[linux.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/psi.h>
#include <linux/uio.h>
#include <linux/sched/task.h>

static struct bio *get_swap_bio(gfp_t gfp_flags,
				struct page *page, bio_end_io_t end_io)
{
	struct bio *bio;

	bio = bio_alloc(gfp_flags, 1);
	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;

		bio_add_page(bio, page, thp_size(page), 0);
	}
	return bio;
}

void end_swap_bio_write(struct bio *bio)
{
	struct page *page = bio_first_page_all(bio);

	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;
	swp_entry_t entry;

	/*
	 * 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 (data_race(!(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;
	entry.val = page_private(page);
	if (disk->fops->swap_slot_free_notify && __swap_count(entry) == 1) {
		unsigned long offset;

		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_first_page_all(bio);
	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);
	if (waiter) {
		blk_wake_io_task(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 tree.  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 = probe_block;
		ret = bmap(inode, &first_block);
		if (ret || !first_block)
			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 = probe_block + block_in_page;
			ret = bmap(inode, &block);
			if (ret || !block)
				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;
	}
	/*
	 * Arch code may have to preserve more data than just the page
	 * contents, e.g. memory tags.
	 */
	ret = arch_prepare_to_swap(page);
	if (ret) {
		set_page_dirty(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, thp_nr_pages(page));
}

#if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP)
static void bio_associate_blkg_from_page(struct bio *bio, struct page *page)
{
	struct cgroup_subsys_state *css;

	if (!page->mem_cgroup)
		return;

	rcu_read_lock();
	css = cgroup_e_css(page->mem_cgroup->css.cgroup, &io_cgrp_subsys);
	bio_associate_blkg_from_css(bio, css);
	rcu_read_unlock();
}
#else
#define bio_associate_blkg_from_page(bio, page)		do { } while (0)
#endif /* CONFIG_MEMCG && CONFIG_BLK_CGROUP */

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 (data_race(sis->flags & SWP_FS_OPS)) {
		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, 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 | REQ_SWAP | wbc_to_write_flags(wbc);
	bio_associate_blkg_from_page(bio, page);
	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;
	unsigned long pflags;

	VM_BUG_ON_PAGE(!PageSwapCache(page) && !synchronous, page);
	VM_BUG_ON_PAGE(!PageLocked(page), page);
	VM_BUG_ON_PAGE(PageUptodate(page), page);

	/*
	 * Count submission time as memory stall. When the device is congested,
	 * or the submitting cgroup IO-throttled, submission can be a
	 * significant part of overall IO time.
	 */
	psi_memstall_enter(&pflags);

	if (frontswap_load(page) == 0) {
		SetPageUptodate(page);
		unlock_page(page);
		goto out;
	}

	if (data_race(sis->flags & SWP_FS_OPS)) {
		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);
		goto out;
	}

	if (sis->flags & SWP_SYNCHRONOUS_IO) {
		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);
			goto out;
		}
	}

	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.
	 */
	bio_set_op_attrs(bio, REQ_OP_READ, 0);
	if (synchronous) {
		bio->bi_opf |= REQ_HIPRI;
		get_task_struct(current);
		bio->bi_private = current;
	}
	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, true))
			blk_io_schedule();
	}
	__set_current_state(TASK_RUNNING);
	bio_put(bio);

