Blame view
mm/swap_state.c
21.5 KB
b24413180 License cleanup: ... |
1 |
// SPDX-License-Identifier: GPL-2.0 |
1da177e4c Linux-2.6.12-rc2 |
2 3 4 5 6 7 8 9 |
/* * linux/mm/swap_state.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds * Swap reorganised 29.12.95, Stephen Tweedie * * Rewritten to use page cache, (C) 1998 Stephen Tweedie */ |
1da177e4c Linux-2.6.12-rc2 |
10 |
#include <linux/mm.h> |
5a0e3ad6a include cleanup: ... |
11 |
#include <linux/gfp.h> |
1da177e4c Linux-2.6.12-rc2 |
12 13 |
#include <linux/kernel_stat.h> #include <linux/swap.h> |
46017e954 swapin_readahead:... |
14 |
#include <linux/swapops.h> |
1da177e4c Linux-2.6.12-rc2 |
15 16 |
#include <linux/init.h> #include <linux/pagemap.h> |
1da177e4c Linux-2.6.12-rc2 |
17 |
#include <linux/backing-dev.h> |
3fb5c298b swap: allow swap ... |
18 |
#include <linux/blkdev.h> |
c484d4104 [PATCH] mm: free_... |
19 |
#include <linux/pagevec.h> |
b20a35035 [PATCH] page migr... |
20 |
#include <linux/migrate.h> |
4b3ef9daa mm/swap: split sw... |
21 |
#include <linux/vmalloc.h> |
67afa38e0 mm/swap: add cach... |
22 |
#include <linux/swap_slots.h> |
38d8b4e6b mm, THP, swap: de... |
23 |
#include <linux/huge_mm.h> |
1da177e4c Linux-2.6.12-rc2 |
24 25 26 27 28 |
#include <asm/pgtable.h> /* * swapper_space is a fiction, retained to simplify the path through |
7eaceacca block: remove per... |
29 |
* vmscan's shrink_page_list. |
1da177e4c Linux-2.6.12-rc2 |
30 |
*/ |
f5e54d6e5 [PATCH] mark addr... |
31 |
static const struct address_space_operations swap_aops = { |
1da177e4c Linux-2.6.12-rc2 |
32 |
.writepage = swap_writepage, |
62c230bc1 mm: add support f... |
33 |
.set_page_dirty = swap_set_page_dirty, |
1c93923cc include/linux/mig... |
34 |
#ifdef CONFIG_MIGRATION |
e965f9630 [PATCH] Direct Mi... |
35 |
.migratepage = migrate_page, |
1c93923cc include/linux/mig... |
36 |
#endif |
1da177e4c Linux-2.6.12-rc2 |
37 |
}; |
4b3ef9daa mm/swap: split sw... |
38 39 |
struct address_space *swapper_spaces[MAX_SWAPFILES]; static unsigned int nr_swapper_spaces[MAX_SWAPFILES]; |
ec560175c mm, swap: VMA bas... |
40 |
bool swap_vma_readahead = true; |
ec560175c mm, swap: VMA bas... |
41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 |
#define SWAP_RA_WIN_SHIFT (PAGE_SHIFT / 2) #define SWAP_RA_HITS_MASK ((1UL << SWAP_RA_WIN_SHIFT) - 1) #define SWAP_RA_HITS_MAX SWAP_RA_HITS_MASK #define SWAP_RA_WIN_MASK (~PAGE_MASK & ~SWAP_RA_HITS_MASK) #define SWAP_RA_HITS(v) ((v) & SWAP_RA_HITS_MASK) #define SWAP_RA_WIN(v) (((v) & SWAP_RA_WIN_MASK) >> SWAP_RA_WIN_SHIFT) #define SWAP_RA_ADDR(v) ((v) & PAGE_MASK) #define SWAP_RA_VAL(addr, win, hits) \ (((addr) & PAGE_MASK) | \ (((win) << SWAP_RA_WIN_SHIFT) & SWAP_RA_WIN_MASK) | \ ((hits) & SWAP_RA_HITS_MASK)) /* Initial readahead hits is 4 to start up with a small window */ #define GET_SWAP_RA_VAL(vma) \ (atomic_long_read(&(vma)->swap_readahead_info) ? : 4) |
1da177e4c Linux-2.6.12-rc2 |
58 59 |
#define INC_CACHE_INFO(x) do { swap_cache_info.x++; } while (0) |
38d8b4e6b mm, THP, swap: de... |
60 |
#define ADD_CACHE_INFO(x, nr) do { swap_cache_info.x += (nr); } while (0) |
1da177e4c Linux-2.6.12-rc2 |
61 62 63 64 65 66 |
static struct { unsigned long add_total; unsigned long del_total; unsigned long find_success; unsigned long find_total; |
1da177e4c Linux-2.6.12-rc2 |
67 |
} swap_cache_info; |
33806f06d swap: make each s... |
68 69 |
unsigned long total_swapcache_pages(void) { |
4b3ef9daa mm/swap: split sw... |
70 |
unsigned int i, j, nr; |
33806f06d swap: make each s... |
71 |
unsigned long ret = 0; |
4b3ef9daa mm/swap: split sw... |
72 |
struct address_space *spaces; |
33806f06d swap: make each s... |
73 |
|
4b3ef9daa mm/swap: split sw... |
74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 |
rcu_read_lock(); for (i = 0; i < MAX_SWAPFILES; i++) { /* * The corresponding entries in nr_swapper_spaces and * swapper_spaces will be reused only after at least * one grace period. So it is impossible for them * belongs to different usage. */ nr = nr_swapper_spaces[i]; spaces = rcu_dereference(swapper_spaces[i]); if (!nr || !spaces) continue; for (j = 0; j < nr; j++) ret += spaces[j].nrpages; } rcu_read_unlock(); |
