[IPv4]: FIB configuration using struct fib_config

Introduces struct fib_config replacing the ugly struct kern_rta
prone to ordering issues. Avoids creating faked netlink messages
for auto generated routes or requests via ioctl.

A new interface net/nexthop.h is added to help navigate through
nexthop configuration arrays.

A new struct nl_info will be used to carry the necessary netlink
information to be used for notifications later on.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/ip_fib.h

@@ -20,25 +20,32 @@
 #include <linux/seq_file.h>
 #include <net/fib_rules.h>
 
-/* WARNING: The ordering of these elements must match ordering
- *          of RTA_* rtnetlink attribute numbers.
- */
-struct kern_rta {
-	void		*rta_dst;
-	void		*rta_src;
-	int		*rta_iif;
-	int		*rta_oif;
-	void		*rta_gw;
-	u32		*rta_priority;
-	void		*rta_prefsrc;
-	struct rtattr	*rta_mx;
-	struct rtattr	*rta_mp;
-	unsigned char	*rta_protoinfo;
-	u32		*rta_flow;
-	struct rta_cacheinfo *rta_ci;
-	struct rta_session *rta_sess;
-	u32		*rta_mp_alg;
-};
+struct fib_config {
+	u8			fc_family;
+	u8			fc_dst_len;
+	u8			fc_src_len;
+	u8			fc_tos;
+	u8			fc_protocol;
+	u8			fc_scope;
+	u8			fc_type;
+	/* 1 byte unused */
+	u32			fc_table;
+	u32			fc_dst;
+	u32			fc_src;
+	u32			fc_gw;
+	int			fc_oif;
+	u32			fc_flags;
+	u32			fc_priority;
+	u32			fc_prefsrc;
+	struct nlattr		*fc_mx;
+	struct rtnexthop	*fc_mp;
+	int			fc_mx_len;
+	int			fc_mp_len;
+	u32			fc_flow;
+	u32			fc_mp_alg;
+	u32			fc_nlflags;
+	struct nl_info		fc_nlinfo;
+ };
 
 struct fib_info;
 
@@ -154,12 +161,8 @@ struct fib_table {
 	u32		tb_id;
 	unsigned	tb_stamp;
 	int		(*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res);
-	int		(*tb_insert)(struct fib_table *table, struct rtmsg *r,
-				     struct kern_rta *rta, struct nlmsghdr *n,
-				     struct netlink_skb_parms *req);
-	int		(*tb_delete)(struct fib_table *table, struct rtmsg *r,
-				     struct kern_rta *rta, struct nlmsghdr *n,
-				     struct netlink_skb_parms *req);
+	int		(*tb_insert)(struct fib_table *, struct fib_config *);
+	int		(*tb_delete)(struct fib_table *, struct fib_config *);
 	int		(*tb_dump)(struct fib_table *table, struct sk_buff *skb,
 				     struct netlink_callback *cb);
 	int		(*tb_flush)(struct fib_table *table);
@@ -228,8 +231,6 @@ struct rtentry;
 extern int ip_fib_check_default(u32 gw, struct net_device *dev);
 extern int fib_sync_down(u32 local, struct net_device *dev, int force);
 extern int fib_sync_up(struct net_device *dev);
-extern int fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm,
-			       struct kern_rta *rta, struct rtentry *r);
 extern u32  __fib_res_prefsrc(struct fib_result *res);
 
 /* Exported by fib_hash.c */

include/net/netlink.h

@@ -192,6 +192,16 @@ struct nla_policy {
 	u16		minlen;
 };
 
