[linux.git] / net / core / sysctl_net_core.c

// SPDX-License-Identifier: GPL-2.0
/* -*- linux-c -*-
 * sysctl_net_core.c: sysctl interface to net core subsystem.
 *
 * Begun April 1, 1996, Mike Shaver.
 * Added /proc/sys/net/core directory entry (empty =) ). [MS]
 */

#include <linux/mm.h>
#include <linux/sysctl.h>
#include <linux/module.h>
#include <linux/socket.h>
#include <linux/netdevice.h>
#include <linux/ratelimit.h>
#include <linux/vmalloc.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/kmemleak.h>

#include <net/ip.h>
#include <net/sock.h>
#include <net/net_ratelimit.h>
#include <net/busy_poll.h>
#include <net/pkt_sched.h>

static int zero = 0;
static int one = 1;
static int two __maybe_unused = 2;
static int min_sndbuf = SOCK_MIN_SNDBUF;
static int min_rcvbuf = SOCK_MIN_RCVBUF;
static int max_skb_frags = MAX_SKB_FRAGS;

static int net_msg_warn;	/* Unused, but still a sysctl */

#ifdef CONFIG_RPS
static int rps_sock_flow_sysctl(struct ctl_table *table, int write,
				void __user *buffer, size_t *lenp, loff_t *ppos)
{
	unsigned int orig_size, size;
	int ret, i;
	struct ctl_table tmp = {
		.data = &size,
		.maxlen = sizeof(size),
		.mode = table->mode
	};
	struct rps_sock_flow_table *orig_sock_table, *sock_table;
	static DEFINE_MUTEX(sock_flow_mutex);

	mutex_lock(&sock_flow_mutex);

	orig_sock_table = rcu_dereference_protected(rps_sock_flow_table,
					lockdep_is_held(&sock_flow_mutex));
	size = orig_size = orig_sock_table ? orig_sock_table->mask + 1 : 0;

	ret = proc_dointvec(&tmp, write, buffer, lenp, ppos);

	if (write) {
		if (size) {
			if (size > 1<<29) {
				/* Enforce limit to prevent overflow */
				mutex_unlock(&sock_flow_mutex);
				return -EINVAL;
			}
			size = roundup_pow_of_two(size);
			if (size != orig_size) {
				sock_table =
				    vmalloc(RPS_SOCK_FLOW_TABLE_SIZE(size));
				if (!sock_table) {
					mutex_unlock(&sock_flow_mutex);
					return -ENOMEM;
				}
				rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1;
				sock_table->mask = size - 1;
			} else
				sock_table = orig_sock_table;

			for (i = 0; i < size; i++)
				sock_table->ents[i] = RPS_NO_CPU;
		} else
			sock_table = NULL;

		if (sock_table != orig_sock_table) {
			rcu_assign_pointer(rps_sock_flow_table, sock_table);
			if (sock_table) {
				static_key_slow_inc(&rps_needed);
				static_key_slow_inc(&rfs_needed);
			}
			if (orig_sock_table) {
				static_key_slow_dec(&rps_needed);
				static_key_slow_dec(&rfs_needed);
				synchronize_rcu();
				vfree(orig_sock_table);
			}
		}
	}

	mutex_unlock(&sock_flow_mutex);

	return ret;
}
#endif /* CONFIG_RPS */

#ifdef CONFIG_NET_FLOW_LIMIT
static DEFINE_MUTEX(flow_limit_update_mutex);

static int flow_limit_cpu_sysctl(struct ctl_table *table, int write,
				 void __user *buffer, size_t *lenp,
				 loff_t *ppos)
{
	struct sd_flow_limit *cur;
	struct softnet_data *sd;
	cpumask_var_t mask;
	int i, len, ret = 0;

	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
		return -ENOMEM;

	if (write) {
		ret = cpumask_parse_user(buffer, *lenp, mask);
		if (ret)
			goto done;

		mutex_lock(&flow_limit_update_mutex);
		len = sizeof(*cur) + netdev_flow_limit_table_len;
		for_each_possible_cpu(i) {
			sd = &per_cpu(softnet_data, i);
			cur = rcu_dereference_protected(sd->flow_limit,
				     lockdep_is_held(&flow_limit_update_mutex));
			if (cur && !cpumask_test_cpu(i, mask)) {
				RCU_INIT_POINTER(sd->flow_limit, NULL);
				synchronize_rcu();
				kfree(cur);
			} else if (!cur && cpumask_test_cpu(i, mask)) {
				cur = kzalloc_node(len, GFP_KERNEL,
						   cpu_to_node(i));
				if (!cur) {
					/* not unwinding previous changes */
					ret = -ENOMEM;
					goto write_unlock;
				}
				cur->num_buckets = netdev_flow_limit_table_len;
				rcu_assign_pointer(sd->flow_limit, cur);
			}
		}
write_unlock:
		mutex_unlock(&flow_limit_update_mutex);
	} else {
		char kbuf[128];

