From: Runyu Xiao Date: Thu, 16 Jul 2026 14:29:58 +0000 (+0800) Subject: raw: annotate lockless match fields in raw_v4_match() X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=18f116931f52e3c3303ad4b15ff41eb89b0e4239;p=thirdparty%2Fkernel%2Flinux.git raw: annotate lockless match fields in raw_v4_match() raw_v4_match() is a lockless match helper under sk_for_each_rcu(). It still reads inet->inet_daddr, inet->inet_rcv_saddr and sk->sk_bound_dev_if with plain loads while bind, connect and bind-to-device paths can update the same match fields concurrently. Annotate only those mutable match fields in raw_v4_match(), and do so at the point of use instead of hoisting the bound-device read before the earlier short-circuit tests. Also annotate the raw bind writer and the shared IPv4 datagram connect writer used by raw sockets, so the address fields updated on bind and connect match explicit WRITE_ONCE() updates. This version intentionally leaves the shared disconnect-side IPv4 writers to follow-up cleanup and limits the writer changes here to the raw bind path and the datagram connect path directly exercised by raw sockets. Fixes: 0daf07e52709 ("raw: convert raw sockets to RCU") Signed-off-by: Runyu Xiao Link: https://patch.msgid.link/20260716142958.3064224-1-runyu.xiao@seu.edu.cn Signed-off-by: Jakub Kicinski --- diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c index 1614593b6d72..7d25519a6cdd 100644 --- a/net/ipv4/datagram.c +++ b/net/ipv4/datagram.c @@ -63,12 +63,12 @@ int __ip4_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int } /* Update addresses before rehashing */ - inet->inet_daddr = fl4->daddr; + WRITE_ONCE(inet->inet_daddr, fl4->daddr); inet->inet_dport = usin->sin_port; if (!inet->inet_saddr) inet->inet_saddr = fl4->saddr; if (!inet->inet_rcv_saddr) { - inet->inet_rcv_saddr = fl4->saddr; + WRITE_ONCE(inet->inet_rcv_saddr, fl4->saddr); if (sk->sk_prot->rehash) sk->sk_prot->rehash(sk); } diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index e9fbab6ad914..7f74d8b95a37 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -118,13 +118,21 @@ bool raw_v4_match(struct net *net, const struct sock *sk, unsigned short num, __be32 raddr, __be32 laddr, int dif, int sdif) { const struct inet_sock *inet = inet_sk(sk); + __be32 daddr, rcv_saddr; - if (net_eq(sock_net(sk), net) && inet->inet_num == num && - !(inet->inet_daddr && inet->inet_daddr != raddr) && - !(inet->inet_rcv_saddr && inet->inet_rcv_saddr != laddr) && - raw_sk_bound_dev_eq(net, sk->sk_bound_dev_if, dif, sdif)) - return true; - return false; + if (!net_eq(sock_net(sk), net) || inet->inet_num != num) + return false; + + daddr = READ_ONCE(inet->inet_daddr); + if (daddr && daddr != raddr) + return false; + + rcv_saddr = READ_ONCE(inet->inet_rcv_saddr); + if (rcv_saddr && rcv_saddr != laddr) + return false; + + return raw_sk_bound_dev_eq(net, READ_ONCE(sk->sk_bound_dev_if), + dif, sdif); } EXPORT_SYMBOL_GPL(raw_v4_match); @@ -722,7 +730,8 @@ static int raw_bind(struct sock *sk, struct sockaddr_unsized *uaddr, chk_addr_ret)) goto out; - inet->inet_rcv_saddr = inet->inet_saddr = addr->sin_addr.s_addr; + inet->inet_saddr = addr->sin_addr.s_addr; + WRITE_ONCE(inet->inet_rcv_saddr, addr->sin_addr.s_addr); if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST) inet->inet_saddr = 0; /* Use device */ sk_dst_reset(sk);