DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors
@ 2021-10-25 11:04 Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 2/7] net/sfc: rename ethdev m-port retrieval helper Ivan Malov
                   ` (6 more replies)
  0 siblings, 7 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko, Andy Moreton, Igor Romanov

VF representors do not own dedicated m-ports and thus cannot
be referred to as traffic endpoints in flow items or actions.

Fixes: a62ec90522a6 ("net/sfc: add port representors infrastructure")
Fixes: f55b61cec94a ("net/sfc: support port representor flow item")

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 drivers/net/sfc/sfc_switch.c | 22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/net/sfc/sfc_switch.c b/drivers/net/sfc/sfc_switch.c
index a28e861de5..265a17f4c4 100644
--- a/drivers/net/sfc/sfc_switch.c
+++ b/drivers/net/sfc/sfc_switch.c
@@ -512,7 +512,7 @@ sfc_mae_clear_switch_port(uint16_t switch_domain_id,
 static int
 sfc_mae_find_switch_port_by_ethdev(uint16_t switch_domain_id,
 				   uint16_t ethdev_port_id,
-				   efx_mport_sel_t *mport_sel)
+				   struct sfc_mae_switch_port **switch_port)
 {
 	struct sfc_mae_switch_domain *domain;
 	struct sfc_mae_switch_port *port;
@@ -528,7 +528,7 @@ sfc_mae_find_switch_port_by_ethdev(uint16_t switch_domain_id,
 
 	TAILQ_FOREACH(port, &domain->ports, switch_domain_ports) {
 		if (port->ethdev_port_id == ethdev_port_id) {
-			*mport_sel = port->ethdev_mport;
+			*switch_port = port;
 			return 0;
 		}
 	}
@@ -541,11 +541,27 @@ sfc_mae_switch_port_by_ethdev(uint16_t switch_domain_id,
 			      uint16_t ethdev_port_id,
 			      efx_mport_sel_t *mport_sel)
 {
+	struct sfc_mae_switch_port *port;
 	int rc;
 
 	rte_spinlock_lock(&sfc_mae_switch.lock);
 	rc = sfc_mae_find_switch_port_by_ethdev(switch_domain_id,
-						ethdev_port_id, mport_sel);
+						ethdev_port_id, &port);
+	if (rc != 0)
+		goto unlock;
+
+	if (port->type != SFC_MAE_SWITCH_PORT_INDEPENDENT) {
+		/*
+		 * The ethdev is a "VF representor". It does not own
+		 * a dedicated m-port suitable for use in flow rules.
+		 */
+		rc = ENOTSUP;
+		goto unlock;
+	}
+
+	*mport_sel = port->ethdev_mport;
+
+unlock:
 	rte_spinlock_unlock(&sfc_mae_switch.lock);
 
 	return rc;
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 2/7] net/sfc: rename ethdev m-port retrieval helper
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 3/7] net/sfc: improve m-port related log messages Ivan Malov
                   ` (5 subsequent siblings)
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

The function in question has an unfortunate name that reads
like finding a SW switch port entry. In fact just one of
the two m-ports is retrieved from that entry.

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 drivers/net/sfc/sfc_mae.c    | 10 +++++-----
 drivers/net/sfc/sfc_switch.c |  6 +++---
 drivers/net/sfc/sfc_switch.h |  6 +++---
 3 files changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index 455744c570..84b13925ff 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -1276,8 +1276,8 @@ sfc_mae_rule_parse_item_port_id(const struct rte_flow_item *item,
 					  "The port ID is too large");
 	}
 
-	rc = sfc_mae_switch_port_by_ethdev(ctx_mae->sa->mae.switch_domain_id,
-					   spec->id, &mport_sel);
+	rc = sfc_mae_switch_get_ethdev_mport(ctx_mae->sa->mae.switch_domain_id,
+					     spec->id, &mport_sel);
 	if (rc != 0) {
 		return rte_flow_error_set(error, rc,
 				RTE_FLOW_ERROR_TYPE_ITEM, item,
@@ -1335,7 +1335,7 @@ sfc_mae_rule_parse_item_port_representor(const struct rte_flow_item *item,
 	if (spec == NULL)
 		return 0;
 
