Blame view
mm/memory.c
107 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 |
/* * linux/mm/memory.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds */ /* * demand-loading started 01.12.91 - seems it is high on the list of * things wanted, and it should be easy to implement. - Linus */ /* * Ok, demand-loading was easy, shared pages a little bit tricker. Shared * pages started 02.12.91, seems to work. - Linus. * * Tested sharing by executing about 30 /bin/sh: under the old kernel it * would have taken more than the 6M I have free, but it worked well as * far as I could see. * * Also corrected some "invalidate()"s - I wasn't doing enough of them. */ /* * Real VM (paging to/from disk) started 18.12.91. Much more work and * thought has to go into this. Oh, well.. * 19.12.91 - works, somewhat. Sometimes I get faults, don't know why. * Found it. Everything seems to work now. * 20.12.91 - Ok, making the swap-device changeable like the root. */ /* * 05.04.94 - Multi-page memory management added for v1.1. * Idea by Alex Bligh (alex@cconcepts.co.uk) * * 16.07.99 - Support of BIGMEM added by Gerhard Wichert, Siemens AG * (Gerhard.Wichert@pdb.siemens.de) * * Aug/Sep 2004 Changed to four level page tables (Andi Kleen) */ #include <linux/kernel_stat.h> #include <linux/mm.h> #include <linux/hugetlb.h> #include <linux/mman.h> #include <linux/swap.h> #include <linux/highmem.h> #include <linux/pagemap.h> |
9a8408951
|
48 |
#include <linux/ksm.h> |
1da177e4c
|
49 |
#include <linux/rmap.h> |
b95f1b31b
|
50 |
#include <linux/export.h> |
0ff922452
|
51 |
#include <linux/delayacct.h> |
1da177e4c
|
52 |
#include <linux/init.h> |
01c8f1c44
|
53 |
#include <linux/pfn_t.h> |
edc79b2a4
|
54 |
#include <linux/writeback.h> |
8a9f3ccd2
|
55 |
#include <linux/memcontrol.h> |
cddb8a5c1
|
56 |
#include <linux/mmu_notifier.h> |
3dc147414
|
57 58 59 |
#include <linux/kallsyms.h> #include <linux/swapops.h> #include <linux/elf.h> |
5a0e3ad6a
|
60 |
#include <linux/gfp.h> |
4daae3b4b
|
61 |
#include <linux/migrate.h> |
2fbc57c53
|
62 |
#include <linux/string.h> |
0abdd7a81
|
63 |
#include <linux/dma-debug.h> |
1592eef01
|
64 |
#include <linux/debugfs.h> |
6b251fc96
|
65 |
#include <linux/userfaultfd_k.h> |
1da177e4c
|
66 |
|
6952b61de
|
67 |
#include <asm/io.h> |
33a709b25
|
68 |
#include <asm/mmu_context.h> |
1da177e4c
|
69 70 71 72 73 |
#include <asm/pgalloc.h> #include <asm/uaccess.h> #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/pgtable.h> |
42b777281
|
74 |
#include "internal.h" |
90572890d
|
75 76 |
#ifdef LAST_CPUPID_NOT_IN_PAGE_FLAGS #warning Unfortunate NUMA and NUMA Balancing config, growing page-frame for last_cpupid. |
75980e97d
|
77 |
#endif |
d41dee369
|
78 |
#ifndef CONFIG_NEED_MULTIPLE_NODES |
1da177e4c
|
79 80 81 82 83 84 85 |
/* use the per-pgdat data instead for discontigmem - mbligh */ unsigned long max_mapnr; struct page *mem_map; EXPORT_SYMBOL(max_mapnr); EXPORT_SYMBOL(mem_map); #endif |
1da177e4c
|
86 87 88 89 90 91 92 93 |
/* * A number of key systems in x86 including ioremap() rely on the assumption * that high_memory defines the upper bound on direct map memory, then end * of ZONE_NORMAL. Under CONFIG_DISCONTIG this means that max_low_pfn and * highstart_pfn must be the same; there must be no gap between ZONE_NORMAL * and ZONE_HIGHMEM. */ void * high_memory; |
1da177e4c
|
94 |
|
1da177e4c
|
95 |
EXPORT_SYMBOL(high_memory); |
1da177e4c
|
96 |
|
32a932332
|
97 98 99 100 101 102 103 104 105 106 107 108 |
/* * Randomize the address space (stacks, mmaps, brk, etc.). * * ( When CONFIG_COMPAT_BRK=y we exclude brk from randomization, * as ancient (libc5 based) binaries can segfault. ) */ int randomize_va_space __read_mostly = #ifdef CONFIG_COMPAT_BRK 1; #else 2; #endif |
a62eaf151
|
109 110 111 112 |
static int __init disable_randmaps(char *s) { randomize_va_space = 0; |
9b41046cd
|
113 |
return 1; |
a62eaf151
|
114 115 |
} __setup("norandmaps", disable_randmaps); |
62eede62d
|
116 |
unsigned long zero_pfn __read_mostly; |
03f6462a3
|
117 |
unsigned long highest_memmap_pfn __read_mostly; |
a13ea5b75
|
118 |
|
0b70068e4
|
119 |
EXPORT_SYMBOL(zero_pfn); |
a13ea5b75
|
120 121 122 123 124 125 126 127 128 |
/* * CONFIG_MMU architectures set up ZERO_PAGE in their paging_init() */ static int __init init_zero_pfn(void) { zero_pfn = page_to_pfn(ZERO_PAGE(0)); return 0; } core_initcall(init_zero_pfn); |
a62eaf151
|
129 |
|
d559db086
|
130 |
|
34e55232e
|
131 |
#if defined(SPLIT_RSS_COUNTING) |
ea48cf786
|
132 |
void sync_mm_rss(struct mm_struct *mm) |
34e55232e
|
133 134 135 136 |
{ int i; for (i = 0; i < NR_MM_COUNTERS; i++) { |
05af2e104
|
137 138 139 |
if (current->rss_stat.count[i]) { add_mm_counter(mm, i, current->rss_stat.count[i]); current->rss_stat.count[i] = 0; |
34e55232e
|
140 141 |
} } |
05af2e104
|
142 |
current->rss_stat.events = 0; |
34e55232e
|
143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 |
} static void add_mm_counter_fast(struct mm_struct *mm, int member, int val) { struct task_struct *task = current; if (likely(task->mm == mm)) task->rss_stat.count[member] += val; else add_mm_counter(mm, member, val); } #define inc_mm_counter_fast(mm, member) add_mm_counter_fast(mm, member, 1) #define dec_mm_counter_fast(mm, member) add_mm_counter_fast(mm, member, -1) /* sync counter once per 64 page faults */ #define TASK_RSS_EVENTS_THRESH (64) static void check_sync_rss_stat(struct task_struct *task) { if (unlikely(task != current)) return; if (unlikely(task->rss_stat.events++ > TASK_RSS_EVENTS_THRESH)) |
ea48cf786
|
164 |
sync_mm_rss(task->mm); |
34e55232e
|
165 |
} |
9547d01bf
|
166 |
#else /* SPLIT_RSS_COUNTING */ |
34e55232e
|
167 168 169 170 171 172 173 |
#define inc_mm_counter_fast(mm, member) inc_mm_counter(mm, member) #define dec_mm_counter_fast(mm, member) dec_mm_counter(mm, member) static void check_sync_rss_stat(struct task_struct *task) { } |
9547d01bf
|
174 175 176 |
#endif /* SPLIT_RSS_COUNTING */ #ifdef HAVE_GENERIC_MMU_GATHER |
ca1d6c7d9
|
177 |
static bool tlb_next_batch(struct mmu_gather *tlb) |
9547d01bf
|
178 179 180 181 182 183 |
{ struct mmu_gather_batch *batch; batch = tlb->active; if (batch->next) { tlb->active = batch->next; |
ca1d6c7d9
|
184 |
return true; |
9547d01bf
|
185 |
} |
53a59fc67
|
186 |
if (tlb->batch_count == MAX_GATHER_BATCH_COUNT) |
ca1d6c7d9
|
187 |
return false; |
53a59fc67
|
188 |
|
9547d01bf
|
189 190 |
batch = (void *)__get_free_pages(GFP_NOWAIT | __GFP_NOWARN, 0); if (!batch) |
ca1d6c7d9
|
191 |
return false; |
9547d01bf
|
192 |
|
53a59fc67
|
193 |
tlb->batch_count++; |
9547d01bf
|
194 195 196 197 198 199 |
batch->next = NULL; batch->nr = 0; batch->max = MAX_GATHER_BATCH; tlb->active->next = batch; tlb->active = batch; |
ca1d6c7d9
|
200 |
return true; |
9547d01bf
|
201 202 203 204 205 206 207 |
} /* tlb_gather_mmu * Called to initialize an (on-stack) mmu_gather structure for page-table * tear-down from @mm. The @fullmm argument is used when @mm is without * users and we're going to destroy the full address space (exit/execve). */ |
2b047252d
|
208 |
void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, unsigned long start, unsigned long end) |
9547d01bf
|
209 210 |
{ tlb->mm = mm; |
2b047252d
|
211 212 |
/* Is it from 0 to ~0? */ tlb->fullmm = !(start | (end+1)); |
1de14c3c5
|
213 |
tlb->need_flush_all = 0; |
9547d01bf
|
214 215 216 217 |
tlb->local.next = NULL; tlb->local.nr = 0; tlb->local.max = ARRAY_SIZE(tlb->__pages); tlb->active = &tlb->local; |
53a59fc67
|
218 |
tlb->batch_count = 0; |
9547d01bf
|
219 220 221 222 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb->batch = NULL; #endif |
fb7332a9f
|
223 224 |
__tlb_reset_range(tlb); |
9547d01bf
|
225 |
} |
1cf35d477
|
226 |
static void tlb_flush_mmu_tlbonly(struct mmu_gather *tlb) |
9547d01bf
|
227 |
{ |
721c21c17
|
228 229 |
if (!tlb->end) return; |
9547d01bf
|
230 |
tlb_flush(tlb); |
34ee645e8
|
231 |
mmu_notifier_invalidate_range(tlb->mm, tlb->start, tlb->end); |
9547d01bf
|
232 233 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb_table_flush(tlb); |
34e55232e
|
234 |
#endif |
fb7332a9f
|
235 |
__tlb_reset_range(tlb); |
1cf35d477
|
236 237 238 239 240 |
} static void tlb_flush_mmu_free(struct mmu_gather *tlb) { struct mmu_gather_batch *batch; |
34e55232e
|
241 |
|
721c21c17
|
242 |
for (batch = &tlb->local; batch && batch->nr; batch = batch->next) { |
9547d01bf
|
243 244 245 246 247 |
free_pages_and_swap_cache(batch->pages, batch->nr); batch->nr = 0; } tlb->active = &tlb->local; } |
1cf35d477
|
248 249 |
void tlb_flush_mmu(struct mmu_gather *tlb) { |
1cf35d477
|
250 251 252 |
tlb_flush_mmu_tlbonly(tlb); tlb_flush_mmu_free(tlb); } |
9547d01bf
|
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 |
/* tlb_finish_mmu * Called at the end of the shootdown operation to free up any resources * that were required. */ void tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end) { struct mmu_gather_batch *batch, *next; tlb_flush_mmu(tlb); /* keep the page table cache within bounds */ check_pgt_cache(); for (batch = tlb->local.next; batch; batch = next) { next = batch->next; free_pages((unsigned long)batch, 0); } tlb->local.next = NULL; } /* __tlb_remove_page * Must perform the equivalent to __free_pte(pte_get_and_clear(ptep)), while * handling the additional races in SMP caused by other CPUs caching valid * mappings in their TLBs. Returns the number of free page slots left. * When out of page slots we must call tlb_flush_mmu(). */ int __tlb_remove_page(struct mmu_gather *tlb, struct page *page) { struct mmu_gather_batch *batch; |
fb7332a9f
|
282 |
VM_BUG_ON(!tlb->end); |
9547d01bf
|
283 |
|
9547d01bf
|
284 285 286 287 288 |
batch = tlb->active; batch->pages[batch->nr++] = page; if (batch->nr == batch->max) { if (!tlb_next_batch(tlb)) return 0; |
0b43c3aab
|
289 |
batch = tlb->active; |
9547d01bf
|
290 |
} |
309381fea
|
291 |
VM_BUG_ON_PAGE(batch->nr > batch->max, page); |
9547d01bf
|
292 293 294 295 296 |
return batch->max - batch->nr; } #endif /* HAVE_GENERIC_MMU_GATHER */ |
267239116
|
297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE /* * See the comment near struct mmu_table_batch. */ static void tlb_remove_table_smp_sync(void *arg) { /* Simply deliver the interrupt */ } static void tlb_remove_table_one(void *table) { /* * This isn't an RCU grace period and hence the page-tables cannot be * assumed to be actually RCU-freed. * * It is however sufficient for software page-table walkers that rely on * IRQ disabling. See the comment near struct mmu_table_batch. */ smp_call_function(tlb_remove_table_smp_sync, NULL, 1); __tlb_remove_table(table); } static void tlb_remove_table_rcu(struct rcu_head *head) { struct mmu_table_batch *batch; int i; batch = container_of(head, struct mmu_table_batch, rcu); for (i = 0; i < batch->nr; i++) __tlb_remove_table(batch->tables[i]); free_page((unsigned long)batch); } void tlb_table_flush(struct mmu_gather *tlb) { struct mmu_table_batch **batch = &tlb->batch; if (*batch) { call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu); *batch = NULL; } } void tlb_remove_table(struct mmu_gather *tlb, void *table) { struct mmu_table_batch **batch = &tlb->batch; |
267239116
|
347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 |
/* * When there's less then two users of this mm there cannot be a * concurrent page-table walk. */ if (atomic_read(&tlb->mm->mm_users) < 2) { __tlb_remove_table(table); return; } if (*batch == NULL) { *batch = (struct mmu_table_batch *)__get_free_page(GFP_NOWAIT | __GFP_NOWARN); if (*batch == NULL) { tlb_remove_table_one(table); return; } (*batch)->nr = 0; } (*batch)->tables[(*batch)->nr++] = table; if ((*batch)->nr == MAX_TABLE_BATCH) tlb_table_flush(tlb); } |
9547d01bf
|
368 |
#endif /* CONFIG_HAVE_RCU_TABLE_FREE */ |
267239116
|
369 |
|
1da177e4c
|
370 |
/* |
1da177e4c
|
371 372 373 |
* Note: this doesn't free the actual pages themselves. That * has been handled earlier when unmapping all the memory regions. */ |
9e1b32caa
|
374 375 |
static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd, unsigned long addr) |
1da177e4c
|
376 |
{ |
2f569afd9
|
377 |
pgtable_t token = pmd_pgtable(*pmd); |
e0da382c9
|
378 |
pmd_clear(pmd); |
9e1b32caa
|
379 |
pte_free_tlb(tlb, token, addr); |
e1f56c89b
|
380 |
atomic_long_dec(&tlb->mm->nr_ptes); |
1da177e4c
|
381 |
} |
e0da382c9
|
382 383 384 |
static inline void free_pmd_range(struct mmu_gather *tlb, pud_t *pud, unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c
|
385 386 387 |
{ pmd_t *pmd; unsigned long next; |
e0da382c9
|
388 |
unsigned long start; |
1da177e4c
|
389 |
|
e0da382c9
|
390 |
start = addr; |
1da177e4c
|
391 |
pmd = pmd_offset(pud, addr); |
1da177e4c
|
392 393 394 395 |
do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
9e1b32caa
|
396 |
free_pte_range(tlb, pmd, addr); |
1da177e4c
|
397 |
} while (pmd++, addr = next, addr != end); |
e0da382c9
|
398 399 400 401 402 403 404 |
start &= PUD_MASK; if (start < floor) return; if (ceiling) { ceiling &= PUD_MASK; if (!ceiling) return; |
1da177e4c
|
405 |
} |
e0da382c9
|
406 407 408 409 410 |
if (end - 1 > ceiling - 1) return; pmd = pmd_offset(pud, start); pud_clear(pud); |
9e1b32caa
|
411 |
pmd_free_tlb(tlb, pmd, start); |
dc6c9a35b
|
412 |
mm_dec_nr_pmds(tlb->mm); |
1da177e4c
|
413 |
} |
e0da382c9
|
414 415 416 |
static inline void free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c
|
417 418 419 |
{ pud_t *pud; unsigned long next; |
e0da382c9
|
420 |
unsigned long start; |
1da177e4c
|
421 |
|
e0da382c9
|
422 |
start = addr; |
1da177e4c
|
423 |
pud = pud_offset(pgd, addr); |
1da177e4c
|
424 425 426 427 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
e0da382c9
|
428 |
free_pmd_range(tlb, pud, addr, next, floor, ceiling); |
1da177e4c
|
429 |
} while (pud++, addr = next, addr != end); |
e0da382c9
|
430 431 432 433 434 435 436 |
start &= PGDIR_MASK; if (start < floor) return; if (ceiling) { ceiling &= PGDIR_MASK; if (!ceiling) return; |
1da177e4c
|
437 |
} |
e0da382c9
|
438 439 440 441 442 |
if (end - 1 > ceiling - 1) return; pud = pud_offset(pgd, start); pgd_clear(pgd); |
9e1b32caa
|
443 |
pud_free_tlb(tlb, pud, start); |
1da177e4c
|
444 445 446 |
} /* |
e0da382c9
|
447 |
* This function frees user-level page tables of a process. |
1da177e4c
|
448 |
*/ |
42b777281
|
449 |
void free_pgd_range(struct mmu_gather *tlb, |
e0da382c9
|
450 451 |
unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c
|
452 453 454 |
{ pgd_t *pgd; unsigned long next; |
e0da382c9
|
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 |
/* * The next few lines have given us lots of grief... * * Why are we testing PMD* at this top level? Because often * there will be no work to do at all, and we'd prefer not to * go all the way down to the bottom just to discover that. * * Why all these "- 1"s? Because 0 represents both the bottom * of the address space and the top of it (using -1 for the * top wouldn't help much: the masks would do the wrong thing). * The rule is that addr 0 and floor 0 refer to the bottom of * the address space, but end 0 and ceiling 0 refer to the top * Comparisons need to use "end - 1" and "ceiling - 1" (though * that end 0 case should be mythical). * * Wherever addr is brought up or ceiling brought down, we must * be careful to reject "the opposite 0" before it confuses the * subsequent tests. But what about where end is brought down * by PMD_SIZE below? no, end can't go down to 0 there. * * Whereas we round start (addr) and ceiling down, by different * masks at different levels, in order to test whether a table * now has no other vmas using it, so can be freed, we don't * bother to round floor or end up - the tests don't need that. */ |
1da177e4c
|
481 |
|
e0da382c9
|
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 |
addr &= PMD_MASK; if (addr < floor) { addr += PMD_SIZE; if (!addr) return; } if (ceiling) { ceiling &= PMD_MASK; if (!ceiling) return; } if (end - 1 > ceiling - 1) end -= PMD_SIZE; if (addr > end - 1) return; |
42b777281
|
497 |
pgd = pgd_offset(tlb->mm, addr); |
1da177e4c
|
498 499 500 501 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
42b777281
|
502 |
free_pud_range(tlb, pgd, addr, next, floor, ceiling); |
1da177e4c
|
503 |
} while (pgd++, addr = next, addr != end); |
e0da382c9
|
504 |
} |
42b777281
|
505 |
void free_pgtables(struct mmu_gather *tlb, struct vm_area_struct *vma, |
3bf5ee956
|
506 |
unsigned long floor, unsigned long ceiling) |
e0da382c9
|
507 508 509 510 |
{ while (vma) { struct vm_area_struct *next = vma->vm_next; unsigned long addr = vma->vm_start; |
8f4f8c164
|
511 |
/* |
25d9e2d15
|
512 513 |
* Hide vma from rmap and truncate_pagecache before freeing * pgtables |
8f4f8c164
|
514 |
*/ |
5beb49305
|
515 |
unlink_anon_vmas(vma); |
8f4f8c164
|
516 |
unlink_file_vma(vma); |
9da61aef0
|
517 |
if (is_vm_hugetlb_page(vma)) { |
3bf5ee956
|
518 |
hugetlb_free_pgd_range(tlb, addr, vma->vm_end, |
e0da382c9
|
519 |
floor, next? next->vm_start: ceiling); |
3bf5ee956
|
520 521 522 523 524 |
} else { /* * Optimization: gather nearby vmas into one call down */ while (next && next->vm_start <= vma->vm_end + PMD_SIZE |
4866920b9
|
525 |
&& !is_vm_hugetlb_page(next)) { |
3bf5ee956
|
526 527 |
vma = next; next = vma->vm_next; |
5beb49305
|
528 |
unlink_anon_vmas(vma); |
8f4f8c164
|
529 |
unlink_file_vma(vma); |
3bf5ee956
|
530 531 532 533 |
} free_pgd_range(tlb, addr, vma->vm_end, floor, next? next->vm_start: ceiling); } |
e0da382c9
|
534 535 |
vma = next; } |
1da177e4c
|
536 |
} |
3ed3a4f0d
|
