Merge bk://kernel.bkbits.net/davem/sparc-2.5

into home.transmeta.com:/home/torvalds/v2.5/linux

Documentation/networking/ip-sysctl.txt

@@ -281,6 +281,14 @@ tcp_rfc1337 - BOOLEAN
 	assassination.   
 	Default: 0
 
+tcp_low_latency - BOOLEAN
+	If set, the TCP stack makes decisions that prefer lower
+	latency as opposed to higher throughput.  By default, this
+	option is not set meaning that higher throughput is preferred.
+	An example of an application where this default should be
+	changed would be a Beowulf compute cluster.
+	Default: 0
+
 ip_local_port_range - 2 INTEGERS
 	Defines the local port range that is used by TCP and UDP to
 	choose the local port. The first number is the first, the 

drivers/net/ppp_generic.c

@@ -962,12 +962,11 @@ ppp_send_frame(struct ppp *ppp, struct sk_buff *skb)
 		if (ppp->pass_filter.filter
 		    && sk_run_filter(skb, ppp->pass_filter.filter,
 				     ppp->pass_filter.len) == 0) {
-			if (ppp->debug & 1) {
+			if (ppp->debug & 1)
 				printk(KERN_DEBUG "PPP: outbound frame not passed\n");
 			kfree_skb(skb);
 			return;
 		}
-		}
 		/* if this packet passes the active filter, record the time */
 		if (!(ppp->active_filter.filter
 		      && sk_run_filter(skb, ppp->active_filter.filter,

include/linux/sysctl.h

@@ -305,7 +305,8 @@ enum
 	NET_IPV4_ICMP_RATELIMIT=89,
 	NET_IPV4_ICMP_RATEMASK=90,
 	NET_TCP_TW_REUSE=91,
-	NET_TCP_FRTO=92
+	NET_TCP_FRTO=92,
+	NET_TCP_LOW_LATENCY=93
 };
 
 enum {

include/net/ip.h

@@ -96,7 +96,7 @@ extern int		ip_mc_output(struct sk_buff *skb);
 extern int		ip_fragment(struct sk_buff *skb, int (*out)(struct sk_buff*));
 extern int		ip_do_nat(struct sk_buff *skb);
 extern void		ip_send_check(struct iphdr *ip);
-extern int		ip_queue_xmit(struct sk_buff *skb);
+extern int		ip_queue_xmit(struct sk_buff *skb, int ipfragok);
 extern void		ip_init(void);
 extern int		ip_append_data(struct sock *sk,
 				       int getfrag(void *from, char *to, int offset, int len,

include/net/route.h

@@ -167,7 +167,7 @@ static inline int ip_route_connect(struct rtable **rp, u32 dst,
 		ip_rt_put(*rp);
 		*rp = NULL;
 	}
-	return ip_route_output_flow(rp, &fl, sk, 0);
+	return ip_route_output_flow(rp, &fl, sk, 1);
 }
 
 static inline int ip_route_newports(struct rtable **rp, u16 sport, u16 dport,

include/net/tcp.h

@@ -474,6 +474,7 @@ extern int sysctl_tcp_app_win;
 extern int sysctl_tcp_adv_win_scale;
 extern int sysctl_tcp_tw_reuse;
 extern int sysctl_tcp_frto;
+extern int sysctl_tcp_low_latency;
 
 extern atomic_t tcp_memory_allocated;
 extern atomic_t tcp_sockets_allocated;
@@ -566,7 +567,8 @@ static inline void tcp_openreq_free(struct open_request *req)
  */
 
 struct tcp_func {
-	int			(*queue_xmit)		(struct sk_buff *skb);
+	int			(*queue_xmit)		(struct sk_buff *skb,
+							 int ipfragok);
 
 	void			(*send_check)		(struct sock *sk,
 							 struct tcphdr *th,
@@ -1348,7 +1350,7 @@ static __inline__ int tcp_prequeue(struct sock *sk, struct sk_buff *skb)
 {
 	struct tcp_opt *tp = tcp_sk(sk);
 
-	if (tp->ucopy.task) {
+	if (!sysctl_tcp_low_latency && tp->ucopy.task) {
 		__skb_queue_tail(&tp->ucopy.prequeue, skb);
 		tp->ucopy.memory += skb->truesize;
 		if (tp->ucopy.memory > sk->rcvbuf) {

net/ipv4/ah.c

@@ -280,6 +280,8 @@ int ah_input(struct xfrm_state *x, struct sk_buff *skb)
 	    pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
 		goto out;
 
