From 096361b15577a583afcc28179a08c75cf95e9dae Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:25 +0000
Subject: [PATCH 01/15] ipv6: add ipv6_devconf_read_txrx cacheline_group

IPv6 TX and RX fast path use the following fields:

- disable_ipv6
- hop_limit
- mtu6
- forwarding
- disable_policy
- proxy_ndp

Place them in a group to increase data locality.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/ipv6.h | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index ef3aa060a289e..383a0ea2ab913 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -3,6 +3,7 @@
 #define _IPV6_H
 
 #include <uapi/linux/ipv6.h>
+#include <linux/cache.h>
 
 #define ipv6_optlen(p)  (((p)->hdrlen+1) << 3)
 #define ipv6_authlen(p) (((p)->hdrlen+2) << 2)
@@ -10,9 +11,16 @@
  * This structure contains configuration options per IPv6 link.
  */
 struct ipv6_devconf {
-	__s32		forwarding;
+	/* RX & TX fastpath fields. */
+	__cacheline_group_begin(ipv6_devconf_read_txrx);
+	__s32		disable_ipv6;
 	__s32		hop_limit;
 	__s32		mtu6;
+	__s32		forwarding;
+	__s32		disable_policy;
+	__s32		proxy_ndp;
+	__cacheline_group_end(ipv6_devconf_read_txrx);
+
 	__s32		accept_ra;
 	__s32		accept_redirects;
 	__s32		autoconf;
@@ -45,7 +53,6 @@ struct ipv6_devconf {
 	__s32		accept_ra_rt_info_max_plen;
 #endif
 #endif
-	__s32		proxy_ndp;
 	__s32		accept_source_route;
 	__s32		accept_ra_from_local;
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
@@ -55,7 +62,6 @@ struct ipv6_devconf {
 #ifdef CONFIG_IPV6_MROUTE
 	atomic_t	mc_forwarding;
 #endif
-	__s32		disable_ipv6;
 	__s32		drop_unicast_in_l2_multicast;
 	__s32		accept_dad;
 	__s32		force_tllao;
@@ -76,7 +82,6 @@ struct ipv6_devconf {
 #endif
 	__u32		enhanced_dad;
 	__u32		addr_gen_mode;
-	__s32		disable_policy;
 	__s32           ndisc_tclass;
 	__s32		rpl_seg_enabled;
 	__u32		ioam6_id;

From d289ab65b89c1d4d88417cb6c03e923f21f95fae Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:26 +0000
Subject: [PATCH 02/15] ipv6: annotate data-races around cnf.disable_ipv6

disable_ipv6 is read locklessly, add appropriate READ_ONCE()
and WRITE_ONCE() annotations.

v2: do not preload net before rtnl_trylock() in
    addrconf_disable_ipv6() (Jiri)

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c   | 9 +++++----
 net/ipv6/ip6_input.c  | 4 ++--
 net/ipv6/ip6_output.c | 2 +-
 3 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 9e949403c1363..c965400fb9340 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -4214,7 +4214,7 @@ static void addrconf_dad_work(struct work_struct *w)
 			if (!ipv6_generate_eui64(addr.s6_addr + 8, idev->dev) &&
 			    ipv6_addr_equal(&ifp->addr, &addr)) {
 				/* DAD failed for link-local based on MAC */
-				idev->cnf.disable_ipv6 = 1;
+				WRITE_ONCE(idev->cnf.disable_ipv6, 1);
 
 				pr_info("%s: IPv6 being disabled!\n",
 					ifp->idev->dev->name);
@@ -6389,7 +6389,8 @@ static void addrconf_disable_change(struct net *net, __s32 newf)
 		idev = __in6_dev_get(dev);
 		if (idev) {
 			int changed = (!idev->cnf.disable_ipv6) ^ (!newf);
-			idev->cnf.disable_ipv6 = newf;
+
+			WRITE_ONCE(idev->cnf.disable_ipv6, newf);
 			if (changed)
 				dev_disable_change(idev);
 		}
@@ -6406,7 +6407,7 @@ static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
 
 	net = (struct net *)table->extra2;
 	old = *p;
-	*p = newf;
+	WRITE_ONCE(*p, newf);
 
 	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
 		rtnl_unlock();
@@ -6414,7 +6415,7 @@ static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
 	}
 
 	if (p == &net->ipv6.devconf_all->disable_ipv6) {
-		net->ipv6.devconf_dflt->disable_ipv6 = newf;
+		WRITE_ONCE(net->ipv6.devconf_dflt->disable_ipv6, newf);
 		addrconf_disable_change(net, newf);
 	} else if ((!newf) ^ (!old))
 		dev_disable_change((struct inet6_dev *)table->extra1);
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index b8378814532ce..1ba97933c74fb 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -168,9 +168,9 @@ static struct sk_buff *ip6_rcv_core(struct sk_buff *skb, struct net_device *dev,
 
 	SKB_DR_SET(reason, NOT_SPECIFIED);
 	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL ||
-	    !idev || unlikely(idev->cnf.disable_ipv6)) {
+	    !idev || unlikely(READ_ONCE(idev->cnf.disable_ipv6))) {
 		__IP6_INC_STATS(net, idev, IPSTATS_MIB_INDISCARDS);
-		if (idev && unlikely(idev->cnf.disable_ipv6))
+		if (idev && unlikely(READ_ONCE(idev->cnf.disable_ipv6)))
 			SKB_DR_SET(reason, IPV6DISABLED);
 		goto drop;
 	}
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 31b86fe661aa6..0559bd0005858 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -234,7 +234,7 @@ int ip6_output(struct net *net, struct sock *sk, struct sk_buff *skb)
 	skb->protocol = htons(ETH_P_IPV6);
 	skb->dev = dev;
 
-	if (unlikely(idev->cnf.disable_ipv6)) {
+	if (unlikely(READ_ONCE(idev->cnf.disable_ipv6))) {
 		IP6_INC_STATS(net, idev, IPSTATS_MIB_OUTDISCARDS);
 		kfree_skb_reason(skb, SKB_DROP_REASON_IPV6DISABLED);
 		return 0;

From 553ced03b227c99f0bf68d86f7b45ed7125563f2 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:27 +0000
Subject: [PATCH 03/15] ipv6: addrconf_disable_ipv6() optimization

Writing over /proc/sys/net/ipv6/conf/default/disable_ipv6
does not need to hold RTNL.

v3: remove a wrong change (Jakub Kicinski feedback)

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index c965400fb9340..6975685f55ea9 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -6399,21 +6399,20 @@ static void addrconf_disable_change(struct net *net, __s32 newf)
 
 static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf)
 {
-	struct net *net;
+	struct net *net = (struct net *)table->extra2;
 	int old;
 
+	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
+		WRITE_ONCE(*p, newf);
+		return 0;
+	}
+
 	if (!rtnl_trylock())
 		return restart_syscall();
 
-	net = (struct net *)table->extra2;
 	old = *p;
 	WRITE_ONCE(*p, newf);
 
-	if (p == &net->ipv6.devconf_dflt->disable_ipv6) {
-		rtnl_unlock();
-		return 0;
-	}
-
 	if (p == &net->ipv6.devconf_all->disable_ipv6) {
 		WRITE_ONCE(net->ipv6.devconf_dflt->disable_ipv6, newf);
 		addrconf_disable_change(net, newf);

From e7135f484994494a38071e3653a83d21d305f50c Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:28 +0000
Subject: [PATCH 04/15] ipv6: annotate data-races around cnf.mtu6

idev->cnf.mtu6 might be read locklessly, add appropriate READ_ONCE()
and WRITE_ONCE() annotations.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/ip6_route.h | 2 +-
 net/ipv6/addrconf.c     | 4 ++--
 net/ipv6/ndisc.c        | 4 ++--
 net/ipv6/route.c        | 6 +++---
 4 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 52a51c69aa9de..a30c6aa9e5cf3 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -332,7 +332,7 @@ static inline unsigned int ip6_dst_mtu_maybe_forward(const struct dst_entry *dst
 	rcu_read_lock();
 	idev = __in6_dev_get(dst->dev);
 	if (idev)
-		mtu = idev->cnf.mtu6;
+		mtu = READ_ONCE(idev->cnf.mtu6);
 	rcu_read_unlock();
 
 out:
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 6975685f55ea9..1b534edc7ca2c 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -3671,7 +3671,7 @@ static int addrconf_notify(struct notifier_block *this, unsigned long event,
 
 		if (idev) {
 			rt6_mtu_change(dev, dev->mtu);
-			idev->cnf.mtu6 = dev->mtu;
+			WRITE_ONCE(idev->cnf.mtu6, dev->mtu);
 			break;
 		}
 
@@ -3763,7 +3763,7 @@ static int addrconf_notify(struct notifier_block *this, unsigned long event,
 			if (idev->cnf.mtu6 != dev->mtu &&
 			    dev->mtu >= IPV6_MIN_MTU) {
 				rt6_mtu_change(dev, dev->mtu);
-				idev->cnf.mtu6 = dev->mtu;
+				WRITE_ONCE(idev->cnf.mtu6, dev->mtu);
 			}
 			WRITE_ONCE(idev->tstamp, jiffies);
 			inet6_ifinfo_notify(RTM_NEWLINK, idev);
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 8523f0595b018..e96d79cd34d27 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1578,8 +1578,8 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 
 		if (mtu < IPV6_MIN_MTU || mtu > skb->dev->mtu) {
 			ND_PRINTK(2, warn, "RA: invalid mtu: %d\n", mtu);
-		} else if (in6_dev->cnf.mtu6 != mtu) {
-			in6_dev->cnf.mtu6 = mtu;
+		} else if (READ_ONCE(in6_dev->cnf.mtu6) != mtu) {
+			WRITE_ONCE(in6_dev->cnf.mtu6, mtu);
 			fib6_metric_set(rt, RTAX_MTU, mtu);
 			rt6_mtu_change(skb->dev, mtu);
 		}
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 707d65bc9c0e5..66c685b0b6199 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1596,7 +1596,7 @@ static unsigned int fib6_mtu(const struct fib6_result *res)
 
 		rcu_read_lock();
 		idev = __in6_dev_get(dev);
-		mtu = idev->cnf.mtu6;
+		mtu = READ_ONCE(idev->cnf.mtu6);
 		rcu_read_unlock();
 	}
 
