From 38e8a5c040d3ec99a8351c688dcdf0f549611565 Mon Sep 17 00:00:00 2001
From: Roi Dayan <roid@mellanox.com>
Date: Mon, 21 Aug 2017 12:04:50 +0300
Subject: net/mlx5e: IPoIB, Fix access to invalid memory address

When cleaning rdma netdevice we need to save the mdev pointer
because priv is released when we release netdev.

This bug was found using the kernel address sanitizer (KASAN).
use-after-free in mlx5_rdma_netdev_free+0xe3/0x100 [mlx5_core]

Fixes: 48935bbb7ae8 ("net/mlx5e: IPoIB, Add netdevice profile skeleton")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
index 8529805..145e392 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
@@ -572,12 +572,13 @@ void mlx5_rdma_netdev_free(struct net_device *netdev)
 {
 	struct mlx5e_priv          *priv    = mlx5i_epriv(netdev);
 	const struct mlx5e_profile *profile = priv->profile;
+	struct mlx5_core_dev       *mdev    = priv->mdev;
 
 	mlx5e_detach_netdev(priv);
 	profile->cleanup(priv);
 	destroy_workqueue(priv->wq);
 	free_netdev(netdev);
 
-	mlx5e_destroy_mdev_resources(priv->mdev);
+	mlx5e_destroy_mdev_resources(mdev);
 }
 EXPORT_SYMBOL(mlx5_rdma_netdev_free);
-- 
cgit v1.1


From 99d3cd27f755d63fd6cf85169eaa873d90769aa5 Mon Sep 17 00:00:00 2001
From: Inbar Karmy <inbark@mellanox.com>
Date: Thu, 24 Aug 2017 17:21:44 +0300
Subject: net/mlx5: Fix FPGA capability location

Currently, FPGA capability is located in (mdev)->caps.hca_cur,
change the location to be (mdev)->caps.fpga,
since hca_cur is reserved for HCA device capabilities.

Fixes: e29341fb3a5b ("net/mlx5: FPGA, Add basic support for Innova")
Signed-off-by: Inbar Karmy <inbark@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.c  | 4 ++--
 drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.h  | 2 +-
 drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c | 3 +--
 3 files changed, 4 insertions(+), 5 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.c
index e37453d..c0fd221 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.c
@@ -71,11 +71,11 @@ int mlx5_fpga_access_reg(struct mlx5_core_dev *dev, u8 size, u64 addr,
 	return 0;
 }
 
-int mlx5_fpga_caps(struct mlx5_core_dev *dev, u32 *caps)
+int mlx5_fpga_caps(struct mlx5_core_dev *dev)
 {
 	u32 in[MLX5_ST_SZ_DW(fpga_cap)] = {0};
 
-	return mlx5_core_access_reg(dev, in, sizeof(in), caps,
+	return mlx5_core_access_reg(dev, in, sizeof(in), dev->caps.fpga,
 				    MLX5_ST_SZ_BYTES(fpga_cap),
 				    MLX5_REG_FPGA_CAP, 0, 0);
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.h b/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.h
index 94bdfd4..d05233c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/cmd.h
@@ -65,7 +65,7 @@ struct mlx5_fpga_qp_counters {
 	u64 rx_total_drop;
 };
 
-int mlx5_fpga_caps(struct mlx5_core_dev *dev, u32 *caps);
+int mlx5_fpga_caps(struct mlx5_core_dev *dev);
 int mlx5_fpga_query(struct mlx5_core_dev *dev, struct mlx5_fpga_query *query);
 int mlx5_fpga_ctrl_op(struct mlx5_core_dev *dev, u8 op);
 int mlx5_fpga_access_reg(struct mlx5_core_dev *dev, u8 size, u64 addr,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c b/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
index 9034e99..dc89703 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
@@ -139,8 +139,7 @@ int mlx5_fpga_device_start(struct mlx5_core_dev *mdev)
 	if (err)
 		goto out;
 
-	err = mlx5_fpga_caps(fdev->mdev,
-			     fdev->mdev->caps.hca_cur[MLX5_CAP_FPGA]);
+	err = mlx5_fpga_caps(fdev->mdev);
 	if (err)
 		goto out;
 
-- 
cgit v1.1


From 16f1c5bb3ed75b3cf3ced537db40f7e1a244debe Mon Sep 17 00:00:00 2001
From: Raed Salem <raeds@mellanox.com>
Date: Sun, 30 Jul 2017 11:02:51 +0300
Subject: net/mlx5: Check device capability for maximum flow counters

Added check for the maximal number of flow counters attached
to rule (FTE).

Fixes: bd5251dbf156b ('net/mlx5_core: Introduce flow steering destination of type counter')
Signed-off-by: Raed Salem <raeds@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c  |  8 ++++++++
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.h | 11 +++++++++++
 2 files changed, 19 insertions(+)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index e0d0efd..36ecc2b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -293,6 +293,9 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
 	}
 
 	if (fte->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) {
+		int max_list_size = BIT(MLX5_CAP_FLOWTABLE_TYPE(dev,
+					log_max_flow_counter,
+					ft->type));
 		int list_size = 0;
 
 		list_for_each_entry(dst, &fte->node.children, node.list) {
@@ -305,12 +308,17 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
 			in_dests += MLX5_ST_SZ_BYTES(dest_format_struct);
 			list_size++;
 		}
+		if (list_size > max_list_size) {
+			err = -EINVAL;
+			goto err_out;
+		}
 
 		MLX5_SET(flow_context, in_flow_context, flow_counter_list_size,
 			 list_size);
 	}
 
 	err = mlx5_cmd_exec(dev, in, inlen, out, sizeof(out));
+err_out:
 	kvfree(in);
 	return err;
 }
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
index 5509a752..48dd789 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
@@ -52,6 +52,7 @@ enum fs_flow_table_type {
 	FS_FT_FDB             = 0X4,
 	FS_FT_SNIFFER_RX	= 0X5,
 	FS_FT_SNIFFER_TX	= 0X6,
+	FS_FT_MAX_TYPE = FS_FT_SNIFFER_TX,
 };
 
 enum fs_flow_table_op_mod {
@@ -260,4 +261,14 @@ void mlx5_cleanup_fs(struct mlx5_core_dev *dev);
 #define fs_for_each_dst(pos, fte)			\
 	fs_list_for_each_entry(pos, &(fte)->node.children)
 
+#define MLX5_CAP_FLOWTABLE_TYPE(mdev, cap, type) (		\
+	(type == FS_FT_NIC_RX) ? MLX5_CAP_FLOWTABLE_NIC_RX(mdev, cap) :		\
+	(type == FS_FT_ESW_EGRESS_ACL) ? MLX5_CAP_ESW_EGRESS_ACL(mdev, cap) :		\
+	(type == FS_FT_ESW_INGRESS_ACL) ? MLX5_CAP_ESW_INGRESS_ACL(mdev, cap) :		\
+	(type == FS_FT_FDB) ? MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, cap) :		\
+	(type == FS_FT_SNIFFER_RX) ? MLX5_CAP_FLOWTABLE_SNIFFER_RX(mdev, cap) :		\
+	(type == FS_FT_SNIFFER_TX) ? MLX5_CAP_FLOWTABLE_SNIFFER_TX(mdev, cap) :		\
+	(BUILD_BUG_ON_ZERO(FS_FT_SNIFFER_TX != FS_FT_MAX_TYPE))\
+	)
+
 #endif
-- 
cgit v1.1


From ace743214ea205c7d433562c5fa24e33bdfda7ab Mon Sep 17 00:00:00 2001
From: Paul Blakey <paulb@mellanox.com>
Date: Tue, 5 Sep 2017 15:05:51 +0300
Subject: net/mlx5e: Fix erroneous freeing of encap header buffer

In case the neighbour for the tunnel destination isn't valid,
we send a neighbour update request but we free the encap
header buffer. This is wrong, because we still need it for
allocating a HW encap entry once the neighbour is available.

Fix that by skipping freeing it if we wait for neighbour.

Fixes: 232c001398ae ('net/mlx5e: Add support to neighbour update flow')
Signed-off-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 | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index da503e6..4e2fc01 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1564,7 +1564,7 @@ static int mlx5e_create_encap_header_ipv4(struct mlx5e_priv *priv,
 		break;
 	default:
 		err = -EOPNOTSUPP;
-		goto out;
+		goto free_encap;
 	}
 	fl4.flowi4_tos = tun_key->tos;
 	fl4.daddr = tun_key->u.ipv4.dst;
@@ -1573,7 +1573,7 @@ static int mlx5e_create_encap_header_ipv4(struct mlx5e_priv *priv,
 	err = mlx5e_route_lookup_ipv4(priv, mirred_dev, &out_dev,
 				      &fl4, &n, &ttl);
 	if (err)
-		goto out;
+		goto free_encap;
 
 	/* used by mlx5e_detach_encap to lookup a neigh hash table
 	 * entry in the neigh hash table when a user deletes a rule
@@ -1590,7 +1590,7 @@ static int mlx5e_create_encap_header_ipv4(struct mlx5e_priv *priv,
 	 */
 	err = mlx5e_rep_encap_entry_attach(netdev_priv(out_dev), e);
 	if (err)
-		goto out;
+		goto free_encap;
 
 	read_lock_bh(&n->lock);
 	nud_state = n->nud_state;
@@ -1630,8 +1630,9 @@ static int mlx5e_create_encap_header_ipv4(struct mlx5e_priv *priv,
 
 destroy_neigh_entry:
 	mlx5e_rep_encap_entry_detach(netdev_priv(e->out_dev), e);
-out:
+free_encap:
 	kfree(encap_header);
+out:
 	if (n)
 		neigh_release(n);
 	return err;
@@ -1668,7 +1669,7 @@ static int mlx5e_create_encap_header_ipv6(struct mlx5e_priv *priv,
 		break;
 	default:
 		err = -EOPNOTSUPP;
-		goto out;
+		goto free_encap;
 	}
 
 	fl6.flowlabel = ip6_make_flowinfo(RT_TOS(tun_key->tos), tun_key->label);
@@ -1678,7 +1679,7 @@ static int mlx5e_create_encap_header_ipv6(struct mlx5e_priv *priv,
 	err = mlx5e_route_lookup_ipv6(priv, mirred_dev, &out_dev,
 				      &fl6, &n, &ttl);
 	if (err)
-		goto out;
+		goto free_encap;
 