+	skb->ip_summed = CHECKSUM_NONE;
+
 	ah = (struct ip_auth_hdr*)skb->data;
 	iph = skb->nh.iph;
 

net/ipv4/esp.c

@@ -481,6 +481,8 @@ int esp_input(struct xfrm_state *x, struct sk_buff *skb)
 	if ((nfrags = skb_cow_data(skb, 0, &trailer)) < 0)
 		goto out;
 
+	skb->ip_summed = CHECKSUM_NONE;
+
 	esph = (struct ip_esp_hdr*)skb->data;
 	iph = skb->nh.iph;
 

net/ipv4/fib_hash.c

@@ -941,7 +941,7 @@ static __inline__ struct fib_node *fib_get_next(struct seq_file *seq)
 
 			if (!iter->zone)
 				goto out;
-			if (iter->zone->fz_next);
+			if (iter->zone->fz_next)
 				break;
 		}
 		

net/ipv4/ip_output.c

@@ -280,7 +280,7 @@ int ip_output(struct sk_buff *skb)
 		return ip_finish_output(skb);
 }
 
-int ip_queue_xmit(struct sk_buff *skb)
+int ip_queue_xmit(struct sk_buff *skb, int ipfragok)
 {
 	struct sock *sk = skb->sk;
 	struct inet_opt *inet = inet_sk(sk);
@@ -337,7 +337,7 @@ int ip_queue_xmit(struct sk_buff *skb)
 	iph = (struct iphdr *) skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0));
 	*((__u16 *)iph)	= htons((4 << 12) | (5 << 8) | (inet->tos & 0xff));
 	iph->tot_len = htons(skb->len);
-	if (ip_dont_fragment(sk, &rt->u.dst))
+	if (ip_dont_fragment(sk, &rt->u.dst) && !ipfragok)
 		iph->frag_off = htons(IP_DF);
 	else
 		iph->frag_off = 0;

net/ipv4/sysctl_net_ipv4.c

@@ -223,6 +223,8 @@ ctl_table ipv4_table[] = {
 	 &sysctl_tcp_tw_reuse, sizeof(int), 0644, NULL, &proc_dointvec},
 	{NET_TCP_FRTO, "tcp_frto",
 	 &sysctl_tcp_frto, sizeof(int), 0644, NULL, &proc_dointvec},
+	{NET_TCP_LOW_LATENCY, "tcp_low_latency",
+	 &sysctl_tcp_low_latency, sizeof(int), 0644, NULL, &proc_dointvec},
 	{0}
 };
 

net/ipv4/tcp_ipv4.c

@@ -75,6 +75,7 @@
 extern int sysctl_ip_dynaddr;
 extern int sysctl_ip_default_ttl;
 int sysctl_tcp_tw_reuse;
+int sysctl_tcp_low_latency;
 
 /* Check TCP sequence numbers in ICMP packets. */
 #define ICMP_MIN_LENGTH 8
@@ -2049,14 +2050,14 @@ int tcp_v4_tw_remember_stamp(struct tcp_tw_bucket *tw)
 struct tcp_func ipv4_specific = {
 	.queue_xmit	=	ip_queue_xmit,
 	.send_check	=	tcp_v4_send_check,
-	.rebuild_header =tcp_v4_rebuild_header,
+	.rebuild_header	=	tcp_v4_rebuild_header,
 	.conn_request	=	tcp_v4_conn_request,
-	.syn_recv_sock =tcp_v4_syn_recv_sock,
-	.remember_stamp =tcp_v4_remember_stamp,
-	.net_header_len =sizeof(struct iphdr),
+	.syn_recv_sock	=	tcp_v4_syn_recv_sock,
+	.remember_stamp	=	tcp_v4_remember_stamp,
+	.net_header_len	=	sizeof(struct iphdr),
 	.setsockopt	=	ip_setsockopt,
 	.getsockopt	=	ip_getsockopt,
-	.addr2sockaddr =v4_addr2sockaddr,
+	.addr2sockaddr	=	v4_addr2sockaddr,
 	.sockaddr_len	=	sizeof(struct sockaddr_in),
 };
 

net/ipv4/tcp_minisocks.c

@@ -952,6 +952,12 @@ struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
 	if (flg & (TCP_FLAG_RST|TCP_FLAG_SYN))
 		goto embryonic_reset;
 
