[linux-2.6-block.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)
{
	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;
}

/*
 * 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);
			kfree_rcu(walk, rcu);
		}
	}
}

/*
 * 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;

	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,
f780bdb7	65	struct cgroup new_cgroup, struct task_struct task)
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
	128	/*
b4046f00	129	* called under devcgroup_mutex
08ce5f16 SH	130	*/
	131	static void dev_whitelist_rm(struct dev_cgroup *dev_cgroup,
	132	struct dev_whitelist_item *wh)
	133	{
	134	struct dev_whitelist_item walk, tmp;
	135
08ce5f16 SH	136	list_for_each_entry_safe(walk, tmp, &dev_cgroup->whitelist, list) {
	137	if (walk->type == DEV_ALL)
	138	goto remove;
	139	if (walk->type != wh->type)
	140	continue;
	141	if (walk->major != ~0 && walk->major != wh->major)
	142	continue;
	143	if (walk->minor != ~0 && walk->minor != wh->minor)
	144	continue;
	145
	146	remove:
	147	walk->access &= ~wh->access;
	148	if (!walk->access) {
4efd1a1b	149	list_del_rcu(&walk->list);
6034f7e6	150	kfree_rcu(walk, rcu);
08ce5f16 SH	151	}
08ce5f16 SH	152	}
08ce5f16 SH	153	}
	154
	155	/*
	156	* called from kernel/cgroup.c with cgroup_lock() held.
	157	*/
	158	static struct cgroup_subsys_state devcgroup_create(struct cgroup_subsys ss,
	159	struct cgroup *cgroup)
	160	{
	161	struct dev_cgroup dev_cgroup, parent_dev_cgroup;
	162	struct cgroup *parent_cgroup;
	163	int ret;
	164
	165	dev_cgroup = kzalloc(sizeof(*dev_cgroup), GFP_KERNEL);
	166	if (!dev_cgroup)
	167	return ERR_PTR(-ENOMEM);
	168	INIT_LIST_HEAD(&dev_cgroup->whitelist);
	169	parent_cgroup = cgroup->parent;
	170
	171	if (parent_cgroup == NULL) {
	172	struct dev_whitelist_item *wh;
	173	wh = kmalloc(sizeof(*wh), GFP_KERNEL);
	174	if (!wh) {
	175	kfree(dev_cgroup);
	176	return ERR_PTR(-ENOMEM);
	177	}
	178	wh->minor = wh->major = ~0;
	179	wh->type = DEV_ALL;
7759fc9d	180	wh->access = ACC_MASK;
08ce5f16 SH	181	list_add(&wh->list, &dev_cgroup->whitelist);
	182	} else {
	183	parent_dev_cgroup = cgroup_to_devcgroup(parent_cgroup);
b4046f00	184	mutex_lock(&devcgroup_mutex);
08ce5f16 SH	185	ret = dev_whitelist_copy(&dev_cgroup->whitelist,
08ce5f16 SH	186	&parent_dev_cgroup->whitelist);
b4046f00	187	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	188	if (ret) {
	189	kfree(dev_cgroup);
	190	return ERR_PTR(ret);
	191	}
	192	}
	193
08ce5f16 SH	194	return &dev_cgroup->css;
	195	}
	196
	197	static void devcgroup_destroy(struct cgroup_subsys *ss,
	198	struct cgroup *cgroup)
	199	{
	200	struct dev_cgroup *dev_cgroup;
	201	struct dev_whitelist_item wh, tmp;
	202
	203	dev_cgroup = cgroup_to_devcgroup(cgroup);
	204	list_for_each_entry_safe(wh, tmp, &dev_cgroup->whitelist, list) {
	205	list_del(&wh->list);
	206	kfree(wh);
	207	}
	208	kfree(dev_cgroup);
	209	}
	210
	211	#define DEVCG_ALLOW 1
	212	#define DEVCG_DENY 2
29486df3 SH	213	#define DEVCG_LIST 3
29486df3 SH	214
17d213f8	215	#define MAJMINLEN 13
29486df3	216	#define ACCLEN 4
08ce5f16 SH	217
	218	static void set_access(char *acc, short access)
	219	{
	220	int idx = 0;
29486df3	221	memset(acc, 0, ACCLEN);
08ce5f16 SH	222	if (access & ACC_READ)
	223	acc[idx++] = 'r';
	224	if (access & ACC_WRITE)
	225	acc[idx++] = 'w';
	226	if (access & ACC_MKNOD)
	227	acc[idx++] = 'm';
	228	}
