Blame view
mm/migrate.c
77.1 KB
b24413180
|
1 |
// SPDX-License-Identifier: GPL-2.0 |
b20a35035
|
2 |
/* |
14e0f9bcc
|
3 |
* Memory Migration functionality - linux/mm/migrate.c |
b20a35035
|
4 5 6 7 8 9 10 11 12 |
* * Copyright (C) 2006 Silicon Graphics, Inc., Christoph Lameter * * Page migration was first developed in the context of the memory hotplug * project. The main authors of the migration code are: * * IWAMOTO Toshihiro <iwamoto@valinux.co.jp> * Hirokazu Takahashi <taka@valinux.co.jp> * Dave Hansen <haveblue@us.ibm.com> |
cde535359
|
13 |
* Christoph Lameter |
b20a35035
|
14 15 16 |
*/ #include <linux/migrate.h> |
b95f1b31b
|
17 |
#include <linux/export.h> |
b20a35035
|
18 |
#include <linux/swap.h> |
0697212a4
|
19 |
#include <linux/swapops.h> |
b20a35035
|
20 |
#include <linux/pagemap.h> |
e23ca00bf
|
21 |
#include <linux/buffer_head.h> |
b20a35035
|
22 |
#include <linux/mm_inline.h> |
b488893a3
|
23 |
#include <linux/nsproxy.h> |
b20a35035
|
24 |
#include <linux/pagevec.h> |
e9995ef97
|
25 |
#include <linux/ksm.h> |
b20a35035
|
26 27 28 29 |
#include <linux/rmap.h> #include <linux/topology.h> #include <linux/cpu.h> #include <linux/cpuset.h> |
04e62a29b
|
30 |
#include <linux/writeback.h> |
742755a1d
|
31 32 |
#include <linux/mempolicy.h> #include <linux/vmalloc.h> |
86c3a7645
|
33 |
#include <linux/security.h> |
42cb14b11
|
34 |
#include <linux/backing-dev.h> |
bda807d44
|
35 |
#include <linux/compaction.h> |
4f5ca2657
|
36 |
#include <linux/syscalls.h> |
7addf4438
|
37 |
#include <linux/compat.h> |
290408d4a
|
38 |
#include <linux/hugetlb.h> |
8e6ac7fab
|
39 |
#include <linux/hugetlb_cgroup.h> |
5a0e3ad6a
|
40 |
#include <linux/gfp.h> |
a520110e4
|
41 |
#include <linux/pagewalk.h> |
df6ad6983
|
42 |
#include <linux/pfn_t.h> |
a5430dda8
|
43 |
#include <linux/memremap.h> |
8315ada7f
|
44 |
#include <linux/userfaultfd_k.h> |
bf6bddf19
|
45 |
#include <linux/balloon_compaction.h> |
f714f4f20
|
46 |
#include <linux/mmu_notifier.h> |
33c3fc71c
|
47 |
#include <linux/page_idle.h> |
d435edca9
|
48 |
#include <linux/page_owner.h> |
6e84f3152
|
49 |
#include <linux/sched/mm.h> |
197e7e521
|
50 |
#include <linux/ptrace.h> |
b20a35035
|
51 |
|
0d1836c36
|
52 |
#include <asm/tlbflush.h> |
7b2a2d4a1
|
53 54 |
#define CREATE_TRACE_POINTS #include <trace/events/migrate.h> |
b20a35035
|
55 |
#include "internal.h" |
b20a35035
|
56 |
/* |
742755a1d
|
57 |
* migrate_prep() needs to be called before we start compiling a list of pages |
748446bb6
|
58 59 |
* to be migrated using isolate_lru_page(). If scheduling work on other CPUs is * undesirable, use migrate_prep_local() |
b20a35035
|
60 61 62 |
*/ int migrate_prep(void) { |
b20a35035
|
63 64 65 66 67 68 69 70 71 72 |
/* * Clear the LRU lists so pages can be isolated. * Note that pages may be moved off the LRU after we have * drained them. Those pages will fail to migrate like other * pages that may be busy. */ lru_add_drain_all(); return 0; } |
748446bb6
|
73 74 75 76 77 78 79 |
/* Do the necessary work of migrate_prep but not if it involves other CPUs */ int migrate_prep_local(void) { lru_add_drain(); return 0; } |
9e5bcd610
|
80 |
int isolate_movable_page(struct page *page, isolate_mode_t mode) |
bda807d44
|
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
{ struct address_space *mapping; /* * Avoid burning cycles with pages that are yet under __free_pages(), * or just got freed under us. * * In case we 'win' a race for a movable page being freed under us and * raise its refcount preventing __free_pages() from doing its job * the put_page() at the end of this block will take care of * release this page, thus avoiding a nasty leakage. */ if (unlikely(!get_page_unless_zero(page))) goto out; /* * Check PageMovable before holding a PG_lock because page's owner * assumes anybody doesn't touch PG_lock of newly allocated page |
8bb4e7a2e
|
99 |
* so unconditionally grabbing the lock ruins page's owner side. |
bda807d44
|
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
*/ if (unlikely(!__PageMovable(page))) goto out_putpage; /* * As movable pages are not isolated from LRU lists, concurrent * compaction threads can race against page migration functions * as well as race against the releasing a page. * * In order to avoid having an already isolated movable page * being (wrongly) re-isolated while it is under migration, * or to avoid attempting to isolate pages being released, * lets be sure we have the page lock * before proceeding with the movable page isolation steps. */ if (unlikely(!trylock_page(page))) goto out_putpage; if (!PageMovable(page) || PageIsolated(page)) goto out_no_isolated; mapping = page_mapping(page); VM_BUG_ON_PAGE(!mapping, page); if (!mapping->a_ops->isolate_page(page, mode)) goto out_no_isolated; /* Driver shouldn't use PG_isolated bit of page->flags */ WARN_ON_ONCE(PageIsolated(page)); __SetPageIsolated(page); unlock_page(page); |
9e5bcd610
|
130 |
return 0; |
bda807d44
|
131 132 133 134 135 136 |
out_no_isolated: unlock_page(page); out_putpage: put_page(page); out: |
9e5bcd610
|
137 |
return -EBUSY; |
bda807d44
|
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 |
} /* It should be called on page which is PG_movable */ void putback_movable_page(struct page *page) { struct address_space *mapping; VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(!PageIsolated(page), page); mapping = page_mapping(page); mapping->a_ops->putback_page(page); __ClearPageIsolated(page); } |
b20a35035
|
153 |
/* |
5733c7d11
|
154 155 156 |
* Put previously isolated pages back onto the appropriate lists * from where they were once taken off for compaction/migration. * |
59c82b70d
|
157 158 159 |
* This function shall be used whenever the isolated pageset has been * built from lru, balloon, hugetlbfs page. See isolate_migratepages_range() * and isolate_huge_page(). |
5733c7d11
|
160 161 162 163 164 165 166 |
*/ void putback_movable_pages(struct list_head *l) { struct page *page; struct page *page2; list_for_each_entry_safe(page, page2, l, lru) { |
31caf665e
|
167 168 169 170 |
if (unlikely(PageHuge(page))) { putback_active_hugepage(page); continue; } |
5733c7d11
|
171 |
list_del(&page->lru); |
bda807d44
|
172 173 174 175 176 |
/* * We isolated non-lru movable page so here we can use * __PageMovable because LRU page's mapping cannot have * PAGE_MAPPING_MOVABLE. */ |
b1123ea6d
|
177 |
if (unlikely(__PageMovable(page))) { |
bda807d44
|
178 179 180 181 182 183 184 185 186 |
VM_BUG_ON_PAGE(!PageIsolated(page), page); lock_page(page); if (PageMovable(page)) putback_movable_page(page); else __ClearPageIsolated(page); unlock_page(page); put_page(page); } else { |
e8db67eb0
|
187 188 |
mod_node_page_state(page_pgdat(page), NR_ISOLATED_ANON + page_is_file_cache(page), -hpage_nr_pages(page)); |
fc280fe87
|
189 |
putback_lru_page(page); |
bda807d44
|
190 |
} |
b20a35035
|
191 |
} |
b20a35035
|
192 |
} |
0697212a4
|
193 194 195 |
/* * Restore a potential migration pte to a working pte entry */ |
e4b822227
|
196 |
static bool remove_migration_pte(struct page *page, struct vm_area_struct *vma, |
e9995ef97
|
197 |
unsigned long addr, void *old) |
0697212a4
|
198 |
{ |
3fe87967c
|
199 200 201 202 203 204 205 206 |
struct page_vma_mapped_walk pvmw = { .page = old, .vma = vma, .address = addr, .flags = PVMW_SYNC | PVMW_MIGRATION, }; struct page *new; pte_t pte; |
0697212a4
|
207 |
swp_entry_t entry; |
0697212a4
|
208 |
|
3fe87967c
|
209 210 |
VM_BUG_ON_PAGE(PageTail(page), page); while (page_vma_mapped_walk(&pvmw)) { |
4b0ece6fa
|
211 212 213 214 215 |
if (PageKsm(page)) new = page; else new = page - pvmw.page->index + linear_page_index(vma, pvmw.address); |
0697212a4
|
216 |
|
616b83715
|
217 218 219 220 221 222 223 224 |
#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION /* PMD-mapped THP migration entry */ if (!pvmw.pte) { VM_BUG_ON_PAGE(PageHuge(page) || !PageTransCompound(page), page); remove_migration_pmd(&pvmw, new); continue; } #endif |
3fe87967c
|
225 226 227 228 |
get_page(new); pte = pte_mkold(mk_pte(new, READ_ONCE(vma->vm_page_prot))); if (pte_swp_soft_dirty(*pvmw.pte)) pte = pte_mksoft_dirty(pte); |
0697212a4
|
229 |
|
3fe87967c
|
230 231 232 233 234 235 |
/* * Recheck VMA as permissions can change since migration started */ entry = pte_to_swp_entry(*pvmw.pte); if (is_write_migration_entry(entry)) pte = maybe_mkwrite(pte, vma); |
d3cb8bf60
|
236 |
|
df6ad6983
|
237 238 239 240 |
if (unlikely(is_zone_device_page(new))) { if (is_device_private_page(new)) { entry = make_device_private_entry(new, pte_write(pte)); pte = swp_entry_to_pte(entry); |
df6ad6983
|
241 |
} |
d2b2c6dd2
|
242 |
} |
a5430dda8
|
243 |
|
3ef8fd7f7
|
244 |
#ifdef CONFIG_HUGETLB_PAGE |
3fe87967c
|
245 246 247 |
if (PageHuge(new)) { pte = pte_mkhuge(pte); pte = arch_make_huge_pte(pte, vma, new, 0); |
383321ab8
|
248 |
set_huge_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte); |
3fe87967c
|
249 250 251 252 |
if (PageAnon(new)) hugepage_add_anon_rmap(new, vma, pvmw.address); else page_dup_rmap(new, true); |
383321ab8
|
253 254 255 256 |
} else #endif { set_pte_at(vma->vm_mm, pvmw.address, pvmw.pte, pte); |
04e62a29b
|
257 |
|
383321ab8
|
258 259 260 261 262 |
if (PageAnon(new)) page_add_anon_rmap(new, vma, pvmw.address, false); else page_add_file_rmap(new, false); } |
3fe87967c
|
263 264 |
if (vma->vm_flags & VM_LOCKED && !PageTransCompound(new)) mlock_vma_page(new); |
e125fe405
|
265 266 |
if (PageTransHuge(page) && PageMlocked(page)) clear_page_mlock(page); |
3fe87967c
|
267 268 269 |
/* No need to invalidate - it was non-present before */ update_mmu_cache(vma, pvmw.address, pvmw.pte); } |
51afb12ba
|
270 |
|
e4b822227
|
271 |
return true; |
0697212a4
|
272 273 274 |
} /* |
04e62a29b
|
275 276 277 |
* Get rid of all migration entries and replace them by * references to the indicated page. */ |
e388466de
|
278 |
void remove_migration_ptes(struct page *old, struct page *new, bool locked) |
04e62a29b
|
279 |
{ |
051ac83ad
|
280 281 282 283 |
struct rmap_walk_control rwc = { .rmap_one = remove_migration_pte, .arg = old, }; |
e388466de
|
284 285 286 287 |
if (locked) rmap_walk_locked(new, &rwc); else rmap_walk(new, &rwc); |
04e62a29b
|
288 289 290 |
} /* |
0697212a4
|
291 292 293 |
* Something used the pte of a page under migration. We need to * get to the page and wait until migration is finished. * When we return from this function the fault will be retried. |
0697212a4
|
294 |
*/ |
e66f17ff7
|
295 |
void __migration_entry_wait(struct mm_struct *mm, pte_t *ptep, |
30dad3092
|
296 |
spinlock_t *ptl) |
0697212a4
|
297 |
{ |
30dad3092
|
298 |
pte_t pte; |
0697212a4
|
299 300 |
swp_entry_t entry; struct page *page; |
30dad3092
|
301 |
spin_lock(ptl); |
0697212a4
|
302 303 304 305 306 307 308 309 310 |
pte = *ptep; if (!is_swap_pte(pte)) goto out; entry = pte_to_swp_entry(pte); if (!is_migration_entry(entry)) goto out; page = migration_entry_to_page(entry); |
e286781d5
|
311 |
/* |
89eb946a7
|
312 |
* Once page cache replacement of page migration started, page_count |
9a1ea439b
|
313 314 |
* is zero; but we must not call put_and_wait_on_page_locked() without * a ref. Use get_page_unless_zero(), and just fault again if it fails. |
e286781d5
|
315 316 317 |
*/ if (!get_page_unless_zero(page)) goto out; |
0697212a4
|
318 |
pte_unmap_unlock(ptep, ptl); |
9a1ea439b
|
319 |
put_and_wait_on_page_locked(page); |
0697212a4
|
320 321 322 323 |
return; out: pte_unmap_unlock(ptep, ptl); } |
30dad3092
|
324 325 326 327 328 329 330 |
void migration_entry_wait(struct mm_struct *mm, pmd_t *pmd, unsigned long address) { spinlock_t *ptl = pte_lockptr(mm, pmd); pte_t *ptep = pte_offset_map(pmd, address); __migration_entry_wait(mm, ptep, ptl); } |
cb900f412
|
331 332 |
void migration_entry_wait_huge(struct vm_area_struct *vma, struct mm_struct *mm, pte_t *pte) |
30dad3092
|
333 |
{ |
cb900f412
|
334 |
spinlock_t *ptl = huge_pte_lockptr(hstate_vma(vma), mm, pte); |
30dad3092
|
335 336 |
__migration_entry_wait(mm, pte, ptl); } |
616b83715
|
337 338 339 340 341 342 343 344 345 346 347 348 349 |
#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION void pmd_migration_entry_wait(struct mm_struct *mm, pmd_t *pmd) { spinlock_t *ptl; struct page *page; ptl = pmd_lock(mm, pmd); if (!is_pmd_migration_entry(*pmd)) goto unlock; page = migration_entry_to_page(pmd_to_swp_entry(*pmd)); if (!get_page_unless_zero(page)) goto unlock; spin_unlock(ptl); |
9a1ea439b
|
350 |
put_and_wait_on_page_locked(page); |
616b83715
|
351 352 353 354 355 |
return; unlock: spin_unlock(ptl); } #endif |
f900482da
|
356 |
static int expected_page_refs(struct address_space *mapping, struct page *page) |
0b3901b38
|
357 358 359 360 361 362 363 364 |
