEN;
-	lro_num_seg = be32_to_cpu(cqe->srqn) >> 24;
 	if (lro_num_seg > 1) {
 		mlx5e_lro_update_hdr(skb, cqe, cqe_bcnt);
 		skb_shinfo(skb)->gso_size = DIV_ROUND_UP(cqe_bcnt, lro_num_seg);
-- 
2.17.0

[net-next 04/16] net/mlx5e: Use bool as return type for mlx5e_xdp_handle

[Saeed Mahameed]

From: Tariq Toukan <tariqt@mellanox.com>

Function returns boolean values, use bool instead of int.

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index e10422b845ca..abc6ca8168d2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -807,9 +807,9 @@ static inline bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
 }
 
 /* returns true if packet was consumed by xdp */
-static inline int mlx5e_xdp_handle(struct mlx5e_rq *rq,
-				   struct mlx5e_dma_info *di,
-				   void *va, u16 *rx_headroom, u32 *len)
+static inline bool mlx5e_xdp_handle(struct mlx5e_rq *rq,
+				    struct mlx5e_dma_info *di,
+				    void *va, u16 *rx_headroom, u32 *len)
 {
 	struct bpf_prog *prog = READ_ONCE(rq->xdp_prog);
 	struct xdp_buff xdp;
-- 
2.17.0

[net-next 05/16] net/mlx5e: Remove double defined DMAC header re-write element

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

The firmware DMAC_47_16 header re-write token was defined twice,
clean it up.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reported-by: Jianbo Liu <jianbol@mellanox.com>
Reviewed-by: Jianbo Liu <jianbol@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 9b0e3ccffa66..969a1768da71 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1579,7 +1579,6 @@ struct mlx5_fields {
 		{MLX5_ACTION_IN_FIELD_OUT_ ## fw_field, size, offsetof(struct pedit_headers, field) + (off)}
 
 static struct mlx5_fields fields[] = {
-	OFFLOAD(DMAC_47_16, 4, eth.h_dest[0], 0),
 	OFFLOAD(DMAC_47_16, 4, eth.h_dest[0], 0),
 	OFFLOAD(DMAC_15_0,  2, eth.h_dest[4], 0),
 	OFFLOAD(SMAC_47_16, 4, eth.h_source[0], 0),
-- 
2.17.0

[net-next 06/16] net/mlx5e: Clean static checker complaints on TC offload and VF reps code

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

Clean warning/check complaints made by checkpatch on en_{tc,rep}.c

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Jianbo Liu <jianbol@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c |  8 ++++----
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c  | 11 +++++------
 2 files changed, 9 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index 876c3e4c6193..a689f4c90fe3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -681,8 +681,8 @@ static int mlx5e_rep_open(struct net_device *dev)
 		goto unlock;
 
 	if (!mlx5_modify_vport_admin_state(priv->mdev,
-			MLX5_QUERY_VPORT_STATE_IN_OP_MOD_ESW_VPORT,
-			rep->vport, MLX5_ESW_VPORT_ADMIN_STATE_UP))
+					   MLX5_QUERY_VPORT_STATE_IN_OP_MOD_ESW_VPORT,
+					   rep->vport, MLX5_ESW_VPORT_ADMIN_STATE_UP))
 		netif_carrier_on(dev);
 
 unlock:
@@ -699,8 +699,8 @@ static int mlx5e_rep_close(struct net_device *dev)
 
 	mutex_lock(&priv->state_lock);
 	mlx5_modify_vport_admin_state(priv->mdev,
-			MLX5_QUERY_VPORT_STATE_IN_OP_MOD_ESW_VPORT,
-			rep->vport, MLX5_ESW_VPORT_ADMIN_STATE_DOWN);
+				      MLX5_QUERY_VPORT_STATE_IN_OP_MOD_ESW_VPORT,
+				      rep->vport, MLX5_ESW_VPORT_ADMIN_STATE_DOWN);
 	ret = mlx5e_close_locked(dev);
 	mutex_unlock(&priv->state_lock);
 	return ret;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 969a1768da71..85ad0948951c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -97,7 +97,7 @@ enum {
 };
 
 #define MLX5E_TC_TABLE_NUM_GROUPS 4
-#define MLX5E_TC_TABLE_MAX_GROUP_SIZE (1 << 16)
+#define MLX5E_TC_TABLE_MAX_GROUP_SIZE BIT(16)
 
 struct mlx5e_hairpin {
 	struct mlx5_hairpin *pair;
@@ -789,7 +789,7 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv,
 	mlx5_del_flow_rules(flow->rule);
 	mlx5_fc_destroy(priv->mdev, counter);
 
