[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/slab.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>
5a0e3ad6	13	#include <linux/slab.h>
47c59803	14	#include <linux/rcupdate.h>
b4046f00	15	#include <linux/mutex.h>
08ce5f16 SH	16
	17	#define ACC_MKNOD 1
	18	#define ACC_READ 2
	19	#define ACC_WRITE 4
	20	#define ACC_MASK (ACC_MKNOD \| ACC_READ \| ACC_WRITE)
	21
	22	#define DEV_BLOCK 1
	23	#define DEV_CHAR 2
	24	#define DEV_ALL 4 /* this represents all devices */
	25
b4046f00 LZ	26	static DEFINE_MUTEX(devcgroup_mutex);
b4046f00 LZ	27
08ce5f16 SH	28	/*
08ce5f16 SH	29	* whitelist locking rules:
b4046f00	30	* hold devcgroup_mutex for update/read.
47c59803	31	* hold rcu_read_lock() for read.
08ce5f16 SH	32	*/
	33
	34	struct dev_whitelist_item {
	35	u32 major, minor;
	36	short type;
	37	short access;
	38	struct list_head list;
4efd1a1b	39	struct rcu_head rcu;
08ce5f16 SH	40	};
	41
	42	struct dev_cgroup {
	43	struct cgroup_subsys_state css;
	44	struct list_head whitelist;
08ce5f16 SH	45	};
08ce5f16 SH	46
b66862f7 PE	47	static inline struct dev_cgroup css_to_devcgroup(struct cgroup_subsys_state s)
	48	{
	49	return container_of(s, struct dev_cgroup, css);
	50	}
	51
08ce5f16 SH	52	static inline struct dev_cgroup cgroup_to_devcgroup(struct cgroup cgroup)
08ce5f16 SH	53	{
b66862f7	54	return css_to_devcgroup(cgroup_subsys_state(cgroup, devices_subsys_id));
08ce5f16 SH	55	}
08ce5f16 SH	56
f92523e3 PM	57	static inline struct dev_cgroup task_devcgroup(struct task_struct task)
	58	{
	59	return css_to_devcgroup(task_subsys_state(task, devices_subsys_id));
	60	}
	61
08ce5f16 SH	62	struct cgroup_subsys devices_subsys;
	63
	64	static int devcgroup_can_attach(struct cgroup_subsys *ss,
be367d09 BB	65	struct cgroup new_cgroup, struct task_struct task,
be367d09 BB	66	bool threadgroup)
08ce5f16 SH	67	{
	68	if (current != task && !capable(CAP_SYS_ADMIN))
	69	return -EPERM;
	70
	71	return 0;
	72	}
	73
	74	/*
b4046f00	75	* called under devcgroup_mutex
08ce5f16 SH	76	*/
	77	static int dev_whitelist_copy(struct list_head dest, struct list_head orig)
	78	{
	79	struct dev_whitelist_item wh, tmp, *new;
	80
	81	list_for_each_entry(wh, orig, list) {
2cdc7241	82	new = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
08ce5f16 SH	83	if (!new)
08ce5f16 SH	84	goto free_and_exit;
08ce5f16 SH	85	list_add_tail(&new->list, dest);
	86	}
	87
	88	return 0;
	89
	90	free_and_exit:
	91	list_for_each_entry_safe(wh, tmp, dest, list) {
	92	list_del(&wh->list);
	93	kfree(wh);
	94	}
	95	return -ENOMEM;
	96	}
	97
	98	/* Stupid prototype - don't bother combining existing entries */
	99	/*
b4046f00	100	* called under devcgroup_mutex
08ce5f16 SH	101	*/
	102	static int dev_whitelist_add(struct dev_cgroup *dev_cgroup,
	103	struct dev_whitelist_item *wh)
	104	{
d1ee2971	105	struct dev_whitelist_item whcopy, walk;
08ce5f16	106
2cdc7241	107	whcopy = kmemdup(wh, sizeof(*wh), GFP_KERNEL);
08ce5f16 SH	108	if (!whcopy)
	109	return -ENOMEM;
	110
d1ee2971 PE	111	list_for_each_entry(walk, &dev_cgroup->whitelist, list) {
	112	if (walk->type != wh->type)
	113	continue;
	114	if (walk->major != wh->major)
	115	continue;
	116	if (walk->minor != wh->minor)
