macvlan: Introduce 'passthru' mode to takeover the underlying device

With the current default 'vepa' mode, a KVM guest using virtio with macvtap backend has the following limitations. - cannot change/add a mac address on the guest virtio-net - cannot create a vlan device on the guest virtio-net - cannot enable promiscuous mode on guest virtio-net To address these limitations, this patch introduces a new mode called 'passthru' when creating a macvlan device which allows takeover of the underlying device and passing it to a guest using virtio with macvtap backend. Only one macvlan device is allowed in passthru mode and it inherits the mac address from the underlying device and sets it in promiscuous mode to receive and forward all the packets. Signed-off-by: Sridhar Samudrala <sri@us.ibm.com> ------------------------------------------------------------------------- Signed-off-by: David S. Miller <davem@davemloft.net>

macvlan: Introduce 'passthru' mode to takeover the underlying device
With the current default 'vepa' mode, a KVM guest using virtio with macvtap backend has the following limitations. - cannot change/add a mac address on the guest virtio-net - cannot create a vlan device on the guest virtio-net - cannot enable promiscuous mode on guest virtio-net To address these limitations, this patch introduces a new mode called 'passthru' when creating a macvlan device which allows takeover of the underlying device and passing it to a guest using virtio with macvtap backend. Only one macvlan device is allowed in passthru mode and it inherits the mac address from the underlying device and sets it in promiscuous mode to receive and forward all the packets. Signed-off-by: Sridhar Samudrala <sri@us.ibm.com> ------------------------------------------------------------------------- Signed-off-by: David S. Miller <davem@davemloft.net>
Sridhar Samudrala · David S. Miller
1 parent e5700c740d
Showing 2 changed files with 33 additions and 1 deletions Side-by-side Diff
drivers/net/macvlan.c
include/linux/if_link.h
@@ -38,6 +38,7 @@
 	struct hlist_head	vlan_hash[MACVLAN_HASH_SIZE];
 	struct list_head	vlans;
 	struct rcu_head		rcu;
+	bool 			passthru;
 };
  
 #define macvlan_port_get_rcu(dev) \
@@ -169,6 +170,7 @@
 			macvlan_broadcast(skb, port, NULL,
 					  MACVLAN_MODE_PRIVATE |
 					  MACVLAN_MODE_VEPA    |
+					  MACVLAN_MODE_PASSTHRU|
 					  MACVLAN_MODE_BRIDGE);
 		else if (src->mode == MACVLAN_MODE_VEPA)
 			/* flood to everyone except source */
@@ -185,7 +187,10 @@
 		return skb;
 	}
  
-	vlan = macvlan_hash_lookup(port, eth->h_dest);
+	if (port->passthru)
+		vlan = list_first_entry(&port->vlans, struct macvlan_dev, list);
+	else
+		vlan = macvlan_hash_lookup(port, eth->h_dest);
 	if (vlan == NULL)
 		return skb;
  
@@ -288,6 +293,11 @@
 	struct net_device *lowerdev = vlan->lowerdev;
 	int err;
  
+	if (vlan->port->passthru) {
+		dev_set_promiscuity(lowerdev, 1);
+		goto hash_add;
+	}
+
 	err = -EBUSY;
 	if (macvlan_addr_busy(vlan->port, dev->dev_addr))
 		goto out;
@@ -300,6 +310,8 @@
 		if (err < 0)
 			goto del_unicast;
 	}
+
+hash_add:
 	macvlan_hash_add(vlan);
 	return 0;
  
  
@@ -314,12 +326,18 @@
 	struct macvlan_dev *vlan = netdev_priv(dev);
 	struct net_device *lowerdev = vlan->lowerdev;
  
+	if (vlan->port->passthru) {
+		dev_set_promiscuity(lowerdev, -1);
+		goto hash_del;
+	}
+
 	dev_mc_unsync(lowerdev, dev);
 	if (dev->flags & IFF_ALLMULTI)
 		dev_set_allmulti(lowerdev, -1);
  
 	dev_uc_del(lowerdev, dev->dev_addr);
  
+hash_del:
 	macvlan_hash_del(vlan);
 	return 0;
 }
@@ -559,6 +577,7 @@
 	if (port == NULL)
 		return -ENOMEM;
  
+	port->passthru = false;
 	port->dev = dev;
 	INIT_LIST_HEAD(&port->vlans);
 	for (i = 0; i < MACVLAN_HASH_SIZE; i++)
@@ -603,6 +622,7 @@
 		case MACVLAN_MODE_PRIVATE:
 		case MACVLAN_MODE_VEPA:
 		case MACVLAN_MODE_BRIDGE:
+		case MACVLAN_MODE_PASSTHRU:
 			break;
 		default:
 			return -EINVAL;
@@ -652,6 +672,10 @@
 	}
 	port = macvlan_port_get(lowerdev);
  
+	/* Only 1 macvlan device can be created in passthru mode */
+	if (port->passthru)
+		return -EINVAL;
+
 	vlan->lowerdev = lowerdev;
 	vlan->dev      = dev;
 	vlan->port     = port;
@@ -661,6 +685,13 @@
 	vlan->mode     = MACVLAN_MODE_VEPA;
 	if (data && data[IFLA_MACVLAN_MODE])
 		vlan->mode = nla_get_u32(data[IFLA_MACVLAN_MODE]);
+
+	if (vlan->mode == MACVLAN_MODE_PASSTHRU) {
+		if (!list_empty(&port->vlans))
+			return -EINVAL;
+		port->passthru = true;
+		memcpy(dev->dev_addr, lowerdev->dev_addr, ETH_ALEN);
+	}
  
 	err = register_netdevice(dev);
 	if (err < 0)
@@ -259,6 +259,7 @@
 	MACVLAN_MODE_PRIVATE = 1, /* don't talk to other macvlans */
 	MACVLAN_MODE_VEPA    = 2, /* talk to other ports through ext bridge */
 	MACVLAN_MODE_BRIDGE  = 4, /* talk to bridge ports directly */
+	MACVLAN_MODE_PASSTHRU = 8,/* take over the underlying device */
 };
  
