Blame view
mm/memory.c
108 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> |
1da177e4c Linux-2.6.12-rc2 |
60 |
|
6952b61de headers: taskstat... |
61 |
#include <asm/io.h> |
1da177e4c Linux-2.6.12-rc2 |
62 63 64 65 66 |
#include <asm/pgalloc.h> #include <asm/uaccess.h> #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/pgtable.h> |
42b777281 mm: remove double... |
67 |
#include "internal.h" |
d41dee369 [PATCH] sparsemem... |
68 |
#ifndef CONFIG_NEED_MULTIPLE_NODES |
1da177e4c Linux-2.6.12-rc2 |
69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 |
/* use the per-pgdat data instead for discontigmem - mbligh */ unsigned long max_mapnr; struct page *mem_map; EXPORT_SYMBOL(max_mapnr); EXPORT_SYMBOL(mem_map); #endif unsigned long num_physpages; /* * 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 |
86 87 88 |
EXPORT_SYMBOL(num_physpages); EXPORT_SYMBOL(high_memory); |
1da177e4c Linux-2.6.12-rc2 |
89 |
|
32a932332 brk randomization... |
90 91 92 93 94 95 96 97 98 99 100 101 |
/* * 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... |
102 103 104 105 |
static int __init disable_randmaps(char *s) { randomize_va_space = 0; |
9b41046cd [PATCH] Don't pas... |
106 |
return 1; |
a62eaf151 [PATCH] x86_64: A... |
107 108 |
} __setup("norandmaps", disable_randmaps); |
62eede62d mm: ZERO_PAGE wit... |
109 |
unsigned long zero_pfn __read_mostly; |
03f6462a3 mm: move highest_... |
110 |
unsigned long highest_memmap_pfn __read_mostly; |
a13ea5b75 mm: reinstate ZER... |
111 112 113 114 115 116 117 118 119 120 |
/* * 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... |
121 |
|
d559db086 mm: clean up mm_c... |
122 |
|
34e55232e mm: avoid false s... |
123 |
#if defined(SPLIT_RSS_COUNTING) |
ea48cf786 mm, counters: fol... |
124 |
void sync_mm_rss(struct mm_struct *mm) |
34e55232e mm: avoid false s... |
125 126 127 128 |
{ int i; for (i = 0; i < NR_MM_COUNTERS; i++) { |
05af2e104 mm, counters: rem... |
129 130 131 |
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... |
132 133 |
} } |
05af2e104 mm, counters: rem... |
134 |
current->rss_stat.events = 0; |
34e55232e mm: avoid false s... |
135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 |
} 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... |
156 |
sync_mm_rss(task->mm); |
34e55232e mm: avoid false s... |
157 |
} |
9547d01bf mm: uninline larg... |
158 |
#else /* SPLIT_RSS_COUNTING */ |
34e55232e mm: avoid false s... |
159 160 161 162 163 164 165 |
#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... |
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 |
#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; } batch = (void *)__get_free_pages(GFP_NOWAIT | __GFP_NOWARN, 0); if (!batch) return 0; 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). */ void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, bool fullmm) { tlb->mm = mm; tlb->fullmm = fullmm; tlb->need_flush = 0; tlb->fast_mode = (num_possible_cpus() == 1); tlb->local.next = NULL; tlb->local.nr = 0; tlb->local.max = ARRAY_SIZE(tlb->__pages); tlb->active = &tlb->local; #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... |
226 |
#endif |
9547d01bf mm: uninline larg... |
227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 |
if (tlb_fast_mode(tlb)) return; 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... |
266 |
VM_BUG_ON(!tlb->need_flush); |
9547d01bf mm: uninline larg... |
267 268 269 270 271 272 273 274 275 276 277 |
if (tlb_fast_mode(tlb)) { free_page_and_swap_cache(page); return 1; /* avoid calling tlb_flush_mmu() */ } 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... |
279 280 281 282 283 284 285 |
} VM_BUG_ON(batch->nr > batch->max); 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 |
362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 |
/* * If a p?d_bad entry is found while walking page tables, report * the error, before resetting entry to p?d_none. Usually (but * very seldom) called out from the p?d_none_or_clear_bad macros. */ void pgd_clear_bad(pgd_t *pgd) { pgd_ERROR(*pgd); pgd_clear(pgd); } void pud_clear_bad(pud_t *pud) { pud_ERROR(*pud); pud_clear(pud); } void pmd_clear_bad(pmd_t *pmd) { pmd_ERROR(*pmd); pmd_clear(pmd); } /* * Note: this doesn't free the actual pages themselves. That * has been handled earlier when unmapping all the memory regions. */ |
9e1b32caa mm: Pass virtual ... |
390 391 |
static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
392 |
{ |
2f569afd9 CONFIG_HIGHPTE vs... |
393 |
pgtable_t token = pmd_pgtable(*pmd); |
e0da382c9 [PATCH] freepgt: ... |
394 |
pmd_clear(pmd); |
9e1b32caa mm: Pass virtual ... |
395 |
pte_free_tlb(tlb, token, addr); |
e0da382c9 [PATCH] freepgt: ... |
396 |
tlb->mm->nr_ptes--; |
1da177e4c Linux-2.6.12-rc2 |
397 |
} |
e0da382c9 [PATCH] freepgt: ... |
398 399 400 |
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 |
401 402 403 |
{ pmd_t *pmd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
404 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
405 |
|
e0da382c9 [PATCH] freepgt: ... |
406 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
407 |
pmd = pmd_offset(pud, addr); |
1da177e4c Linux-2.6.12-rc2 |
408 409 410 411 |
do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
9e1b32caa mm: Pass virtual ... |
412 |
free_pte_range(tlb, pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
413 |
} while (pmd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
414 415 416 417 418 419 420 |
start &= PUD_MASK; if (start < floor) return; if (ceiling) { ceiling &= PUD_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
421 |
} |
e0da382c9 [PATCH] freepgt: ... |
422 423 424 425 426 |
if (end - 1 > ceiling - 1) return; pmd = pmd_offset(pud, start); pud_clear(pud); |
9e1b32caa mm: Pass virtual ... |
427 |
pmd_free_tlb(tlb, pmd, start); |
1da177e4c Linux-2.6.12-rc2 |
428 |
} |
e0da382c9 [PATCH] freepgt: ... |
429 430 431 |
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 |
432 433 434 |
{ pud_t *pud; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
435 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
436 |
|
e0da382c9 [PATCH] freepgt: ... |
437 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
438 |
pud = pud_offset(pgd, addr); |
1da177e4c Linux-2.6.12-rc2 |
439 440 441 442 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
e0da382c9 [PATCH] freepgt: ... |
443 |
free_pmd_range(tlb, pud, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
444 |
} while (pud++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
445 446 447 448 449 450 451 |
start &= PGDIR_MASK; if (start < floor) return; if (ceiling) { ceiling &= PGDIR_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
452 |
} |
e0da382c9 [PATCH] freepgt: ... |
453 454 455 456 457 |
if (end - 1 > ceiling - 1) return; pud = pud_offset(pgd, start); pgd_clear(pgd); |
9e1b32caa mm: Pass virtual ... |
458 |
pud_free_tlb(tlb, pud, start); |
1da177e4c Linux-2.6.12-rc2 |
459 460 461 |
} /* |
e0da382c9 [PATCH] freepgt: ... |
462 463 |
* This function frees user-level page tables of a process. * |
1da177e4c Linux-2.6.12-rc2 |
464 465 |
* Must be called with pagetable lock held. */ |
42b777281 mm: remove double... |
466 |
void free_pgd_range(struct mmu_gather *tlb, |
e0da382c9 [PATCH] freepgt: ... |
467 468 |
unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c Linux-2.6.12-rc2 |
469 470 471 |
{ pgd_t *pgd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 |
/* * 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 |
498 |
|
e0da382c9 [PATCH] freepgt: ... |
499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 |
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... |
514 |
pgd = pgd_offset(tlb->mm, addr); |
1da177e4c Linux-2.6.12-rc2 |
515 516 517 518 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
42b777281 mm: remove double... |
519 |
free_pud_range(tlb, pgd, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
520 |
} while (pgd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
521 |
} |
42b777281 mm: remove double... |
522 |
void free_pgtables(struct mmu_gather *tlb, struct vm_area_struct *vma, |
3bf5ee956 [PATCH] freepgt: ... |
523 |
unsigned long floor, unsigned long ceiling) |
e0da382c9 [PATCH] freepgt: ... |
524 525 526 527 |
{ while (vma) { struct vm_area_struct *next = vma->vm_next; unsigned long addr = vma->vm_start; |
8f4f8c164 [PATCH] mm: unlin... |
528 |
/* |
25d9e2d15 truncate: new hel... |
529 530 |
* Hide vma from rmap and truncate_pagecache before freeing * pgtables |
8f4f8c164 [PATCH] mm: unlin... |
531 |
*/ |
5beb49305 mm: change anon_v... |
532 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
533 |
unlink_file_vma(vma); |
9da61aef0 [PATCH] hugepage:... |
534 |
if (is_vm_hugetlb_page(vma)) { |
3bf5ee956 [PATCH] freepgt: ... |
535 |
hugetlb_free_pgd_range(tlb, addr, vma->vm_end, |
e0da382c9 [PATCH] freepgt: ... |
536 |
floor, next? next->vm_start: ceiling); |
3bf5ee956 [PATCH] freepgt: ... |
537 538 539 540 541 |
} else { /* * Optimization: gather nearby vmas into one call down */ while (next && next->vm_start <= vma->vm_end + PMD_SIZE |
4866920b9 [PATCH] hugepage:... |
542 |
&& !is_vm_hugetlb_page(next)) { |
3bf5ee956 [PATCH] freepgt: ... |
543 544 |
vma = next; next = vma->vm_next; |
5beb49305 mm: change anon_v... |
545 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
546 |
unlink_file_vma(vma); |
3bf5ee956 [PATCH] freepgt: ... |
547 548 549 550 |
} free_pgd_range(tlb, addr, vma->vm_end, floor, next? next->vm_start: ceiling); } |
e0da382c9 [PATCH] freepgt: ... |
551 552 |
vma = next; } |
1da177e4c Linux-2.6.12-rc2 |
553 |
} |
8ac1f8320 thp: pte alloc tr... |
554 555 |
int __pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma, pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
556 |
{ |
2f569afd9 CONFIG_HIGHPTE vs... |
557 |
pgtable_t new = pte_alloc_one(mm, address); |
8ac1f8320 thp: pte alloc tr... |
558 |
int wait_split_huge_page; |
1bb3630e8 [PATCH] mm: ptd_a... |
559 560 |
if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
561 562 563 564 565 566 567 568 569 570 571 572 573 574 |
/* * 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 */ |
c74df32c7 [PATCH] mm: ptd_a... |
575 |
spin_lock(&mm->page_table_lock); |
8ac1f8320 thp: pte alloc tr... |
576 577 |
wait_split_huge_page = 0; if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
1da177e4c Linux-2.6.12-rc2 |
578 |
mm->nr_ptes++; |
1da177e4c Linux-2.6.12-rc2 |
579 |
pmd_populate(mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
580 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
581 582 |
} else if (unlikely(pmd_trans_splitting(*pmd))) wait_split_huge_page = 1; |
c74df32c7 [PATCH] mm: ptd_a... |
583 |
spin_unlock(&mm->page_table_lock); |
2f569afd9 CONFIG_HIGHPTE vs... |
584 585 |
if (new) pte_free(mm, new); |
8ac1f8320 thp: pte alloc tr... |
586 587 |
if (wait_split_huge_page) wait_split_huge_page(vma->anon_vma, pmd); |
1bb3630e8 [PATCH] mm: ptd_a... |
588 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
589 |
} |
1bb3630e8 [PATCH] mm: ptd_a... |
590 |
int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
591 |
{ |
1bb3630e8 [PATCH] mm: ptd_a... |
592 593 594 |
pte_t *new = pte_alloc_one_kernel(&init_mm, address); if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
595 |
smp_wmb(); /* See comment in __pte_alloc */ |
1bb3630e8 [PATCH] mm: ptd_a... |
596 |
spin_lock(&init_mm.page_table_lock); |
8ac1f8320 thp: pte alloc tr... |
597 |
if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
1bb3630e8 [PATCH] mm: ptd_a... |
598 |
pmd_populate_kernel(&init_mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
599 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
600 601 |
} else VM_BUG_ON(pmd_trans_splitting(*pmd)); |
1bb3630e8 [PATCH] mm: ptd_a... |
602 |
spin_unlock(&init_mm.page_table_lock); |
2f569afd9 CONFIG_HIGHPTE vs... |
603 604 |
if (new) pte_free_kernel(&init_mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
605 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
606 |
} |
d559db086 mm: clean up mm_c... |
607 608 609 610 611 612 |
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... |
613 |
{ |
d559db086 mm: clean up mm_c... |
614 |
int i; |
34e55232e mm: avoid false s... |
615 |
if (current->mm == mm) |
05af2e104 mm, counters: rem... |
616 |
sync_mm_rss(mm); |
d559db086 mm: clean up mm_c... |
617 618 619 |
for (i = 0; i < NR_MM_COUNTERS; i++) if (rss[i]) add_mm_counter(mm, i, rss[i]); |
ae8597623 [PATCH] mm: batch... |
620 |
} |
1da177e4c Linux-2.6.12-rc2 |
621 |
/* |
6aab341e0 mm: re-architect ... |
622 623 624 |
* 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... |
625 626 627 |
* * The calling function must still handle the error. */ |
3dc147414 badpage: replace ... |
628 629 |
static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr, pte_t pte, struct page *page) |
b5810039a [PATCH] core remo... |
630 |
{ |
3dc147414 badpage: replace ... |
631 632 633 634 635 |
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... |
636 637 638 639 640 641 642 643 644 645 646 647 648 649 |
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... |
650 651 652 |
printk(KERN_ALERT "BUG: Bad page map: %lu messages suppressed ", |
d936cf9b3 badpage: ratelimi... |
653 654 655 656 657 658 659 |
nr_unshown); nr_unshown = 0; } nr_shown = 0; } if (nr_shown++ == 0) resume = jiffies + 60 * HZ; |
3dc147414 badpage: replace ... |
660 661 662 |
mapping = vma->vm_file ? vma->vm_file->f_mapping : NULL; index = linear_page_index(vma, addr); |
1e9e63650 badpage: KERN_ALE... |
663 664 665 |
printk(KERN_ALERT "BUG: Bad page map in process %s pte:%08llx pmd:%08llx ", |
3dc147414 badpage: replace ... |
666 667 |
current->comm, (long long)pte_val(pte), (long long)pmd_val(*pmd)); |
718a38211 mm: introduce dum... |
668 669 |
if (page) dump_page(page); |
1e9e63650 badpage: KERN_ALE... |
670 |
printk(KERN_ALERT |
3dc147414 badpage: replace ... |
671 672 673 674 675 676 677 |
"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) |
1e9e63650 badpage: KERN_ALE... |
678 679 |
print_symbol(KERN_ALERT "vma->vm_ops->fault: %s ", |
3dc147414 badpage: replace ... |
680 681 |
(unsigned long)vma->vm_ops->fault); if (vma->vm_file && vma->vm_file->f_op) |
1e9e63650 badpage: KERN_ALE... |
682 683 |
print_symbol(KERN_ALERT "vma->vm_file->f_op->mmap: %s ", |