-	rc = sfc_mae_switch_port_by_ethdev(
+	rc = sfc_mae_switch_get_ethdev_mport(
 			ctx_mae->sa->mae.switch_domain_id,
 			spec->port_id, &mport_sel);
 	if (rc != 0) {
@@ -3406,8 +3406,8 @@ sfc_mae_rule_parse_action_port_id(struct sfc_adapter *sa,
 
 	port_id = (conf->original != 0) ? sas->port_id : conf->id;
 
-	rc = sfc_mae_switch_port_by_ethdev(mae->switch_domain_id,
-					   port_id, &mport);
+	rc = sfc_mae_switch_get_ethdev_mport(mae->switch_domain_id,
+					     port_id, &mport);
 	if (rc != 0) {
 		sfc_err(sa, "failed to find MAE switch port SW entry for RTE ethdev port %u: %s",
 			port_id, strerror(rc));
diff --git a/drivers/net/sfc/sfc_switch.c b/drivers/net/sfc/sfc_switch.c
index 265a17f4c4..3f7518fa30 100644
--- a/drivers/net/sfc/sfc_switch.c
+++ b/drivers/net/sfc/sfc_switch.c
@@ -537,9 +537,9 @@ sfc_mae_find_switch_port_by_ethdev(uint16_t switch_domain_id,
 }
 
 int
-sfc_mae_switch_port_by_ethdev(uint16_t switch_domain_id,
-			      uint16_t ethdev_port_id,
-			      efx_mport_sel_t *mport_sel)
+sfc_mae_switch_get_ethdev_mport(uint16_t switch_domain_id,
+				uint16_t ethdev_port_id,
+				efx_mport_sel_t *mport_sel)
 {
 	struct sfc_mae_switch_port *port;
 	int rc;
diff --git a/drivers/net/sfc/sfc_switch.h b/drivers/net/sfc/sfc_switch.h
index 7917141038..a5a0fb4fc5 100644
--- a/drivers/net/sfc/sfc_switch.h
+++ b/drivers/net/sfc/sfc_switch.h
@@ -102,9 +102,9 @@ int sfc_mae_assign_switch_port(uint16_t switch_domain_id,
 int sfc_mae_clear_switch_port(uint16_t switch_domain_id,
 			      uint16_t switch_port_id);
 
-int sfc_mae_switch_port_by_ethdev(uint16_t switch_domain_id,
-				  uint16_t ethdev_port_id,
-				  efx_mport_sel_t *mport_sel);
+int sfc_mae_switch_get_ethdev_mport(uint16_t switch_domain_id,
+				    uint16_t ethdev_port_id,
+				    efx_mport_sel_t *mport_sel);
 
 int sfc_mae_switch_port_id_by_entity(uint16_t switch_domain_id,
 				     const efx_mport_sel_t *entity_mportp,
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 3/7] net/sfc: improve m-port related log messages
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 2/7] net/sfc: rename ethdev m-port retrieval helper Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 4/7] net/sfc: assign correct m-ports to independent switch ports Ivan Malov
                   ` (4 subsequent siblings)
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

Make these messages more specific.

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 drivers/net/sfc/sfc_mae.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index 84b13925ff..a4a22f32c6 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -1281,7 +1281,7 @@ sfc_mae_rule_parse_item_port_id(const struct rte_flow_item *item,
 	if (rc != 0) {
 		return rte_flow_error_set(error, rc,
 				RTE_FLOW_ERROR_TYPE_ITEM, item,
-				"Can't find RTE ethdev by the port ID");
+				"Can't get m-port for the given ethdev");
 	}
 
 	rc = efx_mae_match_spec_mport_set(ctx_mae->match_spec,
@@ -1341,7 +1341,7 @@ sfc_mae_rule_parse_item_port_representor(const struct rte_flow_item *item,
 	if (rc != 0) {
 		return rte_flow_error_set(error, rc,
 				RTE_FLOW_ERROR_TYPE_ITEM, item,
-				"Can't find RTE ethdev by the port ID");
+				"Can't get m-port for the given ethdev");
 	}
 
