[PATCH net-next 0/9] dpaa2-switch: add mirroring support

[PATCH net-next 0/9] dpaa2-switch: add mirroring support

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

This patch set adds per port and per VLAN mirroring in dpaa2-switch.

The first 4 patches are just cosmetic changes. We renamed the
dpaa2_switch_acl_tbl structure into dpaa2_switch_filter_block so that we
can reuse it for filters that do not use the ACL table and reorganized
the addition of trap, redirect and drop filters into a separate
function. All this just to make for a more streamlined addition of the
support for mirroring.

The next 4 patches are actually adding the advertised support. Mirroring
rules can be added in shared blocks, the driver will replicate the same
configuration on all the switch ports part of the same block.

The last patch documents the feature, presents its behavior and
limitations and gives a couple of examples.

Ioana Ciornei (9):
  dpaa2-switch: rename dpaa2_switch_tc_parse_action to specify the ACL
  dpaa2-switch: rename dpaa2_switch_acl_tbl into filter_block
  dpaa2-switch: reorganize dpaa2_switch_cls_flower_replace
  dpaa2-switch: reorganize dpaa2_switch_cls_matchall_replace
  dpaa2-switch: add API for setting up mirroring
  dpaa2-switch: add support for port mirroring
  dpaa2-switch: add VLAN based mirroring
  dpaa2-switch: offload shared block mirror filters when binding to a
    port
  docs: networking: dpaa2: document mirroring support on the switch

 .../freescale/dpaa2/switch-driver.rst         |  43 ++
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 530 +++++++++++++++---
 .../ethernet/freescale/dpaa2/dpaa2-switch.c   | 149 ++---
 .../ethernet/freescale/dpaa2/dpaa2-switch.h   |  44 +-
 .../net/ethernet/freescale/dpaa2/dpsw-cmd.h   |  19 +
 drivers/net/ethernet/freescale/dpaa2/dpsw.c   |  80 +++
 drivers/net/ethernet/freescale/dpaa2/dpsw.h   |  31 +
 7 files changed, 747 insertions(+), 149 deletions(-)

-- 
2.31.1

[PATCH net-next 1/9] dpaa2-switch: rename dpaa2_switch_tc_parse_action to specify the ACL

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Until now, the dpaa2_switch_tc_parse_action() function was used for all
the supported tc actions since all of them were implemented by adding
ACL table entries. In the next commits, the dpaa2-switch driver will
gain mirroring support which is not using the same HW feature.

Make sure that we specify the ACL in the function name so that we make
it clear that it's only used for specific actions.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c        | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index f9451ec5f2cb..639efb3edeec 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -339,10 +339,10 @@ dpaa2_switch_acl_tbl_remove_entry(struct dpaa2_switch_acl_tbl *acl_tbl,
 	return 0;
 }
 
-static int dpaa2_switch_tc_parse_action(struct ethsw_core *ethsw,
-					struct flow_action_entry *cls_act,
-					struct dpsw_acl_result *dpsw_act,
-					struct netlink_ext_ack *extack)
+static int dpaa2_switch_tc_parse_action_acl(struct ethsw_core *ethsw,
+					    struct flow_action_entry *cls_act,
+					    struct dpsw_acl_result *dpsw_act,
+					    struct netlink_ext_ack *extack)
 {
 	int err = 0;
 
@@ -403,8 +403,8 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 		goto free_acl_entry;
 
 	act = &rule->action.entries[0];
-	err = dpaa2_switch_tc_parse_action(ethsw, act,
-					   &acl_entry->cfg.result, extack);
+	err = dpaa2_switch_tc_parse_action_acl(ethsw, act,
+					       &acl_entry->cfg.result, extack);
 	if (err)
 		goto free_acl_entry;
 
@@ -459,8 +459,8 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 		return -ENOMEM;
 
 	act = &cls->rule->action.entries[0];
-	err = dpaa2_switch_tc_parse_action(ethsw, act,
-					   &acl_entry->cfg.result, extack);
+	err = dpaa2_switch_tc_parse_action_acl(ethsw, act,
+					       &acl_entry->cfg.result, extack);
 	if (err)
 		goto free_acl_entry;
 
-- 
2.31.1

[PATCH net-next 2/9] dpaa2-switch: rename dpaa2_switch_acl_tbl into filter_block

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Until now, shared filter blocks were implemented only by ACL tables
shared between ports. Going forward, when the mirroring support will be
added, this will not be true anymore.

Rename the dpaa2_switch_acl_tbl into dpaa2_switch_filter_block so that
we make it clear that the structure is used not only for filters that
use the ACL table but will be used for all the filters that are added in
a block.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     |  98 ++++++-------
 .../ethernet/freescale/dpaa2/dpaa2-switch.c   | 129 +++++++++---------
 .../ethernet/freescale/dpaa2/dpaa2-switch.h   |  28 ++--
 3 files changed, 130 insertions(+), 125 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index 639efb3edeec..80fe09ac9d5f 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -111,11 +111,11 @@ static int dpaa2_switch_flower_parse_key(struct flow_cls_offload *cls,
 	return 0;
 }
 
-int dpaa2_switch_acl_entry_add(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_acl_entry_add(struct dpaa2_switch_filter_block *filter_block,
 			       struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpsw_acl_entry_cfg *acl_entry_cfg = &entry->cfg;
-	struct ethsw_core *ethsw = acl_tbl->ethsw;
+	struct ethsw_core *ethsw = filter_block->ethsw;
 	struct dpsw_acl_key *acl_key = &entry->key;
 	struct device *dev = ethsw->dev;
 	u8 *cmd_buff;
@@ -136,7 +136,7 @@ int dpaa2_switch_acl_entry_add(struct dpaa2_switch_acl_tbl *acl_tbl,
 	}
 
 	err = dpsw_acl_add_entry(ethsw->mc_io, 0, ethsw->dpsw_handle,
-				 acl_tbl->id, acl_entry_cfg);
+				 filter_block->acl_id, acl_entry_cfg);
 
 	dma_unmap_single(dev, acl_entry_cfg->key_iova, sizeof(cmd_buff),
 			 DMA_TO_DEVICE);
@@ -150,12 +150,13 @@ int dpaa2_switch_acl_entry_add(struct dpaa2_switch_acl_tbl *acl_tbl,
 	return 0;
 }
 
-static int dpaa2_switch_acl_entry_remove(struct dpaa2_switch_acl_tbl *acl_tbl,
-					 struct dpaa2_switch_acl_entry *entry)
+static int
+dpaa2_switch_acl_entry_remove(struct dpaa2_switch_filter_block *block,
+			      struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpsw_acl_entry_cfg *acl_entry_cfg = &entry->cfg;
 	struct dpsw_acl_key *acl_key = &entry->key;
-	struct ethsw_core *ethsw = acl_tbl->ethsw;
+	struct ethsw_core *ethsw = block->ethsw;
 	struct device *dev = ethsw->dev;
 	u8 *cmd_buff;
 	int err;
@@ -175,7 +176,7 @@ static int dpaa2_switch_acl_entry_remove(struct dpaa2_switch_acl_tbl *acl_tbl,
 	}
 
 	err = dpsw_acl_remove_entry(ethsw->mc_io, 0, ethsw->dpsw_handle,
-				    acl_tbl->id, acl_entry_cfg);
+				    block->acl_id, acl_entry_cfg);
 
 	dma_unmap_single(dev, acl_entry_cfg->key_iova, sizeof(cmd_buff),
 			 DMA_TO_DEVICE);
@@ -190,19 +191,19 @@ static int dpaa2_switch_acl_entry_remove(struct dpaa2_switch_acl_tbl *acl_tbl,
 }
 
 static int
-dpaa2_switch_acl_entry_add_to_list(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_entry_add_to_list(struct dpaa2_switch_filter_block *block,
 				   struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpaa2_switch_acl_entry *tmp;
 	struct list_head *pos, *n;
 	int index = 0;
 
-	if (list_empty(&acl_tbl->entries)) {
-		list_add(&entry->list, &acl_tbl->entries);
+	if (list_empty(&block->acl_entries)) {
+		list_add(&entry->list, &block->acl_entries);
 		return index;
 	}
 
-	list_for_each_safe(pos, n, &acl_tbl->entries) {
+	list_for_each_safe(pos, n, &block->acl_entries) {
 		tmp = list_entry(pos, struct dpaa2_switch_acl_entry, list);
 		if (entry->prio < tmp->prio)
 			break;
@@ -213,13 +214,13 @@ dpaa2_switch_acl_entry_add_to_list(struct dpaa2_switch_acl_tbl *acl_tbl,
 }
 
 static struct dpaa2_switch_acl_entry*
-dpaa2_switch_acl_entry_get_by_index(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_entry_get_by_index(struct dpaa2_switch_filter_block *block,
 				    int index)
 {
 	struct dpaa2_switch_acl_entry *tmp;
 	int i = 0;
 
-	list_for_each_entry(tmp, &acl_tbl->entries, list) {
+	list_for_each_entry(tmp, &block->acl_entries, list) {
 		if (i == index)
 			return tmp;
 		++i;
@@ -229,37 +230,38 @@ dpaa2_switch_acl_entry_get_by_index(struct dpaa2_switch_acl_tbl *acl_tbl,
 }
 
 static int
-dpaa2_switch_acl_entry_set_precedence(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_entry_set_precedence(struct dpaa2_switch_filter_block *block,
 				      struct dpaa2_switch_acl_entry *entry,
 				      int precedence)
 {
 	int err;
 
