[linux.git] / fs / ext4 / readpage.c

/*
 * linux/fs/ext4/readpage.c
 *
 * Copyright (C) 2002, Linus Torvalds.
 * Copyright (C) 2015, Google, Inc.
 *
 * This was originally taken from fs/mpage.c
 *
 * The intent is the ext4_mpage_readpages() function here is intended
 * to replace mpage_readpages() in the general case, not just for
 * encrypted files.  It has some limitations (see below), where it
 * will fall back to read_block_full_page(), but these limitations
 * should only be hit when page_size != block_size.
 *
 * This will allow us to attach a callback function to support ext4
 * encryption.
 *
 * If anything unusual happens, such as:
 *
 * - encountering a page which has buffers
 * - encountering a page which has a non-hole after a hole
 * - encountering a page with non-contiguous blocks
 *
 * then this code just gives up and calls the buffer_head-based read function.
 * It does handle a page which has holes at the end - that is a common case:
 * the end-of-file on blocksize < PAGE_CACHE_SIZE setups.
 *
 */

#include <linux/kernel.h>
#include <linux/export.h>
#include <linux/mm.h>
#include <linux/kdev_t.h>
#include <linux/gfp.h>
#include <linux/bio.h>
#include <linux/fs.h>
#include <linux/buffer_head.h>
#include <linux/blkdev.h>
#include <linux/highmem.h>
#include <linux/prefetch.h>
#include <linux/mpage.h>
#include <linux/writeback.h>
#include <linux/backing-dev.h>
#include <linux/pagevec.h>
#include <linux/cleancache.h>

#include "ext4.h"

/*
 * Call ext4_decrypt on every single page, reusing the encryption
 * context.
 */
static void completion_pages(struct work_struct *work)
{
#ifdef CONFIG_EXT4_FS_ENCRYPTION
	struct ext4_crypto_ctx *ctx =
		container_of(work, struct ext4_crypto_ctx, r.work);
	struct bio	*bio	= ctx->r.bio;
	struct bio_vec	*bv;
	int		i;

	bio_for_each_segment_all(bv, bio, i) {
		struct page *page = bv->bv_page;

		int ret = ext4_decrypt(ctx, page);
		if (ret) {
			WARN_ON_ONCE(1);
			SetPageError(page);
		} else
			SetPageUptodate(page);
		unlock_page(page);
	}
	ext4_release_crypto_ctx(ctx);
	bio_put(bio);
#else
	BUG();
#endif
}

static inline bool ext4_bio_encrypted(struct bio *bio)
{
#ifdef CONFIG_EXT4_FS_ENCRYPTION
	return unlikely(bio->bi_private != NULL);
#else
	return false;
#endif
}

/*
 * I/O completion handler for multipage BIOs.
 *
 * The mpage code never puts partial pages into a BIO (except for end-of-file).
 * If a page does not map to a contiguous run of blocks then it simply falls
 * back to block_read_full_page().
 *
 * Why is this?  If a page's completion depends on a number of different BIOs
 * which can complete in any order (or at the same time) then determining the
 * status of that page is hard.  See end_buffer_async_read() for the details.
 * There is no point in duplicating all that complexity.
 */
static void mpage_end_io(struct bio *bio)
{
	struct bio_vec *bv;
	int i;

	if (ext4_bio_encrypted(bio)) {
		struct ext4_crypto_ctx *ctx = bio->bi_private;

		if (bio->bi_error) {
			ext4_release_crypto_ctx(ctx);
		} else {
			INIT_WORK(&ctx->r.work, completion_pages);
			ctx->r.bio = bio;
			queue_work(ext4_read_workqueue, &ctx->r.work);
			return;
		}
	}
	bio_for_each_segment_all(bv, bio, i) {
		struct page *page = bv->bv_page;

		if (!bio->bi_error) {
			SetPageUptodate(page);
		} else {
			ClearPageUptodate(page);
			SetPageError(page);
		}
		unlock_page(page);
	}

	bio_put(bio);
}

int ext4_mpage_readpages(struct address_space *mapping,
			 struct list_head *pages, struct page *page,
			 unsigned nr_pages)
{
	struct bio *bio = NULL;
	unsigned page_idx;
	sector_t last_block_in_bio = 0;

