[mirror_ubuntu-bionic-kernel.git] / mm / page_io.c

// SPDX-License-Identifier: GPL-2.0
/*
 *  linux/mm/page_io.c
 *
 *  Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 *
 *  Swap reorganised 29.12.95, 
 *  Asynchronous swapping added 30.12.95. Stephen Tweedie
 *  Removed race in async swapping. 14.4.1996. Bruno Haible
 *  Add swap of shared pages through the page cache. 20.2.1998. Stephen Tweedie
 *  Always use brw_page, life becomes simpler. 12 May 1998 Eric Biederman
 */

#include <linux/mm.h>
#include <linux/kernel_stat.h>
#include <linux/gfp.h>
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/bio.h>
#include <linux/swapops.h>
#include <linux/buffer_head.h>
#include <linux/writeback.h>
#include <linux/frontswap.h>
#include <linux/blkdev.h>
#include <linux/uio.h>
#include <linux/sched/task.h>
#include <asm/pgtable.h>

static struct bio *get_swap_bio(gfp_t gfp_flags,
				struct page *page, bio_end_io_t end_io)
{
	int i, nr = hpage_nr_pages(page);
	struct bio *bio;

	bio = bio_alloc(gfp_flags, nr);
	if (bio) {
		struct block_device *bdev;

		bio->bi_iter.bi_sector = map_swap_page(page, &bdev);
		bio_set_dev(bio, bdev);
		bio->bi_iter.bi_sector <<= PAGE_SHIFT - 9;
		bio->bi_end_io = end_io;

		for (i = 0; i < nr; i++)
			bio_add_page(bio, page + i, PAGE_SIZE, 0);
		VM_BUG_ON(bio->bi_iter.bi_size != PAGE_SIZE * nr);
	}
	return bio;
}

void end_swap_bio_write(struct bio *bio)
{
	struct page *page = bio->bi_io_vec[0].bv_page;

	if (bio->bi_status) {
		SetPageError(page);
		/*
		 * We failed to write the page out to swap-space.
		 * Re-dirty the page in order to avoid it being reclaimed.
		 * Also print a dire warning that things will go BAD (tm)
		 * very quickly.
		 *
		 * Also clear PG_reclaim to avoid rotate_reclaimable_page()
		 */
		set_page_dirty(page);
		pr_alert("Write-error on swap-device (%u:%u:%llu)\n",
			 MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
			 (unsigned long long)bio->bi_iter.bi_sector);
		ClearPageReclaim(page);
	}
	end_page_writeback(page);
	bio_put(bio);
}

static void swap_slot_free_notify(struct page *page)
{
	struct swap_info_struct *sis;
	struct gendisk *disk;

	/*
	 * There is no guarantee that the page is in swap cache - the software
	 * suspend code (at least) uses end_swap_bio_read() against a non-
	 * swapcache page.  So we must check PG_swapcache before proceeding with
	 * this optimization.
	 */
	if (unlikely(!PageSwapCache(page)))
		return;

	sis = page_swap_info(page);
	if (!(sis->flags & SWP_BLKDEV))
		return;

	/*
	 * The swap subsystem performs lazy swap slot freeing,
	 * expecting that the page will be swapped out again.
	 * So we can avoid an unnecessary write if the page
	 * isn't redirtied.
	 * This is good for real swap storage because we can
	 * reduce unnecessary I/O and enhance wear-leveling
	 * if an SSD is used as the as swap device.
	 * But if in-memory swap device (eg zram) is used,
	 * this causes a duplicated copy between uncompressed
	 * data in VM-owned memory and compressed data in
	 * zram-owned memory.  So let's free zram-owned memory
	 * and make the VM-owned decompressed page *dirty*,
	 * so the page should be swapped out somewhere again if
	 * we again wish to reclaim it.
	 */
	disk = sis->bdev->bd_disk;
	if (disk->fops->swap_slot_free_notify) {
		swp_entry_t entry;
		unsigned long offset;

		entry.val = page_private(page);
		offset = swp_offset(entry);

		SetPageDirty(page);
		disk->fops->swap_slot_free_notify(sis->bdev,
				offset);
	}
}

static void end_swap_bio_read(struct bio *bio)
{
	struct page *page = bio->bi_io_vec[0].bv_page;
	struct task_struct *waiter = bio->bi_private;

	if (bio->bi_status) {
		SetPageError(page);
		ClearPageUptodate(page);
		pr_alert("Read-error on swap-device (%u:%u:%llu)\n",
			 MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
			 (unsigned long long)bio->bi_iter.bi_sector);
		goto out;
	}

