Merge branch 'Support-fraglist-GRO-GSO'

Steffen Klassert says:

====================
Support fraglist GRO/GSO

This patchset adds support to do GRO/GSO by chaining packets
of the same flow at the SKB frag_list pointer. This avoids
the overhead to merge payloads into one big packet, and
on the other end, if GSO is needed it avoids the overhead
of splitting the big packet back to the native form.

Patch 1 adds netdev feature flags to enable fraglist GRO,
this implements one of the configuration options discussed
at netconf 2019.

Patch 2 adds a netdev software feature set that defaults to off
and assigns the new fraglist GRO feature flag to it.

Patch 3 adds the core infrastructure to do fraglist GRO/GSO.

Patch 4 enables UDP to use fraglist GRO/GSO if configured.

I have only meaningful forwarding performance measurements.
I did some tests for the local receive path with netperf and iperf,
but in this case the sender that generates the packets is the
bottleneck. So the benchmarks are not that meaningful for the
receive path.

Paolo Abeni did some benchmarks of the local receive path for the
RFC v2 version of this pachset, results can be found here:

https://www.spinics.net/lists/netdev/msg551158.html

I used my IPsec forwarding test setup for the performance measurements:

           ------------         ------------
        -->| router 1 |-------->| router 2 |--
        |  ------------         ------------  |
        |                                     |
        |       --------------------          |
        --------|Spirent Testcenter|<----------
                --------------------

net-next (September 7th 2019):

Single stream UDP frame size 1460 Bytes: 1.161.000 fps (13.5 Gbps).

----------------------------------------------------------------------

net-next (September 7th 2019) + standard UDP GRO/GSO (not implemented
in this patchset):

Single stream UDP frame size 1460 Bytes: 1.801.000 fps (21 Gbps).

----------------------------------------------------------------------

net-next (September 7th 2019) + fraglist UDP GRO/GSO:

Single stream UDP frame size 1460 Bytes: 2.860.000 fps (33.4 Gbps).

=======================================================================

net-next (January 23th 2020):

Single stream UDP frame size 1460 Bytes: 919.000 fps (10.73 Gbps).

----------------------------------------------------------------------

net-next (January 23th 2020) + fraglist UDP GRO/GSO:

Single stream UDP frame size 1460 Bytes: 2.430.000 fps (28.38 Gbps).

-----------------------------------------------------------------------

Changes from RFC v1:

- Add IPv6 support.
- Split patchset to enable UDP GRO by default before adding
  fraglist GRO support.
- Mark fraglist GRO packets as CHECKSUM_NONE.
- Take a refcount on the first segment skb when doing fraglist
  segmentation. With this we can use the same error handling
  path as with standard segmentation.

Changes from RFC v2:

- Add a netdev feature flag to configure listifyed GRO.
- Fix UDP GRO enabling for IPv6.
- Fix a rcu_read_lock() imbalance.
- Fix error path in skb_segment_list().

Changes from RFC v3:

- Rename NETIF_F_GRO_LIST to NETIF_F_GRO_FRAGLIST and add
  NETIF_F_GSO_FRAGLIST.
- Move introduction of SKB_GSO_FRAGLIST to patch 2.
- Use udpv6_encap_needed_key instead of udp_encap_needed_key in IPv6.
- Move some missplaced code from patch 5 to patch 1 where it belongs to.

Changes from RFC v4:

- Drop the 'UDP: enable GRO by default' patch for now. Standard UDP GRO
  is not changed with this patchset.
- Rebase to net-next current.

Changes fom v1 (December 18th):

- Do a full __copy_skb_header instead of tryng to find the really
  needed subset header fields. Thisa can be done later.
- Mark all fraglist GRO packets with CHECKSUM_UNNECESSARY.
- Rebase to net-next current.

