thr3ads.net - Linux Virtualization - [PATCH 0/3] Restore UFO support to virtio

If this information is useful, please help other people find it:
Share via:

Vladislav Yasevich

2015-Jan-26 14:37 UTC

[PATCH 0/3] Restore UFO support to virtio_net devices

commit 3d0ad09412ffe00c9afa201d01effdb6023d09b4
Author: Ben Hutchings <ben at decadent.org.uk>
Date:   Thu Oct 30 18:27:12 2014 +0000

    drivers/net: Disable UFO through virtio

Turned off UFO support to virtio-net based devices due to issues
with IPv6 fragment id generation for UFO packets.  The issue
was that IPv6 UFO/GSO implementation expects the fragment id
to be supplied in skb_shinfo().  However, for packets generated
by the VMs, the fragment id is not supplied which causes all
IPv6 fragments to have the id of 0.

The problem is that turning off UFO support on tap/macvtap
as well as virtio devices caused issues with migrations.  
Migrations would fail when moving a vm from a kernel supporting
expecting UFO to work to the newer kernels that disabled UFO.

This series provides a partial solution to address the migration
issue.  The series reserves a bit in the skb and sets the bit
with the ipv6 fragment id has been generated for the packet.
UFO/GSO code then checks the bit to see if the fragment id
is already present or if a new fragment id needs to be generated.
This solution allows host-originated UFO packets to keep a
better randomized fragment id, as well as generating a randomized
id for VM generated traffic (solving the fragment id 0 issue).

Vladislav Yasevich (3):
  ipv6: Select fragment id during UFO/GSO segmentation if not set.
  Revert "drivers/net, ipv6: Select IPv6 fragment idents for virtio UFO
    packets"
  Revert "drivers/net: Disable UFO through virtio"

 drivers/net/macvtap.c    | 16 ++++++++--------
 drivers/net/tun.c        | 25 +++++++++----------------
 drivers/net/virtio_net.c | 24 ++++++++++--------------
 include/linux/skbuff.h   |  3 ++-
 include/net/ipv6.h       |  2 ++
 net/ipv6/ip6_output.c    |  4 ++--
 net/ipv6/output_core.c   |  9 ++++++++-
 net/ipv6/udp_offload.c   | 10 +++++++++-
 8 files changed, 50 insertions(+), 43 deletions(-)

-- 
1.9.3

Vladislav Yasevich

2015-Jan-26 14:37 UTC

head link

[PATCH 1/3] ipv6: Select fragment id during UFO/GSO segmentation if not set.

If the IPv6 fragment id has not been set and we perform
fragmentation due to UFO, select a new fragment id.
When we store the fragment id into skb_shinfo, set the bit
in the skb so we can re-use the selected id.
This preserves the behavior of UFO packets generated on the
host and solves the issue of id generation for packet sockets
and tap/macvtap devices.

This patch moves ipv6_select_ident() back in to the header file.  
It also provides the helper function that sets skb_shinfo() frag
id and sets the bit.

It also makes sure that we select the fragment id when doing
just gso validation, since it's possible for the packet to
come from an untrusted source (VM) and be forwarded through
a UFO enabled device which will expect the fragment id.

CC: Eric Dumazet <edumazet at google.com>
Signed-off-by: Vladislav Yasevich <vyasevic at redhat.com>
---
 include/linux/skbuff.h |  3 ++-
 include/net/ipv6.h     |  2 ++
 net/ipv6/ip6_output.c  |  4 ++--
 net/ipv6/output_core.c |  9 ++++++++-
 net/ipv6/udp_offload.c | 10 +++++++++-
 5 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 85ab7d7..3ad5203 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -605,7 +605,8 @@ struct sk_buff {
 	__u8			ipvs_property:1;
 	__u8			inner_protocol_type:1;
 	__u8			remcsum_offload:1;
-	/* 3 or 5 bit hole */
+	__u8			ufo_fragid_set:1;
+	/* 2 or 4 bit hole */
 
 #ifdef CONFIG_NET_SCHED
 	__u16			tc_index;	/* traffic control index */
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 4292929..ca6137b 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -671,7 +671,9 @@ static inline int ipv6_addr_diff(const struct in6_addr *a1,
const struct in6_add
 	return __ipv6_addr_diff(a1, a2, sizeof(struct in6_addr));
 }
 
