[PATCH 1/9] [NETLINK]: Extend netlink messaging interface

[PATCH 1/9] [NETLINK]: Extend netlink messaging interface

[Thomas Graf]

[-- Attachment #1: nl_api --]
[-- Type: text/plain, Size: 9890 bytes --]

Adds:
 nlmsg_get_pos()                 return current position in message
 nlmsg_trim()                    trim part of message
 nla_reserve_nohdr(skb, len)     reserve room for an attribute w/o hdr
 nla_put_nohdr(skb, len, data)   add attribute w/o hdr
 nla_find_nested()               find attribute in nested attributes

Fixes nlmsg_new() to take allocation flags and consider size.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/include/net/netlink.h
===================================================================
--- net-2.6.19.orig/include/net/netlink.h
+++ net-2.6.19/include/net/netlink.h
@@ -35,6 +35,8 @@
  *   nlmsg_put()			add a netlink message to an skb
  *   nlmsg_put_answer()			callback based nlmsg_put()
  *   nlmsg_end()			finanlize netlink message
+ *   nlmsg_get_pos()			return current position in message
+ *   nlmsg_trim()			trim part of message
  *   nlmsg_cancel()			cancel message construction
  *   nlmsg_free()			free a netlink message
  *
@@ -80,8 +82,10 @@
  *   struct nlattr			netlink attribtue header
  *
  * Attribute Construction:
- *   nla_reserve(skb, type, len)	reserve skb tailroom for an attribute
+ *   nla_reserve(skb, type, len)	reserve room for an attribute
+ *   nla_reserve_nohdr(skb, len)	reserve room for an attribute w/o hdr
  *   nla_put(skb, type, len, data)	add attribute to skb
+ *   nla_put_nohdr(skb, len, data)	add attribute w/o hdr
  *
  * Attribute Construction for Basic Types:
  *   nla_put_u8(skb, type, value)	add u8 attribute to skb
@@ -139,6 +143,7 @@
  *   nla_next(nla, remaining)		get next netlink attribute
  *   nla_validate()			validate a stream of attributes
  *   nla_find()				find attribute in stream of attributes
+ *   nla_find_nested()			find attribute in nested attributes
  *   nla_parse()			parse and validate stream of attrs
  *   nla_parse_nested()			parse nested attribuets
  *   nla_for_each_attr()		loop over all attributes
@@ -203,12 +208,18 @@ extern int		nla_memcmp(const struct nlat
 extern int		nla_strcmp(const struct nlattr *nla, const char *str);
 extern struct nlattr *	__nla_reserve(struct sk_buff *skb, int attrtype,
 				      int attrlen);
+extern void *		__nla_reserve_nohdr(struct sk_buff *skb, int attrlen);
 extern struct nlattr *	nla_reserve(struct sk_buff *skb, int attrtype,
 				    int attrlen);
+extern void *		nla_reserve_nohdr(struct sk_buff *skb, int attrlen);
 extern void		__nla_put(struct sk_buff *skb, int attrtype,
 				  int attrlen, const void *data);
+extern void		__nla_put_nohdr(struct sk_buff *skb, int attrlen,
+					const void *data);
 extern int		nla_put(struct sk_buff *skb, int attrtype,
 				int attrlen, const void *data);
+extern int		nla_put_nohdr(struct sk_buff *skb, int attrlen,
+				      const void *data);
 
 /**************************************************************************
  * Netlink Messages
@@ -453,12 +464,13 @@ static inline struct nlmsghdr *nlmsg_put
 /**
  * nlmsg_new - Allocate a new netlink message
  * @size: maximum size of message
+ * @flags: the type of memory to allocate.
  *
  * Use NLMSG_GOODSIZE if size isn't know and you need a good default size.
  */
-static inline struct sk_buff *nlmsg_new(int size)
+static inline struct sk_buff *nlmsg_new(int size, gfp_t flags)
 {
-	return alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
+	return alloc_skb(size, flags);
 }
 
 /**
@@ -480,6 +492,32 @@ static inline int nlmsg_end(struct sk_bu
 }
 
 /**
+ * nlmsg_get_pos - return current position in netlink message
+ * @skb: socket buffer the message is stored in
+ *
+ * Returns a pointer to the current tail of the message.
+ */
+static inline void *nlmsg_get_pos(struct sk_buff *skb)
+{
+	return skb->tail;
+}
+
+/**
+ * nlmsg_trim - Trim message to a mark
+ * @skb: socket buffer the message is stored in
+ * @mark: mark to trim to
+ *
+ * Trims the message to the provided mark. Returns -1.
+ */
+static inline int nlmsg_trim(struct sk_buff *skb, void *mark)
+{
+	if (mark)
+		skb_trim(skb, (unsigned char *) mark - skb->data);
+
+	return -1;
+}
+
+/**
  * nlmsg_cancel - Cancel construction of a netlink message
  * @skb: socket buffer the message is stored in
  * @nlh: netlink message header
@@ -489,9 +527,7 @@ static inline int nlmsg_end(struct sk_bu
  */
 static inline int nlmsg_cancel(struct sk_buff *skb, struct nlmsghdr *nlh)
 {
-	skb_trim(skb, (unsigned char *) nlh - skb->data);
-
-	return -1;
+	return nlmsg_trim(skb, nlh);
 }
 
 /**
@@ -631,6 +667,18 @@ static inline struct nlattr *nla_next(co
 }
 
 /**
+ * nla_find_nested - find attribute in a set of nested attributes
+ * @nla: attribute containing the nested attributes
+ * @attrtype: type of attribute to look for
+ *
+ * Returns the first attribute which matches the specified type.
+ */
+static inline struct nlattr *nla_find_nested(struct nlattr *nla, int attrtype)
+{
+	return nla_find(nla_data(nla), nla_len(nla), attrtype);
+}
+
+/**
  * nla_parse_nested - parse nested attributes
  * @tb: destination array with maxtype+1 elements
  * @maxtype: maximum attribute type to be expected
@@ -862,10 +910,7 @@ static inline int nla_nest_end(struct sk
  */
 static inline int nla_nest_cancel(struct sk_buff *skb, struct nlattr *start)
 {
-	if (start)
-		skb_trim(skb, (unsigned char *) start - skb->data);
-
-	return -1;
+	return nlmsg_trim(skb, start);
 }
 
 /**
@@ -880,4 +925,13 @@ static inline int nla_nest_cancel(struct
 	     nla_ok(pos, rem); \
 	     pos = nla_next(pos, &(rem)))
 
+/**
+ * nla_for_each_nested - iterate over nested attributes
+ * @pos: loop counter, set to current attribute
+ * @nla: attribute containing the nested attributes
+ * @rem: initialized to len, holds bytes currently remaining in stream
+ */
+#define nla_for_each_nested(pos, nla, rem) \
+	nla_for_each_attr(pos, nla_data(nla), nla_len(nla), rem)
+
 #endif
Index: net-2.6.19/net/netlink/attr.c
===================================================================
--- net-2.6.19.orig/net/netlink/attr.c
+++ net-2.6.19/net/netlink/attr.c
@@ -255,6 +255,26 @@ struct nlattr *__nla_reserve(struct sk_b
 }
 
 /**
+ * __nla_reserve_nohdr - reserve room for attribute without header
+ * @skb: socket buffer to reserve room on
+ * @attrlen: length of attribute payload
+ *
+ * Reserves room for attribute payload without a header.
+ *
+ * The caller is responsible to ensure that the skb provides enough
+ * tailroom for the payload.
+ */
+void *__nla_reserve_nohdr(struct sk_buff *skb, int attrlen)
+{
+	void *start;
+
+	start = skb_put(skb, NLA_ALIGN(attrlen));
+	memset(start, 0, NLA_ALIGN(attrlen));
+
+	return start;
+}
+
+/**
  * nla_reserve - reserve room for attribute on the skb
  * @skb: socket buffer to reserve room on
  * @attrtype: attribute type
@@ -275,6 +295,24 @@ struct nlattr *nla_reserve(struct sk_buf
 }
 
 /**
+ * nla_reserve - reserve room for attribute without header
+ * @skb: socket buffer to reserve room on
+ * @len: length of attribute payload
+ *
+ * Reserves room for attribute payload without a header.
+ *
+ * Returns NULL if the tailroom of the skb is insufficient to store
+ * the attribute payload.
+ */
+void *nla_reserve_nohdr(struct sk_buff *skb, int attrlen)
+{
+	if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen)))
+		return NULL;
+
+	return __nla_reserve_nohdr(skb, attrlen);
+}
+
+/**
  * __nla_put - Add a netlink attribute to a socket buffer
  * @skb: socket buffer to add attribute to
  * @attrtype: attribute type
@@ -293,6 +331,22 @@ void __nla_put(struct sk_buff *skb, int 
 	memcpy(nla_data(nla), data, attrlen);
 }
 
+/**
+ * __nla_put_nohdr - Add a netlink attribute without header
+ * @skb: socket buffer to add attribute to
+ * @attrlen: length of attribute payload
+ * @data: head of attribute payload
+ *
+ * The caller is responsible to ensure that the skb provides enough
+ * tailroom for the attribute payload.
+ */
+void __nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data)
+{
+	void *start;
+
+	start = __nla_reserve_nohdr(skb, attrlen);
+	memcpy(start, data, attrlen);
+}
 
 /**
  * nla_put - Add a netlink attribute to a socket buffer
@@ -313,15 +367,36 @@ int nla_put(struct sk_buff *skb, int att
 	return 0;
 }
 
+/**
+ * nla_put_nohdr - Add a netlink attribute without header
+ * @skb: socket buffer to add attribute to
+ * @attrlen: length of attribute payload
+ * @data: head of attribute payload
+ *
+ * Returns -1 if the tailroom of the skb is insufficient to store
+ * the attribute payload.
+ */
+int nla_put_nohdr(struct sk_buff *skb, int attrlen, const void *data)
+{
+	if (unlikely(skb_tailroom(skb) < NLA_ALIGN(attrlen)))
+		return -1;
+
+	__nla_put_nohdr(skb, attrlen, data);
+	return 0;
+}
 
 EXPORT_SYMBOL(nla_validate);
 EXPORT_SYMBOL(nla_parse);
 EXPORT_SYMBOL(nla_find);
 EXPORT_SYMBOL(nla_strlcpy);
 EXPORT_SYMBOL(__nla_reserve);
+EXPORT_SYMBOL(__nla_reserve_nohdr);
 EXPORT_SYMBOL(nla_reserve);
+EXPORT_SYMBOL(nla_reserve_nohdr);
 EXPORT_SYMBOL(__nla_put);
+EXPORT_SYMBOL(__nla_put_nohdr);
 EXPORT_SYMBOL(nla_put);
+EXPORT_SYMBOL(nla_put_nohdr);
 EXPORT_SYMBOL(nla_memcpy);
 EXPORT_SYMBOL(nla_memcmp);
 EXPORT_SYMBOL(nla_strcmp);
Index: net-2.6.19/kernel/taskstats.c
===================================================================
--- net-2.6.19.orig/kernel/taskstats.c
+++ net-2.6.19/kernel/taskstats.c
@@ -75,7 +75,7 @@ static int prepare_reply(struct genl_inf
 	/*
 	 * If new attributes are added, please revisit this allocation
 	 */
-	skb = nlmsg_new(size);
+	skb = nlmsg_new(size, GFP_KERNEL);
 	if (!skb)
 		return -ENOMEM;
 