Changes fom v2 (January 24th):

- Do the CHECKSUM_UNNECESSARY setting from IPv4 for IPv6 too.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 200 insertions, 27 deletions

diff --git a/net/core/dev.c b/net/core/dev.c
index c806b078097b..ce8900dbd9ea 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3249,7 +3249,7 @@ struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
 
 	segs = skb_mac_gso_segment(skb, features);
 
-	if (unlikely(skb_needs_check(skb, tx_path) && !IS_ERR(segs)))
+	if (segs != skb && unlikely(skb_needs_check(skb, tx_path) && !IS_ERR(segs)))
 		skb_warn_bad_offload(skb);
 
 	return segs;
@@ -9283,7 +9283,7 @@ int register_netdevice(struct net_device *dev)
 	/* Transfer changeable features to wanted_features and enable
 	 * software offloads (GSO and GRO).
 	 */
-	dev->hw_features |= NETIF_F_SOFT_FEATURES;
+	dev->hw_features |= (NETIF_F_SOFT_FEATURES | NETIF_F_SOFT_FEATURES_OFF);
 	dev->features |= NETIF_F_SOFT_FEATURES;
 
 	if (dev->netdev_ops->ndo_udp_tunnel_add) {
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 48a7029529c9..864cb9e9622f 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -3639,6 +3639,97 @@ static inline skb_frag_t skb_head_frag_to_page_desc(struct sk_buff *frag_skb)
 	return head_frag;
 }
 
+struct sk_buff *skb_segment_list(struct sk_buff *skb,
+				 netdev_features_t features,
+				 unsigned int offset)
+{
+	struct sk_buff *list_skb = skb_shinfo(skb)->frag_list;
+	unsigned int tnl_hlen = skb_tnl_header_len(skb);
+	unsigned int delta_truesize = 0;
+	unsigned int delta_len = 0;
+	struct sk_buff *tail = NULL;
+	struct sk_buff *nskb;
+
+	skb_push(skb, -skb_network_offset(skb) + offset);
+
+	skb_shinfo(skb)->frag_list = NULL;
+
+	do {
+		nskb = list_skb;
+		list_skb = list_skb->next;
+
+		if (!tail)
+			skb->next = nskb;
+		else
+			tail->next = nskb;
+
+		tail = nskb;
+
+		delta_len += nskb->len;
+		delta_truesize += nskb->truesize;
+
+		skb_push(nskb, -skb_network_offset(nskb) + offset);
+
+		 __copy_skb_header(nskb, skb);
+
+		skb_headers_offset_update(nskb, skb_headroom(nskb) - skb_headroom(skb));
+		skb_copy_from_linear_data_offset(skb, -tnl_hlen,
+						 nskb->data - tnl_hlen,
+						 offset + tnl_hlen);
+
+		if (skb_needs_linearize(nskb, features) &&
+		    __skb_linearize(nskb))
+			goto err_linearize;
+
+	} while (list_skb);
+
+	skb->truesize = skb->truesize - delta_truesize;
+	skb->data_len = skb->data_len - delta_len;
+	skb->len = skb->len - delta_len;
+
+	skb_gso_reset(skb);
+
+	skb->prev = tail;
+
+	if (skb_needs_linearize(skb, features) &&
+	    __skb_linearize(skb))
+		goto err_linearize;
+
+	skb_get(skb);
+
+	return skb;
+
+err_linearize:
+	kfree_skb_list(skb->next);
+	skb->next = NULL;
+	return ERR_PTR(-ENOMEM);
+}
+EXPORT_SYMBOL_GPL(skb_segment_list);
+
+int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb)
+{
+	if (unlikely(p->len + skb->len >= 65536))
+		return -E2BIG;
+
+	if (NAPI_GRO_CB(p)->last == p)
+		skb_shinfo(p)->frag_list = skb;
+	else
+		NAPI_GRO_CB(p)->last->next = skb;
+
+	skb_pull(skb, skb_gro_offset(skb));
+
+	NAPI_GRO_CB(p)->last = skb;
+	NAPI_GRO_CB(p)->count++;
+	p->data_len += skb->len;
+	p->truesize += skb->truesize;
+	p->len += skb->len;
+
+	NAPI_GRO_CB(skb)->same_flow = 1;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(skb_gro_receive_list);
+
 /**
  *	skb_segment - Perform protocol segmentation on skb.
  *	@head_skb: buffer to segment
diff --git a/net/ethtool/common.c b/net/ethtool/common.c
index e621b1694d2f..c7b8956c3827 100644
--- a/net/ethtool/common.c
+++ b/net/ethtool/common.c
@@ -59,6 +59,7 @@ const char netdev_features_strings[NETDEV_FEATURE_COUNT][ETH_GSTRING_LEN] = {
 	[NETIF_F_HW_TLS_RECORD_BIT] =	"tls-hw-record",
 	[NETIF_F_HW_TLS_TX_BIT] =	 "tls-hw-tx-offload",
 	[NETIF_F_HW_TLS_RX_BIT] =	 "tls-hw-rx-offload",
+	[NETIF_F_GRO_FRAGLIST_BIT] =	 "rx-gro-list",
 };
 
