Skip to contentSkip to conversation

net: dsa: Implement flow_dissect callback for tag_dsa.

PATCH Applied Rundong Ge 3 participants 5 messages Feb 16, 2019 – Feb 19, 2019

Subsystems (3)
lore
Share & tools
lore

Rundong Ge

author
#0 lore raw

PATCH v2 1/1 net: dsa: Implement flow_dissect callback for tag_dsa.

RPS not work for DSA devices since the 'skb_get_hash' will always get the invalid hash for dsa tagged packets.

"[PATCH] tag_mtk: add flow_dissect callback to the ops struct" introduced the flow_dissect callback to get the right hash for MTK tagged packet. Tag_dsa and tag_edsa also need to implement the callback.

---
 net/dsa/tag_dsa.c  | 9 +++++++++
 net/dsa/tag_edsa.c | 9 +++++++++
 2 files changed, 18 insertions(+)
net/dsa/tag_dsa.c +9 -0
index 8b2f92e..67ff3fa 100644
@@ -146,8 +146,17 @@ static struct sk_buff *dsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
+static int dsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
+				int *offset)
+{
+	*offset = 4;
+	*proto = ((__be16 *)skb->data)[1];
+	return 0;
+}
+
 const struct dsa_device_ops dsa_netdev_ops = {
 	.xmit	= dsa_xmit,
 	.rcv	= dsa_rcv,
+	.flow_dissect   = dsa_tag_flow_dissect,
 	.overhead = DSA_HLEN,
 };
net/dsa/tag_edsa.c +9 -0
index f5b87ee..234585e 100644
@@ -165,8 +165,17 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
 	return skb;
 }
 
+static int edsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
+				 int *offset)
+{
+	*offset = 8;
+	*proto = ((__be16 *)skb->data)[3];
+	return 0;
+}
+
 const struct dsa_device_ops edsa_netdev_ops = {
 	.xmit	= edsa_xmit,
 	.rcv	= edsa_rcv,
+	.flow_dissect   = edsa_tag_flow_dissect,
 	.overhead = EDSA_HLEN,
 };
Signature · 1 line
1.8.3.1
author
#1 lore raw

Hi Andrew

I have tested the L3 forwarding throughput performance of my box (with an intel 4-core processor and each core’s frequency is 2.20GHz). In my test scenario, I generated 200 UDP flows (frame size is 64 bytes) with different src/dst pairs from eth1 to eth0. Eth1 is a slave DSA device from mv88e6190x, eth0 is an intel NIC. Then get the max fps with no frame losing.

The max fps with RPS working is 400235fps, frames are hashed to four cores’ backlog.
And the max fps without RPS is 199686fps.

Rundong

Quoted 62 lines · Rundong Ge <[email protected]> 于2019年2月16日周六 下午4:35写道:

RPS not work for DSA devices since the 'skb_get_hash' will always get the invalid hash for dsa tagged packets.

"[PATCH] tag_mtk: add flow_dissect callback to the ops struct" introduced the flow_dissect callback to get the right hash for MTK tagged packet. Tag_dsa and tag_edsa also need to implement the callback.

---
 net/dsa/tag_dsa.c  | 9 +++++++++
 net/dsa/tag_edsa.c | 9 +++++++++
 2 files changed, 18 insertions(+)
net/dsa/tag_dsa.c +9 -0
index 8b2f92e..67ff3fa 100644
@@ -146,8 +146,17 @@ static struct sk_buff *dsa_rcv(struct sk_buff *skb, struct net_device *dev,
        return skb;
 }
 
+static int dsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
+                               int *offset)
+{
+       *offset = 4;
+       *proto = ((__be16 *)skb->data)[1];
+       return 0;
+}
+
 const struct dsa_device_ops dsa_netdev_ops = {
        .xmit   = dsa_xmit,
        .rcv    = dsa_rcv,
+       .flow_dissect   = dsa_tag_flow_dissect,
        .overhead = DSA_HLEN,
 };
net/dsa/tag_edsa.c +9 -1
index f5b87ee..234585e 100644
@@ -165,8 +165,17 @@ static struct sk_buff *edsa_rcv(struct sk_buff *skb, struct net_device *dev,
        return skb;
 }
 
+static int edsa_tag_flow_dissect(const struct sk_buff *skb, __be16 *proto,
+                                int *offset)
+{
+       *offset = 8;
+       *proto = ((__be16 *)skb->data)[3];
+       return 0;
+}
+
 const struct dsa_device_ops edsa_netdev_ops = {
        .xmit   = edsa_xmit,
        .rcv    = edsa_rcv,
+       .flow_dissect   = edsa_tag_flow_dissect,
        .overhead = EDSA_HLEN,
 };
--

1.8.3.1

maintainer
#2 lore raw
Quoted 9 lines · Rundong Ge

Hi Andrew

I have tested the L3 forwarding throughput performance of my box (with an intel 4-core processor and each core’s frequency is 2.20GHz). In my test scenario, I generated 200 UDP flows (frame size is 64 bytes) with different src/dst pairs from eth1 to eth0. Eth1 is a slave DSA device from mv88e6190x, eth0 is an intel NIC. Then get the max fps with no frame losing.

Quoted 3 lines

The max fps with RPS working is 400235fps, frames are hashed to four cores’ backlog.
And the max fps without RPS is 199686fps.

Nice numbers, thanks.

Andrew

maintainer Reviewed-by on v2 1/1
#3 lore raw
In reply to Rundong Ge
Quoted 9 lines · Rundong Ge

RPS not work for DSA devices since the 'skb_get_hash' will always get the invalid hash for dsa tagged packets.

"[PATCH] tag_mtk: add flow_dissect callback to the ops struct" introduced the flow_dissect callback to get the right hash for MTK tagged packet. Tag_dsa and tag_edsa also need to implement the callback.

Andrew

maintainer Applied
#4 lore raw
In reply to Rundong Ge

From: Rundong Ge <[email protected]>
Date: Sat, 16 Feb 2019 08:35:24 +0000

Quoted 9 lines

RPS not work for DSA devices since the 'skb_get_hash' will always get the invalid hash for dsa tagged packets.

"[PATCH] tag_mtk: add flow_dissect callback to the ops struct" introduced the flow_dissect callback to get the right hash for MTK tagged packet. Tag_dsa and tag_edsa also need to implement the callback.

Applied to net-next.