[PATCH net-next 0/4] net: mvpp2: cls: Allow steering based on vlan tag

[PATCH net-next 0/4] net: mvpp2: cls: Allow steering based on vlan tag

[Maxime Chevallier]

The PPv2 classifier can perform flow steering based on keys extracted
from the VLAN tag. This series adds support for using the vlan id and
the vlan prio as keys, using the ethtool interface.

Patch 1 is a preparatory patch that prevent false-positive matches,
using a dedicated lookup id for the RSS C2 lookup.

Patch 2 allows to separate the flows based on the header fields they
contain. The main goal is to be able to separate tagged traffic from
untagged traffic for flow steering, just as we already do for RSS.

Patch 3 solves an issue we have when extracting fields that aren't full
bytes, such as the vlan tag which is 12 bits wide, or the priority which
is 3 bits wide.

Finally, patch 4 adds support for steering based on both vlan id and
priority, extracted from the outermost tag.

Maxime Chevallier (4):
  net: mvpp2: cls: Use a dedicated lu_type for the RSS lookup
  net: mvpp2: cls: Only select applicable flows of classification
    offload
  net: mvpp2: cls: right-justify the C2 TCAM keys
  net: mvpp2: cls: Add steering based on vlan Id and priority.

 .../net/ethernet/marvell/mvpp2/mvpp2_cls.c    | 118 +++++++++++++-----
 .../net/ethernet/marvell/mvpp2/mvpp2_cls.h    |  28 +++--
 2 files changed, 103 insertions(+), 43 deletions(-)

-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH net-next 1/4] net: mvpp2: cls: Use a dedicated lu_type for the RSS lookup

[Maxime Chevallier]

When performing a TCAM lookup in the C2 engine, it's possible that
multiple entries match the packet. To make sure the correct entry match
when performing a lookup, the Flow Table can set a lookup type, which
will be used in the TCAM lookup, thus preventing such false-positives.

We need to make sure the RSS match doesn't interfere with other
classification lookups, hence we use a dedicated lookup_type for it.

Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
---
 drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c | 4 ++--
 drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h | 5 +++++
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
index e47c00c5f829..8af13316ecb1 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
@@ -596,7 +596,7 @@ static void mvpp2_cls_flow_init(struct mvpp2 *priv,
 
 	mvpp2_cls_flow_eng_set(&fe, MVPP22_CLS_ENGINE_C2);
 	mvpp2_cls_flow_port_id_sel(&fe, true);
-	mvpp2_cls_flow_lu_type_set(&fe, MVPP22_FLOW_ETHERNET);
+	mvpp2_cls_flow_lu_type_set(&fe, MVPP22_CLS_LU_TYPE_ALL);
 
 	/* Add all ports */
 	for (i = 0; i < MVPP2_MAX_PORTS; i++)
@@ -861,7 +861,7 @@ static void mvpp2_port_c2_cls_init(struct mvpp2_port *port)
 
 	/* Match on Lookup Type */
 	c2.tcam[4] |= MVPP22_CLS_C2_TCAM_EN(MVPP22_CLS_C2_LU_TYPE(MVPP2_CLS_LU_TYPE_MASK));
-	c2.tcam[4] |= MVPP22_CLS_C2_LU_TYPE(MVPP22_FLOW_ETHERNET);
+	c2.tcam[4] |= MVPP22_CLS_C2_LU_TYPE(MVPP22_CLS_LU_TYPE_ALL);
 
 	/* Update RSS status after matching this entry */
 	c2.act = MVPP22_CLS_C2_ACT_RSS_EN(MVPP22_C2_UPD_LOCK);
diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
index 26cc1176e758..957f80b31743 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
@@ -180,6 +180,11 @@ enum mvpp2_prs_flow {
 /* LU Type defined for all engines, and specified in the flow table */
 #define MVPP2_CLS_LU_TYPE_MASK			0x3f
 
+enum mvpp2_cls_lu_type {
+	/* rule->loc is used as a lu-type for the entries 0 - 62. */
+	MVPP22_CLS_LU_TYPE_ALL = 63,
+};
+
 #define MVPP2_N_FLOWS		(MVPP2_FL_LAST - MVPP2_FL_START)
 
 struct mvpp2_cls_flow {
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH net-next 2/4] net: mvpp2: cls: Only select applicable flows of classification offload

[Maxime Chevallier]

The way we currently handle classification offload and RSS is by having
dedicated lookup sequences in the flow table, each being selected
depending on several fields being present in the packet header.