33806f06d swap: make each s... |
90 91 |
return ret; } |
579f82901 swap: add a simpl... |
92 |
static atomic_t swapin_readahead_hits = ATOMIC_INIT(4); |
1da177e4c Linux-2.6.12-rc2 |
93 94 |
void show_swap_cache_info(void) { |
33806f06d swap: make each s... |
95 96 |
printk("%lu pages in swap cache ", total_swapcache_pages()); |
2c97b7fc0 mm: print swapcac... |
97 98 |
printk("Swap cache stats: add %lu, delete %lu, find %lu/%lu ", |
1da177e4c Linux-2.6.12-rc2 |
99 |
swap_cache_info.add_total, swap_cache_info.del_total, |
bb63be0a0 tmpfs: move swap_... |
100 |
swap_cache_info.find_success, swap_cache_info.find_total); |
ec8acf20a swap: add per-par... |
101 102 103 |
printk("Free swap = %ldkB ", get_nr_swap_pages() << (PAGE_SHIFT - 10)); |
1da177e4c Linux-2.6.12-rc2 |
104 105 106 107 108 |
printk("Total swap = %lukB ", total_swap_pages << (PAGE_SHIFT - 10)); } /* |
31a563962 mm: add_to_swap_c... |
109 |
* __add_to_swap_cache resembles add_to_page_cache_locked on swapper_space, |
1da177e4c Linux-2.6.12-rc2 |
110 111 |
* but sets SwapCache flag and private instead of mapping and index. */ |
2f772e6ca mm: break up swap... |
112 |
int __add_to_swap_cache(struct page *page, swp_entry_t entry) |
1da177e4c Linux-2.6.12-rc2 |
113 |
{ |
38d8b4e6b mm, THP, swap: de... |
114 |
int error, i, nr = hpage_nr_pages(page); |
33806f06d swap: make each s... |
115 |
struct address_space *address_space; |
38d8b4e6b mm, THP, swap: de... |
116 |
pgoff_t idx = swp_offset(entry); |
1da177e4c Linux-2.6.12-rc2 |
117 |
|
309381fea mm: dump page whe... |
118 119 120 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(PageSwapCache(page), page); VM_BUG_ON_PAGE(!PageSwapBacked(page), page); |
51726b122 mm: replace some ... |
121 |
|
38d8b4e6b mm, THP, swap: de... |
122 |
page_ref_add(page, nr); |
31a563962 mm: add_to_swap_c... |
123 |
SetPageSwapCache(page); |
31a563962 mm: add_to_swap_c... |
124 |
|
33806f06d swap: make each s... |
125 126 |
address_space = swap_address_space(entry); spin_lock_irq(&address_space->tree_lock); |
38d8b4e6b mm, THP, swap: de... |
127 128 129 130 131 132 |
for (i = 0; i < nr; i++) { set_page_private(page + i, entry.val + i); error = radix_tree_insert(&address_space->page_tree, idx + i, page + i); if (unlikely(error)) break; |
31a563962 mm: add_to_swap_c... |
133 |
} |
38d8b4e6b mm, THP, swap: de... |
134 135 136 137 138 |
if (likely(!error)) { address_space->nrpages += nr; __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, nr); ADD_CACHE_INFO(add_total, nr); } else { |
2ca4532a4 mm: add_to_swap_c... |
139 140 141 142 143 144 |
/* * Only the context which have set SWAP_HAS_CACHE flag * would call add_to_swap_cache(). * So add_to_swap_cache() doesn't returns -EEXIST. */ VM_BUG_ON(error == -EEXIST); |
38d8b4e6b mm, THP, swap: de... |
145 146 147 148 149 |
set_page_private(page + i, 0UL); while (i--) { radix_tree_delete(&address_space->page_tree, idx + i); set_page_private(page + i, 0UL); } |
31a563962 mm: add_to_swap_c... |
150 |
ClearPageSwapCache(page); |
38d8b4e6b mm, THP, swap: de... |
151 |
page_ref_sub(page, nr); |
31a563962 mm: add_to_swap_c... |
152 |
} |
38d8b4e6b mm, THP, swap: de... |
153 |
spin_unlock_irq(&address_space->tree_lock); |
31a563962 mm: add_to_swap_c... |
154 155 156 157 158 159 160 161 |
return error; } int add_to_swap_cache(struct page *page, swp_entry_t entry, gfp_t gfp_mask) { int error; |
38d8b4e6b mm, THP, swap: de... |
162 |
error = radix_tree_maybe_preload_order(gfp_mask, compound_order(page)); |
35c754d79 memory controller... |
163 |
if (!error) { |
31a563962 mm: add_to_swap_c... |
164 |
error = __add_to_swap_cache(page, entry); |
1da177e4c Linux-2.6.12-rc2 |
165 |
radix_tree_preload_end(); |
fa1de9008 memcgroup: revert... |
166 |
} |
1da177e4c Linux-2.6.12-rc2 |
167 168 |
return error; } |
1da177e4c Linux-2.6.12-rc2 |
169 170 171 172 173 174 |
/* * This must be called only on pages that have * been verified to be in the swap cache. */ void __delete_from_swap_cache(struct page *page) { |