+void ipv6_select_ident(struct frag_hdr *fhdr, struct rt6_info *rt);
 void ipv6_proxy_select_ident(struct sk_buff *skb);
+void ipv6_skb_set_fragid(struct sk_buff *skb, __be32 frag_id);
 
 int ip6_dst_hoplimit(struct dst_entry *dst);
 
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index ce69a12..b940b3f 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -537,7 +537,7 @@ static void ip6_copy_metadata(struct sk_buff *to, struct
sk_buff *from)
 	skb_copy_secmark(to, from);
 }
 
-static void ipv6_select_ident(struct frag_hdr *fhdr, struct rt6_info *rt)
+void ipv6_select_ident(struct frag_hdr *fhdr, struct rt6_info *rt)
 {
 	static u32 ip6_idents_hashrnd __read_mostly;
 	u32 hash, id;
@@ -1092,7 +1092,7 @@ static inline int ip6_ufo_append_data(struct sock *sk,
 				     sizeof(struct frag_hdr)) & ~7;
 	skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
 	ipv6_select_ident(&fhdr, rt);
-	skb_shinfo(skb)->ip6_frag_id = fhdr.identification;
+	ipv6_skb_set_fragid(skb, fhdr.identification);
 
 append:
 	return skb_append_datato_frags(sk, skb, getfrag, from,
diff --git a/net/ipv6/output_core.c b/net/ipv6/output_core.c
index 97f41a3..68f879b 100644
--- a/net/ipv6/output_core.c
+++ b/net/ipv6/output_core.c
@@ -38,10 +38,17 @@ void ipv6_proxy_select_ident(struct sk_buff *skb)
 	hash = __ipv6_addr_jhash(&addrs[0], hash);
 
 	id = ip_idents_reserve(hash, 1);
-	skb_shinfo(skb)->ip6_frag_id = htonl(id);
+	ipv6_skb_set_fragid(skb, htonl(id));
 }
 EXPORT_SYMBOL_GPL(ipv6_proxy_select_ident);
 
+void ipv6_skb_set_fragid(struct sk_buff *skb, __be32 frag_id)
+{
+	skb_shinfo(skb)->ip6_frag_id = frag_id;
+	skb->ufo_fragid_set = 1;
+}
+EXPORT_SYMBOL_GPL(ipv6_skb_set_fragid);
+
 int ip6_find_1stfragopt(struct sk_buff *skb, u8 **nexthdr)
 {
 	u16 offset = sizeof(struct ipv6hdr);
diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c
index b6aa8ed..7cda88d 100644
--- a/net/ipv6/udp_offload.c
+++ b/net/ipv6/udp_offload.c
@@ -52,6 +52,10 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
 
 		skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss);
 
+		/* Set the IPv6 fragment id if not set yet */
+		if (!skb->ufo_fragid_set)
+			ipv6_proxy_select_ident(skb);
+
 		segs = NULL;
 		goto out;
 	}
@@ -108,7 +112,11 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff
*skb,
 		fptr = (struct frag_hdr *)(skb_network_header(skb) + unfrag_ip6hlen);
 		fptr->nexthdr = nexthdr;
 		fptr->reserved = 0;
-		fptr->identification = skb_shinfo(skb)->ip6_frag_id;
+		if (!skb->ufo_fragid_set)
+			fptr->identification = skb_shinfo(skb)->ip6_frag_id;
+		else
+			ipv6_select_ident(fptr,
+					  (struct rt6_info *)skb_dst(skb));
 
 		/* Fragment the skb. ipv6 header and the remaining fields of the
 		 * fragment header are updated in ipv6_gso_segment()
-- 
1.9.3

Vladislav Yasevich

2015-Jan-26 14:37 UTC

head link

[PATCH 2/3] Revert "drivers/net, ipv6: Select IPv6 fragment idents for virtio UFO packets"

This reverts commit 5188cd44c55db3e92cd9e77a40b5baa7ed4340f7.

Now that GSO layer can track if fragment id has been selected
and can allocate on if necessary, we don't need to do this in
tap and macvtap.  This reverts most of the code and only keeps
the new ipv6 fragment id generation function that is still needed.

It also solves this issues for packet sockets which were
missed before.

