[linux.git] / mm / memfd.c

/*
 * memfd_create system call and file sealing support
 *
 * Code was originally included in shmem.c, and broken out to facilitate
 * use by hugetlbfs as well as tmpfs.
 *
 * This file is released under the GPL.
 */

#include <linux/fs.h>
#include <linux/vfs.h>
#include <linux/pagemap.h>
#include <linux/file.h>
#include <linux/mm.h>
#include <linux/sched/signal.h>
#include <linux/khugepaged.h>
#include <linux/syscalls.h>
#include <linux/hugetlb.h>
#include <linux/shmem_fs.h>
#include <linux/memfd.h>
#include <linux/pid_namespace.h>
#include <uapi/linux/memfd.h>

/*
 * We need a tag: a new tag would expand every xa_node by 8 bytes,
 * so reuse a tag which we firmly believe is never set or cleared on tmpfs
 * or hugetlbfs because they are memory only filesystems.
 */
#define MEMFD_TAG_PINNED        PAGECACHE_TAG_TOWRITE
#define LAST_SCAN               4       /* about 150ms max */

static void memfd_tag_pins(struct xa_state *xas)
{
	struct page *page;
	int latency = 0;
	int cache_count;

	lru_add_drain();

	xas_lock_irq(xas);
	xas_for_each(xas, page, ULONG_MAX) {
		cache_count = 1;
		if (!xa_is_value(page) &&
		    PageTransHuge(page) && !PageHuge(page))
			cache_count = HPAGE_PMD_NR;

		if (!xa_is_value(page) &&
		    page_count(page) - total_mapcount(page) != cache_count)
			xas_set_mark(xas, MEMFD_TAG_PINNED);
		if (cache_count != 1)
			xas_set(xas, page->index + cache_count);

		latency += cache_count;
		if (latency < XA_CHECK_SCHED)
			continue;
		latency = 0;

		xas_pause(xas);
		xas_unlock_irq(xas);
		cond_resched();
		xas_lock_irq(xas);
	}
	xas_unlock_irq(xas);
}

/*
 * Setting SEAL_WRITE requires us to verify there's no pending writer. However,
 * via get_user_pages(), drivers might have some pending I/O without any active
 * user-space mappings (eg., direct-IO, AIO). Therefore, we look at all pages
 * and see whether it has an elevated ref-count. If so, we tag them and wait for
 * them to be dropped.
 * The caller must guarantee that no new user will acquire writable references
 * to those pages to avoid races.
 */
static int memfd_wait_for_pins(struct address_space *mapping)
{
	XA_STATE(xas, &mapping->i_pages, 0);
	struct page *page;
	int error, scan;

	memfd_tag_pins(&xas);

	error = 0;
	for (scan = 0; scan <= LAST_SCAN; scan++) {
		int latency = 0;
		int cache_count;

		if (!xas_marked(&xas, MEMFD_TAG_PINNED))
			break;

		if (!scan)
			lru_add_drain_all();
		else if (schedule_timeout_killable((HZ << scan) / 200))
			scan = LAST_SCAN;

		xas_set(&xas, 0);
		xas_lock_irq(&xas);
		xas_for_each_marked(&xas, page, ULONG_MAX, MEMFD_TAG_PINNED) {
			bool clear = true;

			cache_count = 1;
			if (!xa_is_value(page) &&
			    PageTransHuge(page) && !PageHuge(page))
				cache_count = HPAGE_PMD_NR;

			if (!xa_is_value(page) && cache_count !=
			    page_count(page) - total_mapcount(page)) {
				/*
				 * On the last scan, we clean up all those tags
				 * we inserted; but make a note that we still
				 * found pages pinned.
				 */
				if (scan == LAST_SCAN)
					error = -EBUSY;
				else
					clear = false;
			}
			if (clear)
				xas_clear_mark(&xas, MEMFD_TAG_PINNED);

			latency += cache_count;
			if (latency < XA_CHECK_SCHED)
				continue;
			latency = 0;

			xas_pause(&xas);
			xas_unlock_irq(&xas);
			cond_resched();
			xas_lock_irq(&xas);
		}
		xas_unlock_irq(&xas);
	}

	return error;
}

static unsigned int *memfd_file_seals_ptr(struct file *file)
{
	if (shmem_file(file))
		return &SHMEM_I(file_inode(file))->seals;

#ifdef CONFIG_HUGETLBFS
	if (is_file_hugepages(file))
		return &HUGETLBFS_I(file_inode(file))->seals;
#endif

	return NULL;
}

#define F_ALL_SEALS (F_SEAL_SEAL | \
		     F_SEAL_EXEC | \
		     F_SEAL_SHRINK | \
		     F_SEAL_GROW | \
		     F_SEAL_WRITE | \
		     F_SEAL_FUTURE_WRITE)

static int memfd_add_seals(struct file *file, unsigned int seals)
{
	struct inode *inode = file_inode(file);
	unsigned int *file_seals;
	int error;

	/*
	 * SEALING
	 * Sealing allows multiple parties to share a tmpfs or hugetlbfs file
	 * but restrict access to a specific subset of file operations. Seals
	 * can only be added, but never removed. This way, mutually untrusted
	 * parties can share common memory regions with a well-defined policy.
	 * A malicious peer can thus never perform unwanted operations on a
	 * shared object.
	 *
	 * Seals are only supported on special tmpfs or hugetlbfs files and
	 * always affect the whole underlying inode. Once a seal is set, it
	 * may prevent some kinds of access to the file. Currently, the
	 * following seals are defined:
	 *   SEAL_SEAL: Prevent further seals from being set on this file
	 *   SEAL_SHRINK: Prevent the file from shrinking
	 *   SEAL_GROW: Prevent the file from growing
	 *   SEAL_WRITE: Prevent write access to the file
	 *   SEAL_EXEC: Prevent modification of the exec bits in the file mode
	 *
	 * As we don't require any trust relationship between two parties, we
	 * must prevent seals from being removed. Therefore, sealing a file
	 * only adds a given set of seals to the file, it never touches
	 * existing seals. Furthermore, the "setting seals"-operation can be
	 * sealed itself, which basically prevents any further seal from being
	 * added.
	 *
	 * Semantics of sealing are only defined on volatile files. Only
	 * anonymous tmpfs and hugetlbfs files support sealing. More
	 * importantly, seals are never written to disk. Therefore, there's
	 * no plan to support it on other file types.
	 */

	if (!(file->f_mode & FMODE_WRITE))
		return -EPERM;
	if (seals & ~(unsigned int)F_ALL_SEALS)
		return -EINVAL;

	inode_lock(inode);

	file_seals = memfd_file_seals_ptr(file);
	if (!file_seals) {
		error = -EINVAL;
		goto unlock;
	}

	if (*file_seals & F_SEAL_SEAL) {
		error = -EPERM;
		goto unlock;
	}

	if ((seals & F_SEAL_WRITE) && !(*file_seals & F_SEAL_WRITE)) {
		error = mapping_deny_writable(file->f_mapping);
		if (error)
			goto unlock;

		error = memfd_wait_for_pins(file->f_mapping);
		if (error) {
			mapping_allow_writable(file->f_mapping);
			goto unlock;
		}
	}

	/*
	 * SEAL_EXEC implys SEAL_WRITE, making W^X from the start.
	 */
	if (seals & F_SEAL_EXEC && inode->i_mode & 0111)
		seals |= F_SEAL_SHRINK|F_SEAL_GROW|F_SEAL_WRITE|F_SEAL_FUTURE_WRITE;

	*file_seals |= seals;
	error = 0;

unlock:
	inode_unlock(inode);
	return error;
}

static int memfd_get_seals(struct file *file)
{
	unsigned int *seals = memfd_file_seals_ptr(file);

	return seals ? *seals : -EINVAL;
}

long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int arg)
{
	long error;

	switch (cmd) {
	case F_ADD_SEALS:
		error = memfd_add_seals(file, arg);
		break;
	case F_GET_SEALS:
		error = memfd_get_seals(file);
		break;
	default:
		error = -EINVAL;
		break;
	}

	return error;
}

#define MFD_NAME_PREFIX "memfd:"
#define MFD_NAME_PREFIX_LEN (sizeof(MFD_NAME_PREFIX) - 1)
#define MFD_NAME_MAX_LEN (NAME_MAX - MFD_NAME_PREFIX_LEN)

#define MFD_ALL_FLAGS (MFD_CLOEXEC | MFD_ALLOW_SEALING | MFD_HUGETLB | MFD_NOEXEC_SEAL | MFD_EXEC)

static int check_sysctl_memfd_noexec(unsigned int *flags)
{
#ifdef CONFIG_SYSCTL
	struct pid_namespace *ns = task_active_pid_ns(current);
	int sysctl = pidns_memfd_noexec_scope(ns);

	if (!(*flags & (MFD_EXEC | MFD_NOEXEC_SEAL))) {
		if (sysctl >= MEMFD_NOEXEC_SCOPE_NOEXEC_SEAL)
			*flags |= MFD_NOEXEC_SEAL;
		else
			*flags |= MFD_EXEC;
	}

	if (!(*flags & MFD_NOEXEC_SEAL) && sysctl >= MEMFD_NOEXEC_SCOPE_NOEXEC_ENFORCED) {
		pr_err_ratelimited(
			"%s[%d]: memfd_create() requires MFD_NOEXEC_SEAL with vm.memfd_noexec=%d\n",
			current->comm, task_pid_nr(current), sysctl);
		return -EACCES;
	}
#endif
	return 0;
}

SYSCALL_DEFINE2(memfd_create,
		const char __user *, uname,
		unsigned int, flags)
{
	unsigned int *file_seals;
	struct file *file;
	int fd, error;
	char *name;
	long len;

	if (!(flags & MFD_HUGETLB)) {
		if (flags & ~(unsigned int)MFD_ALL_FLAGS)
			return -EINVAL;
	} else {
		/* Allow huge page size encoding in flags. */
		if (flags & ~(unsigned int)(MFD_ALL_FLAGS |
				(MFD_HUGE_MASK << MFD_HUGE_SHIFT)))
			return -EINVAL;
	}

	/* Invalid if both EXEC and NOEXEC_SEAL are set.*/
	if ((flags & MFD_EXEC) && (flags & MFD_NOEXEC_SEAL))
		return -EINVAL;

	if (!(flags & (MFD_EXEC | MFD_NOEXEC_SEAL))) {
		pr_warn_once(
			"%s[%d]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set\n",
			current->comm, task_pid_nr(current));
	}

	error = check_sysctl_memfd_noexec(&flags);
	if (error < 0)
		return error;

	/* length includes terminating zero */
	len = strnlen_user(uname, MFD_NAME_MAX_LEN + 1);
	if (len <= 0)
		return -EFAULT;
	if (len > MFD_NAME_MAX_LEN + 1)
		return -EINVAL;

	name = kmalloc(len + MFD_NAME_PREFIX_LEN, GFP_KERNEL);
	if (!name)
		return -ENOMEM;

	strcpy(name, MFD_NAME_PREFIX);
	if (copy_from_user(&name[MFD_NAME_PREFIX_LEN], uname, len)) {
		error = -EFAULT;
		goto err_name;
	}

	/* terminating-zero may have changed after strnlen_user() returned */
	if (name[len + MFD_NAME_PREFIX_LEN - 1]) {
		error = -EFAULT;
		goto err_name;
	}

	fd = get_unused_fd_flags((flags & MFD_CLOEXEC) ? O_CLOEXEC : 0);
	if (fd < 0) {
		error = fd;
		goto err_name;
	}

	if (flags & MFD_HUGETLB) {
		file = hugetlb_file_setup(name, 0, VM_NORESERVE,
					HUGETLB_ANONHUGE_INODE,
					(flags >> MFD_HUGE_SHIFT) &
					MFD_HUGE_MASK);
	} else
		file = shmem_file_setup(name, 0, VM_NORESERVE);
	if (IS_ERR(file)) {
		error = PTR_ERR(file);
		goto err_fd;
	}
	file->f_mode |= FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE;
	file->f_flags |= O_LARGEFILE;

	if (flags & MFD_NOEXEC_SEAL) {
		struct inode *inode = file_inode(file);

		inode->i_mode &= ~0111;
		file_seals = memfd_file_seals_ptr(file);
		if (file_seals) {
			*file_seals &= ~F_SEAL_SEAL;
			*file_seals |= F_SEAL_EXEC;
		}
	} else if (flags & MFD_ALLOW_SEALING) {
		/* MFD_EXEC and MFD_ALLOW_SEALING are set */
		file_seals = memfd_file_seals_ptr(file);
		if (file_seals)
			*file_seals &= ~F_SEAL_SEAL;
	}

	fd_install(fd, file);
	kfree(name);
	return fd;

err_fd:
	put_unused_fd(fd);
err_name:
	kfree(name);
	return error;
}
Commit	Line	Data
5d752600 MK	1	/*
	2	* memfd_create system call and file sealing support
	3	*
	4	* Code was originally included in shmem.c, and broken out to facilitate
	5	* use by hugetlbfs as well as tmpfs.
	6	*
	7	* This file is released under the GPL.
	8	*/
	9
	10	#include <linux/fs.h>
	11	#include <linux/vfs.h>
	12	#include <linux/pagemap.h>
	13	#include <linux/file.h>
	14	#include <linux/mm.h>
	15	#include <linux/sched/signal.h>
	16	#include <linux/khugepaged.h>
	17	#include <linux/syscalls.h>
	18	#include <linux/hugetlb.h>
	19	#include <linux/shmem_fs.h>
	20	#include <linux/memfd.h>
105ff533	21	#include <linux/pid_namespace.h>
5d752600 MK	22	#include <uapi/linux/memfd.h>
	23
	24	/*
2313216f	25	* We need a tag: a new tag would expand every xa_node by 8 bytes,
5d752600 MK	26	* so reuse a tag which we firmly believe is never set or cleared on tmpfs
	27	* or hugetlbfs because they are memory only filesystems.
	28	*/
	29	#define MEMFD_TAG_PINNED PAGECACHE_TAG_TOWRITE
	30	#define LAST_SCAN 4 /* about 150ms max */
	31
ef3038a5	32	static void memfd_tag_pins(struct xa_state *xas)
5d752600	33	{
5d752600	34	struct page *page;
f2b277c4 HD	35	int latency = 0;
f2b277c4 HD	36	int cache_count;
5d752600 MK	37
5d752600 MK	38	lru_add_drain();
5d752600	39
ef3038a5 MW	40	xas_lock_irq(xas);
ef3038a5 MW	41	xas_for_each(xas, page, ULONG_MAX) {
f2b277c4 HD	42	cache_count = 1;
	43	if (!xa_is_value(page) &&
	44	PageTransHuge(page) && !PageHuge(page))
	45	cache_count = HPAGE_PMD_NR;
	46
	47	if (!xa_is_value(page) &&
	48	page_count(page) - total_mapcount(page) != cache_count)
ef3038a5	49	xas_set_mark(xas, MEMFD_TAG_PINNED);
f2b277c4 HD	50	if (cache_count != 1)
f2b277c4 HD	51	xas_set(xas, page->index + cache_count);
5d752600	52
f2b277c4 HD	53	latency += cache_count;
f2b277c4 HD	54	if (latency < XA_CHECK_SCHED)
ef3038a5	55	continue;
f2b277c4	56	latency = 0;
ef3038a5 MW	57
	58	xas_pause(xas);
	59	xas_unlock_irq(xas);
	60	cond_resched();
	61	xas_lock_irq(xas);
5d752600	62	}
ef3038a5	63	xas_unlock_irq(xas);
5d752600 MK	64	}
	65
	66	/*
	67	* Setting SEAL_WRITE requires us to verify there's no pending writer. However,
	68	* via get_user_pages(), drivers might have some pending I/O without any active
	69	* user-space mappings (eg., direct-IO, AIO). Therefore, we look at all pages
	70	* and see whether it has an elevated ref-count. If so, we tag them and wait for
	71	* them to be dropped.
	72	* The caller must guarantee that no new user will acquire writable references
	73	* to those pages to avoid races.
	74	*/
	75	static int memfd_wait_for_pins(struct address_space *mapping)
	76	{
2313216f	77	XA_STATE(xas, &mapping->i_pages, 0);
5d752600 MK	78	struct page *page;
	79	int error, scan;
	80
ef3038a5	81	memfd_tag_pins(&xas);
5d752600 MK	82
	83	error = 0;
	84	for (scan = 0; scan <= LAST_SCAN; scan++) {
f2b277c4 HD	85	int latency = 0;
f2b277c4 HD	86	int cache_count;
2313216f MW	87
2313216f MW	88	if (!xas_marked(&xas, MEMFD_TAG_PINNED))
5d752600 MK	89	break;
	90
	91	if (!scan)
	92	lru_add_drain_all();
	93	else if (schedule_timeout_killable((HZ << scan) / 200))
	94	scan = LAST_SCAN;
	95
2313216f MW	96	xas_set(&xas, 0);
	97	xas_lock_irq(&xas);
	98	xas_for_each_marked(&xas, page, ULONG_MAX, MEMFD_TAG_PINNED) {
	99	bool clear = true;
f2b277c4 HD	100
	101	cache_count = 1;
	102	if (!xa_is_value(page) &&
	103	PageTransHuge(page) && !PageHuge(page))
	104	cache_count = HPAGE_PMD_NR;
	105
	106	if (!xa_is_value(page) && cache_count !=
	107	page_count(page) - total_mapcount(page)) {
5d752600 MK	108	/*
	109	* On the last scan, we clean up all those tags
	110	* we inserted; but make a note that we still
	111	* found pages pinned.
	112	*/
2313216f MW	113	if (scan == LAST_SCAN)
	114	error = -EBUSY;
	115	else
	116	clear = false;
5d752600	117	}
2313216f MW	118	if (clear)
2313216f MW	119	xas_clear_mark(&xas, MEMFD_TAG_PINNED);
f2b277c4 HD	120
	121	latency += cache_count;
	122	if (latency < XA_CHECK_SCHED)
2313216f	123	continue;
f2b277c4	124	latency = 0;
5d752600	125
2313216f MW	126	xas_pause(&xas);
	127	xas_unlock_irq(&xas);
	128	cond_resched();
	129	xas_lock_irq(&xas);
5d752600	130	}
2313216f	131	xas_unlock_irq(&xas);
5d752600 MK	132	}
	133
	134	return error;
	135	}
	136
	137	static unsigned int memfd_file_seals_ptr(struct file file)
	138	{
	139	if (shmem_file(file))
	140	return &SHMEM_I(file_inode(file))->seals;
	141
	142	#ifdef CONFIG_HUGETLBFS
	143	if (is_file_hugepages(file))
	144	return &HUGETLBFS_I(file_inode(file))->seals;
	145	#endif
	146
	147	return NULL;
	148	}
	149
	150	#define F_ALL_SEALS (F_SEAL_SEAL \| \
6fd73538	151	F_SEAL_EXEC \| \
5d752600 MK	152	F_SEAL_SHRINK \| \
5d752600 MK	153	F_SEAL_GROW \| \
ab3948f5 JFG	154	F_SEAL_WRITE \| \
ab3948f5 JFG	155	F_SEAL_FUTURE_WRITE)
5d752600 MK	156
	157	static int memfd_add_seals(struct file *file, unsigned int seals)
	158	{
	159	struct inode *inode = file_inode(file);
	160	unsigned int *file_seals;
	161	int error;
	162
	163	/*
	164	* SEALING
	165	* Sealing allows multiple parties to share a tmpfs or hugetlbfs file
	166	* but restrict access to a specific subset of file operations. Seals
	167	* can only be added, but never removed. This way, mutually untrusted
	168	* parties can share common memory regions with a well-defined policy.
	169	* A malicious peer can thus never perform unwanted operations on a
	170	* shared object.
	171	*
	172	* Seals are only supported on special tmpfs or hugetlbfs files and
	173	* always affect the whole underlying inode. Once a seal is set, it
	174	* may prevent some kinds of access to the file. Currently, the
	175	* following seals are defined:
	176	* SEAL_SEAL: Prevent further seals from being set on this file
	177	* SEAL_SHRINK: Prevent the file from shrinking
	178	* SEAL_GROW: Prevent the file from growing
	179	* SEAL_WRITE: Prevent write access to the file
6fd73538	180	* SEAL_EXEC: Prevent modification of the exec bits in the file mode
5d752600 MK	181	*
	182	* As we don't require any trust relationship between two parties, we
	183	* must prevent seals from being removed. Therefore, sealing a file
	184	* only adds a given set of seals to the file, it never touches
	185	* existing seals. Furthermore, the "setting seals"-operation can be
	186	* sealed itself, which basically prevents any further seal from being
	187	* added.
	188	*
	189	* Semantics of sealing are only defined on volatile files. Only
	190	* anonymous tmpfs and hugetlbfs files support sealing. More
	191	* importantly, seals are never written to disk. Therefore, there's
	192	* no plan to support it on other file types.
	193	*/
	194
	195	if (!(file->f_mode & FMODE_WRITE))
	196	return -EPERM;
	197	if (seals & ~(unsigned int)F_ALL_SEALS)
	198	return -EINVAL;
	199
	200	inode_lock(inode);
	201
	202	file_seals = memfd_file_seals_ptr(file);
	203	if (!file_seals) {
	204	error = -EINVAL;
	205	goto unlock;
	206	}
	207
	208	if (*file_seals & F_SEAL_SEAL) {
	209	error = -EPERM;
	210	goto unlock;
	211	}
	212
	213	if ((seals & F_SEAL_WRITE) && !(*file_seals & F_SEAL_WRITE)) {
	214	error = mapping_deny_writable(file->f_mapping);
	215	if (error)
	216	goto unlock;
	217
	218	error = memfd_wait_for_pins(file->f_mapping);
	219	if (error) {
	220	mapping_allow_writable(file->f_mapping);
	221	goto unlock;
	222	}
	223	}
	224
c4f75bc8 JX	225	/*
	226	* SEAL_EXEC implys SEAL_WRITE, making W^X from the start.
	227	*/
	228	if (seals & F_SEAL_EXEC && inode->i_mode & 0111)
	229	seals \|= F_SEAL_SHRINK\|F_SEAL_GROW\|F_SEAL_WRITE\|F_SEAL_FUTURE_WRITE;
	230
5d752600 MK	231	*file_seals \|= seals;
	232	error = 0;
	233
	234	unlock:
	235	inode_unlock(inode);
	236	return error;
	237	}
	238
	239	static int memfd_get_seals(struct file *file)
	240	{
	241	unsigned int *seals = memfd_file_seals_ptr(file);
	242
	243	return seals ? *seals : -EINVAL;
	244	}
	245
f7b8f70b	246	long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int arg)
5d752600 MK	247	{
	248	long error;
	249
	250	switch (cmd) {
	251	case F_ADD_SEALS:
5d752600 MK	252	error = memfd_add_seals(file, arg);
	253	break;
	254	case F_GET_SEALS:
	255	error = memfd_get_seals(file);
	256	break;
	257	default:
	258	error = -EINVAL;
	259	break;
	260	}
	261
	262	return error;
	263	}
	264
	265	#define MFD_NAME_PREFIX "memfd:"
	266	#define MFD_NAME_PREFIX_LEN (sizeof(MFD_NAME_PREFIX) - 1)
	267	#define MFD_NAME_MAX_LEN (NAME_MAX - MFD_NAME_PREFIX_LEN)
	268
105ff533	269	#define MFD_ALL_FLAGS (MFD_CLOEXEC \| MFD_ALLOW_SEALING \| MFD_HUGETLB \| MFD_NOEXEC_SEAL \| MFD_EXEC)
5d752600	270
72de2591 JX	271	static int check_sysctl_memfd_noexec(unsigned int *flags)
	272	{
	273	#ifdef CONFIG_SYSCTL
9876cfe8 AS	274	struct pid_namespace *ns = task_active_pid_ns(current);
9876cfe8 AS	275	int sysctl = pidns_memfd_noexec_scope(ns);
72de2591 JX	276
72de2591 JX	277	if (!(*flags & (MFD_EXEC \| MFD_NOEXEC_SEAL))) {
202e1422	278	if (sysctl >= MEMFD_NOEXEC_SCOPE_NOEXEC_SEAL)
72de2591 JX	279	*flags \|= MFD_NOEXEC_SEAL;
	280	else
	281	*flags \|= MFD_EXEC;
	282	}
	283
202e1422 AS	284	if (!(*flags & MFD_NOEXEC_SEAL) && sysctl >= MEMFD_NOEXEC_SCOPE_NOEXEC_ENFORCED) {
	285	pr_err_ratelimited(
	286	"%s[%d]: memfd_create() requires MFD_NOEXEC_SEAL with vm.memfd_noexec=%d\n",
	287	current->comm, task_pid_nr(current), sysctl);
