Merge with Linux 2.4.3.

Note that mingetty does no longer work with serial console, you have to
switch to another getty like getty_ps.  This commit also includes a
fix for a setitimer bug which did prevent getty_ps from working on
older kernels.

7 files changed, 370 insertions, 15 deletions

diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 8d668ff41..e35e0d6cb 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -1,7 +1,7 @@
 /*
  *	NET3	IP device support routines.
  *
- *	Version: $Id: devinet.c,v 1.40 2001/02/05 06:03:47 davem Exp $
+ *	Version: $Id: devinet.c,v 1.41 2001/02/18 09:26:26 davem Exp $
  *
  *		This program is free software; you can redistribute it and/or
  *		modify it under the terms of the GNU General Public License
@@ -827,10 +827,8 @@ static int inetdev_event(struct notifier_block *this, unsigned long event, void
 	return NOTIFY_DONE;
 }
 
-struct notifier_block ip_netdev_notifier={
-	inetdev_event,
-	NULL,
-	0
+struct notifier_block ip_netdev_notifier = {
+	notifier_call:	inetdev_event,
 };
 
 #ifdef CONFIG_RTNETLINK
diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index afed5862e..5cd316a26 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -213,18 +213,17 @@ static void ip_evictor(void)
 			if (ipq_hash[i] == NULL)
 				continue;
 
-			write_lock(&ipfrag_lock);
+			read_lock(&ipfrag_lock);
 			if ((qp = ipq_hash[i]) != NULL) {
 				/* find the oldest queue for this hash bucket */
 				while (qp->next)
 					qp = qp->next;
-				__ipq_unlink(qp);
-				write_unlock(&ipfrag_lock);
+				atomic_inc(&qp->refcnt);
+				read_unlock(&ipfrag_lock);
 
 				spin_lock(&qp->lock);
-				if (del_timer(&qp->timer))
-					atomic_dec(&qp->refcnt);
-				qp->last_in |= COMPLETE;
+				if (!(qp->last_in&COMPLETE))
+					ipq_kill(qp);
 				spin_unlock(&qp->lock);
 
 				ipq_put(qp);
@@ -232,7 +231,7 @@ static void ip_evictor(void)
 				progress = 1;
 				continue;
 			}
-			write_unlock(&ipfrag_lock);
+			read_unlock(&ipfrag_lock);
 		}
 	} while (progress);
 }
diff --git a/net/ipv4/netfilter/Config.in b/net/ipv4/netfilter/Config.in
index 5887658fb..cb4789fb3 100644
--- a/net/ipv4/netfilter/Config.in
+++ b/net/ipv4/netfilter/Config.in
@@ -20,6 +20,7 @@ if [ "$CONFIG_IP_NF_IPTABLES" != "n" ]; then
   dep_tristate '  netfilter MARK match support' CONFIG_IP_NF_MATCH_MARK $CONFIG_IP_NF_IPTABLES
   dep_tristate '  Multiple port match support' CONFIG_IP_NF_MATCH_MULTIPORT $CONFIG_IP_NF_IPTABLES
   dep_tristate '  TOS match support' CONFIG_IP_NF_MATCH_TOS $CONFIG_IP_NF_IPTABLES
+  dep_tristate '  tcpmss match support' CONFIG_IP_NF_MATCH_TCPMSS $CONFIG_IP_NF_IPTABLES
   if [ "$CONFIG_IP_NF_CONNTRACK" != "n" ]; then
     dep_tristate '  Connection state match support' CONFIG_IP_NF_MATCH_STATE $CONFIG_IP_NF_CONNTRACK $CONFIG_IP_NF_IPTABLES 
   fi
@@ -45,7 +46,7 @@ if [ "$CONFIG_IP_NF_IPTABLES" != "n" ]; then
       # If they want FTP, set to $CONFIG_IP_NF_NAT (m or y), 
       # or $CONFIG_IP_NF_FTP (m or y), whichever is weaker.  Argh.
       if [ "$CONFIG_IP_NF_FTP" = "m" ]; then
-        define_tristate CONFIG_IP_NF_NAT_FTP m
+	define_tristate CONFIG_IP_NF_NAT_FTP m
       else
         if [ "$CONFIG_IP_NF_FTP" = "y" ]; then
           define_tristate CONFIG_IP_NF_NAT_FTP $CONFIG_IP_NF_NAT
@@ -60,6 +61,7 @@ if [ "$CONFIG_IP_NF_IPTABLES" != "n" ]; then
     dep_tristate '    MARK target support' CONFIG_IP_NF_TARGET_MARK $CONFIG_IP_NF_MANGLE
   fi
   dep_tristate '  LOG target support' CONFIG_IP_NF_TARGET_LOG $CONFIG_IP_NF_IPTABLES
+  dep_tristate '  TCPMSS target support' CONFIG_IP_NF_TARGET_TCPMSS $CONFIG_IP_NF_IPTABLES
 fi
 