@@ -3249,8 +3249,8 @@ u32 ip6_mtu_from_fib6(const struct fib6_result *res,
 
 		mtu = IPV6_MIN_MTU;
 		idev = __in6_dev_get(dev);
-		if (idev && idev->cnf.mtu6 > mtu)
-			mtu = idev->cnf.mtu6;
+		if (idev)
+			mtu = max_t(u32, mtu, READ_ONCE(idev->cnf.mtu6));
 	}
 
 	mtu = min_t(unsigned int, mtu, IP6_MAX_MTU);

From e0bb2675fea2783c45bb95d74f00c55156720863 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:29 +0000
Subject: [PATCH 05/15] ipv6: annotate data-races around cnf.hop_limit

idev->cnf.hop_limit and net->ipv6.devconf_all->hop_limit
might be read locklessly, add appropriate READ_ONCE()
and WRITE_ONCE() annotations.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Florian Westphal <fw@strlen.de> # for netfilter parts
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/ethernet/netronome/nfp/flower/action.c | 2 +-
 net/ipv6/ipv6_sockglue.c                           | 2 +-
 net/ipv6/ndisc.c                                   | 2 +-
 net/ipv6/netfilter/nf_reject_ipv6.c                | 4 ++--
 net/ipv6/output_core.c                             | 4 ++--
 net/netfilter/nf_synproxy_core.c                   | 2 +-
 6 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/netronome/nfp/flower/action.c b/drivers/net/ethernet/netronome/nfp/flower/action.c
index 2b383d92d7f57..2c3f629079584 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/action.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/action.c
@@ -460,7 +460,7 @@ nfp_fl_set_tun(struct nfp_app *app, struct nfp_fl_set_tun *set_tun,
 			set_tun->ttl = ip6_dst_hoplimit(dst);
 			dst_release(dst);
 		} else {
-			set_tun->ttl = net->ipv6.devconf_all->hop_limit;
+			set_tun->ttl = READ_ONCE(net->ipv6.devconf_all->hop_limit);
 		}
 #endif
 	} else {
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index 56c3c467f9deb..f61d977ac0528 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -1346,7 +1346,7 @@ int do_ipv6_getsockopt(struct sock *sk, int level, int optname,
 		}
 
 		if (val < 0)
-			val = sock_net(sk)->ipv6.devconf_all->hop_limit;
+			val = READ_ONCE(sock_net(sk)->ipv6.devconf_all->hop_limit);
 		break;
 	}
 
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index e96d79cd34d27..9c9c31268432e 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1423,7 +1423,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	if (in6_dev->cnf.accept_ra_min_hop_limit < 256 &&
 	    ra_msg->icmph.icmp6_hop_limit) {
 		if (in6_dev->cnf.accept_ra_min_hop_limit <= ra_msg->icmph.icmp6_hop_limit) {
-			in6_dev->cnf.hop_limit = ra_msg->icmph.icmp6_hop_limit;
+			WRITE_ONCE(in6_dev->cnf.hop_limit, ra_msg->icmph.icmp6_hop_limit);
 			fib6_metric_set(rt, RTAX_HOPLIMIT,
 					ra_msg->icmph.icmp6_hop_limit);
 		} else {
diff --git a/net/ipv6/netfilter/nf_reject_ipv6.c b/net/ipv6/netfilter/nf_reject_ipv6.c
index 196dd4ecb5e21..dedee264b8f6c 100644
--- a/net/ipv6/netfilter/nf_reject_ipv6.c
+++ b/net/ipv6/netfilter/nf_reject_ipv6.c
@@ -83,7 +83,7 @@ struct sk_buff *nf_reject_skb_v6_tcp_reset(struct net *net,
 
 	skb_reserve(nskb, LL_MAX_HEADER);
 	nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_TCP,
-				     net->ipv6.devconf_all->hop_limit);
+				     READ_ONCE(net->ipv6.devconf_all->hop_limit));
 	nf_reject_ip6_tcphdr_put(nskb, oldskb, oth, otcplen);
 	nip6h->payload_len = htons(nskb->len - sizeof(struct ipv6hdr));
 
@@ -124,7 +124,7 @@ struct sk_buff *nf_reject_skb_v6_unreach(struct net *net,
 
 	skb_reserve(nskb, LL_MAX_HEADER);
 	nip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_ICMPV6,
-				     net->ipv6.devconf_all->hop_limit);
+				     READ_ONCE(net->ipv6.devconf_all->hop_limit));
 
 	skb_reset_transport_header(nskb);
 	icmp6h = skb_put_zero(nskb, sizeof(struct icmp6hdr));
diff --git a/net/ipv6/output_core.c b/net/ipv6/output_core.c
index b5205311f372b..806d4b5dd1e60 100644
--- a/net/ipv6/output_core.c
+++ b/net/ipv6/output_core.c
@@ -111,9 +111,9 @@ int ip6_dst_hoplimit(struct dst_entry *dst)
 		rcu_read_lock();
 		idev = __in6_dev_get(dev);
 		if (idev)
-			hoplimit = idev->cnf.hop_limit;
+			hoplimit = READ_ONCE(idev->cnf.hop_limit);
 		else
-			hoplimit = dev_net(dev)->ipv6.devconf_all->hop_limit;
+			hoplimit = READ_ONCE(dev_net(dev)->ipv6.devconf_all->hop_limit);
 		rcu_read_unlock();
 	}
 	return hoplimit;
diff --git a/net/netfilter/nf_synproxy_core.c b/net/netfilter/nf_synproxy_core.c
index fbbc4fd373495..5b140c12b7dfa 100644
--- a/net/netfilter/nf_synproxy_core.c
+++ b/net/netfilter/nf_synproxy_core.c
@@ -800,7 +800,7 @@ synproxy_build_ip_ipv6(struct net *net, struct sk_buff *skb,
 	skb_reset_network_header(skb);
 	iph = skb_put(skb, sizeof(*iph));
 	ip6_flow_hdr(iph, 0, 0);
-	iph->hop_limit	= net->ipv6.devconf_all->hop_limit;
+	iph->hop_limit	= READ_ONCE(net->ipv6.devconf_all->hop_limit);
 	iph->nexthdr	= IPPROTO_TCP;
 	iph->saddr	= *saddr;
 	iph->daddr	= *daddr;

From 32f754176e889cdfe989ef08ece19859427755df Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:30 +0000
Subject: [PATCH 06/15] ipv6: annotate data-races around cnf.forwarding

idev->cnf.forwarding and net->ipv6.devconf_all->forwarding
might be read locklessly, add appropriate READ_ONCE()
and WRITE_ONCE() annotations.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/usb/cdc_mbim.c |  2 +-
 include/net/ipv6.h         |  8 +++++---
 net/core/filter.c          |  2 +-
 net/ipv6/addrconf.c        | 10 ++++++----
 net/ipv6/ip6_output.c      |  2 +-
 net/ipv6/ndisc.c           | 11 ++++++-----
 net/ipv6/route.c           |  4 ++--
 7 files changed, 22 insertions(+), 17 deletions(-)

diff --git a/drivers/net/usb/cdc_mbim.c b/drivers/net/usb/cdc_mbim.c
index cd4083e0b3b9e..e13e4920ee9b2 100644
--- a/drivers/net/usb/cdc_mbim.c
+++ b/drivers/net/usb/cdc_mbim.c
@@ -339,7 +339,7 @@ static void do_neigh_solicit(struct usbnet *dev, u8 *buf, u16 tci)
 	in6_dev = in6_dev_get(netdev);
 	if (!in6_dev)
 		goto out;
-	is_router = !!in6_dev->cnf.forwarding;
+	is_router = !!READ_ONCE(in6_dev->cnf.forwarding);
 	in6_dev_put(in6_dev);
 
 	/* ipv6_stub != NULL if in6_dev_get returned an inet6_dev */
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index cf25ea21d770d..88a8e554f7a12 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -534,13 +534,15 @@ static inline int ipv6_hopopt_jumbo_remove(struct sk_buff *skb)
 	return 0;
 }
 
-static inline bool ipv6_accept_ra(struct inet6_dev *idev)
+static inline bool ipv6_accept_ra(const struct inet6_dev *idev)
 {
+	s32 accept_ra = READ_ONCE(idev->cnf.accept_ra);
+
 	/* If forwarding is enabled, RA are not accepted unless the special
 	 * hybrid mode (accept_ra=2) is enabled.
 	 */
-	return idev->cnf.forwarding ? idev->cnf.accept_ra == 2 :
-	    idev->cnf.accept_ra;
+	return READ_ONCE(idev->cnf.forwarding) ? accept_ra == 2 :
+		accept_ra;
 }
 
 #define IPV6_FRAG_HIGH_THRESH	(4 * 1024*1024)	/* 4194304 */
diff --git a/net/core/filter.c b/net/core/filter.c
index 358870408a51e..58e8e1a70aa75 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -5988,7 +5988,7 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params,
 		return -ENODEV;
 
 	idev = __in6_dev_get_safely(dev);
-	if (unlikely(!idev || !idev->cnf.forwarding))
+	if (unlikely(!idev || !READ_ONCE(idev->cnf.forwarding)))
 		return BPF_FIB_LKUP_RET_FWD_DISABLED;
 
 	if (flags & BPF_FIB_LOOKUP_OUTPUT) {
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 1b534edc7ca2c..ec99e393f6a91 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -551,7 +551,8 @@ static int inet6_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
 		goto out;
 
 	if ((all || type == NETCONFA_FORWARDING) &&
-	    nla_put_s32(skb, NETCONFA_FORWARDING, devconf->forwarding) < 0)
+	    nla_put_s32(skb, NETCONFA_FORWARDING,
+			READ_ONCE(devconf->forwarding)) < 0)
 		goto nla_put_failure;
 #ifdef CONFIG_IPV6_MROUTE
 	if ((all || type == NETCONFA_MC_FORWARDING) &&
@@ -869,7 +870,8 @@ static void addrconf_forward_change(struct net *net, __s32 newf)
 		idev = __in6_dev_get(dev);
 		if (idev) {
 			int changed = (!idev->cnf.forwarding) ^ (!newf);
-			idev->cnf.forwarding = newf;
+
+			WRITE_ONCE(idev->cnf.forwarding, newf);
 			if (changed)
 				dev_forward_change(idev);
 		}
@@ -886,7 +888,7 @@ static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
 
 	net = (struct net *)table->extra2;
 	old = *p;
-	*p = newf;
+	WRITE_ONCE(*p, newf);
 
 	if (p == &net->ipv6.devconf_dflt->forwarding) {
 		if ((!newf) ^ (!old))
@@ -901,7 +903,7 @@ static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf)
 	if (p == &net->ipv6.devconf_all->forwarding) {
 		int old_dflt = net->ipv6.devconf_dflt->forwarding;
 