		if (*ppos || !*lenp) {
			*lenp = 0;
			goto done;
		}

		cpumask_clear(mask);
		rcu_read_lock();
		for_each_possible_cpu(i) {
			sd = &per_cpu(softnet_data, i);
			if (rcu_dereference(sd->flow_limit))
				cpumask_set_cpu(i, mask);
		}
		rcu_read_unlock();

		len = min(sizeof(kbuf) - 1, *lenp);
		len = scnprintf(kbuf, len, "%*pb", cpumask_pr_args(mask));
		if (!len) {
			*lenp = 0;
			goto done;
		}
		if (len < *lenp)
			kbuf[len++] = '\n';
		if (copy_to_user(buffer, kbuf, len)) {
			ret = -EFAULT;
			goto done;
		}
		*lenp = len;
		*ppos += len;
	}

done:
	free_cpumask_var(mask);
	return ret;
}

static int flow_limit_table_len_sysctl(struct ctl_table *table, int write,
				       void __user *buffer, size_t *lenp,
				       loff_t *ppos)
{
	unsigned int old, *ptr;
	int ret;

	mutex_lock(&flow_limit_update_mutex);

	ptr = table->data;
	old = *ptr;
	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	if (!ret && write && !is_power_of_2(*ptr)) {
		*ptr = old;
		ret = -EINVAL;
	}

	mutex_unlock(&flow_limit_update_mutex);
	return ret;
}
#endif /* CONFIG_NET_FLOW_LIMIT */

#ifdef CONFIG_NET_SCHED
static int set_default_qdisc(struct ctl_table *table, int write,
			     void __user *buffer, size_t *lenp, loff_t *ppos)
{
	char id[IFNAMSIZ];
	struct ctl_table tbl = {
		.data = id,
		.maxlen = IFNAMSIZ,
	};
	int ret;

	qdisc_get_default(id, IFNAMSIZ);

	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
	if (write && ret == 0)
		ret = qdisc_set_default(id);
	return ret;
}
#endif

static int proc_do_dev_weight(struct ctl_table *table, int write,
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int ret;

	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	if (ret != 0)
		return ret;

	dev_rx_weight = weight_p * dev_weight_rx_bias;
	dev_tx_weight = weight_p * dev_weight_tx_bias;

	return ret;
}

static int proc_do_rss_key(struct ctl_table *table, int write,
			   void __user *buffer, size_t *lenp, loff_t *ppos)
{
	struct ctl_table fake_table;
	char buf[NETDEV_RSS_KEY_LEN * 3];

	snprintf(buf, sizeof(buf), "%*phC", NETDEV_RSS_KEY_LEN, netdev_rss_key);
	fake_table.data = buf;
	fake_table.maxlen = sizeof(buf);
	return proc_dostring(&fake_table, write, buffer, lenp, ppos);
}

#ifdef CONFIG_BPF_JIT
static int proc_dointvec_minmax_bpf_enable(struct ctl_table *table, int write,
					   void __user *buffer, size_t *lenp,
					   loff_t *ppos)
{
	int ret, jit_enable = *(int *)table->data;
	struct ctl_table tmp = *table;

	if (write && !capable(CAP_SYS_ADMIN))
		return -EPERM;

	tmp.data = &jit_enable;
	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
	if (write && !ret) {
		if (jit_enable < 2 ||
		    (jit_enable == 2 && bpf_dump_raw_ok())) {
			*(int *)table->data = jit_enable;
			if (jit_enable == 2)
				pr_warn("bpf_jit_enable = 2 was set! NEVER use this in production, only for JIT debugging!\n");
		} else {
			ret = -EPERM;
		}
	}
	return ret;
}

# ifdef CONFIG_HAVE_EBPF_JIT
static int
proc_dointvec_minmax_bpf_restricted(struct ctl_table *table, int write,
				    void __user *buffer, size_t *lenp,
				    loff_t *ppos)
{
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	return proc_dointvec_minmax(table, write, buffer, lenp, ppos);
}
# endif
#endif