 # Backwards compatibility modules: only if you don't build in the others.
diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile
index c40caa75e..5833a87c7 100644
--- a/net/ipv4/netfilter/Makefile
+++ b/net/ipv4/netfilter/Makefile
@@ -54,6 +54,7 @@ obj-$(CONFIG_IP_NF_MATCH_OWNER) += ipt_owner.o
 obj-$(CONFIG_IP_NF_MATCH_TOS) += ipt_tos.o
 obj-$(CONFIG_IP_NF_MATCH_STATE) += ipt_state.o
 obj-$(CONFIG_IP_NF_MATCH_UNCLEAN) += ipt_unclean.o
+obj-$(CONFIG_IP_NF_MATCH_TCPMSS) += ipt_tcpmss.o
 
 # targets
 obj-$(CONFIG_IP_NF_TARGET_REJECT) += ipt_REJECT.o
@@ -63,6 +64,7 @@ obj-$(CONFIG_IP_NF_TARGET_MARK) += ipt_MARK.o
 obj-$(CONFIG_IP_NF_TARGET_MASQUERADE) += ipt_MASQUERADE.o
 obj-$(CONFIG_IP_NF_TARGET_REDIRECT) += ipt_REDIRECT.o
 obj-$(CONFIG_IP_NF_TARGET_LOG) += ipt_LOG.o
+obj-$(CONFIG_IP_NF_TARGET_TCPMSS) += ipt_TCPMSS.o
 