	struct inode *inode = mapping->host;
	const unsigned blkbits = inode->i_blkbits;
	const unsigned blocks_per_page = PAGE_CACHE_SIZE >> blkbits;
	const unsigned blocksize = 1 << blkbits;
	sector_t block_in_file;
	sector_t last_block;
	sector_t last_block_in_file;
	sector_t blocks[MAX_BUF_PER_PAGE];
	unsigned page_block;
	struct block_device *bdev = inode->i_sb->s_bdev;
	int length;
	unsigned relative_block = 0;
	struct ext4_map_blocks map;

	map.m_pblk = 0;
	map.m_lblk = 0;
	map.m_len = 0;
	map.m_flags = 0;

	for (page_idx = 0; nr_pages; page_idx++, nr_pages--) {
		int fully_mapped = 1;
		unsigned first_hole = blocks_per_page;

		prefetchw(&page->flags);
		if (pages) {
			page = list_entry(pages->prev, struct page, lru);
			list_del(&page->lru);
			if (add_to_page_cache_lru(page, mapping,
						  page->index, GFP_KERNEL))
				goto next_page;
		}

		if (page_has_buffers(page))
			goto confused;

		block_in_file = (sector_t)page->index << (PAGE_CACHE_SHIFT - blkbits);
		last_block = block_in_file + nr_pages * blocks_per_page;
		last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits;
		if (last_block > last_block_in_file)
			last_block = last_block_in_file;
		page_block = 0;

		/*
		 * Map blocks using the previous result first.
		 */
		if ((map.m_flags & EXT4_MAP_MAPPED) &&
		    block_in_file > map.m_lblk &&
		    block_in_file < (map.m_lblk + map.m_len)) {
			unsigned map_offset = block_in_file - map.m_lblk;
			unsigned last = map.m_len - map_offset;

			for (relative_block = 0; ; relative_block++) {
				if (relative_block == last) {
					/* needed? */
					map.m_flags &= ~EXT4_MAP_MAPPED;
					break;
				}
				if (page_block == blocks_per_page)
					break;
				blocks[page_block] = map.m_pblk + map_offset +
					relative_block;
				page_block++;
				block_in_file++;
			}
		}

		/*
		 * Then do more ext4_map_blocks() calls until we are
		 * done with this page.
		 */
		while (page_block < blocks_per_page) {
			if (block_in_file < last_block) {
				map.m_lblk = block_in_file;
				map.m_len = last_block - block_in_file;

				if (ext4_map_blocks(NULL, inode, &map, 0) < 0) {
				set_error_page:
					SetPageError(page);
					zero_user_segment(page, 0,
							  PAGE_CACHE_SIZE);
					unlock_page(page);
					goto next_page;
				}
			}
			if ((map.m_flags & EXT4_MAP_MAPPED) == 0) {
				fully_mapped = 0;
				if (first_hole == blocks_per_page)
					first_hole = page_block;
				page_block++;
				block_in_file++;
				continue;
			}
			if (first_hole != blocks_per_page)
				goto confused;		/* hole -> non-hole */

			/* Contiguous blocks? */
			if (page_block && blocks[page_block-1] != map.m_pblk-1)
				goto confused;
			for (relative_block = 0; ; relative_block++) {
				if (relative_block == map.m_len) {
					/* needed? */
					map.m_flags &= ~EXT4_MAP_MAPPED;
					break;
				} else if (page_block == blocks_per_page)
					break;
				blocks[page_block] = map.m_pblk+relative_block;
				page_block++;
				block_in_file++;
			}
		}
		if (first_hole != blocks_per_page) {
			zero_user_segment(page, first_hole << blkbits,
					  PAGE_CACHE_SIZE);
			if (first_hole == 0) {
				SetPageUptodate(page);
				unlock_page(page);
				goto next_page;
			}
		} else if (fully_mapped) {
			SetPageMappedToDisk(page);
		}
		if (fully_mapped && blocks_per_page == 1 &&
		    !PageUptodate(page) && cleancache_get_page(page) == 0) {
			SetPageUptodate(page);
			goto confused;
		}

