page_io.c 9.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 *  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>
15
#include <linux/gfp.h>
L
Linus Torvalds 已提交
16 17 18 19
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/bio.h>
#include <linux/swapops.h>
20
#include <linux/buffer_head.h>
L
Linus Torvalds 已提交
21
#include <linux/writeback.h>
22
#include <linux/frontswap.h>
23
#include <linux/blkdev.h>
24
#include <linux/uio.h>
L
Linus Torvalds 已提交
25 26
#include <asm/pgtable.h>

H
Hugh Dickins 已提交
27
static struct bio *get_swap_bio(gfp_t gfp_flags,
L
Linus Torvalds 已提交
28 29 30 31 32 33
				struct page *page, bio_end_io_t end_io)
{
	struct bio *bio;

	bio = bio_alloc(gfp_flags, 1);
	if (bio) {
34 35
		bio->bi_iter.bi_sector = map_swap_page(page, &bio->bi_bdev);
		bio->bi_iter.bi_sector <<= PAGE_SHIFT - 9;
L
Linus Torvalds 已提交
36
		bio->bi_end_io = end_io;
37 38 39

		bio_add_page(bio, page, PAGE_SIZE, 0);
		BUG_ON(bio->bi_iter.bi_size != PAGE_SIZE);
L
Linus Torvalds 已提交
40 41 42 43
	}
	return bio;
}

44
void end_swap_bio_write(struct bio *bio)
L
Linus Torvalds 已提交
45 46 47
{
	struct page *page = bio->bi_io_vec[0].bv_page;

48
	if (bio->bi_error) {
L
Linus Torvalds 已提交
49
		SetPageError(page);
50 51 52 53 54 55 56 57 58
		/*
		 * 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);
59 60 61 62
		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);
63 64
		ClearPageReclaim(page);
	}
L
Linus Torvalds 已提交
65 66 67 68
	end_page_writeback(page);
	bio_put(bio);
}

69
static void end_swap_bio_read(struct bio *bio)
L
Linus Torvalds 已提交
70 71 72
{
	struct page *page = bio->bi_io_vec[0].bv_page;

73
	if (bio->bi_error) {
L
Linus Torvalds 已提交
74 75
		SetPageError(page);
		ClearPageUptodate(page);
76 77 78 79
		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);
80
		goto out;
L
Linus Torvalds 已提交
81
	}
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

	SetPageUptodate(page);

	/*
	 * 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 (likely(PageSwapCache(page))) {
		struct swap_info_struct *sis;

		sis = page_swap_info(page);
		if (sis->flags & SWP_BLKDEV) {
			/*
			 * 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.
			 */
			struct gendisk *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);
			}
		}
	}

out:
L
Linus Torvalds 已提交
128 129 130 131
	unlock_page(page);
	bio_put(bio);
}

132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218
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;

		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:
219
	pr_err("swapon: swapfile has holes\n");
220 221 222 223
	ret = -EINVAL;
	goto out;
}

L
Linus Torvalds 已提交
224 225 226 227 228 229
/*
 * 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)
{
230
	int ret = 0;
L
Linus Torvalds 已提交
231

232
	if (try_to_free_swap(page)) {
L
Linus Torvalds 已提交
233 234 235
		unlock_page(page);
		goto out;
	}
236
	if (frontswap_store(page) == 0) {
237 238 239 240 241
		set_page_writeback(page);
		unlock_page(page);
		end_page_writeback(page);
		goto out;
	}
242
	ret = __swap_writepage(page, wbc, end_swap_bio_write);
243 244 245 246
out:
	return ret;
}

247 248 249 250 251
static sector_t swap_page_sector(struct page *page)
{
	return (sector_t)__page_file_index(page) << (PAGE_CACHE_SHIFT - 9);
}

252
int __swap_writepage(struct page *page, struct writeback_control *wbc,
253
		bio_end_io_t end_write_func)
254 255
{
	struct bio *bio;
256
	int ret, rw = WRITE;
257
	struct swap_info_struct *sis = page_swap_info(page);
258 259 260 261 262

	if (sis->flags & SWP_FILE) {
		struct kiocb kiocb;
		struct file *swap_file = sis->swap_file;
		struct address_space *mapping = swap_file->f_mapping;
A
Al Viro 已提交
263 264 265 266 267
		struct bio_vec bv = {
			.bv_page = page,
			.bv_len  = PAGE_SIZE,
			.bv_offset = 0
		};
A
Al Viro 已提交
268
		struct iov_iter from;
269

A
Al Viro 已提交
270
		iov_iter_bvec(&from, ITER_BVEC | WRITE, &bv, 1, PAGE_SIZE);
271 272 273
		init_sync_kiocb(&kiocb, swap_file);
		kiocb.ki_pos = page_file_offset(page);

274
		set_page_writeback(page);
275
		unlock_page(page);
276
		ret = mapping->a_ops->direct_IO(&kiocb, &from, kiocb.ki_pos);
277 278 279
		if (ret == PAGE_SIZE) {
			count_vm_event(PSWPOUT);
			ret = 0;
280
		} else {
281 282 283 284 285 286 287 288 289 290
			/*
			 * 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.
			 */
291
			set_page_dirty(page);
292
			ClearPageReclaim(page);
293 294
			pr_err_ratelimited("Write error on dio swapfile (%llu)\n",
					   page_file_offset(page));
295
		}
296
		end_page_writeback(page);
297 298 299
		return ret;
	}

300 301 302 303 304 305 306
	ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc);
	if (!ret) {
		count_vm_event(PSWPOUT);
		return 0;
	}

	ret = 0;
307
	bio = get_swap_bio(GFP_NOIO, page, end_write_func);
L
Linus Torvalds 已提交
308 309 310 311 312 313 314
	if (bio == NULL) {
		set_page_dirty(page);
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
	if (wbc->sync_mode == WB_SYNC_ALL)
J
Jens Axboe 已提交
315
		rw |= REQ_SYNC;
316
	count_vm_event(PSWPOUT);
L
Linus Torvalds 已提交
317 318 319 320 321 322 323
	set_page_writeback(page);
	unlock_page(page);
	submit_bio(rw, bio);
out:
	return ret;
}

324
int swap_readpage(struct page *page)
L
Linus Torvalds 已提交
325 326 327
{
	struct bio *bio;
	int ret = 0;
328
	struct swap_info_struct *sis = page_swap_info(page);
L
Linus Torvalds 已提交
329

330 331
	VM_BUG_ON_PAGE(!PageLocked(page), page);
	VM_BUG_ON_PAGE(PageUptodate(page), page);
332
	if (frontswap_load(page) == 0) {
333 334 335 336
		SetPageUptodate(page);
		unlock_page(page);
		goto out;
	}
337 338 339 340 341 342 343 344 345 346 347

	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;
	}

348 349 350 351 352 353 354
	ret = bdev_read_page(sis->bdev, swap_page_sector(page), page);
	if (!ret) {
		count_vm_event(PSWPIN);
		return 0;
	}

	ret = 0;
H
Hugh Dickins 已提交
355
	bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read);
L
Linus Torvalds 已提交
356 357 358 359 360
	if (bio == NULL) {
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
361
	count_vm_event(PSWPIN);
L
Linus Torvalds 已提交
362 363 364 365
	submit_bio(READ, bio);
out:
	return ret;
}
366 367 368 369 370 371 372 373 374 375 376 377

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;
		return mapping->a_ops->set_page_dirty(page);
	} else {
		return __set_page_dirty_no_writeback(page);
	}
}