 	rc = efx_mae_match_spec_mport_set(ctx_mae->match_spec,
@@ -3409,7 +3409,7 @@ sfc_mae_rule_parse_action_port_id(struct sfc_adapter *sa,
 	rc = sfc_mae_switch_get_ethdev_mport(mae->switch_domain_id,
 					     port_id, &mport);
 	if (rc != 0) {
-		sfc_err(sa, "failed to find MAE switch port SW entry for RTE ethdev port %u: %s",
+		sfc_err(sa, "failed to get m-port for the given ethdev (port_id=%u): %s",
 			port_id, strerror(rc));
 		return rc;
 	}
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 4/7] net/sfc: assign correct m-ports to independent switch ports
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 2/7] net/sfc: rename ethdev m-port retrieval helper Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 3/7] net/sfc: improve m-port related log messages Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 5/7] net/sfc: support represented port flow item Ivan Malov
                   ` (3 subsequent siblings)
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

In accordance with patches [1-4], MAE admin ethdev represents a
network port and not the PF which it sits on. Rework the way
how "ethdev" and "entity" m-ports are assigned in SW switch
port entries of independent ethdevs. Explain in comments.

[1] commit 081e42dab11d ("ethdev: add port representor item to flow API")
[2] commit 49863ae2bf95 ("ethdev: add represented port item to flow API")
[3] commit 8edb6bc0263e ("ethdev: add port representor action
to flow API")
[4] commit 88caad251c8d ("ethdev: add represented port action
to flow API")

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 drivers/net/sfc/sfc_mae.c | 41 ++++++++++++++++++++++++++++++++++++---
 1 file changed, 38 insertions(+), 3 deletions(-)

diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index a4a22f32c6..bd8a913a49 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -23,7 +23,7 @@
 #include "sfc_service.h"
 
 static int
-sfc_mae_assign_entity_mport(struct sfc_adapter *sa,
+sfc_mae_assign_ethdev_mport(struct sfc_adapter *sa,
 			    efx_mport_sel_t *mportp)
 {
 	const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
@@ -32,6 +32,35 @@ sfc_mae_assign_entity_mport(struct sfc_adapter *sa,
 					      mportp);
 }
 
+static int
+sfc_mae_assign_entity_mport(struct sfc_adapter *sa,
+			    efx_mport_sel_t *mportp)
+{
+	const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+	int rc = 0;
+
+	if (encp->enc_mae_admin) {
+		/*
+		 * This ethdev sits on MAE admin PF. The represented
+		 * entity is the network port assigned to that PF.
+		 */
+		rc = efx_mae_mport_by_phy_port(encp->enc_assigned_port, mportp);
+	} else {
+		/*
+		 * This ethdev sits on unprivileged PF / VF. The entity
+		 * represented by the ethdev can change dynamically
+		 * as MAE admin changes default traffic rules.
+		 *
+		 * For the sake of simplicity, do not fill in the m-port
+		 * and assume that flow rules should not be allowed to
+		 * reference the entity represented by this ethdev.
+		 */
+		efx_mae_mport_invalid(mportp);
+	}
+
+	return rc;
+}
+
 static int
 sfc_mae_counter_registry_init(struct sfc_mae_counter_registry *registry,
 			      uint32_t nb_counters_max)
@@ -184,6 +213,7 @@ sfc_mae_attach(struct sfc_adapter *sa)
 	struct sfc_adapter_shared * const sas = sfc_sa2shared(sa);
 	struct sfc_mae_switch_port_request switch_port_request = {0};
 	const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
+	efx_mport_sel_t ethdev_mport;
 	efx_mport_sel_t entity_mport;
 	struct sfc_mae *mae = &sa->mae;
 	struct sfc_mae_bounce_eh *bounce_eh = &mae->bounce_eh;
@@ -218,6 +248,11 @@ sfc_mae_attach(struct sfc_adapter *sa)
 		}
 	}
 