		/*
		 * This page will go to BIO.  Do we need to send this
		 * BIO off first?
		 */
		if (bio && (last_block_in_bio != blocks[0] - 1)) {
		submit_and_realloc:
			submit_bio(READ, bio);
			bio = NULL;
		}
		if (bio == NULL) {
			struct ext4_crypto_ctx *ctx = NULL;

			if (ext4_encrypted_inode(inode) &&
			    S_ISREG(inode->i_mode)) {
				ctx = ext4_get_crypto_ctx(inode);
				if (IS_ERR(ctx))
					goto set_error_page;
			}
			bio = bio_alloc(GFP_KERNEL,
				min_t(int, nr_pages, bio_get_nr_vecs(bdev)));
			if (!bio) {
				if (ctx)
					ext4_release_crypto_ctx(ctx);
				goto set_error_page;
			}
			bio->bi_bdev = bdev;
			bio->bi_iter.bi_sector = blocks[0] << (blkbits - 9);
			bio->bi_end_io = mpage_end_io;
			bio->bi_private = ctx;
		}

		length = first_hole << blkbits;
		if (bio_add_page(bio, page, length, 0) < length)
			goto submit_and_realloc;

		if (((map.m_flags & EXT4_MAP_BOUNDARY) &&
		     (relative_block == map.m_len)) ||
		    (first_hole != blocks_per_page)) {
			submit_bio(READ, bio);
			bio = NULL;
		} else
			last_block_in_bio = blocks[blocks_per_page - 1];
		goto next_page;
	confused:
		if (bio) {
			submit_bio(READ, bio);
			bio = NULL;
		}
		if (!PageUptodate(page))
			block_read_full_page(page, ext4_get_block);
		else
			unlock_page(page);
	next_page:
		if (pages)
			page_cache_release(page);
	}
	BUG_ON(pages && !list_empty(pages));
	if (bio)
		submit_bio(READ, bio);
	return 0;
}
Commit	Line	Data
f64e02fe TT	1	/*
	2	* linux/fs/ext4/readpage.c
	3	*
	4	* Copyright (C) 2002, Linus Torvalds.
	5	* Copyright (C) 2015, Google, Inc.
	6	*
	7	* This was originally taken from fs/mpage.c
	8	*
	9	* The intent is the ext4_mpage_readpages() function here is intended
	10	* to replace mpage_readpages() in the general case, not just for
	11	* encrypted files. It has some limitations (see below), where it
	12	* will fall back to read_block_full_page(), but these limitations
	13	* should only be hit when page_size != block_size.
	14	*
	15	* This will allow us to attach a callback function to support ext4
	16	* encryption.
	17	*
	18	* If anything unusual happens, such as:
	19	*
	20	* - encountering a page which has buffers
	21	* - encountering a page which has a non-hole after a hole
	22	* - encountering a page with non-contiguous blocks
	23	*
	24	* then this code just gives up and calls the buffer_head-based read function.
	25	* It does handle a page which has holes at the end - that is a common case:
	26	* the end-of-file on blocksize < PAGE_CACHE_SIZE setups.
	27	*
	28	*/
	29
	30	#include <linux/kernel.h>
	31	#include <linux/export.h>
	32	#include <linux/mm.h>
	33	#include <linux/kdev_t.h>
	34	#include <linux/gfp.h>
	35	#include <linux/bio.h>
	36	#include <linux/fs.h>
	37	#include <linux/buffer_head.h>
	38	#include <linux/blkdev.h>
	39	#include <linux/highmem.h>
	40	#include <linux/prefetch.h>
	41	#include <linux/mpage.h>
	42	#include <linux/writeback.h>
	43	#include <linux/backing-dev.h>
	44	#include <linux/pagevec.h>
	45	#include <linux/cleancache.h>
	46
	47	#include "ext4.h"
	48
c9c7429c MH	49	/*
	50	* Call ext4_decrypt on every single page, reusing the encryption
	51	* context.
	52	*/
	53	static void completion_pages(struct work_struct *work)
	54	{
	55	#ifdef CONFIG_EXT4_FS_ENCRYPTION
	56	struct ext4_crypto_ctx *ctx =
614def70 TT	57	container_of(work, struct ext4_crypto_ctx, r.work);
614def70 TT	58	struct bio *bio = ctx->r.bio;
c9c7429c MH	59	struct bio_vec *bv;
	60	int i;
	61
	62	bio_for_each_segment_all(bv, bio, i) {
	63	struct page *page = bv->bv_page;
	64
	65	int ret = ext4_decrypt(ctx, page);
	66	if (ret) {
	67	WARN_ON_ONCE(1);