 	/* used by mlx5e_detach_encap to lookup a neigh hash table
 	 * entry in the neigh hash table when a user deletes a rule
@@ -1695,7 +1696,7 @@ static int mlx5e_create_encap_header_ipv6(struct mlx5e_priv *priv,
 	 */
 	err = mlx5e_rep_encap_entry_attach(netdev_priv(out_dev), e);
 	if (err)
-		goto out;
+		goto free_encap;
 
 	read_lock_bh(&n->lock);
 	nud_state = n->nud_state;
@@ -1736,8 +1737,9 @@ static int mlx5e_create_encap_header_ipv6(struct mlx5e_priv *priv,
 
 destroy_neigh_entry:
 	mlx5e_rep_encap_entry_detach(netdev_priv(e->out_dev), e);
-out:
+free_encap:
 	kfree(encap_header);
+out:
 	if (n)
 		neigh_release(n);
 	return err;
-- 
cgit v1.1


From bdd66ac0aeed971d1cb42b3aa0d11b0ea3842e09 Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Sun, 11 Jun 2017 21:13:25 +0300
Subject: net/mlx5e: Disallow TC offloading of unsupported match/action
 combinations

When offloading header re-write, the HW may need to adjust checksums along
the packet. For IP traffic, and a case where we are asked to modify fields in
the IP header, current HW supports that only for TCP and UDP. Enforce it, in
this case fail the offloading attempt for non TCP/UDP packets.

Fixes: d7e75a325cb2 ('net/mlx5e: Add offloading of E-Switch TC pedit (header re-write) actions')
Fixes: 2f4fe4cab073 ('net/mlx5e: Add offloading of NIC TC pedit (header re-write) actions')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 70 +++++++++++++++++++++++++
 1 file changed, 70 insertions(+)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 4e2fc01..d378600 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1317,6 +1317,69 @@ static bool csum_offload_supported(struct mlx5e_priv *priv, u32 action, u32 upda
 	return true;
 }
 
+static bool modify_header_match_supported(struct mlx5_flow_spec *spec,
+					  struct tcf_exts *exts)
+{
+	const struct tc_action *a;
+	bool modify_ip_header;
+	LIST_HEAD(actions);
+	u8 htype, ip_proto;
+	void *headers_v;
+	u16 ethertype;
+	int nkeys, i;
+
+	headers_v = MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers);
+	ethertype = MLX5_GET(fte_match_set_lyr_2_4, headers_v, ethertype);
+
+	/* for non-IP we only re-write MACs, so we're okay */
+	if (ethertype != ETH_P_IP && ethertype != ETH_P_IPV6)
+		goto out_ok;
+
+	modify_ip_header = false;
+	tcf_exts_to_list(exts, &actions);
+	list_for_each_entry(a, &actions, list) {
+		if (!is_tcf_pedit(a))
+			continue;
+
+		nkeys = tcf_pedit_nkeys(a);
+		for (i = 0; i < nkeys; i++) {
+			htype = tcf_pedit_htype(a, i);
+			if (htype == TCA_PEDIT_KEY_EX_HDR_TYPE_IP4 ||
+			    htype == TCA_PEDIT_KEY_EX_HDR_TYPE_IP6) {
+				modify_ip_header = true;
+				break;
+			}
+		}
+	}
+
+	ip_proto = MLX5_GET(fte_match_set_lyr_2_4, headers_v, ip_protocol);
+	if (modify_ip_header && ip_proto != IPPROTO_TCP && ip_proto != IPPROTO_UDP) {
+		pr_info("can't offload re-write of ip proto %d\n", ip_proto);
+		return false;
+	}
+
+out_ok:
+	return true;
+}
+
+static bool actions_match_supported(struct mlx5e_priv *priv,
+				    struct tcf_exts *exts,
+				    struct mlx5e_tc_flow_parse_attr *parse_attr,
+				    struct mlx5e_tc_flow *flow)
+{
+	u32 actions;
+
+	if (flow->flags & MLX5E_TC_FLOW_ESWITCH)
+		actions = flow->esw_attr->action;
+	else
+		actions = flow->nic_attr->action;
+
+	if (actions & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR)
+		return modify_header_match_supported(&parse_attr->spec, exts);
+
+	return true;
+}
+
 static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 				struct mlx5e_tc_flow_parse_attr *parse_attr,
 				struct mlx5e_tc_flow *flow)
@@ -1378,6 +1441,9 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 		return -EINVAL;
 	}
 
+	if (!actions_match_supported(priv, exts, parse_attr, flow))
+		return -EOPNOTSUPP;
+
 	return 0;
 }
 
@@ -1936,6 +2002,10 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 
 		return -EINVAL;
 	}
+
+	if (!actions_match_supported(priv, exts, parse_attr, flow))
+		return -EOPNOTSUPP;
+
 	return err;
 }
 
-- 
cgit v1.1


From b281208911a549e391d92ee6cb680dcd3d71783b Mon Sep 17 00:00:00 2001
From: Vlad Buslov <vladbu@mellanox.com>
Date: Tue, 8 Aug 2017 11:45:28 +0300
Subject: net/mlx5e: Check encap entry state when offloading tunneled flows

Encap entries cached by the driver could be invalidated due to
tunnel destination neighbour state changes.
When attempting to offload a flow that uses a cached encap entry,
we must check the entry validity and defer the offloading
if the entry exists but not valid.

When EAGAIN is returned, the flow offloading to hardware takes place
by the neigh update code when the tunnel destination neighbour
becomes connected.

Fixes: 232c001398ae ("net/mlx5e: Add support to neighbour update flow")
Signed-off-by: Vlad Buslov <vladbu@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 | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index d378600..1aa2028 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -1859,6 +1859,7 @@ vxlan_encap_offload_err:
 		}
 	}
 
+	/* must verify if encap is valid or not */
 	if (found)
 		goto attach_flow;
 
@@ -1885,6 +1886,8 @@ attach_flow:
 	*encap_dev = e->out_dev;
 	if (e->flags & MLX5_ENCAP_ENTRY_VALID)
 		attr->encap_id = e->encap_id;
+	else
+		err = -EAGAIN;
 
 	return err;
 
-- 
cgit v1.1


From b20eab15a1d5091e45022401e75b49948e8be33f Mon Sep 17 00:00:00 2001
From: Gal Pressman <galp@mellanox.com>
Date: Tue, 12 Sep 2017 17:51:12 +0300
Subject: net/mlx5e: Print netdev features correctly in error message

Use the correct formatting for netdev features.

Fixes: 0e405443e803 ("net/mlx5e: Improve set features ndo resiliency")
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index dfc2972..84b013d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -3333,8 +3333,8 @@ static int mlx5e_handle_feature(struct net_device *netdev,
 
 	err = feature_handler(netdev, enable);
 	if (err) {
-		netdev_err(netdev, "%s feature 0x%llx failed err %d\n",
-			   enable ? "Enable" : "Disable", feature, err);
+		netdev_err(netdev, "%s feature %pNF failed, err %d\n",
+			   enable ? "Enable" : "Disable", &feature, err);
 		return err;
 	}
 
-- 
cgit v1.1


From 1456f69ff5fbba48ed5bc86e858e945e693ba0b7 Mon Sep 17 00:00:00 2001
From: Gal Pressman <galp@mellanox.com>
Date: Sun, 10 Sep 2017 10:36:06 +0300
Subject: net/mlx5e: Don't add/remove 802.1ad rules when changing 802.1Q VLAN
 filter

Toggling of C-tag VLAN filter should not affect the "any S-tag" steering rule.

Fixes: 8a271746a264 ("net/mlx5e: Receive s-tagged packets in promiscuous mode")
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_fs.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
index f11fd07..850cdc9 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c
@@ -291,7 +291,7 @@ void mlx5e_enable_vlan_filter(struct mlx5e_priv *priv)
 	priv->fs.vlan.filter_disabled = false;
 	if (priv->netdev->flags & IFF_PROMISC)
 		return;
-	mlx5e_del_any_vid_rules(priv);
+	mlx5e_del_vlan_rule(priv, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0);
 }
 
 void mlx5e_disable_vlan_filter(struct mlx5e_priv *priv)
@@ -302,7 +302,7 @@ void mlx5e_disable_vlan_filter(struct mlx5e_priv *priv)
 	priv->fs.vlan.filter_disabled = true;
 	if (priv->netdev->flags & IFF_PROMISC)
 		return;
-	mlx5e_add_any_vid_rules(priv);
+	mlx5e_add_vlan_rule(priv, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0);
 }
 
 int mlx5e_vlan_rx_add_vid(struct net_device *dev, __always_unused __be16 proto,
-- 
cgit v1.1


From 603e1f5bd3ca76f16688e10040545594d2e91ba4 Mon Sep 17 00:00:00 2001
From: Gal Pressman <galp@mellanox.com>
Date: Wed, 13 Sep 2017 15:37:50 +0300
Subject: net/mlx5e: Fix calculated checksum offloads counters

Instead of calculating the offloads counters, count them explicitly.
The calculations done for these counters would result in bugs in some
cases, for example:
When running TCP traffic over a VXLAN tunnel with TSO enabled the following
counters would increase:
       tx_csum_partial: 1,333,284
       tx_csum_partial_inner: 29,286
       tx4_csum_partial_inner: 384
       tx7_csum_partial_inner: 8
       tx9_csum_partial_inner: 34
       tx10_csum_partial_inner: 26,807
       tx11_csum_partial_inner: 287
       tx12_csum_partial_inner: 27
       tx16_csum_partial_inner: 6
       tx25_csum_partial_inner: 1,733

Seems like tx_csum_partial increased out of nowhere.
The issue is in the following calculation in mlx5e_update_sw_counters:
s->tx_csum_partial = s->tx_packets - tx_offload_none - s->tx_csum_partial_inner;

While tx_packets increases by the number of GSO segments for each SKB,
tx_csum_partial_inner will only increase by one, resulting in wrong
tx_csum_partial counter.

