[NET]: Prepare for zerocopy NFS and IPSEC.

- Import va10-hwchecksum-2.5.36.patch
- Import va11-udpsendfile-2.5.36.patch
- Implement new encapsulation friendly ipv4 output path.

9 files changed, 125 insertions, 11 deletions

diff --git a/include/linux/ip.h b/include/linux/ip.h
index 3ba8e804b0d8..e612f5be9cab 100644
--- a/include/linux/ip.h
+++ b/include/linux/ip.h
@@ -137,8 +137,24 @@ struct inet_opt {
 	int			mc_index;	/* Multicast device index */
 	__u32			mc_addr;
 	struct ip_mc_socklist	*mc_list;	/* Group array */
+	struct page		*sndmsg_page;	/* Cached page for sendmsg */
+	u32			sndmsg_off;	/* Cached offset for sendmsg */
+	/*
+	 * Following members are used to retain the infomation to build
+	 * an ip header on each ip fragmentation while the socket is corked.
+	 */
+	struct {
+		unsigned int		flags;
+		unsigned int		fragsize;
+		struct ip_options	*opt;
+		struct rtable		*rt;
+		int			length; /* Total length of all frames */
+		u32			addr;
+	} cork;
 };
 
+#define IPCORK_OPT	1	/* ip-options has been held in ipcork.opt */
+
 struct ipv6_pinfo;
 
 /* WARNING: don't change the layout of the members in inet_sock! */
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 9b6e6ad08e2a..26467b154cdd 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -765,6 +765,15 @@ static inline int skb_headlen(const struct sk_buff *skb)
 	return skb->len - skb->data_len;
 }
 
+static inline int skb_pagelen(const struct sk_buff *skb)
+{
+	int i, len = 0;
+
+	for (i = (int)skb_shinfo(skb)->nr_frags - 1; i >= 0; i--)
+		len += skb_shinfo(skb)->frags[i].size;
+	return len + skb_headlen(skb);
+}
+
 #define SKB_PAGE_ASSERT(skb) do { if (skb_shinfo(skb)->nr_frags) \
 					BUG(); } while (0)
 #define SKB_FRAG_ASSERT(skb) do { if (skb_shinfo(skb)->frag_list) \
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index 91b66d7ab004..3e43be408a06 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -285,8 +285,6 @@ struct tcp_opt {
 
 	struct tcp_func		*af_specific;	/* Operations which are AF_INET{4,6} specific	*/
 	struct sk_buff		*send_head;	/* Front of stuff to transmit			*/
-	struct page		*sndmsg_page;	/* Cached page for sendmsg			*/
-	u32			sndmsg_off;	/* Cached offset for sendmsg			*/
 
  	__u32	rcv_wnd;	/* Current receiver window		*/
 	__u32	rcv_wup;	/* rcv_nxt on last window update sent	*/
diff --git a/include/linux/udp.h b/include/linux/udp.h
index ab75a1e815db..5bdb970a1b69 100644
--- a/include/linux/udp.h
+++ b/include/linux/udp.h
@@ -17,6 +17,9 @@
 #ifndef _LINUX_UDP_H
 #define _LINUX_UDP_H
 
+#include <asm/byteorder.h>
+#include <net/sock.h>
+#include <linux/ip.h>
 
 struct udphdr {
 	__u16	source;
@@ -25,5 +28,33 @@ struct udphdr {
 	__u16	check;
 };
 
+/* UDP socket options */
+#define UDP_CORK	1	/* Never send partially complete segments */
+
+struct udp_opt {
+	int		pending;	/* Any pending frames ? */
+	unsigned int	corkflag;	/* Cork is required */
+	/*
+	 * Following members retains the infomation to create a UDP header
+	 * when the socket is uncorked.
+	 */
+	u32		saddr;		/* source address */
+	u32		daddr;		/* destination address */
+	__u16		sport;		/* source port */
+	__u16		dport;		/* destination port */
+	__u16		len;		/* total length of pending frames */
+};
+
+/* WARNING: don't change the layout of the members in udp_sock! */
+struct udp_sock {
+	struct sock	  sk;
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+	struct ipv6_pinfo *pinet6;
+#endif
+	struct inet_opt	  inet;
+	struct udp_opt	  udp;
+};
+
+#define udp_sk(__sk) (&((struct udp_sock *)__sk)->udp)
 
 #endif	/* _LINUX_UDP_H */
diff --git a/include/net/dst.h b/include/net/dst.h
index d81eb07bbd5b..c27a2e6f497c 100644
--- a/include/net/dst.h
+++ b/include/net/dst.h
@@ -29,6 +29,7 @@ struct dst_entry
 	struct dst_entry        *next;
 	atomic_t		__refcnt;	/* client references	*/
 	int			__use;
+	struct dst_entry	*child;
 	struct net_device       *dev;
 	int			obsolete;
 	int			flags;
@@ -36,6 +37,8 @@ struct dst_entry
 	unsigned long		lastuse;
 	unsigned long		expires;
 
