Blame view
mm/memory.c
120 KB
1da177e4c Linux-2.6.12-rc2 |
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 ksm: identify Pag... |
48 |
#include <linux/ksm.h> |
1da177e4c Linux-2.6.12-rc2 |
49 |
#include <linux/rmap.h> |
b95f1b31b mm: Map most file... |
50 |
#include <linux/export.h> |
0ff922452 [PATCH] per-task-... |
51 |
#include <linux/delayacct.h> |
1da177e4c Linux-2.6.12-rc2 |
52 |
#include <linux/init.h> |
edc79b2a4 [PATCH] mm: balan... |
53 |
#include <linux/writeback.h> |
8a9f3ccd2 Memory controller... |
54 |
#include <linux/memcontrol.h> |
cddb8a5c1 mmu-notifiers: core |
55 |
#include <linux/mmu_notifier.h> |
3dc147414 badpage: replace ... |
56 57 58 |
#include <linux/kallsyms.h> #include <linux/swapops.h> #include <linux/elf.h> |
5a0e3ad6a include cleanup: ... |
59 |
#include <linux/gfp.h> |
4daae3b4b mm: mempolicy: Us... |
60 |
#include <linux/migrate.h> |
2fbc57c53 mm: use kbasename() |
61 |
#include <linux/string.h> |
0abdd7a81 dma-debug: introd... |
62 |
#include <linux/dma-debug.h> |
1592eef01 mm: add debugfs t... |
63 |
#include <linux/debugfs.h> |
1da177e4c Linux-2.6.12-rc2 |
64 |
|
6952b61de headers: taskstat... |
65 |
#include <asm/io.h> |
1da177e4c Linux-2.6.12-rc2 |
66 67 68 69 70 |
#include <asm/pgalloc.h> #include <asm/uaccess.h> #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/pgtable.h> |
42b777281 mm: remove double... |
71 |
#include "internal.h" |
90572890d mm: numa: Change ... |
72 73 |
#ifdef LAST_CPUPID_NOT_IN_PAGE_FLAGS #warning Unfortunate NUMA and NUMA Balancing config, growing page-frame for last_cpupid. |
75980e97d mm: fold page->_l... |
74 |
#endif |
d41dee369 [PATCH] sparsemem... |
75 |
#ifndef CONFIG_NEED_MULTIPLE_NODES |
1da177e4c Linux-2.6.12-rc2 |
76 77 78 79 80 81 82 |
/* 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 Linux-2.6.12-rc2 |
83 84 85 86 87 88 89 90 |
/* * 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 Linux-2.6.12-rc2 |
91 |
|
1da177e4c Linux-2.6.12-rc2 |
92 |
EXPORT_SYMBOL(high_memory); |
1da177e4c Linux-2.6.12-rc2 |
93 |
|
32a932332 brk randomization... |
94 95 96 97 98 99 100 101 102 103 104 105 |
/* * 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 [PATCH] x86_64: A... |
106 107 108 109 |
static int __init disable_randmaps(char *s) { randomize_va_space = 0; |
9b41046cd [PATCH] Don't pas... |
110 |
return 1; |
a62eaf151 [PATCH] x86_64: A... |
111 112 |
} __setup("norandmaps", disable_randmaps); |
62eede62d mm: ZERO_PAGE wit... |
113 |
unsigned long zero_pfn __read_mostly; |
03f6462a3 mm: move highest_... |
114 |
unsigned long highest_memmap_pfn __read_mostly; |
a13ea5b75 mm: reinstate ZER... |
115 116 117 118 119 120 121 122 123 124 |
/* * 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 [PATCH] x86_64: A... |
125 |
|
d559db086 mm: clean up mm_c... |
126 |
|
34e55232e mm: avoid false s... |
127 |
#if defined(SPLIT_RSS_COUNTING) |
ea48cf786 mm, counters: fol... |
128 |
void sync_mm_rss(struct mm_struct *mm) |
34e55232e mm: avoid false s... |
129 130 131 132 |
{ int i; for (i = 0; i < NR_MM_COUNTERS; i++) { |
05af2e104 mm, counters: rem... |
133 134 135 |
if (current->rss_stat.count[i]) { add_mm_counter(mm, i, current->rss_stat.count[i]); current->rss_stat.count[i] = 0; |
34e55232e mm: avoid false s... |
136 137 |
} } |
05af2e104 mm, counters: rem... |
138 |
current->rss_stat.events = 0; |
34e55232e mm: avoid false s... |
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 |
} 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 mm, counters: fol... |
160 |
sync_mm_rss(task->mm); |
34e55232e mm: avoid false s... |
161 |
} |
9547d01bf mm: uninline larg... |
162 |
#else /* SPLIT_RSS_COUNTING */ |
34e55232e mm: avoid false s... |
163 164 165 166 167 168 169 |
#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 mm: uninline larg... |
170 171 172 173 174 175 176 177 178 179 180 181 182 |
#endif /* SPLIT_RSS_COUNTING */ #ifdef HAVE_GENERIC_MMU_GATHER static int tlb_next_batch(struct mmu_gather *tlb) { struct mmu_gather_batch *batch; batch = tlb->active; if (batch->next) { tlb->active = batch->next; return 1; } |
53a59fc67 mm: limit mmu_gat... |
183 184 |
if (tlb->batch_count == MAX_GATHER_BATCH_COUNT) return 0; |
9547d01bf mm: uninline larg... |
185 186 187 |
batch = (void *)__get_free_pages(GFP_NOWAIT | __GFP_NOWARN, 0); if (!batch) return 0; |
53a59fc67 mm: limit mmu_gat... |
188 |
tlb->batch_count++; |
9547d01bf mm: uninline larg... |
189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 |
batch->next = NULL; batch->nr = 0; batch->max = MAX_GATHER_BATCH; tlb->active->next = batch; tlb->active = batch; return 1; } /* 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 Fix TLB gather vi... |
204 |
void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, unsigned long start, unsigned long end) |
9547d01bf mm: uninline larg... |
205 206 |
{ tlb->mm = mm; |
2b047252d Fix TLB gather vi... |
207 208 |
/* Is it from 0 to ~0? */ tlb->fullmm = !(start | (end+1)); |
1de14c3c5 x86-32: Fix possi... |
209 |
tlb->need_flush_all = 0; |
2b047252d Fix TLB gather vi... |
210 211 |
tlb->start = start; tlb->end = end; |
9547d01bf mm: uninline larg... |
212 |
tlb->need_flush = 0; |
9547d01bf mm: uninline larg... |
213 214 215 216 |
tlb->local.next = NULL; tlb->local.nr = 0; tlb->local.max = ARRAY_SIZE(tlb->__pages); tlb->active = &tlb->local; |
53a59fc67 mm: limit mmu_gat... |
217 |
tlb->batch_count = 0; |
9547d01bf mm: uninline larg... |
218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb->batch = NULL; #endif } void tlb_flush_mmu(struct mmu_gather *tlb) { struct mmu_gather_batch *batch; if (!tlb->need_flush) return; tlb->need_flush = 0; tlb_flush(tlb); #ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb_table_flush(tlb); |
34e55232e mm: avoid false s... |
234 |
#endif |
9547d01bf mm: uninline larg... |
235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 |
for (batch = &tlb->local; batch; batch = batch->next) { free_pages_and_swap_cache(batch->pages, batch->nr); batch->nr = 0; } tlb->active = &tlb->local; } /* 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; |
f21760b15 thp: add tlb_remo... |
271 |
VM_BUG_ON(!tlb->need_flush); |
9547d01bf mm: uninline larg... |
272 |
|
9547d01bf mm: uninline larg... |
273 274 275 276 277 |
batch = tlb->active; batch->pages[batch->nr++] = page; if (batch->nr == batch->max) { if (!tlb_next_batch(tlb)) return 0; |
0b43c3aab mm: __tlb_remove_... |
278 |
batch = tlb->active; |
9547d01bf mm: uninline larg... |
279 |
} |
309381fea mm: dump page whe... |
280 |
VM_BUG_ON_PAGE(batch->nr > batch->max, page); |
9547d01bf mm: uninline larg... |
281 282 283 284 285 |
return batch->max - batch->nr; } #endif /* HAVE_GENERIC_MMU_GATHER */ |
267239116 mm, powerpc: move... |
286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 |
#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; tlb->need_flush = 1; /* * 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 mm: uninline larg... |
360 |
#endif /* CONFIG_HAVE_RCU_TABLE_FREE */ |
267239116 mm, powerpc: move... |
361 |
|
1da177e4c Linux-2.6.12-rc2 |
362 |
/* |
1da177e4c Linux-2.6.12-rc2 |
363 364 365 |
* Note: this doesn't free the actual pages themselves. That * has been handled earlier when unmapping all the memory regions. */ |
9e1b32caa mm: Pass virtual ... |
366 367 |
static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
368 |
{ |
2f569afd9 CONFIG_HIGHPTE vs... |
369 |
pgtable_t token = pmd_pgtable(*pmd); |
e0da382c9 [PATCH] freepgt: ... |
370 |
pmd_clear(pmd); |
9e1b32caa mm: Pass virtual ... |
371 |
pte_free_tlb(tlb, token, addr); |
e1f56c89b mm: convert mm->n... |
372 |
atomic_long_dec(&tlb->mm->nr_ptes); |
1da177e4c Linux-2.6.12-rc2 |
373 |
} |
e0da382c9 [PATCH] freepgt: ... |
374 375 376 |
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 Linux-2.6.12-rc2 |
377 378 379 |
{ pmd_t *pmd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
380 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
381 |
|
e0da382c9 [PATCH] freepgt: ... |
382 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
383 |
pmd = pmd_offset(pud, addr); |
1da177e4c Linux-2.6.12-rc2 |
384 385 386 387 |
do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
9e1b32caa mm: Pass virtual ... |
388 |
free_pte_range(tlb, pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
389 |
} while (pmd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
390 391 392 393 394 395 396 |
start &= PUD_MASK; if (start < floor) return; if (ceiling) { ceiling &= PUD_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
397 |
} |
e0da382c9 [PATCH] freepgt: ... |
398 399 400 401 402 |
if (end - 1 > ceiling - 1) return; pmd = pmd_offset(pud, start); pud_clear(pud); |
9e1b32caa mm: Pass virtual ... |
403 |
pmd_free_tlb(tlb, pmd, start); |
1da177e4c Linux-2.6.12-rc2 |
404 |
} |
e0da382c9 [PATCH] freepgt: ... |
405 406 407 |
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 Linux-2.6.12-rc2 |
408 409 410 |
{ pud_t *pud; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
411 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
412 |
|
e0da382c9 [PATCH] freepgt: ... |
413 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
414 |
pud = pud_offset(pgd, addr); |
1da177e4c Linux-2.6.12-rc2 |
415 416 417 418 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
e0da382c9 [PATCH] freepgt: ... |
419 |
free_pmd_range(tlb, pud, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
420 |
} while (pud++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
421 422 423 424 425 426 427 |
start &= PGDIR_MASK; if (start < floor) return; if (ceiling) { ceiling &= PGDIR_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
428 |
} |
e0da382c9 [PATCH] freepgt: ... |
429 430 431 432 433 |
if (end - 1 > ceiling - 1) return; pud = pud_offset(pgd, start); pgd_clear(pgd); |
9e1b32caa mm: Pass virtual ... |
434 |
pud_free_tlb(tlb, pud, start); |
1da177e4c Linux-2.6.12-rc2 |
435 436 437 |
} /* |
e0da382c9 [PATCH] freepgt: ... |
438 |
* This function frees user-level page tables of a process. |
1da177e4c Linux-2.6.12-rc2 |
439 |
*/ |
42b777281 mm: remove double... |
440 |
void free_pgd_range(struct mmu_gather *tlb, |
e0da382c9 [PATCH] freepgt: ... |
441 442 |
unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c Linux-2.6.12-rc2 |
443 444 445 |
{ pgd_t *pgd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 |
/* * 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 Linux-2.6.12-rc2 |
472 |
|
e0da382c9 [PATCH] freepgt: ... |
473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 |
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 mm: remove double... |
488 |
pgd = pgd_offset(tlb->mm, addr); |
1da177e4c Linux-2.6.12-rc2 |
489 490 491 492 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
42b777281 mm: remove double... |
493 |
free_pud_range(tlb, pgd, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
494 |
} while (pgd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
495 |
} |
42b777281 mm: remove double... |
496 |
void free_pgtables(struct mmu_gather *tlb, struct vm_area_struct *vma, |
3bf5ee956 [PATCH] freepgt: ... |
497 |
unsigned long floor, unsigned long ceiling) |
e0da382c9 [PATCH] freepgt: ... |
498 499 500 501 |
{ while (vma) { struct vm_area_struct *next = vma->vm_next; unsigned long addr = vma->vm_start; |
8f4f8c164 [PATCH] mm: unlin... |
502 |
/* |
25d9e2d15 truncate: new hel... |
503 504 |
* Hide vma from rmap and truncate_pagecache before freeing * pgtables |
8f4f8c164 [PATCH] mm: unlin... |
505 |
*/ |
5beb49305 mm: change anon_v... |
506 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
507 |
unlink_file_vma(vma); |
9da61aef0 [PATCH] hugepage:... |
508 |
if (is_vm_hugetlb_page(vma)) { |
3bf5ee956 [PATCH] freepgt: ... |
509 |
hugetlb_free_pgd_range(tlb, addr, vma->vm_end, |
e0da382c9 [PATCH] freepgt: ... |
510 |
floor, next? next->vm_start: ceiling); |
3bf5ee956 [PATCH] freepgt: ... |
511 512 513 514 515 |
} else { /* * Optimization: gather nearby vmas into one call down */ while (next && next->vm_start <= vma->vm_end + PMD_SIZE |
4866920b9 [PATCH] hugepage:... |
516 |
&& !is_vm_hugetlb_page(next)) { |
3bf5ee956 [PATCH] freepgt: ... |
517 518 |
vma = next; next = vma->vm_next; |
5beb49305 mm: change anon_v... |
519 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
520 |
unlink_file_vma(vma); |
3bf5ee956 [PATCH] freepgt: ... |
521 522 523 524 |
} free_pgd_range(tlb, addr, vma->vm_end, floor, next? next->vm_start: ceiling); } |
e0da382c9 [PATCH] freepgt: ... |
525 526 |
vma = next; } |
1da177e4c Linux-2.6.12-rc2 |
527 |
} |
8ac1f8320 thp: pte alloc tr... |
528 529 |
int __pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma, pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
530 |
{ |
c4088ebdc mm: convert the r... |
531 |
spinlock_t *ptl; |
2f569afd9 CONFIG_HIGHPTE vs... |
532 |
pgtable_t new = pte_alloc_one(mm, address); |
8ac1f8320 thp: pte alloc tr... |
533 |
int wait_split_huge_page; |
1bb3630e8 [PATCH] mm: ptd_a... |
534 535 |
if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
536 537 538 539 540 541 542 543 544 545 546 547 548 549 |
/* * 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 mm: convert the r... |
550 |
ptl = pmd_lock(mm, pmd); |
8ac1f8320 thp: pte alloc tr... |
551 552 |
wait_split_huge_page = 0; if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
e1f56c89b mm: convert mm->n... |
553 |
atomic_long_inc(&mm->nr_ptes); |
1da177e4c Linux-2.6.12-rc2 |
554 |
pmd_populate(mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
555 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
556 557 |
} else if (unlikely(pmd_trans_splitting(*pmd))) wait_split_huge_page = 1; |
c4088ebdc mm: convert the r... |
558 |
spin_unlock(ptl); |
2f569afd9 CONFIG_HIGHPTE vs... |
559 560 |
if (new) pte_free(mm, new); |
8ac1f8320 thp: pte alloc tr... |
561 562 |
if (wait_split_huge_page) wait_split_huge_page(vma->anon_vma, pmd); |
1bb3630e8 [PATCH] mm: ptd_a... |
563 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
564 |
} |
1bb3630e8 [PATCH] mm: ptd_a... |
565 |
int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
566 |
{ |
1bb3630e8 [PATCH] mm: ptd_a... |
567 568 569 |
pte_t *new = pte_alloc_one_kernel(&init_mm, address); if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
570 |
smp_wmb(); /* See comment in __pte_alloc */ |
1bb3630e8 [PATCH] mm: ptd_a... |
571 |
spin_lock(&init_mm.page_table_lock); |
8ac1f8320 thp: pte alloc tr... |
572 |
if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
1bb3630e8 [PATCH] mm: ptd_a... |
573 |
pmd_populate_kernel(&init_mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
574 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
575 576 |
} else VM_BUG_ON(pmd_trans_splitting(*pmd)); |
1bb3630e8 [PATCH] mm: ptd_a... |
577 |
spin_unlock(&init_mm.page_table_lock); |
2f569afd9 CONFIG_HIGHPTE vs... |
578 579 |
if (new) pte_free_kernel(&init_mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
580 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
581 |
} |
d559db086 mm: clean up mm_c... |
582 583 584 585 586 587 |
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 [PATCH] mm: batch... |
588 |
{ |
d559db086 mm: clean up mm_c... |
589 |
int i; |
34e55232e mm: avoid false s... |
590 |
if (current->mm == mm) |
05af2e104 mm, counters: rem... |
591 |
sync_mm_rss(mm); |
d559db086 mm: clean up mm_c... |
592 593 594 |
for (i = 0; i < NR_MM_COUNTERS; i++) if (rss[i]) add_mm_counter(mm, i, rss[i]); |
ae8597623 [PATCH] mm: batch... |
595 |
} |
1da177e4c Linux-2.6.12-rc2 |
596 |
/* |
6aab341e0 mm: re-architect ... |
597 598 599 |
* 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 [PATCH] core remo... |
600 601 602 |
* * The calling function must still handle the error. */ |
3dc147414 badpage: replace ... |
603 604 |
static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr, pte_t pte, struct page *page) |
b5810039a [PATCH] core remo... |
605 |
{ |
3dc147414 badpage: replace ... |
606 607 608 609 610 |
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 badpage: ratelimi... |
611 612 613 614 615 616 617 618 619 620 621 622 623 624 |
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) { |
1e9e63650 badpage: KERN_ALE... |
625 626 627 |
printk(KERN_ALERT "BUG: Bad page map: %lu messages suppressed ", |
d936cf9b3 badpage: ratelimi... |
628 629 630 631 632 633 634 |
nr_unshown); nr_unshown = 0; } nr_shown = 0; } if (nr_shown++ == 0) resume = jiffies + 60 * HZ; |
3dc147414 badpage: replace ... |
635 636 637 |
mapping = vma->vm_file ? vma->vm_file->f_mapping : NULL; index = linear_page_index(vma, addr); |
1e9e63650 badpage: KERN_ALE... |
638 639 640 |
printk(KERN_ALERT "BUG: Bad page map in process %s pte:%08llx pmd:%08llx ", |
3dc147414 badpage: replace ... |
641 642 |
current->comm, (long long)pte_val(pte), (long long)pmd_val(*pmd)); |
718a38211 mm: introduce dum... |
643 |
if (page) |
f0b791a34 mm: print more de... |
644 |
dump_page(page, "bad pte"); |
1e9e63650 badpage: KERN_ALE... |
645 |
printk(KERN_ALERT |
3dc147414 badpage: replace ... |
646 647 648 649 650 651 652 |
"addr:%p vm_flags:%08lx anon_vma:%p mapping:%p index:%lx ", (void *)addr, vma->vm_flags, vma->anon_vma, mapping, index); /* * Choose text because data symbols depend on CONFIG_KALLSYMS_ALL=y */ if (vma->vm_ops) |
071361d34 mm: Convert print... |
653 654 655 |
printk(KERN_ALERT "vma->vm_ops->fault: %pSR ", vma->vm_ops->fault); |
72c2d5319 file->f_op is nev... |
656 |
if (vma->vm_file) |
071361d34 mm: Convert print... |
657 658 659 |
printk(KERN_ALERT "vma->vm_file->f_op->mmap: %pSR ", vma->vm_file->f_op->mmap); |
b5810039a [PATCH] core remo... |
660 |
dump_stack(); |
373d4d099 taint: add explic... |
661 |
add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); |
b5810039a [PATCH] core remo... |
662 |
} |
2ec74c3ef mm: move all mmu ... |
663 |
static inline bool is_cow_mapping(vm_flags_t flags) |
67121172f Allow arbitrary r... |
664 665 666 |
{ return (flags & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE; } |
b5810039a [PATCH] core remo... |
667 |
/* |
7e675137a mm: introduce pte... |
668 |
* vm_normal_page -- This function gets the "struct page" associated with a pte. |
6aab341e0 mm: re-architect ... |
669 |
* |
7e675137a mm: introduce pte... |
670 671 672 |
* "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 mm: introduce VM_... |
673 |
* |
7e675137a mm: introduce pte... |
674 675 676 677 678 679 680 681 |
* 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 mm: re-architect ... |
682 |
* |
b379d7901 mm: introduce VM_... |
683 684 |
* 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 mm: introduce pte... |
685 686 |
* set, and the vm_pgoff will point to the first PFN mapped: thus every special * mapping will always honor the rule |