{ int expected_count = 1; /* * Device public or private pages have an extra refcount as they are * ZONE_DEVICE pages. */ expected_count += is_device_private_page(page); |
f900482da
|
365 |
if (mapping) |
0b3901b38
|
366 367 368 369 |
expected_count += hpage_nr_pages(page) + page_has_private(page); return expected_count; } |
b20a35035
|
370 |
/* |
c3fcf8a5d
|
371 |
* Replace the page in the mapping. |
5b5c7120e
|
372 373 374 375 |
* * The number of remaining references must be: * 1 for anonymous pages without a mapping * 2 for pages with a mapping |
266cf658e
|
376 |
* 3 for pages with a mapping and PagePrivate/PagePrivate2 set. |
b20a35035
|
377 |
*/ |
36bc08cc0
|
378 |
int migrate_page_move_mapping(struct address_space *mapping, |
371096949
|
379 |
struct page *newpage, struct page *page, int extra_count) |
b20a35035
|
380 |
{ |
89eb946a7
|
381 |
XA_STATE(xas, &mapping->i_pages, page_index(page)); |
42cb14b11
|
382 383 |
struct zone *oldzone, *newzone; int dirty; |
f900482da
|
384 |
int expected_count = expected_page_refs(mapping, page) + extra_count; |
8763cb45a
|
385 |
|
6c5240ae7
|
386 |
if (!mapping) { |
0e8c7d0fd
|
387 |
/* Anonymous page without mapping */ |
8e321fefb
|
388 |
if (page_count(page) != expected_count) |
6c5240ae7
|
389 |
return -EAGAIN; |
cf4b769ab
|
390 391 |
/* No turning back from here */ |
cf4b769ab
|
392 393 394 |
newpage->index = page->index; newpage->mapping = page->mapping; if (PageSwapBacked(page)) |
fa9949da5
|
395 |
__SetPageSwapBacked(newpage); |
cf4b769ab
|
396 |
|
78bd52097
|
397 |
return MIGRATEPAGE_SUCCESS; |
6c5240ae7
|
398 |
} |
42cb14b11
|
399 400 |
oldzone = page_zone(page); newzone = page_zone(newpage); |
89eb946a7
|
401 |
xas_lock_irq(&xas); |
89eb946a7
|
402 403 |
if (page_count(page) != expected_count || xas_load(&xas) != page) { xas_unlock_irq(&xas); |
e23ca00bf
|
404 |
return -EAGAIN; |
b20a35035
|
405 |
} |
fe896d187
|
406 |
if (!page_ref_freeze(page, expected_count)) { |
89eb946a7
|
407 |
xas_unlock_irq(&xas); |
e286781d5
|
408 409 |
return -EAGAIN; } |
b20a35035
|
410 |
/* |
cf4b769ab
|
411 412 |
* Now we know that no one else is looking at the page: * no turning back from here. |
b20a35035
|
413 |
*/ |
cf4b769ab
|
414 415 |
newpage->index = page->index; newpage->mapping = page->mapping; |
e71769ae5
|
416 |
page_ref_add(newpage, hpage_nr_pages(page)); /* add cache reference */ |
6326fec11
|
417 418 419 420 421 422 423 424 |
if (PageSwapBacked(page)) { __SetPageSwapBacked(newpage); if (PageSwapCache(page)) { SetPageSwapCache(newpage); set_page_private(newpage, page_private(page)); } } else { VM_BUG_ON_PAGE(PageSwapCache(page), page); |
b20a35035
|
425 |
} |
42cb14b11
|
426 427 428 429 430 431 |
/* Move dirty while page refs frozen and newpage not yet exposed */ dirty = PageDirty(page); if (dirty) { ClearPageDirty(page); SetPageDirty(newpage); } |
89eb946a7
|
432 |
xas_store(&xas, newpage); |
e71769ae5
|
433 434 |
if (PageTransHuge(page)) { int i; |
e71769ae5
|
435 |
|
013567be1
|
436 |
for (i = 1; i < HPAGE_PMD_NR; i++) { |
89eb946a7
|
437 |
xas_next(&xas); |
4101196b1
|
438 |
xas_store(&xas, newpage); |
e71769ae5
|
439 |
} |
e71769ae5
|
440 |
} |
7cf9c2c76
|
441 442 |
/* |
937a94c9d
|
443 444 |
* Drop cache reference from old page by unfreezing * to one less reference. |
7cf9c2c76
|
445 446 |
* We know this isn't the last reference. */ |
e71769ae5
|
447 |
page_ref_unfreeze(page, expected_count - hpage_nr_pages(page)); |
7cf9c2c76
|
448 |
|
89eb946a7
|
449 |
xas_unlock(&xas); |
42cb14b11
|
450 |
/* Leave irq disabled to prevent preemption while updating stats */ |
0e8c7d0fd
|
451 452 453 454 455 456 457 |
/* * If moved to a different zone then also account * the page for that zone. Other VM counters will be * taken care of when we establish references to the * new page and drop references to the old page. * * Note that anonymous pages are accounted for |
4b9d0fab7
|
458 |
* via NR_FILE_PAGES and NR_ANON_MAPPED if they |
0e8c7d0fd
|
459 460 |
* are mapped to swap space. */ |
42cb14b11
|
461 |
if (newzone != oldzone) { |
11fb99898
|
462 463 |
__dec_node_state(oldzone->zone_pgdat, NR_FILE_PAGES); __inc_node_state(newzone->zone_pgdat, NR_FILE_PAGES); |
42cb14b11
|
464 |
if (PageSwapBacked(page) && !PageSwapCache(page)) { |
11fb99898
|
465 466 |
__dec_node_state(oldzone->zone_pgdat, NR_SHMEM); __inc_node_state(newzone->zone_pgdat, NR_SHMEM); |
42cb14b11
|
467 468 |
} if (dirty && mapping_cap_account_dirty(mapping)) { |
11fb99898
|
469 |
__dec_node_state(oldzone->zone_pgdat, NR_FILE_DIRTY); |
5a1c84b40
|
470 |
__dec_zone_state(oldzone, NR_ZONE_WRITE_PENDING); |
11fb99898
|
471 |
__inc_node_state(newzone->zone_pgdat, NR_FILE_DIRTY); |
5a1c84b40
|
472 |
__inc_zone_state(newzone, NR_ZONE_WRITE_PENDING); |
42cb14b11
|
473 |
} |
4b02108ac
|
474 |
} |
42cb14b11
|
475 |
local_irq_enable(); |
b20a35035
|
476 |
|
78bd52097
|
477 |
return MIGRATEPAGE_SUCCESS; |
b20a35035
|
478 |
} |
1118dce77
|
479 |
EXPORT_SYMBOL(migrate_page_move_mapping); |
b20a35035
|
480 481 |
/* |
290408d4a
|
482 483 484 485 486 487 |
* The expected number of remaining references is the same as that * of migrate_page_move_mapping(). */ int migrate_huge_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page) { |
89eb946a7
|
488 |
XA_STATE(xas, &mapping->i_pages, page_index(page)); |
290408d4a
|
489 |
int expected_count; |
290408d4a
|
490 |
|
89eb946a7
|
491 |
xas_lock_irq(&xas); |
290408d4a
|
492 |
expected_count = 2 + page_has_private(page); |
89eb946a7
|
493 494 |
if (page_count(page) != expected_count || xas_load(&xas) != page) { xas_unlock_irq(&xas); |
290408d4a
|
495 496 |
return -EAGAIN; } |
fe896d187
|
497 |
if (!page_ref_freeze(page, expected_count)) { |
89eb946a7
|
498 |
xas_unlock_irq(&xas); |
290408d4a
|
499 500 |
return -EAGAIN; } |
cf4b769ab
|
501 502 |
newpage->index = page->index; newpage->mapping = page->mapping; |
6a93ca8fd
|
503 |
|
290408d4a
|
504 |
get_page(newpage); |
89eb946a7
|
505 |
xas_store(&xas, newpage); |
290408d4a
|
506 |
|
fe896d187
|
507 |
page_ref_unfreeze(page, expected_count - 1); |
290408d4a
|
508 |
|
89eb946a7
|
509 |
xas_unlock_irq(&xas); |
6a93ca8fd
|
510 |
|
78bd52097
|
511 |
return MIGRATEPAGE_SUCCESS; |
290408d4a
|
512 513 514 |
} /* |
30b0a105d
|
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 |
* Gigantic pages are so large that we do not guarantee that page++ pointer * arithmetic will work across the entire page. We need something more * specialized. */ static void __copy_gigantic_page(struct page *dst, struct page *src, int nr_pages) { int i; struct page *dst_base = dst; struct page *src_base = src; for (i = 0; i < nr_pages; ) { cond_resched(); copy_highpage(dst, src); i++; dst = mem_map_next(dst, dst_base, i); src = mem_map_next(src, src_base, i); } } static void copy_huge_page(struct page *dst, struct page *src) { int i; int nr_pages; if (PageHuge(src)) { /* hugetlbfs page */ struct hstate *h = page_hstate(src); nr_pages = pages_per_huge_page(h); if (unlikely(nr_pages > MAX_ORDER_NR_PAGES)) { __copy_gigantic_page(dst, src, nr_pages); return; } } else { /* thp page */ BUG_ON(!PageTransHuge(src)); nr_pages = hpage_nr_pages(src); } for (i = 0; i < nr_pages; i++) { cond_resched(); copy_highpage(dst + i, src + i); } } /* |
b20a35035
|
563 564 |
* Copy the page to its new location */ |
2916ecc0f
|
565 |
void migrate_page_states(struct page *newpage, struct page *page) |
b20a35035
|
566 |
{ |
7851a45cd
|
567 |
int cpupid; |
b20a35035
|
568 569 570 571 572 573 |
if (PageError(page)) SetPageError(newpage); if (PageReferenced(page)) SetPageReferenced(newpage); if (PageUptodate(page)) SetPageUptodate(newpage); |
894bc3104
|
574 |
if (TestClearPageActive(page)) { |
309381fea
|
575 |
VM_BUG_ON_PAGE(PageUnevictable(page), page); |
b20a35035
|
576 |
SetPageActive(newpage); |
418b27ef5
|
577 578 |
} else if (TestClearPageUnevictable(page)) SetPageUnevictable(newpage); |
1899ad18c
|
579 580 |
if (PageWorkingset(page)) SetPageWorkingset(newpage); |
b20a35035
|
581 582 583 584 |
if (PageChecked(page)) SetPageChecked(newpage); if (PageMappedToDisk(page)) SetPageMappedToDisk(newpage); |
42cb14b11
|
585 586 587 |
/* Move dirty on pages not done by migrate_page_move_mapping() */ if (PageDirty(page)) SetPageDirty(newpage); |
b20a35035
|
588 |
|
33c3fc71c
|
589 590 591 592 |
if (page_is_young(page)) set_page_young(newpage); if (page_is_idle(page)) set_page_idle(newpage); |
7851a45cd
|
593 594 595 596 597 598 |
/* * Copy NUMA information to the new page, to prevent over-eager * future migrations of this same page. */ cpupid = page_cpupid_xchg_last(page, -1); page_cpupid_xchg_last(newpage, cpupid); |
e9995ef97
|
599 |
ksm_migrate_page(newpage, page); |
c8d6553b9
|
600 601 602 603 |
/* * Please do not reorder this without considering how mm/ksm.c's * get_ksm_page() depends upon ksm_migrate_page() and PageSwapCache(). */ |
b3b3a99c5
|
604 605 |
if (PageSwapCache(page)) ClearPageSwapCache(page); |
b20a35035
|
606 607 |
ClearPagePrivate(page); set_page_private(page, 0); |
b20a35035
|
608 609 610 611 612 613 614 |
/* * If any waiters have accumulated on the new page then * wake them up. */ if (PageWriteback(newpage)) end_page_writeback(newpage); |
d435edca9
|
615 616 |
copy_page_owner(page, newpage); |
74485cf2b
|
617 618 |
mem_cgroup_migrate(page, newpage); |
b20a35035
|
619 |
} |
2916ecc0f
|
620 621 622 623 624 625 626 627 628 629 630 |
EXPORT_SYMBOL(migrate_page_states); void migrate_page_copy(struct page *newpage, struct page *page) { if (PageHuge(page) || PageTransHuge(page)) copy_huge_page(newpage, page); else copy_highpage(newpage, page); migrate_page_states(newpage, page); } |
1118dce77
|
631 |
EXPORT_SYMBOL(migrate_page_copy); |
b20a35035
|
632 |
|
1d8b85ccf
|
633 634 635 |
/************************************************************ * Migration functions ***********************************************************/ |
b20a35035
|
636 |
/* |
bda807d44
|
637 |
* Common logic to directly migrate a single LRU page suitable for |
266cf658e
|
638 |
* pages that do not use PagePrivate/PagePrivate2. |
b20a35035
|
639 640 641 |
* * Pages are locked upon entry and exit. */ |
2d1db3b11
|
642 |
int migrate_page(struct address_space *mapping, |
a6bc32b89
|
643 644 |
struct page *newpage, struct page *page, enum migrate_mode mode) |
b20a35035
|
645 646 647 648 |
{ int rc; BUG_ON(PageWriteback(page)); /* Writeback must be complete */ |
371096949
|
649 |
rc = migrate_page_move_mapping(mapping, newpage, page, 0); |
b20a35035
|
650 |
|
78bd52097
|
651 |
if (rc != MIGRATEPAGE_SUCCESS) |
b20a35035
|
652 |
return rc; |
2916ecc0f
|
653 654 655 656 |
if (mode != MIGRATE_SYNC_NO_COPY) migrate_page_copy(newpage, page); else migrate_page_states(newpage, page); |
78bd52097
|
657 |
return MIGRATEPAGE_SUCCESS; |
b20a35035
|
658 659 |
} EXPORT_SYMBOL(migrate_page); |
9361401eb
|
660 |
#ifdef CONFIG_BLOCK |
84ade7c15
|
661 662 663 664 665 666 667 668 669 |
/* Returns true if all buffers are successfully locked */ static bool buffer_migrate_lock_buffers(struct buffer_head *head, enum migrate_mode mode) { struct buffer_head *bh = head; /* Simple case, sync compaction */ if (mode != MIGRATE_ASYNC) { do { |
84ade7c15
|
670 671 672 673 674 675 676 677 678 679 |
lock_buffer(bh); bh = bh->b_this_page; } while (bh != head); return true; } /* async case, we cannot block on lock_buffer so use trylock_buffer */ do { |
84ade7c15
|
680 681 682 683 684 685 |
if (!trylock_buffer(bh)) { /* * We failed to lock the buffer and cannot stall in * async migration. Release the taken locks */ struct buffer_head *failed_bh = bh; |
84ade7c15
|
686 687 688 |
bh = head; while (bh != failed_bh) { unlock_buffer(bh); |
84ade7c15
|
689 690 691 692 693 694 695 696 697 |
bh = bh->b_this_page; } return false; } bh = bh->b_this_page; } while (bh != head); return true; } |
89cb0888c
|
698 699 700 |
static int __buffer_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode, bool check_refs) |
1d8b85ccf
|
701 |
{ |
1d8b85ccf
|
702 703 |
struct buffer_head *bh, *head; int rc; |
cc4f11e69
|
704 |
int expected_count; |
1d8b85ccf
|
705 |
|
1d8b85ccf
|
706 |
if (!page_has_buffers(page)) |
a6bc32b89
|
707 |
return migrate_page(mapping, newpage, page, mode); |
1d8b85ccf
|
708 |
|
cc4f11e69
|
709 |
/* Check whether page does not have extra refs before we do more work */ |
f900482da
|
710 |
expected_count = expected_page_refs(mapping, page); |
cc4f11e69
|
711 712 |
if (page_count(page) != expected_count) return -EAGAIN; |
1d8b85ccf
|
713 |
|
cc4f11e69
|
714 715 716 |
head = page_buffers(page); if (!buffer_migrate_lock_buffers(head, mode)) return -EAGAIN; |
1d8b85ccf
|
717 |
|
89cb0888c
|
718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 |