33806f06d swap: make each s... |
175 |
struct address_space *address_space; |
38d8b4e6b mm, THP, swap: de... |
176 177 178 |
int i, nr = hpage_nr_pages(page); swp_entry_t entry; pgoff_t idx; |
33806f06d swap: make each s... |
179 |
|
309381fea mm: dump page whe... |
180 181 182 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageSwapCache(page), page); VM_BUG_ON_PAGE(PageWriteback(page), page); |
1da177e4c Linux-2.6.12-rc2 |
183 |
|
33806f06d swap: make each s... |
184 185 |
entry.val = page_private(page); address_space = swap_address_space(entry); |
38d8b4e6b mm, THP, swap: de... |
186 187 188 189 190 |
idx = swp_offset(entry); for (i = 0; i < nr; i++) { radix_tree_delete(&address_space->page_tree, idx + i); set_page_private(page + i, 0); } |
1da177e4c Linux-2.6.12-rc2 |
191 |
ClearPageSwapCache(page); |
38d8b4e6b mm, THP, swap: de... |
192 193 194 |
address_space->nrpages -= nr; __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, -nr); ADD_CACHE_INFO(del_total, nr); |
1da177e4c Linux-2.6.12-rc2 |
195 196 197 198 199 200 201 202 203 |
} /** * add_to_swap - allocate swap space for a page * @page: page we want to move to swap * * Allocate swap space for the page and add the page to the * swap cache. Caller needs to hold the page lock. */ |
0f0746589 mm, THP, swap: mo... |
204 |
int add_to_swap(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
205 206 |
{ swp_entry_t entry; |
1da177e4c Linux-2.6.12-rc2 |
207 |
int err; |
309381fea mm: dump page whe... |
208 209 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageUptodate(page), page); |
1da177e4c Linux-2.6.12-rc2 |
210 |
|
38d8b4e6b mm, THP, swap: de... |
211 |
entry = get_swap_page(page); |
2ca4532a4 mm: add_to_swap_c... |
212 |
if (!entry.val) |
0f0746589 mm, THP, swap: mo... |
213 |
return 0; |
38d8b4e6b mm, THP, swap: de... |
214 |
if (mem_cgroup_try_charge_swap(page, entry)) |
0f0746589 mm, THP, swap: mo... |
215 |
goto fail; |
3f04f62f9 thp: split_huge_p... |
216 |
|
2ca4532a4 mm: add_to_swap_c... |
217 218 219 220 221 222 223 224 225 |
/* * Radix-tree node allocations from PF_MEMALLOC contexts could * completely exhaust the page allocator. __GFP_NOMEMALLOC * stops emergency reserves from being allocated. * * TODO: this could cause a theoretical memory reclaim * deadlock in the swap out path. */ /* |
854e9ed09 mm: support madvi... |
226 |
* Add it to the swap cache. |
2ca4532a4 mm: add_to_swap_c... |
227 228 229 |
*/ err = add_to_swap_cache(page, entry, __GFP_HIGH|__GFP_NOMEMALLOC|__GFP_NOWARN); |
38d8b4e6b mm, THP, swap: de... |
230 231 |
/* -ENOMEM radix-tree allocation failure */ if (err) |
bd53b714d [PATCH] mm: use _... |
232 |
/* |
2ca4532a4 mm: add_to_swap_c... |
233 234 |
* add_to_swap_cache() doesn't return -EEXIST, so we can safely * clear SWAP_HAS_CACHE flag. |
1da177e4c Linux-2.6.12-rc2 |
235 |
*/ |
0f0746589 mm, THP, swap: mo... |
236 |
goto fail; |
9625456cc mm: fix data corr... |
237 238 239 240 241 242 243 244 245 246 247 |
/* * Normally the page will be dirtied in unmap because its pte should be * dirty. A special case is MADV_FREE page. The page'e pte could have * dirty bit cleared but the page's SwapBacked bit is still set because * clearing the dirty bit and SwapBacked bit has no lock protected. For * such page, unmap will not set dirty bit for it, so page reclaim will * not write the page out. This can cause data corruption when the page * is swap in later. Always setting the dirty bit for the page solves * the problem. */ set_page_dirty(page); |
38d8b4e6b mm, THP, swap: de... |
248 249 |
return 1; |
38d8b4e6b mm, THP, swap: de... |
250 |
fail: |
0f0746589 mm, THP, swap: mo... |
251 |
put_swap_page(page, entry); |
38d8b4e6b mm, THP, swap: de... |
252 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
253 254 255 256 257 258 259 260 261 262 263 |
} /* * This must be called only on pages that have * been verified to be in the swap cache and locked. * It will never put the page into the free list, * the caller has a reference on the page. */ void delete_from_swap_cache(struct page *page) { swp_entry_t entry; |