-	err = dpaa2_switch_acl_entry_remove(acl_tbl, entry);
+	err = dpaa2_switch_acl_entry_remove(block, entry);
 	if (err)
 		return err;
 
 	entry->cfg.precedence = precedence;
-	return dpaa2_switch_acl_entry_add(acl_tbl, entry);
+	return dpaa2_switch_acl_entry_add(block, entry);
 }
 
-static int dpaa2_switch_acl_tbl_add_entry(struct dpaa2_switch_acl_tbl *acl_tbl,
-					  struct dpaa2_switch_acl_entry *entry)
+static int
+dpaa2_switch_acl_tbl_add_entry(struct dpaa2_switch_filter_block *block,
+			       struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpaa2_switch_acl_entry *tmp;
 	int index, i, precedence, err;
 
 	/* Add the new ACL entry to the linked list and get its index */
-	index = dpaa2_switch_acl_entry_add_to_list(acl_tbl, entry);
+	index = dpaa2_switch_acl_entry_add_to_list(block, entry);
 
 	/* Move up in priority the ACL entries to make space
 	 * for the new filter.
 	 */
-	precedence = DPAA2_ETHSW_PORT_MAX_ACL_ENTRIES - acl_tbl->num_rules - 1;
+	precedence = DPAA2_ETHSW_PORT_MAX_ACL_ENTRIES - block->num_acl_rules - 1;
 	for (i = 0; i < index; i++) {
-		tmp = dpaa2_switch_acl_entry_get_by_index(acl_tbl, i);
+		tmp = dpaa2_switch_acl_entry_get_by_index(block, i);
 
-		err = dpaa2_switch_acl_entry_set_precedence(acl_tbl, tmp,
+		err = dpaa2_switch_acl_entry_set_precedence(block, tmp,
 							    precedence);
 		if (err)
 			return err;
@@ -269,19 +271,19 @@ static int dpaa2_switch_acl_tbl_add_entry(struct dpaa2_switch_acl_tbl *acl_tbl,
 
 	/* Add the new entry to hardware */
 	entry->cfg.precedence = precedence;
-	err = dpaa2_switch_acl_entry_add(acl_tbl, entry);
-	acl_tbl->num_rules++;
+	err = dpaa2_switch_acl_entry_add(block, entry);
+	block->num_acl_rules++;
 
 	return err;
 }
 
 static struct dpaa2_switch_acl_entry *
-dpaa2_switch_acl_tbl_find_entry_by_cookie(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_tbl_find_entry_by_cookie(struct dpaa2_switch_filter_block *block,
 					  unsigned long cookie)
 {
 	struct dpaa2_switch_acl_entry *tmp, *n;
 
-	list_for_each_entry_safe(tmp, n, &acl_tbl->entries, list) {
+	list_for_each_entry_safe(tmp, n, &block->acl_entries, list) {
 		if (tmp->cookie == cookie)
 			return tmp;
 	}
@@ -289,13 +291,13 @@ dpaa2_switch_acl_tbl_find_entry_by_cookie(struct dpaa2_switch_acl_tbl *acl_tbl,
 }
 
 static int
-dpaa2_switch_acl_entry_get_index(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_entry_get_index(struct dpaa2_switch_filter_block *block,
 				 struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpaa2_switch_acl_entry *tmp, *n;
 	int index = 0;
 
-	list_for_each_entry_safe(tmp, n, &acl_tbl->entries, list) {
+	list_for_each_entry_safe(tmp, n, &block->acl_entries, list) {
 		if (tmp->cookie == entry->cookie)
 			return index;
 		index++;
@@ -304,20 +306,20 @@ dpaa2_switch_acl_entry_get_index(struct dpaa2_switch_acl_tbl *acl_tbl,
 }
 
 static int
-dpaa2_switch_acl_tbl_remove_entry(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_acl_tbl_remove_entry(struct dpaa2_switch_filter_block *block,
 				  struct dpaa2_switch_acl_entry *entry)
 {
 	struct dpaa2_switch_acl_entry *tmp;
 	int index, i, precedence, err;
 
-	index = dpaa2_switch_acl_entry_get_index(acl_tbl, entry);
+	index = dpaa2_switch_acl_entry_get_index(block, entry);
 
 	/* Remove from hardware the ACL entry */
-	err = dpaa2_switch_acl_entry_remove(acl_tbl, entry);
+	err = dpaa2_switch_acl_entry_remove(block, entry);
 	if (err)
 		return err;
 
-	acl_tbl->num_rules--;
+	block->num_acl_rules--;
 
 	/* Remove it from the list also */
 	list_del(&entry->list);
@@ -325,8 +327,8 @@ dpaa2_switch_acl_tbl_remove_entry(struct dpaa2_switch_acl_tbl *acl_tbl,
 	/* Move down in priority the entries over the deleted one */
 	precedence = entry->cfg.precedence;
 	for (i = index - 1; i >= 0; i--) {
-		tmp = dpaa2_switch_acl_entry_get_by_index(acl_tbl, i);
-		err = dpaa2_switch_acl_entry_set_precedence(acl_tbl, tmp,
+		tmp = dpaa2_switch_acl_entry_get_by_index(block, i);
+		err = dpaa2_switch_acl_entry_set_precedence(block, tmp,
 							    precedence);
 		if (err)
 			return err;
@@ -374,13 +376,13 @@ static int dpaa2_switch_tc_parse_action_acl(struct ethsw_core *ethsw,
 	return err;
 }
 
-int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
 	struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
 	struct netlink_ext_ack *extack = cls->common.extack;
-	struct ethsw_core *ethsw = acl_tbl->ethsw;
 	struct dpaa2_switch_acl_entry *acl_entry;
+	struct ethsw_core *ethsw = block->ethsw;
 	struct flow_action_entry *act;
 	int err;
 
@@ -389,7 +391,7 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 		return -EOPNOTSUPP;
 	}
 
-	if (dpaa2_switch_acl_tbl_is_full(acl_tbl)) {
+	if (dpaa2_switch_acl_tbl_is_full(block)) {
 		NL_SET_ERR_MSG(extack, "Maximum filter capacity reached");
 		return -ENOMEM;
 	}
@@ -411,7 +413,7 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 	acl_entry->prio = cls->common.prio;
 	acl_entry->cookie = cls->cookie;
 
-	err = dpaa2_switch_acl_tbl_add_entry(acl_tbl, acl_entry);
+	err = dpaa2_switch_acl_tbl_add_entry(block, acl_entry);
 	if (err)
 		goto free_acl_entry;
 
@@ -423,23 +425,23 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 	return err;
 }
 
-int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
 	struct dpaa2_switch_acl_entry *entry;
 
-	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(acl_tbl, cls->cookie);
+	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block, cls->cookie);
 	if (!entry)
 		return 0;
 
-	return dpaa2_switch_acl_tbl_remove_entry(acl_tbl, entry);
+	return dpaa2_switch_acl_tbl_remove_entry(block, entry);
 }
 
-int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
 	struct netlink_ext_ack *extack = cls->common.extack;
-	struct ethsw_core *ethsw = acl_tbl->ethsw;
+	struct ethsw_core *ethsw = block->ethsw;
 	struct dpaa2_switch_acl_entry *acl_entry;
 	struct flow_action_entry *act;
 	int err;
@@ -449,7 +451,7 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 		return -EOPNOTSUPP;
 	}
 
-	if (dpaa2_switch_acl_tbl_is_full(acl_tbl)) {
+	if (dpaa2_switch_acl_tbl_is_full(block)) {
 		NL_SET_ERR_MSG(extack, "Maximum filter capacity reached");
 		return -ENOMEM;
 	}
@@ -467,7 +469,7 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 	acl_entry->prio = cls->common.prio;
 	acl_entry->cookie = cls->cookie;
 
-	err = dpaa2_switch_acl_tbl_add_entry(acl_tbl, acl_entry);
+	err = dpaa2_switch_acl_tbl_add_entry(block, acl_entry);
 	if (err)
 		goto free_acl_entry;
 
@@ -479,14 +481,14 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
 	return err;
 }
 
-int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
 	struct dpaa2_switch_acl_entry *entry;
 
-	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(acl_tbl, cls->cookie);
+	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block, cls->cookie);
 	if (!entry)
 		return 0;
 
-	return  dpaa2_switch_acl_tbl_remove_entry(acl_tbl, entry);
+	return  dpaa2_switch_acl_tbl_remove_entry(block, entry);
 }
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
index c233e8786e19..1806012f41d2 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
@@ -41,14 +41,14 @@ static struct dpaa2_switch_fdb *dpaa2_switch_fdb_get_unused(struct ethsw_core *e
 	return NULL;
 }
 
-static struct dpaa2_switch_acl_tbl *
-dpaa2_switch_acl_tbl_get_unused(struct ethsw_core *ethsw)
+static struct dpaa2_switch_filter_block *
+dpaa2_switch_filter_block_get_unused(struct ethsw_core *ethsw)
 {
 	int i;
 
 	for (i = 0; i < ethsw->sw_attr.num_ifs; i++)
-		if (!ethsw->acls[i].in_use)
-			return &ethsw->acls[i];
+		if (!ethsw->filter_blocks[i].in_use)
+			return &ethsw->filter_blocks[i];
 	return NULL;
 }
 