static struct ctl_table net_core_table[] = {
#ifdef CONFIG_NET
	{
		.procname	= "wmem_max",
		.data		= &sysctl_wmem_max,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &min_sndbuf,
	},
	{
		.procname	= "rmem_max",
		.data		= &sysctl_rmem_max,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &min_rcvbuf,
	},
	{
		.procname	= "wmem_default",
		.data		= &sysctl_wmem_default,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &min_sndbuf,
	},
	{
		.procname	= "rmem_default",
		.data		= &sysctl_rmem_default,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &min_rcvbuf,
	},
	{
		.procname	= "dev_weight",
		.data		= &weight_p,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_dev_weight,
	},
	{
		.procname	= "dev_weight_rx_bias",
		.data		= &dev_weight_rx_bias,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_dev_weight,
	},
	{
		.procname	= "dev_weight_tx_bias",
		.data		= &dev_weight_tx_bias,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_do_dev_weight,
	},
	{
		.procname	= "netdev_max_backlog",
		.data		= &netdev_max_backlog,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "netdev_rss_key",
		.data		= &netdev_rss_key,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= proc_do_rss_key,
	},
#ifdef CONFIG_BPF_JIT
	{
		.procname	= "bpf_jit_enable",
		.data		= &bpf_jit_enable,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax_bpf_enable,
# ifdef CONFIG_BPF_JIT_ALWAYS_ON
		.extra1		= &one,
		.extra2		= &one,
# else
		.extra1		= &zero,
		.extra2		= &two,
# endif
	},
# ifdef CONFIG_HAVE_EBPF_JIT
	{
		.procname	= "bpf_jit_harden",
		.data		= &bpf_jit_harden,
		.maxlen		= sizeof(int),
		.mode		= 0600,
		.proc_handler	= proc_dointvec_minmax_bpf_restricted,
		.extra1		= &zero,
		.extra2		= &two,
	},
	{
		.procname	= "bpf_jit_kallsyms",
		.data		= &bpf_jit_kallsyms,
		.maxlen		= sizeof(int),
		.mode		= 0600,
		.proc_handler	= proc_dointvec_minmax_bpf_restricted,
		.extra1		= &zero,
		.extra2		= &one,
	},
# endif
#endif
	{
		.procname	= "netdev_tstamp_prequeue",
		.data		= &netdev_tstamp_prequeue,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "message_cost",
		.data		= &net_ratelimit_state.interval,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_jiffies,
	},
	{
		.procname	= "message_burst",
		.data		= &net_ratelimit_state.burst,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec,
	},
	{
		.procname	= "optmem_max",
		.data		= &sysctl_optmem_max,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "tstamp_allow_data",
		.data		= &sysctl_tstamp_allow_data,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
		.extra2		= &one
	},
#ifdef CONFIG_RPS
	{
		.procname	= "rps_sock_flow_entries",
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= rps_sock_flow_sysctl
	},
#endif
#ifdef CONFIG_NET_FLOW_LIMIT
	{
		.procname	= "flow_limit_cpu_bitmap",
		.mode		= 0644,
		.proc_handler	= flow_limit_cpu_sysctl
	},
	{
		.procname	= "flow_limit_table_len",
		.data		= &netdev_flow_limit_table_len,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= flow_limit_table_len_sysctl
	},
#endif /* CONFIG_NET_FLOW_LIMIT */
#ifdef CONFIG_NET_RX_BUSY_POLL
	{
		.procname	= "busy_poll",
		.data		= &sysctl_net_busy_poll,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
	},
	{
		.procname	= "busy_read",
		.data		= &sysctl_net_busy_read,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
	},
#endif
#ifdef CONFIG_NET_SCHED
	{
		.procname	= "default_qdisc",
		.mode		= 0644,
		.maxlen		= IFNAMSIZ,
		.proc_handler	= set_default_qdisc
	},
#endif
#endif /* CONFIG_NET */
	{
		.procname	= "netdev_budget",
		.data		= &netdev_budget,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "warnings",
		.data		= &net_msg_warn,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec
	},
	{
		.procname	= "max_skb_frags",
		.data		= &sysctl_max_skb_frags,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &one,
		.extra2		= &max_skb_frags,
	},
	{
		.procname	= "netdev_budget_usecs",
		.data		= &netdev_budget_usecs,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= &zero,
	},
	{ }
};

static struct ctl_table netns_core_table[] = {
	{
		.procname	= "somaxconn",
		.data		= &init_net.core.sysctl_somaxconn,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.extra1		= &zero,
		.proc_handler	= proc_dointvec_minmax
	},
	{ }
};

static __net_init int sysctl_core_net_init(struct net *net)
{
	struct ctl_table *tbl;

	tbl = netns_core_table;
	if (!net_eq(net, &init_net)) {
		tbl = kmemdup(tbl, sizeof(netns_core_table), GFP_KERNEL);
		if (tbl == NULL)
			goto err_dup;

		tbl[0].data = &net->core.sysctl_somaxconn;

		/* Don't export any sysctls to unprivileged users */
		if (net->user_ns != &init_user_ns) {
			tbl[0].procname = NULL;
		}
	}

	net->core.sysctl_hdr = register_net_sysctl(net, "net/core", tbl);
	if (net->core.sysctl_hdr == NULL)
		goto err_reg;

	return 0;

err_reg:
	if (tbl != netns_core_table)
		kfree(tbl);
err_dup:
	return -ENOMEM;
}

static __net_exit void sysctl_core_net_exit(struct net *net)
{
	struct ctl_table *tbl;

	tbl = net->core.sysctl_hdr->ctl_table_arg;
	unregister_net_sysctl_table(net->core.sysctl_hdr);
	BUG_ON(tbl == netns_core_table);
	kfree(tbl);
}

static __net_initdata struct pernet_operations sysctl_core_ops = {
	.init = sysctl_core_net_init,
	.exit = sysctl_core_net_exit,
};

