aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorHerbert Xu <herbert@gondor.apana.org.au>2008-12-15 23:41:09 -0800
committerDavid S. Miller <davem@davemloft.net>2008-12-15 23:41:09 -0800
commit73cc19f1556b95976934de236fd9043f7208844f (patch)
tree4ebe390a9b328bba32d149b9f83e998836806f5d
parentnet: Add Generic Receive Offload infrastructure (diff)
downloadlinux-dev-73cc19f1556b95976934de236fd9043f7208844f.tar.xz
linux-dev-73cc19f1556b95976934de236fd9043f7208844f.zip
ipv4: Add GRO infrastructure
This patch adds GRO support for IPv4. The criteria for merging is more stringent than LRO, in particular, we require all fields in the IP header to be identical except for the length, ID and checksum. In addition, the ID must form an arithmetic sequence with a difference of one. The ID requirement might seem overly strict, however, most hardware TSO solutions already obey this rule. Linux itself also obeys this whether GSO is in use or not. In future we could relax this rule by storing the IDs (or rather making sure that we don't drop them when pulling the aggregate skb's tail). Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
-rw-r--r--include/net/protocol.h3
-rw-r--r--net/ipv4/af_inet.c97
2 files changed, 100 insertions, 0 deletions
diff --git a/include/net/protocol.h b/include/net/protocol.h
index 8d024d7cb741..cb2965aa1b62 100644
--- a/include/net/protocol.h
+++ b/include/net/protocol.h
@@ -39,6 +39,9 @@ struct net_protocol {
int (*gso_send_check)(struct sk_buff *skb);
struct sk_buff *(*gso_segment)(struct sk_buff *skb,
int features);
+ struct sk_buff **(*gro_receive)(struct sk_buff **head,
+ struct sk_buff *skb);
+ int (*gro_complete)(struct sk_buff *skb);
unsigned int no_policy:1,
netns_ok:1;
};
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index fe03048c130d..a85595307fa7 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -94,6 +94,7 @@
#include <linux/igmp.h>
#include <linux/inetdevice.h>
#include <linux/netdevice.h>
+#include <net/checksum.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/arp.h>
@@ -1241,6 +1242,100 @@ out:
return segs;
}
+static struct sk_buff **inet_gro_receive(struct sk_buff **head,
+ struct sk_buff *skb)
+{
+ struct net_protocol *ops;
+ struct sk_buff **pp = NULL;
+ struct sk_buff *p;
+ struct iphdr *iph;
+ int flush = 1;
+ int proto;
+ int id;
+
+ if (unlikely(!pskb_may_pull(skb, sizeof(*iph))))
+ goto out;
+
+ iph = ip_hdr(skb);
+ proto = iph->protocol & (MAX_INET_PROTOS - 1);
+
+ rcu_read_lock();
+ ops = rcu_dereference(inet_protos[proto]);
+ if (!ops || !ops->gro_receive)
+ goto out_unlock;
+
+ if (iph->version != 4 || iph->ihl != 5)
+ goto out_unlock;
+
+ if (unlikely(ip_fast_csum((u8 *)iph, iph->ihl)))
+ goto out_unlock;
+
+ flush = ntohs(iph->tot_len) != skb->len ||
+ iph->frag_off != htons(IP_DF);
+ id = ntohs(iph->id);
+
+ for (p = *head; p; p = p->next) {
+ struct iphdr *iph2;
+
+ if (!NAPI_GRO_CB(p)->same_flow)
+ continue;
+
+ iph2 = ip_hdr(p);
+
+ if (iph->protocol != iph2->protocol ||
+ iph->tos != iph2->tos ||
+ memcmp(&iph->saddr, &iph2->saddr, 8)) {
+ NAPI_GRO_CB(p)->same_flow = 0;
+ continue;
+ }
+
+ /* All fields must match except length and checksum. */
+ NAPI_GRO_CB(p)->flush |=
+ memcmp(&iph->frag_off, &iph2->frag_off, 4) ||
+ (u16)(ntohs(iph2->id) + NAPI_GRO_CB(p)->count) != id;
+
+ NAPI_GRO_CB(p)->flush |= flush;
+ }
+
+ NAPI_GRO_CB(skb)->flush |= flush;
+ __skb_pull(skb, sizeof(*iph));
+ skb_reset_transport_header(skb);
+
+ pp = ops->gro_receive(head, skb);
+
+out_unlock:
+ rcu_read_unlock();
+
+out:
+ NAPI_GRO_CB(skb)->flush |= flush;
+
+ return pp;
+}
+
+static int inet_gro_complete(struct sk_buff *skb)
+{
+ struct net_protocol *ops;
+ struct iphdr *iph = ip_hdr(skb);
+ int proto = iph->protocol & (MAX_INET_PROTOS - 1);
+ int err = -ENOSYS;
+ __be16 newlen = htons(skb->len - skb_network_offset(skb));
+
+ csum_replace2(&iph->check, iph->tot_len, newlen);
+ iph->tot_len = newlen;
+
+ rcu_read_lock();
+ ops = rcu_dereference(inet_protos[proto]);
+ if (WARN_ON(!ops || !ops->gro_complete))
+ goto out_unlock;
+
+ err = ops->gro_complete(skb);
+
+out_unlock:
+ rcu_read_unlock();
+
+ return err;
+}
+
int inet_ctl_sock_create(struct sock **sk, unsigned short family,
unsigned short type, unsigned char protocol,
struct net *net)
@@ -1407,6 +1502,8 @@ static struct packet_type ip_packet_type = {
.func = ip_rcv,
.gso_send_check = inet_gso_send_check,
.gso_segment = inet_gso_segment,
+ .gro_receive = inet_gro_receive,
+ .gro_complete = inet_gro_complete,
};
static int __init inet_init(void)