[mirror_ubuntu-jammy-kernel.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 <net/ip.h>
#include <net/sock.h>
#include <net/net_ratelimit.h>
#include <net/busy_poll.h>
#include <net/pkt_sched.h>

static int two = 2;
static int three = 3;
static int int_3600 = 3600;
static int min_sndbuf = SOCK_MIN_SNDBUF;
static int min_rcvbuf = SOCK_MIN_RCVBUF;
static int max_skb_frags = MAX_SKB_FRAGS;
static long long_one __maybe_unused = 1;
static long long_max __maybe_unused = LONG_MAX;

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

int sysctl_fb_tunnels_only_for_init_net __read_mostly = 0;
EXPORT_SYMBOL(sysctl_fb_tunnels_only_for_init_net);

/* 0 - Keep current behavior:
 *     IPv4: inherit all current settings from init_net
 *     IPv6: reset all settings to default
 * 1 - Both inherit all current settings from init_net
 * 2 - Both reset all settings to default
 * 3 - Both inherit all settings from current netns
 */
int sysctl_devconf_inherit_init_net __read_mostly;
EXPORT_SYMBOL(sysctl_devconf_inherit_init_net);

#ifdef CONFIG_RPS
static int rps_sock_flow_sysctl(struct ctl_table *table, int write,
				void *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_branch_inc(&rps_needed);
				static_branch_inc(&rfs_needed);
			}
			if (orig_sock_table) {
				static_branch_dec(&rps_needed);
				static_branch_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 *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(buffer, 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';
		memcpy(buffer, kbuf, len);
		*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 *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 *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 *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 *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 *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(current_cred()))) {
			*(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 *buffer, size_t *lenp, loff_t *ppos)
{
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

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

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

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

static struct ctl_table net_core_table[] = {
	{
		.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		= SYSCTL_ONE,
		.extra2		= SYSCTL_ONE,
# else
		.extra1		= SYSCTL_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		= SYSCTL_ZERO,
		.extra2		= &two,
	},
	{
		.procname	= "bpf_jit_kallsyms",
		.data		= &bpf_jit_kallsyms,
		.maxlen		= sizeof(int),
		.mode		= 0600,
		.proc_handler	= proc_dointvec_minmax_bpf_restricted,
		.extra1		= SYSCTL_ZERO,
		.extra2		= SYSCTL_ONE,
	},
# endif
	{
		.procname	= "bpf_jit_limit",
		.data		= &bpf_jit_limit,
		.maxlen		= sizeof(long),
		.mode		= 0600,
		.proc_handler	= proc_dolongvec_minmax_bpf_restricted,
		.extra1		= &long_one,
		.extra2		= &bpf_jit_limit_max,
	},
#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		= SYSCTL_ZERO,
		.extra2		= SYSCTL_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		= SYSCTL_ZERO,
	},