+/**
+ * struct nl_info - netlink source information
+ * @nlh: Netlink message header of original request
+ * @pid: Netlink PID of requesting application
+ */
+struct nl_info {
+	struct nlmsghdr		*nlh;
+	u32			pid;
+};
+
 extern void		netlink_run_queue(struct sock *sk, unsigned int *qlen,
 					  int (*cb)(struct sk_buff *,
 						    struct nlmsghdr *, int *));

include/net/nexthop.h

+#ifndef __NET_NEXTHOP_H
+#define __NET_NEXTHOP_H
+
+#include <linux/rtnetlink.h>
+#include <net/netlink.h>
+
+static inline int rtnh_ok(const struct rtnexthop *rtnh, int remaining)
+{
+	return remaining >= sizeof(*rtnh) &&
+	       rtnh->rtnh_len >= sizeof(*rtnh) &&
+	       rtnh->rtnh_len <= remaining;
+}
+
+static inline struct rtnexthop *rtnh_next(const struct rtnexthop *rtnh,
+                                         int *remaining)
+{
+	int totlen = NLA_ALIGN(rtnh->rtnh_len);
+
+	*remaining -= totlen;
+	return (struct rtnexthop *) ((char *) rtnh + totlen);
+}
+
+static inline struct nlattr *rtnh_attrs(const struct rtnexthop *rtnh)
+{
+	return (struct nlattr *) ((char *) rtnh + NLA_ALIGN(sizeof(*rtnh)));
+}
+
+static inline int rtnh_attrlen(const struct rtnexthop *rtnh)
+{
+	return rtnh->rtnh_len - NLA_ALIGN(sizeof(*rtnh));
+}
+
+#endif

net/ipv4/fib_hash.c

@@ -379,42 +379,39 @@ static struct fib_node *fib_find_node(struct fn_zone *fz, u32 key)
 	return NULL;
 }
 
-static int
-fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *n, struct netlink_skb_parms *req)
+static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct fn_hash *table = (struct fn_hash *) tb->tb_data;
 	struct fib_node *new_f, *f;
 	struct fib_alias *fa, *new_fa;
 	struct fn_zone *fz;
 	struct fib_info *fi;
-	int z = r->rtm_dst_len;
-	int type = r->rtm_type;
-	u8 tos = r->rtm_tos;
+	u8 tos = cfg->fc_tos;
 	u32 key;
 	int err;
 
-	if (z > 32)
+	if (cfg->fc_dst_len > 32)
 		return -EINVAL;
-	fz = table->fn_zones[z];
-	if (!fz && !(fz = fn_new_zone(table, z)))
+
+	fz = table->fn_zones[cfg->fc_dst_len];
+	if (!fz && !(fz = fn_new_zone(table, cfg->fc_dst_len)))
 		return -ENOBUFS;
 
 	key = 0;
-	if (rta->rta_dst) {
-		u32 dst;
-		memcpy(&dst, rta->rta_dst, 4);
-		if (dst & ~FZ_MASK(fz))
+	if (cfg->fc_dst) {
+		if (cfg->fc_dst & ~FZ_MASK(fz))
 			return -EINVAL;
-		key = fz_key(dst, fz);
+		key = fz_key(cfg->fc_dst, fz);
 	}
 
-	if  ((fi = fib_create_info(r, rta, n, &err)) == NULL)
-		return err;
+	fi = fib_create_info(cfg);
+	if (IS_ERR(fi))
+		return PTR_ERR(fi);
 
 	if (fz->fz_nent > (fz->fz_divisor<<1) &&
 	    fz->fz_divisor < FZ_MAX_DIVISOR &&
-	    (z==32 || (1<<z) > fz->fz_divisor))
+	    (cfg->fc_dst_len == 32 ||
+	     (1 << cfg->fc_dst_len) > fz->fz_divisor))
 		fn_rehash_zone(fz);
 
 	f = fib_find_node(fz, key);
@@ -440,18 +437,18 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		struct fib_alias *fa_orig;
 
 		err = -EEXIST;
-		if (n->nlmsg_flags & NLM_F_EXCL)
+		if (cfg->fc_nlflags & NLM_F_EXCL)
 			goto out;
 
