[mirror_ubuntu-bionic-kernel.git] / net / core / netprio_cgroup.c

/*
 * net/core/netprio_cgroup.c	Priority Control Group
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 *
 * Authors:	Neil Horman <nhorman@tuxdriver.com>
 */

#include <linux/module.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/skbuff.h>
#include <linux/cgroup.h>
#include <linux/rcupdate.h>
#include <linux/atomic.h>
#include <net/rtnetlink.h>
#include <net/pkt_cls.h>
#include <net/sock.h>
#include <net/netprio_cgroup.h>

#define PRIOIDX_SZ 128

static unsigned long prioidx_map[PRIOIDX_SZ];
static DEFINE_SPINLOCK(prioidx_map_lock);
static atomic_t max_prioidx = ATOMIC_INIT(0);

static inline struct cgroup_netprio_state *cgrp_netprio_state(struct cgroup *cgrp)
{
	return container_of(cgroup_subsys_state(cgrp, net_prio_subsys_id),
			    struct cgroup_netprio_state, css);
}

static int get_prioidx(u32 *prio)
{
	unsigned long flags;
	u32 prioidx;

	spin_lock_irqsave(&prioidx_map_lock, flags);
	prioidx = find_first_zero_bit(prioidx_map, sizeof(unsigned long) * PRIOIDX_SZ);
	if (prioidx == sizeof(unsigned long) * PRIOIDX_SZ) {
		spin_unlock_irqrestore(&prioidx_map_lock, flags);
		return -ENOSPC;
	}
	set_bit(prioidx, prioidx_map);
	spin_unlock_irqrestore(&prioidx_map_lock, flags);
	atomic_set(&max_prioidx, prioidx);
	*prio = prioidx;
	return 0;
}

static void put_prioidx(u32 idx)
{
	unsigned long flags;

	spin_lock_irqsave(&prioidx_map_lock, flags);
	clear_bit(idx, prioidx_map);
	spin_unlock_irqrestore(&prioidx_map_lock, flags);
}

static void extend_netdev_table(struct net_device *dev, u32 new_len)
{
	size_t new_size = sizeof(struct netprio_map) +
			   ((sizeof(u32) * new_len));
	struct netprio_map *new_priomap = kzalloc(new_size, GFP_KERNEL);
	struct netprio_map *old_priomap;
	int i;

	old_priomap  = rtnl_dereference(dev->priomap);

	if (!new_priomap) {
		printk(KERN_WARNING "Unable to alloc new priomap!\n");
		return;
	}

	for (i = 0;
	     old_priomap && (i < old_priomap->priomap_len);
	     i++)
		new_priomap->priomap[i] = old_priomap->priomap[i];

	new_priomap->priomap_len = new_len;

	rcu_assign_pointer(dev->priomap, new_priomap);
	if (old_priomap)
		kfree_rcu(old_priomap, rcu);
}

static void update_netdev_tables(void)
{
	struct net_device *dev;
	u32 max_len = atomic_read(&max_prioidx) + 1;
	struct netprio_map *map;

	rtnl_lock();
	for_each_netdev(&init_net, dev) {
		map = rtnl_dereference(dev->priomap);
		if ((!map) ||
		    (map->priomap_len < max_len))
			extend_netdev_table(dev, max_len);
	}
	rtnl_unlock();
}

static struct cgroup_subsys_state *cgrp_create(struct cgroup *cgrp)
{
	struct cgroup_netprio_state *cs;
	int ret;

	cs = kzalloc(sizeof(*cs), GFP_KERNEL);
	if (!cs)
		return ERR_PTR(-ENOMEM);

	if (cgrp->parent && cgrp_netprio_state(cgrp->parent)->prioidx) {
		kfree(cs);
		return ERR_PTR(-EINVAL);
	}

	ret = get_prioidx(&cs->prioidx);
	if (ret != 0) {
		printk(KERN_WARNING "No space in priority index array\n");
		kfree(cs);
		return ERR_PTR(ret);
	}

	return &cs->css;
}

static void cgrp_destroy(struct cgroup *cgrp)
{
	struct cgroup_netprio_state *cs;
	struct net_device *dev;
	struct netprio_map *map;

