[linux.git] / security / device_cgroup.c

/*
 * device_cgroup.c - device cgroup subsystem
 *
 * Copyright 2007 IBM Corp
 */

#include <linux/device_cgroup.h>
#include <linux/cgroup.h>
#include <linux/ctype.h>
#include <linux/list.h>
#include <linux/uaccess.h>
#include <linux/seq_file.h>
#include <linux/rcupdate.h>
#include <linux/mutex.h>

#define ACC_MKNOD 1
#define ACC_READ  2
#define ACC_WRITE 4
#define ACC_MASK (ACC_MKNOD | ACC_READ | ACC_WRITE)

#define DEV_BLOCK 1
#define DEV_CHAR  2
#define DEV_ALL   4  /* this represents all devices */

static DEFINE_MUTEX(devcgroup_mutex);

/*
 * whitelist locking rules:
 * hold devcgroup_mutex for update/read.
 * hold rcu_read_lock() for read.
 */

struct dev_whitelist_item {
	u32 major, minor;
	short type;
	short access;
	struct list_head list;
	struct rcu_head rcu;
};

struct dev_cgroup {
	struct cgroup_subsys_state css;
	struct list_head whitelist;
};

static inline struct dev_cgroup *css_to_devcgroup(struct cgroup_subsys_state *s)
{
	return container_of(s, struct dev_cgroup, css);
}

static inline struct dev_cgroup *cgroup_to_devcgroup(struct cgroup *cgroup)
{
	return css_to_devcgroup(cgroup_subsys_state(cgroup, devices_subsys_id));
}

static inline struct dev_cgroup *task_devcgroup(struct task_struct *task)
{
	return css_to_devcgroup(task_subsys_state(task, devices_subsys_id));
}

struct cgroup_subsys devices_subsys;

static int devcgroup_can_attach(struct cgroup_subsys *ss,
		struct cgroup *new_cgroup, struct task_struct *task,
		bool threadgroup)
{
	if (current != task && !capable(CAP_SYS_ADMIN))
			return -EPERM;

	return 0;
}

/*
 * called under devcgroup_mutex
 */
static int dev_whitelist_copy(struct list_head *dest, struct list_head *orig)
{
	struct dev_whitelist_item *wh, *tmp, *new;

	list_for_each_entry(wh, orig, list) {
		new = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
		if (!new)
			goto free_and_exit;
		list_add_tail(&new->list, dest);
	}

	return 0;

free_and_exit:
	list_for_each_entry_safe(wh, tmp, dest, list) {
		list_del(&wh->list);
		kfree(wh);
	}
	return -ENOMEM;
}

/* Stupid prototype - don't bother combining existing entries */
/*
 * called under devcgroup_mutex
 */
static int dev_whitelist_add(struct dev_cgroup *dev_cgroup,
			struct dev_whitelist_item *wh)
{
	struct dev_whitelist_item *whcopy, *walk;

	whcopy = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
	if (!whcopy)
		return -ENOMEM;

	list_for_each_entry(walk, &dev_cgroup->whitelist, list) {
		if (walk->type != wh->type)
			continue;
		if (walk->major != wh->major)
			continue;
		if (walk->minor != wh->minor)
			continue;

		walk->access |= wh->access;
		kfree(whcopy);
		whcopy = NULL;
	}

	if (whcopy != NULL)
		list_add_tail_rcu(&whcopy->list, &dev_cgroup->whitelist);
	return 0;
}

static void whitelist_item_free(struct rcu_head *rcu)
{
	struct dev_whitelist_item *item;

	item = container_of(rcu, struct dev_whitelist_item, rcu);
	kfree(item);
}

/*
 * called under devcgroup_mutex
 */
static void dev_whitelist_rm(struct dev_cgroup *dev_cgroup,
			struct dev_whitelist_item *wh)
{
	struct dev_whitelist_item *walk, *tmp;

	list_for_each_entry_safe(walk, tmp, &dev_cgroup->whitelist, list) {
		if (walk->type == DEV_ALL)
			goto remove;
		if (walk->type != wh->type)
			continue;
		if (walk->major != ~0 && walk->major != wh->major)
			continue;
		if (walk->minor != ~0 && walk->minor != wh->minor)
			continue;

remove:
		walk->access &= ~wh->access;
		if (!walk->access) {
			list_del_rcu(&walk->list);
			call_rcu(&walk->rcu, whitelist_item_free);
		}
	}
}