537 |
int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) |
1da177e4c
|
538 |
{ |
c4088ebdc
|
539 |
spinlock_t *ptl; |
2f569afd9
|
540 |
pgtable_t new = pte_alloc_one(mm, address); |
1bb3630e8
|
541 542 |
if (!new) return -ENOMEM; |
362a61ad6
|
543 544 545 546 547 548 549 550 551 552 553 554 555 556 |
/* * Ensure all pte setup (eg. pte page lock and page clearing) are * visible before the pte is made visible to other CPUs by being * put into page tables. * * The other side of the story is the pointer chasing in the page * table walking code (when walking the page table without locking; * ie. most of the time). Fortunately, these data accesses consist * of a chain of data-dependent loads, meaning most CPUs (alpha * being the notable exception) will already guarantee loads are * seen in-order. See the alpha page table accessors for the * smp_read_barrier_depends() barriers in page table walking code. */ smp_wmb(); /* Could be smp_wmb__xxx(before|after)_spin_lock */ |
c4088ebdc
|
557 |
ptl = pmd_lock(mm, pmd); |
8ac1f8320
|
558 |
if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
e1f56c89b
|
559 |
atomic_long_inc(&mm->nr_ptes); |
1da177e4c
|
560 |
pmd_populate(mm, pmd, new); |
2f569afd9
|
561 |
new = NULL; |
4b471e889
|
562 |
} |
c4088ebdc
|
563 |
spin_unlock(ptl); |
2f569afd9
|
564 565 |
if (new) pte_free(mm, new); |
1bb3630e8
|
566 |
return 0; |
1da177e4c
|
567 |
} |
1bb3630e8
|
568 |
int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) |
1da177e4c
|
569 |
{ |
1bb3630e8
|
570 571 572 |
pte_t *new = pte_alloc_one_kernel(&init_mm, address); if (!new) return -ENOMEM; |
362a61ad6
|
573 |
smp_wmb(); /* See comment in __pte_alloc */ |
1bb3630e8
|
574 |
spin_lock(&init_mm.page_table_lock); |
8ac1f8320
|
575 |
if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
1bb3630e8
|
576 |
pmd_populate_kernel(&init_mm, pmd, new); |
2f569afd9
|
577 |
new = NULL; |
4b471e889
|
578 |
} |
1bb3630e8
|
579 |
spin_unlock(&init_mm.page_table_lock); |
2f569afd9
|
580 581 |
if (new) pte_free_kernel(&init_mm, new); |
1bb3630e8
|
582 |
return 0; |
1da177e4c
|
583 |
} |
d559db086
|
584 585 586 587 588 589 |
static inline void init_rss_vec(int *rss) { memset(rss, 0, sizeof(int) * NR_MM_COUNTERS); } static inline void add_mm_rss_vec(struct mm_struct *mm, int *rss) |
ae8597623
|
590 |
{ |
d559db086
|
591 |
int i; |
34e55232e
|
592 |
if (current->mm == mm) |
05af2e104
|
593 |
sync_mm_rss(mm); |
d559db086
|
594 595 596 |
for (i = 0; i < NR_MM_COUNTERS; i++) if (rss[i]) add_mm_counter(mm, i, rss[i]); |
ae8597623
|
597 |
} |
1da177e4c
|
598 |
/* |
6aab341e0
|
599 600 601 |
* This function is called to print an error when a bad pte * is found. For example, we might have a PFN-mapped pte in * a region that doesn't allow it. |
b5810039a
|
602 603 604 |
* * The calling function must still handle the error. */ |
3dc147414
|
605 606 |
static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr, pte_t pte, struct page *page) |
b5810039a
|
607 |
{ |
3dc147414
|
608 609 610 611 612 |
pgd_t *pgd = pgd_offset(vma->vm_mm, addr); pud_t *pud = pud_offset(pgd, addr); pmd_t *pmd = pmd_offset(pud, addr); struct address_space *mapping; pgoff_t index; |
d936cf9b3
|
613 614 615 616 617 618 619 620 621 622 623 624 625 626 |
static unsigned long resume; static unsigned long nr_shown; static unsigned long nr_unshown; /* * Allow a burst of 60 reports, then keep quiet for that minute; * or allow a steady drip of one report per second. */ if (nr_shown == 60) { if (time_before(jiffies, resume)) { nr_unshown++; return; } if (nr_unshown) { |
1170532bb
|
627 628 629 |
pr_alert("BUG: Bad page map: %lu messages suppressed ", nr_unshown); |
d936cf9b3
|
630 631 632 633 634 635 |
nr_unshown = 0; } nr_shown = 0; } if (nr_shown++ == 0) resume = jiffies + 60 * HZ; |
3dc147414
|
636 637 638 |
mapping = vma->vm_file ? vma->vm_file->f_mapping : NULL; index = linear_page_index(vma, addr); |
1170532bb
|
639 640 641 642 |
pr_alert("BUG: Bad page map in process %s pte:%08llx pmd:%08llx ", current->comm, (long long)pte_val(pte), (long long)pmd_val(*pmd)); |
718a38211
|
643 |
if (page) |
f0b791a34
|
644 |
dump_page(page, "bad pte"); |
1170532bb
|
645 646 647 |
pr_alert("addr:%p vm_flags:%08lx anon_vma:%p mapping:%p index:%lx ", (void *)addr, vma->vm_flags, vma->anon_vma, mapping, index); |
3dc147414
|
648 649 650 |
/* * Choose text because data symbols depend on CONFIG_KALLSYMS_ALL=y */ |
2682582a6
|
651 652 653 654 655 656 |
pr_alert("file:%pD fault:%pf mmap:%pf readpage:%pf ", vma->vm_file, vma->vm_ops ? vma->vm_ops->fault : NULL, vma->vm_file ? vma->vm_file->f_op->mmap : NULL, mapping ? mapping->a_ops->readpage : NULL); |
b5810039a
|
657 |
dump_stack(); |
373d4d099
|
658 |
add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); |
b5810039a
|
659 660 661 |
} /* |
7e675137a
|
662 |
* vm_normal_page -- This function gets the "struct page" associated with a pte. |
6aab341e0
|
663 |
* |
7e675137a
|
664 665 666 |
* "Special" mappings do not wish to be associated with a "struct page" (either * it doesn't exist, or it exists but they don't want to touch it). In this * case, NULL is returned here. "Normal" mappings do have a struct page. |
b379d7901
|
667 |
* |
7e675137a
|
668 669 670 671 672 673 674 675 |
* There are 2 broad cases. Firstly, an architecture may define a pte_special() * pte bit, in which case this function is trivial. Secondly, an architecture * may not have a spare pte bit, which requires a more complicated scheme, * described below. * * A raw VM_PFNMAP mapping (ie. one that is not COWed) is always considered a * special mapping (even if there are underlying and valid "struct pages"). * COWed pages of a VM_PFNMAP are always normal. |
6aab341e0
|
676 |
* |
b379d7901
|
677 678 |
* The way we recognize COWed pages within VM_PFNMAP mappings is through the * rules set up by "remap_pfn_range()": the vma will have the VM_PFNMAP bit |
7e675137a
|
679 680 |
* set, and the vm_pgoff will point to the first PFN mapped: thus every special * mapping will always honor the rule |
6aab341e0
|
681 682 683 |
* * pfn_of_page == vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT) * |
7e675137a
|
684 685 686 687 688 689 |
* And for normal mappings this is false. * * This restricts such mappings to be a linear translation from virtual address * to pfn. To get around this restriction, we allow arbitrary mappings so long * as the vma is not a COW mapping; in that case, we know that all ptes are * special (because none can have been COWed). |
b379d7901
|
690 |
* |
b379d7901
|
691 |
* |
7e675137a
|
692 |
* In order to support COW of arbitrary special mappings, we have VM_MIXEDMAP. |
b379d7901
|
693 694 695 696 697 698 699 700 701 |
* * VM_MIXEDMAP mappings can likewise contain memory with or without "struct * page" backing, however the difference is that _all_ pages with a struct * page (that is, those where pfn_valid is true) are refcounted and considered * normal pages by the VM. The disadvantage is that pages are refcounted * (which can be slower and simply not an option for some PFNMAP users). The * advantage is that we don't have to follow the strict linearity rule of * PFNMAP mappings in order to support COWable mappings. * |
ee498ed73
|
702 |
*/ |
7e675137a
|
703 704 705 706 707 708 709 |
#ifdef __HAVE_ARCH_PTE_SPECIAL # define HAVE_PTE_SPECIAL 1 #else # define HAVE_PTE_SPECIAL 0 #endif struct page *vm_normal_page(struct vm_area_struct *vma, unsigned long addr, pte_t pte) |
ee498ed73
|
710 |
{ |
22b31eec6
|
711 |
unsigned long pfn = pte_pfn(pte); |
7e675137a
|
712 713 |
if (HAVE_PTE_SPECIAL) { |
b38af4721
|
714 |
if (likely(!pte_special(pte))) |
22b31eec6
|
715 |
goto check_pfn; |
667a0a06c
|
716 717 |
if (vma->vm_ops && vma->vm_ops->find_special_page) return vma->vm_ops->find_special_page(vma, addr); |
a13ea5b75
|
718 719 |
if (vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)) return NULL; |
62eede62d
|
720 |
if (!is_zero_pfn(pfn)) |
22b31eec6
|
721 |
print_bad_pte(vma, addr, pte, NULL); |
7e675137a
|
722 723 724 725 |
return NULL; } /* !HAVE_PTE_SPECIAL case follows: */ |
b379d7901
|
726 727 728 729 730 731 |
if (unlikely(vma->vm_flags & (VM_PFNMAP|VM_MIXEDMAP))) { if (vma->vm_flags & VM_MIXEDMAP) { if (!pfn_valid(pfn)) return NULL; goto out; } else { |
7e675137a
|
732 733 |
unsigned long off; off = (addr - vma->vm_start) >> PAGE_SHIFT; |
b379d7901
|
734 735 736 737 738 |
if (pfn == vma->vm_pgoff + off) return NULL; if (!is_cow_mapping(vma->vm_flags)) return NULL; } |
6aab341e0
|
739 |
} |
b38af4721
|
740 741 |
if (is_zero_pfn(pfn)) return NULL; |
22b31eec6
|
742 743 744 745 746 |
check_pfn: if (unlikely(pfn > highest_memmap_pfn)) { print_bad_pte(vma, addr, pte, NULL); return NULL; } |
6aab341e0
|
747 748 |
/* |
7e675137a
|
749 |
* NOTE! We still have PageReserved() pages in the page tables. |
7e675137a
|
750 |
* eg. VDSO mappings can cause them to exist. |
6aab341e0
|
751 |
*/ |
b379d7901
|
752 |
out: |
6aab341e0
|
753 |
return pfn_to_page(pfn); |
ee498ed73
|
754 755 756 |
} /* |
1da177e4c
|
757 758 759 |
* copy one vm_area from one task to the other. Assumes the page tables * already present in the new task to be cleared in the whole range * covered by this vma. |
1da177e4c
|
760 |
*/ |
570a335b8
|
761 |
static inline unsigned long |
1da177e4c
|
762 |
copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
b5810039a
|
763 |
pte_t *dst_pte, pte_t *src_pte, struct vm_area_struct *vma, |
8c1037627
|
764 |
unsigned long addr, int *rss) |
1da177e4c
|
765 |
{ |
b5810039a
|
766 |
unsigned long vm_flags = vma->vm_flags; |
1da177e4c
|
767 768 |
pte_t pte = *src_pte; struct page *page; |
1da177e4c
|
769 770 771 |
/* pte contains position in swap or file, so copy. */ if (unlikely(!pte_present(pte))) { |
0661a3361
|
772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 |
swp_entry_t entry = pte_to_swp_entry(pte); if (likely(!non_swap_entry(entry))) { if (swap_duplicate(entry) < 0) return entry.val; /* make sure dst_mm is on swapoff's mmlist. */ if (unlikely(list_empty(&dst_mm->mmlist))) { spin_lock(&mmlist_lock); if (list_empty(&dst_mm->mmlist)) list_add(&dst_mm->mmlist, &src_mm->mmlist); spin_unlock(&mmlist_lock); } rss[MM_SWAPENTS]++; } else if (is_migration_entry(entry)) { page = migration_entry_to_page(entry); |
eca56ff90
|
789 |
rss[mm_counter(page)]++; |
0661a3361
|
790 791 792 793 794 795 796 797 798 799 800 801 |
if (is_write_migration_entry(entry) && is_cow_mapping(vm_flags)) { /* * COW mappings require pages in both * parent and child to be set to read. */ make_migration_entry_read(&entry); pte = swp_entry_to_pte(entry); if (pte_swp_soft_dirty(*src_pte)) pte = pte_swp_mksoft_dirty(pte); set_pte_at(src_mm, addr, src_pte, pte); |
0697212a4
|
802 |
} |
1da177e4c
|
803 |
} |
ae8597623
|
804 |
goto out_set_pte; |
1da177e4c
|
805 |
} |
1da177e4c
|
806 807 808 809 |
/* * If it's a COW mapping, write protect it both * in the parent and the child */ |
67121172f
|
810 |
if (is_cow_mapping(vm_flags)) { |
1da177e4c
|
811 |
ptep_set_wrprotect(src_mm, addr, src_pte); |
3dc907951
|
812 |
pte = pte_wrprotect(pte); |
1da177e4c
|
813 814 815 816 817 818 819 820 821 |
} /* * If it's a shared mapping, mark it clean in * the child */ if (vm_flags & VM_SHARED) pte = pte_mkclean(pte); pte = pte_mkold(pte); |
6aab341e0
|
822 823 824 825 |
page = vm_normal_page(vma, addr, pte); if (page) { get_page(page); |
53f9263ba
|
826 |
page_dup_rmap(page, false); |
eca56ff90
|
827 |
rss[mm_counter(page)]++; |
6aab341e0
|
828 |
} |
ae8597623
|
829 830 831 |
out_set_pte: set_pte_at(dst_mm, addr, dst_pte, pte); |
570a335b8
|
832 |
return 0; |
1da177e4c
|
833 |
} |
21bda264f
|
834 |
static int copy_pte_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
71e3aac07
|
835 836 |
pmd_t *dst_pmd, pmd_t *src_pmd, struct vm_area_struct *vma, unsigned long addr, unsigned long end) |
1da177e4c
|
837 |
{ |
c36987e2e
|
838 |
pte_t *orig_src_pte, *orig_dst_pte; |
1da177e4c
|
839 |
pte_t *src_pte, *dst_pte; |
c74df32c7
|
840 |
spinlock_t *src_ptl, *dst_ptl; |
e040f218b
|
841 |
int progress = 0; |
d559db086
|
842 |
int rss[NR_MM_COUNTERS]; |
570a335b8
|
843 |
swp_entry_t entry = (swp_entry_t){0}; |
1da177e4c
|
844 845 |
again: |
d559db086
|
846 |
init_rss_vec(rss); |
c74df32c7
|
847 |
dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl); |
1da177e4c
|
848 849 |
if (!dst_pte) return -ENOMEM; |
ece0e2b64
|
850 |
src_pte = pte_offset_map(src_pmd, addr); |
4c21e2f24
|
851 |
src_ptl = pte_lockptr(src_mm, src_pmd); |
f20dc5f7c
|
852 |
spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING); |
c36987e2e
|
853 854 |
orig_src_pte = src_pte; orig_dst_pte = dst_pte; |
6606c3e0d
|
855 |
arch_enter_lazy_mmu_mode(); |
1da177e4c
|
856 |
|
1da177e4c
|
857 858 859 860 861 |
do { /* * We are holding two locks at this point - either of them * could generate latencies in another task on another CPU. */ |
e040f218b
|
862 863 864 |
if (progress >= 32) { progress = 0; if (need_resched() || |
95c354fe9
|
865 |
spin_needbreak(src_ptl) || spin_needbreak(dst_ptl)) |
e040f218b
|
866 867 |
break; } |
1da177e4c
|
868 869 870 871 |
if (pte_none(*src_pte)) { progress++; continue; } |
570a335b8
|
872 873 874 875 |
entry.val = copy_one_pte(dst_mm, src_mm, dst_pte, src_pte, vma, addr, rss); if (entry.val) break; |
1da177e4c
|
876 877 |
progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); |
1da177e4c
|
878 |
|
6606c3e0d
|
879 |
arch_leave_lazy_mmu_mode(); |
c74df32c7
|
880 |
spin_unlock(src_ptl); |
ece0e2b64
|
881 |
pte_unmap(orig_src_pte); |
d559db086
|
882 |
add_mm_rss_vec(dst_mm, rss); |
c36987e2e
|
883 |
pte_unmap_unlock(orig_dst_pte, dst_ptl); |
c74df32c7
|
884 |
cond_resched(); |
570a335b8
|
885 886 887 888 889 890 |
if (entry.val) { if (add_swap_count_continuation(entry, GFP_KERNEL) < 0) return -ENOMEM; progress = 0; } |
1da177e4c
|
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 |
if (addr != end) goto again; return 0; } static inline int copy_pmd_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, pud_t *dst_pud, pud_t *src_pud, struct vm_area_struct *vma, unsigned long addr, unsigned long end) { pmd_t *src_pmd, *dst_pmd; unsigned long next; dst_pmd = pmd_alloc(dst_mm, dst_pud, addr); if (!dst_pmd) return -ENOMEM; src_pmd = pmd_offset(src_pud, addr); do { next = pmd_addr_end(addr, end); |
5c7fb56e5
|
909 |
if (pmd_trans_huge(*src_pmd) || pmd_devmap(*src_pmd)) { |
71e3aac07
|
910 |
int err; |
14d1a55cd
|
911 |
VM_BUG_ON(next-addr != HPAGE_PMD_SIZE); |
71e3aac07
|
912 913 914 915 916 917 918 919 |
err = copy_huge_pmd(dst_mm, src_mm, dst_pmd, src_pmd, addr, vma); if (err == -ENOMEM) return -ENOMEM; if (!err) continue; /* fall through */ } |
1da177e4c
|
920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 |
if (pmd_none_or_clear_bad(src_pmd)) continue; if (copy_pte_range(dst_mm, src_mm, dst_pmd, src_pmd, vma, addr, next)) return -ENOMEM; } while (dst_pmd++, src_pmd++, addr = next, addr != end); return 0; } static inline int copy_pud_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, pgd_t *dst_pgd, pgd_t *src_pgd, struct vm_area_struct *vma, unsigned long addr, unsigned long end) { pud_t *src_pud, *dst_pud; unsigned long next; dst_pud = pud_alloc(dst_mm, dst_pgd, addr); if (!dst_pud) return -ENOMEM; src_pud = pud_offset(src_pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(src_pud)) continue; if (copy_pmd_range(dst_mm, src_mm, dst_pud, src_pud, vma, addr, next)) return -ENOMEM; } while (dst_pud++, src_pud++, addr = next, addr != end); return 0; } int copy_page_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, struct vm_area_struct *vma) { pgd_t *src_pgd, *dst_pgd; unsigned long next; unsigned long addr = vma->vm_start; unsigned long end = vma->vm_end; |
2ec74c3ef
|
958 959 960 |
unsigned long mmun_start; /* For mmu_notifiers */ unsigned long mmun_end; /* For mmu_notifiers */ bool is_cow; |
cddb8a5c1
|
961 |
int ret; |
1da177e4c
|
962 |
|
d992895ba
|
963 964 965 966 967 968 |
/* * Don't copy ptes where a page fault will fill them correctly. * Fork becomes much lighter when there are big shared or private * readonly mappings. The tradeoff is that copy_page_range is more * efficient than faulting. */ |
0661a3361
|
969 970 971 |
if (!(vma->vm_flags & (VM_HUGETLB | VM_PFNMAP | VM_MIXEDMAP)) && !vma->anon_vma) return 0; |
d992895ba
|
972 |
|
1da177e4c
|
973 974 |
if (is_vm_hugetlb_page(vma)) return copy_hugetlb_page_range(dst_mm, src_mm, vma); |
b3b9c2932
|
975 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) { |
2ab640379
|
976 977 978 979 |
/* * We do not free on error cases below as remove_vma * gets called on error from higher level routine */ |
5180da410
|
980 |
ret = track_pfn_copy(vma); |
2ab640379
|
981 982 983 |
if (ret) return ret; } |
cddb8a5c1
|
984 985 986 987 988 989 |
/* * We need to invalidate the secondary MMU mappings only when * there could be a permission downgrade on the ptes of the * parent mm. And a permission downgrade will only happen if * is_cow_mapping() returns true. */ |
2ec74c3ef
|
990 991 992 993 994 995 |
is_cow = is_cow_mapping(vma->vm_flags); mmun_start = addr; mmun_end = end; if (is_cow) mmu_notifier_invalidate_range_start(src_mm, mmun_start, mmun_end); |
cddb8a5c1
|
996 997 |
ret = 0; |
1da177e4c
|
998 999 1000 1001 1002 1003 |
dst_pgd = pgd_offset(dst_mm, addr); src_pgd = pgd_offset(src_mm, addr); do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(src_pgd)) continue; |
cddb8a5c1
|
1004 1005 1006 1007 1008 |
if (unlikely(copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd, vma, addr, next))) { ret = -ENOMEM; break; } |
