[linux-block.git] / net / mptcp / mptcp_diag.c

// SPDX-License-Identifier: GPL-2.0
/* MPTCP socket monitoring support
 *
 * Copyright (c) 2020 Red Hat
 *
 * Author: Paolo Abeni <pabeni@redhat.com>
 */

#include <linux/kernel.h>
#include <linux/net.h>
#include <linux/inet_diag.h>
#include <net/netlink.h>
#include <uapi/linux/mptcp.h>
#include "protocol.h"

static int sk_diag_dump(struct sock *sk, struct sk_buff *skb,
			struct netlink_callback *cb,
			const struct inet_diag_req_v2 *req,
			struct nlattr *bc, bool net_admin)
{
	if (!inet_diag_bc_sk(bc, sk))
		return 0;

	return inet_sk_diag_fill(sk, inet_csk(sk), skb, cb, req, NLM_F_MULTI,
				 net_admin);
}

static int mptcp_diag_dump_one(struct netlink_callback *cb,
			       const struct inet_diag_req_v2 *req)
{
	struct sk_buff *in_skb = cb->skb;
	struct mptcp_sock *msk = NULL;
	struct sk_buff *rep;
	int err = -ENOENT;
	struct net *net;
	struct sock *sk;

	net = sock_net(in_skb->sk);
	msk = mptcp_token_get_sock(net, req->id.idiag_cookie[0]);
	if (!msk)
		goto out_nosk;

	err = -ENOMEM;
	sk = (struct sock *)msk;
	rep = nlmsg_new(nla_total_size(sizeof(struct inet_diag_msg)) +
			inet_diag_msg_attrs_size() +
			nla_total_size(sizeof(struct mptcp_info)) +
			nla_total_size(sizeof(struct inet_diag_meminfo)) + 64,
			GFP_KERNEL);
	if (!rep)
		goto out;

	err = inet_sk_diag_fill(sk, inet_csk(sk), rep, cb, req, 0,
				netlink_net_capable(in_skb, CAP_NET_ADMIN));
	if (err < 0) {
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(rep);
		goto out;
	}
	err = nlmsg_unicast(net->diag_nlsk, rep, NETLINK_CB(in_skb).portid);

out:
	sock_put(sk);

out_nosk:
	return err;
}

struct mptcp_diag_ctx {
	long s_slot;
	long s_num;
	unsigned int l_slot;
	unsigned int l_num;
};

static void mptcp_diag_dump_listeners(struct sk_buff *skb, struct netlink_callback *cb,
				      const struct inet_diag_req_v2 *r,
				      bool net_admin)
{
	struct inet_diag_dump_data *cb_data = cb->data;
	struct mptcp_diag_ctx *diag_ctx = (void *)cb->ctx;
	struct nlattr *bc = cb_data->inet_diag_nla_bc;
	struct net *net = sock_net(skb->sk);
	int i;

	for (i = diag_ctx->l_slot; i < INET_LHTABLE_SIZE; i++) {
		struct inet_listen_hashbucket *ilb;
		struct hlist_nulls_node *node;
		struct sock *sk;
		int num = 0;

		ilb = &tcp_hashinfo.listening_hash[i];

		rcu_read_lock();
		spin_lock(&ilb->lock);
		sk_nulls_for_each(sk, node, &ilb->nulls_head) {
			const struct mptcp_subflow_context *ctx = mptcp_subflow_ctx(sk);
			struct inet_sock *inet = inet_sk(sk);
			int ret;

			if (num < diag_ctx->l_num)
				goto next_listen;

			if (!ctx || strcmp(inet_csk(sk)->icsk_ulp_ops->name, "mptcp"))
				goto next_listen;

			sk = ctx->conn;
			if (!sk || !net_eq(sock_net(sk), net))
				goto next_listen;

			if (r->sdiag_family != AF_UNSPEC &&
			    sk->sk_family != r->sdiag_family)
				goto next_listen;

			if (r->id.idiag_sport != inet->inet_sport &&
			    r->id.idiag_sport)
				goto next_listen;

			if (!refcount_inc_not_zero(&sk->sk_refcnt))
				goto next_listen;

			ret = sk_diag_dump(sk, skb, cb, r, bc, net_admin);

			sock_put(sk);

			if (ret < 0) {
				spin_unlock(&ilb->lock);
				rcu_read_unlock();
				diag_ctx->l_slot = i;
				diag_ctx->l_num = num;
				return;
			}
			diag_ctx->l_num = num + 1;
			num = 0;
next_listen:
			++num;
		}
		spin_unlock(&ilb->lock);
		rcu_read_unlock();

		cond_resched();
		diag_ctx->l_num = 0;
	}

	diag_ctx->l_num = 0;
	diag_ctx->l_slot = i;
}

static void mptcp_diag_dump(struct sk_buff *skb, struct netlink_callback *cb,
			    const struct inet_diag_req_v2 *r)
{
	bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
	struct mptcp_diag_ctx *diag_ctx = (void *)cb->ctx;
	struct net *net = sock_net(skb->sk);
	struct inet_diag_dump_data *cb_data;
	struct mptcp_sock *msk;
	struct nlattr *bc;