-		net->ipv6.devconf_dflt->forwarding = newf;
+		WRITE_ONCE(net->ipv6.devconf_dflt->forwarding, newf);
 		if ((!newf) ^ (!old_dflt))
 			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
 						     NETCONFA_FORWARDING,
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 0559bd0005858..444be8c84cc57 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -501,7 +501,7 @@ int ip6_forward(struct sk_buff *skb)
 	u32 mtu;
 
 	idev = __in6_dev_get_safely(dev_get_by_index_rcu(net, IP6CB(skb)->iif));
-	if (net->ipv6.devconf_all->forwarding == 0)
+	if (READ_ONCE(net->ipv6.devconf_all->forwarding) == 0)
 		goto error;
 
 	if (skb->pkt_type != PACKET_HOST)
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 9c9c31268432e..1fb5e37bc78be 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -903,7 +903,7 @@ static enum skb_drop_reason ndisc_recv_ns(struct sk_buff *skb)
 		}
 
 		if (ipv6_chk_acast_addr(net, dev, &msg->target) ||
-		    (idev->cnf.forwarding &&
+		    (READ_ONCE(idev->cnf.forwarding) &&
 		     (net->ipv6.devconf_all->proxy_ndp || idev->cnf.proxy_ndp) &&
 		     (is_router = pndisc_is_router(&msg->target, dev)) >= 0)) {
 			if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) &&
@@ -929,7 +929,7 @@ static enum skb_drop_reason ndisc_recv_ns(struct sk_buff *skb)
 	}
 
 	if (is_router < 0)
-		is_router = idev->cnf.forwarding;
+		is_router = READ_ONCE(idev->cnf.forwarding);
 
 	if (dad) {
 		ndisc_send_na(dev, &in6addr_linklocal_allnodes, &msg->target,
@@ -1080,7 +1080,7 @@ static enum skb_drop_reason ndisc_recv_na(struct sk_buff *skb)
 	 * Note that we don't do a (daddr == all-routers-mcast) check.
 	 */
 	new_state = msg->icmph.icmp6_solicited ? NUD_REACHABLE : NUD_STALE;
-	if (!neigh && lladdr && idev && idev->cnf.forwarding) {
+	if (!neigh && lladdr && idev && READ_ONCE(idev->cnf.forwarding)) {
 		if (accept_untracked_na(dev, saddr)) {
 			neigh = neigh_create(&nd_tbl, &msg->target, dev);
 			new_state = NUD_STALE;
@@ -1100,7 +1100,8 @@ static enum skb_drop_reason ndisc_recv_na(struct sk_buff *skb)
 		 * has already sent a NA to us.
 		 */
 		if (lladdr && !memcmp(lladdr, dev->dev_addr, dev->addr_len) &&
-		    net->ipv6.devconf_all->forwarding && net->ipv6.devconf_all->proxy_ndp &&
+		    READ_ONCE(net->ipv6.devconf_all->forwarding) &&
+		    net->ipv6.devconf_all->proxy_ndp &&
 		    pneigh_lookup(&nd_tbl, net, &msg->target, dev, 0)) {
 			/* XXX: idev->cnf.proxy_ndp */
 			goto out;
@@ -1148,7 +1149,7 @@ static enum skb_drop_reason ndisc_recv_rs(struct sk_buff *skb)
 	}
 
 	/* Don't accept RS if we're not in router mode */
-	if (!idev->cnf.forwarding)
+	if (!READ_ONCE(idev->cnf.forwarding))
 		goto out;
 
 	/*
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 66c685b0b6199..6a2b53de48182 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -2220,7 +2220,7 @@ struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table,
 
 	strict |= flags & RT6_LOOKUP_F_IFACE;
 	strict |= flags & RT6_LOOKUP_F_IGNORE_LINKSTATE;
-	if (net->ipv6.devconf_all->forwarding == 0)
+	if (READ_ONCE(net->ipv6.devconf_all->forwarding) == 0)
 		strict |= RT6_LOOKUP_F_REACHABLE;
 
 	rcu_read_lock();
@@ -4149,7 +4149,7 @@ static void rt6_do_redirect(struct dst_entry *dst, struct sock *sk, struct sk_bu
 	in6_dev = __in6_dev_get(skb->dev);
 	if (!in6_dev)
 		return;
-	if (in6_dev->cnf.forwarding || !in6_dev->cnf.accept_redirects)
+	if (READ_ONCE(in6_dev->cnf.forwarding) || !in6_dev->cnf.accept_redirects)
 		return;
 
 	/* RFC2461 8.1:

From ddea75d344dd87cf7f5d79efb1aab80b493f8393 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:31 +0000
Subject: [PATCH 07/15] ipv6: annotate data-races in ndisc_router_discovery()

Annotate reads from in6_dev->cnf.XXX fields, as they could
change concurrently.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/ndisc.c | 33 ++++++++++++++++++---------------
 1 file changed, 18 insertions(+), 15 deletions(-)

diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 1fb5e37bc78be..f6430db249401 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -1319,7 +1319,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	if (old_if_flags != in6_dev->if_flags)
 		send_ifinfo_notify = true;
 
-	if (!in6_dev->cnf.accept_ra_defrtr) {
+	if (!READ_ONCE(in6_dev->cnf.accept_ra_defrtr)) {
 		ND_PRINTK(2, info,
 			  "RA: %s, defrtr is false for dev: %s\n",
 			  __func__, skb->dev->name);
@@ -1327,7 +1327,8 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	}
 
 	lifetime = ntohs(ra_msg->icmph.icmp6_rt_lifetime);
-	if (lifetime != 0 && lifetime < in6_dev->cnf.accept_ra_min_lft) {
+	if (lifetime != 0 &&
+	    lifetime < READ_ONCE(in6_dev->cnf.accept_ra_min_lft)) {
 		ND_PRINTK(2, info,
 			  "RA: router lifetime (%ds) is too short: %s\n",
 			  lifetime, skb->dev->name);
@@ -1338,7 +1339,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	 * accept_ra_from_local is set to true.
 	 */
 	net = dev_net(in6_dev->dev);
-	if (!in6_dev->cnf.accept_ra_from_local &&
+	if (!READ_ONCE(in6_dev->cnf.accept_ra_from_local) &&
 	    ipv6_chk_addr(net, &ipv6_hdr(skb)->saddr, in6_dev->dev, 0)) {
 		ND_PRINTK(2, info,
 			  "RA from local address detected on dev: %s: default router ignored\n",
@@ -1350,7 +1351,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	pref = ra_msg->icmph.icmp6_router_pref;
 	/* 10b is handled as if it were 00b (medium) */
 	if (pref == ICMPV6_ROUTER_PREF_INVALID ||
-	    !in6_dev->cnf.accept_ra_rtr_pref)
+	    !READ_ONCE(in6_dev->cnf.accept_ra_rtr_pref))
 		pref = ICMPV6_ROUTER_PREF_MEDIUM;
 #endif
 	/* routes added from RAs do not use nexthop objects */
@@ -1421,10 +1422,12 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 
 		spin_unlock_bh(&table->tb6_lock);
 	}
-	if (in6_dev->cnf.accept_ra_min_hop_limit < 256 &&
+	if (READ_ONCE(in6_dev->cnf.accept_ra_min_hop_limit) < 256 &&
 	    ra_msg->icmph.icmp6_hop_limit) {
-		if (in6_dev->cnf.accept_ra_min_hop_limit <= ra_msg->icmph.icmp6_hop_limit) {
-			WRITE_ONCE(in6_dev->cnf.hop_limit, ra_msg->icmph.icmp6_hop_limit);
+		if (READ_ONCE(in6_dev->cnf.accept_ra_min_hop_limit) <=
+		    ra_msg->icmph.icmp6_hop_limit) {
+			WRITE_ONCE(in6_dev->cnf.hop_limit,
+				   ra_msg->icmph.icmp6_hop_limit);
 			fib6_metric_set(rt, RTAX_HOPLIMIT,
 					ra_msg->icmph.icmp6_hop_limit);
 		} else {
@@ -1506,7 +1509,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	}
 
 #ifdef CONFIG_IPV6_ROUTE_INFO
-	if (!in6_dev->cnf.accept_ra_from_local &&
+	if (!READ_ONCE(in6_dev->cnf.accept_ra_from_local) &&
 	    ipv6_chk_addr(dev_net(in6_dev->dev), &ipv6_hdr(skb)->saddr,
 			  in6_dev->dev, 0)) {
 		ND_PRINTK(2, info,
@@ -1515,7 +1518,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 		goto skip_routeinfo;
 	}
 
-	if (in6_dev->cnf.accept_ra_rtr_pref && ndopts.nd_opts_ri) {
+	if (READ_ONCE(in6_dev->cnf.accept_ra_rtr_pref) && ndopts.nd_opts_ri) {
 		struct nd_opt_hdr *p;
 		for (p = ndopts.nd_opts_ri;
 		     p;
@@ -1527,14 +1530,14 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 				continue;
 #endif
 			if (ri->prefix_len == 0 &&
-			    !in6_dev->cnf.accept_ra_defrtr)
+			    !READ_ONCE(in6_dev->cnf.accept_ra_defrtr))
 				continue;
 			if (ri->lifetime != 0 &&
-			    ntohl(ri->lifetime) < in6_dev->cnf.accept_ra_min_lft)
+			    ntohl(ri->lifetime) < READ_ONCE(in6_dev->cnf.accept_ra_min_lft))
 				continue;
-			if (ri->prefix_len < in6_dev->cnf.accept_ra_rt_info_min_plen)
+			if (ri->prefix_len < READ_ONCE(in6_dev->cnf.accept_ra_rt_info_min_plen))
 				continue;
-			if (ri->prefix_len > in6_dev->cnf.accept_ra_rt_info_max_plen)
+			if (ri->prefix_len > READ_ONCE(in6_dev->cnf.accept_ra_rt_info_max_plen))
 				continue;
 			rt6_route_rcv(skb->dev, (u8 *)p, (p->nd_opt_len) << 3,
 				      &ipv6_hdr(skb)->saddr);
@@ -1554,7 +1557,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 	}
 #endif
 