-	if (!mlx5e_tc_num_filters(priv) && (priv->fs.tc.t)) {
+	if (!mlx5e_tc_num_filters(priv) && priv->fs.tc.t) {
 		mlx5_destroy_flow_table(priv->fs.tc.t);
 		priv->fs.tc.t = NULL;
 	}
@@ -1765,12 +1765,12 @@ static int parse_tc_pedit_action(struct mlx5e_priv *priv,
 		err = -EOPNOTSUPP; /* can't be all optimistic */
 
 		if (htype == TCA_PEDIT_KEY_EX_HDR_TYPE_NETWORK) {
-			printk(KERN_WARNING "mlx5: legacy pedit isn't offloaded\n");
+			netdev_warn(priv->netdev, "legacy pedit isn't offloaded\n");
 			goto out_err;
 		}
 
 		if (cmd != TCA_PEDIT_KEY_EX_CMD_SET && cmd != TCA_PEDIT_KEY_EX_CMD_ADD) {
-			printk(KERN_WARNING "mlx5: pedit cmd %d isn't offloaded\n", cmd);
+			netdev_warn(priv->netdev, "pedit cmd %d isn't offloaded\n", cmd);
 			goto out_err;
 		}
 
@@ -1794,8 +1794,7 @@ static int parse_tc_pedit_action(struct mlx5e_priv *priv,
 	for (cmd = 0; cmd < __PEDIT_CMD_MAX; cmd++) {
 		cmd_masks = &masks[cmd];
 		if (memcmp(cmd_masks, &zero_masks, sizeof(zero_masks))) {
-			printk(KERN_WARNING "mlx5: attempt to offload an unsupported field (cmd %d)\n",
-			       cmd);
+			netdev_warn(priv->netdev, "attempt to offload an unsupported field (cmd %d)\n", cmd);
 			print_hex_dump(KERN_WARNING, "mask: ", DUMP_PREFIX_ADDRESS,
 				       16, 1, cmd_masks, sizeof(zero_masks), true);
 			err = -EOPNOTSUPP;
-- 
2.17.0

[net-next 07/16] net/mlx5e: Avoid redundant zeroing of offloaded TC flow attributes

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

This is not needed as the attributes are zeroed out on allocation.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Jianbo Liu <jianbol@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 85ad0948951c..7d88e813bad1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1923,7 +1923,6 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 		return -EINVAL;
 
 	attr->flow_tag = MLX5_FS_DEFAULT_FLOW_TAG;
-	attr->action = 0;
 
 	tcf_exts_to_list(exts, &actions);
 	list_for_each_entry(a, &actions, list) {
@@ -2464,7 +2463,6 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	if (!tcf_exts_has_actions(exts))
 		return -EINVAL;
 
-	memset(attr, 0, sizeof(*attr));
 	attr->in_rep = rpriv->rep;
 
 	tcf_exts_to_list(exts, &actions);
-- 
2.17.0

[net-next 08/16] net/mlx5e: Return success when TC offloaded fdb actions parsed ok

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

Reaching here, means we didn't err anywhere, so lets just
return success.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Jianbo Liu <jianbol@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 7d88e813bad1..c15bd7b6a840 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -2458,7 +2458,6 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	const struct tc_action *a;
 	LIST_HEAD(actions);
 	bool encap = false;
-	int err = 0;
 
 	if (!tcf_exts_has_actions(exts))
 		return -EINVAL;
@@ -2474,6 +2473,8 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 		}
 
 		if (is_tcf_pedit(a)) {
+			int err;
+
 			err = parse_tc_pedit_action(priv, a, MLX5_FLOW_NAMESPACE_FDB,
 						    parse_attr);
 			if (err)
@@ -2561,7 +2562,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	if (!actions_match_supported(priv, exts, parse_attr, flow))
 		return -EOPNOTSUPP;
 
-	return err;
+	return 0;
 }
 
 int mlx5e_configure_flower(struct mlx5e_priv *priv,
-- 
2.17.0

[net-next 09/16] net/mlx5e: Use local actions var while processing offloaded TC flow actions

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

Use local actions variable while parsing the actions of offloaded TC flow.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 44 ++++++++++---------
 1 file changed, 24 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index c15bd7b6a840..24b6ca205c74 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1917,6 +1917,7 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	struct mlx5_nic_flow_attr *attr = flow->nic_attr;
 	const struct tc_action *a;
 	LIST_HEAD(actions);
+	u32 action = 0;
 	int err;
 
 	if (!tcf_exts_has_actions(exts))
@@ -1927,10 +1928,10 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	tcf_exts_to_list(exts, &actions);
 	list_for_each_entry(a, &actions, list) {
 		if (is_tcf_gact_shot(a)) {
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_DROP;
+			action |= MLX5_FLOW_CONTEXT_ACTION_DROP;
 			if (MLX5_CAP_FLOWTABLE(priv->mdev,
 					       flow_table_properties_nic_receive.flow_counter))
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
+				action |= MLX5_FLOW_CONTEXT_ACTION_COUNT;
 			continue;
 		}
 
@@ -1940,13 +1941,13 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 			if (err)
 				return err;
 
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_MOD_HDR |
-					MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
+			action |= MLX5_FLOW_CONTEXT_ACTION_MOD_HDR |
+				  MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
 			continue;
 		}
 
 		if (is_tcf_csum(a)) {
-			if (csum_offload_supported(priv, attr->action,
+			if (csum_offload_supported(priv, action,
 						   tcf_csum_update_flags(a)))
 				continue;
 