+	/* ACK sequence verified above, just make sure ACK is
+	 * set.  If ACK not set, just silently drop the packet.
+	 */
+	if (!(flg & TCP_FLAG_ACK))
+		return NULL;
+
 	/* If TCP_DEFER_ACCEPT is set, drop bare ACK. */
 	if (tp->defer_accept && TCP_SKB_CB(skb)->end_seq == req->rcv_isn+1) {
 		req->acked = 1;

net/ipv4/tcp_output.c

@@ -276,7 +276,7 @@ int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb)
 
 		TCP_INC_STATS(TcpOutSegs);
 
-		err = tp->af_specific->queue_xmit(skb);
+		err = tp->af_specific->queue_xmit(skb, 0);
 		if (err <= 0)
 			return err;
 
@@ -786,13 +786,13 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m
 		/* Ok.  We will be able to collapse the packet. */
 		__skb_unlink(next_skb, next_skb->list);
 
+		memcpy(skb_put(skb, next_skb_size), next_skb->data, next_skb_size);
+
 		if (next_skb->ip_summed == CHECKSUM_HW)
 			skb->ip_summed = CHECKSUM_HW;
 
-		if (skb->ip_summed != CHECKSUM_HW) {
-			memcpy(skb_put(skb, next_skb_size), next_skb->data, next_skb_size);
+		if (skb->ip_summed != CHECKSUM_HW)
 			skb->csum = csum_block_add(skb->csum, next_skb->csum, skb_size);
-		}
 
 		/* Update sequence range on original skb. */
 		TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;

net/ipv4/xfrm_user.c

@@ -276,9 +276,11 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr)
 	copy_to_user_state(x, p);
 
 	if (x->aalg)
-		RTA_PUT(skb, XFRMA_ALG_AUTH, sizeof(*(x->aalg)), x->aalg);
+		RTA_PUT(skb, XFRMA_ALG_AUTH,
+			sizeof(*(x->aalg))+(x->aalg->alg_key_len+7)/8, x->aalg);
 	if (x->ealg)
-		RTA_PUT(skb, XFRMA_ALG_CRYPT, sizeof(*(x->ealg)), x->ealg);
+		RTA_PUT(skb, XFRMA_ALG_CRYPT,
+			sizeof(*(x->ealg))+(x->ealg->alg_key_len+7)/8, x->ealg);
 	if (x->calg)
 		RTA_PUT(skb, XFRMA_ALG_COMP, sizeof(*(x->calg)), x->calg);
 
@@ -655,6 +657,7 @@ static int xfrm_dump_policy(struct sk_buff *skb, struct netlink_callback *cb)
 	info.in_skb = cb->skb;
 	info.out_skb = skb;
 	info.nlmsg_seq = cb->nlh->nlmsg_seq;
+	info.this_idx = 0;
 	info.start_idx = cb->args[0];
 	(void) xfrm_policy_walk(dump_one_policy, &info);
 	cb->args[0] = info.this_idx;
@@ -752,7 +755,7 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err
 {
 	struct rtattr *xfrma[XFRMA_MAX];
 	struct xfrm_link *link;
-	int type, min_len, kind;
+	int type, min_len;
 
 	if (!(nlh->nlmsg_flags & NLM_F_REQUEST))
 		return 0;
@@ -768,7 +771,6 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err
 		goto err_einval;
 
 	type -= XFRM_MSG_BASE;
-	kind = (type & 3);
 	link = &xfrm_dispatch[type];
 
 	/* All operations require privileges, even GET */
@@ -777,7 +779,7 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err
 		return -1;
 	}
 
-	if (kind == 2 && (nlh->nlmsg_flags & NLM_F_DUMP)) {
+	if ((type == 2 || type == 5) && (nlh->nlmsg_flags & NLM_F_DUMP)) {
 		u32 rlen;
 
 		if (link->dump == NULL)

net/ipv6/tcp_ipv6.c

@@ -934,11 +934,11 @@ static void tcp_v6_or_free(struct open_request *req)
 }
 
 static struct or_calltable or_ipv6 = {
-	AF_INET6,
-	tcp_v6_send_synack,
-	tcp_v6_or_send_ack,
-	tcp_v6_or_free,
-	tcp_v6_send_reset
+	.family		=	AF_INET6,
+	.rtx_syn_ack	=	tcp_v6_send_synack,
+	.send_ack	=	tcp_v6_or_send_ack,
+	.destructor	=	tcp_v6_or_free,
+	.send_reset	=	tcp_v6_send_reset
 };
 
 static int ipv6_opt_accepted(struct sock *sk, struct sk_buff *skb)
@@ -1746,7 +1746,7 @@ static int tcp_v6_rebuild_header(struct sock *sk)
 	return 0;
 }
 