/*
 * called from kernel/cgroup.c with cgroup_lock() held.
 */
static struct cgroup_subsys_state *devcgroup_create(struct cgroup_subsys *ss,
						struct cgroup *cgroup)
{
	struct dev_cgroup *dev_cgroup, *parent_dev_cgroup;
	struct cgroup *parent_cgroup;
	int ret;

	dev_cgroup = kzalloc(sizeof(*dev_cgroup), GFP_KERNEL);
	if (!dev_cgroup)
		return ERR_PTR(-ENOMEM);
	INIT_LIST_HEAD(&dev_cgroup->whitelist);
	parent_cgroup = cgroup->parent;

	if (parent_cgroup == NULL) {
		struct dev_whitelist_item *wh;
		wh = kmalloc(sizeof(*wh), GFP_KERNEL);
		if (!wh) {
			kfree(dev_cgroup);
			return ERR_PTR(-ENOMEM);
		}
		wh->minor = wh->major = ~0;
		wh->type = DEV_ALL;
		wh->access = ACC_MASK;
		list_add(&wh->list, &dev_cgroup->whitelist);
	} else {
		parent_dev_cgroup = cgroup_to_devcgroup(parent_cgroup);
		mutex_lock(&devcgroup_mutex);
		ret = dev_whitelist_copy(&dev_cgroup->whitelist,
				&parent_dev_cgroup->whitelist);
		mutex_unlock(&devcgroup_mutex);
		if (ret) {
			kfree(dev_cgroup);
			return ERR_PTR(ret);
		}
	}

	return &dev_cgroup->css;
}

static void devcgroup_destroy(struct cgroup_subsys *ss,
			struct cgroup *cgroup)
{
	struct dev_cgroup *dev_cgroup;
	struct dev_whitelist_item *wh, *tmp;

	dev_cgroup = cgroup_to_devcgroup(cgroup);
	list_for_each_entry_safe(wh, tmp, &dev_cgroup->whitelist, list) {
		list_del(&wh->list);
		kfree(wh);
	}
	kfree(dev_cgroup);
}

#define DEVCG_ALLOW 1
#define DEVCG_DENY 2
#define DEVCG_LIST 3

#define MAJMINLEN 13
#define ACCLEN 4

static void set_access(char *acc, short access)
{
	int idx = 0;
	memset(acc, 0, ACCLEN);
	if (access & ACC_READ)
		acc[idx++] = 'r';
	if (access & ACC_WRITE)
		acc[idx++] = 'w';
	if (access & ACC_MKNOD)
		acc[idx++] = 'm';
}

static char type_to_char(short type)
{
	if (type == DEV_ALL)
		return 'a';
	if (type == DEV_CHAR)
		return 'c';
	if (type == DEV_BLOCK)
		return 'b';
	return 'X';
}

static void set_majmin(char *str, unsigned m)
{
	if (m == ~0)
		strcpy(str, "*");
	else
		sprintf(str, "%u", m);
}

static int devcgroup_seq_read(struct cgroup *cgroup, struct cftype *cft,
				struct seq_file *m)
{
	struct dev_cgroup *devcgroup = cgroup_to_devcgroup(cgroup);
	struct dev_whitelist_item *wh;
	char maj[MAJMINLEN], min[MAJMINLEN], acc[ACCLEN];

	rcu_read_lock();
	list_for_each_entry_rcu(wh, &devcgroup->whitelist, list) {
		set_access(acc, wh->access);
		set_majmin(maj, wh->major);
		set_majmin(min, wh->minor);
		seq_printf(m, "%c %s:%s %s\n", type_to_char(wh->type),
			   maj, min, acc);
	}
	rcu_read_unlock();

	return 0;
}

/*
 * may_access_whitelist:
 * does the access granted to dev_cgroup c contain the access
 * requested in whitelist item refwh.
 * return 1 if yes, 0 if no.
 * call with devcgroup_mutex held
 */
static int may_access_whitelist(struct dev_cgroup *c,
				       struct dev_whitelist_item *refwh)
{
	struct dev_whitelist_item *whitem;

	list_for_each_entry(whitem, &c->whitelist, list) {
		if (whitem->type & DEV_ALL)
			return 1;
		if ((refwh->type & DEV_BLOCK) && !(whitem->type & DEV_BLOCK))
			continue;
		if ((refwh->type & DEV_CHAR) && !(whitem->type & DEV_CHAR))
			continue;
		if (whitem->major != ~0 && whitem->major != refwh->major)
			continue;
		if (whitem->minor != ~0 && whitem->minor != refwh->minor)
			continue;
		if (refwh->access & (~whitem->access))
			continue;
		return 1;
	}
	return 0;
}

/*
 * parent_has_perm:
 * when adding a new allow rule to a device whitelist, the rule
 * must be allowed in the parent device
 */
static int parent_has_perm(struct dev_cgroup *childcg,
				  struct dev_whitelist_item *wh)
{
	struct cgroup *pcg = childcg->css.cgroup->parent;
	struct dev_cgroup *parent;

	if (!pcg)
		return 1;
	parent = cgroup_to_devcgroup(pcg);
	return may_access_whitelist(parent, wh);
}

