Skip to contentSkip to conversation

dsa: Allow forwarding of redirected IGMP traffic

PATCH Applied Daniel Mack 3 participants 5 messages Jun 20, 2020 – Jun 24, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 dsa: Allow forwarding of redirected IGMP traffic

The driver for Marvell switches puts all ports in IGMP snooping mode which results in all IGMP/MLD frames that ingress on the ports to be forwarded to the CPU only.

The bridge code in the kernel can then interpret these frames and act upon them, for instance by updating the mdb in the switch to reflect multicast memberships of stations connected to the ports. However, the IGMP/MLD frames must then also be forwarded to other ports of the bridge so external IGMP queriers can track membership reports, and external multicast clients can receive query reports from foreign IGMP queriers.

Currently, this is impossible as the EDSA tagger sets offload_fwd_mark on the skb when it unwraps the tagged frames, and that will make the switchdev layer prevent the skb from egressing on any other port of the same switch.

To fix that, look at the To_CPU code in the DSA header and make forwarding of the frame possible for trapped IGMP packets.

Introduce some #defines for the frame types to make the code a bit more comprehensive.

This was tested on a Marvell 88E6352 variant.

---
v2:
* Limit IGMP handling to TO_CPU frames
* Use #defines for the TO_CPU codes and the frame types

 net/dsa/tag_edsa.c | 37 ++++++++++++++++++++++++++++++++++---
 1 file changed, 34 insertions(+), 3 deletions(-)
net/dsa/tag_edsa.c +34 -3
index e8eaa804ccb9e..d6200ff982007 100644
@@ -13,6 +13,16 @@
 #define DSA_HLEN	4
 #define EDSA_HLEN	8
 
+#define FRAME_TYPE_TO_CPU	0x00
+#define FRAME_TYPE_FORWARD	0x03
+
+#define TO_CPU_CODE_MGMT_TRAP		0x00
+#define TO_CPU_CODE_FRAME2REG		0x01
+#define TO_CPU_CODE_IGMP_MLD_TRAP	0x02
+#define TO_CPU_CODE_POLICY_TRAP		0x03
+#define TO_CPU_CODE_ARP_MIRROR		0x04
+#define TO_CPU_CODE_POLICY_MIRROR	0x05
+
 static struct sk_buff *edsa_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct dsa_port *dp = dsa_slave_to_port(dev);
@@ -77,6 +87,8 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 				struct packet_type *pt)
 {
 	u8 *edsa_header;
+	int frame_type;
+	int code;
 	int source_device;
 	int source_port;
 
@@ -91,8 +103,29 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	/*
 	 * Check that frame type is either TO_CPU or FORWARD.
 	 */
-	if ((edsa_header[0] & 0xc0) != 0x00 && (edsa_header[0] & 0xc0) != 0xc0)
+	frame_type = edsa_header[0] >> 6;
+
+	switch (frame_type) {
+	case FRAME_TYPE_TO_CPU:
+		code = (edsa_header[1] & 0x6) | ((edsa_header[2] >> 4) & 1);
+
+		/*
+		 * Mark the frame to never egress on any port of the same switch
+		 * unless it's a trapped IGMP/MLD packet, in which case the
+		 * bridge might want to forward it.
+		 */
+		if (code != TO_CPU_CODE_IGMP_MLD_TRAP)
+			skb->offload_fwd_mark = 1;
+
+		break;
+
+	case FRAME_TYPE_FORWARD:
+		skb->offload_fwd_mark = 1;
+		break;
+
+	default:
 		return NULL;
+	}
 
 	/*
 	 * Determine source device and port.
@@ -156,8 +189,6 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 			2 * ETH_ALEN);
 	}
 
-	skb->offload_fwd_mark = 1;
-
 	return skb;
 }
 
Signature · 1 line
2.26.2
#1 lore raw

Andrew,

This version should address the comments you had on my initial submission. Does this one look better now?

Thanks,
Daniel

Quoted 105 lines · Daniel Mack

The driver for Marvell switches puts all ports in IGMP snooping mode which results in all IGMP/MLD frames that ingress on the ports to be forwarded to the CPU only.

The bridge code in the kernel can then interpret these frames and act upon them, for instance by updating the mdb in the switch to reflect multicast memberships of stations connected to the ports. However, the IGMP/MLD frames must then also be forwarded to other ports of the bridge so external IGMP queriers can track membership reports, and external multicast clients can receive query reports from foreign IGMP queriers.

Currently, this is impossible as the EDSA tagger sets offload_fwd_mark on the skb when it unwraps the tagged frames, and that will make the switchdev layer prevent the skb from egressing on any other port of the same switch.

To fix that, look at the To_CPU code in the DSA header and make forwarding of the frame possible for trapped IGMP packets.

Introduce some #defines for the frame types to make the code a bit more comprehensive.

This was tested on a Marvell 88E6352 variant.

