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