/*
 * Modify the whitelist using allow/deny rules.
 * CAP_SYS_ADMIN is needed for this.  It's at least separate from CAP_MKNOD
 * so we can give a container CAP_MKNOD to let it create devices but not
 * modify the whitelist.
 * It seems likely we'll want to add a CAP_CONTAINER capability to allow
 * us to also grant CAP_SYS_ADMIN to containers without giving away the
 * device whitelist controls, but for now we'll stick with CAP_SYS_ADMIN
 *
 * Taking rules away is always allowed (given CAP_SYS_ADMIN).  Granting
 * new access is only allowed if you're in the top-level cgroup, or your
 * parent cgroup has the access you're asking for.
 */
static int devcgroup_update_access(struct dev_cgroup *devcgroup,
				   int filetype, const char *buffer)
{
	const char *b;
	char *endp;
	int count;
	struct dev_whitelist_item wh;

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

	memset(&wh, 0, sizeof(wh));
	b = buffer;

	switch (*b) {
	case 'a':
		wh.type = DEV_ALL;
		wh.access = ACC_MASK;
		wh.major = ~0;
		wh.minor = ~0;
		goto handle;
	case 'b':
		wh.type = DEV_BLOCK;
		break;
	case 'c':
		wh.type = DEV_CHAR;
		break;
	default:
		return -EINVAL;
	}
	b++;
	if (!isspace(*b))
		return -EINVAL;
	b++;
	if (*b == '*') {
		wh.major = ~0;
		b++;
	} else if (isdigit(*b)) {
		wh.major = simple_strtoul(b, &endp, 10);
		b = endp;
	} else {
		return -EINVAL;
	}
	if (*b != ':')
		return -EINVAL;
	b++;

	/* read minor */
	if (*b == '*') {
		wh.minor = ~0;
		b++;
	} else if (isdigit(*b)) {
		wh.minor = simple_strtoul(b, &endp, 10);
		b = endp;
	} else {
		return -EINVAL;
	}
	if (!isspace(*b))
		return -EINVAL;
	for (b++, count = 0; count < 3; count++, b++) {
		switch (*b) {
		case 'r':
			wh.access |= ACC_READ;
			break;
		case 'w':
			wh.access |= ACC_WRITE;
			break;
		case 'm':
			wh.access |= ACC_MKNOD;
			break;
		case '\n':
		case '\0':
			count = 3;
			break;
		default:
			return -EINVAL;
		}
	}

handle:
	switch (filetype) {
	case DEVCG_ALLOW:
		if (!parent_has_perm(devcgroup, &wh))
			return -EPERM;
		return dev_whitelist_add(devcgroup, &wh);
	case DEVCG_DENY:
		dev_whitelist_rm(devcgroup, &wh);
		break;
	default:
		return -EINVAL;
	}
	return 0;
}

static int devcgroup_access_write(struct cgroup *cgrp, struct cftype *cft,
				  const char *buffer)
{
	int retval;

	mutex_lock(&devcgroup_mutex);
	retval = devcgroup_update_access(cgroup_to_devcgroup(cgrp),
					 cft->private, buffer);
	mutex_unlock(&devcgroup_mutex);
	return retval;
}

static struct cftype dev_cgroup_files[] = {
	{
		.name = "allow",
		.write_string  = devcgroup_access_write,
		.private = DEVCG_ALLOW,
	},
	{
		.name = "deny",
		.write_string = devcgroup_access_write,
		.private = DEVCG_DENY,
	},
	{
		.name = "list",
		.read_seq_string = devcgroup_seq_read,
		.private = DEVCG_LIST,
	},
};

static int devcgroup_populate(struct cgroup_subsys *ss,
				struct cgroup *cgroup)
{
	return cgroup_add_files(cgroup, ss, dev_cgroup_files,
					ARRAY_SIZE(dev_cgroup_files));
}

struct cgroup_subsys devices_subsys = {
	.name = "devices",
	.can_attach = devcgroup_can_attach,
	.create = devcgroup_create,
	.destroy  = devcgroup_destroy,
	.populate = devcgroup_populate,
	.subsys_id = devices_subsys_id,
};

int devcgroup_inode_permission(struct inode *inode, int mask)
{
	struct dev_cgroup *dev_cgroup;
	struct dev_whitelist_item *wh;

	dev_t device = inode->i_rdev;
	if (!device)
		return 0;
	if (!S_ISBLK(inode->i_mode) && !S_ISCHR(inode->i_mode))
		return 0;

	rcu_read_lock();

	dev_cgroup = task_devcgroup(current);