6aab341e0 mm: re-architect ... |
687 688 689 |
* * pfn_of_page == vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT) * |
7e675137a mm: introduce pte... |
690 691 692 693 694 695 |
* 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 mm: introduce VM_... |
696 |
* |
b379d7901 mm: introduce VM_... |
697 |
* |
7e675137a mm: introduce pte... |
698 |
* In order to support COW of arbitrary special mappings, we have VM_MIXEDMAP. |
b379d7901 mm: introduce VM_... |
699 700 701 702 703 704 705 706 707 |
* * 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 [PATCH] unpaged: ... |
708 |
*/ |
7e675137a mm: introduce pte... |
709 710 711 712 713 714 715 |
#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 [PATCH] unpaged: ... |
716 |
{ |
22b31eec6 badpage: vm_norma... |
717 |
unsigned long pfn = pte_pfn(pte); |
7e675137a mm: introduce pte... |
718 719 |
if (HAVE_PTE_SPECIAL) { |
22b31eec6 badpage: vm_norma... |
720 721 |
if (likely(!pte_special(pte))) goto check_pfn; |
a13ea5b75 mm: reinstate ZER... |
722 723 |
if (vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)) return NULL; |
62eede62d mm: ZERO_PAGE wit... |
724 |
if (!is_zero_pfn(pfn)) |
22b31eec6 badpage: vm_norma... |
725 |
print_bad_pte(vma, addr, pte, NULL); |
7e675137a mm: introduce pte... |
726 727 728 729 |
return NULL; } /* !HAVE_PTE_SPECIAL case follows: */ |
b379d7901 mm: introduce VM_... |
730 731 732 733 734 735 |
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 mm: introduce pte... |
736 737 |
unsigned long off; off = (addr - vma->vm_start) >> PAGE_SHIFT; |
b379d7901 mm: introduce VM_... |
738 739 740 741 742 |
if (pfn == vma->vm_pgoff + off) return NULL; if (!is_cow_mapping(vma->vm_flags)) return NULL; } |
6aab341e0 mm: re-architect ... |
743 |
} |
62eede62d mm: ZERO_PAGE wit... |
744 745 |
if (is_zero_pfn(pfn)) return NULL; |
22b31eec6 badpage: vm_norma... |
746 747 748 749 750 |
check_pfn: if (unlikely(pfn > highest_memmap_pfn)) { print_bad_pte(vma, addr, pte, NULL); return NULL; } |
6aab341e0 mm: re-architect ... |
751 752 |
/* |
7e675137a mm: introduce pte... |
753 |
* NOTE! We still have PageReserved() pages in the page tables. |
7e675137a mm: introduce pte... |
754 |
* eg. VDSO mappings can cause them to exist. |
6aab341e0 mm: re-architect ... |
755 |
*/ |
b379d7901 mm: introduce VM_... |
756 |
out: |
6aab341e0 mm: re-architect ... |
757 |
return pfn_to_page(pfn); |
ee498ed73 [PATCH] unpaged: ... |
758 759 760 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
761 762 763 |
* 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 Linux-2.6.12-rc2 |
764 |
*/ |
570a335b8 swap_info: swap c... |
765 |
static inline unsigned long |
1da177e4c Linux-2.6.12-rc2 |
766 |
copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
b5810039a [PATCH] core remo... |
767 |
pte_t *dst_pte, pte_t *src_pte, struct vm_area_struct *vma, |
8c1037627 [PATCH] mm: copy_... |
768 |
unsigned long addr, int *rss) |
1da177e4c Linux-2.6.12-rc2 |
769 |
{ |
b5810039a [PATCH] core remo... |
770 |
unsigned long vm_flags = vma->vm_flags; |
1da177e4c Linux-2.6.12-rc2 |
771 772 |
pte_t pte = *src_pte; struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
773 774 775 776 |
/* pte contains position in swap or file, so copy. */ if (unlikely(!pte_present(pte))) { if (!pte_file(pte)) { |
0697212a4 [PATCH] Swapless ... |
777 |
swp_entry_t entry = pte_to_swp_entry(pte); |
570a335b8 swap_info: swap c... |
778 779 |
if (swap_duplicate(entry) < 0) return entry.val; |
1da177e4c Linux-2.6.12-rc2 |
780 781 782 |
/* make sure dst_mm is on swapoff's mmlist. */ if (unlikely(list_empty(&dst_mm->mmlist))) { spin_lock(&mmlist_lock); |
f412ac08c [PATCH] mm: fix r... |
783 784 785 |
if (list_empty(&dst_mm->mmlist)) list_add(&dst_mm->mmlist, &src_mm->mmlist); |
1da177e4c Linux-2.6.12-rc2 |
786 787 |
spin_unlock(&mmlist_lock); } |
b084d4353 mm: count swap usage |
788 789 |
if (likely(!non_swap_entry(entry))) rss[MM_SWAPENTS]++; |
9f9f1acd7 mm: fix rss count... |
790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 |
else if (is_migration_entry(entry)) { page = migration_entry_to_page(entry); if (PageAnon(page)) rss[MM_ANONPAGES]++; else rss[MM_FILEPAGES]++; 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); |
c3d16e165 mm: migration: do... |
806 807 |
if (pte_swp_soft_dirty(*src_pte)) pte = pte_swp_mksoft_dirty(pte); |
9f9f1acd7 mm: fix rss count... |
808 809 |
set_pte_at(src_mm, addr, src_pte, pte); } |
0697212a4 [PATCH] Swapless ... |
810 |
} |
1da177e4c Linux-2.6.12-rc2 |
811 |
} |
ae8597623 [PATCH] mm: batch... |
812 |
goto out_set_pte; |
1da177e4c Linux-2.6.12-rc2 |
813 |
} |
1da177e4c Linux-2.6.12-rc2 |
814 815 816 817 |
/* * If it's a COW mapping, write protect it both * in the parent and the child */ |
67121172f Allow arbitrary r... |
818 |
if (is_cow_mapping(vm_flags)) { |
1da177e4c Linux-2.6.12-rc2 |
819 |
ptep_set_wrprotect(src_mm, addr, src_pte); |
3dc907951 [PATCH] paravirt:... |
820 |
pte = pte_wrprotect(pte); |
1da177e4c Linux-2.6.12-rc2 |
821 822 823 824 825 826 827 828 829 |
} /* * 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 mm: re-architect ... |
830 831 832 833 |
page = vm_normal_page(vma, addr, pte); if (page) { get_page(page); |
21333b2b6 ksm: no debug in ... |
834 |
page_dup_rmap(page); |
d559db086 mm: clean up mm_c... |
835 836 837 838 |
if (PageAnon(page)) rss[MM_ANONPAGES]++; else rss[MM_FILEPAGES]++; |
6aab341e0 mm: re-architect ... |
839 |
} |
ae8597623 [PATCH] mm: batch... |
840 841 842 |
out_set_pte: set_pte_at(dst_mm, addr, dst_pte, pte); |
570a335b8 swap_info: swap c... |
843 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
844 |
} |
71e3aac07 thp: transparent ... |
845 846 847 |
int copy_pte_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, pmd_t *dst_pmd, pmd_t *src_pmd, struct vm_area_struct *vma, unsigned long addr, unsigned long end) |
1da177e4c Linux-2.6.12-rc2 |
848 |
{ |
c36987e2e mm: don't call pt... |
849 |
pte_t *orig_src_pte, *orig_dst_pte; |
1da177e4c Linux-2.6.12-rc2 |
850 |
pte_t *src_pte, *dst_pte; |
c74df32c7 [PATCH] mm: ptd_a... |
851 |
spinlock_t *src_ptl, *dst_ptl; |
e040f218b [PATCH] mm: copy_... |
852 |
int progress = 0; |
d559db086 mm: clean up mm_c... |
853 |
int rss[NR_MM_COUNTERS]; |
570a335b8 swap_info: swap c... |
854 |
swp_entry_t entry = (swp_entry_t){0}; |
1da177e4c Linux-2.6.12-rc2 |
855 856 |
again: |
d559db086 mm: clean up mm_c... |
857 |
init_rss_vec(rss); |
c74df32c7 [PATCH] mm: ptd_a... |
858 |
dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl); |
1da177e4c Linux-2.6.12-rc2 |
859 860 |
if (!dst_pte) return -ENOMEM; |
ece0e2b64 mm: remove pte_*m... |
861 |
src_pte = pte_offset_map(src_pmd, addr); |
4c21e2f24 [PATCH] mm: split... |
862 |
src_ptl = pte_lockptr(src_mm, src_pmd); |
f20dc5f7c [PATCH] lockdep: ... |
863 |
spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING); |
c36987e2e mm: don't call pt... |
864 865 |
orig_src_pte = src_pte; orig_dst_pte = dst_pte; |
6606c3e0d [PATCH] paravirt:... |
866 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
867 |
|
1da177e4c Linux-2.6.12-rc2 |
868 869 870 871 872 |
do { /* * We are holding two locks at this point - either of them * could generate latencies in another task on another CPU. */ |
e040f218b [PATCH] mm: copy_... |
873 874 875 |
if (progress >= 32) { progress = 0; if (need_resched() || |
95c354fe9 spinlock: lockbre... |
876 |
spin_needbreak(src_ptl) || spin_needbreak(dst_ptl)) |
e040f218b [PATCH] mm: copy_... |
877 878 |
break; } |
1da177e4c Linux-2.6.12-rc2 |
879 880 881 882 |
if (pte_none(*src_pte)) { progress++; continue; } |
570a335b8 swap_info: swap c... |
883 884 885 886 |
entry.val = copy_one_pte(dst_mm, src_mm, dst_pte, src_pte, vma, addr, rss); if (entry.val) break; |
1da177e4c Linux-2.6.12-rc2 |
887 888 |
progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
889 |
|
6606c3e0d [PATCH] paravirt:... |
890 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
891 |
spin_unlock(src_ptl); |
ece0e2b64 mm: remove pte_*m... |
892 |
pte_unmap(orig_src_pte); |
d559db086 mm: clean up mm_c... |
893 |
add_mm_rss_vec(dst_mm, rss); |
c36987e2e mm: don't call pt... |
894 |
pte_unmap_unlock(orig_dst_pte, dst_ptl); |
c74df32c7 [PATCH] mm: ptd_a... |
895 |
cond_resched(); |
570a335b8 swap_info: swap c... |
896 897 898 899 900 901 |
if (entry.val) { if (add_swap_count_continuation(entry, GFP_KERNEL) < 0) return -ENOMEM; progress = 0; } |
1da177e4c Linux-2.6.12-rc2 |
902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 |
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); |
71e3aac07 thp: transparent ... |
920 921 |
if (pmd_trans_huge(*src_pmd)) { int err; |
14d1a55cd thp: add debug ch... |
922 |
VM_BUG_ON(next-addr != HPAGE_PMD_SIZE); |
71e3aac07 thp: transparent ... |
923 924 925 926 927 928 929 930 |
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 Linux-2.6.12-rc2 |
931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 |
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 mm: move all mmu ... |
969 970 971 |
unsigned long mmun_start; /* For mmu_notifiers */ unsigned long mmun_end; /* For mmu_notifiers */ bool is_cow; |
cddb8a5c1 mmu-notifiers: core |
972 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
973 |
|
d992895ba [PATCH] Lazy page... |
974 975 976 977 978 979 |
/* * 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. */ |
4b6e1e370 mm: kill vma flag... |
980 981 |
if (!(vma->vm_flags & (VM_HUGETLB | VM_NONLINEAR | VM_PFNMAP | VM_MIXEDMAP))) { |
d992895ba [PATCH] Lazy page... |
982 983 984 |
if (!vma->anon_vma) return 0; } |
1da177e4c Linux-2.6.12-rc2 |
985 986 |
if (is_vm_hugetlb_page(vma)) return copy_hugetlb_page_range(dst_mm, src_mm, vma); |
b3b9c2932 mm, x86, pat: rew... |
987 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) { |
2ab640379 x86: PAT: hooks i... |
988 989 990 991 |
/* * We do not free on error cases below as remove_vma * gets called on error from higher level routine */ |
5180da410 x86, pat: separat... |
992 |
ret = track_pfn_copy(vma); |
2ab640379 x86: PAT: hooks i... |
993 994 995 |
if (ret) return ret; } |
cddb8a5c1 mmu-notifiers: core |
996 997 998 999 1000 1001 |
/* * 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 mm: move all mmu ... |
1002 1003 1004 1005 1006 1007 |
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 mmu-notifiers: core |
1008 1009 |
ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
1010 1011 1012 1013 1014 1015 |
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 mmu-notifiers: core |
1016 1017 1018 1019 1020 |
if (unlikely(copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd, vma, addr, next))) { ret = -ENOMEM; break; } |
1da177e4c Linux-2.6.12-rc2 |
1021 |
} while (dst_pgd++, src_pgd++, addr = next, addr != end); |
cddb8a5c1 mmu-notifiers: core |
1022 |
|
2ec74c3ef mm: move all mmu ... |
1023 1024 |
if (is_cow) mmu_notifier_invalidate_range_end(src_mm, mmun_start, mmun_end); |
cddb8a5c1 mmu-notifiers: core |
1025 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1026 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1027 |
static unsigned long zap_pte_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1028 |
struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c Linux-2.6.12-rc2 |
1029 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1030 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1031 |
{ |
b5810039a [PATCH] core remo... |
1032 |
struct mm_struct *mm = tlb->mm; |
d16dfc550 mm: mmu_gather re... |
1033 |
int force_flush = 0; |
d559db086 mm: clean up mm_c... |
1034 |
int rss[NR_MM_COUNTERS]; |
97a894136 mm: Remove i_mmap... |
1035 |
spinlock_t *ptl; |
5f1a19070 mm: fix wrong kun... |
1036 |
pte_t *start_pte; |
97a894136 mm: Remove i_mmap... |
1037 |
pte_t *pte; |
d559db086 mm: clean up mm_c... |
1038 |
|
d16dfc550 mm: mmu_gather re... |
1039 |
again: |
e303297e6 mm: extended batc... |
1040 |
init_rss_vec(rss); |
5f1a19070 mm: fix wrong kun... |
1041 1042 |
start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl); pte = start_pte; |
6606c3e0d [PATCH] paravirt:... |
1043 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
1044 1045 |
do { pte_t ptent = *pte; |
51c6f666f [PATCH] mm: ZAP_B... |
1046 |
if (pte_none(ptent)) { |
1da177e4c Linux-2.6.12-rc2 |
1047 |
continue; |
51c6f666f [PATCH] mm: ZAP_B... |
1048 |
} |
6f5e6b9e6 [PATCH] fix free ... |
1049 |
|
1da177e4c Linux-2.6.12-rc2 |
1050 |
if (pte_present(ptent)) { |
ee498ed73 [PATCH] unpaged: ... |
1051 |
struct page *page; |
51c6f666f [PATCH] mm: ZAP_B... |
1052 |
|
6aab341e0 mm: re-architect ... |
1053 |
page = vm_normal_page(vma, addr, ptent); |
1da177e4c Linux-2.6.12-rc2 |
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 |
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; /* * Each page->index must be checked when * invalidating or truncating nonlinear. */ if (details->nonlinear_vma && (page->index < details->first_index || page->index > details->last_index)) continue; } |
b5810039a [PATCH] core remo... |
1072 |
ptent = ptep_get_and_clear_full(mm, addr, pte, |
a600388d2 [PATCH] x86: ptep... |
1073 |
tlb->fullmm); |
1da177e4c Linux-2.6.12-rc2 |
1074 1075 1076 1077 1078 |
tlb_remove_tlb_entry(tlb, pte, addr); if (unlikely(!page)) continue; if (unlikely(details) && details->nonlinear_vma && linear_page_index(details->nonlinear_vma, |
41bb3476b mm: save soft-dir... |
1079 1080 1081 1082 1083 1084 |
addr) != page->index) { pte_t ptfile = pgoff_to_pte(page->index); if (pte_soft_dirty(ptent)) pte_file_mksoft_dirty(ptfile); set_pte_at(mm, addr, pte, ptfile); } |
1da177e4c Linux-2.6.12-rc2 |
1085 |
if (PageAnon(page)) |
d559db086 mm: clean up mm_c... |
1086 |
rss[MM_ANONPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1087 1088 1089 |
else { if (pte_dirty(ptent)) set_page_dirty(page); |
4917e5d04 mm: more likely r... |
1090 |
if (pte_young(ptent) && |
64363aad5 mm: remove unused... |
1091 |
likely(!(vma->vm_flags & VM_SEQ_READ))) |
bf3f3bc5e mm: don't mark_pa... |
1092 |
mark_page_accessed(page); |
d559db086 mm: clean up mm_c... |
1093 |
rss[MM_FILEPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1094 |
} |
edc315fd2 badpage: remove v... |
1095 |
page_remove_rmap(page); |
3dc147414 badpage: replace ... |
1096 1097 |
if (unlikely(page_mapcount(page) < 0)) print_bad_pte(vma, addr, ptent, page); |
d16dfc550 mm: mmu_gather re... |
1098 1099 1100 |
force_flush = !__tlb_remove_page(tlb, page); if (force_flush) break; |
1da177e4c Linux-2.6.12-rc2 |
1101 1102 1103 1104 1105 1106 1107 1108 |
continue; } /* * If details->check_mapping, we leave swap entries; * if details->nonlinear_vma, we leave file entries. */ if (unlikely(details)) continue; |
2509ef26d badpage: zap prin... |
1109 1110 1111 |
if (pte_file(ptent)) { if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) print_bad_pte(vma, addr, ptent, NULL); |
b084d4353 mm: count swap usage |
1112 1113 1114 1115 1116 |
} else { swp_entry_t entry = pte_to_swp_entry(ptent); if (!non_swap_entry(entry)) rss[MM_SWAPENTS]--; |
9f9f1acd7 mm: fix rss count... |
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 |
else if (is_migration_entry(entry)) { struct page *page; page = migration_entry_to_page(entry); if (PageAnon(page)) rss[MM_ANONPAGES]--; else rss[MM_FILEPAGES]--; } |
b084d4353 mm: count swap usage |
1127 1128 1129 |
if (unlikely(!free_swap_and_cache(entry))) print_bad_pte(vma, addr, ptent, NULL); } |
9888a1cae [PATCH] paravirt:... |
1130 |
pte_clear_not_present_full(mm, addr, pte, tlb->fullmm); |
97a894136 mm: Remove i_mmap... |
1131 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
ae8597623 [PATCH] mm: batch... |
1132 |
|
d559db086 mm: clean up mm_c... |
1133 |
add_mm_rss_vec(mm, rss); |
6606c3e0d [PATCH] paravirt:... |
1134 |
arch_leave_lazy_mmu_mode(); |
5f1a19070 mm: fix wrong kun... |
1135 |
pte_unmap_unlock(start_pte, ptl); |
51c6f666f [PATCH] mm: ZAP_B... |
1136 |
|
d16dfc550 mm: mmu_gather re... |
1137 1138 1139 1140 1141 1142 |
/* * mmu_gather ran out of room to batch pages, we break out of * the PTE lock to avoid doing the potential expensive TLB invalidate * and page-free while holding it. */ if (force_flush) { |
2b047252d Fix TLB gather vi... |
1143 |
unsigned long old_end; |
d16dfc550 mm: mmu_gather re... |
1144 |
force_flush = 0; |
597e1c358 mm/mmu_gather: en... |
1145 |
|
2b047252d Fix TLB gather vi... |
1146 1147 1148 1149 1150 1151 |
/* * Flush the TLB just for the previous segment, * then update the range to be the remaining * TLB range. */ old_end = tlb->end; |
e6c495a96 mm: fix the TLB r... |
1152 |
tlb->end = addr; |
2b047252d Fix TLB gather vi... |
1153 |
|
d16dfc550 mm: mmu_gather re... |
1154 |
tlb_flush_mmu(tlb); |
2b047252d Fix TLB gather vi... |
1155 1156 1157 1158 1159 |
tlb->start = addr; tlb->end = old_end; if (addr != end) |
d16dfc550 mm: mmu_gather re... |
1160 1161 |
goto again; } |
51c6f666f [PATCH] mm: ZAP_B... |
1162 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1163 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1164 |
static inline unsigned long zap_pmd_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1165 |
struct vm_area_struct *vma, pud_t *pud, |
1da177e4c Linux-2.6.12-rc2 |
1166 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1167 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1168 1169 1170 1171 1172 1173 1174 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
71e3aac07 thp: transparent ... |
1175 |
if (pmd_trans_huge(*pmd)) { |
1a5a9906d mm: thp: fix pmd_... |
1176 |
if (next - addr != HPAGE_PMD_SIZE) { |
e0897d75f mm, thp: print us... |
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 |
#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 |
e180377f1 thp: change split... |
1187 |
split_huge_page_pmd(vma, addr, pmd); |
f21760b15 thp: add tlb_remo... |
1188 |
} else if (zap_huge_pmd(tlb, vma, pmd, addr)) |
1a5a9906d mm: thp: fix pmd_... |
1189 |
goto next; |
71e3aac07 thp: transparent ... |
1190 1191 |
/* fall through */ } |
1a5a9906d mm: thp: fix pmd_... |
1192 1193 1194 1195 1196 1197 1198 1199 1200 |
/* * 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 mm: Remove i_mmap... |
1201 |
next = zap_pte_range(tlb, vma, pmd, addr, next, details); |
1a5a9906d mm: thp: fix pmd_... |
1202 |
next: |
97a894136 mm: Remove i_mmap... |
1203 1204 |
cond_resched(); } while (pmd++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1205 1206 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1207 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1208 |
static inline unsigned long zap_pud_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1209 |
struct vm_area_struct *vma, pgd_t *pgd, |
1da177e4c Linux-2.6.12-rc2 |