	SetPageUptodate(page);
	swap_slot_free_notify(page);
out:
	unlock_page(page);
	WRITE_ONCE(bio->bi_private, NULL);
	bio_put(bio);
	wake_up_process(waiter);
	put_task_struct(waiter);
}

int generic_swapfile_activate(struct swap_info_struct *sis,
				struct file *swap_file,
				sector_t *span)
{
	struct address_space *mapping = swap_file->f_mapping;
	struct inode *inode = mapping->host;
	unsigned blocks_per_page;
	unsigned long page_no;
	unsigned blkbits;
	sector_t probe_block;
	sector_t last_block;
	sector_t lowest_block = -1;
	sector_t highest_block = 0;
	int nr_extents = 0;
	int ret;

	blkbits = inode->i_blkbits;
	blocks_per_page = PAGE_SIZE >> blkbits;

	/*
	 * Map all the blocks into the extent list.  This code doesn't try
	 * to be very smart.
	 */
	probe_block = 0;
	page_no = 0;
	last_block = i_size_read(inode) >> blkbits;
	while ((probe_block + blocks_per_page) <= last_block &&
			page_no < sis->max) {
		unsigned block_in_page;
		sector_t first_block;

		cond_resched();

		first_block = bmap(inode, probe_block);
		if (first_block == 0)
			goto bad_bmap;

		/*
		 * It must be PAGE_SIZE aligned on-disk
		 */
		if (first_block & (blocks_per_page - 1)) {
			probe_block++;
			goto reprobe;
		}

		for (block_in_page = 1; block_in_page < blocks_per_page;
					block_in_page++) {
			sector_t block;

			block = bmap(inode, probe_block + block_in_page);
			if (block == 0)
				goto bad_bmap;
			if (block != first_block + block_in_page) {
				/* Discontiguity */
				probe_block++;
				goto reprobe;
			}
		}

		first_block >>= (PAGE_SHIFT - blkbits);
		if (page_no) {	/* exclude the header page */
			if (first_block < lowest_block)
				lowest_block = first_block;
			if (first_block > highest_block)
				highest_block = first_block;
		}

		/*
		 * We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks
		 */
		ret = add_swap_extent(sis, page_no, 1, first_block);
		if (ret < 0)
			goto out;
		nr_extents += ret;
		page_no++;
		probe_block += blocks_per_page;
reprobe:
		continue;
	}
	ret = nr_extents;
	*span = 1 + highest_block - lowest_block;
	if (page_no == 0)
		page_no = 1;	/* force Empty message */
	sis->max = page_no;
	sis->pages = page_no - 1;
	sis->highest_bit = page_no - 1;
out:
	return ret;
bad_bmap:
	pr_err("swapon: swapfile has holes\n");
	ret = -EINVAL;
	goto out;
}

/*
 * We may have stale swap cache pages in memory: notice
 * them here and get rid of the unnecessary final write.
 */
int swap_writepage(struct page *page, struct writeback_control *wbc)
{
	int ret = 0;

	if (try_to_free_swap(page)) {
		unlock_page(page);
		goto out;
	}
	if (frontswap_store(page) == 0) {
		set_page_writeback(page);
		unlock_page(page);
		end_page_writeback(page);
		goto out;
	}
	ret = __swap_writepage(page, wbc, end_swap_bio_write);
out:
	return ret;
}

static sector_t swap_page_sector(struct page *page)
{
	return (sector_t)__page_file_index(page) << (PAGE_SHIFT - 9);
}

static inline void count_swpout_vm_event(struct page *page)
{
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	if (unlikely(PageTransHuge(page)))
		count_vm_event(THP_SWPOUT);
#endif
	count_vm_events(PSWPOUT, hpage_nr_pages(page));
}

int __swap_writepage(struct page *page, struct writeback_control *wbc,
		bio_end_io_t end_write_func)
{
	struct bio *bio;
	int ret;
	struct swap_info_struct *sis = page_swap_info(page);

