[mirror_ubuntu-artful-kernel.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_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"

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)) {
		if (bio->bi_status) {
			fscrypt_release_ctx(bio->bi_private);
		} else {
			fscrypt_decrypt_bio_pages(bio->bi_private, bio);
			return;
		}
	}
	bio_for_each_segment_all(bv, bio, i) {
		struct page *page = bv->bv_page;

		if (!bio->bi_status) {
			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;
	sector_t last_block_in_bio = 0;

	struct inode *inode = mapping->host;
	const unsigned blkbits = inode->i_blkbits;
	const unsigned blocks_per_page = PAGE_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 (; nr_pages; 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,
				  readahead_gfp_mask(mapping)))
				goto next_page;
		}

		if (page_has_buffers(page))
			goto confused;

		block_in_file = (sector_t)page->index << (PAGE_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_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_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(bio);
			bio = NULL;
		}
		if (bio == NULL) {
			struct fscrypt_ctx *ctx = NULL;

			if (ext4_encrypted_inode(inode) &&
			    S_ISREG(inode->i_mode)) {
				ctx = fscrypt_get_ctx(inode, GFP_NOFS);
				if (IS_ERR(ctx))
					goto set_error_page;
			}
			bio = bio_alloc(GFP_KERNEL,
				min_t(int, nr_pages, BIO_MAX_PAGES));
			if (!bio) {
				if (ctx)
					fscrypt_release_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;
			bio_set_op_attrs(bio, REQ_OP_READ, 0);
		}