Fixes: 3d0ad09412ff (drivers/net: Disable UFO through virtio)
Signed-off-by: Vladislav Yasevich <vyasevic at redhat.com>
---
 drivers/net/macvtap.c | 3 ---
 drivers/net/tun.c     | 6 +-----
 2 files changed, 1 insertion(+), 8 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 7df2217..0b86e46 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -17,7 +17,6 @@
 #include <linux/fs.h>
 #include <linux/uio.h>
 
-#include <net/ipv6.h>
 #include <net/net_namespace.h>
 #include <net/rtnetlink.h>
 #include <net/sock.h>
@@ -589,8 +588,6 @@ static int macvtap_skb_from_vnet_hdr(struct macvtap_queue
*q,
 			pr_warn_once("macvtap: %s: using disabled UFO feature; please fix this
program\n",
 				     current->comm);
 			gso_type = SKB_GSO_UDP;
-			if (skb->protocol == htons(ETH_P_IPV6))
-				ipv6_proxy_select_ident(skb);
 			break;
 		default:
 			return -EINVAL;
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 8c8dc16..5ca42b7 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -65,7 +65,6 @@
 #include <linux/nsproxy.h>
 #include <linux/virtio_net.h>
 #include <linux/rcupdate.h>
-#include <net/ipv6.h>
 #include <net/net_namespace.h>
 #include <net/netns/generic.h>
 #include <net/rtnetlink.h>
@@ -1167,8 +1166,6 @@ static ssize_t tun_get_user(struct tun_struct *tun, struct
tun_file *tfile,
 		break;
 	}
 
-	skb_reset_network_header(skb);
-
 	if (gso.gso_type != VIRTIO_NET_HDR_GSO_NONE) {
 		pr_debug("GSO!\n");
 		switch (gso.gso_type & ~VIRTIO_NET_HDR_GSO_ECN) {
@@ -1189,8 +1186,6 @@ static ssize_t tun_get_user(struct tun_struct *tun, struct
tun_file *tfile,
 					    current->comm);
 			}
 			skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
-			if (skb->protocol == htons(ETH_P_IPV6))
-				ipv6_proxy_select_ident(skb);
 			break;
 		}
 		default:
@@ -1221,6 +1216,7 @@ static ssize_t tun_get_user(struct tun_struct *tun, struct
tun_file *tfile,
 		skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG;
 	}
 
+	skb_reset_network_header(skb);
 	skb_probe_transport_header(skb, 0);
 
 	rxhash = skb_get_hash(skb);
-- 
1.9.3

Vladislav Yasevich

2015-Jan-26 14:37 UTC

head link

[PATCH 3/3] Revert "drivers/net: Disable UFO through virtio"

This reverts commit 3d0ad09412ffe00c9afa201d01effdb6023d09b4.

Now that UFO functionality can correctly track the fragment
id has been selected and select a fragment id if necessary,
we can re-enable UFO on tap/macvap and virtio devices.

Signed-off-by: Vladislav Yasevich <vyasevic at redhat.com>
---
 drivers/net/macvtap.c    | 13 ++++++++-----
 drivers/net/tun.c        | 19 ++++++++-----------
 drivers/net/virtio_net.c | 24 ++++++++++--------------
 3 files changed, 26 insertions(+), 30 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 0b86e46..919f4fc 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -80,7 +80,7 @@ static struct cdev macvtap_cdev;
 static const struct proto_ops macvtap_socket_ops;
 
 #define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
-		      NETIF_F_TSO6)
+		      NETIF_F_TSO6 | NETIF_F_UFO)
 #define RX_OFFLOADS (NETIF_F_GRO | NETIF_F_LRO)
 #define TAP_FEATURES (NETIF_F_GSO | NETIF_F_SG)
 
@@ -585,8 +585,6 @@ static int macvtap_skb_from_vnet_hdr(struct macvtap_queue
*q,
 			gso_type = SKB_GSO_TCPV6;
 			break;
 		case VIRTIO_NET_HDR_GSO_UDP:
-			pr_warn_once("macvtap: %s: using disabled UFO feature; please fix this
program\n",
-				     current->comm);
 			gso_type = SKB_GSO_UDP;
 			break;
 		default:
@@ -633,6 +631,8 @@ static void macvtap_skb_to_vnet_hdr(struct macvtap_queue *q,
 			vnet_hdr->gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
 		else if (sinfo->gso_type & SKB_GSO_TCPV6)
 			vnet_hdr->gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
+		else if (sinfo->gso_type & SKB_GSO_UDP)
+			vnet_hdr->gso_type = VIRTIO_NET_HDR_GSO_UDP;
 		else
 			BUG();
 		if (sinfo->gso_type & SKB_GSO_TCP_ECN)
@@ -962,6 +962,9 @@ static int set_offload(struct macvtap_queue *q, unsigned
long arg)
 			if (arg & TUN_F_TSO6)
 				feature_mask |= NETIF_F_TSO6;
 		}
+
+		if (arg & TUN_F_UFO)
+			feature_mask |= NETIF_F_UFO;
 	}
 