@@ -1960,8 +1961,8 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 			    same_hw_devs(priv, netdev_priv(peer_dev))) {
 				parse_attr->mirred_ifindex = peer_dev->ifindex;
 				flow->flags |= MLX5E_TC_FLOW_HAIRPIN;
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
-						MLX5_FLOW_CONTEXT_ACTION_COUNT;
+				action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
+					  MLX5_FLOW_CONTEXT_ACTION_COUNT;
 			} else {
 				netdev_warn(priv->netdev, "device %s not on same HW, can't offload\n",
 					    peer_dev->name);
@@ -1980,13 +1981,14 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 			}
 
 			attr->flow_tag = mark;
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
+			action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
 			continue;
 		}
 
 		return -EINVAL;
 	}
 
+	attr->action = action;
 	if (!actions_match_supported(priv, exts, parse_attr, flow))
 		return -EOPNOTSUPP;
 
@@ -2458,6 +2460,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	const struct tc_action *a;
 	LIST_HEAD(actions);
 	bool encap = false;
+	u32 action = 0;
 
 	if (!tcf_exts_has_actions(exts))
 		return -EINVAL;
@@ -2467,8 +2470,8 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 	tcf_exts_to_list(exts, &actions);
 	list_for_each_entry(a, &actions, list) {
 		if (is_tcf_gact_shot(a)) {
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_DROP |
-					MLX5_FLOW_CONTEXT_ACTION_COUNT;
+			action |= MLX5_FLOW_CONTEXT_ACTION_DROP |
+				  MLX5_FLOW_CONTEXT_ACTION_COUNT;
 			continue;
 		}
 
@@ -2480,12 +2483,12 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 			if (err)
 				return err;
 
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_MOD_HDR;
+			action |= MLX5_FLOW_CONTEXT_ACTION_MOD_HDR;
 			continue;
 		}
 
 		if (is_tcf_csum(a)) {
-			if (csum_offload_supported(priv, attr->action,
+			if (csum_offload_supported(priv, action,
 						   tcf_csum_update_flags(a)))
 				continue;
 
@@ -2500,8 +2503,8 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 
 			if (switchdev_port_same_parent_id(priv->netdev,
 							  out_dev)) {
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
-					MLX5_FLOW_CONTEXT_ACTION_COUNT;
+				action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
+					  MLX5_FLOW_CONTEXT_ACTION_COUNT;
 				out_priv = netdev_priv(out_dev);
 				rpriv = out_priv->ppriv;
 				attr->out_rep = rpriv->rep;
@@ -2509,9 +2512,9 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 				parse_attr->mirred_ifindex = out_dev->ifindex;
 				parse_attr->tun_info = *info;
 				attr->parse_attr = parse_attr;
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_ENCAP |
-					MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
-					MLX5_FLOW_CONTEXT_ACTION_COUNT;
+				action |= MLX5_FLOW_CONTEXT_ACTION_ENCAP |
+					  MLX5_FLOW_CONTEXT_ACTION_FWD_DEST |
+					  MLX5_FLOW_CONTEXT_ACTION_COUNT;
 				/* attr->out_rep is resolved when we handle encap */
 			} else {
 				pr_err("devices %s %s not on same switch HW, can't offload forwarding\n",
@@ -2532,9 +2535,9 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 
 		if (is_tcf_vlan(a)) {
 			if (tcf_vlan_action(a) == TCA_VLAN_ACT_POP) {
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_VLAN_POP;
+				action |= MLX5_FLOW_CONTEXT_ACTION_VLAN_POP;
 			} else if (tcf_vlan_action(a) == TCA_VLAN_ACT_PUSH) {
-				attr->action |= MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH;
+				action |= MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH;
 				attr->vlan_vid = tcf_vlan_push_vid(a);
 				if (mlx5_eswitch_vlan_actions_supported(priv->mdev)) {
 					attr->vlan_prio = tcf_vlan_push_prio(a);
@@ -2552,13 +2555,14 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 		}
 
 		if (is_tcf_tunnel_release(a)) {
-			attr->action |= MLX5_FLOW_CONTEXT_ACTION_DECAP;
+			action |= MLX5_FLOW_CONTEXT_ACTION_DECAP;
 			continue;
 		}
 
 		return -EINVAL;
 	}
 
+	attr->action = action;
 	if (!actions_match_supported(priv, exts, parse_attr, flow))
 		return -EOPNOTSUPP;
 
-- 
2.17.0

[net-next 10/16] net/mlx5e: Properly order min inline mode setup while parsing TC matches

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

Set the initial value to none instead of L2, and set to L2
where the previous initial value was assumed. Make sure to
parse L2 matches before L3 matches and L3 before L4.