1da177e4c
|
1009 |
} while (dst_pgd++, src_pgd++, addr = next, addr != end); |
cddb8a5c1
|
1010 |
|
2ec74c3ef
|
1011 1012 |
if (is_cow) mmu_notifier_invalidate_range_end(src_mm, mmun_start, mmun_end); |
cddb8a5c1
|
1013 |
return ret; |
1da177e4c
|
1014 |
} |
51c6f666f
|
1015 |
static unsigned long zap_pte_range(struct mmu_gather *tlb, |
b5810039a
|
1016 |
struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c
|
1017 |
unsigned long addr, unsigned long end, |
97a894136
|
1018 |
struct zap_details *details) |
1da177e4c
|
1019 |
{ |
b5810039a
|
1020 |
struct mm_struct *mm = tlb->mm; |
d16dfc550
|
1021 |
int force_flush = 0; |
d559db086
|
1022 |
int rss[NR_MM_COUNTERS]; |
97a894136
|
1023 |
spinlock_t *ptl; |
5f1a19070
|
1024 |
pte_t *start_pte; |
97a894136
|
1025 |
pte_t *pte; |
8a5f14a23
|
1026 |
swp_entry_t entry; |
d559db086
|
1027 |
|
d16dfc550
|
1028 |
again: |
e303297e6
|
1029 |
init_rss_vec(rss); |
5f1a19070
|
1030 1031 |
start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl); pte = start_pte; |
6606c3e0d
|
1032 |
arch_enter_lazy_mmu_mode(); |
1da177e4c
|
1033 1034 |
do { pte_t ptent = *pte; |
51c6f666f
|
1035 |
if (pte_none(ptent)) { |
1da177e4c
|
1036 |
continue; |
51c6f666f
|
1037 |
} |
6f5e6b9e6
|
1038 |
|
1da177e4c
|
1039 |
if (pte_present(ptent)) { |
ee498ed73
|
1040 |
struct page *page; |
51c6f666f
|
1041 |
|
6aab341e0
|
1042 |
page = vm_normal_page(vma, addr, ptent); |
1da177e4c
|
1043 1044 1045 1046 1047 1048 1049 1050 1051 |
if (unlikely(details) && page) { /* * unmap_shared_mapping_pages() wants to * invalidate cache without truncating: * unmap shared but keep private pages. */ if (details->check_mapping && details->check_mapping != page->mapping) continue; |
1da177e4c
|
1052 |
} |
b5810039a
|
1053 |
ptent = ptep_get_and_clear_full(mm, addr, pte, |
a600388d2
|
1054 |
tlb->fullmm); |
1da177e4c
|
1055 1056 1057 |
tlb_remove_tlb_entry(tlb, pte, addr); if (unlikely(!page)) continue; |
eca56ff90
|
1058 1059 |
if (!PageAnon(page)) { |
1cf35d477
|
1060 |
if (pte_dirty(ptent)) { |
aac453635
|
1061 1062 1063 1064 1065 1066 |
/* * oom_reaper cannot tear down dirty * pages */ if (unlikely(details && details->ignore_dirty)) continue; |
1cf35d477
|
1067 |
force_flush = 1; |
6237bcd94
|
1068 |
set_page_dirty(page); |
1cf35d477
|
1069 |
} |
4917e5d04
|
1070 |
if (pte_young(ptent) && |
64363aad5
|
1071 |
likely(!(vma->vm_flags & VM_SEQ_READ))) |
bf3f3bc5e
|
1072 |
mark_page_accessed(page); |
6237bcd94
|
1073 |
} |
eca56ff90
|
1074 |
rss[mm_counter(page)]--; |
d281ee614
|
1075 |
page_remove_rmap(page, false); |
3dc147414
|
1076 1077 |
if (unlikely(page_mapcount(page) < 0)) print_bad_pte(vma, addr, ptent, page); |
1cf35d477
|
1078 1079 |
if (unlikely(!__tlb_remove_page(tlb, page))) { force_flush = 1; |
ce9ec37bd
|
1080 |
addr += PAGE_SIZE; |
d16dfc550
|
1081 |
break; |
1cf35d477
|
1082 |
} |
1da177e4c
|
1083 1084 |
continue; } |
aac453635
|
1085 1086 |
/* only check swap_entries if explicitly asked for in details */ if (unlikely(details && !details->check_swap_entries)) |
1da177e4c
|
1087 |
continue; |
b084d4353
|
1088 |
|
8a5f14a23
|
1089 1090 1091 1092 1093 |
entry = pte_to_swp_entry(ptent); if (!non_swap_entry(entry)) rss[MM_SWAPENTS]--; else if (is_migration_entry(entry)) { struct page *page; |
9f9f1acd7
|
1094 |
|
8a5f14a23
|
1095 |
page = migration_entry_to_page(entry); |
eca56ff90
|
1096 |
rss[mm_counter(page)]--; |
b084d4353
|
1097 |
} |
8a5f14a23
|
1098 1099 |
if (unlikely(!free_swap_and_cache(entry))) print_bad_pte(vma, addr, ptent, NULL); |
9888a1cae
|
1100 |
pte_clear_not_present_full(mm, addr, pte, tlb->fullmm); |
97a894136
|
1101 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
ae8597623
|
1102 |
|
d559db086
|
1103 |
add_mm_rss_vec(mm, rss); |
6606c3e0d
|
1104 |
arch_leave_lazy_mmu_mode(); |
51c6f666f
|
1105 |
|
1cf35d477
|
1106 |
/* Do the actual TLB flush before dropping ptl */ |
fb7332a9f
|
1107 |
if (force_flush) |
1cf35d477
|
1108 |
tlb_flush_mmu_tlbonly(tlb); |
1cf35d477
|
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 |
pte_unmap_unlock(start_pte, ptl); /* * If we forced a TLB flush (either due to running out of * batch buffers or because we needed to flush dirty TLB * entries before releasing the ptl), free the batched * memory too. Restart if we didn't do everything. */ if (force_flush) { force_flush = 0; tlb_flush_mmu_free(tlb); |
2b047252d
|
1120 1121 |
if (addr != end) |
d16dfc550
|
1122 1123 |
goto again; } |
51c6f666f
|
1124 |
return addr; |
1da177e4c
|
1125 |
} |
51c6f666f
|
1126 |
static inline unsigned long zap_pmd_range(struct mmu_gather *tlb, |
b5810039a
|
1127 |
struct vm_area_struct *vma, pud_t *pud, |
1da177e4c
|
1128 |
unsigned long addr, unsigned long end, |
97a894136
|
1129 |
struct zap_details *details) |
1da177e4c
|
1130 1131 1132 1133 1134 1135 1136 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
5c7fb56e5
|
1137 |
if (pmd_trans_huge(*pmd) || pmd_devmap(*pmd)) { |
1a5a9906d
|
1138 |
if (next - addr != HPAGE_PMD_SIZE) { |
e0897d75f
|
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 |
#ifdef CONFIG_DEBUG_VM if (!rwsem_is_locked(&tlb->mm->mmap_sem)) { pr_err("%s: mmap_sem is unlocked! addr=0x%lx end=0x%lx vma->vm_start=0x%lx vma->vm_end=0x%lx ", __func__, addr, end, vma->vm_start, vma->vm_end); BUG(); } #endif |
78ddc5347
|
1149 |
split_huge_pmd(vma, pmd, addr); |
f21760b15
|
1150 |
} else if (zap_huge_pmd(tlb, vma, pmd, addr)) |
1a5a9906d
|
1151 |
goto next; |
71e3aac07
|
1152 1153 |
/* fall through */ } |
1a5a9906d
|
1154 1155 1156 1157 1158 1159 1160 1161 1162 |
/* * Here there can be other concurrent MADV_DONTNEED or * trans huge page faults running, and if the pmd is * none or trans huge it can change under us. This is * because MADV_DONTNEED holds the mmap_sem in read * mode. */ if (pmd_none_or_trans_huge_or_clear_bad(pmd)) goto next; |
97a894136
|
1163 |
next = zap_pte_range(tlb, vma, pmd, addr, next, details); |
1a5a9906d
|
1164 |
next: |
97a894136
|
1165 1166 |
cond_resched(); } while (pmd++, addr = next, addr != end); |
51c6f666f
|
1167 1168 |
return addr; |
1da177e4c
|
1169 |
} |
51c6f666f
|
1170 |
static inline unsigned long zap_pud_range(struct mmu_gather *tlb, |
b5810039a
|
1171 |
struct vm_area_struct *vma, pgd_t *pgd, |
1da177e4c
|
1172 |
unsigned long addr, unsigned long end, |
97a894136
|
1173 |
struct zap_details *details) |
1da177e4c
|
1174 1175 1176 1177 1178 1179 1180 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
97a894136
|
1181 |
if (pud_none_or_clear_bad(pud)) |
1da177e4c
|
1182 |
continue; |
97a894136
|
1183 1184 |
next = zap_pmd_range(tlb, vma, pud, addr, next, details); } while (pud++, addr = next, addr != end); |
51c6f666f
|
1185 1186 |
return addr; |
1da177e4c
|
1187 |
} |
aac453635
|
1188 |
void unmap_page_range(struct mmu_gather *tlb, |
038c7aa16
|
1189 1190 1191 |
struct vm_area_struct *vma, unsigned long addr, unsigned long end, struct zap_details *details) |
1da177e4c
|
1192 1193 1194 |
{ pgd_t *pgd; unsigned long next; |
1da177e4c
|
1195 1196 1197 1198 1199 |
BUG_ON(addr >= end); tlb_start_vma(tlb, vma); pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); |
97a894136
|
1200 |
if (pgd_none_or_clear_bad(pgd)) |
1da177e4c
|
1201 |
continue; |
97a894136
|
1202 1203 |
next = zap_pud_range(tlb, vma, pgd, addr, next, details); } while (pgd++, addr = next, addr != end); |
1da177e4c
|
1204 1205 |
tlb_end_vma(tlb, vma); } |
51c6f666f
|
1206 |
|
f5cc4eef9
|
1207 1208 1209 |
static void unmap_single_vma(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e
|
1210 |
unsigned long end_addr, |
f5cc4eef9
|
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 |
struct zap_details *details) { unsigned long start = max(vma->vm_start, start_addr); unsigned long end; if (start >= vma->vm_end) return; end = min(vma->vm_end, end_addr); if (end <= vma->vm_start) return; |
cbc91f71b
|
1221 1222 |
if (vma->vm_file) uprobe_munmap(vma, start, end); |
b3b9c2932
|
1223 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) |
5180da410
|
1224 |
untrack_pfn(vma, 0, 0); |
f5cc4eef9
|
1225 1226 1227 1228 1229 1230 1231 |
if (start != end) { if (unlikely(is_vm_hugetlb_page(vma))) { /* * It is undesirable to test vma->vm_file as it * should be non-null for valid hugetlb area. * However, vm_file will be NULL in the error |
7aa6b4ad5
|
1232 |
* cleanup path of mmap_region. When |
f5cc4eef9
|
1233 |
* hugetlbfs ->mmap method fails, |
7aa6b4ad5
|
1234 |
* mmap_region() nullifies vma->vm_file |
f5cc4eef9
|
1235 1236 1237 1238 |
* before calling this function to clean up. * Since no pte has actually been setup, it is * safe to do nothing in this case. */ |
24669e584
|
1239 |
if (vma->vm_file) { |
83cde9e8b
|
1240 |
i_mmap_lock_write(vma->vm_file->f_mapping); |
d833352a4
|
1241 |
__unmap_hugepage_range_final(tlb, vma, start, end, NULL); |
83cde9e8b
|
1242 |
i_mmap_unlock_write(vma->vm_file->f_mapping); |
24669e584
|
1243 |
} |
f5cc4eef9
|
1244 1245 1246 |
} else unmap_page_range(tlb, vma, start, end, details); } |
1da177e4c
|
1247 |
} |
1da177e4c
|
1248 1249 |
/** * unmap_vmas - unmap a range of memory covered by a list of vma's |
0164f69d0
|
1250 |
* @tlb: address of the caller's struct mmu_gather |
1da177e4c
|
1251 1252 1253 |
* @vma: the starting vma * @start_addr: virtual address at which to start unmapping * @end_addr: virtual address at which to end unmapping |
1da177e4c
|
1254 |
* |
508034a32
|
1255 |
* Unmap all pages in the vma list. |
1da177e4c
|
1256 |
* |
1da177e4c
|
1257 1258 1259 1260 1261 1262 1263 1264 1265 |
* Only addresses between `start' and `end' will be unmapped. * * The VMA list must be sorted in ascending virtual address order. * * unmap_vmas() assumes that the caller will flush the whole unmapped address * range after unmap_vmas() returns. So the only responsibility here is to * ensure that any thus-far unmapped pages are flushed before unmap_vmas() * drops the lock and schedules. */ |
6e8bb0193
|
1266 |
void unmap_vmas(struct mmu_gather *tlb, |
1da177e4c
|
1267 |
struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e
|
1268 |
unsigned long end_addr) |
1da177e4c
|
1269 |
{ |
cddb8a5c1
|
1270 |
struct mm_struct *mm = vma->vm_mm; |
1da177e4c
|
1271 |
|
cddb8a5c1
|
1272 |
mmu_notifier_invalidate_range_start(mm, start_addr, end_addr); |
f5cc4eef9
|
1273 |
for ( ; vma && vma->vm_start < end_addr; vma = vma->vm_next) |
4f74d2c8e
|
1274 |
unmap_single_vma(tlb, vma, start_addr, end_addr, NULL); |
cddb8a5c1
|
1275 |
mmu_notifier_invalidate_range_end(mm, start_addr, end_addr); |
1da177e4c
|
1276 1277 1278 1279 1280 |
} /** * zap_page_range - remove user pages in a given range * @vma: vm_area_struct holding the applicable pages |
eb4546bbb
|
1281 |
* @start: starting address of pages to zap |
1da177e4c
|
1282 |
* @size: number of bytes to zap |
8a5f14a23
|
1283 |
* @details: details of shared cache invalidation |
f5cc4eef9
|
1284 1285 |
* * Caller must protect the VMA list |
1da177e4c
|
1286 |
*/ |
7e027b14d
|
1287 |
void zap_page_range(struct vm_area_struct *vma, unsigned long start, |
1da177e4c
|
1288 1289 1290 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550
|
1291 |
struct mmu_gather tlb; |
7e027b14d
|
1292 |
unsigned long end = start + size; |
1da177e4c
|
1293 |
|
1da177e4c
|
1294 |
lru_add_drain(); |
2b047252d
|
1295 |
tlb_gather_mmu(&tlb, mm, start, end); |
365e9c87a
|
1296 |
update_hiwater_rss(mm); |
7e027b14d
|
1297 1298 |
mmu_notifier_invalidate_range_start(mm, start, end); for ( ; vma && vma->vm_start < end; vma = vma->vm_next) |
4f74d2c8e
|
1299 |
unmap_single_vma(&tlb, vma, start, end, details); |
7e027b14d
|
1300 1301 |
mmu_notifier_invalidate_range_end(mm, start, end); tlb_finish_mmu(&tlb, start, end); |
1da177e4c
|
1302 |
} |
c627f9cc0
|
1303 |
/** |
f5cc4eef9
|
1304 1305 1306 1307 |
* zap_page_range_single - remove user pages in a given range * @vma: vm_area_struct holding the applicable pages * @address: starting address of pages to zap * @size: number of bytes to zap |
8a5f14a23
|
1308 |
* @details: details of shared cache invalidation |
f5cc4eef9
|
1309 1310 |
* * The range must fit into one VMA. |
1da177e4c
|
1311 |
*/ |
f5cc4eef9
|
1312 |
static void zap_page_range_single(struct vm_area_struct *vma, unsigned long address, |
1da177e4c
|
1313 1314 1315 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550
|
1316 |
struct mmu_gather tlb; |
1da177e4c
|
1317 |
unsigned long end = address + size; |
1da177e4c
|
1318 |
|
1da177e4c
|
1319 |
lru_add_drain(); |
2b047252d
|
1320 |
tlb_gather_mmu(&tlb, mm, address, end); |
365e9c87a
|
1321 |
update_hiwater_rss(mm); |
f5cc4eef9
|
1322 |
mmu_notifier_invalidate_range_start(mm, address, end); |
4f74d2c8e
|
1323 |
unmap_single_vma(&tlb, vma, address, end, details); |
f5cc4eef9
|
1324 |
mmu_notifier_invalidate_range_end(mm, address, end); |
d16dfc550
|
1325 |
tlb_finish_mmu(&tlb, address, end); |
1da177e4c
|
1326 |
} |
c627f9cc0
|
1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 |
/** * zap_vma_ptes - remove ptes mapping the vma * @vma: vm_area_struct holding ptes to be zapped * @address: starting address of pages to zap * @size: number of bytes to zap * * This function only unmaps ptes assigned to VM_PFNMAP vmas. * * The entire address range must be fully contained within the vma. * * Returns 0 if successful. */ int zap_vma_ptes(struct vm_area_struct *vma, unsigned long address, unsigned long size) { if (address < vma->vm_start || address + size > vma->vm_end || !(vma->vm_flags & VM_PFNMAP)) return -1; |
f5cc4eef9
|
1345 |
zap_page_range_single(vma, address, size, NULL); |
c627f9cc0
|
1346 1347 1348 |
return 0; } EXPORT_SYMBOL_GPL(zap_vma_ptes); |
25ca1d6c0
|
1349 |
pte_t *__get_locked_pte(struct mm_struct *mm, unsigned long addr, |
920c7a5d0
|
1350 |
spinlock_t **ptl) |
c9cfcddfd
|
1351 1352 1353 1354 |
{ pgd_t * pgd = pgd_offset(mm, addr); pud_t * pud = pud_alloc(mm, pgd, addr); if (pud) { |
49c91fb01
|
1355 |
pmd_t * pmd = pmd_alloc(mm, pud, addr); |
f66055ab6
|
1356 1357 |
if (pmd) { VM_BUG_ON(pmd_trans_huge(*pmd)); |
c9cfcddfd
|
1358 |
return pte_alloc_map_lock(mm, pmd, addr, ptl); |
f66055ab6
|
1359 |
} |
c9cfcddfd
|
1360 1361 1362 |
} return NULL; } |
1da177e4c
|
1363 |
/* |
238f58d89
|
1364 1365 1366 1367 1368 1369 |
* This is the old fallback for page remapping. * * For historical reasons, it only allows reserved pages. Only * old drivers should use this, and they needed to mark their * pages reserved for the old functions anyway. */ |
423bad600
|
1370 1371 |
static int insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot) |
238f58d89
|
1372 |
{ |
423bad600
|
1373 |
struct mm_struct *mm = vma->vm_mm; |
238f58d89
|
1374 |
int retval; |
c9cfcddfd
|
1375 |
pte_t *pte; |
8a9f3ccd2
|
1376 |
spinlock_t *ptl; |
238f58d89
|
1377 |
retval = -EINVAL; |
a145dd411
|
1378 |
if (PageAnon(page)) |
5b4e655e9
|
1379 |
goto out; |
238f58d89
|
1380 1381 |
retval = -ENOMEM; flush_dcache_page(page); |
c9cfcddfd
|
1382 |
pte = get_locked_pte(mm, addr, &ptl); |
238f58d89
|
1383 |
if (!pte) |
5b4e655e9
|
1384 |
goto out; |
238f58d89
|
1385 1386 1387 1388 1389 1390 |
retval = -EBUSY; if (!pte_none(*pte)) goto out_unlock; /* Ok, finally just insert the thing.. */ get_page(page); |
eca56ff90
|
1391 |
inc_mm_counter_fast(mm, mm_counter_file(page)); |
238f58d89
|
1392 1393 1394 1395 |
page_add_file_rmap(page); set_pte_at(mm, addr, pte, mk_pte(page, prot)); retval = 0; |
8a9f3ccd2
|
1396 1397 |
pte_unmap_unlock(pte, ptl); return retval; |
238f58d89
|
1398 1399 1400 1401 1402 |
out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
bfa5bf6d6
|
1403 1404 1405 1406 1407 1408 |
/** * vm_insert_page - insert single page into user vma * @vma: user vma to map to * @addr: target user address of this page * @page: source kernel page * |
a145dd411
|
1409 1410 1411 1412 1413 1414 |
* This allows drivers to insert individual pages they've allocated * into a user vma. * * The page has to be a nice clean _individual_ kernel allocation. * If you allocate a compound page, you need to have marked it as * such (__GFP_COMP), or manually just split the page up yourself |
8dfcc9ba2
|
1415 |
* (see split_page()). |
a145dd411
|
1416 1417 1418 1419 1420 1421 1422 1423 |
* * NOTE! Traditionally this was done with "remap_pfn_range()" which * took an arbitrary page protection parameter. This doesn't allow * that. Your vma protection will have to be set up correctly, which * means that if you want a shared writable mapping, you'd better * ask for a shared writable mapping! * * The page does not need to be reserved. |
4b6e1e370
|
1424 1425 1426 1427 1428 |
* * Usually this function is called from f_op->mmap() handler * under mm->mmap_sem write-lock, so it can change vma->vm_flags. * Caller must set VM_MIXEDMAP on vma if it wants to call this * function from other places, for example from page-fault handler. |
a145dd411
|
1429 |
*/ |
423bad600
|
1430 1431 |
int vm_insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page) |
a145dd411
|
1432 1433 1434 1435 1436 |
{ if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; if (!page_count(page)) return -EINVAL; |
4b6e1e370
|
1437 1438 1439 1440 1441 |
if (!(vma->vm_flags & VM_MIXEDMAP)) { BUG_ON(down_read_trylock(&vma->vm_mm->mmap_sem)); BUG_ON(vma->vm_flags & VM_PFNMAP); vma->vm_flags |= VM_MIXEDMAP; } |
423bad600
|
1442 |
return insert_page(vma, addr, page, vma->vm_page_prot); |
a145dd411
|
1443 |
} |
e3c3374fb
|
1444 |
EXPORT_SYMBOL(vm_insert_page); |
a145dd411
|
1445 |
|
423bad600
|
1446 |
