[linux-2.6-block.git] / net / ipv4 / tcp_memcontrol.c

#include <net/tcp.h>
#include <net/tcp_memcontrol.h>
#include <net/sock.h>
#include <net/ip.h>
#include <linux/nsproxy.h>
#include <linux/memcontrol.h>
#include <linux/module.h>

static u64 tcp_cgroup_read(struct cgroup *cont, struct cftype *cft);
static int tcp_cgroup_write(struct cgroup *cont, struct cftype *cft,
			    const char *buffer);
static int tcp_cgroup_reset(struct cgroup *cont, unsigned int event);

static struct cftype tcp_files[] = {
	{
		.name = "kmem.tcp.limit_in_bytes",
		.write_string = tcp_cgroup_write,
		.read_u64 = tcp_cgroup_read,
		.private = RES_LIMIT,
	},
	{
		.name = "kmem.tcp.usage_in_bytes",
		.read_u64 = tcp_cgroup_read,
		.private = RES_USAGE,
	},
	{
		.name = "kmem.tcp.failcnt",
		.private = RES_FAILCNT,
		.trigger = tcp_cgroup_reset,
		.read_u64 = tcp_cgroup_read,
	},
	{
		.name = "kmem.tcp.max_usage_in_bytes",
		.private = RES_MAX_USAGE,
		.trigger = tcp_cgroup_reset,
		.read_u64 = tcp_cgroup_read,
	},
};

static inline struct tcp_memcontrol *tcp_from_cgproto(struct cg_proto *cg_proto)
{
	return container_of(cg_proto, struct tcp_memcontrol, cg_proto);
}

static void memcg_tcp_enter_memory_pressure(struct sock *sk)
{
	if (sk->sk_cgrp->memory_pressure)
		*sk->sk_cgrp->memory_pressure = 1;
}
EXPORT_SYMBOL(memcg_tcp_enter_memory_pressure);

int tcp_init_cgroup(struct cgroup *cgrp, struct cgroup_subsys *ss)
{
	/*
	 * The root cgroup does not use res_counters, but rather,
	 * rely on the data already collected by the network
	 * subsystem
	 */
	struct res_counter *res_parent = NULL;
	struct cg_proto *cg_proto, *parent_cg;
	struct tcp_memcontrol *tcp;
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
	struct mem_cgroup *parent = parent_mem_cgroup(memcg);
	struct net *net = current->nsproxy->net_ns;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		goto create_files;

	tcp = tcp_from_cgproto(cg_proto);

	tcp->tcp_prot_mem[0] = net->ipv4.sysctl_tcp_mem[0];
	tcp->tcp_prot_mem[1] = net->ipv4.sysctl_tcp_mem[1];
	tcp->tcp_prot_mem[2] = net->ipv4.sysctl_tcp_mem[2];
	tcp->tcp_memory_pressure = 0;

	parent_cg = tcp_prot.proto_cgroup(parent);
	if (parent_cg)
		res_parent = parent_cg->memory_allocated;

	res_counter_init(&tcp->tcp_memory_allocated, res_parent);
	percpu_counter_init(&tcp->tcp_sockets_allocated, 0);

	cg_proto->enter_memory_pressure = memcg_tcp_enter_memory_pressure;
	cg_proto->memory_pressure = &tcp->tcp_memory_pressure;
	cg_proto->sysctl_mem = tcp->tcp_prot_mem;
	cg_proto->memory_allocated = &tcp->tcp_memory_allocated;
	cg_proto->sockets_allocated = &tcp->tcp_sockets_allocated;
	cg_proto->memcg = memcg;

create_files:
	return cgroup_add_files(cgrp, ss, tcp_files,
				ARRAY_SIZE(tcp_files));
}
EXPORT_SYMBOL(tcp_init_cgroup);

void tcp_destroy_cgroup(struct cgroup *cgrp)
{
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
	struct cg_proto *cg_proto;
	struct tcp_memcontrol *tcp;
	u64 val;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return;

	tcp = tcp_from_cgproto(cg_proto);
	percpu_counter_destroy(&tcp->tcp_sockets_allocated);