+	sfc_log_init(sa, "assign ethdev MPORT");
+	rc = sfc_mae_assign_ethdev_mport(sa, &ethdev_mport);
+	if (rc != 0)
+		goto fail_mae_assign_ethdev_mport;
+
 	sfc_log_init(sa, "assign entity MPORT");
 	rc = sfc_mae_assign_entity_mport(sa, &entity_mport);
 	if (rc != 0)
@@ -230,9 +265,8 @@ sfc_mae_attach(struct sfc_adapter *sa)
 
 	sfc_log_init(sa, "assign RTE switch port");
 	switch_port_request.type = SFC_MAE_SWITCH_PORT_INDEPENDENT;
+	switch_port_request.ethdev_mportp = &ethdev_mport;
 	switch_port_request.entity_mportp = &entity_mport;
-	/* RTE ethdev MPORT matches that of the entity for independent ports. */
-	switch_port_request.ethdev_mportp = &entity_mport;
 	switch_port_request.ethdev_port_id = sas->port_id;
 	switch_port_request.port_data.indep.mae_admin =
 		encp->enc_mae_admin == B_TRUE;
@@ -272,6 +306,7 @@ sfc_mae_attach(struct sfc_adapter *sa)
 fail_mae_assign_switch_port:
 fail_mae_assign_switch_domain:
 fail_mae_assign_entity_mport:
+fail_mae_assign_ethdev_mport:
 	if (encp->enc_mae_admin)
 		sfc_mae_counter_registry_fini(&mae->counter_registry);
 
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 5/7] net/sfc: support represented port flow item
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
                   ` (2 preceding siblings ...)
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 4/7] net/sfc: assign correct m-ports to independent switch ports Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 6/7] net/sfc: support port representor related flow actions Ivan Malov
                   ` (2 subsequent siblings)
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

Add support for item REPRESENTED_PORT to match on traffic entering
the embedded switch from the entity represented by the given
ethdev (network port or VF).

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 doc/guides/nics/features/sfc.ini |  1 +
 doc/guides/nics/sfc_efx.rst      |  2 ++
 drivers/net/sfc/sfc_mae.c        | 52 +++++++++++++++++++++++++-------
 drivers/net/sfc/sfc_switch.c     | 29 ++++++++++++++++++
 drivers/net/sfc/sfc_switch.h     |  4 +++
 5 files changed, 77 insertions(+), 11 deletions(-)

diff --git a/doc/guides/nics/features/sfc.ini b/doc/guides/nics/features/sfc.ini
index 7db868e59f..c830426eb2 100644
--- a/doc/guides/nics/features/sfc.ini
+++ b/doc/guides/nics/features/sfc.ini
@@ -53,6 +53,7 @@ port_id              = Y
 port_representor     = Y
 pppoed               = Y
 pppoes               = Y
+represented_port     = Y
 tcp                  = Y
 udp                  = Y
 vf                   = Y
diff --git a/doc/guides/nics/sfc_efx.rst b/doc/guides/nics/sfc_efx.rst
index 843c24991c..8dbd250b3c 100644
--- a/doc/guides/nics/sfc_efx.rst
+++ b/doc/guides/nics/sfc_efx.rst
@@ -194,6 +194,8 @@ Supported pattern items (***transfer*** rules):
 
 - PORT_REPRESENTOR (cannot repeat; conflicts with other traffic source items)
 
+- REPRESENTED_PORT (cannot repeat; conflicts with other traffic source items)
+
 - PORT_ID (cannot repeat; conflicts with other traffic source items)
 
