[PATCH v3 bpf-next 09/15] bpf: Change bpf_setsockopt(SOL_SOCKET) to reuse sk_setsockopt()

[Martin KaFai Lau <kafai@fb.com>]

After the prep work in the previous patches,
this patch removes most of the dup code from bpf_setsockopt(SOL_SOCKET)
and reuses them from sk_setsockopt().

The only exception is SO_RCVLOWAT.  The bpf_setsockopt() does
not always have the sock ptr (sk->sk_socket) and sock->ops->set_rcvlowat
is needed in sk_setsockopt.  tcp_set_rcvlowat is the only
implementation for set_rcvlowat.  bpf_setsockopt() needs
one special handling for SO_RCVLOWAT for tcp_sock, so leave
SO_RCVLOWAT in bpf_setsockopt() for now.

The existing optname white-list is refactored into a new
function sol_socket_setsockopt().

Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
 include/net/sock.h |   2 +
 net/core/filter.c  | 130 +++++++++++++--------------------------------
 net/core/sock.c    |   4 +-
 3 files changed, 42 insertions(+), 94 deletions(-)

diff --git a/include/net/sock.h b/include/net/sock.h
index 72b78c2b6f83..b7e159f9d7bf 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1800,6 +1800,8 @@ void sock_pfree(struct sk_buff *skb);
 #define sock_edemux sock_efree
 #endif
 
+int sk_setsockopt(struct sock *sk, int level, int optname,
+		  sockptr_t optval, unsigned int optlen);
 int sock_setsockopt(struct socket *sock, int level, int op,
 		    sockptr_t optval, unsigned int optlen);
 
diff --git a/net/core/filter.c b/net/core/filter.c
index 01cb4a01b1c1..d1f9f8360f60 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -5013,106 +5013,52 @@ static const struct bpf_func_proto bpf_get_socket_uid_proto = {
 	.arg1_type      = ARG_PTR_TO_CTX,
 };
 
+static int sol_socket_setsockopt(struct sock *sk, int optname,
+				 char *optval, int optlen)
+{
+	switch (optname) {
+	case SO_SNDBUF:
+	case SO_RCVBUF:
+	case SO_KEEPALIVE:
+	case SO_PRIORITY:
+	case SO_REUSEPORT:
+	case SO_RCVLOWAT:
+	case SO_MARK:
+	case SO_MAX_PACING_RATE:
+	case SO_BINDTOIFINDEX:
+	case SO_TXREHASH:
+		if (optlen != sizeof(int))
+			return -EINVAL;
+		break;
+	case SO_BINDTODEVICE:
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (optname == SO_RCVLOWAT) {
+		int val = *(int *)optval;
+
+		if (val < 0)
+			val = INT_MAX;
+		WRITE_ONCE(sk->sk_rcvlowat, val ? : 1);
+		return 0;
+	}
+
+	return sk_setsockopt(sk, SOL_SOCKET, optname,
+			     KERNEL_SOCKPTR(optval), optlen);
+}
+
 static int __bpf_setsockopt(struct sock *sk, int level, int optname,
 			    char *optval, int optlen)
 {
-	char devname[IFNAMSIZ];
-	int val, valbool;
-	struct net *net;
-	int ifindex;
-	int ret = 0;
+	int val, ret = 0;
 
 	if (!sk_fullsock(sk))
 		return -EINVAL;
 
 	if (level == SOL_SOCKET) {
-		if (optlen != sizeof(int) && optname != SO_BINDTODEVICE)
-			return -EINVAL;
-		val = *((int *)optval);
-		valbool = val ? 1 : 0;
-
-		/* Only some socketops are supported */
-		switch (optname) {
-		case SO_RCVBUF:
-			val = min_t(u32, val, sysctl_rmem_max);
-			val = min_t(int, val, INT_MAX / 2);
-			sk->sk_userlocks |= SOCK_RCVBUF_LOCK;
-			WRITE_ONCE(sk->sk_rcvbuf,
-				   max_t(int, val * 2, SOCK_MIN_RCVBUF));
-			break;
-		case SO_SNDBUF:
-			val = min_t(u32, val, sysctl_wmem_max);
-			val = min_t(int, val, INT_MAX / 2);
-			sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
-			WRITE_ONCE(sk->sk_sndbuf,
-				   max_t(int, val * 2, SOCK_MIN_SNDBUF));
-			break;
-		case SO_MAX_PACING_RATE: /* 32bit version */
-			if (val != ~0U)
-				cmpxchg(&sk->sk_pacing_status,
-					SK_PACING_NONE,
-					SK_PACING_NEEDED);
-			sk->sk_max_pacing_rate = (val == ~0U) ?
-						 ~0UL : (unsigned int)val;
-			sk->sk_pacing_rate = min(sk->sk_pacing_rate,
-						 sk->sk_max_pacing_rate);
-			break;
-		case SO_PRIORITY:
-			sk->sk_priority = val;
-			break;
-		case SO_RCVLOWAT:
-			if (val < 0)
-				val = INT_MAX;
-			WRITE_ONCE(sk->sk_rcvlowat, val ? : 1);
-			break;
-		case SO_MARK:
-			if (sk->sk_mark != val) {
-				sk->sk_mark = val;
-				sk_dst_reset(sk);
-			}
-			break;
-		case SO_BINDTODEVICE:
-			optlen = min_t(long, optlen, IFNAMSIZ - 1);
-			strncpy(devname, optval, optlen);
-			devname[optlen] = 0;
-
-			ifindex = 0;
-			if (devname[0] != '\0') {
-				struct net_device *dev;
-
-				ret = -ENODEV;
-
-				net = sock_net(sk);
-				dev = dev_get_by_name(net, devname);
-				if (!dev)
-					break;
-				ifindex = dev->ifindex;
-				dev_put(dev);
-			}
-			fallthrough;
-		case SO_BINDTOIFINDEX:
-			if (optname == SO_BINDTOIFINDEX)
-				ifindex = val;
-			ret = sock_bindtoindex(sk, ifindex, false);
-			break;
-		case SO_KEEPALIVE:
-			if (sk->sk_prot->keepalive)
-				sk->sk_prot->keepalive(sk, valbool);
-			sock_valbool_flag(sk, SOCK_KEEPOPEN, valbool);
-			break;
-		case SO_REUSEPORT:
-			sk->sk_reuseport = valbool;
-			break;
-		case SO_TXREHASH:
-			if (val < -1 || val > 1) {
-				ret = -EINVAL;
-				break;
-			}
-			sk->sk_txrehash = (u8)val;
-			break;
-		default:
-			ret = -EINVAL;
-		}
+		return sol_socket_setsockopt(sk, optname, optval, optlen);
 	} else if (IS_ENABLED(CONFIG_INET) && level == SOL_IP) {
 		if (optlen != sizeof(int) || sk->sk_family != AF_INET)
 			return -EINVAL;
diff --git a/net/core/sock.c b/net/core/sock.c
index 7ea46e4700fd..29f161ab717f 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -1077,8 +1077,8 @@ EXPORT_SYMBOL(sockopt_capable);
  *	at the socket level. Everything here is generic.
  */
 
-static int sk_setsockopt(struct sock *sk, int level, int optname,
-			 sockptr_t optval, unsigned int optlen)
+int sk_setsockopt(struct sock *sk, int level, int optname,
+		  sockptr_t optval, unsigned int optlen)
 {
 	struct so_timestamping timestamping;
 	struct socket *sock = sk->sk_socket;
-- 
2.30.2