3dc147414 badpage: replace ... |
684 |
(unsigned long)vma->vm_file->f_op->mmap); |
b5810039a [PATCH] core remo... |
685 |
dump_stack(); |
3dc147414 badpage: replace ... |
686 |
add_taint(TAINT_BAD_PAGE); |
b5810039a [PATCH] core remo... |
687 |
} |
ca16d140a mm: don't access ... |
688 |
static inline int is_cow_mapping(vm_flags_t flags) |
67121172f Allow arbitrary r... |
689 690 691 |
{ return (flags & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE; } |
62eede62d mm: ZERO_PAGE wit... |
692 693 694 695 696 697 698 699 700 701 702 703 704 |
#ifndef is_zero_pfn static inline int is_zero_pfn(unsigned long pfn) { return pfn == zero_pfn; } #endif #ifndef my_zero_pfn static inline unsigned long my_zero_pfn(unsigned long addr) { return zero_pfn; } #endif |
b5810039a [PATCH] core remo... |
705 |
/* |
7e675137a mm: introduce pte... |
706 |
* vm_normal_page -- This function gets the "struct page" associated with a pte. |
6aab341e0 mm: re-architect ... |
707 |
* |
7e675137a mm: introduce pte... |
708 709 710 |
* "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_... |
711 |
* |
7e675137a mm: introduce pte... |
712 713 714 715 716 717 718 719 |
* 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 ... |
720 |
* |
b379d7901 mm: introduce VM_... |
721 722 |
* 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... |
723 724 |
* 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 ... |
725 726 727 |
* * pfn_of_page == vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT) * |
7e675137a mm: introduce pte... |
728 729 730 731 732 733 |
* 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_... |
734 |
* |
b379d7901 mm: introduce VM_... |
735 |
* |
7e675137a mm: introduce pte... |
736 |
* In order to support COW of arbitrary special mappings, we have VM_MIXEDMAP. |
b379d7901 mm: introduce VM_... |
737 738 739 740 741 742 743 744 745 |
* * 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: ... |
746 |
*/ |
7e675137a mm: introduce pte... |
747 748 749 750 751 752 753 |
#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: ... |
754 |
{ |
22b31eec6 badpage: vm_norma... |
755 |
unsigned long pfn = pte_pfn(pte); |
7e675137a mm: introduce pte... |
756 757 |
if (HAVE_PTE_SPECIAL) { |
22b31eec6 badpage: vm_norma... |
758 759 |
if (likely(!pte_special(pte))) goto check_pfn; |
a13ea5b75 mm: reinstate ZER... |
760 761 |
if (vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)) return NULL; |
62eede62d mm: ZERO_PAGE wit... |
762 |
if (!is_zero_pfn(pfn)) |
22b31eec6 badpage: vm_norma... |
763 |
print_bad_pte(vma, addr, pte, NULL); |
7e675137a mm: introduce pte... |
764 765 766 767 |
return NULL; } /* !HAVE_PTE_SPECIAL case follows: */ |
b379d7901 mm: introduce VM_... |
768 769 770 771 772 773 |
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... |
774 775 |
unsigned long off; off = (addr - vma->vm_start) >> PAGE_SHIFT; |
b379d7901 mm: introduce VM_... |
776 777 778 779 780 |
if (pfn == vma->vm_pgoff + off) return NULL; if (!is_cow_mapping(vma->vm_flags)) return NULL; } |
6aab341e0 mm: re-architect ... |
781 |
} |
62eede62d mm: ZERO_PAGE wit... |
782 783 |
if (is_zero_pfn(pfn)) return NULL; |
22b31eec6 badpage: vm_norma... |
784 785 786 787 788 |
check_pfn: if (unlikely(pfn > highest_memmap_pfn)) { print_bad_pte(vma, addr, pte, NULL); return NULL; } |
6aab341e0 mm: re-architect ... |
789 790 |
/* |
7e675137a mm: introduce pte... |
791 |
* NOTE! We still have PageReserved() pages in the page tables. |
7e675137a mm: introduce pte... |
792 |
* eg. VDSO mappings can cause them to exist. |
6aab341e0 mm: re-architect ... |
793 |
*/ |
b379d7901 mm: introduce VM_... |
794 |
out: |
6aab341e0 mm: re-architect ... |
795 |
return pfn_to_page(pfn); |
ee498ed73 [PATCH] unpaged: ... |
796 797 798 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
799 800 801 |
* 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 |
802 |
*/ |
570a335b8 swap_info: swap c... |
803 |
static inline unsigned long |
1da177e4c Linux-2.6.12-rc2 |
804 |
copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
b5810039a [PATCH] core remo... |
805 |
pte_t *dst_pte, pte_t *src_pte, struct vm_area_struct *vma, |
8c1037627 [PATCH] mm: copy_... |
806 |
unsigned long addr, int *rss) |
1da177e4c Linux-2.6.12-rc2 |
807 |
{ |
b5810039a [PATCH] core remo... |
808 |
unsigned long vm_flags = vma->vm_flags; |
1da177e4c Linux-2.6.12-rc2 |
809 810 |
pte_t pte = *src_pte; struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
811 812 813 814 |
/* pte contains position in swap or file, so copy. */ if (unlikely(!pte_present(pte))) { if (!pte_file(pte)) { |
0697212a4 [PATCH] Swapless ... |
815 |
swp_entry_t entry = pte_to_swp_entry(pte); |
570a335b8 swap_info: swap c... |
816 817 |
if (swap_duplicate(entry) < 0) return entry.val; |
1da177e4c Linux-2.6.12-rc2 |
818 819 820 |
/* 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... |
821 822 823 |
if (list_empty(&dst_mm->mmlist)) list_add(&dst_mm->mmlist, &src_mm->mmlist); |
1da177e4c Linux-2.6.12-rc2 |
824 825 |
spin_unlock(&mmlist_lock); } |
b084d4353 mm: count swap usage |
826 827 |
if (likely(!non_swap_entry(entry))) rss[MM_SWAPENTS]++; |
9f9f1acd7 mm: fix rss count... |
828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 |
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); set_pte_at(src_mm, addr, src_pte, pte); } |
0697212a4 [PATCH] Swapless ... |
846 |
} |
1da177e4c Linux-2.6.12-rc2 |
847 |
} |
ae8597623 [PATCH] mm: batch... |
848 |
goto out_set_pte; |
1da177e4c Linux-2.6.12-rc2 |
849 |
} |
1da177e4c Linux-2.6.12-rc2 |
850 851 852 853 |
/* * If it's a COW mapping, write protect it both * in the parent and the child */ |
67121172f Allow arbitrary r... |
854 |
if (is_cow_mapping(vm_flags)) { |
1da177e4c Linux-2.6.12-rc2 |
855 |
ptep_set_wrprotect(src_mm, addr, src_pte); |
3dc907951 [PATCH] paravirt:... |
856 |
pte = pte_wrprotect(pte); |
1da177e4c Linux-2.6.12-rc2 |
857 858 859 860 861 862 863 864 865 |
} /* * 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 ... |
866 867 868 869 |
page = vm_normal_page(vma, addr, pte); if (page) { get_page(page); |
21333b2b6 ksm: no debug in ... |
870 |
page_dup_rmap(page); |
d559db086 mm: clean up mm_c... |
871 872 873 874 |
if (PageAnon(page)) rss[MM_ANONPAGES]++; else rss[MM_FILEPAGES]++; |
6aab341e0 mm: re-architect ... |
875 |
} |
ae8597623 [PATCH] mm: batch... |
876 877 878 |
out_set_pte: set_pte_at(dst_mm, addr, dst_pte, pte); |
570a335b8 swap_info: swap c... |
879 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
880 |
} |
71e3aac07 thp: transparent ... |
881 882 883 |
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 |
884 |
{ |
c36987e2e mm: don't call pt... |
885 |
pte_t *orig_src_pte, *orig_dst_pte; |
1da177e4c Linux-2.6.12-rc2 |
886 |
pte_t *src_pte, *dst_pte; |
c74df32c7 [PATCH] mm: ptd_a... |
887 |
spinlock_t *src_ptl, *dst_ptl; |
e040f218b [PATCH] mm: copy_... |
888 |
int progress = 0; |
d559db086 mm: clean up mm_c... |
889 |
int rss[NR_MM_COUNTERS]; |
570a335b8 swap_info: swap c... |
890 |
swp_entry_t entry = (swp_entry_t){0}; |
1da177e4c Linux-2.6.12-rc2 |
891 892 |
again: |
d559db086 mm: clean up mm_c... |
893 |
init_rss_vec(rss); |
c74df32c7 [PATCH] mm: ptd_a... |
894 |
dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl); |
1da177e4c Linux-2.6.12-rc2 |
895 896 |
if (!dst_pte) return -ENOMEM; |
ece0e2b64 mm: remove pte_*m... |
897 |
src_pte = pte_offset_map(src_pmd, addr); |
4c21e2f24 [PATCH] mm: split... |
898 |
src_ptl = pte_lockptr(src_mm, src_pmd); |
f20dc5f7c [PATCH] lockdep: ... |
899 |
spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING); |
c36987e2e mm: don't call pt... |
900 901 |
orig_src_pte = src_pte; orig_dst_pte = dst_pte; |
6606c3e0d [PATCH] paravirt:... |
902 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
903 |
|
1da177e4c Linux-2.6.12-rc2 |
904 905 906 907 908 |
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_... |
909 910 911 |
if (progress >= 32) { progress = 0; if (need_resched() || |
95c354fe9 spinlock: lockbre... |
912 |
spin_needbreak(src_ptl) || spin_needbreak(dst_ptl)) |
e040f218b [PATCH] mm: copy_... |
913 914 |
break; } |
1da177e4c Linux-2.6.12-rc2 |
915 916 917 918 |
if (pte_none(*src_pte)) { progress++; continue; } |
570a335b8 swap_info: swap c... |
919 920 921 922 |
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 |
923 924 |
progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
925 |
|
6606c3e0d [PATCH] paravirt:... |
926 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
927 |
spin_unlock(src_ptl); |
ece0e2b64 mm: remove pte_*m... |
928 |
pte_unmap(orig_src_pte); |
d559db086 mm: clean up mm_c... |
929 |
add_mm_rss_vec(dst_mm, rss); |
c36987e2e mm: don't call pt... |
930 |
pte_unmap_unlock(orig_dst_pte, dst_ptl); |
c74df32c7 [PATCH] mm: ptd_a... |
931 |
cond_resched(); |
570a335b8 swap_info: swap c... |
932 933 934 935 936 937 |
if (entry.val) { if (add_swap_count_continuation(entry, GFP_KERNEL) < 0) return -ENOMEM; progress = 0; } |
1da177e4c Linux-2.6.12-rc2 |
938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 |
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 ... |
956 957 |
if (pmd_trans_huge(*src_pmd)) { int err; |
14d1a55cd thp: add debug ch... |
958 |
VM_BUG_ON(next-addr != HPAGE_PMD_SIZE); |
71e3aac07 thp: transparent ... |
959 960 961 962 963 964 965 966 |
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 |
967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 |
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; |
cddb8a5c1 mmu-notifiers: core |
1005 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1006 |
|
d992895ba [PATCH] Lazy page... |
1007 1008 1009 1010 1011 1012 |
/* * 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. */ |
4d7672b46 Make sure we copy... |
1013 |
if (!(vma->vm_flags & (VM_HUGETLB|VM_NONLINEAR|VM_PFNMAP|VM_INSERTPAGE))) { |
d992895ba [PATCH] Lazy page... |
1014 1015 1016 |
if (!vma->anon_vma) return 0; } |
1da177e4c Linux-2.6.12-rc2 |
1017 1018 |
if (is_vm_hugetlb_page(vma)) return copy_hugetlb_page_range(dst_mm, src_mm, vma); |
34801ba9b x86: PAT: move tr... |
1019 |
if (unlikely(is_pfn_mapping(vma))) { |
2ab640379 x86: PAT: hooks i... |
1020 1021 1022 1023 1024 1025 1026 1027 |
/* * We do not free on error cases below as remove_vma * gets called on error from higher level routine */ ret = track_pfn_vma_copy(vma); if (ret) return ret; } |
cddb8a5c1 mmu-notifiers: core |
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 |
/* * 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. */ if (is_cow_mapping(vma->vm_flags)) mmu_notifier_invalidate_range_start(src_mm, addr, end); ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
1038 1039 1040 1041 1042 1043 |
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 |
1044 1045 1046 1047 1048 |
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 |
1049 |
} while (dst_pgd++, src_pgd++, addr = next, addr != end); |
cddb8a5c1 mmu-notifiers: core |
1050 1051 1052 1053 1054 |
if (is_cow_mapping(vma->vm_flags)) mmu_notifier_invalidate_range_end(src_mm, vma->vm_start, end); return ret; |
1da177e4c Linux-2.6.12-rc2 |
1055 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1056 |
static unsigned long zap_pte_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1057 |
struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c Linux-2.6.12-rc2 |
1058 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1059 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1060 |
{ |
b5810039a [PATCH] core remo... |
1061 |
struct mm_struct *mm = tlb->mm; |
d16dfc550 mm: mmu_gather re... |
1062 |
int force_flush = 0; |
d559db086 mm: clean up mm_c... |
1063 |
int rss[NR_MM_COUNTERS]; |
97a894136 mm: Remove i_mmap... |
1064 |
spinlock_t *ptl; |
5f1a19070 mm: fix wrong kun... |
1065 |
pte_t *start_pte; |
97a894136 mm: Remove i_mmap... |
1066 |
pte_t *pte; |
d559db086 mm: clean up mm_c... |
1067 |
|
d16dfc550 mm: mmu_gather re... |
1068 |
again: |
e303297e6 mm: extended batc... |
1069 |
init_rss_vec(rss); |
5f1a19070 mm: fix wrong kun... |
1070 1071 |
start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl); pte = start_pte; |
6606c3e0d [PATCH] paravirt:... |
1072 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
1073 1074 |
do { pte_t ptent = *pte; |
51c6f666f [PATCH] mm: ZAP_B... |
1075 |
if (pte_none(ptent)) { |
1da177e4c Linux-2.6.12-rc2 |
1076 |
continue; |
51c6f666f [PATCH] mm: ZAP_B... |
1077 |
} |
6f5e6b9e6 [PATCH] fix free ... |
1078 |
|
1da177e4c Linux-2.6.12-rc2 |
1079 |
if (pte_present(ptent)) { |
ee498ed73 [PATCH] unpaged: ... |
1080 |
struct page *page; |
51c6f666f [PATCH] mm: ZAP_B... |
1081 |
|
6aab341e0 mm: re-architect ... |
1082 |
page = vm_normal_page(vma, addr, ptent); |
1da177e4c Linux-2.6.12-rc2 |
1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 |
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... |
1101 |
ptent = ptep_get_and_clear_full(mm, addr, pte, |
a600388d2 [PATCH] x86: ptep... |
1102 |
tlb->fullmm); |
1da177e4c Linux-2.6.12-rc2 |
1103 1104 1105 1106 1107 1108 |
tlb_remove_tlb_entry(tlb, pte, addr); if (unlikely(!page)) continue; if (unlikely(details) && details->nonlinear_vma && linear_page_index(details->nonlinear_vma, addr) != page->index) |
b5810039a [PATCH] core remo... |
1109 |
set_pte_at(mm, addr, pte, |
1da177e4c Linux-2.6.12-rc2 |
1110 |
pgoff_to_pte(page->index)); |
1da177e4c Linux-2.6.12-rc2 |
1111 |
if (PageAnon(page)) |
d559db086 mm: clean up mm_c... |
1112 |
rss[MM_ANONPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1113 1114 1115 |
else { if (pte_dirty(ptent)) set_page_dirty(page); |
4917e5d04 mm: more likely r... |
1116 1117 |
if (pte_young(ptent) && likely(!VM_SequentialReadHint(vma))) |
bf3f3bc5e mm: don't mark_pa... |
1118 |
mark_page_accessed(page); |
d559db086 mm: clean up mm_c... |
1119 |
rss[MM_FILEPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1120 |
} |
edc315fd2 badpage: remove v... |
1121 |
page_remove_rmap(page); |
3dc147414 badpage: replace ... |
1122 1123 |
if (unlikely(page_mapcount(page) < 0)) print_bad_pte(vma, addr, ptent, page); |
d16dfc550 mm: mmu_gather re... |
1124 1125 1126 |
force_flush = !__tlb_remove_page(tlb, page); if (force_flush) break; |
1da177e4c Linux-2.6.12-rc2 |
1127 1128 1129 1130 1131 1132 1133 1134 |
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... |
