Merge branch 'inet_csums_part3'

Tom Herbert says:

====================
net: Checksum offload changes - Part III

I am working on overhauling RX checksum offload. Goals of this effort
are:

- Specify what exactly it means when driver returns CHECKSUM_UNNECESSARY
- Preserve CHECKSUM_COMPLETE through encapsulation layers
- Don't do skb_checksum more than once per packet
- Unify GRO and non-GRO csum verification as much as possible
- Unify the checksum functions (checksum_init)
- Simply code

What is in this third patch set:

- Remove sk_no_check from sunrpc (doesn't seem to have any effect)
- Eliminate no_check from protosw. All protocols are using default of
  zero for this
- Split sk_no_check into sk_no_check_tx and sk_no_check_rx
- Make enabling of UDP6 more restrictive and explicit
- Support zero UDP6 checksums in l2tp

V2: Took out vxlan changes to set zero csums in IPv6, this will
    be in a later patch set.
V3: Fixed bug in restricting UDP6 checksums.

Please review carefully and test if possible, mucking with basic
checksum functions is always a little precarious :-)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/scsi/iscsi_tcp.c

@@ -244,7 +244,7 @@ iscsi_sw_tcp_conn_restore_callbacks(struct iscsi_conn *conn)
 	sk->sk_data_ready   = tcp_sw_conn->old_data_ready;
 	sk->sk_state_change = tcp_sw_conn->old_state_change;
 	sk->sk_write_space  = tcp_sw_conn->old_write_space;
-	sk->sk_no_check	 = 0;
+	sk->sk_no_check_tx = 0;
 	write_unlock_bh(&sk->sk_callback_lock);
 }
 

include/linux/udp.h

@@ -47,7 +47,9 @@ struct udp_sock {
 #define udp_portaddr_node	inet.sk.__sk_common.skc_portaddr_node
 	int		 pending;	/* Any pending frames ? */
 	unsigned int	 corkflag;	/* Cork is required */
-  	__u16		 encap_type;	/* Is this an Encapsulation socket? */
+	__u8		 encap_type;	/* Is this an Encapsulation socket? */
+	unsigned char	 no_check6_tx:1,/* Send zero UDP6 checksums on TX? */
+			 no_check6_rx:1;/* Allow zero UDP6 checksums on RX? */
 	/*
 	 * Following member retains the information to create a UDP header
 	 * when the socket is uncorked.
@@ -76,6 +78,26 @@ static inline struct udp_sock *udp_sk(const struct sock *sk)
 	return (struct udp_sock *)sk;
 }
 
+static inline void udp_set_no_check6_tx(struct sock *sk, bool val)
+{
+	udp_sk(sk)->no_check6_tx = val;
+}
+
+static inline void udp_set_no_check6_rx(struct sock *sk, bool val)
+{
+	udp_sk(sk)->no_check6_rx = val;
+}
+
+static inline bool udp_get_no_check6_tx(struct sock *sk)
+{
+	return udp_sk(sk)->no_check6_tx;
+}
+
+static inline bool udp_get_no_check6_rx(struct sock *sk)
+{
+	return udp_sk(sk)->no_check6_rx;
+}
+
 #define udp_portaddr_for_each_entry(__sk, node, list) \
 	hlist_nulls_for_each_entry(__sk, node, list, __sk_common.skc_portaddr_node)
 

include/net/protocol.h

@@ -86,7 +86,6 @@ struct inet_protosw {
 	struct proto	 *prot;
 	const struct proto_ops *ops;
 
-	char             no_check;   /* checksum on rcv/xmit/none? */
 	unsigned char	 flags;      /* See INET_PROTOSW_* below.  */
 };
 #define INET_PROTOSW_REUSE 0x01	     /* Are ports automatically reusable? */

include/net/sock.h

@@ -243,7 +243,8 @@ struct cg_proto;
   *	@sk_sndbuf: size of send buffer in bytes
   *	@sk_flags: %SO_LINGER (l_onoff), %SO_BROADCAST, %SO_KEEPALIVE,
   *		   %SO_OOBINLINE settings, %SO_TIMESTAMPING settings
-  *	@sk_no_check: %SO_NO_CHECK setting, whether or not checkup packets
+  *	@sk_no_check_tx: %SO_NO_CHECK setting, set checksum in TX packets
+  *	@sk_no_check_rx: allow zero checksum in RX packets
   *	@sk_route_caps: route capabilities (e.g. %NETIF_F_TSO)
   *	@sk_route_nocaps: forbidden route capabilities (e.g NETIF_F_GSO_MASK)
   *	@sk_gso_type: GSO type (e.g. %SKB_GSO_TCPV4)
@@ -371,7 +372,8 @@ struct sock {
 	struct sk_buff_head	sk_write_queue;
 	kmemcheck_bitfield_begin(flags);
 	unsigned int		sk_shutdown  : 2,
-				sk_no_check  : 2,
+				sk_no_check_tx : 1,
+				sk_no_check_rx : 1,
 				sk_userlocks : 4,
 				sk_protocol  : 8,
 				sk_type      : 16;

include/net/udp.h

@@ -95,15 +95,6 @@ static inline struct udp_hslot *udp_hashslot2(struct udp_table *table,
 	return &table->hash2[hash & table->mask];
 }
 