-	if (in6_dev->cnf.accept_ra_pinfo && ndopts.nd_opts_pi) {
+	if (READ_ONCE(in6_dev->cnf.accept_ra_pinfo) && ndopts.nd_opts_pi) {
 		struct nd_opt_hdr *p;
 		for (p = ndopts.nd_opts_pi;
 		     p;
@@ -1565,7 +1568,7 @@ static enum skb_drop_reason ndisc_router_discovery(struct sk_buff *skb)
 		}
 	}
 
-	if (ndopts.nd_opts_mtu && in6_dev->cnf.accept_ra_mtu) {
+	if (ndopts.nd_opts_mtu && READ_ONCE(in6_dev->cnf.accept_ra_mtu)) {
 		__be32 n;
 		u32 mtu;
 

From fca34cc075996767fbbdb6252be9ddd21c34c920 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:32 +0000
Subject: [PATCH 08/15] ipv6: annotate data-races around
 idev->cnf.ignore_routes_with_linkdown

idev->cnf.ignore_routes_with_linkdown can be used without any locks,
add appropriate annotations.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/addrconf.h | 2 +-
 net/ipv6/addrconf.c    | 8 ++++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/include/net/addrconf.h b/include/net/addrconf.h
index 30d6f1e84e465..9d06eb945509e 100644
--- a/include/net/addrconf.h
+++ b/include/net/addrconf.h
@@ -417,7 +417,7 @@ static inline bool ip6_ignore_linkdown(const struct net_device *dev)
 	if (unlikely(!idev))
 		return true;
 
-	return !!idev->cnf.ignore_routes_with_linkdown;
+	return !!READ_ONCE(idev->cnf.ignore_routes_with_linkdown);
 }
 
 void inet6_ifa_finish_destroy(struct inet6_ifaddr *ifp);
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index ec99e393f6a91..82b8d90afb6aa 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -566,7 +566,7 @@ static int inet6_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
 
 	if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
 	    nla_put_s32(skb, NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN,
-			devconf->ignore_routes_with_linkdown) < 0)
+			READ_ONCE(devconf->ignore_routes_with_linkdown)) < 0)
 		goto nla_put_failure;
 
 out:
@@ -935,7 +935,7 @@ static void addrconf_linkdown_change(struct net *net, __s32 newf)
 		if (idev) {
 			int changed = (!idev->cnf.ignore_routes_with_linkdown) ^ (!newf);
 
-			idev->cnf.ignore_routes_with_linkdown = newf;
+			WRITE_ONCE(idev->cnf.ignore_routes_with_linkdown, newf);
 			if (changed)
 				inet6_netconf_notify_devconf(dev_net(dev),
 							     RTM_NEWNETCONF,
@@ -956,7 +956,7 @@ static int addrconf_fixup_linkdown(struct ctl_table *table, int *p, int newf)
 
 	net = (struct net *)table->extra2;
 	old = *p;
-	*p = newf;
+	WRITE_ONCE(*p, newf);
 
 	if (p == &net->ipv6.devconf_dflt->ignore_routes_with_linkdown) {
 		if ((!newf) ^ (!old))
@@ -970,7 +970,7 @@ static int addrconf_fixup_linkdown(struct ctl_table *table, int *p, int newf)
 	}
 
 	if (p == &net->ipv6.devconf_all->ignore_routes_with_linkdown) {
-		net->ipv6.devconf_dflt->ignore_routes_with_linkdown = newf;
+		WRITE_ONCE(net->ipv6.devconf_dflt->ignore_routes_with_linkdown, newf);
 		addrconf_linkdown_change(net, newf);
 		if ((!newf) ^ (!old))
 			inet6_netconf_notify_devconf(net,

From e248948a4471c6efca19f7c5e657b30ee67a6485 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:33 +0000
Subject: [PATCH 09/15] ipv6: annotate data-races in rt6_probe()

Use READ_ONCE() while reading idev->cnf.rtr_probe_interval
while its value could be changed.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/route.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 6a2b53de48182..1b897c57c55fe 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -645,14 +645,15 @@ static void rt6_probe(struct fib6_nh *fib6_nh)
 		write_lock_bh(&neigh->lock);
 		if (!(neigh->nud_state & NUD_VALID) &&
 		    time_after(jiffies,
-			       neigh->updated + idev->cnf.rtr_probe_interval)) {
+			       neigh->updated +
+			       READ_ONCE(idev->cnf.rtr_probe_interval))) {
 			work = kmalloc(sizeof(*work), GFP_ATOMIC);
 			if (work)
 				__neigh_set_probe_once(neigh);
 		}
 		write_unlock_bh(&neigh->lock);
 	} else if (time_after(jiffies, last_probe +
-				       idev->cnf.rtr_probe_interval)) {
+				       READ_ONCE(idev->cnf.rtr_probe_interval))) {
 		work = kmalloc(sizeof(*work), GFP_ATOMIC);
 	}
 

From a8fbd4d90720b6c930661ed593d54aba77cec3c2 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:34 +0000
Subject: [PATCH 10/15] ipv6: annotate data-races around devconf->proxy_ndp

devconf->proxy_ndp can be read and written locklessly,
add appropriate annotations.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c   | 3 ++-
 net/ipv6/ip6_output.c | 2 +-
 net/ipv6/ndisc.c      | 5 +++--
 3 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 82b8d90afb6aa..409aa5ddb8569 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -561,7 +561,8 @@ static int inet6_netconf_fill_devconf(struct sk_buff *skb, int ifindex,
 		goto nla_put_failure;
 #endif
 	if ((all || type == NETCONFA_PROXY_NEIGH) &&
-	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH, devconf->proxy_ndp) < 0)
+	    nla_put_s32(skb, NETCONFA_PROXY_NEIGH,
+			READ_ONCE(devconf->proxy_ndp)) < 0)
 		goto nla_put_failure;
 
 	if ((all || type == NETCONFA_IGNORE_ROUTES_WITH_LINKDOWN) &&
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 444be8c84cc57..f08af3f4e54f5 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -552,7 +552,7 @@ int ip6_forward(struct sk_buff *skb)
 	}
 
 	/* XXX: idev->cnf.proxy_ndp? */
-	if (net->ipv6.devconf_all->proxy_ndp &&
+	if (READ_ONCE(net->ipv6.devconf_all->proxy_ndp) &&
 	    pneigh_lookup(&nd_tbl, net, &hdr->daddr, skb->dev, 0)) {
 		int proxied = ip6_forward_proxy_check(skb);
 		if (proxied > 0) {
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index f6430db249401..4114918f12c88 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -904,7 +904,8 @@ static enum skb_drop_reason ndisc_recv_ns(struct sk_buff *skb)
 
 		if (ipv6_chk_acast_addr(net, dev, &msg->target) ||
 		    (READ_ONCE(idev->cnf.forwarding) &&
-		     (net->ipv6.devconf_all->proxy_ndp || idev->cnf.proxy_ndp) &&
+		     (READ_ONCE(net->ipv6.devconf_all->proxy_ndp) ||
+		      READ_ONCE(idev->cnf.proxy_ndp)) &&
 		     (is_router = pndisc_is_router(&msg->target, dev)) >= 0)) {
 			if (!(NEIGH_CB(skb)->flags & LOCALLY_ENQUEUED) &&
 			    skb->pkt_type != PACKET_HOST &&
@@ -1101,7 +1102,7 @@ static enum skb_drop_reason ndisc_recv_na(struct sk_buff *skb)
 		 */
 		if (lladdr && !memcmp(lladdr, dev->dev_addr, dev->addr_len) &&
 		    READ_ONCE(net->ipv6.devconf_all->forwarding) &&
-		    net->ipv6.devconf_all->proxy_ndp &&
+		    READ_ONCE(net->ipv6.devconf_all->proxy_ndp) &&
 		    pneigh_lookup(&nd_tbl, net, &msg->target, dev, 0)) {
 			/* XXX: idev->cnf.proxy_ndp */
 			goto out;

From 624d5aec487cf8c2955d9c5880685714f7fe8e6f Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:35 +0000
Subject: [PATCH 11/15] ipv6: annotate data-races around
 devconf->disable_policy

idev->cnf.disable_policy and net->ipv6.devconf_all->disable_policy
can be read locklessly. Add appropriate annotations on reads
and writes.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c   | 2 +-
 net/ipv6/ip6_output.c | 4 ++--
 net/ipv6/route.c      | 4 ++--
 3 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 409aa5ddb8569..44340ff82da53 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -6699,7 +6699,7 @@ int addrconf_disable_policy(struct ctl_table *ctl, int *valp, int val)
 	if (!rtnl_trylock())
 		return restart_syscall();
 
-	*valp = val;
+	WRITE_ONCE(*valp, val);
 
 	net = (struct net *)ctl->extra2;
 	if (valp == &net->ipv6.devconf_dflt->disable_policy) {
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index f08af3f4e54f5..b9dd3a66e4236 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -513,8 +513,8 @@ int ip6_forward(struct sk_buff *skb)
 	if (skb_warn_if_lro(skb))
 		goto drop;
 
-	if (!net->ipv6.devconf_all->disable_policy &&
-	    (!idev || !idev->cnf.disable_policy) &&
+	if (!READ_ONCE(net->ipv6.devconf_all->disable_policy) &&
+	    (!idev || !READ_ONCE(idev->cnf.disable_policy)) &&
 	    !xfrm6_policy_check(NULL, XFRM_POLICY_FWD, skb)) {
 		__IP6_INC_STATS(net, idev, IPSTATS_MIB_INDISCARDS);
 		goto drop;
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 1b897c57c55fe..a92fcac902aea 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -4584,8 +4584,8 @@ struct fib6_info *addrconf_f6i_alloc(struct net *net,
 		f6i->dst_nocount = true;
 
 		if (!anycast &&
-		    (net->ipv6.devconf_all->disable_policy ||
-		     idev->cnf.disable_policy))
+		    (READ_ONCE(net->ipv6.devconf_all->disable_policy) ||
+		     READ_ONCE(idev->cnf.disable_policy)))
 			f6i->dst_nopolicy = true;
 	}
 

From 45b90ec9a214ff7fbde8190126301a07387c46e5 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:36 +0000
Subject: [PATCH 12/15] ipv6: addrconf_disable_policy() optimization