 /* SR-IOV virtual function management section */
...	...	@@ -38,6 +38,7 @@
38	38	struct hlist_head vlan_hash[MACVLAN_HASH_SIZE];
39	39	struct list_head vlans;
40	40	struct rcu_head rcu;
	41	+ bool passthru;
41	42	};
42	43
43	44	#define macvlan_port_get_rcu(dev) \
...	...	@@ -169,6 +170,7 @@
169	170	macvlan_broadcast(skb, port, NULL,
170	171	MACVLAN_MODE_PRIVATE \|
171	172	MACVLAN_MODE_VEPA \|
	173	+ MACVLAN_MODE_PASSTHRU\|
172	174	MACVLAN_MODE_BRIDGE);
173	175	else if (src->mode == MACVLAN_MODE_VEPA)
174	176	/* flood to everyone except source */
...	...	@@ -185,7 +187,10 @@
185	187	return skb;
186	188	}
187	189
188		- vlan = macvlan_hash_lookup(port, eth->h_dest);
	190	+ if (port->passthru)
	191	+ vlan = list_first_entry(&port->vlans, struct macvlan_dev, list);
	192	+ else
	193	+ vlan = macvlan_hash_lookup(port, eth->h_dest);
189	194	if (vlan == NULL)
190	195	return skb;
191	196
...	...	@@ -288,6 +293,11 @@
288	293	struct net_device *lowerdev = vlan->lowerdev;
289	294	int err;
290	295
	296	+ if (vlan->port->passthru) {
	297	+ dev_set_promiscuity(lowerdev, 1);
	298	+ goto hash_add;
	299	+ }
	300	+
291	301	err = -EBUSY;
292	302	if (macvlan_addr_busy(vlan->port, dev->dev_addr))
293	303	goto out;
...	...	@@ -300,6 +310,8 @@
300	310	if (err < 0)
301	311	goto del_unicast;
302	312	}
	313	+
	314	+hash_add:
303	315	macvlan_hash_add(vlan);
304	316	return 0;
305	317
306	318
...	...	@@ -314,12 +326,18 @@
314	326	struct macvlan_dev *vlan = netdev_priv(dev);
315	327	struct net_device *lowerdev = vlan->lowerdev;
316	328
	329	+ if (vlan->port->passthru) {
	330	+ dev_set_promiscuity(lowerdev, -1);
	331	+ goto hash_del;
	332	+ }
	333	+
317	334	dev_mc_unsync(lowerdev, dev);
318	335	if (dev->flags & IFF_ALLMULTI)
319	336	dev_set_allmulti(lowerdev, -1);
320	337
321	338	dev_uc_del(lowerdev, dev->dev_addr);
322	339
	340	+hash_del:
323	341	macvlan_hash_del(vlan);
324	342	return 0;
325	343	}
...	...	@@ -559,6 +577,7 @@
559	577	if (port == NULL)
560	578	return -ENOMEM;
561	579
	580	+ port->passthru = false;
562	581	port->dev = dev;
563	582	INIT_LIST_HEAD(&port->vlans);
564	583	for (i = 0; i < MACVLAN_HASH_SIZE; i++)
...	...	@@ -603,6 +622,7 @@
603	622	case MACVLAN_MODE_PRIVATE:
604	623	case MACVLAN_MODE_VEPA:
605	624	case MACVLAN_MODE_BRIDGE:
	625	+ case MACVLAN_MODE_PASSTHRU:
606	626	break;
607	627	default:
608	628	return -EINVAL;
...	...	@@ -652,6 +672,10 @@
652	672	}
653	673	port = macvlan_port_get(lowerdev);
654	674
	675	+ /* Only 1 macvlan device can be created in passthru mode */
	676	+ if (port->passthru)
	677	+ return -EINVAL;
	678	+
655	679	vlan->lowerdev = lowerdev;
656	680	vlan->dev = dev;
657	681	vlan->port = port;
...	...	@@ -661,6 +685,13 @@
661	685	vlan->mode = MACVLAN_MODE_VEPA;
662	686	if (data && data[IFLA_MACVLAN_MODE])
663	687	vlan->mode = nla_get_u32(data[IFLA_MACVLAN_MODE]);
	688	+
	689	+ if (vlan->mode == MACVLAN_MODE_PASSTHRU) {
	690	+ if (!list_empty(&port->vlans))
	691	+ return -EINVAL;
	692	+ port->passthru = true;
	693	+ memcpy(dev->dev_addr, lowerdev->dev_addr, ETH_ALEN);
	694	+ }
664	695
665	696	err = register_netdevice(dev);
666	697	if (err < 0)
...	...	@@ -259,6 +259,7 @@
259	259	MACVLAN_MODE_PRIVATE = 1, /* don't talk to other macvlans */
260	260	MACVLAN_MODE_VEPA = 2, /* talk to other ports through ext bridge */
261	261	MACVLAN_MODE_BRIDGE = 4, /* talk to bridge ports directly */
	262	+ MACVLAN_MODE_PASSTHRU = 8,/* take over the underlying device */
262	263	};
263	264
264	265	/* SR-IOV virtual function management section */