if (check_refs) { bool busy; bool invalidated = false; recheck_buffers: busy = false; spin_lock(&mapping->private_lock); bh = head; do { if (atomic_read(&bh->b_count)) { busy = true; break; } bh = bh->b_this_page; } while (bh != head); |
89cb0888c
|
733 734 735 736 737 |
if (busy) { if (invalidated) { rc = -EAGAIN; goto unlock_buffers; } |
ebdf4de56
|
738 |
spin_unlock(&mapping->private_lock); |
89cb0888c
|
739 740 741 742 743 |
invalidate_bh_lrus(); invalidated = true; goto recheck_buffers; } } |
371096949
|
744 |
rc = migrate_page_move_mapping(mapping, newpage, page, 0); |
78bd52097
|
745 |
if (rc != MIGRATEPAGE_SUCCESS) |
cc4f11e69
|
746 |
goto unlock_buffers; |
1d8b85ccf
|
747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 |
ClearPagePrivate(page); set_page_private(newpage, page_private(page)); set_page_private(page, 0); put_page(page); get_page(newpage); bh = head; do { set_bh_page(bh, newpage, bh_offset(bh)); bh = bh->b_this_page; } while (bh != head); SetPagePrivate(newpage); |
2916ecc0f
|
762 763 764 765 |
if (mode != MIGRATE_SYNC_NO_COPY) migrate_page_copy(newpage, page); else migrate_page_states(newpage, page); |
1d8b85ccf
|
766 |
|
cc4f11e69
|
767 768 |
rc = MIGRATEPAGE_SUCCESS; unlock_buffers: |
ebdf4de56
|
769 770 |
if (check_refs) spin_unlock(&mapping->private_lock); |
1d8b85ccf
|
771 772 773 |
bh = head; do { unlock_buffer(bh); |
1d8b85ccf
|
774 775 776 |
bh = bh->b_this_page; } while (bh != head); |
cc4f11e69
|
777 |
return rc; |
1d8b85ccf
|
778 |
} |
89cb0888c
|
779 780 781 782 783 784 785 786 787 788 789 |
/* * Migration function for pages with buffers. This function can only be used * if the underlying filesystem guarantees that no other references to "page" * exist. For example attached buffer heads are accessed only under page lock. */ int buffer_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode) { return __buffer_migrate_page(mapping, newpage, page, mode, false); } |
1d8b85ccf
|
790 |
EXPORT_SYMBOL(buffer_migrate_page); |
89cb0888c
|
791 792 793 794 795 796 797 798 799 800 801 802 |
/* * Same as above except that this variant is more careful and checks that there * are also no buffer head references. This function is the right one for * mappings where buffer heads are directly looked up and referenced (such as * block device mappings). */ int buffer_migrate_page_norefs(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode) { return __buffer_migrate_page(mapping, newpage, page, mode, true); } |
9361401eb
|
803 |
#endif |
1d8b85ccf
|
804 |
|
04e62a29b
|
805 806 807 808 |
/* * Writeback a page to clean the dirty state */ static int writeout(struct address_space *mapping, struct page *page) |
8351a6e47
|
809 |
{ |
04e62a29b
|
810 811 812 813 814 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, .nr_to_write = 1, .range_start = 0, .range_end = LLONG_MAX, |
04e62a29b
|
815 816 817 818 819 820 821 822 823 824 825 |
.for_reclaim = 1 }; int rc; if (!mapping->a_ops->writepage) /* No write method for the address space */ return -EINVAL; if (!clear_page_dirty_for_io(page)) /* Someone else already triggered a write */ return -EAGAIN; |
8351a6e47
|
826 |
/* |
04e62a29b
|
827 828 829 830 831 832 |
* A dirty page may imply that the underlying filesystem has * the page on some queue. So the page must be clean for * migration. Writeout may mean we loose the lock and the * page state is no longer what we checked for earlier. * At this point we know that the migration attempt cannot * be successful. |
8351a6e47
|
833 |
*/ |
e388466de
|
834 |
remove_migration_ptes(page, page, false); |
8351a6e47
|
835 |
|
04e62a29b
|
836 |
rc = mapping->a_ops->writepage(page, &wbc); |
8351a6e47
|
837 |
|
04e62a29b
|
838 839 840 |
if (rc != AOP_WRITEPAGE_ACTIVATE) /* unlocked. Relock */ lock_page(page); |
bda8550de
|
841 |
return (rc < 0) ? -EIO : -EAGAIN; |
04e62a29b
|
842 843 844 845 846 847 |
} /* * Default handling if a filesystem does not provide a migration function. */ static int fallback_migrate_page(struct address_space *mapping, |
a6bc32b89
|
848 |
struct page *newpage, struct page *page, enum migrate_mode mode) |
04e62a29b
|
849 |
{ |
b969c4ab9
|
850 |
if (PageDirty(page)) { |
a6bc32b89
|
851 |
/* Only writeback pages in full synchronous migration */ |
2916ecc0f
|
852 853 854 855 856 |
switch (mode) { case MIGRATE_SYNC: case MIGRATE_SYNC_NO_COPY: break; default: |
b969c4ab9
|
857 |
return -EBUSY; |
2916ecc0f
|
858 |
} |
04e62a29b
|
859 |
return writeout(mapping, page); |
b969c4ab9
|
860 |
} |
8351a6e47
|
861 862 863 864 865 |
/* * Buffers may be managed in a filesystem specific way. * We must have no buffers or drop them. */ |
266cf658e
|
866 |
if (page_has_private(page) && |
8351a6e47
|
867 |
!try_to_release_page(page, GFP_KERNEL)) |
806031bb5
|
868 |
return mode == MIGRATE_SYNC ? -EAGAIN : -EBUSY; |
8351a6e47
|
869 |
|
a6bc32b89
|
870 |
return migrate_page(mapping, newpage, page, mode); |
8351a6e47
|
871 |
} |
1d8b85ccf
|
872 |
/* |
e24f0b8f7
|
873 874 875 876 877 |
* Move a page to a newly allocated page * The page is locked and all ptes have been successfully removed. * * The new page will have replaced the old page if this function * is successful. |
894bc3104
|
878 879 880 |
* * Return value: * < 0 - error code |
78bd52097
|
881 |
* MIGRATEPAGE_SUCCESS - success |
e24f0b8f7
|
882 |
*/ |
3fe2011ff
|
883 |
static int move_to_new_page(struct page *newpage, struct page *page, |
5c3f9a673
|
884 |
enum migrate_mode mode) |
e24f0b8f7
|
885 886 |
{ struct address_space *mapping; |
bda807d44
|
887 888 |
int rc = -EAGAIN; bool is_lru = !__PageMovable(page); |
e24f0b8f7
|
889 |
|
7db7671f8
|
890 891 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageLocked(newpage), newpage); |
e24f0b8f7
|
892 |
|
e24f0b8f7
|
893 |
mapping = page_mapping(page); |
bda807d44
|
894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 |
if (likely(is_lru)) { if (!mapping) rc = migrate_page(mapping, newpage, page, mode); else if (mapping->a_ops->migratepage) /* * Most pages have a mapping and most filesystems * provide a migratepage callback. Anonymous pages * are part of swap space which also has its own * migratepage callback. This is the most common path * for page migration. */ rc = mapping->a_ops->migratepage(mapping, newpage, page, mode); else rc = fallback_migrate_page(mapping, newpage, page, mode); } else { |
e24f0b8f7
|
912 |
/* |
bda807d44
|
913 914 |
* In case of non-lru page, it could be released after * isolation step. In that case, we shouldn't try migration. |
e24f0b8f7
|
915 |
*/ |
bda807d44
|
916 917 918 919 920 921 922 923 924 925 926 927 |
VM_BUG_ON_PAGE(!PageIsolated(page), page); if (!PageMovable(page)) { rc = MIGRATEPAGE_SUCCESS; __ClearPageIsolated(page); goto out; } rc = mapping->a_ops->migratepage(mapping, newpage, page, mode); WARN_ON_ONCE(rc == MIGRATEPAGE_SUCCESS && !PageIsolated(page)); } |
e24f0b8f7
|
928 |
|
5c3f9a673
|
929 930 931 932 933 |
/* * When successful, old pagecache page->mapping must be cleared before * page is freed; but stats require that PageAnon be left as PageAnon. */ if (rc == MIGRATEPAGE_SUCCESS) { |
bda807d44
|
934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 |
if (__PageMovable(page)) { VM_BUG_ON_PAGE(!PageIsolated(page), page); /* * We clear PG_movable under page_lock so any compactor * cannot try to migrate this page. */ __ClearPageIsolated(page); } /* * Anonymous and movable page->mapping will be cleard by * free_pages_prepare so don't reset it here for keeping * the type to work PageAnon, for example. */ if (!PageMappingFlags(page)) |
5c3f9a673
|
950 |
page->mapping = NULL; |
d2b2c6dd2
|
951 |
|
25b2995a3
|
952 |
if (likely(!is_zone_device_page(newpage))) |
d2b2c6dd2
|
953 |
flush_dcache_page(newpage); |
3fe2011ff
|
954 |
} |
bda807d44
|
955 |
out: |
e24f0b8f7
|
956 957 |
return rc; } |
0dabec93d
|
958 |
static int __unmap_and_move(struct page *page, struct page *newpage, |
9c620e2bc
|
959 |
int force, enum migrate_mode mode) |
e24f0b8f7
|
960 |
{ |
0dabec93d
|
961 |
int rc = -EAGAIN; |
2ebba6b7e
|
962 |
int page_was_mapped = 0; |
3f6c82728
|
963 |
struct anon_vma *anon_vma = NULL; |
bda807d44
|
964 |
bool is_lru = !__PageMovable(page); |
95a402c38
|
965 |
|
529ae9aaa
|
966 |
if (!trylock_page(page)) { |
a6bc32b89
|
967 |
if (!force || mode == MIGRATE_ASYNC) |
0dabec93d
|
968 |
goto out; |
3e7d34497
|
969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 |
/* * It's not safe for direct compaction to call lock_page. * For example, during page readahead pages are added locked * to the LRU. Later, when the IO completes the pages are * marked uptodate and unlocked. However, the queueing * could be merging multiple pages for one bio (e.g. * mpage_readpages). If an allocation happens for the * second or third page, the process can end up locking * the same page twice and deadlocking. Rather than * trying to be clever about what pages can be locked, * avoid the use of lock_page for direct compaction * altogether. */ if (current->flags & PF_MEMALLOC) |
0dabec93d
|
984 |
goto out; |
3e7d34497
|
985 |
|
e24f0b8f7
|
986 987 988 989 |
lock_page(page); } if (PageWriteback(page)) { |
11bc82d67
|
990 |
/* |
fed5b64a9
|
991 |
* Only in the case of a full synchronous migration is it |
a6bc32b89
|
992 993 994 |
* necessary to wait for PageWriteback. In the async case, * the retry loop is too short and in the sync-light case, * the overhead of stalling is too much |
11bc82d67
|
995 |
*/ |
2916ecc0f
|
996 997 998 999 1000 |
switch (mode) { case MIGRATE_SYNC: case MIGRATE_SYNC_NO_COPY: break; default: |
11bc82d67
|
1001 |
rc = -EBUSY; |
0a31bc97c
|
1002 |
goto out_unlock; |
11bc82d67
|
1003 1004 |
} if (!force) |
0a31bc97c
|
1005 |
goto out_unlock; |
e24f0b8f7
|
1006 1007 |
wait_on_page_writeback(page); } |
03f15c86c
|
1008 |
|
e24f0b8f7
|
1009 |
/* |
dc386d4d1
|
1010 1011 |
* By try_to_unmap(), page->mapcount goes down to 0 here. In this case, * we cannot notice that anon_vma is freed while we migrates a page. |
1ce82b69e
|
1012 |
* This get_anon_vma() delays freeing anon_vma pointer until the end |
dc386d4d1
|
1013 |
* of migration. File cache pages are no problem because of page_lock() |
989f89c57
|
1014 1015 |
* File Caches may use write_page() or lock_page() in migration, then, * just care Anon page here. |
03f15c86c
|
1016 1017 1018 1019 1020 1021 |
* * Only page_get_anon_vma() understands the subtleties of * getting a hold on an anon_vma from outside one of its mms. * But if we cannot get anon_vma, then we won't need it anyway, * because that implies that the anon page is no longer mapped * (and cannot be remapped so long as we hold the page lock). |
dc386d4d1
|
1022 |
*/ |
03f15c86c
|
1023 |
if (PageAnon(page) && !PageKsm(page)) |
746b18d42
|
1024 |
anon_vma = page_get_anon_vma(page); |
62e1c5530
|
1025 |
|
7db7671f8
|
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 |
/* * Block others from accessing the new page when we get around to * establishing additional references. We are usually the only one * holding a reference to newpage at this point. We used to have a BUG * here if trylock_page(newpage) fails, but would like to allow for * cases where there might be a race with the previous use of newpage. * This is much like races on refcount of oldpage: just don't BUG(). */ if (unlikely(!trylock_page(newpage))) goto out_unlock; |
bda807d44
|
1036 1037 1038 1039 |
if (unlikely(!is_lru)) { rc = move_to_new_page(newpage, page, mode); goto out_unlock_both; } |
dc386d4d1
|
1040 |
/* |
62e1c5530
|
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 |
* Corner case handling: * 1. When a new swap-cache page is read into, it is added to the LRU * and treated as swapcache but it has no rmap yet. * Calling try_to_unmap() against a page->mapping==NULL page will * trigger a BUG. So handle it here. * 2. An orphaned page (see truncate_complete_page) might have * fs-private metadata. The page can be picked up due to memory * offlining. Everywhere else except page reclaim, the page is * invisible to the vm, so the page can not be migrated. So try to * free the metadata, so the page can be freed. |
e24f0b8f7
|
1051 |
*/ |
62e1c5530
|
1052 |
if (!page->mapping) { |
309381fea
|
1053 |
VM_BUG_ON_PAGE(PageAnon(page), page); |
1ce82b69e
|
1054 |
if (page_has_private(page)) { |
62e1c5530
|
1055 |
try_to_free_buffers(page); |
7db7671f8
|
1056 |
goto out_unlock_both; |
62e1c5530
|
1057 |
} |
7db7671f8
|
1058 1059 |
} else if (page_mapped(page)) { /* Establish migration ptes */ |
03f15c86c
|
1060 1061 |
VM_BUG_ON_PAGE(PageAnon(page) && !PageKsm(page) && !anon_vma, page); |
2ebba6b7e
|
1062 |
try_to_unmap(page, |
da1b13ccf
|
1063 |
TTU_MIGRATION|TTU_IGNORE_MLOCK|TTU_IGNORE_ACCESS); |
2ebba6b7e
|
1064 1065 |
page_was_mapped = 1; } |
dc386d4d1
|
1066 |