Index: net-2.6.19/net/netlink/genetlink.c
===================================================================
--- net-2.6.19.orig/net/netlink/genetlink.c
+++ net-2.6.19/net/netlink/genetlink.c
@@ -440,7 +440,7 @@ static struct sk_buff *ctrl_build_msg(st
 	struct sk_buff *skb;
 	int err;
 
-	skb = nlmsg_new(NLMSG_GOODSIZE);
+	skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
 	if (skb == NULL)
 		return ERR_PTR(-ENOBUFS);
 

[PATCH 2/9] [NETLINK]: Convert core netlink handling to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_af --]
[-- Type: text/plain, Size: 3473 bytes --]

Fixes a theoretical memory and locking leak when the size of
the netlink header would exceed the skb tailroom.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/netlink/af_netlink.c
===================================================================
--- net-2.6.19.orig/net/netlink/af_netlink.c
+++ net-2.6.19/net/netlink/af_netlink.c
@@ -1147,7 +1147,7 @@ static int netlink_sendmsg(struct kiocb 
 	if (len > sk->sk_sndbuf - 32)
 		goto out;
 	err = -ENOBUFS;
-	skb = alloc_skb(len, GFP_KERNEL);
+	skb = nlmsg_new(len, GFP_KERNEL);
 	if (skb==NULL)
 		goto out;
 
@@ -1342,19 +1342,18 @@ static int netlink_dump(struct sock *sk)
 	struct netlink_callback *cb;
 	struct sk_buff *skb;
 	struct nlmsghdr *nlh;
-	int len;
+	int len, err = -ENOBUFS;
 	
 	skb = sock_rmalloc(sk, NLMSG_GOODSIZE, 0, GFP_KERNEL);
 	if (!skb)
-		return -ENOBUFS;
+		goto errout;
 
 	spin_lock(&nlk->cb_lock);
 
 	cb = nlk->cb;
 	if (cb == NULL) {
-		spin_unlock(&nlk->cb_lock);
-		kfree_skb(skb);
-		return -EINVAL;
+		err = -EINVAL;
+		goto errout_skb;
 	}
 
 	len = cb->dump(skb, cb);
@@ -1366,8 +1365,12 @@ static int netlink_dump(struct sock *sk)
 		return 0;
 	}
 
-	nlh = NLMSG_NEW_ANSWER(skb, cb, NLMSG_DONE, sizeof(len), NLM_F_MULTI);
-	memcpy(NLMSG_DATA(nlh), &len, sizeof(len));
+	nlh = nlmsg_put_answer(skb, cb, NLMSG_DONE, sizeof(len), NLM_F_MULTI);
+	if (!nlh)
+		goto errout_skb;
+
+	memcpy(nlmsg_data(nlh), &len, sizeof(len));
+
 	skb_queue_tail(&sk->sk_receive_queue, skb);
 	sk->sk_data_ready(sk, skb->len);
 
@@ -1379,8 +1382,11 @@ static int netlink_dump(struct sock *sk)
 	netlink_destroy_callback(cb);
 	return 0;
 
-nlmsg_failure:
-	return -ENOBUFS;
+errout_skb:
+	spin_unlock(&nlk->cb_lock);
+	kfree_skb(skb);
+errout:
+	return err;
 }
 
 int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
@@ -1432,11 +1438,11 @@ void netlink_ack(struct sk_buff *in_skb,
 	int size;
 
 	if (err == 0)
-		size = NLMSG_SPACE(sizeof(struct nlmsgerr));
+		size = nlmsg_total_size(sizeof(*errmsg));
 	else
-		size = NLMSG_SPACE(4 + NLMSG_ALIGN(nlh->nlmsg_len));
+		size = nlmsg_total_size(sizeof(*errmsg) + nlmsg_len(nlh));
 
-	skb = alloc_skb(size, GFP_KERNEL);
+	skb = nlmsg_new(size, GFP_KERNEL);
 	if (!skb) {
 		struct sock *sk;
 
@@ -1452,16 +1458,15 @@ void netlink_ack(struct sk_buff *in_skb,
 
 	rep = __nlmsg_put(skb, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq,
 			  NLMSG_ERROR, sizeof(struct nlmsgerr), 0);
-	errmsg = NLMSG_DATA(rep);
+	errmsg = nlmsg_data(rep);
 	errmsg->error = err;
-	memcpy(&errmsg->msg, nlh, err ? nlh->nlmsg_len : sizeof(struct nlmsghdr));
+	memcpy(&errmsg->msg, nlh, err ? nlh->nlmsg_len : sizeof(*nlh));
 	netlink_unicast(in_skb->sk, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
 }
 
 static int netlink_rcv_skb(struct sk_buff *skb, int (*cb)(struct sk_buff *,
 						     struct nlmsghdr *, int *))
 {
-	unsigned int total_len;
 	struct nlmsghdr *nlh;
 	int err;
 
@@ -1471,8 +1476,6 @@ static int netlink_rcv_skb(struct sk_buf
 		if (nlh->nlmsg_len < NLMSG_HDRLEN || skb->len < nlh->nlmsg_len)
 			return 0;
 
-		total_len = min(NLMSG_ALIGN(nlh->nlmsg_len), skb->len);
-
 		if (cb(skb, nlh, &err) < 0) {
 			/* Not an error, but we have to interrupt processing
 			 * here. Note: that in this case we do not pull
@@ -1484,7 +1487,7 @@ static int netlink_rcv_skb(struct sk_buf
 		} else if (nlh->nlmsg_flags & NLM_F_ACK)
 			netlink_ack(skb, nlh, 0);
 
-		skb_pull(skb, total_len);
+		netlink_queue_skip(nlh, skb);
 	}
 
 	return 0;

[PATCH 3/9] [IPV4]: Convert address addition to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_ipv4_newaddr --]
[-- Type: text/plain, Size: 4143 bytes --]

Adds rtm_to_ifaddr() transforming a netlink message to a
struct in_ifaddr. Fixes various unvalidated netlink attributes
causing memory corruptions when left empty by userspace
applications.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/ipv4/devinet.c
===================================================================
--- net-2.6.19.orig/net/ipv4/devinet.c
+++ net-2.6.19/net/ipv4/devinet.c
@@ -62,6 +62,7 @@
 #include <net/ip.h>
 #include <net/route.h>
 #include <net/ip_fib.h>
+#include <net/netlink.h>
 
 struct ipv4_devconf ipv4_devconf = {
 	.accept_redirects = 1,
@@ -78,6 +79,14 @@ static struct ipv4_devconf ipv4_devconf_
 	.accept_source_route = 1,
 };
 
+static struct nla_policy ifa_ipv4_policy[IFA_MAX+1] __read_mostly = {
+	[IFA_LOCAL]     	= { .type = NLA_U32 },
+	[IFA_ADDRESS]   	= { .type = NLA_U32 },
+	[IFA_BROADCAST] 	= { .type = NLA_U32 },
+	[IFA_ANYCAST]   	= { .type = NLA_U32 },
+	[IFA_LABEL]     	= { .type = NLA_STRING },
+};
+
 static void rtmsg_ifa(int event, struct in_ifaddr *);
 
 static BLOCKING_NOTIFIER_HEAD(inetaddr_chain);
@@ -451,57 +460,90 @@ out:
 	return -EADDRNOTAVAIL;
 }
 
-static int inet_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+static struct in_ifaddr *rtm_to_ifaddr(struct nlmsghdr *nlh)
 {
-	struct rtattr **rta = arg;
+	struct nlattr *tb[IFA_MAX+1];
+	struct in_ifaddr *ifa;
+	struct ifaddrmsg *ifm;
 	struct net_device *dev;
 	struct in_device *in_dev;
-	struct ifaddrmsg *ifm = NLMSG_DATA(nlh);
-	struct in_ifaddr *ifa;
-	int rc = -EINVAL;
+	int err = -EINVAL;
 
-	ASSERT_RTNL();
+	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv4_policy);
+	if (err < 0)
+		goto errout;
 
-	if (ifm->ifa_prefixlen > 32 || !rta[IFA_LOCAL - 1])
-		goto out;
+	ifm = nlmsg_data(nlh);
+	if (ifm->ifa_prefixlen > 32 || tb[IFA_LOCAL] == NULL)
+		goto errout;
 
-	rc = -ENODEV;
-	if ((dev = __dev_get_by_index(ifm->ifa_index)) == NULL)
-		goto out;
+	dev = __dev_get_by_index(ifm->ifa_index);
+	if (dev == NULL) {
+		err = -ENODEV;
+		goto errout;
+	}
 
-	rc = -ENOBUFS;
-	if ((in_dev = __in_dev_get_rtnl(dev)) == NULL) {
+	in_dev = __in_dev_get_rtnl(dev);
+	if (in_dev == NULL) {
 		in_dev = inetdev_init(dev);
-		if (!in_dev)
-			goto out;
+		if (in_dev == NULL) {
+			err = -ENOBUFS;
+			goto errout;
+		}
 	}
 
-	if ((ifa = inet_alloc_ifa()) == NULL)
-		goto out;
+	ifa = inet_alloc_ifa();
+	if (ifa == NULL) {
+		/*
+		 * A potential indev allocation can be left alive, it stays
+		 * assigned to its device and is destroy with it.
+		 */
+		err = -ENOBUFS;
+		goto errout;
+	}
+
+	in_dev_hold(in_dev);
+
+	if (tb[IFA_ADDRESS] == NULL)
+		tb[IFA_ADDRESS] = tb[IFA_LOCAL];
 
-	if (!rta[IFA_ADDRESS - 1])
-		rta[IFA_ADDRESS - 1] = rta[IFA_LOCAL - 1];
-	memcpy(&ifa->ifa_local, RTA_DATA(rta[IFA_LOCAL - 1]), 4);
-	memcpy(&ifa->ifa_address, RTA_DATA(rta[IFA_ADDRESS - 1]), 4);
 	ifa->ifa_prefixlen = ifm->ifa_prefixlen;
 	ifa->ifa_mask = inet_make_mask(ifm->ifa_prefixlen);
-	if (rta[IFA_BROADCAST - 1])
-		memcpy(&ifa->ifa_broadcast,
-		       RTA_DATA(rta[IFA_BROADCAST - 1]), 4);
-	if (rta[IFA_ANYCAST - 1])
-		memcpy(&ifa->ifa_anycast, RTA_DATA(rta[IFA_ANYCAST - 1]), 4);
 	ifa->ifa_flags = ifm->ifa_flags;
 	ifa->ifa_scope = ifm->ifa_scope;
-	in_dev_hold(in_dev);
-	ifa->ifa_dev   = in_dev;
-	if (rta[IFA_LABEL - 1])
-		rtattr_strlcpy(ifa->ifa_label, rta[IFA_LABEL - 1], IFNAMSIZ);
+	ifa->ifa_dev = in_dev;
+
+	ifa->ifa_local = nla_get_u32(tb[IFA_LOCAL]);
+	ifa->ifa_address = nla_get_u32(tb[IFA_ADDRESS]);
+
+	if (tb[IFA_BROADCAST])
+		ifa->ifa_broadcast = nla_get_u32(tb[IFA_BROADCAST]);
+
+	if (tb[IFA_ANYCAST])
+		ifa->ifa_anycast = nla_get_u32(tb[IFA_ANYCAST]);
+
+	if (tb[IFA_LABEL])
+		nla_strlcpy(ifa->ifa_label, tb[IFA_LABEL], IFNAMSIZ);
 	else
 		memcpy(ifa->ifa_label, dev->name, IFNAMSIZ);
 
