af_unix: implement ->psock_update_sk_prot()

unix_proto is special, it is very different from INET proto,
which even does not have a ->close(). We have to add a dummy
one to satisfy sockmap.

And now we can implement unix_bpf_update_proto() to update
sk_prot.

Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jakub Sitnicki <jakub@cloudflare.com>
Cc: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>

MAINTAINERS

@@ -10000,6 +10000,7 @@ F:	net/core/skmsg.c
 F:	net/core/sock_map.c
 F:	net/ipv4/tcp_bpf.c
 F:	net/ipv4/udp_bpf.c
+F:	net/unix/unix_bpf.c
 
 LANTIQ / INTEL Ethernet drivers
 M:	Hauke Mehrtens <hauke@hauke-m.de>

include/net/af_unix.h

@@ -89,4 +89,14 @@ void unix_sysctl_unregister(struct net *net);
 static inline int unix_sysctl_register(struct net *net) { return 0; }
 static inline void unix_sysctl_unregister(struct net *net) {}
 #endif
+
+#ifdef CONFIG_BPF_SYSCALL
+extern struct proto unix_proto;
+
+int unix_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore);
+void __init unix_bpf_build_proto(void);
+#else
+static inline void __init unix_bpf_build_proto(void)
+{}
+#endif
 #endif

net/core/sock_map.c

@@ -1536,6 +1536,7 @@ void sock_map_close(struct sock *sk, long timeout)
 	release_sock(sk);
 	saved_close(sk, timeout);
 }
+EXPORT_SYMBOL_GPL(sock_map_close);
 
 static int sock_map_iter_attach_target(struct bpf_prog *prog,
 				       union bpf_iter_link_info *linfo,

net/unix/Makefile

@@ -7,6 +7,7 @@ obj-$(CONFIG_UNIX)	+= unix.o
 
 unix-y			:= af_unix.o garbage.o
 unix-$(CONFIG_SYSCTL)	+= sysctl_net_unix.o
+unix-$(CONFIG_BPF_SYSCALL) += unix_bpf.o
 
 obj-$(CONFIG_UNIX_DIAG)	+= unix_diag.o
 unix_diag-y		:= diag.o

net/unix/af_unix.c

@@ -771,10 +771,18 @@ static const struct proto_ops unix_seqpacket_ops = {
 	.show_fdinfo =	unix_show_fdinfo,
 };
 
-static struct proto unix_proto = {
+static void unix_close(struct sock *sk, long timeout)
+{
+}
+
+struct proto unix_proto = {
 	.name			= "UNIX",
 	.owner			= THIS_MODULE,
 	.obj_size		= sizeof(struct unix_sock),
+	.close			= unix_close,
+#ifdef CONFIG_BPF_SYSCALL
+	.psock_update_sk_prot	= unix_bpf_update_proto,
+#endif
 };
 
 static struct sock *unix_create1(struct net *net, struct socket *sock, int kern)
@@ -859,6 +867,7 @@ static int unix_release(struct socket *sock)
 		return 0;
 
 	unix_release_sock(sk, 0);
+	sk->sk_prot->close(sk, 0);
 	sock->sk = NULL;
 
 	return 0;
@@ -2957,6 +2966,7 @@ static int __init af_unix_init(void)
 
 	sock_register(&unix_family_ops);
 	register_pernet_subsys(&unix_net_ops);
+	unix_bpf_build_proto();
 out:
 	return rc;
 }

net/unix/unix_bpf.c

@@ -0,0 +1,46 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Cong Wang <cong.wang@bytedance.com> */
+
+#include <linux/skmsg.h>
+#include <net/sock.h>
+#include <net/af_unix.h>
+
+static struct proto *unix_prot_saved __read_mostly;
+static DEFINE_SPINLOCK(unix_prot_lock);
+static struct proto unix_bpf_prot;
+
+static void unix_bpf_rebuild_protos(struct proto *prot, const struct proto *base)
+{
+	*prot        = *base;
+	prot->close  = sock_map_close;
+}
+
+static void unix_bpf_check_needs_rebuild(struct proto *ops)
+{
+	if (unlikely(ops != smp_load_acquire(&unix_prot_saved))) {
+		spin_lock_bh(&unix_prot_lock);
+		if (likely(ops != unix_prot_saved)) {
+			unix_bpf_rebuild_protos(&unix_bpf_prot, ops);
+			smp_store_release(&unix_prot_saved, ops);
+		}
+		spin_unlock_bh(&unix_prot_lock);
+	}
+}
+
+int unix_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore)
+{
+	if (restore) {
+		sk->sk_write_space = psock->saved_write_space;
+		WRITE_ONCE(sk->sk_prot, psock->sk_proto);
+		return 0;
+	}
+
+	unix_bpf_check_needs_rebuild(psock->sk_proto);
+	WRITE_ONCE(sk->sk_prot, &unix_bpf_prot);
+	return 0;
+}
+
+void __init unix_bpf_build_proto(void)
+{
+	unix_bpf_rebuild_protos(&unix_bpf_prot, &unix_proto);
+}