Fixes: bfe6d8d1d433 ("net/mlx5e: Reorganize ethtool statistics")
Signed-off-by: Gal Pressman <galp@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_main.c  | 9 +++------
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c    | 3 +++
 drivers/net/ethernet/mellanox/mlx5/core/en_stats.h | 6 ++++++
 drivers/net/ethernet/mellanox/mlx5/core/en_tx.c    | 1 +
 4 files changed, 13 insertions(+), 6 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
index 84b013d..cc11bbb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
@@ -184,7 +184,6 @@ static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
 	struct mlx5e_sw_stats temp, *s = &temp;
 	struct mlx5e_rq_stats *rq_stats;
 	struct mlx5e_sq_stats *sq_stats;
-	u64 tx_offload_none = 0;
 	int i, j;
 
 	memset(s, 0, sizeof(*s));
@@ -199,6 +198,7 @@ static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
 		s->rx_lro_bytes	+= rq_stats->lro_bytes;
 		s->rx_csum_none	+= rq_stats->csum_none;
 		s->rx_csum_complete += rq_stats->csum_complete;
+		s->rx_csum_unnecessary += rq_stats->csum_unnecessary;
 		s->rx_csum_unnecessary_inner += rq_stats->csum_unnecessary_inner;
 		s->rx_xdp_drop += rq_stats->xdp_drop;
 		s->rx_xdp_tx += rq_stats->xdp_tx;
@@ -229,14 +229,11 @@ static void mlx5e_update_sw_counters(struct mlx5e_priv *priv)
 			s->tx_queue_dropped	+= sq_stats->dropped;
 			s->tx_xmit_more		+= sq_stats->xmit_more;
 			s->tx_csum_partial_inner += sq_stats->csum_partial_inner;
-			tx_offload_none		+= sq_stats->csum_none;
+			s->tx_csum_none		+= sq_stats->csum_none;
+			s->tx_csum_partial	+= sq_stats->csum_partial;
 		}
 	}
 
-	/* Update calculated offload counters */
-	s->tx_csum_partial = s->tx_packets - tx_offload_none - s->tx_csum_partial_inner;
-	s->rx_csum_unnecessary = s->rx_packets - s->rx_csum_none - s->rx_csum_complete;
-
 	s->link_down_events_phy = MLX5_GET(ppcnt_reg,
 				priv->stats.pport.phy_counters,
 				counter_set.phys_layer_cntrs.link_down_events);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index f1dd638..15a1687 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -627,6 +627,7 @@ static inline void mlx5e_handle_csum(struct net_device *netdev,
 
 	if (lro) {
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
+		rq->stats.csum_unnecessary++;
 		return;
 	}
 
@@ -644,7 +645,9 @@ static inline void mlx5e_handle_csum(struct net_device *netdev,
 			skb->csum_level = 1;
 			skb->encapsulation = 1;
 			rq->stats.csum_unnecessary_inner++;
+			return;
 		}
+		rq->stats.csum_unnecessary++;
 		return;
 	}
 csum_none:
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
index 6d199ff..f863721 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h
@@ -68,6 +68,7 @@ struct mlx5e_sw_stats {
 	u64 rx_xdp_drop;
 	u64 rx_xdp_tx;
 	u64 rx_xdp_tx_full;
+	u64 tx_csum_none;
 	u64 tx_csum_partial;
 	u64 tx_csum_partial_inner;
 	u64 tx_queue_stopped;
@@ -108,6 +109,7 @@ static const struct counter_desc sw_stats_desc[] = {
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_xdp_drop) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_xdp_tx) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_xdp_tx_full) },
+	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_none) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_partial) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_csum_partial_inner) },
 	{ MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, tx_queue_stopped) },
@@ -339,6 +341,7 @@ struct mlx5e_rq_stats {
 	u64 packets;
 	u64 bytes;
 	u64 csum_complete;
+	u64 csum_unnecessary;
 	u64 csum_unnecessary_inner;
 	u64 csum_none;
 	u64 lro_packets;
@@ -363,6 +366,7 @@ static const struct counter_desc rq_stats_desc[] = {
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, packets) },
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, bytes) },
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_complete) },
+	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_unnecessary) },
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_unnecessary_inner) },
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_none) },
 	{ MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, xdp_drop) },
@@ -392,6 +396,7 @@ struct mlx5e_sq_stats {
 	u64 tso_bytes;
 	u64 tso_inner_packets;
 	u64 tso_inner_bytes;
+	u64 csum_partial;
 	u64 csum_partial_inner;
 	u64 nop;
 	/* less likely accessed in data path */
@@ -408,6 +413,7 @@ static const struct counter_desc sq_stats_desc[] = {
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tso_bytes) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tso_inner_packets) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, tso_inner_bytes) },
+	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, csum_partial) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, csum_partial_inner) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, nop) },
 	{ MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, csum_none) },
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
index fee43e4..1d6925d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c
@@ -193,6 +193,7 @@ mlx5e_txwqe_build_eseg_csum(struct mlx5e_txqsq *sq, struct sk_buff *skb, struct
 			sq->stats.csum_partial_inner++;
 		} else {
 			eseg->cs_flags |= MLX5_ETH_WQE_L4_CSUM;
+			sq->stats.csum_partial++;
 		}
 	} else
 		sq->stats.csum_none++;
-- 
cgit v1.1


From 480df991b869eff02a004e8fe7707900437cfcd4 Mon Sep 17 00:00:00 2001
From: Matan Barak <matanb@mellanox.com>
Date: Thu, 31 Aug 2017 18:52:14 +0300
Subject: net/mlx5: Fix static checker warning on steering tracepoints code

Fix this sparse complaint:

drivers/net/ethernet/mellanox/mlx5/core/./diag/fs_tracepoint.h:172:1:
	warning: odd constant _Bool cast (ffffffffffffffff becomes 1)

Fixes: d9fea79171ee ('net/mlx5: Add tracepoints')
Signed-off-by: Matan Barak <matanb@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h
index 1e3a6c3..80eef41 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.h
@@ -139,7 +139,7 @@ TRACE_EVENT(mlx5_fs_del_fg,
 	{MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO, "NEXT_PRIO"}
 
 TRACE_EVENT(mlx5_fs_set_fte,
-	    TP_PROTO(const struct fs_fte *fte, bool new_fte),
+	    TP_PROTO(const struct fs_fte *fte, int new_fte),
 	    TP_ARGS(fte, new_fte),
 	    TP_STRUCT__entry(
 		__field(const struct fs_fte *, fte)
@@ -149,7 +149,7 @@ TRACE_EVENT(mlx5_fs_set_fte,
 		__field(u32, action)
 		__field(u32, flow_tag)
 		__field(u8,  mask_enable)
-		__field(bool, new_fte)
+		__field(int, new_fte)
 		__array(u32, mask_outer, MLX5_ST_SZ_DW(fte_match_set_lyr_2_4))
 		__array(u32, mask_inner, MLX5_ST_SZ_DW(fte_match_set_lyr_2_4))
 		__array(u32, mask_misc, MLX5_ST_SZ_DW(fte_match_set_misc))
-- 
cgit v1.1


From 353f59f4d41e9c5798a15c5c52958f25b579a3d5 Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Sun, 24 Sep 2017 09:54:00 +0200
Subject: net/mlx5: Fix wrong indentation in enable SRIOV code

Smatch is screaming:

drivers/net/ethernet/mellanox/mlx5/core/sriov.c:112
	mlx5_device_enable_sriov() warn: inconsistent indenting

fix that.

Fixes: 7ecf6d8ff154 ('IB/mlx5: Restore IB guid/policy for virtual functions')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/sriov.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
index 6c48e99..2a8b529 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c
@@ -109,7 +109,7 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs)
 				mlx5_core_warn(dev,
 					       "failed to restore VF %d settings, err %d\n",
 					       vf, err);
-			continue;
+				continue;
 			}
 		}
 		mlx5_core_dbg(dev, "successfully enabled VF* %d\n", vf);
-- 
cgit v1.1


From 28a04c7b7bbecaab642fcb6a2d7354eb70ea7fbe Mon Sep 17 00:00:00 2001
From: Petr Machata <petrm@mellanox.com>
Date: Mon, 2 Oct 2017 12:14:56 +0200
Subject: mlxsw: spectrum_router: Move VRF refcounting

When creating a new RIF, bumping RIF count of the containing VR is the
last thing to be done. Symmetrically, when destroying a RIF, RIF count
is first dropped and only then the rest of the cleanup proceeds.

That's a problem for loopback RIFs. Those hold two VR references: one
for overlay and one for underlay. mlxsw_sp_rif_destroy() releases the
overlay one, and the deconfigure() callback the underlay one. But if
both overlay and underlay are the same, and if there are no other
artifacts holding the VR alive, this put actually destroys the VR. Later
on, when mlxsw_sp_rif_destroy() calls mlxsw_sp_vr_put() for the same VR,
the VR will already have been released and the kernel crashes with NULL
pointer dereference.

The underlying problem is that the RIF under destruction ends up
referencing the overlay VR much longer than it claims: all the way until
the call to mlxsw_sp_vr_put(). So line up the reference counting
properly to reflect this. Make corresponding changes in
mlxsw_sp_rif_create() as well for symmetry.