1135 1136 1137 |
if (pte_file(ptent)) { if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) print_bad_pte(vma, addr, ptent, NULL); |
b084d4353 mm: count swap usage |
1138 1139 1140 1141 1142 |
} else { swp_entry_t entry = pte_to_swp_entry(ptent); if (!non_swap_entry(entry)) rss[MM_SWAPENTS]--; |
9f9f1acd7 mm: fix rss count... |
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 |
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 |
1153 1154 1155 |
if (unlikely(!free_swap_and_cache(entry))) print_bad_pte(vma, addr, ptent, NULL); } |
9888a1cae [PATCH] paravirt:... |
1156 |
pte_clear_not_present_full(mm, addr, pte, tlb->fullmm); |
97a894136 mm: Remove i_mmap... |
1157 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
ae8597623 [PATCH] mm: batch... |
1158 |
|
d559db086 mm: clean up mm_c... |
1159 |
add_mm_rss_vec(mm, rss); |
6606c3e0d [PATCH] paravirt:... |
1160 |
arch_leave_lazy_mmu_mode(); |
5f1a19070 mm: fix wrong kun... |
1161 |
pte_unmap_unlock(start_pte, ptl); |
51c6f666f [PATCH] mm: ZAP_B... |
1162 |
|
d16dfc550 mm: mmu_gather re... |
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
/* * 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) { force_flush = 0; tlb_flush_mmu(tlb); if (addr != end) goto again; } |
51c6f666f [PATCH] mm: ZAP_B... |
1174 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1175 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1176 |
static inline unsigned long zap_pmd_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1177 |
struct vm_area_struct *vma, pud_t *pud, |
1da177e4c Linux-2.6.12-rc2 |
1178 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1179 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1180 1181 1182 1183 1184 1185 1186 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
71e3aac07 thp: transparent ... |
1187 |
if (pmd_trans_huge(*pmd)) { |
1a5a9906d mm: thp: fix pmd_... |
1188 |
if (next - addr != HPAGE_PMD_SIZE) { |
14d1a55cd thp: add debug ch... |
1189 |
VM_BUG_ON(!rwsem_is_locked(&tlb->mm->mmap_sem)); |
71e3aac07 thp: transparent ... |
1190 |
split_huge_page_pmd(vma->vm_mm, pmd); |
f21760b15 thp: add tlb_remo... |
1191 |
} else if (zap_huge_pmd(tlb, vma, pmd, addr)) |
1a5a9906d mm: thp: fix pmd_... |
1192 |
goto next; |
71e3aac07 thp: transparent ... |
1193 1194 |
/* fall through */ } |
1a5a9906d mm: thp: fix pmd_... |
1195 1196 1197 1198 1199 1200 1201 1202 1203 |
/* * 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... |
1204 |
next = zap_pte_range(tlb, vma, pmd, addr, next, details); |
1a5a9906d mm: thp: fix pmd_... |
1205 |
next: |
97a894136 mm: Remove i_mmap... |
1206 1207 |
cond_resched(); } while (pmd++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1208 1209 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1210 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1211 |
static inline unsigned long zap_pud_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1212 |
struct vm_area_struct *vma, pgd_t *pgd, |
1da177e4c Linux-2.6.12-rc2 |
1213 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1214 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1215 1216 1217 1218 1219 1220 1221 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1222 |
if (pud_none_or_clear_bad(pud)) |
1da177e4c Linux-2.6.12-rc2 |
1223 |
continue; |
97a894136 mm: Remove i_mmap... |
1224 1225 |
next = zap_pmd_range(tlb, vma, pud, addr, next, details); } while (pud++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1226 1227 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1228 |
} |
038c7aa16 VM: unmap_page_ra... |
1229 1230 1231 1232 |
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 |
1233 1234 1235 1236 1237 1238 1239 1240 |
{ pgd_t *pgd; unsigned long next; if (details && !details->check_mapping && !details->nonlinear_vma) details = NULL; BUG_ON(addr >= end); |
569b846df memcg: coalesce u... |
1241 |
mem_cgroup_uncharge_start(); |
1da177e4c Linux-2.6.12-rc2 |
1242 1243 1244 1245 |
tlb_start_vma(tlb, vma); pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1246 |
if (pgd_none_or_clear_bad(pgd)) |
1da177e4c Linux-2.6.12-rc2 |
1247 |
continue; |
97a894136 mm: Remove i_mmap... |
1248 1249 |
next = zap_pud_range(tlb, vma, pgd, addr, next, details); } while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
1250 |
tlb_end_vma(tlb, vma); |
569b846df memcg: coalesce u... |
1251 |
mem_cgroup_uncharge_end(); |
1da177e4c Linux-2.6.12-rc2 |
1252 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1253 |
|
f5cc4eef9 VM: make zap_page... |
1254 1255 1256 |
static void unmap_single_vma(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e vm: remove 'nr_ac... |
1257 |
unsigned long end_addr, |
f5cc4eef9 VM: make zap_page... |
1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 |
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... |
1268 1269 |
if (vma->vm_file) uprobe_munmap(vma, start, end); |
f5cc4eef9 VM: make zap_page... |
1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 |
if (unlikely(is_pfn_mapping(vma))) untrack_pfn_vma(vma, 0, 0); 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 * cleanup path of do_mmap_pgoff. When * hugetlbfs ->mmap method fails, * do_mmap_pgoff() nullifies vma->vm_file * before calling this function to clean up. * Since no pte has actually been setup, it is * safe to do nothing in this case. */ if (vma->vm_file) unmap_hugepage_range(vma, start, end, NULL); } 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 1325 1326 1327 |
} /** * zap_page_range - 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 |
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(); |
d16dfc550 mm: mmu_gather re... |
1339 |
tlb_gather_mmu(&tlb, mm, 0); |
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(); |
d16dfc550 mm: mmu_gather re... |
1364 |
tlb_gather_mmu(&tlb, mm, 0); |
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 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 |
/** * follow_page - look up a page descriptor from a user-virtual address * @vma: vm_area_struct mapping @address * @address: virtual address to look up * @flags: flags modifying lookup behaviour * * @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 |
1404 |
*/ |
6aab341e0 mm: re-architect ... |
1405 |
struct page *follow_page(struct vm_area_struct *vma, unsigned long address, |
deceb6cd1 [PATCH] mm: follo... |
1406 |
unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
1407 1408 1409 1410 1411 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; |
deceb6cd1 [PATCH] mm: follo... |
1412 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
1413 |
struct page *page; |
6aab341e0 mm: re-architect ... |
1414 |
struct mm_struct *mm = vma->vm_mm; |
1da177e4c Linux-2.6.12-rc2 |
1415 |
|
deceb6cd1 [PATCH] mm: follo... |
1416 1417 1418 1419 1420 |
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 |
1421 |
|
deceb6cd1 [PATCH] mm: follo... |
1422 |
page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1423 1424 |
pgd = pgd_offset(mm, address); if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) |
deceb6cd1 [PATCH] mm: follo... |
1425 |
goto no_page_table; |
1da177e4c Linux-2.6.12-rc2 |
1426 1427 |
pud = pud_offset(pgd, address); |
ceb868796 hugetlb: introduc... |
1428 |
if (pud_none(*pud)) |
deceb6cd1 [PATCH] mm: follo... |
1429 |
goto no_page_table; |
8a07651ee thp: transparent ... |
1430 |
if (pud_huge(*pud) && vma->vm_flags & VM_HUGETLB) { |
ceb868796 hugetlb: introduc... |
1431 1432 1433 1434 1435 1436 |
BUG_ON(flags & FOLL_GET); 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 |
1437 |
pmd = pmd_offset(pud, address); |
aeed5fce3 x86: fix PAE pmd_... |
1438 |
if (pmd_none(*pmd)) |
deceb6cd1 [PATCH] mm: follo... |
1439 |
goto no_page_table; |
71e3aac07 thp: transparent ... |
1440 |
if (pmd_huge(*pmd) && vma->vm_flags & VM_HUGETLB) { |
deceb6cd1 [PATCH] mm: follo... |
1441 1442 |
BUG_ON(flags & FOLL_GET); page = follow_huge_pmd(mm, address, pmd, flags & FOLL_WRITE); |
1da177e4c Linux-2.6.12-rc2 |
1443 |
goto out; |
deceb6cd1 [PATCH] mm: follo... |
1444 |
} |
71e3aac07 thp: transparent ... |
1445 |
if (pmd_trans_huge(*pmd)) { |
500d65d47 thp: pmd_trans_hu... |
1446 1447 1448 1449 |
if (flags & FOLL_SPLIT) { split_huge_page_pmd(mm, pmd); goto split_fallthrough; } |
71e3aac07 thp: transparent ... |
1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 |
spin_lock(&mm->page_table_lock); if (likely(pmd_trans_huge(*pmd))) { if (unlikely(pmd_trans_splitting(*pmd))) { spin_unlock(&mm->page_table_lock); wait_split_huge_page(vma->anon_vma, pmd); } else { page = follow_trans_huge_pmd(mm, address, pmd, flags); spin_unlock(&mm->page_table_lock); goto out; } } else spin_unlock(&mm->page_table_lock); /* fall through */ } |
500d65d47 thp: pmd_trans_hu... |
1465 |
split_fallthrough: |
aeed5fce3 x86: fix PAE pmd_... |
1466 1467 |
if (unlikely(pmd_bad(*pmd))) goto no_page_table; |
deceb6cd1 [PATCH] mm: follo... |
1468 |
ptep = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
1469 1470 |
pte = *ptep; |
deceb6cd1 [PATCH] mm: follo... |
1471 |
if (!pte_present(pte)) |
89f5b7da2 Reinstate ZERO_PA... |
1472 |
goto no_page; |
deceb6cd1 [PATCH] mm: follo... |
1473 1474 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
a13ea5b75 mm: reinstate ZER... |
1475 |
|
6aab341e0 mm: re-architect ... |
1476 |
page = vm_normal_page(vma, address, pte); |
a13ea5b75 mm: reinstate ZER... |
1477 1478 |
if (unlikely(!page)) { if ((flags & FOLL_DUMP) || |
62eede62d mm: ZERO_PAGE wit... |
1479 |
!is_zero_pfn(pte_pfn(pte))) |
a13ea5b75 mm: reinstate ZER... |
1480 1481 1482 |
goto bad_page; page = pte_page(pte); } |
1da177e4c Linux-2.6.12-rc2 |
1483 |
|
deceb6cd1 [PATCH] mm: follo... |
1484 |
if (flags & FOLL_GET) |
70b50f94f mm: thp: tail pag... |
1485 |
get_page_foll(page); |
deceb6cd1 [PATCH] mm: follo... |
1486 1487 1488 1489 |
if (flags & FOLL_TOUCH) { if ((flags & FOLL_WRITE) && !pte_dirty(pte) && !PageDirty(page)) set_page_dirty(page); |
bd775c42e mm: add comment w... |
1490 1491 1492 1493 1494 |
/* * 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... |
1495 1496 |
mark_page_accessed(page); } |
a1fde08c7 VM: skip the stac... |
1497 |
if ((flags & FOLL_MLOCK) && (vma->vm_flags & VM_LOCKED)) { |
110d74a92 mm: add FOLL_MLOC... |
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 |
/* * 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 */ /* * Because we lock page here and migration is * blocked by the pte's page reference, we need * only check for file-cache page truncation. */ if (page->mapping) mlock_vma_page(page); unlock_page(page); } } |
deceb6cd1 [PATCH] mm: follo... |
1519 1520 |
unlock: pte_unmap_unlock(ptep, ptl); |
1da177e4c Linux-2.6.12-rc2 |
1521 |
out: |
deceb6cd1 [PATCH] mm: follo... |
1522 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1523 |
|
89f5b7da2 Reinstate ZERO_PA... |
1524 1525 1526 1527 1528 1529 1530 1531 |
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... |
1532 |
|
deceb6cd1 [PATCH] mm: follo... |
1533 1534 1535 |
no_page_table: /* * When core dumping an enormous anonymous area that nobody |
8e4b9a607 mm: FOLL_DUMP rep... |
1536 1537 1538 1539 1540 |
* 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... |
1541 |
*/ |
8e4b9a607 mm: FOLL_DUMP rep... |
1542 1543 1544 |
if ((flags & FOLL_DUMP) && (!vma->vm_ops || !vma->vm_ops->fault)) return ERR_PTR(-EFAULT); |
deceb6cd1 [PATCH] mm: follo... |
1545 |
return page; |
1da177e4c Linux-2.6.12-rc2 |
1546 |
} |
95042f9eb vm: fix mlock() o... |
1547 1548 |
static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr) { |
a09a79f66 Don't lock guardp... |
1549 1550 |
return stack_guard_page_start(vma, addr) || stack_guard_page_end(vma, addr+PAGE_SIZE); |
95042f9eb vm: fix mlock() o... |
1551 |
} |
0014bd990 mm: export __get_... |
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 |
/** * __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. */ |
b291f0003 mlock: mlocked pa... |
1601 |
int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, |
58fa879e1 mm: FOLL flags fo... |
1602 |
unsigned long start, int nr_pages, unsigned int gup_flags, |
53a7706d5 mlock: do not hol... |
1603 1604 |
struct page **pages, struct vm_area_struct **vmas, int *nonblocking) |
1da177e4c Linux-2.6.12-rc2 |
1605 1606 |
{ int i; |
58fa879e1 mm: FOLL flags fo... |
1607 |
unsigned long vm_flags; |
1da177e4c Linux-2.6.12-rc2 |
1608 |
|
9d73777e5 clarify get_user_... |
1609 |
if (nr_pages <= 0) |
900cf086f Be more robust ab... |
1610 |
return 0; |
58fa879e1 mm: FOLL flags fo... |
1611 1612 |
VM_BUG_ON(!!pages != !!(gup_flags & FOLL_GET)); |
1da177e4c Linux-2.6.12-rc2 |
1613 1614 |
/* * Require read or write permissions. |
58fa879e1 mm: FOLL flags fo... |
1615 |
* If FOLL_FORCE is set, we only require the "MAY" flags. |
1da177e4c Linux-2.6.12-rc2 |
1616 |
*/ |
58fa879e1 mm: FOLL flags fo... |
1617 1618 1619 1620 |
vm_flags = (gup_flags & FOLL_WRITE) ? (VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD); vm_flags &= (gup_flags & FOLL_FORCE) ? (VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE); |
1da177e4c Linux-2.6.12-rc2 |
1621 1622 1623 |
i = 0; do { |
deceb6cd1 [PATCH] mm: follo... |
1624 |
struct vm_area_struct *vma; |
1da177e4c Linux-2.6.12-rc2 |
1625 1626 |
vma = find_extend_vma(mm, start); |
e7f22e207 mm: use mm_struct... |
1627 |
if (!vma && in_gate_area(mm, start)) { |
1da177e4c Linux-2.6.12-rc2 |
1628 |
unsigned long pg = start & PAGE_MASK; |
1da177e4c Linux-2.6.12-rc2 |
1629 1630 1631 1632 |
pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; |
b291f0003 mlock: mlocked pa... |
1633 1634 |
/* user gate pages are read-only */ |
58fa879e1 mm: FOLL flags fo... |
1635 |
if (gup_flags & FOLL_WRITE) |
1da177e4c Linux-2.6.12-rc2 |
1636 1637 1638 1639 1640 1641 1642 1643 1644 |
return i ? : -EFAULT; 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... |
1645 1646 |
if (pmd_none(*pmd)) return i ? : -EFAULT; |
f66055ab6 thp: verify pmd_t... |
1647 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c Linux-2.6.12-rc2 |
1648 |
pte = pte_offset_map(pmd, pg); |
690dbe1ce [PATCH] x86_64: a... |
1649 1650 1651 1652 |
if (pte_none(*pte)) { pte_unmap(pte); return i ? : -EFAULT; } |
95042f9eb vm: fix mlock() o... |
1653 |
vma = get_gate_vma(mm); |
1da177e4c Linux-2.6.12-rc2 |
1654 |
if (pages) { |
de51257aa mm: fix ia64 cras... |
1655 |
