[mirror_ubuntu-artful-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)
{
	struct bio *bio;

	bio = bio_alloc(gfp_flags, 1);
	if (bio) {
		bio->bi_iter.bi_sector = map_swap_page(page, &bio->bi_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",
			 imajor(bio->bi_bdev->bd_inode),
			 iminor(bio->bi_bdev->bd_inode),
			 (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",
			 imajor(bio->bi_bdev->bd_inode),
			 iminor(bio->bi_bdev->bd_inode),
			 (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);
}

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 block_device *bdev;

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