From 4598380f9c548aa161eb4e990a1583f0a7d1e0d7 Mon Sep 17 00:00:00 2001 From: Hangbin Liu Date: Thu, 6 Apr 2023 16:23:50 +0800 Subject: 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") Reviewed-by: Jonathan Toppins Acked-by: Jay Vosburgh Signed-off-by: Hangbin Liu Signed-off-by: David S. Miller --- include/net/bonding.h | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'include/net/bonding.h') 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; } -- cgit v1.2.3-70-g09d2 From cb9e6e584d58420df182102674e636fb841dae4c Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Thu, 4 May 2023 11:52:49 +0200 Subject: bonding: add xdp_features support Introduce xdp_features support for bonding driver according to the slave devices attached to the master one. xdp_features is required whenever we want to xdp_redirect traffic into a bond device and then into selected slaves attached to it. Reviewed-by: Simon Horman Acked-by: Jay Vosburgh Fixes: 66c0e13ad236 ("drivers: net: turn on XDP features") Signed-off-by: Lorenzo Bianconi Reviewed-by: Jussi Maki Tested-by: Jussi Maki Signed-off-by: David S. Miller --- drivers/net/bonding/bond_main.c | 29 +++++++++++++++++++++++++++++ drivers/net/bonding/bond_options.c | 2 ++ include/net/bonding.h | 1 + 3 files changed, 32 insertions(+) (limited to 'include/net/bonding.h') diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 710548dbd0c1..3fed888629f7 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1789,6 +1789,26 @@ static void bond_ether_setup(struct net_device *bond_dev) bond_dev->priv_flags &= ~IFF_TX_SKB_SHARING; } +void bond_xdp_set_features(struct net_device *bond_dev) +{ + struct bonding *bond = netdev_priv(bond_dev); + xdp_features_t val = NETDEV_XDP_ACT_MASK; + struct list_head *iter; + struct slave *slave; + + ASSERT_RTNL(); + + if (!bond_xdp_check(bond)) { + xdp_clear_features_flag(bond_dev); + return; + } + + bond_for_each_slave(bond, slave, iter) + val &= slave->dev->xdp_features; + + xdp_set_features_flag(bond_dev, val); +} + /* enslave device to bond device */ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, struct netlink_ext_ack *extack) @@ -2236,6 +2256,8 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, bpf_prog_inc(bond->xdp_prog); } + bond_xdp_set_features(bond_dev); + slave_info(bond_dev, slave_dev, "Enslaving as %s interface with %s link\n", bond_is_active_slave(new_slave) ? "an active" : "a backup", new_slave->link != BOND_LINK_DOWN ? "an up" : "a down"); @@ -2483,6 +2505,7 @@ static int __bond_release_one(struct net_device *bond_dev, if (!netif_is_bond_master(slave_dev)) slave_dev->priv_flags &= ~IFF_BONDING; + bond_xdp_set_features(bond_dev); kobject_put(&slave->kobj); return 0; @@ -3930,6 +3953,9 @@ static int bond_slave_netdev_event(unsigned long event, /* Propagate to master device */ call_netdevice_notifiers(event, slave->bond->dev); break; + case NETDEV_XDP_FEAT_CHANGE: + bond_xdp_set_features(bond_dev); + break; default: break; } @@ -5874,6 +5900,9 @@ void bond_setup(struct net_device *bond_dev) if (BOND_MODE(bond) == BOND_MODE_ACTIVEBACKUP) bond_dev->features |= BOND_XFRM_FEATURES; #endif /* CONFIG_XFRM_OFFLOAD */ + + if (bond_xdp_check(bond)) + bond_dev->xdp_features = NETDEV_XDP_ACT_MASK; } /* Destroy a bonding device. diff --git a/drivers/net/bonding/bond_options.c b/drivers/net/bonding/bond_options.c index f71d5517f829..0498fc6731f8 100644 --- a/drivers/net/bonding/bond_options.c +++ b/drivers/net/bonding/bond_options.c @@ -877,6 +877,8 @@ static int bond_option_mode_set(struct bonding *bond, netdev_update_features(bond->dev); } + bond_xdp_set_features(bond->dev); + return 0; } diff --git a/include/net/bonding.h b/include/net/bonding.h index c3843239517d..a60a24923b55 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -659,6 +659,7 @@ void bond_destroy_sysfs(struct bond_net *net); void bond_prepare_sysfs_group(struct bonding *bond); int bond_sysfs_slave_add(struct slave *slave); void bond_sysfs_slave_del(struct slave *slave); +void bond_xdp_set_features(struct net_device *bond_dev); int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, struct netlink_ext_ack *extack); int bond_release(struct net_device *bond_dev, struct net_device *slave_dev); -- cgit v1.2.3-70-g09d2 From 9949e2efb54eb3001cb2f6512ff3166dddbfb75d Mon Sep 17 00:00:00 2001 From: Hangbin Liu Date: Tue, 9 May 2023 11:11:57 +0800 Subject: bonding: fix send_peer_notif overflow Bonding send_peer_notif was defined as u8. Since commit 07a4ddec3ce9 ("bonding: add an option to specify a delay between peer notifications"). the bond->send_peer_notif will be num_peer_notif multiplied by peer_notif_delay, which is u8 * u32. This would cause the send_peer_notif overflow easily. e.g. ip link add bond0 type bond mode 1 miimon 100 num_grat_arp 30 peer_notify_delay 1000 To fix the overflow, let's set the send_peer_notif to u32 and limit peer_notif_delay to 300s. Reported-by: Liang Li Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2090053 Fixes: 07a4ddec3ce9 ("bonding: add an option to specify a delay between peer notifications") Signed-off-by: Hangbin Liu Signed-off-by: David S. Miller --- drivers/net/bonding/bond_netlink.c | 7 ++++++- drivers/net/bonding/bond_options.c | 8 +++++++- include/net/bonding.h | 2 +- 3 files changed, 14 insertions(+), 3 deletions(-) (limited to 'include/net/bonding.h') diff --git a/drivers/net/bonding/bond_netlink.c b/drivers/net/bonding/bond_netlink.c index c2d080fc4fc4..27cbe148f0db 100644 --- a/drivers/net/bonding/bond_netlink.c +++ b/drivers/net/bonding/bond_netlink.c @@ -84,6 +84,11 @@ nla_put_failure: return -EMSGSIZE; } +/* Limit the max delay range to 300s */ +static struct netlink_range_validation delay_range = { + .max = 300000, +}; + static const struct nla_policy bond_policy[IFLA_BOND_MAX + 1] = { [IFLA_BOND_MODE] = { .type = NLA_U8 }, [IFLA_BOND_ACTIVE_SLAVE] = { .type = NLA_U32 }, @@ -114,7 +119,7 @@ static const struct nla_policy bond_policy[IFLA_BOND_MAX + 1] = { [IFLA_BOND_AD_ACTOR_SYSTEM] = { .type = NLA_BINARY, .len = ETH_ALEN }, [IFLA_BOND_TLB_DYNAMIC_LB] = { .type = NLA_U8 }, - [IFLA_BOND_PEER_NOTIF_DELAY] = { .type = NLA_U32 }, + [IFLA_BOND_PEER_NOTIF_DELAY] = NLA_POLICY_FULL_RANGE(NLA_U32, &delay_range), [IFLA_BOND_MISSED_MAX] = { .type = NLA_U8 }, [IFLA_BOND_NS_IP6_TARGET] = { .type = NLA_NESTED }, }; diff --git a/drivers/net/bonding/bond_options.c b/drivers/net/bonding/bond_options.c index 0498fc6731f8..f3f27f0bd2a6 100644 --- a/drivers/net/bonding/bond_options.c +++ b/drivers/net/bonding/bond_options.c @@ -169,6 +169,12 @@ static const struct bond_opt_value bond_num_peer_notif_tbl[] = { { NULL, -1, 0} }; +static const struct bond_opt_value bond_peer_notif_delay_tbl[] = { + { "off", 0, 0}, + { "maxval", 300000, BOND_VALFLAG_MAX}, + { NULL, -1, 0} +}; + static const struct bond_opt_value bond_primary_reselect_tbl[] = { { "always", BOND_PRI_RESELECT_ALWAYS, BOND_VALFLAG_DEFAULT}, { "better", BOND_PRI_RESELECT_BETTER, 0}, @@ -488,7 +494,7 @@ static const struct bond_option bond_opts[BOND_OPT_LAST] = { .id = BOND_OPT_PEER_NOTIF_DELAY, .name = "peer_notif_delay", .desc = "Delay between each peer notification on failover event, in milliseconds", - .values = bond_intmax_tbl, + .values = bond_peer_notif_delay_tbl, .set = bond_option_peer_notif_delay_set } }; diff --git a/include/net/bonding.h b/include/net/bonding.h index a60a24923b55..0efef2a952b7 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -233,7 +233,7 @@ struct bonding { */ spinlock_t mode_lock; spinlock_t stats_lock; - u8 send_peer_notif; + u32 send_peer_notif; u8 igmp_retrans; #ifdef CONFIG_PROC_FS struct proc_dir_entry *proc_entry; -- cgit v1.2.3-70-g09d2 From ae9b15fbe63447bc1d3bba3769f409d17ca6fdf6 Mon Sep 17 00:00:00 2001 From: Taehee Yoo Date: Wed, 17 May 2023 14:30:10 +0000 Subject: net: fix stack overflow when LRO is disabled for virtual interfaces When the virtual interface's feature is updated, it synchronizes the updated feature for its own lower interface. This propagation logic should be worked as the iteration, not recursively. But it works recursively due to the netdev notification unexpectedly. This problem occurs when it disables LRO only for the team and bonding interface type. team0 | +------+------+-----+-----+ | | | | | team1 team2 