+	unsigned		header_len;	/* more space at head required */
+
 	unsigned		mxlock;
 	unsigned		pmtu;
 	unsigned		window;
@@ -108,18 +111,30 @@ void dst_release(struct dst_entry * dst)
 		atomic_dec(&dst->__refcnt);
 }
 
+/* Children define the path of the packet through the
+ * Linux networking.  Thus, destinations are stackable.
+ */
+
+static inline struct dst_entry *dst_pop(struct dst_entry *dst)
+{
+	struct dst_entry *child = dst_clone(dst->child);
+
+	dst_release(dst);
+	return child;
+}
+
 extern void * dst_alloc(struct dst_ops * ops);
 extern void __dst_free(struct dst_entry * dst);
-extern void dst_destroy(struct dst_entry * dst);
+extern struct dst_entry *dst_destroy(struct dst_entry * dst);
 
-static inline
-void dst_free(struct dst_entry * dst)
+static inline void dst_free(struct dst_entry * dst)
 {
 	if (dst->obsolete > 1)
 		return;
 	if (!atomic_read(&dst->__refcnt)) {
-		dst_destroy(dst);
-		return;
+		dst = dst_destroy(dst);
+		if (!dst)
+			return;
 	}
 	__dst_free(dst);
 }
@@ -155,6 +170,37 @@ static inline void dst_set_expires(struct dst_entry *dst, int timeout)
 		dst->expires = expires;
 }
 
+/* Output packet to network from transport.  */
+static inline int dst_output(struct sk_buff *skb)
+{
+	int err;
+
+	for (;;) {
+		err = skb->dst->output(skb);
+
+		if (likely(err == 0))
+			return err;
+		if (unlikely(err != NET_XMIT_BYPASS))
+			return err;
+	}
+}
+
+/* Input packet from network to transport.  */
+static inline int dst_input(struct sk_buff *skb)
+{
+	int err;
+
+	for (;;) {
+		err = skb->dst->input(skb);
+
+		if (likely(err == 0))
+			return err;
+		/* Oh, Jamal... Seems, I will not forgive you this mess. :-) */
+		if (unlikely(err != NET_XMIT_BYPASS))
+			return err;
+	}
+}
+
 extern void		dst_init(void);
 
 #endif
diff --git a/include/net/ip.h b/include/net/ip.h
index 64f3ca87c050..36de9ce94756 100644
--- a/include/net/ip.h
+++ b/include/net/ip.h
@@ -102,12 +102,26 @@ extern int		ip_build_xmit(struct sock *sk,
 				      int getfrag (const void *,
 						   char *,
 						   unsigned int,
-						   unsigned int),
+						   unsigned int,
+						   struct sk_buff *),
 				      const void *frag,
 				      unsigned length,
 				      struct ipcm_cookie *ipc,
 				      struct rtable *rt,
 				      int flags);
+extern int		ip_append_data(struct sock *sk,
+				       int getfrag(void *from, char *to, int offset, int len,
+						   int odd, struct sk_buff *skb),
+				void *from, int len, int protolen,
+				struct ipcm_cookie *ipc,
+				struct rtable *rt,
+				unsigned int flags);
+extern int		generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb);
+extern ssize_t		ip_append_page(struct sock *sk, struct page *page,
+				int offset, size_t size, int flags);
+extern int		ip_push_pending_frames(struct sock *sk);
+extern void		ip_flush_pending_frames(struct sock *sk);
+
 
 /*
  *	Map a multicast IP onto multicast MAC for type Token Ring.
diff --git a/include/net/sock.h b/include/net/sock.h
index d3706f8ee02c..7240f39e7cf9 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -249,6 +249,8 @@ struct proto {
 					   struct msghdr *msg,
 					int len, int noblock, int flags, 
 					int *addr_len);
+	int			(*sendpage)(struct sock *sk, struct page *page,
+					int offset, size_t size, int flags);
 	int			(*bind)(struct sock *sk, 
 					struct sockaddr *uaddr, int addr_len);
 
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 5b84f5563144..6683b8379936 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1851,7 +1851,7 @@ static inline void tcp_v4_setup_caps(struct sock *sk, struct dst_entry *dst)
 {
 	sk->route_caps = dst->dev->features;
 	if (sk->route_caps & NETIF_F_TSO) {
-		if (sk->no_largesend)
+		if (sk->no_largesend || dst->header_len)
 			sk->route_caps &= ~NETIF_F_TSO;
 	}
 }
diff --git a/include/net/udp.h b/include/net/udp.h
index 7e46b4fb8c0b..63a24bdd3b06 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -76,6 +76,4 @@ extern struct udp_mib udp_statistics[NR_CPUS*2];
 #define UDP_INC_STATS_BH(field)		SNMP_INC_STATS_BH(udp_statistics, field)
 #define UDP_INC_STATS_USER(field) 	SNMP_INC_STATS_USER(udp_statistics, field)
 
-#define udp_sock inet_sock
-
 #endif	/* _UDP_H */