page_io.c 10.1 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>
25
#include <linux/sched/task.h>
L
Linus Torvalds 已提交
26 27
#include <asm/pgtable.h>

H
Hugh Dickins 已提交
28
static struct bio *get_swap_bio(gfp_t gfp_flags,
L
Linus Torvalds 已提交
29 30 31 32 33 34
				struct page *page, bio_end_io_t end_io)
{
	struct bio *bio;

	bio = bio_alloc(gfp_flags, 1);
	if (bio) {
35 36
		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
37
		bio->bi_end_io = end_io;
38 39 40

		bio_add_page(bio, page, PAGE_SIZE, 0);
		BUG_ON(bio->bi_iter.bi_size != PAGE_SIZE);
L
Linus Torvalds 已提交
41 42 43 44
	}
	return bio;
}

45
void end_swap_bio_write(struct bio *bio)
L
Linus Torvalds 已提交
46 47 48
{
	struct page *page = bio->bi_io_vec[0].bv_page;

49
	if (bio->bi_status) {
L
Linus Torvalds 已提交
50
		SetPageError(page);
51 52 53 54 55 56 57 58 59
		/*
		 * 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);
60 61 62 63
		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);
64 65
		ClearPageReclaim(page);
	}
L
Linus Torvalds 已提交
66 67 68 69
	end_page_writeback(page);
	bio_put(bio);
}

M
Minchan Kim 已提交
70 71 72 73 74 75 76 77 78 79 80 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
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);
	}
}

118
static void end_swap_bio_read(struct bio *bio)
L
Linus Torvalds 已提交
119 120
{
	struct page *page = bio->bi_io_vec[0].bv_page;
121
	struct task_struct *waiter = bio->bi_private;
L
Linus Torvalds 已提交
122

123
	if (bio->bi_status) {
L
Linus Torvalds 已提交
124 125
		SetPageError(page);
		ClearPageUptodate(page);
126 127 128 129
		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);
130
		goto out;
L
Linus Torvalds 已提交
131
	}
132 133

	SetPageUptodate(page);
M
Minchan Kim 已提交
134
	swap_slot_free_notify(page);
135
out:
L
Linus Torvalds 已提交
136
	unlock_page(page);
137
	WRITE_ONCE(bio->bi_private, NULL);
L
Linus Torvalds 已提交
138
	bio_put(bio);
139
	wake_up_process(waiter);
140
	put_task_struct(waiter);
L
Linus Torvalds 已提交
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
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;

174 175
		cond_resched();

176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
		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:
232
	pr_err("swapon: swapfile has holes\n");
233 234 235 236
	ret = -EINVAL;
	goto out;
}

L
Linus Torvalds 已提交
237 238 239 240 241 242
/*
 * 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)
{
243
	int ret = 0;
L
Linus Torvalds 已提交
244

245
	if (try_to_free_swap(page)) {
L
Linus Torvalds 已提交
246 247 248
		unlock_page(page);
		goto out;
	}
249
	if (frontswap_store(page) == 0) {
250 251 252 253 254
		set_page_writeback(page);
		unlock_page(page);
		end_page_writeback(page);
		goto out;
	}
255
	ret = __swap_writepage(page, wbc, end_swap_bio_write);
256 257 258 259
out:
	return ret;
}

260 261
static sector_t swap_page_sector(struct page *page)
{
262
	return (sector_t)__page_file_index(page) << (PAGE_SHIFT - 9);
263 264
}

265
int __swap_writepage(struct page *page, struct writeback_control *wbc,
266
		bio_end_io_t end_write_func)
267 268
{
	struct bio *bio;
269
	int ret;
270
	struct swap_info_struct *sis = page_swap_info(page);
271

272
	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
273 274 275 276
	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
277 278 279 280 281
		struct bio_vec bv = {
			.bv_page = page,
			.bv_len  = PAGE_SIZE,
			.bv_offset = 0
		};
A
Al Viro 已提交
282
		struct iov_iter from;
283

