A million changes, tunl0 was only ever used for checking if the ipip module is loaded. So, Instead of trying to modify it to check if it exists, We now just use get tunnel. We improved some errors and added some of our own. We made more changes than I can remember right now, Please enjoy.

src/igs.c

@@ -306,16 +306,16 @@ void init_internet_gateway_search(void)
 	/*
 	 * Bring tunl0 up (just to test if the ipip module is loaded)
 	 */
-	loginfo("Configuring the \"" DEFAULT_TUNL_IF "\" tunnel device");
+	loginfo("Checking if \"" DEFAULT_TUNL_IF "\" exists");
 	if(tunnel_change(0, 0, 0, DEFAULT_TUNL_PREFIX, DEFAULT_TUNL_NUMBER) < 0) {
-		printf("Cannot initialize \"" DEFAULT_TUNL_IF "\". "
+		printf("Cannot read \"" DEFAULT_TUNL_IF "\". "
 			"Is the \"ipip\" kernel module loaded?\n"
 			"  If you don't care about using the shared internet "
 			"connections of the ntk nodes\n"
 			"  around you, disable the \"use_shared_inet\" option "
 			"in netsukuku.conf");
-			//del_resolv_conf("nameserver 127.0.0.1", "/etc/resolv.conf");
-			//exit(1);
+			del_resolv_conf("nameserver 127.0.0.1", "/etc/resolv.conf");
+			exit(1);
 			}
 	ifs_del_all_name(me.cur_ifs, &me.cur_ifs_n, NTK_TUNL_PREFIX);
 	ifs_del_all_name(me.cur_ifs, &me.cur_ifs_n, DEFAULT_TUNL_PREFIX);

src/inet.c

@@ -1055,6 +1055,7 @@ ssize_t inet_sendto(int s, const void *msg, size_t len, int flags,
 	int ret;
 
 	if((err=sendto(s, msg, len, flags, to, tolen))==-1) {
+		error("sendto errno: %d err is: %d", errno, err);
 		switch(errno)
 		{
 			case EMSGSIZE:

src/iptunnel.c

@@ -56,8 +56,7 @@ int tunnel_add(inet_prefix *remote, inet_prefix *local, char *dev,
 int tunnel_change(inet_prefix *remote, inet_prefix *local, char *dev,
 		char *tunl_prefix, int tunl_number)
 {
-	error("The value of siocchgtunnel is: %d", SIOCCHGTUNNEL);
-	return do_add(SIOCCHGTUNNEL, remote, local, dev, tunl_prefix,
+	return do_add(SIOCGETTUNNEL, remote, local, dev, tunl_prefix,
 			tunl_number);
 }
 
@@ -248,7 +247,6 @@ int do_get(char *dev)
 static int do_add(int cmd, inet_prefix *remote, inet_prefix *local, char *dev,
 		char *tunl_prefix, int tunl_number)
 {
-    error("This is the do_add function.");
 	struct ip_tunnel_parm p;
 
 	if (fill_tunnel_parm(cmd, remote, local, dev, tunl_prefix,

src/iptunnel.h

@@ -22,8 +22,8 @@
 #include "if.h"
 
 #define DEFAULT_TUNL_PREFIX	"tunl"
-#define DEFAULT_TUNL_NUMBER	1	/* The permanent tunl0 device */
-#define DEFAULT_TUNL_IF		"tunl1"
+#define DEFAULT_TUNL_NUMBER	0	/* The permanent tunl0 device */
+#define DEFAULT_TUNL_IF		"tunl0"
 #define NTK_TUNL_PREFIX		"ntk_tunl"
 
 /* Usage: printf(TUNL_STRING, TUNL_NUMBER("tunl", x)); */

src/mark.c

@@ -5,7 +5,7 @@
 	     ***********************************************
 	     *******          BEGIN 3/2006          ********
 *************************************************************************
-*                                              				* 
+*                                              				*
 *  This program is free software; you can redistribute it and/or modify	*
 *  it under the terms of the GNU General Public License as published by	*
 *  the Free Software Foundation; either version 2 of the License, or	*
@@ -18,16 +18,16 @@
 *									*
 ************************************************************************/
 