-		if (n->nlmsg_flags & NLM_F_REPLACE) {
+		if (cfg->fc_nlflags & NLM_F_REPLACE) {
 			struct fib_info *fi_drop;
 			u8 state;
 
 			write_lock_bh(&fib_hash_lock);
 			fi_drop = fa->fa_info;
 			fa->fa_info = fi;
-			fa->fa_type = type;
-			fa->fa_scope = r->rtm_scope;
+			fa->fa_type = cfg->fc_type;
+			fa->fa_scope = cfg->fc_scope;
 			state = fa->fa_state;
 			fa->fa_state &= ~FA_S_ACCESSED;
 			fib_hash_genid++;
@@ -474,17 +471,17 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 				break;
 			if (fa->fa_info->fib_priority != fi->fib_priority)
 				break;
-			if (fa->fa_type == type &&
-			    fa->fa_scope == r->rtm_scope &&
+			if (fa->fa_type == cfg->fc_type &&
+			    fa->fa_scope == cfg->fc_scope &&
 			    fa->fa_info == fi)
 				goto out;
 		}
-		if (!(n->nlmsg_flags & NLM_F_APPEND))
+		if (!(cfg->fc_nlflags & NLM_F_APPEND))
 			fa = fa_orig;
 	}
 
 	err = -ENOENT;
-	if (!(n->nlmsg_flags&NLM_F_CREATE))
+	if (!(cfg->fc_nlflags & NLM_F_CREATE))
 		goto out;
 
 	err = -ENOBUFS;
@@ -506,8 +503,8 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	new_fa->fa_info = fi;
 	new_fa->fa_tos = tos;
-	new_fa->fa_type = type;
-	new_fa->fa_scope = r->rtm_scope;
+	new_fa->fa_type = cfg->fc_type;
+	new_fa->fa_scope = cfg->fc_scope;
 	new_fa->fa_state = 0;
 
 	/*
@@ -526,7 +523,8 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		fz->fz_nent++;
 	rt_cache_flush(-1);
 
-	rtmsg_fib(RTM_NEWROUTE, key, new_fa, z, tb->tb_id, n, req);
+	rtmsg_fib(RTM_NEWROUTE, key, new_fa, cfg->fc_dst_len, tb->tb_id,
+		  &cfg->fc_nlinfo);
 	return 0;
 
 out_free_new_fa:
@@ -537,30 +535,25 @@ fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 }
 
 
-static int
-fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *n, struct netlink_skb_parms *req)
+static int fn_hash_delete(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct fn_hash *table = (struct fn_hash*)tb->tb_data;
 	struct fib_node *f;
 	struct fib_alias *fa, *fa_to_delete;
-	int z = r->rtm_dst_len;
 	struct fn_zone *fz;
 	u32 key;
-	u8 tos = r->rtm_tos;
 
-	if (z > 32)
+	if (cfg->fc_dst_len > 32)
 		return -EINVAL;
-	if ((fz  = table->fn_zones[z]) == NULL)
+
+	if ((fz  = table->fn_zones[cfg->fc_dst_len]) == NULL)
 		return -ESRCH;
 
 	key = 0;
-	if (rta->rta_dst) {
-		u32 dst;
-		memcpy(&dst, rta->rta_dst, 4);
-		if (dst & ~FZ_MASK(fz))
+	if (cfg->fc_dst) {
+		if (cfg->fc_dst & ~FZ_MASK(fz))
 			return -EINVAL;
-		key = fz_key(dst, fz);
+		key = fz_key(cfg->fc_dst, fz);
 	}
 
 	f = fib_find_node(fz, key);
@@ -568,7 +561,7 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	if (!f)
 		fa = NULL;
 	else
-		fa = fib_find_alias(&f->fn_alias, tos, 0);
+		fa = fib_find_alias(&f->fn_alias, cfg->fc_tos, 0);
 	if (!fa)
 		return -ESRCH;
 
@@ -577,16 +570,16 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	list_for_each_entry_continue(fa, &f->fn_alias, fa_list) {
 		struct fib_info *fi = fa->fa_info;
 
-		if (fa->fa_tos != tos)
+		if (fa->fa_tos != cfg->fc_tos)
 			break;
 