Writing over /proc/sys/net/ipv6/conf/default/disable_policy
does not need to hold RTNL.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 44340ff82da53..0e7135a206b09 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -6693,20 +6693,19 @@ void addrconf_disable_policy_idev(struct inet6_dev *idev, int val)
 static
 int addrconf_disable_policy(struct ctl_table *ctl, int *valp, int val)
 {
+	struct net *net = (struct net *)ctl->extra2;
 	struct inet6_dev *idev;
-	struct net *net;
+
+	if (valp == &net->ipv6.devconf_dflt->disable_policy) {
+		WRITE_ONCE(*valp, val);
+		return 0;
+	}
 
 	if (!rtnl_trylock())
 		return restart_syscall();
 
 	WRITE_ONCE(*valp, val);
 
-	net = (struct net *)ctl->extra2;
-	if (valp == &net->ipv6.devconf_dflt->disable_policy) {
-		rtnl_unlock();
-		return 0;
-	}
-
 	if (valp == &net->ipv6.devconf_all->disable_policy)  {
 		struct net_device *dev;
 

From 2aba913f99debce6cd1c7f606d69d094d1202559 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:37 +0000
Subject: [PATCH 13/15] ipv6/addrconf: annotate data-races around devconf
 fields (I)

Annotate lockless reads and writes on following devconf fields:

- regen_min_advance
- regen_max_retry
- dad_transmits
- use_tempaddr
- max_addresses
- max_desync_factor
- temp_valid_lft
- rtr_solicits
- rtr_solicit_max_interval
- rtr_solicit_interval
- rtr_solicit_delay
- enhanced_dad
- accept_redirects

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c | 67 +++++++++++++++++++++++++--------------------
 net/ipv6/route.c    |  3 +-
 2 files changed, 39 insertions(+), 31 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 0e7135a206b09..bcfe725f8fbd4 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1359,11 +1359,12 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 	in6_ifa_put(ifp);
 }
 
-static unsigned long ipv6_get_regen_advance(struct inet6_dev *idev)
+static unsigned long ipv6_get_regen_advance(const struct inet6_dev *idev)
 {
-	return idev->cnf.regen_min_advance + idev->cnf.regen_max_retry *
-			idev->cnf.dad_transmits *
-			max(NEIGH_VAR(idev->nd_parms, RETRANS_TIME), HZ/100) / HZ;
+	return READ_ONCE(idev->cnf.regen_min_advance) +
+		READ_ONCE(idev->cnf.regen_max_retry) *
+		READ_ONCE(idev->cnf.dad_transmits) *
+		max(NEIGH_VAR(idev->nd_parms, RETRANS_TIME), HZ/100) / HZ;
 }
 
 static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, bool block)
@@ -1384,7 +1385,7 @@ static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, bool block)
 
 retry:
 	in6_dev_hold(idev);
-	if (idev->cnf.use_tempaddr <= 0) {
+	if (READ_ONCE(idev->cnf.use_tempaddr) <= 0) {
 		write_unlock_bh(&idev->lock);
 		pr_info("%s: use_tempaddr is disabled\n", __func__);
 		in6_dev_put(idev);
@@ -1392,8 +1393,8 @@ static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, bool block)
 		goto out;
 	}
 	spin_lock_bh(&ifp->lock);
-	if (ifp->regen_count++ >= idev->cnf.regen_max_retry) {
-		idev->cnf.use_tempaddr = -1;	/*XXX*/
+	if (ifp->regen_count++ >= READ_ONCE(idev->cnf.regen_max_retry)) {
+		WRITE_ONCE(idev->cnf.use_tempaddr, -1);	/*XXX*/
 		spin_unlock_bh(&ifp->lock);
 		write_unlock_bh(&idev->lock);
 		pr_warn("%s: regeneration time exceeded - disabled temporary address support\n",
@@ -1415,7 +1416,7 @@ static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, bool block)
 	 */
 	cnf_temp_preferred_lft = READ_ONCE(idev->cnf.temp_prefered_lft);
 	max_desync_factor = min_t(long,
-				  idev->cnf.max_desync_factor,
+				  READ_ONCE(idev->cnf.max_desync_factor),
 				  cnf_temp_preferred_lft - regen_advance);
 
 	if (unlikely(idev->desync_factor > max_desync_factor)) {
@@ -1432,7 +1433,7 @@ static int ipv6_create_tempaddr(struct inet6_ifaddr *ifp, bool block)
 
 	memset(&cfg, 0, sizeof(cfg));
 	cfg.valid_lft = min_t(__u32, ifp->valid_lft,
-			      idev->cnf.temp_valid_lft + age);
+			      READ_ONCE(idev->cnf.temp_valid_lft) + age);
 	cfg.preferred_lft = cnf_temp_preferred_lft + age - idev->desync_factor;
 	cfg.preferred_lft = min_t(__u32, if_public_preferred_lft, cfg.preferred_lft);
 	cfg.preferred_lft = min_t(__u32, cfg.valid_lft, cfg.preferred_lft);
@@ -1685,7 +1686,7 @@ static int ipv6_get_saddr_eval(struct net *net,
 		 */
 		int preftmp = dst->prefs & (IPV6_PREFER_SRC_PUBLIC|IPV6_PREFER_SRC_TMP) ?
 				!!(dst->prefs & IPV6_PREFER_SRC_TMP) :
-				score->ifa->idev->cnf.use_tempaddr >= 2;
+				READ_ONCE(score->ifa->idev->cnf.use_tempaddr) >= 2;
 		ret = (!(score->ifa->flags & IFA_F_TEMPORARY)) ^ preftmp;
 		break;
 	    }
@@ -2168,6 +2169,7 @@ void addrconf_dad_failure(struct sk_buff *skb, struct inet6_ifaddr *ifp)
 {
 	struct inet6_dev *idev = ifp->idev;
 	struct net *net = dev_net(idev->dev);
+	int max_addresses;
 
 	if (addrconf_dad_end(ifp)) {
 		in6_ifa_put(ifp);
@@ -2205,9 +2207,9 @@ void addrconf_dad_failure(struct sk_buff *skb, struct inet6_ifaddr *ifp)
 
 		spin_unlock_bh(&ifp->lock);
 
-		if (idev->cnf.max_addresses &&
-		    ipv6_count_addresses(idev) >=
-		    idev->cnf.max_addresses)
+		max_addresses = READ_ONCE(idev->cnf.max_addresses);
+		if (max_addresses &&
+		    ipv6_count_addresses(idev) >= max_addresses)
 			goto lock_errdad;
 
 		net_info_ratelimited("%s: generating new stable privacy address because of DAD conflict\n",
@@ -2604,11 +2606,11 @@ static void manage_tempaddrs(struct inet6_dev *idev,
 		 * (TEMP_PREFERRED_LIFETIME - DESYNC_FACTOR), respectively.
 		 */
 		age = (now - ift->cstamp) / HZ;
-		max_valid = idev->cnf.temp_valid_lft - age;
+		max_valid = READ_ONCE(idev->cnf.temp_valid_lft) - age;
 		if (max_valid < 0)
 			max_valid = 0;
 
-		max_prefered = idev->cnf.temp_prefered_lft -
+		max_prefered = READ_ONCE(idev->cnf.temp_prefered_lft) -
 			       idev->desync_factor - age;
 		if (max_prefered < 0)
 			max_prefered = 0;
@@ -2641,7 +2643,7 @@ static void manage_tempaddrs(struct inet6_dev *idev,
 	if (list_empty(&idev->tempaddr_list) && (valid_lft || prefered_lft))
 		create = true;
 
-	if (create && idev->cnf.use_tempaddr > 0) {
+	if (create && READ_ONCE(idev->cnf.use_tempaddr) > 0) {
 		/* When a new public address is created as described
 		 * in [ADDRCONF], also create a new temporary address.
 		 */
@@ -2669,7 +2671,7 @@ int addrconf_prefix_rcv_add_addr(struct net *net, struct net_device *dev,
 	int create = 0, update_lft = 0;
 
 	if (!ifp && valid_lft) {
-		int max_addresses = in6_dev->cnf.max_addresses;
+		int max_addresses = READ_ONCE(in6_dev->cnf.max_addresses);
 		struct ifa6_config cfg = {
 			.pfx = addr,
 			.plen = pinfo->prefix_len,
@@ -4028,6 +4030,7 @@ static void addrconf_rs_timer(struct timer_list *t)
 	struct inet6_dev *idev = from_timer(idev, t, rs_timer);
 	struct net_device *dev = idev->dev;
 	struct in6_addr lladdr;
+	int rtr_solicits;
 
 	write_lock(&idev->lock);
 	if (idev->dead || !(idev->if_flags & IF_READY))
@@ -4040,7 +4043,9 @@ static void addrconf_rs_timer(struct timer_list *t)
 	if (idev->if_flags & IF_RA_RCVD)
 		goto out;
 
-	if (idev->rs_probes++ < idev->cnf.rtr_solicits || idev->cnf.rtr_solicits < 0) {
+	rtr_solicits = READ_ONCE(idev->cnf.rtr_solicits);
+
+	if (idev->rs_probes++ < rtr_solicits || rtr_solicits < 0) {
 		write_unlock(&idev->lock);
 		if (!ipv6_get_lladdr(dev, &lladdr, IFA_F_TENTATIVE))
 			ndisc_send_rs(dev, &lladdr,
@@ -4050,11 +4055,12 @@ static void addrconf_rs_timer(struct timer_list *t)
 
 		write_lock(&idev->lock);
 		idev->rs_interval = rfc3315_s14_backoff_update(
-			idev->rs_interval, idev->cnf.rtr_solicit_max_interval);
+				idev->rs_interval,
+				READ_ONCE(idev->cnf.rtr_solicit_max_interval));
 		/* The wait after the last probe can be shorter */
 		addrconf_mod_rs_timer(idev, (idev->rs_probes ==
-					     idev->cnf.rtr_solicits) ?
-				      idev->cnf.rtr_solicit_delay :
+					     READ_ONCE(idev->cnf.rtr_solicits)) ?
+				      READ_ONCE(idev->cnf.rtr_solicit_delay) :
 				      idev->rs_interval);
 	} else {
 		/*
@@ -4075,24 +4081,25 @@ static void addrconf_rs_timer(struct timer_list *t)
  */
 static void addrconf_dad_kick(struct inet6_ifaddr *ifp)
 {
-	unsigned long rand_num;
 	struct inet6_dev *idev = ifp->idev;
+	unsigned long rand_num;
 	u64 nonce;
 
 	if (ifp->flags & IFA_F_OPTIMISTIC)
 		rand_num = 0;
 	else
-		rand_num = get_random_u32_below(idev->cnf.rtr_solicit_delay ? : 1);
+		rand_num = get_random_u32_below(
+				READ_ONCE(idev->cnf.rtr_solicit_delay) ? : 1);
 
 	nonce = 0;
-	if (idev->cnf.enhanced_dad ||
-	    dev_net(idev->dev)->ipv6.devconf_all->enhanced_dad) {
+	if (READ_ONCE(idev->cnf.enhanced_dad) ||
+	    READ_ONCE(dev_net(idev->dev)->ipv6.devconf_all->enhanced_dad)) {
 		do
 			get_random_bytes(&nonce, 6);
 		while (nonce == 0);
 	}
 	ifp->dad_nonce = nonce;
-	ifp->dad_probes = idev->cnf.dad_transmits;
+	ifp->dad_probes = READ_ONCE(idev->cnf.dad_transmits);
 	addrconf_mod_dad_work(ifp, rand_num);
 }
 