	117	continue;
	118
	119	walk->access \|= wh->access;
	120	kfree(whcopy);
	121	whcopy = NULL;
	122	}
	123
	124	if (whcopy != NULL)
4efd1a1b	125	list_add_tail_rcu(&whcopy->list, &dev_cgroup->whitelist);
08ce5f16 SH	126	return 0;
	127	}
	128
4efd1a1b PE	129	static void whitelist_item_free(struct rcu_head *rcu)
	130	{
	131	struct dev_whitelist_item *item;
	132
	133	item = container_of(rcu, struct dev_whitelist_item, rcu);
	134	kfree(item);
	135	}
	136
08ce5f16	137	/*
b4046f00	138	* called under devcgroup_mutex
08ce5f16 SH	139	*/
	140	static void dev_whitelist_rm(struct dev_cgroup *dev_cgroup,
	141	struct dev_whitelist_item *wh)
	142	{
	143	struct dev_whitelist_item walk, tmp;
	144
08ce5f16 SH	145	list_for_each_entry_safe(walk, tmp, &dev_cgroup->whitelist, list) {
	146	if (walk->type == DEV_ALL)
	147	goto remove;
	148	if (walk->type != wh->type)
	149	continue;
	150	if (walk->major != ~0 && walk->major != wh->major)
	151	continue;
	152	if (walk->minor != ~0 && walk->minor != wh->minor)
	153	continue;
	154
	155	remove:
	156	walk->access &= ~wh->access;
	157	if (!walk->access) {
4efd1a1b PE	158	list_del_rcu(&walk->list);
4efd1a1b PE	159	call_rcu(&walk->rcu, whitelist_item_free);
08ce5f16 SH	160	}
08ce5f16 SH	161	}
08ce5f16 SH	162	}
	163
	164	/*
	165	* called from kernel/cgroup.c with cgroup_lock() held.
	166	*/
	167	static struct cgroup_subsys_state devcgroup_create(struct cgroup_subsys ss,
	168	struct cgroup *cgroup)
	169	{
	170	struct dev_cgroup dev_cgroup, parent_dev_cgroup;
	171	struct cgroup *parent_cgroup;
	172	int ret;
	173
	174	dev_cgroup = kzalloc(sizeof(*dev_cgroup), GFP_KERNEL);
	175	if (!dev_cgroup)
	176	return ERR_PTR(-ENOMEM);
	177	INIT_LIST_HEAD(&dev_cgroup->whitelist);
	178	parent_cgroup = cgroup->parent;
	179
	180	if (parent_cgroup == NULL) {
	181	struct dev_whitelist_item *wh;
	182	wh = kmalloc(sizeof(*wh), GFP_KERNEL);
	183	if (!wh) {
	184	kfree(dev_cgroup);
	185	return ERR_PTR(-ENOMEM);
	186	}
	187	wh->minor = wh->major = ~0;
	188	wh->type = DEV_ALL;
7759fc9d	189	wh->access = ACC_MASK;
08ce5f16 SH	190	list_add(&wh->list, &dev_cgroup->whitelist);
	191	} else {
	192	parent_dev_cgroup = cgroup_to_devcgroup(parent_cgroup);
b4046f00	193	mutex_lock(&devcgroup_mutex);
08ce5f16 SH	194	ret = dev_whitelist_copy(&dev_cgroup->whitelist,
08ce5f16 SH	195	&parent_dev_cgroup->whitelist);
b4046f00	196	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	197	if (ret) {
	198	kfree(dev_cgroup);
	199	return ERR_PTR(ret);
	200	}
	201	}
	202
08ce5f16 SH	203	return &dev_cgroup->css;
	204	}
	205
	206	static void devcgroup_destroy(struct cgroup_subsys *ss,
	207	struct cgroup *cgroup)
	208	{
	209	struct dev_cgroup *dev_cgroup;
	210	struct dev_whitelist_item wh, tmp;
	211
	212	dev_cgroup = cgroup_to_devcgroup(cgroup);
	213	list_for_each_entry_safe(wh, tmp, &dev_cgroup->whitelist, list) {
	214	list_del(&wh->list);
	215	kfree(wh);
	216	}
	217	kfree(dev_cgroup);
	218	}
	219
	220	#define DEVCG_ALLOW 1
	221	#define DEVCG_DENY 2
29486df3 SH	222	#define DEVCG_LIST 3
29486df3 SH	223
17d213f8	224	#define MAJMINLEN 13
29486df3	225	#define ACCLEN 4
08ce5f16 SH	226
	227	static void set_access(char *acc, short access)