We need to make sure the classification operation we want to perform can
be done in each flow we want to insert it into. As an example,
classifying on VLAN tag can only be done on flows used for tagged
traffic.

This commit makes sure we don't insert rules in flows we aren't
compatible with.

Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
---
 drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
index 8af13316ecb1..7cd9d6da0319 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
@@ -1201,6 +1201,9 @@ static int mvpp2_port_flt_rfs_rule_insert(struct mvpp2_port *port,
 		if (!flow)
 			return 0;
 
+		if ((rule->hek_fields & flow->supported_hash_opts) != rule->hek_fields)
+			continue;
+
 		index = MVPP2_CLS_FLT_C2_RFS(port->id, flow->flow_id, rule->loc);
 
 		mvpp2_cls_flow_read(priv, index, &fe);
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH net-next 3/4] net: mvpp2: cls: right-justify the C2 TCAM keys

[Maxime Chevallier]

The C2 TCAM used for classification uses a key (Header Extracted Key)
built by concatenating several fields extracted from the packet header.

After a lot of trial-and-error and some guess work, it seems the HEK is
right justified, with the first fields being stored in the MSB, then
concatenated up until the LSB.

Until now, this doesn't cause any issue since all HEK fields we use are
full bytes. However this is an issue for the upcoming VLAN id and pri
extraction, which aren't full bytes.

Rework the way we built that TCAM key, by changing the order in which we
append the fields.

Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
---
 drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
index 7cd9d6da0319..c4c467f5f4f6 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
@@ -1081,13 +1081,13 @@ static int mvpp2_port_c2_tcam_rule_add(struct mvpp2_port *port,
 
 	rule->c2_index = c2.index;
 
-	c2.tcam[0] = (rule->c2_tcam & 0xffff) |
+	c2.tcam[3] = (rule->c2_tcam & 0xffff) |
 		     ((rule->c2_tcam_mask & 0xffff) << 16);
-	c2.tcam[1] = ((rule->c2_tcam >> 16) & 0xffff) |
+	c2.tcam[2] = ((rule->c2_tcam >> 16) & 0xffff) |
 		     (((rule->c2_tcam_mask >> 16) & 0xffff) << 16);
-	c2.tcam[2] = ((rule->c2_tcam >> 32) & 0xffff) |
+	c2.tcam[1] = ((rule->c2_tcam >> 32) & 0xffff) |
 		     (((rule->c2_tcam_mask >> 32) & 0xffff) << 16);
-	c2.tcam[3] = ((rule->c2_tcam >> 48) & 0xffff) |
+	c2.tcam[0] = ((rule->c2_tcam >> 48) & 0xffff) |
 		     (((rule->c2_tcam_mask >> 48) & 0xffff) << 16);
 
 	pmap = BIT(port->id);
@@ -1222,7 +1222,7 @@ static int mvpp2_port_flt_rfs_rule_insert(struct mvpp2_port *port,
 static int mvpp2_cls_c2_build_match(struct mvpp2_rfs_rule *rule)
 {
 	struct flow_rule *flow = rule->flow;
-	int offs = 64;
+	int offs = 0;
 
 	if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_PORTS)) {
 		struct flow_match_ports match;
@@ -1230,18 +1230,18 @@ static int mvpp2_cls_c2_build_match(struct mvpp2_rfs_rule *rule)
 		flow_rule_match_ports(flow, &match);
 		if (match.mask->src) {
 			rule->hek_fields |= MVPP22_CLS_HEK_OPT_L4SIP;
-			offs -= mvpp2_cls_hek_field_size(MVPP22_CLS_HEK_OPT_L4SIP);
 
 			rule->c2_tcam |= ((u64)ntohs(match.key->src)) << offs;
 			rule->c2_tcam_mask |= ((u64)ntohs(match.mask->src)) << offs;
+			offs += mvpp2_cls_hek_field_size(MVPP22_CLS_HEK_OPT_L4SIP);
 		}
 
 		if (match.mask->dst) {
 			rule->hek_fields |= MVPP22_CLS_HEK_OPT_L4DIP;
-			offs -= mvpp2_cls_hek_field_size(MVPP22_CLS_HEK_OPT_L4DIP);
 
 			rule->c2_tcam |= ((u64)ntohs(match.key->dst)) << offs;
 			rule->c2_tcam_mask |= ((u64)ntohs(match.mask->dst)) << offs;
+			offs += mvpp2_cls_hek_field_size(MVPP22_CLS_HEK_OPT_L4DIP);
 		}
 	}
 
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH net-next 4/4] net: mvpp2: cls: Add steering based on vlan Id and priority.