-		if ((!r->rtm_type ||
-		     fa->fa_type == r->rtm_type) &&
-		    (r->rtm_scope == RT_SCOPE_NOWHERE ||
-		     fa->fa_scope == r->rtm_scope) &&
-		    (!r->rtm_protocol ||
-		     fi->fib_protocol == r->rtm_protocol) &&
-		    fib_nh_match(r, n, rta, fi) == 0) {
+		if ((!cfg->fc_type ||
+		     fa->fa_type == cfg->fc_type) &&
+		    (cfg->fc_scope == RT_SCOPE_NOWHERE ||
+		     fa->fa_scope == cfg->fc_scope) &&
+		    (!cfg->fc_protocol ||
+		     fi->fib_protocol == cfg->fc_protocol) &&
+		    fib_nh_match(cfg, fi) == 0) {
 			fa_to_delete = fa;
 			break;
 		}
@@ -596,7 +589,8 @@ fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		int kill_fn;
 
 		fa = fa_to_delete;
-		rtmsg_fib(RTM_DELROUTE, key, fa, z, tb->tb_id, n, req);
+		rtmsg_fib(RTM_DELROUTE, key, fa, cfg->fc_dst_len,
+			  tb->tb_id, &cfg->fc_nlinfo);
 
 		kill_fn = 0;
 		write_lock_bh(&fib_hash_lock);

net/ipv4/fib_lookup.h

@@ -23,19 +23,14 @@ extern int fib_semantic_match(struct list_head *head,
 			      struct fib_result *res, __u32 zone, __u32 mask,
 				int prefixlen);
 extern void fib_release_info(struct fib_info *);
-extern struct fib_info *fib_create_info(const struct rtmsg *r,
-					struct kern_rta *rta,
-					const struct nlmsghdr *,
-					int *err);
-extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *,
-			struct kern_rta *rta, struct fib_info *fi);
+extern struct fib_info *fib_create_info(struct fib_config *cfg);
+extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi);
 extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
 			 u32 tb_id, u8 type, u8 scope, void *dst,
 			 int dst_len, u8 tos, struct fib_info *fi,
 			 unsigned int);
 extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa,
-		      int z, u32 tb_id,
-		      struct nlmsghdr *n, struct netlink_skb_parms *req);
+		      int dst_len, u32 tb_id, struct nl_info *info);
 extern struct fib_alias *fib_find_alias(struct list_head *fah,
 					u8 tos, u32 prio);
 extern int fib_detect_death(struct fib_info *fi, int order,

net/ipv4/fib_trie.c

@@ -1124,17 +1124,14 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen)
 	return fa_head;
 }
 
-static int
-fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-	       struct nlmsghdr *nlhdr, struct netlink_skb_parms *req)
+static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct trie *t = (struct trie *) tb->tb_data;
 	struct fib_alias *fa, *new_fa;
 	struct list_head *fa_head = NULL;
 	struct fib_info *fi;
-	int plen = r->rtm_dst_len;
-	int type = r->rtm_type;
-	u8 tos = r->rtm_tos;
+	int plen = cfg->fc_dst_len;
+	u8 tos = cfg->fc_tos;
 	u32 key, mask;
 	int err;
 	struct leaf *l;
@@ -1142,11 +1139,7 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 	if (plen > 32)
 		return -EINVAL;
 
-	key = 0;
-	if (rta->rta_dst)
-		memcpy(&key, rta->rta_dst, 4);
-
-	key = ntohl(key);
+	key = ntohl(cfg->fc_dst);
 
 	pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen);
 
@@ -1157,10 +1150,11 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	key = key & mask;
 
-	fi = fib_create_info(r, rta, nlhdr, &err);
-
-	if (!fi)
+	fi = fib_create_info(cfg);
+	if (IS_ERR(fi)) {
+		err = PTR_ERR(fi);
 		goto err;
+	}
 
 	l = fib_find_node(t, key);
 	fa = NULL;
@@ -1185,10 +1179,10 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		struct fib_alias *fa_orig;
 
 		err = -EEXIST;
-		if (nlhdr->nlmsg_flags & NLM_F_EXCL)
+		if (cfg->fc_nlflags & NLM_F_EXCL)
 			goto out;
 
-		if (nlhdr->nlmsg_flags & NLM_F_REPLACE) {
+		if (cfg->fc_nlflags & NLM_F_REPLACE) {
 			struct fib_info *fi_drop;
 			u8 state;
 