 const char
diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c
index b25e42100ceb..1a98583a79f4 100644
--- a/net/ipv4/udp_offload.c
+++ b/net/ipv4/udp_offload.c
@@ -184,6 +184,20 @@ out_unlock:
 }
 EXPORT_SYMBOL(skb_udp_tunnel_segment);
 
+static struct sk_buff *__udp_gso_segment_list(struct sk_buff *skb,
+					      netdev_features_t features)
+{
+	unsigned int mss = skb_shinfo(skb)->gso_size;
+
+	skb = skb_segment_list(skb, features, skb_mac_header_len(skb));
+	if (IS_ERR(skb))
+		return skb;
+
+	udp_hdr(skb)->len = htons(sizeof(struct udphdr) + mss);
+
+	return skb;
+}
+
 struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
 				  netdev_features_t features)
 {
@@ -196,6 +210,9 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
 	__sum16 check;
 	__be16 newlen;
 
+	if (skb_shinfo(gso_skb)->gso_type & SKB_GSO_FRAGLIST)
+		return __udp_gso_segment_list(gso_skb, features);
+
 	mss = skb_shinfo(gso_skb)->gso_size;
 	if (gso_skb->len <= sizeof(*uh) + mss)
 		return ERR_PTR(-EINVAL);
@@ -354,6 +371,7 @@ static struct sk_buff *udp_gro_receive_segment(struct list_head *head,
 	struct udphdr *uh2;
 	struct sk_buff *p;
 	unsigned int ulen;
+	int ret = 0;
 
 	/* requires non zero csum, for symmetry with GSO */
 	if (!uh->check) {
@@ -369,7 +387,6 @@ static struct sk_buff *udp_gro_receive_segment(struct list_head *head,
 	}
 	/* pull encapsulating udp header */
 	skb_gro_pull(skb, sizeof(struct udphdr));
-	skb_gro_postpull_rcsum(skb, uh, sizeof(struct udphdr));
 
 	list_for_each_entry(p, head, list) {
 		if (!NAPI_GRO_CB(p)->same_flow)
@@ -383,14 +400,40 @@ static struct sk_buff *udp_gro_receive_segment(struct list_head *head,
 			continue;
 		}
 
+		if (NAPI_GRO_CB(skb)->is_flist != NAPI_GRO_CB(p)->is_flist) {
+			NAPI_GRO_CB(skb)->flush = 1;
+			return p;
+		}
+
 		/* Terminate the flow on len mismatch or if it grow "too much".
 		 * Under small packet flood GRO count could elsewhere grow a lot
 		 * leading to excessive truesize values.
 		 * On len mismatch merge the first packet shorter than gso_size,
 		 * otherwise complete the GRO packet.
 		 */
-		if (ulen > ntohs(uh2->len) || skb_gro_receive(p, skb) ||
-		    ulen != ntohs(uh2->len) ||
+		if (ulen > ntohs(uh2->len)) {
+			pp = p;
+		} else {
+			if (NAPI_GRO_CB(skb)->is_flist) {
+				if (!pskb_may_pull(skb, skb_gro_offset(skb))) {
+					NAPI_GRO_CB(skb)->flush = 1;
+					return NULL;
+				}
+				if ((skb->ip_summed != p->ip_summed) ||
+				    (skb->csum_level != p->csum_level)) {
+					NAPI_GRO_CB(skb)->flush = 1;
+					return NULL;
+				}
+				ret = skb_gro_receive_list(p, skb);
+			} else {
+				skb_gro_postpull_rcsum(skb, uh,
+						       sizeof(struct udphdr));
+
+				ret = skb_gro_receive(p, skb);
+			}
+		}
+
+		if (ret || ulen != ntohs(uh2->len) ||
 		    NAPI_GRO_CB(p)->count >= UDP_GRO_CNT_MAX)
 			pp = p;
 