This is a pre-step to get the match level for more purposes
other than the validating the needed vs. actual inline level.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 129 +++++++++++-------
 1 file changed, 76 insertions(+), 53 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 24b6ca205c74..78f5c47affb4 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1201,7 +1201,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 	u16 addr_type = 0;
 	u8 ip_proto = 0;
 
-	*min_inline = MLX5_INLINE_MODE_L2;
+	*min_inline = MLX5_INLINE_MODE_NONE;
 
 	if (f->dissector->used_keys &
 	    ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) |
@@ -1251,58 +1251,6 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 					 inner_headers);
 	}
 
-	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_CONTROL)) {
-		struct flow_dissector_key_control *key =
-			skb_flow_dissector_target(f->dissector,
-						  FLOW_DISSECTOR_KEY_CONTROL,
-						  f->key);
-
-		struct flow_dissector_key_control *mask =
-			skb_flow_dissector_target(f->dissector,
-						  FLOW_DISSECTOR_KEY_CONTROL,
-						  f->mask);
-		addr_type = key->addr_type;
-
-		/* the HW doesn't support frag first/later */
-		if (mask->flags & FLOW_DIS_FIRST_FRAG)
-			return -EOPNOTSUPP;
-
-		if (mask->flags & FLOW_DIS_IS_FRAGMENT) {
-			MLX5_SET(fte_match_set_lyr_2_4, headers_c, frag, 1);
-			MLX5_SET(fte_match_set_lyr_2_4, headers_v, frag,
-				 key->flags & FLOW_DIS_IS_FRAGMENT);
-
-			/* the HW doesn't need L3 inline to match on frag=no */
-			if (key->flags & FLOW_DIS_IS_FRAGMENT)
-				*min_inline = MLX5_INLINE_MODE_IP;
-		}
-	}
-
-	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
-		struct flow_dissector_key_basic *key =
-			skb_flow_dissector_target(f->dissector,
-						  FLOW_DISSECTOR_KEY_BASIC,
-						  f->key);
-		struct flow_dissector_key_basic *mask =
-			skb_flow_dissector_target(f->dissector,
-						  FLOW_DISSECTOR_KEY_BASIC,
-						  f->mask);
-		ip_proto = key->ip_proto;
-
-		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ethertype,
-			 ntohs(mask->n_proto));
-		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ethertype,
-			 ntohs(key->n_proto));
-
-		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol,
-			 mask->ip_proto);
-		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
-			 key->ip_proto);
-
-		if (mask->ip_proto)
-			*min_inline = MLX5_INLINE_MODE_IP;
-	}
-
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_ETH_ADDRS)) {
 		struct flow_dissector_key_eth_addrs *key =
 			skb_flow_dissector_target(f->dissector,
@@ -1326,6 +1274,9 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 					     smac_47_16),
 				key->src);
+
+		if (!is_zero_ether_addr(mask->src) || !is_zero_ether_addr(mask->dst))
+			*min_inline = MLX5_INLINE_MODE_L2;
 	}
 
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_VLAN)) {
@@ -1346,9 +1297,79 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 
 			MLX5_SET(fte_match_set_lyr_2_4, headers_c, first_prio, mask->vlan_priority);
 			MLX5_SET(fte_match_set_lyr_2_4, headers_v, first_prio, key->vlan_priority);
+
+			*min_inline = MLX5_INLINE_MODE_L2;
 		}
 	}
 
+	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
+		struct flow_dissector_key_basic *key =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_BASIC,
+						  f->key);
+		struct flow_dissector_key_basic *mask =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_BASIC,
+						  f->mask);
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ethertype,
+			 ntohs(mask->n_proto));
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ethertype,
+			 ntohs(key->n_proto));
+
+		if (mask->n_proto)
+			*min_inline = MLX5_INLINE_MODE_L2;
+	}
+
+	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_CONTROL)) {
+		struct flow_dissector_key_control *key =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_CONTROL,
+						  f->key);
+
+		struct flow_dissector_key_control *mask =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_CONTROL,
+						  f->mask);
+		addr_type = key->addr_type;
+
+		/* the HW doesn't support frag first/later */
+		if (mask->flags & FLOW_DIS_FIRST_FRAG)
+			return -EOPNOTSUPP;
+
+		if (mask->flags & FLOW_DIS_IS_FRAGMENT) {
+			MLX5_SET(fte_match_set_lyr_2_4, headers_c, frag, 1);
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v, frag,
+				 key->flags & FLOW_DIS_IS_FRAGMENT);
+
+			/* the HW doesn't need L3 inline to match on frag=no */
+			if (!(key->flags & FLOW_DIS_IS_FRAGMENT))
+				*min_inline = MLX5_INLINE_MODE_L2;
+	/* ***  L2 attributes parsing up to here *** */
+			else
+				*min_inline = MLX5_INLINE_MODE_IP;
+		}
+	}
+
+	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
+		struct flow_dissector_key_basic *key =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_BASIC,
+						  f->key);
+		struct flow_dissector_key_basic *mask =
+			skb_flow_dissector_target(f->dissector,
+						  FLOW_DISSECTOR_KEY_BASIC,
+						  f->mask);
+		ip_proto = key->ip_proto;
+
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol,
+			 mask->ip_proto);
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
+			 key->ip_proto);
+
+		if (mask->ip_proto)
+			*min_inline = MLX5_INLINE_MODE_IP;
+	}
+
 	if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
 		struct flow_dissector_key_ipv4_addrs *key =
 			skb_flow_dissector_target(f->dissector,
@@ -1433,6 +1454,8 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			*min_inline = MLX5_INLINE_MODE_IP;
 	}
 