Fixes: 6ddb7426a7d4 ("mlxsw: spectrum_router: Introduce loopback RIFs")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 2cfb3f5..3917b4d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -5068,6 +5068,7 @@ mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp,
 	vr = mlxsw_sp_vr_get(mlxsw_sp, tb_id ? : RT_TABLE_MAIN);
 	if (IS_ERR(vr))
 		return ERR_CAST(vr);
+	vr->rif_count++;
 
 	err = mlxsw_sp_rif_index_alloc(mlxsw_sp, &rif_index);
 	if (err)
@@ -5099,7 +5100,6 @@ mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp,
 
 	mlxsw_sp_rif_counters_alloc(rif);
 	mlxsw_sp->router->rifs[rif_index] = rif;
-	vr->rif_count++;
 
 	return rif;
 
@@ -5110,6 +5110,7 @@ err_fid_get:
 	kfree(rif);
 err_rif_alloc:
 err_rif_index_alloc:
+	vr->rif_count--;
 	mlxsw_sp_vr_put(vr);
 	return ERR_PTR(err);
 }
@@ -5124,7 +5125,6 @@ void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif)
 	mlxsw_sp_router_rif_gone_sync(mlxsw_sp, rif);
 	vr = &mlxsw_sp->router->vrs[rif->vr_id];
 
-	vr->rif_count--;
 	mlxsw_sp->router->rifs[rif->rif_index] = NULL;
 	mlxsw_sp_rif_counters_free(rif);
 	ops->deconfigure(rif);
@@ -5132,6 +5132,7 @@ void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif)
 		/* Loopback RIFs are not associated with a FID. */
 		mlxsw_sp_fid_put(fid);
 	kfree(rif);
+	vr->rif_count--;
 	mlxsw_sp_vr_put(vr);
 }
 
-- 
cgit v1.1


From de0f43c01a4b5d408a5c087c8a92ac1739938f8b Mon Sep 17 00:00:00 2001
From: Petr Machata <petrm@mellanox.com>
Date: Mon, 2 Oct 2017 12:14:57 +0200
Subject: mlxsw: spectrum_router: Track RIF of IPIP next hops

When considering whether to set RTNH_F_OFFLOAD flag on an IPv6 route,
mlxsw_sp_fib6_entry_offload_set() looks up the mlxsw_sp_nexthop
corresponding to a given route, and decides based on whether the next
hop's offloaded flag was set. When looking for the matching next hop, it
also takes into account the device of the route, which must match next
hop's RIF.

IPIP next hops however hitherto didn't set the RIF. As a result, IPv6
routes forwarding traffic to IP-in-IP netdevices are never marked as
offloaded, even when they actually are.

Thus track RIF of IPIP next hops the same way as that of ETHERNET next
hops.

Fixes: 8f28a3097645 ("mlxsw: spectrum_router: Support IPv6 overlay encap")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 3917b4d..032089e 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -2723,6 +2723,7 @@ static void mlxsw_sp_nexthop_type_fini(struct mlxsw_sp *mlxsw_sp,
 		mlxsw_sp_nexthop_rif_fini(nh);
 		break;
 	case MLXSW_SP_NEXTHOP_TYPE_IPIP:
+		mlxsw_sp_nexthop_rif_fini(nh);
 		mlxsw_sp_nexthop_ipip_fini(mlxsw_sp, nh);
 		break;
 	}
@@ -2742,7 +2743,11 @@ static int mlxsw_sp_nexthop4_type_init(struct mlxsw_sp *mlxsw_sp,
 	    router->ipip_ops_arr[ipipt]->can_offload(mlxsw_sp, dev,
 						     MLXSW_SP_L3_PROTO_IPV4)) {
 		nh->type = MLXSW_SP_NEXTHOP_TYPE_IPIP;
-		return mlxsw_sp_nexthop_ipip_init(mlxsw_sp, ipipt, nh, dev);
+		err = mlxsw_sp_nexthop_ipip_init(mlxsw_sp, ipipt, nh, dev);
+		if (err)
+			return err;
+		mlxsw_sp_nexthop_rif_init(nh, &nh->ipip_entry->ol_lb->common);
+		return 0;
 	}
 
 	nh->type = MLXSW_SP_NEXTHOP_TYPE_ETH;
@@ -4009,7 +4014,11 @@ static int mlxsw_sp_nexthop6_type_init(struct mlxsw_sp *mlxsw_sp,
 	    router->ipip_ops_arr[ipipt]->can_offload(mlxsw_sp, dev,
 						     MLXSW_SP_L3_PROTO_IPV6)) {
 		nh->type = MLXSW_SP_NEXTHOP_TYPE_IPIP;
-		return mlxsw_sp_nexthop_ipip_init(mlxsw_sp, ipipt, nh, dev);
+		err = mlxsw_sp_nexthop_ipip_init(mlxsw_sp, ipipt, nh, dev);
+		if (err)
+			return err;
+		mlxsw_sp_nexthop_rif_init(nh, &nh->ipip_entry->ol_lb->common);
+		return 0;
 	}
 
 	nh->type = MLXSW_SP_NEXTHOP_TYPE_ETH;
-- 
cgit v1.1


From a69518cf0b4cbf02c6bc1239cdeb8750a9eb8077 Mon Sep 17 00:00:00 2001
From: Ido Schimmel <idosch@mellanox.com>
Date: Sun, 8 Oct 2017 11:53:26 +0200
Subject: mlxsw: spectrum_router: Avoid expensive lookup during route removal

In commit fc922bb0dd94 ("mlxsw: spectrum_router: Use one LPM tree for
all virtual routers") I increased the scale of supported VRFs by having
all of them share the same LPM tree.

In order to avoid look-ups for prefix lengths that don't exist, each
route removal would trigger an aggregation across all the active virtual
routers to see which prefix lengths are in use and which aren't and
structure the tree accordingly.

With the way the data structures are currently laid out, this is a very
expensive operation. When preformed repeatedly - due to the invocation
of the abort mechanism - and with enough VRFs, this can result in a hung
task.

For now, avoid this optimization until it can be properly re-added in
net-next.

Fixes: fc922bb0dd94 ("mlxsw: spectrum_router: Use one LPM tree for all virtual routers")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reported-by: David Ahern <dsa@cumulusnetworks.com>
Tested-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 14 --------------
 1 file changed, 14 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 032089e..c16718d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -3505,20 +3505,6 @@ static int mlxsw_sp_fib_lpm_tree_link(struct mlxsw_sp *mlxsw_sp,
 static void mlxsw_sp_fib_lpm_tree_unlink(struct mlxsw_sp *mlxsw_sp,
 					 struct mlxsw_sp_fib *fib)
 {
-	struct mlxsw_sp_prefix_usage req_prefix_usage = {{ 0 } };
-	struct mlxsw_sp_lpm_tree *lpm_tree;
-
-	/* Aggregate prefix lengths across all virtual routers to make
-	 * sure we only have used prefix lengths in the LPM tree.
-	 */
-	mlxsw_sp_vrs_prefixes(mlxsw_sp, fib->proto, &req_prefix_usage);
-	lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage,
-					 fib->proto);
-	if (IS_ERR(lpm_tree))
-		goto err_tree_get;
-	mlxsw_sp_vrs_lpm_tree_replace(mlxsw_sp, fib, lpm_tree);
-
-err_tree_get:
 	if (!mlxsw_sp_prefix_usage_none(&fib->prefix_usage))
 		return;
 	mlxsw_sp_vr_lpm_tree_unbind(mlxsw_sp, fib);
-- 
cgit v1.1


From d965465b60bad79d0b067f1009ba80ae76a6561a Mon Sep 17 00:00:00 2001
From: Ido Schimmel <idosch@mellanox.com>
Date: Mon, 16 Oct 2017 16:28:28 +0200
Subject: mlxsw: core: Fix possible deadlock

When an EMAD is transmitted, a timeout work item is scheduled with a
delay of 200ms, so that another EMAD will be retried until a maximum of
five retries.

In certain situations, it's possible for the function waiting on the
EMAD to be associated with a work item that is queued on the same
workqueue (`mlxsw_core`) as the timeout work item. This results in
flushing a work item on the same workqueue.

According to commit e159489baa71 ("workqueue: relax lockdep annotation
on flush_work()") the above may lead to a deadlock in case the workqueue
has only one worker active or if the system in under memory pressure and
the rescue worker is in use. The latter explains the very rare and
random nature of the lockdep splats we have been seeing:

[   52.730240] ============================================
[   52.736179] WARNING: possible recursive locking detected
[   52.742119] 4.14.0-rc3jiri+ #4 Not tainted
[   52.746697] --------------------------------------------
[   52.752635] kworker/1:3/599 is trying to acquire lock:
[   52.758378]  (mlxsw_core_driver_name){+.+.}, at: [<ffffffff811c4fa4>] flush_work+0x3a4/0x5e0
[   52.767837]
               but task is already holding lock:
[   52.774360]  (mlxsw_core_driver_name){+.+.}, at: [<ffffffff811c65c4>] process_one_work+0x7d4/0x12f0
[   52.784495]
               other info that might help us debug this:
[   52.791794]  Possible unsafe locking scenario:
[   52.798413]        CPU0
[   52.801144]        ----
[   52.803875]   lock(mlxsw_core_driver_name);
[   52.808556]   lock(mlxsw_core_driver_name);
[   52.813236]
                *** DEADLOCK ***
[   52.819857]  May be due to missing lock nesting notation
[   52.827450] 3 locks held by kworker/1:3/599:
[   52.832221]  #0:  (mlxsw_core_driver_name){+.+.}, at: [<ffffffff811c65c4>] process_one_work+0x7d4/0x12f0
[   52.842846]  #1:  ((&(&bridge->fdb_notify.dw)->work)){+.+.}, at: [<ffffffff811c65c4>] process_one_work+0x7d4/0x12f0
[   52.854537]  #2:  (rtnl_mutex){+.+.}, at: [<ffffffff822ad8e7>] rtnl_lock+0x17/0x20
[   52.863021]
               stack backtrace:
[   52.867890] CPU: 1 PID: 599 Comm: kworker/1:3 Not tainted 4.14.0-rc3jiri+ #4
[   52.875773] Hardware name: Mellanox Technologies Ltd. "MSN2100-CB2F"/"SA001017", BIOS 5.6.5 06/07/2016
[   52.886267] Workqueue: mlxsw_core mlxsw_sp_fdb_notify_work [mlxsw_spectrum]
[   52.894060] Call Trace:
[   52.909122]  __lock_acquire+0xf6f/0x2a10
[   53.025412]  lock_acquire+0x158/0x440
[   53.047557]  flush_work+0x3c4/0x5e0
[   53.087571]  __cancel_work_timer+0x3ca/0x5e0
[   53.177051]  cancel_delayed_work_sync+0x13/0x20
[   53.182142]  mlxsw_reg_trans_bulk_wait+0x12d/0x7a0 [mlxsw_core]
[   53.194571]  mlxsw_core_reg_access+0x586/0x990 [mlxsw_core]
[   53.225365]  mlxsw_reg_query+0x10/0x20 [mlxsw_core]
[   53.230882]  mlxsw_sp_fdb_notify_work+0x2a3/0x9d0 [mlxsw_spectrum]
[   53.237801]  process_one_work+0x8f1/0x12f0
[   53.321804]  worker_thread+0x1fd/0x10c0
[   53.435158]  kthread+0x28e/0x370
[   53.448703]  ret_from_fork+0x2a/0x40
[   53.453017] mlxsw_spectrum 0000:01:00.0: EMAD retries (2/5) (tid=bf4549b100000774)
[   53.453119] mlxsw_spectrum 0000:01:00.0: EMAD retries (5/5) (tid=bf4549b100000770)
[   53.453132] mlxsw_spectrum 0000:01:00.0: EMAD reg access failed (tid=bf4549b100000770,reg_id=200b(sfn),type=query,status=0(operation performed))
[   53.453143] mlxsw_spectrum 0000:01:00.0: Failed to get FDB notifications