	BUILD_BUG_ON(sizeof(cb->ctx) < sizeof(*diag_ctx));

	cb_data = cb->data;
	bc = cb_data->inet_diag_nla_bc;

	while ((msk = mptcp_token_iter_next(net, &diag_ctx->s_slot,
					    &diag_ctx->s_num)) != NULL) {
		struct inet_sock *inet = (struct inet_sock *)msk;
		struct sock *sk = (struct sock *)msk;
		int ret = 0;

		if (!(r->idiag_states & (1 << sk->sk_state)))
			goto next;
		if (r->sdiag_family != AF_UNSPEC &&
		    sk->sk_family != r->sdiag_family)
			goto next;
		if (r->id.idiag_sport != inet->inet_sport &&
		    r->id.idiag_sport)
			goto next;
		if (r->id.idiag_dport != inet->inet_dport &&
		    r->id.idiag_dport)
			goto next;

		ret = sk_diag_dump(sk, skb, cb, r, bc, net_admin);
next:
		sock_put(sk);
		if (ret < 0) {
			/* will retry on the same position */
			diag_ctx->s_num--;
			break;
		}
		cond_resched();
	}

	if ((r->idiag_states & TCPF_LISTEN) && r->id.idiag_dport == 0)
		mptcp_diag_dump_listeners(skb, cb, r, net_admin);
}

static void mptcp_diag_get_info(struct sock *sk, struct inet_diag_msg *r,
				void *_info)
{
	struct mptcp_sock *msk = mptcp_sk(sk);
	struct mptcp_info *info = _info;

	r->idiag_rqueue = sk_rmem_alloc_get(sk);
	r->idiag_wqueue = sk_wmem_alloc_get(sk);

	if (inet_sk_state_load(sk) == TCP_LISTEN) {
		struct sock *lsk = READ_ONCE(msk->first);

		if (lsk) {
			/* override with settings from tcp listener,
			 * so Send-Q will show accept queue.
			 */
			r->idiag_rqueue = READ_ONCE(lsk->sk_ack_backlog);
			r->idiag_wqueue = READ_ONCE(lsk->sk_max_ack_backlog);
		}
	}

	if (!info)
		return;

	mptcp_diag_fill_info(msk, info);
}

static const struct inet_diag_handler mptcp_diag_handler = {
	.dump		 = mptcp_diag_dump,
	.dump_one	 = mptcp_diag_dump_one,
	.idiag_get_info  = mptcp_diag_get_info,
	.idiag_type	 = IPPROTO_MPTCP,
	.idiag_info_size = sizeof(struct mptcp_info),
};

static int __init mptcp_diag_init(void)
{
	return inet_diag_register(&mptcp_diag_handler);
}

static void __exit mptcp_diag_exit(void)
{
	inet_diag_unregister(&mptcp_diag_handler);
}