static int insert_pfn(struct vm_area_struct *vma, unsigned long addr, |
01c8f1c44
|
1447 |
pfn_t pfn, pgprot_t prot) |
423bad600
|
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 |
{ struct mm_struct *mm = vma->vm_mm; int retval; pte_t *pte, entry; spinlock_t *ptl; retval = -ENOMEM; pte = get_locked_pte(mm, addr, &ptl); if (!pte) goto out; retval = -EBUSY; if (!pte_none(*pte)) goto out_unlock; /* Ok, finally just insert the thing.. */ |
01c8f1c44
|
1463 1464 1465 1466 |
if (pfn_t_devmap(pfn)) entry = pte_mkdevmap(pfn_t_pte(pfn, prot)); else entry = pte_mkspecial(pfn_t_pte(pfn, prot)); |
423bad600
|
1467 |
set_pte_at(mm, addr, pte, entry); |
4b3073e1c
|
1468 |
update_mmu_cache(vma, addr, pte); /* XXX: why not for insert_page? */ |
423bad600
|
1469 1470 1471 1472 1473 1474 1475 |
retval = 0; out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
e0dc0d8f4
|
1476 1477 1478 1479 1480 1481 |
/** * vm_insert_pfn - insert single pfn into user vma * @vma: user vma to map to * @addr: target user address of this page * @pfn: source kernel pfn * |
c462f179e
|
1482 |
* Similar to vm_insert_page, this allows drivers to insert individual pages |
e0dc0d8f4
|
1483 1484 1485 1486 |
* they've allocated into a user vma. Same comments apply. * * This function should only be called from a vm_ops->fault handler, and * in that case the handler should return NULL. |
0d71d10a4
|
1487 1488 1489 1490 1491 |
* * vma cannot be a COW mapping. * * As this is called only for pages that do not currently exist, we * do not need to flush old virtual caches or the TLB. |
e0dc0d8f4
|
1492 1493 |
*/ int vm_insert_pfn(struct vm_area_struct *vma, unsigned long addr, |
423bad600
|
1494 |
unsigned long pfn) |
e0dc0d8f4
|
1495 |
{ |
1745cbc5d
|
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 |
return vm_insert_pfn_prot(vma, addr, pfn, vma->vm_page_prot); } EXPORT_SYMBOL(vm_insert_pfn); /** * vm_insert_pfn_prot - insert single pfn into user vma with specified pgprot * @vma: user vma to map to * @addr: target user address of this page * @pfn: source kernel pfn * @pgprot: pgprot flags for the inserted page * * This is exactly like vm_insert_pfn, except that it allows drivers to * to override pgprot on a per-page basis. * * This only makes sense for IO mappings, and it makes no sense for * cow mappings. In general, using multiple vmas is preferable; * vm_insert_pfn_prot should only be used if using multiple VMAs is * impractical. */ int vm_insert_pfn_prot(struct vm_area_struct *vma, unsigned long addr, unsigned long pfn, pgprot_t pgprot) { |
2ab640379
|
1518 |
int ret; |
7e675137a
|
1519 1520 1521 1522 1523 1524 |
/* * Technically, architectures with pte_special can avoid all these * restrictions (same for remap_pfn_range). However we would like * consistency in testing and feature parity among all, so we should * try to keep these invariants in place for everybody. */ |
b379d7901
|
1525 1526 1527 1528 1529 |
BUG_ON(!(vma->vm_flags & (VM_PFNMAP|VM_MIXEDMAP))); BUG_ON((vma->vm_flags & (VM_PFNMAP|VM_MIXEDMAP)) == (VM_PFNMAP|VM_MIXEDMAP)); BUG_ON((vma->vm_flags & VM_PFNMAP) && is_cow_mapping(vma->vm_flags)); BUG_ON((vma->vm_flags & VM_MIXEDMAP) && pfn_valid(pfn)); |
e0dc0d8f4
|
1530 |
|
423bad600
|
1531 1532 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
f25748e3c
|
1533 |
if (track_pfn_insert(vma, &pgprot, __pfn_to_pfn_t(pfn, PFN_DEV))) |
2ab640379
|
1534 |
return -EINVAL; |
01c8f1c44
|
1535 |
ret = insert_pfn(vma, addr, __pfn_to_pfn_t(pfn, PFN_DEV), pgprot); |
2ab640379
|
1536 |
|
2ab640379
|
1537 |
return ret; |
423bad600
|
1538 |
} |
1745cbc5d
|
1539 |
EXPORT_SYMBOL(vm_insert_pfn_prot); |
e0dc0d8f4
|
1540 |
|
423bad600
|
1541 |
int vm_insert_mixed(struct vm_area_struct *vma, unsigned long addr, |
01c8f1c44
|
1542 |
pfn_t pfn) |
423bad600
|
1543 1544 |
{ BUG_ON(!(vma->vm_flags & VM_MIXEDMAP)); |
e0dc0d8f4
|
1545 |
|
423bad600
|
1546 1547 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
e0dc0d8f4
|
1548 |
|
423bad600
|
1549 1550 1551 1552 |
/* * If we don't have pte special, then we have to use the pfn_valid() * based VM_MIXEDMAP scheme (see vm_normal_page), and thus we *must* * refcount the page if pfn_valid is true (hence insert_page rather |
62eede62d
|
1553 1554 |
* than insert_pfn). If a zero_pfn were inserted into a VM_MIXEDMAP * without pte special, it would there be refcounted as a normal page. |
423bad600
|
1555 |
*/ |
03fc2da63
|
1556 |
if (!HAVE_PTE_SPECIAL && !pfn_t_devmap(pfn) && pfn_t_valid(pfn)) { |
423bad600
|
1557 |
struct page *page; |
03fc2da63
|
1558 1559 1560 1561 1562 1563 |
/* * At this point we are committed to insert_page() * regardless of whether the caller specified flags that * result in pfn_t_has_page() == false. */ page = pfn_to_page(pfn_t_to_pfn(pfn)); |
423bad600
|
1564 1565 1566 |
return insert_page(vma, addr, page, vma->vm_page_prot); } return insert_pfn(vma, addr, pfn, vma->vm_page_prot); |
e0dc0d8f4
|
1567 |
} |
423bad600
|
1568 |
EXPORT_SYMBOL(vm_insert_mixed); |
e0dc0d8f4
|
1569 |
|
a145dd411
|
1570 |
/* |
1da177e4c
|
1571 1572 1573 1574 1575 1576 1577 1578 1579 |
* maps a range of physical memory into the requested pages. the old * mappings are removed. any references to nonexistent pages results * in null mappings (currently treated as "copy-on-access") */ static int remap_pte_range(struct mm_struct *mm, pmd_t *pmd, unsigned long addr, unsigned long end, unsigned long pfn, pgprot_t prot) { pte_t *pte; |
c74df32c7
|
1580 |
spinlock_t *ptl; |
1da177e4c
|
1581 |
|
c74df32c7
|
1582 |
pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); |
1da177e4c
|
1583 1584 |
if (!pte) return -ENOMEM; |
6606c3e0d
|
1585 |
arch_enter_lazy_mmu_mode(); |
1da177e4c
|
1586 1587 |
do { BUG_ON(!pte_none(*pte)); |
7e675137a
|
1588 |
set_pte_at(mm, addr, pte, pte_mkspecial(pfn_pte(pfn, prot))); |
1da177e4c
|
1589 1590 |
pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); |
6606c3e0d
|
1591 |
arch_leave_lazy_mmu_mode(); |
c74df32c7
|
1592 |
pte_unmap_unlock(pte - 1, ptl); |
1da177e4c
|
1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 |
return 0; } static inline int remap_pmd_range(struct mm_struct *mm, pud_t *pud, unsigned long addr, unsigned long end, unsigned long pfn, pgprot_t prot) { pmd_t *pmd; unsigned long next; pfn -= addr >> PAGE_SHIFT; pmd = pmd_alloc(mm, pud, addr); if (!pmd) return -ENOMEM; |
f66055ab6
|
1607 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c
|
1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 |
do { next = pmd_addr_end(addr, end); if (remap_pte_range(mm, pmd, addr, next, pfn + (addr >> PAGE_SHIFT), prot)) return -ENOMEM; } while (pmd++, addr = next, addr != end); return 0; } static inline int remap_pud_range(struct mm_struct *mm, pgd_t *pgd, unsigned long addr, unsigned long end, unsigned long pfn, pgprot_t prot) { pud_t *pud; unsigned long next; pfn -= addr >> PAGE_SHIFT; pud = pud_alloc(mm, pgd, addr); if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); if (remap_pmd_range(mm, pud, addr, next, pfn + (addr >> PAGE_SHIFT), prot)) return -ENOMEM; } while (pud++, addr = next, addr != end); return 0; } |
bfa5bf6d6
|
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 |
/** * remap_pfn_range - remap kernel memory to userspace * @vma: user vma to map to * @addr: target user address to start at * @pfn: physical address of kernel memory * @size: size of map area * @prot: page protection flags for this mapping * * Note: this is only safe if the mm semaphore is held when called. */ |
1da177e4c
|
1646 1647 1648 1649 1650 |
int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr, unsigned long pfn, unsigned long size, pgprot_t prot) { pgd_t *pgd; unsigned long next; |
2d15cab85
|
1651 |
unsigned long end = addr + PAGE_ALIGN(size); |
1da177e4c
|
1652 1653 1654 1655 1656 1657 1658 1659 |
struct mm_struct *mm = vma->vm_mm; int err; /* * Physically remapped pages are special. Tell the * rest of the world about it: * VM_IO tells people not to look at these pages * (accesses can have side effects). |
6aab341e0
|
1660 1661 1662 |
* VM_PFNMAP tells the core MM that the base pages are just * raw PFN mappings, and do not have a "struct page" associated * with them. |
314e51b98
|
1663 1664 1665 1666 |
* VM_DONTEXPAND * Disable vma merging and expanding with mremap(). * VM_DONTDUMP * Omit vma from core dump, even when VM_IO turned off. |
fb155c161
|
1667 1668 1669 1670 |
* * There's a horrible special case to handle copy-on-write * behaviour that some programs depend on. We mark the "original" * un-COW'ed pages by matching them up with "vma->vm_pgoff". |
b3b9c2932
|
1671 |
* See vm_normal_page() for details. |
1da177e4c
|
1672 |
*/ |
b3b9c2932
|
1673 1674 1675 |
if (is_cow_mapping(vma->vm_flags)) { if (addr != vma->vm_start || end != vma->vm_end) return -EINVAL; |
fb155c161
|
1676 |
vma->vm_pgoff = pfn; |
b3b9c2932
|
1677 1678 1679 1680 |
} err = track_pfn_remap(vma, &prot, pfn, addr, PAGE_ALIGN(size)); if (err) |
3c8bb73ac
|
1681 |
return -EINVAL; |
fb155c161
|
1682 |
|
314e51b98
|
1683 |
vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP; |
1da177e4c
|
1684 1685 1686 1687 1688 |
BUG_ON(addr >= end); pfn -= addr >> PAGE_SHIFT; pgd = pgd_offset(mm, addr); flush_cache_range(vma, addr, end); |
1da177e4c
|
1689 1690 1691 1692 1693 1694 1695 |
do { next = pgd_addr_end(addr, end); err = remap_pud_range(mm, pgd, addr, next, pfn + (addr >> PAGE_SHIFT), prot); if (err) break; } while (pgd++, addr = next, addr != end); |
2ab640379
|
1696 1697 |
if (err) |
5180da410
|
1698 |
untrack_pfn(vma, pfn, PAGE_ALIGN(size)); |
2ab640379
|
1699 |
|
1da177e4c
|
1700 1701 1702 |
return err; } EXPORT_SYMBOL(remap_pfn_range); |
b4cbb197c
|
1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 |
/** * vm_iomap_memory - remap memory to userspace * @vma: user vma to map to * @start: start of area * @len: size of area * * This is a simplified io_remap_pfn_range() for common driver use. The * driver just needs to give us the physical memory range to be mapped, * we'll figure out the rest from the vma information. * * NOTE! Some drivers might want to tweak vma->vm_page_prot first to get * whatever write-combining details or similar. */ int vm_iomap_memory(struct vm_area_struct *vma, phys_addr_t start, unsigned long len) { unsigned long vm_len, pfn, pages; /* Check that the physical memory area passed in looks valid */ if (start + len < start) return -EINVAL; /* * You *really* shouldn't map things that aren't page-aligned, * but we've historically allowed it because IO memory might * just have smaller alignment. */ len += start & ~PAGE_MASK; pfn = start >> PAGE_SHIFT; pages = (len + ~PAGE_MASK) >> PAGE_SHIFT; if (pfn + pages < pfn) return -EINVAL; /* We start the mapping 'vm_pgoff' pages into the area */ if (vma->vm_pgoff > pages) return -EINVAL; pfn += vma->vm_pgoff; pages -= vma->vm_pgoff; /* Can we fit all of the mapping? */ vm_len = vma->vm_end - vma->vm_start; if (vm_len >> PAGE_SHIFT > pages) return -EINVAL; /* Ok, let it rip */ return io_remap_pfn_range(vma, vma->vm_start, pfn, vm_len, vma->vm_page_prot); } EXPORT_SYMBOL(vm_iomap_memory); |
aee16b3ce
|
1749 1750 1751 1752 1753 1754 |
static int apply_to_pte_range(struct mm_struct *mm, pmd_t *pmd, unsigned long addr, unsigned long end, pte_fn_t fn, void *data) { pte_t *pte; int err; |
2f569afd9
|
1755 |
pgtable_t token; |
949099148
|
1756 |
spinlock_t *uninitialized_var(ptl); |
aee16b3ce
|
1757 1758 1759 1760 1761 1762 1763 1764 |
pte = (mm == &init_mm) ? pte_alloc_kernel(pmd, addr) : pte_alloc_map_lock(mm, pmd, addr, &ptl); if (!pte) return -ENOMEM; BUG_ON(pmd_huge(*pmd)); |
38e0edb15
|
1765 |
arch_enter_lazy_mmu_mode(); |
2f569afd9
|
1766 |
token = pmd_pgtable(*pmd); |
aee16b3ce
|
1767 1768 |
do { |
c36987e2e
|
1769 |
err = fn(pte++, token, addr, data); |
aee16b3ce
|
1770 1771 |
if (err) break; |
c36987e2e
|
1772 |
} while (addr += PAGE_SIZE, addr != end); |
aee16b3ce
|
1773 |
|
38e0edb15
|
1774 |
arch_leave_lazy_mmu_mode(); |
aee16b3ce
|
1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 |
if (mm != &init_mm) pte_unmap_unlock(pte-1, ptl); return err; } static int apply_to_pmd_range(struct mm_struct *mm, pud_t *pud, unsigned long addr, unsigned long end, pte_fn_t fn, void *data) { pmd_t *pmd; unsigned long next; int err; |
ceb868796
|
1787 |
BUG_ON(pud_huge(*pud)); |
aee16b3ce
|
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 |
pmd = pmd_alloc(mm, pud, addr); if (!pmd) return -ENOMEM; do { next = pmd_addr_end(addr, end); err = apply_to_pte_range(mm, pmd, addr, next, fn, data); if (err) break; } while (pmd++, addr = next, addr != end); return err; } static int apply_to_pud_range(struct mm_struct *mm, pgd_t *pgd, unsigned long addr, unsigned long end, pte_fn_t fn, void *data) { pud_t *pud; unsigned long next; int err; pud = pud_alloc(mm, pgd, addr); if (!pud) return -ENOMEM; do { next = pud_addr_end(addr, end); err = apply_to_pmd_range(mm, pud, addr, next, fn, data); if (err) break; } while (pud++, addr = next, addr != end); return err; } /* * Scan a region of virtual memory, filling in page tables as necessary * and calling a provided function on each leaf page table. */ int apply_to_page_range(struct mm_struct *mm, unsigned long addr, unsigned long size, pte_fn_t fn, void *data) { pgd_t *pgd; unsigned long next; |
57250a5bf
|
1829 |
unsigned long end = addr + size; |
aee16b3ce
|
1830 |
int err; |
9cb65bc3b
|
1831 1832 |
if (WARN_ON(addr >= end)) return -EINVAL; |
aee16b3ce
|
1833 1834 1835 1836 1837 1838 1839 |
pgd = pgd_offset(mm, addr); do { next = pgd_addr_end(addr, end); err = apply_to_pud_range(mm, pgd, addr, next, fn, data); if (err) break; } while (pgd++, addr = next, addr != end); |
57250a5bf
|
1840 |
|
aee16b3ce
|
1841 1842 1843 |
return err; } EXPORT_SYMBOL_GPL(apply_to_page_range); |
1da177e4c
|
1844 |
/* |
9b4bdd2ff
|
1845 1846 1847 1848 1849 |
* handle_pte_fault chooses page fault handler according to an entry which was * read non-atomically. Before making any commitment, on those architectures * or configurations (e.g. i386 with PAE) which might give a mix of unmatched * parts, do_swap_page must check under lock before unmapping the pte and * proceeding (but do_wp_page is only called after already making such a check; |
a335b2e17
|
1850 |
* and do_anonymous_page can safely check later on). |
8f4e2101f
|
1851 |
*/ |
4c21e2f24
|
1852 |
static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd, |
8f4e2101f
|
1853 1854 1855 1856 1857 |
pte_t *page_table, pte_t orig_pte) { int same = 1; #if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT) if (sizeof(pte_t) > sizeof(unsigned long)) { |
4c21e2f24
|
1858 1859 |
spinlock_t *ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
8f4e2101f
|
1860 |
same = pte_same(*page_table, orig_pte); |
4c21e2f24
|
1861 |
spin_unlock(ptl); |
8f4e2101f
|
1862 1863 1864 1865 1866 |
} #endif pte_unmap(page_table); return same; } |
9de455b20
|
1867 |
static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma) |
6aab341e0
|
1868 |
{ |
0abdd7a81
|
1869 |
debug_dma_assert_idle(src); |
6aab341e0
|
1870 1871 1872 1873 1874 1875 1876 |
/* * If the source page was a PFN mapping, we don't have * a "struct page" for it. We do a best-effort copy by * just copying from the original user address. If that * fails, we just zero-fill it. Live with it. */ if (unlikely(!src)) { |
9b04c5fec
|
1877 |
void *kaddr = kmap_atomic(dst); |
5d2a2dbbc
|
1878 1879 1880 1881 1882 1883 1884 1885 1886 |
void __user *uaddr = (void __user *)(va & PAGE_MASK); /* * This really shouldn't fail, because the page is there * in the page tables. But it might just be unreadable, * in which case we just give up and fill the result with * zeroes. */ if (__copy_from_user_inatomic(kaddr, uaddr, PAGE_SIZE)) |
3ecb01df3
|
1887 |
clear_page(kaddr); |
9b04c5fec
|
1888 |
kunmap_atomic(kaddr); |
c4ec7b0de
|
1889 |
flush_dcache_page(dst); |
0ed361dec
|
1890 1891 |
} else copy_user_highpage(dst, src, va, vma); |
6aab341e0
|
1892 |
} |
c20cd45eb
|
1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 |
static gfp_t __get_fault_gfp_mask(struct vm_area_struct *vma) { struct file *vm_file = vma->vm_file; if (vm_file) return mapping_gfp_mask(vm_file->f_mapping) | __GFP_FS | __GFP_IO; /* * Special mappings (e.g. VDSO) do not have any file so fake * a default GFP_KERNEL for them. */ return GFP_KERNEL; } |
1da177e4c
|
1906 |
/* |
fb09a4642
|
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 |
* Notify the address space that the page is about to become writable so that * it can prohibit this or wait for the page to get into an appropriate state. * * We do this without the lock held, so that it can sleep if it needs to. */ static int do_page_mkwrite(struct vm_area_struct *vma, struct page *page, unsigned long address) { struct vm_fault vmf; int ret; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = page->index; vmf.flags = FAULT_FLAG_WRITE|FAULT_FLAG_MKWRITE; |
c20cd45eb
|
1921 |
vmf.gfp_mask = __get_fault_gfp_mask(vma); |
fb09a4642
|
1922 |
vmf.page = page; |
2e4cdab05
|
1923 |
vmf.cow_page = NULL; |
fb09a4642
|
1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 |
ret = vma->vm_ops->page_mkwrite(vma, &vmf); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) return ret; if (unlikely(!(ret & VM_FAULT_LOCKED))) { lock_page(page); if (!page->mapping) { unlock_page(page); return 0; /* retry */ } ret |= VM_FAULT_LOCKED; } else VM_BUG_ON_PAGE(!PageLocked(page), page); return ret; } /* |
4e047f897
|
1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 |