out:
	psi_memstall_leave(&pflags);
	return ret;
}

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

	if (data_race(sis->flags & SWP_FS_OPS)) {
		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>
93779069	25	#include <linux/psi.h>
e2e40f2c	26	#include <linux/uio.h>
b0ba2d0f	27	#include <linux/sched/task.h>
1da177e4	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)
	31	{
	32	struct bio *bio;
	33
1a5f439c	34	bio = bio_alloc(gfp_flags, 1);
1da177e4	35	if (bio) {
74d46992 CH	36	struct block_device *bdev;
	37
	38	bio->bi_iter.bi_sector = map_swap_page(page, &bdev);
	39	bio_set_dev(bio, bdev);
4f024f37	40	bio->bi_iter.bi_sector <<= PAGE_SHIFT - 9;
1da177e4	41	bio->bi_end_io = end_io;
6cf66b4c	42
af3bbc12	43	bio_add_page(bio, page, thp_size(page), 0);
1da177e4 LT	44	}
	45	return bio;
	46	}
	47
4246a0b6	48	void end_swap_bio_write(struct bio *bio)
1da177e4	49	{
263663cd	50	struct page *page = bio_first_page_all(bio);
1da177e4	51
4e4cbee9	52	if (bio->bi_status) {
1da177e4	53	SetPageError(page);
6ddab3b9 PZ	54	/*
	55	* We failed to write the page out to swap-space.
	56	* Re-dirty the page in order to avoid it being reclaimed.
	57	* Also print a dire warning that things will go BAD (tm)
	58	* very quickly.
	59	*
	60	* Also clear PG_reclaim to avoid rotate_reclaimable_page()
	61	*/
	62	set_page_dirty(page);
1170532b	63	pr_alert("Write-error on swap-device (%u:%u:%llu)\n",
74d46992	64	MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
1170532b	65	(unsigned long long)bio->bi_iter.bi_sector);
6ddab3b9 PZ	66	ClearPageReclaim(page);
6ddab3b9 PZ	67	}
1da177e4 LT	68	end_page_writeback(page);
1da177e4 LT	69	bio_put(bio);
1da177e4 LT	70	}
1da177e4 LT	71
3f2b1a04 MK	72	static void swap_slot_free_notify(struct page *page)
	73	{
	74	struct swap_info_struct *sis;
	75	struct gendisk *disk;
5df373e9	76	swp_entry_t entry;
3f2b1a04 MK	77
	78	/*
	79	* There is no guarantee that the page is in swap cache - the software
	80	* suspend code (at least) uses end_swap_bio_read() against a non-
	81	* swapcache page. So we must check PG_swapcache before proceeding with
	82	* this optimization.
	83	*/
	84	if (unlikely(!PageSwapCache(page)))
	85	return;
	86
	87	sis = page_swap_info(page);
7b37e226	88	if (data_race(!(sis->flags & SWP_BLKDEV)))
3f2b1a04 MK	89	return;
	90
	91	/*
	92	* The swap subsystem performs lazy swap slot freeing,
	93	* expecting that the page will be swapped out again.
	94	* So we can avoid an unnecessary write if the page
	95	* isn't redirtied.
	96	* This is good for real swap storage because we can
	97	* reduce unnecessary I/O and enhance wear-leveling
	98	* if an SSD is used as the as swap device.
	99	* But if in-memory swap device (eg zram) is used,
	100	* this causes a duplicated copy between uncompressed
	101	* data in VM-owned memory and compressed data in
	102	* zram-owned memory. So let's free zram-owned memory
	103	* and make the VM-owned decompressed page dirty,
	104	* so the page should be swapped out somewhere again if
	105	* we again wish to reclaim it.
	106	*/
	107	disk = sis->bdev->bd_disk;
5df373e9 VM	108	entry.val = page_private(page);
5df373e9 VM	109	if (disk->fops->swap_slot_free_notify && __swap_count(entry) == 1) {
3f2b1a04 MK	110	unsigned long offset;
3f2b1a04 MK	111
3f2b1a04 MK	112	offset = swp_offset(entry);
	113
	114	SetPageDirty(page);
	115	disk->fops->swap_slot_free_notify(sis->bdev,
	116	offset);
	117	}
	118	}
	119
4246a0b6	120	static void end_swap_bio_read(struct bio *bio)
1da177e4	121	{
263663cd	122	struct page *page = bio_first_page_all(bio);
23955622	123	struct task_struct *waiter = bio->bi_private;