struct page *page; |
95042f9eb vm: fix mlock() o... |
1656 |
page = vm_normal_page(vma, start, *pte); |
de51257aa mm: fix ia64 cras... |
1657 1658 1659 1660 1661 1662 1663 1664 1665 |
if (!page) { if (!(gup_flags & FOLL_DUMP) && is_zero_pfn(pte_pfn(*pte))) page = pte_page(*pte); else { pte_unmap(pte); return i ? : -EFAULT; } } |
6aab341e0 mm: re-architect ... |
1666 |
pages[i] = page; |
de51257aa mm: fix ia64 cras... |
1667 |
get_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1668 1669 |
} pte_unmap(pte); |
95042f9eb vm: fix mlock() o... |
1670 |
goto next_page; |
1da177e4c Linux-2.6.12-rc2 |
1671 |
} |
b291f0003 mlock: mlocked pa... |
1672 1673 |
if (!vma || (vma->vm_flags & (VM_IO | VM_PFNMAP)) || |
1c3aff1ce mm: remove unused... |
1674 |
!(vm_flags & vma->vm_flags)) |
1da177e4c Linux-2.6.12-rc2 |
1675 |
return i ? : -EFAULT; |
2a15efc95 mm: follow_hugetl... |
1676 1677 |
if (is_vm_hugetlb_page(vma)) { i = follow_hugetlb_page(mm, vma, pages, vmas, |
58fa879e1 mm: FOLL flags fo... |
1678 |
&start, &nr_pages, i, gup_flags); |
2a15efc95 mm: follow_hugetl... |
1679 1680 |
continue; } |
deceb6cd1 [PATCH] mm: follo... |
1681 |
|
1da177e4c Linux-2.6.12-rc2 |
1682 |
do { |
08ef47293 [PATCH] get_user_... |
1683 |
struct page *page; |
58fa879e1 mm: FOLL flags fo... |
1684 |
unsigned int foll_flags = gup_flags; |
1da177e4c Linux-2.6.12-rc2 |
1685 |
|
462e00cc7 oom: stop allocat... |
1686 |
/* |
4779280d1 mm: make get_user... |
1687 |
* If we have a pending SIGKILL, don't keep faulting |
1c3aff1ce mm: remove unused... |
1688 |
* pages and potentially allocating memory. |
462e00cc7 oom: stop allocat... |
1689 |
*/ |
1c3aff1ce mm: remove unused... |
1690 |
if (unlikely(fatal_signal_pending(current))) |
4779280d1 mm: make get_user... |
1691 |
return i ? i : -ERESTARTSYS; |
462e00cc7 oom: stop allocat... |
1692 |
|
deceb6cd1 [PATCH] mm: follo... |
1693 |
cond_resched(); |
6aab341e0 mm: re-architect ... |
1694 |
while (!(page = follow_page(vma, start, foll_flags))) { |
deceb6cd1 [PATCH] mm: follo... |
1695 |
int ret; |
53a7706d5 mlock: do not hol... |
1696 |
unsigned int fault_flags = 0; |
a09a79f66 Don't lock guardp... |
1697 1698 1699 1700 1701 |
/* 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... |
1702 1703 1704 1705 |
if (foll_flags & FOLL_WRITE) fault_flags |= FAULT_FLAG_WRITE; if (nonblocking) fault_flags |= FAULT_FLAG_ALLOW_RETRY; |
318b275fb mm: allow GUP to ... |
1706 1707 |
if (foll_flags & FOLL_NOWAIT) fault_flags |= (FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT); |
d06063cc2 Move FAULT_FLAG_x... |
1708 |
|
d26ed650d mm: don't rely on... |
1709 |
ret = handle_mm_fault(mm, vma, start, |
53a7706d5 mlock: do not hol... |
1710 |
fault_flags); |
d26ed650d mm: don't rely on... |
1711 |
|
83c54070e mm: fault feedbac... |
1712 1713 1714 |
if (ret & VM_FAULT_ERROR) { if (ret & VM_FAULT_OOM) return i ? i : -ENOMEM; |
69ebb83e1 mm: make __get_us... |
1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 |
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) |
83c54070e mm: fault feedbac... |
1725 1726 1727 |
return i ? i : -EFAULT; BUG(); } |
e7f22e207 mm: use mm_struct... |
1728 1729 1730 1731 1732 1733 1734 |
if (tsk) { if (ret & VM_FAULT_MAJOR) tsk->maj_flt++; else tsk->min_flt++; } |
83c54070e mm: fault feedbac... |
1735 |
|
53a7706d5 mlock: do not hol... |
1736 |
if (ret & VM_FAULT_RETRY) { |
318b275fb mm: allow GUP to ... |
1737 1738 |
if (nonblocking) *nonblocking = 0; |
53a7706d5 mlock: do not hol... |
1739 1740 |
return i; } |
a68d2ebc1 Fix up recent get... |
1741 |
/* |
83c54070e mm: fault feedbac... |
1742 1743 1744 1745 |
* 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... |
1746 1747 1748 1749 1750 1751 |
* 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... |
1752 |
*/ |
878b63ac8 mm: gup persist f... |
1753 1754 |
if ((ret & VM_FAULT_WRITE) && !(vma->vm_flags & VM_WRITE)) |
deceb6cd1 [PATCH] mm: follo... |
1755 |
foll_flags &= ~FOLL_WRITE; |
83c54070e mm: fault feedbac... |
1756 |
|
7f7bbbe50 [PATCH] page faul... |
1757 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1758 |
} |
89f5b7da2 Reinstate ZERO_PA... |
1759 1760 |
if (IS_ERR(page)) return i ? i : PTR_ERR(page); |
1da177e4c Linux-2.6.12-rc2 |
1761 |
if (pages) { |
08ef47293 [PATCH] get_user_... |
1762 |
pages[i] = page; |
03beb0766 [PATCH] Add API f... |
1763 |
|
a6f36be32 [ARM] pass vma fo... |
1764 |
flush_anon_page(vma, page, start); |
08ef47293 [PATCH] get_user_... |
1765 |
flush_dcache_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1766 |
} |
95042f9eb vm: fix mlock() o... |
1767 |
next_page: |
1da177e4c Linux-2.6.12-rc2 |
1768 1769 1770 1771 |
if (vmas) vmas[i] = vma; i++; start += PAGE_SIZE; |
9d73777e5 clarify get_user_... |
1772 1773 1774 |
nr_pages--; } while (nr_pages && start < vma->vm_end); } while (nr_pages); |
1da177e4c Linux-2.6.12-rc2 |
1775 1776 |
return i; } |
0014bd990 mm: export __get_... |
1777 |
EXPORT_SYMBOL(__get_user_pages); |
b291f0003 mlock: mlocked pa... |
1778 |
|
2efaca927 mm/futex: fix fut... |
1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 |
/* * 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_... |
1836 |
* get_user_pages() - pin user pages in memory |
e7f22e207 mm: use mm_struct... |
1837 1838 |
* @tsk: the task_struct to use for page fault accounting, or * NULL if faults are not to be recorded. |
d2bf6be8a mm: clean up get_... |
1839 1840 |
* @mm: mm_struct of target mm * @start: starting user address |
9d73777e5 clarify get_user_... |
1841 |
* @nr_pages: number of pages from start to pin |
d2bf6be8a mm: clean up get_... |
1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 |
* @write: whether pages will be written to by the caller * @force: whether to force write access even if user mapping is * readonly. This will result in the page being COWed even * in MAP_SHARED mappings. You do not want this. * @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_... |
1853 |
* requested. If nr_pages is 0 or negative, returns 0. If no pages |
d2bf6be8a mm: clean up get_... |
1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 |
* 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. */ |
b291f0003 mlock: mlocked pa... |
1886 |
int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, |
9d73777e5 clarify get_user_... |
1887 |
unsigned long start, int nr_pages, int write, int force, |
b291f0003 mlock: mlocked pa... |
1888 1889 |
struct page **pages, struct vm_area_struct **vmas) { |
58fa879e1 mm: FOLL flags fo... |
1890 |
int flags = FOLL_TOUCH; |
b291f0003 mlock: mlocked pa... |
1891 |
|
58fa879e1 mm: FOLL flags fo... |
1892 1893 |
if (pages) flags |= FOLL_GET; |
b291f0003 mlock: mlocked pa... |
1894 |
if (write) |
58fa879e1 mm: FOLL flags fo... |
1895 |
flags |= FOLL_WRITE; |
b291f0003 mlock: mlocked pa... |
1896 |
if (force) |
58fa879e1 mm: FOLL flags fo... |
1897 |
flags |= FOLL_FORCE; |
b291f0003 mlock: mlocked pa... |
1898 |
|
53a7706d5 mlock: do not hol... |
1899 1900 |
return __get_user_pages(tsk, mm, start, nr_pages, flags, pages, vmas, NULL); |
b291f0003 mlock: mlocked pa... |
1901 |
} |
1da177e4c Linux-2.6.12-rc2 |
1902 |
EXPORT_SYMBOL(get_user_pages); |
f3e8fccd0 mm: add get_dump_... |
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 |
/** * 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... |
1924 1925 |
FOLL_FORCE | FOLL_DUMP | FOLL_GET, &page, &vma, NULL) < 1) |
f3e8fccd0 mm: add get_dump_... |
1926 |
return NULL; |
f3e8fccd0 mm: add get_dump_... |
1927 1928 1929 1930 |
flush_cache_page(vma, addr, page_to_pfn(page)); return page; } #endif /* CONFIG_ELF_CORE */ |
25ca1d6c0 mm: wrap get_lock... |
1931 |
pte_t *__get_locked_pte(struct mm_struct *mm, unsigned long addr, |
920c7a5d0 mm: remove fastca... |
1932 |
spinlock_t **ptl) |
c9cfcddfd VM: add common he... |
1933 1934 1935 1936 |
{ pgd_t * pgd = pgd_offset(mm, addr); pud_t * pud = pud_alloc(mm, pgd, addr); if (pud) { |
49c91fb01 [PATCH] VM: Fix t... |
1937 |
pmd_t * pmd = pmd_alloc(mm, pud, addr); |
f66055ab6 thp: verify pmd_t... |
1938 1939 |
if (pmd) { VM_BUG_ON(pmd_trans_huge(*pmd)); |
c9cfcddfd VM: add common he... |
1940 |
return pte_alloc_map_lock(mm, pmd, addr, ptl); |
f66055ab6 thp: verify pmd_t... |
1941 |
} |
c9cfcddfd VM: add common he... |
1942 1943 1944 |
} return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
1945 |
/* |
238f58d89 Support strange d... |
1946 1947 1948 1949 1950 1951 |
* 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... |
1952 1953 |
static int insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot) |
238f58d89 Support strange d... |
1954 |
{ |
423bad600 mm: add vm_insert... |
1955 |
struct mm_struct *mm = vma->vm_mm; |
238f58d89 Support strange d... |
1956 |
int retval; |
c9cfcddfd VM: add common he... |
1957 |
pte_t *pte; |
8a9f3ccd2 Memory controller... |
1958 |
spinlock_t *ptl; |
238f58d89 Support strange d... |
1959 |
retval = -EINVAL; |
a145dd411 VM: add "vm_inser... |
1960 |
if (PageAnon(page)) |
5b4e655e9 memcg: avoid acco... |
1961 |
goto out; |
238f58d89 Support strange d... |
1962 1963 |
retval = -ENOMEM; flush_dcache_page(page); |
c9cfcddfd VM: add common he... |
1964 |
pte = get_locked_pte(mm, addr, &ptl); |
238f58d89 Support strange d... |
1965 |
if (!pte) |
5b4e655e9 memcg: avoid acco... |
1966 |
goto out; |
238f58d89 Support strange d... |
1967 1968 1969 1970 1971 1972 |
retval = -EBUSY; if (!pte_none(*pte)) goto out_unlock; /* Ok, finally just insert the thing.. */ get_page(page); |
34e55232e mm: avoid false s... |
1973 |
inc_mm_counter_fast(mm, MM_FILEPAGES); |
238f58d89 Support strange d... |
1974 1975 1976 1977 |
page_add_file_rmap(page); set_pte_at(mm, addr, pte, mk_pte(page, prot)); retval = 0; |
8a9f3ccd2 Memory controller... |
1978 1979 |
pte_unmap_unlock(pte, ptl); return retval; |
238f58d89 Support strange d... |
1980 1981 1982 1983 1984 |
out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
bfa5bf6d6 [PATCH] Add kerne... |
1985 1986 1987 1988 1989 1990 |
/** * 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... |
1991 1992 1993 1994 1995 1996 |
* 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... |
1997 |
* (see split_page()). |
a145dd411 VM: add "vm_inser... |
1998 1999 2000 2001 2002 2003 2004 2005 2006 |
* * 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. */ |
423bad600 mm: add vm_insert... |
2007 2008 |
int vm_insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page) |
a145dd411 VM: add "vm_inser... |
2009 2010 2011 2012 2013 |
{ if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; if (!page_count(page)) return -EINVAL; |
4d7672b46 Make sure we copy... |
2014 |
vma->vm_flags |= VM_INSERTPAGE; |
423bad600 mm: add vm_insert... |
2015 |
return insert_page(vma, addr, page, vma->vm_page_prot); |
a145dd411 VM: add "vm_inser... |
2016 |
} |
e3c3374fb Make vm_insert_pa... |
2017 |
EXPORT_SYMBOL(vm_insert_page); |
a145dd411 VM: add "vm_inser... |
2018 |
|
423bad600 mm: add vm_insert... |
2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 |
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... |
2038 |
update_mmu_cache(vma, addr, pte); /* XXX: why not for insert_page? */ |
423bad600 mm: add vm_insert... |
2039 2040 2041 2042 2043 2044 2045 |
retval = 0; out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
e0dc0d8f4 [PATCH] add vm_in... |
2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 |
/** * 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 * * Similar to vm_inert_page, this allows drivers to insert individual pages * 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 |
2057 2058 2059 2060 2061 |
* * 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... |
2062 2063 |
*/ int vm_insert_pfn(struct vm_area_struct *vma, unsigned long addr, |
423bad600 mm: add vm_insert... |
2064 |
unsigned long pfn) |
e0dc0d8f4 [PATCH] add vm_in... |
2065 |
{ |
2ab640379 x86: PAT: hooks i... |
2066 |
int ret; |
e4b866ed1 x86 PAT: change t... |
2067 |
pgprot_t pgprot = vma->vm_page_prot; |
7e675137a mm: introduce pte... |
2068 2069 2070 2071 2072 2073 |
/* * 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_... |
2074 2075 2076 2077 2078 |
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... |
2079 |
|
423bad600 mm: add vm_insert... |
2080 2081 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
e4b866ed1 x86 PAT: change t... |
2082 |
if (track_pfn_vma_new(vma, &pgprot, pfn, PAGE_SIZE)) |
2ab640379 x86: PAT: hooks i... |
2083 |
return -EINVAL; |
e4b866ed1 x86 PAT: change t... |
2084 |
ret = insert_pfn(vma, addr, pfn, pgprot); |
2ab640379 x86: PAT: hooks i... |
2085 2086 2087 2088 2089 |
if (ret) untrack_pfn_vma(vma, pfn, PAGE_SIZE); return ret; |
423bad600 mm: add vm_insert... |
2090 2091 |
} EXPORT_SYMBOL(vm_insert_pfn); |
e0dc0d8f4 [PATCH] add vm_in... |
2092 |
|
423bad600 mm: add vm_insert... |
2093 2094 2095 2096 |
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... |
2097 |
|
423bad600 mm: add vm_insert... |
2098 2099 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
e0dc0d8f4 [PATCH] add vm_in... |
2100 |
|
423bad600 mm: add vm_insert... |
2101 2102 2103 2104 |
/* * 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... |
2105 2106 |
* 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... |
2107 2108 2109 2110 2111 2112 2113 2114 |
*/ 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... |
2115 |
} |
423bad600 mm: add vm_insert... |
2116 |
EXPORT_SYMBOL(vm_insert_mixed); |
e0dc0d8f4 [PATCH] add vm_in... |
2117 |
|
a145dd411 VM: add "vm_inser... |
2118 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2119 2120 2121 2122 2123 2124 2125 2126 2127 |
* 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... |
2128 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
2129 |
|
c74df32c7 [PATCH] mm: ptd_a... |
2130 |
pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
2131 2132 |
if (!pte) return -ENOMEM; |
6606c3e0d [PATCH] paravirt:... |
2133 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
2134 2135 |
do { BUG_ON(!pte_none(*pte)); |
7e675137a mm: introduce pte... |
2136 |
set_pte_at(mm, addr, pte, pte_mkspecial(pfn_pte(pfn, prot))); |
1da177e4c Linux-2.6.12-rc2 |
2137 2138 |
pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); |
6606c3e0d [PATCH] paravirt:... |
2139 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
2140 |