-	rc = inet_insert_ifa(ifa);
-out:
-	return rc;
+	return ifa;
+
+errout:
+	return ERR_PTR(err);
+}
+
+static int inet_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+{
+	struct in_ifaddr *ifa;
+
+	ASSERT_RTNL();
+
+	ifa = rtm_to_ifaddr(nlh);
+	if (IS_ERR(ifa))
+		return PTR_ERR(ifa);
+
+	return inet_insert_ifa(ifa);
 }
 
 /*

[PATCH 4/9] [IPV4]: Convert address deletion to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_ipv4_deladdr --]
[-- Type: text/plain, Size: 2033 bytes --]

Fixes various unvalidated netlink attributes causing
memory corruptions when left empty by userspace.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/ipv4/devinet.c
===================================================================
--- net-2.6.19.orig/net/ipv4/devinet.c
+++ net-2.6.19/net/ipv4/devinet.c
@@ -430,34 +430,48 @@ struct in_ifaddr *inet_ifa_byprefix(stru
 
 static int inet_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct rtattr **rta = arg;
+	struct nlattr *tb[IFA_MAX+1];
 	struct in_device *in_dev;
-	struct ifaddrmsg *ifm = NLMSG_DATA(nlh);
+	struct ifaddrmsg *ifm;
 	struct in_ifaddr *ifa, **ifap;
+	int err = -EINVAL;
 
 	ASSERT_RTNL();
 
-	if ((in_dev = inetdev_by_index(ifm->ifa_index)) == NULL)
-		goto out;
+	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFA_MAX, ifa_ipv4_policy);
+	if (err < 0)
+		goto errout;
+
+	ifm = nlmsg_data(nlh);
+	in_dev = inetdev_by_index(ifm->ifa_index);
+	if (in_dev == NULL) {
+		err = -ENODEV;
+		goto errout;
+	}
+
 	__in_dev_put(in_dev);
 
 	for (ifap = &in_dev->ifa_list; (ifa = *ifap) != NULL;
 	     ifap = &ifa->ifa_next) {
-		if ((rta[IFA_LOCAL - 1] &&
-		     memcmp(RTA_DATA(rta[IFA_LOCAL - 1]),
-			    &ifa->ifa_local, 4)) ||
-		    (rta[IFA_LABEL - 1] &&
-		     rtattr_strcmp(rta[IFA_LABEL - 1], ifa->ifa_label)) ||
-		    (rta[IFA_ADDRESS - 1] &&
-		     (ifm->ifa_prefixlen != ifa->ifa_prefixlen ||
-		      !inet_ifa_match(*(u32*)RTA_DATA(rta[IFA_ADDRESS - 1]),
-			      	      ifa))))
+		if (tb[IFA_LOCAL] &&
+		    ifa->ifa_local != nla_get_u32(tb[IFA_LOCAL]))
 			continue;
+
+		if (tb[IFA_LABEL] && nla_strcmp(tb[IFA_LABEL], ifa->ifa_label))
+			continue;
+
+		if (tb[IFA_ADDRESS] &&
+		    (ifm->ifa_prefixlen != ifa->ifa_prefixlen ||
+		    !inet_ifa_match(nla_get_u32(tb[IFA_ADDRESS]), ifa)))
+			continue;
+
 		inet_del_ifa(in_dev, ifap, 1);
 		return 0;
 	}
-out:
-	return -EADDRNOTAVAIL;
+
+	err = -EADDRNOTAVAIL;
+errout:
+	return err;
 }
 
 static struct in_ifaddr *rtm_to_ifaddr(struct nlmsghdr *nlh)

[PATCH 5/9] [IPV4]: Convert address dumping to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_ipv4_dumpaddr --]
[-- Type: text/plain, Size: 2422 bytes --]

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/ipv4/devinet.c
===================================================================
--- net-2.6.19.orig/net/ipv4/devinet.c
+++ net-2.6.19/net/ipv4/devinet.c
@@ -1112,32 +1112,37 @@ static int inet_fill_ifaddr(struct sk_bu
 {
 	struct ifaddrmsg *ifm;
 	struct nlmsghdr  *nlh;
-	unsigned char	 *b = skb->tail;
 
-	nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*ifm), flags);
-	ifm = NLMSG_DATA(nlh);
+	nlh = nlmsg_put(skb, pid, seq, event, sizeof(*ifm), flags);
+	if (nlh == NULL)
+		return -ENOBUFS;
+
+	ifm = nlmsg_data(nlh);
 	ifm->ifa_family = AF_INET;
 	ifm->ifa_prefixlen = ifa->ifa_prefixlen;
 	ifm->ifa_flags = ifa->ifa_flags|IFA_F_PERMANENT;
 	ifm->ifa_scope = ifa->ifa_scope;
 	ifm->ifa_index = ifa->ifa_dev->dev->ifindex;
+
 	if (ifa->ifa_address)
-		RTA_PUT(skb, IFA_ADDRESS, 4, &ifa->ifa_address);
+		NLA_PUT_U32(skb, IFA_ADDRESS, ifa->ifa_address);
+
 	if (ifa->ifa_local)
-		RTA_PUT(skb, IFA_LOCAL, 4, &ifa->ifa_local);
+		NLA_PUT_U32(skb, IFA_LOCAL, ifa->ifa_local);
+
 	if (ifa->ifa_broadcast)
-		RTA_PUT(skb, IFA_BROADCAST, 4, &ifa->ifa_broadcast);
+		NLA_PUT_U32(skb, IFA_BROADCAST, ifa->ifa_broadcast);
+
 	if (ifa->ifa_anycast)
-		RTA_PUT(skb, IFA_ANYCAST, 4, &ifa->ifa_anycast);
+		NLA_PUT_U32(skb, IFA_ANYCAST, ifa->ifa_anycast);
+
 	if (ifa->ifa_label[0])
-		RTA_PUT(skb, IFA_LABEL, IFNAMSIZ, &ifa->ifa_label);
-	nlh->nlmsg_len = skb->tail - b;
-	return skb->len;
+		NLA_PUT_STRING(skb, IFA_LABEL, ifa->ifa_label);
+
+	return nlmsg_end(skb, nlh);
 
-nlmsg_failure:
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
+nla_put_failure:
+	return nlmsg_cancel(skb, nlh);
 }
 
 static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
@@ -1185,17 +1190,16 @@ done:
 
 static void rtmsg_ifa(int event, struct in_ifaddr* ifa)
 {
-	int size = NLMSG_SPACE(sizeof(struct ifaddrmsg) + 128);
-	struct sk_buff *skb = alloc_skb(size, GFP_KERNEL);
+	struct sk_buff *skb;
 
-	if (!skb)
+	skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
+	if (skb == NULL)
 		netlink_set_err(rtnl, 0, RTNLGRP_IPV4_IFADDR, ENOBUFS);
 	else if (inet_fill_ifaddr(skb, ifa, 0, 0, event, 0) < 0) {
 		kfree_skb(skb);
 		netlink_set_err(rtnl, 0, RTNLGRP_IPV4_IFADDR, EINVAL);
-	} else {
+	} else
 		netlink_broadcast(rtnl, skb, 0, RTNLGRP_IPV4_IFADDR, GFP_KERNEL);
-	}
 }
 
 static struct rtnetlink_link inet_rtnetlink_table[RTM_NR_MSGTYPES] = {

[PATCH 6/9] [IPv4]: Move interface address bits to linux/if_addr.h

[Thomas Graf]

[-- Attachment #1: nl_if_addr --]
[-- Type: text/plain, Size: 5364 bytes --]

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/include/linux/if_addr.h
===================================================================
--- /dev/null
+++ net-2.6.19/include/linux/if_addr.h
@@ -0,0 +1,53 @@
+#ifndef __LINUX_IF_ADDR_H
+#define __LINUX_IF_ADDR_H
+
+#include <linux/netlink.h>
+
+struct ifaddrmsg
+{
+	__u8		ifa_family;
+	__u8		ifa_prefixlen;	/* The prefix length		*/
+	__u8		ifa_flags;	/* Flags			*/
+	__u8		ifa_scope;	/* Address scope		*/
+	__u32		ifa_index;	/* Link index			*/
+};
+
+/*
+ * Important comment:
+ * IFA_ADDRESS is prefix address, rather than local interface address.
+ * It makes no difference for normally configured broadcast interfaces,
+ * but for point-to-point IFA_ADDRESS is DESTINATION address,
+ * local address is supplied in IFA_LOCAL attribute.
+ */
+enum
+{
+	IFA_UNSPEC,
+	IFA_ADDRESS,
+	IFA_LOCAL,
+	IFA_LABEL,
+	IFA_BROADCAST,
+	IFA_ANYCAST,
+	IFA_CACHEINFO,
+	IFA_MULTICAST,
+	__IFA_MAX,
+};
+
+#define IFA_MAX (__IFA_MAX - 1)
+
+/* ifa_flags */
+#define IFA_F_SECONDARY		0x01
+#define IFA_F_TEMPORARY		IFA_F_SECONDARY
+
+#define IFA_F_DEPRECATED	0x20
+#define IFA_F_TENTATIVE		0x40
+#define IFA_F_PERMANENT		0x80
+
+struct ifa_cacheinfo
+{
+	__u32	ifa_prefered;
+	__u32	ifa_valid;
+	__u32	cstamp; /* created timestamp, hundredths of seconds */
+	__u32	tstamp; /* updated timestamp, hundredths of seconds */
+};
+
+#endif
Index: net-2.6.19/net/decnet/dn_dev.c
===================================================================
--- net-2.6.19.orig/net/decnet/dn_dev.c
+++ net-2.6.19/net/decnet/dn_dev.c
@@ -34,6 +34,7 @@
 #include <linux/seq_file.h>
 #include <linux/timer.h>
 #include <linux/string.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/if_ether.h>
 #include <linux/skbuff.h>
Index: net-2.6.19/net/ipv4/devinet.c
===================================================================
--- net-2.6.19.orig/net/ipv4/devinet.c
+++ net-2.6.19/net/ipv4/devinet.c
@@ -43,6 +43,7 @@
 #include <linux/in.h>
 #include <linux/errno.h>
 #include <linux/interrupt.h>
+#include <linux/if_addr.h>
 #include <linux/if_ether.h>
 #include <linux/inet.h>
 #include <linux/netdevice.h>
Index: net-2.6.19/net/ipv4/fib_frontend.c
===================================================================
--- net-2.6.19.orig/net/ipv4/fib_frontend.c
+++ net-2.6.19/net/ipv4/fib_frontend.c
@@ -32,6 +32,7 @@
 #include <linux/inet.h>
 #include <linux/inetdevice.h>
 #include <linux/netdevice.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/skbuff.h>
 #include <linux/netlink.h>
Index: net-2.6.19/net/ipv6/addrconf.c
===================================================================
--- net-2.6.19.orig/net/ipv6/addrconf.c
+++ net-2.6.19/net/ipv6/addrconf.c
@@ -48,6 +48,7 @@
 #include <linux/net.h>
 #include <linux/in6.h>
 #include <linux/netdevice.h>
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/if_arcnet.h>
 #include <linux/if_infiniband.h>
Index: net-2.6.19/net/ipv6/ndisc.c
===================================================================
--- net-2.6.19.orig/net/ipv6/ndisc.c
+++ net-2.6.19/net/ipv6/ndisc.c
@@ -62,6 +62,7 @@
 #include <linux/sysctl.h>
 #endif
 