 # backwards compatibility 
 obj-$(CONFIG_IP_NF_COMPAT_IPCHAINS) += ipchains.o
diff --git a/net/ipv4/netfilter/ipt_TCPMSS.c b/net/ipv4/netfilter/ipt_TCPMSS.c
new file mode 100644
index 000000000..9dd50740a
--- /dev/null
+++ b/net/ipv4/netfilter/ipt_TCPMSS.c
@@ -0,0 +1,245 @@
+/*
+ * This is a module which is used for setting the MSS option in TCP packets.
+ *
+ * Copyright (c) 2000 Marc Boucher
+ */
+#include <linux/module.h>
+#include <linux/skbuff.h>
+
+#include <linux/ip.h>
+#include <net/tcp.h>
+
+#include <linux/netfilter_ipv4/ip_tables.h>
+#include <linux/netfilter_ipv4/ipt_TCPMSS.h>
+
+#if 0
+#define DEBUGP printk
+#else
+#define DEBUGP(format, args...)
+#endif
+
+static u_int16_t
+cheat_check(u_int32_t oldvalinv, u_int32_t newval, u_int16_t oldcheck)
+{
+	u_int32_t diffs[] = { oldvalinv, newval };
+	return csum_fold(csum_partial((char *)diffs, sizeof(diffs),
+                                      oldcheck^0xFFFF));
+}
+
+static inline unsigned int
+optlen(const u_int8_t *opt, unsigned int offset)
+{
+	/* Beware zero-length options: make finite progress */
+	if (opt[offset] <= TCPOPT_NOP || opt[offset+1] == 0) return 1;
+	else return opt[offset+1];
+}
+
+static unsigned int
+ipt_tcpmss_target(struct sk_buff **pskb,
+		  unsigned int hooknum,
+		  const struct net_device *in,
+		  const struct net_device *out,
+		  const void *targinfo,
+		  void *userinfo)
+{
+	const struct ipt_tcpmss_info *tcpmssinfo = targinfo;
+	struct tcphdr *tcph;
+	struct iphdr *iph = (*pskb)->nh.iph;
+	u_int16_t tcplen, newtotlen, oldval, newmss;
+	unsigned int i;
+	u_int8_t *opt;
+
+	tcplen = (*pskb)->len - iph->ihl*4;
+
+	tcph = (void *)iph + iph->ihl*4;
+
+	/* Since it passed flags test in tcp match, we know it is is
+	   not a fragment, and has data >= tcp header length.  SYN
+	   packets should not contain data: if they did, then we risk
+	   running over MTU, sending Frag Needed and breaking things
+	   badly. --RR */
+	if (tcplen != tcph->doff*4) {
+		if (net_ratelimit())
+			printk(KERN_ERR
+			       "ipt_tcpmss_target: bad length (%d bytes)\n",
+			       (*pskb)->len);
+		return NF_DROP;
+	}
+
+	if(tcpmssinfo->mss == IPT_TCPMSS_CLAMP_PMTU) {
+		if(!(*pskb)->dst) {
+			if (net_ratelimit())
+				printk(KERN_ERR
+			       		"ipt_tcpmss_target: no dst?! can't determine path-MTU\n");
+			return NF_DROP; /* or IPT_CONTINUE ?? */
+		}
+
+		if((*pskb)->dst->pmtu <= (sizeof(struct iphdr) + sizeof(struct tcphdr))) {
+			if (net_ratelimit())
+				printk(KERN_ERR
+		       			"ipt_tcpmss_target: unknown or invalid path-MTU (%d)\n", (*pskb)->dst->pmtu);
+			return NF_DROP; /* or IPT_CONTINUE ?? */
+		}
+
+		newmss = (*pskb)->dst->pmtu - sizeof(struct iphdr) - sizeof(struct tcphdr);
+	} else
+		newmss = tcpmssinfo->mss;
+
+ 	opt = (u_int8_t *)tcph;
+	for (i = sizeof(struct tcphdr); i < tcph->doff*4; i += optlen(opt, i)){
+		if ((opt[i] == TCPOPT_MSS) &&
+		    ((tcph->doff*4 - i) >= TCPOLEN_MSS) &&
+		    (opt[i+1] == TCPOLEN_MSS)) {
+			u_int16_t oldmss;
+
+			oldmss = (opt[i+2] << 8) | opt[i+3];
+
+			if((tcpmssinfo->mss == IPT_TCPMSS_CLAMP_PMTU) &&
+				(oldmss <= newmss))
+					return IPT_CONTINUE;
+
+			opt[i+2] = (newmss & 0xff00) >> 8;
+			opt[i+3] = (newmss & 0x00ff);
+
+			tcph->check = cheat_check(htons(oldmss)^0xFFFF,
+						  htons(newmss),
+						  tcph->check);
+
+			DEBUGP(KERN_INFO "ipt_tcpmss_target: %u.%u.%u.%u:%hu"
+			       "->%u.%u.%u.%u:%hu changed TCP MSS option"
+			       " (from %u to %u)\n", 
+			       NIPQUAD((*pskb)->nh.iph->saddr),
+			       ntohs(tcph->source),
+			       NIPQUAD((*pskb)->nh.iph->daddr),