pte_unmap_unlock(pte - 1, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 |
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... |
2155 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c Linux-2.6.12-rc2 |
2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 |
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... |
2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 |
/** * 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 |
2194 2195 2196 2197 2198 |
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... |
2199 |
unsigned long end = addr + PAGE_ALIGN(size); |
1da177e4c Linux-2.6.12-rc2 |
2200 2201 2202 2203 2204 2205 2206 2207 |
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). |
0b14c179a [PATCH] unpaged: ... |
2208 2209 2210 2211 2212 |
* VM_RESERVED is specified all over the place, because * in 2.4 it kept swapout's vma scan off this vma; but * in 2.6 the LRU scan won't even find its pages, so this * flag means no more than count its pages in reserved_vm, * and omit it from core dump, even when VM_IO turned off. |
6aab341e0 mm: re-architect ... |
2213 2214 2215 |
* 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. |
fb155c161 Allow arbitrary s... |
2216 2217 2218 2219 |
* * 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". |
1da177e4c Linux-2.6.12-rc2 |
2220 |
*/ |
4bb9c5c02 VM, x86, PAT: Cha... |
2221 |
if (addr == vma->vm_start && end == vma->vm_end) { |
fb155c161 Allow arbitrary s... |
2222 |
vma->vm_pgoff = pfn; |
895791dac VM, x86, PAT: add... |
2223 |
vma->vm_flags |= VM_PFN_AT_MMAP; |
4bb9c5c02 VM, x86, PAT: Cha... |
2224 |
} else if (is_cow_mapping(vma->vm_flags)) |
3c8bb73ac x86: PAT: store v... |
2225 |
return -EINVAL; |
fb155c161 Allow arbitrary s... |
2226 |
|
6aab341e0 mm: re-architect ... |
2227 |
vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP; |
1da177e4c Linux-2.6.12-rc2 |
2228 |
|
e4b866ed1 x86 PAT: change t... |
2229 |
err = track_pfn_vma_new(vma, &prot, pfn, PAGE_ALIGN(size)); |
a36706131 x86 PAT: remove P... |
2230 2231 2232 2233 2234 2235 |
if (err) { /* * To indicate that track_pfn related cleanup is not * needed from higher level routine calling unmap_vmas */ vma->vm_flags &= ~(VM_IO | VM_RESERVED | VM_PFNMAP); |
895791dac VM, x86, PAT: add... |
2236 |
vma->vm_flags &= ~VM_PFN_AT_MMAP; |
2ab640379 x86: PAT: hooks i... |
2237 |
return -EINVAL; |
a36706131 x86 PAT: remove P... |
2238 |
} |
2ab640379 x86: PAT: hooks i... |
2239 |
|
1da177e4c Linux-2.6.12-rc2 |
2240 2241 2242 2243 |
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 |
2244 2245 2246 2247 2248 2249 2250 |
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... |
2251 2252 2253 |
if (err) untrack_pfn_vma(vma, pfn, PAGE_ALIGN(size)); |
1da177e4c Linux-2.6.12-rc2 |
2254 2255 2256 |
return err; } EXPORT_SYMBOL(remap_pfn_range); |
aee16b3ce Add apply_to_page... |
2257 2258 2259 2260 2261 2262 |
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... |
2263 |
pgtable_t token; |
949099148 Add unitialized_v... |
2264 |
spinlock_t *uninitialized_var(ptl); |
aee16b3ce Add apply_to_page... |
2265 2266 2267 2268 2269 2270 2271 2272 |
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... |
2273 |
arch_enter_lazy_mmu_mode(); |
2f569afd9 CONFIG_HIGHPTE vs... |
2274 |
token = pmd_pgtable(*pmd); |
aee16b3ce Add apply_to_page... |
2275 2276 |
do { |
c36987e2e mm: don't call pt... |
2277 |
err = fn(pte++, token, addr, data); |
aee16b3ce Add apply_to_page... |
2278 2279 |
if (err) break; |
c36987e2e mm: don't call pt... |
2280 |
} while (addr += PAGE_SIZE, addr != end); |
aee16b3ce Add apply_to_page... |
2281 |
|
38e0edb15 mm/apply_to_range... |
2282 |
arch_leave_lazy_mmu_mode(); |
aee16b3ce Add apply_to_page... |
2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 |
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... |
2295 |
BUG_ON(pud_huge(*pud)); |
aee16b3ce Add apply_to_page... |
2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 |
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... |
2337 |
unsigned long end = addr + size; |
aee16b3ce Add apply_to_page... |
2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 |
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... |
2348 |
|
aee16b3ce Add apply_to_page... |
2349 2350 2351 |
return err; } EXPORT_SYMBOL_GPL(apply_to_page_range); |
1da177e4c Linux-2.6.12-rc2 |
2352 |
/* |
8f4e2101f [PATCH] mm: page ... |
2353 2354 2355 |
* 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... |
2356 |
* might give a mix of unmatched parts, do_swap_page and do_nonlinear_fault |
8f4e2101f [PATCH] mm: page ... |
2357 2358 |
* 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... |
2359 |
* and do_anonymous_page can safely check later on). |
8f4e2101f [PATCH] mm: page ... |
2360 |
*/ |
4c21e2f24 [PATCH] mm: split... |
2361 |
static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd, |
8f4e2101f [PATCH] mm: page ... |
2362 2363 2364 2365 2366 |
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... |
2367 2368 |
spinlock_t *ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
8f4e2101f [PATCH] mm: page ... |
2369 |
same = pte_same(*page_table, orig_pte); |
4c21e2f24 [PATCH] mm: split... |
2370 |
spin_unlock(ptl); |
8f4e2101f [PATCH] mm: page ... |
2371 2372 2373 2374 2375 |
} #endif pte_unmap(page_table); return same; } |
9de455b20 [PATCH] Pass vma ... |
2376 |
static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma) |
6aab341e0 mm: re-architect ... |
2377 2378 2379 2380 2381 2382 2383 2384 |
{ /* * 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... |
2385 |
void *kaddr = kmap_atomic(dst); |
5d2a2dbbc cow_user_page: fi... |
2386 2387 2388 2389 2390 2391 2392 2393 2394 |
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()/... |
2395 |
clear_page(kaddr); |
9b04c5fec mm: remove the se... |
2396 |
kunmap_atomic(kaddr); |
c4ec7b0de [PATCH] mm: D-cac... |
2397 |
flush_dcache_page(dst); |
0ed361dec mm: fix PageUptod... |
2398 2399 |
} else copy_user_highpage(dst, src, va, vma); |
6aab341e0 mm: re-architect ... |
2400 |
} |
1da177e4c Linux-2.6.12-rc2 |
2401 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2402 2403 2404 2405 |
* 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 |
2406 2407 2408 2409 2410 2411 2412 2413 2414 |
* 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 ... |
2415 2416 2417 |
* 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 |
2418 |
*/ |
65500d234 [PATCH] mm: page ... |
2419 2420 |
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 ... |
2421 |
spinlock_t *ptl, pte_t orig_pte) |
e6219ec81 mm: add lock rele... |
2422 |
__releases(ptl) |
1da177e4c Linux-2.6.12-rc2 |
2423 |
{ |
e5bbe4dfc [PATCH] pfnmap: r... |
2424 |
struct page *old_page, *new_page; |
1da177e4c Linux-2.6.12-rc2 |
2425 |
pte_t entry; |
b009c024f do_wp_page: remov... |
2426 |
int ret = 0; |
a200ee182 mm: set_page_dirt... |
2427 |
int page_mkwrite = 0; |
d08b3851d [PATCH] mm: track... |
2428 |
struct page *dirty_page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2429 |
|
6aab341e0 mm: re-architect ... |
2430 |
old_page = vm_normal_page(vma, address, orig_pte); |
251b97f55 mm: dirty page ac... |
2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 |
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 ... |
2442 |
goto gotten; |
251b97f55 mm: dirty page ac... |
2443 |
} |
1da177e4c Linux-2.6.12-rc2 |
2444 |
|
d08b3851d [PATCH] mm: track... |
2445 |
/* |
ee6a64578 [PATCH] mm: fixup... |
2446 2447 |
* Take out anonymous pages first, anonymous shared vmas are * not dirty accountable. |
d08b3851d [PATCH] mm: track... |
2448 |
*/ |
9a8408951 ksm: identify Pag... |
2449 |
if (PageAnon(old_page) && !PageKsm(old_page)) { |
ab967d860 mm: wp lock page ... |
2450 2451 2452 2453 2454 2455 2456 2457 |
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 ... |
2458 2459 2460 |
goto unlock; } page_cache_release(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2461 |
} |
b009c024f do_wp_page: remov... |
2462 |
if (reuse_swap_page(old_page)) { |
c44b67432 rmap: move exclus... |
2463 2464 2465 2466 2467 2468 |
/* * 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... |
2469 2470 2471 |
unlock_page(old_page); goto reuse; } |
ab967d860 mm: wp lock page ... |
2472 |
unlock_page(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2473 |
} else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) == |
d08b3851d [PATCH] mm: track... |
2474 |
(VM_WRITE|VM_SHARED))) { |
ee6a64578 [PATCH] mm: fixup... |
2475 2476 2477 2478 2479 |
/* * 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_... |
2480 |
if (vma->vm_ops && vma->vm_ops->page_mkwrite) { |
c2ec175c3 mm: page_mkwrite ... |
2481 2482 2483 2484 2485 2486 2487 2488 |
struct vm_fault vmf; int tmp; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = old_page->index; vmf.flags = FAULT_FLAG_WRITE|FAULT_FLAG_MKWRITE; vmf.page = old_page; |
9637a5efd [PATCH] add page_... |
2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 |
/* * 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. */ page_cache_get(old_page); pte_unmap_unlock(page_table, ptl); |
c2ec175c3 mm: page_mkwrite ... |
2499 2500 2501 2502 |
tmp = vma->vm_ops->page_mkwrite(vma, &vmf); if (unlikely(tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) { ret = tmp; |
9637a5efd [PATCH] add page_... |
2503 |
goto unwritable_page; |
c2ec175c3 mm: page_mkwrite ... |
2504 |
} |
b827e496c mm: close page_mk... |
2505 2506 2507 2508 2509 2510 2511 2512 2513 |
if (unlikely(!(tmp & VM_FAULT_LOCKED))) { lock_page(old_page); if (!old_page->mapping) { ret = 0; /* retry the fault */ unlock_page(old_page); goto unwritable_page; } } else VM_BUG_ON(!PageLocked(old_page)); |
9637a5efd [PATCH] add page_... |
2514 |
|
9637a5efd [PATCH] add page_... |
2515 2516 2517 2518 2519 2520 2521 2522 |
/* * 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... |
2523 2524 |
if (!pte_same(*page_table, orig_pte)) { unlock_page(old_page); |
9637a5efd [PATCH] add page_... |
2525 |
goto unlock; |
b827e496c mm: close page_mk... |
2526 |
} |
a200ee182 mm: set_page_dirt... |
2527 2528 |
page_mkwrite = 1; |
1da177e4c Linux-2.6.12-rc2 |
2529 |
} |
d08b3851d [PATCH] mm: track... |
2530 2531 |
dirty_page = old_page; get_page(dirty_page); |
9637a5efd [PATCH] add page_... |
2532 |
|
251b97f55 mm: dirty page ac... |
2533 |
reuse: |
9637a5efd [PATCH] add page_... |
2534 2535 2536 |
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... |
2537 |
if (ptep_set_access_flags(vma, address, page_table, entry,1)) |
4b3073e1c MM: Pass a PTE po... |
2538 |
update_mmu_cache(vma, address, page_table); |
72ddc8f72 do_wp_page: clari... |
2539 |
pte_unmap_unlock(page_table, ptl); |
9637a5efd [PATCH] add page_... |
2540 |
ret |= VM_FAULT_WRITE; |
72ddc8f72 do_wp_page: clari... |
2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 |
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. * |
a335b2e17 mm: Fix out-of-da... |
2551 |
* __do_fault is protected similarly. |
72ddc8f72 do_wp_page: clari... |
2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 |
*/ if (!page_mkwrite) { wait_on_page_locked(dirty_page); set_page_dirty_balance(dirty_page, page_mkwrite); } 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); } } /* file_update_time outside page_lock */ if (vma->vm_file) file_update_time(vma->vm_file); return ret; |
1da177e4c Linux-2.6.12-rc2 |
2578 |
} |
1da177e4c Linux-2.6.12-rc2 |
2579 2580 2581 2582 |
/* * Ok, we need to copy. Oh, well.. */ |
b5810039a [PATCH] core remo... |
2583 |
page_cache_get(old_page); |
920fc356f [PATCH] unpaged: ... |
2584 |
gotten: |
8f4e2101f [PATCH] mm: page ... |
2585 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2586 2587 |
if (unlikely(anon_vma_prepare(vma))) |
65500d234 [PATCH] mm: page ... |
2588 |
goto oom; |
a13ea5b75 mm: reinstate ZER... |
2589 |
|
62eede62d mm: ZERO_PAGE wit... |
2590 |
if (is_zero_pfn(pte_pfn(orig_pte))) { |
a13ea5b75 mm: reinstate ZER... |
2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 |
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); |
2c26fdd70 memcg: revert gfp... |
2601 |
if (mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL)) |
8a9f3ccd2 Memory controller... |
2602 |
goto oom_free_new; |
1da177e4c Linux-2.6.12-rc2 |
2603 2604 2605 |
/* * Re-check the pte - we dropped the lock */ |
8f4e2101f [PATCH] mm: page ... |
2606 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
65500d234 [PATCH] mm: page ... |
2607 |
if (likely(pte_same(*page_table, orig_pte))) { |
920fc356f [PATCH] unpaged: ... |
2608 |
if (old_page) { |
920fc356f [PATCH] unpaged: ... |
2609 |
if (!PageAnon(old_page)) { |
34e55232e mm: avoid false s... |
2610 2611 |
dec_mm_counter_fast(mm, MM_FILEPAGES); inc_mm_counter_fast(mm, MM_ANONPAGES); |
920fc356f [PATCH] unpaged: ... |
2612 2613 |
} } else |
34e55232e mm: avoid false s... |
2614 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
eca351336 [PATCH] Fix missi... |
2615 |
flush_cache_page(vma, address, pte_pfn(orig_pte)); |
65500d234 [PATCH] mm: page ... |
2616 2617 |
entry = mk_pte(new_page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
4ce072f1f [PATCH] mm: fix a... |
2618 2619 2620 2621 2622 2623 |
/* * 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... |
2624 |
ptep_clear_flush(vma, address, page_table); |
9617d95e6 [PATCH] mm: rmap ... |
2625 |
page_add_new_anon_rmap(new_page, vma, address); |
828502d30 ksm: add mmu_noti... |
2626 2627 2628 2629 2630 2631 |
/* * 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... |
2632 |
update_mmu_cache(vma, address, page_table); |
945754a17 mm: fix race in C... |
2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 |
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... |
2656 |
page_remove_rmap(old_page); |
945754a17 mm: fix race in C... |
2657 |
} |
1da177e4c Linux-2.6.12-rc2 |
2658 2659 |
/* Free the old page.. */ new_page = old_page; |
f33ea7f40 [PATCH] fix get_u... |
2660 |
ret |= VM_FAULT_WRITE; |
8a9f3ccd2 Memory controller... |
2661 2662 |
} else mem_cgroup_uncharge_page(new_page); |
920fc356f [PATCH] unpaged: ... |
2663 2664 |
if (new_page) page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2665 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2666 |
pte_unmap_unlock(page_table, ptl); |
e15f8c01a mlock: fix race w... |
2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 |