	{
		.procname	= "busy_read",
		.data		= &sysctl_net_busy_read,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ZERO,
	},
#endif
#ifdef CONFIG_NET_SCHED
	{
		.procname	= "default_qdisc",
		.mode		= 0644,
		.maxlen		= IFNAMSIZ,
		.proc_handler	= set_default_qdisc
	},
#endif
	{
		.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		= SYSCTL_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		= SYSCTL_ZERO,
	},
	{
		.procname	= "fb_tunnels_only_for_init_net",
		.data		= &sysctl_fb_tunnels_only_for_init_net,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ZERO,
		.extra2		= &two,
	},
	{
		.procname	= "devconf_inherit_init_net",
		.data		= &sysctl_devconf_inherit_init_net,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ZERO,
		.extra2		= &three,
	},
	{
		.procname	= "high_order_alloc_disable",
		.data		= &net_high_order_alloc_disable_key.key,
		.maxlen         = sizeof(net_high_order_alloc_disable_key),
		.mode		= 0644,
		.proc_handler	= proc_do_static_key,
	},
	{
		.procname	= "gro_normal_batch",
		.data		= &gro_normal_batch,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ONE,
	},
	{
		.procname	= "netdev_unregister_timeout_secs",
		.data		= &netdev_unregister_timeout_secs,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= proc_dointvec_minmax,
		.extra1		= SYSCTL_ONE,
		.extra2		= &int_3600,
	},
	{ }
};

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

static int __init fb_tunnels_only_for_init_net_sysctl_setup(char *str)
{
	/* fallback tunnels for initns only */
	if (!strncmp(str, "initns", 6))
		sysctl_fb_tunnels_only_for_init_net = 1;
	/* no fallback tunnels anywhere */
	else if (!strncmp(str, "none", 4))
		sysctl_fb_tunnels_only_for_init_net = 2;

	return 1;
}
__setup("fb_tunnels=", fb_tunnels_only_for_init_net_sysctl_setup);

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>
3fff4c42	18
63d819ca	19	#include <net/ip.h>
20380731	20	#include <net/sock.h>
c5c177b4	21	#include <net/net_ratelimit.h>
076bb0c8	22	#include <net/busy_poll.h>
6da7c8fc	23	#include <net/pkt_sched.h>
1da177e4	24
316cdaa1	25	static int two = 2;
9efd6a3c	26	static int three = 3;
5aa3afe1	27	static int int_3600 = 3600;
b1cb59cf AK	28	static int min_sndbuf = SOCK_MIN_SNDBUF;
b1cb59cf AK	29	static int min_rcvbuf = SOCK_MIN_RCVBUF;
5f74f82e	30	static int max_skb_frags = MAX_SKB_FRAGS;
fdadd049 DB	31	static long long_one __maybe_unused = 1;
fdadd049 DB	32	static long long_max __maybe_unused = LONG_MAX;
cdda8891	33
ba7a46f1 JP	34	static int net_msg_warn; /* Unused, but still a sysctl */
ba7a46f1 JP	35
79134e6c ED	36	int sysctl_fb_tunnels_only_for_init_net __read_mostly = 0;
	37	EXPORT_SYMBOL(sysctl_fb_tunnels_only_for_init_net);
	38
856c395c CW	39	/* 0 - Keep current behavior:
	40	* IPv4: inherit all current settings from init_net
	41	* IPv6: reset all settings to default
	42	* 1 - Both inherit all current settings from init_net
	43	* 2 - Both reset all settings to default
9efd6a3c	44	* 3 - Both inherit all settings from current netns
856c395c CW	45	*/
	46	int sysctl_devconf_inherit_init_net __read_mostly;
