Blame view
mm/page_io.c
11.5 KB
b24413180
|
1 |
// SPDX-License-Identifier: GPL-2.0 |
1da177e4c
|
2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
/* * 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> |
5a0e3ad6a
|
16 |
#include <linux/gfp.h> |
1da177e4c
|
17 18 19 20 |
#include <linux/pagemap.h> #include <linux/swap.h> #include <linux/bio.h> #include <linux/swapops.h> |
62c230bc1
|
21 |
#include <linux/buffer_head.h> |
1da177e4c
|
22 |
#include <linux/writeback.h> |
38b5faf4b
|
23 |
#include <linux/frontswap.h> |
b430e9d1c
|
24 |
#include <linux/blkdev.h> |
937790699
|
25 |
#include <linux/psi.h> |
e2e40f2c1
|
26 |
#include <linux/uio.h> |
b0ba2d0fa
|
27 |
#include <linux/sched/task.h> |
1da177e4c
|
28 |
|
f29ad6a99
|
29 |
static struct bio *get_swap_bio(gfp_t gfp_flags, |
1da177e4c
|
30 31 32 |
struct page *page, bio_end_io_t end_io) { struct bio *bio; |
1a5f439c7
|
33 |
bio = bio_alloc(gfp_flags, 1); |
1da177e4c
|
34 |
if (bio) { |
74d46992e
|
35 36 37 38 |
struct block_device *bdev; bio->bi_iter.bi_sector = map_swap_page(page, &bdev); bio_set_dev(bio, bdev); |
4f024f379
|
39 |
bio->bi_iter.bi_sector <<= PAGE_SHIFT - 9; |
1da177e4c
|
40 |
bio->bi_end_io = end_io; |
6cf66b4ca
|
41 |
|
af3bbc12d
|
42 |
bio_add_page(bio, page, thp_size(page), 0); |
1da177e4c
|
43 44 45 |
} return bio; } |
4246a0b63
|
46 |
void end_swap_bio_write(struct bio *bio) |
1da177e4c
|
47 |
{ |
263663cd3
|
48 |
struct page *page = bio_first_page_all(bio); |
1da177e4c
|
49 |
|
4e4cbee93
|
50 |
if (bio->bi_status) { |
1da177e4c
|
51 |
SetPageError(page); |
6ddab3b9e
|
52 53 54 55 56 57 58 59 60 |
/* * 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); |
1170532bb
|
61 62 |
pr_alert("Write-error on swap-device (%u:%u:%llu) ", |
74d46992e
|
63 |
MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)), |
1170532bb
|
64 |
(unsigned long long)bio->bi_iter.bi_sector); |
6ddab3b9e
|
65 66 |
ClearPageReclaim(page); } |
1da177e4c
|
67 68 |
end_page_writeback(page); bio_put(bio); |
1da177e4c
|
69 |
} |
3f2b1a04f
|
70 71 72 73 |
static void swap_slot_free_notify(struct page *page) { struct swap_info_struct *sis; struct gendisk *disk; |
5df373e95
|
74 |
swp_entry_t entry; |
3f2b1a04f
|
75 76 77 78 79 80 81 82 83 84 85 |
/* * 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); |
7b37e2267
|
86 |
if (data_race(!(sis->flags & SWP_BLKDEV))) |
3f2b1a04f
|
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 |
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; |
5df373e95
|
106 107 |
entry.val = page_private(page); if (disk->fops->swap_slot_free_notify && __swap_count(entry) == 1) { |
3f2b1a04f
|
108 |
unsigned long offset; |
3f2b1a04f
|
109 110 111 112 113 114 115 |
offset = swp_offset(entry); SetPageDirty(page); disk->fops->swap_slot_free_notify(sis->bdev, offset); } } |
4246a0b63
|
116 |
static void end_swap_bio_read(struct bio *bio) |
1da177e4c
|
117 |
{ |
263663cd3
|
118 |
struct page *page = bio_first_page_all(bio); |
23955622f
|
119 |
struct task_struct *waiter = bio->bi_private; |