1da177e4	124
4e4cbee9	125	if (bio->bi_status) {
1da177e4 LT	126	SetPageError(page);
1da177e4 LT	127	ClearPageUptodate(page);
1170532b	128	pr_alert("Read-error on swap-device (%u:%u:%llu)\n",
74d46992	129	MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)),
1170532b	130	(unsigned long long)bio->bi_iter.bi_sector);
b430e9d1	131	goto out;
1da177e4	132	}
b430e9d1 MK	133
b430e9d1 MK	134	SetPageUptodate(page);
3f2b1a04	135	swap_slot_free_notify(page);
b430e9d1	136	out:
1da177e4	137	unlock_page(page);
23955622	138	WRITE_ONCE(bio->bi_private, NULL);
1da177e4	139	bio_put(bio);
87518530 ON	140	if (waiter) {
	141	blk_wake_io_task(waiter);
	142	put_task_struct(waiter);
	143	}
1da177e4 LT	144	}
1da177e4 LT	145
a509bc1a MG	146	int generic_swapfile_activate(struct swap_info_struct *sis,
	147	struct file *swap_file,
	148	sector_t *span)
	149	{
	150	struct address_space *mapping = swap_file->f_mapping;
	151	struct inode *inode = mapping->host;
	152	unsigned blocks_per_page;
	153	unsigned long page_no;
	154	unsigned blkbits;
	155	sector_t probe_block;
	156	sector_t last_block;
	157	sector_t lowest_block = -1;
	158	sector_t highest_block = 0;
	159	int nr_extents = 0;
	160	int ret;
	161
	162	blkbits = inode->i_blkbits;
	163	blocks_per_page = PAGE_SIZE >> blkbits;
	164
	165	/*
4efaceb1	166	* Map all the blocks into the extent tree. This code doesn't try
a509bc1a MG	167	* to be very smart.
	168	*/
	169	probe_block = 0;
	170	page_no = 0;
	171	last_block = i_size_read(inode) >> blkbits;
	172	while ((probe_block + blocks_per_page) <= last_block &&
	173	page_no < sis->max) {
	174	unsigned block_in_page;
	175	sector_t first_block;
	176
7e4411bf MP	177	cond_resched();
7e4411bf MP	178
30460e1e CM	179	first_block = probe_block;
	180	ret = bmap(inode, &first_block);
	181	if (ret \|\| !first_block)
a509bc1a MG	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
30460e1e CM	196	block = probe_block + block_in_page;
	197	ret = bmap(inode, &block);
	198	if (ret \|\| !block)
a509bc1a	199	goto bad_bmap;
30460e1e	200
a509bc1a MG	201	if (block != first_block + block_in_page) {
	202	/* Discontiguity */
	203	probe_block++;
	204	goto reprobe;
	205	}
	206	}
	207
	208	first_block >>= (PAGE_SHIFT - blkbits);
	209	if (page_no) { /* exclude the header page */
	210	if (first_block < lowest_block)
	211	lowest_block = first_block;
	212	if (first_block > highest_block)
	213	highest_block = first_block;
	214	}
	215
	216	/*
	217	* We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks
	218	*/
	219	ret = add_swap_extent(sis, page_no, 1, first_block);
	220	if (ret < 0)
	221	goto out;
	222	nr_extents += ret;
	223	page_no++;
	224	probe_block += blocks_per_page;
	225	reprobe:
	226	continue;
	227	}
	228	ret = nr_extents;
	229	*span = 1 + highest_block - lowest_block;
	230	if (page_no == 0)
	231	page_no = 1; /* force Empty message */
	232	sis->max = page_no;
	233	sis->pages = page_no - 1;
	234	sis->highest_bit = page_no - 1;
	235	out:
	236	return ret;
	237	bad_bmap:
1170532b	238	pr_err("swapon: swapfile has holes\n");
a509bc1a MG	239	ret = -EINVAL;
	240	goto out;
	241	}
	242
1da177e4 LT	243	/*
	244	* We may have stale swap cache pages in memory: notice
	245	* them here and get rid of the unnecessary final write.
	246	*/
	247	int swap_writepage(struct page page, struct writeback_control wbc)
	248	{
2f772e6c	249	int ret = 0;
1da177e4	250
a2c43eed	251	if (try_to_free_swap(page)) {
1da177e4 LT	252	unlock_page(page);
	253	goto out;
	254	}
8a84802e SP	255	/*
	256	* Arch code may have to preserve more data than just the page
	257	* contents, e.g. memory tags.