 - PHY_PORT (cannot repeat; conflicts with other traffic source items)
diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index bd8a913a49..8ec4036275 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -1333,9 +1333,9 @@ sfc_mae_rule_parse_item_port_id(const struct rte_flow_item *item,
 }
 
 static int
-sfc_mae_rule_parse_item_port_representor(const struct rte_flow_item *item,
-					 struct sfc_flow_parse_ctx *ctx,
-					 struct rte_flow_error *error)
+sfc_mae_rule_parse_item_ethdev_based(const struct rte_flow_item *item,
+				     struct sfc_flow_parse_ctx *ctx,
+				     struct rte_flow_error *error)
 {
 	struct sfc_mae_parse_ctx *ctx_mae = ctx->mae;
 	const struct rte_flow_item_ethdev supp_mask = {
@@ -1363,20 +1363,38 @@ sfc_mae_rule_parse_item_port_representor(const struct rte_flow_item *item,
 	if (mask->port_id != supp_mask.port_id) {
 		return rte_flow_error_set(error, EINVAL,
 				RTE_FLOW_ERROR_TYPE_ITEM, item,
-				"Bad mask in the PORT_REPRESENTOR pattern item");
+				"Bad mask in the ethdev-based pattern item");
 	}
 
 	/* If "spec" is not set, could be any port ID */
 	if (spec == NULL)
 		return 0;
 
-	rc = sfc_mae_switch_get_ethdev_mport(
-			ctx_mae->sa->mae.switch_domain_id,
-			spec->port_id, &mport_sel);
-	if (rc != 0) {
-		return rte_flow_error_set(error, rc,
+	switch (item->type) {
+	case RTE_FLOW_ITEM_TYPE_PORT_REPRESENTOR:
+		rc = sfc_mae_switch_get_ethdev_mport(
+				ctx_mae->sa->mae.switch_domain_id,
+				spec->port_id, &mport_sel);
+		if (rc != 0) {
+			return rte_flow_error_set(error, rc,
+					RTE_FLOW_ERROR_TYPE_ITEM, item,
+					"Can't get m-port for the given ethdev");
+		}
+		break;
+	case RTE_FLOW_ITEM_TYPE_REPRESENTED_PORT:
+		rc = sfc_mae_switch_get_entity_mport(
+				ctx_mae->sa->mae.switch_domain_id,
+				spec->port_id, &mport_sel);
+		if (rc != 0) {
+			return rte_flow_error_set(error, rc,
+					RTE_FLOW_ERROR_TYPE_ITEM, item,
+					"Can't get m-port for the given ethdev");
+		}
+		break;
+	default:
+		return rte_flow_error_set(error, EINVAL,
 				RTE_FLOW_ERROR_TYPE_ITEM, item,
-				"Can't get m-port for the given ethdev");
+				"Unsupported ethdev-based flow item");
 	}
 
 	rc = efx_mae_match_spec_mport_set(ctx_mae->match_spec,
@@ -2329,7 +2347,19 @@ static const struct sfc_flow_item sfc_flow_items[] = {
 		.prev_layer = SFC_FLOW_ITEM_ANY_LAYER,
 		.layer = SFC_FLOW_ITEM_ANY_LAYER,
 		.ctx_type = SFC_FLOW_PARSE_CTX_MAE,
-		.parse = sfc_mae_rule_parse_item_port_representor,
+		.parse = sfc_mae_rule_parse_item_ethdev_based,
+	},
+	{
+		.type = RTE_FLOW_ITEM_TYPE_REPRESENTED_PORT,
+		.name = "REPRESENTED_PORT",
+		/*
+		 * In terms of RTE flow, this item is a META one,
+		 * and its position in the pattern is don't care.
+		 */
+		.prev_layer = SFC_FLOW_ITEM_ANY_LAYER,
+		.layer = SFC_FLOW_ITEM_ANY_LAYER,
+		.ctx_type = SFC_FLOW_PARSE_CTX_MAE,
+		.parse = sfc_mae_rule_parse_item_ethdev_based,
 	},
 	{
 		.type = RTE_FLOW_ITEM_TYPE_PHY_PORT,
diff --git a/drivers/net/sfc/sfc_switch.c b/drivers/net/sfc/sfc_switch.c
index 3f7518fa30..5c10e8fc74 100644
--- a/drivers/net/sfc/sfc_switch.c
+++ b/drivers/net/sfc/sfc_switch.c
@@ -567,6 +567,35 @@ sfc_mae_switch_get_ethdev_mport(uint16_t switch_domain_id,
 	return rc;
 }
 