	229
	230	static char type_to_char(short type)
	231	{
	232	if (type == DEV_ALL)
	233	return 'a';
	234	if (type == DEV_CHAR)
	235	return 'c';
	236	if (type == DEV_BLOCK)
	237	return 'b';
	238	return 'X';
	239	}
	240
29486df3	241	static void set_majmin(char *str, unsigned m)
08ce5f16	242	{
08ce5f16	243	if (m == ~0)
7759fc9d	244	strcpy(str, "*");
08ce5f16	245	else
7759fc9d	246	sprintf(str, "%u", m);
08ce5f16 SH	247	}
08ce5f16 SH	248
29486df3 SH	249	static int devcgroup_seq_read(struct cgroup cgroup, struct cftype cft,
29486df3 SH	250	struct seq_file *m)
08ce5f16	251	{
29486df3	252	struct dev_cgroup *devcgroup = cgroup_to_devcgroup(cgroup);
08ce5f16	253	struct dev_whitelist_item *wh;
29486df3	254	char maj[MAJMINLEN], min[MAJMINLEN], acc[ACCLEN];
08ce5f16	255
4efd1a1b PE	256	rcu_read_lock();
4efd1a1b PE	257	list_for_each_entry_rcu(wh, &devcgroup->whitelist, list) {
08ce5f16	258	set_access(acc, wh->access);
29486df3 SH	259	set_majmin(maj, wh->major);
	260	set_majmin(min, wh->minor);
	261	seq_printf(m, "%c %s:%s %s\n", type_to_char(wh->type),
	262	maj, min, acc);
08ce5f16	263	}
4efd1a1b	264	rcu_read_unlock();
08ce5f16	265
29486df3	266	return 0;
08ce5f16 SH	267	}
	268
	269	/*
	270	* may_access_whitelist:
	271	* does the access granted to dev_cgroup c contain the access
	272	* requested in whitelist item refwh.
	273	* return 1 if yes, 0 if no.
b4046f00	274	* call with devcgroup_mutex held
08ce5f16 SH	275	*/
	276	static int may_access_whitelist(struct dev_cgroup *c,
	277	struct dev_whitelist_item *refwh)
	278	{
	279	struct dev_whitelist_item *whitem;
	280
	281	list_for_each_entry(whitem, &c->whitelist, list) {
	282	if (whitem->type & DEV_ALL)
	283	return 1;
	284	if ((refwh->type & DEV_BLOCK) && !(whitem->type & DEV_BLOCK))
	285	continue;
	286	if ((refwh->type & DEV_CHAR) && !(whitem->type & DEV_CHAR))
	287	continue;
	288	if (whitem->major != ~0 && whitem->major != refwh->major)
	289	continue;
	290	if (whitem->minor != ~0 && whitem->minor != refwh->minor)
	291	continue;
ec229e83	292	if (refwh->access & (~whitem->access))
08ce5f16 SH	293	continue;
	294	return 1;
	295	}
	296	return 0;
	297	}
	298
	299	/*
	300	* parent_has_perm:
	301	* when adding a new allow rule to a device whitelist, the rule
	302	* must be allowed in the parent device
	303	*/
f92523e3	304	static int parent_has_perm(struct dev_cgroup *childcg,
08ce5f16 SH	305	struct dev_whitelist_item *wh)
08ce5f16 SH	306	{
f92523e3	307	struct cgroup *pcg = childcg->css.cgroup->parent;
08ce5f16	308	struct dev_cgroup *parent;
08ce5f16 SH	309
	310	if (!pcg)
	311	return 1;
	312	parent = cgroup_to_devcgroup(pcg);
47c59803	313	return may_access_whitelist(parent, wh);
08ce5f16 SH	314	}
	315
	316	/*
	317	* Modify the whitelist using allow/deny rules.
	318	* CAP_SYS_ADMIN is needed for this. It's at least separate from CAP_MKNOD
	319	* so we can give a container CAP_MKNOD to let it create devices but not
	320	* modify the whitelist.
	321	* It seems likely we'll want to add a CAP_CONTAINER capability to allow
	322	* us to also grant CAP_SYS_ADMIN to containers without giving away the
	323	* device whitelist controls, but for now we'll stick with CAP_SYS_ADMIN
	324	*
	325	* Taking rules away is always allowed (given CAP_SYS_ADMIN). Granting
	326	* new access is only allowed if you're in the top-level cgroup, or your
	327	* parent cgroup has the access you're asking for.
	328	*/
f92523e3 PM	329	static int devcgroup_update_access(struct dev_cgroup *devcgroup,
f92523e3 PM	330	int filetype, const char *buffer)
08ce5f16	331	{
f92523e3	332	const char *b;
7759fc9d	333	char *endp;
c012a54a	334	int count;
08ce5f16 SH	335	struct dev_whitelist_item wh;
	336
	337	if (!capable(CAP_SYS_ADMIN))
	338	return -EPERM;
	339
08ce5f16 SH	340	memset(&wh, 0, sizeof(wh));
	341	b = buffer;
	342
	343	switch (*b) {
	344	case 'a':
	345	wh.type = DEV_ALL;
	346	wh.access = ACC_MASK;
d823f6bf LZ	347	wh.major = ~0;
d823f6bf LZ	348	wh.minor = ~0;
08ce5f16 SH	349	goto handle;
	350	case 'b':
	351	wh.type = DEV_BLOCK;
	352	break;
	353	case 'c':
	354	wh.type = DEV_CHAR;
	355	break;
	356	default:
f92523e3	357	return -EINVAL;
08ce5f16 SH	358	}
08ce5f16 SH	359	b++;
f92523e3 PM	360	if (!isspace(*b))
f92523e3 PM	361	return -EINVAL;
08ce5f16 SH	362	b++;
	363	if (b == '') {
	364	wh.major = ~0;
	365	b++;
	366	} else if (isdigit(*b)) {
7759fc9d LZ	367	wh.major = simple_strtoul(b, &endp, 10);
7759fc9d LZ	368	b = endp;
08ce5f16	369	} else {
f92523e3	370	return -EINVAL;
08ce5f16	371	}
f92523e3 PM	372	if (*b != ':')
f92523e3 PM	373	return -EINVAL;
08ce5f16 SH	374	b++;
	375
	376	/* read minor */
	377	if (b == '') {
	378	wh.minor = ~0;
	379	b++;
	380	} else if (isdigit(*b)) {
7759fc9d LZ	381	wh.minor = simple_strtoul(b, &endp, 10);
7759fc9d LZ	382	b = endp;
08ce5f16	383	} else {
f92523e3	384	return -EINVAL;
08ce5f16	385	}
f92523e3 PM	386	if (!isspace(*b))
f92523e3 PM	387	return -EINVAL;
08ce5f16 SH	388	for (b++, count = 0; count < 3; count++, b++) {
	389	switch (*b) {
	390	case 'r':
	391	wh.access \|= ACC_READ;
	392	break;
	393	case 'w':
	394	wh.access \|= ACC_WRITE;
	395	break;
	396	case 'm':
	397	wh.access \|= ACC_MKNOD;
	398	break;
	399	case '\n':
	400	case '\0':
	401	count = 3;
	402	break;
	403	default:
f92523e3	404	return -EINVAL;
08ce5f16 SH	405	}
	406	}
	407
	408	handle:
08ce5f16 SH	409	switch (filetype) {
08ce5f16 SH	410	case DEVCG_ALLOW:
f92523e3 PM	411	if (!parent_has_perm(devcgroup, &wh))
	412	return -EPERM;
	413	return dev_whitelist_add(devcgroup, &wh);
08ce5f16 SH	414	case DEVCG_DENY:
	415	dev_whitelist_rm(devcgroup, &wh);
	416	break;
	417	default:
f92523e3	418	return -EINVAL;
08ce5f16	419	}
f92523e3 PM	420	return 0;
f92523e3 PM	421	}
08ce5f16	422
f92523e3 PM	423	static int devcgroup_access_write(struct cgroup cgrp, struct cftype cft,
	424	const char *buffer)
	425	{
	426	int retval;
b4046f00 LZ	427
b4046f00 LZ	428	mutex_lock(&devcgroup_mutex);
f92523e3 PM	429	retval = devcgroup_update_access(cgroup_to_devcgroup(cgrp),
f92523e3 PM	430	cft->private, buffer);
b4046f00	431	mutex_unlock(&devcgroup_mutex);
08ce5f16 SH	432	return retval;
	433	}
	434
	435	static struct cftype dev_cgroup_files[] = {
	436	{
	437	.name = "allow",
f92523e3	438	.write_string = devcgroup_access_write,
08ce5f16 SH	439	.private = DEVCG_ALLOW,