@@ -1200,8 +1194,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 			fi_drop = fa->fa_info;
 			new_fa->fa_tos = fa->fa_tos;
 			new_fa->fa_info = fi;
-			new_fa->fa_type = type;
-			new_fa->fa_scope = r->rtm_scope;
+			new_fa->fa_type = cfg->fc_type;
+			new_fa->fa_scope = cfg->fc_scope;
 			state = fa->fa_state;
 			new_fa->fa_state &= ~FA_S_ACCESSED;
 
@@ -1224,17 +1218,17 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 				break;
 			if (fa->fa_info->fib_priority != fi->fib_priority)
 				break;
-			if (fa->fa_type == type &&
-			    fa->fa_scope == r->rtm_scope &&
+			if (fa->fa_type == cfg->fc_type &&
+			    fa->fa_scope == cfg->fc_scope &&
 			    fa->fa_info == fi) {
 				goto out;
 			}
 		}
-		if (!(nlhdr->nlmsg_flags & NLM_F_APPEND))
+		if (!(cfg->fc_nlflags & NLM_F_APPEND))
 			fa = fa_orig;
 	}
 	err = -ENOENT;
-	if (!(nlhdr->nlmsg_flags & NLM_F_CREATE))
+	if (!(cfg->fc_nlflags & NLM_F_CREATE))
 		goto out;
 
 	err = -ENOBUFS;
@@ -1244,8 +1238,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 
 	new_fa->fa_info = fi;
 	new_fa->fa_tos = tos;
-	new_fa->fa_type = type;
-	new_fa->fa_scope = r->rtm_scope;
+	new_fa->fa_type = cfg->fc_type;
+	new_fa->fa_scope = cfg->fc_scope;
 	new_fa->fa_state = 0;
 	/*
 	 * Insert new entry to the list.
@@ -1262,7 +1256,8 @@ fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 			  (fa ? &fa->fa_list : fa_head));
 
 	rt_cache_flush(-1);
-	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, nlhdr, req);
+	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id,
+		  &cfg->fc_nlinfo);
 succeeded:
 	return 0;
 
@@ -1548,28 +1543,21 @@ static int trie_leaf_remove(struct trie *t, t_key key)
 	return 1;
 }
 
-static int
-fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
-		struct nlmsghdr *nlhdr, struct netlink_skb_parms *req)
+static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg)
 {
 	struct trie *t = (struct trie *) tb->tb_data;
 	u32 key, mask;
-	int plen = r->rtm_dst_len;
-	u8 tos = r->rtm_tos;
+	int plen = cfg->fc_dst_len;
+	u8 tos = cfg->fc_tos;
 	struct fib_alias *fa, *fa_to_delete;
 	struct list_head *fa_head;
 	struct leaf *l;
 	struct leaf_info *li;
 
-
 	if (plen > 32)
 		return -EINVAL;
 
-	key = 0;
-	if (rta->rta_dst)
-		memcpy(&key, rta->rta_dst, 4);
-
-	key = ntohl(key);
+	key = ntohl(cfg->fc_dst);
 	mask = ntohl(inet_make_mask(plen));
 
 	if (key & ~mask)
@@ -1598,13 +1586,12 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		if (fa->fa_tos != tos)
 			break;
 
-		if ((!r->rtm_type ||
-		     fa->fa_type == r->rtm_type) &&
-		    (r->rtm_scope == RT_SCOPE_NOWHERE ||
-		     fa->fa_scope == r->rtm_scope) &&
-		    (!r->rtm_protocol ||
-		     fi->fib_protocol == r->rtm_protocol) &&
-		    fib_nh_match(r, nlhdr, rta, fi) == 0) {
+		if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) &&
+		    (cfg->fc_scope == RT_SCOPE_NOWHERE ||
+		     fa->fa_scope == cfg->fc_scope) &&
+		    (!cfg->fc_protocol ||
+		     fi->fib_protocol == cfg->fc_protocol) &&
+		    fib_nh_match(cfg, fi) == 0) {
 			fa_to_delete = fa;
 			break;
 		}
@@ -1614,7 +1601,8 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta,
 		return -ESRCH;
 
 	fa = fa_to_delete;
-	rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req);
+	rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id,
+		  &cfg->fc_nlinfo);
 
 	l = fib_find_node(t, key);
 	li = find_leaf_info(l, plen);