[Maxime Chevallier]

This commit allows using the vlan Id and priority as parts of the key
for classification offload. These fields are extracted from the
outermost tag, if multiple tags are present.

Vlan Id and priority are considered as 2 different fields by the
classifier, however the fields are both appended in the Header Extracted
Key in the same layout as they are found in the tags. This means that
when steering only based on the prio, a 16-bit slot is still taken in
the HEK.

The classifier doesn't allow extracting the DEI bit from the tag, so we
explicitly prevent user from using this bit in the key.

This commit adds the vlan priotity as a compatible HEK field for
tagged traffic, meaning that we limit the possibility of extracting this
field only to the flows that contain tagged traffic.

Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
---
 .../net/ethernet/marvell/mvpp2/mvpp2_cls.c    | 97 ++++++++++++++-----
 .../net/ethernet/marvell/mvpp2/mvpp2_cls.h    | 23 +++--
 2 files changed, 86 insertions(+), 34 deletions(-)

diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
index c4c467f5f4f6..b195fb5d61f4 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.c
@@ -44,17 +44,17 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* TCP over IPv4 flows, Not fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4 | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OPT | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OTHER | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -79,17 +79,17 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* TCP over IPv4 flows, fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4 | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OPT | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP4, MVPP2_FL_IP4_TCP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OTHER | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -114,17 +114,17 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* UDP over IPv4 flows, Not fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4 | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OPT | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_NF_TAG,
-		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OTHER | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -149,17 +149,17 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* UDP over IPv4 flows, fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4 | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OPT | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP4, MVPP2_FL_IP4_UDP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OTHER | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -178,12 +178,12 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* TCP over IPv6 flows, not fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP6, MVPP2_FL_IP6_TCP_NF_TAG,
-		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6 | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP6, MVPP2_FL_IP6_TCP_NF_TAG,
-		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6_EXT | MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -202,13 +202,13 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* TCP over IPv6 flows, fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP6, MVPP2_FL_IP6_TCP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6 | MVPP2_PRS_RI_IP_FRAG_TRUE |
 		       MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_TCP6, MVPP2_FL_IP6_TCP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6_EXT | MVPP2_PRS_RI_IP_FRAG_TRUE |
 		       MVPP2_PRS_RI_L4_TCP,
 		       MVPP2_PRS_IP_MASK),
@@ -228,12 +228,12 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* UDP over IPv6 flows, not fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP6, MVPP2_FL_IP6_UDP_NF_TAG,
-		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6 | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP6, MVPP2_FL_IP6_UDP_NF_TAG,
-		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_5T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6_EXT | MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
@@ -252,13 +252,13 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* UDP over IPv6 flows, fragmented, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP6, MVPP2_FL_IP6_UDP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6 | MVPP2_PRS_RI_IP_FRAG_TRUE |
 		       MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
 
 	MVPP2_DEF_FLOW(MVPP22_FLOW_UDP6, MVPP2_FL_IP6_UDP_FRAG_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6_EXT | MVPP2_PRS_RI_IP_FRAG_TRUE |
 		       MVPP2_PRS_RI_L4_UDP,
 		       MVPP2_PRS_IP_MASK),
@@ -279,15 +279,15 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* IPv4 flows, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_IP4, MVPP2_FL_IP4_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4,
 		       MVPP2_PRS_RI_L3_PROTO_MASK),
 	MVPP2_DEF_FLOW(MVPP22_FLOW_IP4, MVPP2_FL_IP4_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OPT,
 		       MVPP2_PRS_RI_L3_PROTO_MASK),
 	MVPP2_DEF_FLOW(MVPP22_FLOW_IP4, MVPP2_FL_IP4_TAG,
-		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP4_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP4_OTHER,
 		       MVPP2_PRS_RI_L3_PROTO_MASK),
 