		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(bio);
			bio = NULL;
		} else
			last_block_in_bio = blocks[blocks_per_page - 1];
		goto next_page;
	confused:
		if (bio) {
			submit_bio(bio);
			bio = NULL;
		}
		if (!PageUptodate(page))
			block_read_full_page(page, ext4_get_block);
		else
			unlock_page(page);
	next_page:
		if (pages)
			put_page(page);
	}
	BUG_ON(pages && !list_empty(pages));
	if (bio)
		submit_bio(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:
ea1754a0	26	* the end-of-file on blocksize < PAGE_SIZE setups.
f64e02fe TT	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	static inline bool ext4_bio_encrypted(struct bio *bio)
	50	{
	51	#ifdef CONFIG_EXT4_FS_ENCRYPTION
	52	return unlikely(bio->bi_private != NULL);
	53	#else
	54	return false;
	55	#endif
	56	}
	57
f64e02fe TT	58	/*
	59	* I/O completion handler for multipage BIOs.
	60	*
	61	* The mpage code never puts partial pages into a BIO (except for end-of-file).
	62	* If a page does not map to a contiguous run of blocks then it simply falls
	63	* back to block_read_full_page().
	64	*
	65	* Why is this? If a page's completion depends on a number of different BIOs
	66	* which can complete in any order (or at the same time) then determining the
	67	* status of that page is hard. See end_buffer_async_read() for the details.
	68	* There is no point in duplicating all that complexity.
	69	*/
4246a0b6	70	static void mpage_end_io(struct bio *bio)
f64e02fe TT	71	{
	72	struct bio_vec *bv;
	73	int i;
	74
c9c7429c	75	if (ext4_bio_encrypted(bio)) {
4e4cbee9	76	if (bio->bi_status) {
a7550b30	77	fscrypt_release_ctx(bio->bi_private);
c9c7429c	78	} else {
a7550b30	79	fscrypt_decrypt_bio_pages(bio->bi_private, bio);
c9c7429c MH	80	return;
	81	}
	82	}
f64e02fe TT	83	bio_for_each_segment_all(bv, bio, i) {
	84	struct page *page = bv->bv_page;
	85
4e4cbee9	86	if (!bio->bi_status) {
f64e02fe TT	87	SetPageUptodate(page);
	88	} else {
	89	ClearPageUptodate(page);
	90	SetPageError(page);
	91	}
	92	unlock_page(page);
	93	}
	94
	95	bio_put(bio);
	96	}
	97
	98	int ext4_mpage_readpages(struct address_space *mapping,
	99	struct list_head pages, struct page page,
	100	unsigned nr_pages)
	101	{
	102	struct bio *bio = NULL;
f64e02fe TT	103	sector_t last_block_in_bio = 0;
	104
	105	struct inode *inode = mapping->host;
	106	const unsigned blkbits = inode->i_blkbits;
09cbfeaf	107	const unsigned blocks_per_page = PAGE_SIZE >> blkbits;
f64e02fe TT	108	const unsigned blocksize = 1 << blkbits;
	109	sector_t block_in_file;
	110	sector_t last_block;
	111	sector_t last_block_in_file;
	112	sector_t blocks[MAX_BUF_PER_PAGE];
	113	unsigned page_block;
	114	struct block_device *bdev = inode->i_sb->s_bdev;
	115	int length;
	116	unsigned relative_block = 0;
	117	struct ext4_map_blocks map;
	118
	119	map.m_pblk = 0;
	120	map.m_lblk = 0;
	121	map.m_len = 0;
	122	map.m_flags = 0;
	123
de9e9181	124	for (; nr_pages; nr_pages--) {
f64e02fe TT	125	int fully_mapped = 1;
	126	unsigned first_hole = blocks_per_page;
	127
	128	prefetchw(&page->flags);
	129	if (pages) {
	130	page = list_entry(pages->prev, struct page, lru);
	131	list_del(&page->lru);
063d99b4	132	if (add_to_page_cache_lru(page, mapping, page->index,
8a5c743e	133	readahead_gfp_mask(mapping)))
f64e02fe TT	134	goto next_page;
	135	}
	136
	137	if (page_has_buffers(page))
	138	goto confused;
	139
09cbfeaf	140	block_in_file = (sector_t)page->index << (PAGE_SHIFT - blkbits);
f64e02fe TT	141	last_block = block_in_file + nr_pages * blocks_per_page;
	142	last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits;
	143	if (last_block > last_block_in_file)
	144	last_block = last_block_in_file;
	145	page_block = 0;
	146
	147	/*
	148	* Map blocks using the previous result first.
	149	*/
	150	if ((map.m_flags & EXT4_MAP_MAPPED) &&
	151	block_in_file > map.m_lblk &&
	152	block_in_file < (map.m_lblk + map.m_len)) {
	153	unsigned map_offset = block_in_file - map.m_lblk;
