[pull request][net 00/10] Mellanox, mlx5 fixes 2019-03-11

[pull request][net 00/10] Mellanox, mlx5 fixes 2019-03-11

[Saeed Mahameed]

Hi Dave,

This series introduces some fixes to mlx5 driver.
For more information please see tag log below.

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

For -stable v5.0:
('net/mlx5e: Fix access to non-existing receive queue')
('net/mlx5e: Properly get the PF number phys port name ndo')
('net/mlx5: Fix multiple updates of steering rules in parallel')
('net/mlx5: Avoid panic when setting vport mac, getting vport config')
('net/mlx5: Avoid panic when setting vport rate')
('net/mlx5e: IPoIB, Fix RX checksum statistics update')

Thanks,
Saeed.

---
The following changes since commit a843dc4ebaecd15fca1f4d35a97210f72ea1473b:

  net: sit: fix UBSAN Undefined behaviour in check_6rd (2019-03-11 10:32:45 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux.git tags/mlx5-fixes-2019-03-11

for you to fetch changes up to 24319258660a84dd77f4be026a55b10a12524919:

  net/mlx5: Avoid panic when setting vport rate (2019-03-11 11:32:40 -0700)

----------------------------------------------------------------
mlx5-fixes-2019-03-11

----------------------------------------------------------------
Eli Britstein (4):
      net/mlx5: Fix multiple updates of steering rules in parallel
      net/mlx5: Add modify FTE helper function
      net/mlx5: Add a locked flag to node removal functions
      net/mlx5: Consolidate update FTE for all removal changes

Feras Daoud (1):
      net/mlx5e: IPoIB, Fix RX checksum statistics update

Roi Dayan (2):
      net/mlx5e: Properly get the PF number phys port name ndo
      net/mlx5: Remove redundant lag function to get pf num

Tariq Toukan (1):
      net/mlx5e: Fix access to non-existing receive queue

Tonghao Zhang (2):
      net/mlx5: Avoid panic when setting vport mac, getting vport config
      net/mlx5: Avoid panic when setting vport rate

 .../net/ethernet/mellanox/mlx5/core/en_ethtool.c   |   3 +
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c   |  13 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c    |  11 +-
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c  |  17 ++-
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c  | 161 +++++++++++++--------
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.h  |   1 +
 drivers/net/ethernet/mellanox/mlx5/core/lag.c      |  21 ---
 .../net/ethernet/mellanox/mlx5/core/mlx5_core.h    |   2 -
 8 files changed, 128 insertions(+), 101 deletions(-)

[net 01/10] net/mlx5: Fix multiple updates of steering rules in parallel

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

There might be a condition where the fte found is not active yet. In
this case we should not use it, but continue to search for another, or
allocate a new one.

Fixes: bd71b08ec2ee ("net/mlx5: Support multiple updates of steering rules in parallel")
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index f2cfa012315e..86f986308a90 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -491,6 +491,7 @@ static void del_hw_fte(struct fs_node *node)
 			mlx5_core_warn(dev,
 				       "flow steering can't delete fte in index %d of flow group id %d\n",
 				       fte->index, fg->id);
+		node->active = 0;
 	}
 }
 
@@ -1601,6 +1602,11 @@ lookup_fte_locked(struct mlx5_flow_group *g,
 		fte_tmp = NULL;
 		goto out;
 	}
+	if (!fte_tmp->node.active) {
+		tree_put_node(&fte_tmp->node);
+		fte_tmp = NULL;
+		goto out;
+	}
 
 	nested_down_write_ref_node(&fte_tmp->node, FS_LOCK_CHILD);
 out:
-- 
2.20.1

[net 02/10] net/mlx5: Add modify FTE helper function

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Add modify FTE helper function and use it when deleting a rule, as a
pre-step towards consolidated FTE modification, with no functional
change.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 50 +++++++++++--------
 .../net/ethernet/mellanox/mlx5/core/fs_core.h |  1 +
 2 files changed, 30 insertions(+), 21 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 86f986308a90..5ff8c5f44b35 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -420,22 +420,37 @@ static void del_sw_flow_table(struct fs_node *node)
 	kfree(ft);
 }
 
-static void del_sw_hw_rule(struct fs_node *node)
+static void modify_fte(struct fs_fte *fte)
 {
 	struct mlx5_flow_root_namespace *root;
-	struct mlx5_flow_rule *rule;
 	struct mlx5_flow_table *ft;
 	struct mlx5_flow_group *fg;
-	struct fs_fte *fte;
-	int modify_mask;
-	struct mlx5_core_dev *dev = get_dev(node);
+	struct mlx5_core_dev *dev;
 	int err;
-	bool update_fte = false;
 