	68	SetPageError(page);
	69	} else
	70	SetPageUptodate(page);
	71	unlock_page(page);
	72	}
	73	ext4_release_crypto_ctx(ctx);
	74	bio_put(bio);
	75	#else
	76	BUG();
	77	#endif
	78	}
	79
	80	static inline bool ext4_bio_encrypted(struct bio *bio)
	81	{
	82	#ifdef CONFIG_EXT4_FS_ENCRYPTION
	83	return unlikely(bio->bi_private != NULL);
	84	#else
	85	return false;
	86	#endif
	87	}
	88
f64e02fe TT	89	/*
	90	* I/O completion handler for multipage BIOs.
	91	*
	92	* The mpage code never puts partial pages into a BIO (except for end-of-file).
	93	* If a page does not map to a contiguous run of blocks then it simply falls
	94	* back to block_read_full_page().
	95	*
	96	* Why is this? If a page's completion depends on a number of different BIOs
	97	* which can complete in any order (or at the same time) then determining the
	98	* status of that page is hard. See end_buffer_async_read() for the details.
	99	* There is no point in duplicating all that complexity.
	100	*/
4246a0b6	101	static void mpage_end_io(struct bio *bio)
f64e02fe TT	102	{
	103	struct bio_vec *bv;
	104	int i;
	105
c9c7429c MH	106	if (ext4_bio_encrypted(bio)) {
	107	struct ext4_crypto_ctx *ctx = bio->bi_private;
	108
4246a0b6	109	if (bio->bi_error) {
c9c7429c MH	110	ext4_release_crypto_ctx(ctx);
c9c7429c MH	111	} else {
614def70 TT	112	INIT_WORK(&ctx->r.work, completion_pages);
	113	ctx->r.bio = bio;
	114	queue_work(ext4_read_workqueue, &ctx->r.work);
c9c7429c MH	115	return;
	116	}
	117	}
f64e02fe TT	118	bio_for_each_segment_all(bv, bio, i) {
	119	struct page *page = bv->bv_page;
	120
4246a0b6	121	if (!bio->bi_error) {
f64e02fe TT	122	SetPageUptodate(page);
	123	} else {
	124	ClearPageUptodate(page);
	125	SetPageError(page);
	126	}
	127	unlock_page(page);
	128	}
	129
	130	bio_put(bio);
	131	}
	132
	133	int ext4_mpage_readpages(struct address_space *mapping,
	134	struct list_head pages, struct page page,
	135	unsigned nr_pages)
	136	{
	137	struct bio *bio = NULL;
	138	unsigned page_idx;
	139	sector_t last_block_in_bio = 0;
	140
	141	struct inode *inode = mapping->host;
	142	const unsigned blkbits = inode->i_blkbits;
	143	const unsigned blocks_per_page = PAGE_CACHE_SIZE >> blkbits;
	144	const unsigned blocksize = 1 << blkbits;
	145	sector_t block_in_file;
	146	sector_t last_block;
	147	sector_t last_block_in_file;
	148	sector_t blocks[MAX_BUF_PER_PAGE];
	149	unsigned page_block;
	150	struct block_device *bdev = inode->i_sb->s_bdev;
	151	int length;
	152	unsigned relative_block = 0;
	153	struct ext4_map_blocks map;
	154
	155	map.m_pblk = 0;
	156	map.m_lblk = 0;
	157	map.m_len = 0;
	158	map.m_flags = 0;
	159
	160	for (page_idx = 0; nr_pages; page_idx++, nr_pages--) {
	161	int fully_mapped = 1;
	162	unsigned first_hole = blocks_per_page;
	163
	164	prefetchw(&page->flags);
	165	if (pages) {
	166	page = list_entry(pages->prev, struct page, lru);
	167	list_del(&page->lru);
	168	if (add_to_page_cache_lru(page, mapping,
	169	page->index, GFP_KERNEL))
	170	goto next_page;
	171	}
	172
	173	if (page_has_buffers(page))
	174	goto confused;
	175
	176	block_in_file = (sector_t)page->index << (PAGE_CACHE_SHIFT - blkbits);
	177	last_block = block_in_file + nr_pages * blocks_per_page;
	178	last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits;
	179	if (last_block > last_block_in_file)
	180	last_block = last_block_in_file;
	181	page_block = 0;
	182
	183	/*
	184	* Map blocks using the previous result first.
	185	*/
186	if ((map.m_flags & EXT4_MAP_MAPPED) &&
187	block_in_file > map.m_lblk &&
188	block_in_file < (map.m_lblk + map.m_len)) {
189	unsigned map_offset = block_in_file - map.m_lblk;
190	unsigned last = map.m_len - map_offset;