1da177e4c
|
120 |
|
4e4cbee93
|
121 |
if (bio->bi_status) { |
1da177e4c
|
122 123 |
SetPageError(page); ClearPageUptodate(page); |
1170532bb
|
124 125 |
pr_alert("Read-error on swap-device (%u:%u:%llu) ", |
74d46992e
|
126 |
MAJOR(bio_dev(bio)), MINOR(bio_dev(bio)), |
1170532bb
|
127 |
(unsigned long long)bio->bi_iter.bi_sector); |
b430e9d1c
|
128 |
goto out; |
1da177e4c
|
129 |
} |
b430e9d1c
|
130 131 |
SetPageUptodate(page); |
3f2b1a04f
|
132 |
swap_slot_free_notify(page); |
b430e9d1c
|
133 |
out: |
1da177e4c
|
134 |
unlock_page(page); |
23955622f
|
135 |
WRITE_ONCE(bio->bi_private, NULL); |
1da177e4c
|
136 |
bio_put(bio); |
875185309
|
137 138 139 140 |
if (waiter) { blk_wake_io_task(waiter); put_task_struct(waiter); } |
1da177e4c
|
141 |
} |
a509bc1a9
|
142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 |
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; /* |
4efaceb1c
|
162 |
* Map all the blocks into the extent tree. This code doesn't try |
a509bc1a9
|
163 164 165 166 167 168 169 170 171 |
* 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; |
7e4411bfe
|
172 |
cond_resched(); |
30460e1ea
|
173 174 175 |
first_block = probe_block; ret = bmap(inode, &first_block); if (ret || !first_block) |
a509bc1a9
|
176 177 178 179 180 181 182 183 184 185 186 187 188 |
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; |
30460e1ea
|
189 190 191 |
block = probe_block + block_in_page; ret = bmap(inode, &block); if (ret || !block) |
a509bc1a9
|
192 |
goto bad_bmap; |
30460e1ea
|
193 |
|
a509bc1a9
|
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 |
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: |
1170532bb
|
231 232 |
pr_err("swapon: swapfile has holes "); |
a509bc1a9
|
233 234 235 |
ret = -EINVAL; goto out; } |
1da177e4c
|
236 237 238 239 240 241 |
/* * 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) { |
2f772e6ca
|
242 |
int ret = 0; |
1da177e4c
|
243 |
|
a2c43eed8
|
244 |
if (try_to_free_swap(page)) { |
1da177e4c
|
245 246 247 |
unlock_page(page); goto out; } |
8a84802e2
|
248 249 250 251 252 253 254 255 256 257 |
/* * Arch code may have to preserve more data than just the page * contents, e.g. memory tags. */ ret = arch_prepare_to_swap(page); if (ret) { set_page_dirty(page); unlock_page(page); goto out; } |
165c8aed5
|
258 |
if (frontswap_store(page) == 0) { |
38b5faf4b
|
259 260 261 262 263 |
set_page_writeback(page); unlock_page(page); end_page_writeback(page); goto out; } |
1eec6702a
|
264 |
ret = __swap_writepage(page, wbc, end_swap_bio_write); |
2f772e6ca
|
265 266 267 |
out: return ret; } |
dd6bd0d9c
|
268 269 |
static sector_t swap_page_sector(struct page *page) { |
09cbfeaf1
|
270 |
return (sector_t)__page_file_index(page) << (PAGE_SHIFT - 9); |
dd6bd0d9c
|
271 |
} |
225311a46
|
272 273 274 275 276 277 |
static inline void count_swpout_vm_event(struct page *page) { #ifdef CONFIG_TRANSPARENT_HUGEPAGE if (unlikely(PageTransHuge(page))) count_vm_event(THP_SWPOUT); #endif |
6c357848b
|
278 |