-	fs_get_obj(rule, node);
-	fs_get_obj(fte, rule->node.parent);
+	if (!fte->modify_mask)
+		return;
+
 	fs_get_obj(fg, fte->node.parent);
 	fs_get_obj(ft, fg->node.parent);
+	dev = get_dev(&fte->node);
+
+	root = find_root(&ft->node);
+	err = root->cmds->update_fte(dev, ft, fg->id, fte->modify_mask, fte);
+	if (err)
+		mlx5_core_warn(dev,
+			       "%s can't del rule fg id=%d fte_index=%d\n",
+			       __func__, fg->id, fte->index);
+	fte->modify_mask = 0;
+}
+
+static void del_sw_hw_rule(struct fs_node *node)
+{
+	struct mlx5_flow_rule *rule;
+	struct fs_fte *fte;
+
+	fs_get_obj(rule, node);
+	fs_get_obj(fte, rule->node.parent);
 	trace_mlx5_fs_del_rule(rule);
 	if (rule->sw_action == MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO) {
 		mutex_lock(&rule->dest_attr.ft->lock);
@@ -445,27 +460,20 @@ static void del_sw_hw_rule(struct fs_node *node)
 
 	if (rule->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_COUNTER  &&
 	    --fte->dests_size) {
-		modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION) |
-			      BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_FLOW_COUNTERS);
+		fte->modify_mask |=
+			BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_ACTION) |
+			BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_FLOW_COUNTERS);
 		fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_COUNT;
-		update_fte = true;
 		goto out;
 	}
 
 	if ((fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) &&
 	    --fte->dests_size) {
-		modify_mask = BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_DESTINATION_LIST);
-		update_fte = true;
+		fte->modify_mask |=
+			BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_DESTINATION_LIST);
 	}
 out:
-	root = find_root(&ft->node);
-	if (update_fte && fte->dests_size) {
-		err = root->cmds->update_fte(dev, ft, fg->id, modify_mask, fte);
-		if (err)
-			mlx5_core_warn(dev,
-				       "%s can't del rule fg id=%d fte_index=%d\n",
-				       __func__, fg->id, fte->index);
-	}
+	modify_fte(fte);
 	kfree(rule);
 }
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
index 2dc86347af58..87de0e4d9124 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h
@@ -172,6 +172,7 @@ struct fs_fte {
 	enum fs_fte_status		status;
 	struct mlx5_fc			*counter;
 	struct rhash_head		hash;
+	int				modify_mask;
 };
 
 /* Type of children is mlx5_flow_table/namespace */
-- 
2.20.1

[net 03/10] net/mlx5: Add a locked flag to node removal functions

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

Add a locked flag to the node removal functions to signal if the
parent is already locked from the caller function or not as a pre-step
towards outside lock. Currently always use false with no functional
change.

Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 86 ++++++++++---------
 1 file changed, 44 insertions(+), 42 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 5ff8c5f44b35..42a6952ff9d0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -263,10 +263,11 @@ static void nested_down_write_ref_node(struct fs_node *node,
 	}
 }
 
-static void down_write_ref_node(struct fs_node *node)
+static void down_write_ref_node(struct fs_node *node, bool locked)
 {
 	if (node) {
-		down_write(&node->lock);
+		if (!locked)
+			down_write(&node->lock);
 		refcount_inc(&node->refcount);
 	}
 }
@@ -277,13 +278,14 @@ static void up_read_ref_node(struct fs_node *node)
 	up_read(&node->lock);
 }
 
-static void up_write_ref_node(struct fs_node *node)
+static void up_write_ref_node(struct fs_node *node, bool locked)
 {
 	refcount_dec(&node->refcount);
-	up_write(&node->lock);
+	if (!locked)
+		up_write(&node->lock);
 }
 
-static void tree_put_node(struct fs_node *node)
+static void tree_put_node(struct fs_node *node, bool locked)
 {
 	struct fs_node *parent_node = node->parent;
 
@@ -294,27 +296,27 @@ static void tree_put_node(struct fs_node *node)
 			/* Only root namespace doesn't have parent and we just
 			 * need to free its node.
 			 */
-			down_write_ref_node(parent_node);
+			down_write_ref_node(parent_node, locked);
 			list_del_init(&node->list);
 			if (node->del_sw_func)
 				node->del_sw_func(node);
-			up_write_ref_node(parent_node);
+			up_write_ref_node(parent_node, locked);
 		} else {
 			kfree(node);
 		}
 		node = NULL;
 	}
 	if (!node && parent_node)
-		tree_put_node(parent_node);
+		tree_put_node(parent_node, locked);
 }
 
-static int tree_remove_node(struct fs_node *node)
+static int tree_remove_node(struct fs_node *node, bool locked)
 {
 	if (refcount_read(&node->refcount) > 1) {
 		refcount_dec(&node->refcount);
 		return -EEXIST;
 	}
-	tree_put_node(node);
+	tree_put_node(node, locked);
 	return 0;
 }
 