	cs = cgrp_netprio_state(cgrp);
	rtnl_lock();
	for_each_netdev(&init_net, dev) {
		map = rtnl_dereference(dev->priomap);
		if (map)
			map->priomap[cs->prioidx] = 0;
	}
	rtnl_unlock();
	put_prioidx(cs->prioidx);
	kfree(cs);
}

static u64 read_prioidx(struct cgroup *cgrp, struct cftype *cft)
{
	return (u64)cgrp_netprio_state(cgrp)->prioidx;
}

static int read_priomap(struct cgroup *cont, struct cftype *cft,
			struct cgroup_map_cb *cb)
{
	struct net_device *dev;
	u32 prioidx = cgrp_netprio_state(cont)->prioidx;
	u32 priority;
	struct netprio_map *map;

	rcu_read_lock();
	for_each_netdev_rcu(&init_net, dev) {
		map = rcu_dereference(dev->priomap);
		priority = map ? map->priomap[prioidx] : 0;
		cb->fill(cb, dev->name, priority);
	}
	rcu_read_unlock();
	return 0;
}

static int write_priomap(struct cgroup *cgrp, struct cftype *cft,
			 const char *buffer)
{
	char *devname = kstrdup(buffer, GFP_KERNEL);
	int ret = -EINVAL;
	u32 prioidx = cgrp_netprio_state(cgrp)->prioidx;
	unsigned long priority;
	char *priostr;
	struct net_device *dev;
	struct netprio_map *map;

	if (!devname)
		return -ENOMEM;

	/*
	 * Minimally sized valid priomap string
	 */
	if (strlen(devname) < 3)
		goto out_free_devname;

	priostr = strstr(devname, " ");
	if (!priostr)
		goto out_free_devname;

	/*
	 *Separate the devname from the associated priority
	 *and advance the priostr poitner to the priority value
	 */
	*priostr = '\0';
	priostr++;

	/*
	 * If the priostr points to NULL, we're at the end of the passed
	 * in string, and its not a valid write
	 */
	if (*priostr == '\0')
		goto out_free_devname;

	ret = kstrtoul(priostr, 10, &priority);
	if (ret < 0)
		goto out_free_devname;

	ret = -ENODEV;

	dev = dev_get_by_name(&init_net, devname);
	if (!dev)
		goto out_free_devname;

	update_netdev_tables();
	ret = 0;
	rcu_read_lock();
	map = rcu_dereference(dev->priomap);
	if (map)
		map->priomap[prioidx] = priority;
	rcu_read_unlock();
	dev_put(dev);

out_free_devname:
	kfree(devname);
	return ret;
}

static struct cftype ss_files[] = {
	{
		.name = "prioidx",
		.read_u64 = read_prioidx,
	},
	{
		.name = "ifpriomap",
		.read_map = read_priomap,
		.write_string = write_priomap,
	},
};

static int cgrp_populate(struct cgroup_subsys *ss, struct cgroup *cgrp)
{
	return cgroup_add_files(cgrp, ss, ss_files, ARRAY_SIZE(ss_files));
}

struct cgroup_subsys net_prio_subsys = {
	.name		= "net_prio",
	.create		= cgrp_create,
	.destroy	= cgrp_destroy,
	.populate	= cgrp_populate,
#ifdef CONFIG_NETPRIO_CGROUP
	.subsys_id	= net_prio_subsys_id,
#endif
	.module		= THIS_MODULE
};

static int netprio_device_event(struct notifier_block *unused,
				unsigned long event, void *ptr)
{
	struct net_device *dev = ptr;
	struct netprio_map *old;

	/*
	 * Note this is called with rtnl_lock held so we have update side
	 * protection on our rcu assignments
	 */

	switch (event) {
	case NETDEV_UNREGISTER:
		old = rtnl_dereference(dev->priomap);
		RCU_INIT_POINTER(dev->priomap, NULL);
		if (old)
			kfree_rcu(old, rcu);
		break;
	}
	return NOTIFY_DONE;
}

static struct notifier_block netprio_device_notifier = {
	.notifier_call = netprio_device_event
};

static int __init init_cgroup_netprio(void)
{
	int ret;

	ret = cgroup_load_subsys(&net_prio_subsys);
	if (ret)
		goto out;
#ifndef CONFIG_NETPRIO_CGROUP
	smp_wmb();
	net_prio_subsys_id = net_prio_subsys.subsys_id;
#endif

	register_netdevice_notifier(&netprio_device_notifier);

out:
	return ret;
}

static void __exit exit_cgroup_netprio(void)
{
	struct netprio_map *old;
	struct net_device *dev;

	unregister_netdevice_notifier(&netprio_device_notifier);

	cgroup_unload_subsys(&net_prio_subsys);