Fix this by creating another workqueue for EMAD timeouts, thereby
preventing the situation of a work item trying to flush a work item
queued on the same workqueue.

Fixes: caf7297e7ab5f ("mlxsw: core: Introduce support for asynchronous EMAD register access")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reported-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/core.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c
index 9d5e7cf..f3315bc 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/core.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/core.c
@@ -96,6 +96,7 @@ struct mlxsw_core {
 	const struct mlxsw_bus *bus;
 	void *bus_priv;
 	const struct mlxsw_bus_info *bus_info;
+	struct workqueue_struct *emad_wq;
 	struct list_head rx_listener_list;
 	struct list_head event_listener_list;
 	struct {
@@ -465,7 +466,7 @@ static void mlxsw_emad_trans_timeout_schedule(struct mlxsw_reg_trans *trans)
 {
 	unsigned long timeout = msecs_to_jiffies(MLXSW_EMAD_TIMEOUT_MS);
 
-	mlxsw_core_schedule_dw(&trans->timeout_dw, timeout);
+	queue_delayed_work(trans->core->emad_wq, &trans->timeout_dw, timeout);
 }
 
 static int mlxsw_emad_transmit(struct mlxsw_core *mlxsw_core,
@@ -587,12 +588,18 @@ static const struct mlxsw_listener mlxsw_emad_rx_listener =
 
 static int mlxsw_emad_init(struct mlxsw_core *mlxsw_core)
 {
+	struct workqueue_struct *emad_wq;
 	u64 tid;
 	int err;
 
 	if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX))
 		return 0;
 
+	emad_wq = alloc_workqueue("mlxsw_core_emad", WQ_MEM_RECLAIM, 0);
+	if (!emad_wq)
+		return -ENOMEM;
+	mlxsw_core->emad_wq = emad_wq;
+
 	/* Set the upper 32 bits of the transaction ID field to a random
 	 * number. This allows us to discard EMADs addressed to other
 	 * devices.
@@ -619,6 +626,7 @@ static int mlxsw_emad_init(struct mlxsw_core *mlxsw_core)
 err_emad_trap_set:
 	mlxsw_core_trap_unregister(mlxsw_core, &mlxsw_emad_rx_listener,
 				   mlxsw_core);
+	destroy_workqueue(mlxsw_core->emad_wq);
 	return err;
 }
 
@@ -631,6 +639,7 @@ static void mlxsw_emad_fini(struct mlxsw_core *mlxsw_core)
 	mlxsw_core->emad.use_emad = false;
 	mlxsw_core_trap_unregister(mlxsw_core, &mlxsw_emad_rx_listener,
 				   mlxsw_core);
+	destroy_workqueue(mlxsw_core->emad_wq);
 }
 
 static struct sk_buff *mlxsw_emad_alloc(const struct mlxsw_core *mlxsw_core,
-- 
cgit v1.1


From 14aefd9011f14ecf1f821fcd1754f009f4ab3df9 Mon Sep 17 00:00:00 2001
From: Petr Machata <petrm@mellanox.com>
Date: Fri, 20 Oct 2017 09:16:15 +0200
Subject: mlxsw: reg: Add Tunneling IPinIP General Configuration Register

The TIGCR register is used for setting up the IPinIP Tunnel
configuration.

Fixes: ee954d1a91b2 ("mlxsw: spectrum_router: Support GRE tunnels")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/reg.h | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index cc27c5d..4afc848 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -6401,6 +6401,36 @@ static inline void mlxsw_reg_mgpc_pack(char *payload, u32 counter_index,
 	mlxsw_reg_mgpc_opcode_set(payload, opcode);
 }
 
+/* TIGCR - Tunneling IPinIP General Configuration Register
+ * -------------------------------------------------------
+ * The TIGCR register is used for setting up the IPinIP Tunnel configuration.
+ */
+#define MLXSW_REG_TIGCR_ID 0xA801
+#define MLXSW_REG_TIGCR_LEN 0x10
+
+MLXSW_REG_DEFINE(tigcr, MLXSW_REG_TIGCR_ID, MLXSW_REG_TIGCR_LEN);
+
+/* reg_tigcr_ipip_ttlc
+ * For IPinIP Tunnel encapsulation: whether to copy the ttl from the packet
+ * header.
+ * Access: RW
+ */
+MLXSW_ITEM32(reg, tigcr, ttlc, 0x04, 8, 1);
+
+/* reg_tigcr_ipip_ttl_uc
+ * The TTL for IPinIP Tunnel encapsulation of unicast packets if
+ * reg_tigcr_ipip_ttlc is unset.
+ * Access: RW
+ */
+MLXSW_ITEM32(reg, tigcr, ttl_uc, 0x04, 0, 8);
+
+static inline void mlxsw_reg_tigcr_pack(char *payload, bool ttlc, u8 ttl_uc)
+{
+	MLXSW_REG_ZERO(tigcr, payload);
+	mlxsw_reg_tigcr_ttlc_set(payload, ttlc);
+	mlxsw_reg_tigcr_ttl_uc_set(payload, ttl_uc);
+}
+
 /* SBPR - Shared Buffer Pools Register
  * -----------------------------------
  * The SBPR configures and retrieves the shared buffer pools and configuration.
@@ -6881,6 +6911,7 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = {
 	MLXSW_REG(mcc),
 	MLXSW_REG(mcda),
 	MLXSW_REG(mgpc),
+	MLXSW_REG(tigcr),
 	MLXSW_REG(sbpr),
 	MLXSW_REG(sbcm),
 	MLXSW_REG(sbpm),
-- 
cgit v1.1


From dcbda2820ff91a692338fed2c99bb9b1af37a05a Mon Sep 17 00:00:00 2001
From: Petr Machata <petrm@mellanox.com>
Date: Fri, 20 Oct 2017 09:16:16 +0200
Subject: mlxsw: spectrum_router: Configure TIGCR on init

Spectrum tunnels do not default to ttl of "inherit" like the Linux ones
do. Configure TIGCR on router init so that the TTL of tunnel packets is
copied from the overlay packets.

Fixes: ee954d1a91b2 ("mlxsw: spectrum_router: Support GRE tunnels")
Signed-off-by: Petr Machata <petrm@mellanox.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index c16718d..5189022a 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -5896,11 +5896,20 @@ static void mlxsw_sp_rifs_fini(struct mlxsw_sp *mlxsw_sp)
 	kfree(mlxsw_sp->router->rifs);
 }
 
+static int
+mlxsw_sp_ipip_config_tigcr(struct mlxsw_sp *mlxsw_sp)
+{
+	char tigcr_pl[MLXSW_REG_TIGCR_LEN];
+
+	mlxsw_reg_tigcr_pack(tigcr_pl, true, 0);
+	return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tigcr), tigcr_pl);
+}
+
 static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp)
 {
 	mlxsw_sp->router->ipip_ops_arr = mlxsw_sp_ipip_ops_arr;
 	INIT_LIST_HEAD(&mlxsw_sp->router->ipip_list);
-	return 0;
+	return mlxsw_sp_ipip_config_tigcr(mlxsw_sp);
 }
 
 static void mlxsw_sp_ipips_fini(struct mlxsw_sp *mlxsw_sp)
-- 
cgit v1.1


From 6377ed0bbae6fa28853e1679d068a9106c8a8908 Mon Sep 17 00:00:00 2001
From: Moshe Shemesh <moshe@mellanox.com>
Date: Thu, 19 Oct 2017 14:14:29 +0300
Subject: net/mlx5: Fix health work queue spin lock to IRQ safe

spin_lock/unlock of health->wq_lock should be IRQ safe.
It was changed to spin_lock_irqsave since adding commit 0179720d6be2
("net/mlx5: Introduce trigger_health_work function") which uses
spin_lock from asynchronous event (IRQ) context.
Thus, all spin_lock/unlock of health->wq_lock should have been moved
to IRQ safe mode.
However, one occurrence on new code using this lock missed that
change, resulting in possible deadlock:
  kernel: Possible unsafe locking scenario:
  kernel:       CPU0
  kernel:       ----
  kernel:  lock(&(&health->wq_lock)->rlock);
  kernel:  <Interrupt>
  kernel:    lock(&(&health->wq_lock)->rlock);
  kernel: #012 *** DEADLOCK ***

Fixes: 2a0165a034ac ("net/mlx5: Cancel delayed recovery work when unloading the driver")
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/health.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c
index 8aea0a0..db86e15 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c
@@ -356,10 +356,11 @@ void mlx5_drain_health_wq(struct mlx5_core_dev *dev)
 void mlx5_drain_health_recovery(struct mlx5_core_dev *dev)
 {
 	struct mlx5_core_health *health = &dev->priv.health;
+	unsigned long flags;
 
-	spin_lock(&health->wq_lock);
+	spin_lock_irqsave(&health->wq_lock, flags);
 	set_bit(MLX5_DROP_NEW_RECOVERY_WORK, &health->flags);
-	spin_unlock(&health->wq_lock);
+	spin_unlock_irqrestore(&health->wq_lock, flags);
 	cancel_delayed_work_sync(&dev->priv.health.recover_work);
 }
 
-- 
cgit v1.1


From 4ca637a20a524cd8ddbca696f12bfa92111c96e3 Mon Sep 17 00:00:00 2001
From: Huy Nguyen <huyn@mellanox.com>
Date: Wed, 4 Oct 2017 17:58:21 -0500
Subject: net/mlx5: Delay events till mlx5 interface's add complete for pci
 resume

mlx5_ib_add is called during mlx5_pci_resume after a pci error.
Before mlx5_ib_add completes, there are multiple events which trigger
function mlx5_ib_event. This cause kernel panic because mlx5_ib_event
accesses unitialized resources.