@@ -867,7 +869,7 @@ static int _mlx5_modify_rule_destination(struct mlx5_flow_rule *rule,
 	fs_get_obj(fte, rule->node.parent);
 	if (!(fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST))
 		return -EINVAL;
-	down_write_ref_node(&fte->node);
+	down_write_ref_node(&fte->node, false);
 	fs_get_obj(fg, fte->node.parent);
 	fs_get_obj(ft, fg->node.parent);
 
@@ -875,7 +877,7 @@ static int _mlx5_modify_rule_destination(struct mlx5_flow_rule *rule,
 	root = find_root(&ft->node);
 	err = root->cmds->update_fte(get_dev(&ft->node), ft, fg->id,
 				     modify_mask, fte);
-	up_write_ref_node(&fte->node);
+	up_write_ref_node(&fte->node, false);
 
 	return err;
 }
@@ -1025,11 +1027,11 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa
 	if (err)
 		goto destroy_ft;
 	ft->node.active = true;
-	down_write_ref_node(&fs_prio->node);
+	down_write_ref_node(&fs_prio->node, false);
 	tree_add_node(&ft->node, &fs_prio->node);
 	list_add_flow_table(ft, fs_prio);
 	fs_prio->num_ft++;
-	up_write_ref_node(&fs_prio->node);
+	up_write_ref_node(&fs_prio->node, false);
 	mutex_unlock(&root->chain_lock);
 	trace_mlx5_fs_add_ft(ft);
 	return ft;
@@ -1123,17 +1125,17 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft,
 	if (ft->autogroup.active)
 		return ERR_PTR(-EPERM);
 
-	down_write_ref_node(&ft->node);
+	down_write_ref_node(&ft->node, false);
 	fg = alloc_insert_flow_group(ft, match_criteria_enable, match_criteria,
 				     start_index, end_index,
 				     ft->node.children.prev);
-	up_write_ref_node(&ft->node);
+	up_write_ref_node(&ft->node, false);
 	if (IS_ERR(fg))
 		return fg;
 
 	err = root->cmds->create_flow_group(dev, ft, fg_in, &fg->id);
 	if (err) {
-		tree_put_node(&fg->node);
+		tree_put_node(&fg->node, false);
 		return ERR_PTR(err);
 	}
 	trace_mlx5_fs_add_fg(fg);
@@ -1530,10 +1532,10 @@ static void free_match_list(struct match_list_head *head)
 		struct match_list *iter, *match_tmp;
 
 		list_del(&head->first.list);
-		tree_put_node(&head->first.g->node);
+		tree_put_node(&head->first.g->node, false);
 		list_for_each_entry_safe(iter, match_tmp, &head->list,
 					 list) {
-			tree_put_node(&iter->g->node);
+			tree_put_node(&iter->g->node, false);
 			list_del(&iter->list);
 			kfree(iter);
 		}
@@ -1611,7 +1613,7 @@ lookup_fte_locked(struct mlx5_flow_group *g,
 		goto out;
 	}
 	if (!fte_tmp->node.active) {
-		tree_put_node(&fte_tmp->node);
+		tree_put_node(&fte_tmp->node, false);
 		fte_tmp = NULL;
 		goto out;
 	}
@@ -1619,7 +1621,7 @@ lookup_fte_locked(struct mlx5_flow_group *g,
 	nested_down_write_ref_node(&fte_tmp->node, FS_LOCK_CHILD);
 out:
 	if (take_write)
-		up_write_ref_node(&g->node);
+		up_write_ref_node(&g->node, false);
 	else
 		up_read_ref_node(&g->node);
 	return fte_tmp;
@@ -1661,8 +1663,8 @@ try_add_to_existing_fg(struct mlx5_flow_table *ft,
 			continue;
 		rule = add_rule_fg(g, spec->match_value,
 				   flow_act, dest, dest_num, fte_tmp);
-		up_write_ref_node(&fte_tmp->node);
-		tree_put_node(&fte_tmp->node);
+		up_write_ref_node(&fte_tmp->node, false);
+		tree_put_node(&fte_tmp->node, false);
 		kmem_cache_free(steering->ftes_cache, fte);
 		return rule;
 	}
@@ -1698,7 +1700,7 @@ try_add_to_existing_fg(struct mlx5_flow_table *ft,
 
 		err = insert_fte(g, fte);
 		if (err) {
-			up_write_ref_node(&g->node);
+			up_write_ref_node(&g->node, false);
 			if (err == -ENOSPC)
 				continue;
 			kmem_cache_free(steering->ftes_cache, fte);
@@ -1706,11 +1708,11 @@ try_add_to_existing_fg(struct mlx5_flow_table *ft,
 		}
 