+#include <linux/if_addr.h>
 #include <linux/if_arp.h>
 #include <linux/ipv6.h>
 #include <linux/icmpv6.h>
Index: net-2.6.19/net/core/rtnetlink.c
===================================================================
--- net-2.6.19.orig/net/core/rtnetlink.c
+++ net-2.6.19/net/core/rtnetlink.c
@@ -35,6 +35,7 @@
 #include <linux/init.h>
 #include <linux/security.h>
 #include <linux/mutex.h>
+#include <linux/if_addr.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
Index: net-2.6.19/include/linux/rtnetlink.h
===================================================================
--- net-2.6.19.orig/include/linux/rtnetlink.h
+++ net-2.6.19/include/linux/rtnetlink.h
@@ -384,62 +384,6 @@ struct rta_session
 };
 
 
-/*********************************************************
- *		Interface address.
- ****/
-
-struct ifaddrmsg
-{
-	unsigned char	ifa_family;
-	unsigned char	ifa_prefixlen;	/* The prefix length		*/
-	unsigned char	ifa_flags;	/* Flags			*/
-	unsigned char	ifa_scope;	/* See above			*/
-	int		ifa_index;	/* Link index			*/
-};
-
-enum
-{
-	IFA_UNSPEC,
-	IFA_ADDRESS,
-	IFA_LOCAL,
-	IFA_LABEL,
-	IFA_BROADCAST,
-	IFA_ANYCAST,
-	IFA_CACHEINFO,
-	IFA_MULTICAST,
-	__IFA_MAX
-};
-
-#define IFA_MAX (__IFA_MAX - 1)
-
-/* ifa_flags */
-
-#define IFA_F_SECONDARY		0x01
-#define IFA_F_TEMPORARY		IFA_F_SECONDARY
-
-#define IFA_F_DEPRECATED	0x20
-#define IFA_F_TENTATIVE		0x40
-#define IFA_F_PERMANENT		0x80
-
-struct ifa_cacheinfo
-{
-	__u32	ifa_prefered;
-	__u32	ifa_valid;
-	__u32	cstamp; /* created timestamp, hundredths of seconds */
-	__u32	tstamp; /* updated timestamp, hundredths of seconds */
-};
-
-
-#define IFA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifaddrmsg))))
-#define IFA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifaddrmsg))
-
-/*
-   Important comment:
-   IFA_ADDRESS is prefix address, rather than local interface address.
-   It makes no difference for normally configured broadcast interfaces,
-   but for point-to-point IFA_ADDRESS is DESTINATION address,
-   local address is supplied in IFA_LOCAL attribute.
- */
 
 /**************************************************************
  *		Neighbour discovery.

[PATCH 7/9] [NET]: Convert link modification to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_link --]
[-- Type: text/plain, Size: 7908 bytes --]

Transforms do_setlink() into rtnl_setlink() using the new
netlink api. A warning message printed to the console is
added in the event that a change request fails while part
of the change request has been comitted already. The ioctl()
based nature of net devices makes it almost impossible to
move on to atomic netlink operations without obsoleting
some of the functionality.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/core/rtnetlink.c
===================================================================
--- net-2.6.19.orig/net/core/rtnetlink.c
+++ net-2.6.19/net/core/rtnetlink.c
@@ -336,52 +336,69 @@ static int rtnetlink_dump_ifinfo(struct 
 	return skb->len;
 }
 
-static int do_setlink(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+static struct nla_policy ifla_policy[IFLA_MAX+1] __read_mostly = {
+	[IFLA_IFNAME]		= { .type = NLA_STRING },
+	[IFLA_MAP]		= { .minlen = sizeof(struct rtnl_link_ifmap) },
+	[IFLA_MTU]		= { .type = NLA_U32 },
+	[IFLA_TXQLEN]		= { .type = NLA_U32 },
+	[IFLA_WEIGHT]		= { .type = NLA_U32 },
+	[IFLA_OPERSTATE]	= { .type = NLA_U8 },
+	[IFLA_LINKMODE]		= { .type = NLA_U8 },
+};
+
+static int rtnl_setlink(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
 {
-	struct ifinfomsg  *ifm = NLMSG_DATA(nlh);
-	struct rtattr    **ida = arg;
+	struct ifinfomsg *ifm;
 	struct net_device *dev;
-	int err, send_addr_notify = 0;
+	int err, send_addr_notify = 0, modified = 0;
+	struct nlattr *tb[IFLA_MAX+1];
+	char ifname[IFNAMSIZ];
+
+	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFLA_MAX, ifla_policy);
+	if (err < 0)
+		goto errout;
+
+	if (tb[IFLA_IFNAME] &&
+	    nla_strlcpy(ifname, tb[IFLA_IFNAME], IFNAMSIZ) >= IFNAMSIZ)
+		return -EINVAL;
 
+	err = -EINVAL;
+	ifm = nlmsg_data(nlh);
 	if (ifm->ifi_index >= 0)
 		dev = dev_get_by_index(ifm->ifi_index);
-	else if (ida[IFLA_IFNAME - 1]) {
-		char ifname[IFNAMSIZ];
-
-		if (rtattr_strlcpy(ifname, ida[IFLA_IFNAME - 1],
-		                   IFNAMSIZ) >= IFNAMSIZ)
-			return -EINVAL;
+	else if (tb[IFLA_IFNAME])
 		dev = dev_get_by_name(ifname);
-	} else
-		return -EINVAL;
+	else
+		goto errout;
 
-	if (!dev)
-		return -ENODEV;
+	if (dev == NULL) {
+		err = -ENODEV;
+		goto errout;
+	}
 
-	err = -EINVAL;
+	if (tb[IFLA_ADDRESS] &&
+	    nla_len(tb[IFLA_ADDRESS]) < dev->addr_len)
+		goto errout_dev;
 
-	if (ifm->ifi_flags)
-		dev_change_flags(dev, ifm->ifi_flags);
+	if (tb[IFLA_BROADCAST] &&
+	    nla_len(tb[IFLA_BROADCAST]) < dev->addr_len)
+		goto errout_dev;
 
-	if (ida[IFLA_MAP - 1]) {
+	if (tb[IFLA_MAP]) {
 		struct rtnl_link_ifmap *u_map;
 		struct ifmap k_map;
 
 		if (!dev->set_config) {
 			err = -EOPNOTSUPP;
-			goto out;
+			goto errout_dev;
 		}
 
 		if (!netif_device_present(dev)) {
 			err = -ENODEV;
-			goto out;
+			goto errout_dev;
 		}
-		
-		if (ida[IFLA_MAP - 1]->rta_len != RTA_LENGTH(sizeof(*u_map)))
-			goto out;
-
-		u_map = RTA_DATA(ida[IFLA_MAP - 1]);
 
+		u_map = nla_data(tb[IFLA_MAP]);
 		k_map.mem_start = (unsigned long) u_map->mem_start;
 		k_map.mem_end = (unsigned long) u_map->mem_end;
 		k_map.base_addr = (unsigned short) u_map->base_addr;
@@ -390,106 +407,97 @@ static int do_setlink(struct sk_buff *sk
 		k_map.port = (unsigned char) u_map->port;
 
 		err = dev->set_config(dev, &k_map);
+		if (err < 0)
+			goto errout_dev;
 
-		if (err)
-			goto out;
+		modified = 1;
 	}
 
-	if (ida[IFLA_ADDRESS - 1]) {
+	if (tb[IFLA_ADDRESS]) {
 		if (!dev->set_mac_address) {
 			err = -EOPNOTSUPP;
-			goto out;
+			goto errout_dev;
 		}
+
 		if (!netif_device_present(dev)) {
 			err = -ENODEV;
-			goto out;
+			goto errout_dev;
 		}
-		if (ida[IFLA_ADDRESS - 1]->rta_len != RTA_LENGTH(dev->addr_len))
-			goto out;
 
-		err = dev->set_mac_address(dev, RTA_DATA(ida[IFLA_ADDRESS - 1]));
-		if (err)
-			goto out;
+		err = dev->set_mac_address(dev, nla_data(tb[IFLA_ADDRESS]));
+		if (err < 0)
+			goto errout_dev;
 		send_addr_notify = 1;
+		modified = 1;
 	}
 
-	if (ida[IFLA_BROADCAST - 1]) {
-		if (ida[IFLA_BROADCAST - 1]->rta_len != RTA_LENGTH(dev->addr_len))
-			goto out;
-		memcpy(dev->broadcast, RTA_DATA(ida[IFLA_BROADCAST - 1]),
-		       dev->addr_len);
-		send_addr_notify = 1;
+	if (tb[IFLA_MTU]) {
+		err = dev_set_mtu(dev, nla_get_u32(tb[IFLA_MTU]));
+		if (err < 0)
+			goto errout_dev;
+		modified = 1;
+	}
+
+	/*
+	 * Interface selected by interface index but interface
+	 * name provided implies that a name change has been
+	 * requested.
+	 */
+	if (ifm->ifi_index >= 0 && ifname[0]) {
+		err = dev_change_name(dev, ifname);
+		if (err < 0)
+			goto errout_dev;
+		modified = 1;
 	}
 
-	if (ida[IFLA_MTU - 1]) {
-		if (ida[IFLA_MTU - 1]->rta_len != RTA_LENGTH(sizeof(u32)))
-			goto out;
-		err = dev_set_mtu(dev, *((u32 *) RTA_DATA(ida[IFLA_MTU - 1])));
-
-		if (err)
-			goto out;
-
+#ifdef CONFIG_NET_WIRELESS_RTNETLINK
+	if (tb[IFLA_WIRELESS]) {
+		/* Call Wireless Extensions.
+		 * Various stuff checked in there... */
+		err = wireless_rtnetlink_set(dev, nla_data(tb[IFLA_WIRELESS]),
+					     nla_len(tb[IFLA_WIRELESS]));
+		if (err < 0)
+			goto errout_dev;
 	}
+#endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
 
-	if (ida[IFLA_TXQLEN - 1]) {
-		if (ida[IFLA_TXQLEN - 1]->rta_len != RTA_LENGTH(sizeof(u32)))
-			goto out;
-
-		dev->tx_queue_len = *((u32 *) RTA_DATA(ida[IFLA_TXQLEN - 1]));
+	if (tb[IFLA_BROADCAST]) {
+		nla_memcpy(dev->broadcast, tb[IFLA_BROADCAST], dev->addr_len);
+		send_addr_notify = 1;
 	}
 
-	if (ida[IFLA_WEIGHT - 1]) {
-		if (ida[IFLA_WEIGHT - 1]->rta_len != RTA_LENGTH(sizeof(u32)))
-			goto out;
-
-		dev->weight = *((u32 *) RTA_DATA(ida[IFLA_WEIGHT - 1]));
-	}
+	if (ifm->ifi_flags)
+		dev_change_flags(dev, ifm->ifi_flags);
 