	258	*/
	259	ret = arch_prepare_to_swap(page);
	260	if (ret) {
	261	set_page_dirty(page);
	262	unlock_page(page);
	263	goto out;
	264	}
165c8aed	265	if (frontswap_store(page) == 0) {
38b5faf4 DM	266	set_page_writeback(page);
	267	unlock_page(page);
	268	end_page_writeback(page);
	269	goto out;
	270	}
1eec6702	271	ret = __swap_writepage(page, wbc, end_swap_bio_write);
2f772e6c SJ	272	out:
	273	return ret;
	274	}
	275
dd6bd0d9 MW	276	static sector_t swap_page_sector(struct page *page)
dd6bd0d9 MW	277	{
09cbfeaf	278	return (sector_t)__page_file_index(page) << (PAGE_SHIFT - 9);
dd6bd0d9 MW	279	}
dd6bd0d9 MW	280
225311a4 YH	281	static inline void count_swpout_vm_event(struct page *page)
	282	{
	283	#ifdef CONFIG_TRANSPARENT_HUGEPAGE
	284	if (unlikely(PageTransHuge(page)))
	285	count_vm_event(THP_SWPOUT);
	286	#endif
6c357848	287	count_vm_events(PSWPOUT, thp_nr_pages(page));
225311a4 YH	288	}
225311a4 YH	289
a18b9b15 CH	290	#if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP)
	291	static void bio_associate_blkg_from_page(struct bio bio, struct page page)
	292	{
	293	struct cgroup_subsys_state *css;
	294
	295	if (!page->mem_cgroup)
	296	return;
	297
	298	rcu_read_lock();
	299	css = cgroup_e_css(page->mem_cgroup->css.cgroup, &io_cgrp_subsys);
	300	bio_associate_blkg_from_css(bio, css);
	301	rcu_read_unlock();
	302	}
	303	#else
	304	#define bio_associate_blkg_from_page(bio, page) do { } while (0)
	305	#endif /* CONFIG_MEMCG && CONFIG_BLK_CGROUP */
	306
1eec6702	307	int __swap_writepage(struct page page, struct writeback_control wbc,
4246a0b6	308	bio_end_io_t end_write_func)
2f772e6c SJ	309	{
2f772e6c SJ	310	struct bio *bio;
4e49ea4a	311	int ret;
2f772e6c	312	struct swap_info_struct *sis = page_swap_info(page);
62c230bc	313
cc30c5d6	314	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
32646315	315	if (data_race(sis->flags & SWP_FS_OPS)) {
62c230bc MG	316	struct kiocb kiocb;
	317	struct file *swap_file = sis->swap_file;
	318	struct address_space *mapping = swap_file->f_mapping;
62a8067a AV	319	struct bio_vec bv = {
	320	.bv_page = page,
	321	.bv_len = PAGE_SIZE,
	322	.bv_offset = 0
	323	};
05afcb77	324	struct iov_iter from;
62c230bc	325
aa563d7b	326	iov_iter_bvec(&from, WRITE, &bv, 1, PAGE_SIZE);
62c230bc MG	327	init_sync_kiocb(&kiocb, swap_file);
62c230bc MG	328	kiocb.ki_pos = page_file_offset(page);
62c230bc	329
0cdc444a	330	set_page_writeback(page);
62c230bc	331	unlock_page(page);
c8b8e32d	332	ret = mapping->a_ops->direct_IO(&kiocb, &from);
62c230bc MG	333	if (ret == PAGE_SIZE) {
	334	count_vm_event(PSWPOUT);
	335	ret = 0;
2d30d31e	336	} else {
0cdc444a MG	337	/*
	338	* In the case of swap-over-nfs, this can be a
	339	* temporary failure if the system has limited
	340	* memory for allocating transmit buffers.
	341	* Mark the page dirty and avoid
	342	* rotate_reclaimable_page but rate-limit the
	343	* messages but do not flag PageError like
	344	* the normal direct-to-bio case as it could
	345	* be temporary.
	346	*/
2d30d31e	347	set_page_dirty(page);
0cdc444a	348	ClearPageReclaim(page);
1170532b JP	349	pr_err_ratelimited("Write error on dio swapfile (%llu)\n",
1170532b JP	350	page_file_offset(page));
62c230bc	351	}
0cdc444a	352	end_page_writeback(page);
62c230bc MG	353	return ret;
	354	}
	355
dd6bd0d9 MW	356	ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc);
dd6bd0d9 MW	357	if (!ret) {
225311a4	358	count_swpout_vm_event(page);
dd6bd0d9 MW	359	return 0;
	360	}
	361
	362	ret = 0;
1eec6702	363	bio = get_swap_bio(GFP_NOIO, page, end_write_func);
1da177e4 LT	364	if (bio == NULL) {
	365	set_page_dirty(page);