 		nested_down_write_ref_node(&fte->node, FS_LOCK_CHILD);
-		up_write_ref_node(&g->node);
+		up_write_ref_node(&g->node, false);
 		rule = add_rule_fg(g, spec->match_value,
 				   flow_act, dest, dest_num, fte);
-		up_write_ref_node(&fte->node);
-		tree_put_node(&fte->node);
+		up_write_ref_node(&fte->node, false);
+		tree_put_node(&fte->node, false);
 		return rule;
 	}
 	rule = ERR_PTR(-ENOENT);
@@ -1752,7 +1754,7 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft,
 	err = build_match_list(&match_head, ft, spec);
 	if (err) {
 		if (take_write)
-			up_write_ref_node(&ft->node);
+			up_write_ref_node(&ft->node, false);
 		else
 			up_read_ref_node(&ft->node);
 		return ERR_PTR(err);
@@ -1767,7 +1769,7 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft,
 	if (!IS_ERR(rule) ||
 	    (PTR_ERR(rule) != -ENOENT && PTR_ERR(rule) != -EAGAIN)) {
 		if (take_write)
-			up_write_ref_node(&ft->node);
+			up_write_ref_node(&ft->node, false);
 		return rule;
 	}
 
@@ -1783,12 +1785,12 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft,
 	g = alloc_auto_flow_group(ft, spec);
 	if (IS_ERR(g)) {
 		rule = ERR_CAST(g);
-		up_write_ref_node(&ft->node);
+		up_write_ref_node(&ft->node, false);
 		return rule;
 	}
 
 	nested_down_write_ref_node(&g->node, FS_LOCK_PARENT);
-	up_write_ref_node(&ft->node);
+	up_write_ref_node(&ft->node, false);
 
 	err = create_auto_flow_group(ft, g);
 	if (err)
@@ -1807,17 +1809,17 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft,
 	}
 
 	nested_down_write_ref_node(&fte->node, FS_LOCK_CHILD);
-	up_write_ref_node(&g->node);
+	up_write_ref_node(&g->node, false);
 	rule = add_rule_fg(g, spec->match_value, flow_act, dest,
 			   dest_num, fte);
-	up_write_ref_node(&fte->node);
-	tree_put_node(&fte->node);
-	tree_put_node(&g->node);
+	up_write_ref_node(&fte->node, false);
+	tree_put_node(&fte->node, false);
+	tree_put_node(&g->node, false);
 	return rule;
 
 err_release_fg:
-	up_write_ref_node(&g->node);
-	tree_put_node(&g->node);
+	up_write_ref_node(&g->node, false);
+	tree_put_node(&g->node, false);
 	return ERR_PTR(err);
 }
 
@@ -1883,7 +1885,7 @@ void mlx5_del_flow_rules(struct mlx5_flow_handle *handle)
 	int i;
 
 	for (i = handle->num_rules - 1; i >= 0; i--)
-		tree_remove_node(&handle->rule[i]->node);
+		tree_remove_node(&handle->rule[i]->node, false);
 	kfree(handle);
 }
 EXPORT_SYMBOL(mlx5_del_flow_rules);
@@ -1986,7 +1988,7 @@ int mlx5_destroy_flow_table(struct mlx5_flow_table *ft)
 		mutex_unlock(&root->chain_lock);
 		return err;
 	}
-	if (tree_remove_node(&ft->node))
+	if (tree_remove_node(&ft->node, false))
 		mlx5_core_warn(get_dev(&ft->node), "Flow table %d wasn't destroyed, refcount > 1\n",
 			       ft->id);
 	mutex_unlock(&root->chain_lock);
@@ -1997,7 +1999,7 @@ EXPORT_SYMBOL(mlx5_destroy_flow_table);
 
 void mlx5_destroy_flow_group(struct mlx5_flow_group *fg)
 {
-	if (tree_remove_node(&fg->node))
+	if (tree_remove_node(&fg->node, false))
 		mlx5_core_warn(get_dev(&fg->node), "Flow group %d wasn't destroyed, refcount > 1\n",
 			       fg->id);
 }
@@ -2381,8 +2383,8 @@ static void clean_tree(struct fs_node *node)
 		tree_get_node(node);
 		list_for_each_entry_safe(iter, temp, &node->children, list)
 			clean_tree(iter);
-		tree_put_node(node);
-		tree_remove_node(node);
+		tree_put_node(node, false);
+		tree_remove_node(node, false);
 	}
 }
 
-- 
2.20.1

[net 04/10] net/mlx5: Consolidate update FTE for all removal changes

[Saeed Mahameed]

From: Eli Britstein <elibr@mellanox.com>

With commit a18e879d4e45 ("net/mlx5e: Annul encap action ordering
requirement") and a use-case of e-switch remote mirroring, the
incremental/stepped FTE removal process done by the fs core got us to
illegal transient states and FW errors:

SET_FLOW_TABLE_ENTRY(0x936) op_mod(0x0) failed, status bad
parameter(0x3), syndrome (0x9c2e40)

To avoid that and improve FTE removal performance, aggregate the FTE's
updates that should be applied. Remove the FTE if it is empty, or apply
one FW update command with the aggregated updates.