	154	unsigned last = map.m_len - map_offset;
	155
	156	for (relative_block = 0; ; relative_block++) {
	157	if (relative_block == last) {
	158	/* needed? */
	159	map.m_flags &= ~EXT4_MAP_MAPPED;
	160	break;
	161	}
	162	if (page_block == blocks_per_page)
	163	break;
	164	blocks[page_block] = map.m_pblk + map_offset +
	165	relative_block;
	166	page_block++;
	167	block_in_file++;
	168	}
	169	}
	170
	171	/*
	172	* Then do more ext4_map_blocks() calls until we are
	173	* done with this page.
	174	*/
	175	while (page_block < blocks_per_page) {
	176	if (block_in_file < last_block) {
	177	map.m_lblk = block_in_file;
	178	map.m_len = last_block - block_in_file;
	179
	180	if (ext4_map_blocks(NULL, inode, &map, 0) < 0) {
	181	set_error_page:
	182	SetPageError(page);
	183	zero_user_segment(page, 0,
09cbfeaf	184	PAGE_SIZE);
f64e02fe TT	185	unlock_page(page);
	186	goto next_page;
	187	}
	188	}
	189	if ((map.m_flags & EXT4_MAP_MAPPED) == 0) {
	190	fully_mapped = 0;
	191	if (first_hole == blocks_per_page)
	192	first_hole = page_block;
	193	page_block++;
	194	block_in_file++;
	195	continue;
	196	}
	197	if (first_hole != blocks_per_page)
	198	goto confused; /* hole -> non-hole */
	199
	200	/* Contiguous blocks? */
	201	if (page_block && blocks[page_block-1] != map.m_pblk-1)
	202	goto confused;
	203	for (relative_block = 0; ; relative_block++) {
	204	if (relative_block == map.m_len) {
	205	/* needed? */
	206	map.m_flags &= ~EXT4_MAP_MAPPED;
	207	break;
	208	} else if (page_block == blocks_per_page)
	209	break;
	210	blocks[page_block] = map.m_pblk+relative_block;
	211	page_block++;
	212	block_in_file++;
	213	}
	214	}
	215	if (first_hole != blocks_per_page) {
	216	zero_user_segment(page, first_hole << blkbits,
09cbfeaf	217	PAGE_SIZE);
f64e02fe TT	218	if (first_hole == 0) {
	219	SetPageUptodate(page);
	220	unlock_page(page);
	221	goto next_page;
	222	}
	223	} else if (fully_mapped) {
	224	SetPageMappedToDisk(page);
	225	}
	226	if (fully_mapped && blocks_per_page == 1 &&
	227	!PageUptodate(page) && cleancache_get_page(page) == 0) {
	228	SetPageUptodate(page);
	229	goto confused;
	230	}
	231
	232	/*
	233	* This page will go to BIO. Do we need to send this
	234	* BIO off first?
	235	*/
	236	if (bio && (last_block_in_bio != blocks[0] - 1)) {
	237	submit_and_realloc:
4e49ea4a	238	submit_bio(bio);
f64e02fe TT	239	bio = NULL;
	240	}
	241	if (bio == NULL) {
a7550b30	242	struct fscrypt_ctx *ctx = NULL;
c9c7429c MH	243
	244	if (ext4_encrypted_inode(inode) &&
	245	S_ISREG(inode->i_mode)) {
a7550b30	246	ctx = fscrypt_get_ctx(inode, GFP_NOFS);
c9c7429c MH	247	if (IS_ERR(ctx))
	248	goto set_error_page;
	249	}
f64e02fe	250	bio = bio_alloc(GFP_KERNEL,
b54ffb73	251	min_t(int, nr_pages, BIO_MAX_PAGES));
c9c7429c MH	252	if (!bio) {
c9c7429c MH	253	if (ctx)
a7550b30	254	fscrypt_release_ctx(ctx);
f64e02fe	255	goto set_error_page;
c9c7429c	256	}
f64e02fe TT	257	bio->bi_bdev = bdev;
	258	bio->bi_iter.bi_sector = blocks[0] << (blkbits - 9);
	259	bio->bi_end_io = mpage_end_io;
c9c7429c	260	bio->bi_private = ctx;
95fe6c1a	261	bio_set_op_attrs(bio, REQ_OP_READ, 0);
f64e02fe TT	262	}
	263
	264	length = first_hole << blkbits;
	265	if (bio_add_page(bio, page, length, 0) < length)
	266	goto submit_and_realloc;
	267
	268	if (((map.m_flags & EXT4_MAP_BOUNDARY) &&
	269	(relative_block == map.m_len)) \|\|
	270	(first_hole != blocks_per_page)) {
4e49ea4a	271	submit_bio(bio);
f64e02fe TT	272	bio = NULL;
	273	} else
	274	last_block_in_bio = blocks[blocks_per_page - 1];
	275	goto next_page;
	276	confused:
	277	if (bio) {
4e49ea4a	278	submit_bio(bio);
f64e02fe TT	279	bio = NULL;
	280	}
	281	if (!PageUptodate(page))
	282	block_read_full_page(page, ext4_get_block);
	283	else
	284	unlock_page(page);
	285	next_page:
	286	if (pages)
09cbfeaf	287	put_page(page);
f64e02fe TT	288	}
	289	BUG_ON(pages && !list_empty(pages));
	290	if (bio)
4e49ea4a	291	submit_bio(bio);
f64e02fe TT	292	return 0;
f64e02fe TT	293	}