#ifndef CONFIG_NETPRIO_CGROUP
	net_prio_subsys_id = -1;
	synchronize_rcu();
#endif

	rtnl_lock();
	for_each_netdev(&init_net, dev) {
		old = rtnl_dereference(dev->priomap);
		RCU_INIT_POINTER(dev->priomap, NULL);
		if (old)
			kfree_rcu(old, rcu);
	}
	rtnl_unlock();
}

module_init(init_cgroup_netprio);
module_exit(exit_cgroup_netprio);
MODULE_LICENSE("GPL v2");
Commit	Line	Data
5bc1421e NH	1	/*
	2	* net/core/netprio_cgroup.c Priority Control Group
	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of the GNU General Public License
	6	* as published by the Free Software Foundation; either version
	7	* 2 of the License, or (at your option) any later version.
	8	*
	9	* Authors: Neil Horman <nhorman@tuxdriver.com>
	10	*/
	11
	12	#include <linux/module.h>
	13	#include <linux/slab.h>
	14	#include <linux/types.h>
	15	#include <linux/string.h>
	16	#include <linux/errno.h>
	17	#include <linux/skbuff.h>
	18	#include <linux/cgroup.h>
	19	#include <linux/rcupdate.h>
	20	#include <linux/atomic.h>
	21	#include <net/rtnetlink.h>
	22	#include <net/pkt_cls.h>
	23	#include <net/sock.h>
	24	#include <net/netprio_cgroup.h>
	25
5bc1421e NH	26	#define PRIOIDX_SZ 128
	27
	28	static unsigned long prioidx_map[PRIOIDX_SZ];
	29	static DEFINE_SPINLOCK(prioidx_map_lock);
	30	static atomic_t max_prioidx = ATOMIC_INIT(0);
	31
	32	static inline struct cgroup_netprio_state cgrp_netprio_state(struct cgroup cgrp)
	33	{
	34	return container_of(cgroup_subsys_state(cgrp, net_prio_subsys_id),
	35	struct cgroup_netprio_state, css);
	36	}
	37
	38	static int get_prioidx(u32 *prio)
	39	{
	40	unsigned long flags;
	41	u32 prioidx;
	42
	43	spin_lock_irqsave(&prioidx_map_lock, flags);
	44	prioidx = find_first_zero_bit(prioidx_map, sizeof(unsigned long) * PRIOIDX_SZ);
5962b35c NH	45	if (prioidx == sizeof(unsigned long) * PRIOIDX_SZ) {
	46	spin_unlock_irqrestore(&prioidx_map_lock, flags);
	47	return -ENOSPC;
	48	}
5bc1421e NH	49	set_bit(prioidx, prioidx_map);
5bc1421e NH	50	spin_unlock_irqrestore(&prioidx_map_lock, flags);
5bc1421e NH	51	atomic_set(&max_prioidx, prioidx);
	52	*prio = prioidx;
	53	return 0;
	54	}
	55
	56	static void put_prioidx(u32 idx)
	57	{
	58	unsigned long flags;
	59
	60	spin_lock_irqsave(&prioidx_map_lock, flags);
	61	clear_bit(idx, prioidx_map);
	62	spin_unlock_irqrestore(&prioidx_map_lock, flags);
	63	}
	64
	65	static void extend_netdev_table(struct net_device *dev, u32 new_len)
	66	{
	67	size_t new_size = sizeof(struct netprio_map) +
	68	((sizeof(u32) * new_len));
	69	struct netprio_map *new_priomap = kzalloc(new_size, GFP_KERNEL);
	70	struct netprio_map *old_priomap;
	71	int i;
	72
	73	old_priomap = rtnl_dereference(dev->priomap);
	74
	75	if (!new_priomap) {
	76	printk(KERN_WARNING "Unable to alloc new priomap!\n");
	77	return;
	78	}
	79
	80	for (i = 0;
	81	old_priomap && (i < old_priomap->priomap_len);
	82	i++)
	83	new_priomap->priomap[i] = old_priomap->priomap[i];
	84
	85	new_priomap->priomap_len = new_len;
	86
	87	rcu_assign_pointer(dev->priomap, new_priomap);
	88	if (old_priomap)
