Blame view
mm/memory.c
103 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 |
/* * linux/mm/memory.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds */ /* * demand-loading started 01.12.91 - seems it is high on the list of * things wanted, and it should be easy to implement. - Linus */ /* * Ok, demand-loading was easy, shared pages a little bit tricker. Shared * pages started 02.12.91, seems to work. - Linus. * * Tested sharing by executing about 30 /bin/sh: under the old kernel it * would have taken more than the 6M I have free, but it worked well as * far as I could see. * * Also corrected some "invalidate()"s - I wasn't doing enough of them. */ /* * Real VM (paging to/from disk) started 18.12.91. Much more work and * thought has to go into this. Oh, well.. * 19.12.91 - works, somewhat. Sometimes I get faults, don't know why. * Found it. Everything seems to work now. * 20.12.91 - Ok, making the swap-device changeable like the root. */ /* * 05.04.94 - Multi-page memory management added for v1.1. * Idea by Alex Bligh (alex@cconcepts.co.uk) * * 16.07.99 - Support of BIGMEM added by Gerhard Wichert, Siemens AG * (Gerhard.Wichert@pdb.siemens.de) * * Aug/Sep 2004 Changed to four level page tables (Andi Kleen) */ #include <linux/kernel_stat.h> #include <linux/mm.h> #include <linux/hugetlb.h> #include <linux/mman.h> #include <linux/swap.h> #include <linux/highmem.h> #include <linux/pagemap.h> |
9a8408951 ksm: identify Pag... |
48 |
#include <linux/ksm.h> |
1da177e4c Linux-2.6.12-rc2 |
49 |
#include <linux/rmap.h> |
b95f1b31b mm: Map most file... |
50 |
#include <linux/export.h> |
0ff922452 [PATCH] per-task-... |
51 |
#include <linux/delayacct.h> |
1da177e4c Linux-2.6.12-rc2 |
52 |
#include <linux/init.h> |
edc79b2a4 [PATCH] mm: balan... |
53 |
#include <linux/writeback.h> |
8a9f3ccd2 Memory controller... |
54 |
#include <linux/memcontrol.h> |
cddb8a5c1 mmu-notifiers: core |
55 |
#include <linux/mmu_notifier.h> |
3dc147414 badpage: replace ... |
56 57 58 |
#include <linux/kallsyms.h> #include <linux/swapops.h> #include <linux/elf.h> |
5a0e3ad6a include cleanup: ... |
59 |
#include <linux/gfp.h> |
4daae3b4b mm: mempolicy: Us... |
60 |
#include <linux/migrate.h> |
2fbc57c53 mm: use kbasename() |
61 |
#include <linux/string.h> |
0abdd7a81 dma-debug: introd... |
62 |
#include <linux/dma-debug.h> |
1592eef01 mm: add debugfs t... |
63 |
#include <linux/debugfs.h> |
1da177e4c Linux-2.6.12-rc2 |
64 |
|
6952b61de headers: taskstat... |
65 |
#include <asm/io.h> |
1da177e4c Linux-2.6.12-rc2 |
66 67 68 69 70 |
#include <asm/pgalloc.h> #include <asm/uaccess.h> #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/pgtable.h> |
42b777281 mm: remove double... |
71 |
#include "internal.h" |
90572890d mm: numa: Change ... |
72 73 |
#ifdef LAST_CPUPID_NOT_IN_PAGE_FLAGS #warning Unfortunate NUMA and NUMA Balancing config, growing page-frame for last_cpupid. |
75980e97d mm: fold page->_l... |
74 |
#endif |
d41dee369 [PATCH] sparsemem... |
75 |
#ifndef CONFIG_NEED_MULTIPLE_NODES |
1da177e4c Linux-2.6.12-rc2 |
76 77 78 79 80 81 82 |
/* use the per-pgdat data instead for discontigmem - mbligh */ unsigned long max_mapnr; struct page *mem_map; EXPORT_SYMBOL(max_mapnr); EXPORT_SYMBOL(mem_map); #endif |
1da177e4c Linux-2.6.12-rc2 |
83 84 85 86 87 88 89 90 |
/* * A number of key systems in x86 including ioremap() rely on the assumption * that high_memory defines the upper bound on direct map memory, then end * of ZONE_NORMAL. Under CONFIG_DISCONTIG this means that max_low_pfn and * highstart_pfn must be the same; there must be no gap between ZONE_NORMAL * and ZONE_HIGHMEM. */ void * high_memory; |
1da177e4c Linux-2.6.12-rc2 |
91 |
|
1da177e4c Linux-2.6.12-rc2 |
92 |
EXPORT_SYMBOL(high_memory); |
1da177e4c Linux-2.6.12-rc2 |
93 |
|
32a932332 brk randomization... |
94 95 96 97 98 99 100 101 102 103 104 105 |
/* * Randomize the address space (stacks, mmaps, brk, etc.). * * ( When CONFIG_COMPAT_BRK=y we exclude brk from randomization, * as ancient (libc5 based) binaries can segfault. ) */ int randomize_va_space __read_mostly = #ifdef CONFIG_COMPAT_BRK 1; #else 2; #endif |
a62eaf151 [PATCH] x86_64: A... |
106 107 108 109 |
static int __init disable_randmaps(char *s) { randomize_va_space = 0; |
9b41046cd [PATCH] Don't pas... |
110 |
return 1; |
a62eaf151 [PATCH] x86_64: A... |
111 112 |
} __setup("norandmaps", disable_randmaps); |
62eede62d mm: ZERO_PAGE wit... |
113 |
unsigned long zero_pfn __read_mostly; |
03f6462a3 mm: move highest_... |
114 |
unsigned long highest_memmap_pfn __read_mostly; |
a13ea5b75 mm: reinstate ZER... |
115 |
|
0b70068e4 mm: export symbol... |
116 |
EXPORT_SYMBOL(zero_pfn); |
a13ea5b75 mm: reinstate ZER... |
117 118 119 120 121 122 123 124 125 |
/* * 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... |
126 |
|
d559db086 mm: clean up mm_c... |
127 |
|
34e55232e mm: avoid false s... |
128 |
#if defined(SPLIT_RSS_COUNTING) |
ea48cf786 mm, counters: fol... |
129 |
void sync_mm_rss(struct mm_struct *mm) |
34e55232e mm: avoid false s... |
130 131 132 133 |
{ int i; for (i = 0; i < NR_MM_COUNTERS; i++) { |
05af2e104 mm, counters: rem... |
134 135 136 |
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... |
137 138 |
} } |
05af2e104 mm, counters: rem... |
139 |
current->rss_stat.events = 0; |
34e55232e mm: avoid false s... |
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 |
} 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... |
161 |
sync_mm_rss(task->mm); |
34e55232e mm: avoid false s... |
162 |
} |
9547d01bf mm: uninline larg... |
163 |
#else /* SPLIT_RSS_COUNTING */ |
34e55232e mm: avoid false s... |
164 165 166 167 168 169 170 |
#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... |
171 172 173 174 175 176 177 178 179 180 181 182 183 |
#endif /* SPLIT_RSS_COUNTING */ #ifdef HAVE_GENERIC_MMU_GATHER static int tlb_next_batch(struct mmu_gather *tlb) { struct mmu_gather_batch *batch; batch = tlb->active; if (batch->next) { tlb->active = batch->next; return 1; } |
53a59fc67 mm: limit mmu_gat... |
184 185 |
if (tlb->batch_count == MAX_GATHER_BATCH_COUNT) return 0; |
9547d01bf mm: uninline larg... |
186 187 188 |
batch = (void *)__get_free_pages(GFP_NOWAIT | __GFP_NOWARN, 0); if (!batch) return 0; |
53a59fc67 mm: limit mmu_gat... |
189 |
tlb->batch_count++; |
9547d01bf mm: uninline larg... |
190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 |
batch->next = NULL; batch->nr = 0; batch->max = MAX_GATHER_BATCH; tlb->active->next = batch; tlb->active = batch; return 1; } /* tlb_gather_mmu * Called to initialize an (on-stack) mmu_gather structure for page-table * tear-down from @mm. The @fullmm argument is used when @mm is without * users and we're going to destroy the full address space (exit/execve). */ |
2b047252d Fix TLB gather vi... |
205 |
void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, unsigned long start, unsigned long end) |
9547d01bf mm: uninline larg... |
206 207 |
{ tlb->mm = mm; |
2b047252d Fix TLB gather vi... |
208 209 |
/* Is it from 0 to ~0? */ tlb->fullmm = !(start | (end+1)); |
1de14c3c5 x86-32: Fix possi... |
210 |
tlb->need_flush_all = 0; |
9547d01bf mm: uninline larg... |
211 212 213 214 |
tlb->local.next = NULL; tlb->local.nr = 0; tlb->local.max = ARRAY_SIZE(tlb->__pages); tlb->active = &tlb->local; |
53a59fc67 mm: limit mmu_gat... |
215 |
tlb->batch_count = 0; |
9547d01bf mm: uninline larg... |
216 217 218 219 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb->batch = NULL; #endif |
fb7332a9f mmu_gather: move ... |
220 221 |
__tlb_reset_range(tlb); |
9547d01bf mm: uninline larg... |
222 |
} |
1cf35d477 mm: split 'tlb_fl... |
223 |
static void tlb_flush_mmu_tlbonly(struct mmu_gather *tlb) |
9547d01bf mm: uninline larg... |
224 |
{ |
9547d01bf mm: uninline larg... |
225 |
tlb_flush(tlb); |
34ee645e8 mmu_notifier: cal... |
226 |
mmu_notifier_invalidate_range(tlb->mm, tlb->start, tlb->end); |
9547d01bf mm: uninline larg... |
227 228 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb_table_flush(tlb); |
34e55232e mm: avoid false s... |
229 |
#endif |
fb7332a9f mmu_gather: move ... |
230 |
__tlb_reset_range(tlb); |
1cf35d477 mm: split 'tlb_fl... |
231 232 233 234 235 |
} static void tlb_flush_mmu_free(struct mmu_gather *tlb) { struct mmu_gather_batch *batch; |
34e55232e mm: avoid false s... |
236 |
|
9547d01bf mm: uninline larg... |
237 238 239 240 241 242 |
for (batch = &tlb->local; batch; batch = batch->next) { free_pages_and_swap_cache(batch->pages, batch->nr); batch->nr = 0; } tlb->active = &tlb->local; } |
1cf35d477 mm: split 'tlb_fl... |
243 244 |
void tlb_flush_mmu(struct mmu_gather *tlb) { |
f045bbb9f mmu_gather: fix o... |
245 246 |
if (!tlb->end) return; |
1cf35d477 mm: split 'tlb_fl... |
247 248 249 |
tlb_flush_mmu_tlbonly(tlb); tlb_flush_mmu_free(tlb); } |
9547d01bf mm: uninline larg... |