-	if (ida[IFLA_OPERSTATE - 1]) {
-		if (ida[IFLA_OPERSTATE - 1]->rta_len != RTA_LENGTH(sizeof(u8)))
-			goto out;
+	if (tb[IFLA_TXQLEN])
+		dev->tx_queue_len = nla_get_u32(tb[IFLA_TXQLEN]);
 
-		set_operstate(dev, *((u8 *) RTA_DATA(ida[IFLA_OPERSTATE - 1])));
-	}
+	if (tb[IFLA_WEIGHT])
+		dev->weight = nla_get_u32(tb[IFLA_WEIGHT]);
 
-	if (ida[IFLA_LINKMODE - 1]) {
-		if (ida[IFLA_LINKMODE - 1]->rta_len != RTA_LENGTH(sizeof(u8)))
-			goto out;
+	if (tb[IFLA_OPERSTATE])
+		set_operstate(dev, nla_get_u8(tb[IFLA_OPERSTATE]));
 
+	if (tb[IFLA_LINKMODE]) {
 		write_lock_bh(&dev_base_lock);
-		dev->link_mode = *((u8 *) RTA_DATA(ida[IFLA_LINKMODE - 1]));
+		dev->link_mode = nla_get_u8(tb[IFLA_LINKMODE]);
 		write_unlock_bh(&dev_base_lock);
 	}
 
-	if (ifm->ifi_index >= 0 && ida[IFLA_IFNAME - 1]) {
-		char ifname[IFNAMSIZ];
-
-		if (rtattr_strlcpy(ifname, ida[IFLA_IFNAME - 1],
-		                   IFNAMSIZ) >= IFNAMSIZ)
-			goto out;
-		err = dev_change_name(dev, ifname);
-		if (err)
-			goto out;
-	}
-
-#ifdef CONFIG_NET_WIRELESS_RTNETLINK
-	if (ida[IFLA_WIRELESS - 1]) {
-
-		/* Call Wireless Extensions.
-		 * Various stuff checked in there... */
-		err = wireless_rtnetlink_set(dev, RTA_DATA(ida[IFLA_WIRELESS - 1]), ida[IFLA_WIRELESS - 1]->rta_len);
-		if (err)
-			goto out;
-	}
-#endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
-
 	err = 0;
 
-out:
+errout_dev:
+	if (err < 0 && modified && net_ratelimit())
+		printk(KERN_WARNING "A link change request failed with "
+		       "some changes comitted already. Interface %s may "
+		       "have been left with an inconsistent configuration, "
+		       "please check.\n", dev->name);
+
 	if (send_addr_notify)
 		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
 
 	dev_put(dev);
+errout:
 	return err;
 }
 
@@ -740,7 +748,7 @@ static struct rtnetlink_link link_rtnetl
 					 .doit   = do_getlink,
 #endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
 					 .dumpit = rtnetlink_dump_ifinfo },
-	[RTM_SETLINK     - RTM_BASE] = { .doit   = do_setlink		 },
+	[RTM_SETLINK     - RTM_BASE] = { .doit   = rtnl_setlink		 },
 	[RTM_GETADDR     - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
 	[RTM_GETROUTE    - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
 	[RTM_NEWNEIGH    - RTM_BASE] = { .doit   = neigh_add		 },

[PATCH 8/9] [NET]: Convert link dumping to new netlink api

[Thomas Graf]

[-- Attachment #1: nl_link_dump --]
[-- Type: text/plain, Size: 11635 bytes --]

Transforms netlink code to dump link tables to use the new
netlink api. Makes rtnl_getlink() available regardless of the
availability of the wireless extensions.

Adding copy_rtnl_link_stats() avoids the structural dependency
of struct rtnl_link_stats on struct net_device_stats and thus
avoids troubles later on.

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/net/core/rtnetlink.c
===================================================================
--- net-2.6.19.orig/net/core/rtnetlink.c
+++ net-2.6.19/net/core/rtnetlink.c
@@ -218,41 +218,73 @@ static void set_operstate(struct net_dev
 	}
 }
 
-static int rtnetlink_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
-				 int type, u32 pid, u32 seq, u32 change, 
-				 unsigned int flags)
-{
-	struct ifinfomsg *r;
-	struct nlmsghdr  *nlh;
-	unsigned char	 *b = skb->tail;
-
-	nlh = NLMSG_NEW(skb, pid, seq, type, sizeof(*r), flags);
-	r = NLMSG_DATA(nlh);
-	r->ifi_family = AF_UNSPEC;
-	r->__ifi_pad = 0;
-	r->ifi_type = dev->type;
-	r->ifi_index = dev->ifindex;
-	r->ifi_flags = dev_get_flags(dev);
-	r->ifi_change = change;
+static void copy_rtnl_link_stats(struct rtnl_link_stats *a,
+				 struct net_device_stats *b)
+{
+	a->rx_packets = b->rx_packets;
+	a->tx_packets = b->tx_packets;
+	a->rx_bytes = b->rx_bytes;
+	a->tx_bytes = b->tx_bytes;
+	a->rx_errors = b->rx_errors;
+	a->tx_errors = b->tx_errors;
+	a->rx_dropped = b->rx_dropped;
+	a->tx_dropped = b->tx_dropped;
+
+	a->multicast = b->multicast;
+	a->collisions = b->collisions;
+
+	a->rx_length_errors = b->rx_length_errors;
+	a->rx_over_errors = b->rx_over_errors;
+	a->rx_crc_errors = b->rx_crc_errors;
+	a->rx_frame_errors = b->rx_frame_errors;
+	a->rx_fifo_errors = b->rx_fifo_errors;
+	a->rx_missed_errors = b->rx_missed_errors;
+
+	a->tx_aborted_errors = b->tx_aborted_errors;
+	a->tx_carrier_errors = b->tx_carrier_errors;
+	a->tx_fifo_errors = b->tx_fifo_errors;
+	a->tx_heartbeat_errors = b->tx_heartbeat_errors;
+	a->tx_window_errors = b->tx_window_errors;
 
-	RTA_PUT(skb, IFLA_IFNAME, strlen(dev->name)+1, dev->name);
+	a->rx_compressed = b->rx_compressed;
+	a->tx_compressed = b->tx_compressed;
+};
 
-	if (1) {
-		u32 txqlen = dev->tx_queue_len;
-		RTA_PUT(skb, IFLA_TXQLEN, sizeof(txqlen), &txqlen);
-	}
+static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
+			    void *iwbuf, int iwbuflen, int type, u32 pid,
+			    u32 seq, u32 change, unsigned int flags)
+{
+	struct ifinfomsg *ifm;
+	struct nlmsghdr *nlh;
 
-	if (1) {
-		u32 weight = dev->weight;
-		RTA_PUT(skb, IFLA_WEIGHT, sizeof(weight), &weight);
-	}
+	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*ifm), flags);
+	if (nlh == NULL)
+		return -ENOBUFS;
 
-	if (1) {
-		u8 operstate = netif_running(dev)?dev->operstate:IF_OPER_DOWN;
-		u8 link_mode = dev->link_mode;
-		RTA_PUT(skb, IFLA_OPERSTATE, sizeof(operstate), &operstate);
-		RTA_PUT(skb, IFLA_LINKMODE, sizeof(link_mode), &link_mode);
-	}
+	ifm = nlmsg_data(nlh);
+	ifm->ifi_family = AF_UNSPEC;
+	ifm->__ifi_pad = 0;
+	ifm->ifi_type = dev->type;
+	ifm->ifi_index = dev->ifindex;
+	ifm->ifi_flags = dev_get_flags(dev);
+	ifm->ifi_change = change;
+
+	NLA_PUT_STRING(skb, IFLA_IFNAME, dev->name);
+	NLA_PUT_U32(skb, IFLA_TXQLEN, dev->tx_queue_len);
+	NLA_PUT_U32(skb, IFLA_WEIGHT, dev->weight);
+	NLA_PUT_U8(skb, IFLA_OPERSTATE,
+		   netif_running(dev) ? dev->operstate : IF_OPER_DOWN);
+	NLA_PUT_U8(skb, IFLA_LINKMODE, dev->link_mode);
+	NLA_PUT_U32(skb, IFLA_MTU, dev->mtu);
+
+	if (dev->ifindex != dev->iflink)
+		NLA_PUT_U32(skb, IFLA_LINK, dev->iflink);
+
+	if (dev->master)
+		NLA_PUT_U32(skb, IFLA_MASTER, dev->master->ifindex);
+
+	if (dev->qdisc_sleeping)
+		NLA_PUT_STRING(skb, IFLA_QDISC, dev->qdisc_sleeping->ops->id);
 
 	if (1) {
 		struct rtnl_link_ifmap map = {
@@ -263,58 +295,38 @@ static int rtnetlink_fill_ifinfo(struct 
 			.dma         = dev->dma,
 			.port        = dev->if_port,
 		};
-		RTA_PUT(skb, IFLA_MAP, sizeof(map), &map);
+		NLA_PUT(skb, IFLA_MAP, sizeof(map), &map);
 	}
 
 	if (dev->addr_len) {
-		RTA_PUT(skb, IFLA_ADDRESS, dev->addr_len, dev->dev_addr);
-		RTA_PUT(skb, IFLA_BROADCAST, dev->addr_len, dev->broadcast);
-	}
-
-	if (1) {
-		u32 mtu = dev->mtu;
-		RTA_PUT(skb, IFLA_MTU, sizeof(mtu), &mtu);
-	}
-
-	if (dev->ifindex != dev->iflink) {
-		u32 iflink = dev->iflink;
-		RTA_PUT(skb, IFLA_LINK, sizeof(iflink), &iflink);
-	}
-
-	if (dev->qdisc_sleeping)
-		RTA_PUT(skb, IFLA_QDISC,
-			strlen(dev->qdisc_sleeping->ops->id) + 1,
-			dev->qdisc_sleeping->ops->id);
-	
-	if (dev->master) {
-		u32 master = dev->master->ifindex;
-		RTA_PUT(skb, IFLA_MASTER, sizeof(master), &master);
+		NLA_PUT(skb, IFLA_ADDRESS, dev->addr_len, dev->dev_addr);
+		NLA_PUT(skb, IFLA_BROADCAST, dev->addr_len, dev->broadcast);
 	}
 
 	if (dev->get_stats) {
-		unsigned long *stats = (unsigned long*)dev->get_stats(dev);
+		struct net_device_stats *stats = dev->get_stats(dev);
 		if (stats) {
-			struct rtattr  *a;
-			__u32	       *s;
-			int		i;
-			int		n = sizeof(struct rtnl_link_stats)/4;
-
-			a = __RTA_PUT(skb, IFLA_STATS, n*4);
-			s = RTA_DATA(a);
-			for (i=0; i<n; i++)
-				s[i] = stats[i];
+			struct nlattr *attr;
+
+			attr = nla_reserve(skb, IFLA_STATS,
+					   sizeof(struct rtnl_link_stats));
+			if (attr == NULL)
+				goto nla_put_failure;
+
+			copy_rtnl_link_stats(nla_data(attr), stats);
 		}
 	}
-	nlh->nlmsg_len = skb->tail - b;
-	return skb->len;
 