static __init int sysctl_core_init(void)
{
	register_net_sysctl(&init_net, "net/core", net_core_table);
	return register_pernet_subsys(&sysctl_core_ops);
}

fs_initcall(sysctl_core_init);
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/* -- linux-c --
	3	* sysctl_net_core.c: sysctl interface to net core subsystem.
	4	*
	5	* Begun April 1, 1996, Mike Shaver.
	6	* Added /proc/sys/net/core directory entry (empty =) ). [MS]
	7	*/
	8
	9	#include <linux/mm.h>
	10	#include <linux/sysctl.h>
1da177e4	11	#include <linux/module.h>
20380731	12	#include <linux/socket.h>
a37ae408	13	#include <linux/netdevice.h>
3fff4c42	14	#include <linux/ratelimit.h>
fec5e652	15	#include <linux/vmalloc.h>
33eb9cfc	16	#include <linux/init.h>
5a0e3ad6	17	#include <linux/slab.h>
7426a564	18	#include <linux/kmemleak.h>
3fff4c42	19
63d819ca	20	#include <net/ip.h>
20380731	21	#include <net/sock.h>
c5c177b4	22	#include <net/net_ratelimit.h>
076bb0c8	23	#include <net/busy_poll.h>
6da7c8fc	24	#include <net/pkt_sched.h>
1da177e4	25
5f671d6b	26	static int zero = 0;
cdda8891	27	static int one = 1;
fa9dd599	28	static int two __maybe_unused = 2;
b1cb59cf AK	29	static int min_sndbuf = SOCK_MIN_SNDBUF;
b1cb59cf AK	30	static int min_rcvbuf = SOCK_MIN_RCVBUF;
5f74f82e	31	static int max_skb_frags = MAX_SKB_FRAGS;
cdda8891	32
ba7a46f1 JP	33	static int net_msg_warn; /* Unused, but still a sysctl */
ba7a46f1 JP	34
fec5e652	35	#ifdef CONFIG_RPS
fe2c6338	36	static int rps_sock_flow_sysctl(struct ctl_table *table, int write,
fec5e652 TH	37	void __user buffer, size_t lenp, loff_t *ppos)
	38	{
	39	unsigned int orig_size, size;
	40	int ret, i;
fe2c6338	41	struct ctl_table tmp = {
fec5e652 TH	42	.data = &size,
	43	.maxlen = sizeof(size),
	44	.mode = table->mode
	45	};
	46	struct rps_sock_flow_table orig_sock_table, sock_table;
	47	static DEFINE_MUTEX(sock_flow_mutex);
	48
	49	mutex_lock(&sock_flow_mutex);
	50
6e3f7faf ED	51	orig_sock_table = rcu_dereference_protected(rps_sock_flow_table,
6e3f7faf ED	52	lockdep_is_held(&sock_flow_mutex));
fec5e652 TH	53	size = orig_size = orig_sock_table ? orig_sock_table->mask + 1 : 0;
	54
	55	ret = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	56
	57	if (write) {
	58	if (size) {
93c1af6c	59	if (size > 1<<29) {
fec5e652 TH	60	/* Enforce limit to prevent overflow */
	61	mutex_unlock(&sock_flow_mutex);
	62	return -EINVAL;
	63	}
	64	size = roundup_pow_of_two(size);
	65	if (size != orig_size) {
	66	sock_table =
	67	vmalloc(RPS_SOCK_FLOW_TABLE_SIZE(size));
	68	if (!sock_table) {
	69	mutex_unlock(&sock_flow_mutex);
	70	return -ENOMEM;
	71	}
567e4b79	72	rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1;
fec5e652 TH	73	sock_table->mask = size - 1;
	74	} else
	75	sock_table = orig_sock_table;
	76
	77	for (i = 0; i < size; i++)
	78	sock_table->ents[i] = RPS_NO_CPU;
	79	} else
	80	sock_table = NULL;
	81
	82	if (sock_table != orig_sock_table) {
	83	rcu_assign_pointer(rps_sock_flow_table, sock_table);
13bfff25	84	if (sock_table) {
c5905afb	85	static_key_slow_inc(&rps_needed);
13bfff25 ED	86	static_key_slow_inc(&rfs_needed);
13bfff25 ED	87	}
adc9300e	88	if (orig_sock_table) {
c5905afb	89	static_key_slow_dec(&rps_needed);
13bfff25	90	static_key_slow_dec(&rfs_needed);
adc9300e ED	91	synchronize_rcu();
	92	vfree(orig_sock_table);
	93	}
fec5e652 TH	94	}
	95	}
	96
	97	mutex_unlock(&sock_flow_mutex);
	98
	99	return ret;
	100	}
	101	#endif /* CONFIG_RPS */
	102
99bbc707 WB	103	#ifdef CONFIG_NET_FLOW_LIMIT