1210 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1211 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1212 1213 1214 1215 1216 1217 1218 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1219 |
if (pud_none_or_clear_bad(pud)) |
1da177e4c Linux-2.6.12-rc2 |
1220 |
continue; |
97a894136 mm: Remove i_mmap... |
1221 1222 |
next = zap_pmd_range(tlb, vma, pud, addr, next, details); } while (pud++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1223 1224 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1225 |
} |
038c7aa16 VM: unmap_page_ra... |
1226 1227 1228 1229 |
static void unmap_page_range(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long addr, unsigned long end, struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1230 1231 1232 1233 1234 1235 1236 1237 |
{ pgd_t *pgd; unsigned long next; if (details && !details->check_mapping && !details->nonlinear_vma) details = NULL; BUG_ON(addr >= end); |
569b846df memcg: coalesce u... |
1238 |
mem_cgroup_uncharge_start(); |
1da177e4c Linux-2.6.12-rc2 |
1239 1240 1241 1242 |
tlb_start_vma(tlb, vma); pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1243 |
if (pgd_none_or_clear_bad(pgd)) |
1da177e4c Linux-2.6.12-rc2 |
1244 |
continue; |
97a894136 mm: Remove i_mmap... |
1245 1246 |
next = zap_pud_range(tlb, vma, pgd, addr, next, details); } while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
1247 |
tlb_end_vma(tlb, vma); |
569b846df memcg: coalesce u... |
1248 |
mem_cgroup_uncharge_end(); |
1da177e4c Linux-2.6.12-rc2 |
1249 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1250 |
|
f5cc4eef9 VM: make zap_page... |
1251 1252 1253 |
static void unmap_single_vma(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e vm: remove 'nr_ac... |
1254 |
unsigned long end_addr, |
f5cc4eef9 VM: make zap_page... |
1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 |
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 uprobes/core: Dec... |
1265 1266 |
if (vma->vm_file) uprobe_munmap(vma, start, end); |
b3b9c2932 mm, x86, pat: rew... |
1267 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) |
5180da410 x86, pat: separat... |
1268 |
untrack_pfn(vma, 0, 0); |
f5cc4eef9 VM: make zap_page... |
1269 1270 1271 1272 1273 1274 1275 |
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 mm/memory.c: upda... |
1276 |
* cleanup path of mmap_region. When |
f5cc4eef9 VM: make zap_page... |
1277 |
* hugetlbfs ->mmap method fails, |
7aa6b4ad5 mm/memory.c: upda... |
1278 |
* mmap_region() nullifies vma->vm_file |
f5cc4eef9 VM: make zap_page... |
1279 1280 1281 1282 |
* before calling this function to clean up. * Since no pte has actually been setup, it is * safe to do nothing in this case. */ |
24669e584 hugetlb: use mmu_... |
1283 1284 |
if (vma->vm_file) { mutex_lock(&vma->vm_file->f_mapping->i_mmap_mutex); |
d833352a4 mm: hugetlbfs: cl... |
1285 |
__unmap_hugepage_range_final(tlb, vma, start, end, NULL); |
24669e584 hugetlb: use mmu_... |
1286 1287 |
mutex_unlock(&vma->vm_file->f_mapping->i_mmap_mutex); } |
f5cc4eef9 VM: make zap_page... |
1288 1289 1290 |
} else unmap_page_range(tlb, vma, start, end, details); } |
1da177e4c Linux-2.6.12-rc2 |
1291 |
} |
1da177e4c Linux-2.6.12-rc2 |
1292 1293 |
/** * unmap_vmas - unmap a range of memory covered by a list of vma's |
0164f69d0 mm/memory.c: fix ... |
1294 |
* @tlb: address of the caller's struct mmu_gather |
1da177e4c Linux-2.6.12-rc2 |
1295 1296 1297 |
* @vma: the starting vma * @start_addr: virtual address at which to start unmapping * @end_addr: virtual address at which to end unmapping |
1da177e4c Linux-2.6.12-rc2 |
1298 |
* |
508034a32 [PATCH] mm: unmap... |
1299 |
* Unmap all pages in the vma list. |
1da177e4c Linux-2.6.12-rc2 |
1300 |
* |
1da177e4c Linux-2.6.12-rc2 |
1301 1302 1303 1304 1305 1306 1307 1308 1309 |
* 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 VM: make unmap_vm... |
1310 |
void unmap_vmas(struct mmu_gather *tlb, |
1da177e4c Linux-2.6.12-rc2 |
1311 |
struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e vm: remove 'nr_ac... |
1312 |
unsigned long end_addr) |
1da177e4c Linux-2.6.12-rc2 |
1313 |
{ |
cddb8a5c1 mmu-notifiers: core |
1314 |
struct mm_struct *mm = vma->vm_mm; |
1da177e4c Linux-2.6.12-rc2 |
1315 |
|
cddb8a5c1 mmu-notifiers: core |
1316 |
mmu_notifier_invalidate_range_start(mm, start_addr, end_addr); |
f5cc4eef9 VM: make zap_page... |
1317 |
for ( ; vma && vma->vm_start < end_addr; vma = vma->vm_next) |
4f74d2c8e vm: remove 'nr_ac... |
1318 |
unmap_single_vma(tlb, vma, start_addr, end_addr, NULL); |
cddb8a5c1 mmu-notifiers: core |
1319 |
mmu_notifier_invalidate_range_end(mm, start_addr, end_addr); |
1da177e4c Linux-2.6.12-rc2 |
1320 1321 1322 1323 1324 |
} /** * zap_page_range - remove user pages in a given range * @vma: vm_area_struct holding the applicable pages |
eb4546bbb mm/memory.c: fix ... |
1325 |
* @start: starting address of pages to zap |
1da177e4c Linux-2.6.12-rc2 |
1326 1327 |
* @size: number of bytes to zap * @details: details of nonlinear truncation or shared cache invalidation |
f5cc4eef9 VM: make zap_page... |
1328 1329 |
* * Caller must protect the VMA list |
1da177e4c Linux-2.6.12-rc2 |
1330 |
*/ |
7e027b14d vm: simplify unma... |
1331 |
void zap_page_range(struct vm_area_struct *vma, unsigned long start, |
1da177e4c Linux-2.6.12-rc2 |
1332 1333 1334 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550 mm: mmu_gather re... |
1335 |
struct mmu_gather tlb; |
7e027b14d vm: simplify unma... |
1336 |
unsigned long end = start + size; |
1da177e4c Linux-2.6.12-rc2 |
1337 |
|
1da177e4c Linux-2.6.12-rc2 |
1338 |
lru_add_drain(); |
2b047252d Fix TLB gather vi... |
1339 |
tlb_gather_mmu(&tlb, mm, start, end); |
365e9c87a [PATCH] mm: updat... |
1340 |
update_hiwater_rss(mm); |
7e027b14d vm: simplify unma... |
1341 1342 |
mmu_notifier_invalidate_range_start(mm, start, end); for ( ; vma && vma->vm_start < end; vma = vma->vm_next) |
4f74d2c8e vm: remove 'nr_ac... |
1343 |
unmap_single_vma(&tlb, vma, start, end, details); |
7e027b14d vm: simplify unma... |
1344 1345 |
mmu_notifier_invalidate_range_end(mm, start, end); tlb_finish_mmu(&tlb, start, end); |
1da177e4c Linux-2.6.12-rc2 |
1346 |
} |
c627f9cc0 mm: add zap_vma_p... |
1347 |
/** |
f5cc4eef9 VM: make zap_page... |
1348 1349 1350 1351 1352 1353 1354 |
* 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 * @details: details of nonlinear truncation or shared cache invalidation * * The range must fit into one VMA. |
1da177e4c Linux-2.6.12-rc2 |
1355 |
*/ |
f5cc4eef9 VM: make zap_page... |
1356 |
static void zap_page_range_single(struct vm_area_struct *vma, unsigned long address, |
1da177e4c Linux-2.6.12-rc2 |
1357 1358 1359 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550 mm: mmu_gather re... |
1360 |
struct mmu_gather tlb; |
1da177e4c Linux-2.6.12-rc2 |
1361 |
unsigned long end = address + size; |
1da177e4c Linux-2.6.12-rc2 |
1362 |
|
1da177e4c Linux-2.6.12-rc2 |
1363 |
lru_add_drain(); |
2b047252d Fix TLB gather vi... |
1364 |
tlb_gather_mmu(&tlb, mm, address, end); |
365e9c87a [PATCH] mm: updat... |
1365 |
update_hiwater_rss(mm); |
f5cc4eef9 VM: make zap_page... |
1366 |
mmu_notifier_invalidate_range_start(mm, address, end); |
4f74d2c8e vm: remove 'nr_ac... |
1367 |
unmap_single_vma(&tlb, vma, address, end, details); |
f5cc4eef9 VM: make zap_page... |
1368 |
mmu_notifier_invalidate_range_end(mm, address, end); |
d16dfc550 mm: mmu_gather re... |
1369 |
tlb_finish_mmu(&tlb, address, end); |
1da177e4c Linux-2.6.12-rc2 |
1370 |
} |
c627f9cc0 mm: add zap_vma_p... |
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 |
/** * 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 VM: make zap_page... |
1389 |
zap_page_range_single(vma, address, size, NULL); |
c627f9cc0 mm: add zap_vma_p... |
1390 1391 1392 |
return 0; } EXPORT_SYMBOL_GPL(zap_vma_ptes); |
142762bd8 mm: document foll... |
1393 |
/** |
240aadeed mm: accelerate mm... |
1394 |
* follow_page_mask - look up a page descriptor from a user-virtual address |
142762bd8 mm: document foll... |
1395 1396 1397 |
* @vma: vm_area_struct mapping @address * @address: virtual address to look up * @flags: flags modifying lookup behaviour |
240aadeed mm: accelerate mm... |
1398 |
* @page_mask: on output, *page_mask is set according to the size of the page |
142762bd8 mm: document foll... |
1399 1400 1401 1402 1403 1404 |
* * @flags can have FOLL_ flags set, defined in <linux/mm.h> * * Returns the mapped (struct page *), %NULL if no mapping exists, or * an error pointer if there is a mapping to something not represented * by a page descriptor (see also vm_normal_page()). |
1da177e4c Linux-2.6.12-rc2 |
1405 |
*/ |
240aadeed mm: accelerate mm... |
1406 1407 1408 |
struct page *follow_page_mask(struct vm_area_struct *vma, unsigned long address, unsigned int flags, unsigned int *page_mask) |
1da177e4c Linux-2.6.12-rc2 |
1409 1410 1411 1412 1413 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; |
deceb6cd1 [PATCH] mm: follo... |
1414 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
1415 |
struct page *page; |
6aab341e0 mm: re-architect ... |
1416 |
struct mm_struct *mm = vma->vm_mm; |
1da177e4c Linux-2.6.12-rc2 |
1417 |
|
240aadeed mm: accelerate mm... |
1418 |
*page_mask = 0; |
deceb6cd1 [PATCH] mm: follo... |
1419 1420 1421 1422 1423 |
page = follow_huge_addr(mm, address, flags & FOLL_WRITE); if (!IS_ERR(page)) { BUG_ON(flags & FOLL_GET); goto out; } |
1da177e4c Linux-2.6.12-rc2 |
1424 |
|
deceb6cd1 [PATCH] mm: follo... |
1425 |
page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1426 1427 |
pgd = pgd_offset(mm, address); if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) |
deceb6cd1 [PATCH] mm: follo... |
1428 |
goto no_page_table; |
1da177e4c Linux-2.6.12-rc2 |
1429 1430 |
pud = pud_offset(pgd, address); |
ceb868796 hugetlb: introduc... |
1431 |
if (pud_none(*pud)) |
deceb6cd1 [PATCH] mm: follo... |
1432 |
goto no_page_table; |
8a07651ee thp: transparent ... |
1433 |
if (pud_huge(*pud) && vma->vm_flags & VM_HUGETLB) { |
e632a938d mm: migrate: add ... |
1434 1435 |
if (flags & FOLL_GET) goto out; |
ceb868796 hugetlb: introduc... |
1436 1437 1438 1439 1440 |
page = follow_huge_pud(mm, address, pud, flags & FOLL_WRITE); goto out; } if (unlikely(pud_bad(*pud))) goto no_page_table; |
1da177e4c Linux-2.6.12-rc2 |
1441 |
pmd = pmd_offset(pud, address); |
aeed5fce3 x86: fix PAE pmd_... |
1442 |
if (pmd_none(*pmd)) |
deceb6cd1 [PATCH] mm: follo... |
1443 |
goto no_page_table; |
71e3aac07 thp: transparent ... |
1444 |
if (pmd_huge(*pmd) && vma->vm_flags & VM_HUGETLB) { |
deceb6cd1 [PATCH] mm: follo... |
1445 |
page = follow_huge_pmd(mm, address, pmd, flags & FOLL_WRITE); |
e632a938d mm: migrate: add ... |
1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 |
if (flags & FOLL_GET) { /* * Refcount on tail pages are not well-defined and * shouldn't be taken. The caller should handle a NULL * return when trying to follow tail pages. */ if (PageHead(page)) get_page(page); else { page = NULL; goto out; } } |
1da177e4c Linux-2.6.12-rc2 |
1459 |
goto out; |
deceb6cd1 [PATCH] mm: follo... |
1460 |
} |
0b9d70529 mm: numa: Support... |
1461 1462 |
if ((flags & FOLL_NUMA) && pmd_numa(*pmd)) goto no_page_table; |
71e3aac07 thp: transparent ... |
1463 |
if (pmd_trans_huge(*pmd)) { |
500d65d47 thp: pmd_trans_hu... |
1464 |
if (flags & FOLL_SPLIT) { |
e180377f1 thp: change split... |
1465 |
split_huge_page_pmd(vma, address, pmd); |
500d65d47 thp: pmd_trans_hu... |
1466 1467 |
goto split_fallthrough; } |
c4088ebdc mm: convert the r... |
1468 |
ptl = pmd_lock(mm, pmd); |
71e3aac07 thp: transparent ... |
1469 1470 |
if (likely(pmd_trans_huge(*pmd))) { if (unlikely(pmd_trans_splitting(*pmd))) { |
c4088ebdc mm: convert the r... |
1471 |
spin_unlock(ptl); |
71e3aac07 thp: transparent ... |
1472 1473 |
wait_split_huge_page(vma->anon_vma, pmd); } else { |
b676b293f mm, thp: fix mapp... |
1474 |
page = follow_trans_huge_pmd(vma, address, |
71e3aac07 thp: transparent ... |
1475 |
pmd, flags); |
c4088ebdc mm: convert the r... |
1476 |
spin_unlock(ptl); |
240aadeed mm: accelerate mm... |
1477 |
*page_mask = HPAGE_PMD_NR - 1; |
71e3aac07 thp: transparent ... |
1478 1479 1480 |
goto out; } } else |
c4088ebdc mm: convert the r... |
1481 |
spin_unlock(ptl); |
71e3aac07 thp: transparent ... |
1482 1483 |
/* fall through */ } |
500d65d47 thp: pmd_trans_hu... |
1484 |
split_fallthrough: |
aeed5fce3 x86: fix PAE pmd_... |
1485 1486 |
if (unlikely(pmd_bad(*pmd))) goto no_page_table; |
deceb6cd1 [PATCH] mm: follo... |
1487 |
ptep = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
1488 1489 |
pte = *ptep; |
5117b3b83 mm,ksm: FOLL_MIGR... |
1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 |
if (!pte_present(pte)) { swp_entry_t entry; /* * KSM's break_ksm() relies upon recognizing a ksm page * even while it is being migrated, so for that case we * need migration_entry_wait(). */ if (likely(!(flags & FOLL_MIGRATION))) goto no_page; if (pte_none(pte) || pte_file(pte)) goto no_page; entry = pte_to_swp_entry(pte); if (!is_migration_entry(entry)) goto no_page; pte_unmap_unlock(ptep, ptl); migration_entry_wait(mm, pmd, address); goto split_fallthrough; } |
0b9d70529 mm: numa: Support... |
1508 1509 |
if ((flags & FOLL_NUMA) && pte_numa(pte)) goto no_page; |
deceb6cd1 [PATCH] mm: follo... |
1510 1511 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
a13ea5b75 mm: reinstate ZER... |
1512 |
|
6aab341e0 mm: re-architect ... |
1513 |
page = vm_normal_page(vma, address, pte); |
a13ea5b75 mm: reinstate ZER... |
1514 1515 |
if (unlikely(!page)) { if ((flags & FOLL_DUMP) || |
62eede62d mm: ZERO_PAGE wit... |
1516 |
!is_zero_pfn(pte_pfn(pte))) |
a13ea5b75 mm: reinstate ZER... |
1517 1518 1519 |
goto bad_page; page = pte_page(pte); } |
1da177e4c Linux-2.6.12-rc2 |
1520 |
|
deceb6cd1 [PATCH] mm: follo... |
1521 |
if (flags & FOLL_GET) |
70b50f94f mm: thp: tail pag... |
1522 |
get_page_foll(page); |
deceb6cd1 [PATCH] mm: follo... |
1523 1524 1525 1526 |
if (flags & FOLL_TOUCH) { if ((flags & FOLL_WRITE) && !pte_dirty(pte) && !PageDirty(page)) set_page_dirty(page); |
bd775c42e mm: add comment w... |
1527 1528 1529 1530 1531 |
/* * pte_mkyoung() would be more correct here, but atomic care * is needed to avoid losing the dirty bit: it is easier to use * mark_page_accessed(). */ |
deceb6cd1 [PATCH] mm: follo... |
1532 1533 |
mark_page_accessed(page); } |
a1fde08c7 VM: skip the stac... |
1534 |
if ((flags & FOLL_MLOCK) && (vma->vm_flags & VM_LOCKED)) { |
110d74a92 mm: add FOLL_MLOC... |
1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 |
/* * The preliminary mapping check is mainly to avoid the * pointless overhead of lock_page on the ZERO_PAGE * which might bounce very badly if there is contention. * * If the page is already locked, we don't need to * handle it now - vmscan will handle it later if and * when it attempts to reclaim the page. */ if (page->mapping && trylock_page(page)) { lru_add_drain(); /* push cached pages to LRU */ /* |
e6c509f85 mm: use clear_pag... |
1547 1548 1549 1550 |
* Because we lock page here, and migration is * blocked by the pte's page reference, and we * know the page is still mapped, we don't even * need to check for file-cache page truncation. |
110d74a92 mm: add FOLL_MLOC... |
1551 |
*/ |
e6c509f85 mm: use clear_pag... |
1552 |
mlock_vma_page(page); |
110d74a92 mm: add FOLL_MLOC... |
1553 1554 1555 |
unlock_page(page); } } |
deceb6cd1 [PATCH] mm: follo... |
1556 1557 |
unlock: pte_unmap_unlock(ptep, ptl); |
1da177e4c Linux-2.6.12-rc2 |
1558 |
out: |
deceb6cd1 [PATCH] mm: follo... |
1559 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1560 |
|
89f5b7da2 Reinstate ZERO_PA... |
1561 1562 1563 1564 1565 1566 1567 1568 |
bad_page: pte_unmap_unlock(ptep, ptl); return ERR_PTR(-EFAULT); no_page: pte_unmap_unlock(ptep, ptl); if (!pte_none(pte)) return page; |
8e4b9a607 mm: FOLL_DUMP rep... |
1569 |
|
deceb6cd1 [PATCH] mm: follo... |
1570 1571 1572 |
no_page_table: /* * When core dumping an enormous anonymous area that nobody |
8e4b9a607 mm: FOLL_DUMP rep... |
1573 1574 1575 1576 1577 |
* has touched so far, we don't want to allocate unnecessary pages or * page tables. Return error instead of NULL to skip handle_mm_fault, * then get_dump_page() will return NULL to leave a hole in the dump. * But we can only make this optimization where a hole would surely * be zero-filled if handle_mm_fault() actually did handle it. |
deceb6cd1 [PATCH] mm: follo... |
1578 |
*/ |
8e4b9a607 mm: FOLL_DUMP rep... |
1579 1580 1581 |
if ((flags & FOLL_DUMP) && (!vma->vm_ops || !vma->vm_ops->fault)) return ERR_PTR(-EFAULT); |
deceb6cd1 [PATCH] mm: follo... |
1582 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1583 |
} |
95042f9eb vm: fix mlock() o... |
1584 1585 |
static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr) { |
a09a79f66 Don't lock guardp... |
1586 1587 |
return stack_guard_page_start(vma, addr) || stack_guard_page_end(vma, addr+PAGE_SIZE); |
95042f9eb vm: fix mlock() o... |
1588 |
} |
0014bd990 mm: export __get_... |
1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 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 1636 1637 |
/** * __get_user_pages() - pin user pages in memory * @tsk: task_struct of target task * @mm: mm_struct of target mm * @start: starting user address * @nr_pages: number of pages from start to pin * @gup_flags: flags modifying pin behaviour * @pages: array that receives pointers to the pages pinned. * Should be at least nr_pages long. Or NULL, if caller * only intends to ensure the pages are faulted in. * @vmas: array of pointers to vmas corresponding to each page. * Or NULL if the caller does not require them. * @nonblocking: whether waiting for disk IO or mmap_sem contention * * Returns number of pages pinned. This may be fewer than the number * requested. If nr_pages is 0 or negative, returns 0. If no pages * were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. vmas will only * remain valid while mmap_sem is held. * * Must be called with mmap_sem held for read or write. * * __get_user_pages walks a process's page tables and takes a reference to * each struct page that each user address corresponds to at a given * instant. That is, it takes the page that would be accessed if a user * thread accesses the given user virtual address at that instant. * * This does not guarantee that the page exists in the user mappings when * __get_user_pages returns, and there may even be a completely different * page there in some cases (eg. if mmapped pagecache has been invalidated * and subsequently re faulted). However it does guarantee that the page * won't be freed completely. And mostly callers simply care that the page * contains data that was valid *at some point in time*. Typically, an IO * or similar operation cannot guarantee anything stronger anyway because * locks can't be held over the syscall boundary. * * If @gup_flags & FOLL_WRITE == 0, the page must not be written to. If * the page is written to, set_page_dirty (or set_page_dirty_lock, as * appropriate) must be called after the page is finished with, and * before put_page is called. * * If @nonblocking != NULL, __get_user_pages will not wait for disk IO * or mmap_sem contention, and if waiting is needed to pin all pages, * *@nonblocking will be set to 0. * * In most cases, get_user_pages or get_user_pages_fast should be used * instead of __get_user_pages. __get_user_pages should be used only if * you need some special @gup_flags. */ |