-nlmsg_failure:
-rtattr_failure:
-	skb_trim(skb, b - skb->data);
-	return -1;
+	if (iwbuf)
+		NLA_PUT(skb, IFLA_WIRELESS, iwbuflen, iwbuf);
+
+	return nlmsg_end(skb, nlh);
+
+nla_put_failure:
+	return nlmsg_cancel(skb, nlh);
 }
 
-static int rtnetlink_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
+static int rtnl_dump_ifinfo(struct sk_buff *skb, struct netlink_callback *cb)
 {
 	int idx;
 	int s_idx = cb->args[0];
@@ -324,10 +336,9 @@ static int rtnetlink_dump_ifinfo(struct 
 	for (dev=dev_base, idx=0; dev; dev = dev->next, idx++) {
 		if (idx < s_idx)
 			continue;
-		if (rtnetlink_fill_ifinfo(skb, dev, RTM_NEWLINK,
-					  NETLINK_CB(cb->skb).pid,
-					  cb->nlh->nlmsg_seq, 0,
-					  NLM_F_MULTI) <= 0)
+		if (rtnl_fill_ifinfo(skb, dev, NULL, 0, RTM_NEWLINK,
+				     NETLINK_CB(cb->skb).pid,
+				     cb->nlh->nlmsg_seq, 0, NLM_F_MULTI) <= 0)
 			break;
 	}
 	read_unlock(&dev_base_lock);
@@ -501,84 +512,69 @@ errout:
 	return err;
 }
 
-#ifdef CONFIG_NET_WIRELESS_RTNETLINK
-static int do_getlink(struct sk_buff *in_skb, struct nlmsghdr* in_nlh, void *arg)
+static int rtnl_getlink(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 {
-	struct ifinfomsg  *ifm = NLMSG_DATA(in_nlh);
-	struct rtattr    **ida = arg;
-	struct net_device *dev;
-	struct ifinfomsg *r;
-	struct nlmsghdr  *nlh;
-	int err = -ENOBUFS;
-	struct sk_buff *skb;
-	unsigned char	 *b;
-	char *iw_buf = NULL;
+	struct ifinfomsg *ifm;
+	struct nlattr *tb[IFLA_MAX+1];
+	struct net_device *dev = NULL;
+	struct sk_buff *nskb;
+	char *iw_buf = NULL, *iw = NULL;
 	int iw_buf_len = 0;
+	int err, payload;
 
-	if (ifm->ifi_index >= 0)
+	err = nlmsg_parse(nlh, sizeof(*ifm), tb, IFLA_MAX, ifla_policy);
+	if (err < 0)
+		goto errout;
+
+	ifm = nlmsg_data(nlh);
+	if (ifm->ifi_index >= 0) {
 		dev = dev_get_by_index(ifm->ifi_index);
-	else
+		if (dev == NULL)
+			return -ENODEV;
+	} else
 		return -EINVAL;
-	if (!dev)
-		return -ENODEV;
 
-#ifdef CONFIG_NET_WIRELESS_RTNETLINK
-	if (ida[IFLA_WIRELESS - 1]) {
 
+#ifdef CONFIG_NET_WIRELESS_RTNETLINK
+	if (tb[IFLA_WIRELESS]) {
 		/* Call Wireless Extensions. We need to know the size before
 		 * we can alloc. Various stuff checked in there... */
-		err = wireless_rtnetlink_get(dev, RTA_DATA(ida[IFLA_WIRELESS - 1]), ida[IFLA_WIRELESS - 1]->rta_len, &iw_buf, &iw_buf_len);
-		if (err)
-			goto out;
+		err = wireless_rtnetlink_get(dev, nla_data(tb[IFLA_WIRELESS]),
+					     nla_len(tb[IFLA_WIRELESS]),
+					     &iw_buf, &iw_buf_len);
+		if (err < 0)
+			goto errout;
+
+		iw += IW_EV_POINT_OFF;
 	}
 #endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
 
-	/* Create a skb big enough to include all the data.
-	 * Some requests are way bigger than 4k... Jean II */
-	skb = alloc_skb((NLMSG_LENGTH(sizeof(*r))) + (RTA_SPACE(iw_buf_len)),
-			GFP_KERNEL);
-	if (!skb)
-		goto out;
-	b = skb->tail;
-
-	/* Put in the message the usual good stuff */
-	nlh = NLMSG_PUT(skb, NETLINK_CB(in_skb).pid, in_nlh->nlmsg_seq,
-			RTM_NEWLINK, sizeof(*r));
-	r = NLMSG_DATA(nlh);
-	r->ifi_family = AF_UNSPEC;
-	r->__ifi_pad = 0;
-	r->ifi_type = dev->type;
-	r->ifi_index = dev->ifindex;
-	r->ifi_flags = dev->flags;
-	r->ifi_change = 0;
-
-	/* Put the wireless payload if it exist */
-	if(iw_buf != NULL)
-		RTA_PUT(skb, IFLA_WIRELESS, iw_buf_len,
-			iw_buf + IW_EV_POINT_OFF);
-
-	nlh->nlmsg_len = skb->tail - b;
+	payload = NLMSG_ALIGN(sizeof(struct ifinfomsg) +
+			      nla_total_size(iw_buf_len));
+	nskb = nlmsg_new(nlmsg_total_size(payload), GFP_KERNEL);
+	if (nskb == NULL) {
+		err = -ENOBUFS;
+		goto errout;
+	}
 
-	/* Needed ? */
-	NETLINK_CB(skb).dst_pid = NETLINK_CB(in_skb).pid;
+	err = rtnl_fill_ifinfo(nskb, dev, iw, iw_buf_len, RTM_NEWLINK,
+			       NETLINK_CB(skb).pid, nlh->nlmsg_seq, 0, 0);
+	if (err <= 0) {
+		kfree_skb(skb);
+		goto errout;
+	}
 
-	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
+	err = netlink_unicast(rtnl, skb, NETLINK_CB(skb).pid, MSG_DONTWAIT);
 	if (err > 0)
 		err = 0;
-out:
-	if(iw_buf != NULL)
-		kfree(iw_buf);
+errout:
+	kfree(iw_buf);
 	dev_put(dev);
-	return err;
 
-rtattr_failure:
-nlmsg_failure:
-	kfree_skb(skb);
-	goto out;
+	return err;
 }
-#endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
 
-static int rtnetlink_dump_all(struct sk_buff *skb, struct netlink_callback *cb)
+static int rtnl_dump_all(struct sk_buff *skb, struct netlink_callback *cb)
 {
 	int idx;
 	int s_idx = cb->family;
@@ -609,11 +605,11 @@ void rtmsg_ifinfo(int type, struct net_d
 			       sizeof(struct rtnl_link_ifmap) +
 			       sizeof(struct rtnl_link_stats) + 128);
 
-	skb = alloc_skb(size, GFP_KERNEL);
+	skb = nlmsg_new(size, GFP_KERNEL);
 	if (!skb)
 		return;
 
-	if (rtnetlink_fill_ifinfo(skb, dev, type, 0, 0, change, 0) < 0) {
+	if (rtnl_fill_ifinfo(skb, dev, NULL, 0, type, 0, 0, change, 0) < 0) {
 		kfree_skb(skb);
 		return;
 	}
@@ -743,14 +739,11 @@ static void rtnetlink_rcv(struct sock *s
 
 static struct rtnetlink_link link_rtnetlink_table[RTM_NR_MSGTYPES] =
 {
-	[RTM_GETLINK     - RTM_BASE] = {
-#ifdef CONFIG_NET_WIRELESS_RTNETLINK
-					 .doit   = do_getlink,
-#endif	/* CONFIG_NET_WIRELESS_RTNETLINK */
-					 .dumpit = rtnetlink_dump_ifinfo },
+	[RTM_GETLINK     - RTM_BASE] = { .doit   = rtnl_getlink,
+					 .dumpit = rtnl_dump_ifinfo	 },
 	[RTM_SETLINK     - RTM_BASE] = { .doit   = rtnl_setlink		 },
-	[RTM_GETADDR     - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
-	[RTM_GETROUTE    - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
+	[RTM_GETADDR     - RTM_BASE] = { .dumpit = rtnl_dump_all	 },
+	[RTM_GETROUTE    - RTM_BASE] = { .dumpit = rtnl_dump_all	 },
 	[RTM_NEWNEIGH    - RTM_BASE] = { .doit   = neigh_add		 },
 	[RTM_DELNEIGH    - RTM_BASE] = { .doit   = neigh_delete		 },
 	[RTM_GETNEIGH    - RTM_BASE] = { .dumpit = neigh_dump_info	 },
@@ -758,7 +751,7 @@ static struct rtnetlink_link link_rtnetl
 	[RTM_NEWRULE     - RTM_BASE] = { .doit   = fib_nl_newrule	 },
 	[RTM_DELRULE     - RTM_BASE] = { .doit   = fib_nl_delrule	 },
 #endif
-	[RTM_GETRULE     - RTM_BASE] = { .dumpit = rtnetlink_dump_all	 },
+	[RTM_GETRULE     - RTM_BASE] = { .dumpit = rtnl_dump_all	 },
 	[RTM_GETNEIGHTBL - RTM_BASE] = { .dumpit = neightbl_dump_info	 },
 	[RTM_SETNEIGHTBL - RTM_BASE] = { .doit   = neightbl_set		 },
 };

[PATCH 9/9] [NET]: Move netlink interface bits to linux/if.h

[Thomas Graf]

[-- Attachment #1: nl_if --]
[-- Type: text/plain, Size: 8238 bytes --]

Signed-off-by: Thomas Graf <tgraf@suug.ch>

Index: net-2.6.19/include/linux/if.h
===================================================================
--- net-2.6.19.orig/include/linux/if.h
+++ net-2.6.19/include/linux/if.h
@@ -212,5 +212,134 @@ struct ifconf 
 #define	ifc_buf	ifc_ifcu.ifcu_buf		/* buffer address	*/
 #define	ifc_req	ifc_ifcu.ifcu_req		/* array of structures	*/
 