	47	EXPORT_SYMBOL(sysctl_devconf_inherit_init_net);
	48
fec5e652	49	#ifdef CONFIG_RPS
fe2c6338	50	static int rps_sock_flow_sysctl(struct ctl_table *table, int write,
32927393	51	void buffer, size_t lenp, loff_t *ppos)
fec5e652 TH	52	{
	53	unsigned int orig_size, size;
	54	int ret, i;
fe2c6338	55	struct ctl_table tmp = {
fec5e652 TH	56	.data = &size,
	57	.maxlen = sizeof(size),
	58	.mode = table->mode
	59	};
	60	struct rps_sock_flow_table orig_sock_table, sock_table;
	61	static DEFINE_MUTEX(sock_flow_mutex);
	62
	63	mutex_lock(&sock_flow_mutex);
	64
6e3f7faf ED	65	orig_sock_table = rcu_dereference_protected(rps_sock_flow_table,
6e3f7faf ED	66	lockdep_is_held(&sock_flow_mutex));
fec5e652 TH	67	size = orig_size = orig_sock_table ? orig_sock_table->mask + 1 : 0;
	68
	69	ret = proc_dointvec(&tmp, write, buffer, lenp, ppos);
	70
	71	if (write) {
	72	if (size) {
93c1af6c	73	if (size > 1<<29) {
fec5e652 TH	74	/* Enforce limit to prevent overflow */
	75	mutex_unlock(&sock_flow_mutex);
	76	return -EINVAL;
	77	}
	78	size = roundup_pow_of_two(size);
	79	if (size != orig_size) {
	80	sock_table =
	81	vmalloc(RPS_SOCK_FLOW_TABLE_SIZE(size));
	82	if (!sock_table) {
	83	mutex_unlock(&sock_flow_mutex);
	84	return -ENOMEM;
	85	}
567e4b79	86	rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1;
fec5e652 TH	87	sock_table->mask = size - 1;
	88	} else
	89	sock_table = orig_sock_table;
	90
	91	for (i = 0; i < size; i++)
	92	sock_table->ents[i] = RPS_NO_CPU;
	93	} else
	94	sock_table = NULL;
	95
	96	if (sock_table != orig_sock_table) {
	97	rcu_assign_pointer(rps_sock_flow_table, sock_table);
13bfff25	98	if (sock_table) {
dc05360f ED	99	static_branch_inc(&rps_needed);
dc05360f ED	100	static_branch_inc(&rfs_needed);
13bfff25	101	}
adc9300e	102	if (orig_sock_table) {
dc05360f ED	103	static_branch_dec(&rps_needed);
dc05360f ED	104	static_branch_dec(&rfs_needed);
adc9300e ED	105	synchronize_rcu();
	106	vfree(orig_sock_table);
	107	}
fec5e652 TH	108	}
	109	}
	110
	111	mutex_unlock(&sock_flow_mutex);
	112
	113	return ret;
	114	}
	115	#endif /* CONFIG_RPS */
	116
99bbc707 WB	117	#ifdef CONFIG_NET_FLOW_LIMIT
	118	static DEFINE_MUTEX(flow_limit_update_mutex);
	119
fe2c6338	120	static int flow_limit_cpu_sysctl(struct ctl_table *table, int write,
32927393	121	void buffer, size_t lenp, loff_t *ppos)
99bbc707 WB	122	{
	123	struct sd_flow_limit *cur;
	124	struct softnet_data *sd;
	125	cpumask_var_t mask;
	126	int i, len, ret = 0;
	127
	128	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
	129	return -ENOMEM;
	130
	131	if (write) {
56965ac7	132	ret = cpumask_parse(buffer, mask);
99bbc707 WB	133	if (ret)
	134	goto done;
	135
	136	mutex_lock(&flow_limit_update_mutex);
	137	len = sizeof(*cur) + netdev_flow_limit_table_len;
	138	for_each_possible_cpu(i) {
	139	sd = &per_cpu(softnet_data, i);
	140	cur = rcu_dereference_protected(sd->flow_limit,
	141	lockdep_is_held(&flow_limit_update_mutex));
	142	if (cur && !cpumask_test_cpu(i, mask)) {
	143	RCU_INIT_POINTER(sd->flow_limit, NULL);
	144	synchronize_rcu();
	145	kfree(cur);
	146	} else if (!cur && cpumask_test_cpu(i, mask)) {