28a35716d mm: use long type... |
1638 1639 1640 1641 |
long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, struct vm_area_struct **vmas, int *nonblocking) |
1da177e4c Linux-2.6.12-rc2 |
1642 |
{ |
28a35716d mm: use long type... |
1643 |
long i; |
58fa879e1 mm: FOLL flags fo... |
1644 |
unsigned long vm_flags; |
240aadeed mm: accelerate mm... |
1645 |
unsigned int page_mask; |
1da177e4c Linux-2.6.12-rc2 |
1646 |
|
28a35716d mm: use long type... |
1647 |
if (!nr_pages) |
900cf086f Be more robust ab... |
1648 |
return 0; |
58fa879e1 mm: FOLL flags fo... |
1649 1650 |
VM_BUG_ON(!!pages != !!(gup_flags & FOLL_GET)); |
0b9d70529 mm: numa: Support... |
1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 |
/* * If FOLL_FORCE and FOLL_NUMA are both set, handle_mm_fault * would be called on PROT_NONE ranges. We must never invoke * handle_mm_fault on PROT_NONE ranges or the NUMA hinting * page faults would unprotect the PROT_NONE ranges if * _PAGE_NUMA and _PAGE_PROTNONE are sharing the same pte/pmd * bitflag. So to avoid that, don't set FOLL_NUMA if * FOLL_FORCE is set. */ if (!(gup_flags & FOLL_FORCE)) gup_flags |= FOLL_NUMA; |
1da177e4c Linux-2.6.12-rc2 |
1662 1663 1664 |
i = 0; do { |
deceb6cd1 [PATCH] mm: follo... |
1665 |
struct vm_area_struct *vma; |
1da177e4c Linux-2.6.12-rc2 |
1666 1667 |
vma = find_extend_vma(mm, start); |
e7f22e207 mm: use mm_struct... |
1668 |
if (!vma && in_gate_area(mm, start)) { |
1da177e4c Linux-2.6.12-rc2 |
1669 |
unsigned long pg = start & PAGE_MASK; |
1da177e4c Linux-2.6.12-rc2 |
1670 1671 1672 1673 |
pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; |
b291f0003 mlock: mlocked pa... |
1674 1675 |
/* user gate pages are read-only */ |
58fa879e1 mm: FOLL flags fo... |
1676 |
if (gup_flags & FOLL_WRITE) |
cda540ace mm: get_user_page... |
1677 |
goto efault; |
1da177e4c Linux-2.6.12-rc2 |
1678 1679 1680 1681 1682 1683 1684 1685 |
if (pg > TASK_SIZE) pgd = pgd_offset_k(pg); else pgd = pgd_offset_gate(mm, pg); BUG_ON(pgd_none(*pgd)); pud = pud_offset(pgd, pg); BUG_ON(pud_none(*pud)); pmd = pmd_offset(pud, pg); |
690dbe1ce [PATCH] x86_64: a... |
1686 |
if (pmd_none(*pmd)) |
cda540ace mm: get_user_page... |
1687 |
goto efault; |
f66055ab6 thp: verify pmd_t... |
1688 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c Linux-2.6.12-rc2 |
1689 |
pte = pte_offset_map(pmd, pg); |
690dbe1ce [PATCH] x86_64: a... |
1690 1691 |
if (pte_none(*pte)) { pte_unmap(pte); |
cda540ace mm: get_user_page... |
1692 |
goto efault; |
690dbe1ce [PATCH] x86_64: a... |
1693 |
} |
95042f9eb vm: fix mlock() o... |
1694 |
vma = get_gate_vma(mm); |
1da177e4c Linux-2.6.12-rc2 |
1695 |
if (pages) { |
de51257aa mm: fix ia64 cras... |
1696 |
struct page *page; |
95042f9eb vm: fix mlock() o... |
1697 |
page = vm_normal_page(vma, start, *pte); |
de51257aa mm: fix ia64 cras... |
1698 1699 1700 1701 1702 1703 |
if (!page) { if (!(gup_flags & FOLL_DUMP) && is_zero_pfn(pte_pfn(*pte))) page = pte_page(*pte); else { pte_unmap(pte); |
cda540ace mm: get_user_page... |
1704 |
goto efault; |
de51257aa mm: fix ia64 cras... |
1705 1706 |
} } |
6aab341e0 mm: re-architect ... |
1707 |
pages[i] = page; |
de51257aa mm: fix ia64 cras... |
1708 |
get_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1709 1710 |
} pte_unmap(pte); |
240aadeed mm: accelerate mm... |
1711 |
page_mask = 0; |
95042f9eb vm: fix mlock() o... |
1712 |
goto next_page; |
1da177e4c Linux-2.6.12-rc2 |
1713 |
} |
cda540ace mm: get_user_page... |
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 1749 |
if (!vma) goto efault; vm_flags = vma->vm_flags; if (vm_flags & (VM_IO | VM_PFNMAP)) goto efault; if (gup_flags & FOLL_WRITE) { if (!(vm_flags & VM_WRITE)) { if (!(gup_flags & FOLL_FORCE)) goto efault; /* * We used to let the write,force case do COW * in a VM_MAYWRITE VM_SHARED !VM_WRITE vma, so * ptrace could set a breakpoint in a read-only * mapping of an executable, without corrupting * the file (yet only when that file had been * opened for writing!). Anon pages in shared * mappings are surprising: now just reject it. */ if (!is_cow_mapping(vm_flags)) { WARN_ON_ONCE(vm_flags & VM_MAYWRITE); goto efault; } } } else { if (!(vm_flags & VM_READ)) { if (!(gup_flags & FOLL_FORCE)) goto efault; /* * Is there actually any vma we can reach here * which does not have VM_MAYREAD set? */ if (!(vm_flags & VM_MAYREAD)) goto efault; } } |
1da177e4c Linux-2.6.12-rc2 |
1750 |
|
2a15efc95 mm: follow_hugetl... |
1751 1752 |
if (is_vm_hugetlb_page(vma)) { i = follow_hugetlb_page(mm, vma, pages, vmas, |
58fa879e1 mm: FOLL flags fo... |
1753 |
&start, &nr_pages, i, gup_flags); |
2a15efc95 mm: follow_hugetl... |
1754 1755 |
continue; } |
deceb6cd1 [PATCH] mm: follo... |
1756 |
|
1da177e4c Linux-2.6.12-rc2 |
1757 |
do { |
08ef47293 [PATCH] get_user_... |
1758 |
struct page *page; |
58fa879e1 mm: FOLL flags fo... |
1759 |
unsigned int foll_flags = gup_flags; |
240aadeed mm: accelerate mm... |
1760 |
unsigned int page_increm; |
1da177e4c Linux-2.6.12-rc2 |
1761 |
|
462e00cc7 oom: stop allocat... |
1762 |
/* |
4779280d1 mm: make get_user... |
1763 |
* If we have a pending SIGKILL, don't keep faulting |
1c3aff1ce mm: remove unused... |
1764 |
* pages and potentially allocating memory. |
462e00cc7 oom: stop allocat... |
1765 |
*/ |
1c3aff1ce mm: remove unused... |
1766 |
if (unlikely(fatal_signal_pending(current))) |
4779280d1 mm: make get_user... |
1767 |
return i ? i : -ERESTARTSYS; |
462e00cc7 oom: stop allocat... |
1768 |
|
deceb6cd1 [PATCH] mm: follo... |
1769 |
cond_resched(); |
240aadeed mm: accelerate mm... |
1770 1771 |
while (!(page = follow_page_mask(vma, start, foll_flags, &page_mask))) { |
deceb6cd1 [PATCH] mm: follo... |
1772 |
int ret; |
53a7706d5 mlock: do not hol... |
1773 |
unsigned int fault_flags = 0; |
a09a79f66 Don't lock guardp... |
1774 1775 1776 1777 1778 |
/* For mlock, just skip the stack guard page. */ if (foll_flags & FOLL_MLOCK) { if (stack_guard_page(vma, start)) goto next_page; } |
53a7706d5 mlock: do not hol... |
1779 1780 1781 1782 |
if (foll_flags & FOLL_WRITE) fault_flags |= FAULT_FLAG_WRITE; if (nonblocking) fault_flags |= FAULT_FLAG_ALLOW_RETRY; |
318b275fb mm: allow GUP to ... |
1783 1784 |
if (foll_flags & FOLL_NOWAIT) fault_flags |= (FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT); |
d06063cc2 Move FAULT_FLAG_x... |
1785 |
|
d26ed650d mm: don't rely on... |
1786 |
ret = handle_mm_fault(mm, vma, start, |
53a7706d5 mlock: do not hol... |
1787 |
fault_flags); |
d26ed650d mm: don't rely on... |
1788 |
|
83c54070e mm: fault feedbac... |
1789 1790 1791 |
if (ret & VM_FAULT_ERROR) { if (ret & VM_FAULT_OOM) return i ? i : -ENOMEM; |
69ebb83e1 mm: make __get_us... |
1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 |
if (ret & (VM_FAULT_HWPOISON | VM_FAULT_HWPOISON_LARGE)) { if (i) return i; else if (gup_flags & FOLL_HWPOISON) return -EHWPOISON; else return -EFAULT; } if (ret & VM_FAULT_SIGBUS) |
cda540ace mm: get_user_page... |
1802 |
goto efault; |
83c54070e mm: fault feedbac... |
1803 1804 |
BUG(); } |
e7f22e207 mm: use mm_struct... |
1805 1806 1807 1808 1809 1810 1811 |
if (tsk) { if (ret & VM_FAULT_MAJOR) tsk->maj_flt++; else tsk->min_flt++; } |
83c54070e mm: fault feedbac... |
1812 |
|
53a7706d5 mlock: do not hol... |
1813 |
if (ret & VM_FAULT_RETRY) { |
318b275fb mm: allow GUP to ... |
1814 1815 |
if (nonblocking) *nonblocking = 0; |
53a7706d5 mlock: do not hol... |
1816 1817 |
return i; } |
a68d2ebc1 Fix up recent get... |
1818 |
/* |
83c54070e mm: fault feedbac... |
1819 1820 1821 1822 |
* The VM_FAULT_WRITE bit tells us that * do_wp_page has broken COW when necessary, * even if maybe_mkwrite decided not to set * pte_write. We can thus safely do subsequent |
878b63ac8 mm: gup persist f... |
1823 1824 1825 1826 1827 1828 |
* page lookups as if they were reads. But only * do so when looping for pte_write is futile: * in some cases userspace may also be wanting * to write to the gotten user page, which a * read fault here might prevent (a readonly * page might get reCOWed by userspace write). |
a68d2ebc1 Fix up recent get... |
1829 |
*/ |
878b63ac8 mm: gup persist f... |
1830 1831 |
if ((ret & VM_FAULT_WRITE) && !(vma->vm_flags & VM_WRITE)) |
deceb6cd1 [PATCH] mm: follo... |
1832 |
foll_flags &= ~FOLL_WRITE; |
83c54070e mm: fault feedbac... |
1833 |
|
7f7bbbe50 [PATCH] page faul... |
1834 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1835 |
} |
89f5b7da2 Reinstate ZERO_PA... |
1836 1837 |
if (IS_ERR(page)) return i ? i : PTR_ERR(page); |
1da177e4c Linux-2.6.12-rc2 |
1838 |
if (pages) { |
08ef47293 [PATCH] get_user_... |
1839 |
pages[i] = page; |
03beb0766 [PATCH] Add API f... |
1840 |
|
a6f36be32 [ARM] pass vma fo... |
1841 |
flush_anon_page(vma, page, start); |
08ef47293 [PATCH] get_user_... |
1842 |
flush_dcache_page(page); |
240aadeed mm: accelerate mm... |
1843 |
page_mask = 0; |
1da177e4c Linux-2.6.12-rc2 |
1844 |
} |
95042f9eb vm: fix mlock() o... |
1845 |
next_page: |
240aadeed mm: accelerate mm... |
1846 |
if (vmas) { |
1da177e4c Linux-2.6.12-rc2 |
1847 |
vmas[i] = vma; |
240aadeed mm: accelerate mm... |
1848 1849 1850 1851 1852 1853 1854 1855 |
page_mask = 0; } page_increm = 1 + (~(start >> PAGE_SHIFT) & page_mask); if (page_increm > nr_pages) page_increm = nr_pages; i += page_increm; start += page_increm * PAGE_SIZE; nr_pages -= page_increm; |
9d73777e5 clarify get_user_... |
1856 1857 |
} while (nr_pages && start < vma->vm_end); } while (nr_pages); |
1da177e4c Linux-2.6.12-rc2 |
1858 |
return i; |
cda540ace mm: get_user_page... |
1859 1860 |
efault: return i ? : -EFAULT; |
1da177e4c Linux-2.6.12-rc2 |
1861 |
} |
0014bd990 mm: export __get_... |
1862 |
EXPORT_SYMBOL(__get_user_pages); |
b291f0003 mlock: mlocked pa... |
1863 |
|
2efaca927 mm/futex: fix fut... |
1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 |
/* * fixup_user_fault() - manually resolve a user page fault * @tsk: the task_struct to use for page fault accounting, or * NULL if faults are not to be recorded. * @mm: mm_struct of target mm * @address: user address * @fault_flags:flags to pass down to handle_mm_fault() * * This is meant to be called in the specific scenario where for locking reasons * we try to access user memory in atomic context (within a pagefault_disable() * section), this returns -EFAULT, and we want to resolve the user fault before * trying again. * * Typically this is meant to be used by the futex code. * * The main difference with get_user_pages() is that this function will * unconditionally call handle_mm_fault() which will in turn perform all the * necessary SW fixup of the dirty and young bits in the PTE, while * handle_mm_fault() only guarantees to update these in the struct page. * * This is important for some architectures where those bits also gate the * access permission to the page because they are maintained in software. On * such architectures, gup() will not be enough to make a subsequent access * succeed. * * This should be called with the mm_sem held for read. */ int fixup_user_fault(struct task_struct *tsk, struct mm_struct *mm, unsigned long address, unsigned int fault_flags) { struct vm_area_struct *vma; int ret; vma = find_extend_vma(mm, address); if (!vma || address < vma->vm_start) return -EFAULT; ret = handle_mm_fault(mm, vma, address, fault_flags); if (ret & VM_FAULT_ERROR) { if (ret & VM_FAULT_OOM) return -ENOMEM; if (ret & (VM_FAULT_HWPOISON | VM_FAULT_HWPOISON_LARGE)) return -EHWPOISON; if (ret & VM_FAULT_SIGBUS) return -EFAULT; BUG(); } if (tsk) { if (ret & VM_FAULT_MAJOR) tsk->maj_flt++; else tsk->min_flt++; } return 0; } /* |
d2bf6be8a mm: clean up get_... |
1921 |
* get_user_pages() - pin user pages in memory |
e7f22e207 mm: use mm_struct... |
1922 1923 |
* @tsk: the task_struct to use for page fault accounting, or * NULL if faults are not to be recorded. |
d2bf6be8a mm: clean up get_... |
1924 1925 |
* @mm: mm_struct of target mm * @start: starting user address |
9d73777e5 clarify get_user_... |
1926 |
* @nr_pages: number of pages from start to pin |
d2bf6be8a mm: clean up get_... |
1927 |
* @write: whether pages will be written to by the caller |
cda540ace mm: get_user_page... |
1928 1929 |
* @force: whether to force access even when user mapping is currently * protected (but never forces write access to shared mapping). |
d2bf6be8a mm: clean up get_... |
1930 1931 1932 1933 1934 1935 1936 |
* @pages: array that receives pointers to the pages pinned. * Should be at least nr_pages long. Or NULL, if caller * only intends to ensure the pages are faulted in. * @vmas: array of pointers to vmas corresponding to each page. * Or NULL if the caller does not require them. * * Returns number of pages pinned. This may be fewer than the number |
9d73777e5 clarify get_user_... |
1937 |
* requested. If nr_pages is 0 or negative, returns 0. If no pages |
d2bf6be8a mm: clean up get_... |
1938 1939 1940 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 |
* were pinned, returns -errno. Each page returned must be released * with a put_page() call when it is finished with. vmas will only * remain valid while mmap_sem is held. * * Must be called with mmap_sem held for read or write. * * get_user_pages walks a process's page tables and takes a reference to * each struct page that each user address corresponds to at a given * instant. That is, it takes the page that would be accessed if a user * thread accesses the given user virtual address at that instant. * * This does not guarantee that the page exists in the user mappings when * get_user_pages returns, and there may even be a completely different * page there in some cases (eg. if mmapped pagecache has been invalidated * and subsequently re faulted). However it does guarantee that the page * won't be freed completely. And mostly callers simply care that the page * contains data that was valid *at some point in time*. Typically, an IO * or similar operation cannot guarantee anything stronger anyway because * locks can't be held over the syscall boundary. * * If write=0, the page must not be written to. If the page is written to, * set_page_dirty (or set_page_dirty_lock, as appropriate) must be called * after the page is finished with, and before put_page is called. * * get_user_pages is typically used for fewer-copy IO operations, to get a * handle on the memory by some means other than accesses via the user virtual * addresses. The pages may be submitted for DMA to devices or accessed via * their kernel linear mapping (via the kmap APIs). Care should be taken to * use the correct cache flushing APIs. * * See also get_user_pages_fast, for performance critical applications. */ |
28a35716d mm: use long type... |
1970 1971 1972 |
long get_user_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, unsigned long nr_pages, int write, int force, struct page **pages, struct vm_area_struct **vmas) |
b291f0003 mlock: mlocked pa... |
1973 |
{ |
58fa879e1 mm: FOLL flags fo... |
1974 |
int flags = FOLL_TOUCH; |
b291f0003 mlock: mlocked pa... |
1975 |
|
58fa879e1 mm: FOLL flags fo... |
1976 1977 |
if (pages) flags |= FOLL_GET; |
b291f0003 mlock: mlocked pa... |
1978 |
if (write) |
58fa879e1 mm: FOLL flags fo... |
1979 |
flags |= FOLL_WRITE; |
b291f0003 mlock: mlocked pa... |
1980 |
if (force) |
58fa879e1 mm: FOLL flags fo... |
1981 |
flags |= FOLL_FORCE; |
b291f0003 mlock: mlocked pa... |
1982 |
|
53a7706d5 mlock: do not hol... |
1983 1984 |
return __get_user_pages(tsk, mm, start, nr_pages, flags, pages, vmas, NULL); |
b291f0003 mlock: mlocked pa... |
1985 |
} |
1da177e4c Linux-2.6.12-rc2 |
1986 |
EXPORT_SYMBOL(get_user_pages); |
f3e8fccd0 mm: add get_dump_... |
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 |
/** * get_dump_page() - pin user page in memory while writing it to core dump * @addr: user address * * Returns struct page pointer of user page pinned for dump, * to be freed afterwards by page_cache_release() or put_page(). * * Returns NULL on any kind of failure - a hole must then be inserted into * the corefile, to preserve alignment with its headers; and also returns * NULL wherever the ZERO_PAGE, or an anonymous pte_none, has been found - * allowing a hole to be left in the corefile to save diskspace. * * Called without mmap_sem, but after all other threads have been killed. */ #ifdef CONFIG_ELF_CORE struct page *get_dump_page(unsigned long addr) { struct vm_area_struct *vma; struct page *page; if (__get_user_pages(current, current->mm, addr, 1, |
53a7706d5 mlock: do not hol... |
2008 2009 |
FOLL_FORCE | FOLL_DUMP | FOLL_GET, &page, &vma, NULL) < 1) |
f3e8fccd0 mm: add get_dump_... |
2010 |
return NULL; |
f3e8fccd0 mm: add get_dump_... |
2011 2012 2013 2014 |
flush_cache_page(vma, addr, page_to_pfn(page)); return page; } #endif /* CONFIG_ELF_CORE */ |
25ca1d6c0 mm: wrap get_lock... |
2015 |
pte_t *__get_locked_pte(struct mm_struct *mm, unsigned long addr, |
920c7a5d0 mm: remove fastca... |
2016 |
spinlock_t **ptl) |
c9cfcddfd VM: add common he... |
2017 2018 2019 2020 |
{ pgd_t * pgd = pgd_offset(mm, addr); pud_t * pud = pud_alloc(mm, pgd, addr); if (pud) { |
49c91fb01 [PATCH] VM: Fix t... |
2021 |
pmd_t * pmd = pmd_alloc(mm, pud, addr); |
f66055ab6 thp: verify pmd_t... |
2022 2023 |
if (pmd) { VM_BUG_ON(pmd_trans_huge(*pmd)); |
c9cfcddfd VM: add common he... |
2024 |
return pte_alloc_map_lock(mm, pmd, addr, ptl); |
f66055ab6 thp: verify pmd_t... |
2025 |
} |
c9cfcddfd VM: add common he... |
2026 2027 2028 |
} return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
2029 |
/* |
238f58d89 Support strange d... |
2030 2031 2032 2033 2034 2035 |
* 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 mm: add vm_insert... |
2036 2037 |
static int insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot) |
238f58d89 Support strange d... |
2038 |
{ |
423bad600 mm: add vm_insert... |
2039 |
struct mm_struct *mm = vma->vm_mm; |
238f58d89 Support strange d... |
2040 |
int retval; |
c9cfcddfd VM: add common he... |
2041 |
pte_t *pte; |
8a9f3ccd2 Memory controller... |