250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 |
/* tlb_finish_mmu * Called at the end of the shootdown operation to free up any resources * that were required. */ void tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end) { struct mmu_gather_batch *batch, *next; tlb_flush_mmu(tlb); /* keep the page table cache within bounds */ check_pgt_cache(); for (batch = tlb->local.next; batch; batch = next) { next = batch->next; free_pages((unsigned long)batch, 0); } tlb->local.next = NULL; } /* __tlb_remove_page * Must perform the equivalent to __free_pte(pte_get_and_clear(ptep)), while * handling the additional races in SMP caused by other CPUs caching valid * mappings in their TLBs. Returns the number of free page slots left. * When out of page slots we must call tlb_flush_mmu(). */ int __tlb_remove_page(struct mmu_gather *tlb, struct page *page) { struct mmu_gather_batch *batch; |
fb7332a9f mmu_gather: move ... |
279 |
VM_BUG_ON(!tlb->end); |
9547d01bf mm: uninline larg... |
280 |
|
9547d01bf mm: uninline larg... |
281 282 283 284 285 |
batch = tlb->active; batch->pages[batch->nr++] = page; if (batch->nr == batch->max) { if (!tlb_next_batch(tlb)) return 0; |
0b43c3aab mm: __tlb_remove_... |
286 |
batch = tlb->active; |
9547d01bf mm: uninline larg... |
287 |
} |
309381fea mm: dump page whe... |
288 |
VM_BUG_ON_PAGE(batch->nr > batch->max, page); |
9547d01bf mm: uninline larg... |
289 290 291 292 293 |
return batch->max - batch->nr; } #endif /* HAVE_GENERIC_MMU_GATHER */ |
267239116 mm, powerpc: move... |
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 |
#ifdef CONFIG_HAVE_RCU_TABLE_FREE /* * See the comment near struct mmu_table_batch. */ static void tlb_remove_table_smp_sync(void *arg) { /* Simply deliver the interrupt */ } static void tlb_remove_table_one(void *table) { /* * This isn't an RCU grace period and hence the page-tables cannot be * assumed to be actually RCU-freed. * * It is however sufficient for software page-table walkers that rely on * IRQ disabling. See the comment near struct mmu_table_batch. */ smp_call_function(tlb_remove_table_smp_sync, NULL, 1); __tlb_remove_table(table); } static void tlb_remove_table_rcu(struct rcu_head *head) { struct mmu_table_batch *batch; int i; batch = container_of(head, struct mmu_table_batch, rcu); for (i = 0; i < batch->nr; i++) __tlb_remove_table(batch->tables[i]); free_page((unsigned long)batch); } void tlb_table_flush(struct mmu_gather *tlb) { struct mmu_table_batch **batch = &tlb->batch; if (*batch) { call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu); *batch = NULL; } } void tlb_remove_table(struct mmu_gather *tlb, void *table) { struct mmu_table_batch **batch = &tlb->batch; |
267239116 mm, powerpc: move... |
344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 |
/* * 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... |
365 |
#endif /* CONFIG_HAVE_RCU_TABLE_FREE */ |
267239116 mm, powerpc: move... |
366 |
|
1da177e4c Linux-2.6.12-rc2 |
367 |
/* |
1da177e4c Linux-2.6.12-rc2 |
368 369 370 |
* Note: this doesn't free the actual pages themselves. That * has been handled earlier when unmapping all the memory regions. */ |
9e1b32caa mm: Pass virtual ... |
371 372 |
static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd, unsigned long addr) |
1da177e4c Linux-2.6.12-rc2 |
373 |
{ |
2f569afd9 CONFIG_HIGHPTE vs... |
374 |
pgtable_t token = pmd_pgtable(*pmd); |
e0da382c9 [PATCH] freepgt: ... |
375 |
pmd_clear(pmd); |
9e1b32caa mm: Pass virtual ... |
376 |
pte_free_tlb(tlb, token, addr); |
e1f56c89b mm: convert mm->n... |
377 |
atomic_long_dec(&tlb->mm->nr_ptes); |
1da177e4c Linux-2.6.12-rc2 |
378 |
} |
e0da382c9 [PATCH] freepgt: ... |
379 380 381 |
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 |
382 383 384 |
{ pmd_t *pmd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
385 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
386 |
|
e0da382c9 [PATCH] freepgt: ... |
387 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
388 |
pmd = pmd_offset(pud, addr); |
1da177e4c Linux-2.6.12-rc2 |
389 390 391 392 |
do { next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; |
9e1b32caa mm: Pass virtual ... |
393 |
free_pte_range(tlb, pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
394 |
} while (pmd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
395 396 397 398 399 400 401 |
start &= PUD_MASK; if (start < floor) return; if (ceiling) { ceiling &= PUD_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
402 |
} |
e0da382c9 [PATCH] freepgt: ... |
403 404 405 406 407 |
if (end - 1 > ceiling - 1) return; pmd = pmd_offset(pud, start); pud_clear(pud); |
9e1b32caa mm: Pass virtual ... |
408 |
pmd_free_tlb(tlb, pmd, start); |
1da177e4c Linux-2.6.12-rc2 |
409 |
} |
e0da382c9 [PATCH] freepgt: ... |
410 411 412 |
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 |
413 414 415 |
{ pud_t *pud; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
416 |
unsigned long start; |
1da177e4c Linux-2.6.12-rc2 |
417 |
|
e0da382c9 [PATCH] freepgt: ... |
418 |
start = addr; |
1da177e4c Linux-2.6.12-rc2 |
419 |
pud = pud_offset(pgd, addr); |
1da177e4c Linux-2.6.12-rc2 |
420 421 422 423 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
e0da382c9 [PATCH] freepgt: ... |
424 |
free_pmd_range(tlb, pud, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
425 |
} while (pud++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
426 427 428 429 430 431 432 |
start &= PGDIR_MASK; if (start < floor) return; if (ceiling) { ceiling &= PGDIR_MASK; if (!ceiling) return; |
1da177e4c Linux-2.6.12-rc2 |
433 |
} |
e0da382c9 [PATCH] freepgt: ... |
434 435 436 437 438 |
if (end - 1 > ceiling - 1) return; pud = pud_offset(pgd, start); pgd_clear(pgd); |
9e1b32caa mm: Pass virtual ... |
439 |
pud_free_tlb(tlb, pud, start); |
1da177e4c Linux-2.6.12-rc2 |
440 441 442 |
} /* |
e0da382c9 [PATCH] freepgt: ... |
443 |
* This function frees user-level page tables of a process. |
1da177e4c Linux-2.6.12-rc2 |
444 |
*/ |
42b777281 mm: remove double... |
445 |
void free_pgd_range(struct mmu_gather *tlb, |
e0da382c9 [PATCH] freepgt: ... |
446 447 |
unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) |
1da177e4c Linux-2.6.12-rc2 |
448 449 450 |
{ pgd_t *pgd; unsigned long next; |
e0da382c9 [PATCH] freepgt: ... |
451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 |
/* * 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 |
477 |
|
e0da382c9 [PATCH] freepgt: ... |
478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 |
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... |
493 |
pgd = pgd_offset(tlb->mm, addr); |
1da177e4c Linux-2.6.12-rc2 |
494 495 496 497 |
do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
42b777281 mm: remove double... |
498 |
free_pud_range(tlb, pgd, addr, next, floor, ceiling); |
1da177e4c Linux-2.6.12-rc2 |
499 |
} while (pgd++, addr = next, addr != end); |
e0da382c9 [PATCH] freepgt: ... |
500 |
} |
42b777281 mm: remove double... |
501 |
void free_pgtables(struct mmu_gather *tlb, struct vm_area_struct *vma, |
3bf5ee956 [PATCH] freepgt: ... |
502 |
unsigned long floor, unsigned long ceiling) |
e0da382c9 [PATCH] freepgt: ... |
503 504 505 506 |
{ while (vma) { struct vm_area_struct *next = vma->vm_next; unsigned long addr = vma->vm_start; |
8f4f8c164 [PATCH] mm: unlin... |
507 |
/* |
25d9e2d15 truncate: new hel... |
508 509 |
* Hide vma from rmap and truncate_pagecache before freeing * pgtables |
8f4f8c164 [PATCH] mm: unlin... |
510 |
*/ |
5beb49305 mm: change anon_v... |
511 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
512 |
unlink_file_vma(vma); |
9da61aef0 [PATCH] hugepage:... |
513 |
if (is_vm_hugetlb_page(vma)) { |
3bf5ee956 [PATCH] freepgt: ... |
514 |
hugetlb_free_pgd_range(tlb, addr, vma->vm_end, |
e0da382c9 [PATCH] freepgt: ... |
515 |
floor, next? next->vm_start: ceiling); |
3bf5ee956 [PATCH] freepgt: ... |
516 517 518 519 520 |
} else { /* * Optimization: gather nearby vmas into one call down */ while (next && next->vm_start <= vma->vm_end + PMD_SIZE |
4866920b9 [PATCH] hugepage:... |
521 |
&& !is_vm_hugetlb_page(next)) { |
3bf5ee956 [PATCH] freepgt: ... |
522 523 |
vma = next; next = vma->vm_next; |
5beb49305 mm: change anon_v... |
524 |
unlink_anon_vmas(vma); |
8f4f8c164 [PATCH] mm: unlin... |
525 |
unlink_file_vma(vma); |
3bf5ee956 [PATCH] freepgt: ... |
526 527 528 529 |
} free_pgd_range(tlb, addr, vma->vm_end, floor, next? next->vm_start: ceiling); } |
e0da382c9 [PATCH] freepgt: ... |
530 531 |
vma = next; } |
1da177e4c Linux-2.6.12-rc2 |
532 |
} |
8ac1f8320 thp: pte alloc tr... |
533 534 |
int __pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma, pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
535 |
{ |
c4088ebdc mm: convert the r... |
536 |
spinlock_t *ptl; |
2f569afd9 CONFIG_HIGHPTE vs... |
537 |
pgtable_t new = pte_alloc_one(mm, address); |
8ac1f8320 thp: pte alloc tr... |