* Handle write page faults for pages that can be reused in the current vma * * This can happen either due to the mapping being with the VM_SHARED flag, * or due to us being the last reference standing to the page. In either * case, all we need to do here is to mark the page as writable and update * any related book-keeping. */ static inline int wp_page_reuse(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, spinlock_t *ptl, pte_t orig_pte, struct page *page, int page_mkwrite, int dirty_shared) __releases(ptl) { pte_t entry; /* * Clear the pages cpupid information as the existing * information potentially belongs to a now completely * unrelated process. */ if (page) page_cpupid_xchg_last(page, (1 << LAST_CPUPID_SHIFT) - 1); flush_cache_page(vma, address, pte_pfn(orig_pte)); entry = pte_mkyoung(orig_pte); entry = maybe_mkwrite(pte_mkdirty(entry), vma); if (ptep_set_access_flags(vma, address, page_table, entry, 1)) update_mmu_cache(vma, address, page_table); pte_unmap_unlock(page_table, ptl); if (dirty_shared) { struct address_space *mapping; int dirtied; if (!page_mkwrite) lock_page(page); dirtied = set_page_dirty(page); VM_BUG_ON_PAGE(PageAnon(page), page); mapping = page->mapping; unlock_page(page); |
09cbfeaf1
|
1982 |
put_page(page); |
4e047f897
|
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 |
if ((dirtied || page_mkwrite) && mapping) { /* * Some device drivers do not set page.mapping * but still dirty their pages */ balance_dirty_pages_ratelimited(mapping); } if (!page_mkwrite) file_update_time(vma->vm_file); } return VM_FAULT_WRITE; } /* |
2f38ab2c3
|
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 |
* Handle the case of a page which we actually need to copy to a new page. * * Called with mmap_sem locked and the old page referenced, but * without the ptl held. * * High level logic flow: * * - Allocate a page, copy the content of the old page to the new one. * - Handle book keeping and accounting - cgroups, mmu-notifiers, etc. * - Take the PTL. If the pte changed, bail out and release the allocated page * - If the pte is still the way we remember it, update the page table and all * relevant references. This includes dropping the reference the page-table * held to the old page, as well as updating the rmap. * - In any case, unlock the PTL and drop the reference we took to the old page. */ static int wp_page_copy(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, pte_t orig_pte, struct page *old_page) { struct page *new_page = NULL; spinlock_t *ptl = NULL; pte_t entry; int page_copied = 0; const unsigned long mmun_start = address & PAGE_MASK; /* For mmu_notifiers */ const unsigned long mmun_end = mmun_start + PAGE_SIZE; /* For mmu_notifiers */ struct mem_cgroup *memcg; if (unlikely(anon_vma_prepare(vma))) goto oom; if (is_zero_pfn(pte_pfn(orig_pte))) { new_page = alloc_zeroed_user_highpage_movable(vma, address); if (!new_page) goto oom; } else { new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); if (!new_page) goto oom; cow_user_page(new_page, old_page, address, vma); } |
2f38ab2c3
|
2040 |
|
f627c2f53
|
2041 |
if (mem_cgroup_try_charge(new_page, mm, GFP_KERNEL, &memcg, false)) |
2f38ab2c3
|
2042 |
goto oom_free_new; |
eb3c24f30
|
2043 |
__SetPageUptodate(new_page); |
2f38ab2c3
|
2044 2045 2046 2047 2048 2049 2050 2051 2052 |
mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end); /* * Re-check the pte - we dropped the lock */ page_table = pte_offset_map_lock(mm, pmd, address, &ptl); if (likely(pte_same(*page_table, orig_pte))) { if (old_page) { if (!PageAnon(old_page)) { |
eca56ff90
|
2053 2054 |
dec_mm_counter_fast(mm, mm_counter_file(old_page)); |
2f38ab2c3
|
2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 |
inc_mm_counter_fast(mm, MM_ANONPAGES); } } else { inc_mm_counter_fast(mm, MM_ANONPAGES); } flush_cache_page(vma, address, pte_pfn(orig_pte)); entry = mk_pte(new_page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); /* * Clear the pte entry and flush it first, before updating the * pte with the new entry. This will avoid a race condition * seen in the presence of one thread doing SMC and another * thread doing COW. */ ptep_clear_flush_notify(vma, address, page_table); |
d281ee614
|
2070 |
page_add_new_anon_rmap(new_page, vma, address, false); |
f627c2f53
|
2071 |
mem_cgroup_commit_charge(new_page, memcg, false, false); |
2f38ab2c3
|
2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 |
lru_cache_add_active_or_unevictable(new_page, vma); /* * We call the notify macro here because, when using secondary * mmu page tables (such as kvm shadow page tables), we want the * new page to be mapped directly into the secondary page table. */ set_pte_at_notify(mm, address, page_table, entry); update_mmu_cache(vma, address, page_table); if (old_page) { /* * Only after switching the pte to the new page may * we remove the mapcount here. Otherwise another * process may come and find the rmap count decremented * before the pte is switched to the new page, and * "reuse" the old page writing into it while our pte * here still points into it and can be read by other * threads. * * The critical issue is to order this * page_remove_rmap with the ptp_clear_flush above. * Those stores are ordered by (if nothing else,) * the barrier present in the atomic_add_negative * in page_remove_rmap. * * Then the TLB flush in ptep_clear_flush ensures that * no process can access the old page before the * decremented mapcount is visible. And the old page * cannot be reused until after the decremented * mapcount is visible. So transitively, TLBs to * old page will be flushed before it can be reused. */ |
d281ee614
|
2103 |
page_remove_rmap(old_page, false); |
2f38ab2c3
|
2104 2105 2106 2107 2108 2109 |
} /* Free the old page.. */ new_page = old_page; page_copied = 1; } else { |
f627c2f53
|
2110 |
mem_cgroup_cancel_charge(new_page, memcg, false); |
2f38ab2c3
|
2111 2112 2113 |
} if (new_page) |
09cbfeaf1
|
2114 |
put_page(new_page); |
2f38ab2c3
|
2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 |
pte_unmap_unlock(page_table, ptl); mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end); if (old_page) { /* * Don't let another task, with possibly unlocked vma, * keep the mlocked page. */ if (page_copied && (vma->vm_flags & VM_LOCKED)) { lock_page(old_page); /* LRU manipulation */ |
e90309c9f
|
2125 2126 |
if (PageMlocked(old_page)) munlock_vma_page(old_page); |
2f38ab2c3
|
2127 2128 |
unlock_page(old_page); } |
09cbfeaf1
|
2129 |
put_page(old_page); |
2f38ab2c3
|
2130 2131 2132 |
} return page_copied ? VM_FAULT_WRITE : 0; oom_free_new: |
09cbfeaf1
|
2133 |
put_page(new_page); |
2f38ab2c3
|
2134 2135 |
oom: if (old_page) |
09cbfeaf1
|
2136 |
put_page(old_page); |
2f38ab2c3
|
2137 2138 |
return VM_FAULT_OOM; } |
dd9061846
|
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 |
/* * Handle write page faults for VM_MIXEDMAP or VM_PFNMAP for a VM_SHARED * mapping */ static int wp_pfn_shared(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, spinlock_t *ptl, pte_t orig_pte, pmd_t *pmd) { if (vma->vm_ops && vma->vm_ops->pfn_mkwrite) { struct vm_fault vmf = { .page = NULL, .pgoff = linear_page_index(vma, address), .virtual_address = (void __user *)(address & PAGE_MASK), .flags = FAULT_FLAG_WRITE | FAULT_FLAG_MKWRITE, }; int ret; pte_unmap_unlock(page_table, ptl); ret = vma->vm_ops->pfn_mkwrite(vma, &vmf); if (ret & VM_FAULT_ERROR) return ret; page_table = pte_offset_map_lock(mm, pmd, address, &ptl); /* * We might have raced with another page fault while we * released the pte_offset_map_lock. */ if (!pte_same(*page_table, orig_pte)) { pte_unmap_unlock(page_table, ptl); return 0; } } return wp_page_reuse(mm, vma, address, page_table, ptl, orig_pte, NULL, 0, 0); } |
93e478d4c
|
2174 2175 2176 2177 2178 2179 2180 |
static int wp_page_shared(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, spinlock_t *ptl, pte_t orig_pte, struct page *old_page) __releases(ptl) { int page_mkwrite = 0; |
09cbfeaf1
|
2181 |
get_page(old_page); |
93e478d4c
|
2182 |
|
93e478d4c
|
2183 2184 2185 2186 2187 2188 2189 |
if (vma->vm_ops && vma->vm_ops->page_mkwrite) { int tmp; pte_unmap_unlock(page_table, ptl); tmp = do_page_mkwrite(vma, old_page, address); if (unlikely(!tmp || (tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) { |
09cbfeaf1
|
2190 |
put_page(old_page); |
93e478d4c
|
2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 |
return tmp; } /* * Since we dropped the lock we need to revalidate * the PTE as someone else may have changed it. If * they did, we just return, as we can count on the * MMU to tell us if they didn't also make it writable. */ page_table = pte_offset_map_lock(mm, pmd, address, &ptl); if (!pte_same(*page_table, orig_pte)) { unlock_page(old_page); pte_unmap_unlock(page_table, ptl); |
09cbfeaf1
|
2204 |
put_page(old_page); |
93e478d4c
|
2205 2206 2207 2208 2209 2210 2211 2212 |
return 0; } page_mkwrite = 1; } return wp_page_reuse(mm, vma, address, page_table, ptl, orig_pte, old_page, page_mkwrite, 1); } |
2f38ab2c3
|
2213 |
/* |
1da177e4c
|
2214 2215 2216 2217 |
* This routine handles present pages, when users try to write * to a shared page. It is done by copying the page to a new address * and decrementing the shared-page counter for the old page. * |
1da177e4c
|
2218 2219 2220 2221 2222 2223 2224 2225 2226 |
* Note that this routine assumes that the protection checks have been * done by the caller (the low-level page fault routine in most cases). * Thus we can safely just mark it writable once we've done any necessary * COW. * * We also mark the page dirty at this point even though the page will * change only once the write actually happens. This avoids a few races, * and potentially makes it more efficient. * |
8f4e2101f
|
2227 2228 2229 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), with pte both mapped and locked. * We return with mmap_sem still held, but pte unmapped and unlocked. |
1da177e4c
|
2230 |
*/ |
65500d234
|
2231 2232 |
static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, |
8f4e2101f
|
2233 |
spinlock_t *ptl, pte_t orig_pte) |
e6219ec81
|
2234 |
__releases(ptl) |
1da177e4c
|
2235 |
{ |
2f38ab2c3
|
2236 |
struct page *old_page; |
1da177e4c
|
2237 |
|
6aab341e0
|
2238 |
old_page = vm_normal_page(vma, address, orig_pte); |
251b97f55
|
2239 2240 |
if (!old_page) { /* |
64e455079
|
2241 2242 |
* VM_MIXEDMAP !pfn_valid() case, or VM_SOFTDIRTY clear on a * VM_PFNMAP VMA. |
251b97f55
|
2243 2244 |
* * We should not cow pages in a shared writeable mapping. |
dd9061846
|
2245 |
* Just mark the pages writable and/or call ops->pfn_mkwrite. |
251b97f55
|
2246 2247 2248 |
*/ if ((vma->vm_flags & (VM_WRITE|VM_SHARED)) == (VM_WRITE|VM_SHARED)) |
dd9061846
|
2249 2250 |
return wp_pfn_shared(mm, vma, address, page_table, ptl, orig_pte, pmd); |
2f38ab2c3
|
2251 2252 2253 2254 |
pte_unmap_unlock(page_table, ptl); return wp_page_copy(mm, vma, address, page_table, pmd, orig_pte, old_page); |
251b97f55
|
2255 |
} |
1da177e4c
|
2256 |
|
d08b3851d
|
2257 |
/* |
ee6a64578
|
2258 2259 |
* Take out anonymous pages first, anonymous shared vmas are * not dirty accountable. |
d08b3851d
|
2260 |
*/ |
9a8408951
|
2261 |
if (PageAnon(old_page) && !PageKsm(old_page)) { |
ab967d860
|
2262 |
if (!trylock_page(old_page)) { |
09cbfeaf1
|
2263 |
get_page(old_page); |
ab967d860
|
2264 2265 2266 2267 2268 2269 |
pte_unmap_unlock(page_table, ptl); lock_page(old_page); page_table = pte_offset_map_lock(mm, pmd, address, &ptl); if (!pte_same(*page_table, orig_pte)) { unlock_page(old_page); |
287668052
|
2270 |
pte_unmap_unlock(page_table, ptl); |
09cbfeaf1
|
2271 |
put_page(old_page); |
287668052
|
2272 |
return 0; |
ab967d860
|
2273 |
} |
09cbfeaf1
|
2274 |
put_page(old_page); |
ee6a64578
|
2275 |
} |
b009c024f
|
2276 |
if (reuse_swap_page(old_page)) { |
c44b67432
|
2277 2278 2279 2280 2281 2282 |
/* * The page is all ours. Move it to our anon_vma so * the rmap code will not search our parent or siblings. * Protected against the rmap code by the page lock. */ page_move_anon_rmap(old_page, vma, address); |
b009c024f
|
2283 |
unlock_page(old_page); |
4e047f897
|
2284 2285 |
return wp_page_reuse(mm, vma, address, page_table, ptl, orig_pte, old_page, 0, 0); |
b009c024f
|
2286 |
} |
ab967d860
|
2287 |
unlock_page(old_page); |
ee6a64578
|
2288 |
} else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) == |
d08b3851d
|
2289 |
(VM_WRITE|VM_SHARED))) { |
93e478d4c
|
2290 2291 |
return wp_page_shared(mm, vma, address, page_table, pmd, ptl, orig_pte, old_page); |
1da177e4c
|
2292 |
} |
1da177e4c
|
2293 2294 2295 2296 |
/* * Ok, we need to copy. Oh, well.. */ |
09cbfeaf1
|
2297 |
get_page(old_page); |
287668052
|
2298 |
|
8f4e2101f
|
2299 |
pte_unmap_unlock(page_table, ptl); |
2f38ab2c3
|
2300 2301 |
return wp_page_copy(mm, vma, address, page_table, pmd, orig_pte, old_page); |
1da177e4c
|
2302 |
} |
97a894136
|
2303 |
static void unmap_mapping_range_vma(struct vm_area_struct *vma, |
1da177e4c
|
2304 2305 2306 |
unsigned long start_addr, unsigned long end_addr, struct zap_details *details) { |
f5cc4eef9
|
2307 |
zap_page_range_single(vma, start_addr, end_addr - start_addr, details); |
1da177e4c
|
2308 |
} |
6b2dbba8b
|
2309 |
static inline void unmap_mapping_range_tree(struct rb_root *root, |
1da177e4c
|
2310 2311 2312 |
struct zap_details *details) { struct vm_area_struct *vma; |
1da177e4c
|
2313 |
pgoff_t vba, vea, zba, zea; |
6b2dbba8b
|
2314 |
vma_interval_tree_foreach(vma, root, |
1da177e4c
|
2315 |
details->first_index, details->last_index) { |
1da177e4c
|
2316 2317 |
vba = vma->vm_pgoff; |
d6e932177
|
2318 |
vea = vba + vma_pages(vma) - 1; |
1da177e4c
|
2319 2320 2321 2322 2323 2324 |
zba = details->first_index; if (zba < vba) zba = vba; zea = details->last_index; if (zea > vea) zea = vea; |
97a894136
|
2325 |
unmap_mapping_range_vma(vma, |
1da177e4c
|
2326 2327 |
((zba - vba) << PAGE_SHIFT) + vma->vm_start, ((zea - vba + 1) << PAGE_SHIFT) + vma->vm_start, |
97a894136
|
2328 |
details); |
1da177e4c
|
2329 2330 |
} } |
1da177e4c
|
2331 |
/** |
8a5f14a23
|
2332 2333 2334 2335 |
* unmap_mapping_range - unmap the portion of all mmaps in the specified * address_space corresponding to the specified page range in the underlying * file. * |
3d41088fa
|
2336 |
* @mapping: the address space containing mmaps to be unmapped. |
1da177e4c
|
2337 2338 |
* @holebegin: byte in first page to unmap, relative to the start of * the underlying file. This will be rounded down to a PAGE_SIZE |
25d9e2d15
|
2339 |
* boundary. Note that this is different from truncate_pagecache(), which |
1da177e4c
|
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 |
* must keep the partial page. In contrast, we must get rid of * partial pages. * @holelen: size of prospective hole in bytes. This will be rounded * up to a PAGE_SIZE boundary. A holelen of zero truncates to the * end of the file. * @even_cows: 1 when truncating a file, unmap even private COWed pages; * but 0 when invalidating pagecache, don't throw away private data. */ void unmap_mapping_range(struct address_space *mapping, loff_t const holebegin, loff_t const holelen, int even_cows) { |
aac453635
|
2351 |
struct zap_details details = { }; |
1da177e4c
|
2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 |
pgoff_t hba = holebegin >> PAGE_SHIFT; pgoff_t hlen = (holelen + PAGE_SIZE - 1) >> PAGE_SHIFT; /* Check for overflow. */ if (sizeof(holelen) > sizeof(hlen)) { long long holeend = (holebegin + holelen + PAGE_SIZE - 1) >> PAGE_SHIFT; if (holeend & ~(long long)ULONG_MAX) hlen = ULONG_MAX - hba + 1; } details.check_mapping = even_cows? NULL: mapping; |
1da177e4c
|
2364 2365 2366 2367 |
details.first_index = hba; details.last_index = hba + hlen - 1; if (details.last_index < details.first_index) details.last_index = ULONG_MAX; |
1da177e4c
|
2368 |
|
0f90cc660
|
2369 2370 |
/* DAX uses i_mmap_lock to serialise file truncate vs page fault */ |
46c043ede
|
2371 |
i_mmap_lock_write(mapping); |
6b2dbba8b
|
2372 |
if (unlikely(!RB_EMPTY_ROOT(&mapping->i_mmap))) |
1da177e4c
|
2373 |
unmap_mapping_range_tree(&mapping->i_mmap, &details); |
46c043ede
|
2374 |
i_mmap_unlock_write(mapping); |
1da177e4c
|
2375 2376 |
} EXPORT_SYMBOL(unmap_mapping_range); |
1da177e4c
|
2377 |
/* |
8f4e2101f
|
2378 2379 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. |
9a95f3cf7
|
2380 2381 2382 2383 |
* We return with pte unmapped and unlocked. * * We return with the mmap_sem locked or unlocked in the same cases * as does filemap_fault(). |
1da177e4c
|
2384 |
*/ |
65500d234
|
2385 2386 |
static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f
|
2387 |
unsigned int flags, pte_t orig_pte) |
1da177e4c
|
2388 |
{ |
8f4e2101f
|
2389 |
spinlock_t *ptl; |
56f31801c
|
2390 |
struct page *page, *swapcache; |
00501b531
|
2391 |
struct mem_cgroup *memcg; |
65500d234
|
2392 |
swp_entry_t entry; |
1da177e4c
|
2393 |
pte_t pte; |
d065bd810
|
2394 |
int locked; |
ad8c2ee80
|
2395 |
int exclusive = 0; |
83c54070e
|
2396 |
int ret = 0; |
1da177e4c
|
2397 |
|
4c21e2f24
|
2398 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
8f4e2101f
|
2399 |
goto out; |
65500d234
|
2400 2401 |
entry = pte_to_swp_entry(orig_pte); |
d1737fdbe
|
2402 2403 2404 2405 2406 2407 2408 |
if (unlikely(non_swap_entry(entry))) { if (is_migration_entry(entry)) { migration_entry_wait(mm, pmd, address); } else if (is_hwpoison_entry(entry)) { ret = VM_FAULT_HWPOISON; } else { print_bad_pte(vma, address, orig_pte, NULL); |
d99be1a8e
|
2409 |