@@ -1127,28 +1127,28 @@ static netdev_tx_t dpaa2_switch_port_tx(struct sk_buff *skb,
 }
 
 static int
-dpaa2_switch_setup_tc_cls_flower(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_setup_tc_cls_flower(struct dpaa2_switch_filter_block *filter_block,
 				 struct flow_cls_offload *f)
 {
 	switch (f->command) {
 	case FLOW_CLS_REPLACE:
-		return dpaa2_switch_cls_flower_replace(acl_tbl, f);
+		return dpaa2_switch_cls_flower_replace(filter_block, f);
 	case FLOW_CLS_DESTROY:
-		return dpaa2_switch_cls_flower_destroy(acl_tbl, f);
+		return dpaa2_switch_cls_flower_destroy(filter_block, f);
 	default:
 		return -EOPNOTSUPP;
 	}
 }
 
 static int
-dpaa2_switch_setup_tc_cls_matchall(struct dpaa2_switch_acl_tbl *acl_tbl,
+dpaa2_switch_setup_tc_cls_matchall(struct dpaa2_switch_filter_block *block,
 				   struct tc_cls_matchall_offload *f)
 {
 	switch (f->command) {
 	case TC_CLSMATCHALL_REPLACE:
-		return dpaa2_switch_cls_matchall_replace(acl_tbl, f);
+		return dpaa2_switch_cls_matchall_replace(block, f);
 	case TC_CLSMATCHALL_DESTROY:
-		return dpaa2_switch_cls_matchall_destroy(acl_tbl, f);
+		return dpaa2_switch_cls_matchall_destroy(block, f);
 	default:
 		return -EOPNOTSUPP;
 	}
@@ -1170,106 +1170,108 @@ static int dpaa2_switch_port_setup_tc_block_cb_ig(enum tc_setup_type type,
 
 static LIST_HEAD(dpaa2_switch_block_cb_list);
 
-static int dpaa2_switch_port_acl_tbl_bind(struct ethsw_port_priv *port_priv,
-					  struct dpaa2_switch_acl_tbl *acl_tbl)
+static int
+dpaa2_switch_port_acl_tbl_bind(struct ethsw_port_priv *port_priv,
+			       struct dpaa2_switch_filter_block *block)
 {
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
 	struct net_device *netdev = port_priv->netdev;
 	struct dpsw_acl_if_cfg acl_if_cfg;
 	int err;
 
-	if (port_priv->acl_tbl)
+	if (port_priv->filter_block)
 		return -EINVAL;
 
 	acl_if_cfg.if_id[0] = port_priv->idx;
 	acl_if_cfg.num_ifs = 1;
 	err = dpsw_acl_add_if(ethsw->mc_io, 0, ethsw->dpsw_handle,
-			      acl_tbl->id, &acl_if_cfg);
+			      block->acl_id, &acl_if_cfg);
 	if (err) {
 		netdev_err(netdev, "dpsw_acl_add_if err %d\n", err);
 		return err;
 	}
 
-	acl_tbl->ports |= BIT(port_priv->idx);
-	port_priv->acl_tbl = acl_tbl;
+	block->ports |= BIT(port_priv->idx);
+	port_priv->filter_block = block;
 
 	return 0;
 }
 
 static int
 dpaa2_switch_port_acl_tbl_unbind(struct ethsw_port_priv *port_priv,
-				 struct dpaa2_switch_acl_tbl *acl_tbl)
+				 struct dpaa2_switch_filter_block *block)
 {
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
 	struct net_device *netdev = port_priv->netdev;
 	struct dpsw_acl_if_cfg acl_if_cfg;
 	int err;
 
-	if (port_priv->acl_tbl != acl_tbl)
+	if (port_priv->filter_block != block)
 		return -EINVAL;
 
 	acl_if_cfg.if_id[0] = port_priv->idx;
 	acl_if_cfg.num_ifs = 1;
 	err = dpsw_acl_remove_if(ethsw->mc_io, 0, ethsw->dpsw_handle,
-				 acl_tbl->id, &acl_if_cfg);
+				 block->acl_id, &acl_if_cfg);
 	if (err) {
 		netdev_err(netdev, "dpsw_acl_add_if err %d\n", err);
 		return err;
 	}
 
-	acl_tbl->ports &= ~BIT(port_priv->idx);
-	port_priv->acl_tbl = NULL;
+	block->ports &= ~BIT(port_priv->idx);
+	port_priv->filter_block = NULL;
 	return 0;
 }
 
 static int dpaa2_switch_port_block_bind(struct ethsw_port_priv *port_priv,
-					struct dpaa2_switch_acl_tbl *acl_tbl)
+					struct dpaa2_switch_filter_block *block)
 {
-	struct dpaa2_switch_acl_tbl *old_acl_tbl = port_priv->acl_tbl;
+	struct dpaa2_switch_filter_block *old_block = port_priv->filter_block;
 	int err;
 
 	/* If the port is already bound to this ACL table then do nothing. This
 	 * can happen when this port is the first one to join a tc block
 	 */
-	if (port_priv->acl_tbl == acl_tbl)
+	if (port_priv->filter_block == block)
 		return 0;
 
-	err = dpaa2_switch_port_acl_tbl_unbind(port_priv, old_acl_tbl);
+	err = dpaa2_switch_port_acl_tbl_unbind(port_priv, old_block);
 	if (err)
 		return err;
 
 	/* Mark the previous ACL table as being unused if this was the last
 	 * port that was using it.
 	 */
-	if (old_acl_tbl->ports == 0)
-		old_acl_tbl->in_use = false;
+	if (old_block->ports == 0)
+		old_block->in_use = false;
 
-	return dpaa2_switch_port_acl_tbl_bind(port_priv, acl_tbl);
+	return dpaa2_switch_port_acl_tbl_bind(port_priv, block);
 }
 
-static int dpaa2_switch_port_block_unbind(struct ethsw_port_priv *port_priv,
-					  struct dpaa2_switch_acl_tbl *acl_tbl)
+static int
+dpaa2_switch_port_block_unbind(struct ethsw_port_priv *port_priv,
+			       struct dpaa2_switch_filter_block *block)
 {
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
-	struct dpaa2_switch_acl_tbl *new_acl_tbl;
+	struct dpaa2_switch_filter_block *new_block;
 	int err;
 
 	/* We are the last port that leaves a block (an ACL table).
 	 * We'll continue to use this table.
 	 */
-	if (acl_tbl->ports == BIT(port_priv->idx))
+	if (block->ports == BIT(port_priv->idx))
 		return 0;
 
-	err = dpaa2_switch_port_acl_tbl_unbind(port_priv, acl_tbl);
+	err = dpaa2_switch_port_acl_tbl_unbind(port_priv, block);
 	if (err)
 		return err;
 
-	if (acl_tbl->ports == 0)
-		acl_tbl->in_use = false;
+	if (block->ports == 0)
+		block->in_use = false;
 
-	new_acl_tbl = dpaa2_switch_acl_tbl_get_unused(ethsw);
-	new_acl_tbl->in_use = true;
-	return dpaa2_switch_port_acl_tbl_bind(port_priv, new_acl_tbl);
+	new_block = dpaa2_switch_filter_block_get_unused(ethsw);
+	new_block->in_use = true;
+	return dpaa2_switch_port_acl_tbl_bind(port_priv, new_block);
 }
 
 static int dpaa2_switch_setup_tc_block_bind(struct net_device *netdev,
@@ -1277,7 +1279,7 @@ static int dpaa2_switch_setup_tc_block_bind(struct net_device *netdev,
 {
 	struct ethsw_port_priv *port_priv = netdev_priv(netdev);
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
-	struct dpaa2_switch_acl_tbl *acl_tbl;
+	struct dpaa2_switch_filter_block *filter_block;
 	struct flow_block_cb *block_cb;
 	bool register_block = false;
 	int err;
@@ -1287,24 +1289,24 @@ static int dpaa2_switch_setup_tc_block_bind(struct net_device *netdev,
 					ethsw);
 
 	if (!block_cb) {
-		/* If the ACL table is not already known, then this port must
-		 * be the first to join it. In this case, we can just continue
-		 * to use our private table
+		/* If the filter block is not already known, then this port
+		 * must be the first to join it. In this case, we can just
+		 * continue to use our private table
 		 */
-		acl_tbl = port_priv->acl_tbl;
+		filter_block = port_priv->filter_block;
 
 		block_cb = flow_block_cb_alloc(dpaa2_switch_port_setup_tc_block_cb_ig,
-					       ethsw, acl_tbl, NULL);
+					       ethsw, filter_block, NULL);
 		if (IS_ERR(block_cb))
 			return PTR_ERR(block_cb);
 
 		register_block = true;
 	} else {
-		acl_tbl = flow_block_cb_priv(block_cb);
+		filter_block = flow_block_cb_priv(block_cb);
 	}
 
 	flow_block_cb_incref(block_cb);
-	err = dpaa2_switch_port_block_bind(port_priv, acl_tbl);
+	err = dpaa2_switch_port_block_bind(port_priv, filter_block);
 	if (err)
 		goto err_block_bind;
 
@@ -1327,7 +1329,7 @@ static void dpaa2_switch_setup_tc_block_unbind(struct net_device *netdev,
 {
 	struct ethsw_port_priv *port_priv = netdev_priv(netdev);
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
-	struct dpaa2_switch_acl_tbl *acl_tbl;
+	struct dpaa2_switch_filter_block *filter_block;
 	struct flow_block_cb *block_cb;
 	int err;
 