Fixes: a18e879d4e45 ("net/mlx5e: Annul encap action ordering requirement")
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/fs_core.c | 31 +++++++++++++++----
 1 file changed, 25 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 42a6952ff9d0..0be3eb86dd84 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -430,9 +430,6 @@ static void modify_fte(struct fs_fte *fte)
 	struct mlx5_core_dev *dev;
 	int err;
 
-	if (!fte->modify_mask)
-		return;
-
 	fs_get_obj(fg, fte->node.parent);
 	fs_get_obj(ft, fg->node.parent);
 	dev = get_dev(&fte->node);
@@ -475,7 +472,6 @@ static void del_sw_hw_rule(struct fs_node *node)
 			BIT(MLX5_SET_FTE_MODIFY_ENABLE_MASK_DESTINATION_LIST);
 	}
 out:
-	modify_fte(fte);
 	kfree(rule);
 }
 
@@ -602,7 +598,7 @@ static struct fs_fte *alloc_fte(struct mlx5_flow_table *ft,
 	fte->node.type =  FS_TYPE_FLOW_ENTRY;
 	fte->action = *flow_act;
 
-	tree_init_node(&fte->node, del_hw_fte, del_sw_fte);
+	tree_init_node(&fte->node, NULL, del_sw_fte);
 
 	return fte;
 }
@@ -1882,10 +1878,33 @@ EXPORT_SYMBOL(mlx5_add_flow_rules);
 
 void mlx5_del_flow_rules(struct mlx5_flow_handle *handle)
 {
+	struct fs_fte *fte;
 	int i;
 
+	/* In order to consolidate the HW changes we lock the FTE for other
+	 * changes, and increase its refcount, in order not to perform the
+	 * "del" functions of the FTE. Will handle them here.
+	 * The removal of the rules is done under locked FTE.
+	 * After removing all the handle's rules, if there are remaining
+	 * rules, it means we just need to modify the FTE in FW, and
+	 * unlock/decrease the refcount we increased before.
+	 * Otherwise, it means the FTE should be deleted. First delete the
+	 * FTE in FW. Then, unlock the FTE, and proceed the tree_put_node of
+	 * the FTE, which will handle the last decrease of the refcount, as
+	 * well as required handling of its parent.
+	 */
+	fs_get_obj(fte, handle->rule[0]->node.parent);
+	down_write_ref_node(&fte->node, false);
 	for (i = handle->num_rules - 1; i >= 0; i--)
-		tree_remove_node(&handle->rule[i]->node, false);
+		tree_remove_node(&handle->rule[i]->node, true);
+	if (fte->modify_mask && fte->dests_size) {
+		modify_fte(fte);
+		up_write_ref_node(&fte->node, false);
+	} else {
+		del_hw_fte(&fte->node);
+		up_write(&fte->node.lock);
+		tree_put_node(&fte->node, false);
+	}
 	kfree(handle);
 }
 EXPORT_SYMBOL(mlx5_del_flow_rules);
-- 
2.20.1

[net 05/10] net/mlx5e: Properly get the PF number phys port name ndo

[Saeed Mahameed]

From: Roi Dayan <roid@mellanox.com>

Currently, we fail to retrieve the PF number in some cases (e.g
single ported cards, lag capability), this further results in a
call trace issued by the rtnetlink code, since the error value
is not -EOPNOTSUPP. Change the implementation to be independent
from the lag code and function properly on both two ports and
single ported cards.

Call Trace:

[  194.525057] mlx5_core 0000:82:00.0: mlx5_lag_get_pf_num:605:(pid 837): no lag device, can't get pf num
[  194.525804] WARNING: CPU: 7 PID: 837 at net/core/rtnetlink.c:3457 rtmsg_ifinfo_build_skb+0x131/0x160
[  194.529952] CPU: 7 PID: 837 Comm: kworker/7:3 Tainted: G        W  O      5.0.0-rc7+ #3
[  194.531307] Workqueue: events linkwatch_event
[  194.531697] RIP: 0010:rtmsg_ifinfo_build_skb+0x131/0x160
[  194.545007] Call Trace:
[  194.545406]  rtmsg_ifinfo_event.part.29+0x1b/0xb0
[  194.545810]  rtmsg_ifinfo+0x51/0x80
[  194.546209]  netdev_state_change+0xc7/0x110
[  194.546608]  ? dev_valid_name+0x1b0/0x1b0
[  194.547010]  ? __local_bh_enable_ip+0xef/0x1d0
[  194.547411]  ? lockdep_hardirqs_on+0x3ea/0x560
[  194.547811]  ? linkwatch_do_dev+0x9b/0x100
[  194.548207]  linkwatch_do_dev+0x9b/0x100
[  194.548605]  __linkwatch_run_queue+0x244/0x430
[  194.549014]  ? linkwatch_schedule_work+0x100/0x100
[  194.549412]  ? lock_acquire+0x10f/0x2d0
[  194.549816]  linkwatch_event+0x3f/0x50
[  194.550212]  process_one_work+0x7d3/0x1460