+/* The struct should be in sync with struct net_device_stats */
+struct rtnl_link_stats
+{
+	__u32	rx_packets;		/* total packets received	*/
+	__u32	tx_packets;		/* total packets transmitted	*/
+	__u32	rx_bytes;		/* total bytes received 	*/
+	__u32	tx_bytes;		/* total bytes transmitted	*/
+	__u32	rx_errors;		/* bad packets received		*/
+	__u32	tx_errors;		/* packet transmit problems	*/
+	__u32	rx_dropped;		/* no space in linux buffers	*/
+	__u32	tx_dropped;		/* no space available in linux	*/
+	__u32	multicast;		/* multicast packets received	*/
+	__u32	collisions;
+
+	/* detailed rx_errors: */
+	__u32	rx_length_errors;
+	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
+	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
+	__u32	rx_frame_errors;	/* recv'd frame alignment error */
+	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
+	__u32	rx_missed_errors;	/* receiver missed packet	*/
+
+	/* detailed tx_errors */
+	__u32	tx_aborted_errors;
+	__u32	tx_carrier_errors;
+	__u32	tx_fifo_errors;
+	__u32	tx_heartbeat_errors;
+	__u32	tx_window_errors;
+
+	/* for cslip etc */
+	__u32	rx_compressed;
+	__u32	tx_compressed;
+};
+
+/* The struct should be in sync with struct ifmap */
+struct rtnl_link_ifmap
+{
+	__u64	mem_start;
+	__u64	mem_end;
+	__u64	base_addr;
+	__u16	irq;
+	__u8	dma;
+	__u8	port;
+};
+
+enum
+{
+	IFLA_UNSPEC,
+	IFLA_ADDRESS,
+	IFLA_BROADCAST,
+	IFLA_IFNAME,
+	IFLA_MTU,
+	IFLA_LINK,
+	IFLA_QDISC,
+	IFLA_STATS,
+	IFLA_COST,
+#define IFLA_COST IFLA_COST
+	IFLA_PRIORITY,
+#define IFLA_PRIORITY IFLA_PRIORITY
+	IFLA_MASTER,
+#define IFLA_MASTER IFLA_MASTER
+	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
+#define IFLA_WIRELESS IFLA_WIRELESS
+	IFLA_PROTINFO,		/* Protocol specific information for a link */
+#define IFLA_PROTINFO IFLA_PROTINFO
+	IFLA_TXQLEN,
+#define IFLA_TXQLEN IFLA_TXQLEN
+	IFLA_MAP,
+#define IFLA_MAP IFLA_MAP
+	IFLA_WEIGHT,
+#define IFLA_WEIGHT IFLA_WEIGHT
+	IFLA_OPERSTATE,
+	IFLA_LINKMODE,
+	__IFLA_MAX
+};
+
+
+#define IFLA_MAX (__IFLA_MAX - 1)
+
+/* ifi_flags.
+
+   IFF_* flags.
+
+   The only change is:
+   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
+   more not changeable by user. They describe link media
+   characteristics and set by device driver.
+
+   Comments:
+   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
+   - If neither of these three flags are set;
+     the interface is NBMA.
+
+   - IFF_MULTICAST does not mean anything special:
+   multicasts can be used on all not-NBMA links.
+   IFF_MULTICAST means that this media uses special encapsulation
+   for multicast frames. Apparently, all IFF_POINTOPOINT and
+   IFF_BROADCAST devices are able to use multicasts too.
+ */
+
+/* IFLA_LINK.
+   For usual devices it is equal ifi_index.
+   If it is a "virtual interface" (f.e. tunnel), ifi_link
+   can point to real physical interface (f.e. for bandwidth calculations),
+   or maybe 0, what means, that real media is unknown (usual
+   for IPIP tunnels, when route to endpoint is allowed to change)
+ */
+
+/* Subtype attributes for IFLA_PROTINFO */
+enum
+{
+	IFLA_INET6_UNSPEC,
+	IFLA_INET6_FLAGS,	/* link flags			*/
+	IFLA_INET6_CONF,	/* sysctl parameters		*/
+	IFLA_INET6_STATS,	/* statistics			*/
+	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
+	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
+	__IFLA_INET6_MAX
+};
+
+#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
+
+struct ifla_cacheinfo
+{
+	__u32	max_reasm_len;
+	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
+	__u32	reachable_time;
+	__u32	retrans_time;
+};
 
 #endif /* _LINUX_IF_H */
Index: net-2.6.19/include/linux/rtnetlink.h
===================================================================
--- net-2.6.19.orig/include/linux/rtnetlink.h
+++ net-2.6.19/include/linux/rtnetlink.h
@@ -2,6 +2,7 @@
 #define __LINUX_RTNETLINK_H
 
 #include <linux/netlink.h>
+#include <linux/if.h>
 
 /****
  *		Routing/neighbour discovery messages.
@@ -607,138 +608,6 @@ struct prefix_cacheinfo
 	__u32	valid_time;
 };
 
-/* The struct should be in sync with struct net_device_stats */
-struct rtnl_link_stats
-{
-	__u32	rx_packets;		/* total packets received	*/
-	__u32	tx_packets;		/* total packets transmitted	*/
-	__u32	rx_bytes;		/* total bytes received 	*/
-	__u32	tx_bytes;		/* total bytes transmitted	*/
-	__u32	rx_errors;		/* bad packets received		*/
-	__u32	tx_errors;		/* packet transmit problems	*/
-	__u32	rx_dropped;		/* no space in linux buffers	*/
-	__u32	tx_dropped;		/* no space available in linux	*/
-	__u32	multicast;		/* multicast packets received	*/
-	__u32	collisions;
-
-	/* detailed rx_errors: */
-	__u32	rx_length_errors;
-	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
-	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
-	__u32	rx_frame_errors;	/* recv'd frame alignment error */
-	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
-	__u32	rx_missed_errors;	/* receiver missed packet	*/
-
-	/* detailed tx_errors */
-	__u32	tx_aborted_errors;
-	__u32	tx_carrier_errors;
-	__u32	tx_fifo_errors;
-	__u32	tx_heartbeat_errors;
-	__u32	tx_window_errors;
-	
-	/* for cslip etc */
-	__u32	rx_compressed;
-	__u32	tx_compressed;
-};
-
-/* The struct should be in sync with struct ifmap */
-struct rtnl_link_ifmap
-{
-	__u64	mem_start;
-	__u64	mem_end;
-	__u64	base_addr;
-	__u16	irq;
-	__u8	dma;
-	__u8	port;
-};
-
-enum
-{
-	IFLA_UNSPEC,
-	IFLA_ADDRESS,
-	IFLA_BROADCAST,
-	IFLA_IFNAME,
-	IFLA_MTU,
-	IFLA_LINK,
-	IFLA_QDISC,
-	IFLA_STATS,
-	IFLA_COST,
-#define IFLA_COST IFLA_COST
-	IFLA_PRIORITY,
-#define IFLA_PRIORITY IFLA_PRIORITY
-	IFLA_MASTER,
-#define IFLA_MASTER IFLA_MASTER
-	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
-#define IFLA_WIRELESS IFLA_WIRELESS
-	IFLA_PROTINFO,		/* Protocol specific information for a link */
-#define IFLA_PROTINFO IFLA_PROTINFO
-	IFLA_TXQLEN,
-#define IFLA_TXQLEN IFLA_TXQLEN
-	IFLA_MAP,
-#define IFLA_MAP IFLA_MAP
-	IFLA_WEIGHT,
-#define IFLA_WEIGHT IFLA_WEIGHT
-	IFLA_OPERSTATE,
-	IFLA_LINKMODE,
-	__IFLA_MAX
-};
-
-
-#define IFLA_MAX (__IFLA_MAX - 1)
-
-#define IFLA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifinfomsg))))
-#define IFLA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifinfomsg))
-
-/* ifi_flags.
-
-   IFF_* flags.
-
-   The only change is:
-   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
-   more not changeable by user. They describe link media
-   characteristics and set by device driver.
-
-   Comments:
-   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
-   - If neither of these three flags are set;
-     the interface is NBMA.
-
-   - IFF_MULTICAST does not mean anything special:
-   multicasts can be used on all not-NBMA links.
-   IFF_MULTICAST means that this media uses special encapsulation
-   for multicast frames. Apparently, all IFF_POINTOPOINT and
-   IFF_BROADCAST devices are able to use multicasts too.
- */
-
-/* IFLA_LINK.
-   For usual devices it is equal ifi_index.
-   If it is a "virtual interface" (f.e. tunnel), ifi_link
-   can point to real physical interface (f.e. for bandwidth calculations),
-   or maybe 0, what means, that real media is unknown (usual
-   for IPIP tunnels, when route to endpoint is allowed to change)
- */
-
-/* Subtype attributes for IFLA_PROTINFO */
-enum
-{
-	IFLA_INET6_UNSPEC,
-	IFLA_INET6_FLAGS,	/* link flags			*/
-	IFLA_INET6_CONF,	/* sysctl parameters		*/
-	IFLA_INET6_STATS,	/* statistics			*/
-	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
-	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
-	__IFLA_INET6_MAX
-};
-
-#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
-
-struct ifla_cacheinfo
-{
-	__u32	max_reasm_len;
-	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
-	__u32	reachable_time;
-	__u32	retrans_time;
-};
 
 /*****************************************************************
  *		Traffic control messages.

[PATCHSET] Begin of transformations to new netlink api

[Thomas Graf]

First batch of patches converting existing netlink code to use
the new api. Fixing lots of unvalidated accesses to attributes.

Re: [PATCHSET] Begin of transformations to new netlink api

[David Miller]

From: Thomas Graf <tgraf@suug.ch>
Date: Sat, 05 Aug 2006 02:34:52 +0200

> First batch of patches converting existing netlink code to use
> the new api. Fixing lots of unvalidated accesses to attributes.

Looks very nice, all applied to net-2.6.19, thanks Thomas.

I've also rebased the net-2.6.19 GIT tree.

Re: [PATCH 9/9] [NET]: Move netlink interface bits to linux/if.h

[David Miller]

From: Stephen Hemminger <shemminger@osdl.org>
Date: Sat, 05 Aug 2006 22:42:18 -0700

> Thomas Graf wrote:
> > -/* The struct should be in sync with struct net_device_stats */
> > -struct rtnl_link_stats
> > -{
> > -	__u32	rx_packets;		/* total packets received	*/
> > -	__u32	tx_packets;		/* total packets transmitted	*/
> > -	__u32	rx_bytes;		/* total bytes received 	*/
> > -	__u32	tx_bytes;		/* total bytes transmitted	*/
> >   
> 
> Make the packets and byte counts __u64 so if/when we extend net_device 
> stats to bigger
> values, the message format doesn't have to change.

It's an existing ABI that he is just moving from one place
to another, we are not at liberty to change it's layout.

We chould create a new one, but that makes no sense until
the device layer actually has the means by which to maintain
u64 values.