@@ -1337,8 +1339,8 @@ static void dpaa2_switch_setup_tc_block_unbind(struct net_device *netdev,
 	if (!block_cb)
 		return;
 
-	acl_tbl = flow_block_cb_priv(block_cb);
-	err = dpaa2_switch_port_block_unbind(port_priv, acl_tbl);
+	filter_block = flow_block_cb_priv(block_cb);
+	err = dpaa2_switch_port_block_unbind(port_priv, filter_block);
 	if (!err && !flow_block_cb_decref(block_cb)) {
 		flow_block_cb_remove(block_cb, f);
 		list_del(&block_cb->driver_list);
@@ -2991,7 +2993,7 @@ static int dpaa2_switch_port_trap_mac_addr(struct ethsw_port_priv *port_priv,
 	acl_entry.cfg.precedence = 0;
 	acl_entry.cfg.result.action = DPSW_ACL_ACTION_REDIRECT_TO_CTRL_IF;
 
-	return dpaa2_switch_acl_entry_add(port_priv->acl_tbl, &acl_entry);
+	return dpaa2_switch_acl_entry_add(port_priv->filter_block, &acl_entry);
 }
 
 static int dpaa2_switch_port_init(struct ethsw_port_priv *port_priv, u16 port)
@@ -3004,7 +3006,7 @@ static int dpaa2_switch_port_init(struct ethsw_port_priv *port_priv, u16 port)
 	};
 	struct net_device *netdev = port_priv->netdev;
 	struct ethsw_core *ethsw = port_priv->ethsw_data;
-	struct dpaa2_switch_acl_tbl *acl_tbl;
+	struct dpaa2_switch_filter_block *filter_block;
 	struct dpsw_fdb_cfg fdb_cfg = {0};
 	struct dpsw_if_attr dpsw_if_attr;
 	struct dpaa2_switch_fdb *fdb;
@@ -3059,14 +3061,14 @@ static int dpaa2_switch_port_init(struct ethsw_port_priv *port_priv, u16 port)
 		return err;
 	}
 
-	acl_tbl = dpaa2_switch_acl_tbl_get_unused(ethsw);
-	acl_tbl->ethsw = ethsw;
-	acl_tbl->id = acl_tbl_id;
-	acl_tbl->in_use = true;
-	acl_tbl->num_rules = 0;
-	INIT_LIST_HEAD(&acl_tbl->entries);
+	filter_block = dpaa2_switch_filter_block_get_unused(ethsw);
+	filter_block->ethsw = ethsw;
+	filter_block->acl_id = acl_tbl_id;
+	filter_block->in_use = true;
+	filter_block->num_acl_rules = 0;
+	INIT_LIST_HEAD(&filter_block->acl_entries);
 
-	err = dpaa2_switch_port_acl_tbl_bind(port_priv, acl_tbl);
+	err = dpaa2_switch_port_acl_tbl_bind(port_priv, filter_block);
 	if (err)
 		return err;
 
@@ -3120,7 +3122,7 @@ static int dpaa2_switch_remove(struct fsl_mc_device *sw_dev)
 	}
 
 	kfree(ethsw->fdbs);
-	kfree(ethsw->acls);
+	kfree(ethsw->filter_blocks);
 	kfree(ethsw->ports);
 
 	dpaa2_switch_takedown(sw_dev);
@@ -3248,9 +3250,10 @@ static int dpaa2_switch_probe(struct fsl_mc_device *sw_dev)
 		goto err_free_ports;
 	}
 
-	ethsw->acls = kcalloc(ethsw->sw_attr.num_ifs, sizeof(*ethsw->acls),
-			      GFP_KERNEL);
-	if (!ethsw->acls) {
+	ethsw->filter_blocks = kcalloc(ethsw->sw_attr.num_ifs,
+				       sizeof(*ethsw->filter_blocks),
+				       GFP_KERNEL);
+	if (!ethsw->filter_blocks) {
 		err = -ENOMEM;
 		goto err_free_fdbs;
 	}
@@ -3303,7 +3306,7 @@ static int dpaa2_switch_probe(struct fsl_mc_device *sw_dev)
 err_free_netdev:
 	for (i--; i >= 0; i--)
 		free_netdev(ethsw->ports[i]->netdev);
-	kfree(ethsw->acls);
+	kfree(ethsw->filter_blocks);
 err_free_fdbs:
 	kfree(ethsw->fdbs);
 err_free_ports:
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
index bdef71f234cb..296a09eb7a9a 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
@@ -113,20 +113,20 @@ struct dpaa2_switch_acl_entry {
 	struct dpsw_acl_key	key;
 };
 
-struct dpaa2_switch_acl_tbl {
-	struct list_head	entries;
+struct dpaa2_switch_filter_block {
 	struct ethsw_core	*ethsw;
 	u64			ports;
-
-	u16			id;
-	u8			num_rules;
 	bool			in_use;
+
+	struct list_head	acl_entries;
+	u16			acl_id;
+	u8			num_acl_rules;
 };
 
 static inline bool
-dpaa2_switch_acl_tbl_is_full(struct dpaa2_switch_acl_tbl *acl_tbl)
+dpaa2_switch_acl_tbl_is_full(struct dpaa2_switch_filter_block *filter_block)
 {
-	if ((acl_tbl->num_rules + DPAA2_ETHSW_PORT_DEFAULT_TRAPS) >=
+	if ((filter_block->num_acl_rules + DPAA2_ETHSW_PORT_DEFAULT_TRAPS) >=
 	    DPAA2_ETHSW_PORT_MAX_ACL_ENTRIES)
 		return true;
 	return false;
@@ -149,7 +149,7 @@ struct ethsw_port_priv {
 	bool			ucast_flood;
 	bool			learn_ena;
 
-	struct dpaa2_switch_acl_tbl *acl_tbl;
+	struct dpaa2_switch_filter_block *filter_block;
 };
 
 /* Switch data */
@@ -175,7 +175,7 @@ struct ethsw_core {
 	int				napi_users;
 
 	struct dpaa2_switch_fdb		*fdbs;
-	struct dpaa2_switch_acl_tbl	*acls;
+	struct dpaa2_switch_filter_block *filter_blocks;
 };
 
 static inline int dpaa2_switch_get_index(struct ethsw_core *ethsw,
@@ -229,18 +229,18 @@ typedef int dpaa2_switch_fdb_cb_t(struct ethsw_port_priv *port_priv,
 
 /* TC offload */
 
-int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls);
 
-int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls);
 
-int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls);
 
-int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls);
 
-int dpaa2_switch_acl_entry_add(struct dpaa2_switch_acl_tbl *acl_tbl,
+int dpaa2_switch_acl_entry_add(struct dpaa2_switch_filter_block *block,
 			       struct dpaa2_switch_acl_entry *entry);
 #endif	/* __ETHSW_H */
-- 
2.31.1

[PATCH net-next 3/9] dpaa2-switch: reorganize dpaa2_switch_cls_flower_replace

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Extract the necessary steps to offload a filter by using the ACL table
in a separate function - dpaa2_switch_cls_flower_replace_acl().
This is intended to help with the code readability when the mirroring
support is added.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 32 +++++++++++++++----
 1 file changed, 26 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index 80fe09ac9d5f..38a321be58ff 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -376,7 +376,8 @@ static int dpaa2_switch_tc_parse_action_acl(struct ethsw_core *ethsw,
 	return err;
 }
 
-int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
+static int
+dpaa2_switch_cls_flower_replace_acl(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
 	struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
@@ -386,11 +387,6 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 	struct flow_action_entry *act;
 	int err;
 
-	if (!flow_offload_has_one_action(&rule->action)) {
-		NL_SET_ERR_MSG(extack, "Only singular actions are supported");
-		return -EOPNOTSUPP;
-	}
-
 	if (dpaa2_switch_acl_tbl_is_full(block)) {
 		NL_SET_ERR_MSG(extack, "Maximum filter capacity reached");
 		return -ENOMEM;
@@ -425,6 +421,30 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 	return err;
 }
 
+int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
+				    struct flow_cls_offload *cls)
+{
+	struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct flow_action_entry *act;
+
+	if (!flow_offload_has_one_action(&rule->action)) {
+		NL_SET_ERR_MSG(extack, "Only singular actions are supported");
+		return -EOPNOTSUPP;
+	}
+
+	act = &rule->action.entries[0];
+	switch (act->id) {
+	case FLOW_ACTION_REDIRECT:
+	case FLOW_ACTION_TRAP:
+	case FLOW_ACTION_DROP:
+		return dpaa2_switch_cls_flower_replace_acl(block, cls);
+	default:
+		NL_SET_ERR_MSG_MOD(extack, "Action not supported");
+		return -EOPNOTSUPP;
+	}
+}
+
 int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
-- 
2.31.1

[PATCH net-next 4/9] dpaa2-switch: reorganize dpaa2_switch_cls_matchall_replace

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Extract the necessary steps to offload a filter by using the ACL table
in a separate function - dpaa2_switch_cls_matchall_replace_acl().