2042 |
spinlock_t *ptl; |
238f58d89 Support strange d... |
2043 |
retval = -EINVAL; |
a145dd411 VM: add "vm_inser... |
2044 |
if (PageAnon(page)) |
5b4e655e9 memcg: avoid acco... |
2045 |
goto out; |
238f58d89 Support strange d... |
2046 2047 |
retval = -ENOMEM; flush_dcache_page(page); |
c9cfcddfd VM: add common he... |
2048 |
pte = get_locked_pte(mm, addr, &ptl); |
238f58d89 Support strange d... |
2049 |
if (!pte) |
5b4e655e9 memcg: avoid acco... |
2050 |
goto out; |
238f58d89 Support strange d... |
2051 2052 2053 2054 2055 2056 |
retval = -EBUSY; if (!pte_none(*pte)) goto out_unlock; /* Ok, finally just insert the thing.. */ get_page(page); |
34e55232e mm: avoid false s... |
2057 |
inc_mm_counter_fast(mm, MM_FILEPAGES); |
238f58d89 Support strange d... |
2058 2059 2060 2061 |
page_add_file_rmap(page); set_pte_at(mm, addr, pte, mk_pte(page, prot)); retval = 0; |
8a9f3ccd2 Memory controller... |
2062 2063 |
pte_unmap_unlock(pte, ptl); return retval; |
238f58d89 Support strange d... |
2064 2065 2066 2067 2068 |
out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
bfa5bf6d6 [PATCH] Add kerne... |
2069 2070 2071 2072 2073 2074 |
/** * 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 VM: add "vm_inser... |
2075 2076 2077 2078 2079 2080 |
* 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 [PATCH] mm: split... |
2081 |
* (see split_page()). |
a145dd411 VM: add "vm_inser... |
2082 2083 2084 2085 2086 2087 2088 2089 |
* * 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 mm: kill vma flag... |
2090 2091 2092 2093 2094 |
* * 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 VM: add "vm_inser... |
2095 |
*/ |
423bad600 mm: add vm_insert... |
2096 2097 |
int vm_insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page) |
a145dd411 VM: add "vm_inser... |
2098 2099 2100 2101 2102 |
{ if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; if (!page_count(page)) return -EINVAL; |
4b6e1e370 mm: kill vma flag... |
2103 2104 2105 2106 2107 |
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 mm: add vm_insert... |
2108 |
return insert_page(vma, addr, page, vma->vm_page_prot); |
a145dd411 VM: add "vm_inser... |
2109 |
} |
e3c3374fb Make vm_insert_pa... |
2110 |
EXPORT_SYMBOL(vm_insert_page); |
a145dd411 VM: add "vm_inser... |
2111 |
|
423bad600 mm: add vm_insert... |
2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 |
static int insert_pfn(struct vm_area_struct *vma, unsigned long addr, unsigned long pfn, pgprot_t prot) { 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.. */ entry = pte_mkspecial(pfn_pte(pfn, prot)); set_pte_at(mm, addr, pte, entry); |
4b3073e1c MM: Pass a PTE po... |
2131 |
update_mmu_cache(vma, addr, pte); /* XXX: why not for insert_page? */ |
423bad600 mm: add vm_insert... |
2132 2133 2134 2135 2136 2137 2138 |
retval = 0; out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
e0dc0d8f4 [PATCH] add vm_in... |
2139 2140 2141 2142 2143 2144 |
/** * 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 mm/memory.c: fix ... |
2145 |
* Similar to vm_insert_page, this allows drivers to insert individual pages |
e0dc0d8f4 [PATCH] add vm_in... |
2146 2147 2148 2149 |
* 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 mm: remove nopfn |
2150 2151 2152 2153 2154 |
* * 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 [PATCH] add vm_in... |
2155 2156 |
*/ int vm_insert_pfn(struct vm_area_struct *vma, unsigned long addr, |
423bad600 mm: add vm_insert... |
2157 |
unsigned long pfn) |
e0dc0d8f4 [PATCH] add vm_in... |
2158 |
{ |
2ab640379 x86: PAT: hooks i... |
2159 |
int ret; |
e4b866ed1 x86 PAT: change t... |
2160 |
pgprot_t pgprot = vma->vm_page_prot; |
7e675137a mm: introduce pte... |
2161 2162 2163 2164 2165 2166 |
/* * 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 mm: introduce VM_... |
2167 2168 2169 2170 2171 |
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 [PATCH] add vm_in... |
2172 |
|
423bad600 mm: add vm_insert... |
2173 2174 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
5180da410 x86, pat: separat... |
2175 |
if (track_pfn_insert(vma, &pgprot, pfn)) |
2ab640379 x86: PAT: hooks i... |
2176 |
return -EINVAL; |
e4b866ed1 x86 PAT: change t... |
2177 |
ret = insert_pfn(vma, addr, pfn, pgprot); |
2ab640379 x86: PAT: hooks i... |
2178 |
|
2ab640379 x86: PAT: hooks i... |
2179 |
return ret; |
423bad600 mm: add vm_insert... |
2180 2181 |
} EXPORT_SYMBOL(vm_insert_pfn); |
e0dc0d8f4 [PATCH] add vm_in... |
2182 |
|
423bad600 mm: add vm_insert... |
2183 2184 2185 2186 |
int vm_insert_mixed(struct vm_area_struct *vma, unsigned long addr, unsigned long pfn) { BUG_ON(!(vma->vm_flags & VM_MIXEDMAP)); |
e0dc0d8f4 [PATCH] add vm_in... |
2187 |
|
423bad600 mm: add vm_insert... |
2188 2189 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
e0dc0d8f4 [PATCH] add vm_in... |
2190 |
|
423bad600 mm: add vm_insert... |
2191 2192 2193 2194 |
/* * 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 mm: ZERO_PAGE wit... |
2195 2196 |
* 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 mm: add vm_insert... |
2197 2198 2199 2200 2201 2202 2203 2204 |
*/ if (!HAVE_PTE_SPECIAL && pfn_valid(pfn)) { struct page *page; page = pfn_to_page(pfn); return insert_page(vma, addr, page, vma->vm_page_prot); } return insert_pfn(vma, addr, pfn, vma->vm_page_prot); |
e0dc0d8f4 [PATCH] add vm_in... |
2205 |
} |
423bad600 mm: add vm_insert... |
2206 |
EXPORT_SYMBOL(vm_insert_mixed); |
e0dc0d8f4 [PATCH] add vm_in... |
2207 |
|
a145dd411 VM: add "vm_inser... |
2208 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2209 2210 2211 2212 2213 2214 2215 2216 2217 |
* 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 [PATCH] mm: ptd_a... |
2218 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
2219 |
|
c74df32c7 [PATCH] mm: ptd_a... |
2220 |
pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
2221 2222 |
if (!pte) return -ENOMEM; |
6606c3e0d [PATCH] paravirt:... |
2223 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
2224 2225 |
do { BUG_ON(!pte_none(*pte)); |
7e675137a mm: introduce pte... |
2226 |
set_pte_at(mm, addr, pte, pte_mkspecial(pfn_pte(pfn, prot))); |
1da177e4c Linux-2.6.12-rc2 |
2227 2228 |
pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); |
6606c3e0d [PATCH] paravirt:... |
2229 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
2230 |
pte_unmap_unlock(pte - 1, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 |
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 thp: verify pmd_t... |
2245 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c Linux-2.6.12-rc2 |
2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 |
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 [PATCH] Add kerne... |
2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 |
/** * 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 Linux-2.6.12-rc2 |
2284 2285 2286 2287 2288 |
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 [PATCH] mm: fix r... |
2289 |
unsigned long end = addr + PAGE_ALIGN(size); |
1da177e4c Linux-2.6.12-rc2 |
2290 2291 2292 2293 2294 2295 2296 2297 |
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 mm: re-architect ... |
2298 2299 2300 |
* 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 mm: kill vma flag... |
2301 2302 2303 2304 |
* VM_DONTEXPAND * Disable vma merging and expanding with mremap(). * VM_DONTDUMP * Omit vma from core dump, even when VM_IO turned off. |
fb155c161 Allow arbitrary s... |
2305 2306 2307 2308 |
* * 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 mm, x86, pat: rew... |
2309 |
* See vm_normal_page() for details. |
1da177e4c Linux-2.6.12-rc2 |
2310 |
*/ |
b3b9c2932 mm, x86, pat: rew... |
2311 2312 2313 |
if (is_cow_mapping(vma->vm_flags)) { if (addr != vma->vm_start || end != vma->vm_end) return -EINVAL; |
fb155c161 Allow arbitrary s... |
2314 |
vma->vm_pgoff = pfn; |
b3b9c2932 mm, x86, pat: rew... |
2315 2316 2317 2318 |
} err = track_pfn_remap(vma, &prot, pfn, addr, PAGE_ALIGN(size)); if (err) |
3c8bb73ac x86: PAT: store v... |
2319 |
return -EINVAL; |
fb155c161 Allow arbitrary s... |
2320 |
|
314e51b98 mm: kill vma flag... |
2321 |
vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP; |
1da177e4c Linux-2.6.12-rc2 |
2322 2323 2324 2325 2326 |
BUG_ON(addr >= end); pfn -= addr >> PAGE_SHIFT; pgd = pgd_offset(mm, addr); flush_cache_range(vma, addr, end); |
1da177e4c Linux-2.6.12-rc2 |
2327 2328 2329 2330 2331 2332 2333 |
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 x86: PAT: hooks i... |
2334 2335 |
if (err) |
5180da410 x86, pat: separat... |
2336 |
untrack_pfn(vma, pfn, PAGE_ALIGN(size)); |
2ab640379 x86: PAT: hooks i... |
2337 |
|
1da177e4c Linux-2.6.12-rc2 |
2338 2339 2340 |
return err; } EXPORT_SYMBOL(remap_pfn_range); |
b4cbb197c vm: add vm_iomap_... |
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 |
/** * 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 Add apply_to_page... |
2387 2388 2389 2390 2391 2392 |
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 CONFIG_HIGHPTE vs... |
2393 |
pgtable_t token; |
949099148 Add unitialized_v... |
2394 |
spinlock_t *uninitialized_var(ptl); |
aee16b3ce Add apply_to_page... |
2395 2396 2397 2398 2399 2400 2401 2402 |
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 mm/apply_to_range... |
2403 |
arch_enter_lazy_mmu_mode(); |
2f569afd9 CONFIG_HIGHPTE vs... |
2404 |
token = pmd_pgtable(*pmd); |
aee16b3ce Add apply_to_page... |
2405 2406 |
do { |
c36987e2e mm: don't call pt... |
2407 |
err = fn(pte++, token, addr, data); |
aee16b3ce Add apply_to_page... |
2408 2409 |
if (err) break; |
c36987e2e mm: don't call pt... |
2410 |
} while (addr += PAGE_SIZE, addr != end); |
aee16b3ce Add apply_to_page... |
2411 |
|
38e0edb15 mm/apply_to_range... |
2412 |
arch_leave_lazy_mmu_mode(); |
aee16b3ce Add apply_to_page... |
2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 |
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 hugetlb: introduc... |
2425 |
BUG_ON(pud_huge(*pud)); |
aee16b3ce Add apply_to_page... |
2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 |
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 mmu-notifiers: re... |
2467 |
unsigned long end = addr + size; |
aee16b3ce Add apply_to_page... |
2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 |
int err; BUG_ON(addr >= end); 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 mmu-notifiers: re... |
2478 |
|
aee16b3ce Add apply_to_page... |
2479 2480 2481 |
return err; } EXPORT_SYMBOL_GPL(apply_to_page_range); |
1da177e4c Linux-2.6.12-rc2 |
2482 |
/* |
8f4e2101f [PATCH] mm: page ... |
2483 2484 2485 |
* 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 |
a335b2e17 mm: Fix out-of-da... |
2486 |
* might give a mix of unmatched parts, do_swap_page and do_nonlinear_fault |
8f4e2101f [PATCH] mm: page ... |
2487 2488 |
* must check under lock before unmapping the pte and proceeding * (but do_wp_page is only called after already making such a check; |
a335b2e17 mm: Fix out-of-da... |
2489 |
* and do_anonymous_page can safely check later on). |
8f4e2101f [PATCH] mm: page ... |
2490 |
*/ |
4c21e2f24 [PATCH] mm: split... |
2491 |
static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd, |
8f4e2101f [PATCH] mm: page ... |
2492 2493 2494 2495 2496 |
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 [PATCH] mm: split... |
2497 2498 |
spinlock_t *ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
8f4e2101f [PATCH] mm: page ... |
2499 |
same = pte_same(*page_table, orig_pte); |
4c21e2f24 [PATCH] mm: split... |
2500 |
spin_unlock(ptl); |
8f4e2101f [PATCH] mm: page ... |
2501 2502 2503 2504 2505 |
} #endif pte_unmap(page_table); return same; } |
9de455b20 [PATCH] Pass vma ... |
2506 |
static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma) |
6aab341e0 mm: re-architect ... |
2507 |
{ |
0abdd7a81 dma-debug: introd... |
2508 |
debug_dma_assert_idle(src); |
6aab341e0 mm: re-architect ... |
2509 2510 2511 2512 2513 2514 2515 |
/* * 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 mm: remove the se... |
2516 |
void *kaddr = kmap_atomic(dst); |
5d2a2dbbc cow_user_page: fi... |
2517 2518 2519 2520 2521 2522 2523 2524 2525 |
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 use clear_page()/... |
2526 |
clear_page(kaddr); |
9b04c5fec mm: remove the se... |
2527 |
kunmap_atomic(kaddr); |
c4ec7b0de [PATCH] mm: D-cac... |
2528 |
flush_dcache_page(dst); |
0ed361dec mm: fix PageUptod... |
2529 2530 |
} else copy_user_highpage(dst, src, va, vma); |
6aab341e0 mm: re-architect ... |
2531 |
} |
1da177e4c Linux-2.6.12-rc2 |
2532 |
/* |
fb09a4642 mm: consolidate c... |
2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 |
* 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; vmf.page = page; 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; } /* |
1da177e4c Linux-2.6.12-rc2 |
2565 2566 2567 2568 |
* 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 Linux-2.6.12-rc2 |
2569 2570 2571 2572 2573 2574 2575 2576 2577 |
* 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 [PATCH] mm: page ... |
2578 2579 2580 |
* 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 Linux-2.6.12-rc2 |
2581 |
*/ |
65500d234 [PATCH] mm: page ... |
2582 2583 |
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 [PATCH] mm: page ... |
2584 |
spinlock_t *ptl, pte_t orig_pte) |
e6219ec81 mm: add lock rele... |
2585 |
__releases(ptl) |
1da177e4c Linux-2.6.12-rc2 |
2586 |
{ |
2ec74c3ef mm: move all mmu ... |
2587 |
struct page *old_page, *new_page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2588 |
pte_t entry; |
b009c024f do_wp_page: remov... |
2589 |
int ret = 0; |
a200ee182 mm: set_page_dirt... |
2590 |
int page_mkwrite = 0; |
d08b3851d [PATCH] mm: track... |
2591 |
struct page *dirty_page = NULL; |
1756954c6 mm: fix build war... |
2592 2593 |
unsigned long mmun_start = 0; /* For mmu_notifiers */ unsigned long mmun_end = 0; /* For mmu_notifiers */ |
1da177e4c Linux-2.6.12-rc2 |
2594 |
|
6aab341e0 mm: re-architect ... |
2595 |
old_page = vm_normal_page(vma, address, orig_pte); |
251b97f55 mm: dirty page ac... |
2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 |
if (!old_page) { /* * VM_MIXEDMAP !pfn_valid() case * * We should not cow pages in a shared writeable mapping. * Just mark the pages writable as we can't do any dirty * accounting on raw pfn maps. */ if ((vma->vm_flags & (VM_WRITE|VM_SHARED)) == (VM_WRITE|VM_SHARED)) goto reuse; |
6aab341e0 mm: re-architect ... |
2607 |
goto gotten; |
251b97f55 mm: dirty page ac... |
2608 |
} |
1da177e4c Linux-2.6.12-rc2 |
2609 |
|
d08b3851d [PATCH] mm: track... |
2610 |
/* |
ee6a64578 [PATCH] mm: fixup... |
2611 2612 |
* Take out anonymous pages first, anonymous shared vmas are * not dirty accountable. |
d08b3851d [PATCH] mm: track... |
2613 |
*/ |
9a8408951 ksm: identify Pag... |
2614 |
if (PageAnon(old_page) && !PageKsm(old_page)) { |
ab967d860 mm: wp lock page ... |
2615 2616 2617 2618 2619 2620 2621 2622 |
if (!trylock_page(old_page)) { page_cache_get(old_page); 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); |
ab967d860 mm: wp lock page ... |
2623 2624 2625 |
goto unlock; } page_cache_release(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2626 |
} |
b009c024f do_wp_page: remov... |
2627 |
if (reuse_swap_page(old_page)) { |
c44b67432 rmap: move exclus... |
2628 2629 2630 2631 2632 2633 |
/* * 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 do_wp_page: remov... |
2634 2635 2636 |
unlock_page(old_page); goto reuse; } |
ab967d860 mm: wp lock page ... |
2637 |
unlock_page(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2638 |
} else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) == |
d08b3851d [PATCH] mm: track... |
2639 |
(VM_WRITE|VM_SHARED))) { |
ee6a64578 [PATCH] mm: fixup... |
2640 2641 2642 2643 2644 |
/* * Only catch write-faults on shared writable pages, * read-only shared pages can get COWed by * get_user_pages(.write=1, .force=1). */ |
9637a5efd [PATCH] add page_... |
2645 |
if (vma->vm_ops && vma->vm_ops->page_mkwrite) { |
c2ec175c3 mm: page_mkwrite ... |
2646 |
int tmp; |
9637a5efd [PATCH] add page_... |
2647 2648 |
page_cache_get(old_page); pte_unmap_unlock(page_table, ptl); |
fb09a4642 mm: consolidate c... |
2649 2650 2651 2652 2653 |
tmp = do_page_mkwrite(vma, old_page, address); if (unlikely(!tmp || (tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) { page_cache_release(old_page); return tmp; |
c2ec175c3 mm: page_mkwrite ... |
2654 |
} |
9637a5efd [PATCH] add page_... |
2655 2656 2657 2658 2659 2660 2661 2662 |
/* * 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); |
b827e496c mm: close page_mk... |
2663 2664 |
if (!pte_same(*page_table, orig_pte)) { unlock_page(old_page); |
9637a5efd [PATCH] add page_... |
2665 |
goto unlock; |
b827e496c mm: close page_mk... |
2666 |
} |
a200ee182 mm: set_page_dirt... |
2667 2668 |
page_mkwrite = 1; |
1da177e4c Linux-2.6.12-rc2 |
2669 |
} |
d08b3851d [PATCH] mm: track... |
2670 2671 |
dirty_page = old_page; get_page(dirty_page); |
9637a5efd [PATCH] add page_... |
2672 |
|
251b97f55 mm: dirty page ac... |
2673 |
reuse: |
8c8a743c5 sched/numa: Use {... |
2674 2675 2676 2677 2678 2679 2680 |
/* * Clear the pages cpupid information as the existing * information potentially belongs to a now completely * unrelated process. */ if (old_page) page_cpupid_xchg_last(old_page, (1 << LAST_CPUPID_SHIFT) - 1); |
9637a5efd [PATCH] add page_... |
2681 2682 2683 |
flush_cache_page(vma, address, pte_pfn(orig_pte)); entry = pte_mkyoung(orig_pte); entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
954ffcb35 flush icache befo... |
2684 |
if (ptep_set_access_flags(vma, address, page_table, entry,1)) |
4b3073e1c MM: Pass a PTE po... |
2685 |
update_mmu_cache(vma, address, page_table); |
72ddc8f72 do_wp_page: clari... |
2686 |
pte_unmap_unlock(page_table, ptl); |
9637a5efd [PATCH] add page_... |
2687 |
ret |= VM_FAULT_WRITE; |
72ddc8f72 do_wp_page: clari... |
2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 |
if (!dirty_page) return ret; /* * Yes, Virginia, this is actually required to prevent a race * with clear_page_dirty_for_io() from clearing the page dirty * bit after it clear all dirty ptes, but before a racing * do_wp_page installs a dirty pte. * |
f0c6d4d29 mm: introduce do_... |
2698 |
* do_shared_fault is protected similarly. |
72ddc8f72 do_wp_page: clari... |
2699 2700 2701 |
*/ if (!page_mkwrite) { wait_on_page_locked(dirty_page); |
ed6d7c8e5 mm: remove unused... |
2702 |
set_page_dirty_balance(dirty_page); |
41c4d25f7 mm: Update file t... |
2703 2704 2705 |
/* file_update_time outside page_lock */ if (vma->vm_file) file_update_time(vma->vm_file); |
72ddc8f72 do_wp_page: clari... |
2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 |