-/* 
+/*
  * This code is written with my blood.
  * My hand was hurt. The keyboard was red.
  * In this code you can find my sacrifice.
  *
  * This code is a netfilter iptc library.
- * iptc is very bad documented: wisdom and 
- * debuggers was my friends to understand 
- * netfilter behavior. 
- * I hope you'll never need to code netfilter 
+ * iptc is very bad documented: wisdom and
+ * debuggers was my friends to understand
+ * netfilter behavior.
+ * I hope you'll never need to code netfilter
  * apps.
  * Memory dumpers are with you.
  */
@@ -47,7 +47,7 @@ static int clean_on_exit;
 static rule_store rr,fr,dr;
 static int dumped;
 
-/* Table init: is too easy for comments. 
+/* Table init: is too easy for comments.
  * Returns:
  * 	0
  * 	-1
@@ -62,8 +62,8 @@ int table_init(const char *table, iptc_handle_t *t)
 	return 0;
 
 }
-/* 
- * insert the rule -rule- on chain -chain- 
+/*
+ * insert the rule -rule- on chain -chain-
  * at the position pos.
  * Returns:
  * 	0
@@ -79,7 +79,7 @@ int insert_rule(const char *rule,iptc_handle_t *t,const char *chain,int pos)
 	}
 	return 0;
 }
-/* 
+/*
  * append the rule -rule- on chain -chain-.
  * Returns:
  * 	0
@@ -105,7 +105,7 @@ int commit_rules(iptc_handle_t *t)
 {
 	int res;
 	res=iptc_commit(t);
-	error("This is the value of res: %i This is the value of t is: %s", res, t);
+	error("This is the value of res: %i This is the value of t is: %p", res, t);
 	if (!res) {
 		error("In commit_rules: %s.",iptc_strerror(errno));
 		err_ret(ERR_NETCOM,-1);
@@ -114,13 +114,13 @@ int commit_rules(iptc_handle_t *t)
 }
 
 
-/* 
+/*
  * Put in -rule- the netfilter rule:
- * 
+ *
  *  -A OUTPUT -o ntk_tunl+ -m conntrack  \
  *  --ctstate RELATED,ESTABLISHED -j CONNMARK \
  *  --restore-mark
- *  
+ *
  * -rule- has to be RESTORE_OUTPUT_RULE_SZ-sized
  */
 void restore_output_rule_init(char *rule)
@@ -132,7 +132,7 @@ void restore_output_rule_init(char *rule)
 	struct ipt_connmark_target_info *icmi;
 
 	memset(rule,0,RESTORE_OUTPUT_RULE_SZ);
-	
+
 	ee=(struct ipt_entry*)(rule);
 	em=(struct ipt_entry_match*)(rule+OFFSET_MATCH);
 	ici=(struct ipt_conntrack_info*)(rule+OFFSET_MATCH_INFO);
@@ -141,14 +141,14 @@ void restore_output_rule_init(char *rule)
 
 	ee->next_offset=RESTORE_OUTPUT_RULE_SZ;
 	ee->target_offset=OFFSET_TARGET;
-	
+
 	snprintf(ee->ip.outiface,IFNAMSIZ,"%s+",NTK_TUNL_PREFIX);
 	memset(ee->ip.outiface_mask,0xFF,strlen(ee->ip.outiface)-1);
 
 	strcpy(em->u.user.name,MOD_CONNTRACK);
 	em->u.match_size=MATCH_SZ;;
 	em->u.user.match_size=em->u.match_size;
-	
+
 	et->u.target_size=TARGET_SZ;
 	et->u.user.target_size=et->u.target_size;
 	strcpy(et->u.user.name,MOD_CONNMARK);
@@ -160,12 +160,12 @@ void restore_output_rule_init(char *rule)
 	icmi->mode=IPT_CONNMARK_RESTORE;
 	icmi->mask= 0xffffffffUL;
 }
-/* 
+/*
  * Put in -rule- the netfilter rule:
- * 
- *  -A POSTROUTING -o ntk_tunl+ -m conntrack 
+ *
+ *  -A POSTROUTING -o ntk_tunl+ -m conntrack
  *  --ctstate NEW -j ntk_mark_chain
- *  
+ *
  * -rule- has to be NTK_FORWARD_RULE_SZ-sized
  */
 void ntk_forward_rule_init(char *rule)
@@ -174,9 +174,9 @@ void ntk_forward_rule_init(char *rule)
 	struct ipt_entry_match *em;
 	struct ipt_entry_target *et;
 	struct ipt_conntrack_info *ici;
-	
+
 	memset(rule,0,NTK_FORWARD_RULE_SZ);
-	
+
 	ee=(struct ipt_entry*)(rule);
 	em=(struct ipt_entry_match*)(rule+IPT_ENTRY_SZ);
 	ici=(struct ipt_conntrack_info*)(rule+OFFSET_MATCH_INFO);
@@ -198,13 +198,13 @@ void ntk_forward_rule_init(char *rule)
 	et->u.user.target_size=et->u.target_size;
 	strcpy(et->u.user.name,NTK_MARK_CHAIN);
 }
-/* 
+/*
  * Put in -rule- the netfilter rule:
- * 
- * 
+ *
+ *
  *  -A ntk_mark_chain -o ntk_tunl<outiface_num>
  *  -j CONNMARK --set-mark <outiface_num>
- *  
+ *
  * -rule- has to be MARK_RULE_SZ-sized
  */
 void mark_rule_init(char *rule,char *outiface,int outiface_num)
@@ -214,7 +214,7 @@ void mark_rule_init(char *rule,char *outiface,int outiface_num)
 	struct ipt_connmark_target_info *icmi;
 
