[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20230329101859.3458449-2-liuhangbin@gmail.com>
Date: Wed, 29 Mar 2023 18:18:57 +0800
From: Hangbin Liu <liuhangbin@...il.com>
To: netdev@...r.kernel.org
Cc: Jay Vosburgh <j.vosburgh@...il.com>,
"David S . Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Jonathan Toppins <jtoppins@...hat.com>,
Paolo Abeni <pabeni@...hat.com>,
Eric Dumazet <edumazet@...gle.com>,
Liang Li <liali@...hat.com>, Hangbin Liu <liuhangbin@...il.com>
Subject: [PATCH net 1/3] bonding: fix ns validation on backup slaves
When arp_validate is set to 2, 3, or 6, validation is performed for
backup slaves as well. As stated in the bond documentation, validation
involves checking the broadcast ARP request sent out via the active
slave. This helps determine which slaves are more likely to function in
the event of an active slave failure.
However, when the target is an IPv6 address, the NS message sent from
the active interface is not checked on backup slaves. Additionally,
based on the bond_arp_rcv() rule b, we must reverse the saddr and daddr
when checking the NS message.
Note that when checking the NS message, the destination address is a
multicast address. Therefore, we must convert the target address to
solicited multicast in the bond_get_targets_ip6() function.
Prior to the fix, the backup slaves had a mii status of "down", but
after the fix, all of the slaves' mii status was updated to "UP".
Fixes: 4e24be018eb9 ("bonding: add new parameter ns_targets")
Signed-off-by: Hangbin Liu <liuhangbin@...il.com>
---
drivers/net/bonding/bond_main.c | 5 +++--
include/net/bonding.h | 8 ++++++--
2 files changed, 9 insertions(+), 4 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 236e5219c811..8cc9a74789b7 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -3269,7 +3269,8 @@ static int bond_na_rcv(const struct sk_buff *skb, struct bonding *bond,
combined = skb_header_pointer(skb, 0, sizeof(_combined), &_combined);
if (!combined || combined->ip6.nexthdr != NEXTHDR_ICMP ||
- combined->icmp6.icmp6_type != NDISC_NEIGHBOUR_ADVERTISEMENT)
+ (combined->icmp6.icmp6_type != NDISC_NEIGHBOUR_SOLICITATION &&
+ combined->icmp6.icmp6_type != NDISC_NEIGHBOUR_ADVERTISEMENT))
goto out;
saddr = &combined->ip6.saddr;
@@ -3291,7 +3292,7 @@ static int bond_na_rcv(const struct sk_buff *skb, struct bonding *bond,
else if (curr_active_slave &&
time_after(slave_last_rx(bond, curr_active_slave),
curr_active_slave->last_link_up))
- bond_validate_na(bond, slave, saddr, daddr);
+ bond_validate_na(bond, slave, daddr, saddr);
else if (curr_arp_slave &&
bond_time_in_interval(bond, slave_last_tx(curr_arp_slave), 1))
bond_validate_na(bond, slave, saddr, daddr);
diff --git a/include/net/bonding.h b/include/net/bonding.h
index ea36ab7f9e72..c3843239517d 100644
--- a/include/net/bonding.h
+++ b/include/net/bonding.h
@@ -761,13 +761,17 @@ static inline int bond_get_targets_ip(__be32 *targets, __be32 ip)
#if IS_ENABLED(CONFIG_IPV6)
static inline int bond_get_targets_ip6(struct in6_addr *targets, struct in6_addr *ip)
{
+ struct in6_addr mcaddr;
int i;
- for (i = 0; i < BOND_MAX_NS_TARGETS; i++)
- if (ipv6_addr_equal(&targets[i], ip))
+ for (i = 0; i < BOND_MAX_NS_TARGETS; i++) {
+ addrconf_addr_solict_mult(&targets[i], &mcaddr);
+ if ((ipv6_addr_equal(&targets[i], ip)) ||
+ (ipv6_addr_equal(&mcaddr, ip)))
return i;
else if (ipv6_addr_any(&targets[i]))
break;
+ }
return -1;
}
--
2.38.1
Powered by blists - more mailing lists