module_init(mptcp_diag_init);
module_exit(mptcp_diag_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS_NET_PF_PROTO_TYPE(PF_NETLINK, NETLINK_SOCK_DIAG, 2-262 /* AF_INET - IPPROTO_MPTCP */);
Commit	Line	Data
ac3b45f6 PA	1	// SPDX-License-Identifier: GPL-2.0
	2	/* MPTCP socket monitoring support
	3	*
	4	* Copyright (c) 2020 Red Hat
	5	*
	6	* Author: Paolo Abeni <pabeni@redhat.com>
	7	*/
	8
	9	#include <linux/kernel.h>
	10	#include <linux/net.h>
	11	#include <linux/inet_diag.h>
	12	#include <net/netlink.h>
	13	#include <uapi/linux/mptcp.h>
	14	#include "protocol.h"
	15
	16	static int sk_diag_dump(struct sock sk, struct sk_buff skb,
	17	struct netlink_callback *cb,
	18	const struct inet_diag_req_v2 *req,
	19	struct nlattr *bc, bool net_admin)
	20	{
	21	if (!inet_diag_bc_sk(bc, sk))
	22	return 0;
	23
	24	return inet_sk_diag_fill(sk, inet_csk(sk), skb, cb, req, NLM_F_MULTI,
	25	net_admin);
	26	}
	27
	28	static int mptcp_diag_dump_one(struct netlink_callback *cb,
	29	const struct inet_diag_req_v2 *req)
	30	{
	31	struct sk_buff *in_skb = cb->skb;
	32	struct mptcp_sock *msk = NULL;
	33	struct sk_buff *rep;
	34	int err = -ENOENT;
	35	struct net *net;
	36	struct sock *sk;
	37
	38	net = sock_net(in_skb->sk);
ea1300b9	39	msk = mptcp_token_get_sock(net, req->id.idiag_cookie[0]);
ac3b45f6 PA	40	if (!msk)
	41	goto out_nosk;
	42
	43	err = -ENOMEM;
	44	sk = (struct sock *)msk;
	45	rep = nlmsg_new(nla_total_size(sizeof(struct inet_diag_msg)) +
	46	inet_diag_msg_attrs_size() +
	47	nla_total_size(sizeof(struct mptcp_info)) +
	48	nla_total_size(sizeof(struct inet_diag_meminfo)) + 64,
	49	GFP_KERNEL);
	50	if (!rep)
	51	goto out;
	52
	53	err = inet_sk_diag_fill(sk, inet_csk(sk), rep, cb, req, 0,
	54	netlink_net_capable(in_skb, CAP_NET_ADMIN));
	55	if (err < 0) {
	56	WARN_ON(err == -EMSGSIZE);
	57	kfree_skb(rep);
	58	goto out;
	59	}
01757f53 YD	60	err = nlmsg_unicast(net->diag_nlsk, rep, NETLINK_CB(in_skb).portid);
01757f53 YD	61
ac3b45f6 PA	62	out:
	63	sock_put(sk);
	64
	65	out_nosk:
	66	return err;
	67	}
	68
6b9ea5c8 FW	69	struct mptcp_diag_ctx {
	70	long s_slot;
	71	long s_num;
4fa39b70 FW	72	unsigned int l_slot;
4fa39b70 FW	73	unsigned int l_num;
6b9ea5c8 FW	74	};
6b9ea5c8 FW	75
4fa39b70 FW	76	static void mptcp_diag_dump_listeners(struct sk_buff skb, struct netlink_callback cb,
	77	const struct inet_diag_req_v2 *r,
	78	bool net_admin)
	79	{
	80	struct inet_diag_dump_data *cb_data = cb->data;
	81	struct mptcp_diag_ctx diag_ctx = (void )cb->ctx;
	82	struct nlattr *bc = cb_data->inet_diag_nla_bc;
	83	struct net *net = sock_net(skb->sk);
	84	int i;
	85
	86	for (i = diag_ctx->l_slot; i < INET_LHTABLE_SIZE; i++) {
	87	struct inet_listen_hashbucket *ilb;
	88	struct hlist_nulls_node *node;
	89	struct sock *sk;
	90	int num = 0;
	91
	92	ilb = &tcp_hashinfo.listening_hash[i];
	93
	94	rcu_read_lock();
	95	spin_lock(&ilb->lock);
	96	sk_nulls_for_each(sk, node, &ilb->nulls_head) {
	97	const struct mptcp_subflow_context *ctx = mptcp_subflow_ctx(sk);
	98	struct inet_sock *inet = inet_sk(sk);
	99	int ret;
	100
	101	if (num < diag_ctx->l_num)
	102	goto next_listen;
	103
	104	if (!ctx \|\| strcmp(inet_csk(sk)->icsk_ulp_ops->name, "mptcp"))
	105	goto next_listen;
	106
	107	sk = ctx->conn;
	108	if (!sk \|\| !net_eq(sock_net(sk), net))
	109	goto next_listen;
	110
	111	if (r->sdiag_family != AF_UNSPEC &&
	112	sk->sk_family != r->sdiag_family)
	113	goto next_listen;
	114
	115	if (r->id.idiag_sport != inet->inet_sport &&
	116	r->id.idiag_sport)
	117	goto next_listen;
	118
	119	if (!refcount_inc_not_zero(&sk->sk_refcnt))
	120	goto next_listen;