+	/* ***  L3 attributes parsing up to here *** */
+
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_PORTS)) {
 		struct flow_dissector_key_ports *key =
 			skb_flow_dissector_target(f->dissector,
-- 
2.17.0

[net-next 11/16] net/mlx5e: Get the required HW match level while parsing TC flow matches

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

Introduce levels of matching on headers of offloaded flows
(none, L2, L3, L4) that follow the inline mode levels.

This is pre-step for us to offload flows without any
matches on headers.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_tc.c   | 34 +++++++++----------
 .../net/ethernet/mellanox/mlx5/core/eswitch.h |  7 ++++
 2 files changed, 24 insertions(+), 17 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 78f5c47affb4..e84bcea8b071 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1192,7 +1192,7 @@ static int parse_tunnel_attr(struct mlx5e_priv *priv,
 static int __parse_cls_flower(struct mlx5e_priv *priv,
 			      struct mlx5_flow_spec *spec,
 			      struct tc_cls_flower_offload *f,
-			      u8 *min_inline)
+			      u8 *match_level)
 {
 	void *headers_c = MLX5_ADDR_OF(fte_match_param, spec->match_criteria,
 				       outer_headers);
@@ -1201,7 +1201,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 	u16 addr_type = 0;
 	u8 ip_proto = 0;
 
-	*min_inline = MLX5_INLINE_MODE_NONE;
+	*match_level = MLX5_MATCH_NONE;
 
 	if (f->dissector->used_keys &
 	    ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) |
@@ -1276,7 +1276,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 				key->src);
 
 		if (!is_zero_ether_addr(mask->src) || !is_zero_ether_addr(mask->dst))
-			*min_inline = MLX5_INLINE_MODE_L2;
+			*match_level = MLX5_MATCH_L2;
 	}
 
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_VLAN)) {
@@ -1298,7 +1298,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			MLX5_SET(fte_match_set_lyr_2_4, headers_c, first_prio, mask->vlan_priority);
 			MLX5_SET(fte_match_set_lyr_2_4, headers_v, first_prio, key->vlan_priority);
 
-			*min_inline = MLX5_INLINE_MODE_L2;
+			*match_level = MLX5_MATCH_L2;
 		}
 	}
 
@@ -1317,7 +1317,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			 ntohs(key->n_proto));
 
 		if (mask->n_proto)
-			*min_inline = MLX5_INLINE_MODE_L2;
+			*match_level = MLX5_MATCH_L2;
 	}
 
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_CONTROL)) {
@@ -1343,10 +1343,10 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 
 			/* the HW doesn't need L3 inline to match on frag=no */
 			if (!(key->flags & FLOW_DIS_IS_FRAGMENT))
-				*min_inline = MLX5_INLINE_MODE_L2;
+				*match_level = MLX5_INLINE_MODE_L2;
 	/* ***  L2 attributes parsing up to here *** */
 			else
-				*min_inline = MLX5_INLINE_MODE_IP;
+				*match_level = MLX5_INLINE_MODE_IP;
 		}
 	}
 
@@ -1367,7 +1367,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			 key->ip_proto);
 
 		if (mask->ip_proto)
-			*min_inline = MLX5_INLINE_MODE_IP;
+			*match_level = MLX5_MATCH_L3;
 	}
 
 	if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
@@ -1394,7 +1394,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 		       &key->dst, sizeof(key->dst));
 
 		if (mask->src || mask->dst)
-			*min_inline = MLX5_INLINE_MODE_IP;
+			*match_level = MLX5_MATCH_L3;
 	}
 
 	if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) {
@@ -1423,7 +1423,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 
 		if (ipv6_addr_type(&mask->src) != IPV6_ADDR_ANY ||
 		    ipv6_addr_type(&mask->dst) != IPV6_ADDR_ANY)
-			*min_inline = MLX5_INLINE_MODE_IP;
+			*match_level = MLX5_MATCH_L3;
 	}
 
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_IP)) {
@@ -1451,7 +1451,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			return -EOPNOTSUPP;
 
 		if (mask->tos || mask->ttl)
-			*min_inline = MLX5_INLINE_MODE_IP;
+			*match_level = MLX5_MATCH_L3;
 	}
 
 	/* ***  L3 attributes parsing up to here *** */
@@ -1496,7 +1496,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 		}
 
 		if (mask->src || mask->dst)
-			*min_inline = MLX5_INLINE_MODE_TCP_UDP;
+			*match_level = MLX5_MATCH_L4;
 	}
 