	89	kfree_rcu(old_priomap, rcu);
	90	}
	91
	92	static void update_netdev_tables(void)
	93	{
	94	struct net_device *dev;
a87dfe14	95	u32 max_len = atomic_read(&max_prioidx) + 1;
5bc1421e NH	96	struct netprio_map *map;
	97
	98	rtnl_lock();
	99	for_each_netdev(&init_net, dev) {
	100	map = rtnl_dereference(dev->priomap);
	101	if ((!map) \|\|
	102	(map->priomap_len < max_len))
	103	extend_netdev_table(dev, max_len);
	104	}
	105	rtnl_unlock();
	106	}
	107
761b3ef5	108	static struct cgroup_subsys_state cgrp_create(struct cgroup cgrp)
5bc1421e NH	109	{
	110	struct cgroup_netprio_state *cs;
	111	int ret;
	112
	113	cs = kzalloc(sizeof(*cs), GFP_KERNEL);
	114	if (!cs)
	115	return ERR_PTR(-ENOMEM);
	116
	117	if (cgrp->parent && cgrp_netprio_state(cgrp->parent)->prioidx) {
	118	kfree(cs);
	119	return ERR_PTR(-EINVAL);
	120	}
	121
	122	ret = get_prioidx(&cs->prioidx);
	123	if (ret != 0) {
	124	printk(KERN_WARNING "No space in priority index array\n");
	125	kfree(cs);
	126	return ERR_PTR(ret);
	127	}
	128
	129	return &cs->css;
	130	}
	131
761b3ef5	132	static void cgrp_destroy(struct cgroup *cgrp)
5bc1421e NH	133	{
	134	struct cgroup_netprio_state *cs;
	135	struct net_device *dev;
	136	struct netprio_map *map;
	137
	138	cs = cgrp_netprio_state(cgrp);
	139	rtnl_lock();
	140	for_each_netdev(&init_net, dev) {
	141	map = rtnl_dereference(dev->priomap);
	142	if (map)
	143	map->priomap[cs->prioidx] = 0;
	144	}
	145	rtnl_unlock();
	146	put_prioidx(cs->prioidx);
	147	kfree(cs);
	148	}
	149
	150	static u64 read_prioidx(struct cgroup cgrp, struct cftype cft)
	151	{
	152	return (u64)cgrp_netprio_state(cgrp)->prioidx;
	153	}
	154
	155	static int read_priomap(struct cgroup cont, struct cftype cft,
	156	struct cgroup_map_cb *cb)
	157	{
	158	struct net_device *dev;
	159	u32 prioidx = cgrp_netprio_state(cont)->prioidx;
	160	u32 priority;
	161	struct netprio_map *map;
	162
	163	rcu_read_lock();
	164	for_each_netdev_rcu(&init_net, dev) {
	165	map = rcu_dereference(dev->priomap);
	166	priority = map ? map->priomap[prioidx] : 0;
	167	cb->fill(cb, dev->name, priority);
	168	}
	169	rcu_read_unlock();
	170	return 0;
	171	}
	172
	173	static int write_priomap(struct cgroup cgrp, struct cftype cft,
	174	const char *buffer)
	175	{
	176	char *devname = kstrdup(buffer, GFP_KERNEL);
	177	int ret = -EINVAL;
	178	u32 prioidx = cgrp_netprio_state(cgrp)->prioidx;
	179	unsigned long priority;
	180	char *priostr;
	181	struct net_device *dev;
	182	struct netprio_map *map;
	183
	184	if (!devname)
	185	return -ENOMEM;
	186
	187	/*
	188	* Minimally sized valid priomap string
	189	*/
	190	if (strlen(devname) < 3)
	191	goto out_free_devname;
	192
	193	priostr = strstr(devname, " ");
	194	if (!priostr)
	195	goto out_free_devname;
	196
197	/*
198	*Separate the devname from the associated priority
199	*and advance the priostr poitner to the priority value
200	*/
201	*priostr = '\0';
202	priostr++;
203
204	/*
205	* If the priostr points to NULL, we're at the end of the passed
206	* in string, and its not a valid write