33806f06d swap: make each s... |
264 |
struct address_space *address_space; |
1da177e4c Linux-2.6.12-rc2 |
265 |
|
4c21e2f24 [PATCH] mm: split... |
266 |
entry.val = page_private(page); |
1da177e4c Linux-2.6.12-rc2 |
267 |
|
33806f06d swap: make each s... |
268 269 |
address_space = swap_address_space(entry); spin_lock_irq(&address_space->tree_lock); |
1da177e4c Linux-2.6.12-rc2 |
270 |
__delete_from_swap_cache(page); |
33806f06d swap: make each s... |
271 |
spin_unlock_irq(&address_space->tree_lock); |
1da177e4c Linux-2.6.12-rc2 |
272 |
|
75f6d6d29 mm, THP, swap: un... |
273 |
put_swap_page(page, entry); |
38d8b4e6b mm, THP, swap: de... |
274 |
page_ref_sub(page, hpage_nr_pages(page)); |
1da177e4c Linux-2.6.12-rc2 |
275 |
} |
1da177e4c Linux-2.6.12-rc2 |
276 277 278 279 |
/* * If we are the only user, then try to free up the swap cache. * * Its ok to check for PageSwapCache without the page lock |
a2c43eed8 mm: try_to_free_s... |
280 281 |
* here because we are going to recheck again inside * try_to_free_swap() _with_ the lock. |
1da177e4c Linux-2.6.12-rc2 |
282 283 284 285 |
* - Marcelo */ static inline void free_swap_cache(struct page *page) { |
a2c43eed8 mm: try_to_free_s... |
286 287 |
if (PageSwapCache(page) && !page_mapped(page) && trylock_page(page)) { try_to_free_swap(page); |
1da177e4c Linux-2.6.12-rc2 |
288 289 290 291 292 293 |
unlock_page(page); } } /* * Perform a free_page(), also freeing any swap cache associated with |
b8072f099 [PATCH] mm: updat... |
294 |
* this page if it is the last user of the page. |
1da177e4c Linux-2.6.12-rc2 |
295 296 297 298 |
*/ void free_page_and_swap_cache(struct page *page) { free_swap_cache(page); |
6fcb52a56 thp: reduce usage... |
299 |
if (!is_huge_zero_page(page)) |
770a53702 mm: thp: broken p... |
300 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
301 302 303 304 305 306 307 308 |
} /* * Passed an array of pages, drop them all from swapcache and then release * them. They are removed from the LRU and freed if this is their last use. */ void free_pages_and_swap_cache(struct page **pages, int nr) { |
1da177e4c Linux-2.6.12-rc2 |
309 |
struct page **pagep = pages; |
aabfb5729 mm: memcontrol: d... |
310 |
int i; |
1da177e4c Linux-2.6.12-rc2 |
311 312 |
lru_add_drain(); |
aabfb5729 mm: memcontrol: d... |
313 314 315 |
for (i = 0; i < nr; i++) free_swap_cache(pagep[i]); release_pages(pagep, nr, false); |
1da177e4c Linux-2.6.12-rc2 |
316 317 318 319 320 321 322 323 |
} /* * Lookup a swap entry in the swap cache. A found page will be returned * unlocked and with its refcount incremented - we rely on the kernel * lock getting page table operations atomic even if we drop the page * lock before returning. */ |
ec560175c mm, swap: VMA bas... |
324 325 |
struct page *lookup_swap_cache(swp_entry_t entry, struct vm_area_struct *vma, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
326 327 |
{ struct page *page; |
ec560175c mm, swap: VMA bas... |
328 329 |
unsigned long ra_info; int win, hits, readahead; |
1da177e4c Linux-2.6.12-rc2 |
330 |
|
f6ab1f7f6 mm, swap: use off... |
331 |
page = find_get_page(swap_address_space(entry), swp_offset(entry)); |
1da177e4c Linux-2.6.12-rc2 |
332 |
|
ec560175c mm, swap: VMA bas... |
333 334 |
INC_CACHE_INFO(find_total); if (page) { |
1da177e4c Linux-2.6.12-rc2 |
335 |
INC_CACHE_INFO(find_success); |
ec560175c mm, swap: VMA bas... |
336 337 338 339 340 341 342 343 344 345 346 347 348 |
if (unlikely(PageTransCompound(page))) return page; readahead = TestClearPageReadahead(page); if (vma) { ra_info = GET_SWAP_RA_VAL(vma); win = SWAP_RA_WIN(ra_info); hits = SWAP_RA_HITS(ra_info); if (readahead) hits = min_t(int, hits + 1, SWAP_RA_HITS_MAX); atomic_long_set(&vma->swap_readahead_info, SWAP_RA_VAL(addr, win, hits)); } if (readahead) { |
cbc65df24 mm, swap: add swa... |
349 |
count_vm_event(SWAP_RA_HIT); |
ec560175c mm, swap: VMA bas... |
350 351 |
if (!vma) atomic_inc(&swapin_readahead_hits); |
cbc65df24 mm, swap: add swa... |
352 |
} |
579f82901 swap: add a simpl... |