538 |
int wait_split_huge_page; |
1bb3630e8 [PATCH] mm: ptd_a... |
539 540 |
if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
541 542 543 544 545 546 547 548 549 550 551 552 553 554 |
/* * Ensure all pte setup (eg. pte page lock and page clearing) are * visible before the pte is made visible to other CPUs by being * put into page tables. * * The other side of the story is the pointer chasing in the page * table walking code (when walking the page table without locking; * ie. most of the time). Fortunately, these data accesses consist * of a chain of data-dependent loads, meaning most CPUs (alpha * being the notable exception) will already guarantee loads are * seen in-order. See the alpha page table accessors for the * smp_read_barrier_depends() barriers in page table walking code. */ smp_wmb(); /* Could be smp_wmb__xxx(before|after)_spin_lock */ |
c4088ebdc mm: convert the r... |
555 |
ptl = pmd_lock(mm, pmd); |
8ac1f8320 thp: pte alloc tr... |
556 557 |
wait_split_huge_page = 0; if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
e1f56c89b mm: convert mm->n... |
558 |
atomic_long_inc(&mm->nr_ptes); |
1da177e4c Linux-2.6.12-rc2 |
559 |
pmd_populate(mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
560 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
561 562 |
} else if (unlikely(pmd_trans_splitting(*pmd))) wait_split_huge_page = 1; |
c4088ebdc mm: convert the r... |
563 |
spin_unlock(ptl); |
2f569afd9 CONFIG_HIGHPTE vs... |
564 565 |
if (new) pte_free(mm, new); |
8ac1f8320 thp: pte alloc tr... |
566 567 |
if (wait_split_huge_page) wait_split_huge_page(vma->anon_vma, pmd); |
1bb3630e8 [PATCH] mm: ptd_a... |
568 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
569 |
} |
1bb3630e8 [PATCH] mm: ptd_a... |
570 |
int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
571 |
{ |
1bb3630e8 [PATCH] mm: ptd_a... |
572 573 574 |
pte_t *new = pte_alloc_one_kernel(&init_mm, address); if (!new) return -ENOMEM; |
362a61ad6 fix SMP data race... |
575 |
smp_wmb(); /* See comment in __pte_alloc */ |
1bb3630e8 [PATCH] mm: ptd_a... |
576 |
spin_lock(&init_mm.page_table_lock); |
8ac1f8320 thp: pte alloc tr... |
577 |
if (likely(pmd_none(*pmd))) { /* Has another populated it ? */ |
1bb3630e8 [PATCH] mm: ptd_a... |
578 |
pmd_populate_kernel(&init_mm, pmd, new); |
2f569afd9 CONFIG_HIGHPTE vs... |
579 |
new = NULL; |
8ac1f8320 thp: pte alloc tr... |
580 581 |
} else VM_BUG_ON(pmd_trans_splitting(*pmd)); |
1bb3630e8 [PATCH] mm: ptd_a... |
582 |
spin_unlock(&init_mm.page_table_lock); |
2f569afd9 CONFIG_HIGHPTE vs... |
583 584 |
if (new) pte_free_kernel(&init_mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
585 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
586 |
} |
d559db086 mm: clean up mm_c... |
587 588 589 590 591 592 |
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... |
593 |
{ |
d559db086 mm: clean up mm_c... |
594 |
int i; |
34e55232e mm: avoid false s... |
595 |
if (current->mm == mm) |
05af2e104 mm, counters: rem... |
596 |
sync_mm_rss(mm); |
d559db086 mm: clean up mm_c... |
597 598 599 |
for (i = 0; i < NR_MM_COUNTERS; i++) if (rss[i]) add_mm_counter(mm, i, rss[i]); |
ae8597623 [PATCH] mm: batch... |
600 |
} |
1da177e4c Linux-2.6.12-rc2 |
601 |
/* |
6aab341e0 mm: re-architect ... |
602 603 604 |
* 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... |
605 606 607 |
* * The calling function must still handle the error. */ |
3dc147414 badpage: replace ... |
608 609 |
static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr, pte_t pte, struct page *page) |
b5810039a [PATCH] core remo... |
610 |
{ |
3dc147414 badpage: replace ... |
611 612 613 614 615 |
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... |
616 617 618 619 620 621 622 623 624 625 626 627 628 629 |
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... |
630 631 632 |
printk(KERN_ALERT "BUG: Bad page map: %lu messages suppressed ", |
d936cf9b3 badpage: ratelimi... |
633 634 635 636 637 638 639 |
nr_unshown); nr_unshown = 0; } nr_shown = 0; } if (nr_shown++ == 0) resume = jiffies + 60 * HZ; |
3dc147414 badpage: replace ... |
640 641 642 |
mapping = vma->vm_file ? vma->vm_file->f_mapping : NULL; index = linear_page_index(vma, addr); |
1e9e63650 badpage: KERN_ALE... |
643 644 645 |
printk(KERN_ALERT "BUG: Bad page map in process %s pte:%08llx pmd:%08llx ", |
3dc147414 badpage: replace ... |
646 647 |
current->comm, (long long)pte_val(pte), (long long)pmd_val(*pmd)); |
718a38211 mm: introduce dum... |
648 |
if (page) |
f0b791a34 mm: print more de... |
649 |
dump_page(page, "bad pte"); |
1e9e63650 badpage: KERN_ALE... |
650 |
printk(KERN_ALERT |
3dc147414 badpage: replace ... |
651 652 653 654 655 656 657 |
"addr:%p vm_flags:%08lx anon_vma:%p mapping:%p index:%lx ", (void *)addr, vma->vm_flags, vma->anon_vma, mapping, index); /* * Choose text because data symbols depend on CONFIG_KALLSYMS_ALL=y */ if (vma->vm_ops) |
071361d34 mm: Convert print... |
658 659 660 |
printk(KERN_ALERT "vma->vm_ops->fault: %pSR ", vma->vm_ops->fault); |
72c2d5319 file->f_op is nev... |
661 |
if (vma->vm_file) |
071361d34 mm: Convert print... |
662 663 664 |
printk(KERN_ALERT "vma->vm_file->f_op->mmap: %pSR ", vma->vm_file->f_op->mmap); |
b5810039a [PATCH] core remo... |
665 |
dump_stack(); |
373d4d099 taint: add explic... |
666 |
add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); |
b5810039a [PATCH] core remo... |
667 668 669 |
} /* |
7e675137a mm: introduce pte... |
670 |
* vm_normal_page -- This function gets the "struct page" associated with a pte. |
6aab341e0 mm: re-architect ... |
671 |
* |
7e675137a mm: introduce pte... |
672 673 674 |
* "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_... |
675 |
* |
7e675137a mm: introduce pte... |
676 677 678 679 680 681 682 683 |
* 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 ... |
684 |
* |
b379d7901 mm: introduce VM_... |
685 686 |
* 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... |
687 688 |
* 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 ... |
689 690 691 |
* * pfn_of_page == vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT) * |
7e675137a mm: introduce pte... |
692 693 694 695 696 697 |
* 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_... |
698 |
* |
b379d7901 mm: introduce VM_... |
699 |
* |
7e675137a mm: introduce pte... |
700 |
* In order to support COW of arbitrary special mappings, we have VM_MIXEDMAP. |
b379d7901 mm: introduce VM_... |
701 702 703 704 705 706 707 708 709 |
* * 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: ... |
710 |
*/ |
7e675137a mm: introduce pte... |
711 712 713 714 715 716 717 |
#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: ... |
718 |
{ |
22b31eec6 badpage: vm_norma... |
719 |
unsigned long pfn = pte_pfn(pte); |
7e675137a mm: introduce pte... |
720 721 |
if (HAVE_PTE_SPECIAL) { |
b38af4721 x86,mm: fix pte_s... |
722 |
if (likely(!pte_special(pte))) |
22b31eec6 badpage: vm_norma... |
723 |
goto check_pfn; |
a13ea5b75 mm: reinstate ZER... |
724 725 |
if (vma->vm_flags & (VM_PFNMAP | VM_MIXEDMAP)) return NULL; |
62eede62d mm: ZERO_PAGE wit... |
726 |
if (!is_zero_pfn(pfn)) |
22b31eec6 badpage: vm_norma... |
727 |
print_bad_pte(vma, addr, pte, NULL); |
7e675137a mm: introduce pte... |
728 729 730 731 |
return NULL; } /* !HAVE_PTE_SPECIAL case follows: */ |
b379d7901 mm: introduce VM_... |
732 733 734 735 736 737 |
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... |
738 739 |
unsigned long off; off = (addr - vma->vm_start) >> PAGE_SHIFT; |
b379d7901 mm: introduce VM_... |
740 741 742 743 744 |
if (pfn == vma->vm_pgoff + off) return NULL; if (!is_cow_mapping(vma->vm_flags)) return NULL; } |
6aab341e0 mm: re-architect ... |
745 |
} |
b38af4721 x86,mm: fix pte_s... |
746 747 |
if (is_zero_pfn(pfn)) return NULL; |
22b31eec6 badpage: vm_norma... |
748 749 750 751 752 |
check_pfn: if (unlikely(pfn > highest_memmap_pfn)) { print_bad_pte(vma, addr, pte, NULL); return NULL; } |
6aab341e0 mm: re-architect ... |
753 754 |
/* |
7e675137a mm: introduce pte... |
755 |
* NOTE! We still have PageReserved() pages in the page tables. |
7e675137a mm: introduce pte... |
756 |
* eg. VDSO mappings can cause them to exist. |
6aab341e0 mm: re-architect ... |
757 |
*/ |
b379d7901 mm: introduce VM_... |
758 |
out: |
6aab341e0 mm: re-architect ... |
759 |
return pfn_to_page(pfn); |
ee498ed73 [PATCH] unpaged: ... |
760 761 762 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
763 764 765 |
* 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 |
766 |
*/ |
570a335b8 swap_info: swap c... |
767 |
static inline unsigned long |
1da177e4c Linux-2.6.12-rc2 |
768 |
copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
b5810039a [PATCH] core remo... |
769 |
pte_t *dst_pte, pte_t *src_pte, struct vm_area_struct *vma, |
8c1037627 [PATCH] mm: copy_... |
770 |