207	*/
208	if (*priostr == '\0')
209	goto out_free_devname;
210
211	ret = kstrtoul(priostr, 10, &priority);
212	if (ret < 0)
213	goto out_free_devname;
214
215	ret = -ENODEV;
216
217	dev = dev_get_by_name(&init_net, devname);
218	if (!dev)
219	goto out_free_devname;
220
221	update_netdev_tables();
222	ret = 0;
223	rcu_read_lock();
224	map = rcu_dereference(dev->priomap);
225	if (map)
226	map->priomap[prioidx] = priority;
227	rcu_read_unlock();
228	dev_put(dev);
229
230	out_free_devname:
231	kfree(devname);
232	return ret;
233	}
234
235	static struct cftype ss_files[] = {
236	{
237	.name = "prioidx",
238	.read_u64 = read_prioidx,
239	},
240	{
241	.name = "ifpriomap",
242	.read_map = read_priomap,
243	.write_string = write_priomap,
244	},
245	};
246
247	static int cgrp_populate(struct cgroup_subsys ss, struct cgroup cgrp)
248	{
249	return cgroup_add_files(cgrp, ss, ss_files, ARRAY_SIZE(ss_files));
250	}
251
676f7c8f TH	252	struct cgroup_subsys net_prio_subsys = {
	253	.name = "net_prio",
	254	.create = cgrp_create,
	255	.destroy = cgrp_destroy,
	256	.populate = cgrp_populate,
	257	#ifdef CONFIG_NETPRIO_CGROUP
	258	.subsys_id = net_prio_subsys_id,
	259	#endif
	260	.module = THIS_MODULE
	261	};
	262
5bc1421e NH	263	static int netprio_device_event(struct notifier_block *unused,
	264	unsigned long event, void *ptr)
	265	{
	266	struct net_device *dev = ptr;
	267	struct netprio_map *old;
5bc1421e NH	268
	269	/*
	270	* Note this is called with rtnl_lock held so we have update side
	271	* protection on our rcu assignments
	272	*/
	273
	274	switch (event) {
5bc1421e NH	275	case NETDEV_UNREGISTER:
5bc1421e NH	276	old = rtnl_dereference(dev->priomap);
2cfa5a04	277	RCU_INIT_POINTER(dev->priomap, NULL);
5bc1421e NH	278	if (old)
	279	kfree_rcu(old, rcu);
	280	break;
	281	}
	282	return NOTIFY_DONE;
	283	}
	284
	285	static struct notifier_block netprio_device_notifier = {
	286	.notifier_call = netprio_device_event
	287	};
	288
	289	static int __init init_cgroup_netprio(void)
	290	{
	291	int ret;
	292
	293	ret = cgroup_load_subsys(&net_prio_subsys);
	294	if (ret)
	295	goto out;
	296	#ifndef CONFIG_NETPRIO_CGROUP
	297	smp_wmb();
	298	net_prio_subsys_id = net_prio_subsys.subsys_id;
	299	#endif
	300
	301	register_netdevice_notifier(&netprio_device_notifier);
	302
	303	out:
	304	return ret;
	305	}
	306
	307	static void __exit exit_cgroup_netprio(void)
	308	{
	309	struct netprio_map *old;
	310	struct net_device *dev;
	311
	312	unregister_netdevice_notifier(&netprio_device_notifier);
	313
	314	cgroup_unload_subsys(&net_prio_subsys);
	315
	316	#ifndef CONFIG_NETPRIO_CGROUP
	317	net_prio_subsys_id = -1;
	318	synchronize_rcu();
	319	#endif
	320
	321	rtnl_lock();
	322	for_each_netdev(&init_net, dev) {
	323	old = rtnl_dereference(dev->priomap);
2cfa5a04	324	RCU_INIT_POINTER(dev->priomap, NULL);
5bc1421e NH	325	if (old)
	326	kfree_rcu(old, rcu);
	327	}
	328	rtnl_unlock();
	329	}
	330
	331	module_init(init_cgroup_netprio);
	332	module_exit(exit_cgroup_netprio);
	333	MODULE_LICENSE("GPL v2");