353 |
} |
1da177e4c Linux-2.6.12-rc2 |
354 355 |
return page; } |
5b999aadb mm: swap: zswap: ... |
356 357 358 |
struct page *__read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, struct vm_area_struct *vma, unsigned long addr, bool *new_page_allocated) |
1da177e4c Linux-2.6.12-rc2 |
359 360 |
{ struct page *found_page, *new_page = NULL; |
5b999aadb mm: swap: zswap: ... |
361 |
struct address_space *swapper_space = swap_address_space(entry); |
1da177e4c Linux-2.6.12-rc2 |
362 |
int err; |
5b999aadb mm: swap: zswap: ... |
363 |
*new_page_allocated = false; |
1da177e4c Linux-2.6.12-rc2 |
364 365 366 367 368 369 370 |
do { /* * First check the swap cache. Since this is normally * called after lookup_swap_cache() failed, re-calling * that would confuse statistics. */ |
f6ab1f7f6 mm, swap: use off... |
371 |
found_page = find_get_page(swapper_space, swp_offset(entry)); |
1da177e4c Linux-2.6.12-rc2 |
372 373 |
if (found_page) break; |
ba81f8384 mm/swap: skip rea... |
374 375 376 377 378 379 380 381 382 383 |
/* * Just skip read ahead for unused swap slot. * During swap_off when swap_slot_cache is disabled, * we have to handle the race between putting * swap entry in swap cache and marking swap slot * as SWAP_HAS_CACHE. That's done in later part of code or * else swap_off will be aborted if we return NULL. */ if (!__swp_swapcount(entry) && swap_slot_cache_enabled) break; |
e8c26ab60 mm/swap: skip rea... |
384 |
|
1da177e4c Linux-2.6.12-rc2 |
385 386 387 388 |
/* * Get a new page to read into from swap. */ if (!new_page) { |
02098feaa swapin needs gfp_... |
389 |
new_page = alloc_page_vma(gfp_mask, vma, addr); |
1da177e4c Linux-2.6.12-rc2 |
390 391 392 393 394 |
if (!new_page) break; /* Out of memory */ } /* |
31a563962 mm: add_to_swap_c... |
395 396 |
* call radix_tree_preload() while we can wait. */ |
5e4c0d974 lib/radix-tree.c:... |
397 |
err = radix_tree_maybe_preload(gfp_mask & GFP_KERNEL); |
31a563962 mm: add_to_swap_c... |
398 399 400 401 |
if (err) break; /* |
f000944d0 tmpfs: shuffle ad... |
402 403 |
* Swap entry may have been freed since our caller observed it. */ |
355cfa73d mm: modify swap_m... |
404 |
err = swapcache_prepare(entry); |
cbab0e4ee swap: avoid read_... |
405 |
if (err == -EEXIST) { |
31a563962 mm: add_to_swap_c... |
406 |
radix_tree_preload_end(); |
cbab0e4ee swap: avoid read_... |
407 408 409 |
/* * We might race against get_swap_page() and stumble * across a SWAP_HAS_CACHE swap_map entry whose page |
9c1cc2e4f mm, swap: fix com... |
410 |
* has not been brought into the swapcache yet. |
cbab0e4ee swap: avoid read_... |
411 412 |
*/ cond_resched(); |
355cfa73d mm: modify swap_m... |
413 |
continue; |
31a563962 mm: add_to_swap_c... |
414 415 416 |
} if (err) { /* swp entry is obsolete ? */ radix_tree_preload_end(); |
f000944d0 tmpfs: shuffle ad... |
417 |
break; |
31a563962 mm: add_to_swap_c... |
418 |
} |
f000944d0 tmpfs: shuffle ad... |
419 |
|
2ca4532a4 mm: add_to_swap_c... |
420 |
/* May fail (-ENOMEM) if radix-tree node allocation failed. */ |
48c935ad8 page-flags: defin... |
421 |
__SetPageLocked(new_page); |
fa9949da5 mm: use __SetPage... |
422 |
__SetPageSwapBacked(new_page); |
31a563962 mm: add_to_swap_c... |
423 |
err = __add_to_swap_cache(new_page, entry); |
529ae9aaa mm: rename page t... |
424 |
if (likely(!err)) { |
31a563962 mm: add_to_swap_c... |
425 |
radix_tree_preload_end(); |
1da177e4c Linux-2.6.12-rc2 |
426 427 428 |
/* * Initiate read into locked page and return. */ |
c5fdae469 vmscan: add newly... |
429 |
lru_cache_add_anon(new_page); |
5b999aadb mm: swap: zswap: ... |
430 |
*new_page_allocated = true; |
1da177e4c Linux-2.6.12-rc2 |
431 432 |
return new_page; } |
31a563962 mm: add_to_swap_c... |
433 |
radix_tree_preload_end(); |
48c935ad8 page-flags: defin... |
434 |
__ClearPageLocked(new_page); |
2ca4532a4 mm: add_to_swap_c... |
435 436 437 438 |
/* * add_to_swap_cache() doesn't return -EEXIST, so we can safely * clear SWAP_HAS_CACHE flag. */ |
75f6d6d29 mm, THP, swap: un... |
439 |
put_swap_page(new_page, entry); |
f000944d0 tmpfs: shuffle ad... |
