Commit 3eb45036 authored by Santosh Shilimkar's avatar Santosh Shilimkar
Browse files

rds: add type of service(tos) infrastructure 



RDS Service type (TOS) is user-defined and needs to be configured
via RDS IOCTL interface. It must be set before initiating any
traffic and once set the TOS can not be changed. All out-going
traffic from the socket will be associated with its TOS.

Reviewed-by: default avatarSowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: default avatarSantosh Shilimkar <santosh.shilimkar@oracle.com>
[yanjun.zhu@oracle.com: Adapted original patch with ipv6 changes]
Signed-off-by: default avatarZhu Yanjun <yanjun.zhu@oracle.com>
parent d021fabf

include/uapi/linux/rds.h

+11 −0
Original line number Original line Diff line number Diff line
@@ -69,6 +69,12 @@ 
#define RDS_TRANS_COUNT	3

#define RDS_TRANS_COUNT	3

#define	RDS_TRANS_NONE	(~0)

#define	RDS_TRANS_NONE	(~0)





/* IOCTLS commands for SOL_RDS */

#define SIOCRDSSETTOS		(SIOCPROTOPRIVATE)

#define SIOCRDSGETTOS		(SIOCPROTOPRIVATE + 1)



typedef __u8	rds_tos_t;



/*

/*

 * Control message types for SOL_RDS.

 * Control message types for SOL_RDS.

 *

 *
@@ -149,6 +155,7 @@ struct rds_info_connection { 
	__be32		faddr;

	__be32		faddr;

	__u8		transport[TRANSNAMSIZ];		/* null term ascii */

	__u8		transport[TRANSNAMSIZ];		/* null term ascii */

	__u8		flags;

	__u8		flags;

	__u8		tos;

} __attribute__((packed));

} __attribute__((packed));





struct rds6_info_connection {

struct rds6_info_connection {
@@ -171,6 +178,7 @@ struct rds_info_message { 
	__be16		lport;

	__be16		lport;

	__be16		fport;

	__be16		fport;

	__u8		flags;

	__u8		flags;

	__u8		tos;

} __attribute__((packed));

} __attribute__((packed));





struct rds6_info_message {

struct rds6_info_message {
@@ -214,6 +222,7 @@ struct rds_info_tcp_socket { 
	__u32           last_sent_nxt;

	__u32           last_sent_nxt;

	__u32           last_expected_una;

	__u32           last_expected_una;

	__u32           last_seen_una;

	__u32           last_seen_una;

	__u8		tos;

} __attribute__((packed));

} __attribute__((packed));





struct rds6_info_tcp_socket {

struct rds6_info_tcp_socket {
@@ -240,6 +249,7 @@ struct rds_info_rdma_connection { 
	__u32		max_send_sge;

	__u32		max_send_sge;

	__u32		rdma_mr_max;

	__u32		rdma_mr_max;

	__u32		rdma_mr_size;

	__u32		rdma_mr_size;

	__u8		tos;

};

};





struct rds6_info_rdma_connection {

struct rds6_info_rdma_connection {
@@ -253,6 +263,7 @@ struct rds6_info_rdma_connection { 
	__u32		max_send_sge;

	__u32		max_send_sge;

	__u32		rdma_mr_max;

	__u32		rdma_mr_max;

	__u32		rdma_mr_size;

	__u32		rdma_mr_size;

	__u8		tos;

};

};





/* RDS message Receive Path Latency points */

/* RDS message Receive Path Latency points */

net/rds/af_rds.c

+34 −1
Original line number Original line Diff line number Diff line
@@ -254,9 +254,40 @@ static __poll_t rds_poll(struct file *file, struct socket *sock, 




static int rds_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)

static int rds_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)

{

{

	struct rds_sock *rs = rds_sk_to_rs(sock->sk);

	rds_tos_t tos;



	switch (cmd) {

	case SIOCRDSSETTOS:

		if (get_user(tos, (rds_tos_t __user *)arg))

			return -EFAULT;



		if (rs->rs_transport &&

		    rs->rs_transport->t_type == RDS_TRANS_TCP)

			tos = 0;



		spin_lock_bh(&rds_sock_lock);

		if (rs->rs_tos || rs->rs_conn) {

			spin_unlock_bh(&rds_sock_lock);

			return -EINVAL;