	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
	if (sis->flags & SWP_FILE) {
		struct kiocb kiocb;
		struct file *swap_file = sis->swap_file;
		struct address_space *mapping = swap_file->f_mapping;
		struct bio_vec bv = {
			.bv_page = page,
			.bv_len  = PAGE_SIZE,
			.bv_offset = 0
		};
		struct iov_iter from;

		iov_iter_bvec(&from, ITER_BVEC | WRITE, &bv, 1, PAGE_SIZE);
		init_sync_kiocb(&kiocb, swap_file);
		kiocb.ki_pos = page_file_offset(page);

		set_page_writeback(page);
		unlock_page(page);
		ret = mapping->a_ops->direct_IO(&kiocb, &from);
		if (ret == PAGE_SIZE) {
			count_vm_event(PSWPOUT);
			ret = 0;
		} else {
			/*
			 * In the case of swap-over-nfs, this can be a
			 * temporary failure if the system has limited
			 * memory for allocating transmit buffers.
			 * Mark the page dirty and avoid
			 * rotate_reclaimable_page but rate-limit the
			 * messages but do not flag PageError like
			 * the normal direct-to-bio case as it could
			 * be temporary.
			 */
			set_page_dirty(page);
			ClearPageReclaim(page);
			pr_err_ratelimited("Write error on dio swapfile (%llu)\n",
					   page_file_offset(page));
		}
		end_page_writeback(page);
		return ret;
	}

	ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc);
	if (!ret) {
		count_swpout_vm_event(page);
		return 0;
	}

	ret = 0;
	bio = get_swap_bio(GFP_NOIO, page, end_write_func);
	if (bio == NULL) {
		set_page_dirty(page);
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
	bio->bi_opf = REQ_OP_WRITE | wbc_to_write_flags(wbc);
	count_swpout_vm_event(page);
	set_page_writeback(page);
	unlock_page(page);
	submit_bio(bio);
out:
	return ret;
}

int swap_readpage(struct page *page, bool synchronous)
{
	struct bio *bio;
	int ret = 0;
	struct swap_info_struct *sis = page_swap_info(page);
	blk_qc_t qc;
	struct gendisk *disk;

	VM_BUG_ON_PAGE(!PageSwapCache(page) && !synchronous, page);
	VM_BUG_ON_PAGE(!PageLocked(page), page);
	VM_BUG_ON_PAGE(PageUptodate(page), page);
	if (frontswap_load(page) == 0) {
		SetPageUptodate(page);
		unlock_page(page);
		goto out;
	}

	if (sis->flags & SWP_FILE) {
		struct file *swap_file = sis->swap_file;
		struct address_space *mapping = swap_file->f_mapping;

		ret = mapping->a_ops->readpage(swap_file, page);
		if (!ret)
			count_vm_event(PSWPIN);
		return ret;
	}

	ret = bdev_read_page(sis->bdev, swap_page_sector(page), page);
	if (!ret) {
		if (trylock_page(page)) {
			swap_slot_free_notify(page);
			unlock_page(page);
		}

		count_vm_event(PSWPIN);
		return 0;
	}

	ret = 0;
	bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read);
	if (bio == NULL) {
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
	disk = bio->bi_disk;
	/*
	 * Keep this task valid during swap readpage because the oom killer may
	 * attempt to access it in the page fault retry time check.
	 */
	get_task_struct(current);
	bio->bi_private = current;
	bio_set_op_attrs(bio, REQ_OP_READ, 0);
	count_vm_event(PSWPIN);
	bio_get(bio);
	qc = submit_bio(bio);
	while (synchronous) {
		set_current_state(TASK_UNINTERRUPTIBLE);
		if (!READ_ONCE(bio->bi_private))
			break;

		if (!blk_poll(disk->queue, qc))
			break;
	}
	__set_current_state(TASK_RUNNING);
	bio_put(bio);

out:
	return ret;
}

int swap_set_page_dirty(struct page *page)
{
	struct swap_info_struct *sis = page_swap_info(page);

	if (sis->flags & SWP_FILE) {
		struct address_space *mapping = sis->swap_file->f_mapping;