Fixes: c12ecc230564 ("net/mlx5e: Move to use common phys port names for vport representors")
Signed-off-by: Roi Dayan <roid@mellanox.com>
Acked-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
index a1a3e2774989..a66b6ed80b30 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c
@@ -1129,16 +1129,17 @@ static int mlx5e_rep_get_phys_port_name(struct net_device *dev,
 	struct mlx5e_priv *priv = netdev_priv(dev);
 	struct mlx5e_rep_priv *rpriv = priv->ppriv;
 	struct mlx5_eswitch_rep *rep = rpriv->rep;
-	int ret, pf_num;
+	unsigned int fn;
+	int ret;
 
-	ret = mlx5_lag_get_pf_num(priv->mdev, &pf_num);
-	if (ret)
-		return ret;
+	fn = PCI_FUNC(priv->mdev->pdev->devfn);
+	if (fn >= MLX5_MAX_PORTS)
+		return -EOPNOTSUPP;
 
 	if (rep->vport == MLX5_VPORT_UPLINK)
-		ret = snprintf(buf, len, "p%d", pf_num);
+		ret = snprintf(buf, len, "p%d", fn);
 	else
-		ret = snprintf(buf, len, "pf%dvf%d", pf_num, rep->vport - 1);
+		ret = snprintf(buf, len, "pf%dvf%d", fn, rep->vport - 1);
 
 	if (ret >= len)
 		return -EOPNOTSUPP;
-- 
2.20.1

[net 06/10] net/mlx5: Remove redundant lag function to get pf num

[Saeed Mahameed]

From: Roi Dayan <roid@mellanox.com>

The function is not being used.

Signed-off-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/lag.c | 21 -------------------
 .../ethernet/mellanox/mlx5/core/mlx5_core.h   |  2 --
 2 files changed, 23 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag.c
index 48aa6e030bcf..959605559858 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/lag.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/lag.c
@@ -595,27 +595,6 @@ void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev)
 			      err);
 }
 
-int mlx5_lag_get_pf_num(struct mlx5_core_dev *dev, int *pf_num)
-{
-	struct mlx5_lag *ldev;
-	int n;
-
-	ldev = mlx5_lag_dev_get(dev);
-	if (!ldev) {
-		mlx5_core_warn(dev, "no lag device, can't get pf num\n");
-		return -EINVAL;
-	}
-
-	for (n = 0; n < MLX5_MAX_PORTS; n++)
-		if (ldev->pf[n].dev == dev) {
-			*pf_num = n;
-			return 0;
-		}
-
-	mlx5_core_warn(dev, "wasn't able to locate pf in the lag device\n");
-	return -EINVAL;
-}
-
 /* Must be called with intf_mutex held */
 void mlx5_lag_remove(struct mlx5_core_dev *dev)
 {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index 9529cf9623e3..7b331674622c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -188,8 +188,6 @@ static inline int mlx5_lag_is_lacp_owner(struct mlx5_core_dev *dev)
 		    MLX5_CAP_GEN(dev, lag_master);
 }
 
-int mlx5_lag_get_pf_num(struct mlx5_core_dev *dev, int *pf_num);
-
 void mlx5_reload_interface(struct mlx5_core_dev *mdev, int protocol);
 void mlx5_lag_update(struct mlx5_core_dev *dev);
 
-- 
2.20.1

[net 07/10] net/mlx5e: IPoIB, Fix RX checksum statistics update

[Saeed Mahameed]

From: Feras Daoud <ferasda@mellanox.com>

Update the RX checksum only if the feature is enabled.

Fixes: 9d6bd752c63c ("net/mlx5e: IPoIB, RX handler")
Signed-off-by: Feras Daoud <ferasda@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 11 ++++++++---
 1 file changed, 8 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 be396e5e4e39..3dde5c7e0739 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1295,8 +1295,14 @@ static inline void mlx5i_complete_rx_cqe(struct mlx5e_rq *rq,
 
 	skb->protocol = *((__be16 *)(skb->data));
 
-	skb->ip_summed = CHECKSUM_COMPLETE;
-	skb->csum = csum_unfold((__force __sum16)cqe->check_sum);
+	if (netdev->features & NETIF_F_RXCSUM) {
+		skb->ip_summed = CHECKSUM_COMPLETE;
+		skb->csum = csum_unfold((__force __sum16)cqe->check_sum);
+		stats->csum_complete++;
+	} else {
+		skb->ip_summed = CHECKSUM_NONE;
+		stats->csum_none++;
+	}
 
 	if (unlikely(mlx5e_rx_hw_stamp(tstamp)))
 		skb_hwtstamps(skb)->hwtstamp =
@@ -1315,7 +1321,6 @@ static inline void mlx5i_complete_rx_cqe(struct mlx5e_rq *rq,
 
 	skb->dev = netdev;
 
-	stats->csum_complete++;
 	stats->packets++;
 	stats->bytes += cqe_bcnt;
 }
-- 
2.20.1

[net 08/10] net/mlx5e: Fix access to non-existing receive queue

[Saeed Mahameed]

From: Tariq Toukan <tariqt@mellanox.com>

In case number of channels is changed while interface is down,
RSS indirection table is mistakenly not modified accordingly,
causing access to out-of-range non-existing object.