-/* Note: this must match 'valbool' in sock_setsockopt */
-#define UDP_CSUM_NOXMIT		1
-
-/* Used by SunRPC/xprt layer. */
-#define UDP_CSUM_NORCV		2
-
-/* Default, as per the RFC, is to always do csums. */
-#define UDP_CSUM_DEFAULT	0
-
 extern struct proto udp_prot;
 
 extern atomic_long_t udp_memory_allocated;

include/uapi/linux/l2tp.h

@@ -124,6 +124,8 @@ enum {
 	L2TP_ATTR_STATS,		/* nested */
 	L2TP_ATTR_IP6_SADDR,		/* struct in6_addr */
 	L2TP_ATTR_IP6_DADDR,		/* struct in6_addr */
+	L2TP_ATTR_UDP_ZERO_CSUM6_TX,	/* u8 */
+	L2TP_ATTR_UDP_ZERO_CSUM6_RX,	/* u8 */
 	__L2TP_ATTR_MAX,
 };
 

include/uapi/linux/udp.h

@@ -29,6 +29,8 @@ struct udphdr {
 /* UDP socket options */
 #define UDP_CORK	1	/* Never send partially complete segments */
 #define UDP_ENCAP	100	/* Set the socket to accept encapsulated packets */
+#define UDP_NO_CHECK6_TX 101	/* Disable sending checksum for UDP6X */
+#define UDP_NO_CHECK6_RX 102	/* Disable accpeting checksum for UDP6 */
 
 /* UDP encapsulation types */
 #define UDP_ENCAP_ESPINUDP_NON_IKE	1 /* draft-ietf-ipsec-nat-t-ike-00/01 */

net/appletalk/ddp.c

@@ -1669,7 +1669,7 @@ static int atalk_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr
 		goto out;
 	}
 
-	if (sk->sk_no_check == 1)
+	if (sk->sk_no_check_tx)
 		ddp->deh_sum = 0;
 	else
 		ddp->deh_sum = atalk_checksum(skb, len + sizeof(*ddp));

net/core/sock.c

@@ -784,7 +784,7 @@ int sock_setsockopt(struct socket *sock, int level, int optname,
 		break;
 
 	case SO_NO_CHECK:
-		sk->sk_no_check = valbool;
+		sk->sk_no_check_tx = valbool;
 		break;
 
 	case SO_PRIORITY:
@@ -1064,7 +1064,7 @@ int sock_getsockopt(struct socket *sock, int level, int optname,
 		break;
 
 	case SO_NO_CHECK:
-		v.val = sk->sk_no_check;
+		v.val = sk->sk_no_check_tx;
 		break;
 
 	case SO_PRIORITY:

net/dccp/ipv4.c

@@ -1024,7 +1024,6 @@ static struct inet_protosw dccp_v4_protosw = {
 	.protocol	= IPPROTO_DCCP,
 	.prot		= &dccp_v4_prot,
 	.ops		= &inet_dccp_ops,
-	.no_check	= 0,
 	.flags		= INET_PROTOSW_ICSK,
 };
 

net/decnet/af_decnet.c

@@ -481,7 +481,7 @@ static struct sock *dn_alloc_sock(struct net *net, struct socket *sock, gfp_t gf
 
 	sk->sk_backlog_rcv = dn_nsp_backlog_rcv;
 	sk->sk_destruct    = dn_destruct;
-	sk->sk_no_check    = 1;
+	sk->sk_no_check_tx = 1;
 	sk->sk_family      = PF_DECnet;
 	sk->sk_protocol    = 0;
 	sk->sk_allocation  = gfp;

net/ipv4/af_inet.c

@@ -254,7 +254,6 @@ static int inet_create(struct net *net, struct socket *sock, int protocol,
 	struct inet_sock *inet;
 	struct proto *answer_prot;
 	unsigned char answer_flags;
-	char answer_no_check;
 	int try_loading_module = 0;
 	int err;
 
@@ -312,7 +311,6 @@ static int inet_create(struct net *net, struct socket *sock, int protocol,
 
 	sock->ops = answer->ops;
 	answer_prot = answer->prot;
-	answer_no_check = answer->no_check;
 	answer_flags = answer->flags;
 	rcu_read_unlock();
 