unsigned long addr, int *rss) |
1da177e4c Linux-2.6.12-rc2 |
771 |
{ |
b5810039a [PATCH] core remo... |
772 |
unsigned long vm_flags = vma->vm_flags; |
1da177e4c Linux-2.6.12-rc2 |
773 774 |
pte_t pte = *src_pte; struct page *page; |
1da177e4c Linux-2.6.12-rc2 |
775 776 777 778 |
/* pte contains position in swap or file, so copy. */ if (unlikely(!pte_present(pte))) { if (!pte_file(pte)) { |
0697212a4 [PATCH] Swapless ... |
779 |
swp_entry_t entry = pte_to_swp_entry(pte); |
2022b4d18 mm: fix swapoff h... |
780 781 782 783 784 785 786 787 788 789 790 791 |
if (likely(!non_swap_entry(entry))) { if (swap_duplicate(entry) < 0) return entry.val; /* make sure dst_mm is on swapoff's mmlist. */ if (unlikely(list_empty(&dst_mm->mmlist))) { spin_lock(&mmlist_lock); if (list_empty(&dst_mm->mmlist)) list_add(&dst_mm->mmlist, &src_mm->mmlist); spin_unlock(&mmlist_lock); } |
b084d4353 mm: count swap usage |
792 |
rss[MM_SWAPENTS]++; |
2022b4d18 mm: fix swapoff h... |
793 |
} else if (is_migration_entry(entry)) { |
9f9f1acd7 mm: fix rss count... |
794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 |
page = migration_entry_to_page(entry); if (PageAnon(page)) rss[MM_ANONPAGES]++; else rss[MM_FILEPAGES]++; if (is_write_migration_entry(entry) && is_cow_mapping(vm_flags)) { /* * COW mappings require pages in both * parent and child to be set to read. */ make_migration_entry_read(&entry); pte = swp_entry_to_pte(entry); |
c3d16e165 mm: migration: do... |
809 810 |
if (pte_swp_soft_dirty(*src_pte)) pte = pte_swp_mksoft_dirty(pte); |
9f9f1acd7 mm: fix rss count... |
811 812 |
set_pte_at(src_mm, addr, src_pte, pte); } |
0697212a4 [PATCH] Swapless ... |
813 |
} |
1da177e4c Linux-2.6.12-rc2 |
814 |
} |
ae8597623 [PATCH] mm: batch... |
815 |
goto out_set_pte; |
1da177e4c Linux-2.6.12-rc2 |
816 |
} |
1da177e4c Linux-2.6.12-rc2 |
817 818 819 820 |
/* * If it's a COW mapping, write protect it both * in the parent and the child */ |
67121172f Allow arbitrary r... |
821 |
if (is_cow_mapping(vm_flags)) { |
1da177e4c Linux-2.6.12-rc2 |
822 |
ptep_set_wrprotect(src_mm, addr, src_pte); |
3dc907951 [PATCH] paravirt:... |
823 |
pte = pte_wrprotect(pte); |
1da177e4c Linux-2.6.12-rc2 |
824 825 826 827 828 829 830 831 832 |
} /* * 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 ... |
833 834 835 836 |
page = vm_normal_page(vma, addr, pte); if (page) { get_page(page); |
21333b2b6 ksm: no debug in ... |
837 |
page_dup_rmap(page); |
d559db086 mm: clean up mm_c... |
838 839 840 841 |
if (PageAnon(page)) rss[MM_ANONPAGES]++; else rss[MM_FILEPAGES]++; |
6aab341e0 mm: re-architect ... |
842 |
} |
ae8597623 [PATCH] mm: batch... |
843 844 845 |
out_set_pte: set_pte_at(dst_mm, addr, dst_pte, pte); |
570a335b8 swap_info: swap c... |
846 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
847 |
} |
21bda264f mm: make copy_pte... |
848 |
static int copy_pte_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, |
71e3aac07 thp: transparent ... |
849 850 |
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 |
851 |
{ |
c36987e2e mm: don't call pt... |
852 |
pte_t *orig_src_pte, *orig_dst_pte; |
1da177e4c Linux-2.6.12-rc2 |
853 |
pte_t *src_pte, *dst_pte; |
c74df32c7 [PATCH] mm: ptd_a... |
854 |
spinlock_t *src_ptl, *dst_ptl; |
e040f218b [PATCH] mm: copy_... |
855 |
int progress = 0; |
d559db086 mm: clean up mm_c... |
856 |
int rss[NR_MM_COUNTERS]; |
570a335b8 swap_info: swap c... |
857 |
swp_entry_t entry = (swp_entry_t){0}; |
1da177e4c Linux-2.6.12-rc2 |
858 859 |
again: |
d559db086 mm: clean up mm_c... |
860 |
init_rss_vec(rss); |
c74df32c7 [PATCH] mm: ptd_a... |
861 |
dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl); |
1da177e4c Linux-2.6.12-rc2 |
862 863 |
if (!dst_pte) return -ENOMEM; |
ece0e2b64 mm: remove pte_*m... |
864 |
src_pte = pte_offset_map(src_pmd, addr); |
4c21e2f24 [PATCH] mm: split... |
865 |
src_ptl = pte_lockptr(src_mm, src_pmd); |
f20dc5f7c [PATCH] lockdep: ... |
866 |
spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING); |
c36987e2e mm: don't call pt... |
867 868 |
orig_src_pte = src_pte; orig_dst_pte = dst_pte; |
6606c3e0d [PATCH] paravirt:... |
869 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
870 |
|
1da177e4c Linux-2.6.12-rc2 |
871 872 873 874 875 |
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_... |
876 877 878 |
if (progress >= 32) { progress = 0; if (need_resched() || |
95c354fe9 spinlock: lockbre... |
879 |
spin_needbreak(src_ptl) || spin_needbreak(dst_ptl)) |
e040f218b [PATCH] mm: copy_... |
880 881 |
break; } |
1da177e4c Linux-2.6.12-rc2 |
882 883 884 885 |
if (pte_none(*src_pte)) { progress++; continue; } |
570a335b8 swap_info: swap c... |
886 887 888 889 |
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 |
890 891 |
progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
892 |
|
6606c3e0d [PATCH] paravirt:... |
893 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
894 |
spin_unlock(src_ptl); |
ece0e2b64 mm: remove pte_*m... |
895 |
pte_unmap(orig_src_pte); |
d559db086 mm: clean up mm_c... |
896 |
add_mm_rss_vec(dst_mm, rss); |
c36987e2e mm: don't call pt... |
897 |
pte_unmap_unlock(orig_dst_pte, dst_ptl); |
c74df32c7 [PATCH] mm: ptd_a... |
898 |
cond_resched(); |
570a335b8 swap_info: swap c... |
899 900 901 902 903 904 |
if (entry.val) { if (add_swap_count_continuation(entry, GFP_KERNEL) < 0) return -ENOMEM; progress = 0; } |
1da177e4c Linux-2.6.12-rc2 |
905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 |
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 ... |
923 924 |
if (pmd_trans_huge(*src_pmd)) { int err; |
14d1a55cd thp: add debug ch... |
925 |
VM_BUG_ON(next-addr != HPAGE_PMD_SIZE); |
71e3aac07 thp: transparent ... |
926 927 928 929 930 931 932 933 |
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 |
934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 |
if (pmd_none_or_clear_bad(src_pmd)) continue; if (copy_pte_range(dst_mm, src_mm, dst_pmd, src_pmd, vma, addr, next)) return -ENOMEM; } while (dst_pmd++, src_pmd++, addr = next, addr != end); return 0; } static inline int copy_pud_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, pgd_t *dst_pgd, pgd_t *src_pgd, struct vm_area_struct *vma, unsigned long addr, unsigned long end) { pud_t *src_pud, *dst_pud; unsigned long next; dst_pud = pud_alloc(dst_mm, dst_pgd, addr); if (!dst_pud) return -ENOMEM; src_pud = pud_offset(src_pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(src_pud)) continue; if (copy_pmd_range(dst_mm, src_mm, dst_pud, src_pud, vma, addr, next)) return -ENOMEM; } while (dst_pud++, src_pud++, addr = next, addr != end); return 0; } int copy_page_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, struct vm_area_struct *vma) { pgd_t *src_pgd, *dst_pgd; unsigned long next; unsigned long addr = vma->vm_start; unsigned long end = vma->vm_end; |
2ec74c3ef mm: move all mmu ... |
972 973 974 |
unsigned long mmun_start; /* For mmu_notifiers */ unsigned long mmun_end; /* For mmu_notifiers */ bool is_cow; |
cddb8a5c1 mmu-notifiers: core |
975 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
976 |
|
d992895ba [PATCH] Lazy page... |
977 978 979 980 981 982 |
/* * Don't copy ptes where a page fault will fill them correctly. * Fork becomes much lighter when there are big shared or private * readonly mappings. The tradeoff is that copy_page_range is more * efficient than faulting. */ |
4b6e1e370 mm: kill vma flag... |
983 984 |
if (!(vma->vm_flags & (VM_HUGETLB | VM_NONLINEAR | VM_PFNMAP | VM_MIXEDMAP))) { |
d992895ba [PATCH] Lazy page... |
985 986 987 |
if (!vma->anon_vma) return 0; } |
1da177e4c Linux-2.6.12-rc2 |
988 989 |
if (is_vm_hugetlb_page(vma)) return copy_hugetlb_page_range(dst_mm, src_mm, vma); |
b3b9c2932 mm, x86, pat: rew... |
990 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) { |
2ab640379 x86: PAT: hooks i... |
991 992 993 994 |
/* * We do not free on error cases below as remove_vma * gets called on error from higher level routine */ |
5180da410 x86, pat: separat... |
995 |
ret = track_pfn_copy(vma); |
2ab640379 x86: PAT: hooks i... |
996 997 998 |
if (ret) return ret; } |
cddb8a5c1 mmu-notifiers: core |
999 1000 1001 1002 1003 1004 |
/* * We need to invalidate the secondary MMU mappings only when * there could be a permission downgrade on the ptes of the * parent mm. And a permission downgrade will only happen if * is_cow_mapping() returns true. */ |
2ec74c3ef mm: move all mmu ... |
1005 1006 1007 1008 1009 1010 |
is_cow = is_cow_mapping(vma->vm_flags); mmun_start = addr; mmun_end = end; if (is_cow) mmu_notifier_invalidate_range_start(src_mm, mmun_start, mmun_end); |
cddb8a5c1 mmu-notifiers: core |
1011 1012 |
ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
1013 1014 1015 1016 1017 1018 |
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 |
1019 1020 1021 1022 1023 |
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 |
1024 |