	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
		if (wh->type & DEV_ALL)
			goto found;
		if ((wh->type & DEV_BLOCK) && !S_ISBLK(inode->i_mode))
			continue;
		if ((wh->type & DEV_CHAR) && !S_ISCHR(inode->i_mode))
			continue;
		if (wh->major != ~0 && wh->major != imajor(inode))
			continue;
		if (wh->minor != ~0 && wh->minor != iminor(inode))
			continue;

		if ((mask & MAY_WRITE) && !(wh->access & ACC_WRITE))
			continue;
		if ((mask & MAY_READ) && !(wh->access & ACC_READ))
			continue;
found:
		rcu_read_unlock();
		return 0;
	}

	rcu_read_unlock();

	return -EPERM;
}

int devcgroup_inode_mknod(int mode, dev_t dev)
{
	struct dev_cgroup *dev_cgroup;
	struct dev_whitelist_item *wh;

	if (!S_ISBLK(mode) && !S_ISCHR(mode))
		return 0;

	rcu_read_lock();

	dev_cgroup = task_devcgroup(current);

	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
		if (wh->type & DEV_ALL)
			goto found;
		if ((wh->type & DEV_BLOCK) && !S_ISBLK(mode))
			continue;
		if ((wh->type & DEV_CHAR) && !S_ISCHR(mode))
			continue;
		if (wh->major != ~0 && wh->major != MAJOR(dev))
			continue;
		if (wh->minor != ~0 && wh->minor != MINOR(dev))
			continue;

		if (!(wh->access & ACC_MKNOD))
			continue;
found:
		rcu_read_unlock();
		return 0;
	}

	rcu_read_unlock();