This is intended to help with the code readability when the mirroring
support is added.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 31 +++++++++++++++----
 1 file changed, 25 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index 38a321be58ff..637291060fd5 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -457,7 +457,8 @@ int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_filter_block *block,
 	return dpaa2_switch_acl_tbl_remove_entry(block, entry);
 }
 
-int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
+static int
+dpaa2_switch_cls_matchall_replace_acl(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
 	struct netlink_ext_ack *extack = cls->common.extack;
@@ -466,11 +467,6 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 	struct flow_action_entry *act;
 	int err;
 
-	if (!flow_offload_has_one_action(&cls->rule->action)) {
-		NL_SET_ERR_MSG(extack, "Only singular actions are supported");
-		return -EOPNOTSUPP;
-	}
-
 	if (dpaa2_switch_acl_tbl_is_full(block)) {
 		NL_SET_ERR_MSG(extack, "Maximum filter capacity reached");
 		return -ENOMEM;
@@ -501,6 +497,29 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 	return err;
 }
 
+int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
+				      struct tc_cls_matchall_offload *cls)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct flow_action_entry *act;
+
+	if (!flow_offload_has_one_action(&cls->rule->action)) {
+		NL_SET_ERR_MSG(extack, "Only singular actions are supported");
+		return -EOPNOTSUPP;
+	}
+
+	act = &cls->rule->action.entries[0];
+	switch (act->id) {
+	case FLOW_ACTION_REDIRECT:
+	case FLOW_ACTION_TRAP:
+	case FLOW_ACTION_DROP:
+		return dpaa2_switch_cls_matchall_replace_acl(block, cls);
+	default:
+		NL_SET_ERR_MSG_MOD(extack, "Action not supported");
+		return -EOPNOTSUPP;
+	}
+}
+
 int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
-- 
2.31.1

[PATCH net-next 5/9] dpaa2-switch: add API for setting up mirroring

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Add the necessary MC API for setting up and configuring the mirroring
feature on the DPSW DPAA2 object.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../net/ethernet/freescale/dpaa2/dpsw-cmd.h   | 19 +++++
 drivers/net/ethernet/freescale/dpaa2/dpsw.c   | 80 +++++++++++++++++++
 drivers/net/ethernet/freescale/dpaa2/dpsw.h   | 31 +++++++
 3 files changed, 130 insertions(+)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpsw-cmd.h b/drivers/net/ethernet/freescale/dpaa2/dpsw-cmd.h
index cb13e740f72b..397d55f2bd99 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpsw-cmd.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpsw-cmd.h
@@ -39,11 +39,16 @@
 #define DPSW_CMDID_GET_IRQ_STATUS           DPSW_CMD_ID(0x016)
 #define DPSW_CMDID_CLEAR_IRQ_STATUS         DPSW_CMD_ID(0x017)
 
+#define DPSW_CMDID_SET_REFLECTION_IF        DPSW_CMD_ID(0x022)
+
 #define DPSW_CMDID_IF_SET_TCI               DPSW_CMD_ID(0x030)
 #define DPSW_CMDID_IF_SET_STP               DPSW_CMD_ID(0x031)
 
 #define DPSW_CMDID_IF_GET_COUNTER           DPSW_CMD_V2(0x034)
 
+#define DPSW_CMDID_IF_ADD_REFLECTION        DPSW_CMD_ID(0x037)
+#define DPSW_CMDID_IF_REMOVE_REFLECTION     DPSW_CMD_ID(0x038)
+
 #define DPSW_CMDID_IF_ENABLE                DPSW_CMD_ID(0x03D)
 #define DPSW_CMDID_IF_DISABLE               DPSW_CMD_ID(0x03E)
 
@@ -533,5 +538,19 @@ struct dpsw_cmd_acl_entry {
 	__le64 pad2[4];
 	__le64 key_iova;
 };
+
+struct dpsw_cmd_set_reflection_if {
+	__le16 if_id;
+};
+
+#define DPSW_FILTER_SHIFT	0
+#define DPSW_FILTER_SIZE	2
+
+struct dpsw_cmd_if_reflection {
+	__le16 if_id;
+	__le16 vlan_id;
+	/* only 2 bits from the LSB */
+	u8 filter;
+};
 #pragma pack(pop)
 #endif /* __FSL_DPSW_CMD_H */
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpsw.c b/drivers/net/ethernet/freescale/dpaa2/dpsw.c
index 6352d6d1ecba..ab921d75deb2 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpsw.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpsw.c
@@ -1579,3 +1579,83 @@ int dpsw_acl_remove_entry(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
 	/* send command to mc*/
 	return mc_send_command(mc_io, &cmd);
 }
+
+/**
+ * dpsw_set_reflection_if() - Set target interface for traffic mirrored
+ * @mc_io:	Pointer to MC portal's I/O object
+ * @cmd_flags:	Command flags; one or more of 'MC_CMD_FLAG_'
+ * @token:	Token of DPSW object
+ * @if_id:	Interface Id
+ *
+ * Only one mirroring destination is allowed per switch
+ *
+ * Return:	Completion status. '0' on Success; Error code otherwise.
+ */
+int dpsw_set_reflection_if(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			   u16 if_id)
+{
+	struct dpsw_cmd_set_reflection_if *cmd_params;
+	struct fsl_mc_command cmd = { 0 };
+
+	cmd.header = mc_encode_cmd_header(DPSW_CMDID_SET_REFLECTION_IF,
+					  cmd_flags,
+					  token);
+	cmd_params = (struct dpsw_cmd_set_reflection_if *)cmd.params;
+	cmd_params->if_id = cpu_to_le16(if_id);
+
+	return mc_send_command(mc_io, &cmd);
+}
+
+/**
+ * dpsw_if_add_reflection() - Setup mirroring rule
+ * @mc_io:	Pointer to MC portal's I/O object
+ * @cmd_flags:	Command flags; one or more of 'MC_CMD_FLAG_'
+ * @token:	Token of DPSW object
+ * @if_id:	Interface Identifier
+ * @cfg:	Reflection configuration
+ *
+ * Return:	Completion status. '0' on Success; Error code otherwise.
+ */
+int dpsw_if_add_reflection(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			   u16 if_id, const struct dpsw_reflection_cfg *cfg)
+{
+	struct dpsw_cmd_if_reflection *cmd_params;
+	struct fsl_mc_command cmd = { 0 };
+
+	cmd.header = mc_encode_cmd_header(DPSW_CMDID_IF_ADD_REFLECTION,
+					  cmd_flags,
+					  token);
+	cmd_params = (struct dpsw_cmd_if_reflection *)cmd.params;
+	cmd_params->if_id = cpu_to_le16(if_id);
+	cmd_params->vlan_id = cpu_to_le16(cfg->vlan_id);
+	dpsw_set_field(cmd_params->filter, FILTER, cfg->filter);
+
+	return mc_send_command(mc_io, &cmd);
+}
+
+/**
+ * dpsw_if_remove_reflection() - Remove mirroring rule
+ * @mc_io:	Pointer to MC portal's I/O object
+ * @cmd_flags:	Command flags; one or more of 'MC_CMD_FLAG_'
+ * @token:	Token of DPSW object
+ * @if_id:	Interface Identifier
+ * @cfg:	Reflection configuration
+ *
+ * Return:	Completion status. '0' on Success; Error code otherwise.
+ */
+int dpsw_if_remove_reflection(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			      u16 if_id, const struct dpsw_reflection_cfg *cfg)
+{
+	struct dpsw_cmd_if_reflection *cmd_params;
+	struct fsl_mc_command cmd = { 0 };
+
+	cmd.header = mc_encode_cmd_header(DPSW_CMDID_IF_REMOVE_REFLECTION,
+					  cmd_flags,
+					  token);
+	cmd_params = (struct dpsw_cmd_if_reflection *)cmd.params;
+	cmd_params->if_id = cpu_to_le16(if_id);
+	cmd_params->vlan_id = cpu_to_le16(cfg->vlan_id);
+	dpsw_set_field(cmd_params->filter, FILTER, cfg->filter);
+
+	return mc_send_command(mc_io, &cmd);
+}
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpsw.h b/drivers/net/ethernet/freescale/dpaa2/dpsw.h
index 5ef221a25b02..892df905b876 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpsw.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpsw.h
@@ -752,4 +752,35 @@ int dpsw_acl_add_entry(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
 
 int dpsw_acl_remove_entry(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
 			  u16 acl_id, const struct dpsw_acl_entry_cfg *cfg);
+
+/**
+ * enum dpsw_reflection_filter - Filter type for frames to be reflected
+ * @DPSW_REFLECTION_FILTER_INGRESS_ALL: Reflect all frames
+ * @DPSW_REFLECTION_FILTER_INGRESS_VLAN: Reflect only frames that belong to
+ *	the particular VLAN defined by vid parameter
+ *
+ */
+enum dpsw_reflection_filter {
+	DPSW_REFLECTION_FILTER_INGRESS_ALL = 0,
+	DPSW_REFLECTION_FILTER_INGRESS_VLAN = 1
+};
+
+/**
+ * struct dpsw_reflection_cfg - Structure representing the mirroring config
+ * @filter: Filter type for frames to be mirrored
+ * @vlan_id: VLAN ID to mirror; valid only when the type is DPSW_INGRESS_VLAN
+ */
+struct dpsw_reflection_cfg {
+	enum dpsw_reflection_filter filter;
+	u16 vlan_id;
+};
+
+int dpsw_set_reflection_if(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			   u16 if_id);
+
+int dpsw_if_add_reflection(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			   u16 if_id, const struct dpsw_reflection_cfg *cfg);
+
+int dpsw_if_remove_reflection(struct fsl_mc_io *mc_io, u32 cmd_flags, u16 token,
+			      u16 if_id, const struct dpsw_reflection_cfg *cfg);
 #endif /* __FSL_DPSW_H */
-- 
2.31.1

[PATCH net-next 6/9] dpaa2-switch: add support for port mirroring

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Add support for per port mirroring for the DPAA2 switch. We support
only single mirror port, therefore we allow mirroring rules only as long
as the destination port is always the same.