} while (dst_pgd++, src_pgd++, addr = next, addr != end); |
cddb8a5c1 mmu-notifiers: core |
1025 |
|
2ec74c3ef mm: move all mmu ... |
1026 1027 |
if (is_cow) mmu_notifier_invalidate_range_end(src_mm, mmun_start, mmun_end); |
cddb8a5c1 mmu-notifiers: core |
1028 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1029 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1030 |
static unsigned long zap_pte_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1031 |
struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c Linux-2.6.12-rc2 |
1032 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1033 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1034 |
{ |
b5810039a [PATCH] core remo... |
1035 |
struct mm_struct *mm = tlb->mm; |
d16dfc550 mm: mmu_gather re... |
1036 |
int force_flush = 0; |
d559db086 mm: clean up mm_c... |
1037 |
int rss[NR_MM_COUNTERS]; |
97a894136 mm: Remove i_mmap... |
1038 |
spinlock_t *ptl; |
5f1a19070 mm: fix wrong kun... |
1039 |
pte_t *start_pte; |
97a894136 mm: Remove i_mmap... |
1040 |
pte_t *pte; |
d559db086 mm: clean up mm_c... |
1041 |
|
d16dfc550 mm: mmu_gather re... |
1042 |
again: |
e303297e6 mm: extended batc... |
1043 |
init_rss_vec(rss); |
5f1a19070 mm: fix wrong kun... |
1044 1045 |
start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl); pte = start_pte; |
6606c3e0d [PATCH] paravirt:... |
1046 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
1047 1048 |
do { pte_t ptent = *pte; |
51c6f666f [PATCH] mm: ZAP_B... |
1049 |
if (pte_none(ptent)) { |
1da177e4c Linux-2.6.12-rc2 |
1050 |
continue; |
51c6f666f [PATCH] mm: ZAP_B... |
1051 |
} |
6f5e6b9e6 [PATCH] fix free ... |
1052 |
|
1da177e4c Linux-2.6.12-rc2 |
1053 |
if (pte_present(ptent)) { |
ee498ed73 [PATCH] unpaged: ... |
1054 |
struct page *page; |
51c6f666f [PATCH] mm: ZAP_B... |
1055 |
|
6aab341e0 mm: re-architect ... |
1056 |
page = vm_normal_page(vma, addr, ptent); |
1da177e4c Linux-2.6.12-rc2 |
1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 |
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... |
1075 |
ptent = ptep_get_and_clear_full(mm, addr, pte, |
a600388d2 [PATCH] x86: ptep... |
1076 |
tlb->fullmm); |
1da177e4c Linux-2.6.12-rc2 |
1077 1078 1079 1080 1081 |
tlb_remove_tlb_entry(tlb, pte, addr); if (unlikely(!page)) continue; if (unlikely(details) && details->nonlinear_vma && linear_page_index(details->nonlinear_vma, |
41bb3476b mm: save soft-dir... |
1082 1083 1084 |
addr) != page->index) { pte_t ptfile = pgoff_to_pte(page->index); if (pte_soft_dirty(ptent)) |
dbab31aa2 mm: softdirty: ke... |
1085 |
ptfile = pte_file_mksoft_dirty(ptfile); |
41bb3476b mm: save soft-dir... |
1086 1087 |
set_pte_at(mm, addr, pte, ptfile); } |
1da177e4c Linux-2.6.12-rc2 |
1088 |
if (PageAnon(page)) |
d559db086 mm: clean up mm_c... |
1089 |
rss[MM_ANONPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1090 |
else { |
1cf35d477 mm: split 'tlb_fl... |
1091 1092 |
if (pte_dirty(ptent)) { force_flush = 1; |
6237bcd94 [PATCH] mm: zap_p... |
1093 |
set_page_dirty(page); |
1cf35d477 mm: split 'tlb_fl... |
1094 |
} |
4917e5d04 mm: more likely r... |
1095 |
if (pte_young(ptent) && |
64363aad5 mm: remove unused... |
1096 |
likely(!(vma->vm_flags & VM_SEQ_READ))) |
bf3f3bc5e mm: don't mark_pa... |
1097 |
mark_page_accessed(page); |
d559db086 mm: clean up mm_c... |
1098 |
rss[MM_FILEPAGES]--; |
6237bcd94 [PATCH] mm: zap_p... |
1099 |
} |
edc315fd2 badpage: remove v... |
1100 |
page_remove_rmap(page); |
3dc147414 badpage: replace ... |
1101 1102 |
if (unlikely(page_mapcount(page) < 0)) print_bad_pte(vma, addr, ptent, page); |
1cf35d477 mm: split 'tlb_fl... |
1103 1104 |
if (unlikely(!__tlb_remove_page(tlb, page))) { force_flush = 1; |
ce9ec37bd zap_pte_range: up... |
1105 |
addr += PAGE_SIZE; |
d16dfc550 mm: mmu_gather re... |
1106 |
break; |
1cf35d477 mm: split 'tlb_fl... |
1107 |
} |
1da177e4c Linux-2.6.12-rc2 |
1108 1109 1110 1111 1112 1113 1114 1115 |
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... |
1116 1117 1118 |
if (pte_file(ptent)) { if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) print_bad_pte(vma, addr, ptent, NULL); |
b084d4353 mm: count swap usage |
1119 1120 1121 1122 1123 |
} else { swp_entry_t entry = pte_to_swp_entry(ptent); if (!non_swap_entry(entry)) rss[MM_SWAPENTS]--; |
9f9f1acd7 mm: fix rss count... |
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 |
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 |
1134 1135 1136 |
if (unlikely(!free_swap_and_cache(entry))) print_bad_pte(vma, addr, ptent, NULL); } |
9888a1cae [PATCH] paravirt:... |
1137 |
pte_clear_not_present_full(mm, addr, pte, tlb->fullmm); |
97a894136 mm: Remove i_mmap... |
1138 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
ae8597623 [PATCH] mm: batch... |
1139 |
|
d559db086 mm: clean up mm_c... |
1140 |
add_mm_rss_vec(mm, rss); |
6606c3e0d [PATCH] paravirt:... |
1141 |
arch_leave_lazy_mmu_mode(); |
51c6f666f [PATCH] mm: ZAP_B... |
1142 |
|
1cf35d477 mm: split 'tlb_fl... |
1143 |
/* Do the actual TLB flush before dropping ptl */ |
fb7332a9f mmu_gather: move ... |
1144 |
if (force_flush) |
1cf35d477 mm: split 'tlb_fl... |
1145 |
tlb_flush_mmu_tlbonly(tlb); |
1cf35d477 mm: split 'tlb_fl... |
1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 |
pte_unmap_unlock(start_pte, ptl); /* * If we forced a TLB flush (either due to running out of * batch buffers or because we needed to flush dirty TLB * entries before releasing the ptl), free the batched * memory too. Restart if we didn't do everything. */ if (force_flush) { force_flush = 0; tlb_flush_mmu_free(tlb); |
2b047252d Fix TLB gather vi... |
1157 1158 |
if (addr != end) |
d16dfc550 mm: mmu_gather re... |
1159 1160 |
goto again; } |
51c6f666f [PATCH] mm: ZAP_B... |
1161 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1162 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1163 |
static inline unsigned long zap_pmd_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1164 |
struct vm_area_struct *vma, pud_t *pud, |
1da177e4c Linux-2.6.12-rc2 |
1165 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1166 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1167 1168 1169 1170 1171 1172 1173 |
{ pmd_t *pmd; unsigned long next; pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
71e3aac07 thp: transparent ... |
1174 |
if (pmd_trans_huge(*pmd)) { |
1a5a9906d mm: thp: fix pmd_... |
1175 |
if (next - addr != HPAGE_PMD_SIZE) { |
e0897d75f mm, thp: print us... |
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 |
#ifdef CONFIG_DEBUG_VM if (!rwsem_is_locked(&tlb->mm->mmap_sem)) { pr_err("%s: mmap_sem is unlocked! addr=0x%lx end=0x%lx vma->vm_start=0x%lx vma->vm_end=0x%lx ", __func__, addr, end, vma->vm_start, vma->vm_end); BUG(); } #endif |
e180377f1 thp: change split... |
1186 |
split_huge_page_pmd(vma, addr, pmd); |
f21760b15 thp: add tlb_remo... |
1187 |
} else if (zap_huge_pmd(tlb, vma, pmd, addr)) |
1a5a9906d mm: thp: fix pmd_... |
1188 |
goto next; |
71e3aac07 thp: transparent ... |
1189 1190 |
/* fall through */ } |
1a5a9906d mm: thp: fix pmd_... |
1191 1192 1193 1194 1195 1196 1197 1198 1199 |
/* * 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... |
1200 |
next = zap_pte_range(tlb, vma, pmd, addr, next, details); |
1a5a9906d mm: thp: fix pmd_... |
1201 |
next: |
97a894136 mm: Remove i_mmap... |
1202 1203 |
cond_resched(); } while (pmd++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1204 1205 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1206 |
} |
51c6f666f [PATCH] mm: ZAP_B... |
1207 |
static inline unsigned long zap_pud_range(struct mmu_gather *tlb, |
b5810039a [PATCH] core remo... |
1208 |
struct vm_area_struct *vma, pgd_t *pgd, |
1da177e4c Linux-2.6.12-rc2 |
1209 |
unsigned long addr, unsigned long end, |
97a894136 mm: Remove i_mmap... |
1210 |
struct zap_details *details) |
1da177e4c Linux-2.6.12-rc2 |
1211 1212 1213 1214 1215 1216 1217 |
{ pud_t *pud; unsigned long next; pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1218 |
if (pud_none_or_clear_bad(pud)) |
1da177e4c Linux-2.6.12-rc2 |
1219 |
continue; |
97a894136 mm: Remove i_mmap... |
1220 1221 |
next = zap_pmd_range(tlb, vma, pud, addr, next, details); } while (pud++, addr = next, addr != end); |
51c6f666f [PATCH] mm: ZAP_B... |
1222 1223 |
return addr; |
1da177e4c Linux-2.6.12-rc2 |
1224 |
} |
038c7aa16 VM: unmap_page_ra... |
1225 1226 1227 1228 |
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 |
1229 1230 1231 1232 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); tlb_start_vma(tlb, vma); pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); |
97a894136 mm: Remove i_mmap... |
1241 |
if (pgd_none_or_clear_bad(pgd)) |
1da177e4c Linux-2.6.12-rc2 |
1242 |
continue; |
97a894136 mm: Remove i_mmap... |
1243 1244 |
next = zap_pud_range(tlb, vma, pgd, addr, next, details); } while (pgd++, addr = next, addr != end); |