	return -EPERM;
}
Commit	Line	Data
08ce5f16	1	/*
47c59803	2	* device_cgroup.c - device cgroup subsystem
08ce5f16 SH	3	*
	4	* Copyright 2007 IBM Corp
	5	*/
	6
	7	#include <linux/device_cgroup.h>
	8	#include <linux/cgroup.h>
	9	#include <linux/ctype.h>
	10	#include <linux/list.h>
	11	#include <linux/uaccess.h>
29486df3	12	#include <linux/seq_file.h>
47c59803	13	#include <linux/rcupdate.h>
b4046f00	14	#include <linux/mutex.h>
08ce5f16 SH	15
	16	#define ACC_MKNOD 1
	17	#define ACC_READ 2
	18	#define ACC_WRITE 4
	19	#define ACC_MASK (ACC_MKNOD \| ACC_READ \| ACC_WRITE)
	20
	21	#define DEV_BLOCK 1
	22	#define DEV_CHAR 2
	23	#define DEV_ALL 4 /* this represents all devices */
	24
b4046f00 LZ	25	static DEFINE_MUTEX(devcgroup_mutex);
b4046f00 LZ	26
08ce5f16 SH	27	/*
08ce5f16 SH	28	* whitelist locking rules:
b4046f00	29	* hold devcgroup_mutex for update/read.
47c59803	30	* hold rcu_read_lock() for read.
08ce5f16 SH	31	*/
	32
	33	struct dev_whitelist_item {
	34	u32 major, minor;
	35	short type;
	36	short access;
	37	struct list_head list;
4efd1a1b	38	struct rcu_head rcu;
08ce5f16 SH	39	};
	40
	41	struct dev_cgroup {
	42	struct cgroup_subsys_state css;
	43	struct list_head whitelist;
08ce5f16 SH	44	};
08ce5f16 SH	45
b66862f7 PE	46	static inline struct dev_cgroup css_to_devcgroup(struct cgroup_subsys_state s)
	47	{
	48	return container_of(s, struct dev_cgroup, css);
	49	}
	50
08ce5f16 SH	51	static inline struct dev_cgroup cgroup_to_devcgroup(struct cgroup cgroup)
08ce5f16 SH	52	{
b66862f7	53	return css_to_devcgroup(cgroup_subsys_state(cgroup, devices_subsys_id));
08ce5f16 SH	54	}
08ce5f16 SH	55
f92523e3 PM	56	static inline struct dev_cgroup task_devcgroup(struct task_struct task)
	57	{
	58	return css_to_devcgroup(task_subsys_state(task, devices_subsys_id));
	59	}
	60
08ce5f16 SH	61	struct cgroup_subsys devices_subsys;
	62
	63	static int devcgroup_can_attach(struct cgroup_subsys *ss,
be367d09 BB	64	struct cgroup new_cgroup, struct task_struct task,
be367d09 BB	65	bool threadgroup)
08ce5f16 SH	66	{
	67	if (current != task && !capable(CAP_SYS_ADMIN))
	68	return -EPERM;
	69
	70	return 0;
	71	}
	72
	73	/*
b4046f00	74	* called under devcgroup_mutex
08ce5f16 SH	75	*/
	76	static int dev_whitelist_copy(struct list_head dest, struct list_head orig)
	77	{
	78	struct dev_whitelist_item wh, tmp, *new;
	79
	80	list_for_each_entry(wh, orig, list) {
2cdc7241	81	new = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
08ce5f16 SH	82	if (!new)
08ce5f16 SH	83	goto free_and_exit;
08ce5f16 SH	84	list_add_tail(&new->list, dest);
	85	}
	86
	87	return 0;
	88
	89	free_and_exit:
	90	list_for_each_entry_safe(wh, tmp, dest, list) {
	91	list_del(&wh->list);
	92	kfree(wh);
	93	}
	94	return -ENOMEM;
	95	}
	96
	97	/* Stupid prototype - don't bother combining existing entries */
	98	/*
b4046f00	99	* called under devcgroup_mutex
08ce5f16 SH	100	*/
	101	static int dev_whitelist_add(struct dev_cgroup *dev_cgroup,
	102	struct dev_whitelist_item *wh)
	103	{
d1ee2971	104	struct dev_whitelist_item whcopy, walk;
08ce5f16	105
2cdc7241	106	whcopy = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
08ce5f16 SH	107	if (!whcopy)
	108	return -ENOMEM;
	109
d1ee2971 PE	110	list_for_each_entry(walk, &dev_cgroup->whitelist, list) {
	111	if (walk->type != wh->type)
	112	continue;
	113	if (walk->major != wh->major)
	114	continue;
	115	if (walk->minor != wh->minor)
	116	continue;
	117
	118	walk->access \|= wh->access;
	119	kfree(whcopy);
	120	whcopy = NULL;
	121	}
	122
	123	if (whcopy != NULL)
4efd1a1b	124	list_add_tail_rcu(&whcopy->list, &dev_cgroup->whitelist);
08ce5f16 SH	125	return 0;
	126	}
	127
4efd1a1b PE	128	static void whitelist_item_free(struct rcu_head *rcu)
	129	{
	130	struct dev_whitelist_item *item;
	131
	132	item = container_of(rcu, struct dev_whitelist_item, rcu);
	133	kfree(item);
	134	}
	135
08ce5f16	136	/*
b4046f00	137	* called under devcgroup_mutex
08ce5f16 SH	138	*/
	139	static void dev_whitelist_rm(struct dev_cgroup *dev_cgroup,
	140	struct dev_whitelist_item *wh)