5b59d467 ED	147	cur = kzalloc_node(len, GFP_KERNEL,
5b59d467 ED	148	cpu_to_node(i));
99bbc707 WB	149	if (!cur) {
	150	/* not unwinding previous changes */
	151	ret = -ENOMEM;
	152	goto write_unlock;
	153	}
	154	cur->num_buckets = netdev_flow_limit_table_len;
	155	rcu_assign_pointer(sd->flow_limit, cur);
	156	}
	157	}
	158	write_unlock:
	159	mutex_unlock(&flow_limit_update_mutex);
	160	} else {
5f121b9a WB	161	char kbuf[128];
5f121b9a WB	162
99bbc707 WB	163	if (ppos \|\| !lenp) {
	164	*lenp = 0;
	165	goto done;
	166	}
	167
	168	cpumask_clear(mask);
	169	rcu_read_lock();
	170	for_each_possible_cpu(i) {
	171	sd = &per_cpu(softnet_data, i);
	172	if (rcu_dereference(sd->flow_limit))
	173	cpumask_set_cpu(i, mask);
	174	}
	175	rcu_read_unlock();
	176
5f121b9a	177	len = min(sizeof(kbuf) - 1, *lenp);
f0906827	178	len = scnprintf(kbuf, len, "%*pb", cpumask_pr_args(mask));
5f121b9a WB	179	if (!len) {
	180	*lenp = 0;
	181	goto done;
	182	}
	183	if (len < *lenp)
	184	kbuf[len++] = '\n';
32927393	185	memcpy(buffer, kbuf, len);
5f121b9a WB	186	*lenp = len;
5f121b9a WB	187	*ppos += len;
99bbc707 WB	188	}
	189
	190	done:
	191	free_cpumask_var(mask);
	192	return ret;
	193	}
	194
fe2c6338	195	static int flow_limit_table_len_sysctl(struct ctl_table *table, int write,
32927393	196	void buffer, size_t lenp, loff_t *ppos)
99bbc707 WB	197	{
	198	unsigned int old, *ptr;
	199	int ret;
	200
	201	mutex_lock(&flow_limit_update_mutex);
	202
	203	ptr = table->data;
	204	old = *ptr;
	205	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	206	if (!ret && write && !is_power_of_2(*ptr)) {
	207	*ptr = old;
	208	ret = -EINVAL;
	209	}
	210
	211	mutex_unlock(&flow_limit_update_mutex);
	212	return ret;
	213	}
	214	#endif /* CONFIG_NET_FLOW_LIMIT */
	215
6da7c8fc	216	#ifdef CONFIG_NET_SCHED
6da7c8fc	217	static int set_default_qdisc(struct ctl_table *table, int write,
32927393	218	void buffer, size_t lenp, loff_t *ppos)
6da7c8fc	219	{
	220	char id[IFNAMSIZ];
	221	struct ctl_table tbl = {
	222	.data = id,
	223	.maxlen = IFNAMSIZ,
	224	};
	225	int ret;
	226
	227	qdisc_get_default(id, IFNAMSIZ);
	228
	229	ret = proc_dostring(&tbl, write, buffer, lenp, ppos);
	230	if (write && ret == 0)
	231	ret = qdisc_set_default(id);
	232	return ret;
	233	}
	234	#endif
	235
3d48b53f	236	static int proc_do_dev_weight(struct ctl_table *table, int write,
32927393	237	void buffer, size_t lenp, loff_t *ppos)
3d48b53f MT	238	{
	239	int ret;
	240
	241	ret = proc_dointvec(table, write, buffer, lenp, ppos);
	242	if (ret != 0)
	243	return ret;
	244
	245	dev_rx_weight = weight_p * dev_weight_rx_bias;
	246	dev_tx_weight = weight_p * dev_weight_tx_bias;
	247
	248	return ret;
	249	}
	250
960fb622	251	static int proc_do_rss_key(struct ctl_table *table, int write,
32927393	252	void buffer, size_t lenp, loff_t *ppos)
960fb622 ED	253	{
	254	struct ctl_table fake_table;
	255	char buf[NETDEV_RSS_KEY_LEN * 3];
	256
	257	snprintf(buf, sizeof(buf), "%*phC", NETDEV_RSS_KEY_LEN, netdev_rss_key);
	258	fake_table.data = buf;
	259	fake_table.maxlen = sizeof(buf);
	260	return proc_dostring(&fake_table, write, buffer, lenp, ppos);