	228	{
	229	int idx = 0;
29486df3	230	memset(acc, 0, ACCLEN);
08ce5f16 SH	231	if (access & ACC_READ)
	232	acc[idx++] = 'r';
	233	if (access & ACC_WRITE)
	234	acc[idx++] = 'w';
	235	if (access & ACC_MKNOD)
	236	acc[idx++] = 'm';
	237	}
	238
	239	static char type_to_char(short type)
	240	{
	241	if (type == DEV_ALL)
	242	return 'a';
	243	if (type == DEV_CHAR)
	244	return 'c';
	245	if (type == DEV_BLOCK)
	246	return 'b';
	247	return 'X';
	248	}
	249
29486df3	250	static void set_majmin(char *str, unsigned m)
08ce5f16	251	{
08ce5f16	252	if (m == ~0)
7759fc9d	253	strcpy(str, "*");
08ce5f16	254	else
7759fc9d	255	sprintf(str, "%u", m);
08ce5f16 SH	256	}
08ce5f16 SH	257
29486df3 SH	258	static int devcgroup_seq_read(struct cgroup cgroup, struct cftype cft,
29486df3 SH	259	struct seq_file *m)
08ce5f16	260	{
29486df3	261	struct dev_cgroup *devcgroup = cgroup_to_devcgroup(cgroup);
08ce5f16	262	struct dev_whitelist_item *wh;
29486df3	263	char maj[MAJMINLEN], min[MAJMINLEN], acc[ACCLEN];
08ce5f16	264
4efd1a1b PE	265	rcu_read_lock();
4efd1a1b PE	266	list_for_each_entry_rcu(wh, &devcgroup->whitelist, list) {
08ce5f16	267	set_access(acc, wh->access);
29486df3 SH	268	set_majmin(maj, wh->major);
	269	set_majmin(min, wh->minor);
	270	seq_printf(m, "%c %s:%s %s\n", type_to_char(wh->type),
	271	maj, min, acc);
08ce5f16	272	}
4efd1a1b	273	rcu_read_unlock();
08ce5f16	274
29486df3	275	return 0;
08ce5f16 SH	276	}
	277
	278	/*
	279	* may_access_whitelist:
	280	* does the access granted to dev_cgroup c contain the access
	281	* requested in whitelist item refwh.
	282	* return 1 if yes, 0 if no.
b4046f00	283	* call with devcgroup_mutex held
08ce5f16 SH	284	*/
	285	static int may_access_whitelist(struct dev_cgroup *c,
	286	struct dev_whitelist_item *refwh)
	287	{
	288	struct dev_whitelist_item *whitem;
	289
	290	list_for_each_entry(whitem, &c->whitelist, list) {
	291	if (whitem->type & DEV_ALL)
	292	return 1;
	293	if ((refwh->type & DEV_BLOCK) && !(whitem->type & DEV_BLOCK))
	294	continue;
	295	if ((refwh->type & DEV_CHAR) && !(whitem->type & DEV_CHAR))
	296	continue;
	297	if (whitem->major != ~0 && whitem->major != refwh->major)
	298	continue;
	299	if (whitem->minor != ~0 && whitem->minor != refwh->minor)
	300	continue;
ec229e83	301	if (refwh->access & (~whitem->access))
08ce5f16 SH	302	continue;
	303	return 1;
	304	}
	305	return 0;
	306	}
	307
	308	/*
	309	* parent_has_perm:
	310	* when adding a new allow rule to a device whitelist, the rule
	311	* must be allowed in the parent device
	312	*/
f92523e3	313	static int parent_has_perm(struct dev_cgroup *childcg,
08ce5f16 SH	314	struct dev_whitelist_item *wh)
08ce5f16 SH	315	{
f92523e3	316	struct cgroup *pcg = childcg->css.cgroup->parent;
08ce5f16	317	struct dev_cgroup *parent;
08ce5f16 SH	318
	319	if (!pcg)
	320	return 1;
	321	parent = cgroup_to_devcgroup(pcg);
47c59803	322	return may_access_whitelist(parent, wh);
08ce5f16 SH	323	}
	324
	325	/*
	326	* Modify the whitelist using allow/deny rules.
	327	* CAP_SYS_ADMIN is needed for this. It's at least separate from CAP_MKNOD
	328	* so we can give a container CAP_MKNOD to let it create devices but not
	329	* modify the whitelist.
