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