} put_page(dirty_page); if (page_mkwrite) { struct address_space *mapping = dirty_page->mapping; set_page_dirty(dirty_page); unlock_page(dirty_page); page_cache_release(dirty_page); if (mapping) { /* * Some device drivers do not set page.mapping * but still dirty their pages */ balance_dirty_pages_ratelimited(mapping); } } |
72ddc8f72 do_wp_page: clari... |
2722 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2723 |
} |
1da177e4c Linux-2.6.12-rc2 |
2724 2725 2726 2727 |
/* * Ok, we need to copy. Oh, well.. */ |
b5810039a [PATCH] core remo... |
2728 |
page_cache_get(old_page); |
920fc356f [PATCH] unpaged: ... |
2729 |
gotten: |
8f4e2101f [PATCH] mm: page ... |
2730 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2731 2732 |
if (unlikely(anon_vma_prepare(vma))) |
65500d234 [PATCH] mm: page ... |
2733 |
goto oom; |
a13ea5b75 mm: reinstate ZER... |
2734 |
|
62eede62d mm: ZERO_PAGE wit... |
2735 |
if (is_zero_pfn(pte_pfn(orig_pte))) { |
a13ea5b75 mm: reinstate ZER... |
2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 |
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); } __SetPageUptodate(new_page); |
d715ae08f memcg: rename hig... |
2746 |
if (mem_cgroup_charge_anon(new_page, mm, GFP_KERNEL)) |
8a9f3ccd2 Memory controller... |
2747 |
goto oom_free_new; |
6bdb913f0 mm: wrap calls to... |
2748 |
mmun_start = address & PAGE_MASK; |
1756954c6 mm: fix build war... |
2749 |
mmun_end = mmun_start + PAGE_SIZE; |
6bdb913f0 mm: wrap calls to... |
2750 |
mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end); |
1da177e4c Linux-2.6.12-rc2 |
2751 2752 2753 |
/* * Re-check the pte - we dropped the lock */ |
8f4e2101f [PATCH] mm: page ... |
2754 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
65500d234 [PATCH] mm: page ... |
2755 |
if (likely(pte_same(*page_table, orig_pte))) { |
920fc356f [PATCH] unpaged: ... |
2756 |
if (old_page) { |
920fc356f [PATCH] unpaged: ... |
2757 |
if (!PageAnon(old_page)) { |
34e55232e mm: avoid false s... |
2758 2759 |
dec_mm_counter_fast(mm, MM_FILEPAGES); inc_mm_counter_fast(mm, MM_ANONPAGES); |
920fc356f [PATCH] unpaged: ... |
2760 2761 |
} } else |
34e55232e mm: avoid false s... |
2762 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
eca351336 [PATCH] Fix missi... |
2763 |
flush_cache_page(vma, address, pte_pfn(orig_pte)); |
65500d234 [PATCH] mm: page ... |
2764 2765 |
entry = mk_pte(new_page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
4ce072f1f [PATCH] mm: fix a... |
2766 2767 2768 2769 2770 2771 |
/* * 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. */ |
828502d30 ksm: add mmu_noti... |
2772 |
ptep_clear_flush(vma, address, page_table); |
9617d95e6 [PATCH] mm: rmap ... |
2773 |
page_add_new_anon_rmap(new_page, vma, address); |
828502d30 ksm: add mmu_noti... |
2774 2775 2776 2777 2778 2779 |
/* * 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); |
4b3073e1c MM: Pass a PTE po... |
2780 |
update_mmu_cache(vma, address, page_table); |
945754a17 mm: fix race in C... |
2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 |
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. */ |
edc315fd2 badpage: remove v... |
2804 |
page_remove_rmap(old_page); |
945754a17 mm: fix race in C... |
2805 |
} |
1da177e4c Linux-2.6.12-rc2 |
2806 2807 |
/* Free the old page.. */ new_page = old_page; |
f33ea7f40 [PATCH] fix get_u... |
2808 |
ret |= VM_FAULT_WRITE; |
8a9f3ccd2 Memory controller... |
2809 2810 |
} else mem_cgroup_uncharge_page(new_page); |
6bdb913f0 mm: wrap calls to... |
2811 2812 |
if (new_page) page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2813 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2814 |
pte_unmap_unlock(page_table, ptl); |
1756954c6 mm: fix build war... |
2815 |
if (mmun_end > mmun_start) |
6bdb913f0 mm: wrap calls to... |
2816 |
mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end); |
e15f8c01a mlock: fix race w... |
2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 |
if (old_page) { /* * Don't let another task, with possibly unlocked vma, * keep the mlocked page. */ if ((ret & VM_FAULT_WRITE) && (vma->vm_flags & VM_LOCKED)) { lock_page(old_page); /* LRU manipulation */ munlock_vma_page(old_page); unlock_page(old_page); } page_cache_release(old_page); } |
f33ea7f40 [PATCH] fix get_u... |
2829 |
return ret; |
8a9f3ccd2 Memory controller... |
2830 |
oom_free_new: |
6dbf6d3bb memcg: page_cache... |
2831 |
page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2832 |
oom: |
66521d5aa mm/memory.c: remo... |
2833 |
if (old_page) |
920fc356f [PATCH] unpaged: ... |
2834 |
page_cache_release(old_page); |
1da177e4c Linux-2.6.12-rc2 |
2835 2836 |
return VM_FAULT_OOM; } |
97a894136 mm: Remove i_mmap... |
2837 |
static void unmap_mapping_range_vma(struct vm_area_struct *vma, |
1da177e4c Linux-2.6.12-rc2 |
2838 2839 2840 |
unsigned long start_addr, unsigned long end_addr, struct zap_details *details) { |
f5cc4eef9 VM: make zap_page... |
2841 |
zap_page_range_single(vma, start_addr, end_addr - start_addr, details); |
1da177e4c Linux-2.6.12-rc2 |
2842 |
} |
6b2dbba8b mm: replace vma p... |
2843 |
static inline void unmap_mapping_range_tree(struct rb_root *root, |
1da177e4c Linux-2.6.12-rc2 |
2844 2845 2846 |
struct zap_details *details) { struct vm_area_struct *vma; |
1da177e4c Linux-2.6.12-rc2 |
2847 |
pgoff_t vba, vea, zba, zea; |
6b2dbba8b mm: replace vma p... |
2848 |
vma_interval_tree_foreach(vma, root, |
1da177e4c Linux-2.6.12-rc2 |
2849 |
details->first_index, details->last_index) { |
1da177e4c Linux-2.6.12-rc2 |
2850 2851 |
vba = vma->vm_pgoff; |
d6e932177 mm: use vma_pages... |
2852 |
vea = vba + vma_pages(vma) - 1; |
1da177e4c Linux-2.6.12-rc2 |
2853 2854 2855 2856 2857 2858 2859 |
/* Assume for now that PAGE_CACHE_SHIFT == PAGE_SHIFT */ zba = details->first_index; if (zba < vba) zba = vba; zea = details->last_index; if (zea > vea) zea = vea; |
97a894136 mm: Remove i_mmap... |
2860 |
unmap_mapping_range_vma(vma, |
1da177e4c Linux-2.6.12-rc2 |
2861 2862 |
((zba - vba) << PAGE_SHIFT) + vma->vm_start, ((zea - vba + 1) << PAGE_SHIFT) + vma->vm_start, |
97a894136 mm: Remove i_mmap... |
2863 |
details); |
1da177e4c Linux-2.6.12-rc2 |
2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 |
} } static inline void unmap_mapping_range_list(struct list_head *head, struct zap_details *details) { struct vm_area_struct *vma; /* * In nonlinear VMAs there is no correspondence between virtual address * offset and file offset. So we must perform an exhaustive search * across *all* the pages in each nonlinear VMA, not just the pages * whose virtual address lies outside the file truncation point. */ |
6b2dbba8b mm: replace vma p... |
2878 |
list_for_each_entry(vma, head, shared.nonlinear) { |
1da177e4c Linux-2.6.12-rc2 |
2879 |
details->nonlinear_vma = vma; |
97a894136 mm: Remove i_mmap... |
2880 |
unmap_mapping_range_vma(vma, vma->vm_start, vma->vm_end, details); |
1da177e4c Linux-2.6.12-rc2 |
2881 2882 2883 2884 |
} } /** |
72fd4a35a [PATCH] Numerous ... |
2885 |
* 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 [PATCH] DocBook: ... |
2886 |
* @mapping: the address space containing mmaps to be unmapped. |
1da177e4c Linux-2.6.12-rc2 |
2887 2888 |
* @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 truncate: new hel... |
2889 |
* boundary. Note that this is different from truncate_pagecache(), which |
1da177e4c Linux-2.6.12-rc2 |
2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 |
* 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) { struct zap_details details; 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; details.nonlinear_vma = NULL; details.first_index = hba; details.last_index = hba + hlen - 1; if (details.last_index < details.first_index) details.last_index = ULONG_MAX; |
1da177e4c Linux-2.6.12-rc2 |
2919 |
|
1da177e4c Linux-2.6.12-rc2 |
2920 |
|
3d48ae45e mm: Convert i_mma... |
2921 |
mutex_lock(&mapping->i_mmap_mutex); |
6b2dbba8b mm: replace vma p... |
2922 |
if (unlikely(!RB_EMPTY_ROOT(&mapping->i_mmap))) |
1da177e4c Linux-2.6.12-rc2 |
2923 2924 2925 |
unmap_mapping_range_tree(&mapping->i_mmap, &details); if (unlikely(!list_empty(&mapping->i_mmap_nonlinear))) unmap_mapping_range_list(&mapping->i_mmap_nonlinear, &details); |
3d48ae45e mm: Convert i_mma... |
2926 |
mutex_unlock(&mapping->i_mmap_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2927 2928 |
} EXPORT_SYMBOL(unmap_mapping_range); |
1da177e4c Linux-2.6.12-rc2 |
2929 |
/* |
8f4e2101f [PATCH] mm: page ... |
2930 2931 2932 |
* 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 Linux-2.6.12-rc2 |
2933 |
*/ |
65500d234 [PATCH] mm: page ... |
2934 2935 |
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 Remove internal u... |
2936 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
2937 |
{ |
8f4e2101f [PATCH] mm: page ... |
2938 |
spinlock_t *ptl; |
56f31801c mm: cleanup "swap... |
2939 |
struct page *page, *swapcache; |
65500d234 [PATCH] mm: page ... |
2940 |
swp_entry_t entry; |
1da177e4c Linux-2.6.12-rc2 |
2941 |
pte_t pte; |
d065bd810 mm: retry page fa... |
2942 |
int locked; |
56039efa1 memcg: fix ugly i... |
2943 |
struct mem_cgroup *ptr; |
ad8c2ee80 rmap: add exclusi... |
2944 |
int exclusive = 0; |
83c54070e mm: fault feedbac... |
2945 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2946 |
|
4c21e2f24 [PATCH] mm: split... |
2947 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
8f4e2101f [PATCH] mm: page ... |
2948 |
goto out; |
65500d234 [PATCH] mm: page ... |
2949 2950 |
entry = pte_to_swp_entry(orig_pte); |
d1737fdbe HWPOISON: Add bas... |
2951 2952 2953 2954 2955 2956 2957 |
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 mm: sigbus instea... |
2958 |
ret = VM_FAULT_SIGBUS; |
d1737fdbe HWPOISON: Add bas... |
2959 |
} |
0697212a4 [PATCH] Swapless ... |
2960 2961 |
goto out; } |
0ff922452 [PATCH] per-task-... |
2962 |
delayacct_set_flag(DELAYACCT_PF_SWAPIN); |
1da177e4c Linux-2.6.12-rc2 |
2963 2964 |
page = lookup_swap_cache(entry); if (!page) { |
02098feaa swapin needs gfp_... |
2965 2966 |
page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, vma, address); |
1da177e4c Linux-2.6.12-rc2 |
2967 2968 |
if (!page) { /* |
8f4e2101f [PATCH] mm: page ... |
2969 2970 |
* Back out if somebody else faulted in this pte * while we released the pte lock. |
1da177e4c Linux-2.6.12-rc2 |
2971 |
*/ |
8f4e2101f [PATCH] mm: page ... |
2972 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
2973 2974 |
if (likely(pte_same(*page_table, orig_pte))) ret = VM_FAULT_OOM; |
0ff922452 [PATCH] per-task-... |
2975 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
65500d234 [PATCH] mm: page ... |
2976 |
goto unlock; |
1da177e4c Linux-2.6.12-rc2 |
2977 2978 2979 2980 |
} /* Had to read the page from swap area: Major fault */ ret = VM_FAULT_MAJOR; |
f8891e5e1 [PATCH] Light wei... |
2981 |
count_vm_event(PGMAJFAULT); |
456f998ec memcg: add the pa... |
2982 |
mem_cgroup_count_vm_event(mm, PGMAJFAULT); |
d1737fdbe HWPOISON: Add bas... |
2983 |
} else if (PageHWPoison(page)) { |
71f72525d HWPOISON: comment... |
2984 2985 2986 2987 |
/* * hwpoisoned dirty swapcache pages are kept for killing * owner processes (which may be unknown at hwpoison time) */ |
d1737fdbe HWPOISON: Add bas... |
2988 2989 |
ret = VM_FAULT_HWPOISON; delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
56f31801c mm: cleanup "swap... |
2990 |
swapcache = page; |
4779cb31c HWPOISON: Fix pag... |
2991 |
goto out_release; |
1da177e4c Linux-2.6.12-rc2 |
2992 |
} |
56f31801c mm: cleanup "swap... |
2993 |
swapcache = page; |
d065bd810 mm: retry page fa... |
2994 |
locked = lock_page_or_retry(page, mm, flags); |
e709ffd61 mm: remove swap t... |
2995 |
|
073e587ec memcg: move charg... |
2996 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
d065bd810 mm: retry page fa... |
2997 2998 2999 3000 |
if (!locked) { ret |= VM_FAULT_RETRY; goto out_release; } |
073e587ec memcg: move charg... |
3001 |
|
4969c1192 mm: fix swapin ra... |
3002 |
/* |
31c4a3d3a mm: further fix s... |
3003 3004 3005 3006 |
* 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 mm: fix swapin ra... |
3007 |
*/ |
31c4a3d3a mm: further fix s... |
3008 |
if (unlikely(!PageSwapCache(page) || page_private(page) != entry.val)) |
4969c1192 mm: fix swapin ra... |
3009 |
goto out_page; |
cbf86cfe0 ksm: remove old s... |
3010 3011 3012 3013 |
page = ksm_might_need_to_copy(page, vma, address); if (unlikely(!page)) { ret = VM_FAULT_OOM; page = swapcache; |
cbf86cfe0 ksm: remove old s... |
3014 |
goto out_page; |
5ad646880 ksm: let shared p... |
3015 |
} |
2c26fdd70 memcg: revert gfp... |
3016 |
if (mem_cgroup_try_charge_swapin(mm, page, GFP_KERNEL, &ptr)) { |
8a9f3ccd2 Memory controller... |
3017 |
ret = VM_FAULT_OOM; |
bc43f75cd mm: fix pageref l... |
3018 |
goto out_page; |
8a9f3ccd2 Memory controller... |
3019 |
} |
1da177e4c Linux-2.6.12-rc2 |
3020 |
/* |
8f4e2101f [PATCH] mm: page ... |
3021 |
* Back out if somebody else already faulted in this pte. |
1da177e4c Linux-2.6.12-rc2 |
3022 |
*/ |
8f4e2101f [PATCH] mm: page ... |
3023 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
9e9bef07c [PATCH] mm: do_sw... |
3024 |
if (unlikely(!pte_same(*page_table, orig_pte))) |
b81074800 [PATCH] do_swap_p... |
3025 |
goto out_nomap; |
b81074800 [PATCH] do_swap_p... |
3026 3027 3028 3029 |
if (unlikely(!PageUptodate(page))) { ret = VM_FAULT_SIGBUS; goto out_nomap; |
1da177e4c Linux-2.6.12-rc2 |
3030 |
} |
8c7c6e34a memcg: mem+swap c... |
3031 3032 3033 3034 3035 3036 3037 3038 |
/* * 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. |
03f3c4336 memcg: fix swap a... |
3039 3040 3041 3042 |
* Because delete_from_swap_page() may be called by reuse_swap_page(), * mem_cgroup_commit_charge_swapin() may not be able to find swp_entry * in page->private. In this case, a record in swap_cgroup is silently * discarded at swap_free(). |
8c7c6e34a memcg: mem+swap c... |
3043 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
3044 |
|
34e55232e mm: avoid false s... |
3045 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
b084d4353 mm: count swap usage |
3046 |
dec_mm_counter_fast(mm, MM_SWAPENTS); |
1da177e4c Linux-2.6.12-rc2 |
3047 |
pte = mk_pte(page, vma->vm_page_prot); |
30c9f3a9f Remove internal u... |
3048 |
if ((flags & FAULT_FLAG_WRITE) && reuse_swap_page(page)) { |
1da177e4c Linux-2.6.12-rc2 |
3049 |
pte = maybe_mkwrite(pte_mkdirty(pte), vma); |
30c9f3a9f Remove internal u... |
3050 |
flags &= ~FAULT_FLAG_WRITE; |
9a5b489b8 mm: set VM_FAULT_... |
3051 |
ret |= VM_FAULT_WRITE; |
ad8c2ee80 rmap: add exclusi... |
3052 |
exclusive = 1; |
1da177e4c Linux-2.6.12-rc2 |
3053 |
} |
1da177e4c Linux-2.6.12-rc2 |
3054 |
flush_icache_page(vma, page); |
179ef71cb mm: save soft-dir... |
3055 3056 |
if (pte_swp_soft_dirty(orig_pte)) pte = pte_mksoft_dirty(pte); |
1da177e4c Linux-2.6.12-rc2 |
3057 |
set_pte_at(mm, address, page_table, pte); |
56f31801c mm: cleanup "swap... |
3058 |
if (page == swapcache) |
af34770e5 mm: reduce rmap o... |
3059 |
do_page_add_anon_rmap(page, vma, address, exclusive); |
56f31801c mm: cleanup "swap... |
3060 3061 |
else /* ksm created a completely new copy */ page_add_new_anon_rmap(page, vma, address); |
03f3c4336 memcg: fix swap a... |
3062 3063 |
/* It's better to call commit-charge after rmap is established */ mem_cgroup_commit_charge_swapin(page, ptr); |
1da177e4c Linux-2.6.12-rc2 |
3064 |
|
c475a8ab6 [PATCH] can_share... |
3065 |
swap_free(entry); |
b291f0003 mlock: mlocked pa... |
3066 |
if (vm_swap_full() || (vma->vm_flags & VM_LOCKED) || PageMlocked(page)) |
a2c43eed8 mm: try_to_free_s... |
3067 |
try_to_free_swap(page); |
c475a8ab6 [PATCH] can_share... |
3068 |
unlock_page(page); |
56f31801c mm: cleanup "swap... |
3069 |
if (page != swapcache) { |
4969c1192 mm: fix swapin ra... |
3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 |
/* * 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); page_cache_release(swapcache); } |
c475a8ab6 [PATCH] can_share... |
3081 |
|
30c9f3a9f Remove internal u... |
3082 |
if (flags & FAULT_FLAG_WRITE) { |
61469f1d5 memcg: when do_sw... |
3083 3084 3085 |
ret |= do_wp_page(mm, vma, address, page_table, pmd, ptl, pte); if (ret & VM_FAULT_ERROR) ret &= VM_FAULT_ERROR; |
1da177e4c Linux-2.6.12-rc2 |
3086 3087 3088 3089 |
goto out; } /* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
3090 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
3091 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
3092 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
3093 3094 |
out: return ret; |
b81074800 [PATCH] do_swap_p... |
3095 |
out_nomap: |
7a81b88cb memcg: introduce ... |
3096 |
mem_cgroup_cancel_charge_swapin(ptr); |
8f4e2101f [PATCH] mm: page ... |
3097 |
pte_unmap_unlock(page_table, ptl); |
bc43f75cd mm: fix pageref l... |
3098 |
out_page: |
b81074800 [PATCH] do_swap_p... |
3099 |
unlock_page(page); |
4779cb31c HWPOISON: Fix pag... |
3100 |
out_release: |
b81074800 [PATCH] do_swap_p... |
3101 |
page_cache_release(page); |
56f31801c mm: cleanup "swap... |
3102 |
if (page != swapcache) { |
4969c1192 mm: fix swapin ra... |
3103 3104 3105 |
unlock_page(swapcache); page_cache_release(swapcache); } |
65500d234 [PATCH] mm: page ... |
3106 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
3107 3108 3109 |
} /* |
8ca3eb080 guard page for st... |
3110 3111 |
* This is like a special single-page "expand_{down|up}wards()", * except we must first make sure that 'address{-|+}PAGE_SIZE' |
320b2b8de mm: keep a guard ... |
3112 |
* doesn't hit another vma. |
320b2b8de mm: keep a guard ... |
3113 3114 3115 3116 3117 |
*/ 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 mm: make stack gu... |
3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 |
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 mm: keep a guard ... |
3128 |
|
d05f3169c mm: make expand_d... |
3129 |
expand_downwards(vma, address - PAGE_SIZE); |
320b2b8de mm: keep a guard ... |
3130 |
} |
8ca3eb080 guard page for st... |
3131 3132 3133 3134 3135 3136 3137 3138 3139 |
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; expand_upwards(vma, address + PAGE_SIZE); } |
320b2b8de mm: keep a guard ... |
3140 3141 3142 3143 |
return 0; } /* |
8f4e2101f [PATCH] mm: page ... |
3144 3145 3146 |
* 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 Linux-2.6.12-rc2 |
3147 |
*/ |
65500d234 [PATCH] mm: page ... |
3148 3149 |
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 Remove internal u... |
3150 |
unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3151 |
{ |
8f4e2101f [PATCH] mm: page ... |
3152 3153 |
struct page *page; spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
3154 |
pte_t entry; |
1da177e4c Linux-2.6.12-rc2 |
3155 |
|
11ac55247 mm: fix page tabl... |