Unlike all the actions (drop, redirect, trap) already supported by the
dpaa2-switch driver, adding mirroring filters in shared blocks is not
achieved by a singular ACL entry added in a table shared by the ports.
This is why, when a new mirror filter is added in a block we have to got
through all the switch ports sharing it and configure the filter
individually on all.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 171 +++++++++++++++++-
 .../ethernet/freescale/dpaa2/dpaa2-switch.c   |   6 +
 .../ethernet/freescale/dpaa2/dpaa2-switch.h   |  10 +
 3 files changed, 182 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index 637291060fd5..efd6d58ca191 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -305,6 +305,19 @@ dpaa2_switch_acl_entry_get_index(struct dpaa2_switch_filter_block *block,
 	return -ENOENT;
 }
 
+static struct dpaa2_switch_mirror_entry *
+dpaa2_switch_mirror_find_entry_by_cookie(struct dpaa2_switch_filter_block *block,
+					 unsigned long cookie)
+{
+	struct dpaa2_switch_mirror_entry *tmp, *n;
+
+	list_for_each_entry_safe(tmp, n, &block->mirror_entries, list) {
+		if (tmp->cookie == cookie)
+			return tmp;
+	}
+	return NULL;
+}
+
 static int
 dpaa2_switch_acl_tbl_remove_entry(struct dpaa2_switch_filter_block *block,
 				  struct dpaa2_switch_acl_entry *entry)
@@ -376,6 +389,83 @@ static int dpaa2_switch_tc_parse_action_acl(struct ethsw_core *ethsw,
 	return err;
 }
 
+static int
+dpaa2_switch_block_add_mirror(struct dpaa2_switch_filter_block *block,
+			      struct dpaa2_switch_mirror_entry *entry,
+			      u16 to, struct netlink_ext_ack *extack)
+{
+	unsigned long block_ports = block->ports;
+	struct ethsw_core *ethsw = block->ethsw;
+	unsigned long ports_added = 0;
+	bool mirror_port_enabled;
+	int err, port;
+
+	/* Setup the mirroring port */
+	mirror_port_enabled = (ethsw->mirror_port != ethsw->sw_attr.num_ifs);
+	if (!mirror_port_enabled) {
+		err = dpsw_set_reflection_if(ethsw->mc_io, 0,
+					     ethsw->dpsw_handle, to);
+		if (err)
+			return err;
+		ethsw->mirror_port = to;
+	}
+
+	/* Setup the same egress mirroring configuration on all the switch
+	 * ports that share the same filter block.
+	 */
+	for_each_set_bit(port, &block_ports, ethsw->sw_attr.num_ifs) {
+		err = dpsw_if_add_reflection(ethsw->mc_io, 0,
+					     ethsw->dpsw_handle,
+					     port, &entry->cfg);
+		if (err)
+			goto err_remove_filters;
+
+		ports_added |= BIT(port);
+	}
+
+	list_add(&entry->list, &block->mirror_entries);
+
+	return 0;
+
+err_remove_filters:
+	for_each_set_bit(port, &ports_added, ethsw->sw_attr.num_ifs) {
+		dpsw_if_remove_reflection(ethsw->mc_io, 0, ethsw->dpsw_handle,
+					  port, &entry->cfg);
+	}
+
+	if (!mirror_port_enabled)
+		ethsw->mirror_port = ethsw->sw_attr.num_ifs;
+
+	return err;
+}
+
+static int
+dpaa2_switch_block_remove_mirror(struct dpaa2_switch_filter_block *block,
+				 struct dpaa2_switch_mirror_entry *entry)
+{
+	struct dpsw_reflection_cfg *cfg = &entry->cfg;
+	unsigned long block_ports = block->ports;
+	struct ethsw_core *ethsw = block->ethsw;
+	int port;
+
+	/* Remove this mirroring configuration from all the ports belonging to
+	 * the filter block.
+	 */
+	for_each_set_bit(port, &block_ports, ethsw->sw_attr.num_ifs)
+		dpsw_if_remove_reflection(ethsw->mc_io, 0, ethsw->dpsw_handle,
+					  port, cfg);
+
+	/* Also remove it from the list of mirror filters */
+	list_del(&entry->list);
+	kfree(entry);
+
+	/* If this was the last mirror filter, then unset the mirror port */
+	if (list_empty(&block->mirror_entries))
+		ethsw->mirror_port =  ethsw->sw_attr.num_ifs;
+
+	return 0;
+}
+
 static int
 dpaa2_switch_cls_flower_replace_acl(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
@@ -497,6 +587,64 @@ dpaa2_switch_cls_matchall_replace_acl(struct dpaa2_switch_filter_block *block,
 	return err;
 }
 
+static int
+dpaa2_switch_cls_matchall_replace_mirror(struct dpaa2_switch_filter_block *block,
+					 struct tc_cls_matchall_offload *cls)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct dpaa2_switch_mirror_entry *mirror_entry;
+	struct ethsw_core *ethsw = block->ethsw;
+	struct dpaa2_switch_mirror_entry *tmp;
+	struct flow_action_entry *cls_act;
+	struct list_head *pos, *n;
+	bool mirror_port_enabled;
+	u16 if_id;
+
+	mirror_port_enabled = (ethsw->mirror_port != ethsw->sw_attr.num_ifs);
+	cls_act = &cls->rule->action.entries[0];
+
+	/* Offload rules only when the destination is a DPAA2 switch port */
+	if (!dpaa2_switch_port_dev_check(cls_act->dev)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Destination not a DPAA2 switch port");
+		return -EOPNOTSUPP;
+	}
+	if_id = dpaa2_switch_get_index(ethsw, cls_act->dev);
+
+	/* We have a single mirror port but can configure egress mirroring on
+	 * all the other switch ports. We need to allow mirroring rules only
+	 * when the destination port is the same.
+	 */
+	if (mirror_port_enabled && ethsw->mirror_port != if_id) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Multiple mirror ports not supported");
+		return -EBUSY;
+	}
+
+	/* Make sure that we don't already have a mirror rule with the same
+	 * configuration. One matchall rule per block is the maximum.
+	 */
+	list_for_each_safe(pos, n, &block->mirror_entries) {
+		tmp = list_entry(pos, struct dpaa2_switch_mirror_entry, list);
+
+		if (tmp->cfg.filter == DPSW_REFLECTION_FILTER_INGRESS_ALL) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Matchall mirror filter already installed");
+			return -EBUSY;
+		}
+	}
+
+	mirror_entry = kzalloc(sizeof(*mirror_entry), GFP_KERNEL);
+	if (!mirror_entry)
+		return -ENOMEM;
+
+	mirror_entry->cfg.filter = DPSW_REFLECTION_FILTER_INGRESS_ALL;
+	mirror_entry->cookie = cls->cookie;
+
+	return dpaa2_switch_block_add_mirror(block, mirror_entry, if_id,
+					     extack);
+}
+
 int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
@@ -514,6 +662,8 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 	case FLOW_ACTION_TRAP:
 	case FLOW_ACTION_DROP:
 		return dpaa2_switch_cls_matchall_replace_acl(block, cls);
+	case FLOW_ACTION_MIRRED:
+		return dpaa2_switch_cls_matchall_replace_mirror(block, cls);
 	default:
 		NL_SET_ERR_MSG_MOD(extack, "Action not supported");
 		return -EOPNOTSUPP;
@@ -523,11 +673,22 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
-	struct dpaa2_switch_acl_entry *entry;
+	struct dpaa2_switch_mirror_entry *mirror_entry;
+	struct dpaa2_switch_acl_entry *acl_entry;
 
-	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block, cls->cookie);
-	if (!entry)
-		return 0;
+	/* If this filter is a an ACL one, remove it */
+	acl_entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block,
+							      cls->cookie);
+	if (acl_entry)
+		return dpaa2_switch_acl_tbl_remove_entry(block,
+							 acl_entry);
 
-	return  dpaa2_switch_acl_tbl_remove_entry(block, entry);
+	/* If not, then it has to be a mirror */
+	mirror_entry = dpaa2_switch_mirror_find_entry_by_cookie(block,
+								cls->cookie);
+	if (mirror_entry)
+		return dpaa2_switch_block_remove_mirror(block,
+							mirror_entry);
+
+	return 0;
 }
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
index 1806012f41d2..3857d9093623 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
@@ -3067,6 +3067,7 @@ static int dpaa2_switch_port_init(struct ethsw_port_priv *port_priv, u16 port)
 	filter_block->in_use = true;
 	filter_block->num_acl_rules = 0;
 	INIT_LIST_HEAD(&filter_block->acl_entries);
+	INIT_LIST_HEAD(&filter_block->mirror_entries);
 
 	err = dpaa2_switch_port_acl_tbl_bind(port_priv, filter_block);
 	if (err)
@@ -3284,6 +3285,11 @@ static int dpaa2_switch_probe(struct fsl_mc_device *sw_dev)
 	if (err)
 		goto err_stop;
 
+	/* By convention, if the mirror port is equal to the number of switch
+	 * interfaces, then mirroring of any kind is disabled.
+	 */
+	ethsw->mirror_port =  ethsw->sw_attr.num_ifs;
+
 	/* Register the netdev only when the entire setup is done and the
 	 * switch port interfaces are ready to receive traffic
 	 */
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
index 296a09eb7a9a..79e8a40f97f7 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
@@ -113,6 +113,13 @@ struct dpaa2_switch_acl_entry {
 	struct dpsw_acl_key	key;
 };
 
+struct dpaa2_switch_mirror_entry {
+	struct list_head	list;
+	struct dpsw_reflection_cfg cfg;
+	unsigned long		cookie;
+	u16 if_id;
+};
+
 struct dpaa2_switch_filter_block {
 	struct ethsw_core	*ethsw;
 	u64			ports;
@@ -121,6 +128,8 @@ struct dpaa2_switch_filter_block {
 	struct list_head	acl_entries;
 	u16			acl_id;
 	u8			num_acl_rules;
+
+	struct list_head	mirror_entries;
 };
 
 static inline bool
@@ -176,6 +185,7 @@ struct ethsw_core {
 
 	struct dpaa2_switch_fdb		*fdbs;
 	struct dpaa2_switch_filter_block *filter_blocks;
+	u16				mirror_port;
 };
 
 static inline int dpaa2_switch_get_index(struct ethsw_core *ethsw,
-- 
2.31.1

[PATCH net-next 7/9] dpaa2-switch: add VLAN based mirroring

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Using the infrastructure added in the previous patch, extend tc-flower
support with FLOW_ACTION_MIRRED based on VLAN.