		VM_BUG_ON_PAGE(!PageSwapCache(page), page);
		return mapping->a_ops->set_page_dirty(page);
	} else {
		return __set_page_dirty_no_writeback(page);
	}
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
	3	* linux/mm/page_io.c
	4	*
	5	* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
	6	*
	7	* Swap reorganised 29.12.95,
	8	* Asynchronous swapping added 30.12.95. Stephen Tweedie
	9	* Removed race in async swapping. 14.4.1996. Bruno Haible
	10	* Add swap of shared pages through the page cache. 20.2.1998. Stephen Tweedie
	11	* Always use brw_page, life becomes simpler. 12 May 1998 Eric Biederman
	12	*/
	13
	14	#include <linux/mm.h>
	15	#include <linux/kernel_stat.h>
5a0e3ad6	16	#include <linux/gfp.h>
1da177e4 LT	17	#include <linux/pagemap.h>
	18	#include <linux/swap.h>
	19	#include <linux/bio.h>
	20	#include <linux/swapops.h>
62c230bc	21	#include <linux/buffer_head.h>
1da177e4	22	#include <linux/writeback.h>
38b5faf4	23	#include <linux/frontswap.h>
b430e9d1	24	#include <linux/blkdev.h>
e2e40f2c	25	#include <linux/uio.h>
b0ba2d0f	26	#include <linux/sched/task.h>
1da177e4 LT	27	#include <asm/pgtable.h>
1da177e4 LT	28
f29ad6a9	29	static struct bio *get_swap_bio(gfp_t gfp_flags,
1da177e4 LT	30	struct page *page, bio_end_io_t end_io)
1da177e4 LT	31	{
225311a4	32	int i, nr = hpage_nr_pages(page);
1da177e4 LT	33	struct bio *bio;
1da177e4 LT	34
225311a4	35	bio = bio_alloc(gfp_flags, nr);
1da177e4	36	if (bio) {
74d46992 CH	37	struct block_device *bdev;
	38
	39	bio->bi_iter.bi_sector = map_swap_page(page, &bdev);
	40	bio_set_dev(bio, bdev);
4f024f37	41	bio->bi_iter.bi_sector <<= PAGE_SHIFT - 9;
1da177e4	42	bio->bi_end_io = end_io;
6cf66b4c	43
225311a4 HY	44	for (i = 0; i < nr; i++)
	45	bio_add_page(bio, page + i, PAGE_SIZE, 0);
	46	VM_BUG_ON(bio->bi_iter.bi_size != PAGE_SIZE * nr);
1da177e4 LT	47	}
	48	return bio;
	49	}
	50
4246a0b6	51	void end_swap_bio_write(struct bio *bio)
1da177e4	52	{
1da177e4 LT	53	struct page *page = bio->bi_io_vec[0].bv_page;
1da177e4 LT	54
4e4cbee9	55	if (bio->bi_status) {
1da177e4	56	SetPageError(page);
6ddab3b9 PZ	57	/*
	58	* We failed to write the page out to swap-space.
	59	* Re-dirty the page in order to avoid it being reclaimed.
	60	* Also print a dire warning that things will go BAD (tm)
	61	* very quickly.
	62	*
	63	* Also clear PG_reclaim to avoid rotate_reclaimable_page()
	64	*/
	65	set_page_dirty(page);
1170532b	66	pr_alert("Write-error on swap-device (%u:%u:%llu)\n",
74d46992	67	MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
1170532b	68	(unsigned long long)bio->bi_iter.bi_sector);
6ddab3b9 PZ	69	ClearPageReclaim(page);
6ddab3b9 PZ	70	}
1da177e4 LT	71	end_page_writeback(page);
1da177e4 LT	72	bio_put(bio);
1da177e4 LT	73	}
1da177e4 LT	74
3f2b1a04 MK	75	static void swap_slot_free_notify(struct page *page)
	76	{
	77	struct swap_info_struct *sis;
	78	struct gendisk *disk;
	79
	80	/*
	81	* There is no guarantee that the page is in swap cache - the software
	82	* suspend code (at least) uses end_swap_bio_read() against a non-
	83	* swapcache page. So we must check PG_swapcache before proceeding with
	84	* this optimization.
	85	*/
	86	if (unlikely(!PageSwapCache(page)))
	87	return;
	88
	89	sis = page_swap_info(page);
	90	if (!(sis->flags & SWP_BLKDEV))
	91	return;
	92
	93	/*
	94	* The swap subsystem performs lazy swap slot freeing,
	95	* expecting that the page will be swapped out again.
	96	* So we can avoid an unnecessary write if the page
	97	* isn't redirtied.