191
192	for (relative_block = 0; ; relative_block++) {
193	if (relative_block == last) {
194	/* needed? */
195	map.m_flags &= ~EXT4_MAP_MAPPED;
196	break;
197	}
198	if (page_block == blocks_per_page)
199	break;
200	blocks[page_block] = map.m_pblk + map_offset +
201	relative_block;
202	page_block++;
203	block_in_file++;
204	}
205	}
206
207	/*
208	* Then do more ext4_map_blocks() calls until we are
209	* done with this page.
210	*/
211	while (page_block < blocks_per_page) {
212	if (block_in_file < last_block) {
213	map.m_lblk = block_in_file;
214	map.m_len = last_block - block_in_file;
215
216	if (ext4_map_blocks(NULL, inode, &map, 0) < 0) {
217	set_error_page:
218	SetPageError(page);
219	zero_user_segment(page, 0,
220	PAGE_CACHE_SIZE);
221	unlock_page(page);
222	goto next_page;
223	}
224	}
225	if ((map.m_flags & EXT4_MAP_MAPPED) == 0) {
226	fully_mapped = 0;
227	if (first_hole == blocks_per_page)
228	first_hole = page_block;
229	page_block++;
230	block_in_file++;
231	continue;
232	}
233	if (first_hole != blocks_per_page)
234	goto confused; /* hole -> non-hole */
235
236	/* Contiguous blocks? */
237	if (page_block && blocks[page_block-1] != map.m_pblk-1)
238	goto confused;
239	for (relative_block = 0; ; relative_block++) {
240	if (relative_block == map.m_len) {
241	/* needed? */
242	map.m_flags &= ~EXT4_MAP_MAPPED;
243	break;
244	} else if (page_block == blocks_per_page)
245	break;
246	blocks[page_block] = map.m_pblk+relative_block;
247	page_block++;
248	block_in_file++;
249	}
250	}
251	if (first_hole != blocks_per_page) {
252	zero_user_segment(page, first_hole << blkbits,
253	PAGE_CACHE_SIZE);
254	if (first_hole == 0) {
255	SetPageUptodate(page);
256	unlock_page(page);
257	goto next_page;
258	}
259	} else if (fully_mapped) {
260	SetPageMappedToDisk(page);
261	}
262	if (fully_mapped && blocks_per_page == 1 &&
263	!PageUptodate(page) && cleancache_get_page(page) == 0) {
264	SetPageUptodate(page);
265	goto confused;
266	}
267
268	/*
269	* This page will go to BIO. Do we need to send this
270	* BIO off first?
271	*/
272	if (bio && (last_block_in_bio != blocks[0] - 1)) {
273	submit_and_realloc:
274	submit_bio(READ, bio);
275	bio = NULL;
276	}
277	if (bio == NULL) {
c9c7429c MH	278	struct ext4_crypto_ctx *ctx = NULL;
	279
	280	if (ext4_encrypted_inode(inode) &&
	281	S_ISREG(inode->i_mode)) {
	282	ctx = ext4_get_crypto_ctx(inode);
	283	if (IS_ERR(ctx))
	284	goto set_error_page;
	285	}
f64e02fe TT	286	bio = bio_alloc(GFP_KERNEL,
f64e02fe TT	287	min_t(int, nr_pages, bio_get_nr_vecs(bdev)));
c9c7429c MH	288	if (!bio) {
	289	if (ctx)
	290	ext4_release_crypto_ctx(ctx);
f64e02fe	291	goto set_error_page;
c9c7429c	292	}
f64e02fe TT	293	bio->bi_bdev = bdev;
	294	bio->bi_iter.bi_sector = blocks[0] << (blkbits - 9);
	295	bio->bi_end_io = mpage_end_io;
c9c7429c	296	bio->bi_private = ctx;
f64e02fe TT	297	}
	298
	299	length = first_hole << blkbits;
	300	if (bio_add_page(bio, page, length, 0) < length)
	301	goto submit_and_realloc;
	302
	303	if (((map.m_flags & EXT4_MAP_BOUNDARY) &&
	304	(relative_block == map.m_len)) \|\|
	305	(first_hole != blocks_per_page)) {
	306	submit_bio(READ, bio);
	307	bio = NULL;
	308	} else
	309	last_block_in_bio = blocks[blocks_per_page - 1];
	310	goto next_page;
	311	confused:
	312	if (bio) {
	313	submit_bio(READ, bio);
	314	bio = NULL;
	315	}
	316	if (!PageUptodate(page))
	317	block_read_full_page(page, ext4_get_block);
	318	else
	319	unlock_page(page);
	320	next_page:
	321	if (pages)
	322	page_cache_release(page);
	323	}
	324	BUG_ON(pages && !list_empty(pages));
	325	if (bio)
	326	submit_bio(READ, bio);
	327	return 0;
	328	}