ret = VM_FAULT_SIGBUS; |
d1737fdbe
|
2410 |
} |
0697212a4
|
2411 2412 |
goto out; } |
0ff922452
|
2413 |
delayacct_set_flag(DELAYACCT_PF_SWAPIN); |
1da177e4c
|
2414 2415 |
page = lookup_swap_cache(entry); if (!page) { |
02098feaa
|
2416 2417 |
page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, vma, address); |
1da177e4c
|
2418 2419 |
if (!page) { /* |
8f4e2101f
|
2420 2421 |
* Back out if somebody else faulted in this pte * while we released the pte lock. |
1da177e4c
|
2422 |
*/ |
8f4e2101f
|
2423 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c
|
2424 2425 |
if (likely(pte_same(*page_table, orig_pte))) ret = VM_FAULT_OOM; |
0ff922452
|
2426 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
65500d234
|
2427 |
goto unlock; |
1da177e4c
|
2428 2429 2430 2431 |
} /* Had to read the page from swap area: Major fault */ ret = VM_FAULT_MAJOR; |
f8891e5e1
|
2432 |
count_vm_event(PGMAJFAULT); |
456f998ec
|
2433 |
mem_cgroup_count_vm_event(mm, PGMAJFAULT); |
d1737fdbe
|
2434 |
} else if (PageHWPoison(page)) { |
71f72525d
|
2435 2436 2437 2438 |
/* * hwpoisoned dirty swapcache pages are kept for killing * owner processes (which may be unknown at hwpoison time) */ |
d1737fdbe
|
2439 2440 |
ret = VM_FAULT_HWPOISON; delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
56f31801c
|
2441 |
swapcache = page; |
4779cb31c
|
2442 |
goto out_release; |
1da177e4c
|
2443 |
} |
56f31801c
|
2444 |
swapcache = page; |
d065bd810
|
2445 |
locked = lock_page_or_retry(page, mm, flags); |
e709ffd61
|
2446 |
|
073e587ec
|
2447 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
d065bd810
|
2448 2449 2450 2451 |
if (!locked) { ret |= VM_FAULT_RETRY; goto out_release; } |
073e587ec
|
2452 |
|
4969c1192
|
2453 |
/* |
31c4a3d3a
|
2454 2455 2456 2457 |
* Make sure try_to_free_swap or reuse_swap_page or swapoff did not * release the swapcache from under us. The page pin, and pte_same * test below, are not enough to exclude that. Even if it is still * swapcache, we need to check that the page's swap has not changed. |
4969c1192
|
2458 |
*/ |
31c4a3d3a
|
2459 |
if (unlikely(!PageSwapCache(page) || page_private(page) != entry.val)) |
4969c1192
|
2460 |
goto out_page; |
cbf86cfe0
|
2461 2462 2463 2464 |
page = ksm_might_need_to_copy(page, vma, address); if (unlikely(!page)) { ret = VM_FAULT_OOM; page = swapcache; |
cbf86cfe0
|
2465 |
goto out_page; |
5ad646880
|
2466 |
} |
f627c2f53
|
2467 |
if (mem_cgroup_try_charge(page, mm, GFP_KERNEL, &memcg, false)) { |
8a9f3ccd2
|
2468 |
ret = VM_FAULT_OOM; |
bc43f75cd
|
2469 |
goto out_page; |
8a9f3ccd2
|
2470 |
} |
1da177e4c
|
2471 |
/* |
8f4e2101f
|
2472 |
* Back out if somebody else already faulted in this pte. |
1da177e4c
|
2473 |
*/ |
8f4e2101f
|
2474 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
9e9bef07c
|
2475 |
if (unlikely(!pte_same(*page_table, orig_pte))) |
b81074800
|
2476 |
goto out_nomap; |
b81074800
|
2477 2478 2479 2480 |
if (unlikely(!PageUptodate(page))) { ret = VM_FAULT_SIGBUS; goto out_nomap; |
1da177e4c
|
2481 |
} |
8c7c6e34a
|
2482 2483 2484 2485 2486 2487 2488 2489 |
/* * The page isn't present yet, go ahead with the fault. * * Be careful about the sequence of operations here. * To get its accounting right, reuse_swap_page() must be called * while the page is counted on swap but not yet in mapcount i.e. * before page_add_anon_rmap() and swap_free(); try_to_free_swap() * must be called after the swap_free(), or it will never succeed. |
8c7c6e34a
|
2490 |
*/ |
1da177e4c
|
2491 |
|
34e55232e
|
2492 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
b084d4353
|
2493 |
dec_mm_counter_fast(mm, MM_SWAPENTS); |
1da177e4c
|
2494 |
pte = mk_pte(page, vma->vm_page_prot); |
30c9f3a9f
|
2495 |
if ((flags & FAULT_FLAG_WRITE) && reuse_swap_page(page)) { |
1da177e4c
|
2496 |
pte = maybe_mkwrite(pte_mkdirty(pte), vma); |
30c9f3a9f
|
2497 |
flags &= ~FAULT_FLAG_WRITE; |
9a5b489b8
|
2498 |
ret |= VM_FAULT_WRITE; |
d281ee614
|
2499 |
exclusive = RMAP_EXCLUSIVE; |
1da177e4c
|
2500 |
} |
1da177e4c
|
2501 |
flush_icache_page(vma, page); |
179ef71cb
|
2502 2503 |
if (pte_swp_soft_dirty(orig_pte)) pte = pte_mksoft_dirty(pte); |
1da177e4c
|
2504 |
set_pte_at(mm, address, page_table, pte); |
00501b531
|
2505 |
if (page == swapcache) { |
af34770e5
|
2506 |
do_page_add_anon_rmap(page, vma, address, exclusive); |
f627c2f53
|
2507 |
mem_cgroup_commit_charge(page, memcg, true, false); |
00501b531
|
2508 |
} else { /* ksm created a completely new copy */ |
d281ee614
|
2509 |
page_add_new_anon_rmap(page, vma, address, false); |
f627c2f53
|
2510 |
mem_cgroup_commit_charge(page, memcg, false, false); |
00501b531
|
2511 2512 |
lru_cache_add_active_or_unevictable(page, vma); } |
1da177e4c
|
2513 |
|
c475a8ab6
|
2514 |
swap_free(entry); |
5ccc5abaa
|
2515 2516 |
if (mem_cgroup_swap_full(page) || (vma->vm_flags & VM_LOCKED) || PageMlocked(page)) |
a2c43eed8
|
2517 |
try_to_free_swap(page); |
c475a8ab6
|
2518 |
unlock_page(page); |
56f31801c
|
2519 |
if (page != swapcache) { |
4969c1192
|
2520 2521 2522 2523 2524 2525 2526 2527 2528 |
/* * Hold the lock to avoid the swap entry to be reused * until we take the PT lock for the pte_same() check * (to avoid false positives from pte_same). For * further safety release the lock after the swap_free * so that the swap count won't change under a * parallel locked swapcache. */ unlock_page(swapcache); |
09cbfeaf1
|
2529 |
put_page(swapcache); |
4969c1192
|
2530 |
} |
c475a8ab6
|
2531 |
|
30c9f3a9f
|
2532 |
if (flags & FAULT_FLAG_WRITE) { |
61469f1d5
|
2533 2534 2535 |
ret |= do_wp_page(mm, vma, address, page_table, pmd, ptl, pte); if (ret & VM_FAULT_ERROR) ret &= VM_FAULT_ERROR; |
1da177e4c
|
2536 2537 2538 2539 |
goto out; } /* No need to invalidate - it was non-present before */ |
4b3073e1c
|
2540 |
update_mmu_cache(vma, address, page_table); |
65500d234
|
2541 |
unlock: |
8f4e2101f
|
2542 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c
|
2543 2544 |
out: return ret; |
b81074800
|
2545 |
out_nomap: |
f627c2f53
|
2546 |
mem_cgroup_cancel_charge(page, memcg, false); |
8f4e2101f
|
2547 |
pte_unmap_unlock(page_table, ptl); |
bc43f75cd
|
2548 |
out_page: |
b81074800
|
2549 |
unlock_page(page); |
4779cb31c
|
2550 |
out_release: |
09cbfeaf1
|
2551 |
put_page(page); |
56f31801c
|
2552 |
if (page != swapcache) { |
4969c1192
|
2553 |
unlock_page(swapcache); |
09cbfeaf1
|
2554 |
put_page(swapcache); |
4969c1192
|
2555 |
} |
65500d234
|
2556 |
return ret; |
1da177e4c
|
2557 2558 2559 |
} /* |
8ca3eb080
|
2560 2561 |
* This is like a special single-page "expand_{down|up}wards()", * except we must first make sure that 'address{-|+}PAGE_SIZE' |
320b2b8de
|
2562 |
* doesn't hit another vma. |
320b2b8de
|
2563 2564 2565 2566 2567 |
*/ static inline int check_stack_guard_page(struct vm_area_struct *vma, unsigned long address) { address &= PAGE_MASK; if ((vma->vm_flags & VM_GROWSDOWN) && address == vma->vm_start) { |
0e8e50e20
|
2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 |
struct vm_area_struct *prev = vma->vm_prev; /* * Is there a mapping abutting this one below? * * That's only ok if it's the same stack mapping * that has gotten split.. */ if (prev && prev->vm_end == address) return prev->vm_flags & VM_GROWSDOWN ? 0 : -ENOMEM; |
320b2b8de
|
2578 |
|
fee7e49d4
|
2579 |
return expand_downwards(vma, address - PAGE_SIZE); |
320b2b8de
|
2580 |
} |
8ca3eb080
|
2581 2582 2583 2584 2585 2586 |
if ((vma->vm_flags & VM_GROWSUP) && address + PAGE_SIZE == vma->vm_end) { struct vm_area_struct *next = vma->vm_next; /* As VM_GROWSDOWN but s/below/above/ */ if (next && next->vm_start == address + PAGE_SIZE) return next->vm_flags & VM_GROWSUP ? 0 : -ENOMEM; |
fee7e49d4
|
2587 |
return expand_upwards(vma, address + PAGE_SIZE); |
8ca3eb080
|
2588 |
} |
320b2b8de
|
2589 2590 2591 2592 |
return 0; } /* |
8f4e2101f
|
2593 2594 2595 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. * We return with mmap_sem still held, but pte unmapped and unlocked. |
1da177e4c
|
2596 |
*/ |
65500d234
|
2597 2598 |
static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f
|
2599 |
unsigned int flags) |
1da177e4c
|
2600 |
{ |
00501b531
|
2601 |
struct mem_cgroup *memcg; |
8f4e2101f
|
2602 2603 |
struct page *page; spinlock_t *ptl; |
1da177e4c
|
2604 |
pte_t entry; |
1da177e4c
|
2605 |
|
11ac55247
|
2606 |
pte_unmap(page_table); |
6b7339f4c
|
2607 2608 2609 |
/* File mapping without ->vm_ops ? */ if (vma->vm_flags & VM_SHARED) return VM_FAULT_SIGBUS; |
11ac55247
|
2610 2611 |
/* Check if we need to add a guard page to the stack */ if (check_stack_guard_page(vma, address) < 0) |
9c145c56d
|
2612 |
return VM_FAULT_SIGSEGV; |
320b2b8de
|
2613 |
|
11ac55247
|
2614 |
/* Use the zero-page for reads */ |
593befa6a
|
2615 |
if (!(flags & FAULT_FLAG_WRITE) && !mm_forbids_zeropage(mm)) { |
62eede62d
|
2616 2617 |
entry = pte_mkspecial(pfn_pte(my_zero_pfn(address), vma->vm_page_prot)); |
11ac55247
|
2618 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
a13ea5b75
|
2619 2620 |
if (!pte_none(*page_table)) goto unlock; |
6b251fc96
|
2621 2622 2623 2624 2625 2626 |
/* Deliver the page fault to userland, check inside PT lock */ if (userfaultfd_missing(vma)) { pte_unmap_unlock(page_table, ptl); return handle_userfault(vma, address, flags, VM_UFFD_MISSING); } |
a13ea5b75
|
2627 2628 |
goto setpte; } |
557ed1fa2
|
2629 |
/* Allocate our own private page. */ |
557ed1fa2
|
2630 2631 2632 2633 2634 |
if (unlikely(anon_vma_prepare(vma))) goto oom; page = alloc_zeroed_user_highpage_movable(vma, address); if (!page) goto oom; |
eb3c24f30
|
2635 |
|
f627c2f53
|
2636 |
if (mem_cgroup_try_charge(page, mm, GFP_KERNEL, &memcg, false)) |
eb3c24f30
|
2637 |
goto oom_free_page; |
52f37629f
|
2638 2639 2640 2641 2642 |
/* * The memory barrier inside __SetPageUptodate makes sure that * preceeding stores to the page contents become visible before * the set_pte_at() write. */ |
0ed361dec
|
2643 |
__SetPageUptodate(page); |
8f4e2101f
|
2644 |
|
557ed1fa2
|
2645 |
entry = mk_pte(page, vma->vm_page_prot); |
1ac0cb5d0
|
2646 2647 |
if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); |
1da177e4c
|
2648 |
|
557ed1fa2
|
2649 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1c2fb7a4c
|
2650 |
if (!pte_none(*page_table)) |
557ed1fa2
|
2651 |
goto release; |
9ba692948
|
2652 |
|
6b251fc96
|
2653 2654 2655 |
/* Deliver the page fault to userland, check inside PT lock */ if (userfaultfd_missing(vma)) { pte_unmap_unlock(page_table, ptl); |
f627c2f53
|
2656 |
mem_cgroup_cancel_charge(page, memcg, false); |
09cbfeaf1
|
2657 |
put_page(page); |
6b251fc96
|
2658 2659 2660 |
return handle_userfault(vma, address, flags, VM_UFFD_MISSING); } |
34e55232e
|
2661 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
d281ee614
|
2662 |
page_add_new_anon_rmap(page, vma, address, false); |
f627c2f53
|
2663 |
mem_cgroup_commit_charge(page, memcg, false, false); |
00501b531
|
2664 |
lru_cache_add_active_or_unevictable(page, vma); |
a13ea5b75
|
2665 |
setpte: |
65500d234
|
2666 |
set_pte_at(mm, address, page_table, entry); |
1da177e4c
|
2667 2668 |
/* No need to invalidate - it was non-present before */ |
4b3073e1c
|
2669 |
update_mmu_cache(vma, address, page_table); |
65500d234
|
2670 |
unlock: |
8f4e2101f
|
2671 |
pte_unmap_unlock(page_table, ptl); |
83c54070e
|
2672 |
return 0; |
8f4e2101f
|
2673 |
release: |
f627c2f53
|
2674 |
mem_cgroup_cancel_charge(page, memcg, false); |
09cbfeaf1
|
2675 |
put_page(page); |
8f4e2101f
|
2676 |
goto unlock; |
8a9f3ccd2
|
2677 |
oom_free_page: |
09cbfeaf1
|
2678 |
put_page(page); |
65500d234
|
2679 |
oom: |
1da177e4c
|
2680 2681 |
return VM_FAULT_OOM; } |
9a95f3cf7
|
2682 2683 2684 2685 2686 |
/* * The mmap_sem must have been held on entry, and may have been * released depending on flags and vma->vm_ops->fault() return value. * See filemap_fault() and __lock_page_retry(). */ |
7eae74af3
|
2687 |
static int __do_fault(struct vm_area_struct *vma, unsigned long address, |
2e4cdab05
|
2688 2689 |
pgoff_t pgoff, unsigned int flags, struct page *cow_page, struct page **page) |
7eae74af3
|
2690 2691 2692 2693 2694 2695 2696 2697 |
{ struct vm_fault vmf; int ret; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = pgoff; vmf.flags = flags; vmf.page = NULL; |
c20cd45eb
|
2698 |
vmf.gfp_mask = __get_fault_gfp_mask(vma); |
2e4cdab05
|
2699 |
vmf.cow_page = cow_page; |
7eae74af3
|
2700 2701 2702 2703 |
ret = vma->vm_ops->fault(vma, &vmf); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) return ret; |
2e4cdab05
|
2704 2705 |
if (!vmf.page) goto out; |
7eae74af3
|
2706 2707 2708 2709 |
if (unlikely(PageHWPoison(vmf.page))) { if (ret & VM_FAULT_LOCKED) unlock_page(vmf.page); |
09cbfeaf1
|
2710 |
put_page(vmf.page); |
7eae74af3
|
2711 2712 2713 2714 2715 2716 2717 |
return VM_FAULT_HWPOISON; } if (unlikely(!(ret & VM_FAULT_LOCKED))) lock_page(vmf.page); else VM_BUG_ON_PAGE(!PageLocked(vmf.page), vmf.page); |
2e4cdab05
|
2718 |
out: |
7eae74af3
|
2719 2720 2721 |
*page = vmf.page; return ret; } |
8c6e50b02
|
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 |
/** * do_set_pte - setup new PTE entry for given page and add reverse page mapping. * * @vma: virtual memory area * @address: user virtual address * @page: page to map * @pte: pointer to target page table entry * @write: true, if new entry is writable * @anon: true, if it's anonymous page * * Caller must hold page table lock relevant for @pte. * * Target users are page handler itself and implementations of * vm_ops->map_pages. */ void do_set_pte(struct vm_area_struct *vma, unsigned long address, |
3bb977946
|
2738 2739 2740 2741 2742 2743 2744 2745 |
struct page *page, pte_t *pte, bool write, bool anon) { pte_t entry; flush_icache_page(vma, page); entry = mk_pte(page, vma->vm_page_prot); if (write) entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
3bb977946
|
2746 2747 |
if (anon) { inc_mm_counter_fast(vma->vm_mm, MM_ANONPAGES); |
d281ee614
|
2748 |
page_add_new_anon_rmap(page, vma, address, false); |
3bb977946
|
2749 |
} else { |
eca56ff90
|
2750 |
inc_mm_counter_fast(vma->vm_mm, mm_counter_file(page)); |
3bb977946
|
2751 2752 2753 2754 2755 2756 2757 |
page_add_file_rmap(page); } set_pte_at(vma->vm_mm, address, pte, entry); /* no need to invalidate: a not-present page won't be cached */ update_mmu_cache(vma, address, pte); } |
3a91053ae
|
2758 2759 |
static unsigned long fault_around_bytes __read_mostly = rounddown_pow_of_two(65536); |
a9b0f8618
|
2760 |
|
a9b0f8618
|
2761 2762 |
#ifdef CONFIG_DEBUG_FS static int fault_around_bytes_get(void *data, u64 *val) |
1592eef01
|
2763 |
{ |
a9b0f8618
|
2764 |
*val = fault_around_bytes; |
1592eef01
|
2765 2766 |
return 0; } |
b4903d6e8
|
2767 2768 2769 2770 2771 |
/* * fault_around_pages() and fault_around_mask() expects fault_around_bytes * rounded down to nearest page order. It's what do_fault_around() expects to * see. */ |
a9b0f8618
|
2772 |
static int fault_around_bytes_set(void *data, u64 val) |
1592eef01
|
2773 |
{ |
a9b0f8618
|
2774 |
if (val / PAGE_SIZE > PTRS_PER_PTE) |
1592eef01
|
2775 |
return -EINVAL; |
b4903d6e8
|
2776 2777 2778 2779 |
if (val > PAGE_SIZE) fault_around_bytes = rounddown_pow_of_two(val); else fault_around_bytes = PAGE_SIZE; /* rounddown_pow_of_two(0) is undefined */ |
1592eef01
|
2780 2781 |
return 0; } |
a9b0f8618
|
2782 2783 2784 |
DEFINE_SIMPLE_ATTRIBUTE(fault_around_bytes_fops, fault_around_bytes_get, fault_around_bytes_set, "%llu "); |
1592eef01
|
2785 2786 2787 2788 |
static int __init fault_around_debugfs(void) { void *ret; |
a9b0f8618
|
2789 2790 |
ret = debugfs_create_file("fault_around_bytes", 0644, NULL, NULL, &fault_around_bytes_fops); |
1592eef01
|
2791 |
if (!ret) |
a9b0f8618
|
2792 |
pr_warn("Failed to create fault_around_bytes in debugfs"); |
1592eef01
|
2793 2794 2795 |
return 0; } late_initcall(fault_around_debugfs); |
1592eef01
|
2796 |
#endif |
8c6e50b02
|
2797 |
|
1fdb412bd
|
2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 |
/* * do_fault_around() tries to map few pages around the fault address. The hope * is that the pages will be needed soon and this will lower the number of * faults to handle. * * It uses vm_ops->map_pages() to map the pages, which skips the page if it's * not ready to be mapped: not up-to-date, locked, etc. * * This function is called with the page table lock taken. In the split ptlock * case the page table lock only protects only those entries which belong to * the page table corresponding to the fault address. * * This function doesn't cross the VMA boundaries, in order to call map_pages() * only once. * * fault_around_pages() defines how many pages we'll try to map. * do_fault_around() expects it to return a power of two less than or equal to * PTRS_PER_PTE. * * The virtual address of the area that we map is naturally aligned to the * fault_around_pages() value (and therefore to page order). This way it's * easier to guarantee that we don't cross page table boundaries. */ |
8c6e50b02
|
2821 2822 2823 |
static void do_fault_around(struct vm_area_struct *vma, unsigned long address, pte_t *pte, pgoff_t pgoff, unsigned int flags) { |
aecd6f442
|
2824 |
unsigned long start_addr, nr_pages, mask; |
8c6e50b02
|
2825 2826 2827 |
pgoff_t max_pgoff; struct vm_fault vmf; int off; |
4db0c3c29
|
2828 |
nr_pages = READ_ONCE(fault_around_bytes) >> PAGE_SHIFT; |
aecd6f442
|
2829 2830 2831 |
mask = ~(nr_pages * PAGE_SIZE - 1) & PAGE_MASK; start_addr = max(address & mask, vma->vm_start); |
8c6e50b02
|
2832 2833 2834 2835 2836 2837 |