	141	{
	142	struct dev_whitelist_item walk, tmp;
	143
08ce5f16 SH	144	list_for_each_entry_safe(walk, tmp, &dev_cgroup->whitelist, list) {
	145	if (walk->type == DEV_ALL)
	146	goto remove;
	147	if (walk->type != wh->type)
	148	continue;
	149	if (walk->major != ~0 && walk->major != wh->major)
	150	continue;
	151	if (walk->minor != ~0 && walk->minor != wh->minor)
	152	continue;
	153
	154	remove:
	155	walk->access &= ~wh->access;
	156	if (!walk->access) {
4efd1a1b PE	157	list_del_rcu(&walk->list);
4efd1a1b PE	158	call_rcu(&walk->rcu, whitelist_item_free);
08ce5f16 SH	159	}
08ce5f16 SH	160	}
08ce5f16 SH	161	}
	162
	163	/*
	164	* called from kernel/cgroup.c with cgroup_lock() held.
	165	*/
	166	static struct cgroup_subsys_state devcgroup_create(struct cgroup_subsys ss,
	167	struct cgroup *cgroup)
	168	{
	169	struct dev_cgroup dev_cgroup, parent_dev_cgroup;
	170	struct cgroup *parent_cgroup;
	171	int ret;
	172
	173	dev_cgroup = kzalloc(sizeof(*dev_cgroup), GFP_KERNEL);
	174	if (!dev_cgroup)
	175	return ERR_PTR(-ENOMEM);
	176	INIT_LIST_HEAD(&dev_cgroup->whitelist);
	177	parent_cgroup = cgroup->parent;
	178
	179	if (parent_cgroup == NULL) {
	180	struct dev_whitelist_item *wh;
	181	wh = kmalloc(sizeof(*wh), GFP_KERNEL);
	182	if (!wh) {
	183	kfree(dev_cgroup);
	184	return ERR_PTR(-ENOMEM);
	185	}
	186	wh->minor = wh->major = ~0;
	187	wh->type = DEV_ALL;
7759fc9d	188	wh->access = ACC_MASK;
08ce5f16 SH	189	list_add(&wh->list, &dev_cgroup->whitelist);
	190	} else {
	191	parent_dev_cgroup = cgroup_to_devcgroup(parent_cgroup);
b4046f00	192	mutex_lock(&devcgroup_mutex);
08ce5f16 SH	193	ret = dev_whitelist_copy(&dev_cgroup->whitelist,
08ce5f16 SH	194	&parent_dev_cgroup->whitelist);
b4046f00	195	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	196	if (ret) {
	197	kfree(dev_cgroup);
	198	return ERR_PTR(ret);
	199	}
	200	}
	201
08ce5f16 SH	202	return &dev_cgroup->css;
	203	}
	204
	205	static void devcgroup_destroy(struct cgroup_subsys *ss,
	206	struct cgroup *cgroup)
	207	{
	208	struct dev_cgroup *dev_cgroup;
	209	struct dev_whitelist_item wh, tmp;
	210
	211	dev_cgroup = cgroup_to_devcgroup(cgroup);
	212	list_for_each_entry_safe(wh, tmp, &dev_cgroup->whitelist, list) {
	213	list_del(&wh->list);
	214	kfree(wh);
	215	}
	216	kfree(dev_cgroup);
	217	}
	218
	219	#define DEVCG_ALLOW 1
	220	#define DEVCG_DENY 2
29486df3 SH	221	#define DEVCG_LIST 3
29486df3 SH	222
17d213f8	223	#define MAJMINLEN 13
29486df3	224	#define ACCLEN 4
08ce5f16 SH	225
	226	static void set_access(char *acc, short access)
	227	{
	228	int idx = 0;
29486df3	229	memset(acc, 0, ACCLEN);
08ce5f16 SH	230	if (access & ACC_READ)
	231	acc[idx++] = 'r';
	232	if (access & ACC_WRITE)
	233	acc[idx++] = 'w';
	234	if (access & ACC_MKNOD)
	235	acc[idx++] = 'm';
	236	}
	237
	238	static char type_to_char(short type)
	239	{
	240	if (type == DEV_ALL)
	241	return 'a';
	242	if (type == DEV_CHAR)
	243	return 'c';
	244	if (type == DEV_BLOCK)
	245	return 'b';
	246	return 'X';
	247	}
	248
29486df3	249	static void set_majmin(char *str, unsigned m)
08ce5f16	250	{
08ce5f16	251	if (m == ~0)
7759fc9d	252	strcpy(str, "*");
08ce5f16	253	else
7759fc9d	254	sprintf(str, "%u", m);
08ce5f16 SH	255	}
08ce5f16 SH	256
29486df3 SH	257	static int devcgroup_seq_read(struct cgroup cgroup, struct cftype cft,
29486df3 SH	258	struct seq_file *m)
08ce5f16	259	{
29486df3	260	struct dev_cgroup *devcgroup = cgroup_to_devcgroup(cgroup);
08ce5f16	261	struct dev_whitelist_item *wh;
29486df3	262	char maj[MAJMINLEN], min[MAJMINLEN], acc[ACCLEN];
08ce5f16	263
4efd1a1b PE	264	rcu_read_lock();
4efd1a1b PE	265	list_for_each_entry_rcu(wh, &devcgroup->whitelist, list) {
08ce5f16	266	set_access(acc, wh->access);
29486df3 SH	267	set_majmin(maj, wh->major);
	268	set_majmin(min, wh->minor);
	269	seq_printf(m, "%c %s:%s %s\n", type_to_char(wh->type),
	270	maj, min, acc);
08ce5f16	271	}
4efd1a1b	272	rcu_read_unlock();
08ce5f16	273
29486df3	274	return 0;
08ce5f16 SH	275	}
	276
	277	/*
	278	* may_access_whitelist:
	279	* does the access granted to dev_cgroup c contain the access
	280	* requested in whitelist item refwh.
	281	* return 1 if yes, 0 if no.
b4046f00	282	* call with devcgroup_mutex held
08ce5f16 SH	283	*/
	284	static int may_access_whitelist(struct dev_cgroup *c,
	285	struct dev_whitelist_item *refwh)
	286	{
	287	struct dev_whitelist_item *whitem;
	288
	289	list_for_each_entry(whitem, &c->whitelist, list) {
	290	if (whitem->type & DEV_ALL)
	291	return 1;
	292	if ((refwh->type & DEV_BLOCK) && !(whitem->type & DEV_BLOCK))
	293	continue;
	294	if ((refwh->type & DEV_CHAR) && !(whitem->type & DEV_CHAR))
	295	continue;
	296	if (whitem->major != ~0 && whitem->major != refwh->major)
	297	continue;
	298	if (whitem->minor != ~0 && whitem->minor != refwh->minor)
	299	continue;
ec229e83	300	if (refwh->access & (~whitem->access))
08ce5f16 SH	301	continue;
	302	return 1;
	303	}
	304	return 0;
	305	}
	306
	307	/*
	308	* parent_has_perm:
	309	* when adding a new allow rule to a device whitelist, the rule
	310	* must be allowed in the parent device
	311	*/
f92523e3	312	static int parent_has_perm(struct dev_cgroup *childcg,
08ce5f16 SH	313	struct dev_whitelist_item *wh)
08ce5f16 SH	314	{
f92523e3	315	struct cgroup *pcg = childcg->css.cgroup->parent;
08ce5f16	316	struct dev_cgroup *parent;
08ce5f16 SH	317
	318	if (!pcg)
	319	return 1;
	320	parent = cgroup_to_devcgroup(pcg);
47c59803	321	return may_access_whitelist(parent, wh);
08ce5f16 SH	322	}
	323
	324	/*
	325	* Modify the whitelist using allow/deny rules.
	326	* CAP_SYS_ADMIN is needed for this. It's at least separate from CAP_MKNOD
	327	* so we can give a container CAP_MKNOD to let it create devices but not
	328	* modify the whitelist.
	329	* It seems likely we'll want to add a CAP_CONTAINER capability to allow
	330	* us to also grant CAP_SYS_ADMIN to containers without giving away the
	331	* device whitelist controls, but for now we'll stick with CAP_SYS_ADMIN
	332	*
	333	* Taking rules away is always allowed (given CAP_SYS_ADMIN). Granting
	334	* new access is only allowed if you're in the top-level cgroup, or your
	335	* parent cgroup has the access you're asking for.
	336	*/
f92523e3 PM	337	static int devcgroup_update_access(struct dev_cgroup *devcgroup,
f92523e3 PM	338	int filetype, const char *buffer)
08ce5f16	339	{
f92523e3	340	const char *b;
7759fc9d	341	char *endp;
c012a54a	342	int count;
08ce5f16 SH	343	struct dev_whitelist_item wh;
	344
	345	if (!capable(CAP_SYS_ADMIN))
	346	return -EPERM;
	347
08ce5f16 SH	348	memset(&wh, 0, sizeof(wh));
	349	b = buffer;
	350
	351	switch (*b) {
	352	case 'a':
	353	wh.type = DEV_ALL;
	354	wh.access = ACC_MASK;
d823f6bf LZ	355	wh.major = ~0;
d823f6bf LZ	356	wh.minor = ~0;
08ce5f16 SH	357	goto handle;
	358	case 'b':
	359	wh.type = DEV_BLOCK;
	360	break;
	361	case 'c':
	362	wh.type = DEV_CHAR;
	363	break;
	364	default:
f92523e3	365	return -EINVAL;
08ce5f16 SH	366	}
08ce5f16 SH	367	b++;
f92523e3 PM	368	if (!isspace(*b))
f92523e3 PM	369	return -EINVAL;
08ce5f16 SH	370	b++;
	371	if (b == '') {
	372	wh.major = ~0;
	373	b++;
	374	} else if (isdigit(*b)) {
7759fc9d LZ	375	wh.major = simple_strtoul(b, &endp, 10);
7759fc9d LZ	376	b = endp;
08ce5f16	377	} else {
f92523e3	378	return -EINVAL;
08ce5f16	379	}
f92523e3 PM	380	if (*b != ':')
f92523e3 PM	381	return -EINVAL;
08ce5f16 SH	382	b++;
	383
	384	/* read minor */
	385	if (b == '') {
	386	wh.minor = ~0;
	387	b++;
	388	} else if (isdigit(*b)) {
7759fc9d LZ	389	wh.minor = simple_strtoul(b, &endp, 10);
7759fc9d LZ	390	b = endp;
08ce5f16	391	} else {
f92523e3	392	return -EINVAL;
08ce5f16	393	}
f92523e3 PM	394	if (!isspace(*b))
f92523e3 PM	395	return -EINVAL;
08ce5f16 SH	396	for (b++, count = 0; count < 3; count++, b++) {
	397	switch (*b) {
	398	case 'r':
	399	wh.access \|= ACC_READ;
	400	break;
	401	case 'w':
	402	wh.access \|= ACC_WRITE;
	403	break;
	404	case 'm':
	405	wh.access \|= ACC_MKNOD;
	406	break;
	407	case '\n':
	408	case '\0':
	409	count = 3;
	410	break;
	411	default:
f92523e3	412	return -EINVAL;