	261	}
	262
2e4a3098 DB	263	#ifdef CONFIG_BPF_JIT
2e4a3098 DB	264	static int proc_dointvec_minmax_bpf_enable(struct ctl_table *table, int write,
32927393	265	void buffer, size_t lenp,
2e4a3098 DB	266	loff_t *ppos)
	267	{
	268	int ret, jit_enable = (int )table->data;
	269	struct ctl_table tmp = *table;
	270
	271	if (write && !capable(CAP_SYS_ADMIN))
	272	return -EPERM;
	273
	274	tmp.data = &jit_enable;
	275	ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
	276	if (write && !ret) {
	277	if (jit_enable < 2 \|\|
63960260	278	(jit_enable == 2 && bpf_dump_raw_ok(current_cred()))) {
2e4a3098 DB	279	(int )table->data = jit_enable;
	280	if (jit_enable == 2)
	281	pr_warn("bpf_jit_enable = 2 was set! NEVER use this in production, only for JIT debugging!\n");
	282	} else {
	283	ret = -EPERM;
	284	}
	285	}
	286	return ret;
	287	}
	288
1148f9ad	289	# ifdef CONFIG_HAVE_EBPF_JIT
2e4a3098 DB	290	static int
2e4a3098 DB	291	proc_dointvec_minmax_bpf_restricted(struct ctl_table *table, int write,
32927393	292	void buffer, size_t lenp, loff_t *ppos)
2e4a3098 DB	293	{
	294	if (!capable(CAP_SYS_ADMIN))
	295	return -EPERM;
	296
	297	return proc_dointvec_minmax(table, write, buffer, lenp, ppos);
	298	}
1148f9ad	299	# endif /* CONFIG_HAVE_EBPF_JIT */
fdadd049 DB	300
	301	static int
	302	proc_dolongvec_minmax_bpf_restricted(struct ctl_table *table, int write,
32927393	303	void buffer, size_t lenp, loff_t *ppos)
fdadd049 DB	304	{
	305	if (!capable(CAP_SYS_ADMIN))
	306	return -EPERM;
	307
	308	return proc_doulongvec_minmax(table, write, buffer, lenp, ppos);
	309	}
2e4a3098 DB	310	#endif
2e4a3098 DB	311
33eb9cfc	312	static struct ctl_table net_core_table[] = {
1da177e4	313	{
1da177e4 LT	314	.procname = "wmem_max",
	315	.data = &sysctl_wmem_max,
	316	.maxlen = sizeof(int),
	317	.mode = 0644,
cdda8891	318	.proc_handler = proc_dointvec_minmax,
b1cb59cf	319	.extra1 = &min_sndbuf,
1da177e4 LT	320	},
1da177e4 LT	321	{
1da177e4 LT	322	.procname = "rmem_max",
	323	.data = &sysctl_rmem_max,
	324	.maxlen = sizeof(int),
	325	.mode = 0644,
cdda8891	326	.proc_handler = proc_dointvec_minmax,
b1cb59cf	327	.extra1 = &min_rcvbuf,
1da177e4 LT	328	},
1da177e4 LT	329	{
1da177e4 LT	330	.procname = "wmem_default",
	331	.data = &sysctl_wmem_default,
	332	.maxlen = sizeof(int),
	333	.mode = 0644,
cdda8891	334	.proc_handler = proc_dointvec_minmax,
b1cb59cf	335	.extra1 = &min_sndbuf,
1da177e4 LT	336	},
1da177e4 LT	337	{
1da177e4 LT	338	.procname = "rmem_default",
	339	.data = &sysctl_rmem_default,
	340	.maxlen = sizeof(int),
	341	.mode = 0644,
cdda8891	342	.proc_handler = proc_dointvec_minmax,
b1cb59cf	343	.extra1 = &min_rcvbuf,
1da177e4 LT	344	},
1da177e4 LT	345	{
1da177e4 LT	346	.procname = "dev_weight",
	347	.data = &weight_p,
	348	.maxlen = sizeof(int),
	349	.mode = 0644,
3d48b53f MT	350	.proc_handler = proc_do_dev_weight,
	351	},
	352	{
	353	.procname = "dev_weight_rx_bias",
	354	.data = &dev_weight_rx_bias,
	355	.maxlen = sizeof(int),
	356	.mode = 0644,
	357	.proc_handler = proc_do_dev_weight,
	358	},
	359	{
	360	.procname = "dev_weight_tx_bias",