	104	static DEFINE_MUTEX(flow_limit_update_mutex);
	105
fe2c6338	106	static int flow_limit_cpu_sysctl(struct ctl_table *table, int write,
99bbc707 WB	107	void __user buffer, size_t lenp,
	108	loff_t *ppos)
	109	{
	110	struct sd_flow_limit *cur;
	111	struct softnet_data *sd;
	112	cpumask_var_t mask;
	113	int i, len, ret = 0;
	114
	115	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
	116	return -ENOMEM;
	117
	118	if (write) {
	119	ret = cpumask_parse_user(buffer, *lenp, mask);
	120	if (ret)
	121	goto done;
	122
	123	mutex_lock(&flow_limit_update_mutex);
	124	len = sizeof(*cur) + netdev_flow_limit_table_len;
	125	for_each_possible_cpu(i) {
	126	sd = &per_cpu(softnet_data, i);
	127	cur = rcu_dereference_protected(sd->flow_limit,
	128	lockdep_is_held(&flow_limit_update_mutex));
	129	if (cur && !cpumask_test_cpu(i, mask)) {
	130	RCU_INIT_POINTER(sd->flow_limit, NULL);
	131	synchronize_rcu();
	132	kfree(cur);
	133	} else if (!cur && cpumask_test_cpu(i, mask)) {
5b59d467 ED	134	cur = kzalloc_node(len, GFP_KERNEL,
5b59d467 ED	135	cpu_to_node(i));
99bbc707 WB	136	if (!cur) {
	137	/* not unwinding previous changes */
	138	ret = -ENOMEM;
	139	goto write_unlock;
	140	}
	141	cur->num_buckets = netdev_flow_limit_table_len;
	142	rcu_assign_pointer(sd->flow_limit, cur);
	143	}
	144	}
	145	write_unlock:
	146	mutex_unlock(&flow_limit_update_mutex);
	147	} else {
5f121b9a WB	148	char kbuf[128];
5f121b9a WB	149
99bbc707 WB	150	if (ppos \|\| !lenp) {
	151	*lenp = 0;
	152	goto done;
	153	}
	154
	155	cpumask_clear(mask);
	156	rcu_read_lock();
	157	for_each_possible_cpu(i) {
	158	sd = &per_cpu(softnet_data, i);
	159	if (rcu_dereference(sd->flow_limit))
	160	cpumask_set_cpu(i, mask);
	161	}
	162	rcu_read_unlock();
	163
5f121b9a	164	len = min(sizeof(kbuf) - 1, *lenp);
f0906827	165	len = scnprintf(kbuf, len, "%*pb", cpumask_pr_args(mask));
5f121b9a WB	166	if (!len) {
	167	*lenp = 0;
	168	goto done;
	169	}
	170	if (len < *lenp)
	171	kbuf[len++] = '\n';
	172	if (copy_to_user(buffer, kbuf, len)) {
	173	ret = -EFAULT;
	174	goto done;
	175	}
	176	*lenp = len;
	177	*ppos += len;
99bbc707 WB	178	}
	179
	180	done:
	181	free_cpumask_var(mask);
	182	return ret;
	183	}
	184
fe2c6338	185	static int flow_limit_table_len_sysctl(struct ctl_table *table, int write,
99bbc707 WB	186	void __user buffer, size_t lenp,
	187	loff_t *ppos)
	188	{
	189	unsigned int old, *ptr;
	190	int ret;
	191
	192	mutex_lock(&flow_limit_update_mutex);
	193
	194	ptr = table->data;
	195	old = *ptr;
	196	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	197	if (!ret && write && !is_power_of_2(*ptr)) {
	198	*ptr = old;
	199	ret = -EINVAL;
	200	}
	201
	202	mutex_unlock(&flow_limit_update_mutex);
	203	return ret;
	204	}
	205	#endif /* CONFIG_NET_FLOW_LIMIT */
	206
6da7c8fc	207	#ifdef CONFIG_NET_SCHED
	208	static int set_default_qdisc(struct ctl_table *table, int write,
	209	void __user buffer, size_t lenp, loff_t *ppos)
	210	{
	211	char id[IFNAMSIZ];
	212	struct ctl_table tbl = {
	213	.data = id,
	214	.maxlen = IFNAMSIZ,
	215	};
	216	int ret;
	217
	218	qdisc_get_default(id, IFNAMSIZ);
	219
	220	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
	221	if (write && ret == 0)
	222	ret = qdisc_set_default(id);
	223	return ret;
	224	}
	225	#endif
	226
3d48b53f MT	227	static int proc_do_dev_weight(struct ctl_table *table, int write,
	228	void __user buffer, size_t lenp, loff_t *ppos)
	229	{
	230	int ret;
	231