The fix is to extend Erez Shitrit's patch <97834eba7c19>
("net/mlx5: Delay events till ib registration ends") to cover
the pci resume code path.

Trace:
mlx5_core 0001:01:00.6: mlx5_pci_resume was called
mlx5_core 0001:01:00.6: firmware version: 16.20.1011
mlx5_core 0001:01:00.6: mlx5_attach_interface:164:(pid 779):
mlx5_ib_event:2996:(pid 34777): warning: event on port 1
mlx5_ib_event:2996:(pid 34782): warning: event on port 1
Unable to handle kernel paging request for data at address 0x0001c104
Faulting instruction address: 0xd000000008f411fc
Oops: Kernel access of bad area, sig: 11 [#1]
...
...
Call Trace:
[c000000fff77bb70] [d000000008f4119c] mlx5_ib_event+0x64/0x470 [mlx5_ib] (unreliable)
[c000000fff77bc60] [d000000008e67130] mlx5_core_event+0xb8/0x210 [mlx5_core]
[c000000fff77bd10] [d000000008e4bd00] mlx5_eq_int+0x528/0x860[mlx5_core]

Fixes: 97834eba7c19 ("net/mlx5: Delay events till ib registration ends")
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/dev.c | 70 ++++++++++++++++-----------
 1 file changed, 41 insertions(+), 29 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/dev.c b/drivers/net/ethernet/mellanox/mlx5/core/dev.c
index ff60cf7..fc28171 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/dev.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/dev.c
@@ -77,35 +77,41 @@ static void add_delayed_event(struct mlx5_priv *priv,
 	list_add_tail(&delayed_event->list, &priv->waiting_events_list);
 }
 
-static void fire_delayed_event_locked(struct mlx5_device_context *dev_ctx,
-				      struct mlx5_core_dev *dev,
-				      struct mlx5_priv *priv)
+static void delayed_event_release(struct mlx5_device_context *dev_ctx,
+				  struct mlx5_priv *priv)
 {
+	struct mlx5_core_dev *dev = container_of(priv, struct mlx5_core_dev, priv);
 	struct mlx5_delayed_event *de;
 	struct mlx5_delayed_event *n;
+	struct list_head temp;
 
-	/* stop delaying events */
-	priv->is_accum_events = false;
+	INIT_LIST_HEAD(&temp);
+
+	spin_lock_irq(&priv->ctx_lock);
 
-	/* fire all accumulated events before new event comes */
-	list_for_each_entry_safe(de, n, &priv->waiting_events_list, list) {
+	priv->is_accum_events = false;
+	list_splice_init(&priv->waiting_events_list, &temp);
+	if (!dev_ctx->context)
+		goto out;
+	list_for_each_entry_safe(de, n, &priv->waiting_events_list, list)
 		dev_ctx->intf->event(dev, dev_ctx->context, de->event, de->param);
+
+out:
+	spin_unlock_irq(&priv->ctx_lock);
+
+	list_for_each_entry_safe(de, n, &temp, list) {
 		list_del(&de->list);
 		kfree(de);
 	}
 }
 
-static void cleanup_delayed_evets(struct mlx5_priv *priv)
+/* accumulating events that can come after mlx5_ib calls to
+ * ib_register_device, till adding that interface to the events list.
+ */
+static void delayed_event_start(struct mlx5_priv *priv)
 {
-	struct mlx5_delayed_event *de;
-	struct mlx5_delayed_event *n;
-
 	spin_lock_irq(&priv->ctx_lock);
-	priv->is_accum_events = false;
-	list_for_each_entry_safe(de, n, &priv->waiting_events_list, list) {
-		list_del(&de->list);
-		kfree(de);
-	}
+	priv->is_accum_events = true;
 	spin_unlock_irq(&priv->ctx_lock);
 }
 
@@ -122,11 +128,8 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv)
 		return;
 
 	dev_ctx->intf = intf;
-	/* accumulating events that can come after mlx5_ib calls to
-	 * ib_register_device, till adding that interface to the events list.
-	 */
 
-	priv->is_accum_events = true;
+	delayed_event_start(priv);
 
 	dev_ctx->context = intf->add(dev);
 	set_bit(MLX5_INTERFACE_ADDED, &dev_ctx->state);
@@ -137,8 +140,6 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv)
 		spin_lock_irq(&priv->ctx_lock);
 		list_add_tail(&dev_ctx->list, &priv->ctx_list);
 
-		fire_delayed_event_locked(dev_ctx, dev, priv);
-
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
 		if (dev_ctx->intf->pfault) {
 			if (priv->pfault) {
@@ -150,11 +151,12 @@ void mlx5_add_device(struct mlx5_interface *intf, struct mlx5_priv *priv)
 		}
 #endif
 		spin_unlock_irq(&priv->ctx_lock);
-	} else {
-		kfree(dev_ctx);
-		 /* delete all accumulated events */
-		cleanup_delayed_evets(priv);
 	}
+
+	delayed_event_release(dev_ctx, priv);
+
+	if (!dev_ctx->context)
+		kfree(dev_ctx);
 }
 
 static struct mlx5_device_context *mlx5_get_device(struct mlx5_interface *intf,
@@ -205,17 +207,21 @@ static void mlx5_attach_interface(struct mlx5_interface *intf, struct mlx5_priv
 	if (!dev_ctx)
 		return;
 
+	delayed_event_start(priv);
 	if (intf->attach) {
 		if (test_bit(MLX5_INTERFACE_ATTACHED, &dev_ctx->state))
-			return;
+			goto out;
 		intf->attach(dev, dev_ctx->context);
 		set_bit(MLX5_INTERFACE_ATTACHED, &dev_ctx->state);
 	} else {
 		if (test_bit(MLX5_INTERFACE_ADDED, &dev_ctx->state))
-			return;
+			goto out;
 		dev_ctx->context = intf->add(dev);
 		set_bit(MLX5_INTERFACE_ADDED, &dev_ctx->state);
 	}
+
+out:
+	delayed_event_release(dev_ctx, priv);
 }
 
 void mlx5_attach_device(struct mlx5_core_dev *dev)
@@ -414,8 +420,14 @@ void mlx5_core_event(struct mlx5_core_dev *dev, enum mlx5_dev_event event,
 	if (priv->is_accum_events)
 		add_delayed_event(priv, dev, event, param);
 
+	/* After mlx5_detach_device, the dev_ctx->intf is still set and dev_ctx is
+	 * still in priv->ctx_list. In this case, only notify the dev_ctx if its
+	 * ADDED or ATTACHED bit are set.
+	 */
 	list_for_each_entry(dev_ctx, &priv->ctx_list, list)
-		if (dev_ctx->intf->event)
+		if (dev_ctx->intf->event &&
+		    (test_bit(MLX5_INTERFACE_ADDED, &dev_ctx->state) ||
+		     test_bit(MLX5_INTERFACE_ATTACHED, &dev_ctx->state)))
 			dev_ctx->intf->event(dev, dev_ctx->context, event, param);
 
 	spin_unlock_irqrestore(&priv->ctx_lock, flags);
-- 
cgit v1.1


From 3c37745ec614ff048d5dce38f976804b05d307ee Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Tue, 17 Oct 2017 12:33:43 +0200
Subject: net/mlx5e: Properly deal with encap flows add/del under neigh update

Currently, the encap action offload is handled in the actions parse
function and not in mlx5e_tc_add_fdb_flow() where we deal with all
the other aspects of offloading actions (vlan, modify header) and
the rule itself.

When the neigh update code (mlx5e_tc_encap_flows_add()) recreates the
encap entry and offloads the related flows, we wrongly call again into
mlx5e_tc_add_fdb_flow(), this for itself would cause us to handle
again the offloading of vlans and header re-write which puts things
in non consistent state and step on freed memory (e.g the modify
header parse buffer which is already freed).

Since on error, mlx5e_tc_add_fdb_flow() detaches and may release the
encap entry, it causes a corruption at the neigh update code which goes
over the list of flows associated with this encap entry, or double free
when the tc flow is later deleted by user-space.

When neigh update (mlx5e_tc_encap_flows_del()) unoffloads the flows related
to an encap entry which is now invalid, we do a partial repeat of the eswitch
flow removal code which is wrong too.

To fix things up we do the following:

(1) handle the encap action offload in the eswitch flow add function
    mlx5e_tc_add_fdb_flow() as done for the other actions and the rule itself.

(2) modify the neigh update code (mlx5e_tc_encap_flows_add/del) to only
    deal with the encap entry and rules delete/add and not with any of
    the other offloaded actions.

