bonding: fix lladdr finding and confirmation
Subsystems (4)
v2 cover bonding: fix lladdr finding and confirmation
This patch set fixed 3 issues when setting lladdr as bonding IPv6 target. Please see each patch for the details.
v2: separate the patch to 3 parts
Hangbin Liu (3):
bonding: use unspecified address if no available link local address
bonding: add all node mcast address when slave up
bonding: accept unsolicited NA message
drivers/net/bonding/bond_main.c | 20 +++++++++++++++----- net/ipv6/addrconf.c | 8 ++++++-- 2 files changed, 21 insertions(+), 7 deletions(-)
Signature · 1 line
2.37.1
PATCH v2 1/3 bonding: use unspecified address if no available link local address
When ns_ip6_target was set, the ipv6_dev_get_saddr() will be called to get available source address and send IPv6 neighbor solicit message.
If the target is global address, ipv6_dev_get_saddr() will get any available src address. But if the target is link local address, ipv6_dev_get_saddr() will only get available address from our interface, i.e. the corresponding bond interface.
But before bond interface up, all the address is tentative, while ipv6_dev_get_saddr() will ignore tentative address. This makes we can't find available link local src address, then bond_ns_send() will not be called and no NS message was sent. Finally bond interface will keep in down state.
Fix this by sending NS with unspecified address if there is no available source address.
--- drivers/net/bonding/bond_main.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/net/bonding/bond_main.c +3 -0
@@ -3167,6 +3167,9 @@ static void bond_ns_send_all(struct bonding *bond, struct slave *slave) found:if (!ipv6_dev_get_saddr (dev_net (dst->dev), dst->dev, &targets[i],0 , &saddr))bond_ns_send (slave, &targets[i], &saddr, tags); +else +bond_ns_send (slave, &targets[i], &in6addr_any, tags); +dst_release (dst);kfree (tags); }
Signature · 1 line
2.37.1
PATCH v2 2/3 bonding: add all node mcast address when slave up
In reply to Hangbin LiuWhen a link is enslave to bond, it need to set the interface down first. This makes the slave remove mac multicast address 33:33:00:00:00:01(The IPv6 multicast address ff02::1 is kept even when the interface down). When bond set the slave up, ipv6_mc_up() was not called due to commit c2edacf80e15 ("bonding / ipv6: no addrconf for slaves separately from master").
This is not an issue before we adding the lladdr target feature for bonding, as the mac multicast address will be added back when bond interface up and join group ff02::1.
But after adding lladdr target feature for bonding. When user set a lladdr target, the unsolicited NA message with all-nodes multicast dest will be dropped as the slave interface never add 33:33:00:00:00:01 back.
Fix this by calling ipv6_mc_up() to add 33:33:00:00:00:01 back when the slave interface up.
--- net/ipv6/addrconf.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
net/ipv6/addrconf.c +6 -2
@@ -3557,11 +3557,15 @@ static int addrconf_notify(struct notifier_block *this, unsigned long event, fallthrough;case NETDEV_UP :case NETDEV_CHANGE : -if (dev->flags &IFF_SLAVE ) +if (idev && idev->cnf.disable_ipv6)break ; -if (idev && idev->cnf.disable_ipv6) +if (dev->flags &IFF_SLAVE ) { +if (event ==NETDEV_UP && !IS_ERR_OR_NULL (idev) && + dev->flags &IFF_UP && dev->flags &IFF_MULTICAST ) +ipv6_mc_up (idev);break ; + }if (event ==NETDEV_UP ) {/* restore routes for permanent addresses */
Signature · 1 line
2.37.1
PATCH v2 3/3 bonding: accept unsolicited NA message
In reply to Hangbin LiuThe unsolicited NA message with all-nodes multicast dest address should be valid, as this also means the link could reach the target.
Also rename bond_validate_ns() to bond_validate_na().
--- drivers/net/bonding/bond_main.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-)
drivers/net/bonding/bond_main.c +12 -5
@@ -3201,12 +3201,19 @@ static bool bond_has_this_ip6(struct bonding *bond, struct in6_addr *addr)return ret; } -static void bond_validate_ns (struct bonding *bond,struct slave *slave, +static void bond_validate_na (struct bonding *bond,struct slave *slave,struct in6_addr *saddr,struct in6_addr *daddr) {int i; -if (ipv6_addr_any (saddr) || !bond_has_this_ip6 (bond, daddr)) { +/* Ignore NAs that: +* 1. Source address is unspecified address. +* 2. Dest address is neither all-nodes multicast address nor +* exist on bond interface. +*/ +if (ipv6_addr_any (saddr) || + (!ipv6_addr_equal (daddr, &in6addr_linklocal_allnodes) && + !bond_has_this_ip6 (bond, daddr))) {slave_dbg (bond->dev, slave->dev,"%s: sip %pI6c tip %pI6c not found\n" , __func__, saddr, daddr);return ;
@@ -3249,14 +3256,14 @@ static int bond_na_rcv(const struct sk_buff *skb, struct bonding *bond,* see bond_arp_rcv(). */ if (bond_is_active_slave (slave)) -bond_validate_ns (bond, slave, saddr, daddr); +bond_validate_na (bond, slave, saddr, daddr);else if (curr_active_slave &&time_after (slave_last_rx (bond, curr_active_slave), curr_active_slave->last_link_up)) -bond_validate_ns (bond, slave, saddr, daddr); +bond_validate_na (bond, slave, saddr, daddr);else if (curr_arp_slave &&bond_time_in_interval (bond,slave_last_tx (curr_arp_slave),1 )) -bond_validate_ns (bond, slave, saddr, daddr); +bond_validate_na (bond, slave, saddr, daddr); out:return RX_HANDLER_ANOTHER ;
Signature · 1 line
2.37.1
Jakub Kicinski
maintainerQuoted 2 lines · On Tue, 30 Aug 2022 17:37:19 +0800 Hangbin Liu wrote:
This patch set fixed 3 issues when setting lladdr as bonding IPv6 target. Please see each patch for the details.
Hi Jay, is this one still on your radar?
This series was applied to netdev/net.git (master) by David S. Miller <[email protected]>:
Hello:
This series was applied to netdev/net.git (master) by David S. Miller <[email protected]>:
Quoted 11 lines · On Tue, 30 Aug 2022 17:37:19 +0800 you wrote:
This patch set fixed 3 issues when setting lladdr as bonding IPv6 target. Please see each patch for the details.
v2: separate the patch to 3 parts
Hangbin Liu (3):
bonding: use unspecified address if no available link local address
bonding: add all node mcast address when slave up
bonding: accept unsolicited NA message[...]
Here is the summary with links:
- [PATCHv2,net,1/3] bonding: use unspecified address if no available link local address
https://git.kernel.org/netdev/net/c/b7f14132bf58 - [PATCHv2,net,2/3] bonding: add all node mcast address when slave up https://git.kernel.org/netdev/net/c/fd16eb948ea8 - [PATCHv2,net,3/3] bonding: accept unsolicited NA message https://git.kernel.org/netdev/net/c/592335a4164c
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html