 	memset(rule,0,MARK_RULE_SZ);
-	
+
 	ee=(struct ipt_entry*)(rule);
 	et=(struct ipt_entry_target*)(rule+IPT_ENTRY_SZ);
 	icmi=(struct ipt_connmark_target_info*)(rule+IPT_ENTRY_SZ+IPT_ENTRY_TARGET_SZ);
@@ -232,13 +232,13 @@ void mark_rule_init(char *rule,char *outiface,int outiface_num)
 	memset(ee->ip.outiface_mask,0xFF,strlen(ee->ip.outiface));
 	icmi->mark=outiface_num+1;
 }
-/* 
+/*
  * Put in -rule- the netfilter rule:
- * 
- * 
+ *
+ *
  *  -A PREROUTING -o ntk_tunl+ \
  *  -j CONNMARK --set-mark 25
- *  
+ *
  * -rule- has to be IGW_FILTER_RULE_SZ-sized
  */
 void igw_mark_rule_init(char *rule)
@@ -250,7 +250,7 @@ void igw_mark_rule_init(char *rule)
 	memset(rule,0,IGW_FILTER_RULE_SZ);
 	e=(struct ipt_entry*)rule;
 	et=(struct ipt_entry_target*)(rule+IPT_ENTRY_SZ);
-	
+
 	e->next_offset=IGW_FILTER_RULE_SZ;
 	e->target_offset=IPT_ENTRY_SZ;
 	snprintf(e->ip.iniface,IFNAMSIZ,"%s+",NTK_TUNL_PREFIX);
@@ -263,7 +263,7 @@ void igw_mark_rule_init(char *rule)
 	memcpy(et->data,&res,4);
 }
 /*
- * Build the chain ntk_mark_chain on 
+ * Build the chain ntk_mark_chain on
  * mangle table.
  */
 int ntk_mark_chain_init(iptc_handle_t *t)
@@ -271,14 +271,14 @@ int ntk_mark_chain_init(iptc_handle_t *t)
 	int res;
 	res=iptc_is_chain(NTK_MARK_CHAIN,*t);
 	if (res) {
-		debug(DBG_NORMAL,"In mark_init: bizarre, ntk mangle" 
+		debug(DBG_NORMAL,"In mark_init: bizarre, ntk mangle"
 				 "chain is present yet. it will be flushed.");
 		res=iptc_flush_entries(NTK_MARK_CHAIN,t);
-		if (!res) 
+		if (!res)
 			goto dontwork;
 	} else {
 		res=iptc_create_chain(NTK_MARK_CHAIN,t);
-		if (!res) 
+		if (!res)
 			goto dontwork;
 	}
 	return 0;
@@ -369,7 +369,7 @@ int load_dump_rules()
 	if (!dumped)
 		return 0;
 	fd=open("/usr/share/netsukuku/mark_rules",O_RDONLY );
-	if (fd==-1) 
+	if (fd==-1)
 		return -1;
 	read(fd,&d_rr,sizeof(rule_store));
 	read(fd,&d_fr,sizeof(rule_store));
@@ -439,9 +439,9 @@ int mark_init(int igw)
 		error(err_str);
 		error("Unable to create netfilter forwarding rule.");
 		goto cannot_init;
-	}	
+	}
 	if (igw) {
-		death_loop_rule=1; 
+		death_loop_rule=1;
 		igw_mark_rule_init(rule);
 		res=insert_rule(rule,&t,CHAIN_PREROUTING,0);
 		if (res) {
@@ -449,7 +449,7 @@ int mark_init(int igw)
 			error("Unable to create netfilter igw death loop rule.");
 			death_loop_rule=0;
 			goto cannot_init;
-		}  
+		}
 	}
 	else
 		death_loop_rule=0;
@@ -481,12 +481,12 @@ cannot_init:
 	err_ret(ERR_MRKINI,-1);
 