Tested with:

tc qdisc add dev eth8 ingress_block 1 clsact
tc filter add block 1 ingress protocol 802.1q flower skip_sw \
	vlan_id 100 action mirred egress mirror dev eth6
tc filter del block 1 ingress pref 49152

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 143 +++++++++++++++++-
 1 file changed, 138 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index efd6d58ca191..3c4f5ada12fd 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -396,7 +396,9 @@ dpaa2_switch_block_add_mirror(struct dpaa2_switch_filter_block *block,
 {
 	unsigned long block_ports = block->ports;
 	struct ethsw_core *ethsw = block->ethsw;
+	struct ethsw_port_priv *port_priv;
 	unsigned long ports_added = 0;
+	u16 vlan = entry->cfg.vlan_id;
 	bool mirror_port_enabled;
 	int err, port;
 
@@ -414,6 +416,19 @@ dpaa2_switch_block_add_mirror(struct dpaa2_switch_filter_block *block,
 	 * ports that share the same filter block.
 	 */
 	for_each_set_bit(port, &block_ports, ethsw->sw_attr.num_ifs) {
+		port_priv = ethsw->ports[port];
+
+		/* We cannot add a per VLAN mirroring rule if the VLAN in
+		 * question is not installed on the switch port.
+		 */
+		if (entry->cfg.filter == DPSW_REFLECTION_FILTER_INGRESS_VLAN &&
+		    !(port_priv->vlans[vlan] & ETHSW_VLAN_MEMBER)) {
+			NL_SET_ERR_MSG(extack,
+				       "VLAN must be installed on the switch port");
+			err = -EINVAL;
+			goto err_remove_filters;
+		}
+
 		err = dpsw_if_add_reflection(ethsw->mc_io, 0,
 					     ethsw->dpsw_handle,
 					     port, &entry->cfg);
@@ -511,6 +526,112 @@ dpaa2_switch_cls_flower_replace_acl(struct dpaa2_switch_filter_block *block,
 	return err;
 }
 
+static int dpaa2_switch_flower_parse_mirror_key(struct flow_cls_offload *cls,
+						u16 *vlan)
+{
+	struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
+	struct flow_dissector *dissector = rule->match.dissector;
+	struct netlink_ext_ack *extack = cls->common.extack;
+
+	if (dissector->used_keys &
+	    ~(BIT(FLOW_DISSECTOR_KEY_BASIC) |
+	      BIT(FLOW_DISSECTOR_KEY_CONTROL) |
+	      BIT(FLOW_DISSECTOR_KEY_VLAN))) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Mirroring is supported only per VLAN");
+		return -EOPNOTSUPP;
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) {
+		struct flow_match_vlan match;
+
+		flow_rule_match_vlan(rule, &match);
+
+		if (match.mask->vlan_priority != 0 ||
+		    match.mask->vlan_dei != 0) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Only matching on VLAN ID supported");
+			return -EOPNOTSUPP;
+		}
+
+		if (match.mask->vlan_id != 0xFFF) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Masked matching not supported");
+			return -EOPNOTSUPP;
+		}
+
+		*vlan = (u16)match.key->vlan_id;
+	}
+
+	return 0;
+}
+
+static int
+dpaa2_switch_cls_flower_replace_mirror(struct dpaa2_switch_filter_block *block,
+				       struct flow_cls_offload *cls)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct dpaa2_switch_mirror_entry *mirror_entry;
+	struct ethsw_core *ethsw = block->ethsw;
+	struct dpaa2_switch_mirror_entry *tmp;
+	struct flow_action_entry *cls_act;
+	struct list_head *pos, *n;
+	bool mirror_port_enabled;
+	u16 if_id, vlan;
+	int err;
+
+	mirror_port_enabled = (ethsw->mirror_port != ethsw->sw_attr.num_ifs);
+	cls_act = &cls->rule->action.entries[0];
+
+	/* Offload rules only when the destination is a DPAA2 switch port */
+	if (!dpaa2_switch_port_dev_check(cls_act->dev)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Destination not a DPAA2 switch port");
+		return -EOPNOTSUPP;
+	}
+	if_id = dpaa2_switch_get_index(ethsw, cls_act->dev);
+
+	/* We have a single mirror port but can configure egress mirroring on
+	 * all the other switch ports. We need to allow mirroring rules only
+	 * when the destination port is the same.
+	 */
+	if (mirror_port_enabled && ethsw->mirror_port != if_id) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Multiple mirror ports not supported");
+		return -EBUSY;
+	}
+
+	/* Parse the key */
+	err = dpaa2_switch_flower_parse_mirror_key(cls, &vlan);
+	if (err)
+		return err;
+
+	/* Make sure that we don't already have a mirror rule with the same
+	 * configuration.
+	 */
+	list_for_each_safe(pos, n, &block->mirror_entries) {
+		tmp = list_entry(pos, struct dpaa2_switch_mirror_entry, list);
+
+		if (tmp->cfg.filter == DPSW_REFLECTION_FILTER_INGRESS_VLAN &&
+		    tmp->cfg.vlan_id == vlan) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "VLAN mirror filter already installed");
+			return -EBUSY;
+		}
+	}
+
+	mirror_entry = kzalloc(sizeof(*mirror_entry), GFP_KERNEL);
+	if (!mirror_entry)
+		return -ENOMEM;
+
+	mirror_entry->cfg.filter = DPSW_REFLECTION_FILTER_INGRESS_VLAN;
+	mirror_entry->cfg.vlan_id = vlan;
+	mirror_entry->cookie = cls->cookie;
+
+	return dpaa2_switch_block_add_mirror(block, mirror_entry, if_id,
+					     extack);
+}
+
 int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
@@ -529,6 +650,8 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 	case FLOW_ACTION_TRAP:
 	case FLOW_ACTION_DROP:
 		return dpaa2_switch_cls_flower_replace_acl(block, cls);
+	case FLOW_ACTION_MIRRED:
+		return dpaa2_switch_cls_flower_replace_mirror(block, cls);
 	default:
 		NL_SET_ERR_MSG_MOD(extack, "Action not supported");
 		return -EOPNOTSUPP;
@@ -538,13 +661,23 @@ int dpaa2_switch_cls_flower_replace(struct dpaa2_switch_filter_block *block,
 int dpaa2_switch_cls_flower_destroy(struct dpaa2_switch_filter_block *block,
 				    struct flow_cls_offload *cls)
 {
-	struct dpaa2_switch_acl_entry *entry;
+	struct dpaa2_switch_mirror_entry *mirror_entry;
+	struct dpaa2_switch_acl_entry *acl_entry;
 
-	entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block, cls->cookie);
-	if (!entry)
-		return 0;
+	/* If this filter is a an ACL one, remove it */
+	acl_entry = dpaa2_switch_acl_tbl_find_entry_by_cookie(block,
+							      cls->cookie);
+	if (acl_entry)
+		return dpaa2_switch_acl_tbl_remove_entry(block, acl_entry);
+
+	/* If not, then it has to be a mirror */
+	mirror_entry = dpaa2_switch_mirror_find_entry_by_cookie(block,
+								cls->cookie);
+	if (mirror_entry)
+		return dpaa2_switch_block_remove_mirror(block,
+							mirror_entry);
 
-	return dpaa2_switch_acl_tbl_remove_entry(block, entry);
+	return 0;
 }
 
 static int
-- 
2.31.1

[PATCH net-next 8/9] dpaa2-switch: offload shared block mirror filters when binding to a port

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

When mirroring rules are added in shared filter blocks, the same
mirroring rule has to be configured on all the switch ports that are
part of the same block.

In case a switch port joins a shared block after mirroring filters have
been already added to it, then all the mirror rules should be offloaded
to the port. The reverse, removal of mirroring rules, has to be done at
block unbind.

