Skip to contentSkip to conversation

ieee802154: Fix EUI-64 station address validation.

RFC PATCH In review Lennert Buytenhek 3 participants 4 messages May 28, 2015 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): 53cacc49e75b59f7b6458075b032be16f1133abb.1432816524.git.buytenh@wantstofly.org

#1 lore raw

RFC PATCH 2/2 ieee802154: Fix EUI-64 station address validation.

Refuse to allow setting an EUI-64 group address as an interface address, as those are not valid station addresses.

---
 include/linux/ieee802154.h | 10 ++++------
 net/mac802154/iface.c      |  4 ++--
 2 files changed, 6 insertions(+), 8 deletions(-)
include/linux/ieee802154.h +4 -6
index 552210d..1dc1f4e 100644
@@ -225,15 +225,13 @@ static inline bool ieee802154_is_valid_psdu_len(const u8 len)
  * ieee802154_is_valid_psdu_len - check if extended addr is valid
  * @addr: extended addr to check
  */
-static inline bool ieee802154_is_valid_extended_addr(const __le64 addr)
+static inline bool ieee802154_is_valid_extended_unicast_addr(const __le64 addr)
 {
-	/* These EUI-64 addresses are reserved by IEEE. 0xffffffffffffffff
-	 * is used internally as extended to short address broadcast mapping.
-	 * This is currently a workaround because neighbor discovery can't
-	 * deal with short addresses types right now.
+	/* Bail out if the address is all zero, or if the group
+	 * address bit is set.
 	 */
 	return ((addr != cpu_to_le64(0x0000000000000000ULL)) &&
-		(addr != cpu_to_le64(0xffffffffffffffffULL)));
+		!(addr & cpu_to_le64(0x0100000000000000ULL)));
 }
 
 /**
net/mac802154/iface.c +2 -2
index b544b5d..6ac0239 100644
@@ -126,7 +126,7 @@ static int mac802154_wpan_mac_addr(struct net_device *dev, void *p)
 		return -EBUSY;
 
 	ieee802154_be64_to_le64(&extended_addr, addr->sa_data);
-	if (!ieee802154_is_valid_extended_addr(extended_addr))
+	if (!ieee802154_is_valid_extended_unicast_addr(extended_addr))
 		return -EINVAL;
 
 	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
@@ -539,7 +539,7 @@ ieee802154_if_add(struct ieee802154_local *local, const char *name,
 	switch (type) {
 	case NL802154_IFTYPE_NODE:
 		ndev->type = ARPHRD_IEEE802154;
-		if (ieee802154_is_valid_extended_addr(extended_addr))
+		if (ieee802154_is_valid_extended_unicast_addr(extended_addr))
 			ieee802154_le64_to_be64(ndev->dev_addr, &extended_addr);
 		else
 			memcpy(ndev->dev_addr, ndev->perm_addr,
Signature · 1 line
2.4.1
maintainer
#2 lore raw
Quoted 3 lines · Lennert Buytenhek

Refuse to allow setting an EUI-64 group address as an interface address, as those are not valid station addresses.

Read my other review, this will _maybe_ show that extended addr are always a unicast address (the full address range).

- Alex

#4 lore raw
In reply to Lennert Buytenhek

Hi Lennert,

Quoted 8 lines

Refuse to allow setting an EUI-64 group address as an interface address, as those are not valid station addresses.

---
include/linux/ieee802154.h | 10 ++++------
net/mac802154/iface.c | 4 ++--

2 files changed, 6 insertions(+), 8 deletions(-)

patch has been applied to bluetooth-next tree.

Regards

Marcel