-static int tcp_v6_xmit(struct sk_buff *skb)
+static int tcp_v6_xmit(struct sk_buff *skb, int ipfragok)
 {
 	struct sock *sk = skb->sk;
 	struct inet_opt *inet = inet_sk(sk);
@@ -1812,18 +1812,18 @@ static int tcp_v6_remember_stamp(struct sock *sk)
 }
 
 static struct tcp_func ipv6_specific = {
-	tcp_v6_xmit,
-	tcp_v6_send_check,
-	tcp_v6_rebuild_header,
-	tcp_v6_conn_request,
-	tcp_v6_syn_recv_sock,
-	tcp_v6_remember_stamp,
-	sizeof(struct ipv6hdr),
-
-	ipv6_setsockopt,
-	ipv6_getsockopt,
-	v6_addr2sockaddr,
-	sizeof(struct sockaddr_in6)
+	.queue_xmit	=	tcp_v6_xmit,
+	.send_check	=	tcp_v6_send_check,
+	.rebuild_header	=	tcp_v6_rebuild_header,
+	.conn_request	=	tcp_v6_conn_request,
+	.syn_recv_sock	=	tcp_v6_syn_recv_sock,
+	.remember_stamp	=	tcp_v6_remember_stamp,
+	.net_header_len	=	sizeof(struct ipv6hdr),
+
+	.setsockopt	=	ipv6_setsockopt,
+	.getsockopt	=	ipv6_getsockopt,
+	.addr2sockaddr	=	v6_addr2sockaddr,
+	.sockaddr_len	=	sizeof(struct sockaddr_in6)
 };
 
 /*
@@ -1831,18 +1831,18 @@ static struct tcp_func ipv6_specific = {
  */
 
 static struct tcp_func ipv6_mapped = {
-	ip_queue_xmit,
-	tcp_v4_send_check,
-	tcp_v4_rebuild_header,
-	tcp_v6_conn_request,
-	tcp_v6_syn_recv_sock,
-	tcp_v4_remember_stamp,
-	sizeof(struct iphdr),
-
-	ipv6_setsockopt,
-	ipv6_getsockopt,
-	v6_addr2sockaddr,
-	sizeof(struct sockaddr_in6)
+	.queue_xmit	=	ip_queue_xmit,
+	.send_check	=	tcp_v4_send_check,
+	.rebuild_header	=	tcp_v4_rebuild_header,
+	.conn_request	=	tcp_v6_conn_request,
+	.syn_recv_sock	=	tcp_v6_syn_recv_sock,
+	.remember_stamp	=	tcp_v4_remember_stamp,
+	.net_header_len	=	sizeof(struct iphdr),
+
+	.setsockopt	=	ipv6_setsockopt,
+	.getsockopt	=	ipv6_getsockopt,
+	.addr2sockaddr	=	v6_addr2sockaddr,
+	.sockaddr_len	=	sizeof(struct sockaddr_in6)
 };
 
 
@@ -2188,7 +2188,7 @@ static struct inet_protosw tcpv6_protosw = {
 	.protocol	=	IPPROTO_TCP,
 	.prot		=	&tcpv6_prot,
 	.ops		=	&inet6_stream_ops,
-	.capability =-1,
+	.capability	=	-1,
 	.no_check	=	0,
 	.flags		=	INET_PROTOSW_PERMANENT,
 };

net/netsyms.c

@@ -248,6 +248,8 @@ EXPORT_SYMBOL(inet_unregister_protosw);
 EXPORT_SYMBOL(ip_route_output_key);
 EXPORT_SYMBOL(ip_route_input);
 EXPORT_SYMBOL(icmp_send);
+EXPORT_SYMBOL(icmp_statistics);
+EXPORT_SYMBOL(icmp_err_convert);
 EXPORT_SYMBOL(ip_options_compile);
 EXPORT_SYMBOL(ip_options_undo);
 EXPORT_SYMBOL(arp_send);
@@ -427,6 +429,7 @@ EXPORT_SYMBOL(sysctl_tcp_wmem);
 EXPORT_SYMBOL(sysctl_tcp_ecn);
 EXPORT_SYMBOL(tcp_cwnd_application_limited);
 EXPORT_SYMBOL(tcp_sendpage);
+EXPORT_SYMBOL(sysctl_tcp_low_latency);
 
 EXPORT_SYMBOL(tcp_write_xmit);