 	/* tun/tap driver inverts the usage for TSO offloads, where
@@ -972,7 +975,7 @@ static int set_offload(struct macvtap_queue *q, unsigned
long arg)
 	 * When user space turns off TSO, we turn off GSO/LRO so that
 	 * user-space will not receive TSO frames.
 	 */
-	if (feature_mask & (NETIF_F_TSO | NETIF_F_TSO6))
+	if (feature_mask & (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_UFO))
 		features |= RX_OFFLOADS;
 	else
 		features &= ~RX_OFFLOADS;
@@ -1087,7 +1090,7 @@ static long macvtap_ioctl(struct file *file, unsigned int
cmd,
 	case TUNSETOFFLOAD:
 		/* let the user check for future flags */
 		if (arg & ~(TUN_F_CSUM | TUN_F_TSO4 | TUN_F_TSO6 |
-			    TUN_F_TSO_ECN))
+			    TUN_F_TSO_ECN | TUN_F_UFO))
 			return -EINVAL;
 
 		rtnl_lock();
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 5ca42b7..10f9e40 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -186,7 +186,7 @@ struct tun_struct {
 	struct net_device	*dev;
 	netdev_features_t	set_features;
 #define TUN_USER_FEATURES (NETIF_F_HW_CSUM|NETIF_F_TSO_ECN|NETIF_F_TSO| \
-			  NETIF_F_TSO6)
+			  NETIF_F_TSO6|NETIF_F_UFO)
 
 	int			vnet_hdr_sz;
 	int			sndbuf;
@@ -1176,18 +1176,8 @@ static ssize_t tun_get_user(struct tun_struct *tun,
struct tun_file *tfile,
 			skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
 			break;
 		case VIRTIO_NET_HDR_GSO_UDP:
-		{
-			static bool warned;
-
-			if (!warned) {
-				warned = true;
-				netdev_warn(tun->dev,
-					    "%s: using disabled UFO feature; please fix this program\n",
-					    current->comm);
-			}
 			skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
 			break;
-		}
 		default:
 			tun->dev->stats.rx_frame_errors++;
 			kfree_skb(skb);
@@ -1294,6 +1284,8 @@ static ssize_t tun_put_user(struct tun_struct *tun,
 				gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
 			else if (sinfo->gso_type & SKB_GSO_TCPV6)
 				gso.gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
+			else if (sinfo->gso_type & SKB_GSO_UDP)
+				gso.gso_type = VIRTIO_NET_HDR_GSO_UDP;
 			else {
 				pr_err("unexpected GSO type: "
 				       "0x%x, gso_size %d, hdr_len %d\n",
@@ -1742,6 +1734,11 @@ static int set_offload(struct tun_struct *tun, unsigned
long arg)
 				features |= NETIF_F_TSO6;
 			arg &= ~(TUN_F_TSO4|TUN_F_TSO6);
 		}
+
+		if (arg & TUN_F_UFO) {
+			features |= NETIF_F_UFO;
+			arg &= ~TUN_F_UFO;
+		}
 	}
 
 	/* This gives the user a way to test for new features in future by
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 5ca9771..059fdf1 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -490,17 +490,8 @@ static void receive_buf(struct virtnet_info *vi, struct
receive_queue *rq,
 			skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
 			break;
 		case VIRTIO_NET_HDR_GSO_UDP:
-		{
-			static bool warned;
-
-			if (!warned) {
-				warned = true;
-				netdev_warn(dev,
-					    "host using disabled UFO feature; please fix it\n");
-			}
 			skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
 			break;
-		}
 		case VIRTIO_NET_HDR_GSO_TCPV6:
 			skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;
 			break;
@@ -888,6 +879,8 @@ static int xmit_skb(struct send_queue *sq, struct sk_buff
*skb)
 			hdr->hdr.gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
 		else if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6)
 			hdr->hdr.gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
+		else if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP)
+			hdr->hdr.gso_type = VIRTIO_NET_HDR_GSO_UDP;
 		else
 			BUG();
 		if (skb_shinfo(skb)->gso_type & SKB_GSO_TCP_ECN)
@@ -1748,7 +1741,7 @@ static int virtnet_probe(struct virtio_device *vdev)
 			dev->features |= NETIF_F_HW_CSUM|NETIF_F_SG|NETIF_F_FRAGLIST;
 