 	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_TCP)) {
@@ -1515,7 +1515,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv,
 			 ntohs(key->flags));
 
 		if (mask->flags)
-			*min_inline = MLX5_INLINE_MODE_TCP_UDP;
+			*match_level = MLX5_MATCH_L4;
 	}
 
 	return 0;
@@ -1530,19 +1530,19 @@ static int parse_cls_flower(struct mlx5e_priv *priv,
 	struct mlx5_eswitch *esw = dev->priv.eswitch;
 	struct mlx5e_rep_priv *rpriv = priv->ppriv;
 	struct mlx5_eswitch_rep *rep;
-	u8 min_inline;
+	u8 match_level;
 	int err;
 
-	err = __parse_cls_flower(priv, spec, f, &min_inline);
+	err = __parse_cls_flower(priv, spec, f, &match_level);
 
 	if (!err && (flow->flags & MLX5E_TC_FLOW_ESWITCH)) {
 		rep = rpriv->rep;
 		if (rep->vport != FDB_UPLINK_VPORT &&
 		    (esw->offloads.inline_mode != MLX5_INLINE_MODE_NONE &&
-		    esw->offloads.inline_mode < min_inline)) {
+		    esw->offloads.inline_mode < match_level)) {
 			netdev_warn(priv->netdev,
 				    "Flow is not offloaded due to min inline setting, required %d actual %d\n",
-				    min_inline, esw->offloads.inline_mode);
+				    match_level, esw->offloads.inline_mode);
 			return -EOPNOTSUPP;
 		}
 	}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index 4cd773fa55e3..efae77dd1e35 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -227,6 +227,13 @@ enum {
 	SET_VLAN_INSERT	= BIT(1)
 };
 
+enum mlx5_flow_match_level {
+	MLX5_MATCH_NONE	= MLX5_INLINE_MODE_NONE,
+	MLX5_MATCH_L2	= MLX5_INLINE_MODE_L2,
+	MLX5_MATCH_L3	= MLX5_INLINE_MODE_IP,
+	MLX5_MATCH_L4	= MLX5_INLINE_MODE_TCP_UDP,
+};
+
 struct mlx5_esw_flow_attr {
 	struct mlx5_eswitch_rep *in_rep;
 	struct mlx5_eswitch_rep *out_rep;
-- 
2.17.0

[net-next 12/16] net/mlx5e: Support offloaded TC flows with no matches on headers

[Saeed Mahameed]

From: Or Gerlitz <ogerlitz@mellanox.com>

For example:
    tc filter add dev ens2f0_0 parent ffff: flower skip_sw action drop

Note that for eswitch flows, we still always match on the source port.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c        | 10 +++++++++-
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.h      |  1 +
 .../net/ethernet/mellanox/mlx5/core/eswitch_offloads.c |  8 ++++++--
 3 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index e84bcea8b071..1dc24e3a0841 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -58,6 +58,7 @@ struct mlx5_nic_flow_attr {
 	u32 flow_tag;
 	u32 mod_hdr_id;
 	u32 hairpin_tirn;
+	u8 match_level;
 	struct mlx5_flow_table	*hairpin_ft;
 };
 
@@ -753,7 +754,9 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv,
 		table_created = true;
 	}
 
-	parse_attr->spec.match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
+	if (attr->match_level != MLX5_MATCH_NONE)
+		parse_attr->spec.match_criteria_enable = MLX5_MATCH_OUTER_HEADERS;
+
 	rule = mlx5_add_flow_rules(priv->fs.tc.t, &parse_attr->spec,
 				   &flow_act, dest, dest_ix);
 
@@ -1547,6 +1550,11 @@ static int parse_cls_flower(struct mlx5e_priv *priv,
 		}
 	}
 
+	if (flow->flags & MLX5E_TC_FLOW_ESWITCH)
+		flow->esw_attr->match_level = match_level;
+	else
+		flow->nic_attr->match_level = match_level;
+
 	return err;
 }
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
index efae77dd1e35..edf47a4d549e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h
@@ -245,6 +245,7 @@ struct mlx5_esw_flow_attr {
 	bool	vlan_handled;
 	u32	encap_id;
 	u32	mod_hdr_id;
+	u8	match_level;
 	struct mlx5e_tc_flow_parse_attr *parse_attr;
 };
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
index 35e256eb2f6e..8dd0eca03202 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c
@@ -91,8 +91,12 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw,
 	misc = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, misc_parameters);
 	MLX5_SET_TO_ONES(fte_match_set_misc, misc, source_port);
 
-	spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS |
-				      MLX5_MATCH_MISC_PARAMETERS;
+	if (attr->match_level == MLX5_MATCH_NONE)
+		spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS;
+	else
+		spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS |
+					      MLX5_MATCH_MISC_PARAMETERS;
+
 	if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_DECAP)
 		spec->match_criteria_enable |= MLX5_MATCH_INNER_HEADERS;
 
-- 
2.17.0

[net-next 13/16] net/mlx5e: Report all channels with min RX WQEs timeout

[Saeed Mahameed]

From: Eran Ben Elisha <eranbe@mellanox.com>

Report all channels which got timeout on posting the minimal number of
RX WQEs and not only the first one. Avoid busy wait on every channel,
when one of the RQs check got timeout, poll once for the remaining RQs.