1da177e4c Linux-2.6.12-rc2 |
1245 1246 |
tlb_end_vma(tlb, vma); } |
51c6f666f [PATCH] mm: ZAP_B... |
1247 |
|
f5cc4eef9 VM: make zap_page... |
1248 1249 1250 |
static void unmap_single_vma(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e vm: remove 'nr_ac... |
1251 |
unsigned long end_addr, |
f5cc4eef9 VM: make zap_page... |
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 |
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... |
1262 1263 |
if (vma->vm_file) uprobe_munmap(vma, start, end); |
b3b9c2932 mm, x86, pat: rew... |
1264 |
if (unlikely(vma->vm_flags & VM_PFNMAP)) |
5180da410 x86, pat: separat... |
1265 |
untrack_pfn(vma, 0, 0); |
f5cc4eef9 VM: make zap_page... |
1266 1267 1268 1269 1270 1271 1272 |
if (start != end) { if (unlikely(is_vm_hugetlb_page(vma))) { /* * It is undesirable to test vma->vm_file as it * should be non-null for valid hugetlb area. * However, vm_file will be NULL in the error |
7aa6b4ad5 mm/memory.c: upda... |
1273 |
* cleanup path of mmap_region. When |
f5cc4eef9 VM: make zap_page... |
1274 |
* hugetlbfs ->mmap method fails, |
7aa6b4ad5 mm/memory.c: upda... |
1275 |
* mmap_region() nullifies vma->vm_file |
f5cc4eef9 VM: make zap_page... |
1276 1277 1278 1279 |
* before calling this function to clean up. * Since no pte has actually been setup, it is * safe to do nothing in this case. */ |
24669e584 hugetlb: use mmu_... |
1280 |
if (vma->vm_file) { |
83cde9e8b mm: use new helpe... |
1281 |
i_mmap_lock_write(vma->vm_file->f_mapping); |
d833352a4 mm: hugetlbfs: cl... |
1282 |
__unmap_hugepage_range_final(tlb, vma, start, end, NULL); |
83cde9e8b mm: use new helpe... |
1283 |
i_mmap_unlock_write(vma->vm_file->f_mapping); |
24669e584 hugetlb: use mmu_... |
1284 |
} |
f5cc4eef9 VM: make zap_page... |
1285 1286 1287 |
} else unmap_page_range(tlb, vma, start, end, details); } |
1da177e4c Linux-2.6.12-rc2 |
1288 |
} |
1da177e4c Linux-2.6.12-rc2 |
1289 1290 |
/** * unmap_vmas - unmap a range of memory covered by a list of vma's |
0164f69d0 mm/memory.c: fix ... |
1291 |
* @tlb: address of the caller's struct mmu_gather |
1da177e4c Linux-2.6.12-rc2 |
1292 1293 1294 |
* @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 |
1295 |
* |
508034a32 [PATCH] mm: unmap... |
1296 |
* Unmap all pages in the vma list. |
1da177e4c Linux-2.6.12-rc2 |
1297 |
* |
1da177e4c Linux-2.6.12-rc2 |
1298 1299 1300 1301 1302 1303 1304 1305 1306 |
* 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... |
1307 |
void unmap_vmas(struct mmu_gather *tlb, |
1da177e4c Linux-2.6.12-rc2 |
1308 |
struct vm_area_struct *vma, unsigned long start_addr, |
4f74d2c8e vm: remove 'nr_ac... |
1309 |
unsigned long end_addr) |
1da177e4c Linux-2.6.12-rc2 |
1310 |
{ |
cddb8a5c1 mmu-notifiers: core |
1311 |
struct mm_struct *mm = vma->vm_mm; |
1da177e4c Linux-2.6.12-rc2 |
1312 |
|
cddb8a5c1 mmu-notifiers: core |
1313 |
mmu_notifier_invalidate_range_start(mm, start_addr, end_addr); |
f5cc4eef9 VM: make zap_page... |
1314 |
for ( ; vma && vma->vm_start < end_addr; vma = vma->vm_next) |
4f74d2c8e vm: remove 'nr_ac... |
1315 |
unmap_single_vma(tlb, vma, start_addr, end_addr, NULL); |
cddb8a5c1 mmu-notifiers: core |
1316 |
mmu_notifier_invalidate_range_end(mm, start_addr, end_addr); |
1da177e4c Linux-2.6.12-rc2 |
1317 1318 1319 1320 1321 |
} /** * zap_page_range - remove user pages in a given range * @vma: vm_area_struct holding the applicable pages |
eb4546bbb mm/memory.c: fix ... |
1322 |
* @start: starting address of pages to zap |
1da177e4c Linux-2.6.12-rc2 |
1323 1324 |
* @size: number of bytes to zap * @details: details of nonlinear truncation or shared cache invalidation |
f5cc4eef9 VM: make zap_page... |
1325 1326 |
* * Caller must protect the VMA list |
1da177e4c Linux-2.6.12-rc2 |
1327 |
*/ |
7e027b14d vm: simplify unma... |
1328 |
void zap_page_range(struct vm_area_struct *vma, unsigned long start, |
1da177e4c Linux-2.6.12-rc2 |
1329 1330 1331 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550 mm: mmu_gather re... |
1332 |
struct mmu_gather tlb; |
7e027b14d vm: simplify unma... |
1333 |
unsigned long end = start + size; |
1da177e4c Linux-2.6.12-rc2 |
1334 |
|
1da177e4c Linux-2.6.12-rc2 |
1335 |
lru_add_drain(); |
2b047252d Fix TLB gather vi... |
1336 |
tlb_gather_mmu(&tlb, mm, start, end); |
365e9c87a [PATCH] mm: updat... |
1337 |
update_hiwater_rss(mm); |
7e027b14d vm: simplify unma... |
1338 1339 |
mmu_notifier_invalidate_range_start(mm, start, end); for ( ; vma && vma->vm_start < end; vma = vma->vm_next) |
4f74d2c8e vm: remove 'nr_ac... |
1340 |
unmap_single_vma(&tlb, vma, start, end, details); |
7e027b14d vm: simplify unma... |
1341 1342 |
mmu_notifier_invalidate_range_end(mm, start, end); tlb_finish_mmu(&tlb, start, end); |
1da177e4c Linux-2.6.12-rc2 |
1343 |
} |
c627f9cc0 mm: add zap_vma_p... |
1344 |
/** |
f5cc4eef9 VM: make zap_page... |
1345 1346 1347 1348 1349 1350 1351 |
* 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 |
1352 |
*/ |
f5cc4eef9 VM: make zap_page... |
1353 |
static void zap_page_range_single(struct vm_area_struct *vma, unsigned long address, |
1da177e4c Linux-2.6.12-rc2 |
1354 1355 1356 |
unsigned long size, struct zap_details *details) { struct mm_struct *mm = vma->vm_mm; |
d16dfc550 mm: mmu_gather re... |
1357 |
struct mmu_gather tlb; |
1da177e4c Linux-2.6.12-rc2 |
1358 |
unsigned long end = address + size; |
1da177e4c Linux-2.6.12-rc2 |
1359 |
|
1da177e4c Linux-2.6.12-rc2 |
1360 |
lru_add_drain(); |
2b047252d Fix TLB gather vi... |
1361 |
tlb_gather_mmu(&tlb, mm, address, end); |
365e9c87a [PATCH] mm: updat... |
1362 |
update_hiwater_rss(mm); |
f5cc4eef9 VM: make zap_page... |
1363 |
mmu_notifier_invalidate_range_start(mm, address, end); |
4f74d2c8e vm: remove 'nr_ac... |
1364 |
unmap_single_vma(&tlb, vma, address, end, details); |
f5cc4eef9 VM: make zap_page... |
1365 |
mmu_notifier_invalidate_range_end(mm, address, end); |
d16dfc550 mm: mmu_gather re... |
1366 |
tlb_finish_mmu(&tlb, address, end); |
1da177e4c Linux-2.6.12-rc2 |
1367 |
} |
c627f9cc0 mm: add zap_vma_p... |
1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 |
/** * 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... |
1386 |
zap_page_range_single(vma, address, size, NULL); |
c627f9cc0 mm: add zap_vma_p... |
1387 1388 1389 |
return 0; } EXPORT_SYMBOL_GPL(zap_vma_ptes); |
25ca1d6c0 mm: wrap get_lock... |
1390 |
pte_t *__get_locked_pte(struct mm_struct *mm, unsigned long addr, |
920c7a5d0 mm: remove fastca... |
1391 |
spinlock_t **ptl) |
c9cfcddfd VM: add common he... |
1392 1393 1394 1395 |
{ pgd_t * pgd = pgd_offset(mm, addr); pud_t * pud = pud_alloc(mm, pgd, addr); if (pud) { |
49c91fb01 [PATCH] VM: Fix t... |
1396 |
pmd_t * pmd = pmd_alloc(mm, pud, addr); |
f66055ab6 thp: verify pmd_t... |
1397 1398 |
if (pmd) { VM_BUG_ON(pmd_trans_huge(*pmd)); |
c9cfcddfd VM: add common he... |
1399 |
return pte_alloc_map_lock(mm, pmd, addr, ptl); |
f66055ab6 thp: verify pmd_t... |
1400 |
} |
c9cfcddfd VM: add common he... |
1401 1402 1403 |
} return NULL; } |
1da177e4c Linux-2.6.12-rc2 |
1404 |
/* |
238f58d89 Support strange d... |
1405 1406 1407 1408 1409 1410 |
* 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... |
1411 1412 |
static int insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot) |
238f58d89 Support strange d... |
1413 |
{ |
423bad600 mm: add vm_insert... |
1414 |
struct mm_struct *mm = vma->vm_mm; |
238f58d89 Support strange d... |
1415 |
int retval; |
c9cfcddfd VM: add common he... |
1416 |
pte_t *pte; |
8a9f3ccd2 Memory controller... |
1417 |
spinlock_t *ptl; |
238f58d89 Support strange d... |
1418 |
retval = -EINVAL; |
a145dd411 VM: add "vm_inser... |
1419 |
if (PageAnon(page)) |
5b4e655e9 memcg: avoid acco... |
1420 |
goto out; |
238f58d89 Support strange d... |
1421 1422 |
retval = -ENOMEM; flush_dcache_page(page); |
c9cfcddfd VM: add common he... |
1423 |
pte = get_locked_pte(mm, addr, &ptl); |
238f58d89 Support strange d... |
1424 |
if (!pte) |
5b4e655e9 memcg: avoid acco... |
1425 |
goto out; |
238f58d89 Support strange d... |
1426 1427 1428 1429 1430 1431 |
retval = -EBUSY; if (!pte_none(*pte)) goto out_unlock; /* Ok, finally just insert the thing.. */ get_page(page); |
34e55232e mm: avoid false s... |
1432 |
inc_mm_counter_fast(mm, MM_FILEPAGES); |
238f58d89 Support strange d... |
1433 1434 1435 1436 |
page_add_file_rmap(page); set_pte_at(mm, addr, pte, mk_pte(page, prot)); retval = 0; |
8a9f3ccd2 Memory controller... |
1437 1438 |
pte_unmap_unlock(pte, ptl); return retval; |
238f58d89 Support strange d... |
1439 1440 1441 1442 1443 |
out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
bfa5bf6d6 [PATCH] Add kerne... |
1444 1445 1446 1447 1448 1449 |
/** * 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... |
1450 1451 1452 1453 1454 1455 |
* 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... |
1456 |
* (see split_page()). |