		}

		rs->rs_tos = tos;

		spin_unlock_bh(&rds_sock_lock);

		break;

	case SIOCRDSGETTOS:

		spin_lock_bh(&rds_sock_lock);

		tos = rs->rs_tos;

		spin_unlock_bh(&rds_sock_lock);

		if (put_user(tos, (rds_tos_t __user *)arg))

			return -EFAULT;

		break;

	default:

		return -ENOIOCTLCMD;

		return -ENOIOCTLCMD;

	}

	}





	return 0;

}



static int rds_cancel_sent_to(struct rds_sock *rs, char __user *optval,

static int rds_cancel_sent_to(struct rds_sock *rs, char __user *optval,

			      int len)

			      int len)

{

{
@@ -650,6 +681,8 @@ static int __rds_create(struct socket *sock, struct sock *sk, int protocol) 
	spin_lock_init(&rs->rs_rdma_lock);

	spin_lock_init(&rs->rs_rdma_lock);

	rs->rs_rdma_keys = RB_ROOT;

	rs->rs_rdma_keys = RB_ROOT;

	rs->rs_rx_traces = 0;

	rs->rs_rx_traces = 0;

	rs->rs_tos = 0;

	rs->rs_conn = NULL;





	spin_lock_bh(&rds_sock_lock);

	spin_lock_bh(&rds_sock_lock);

	list_add_tail(&rs->rs_item, &rds_sock_list);

	list_add_tail(&rs->rs_item, &rds_sock_list);

net/rds/connection.c

+11 −9
Original line number Original line Diff line number Diff line
@@ -84,7 +84,7 @@ static struct rds_connection *rds_conn_lookup(struct net *net, 
					      const struct in6_addr *laddr,

					      const struct in6_addr *laddr,

					      const struct in6_addr *faddr,

					      const struct in6_addr *faddr,

					      struct rds_transport *trans,

					      struct rds_transport *trans,

					      int dev_if)

					      u8 tos, int dev_if)

{

{

	struct rds_connection *conn, *ret = NULL;

	struct rds_connection *conn, *ret = NULL;
@@ -92,6 +92,7 @@ static struct rds_connection *rds_conn_lookup(struct net *net, 
		if (ipv6_addr_equal(&conn->c_faddr, faddr) &&

		if (ipv6_addr_equal(&conn->c_faddr, faddr) &&

		    ipv6_addr_equal(&conn->c_laddr, laddr) &&

		    ipv6_addr_equal(&conn->c_laddr, laddr) &&

		    conn->c_trans == trans &&

		    conn->c_trans == trans &&

		    conn->c_tos == tos &&

		    net == rds_conn_net(conn) &&

		    net == rds_conn_net(conn) &&

		    conn->c_dev_if == dev_if) {

		    conn->c_dev_if == dev_if) {

			ret = conn;

			ret = conn;
@@ -160,7 +161,7 @@ static struct rds_connection *__rds_conn_create(struct net *net, 
						const struct in6_addr *laddr,

						const struct in6_addr *laddr,

						const struct in6_addr *faddr,

						const struct in6_addr *faddr,

						struct rds_transport *trans,

						struct rds_transport *trans,

						gfp_t gfp,

						gfp_t gfp, u8 tos,

						int is_outgoing,

						int is_outgoing,

						int dev_if)

						int dev_if)

{

{
@@ -172,7 +173,7 @@ static struct rds_connection *__rds_conn_create(struct net *net, 
	int npaths = (trans->t_mp_capable ? RDS_MPATH_WORKERS : 1);

	int npaths = (trans->t_mp_capable ? RDS_MPATH_WORKERS : 1);





	rcu_read_lock();

	rcu_read_lock();

	conn = rds_conn_lookup(net, head, laddr, faddr, trans, dev_if);

	conn = rds_conn_lookup(net, head, laddr, faddr, trans, tos, dev_if);

	if (conn &&

	if (conn &&

	    conn->c_loopback &&

	    conn->c_loopback &&

	    conn->c_trans != &rds_loop_transport &&

	    conn->c_trans != &rds_loop_transport &&
@@ -206,6 +207,7 @@ static struct rds_connection *__rds_conn_create(struct net *net, 
	conn->c_isv6 = !ipv6_addr_v4mapped(laddr);

	conn->c_isv6 = !ipv6_addr_v4mapped(laddr);

	conn->c_faddr = *faddr;

	conn->c_faddr = *faddr;

	conn->c_dev_if = dev_if;

	conn->c_dev_if = dev_if;

	conn->c_tos = tos;