@@ -401,36 +444,29 @@ static struct sk_buff *udp_gro_receive_segment(struct list_head *head,
 	return NULL;
 }
 
-INDIRECT_CALLABLE_DECLARE(struct sock *udp6_lib_lookup_skb(struct sk_buff *skb,
-						   __be16 sport, __be16 dport));
 struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb,
-				struct udphdr *uh, udp_lookup_t lookup)
+				struct udphdr *uh, struct sock *sk)
 {
 	struct sk_buff *pp = NULL;
 	struct sk_buff *p;
 	struct udphdr *uh2;
 	unsigned int off = skb_gro_offset(skb);
 	int flush = 1;
-	struct sock *sk;
 
-	rcu_read_lock();
-	sk = INDIRECT_CALL_INET(lookup, udp6_lib_lookup_skb,
-				udp4_lib_lookup_skb, skb, uh->source, uh->dest);
-	if (!sk)
-		goto out_unlock;
+	if (skb->dev->features & NETIF_F_GRO_FRAGLIST)
+		NAPI_GRO_CB(skb)->is_flist = sk ? !udp_sk(sk)->gro_enabled: 1;
 
-	if (udp_sk(sk)->gro_enabled) {
+	if ((sk && udp_sk(sk)->gro_enabled) || NAPI_GRO_CB(skb)->is_flist) {
 		pp = call_gro_receive(udp_gro_receive_segment, head, skb);
-		rcu_read_unlock();
 		return pp;
 	}
 
-	if (NAPI_GRO_CB(skb)->encap_mark ||
+	if (!sk || NAPI_GRO_CB(skb)->encap_mark ||
 	    (skb->ip_summed != CHECKSUM_PARTIAL &&
 	     NAPI_GRO_CB(skb)->csum_cnt == 0 &&
 	     !NAPI_GRO_CB(skb)->csum_valid) ||
 	    !udp_sk(sk)->gro_receive)
-		goto out_unlock;
+		goto out;
 
 	/* mark that this skb passed once through the tunnel gro layer */
 	NAPI_GRO_CB(skb)->encap_mark = 1;
@@ -457,8 +493,7 @@ struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb,
 	skb_gro_postpull_rcsum(skb, uh, sizeof(struct udphdr));
 	pp = call_gro_receive_sk(udp_sk(sk)->gro_receive, sk, head, skb);
 
-out_unlock:
-	rcu_read_unlock();
+out:
 	skb_gro_flush_final(skb, pp, flush);
 	return pp;
 }
@@ -468,8 +503,10 @@ INDIRECT_CALLABLE_SCOPE
 struct sk_buff *udp4_gro_receive(struct list_head *head, struct sk_buff *skb)
 {
 	struct udphdr *uh = udp_gro_udphdr(skb);
+	struct sk_buff *pp;
+	struct sock *sk;
 
-	if (unlikely(!uh) || !static_branch_unlikely(&udp_encap_needed_key))
+	if (unlikely(!uh))
 		goto flush;
 
 	/* Don't bother verifying checksum if we're going to flush anyway. */
@@ -484,7 +521,11 @@ struct sk_buff *udp4_gro_receive(struct list_head *head, struct sk_buff *skb)
 					     inet_gro_compute_pseudo);
 skip:
 	NAPI_GRO_CB(skb)->is_ipv6 = 0;
-	return udp_gro_receive(head, skb, uh, udp4_lib_lookup_skb);
+	rcu_read_lock();
+	sk = static_branch_unlikely(&udp_encap_needed_key) ? udp4_lib_lookup_skb(skb, uh->source, uh->dest) : NULL;
+	pp = udp_gro_receive(head, skb, uh, sk);
+	rcu_read_unlock();
+	return pp;
 
 flush:
 	NAPI_GRO_CB(skb)->flush = 1;
@@ -517,9 +558,7 @@ int udp_gro_complete(struct sk_buff *skb, int nhoff,
 	rcu_read_lock();
 	sk = INDIRECT_CALL_INET(lookup, udp6_lib_lookup_skb,
 				udp4_lib_lookup_skb, skb, uh->source, uh->dest);
-	if (sk && udp_sk(sk)->gro_enabled) {
-		err = udp_gro_complete_segment(skb);
-	} else if (sk && udp_sk(sk)->gro_complete) {
+	if (sk && udp_sk(sk)->gro_complete) {
 		skb_shinfo(skb)->gso_type = uh->check ? SKB_GSO_UDP_TUNNEL_CSUM
 					: SKB_GSO_UDP_TUNNEL;
 