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... |
2679 |
return ret; |
8a9f3ccd2 Memory controller... |
2680 |
oom_free_new: |
6dbf6d3bb memcg: page_cache... |
2681 |
page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2682 |
oom: |
b827e496c mm: close page_mk... |
2683 2684 2685 2686 2687 |
if (old_page) { if (page_mkwrite) { unlock_page(old_page); page_cache_release(old_page); } |
920fc356f [PATCH] unpaged: ... |
2688 |
page_cache_release(old_page); |
b827e496c mm: close page_mk... |
2689 |
} |
1da177e4c Linux-2.6.12-rc2 |
2690 |
return VM_FAULT_OOM; |
9637a5efd [PATCH] add page_... |
2691 2692 2693 |
unwritable_page: page_cache_release(old_page); |
c2ec175c3 mm: page_mkwrite ... |
2694 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2695 |
} |
97a894136 mm: Remove i_mmap... |
2696 |
static void unmap_mapping_range_vma(struct vm_area_struct *vma, |
1da177e4c Linux-2.6.12-rc2 |
2697 2698 2699 |
unsigned long start_addr, unsigned long end_addr, struct zap_details *details) { |
f5cc4eef9 VM: make zap_page... |
2700 |
zap_page_range_single(vma, start_addr, end_addr - start_addr, details); |
1da177e4c Linux-2.6.12-rc2 |
2701 2702 2703 2704 2705 2706 2707 2708 |
} static inline void unmap_mapping_range_tree(struct prio_tree_root *root, struct zap_details *details) { struct vm_area_struct *vma; struct prio_tree_iter iter; pgoff_t vba, vea, zba, zea; |
1da177e4c Linux-2.6.12-rc2 |
2709 2710 |
vma_prio_tree_foreach(vma, &iter, root, details->first_index, details->last_index) { |
1da177e4c Linux-2.6.12-rc2 |
2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 |
vba = vma->vm_pgoff; vea = vba + ((vma->vm_end - vma->vm_start) >> PAGE_SHIFT) - 1; /* 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... |
2721 |
unmap_mapping_range_vma(vma, |
1da177e4c Linux-2.6.12-rc2 |
2722 2723 |
((zba - vba) << PAGE_SHIFT) + vma->vm_start, ((zea - vba + 1) << PAGE_SHIFT) + vma->vm_start, |
97a894136 mm: Remove i_mmap... |
2724 |
details); |
1da177e4c Linux-2.6.12-rc2 |
2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 |
} } 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. */ |
1da177e4c Linux-2.6.12-rc2 |
2739 |
list_for_each_entry(vma, head, shared.vm_set.list) { |
1da177e4c Linux-2.6.12-rc2 |
2740 |
details->nonlinear_vma = vma; |
97a894136 mm: Remove i_mmap... |
2741 |
unmap_mapping_range_vma(vma, vma->vm_start, vma->vm_end, details); |
1da177e4c Linux-2.6.12-rc2 |
2742 2743 2744 2745 |
} } /** |
72fd4a35a [PATCH] Numerous ... |
2746 |
* 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: ... |
2747 |
* @mapping: the address space containing mmaps to be unmapped. |
1da177e4c Linux-2.6.12-rc2 |
2748 2749 |
* @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... |
2750 |
* boundary. Note that this is different from truncate_pagecache(), which |
1da177e4c Linux-2.6.12-rc2 |
2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 |
* 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 |
2780 |
|
1da177e4c Linux-2.6.12-rc2 |
2781 |
|
3d48ae45e mm: Convert i_mma... |
2782 |
mutex_lock(&mapping->i_mmap_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2783 2784 2785 2786 |
if (unlikely(!prio_tree_empty(&mapping->i_mmap))) 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... |
2787 |
mutex_unlock(&mapping->i_mmap_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2788 2789 |
} EXPORT_SYMBOL(unmap_mapping_range); |
1da177e4c Linux-2.6.12-rc2 |
2790 |
/* |
8f4e2101f [PATCH] mm: page ... |
2791 2792 2793 |
* 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 |
2794 |
*/ |
65500d234 [PATCH] mm: page ... |
2795 2796 |
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... |
2797 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
2798 |
{ |
8f4e2101f [PATCH] mm: page ... |
2799 |
spinlock_t *ptl; |
4969c1192 mm: fix swapin ra... |
2800 |
struct page *page, *swapcache = NULL; |
65500d234 [PATCH] mm: page ... |
2801 |
swp_entry_t entry; |
1da177e4c Linux-2.6.12-rc2 |
2802 |
pte_t pte; |
d065bd810 mm: retry page fa... |
2803 |
int locked; |
56039efa1 memcg: fix ugly i... |
2804 |
struct mem_cgroup *ptr; |
ad8c2ee80 rmap: add exclusi... |
2805 |
int exclusive = 0; |
83c54070e mm: fault feedbac... |
2806 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2807 |
|
4c21e2f24 [PATCH] mm: split... |
2808 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
8f4e2101f [PATCH] mm: page ... |
2809 |
goto out; |
65500d234 [PATCH] mm: page ... |
2810 2811 |
entry = pte_to_swp_entry(orig_pte); |
d1737fdbe HWPOISON: Add bas... |
2812 2813 2814 2815 2816 2817 2818 |
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... |
2819 |
ret = VM_FAULT_SIGBUS; |
d1737fdbe HWPOISON: Add bas... |
2820 |
} |
0697212a4 [PATCH] Swapless ... |
2821 2822 |
goto out; } |
0ff922452 [PATCH] per-task-... |
2823 |
delayacct_set_flag(DELAYACCT_PF_SWAPIN); |
1da177e4c Linux-2.6.12-rc2 |
2824 2825 |
page = lookup_swap_cache(entry); if (!page) { |
02098feaa swapin needs gfp_... |
2826 2827 |
page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, vma, address); |
1da177e4c Linux-2.6.12-rc2 |
2828 2829 |
if (!page) { /* |
8f4e2101f [PATCH] mm: page ... |
2830 2831 |
* Back out if somebody else faulted in this pte * while we released the pte lock. |
1da177e4c Linux-2.6.12-rc2 |
2832 |
*/ |
8f4e2101f [PATCH] mm: page ... |
2833 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
2834 2835 |
if (likely(pte_same(*page_table, orig_pte))) ret = VM_FAULT_OOM; |
0ff922452 [PATCH] per-task-... |
2836 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
65500d234 [PATCH] mm: page ... |
2837 |
goto unlock; |
1da177e4c Linux-2.6.12-rc2 |
2838 2839 2840 2841 |
} /* Had to read the page from swap area: Major fault */ ret = VM_FAULT_MAJOR; |
f8891e5e1 [PATCH] Light wei... |
2842 |
count_vm_event(PGMAJFAULT); |
456f998ec memcg: add the pa... |
2843 |
mem_cgroup_count_vm_event(mm, PGMAJFAULT); |
d1737fdbe HWPOISON: Add bas... |
2844 |
} else if (PageHWPoison(page)) { |
71f72525d HWPOISON: comment... |
2845 2846 2847 2848 |
/* * hwpoisoned dirty swapcache pages are kept for killing * owner processes (which may be unknown at hwpoison time) */ |
d1737fdbe HWPOISON: Add bas... |
2849 2850 |
ret = VM_FAULT_HWPOISON; delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
4779cb31c HWPOISON: Fix pag... |
2851 |
goto out_release; |
1da177e4c Linux-2.6.12-rc2 |
2852 |
} |
d065bd810 mm: retry page fa... |
2853 |
locked = lock_page_or_retry(page, mm, flags); |
e709ffd61 mm: remove swap t... |
2854 |
|
073e587ec memcg: move charg... |
2855 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
d065bd810 mm: retry page fa... |
2856 2857 2858 2859 |
if (!locked) { ret |= VM_FAULT_RETRY; goto out_release; } |
073e587ec memcg: move charg... |
2860 |
|
4969c1192 mm: fix swapin ra... |
2861 |
/* |
31c4a3d3a mm: further fix s... |
2862 2863 2864 2865 |
* 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... |
2866 |
*/ |
31c4a3d3a mm: further fix s... |
2867 |
if (unlikely(!PageSwapCache(page) || page_private(page) != entry.val)) |
4969c1192 mm: fix swapin ra... |
2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 |
goto out_page; if (ksm_might_need_to_copy(page, vma, address)) { swapcache = page; page = ksm_does_need_to_copy(page, vma, address); if (unlikely(!page)) { ret = VM_FAULT_OOM; page = swapcache; swapcache = NULL; goto out_page; } |
5ad646880 ksm: let shared p... |
2880 |
} |
2c26fdd70 memcg: revert gfp... |
2881 |
if (mem_cgroup_try_charge_swapin(mm, page, GFP_KERNEL, &ptr)) { |
8a9f3ccd2 Memory controller... |
2882 |
ret = VM_FAULT_OOM; |
bc43f75cd mm: fix pageref l... |
2883 |
goto out_page; |
8a9f3ccd2 Memory controller... |
2884 |
} |
1da177e4c Linux-2.6.12-rc2 |
2885 |
/* |
8f4e2101f [PATCH] mm: page ... |
2886 |
* Back out if somebody else already faulted in this pte. |
1da177e4c Linux-2.6.12-rc2 |
2887 |
*/ |
8f4e2101f [PATCH] mm: page ... |
2888 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
9e9bef07c [PATCH] mm: do_sw... |
2889 |
if (unlikely(!pte_same(*page_table, orig_pte))) |
b81074800 [PATCH] do_swap_p... |
2890 |
goto out_nomap; |
b81074800 [PATCH] do_swap_p... |
2891 2892 2893 2894 |
if (unlikely(!PageUptodate(page))) { ret = VM_FAULT_SIGBUS; goto out_nomap; |
1da177e4c Linux-2.6.12-rc2 |
2895 |
} |
8c7c6e34a memcg: mem+swap c... |
2896 2897 2898 2899 2900 2901 2902 2903 |
/* * 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... |
2904 2905 2906 2907 |
* 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... |
2908 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2909 |
|
34e55232e mm: avoid false s... |
2910 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
b084d4353 mm: count swap usage |
2911 |
dec_mm_counter_fast(mm, MM_SWAPENTS); |
1da177e4c Linux-2.6.12-rc2 |
2912 |
pte = mk_pte(page, vma->vm_page_prot); |
30c9f3a9f Remove internal u... |
2913 |
if ((flags & FAULT_FLAG_WRITE) && reuse_swap_page(page)) { |
1da177e4c Linux-2.6.12-rc2 |
2914 |
pte = maybe_mkwrite(pte_mkdirty(pte), vma); |
30c9f3a9f Remove internal u... |
2915 |
flags &= ~FAULT_FLAG_WRITE; |
9a5b489b8 mm: set VM_FAULT_... |
2916 |
ret |= VM_FAULT_WRITE; |
ad8c2ee80 rmap: add exclusi... |
2917 |
exclusive = 1; |
1da177e4c Linux-2.6.12-rc2 |
2918 |
} |
1da177e4c Linux-2.6.12-rc2 |
2919 2920 |
flush_icache_page(vma, page); set_pte_at(mm, address, page_table, pte); |
ad8c2ee80 rmap: add exclusi... |
2921 |
do_page_add_anon_rmap(page, vma, address, exclusive); |
03f3c4336 memcg: fix swap a... |
2922 2923 |
/* It's better to call commit-charge after rmap is established */ mem_cgroup_commit_charge_swapin(page, ptr); |
1da177e4c Linux-2.6.12-rc2 |
2924 |
|
c475a8ab6 [PATCH] can_share... |
2925 |
swap_free(entry); |
b291f0003 mlock: mlocked pa... |
2926 |
if (vm_swap_full() || (vma->vm_flags & VM_LOCKED) || PageMlocked(page)) |
a2c43eed8 mm: try_to_free_s... |
2927 |
try_to_free_swap(page); |
c475a8ab6 [PATCH] can_share... |
2928 |
unlock_page(page); |
4969c1192 mm: fix swapin ra... |
2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 |
if (swapcache) { /* * 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... |
2941 |
|
30c9f3a9f Remove internal u... |
2942 |
if (flags & FAULT_FLAG_WRITE) { |
61469f1d5 memcg: when do_sw... |
2943 2944 2945 |
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 |
2946 2947 2948 2949 |
goto out; } /* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
2950 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
2951 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2952 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2953 2954 |
out: return ret; |
b81074800 [PATCH] do_swap_p... |
2955 |
out_nomap: |
7a81b88cb memcg: introduce ... |
2956 |
mem_cgroup_cancel_charge_swapin(ptr); |
8f4e2101f [PATCH] mm: page ... |
2957 |
pte_unmap_unlock(page_table, ptl); |
bc43f75cd mm: fix pageref l... |
2958 |
out_page: |
b81074800 [PATCH] do_swap_p... |
2959 |
unlock_page(page); |
4779cb31c HWPOISON: Fix pag... |
2960 |
out_release: |
b81074800 [PATCH] do_swap_p... |
2961 |
page_cache_release(page); |
4969c1192 mm: fix swapin ra... |
2962 2963 2964 2965 |
if (swapcache) { unlock_page(swapcache); page_cache_release(swapcache); } |
65500d234 [PATCH] mm: page ... |
2966 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2967 2968 2969 |
} /* |
8ca3eb080 guard page for st... |
2970 2971 |
* 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 ... |
2972 |
* doesn't hit another vma. |
320b2b8de mm: keep a guard ... |
2973 2974 2975 2976 2977 |
*/ 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... |
2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 |
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 ... |
2988 |
|
d05f3169c mm: make expand_d... |
2989 |
expand_downwards(vma, address - PAGE_SIZE); |
320b2b8de mm: keep a guard ... |
2990 |
} |
8ca3eb080 guard page for st... |
2991 2992 2993 2994 2995 2996 2997 2998 2999 |
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 ... |
3000 3001 3002 3003 |
return 0; } /* |
8f4e2101f [PATCH] mm: page ... |
3004 3005 3006 |
* 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 |
3007 |
*/ |
65500d234 [PATCH] mm: page ... |
3008 3009 |
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... |
3010 |
unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3011 |
{ |
8f4e2101f [PATCH] mm: page ... |
3012 3013 |
struct page *page; spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
3014 |
pte_t entry; |
1da177e4c Linux-2.6.12-rc2 |
3015 |
|
11ac55247 mm: fix page tabl... |
3016 3017 3018 3019 |
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 ... |
3020 |
return VM_FAULT_SIGBUS; |
11ac55247 mm: fix page tabl... |
3021 |
/* Use the zero-page for reads */ |
62eede62d mm: ZERO_PAGE wit... |
3022 3023 3024 |
if (!(flags & FAULT_FLAG_WRITE)) { entry = pte_mkspecial(pfn_pte(my_zero_pfn(address), vma->vm_page_prot)); |
11ac55247 mm: fix page tabl... |
3025 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
a13ea5b75 mm: reinstate ZER... |
3026 3027 3028 3029 |
if (!pte_none(*page_table)) goto unlock; goto setpte; } |
557ed1fa2 remove ZERO_PAGE |
3030 |
/* Allocate our own private page. */ |
557ed1fa2 remove ZERO_PAGE |
3031 3032 3033 3034 3035 |
if (unlikely(anon_vma_prepare(vma))) goto oom; page = alloc_zeroed_user_highpage_movable(vma, address); if (!page) goto oom; |
0ed361dec mm: fix PageUptod... |
3036 |
__SetPageUptodate(page); |
8f4e2101f [PATCH] mm: page ... |
3037 |
|
2c26fdd70 memcg: revert gfp... |
3038 |
if (mem_cgroup_newpage_charge(page, mm, GFP_KERNEL)) |
8a9f3ccd2 Memory controller... |
3039 |
goto oom_free_page; |
557ed1fa2 remove ZERO_PAGE |
3040 |
entry = mk_pte(page, vma->vm_page_prot); |
1ac0cb5d0 mm: fix anonymous... |
3041 3042 |
if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); |
1da177e4c Linux-2.6.12-rc2 |
3043 |
|
557ed1fa2 remove ZERO_PAGE |
3044 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1c2fb7a4c ksm: fix deadlock... |
3045 |
if (!pte_none(*page_table)) |
557ed1fa2 remove ZERO_PAGE |
3046 |
goto release; |
9ba692948 ksm: fix oom dead... |
3047 |
|
34e55232e mm: avoid false s... |
3048 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
557ed1fa2 remove ZERO_PAGE |
3049 |
page_add_new_anon_rmap(page, vma, address); |
a13ea5b75 mm: reinstate ZER... |
3050 |
setpte: |