	330	* It seems likely we'll want to add a CAP_CONTAINER capability to allow
	331	* us to also grant CAP_SYS_ADMIN to containers without giving away the
	332	* device whitelist controls, but for now we'll stick with CAP_SYS_ADMIN
	333	*
	334	* Taking rules away is always allowed (given CAP_SYS_ADMIN). Granting
	335	* new access is only allowed if you're in the top-level cgroup, or your
	336	* parent cgroup has the access you're asking for.
	337	*/
f92523e3 PM	338	static int devcgroup_update_access(struct dev_cgroup *devcgroup,
f92523e3 PM	339	int filetype, const char *buffer)
08ce5f16	340	{
f92523e3	341	const char *b;
7759fc9d	342	char *endp;
c012a54a	343	int count;
08ce5f16 SH	344	struct dev_whitelist_item wh;
	345
	346	if (!capable(CAP_SYS_ADMIN))
	347	return -EPERM;
	348
08ce5f16 SH	349	memset(&wh, 0, sizeof(wh));
	350	b = buffer;
	351
	352	switch (*b) {
	353	case 'a':
	354	wh.type = DEV_ALL;
	355	wh.access = ACC_MASK;
d823f6bf LZ	356	wh.major = ~0;
d823f6bf LZ	357	wh.minor = ~0;
08ce5f16 SH	358	goto handle;
	359	case 'b':
	360	wh.type = DEV_BLOCK;
	361	break;
	362	case 'c':
	363	wh.type = DEV_CHAR;
	364	break;
	365	default:
f92523e3	366	return -EINVAL;
08ce5f16 SH	367	}
08ce5f16 SH	368	b++;
f92523e3 PM	369	if (!isspace(*b))
f92523e3 PM	370	return -EINVAL;
08ce5f16 SH	371	b++;
	372	if (b == '') {
	373	wh.major = ~0;
	374	b++;
	375	} else if (isdigit(*b)) {
7759fc9d LZ	376	wh.major = simple_strtoul(b, &endp, 10);
7759fc9d LZ	377	b = endp;
08ce5f16	378	} else {
f92523e3	379	return -EINVAL;
08ce5f16	380	}
f92523e3 PM	381	if (*b != ':')
f92523e3 PM	382	return -EINVAL;
08ce5f16 SH	383	b++;
	384
	385	/* read minor */
	386	if (b == '') {
	387	wh.minor = ~0;
	388	b++;
	389	} else if (isdigit(*b)) {
7759fc9d LZ	390	wh.minor = simple_strtoul(b, &endp, 10);
7759fc9d LZ	391	b = endp;
08ce5f16	392	} else {
f92523e3	393	return -EINVAL;
08ce5f16	394	}
f92523e3 PM	395	if (!isspace(*b))
f92523e3 PM	396	return -EINVAL;
08ce5f16 SH	397	for (b++, count = 0; count < 3; count++, b++) {
	398	switch (*b) {
	399	case 'r':
	400	wh.access \|= ACC_READ;
	401	break;
	402	case 'w':
	403	wh.access \|= ACC_WRITE;
	404	break;
	405	case 'm':
	406	wh.access \|= ACC_MKNOD;
	407	break;
	408	case '\n':
	409	case '\0':
	410	count = 3;
	411	break;
	412	default:
f92523e3	413	return -EINVAL;
08ce5f16 SH	414	}
	415	}
	416
	417	handle:
08ce5f16 SH	418	switch (filetype) {
08ce5f16 SH	419	case DEVCG_ALLOW:
f92523e3 PM	420	if (!parent_has_perm(devcgroup, &wh))
	421	return -EPERM;
	422	return dev_whitelist_add(devcgroup, &wh);
08ce5f16 SH	423	case DEVCG_DENY:
	424	dev_whitelist_rm(devcgroup, &wh);
	425	break;
	426	default:
f92523e3	427	return -EINVAL;
08ce5f16	428	}
f92523e3 PM	429	return 0;
f92523e3 PM	430	}
08ce5f16	431
f92523e3 PM	432	static int devcgroup_access_write(struct cgroup cgrp, struct cftype cft,
	433	const char *buffer)
	434	{
	435	int retval;
b4046f00 LZ	436
b4046f00 LZ	437	mutex_lock(&devcgroup_mutex);
f92523e3 PM	438	retval = devcgroup_update_access(cgroup_to_devcgroup(cgrp),
f92523e3 PM	439	cft->private, buffer);
b4046f00	440	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	441	return retval;
	442	}
	443