#if IS_ENABLED(CONFIG_IPV6)

#if IS_ENABLED(CONFIG_IPV6)

	/* If the local address is link local, set c_bound_if to be the

	/* If the local address is link local, set c_bound_if to be the
@@ -298,7 +300,7 @@ static struct rds_connection *__rds_conn_create(struct net *net, 
		struct rds_connection *found;

		struct rds_connection *found;





		found = rds_conn_lookup(net, head, laddr, faddr, trans,

		found = rds_conn_lookup(net, head, laddr, faddr, trans,

					dev_if);

					tos, dev_if);

		if (found) {

		if (found) {

			struct rds_conn_path *cp;

			struct rds_conn_path *cp;

			int i;

			int i;
@@ -333,10 +335,10 @@ out: 
struct rds_connection *rds_conn_create(struct net *net,

struct rds_connection *rds_conn_create(struct net *net,

				       const struct in6_addr *laddr,

				       const struct in6_addr *laddr,

				       const struct in6_addr *faddr,

				       const struct in6_addr *faddr,

				       struct rds_transport *trans, gfp_t gfp,

				       struct rds_transport *trans, u8 tos,

				       int dev_if)

				       gfp_t gfp, int dev_if)

{

{

	return __rds_conn_create(net, laddr, faddr, trans, gfp, 0, dev_if);

	return __rds_conn_create(net, laddr, faddr, trans, gfp, tos, 0, dev_if);

}

}

EXPORT_SYMBOL_GPL(rds_conn_create);

EXPORT_SYMBOL_GPL(rds_conn_create);
@@ -344,9 +346,9 @@ struct rds_connection *rds_conn_create_outgoing(struct net *net, 
						const struct in6_addr *laddr,

						const struct in6_addr *laddr,

						const struct in6_addr *faddr,

						const struct in6_addr *faddr,

						struct rds_transport *trans,

						struct rds_transport *trans,

						gfp_t gfp, int dev_if)

						u8 tos, gfp_t gfp, int dev_if)

{

{

	return __rds_conn_create(net, laddr, faddr, trans, gfp, 1, dev_if);

	return __rds_conn_create(net, laddr, faddr, trans, gfp, tos, 1, dev_if);

}

}

EXPORT_SYMBOL_GPL(rds_conn_create_outgoing);

EXPORT_SYMBOL_GPL(rds_conn_create_outgoing);

net/rds/ib.c

+1 −0
Original line number Original line Diff line number Diff line
@@ -301,6 +301,7 @@ static int rds_ib_conn_info_visitor(struct rds_connection *conn, 




	iinfo->src_addr = conn->c_laddr.s6_addr32[3];

	iinfo->src_addr = conn->c_laddr.s6_addr32[3];

	iinfo->dst_addr = conn->c_faddr.s6_addr32[3];

	iinfo->dst_addr = conn->c_faddr.s6_addr32[3];

	iinfo->tos = conn->c_tos;





	memset(&iinfo->src_gid, 0, sizeof(iinfo->src_gid));

	memset(&iinfo->src_gid, 0, sizeof(iinfo->src_gid));

	memset(&iinfo->dst_gid, 0, sizeof(iinfo->dst_gid));

	memset(&iinfo->dst_gid, 0, sizeof(iinfo->dst_gid));

net/rds/ib_cm.c

+1 −1
Original line number Original line Diff line number Diff line
@@ -786,7 +786,7 @@ int rds_ib_cm_handle_connect(struct rdma_cm_id *cm_id, 




	/* RDS/IB is not currently netns aware, thus init_net */

	/* RDS/IB is not currently netns aware, thus init_net */

	conn = rds_conn_create(&init_net, daddr6, saddr6,

	conn = rds_conn_create(&init_net, daddr6, saddr6,

			       &rds_ib_transport, GFP_KERNEL, ifindex);

			       &rds_ib_transport, 0, GFP_KERNEL, ifindex);

	if (IS_ERR(conn)) {

	if (IS_ERR(conn)) {

		rdsdebug("rds_conn_create failed (%ld)\n", PTR_ERR(conn));

		rdsdebug("rds_conn_create failed (%ld)\n", PTR_ERR(conn));

		conn = NULL;

		conn = NULL;

CRA Git | Maintained and supported by SUSTech CRA and CCSE