A
Al Viro 已提交
284
		iov_iter_bvec(&from, ITER_BVEC | WRITE, &bv, 1, PAGE_SIZE);
285 286 287
		init_sync_kiocb(&kiocb, swap_file);
		kiocb.ki_pos = page_file_offset(page);

288
		set_page_writeback(page);
289
		unlock_page(page);
290
		ret = mapping->a_ops->direct_IO(&kiocb, &from);
291 292 293
		if (ret == PAGE_SIZE) {
			count_vm_event(PSWPOUT);
			ret = 0;
294
		} else {
295 296 297 298 299 300 301 302 303 304
			/*
			 * 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.
			 */
305
			set_page_dirty(page);
306
			ClearPageReclaim(page);
307 308
			pr_err_ratelimited("Write error on dio swapfile (%llu)\n",
					   page_file_offset(page));
309
		}
310
		end_page_writeback(page);
311 312 313
		return ret;
	}

314 315 316 317 318 319 320
	ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc);
	if (!ret) {
		count_vm_event(PSWPOUT);
		return 0;
	}

	ret = 0;
321
	bio = get_swap_bio(GFP_NOIO, page, end_write_func);
L
Linus Torvalds 已提交
322 323 324 325 326 327
	if (bio == NULL) {
		set_page_dirty(page);
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
J
Jens Axboe 已提交
328
	bio->bi_opf = REQ_OP_WRITE | wbc_to_write_flags(wbc);
329
	count_vm_event(PSWPOUT);
L
Linus Torvalds 已提交
330 331
	set_page_writeback(page);
	unlock_page(page);
332
	submit_bio(bio);
L
Linus Torvalds 已提交
333 334 335 336
out:
	return ret;
}

337
int swap_readpage(struct page *page, bool do_poll)
L
Linus Torvalds 已提交
338 339 340
{
	struct bio *bio;
	int ret = 0;
341
	struct swap_info_struct *sis = page_swap_info(page);
342 343
	blk_qc_t qc;
	struct block_device *bdev;
L
Linus Torvalds 已提交
344

345
	VM_BUG_ON_PAGE(!PageSwapCache(page), page);
346 347
	VM_BUG_ON_PAGE(!PageLocked(page), page);
	VM_BUG_ON_PAGE(PageUptodate(page), page);
348
	if (frontswap_load(page) == 0) {
349 350 351 352
		SetPageUptodate(page);
		unlock_page(page);
		goto out;
	}
353 354 355 356 357 358 359 360 361 362 363

	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;
	}

364 365
	ret = bdev_read_page(sis->bdev, swap_page_sector(page), page);
	if (!ret) {
366 367 368 369 370
		if (trylock_page(page)) {
			swap_slot_free_notify(page);
			unlock_page(page);
		}

371 372 373 374 375
		count_vm_event(PSWPIN);
		return 0;
	}

	ret = 0;
H
Hugh Dickins 已提交
376
	bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read);
L
Linus Torvalds 已提交
377 378 379 380 381
	if (bio == NULL) {
		unlock_page(page);
		ret = -ENOMEM;
		goto out;
	}
382
	bdev = bio->bi_bdev;
383 384 385 386 387
	/*
	 * 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);
388
	bio->bi_private = current;
389
	bio_set_op_attrs(bio, REQ_OP_READ, 0);
390
	count_vm_event(PSWPIN);
391 392 393 394 395 396 397 398 399 400 401 402 403
	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);

L
Linus Torvalds 已提交
404 405 406
out:
	return ret;
}
407 408 409 410 411 412 413

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;
414 415

		VM_BUG_ON_PAGE(!PageSwapCache(page), page);
416 417 418 419 420
		return mapping->a_ops->set_page_dirty(page);
	} else {
		return __set_page_dirty_no_writeback(page);
	}
}