+int
+sfc_mae_switch_get_entity_mport(uint16_t switch_domain_id,
+				uint16_t ethdev_port_id,
+				efx_mport_sel_t *mport_sel)
+{
+	static struct sfc_mae_switch_port *port;
+	int rc;
+
+	rte_spinlock_lock(&sfc_mae_switch.lock);
+	rc = sfc_mae_find_switch_port_by_ethdev(switch_domain_id,
+						ethdev_port_id, &port);
+	if (rc != 0)
+		goto unlock;
+
+	if (port->type == SFC_MAE_SWITCH_PORT_INDEPENDENT &&
+	    !port->data.indep.mae_admin) {
+		/* See sfc_mae_assign_entity_mport() */
+		rc = ENOTSUP;
+		goto unlock;
+	}
+
+	*mport_sel = port->entity_mport;
+
+unlock:
+	rte_spinlock_unlock(&sfc_mae_switch.lock);
+
+	return rc;
+}
+
 int
 sfc_mae_switch_port_id_by_entity(uint16_t switch_domain_id,
 				 const efx_mport_sel_t *entity_mportp,
diff --git a/drivers/net/sfc/sfc_switch.h b/drivers/net/sfc/sfc_switch.h
index a5a0fb4fc5..62aea9b785 100644
--- a/drivers/net/sfc/sfc_switch.h
+++ b/drivers/net/sfc/sfc_switch.h
@@ -106,6 +106,10 @@ int sfc_mae_switch_get_ethdev_mport(uint16_t switch_domain_id,
 				    uint16_t ethdev_port_id,
 				    efx_mport_sel_t *mport_sel);
 
+int sfc_mae_switch_get_entity_mport(uint16_t switch_domain_id,
+				    uint16_t ethdev_port_id,
+				    efx_mport_sel_t *mport_sel);
+
 int sfc_mae_switch_port_id_by_entity(uint16_t switch_domain_id,
 				     const efx_mport_sel_t *entity_mportp,
 				     enum sfc_mae_switch_port_type type,
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 6/7] net/sfc: support port representor related flow actions
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
                   ` (3 preceding siblings ...)
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 5/7] net/sfc: support represented port flow item Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 7/7] net/sfc: ignore direction attributes in transfer flows Ivan Malov
  2021-11-02 18:27 ` [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ferruh Yigit
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

Add support for actions PORT_REPRESENTOR and REPRESENTED_PORT.

The former should be used instead of ambiguous PORT_ID.

The latter sends traffic to the entity represented by
the given ethdev (network port or VF).

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 doc/guides/nics/features/sfc.ini |  2 +
 doc/guides/nics/sfc_efx.rst      |  4 ++
 drivers/net/sfc/sfc_mae.c        | 66 ++++++++++++++++++++++++++++++++
 3 files changed, 72 insertions(+)

diff --git a/doc/guides/nics/features/sfc.ini b/doc/guides/nics/features/sfc.ini
index c830426eb2..0d785f4765 100644
--- a/doc/guides/nics/features/sfc.ini
+++ b/doc/guides/nics/features/sfc.ini
@@ -73,6 +73,8 @@ of_set_vlan_vid      = Y
 pf                   = Y
 phy_port             = Y
 port_id              = Y
+port_representor     = Y
+represented_port     = Y
 queue                = Y
 rss                  = Y
 vf                   = Y
diff --git a/doc/guides/nics/sfc_efx.rst b/doc/guides/nics/sfc_efx.rst
index 8dbd250b3c..960e25bf98 100644
--- a/doc/guides/nics/sfc_efx.rst
+++ b/doc/guides/nics/sfc_efx.rst
@@ -248,6 +248,10 @@ Supported actions (***transfer*** rules):
 