a145dd411 VM: add "vm_inser... |
1457 1458 1459 1460 1461 1462 1463 1464 |
* * NOTE! Traditionally this was done with "remap_pfn_range()" which * took an arbitrary page protection parameter. This doesn't allow * that. Your vma protection will have to be set up correctly, which * means that if you want a shared writable mapping, you'd better * ask for a shared writable mapping! * * The page does not need to be reserved. |
4b6e1e370 mm: kill vma flag... |
1465 1466 1467 1468 1469 |
* * Usually this function is called from f_op->mmap() handler * under mm->mmap_sem write-lock, so it can change vma->vm_flags. * Caller must set VM_MIXEDMAP on vma if it wants to call this * function from other places, for example from page-fault handler. |
a145dd411 VM: add "vm_inser... |
1470 |
*/ |
423bad600 mm: add vm_insert... |
1471 1472 |
int vm_insert_page(struct vm_area_struct *vma, unsigned long addr, struct page *page) |
a145dd411 VM: add "vm_inser... |
1473 1474 1475 1476 1477 |
{ if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; if (!page_count(page)) return -EINVAL; |
4b6e1e370 mm: kill vma flag... |
1478 1479 1480 1481 1482 |
if (!(vma->vm_flags & VM_MIXEDMAP)) { BUG_ON(down_read_trylock(&vma->vm_mm->mmap_sem)); BUG_ON(vma->vm_flags & VM_PFNMAP); vma->vm_flags |= VM_MIXEDMAP; } |
423bad600 mm: add vm_insert... |
1483 |
return insert_page(vma, addr, page, vma->vm_page_prot); |
a145dd411 VM: add "vm_inser... |
1484 |
} |
e3c3374fb Make vm_insert_pa... |
1485 |
EXPORT_SYMBOL(vm_insert_page); |
a145dd411 VM: add "vm_inser... |
1486 |
|
423bad600 mm: add vm_insert... |
1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 |
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... |
1506 |
update_mmu_cache(vma, addr, pte); /* XXX: why not for insert_page? */ |
423bad600 mm: add vm_insert... |
1507 1508 1509 1510 1511 1512 1513 |
retval = 0; out_unlock: pte_unmap_unlock(pte, ptl); out: return retval; } |
e0dc0d8f4 [PATCH] add vm_in... |
1514 1515 1516 1517 1518 1519 |
/** * vm_insert_pfn - insert single pfn into user vma * @vma: user vma to map to * @addr: target user address of this page * @pfn: source kernel pfn * |
c462f179e mm/memory.c: fix ... |
1520 |
* Similar to vm_insert_page, this allows drivers to insert individual pages |
e0dc0d8f4 [PATCH] add vm_in... |
1521 1522 1523 1524 |
* 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 |
1525 1526 1527 1528 1529 |
* * 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... |
1530 1531 |
*/ int vm_insert_pfn(struct vm_area_struct *vma, unsigned long addr, |
423bad600 mm: add vm_insert... |
1532 |
unsigned long pfn) |
e0dc0d8f4 [PATCH] add vm_in... |
1533 |
{ |
2ab640379 x86: PAT: hooks i... |
1534 |
int ret; |
e4b866ed1 x86 PAT: change t... |
1535 |
pgprot_t pgprot = vma->vm_page_prot; |
7e675137a mm: introduce pte... |
1536 1537 1538 1539 1540 1541 |
/* * 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_... |
1542 1543 1544 1545 1546 |
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... |
1547 |
|
423bad600 mm: add vm_insert... |
1548 1549 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
5180da410 x86, pat: separat... |
1550 |
if (track_pfn_insert(vma, &pgprot, pfn)) |
2ab640379 x86: PAT: hooks i... |
1551 |
return -EINVAL; |
e4b866ed1 x86 PAT: change t... |
1552 |
ret = insert_pfn(vma, addr, pfn, pgprot); |
2ab640379 x86: PAT: hooks i... |
1553 |
|
2ab640379 x86: PAT: hooks i... |
1554 |
return ret; |
423bad600 mm: add vm_insert... |
1555 1556 |
} EXPORT_SYMBOL(vm_insert_pfn); |
e0dc0d8f4 [PATCH] add vm_in... |
1557 |
|
423bad600 mm: add vm_insert... |
1558 1559 1560 1561 |
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... |
1562 |
|
423bad600 mm: add vm_insert... |
1563 1564 |
if (addr < vma->vm_start || addr >= vma->vm_end) return -EFAULT; |
e0dc0d8f4 [PATCH] add vm_in... |
1565 |
|
423bad600 mm: add vm_insert... |
1566 1567 1568 1569 |
/* * 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... |
1570 1571 |
* 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... |
1572 1573 1574 1575 1576 1577 1578 1579 |
*/ 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... |
1580 |
} |
423bad600 mm: add vm_insert... |
1581 |
EXPORT_SYMBOL(vm_insert_mixed); |
e0dc0d8f4 [PATCH] add vm_in... |
1582 |
|
a145dd411 VM: add "vm_inser... |
1583 |
/* |
1da177e4c Linux-2.6.12-rc2 |
1584 1585 1586 1587 1588 1589 1590 1591 1592 |
* 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... |
1593 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
1594 |
|
c74df32c7 [PATCH] mm: ptd_a... |
1595 |
pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
1596 1597 |
if (!pte) return -ENOMEM; |
6606c3e0d [PATCH] paravirt:... |
1598 |
arch_enter_lazy_mmu_mode(); |
1da177e4c Linux-2.6.12-rc2 |
1599 1600 |
do { BUG_ON(!pte_none(*pte)); |
7e675137a mm: introduce pte... |
1601 |
set_pte_at(mm, addr, pte, pte_mkspecial(pfn_pte(pfn, prot))); |
1da177e4c Linux-2.6.12-rc2 |
1602 1603 |
pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); |
6606c3e0d [PATCH] paravirt:... |
1604 |
arch_leave_lazy_mmu_mode(); |
c74df32c7 [PATCH] mm: ptd_a... |
1605 |
pte_unmap_unlock(pte - 1, ptl); |
1da177e4c Linux-2.6.12-rc2 |
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 |
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... |
1620 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
1da177e4c Linux-2.6.12-rc2 |
1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 |
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... |
1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 |
/** * 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 |
1659 1660 1661 1662 1663 |
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... |
1664 |
unsigned long end = addr + PAGE_ALIGN(size); |
1da177e4c Linux-2.6.12-rc2 |
1665 1666 1667 1668 1669 1670 1671 1672 |
struct mm_struct *mm = vma->vm_mm; int err; /* * Physically remapped pages are special. Tell the * rest of the world about it: * VM_IO tells people not to look at these pages * (accesses can have side effects). |
6aab341e0 mm: re-architect ... |
1673 1674 1675 |
* VM_PFNMAP tells the core MM that the base pages are just * raw PFN mappings, and do not have a "struct page" associated * with them. |
314e51b98 mm: kill vma flag... |
1676 1677 1678 1679 |
* VM_DONTEXPAND * Disable vma merging and expanding with mremap(). * VM_DONTDUMP * Omit vma from core dump, even when VM_IO turned off. |
fb155c161 Allow arbitrary s... |
1680 1681 1682 1683 |
* * There's a horrible special case to handle copy-on-write * behaviour that some programs depend on. We mark the "original" * un-COW'ed pages by matching them up with "vma->vm_pgoff". |
b3b9c2932 mm, x86, pat: rew... |
1684 |
* See vm_normal_page() for details. |
1da177e4c Linux-2.6.12-rc2 |
1685 |
*/ |
b3b9c2932 mm, x86, pat: rew... |
1686 1687 1688 |
if (is_cow_mapping(vma->vm_flags)) { if (addr != vma->vm_start || end != vma->vm_end) return -EINVAL; |
fb155c161 Allow arbitrary s... |
1689 |
vma->vm_pgoff = pfn; |
b3b9c2932 mm, x86, pat: rew... |
1690 1691 1692 1693 |
} err = track_pfn_remap(vma, &prot, pfn, addr, PAGE_ALIGN(size)); if (err) |
3c8bb73ac x86: PAT: store v... |
1694 |
return -EINVAL; |
fb155c161 Allow arbitrary s... |
1695 |
|
314e51b98 mm: kill vma flag... |
1696 |
vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP; |
1da177e4c Linux-2.6.12-rc2 |
1697 1698 1699 1700 1701 |
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 |
1702 1703 1704 1705 1706 1707 1708 |
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... |
1709 1710 |
if (err) |
5180da410 x86, pat: separat... |
1711 |
untrack_pfn(vma, pfn, PAGE_ALIGN(size)); |
2ab640379 x86: PAT: hooks i... |
1712 |
|
1da177e4c Linux-2.6.12-rc2 |
1713 1714 1715 |
return err; } EXPORT_SYMBOL(remap_pfn_range); |
b4cbb197c vm: add vm_iomap_... |
1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 |
/** * vm_iomap_memory - remap memory to userspace * @vma: user vma to map to * @start: start of area * @len: size of area * * This is a simplified io_remap_pfn_range() for common driver use. The * driver just needs to give us the physical memory range to be mapped, * we'll figure out the rest from the vma information. * * NOTE! Some drivers might want to tweak vma->vm_page_prot first to get * whatever write-combining details or similar. */ int vm_iomap_memory(struct vm_area_struct *vma, phys_addr_t start, unsigned long len) { unsigned long vm_len, pfn, pages; /* Check that the physical memory area passed in looks valid */ if (start + len < start) return -EINVAL; /* * You *really* shouldn't map things that aren't page-aligned, * but we've historically allowed it because IO memory might * just have smaller alignment. */ len += start & ~PAGE_MASK; pfn = start >> PAGE_SHIFT; pages = (len + ~PAGE_MASK) >> PAGE_SHIFT; if (pfn + pages < pfn) return -EINVAL; /* We start the mapping 'vm_pgoff' pages into the area */ if (vma->vm_pgoff > pages) return -EINVAL; pfn += vma->vm_pgoff; pages -= vma->vm_pgoff; /* Can we fit all of the mapping? */ vm_len = vma->vm_end - vma->vm_start; if (vm_len >> PAGE_SHIFT > pages) return -EINVAL; /* Ok, let it rip */ return io_remap_pfn_range(vma, vma->vm_start, pfn, vm_len, vma->vm_page_prot); } EXPORT_SYMBOL(vm_iomap_memory); |