count_vm_events(PSWPOUT, thp_nr_pages(page)); |
225311a46
|
279 |
} |
a18b9b159
|
280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 |
#if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP) static void bio_associate_blkg_from_page(struct bio *bio, struct page *page) { struct cgroup_subsys_state *css; if (!page->mem_cgroup) return; rcu_read_lock(); css = cgroup_e_css(page->mem_cgroup->css.cgroup, &io_cgrp_subsys); bio_associate_blkg_from_css(bio, css); rcu_read_unlock(); } #else #define bio_associate_blkg_from_page(bio, page) do { } while (0) #endif /* CONFIG_MEMCG && CONFIG_BLK_CGROUP */ |
1eec6702a
|
296 |
int __swap_writepage(struct page *page, struct writeback_control *wbc, |
4246a0b63
|
297 |
bio_end_io_t end_write_func) |
2f772e6ca
|
298 299 |
{ struct bio *bio; |
4e49ea4a3
|
300 |
int ret; |
2f772e6ca
|
301 |
struct swap_info_struct *sis = page_swap_info(page); |
62c230bc1
|
302 |
|
cc30c5d64
|
303 |
VM_BUG_ON_PAGE(!PageSwapCache(page), page); |
326463154
|
304 |
if (data_race(sis->flags & SWP_FS_OPS)) { |
62c230bc1
|
305 306 307 |
struct kiocb kiocb; struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; |
62a8067a7
|
308 309 310 311 312 |
struct bio_vec bv = { .bv_page = page, .bv_len = PAGE_SIZE, .bv_offset = 0 }; |
05afcb77e
|
313 |
struct iov_iter from; |
62c230bc1
|
314 |
|
aa563d7bc
|
315 |
iov_iter_bvec(&from, WRITE, &bv, 1, PAGE_SIZE); |
62c230bc1
|
316 317 |
init_sync_kiocb(&kiocb, swap_file); kiocb.ki_pos = page_file_offset(page); |
62c230bc1
|
318 |
|
0cdc444a6
|
319 |
set_page_writeback(page); |
62c230bc1
|
320 |
unlock_page(page); |
c8b8e32d7
|
321 |
ret = mapping->a_ops->direct_IO(&kiocb, &from); |
62c230bc1
|
322 323 324 |
if (ret == PAGE_SIZE) { count_vm_event(PSWPOUT); ret = 0; |
2d30d31ea
|
325 |
} else { |
0cdc444a6
|
326 327 328 329 330 331 332 333 334 335 |
/* * 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. */ |
2d30d31ea
|
336 |
set_page_dirty(page); |
0cdc444a6
|
337 |
ClearPageReclaim(page); |
1170532bb
|
338 339 340 |
pr_err_ratelimited("Write error on dio swapfile (%llu) ", page_file_offset(page)); |
62c230bc1
|
341 |
} |
0cdc444a6
|
342 |
end_page_writeback(page); |
62c230bc1
|
343 344 |
return ret; } |
dd6bd0d9c
|
345 346 |
ret = bdev_write_page(sis->bdev, swap_page_sector(page), page, wbc); if (!ret) { |
225311a46
|
347 |
count_swpout_vm_event(page); |
dd6bd0d9c
|
348 349 |
return 0; } |
1eec6702a
|
350 |
bio = get_swap_bio(GFP_NOIO, page, end_write_func); |
1da177e4c
|
351 352 353 |
if (bio == NULL) { set_page_dirty(page); unlock_page(page); |
548d9782b
|
354 |
return -ENOMEM; |
1da177e4c
|
355 |
} |
0d1e0c7cd
|
356 |
bio->bi_opf = REQ_OP_WRITE | REQ_SWAP | wbc_to_write_flags(wbc); |
6a7f6d86a
|
357 |
bio_associate_blkg_from_page(bio, page); |
225311a46
|
358 |
count_swpout_vm_event(page); |
1da177e4c
|
359 360 |
set_page_writeback(page); unlock_page(page); |
4e49ea4a3
|
361 |
submit_bio(bio); |
548d9782b
|
362 363 |
return 0; |
1da177e4c
|
364 |
} |
0bcac06f2
|
365 |
int swap_readpage(struct page *page, bool synchronous) |
