[PATCH net-next v1 5/6] net: dsa: tag: mt7628: add VLAN awareness support

[Joris Vaisvila <joey@tinyisr.com>]

Support VLAN-aware bridges in the MT7628 tagger.

VLAN-aware bridge traffic already contains the VLAN tag used for
forwarding, so modify the TPID to contain the destination port mask
instead of adding a new tag.

On receive, strip the MT7628 tag, use the VID to find the source port if
the switch reports port 0 on VLAN-aware traffic, then restore the VLAN
tag as a hardware-accelerated VLAN tag.

Signed-off-by: Joris Vaisvila <joey@tinyisr.com>
---
 net/dsa/tag_mt7628.c | 62 +++++++++++++++++++++++++++++++++++---------
 1 file changed, 50 insertions(+), 12 deletions(-)

diff --git a/net/dsa/tag_mt7628.c b/net/dsa/tag_mt7628.c
index fab9d4f04b7a..7c71e8b08648 100644
--- a/net/dsa/tag_mt7628.c
+++ b/net/dsa/tag_mt7628.c
@@ -6,6 +6,7 @@
 
 #include <linux/etherdevice.h>
 #include <linux/dsa/8021q.h>
+#include <linux/if_vlan.h>
 #include <net/dsa.h>
 
 #include "tag.h"
@@ -15,13 +16,16 @@
  * On TX the lower 6 bits encode the destination port bitmask.
  * On RX the lower 3 bits encode the source port number.
  *
- * The switch hardware will not modify the TPID of an incoming packet if it is
- * already VLAN tagged. To work around this the switch is configured to always
- * append a tag_8021q standalone VLAN tag for each port. That means we can
- * safely strip the outer VLAN tag after parsing it.
+ * The switch can only use VLANs for forwarding control. VLAN-unaware bridges
+ * are simulated using tag_8021q and double tagging, while VLAN-aware bridges
+ * use the VLANs configured by the bridge directly.
  *
- * A VLAN tag is constructed on egress to target the standalone or bridge
- * VLAN and destination port.
+ * On egress, the tagger either adds a new MT7628 tag that contains the
+ * standalone or bridge tag_8021q VLAN and destination port mask, or modifies
+ * an existing VLAN tag to contain the destination port mask.
+ *
+ * On ingress, the VLAN tag is restored after stripping the MT7628 tag, if it
+ * is not a tag_8021q VLAN.
  */
 
 #define MT7628_TAG_NAME "mt7628"
@@ -34,10 +38,25 @@ static struct sk_buff *mt7628_tag_xmit(struct sk_buff *skb,
 				       struct net_device *dev)
 {
 	struct dsa_port *dp;
+	u16 xmit_tpid;
 	u16 xmit_vlan;
 	__be16 *tag;
 
+	xmit_tpid =
+	    ETH_P_8021Q | FIELD_PREP(MT7628_TAG_TX_PORT,
+				     dsa_xmit_port_mask(skb, dev));
 	dp = dsa_user_to_port(dev);
+	if (skb->offload_fwd_mark &&
+	    br_vlan_enabled(dsa_port_bridge_dev_get(dp))) {
+		/*
+		 * On VLAN aware ports only modify the TPID to contain the
+		 * MT7628 egress port metadata, instead of adding a new vlan tag
+		 */
+		tag = dsa_etype_header_pos_tx(skb);
+		tag[0] = htons(xmit_tpid);
+		return skb;
+	}
+
 	xmit_vlan = skb->offload_fwd_mark ?
 	    dsa_tag_8021q_bridge_vid(dsa_port_bridge_num_get(dp)) :
 	    dsa_tag_8021q_standalone_vid(dp);
@@ -47,9 +66,7 @@ static struct sk_buff *mt7628_tag_xmit(struct sk_buff *skb,
 
 	tag = dsa_etype_header_pos_tx(skb);
 
-	tag[0] = htons(ETH_P_8021Q |
-		       FIELD_PREP(MT7628_TAG_TX_PORT,
-				  dsa_xmit_port_mask(skb, dev)));
+	tag[0] = htons(xmit_tpid);
 	tag[1] = htons(xmit_vlan);
 
 	return skb;
@@ -58,7 +75,10 @@ static struct sk_buff *mt7628_tag_xmit(struct sk_buff *skb,
 static struct sk_buff *mt7628_tag_rcv(struct sk_buff *skb,
 				      struct net_device *dev)
 {
+	unsigned int source_port;
+	bool is_dsa_8021q;
 	__be16 *phdr;
+	u16 tci, vid;
 
 	if (unlikely(!pskb_may_pull(skb, MT7628_TAG_LEN))) {
 		kfree_skb(skb);
@@ -66,9 +86,23 @@ static struct sk_buff *mt7628_tag_rcv(struct sk_buff *skb,
 	}
 
 	phdr = dsa_etype_header_pos_rx(skb);
-	skb->dev =
-	    dsa_conduit_find_user(dev, 0,
-				  FIELD_GET(MT7628_TAG_RX_PORT, ntohs(*phdr)));
+	source_port = FIELD_GET(MT7628_TAG_RX_PORT, ntohs(*phdr));
+	tci = ntohs(phdr[1]);
+	vid = tci & VLAN_VID_MASK;
+	is_dsa_8021q = vid_is_dsa_8021q(vid);
+
+	/*
+	 * The source port info is only encoded in the TPID field for packets
+	 * where the VLAN tag is inserted by the PVID mechanism. With VLAN
+	 * filtering enabled, VLAN-tagged ingress packets appear as if they're
+	 * originating on port 0. Use the VID to identify the bridge port in
+	 * this case.
+	 */
+	if (source_port == 0 && !is_dsa_8021q)
+		skb->dev = dsa_find_designated_bridge_port_by_vid(dev, vid);
+	else
+		skb->dev = dsa_conduit_find_user(dev, 0, source_port);
+
 	if (!skb->dev) {
 		kfree_skb(skb);
 		return NULL;
@@ -77,6 +111,10 @@ static struct sk_buff *mt7628_tag_rcv(struct sk_buff *skb,
 	skb_pull_rcsum(skb, MT7628_TAG_LEN);
 	dsa_strip_etype_header(skb, MT7628_TAG_LEN);
 	dsa_default_offload_fwd_mark(skb);
+
+	if (!is_dsa_8021q)
+		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), tci);
+
 	return skb;
 }
 
-- 
2.55.0