|
e6a1530d6
|
1067 |
if (!page_mapped(page)) |
5c3f9a673
|
1068 |
rc = move_to_new_page(newpage, page, mode); |
e24f0b8f7
|
1069 |
|
5c3f9a673
|
1070 1071 |
if (page_was_mapped) remove_migration_ptes(page, |
e388466de
|
1072 |
rc == MIGRATEPAGE_SUCCESS ? newpage : page, false); |
3f6c82728
|
1073 |
|
7db7671f8
|
1074 1075 1076 |
out_unlock_both: unlock_page(newpage); out_unlock: |
3f6c82728
|
1077 |
/* Drop an anon_vma reference if we took one */ |
76545066c
|
1078 |
if (anon_vma) |
9e60109f1
|
1079 |
put_anon_vma(anon_vma); |
e24f0b8f7
|
1080 |
unlock_page(page); |
0dabec93d
|
1081 |
out: |
c6c919eb9
|
1082 1083 1084 1085 |
/* * If migration is successful, decrease refcount of the newpage * which will not free the page because new page owner increased * refcounter. As well, if it is LRU page, add the page to LRU |
e0a352fab
|
1086 1087 1088 1089 |
* list in here. Use the old state of the isolated source page to * determine if we migrated a LRU page. newpage was already unlocked * and possibly modified by its owner - don't rely on the page * state. |
c6c919eb9
|
1090 1091 |
*/ if (rc == MIGRATEPAGE_SUCCESS) { |
e0a352fab
|
1092 |
if (unlikely(!is_lru)) |
c6c919eb9
|
1093 1094 1095 1096 |
put_page(newpage); else putback_lru_page(newpage); } |
0dabec93d
|
1097 1098 |
return rc; } |
95a402c38
|
1099 |
|
0dabec93d
|
1100 |
/* |
ef2a5153b
|
1101 1102 1103 |
* gcc 4.7 and 4.8 on arm get an ICEs when inlining unmap_and_move(). Work * around it. */ |
815f0ddb3
|
1104 1105 |
#if defined(CONFIG_ARM) && \ defined(GCC_VERSION) && GCC_VERSION < 40900 && GCC_VERSION >= 40700 |
ef2a5153b
|
1106 1107 1108 1109 1110 1111 |
#define ICE_noinline noinline #else #define ICE_noinline #endif /* |
0dabec93d
|
1112 1113 1114 |
* Obtain the lock on page, remove all ptes and migrate the page * to the newly allocated page in newpage. */ |
ef2a5153b
|
1115 1116 1117 |
static ICE_noinline int unmap_and_move(new_page_t get_new_page, free_page_t put_new_page, unsigned long private, struct page *page, |
add05cece
|
1118 1119 |
int force, enum migrate_mode mode, enum migrate_reason reason) |
0dabec93d
|
1120 |
{ |
2def7424c
|
1121 |
int rc = MIGRATEPAGE_SUCCESS; |
2def7424c
|
1122 |
struct page *newpage; |
0dabec93d
|
1123 |
|
94723aafb
|
1124 1125 |
if (!thp_migration_supported() && PageTransHuge(page)) return -ENOMEM; |
666feb21a
|
1126 |
newpage = get_new_page(page, private); |
0dabec93d
|
1127 1128 1129 1130 1131 |
if (!newpage) return -ENOMEM; if (page_count(page) == 1) { /* page was freed from under us. So we are done. */ |
c6c919eb9
|
1132 1133 |
ClearPageActive(page); ClearPageUnevictable(page); |
bda807d44
|
1134 1135 1136 1137 1138 1139 |
if (unlikely(__PageMovable(page))) { lock_page(page); if (!PageMovable(page)) __ClearPageIsolated(page); unlock_page(page); } |
c6c919eb9
|
1140 1141 1142 1143 |
if (put_new_page) put_new_page(newpage, private); else put_page(newpage); |
0dabec93d
|
1144 1145 |
goto out; } |
9c620e2bc
|
1146 |
rc = __unmap_and_move(page, newpage, force, mode); |
c6c919eb9
|
1147 |
if (rc == MIGRATEPAGE_SUCCESS) |
7cd12b4ab
|
1148 |
set_page_owner_migrate_reason(newpage, reason); |
bf6bddf19
|
1149 |
|
0dabec93d
|
1150 |
out: |
e24f0b8f7
|
1151 |
if (rc != -EAGAIN) { |
0dabec93d
|
1152 1153 1154 1155 1156 1157 1158 |
/* * A page that has been migrated has all references * removed and will be freed. A page that has not been * migrated will have kepts its references and be * restored. */ list_del(&page->lru); |
6afcf8ef0
|
1159 1160 1161 1162 1163 1164 1165 |
/* * Compaction can migrate also non-LRU pages which are * not accounted to NR_ISOLATED_*. They can be recognized * as __PageMovable */ if (likely(!__PageMovable(page))) |
e8db67eb0
|
1166 1167 |
mod_node_page_state(page_pgdat(page), NR_ISOLATED_ANON + page_is_file_cache(page), -hpage_nr_pages(page)); |
c6c919eb9
|
1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 |
} /* * If migration is successful, releases reference grabbed during * isolation. Otherwise, restore the page to right list unless * we want to retry. */ if (rc == MIGRATEPAGE_SUCCESS) { put_page(page); if (reason == MR_MEMORY_FAILURE) { |
d7e69488b
|
1178 |
/* |
c6c919eb9
|
1179 1180 1181 |
* Set PG_HWPoison on just freed page * intentionally. Although it's rather weird, * it's how HWPoison flag works at the moment. |
d7e69488b
|
1182 |
*/ |
d4ae9916e
|
1183 |
if (set_hwpoison_free_buddy_page(page)) |
da1b13ccf
|
1184 |
num_poisoned_pages_inc(); |
c6c919eb9
|
1185 1186 |
} } else { |
bda807d44
|
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 |
if (rc != -EAGAIN) { if (likely(!__PageMovable(page))) { putback_lru_page(page); goto put_new; } lock_page(page); if (PageMovable(page)) putback_movable_page(page); else __ClearPageIsolated(page); unlock_page(page); put_page(page); } put_new: |
c6c919eb9
|
1202 1203 1204 1205 |
if (put_new_page) put_new_page(newpage, private); else put_page(newpage); |
e24f0b8f7
|
1206 |
} |
68711a746
|
1207 |
|
e24f0b8f7
|
1208 1209 1210 1211 |
return rc; } /* |
290408d4a
|
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 |
* Counterpart of unmap_and_move_page() for hugepage migration. * * This function doesn't wait the completion of hugepage I/O * because there is no race between I/O and migration for hugepage. * Note that currently hugepage I/O occurs only in direct I/O * where no lock is held and PG_writeback is irrelevant, * and writeback status of all subpages are counted in the reference * count of the head page (i.e. if all subpages of a 2MB hugepage are * under direct I/O, the reference of the head page is 512 and a bit more.) * This means that when we try to migrate hugepage whose subpages are * doing direct I/O, some references remain after try_to_unmap() and * hugepage migration fails without data corruption. * * There is also no race when direct I/O is issued on the page under migration, * because then pte is replaced with migration swap entry and direct I/O code * will wait in the page fault for migration to complete. */ static int unmap_and_move_huge_page(new_page_t get_new_page, |
68711a746
|
1230 1231 |
free_page_t put_new_page, unsigned long private, struct page *hpage, int force, |
7cd12b4ab
|
1232 |
enum migrate_mode mode, int reason) |
290408d4a
|
1233 |
{ |
2def7424c
|
1234 |
int rc = -EAGAIN; |
2ebba6b7e
|
1235 |
int page_was_mapped = 0; |
32665f2bb
|
1236 |
struct page *new_hpage; |
290408d4a
|
1237 |
struct anon_vma *anon_vma = NULL; |
83467efbd
|
1238 |
/* |
7ed2c31da
|
1239 |
* Migratability of hugepages depends on architectures and their size. |
83467efbd
|
1240 1241 1242 1243 1244 |
* This check is necessary because some callers of hugepage migration * like soft offline and memory hotremove don't walk through page * tables or check whether the hugepage is pmd-based or not before * kicking migration. */ |
100873d7a
|
1245 |
if (!hugepage_migration_supported(page_hstate(hpage))) { |
32665f2bb
|
1246 |
putback_active_hugepage(hpage); |
83467efbd
|
1247 |
return -ENOSYS; |
32665f2bb
|
1248 |
} |
83467efbd
|
1249 |
|
666feb21a
|
1250 |
new_hpage = get_new_page(hpage, private); |
290408d4a
|
1251 1252 |
if (!new_hpage) return -ENOMEM; |
290408d4a
|
1253 |
if (!trylock_page(hpage)) { |
2916ecc0f
|
1254 |
if (!force) |
290408d4a
|
1255 |
goto out; |
2916ecc0f
|
1256 1257 1258 1259 1260 1261 1262 |
switch (mode) { case MIGRATE_SYNC: case MIGRATE_SYNC_NO_COPY: break; default: goto out; } |
290408d4a
|
1263 1264 |
lock_page(hpage); } |
cb6acd01e
|
1265 1266 1267 1268 1269 1270 1271 1272 1273 |
/* * Check for pages which are in the process of being freed. Without * page_mapping() set, hugetlbfs specific move page routine will not * be called and we could leak usage counts for subpools. */ if (page_private(hpage) && !page_mapping(hpage)) { rc = -EBUSY; goto out_unlock; } |
746b18d42
|
1274 1275 |
if (PageAnon(hpage)) anon_vma = page_get_anon_vma(hpage); |
290408d4a
|
1276 |
|
7db7671f8
|
1277 1278 |
if (unlikely(!trylock_page(new_hpage))) goto put_anon; |
2ebba6b7e
|
1279 1280 |
if (page_mapped(hpage)) { try_to_unmap(hpage, |
ddeaab32a
|
1281 |
TTU_MIGRATION|TTU_IGNORE_MLOCK|TTU_IGNORE_ACCESS); |
2ebba6b7e
|
1282 1283 |
page_was_mapped = 1; } |
290408d4a
|
1284 1285 |
if (!page_mapped(hpage)) |
5c3f9a673
|
1286 |
rc = move_to_new_page(new_hpage, hpage, mode); |
290408d4a
|
1287 |
|
5c3f9a673
|
1288 1289 |
if (page_was_mapped) remove_migration_ptes(hpage, |
e388466de
|
1290 |
rc == MIGRATEPAGE_SUCCESS ? new_hpage : hpage, false); |
290408d4a
|
1291 |
|
7db7671f8
|
1292 1293 1294 |
unlock_page(new_hpage); put_anon: |
fd4a4663d
|
1295 |
if (anon_vma) |
9e60109f1
|
1296 |
put_anon_vma(anon_vma); |
8e6ac7fab
|
1297 |
|
2def7424c
|
1298 |
if (rc == MIGRATEPAGE_SUCCESS) { |
ab5ac90ae
|
1299 |
move_hugetlb_state(hpage, new_hpage, reason); |
2def7424c
|
1300 1301 |
put_new_page = NULL; } |
8e6ac7fab
|
1302 |
|
cb6acd01e
|
1303 |
out_unlock: |
290408d4a
|
1304 |
unlock_page(hpage); |
09761333e
|
1305 |
out: |
b8ec1cee5
|
1306 1307 |
if (rc != -EAGAIN) putback_active_hugepage(hpage); |
68711a746
|
1308 1309 1310 1311 1312 1313 |
/* * If migration was not successful and there's a freeing callback, use * it. Otherwise, put_page() will drop the reference grabbed during * isolation. */ |
2def7424c
|
1314 |
if (put_new_page) |
68711a746
|
1315 1316 |
put_new_page(new_hpage, private); else |
3aaa76e12
|
1317 |
putback_active_hugepage(new_hpage); |
68711a746
|
1318 |
|
290408d4a
|
1319 1320 1321 1322 |
return rc; } /* |
c73e5c9c5
|
1323 1324 |
* migrate_pages - migrate the pages specified in a list, to the free pages * supplied as the target for the page migration |
b20a35035
|
1325 |
* |
c73e5c9c5
|
1326 1327 1328 |
* @from: The list of pages to be migrated. * @get_new_page: The function used to allocate free pages to be used * as the target of the page migration. |
68711a746
|
1329 1330 |
* @put_new_page: The function used to free target pages if migration * fails, or NULL if no special handling is necessary. |
c73e5c9c5
|
1331 1332 1333 1334 |
* @private: Private data to be passed on to get_new_page() * @mode: The migration mode that specifies the constraints for * page migration, if any. * @reason: The reason for page migration. |
b20a35035
|
1335 |
* |
c73e5c9c5
|
1336 1337 |
* The function returns after 10 attempts or if no pages are movable any more * because the list has become empty or no retryable pages exist any more. |
14e0f9bcc
|
1338 |
* The caller should call putback_movable_pages() to return pages to the LRU |
28bd65781
|
1339 |
* or free list only if ret != 0. |
b20a35035
|
1340 |
* |
c73e5c9c5
|
1341 |
* Returns the number of pages that were not migrated, or an error code. |
b20a35035
|
1342 |
*/ |
9c620e2bc
|
1343 |
int migrate_pages(struct list_head *from, new_page_t get_new_page, |
68711a746
|
1344 1345 |
free_page_t put_new_page, unsigned long private, enum migrate_mode mode, int reason) |
b20a35035
|
1346 |
{ |
e24f0b8f7
|
1347 |
int retry = 1; |
b20a35035
|
1348 |
int nr_failed = 0; |
5647bc293
|
1349 |
int nr_succeeded = 0; |
b20a35035
|
1350 1351 1352 1353 1354 1355 1356 1357 |
int pass = 0; struct page *page; struct page *page2; int swapwrite = current->flags & PF_SWAPWRITE; int rc; if (!swapwrite) current->flags |= PF_SWAPWRITE; |
e24f0b8f7
|
1358 1359 |
for(pass = 0; pass < 10 && retry; pass++) { retry = 0; |
b20a35035
|
1360 |
|
e24f0b8f7
|
1361 |
list_for_each_entry_safe(page, page2, from, lru) { |
94723aafb
|
1362 |
retry: |
e24f0b8f7
|
1363 |
cond_resched(); |
2d1db3b11
|
1364 |
|
31caf665e
|
1365 1366 |
if (PageHuge(page)) rc = unmap_and_move_huge_page(get_new_page, |
68711a746
|
1367 |
put_new_page, private, page, |
7cd12b4ab
|
1368 |
pass > 2, mode, reason); |
31caf665e
|
1369 |
else |
68711a746
|
1370 |
rc = unmap_and_move(get_new_page, put_new_page, |
add05cece
|
1371 1372 |
private, page, pass > 2, mode, reason); |
2d1db3b11
|
1373 |
|
e24f0b8f7
|
1374 |
switch(rc) { |
95a402c38
|
1375 |
case -ENOMEM: |
94723aafb
|
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
/* * THP migration might be unsupported or the * allocation could've failed so we should * retry on the same page with the THP split * to base pages. * * Head page is retried immediately and tail * pages are added to the tail of the list so * we encounter them after the rest of the list * is processed. */ |
e6112fc30
|
1387 |
if (PageTransHuge(page) && !PageHuge(page)) { |
94723aafb
|
1388 1389 1390 1391 1392 1393 1394 1395 |
lock_page(page); rc = split_huge_page_to_list(page, from); unlock_page(page); if (!rc) { list_safe_reset_next(page, page2, lru); goto retry; } } |
dfef2ef40
|
1396 |
nr_failed++; |
95a402c38
|
1397 |
goto out; |
e24f0b8f7
|
1398 |
case -EAGAIN: |
2d1db3b11
|