team3 ... team200 If team0's LRO feature is updated, it generates the NETDEV_FEAT_CHANGE event to its own lower interfaces(team1 ~ team200). It is worked by netdev_sync_lower_features(). So, the NETDEV_FEAT_CHANGE notification logic of each lower interface work iteratively. But generated NETDEV_FEAT_CHANGE event is also sent to the upper interface too. upper interface(team0) generates the NETDEV_FEAT_CHANGE event for its own lower interfaces again. lower and upper interfaces receive this event and generate this event again and again. So, the stack overflow occurs. But it is not the infinite loop issue. Because the netdev_sync_lower_features() updates features before generating the NETDEV_FEAT_CHANGE event. Already synchronized lower interfaces skip notification logic. So, it is just the problem that iteration logic is changed to the recursive unexpectedly due to the notification mechanism. Reproducer: ip link add team0 type team ethtool -K team0 lro on for i in {1..200} do ip link add team$i master team0 type team ethtool -K team$i lro on done ethtool -K team0 lro off In order to fix it, the notifier_ctx member of bonding/team is introduced. Reported-by: syzbot+60748c96cf5c6df8e581@syzkaller.appspotmail.com Fixes: fd867d51f889 ("net/core: generic support for disabling netdev features down stack") Signed-off-by: Taehee Yoo Reviewed-by: Eric Dumazet Reviewed-by: Nikolay Aleksandrov Link: https://lore.kernel.org/r/20230517143010.3596250-1-ap420073@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/bonding/bond_main.c | 8 +++++++- drivers/net/team/team.c | 7 ++++++- include/linux/if_team.h | 1 + include/net/bonding.h | 1 + 4 files changed, 15 insertions(+), 2 deletions(-) (limited to 'include/net/bonding.h') diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 3fed888629f7..edbaa1444f8e 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -3947,7 +3947,11 @@ static int bond_slave_netdev_event(unsigned long event, unblock_netpoll_tx(); break; case NETDEV_FEAT_CHANGE: - bond_compute_features(bond); + if (!bond->notifier_ctx) { + bond->notifier_ctx = true; + bond_compute_features(bond); + bond->notifier_ctx = false; + } break; case NETDEV_RESEND_IGMP: /* Propagate to master device */ @@ -6342,6 +6346,8 @@ static int bond_init(struct net_device *bond_dev) if (!bond->wq) return -ENOMEM; + bond->notifier_ctx = false; + spin_lock_init(&bond->stats_lock); netdev_lockdep_set_classes(bond_dev); diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c index d10606f257c4..555b0b1e9a78 100644 --- a/drivers/net/team/team.c +++ b/drivers/net/team/team.c @@ -1629,6 +1629,7 @@ static int team_init(struct net_device *dev) team->dev = dev; team_set_no_mode(team); + team->notifier_ctx = false; team->pcpu_stats = netdev_alloc_pcpu_stats(struct team_pcpu_stats); if (!team->pcpu_stats) @@ -3022,7 +3023,11 @@ static int team_device_event(struct notifier_block *unused, team_del_slave(port->team->dev, dev); break; case NETDEV_FEAT_CHANGE: - team_compute_features(port->team); + if (!port->team->notifier_ctx) { + port->team->notifier_ctx = true; + team_compute_features(port->team); + port->team->notifier_ctx = false; + } break; case NETDEV_PRECHANGEMTU: /* Forbid to change mtu of underlaying device */ diff --git a/include/linux/if_team.h b/include/linux/if_team.h index fc985e5c739d..8de6b6e67829 100644 --- a/include/linux/if_team.h +++ b/include/linux/if_team.h @@ -208,6 +208,7 @@ struct team { bool queue_override_enabled; struct list_head *qom_lists; /* array of queue override mapping lists */ bool port_mtu_change_allowed; + bool notifier_ctx; struct { unsigned int count; unsigned int interval; /* in ms */ diff --git a/include/net/bonding.h b/include/net/bonding.h index 0efef2a952b7..59955ac33157 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -221,6 +221,7 @@ struct bonding { struct bond_up_slave __rcu *usable_slaves; struct bond_up_slave __rcu *all_slaves; bool force_primary; + bool notifier_ctx; s32 slave_cnt; /* never change this value outside the attach/detach wrappers */ int (*recv_probe)(const struct sk_buff *, struct bonding *, struct slave *); -- cgit v1.2.3-70-g09d2