@@ -529,6 +568,8 @@ int udp_gro_complete(struct sk_buff *skb, int nhoff,
 		skb->encapsulation = 1;
 		err = udp_sk(sk)->gro_complete(sk, skb,
 				nhoff + sizeof(struct udphdr));
+	} else {
+		err = udp_gro_complete_segment(skb);
 	}
 	rcu_read_unlock();
 
@@ -544,6 +585,23 @@ INDIRECT_CALLABLE_SCOPE int udp4_gro_complete(struct sk_buff *skb, int nhoff)
 	const struct iphdr *iph = ip_hdr(skb);
 	struct udphdr *uh = (struct udphdr *)(skb->data + nhoff);
 
+	if (NAPI_GRO_CB(skb)->is_flist) {
+		uh->len = htons(skb->len - nhoff);
+
+		skb_shinfo(skb)->gso_type |= (SKB_GSO_FRAGLIST|SKB_GSO_UDP_L4);
+		skb_shinfo(skb)->gso_segs = NAPI_GRO_CB(skb)->count;
+
+		if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
+			if (skb->csum_level < SKB_MAX_CSUM_LEVEL)
+				skb->csum_level++;
+		} else {
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+			skb->csum_level = 0;
+		}
+
+		return 0;
+	}
+
 	if (uh->check)
 		uh->check = ~udp_v4_check(skb->len - nhoff, iph->saddr,
 					  iph->daddr, 0);
diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c
index f0d5fc27d0b5..584157a07759 100644
--- a/net/ipv6/udp_offload.c
+++ b/net/ipv6/udp_offload.c
@@ -115,8 +115,10 @@ INDIRECT_CALLABLE_SCOPE
 struct sk_buff *udp6_gro_receive(struct list_head *head, struct sk_buff *skb)
 {
 	struct udphdr *uh = udp_gro_udphdr(skb);
+	struct sk_buff *pp;
+	struct sock *sk;
 
-	if (unlikely(!uh) || !static_branch_unlikely(&udpv6_encap_needed_key))
+	if (unlikely(!uh))
 		goto flush;
 
 	/* Don't bother verifying checksum if we're going to flush anyway. */
@@ -132,7 +134,11 @@ struct sk_buff *udp6_gro_receive(struct list_head *head, struct sk_buff *skb)
 
 skip:
 	NAPI_GRO_CB(skb)->is_ipv6 = 1;
-	return udp_gro_receive(head, skb, uh, udp6_lib_lookup_skb);
+	rcu_read_lock();
+	sk = static_branch_unlikely(&udpv6_encap_needed_key) ? udp6_lib_lookup_skb(skb, uh->source, uh->dest) : NULL;
+	pp = udp_gro_receive(head, skb, uh, sk);
+	rcu_read_unlock();
+	return pp;
 
 flush:
 	NAPI_GRO_CB(skb)->flush = 1;
@@ -144,6 +150,23 @@ INDIRECT_CALLABLE_SCOPE int udp6_gro_complete(struct sk_buff *skb, int nhoff)
 	const struct ipv6hdr *ipv6h = ipv6_hdr(skb);
 	struct udphdr *uh = (struct udphdr *)(skb->data + nhoff);
 
+	if (NAPI_GRO_CB(skb)->is_flist) {
+		uh->len = htons(skb->len - nhoff);
+
+		skb_shinfo(skb)->gso_type |= (SKB_GSO_FRAGLIST|SKB_GSO_UDP_L4);
+		skb_shinfo(skb)->gso_segs = NAPI_GRO_CB(skb)->count;
+
+		if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
+			if (skb->csum_level < SKB_MAX_CSUM_LEVEL)
+				skb->csum_level++;
+		} else {
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+			skb->csum_level = 0;
+		}
+
+		return 0;
+	}
+
 	if (uh->check)
 		uh->check = ~udp_v6_check(skb->len - nhoff, &ipv6h->saddr,
 					  &ipv6h->daddr, 0);