1399 |
retry++; |
e24f0b8f7
|
1400 |
break; |
78bd52097
|
1401 |
case MIGRATEPAGE_SUCCESS: |
5647bc293
|
1402 |
nr_succeeded++; |
e24f0b8f7
|
1403 1404 |
break; default: |
354a33633
|
1405 1406 1407 1408 1409 1410 |
/* * Permanent failure (-EBUSY, -ENOSYS, etc.): * unlike -EAGAIN case, the failed page is * removed from migration page list and not * retried in the next outer loop. */ |
2d1db3b11
|
1411 |
nr_failed++; |
e24f0b8f7
|
1412 |
break; |
2d1db3b11
|
1413 |
} |
b20a35035
|
1414 1415 |
} } |
f2f81fb2b
|
1416 1417 |
nr_failed += retry; rc = nr_failed; |
95a402c38
|
1418 |
out: |
5647bc293
|
1419 1420 1421 1422 |
if (nr_succeeded) count_vm_events(PGMIGRATE_SUCCESS, nr_succeeded); if (nr_failed) count_vm_events(PGMIGRATE_FAIL, nr_failed); |
7b2a2d4a1
|
1423 |
trace_mm_migrate_pages(nr_succeeded, nr_failed, mode, reason); |
b20a35035
|
1424 1425 |
if (!swapwrite) current->flags &= ~PF_SWAPWRITE; |
78bd52097
|
1426 |
return rc; |
b20a35035
|
1427 |
} |
95a402c38
|
1428 |
|
742755a1d
|
1429 |
#ifdef CONFIG_NUMA |
742755a1d
|
1430 |
|
a49bd4d71
|
1431 |
static int store_status(int __user *status, int start, int value, int nr) |
742755a1d
|
1432 |
{ |
a49bd4d71
|
1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 |
while (nr-- > 0) { if (put_user(value, status + start)) return -EFAULT; start++; } return 0; } static int do_move_pages_to_node(struct mm_struct *mm, struct list_head *pagelist, int node) { int err; if (list_empty(pagelist)) return 0; err = migrate_pages(pagelist, alloc_new_node_page, NULL, node, MIGRATE_SYNC, MR_SYSCALL); if (err) putback_movable_pages(pagelist); return err; |
742755a1d
|
1455 1456 1457 |
} /* |
a49bd4d71
|
1458 1459 |
* Resolves the given address to a struct page, isolates it from the LRU and * puts it to the given pagelist. |
366b5dce6
|
1460 1461 1462 1463 1464 |
* Returns: * errno - if the page cannot be found/isolated * 0 - when it doesn't have to be migrated because it is already on the * target node * 1 - when it has been queued |
742755a1d
|
1465 |
*/ |
a49bd4d71
|
1466 1467 |
static int add_page_for_migration(struct mm_struct *mm, unsigned long addr, int node, struct list_head *pagelist, bool migrate_all) |
742755a1d
|
1468 |
{ |
a49bd4d71
|
1469 1470 1471 |
struct vm_area_struct *vma; struct page *page; unsigned int follflags; |
742755a1d
|
1472 |
int err; |
742755a1d
|
1473 1474 |
down_read(&mm->mmap_sem); |
a49bd4d71
|
1475 1476 1477 1478 |
err = -EFAULT; vma = find_vma(mm, addr); if (!vma || addr < vma->vm_start || !vma_migratable(vma)) goto out; |
742755a1d
|
1479 |
|
a49bd4d71
|
1480 1481 |
/* FOLL_DUMP to ignore special (like zero) pages */ follflags = FOLL_GET | FOLL_DUMP; |
a49bd4d71
|
1482 |
page = follow_page(vma, addr, follflags); |
89f5b7da2
|
1483 |
|
a49bd4d71
|
1484 1485 1486 |
err = PTR_ERR(page); if (IS_ERR(page)) goto out; |
89f5b7da2
|
1487 |
|
a49bd4d71
|
1488 1489 1490 |
err = -ENOENT; if (!page) goto out; |
742755a1d
|
1491 |
|
a49bd4d71
|
1492 1493 1494 |
err = 0; if (page_to_nid(page) == node) goto out_putpage; |
742755a1d
|
1495 |
|
a49bd4d71
|
1496 1497 1498 |
err = -EACCES; if (page_mapcount(page) > 1 && !migrate_all) goto out_putpage; |
742755a1d
|
1499 |
|
a49bd4d71
|
1500 1501 1502 |
if (PageHuge(page)) { if (PageHead(page)) { isolate_huge_page(page, pagelist); |
366b5dce6
|
1503 |
err = 1; |
e632a938d
|
1504 |
} |
a49bd4d71
|
1505 1506 |
} else { struct page *head; |
e632a938d
|
1507 |
|
e8db67eb0
|
1508 1509 |
head = compound_head(page); err = isolate_lru_page(head); |
cf608ac19
|
1510 |
if (err) |
a49bd4d71
|
1511 |
goto out_putpage; |
742755a1d
|
1512 |
|
366b5dce6
|
1513 |
err = 1; |
a49bd4d71
|
1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 |
list_add_tail(&head->lru, pagelist); mod_node_page_state(page_pgdat(head), NR_ISOLATED_ANON + page_is_file_cache(head), hpage_nr_pages(head)); } out_putpage: /* * Either remove the duplicate refcount from * isolate_lru_page() or drop the page ref if it was * not isolated. */ put_page(page); out: |
742755a1d
|
1527 1528 1529 1530 1531 |
up_read(&mm->mmap_sem); return err; } /* |
5e9a0f023
|
1532 1533 1534 |
* Migrate an array of page address onto an array of nodes and fill * the corresponding array of status. */ |
3268c63ed
|
1535 |
static int do_pages_move(struct mm_struct *mm, nodemask_t task_nodes, |
5e9a0f023
|
1536 1537 1538 1539 1540 |
unsigned long nr_pages, const void __user * __user *pages, const int __user *nodes, int __user *status, int flags) { |
a49bd4d71
|
1541 1542 1543 1544 |
int current_node = NUMA_NO_NODE; LIST_HEAD(pagelist); int start, i; int err = 0, err1; |
35282a2de
|
1545 1546 |
migrate_prep(); |
a49bd4d71
|
1547 1548 1549 1550 |
for (i = start = 0; i < nr_pages; i++) { const void __user *p; unsigned long addr; int node; |
3140a2273
|
1551 |
|
a49bd4d71
|
1552 1553 1554 1555 1556 |
err = -EFAULT; if (get_user(p, pages + i)) goto out_flush; if (get_user(node, nodes + i)) goto out_flush; |
057d33891
|
1557 |
addr = (unsigned long)untagged_addr(p); |
a49bd4d71
|
1558 1559 1560 1561 1562 1563 |
err = -ENODEV; if (node < 0 || node >= MAX_NUMNODES) goto out_flush; if (!node_state(node, N_MEMORY)) goto out_flush; |
5e9a0f023
|
1564 |
|
a49bd4d71
|
1565 1566 1567 1568 1569 1570 1571 1572 1573 |
err = -EACCES; if (!node_isset(node, task_nodes)) goto out_flush; if (current_node == NUMA_NO_NODE) { current_node = node; start = i; } else if (node != current_node) { err = do_move_pages_to_node(mm, &pagelist, current_node); |
d364e9b37
|
1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 |
if (err) { /* * Positive err means the number of failed * pages to migrate. Since we are going to * abort and return the number of non-migrated * pages, so need to incude the rest of the * nr_pages that have not been attempted as * well. */ if (err > 0) err += nr_pages - i - 1; |
a49bd4d71
|
1585 |
goto out; |
d364e9b37
|
1586 |
} |
a49bd4d71
|
1587 1588 1589 1590 1591 |
err = store_status(status, start, current_node, i - start); if (err) goto out; start = i; current_node = node; |
3140a2273
|
1592 |
} |
a49bd4d71
|
1593 1594 1595 1596 1597 1598 |
/* * Errors in the page lookup or isolation are not fatal and we simply * report them via status */ err = add_page_for_migration(mm, addr, current_node, &pagelist, flags & MPOL_MF_MOVE_ALL); |
366b5dce6
|
1599 1600 1601 1602 1603 1604 |
if (!err) { /* The page is already on the target node */ err = store_status(status, i, current_node, 1); if (err) goto out_flush; |
a49bd4d71
|
1605 |
continue; |
366b5dce6
|
1606 1607 1608 1609 |
} else if (err > 0) { /* The page is successfully queued for migration */ continue; } |
3140a2273
|
1610 |
|
a49bd4d71
|
1611 1612 1613 |
err = store_status(status, i, err, 1); if (err) goto out_flush; |
5e9a0f023
|
1614 |
|
a49bd4d71
|
1615 |
err = do_move_pages_to_node(mm, &pagelist, current_node); |
d364e9b37
|
1616 1617 1618 |
if (err) { if (err > 0) err += nr_pages - i - 1; |
a49bd4d71
|
1619 |
goto out; |
d364e9b37
|
1620 |
} |
a49bd4d71
|
1621 1622 1623 1624 1625 1626 |
if (i > start) { err = store_status(status, start, current_node, i - start); if (err) goto out; } current_node = NUMA_NO_NODE; |
3140a2273
|
1627 |
} |
a49bd4d71
|
1628 |
out_flush: |
8f175cf5c
|
1629 1630 |
if (list_empty(&pagelist)) return err; |
a49bd4d71
|
1631 1632 |
/* Make sure we do not overwrite the existing error */ err1 = do_move_pages_to_node(mm, &pagelist, current_node); |
d364e9b37
|
1633 1634 1635 1636 1637 1638 1639 |
/* * Don't have to report non-attempted pages here since: * - If the above loop is done gracefully all pages have been * attempted. * - If the above loop is aborted it means a fatal error * happened, should return ret. */ |
a49bd4d71
|
1640 1641 |
if (!err1) err1 = store_status(status, start, current_node, i - start); |
52cf138f5
|
1642 |
if (err >= 0) |
a49bd4d71
|
1643 |
err = err1; |
5e9a0f023
|
1644 1645 1646 1647 1648 |
out: return err; } /* |
2f007e74b
|
1649 |
* Determine the nodes of an array of pages and store it in an array of status. |
742755a1d
|
1650 |
*/ |
80bba1290
|
1651 1652 |
static void do_pages_stat_array(struct mm_struct *mm, unsigned long nr_pages, const void __user **pages, int *status) |
742755a1d
|
1653 |
{ |
2f007e74b
|
1654 |
unsigned long i; |
2f007e74b
|
1655 |
|
742755a1d
|
1656 |
down_read(&mm->mmap_sem); |
2f007e74b
|
1657 |
for (i = 0; i < nr_pages; i++) { |
80bba1290
|
1658 |
unsigned long addr = (unsigned long)(*pages); |
742755a1d
|
1659 1660 |
struct vm_area_struct *vma; struct page *page; |
c095adbc2
|
1661 |
int err = -EFAULT; |
2f007e74b
|
1662 1663 |
vma = find_vma(mm, addr); |
70384dc6d
|
1664 |
if (!vma || addr < vma->vm_start) |
742755a1d
|
1665 |
goto set_status; |
d899844e9
|
1666 1667 |
/* FOLL_DUMP to ignore special (like zero) pages */ page = follow_page(vma, addr, FOLL_DUMP); |
89f5b7da2
|
1668 1669 1670 1671 |
err = PTR_ERR(page); if (IS_ERR(page)) goto set_status; |
d899844e9
|
1672 |
err = page ? page_to_nid(page) : -ENOENT; |
742755a1d
|
1673 |
set_status: |
80bba1290
|
1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 |
*status = err; pages++; status++; } up_read(&mm->mmap_sem); } /* * Determine the nodes of a user array of pages and store it in * a user array of status. */ static int do_pages_stat(struct mm_struct *mm, unsigned long nr_pages, const void __user * __user *pages, int __user *status) { #define DO_PAGES_STAT_CHUNK_NR 16 const void __user *chunk_pages[DO_PAGES_STAT_CHUNK_NR]; int chunk_status[DO_PAGES_STAT_CHUNK_NR]; |
80bba1290
|
1694 |
|
87b8d1ade
|
1695 1696 |
while (nr_pages) { unsigned long chunk_nr; |
80bba1290
|
1697 |
|
87b8d1ade
|
1698 1699 1700 1701 1702 1703 |
chunk_nr = nr_pages; if (chunk_nr > DO_PAGES_STAT_CHUNK_NR) chunk_nr = DO_PAGES_STAT_CHUNK_NR; if (copy_from_user(chunk_pages, pages, chunk_nr * sizeof(*chunk_pages))) break; |
80bba1290
|
1704 1705 |
do_pages_stat_array(mm, chunk_nr, chunk_pages, chunk_status); |
87b8d1ade
|
1706 1707 |
if (copy_to_user(status, chunk_status, chunk_nr * sizeof(*status))) break; |
742755a1d
|
1708 |
|
87b8d1ade
|
1709 1710 1711 1712 1713 |
pages += chunk_nr; status += chunk_nr; nr_pages -= chunk_nr; } return nr_pages ? -EFAULT : 0; |
742755a1d
|
1714 1715 1716 1717 1718 1719 |
} /* * Move a list of pages in the address space of the currently executing * process. */ |
7addf4438
|
1720 1721 1722 1723 |
static int kernel_move_pages(pid_t pid, unsigned long nr_pages, const void __user * __user *pages, const int __user *nodes, int __user *status, int flags) |
742755a1d
|
1724 |
{ |
742755a1d
|
1725 |
struct task_struct *task; |
742755a1d
|
1726 |
struct mm_struct *mm; |
5e9a0f023
|
1727 |
int err; |
3268c63ed
|
1728 |
nodemask_t task_nodes; |
742755a1d
|
1729 1730 1731 1732 1733 1734 1735 1736 1737 |
/* Check flags */ if (flags & ~(MPOL_MF_MOVE|MPOL_MF_MOVE_ALL)) return -EINVAL; if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) return -EPERM; /* Find the mm_struct */ |
a879bf582
|
1738 |
rcu_read_lock(); |
228ebcbe6
|
1739 |
task = pid ? find_task_by_vpid(pid) : current; |
742755a1d
|
1740 |
if (!task) { |
a879bf582
|
1741 |
rcu_read_unlock(); |
742755a1d
|
1742 1743 |
return -ESRCH; } |
3268c63ed
|
1744 |
get_task_struct(task); |
742755a1d
|
1745 1746 1747 |
/* * Check if this process has the right to modify the specified |
197e7e521
|
1748 |
* process. Use the regular "ptrace_may_access()" checks. |
742755a1d
|
1749 |
*/ |
197e7e521
|
1750 |
if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS)) { |
c69e8d9c0
|
1751 |
rcu_read_unlock(); |
742755a1d
|
1752 |
err = -EPERM; |
5e9a0f023
|
1753 |
goto out; |
742755a1d
|
1754 |
} |
c69e8d9c0
|
1755 |
rcu_read_unlock(); |
742755a1d
|
1756 |
|
86c3a7645
|
1757 1758 |
err = security_task_movememory(task); if (err) |
5e9a0f023
|
1759 |
goto out; |
86c3a7645
|
1760 |
|
3268c63ed
|
1761 1762 1763 |
task_nodes = cpuset_mems_allowed(task); mm = get_task_mm(task); put_task_struct(task); |
6e8b09eaf
|
1764 1765 1766 1767 1768 1769 1770 1771 |
if (!mm) return -EINVAL; if (nodes) err = do_pages_move(mm, task_nodes, nr_pages, pages, nodes, status, flags); else err = do_pages_stat(mm, nr_pages, pages, status); |
742755a1d
|
1772 |
|
742755a1d
|
1773 1774 |
mmput(mm); return err; |
3268c63ed
|
1775 1776 1777 1778 |
out: put_task_struct(task); return err; |
742755a1d
|
1779 |
} |
742755a1d
|
1780 |
|
7addf4438
|
1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 |