1da177e4c
|
366 367 368 |
{ struct bio *bio; int ret = 0; |
62c230bc1
|
369 |
struct swap_info_struct *sis = page_swap_info(page); |
23955622f
|
370 |
blk_qc_t qc; |
74d46992e
|
371 |
struct gendisk *disk; |
937790699
|
372 |
unsigned long pflags; |
1da177e4c
|
373 |
|
0bcac06f2
|
374 |
VM_BUG_ON_PAGE(!PageSwapCache(page) && !synchronous, page); |
309381fea
|
375 376 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(PageUptodate(page), page); |
937790699
|
377 378 379 380 381 382 383 |
/* * Count submission time as memory stall. When the device is congested, * or the submitting cgroup IO-throttled, submission can be a * significant part of overall IO time. */ psi_memstall_enter(&pflags); |
165c8aed5
|
384 |
if (frontswap_load(page) == 0) { |
38b5faf4b
|
385 386 387 388 |
SetPageUptodate(page); unlock_page(page); goto out; } |
62c230bc1
|
389 |
|
326463154
|
390 |
if (data_race(sis->flags & SWP_FS_OPS)) { |
62c230bc1
|
391 392 393 394 395 396 |
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); |
937790699
|
397 |
goto out; |
62c230bc1
|
398 |
} |
5115db10a
|
399 400 401 402 403 404 405 |
if (sis->flags & SWP_SYNCHRONOUS_IO) { 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); } |
b06bad17c
|
406 |
|
5115db10a
|
407 408 409 |
count_vm_event(PSWPIN); goto out; } |
dd6bd0d9c
|
410 411 412 |
} ret = 0; |
f29ad6a99
|
413 |
bio = get_swap_bio(GFP_KERNEL, page, end_swap_bio_read); |
1da177e4c
|
414 415 416 417 418 |
if (bio == NULL) { unlock_page(page); ret = -ENOMEM; goto out; } |
74d46992e
|
419 |
disk = bio->bi_disk; |
b0ba2d0fa
|
420 421 422 423 |
/* * Keep this task valid during swap readpage because the oom killer may * attempt to access it in the page fault retry time check. */ |
95fe6c1a2
|
424 |
bio_set_op_attrs(bio, REQ_OP_READ, 0); |
875185309
|
425 |
if (synchronous) { |
b685a7350
|
426 |
bio->bi_opf |= REQ_HIPRI; |
875185309
|
427 428 429 |
get_task_struct(current); bio->bi_private = current; } |
f8891e5e1
|
430 |
count_vm_event(PSWPIN); |
23955622f
|
431 432 |
bio_get(bio); qc = submit_bio(bio); |
0bcac06f2
|
433 |
while (synchronous) { |
1ac5cd497
|
434 |
set_current_state(TASK_UNINTERRUPTIBLE); |
23955622f
|
435 436 |
if (!READ_ONCE(bio->bi_private)) break; |
0a1b8b87d
|
437 |
if (!blk_poll(disk->queue, qc, true)) |
0f190a7ab
|
438 |
blk_io_schedule(); |
23955622f
|
439 440 441 |
} __set_current_state(TASK_RUNNING); bio_put(bio); |
1da177e4c
|
442 |
out: |
937790699
|
443 |
psi_memstall_leave(&pflags); |
1da177e4c
|
444 445 |
return ret; } |
62c230bc1
|
446 447 448 449 |
int swap_set_page_dirty(struct page *page) { struct swap_info_struct *sis = page_swap_info(page); |
326463154
|
450 |
if (data_race(sis->flags & SWP_FS_OPS)) { |
62c230bc1
|
451 |
struct address_space *mapping = sis->swap_file->f_mapping; |
cc30c5d64
|
452 453 |
VM_BUG_ON_PAGE(!PageSwapCache(page), page); |
62c230bc1
|
454 455 456 457 458 |
return mapping->a_ops->set_page_dirty(page); } else { return __set_page_dirty_no_writeback(page); } } |