@@ -303,11 +303,11 @@ static const struct mvpp2_cls_flow cls_flows[MVPP2_N_PRS_FLOWS] = {
 
 	/* IPv6 flows, with vlan tag */
 	MVPP2_DEF_FLOW(MVPP22_FLOW_IP6, MVPP2_FL_IP6_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6,
 		       MVPP2_PRS_RI_L3_PROTO_MASK),
 	MVPP2_DEF_FLOW(MVPP22_FLOW_IP6, MVPP2_FL_IP6_TAG,
-		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_OPT_VLAN,
+		       MVPP22_CLS_HEK_IP6_2T | MVPP22_CLS_HEK_TAGGED,
 		       MVPP2_PRS_RI_L3_IP6,
 		       MVPP2_PRS_RI_L3_PROTO_MASK),
 
@@ -655,6 +655,9 @@ static int mvpp2_flow_set_hek_fields(struct mvpp2_cls_flow_entry *fe,
 		case MVPP22_CLS_HEK_OPT_VLAN:
 			field_id = MVPP22_CLS_FIELD_VLAN;
 			break;
+		case MVPP22_CLS_HEK_OPT_VLAN_PRI:
+			field_id = MVPP22_CLS_FIELD_VLAN_PRI;
+			break;
 		case MVPP22_CLS_HEK_OPT_IP4SA:
 			field_id = MVPP22_CLS_FIELD_IP4SA;
 			break;
@@ -689,6 +692,10 @@ static int mvpp2_cls_hek_field_size(u32 field)
 	switch (field) {
 	case MVPP22_CLS_HEK_OPT_MAC_DA:
 		return 48;
+	case MVPP22_CLS_HEK_OPT_VLAN:
+		return 12;
+	case MVPP22_CLS_HEK_OPT_VLAN_PRI:
+		return 3;
 	case MVPP22_CLS_HEK_OPT_IP4SA:
 	case MVPP22_CLS_HEK_OPT_IP4DA:
 		return 32;
@@ -777,6 +784,9 @@ u16 mvpp2_flow_get_hek_fields(struct mvpp2_cls_flow_entry *fe)
 		case MVPP22_CLS_FIELD_VLAN:
 			hash_opts |= MVPP22_CLS_HEK_OPT_VLAN;
 			break;
+		case MVPP22_CLS_FIELD_VLAN_PRI:
+			hash_opts |= MVPP22_CLS_HEK_OPT_VLAN_PRI;
+			break;
 		case MVPP22_CLS_FIELD_L3_PROTO:
 			hash_opts |= MVPP22_CLS_HEK_OPT_L3_PROTO;
 			break;
@@ -1224,6 +1234,43 @@ static int mvpp2_cls_c2_build_match(struct mvpp2_rfs_rule *rule)
 	struct flow_rule *flow = rule->flow;
 	int offs = 0;
 
+	/* The order of insertion in C2 tcam must match the order in which
+	 * the fields are found in the header
+	 */
+	if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_VLAN)) {
+		struct flow_match_vlan match;
+
+		flow_rule_match_vlan(flow, &match);
+		if (match.mask->vlan_id) {
+			rule->hek_fields |= MVPP22_CLS_HEK_OPT_VLAN;
+
+			rule->c2_tcam |= ((u64)match.key->vlan_id) << offs;
+			rule->c2_tcam_mask |= ((u64)match.mask->vlan_id) << offs;
+
+			/* Don't update the offset yet */
+		}
+
+		if (match.mask->vlan_priority) {
+			rule->hek_fields |= MVPP22_CLS_HEK_OPT_VLAN_PRI;
+
+			/* VLAN pri is always at offset 13 relative to the
+			 * current offset
+			 */
+			rule->c2_tcam |= ((u64)match.key->vlan_priority) <<
+				(offs + 13);
+			rule->c2_tcam_mask |= ((u64)match.mask->vlan_priority) <<
+				(offs + 13);
+		}
+
+		if (match.mask->vlan_dei)
+			return -EOPNOTSUPP;
+
+		/* vlan id and prio always seem to take a full 16-bit slot in
+		 * the Header Extracted Key.
+		 */
+		offs += 16;
+	}
+
 	if (flow_rule_match_key(flow, FLOW_DISSECTOR_KEY_PORTS)) {
 		struct flow_match_ports match;
 
diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
index 957f80b31743..8867f25afab4 100644
--- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
+++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_cls.h
@@ -33,15 +33,16 @@ enum mvpp2_cls_engine {
 };
 