65500d234 [PATCH] mm: page ... |
3051 |
set_pte_at(mm, address, page_table, entry); |
1da177e4c Linux-2.6.12-rc2 |
3052 3053 |
/* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
3054 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
3055 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
3056 |
pte_unmap_unlock(page_table, ptl); |
83c54070e mm: fault feedbac... |
3057 |
return 0; |
8f4e2101f [PATCH] mm: page ... |
3058 |
release: |
8a9f3ccd2 Memory controller... |
3059 |
mem_cgroup_uncharge_page(page); |
8f4e2101f [PATCH] mm: page ... |
3060 3061 |
page_cache_release(page); goto unlock; |
8a9f3ccd2 Memory controller... |
3062 |
oom_free_page: |
6dbf6d3bb memcg: page_cache... |
3063 |
page_cache_release(page); |
65500d234 [PATCH] mm: page ... |
3064 |
oom: |
1da177e4c Linux-2.6.12-rc2 |
3065 3066 3067 3068 |
return VM_FAULT_OOM; } /* |
54cb8821d mm: merge populat... |
3069 |
* __do_fault() tries to create a new page mapping. It aggressively |
1da177e4c Linux-2.6.12-rc2 |
3070 |
* tries to share with existing pages, but makes a separate copy if |
54cb8821d mm: merge populat... |
3071 3072 |
* the FAULT_FLAG_WRITE is set in the flags parameter in order to avoid * the next page fault. |
1da177e4c Linux-2.6.12-rc2 |
3073 3074 3075 3076 |
* * As this is called only for pages that do not currently exist, we * do not need to flush old virtual caches or the TLB. * |
8f4e2101f [PATCH] mm: page ... |
3077 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, |
16abfa086 Fix sys_remap_fil... |
3078 |
* but allow concurrent faults), and pte neither mapped nor locked. |
8f4e2101f [PATCH] mm: page ... |
3079 |
* We return with mmap_sem still held, but pte unmapped and unlocked. |
1da177e4c Linux-2.6.12-rc2 |
3080 |
*/ |
54cb8821d mm: merge populat... |
3081 |
static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
16abfa086 Fix sys_remap_fil... |
3082 |
unsigned long address, pmd_t *pmd, |
54cb8821d mm: merge populat... |
3083 |
pgoff_t pgoff, unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
3084 |
{ |
16abfa086 Fix sys_remap_fil... |
3085 |
pte_t *page_table; |
8f4e2101f [PATCH] mm: page ... |
3086 |
spinlock_t *ptl; |
d0217ac04 mm: fault feedbac... |
3087 |
struct page *page; |
1d65f86db mm: preallocate p... |
3088 |
struct page *cow_page; |
1da177e4c Linux-2.6.12-rc2 |
3089 |
pte_t entry; |
1da177e4c Linux-2.6.12-rc2 |
3090 |
int anon = 0; |
d08b3851d [PATCH] mm: track... |
3091 |
struct page *dirty_page = NULL; |
d0217ac04 mm: fault feedbac... |
3092 3093 |
struct vm_fault vmf; int ret; |
a200ee182 mm: set_page_dirt... |
3094 |
int page_mkwrite = 0; |
54cb8821d mm: merge populat... |
3095 |
|
1d65f86db mm: preallocate p... |
3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 |
/* * If we do COW later, allocate page befor taking lock_page() * on the file cache page. This will reduce lock holding time. */ if ((flags & FAULT_FLAG_WRITE) && !(vma->vm_flags & VM_SHARED)) { if (unlikely(anon_vma_prepare(vma))) return VM_FAULT_OOM; cow_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); if (!cow_page) return VM_FAULT_OOM; if (mem_cgroup_newpage_charge(cow_page, mm, GFP_KERNEL)) { page_cache_release(cow_page); return VM_FAULT_OOM; } } else cow_page = NULL; |
d0217ac04 mm: fault feedbac... |
3115 3116 3117 3118 |
vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = pgoff; vmf.flags = flags; vmf.page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
3119 |
|
3c18ddd16 mm: remove nopage |
3120 |
ret = vma->vm_ops->fault(vma, &vmf); |
d065bd810 mm: retry page fa... |
3121 3122 |
if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) |
1d65f86db mm: preallocate p... |
3123 |
goto uncharge_out; |
1da177e4c Linux-2.6.12-rc2 |
3124 |
|
a3b947eac HWPOISON: Add poi... |
3125 3126 3127 |
if (unlikely(PageHWPoison(vmf.page))) { if (ret & VM_FAULT_LOCKED) unlock_page(vmf.page); |
1d65f86db mm: preallocate p... |
3128 3129 |
ret = VM_FAULT_HWPOISON; goto uncharge_out; |
a3b947eac HWPOISON: Add poi... |
3130 |
} |
d00806b18 mm: fix fault vs ... |
3131 |
/* |
d0217ac04 mm: fault feedbac... |
3132 |
* For consistency in subsequent calls, make the faulted page always |
d00806b18 mm: fix fault vs ... |
3133 3134 |
* locked. */ |
83c54070e mm: fault feedbac... |
3135 |
if (unlikely(!(ret & VM_FAULT_LOCKED))) |
d0217ac04 mm: fault feedbac... |
3136 |
lock_page(vmf.page); |
54cb8821d mm: merge populat... |
3137 |
else |
d0217ac04 mm: fault feedbac... |
3138 |
VM_BUG_ON(!PageLocked(vmf.page)); |
d00806b18 mm: fix fault vs ... |
3139 |
|
1da177e4c Linux-2.6.12-rc2 |
3140 3141 3142 |
/* * Should we do an early C-O-W break? */ |
d0217ac04 mm: fault feedbac... |
3143 |
page = vmf.page; |
54cb8821d mm: merge populat... |
3144 |
if (flags & FAULT_FLAG_WRITE) { |
9637a5efd [PATCH] add page_... |
3145 |
if (!(vma->vm_flags & VM_SHARED)) { |
1d65f86db mm: preallocate p... |
3146 |
page = cow_page; |
54cb8821d mm: merge populat... |
3147 |
anon = 1; |
d0217ac04 mm: fault feedbac... |
3148 |
copy_user_highpage(page, vmf.page, address, vma); |
0ed361dec mm: fix PageUptod... |
3149 |
__SetPageUptodate(page); |
9637a5efd [PATCH] add page_... |
3150 |
} else { |
54cb8821d mm: merge populat... |
3151 3152 |
/* * If the page will be shareable, see if the backing |
9637a5efd [PATCH] add page_... |
3153 |
* address space wants to know that the page is about |
54cb8821d mm: merge populat... |
3154 3155 |
* to become writable */ |
696761476 ocfs2: release pa... |
3156 |
if (vma->vm_ops->page_mkwrite) { |
c2ec175c3 mm: page_mkwrite ... |
3157 |
int tmp; |
696761476 ocfs2: release pa... |
3158 |
unlock_page(page); |
b827e496c mm: close page_mk... |
3159 |
vmf.flags = FAULT_FLAG_WRITE|FAULT_FLAG_MKWRITE; |
c2ec175c3 mm: page_mkwrite ... |
3160 3161 3162 3163 |
tmp = vma->vm_ops->page_mkwrite(vma, &vmf); if (unlikely(tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) { ret = tmp; |
b827e496c mm: close page_mk... |
3164 |
goto unwritable_page; |
d0217ac04 mm: fault feedbac... |
3165 |
} |
b827e496c mm: close page_mk... |
3166 3167 3168 3169 3170 3171 3172 3173 3174 |
if (unlikely(!(tmp & VM_FAULT_LOCKED))) { lock_page(page); if (!page->mapping) { ret = 0; /* retry the fault */ unlock_page(page); goto unwritable_page; } } else VM_BUG_ON(!PageLocked(page)); |
a200ee182 mm: set_page_dirt... |
3175 |
page_mkwrite = 1; |
9637a5efd [PATCH] add page_... |
3176 3177 |
} } |
54cb8821d mm: merge populat... |
3178 |
|
1da177e4c Linux-2.6.12-rc2 |
3179 |
} |
8f4e2101f [PATCH] mm: page ... |
3180 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
3181 3182 3183 3184 3185 3186 |
/* * This silly early PAGE_DIRTY setting removes a race * due to the bad i386 page protection. But it's valid * for other architectures too. * |
30c9f3a9f Remove internal u... |
3187 |
* Note that if FAULT_FLAG_WRITE is set, we either now have |
1da177e4c Linux-2.6.12-rc2 |
3188 3189 3190 3191 3192 |
* an exclusive copy of the page, or this is a shared mapping, * so we can make it writable and dirty to avoid having to * handle that later. */ /* Only go through if we didn't race with anybody else... */ |
1c2fb7a4c ksm: fix deadlock... |
3193 |
if (likely(pte_same(*page_table, orig_pte))) { |
d00806b18 mm: fix fault vs ... |
3194 3195 |
flush_icache_page(vma, page); entry = mk_pte(page, vma->vm_page_prot); |
54cb8821d mm: merge populat... |
3196 |
if (flags & FAULT_FLAG_WRITE) |
1da177e4c Linux-2.6.12-rc2 |
3197 |
entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
1da177e4c Linux-2.6.12-rc2 |
3198 |
if (anon) { |
34e55232e mm: avoid false s... |
3199 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
64d6519dd swap: cull unevic... |
3200 |
page_add_new_anon_rmap(page, vma, address); |
f57e88a8d [PATCH] unpaged: ... |
3201 |
} else { |
34e55232e mm: avoid false s... |
3202 |
inc_mm_counter_fast(mm, MM_FILEPAGES); |
d00806b18 mm: fix fault vs ... |
3203 |
page_add_file_rmap(page); |
54cb8821d mm: merge populat... |
3204 |
if (flags & FAULT_FLAG_WRITE) { |
d00806b18 mm: fix fault vs ... |
3205 |
dirty_page = page; |
d08b3851d [PATCH] mm: track... |
3206 3207 |
get_page(dirty_page); } |
4294621f4 [PATCH] mm: rss =... |
3208 |
} |
64d6519dd swap: cull unevic... |
3209 |
set_pte_at(mm, address, page_table, entry); |
d00806b18 mm: fix fault vs ... |
3210 3211 |
/* no need to invalidate: a not-present page won't be cached */ |
4b3073e1c MM: Pass a PTE po... |
3212 |
update_mmu_cache(vma, address, page_table); |
1da177e4c Linux-2.6.12-rc2 |
3213 |
} else { |
1d65f86db mm: preallocate p... |
3214 3215 |
if (cow_page) mem_cgroup_uncharge_page(cow_page); |
d00806b18 mm: fix fault vs ... |
3216 3217 3218 |
if (anon) page_cache_release(page); else |
54cb8821d mm: merge populat... |
3219 |
anon = 1; /* no anon but release faulted_page */ |
1da177e4c Linux-2.6.12-rc2 |
3220 |
} |
8f4e2101f [PATCH] mm: page ... |
3221 |
pte_unmap_unlock(page_table, ptl); |
d00806b18 mm: fix fault vs ... |
3222 |
|
b827e496c mm: close page_mk... |
3223 3224 |
if (dirty_page) { struct address_space *mapping = page->mapping; |
8f7b3d156 Update ctime and ... |
3225 |
|
b827e496c mm: close page_mk... |
3226 3227 3228 |
if (set_page_dirty(dirty_page)) page_mkwrite = 1; unlock_page(dirty_page); |
d08b3851d [PATCH] mm: track... |
3229 |
put_page(dirty_page); |
b827e496c mm: close page_mk... |
3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 |
if (page_mkwrite && mapping) { /* * Some device drivers do not set page.mapping but still * dirty their pages */ balance_dirty_pages_ratelimited(mapping); } /* file_update_time outside page_lock */ if (vma->vm_file) file_update_time(vma->vm_file); } else { unlock_page(vmf.page); if (anon) page_cache_release(vmf.page); |
d08b3851d [PATCH] mm: track... |
3245 |
} |
d00806b18 mm: fix fault vs ... |
3246 |
|
83c54070e mm: fault feedbac... |
3247 |
return ret; |
b827e496c mm: close page_mk... |
3248 3249 3250 3251 |
unwritable_page: page_cache_release(page); return ret; |
1d65f86db mm: preallocate p... |
3252 3253 3254 3255 3256 3257 3258 |
uncharge_out: /* fs's fault handler get error */ if (cow_page) { mem_cgroup_uncharge_page(cow_page); page_cache_release(cow_page); } return ret; |
54cb8821d mm: merge populat... |
3259 |
} |
d00806b18 mm: fix fault vs ... |
3260 |
|
54cb8821d mm: merge populat... |
3261 3262 |
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... |
3263 |
unsigned int flags, pte_t orig_pte) |
54cb8821d mm: merge populat... |
3264 3265 |
{ pgoff_t pgoff = (((address & PAGE_MASK) |
0da7e01f5 calculation of pg... |
3266 |
- vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; |
54cb8821d mm: merge populat... |
3267 |
|
16abfa086 Fix sys_remap_fil... |
3268 3269 |
pte_unmap(page_table); return __do_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
54cb8821d mm: merge populat... |
3270 |
} |
f4b81804a [PATCH] do_no_pfn() |
3271 |
/* |
1da177e4c Linux-2.6.12-rc2 |
3272 3273 3274 |
* 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 ... |
3275 3276 3277 3278 |
* * 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 |
3279 |
*/ |
d0217ac04 mm: fault feedbac... |
3280 |
static int do_nonlinear_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
65500d234 [PATCH] mm: page ... |
3281 |
unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f Remove internal u... |
3282 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
3283 |
{ |
65500d234 [PATCH] mm: page ... |
3284 |
pgoff_t pgoff; |
1da177e4c Linux-2.6.12-rc2 |
3285 |
|
30c9f3a9f Remove internal u... |
3286 |
flags |= FAULT_FLAG_NONLINEAR; |
4c21e2f24 [PATCH] mm: split... |
3287 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
83c54070e mm: fault feedbac... |
3288 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3289 |
|
2509ef26d badpage: zap prin... |
3290 |
if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) { |
65500d234 [PATCH] mm: page ... |
3291 3292 3293 |
/* * Page table corrupted: show pte and kill process. */ |
3dc147414 badpage: replace ... |
3294 |
print_bad_pte(vma, address, orig_pte, NULL); |
d99be1a8e mm: sigbus instea... |
3295 |
return VM_FAULT_SIGBUS; |
65500d234 [PATCH] mm: page ... |
3296 |
} |
65500d234 [PATCH] mm: page ... |
3297 3298 |
pgoff = pte_to_pgoff(orig_pte); |
16abfa086 Fix sys_remap_fil... |
3299 |
return __do_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
1da177e4c Linux-2.6.12-rc2 |
3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 |
} /* * 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... |
3311 3312 3313 |
* 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 |
3314 |
*/ |
71e3aac07 thp: transparent ... |
3315 3316 3317 |
int handle_pte_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *pte, pmd_t *pmd, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3318 3319 |
{ pte_t entry; |
8f4e2101f [PATCH] mm: page ... |
3320 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
3321 |
|
8dab5241d Rework ptep_set_a... |
3322 |
entry = *pte; |
1da177e4c Linux-2.6.12-rc2 |
3323 |
if (!pte_present(entry)) { |
65500d234 [PATCH] mm: page ... |
3324 |
if (pte_none(entry)) { |
f4b81804a [PATCH] do_no_pfn() |
3325 |
if (vma->vm_ops) { |
3c18ddd16 mm: remove nopage |
3326 |
if (likely(vma->vm_ops->fault)) |
54cb8821d mm: merge populat... |
3327 |
return do_linear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3328 |
pte, pmd, flags, entry); |
f4b81804a [PATCH] do_no_pfn() |
3329 3330 |
} return do_anonymous_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3331 |
pte, pmd, flags); |
65500d234 [PATCH] mm: page ... |
3332 |
} |
1da177e4c Linux-2.6.12-rc2 |
3333 |
if (pte_file(entry)) |
d0217ac04 mm: fault feedbac... |
3334 |
return do_nonlinear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3335 |
pte, pmd, flags, entry); |
65500d234 [PATCH] mm: page ... |
3336 |
return do_swap_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3337 |
pte, pmd, flags, entry); |
1da177e4c Linux-2.6.12-rc2 |
3338 |
} |
4c21e2f24 [PATCH] mm: split... |
3339 |
ptl = pte_lockptr(mm, pmd); |
8f4e2101f [PATCH] mm: page ... |
3340 3341 3342 |
spin_lock(ptl); if (unlikely(!pte_same(*pte, entry))) goto unlock; |
30c9f3a9f Remove internal u... |
3343 |
if (flags & FAULT_FLAG_WRITE) { |
1da177e4c Linux-2.6.12-rc2 |
3344 |
if (!pte_write(entry)) |
8f4e2101f [PATCH] mm: page ... |
3345 3346 |
return do_wp_page(mm, vma, address, pte, pmd, ptl, entry); |