	val = res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);

	if (val != RESOURCE_MAX)
		static_key_slow_dec(&memcg_socket_limit_enabled);
}
EXPORT_SYMBOL(tcp_destroy_cgroup);

static int tcp_update_limit(struct mem_cgroup *memcg, u64 val)
{
	struct net *net = current->nsproxy->net_ns;
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;
	u64 old_lim;
	int i;
	int ret;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return -EINVAL;

	if (val > RESOURCE_MAX)
		val = RESOURCE_MAX;

	tcp = tcp_from_cgproto(cg_proto);

	old_lim = res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);
	ret = res_counter_set_limit(&tcp->tcp_memory_allocated, val);
	if (ret)
		return ret;

	for (i = 0; i < 3; i++)
		tcp->tcp_prot_mem[i] = min_t(long, val >> PAGE_SHIFT,
					     net->ipv4.sysctl_tcp_mem[i]);

	if (val == RESOURCE_MAX && old_lim != RESOURCE_MAX)
		static_key_slow_dec(&memcg_socket_limit_enabled);
	else if (old_lim == RESOURCE_MAX && val != RESOURCE_MAX)
		static_key_slow_inc(&memcg_socket_limit_enabled);

	return 0;
}

static int tcp_cgroup_write(struct cgroup *cont, struct cftype *cft,
			    const char *buffer)
{
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
	unsigned long long val;
	int ret = 0;

	switch (cft->private) {
	case RES_LIMIT:
		/* see memcontrol.c */
		ret = res_counter_memparse_write_strategy(buffer, &val);
		if (ret)
			break;
		ret = tcp_update_limit(memcg, val);
		break;
	default:
		ret = -EINVAL;
		break;
	}
	return ret;
}

static u64 tcp_read_stat(struct mem_cgroup *memcg, int type, u64 default_val)
{
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return default_val;

	tcp = tcp_from_cgproto(cg_proto);
	return res_counter_read_u64(&tcp->tcp_memory_allocated, type);
}

static u64 tcp_read_usage(struct mem_cgroup *memcg)
{
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return atomic_long_read(&tcp_memory_allocated) << PAGE_SHIFT;

	tcp = tcp_from_cgproto(cg_proto);
	return res_counter_read_u64(&tcp->tcp_memory_allocated, RES_USAGE);
}

static u64 tcp_cgroup_read(struct cgroup *cont, struct cftype *cft)
{
	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
	u64 val;

	switch (cft->private) {
	case RES_LIMIT:
		val = tcp_read_stat(memcg, RES_LIMIT, RESOURCE_MAX);
		break;
	case RES_USAGE:
		val = tcp_read_usage(memcg);
		break;
	case RES_FAILCNT:
	case RES_MAX_USAGE:
		val = tcp_read_stat(memcg, cft->private, 0);
		break;
	default:
		BUG();
	}
	return val;
}

static int tcp_cgroup_reset(struct cgroup *cont, unsigned int event)
{
	struct mem_cgroup *memcg;
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;

	memcg = mem_cgroup_from_cont(cont);
	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return 0;
	tcp = tcp_from_cgproto(cg_proto);

	switch (event) {
	case RES_MAX_USAGE:
		res_counter_reset_max(&tcp->tcp_memory_allocated);
		break;
	case RES_FAILCNT:
		res_counter_reset_failcnt(&tcp->tcp_memory_allocated);
		break;
	}

	return 0;
}

unsigned long long tcp_max_memory(const struct mem_cgroup *memcg)
{
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;

	cg_proto = tcp_prot.proto_cgroup((struct mem_cgroup *)memcg);
	if (!cg_proto)
		return 0;

	tcp = tcp_from_cgproto(cg_proto);
	return res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);
}

void tcp_prot_mem(struct mem_cgroup *memcg, long val, int idx)
{
	struct tcp_memcontrol *tcp;
	struct cg_proto *cg_proto;

	cg_proto = tcp_prot.proto_cgroup(memcg);
	if (!cg_proto)
		return;