72de2591 JX	288	return -EACCES;
	289	}
	290	#endif
72de2591 JX	291	return 0;
	292	}
	293
5d752600 MK	294	SYSCALL_DEFINE2(memfd_create,
	295	const char __user *, uname,
	296	unsigned int, flags)
	297	{
	298	unsigned int *file_seals;
	299	struct file *file;
	300	int fd, error;
	301	char *name;
	302	long len;
	303
	304	if (!(flags & MFD_HUGETLB)) {
	305	if (flags & ~(unsigned int)MFD_ALL_FLAGS)
	306	return -EINVAL;
	307	} else {
	308	/* Allow huge page size encoding in flags. */
	309	if (flags & ~(unsigned int)(MFD_ALL_FLAGS \|
	310	(MFD_HUGE_MASK << MFD_HUGE_SHIFT)))
	311	return -EINVAL;
	312	}
	313
105ff533 JX	314	/* Invalid if both EXEC and NOEXEC_SEAL are set.*/
	315	if ((flags & MFD_EXEC) && (flags & MFD_NOEXEC_SEAL))
	316	return -EINVAL;
	317
	318	if (!(flags & (MFD_EXEC \| MFD_NOEXEC_SEAL))) {
2562d67b	319	pr_warn_once(
202e1422 AS	320	"%s[%d]: memfd_create() called without MFD_EXEC or MFD_NOEXEC_SEAL set\n",
202e1422 AS	321	current->comm, task_pid_nr(current));
105ff533 JX	322	}
105ff533 JX	323
202e1422 AS	324	error = check_sysctl_memfd_noexec(&flags);
	325	if (error < 0)
	326	return error;
72de2591	327
5d752600 MK	328	/* length includes terminating zero */
	329	len = strnlen_user(uname, MFD_NAME_MAX_LEN + 1);
	330	if (len <= 0)
	331	return -EFAULT;
	332	if (len > MFD_NAME_MAX_LEN + 1)
	333	return -EINVAL;
	334
	335	name = kmalloc(len + MFD_NAME_PREFIX_LEN, GFP_KERNEL);
	336	if (!name)
	337	return -ENOMEM;
	338
	339	strcpy(name, MFD_NAME_PREFIX);
	340	if (copy_from_user(&name[MFD_NAME_PREFIX_LEN], uname, len)) {
	341	error = -EFAULT;
	342	goto err_name;
	343	}
	344
	345	/* terminating-zero may have changed after strnlen_user() returned */
	346	if (name[len + MFD_NAME_PREFIX_LEN - 1]) {
	347	error = -EFAULT;
	348	goto err_name;
	349	}
	350
	351	fd = get_unused_fd_flags((flags & MFD_CLOEXEC) ? O_CLOEXEC : 0);
	352	if (fd < 0) {
	353	error = fd;
	354	goto err_name;
	355	}
	356
	357	if (flags & MFD_HUGETLB) {
83c1fd76	358	file = hugetlb_file_setup(name, 0, VM_NORESERVE,
5d752600 MK	359	HUGETLB_ANONHUGE_INODE,
	360	(flags >> MFD_HUGE_SHIFT) &
	361	MFD_HUGE_MASK);
	362	} else
	363	file = shmem_file_setup(name, 0, VM_NORESERVE);
	364	if (IS_ERR(file)) {
	365	error = PTR_ERR(file);
	366	goto err_fd;
	367	}
	368	file->f_mode \|= FMODE_LSEEK \| FMODE_PREAD \| FMODE_PWRITE;
c9c554f2	369	file->f_flags \|= O_LARGEFILE;
5d752600	370
105ff533 JX	371	if (flags & MFD_NOEXEC_SEAL) {
	372	struct inode *inode = file_inode(file);
	373
	374	inode->i_mode &= ~0111;
	375	file_seals = memfd_file_seals_ptr(file);
935d44ac RS	376	if (file_seals) {
	377	*file_seals &= ~F_SEAL_SEAL;
	378	*file_seals \|= F_SEAL_EXEC;
	379	}
105ff533 JX	380	} else if (flags & MFD_ALLOW_SEALING) {
105ff533 JX	381	/* MFD_EXEC and MFD_ALLOW_SEALING are set */
5d752600	382	file_seals = memfd_file_seals_ptr(file);
935d44ac RS	383	if (file_seals)
935d44ac RS	384	*file_seals &= ~F_SEAL_SEAL;
5d752600 MK	385	}
	386
	387	fd_install(fd, file);
	388	kfree(name);
	389	return fd;
	390
	391	err_fd:
	392	put_unused_fd(fd);
	393	err_name:
	394	kfree(name);
	395	return error;
	396	}