08ce5f16 SH	413	}
	414	}
	415
	416	handle:
08ce5f16 SH	417	switch (filetype) {
08ce5f16 SH	418	case DEVCG_ALLOW:
f92523e3 PM	419	if (!parent_has_perm(devcgroup, &wh))
	420	return -EPERM;
	421	return dev_whitelist_add(devcgroup, &wh);
08ce5f16 SH	422	case DEVCG_DENY:
	423	dev_whitelist_rm(devcgroup, &wh);
	424	break;
	425	default:
f92523e3	426	return -EINVAL;
08ce5f16	427	}
f92523e3 PM	428	return 0;
f92523e3 PM	429	}
08ce5f16	430
f92523e3 PM	431	static int devcgroup_access_write(struct cgroup cgrp, struct cftype cft,
	432	const char *buffer)
	433	{
	434	int retval;
b4046f00 LZ	435
b4046f00 LZ	436	mutex_lock(&devcgroup_mutex);
f92523e3 PM	437	retval = devcgroup_update_access(cgroup_to_devcgroup(cgrp),
f92523e3 PM	438	cft->private, buffer);
b4046f00	439	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	440	return retval;
	441	}
	442
	443	static struct cftype dev_cgroup_files[] = {
	444	{
	445	.name = "allow",
f92523e3	446	.write_string = devcgroup_access_write,
08ce5f16 SH	447	.private = DEVCG_ALLOW,
	448	},
	449	{
	450	.name = "deny",
f92523e3	451	.write_string = devcgroup_access_write,
08ce5f16 SH	452	.private = DEVCG_DENY,
08ce5f16 SH	453	},
29486df3 SH	454	{
	455	.name = "list",
	456	.read_seq_string = devcgroup_seq_read,
	457	.private = DEVCG_LIST,
	458	},
08ce5f16 SH	459	};
	460
	461	static int devcgroup_populate(struct cgroup_subsys *ss,
	462	struct cgroup *cgroup)
	463	{
	464	return cgroup_add_files(cgroup, ss, dev_cgroup_files,
	465	ARRAY_SIZE(dev_cgroup_files));
	466	}
	467
	468	struct cgroup_subsys devices_subsys = {
	469	.name = "devices",
	470	.can_attach = devcgroup_can_attach,
	471	.create = devcgroup_create,
	472	.destroy = devcgroup_destroy,
	473	.populate = devcgroup_populate,
	474	.subsys_id = devices_subsys_id,
	475	};
	476
	477	int devcgroup_inode_permission(struct inode *inode, int mask)
	478	{
08ce5f16 SH	479	struct dev_cgroup *dev_cgroup;
	480	struct dev_whitelist_item *wh;
	481
	482	dev_t device = inode->i_rdev;
	483	if (!device)
	484	return 0;
	485	if (!S_ISBLK(inode->i_mode) && !S_ISCHR(inode->i_mode))
	486	return 0;
08ce5f16	487
4efd1a1b	488	rcu_read_lock();
36fd71d2 LZ	489
	490	dev_cgroup = task_devcgroup(current);
	491
4efd1a1b	492	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	493	if (wh->type & DEV_ALL)
cd500819	494	goto found;
08ce5f16 SH	495	if ((wh->type & DEV_BLOCK) && !S_ISBLK(inode->i_mode))
	496	continue;
	497	if ((wh->type & DEV_CHAR) && !S_ISCHR(inode->i_mode))
	498	continue;
	499	if (wh->major != ~0 && wh->major != imajor(inode))
	500	continue;
	501	if (wh->minor != ~0 && wh->minor != iminor(inode))
	502	continue;
cd500819	503
08ce5f16 SH	504	if ((mask & MAY_WRITE) && !(wh->access & ACC_WRITE))
	505	continue;
	506	if ((mask & MAY_READ) && !(wh->access & ACC_READ))
	507	continue;
cd500819	508	found:
4efd1a1b	509	rcu_read_unlock();
08ce5f16 SH	510	return 0;
08ce5f16 SH	511	}
36fd71d2	512
4efd1a1b	513	rcu_read_unlock();
08ce5f16 SH	514
	515	return -EPERM;
	516	}
	517
	518	int devcgroup_inode_mknod(int mode, dev_t dev)
	519	{
08ce5f16 SH	520	struct dev_cgroup *dev_cgroup;
	521	struct dev_whitelist_item *wh;
	522
0b82ac37 SH	523	if (!S_ISBLK(mode) && !S_ISCHR(mode))
	524	return 0;
	525
4efd1a1b	526	rcu_read_lock();
36fd71d2 LZ	527
	528	dev_cgroup = task_devcgroup(current);
	529
116e0575	530	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	531	if (wh->type & DEV_ALL)
cd500819	532	goto found;
08ce5f16 SH	533	if ((wh->type & DEV_BLOCK) && !S_ISBLK(mode))
	534	continue;
	535	if ((wh->type & DEV_CHAR) && !S_ISCHR(mode))
	536	continue;
	537	if (wh->major != ~0 && wh->major != MAJOR(dev))
	538	continue;
	539	if (wh->minor != ~0 && wh->minor != MINOR(dev))
	540	continue;
cd500819	541
08ce5f16 SH	542	if (!(wh->access & ACC_MKNOD))
08ce5f16 SH	543	continue;
cd500819	544	found:
4efd1a1b	545	rcu_read_unlock();
08ce5f16 SH	546	return 0;
08ce5f16 SH	547	}
36fd71d2	548
4efd1a1b	549	rcu_read_unlock();
36fd71d2	550
08ce5f16 SH	551	return -EPERM;
08ce5f16 SH	552	}