Fix by updating the RSS indireciton table also in the early
return flow of interface down.

Fixes: fb35c534b788 ("net/mlx5e: Fix NULL pointer derefernce in set channels error flow")
Fixes: bbeb53b8b2c9 ("net/mlx5e: Move RSS params to a dedicated struct")
Reported-by: Or Gerlitz <ogerlitz@mellanox.com>
Tested-by: Maria Pasechnik <mariap@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
index 0804b478ad19..a0987cc5fe4a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c
@@ -424,6 +424,9 @@ int mlx5e_ethtool_set_channels(struct mlx5e_priv *priv,
 
 	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
 		priv->channels.params = new_channels.params;
+		if (!netif_is_rxfh_configured(priv->netdev))
+			mlx5e_build_default_indir_rqt(priv->rss_params.indirection_rqt,
+						      MLX5E_INDIR_RQT_SIZE, count);
 		goto out;
 	}
 
-- 
2.20.1

[net 09/10] net/mlx5: Avoid panic when setting vport mac, getting vport config

[Saeed Mahameed]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

If we try to set VFs mac address on a VF (not PF) net device,
the kernel will be crash. The commands are show as below:

$ echo 2 > /sys/class/net/$MLX_PF0/device/sriov_numvfs
$ ip link set $MLX_VF0 vf 0 mac 00:11:22:33:44:00

[exception RIP: mlx5_eswitch_set_vport_mac+41]
[ffffb8b7079e3688] do_setlink at ffffffff8f67f85b
[ffffb8b7079e37a8] __rtnl_newlink at ffffffff8f683778
[ffffb8b7079e3b68] rtnl_newlink at ffffffff8f683a63
[ffffb8b7079e3b90] rtnetlink_rcv_msg at ffffffff8f67d812
[ffffb8b7079e3c10] netlink_rcv_skb at ffffffff8f6b88ab
[ffffb8b7079e3c60] netlink_unicast at ffffffff8f6b808f
[ffffb8b7079e3ca0] netlink_sendmsg at ffffffff8f6b8412
[ffffb8b7079e3d18] sock_sendmsg at ffffffff8f6452f6
[ffffb8b7079e3d30] ___sys_sendmsg at ffffffff8f645860
[ffffb8b7079e3eb0] __sys_sendmsg at ffffffff8f647a38
[ffffb8b7079e3f38] do_syscall_64 at ffffffff8f00401b
[ffffb8b7079e3f50] entry_SYSCALL_64_after_hwframe at ffffffff8f80008c

and

[exception RIP: mlx5_eswitch_get_vport_config+12]
[ffffa70607e57678] mlx5e_get_vf_config at ffffffffc03c7f8f [mlx5_core]
[ffffa70607e57688] do_setlink at ffffffffbc67fa59
[ffffa70607e577a8] __rtnl_newlink at ffffffffbc683778
[ffffa70607e57b68] rtnl_newlink at ffffffffbc683a63
[ffffa70607e57b90] rtnetlink_rcv_msg at ffffffffbc67d812
[ffffa70607e57c10] netlink_rcv_skb at ffffffffbc6b88ab
[ffffa70607e57c60] netlink_unicast at ffffffffbc6b808f
[ffffa70607e57ca0] netlink_sendmsg at ffffffffbc6b8412
[ffffa70607e57d18] sock_sendmsg at ffffffffbc6452f6
[ffffa70607e57d30] ___sys_sendmsg at ffffffffbc645860
[ffffa70607e57eb0] __sys_sendmsg at ffffffffbc647a38
[ffffa70607e57f38] do_syscall_64 at ffffffffbc00401b
[ffffa70607e57f50] entry_SYSCALL_64_after_hwframe at ffffffffbc80008c

Fixes: a8d70a054a718 ("net/mlx5: E-Switch, Disallow vlan/spoofcheck setup if not being esw manager")
Cc: Eli Cohen <eli@mellanox.com>
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index d0b28251abf2..e582a42005f1 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -1931,7 +1931,7 @@ int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw,
 	u64 node_guid;
 	int err = 0;
 
-	if (!MLX5_CAP_GEN(esw->dev, vport_group_manager))
+	if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager))
 		return -EPERM;
 	if (!LEGAL_VPORT(esw, vport) || is_multicast_ether_addr(mac))
 		return -EINVAL;
@@ -2005,7 +2005,7 @@ int mlx5_eswitch_get_vport_config(struct mlx5_eswitch *esw,
 {
 	struct mlx5_vport *evport;
 