 		if (virtio_has_feature(vdev, VIRTIO_NET_F_GSO)) {
-			dev->hw_features |= NETIF_F_TSO
+			dev->hw_features |= NETIF_F_TSO | NETIF_F_UFO
 				| NETIF_F_TSO_ECN | NETIF_F_TSO6;
 		}
 		/* Individual feature bits: what can host handle? */
@@ -1758,9 +1751,11 @@ static int virtnet_probe(struct virtio_device *vdev)
 			dev->hw_features |= NETIF_F_TSO6;
 		if (virtio_has_feature(vdev, VIRTIO_NET_F_HOST_ECN))
 			dev->hw_features |= NETIF_F_TSO_ECN;
+		if (virtio_has_feature(vdev, VIRTIO_NET_F_HOST_UFO))
+			dev->hw_features |= NETIF_F_UFO;
 
 		if (gso)
-			dev->features |= dev->hw_features & NETIF_F_ALL_TSO;
+			dev->features |= dev->hw_features & (NETIF_F_ALL_TSO|NETIF_F_UFO);
 		/* (!csum && gso) case will be fixed by register_netdev() */
 	}
 	if (virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_CSUM))
@@ -1798,7 +1793,8 @@ static int virtnet_probe(struct virtio_device *vdev)
 	/* If we can receive ANY GSO packets, we must allocate large ones. */
 	if (virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_TSO4) ||
 	    virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_TSO6) ||
-	    virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_ECN))
+	    virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_ECN) ||
+	    virtio_has_feature(vdev, VIRTIO_NET_F_GUEST_UFO))
 		vi->big_packets = true;
 
 	if (virtio_has_feature(vdev, VIRTIO_NET_F_MRG_RXBUF))
@@ -1994,9 +1990,9 @@ static struct virtio_device_id id_table[] = {
 static unsigned int features[] = {
 	VIRTIO_NET_F_CSUM, VIRTIO_NET_F_GUEST_CSUM,
 	VIRTIO_NET_F_GSO, VIRTIO_NET_F_MAC,
-	VIRTIO_NET_F_HOST_TSO4, VIRTIO_NET_F_HOST_TSO6,
+	VIRTIO_NET_F_HOST_TSO4, VIRTIO_NET_F_HOST_UFO, VIRTIO_NET_F_HOST_TSO6,
 	VIRTIO_NET_F_HOST_ECN, VIRTIO_NET_F_GUEST_TSO4, VIRTIO_NET_F_GUEST_TSO6,
-	VIRTIO_NET_F_GUEST_ECN,
+	VIRTIO_NET_F_GUEST_ECN, VIRTIO_NET_F_GUEST_UFO,
 	VIRTIO_NET_F_MRG_RXBUF, VIRTIO_NET_F_STATUS, VIRTIO_NET_F_CTRL_VQ,
 	VIRTIO_NET_F_CTRL_RX, VIRTIO_NET_F_CTRL_VLAN,
 	VIRTIO_NET_F_GUEST_ANNOUNCE, VIRTIO_NET_F_MQ,
-- 
1.9.3