	tcp = tcp_from_cgproto(cg_proto);

	tcp->tcp_prot_mem[idx] = val;
}
Commit	Line	Data
d1a4c0b3 GC	1	#include <net/tcp.h>
	2	#include <net/tcp_memcontrol.h>
	3	#include <net/sock.h>
3dc43e3e GC	4	#include <net/ip.h>
3dc43e3e GC	5	#include <linux/nsproxy.h>
d1a4c0b3 GC	6	#include <linux/memcontrol.h>
	7	#include <linux/module.h>
	8
3aaabe23 GC	9	static u64 tcp_cgroup_read(struct cgroup cont, struct cftype cft);
	10	static int tcp_cgroup_write(struct cgroup cont, struct cftype cft,
	11	const char *buffer);
ffea59e5	12	static int tcp_cgroup_reset(struct cgroup *cont, unsigned int event);
3aaabe23 GC	13
	14	static struct cftype tcp_files[] = {
	15	{
	16	.name = "kmem.tcp.limit_in_bytes",
	17	.write_string = tcp_cgroup_write,
	18	.read_u64 = tcp_cgroup_read,
	19	.private = RES_LIMIT,
	20	},
5a6dd343 GC	21	{
	22	.name = "kmem.tcp.usage_in_bytes",
	23	.read_u64 = tcp_cgroup_read,
	24	.private = RES_USAGE,
	25	},
ffea59e5 GC	26	{
	27	.name = "kmem.tcp.failcnt",
	28	.private = RES_FAILCNT,
	29	.trigger = tcp_cgroup_reset,
	30	.read_u64 = tcp_cgroup_read,
	31	},
0850f0f5 GC	32	{
	33	.name = "kmem.tcp.max_usage_in_bytes",
	34	.private = RES_MAX_USAGE,
	35	.trigger = tcp_cgroup_reset,
	36	.read_u64 = tcp_cgroup_read,
	37	},
3aaabe23 GC	38	};
3aaabe23 GC	39
d1a4c0b3 GC	40	static inline struct tcp_memcontrol tcp_from_cgproto(struct cg_proto cg_proto)
	41	{
	42	return container_of(cg_proto, struct tcp_memcontrol, cg_proto);
	43	}
	44
	45	static void memcg_tcp_enter_memory_pressure(struct sock *sk)
	46	{
c48e074c	47	if (sk->sk_cgrp->memory_pressure)
d1a4c0b3 GC	48	*sk->sk_cgrp->memory_pressure = 1;
	49	}
	50	EXPORT_SYMBOL(memcg_tcp_enter_memory_pressure);
	51
	52	int tcp_init_cgroup(struct cgroup cgrp, struct cgroup_subsys ss)
	53	{
	54	/*
	55	* The root cgroup does not use res_counters, but rather,
	56	* rely on the data already collected by the network
	57	* subsystem
	58	*/
	59	struct res_counter *res_parent = NULL;
	60	struct cg_proto cg_proto, parent_cg;
	61	struct tcp_memcontrol *tcp;
	62	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
	63	struct mem_cgroup *parent = parent_mem_cgroup(memcg);
3dc43e3e	64	struct net *net = current->nsproxy->net_ns;
d1a4c0b3 GC	65
	66	cg_proto = tcp_prot.proto_cgroup(memcg);
	67	if (!cg_proto)
3aaabe23	68	goto create_files;
d1a4c0b3 GC	69
	70	tcp = tcp_from_cgproto(cg_proto);
	71
3dc43e3e GC	72	tcp->tcp_prot_mem[0] = net->ipv4.sysctl_tcp_mem[0];
	73	tcp->tcp_prot_mem[1] = net->ipv4.sysctl_tcp_mem[1];
	74	tcp->tcp_prot_mem[2] = net->ipv4.sysctl_tcp_mem[2];
d1a4c0b3 GC	75	tcp->tcp_memory_pressure = 0;
	76
	77	parent_cg = tcp_prot.proto_cgroup(parent);
	78	if (parent_cg)
	79	res_parent = parent_cg->memory_allocated;
	80
	81	res_counter_init(&tcp->tcp_memory_allocated, res_parent);