440 |
} while (err != -ENOMEM); |
1da177e4c Linux-2.6.12-rc2 |
441 442 |
if (new_page) |
09cbfeaf1 mm, fs: get rid o... |
443 |
put_page(new_page); |
1da177e4c Linux-2.6.12-rc2 |
444 445 |
return found_page; } |
46017e954 swapin_readahead:... |
446 |
|
5b999aadb mm: swap: zswap: ... |
447 448 449 450 451 452 453 |
/* * Locate a page of swap in physical memory, reserving swap cache space * and reading the disk if it is not already cached. * A failure return means that either the page allocation failed or that * the swap entry is no longer in use. */ struct page *read_swap_cache_async(swp_entry_t entry, gfp_t gfp_mask, |
23955622f swap: add block i... |
454 |
struct vm_area_struct *vma, unsigned long addr, bool do_poll) |
5b999aadb mm: swap: zswap: ... |
455 456 457 458 459 460 |
{ bool page_was_allocated; struct page *retpage = __read_swap_cache_async(entry, gfp_mask, vma, addr, &page_was_allocated); if (page_was_allocated) |
23955622f swap: add block i... |
461 |
swap_readpage(retpage, do_poll); |
5b999aadb mm: swap: zswap: ... |
462 463 464 |
return retpage; } |
ec560175c mm, swap: VMA bas... |
465 466 467 468 469 |
static unsigned int __swapin_nr_pages(unsigned long prev_offset, unsigned long offset, int hits, int max_pages, int prev_win) |
579f82901 swap: add a simpl... |
470 |
{ |
ec560175c mm, swap: VMA bas... |
471 |
unsigned int pages, last_ra; |
579f82901 swap: add a simpl... |
472 473 474 475 476 477 |
/* * This heuristic has been found to work well on both sequential and * random loads, swapping to hard disk or to SSD: please don't ask * what the "+ 2" means, it just happens to work well, that's all. */ |
ec560175c mm, swap: VMA bas... |
478 |
pages = hits + 2; |
579f82901 swap: add a simpl... |
479 480 481 482 483 484 485 486 |
if (pages == 2) { /* * We can have no readahead hits to judge by: but must not get * stuck here forever, so check for an adjacent offset instead * (and don't even bother to check whether swap type is same). */ if (offset != prev_offset + 1 && offset != prev_offset - 1) pages = 1; |
579f82901 swap: add a simpl... |
487 488 489 490 491 492 493 494 495 496 497 |
} else { unsigned int roundup = 4; while (roundup < pages) roundup <<= 1; pages = roundup; } if (pages > max_pages) pages = max_pages; /* Don't shrink readahead too fast */ |
ec560175c mm, swap: VMA bas... |
498 |
last_ra = prev_win / 2; |
579f82901 swap: add a simpl... |
499 500 |
if (pages < last_ra) pages = last_ra; |
ec560175c mm, swap: VMA bas... |
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 |
return pages; } static unsigned long swapin_nr_pages(unsigned long offset) { static unsigned long prev_offset; unsigned int hits, pages, max_pages; static atomic_t last_readahead_pages; max_pages = 1 << READ_ONCE(page_cluster); if (max_pages <= 1) return 1; hits = atomic_xchg(&swapin_readahead_hits, 0); pages = __swapin_nr_pages(prev_offset, offset, hits, max_pages, atomic_read(&last_readahead_pages)); if (!hits) prev_offset = offset; |
579f82901 swap: add a simpl... |
520 521 522 523 |
atomic_set(&last_readahead_pages, pages); return pages; } |
46017e954 swapin_readahead:... |
524 525 526 |
/** * swapin_readahead - swap in pages in hope we need them soon * @entry: swap entry of this memory |
7682486b3 mm: fix various k... |
527 |
* @gfp_mask: memory allocation flags |
46017e954 swapin_readahead:... |
528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 |
* @vma: user vma this address belongs to * @addr: target address for mempolicy * * Returns the struct page for entry and addr, after queueing swapin. * * Primitive swap readahead code. We simply read an aligned block of * (1 << page_cluster) entries in the swap area. This method is chosen * because it doesn't cost us any seek time. We also make sure to queue * the 'original' request together with the readahead ones... * * This has been extended to use the NUMA policies from the mm triggering * the readahead. * * Caller must hold down_read on the vma->vm_mm if vma is not NULL. */ |
02098feaa swapin needs gfp_... |
543 |
struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask, |
46017e954 swapin_readahead:... |
544 545 |
struct vm_area_struct *vma, unsigned long addr) { |
46017e954 swapin_readahead:... |
546 |
struct page *page; |
579f82901 swap: add a simpl... |
547 548 |
unsigned long entry_offset = swp_offset(entry); unsigned long offset = entry_offset; |
67f96aa25 mm: make swapin r... |
549 |
unsigned long start_offset, end_offset; |
579f82901 swap: add a simpl... |
550 |
unsigned long mask; |
3fb5c298b swap: allow swap ... |
551 |
struct blk_plug plug; |
c4fa63092 mm, swap: fix swa... |
552 |
bool do_poll = true, page_allocated; |
46017e954 swapin_readahead:... |
553 |
|
579f82901 swap: add a simpl... |
554 555 556 |
mask = swapin_nr_pages(offset) - 1; if (!mask) goto skip; |
23955622f swap: add block i... |
557 |
do_poll = false; |
67f96aa25 mm: make swapin r... |
558 559 560 561 562 |
/* Read a page_cluster sized and aligned cluster around offset. */ start_offset = offset & ~mask; end_offset = offset | mask; if (!start_offset) /* First page is swap header. */ start_offset++; |
3fb5c298b swap: allow swap ... |
563 |
blk_start_plug(&plug); |
67f96aa25 mm: make swapin r... |
564 |
for (offset = start_offset; offset <= end_offset ; offset++) { |
46017e954 swapin_readahead:... |
565 |
/* Ok, do the async read-ahead now */ |
c4fa63092 mm, swap: fix swa... |
566 567 568 |
page = __read_swap_cache_async( swp_entry(swp_type(entry), offset), gfp_mask, vma, addr, &page_allocated); |
46017e954 swapin_readahead:... |
569 |
if (!page) |
67f96aa25 mm: make swapin r... |
570 |
continue; |
c4fa63092 mm, swap: fix swa... |
571 572 573 574 575 576 577 |
if (page_allocated) { swap_readpage(page, false); if (offset != entry_offset && likely(!PageTransCompound(page))) { SetPageReadahead(page); count_vm_event(SWAP_RA); } |
cbc65df24 mm, swap: add swa... |
578 |
} |
09cbfeaf1 mm, fs: get rid o... |
579 |
put_page(page); |
46017e954 swapin_readahead:... |
580 |
} |
3fb5c298b swap: allow swap ... |
581 |
blk_finish_plug(&plug); |
46017e954 swapin_readahead:... |
582 |
lru_add_drain(); /* Push any new pages onto the LRU now */ |
579f82901 swap: add a simpl... |
583 |
skip: |
23955622f swap: add block i... |
584 |
return read_swap_cache_async(entry, gfp_mask, vma, addr, do_poll); |
46017e954 swapin_readahead:... |
585 |
} |
4b3ef9daa mm/swap: split sw... |
586 587 588 589 590 591 592 |
int init_swap_address_space(unsigned int type, unsigned long nr_pages) { struct address_space *spaces, *space; unsigned int i, nr; nr = DIV_ROUND_UP(nr_pages, SWAP_ADDRESS_SPACE_PAGES); |
54f180d3c mm, swap: use kvz... |
593 |
spaces = kvzalloc(sizeof(struct address_space) * nr, GFP_KERNEL); |
4b3ef9daa mm/swap: split sw... |
594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 |
if (!spaces) return -ENOMEM; for (i = 0; i < nr; i++) { space = spaces + i; INIT_RADIX_TREE(&space->page_tree, GFP_ATOMIC|__GFP_NOWARN); atomic_set(&space->i_mmap_writable, 0); space->a_ops = &swap_aops; /* swap cache doesn't use writeback related tags */ mapping_set_no_writeback_tags(space); spin_lock_init(&space->tree_lock); } nr_swapper_spaces[type] = nr; rcu_assign_pointer(swapper_spaces[type], spaces); return 0; } void exit_swap_address_space(unsigned int type) { struct address_space *spaces; spaces = swapper_spaces[type]; nr_swapper_spaces[type] = 0; rcu_assign_pointer(swapper_spaces[type], NULL); synchronize_rcu(); kvfree(spaces); } |
ec560175c mm, swap: VMA bas... |
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 |
static inline void swap_ra_clamp_pfn(struct vm_area_struct *vma, unsigned long faddr, unsigned long lpfn, unsigned long rpfn, unsigned long *start, unsigned long *end) { *start = max3(lpfn, PFN_DOWN(vma->vm_start), PFN_DOWN(faddr & PMD_MASK)); *end = min3(rpfn, PFN_DOWN(vma->vm_end), PFN_DOWN((faddr & PMD_MASK) + PMD_SIZE)); } struct page *swap_readahead_detect(struct vm_fault *vmf, struct vma_swap_readahead *swap_ra) { struct vm_area_struct *vma = vmf->vma; unsigned long swap_ra_info; struct page *page; swp_entry_t entry; unsigned long faddr, pfn, fpfn; unsigned long start, end; pte_t *pte; unsigned int max_win, hits, prev_win, win, left; #ifndef CONFIG_64BIT pte_t *tpte; #endif |