off = ((address - start_addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1); pte -= off; pgoff -= off; /* * max_pgoff is either end of page table or end of vma |
850e9c69c
|
2838 |
* or fault_around_pages() from pgoff, depending what is nearest. |
8c6e50b02
|
2839 2840 2841 2842 |
*/ max_pgoff = pgoff - ((start_addr >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) + PTRS_PER_PTE - 1; max_pgoff = min3(max_pgoff, vma_pages(vma) + vma->vm_pgoff - 1, |
aecd6f442
|
2843 |
pgoff + nr_pages - 1); |
8c6e50b02
|
2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 |
/* Check if it makes any sense to call ->map_pages */ while (!pte_none(*pte)) { if (++pgoff > max_pgoff) return; start_addr += PAGE_SIZE; if (start_addr >= vma->vm_end) return; pte++; } vmf.virtual_address = (void __user *) start_addr; vmf.pte = pte; vmf.pgoff = pgoff; vmf.max_pgoff = max_pgoff; vmf.flags = flags; |
c20cd45eb
|
2860 |
vmf.gfp_mask = __get_fault_gfp_mask(vma); |
8c6e50b02
|
2861 2862 |
vma->vm_ops->map_pages(vma, &vmf); } |
e655fb290
|
2863 2864 2865 2866 2867 2868 |
static int do_read_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, pgoff_t pgoff, unsigned int flags, pte_t orig_pte) { struct page *fault_page; spinlock_t *ptl; |
3bb977946
|
2869 |
pte_t *pte; |
8c6e50b02
|
2870 2871 2872 2873 2874 2875 2876 |
int ret = 0; /* * Let's call ->map_pages() first and use ->fault() as fallback * if page by the offset is not ready to be mapped (cold cache or * something). */ |
9b4bdd2ff
|
2877 |
if (vma->vm_ops->map_pages && fault_around_bytes >> PAGE_SHIFT > 1) { |
8c6e50b02
|
2878 2879 2880 2881 2882 2883 |
pte = pte_offset_map_lock(mm, pmd, address, &ptl); do_fault_around(vma, address, pte, pgoff, flags); if (!pte_same(*pte, orig_pte)) goto unlock_out; pte_unmap_unlock(pte, ptl); } |
e655fb290
|
2884 |
|
2e4cdab05
|
2885 |
ret = __do_fault(vma, address, pgoff, flags, NULL, &fault_page); |
e655fb290
|
2886 2887 2888 2889 2890 2891 2892 |
if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) return ret; pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); unlock_page(fault_page); |
09cbfeaf1
|
2893 |
put_page(fault_page); |
e655fb290
|
2894 2895 |
return ret; } |
3bb977946
|
2896 |
do_set_pte(vma, address, fault_page, pte, false, false); |
e655fb290
|
2897 |
unlock_page(fault_page); |
8c6e50b02
|
2898 2899 |
unlock_out: pte_unmap_unlock(pte, ptl); |
e655fb290
|
2900 2901 |
return ret; } |
ec47c3b95
|
2902 2903 2904 2905 2906 |
static int do_cow_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, pgoff_t pgoff, unsigned int flags, pte_t orig_pte) { struct page *fault_page, *new_page; |
00501b531
|
2907 |
struct mem_cgroup *memcg; |
ec47c3b95
|
2908 |
spinlock_t *ptl; |
3bb977946
|
2909 |
pte_t *pte; |
ec47c3b95
|
2910 2911 2912 2913 2914 2915 2916 2917 |
int ret; if (unlikely(anon_vma_prepare(vma))) return VM_FAULT_OOM; new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); if (!new_page) return VM_FAULT_OOM; |
f627c2f53
|
2918 |
if (mem_cgroup_try_charge(new_page, mm, GFP_KERNEL, &memcg, false)) { |
09cbfeaf1
|
2919 |
put_page(new_page); |
ec47c3b95
|
2920 2921 |
return VM_FAULT_OOM; } |
2e4cdab05
|
2922 |
ret = __do_fault(vma, address, pgoff, flags, new_page, &fault_page); |
ec47c3b95
|
2923 2924 |
if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) goto uncharge_out; |
2e4cdab05
|
2925 2926 |
if (fault_page) copy_user_highpage(new_page, fault_page, address, vma); |
ec47c3b95
|
2927 2928 2929 2930 2931 |
__SetPageUptodate(new_page); pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); |
2e4cdab05
|
2932 2933 |
if (fault_page) { unlock_page(fault_page); |
09cbfeaf1
|
2934 |
put_page(fault_page); |
2e4cdab05
|
2935 2936 2937 |
} else { /* * The fault handler has no page to lock, so it holds |
0df9d41ab
|
2938 |
* i_mmap_lock for read to protect against truncate. |
2e4cdab05
|
2939 |
*/ |
0df9d41ab
|
2940 |
i_mmap_unlock_read(vma->vm_file->f_mapping); |
2e4cdab05
|
2941 |
} |
ec47c3b95
|
2942 2943 |
goto uncharge_out; } |
3bb977946
|
2944 |
do_set_pte(vma, address, new_page, pte, true, true); |
f627c2f53
|
2945 |
mem_cgroup_commit_charge(new_page, memcg, false, false); |
00501b531
|
2946 |
lru_cache_add_active_or_unevictable(new_page, vma); |
ec47c3b95
|
2947 |
pte_unmap_unlock(pte, ptl); |
2e4cdab05
|
2948 2949 |
if (fault_page) { unlock_page(fault_page); |
09cbfeaf1
|
2950 |
put_page(fault_page); |
2e4cdab05
|
2951 2952 2953 |
} else { /* * The fault handler has no page to lock, so it holds |
0df9d41ab
|
2954 |
* i_mmap_lock for read to protect against truncate. |
2e4cdab05
|
2955 |
*/ |
0df9d41ab
|
2956 |
i_mmap_unlock_read(vma->vm_file->f_mapping); |
2e4cdab05
|
2957 |
} |
ec47c3b95
|
2958 2959 |
return ret; uncharge_out: |
f627c2f53
|
2960 |
mem_cgroup_cancel_charge(new_page, memcg, false); |
09cbfeaf1
|
2961 |
put_page(new_page); |
ec47c3b95
|
2962 2963 |
return ret; } |
f0c6d4d29
|
2964 |
static int do_shared_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
16abfa086
|
2965 |
unsigned long address, pmd_t *pmd, |
54cb8821d
|
2966 |
pgoff_t pgoff, unsigned int flags, pte_t orig_pte) |
1da177e4c
|
2967 |
{ |
f0c6d4d29
|
2968 2969 |
struct page *fault_page; struct address_space *mapping; |
8f4e2101f
|
2970 |
spinlock_t *ptl; |
3bb977946
|
2971 |
pte_t *pte; |
f0c6d4d29
|
2972 |
int dirtied = 0; |
f0c6d4d29
|
2973 |
int ret, tmp; |
1d65f86db
|
2974 |
|
2e4cdab05
|
2975 |
ret = __do_fault(vma, address, pgoff, flags, NULL, &fault_page); |
7eae74af3
|
2976 |
if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) |
f0c6d4d29
|
2977 |
return ret; |
1da177e4c
|
2978 2979 |
/* |
f0c6d4d29
|
2980 2981 |
* Check if the backing address space wants to know that the page is * about to become writable |
1da177e4c
|
2982 |
*/ |
fb09a4642
|
2983 2984 2985 2986 2987 |
if (vma->vm_ops->page_mkwrite) { unlock_page(fault_page); tmp = do_page_mkwrite(vma, fault_page, address); if (unlikely(!tmp || (tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) { |
09cbfeaf1
|
2988 |
put_page(fault_page); |
fb09a4642
|
2989 |
return tmp; |
4294621f4
|
2990 |
} |
fb09a4642
|
2991 |
} |
f0c6d4d29
|
2992 2993 2994 2995 |
pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); unlock_page(fault_page); |
09cbfeaf1
|
2996 |
put_page(fault_page); |
f0c6d4d29
|
2997 |
return ret; |
1da177e4c
|
2998 |
} |
3bb977946
|
2999 |
do_set_pte(vma, address, fault_page, pte, true, false); |
f0c6d4d29
|
3000 |
pte_unmap_unlock(pte, ptl); |
b827e496c
|
3001 |
|
f0c6d4d29
|
3002 3003 |
if (set_page_dirty(fault_page)) dirtied = 1; |
d82fa87d2
|
3004 3005 3006 3007 3008 3009 |
/* * Take a local copy of the address_space - page.mapping may be zeroed * by truncate after unlock_page(). The address_space itself remains * pinned by vma->vm_file's reference. We rely on unlock_page()'s * release semantics to prevent the compiler from undoing this copying. */ |
1c290f642
|
3010 |
mapping = page_rmapping(fault_page); |
f0c6d4d29
|
3011 3012 3013 3014 3015 3016 3017 |
unlock_page(fault_page); if ((dirtied || vma->vm_ops->page_mkwrite) && mapping) { /* * Some device drivers do not set page.mapping but still * dirty their pages */ balance_dirty_pages_ratelimited(mapping); |
d08b3851d
|
3018 |
} |
d00806b18
|
3019 |
|
74ec67511
|
3020 |
if (!vma->vm_ops->page_mkwrite) |
f0c6d4d29
|
3021 |
file_update_time(vma->vm_file); |
b827e496c
|
3022 |
|
1d65f86db
|
3023 |
return ret; |
54cb8821d
|
3024 |
} |
d00806b18
|
3025 |
|
9a95f3cf7
|
3026 3027 3028 3029 3030 3031 |
/* * We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults). * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). */ |
9b4bdd2ff
|
3032 |
static int do_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
54cb8821d
|
3033 |
unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f
|
3034 |
unsigned int flags, pte_t orig_pte) |
54cb8821d
|
3035 |
{ |
88193f7ce
|
3036 |
pgoff_t pgoff = linear_page_index(vma, address); |
54cb8821d
|
3037 |
|
16abfa086
|
3038 |
pte_unmap(page_table); |
6b7339f4c
|
3039 3040 3041 |
/* The VMA was not fully populated on mmap() or missing VM_DONTEXPAND */ if (!vma->vm_ops->fault) return VM_FAULT_SIGBUS; |
e655fb290
|
3042 3043 3044 |
if (!(flags & FAULT_FLAG_WRITE)) return do_read_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
ec47c3b95
|
3045 3046 3047 |
if (!(vma->vm_flags & VM_SHARED)) return do_cow_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
f0c6d4d29
|
3048 |
return do_shared_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
54cb8821d
|
3049 |
} |
b19a99392
|
3050 |
static int numa_migrate_prep(struct page *page, struct vm_area_struct *vma, |
04bb2f947
|
3051 3052 |
unsigned long addr, int page_nid, int *flags) |
9532fec11
|
3053 3054 3055 3056 |
{ get_page(page); count_vm_numa_event(NUMA_HINT_FAULTS); |
04bb2f947
|
3057 |
if (page_nid == numa_node_id()) { |
9532fec11
|
3058 |
count_vm_numa_event(NUMA_HINT_FAULTS_LOCAL); |
04bb2f947
|
3059 3060 |
*flags |= TNF_FAULT_LOCAL; } |
9532fec11
|
3061 3062 3063 |
return mpol_misplaced(page, vma, addr); } |
b19a99392
|
3064 |
static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, |
d10e63f29
|
3065 3066 |
unsigned long addr, pte_t pte, pte_t *ptep, pmd_t *pmd) { |
4daae3b4b
|
3067 |
struct page *page = NULL; |
d10e63f29
|
3068 |
spinlock_t *ptl; |
8191acbd3
|
3069 |
int page_nid = -1; |
90572890d
|
3070 |
int last_cpupid; |
cbee9f88e
|
3071 |
int target_nid; |
b8593bfda
|
3072 |
bool migrated = false; |
b191f9b10
|
3073 |
bool was_writable = pte_write(pte); |
6688cc054
|
3074 |
int flags = 0; |
d10e63f29
|
3075 |
|
c0e7cad9f
|
3076 3077 |
/* A PROT_NONE fault should not end up here */ BUG_ON(!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE))); |
d10e63f29
|
3078 3079 3080 3081 3082 |
/* * The "pte" at this point cannot be used safely without * validation through pte_unmap_same(). It's of NUMA type but * the pfn may be screwed if the read is non atomic. * |
4d9424669
|
3083 3084 3085 |
* We can safely just do a "set_pte_at()", because the old * page table entry is not accessible, so there would be no * concurrent hardware modifications to the PTE. |
d10e63f29
|
3086 3087 3088 |
*/ ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
4daae3b4b
|
3089 3090 3091 3092 |
if (unlikely(!pte_same(*ptep, pte))) { pte_unmap_unlock(ptep, ptl); goto out; } |
4d9424669
|
3093 3094 3095 |
/* Make it present again */ pte = pte_modify(pte, vma->vm_page_prot); pte = pte_mkyoung(pte); |
b191f9b10
|
3096 3097 |
if (was_writable) pte = pte_mkwrite(pte); |
d10e63f29
|
3098 3099 3100 3101 3102 3103 3104 3105 |
set_pte_at(mm, addr, ptep, pte); update_mmu_cache(vma, addr, ptep); page = vm_normal_page(vma, addr, pte); if (!page) { pte_unmap_unlock(ptep, ptl); return 0; } |
e81c48024
|
3106 3107 3108 3109 3110 |
/* TODO: handle PTE-mapped THP */ if (PageCompound(page)) { pte_unmap_unlock(ptep, ptl); return 0; } |
6688cc054
|
3111 |
/* |
bea66fbd1
|
3112 3113 3114 3115 3116 3117 |
* Avoid grouping on RO pages in general. RO pages shouldn't hurt as * much anyway since they can be in shared cache state. This misses * the case where a mapping is writable but the process never writes * to it but pte_write gets cleared during protection updates and * pte_dirty has unpredictable behaviour between PTE scan updates, * background writeback, dirty balancing and application behaviour. |
6688cc054
|
3118 |
*/ |
bea66fbd1
|
3119 |
if (!(vma->vm_flags & VM_WRITE)) |
6688cc054
|
3120 |
flags |= TNF_NO_GROUP; |
dabe1d992
|
3121 3122 3123 3124 3125 3126 |
/* * Flag if the page is shared between multiple address spaces. This * is later used when determining whether to group tasks together */ if (page_mapcount(page) > 1 && (vma->vm_flags & VM_SHARED)) flags |= TNF_SHARED; |
90572890d
|
3127 |
last_cpupid = page_cpupid_last(page); |
8191acbd3
|
3128 |
page_nid = page_to_nid(page); |
04bb2f947
|
3129 |
target_nid = numa_migrate_prep(page, vma, addr, page_nid, &flags); |
d10e63f29
|
3130 |
pte_unmap_unlock(ptep, ptl); |
4daae3b4b
|
3131 |
if (target_nid == -1) { |
4daae3b4b
|
3132 3133 3134 3135 3136 |
put_page(page); goto out; } /* Migrate to the requested node */ |
1bc115d87
|
3137 |
migrated = migrate_misplaced_page(page, vma, target_nid); |
6688cc054
|
3138 |
if (migrated) { |
8191acbd3
|
3139 |
page_nid = target_nid; |
6688cc054
|
3140 |
flags |= TNF_MIGRATED; |
074c23817
|
3141 3142 |
} else flags |= TNF_MIGRATE_FAIL; |
4daae3b4b
|
3143 3144 |
out: |
8191acbd3
|
3145 |
if (page_nid != -1) |
6688cc054
|
3146 |
task_numa_fault(last_cpupid, page_nid, 1, flags); |
d10e63f29
|
3147 3148 |
return 0; } |
b96375f74
|
3149 3150 3151 |
static int create_huge_pmd(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, unsigned int flags) { |
fb6dd5fa4
|
3152 |
if (vma_is_anonymous(vma)) |
b96375f74
|
3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 |
return do_huge_pmd_anonymous_page(mm, vma, address, pmd, flags); if (vma->vm_ops->pmd_fault) return vma->vm_ops->pmd_fault(vma, address, pmd, flags); return VM_FAULT_FALLBACK; } static int wp_huge_pmd(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, pmd_t orig_pmd, unsigned int flags) { |
fb6dd5fa4
|
3163 |
if (vma_is_anonymous(vma)) |
b96375f74
|
3164 3165 3166 3167 3168 |
return do_huge_pmd_wp_page(mm, vma, address, pmd, orig_pmd); if (vma->vm_ops->pmd_fault) return vma->vm_ops->pmd_fault(vma, address, pmd, flags); return VM_FAULT_FALLBACK; } |
1da177e4c
|
3169 3170 3171 3172 3173 3174 3175 3176 3177 |
/* * These routines also need to handle stuff like marking pages dirty * and/or accessed for architectures that don't do it in hardware (most * RISC architectures). The early dirtying is also good on the i386. * * There is also a hook called "update_mmu_cache()" that architectures * with external mmu caches can use to update those (ie the Sparc or * PowerPC hashed page tables that act as extended TLBs). * |
c74df32c7
|
3178 3179 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. |
9a95f3cf7
|
3180 3181 3182 3183 |
* We return with pte unmapped and unlocked. * * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). |
1da177e4c
|
3184 |
*/ |
c02925540
|
3185 |
static int handle_pte_fault(struct mm_struct *mm, |
71e3aac07
|
3186 3187 |
struct vm_area_struct *vma, unsigned long address, pte_t *pte, pmd_t *pmd, unsigned int flags) |
1da177e4c
|
3188 3189 |
{ pte_t entry; |
8f4e2101f
|
3190 |
spinlock_t *ptl; |
1da177e4c
|
3191 |
|
e37c69827
|
3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 |
/* * some architectures can have larger ptes than wordsize, * e.g.ppc44x-defconfig has CONFIG_PTE_64BIT=y and CONFIG_32BIT=y, * so READ_ONCE or ACCESS_ONCE cannot guarantee atomic accesses. * The code below just needs a consistent view for the ifs and * we later double check anyway with the ptl lock held. So here * a barrier will do. */ entry = *pte; barrier(); |
1da177e4c
|
3202 |
if (!pte_present(entry)) { |
65500d234
|
3203 |
if (pte_none(entry)) { |
b53306285
|
3204 3205 3206 3207 |
if (vma_is_anonymous(vma)) return do_anonymous_page(mm, vma, address, pte, pmd, flags); else |
6b7339f4c
|
3208 3209 |
return do_fault(mm, vma, address, pte, pmd, flags, entry); |
65500d234
|
3210 |
} |
65500d234
|
3211 |
return do_swap_page(mm, vma, address, |
30c9f3a9f
|
3212 |
pte, pmd, flags, entry); |
1da177e4c
|
3213 |
} |
8a0516ed8
|
3214 |
if (pte_protnone(entry)) |
d10e63f29
|
3215 |
return do_numa_page(mm, vma, address, entry, pte, pmd); |
4c21e2f24
|
3216 |
ptl = pte_lockptr(mm, pmd); |
8f4e2101f
|
3217 3218 3219 |
spin_lock(ptl); if (unlikely(!pte_same(*pte, entry))) goto unlock; |
30c9f3a9f
|
3220 |
if (flags & FAULT_FLAG_WRITE) { |
1da177e4c
|
3221 |
if (!pte_write(entry)) |
8f4e2101f
|
3222 3223 |
return do_wp_page(mm, vma, address, pte, pmd, ptl, entry); |
1da177e4c
|
3224 3225 3226 |
entry = pte_mkdirty(entry); } entry = pte_mkyoung(entry); |
30c9f3a9f
|
3227 |
if (ptep_set_access_flags(vma, address, pte, entry, flags & FAULT_FLAG_WRITE)) { |
4b3073e1c
|
3228 |
update_mmu_cache(vma, address, pte); |
1a44e1490
|
3229 3230 3231 3232 3233 3234 3235 |
} else { /* * This is needed only for protection faults but the arch code * is not yet telling us if this is a protection fault or not. * This still avoids useless tlb flushes for .text page faults * with threads. */ |
30c9f3a9f
|
3236 |
if (flags & FAULT_FLAG_WRITE) |
61c77326d
|
3237 |
flush_tlb_fix_spurious_fault(vma, address); |
1a44e1490
|
3238 |
} |
8f4e2101f
|
3239 3240 |
unlock: pte_unmap_unlock(pte, ptl); |
83c54070e
|
3241 |
return 0; |
1da177e4c
|
3242 3243 3244 3245 |
} /* * By the time we get here, we already hold the mm semaphore |
9a95f3cf7
|
3246 3247 3248 |
* * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). |
1da177e4c
|
3249 |
*/ |
519e52473
|
3250 3251 |
static int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, unsigned int flags) |
1da177e4c
|
3252 3253 3254 3255 3256 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; |
1b2ee1266
|
3257 |
if (!arch_vma_access_permitted(vma, flags & FAULT_FLAG_WRITE, |
d61172b4b
|
3258 |
flags & FAULT_FLAG_INSTRUCTION, |
1b2ee1266
|
3259 |
flags & FAULT_FLAG_REMOTE)) |
33a709b25
|
3260 |
return VM_FAULT_SIGSEGV; |
ac9b9c667
|
3261 |
if (unlikely(is_vm_hugetlb_page(vma))) |
30c9f3a9f
|
3262 |
return hugetlb_fault(mm, vma, address, flags); |
1da177e4c
|
3263 |
|
1da177e4c
|
3264 |
pgd = pgd_offset(mm, address); |
1da177e4c
|
3265 3266 |
pud = pud_alloc(mm, pgd, address); if (!pud) |
c74df32c7
|
3267 |
return VM_FAULT_OOM; |
1da177e4c
|
3268 3269 |
pmd = pmd_alloc(mm, pud, address); if (!pmd) |
c74df32c7