	82	percpu_counter_init(&tcp->tcp_sockets_allocated, 0);
	83
	84	cg_proto->enter_memory_pressure = memcg_tcp_enter_memory_pressure;
	85	cg_proto->memory_pressure = &tcp->tcp_memory_pressure;
	86	cg_proto->sysctl_mem = tcp->tcp_prot_mem;
	87	cg_proto->memory_allocated = &tcp->tcp_memory_allocated;
	88	cg_proto->sockets_allocated = &tcp->tcp_sockets_allocated;
	89	cg_proto->memcg = memcg;
	90
3aaabe23 GC	91	create_files:
	92	return cgroup_add_files(cgrp, ss, tcp_files,
	93	ARRAY_SIZE(tcp_files));
d1a4c0b3 GC	94	}
	95	EXPORT_SYMBOL(tcp_init_cgroup);
	96
761b3ef5	97	void tcp_destroy_cgroup(struct cgroup *cgrp)
d1a4c0b3 GC	98	{
	99	struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp);
	100	struct cg_proto *cg_proto;
	101	struct tcp_memcontrol *tcp;
3aaabe23	102	u64 val;
d1a4c0b3 GC	103
	104	cg_proto = tcp_prot.proto_cgroup(memcg);
	105	if (!cg_proto)
	106	return;
	107
	108	tcp = tcp_from_cgproto(cg_proto);
	109	percpu_counter_destroy(&tcp->tcp_sockets_allocated);
3aaabe23	110
1398eee0	111	val = res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);
3aaabe23 GC	112
3aaabe23 GC	113	if (val != RESOURCE_MAX)
c5905afb	114	static_key_slow_dec(&memcg_socket_limit_enabled);
d1a4c0b3 GC	115	}
d1a4c0b3 GC	116	EXPORT_SYMBOL(tcp_destroy_cgroup);
3aaabe23 GC	117
	118	static int tcp_update_limit(struct mem_cgroup *memcg, u64 val)
	119	{
	120	struct net *net = current->nsproxy->net_ns;
	121	struct tcp_memcontrol *tcp;
	122	struct cg_proto *cg_proto;
	123	u64 old_lim;
	124	int i;
	125	int ret;
	126
	127	cg_proto = tcp_prot.proto_cgroup(memcg);
	128	if (!cg_proto)
	129	return -EINVAL;
	130
	131	if (val > RESOURCE_MAX)
	132	val = RESOURCE_MAX;
	133
	134	tcp = tcp_from_cgproto(cg_proto);
	135
	136	old_lim = res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);
	137	ret = res_counter_set_limit(&tcp->tcp_memory_allocated, val);
	138	if (ret)
	139	return ret;
	140
	141	for (i = 0; i < 3; i++)
	142	tcp->tcp_prot_mem[i] = min_t(long, val >> PAGE_SHIFT,
	143	net->ipv4.sysctl_tcp_mem[i]);
	144
	145	if (val == RESOURCE_MAX && old_lim != RESOURCE_MAX)
c5905afb	146	static_key_slow_dec(&memcg_socket_limit_enabled);
3aaabe23	147	else if (old_lim == RESOURCE_MAX && val != RESOURCE_MAX)
c5905afb	148	static_key_slow_inc(&memcg_socket_limit_enabled);
3aaabe23 GC	149
	150	return 0;
	151	}
	152
	153	static int tcp_cgroup_write(struct cgroup cont, struct cftype cft,
	154	const char *buffer)
	155	{
	156	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
	157	unsigned long long val;
	158	int ret = 0;
	159
	160	switch (cft->private) {
	161	case RES_LIMIT:
	162	/* see memcontrol.c */
	163	ret = res_counter_memparse_write_strategy(buffer, &val);
	164	if (ret)
	165	break;
	166	ret = tcp_update_limit(memcg, val);
	167	break;
	168	default:
	169	ret = -EINVAL;
	170	break;
	171	}
	172	return ret;
	173	}
	174
	175	static u64 tcp_read_stat(struct mem_cgroup *memcg, int type, u64 default_val)
	176	{
	177	struct tcp_memcontrol *tcp;
	178	struct cg_proto *cg_proto;
	179
	180	cg_proto = tcp_prot.proto_cgroup(memcg);
	181	if (!cg_proto)
	182	return default_val;
	183
	184	tcp = tcp_from_cgproto(cg_proto);
	185	return res_counter_read_u64(&tcp->tcp_memory_allocated, type);
	186	}
	187
5a6dd343 GC	188	static u64 tcp_read_usage(struct mem_cgroup *memcg)
	189	{
	190	struct tcp_memcontrol *tcp;
	191	struct cg_proto *cg_proto;
	192
	193	cg_proto = tcp_prot.proto_cgroup(memcg);
	194	if (!cg_proto)
	195	return atomic_long_read(&tcp_memory_allocated) << PAGE_SHIFT;
	196
	197	tcp = tcp_from_cgproto(cg_proto);
	198	return res_counter_read_u64(&tcp->tcp_memory_allocated, RES_USAGE);
	199	}
	200
3aaabe23 GC	201	static u64 tcp_cgroup_read(struct cgroup cont, struct cftype cft)
	202	{
	203	struct mem_cgroup *memcg = mem_cgroup_from_cont(cont);
	204	u64 val;
	205
	206	switch (cft->private) {
	207	case RES_LIMIT:
	208	val = tcp_read_stat(memcg, RES_LIMIT, RESOURCE_MAX);
	209	break;
5a6dd343 GC	210	case RES_USAGE:
	211	val = tcp_read_usage(memcg);
	212	break;
ffea59e5	213	case RES_FAILCNT:
0850f0f5 GC	214	case RES_MAX_USAGE:
0850f0f5 GC	215	val = tcp_read_stat(memcg, cft->private, 0);
ffea59e5	216	break;
3aaabe23 GC	217	default:
	218	BUG();
	219	}
	220	return val;
	221	}
	222
ffea59e5 GC	223	static int tcp_cgroup_reset(struct cgroup *cont, unsigned int event)
	224	{
	225	struct mem_cgroup *memcg;
	226	struct tcp_memcontrol *tcp;
	227	struct cg_proto *cg_proto;
	228
	229	memcg = mem_cgroup_from_cont(cont);
	230	cg_proto = tcp_prot.proto_cgroup(memcg);
	231	if (!cg_proto)
	232	return 0;
	233	tcp = tcp_from_cgproto(cg_proto);
	234
	235	switch (event) {
0850f0f5 GC	236	case RES_MAX_USAGE:
	237	res_counter_reset_max(&tcp->tcp_memory_allocated);
	238	break;
ffea59e5 GC	239	case RES_FAILCNT:
	240	res_counter_reset_failcnt(&tcp->tcp_memory_allocated);
	241	break;
	242	}
	243
	244	return 0;
	245	}
	246
3aaabe23 GC	247	unsigned long long tcp_max_memory(const struct mem_cgroup *memcg)
	248	{
	249	struct tcp_memcontrol *tcp;
	250	struct cg_proto *cg_proto;
	251
	252	cg_proto = tcp_prot.proto_cgroup((struct mem_cgroup *)memcg);
	253	if (!cg_proto)
	254	return 0;
	255
	256	tcp = tcp_from_cgproto(cg_proto);
	257	return res_counter_read_u64(&tcp->tcp_memory_allocated, RES_LIMIT);
	258	}
	259
	260	void tcp_prot_mem(struct mem_cgroup *memcg, long val, int idx)
	261	{
	262	struct tcp_memcontrol *tcp;
	263	struct cg_proto *cg_proto;
	264
	265	cg_proto = tcp_prot.proto_cgroup(memcg);
	266	if (!cg_proto)
	267	return;
	268
	269	tcp = tcp_from_cgproto(cg_proto);
	270
	271	tcp->tcp_prot_mem[idx] = val;
	272	}