61b639723 mm, swap: use pag... |
649 650 651 652 653 654 |
max_win = 1 << min_t(unsigned int, READ_ONCE(page_cluster), SWAP_RA_ORDER_CEILING); if (max_win == 1) { swap_ra->win = 1; return NULL; } |
ec560175c mm, swap: VMA bas... |
655 656 657 658 659 660 661 |
faddr = vmf->address; entry = pte_to_swp_entry(vmf->orig_pte); if ((unlikely(non_swap_entry(entry)))) return NULL; page = lookup_swap_cache(entry, vma, faddr); if (page) return page; |
ec560175c mm, swap: VMA bas... |
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 |
fpfn = PFN_DOWN(faddr); swap_ra_info = GET_SWAP_RA_VAL(vma); pfn = PFN_DOWN(SWAP_RA_ADDR(swap_ra_info)); prev_win = SWAP_RA_WIN(swap_ra_info); hits = SWAP_RA_HITS(swap_ra_info); swap_ra->win = win = __swapin_nr_pages(pfn, fpfn, hits, max_win, prev_win); atomic_long_set(&vma->swap_readahead_info, SWAP_RA_VAL(faddr, win, 0)); if (win == 1) return NULL; /* Copy the PTEs because the page table may be unmapped */ if (fpfn == pfn + 1) swap_ra_clamp_pfn(vma, faddr, fpfn, fpfn + win, &start, &end); else if (pfn == fpfn + 1) swap_ra_clamp_pfn(vma, faddr, fpfn - win + 1, fpfn + 1, &start, &end); else { left = (win - 1) / 2; swap_ra_clamp_pfn(vma, faddr, fpfn - left, fpfn + win - left, &start, &end); } swap_ra->nr_pte = end - start; swap_ra->offset = fpfn - start; pte = vmf->pte - swap_ra->offset; #ifdef CONFIG_64BIT swap_ra->ptes = pte; #else tpte = swap_ra->ptes; for (pfn = start; pfn != end; pfn++) *tpte++ = *pte++; #endif return NULL; } struct page *do_swap_page_readahead(swp_entry_t fentry, gfp_t gfp_mask, struct vm_fault *vmf, struct vma_swap_readahead *swap_ra) { struct blk_plug plug; struct vm_area_struct *vma = vmf->vma; struct page *page; pte_t *pte, pentry; swp_entry_t entry; unsigned int i; bool page_allocated; if (swap_ra->win == 1) goto skip; blk_start_plug(&plug); for (i = 0, pte = swap_ra->ptes; i < swap_ra->nr_pte; i++, pte++) { pentry = *pte; if (pte_none(pentry)) continue; if (pte_present(pentry)) continue; entry = pte_to_swp_entry(pentry); if (unlikely(non_swap_entry(entry))) continue; page = __read_swap_cache_async(entry, gfp_mask, vma, vmf->address, &page_allocated); if (!page) continue; if (page_allocated) { swap_readpage(page, false); if (i != swap_ra->offset && likely(!PageTransCompound(page))) { SetPageReadahead(page); count_vm_event(SWAP_RA); } } put_page(page); } blk_finish_plug(&plug); lru_add_drain(); skip: return read_swap_cache_async(fentry, gfp_mask, vma, vmf->address, swap_ra->win == 1); } |
d9bfcfdc4 mm, swap: add sys... |
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 |
#ifdef CONFIG_SYSFS static ssize_t vma_ra_enabled_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) { return sprintf(buf, "%s ", swap_vma_readahead ? "true" : "false"); } static ssize_t vma_ra_enabled_store(struct kobject *kobj, struct kobj_attribute *attr, const char *buf, size_t count) { if (!strncmp(buf, "true", 4) || !strncmp(buf, "1", 1)) swap_vma_readahead = true; else if (!strncmp(buf, "false", 5) || !strncmp(buf, "0", 1)) swap_vma_readahead = false; else return -EINVAL; return count; } static struct kobj_attribute vma_ra_enabled_attr = __ATTR(vma_ra_enabled, 0644, vma_ra_enabled_show, vma_ra_enabled_store); |
d9bfcfdc4 mm, swap: add sys... |
770 771 |
static struct attribute *swap_attrs[] = { &vma_ra_enabled_attr.attr, |
d9bfcfdc4 mm, swap: add sys... |
772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 |
NULL, }; static struct attribute_group swap_attr_group = { .attrs = swap_attrs, }; static int __init swap_init_sysfs(void) { int err; struct kobject *swap_kobj; swap_kobj = kobject_create_and_add("swap", mm_kobj); if (!swap_kobj) { pr_err("failed to create swap kobject "); return -ENOMEM; } err = sysfs_create_group(swap_kobj, &swap_attr_group); if (err) { pr_err("failed to register swap group "); goto delete_obj; } return 0; delete_obj: kobject_put(swap_kobj); return err; } subsys_initcall(swap_init_sysfs); #endif |