 #define MVPP22_CLS_HEK_OPT_MAC_DA	BIT(0)
-#define MVPP22_CLS_HEK_OPT_VLAN		BIT(1)
-#define MVPP22_CLS_HEK_OPT_L3_PROTO	BIT(2)
-#define MVPP22_CLS_HEK_OPT_IP4SA	BIT(3)
-#define MVPP22_CLS_HEK_OPT_IP4DA	BIT(4)
-#define MVPP22_CLS_HEK_OPT_IP6SA	BIT(5)
-#define MVPP22_CLS_HEK_OPT_IP6DA	BIT(6)
-#define MVPP22_CLS_HEK_OPT_L4SIP	BIT(7)
-#define MVPP22_CLS_HEK_OPT_L4DIP	BIT(8)
-#define MVPP22_CLS_HEK_N_FIELDS		9
+#define MVPP22_CLS_HEK_OPT_VLAN_PRI	BIT(1)
+#define MVPP22_CLS_HEK_OPT_VLAN		BIT(2)
+#define MVPP22_CLS_HEK_OPT_L3_PROTO	BIT(3)
+#define MVPP22_CLS_HEK_OPT_IP4SA	BIT(4)
+#define MVPP22_CLS_HEK_OPT_IP4DA	BIT(5)
+#define MVPP22_CLS_HEK_OPT_IP6SA	BIT(6)
+#define MVPP22_CLS_HEK_OPT_IP6DA	BIT(7)
+#define MVPP22_CLS_HEK_OPT_L4SIP	BIT(8)
+#define MVPP22_CLS_HEK_OPT_L4DIP	BIT(9)
+#define MVPP22_CLS_HEK_N_FIELDS		10
 
 #define MVPP22_CLS_HEK_L4_OPTS	(MVPP22_CLS_HEK_OPT_L4SIP | \
 				 MVPP22_CLS_HEK_OPT_L4DIP)
@@ -59,8 +60,12 @@ enum mvpp2_cls_engine {
 #define MVPP22_CLS_HEK_IP6_5T	(MVPP22_CLS_HEK_IP6_2T | \
 				 MVPP22_CLS_HEK_L4_OPTS)
 
+#define MVPP22_CLS_HEK_TAGGED	(MVPP22_CLS_HEK_OPT_VLAN | \
+				 MVPP22_CLS_HEK_OPT_VLAN_PRI)
+
 enum mvpp2_cls_field_id {
 	MVPP22_CLS_FIELD_MAC_DA = 0x03,
+	MVPP22_CLS_FIELD_VLAN_PRI = 0x05,
 	MVPP22_CLS_FIELD_VLAN = 0x06,
 	MVPP22_CLS_FIELD_L3_PROTO = 0x0f,
 	MVPP22_CLS_FIELD_IP4SA = 0x10,
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH net-next 0/4] net: mvpp2: cls: Allow steering based on vlan tag

[David Miller]

From: Maxime Chevallier <maxime.chevallier@bootlin.com>
Date: Tue, 18 Jun 2019 16:55:15 +0200

> The PPv2 classifier can perform flow steering based on keys extracted
> from the VLAN tag. This series adds support for using the vlan id and
> the vlan prio as keys, using the ethtool interface.
> 
> Patch 1 is a preparatory patch that prevent false-positive matches,
> using a dedicated lookup id for the RSS C2 lookup.
> 
> Patch 2 allows to separate the flows based on the header fields they
> contain. The main goal is to be able to separate tagged traffic from
> untagged traffic for flow steering, just as we already do for RSS.
> 
> Patch 3 solves an issue we have when extracting fields that aren't full
> bytes, such as the vlan tag which is 12 bits wide, or the priority which
> is 3 bits wide.
> 
> Finally, patch 4 adds support for steering based on both vlan id and
> priority, extracted from the outermost tag.

Series applied, thanks.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel