[linux-2.6-block.git] / fs / f2fs / verity.c

// SPDX-License-Identifier: GPL-2.0
/*
 * fs/f2fs/verity.c: fs-verity support for f2fs
 *
 * Copyright 2019 Google LLC
 */

/*
 * Implementation of fsverity_operations for f2fs.
 *
 * Like ext4, f2fs stores the verity metadata (Merkle tree and
 * fsverity_descriptor) past the end of the file, starting at the first 64K
 * boundary beyond i_size.  This approach works because (a) verity files are
 * readonly, and (b) pages fully beyond i_size aren't visible to userspace but
 * can be read/written internally by f2fs with only some relatively small
 * changes to f2fs.  Extended attributes cannot be used because (a) f2fs limits
 * the total size of an inode's xattr entries to 4096 bytes, which wouldn't be
 * enough for even a single Merkle tree block, and (b) f2fs encryption doesn't
 * encrypt xattrs, yet the verity metadata *must* be encrypted when the file is
 * because it contains hashes of the plaintext data.
 *
 * Using a 64K boundary rather than a 4K one keeps things ready for
 * architectures with 64K pages, and it doesn't necessarily waste space on-disk
 * since there can be a hole between i_size and the start of the Merkle tree.
 */

#include <linux/f2fs_fs.h>

#include "f2fs.h"
#include "xattr.h"

#define F2FS_VERIFY_VER	(1)

static inline loff_t f2fs_verity_metadata_pos(const struct inode *inode)
{
	return round_up(inode->i_size, 65536);
}

/*
 * Read some verity metadata from the inode.  __vfs_read() can't be used because
 * we need to read beyond i_size.
 */
static int pagecache_read(struct inode *inode, void *buf, size_t count,
			  loff_t pos)
{
	while (count) {
		size_t n = min_t(size_t, count,
				 PAGE_SIZE - offset_in_page(pos));
		struct page *page;

		page = read_mapping_page(inode->i_mapping, pos >> PAGE_SHIFT,
					 NULL);
		if (IS_ERR(page))
			return PTR_ERR(page);

		memcpy_from_page(buf, page, offset_in_page(pos), n);

		put_page(page);

		buf += n;
		pos += n;
		count -= n;
	}
	return 0;
}

/*
 * Write some verity metadata to the inode for FS_IOC_ENABLE_VERITY.
 * kernel_write() can't be used because the file descriptor is readonly.
 */
static int pagecache_write(struct inode *inode, const void *buf, size_t count,
			   loff_t pos)
{
	struct address_space *mapping = inode->i_mapping;
	const struct address_space_operations *aops = mapping->a_ops;

	if (pos + count > inode->i_sb->s_maxbytes)
		return -EFBIG;

	while (count) {
		size_t n = min_t(size_t, count,
				 PAGE_SIZE - offset_in_page(pos));
		struct page *page;
		void *fsdata = NULL;
		int res;

		res = aops->write_begin(NULL, mapping, pos, n, &page, &fsdata);
		if (res)
			return res;

		memcpy_to_page(page, offset_in_page(pos), buf, n);

		res = aops->write_end(NULL, mapping, pos, n, n, page, fsdata);
		if (res < 0)
			return res;
		if (res != n)
			return -EIO;

		buf += n;
		pos += n;
		count -= n;
	}
	return 0;
}

/*
 * Format of f2fs verity xattr.  This points to the location of the verity
 * descriptor within the file data rather than containing it directly because
 * the verity descriptor *must* be encrypted when f2fs encryption is used.  But,
 * f2fs encryption does not encrypt xattrs.
 */
struct fsverity_descriptor_location {
	__le32 version;
	__le32 size;
	__le64 pos;
};

static int f2fs_begin_enable_verity(struct file *filp)
{
	struct inode *inode = file_inode(filp);
	int err;

	if (f2fs_verity_in_progress(inode))
		return -EBUSY;

	if (f2fs_is_atomic_file(inode))
		return -EOPNOTSUPP;

	/*
	 * Since the file was opened readonly, we have to initialize the quotas
	 * here and not rely on ->open() doing it.  This must be done before
	 * evicting the inline data.
	 */
	err = f2fs_dquot_initialize(inode);
	if (err)
		return err;

	err = f2fs_convert_inline_inode(inode);
	if (err)
		return err;

	set_inode_flag(inode, FI_VERITY_IN_PROGRESS);
	return 0;
}

static int f2fs_end_enable_verity(struct file *filp, const void *desc,
				  size_t desc_size, u64 merkle_tree_size)
{
	struct inode *inode = file_inode(filp);
	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
	u64 desc_pos = f2fs_verity_metadata_pos(inode) + merkle_tree_size;
	struct fsverity_descriptor_location dloc = {
		.version = cpu_to_le32(F2FS_VERIFY_VER),
		.size = cpu_to_le32(desc_size),
		.pos = cpu_to_le64(desc_pos),
	};
	int err = 0, err2 = 0;

	/*
	 * If an error already occurred (which fs/verity/ signals by passing
	 * desc == NULL), then only clean-up is needed.
	 */
	if (desc == NULL)
		goto cleanup;

	/* Append the verity descriptor. */
	err = pagecache_write(inode, desc, desc_size, desc_pos);
	if (err)
		goto cleanup;

	/*
	 * Write all pages (both data and verity metadata).  Note that this must
	 * happen before clearing FI_VERITY_IN_PROGRESS; otherwise pages beyond
	 * i_size won't be written properly.  For crash consistency, this also
	 * must happen before the verity inode flag gets persisted.
	 */
	err = filemap_write_and_wait(inode->i_mapping);
	if (err)
		goto cleanup;

	/* Set the verity xattr. */
	err = f2fs_setxattr(inode, F2FS_XATTR_INDEX_VERITY,
			    F2FS_XATTR_NAME_VERITY, &dloc, sizeof(dloc),
			    NULL, XATTR_CREATE);
	if (err)
		goto cleanup;

	/* Finally, set the verity inode flag. */
	file_set_verity(inode);
	f2fs_set_inode_flags(inode);
	f2fs_mark_inode_dirty_sync(inode, true);

	clear_inode_flag(inode, FI_VERITY_IN_PROGRESS);
	return 0;

cleanup:
	/*
	 * Verity failed to be enabled, so clean up by truncating any verity
	 * metadata that was written beyond i_size (both from cache and from
	 * disk) and clearing FI_VERITY_IN_PROGRESS.
	 *
	 * Taking i_gc_rwsem[WRITE] is needed to stop f2fs garbage collection
	 * from re-instantiating cached pages we are truncating (since unlike
	 * normal file accesses, garbage collection isn't limited by i_size).
	 */
	f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
	truncate_inode_pages(inode->i_mapping, inode->i_size);
	err2 = f2fs_truncate(inode);
	if (err2) {
		f2fs_err(sbi, "Truncating verity metadata failed (errno=%d)",
			 err2);
		set_sbi_flag(sbi, SBI_NEED_FSCK);
	}
	f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
	clear_inode_flag(inode, FI_VERITY_IN_PROGRESS);
	return err ?: err2;
}

static int f2fs_get_verity_descriptor(struct inode *inode, void *buf,
				      size_t buf_size)
{
	struct fsverity_descriptor_location dloc;
	int res;
	u32 size;
	u64 pos;

	/* Get the descriptor location */
	res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_VERITY,
			    F2FS_XATTR_NAME_VERITY, &dloc, sizeof(dloc), NULL);
	if (res < 0 && res != -ERANGE)
		return res;
	if (res != sizeof(dloc) || dloc.version != cpu_to_le32(F2FS_VERIFY_VER)) {
		f2fs_warn(F2FS_I_SB(inode), "unknown verity xattr format");
		return -EINVAL;
	}
	size = le32_to_cpu(dloc.size);
	pos = le64_to_cpu(dloc.pos);

	/* Get the descriptor */
	if (pos + size < pos || pos + size > inode->i_sb->s_maxbytes ||
	    pos < f2fs_verity_metadata_pos(inode) || size > INT_MAX) {
		f2fs_warn(F2FS_I_SB(inode), "invalid verity xattr");
		f2fs_handle_error(F2FS_I_SB(inode),
				ERROR_CORRUPTED_VERITY_XATTR);
		return -EFSCORRUPTED;
	}
	if (buf_size) {
		if (size > buf_size)
			return -ERANGE;
		res = pagecache_read(inode, buf, size, pos);
		if (res)
			return res;
	}
	return size;
}

static struct page *f2fs_read_merkle_tree_page(struct inode *inode,
					       pgoff_t index,
					       unsigned long num_ra_pages)
{
	struct folio *folio;

	index += f2fs_verity_metadata_pos(inode) >> PAGE_SHIFT;

	folio = __filemap_get_folio(inode->i_mapping, index, FGP_ACCESSED, 0);
	if (IS_ERR(folio) || !folio_test_uptodate(folio)) {
		DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index);

		if (!IS_ERR(folio))
			folio_put(folio);
		else if (num_ra_pages > 1)
			page_cache_ra_unbounded(&ractl, num_ra_pages, 0);
		folio = read_mapping_folio(inode->i_mapping, index, NULL);
		if (IS_ERR(folio))
			return ERR_CAST(folio);
	}
	return folio_file_page(folio, index);
}

static int f2fs_write_merkle_tree_block(struct inode *inode, const void *buf,
					u64 pos, unsigned int size)
{
	pos += f2fs_verity_metadata_pos(inode);

	return pagecache_write(inode, buf, size, pos);
}

const struct fsverity_operations f2fs_verityops = {
	.begin_enable_verity	= f2fs_begin_enable_verity,
	.end_enable_verity	= f2fs_end_enable_verity,
	.get_verity_descriptor	= f2fs_get_verity_descriptor,
	.read_merkle_tree_page	= f2fs_read_merkle_tree_page,
	.write_merkle_tree_block = f2fs_write_merkle_tree_block,
};
Commit	Line	Data
95ae251f EB	1	// SPDX-License-Identifier: GPL-2.0
	2	/*
	3	* fs/f2fs/verity.c: fs-verity support for f2fs
	4	*
	5	* Copyright 2019 Google LLC
	6	*/
	7
	8	/*
	9	* Implementation of fsverity_operations for f2fs.
	10	*
	11	* Like ext4, f2fs stores the verity metadata (Merkle tree and
	12	* fsverity_descriptor) past the end of the file, starting at the first 64K
	13	* boundary beyond i_size. This approach works because (a) verity files are
	14	* readonly, and (b) pages fully beyond i_size aren't visible to userspace but
	15	* can be read/written internally by f2fs with only some relatively small
	16	* changes to f2fs. Extended attributes cannot be used because (a) f2fs limits
	17	* the total size of an inode's xattr entries to 4096 bytes, which wouldn't be
	18	* enough for even a single Merkle tree block, and (b) f2fs encryption doesn't
	19	* encrypt xattrs, yet the verity metadata must be encrypted when the file is
	20	* because it contains hashes of the plaintext data.
	21	*
	22	* Using a 64K boundary rather than a 4K one keeps things ready for
	23	* architectures with 64K pages, and it doesn't necessarily waste space on-disk
	24	* since there can be a hole between i_size and the start of the Merkle tree.
	25	*/
	26
	27	#include <linux/f2fs_fs.h>
	28
	29	#include "f2fs.h"
	30	#include "xattr.h"
	31
8fa41016 JQ	32	#define F2FS_VERIFY_VER (1)
8fa41016 JQ	33
95ae251f EB	34	static inline loff_t f2fs_verity_metadata_pos(const struct inode *inode)
	35	{
	36	return round_up(inode->i_size, 65536);
	37	}
	38
	39	/*
	40	* Read some verity metadata from the inode. __vfs_read() can't be used because
	41	* we need to read beyond i_size.
	42	*/
	43	static int pagecache_read(struct inode inode, void buf, size_t count,
	44	loff_t pos)
	45	{
	46	while (count) {
	47	size_t n = min_t(size_t, count,
	48	PAGE_SIZE - offset_in_page(pos));
	49	struct page *page;
95ae251f EB	50
	51	page = read_mapping_page(inode->i_mapping, pos >> PAGE_SHIFT,
	52	NULL);
	53	if (IS_ERR(page))
	54	return PTR_ERR(page);
	55
b87846bd	56	memcpy_from_page(buf, page, offset_in_page(pos), n);
95ae251f EB	57
	58	put_page(page);
	59
	60	buf += n;
	61	pos += n;
	62	count -= n;
	63	}
	64	return 0;
	65	}
	66
	67	/*
	68	* Write some verity metadata to the inode for FS_IOC_ENABLE_VERITY.
	69	* kernel_write() can't be used because the file descriptor is readonly.
	70	*/
	71	static int pagecache_write(struct inode inode, const void buf, size_t count,
	72	loff_t pos)
	73	{
6e0ee044 MWO	74	struct address_space *mapping = inode->i_mapping;
	75	const struct address_space_operations *aops = mapping->a_ops;
	76
95ae251f EB	77	if (pos + count > inode->i_sb->s_maxbytes)
	78	return -EFBIG;
	79
	80	while (count) {
	81	size_t n = min_t(size_t, count,
	82	PAGE_SIZE - offset_in_page(pos));
	83	struct page *page;
b1b98967	84	void *fsdata = NULL;
95ae251f EB	85	int res;
95ae251f EB	86
6e0ee044	87	res = aops->write_begin(NULL, mapping, pos, n, &page, &fsdata);
95ae251f EB	88	if (res)
	89	return res;
	90
b87846bd	91	memcpy_to_page(page, offset_in_page(pos), buf, n);
95ae251f	92
6e0ee044	93	res = aops->write_end(NULL, mapping, pos, n, n, page, fsdata);
95ae251f EB	94	if (res < 0)
	95	return res;
	96	if (res != n)
	97	return -EIO;
	98
	99	buf += n;
	100	pos += n;
	101	count -= n;
	102	}
	103	return 0;
	104	}
	105
	106	/*
	107	* Format of f2fs verity xattr. This points to the location of the verity
	108	* descriptor within the file data rather than containing it directly because
	109	* the verity descriptor must be encrypted when f2fs encryption is used. But,
	110	* f2fs encryption does not encrypt xattrs.
	111	*/
	112	struct fsverity_descriptor_location {
	113	__le32 version;
	114	__le32 size;
	115	__le64 pos;
	116	};
	117
	118	static int f2fs_begin_enable_verity(struct file *filp)
	119	{
	120	struct inode *inode = file_inode(filp);
	121	int err;
	122
	123	if (f2fs_verity_in_progress(inode))
	124	return -EBUSY;
	125
7bc155fe	126	if (f2fs_is_atomic_file(inode))
95ae251f EB	127	return -EOPNOTSUPP;
	128
	129	/*
	130	* Since the file was opened readonly, we have to initialize the quotas
	131	* here and not rely on ->open() doing it. This must be done before
	132	* evicting the inline data.
	133	*/
10a26878	134	err = f2fs_dquot_initialize(inode);
95ae251f EB	135	if (err)
	136	return err;
	137
	138	err = f2fs_convert_inline_inode(inode);
	139	if (err)
	140	return err;
	141
	142	set_inode_flag(inode, FI_VERITY_IN_PROGRESS);
	143	return 0;
	144	}
	145
	146	static int f2fs_end_enable_verity(struct file filp, const void desc,
	147	size_t desc_size, u64 merkle_tree_size)
	148	{
	149	struct inode *inode = file_inode(filp);
3c031542	150	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
95ae251f EB	151	u64 desc_pos = f2fs_verity_metadata_pos(inode) + merkle_tree_size;
95ae251f EB	152	struct fsverity_descriptor_location dloc = {
8fa41016	153	.version = cpu_to_le32(F2FS_VERIFY_VER),
95ae251f EB	154	.size = cpu_to_le32(desc_size),
	155	.pos = cpu_to_le64(desc_pos),
	156	};
3c031542	157	int err = 0, err2 = 0;
95ae251f	158
3c031542 EB	159	/*
	160	* If an error already occurred (which fs/verity/ signals by passing
	161	* desc == NULL), then only clean-up is needed.
	162	*/
	163	if (desc == NULL)
	164	goto cleanup;
95ae251f	165
3c031542 EB	166	/* Append the verity descriptor. */
	167	err = pagecache_write(inode, desc, desc_size, desc_pos);
	168	if (err)
	169	goto cleanup;
	170
	171	/*
	172	* Write all pages (both data and verity metadata). Note that this must
	173	* happen before clearing FI_VERITY_IN_PROGRESS; otherwise pages beyond
	174	* i_size won't be written properly. For crash consistency, this also
	175	* must happen before the verity inode flag gets persisted.
	176	*/
	177	err = filemap_write_and_wait(inode->i_mapping);
	178	if (err)
	179	goto cleanup;
	180
	181	/* Set the verity xattr. */
	182	err = f2fs_setxattr(inode, F2FS_XATTR_INDEX_VERITY,
	183	F2FS_XATTR_NAME_VERITY, &dloc, sizeof(dloc),
	184	NULL, XATTR_CREATE);
	185	if (err)
	186	goto cleanup;
95ae251f	187
3c031542 EB	188	/* Finally, set the verity inode flag. */
	189	file_set_verity(inode);
	190	f2fs_set_inode_flags(inode);
	191	f2fs_mark_inode_dirty_sync(inode, true);
95ae251f EB	192
95ae251f EB	193	clear_inode_flag(inode, FI_VERITY_IN_PROGRESS);
3c031542	194	return 0;
95ae251f	195
3c031542 EB	196	cleanup:
	197	/*
	198	* Verity failed to be enabled, so clean up by truncating any verity
	199	* metadata that was written beyond i_size (both from cache and from