	98	* This is good for real swap storage because we can
	99	* reduce unnecessary I/O and enhance wear-leveling
	100	* if an SSD is used as the as swap device.
	101	* But if in-memory swap device (eg zram) is used,
	102	* this causes a duplicated copy between uncompressed
	103	* data in VM-owned memory and compressed data in
	104	* zram-owned memory. So let's free zram-owned memory
	105	* and make the VM-owned decompressed page dirty,
	106	* so the page should be swapped out somewhere again if
	107	* we again wish to reclaim it.
	108	*/
	109	disk = sis->bdev->bd_disk;
	110	if (disk->fops->swap_slot_free_notify) {
	111	swp_entry_t entry;
	112	unsigned long offset;
	113
	114	entry.val = page_private(page);
	115	offset = swp_offset(entry);
	116
	117	SetPageDirty(page);
	118	disk->fops->swap_slot_free_notify(sis->bdev,
	119	offset);
	120	}
	121	}
	122
4246a0b6	123	static void end_swap_bio_read(struct bio *bio)
1da177e4	124	{
1da177e4	125	struct page *page = bio->bi_io_vec[0].bv_page;
23955622	126	struct task_struct *waiter = bio->bi_private;
1da177e4	127
4e4cbee9	128	if (bio->bi_status) {
1da177e4 LT	129	SetPageError(page);
1da177e4 LT	130	ClearPageUptodate(page);
1170532b	131	pr_alert("Read-error on swap-device (%u:%u:%llu)\n",
74d46992	132	MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
1170532b	133	(unsigned long long)bio->bi_iter.bi_sector);
b430e9d1	134	goto out;
1da177e4	135	}
b430e9d1 MK	136
b430e9d1 MK	137	SetPageUptodate(page);
3f2b1a04	138	swap_slot_free_notify(page);
b430e9d1	139	out:
1da177e4	140	unlock_page(page);
23955622	141	WRITE_ONCE(bio->bi_private, NULL);
1da177e4	142	bio_put(bio);
23955622	143	wake_up_process(waiter);
b0ba2d0f	144	put_task_struct(waiter);
1da177e4 LT	145	}
1da177e4 LT	146
a509bc1a MG	147	int generic_swapfile_activate(struct swap_info_struct *sis,
	148	struct file *swap_file,
	149	sector_t *span)
	150	{
	151	struct address_space *mapping = swap_file->f_mapping;
	152	struct inode *inode = mapping->host;
	153	unsigned blocks_per_page;
	154	unsigned long page_no;
	155	unsigned blkbits;
	156	sector_t probe_block;
	157	sector_t last_block;
	158	sector_t lowest_block = -1;
	159	sector_t highest_block = 0;
	160	int nr_extents = 0;
	161	int ret;
	162
	163	blkbits = inode->i_blkbits;
	164	blocks_per_page = PAGE_SIZE >> blkbits;
	165
	166	/*
	167	* Map all the blocks into the extent list. This code doesn't try
	168	* to be very smart.
	169	*/
	170	probe_block = 0;
	171	page_no = 0;
	172	last_block = i_size_read(inode) >> blkbits;
	173	while ((probe_block + blocks_per_page) <= last_block &&
	174	page_no < sis->max) {
	175	unsigned block_in_page;
	176	sector_t first_block;
	177
7e4411bf MP	178	cond_resched();
7e4411bf MP	179
a509bc1a MG	180	first_block = bmap(inode, probe_block);
	181	if (first_block == 0)
	182	goto bad_bmap;
	183
	184	/*
	185	* It must be PAGE_SIZE aligned on-disk
	186	*/
	187	if (first_block & (blocks_per_page - 1)) {
	188	probe_block++;
	189	goto reprobe;
	190	}
	191
	192	for (block_in_page = 1; block_in_page < blocks_per_page;
	193	block_in_page++) {
	194	sector_t block;
	195
	196	block = bmap(inode, probe_block + block_in_page);
	197	if (block == 0)
	198	goto bad_bmap;
	199	if (block != first_block + block_in_page) {
	200	/* Discontiguity */
	201	probe_block++;
	202	goto reprobe;
	203	}
	204	}
	205
	206	first_block >>= (PAGE_SHIFT - blkbits);
	207	if (page_no) { /* exclude the header page */
	208	if (first_block < lowest_block)