---
v2:
* Limit IGMP handling to TO_CPU frames
* Use #defines for the TO_CPU codes and the frame types

 net/dsa/tag_edsa.c | 37 ++++++++++++++++++++++++++++++++++---
 1 file changed, 34 insertions(+), 3 deletions(-)
net/dsa/tag_edsa.c +34 -3
index e8eaa804ccb9e..d6200ff982007 100644
@@ -13,6 +13,16 @@
 #define DSA_HLEN	4
 #define EDSA_HLEN	8
 
+#define FRAME_TYPE_TO_CPU	0x00
+#define FRAME_TYPE_FORWARD	0x03
+
+#define TO_CPU_CODE_MGMT_TRAP		0x00
+#define TO_CPU_CODE_FRAME2REG		0x01
+#define TO_CPU_CODE_IGMP_MLD_TRAP	0x02
+#define TO_CPU_CODE_POLICY_TRAP		0x03
+#define TO_CPU_CODE_ARP_MIRROR		0x04
+#define TO_CPU_CODE_POLICY_MIRROR	0x05
+
 static struct sk_buff *edsa_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct dsa_port *dp = dsa_slave_to_port(dev);
@@ -77,6 +87,8 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 				struct packet_type *pt)
 {
 	u8 *edsa_header;
+	int frame_type;
+	int code;
 	int source_device;
 	int source_port;
 
@@ -91,8 +103,29 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	/*
 	 * Check that frame type is either TO_CPU or FORWARD.
 	 */
-	if ((edsa_header[0] & 0xc0) != 0x00 && (edsa_header[0] & 0xc0) != 0xc0)
+	frame_type = edsa_header[0] >> 6;
+
+	switch (frame_type) {
+	case FRAME_TYPE_TO_CPU:
+		code = (edsa_header[1] & 0x6) | ((edsa_header[2] >> 4) & 1);
+
+		/*
+		 * Mark the frame to never egress on any port of the same switch
+		 * unless it's a trapped IGMP/MLD packet, in which case the
+		 * bridge might want to forward it.
+		 */
+		if (code != TO_CPU_CODE_IGMP_MLD_TRAP)
+			skb->offload_fwd_mark = 1;
+
+		break;
+
+	case FRAME_TYPE_FORWARD:
+		skb->offload_fwd_mark = 1;
+		break;
+
+	default:
 		return NULL;
+	}
 
 	/*
 	 * Determine source device and port.
@@ -156,8 +189,6 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 			2 * ETH_ALEN);
 	}
 
-	skb->offload_fwd_mark = 1;
-
 	return skb;
 }
 
maintainer
#2 lore raw
Quoted 4 lines · Daniel Mack

Andrew,

This version should address the comments you had on my initial submission. Does this one look better now?

Hi Daniel

It does look better. I'm just trying to find some time to test it a little.

Andrew

maintainer Reviewed-by Tested-by on v2 1/1
#3 lore raw
In reply to Daniel Mack
Quoted 20 lines · Daniel Mack

The driver for Marvell switches puts all ports in IGMP snooping mode which results in all IGMP/MLD frames that ingress on the ports to be forwarded to the CPU only.

The bridge code in the kernel can then interpret these frames and act upon them, for instance by updating the mdb in the switch to reflect multicast memberships of stations connected to the ports. However, the IGMP/MLD frames must then also be forwarded to other ports of the bridge so external IGMP queriers can track membership reports, and external multicast clients can receive query reports from foreign IGMP queriers.

Currently, this is impossible as the EDSA tagger sets offload_fwd_mark on the skb when it unwraps the tagged frames, and that will make the switchdev layer prevent the skb from egressing on any other port of the same switch.

To fix that, look at the To_CPU code in the DSA header and make forwarding of the frame possible for trapped IGMP packets.

Quoted 6 lines

Introduce some #defines for the frame types to make the code a bit more comprehensive.

This was tested on a Marvell 88E6352 variant.

The testing was simple regression testing, not IGMP specific.

Andrew

maintainer Applied
#4 lore raw
In reply to Daniel Mack

From: Daniel Mack <[email protected]>
Date: Sat, 20 Jun 2020 21:39:25 +0200

Quoted 20 lines

The driver for Marvell switches puts all ports in IGMP snooping mode which results in all IGMP/MLD frames that ingress on the ports to be forwarded to the CPU only.

The bridge code in the kernel can then interpret these frames and act upon them, for instance by updating the mdb in the switch to reflect multicast memberships of stations connected to the ports. However, the IGMP/MLD frames must then also be forwarded to other ports of the bridge so external IGMP queriers can track membership reports, and external multicast clients can receive query reports from foreign IGMP queriers.

Currently, this is impossible as the EDSA tagger sets offload_fwd_mark on the skb when it unwraps the tagged frames, and that will make the switchdev layer prevent the skb from egressing on any other port of the same switch.

To fix that, look at the To_CPU code in the DSA header and make forwarding of the frame possible for trapped IGMP packets.

Quoted 6 lines

Introduce some #defines for the frame types to make the code a bit more comprehensive.

This was tested on a Marvell 88E6352 variant.

Applied, thank you.