Re: [PATCH 9/9] [NET]: Move netlink interface bits to linux/if.h

[Stephen Hemminger]

Thomas Graf wrote:
> Signed-off-by: Thomas Graf <tgraf@suug.ch>
>
> Index: net-2.6.19/include/linux/if.h
> ===================================================================
> --- net-2.6.19.orig/include/linux/if.h
> +++ net-2.6.19/include/linux/if.h
> @@ -212,5 +212,134 @@ struct ifconf 
>  #define	ifc_buf	ifc_ifcu.ifcu_buf		/* buffer address	*/
>  #define	ifc_req	ifc_ifcu.ifcu_req		/* array of structures	*/
>  
> +/* The struct should be in sync with struct net_device_stats */
> +struct rtnl_link_stats
> +{
> +	__u32	rx_packets;		/* total packets received	*/
> +	__u32	tx_packets;		/* total packets transmitted	*/
> +	__u32	rx_bytes;		/* total bytes received 	*/
> +	__u32	tx_bytes;		/* total bytes transmitted	*/
> +	__u32	rx_errors;		/* bad packets received		*/
> +	__u32	tx_errors;		/* packet transmit problems	*/
> +	__u32	rx_dropped;		/* no space in linux buffers	*/
> +	__u32	tx_dropped;		/* no space available in linux	*/
> +	__u32	multicast;		/* multicast packets received	*/
> +	__u32	collisions;
> +
> +	/* detailed rx_errors: */
> +	__u32	rx_length_errors;
> +	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
> +	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
> +	__u32	rx_frame_errors;	/* recv'd frame alignment error */
> +	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
> +	__u32	rx_missed_errors;	/* receiver missed packet	*/
> +
> +	/* detailed tx_errors */
> +	__u32	tx_aborted_errors;
> +	__u32	tx_carrier_errors;
> +	__u32	tx_fifo_errors;
> +	__u32	tx_heartbeat_errors;
> +	__u32	tx_window_errors;
> +
> +	/* for cslip etc */
> +	__u32	rx_compressed;
> +	__u32	tx_compressed;
> +};
> +
> +/* The struct should be in sync with struct ifmap */
> +struct rtnl_link_ifmap
> +{
> +	__u64	mem_start;
> +	__u64	mem_end;
> +	__u64	base_addr;
> +	__u16	irq;
> +	__u8	dma;
> +	__u8	port;
> +};
> +
> +enum
> +{
> +	IFLA_UNSPEC,
> +	IFLA_ADDRESS,
> +	IFLA_BROADCAST,
> +	IFLA_IFNAME,
> +	IFLA_MTU,
> +	IFLA_LINK,
> +	IFLA_QDISC,
> +	IFLA_STATS,
> +	IFLA_COST,
> +#define IFLA_COST IFLA_COST
> +	IFLA_PRIORITY,
> +#define IFLA_PRIORITY IFLA_PRIORITY
> +	IFLA_MASTER,
> +#define IFLA_MASTER IFLA_MASTER
> +	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
> +#define IFLA_WIRELESS IFLA_WIRELESS
> +	IFLA_PROTINFO,		/* Protocol specific information for a link */
> +#define IFLA_PROTINFO IFLA_PROTINFO
> +	IFLA_TXQLEN,
> +#define IFLA_TXQLEN IFLA_TXQLEN
> +	IFLA_MAP,
> +#define IFLA_MAP IFLA_MAP
> +	IFLA_WEIGHT,
> +#define IFLA_WEIGHT IFLA_WEIGHT
> +	IFLA_OPERSTATE,
> +	IFLA_LINKMODE,
> +	__IFLA_MAX
> +};
> +
> +
> +#define IFLA_MAX (__IFLA_MAX - 1)
> +
> +/* ifi_flags.
> +
> +   IFF_* flags.
> +
> +   The only change is:
> +   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
> +   more not changeable by user. They describe link media
> +   characteristics and set by device driver.
> +
> +   Comments:
> +   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
> +   - If neither of these three flags are set;
> +     the interface is NBMA.
> +
> +   - IFF_MULTICAST does not mean anything special:
> +   multicasts can be used on all not-NBMA links.
> +   IFF_MULTICAST means that this media uses special encapsulation
> +   for multicast frames. Apparently, all IFF_POINTOPOINT and
> +   IFF_BROADCAST devices are able to use multicasts too.
> + */
> +
> +/* IFLA_LINK.
> +   For usual devices it is equal ifi_index.
> +   If it is a "virtual interface" (f.e. tunnel), ifi_link
> +   can point to real physical interface (f.e. for bandwidth calculations),
> +   or maybe 0, what means, that real media is unknown (usual
> +   for IPIP tunnels, when route to endpoint is allowed to change)
> + */
> +
> +/* Subtype attributes for IFLA_PROTINFO */
> +enum
> +{
> +	IFLA_INET6_UNSPEC,
> +	IFLA_INET6_FLAGS,	/* link flags			*/
> +	IFLA_INET6_CONF,	/* sysctl parameters		*/
> +	IFLA_INET6_STATS,	/* statistics			*/
> +	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
> +	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
> +	__IFLA_INET6_MAX
> +};
> +
> +#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
> +
> +struct ifla_cacheinfo
> +{
> +	__u32	max_reasm_len;
> +	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
> +	__u32	reachable_time;
> +	__u32	retrans_time;
> +};
>  
>  #endif /* _LINUX_IF_H */
> Index: net-2.6.19/include/linux/rtnetlink.h
> ===================================================================
> --- net-2.6.19.orig/include/linux/rtnetlink.h
> +++ net-2.6.19/include/linux/rtnetlink.h
> @@ -2,6 +2,7 @@
>  #define __LINUX_RTNETLINK_H
>  
>  #include <linux/netlink.h>
> +#include <linux/if.h>
>  
>  /****
>   *		Routing/neighbour discovery messages.
> @@ -607,138 +608,6 @@ struct prefix_cacheinfo
>  	__u32	valid_time;
>  };
>  
> -/* The struct should be in sync with struct net_device_stats */
> -struct rtnl_link_stats
> -{
> -	__u32	rx_packets;		/* total packets received	*/
> -	__u32	tx_packets;		/* total packets transmitted	*/
> -	__u32	rx_bytes;		/* total bytes received 	*/
> -	__u32	tx_bytes;		/* total bytes transmitted	*/
>   

Make the packets and byte counts __u64 so if/when we extend net_device 
stats to bigger
values, the message format doesn't have to change.
> -	__u32	rx_errors;		/* bad packets received		*/
> -	__u32	tx_errors;		/* packet transmit problems	*/
> -	__u32	rx_dropped;		/* no space in linux buffers	*/
> -	__u32	tx_dropped;		/* no space available in linux	*/
> -	__u32	multicast;		/* multicast packets received	*/
> -	__u32	collisions;
> -
> -	/* detailed rx_errors: */
> -	__u32	rx_length_errors;
> -	__u32	rx_over_errors;		/* receiver ring buff overflow	*/
> -	__u32	rx_crc_errors;		/* recved pkt with crc error	*/
> -	__u32	rx_frame_errors;	/* recv'd frame alignment error */
> -	__u32	rx_fifo_errors;		/* recv'r fifo overrun		*/
> -	__u32	rx_missed_errors;	/* receiver missed packet	*/
> -
> -	/* detailed tx_errors */
> -	__u32	tx_aborted_errors;
> -	__u32	tx_carrier_errors;
> -	__u32	tx_fifo_errors;
> -	__u32	tx_heartbeat_errors;
> -	__u32	tx_window_errors;
> -	
> -	/* for cslip etc */
> -	__u32	rx_compressed;
> -	__u32	tx_compressed;
> -};
> -
> -/* The struct should be in sync with struct ifmap */
> -struct rtnl_link_ifmap
> -{
> -	__u64	mem_start;
> -	__u64	mem_end;
> -	__u64	base_addr;
> -	__u16	irq;
> -	__u8	dma;
> -	__u8	port;
> -};
> -
> -enum
> -{
> -	IFLA_UNSPEC,
> -	IFLA_ADDRESS,
> -	IFLA_BROADCAST,
> -	IFLA_IFNAME,
> -	IFLA_MTU,
> -	IFLA_LINK,
> -	IFLA_QDISC,
> -	IFLA_STATS,
> -	IFLA_COST,
> -#define IFLA_COST IFLA_COST
> -	IFLA_PRIORITY,
> -#define IFLA_PRIORITY IFLA_PRIORITY
> -	IFLA_MASTER,
> -#define IFLA_MASTER IFLA_MASTER
> -	IFLA_WIRELESS,		/* Wireless Extension event - see wireless.h */
> -#define IFLA_WIRELESS IFLA_WIRELESS
> -	IFLA_PROTINFO,		/* Protocol specific information for a link */
> -#define IFLA_PROTINFO IFLA_PROTINFO
> -	IFLA_TXQLEN,
> -#define IFLA_TXQLEN IFLA_TXQLEN
> -	IFLA_MAP,
> -#define IFLA_MAP IFLA_MAP
> -	IFLA_WEIGHT,
> -#define IFLA_WEIGHT IFLA_WEIGHT
> -	IFLA_OPERSTATE,
> -	IFLA_LINKMODE,
> -	__IFLA_MAX
> -};
> -
> -
> -#define IFLA_MAX (__IFLA_MAX - 1)
> -
> -#define IFLA_RTA(r)  ((struct rtattr*)(((char*)(r)) + NLMSG_ALIGN(sizeof(struct ifinfomsg))))
> -#define IFLA_PAYLOAD(n) NLMSG_PAYLOAD(n,sizeof(struct ifinfomsg))
> -
> -/* ifi_flags.
> -
> -   IFF_* flags.
> -
> -   The only change is:
> -   IFF_LOOPBACK, IFF_BROADCAST and IFF_POINTOPOINT are
> -   more not changeable by user. They describe link media
> -   characteristics and set by device driver.
> -
> -   Comments:
> -   - Combination IFF_BROADCAST|IFF_POINTOPOINT is invalid
> -   - If neither of these three flags are set;
> -     the interface is NBMA.
> -
> -   - IFF_MULTICAST does not mean anything special:
> -   multicasts can be used on all not-NBMA links.
> -   IFF_MULTICAST means that this media uses special encapsulation
> -   for multicast frames. Apparently, all IFF_POINTOPOINT and
> -   IFF_BROADCAST devices are able to use multicasts too.
> - */
> -
> -/* IFLA_LINK.
> -   For usual devices it is equal ifi_index.
> -   If it is a "virtual interface" (f.e. tunnel), ifi_link
> -   can point to real physical interface (f.e. for bandwidth calculations),
> -   or maybe 0, what means, that real media is unknown (usual
> -   for IPIP tunnels, when route to endpoint is allowed to change)
> - */
> -
> -/* Subtype attributes for IFLA_PROTINFO */
> -enum
> -{
> -	IFLA_INET6_UNSPEC,
> -	IFLA_INET6_FLAGS,	/* link flags			*/
> -	IFLA_INET6_CONF,	/* sysctl parameters		*/
> -	IFLA_INET6_STATS,	/* statistics			*/
> -	IFLA_INET6_MCAST,	/* MC things. What of them?	*/
> -	IFLA_INET6_CACHEINFO,	/* time values and max reasm size */
> -	__IFLA_INET6_MAX
> -};
> -
> -#define IFLA_INET6_MAX	(__IFLA_INET6_MAX - 1)
> -
> -struct ifla_cacheinfo
> -{
> -	__u32	max_reasm_len;
> -	__u32	tstamp;		/* ipv6InterfaceTable updated timestamp */
> -	__u32	reachable_time;
> -	__u32	retrans_time;
> -};
>  
>  /*****************************************************************
>   *		Traffic control messages.
>
> -
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>   

Re: [PATCH 9/9] [NET]: Move netlink interface bits to linux/if.h

[Stephen Hemminger]

On Fri, 04 Aug 2006 23:23:14 -0700 (PDT)
David Miller <davem@davemloft.net> wrote:

> From: Stephen Hemminger <shemminger@osdl.org>
> Date: Sat, 05 Aug 2006 22:42:18 -0700
> 
> > Thomas Graf wrote:
> > > -/* The struct should be in sync with struct net_device_stats */
> > > -struct rtnl_link_stats
> > > -{
> > > -	__u32	rx_packets;		/* total packets received	*/
> > > -	__u32	tx_packets;		/* total packets transmitted	*/
> > > -	__u32	rx_bytes;		/* total bytes received 	*/
> > > -	__u32	tx_bytes;		/* total bytes transmitted	*/
> > >   
> > 
> > Make the packets and byte counts __u64 so if/when we extend net_device 
> > stats to bigger
> > values, the message format doesn't have to change.
> 
> It's an existing ABI that he is just moving from one place
> to another, we are not at liberty to change it's layout.

Yeah that makes sense, I thought it was a new interface.

-- 
Stephen Hemminger <shemminger@osdl.org>
"And in the Packet there writ down that doome"