@@ -324,7 +322,6 @@ static int inet_create(struct net *net, struct socket *sock, int protocol,
 		goto out;
 
 	err = 0;
-	sk->sk_no_check = answer_no_check;
 	if (INET_PROTOSW_REUSE & answer_flags)
 		sk->sk_reuse = SK_CAN_REUSE;
 
@@ -1002,7 +999,6 @@ static struct inet_protosw inetsw_array[] =
 		.protocol =   IPPROTO_TCP,
 		.prot =       &tcp_prot,
 		.ops =        &inet_stream_ops,
-		.no_check =   0,
 		.flags =      INET_PROTOSW_PERMANENT |
 			      INET_PROTOSW_ICSK,
 	},
@@ -1012,7 +1008,6 @@ static struct inet_protosw inetsw_array[] =
 		.protocol =   IPPROTO_UDP,
 		.prot =       &udp_prot,
 		.ops =        &inet_dgram_ops,
-		.no_check =   UDP_CSUM_DEFAULT,
 		.flags =      INET_PROTOSW_PERMANENT,
        },
 
@@ -1021,7 +1016,6 @@ static struct inet_protosw inetsw_array[] =
 		.protocol =   IPPROTO_ICMP,
 		.prot =       &ping_prot,
 		.ops =        &inet_dgram_ops,
-		.no_check =   UDP_CSUM_DEFAULT,
 		.flags =      INET_PROTOSW_REUSE,
        },
 
@@ -1030,7 +1024,6 @@ static struct inet_protosw inetsw_array[] =
 	       .protocol =   IPPROTO_IP,	/* wild card */
 	       .prot =       &raw_prot,
 	       .ops =        &inet_sockraw_ops,
-	       .no_check =   UDP_CSUM_DEFAULT,
 	       .flags =      INET_PROTOSW_REUSE,
        }
 };

net/ipv4/udp.c

@@ -785,7 +785,7 @@ static int udp_send_skb(struct sk_buff *skb, struct flowi4 *fl4)
 	if (is_udplite)  				 /*     UDP-Lite      */
 		csum = udplite_csum(skb);
 
-	else if (sk->sk_no_check == UDP_CSUM_NOXMIT) {   /* UDP csum disabled */
+	else if (sk->sk_no_check_tx) {   /* UDP csum disabled */
 
 		skb->ip_summed = CHECKSUM_NONE;
 		goto send;
@@ -1968,7 +1968,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		       int (*push_pending_frames)(struct sock *))
 {
 	struct udp_sock *up = udp_sk(sk);
-	int val;
+	int val, valbool;
 	int err = 0;
 	int is_udplite = IS_UDPLITE(sk);
 
@@ -1978,6 +1978,8 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 	if (get_user(val, (int __user *)optval))
 		return -EFAULT;
 
+	valbool = val ? 1 : 0;
+
 	switch (optname) {
 	case UDP_CORK:
 		if (val != 0) {
@@ -2007,6 +2009,14 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 		}
 		break;
 
+	case UDP_NO_CHECK6_TX:
+		up->no_check6_tx = valbool;
+		break;
+
+	case UDP_NO_CHECK6_RX:
+		up->no_check6_rx = valbool;
+		break;
+
 	/*
 	 * 	UDP-Lite's partial checksum coverage (RFC 3828).
 	 */
@@ -2089,6 +2099,14 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname,
 		val = up->encap_type;
 		break;
 
+	case UDP_NO_CHECK6_TX:
+		val = up->no_check6_tx;
+		break;
+
+	case UDP_NO_CHECK6_RX:
+		val = up->no_check6_rx;
+		break;
+
 	/* The following two cannot be changed on UDP sockets, the return is
 	 * always 0 (which corresponds to the full checksum coverage of UDP). */
 	case UDPLITE_SEND_CSCOV:

net/ipv4/udplite.c

@@ -70,7 +70,6 @@ static struct inet_protosw udplite4_protosw = {
 	.protocol	=  IPPROTO_UDPLITE,
 	.prot		=  &udplite_prot,
 	.ops		=  &inet_dgram_ops,
-	.no_check	=  0,		/* must checksum (RFC 3828) */
 	.flags		=  INET_PROTOSW_PERMANENT,
 };
 

net/ipv6/af_inet6.c

@@ -106,7 +106,6 @@ static int inet6_create(struct net *net, struct socket *sock, int protocol,
 	struct inet_protosw *answer;
 	struct proto *answer_prot;
 	unsigned char answer_flags;
-	char answer_no_check;
 	int try_loading_module = 0;
 	int err;
 
@@ -162,7 +161,6 @@ static int inet6_create(struct net *net, struct socket *sock, int protocol,
 
 	sock->ops = answer->ops;
 	answer_prot = answer->prot;
-	answer_no_check = answer->no_check;
 	answer_flags = answer->flags;
 	rcu_read_unlock();
 