 - VF
 
+- PORT_REPRESENTOR
+
+- REPRESENTED_PORT
+
 - PORT_ID
 
 - COUNT
diff --git a/drivers/net/sfc/sfc_mae.c b/drivers/net/sfc/sfc_mae.c
index 8ec4036275..411f2ac27e 100644
--- a/drivers/net/sfc/sfc_mae.c
+++ b/drivers/net/sfc/sfc_mae.c
@@ -3488,6 +3488,58 @@ sfc_mae_rule_parse_action_port_id(struct sfc_adapter *sa,
 	return rc;
 }
 
+static int
+sfc_mae_rule_parse_action_port_representor(struct sfc_adapter *sa,
+		const struct rte_flow_action_ethdev *conf,
+		efx_mae_actions_t *spec)
+{
+	struct sfc_mae *mae = &sa->mae;
+	efx_mport_sel_t mport;
+	int rc;
+
+	rc = sfc_mae_switch_get_ethdev_mport(mae->switch_domain_id,
+					     conf->port_id, &mport);
+	if (rc != 0) {
+		sfc_err(sa, "failed to get m-port for the given ethdev (port_id=%u): %s",
+			conf->port_id, strerror(rc));
+		return rc;
+	}
+
+	rc = efx_mae_action_set_populate_deliver(spec, &mport);
+	if (rc != 0) {
+		sfc_err(sa, "failed to request action DELIVER with m-port selector 0x%08x: %s",
+			mport.sel, strerror(rc));
+	}
+
+	return rc;
+}
+
+static int
+sfc_mae_rule_parse_action_represented_port(struct sfc_adapter *sa,
+		const struct rte_flow_action_ethdev *conf,
+		efx_mae_actions_t *spec)
+{
+	struct sfc_mae *mae = &sa->mae;
+	efx_mport_sel_t mport;
+	int rc;
+
+	rc = sfc_mae_switch_get_entity_mport(mae->switch_domain_id,
+					     conf->port_id, &mport);
+	if (rc != 0) {
+		sfc_err(sa, "failed to get m-port for the given ethdev (port_id=%u): %s",
+			conf->port_id, strerror(rc));
+		return rc;
+	}
+
+	rc = efx_mae_action_set_populate_deliver(spec, &mport);
+	if (rc != 0) {
+		sfc_err(sa, "failed to request action DELIVER with m-port selector 0x%08x: %s",
+			mport.sel, strerror(rc));
+	}
+
+	return rc;
+}
+
 static const char * const action_names[] = {
 	[RTE_FLOW_ACTION_TYPE_VXLAN_DECAP] = "VXLAN_DECAP",
 	[RTE_FLOW_ACTION_TYPE_OF_POP_VLAN] = "OF_POP_VLAN",
@@ -3501,6 +3553,8 @@ static const char * const action_names[] = {
 	[RTE_FLOW_ACTION_TYPE_PF] = "PF",
 	[RTE_FLOW_ACTION_TYPE_VF] = "VF",
 	[RTE_FLOW_ACTION_TYPE_PORT_ID] = "PORT_ID",
+	[RTE_FLOW_ACTION_TYPE_PORT_REPRESENTOR] = "PORT_REPRESENTOR",
+	[RTE_FLOW_ACTION_TYPE_REPRESENTED_PORT] = "REPRESENTED_PORT",
 	[RTE_FLOW_ACTION_TYPE_DROP] = "DROP",
 	[RTE_FLOW_ACTION_TYPE_JUMP] = "JUMP",
 };
@@ -3609,6 +3663,18 @@ sfc_mae_rule_parse_action(struct sfc_adapter *sa,
 				       bundle->actions_mask);
 		rc = sfc_mae_rule_parse_action_port_id(sa, action->conf, spec);
 		break;
+	case RTE_FLOW_ACTION_TYPE_PORT_REPRESENTOR:
+		SFC_BUILD_SET_OVERFLOW(RTE_FLOW_ACTION_TYPE_PORT_REPRESENTOR,
+				       bundle->actions_mask);
+		rc = sfc_mae_rule_parse_action_port_representor(sa,
+				action->conf, spec);
+		break;
+	case RTE_FLOW_ACTION_TYPE_REPRESENTED_PORT:
+		SFC_BUILD_SET_OVERFLOW(RTE_FLOW_ACTION_TYPE_REPRESENTED_PORT,
+				       bundle->actions_mask);
+		rc = sfc_mae_rule_parse_action_represented_port(sa,
+				action->conf, spec);
+		break;
 	case RTE_FLOW_ACTION_TYPE_DROP:
 		SFC_BUILD_SET_OVERFLOW(RTE_FLOW_ACTION_TYPE_DROP,
 				       bundle->actions_mask);
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [dpdk-dev] [PATCH 7/7] net/sfc: ignore direction attributes in transfer flows
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
                   ` (4 preceding siblings ...)
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 6/7] net/sfc: support port representor related flow actions Ivan Malov
@ 2021-10-25 11:04 ` Ivan Malov
  2021-11-02 18:27 ` [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ferruh Yigit
  6 siblings, 0 replies; 8+ messages in thread
From: Ivan Malov @ 2021-10-25 11:04 UTC (permalink / raw)
  To: dev; +Cc: Andrew Rybchenko