@@ -4331,7 +4338,7 @@ static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
 	send_mld = ifp->scope == IFA_LINK && ipv6_lonely_lladdr(ifp);
 	send_rs = send_mld &&
 		  ipv6_accept_ra(ifp->idev) &&
-		  ifp->idev->cnf.rtr_solicits != 0 &&
+		  READ_ONCE(ifp->idev->cnf.rtr_solicits) != 0 &&
 		  (dev->flags & IFF_LOOPBACK) == 0 &&
 		  (dev->type != ARPHRD_TUNNEL) &&
 		  !netif_is_team_port(dev);
@@ -4366,7 +4373,7 @@ static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
 		write_lock_bh(&ifp->idev->lock);
 		spin_lock(&ifp->lock);
 		ifp->idev->rs_interval = rfc3315_s14_backoff_init(
-			ifp->idev->cnf.rtr_solicit_interval);
+			READ_ONCE(ifp->idev->cnf.rtr_solicit_interval));
 		ifp->idev->rs_probes = 1;
 		ifp->idev->if_flags |= IF_RS_SENT;
 		addrconf_mod_rs_timer(ifp->idev, ifp->idev->rs_interval);
@@ -5915,7 +5922,7 @@ static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token,
 		return -EINVAL;
 	}
 
-	if (idev->cnf.rtr_solicits == 0) {
+	if (READ_ONCE(idev->cnf.rtr_solicits) == 0) {
 		NL_SET_ERR_MSG(extack,
 			       "Router solicitation is disabled on device");
 		return -EINVAL;
@@ -5948,7 +5955,7 @@ static int inet6_set_iftoken(struct inet6_dev *idev, struct in6_addr *token,
 	if (update_rs) {
 		idev->if_flags |= IF_RS_SENT;
 		idev->rs_interval = rfc3315_s14_backoff_init(
-			idev->cnf.rtr_solicit_interval);
+			READ_ONCE(idev->cnf.rtr_solicit_interval));
 		idev->rs_probes = 1;
 		addrconf_mod_rs_timer(idev, idev->rs_interval);
 	}
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index a92fcac902aea..2cecb1c5a58f6 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -4150,7 +4150,8 @@ static void rt6_do_redirect(struct dst_entry *dst, struct sock *sk, struct sk_bu
 	in6_dev = __in6_dev_get(skb->dev);
 	if (!in6_dev)
 		return;
-	if (READ_ONCE(in6_dev->cnf.forwarding) || !in6_dev->cnf.accept_redirects)
+	if (READ_ONCE(in6_dev->cnf.forwarding) ||
+	    !READ_ONCE(in6_dev->cnf.accept_redirects))
 		return;
 
 	/* RFC2461 8.1:

From 2f0ff05a44302c91af54a5f9efe1b65b7681540e Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:38 +0000
Subject: [PATCH 14/15] ipv6/addrconf: annotate data-races around devconf
 fields (II)

Final (?) round of this series.

Annotate lockless reads on following devconf fields,
because they be changed concurrently from /proc/net/ipv6/conf.

- accept_dad
- optimistic_dad
- use_optimistic
- use_oif_addrs_only
- ra_honor_pio_life
- keep_addr_on_down
- ndisc_notify
- ndisc_evict_nocarrier
- suppress_frag_ndisc
- addr_gen_mode
- seg6_enabled
- ioam6_enabled
- ioam6_id
- ioam6_id_wide
- drop_unicast_in_l2_multicast
- mldv[12]_unsolicited_report_interval
- force_mld_version
- force_tllao
- accept_untracked_na
- drop_unsolicited_na
- accept_source_route

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c  | 49 +++++++++++++++++++++++---------------------
 net/ipv6/exthdrs.c   | 16 ++++++++-------
 net/ipv6/ioam6.c     |  8 ++++----
 net/ipv6/ip6_input.c |  2 +-
 net/ipv6/mcast.c     | 14 ++++++-------
 net/ipv6/ndisc.c     | 18 ++++++++--------
 net/ipv6/seg6_hmac.c |  8 +++++---
 7 files changed, 61 insertions(+), 54 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index bcfe725f8fbd4..436d9a6e6a82e 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -1557,15 +1557,17 @@ static inline int ipv6_saddr_preferred(int type)
 	return 0;
 }
 
-static bool ipv6_use_optimistic_addr(struct net *net,
-				     struct inet6_dev *idev)
+static bool ipv6_use_optimistic_addr(const struct net *net,
+				     const struct inet6_dev *idev)
 {
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	if (!idev)
 		return false;
-	if (!net->ipv6.devconf_all->optimistic_dad && !idev->cnf.optimistic_dad)
+	if (!READ_ONCE(net->ipv6.devconf_all->optimistic_dad) &&
+	    !READ_ONCE(idev->cnf.optimistic_dad))
 		return false;
-	if (!net->ipv6.devconf_all->use_optimistic && !idev->cnf.use_optimistic)
+	if (!READ_ONCE(net->ipv6.devconf_all->use_optimistic) &&
+	    !READ_ONCE(idev->cnf.use_optimistic))
 		return false;
 
 	return true;
@@ -1574,13 +1576,14 @@ static bool ipv6_use_optimistic_addr(struct net *net,
 #endif
 }
 
-static bool ipv6_allow_optimistic_dad(struct net *net,
-				      struct inet6_dev *idev)
+static bool ipv6_allow_optimistic_dad(const struct net *net,
+				      const struct inet6_dev *idev)
 {
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
 	if (!idev)
 		return false;
-	if (!net->ipv6.devconf_all->optimistic_dad && !idev->cnf.optimistic_dad)
+	if (!READ_ONCE(net->ipv6.devconf_all->optimistic_dad) &&
+	    !READ_ONCE(idev->cnf.optimistic_dad))
 		return false;
 
 	return true;
@@ -1862,7 +1865,7 @@ int ipv6_dev_get_saddr(struct net *net, const struct net_device *dst_dev,
 		idev = __in6_dev_get(dst_dev);
 		if ((dst_type & IPV6_ADDR_MULTICAST) ||
 		    dst.scope <= IPV6_ADDR_SCOPE_LINKLOCAL ||
-		    (idev && idev->cnf.use_oif_addrs_only)) {
+		    (idev && READ_ONCE(idev->cnf.use_oif_addrs_only))) {
 			use_oif_addr = true;
 		}
 	}
@@ -2683,8 +2686,8 @@ int addrconf_prefix_rcv_add_addr(struct net *net, struct net_device *dev,
 		};
 
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
-		if ((net->ipv6.devconf_all->optimistic_dad ||
-		     in6_dev->cnf.optimistic_dad) &&
+		if ((READ_ONCE(net->ipv6.devconf_all->optimistic_dad) ||
+		     READ_ONCE(in6_dev->cnf.optimistic_dad)) &&
 		    !net->ipv6.devconf_all->forwarding && sllao)
 			cfg.ifa_flags |= IFA_F_OPTIMISTIC;
 #endif
@@ -2733,7 +2736,7 @@ int addrconf_prefix_rcv_add_addr(struct net *net, struct net_device *dev,
 		 */
 		update_lft = !create && stored_lft;
 
-		if (update_lft && !in6_dev->cnf.ra_honor_pio_life) {
+		if (update_lft && !READ_ONCE(in6_dev->cnf.ra_honor_pio_life)) {
 			const u32 minimum_lft = min_t(u32,
 				stored_lft, MIN_VALID_LIFETIME);
 			valid_lft = max(valid_lft, minimum_lft);
@@ -3317,8 +3320,8 @@ void addrconf_add_linklocal(struct inet6_dev *idev,
 	struct inet6_ifaddr *ifp;
 
 #ifdef CONFIG_IPV6_OPTIMISTIC_DAD
-	if ((dev_net(idev->dev)->ipv6.devconf_all->optimistic_dad ||
-	     idev->cnf.optimistic_dad) &&
+	if ((READ_ONCE(dev_net(idev->dev)->ipv6.devconf_all->optimistic_dad) ||
+	     READ_ONCE(idev->cnf.optimistic_dad)) &&
 	    !dev_net(idev->dev)->ipv6.devconf_all->forwarding)
 		cfg.ifa_flags |= IFA_F_OPTIMISTIC;
 #endif
@@ -3890,10 +3893,10 @@ static int addrconf_ifdown(struct net_device *dev, bool unregister)
 	 */
 	if (!unregister && !idev->cnf.disable_ipv6) {
 		/* aggregate the system setting and interface setting */
-		int _keep_addr = net->ipv6.devconf_all->keep_addr_on_down;
+		int _keep_addr = READ_ONCE(net->ipv6.devconf_all->keep_addr_on_down);
 
 		if (!_keep_addr)
-			_keep_addr = idev->cnf.keep_addr_on_down;
+			_keep_addr = READ_ONCE(idev->cnf.keep_addr_on_down);
 
 		keep_addr = (_keep_addr > 0);
 	}
@@ -4119,8 +4122,8 @@ static void addrconf_dad_begin(struct inet6_ifaddr *ifp)
 
 	net = dev_net(dev);
 	if (dev->flags&(IFF_NOARP|IFF_LOOPBACK) ||
-	    (net->ipv6.devconf_all->accept_dad < 1 &&
-	     idev->cnf.accept_dad < 1) ||
+	    (READ_ONCE(net->ipv6.devconf_all->accept_dad) < 1 &&
+	     READ_ONCE(idev->cnf.accept_dad) < 1) ||
 	    !(ifp->flags&IFA_F_TENTATIVE) ||
 	    ifp->flags & IFA_F_NODAD) {
 		bool send_na = false;
@@ -4212,8 +4215,8 @@ static void addrconf_dad_work(struct work_struct *w)
 		action = DAD_ABORT;
 		ifp->state = INET6_IFADDR_STATE_POSTDAD;
 