SYSCALL_DEFINE6(move_pages, pid_t, pid, unsigned long, nr_pages, const void __user * __user *, pages, const int __user *, nodes, int __user *, status, int, flags) { return kernel_move_pages(pid, nr_pages, pages, nodes, status, flags); } #ifdef CONFIG_COMPAT COMPAT_SYSCALL_DEFINE6(move_pages, pid_t, pid, compat_ulong_t, nr_pages, compat_uptr_t __user *, pages32, const int __user *, nodes, int __user *, status, int, flags) { const void __user * __user *pages; int i; pages = compat_alloc_user_space(nr_pages * sizeof(void *)); for (i = 0; i < nr_pages; i++) { compat_uptr_t p; if (get_user(p, pages32 + i) || put_user(compat_ptr(p), pages + i)) return -EFAULT; } return kernel_move_pages(pid, nr_pages, pages, nodes, status, flags); } #endif /* CONFIG_COMPAT */ |
7039e1dbe
|
1810 1811 1812 1813 1814 1815 |
#ifdef CONFIG_NUMA_BALANCING /* * Returns true if this is a safe migration target node for misplaced NUMA * pages. Currently it only checks the watermarks which crude */ static bool migrate_balanced_pgdat(struct pglist_data *pgdat, |
3abef4e6c
|
1816 |
unsigned long nr_migrate_pages) |
7039e1dbe
|
1817 1818 |
{ int z; |
599d0c954
|
1819 |
|
7039e1dbe
|
1820 1821 1822 1823 1824 |
for (z = pgdat->nr_zones - 1; z >= 0; z--) { struct zone *zone = pgdat->node_zones + z; if (!populated_zone(zone)) continue; |
7039e1dbe
|
1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 |
/* Avoid waking kswapd by allocating pages_to_migrate pages. */ if (!zone_watermark_ok(zone, 0, high_wmark_pages(zone) + nr_migrate_pages, 0, 0)) continue; return true; } return false; } static struct page *alloc_misplaced_dst_page(struct page *page, |
666feb21a
|
1837 |
unsigned long data) |
7039e1dbe
|
1838 1839 1840 |
{ int nid = (int) data; struct page *newpage; |
96db800f5
|
1841 |
newpage = __alloc_pages_node(nid, |
e97ca8e5b
|
1842 1843 1844 |
(GFP_HIGHUSER_MOVABLE | __GFP_THISNODE | __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN) & |
8479eba77
|
1845 |
~__GFP_RECLAIM, 0); |
bac0382c6
|
1846 |
|
7039e1dbe
|
1847 1848 |
return newpage; } |
1c30e0177
|
1849 |
static int numamigrate_isolate_page(pg_data_t *pgdat, struct page *page) |
b32967ff1
|
1850 |
{ |
340ef3902
|
1851 |
int page_lru; |
a8f607721
|
1852 |
|
309381fea
|
1853 |
VM_BUG_ON_PAGE(compound_order(page) && !PageTransHuge(page), page); |
3abef4e6c
|
1854 |
|
7039e1dbe
|
1855 |
/* Avoid migrating to a node that is nearly full */ |
d8c6546b1
|
1856 |
if (!migrate_balanced_pgdat(pgdat, compound_nr(page))) |
340ef3902
|
1857 |
return 0; |
7039e1dbe
|
1858 |
|
340ef3902
|
1859 1860 |
if (isolate_lru_page(page)) return 0; |
7039e1dbe
|
1861 |
|
340ef3902
|
1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 |
/* * migrate_misplaced_transhuge_page() skips page migration's usual * check on page_count(), so we must do it here, now that the page * has been isolated: a GUP pin, or any other pin, prevents migration. * The expected page count is 3: 1 for page's mapcount and 1 for the * caller's pin and 1 for the reference taken by isolate_lru_page(). */ if (PageTransHuge(page) && page_count(page) != 3) { putback_lru_page(page); return 0; |
7039e1dbe
|
1872 |
} |
340ef3902
|
1873 |
page_lru = page_is_file_cache(page); |
599d0c954
|
1874 |
mod_node_page_state(page_pgdat(page), NR_ISOLATED_ANON + page_lru, |
340ef3902
|
1875 |
hpage_nr_pages(page)); |
149c33e1c
|
1876 |
/* |
340ef3902
|
1877 1878 1879 |
* Isolating the page has taken another reference, so the * caller's reference can be safely dropped without the page * disappearing underneath us during migration. |
149c33e1c
|
1880 1881 |
*/ put_page(page); |
340ef3902
|
1882 |
return 1; |
b32967ff1
|
1883 |
} |
de466bd62
|
1884 1885 1886 1887 1888 |
bool pmd_trans_migrating(pmd_t pmd) { struct page *page = pmd_page(pmd); return PageLocked(page); } |
b32967ff1
|
1889 1890 1891 1892 1893 |
/* * Attempt to migrate a misplaced page to the specified destination * node. Caller is expected to have an elevated reference count on * the page that will be dropped by this function before returning. */ |
1bc115d87
|
1894 1895 |
int migrate_misplaced_page(struct page *page, struct vm_area_struct *vma, int node) |
b32967ff1
|
1896 1897 |
{ pg_data_t *pgdat = NODE_DATA(node); |
340ef3902
|
1898 |
int isolated; |
b32967ff1
|
1899 1900 1901 1902 |
int nr_remaining; LIST_HEAD(migratepages); /* |
1bc115d87
|
1903 1904 |
* Don't migrate file pages that are mapped in multiple processes * with execute permissions as they are probably shared libraries. |
b32967ff1
|
1905 |
*/ |
1bc115d87
|
1906 1907 |
if (page_mapcount(page) != 1 && page_is_file_cache(page) && (vma->vm_flags & VM_EXEC)) |
b32967ff1
|
1908 |
goto out; |
b32967ff1
|
1909 1910 |
/* |
09a913a7a
|
1911 1912 1913 1914 1915 |
* Also do not migrate dirty pages as not all filesystems can move * dirty pages in MIGRATE_ASYNC mode which is a waste of cycles. */ if (page_is_file_cache(page) && PageDirty(page)) goto out; |
b32967ff1
|
1916 1917 1918 1919 1920 |
isolated = numamigrate_isolate_page(pgdat, page); if (!isolated) goto out; list_add(&page->lru, &migratepages); |
9c620e2bc
|
1921 |
nr_remaining = migrate_pages(&migratepages, alloc_misplaced_dst_page, |
68711a746
|
1922 1923 |
NULL, node, MIGRATE_ASYNC, MR_NUMA_MISPLACED); |
b32967ff1
|
1924 |
if (nr_remaining) { |
59c82b70d
|
1925 1926 |
if (!list_empty(&migratepages)) { list_del(&page->lru); |
599d0c954
|
1927 |
dec_node_page_state(page, NR_ISOLATED_ANON + |
59c82b70d
|
1928 1929 1930 |
page_is_file_cache(page)); putback_lru_page(page); } |
b32967ff1
|
1931 1932 1933 |
isolated = 0; } else count_vm_numa_event(NUMA_PAGE_MIGRATE); |
7039e1dbe
|
1934 |
BUG_ON(!list_empty(&migratepages)); |
7039e1dbe
|
1935 |
return isolated; |
340ef3902
|
1936 1937 1938 1939 |
out: put_page(page); return 0; |
7039e1dbe
|
1940 |
} |
220018d38
|
1941 |
#endif /* CONFIG_NUMA_BALANCING */ |
b32967ff1
|
1942 |
|
220018d38
|
1943 |
#if defined(CONFIG_NUMA_BALANCING) && defined(CONFIG_TRANSPARENT_HUGEPAGE) |
340ef3902
|
1944 1945 1946 1947 |
/* * Migrates a THP to a given target node. page must be locked and is unlocked * before returning. */ |
b32967ff1
|
1948 1949 1950 1951 1952 1953 |
int migrate_misplaced_transhuge_page(struct mm_struct *mm, struct vm_area_struct *vma, pmd_t *pmd, pmd_t entry, unsigned long address, struct page *page, int node) { |
c4088ebdc
|
1954 |
spinlock_t *ptl; |
b32967ff1
|
1955 1956 1957 |
pg_data_t *pgdat = NODE_DATA(node); int isolated = 0; struct page *new_page = NULL; |
b32967ff1
|
1958 |
int page_lru = page_is_file_cache(page); |
7066f0f93
|
1959 |
unsigned long start = address & HPAGE_PMD_MASK; |
b32967ff1
|
1960 |
|
b32967ff1
|
1961 |
new_page = alloc_pages_node(node, |
251603549
|
1962 |
(GFP_TRANSHUGE_LIGHT | __GFP_THISNODE), |
e97ca8e5b
|
1963 |
HPAGE_PMD_ORDER); |
340ef3902
|
1964 1965 |
if (!new_page) goto out_fail; |
9a982250f
|
1966 |
prep_transhuge_page(new_page); |
340ef3902
|
1967 |
|
b32967ff1
|
1968 |
isolated = numamigrate_isolate_page(pgdat, page); |
340ef3902
|
1969 |
if (!isolated) { |
b32967ff1
|
1970 |
put_page(new_page); |
340ef3902
|
1971 |
goto out_fail; |
b32967ff1
|
1972 |
} |
b0943d61b
|
1973 |
|
b32967ff1
|
1974 |
/* Prepare a page as a migration target */ |
48c935ad8
|
1975 |
__SetPageLocked(new_page); |
d44d363f6
|
1976 1977 |
if (PageSwapBacked(page)) __SetPageSwapBacked(new_page); |
b32967ff1
|
1978 1979 1980 1981 |
/* anon mapping, we can simply copy page->mapping to the new page: */ new_page->mapping = page->mapping; new_page->index = page->index; |
7eef5f97c
|
1982 1983 |
/* flush the cache before copying using the kernel virtual address */ flush_cache_range(vma, start, start + HPAGE_PMD_SIZE); |
b32967ff1
|
1984 1985 1986 1987 |
migrate_page_copy(new_page, page); WARN_ON(PageLRU(new_page)); /* Recheck the target PMD */ |
c4088ebdc
|
1988 |
ptl = pmd_lock(mm, pmd); |
f4e177d12
|
1989 |
if (unlikely(!pmd_same(*pmd, entry) || !page_ref_freeze(page, 2))) { |
c4088ebdc
|
1990 |
spin_unlock(ptl); |
b32967ff1
|
1991 1992 1993 1994 1995 1996 |
/* Reverse changes made by migrate_page_copy() */ if (TestClearPageActive(new_page)) SetPageActive(page); if (TestClearPageUnevictable(new_page)) SetPageUnevictable(page); |
b32967ff1
|
1997 1998 1999 |
unlock_page(new_page); put_page(new_page); /* Free it */ |
a54a407fb
|
2000 2001 |
/* Retake the callers reference and putback on LRU */ get_page(page); |
b32967ff1
|
2002 |
putback_lru_page(page); |
599d0c954
|
2003 |
mod_node_page_state(page_pgdat(page), |
a54a407fb
|
2004 |
NR_ISOLATED_ANON + page_lru, -HPAGE_PMD_NR); |
eb4489f69
|
2005 2006 |
goto out_unlock; |
b32967ff1
|
2007 |
} |
101024596
|
2008 |
entry = mk_huge_pmd(new_page, vma->vm_page_prot); |
f55e1014f
|
2009 |
entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma); |
b32967ff1
|
2010 |
|
2b4847e73
|
2011 |
/* |
d7c339341
|
2012 2013 2014 2015 2016 2017 |
* Overwrite the old entry under pagetable lock and establish * the new PTE. Any parallel GUP will either observe the old * page blocking on the page lock, block on the page table * lock or observe the new page. The SetPageUptodate on the * new page and page_add_new_anon_rmap guarantee the copy is * visible before the pagetable update. |
2b4847e73
|
2018 |
*/ |
7066f0f93
|
2019 |
page_add_anon_rmap(new_page, vma, start, true); |
d7c339341
|
2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 |
/* * At this point the pmd is numa/protnone (i.e. non present) and the TLB * has already been flushed globally. So no TLB can be currently * caching this non present pmd mapping. There's no need to clear the * pmd before doing set_pmd_at(), nor to flush the TLB after * set_pmd_at(). Clearing the pmd here would introduce a race * condition against MADV_DONTNEED, because MADV_DONTNEED only holds the * mmap_sem for reading. If the pmd is set to NULL at any given time, * MADV_DONTNEED won't wait on the pmd lock and it'll skip clearing this * pmd. */ |
7066f0f93
|
2031 |
set_pmd_at(mm, start, pmd, entry); |
ce4a9cc57
|
2032 |
update_mmu_cache_pmd(vma, address, &entry); |
2b4847e73
|
2033 |
|
f4e177d12
|
2034 |
page_ref_unfreeze(page, 2); |
51afb12ba
|
2035 |
mlock_migrate_page(new_page, page); |
d281ee614
|
2036 |
page_remove_rmap(page, true); |
7cd12b4ab
|
2037 |
set_page_owner_migrate_reason(new_page, MR_NUMA_MISPLACED); |
2b4847e73
|
2038 |
|
c4088ebdc
|
2039 |
spin_unlock(ptl); |
b32967ff1
|
2040 |
|
11de9927f
|
2041 2042 2043 |
/* Take an "isolate" reference and put new page on the LRU. */ get_page(new_page); putback_lru_page(new_page); |
b32967ff1
|
2044 2045 2046 2047 2048 2049 2050 |
unlock_page(new_page); unlock_page(page); put_page(page); /* Drop the rmap reference */ put_page(page); /* Drop the LRU isolation reference */ count_vm_events(PGMIGRATE_SUCCESS, HPAGE_PMD_NR); count_vm_numa_events(NUMA_PAGE_MIGRATE, HPAGE_PMD_NR); |
599d0c954
|
2051 |
mod_node_page_state(page_pgdat(page), |
b32967ff1
|
2052 2053 2054 |
NR_ISOLATED_ANON + page_lru, -HPAGE_PMD_NR); return isolated; |
340ef3902
|
2055 2056 |
out_fail: count_vm_events(PGMIGRATE_FAIL, HPAGE_PMD_NR); |
2b4847e73
|
2057 2058 |
ptl = pmd_lock(mm, pmd); if (pmd_same(*pmd, entry)) { |
4d9424669
|
2059 |
entry = pmd_modify(entry, vma->vm_page_prot); |
7066f0f93
|
2060 |
set_pmd_at(mm, start, pmd, entry); |
2b4847e73
|
2061 2062 2063 |
update_mmu_cache_pmd(vma, address, &entry); } spin_unlock(ptl); |
a54a407fb
|
2064 |
|
eb4489f69
|
2065 |
out_unlock: |
340ef3902
|
2066 |
unlock_page(page); |
b32967ff1
|
2067 |
put_page(page); |
b32967ff1
|
2068 2069 |
return 0; } |
7039e1dbe
|
2070 2071 2072 |
#endif /* CONFIG_NUMA_BALANCING */ #endif /* CONFIG_NUMA */ |
8763cb45a
|
2073 |
|
9b2ed9cb9
|
2074 |
#ifdef CONFIG_DEVICE_PRIVATE |
8763cb45a
|
2075 2076 2077 2078 2079 2080 2081 2082 |
static int migrate_vma_collect_hole(unsigned long start, unsigned long end, struct mm_walk *walk) { struct migrate_vma *migrate = walk->private; unsigned long addr; for (addr = start & PAGE_MASK; addr < end; addr += PAGE_SIZE) { |
e20d103b6
|
2083 |
migrate->src[migrate->npages] = MIGRATE_PFN_MIGRATE; |
8315ada7f
|
2084 |
migrate->dst[migrate->npages] = 0; |
e20d103b6
|
2085 |
migrate->npages++; |
8315ada7f
|
2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 |
migrate->cpages++; } return 0; } static int migrate_vma_collect_skip(unsigned long start, unsigned long end, struct mm_walk *walk) { struct migrate_vma *migrate = walk->private; unsigned long addr; for (addr = start & PAGE_MASK; addr < end; addr += PAGE_SIZE) { |
8763cb45a
|
2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 |
migrate->dst[migrate->npages] = 0; migrate->src[migrate->npages++] = 0; } return 0; } static int migrate_vma_collect_pmd(pmd_t *pmdp, unsigned long start, unsigned long end, struct mm_walk *walk) { struct migrate_vma *migrate = walk->private; struct vm_area_struct *vma = walk->vma; struct mm_struct *mm = vma->vm_mm; |
8c3328f1f
|
2115 |
unsigned long addr = start, unmapped = 0; |
8763cb45a
|
2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 |