	440	},
	441	{
	442	.name = "deny",
f92523e3	443	.write_string = devcgroup_access_write,
08ce5f16 SH	444	.private = DEVCG_DENY,
08ce5f16 SH	445	},
29486df3 SH	446	{
	447	.name = "list",
	448	.read_seq_string = devcgroup_seq_read,
	449	.private = DEVCG_LIST,
	450	},
08ce5f16 SH	451	};
	452
	453	static int devcgroup_populate(struct cgroup_subsys *ss,
	454	struct cgroup *cgroup)
	455	{
	456	return cgroup_add_files(cgroup, ss, dev_cgroup_files,
	457	ARRAY_SIZE(dev_cgroup_files));
	458	}
	459
	460	struct cgroup_subsys devices_subsys = {
	461	.name = "devices",
	462	.can_attach = devcgroup_can_attach,
	463	.create = devcgroup_create,
c5b60b5e	464	.destroy = devcgroup_destroy,
08ce5f16 SH	465	.populate = devcgroup_populate,
	466	.subsys_id = devices_subsys_id,
	467	};
	468
482e0cd3	469	int __devcgroup_inode_permission(struct inode *inode, int mask)
08ce5f16	470	{
08ce5f16 SH	471	struct dev_cgroup *dev_cgroup;
	472	struct dev_whitelist_item *wh;
	473
4efd1a1b	474	rcu_read_lock();
36fd71d2 LZ	475
	476	dev_cgroup = task_devcgroup(current);
	477
4efd1a1b	478	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	479	if (wh->type & DEV_ALL)
cd500819	480	goto found;
08ce5f16 SH	481	if ((wh->type & DEV_BLOCK) && !S_ISBLK(inode->i_mode))
	482	continue;
	483	if ((wh->type & DEV_CHAR) && !S_ISCHR(inode->i_mode))
	484	continue;
	485	if (wh->major != ~0 && wh->major != imajor(inode))
	486	continue;
	487	if (wh->minor != ~0 && wh->minor != iminor(inode))
	488	continue;
cd500819	489
08ce5f16 SH	490	if ((mask & MAY_WRITE) && !(wh->access & ACC_WRITE))
	491	continue;
	492	if ((mask & MAY_READ) && !(wh->access & ACC_READ))
	493	continue;
cd500819	494	found:
4efd1a1b	495	rcu_read_unlock();
08ce5f16 SH	496	return 0;
08ce5f16 SH	497	}
36fd71d2	498
4efd1a1b	499	rcu_read_unlock();
08ce5f16 SH	500
	501	return -EPERM;
	502	}
	503
	504	int devcgroup_inode_mknod(int mode, dev_t dev)
	505	{
08ce5f16 SH	506	struct dev_cgroup *dev_cgroup;
	507	struct dev_whitelist_item *wh;
	508
0b82ac37 SH	509	if (!S_ISBLK(mode) && !S_ISCHR(mode))
	510	return 0;
	511
4efd1a1b	512	rcu_read_lock();
36fd71d2 LZ	513
	514	dev_cgroup = task_devcgroup(current);
	515
116e0575	516	list_for_each_entry_rcu(wh, &dev_cgroup->whitelist, list) {
08ce5f16	517	if (wh->type & DEV_ALL)
cd500819	518	goto found;
08ce5f16 SH	519	if ((wh->type & DEV_BLOCK) && !S_ISBLK(mode))
	520	continue;
	521	if ((wh->type & DEV_CHAR) && !S_ISCHR(mode))
	522	continue;
	523	if (wh->major != ~0 && wh->major != MAJOR(dev))
	524	continue;
	525	if (wh->minor != ~0 && wh->minor != MINOR(dev))
	526	continue;
cd500819	527
08ce5f16 SH	528	if (!(wh->access & ACC_MKNOD))
08ce5f16 SH	529	continue;
cd500819	530	found:
4efd1a1b	531	rcu_read_unlock();
08ce5f16 SH	532	return 0;
08ce5f16 SH	533	}
36fd71d2	534
4efd1a1b	535	rcu_read_unlock();
36fd71d2	536
08ce5f16 SH	537	return -EPERM;
08ce5f16 SH	538	}