	209	lowest_block = first_block;
	210	if (first_block > highest_block)
	211	highest_block = first_block;
	212	}
	213
	214	/*
	215	* We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks
	216	*/
	217	ret = add_swap_extent(sis, page_no, 1, first_block);
	218	if (ret < 0)
	219	goto out;
	220	nr_extents += ret;
	221	page_no++;
	222	probe_block += blocks_per_page;
	223	reprobe:
	224	continue;
	225	}
	226	ret = nr_extents;
	227	*span = 1 + highest_block - lowest_block;
	228	if (page_no == 0)
	229	page_no = 1; /* force Empty message */
	230	sis->max = page_no;
	231	sis->pages = page_no - 1;
	232	sis->highest_bit = page_no - 1;
	233	out:
	234	return ret;
	235	bad_bmap:
1170532b	236	pr_err("swapon: swapfile has holes\n");
a509bc1a MG	237	ret = -EINVAL;
	238	goto out;
	239	}
	240
1da177e4 LT	241	/*
	242	* We may have stale swap cache pages in memory: notice
	243	* them here and get rid of the unnecessary final write.
	244	*/
	245	int swap_writepage(struct page page, struct writeback_control wbc)
	246	{
2f772e6c	247	int ret = 0;
1da177e4	248
a2c43eed	249	if (try_to_free_swap(page)) {
1da177e4 LT	250	unlock_page(page);
	251	goto out;
	252	}
165c8aed	253	if (frontswap_store(page) == 0) {
38b5faf4 DM	254	set_page_writeback(page);
	255	unlock_page(page);
	256	end_page_writeback(page);
	257	goto out;
	258	}
1eec6702	259	ret = __swap_writepage(page, wbc, end_swap_bio_write);
2f772e6c SJ	260	out:
	261	return ret;
	262	}
	263
dd6bd0d9 MW	264	static sector_t swap_page_sector(struct page *page)
dd6bd0d9 MW	265	{
09cbfeaf	266	return (sector_t)__page_file_index(page) << (PAGE_SHIFT - 9);
dd6bd0d9 MW	267	}
dd6bd0d9 MW	268
225311a4 HY	269	static inline void count_swpout_vm_event(struct page *page)
	270	{
	271	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	272	if (unlikely(PageTransHuge(page)))
	273	count_vm_event(THP_SWPOUT);
	274	#endif
	275	count_vm_events(PSWPOUT, hpage_nr_pages(page));
	276	}
	277
1eec6702	278	int __swap_writepage(struct page page, struct writeback_control wbc,
4246a0b6	279	bio_end_io_t end_write_func)
2f772e6c SJ	280	{
2f772e6c SJ	281	struct bio *bio;
4e49ea4a	282	int ret;
2f772e6c	283	struct swap_info_struct *sis = page_swap_info(page);
62c230bc	284
cc30c5d6	285	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
62c230bc MG	286	if (sis->flags & SWP_FILE) {
	287	struct kiocb kiocb;
	288	struct file *swap_file = sis->swap_file;
	289	struct address_space *mapping = swap_file->f_mapping;
62a8067a AV	290	struct bio_vec bv = {
	291	.bv_page = page,
	292	.bv_len = PAGE_SIZE,
	293	.bv_offset = 0
	294	};
05afcb77	295	struct iov_iter from;
62c230bc	296
05afcb77	297	iov_iter_bvec(&from, ITER_BVEC \| WRITE, &bv, 1, PAGE_SIZE);
62c230bc MG	298	init_sync_kiocb(&kiocb, swap_file);
62c230bc MG	299	kiocb.ki_pos = page_file_offset(page);
62c230bc	300
0cdc444a	301	set_page_writeback(page);
62c230bc	302	unlock_page(page);
c8b8e32d	303	ret = mapping->a_ops->direct_IO(&kiocb, &from);
62c230bc MG	304	if (ret == PAGE_SIZE) {
	305	count_vm_event(PSWPOUT);
	306	ret = 0;
2d30d31e	307	} else {
0cdc444a MG	308	/*
	309	* In the case of swap-over-nfs, this can be a
	310	* temporary failure if the system has limited
	311	* memory for allocating transmit buffers.
	312	* Mark the page dirty and avoid
	313	* rotate_reclaimable_page but rate-limit the
	314	* messages but do not flag PageError like
	315	* the normal direct-to-bio case as it could
	316	* be temporary.
	317	*/
2d30d31e	318	set_page_dirty(page);