	121
	122	ret = sk_diag_dump(sk, skb, cb, r, bc, net_admin);
	123
	124	sock_put(sk);
	125
	126	if (ret < 0) {
	127	spin_unlock(&ilb->lock);
	128	rcu_read_unlock();
	129	diag_ctx->l_slot = i;
	130	diag_ctx->l_num = num;
	131	return;
	132	}
	133	diag_ctx->l_num = num + 1;
	134	num = 0;
	135	next_listen:
	136	++num;
	137	}
	138	spin_unlock(&ilb->lock);
	139	rcu_read_unlock();
140
141	cond_resched();
142	diag_ctx->l_num = 0;
143	}
144
145	diag_ctx->l_num = 0;
146	diag_ctx->l_slot = i;
147	}
148
ac3b45f6 PA	149	static void mptcp_diag_dump(struct sk_buff skb, struct netlink_callback cb,
	150	const struct inet_diag_req_v2 *r)
	151	{
	152	bool net_admin = netlink_net_capable(cb->skb, CAP_NET_ADMIN);
6b9ea5c8	153	struct mptcp_diag_ctx diag_ctx = (void )cb->ctx;
ac3b45f6 PA	154	struct net *net = sock_net(skb->sk);
	155	struct inet_diag_dump_data *cb_data;
	156	struct mptcp_sock *msk;
	157	struct nlattr *bc;
	158
6b9ea5c8 FW	159	BUILD_BUG_ON(sizeof(cb->ctx) < sizeof(*diag_ctx));
6b9ea5c8 FW	160
ac3b45f6 PA	161	cb_data = cb->data;
	162	bc = cb_data->inet_diag_nla_bc;
	163
6b9ea5c8 FW	164	while ((msk = mptcp_token_iter_next(net, &diag_ctx->s_slot,
6b9ea5c8 FW	165	&diag_ctx->s_num)) != NULL) {
ac3b45f6 PA	166	struct inet_sock inet = (struct inet_sock )msk;
	167	struct sock sk = (struct sock )msk;
	168	int ret = 0;
	169
	170	if (!(r->idiag_states & (1 << sk->sk_state)))
	171	goto next;
	172	if (r->sdiag_family != AF_UNSPEC &&
	173	sk->sk_family != r->sdiag_family)
	174	goto next;
	175	if (r->id.idiag_sport != inet->inet_sport &&
	176	r->id.idiag_sport)
	177	goto next;
	178	if (r->id.idiag_dport != inet->inet_dport &&
	179	r->id.idiag_dport)
	180	goto next;
	181
	182	ret = sk_diag_dump(sk, skb, cb, r, bc, net_admin);
	183	next:
	184	sock_put(sk);
	185	if (ret < 0) {
	186	/* will retry on the same position */
6b9ea5c8	187	diag_ctx->s_num--;
ac3b45f6 PA	188	break;
	189	}
	190	cond_resched();
	191	}
4fa39b70 FW	192
	193	if ((r->idiag_states & TCPF_LISTEN) && r->id.idiag_dport == 0)
	194	mptcp_diag_dump_listeners(skb, cb, r, net_admin);
ac3b45f6 PA	195	}
	196
	197	static void mptcp_diag_get_info(struct sock sk, struct inet_diag_msg r,
	198	void *_info)
	199	{
	200	struct mptcp_sock *msk = mptcp_sk(sk);
	201	struct mptcp_info *info = _info;
ac3b45f6 PA	202
	203	r->idiag_rqueue = sk_rmem_alloc_get(sk);
	204	r->idiag_wqueue = sk_wmem_alloc_get(sk);
4fa39b70 FW	205
	206	if (inet_sk_state_load(sk) == TCP_LISTEN) {
	207	struct sock *lsk = READ_ONCE(msk->first);
	208
	209	if (lsk) {
	210	/* override with settings from tcp listener,
	211	* so Send-Q will show accept queue.
	212	*/
	213	r->idiag_rqueue = READ_ONCE(lsk->sk_ack_backlog);
	214	r->idiag_wqueue = READ_ONCE(lsk->sk_max_ack_backlog);
	215	}
	216	}
	217
ac3b45f6 PA	218	if (!info)
	219	return;
	220
61bc6e82	221	mptcp_diag_fill_info(msk, info);
ac3b45f6 PA	222	}
	223
	224	static const struct inet_diag_handler mptcp_diag_handler = {
	225	.dump = mptcp_diag_dump,
	226	.dump_one = mptcp_diag_dump_one,
	227	.idiag_get_info = mptcp_diag_get_info,
	228	.idiag_type = IPPROTO_MPTCP,
	229	.idiag_info_size = sizeof(struct mptcp_info),
	230	};
	231
	232	static int __init mptcp_diag_init(void)
	233	{
	234	return inet_diag_register(&mptcp_diag_handler);
	235	}
	236
	237	static void __exit mptcp_diag_exit(void)
	238	{
	239	inet_diag_unregister(&mptcp_diag_handler);
	240	}
	241
	242	module_init(mptcp_diag_init);
	243	module_exit(mptcp_diag_exit);
	244	MODULE_LICENSE("GPL");
	245	MODULE_ALIAS_NET_PF_PROTO_TYPE(PF_NETLINK, NETLINK_SOCK_DIAG, 2-262 /* AF_INET - IPPROTO_MPTCP */);