	361	.data = &dev_weight_tx_bias,
	362	.maxlen = sizeof(int),
	363	.mode = 0644,
	364	.proc_handler = proc_do_dev_weight,
1da177e4 LT	365	},
1da177e4 LT	366	{
1da177e4 LT	367	.procname = "netdev_max_backlog",
	368	.data = &netdev_max_backlog,
	369	.maxlen = sizeof(int),
	370	.mode = 0644,
6d9f239a	371	.proc_handler = proc_dointvec
1da177e4	372	},
960fb622 ED	373	{
	374	.procname = "netdev_rss_key",
	375	.data = &netdev_rss_key,
	376	.maxlen = sizeof(int),
	377	.mode = 0444,
	378	.proc_handler = proc_do_rss_key,
	379	},
0a14842f ED	380	#ifdef CONFIG_BPF_JIT
	381	{
	382	.procname = "bpf_jit_enable",
	383	.data = &bpf_jit_enable,
	384	.maxlen = sizeof(int),
	385	.mode = 0644,
2e4a3098	386	.proc_handler = proc_dointvec_minmax_bpf_enable,
fa9dd599	387	# ifdef CONFIG_BPF_JIT_ALWAYS_ON
eec4844f MC	388	.extra1 = SYSCTL_ONE,
eec4844f MC	389	.extra2 = SYSCTL_ONE,
fa9dd599	390	# else
eec4844f	391	.extra1 = SYSCTL_ZERO,
fa9dd599 DB	392	.extra2 = &two,
fa9dd599 DB	393	# endif
0a14842f	394	},
4f3446bb DB	395	# ifdef CONFIG_HAVE_EBPF_JIT
	396	{
	397	.procname = "bpf_jit_harden",
	398	.data = &bpf_jit_harden,
	399	.maxlen = sizeof(int),
	400	.mode = 0600,
2e4a3098	401	.proc_handler = proc_dointvec_minmax_bpf_restricted,
eec4844f	402	.extra1 = SYSCTL_ZERO,
fa9dd599	403	.extra2 = &two,
4f3446bb	404	},
74451e66 DB	405	{
	406	.procname = "bpf_jit_kallsyms",
	407	.data = &bpf_jit_kallsyms,
	408	.maxlen = sizeof(int),
	409	.mode = 0600,
2e4a3098	410	.proc_handler = proc_dointvec_minmax_bpf_restricted,
eec4844f MC	411	.extra1 = SYSCTL_ZERO,
eec4844f MC	412	.extra2 = SYSCTL_ONE,
74451e66	413	},
4f3446bb	414	# endif
ede95a63 DB	415	{
	416	.procname = "bpf_jit_limit",
	417	.data = &bpf_jit_limit,
fdadd049	418	.maxlen = sizeof(long),
ede95a63	419	.mode = 0600,
fdadd049 DB	420	.proc_handler = proc_dolongvec_minmax_bpf_restricted,
fdadd049 DB	421	.extra1 = &long_one,
fadb7ff1	422	.extra2 = &bpf_jit_limit_max,
ede95a63	423	},
0a14842f	424	#endif
3b098e2d ED	425	{
	426	.procname = "netdev_tstamp_prequeue",
	427	.data = &netdev_tstamp_prequeue,
	428	.maxlen = sizeof(int),
	429	.mode = 0644,
	430	.proc_handler = proc_dointvec
	431	},
1da177e4	432	{
1da177e4	433	.procname = "message_cost",
717115e1	434	.data = &net_ratelimit_state.interval,
1da177e4 LT	435	.maxlen = sizeof(int),
1da177e4 LT	436	.mode = 0644,
6d9f239a	437	.proc_handler = proc_dointvec_jiffies,
1da177e4 LT	438	},
1da177e4 LT	439	{
1da177e4	440	.procname = "message_burst",
717115e1	441	.data = &net_ratelimit_state.burst,
1da177e4 LT	442	.maxlen = sizeof(int),
1da177e4 LT	443	.mode = 0644,
6d9f239a	444	.proc_handler = proc_dointvec,
1da177e4 LT	445	},
1da177e4 LT	446	{
1da177e4 LT	447	.procname = "optmem_max",
	448	.data = &sysctl_optmem_max,
	449	.maxlen = sizeof(int),
	450	.mode = 0644,
6d9f239a	451	.proc_handler = proc_dointvec
1da177e4	452	},
b245be1f WB	453	{
	454	.procname = "tstamp_allow_data",
	455	.data = &sysctl_tstamp_allow_data,
	456	.maxlen = sizeof(int),
	457	.mode = 0644,
	458	.proc_handler = proc_dointvec_minmax,
eec4844f MC	459	.extra1 = SYSCTL_ZERO,
eec4844f MC	460	.extra2 = SYSCTL_ONE
b245be1f	461	},