1da177e4c Linux-2.6.12-rc2 |
3347 3348 3349 |
entry = pte_mkdirty(entry); } entry = pte_mkyoung(entry); |
30c9f3a9f Remove internal u... |
3350 |
if (ptep_set_access_flags(vma, address, pte, entry, flags & FAULT_FLAG_WRITE)) { |
4b3073e1c MM: Pass a PTE po... |
3351 |
update_mmu_cache(vma, address, pte); |
1a44e1490 [PATCH] .text pag... |
3352 3353 3354 3355 3356 3357 3358 |
} 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... |
3359 |
if (flags & FAULT_FLAG_WRITE) |
61c77326d x86, mm: Avoid un... |
3360 |
flush_tlb_fix_spurious_fault(vma, address); |
1a44e1490 [PATCH] .text pag... |
3361 |
} |
8f4e2101f [PATCH] mm: page ... |
3362 3363 |
unlock: pte_unmap_unlock(pte, ptl); |
83c54070e mm: fault feedbac... |
3364 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3365 3366 3367 3368 3369 |
} /* * By the time we get here, we already hold the mm semaphore */ |
83c54070e mm: fault feedbac... |
3370 |
int handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
d06063cc2 Move FAULT_FLAG_x... |
3371 |
unsigned long address, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3372 3373 3374 3375 3376 3377 3378 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; __set_current_state(TASK_RUNNING); |
f8891e5e1 [PATCH] Light wei... |
3379 |
count_vm_event(PGFAULT); |
456f998ec memcg: add the pa... |
3380 |
mem_cgroup_count_vm_event(mm, PGFAULT); |
1da177e4c Linux-2.6.12-rc2 |
3381 |
|
34e55232e mm: avoid false s... |
3382 3383 |
/* do counter updates before entering really critical section. */ check_sync_rss_stat(current); |
ac9b9c667 [PATCH] Fix handl... |
3384 |
if (unlikely(is_vm_hugetlb_page(vma))) |
30c9f3a9f Remove internal u... |
3385 |
return hugetlb_fault(mm, vma, address, flags); |
1da177e4c Linux-2.6.12-rc2 |
3386 |
|
1f1d06c34 thp, memcg: split... |
3387 |
retry: |
1da177e4c Linux-2.6.12-rc2 |
3388 |
pgd = pgd_offset(mm, address); |
1da177e4c Linux-2.6.12-rc2 |
3389 3390 |
pud = pud_alloc(mm, pgd, address); if (!pud) |
c74df32c7 [PATCH] mm: ptd_a... |
3391 |
return VM_FAULT_OOM; |
1da177e4c Linux-2.6.12-rc2 |
3392 3393 |
pmd = pmd_alloc(mm, pud, address); if (!pmd) |
c74df32c7 [PATCH] mm: ptd_a... |
3394 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3395 3396 3397 3398 3399 3400 |
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma)) { if (!vma->vm_ops) return do_huge_pmd_anonymous_page(mm, vma, address, pmd, flags); } else { pmd_t orig_pmd = *pmd; |
1f1d06c34 thp, memcg: split... |
3401 |
int ret; |
71e3aac07 thp: transparent ... |
3402 3403 3404 3405 |
barrier(); if (pmd_trans_huge(orig_pmd)) { if (flags & FAULT_FLAG_WRITE && !pmd_write(orig_pmd) && |
1f1d06c34 thp, memcg: split... |
3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 |
!pmd_trans_splitting(orig_pmd)) { ret = do_huge_pmd_wp_page(mm, vma, address, pmd, orig_pmd); /* * If COW results in an oom, the huge pmd will * have been split, so retry the fault on the * pte for a smaller charge. */ if (unlikely(ret & VM_FAULT_OOM)) goto retry; return ret; } |
71e3aac07 thp: transparent ... |
3418 3419 3420 3421 3422 3423 3424 3425 3426 |
return 0; } } /* * 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. */ |
cc03638df mm: check if PTE ... |
3427 |
if (unlikely(pmd_none(*pmd)) && __pte_alloc(mm, vma, pmd, address)) |
c74df32c7 [PATCH] mm: ptd_a... |
3428 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 |
/* 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 |
3439 |
|
30c9f3a9f Remove internal u... |
3440 |
return handle_pte_fault(mm, vma, address, pte, pmd, flags); |
1da177e4c Linux-2.6.12-rc2 |
3441 3442 3443 3444 3445 |
} #ifndef __PAGETABLE_PUD_FOLDED /* * Allocate page upper directory. |
872fec16d [PATCH] mm: init_... |
3446 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3447 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3448 |
int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3449 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3450 3451 |
pud_t *new = pud_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3452 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3453 |
|
362a61ad6 fix SMP data race... |
3454 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3455 |
spin_lock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3456 |
if (pgd_present(*pgd)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3457 |
pud_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3458 3459 |
else pgd_populate(mm, pgd, new); |
c74df32c7 [PATCH] mm: ptd_a... |
3460 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3461 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3462 3463 3464 3465 3466 3467 |
} #endif /* __PAGETABLE_PUD_FOLDED */ #ifndef __PAGETABLE_PMD_FOLDED /* * Allocate page middle directory. |
872fec16d [PATCH] mm: init_... |
3468 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3469 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3470 |
int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3471 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3472 3473 |
pmd_t *new = pmd_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3474 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3475 |
|
362a61ad6 fix SMP data race... |
3476 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3477 |
spin_lock(&mm->page_table_lock); |
1da177e4c Linux-2.6.12-rc2 |
3478 |
#ifndef __ARCH_HAS_4LEVEL_HACK |
1bb3630e8 [PATCH] mm: ptd_a... |
3479 |
if (pud_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3480 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3481 3482 |
else pud_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3483 |
#else |
1bb3630e8 [PATCH] mm: ptd_a... |
3484 |
if (pgd_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3485 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3486 3487 |
else pgd_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3488 |
#endif /* __ARCH_HAS_4LEVEL_HACK */ |
c74df32c7 [PATCH] mm: ptd_a... |
3489 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3490 |
return 0; |
e0f39591c [PATCH] Workaroun... |
3491 |
} |
1da177e4c Linux-2.6.12-rc2 |
3492 3493 3494 3495 3496 3497 3498 3499 3500 |
#endif /* __PAGETABLE_PMD_FOLDED */ int make_pages_present(unsigned long addr, unsigned long end) { int ret, len, write; struct vm_area_struct * vma; vma = find_vma(current->mm, addr); if (!vma) |
a477097d9 mlock() fix retur... |
3501 |
return -ENOMEM; |
5ecfda041 mlock: avoid dirt... |
3502 3503 3504 3505 3506 3507 |
/* * We want to touch writable mappings with a write fault in order * to break COW, except for shared mappings because these don't COW * and we would not want to dirty them for nothing. */ write = (vma->vm_flags & (VM_WRITE | VM_SHARED)) == VM_WRITE; |
5bcb28b13 BUG_ON() Conversi... |
3508 3509 |
BUG_ON(addr >= end); BUG_ON(end > vma->vm_end); |
68e116a3b MM: use DIV_ROUND... |
3510 |
len = DIV_ROUND_UP(end, PAGE_SIZE) - addr/PAGE_SIZE; |
1da177e4c Linux-2.6.12-rc2 |
3511 3512 |
ret = get_user_pages(current, current->mm, addr, len, write, 0, NULL, NULL); |
c11d69d8c mlock: revert mai... |
3513 |
if (ret < 0) |
1da177e4c Linux-2.6.12-rc2 |
3514 |
return ret; |
9978ad583 mlock: make mlock... |
3515 |
return ret == len ? 0 : -EFAULT; |
1da177e4c Linux-2.6.12-rc2 |
3516 |
} |
1da177e4c Linux-2.6.12-rc2 |
3517 3518 3519 |
#if !defined(__HAVE_ARCH_GATE_AREA) #if defined(AT_SYSINFO_EHDR) |
5ce7852cd [PATCH] mm/filema... |
3520 |
static struct vm_area_struct gate_vma; |
1da177e4c Linux-2.6.12-rc2 |
3521 3522 3523 3524 3525 3526 |
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_... |
3527 3528 |
gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC; gate_vma.vm_page_prot = __P101; |
909af768e coredump: remove ... |
3529 |
|
1da177e4c Linux-2.6.12-rc2 |
3530 3531 3532 3533 |
return 0; } __initcall(gate_vma_init); #endif |
31db58b3a mm: arch: make ge... |
3534 |
struct vm_area_struct *get_gate_vma(struct mm_struct *mm) |
1da177e4c Linux-2.6.12-rc2 |
3535 3536 3537 3538 3539 3540 3541 |
{ #ifdef AT_SYSINFO_EHDR return &gate_vma; #else return NULL; #endif } |
cae5d3903 mm: arch: rename ... |
3542 |
int in_gate_area_no_mm(unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
3543 3544 3545 3546 3547 3548 3549 3550 3551 |
{ #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... |
3552 |
|
1b36ba815 mm: wrap follow_p... |
3553 |
static int __follow_pte(struct mm_struct *mm, unsigned long address, |
f8ad0f499 mm: introduce fol... |
3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 |
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... |
3570 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
f8ad0f499 mm: introduce fol... |
3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 |
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... |
3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 |
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... |
3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 |
/** * 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... |
3628 |
#ifdef CONFIG_HAVE_IOREMAP_PROT |
d87fe6607 x86: PAT: modify ... |
3629 3630 3631 |
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... |
3632 |
{ |
03668a4de mm: use generic f... |
3633 |
int ret = -EINVAL; |
28b2ee20c access_process_vm... |
3634 3635 |
pte_t *ptep, pte; spinlock_t *ptl; |
28b2ee20c access_process_vm... |
3636 |
|
d87fe6607 x86: PAT: modify ... |
3637 3638 |
if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) goto out; |
28b2ee20c access_process_vm... |
3639 |
|
03668a4de mm: use generic f... |
3640 |
if (follow_pte(vma->vm_mm, address, &ptep, &ptl)) |
d87fe6607 x86: PAT: modify ... |
3641 |
goto out; |
28b2ee20c access_process_vm... |
3642 |
pte = *ptep; |
03668a4de mm: use generic f... |
3643 |
|
28b2ee20c access_process_vm... |
3644 3645 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
28b2ee20c access_process_vm... |
3646 3647 |
*prot = pgprot_val(pte_pgprot(pte)); |
03668a4de mm: use generic f... |
3648 |
*phys = (resource_size_t)pte_pfn(pte) << PAGE_SHIFT; |
28b2ee20c access_process_vm... |
3649 |
|
03668a4de mm: use generic f... |
3650 |
ret = 0; |
28b2ee20c access_process_vm... |
3651 3652 3653 |
unlock: pte_unmap_unlock(ptep, ptl); out: |
d87fe6607 x86: PAT: modify ... |
3654 |
return ret; |
28b2ee20c access_process_vm... |
3655 3656 3657 3658 3659 3660 3661 |
} 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 __... |
3662 |
void __iomem *maddr; |
28b2ee20c access_process_vm... |
3663 |
int offset = addr & (PAGE_SIZE-1); |
d87fe6607 x86: PAT: modify ... |
3664 |
if (follow_phys(vma, addr, write, &prot, &phys_addr)) |
28b2ee20c access_process_vm... |
3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 |
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; } #endif |
0ec76a110 [PATCH] NOMMU: Ch... |
3677 |
/* |
206cb6365 mm: factor out ma... |
3678 3679 |
* Access another process' address space as given in mm. If non-NULL, use the * given task for page fault accounting. |
0ec76a110 [PATCH] NOMMU: Ch... |
3680 |
*/ |
206cb6365 mm: factor out ma... |
3681 3682 |
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... |
3683 |
{ |
0ec76a110 [PATCH] NOMMU: Ch... |
3684 |
struct vm_area_struct *vma; |
0ec76a110 [PATCH] NOMMU: Ch... |
3685 |
void *old_buf = buf; |
0ec76a110 [PATCH] NOMMU: Ch... |
3686 |
down_read(&mm->mmap_sem); |
183ff22bb spelling fixes: mm/ |
3687 |
/* ignore errors, just check how much was successfully transferred */ |
0ec76a110 [PATCH] NOMMU: Ch... |
3688 3689 3690 |
while (len) { int bytes, ret, offset; void *maddr; |
28b2ee20c access_process_vm... |
3691 |
struct page *page = NULL; |
0ec76a110 [PATCH] NOMMU: Ch... |
3692 3693 3694 |
ret = get_user_pages(tsk, mm, addr, 1, write, 1, &page, &vma); |
28b2ee20c access_process_vm... |
3695 3696 3697 3698 3699 3700 3701 |
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... |
3702 |
if (!vma || vma->vm_start > addr) |
28b2ee20c access_process_vm... |
3703 3704 3705 3706 3707 3708 3709 3710 |
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... |
3711 |
} else { |
28b2ee20c access_process_vm... |
3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 |
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... |
3728 |
} |
0ec76a110 [PATCH] NOMMU: Ch... |
3729 3730 3731 3732 3733 |
len -= bytes; buf += bytes; addr += bytes; } up_read(&mm->mmap_sem); |
0ec76a110 [PATCH] NOMMU: Ch... |
3734 3735 3736 |
return buf - old_buf; } |
03252919b x86: print which ... |
3737 |
|
5ddd36b9c mm: implement acc... |
3738 |
/** |
ae91dbfc9 mm: fix memory.c ... |
3739 |
* access_remote_vm - access another process' address space |
5ddd36b9c mm: implement acc... |
3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 |
* @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 ... |
3753 |
/* |
206cb6365 mm: factor out ma... |
3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 |
* 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 ... |
3773 3774 3775 3776 3777 3778 3779 |
/* * 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... |
3780 3781 3782 3783 3784 3785 |
/* * Do not print if we are in atomic * contexts (in exception stacks, etc.): */ if (preempt_count()) return; |
03252919b x86: print which ... |
3786 3787 3788 3789 3790 3791 3792 |
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) { char *p, *s; |
cf28b4863 d_path: Make d_pa... |
3793 |
p = d_path(&f->f_path, buf, PAGE_SIZE); |
03252919b x86: print which ... |
3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 |
if (IS_ERR(p)) p = "?"; s = strrchr(p, '/'); if (s) p = s+1; printk("%s%s[%lx+%lx]", prefix, p, vma->vm_start, vma->vm_end - vma->vm_start); free_page((unsigned long)buf); } } up_read(¤t->mm->mmap_sem); } |
3ee1afa30 x86: some lock an... |
3807 3808 3809 3810 |
#ifdef CONFIG_PROVE_LOCKING void might_fault(void) { |
95156f005 lockdep, mm: fix ... |
3811 3812 3813 3814 3815 3816 3817 3818 |
/* * 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... |
3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 |
might_sleep(); /* * 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. */ if (!in_atomic() && current->mm) might_lock_read(¤t->mm->mmap_sem); } EXPORT_SYMBOL(might_fault); #endif |
47ad8475c thp: clear_copy_h... |
3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 |
#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 */ |