-		if ((dev_net(idev->dev)->ipv6.devconf_all->accept_dad > 1 ||
-		     idev->cnf.accept_dad > 1) &&
+		if ((READ_ONCE(dev_net(idev->dev)->ipv6.devconf_all->accept_dad) > 1 ||
+		     READ_ONCE(idev->cnf.accept_dad) > 1) &&
 		    !idev->cnf.disable_ipv6 &&
 		    !(ifp->flags & IFA_F_STABLE_PRIVACY)) {
 			struct in6_addr addr;
@@ -4352,8 +4355,8 @@ static void addrconf_dad_completed(struct inet6_ifaddr *ifp, bool bump_id,
 
 	/* send unsolicited NA if enabled */
 	if (send_na &&
-	    (ifp->idev->cnf.ndisc_notify ||
-	     dev_net(dev)->ipv6.devconf_all->ndisc_notify)) {
+	    (READ_ONCE(ifp->idev->cnf.ndisc_notify) ||
+	     READ_ONCE(dev_net(dev)->ipv6.devconf_all->ndisc_notify))) {
 		ndisc_send_na(dev, &in6addr_linklocal_allnodes, &ifp->addr,
 			      /*router=*/ !!ifp->idev->cnf.forwarding,
 			      /*solicited=*/ false, /*override=*/ true,
@@ -6541,7 +6544,7 @@ static int addrconf_sysctl_addr_gen_mode(struct ctl_table *ctl, int write,
 		} else if (&net->ipv6.devconf_all->addr_gen_mode == ctl->data) {
 			struct net_device *dev;
 
-			net->ipv6.devconf_dflt->addr_gen_mode = new_val;
+			WRITE_ONCE(net->ipv6.devconf_dflt->addr_gen_mode, new_val);
 			for_each_netdev(net, dev) {
 				idev = __in6_dev_get(dev);
 				if (idev &&
@@ -6553,7 +6556,7 @@ static int addrconf_sysctl_addr_gen_mode(struct ctl_table *ctl, int write,
 			}
 		}
 
-		*((u32 *)ctl->data) = new_val;
+		WRITE_ONCE(*((u32 *)ctl->data), new_val);
 	}
 
 out:
diff --git a/net/ipv6/exthdrs.c b/net/ipv6/exthdrs.c
index 727792907d6cb..26f1e4a5ade05 100644
--- a/net/ipv6/exthdrs.c
+++ b/net/ipv6/exthdrs.c
@@ -379,9 +379,8 @@ static int ipv6_srh_rcv(struct sk_buff *skb)
 
 	idev = __in6_dev_get(skb->dev);
 
-	accept_seg6 = net->ipv6.devconf_all->seg6_enabled;
-	if (accept_seg6 > idev->cnf.seg6_enabled)
-		accept_seg6 = idev->cnf.seg6_enabled;
+	accept_seg6 = min(READ_ONCE(net->ipv6.devconf_all->seg6_enabled),
+			  READ_ONCE(idev->cnf.seg6_enabled));
 
 	if (!accept_seg6) {
 		kfree_skb(skb);
@@ -655,10 +654,13 @@ static int ipv6_rthdr_rcv(struct sk_buff *skb)
 	struct ipv6_rt_hdr *hdr;
 	struct rt0_hdr *rthdr;
 	struct net *net = dev_net(skb->dev);
-	int accept_source_route = net->ipv6.devconf_all->accept_source_route;
+	int accept_source_route;
 
-	if (idev && accept_source_route > idev->cnf.accept_source_route)
-		accept_source_route = idev->cnf.accept_source_route;
+	accept_source_route = READ_ONCE(net->ipv6.devconf_all->accept_source_route);
+
+	if (idev)
+		accept_source_route = min(accept_source_route,
+					  READ_ONCE(idev->cnf.accept_source_route));
 
 	if (!pskb_may_pull(skb, skb_transport_offset(skb) + 8) ||
 	    !pskb_may_pull(skb, (skb_transport_offset(skb) +
@@ -919,7 +921,7 @@ static bool ipv6_hop_ioam(struct sk_buff *skb, int optoff)
 		goto drop;
 
 	/* Ignore if IOAM is not enabled on ingress */
-	if (!__in6_dev_get(skb->dev)->cnf.ioam6_enabled)
+	if (!READ_ONCE(__in6_dev_get(skb->dev)->cnf.ioam6_enabled))
 		goto ignore;
 
 	/* Truncated Option header */
diff --git a/net/ipv6/ioam6.c b/net/ipv6/ioam6.c
index 5fa923f066321..08c9295130657 100644
--- a/net/ipv6/ioam6.c
+++ b/net/ipv6/ioam6.c
@@ -727,7 +727,7 @@ static void __ioam6_fill_trace_data(struct sk_buff *skb,
 		if (!skb->dev)
 			raw16 = IOAM6_U16_UNAVAILABLE;
 		else
-			raw16 = (__force u16)__in6_dev_get(skb->dev)->cnf.ioam6_id;
+			raw16 = (__force u16)READ_ONCE(__in6_dev_get(skb->dev)->cnf.ioam6_id);
 
 		*(__be16 *)data = cpu_to_be16(raw16);
 		data += sizeof(__be16);
@@ -735,7 +735,7 @@ static void __ioam6_fill_trace_data(struct sk_buff *skb,
 		if (skb_dst(skb)->dev->flags & IFF_LOOPBACK)
 			raw16 = IOAM6_U16_UNAVAILABLE;
 		else
-			raw16 = (__force u16)__in6_dev_get(skb_dst(skb)->dev)->cnf.ioam6_id;
+			raw16 = (__force u16)READ_ONCE(__in6_dev_get(skb_dst(skb)->dev)->cnf.ioam6_id);
 
 		*(__be16 *)data = cpu_to_be16(raw16);
 		data += sizeof(__be16);
@@ -822,7 +822,7 @@ static void __ioam6_fill_trace_data(struct sk_buff *skb,
 		if (!skb->dev)
 			raw32 = IOAM6_U32_UNAVAILABLE;
 		else
-			raw32 = __in6_dev_get(skb->dev)->cnf.ioam6_id_wide;
+			raw32 = READ_ONCE(__in6_dev_get(skb->dev)->cnf.ioam6_id_wide);
 
 		*(__be32 *)data = cpu_to_be32(raw32);
 		data += sizeof(__be32);
@@ -830,7 +830,7 @@ static void __ioam6_fill_trace_data(struct sk_buff *skb,
 		if (skb_dst(skb)->dev->flags & IFF_LOOPBACK)
 			raw32 = IOAM6_U32_UNAVAILABLE;
 		else
-			raw32 = __in6_dev_get(skb_dst(skb)->dev)->cnf.ioam6_id_wide;
+			raw32 = READ_ONCE(__in6_dev_get(skb_dst(skb)->dev)->cnf.ioam6_id_wide);
 
 		*(__be32 *)data = cpu_to_be32(raw32);
 		data += sizeof(__be32);
diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c
index 1ba97933c74fb..133610a49da6b 100644
--- a/net/ipv6/ip6_input.c
+++ b/net/ipv6/ip6_input.c
@@ -236,7 +236,7 @@ static struct sk_buff *ip6_rcv_core(struct sk_buff *skb, struct net_device *dev,
 	if (!ipv6_addr_is_multicast(&hdr->daddr) &&
 	    (skb->pkt_type == PACKET_BROADCAST ||
 	     skb->pkt_type == PACKET_MULTICAST) &&
-	    idev->cnf.drop_unicast_in_l2_multicast) {
+	    READ_ONCE(idev->cnf.drop_unicast_in_l2_multicast)) {
 		SKB_DR_SET(reason, UNICAST_IN_L2_MULTICAST);
 		goto err;
 	}
diff --git a/net/ipv6/mcast.c b/net/ipv6/mcast.c
index 76ee1615ff2a0..7ba01d8cfbae8 100644
--- a/net/ipv6/mcast.c
+++ b/net/ipv6/mcast.c
@@ -159,9 +159,9 @@ static int unsolicited_report_interval(struct inet6_dev *idev)
 	int iv;
 
 	if (mld_in_v1_mode(idev))
-		iv = idev->cnf.mldv1_unsolicited_report_interval;
+		iv = READ_ONCE(idev->cnf.mldv1_unsolicited_report_interval);
 	else
-		iv = idev->cnf.mldv2_unsolicited_report_interval;
+		iv = READ_ONCE(idev->cnf.mldv2_unsolicited_report_interval);
 
 	return iv > 0 ? iv : 1;
 }
@@ -1202,15 +1202,15 @@ static bool mld_marksources(struct ifmcaddr6 *pmc, int nsrcs,
 
 static int mld_force_mld_version(const struct inet6_dev *idev)
 {
+	const struct net *net = dev_net(idev->dev);
+	int all_force;
+
+	all_force = READ_ONCE(net->ipv6.devconf_all->force_mld_version);
 	/* Normally, both are 0 here. If enforcement to a particular is
 	 * being used, individual device enforcement will have a lower
 	 * precedence over 'all' device (.../conf/all/force_mld_version).
 	 */
-
-	if (dev_net(idev->dev)->ipv6.devconf_all->force_mld_version != 0)
-		return dev_net(idev->dev)->ipv6.devconf_all->force_mld_version;
-	else
-		return idev->cnf.force_mld_version;
+	return all_force ?: READ_ONCE(idev->cnf.force_mld_version);
 }
 
 static bool mld_in_v2_mode_only(const struct inet6_dev *idev)
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index 4114918f12c88..ae134634c323c 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -451,7 +451,7 @@ static void ip6_nd_hdr(struct sk_buff *skb,
 
 	rcu_read_lock();
 	idev = __in6_dev_get(skb->dev);
-	tclass = idev ? idev->cnf.ndisc_tclass : 0;
+	tclass = idev ? READ_ONCE(idev->cnf.ndisc_tclass) : 0;
 	rcu_read_unlock();
 
 	skb_push(skb, sizeof(*hdr));
@@ -535,7 +535,7 @@ void ndisc_send_na(struct net_device *dev, const struct in6_addr *daddr,
 		src_addr = solicited_addr;
 		if (ifp->flags & IFA_F_OPTIMISTIC)
 			override = false;
-		inc_opt |= ifp->idev->cnf.force_tllao;
+		inc_opt |= READ_ONCE(ifp->idev->cnf.force_tllao);
 		in6_ifa_put(ifp);
 	} else {
 		if (ipv6_dev_get_saddr(dev_net(dev), dev, daddr,
@@ -974,7 +974,7 @@ static int accept_untracked_na(struct net_device *dev, struct in6_addr *saddr)
 {
 	struct inet6_dev *idev = __in6_dev_get(dev);
 