spinlock_t *ptl; pte_t *ptep; again: if (pmd_none(*pmdp)) return migrate_vma_collect_hole(start, end, walk); if (pmd_trans_huge(*pmdp)) { struct page *page; ptl = pmd_lock(mm, pmdp); if (unlikely(!pmd_trans_huge(*pmdp))) { spin_unlock(ptl); goto again; } page = pmd_page(*pmdp); if (is_huge_zero_page(page)) { spin_unlock(ptl); split_huge_pmd(vma, pmdp, addr); if (pmd_trans_unstable(pmdp)) |
8315ada7f
|
2137 |
return migrate_vma_collect_skip(start, end, |
8763cb45a
|
2138 2139 2140 2141 2142 2143 2144 |
walk); } else { int ret; get_page(page); spin_unlock(ptl); if (unlikely(!trylock_page(page))) |
8315ada7f
|
2145 |
return migrate_vma_collect_skip(start, end, |
8763cb45a
|
2146 2147 2148 2149 |
walk); ret = split_huge_page(page); unlock_page(page); put_page(page); |
8315ada7f
|
2150 2151 2152 2153 |
if (ret) return migrate_vma_collect_skip(start, end, walk); if (pmd_none(*pmdp)) |
8763cb45a
|
2154 2155 2156 2157 2158 2159 |
return migrate_vma_collect_hole(start, end, walk); } } if (unlikely(pmd_bad(*pmdp))) |
8315ada7f
|
2160 |
return migrate_vma_collect_skip(start, end, walk); |
8763cb45a
|
2161 2162 |
ptep = pte_offset_map_lock(mm, pmdp, addr, &ptl); |
8c3328f1f
|
2163 |
arch_enter_lazy_mmu_mode(); |
8763cb45a
|
2164 2165 2166 |
for (; addr < end; addr += PAGE_SIZE, ptep++) { unsigned long mpfn, pfn; struct page *page; |
8c3328f1f
|
2167 |
swp_entry_t entry; |
8763cb45a
|
2168 2169 2170 |
pte_t pte; pte = *ptep; |
8763cb45a
|
2171 |
|
a5430dda8
|
2172 |
if (pte_none(pte)) { |
8315ada7f
|
2173 2174 |
mpfn = MIGRATE_PFN_MIGRATE; migrate->cpages++; |
8763cb45a
|
2175 2176 |
goto next; } |
a5430dda8
|
2177 |
if (!pte_present(pte)) { |
276f756d7
|
2178 |
mpfn = 0; |
a5430dda8
|
2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 |
/* * Only care about unaddressable device page special * page table entry. Other special swap entries are not * migratable, and we ignore regular swapped page. */ entry = pte_to_swp_entry(pte); if (!is_device_private_entry(entry)) goto next; page = device_private_entry_to_page(entry); |
06d462beb
|
2190 2191 |
mpfn = migrate_pfn(page_to_pfn(page)) | MIGRATE_PFN_MIGRATE; |
a5430dda8
|
2192 2193 2194 |
if (is_write_device_private_entry(entry)) mpfn |= MIGRATE_PFN_WRITE; } else { |
276f756d7
|
2195 |
pfn = pte_pfn(pte); |
8315ada7f
|
2196 2197 2198 |
if (is_zero_pfn(pfn)) { mpfn = MIGRATE_PFN_MIGRATE; migrate->cpages++; |
8315ada7f
|
2199 2200 |
goto next; } |
25b2995a3
|
2201 |
page = vm_normal_page(migrate->vma, addr, pte); |
a5430dda8
|
2202 2203 2204 |
mpfn = migrate_pfn(pfn) | MIGRATE_PFN_MIGRATE; mpfn |= pte_write(pte) ? MIGRATE_PFN_WRITE : 0; } |
8763cb45a
|
2205 |
/* FIXME support THP */ |
8763cb45a
|
2206 |
if (!page || !page->mapping || PageTransCompound(page)) { |
276f756d7
|
2207 |
mpfn = 0; |
8763cb45a
|
2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 |
goto next; } /* * By getting a reference on the page we pin it and that blocks * any kind of migration. Side effect is that it "freezes" the * pte. * * We drop this reference after isolating the page from the lru * for non device page (device page are not on the lru and thus * can't be dropped from it). */ get_page(page); migrate->cpages++; |
8763cb45a
|
2222 |
|
8c3328f1f
|
2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 |
/* * Optimize for the common case where page is only mapped once * in one process. If we can lock the page, then we can safely * set up a special migration page table entry now. */ if (trylock_page(page)) { pte_t swp_pte; mpfn |= MIGRATE_PFN_LOCKED; ptep_get_and_clear(mm, addr, ptep); /* Setup special migration page table entry */ |
07707125a
|
2235 2236 |
entry = make_migration_entry(page, mpfn & MIGRATE_PFN_WRITE); |
8c3328f1f
|
2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 |
swp_pte = swp_entry_to_pte(entry); if (pte_soft_dirty(pte)) swp_pte = pte_swp_mksoft_dirty(swp_pte); set_pte_at(mm, addr, ptep, swp_pte); /* * This is like regular unmap: we remove the rmap and * drop page refcount. Page won't be freed, as we took * a reference just above. */ page_remove_rmap(page, false); put_page(page); |
a5430dda8
|
2249 2250 2251 |
if (pte_present(pte)) unmapped++; |
8c3328f1f
|
2252 |
} |
8763cb45a
|
2253 |
next: |
a5430dda8
|
2254 |
migrate->dst[migrate->npages] = 0; |
8763cb45a
|
2255 2256 |
migrate->src[migrate->npages++] = mpfn; } |
8c3328f1f
|
2257 |
arch_leave_lazy_mmu_mode(); |
8763cb45a
|
2258 |
pte_unmap_unlock(ptep - 1, ptl); |
8c3328f1f
|
2259 2260 2261 |
/* Only flush the TLB if we actually modified any entries */ if (unmapped) flush_tlb_range(walk->vma, start, end); |
8763cb45a
|
2262 2263 |
return 0; } |
7b86ac337
|
2264 2265 2266 2267 |
static const struct mm_walk_ops migrate_vma_walk_ops = { .pmd_entry = migrate_vma_collect_pmd, .pte_hole = migrate_vma_collect_hole, }; |
8763cb45a
|
2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 |
/* * migrate_vma_collect() - collect pages over a range of virtual addresses * @migrate: migrate struct containing all migration information * * This will walk the CPU page table. For each virtual address backed by a * valid page, it updates the src array and takes a reference on the page, in * order to pin the page until we lock it and unmap it. */ static void migrate_vma_collect(struct migrate_vma *migrate) { |
ac46d4f3c
|
2278 |
struct mmu_notifier_range range; |
8763cb45a
|
2279 |
|
7b86ac337
|
2280 2281 |
mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, NULL, migrate->vma->vm_mm, migrate->start, migrate->end); |
ac46d4f3c
|
2282 |
mmu_notifier_invalidate_range_start(&range); |
8763cb45a
|
2283 |
|
7b86ac337
|
2284 2285 2286 2287 |
walk_page_range(migrate->vma->vm_mm, migrate->start, migrate->end, &migrate_vma_walk_ops, migrate); mmu_notifier_invalidate_range_end(&range); |
8763cb45a
|
2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 |
migrate->end = migrate->start + (migrate->npages << PAGE_SHIFT); } /* * migrate_vma_check_page() - check if page is pinned or not * @page: struct page to check * * Pinned pages cannot be migrated. This is the same test as in * migrate_page_move_mapping(), except that here we allow migration of a * ZONE_DEVICE page. */ static bool migrate_vma_check_page(struct page *page) { /* * One extra ref because caller holds an extra reference, either from * isolate_lru_page() for a regular page, or migrate_vma_collect() for * a device page. */ int extra = 1; /* * FIXME support THP (transparent huge page), it is bit more complex to * check them than regular pages, because they can be mapped with a pmd * or with a pte (split pte mapping). */ if (PageCompound(page)) return false; |
a5430dda8
|
2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 |
/* Page from ZONE_DEVICE have one extra reference */ if (is_zone_device_page(page)) { /* * Private page can never be pin as they have no valid pte and * GUP will fail for those. Yet if there is a pending migration * a thread might try to wait on the pte migration entry and * will bump the page reference count. Sadly there is no way to * differentiate a regular pin from migration wait. Hence to * avoid 2 racing thread trying to migrate back to CPU to enter * infinite loop (one stoping migration because the other is * waiting on pte migration entry). We always return true here. * * FIXME proper solution is to rework migration_entry_wait() so * it does not need to take a reference on page. */ |
25b2995a3
|
2330 |
return is_device_private_page(page); |
a5430dda8
|
2331 |
} |
df6ad6983
|
2332 2333 2334 |
/* For file back page */ if (page_mapping(page)) extra += 1 + page_has_private(page); |
8763cb45a
|
2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 |
if ((page_count(page) - extra) > page_mapcount(page)) return false; return true; } /* * migrate_vma_prepare() - lock pages and isolate them from the lru * @migrate: migrate struct containing all migration information * * This locks pages that have been collected by migrate_vma_collect(). Once each * page is locked it is isolated from the lru (for non-device pages). Finally, * the ref taken by migrate_vma_collect() is dropped, as locked pages cannot be * migrated by concurrent kernel threads. */ static void migrate_vma_prepare(struct migrate_vma *migrate) { const unsigned long npages = migrate->npages; |
8c3328f1f
|
2353 2354 |
const unsigned long start = migrate->start; unsigned long addr, i, restore = 0; |
8763cb45a
|
2355 |
bool allow_drain = true; |
8763cb45a
|
2356 2357 2358 2359 2360 |
lru_add_drain(); for (i = 0; (i < npages) && migrate->cpages; i++) { struct page *page = migrate_pfn_to_page(migrate->src[i]); |
8c3328f1f
|
2361 |
bool remap = true; |
8763cb45a
|
2362 2363 2364 |
if (!page) continue; |
8c3328f1f
|
2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 |
if (!(migrate->src[i] & MIGRATE_PFN_LOCKED)) { /* * Because we are migrating several pages there can be * a deadlock between 2 concurrent migration where each * are waiting on each other page lock. * * Make migrate_vma() a best effort thing and backoff * for any page we can not lock right away. */ if (!trylock_page(page)) { migrate->src[i] = 0; migrate->cpages--; put_page(page); continue; } remap = false; migrate->src[i] |= MIGRATE_PFN_LOCKED; |
8763cb45a
|
2382 |
} |
8763cb45a
|
2383 |
|
a5430dda8
|
2384 2385 2386 2387 2388 2389 2390 |
/* ZONE_DEVICE pages are not on LRU */ if (!is_zone_device_page(page)) { if (!PageLRU(page) && allow_drain) { /* Drain CPU's pagevec */ lru_add_drain_all(); allow_drain = false; } |
8763cb45a
|
2391 |
|
a5430dda8
|
2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 |
if (isolate_lru_page(page)) { if (remap) { migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; migrate->cpages--; restore++; } else { migrate->src[i] = 0; unlock_page(page); migrate->cpages--; put_page(page); } continue; |
8c3328f1f
|
2404 |
} |
a5430dda8
|
2405 2406 2407 |
/* Drop the reference we took in collect */ put_page(page); |
8763cb45a
|
2408 2409 2410 |
} if (!migrate_vma_check_page(page)) { |
8c3328f1f
|
2411 2412 2413 2414 |
if (remap) { migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; migrate->cpages--; restore++; |
8763cb45a
|
2415 |
|
a5430dda8
|
2416 2417 2418 2419 |
if (!is_zone_device_page(page)) { get_page(page); putback_lru_page(page); } |
8c3328f1f
|
2420 2421 2422 2423 |
} else { migrate->src[i] = 0; unlock_page(page); migrate->cpages--; |
a5430dda8
|
2424 2425 2426 2427 |
if (!is_zone_device_page(page)) putback_lru_page(page); else put_page(page); |
8c3328f1f
|
2428 |
} |
8763cb45a
|
2429 2430 |
} } |
8c3328f1f
|
2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 |
for (i = 0, addr = start; i < npages && restore; i++, addr += PAGE_SIZE) { struct page *page = migrate_pfn_to_page(migrate->src[i]); if (!page || (migrate->src[i] & MIGRATE_PFN_MIGRATE)) continue; remove_migration_pte(page, migrate->vma, addr, page); migrate->src[i] = 0; unlock_page(page); put_page(page); restore--; } |
8763cb45a
|
2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 |
} /* * migrate_vma_unmap() - replace page mapping with special migration pte entry * @migrate: migrate struct containing all migration information * * Replace page mapping (CPU page table pte) with a special migration pte entry * and check again if it has been pinned. Pinned pages are restored because we * cannot migrate them. * * This is the last step before we call the device driver callback to allocate * destination memory and copy contents of original page over to new page. */ static void migrate_vma_unmap(struct migrate_vma *migrate) { int flags = TTU_MIGRATION | TTU_IGNORE_MLOCK | TTU_IGNORE_ACCESS; const unsigned long npages = migrate->npages; const unsigned long start = migrate->start; unsigned long addr, i, restore = 0; for (i = 0; i < npages; i++) { struct page *page = migrate_pfn_to_page(migrate->src[i]); if (!page || !(migrate->src[i] & MIGRATE_PFN_MIGRATE)) continue; |
8c3328f1f
|
2470 2471 2472 2473 |
if (page_mapped(page)) { try_to_unmap(page, flags); if (page_mapped(page)) goto restore; |
8763cb45a
|
2474 |
} |
8c3328f1f
|
2475 2476 2477 2478 2479 2480 2481 2482 |
if (migrate_vma_check_page(page)) continue; restore: migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; migrate->cpages--; restore++; |
8763cb45a
|
2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 |
} for (addr = start, i = 0; i < npages && restore; addr += PAGE_SIZE, i++) { struct page *page = migrate_pfn_to_page(migrate->src[i]); if (!page || (migrate->src[i] & MIGRATE_PFN_MIGRATE)) continue; remove_migration_ptes(page, page, false); migrate->src[i] = 0; unlock_page(page); restore--; |
a5430dda8
|
2496 2497 2498 2499 |
if (is_zone_device_page(page)) put_page(page); else putback_lru_page(page); |
8763cb45a
|
2500 2501 |
} } |
