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

/*
 *  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 <asm/pgtable.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, PAGE_SIZE, 0);
		BUG_ON(bio->bi_iter.bi_size != PAGE_SIZE);
	}
	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);
}

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);
}

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_vm_event(PSWPOUT);
		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_vm_event(PSWPOUT);
	set_page_writeback(page);
	unlock_page(page);
	submit_bio(bio);
out:
	return ret;
}

int swap_readpage(struct page *page, bool do_poll)
{
	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), 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;
	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 (do_poll) {
		set_current_state(TASK_UNINTERRUPTIBLE);
		if (!READ_ONCE(bio->bi_private))
			break;

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