	232	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	233	if (ret != 0)
	234	return ret;
	235
	236	dev_rx_weight = weight_p * dev_weight_rx_bias;
	237	dev_tx_weight = weight_p * dev_weight_tx_bias;
	238
	239	return ret;
	240	}
	241
960fb622 ED	242	static int proc_do_rss_key(struct ctl_table *table, int write,
	243	void __user buffer, size_t lenp, loff_t *ppos)
	244	{
	245	struct ctl_table fake_table;
	246	char buf[NETDEV_RSS_KEY_LEN * 3];
	247
	248	snprintf(buf, sizeof(buf), "%*phC", NETDEV_RSS_KEY_LEN, netdev_rss_key);
	249	fake_table.data = buf;
	250	fake_table.maxlen = sizeof(buf);
	251	return proc_dostring(&fake_table, write, buffer, lenp, ppos);
	252	}
	253
2e4a3098 DB	254	#ifdef CONFIG_BPF_JIT
	255	static int proc_dointvec_minmax_bpf_enable(struct ctl_table *table, int write,
	256	void __user buffer, size_t lenp,
	257	loff_t *ppos)
	258	{
	259	int ret, jit_enable = (int )table->data;
	260	struct ctl_table tmp = *table;
	261
	262	if (write && !capable(CAP_SYS_ADMIN))
	263	return -EPERM;
	264
	265	tmp.data = &jit_enable;
	266	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
	267	if (write && !ret) {
	268	if (jit_enable < 2 \|\|
	269	(jit_enable == 2 && bpf_dump_raw_ok())) {
	270	(int )table->data = jit_enable;
	271	if (jit_enable == 2)
	272	pr_warn("bpf_jit_enable = 2 was set! NEVER use this in production, only for JIT debugging!\n");
	273	} else {
	274	ret = -EPERM;
	275	}
	276	}
	277	return ret;
	278	}
	279
	280	# ifdef CONFIG_HAVE_EBPF_JIT
	281	static int
	282	proc_dointvec_minmax_bpf_restricted(struct ctl_table *table, int write,
	283	void __user buffer, size_t lenp,
	284	loff_t *ppos)
	285	{
	286	if (!capable(CAP_SYS_ADMIN))
	287	return -EPERM;
	288
	289	return proc_dointvec_minmax(table, write, buffer, lenp, ppos);
	290	}
	291	# endif
	292	#endif
	293
33eb9cfc	294	static struct ctl_table net_core_table[] = {
1da177e4 LT	295	#ifdef CONFIG_NET
1da177e4 LT	296	{
1da177e4 LT	297	.procname = "wmem_max",
	298	.data = &sysctl_wmem_max,
	299	.maxlen = sizeof(int),
	300	.mode = 0644,
cdda8891	301	.proc_handler = proc_dointvec_minmax,
b1cb59cf	302	.extra1 = &min_sndbuf,
1da177e4 LT	303	},
1da177e4 LT	304	{
1da177e4 LT	305	.procname = "rmem_max",
	306	.data = &sysctl_rmem_max,
	307	.maxlen = sizeof(int),
	308	.mode = 0644,
cdda8891	309	.proc_handler = proc_dointvec_minmax,
b1cb59cf	310	.extra1 = &min_rcvbuf,
1da177e4 LT	311	},
1da177e4 LT	312	{
1da177e4 LT	313	.procname = "wmem_default",
	314	.data = &sysctl_wmem_default,
	315	.maxlen = sizeof(int),
	316	.mode = 0644,
cdda8891	317	.proc_handler = proc_dointvec_minmax,
b1cb59cf	318	.extra1 = &min_sndbuf,
1da177e4 LT	319	},
1da177e4 LT	320	{
1da177e4 LT	321	.procname = "rmem_default",
	322	.data = &sysctl_rmem_default,
	323	.maxlen = sizeof(int),
	324	.mode = 0644,
cdda8891	325	.proc_handler = proc_dointvec_minmax,
b1cb59cf	326	.extra1 = &min_rcvbuf,
1da177e4 LT	327	},
1da177e4 LT	328	{
1da177e4 LT	329	.procname = "dev_weight",
	330	.data = &weight_p,
	331	.maxlen = sizeof(int),
	332	.mode = 0644,
3d48b53f MT	333	.proc_handler = proc_do_dev_weight,
	334	},
	335	{
	336	.procname = "dev_weight_rx_bias",
	337	.data = &dev_weight_rx_bias,
	338	.maxlen = sizeof(int),
	339	.mode = 0644,
	340	.proc_handler = proc_do_dev_weight,
	341	},
	342	{
	343	.procname = "dev_weight_tx_bias",
	344	.data = &dev_weight_tx_bias,
	345	.maxlen = sizeof(int),
	346	.mode = 0644,
	347	.proc_handler = proc_do_dev_weight,
1da177e4 LT	348	},
1da177e4 LT	349	{