	200	* disk) and clearing FI_VERITY_IN_PROGRESS.
	201	*
	202	* Taking i_gc_rwsem[WRITE] is needed to stop f2fs garbage collection
	203	* from re-instantiating cached pages we are truncating (since unlike
	204	* normal file accesses, garbage collection isn't limited by i_size).
	205	*/
e4544b63	206	f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
3c031542 EB	207	truncate_inode_pages(inode->i_mapping, inode->i_size);
	208	err2 = f2fs_truncate(inode);
	209	if (err2) {
	210	f2fs_err(sbi, "Truncating verity metadata failed (errno=%d)",
	211	err2);
	212	set_sbi_flag(sbi, SBI_NEED_FSCK);
95ae251f	213	}
e4544b63	214	f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
3c031542 EB	215	clear_inode_flag(inode, FI_VERITY_IN_PROGRESS);
3c031542 EB	216	return err ?: err2;
95ae251f EB	217	}
	218
	219	static int f2fs_get_verity_descriptor(struct inode inode, void buf,
	220	size_t buf_size)
	221	{
	222	struct fsverity_descriptor_location dloc;
	223	int res;
	224	u32 size;
	225	u64 pos;
	226
	227	/* Get the descriptor location */
	228	res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_VERITY,
	229	F2FS_XATTR_NAME_VERITY, &dloc, sizeof(dloc), NULL);
	230	if (res < 0 && res != -ERANGE)
	231	return res;
8fa41016	232	if (res != sizeof(dloc) \|\| dloc.version != cpu_to_le32(F2FS_VERIFY_VER)) {
95ae251f EB	233	f2fs_warn(F2FS_I_SB(inode), "unknown verity xattr format");
	234	return -EINVAL;
	235	}
	236	size = le32_to_cpu(dloc.size);
	237	pos = le64_to_cpu(dloc.pos);
	238
	239	/* Get the descriptor */
	240	if (pos + size < pos \|\| pos + size > inode->i_sb->s_maxbytes \|\|
	241	pos < f2fs_verity_metadata_pos(inode) \|\| size > INT_MAX) {
	242	f2fs_warn(F2FS_I_SB(inode), "invalid verity xattr");
95fa90c9 CY	243	f2fs_handle_error(F2FS_I_SB(inode),
95fa90c9 CY	244	ERROR_CORRUPTED_VERITY_XATTR);
95ae251f EB	245	return -EFSCORRUPTED;
	246	}
	247	if (buf_size) {
	248	if (size > buf_size)
	249	return -ERANGE;
	250	res = pagecache_read(inode, buf, size, pos);
	251	if (res)
	252	return res;
	253	}
	254	return size;
	255	}
	256
	257	static struct page f2fs_read_merkle_tree_page(struct inode inode,
fd39073d EB	258	pgoff_t index,
fd39073d EB	259	unsigned long num_ra_pages)
95ae251f	260	{
1ff61a32	261	struct folio *folio;
fd39073d	262
95ae251f EB	263	index += f2fs_verity_metadata_pos(inode) >> PAGE_SHIFT;
95ae251f EB	264
1ff61a32 H	265	folio = __filemap_get_folio(inode->i_mapping, index, FGP_ACCESSED, 0);
1ff61a32 H	266	if (IS_ERR(folio) \|\| !folio_test_uptodate(folio)) {
4fa0e3ff MWO	267	DEFINE_READAHEAD(ractl, NULL, NULL, inode->i_mapping, index);
4fa0e3ff MWO	268
1ff61a32 H	269	if (!IS_ERR(folio))
1ff61a32 H	270	folio_put(folio);
fd39073d	271	else if (num_ra_pages > 1)
73bb49da	272	page_cache_ra_unbounded(&ractl, num_ra_pages, 0);
1ff61a32 H	273	folio = read_mapping_folio(inode->i_mapping, index, NULL);
	274	if (IS_ERR(folio))
	275	return ERR_CAST(folio);
fd39073d	276	}
1ff61a32	277	return folio_file_page(folio, index);
95ae251f EB	278	}
	279
	280	static int f2fs_write_merkle_tree_block(struct inode inode, const void buf,
72ea15f0	281	u64 pos, unsigned int size)
95ae251f	282	{
72ea15f0	283	pos += f2fs_verity_metadata_pos(inode);
95ae251f	284
72ea15f0	285	return pagecache_write(inode, buf, size, pos);
95ae251f EB	286	}
	287
	288	const struct fsverity_operations f2fs_verityops = {
	289	.begin_enable_verity = f2fs_begin_enable_verity,
	290	.end_enable_verity = f2fs_end_enable_verity,
	291	.get_verity_descriptor = f2fs_get_verity_descriptor,
	292	.read_merkle_tree_page = f2fs_read_merkle_tree_page,
	293	.write_merkle_tree_block = f2fs_write_merkle_tree_block,
	294	};