Michael S. Tsirkin

2015-Jan-26 15:28 UTC

head link

[PATCH 0/3] Restore UFO support to virtio_net devices

On Mon, Jan 26, 2015 at 09:37:03AM -0500, Vladislav Yasevich
wrote:> commit 3d0ad09412ffe00c9afa201d01effdb6023d09b4
> Author: Ben Hutchings <ben at decadent.org.uk>
> Date:   Thu Oct 30 18:27:12 2014 +0000
> 
>     drivers/net: Disable UFO through virtio
> 
> Turned off UFO support to virtio-net based devices due to issues
> with IPv6 fragment id generation for UFO packets.  The issue
> was that IPv6 UFO/GSO implementation expects the fragment id
> to be supplied in skb_shinfo().  However, for packets generated
> by the VMs, the fragment id is not supplied which causes all
> IPv6 fragments to have the id of 0.
> 
> The problem is that turning off UFO support on tap/macvtap
> as well as virtio devices caused issues with migrations.  
> Migrations would fail when moving a vm from a kernel supporting
> expecting UFO to work to the newer kernels that disabled UFO.
> 
> This series provides a partial solution to address the migration
> issue.  The series reserves a bit in the skb and sets the bit
> with the ipv6 fragment id has been generated for the packet.
> UFO/GSO code then checks the bit to see if the fragment id
> is already present or if a new fragment id needs to be generated.
> This solution allows host-originated UFO packets to keep a
> better randomized fragment id, as well as generating a randomized
> id for VM generated traffic (solving the fragment id 0 issue).
> 
> Vladislav Yasevich (3):
>   ipv6: Select fragment id during UFO/GSO segmentation if not set.
>   Revert "drivers/net, ipv6: Select IPv6 fragment idents for virtio
UFO
>     packets"
>   Revert "drivers/net: Disable UFO through virtio"
> 
>  drivers/net/macvtap.c    | 16 ++++++++--------
>  drivers/net/tun.c        | 25 +++++++++----------------
>  drivers/net/virtio_net.c | 24 ++++++++++--------------
>  include/linux/skbuff.h   |  3 ++-
>  include/net/ipv6.h       |  2 ++
>  net/ipv6/ip6_output.c    |  4 ++--
>  net/ipv6/output_core.c   |  9 ++++++++-
>  net/ipv6/udp_offload.c   | 10 +++++++++-
>  8 files changed, 50 insertions(+), 43 deletions(-)
Should also help a lot to put UDP performance where it was.

Acked-by: Michael S. Tsirkin <mst at redhat.com>


> -- 
> 1.9.3

Michael S. Tsirkin

2015-Jan-26 15:32 UTC

head link

[PATCH 0/3] Restore UFO support to virtio_net devices

On Mon, Jan 26, 2015 at 05:28:42PM +0200, Michael S. Tsirkin
wrote:> On Mon, Jan 26, 2015 at 09:37:03AM -0500, Vladislav Yasevich wrote:
> > commit 3d0ad09412ffe00c9afa201d01effdb6023d09b4
> > Author: Ben Hutchings <ben at decadent.org.uk>
> > Date:   Thu Oct 30 18:27:12 2014 +0000
> > 
> >     drivers/net: Disable UFO through virtio
> > 
> > Turned off UFO support to virtio-net based devices due to issues
> > with IPv6 fragment id generation for UFO packets.  The issue
> > was that IPv6 UFO/GSO implementation expects the fragment id
> > to be supplied in skb_shinfo().  However, for packets generated
> > by the VMs, the fragment id is not supplied which causes all
> > IPv6 fragments to have the id of 0.
> > 
> > The problem is that turning off UFO support on tap/macvtap
> > as well as virtio devices caused issues with migrations.  
> > Migrations would fail when moving a vm from a kernel supporting
> > expecting UFO to work to the newer kernels that disabled UFO.
> > 
> > This series provides a partial solution to address the migration
> > issue.  The series reserves a bit in the skb and sets the bit
> > with the ipv6 fragment id has been generated for the packet.
> > UFO/GSO code then checks the bit to see if the fragment id
> > is already present or if a new fragment id needs to be generated.
> > This solution allows host-originated UFO packets to keep a
> > better randomized fragment id, as well as generating a randomized
> > id for VM generated traffic (solving the fragment id 0 issue).
> > 
> > Vladislav Yasevich (3):
> >   ipv6: Select fragment id during UFO/GSO segmentation if not set.
> >   Revert "drivers/net, ipv6: Select IPv6 fragment idents for
virtio UFO
> >     packets"
> >   Revert "drivers/net: Disable UFO through virtio"
> > 
> >  drivers/net/macvtap.c    | 16 ++++++++--------
> >  drivers/net/tun.c        | 25 +++++++++----------------
> >  drivers/net/virtio_net.c | 24 ++++++++++--------------
> >  include/linux/skbuff.h   |  3 ++-
> >  include/net/ipv6.h       |  2 ++
> >  net/ipv6/ip6_output.c    |  4 ++--
> >  net/ipv6/output_core.c   |  9 ++++++++-
> >  net/ipv6/udp_offload.c   | 10 +++++++++-
> >  8 files changed, 50 insertions(+), 43 deletions(-)
> 
> Should also help a lot to put UDP performance where it was.
> 
> Acked-by: Michael S. Tsirkin <mst at redhat.com>
> 
BTW these might also be a good stable candidate.
> 
> > -- 
> > 1.9.3

Ben Hutchings

2015-Jan-27 02:47 UTC

head link

[PATCH 1/3] ipv6: Select fragment id during UFO/GSO segmentation if not set.