1da177e4 LT	350	.procname = "netdev_max_backlog",
	351	.data = &netdev_max_backlog,
	352	.maxlen = sizeof(int),
	353	.mode = 0644,
6d9f239a	354	.proc_handler = proc_dointvec
1da177e4	355	},
960fb622 ED	356	{
	357	.procname = "netdev_rss_key",
	358	.data = &netdev_rss_key,
	359	.maxlen = sizeof(int),
	360	.mode = 0444,
	361	.proc_handler = proc_do_rss_key,
	362	},
0a14842f ED	363	#ifdef CONFIG_BPF_JIT
	364	{
	365	.procname = "bpf_jit_enable",
	366	.data = &bpf_jit_enable,
	367	.maxlen = sizeof(int),
	368	.mode = 0644,
2e4a3098	369	.proc_handler = proc_dointvec_minmax_bpf_enable,
fa9dd599	370	# ifdef CONFIG_BPF_JIT_ALWAYS_ON
290af866 AS	371	.extra1 = &one,
290af866 AS	372	.extra2 = &one,
fa9dd599 DB	373	# else
	374	.extra1 = &zero,
	375	.extra2 = &two,
	376	# endif
0a14842f	377	},
4f3446bb DB	378	# ifdef CONFIG_HAVE_EBPF_JIT
	379	{
	380	.procname = "bpf_jit_harden",
	381	.data = &bpf_jit_harden,
	382	.maxlen = sizeof(int),
	383	.mode = 0600,
2e4a3098	384	.proc_handler = proc_dointvec_minmax_bpf_restricted,
fa9dd599 DB	385	.extra1 = &zero,
fa9dd599 DB	386	.extra2 = &two,
4f3446bb	387	},
74451e66 DB	388	{
	389	.procname = "bpf_jit_kallsyms",
	390	.data = &bpf_jit_kallsyms,
	391	.maxlen = sizeof(int),
	392	.mode = 0600,
2e4a3098	393	.proc_handler = proc_dointvec_minmax_bpf_restricted,
fa9dd599 DB	394	.extra1 = &zero,
fa9dd599 DB	395	.extra2 = &one,
74451e66	396	},
4f3446bb	397	# endif
0a14842f	398	#endif
3b098e2d ED	399	{
	400	.procname = "netdev_tstamp_prequeue",
	401	.data = &netdev_tstamp_prequeue,
	402	.maxlen = sizeof(int),
	403	.mode = 0644,
	404	.proc_handler = proc_dointvec
	405	},
1da177e4	406	{
1da177e4	407	.procname = "message_cost",
717115e1	408	.data = &net_ratelimit_state.interval,
1da177e4 LT	409	.maxlen = sizeof(int),
1da177e4 LT	410	.mode = 0644,
6d9f239a	411	.proc_handler = proc_dointvec_jiffies,
1da177e4 LT	412	},
1da177e4 LT	413	{
1da177e4	414	.procname = "message_burst",
717115e1	415	.data = &net_ratelimit_state.burst,
1da177e4 LT	416	.maxlen = sizeof(int),
1da177e4 LT	417	.mode = 0644,
6d9f239a	418	.proc_handler = proc_dointvec,
1da177e4 LT	419	},
1da177e4 LT	420	{
1da177e4 LT	421	.procname = "optmem_max",
	422	.data = &sysctl_optmem_max,
	423	.maxlen = sizeof(int),
	424	.mode = 0644,
6d9f239a	425	.proc_handler = proc_dointvec
1da177e4	426	},
b245be1f WB	427	{
	428	.procname = "tstamp_allow_data",
	429	.data = &sysctl_tstamp_allow_data,
	430	.maxlen = sizeof(int),
	431	.mode = 0644,
	432	.proc_handler = proc_dointvec_minmax,
	433	.extra1 = &zero,
	434	.extra2 = &one
	435	},
fec5e652 TH	436	#ifdef CONFIG_RPS
	437	{
	438	.procname = "rps_sock_flow_entries",
	439	.maxlen = sizeof(int),
	440	.mode = 0644,
	441	.proc_handler = rps_sock_flow_sysctl
	442	},
	443	#endif
99bbc707 WB	444	#ifdef CONFIG_NET_FLOW_LIMIT
	445	{
	446	.procname = "flow_limit_cpu_bitmap",
	447	.mode = 0644,
	448	.proc_handler = flow_limit_cpu_sysctl
	449	},
	450	{
	451	.procname = "flow_limit_table_len",
	452	.data = &netdev_flow_limit_table_len,
	453	.maxlen = sizeof(int),
	454	.mode = 0644,
	455	.proc_handler = flow_limit_table_len_sysctl
	456	},
	457	#endif /* CONFIG_NET_FLOW_LIMIT */
e0d1095a	458	#ifdef CONFIG_NET_RX_BUSY_POLL
06021292	459	{
64b0dc51 ET	460	.procname = "busy_poll",
64b0dc51 ET	461	.data = &sysctl_net_busy_poll,
eb6db622	462	.maxlen = sizeof(unsigned int),
06021292	463	.mode = 0644,
95f25521 AD	464	.proc_handler = proc_dointvec_minmax,
95f25521 AD	465	.extra1 = &zero,
06021292	466	},