3156 3157 3158 3159 |
pte_unmap(page_table); /* Check if we need to add a guard page to the stack */ if (check_stack_guard_page(vma, address) < 0) |
320b2b8de mm: keep a guard ... |
3160 |
return VM_FAULT_SIGBUS; |
11ac55247 mm: fix page tabl... |
3161 |
/* Use the zero-page for reads */ |
62eede62d mm: ZERO_PAGE wit... |
3162 3163 3164 |
if (!(flags & FAULT_FLAG_WRITE)) { entry = pte_mkspecial(pfn_pte(my_zero_pfn(address), vma->vm_page_prot)); |
11ac55247 mm: fix page tabl... |
3165 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
a13ea5b75 mm: reinstate ZER... |
3166 3167 3168 3169 |
if (!pte_none(*page_table)) goto unlock; goto setpte; } |
557ed1fa2 remove ZERO_PAGE |
3170 |
/* Allocate our own private page. */ |
557ed1fa2 remove ZERO_PAGE |
3171 3172 3173 3174 3175 |
if (unlikely(anon_vma_prepare(vma))) goto oom; page = alloc_zeroed_user_highpage_movable(vma, address); if (!page) goto oom; |
52f37629f THP: fix comment ... |
3176 3177 3178 3179 3180 |
/* * The memory barrier inside __SetPageUptodate makes sure that * preceeding stores to the page contents become visible before * the set_pte_at() write. */ |
0ed361dec mm: fix PageUptod... |
3181 |
__SetPageUptodate(page); |
8f4e2101f [PATCH] mm: page ... |
3182 |
|
d715ae08f memcg: rename hig... |
3183 |
if (mem_cgroup_charge_anon(page, mm, GFP_KERNEL)) |
8a9f3ccd2 Memory controller... |
3184 |
goto oom_free_page; |
557ed1fa2 remove ZERO_PAGE |
3185 |
entry = mk_pte(page, vma->vm_page_prot); |
1ac0cb5d0 mm: fix anonymous... |
3186 3187 |
if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); |
1da177e4c Linux-2.6.12-rc2 |
3188 |
|
557ed1fa2 remove ZERO_PAGE |
3189 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1c2fb7a4c ksm: fix deadlock... |
3190 |
if (!pte_none(*page_table)) |
557ed1fa2 remove ZERO_PAGE |
3191 |
goto release; |
9ba692948 ksm: fix oom dead... |
3192 |
|
34e55232e mm: avoid false s... |
3193 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
557ed1fa2 remove ZERO_PAGE |
3194 |
page_add_new_anon_rmap(page, vma, address); |
a13ea5b75 mm: reinstate ZER... |
3195 |
setpte: |
65500d234 [PATCH] mm: page ... |
3196 |
set_pte_at(mm, address, page_table, entry); |
1da177e4c Linux-2.6.12-rc2 |
3197 3198 |
/* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
3199 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
3200 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
3201 |
pte_unmap_unlock(page_table, ptl); |
83c54070e mm: fault feedbac... |
3202 |
return 0; |
8f4e2101f [PATCH] mm: page ... |
3203 |
release: |
8a9f3ccd2 Memory controller... |
3204 |
mem_cgroup_uncharge_page(page); |
8f4e2101f [PATCH] mm: page ... |
3205 3206 |
page_cache_release(page); goto unlock; |
8a9f3ccd2 Memory controller... |
3207 |
oom_free_page: |
6dbf6d3bb memcg: page_cache... |
3208 |
page_cache_release(page); |
65500d234 [PATCH] mm: page ... |
3209 |
oom: |
1da177e4c Linux-2.6.12-rc2 |
3210 3211 |
return VM_FAULT_OOM; } |
7eae74af3 mm: do_fault(): e... |
3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 |
static int __do_fault(struct vm_area_struct *vma, unsigned long address, pgoff_t pgoff, unsigned int flags, struct page **page) { struct vm_fault vmf; int ret; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = pgoff; vmf.flags = flags; vmf.page = NULL; ret = vma->vm_ops->fault(vma, &vmf); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) return ret; if (unlikely(PageHWPoison(vmf.page))) { if (ret & VM_FAULT_LOCKED) unlock_page(vmf.page); page_cache_release(vmf.page); return VM_FAULT_HWPOISON; } if (unlikely(!(ret & VM_FAULT_LOCKED))) lock_page(vmf.page); else VM_BUG_ON_PAGE(!PageLocked(vmf.page), vmf.page); *page = vmf.page; return ret; } |
8c6e50b02 mm: introduce vm_... |
3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 |
/** * 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 mm: consolidate c... |
3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 |
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); else if (pte_file(*pte) && pte_file_soft_dirty(*pte)) pte_mksoft_dirty(entry); if (anon) { inc_mm_counter_fast(vma->vm_mm, MM_ANONPAGES); page_add_new_anon_rmap(page, vma, address); } else { inc_mm_counter_fast(vma->vm_mm, MM_FILEPAGES); 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); } |
8c6e50b02 mm: introduce vm_... |
3280 |
#define FAULT_AROUND_ORDER 4 |
1592eef01 mm: add debugfs t... |
3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 |
#ifdef CONFIG_DEBUG_FS static unsigned int fault_around_order = FAULT_AROUND_ORDER; static int fault_around_order_get(void *data, u64 *val) { *val = fault_around_order; return 0; } static int fault_around_order_set(void *data, u64 val) { BUILD_BUG_ON((1UL << FAULT_AROUND_ORDER) > PTRS_PER_PTE); if (1UL << val > PTRS_PER_PTE) return -EINVAL; fault_around_order = val; return 0; } DEFINE_SIMPLE_ATTRIBUTE(fault_around_order_fops, fault_around_order_get, fault_around_order_set, "%llu "); static int __init fault_around_debugfs(void) { void *ret; ret = debugfs_create_file("fault_around_order", 0644, NULL, NULL, &fault_around_order_fops); if (!ret) pr_warn("Failed to create fault_around_order in debugfs"); return 0; } late_initcall(fault_around_debugfs); static inline unsigned long fault_around_pages(void) { return 1UL << fault_around_order; } static inline unsigned long fault_around_mask(void) { return ~((1UL << (PAGE_SHIFT + fault_around_order)) - 1); } #else static inline unsigned long fault_around_pages(void) { unsigned long nr_pages; nr_pages = 1UL << FAULT_AROUND_ORDER; BUILD_BUG_ON(nr_pages > PTRS_PER_PTE); return nr_pages; } static inline unsigned long fault_around_mask(void) { return ~((1UL << (PAGE_SHIFT + FAULT_AROUND_ORDER)) - 1); } #endif |
8c6e50b02 mm: introduce vm_... |
3339 3340 3341 3342 3343 3344 3345 3346 |
static void do_fault_around(struct vm_area_struct *vma, unsigned long address, pte_t *pte, pgoff_t pgoff, unsigned int flags) { unsigned long start_addr; pgoff_t max_pgoff; struct vm_fault vmf; int off; |
1592eef01 mm: add debugfs t... |
3347 |
start_addr = max(address & fault_around_mask(), vma->vm_start); |
8c6e50b02 mm: introduce vm_... |
3348 3349 3350 3351 3352 3353 |
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 |
1592eef01 mm: add debugfs t... |
3354 |
* or fault_around_pages() from pgoff, depending what is neast. |
8c6e50b02 mm: introduce vm_... |
3355 3356 3357 3358 |
*/ 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, |
1592eef01 mm: add debugfs t... |
3359 |
pgoff + fault_around_pages() - 1); |
8c6e50b02 mm: introduce vm_... |
3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 |
/* 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; vma->vm_ops->map_pages(vma, &vmf); } |
e655fb290 mm: introduce do_... |
3378 3379 3380 3381 3382 3383 |
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 mm: consolidate c... |
3384 |
pte_t *pte; |
8c6e50b02 mm: introduce vm_... |
3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 |
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). */ if (vma->vm_ops->map_pages) { 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 mm: introduce do_... |
3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 |
ret = __do_fault(vma, address, pgoff, flags, &fault_page); 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); page_cache_release(fault_page); return ret; } |
3bb977946 mm: consolidate c... |
3411 |
do_set_pte(vma, address, fault_page, pte, false, false); |
e655fb290 mm: introduce do_... |
3412 |
unlock_page(fault_page); |
8c6e50b02 mm: introduce vm_... |
3413 3414 |
unlock_out: pte_unmap_unlock(pte, ptl); |
e655fb290 mm: introduce do_... |
3415 3416 |
return ret; } |
ec47c3b95 mm: introduce do_... |
3417 3418 3419 3420 3421 3422 |
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; spinlock_t *ptl; |
3bb977946 mm: consolidate c... |
3423 |
pte_t *pte; |
ec47c3b95 mm: introduce do_... |
3424 3425 3426 3427 3428 3429 3430 3431 |
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; |
d715ae08f memcg: rename hig... |
3432 |
if (mem_cgroup_charge_anon(new_page, mm, GFP_KERNEL)) { |
ec47c3b95 mm: introduce do_... |
3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 |
page_cache_release(new_page); return VM_FAULT_OOM; } ret = __do_fault(vma, address, pgoff, flags, &fault_page); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) goto uncharge_out; copy_user_highpage(new_page, fault_page, address, vma); __SetPageUptodate(new_page); 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); page_cache_release(fault_page); goto uncharge_out; } |
3bb977946 mm: consolidate c... |
3451 |
do_set_pte(vma, address, new_page, pte, true, true); |
ec47c3b95 mm: introduce do_... |
3452 3453 3454 3455 3456 3457 3458 3459 3460 |
pte_unmap_unlock(pte, ptl); unlock_page(fault_page); page_cache_release(fault_page); return ret; uncharge_out: mem_cgroup_uncharge_page(new_page); page_cache_release(new_page); return ret; } |
f0c6d4d29 mm: introduce do_... |
3461 |
static int do_shared_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
16abfa086 Fix sys_remap_fil... |
3462 |
unsigned long address, pmd_t *pmd, |
54cb8821d mm: merge populat... |
3463 |
pgoff_t pgoff, unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
3464 |
{ |
f0c6d4d29 mm: introduce do_... |
3465 3466 |
struct page *fault_page; struct address_space *mapping; |
8f4e2101f [PATCH] mm: page ... |
3467 |
spinlock_t *ptl; |
3bb977946 mm: consolidate c... |
3468 |
pte_t *pte; |
f0c6d4d29 mm: introduce do_... |
3469 |
int dirtied = 0; |
f0c6d4d29 mm: introduce do_... |
3470 |
int ret, tmp; |
1d65f86db mm: preallocate p... |
3471 |
|
7eae74af3 mm: do_fault(): e... |
3472 3473 |
ret = __do_fault(vma, address, pgoff, flags, &fault_page); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) |
f0c6d4d29 mm: introduce do_... |
3474 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
3475 3476 |
/* |
f0c6d4d29 mm: introduce do_... |
3477 3478 |
* Check if the backing address space wants to know that the page is * about to become writable |
1da177e4c Linux-2.6.12-rc2 |
3479 |
*/ |
fb09a4642 mm: consolidate c... |
3480 3481 3482 3483 3484 |
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)))) { |
f0c6d4d29 mm: introduce do_... |
3485 |
page_cache_release(fault_page); |
fb09a4642 mm: consolidate c... |
3486 |
return tmp; |
4294621f4 [PATCH] mm: rss =... |
3487 |
} |
fb09a4642 mm: consolidate c... |
3488 |
} |
f0c6d4d29 mm: introduce do_... |
3489 3490 3491 3492 3493 3494 |
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); page_cache_release(fault_page); return ret; |
1da177e4c Linux-2.6.12-rc2 |
3495 |
} |
3bb977946 mm: consolidate c... |
3496 |
do_set_pte(vma, address, fault_page, pte, true, false); |
f0c6d4d29 mm: introduce do_... |
3497 |
pte_unmap_unlock(pte, ptl); |
b827e496c mm: close page_mk... |
3498 |
|
f0c6d4d29 mm: introduce do_... |
3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 |
if (set_page_dirty(fault_page)) dirtied = 1; mapping = fault_page->mapping; 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 [PATCH] mm: track... |
3509 |
} |
d00806b18 mm: fix fault vs ... |
3510 |
|
f0c6d4d29 mm: introduce do_... |
3511 3512 3513 |
/* file_update_time outside page_lock */ if (vma->vm_file && !vma->vm_ops->page_mkwrite) file_update_time(vma->vm_file); |
b827e496c mm: close page_mk... |
3514 |
|
1d65f86db mm: preallocate p... |
3515 |
return ret; |
54cb8821d mm: merge populat... |
3516 |
} |
d00806b18 mm: fix fault vs ... |
3517 |
|
54cb8821d mm: merge populat... |
3518 3519 |
static int do_linear_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f Remove internal u... |
3520 |
unsigned int flags, pte_t orig_pte) |
54cb8821d mm: merge populat... |
3521 3522 |
{ pgoff_t pgoff = (((address & PAGE_MASK) |
0da7e01f5 calculation of pg... |
3523 |
- vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; |
54cb8821d mm: merge populat... |
3524 |
|
16abfa086 Fix sys_remap_fil... |
3525 |
pte_unmap(page_table); |
e655fb290 mm: introduce do_... |
3526 3527 3528 |
if (!(flags & FAULT_FLAG_WRITE)) return do_read_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
ec47c3b95 mm: introduce do_... |
3529 3530 3531 |
if (!(vma->vm_flags & VM_SHARED)) return do_cow_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
f0c6d4d29 mm: introduce do_... |
3532 |
return do_shared_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
54cb8821d mm: merge populat... |
3533 |
} |
f4b81804a [PATCH] do_no_pfn() |
3534 |
/* |
1da177e4c Linux-2.6.12-rc2 |
3535 3536 3537 |
* Fault of a previously existing named mapping. Repopulate the pte * from the encoded file_pte if possible. This enables swappable * nonlinear vmas. |
8f4e2101f [PATCH] mm: page ... |
3538 3539 3540 3541 |
* * 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 Linux-2.6.12-rc2 |
3542 |
*/ |
d0217ac04 mm: fault feedbac... |
3543 |
static int do_nonlinear_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
65500d234 [PATCH] mm: page ... |
3544 |
unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f Remove internal u... |
3545 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
3546 |
{ |
65500d234 [PATCH] mm: page ... |
3547 |
pgoff_t pgoff; |
1da177e4c Linux-2.6.12-rc2 |
3548 |
|
30c9f3a9f Remove internal u... |
3549 |
flags |= FAULT_FLAG_NONLINEAR; |
4c21e2f24 [PATCH] mm: split... |
3550 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
83c54070e mm: fault feedbac... |
3551 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3552 |
|
2509ef26d badpage: zap prin... |
3553 |
if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) { |
65500d234 [PATCH] mm: page ... |
3554 3555 3556 |
/* * Page table corrupted: show pte and kill process. */ |
3dc147414 badpage: replace ... |
3557 |
print_bad_pte(vma, address, orig_pte, NULL); |
d99be1a8e mm: sigbus instea... |
3558 |
return VM_FAULT_SIGBUS; |
65500d234 [PATCH] mm: page ... |
3559 |
} |
65500d234 [PATCH] mm: page ... |
3560 3561 |
pgoff = pte_to_pgoff(orig_pte); |
e655fb290 mm: introduce do_... |
3562 3563 3564 |
if (!(flags & FAULT_FLAG_WRITE)) return do_read_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
ec47c3b95 mm: introduce do_... |
3565 3566 3567 |
if (!(vma->vm_flags & VM_SHARED)) return do_cow_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
f0c6d4d29 mm: introduce do_... |
3568 |
return do_shared_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
1da177e4c Linux-2.6.12-rc2 |
3569 |
} |
b19a99392 mm/memory.c: mark... |
3570 |
static int numa_migrate_prep(struct page *page, struct vm_area_struct *vma, |
04bb2f947 sched/numa: Adjus... |
3571 3572 |
unsigned long addr, int page_nid, int *flags) |
9532fec11 mm: numa: Migrate... |
3573 3574 3575 3576 |
{ get_page(page); count_vm_numa_event(NUMA_HINT_FAULTS); |
04bb2f947 sched/numa: Adjus... |
3577 |
if (page_nid == numa_node_id()) { |
9532fec11 mm: numa: Migrate... |
3578 |
count_vm_numa_event(NUMA_HINT_FAULTS_LOCAL); |
04bb2f947 sched/numa: Adjus... |
3579 3580 |
*flags |= TNF_FAULT_LOCAL; } |
9532fec11 mm: numa: Migrate... |
3581 3582 3583 |
return mpol_misplaced(page, vma, addr); } |
b19a99392 mm/memory.c: mark... |
3584 |
static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, |
d10e63f29 mm: numa: Create ... |
3585 3586 |
unsigned long addr, pte_t pte, pte_t *ptep, pmd_t *pmd) { |
4daae3b4b mm: mempolicy: Us... |
3587 |
struct page *page = NULL; |
d10e63f29 mm: numa: Create ... |
3588 |
spinlock_t *ptl; |
8191acbd3 mm: numa: Sanitiz... |
3589 |
int page_nid = -1; |
90572890d mm: numa: Change ... |
3590 |
int last_cpupid; |
cbee9f88e mm: numa: Add fau... |
3591 |
int target_nid; |
b8593bfda mm: sched: Adapt ... |
3592 |
bool migrated = false; |
6688cc054 mm: numa: Do not ... |
3593 |
int flags = 0; |
d10e63f29 mm: numa: Create ... |
3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 |
/* * 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. * * ptep_modify_prot_start is not called as this is clearing * the _PAGE_NUMA bit and it is not really expected that there * would be concurrent hardware modifications to the PTE. */ ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
4daae3b4b mm: mempolicy: Us... |
3606 3607 3608 3609 |
if (unlikely(!pte_same(*ptep, pte))) { pte_unmap_unlock(ptep, ptl); goto out; } |
d10e63f29 mm: numa: Create ... |
3610 3611 3612 3613 3614 3615 3616 3617 3618 |
pte = pte_mknonnuma(pte); 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; } |
a1a46184e mm: numa: Do not ... |
3619 |
BUG_ON(is_zero_pfn(page_to_pfn(page))); |
d10e63f29 mm: numa: Create ... |
3620 |
|
6688cc054 mm: numa: Do not ... |
3621 3622 3623 3624 3625 3626 3627 |
/* * Avoid grouping on DSO/COW pages in specific and RO pages * in general, RO pages shouldn't hurt as much anyway since * they can be in shared cache state. */ if (!pte_write(pte)) flags |= TNF_NO_GROUP; |
dabe1d992 sched/numa: Be mo... |
3628 3629 3630 3631 3632 3633 |
/* * 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 mm: numa: Change ... |
3634 |
last_cpupid = page_cpupid_last(page); |
8191acbd3 mm: numa: Sanitiz... |
3635 |
page_nid = page_to_nid(page); |
04bb2f947 sched/numa: Adjus... |
3636 |
target_nid = numa_migrate_prep(page, vma, addr, page_nid, &flags); |
d10e63f29 mm: numa: Create ... |
3637 |
pte_unmap_unlock(ptep, ptl); |
4daae3b4b mm: mempolicy: Us... |
3638 |
if (target_nid == -1) { |
4daae3b4b mm: mempolicy: Us... |
3639 3640 3641 3642 3643 |
put_page(page); goto out; } /* Migrate to the requested node */ |
1bc115d87 mm: numa: Scan pa... |
3644 |
migrated = migrate_misplaced_page(page, vma, target_nid); |
6688cc054 mm: numa: Do not ... |
3645 |
if (migrated) { |
8191acbd3 mm: numa: Sanitiz... |
3646 |
page_nid = target_nid; |
6688cc054 mm: numa: Do not ... |
3647 3648 |
flags |= TNF_MIGRATED; } |
4daae3b4b mm: mempolicy: Us... |
3649 3650 |
out: |
8191acbd3 mm: numa: Sanitiz... |
3651 |
if (page_nid != -1) |
6688cc054 mm: numa: Do not ... |
3652 |
task_numa_fault(last_cpupid, page_nid, 1, flags); |
d10e63f29 mm: numa: Create ... |
3653 3654 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
3655 3656 3657 3658 3659 3660 3661 3662 3663 |
/* * 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 [PATCH] mm: ptd_a... |
3664 3665 3666 |
* 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 Linux-2.6.12-rc2 |
3667 |
*/ |
c02925540 thp: consolidate ... |
3668 |
static int handle_pte_fault(struct mm_struct *mm, |
71e3aac07 thp: transparent ... |
3669 3670 |
struct vm_area_struct *vma, unsigned long address, pte_t *pte, pmd_t *pmd, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3671 3672 |
{ pte_t entry; |
8f4e2101f [PATCH] mm: page ... |
3673 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
3674 |
|
8dab5241d Rework ptep_set_a... |
3675 |