In addition, add channel index to log when failed to get min RX WQEs
This info is needed in order to debug in case of dysfunctional channel.

Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 23 +++++++++----------
 1 file changed, 11 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 0f2b66b41899..b0d6dde2754f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -747,23 +747,24 @@ static void mlx5e_destroy_rq(struct mlx5e_rq *rq)
 	mlx5_core_destroy_rq(rq->mdev, rq->rqn);
 }
 
-static int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq)
+static int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq, int wait_time)
 {
-	unsigned long exp_time = jiffies + msecs_to_jiffies(20000);
+	unsigned long exp_time = jiffies + msecs_to_jiffies(wait_time);
 	struct mlx5e_channel *c = rq->channel;
 
 	struct mlx5_wq_ll *wq = &rq->wq;
 	u16 min_wqes = mlx5_min_rx_wqes(rq->wq_type, mlx5_wq_ll_get_size(wq));
 
-	while (time_before(jiffies, exp_time)) {
+	do {
 		if (wq->cur_sz >= min_wqes)
 			return 0;
 
 		msleep(20);
-	}
+	} while (time_before(jiffies, exp_time));
+
+	netdev_warn(c->netdev, "Failed to get min RX wqes on Channel[%d] RQN[0x%x] wq cur_sz(%d) min_rx_wqes(%d)\n",
+		    c->ix, rq->rqn, wq->cur_sz, min_wqes);
 
-	netdev_warn(c->netdev, "Failed to get min RX wqes on RQN[0x%x] wq cur_sz(%d) min_rx_wqes(%d)\n",
-		    rq->rqn, wq->cur_sz, min_wqes);
 	return -ETIMEDOUT;
 }
 
@@ -2128,13 +2129,11 @@ static int mlx5e_wait_channels_min_rx_wqes(struct mlx5e_channels *chs)
 	int err = 0;
 	int i;
 
-	for (i = 0; i < chs->num; i++) {
-		err = mlx5e_wait_for_min_rx_wqes(&chs->c[i]->rq);
-		if (err)
-			break;
-	}
+	for (i = 0; i < chs->num; i++)
+		err |= mlx5e_wait_for_min_rx_wqes(&chs->c[i]->rq,
+						  err ? 0 : 20000);
 
-	return err;
+	return err ? -ETIMEDOUT : 0;
 }
 
 static void mlx5e_deactivate_channels(struct mlx5e_channels *chs)
-- 
2.17.0

[net-next 14/16] net/mlx5e: Use __set_bit for adaptive-moderation bit in RQ state

[Saeed Mahameed]

From: Gal Pressman <galp@mellanox.com>

Make the code more clear by replacing the existing code with __set_bit.

Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index b0d6dde2754f..417bf2e8ab85 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -820,7 +820,7 @@ static int mlx5e_open_rq(struct mlx5e_channel *c,
 		goto err_destroy_rq;
 
 	if (params->rx_dim_enabled)
-		c->rq.state |= BIT(MLX5E_RQ_STATE_AM);
+		__set_bit(MLX5E_RQ_STATE_AM, &c->rq.state);
 
 	return 0;
 
-- 
2.17.0

[net-next 15/16] net/mlx5e: Use test bit in en accel xmit flow

[Saeed Mahameed]

From: Gal Pressman <galp@mellanox.com>

Replace (mask & bit) check with test_bit.

Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h
index 68fcb40a2847..f20074dbef32 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h
@@ -49,7 +49,7 @@ static inline struct sk_buff *mlx5e_accel_handle_tx(struct sk_buff *skb,
 						    u16 *pi)
 {
 #ifdef CONFIG_MLX5_EN_TLS
-	if (sq->state & BIT(MLX5E_SQ_STATE_TLS)) {
+	if (test_bit(MLX5E_SQ_STATE_TLS, &sq->state)) {
 		skb = mlx5e_tls_handle_tx_skb(dev, sq, skb, wqe, pi);
 		if (unlikely(!skb))
 			return NULL;
@@ -57,7 +57,7 @@ static inline struct sk_buff *mlx5e_accel_handle_tx(struct sk_buff *skb,
 #endif
 
 #ifdef CONFIG_MLX5_EN_IPSEC
-	if (sq->state & BIT(MLX5E_SQ_STATE_IPSEC)) {
+	if (test_bit(MLX5E_SQ_STATE_IPSEC, &sq->state)) {
 		skb = mlx5e_ipsec_handle_tx_skb(dev, *wqe, skb);
 		if (unlikely(!skb))
 			return NULL;
-- 
2.17.0

[net-next 16/16] net/mlx5e: Remove MLX5E_TEST_BIT macro

[Saeed Mahameed]

From: Gal Pressman <galp@mellanox.com>

MLX5E_TEST_BIT macro is the same as the already existent test_bit,
remove it and replace all usages.

Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en.h      |  2 --
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c   | 10 +++++-----
 drivers/net/ethernet/mellanox/mlx5/core/en_tx.c   |  2 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c |  4 ++--
 4 files changed, 8 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
index 9cc07da09b70..7c930088e96e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
@@ -288,8 +288,6 @@ enum {
 	MLX5E_RQ_STATE_AM,
 };
 
-#define MLX5E_TEST_BIT(state, nr) (state & BIT(nr))
-
 struct mlx5e_cq {
 	/* data path - accessed per cqe */
 	struct mlx5_cqwq           wq;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index abc6ca8168d2..53f72923b164 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -450,7 +450,7 @@ bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq)
 	struct mlx5_wq_ll *wq = &rq->wq;
 	int err;
 
-	if (unlikely(!MLX5E_TEST_BIT(rq->state, MLX5E_RQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state)))
 		return false;
 
 	if (mlx5_wq_ll_is_full(wq))
@@ -508,7 +508,7 @@ static void mlx5e_poll_ico_cq(struct mlx5e_cq *cq, struct mlx5e_rq *rq)
 	struct mlx5e_icosq *sq = container_of(cq, struct mlx5e_icosq, cq);
 	struct mlx5_cqe64 *cqe;
 
-	if (unlikely(!MLX5E_TEST_BIT(sq->state, MLX5E_SQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_SQ_STATE_ENABLED, &sq->state)))
 		return;
 
 	cqe = mlx5_cqwq_get_cqe(&cq->wq);
@@ -525,7 +525,7 @@ bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq)
 {
 	struct mlx5_wq_ll *wq = &rq->wq;
 
-	if (unlikely(!MLX5E_TEST_BIT(rq->state, MLX5E_RQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state)))
 		return false;
 
 	mlx5e_poll_ico_cq(&rq->channel->icosq.cq, rq);
@@ -1132,7 +1132,7 @@ int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget)
 	struct mlx5_cqe64 *cqe;
 	int work_done = 0;
 
-	if (unlikely(!MLX5E_TEST_BIT(rq->state, MLX5E_RQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state)))
 		return 0;
 
 	if (cq->decmprs_left)
@@ -1185,7 +1185,7 @@ bool mlx5e_poll_xdpsq_cq(struct mlx5e_cq *cq)
 
 	sq = container_of(cq, struct mlx5e_xdpsq, cq);
 
-	if (unlikely(!MLX5E_TEST_BIT(sq->state, MLX5E_SQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_SQ_STATE_ENABLED, &sq->state)))
 		return false;
 
 	cqe = mlx5_cqwq_get_cqe(&cq->wq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index 047614d2eda2..2d3f17da5f5c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -450,7 +450,7 @@ bool mlx5e_poll_tx_cq(struct mlx5e_cq *cq, int napi_budget)
 
 	sq = container_of(cq, struct mlx5e_txqsq, cq);
 
-	if (unlikely(!MLX5E_TEST_BIT(sq->state, MLX5E_SQ_STATE_ENABLED)))
+	if (unlikely(!test_bit(MLX5E_SQ_STATE_ENABLED, &sq->state)))
 		return false;
 
 	cqe = mlx5_cqwq_get_cqe(&cq->wq);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
index 900661d45c8a..5d6f9ce2bf80 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c
@@ -48,7 +48,7 @@ static void mlx5e_handle_tx_dim(struct mlx5e_txqsq *sq)
 {
 	struct net_dim_sample dim_sample;
 
-	if (unlikely(!MLX5E_TEST_BIT(sq->state, MLX5E_SQ_STATE_AM)))
+	if (unlikely(!test_bit(MLX5E_SQ_STATE_AM, &sq->state)))
 		return;
 
 	net_dim_sample(sq->cq.event_ctr, sq->stats.packets, sq->stats.bytes,
@@ -60,7 +60,7 @@ static void mlx5e_handle_rx_dim(struct mlx5e_rq *rq)
 {
 	struct net_dim_sample dim_sample;
 
-	if (unlikely(!MLX5E_TEST_BIT(rq->state, MLX5E_RQ_STATE_AM)))
+	if (unlikely(!test_bit(MLX5E_RQ_STATE_AM, &rq->state)))
 		return;
 
 	net_dim_sample(rq->cq.event_ctr, rq->stats.packets, rq->stats.bytes,
-- 
2.17.0

Re: [pull request][net-next 00/16] Mellanox, mlx5e updates 2018-05-14

[David Miller]

From: Saeed Mahameed <saeedm@mellanox.com>
Date: Mon, 14 May 2018 15:19:18 -0700

> The following series contains misc updates for mlx5e netdevice driver.
> For more information please see tag log below.
> 
> Please pull and let me know if there's any problem.

Pulled, thanks Saeed.