+			       ntohs(tcph->dest),
+			       oldmss, newmss);
+			goto retmodified;
+		}
+	}
+
+	/*
+	 * MSS Option not found ?! add it..
+	 */
+	if (skb_tailroom((*pskb)) < TCPOLEN_MSS) {
+		struct sk_buff *newskb;
+
+		newskb = skb_copy_expand(*pskb, skb_headroom(*pskb),
+					 TCPOLEN_MSS, GFP_ATOMIC);
+		if (!newskb) {
+			if (net_ratelimit())
+				printk(KERN_ERR "ipt_tcpmss_target:"
+				       " unable to allocate larger skb\n");
+			return NF_DROP;
+		}
+
+		kfree_skb(*pskb);
+		*pskb = newskb;
+		iph = (*pskb)->nh.iph;
+		tcph = (void *)iph + iph->ihl*4;
+	}
+
+	skb_put((*pskb), TCPOLEN_MSS);
+
+ 	opt = (u_int8_t *)tcph + sizeof(struct tcphdr);
+	memmove(opt + TCPOLEN_MSS, opt, tcplen - sizeof(struct tcphdr));
+
+	tcph->check = cheat_check(htons(tcplen) ^ 0xFFFF,
+				  htons(tcplen + TCPOLEN_MSS), tcph->check);
+	tcplen += TCPOLEN_MSS;
+
+	opt[0] = TCPOPT_MSS;
+	opt[1] = TCPOLEN_MSS;
+	opt[2] = (newmss & 0xff00) >> 8;
+	opt[3] = (newmss & 0x00ff);
+
+	tcph->check = cheat_check(~0, *((u_int32_t *)opt), tcph->check);
+
+	oldval = ((u_int16_t *)tcph)[6];
+	tcph->doff += TCPOLEN_MSS/4;
+	tcph->check = cheat_check(oldval ^ 0xFFFF,
+				  ((u_int16_t *)tcph)[6], tcph->check);
+
+	newtotlen = htons(ntohs(iph->tot_len) + TCPOLEN_MSS);
+	iph->check = cheat_check(iph->tot_len ^ 0xFFFF,
+				 newtotlen, iph->check);
+	iph->tot_len = newtotlen;
+
+	DEBUGP(KERN_INFO "ipt_tcpmss_target: %u.%u.%u.%u:%hu"
+	       "->%u.%u.%u.%u:%hu added TCP MSS option (%u)\n",
+	       NIPQUAD((*pskb)->nh.iph->saddr),
+	       ntohs(tcph->source),
+	       NIPQUAD((*pskb)->nh.iph->daddr),
+	       ntohs(tcph->dest),
+	       newmss);
+
+ retmodified:
+	/* If we had a hardware checksum before, it's now invalid */
+	(*pskb)->ip_summed = CHECKSUM_NONE;
+	(*pskb)->nfcache |= NFC_UNKNOWN | NFC_ALTERED;
+	return IPT_CONTINUE;
+}
+
+#define TH_SYN 0x02
+
+static inline int find_syn_match(const struct ipt_entry_match *m)
+{
+	const struct ipt_tcp *tcpinfo = (const struct ipt_tcp *)m->data;
+
+	if (strcmp(m->u.kernel.match->name, "tcp") == 0
+	    && (tcpinfo->flg_cmp & TH_SYN)
+	    && !(tcpinfo->invflags & IPT_TCP_INV_FLAGS))
+		return 1;
+
+	return 0;
+}
+
+/* Must specify -p tcp --syn/--tcp-flags SYN */
+static int
+ipt_tcpmss_checkentry(const char *tablename,
+		      const struct ipt_entry *e,
+		      void *targinfo,
+		      unsigned int targinfosize,
+		      unsigned int hook_mask)
+{
+	const struct ipt_tcpmss_info *tcpmssinfo = targinfo;
+
+	if (targinfosize != IPT_ALIGN(sizeof(struct ipt_tcpmss_info))) {
+		DEBUGP("ipt_tcpmss_checkentry: targinfosize %u != %u\n",
+		       targinfosize, IPT_ALIGN(sizeof(struct ipt_tcpmss_info)));
+		return 0;
+	}
+
+
+	if((tcpmssinfo->mss == IPT_TCPMSS_CLAMP_PMTU) && 
+			((hook_mask & ~((1 << NF_IP_FORWARD)
+			   	| (1 << NF_IP_LOCAL_OUT)
+			   	| (1 << NF_IP_POST_ROUTING))) != 0)) {
+		printk("TCPMSS: path-MTU clamping only supported in FORWARD, OUTPUT and POSTROUTING hooks\n");
+		return 0;
+	}
+
+	if (e->ip.proto == IPPROTO_TCP
+	    && !(e->ip.invflags & IPT_INV_PROTO)
+	    && IPT_MATCH_ITERATE(e, find_syn_match))
+		return 1;
+
+	printk("TCPMSS: Only works on TCP SYN packets\n");
+	return 0;
+}
+
+static struct ipt_target ipt_tcpmss_reg
+= { { NULL, NULL }, "TCPMSS",
+    ipt_tcpmss_target, ipt_tcpmss_checkentry, NULL, THIS_MODULE };
+
+static int __init init(void)
+{
+	return ipt_register_target(&ipt_tcpmss_reg);
+}
+
+static void __exit fini(void)
+{
+	ipt_unregister_target(&ipt_tcpmss_reg);
+}
+
+module_init(init);
+module_exit(fini);