	444	static struct cftype dev_cgroup_files[] = {
	445	{
	446	.name = "allow",
f92523e3	447	.write_string = devcgroup_access_write,
08ce5f16 SH	448	.private = DEVCG_ALLOW,
	449	},
	450	{
	451	.name = "deny",
f92523e3	452	.write_string = devcgroup_access_write,
08ce5f16 SH	453	.private = DEVCG_DENY,
08ce5f16 SH	454	},
29486df3 SH	455	{
	456	.name = "list",
	457	.read_seq_string = devcgroup_seq_read,
	458	.private = DEVCG_LIST,
	459	},
08ce5f16 SH	460	};
	461
	462	static int devcgroup_populate(struct cgroup_subsys *ss,
	463	struct cgroup *cgroup)
	464	{
	465	return cgroup_add_files(cgroup, ss, dev_cgroup_files,
	466	ARRAY_SIZE(dev_cgroup_files));
	467	}
	468
	469	struct cgroup_subsys devices_subsys = {
	470	.name = "devices",
	471	.can_attach = devcgroup_can_attach,
	472	.create = devcgroup_create,
c5b60b5e	473	.destroy = devcgroup_destroy,
08ce5f16 SH	474	.populate = devcgroup_populate,
	475	.subsys_id = devices_subsys_id,
	476	};
	477
	478	int devcgroup_inode_permission(struct inode *inode, int mask)
	479	{
08ce5f16 SH	480	struct dev_cgroup *dev_cgroup;
	481	struct dev_whitelist_item *wh;
	482
	483	dev_t device = inode->i_rdev;
	484	if (!device)
	485	return 0;
	486	if (!S_ISBLK(inode->i_mode) && !S_ISCHR(inode->i_mode))
	487	return 0;
08ce5f16	488
4efd1a1b	489	rcu_read_lock();
36fd71d2 LZ	490
	491	dev_cgroup = task_devcgroup(current);
	492
4efd1a1b	493	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	494	if (wh->type & DEV_ALL)
cd500819	495	goto found;
08ce5f16 SH	496	if ((wh->type & DEV_BLOCK) && !S_ISBLK(inode->i_mode))
	497	continue;
	498	if ((wh->type & DEV_CHAR) && !S_ISCHR(inode->i_mode))
	499	continue;
	500	if (wh->major != ~0 && wh->major != imajor(inode))
	501	continue;
	502	if (wh->minor != ~0 && wh->minor != iminor(inode))
	503	continue;
cd500819	504
08ce5f16 SH	505	if ((mask & MAY_WRITE) && !(wh->access & ACC_WRITE))
	506	continue;
	507	if ((mask & MAY_READ) && !(wh->access & ACC_READ))
	508	continue;
cd500819	509	found:
4efd1a1b	510	rcu_read_unlock();
08ce5f16 SH	511	return 0;
08ce5f16 SH	512	}
36fd71d2	513
4efd1a1b	514	rcu_read_unlock();
08ce5f16 SH	515
	516	return -EPERM;
	517	}
	518
	519	int devcgroup_inode_mknod(int mode, dev_t dev)
	520	{
08ce5f16 SH	521	struct dev_cgroup *dev_cgroup;
	522	struct dev_whitelist_item *wh;
	523
0b82ac37 SH	524	if (!S_ISBLK(mode) && !S_ISCHR(mode))
	525	return 0;
	526
4efd1a1b	527	rcu_read_lock();
36fd71d2 LZ	528
	529	dev_cgroup = task_devcgroup(current);
	530
116e0575	531	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	532	if (wh->type & DEV_ALL)
cd500819	533	goto found;
08ce5f16 SH	534	if ((wh->type & DEV_BLOCK) && !S_ISBLK(mode))
	535	continue;
	536	if ((wh->type & DEV_CHAR) && !S_ISCHR(mode))
	537	continue;
	538	if (wh->major != ~0 && wh->major != MAJOR(dev))
	539	continue;
	540	if (wh->minor != ~0 && wh->minor != MINOR(dev))
	541	continue;
cd500819	542
08ce5f16 SH	543	if (!(wh->access & ACC_MKNOD))
08ce5f16 SH	544	continue;
cd500819	545	found:
4efd1a1b	546	rcu_read_unlock();
08ce5f16 SH	547	return 0;
08ce5f16 SH	548	}
36fd71d2	549
4efd1a1b	550	rcu_read_unlock();
36fd71d2	551
08ce5f16 SH	552	return -EPERM;
08ce5f16 SH	553	}