fec5e652 TH	462	#ifdef CONFIG_RPS
	463	{
	464	.procname = "rps_sock_flow_entries",
	465	.maxlen = sizeof(int),
	466	.mode = 0644,
	467	.proc_handler = rps_sock_flow_sysctl
	468	},
	469	#endif
99bbc707 WB	470	#ifdef CONFIG_NET_FLOW_LIMIT
	471	{
	472	.procname = "flow_limit_cpu_bitmap",
	473	.mode = 0644,
	474	.proc_handler = flow_limit_cpu_sysctl
	475	},
	476	{
	477	.procname = "flow_limit_table_len",
	478	.data = &netdev_flow_limit_table_len,
	479	.maxlen = sizeof(int),
	480	.mode = 0644,
	481	.proc_handler = flow_limit_table_len_sysctl
	482	},
	483	#endif /* CONFIG_NET_FLOW_LIMIT */
e0d1095a	484	#ifdef CONFIG_NET_RX_BUSY_POLL
06021292	485	{
64b0dc51 ET	486	.procname = "busy_poll",
64b0dc51 ET	487	.data = &sysctl_net_busy_poll,
eb6db622	488	.maxlen = sizeof(unsigned int),
06021292	489	.mode = 0644,
95f25521	490	.proc_handler = proc_dointvec_minmax,
eec4844f	491	.extra1 = SYSCTL_ZERO,
06021292	492	},
2d48d67f	493	{
64b0dc51 ET	494	.procname = "busy_read",
64b0dc51 ET	495	.data = &sysctl_net_busy_read,
2d48d67f ET	496	.maxlen = sizeof(unsigned int),
2d48d67f ET	497	.mode = 0644,
95f25521	498	.proc_handler = proc_dointvec_minmax,
eec4844f	499	.extra1 = SYSCTL_ZERO,
2d48d67f	500	},
6da7c8fc	501	#endif
	502	#ifdef CONFIG_NET_SCHED
	503	{
	504	.procname = "default_qdisc",
	505	.mode = 0644,
	506	.maxlen = IFNAMSIZ,
	507	.proc_handler = set_default_qdisc
	508	},
06021292	509	#endif
51b0bded	510	{
51b0bded SH	511	.procname = "netdev_budget",
	512	.data = &netdev_budget,
	513	.maxlen = sizeof(int),
	514	.mode = 0644,
6d9f239a	515	.proc_handler = proc_dointvec
51b0bded	516	},
a2a316fd	517	{
a2a316fd SH	518	.procname = "warnings",
	519	.data = &net_msg_warn,
	520	.maxlen = sizeof(int),
	521	.mode = 0644,
6d9f239a	522	.proc_handler = proc_dointvec
a2a316fd	523	},
5f74f82e HWR	524	{
	525	.procname = "max_skb_frags",
	526	.data = &sysctl_max_skb_frags,
	527	.maxlen = sizeof(int),
	528	.mode = 0644,
	529	.proc_handler = proc_dointvec_minmax,
eec4844f	530	.extra1 = SYSCTL_ONE,
5f74f82e HWR	531	.extra2 = &max_skb_frags,
5f74f82e HWR	532	},
7acf8a1e MW	533	{
	534	.procname = "netdev_budget_usecs",
	535	.data = &netdev_budget_usecs,
	536	.maxlen = sizeof(unsigned int),
	537	.mode = 0644,
	538	.proc_handler = proc_dointvec_minmax,
eec4844f	539	.extra1 = SYSCTL_ZERO,
7acf8a1e	540	},
79134e6c ED	541	{
	542	.procname = "fb_tunnels_only_for_init_net",
	543	.data = &sysctl_fb_tunnels_only_for_init_net,
	544	.maxlen = sizeof(int),
	545	.mode = 0644,
	546	.proc_handler = proc_dointvec_minmax,
eec4844f	547	.extra1 = SYSCTL_ZERO,
316cdaa1	548	.extra2 = &two,
79134e6c	549	},
856c395c CW	550	{
	551	.procname = "devconf_inherit_init_net",
	552	.data = &sysctl_devconf_inherit_init_net,
	553	.maxlen = sizeof(int),
	554	.mode = 0644,
	555	.proc_handler = proc_dointvec_minmax,
eec4844f	556	.extra1 = SYSCTL_ZERO,
9efd6a3c	557	.extra2 = &three,
856c395c	558	},
ce27ec60 ED	559	{
	560	.procname = "high_order_alloc_disable",
	561	.data = &net_high_order_alloc_disable_key.key,
	562	.maxlen = sizeof(net_high_order_alloc_disable_key),
	563	.mode = 0644,