 }
-/* 
+/*
  * Count the number of rules in ntk_mangle_chain.
  *
- * Returns the number of rules present in 
+ * Returns the number of rules present in
  * this chain.
- */ 
+ */
 int count_ntk_mark_chain(iptc_handle_t *t)
 {
 	int nchain=0;
@@ -533,7 +533,7 @@ int create_mark_rules(int n)
 	if (nchain==-1) {
 		error("In create_mark_rules: can not read ntk_mark_chain.");
 		err_ret(ERR_NETRUL,-1);
-	} 
+	}
 	if (nchain>=n) {
 		debug(DBG_NORMAL,"In create_mark_rules: rules present yet.");
 		return 0;
@@ -555,7 +555,7 @@ int create_mark_rules(int n)
 	return 0;
 }
 /*
- * Deltion function: 
+ * Deltion function:
  * this delete the chain ntk_mark_chain
  * Returns:
  * 	0
@@ -563,21 +563,21 @@ int create_mark_rules(int n)
  */
 
 int delete_ntk_forward_chain(iptc_handle_t *t)
-{	
+{
 	int res;
 
 	res=iptc_is_chain(NTK_MARK_CHAIN,*t);
 	if (!res)
 		return 0;
 	res=iptc_flush_entries(NTK_MARK_CHAIN,t);
-        if (!res) 
+        if (!res)
 		goto cannot_delete;
 	res=iptc_delete_chain(NTK_MARK_CHAIN,t);
-	if (!res) 
+	if (!res)
 		goto cannot_delete;
 	return 0;
-        	
-cannot_delete:	
+
+cannot_delete:
 	error("In delete_ntk_forward_chain: -> %s", iptc_strerror(errno));
 	err_ret(ERR_NETDEL,-1);
 }
@@ -596,7 +596,7 @@ int delete_first_rule(iptc_handle_t *t,const char *chain)
 	if (!res)
 		goto cannot_delete;
 	return 0;
-cannot_delete:	
+cannot_delete:
 	error("In delete_first_rule: -> %s", iptc_strerror(errno));
 	err_ret(ERR_NETDEL,-1);
 }
@@ -624,7 +624,7 @@ int rule_position(rule_store *rule,iptc_handle_t *t)
 	}
 	return found?count:-1;
 }
-/* 
+/*
  * Delete rule -rule.rule- on chain rule.chain.
  * Returns
  * 	0 if deletion is Ok or if nothing
@@ -646,7 +646,7 @@ int delete_rule(rule_store *rule,iptc_handle_t *t)
 	}
 	return 0;
 }
-/* 
+/*
  * clean the rules committed by:
  * 	- mark_init
  * 	- create_mark_rules()
@@ -654,7 +654,7 @@ int delete_rule(rule_store *rule,iptc_handle_t *t)
  * 	0
  * 	-1
  */
-		
+
 int mark_close()
 {
 	iptc_handle_t t;
@@ -666,7 +666,7 @@ int mark_close()
 	}
 	load_dump_rules();
 	res=table_init(MANGLE_TABLE,&t);
-	if (res) 
+	if (res)
 		goto reset_error;
 	res=0;
 	res+=delete_rule(&rr,&t);
@@ -675,13 +675,13 @@ int mark_close()
 		debug(DBG_INSANE,"In mark_close: I'm an IGW: deleting death loop rule.");
 		res+=delete_rule(&dr,&t);
 	}
-	if (res) 
+	if (res)
 		goto reset_error;
 	res=delete_ntk_forward_chain(&t);
 	if (res)
 		goto reset_error;
 	res=commit_rules(&t);
-	if (res) 
+	if (res)
 		goto reset_error;
 	debug(DBG_NORMAL,"Netfilter completely restored.");
 	return 0;