On Mon, 2015-01-26 at 09:37 -0500, Vladislav Yasevich
wrote:> If the IPv6 fragment id has not been set and we perform
> fragmentation due to UFO, select a new fragment id.
> When we store the fragment id into skb_shinfo, set the bit
> in the skb so we can re-use the selected id.
> This preserves the behavior of UFO packets generated on the
> host and solves the issue of id generation for packet sockets
> and tap/macvtap devices.
> 
> This patch moves ipv6_select_ident() back in to the header file.  
> It also provides the helper function that sets skb_shinfo() frag
> id and sets the bit.
> 
> It also makes sure that we select the fragment id when doing
> just gso validation, since it's possible for the packet to
> come from an untrusted source (VM) and be forwarded through
> a UFO enabled device which will expect the fragment id.
> 
> CC: Eric Dumazet <edumazet at google.com>
> Signed-off-by: Vladislav Yasevich <vyasevic at redhat.com>
> ---
>  include/linux/skbuff.h |  3 ++-
>  include/net/ipv6.h     |  2 ++
>  net/ipv6/ip6_output.c  |  4 ++--
>  net/ipv6/output_core.c |  9 ++++++++-
>  net/ipv6/udp_offload.c | 10 +++++++++-
>  5 files changed, 23 insertions(+), 5 deletions(-)
> 
> diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
> index 85ab7d7..3ad5203 100644
> --- a/include/linux/skbuff.h
> +++ b/include/linux/skbuff.h
> @@ -605,7 +605,8 @@ struct sk_buff {
>  	__u8			ipvs_property:1;
>  	__u8			inner_protocol_type:1;
>  	__u8			remcsum_offload:1;
> -	/* 3 or 5 bit hole */
> +	__u8			ufo_fragid_set:1;[...]

Doesn't the flag belong in struct skb_shared_info, rather than struct
sk_buff?  Otherwise this looks fine.

Ben.

-- 
Ben Hutchings
When in doubt, use brute force. - Ken Thompson
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 811 bytes
Desc: This is a digitally signed message part
URL:
<http://lists.linuxfoundation.org/pipermail/virtualization/attachments/20150127/da64381a/attachment.sig>

Ben Hutchings

2015-Jan-27 02:51 UTC

head link

[PATCH 3/3] Revert "drivers/net: Disable UFO through virtio"

On Mon, 2015-01-26 at 09:37 -0500, Vladislav Yasevich
wrote:> This reverts commit 3d0ad09412ffe00c9afa201d01effdb6023d09b4.
> 
> Now that UFO functionality can correctly track the fragment
> id has been selected and select a fragment id if necessary,
> we can re-enable UFO on tap/macvap and virtio devices.
> 
> Signed-off-by: Vladislav Yasevich <vyasevic at redhat.com>[...]

Acked-by: Ben Hutchings <ben at decadent.org.uk>

I think this is sensible even before your other patches.

Ben.

-- 
Ben Hutchings
When in doubt, use brute force. - Ken Thompson
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 811 bytes
Desc: This is a digitally signed message part
URL:
<http://lists.linuxfoundation.org/pipermail/virtualization/attachments/20150127/2262eaa8/attachment.sig>

Maybe Matching Threads

Search for more possibly parallel threads

Linux Virtualization - Jan 2015 - [PATCH 0/3] Restore UFO support to virtio_net devices

[PATCH 0/3] Restore UFO support to virtio_net devices

[PATCH 1/3] ipv6: Select fragment id during UFO/GSO segmentation if not set.

[PATCH 2/3] Revert "drivers/net, ipv6: Select IPv6 fragment idents for virtio UFO packets"

[PATCH 3/3] Revert "drivers/net: Disable UFO through virtio"

[PATCH 0/3] Restore UFO support to virtio_net devices

[PATCH 0/3] Restore UFO support to virtio_net devices

[PATCH 1/3] ipv6: Select fragment id during UFO/GSO segmentation if not set.

[PATCH 3/3] Revert "drivers/net: Disable UFO through virtio"

Maybe Matching Threads