	564	.proc_handler = proc_do_static_key,
	565	},
323ebb61 EC	566	{
	567	.procname = "gro_normal_batch",
	568	.data = &gro_normal_batch,
	569	.maxlen = sizeof(unsigned int),
	570	.mode = 0644,
	571	.proc_handler = proc_dointvec_minmax,
	572	.extra1 = SYSCTL_ONE,
	573	},
5aa3afe1 DV	574	{
	575	.procname = "netdev_unregister_timeout_secs",
	576	.data = &netdev_unregister_timeout_secs,
	577	.maxlen = sizeof(unsigned int),
	578	.mode = 0644,
	579	.proc_handler = proc_dointvec_minmax,
6c996e19	580	.extra1 = SYSCTL_ONE,
5aa3afe1 DV	581	.extra2 = &int_3600,
5aa3afe1 DV	582	},
f8572d8f	583	{ }
1da177e4	584	};
33eb9cfc	585
d5a4502e PE	586	static struct ctl_table netns_core_table[] = {
d5a4502e PE	587	{
d5a4502e PE	588	.procname = "somaxconn",
	589	.data = &init_net.core.sysctl_somaxconn,
	590	.maxlen = sizeof(int),
	591	.mode = 0644,
eec4844f	592	.extra1 = SYSCTL_ZERO,
5f671d6b	593	.proc_handler = proc_dointvec_minmax
d5a4502e	594	},
f8572d8f	595	{ }
d5a4502e PE	596	};
d5a4502e PE	597
316cdaa1 MB	598	static int __init fb_tunnels_only_for_init_net_sysctl_setup(char *str)
	599	{
	600	/* fallback tunnels for initns only */
	601	if (!strncmp(str, "initns", 6))
	602	sysctl_fb_tunnels_only_for_init_net = 1;
	603	/* no fallback tunnels anywhere */
	604	else if (!strncmp(str, "none", 4))
	605	sysctl_fb_tunnels_only_for_init_net = 2;
	606
	607	return 1;
	608	}
	609	__setup("fb_tunnels=", fb_tunnels_only_for_init_net_sysctl_setup);
	610
024626e3	611	static __net_init int sysctl_core_net_init(struct net *net)
33eb9cfc	612	{
d5a4502e	613	struct ctl_table *tbl;
024626e3	614
d5a4502e	615	tbl = netns_core_table;
09ad9bc7	616	if (!net_eq(net, &init_net)) {
d5a4502e	617	tbl = kmemdup(tbl, sizeof(netns_core_table), GFP_KERNEL);
024626e3 PE	618	if (tbl == NULL)
	619	goto err_dup;
	620
d5a4502e	621	tbl[0].data = &net->core.sysctl_somaxconn;
464dc801 EB	622
	623	/* Don't export any sysctls to unprivileged users */
	624	if (net->user_ns != &init_user_ns) {
	625	tbl[0].procname = NULL;
	626	}
024626e3 PE	627	}
024626e3 PE	628
ec8f23ce	629	net->core.sysctl_hdr = register_net_sysctl(net, "net/core", tbl);
8efa6e93	630	if (net->core.sysctl_hdr == NULL)
024626e3	631	goto err_reg;
33eb9cfc	632
024626e3 PE	633	return 0;
	634
	635	err_reg:
d5a4502e	636	if (tbl != netns_core_table)
024626e3 PE	637	kfree(tbl);
	638	err_dup:
	639	return -ENOMEM;
	640	}
	641
	642	static __net_exit void sysctl_core_net_exit(struct net *net)
	643	{
	644	struct ctl_table *tbl;
	645
8efa6e93 PE	646	tbl = net->core.sysctl_hdr->ctl_table_arg;
8efa6e93 PE	647	unregister_net_sysctl_table(net->core.sysctl_hdr);
d5a4502e	648	BUG_ON(tbl == netns_core_table);
024626e3 PE	649	kfree(tbl);
	650	}
	651
	652	static __net_initdata struct pernet_operations sysctl_core_ops = {
	653	.init = sysctl_core_net_init,
	654	.exit = sysctl_core_net_exit,
	655	};
	656
	657	static __init int sysctl_core_init(void)
	658	{
43444757	659	register_net_sysctl(&init_net, "net/core", net_core_table);
024626e3	660	return register_pernet_subsys(&sysctl_core_ops);
33eb9cfc PE	661	}
33eb9cfc PE	662
b27aeadb	663	fs_initcall(sysctl_core_init);