Fixes: 232c001398ae ('net/mlx5e: Add support to neighbour update flow')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 89 +++++++++++++++----------
 1 file changed, 54 insertions(+), 35 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
index 1aa2028..9ba1f720 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c
@@ -78,9 +78,11 @@ struct mlx5e_tc_flow {
 };
 
 struct mlx5e_tc_flow_parse_attr {
+	struct ip_tunnel_info tun_info;
 	struct mlx5_flow_spec spec;
 	int num_mod_hdr_actions;
 	void *mod_hdr_actions;
+	int mirred_ifindex;
 };
 
 enum {
@@ -322,6 +324,12 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv,
 static void mlx5e_detach_encap(struct mlx5e_priv *priv,
 			       struct mlx5e_tc_flow *flow);
 
+static int mlx5e_attach_encap(struct mlx5e_priv *priv,
+			      struct ip_tunnel_info *tun_info,
+			      struct net_device *mirred_dev,
+			      struct net_device **encap_dev,
+			      struct mlx5e_tc_flow *flow);
+
 static struct mlx5_flow_handle *
 mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
 		      struct mlx5e_tc_flow_parse_attr *parse_attr,
@@ -329,9 +337,27 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
 {
 	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
 	struct mlx5_esw_flow_attr *attr = flow->esw_attr;
-	struct mlx5_flow_handle *rule;
+	struct net_device *out_dev, *encap_dev = NULL;
+	struct mlx5_flow_handle *rule = NULL;
+	struct mlx5e_rep_priv *rpriv;
+	struct mlx5e_priv *out_priv;
 	int err;
 
+	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_ENCAP) {
+		out_dev = __dev_get_by_index(dev_net(priv->netdev),
+					     attr->parse_attr->mirred_ifindex);
+		err = mlx5e_attach_encap(priv, &parse_attr->tun_info,
+					 out_dev, &encap_dev, flow);
+		if (err) {
+			rule = ERR_PTR(err);
+			if (err != -EAGAIN)
+				goto err_attach_encap;
+		}
+		out_priv = netdev_priv(encap_dev);
+		rpriv = out_priv->ppriv;
+		attr->out_rep = rpriv->rep;
+	}
+
 	err = mlx5_eswitch_add_vlan_action(esw, attr);
 	if (err) {
 		rule = ERR_PTR(err);
@@ -347,10 +373,14 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv,
 		}
 	}
 
-	rule = mlx5_eswitch_add_offloaded_rule(esw, &parse_attr->spec, attr);
-	if (IS_ERR(rule))
-		goto err_add_rule;
-
+	/* we get here if (1) there's no error (rule being null) or when
+	 * (2) there's an encap action and we're on -EAGAIN (no valid neigh)
+	 */
+	if (rule != ERR_PTR(-EAGAIN)) {
+		rule = mlx5_eswitch_add_offloaded_rule(esw, &parse_attr->spec, attr);
+		if (IS_ERR(rule))
+			goto err_add_rule;
+	}
 	return rule;
 
 err_add_rule:
@@ -361,6 +391,7 @@ err_mod_hdr:
 err_add_vlan:
 	if (attr->action & MLX5_FLOW_CONTEXT_ACTION_ENCAP)
 		mlx5e_detach_encap(priv, flow);
+err_attach_encap:
 	return rule;
 }
 
@@ -389,6 +420,8 @@ static void mlx5e_tc_del_fdb_flow(struct mlx5e_priv *priv,
 void mlx5e_tc_encap_flows_add(struct mlx5e_priv *priv,
 			      struct mlx5e_encap_entry *e)
 {
+	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
+	struct mlx5_esw_flow_attr *esw_attr;
 	struct mlx5e_tc_flow *flow;
 	int err;
 
@@ -404,10 +437,9 @@ void mlx5e_tc_encap_flows_add(struct mlx5e_priv *priv,
 	mlx5e_rep_queue_neigh_stats_work(priv);
 
 	list_for_each_entry(flow, &e->flows, encap) {
-		flow->esw_attr->encap_id = e->encap_id;
-		flow->rule = mlx5e_tc_add_fdb_flow(priv,
-						   flow->esw_attr->parse_attr,
-						   flow);
+		esw_attr = flow->esw_attr;
+		esw_attr->encap_id = e->encap_id;
+		flow->rule = mlx5_eswitch_add_offloaded_rule(esw, &esw_attr->parse_attr->spec, esw_attr);
 		if (IS_ERR(flow->rule)) {
 			err = PTR_ERR(flow->rule);
 			mlx5_core_warn(priv->mdev, "Failed to update cached encapsulation flow, %d\n",
@@ -421,15 +453,13 @@ void mlx5e_tc_encap_flows_add(struct mlx5e_priv *priv,
 void mlx5e_tc_encap_flows_del(struct mlx5e_priv *priv,
 			      struct mlx5e_encap_entry *e)
 {
+	struct mlx5_eswitch *esw = priv->mdev->priv.eswitch;
 	struct mlx5e_tc_flow *flow;
-	struct mlx5_fc *counter;
 
 	list_for_each_entry(flow, &e->flows, encap) {
 		if (flow->flags & MLX5E_TC_FLOW_OFFLOADED) {
 			flow->flags &= ~MLX5E_TC_FLOW_OFFLOADED;
-			counter = mlx5_flow_rule_counter(flow->rule);
-			mlx5_del_flow_rules(flow->rule);
-			mlx5_fc_destroy(priv->mdev, counter);
+			mlx5_eswitch_del_offloaded_rule(esw, flow->rule, flow->esw_attr);
 		}
 	}
 
@@ -1942,7 +1972,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 
 		if (is_tcf_mirred_egress_redirect(a)) {
 			int ifindex = tcf_mirred_ifindex(a);
-			struct net_device *out_dev, *encap_dev = NULL;
+			struct net_device *out_dev;
 			struct mlx5e_priv *out_priv;
 
 			out_dev = __dev_get_by_index(dev_net(priv->netdev), ifindex);
@@ -1955,17 +1985,13 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts,
 				rpriv = out_priv->ppriv;
 				attr->out_rep = rpriv->rep;
 			} else if (encap) {
-				err = mlx5e_attach_encap(priv, info,
-							 out_dev, &encap_dev, flow);
-				if (err && err != -EAGAIN)
-					return err;
+				parse_attr->mirred_ifindex = 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;
-				out_priv = netdev_priv(encap_dev);
-				rpriv = out_priv->ppriv;
-				attr->out_rep = rpriv->rep;
-				attr->parse_attr = parse_attr;
+				/* 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",
 				       priv->netdev->name, out_dev->name);
@@ -2047,7 +2073,7 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
 	if (flow->flags & MLX5E_TC_FLOW_ESWITCH) {
 		err = parse_tc_fdb_actions(priv, f->exts, parse_attr, flow);
 		if (err < 0)
-			goto err_handle_encap_flow;
+			goto err_free;
 		flow->rule = mlx5e_tc_add_fdb_flow(priv, parse_attr, flow);
 	} else {
 		err = parse_tc_nic_actions(priv, f->exts, parse_attr, flow);
@@ -2058,10 +2084,13 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
 
 	if (IS_ERR(flow->rule)) {
 		err = PTR_ERR(flow->rule);
-		goto err_free;
+		if (err != -EAGAIN)
+			goto err_free;
 	}
 
-	flow->flags |= MLX5E_TC_FLOW_OFFLOADED;
+	if (err != -EAGAIN)
+		flow->flags |= MLX5E_TC_FLOW_OFFLOADED;
+
 	err = rhashtable_insert_fast(&tc->ht, &flow->node,
 				     tc->ht_params);
 	if (err)
@@ -2075,16 +2104,6 @@ int mlx5e_configure_flower(struct mlx5e_priv *priv,
 err_del_rule:
 	mlx5e_tc_del_flow(priv, flow);
 
-err_handle_encap_flow:
-	if (err == -EAGAIN) {
-		err = rhashtable_insert_fast(&tc->ht, &flow->node,
-					     tc->ht_params);
-		if (err)
-			mlx5e_tc_del_flow(priv, flow);
-		else
-			return 0;
-	}
-
 err_free:
 	kvfree(parse_attr);
 	kfree(flow);
-- 
cgit v1.1


From be0f161ef141e4df368aa3f417a1c2ab9c362e75 Mon Sep 17 00:00:00 2001
From: Huy Nguyen <huyn@mellanox.com>
Date: Thu, 28 Sep 2017 15:33:50 -0500
Subject: net/mlx5e: DCBNL, Implement tc with ets type and zero bandwidth

Previously, tc with ets type and zero bandwidth is not accepted
by driver. This behavior does not follow the IEEE802.1qaz spec.

If there are tcs with ets type and zero bandwidth, these tcs are
assigned to the lowest priority tc_group #0. We equally distribute
100% bw of the tc_group #0 to these zero bandwidth ets tcs.
Also, the non zero bandwidth ets tcs are assigned to tc_group #1.

If there is no zero bandwidth ets tc, the non zero bandwidth ets tcs
are assigned to tc_group #0.

Fixes: cdcf11212b22 ("net/mlx5e: Validate BW weight values of ETS")
Signed-off-by: Huy Nguyen <huyn@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c | 113 +++++++++++++++------
 drivers/net/ethernet/mellanox/mlx5/core/port.c     |  21 ++++
 2 files changed, 104 insertions(+), 30 deletions(-)

(limited to 'drivers/net/ethernet/mellanox')

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
index c1d384f..51c4cc0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_dcbnl.c
@@ -41,6 +41,11 @@
 #define MLX5E_CEE_STATE_UP    1
 #define MLX5E_CEE_STATE_DOWN  0
 
+enum {
+	MLX5E_VENDOR_TC_GROUP_NUM = 7,
+	MLX5E_LOWEST_PRIO_GROUP   = 0,
+};
+
 /* If dcbx mode is non-host set the dcbx mode to host.
  */
 static int mlx5e_dcbnl_set_dcbx_mode(struct mlx5e_priv *priv,
@@ -85,6 +90,9 @@ static int mlx5e_dcbnl_ieee_getets(struct net_device *netdev,
 {
 	struct mlx5e_priv *priv = netdev_priv(netdev);
 	struct mlx5_core_dev *mdev = priv->mdev;
+	u8 tc_group[IEEE_8021QAZ_MAX_TCS];
+	bool is_tc_group_6_exist = false;
+	bool is_zero_bw_ets_tc = false;
 	int err = 0;
 	int i;
 