	366	unlock_page(page);
	367	ret = -ENOMEM;
	368	goto out;
	369	}
0d1e0c7c	370	bio->bi_opf = REQ_OP_WRITE \| REQ_SWAP \| wbc_to_write_flags(wbc);
6a7f6d86	371	bio_associate_blkg_from_page(bio, page);
225311a4	372	count_swpout_vm_event(page);
1da177e4 LT	373	set_page_writeback(page);
1da177e4 LT	374	unlock_page(page);
4e49ea4a	375	submit_bio(bio);
1da177e4 LT	376	out:
	377	return ret;
	378	}
	379
0bcac06f	380	int swap_readpage(struct page *page, bool synchronous)
1da177e4 LT	381	{
	382	struct bio *bio;
	383	int ret = 0;
62c230bc	384	struct swap_info_struct *sis = page_swap_info(page);
23955622	385	blk_qc_t qc;
74d46992	386	struct gendisk *disk;
93779069	387	unsigned long pflags;
1da177e4	388
0bcac06f	389	VM_BUG_ON_PAGE(!PageSwapCache(page) && !synchronous, page);
309381fe SL	390	VM_BUG_ON_PAGE(!PageLocked(page), page);
309381fe SL	391	VM_BUG_ON_PAGE(PageUptodate(page), page);
93779069 MK	392
	393	/*
	394	* Count submission time as memory stall. When the device is congested,
	395	* or the submitting cgroup IO-throttled, submission can be a
	396	* significant part of overall IO time.
	397	*/
	398	psi_memstall_enter(&pflags);
	399
165c8aed	400	if (frontswap_load(page) == 0) {
38b5faf4 DM	401	SetPageUptodate(page);
	402	unlock_page(page);
	403	goto out;
	404	}
62c230bc	405
32646315	406	if (data_race(sis->flags & SWP_FS_OPS)) {
62c230bc MG	407	struct file *swap_file = sis->swap_file;
	408	struct address_space *mapping = swap_file->f_mapping;
	409
	410	ret = mapping->a_ops->readpage(swap_file, page);
	411	if (!ret)
	412	count_vm_event(PSWPIN);
93779069	413	goto out;
62c230bc MG	414	}
62c230bc MG	415
5115db10 CH	416	if (sis->flags & SWP_SYNCHRONOUS_IO) {
	417	ret = bdev_read_page(sis->bdev, swap_page_sector(page), page);
	418	if (!ret) {
	419	if (trylock_page(page)) {
	420	swap_slot_free_notify(page);
	421	unlock_page(page);
	422	}
b06bad17	423
5115db10 CH	424	count_vm_event(PSWPIN);
	425	goto out;
	426	}
dd6bd0d9 MW	427	}
	428
	429	ret = 0;
f29ad6a9	430	bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read);
1da177e4 LT	431	if (bio == NULL) {
	432	unlock_page(page);
	433	ret = -ENOMEM;
	434	goto out;
	435	}
74d46992	436	disk = bio->bi_disk;
b0ba2d0f TH	437	/*
	438	* Keep this task valid during swap readpage because the oom killer may
	439	* attempt to access it in the page fault retry time check.
	440	*/
95fe6c1a	441	bio_set_op_attrs(bio, REQ_OP_READ, 0);
87518530	442	if (synchronous) {
b685a735	443	bio->bi_opf \|= REQ_HIPRI;
87518530 ON	444	get_task_struct(current);
	445	bio->bi_private = current;
	446	}
f8891e5e	447	count_vm_event(PSWPIN);
23955622 SL	448	bio_get(bio);
23955622 SL	449	qc = submit_bio(bio);
0bcac06f	450	while (synchronous) {
1ac5cd49	451	set_current_state(TASK_UNINTERRUPTIBLE);
23955622 SL	452	if (!READ_ONCE(bio->bi_private))
	453	break;
	454
0a1b8b87	455	if (!blk_poll(disk->queue, qc, true))
0f190a7a	456	blk_io_schedule();
23955622 SL	457	}
	458	__set_current_state(TASK_RUNNING);
	459	bio_put(bio);
	460
1da177e4	461	out:
93779069	462	psi_memstall_leave(&pflags);
1da177e4 LT	463	return ret;
1da177e4 LT	464	}
62c230bc MG	465
	466	int swap_set_page_dirty(struct page *page)
	467	{
	468	struct swap_info_struct *sis = page_swap_info(page);
	469
32646315	470	if (data_race(sis->flags & SWP_FS_OPS)) {
62c230bc	471	struct address_space *mapping = sis->swap_file->f_mapping;
cc30c5d6 AM	472
cc30c5d6 AM	473	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
62c230bc MG	474	return mapping->a_ops->set_page_dirty(page);
	475	} else {
	476	return __set_page_dirty_no_writeback(page);
	477	}
	478	}