@@ -176,7 +174,6 @@ static int inet6_create(struct net *net, struct socket *sock, int protocol,
 	sock_init_data(sock, sk);
 
 	err = 0;
-	sk->sk_no_check = answer_no_check;
 	if (INET_PROTOSW_REUSE & answer_flags)
 		sk->sk_reuse = SK_CAN_REUSE;
 

net/ipv6/ping.c

@@ -51,7 +51,6 @@ static struct inet_protosw pingv6_protosw = {
 	.protocol =  IPPROTO_ICMPV6,
 	.prot =      &pingv6_prot,
 	.ops =       &inet6_dgram_ops,
-	.no_check =  UDP_CSUM_DEFAULT,
 	.flags =     INET_PROTOSW_REUSE,
 };
 

net/ipv6/raw.c

@@ -1322,7 +1322,6 @@ static struct inet_protosw rawv6_protosw = {
 	.protocol	= IPPROTO_IP,	/* wild card */
 	.prot		= &rawv6_prot,
 	.ops		= &inet6_sockraw_ops,
-	.no_check	= UDP_CSUM_DEFAULT,
 	.flags		= INET_PROTOSW_REUSE,
 };
 

net/ipv6/tcp_ipv6.c

@@ -1992,7 +1992,6 @@ static struct inet_protosw tcpv6_protosw = {
 	.protocol	=	IPPROTO_TCP,
 	.prot		=	&tcpv6_prot,
 	.ops		=	&inet6_stream_ops,
-	.no_check	=	0,
 	.flags		=	INET_PROTOSW_PERMANENT |
 				INET_PROTOSW_ICSK,
 };

net/ipv6/udp.c

@@ -794,10 +794,10 @@ static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
 	dif = inet6_iif(skb);
 	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
 	while (sk) {
-		/* If zero checksum and sk_no_check is not on for
+		/* If zero checksum and no_check is not on for
 		 * the socket then skip it.
 		 */
-		if (uh->check || sk->sk_no_check)
+		if (uh->check || udp_sk(sk)->no_check6_rx)
 			stack[count++] = sk;
 
 		sk = udp_v6_mcast_next(net, sk_nulls_next(sk), uh->dest, daddr,
@@ -887,7 +887,7 @@ int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
 	if (sk != NULL) {
 		int ret;
 
-		if (!uh->check && !sk->sk_no_check) {
+		if (!uh->check && !udp_sk(sk)->no_check6_rx) {
 			sock_put(sk);
 			udp6_csum_zero_error(skb);
 			goto csum_error;
@@ -1037,7 +1037,7 @@ static int udp_v6_push_pending_frames(struct sock *sk)
 
 	if (is_udplite)
 		csum = udplite_csum_outgoing(sk, skb);
-	else if (sk->sk_no_check == UDP_CSUM_NOXMIT) {   /* UDP csum disabled */
+	else if (up->no_check6_tx) {   /* UDP csum disabled */
 		skb->ip_summed = CHECKSUM_NONE;
 		goto send;
 	} else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
@@ -1507,7 +1507,6 @@ static struct inet_protosw udpv6_protosw = {
 	.protocol =  IPPROTO_UDP,
 	.prot =      &udpv6_prot,
 	.ops =       &inet6_dgram_ops,
-	.no_check =  UDP_CSUM_DEFAULT,
 	.flags =     INET_PROTOSW_PERMANENT,
 };
 

net/ipv6/udplite.c

@@ -64,7 +64,6 @@ static struct inet_protosw udplite6_protosw = {
 	.protocol	= IPPROTO_UDPLITE,
 	.prot		= &udplitev6_prot,
 	.ops		= &inet6_dgram_ops,
-	.no_check	= 0,
 	.flags		= INET_PROTOSW_PERMANENT,
 };
 

net/ipx/af_ipx.c

@@ -1353,7 +1353,7 @@ static int ipx_create(struct net *net, struct socket *sock, int protocol,
 
 	sk_refcnt_debug_inc(sk);
 	sock_init_data(sock, sk);
-	sk->sk_no_check = 1;		/* Checksum off by default */
+	sk->sk_no_check_tx = 1;		/* Checksum off by default */
 	sock->ops = &ipx_dgram_ops;
 	rc = 0;
 out:

net/ipx/ipx_route.c

@@ -236,7 +236,8 @@ int ipxrtr_route_packet(struct sock *sk, struct sockaddr_ipx *usipx,
 	}
 
 	/* Apply checksum. Not allowed on 802.3 links. */
-	if (sk->sk_no_check || intrfc->if_dlink_type == htons(IPX_FRAME_8023))
+	if (sk->sk_no_check_tx ||
+	    intrfc->if_dlink_type == htons(IPX_FRAME_8023))
 		ipx->ipx_checksum = htons(0xFFFF);
 	else
 		ipx->ipx_checksum = ipx_cksum(ipx, len + sizeof(struct ipxhdr));