|
3270 |
return VM_FAULT_OOM; |
71e3aac07
|
3271 |
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma)) { |
b96375f74
|
3272 |
int ret = create_huge_pmd(mm, vma, address, pmd, flags); |
c02925540
|
3273 3274 |
if (!(ret & VM_FAULT_FALLBACK)) return ret; |
71e3aac07
|
3275 3276 |
} else { pmd_t orig_pmd = *pmd; |
1f1d06c34
|
3277 |
int ret; |
71e3aac07
|
3278 |
barrier(); |
5c7fb56e5
|
3279 |
if (pmd_trans_huge(orig_pmd) || pmd_devmap(orig_pmd)) { |
a1dd450bc
|
3280 |
unsigned int dirty = flags & FAULT_FLAG_WRITE; |
8a0516ed8
|
3281 |
if (pmd_protnone(orig_pmd)) |
4daae3b4b
|
3282 |
return do_huge_pmd_numa_page(mm, vma, address, |
d10e63f29
|
3283 |
orig_pmd, pmd); |
3d59eebc5
|
3284 |
if (dirty && !pmd_write(orig_pmd)) { |
b96375f74
|
3285 3286 |
ret = wp_huge_pmd(mm, vma, address, pmd, orig_pmd, flags); |
9845cbbd1
|
3287 3288 |
if (!(ret & VM_FAULT_FALLBACK)) return ret; |
a1dd450bc
|
3289 3290 3291 |
} else { huge_pmd_set_accessed(mm, vma, address, pmd, orig_pmd, dirty); |
9845cbbd1
|
3292 |
return 0; |
1f1d06c34
|
3293 |
} |
71e3aac07
|
3294 3295 3296 3297 |
} } /* |
3ed3a4f0d
|
3298 |
* Use pte_alloc() instead of pte_alloc_map, because we can't |
71e3aac07
|
3299 3300 3301 |
* run pte_offset_map on the pmd, if an huge pmd could * materialize from under us from a different thread. */ |
3ed3a4f0d
|
3302 |
if (unlikely(pte_alloc(mm, pmd, address))) |
c74df32c7
|
3303 |
return VM_FAULT_OOM; |
ad33bb04b
|
3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 |
/* * If a huge pmd materialized under us just retry later. Use * pmd_trans_unstable() instead of pmd_trans_huge() to ensure the pmd * didn't become pmd_trans_huge under us and then back to pmd_none, as * a result of MADV_DONTNEED running immediately after a huge pmd fault * in a different thread of this mm, in turn leading to a misleading * pmd_trans_huge() retval. All we have to ensure is that it is a * regular pmd that we can walk with pte_offset_map() and we can do that * through an atomic read in C, which is what pmd_trans_unstable() * provides. */ if (unlikely(pmd_trans_unstable(pmd) || pmd_devmap(*pmd))) |
71e3aac07
|
3316 3317 3318 3319 3320 3321 3322 3323 |
return 0; /* * A regular pmd is established and it can't morph into a huge pmd * from under us anymore at this point because we hold the mmap_sem * read mode and khugepaged takes it in write mode. So now it's * safe to run pte_offset_map(). */ pte = pte_offset_map(pmd, address); |
1da177e4c
|
3324 |
|
30c9f3a9f
|
3325 |
return handle_pte_fault(mm, vma, address, pte, pmd, flags); |
1da177e4c
|
3326 |
} |
9a95f3cf7
|
3327 3328 3329 3330 3331 3332 |
/* * By the time we get here, we already hold the mm semaphore * * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). */ |
519e52473
|
3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 |
int handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, unsigned int flags) { int ret; __set_current_state(TASK_RUNNING); count_vm_event(PGFAULT); mem_cgroup_count_vm_event(mm, PGFAULT); /* do counter updates before entering really critical section. */ check_sync_rss_stat(current); /* * Enable the memcg OOM handling for faults triggered in user * space. Kernel faults are handled more gracefully. */ if (flags & FAULT_FLAG_USER) |
494264208
|
3351 |
mem_cgroup_oom_enable(); |
519e52473
|
3352 3353 |
ret = __handle_mm_fault(mm, vma, address, flags); |
494264208
|
3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 |
if (flags & FAULT_FLAG_USER) { mem_cgroup_oom_disable(); /* * The task may have entered a memcg OOM situation but * if the allocation error was handled gracefully (no * VM_FAULT_OOM), there is no need to kill anything. * Just clean up the OOM state peacefully. */ if (task_in_memcg_oom(current) && !(ret & VM_FAULT_OOM)) mem_cgroup_oom_synchronize(false); } |
3812c8c8f
|
3365 |
|
519e52473
|
3366 3367 |
return ret; } |
e1d6d01ab
|
3368 |
EXPORT_SYMBOL_GPL(handle_mm_fault); |
519e52473
|
3369 |
|
1da177e4c
|
3370 3371 3372 |
#ifndef __PAGETABLE_PUD_FOLDED /* * Allocate page upper directory. |
872fec16d
|
3373 |
* We've already handled the fast-path in-line. |
1da177e4c
|
3374 |
*/ |
1bb3630e8
|
3375 |
int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) |
1da177e4c
|
3376 |
{ |
c74df32c7
|
3377 3378 |
pud_t *new = pud_alloc_one(mm, address); if (!new) |
1bb3630e8
|
3379 |
return -ENOMEM; |
1da177e4c
|
3380 |
|
362a61ad6
|
3381 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d
|
3382 |
spin_lock(&mm->page_table_lock); |
1bb3630e8
|
3383 |
if (pgd_present(*pgd)) /* Another has populated it */ |
5e5419734
|
3384 |
pud_free(mm, new); |
1bb3630e8
|
3385 3386 |
else pgd_populate(mm, pgd, new); |
c74df32c7
|
3387 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8
|
3388 |
return 0; |
1da177e4c
|
3389 3390 3391 3392 3393 3394 |
} #endif /* __PAGETABLE_PUD_FOLDED */ #ifndef __PAGETABLE_PMD_FOLDED /* * Allocate page middle directory. |
872fec16d
|
3395 |
* We've already handled the fast-path in-line. |
1da177e4c
|
3396 |
*/ |
1bb3630e8
|
3397 |
int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) |
1da177e4c
|
3398 |
{ |
c74df32c7
|
3399 3400 |
pmd_t *new = pmd_alloc_one(mm, address); if (!new) |
1bb3630e8
|
3401 |
return -ENOMEM; |
1da177e4c
|
3402 |
|
362a61ad6
|
3403 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d
|
3404 |
spin_lock(&mm->page_table_lock); |
1da177e4c
|
3405 |
#ifndef __ARCH_HAS_4LEVEL_HACK |
dc6c9a35b
|
3406 3407 |
if (!pud_present(*pud)) { mm_inc_nr_pmds(mm); |
1bb3630e8
|
3408 |
pud_populate(mm, pud, new); |
dc6c9a35b
|
3409 |
} else /* Another has populated it */ |
5e5419734
|
3410 |
pmd_free(mm, new); |
dc6c9a35b
|
3411 3412 3413 |
#else if (!pgd_present(*pud)) { mm_inc_nr_pmds(mm); |
1bb3630e8
|
3414 |
pgd_populate(mm, pud, new); |
dc6c9a35b
|
3415 3416 |
} else /* Another has populated it */ pmd_free(mm, new); |
1da177e4c
|
3417 |
#endif /* __ARCH_HAS_4LEVEL_HACK */ |
c74df32c7
|
3418 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8
|
3419 |
return 0; |
e0f39591c
|
3420 |
} |
1da177e4c
|
3421 |
#endif /* __PAGETABLE_PMD_FOLDED */ |
1b36ba815
|
3422 |
static int __follow_pte(struct mm_struct *mm, unsigned long address, |
f8ad0f499
|
3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 |
pte_t **ptepp, spinlock_t **ptlp) { pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *ptep; pgd = pgd_offset(mm, address); if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) goto out; pud = pud_offset(pgd, address); if (pud_none(*pud) || unlikely(pud_bad(*pud))) goto out; pmd = pmd_offset(pud, address); |
f66055ab6
|
3439 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
f8ad0f499
|
3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 |
if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd))) goto out; /* We cannot handle huge page PFN maps. Luckily they don't exist. */ if (pmd_huge(*pmd)) goto out; ptep = pte_offset_map_lock(mm, pmd, address, ptlp); if (!ptep) goto out; if (!pte_present(*ptep)) goto unlock; *ptepp = ptep; return 0; unlock: pte_unmap_unlock(ptep, *ptlp); out: return -EINVAL; } |
1b36ba815
|
3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 |
static inline int follow_pte(struct mm_struct *mm, unsigned long address, pte_t **ptepp, spinlock_t **ptlp) { int res; /* (void) is needed to make gcc happy */ (void) __cond_lock(*ptlp, !(res = __follow_pte(mm, address, ptepp, ptlp))); return res; } |
3b6748e2d
|
3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 |
/** * follow_pfn - look up PFN at a user virtual address * @vma: memory mapping * @address: user virtual address * @pfn: location to store found PFN * * Only IO mappings and raw PFN mappings are allowed. * * Returns zero and the pfn at @pfn on success, -ve otherwise. */ int follow_pfn(struct vm_area_struct *vma, unsigned long address, unsigned long *pfn) { int ret = -EINVAL; spinlock_t *ptl; pte_t *ptep; if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) return ret; ret = follow_pte(vma->vm_mm, address, &ptep, &ptl); if (ret) return ret; *pfn = pte_pfn(*ptep); pte_unmap_unlock(ptep, ptl); return 0; } EXPORT_SYMBOL(follow_pfn); |
28b2ee20c
|
3497 |
#ifdef CONFIG_HAVE_IOREMAP_PROT |
d87fe6607
|
3498 3499 3500 |
int follow_phys(struct vm_area_struct *vma, unsigned long address, unsigned int flags, unsigned long *prot, resource_size_t *phys) |
28b2ee20c
|
3501 |
{ |
03668a4de
|
3502 |
int ret = -EINVAL; |
28b2ee20c
|
3503 3504 |
pte_t *ptep, pte; spinlock_t *ptl; |
28b2ee20c
|
3505 |
|
d87fe6607
|
3506 3507 |
if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) goto out; |
28b2ee20c
|
3508 |
|
03668a4de
|
3509 |
if (follow_pte(vma->vm_mm, address, &ptep, &ptl)) |
d87fe6607
|
3510 |
goto out; |
28b2ee20c
|
3511 |
pte = *ptep; |
03668a4de
|
3512 |
|
28b2ee20c
|
3513 3514 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
28b2ee20c
|
3515 3516 |
*prot = pgprot_val(pte_pgprot(pte)); |
03668a4de
|
3517 |
*phys = (resource_size_t)pte_pfn(pte) << PAGE_SHIFT; |
28b2ee20c
|
3518 |
|
03668a4de
|
3519 |
ret = 0; |
28b2ee20c
|
3520 3521 3522 |
unlock: pte_unmap_unlock(ptep, ptl); out: |
d87fe6607
|
3523 |
return ret; |
28b2ee20c
|
3524 3525 3526 3527 3528 3529 3530 |
} int generic_access_phys(struct vm_area_struct *vma, unsigned long addr, void *buf, int len, int write) { resource_size_t phys_addr; unsigned long prot = 0; |
2bc7273b0
|
3531 |
void __iomem *maddr; |
28b2ee20c
|
3532 |
int offset = addr & (PAGE_SIZE-1); |
d87fe6607
|
3533 |
if (follow_phys(vma, addr, write, &prot, &phys_addr)) |
28b2ee20c
|
3534 |
return -EINVAL; |
9cb12d7b4
|
3535 |
maddr = ioremap_prot(phys_addr, PAGE_ALIGN(len + offset), prot); |
28b2ee20c
|
3536 3537 3538 3539 3540 3541 3542 3543 |
if (write) memcpy_toio(maddr + offset, buf, len); else memcpy_fromio(buf, maddr + offset, len); iounmap(maddr); return len; } |
5a73633ef
|
3544 |
EXPORT_SYMBOL_GPL(generic_access_phys); |
28b2ee20c
|
3545 |
#endif |
0ec76a110
|
3546 |
/* |
206cb6365
|
3547 3548 |
* Access another process' address space as given in mm. If non-NULL, use the * given task for page fault accounting. |
0ec76a110
|
3549 |
*/ |
206cb6365
|
3550 3551 |
static int __access_remote_vm(struct task_struct *tsk, struct mm_struct *mm, unsigned long addr, void *buf, int len, int write) |
0ec76a110
|
3552 |
{ |
0ec76a110
|
3553 |
struct vm_area_struct *vma; |
0ec76a110
|
3554 |
void *old_buf = buf; |
0ec76a110
|
3555 |
down_read(&mm->mmap_sem); |
183ff22bb
|
3556 |
/* ignore errors, just check how much was successfully transferred */ |
0ec76a110
|
3557 3558 3559 |
while (len) { int bytes, ret, offset; void *maddr; |
28b2ee20c
|
3560 |
struct page *page = NULL; |
0ec76a110
|
3561 |
|
1e9877902
|
3562 |
ret = get_user_pages_remote(tsk, mm, addr, 1, |
0ec76a110
|
3563 |
write, 1, &page, &vma); |
28b2ee20c
|
3564 |
if (ret <= 0) { |
dbffcd03d
|
3565 3566 3567 |
#ifndef CONFIG_HAVE_IOREMAP_PROT break; #else |
28b2ee20c
|
3568 3569 3570 3571 |
/* * Check if this is a VM_IO | VM_PFNMAP VMA, which * we can access using slightly different code. */ |
28b2ee20c
|
3572 |
vma = find_vma(mm, addr); |
fe936dfc2
|
3573 |
if (!vma || vma->vm_start > addr) |
28b2ee20c
|
3574 3575 3576 3577 3578 |
break; if (vma->vm_ops && vma->vm_ops->access) ret = vma->vm_ops->access(vma, addr, buf, len, write); if (ret <= 0) |
28b2ee20c
|
3579 3580 |
break; bytes = ret; |
dbffcd03d
|
3581 |
#endif |
0ec76a110
|
3582 |
} else { |
28b2ee20c
|
3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 |
bytes = len; offset = addr & (PAGE_SIZE-1); if (bytes > PAGE_SIZE-offset) bytes = PAGE_SIZE-offset; maddr = kmap(page); if (write) { copy_to_user_page(vma, page, addr, maddr + offset, buf, bytes); set_page_dirty_lock(page); } else { copy_from_user_page(vma, page, addr, buf, maddr + offset, bytes); } kunmap(page); |
09cbfeaf1
|
3598 |
put_page(page); |
0ec76a110
|
3599 |
} |
0ec76a110
|
3600 3601 3602 3603 3604 |
len -= bytes; buf += bytes; addr += bytes; } up_read(&mm->mmap_sem); |
0ec76a110
|
3605 3606 3607 |
return buf - old_buf; } |
03252919b
|
3608 |
|
5ddd36b9c
|
3609 |
/** |
ae91dbfc9
|
3610 |
* access_remote_vm - access another process' address space |
5ddd36b9c
|
3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 |
* @mm: the mm_struct of the target address space * @addr: start address to access * @buf: source or destination buffer * @len: number of bytes to transfer * @write: whether the access is a write * * The caller must hold a reference on @mm. */ int access_remote_vm(struct mm_struct *mm, unsigned long addr, void *buf, int len, int write) { return __access_remote_vm(NULL, mm, addr, buf, len, write); } |
03252919b
|
3624 |
/* |
206cb6365
|
3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 |
* Access another process' address space. * Source/target buffer must be kernel space, * Do not walk the page table directly, use get_user_pages */ int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write) { struct mm_struct *mm; int ret; mm = get_task_mm(tsk); if (!mm) return 0; ret = __access_remote_vm(tsk, mm, addr, buf, len, write); mmput(mm); return ret; } |
03252919b
|
3644 3645 3646 3647 3648 3649 3650 |
/* * Print the name of a VMA. */ void print_vma_addr(char *prefix, unsigned long ip) { struct mm_struct *mm = current->mm; struct vm_area_struct *vma; |
e8bff74af
|
3651 3652 3653 3654 3655 3656 |
/* * Do not print if we are in atomic * contexts (in exception stacks, etc.): */ if (preempt_count()) return; |
03252919b
|
3657 3658 3659 3660 3661 3662 |
down_read(&mm->mmap_sem); vma = find_vma(mm, ip); if (vma && vma->vm_file) { struct file *f = vma->vm_file; char *buf = (char *)__get_free_page(GFP_KERNEL); if (buf) { |
2fbc57c53
|
3663 |
char *p; |
03252919b
|
3664 |
|
9bf39ab2a
|
3665 |
p = file_path(f, buf, PAGE_SIZE); |
03252919b
|
3666 3667 |
if (IS_ERR(p)) p = "?"; |
2fbc57c53
|
3668 |
printk("%s%s[%lx+%lx]", prefix, kbasename(p), |
03252919b
|
3669 3670 3671 3672 3673 |
vma->vm_start, vma->vm_end - vma->vm_start); free_page((unsigned long)buf); } } |
51a07e50b
|
3674 |
up_read(&mm->mmap_sem); |
03252919b
|
3675 |
} |
3ee1afa30
|
3676 |
|
662bbcb27
|
3677 |
#if defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_DEBUG_ATOMIC_SLEEP) |
9ec23531f
|
3678 |
void __might_fault(const char *file, int line) |
3ee1afa30
|
3679 |
{ |
95156f005
|
3680 3681 3682 3683 3684 3685 3686 3687 |
/* * Some code (nfs/sunrpc) uses socket ops on kernel memory while * holding the mmap_sem, this is safe because kernel memory doesn't * get paged out, therefore we'll never actually fault, and the * below annotations will generate false positives. */ if (segment_eq(get_fs(), KERNEL_DS)) return; |
9ec23531f
|
3688 |
if (pagefault_disabled()) |
662bbcb27
|
3689 |
return; |
9ec23531f
|
3690 3691 |
__might_sleep(file, line, 0); #if defined(CONFIG_DEBUG_ATOMIC_SLEEP) |
662bbcb27
|
3692 |
if (current->mm) |
3ee1afa30
|
3693 |
might_lock_read(¤t->mm->mmap_sem); |
9ec23531f
|
3694 |
#endif |
3ee1afa30
|
3695 |
} |
9ec23531f
|
3696 |
EXPORT_SYMBOL(__might_fault); |
3ee1afa30
|
3697 |
#endif |
47ad8475c
|
3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 |
#if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_HUGETLBFS) static void clear_gigantic_page(struct page *page, unsigned long addr, unsigned int pages_per_huge_page) { int i; struct page *p = page; might_sleep(); for (i = 0; i < pages_per_huge_page; i++, p = mem_map_next(p, page, i)) { cond_resched(); clear_user_highpage(p, addr + i * PAGE_SIZE); } } void clear_huge_page(struct page *page, unsigned long addr, unsigned int pages_per_huge_page) { int i; if (unlikely(pages_per_huge_page > MAX_ORDER_NR_PAGES)) { clear_gigantic_page(page, addr, pages_per_huge_page); return; } might_sleep(); for (i = 0; i < pages_per_huge_page; i++) { cond_resched(); clear_user_highpage(page + i, addr + i * PAGE_SIZE); } } static void copy_user_gigantic_page(struct page *dst, struct page *src, unsigned long addr, struct vm_area_struct *vma, unsigned int pages_per_huge_page) { int i; struct page *dst_base = dst; struct page *src_base = src; for (i = 0; i < pages_per_huge_page; ) { cond_resched(); copy_user_highpage(dst, src, addr + i*PAGE_SIZE, vma); i++; dst = mem_map_next(dst, dst_base, i); src = mem_map_next(src, src_base, i); } } void copy_user_huge_page(struct page *dst, struct page *src, unsigned long addr, struct vm_area_struct *vma, unsigned int pages_per_huge_page) { int i; if (unlikely(pages_per_huge_page > MAX_ORDER_NR_PAGES)) { copy_user_gigantic_page(dst, src, addr, vma, pages_per_huge_page); return; } might_sleep(); for (i = 0; i < pages_per_huge_page; i++) { cond_resched(); copy_user_highpage(dst + i, src + i, addr + i*PAGE_SIZE, vma); } } #endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */ |
49076ec2c
|
3769 |
|
40b64acd1
|
3770 |
#if USE_SPLIT_PTE_PTLOCKS && ALLOC_SPLIT_PTLOCKS |
b35f1819a
|
3771 3772 3773 3774 3775 3776 3777 3778 |
static struct kmem_cache *page_ptl_cachep; void __init ptlock_cache_init(void) { page_ptl_cachep = kmem_cache_create("page->ptl", sizeof(spinlock_t), 0, SLAB_PANIC, NULL); } |
539edb584
|
3779 |
bool ptlock_alloc(struct page *page) |
49076ec2c
|
3780 3781 |
{ spinlock_t *ptl; |
b35f1819a
|
3782 |
ptl = kmem_cache_alloc(page_ptl_cachep, GFP_KERNEL); |
49076ec2c
|
3783 3784 |
if (!ptl) return false; |
539edb584
|
3785 |
page->ptl = ptl; |
49076ec2c
|
3786 3787 |
return true; } |
539edb584
|
3788 |
void ptlock_free(struct page *page) |
49076ec2c
|
3789 |
{ |
b35f1819a
|
3790 |
kmem_cache_free(page_ptl_cachep, page->ptl); |
49076ec2c
|
3791 3792 |
} #endif |