-	switch (idev->cnf.accept_untracked_na) {
+	switch (READ_ONCE(idev->cnf.accept_untracked_na)) {
 	case 0: /* Don't accept untracked na (absent in neighbor cache) */
 		return 0;
 	case 1: /* Create new entries from na if currently untracked */
@@ -1025,7 +1025,7 @@ static enum skb_drop_reason ndisc_recv_na(struct sk_buff *skb)
 	 * drop_unsolicited_na takes precedence over accept_untracked_na
 	 */
 	if (!msg->icmph.icmp6_solicited && idev &&
-	    idev->cnf.drop_unsolicited_na)
+	    READ_ONCE(idev->cnf.drop_unsolicited_na))
 		return reason;
 
 	if (!ndisc_parse_options(dev, msg->opt, ndoptlen, &ndopts))
@@ -1818,7 +1818,7 @@ static bool ndisc_suppress_frag_ndisc(struct sk_buff *skb)
 	if (!idev)
 		return true;
 	if (IP6CB(skb)->flags & IP6SKB_FRAGMENTED &&
-	    idev->cnf.suppress_frag_ndisc) {
+	    READ_ONCE(idev->cnf.suppress_frag_ndisc)) {
 		net_warn_ratelimited("Received fragmented ndisc packet. Carefully consider disabling suppress_frag_ndisc.\n");
 		return true;
 	}
@@ -1895,8 +1895,8 @@ static int ndisc_netdev_event(struct notifier_block *this, unsigned long event,
 		idev = in6_dev_get(dev);
 		if (!idev)
 			break;
-		if (idev->cnf.ndisc_notify ||
-		    net->ipv6.devconf_all->ndisc_notify)
+		if (READ_ONCE(idev->cnf.ndisc_notify) ||
+		    READ_ONCE(net->ipv6.devconf_all->ndisc_notify))
 			ndisc_send_unsol_na(dev);
 		in6_dev_put(idev);
 		break;
@@ -1905,8 +1905,8 @@ static int ndisc_netdev_event(struct notifier_block *this, unsigned long event,
 		if (!idev)
 			evict_nocarrier = true;
 		else {
-			evict_nocarrier = idev->cnf.ndisc_evict_nocarrier &&
-					  net->ipv6.devconf_all->ndisc_evict_nocarrier;
+			evict_nocarrier = READ_ONCE(idev->cnf.ndisc_evict_nocarrier) &&
+					  READ_ONCE(net->ipv6.devconf_all->ndisc_evict_nocarrier);
 			in6_dev_put(idev);
 		}
 
diff --git a/net/ipv6/seg6_hmac.c b/net/ipv6/seg6_hmac.c
index d43c50a7310d6..861e0366f549d 100644
--- a/net/ipv6/seg6_hmac.c
+++ b/net/ipv6/seg6_hmac.c
@@ -241,6 +241,7 @@ bool seg6_hmac_validate_skb(struct sk_buff *skb)
 	struct sr6_tlv_hmac *tlv;
 	struct ipv6_sr_hdr *srh;
 	struct inet6_dev *idev;
+	int require_hmac;
 
 	idev = __in6_dev_get(skb->dev);
 
@@ -248,16 +249,17 @@ bool seg6_hmac_validate_skb(struct sk_buff *skb)
 
 	tlv = seg6_get_tlv_hmac(srh);
 
+	require_hmac = READ_ONCE(idev->cnf.seg6_require_hmac);
 	/* mandatory check but no tlv */
-	if (idev->cnf.seg6_require_hmac > 0 && !tlv)
+	if (require_hmac > 0 && !tlv)
 		return false;
 
 	/* no check */
-	if (idev->cnf.seg6_require_hmac < 0)
+	if (require_hmac < 0)
 		return true;
 
 	/* check only if present */
-	if (idev->cnf.seg6_require_hmac == 0 && !tlv)
+	if (require_hmac == 0 && !tlv)
 		return true;
 
 	/* now, seg6_require_hmac >= 0 && tlv */

From 2a02f8379bde89472cddb49dc8e55a17e0b68eed Mon Sep 17 00:00:00 2001
From: Eric Dumazet <edumazet@google.com>
Date: Wed, 28 Feb 2024 13:54:39 +0000
Subject: [PATCH 15/15] ipv6: use xa_array iterator to implement
 inet6_netconf_dump_devconf()

1) inet6_netconf_dump_devconf() can run under RCU protection
   instead of RTNL.

2) properly return 0 at the end of a dump, avoiding an
   an extra recvmsg() system call.

3) Do not use inet6_base_seq() anymore, for_each_netdev_dump()
   has nice properties. Restarting a GETDEVCONF dump if a device has
   been added/removed or if net->ipv6.dev_addr_genid has changed is moot.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 net/ipv6/addrconf.c | 102 +++++++++++++++++++-------------------------
 1 file changed, 44 insertions(+), 58 deletions(-)

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 436d9a6e6a82e..2f84e6ecf19f4 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -727,17 +727,18 @@ static u32 inet6_base_seq(const struct net *net)
 	return res;
 }
 
-
 static int inet6_netconf_dump_devconf(struct sk_buff *skb,
 				      struct netlink_callback *cb)
 {
 	const struct nlmsghdr *nlh = cb->nlh;
 	struct net *net = sock_net(skb->sk);
-	int h, s_h;
-	int idx, s_idx;
+	struct {
+		unsigned long ifindex;
+		unsigned int all_default;
+	} *ctx = (void *)cb->ctx;
 	struct net_device *dev;
 	struct inet6_dev *idev;
-	struct hlist_head *head;
+	int err = 0;
 
 	if (cb->strict_check) {
 		struct netlink_ext_ack *extack = cb->extack;
@@ -754,64 +755,48 @@ static int inet6_netconf_dump_devconf(struct sk_buff *skb,
 		}
 	}
 
-	s_h = cb->args[0];
-	s_idx = idx = cb->args[1];
-
-	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
-		idx = 0;
-		head = &net->dev_index_head[h];
-		rcu_read_lock();
-		cb->seq = inet6_base_seq(net);
-		hlist_for_each_entry_rcu(dev, head, index_hlist) {
-			if (idx < s_idx)
-				goto cont;
-			idev = __in6_dev_get(dev);
-			if (!idev)
-				goto cont;
-
-			if (inet6_netconf_fill_devconf(skb, dev->ifindex,
-						       &idev->cnf,
-						       NETLINK_CB(cb->skb).portid,
-						       nlh->nlmsg_seq,
-						       RTM_NEWNETCONF,
-						       NLM_F_MULTI,
-						       NETCONFA_ALL) < 0) {
-				rcu_read_unlock();
-				goto done;
-			}
-			nl_dump_check_consistent(cb, nlmsg_hdr(skb));
-cont:
-			idx++;
-		}
-		rcu_read_unlock();
+	rcu_read_lock();
+	for_each_netdev_dump(net, dev, ctx->ifindex) {
+		idev = __in6_dev_get(dev);
+		if (!idev)
+			continue;
+		err = inet6_netconf_fill_devconf(skb, dev->ifindex,
+					         &idev->cnf,
+						 NETLINK_CB(cb->skb).portid,
+						 nlh->nlmsg_seq,
+						 RTM_NEWNETCONF,
+						 NLM_F_MULTI,
+						 NETCONFA_ALL);
+		if (err < 0)
+			goto done;
 	}
-	if (h == NETDEV_HASHENTRIES) {
-		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
-					       net->ipv6.devconf_all,
-					       NETLINK_CB(cb->skb).portid,
-					       nlh->nlmsg_seq,
-					       RTM_NEWNETCONF, NLM_F_MULTI,
-					       NETCONFA_ALL) < 0)
+	if (ctx->all_default == 0) {
+		err = inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
+						 net->ipv6.devconf_all,
+						 NETLINK_CB(cb->skb).portid,
+						 nlh->nlmsg_seq,
+						 RTM_NEWNETCONF, NLM_F_MULTI,
+						 NETCONFA_ALL);
+		if (err < 0)
 			goto done;
-		else
-			h++;
-	}
-	if (h == NETDEV_HASHENTRIES + 1) {
-		if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
-					       net->ipv6.devconf_dflt,
-					       NETLINK_CB(cb->skb).portid,
-					       nlh->nlmsg_seq,
-					       RTM_NEWNETCONF, NLM_F_MULTI,
-					       NETCONFA_ALL) < 0)
+		ctx->all_default++;
+	}
+	if (ctx->all_default == 1) {
+		err = inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
+						 net->ipv6.devconf_dflt,
+						 NETLINK_CB(cb->skb).portid,
+						 nlh->nlmsg_seq,
+						 RTM_NEWNETCONF, NLM_F_MULTI,
+						 NETCONFA_ALL);
+		if (err < 0)
 			goto done;
-		else
-			h++;
+		ctx->all_default++;
 	}
 done:
-	cb->args[0] = h;
-	cb->args[1] = idx;
-
-	return skb->len;
+	if (err < 0 && likely(skb->len))
+		err = skb->len;
+	rcu_read_unlock();
+	return err;
 }
 
 #ifdef CONFIG_SYSCTL
@@ -7503,7 +7488,8 @@ int __init addrconf_init(void)
 	err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETNETCONF,
 				   inet6_netconf_get_devconf,
 				   inet6_netconf_dump_devconf,
-				   RTNL_FLAG_DOIT_UNLOCKED);
+				   RTNL_FLAG_DOIT_UNLOCKED |
+				   RTNL_FLAG_DUMP_UNLOCKED);
 	if (err < 0)
 		goto errout;
 	err = ipv6_addr_label_rtnl_register();