0cdc444a	319	ClearPageReclaim(page);
1170532b JP	320	pr_err_ratelimited("Write error on dio swapfile (%llu)\n",
1170532b JP	321	page_file_offset(page));
62c230bc	322	}
0cdc444a	323	end_page_writeback(page);
62c230bc MG	324	return ret;
	325	}
	326
dd6bd0d9 MW	327	ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc);
dd6bd0d9 MW	328	if (!ret) {
225311a4	329	count_swpout_vm_event(page);
dd6bd0d9 MW	330	return 0;
	331	}
	332
	333	ret = 0;
1eec6702	334	bio = get_swap_bio(GFP_NOIO, page, end_write_func);
1da177e4 LT	335	if (bio == NULL) {
	336	set_page_dirty(page);
	337	unlock_page(page);
	338	ret = -ENOMEM;
	339	goto out;
	340	}
7637241e	341	bio->bi_opf = REQ_OP_WRITE \| wbc_to_write_flags(wbc);
225311a4	342	count_swpout_vm_event(page);
1da177e4 LT	343	set_page_writeback(page);
1da177e4 LT	344	unlock_page(page);
4e49ea4a	345	submit_bio(bio);
1da177e4 LT	346	out:
	347	return ret;
	348	}
	349
0bcac06f	350	int swap_readpage(struct page *page, bool synchronous)
1da177e4 LT	351	{
	352	struct bio *bio;
	353	int ret = 0;
62c230bc	354	struct swap_info_struct *sis = page_swap_info(page);
23955622	355	blk_qc_t qc;
74d46992	356	struct gendisk *disk;
1da177e4	357
0bcac06f	358	VM_BUG_ON_PAGE(!PageSwapCache(page) && !synchronous, page);
309381fe SL	359	VM_BUG_ON_PAGE(!PageLocked(page), page);
309381fe SL	360	VM_BUG_ON_PAGE(PageUptodate(page), page);
165c8aed	361	if (frontswap_load(page) == 0) {
38b5faf4 DM	362	SetPageUptodate(page);
	363	unlock_page(page);
	364	goto out;
	365	}
62c230bc MG	366
	367	if (sis->flags & SWP_FILE) {
	368	struct file *swap_file = sis->swap_file;
	369	struct address_space *mapping = swap_file->f_mapping;
	370
	371	ret = mapping->a_ops->readpage(swap_file, page);
	372	if (!ret)
	373	count_vm_event(PSWPIN);
	374	return ret;
	375	}
	376
dd6bd0d9 MW	377	ret = bdev_read_page(sis->bdev, swap_page_sector(page), page);
dd6bd0d9 MW	378	if (!ret) {
b06bad17 MK	379	if (trylock_page(page)) {
	380	swap_slot_free_notify(page);
	381	unlock_page(page);
	382	}
	383
dd6bd0d9 MW	384	count_vm_event(PSWPIN);
	385	return 0;
	386	}
	387
	388	ret = 0;
f29ad6a9	389	bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read);
1da177e4 LT	390	if (bio == NULL) {
	391	unlock_page(page);
	392	ret = -ENOMEM;
	393	goto out;
	394	}
74d46992	395	disk = bio->bi_disk;
b0ba2d0f TH	396	/*
	397	* Keep this task valid during swap readpage because the oom killer may
	398	* attempt to access it in the page fault retry time check.
	399	*/
	400	get_task_struct(current);
23955622	401	bio->bi_private = current;
95fe6c1a	402	bio_set_op_attrs(bio, REQ_OP_READ, 0);
f8891e5e	403	count_vm_event(PSWPIN);
23955622 SL	404	bio_get(bio);
23955622 SL	405	qc = submit_bio(bio);
0bcac06f	406	while (synchronous) {
23955622 SL	407	set_current_state(TASK_UNINTERRUPTIBLE);
	408	if (!READ_ONCE(bio->bi_private))
	409	break;
	410
ea435e1b	411	if (!blk_poll(disk->queue, qc))
23955622 SL	412	break;
	413	}
	414	__set_current_state(TASK_RUNNING);
	415	bio_put(bio);
	416
1da177e4 LT	417	out:
	418	return ret;
	419	}
62c230bc MG	420
	421	int swap_set_page_dirty(struct page *page)
	422	{
	423	struct swap_info_struct *sis = page_swap_info(page);
	424
	425	if (sis->flags & SWP_FILE) {
	426	struct address_space *mapping = sis->swap_file->f_mapping;
cc30c5d6 AM	427
cc30c5d6 AM	428	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
62c230bc MG	429	return mapping->a_ops->set_page_dirty(page);
	430	} else {
	431	return __set_page_dirty_no_writeback(page);
	432	}
	433	}