2d48d67f	467	{
64b0dc51 ET	468	.procname = "busy_read",
64b0dc51 ET	469	.data = &sysctl_net_busy_read,
2d48d67f ET	470	.maxlen = sizeof(unsigned int),
2d48d67f ET	471	.mode = 0644,
95f25521 AD	472	.proc_handler = proc_dointvec_minmax,
95f25521 AD	473	.extra1 = &zero,
2d48d67f	474	},
6da7c8fc	475	#endif
	476	#ifdef CONFIG_NET_SCHED
	477	{
	478	.procname = "default_qdisc",
	479	.mode = 0644,
	480	.maxlen = IFNAMSIZ,
	481	.proc_handler = set_default_qdisc
	482	},
06021292	483	#endif
1da177e4	484	#endif /* CONFIG_NET */
51b0bded	485	{
51b0bded SH	486	.procname = "netdev_budget",
	487	.data = &netdev_budget,
	488	.maxlen = sizeof(int),
	489	.mode = 0644,
6d9f239a	490	.proc_handler = proc_dointvec
51b0bded	491	},
a2a316fd	492	{
a2a316fd SH	493	.procname = "warnings",
	494	.data = &net_msg_warn,
	495	.maxlen = sizeof(int),
	496	.mode = 0644,
6d9f239a	497	.proc_handler = proc_dointvec
a2a316fd	498	},
5f74f82e HWR	499	{
	500	.procname = "max_skb_frags",
	501	.data = &sysctl_max_skb_frags,
	502	.maxlen = sizeof(int),
	503	.mode = 0644,
	504	.proc_handler = proc_dointvec_minmax,
	505	.extra1 = &one,
	506	.extra2 = &max_skb_frags,
	507	},
7acf8a1e MW	508	{
	509	.procname = "netdev_budget_usecs",
	510	.data = &netdev_budget_usecs,
	511	.maxlen = sizeof(unsigned int),
	512	.mode = 0644,
	513	.proc_handler = proc_dointvec_minmax,
	514	.extra1 = &zero,
	515	},
f8572d8f	516	{ }
1da177e4	517	};
33eb9cfc	518
d5a4502e PE	519	static struct ctl_table netns_core_table[] = {
d5a4502e PE	520	{
d5a4502e PE	521	.procname = "somaxconn",
	522	.data = &init_net.core.sysctl_somaxconn,
	523	.maxlen = sizeof(int),
	524	.mode = 0644,
5f671d6b	525	.extra1 = &zero,
5f671d6b	526	.proc_handler = proc_dointvec_minmax
d5a4502e	527	},
f8572d8f	528	{ }
d5a4502e PE	529	};
d5a4502e PE	530
024626e3	531	static __net_init int sysctl_core_net_init(struct net *net)
33eb9cfc	532	{
d5a4502e	533	struct ctl_table *tbl;
024626e3	534
d5a4502e	535	tbl = netns_core_table;
09ad9bc7	536	if (!net_eq(net, &init_net)) {
d5a4502e	537	tbl = kmemdup(tbl, sizeof(netns_core_table), GFP_KERNEL);
024626e3 PE	538	if (tbl == NULL)
	539	goto err_dup;
	540
d5a4502e	541	tbl[0].data = &net->core.sysctl_somaxconn;
464dc801 EB	542
	543	/* Don't export any sysctls to unprivileged users */
	544	if (net->user_ns != &init_user_ns) {
	545	tbl[0].procname = NULL;
	546	}
024626e3 PE	547	}
024626e3 PE	548
ec8f23ce	549	net->core.sysctl_hdr = register_net_sysctl(net, "net/core", tbl);
8efa6e93	550	if (net->core.sysctl_hdr == NULL)
024626e3	551	goto err_reg;
33eb9cfc	552
024626e3 PE	553	return 0;
	554
	555	err_reg:
d5a4502e	556	if (tbl != netns_core_table)
024626e3 PE	557	kfree(tbl);
	558	err_dup:
	559	return -ENOMEM;
	560	}
	561
	562	static __net_exit void sysctl_core_net_exit(struct net *net)
	563	{
	564	struct ctl_table *tbl;
	565
8efa6e93 PE	566	tbl = net->core.sysctl_hdr->ctl_table_arg;
8efa6e93 PE	567	unregister_net_sysctl_table(net->core.sysctl_hdr);
d5a4502e	568	BUG_ON(tbl == netns_core_table);
024626e3 PE	569	kfree(tbl);
	570	}
	571
	572	static __net_initdata struct pernet_operations sysctl_core_ops = {
	573	.init = sysctl_core_net_init,
	574	.exit = sysctl_core_net_exit,
	575	};
	576
	577	static __init int sysctl_core_init(void)
	578	{
43444757	579	register_net_sysctl(&init_net, "net/core", net_core_table);
024626e3	580	return register_pernet_subsys(&sysctl_core_ops);
33eb9cfc PE	581	}
33eb9cfc PE	582
b27aeadb	583	fs_initcall(sysctl_core_init);