entry = *pte; |
1da177e4c Linux-2.6.12-rc2 |
3676 |
if (!pte_present(entry)) { |
65500d234 [PATCH] mm: page ... |
3677 |
if (pte_none(entry)) { |
f4b81804a [PATCH] do_no_pfn() |
3678 |
if (vma->vm_ops) { |
3c18ddd16 mm: remove nopage |
3679 |
if (likely(vma->vm_ops->fault)) |
54cb8821d mm: merge populat... |
3680 |
return do_linear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3681 |
pte, pmd, flags, entry); |
f4b81804a [PATCH] do_no_pfn() |
3682 3683 |
} return do_anonymous_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3684 |
pte, pmd, flags); |
65500d234 [PATCH] mm: page ... |
3685 |
} |
1da177e4c Linux-2.6.12-rc2 |
3686 |
if (pte_file(entry)) |
d0217ac04 mm: fault feedbac... |
3687 |
return do_nonlinear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3688 |
pte, pmd, flags, entry); |
65500d234 [PATCH] mm: page ... |
3689 |
return do_swap_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3690 |
pte, pmd, flags, entry); |
1da177e4c Linux-2.6.12-rc2 |
3691 |
} |
d10e63f29 mm: numa: Create ... |
3692 3693 |
if (pte_numa(entry)) return do_numa_page(mm, vma, address, entry, pte, pmd); |
4c21e2f24 [PATCH] mm: split... |
3694 |
ptl = pte_lockptr(mm, pmd); |
8f4e2101f [PATCH] mm: page ... |
3695 3696 3697 |
spin_lock(ptl); if (unlikely(!pte_same(*pte, entry))) goto unlock; |
30c9f3a9f Remove internal u... |
3698 |
if (flags & FAULT_FLAG_WRITE) { |
1da177e4c Linux-2.6.12-rc2 |
3699 |
if (!pte_write(entry)) |
8f4e2101f [PATCH] mm: page ... |
3700 3701 |
return do_wp_page(mm, vma, address, pte, pmd, ptl, entry); |
1da177e4c Linux-2.6.12-rc2 |
3702 3703 3704 |
entry = pte_mkdirty(entry); } entry = pte_mkyoung(entry); |
30c9f3a9f Remove internal u... |
3705 |
if (ptep_set_access_flags(vma, address, pte, entry, flags & FAULT_FLAG_WRITE)) { |
4b3073e1c MM: Pass a PTE po... |
3706 |
update_mmu_cache(vma, address, pte); |
1a44e1490 [PATCH] .text pag... |
3707 3708 3709 3710 3711 3712 3713 |
} 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 Remove internal u... |
3714 |
if (flags & FAULT_FLAG_WRITE) |
61c77326d x86, mm: Avoid un... |
3715 |
flush_tlb_fix_spurious_fault(vma, address); |
1a44e1490 [PATCH] .text pag... |
3716 |
} |
8f4e2101f [PATCH] mm: page ... |
3717 3718 |
unlock: pte_unmap_unlock(pte, ptl); |
83c54070e mm: fault feedbac... |
3719 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3720 3721 3722 3723 3724 |
} /* * By the time we get here, we already hold the mm semaphore */ |
519e52473 mm: memcg: enable... |
3725 3726 |
static int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3727 3728 3729 3730 3731 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; |
ac9b9c667 [PATCH] Fix handl... |
3732 |
if (unlikely(is_vm_hugetlb_page(vma))) |
30c9f3a9f Remove internal u... |
3733 |
return hugetlb_fault(mm, vma, address, flags); |
1da177e4c Linux-2.6.12-rc2 |
3734 |
|
1da177e4c Linux-2.6.12-rc2 |
3735 |
pgd = pgd_offset(mm, address); |
1da177e4c Linux-2.6.12-rc2 |
3736 3737 |
pud = pud_alloc(mm, pgd, address); if (!pud) |
c74df32c7 [PATCH] mm: ptd_a... |
3738 |
return VM_FAULT_OOM; |
1da177e4c Linux-2.6.12-rc2 |
3739 3740 |
pmd = pmd_alloc(mm, pud, address); if (!pmd) |
c74df32c7 [PATCH] mm: ptd_a... |
3741 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3742 |
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma)) { |
c02925540 thp: consolidate ... |
3743 |
int ret = VM_FAULT_FALLBACK; |
71e3aac07 thp: transparent ... |
3744 |
if (!vma->vm_ops) |
c02925540 thp: consolidate ... |
3745 3746 3747 3748 |
ret = do_huge_pmd_anonymous_page(mm, vma, address, pmd, flags); if (!(ret & VM_FAULT_FALLBACK)) return ret; |
71e3aac07 thp: transparent ... |
3749 3750 |
} else { pmd_t orig_pmd = *pmd; |
1f1d06c34 thp, memcg: split... |
3751 |
int ret; |
71e3aac07 thp: transparent ... |
3752 3753 |
barrier(); if (pmd_trans_huge(orig_pmd)) { |
a1dd450bc mm: thp: set the ... |
3754 |
unsigned int dirty = flags & FAULT_FLAG_WRITE; |
e53289c0c mm: reinstante dr... |
3755 3756 3757 3758 3759 3760 3761 |
/* * If the pmd is splitting, return and retry the * the fault. Alternative: wait until the split * is done, and goto retry. */ if (pmd_trans_splitting(orig_pmd)) return 0; |
3d59eebc5 Merge tag 'balanc... |
3762 |
if (pmd_numa(orig_pmd)) |
4daae3b4b mm: mempolicy: Us... |
3763 |
return do_huge_pmd_numa_page(mm, vma, address, |
d10e63f29 mm: numa: Create ... |
3764 |
orig_pmd, pmd); |
3d59eebc5 Merge tag 'balanc... |
3765 |
if (dirty && !pmd_write(orig_pmd)) { |
1f1d06c34 thp, memcg: split... |
3766 3767 |
ret = do_huge_pmd_wp_page(mm, vma, address, pmd, orig_pmd); |
9845cbbd1 mm, thp: fix infi... |
3768 3769 |
if (!(ret & VM_FAULT_FALLBACK)) return ret; |
a1dd450bc mm: thp: set the ... |
3770 3771 3772 |
} else { huge_pmd_set_accessed(mm, vma, address, pmd, orig_pmd, dirty); |
9845cbbd1 mm, thp: fix infi... |
3773 |
return 0; |
1f1d06c34 thp, memcg: split... |
3774 |
} |
71e3aac07 thp: transparent ... |
3775 3776 |
} } |
0f19c1792 mm: numa: Do not ... |
3777 3778 |
/* THP should already have been handled */ BUG_ON(pmd_numa(*pmd)); |
d10e63f29 mm: numa: Create ... |
3779 |
|
71e3aac07 thp: transparent ... |
3780 3781 3782 3783 3784 |
/* * Use __pte_alloc instead of pte_alloc_map, because we can't * run pte_offset_map on the pmd, if an huge pmd could * materialize from under us from a different thread. */ |
4fd017708 mm: Check if PTE ... |
3785 3786 |
if (unlikely(pmd_none(*pmd)) && unlikely(__pte_alloc(mm, vma, pmd, address))) |
c74df32c7 [PATCH] mm: ptd_a... |
3787 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 |
/* if an huge pmd materialized from under us just retry later */ if (unlikely(pmd_trans_huge(*pmd))) 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 Linux-2.6.12-rc2 |
3798 |
|
30c9f3a9f Remove internal u... |
3799 |
return handle_pte_fault(mm, vma, address, pte, pmd, flags); |
1da177e4c Linux-2.6.12-rc2 |
3800 |
} |
519e52473 mm: memcg: enable... |
3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 |
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 mm: memcg: handle... |
3819 |
mem_cgroup_oom_enable(); |
519e52473 mm: memcg: enable... |
3820 3821 |
ret = __handle_mm_fault(mm, vma, address, flags); |
494264208 mm: memcg: handle... |
3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 |
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 mm: memcg: do not... |
3833 |
|
519e52473 mm: memcg: enable... |
3834 3835 |
return ret; } |
1da177e4c Linux-2.6.12-rc2 |
3836 3837 3838 |
#ifndef __PAGETABLE_PUD_FOLDED /* * Allocate page upper directory. |
872fec16d [PATCH] mm: init_... |
3839 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3840 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3841 |
int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3842 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3843 3844 |
pud_t *new = pud_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3845 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3846 |
|
362a61ad6 fix SMP data race... |
3847 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3848 |
spin_lock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3849 |
if (pgd_present(*pgd)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3850 |
pud_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3851 3852 |
else pgd_populate(mm, pgd, new); |
c74df32c7 [PATCH] mm: ptd_a... |
3853 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3854 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3855 3856 3857 3858 3859 3860 |
} #endif /* __PAGETABLE_PUD_FOLDED */ #ifndef __PAGETABLE_PMD_FOLDED /* * Allocate page middle directory. |
872fec16d [PATCH] mm: init_... |
3861 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3862 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3863 |
int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3864 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3865 3866 |
pmd_t *new = pmd_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3867 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3868 |
|
362a61ad6 fix SMP data race... |
3869 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3870 |
spin_lock(&mm->page_table_lock); |
1da177e4c Linux-2.6.12-rc2 |
3871 |
#ifndef __ARCH_HAS_4LEVEL_HACK |
1bb3630e8 [PATCH] mm: ptd_a... |
3872 |
if (pud_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3873 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3874 3875 |
else pud_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3876 |
#else |
1bb3630e8 [PATCH] mm: ptd_a... |
3877 |
if (pgd_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3878 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3879 3880 |
else pgd_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3881 |
#endif /* __ARCH_HAS_4LEVEL_HACK */ |
c74df32c7 [PATCH] mm: ptd_a... |
3882 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3883 |
return 0; |
e0f39591c [PATCH] Workaroun... |
3884 |
} |
1da177e4c Linux-2.6.12-rc2 |
3885 |
#endif /* __PAGETABLE_PMD_FOLDED */ |
1da177e4c Linux-2.6.12-rc2 |
3886 3887 3888 |
#if !defined(__HAVE_ARCH_GATE_AREA) #if defined(AT_SYSINFO_EHDR) |
5ce7852cd [PATCH] mm/filema... |
3889 |
static struct vm_area_struct gate_vma; |
1da177e4c Linux-2.6.12-rc2 |
3890 3891 3892 3893 3894 3895 |
static int __init gate_vma_init(void) { gate_vma.vm_mm = NULL; gate_vma.vm_start = FIXADDR_USER_START; gate_vma.vm_end = FIXADDR_USER_END; |
b6558c4a2 [PATCH] Fix gate_... |
3896 3897 |
gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC; gate_vma.vm_page_prot = __P101; |
909af768e coredump: remove ... |
3898 |
|
1da177e4c Linux-2.6.12-rc2 |
3899 3900 3901 3902 |
return 0; } __initcall(gate_vma_init); #endif |
31db58b3a mm: arch: make ge... |
3903 |
struct vm_area_struct *get_gate_vma(struct mm_struct *mm) |
1da177e4c Linux-2.6.12-rc2 |
3904 3905 3906 3907 3908 3909 3910 |
{ #ifdef AT_SYSINFO_EHDR return &gate_vma; #else return NULL; #endif } |
cae5d3903 mm: arch: rename ... |
3911 |
int in_gate_area_no_mm(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
3912 3913 3914 3915 3916 3917 3918 3919 3920 |
{ #ifdef AT_SYSINFO_EHDR if ((addr >= FIXADDR_USER_START) && (addr < FIXADDR_USER_END)) return 1; #endif return 0; } #endif /* __HAVE_ARCH_GATE_AREA */ |
0ec76a110 [PATCH] NOMMU: Ch... |
3921 |
|
1b36ba815 mm: wrap follow_p... |
3922 |
static int __follow_pte(struct mm_struct *mm, unsigned long address, |
f8ad0f499 mm: introduce fol... |
3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 |
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 thp: verify pmd_t... |
3939 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
f8ad0f499 mm: introduce fol... |
3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 |
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 mm: wrap follow_p... |
3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 |
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 mm: introduce fol... |
3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 |
/** * 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 access_process_vm... |
3997 |
#ifdef CONFIG_HAVE_IOREMAP_PROT |
d87fe6607 x86: PAT: modify ... |
3998 3999 4000 |
int follow_phys(struct vm_area_struct *vma, unsigned long address, unsigned int flags, unsigned long *prot, resource_size_t *phys) |
28b2ee20c access_process_vm... |
4001 |
{ |
03668a4de mm: use generic f... |
4002 |
int ret = -EINVAL; |
28b2ee20c access_process_vm... |
4003 4004 |
pte_t *ptep, pte; spinlock_t *ptl; |
28b2ee20c access_process_vm... |
4005 |
|
d87fe6607 x86: PAT: modify ... |
4006 4007 |
if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) goto out; |
28b2ee20c access_process_vm... |
4008 |
|
03668a4de mm: use generic f... |
4009 |
if (follow_pte(vma->vm_mm, address, &ptep, &ptl)) |
d87fe6607 x86: PAT: modify ... |
4010 |
goto out; |
28b2ee20c access_process_vm... |
4011 |
pte = *ptep; |
03668a4de mm: use generic f... |
4012 |
|
28b2ee20c access_process_vm... |
4013 4014 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
28b2ee20c access_process_vm... |
4015 4016 |
*prot = pgprot_val(pte_pgprot(pte)); |
03668a4de mm: use generic f... |
4017 |
*phys = (resource_size_t)pte_pfn(pte) << PAGE_SHIFT; |
28b2ee20c access_process_vm... |
4018 |
|
03668a4de mm: use generic f... |
4019 |
ret = 0; |
28b2ee20c access_process_vm... |
4020 4021 4022 |
unlock: pte_unmap_unlock(ptep, ptl); out: |
d87fe6607 x86: PAT: modify ... |
4023 |
return ret; |
28b2ee20c access_process_vm... |
4024 4025 4026 4027 4028 4029 4030 |
} 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 mm: make maddr __... |
4031 |
void __iomem *maddr; |
28b2ee20c access_process_vm... |
4032 |
int offset = addr & (PAGE_SIZE-1); |
d87fe6607 x86: PAT: modify ... |
4033 |
if (follow_phys(vma, addr, write, &prot, &phys_addr)) |
28b2ee20c access_process_vm... |
4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 |
return -EINVAL; maddr = ioremap_prot(phys_addr, PAGE_SIZE, prot); if (write) memcpy_toio(maddr + offset, buf, len); else memcpy_fromio(buf, maddr + offset, len); iounmap(maddr); return len; } |
5a73633ef mm: make generic_... |
4045 |
EXPORT_SYMBOL_GPL(generic_access_phys); |
28b2ee20c access_process_vm... |
4046 |
#endif |
0ec76a110 [PATCH] NOMMU: Ch... |
4047 |
/* |
206cb6365 mm: factor out ma... |
4048 4049 |
* Access another process' address space as given in mm. If non-NULL, use the * given task for page fault accounting. |
0ec76a110 [PATCH] NOMMU: Ch... |
4050 |
*/ |
206cb6365 mm: factor out ma... |
4051 4052 |
static int __access_remote_vm(struct task_struct *tsk, struct mm_struct *mm, unsigned long addr, void *buf, int len, int write) |
0ec76a110 [PATCH] NOMMU: Ch... |
4053 |
{ |
0ec76a110 [PATCH] NOMMU: Ch... |
4054 |
struct vm_area_struct *vma; |
0ec76a110 [PATCH] NOMMU: Ch... |
4055 |
void *old_buf = buf; |
0ec76a110 [PATCH] NOMMU: Ch... |
4056 |
down_read(&mm->mmap_sem); |
183ff22bb spelling fixes: mm/ |
4057 |
/* ignore errors, just check how much was successfully transferred */ |
0ec76a110 [PATCH] NOMMU: Ch... |
4058 4059 4060 |
while (len) { int bytes, ret, offset; void *maddr; |
28b2ee20c access_process_vm... |
4061 |
struct page *page = NULL; |
0ec76a110 [PATCH] NOMMU: Ch... |
4062 4063 4064 |
ret = get_user_pages(tsk, mm, addr, 1, write, 1, &page, &vma); |
28b2ee20c access_process_vm... |
4065 4066 4067 4068 4069 4070 4071 |
if (ret <= 0) { /* * Check if this is a VM_IO | VM_PFNMAP VMA, which * we can access using slightly different code. */ #ifdef CONFIG_HAVE_IOREMAP_PROT vma = find_vma(mm, addr); |
fe936dfc2 mm: check that we... |
4072 |
if (!vma || vma->vm_start > addr) |
28b2ee20c access_process_vm... |
4073 4074 4075 4076 4077 4078 4079 4080 |
break; if (vma->vm_ops && vma->vm_ops->access) ret = vma->vm_ops->access(vma, addr, buf, len, write); if (ret <= 0) #endif break; bytes = ret; |
0ec76a110 [PATCH] NOMMU: Ch... |
4081 |
} else { |
28b2ee20c access_process_vm... |
4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 |
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); page_cache_release(page); |
0ec76a110 [PATCH] NOMMU: Ch... |
4098 |
} |
0ec76a110 [PATCH] NOMMU: Ch... |
4099 4100 4101 4102 4103 |
len -= bytes; buf += bytes; addr += bytes; } up_read(&mm->mmap_sem); |
0ec76a110 [PATCH] NOMMU: Ch... |
4104 4105 4106 |
return buf - old_buf; } |
03252919b x86: print which ... |
4107 |
|
5ddd36b9c mm: implement acc... |
4108 |
/** |
ae91dbfc9 mm: fix memory.c ... |
4109 |
* access_remote_vm - access another process' address space |
5ddd36b9c mm: implement acc... |
4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 |
* @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 x86: print which ... |
4123 |
/* |
206cb6365 mm: factor out ma... |
4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 |
* 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 x86: print which ... |
4143 4144 4145 4146 4147 4148 4149 |
/* * 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 x86: fix "BUG: sl... |
4150 4151 4152 4153 4154 4155 |
/* * Do not print if we are in atomic * contexts (in exception stacks, etc.): */ if (preempt_count()) return; |
03252919b x86: print which ... |
4156 4157 4158 4159 4160 4161 |
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 mm: use kbasename() |
4162 |
char *p; |
03252919b x86: print which ... |
4163 |
|
cf28b4863 d_path: Make d_pa... |
4164 |
p = d_path(&f->f_path, buf, PAGE_SIZE); |
03252919b x86: print which ... |
4165 4166 |
if (IS_ERR(p)) p = "?"; |
2fbc57c53 mm: use kbasename() |
4167 |
printk("%s%s[%lx+%lx]", prefix, kbasename(p), |
03252919b x86: print which ... |
4168 4169 4170 4171 4172 |
vma->vm_start, vma->vm_end - vma->vm_start); free_page((unsigned long)buf); } } |
51a07e50b mm/memory.c:print... |
4173 |
up_read(&mm->mmap_sem); |
03252919b x86: print which ... |
4174 |
} |
3ee1afa30 x86: some lock an... |
4175 |
|
662bbcb27 mm, sched: Allow ... |
4176 |
#if defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_DEBUG_ATOMIC_SLEEP) |
3ee1afa30 x86: some lock an... |
4177 4178 |
void might_fault(void) { |
95156f005 lockdep, mm: fix ... |
4179 4180 4181 4182 4183 4184 4185 4186 |
/* * 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; |
3ee1afa30 x86: some lock an... |
4187 4188 4189 4190 4191 |
/* * it would be nicer only to annotate paths which are not under * pagefault_disable, however that requires a larger audit and * providing helpers like get_user_atomic. */ |
662bbcb27 mm, sched: Allow ... |
4192 4193 4194 4195 4196 4197 |
if (in_atomic()) return; __might_sleep(__FILE__, __LINE__, 0); if (current->mm) |
3ee1afa30 x86: some lock an... |
4198 4199 4200 4201 |
might_lock_read(¤t->mm->mmap_sem); } EXPORT_SYMBOL(might_fault); #endif |
47ad8475c thp: clear_copy_h... |
4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 |
#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 mm: dynamically a... |
4273 |
|
40b64acd1 mm: fix build of ... |
4274 |
#if USE_SPLIT_PTE_PTLOCKS && ALLOC_SPLIT_PTLOCKS |
b35f1819a mm: create a sepa... |
4275 4276 4277 4278 4279 4280 4281 4282 |
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 mm: properly sepa... |
4283 |
bool ptlock_alloc(struct page *page) |
49076ec2c mm: dynamically a... |
4284 4285 |
{ spinlock_t *ptl; |
b35f1819a mm: create a sepa... |
4286 |
ptl = kmem_cache_alloc(page_ptl_cachep, GFP_KERNEL); |
49076ec2c mm: dynamically a... |
4287 4288 |
if (!ptl) return false; |
539edb584 mm: properly sepa... |
4289 |
page->ptl = ptl; |
49076ec2c mm: dynamically a... |
4290 4291 |
return true; } |
539edb584 mm: properly sepa... |
4292 |
void ptlock_free(struct page *page) |
49076ec2c mm: dynamically a... |
4293 |
{ |
b35f1819a mm: create a sepa... |
4294 |
kmem_cache_free(page_ptl_cachep, page->ptl); |
49076ec2c mm: dynamically a... |
4295 4296 |
} #endif |