diff --git a/net/ipv4/netfilter/ipt_tcpmss.c b/net/ipv4/netfilter/ipt_tcpmss.c
new file mode 100644
index 000000000..7dae37a33
--- /dev/null
+++ b/net/ipv4/netfilter/ipt_tcpmss.c
@@ -0,0 +1,108 @@
+/* Kernel module to match TCP MSS values. */
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <net/tcp.h>
+
+#include <linux/netfilter_ipv4/ipt_tcpmss.h>
+#include <linux/netfilter_ipv4/ip_tables.h>
+
+#define TH_SYN 0x02
+
+/* Returns 1 if the mss option is set and matched by the range, 0 otherwise */
+static inline int
+mssoption_match(u_int16_t min, u_int16_t max,
+		const struct tcphdr *tcp,
+		u_int16_t datalen,
+		int invert,
+		int *hotdrop)
+{
+	unsigned int i;
+	const u_int8_t *opt = (u_int8_t *)tcp;
+
+	/* If we don't have the whole header, drop packet. */
+	if (tcp->doff * 4 > datalen) {
+		*hotdrop = 1;
+		return 0;
+	}
+
+	for (i = sizeof(struct tcphdr); i < tcp->doff * 4; ) {
+		if ((opt[i] == TCPOPT_MSS)
+		    && ((tcp->doff * 4 - i) >= TCPOLEN_MSS)
+		    && (opt[i+1] == TCPOLEN_MSS)) {
+			u_int16_t mssval;
+
+			mssval = (opt[i+2] << 8) | opt[i+3];
+			
+			return (mssval >= min && mssval <= max) ^ invert;
+		}
+		if (opt[i] < 2) i++;
+		else i += opt[i+1]?:1;
+	}
+
+	return invert;
+}
+
+static int
+match(const struct sk_buff *skb,
+      const struct net_device *in,
+      const struct net_device *out,
+      const void *matchinfo,
+      int offset,
+      const void *hdr,
+      u_int16_t datalen,
+      int *hotdrop)
+{
+	const struct ipt_tcpmss_match_info *info = matchinfo;
+	const struct tcphdr *tcph = (void *)skb->nh.iph + skb->nh.iph->ihl*4;
+
+	return mssoption_match(info->mss_min, info->mss_max, tcph,
+			       skb->len - skb->nh.iph->ihl*4,
+			       info->invert, hotdrop);
+}
+
+static inline int find_syn_match(const struct ipt_entry_match *m)
+{
+	const struct ipt_tcp *tcpinfo = (const struct ipt_tcp *)m->data;
+
+	if (strcmp(m->u.kernel.match->name, "tcp") == 0
+	    && (tcpinfo->flg_cmp & TH_SYN)
+	    && !(tcpinfo->invflags & IPT_TCP_INV_FLAGS))
+		return 1;
+
+	return 0;
+}
+
+static int
+checkentry(const char *tablename,
+           const struct ipt_ip *ip,
+           void *matchinfo,
+           unsigned int matchsize,
+           unsigned int hook_mask)
+{
+	if (matchsize != IPT_ALIGN(sizeof(struct ipt_tcpmss_match_info)))
+		return 0;
+
+	/* Must specify -p tcp */
+	if (ip->proto != IPPROTO_TCP || (ip->invflags & IPT_INV_PROTO)) {
+		printk("tcpmss: Only works on TCP packets\n");
+		return 0;
+	}
+
+	return 1;
+}
+
+static struct ipt_match tcpmss_match
+= { { NULL, NULL }, "tcpmss", &match, &checkentry, NULL, THIS_MODULE };
+
+static int __init init(void)
+{
+	return ipt_register_match(&tcpmss_match);
+}
+
+static void __exit fini(void)
+{
+	ipt_unregister_match(&tcpmss_match);
+}
+
+module_init(init);
+module_exit(fini);
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index 68d536e10..25f0b0c1e 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -1,7 +1,7 @@
 /*
  * sysctl_net_ipv4.c: sysctl interface to net IPV4 subsystem.
  *
- * $Id: sysctl_net_ipv4.c,v 1.47 2000/10/19 15:51:02 davem Exp $
+ * $Id: sysctl_net_ipv4.c,v 1.48 2001/02/23 01:39:05 davem Exp $
  *
  * Begun April 1, 1996, Mike Shaver.
  * Added /proc/sys/net/ipv4 directory entry (empty =) ). [MS]
@@ -47,10 +47,11 @@ extern int inet_peer_maxttl;
 extern int inet_peer_gc_mintime;
 extern int inet_peer_gc_maxtime;
 
+#ifdef CONFIG_SYSCTL
 static int tcp_retr1_max = 255; 
-
 static int ip_local_port_range_min[] = { 1, 1 };
 static int ip_local_port_range_max[] = { 65535, 65535 };
+#endif
 
 struct ipv4_config ipv4_config;