@@ -96,37 +104,64 @@ static int mlx5e_dcbnl_ieee_getets(struct net_device *netdev,
 		err = mlx5_query_port_prio_tc(mdev, i, &ets->prio_tc[i]);
 		if (err)
 			return err;
-	}
 
-	for (i = 0; i < ets->ets_cap; i++) {
+		err = mlx5_query_port_tc_group(mdev, i, &tc_group[i]);
+		if (err)
+			return err;
+
 		err = mlx5_query_port_tc_bw_alloc(mdev, i, &ets->tc_tx_bw[i]);
 		if (err)
 			return err;
+
+		if (ets->tc_tx_bw[i] < MLX5E_MAX_BW_ALLOC &&
+		    tc_group[i] == (MLX5E_LOWEST_PRIO_GROUP + 1))
+			is_zero_bw_ets_tc = true;
+
+		if (tc_group[i] == (MLX5E_VENDOR_TC_GROUP_NUM - 1))
+			is_tc_group_6_exist = true;
+	}
+
+	/* Report 0% ets tc if exits*/
+	if (is_zero_bw_ets_tc) {
+		for (i = 0; i < ets->ets_cap; i++)
+			if (tc_group[i] == MLX5E_LOWEST_PRIO_GROUP)
+				ets->tc_tx_bw[i] = 0;
+	}
+
+	/* Update tc_tsa based on fw setting*/
+	for (i = 0; i < ets->ets_cap; i++) {
 		if (ets->tc_tx_bw[i] < MLX5E_MAX_BW_ALLOC)
 			priv->dcbx.tc_tsa[i] = IEEE_8021QAZ_TSA_ETS;
+		else if (tc_group[i] == MLX5E_VENDOR_TC_GROUP_NUM &&
+			 !is_tc_group_6_exist)
+			priv->dcbx.tc_tsa[i] = IEEE_8021QAZ_TSA_VENDOR;
 	}
-
 	memcpy(ets->tc_tsa, priv->dcbx.tc_tsa, sizeof(ets->tc_tsa));
 
 	return err;
 }
 
-enum {
-	MLX5E_VENDOR_TC_GROUP_NUM = 7,
-	MLX5E_ETS_TC_GROUP_NUM    = 0,
-};
-
 static void mlx5e_build_tc_group(struct ieee_ets *ets, u8 *tc_group, int max_tc)
 {
 	bool any_tc_mapped_to_ets = false;
+	bool ets_zero_bw = false;
 	int strict_group;
 	int i;
 
-	for (i = 0; i <= max_tc; i++)
-		if (ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS)
+	for (i = 0; i <= max_tc; i++) {
+		if (ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS) {
 			any_tc_mapped_to_ets = true;
+			if (!ets->tc_tx_bw[i])
+				ets_zero_bw = true;
+		}
+	}
 
-	strict_group = any_tc_mapped_to_ets ? 1 : 0;
+	/* strict group has higher priority than ets group */
+	strict_group = MLX5E_LOWEST_PRIO_GROUP;
+	if (any_tc_mapped_to_ets)
+		strict_group++;
+	if (ets_zero_bw)
+		strict_group++;
 
 	for (i = 0; i <= max_tc; i++) {
 		switch (ets->tc_tsa[i]) {
@@ -137,7 +172,9 @@ static void mlx5e_build_tc_group(struct ieee_ets *ets, u8 *tc_group, int max_tc)
 			tc_group[i] = strict_group++;
 			break;
 		case IEEE_8021QAZ_TSA_ETS:
-			tc_group[i] = MLX5E_ETS_TC_GROUP_NUM;
+			tc_group[i] = MLX5E_LOWEST_PRIO_GROUP;
+			if (ets->tc_tx_bw[i] && ets_zero_bw)
+				tc_group[i] = MLX5E_LOWEST_PRIO_GROUP + 1;
 			break;
 		}
 	}
@@ -146,9 +183,23 @@ static void mlx5e_build_tc_group(struct ieee_ets *ets, u8 *tc_group, int max_tc)
 static void mlx5e_build_tc_tx_bw(struct ieee_ets *ets, u8 *tc_tx_bw,
 				 u8 *tc_group, int max_tc)
 {
+	int bw_for_ets_zero_bw_tc = 0;
+	int last_ets_zero_bw_tc = -1;
+	int num_ets_zero_bw = 0;
 	int i;
 
 	for (i = 0; i <= max_tc; i++) {
+		if (ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS &&
+		    !ets->tc_tx_bw[i]) {
+			num_ets_zero_bw++;
+			last_ets_zero_bw_tc = i;
+		}
+	}
+
+	if (num_ets_zero_bw)
+		bw_for_ets_zero_bw_tc = MLX5E_MAX_BW_ALLOC / num_ets_zero_bw;
+
+	for (i = 0; i <= max_tc; i++) {
 		switch (ets->tc_tsa[i]) {
 		case IEEE_8021QAZ_TSA_VENDOR:
 			tc_tx_bw[i] = MLX5E_MAX_BW_ALLOC;
@@ -157,12 +208,26 @@ static void mlx5e_build_tc_tx_bw(struct ieee_ets *ets, u8 *tc_tx_bw,
 			tc_tx_bw[i] = MLX5E_MAX_BW_ALLOC;
 			break;
 		case IEEE_8021QAZ_TSA_ETS:
-			tc_tx_bw[i] = ets->tc_tx_bw[i];
+			tc_tx_bw[i] = ets->tc_tx_bw[i] ?
+				      ets->tc_tx_bw[i] :
+				      bw_for_ets_zero_bw_tc;
 			break;
 		}
 	}
+
+	/* Make sure the total bw for ets zero bw group is 100% */
+	if (last_ets_zero_bw_tc != -1)
+		tc_tx_bw[last_ets_zero_bw_tc] +=
+			MLX5E_MAX_BW_ALLOC % num_ets_zero_bw;
 }
 
+/* If there are ETS BW 0,
+ *   Set ETS group # to 1 for all ETS non zero BW tcs. Their sum must be 100%.
+ *   Set group #0 to all the ETS BW 0 tcs and
+ *     equally splits the 100% BW between them
+ *   Report both group #0 and #1 as ETS type.
+ *     All the tcs in group #0 will be reported with 0% BW.
+ */
 int mlx5e_dcbnl_ieee_setets_core(struct mlx5e_priv *priv, struct ieee_ets *ets)
 {
 	struct mlx5_core_dev *mdev = priv->mdev;
@@ -188,7 +253,6 @@ int mlx5e_dcbnl_ieee_setets_core(struct mlx5e_priv *priv, struct ieee_ets *ets)
 		return err;
 
 	memcpy(priv->dcbx.tc_tsa, ets->tc_tsa, sizeof(ets->tc_tsa));
-
 	return err;
 }
 
@@ -209,17 +273,9 @@ static int mlx5e_dbcnl_validate_ets(struct net_device *netdev,
 	}
 
 	/* Validate Bandwidth Sum */
-	for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) {
-		if (ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS) {
-			if (!ets->tc_tx_bw[i]) {
-				netdev_err(netdev,
-					   "Failed to validate ETS: BW 0 is illegal\n");
-				return -EINVAL;
-			}
-
+	for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++)
+		if (ets->tc_tsa[i] == IEEE_8021QAZ_TSA_ETS)
 			bw_sum += ets->tc_tx_bw[i];
-		}
-	}
 
 	if (bw_sum != 0 && bw_sum != 100) {
 		netdev_err(netdev,
@@ -533,8 +589,7 @@ static void mlx5e_dcbnl_getpgtccfgtx(struct net_device *netdev,
 static void mlx5e_dcbnl_getpgbwgcfgtx(struct net_device *netdev,
 				      int pgid, u8 *bw_pct)
 {
-	struct mlx5e_priv *priv = netdev_priv(netdev);
-	struct mlx5_core_dev *mdev = priv->mdev;
+	struct ieee_ets ets;
 
 	if (pgid >= CEE_DCBX_MAX_PGS) {
 		netdev_err(netdev,
@@ -542,8 +597,8 @@ static void mlx5e_dcbnl_getpgbwgcfgtx(struct net_device *netdev,
 		return;
 	}
 
-	if (mlx5_query_port_tc_bw_alloc(mdev, pgid, bw_pct))
-		*bw_pct = 0;
+	mlx5e_dcbnl_ieee_getets(netdev, &ets);
+	*bw_pct = ets.tc_tx_bw[pgid];
 }
 
 static void mlx5e_dcbnl_setpfccfg(struct net_device *netdev,
@@ -739,8 +794,6 @@ static void mlx5e_ets_init(struct mlx5e_priv *priv)
 		ets.prio_tc[i] = i;
 	}
 
-	memcpy(priv->dcbx.tc_tsa, ets.tc_tsa, sizeof(ets.tc_tsa));
-
 	/* tclass[prio=0]=1, tclass[prio=1]=0, tclass[prio=i]=i (for i>1) */
 	ets.prio_tc[0] = 1;
 	ets.prio_tc[1] = 0;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c
index 1975d43..e07061f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c
@@ -677,6 +677,27 @@ int mlx5_set_port_tc_group(struct mlx5_core_dev *mdev, u8 *tc_group)
 }
 EXPORT_SYMBOL_GPL(mlx5_set_port_tc_group);
 
+int mlx5_query_port_tc_group(struct mlx5_core_dev *mdev,
+			     u8 tc, u8 *tc_group)
+{
+	u32 out[MLX5_ST_SZ_DW(qetc_reg)];
+	void *ets_tcn_conf;
+	int err;
+
+	err = mlx5_query_port_qetcr_reg(mdev, out, sizeof(out));
+	if (err)
+		return err;
+
+	ets_tcn_conf = MLX5_ADDR_OF(qetc_reg, out,
+				    tc_configuration[tc]);
+
+	*tc_group = MLX5_GET(ets_tcn_config_reg, ets_tcn_conf,
+			     group);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mlx5_query_port_tc_group);
+
 int mlx5_set_port_tc_bw_alloc(struct mlx5_core_dev *mdev, u8 *tc_bw)
 {
 	u32 in[MLX5_ST_SZ_DW(qetc_reg)] = {0};
-- 
cgit v1.1