-	if (!MLX5_CAP_GEN(esw->dev, vport_group_manager))
+	if (!esw || !MLX5_CAP_GEN(esw->dev, vport_group_manager))
 		return -EPERM;
 	if (!LEGAL_VPORT(esw, vport))
 		return -EINVAL;
-- 
2.20.1

[net 10/10] net/mlx5: Avoid panic when setting vport rate

[Saeed Mahameed]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

If we try to set VFs rate on a VF (not PF) net device, the kernel
will be crash. The commands are show as below:

$ echo 2 > /sys/class/net/$MLX_PF0/device/sriov_numvfs
$ ip link set $MLX_VF0 vf 0 max_tx_rate 2 min_tx_rate 1

If not applied the first patch ("net/mlx5: Avoid panic when setting
vport mac, getting vport config"), the command:

$ ip link set $MLX_VF0 vf 0 rate 100

can also crash the kernel.

[ 1650.006388] RIP: 0010:mlx5_eswitch_set_vport_rate+0x1f/0x260 [mlx5_core]
[ 1650.007092]  do_setlink+0x982/0xd20
[ 1650.007129]  __rtnl_newlink+0x528/0x7d0
[ 1650.007374]  rtnl_newlink+0x43/0x60
[ 1650.007407]  rtnetlink_rcv_msg+0x2a2/0x320
[ 1650.007484]  netlink_rcv_skb+0xcb/0x100
[ 1650.007519]  netlink_unicast+0x17f/0x230
[ 1650.007554]  netlink_sendmsg+0x2d2/0x3d0
[ 1650.007592]  sock_sendmsg+0x36/0x50
[ 1650.007625]  ___sys_sendmsg+0x280/0x2a0
[ 1650.007963]  __sys_sendmsg+0x58/0xa0
[ 1650.007998]  do_syscall_64+0x5b/0x180
[ 1650.009438]  entry_SYSCALL_64_after_hwframe+0x44/0xa9

Fixes: c9497c98901c ("net/mlx5: Add support for setting VF min rate")
Cc: Mohamad Haj Yahia <mohamad@mellanox.com>
Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
index e582a42005f1..ecd2c747f726 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c
@@ -2297,19 +2297,24 @@ static int normalize_vports_min_rate(struct mlx5_eswitch *esw, u32 divider)
 int mlx5_eswitch_set_vport_rate(struct mlx5_eswitch *esw, int vport,
 				u32 max_rate, u32 min_rate)
 {
-	u32 fw_max_bw_share = MLX5_CAP_QOS(esw->dev, max_tsar_bw_share);
-	bool min_rate_supported = MLX5_CAP_QOS(esw->dev, esw_bw_share) &&
-					fw_max_bw_share >= MLX5_MIN_BW_SHARE;
-	bool max_rate_supported = MLX5_CAP_QOS(esw->dev, esw_rate_limit);
 	struct mlx5_vport *evport;
+	u32 fw_max_bw_share;
 	u32 previous_min_rate;
 	u32 divider;
+	bool min_rate_supported;
+	bool max_rate_supported;
 	int err = 0;
 
 	if (!ESW_ALLOWED(esw))
 		return -EPERM;
 	if (!LEGAL_VPORT(esw, vport))
 		return -EINVAL;
+
+	fw_max_bw_share = MLX5_CAP_QOS(esw->dev, max_tsar_bw_share);
+	min_rate_supported = MLX5_CAP_QOS(esw->dev, esw_bw_share) &&
+				fw_max_bw_share >= MLX5_MIN_BW_SHARE;
+	max_rate_supported = MLX5_CAP_QOS(esw->dev, esw_rate_limit);
+
 	if ((min_rate && !min_rate_supported) || (max_rate && !max_rate_supported))
 		return -EOPNOTSUPP;
 
-- 
2.20.1

Re: [pull request][net 00/10] Mellanox, mlx5 fixes 2019-03-11

[David Miller]

From: Saeed Mahameed <saeedm@mellanox.com>
Date: Mon, 11 Mar 2019 11:39:19 -0700

> This series introduces some fixes to mlx5 driver.
> For more information please see tag log below.
> 
> Please pull and let me know if there is any problem.
> 
> For -stable v5.0:
> ('net/mlx5e: Fix access to non-existing receive queue')
> ('net/mlx5e: Properly get the PF number phys port name ndo')
> ('net/mlx5: Fix multiple updates of steering rules in parallel')
> ('net/mlx5: Avoid panic when setting vport mac, getting vport config')
> ('net/mlx5: Avoid panic when setting vport rate')
> ('net/mlx5e: IPoIB, Fix RX checksum statistics update')

Pulled, thanks Saeed.