For this purpose, the dpaa2_switch_block_offload_mirror() and
dpaa2_switch_block_unoffload_mirror() functions are added and called
upon binding and unbinding a switch port to/from a block.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/dpaa2-switch-flower.c     | 51 +++++++++++++++++++
 .../ethernet/freescale/dpaa2/dpaa2-switch.c   | 14 +++++
 .../ethernet/freescale/dpaa2/dpaa2-switch.h   |  6 +++
 3 files changed, 71 insertions(+)

diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
index 3c4f5ada12fd..d6eefbbf163f 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch-flower.c
@@ -803,6 +803,57 @@ int dpaa2_switch_cls_matchall_replace(struct dpaa2_switch_filter_block *block,
 	}
 }
 
+int dpaa2_switch_block_offload_mirror(struct dpaa2_switch_filter_block *block,
+				      struct ethsw_port_priv *port_priv)
+{
+	struct ethsw_core *ethsw = port_priv->ethsw_data;
+	struct dpaa2_switch_mirror_entry *tmp;
+	int err;
+
+	list_for_each_entry(tmp, &block->mirror_entries, list) {
+		err = dpsw_if_add_reflection(ethsw->mc_io, 0,
+					     ethsw->dpsw_handle,
+					     port_priv->idx, &tmp->cfg);
+		if (err)
+			goto unwind_add;
+	}
+
+	return 0;
+
+unwind_add:
+	list_for_each_entry(tmp, &block->mirror_entries, list)
+		dpsw_if_remove_reflection(ethsw->mc_io, 0,
+					  ethsw->dpsw_handle,
+					  port_priv->idx, &tmp->cfg);
+
+	return err;
+}
+
+int dpaa2_switch_block_unoffload_mirror(struct dpaa2_switch_filter_block *block,
+					struct ethsw_port_priv *port_priv)
+{
+	struct ethsw_core *ethsw = port_priv->ethsw_data;
+	struct dpaa2_switch_mirror_entry *tmp;
+	int err;
+
+	list_for_each_entry(tmp, &block->mirror_entries, list) {
+		err = dpsw_if_remove_reflection(ethsw->mc_io, 0,
+						ethsw->dpsw_handle,
+						port_priv->idx, &tmp->cfg);
+		if (err)
+			goto unwind_remove;
+	}
+
+	return 0;
+
+unwind_remove:
+	list_for_each_entry(tmp, &block->mirror_entries, list)
+		dpsw_if_add_reflection(ethsw->mc_io, 0, ethsw->dpsw_handle,
+				       port_priv->idx, &tmp->cfg);
+
+	return err;
+}
+
 int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 				      struct tc_cls_matchall_offload *cls)
 {
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
index 3857d9093623..71129724d9ca 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.c
@@ -1229,6 +1229,13 @@ static int dpaa2_switch_port_block_bind(struct ethsw_port_priv *port_priv,
 	struct dpaa2_switch_filter_block *old_block = port_priv->filter_block;
 	int err;
 
+	/* Offload all the mirror entries found in the block on this new port
+	 * joining it.
+	 */
+	err = dpaa2_switch_block_offload_mirror(block, port_priv);
+	if (err)
+		return err;
+
 	/* If the port is already bound to this ACL table then do nothing. This
 	 * can happen when this port is the first one to join a tc block
 	 */
@@ -1256,6 +1263,13 @@ dpaa2_switch_port_block_unbind(struct ethsw_port_priv *port_priv,
 	struct dpaa2_switch_filter_block *new_block;
 	int err;
 
+	/* Unoffload all the mirror entries found in the block from the
+	 * port leaving it.
+	 */
+	err = dpaa2_switch_block_unoffload_mirror(block, port_priv);
+	if (err)
+		return err;
+
 	/* We are the last port that leaves a block (an ACL table).
 	 * We'll continue to use this table.
 	 */
diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
index 79e8a40f97f7..f69d940f3c5b 100644
--- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
+++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-switch.h
@@ -253,4 +253,10 @@ int dpaa2_switch_cls_matchall_destroy(struct dpaa2_switch_filter_block *block,
 
 int dpaa2_switch_acl_entry_add(struct dpaa2_switch_filter_block *block,
 			       struct dpaa2_switch_acl_entry *entry);
+
+int dpaa2_switch_block_offload_mirror(struct dpaa2_switch_filter_block *block,
+				      struct ethsw_port_priv *port_priv);
+
+int dpaa2_switch_block_unoffload_mirror(struct dpaa2_switch_filter_block *block,
+					struct ethsw_port_priv *port_priv);
 #endif	/* __ETHSW_H */
-- 
2.31.1

[PATCH net-next 9/9] docs: networking: dpaa2: document mirroring support on the switch

[Ioana Ciornei]

From: Ioana Ciornei <ioana.ciornei@nxp.com>

Document the mirroring capabilities of the dpaa2-switch driver,
any restrictions that are imposed and some example commands.

Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
---
 .../freescale/dpaa2/switch-driver.rst         | 43 +++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/Documentation/networking/device_drivers/ethernet/freescale/dpaa2/switch-driver.rst b/Documentation/networking/device_drivers/ethernet/freescale/dpaa2/switch-driver.rst
index 863ca6bd8318..8bf411b857d4 100644
--- a/Documentation/networking/device_drivers/ethernet/freescale/dpaa2/switch-driver.rst
+++ b/Documentation/networking/device_drivers/ethernet/freescale/dpaa2/switch-driver.rst
@@ -172,3 +172,46 @@ Example 4: Use a single shared filter block on both eth5 and eth6::
                 action trap
         $ tc filter add block 1 ingress protocol ipv4 flower src_ip 192.168.1.1 skip_sw \
                 action mirred egress redirect dev eth3
+
+Mirroring
+~~~~~~~~~
+
+The DPAA2 switch supports only per port mirroring and per VLAN mirroring.
+Adding mirroring filters in shared blocks is also supported.
+
+When using the tc-flower classifier with the 802.1q protocol, only the
+''vlan_id'' key will be accepted. Mirroring based on any other fields from the
+802.1q protocol will be rejected::
+
+        $ tc qdisc add dev eth8 ingress_block 1 clsact
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_prio 3 action mirred egress mirror dev eth6
+        Error: fsl_dpaa2_switch: Only matching on VLAN ID supported.
+        We have an error talking to the kernel
+
+If a mirroring VLAN filter is requested on a port, the VLAN must to be
+installed on the switch port in question either using ''bridge'' or by creating
+a VLAN upper device if the switch port is used as a standalone interface::
+
+        $ tc qdisc add dev eth8 ingress_block 1 clsact
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_id 200 action mirred egress mirror dev eth6
+        Error: VLAN must be installed on the switch port.
+        We have an error talking to the kernel
+
+        $ bridge vlan add vid 200 dev eth8
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_id 200 action mirred egress mirror dev eth6
+
+        $ ip link add link eth8 name eth8.200 type vlan id 200
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_id 200 action mirred egress mirror dev eth6
+
+Also, it should be noted that the mirrored traffic will be subject to the same
+egress restrictions as any other traffic. This means that when a mirrored
+packet will reach the mirror port, if the VLAN found in the packet is not
+installed on the port it will get dropped.
+
+The DPAA2 switch supports only a single mirroring destination, thus multiple
+mirror rules can be installed but their ''to'' port has to be the same::
+
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_id 200 action mirred egress mirror dev eth6
+        $ tc filter add block 1 ingress protocol 802.1q flower skip_sw vlan_id 100 action mirred egress mirror dev eth7
+        Error: fsl_dpaa2_switch: Multiple mirror ports not supported.
+        We have an error talking to the kernel
-- 
2.31.1

Re: [PATCH net-next 0/9] dpaa2-switch: add mirroring support

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (refs/heads/master):

On Thu, 29 Jul 2021 20:18:52 +0300 you wrote:
> From: Ioana Ciornei <ioana.ciornei@nxp.com>
> 
> This patch set adds per port and per VLAN mirroring in dpaa2-switch.
> 
> The first 4 patches are just cosmetic changes. We renamed the
> dpaa2_switch_acl_tbl structure into dpaa2_switch_filter_block so that we
> can reuse it for filters that do not use the ACL table and reorganized
> the addition of trap, redirect and drop filters into a separate
> function. All this just to make for a more streamlined addition of the
> support for mirroring.
> 
> [...]

Here is the summary with links:
  - [net-next,1/9] dpaa2-switch: rename dpaa2_switch_tc_parse_action to specify the ACL
    https://git.kernel.org/netdev/net-next/c/3b5d8b448602
  - [net-next,2/9] dpaa2-switch: rename dpaa2_switch_acl_tbl into filter_block
    https://git.kernel.org/netdev/net-next/c/adcb7aa335af
  - [net-next,3/9] dpaa2-switch: reorganize dpaa2_switch_cls_flower_replace
    https://git.kernel.org/netdev/net-next/c/c5f6d490c578
  - [net-next,4/9] dpaa2-switch: reorganize dpaa2_switch_cls_matchall_replace
    https://git.kernel.org/netdev/net-next/c/3fa5514a2966
  - [net-next,5/9] dpaa2-switch: add API for setting up mirroring
    https://git.kernel.org/netdev/net-next/c/cbc2a8893b59
  - [net-next,6/9] dpaa2-switch: add support for port mirroring
    https://git.kernel.org/netdev/net-next/c/e0ead825a1f1
  - [net-next,7/9] dpaa2-switch: add VLAN based mirroring
    https://git.kernel.org/netdev/net-next/c/0f3faece5808
  - [net-next,8/9] dpaa2-switch: offload shared block mirror filters when binding to a port
    https://git.kernel.org/netdev/net-next/c/7a91f9078d4f
  - [net-next,9/9] docs: networking: dpaa2: document mirroring support on the switch
    https://git.kernel.org/netdev/net-next/c/d1626a1c273d

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