a7d1f22bb
|
2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 |
/** * migrate_vma_setup() - prepare to migrate a range of memory * @args: contains the vma, start, and and pfns arrays for the migration * * Returns: negative errno on failures, 0 when 0 or more pages were migrated * without an error. * * Prepare to migrate a range of memory virtual address range by collecting all * the pages backing each virtual address in the range, saving them inside the * src array. Then lock those pages and unmap them. Once the pages are locked * and unmapped, check whether each page is pinned or not. Pages that aren't * pinned have the MIGRATE_PFN_MIGRATE flag set (by this function) in the * corresponding src array entry. Then restores any pages that are pinned, by * remapping and unlocking those pages. * * The caller should then allocate destination memory and copy source memory to * it for all those entries (ie with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE * flag set). Once these are allocated and copied, the caller must update each * corresponding entry in the dst array with the pfn value of the destination * page and with the MIGRATE_PFN_VALID and MIGRATE_PFN_LOCKED flags set * (destination pages must have their struct pages locked, via lock_page()). * * Note that the caller does not have to migrate all the pages that are marked * with MIGRATE_PFN_MIGRATE flag in src array unless this is a migration from * device memory to system memory. If the caller cannot migrate a device page * back to system memory, then it must return VM_FAULT_SIGBUS, which has severe * consequences for the userspace process, so it must be avoided if at all * possible. * * For empty entries inside CPU page table (pte_none() or pmd_none() is true) we * do set MIGRATE_PFN_MIGRATE flag inside the corresponding source array thus * allowing the caller to allocate device memory for those unback virtual * address. For this the caller simply has to allocate device memory and * properly set the destination entry like for regular migration. Note that * this can still fails and thus inside the device driver must check if the * migration was successful for those entries after calling migrate_vma_pages() * just like for regular migration. * * After that, the callers must call migrate_vma_pages() to go over each entry * in the src array that has the MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag * set. If the corresponding entry in dst array has MIGRATE_PFN_VALID flag set, * then migrate_vma_pages() to migrate struct page information from the source * struct page to the destination struct page. If it fails to migrate the * struct page information, then it clears the MIGRATE_PFN_MIGRATE flag in the * src array. * * At this point all successfully migrated pages have an entry in the src * array with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag set and the dst * array entry with MIGRATE_PFN_VALID flag set. * * Once migrate_vma_pages() returns the caller may inspect which pages were * successfully migrated, and which were not. Successfully migrated pages will * have the MIGRATE_PFN_MIGRATE flag set for their src array entry. * * It is safe to update device page table after migrate_vma_pages() because * both destination and source page are still locked, and the mmap_sem is held * in read mode (hence no one can unmap the range being migrated). * * Once the caller is done cleaning up things and updating its page table (if it * chose to do so, this is not an obligation) it finally calls * migrate_vma_finalize() to update the CPU page table to point to new pages * for successfully migrated pages or otherwise restore the CPU page table to * point to the original source pages. */ int migrate_vma_setup(struct migrate_vma *args) { long nr_pages = (args->end - args->start) >> PAGE_SHIFT; args->start &= PAGE_MASK; args->end &= PAGE_MASK; if (!args->vma || is_vm_hugetlb_page(args->vma) || (args->vma->vm_flags & VM_SPECIAL) || vma_is_dax(args->vma)) return -EINVAL; if (nr_pages <= 0) return -EINVAL; if (args->start < args->vma->vm_start || args->start >= args->vma->vm_end) return -EINVAL; if (args->end <= args->vma->vm_start || args->end > args->vma->vm_end) return -EINVAL; if (!args->src || !args->dst) return -EINVAL; memset(args->src, 0, sizeof(*args->src) * nr_pages); args->cpages = 0; args->npages = 0; migrate_vma_collect(args); if (args->cpages) migrate_vma_prepare(args); if (args->cpages) migrate_vma_unmap(args); /* * At this point pages are locked and unmapped, and thus they have * stable content and can safely be copied to destination memory that * is allocated by the drivers. */ return 0; } EXPORT_SYMBOL(migrate_vma_setup); |
8315ada7f
|
2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 |
static void migrate_vma_insert_page(struct migrate_vma *migrate, unsigned long addr, struct page *page, unsigned long *src, unsigned long *dst) { struct vm_area_struct *vma = migrate->vma; struct mm_struct *mm = vma->vm_mm; struct mem_cgroup *memcg; bool flush = false; spinlock_t *ptl; pte_t entry; pgd_t *pgdp; p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; /* Only allow populating anonymous memory */ if (!vma_is_anonymous(vma)) goto abort; pgdp = pgd_offset(mm, addr); p4dp = p4d_alloc(mm, pgdp, addr); if (!p4dp) goto abort; pudp = pud_alloc(mm, p4dp, addr); if (!pudp) goto abort; pmdp = pmd_alloc(mm, pudp, addr); if (!pmdp) goto abort; if (pmd_trans_huge(*pmdp) || pmd_devmap(*pmdp)) goto abort; /* * Use pte_alloc() instead of pte_alloc_map(). We can't run * pte_offset_map() on pmds where a huge pmd might be created * from a different thread. * * pte_alloc_map() is safe to use under down_write(mmap_sem) or when * parallel threads are excluded by other means. * * Here we only have down_read(mmap_sem). */ |
4cf589249
|
2651 |
if (pte_alloc(mm, pmdp)) |
8315ada7f
|
2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 |
goto abort; /* See the comment in pte_alloc_one_map() */ if (unlikely(pmd_trans_unstable(pmdp))) goto abort; if (unlikely(anon_vma_prepare(vma))) goto abort; if (mem_cgroup_try_charge(page, vma->vm_mm, GFP_KERNEL, &memcg, false)) goto abort; /* * The memory barrier inside __SetPageUptodate makes sure that * preceding stores to the page contents become visible before * the set_pte_at() write. */ __SetPageUptodate(page); |
df6ad6983
|
2669 2670 2671 2672 2673 2674 |
if (is_zone_device_page(page)) { if (is_device_private_page(page)) { swp_entry_t swp_entry; swp_entry = make_device_private_entry(page, vma->vm_flags & VM_WRITE); entry = swp_entry_to_pte(swp_entry); |
df6ad6983
|
2675 |
} |
8315ada7f
|
2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 |
} else { entry = mk_pte(page, vma->vm_page_prot); if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); } ptep = pte_offset_map_lock(mm, pmdp, addr, &ptl); if (pte_present(*ptep)) { unsigned long pfn = pte_pfn(*ptep); if (!is_zero_pfn(pfn)) { pte_unmap_unlock(ptep, ptl); mem_cgroup_cancel_charge(page, memcg, false); goto abort; } flush = true; } else if (!pte_none(*ptep)) { pte_unmap_unlock(ptep, ptl); mem_cgroup_cancel_charge(page, memcg, false); goto abort; } /* * Check for usefaultfd but do not deliver the fault. Instead, * just back off. */ if (userfaultfd_missing(vma)) { pte_unmap_unlock(ptep, ptl); mem_cgroup_cancel_charge(page, memcg, false); goto abort; } inc_mm_counter(mm, MM_ANONPAGES); page_add_new_anon_rmap(page, vma, addr, false); mem_cgroup_commit_charge(page, memcg, false, false); if (!is_zone_device_page(page)) lru_cache_add_active_or_unevictable(page, vma); get_page(page); if (flush) { flush_cache_page(vma, addr, pte_pfn(*ptep)); ptep_clear_flush_notify(vma, addr, ptep); set_pte_at_notify(mm, addr, ptep, entry); update_mmu_cache(vma, addr, ptep); } else { /* No need to invalidate - it was non-present before */ set_pte_at(mm, addr, ptep, entry); update_mmu_cache(vma, addr, ptep); } pte_unmap_unlock(ptep, ptl); *src = MIGRATE_PFN_MIGRATE; return; abort: *src &= ~MIGRATE_PFN_MIGRATE; } |
a7d1f22bb
|
2734 |
/** |
8763cb45a
|
2735 2736 2737 2738 2739 2740 2741 |
* migrate_vma_pages() - migrate meta-data from src page to dst page * @migrate: migrate struct containing all migration information * * This migrates struct page meta-data from source struct page to destination * struct page. This effectively finishes the migration from source page to the * destination page. */ |
a7d1f22bb
|
2742 |
void migrate_vma_pages(struct migrate_vma *migrate) |
8763cb45a
|
2743 2744 2745 |
{ const unsigned long npages = migrate->npages; const unsigned long start = migrate->start; |
ac46d4f3c
|
2746 2747 |
struct mmu_notifier_range range; unsigned long addr, i; |
8315ada7f
|
2748 |
bool notified = false; |
8763cb45a
|
2749 2750 2751 2752 2753 2754 |
for (i = 0, addr = start; i < npages; addr += PAGE_SIZE, i++) { struct page *newpage = migrate_pfn_to_page(migrate->dst[i]); struct page *page = migrate_pfn_to_page(migrate->src[i]); struct address_space *mapping; int r; |
8315ada7f
|
2755 2756 |
if (!newpage) { migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; |
8763cb45a
|
2757 |
continue; |
8315ada7f
|
2758 2759 2760 2761 2762 2763 2764 |
} if (!page) { if (!(migrate->src[i] & MIGRATE_PFN_MIGRATE)) { continue; } if (!notified) { |
8315ada7f
|
2765 |
notified = true; |
ac46d4f3c
|
2766 2767 |
mmu_notifier_range_init(&range, |
7269f9999
|
2768 |
MMU_NOTIFY_CLEAR, 0, |
6f4f13e8d
|
2769 |
NULL, |
ac46d4f3c
|
2770 2771 2772 |
migrate->vma->vm_mm, addr, migrate->end); mmu_notifier_invalidate_range_start(&range); |
8315ada7f
|
2773 2774 2775 2776 |
} migrate_vma_insert_page(migrate, addr, newpage, &migrate->src[i], &migrate->dst[i]); |
8763cb45a
|
2777 |
continue; |
8315ada7f
|
2778 |
} |
8763cb45a
|
2779 2780 |
mapping = page_mapping(page); |
a5430dda8
|
2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 |
if (is_zone_device_page(newpage)) { if (is_device_private_page(newpage)) { /* * For now only support private anonymous when * migrating to un-addressable device memory. */ if (mapping) { migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; continue; } |
25b2995a3
|
2791 |
} else { |
a5430dda8
|
2792 2793 2794 2795 2796 2797 2798 2799 |
/* * Other types of ZONE_DEVICE page are not * supported. */ migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; continue; } } |
8763cb45a
|
2800 2801 2802 2803 |
r = migrate_page(mapping, newpage, page, MIGRATE_SYNC_NO_COPY); if (r != MIGRATEPAGE_SUCCESS) migrate->src[i] &= ~MIGRATE_PFN_MIGRATE; } |
8315ada7f
|
2804 |
|
4645b9fe8
|
2805 2806 2807 2808 2809 |
/* * No need to double call mmu_notifier->invalidate_range() callback as * the above ptep_clear_flush_notify() inside migrate_vma_insert_page() * did already call it. */ |
8315ada7f
|
2810 |
if (notified) |
ac46d4f3c
|
2811 |
mmu_notifier_invalidate_range_only_end(&range); |
8763cb45a
|
2812 |
} |
a7d1f22bb
|
2813 |
EXPORT_SYMBOL(migrate_vma_pages); |
8763cb45a
|
2814 |
|
a7d1f22bb
|
2815 |
/** |
8763cb45a
|
2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 |
* migrate_vma_finalize() - restore CPU page table entry * @migrate: migrate struct containing all migration information * * This replaces the special migration pte entry with either a mapping to the * new page if migration was successful for that page, or to the original page * otherwise. * * This also unlocks the pages and puts them back on the lru, or drops the extra * refcount, for device pages. */ |
a7d1f22bb
|
2826 |
void migrate_vma_finalize(struct migrate_vma *migrate) |
8763cb45a
|
2827 2828 2829 2830 2831 2832 2833 |
{ const unsigned long npages = migrate->npages; unsigned long i; for (i = 0; i < npages; i++) { struct page *newpage = migrate_pfn_to_page(migrate->dst[i]); struct page *page = migrate_pfn_to_page(migrate->src[i]); |
8315ada7f
|
2834 2835 2836 2837 2838 |
if (!page) { if (newpage) { unlock_page(newpage); put_page(newpage); } |
8763cb45a
|
2839 |
continue; |
8315ada7f
|
2840 |
} |
8763cb45a
|
2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 |
if (!(migrate->src[i] & MIGRATE_PFN_MIGRATE) || !newpage) { if (newpage) { unlock_page(newpage); put_page(newpage); } newpage = page; } remove_migration_ptes(page, newpage, false); unlock_page(page); migrate->cpages--; |
a5430dda8
|
2852 2853 2854 2855 |
if (is_zone_device_page(page)) put_page(page); else putback_lru_page(page); |
8763cb45a
|
2856 2857 2858 |
if (newpage != page) { unlock_page(newpage); |
a5430dda8
|
2859 2860 2861 2862 |
if (is_zone_device_page(newpage)) put_page(newpage); else putback_lru_page(newpage); |
8763cb45a
|
2863 2864 2865 |
} } } |
a7d1f22bb
|
2866 |
EXPORT_SYMBOL(migrate_vma_finalize); |
9b2ed9cb9
|
2867 |
#endif /* CONFIG_DEVICE_PRIVATE */ |