[1] has deprecated the use of direction attributes in "transfer"
flows. Ignore them during the transition period.

[1] commit 9d2a349b388a ("ethdev: deprecate direction attributes
in transfer flows")

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
---
 drivers/net/sfc/sfc_flow.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/sfc/sfc_flow.c b/drivers/net/sfc/sfc_flow.c
index be2dfe778a..fc74c8035e 100644
--- a/drivers/net/sfc/sfc_flow.c
+++ b/drivers/net/sfc/sfc_flow.c
@@ -1252,13 +1252,13 @@ sfc_flow_parse_attr(struct sfc_adapter *sa,
 				   "Groups are not supported");
 		return -rte_errno;
 	}
-	if (attr->egress != 0) {
+	if (attr->egress != 0 && attr->transfer == 0) {
 		rte_flow_error_set(error, ENOTSUP,
 				   RTE_FLOW_ERROR_TYPE_ATTR_EGRESS, attr,
 				   "Egress is not supported");
 		return -rte_errno;
 	}
-	if (attr->ingress == 0) {
+	if (attr->ingress == 0 && attr->transfer == 0) {
 		rte_flow_error_set(error, ENOTSUP,
 				   RTE_FLOW_ERROR_TYPE_ATTR_INGRESS, attr,
 				   "Ingress is compulsory");
-- 
2.20.1


^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors
  2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
                   ` (5 preceding siblings ...)
  2021-10-25 11:04 ` [dpdk-dev] [PATCH 7/7] net/sfc: ignore direction attributes in transfer flows Ivan Malov
@ 2021-11-02 18:27 ` Ferruh Yigit
  6 siblings, 0 replies; 8+ messages in thread
From: Ferruh Yigit @ 2021-11-02 18:27 UTC (permalink / raw)
  To: Ivan Malov, dev; +Cc: Andrew Rybchenko, Andy Moreton, Igor Romanov

On 10/25/2021 12:04 PM, Ivan Malov wrote:
> VF representors do not own dedicated m-ports and thus cannot
> be referred to as traffic endpoints in flow items or actions.
> 
> Fixes: a62ec90522a6 ("net/sfc: add port representors infrastructure")
> Fixes: f55b61cec94a ("net/sfc: support port representor flow item")
> 
> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>


Series applied to dpdk-next-net/main, thanks.

^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2021-11-02 18:27 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-10-25 11:04 [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 2/7] net/sfc: rename ethdev m-port retrieval helper Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 3/7] net/sfc: improve m-port related log messages Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 4/7] net/sfc: assign correct m-ports to independent switch ports Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 5/7] net/sfc: support represented port flow item Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 6/7] net/sfc: support port representor related flow actions Ivan Malov
2021-10-25 11:04 ` [dpdk-dev] [PATCH 7/7] net/sfc: ignore direction attributes in transfer flows Ivan Malov
2021-11-02 18:27 ` [dpdk-dev] [PATCH 1/7] net/sfc: do not allow flow rules to refer to VF representors Ferruh Yigit

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).