aee16b3ce Add apply_to_page... |
1762 1763 1764 1765 1766 1767 |
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... |
1768 |
pgtable_t token; |
949099148 Add unitialized_v... |
1769 |
spinlock_t *uninitialized_var(ptl); |
aee16b3ce Add apply_to_page... |
1770 1771 1772 1773 1774 1775 1776 1777 |
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... |
1778 |
arch_enter_lazy_mmu_mode(); |
2f569afd9 CONFIG_HIGHPTE vs... |
1779 |
token = pmd_pgtable(*pmd); |
aee16b3ce Add apply_to_page... |
1780 1781 |
do { |
c36987e2e mm: don't call pt... |
1782 |
err = fn(pte++, token, addr, data); |
aee16b3ce Add apply_to_page... |
1783 1784 |
if (err) break; |
c36987e2e mm: don't call pt... |
1785 |
} while (addr += PAGE_SIZE, addr != end); |
aee16b3ce Add apply_to_page... |
1786 |
|
38e0edb15 mm/apply_to_range... |
1787 |
arch_leave_lazy_mmu_mode(); |
aee16b3ce Add apply_to_page... |
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 |
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... |
1800 |
BUG_ON(pud_huge(*pud)); |
aee16b3ce Add apply_to_page... |
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 1836 1837 1838 1839 1840 1841 |
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... |
1842 |
unsigned long end = addr + size; |
aee16b3ce Add apply_to_page... |
1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 |
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... |
1853 |
|
aee16b3ce Add apply_to_page... |
1854 1855 1856 |
return err; } EXPORT_SYMBOL_GPL(apply_to_page_range); |
1da177e4c Linux-2.6.12-rc2 |
1857 |
/* |
8f4e2101f [PATCH] mm: page ... |
1858 1859 1860 |
* 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... |
1861 |
* might give a mix of unmatched parts, do_swap_page and do_nonlinear_fault |
8f4e2101f [PATCH] mm: page ... |
1862 1863 |
* 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... |
1864 |
* and do_anonymous_page can safely check later on). |
8f4e2101f [PATCH] mm: page ... |
1865 |
*/ |
4c21e2f24 [PATCH] mm: split... |
1866 |
static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd, |
8f4e2101f [PATCH] mm: page ... |
1867 1868 1869 1870 1871 |
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... |
1872 1873 |
spinlock_t *ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
8f4e2101f [PATCH] mm: page ... |
1874 |
same = pte_same(*page_table, orig_pte); |
4c21e2f24 [PATCH] mm: split... |
1875 |
spin_unlock(ptl); |
8f4e2101f [PATCH] mm: page ... |
1876 1877 1878 1879 1880 |
} #endif pte_unmap(page_table); return same; } |
9de455b20 [PATCH] Pass vma ... |
1881 |
static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma) |
6aab341e0 mm: re-architect ... |
1882 |
{ |
0abdd7a81 dma-debug: introd... |
1883 |
debug_dma_assert_idle(src); |
6aab341e0 mm: re-architect ... |
1884 1885 1886 1887 1888 1889 1890 |
/* * 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... |
1891 |
void *kaddr = kmap_atomic(dst); |
5d2a2dbbc cow_user_page: fi... |
1892 1893 1894 1895 1896 1897 1898 1899 1900 |
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()/... |
1901 |
clear_page(kaddr); |
9b04c5fec mm: remove the se... |
1902 |
kunmap_atomic(kaddr); |
c4ec7b0de [PATCH] mm: D-cac... |
1903 |
flush_dcache_page(dst); |
0ed361dec mm: fix PageUptod... |
1904 1905 |
} else copy_user_highpage(dst, src, va, vma); |
6aab341e0 mm: re-architect ... |
1906 |
} |
1da177e4c Linux-2.6.12-rc2 |
1907 |
/* |
fb09a4642 mm: consolidate c... |
1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 |
* Notify the address space that the page is about to become writable so that * it can prohibit this or wait for the page to get into an appropriate state. * * We do this without the lock held, so that it can sleep if it needs to. */ static int do_page_mkwrite(struct vm_area_struct *vma, struct page *page, unsigned long address) { struct vm_fault vmf; int ret; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = page->index; vmf.flags = FAULT_FLAG_WRITE|FAULT_FLAG_MKWRITE; vmf.page = page; ret = vma->vm_ops->page_mkwrite(vma, &vmf); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) return ret; if (unlikely(!(ret & VM_FAULT_LOCKED))) { lock_page(page); if (!page->mapping) { unlock_page(page); return 0; /* retry */ } ret |= VM_FAULT_LOCKED; } else VM_BUG_ON_PAGE(!PageLocked(page), page); return ret; } /* |
1da177e4c Linux-2.6.12-rc2 |
1940 1941 1942 1943 |
* 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 |
1944 1945 1946 1947 1948 1949 1950 1951 1952 |
* 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 ... |
1953 1954 1955 |
* 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 |
1956 |
*/ |
65500d234 [PATCH] mm: page ... |
1957 1958 |
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 ... |
1959 |
spinlock_t *ptl, pte_t orig_pte) |
e6219ec81 mm: add lock rele... |
1960 |
__releases(ptl) |
1da177e4c Linux-2.6.12-rc2 |
1961 |
{ |
2ec74c3ef mm: move all mmu ... |
1962 |
struct page *old_page, *new_page = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1963 |
pte_t entry; |
b009c024f do_wp_page: remov... |
1964 |
int ret = 0; |
a200ee182 mm: set_page_dirt... |
1965 |
int page_mkwrite = 0; |
d08b3851d [PATCH] mm: track... |
1966 |
struct page *dirty_page = NULL; |
1756954c6 mm: fix build war... |
1967 1968 |
unsigned long mmun_start = 0; /* For mmu_notifiers */ unsigned long mmun_end = 0; /* For mmu_notifiers */ |
00501b531 mm: memcontrol: r... |
1969 |
struct mem_cgroup *memcg; |
1da177e4c Linux-2.6.12-rc2 |
1970 |
|
6aab341e0 mm: re-architect ... |
1971 |
old_page = vm_normal_page(vma, address, orig_pte); |
251b97f55 mm: dirty page ac... |
1972 1973 |
if (!old_page) { /* |
64e455079 mm: softdirty: en... |
1974 1975 |
* VM_MIXEDMAP !pfn_valid() case, or VM_SOFTDIRTY clear on a * VM_PFNMAP VMA. |
251b97f55 mm: dirty page ac... |
1976 1977 1978 1979 1980 1981 1982 1983 |
* * 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 ... |
1984 |
goto gotten; |
251b97f55 mm: dirty page ac... |
1985 |
} |
1da177e4c Linux-2.6.12-rc2 |
1986 |
|
d08b3851d [PATCH] mm: track... |
1987 |
/* |
ee6a64578 [PATCH] mm: fixup... |
1988 1989 |
* Take out anonymous pages first, anonymous shared vmas are * not dirty accountable. |
d08b3851d [PATCH] mm: track... |
1990 |
*/ |
9a8408951 ksm: identify Pag... |
1991 |
if (PageAnon(old_page) && !PageKsm(old_page)) { |
ab967d860 mm: wp lock page ... |
1992 1993 1994 1995 1996 1997 1998 1999 |
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 ... |
2000 2001 2002 |
goto unlock; } page_cache_release(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2003 |
} |
b009c024f do_wp_page: remov... |
2004 |
if (reuse_swap_page(old_page)) { |
c44b67432 rmap: move exclus... |
2005 2006 2007 2008 2009 2010 |
/* * 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... |
2011 2012 2013 |
unlock_page(old_page); goto reuse; } |
ab967d860 mm: wp lock page ... |
2014 |
unlock_page(old_page); |
ee6a64578 [PATCH] mm: fixup... |
2015 |
} else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) == |
d08b3851d [PATCH] mm: track... |
2016 |
(VM_WRITE|VM_SHARED))) { |
ee6a64578 [PATCH] mm: fixup... |
2017 2018 2019 2020 2021 |
/* * 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_... |
2022 |
if (vma->vm_ops && vma->vm_ops->page_mkwrite) { |
c2ec175c3 mm: page_mkwrite ... |
2023 |
int tmp; |
9637a5efd [PATCH] add page_... |
2024 2025 |
page_cache_get(old_page); pte_unmap_unlock(page_table, ptl); |
fb09a4642 mm: consolidate c... |
2026 2027 2028 2029 2030 |
tmp = do_page_mkwrite(vma, old_page, address); if (unlikely(!tmp || (tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) { page_cache_release(old_page); return tmp; |
c2ec175c3 mm: page_mkwrite ... |
2031 |
} |
9637a5efd [PATCH] add page_... |
2032 2033 2034 2035 2036 2037 2038 2039 |
/* * 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... |
2040 2041 |
if (!pte_same(*page_table, orig_pte)) { unlock_page(old_page); |
9637a5efd [PATCH] add page_... |
2042 |
goto unlock; |
b827e496c mm: close page_mk... |
2043 |
} |
a200ee182 mm: set_page_dirt... |
2044 2045 |
page_mkwrite = 1; |
1da177e4c Linux-2.6.12-rc2 |
2046 |
} |
d08b3851d [PATCH] mm: track... |
2047 2048 |
dirty_page = old_page; get_page(dirty_page); |
9637a5efd [PATCH] add page_... |
2049 |
|
251b97f55 mm: dirty page ac... |
2050 |
reuse: |
8c8a743c5 sched/numa: Use {... |
2051 2052 2053 2054 2055 2056 2057 |
/* * Clear the pages cpupid information as the existing * information potentially belongs to a now completely * unrelated process. */ if (old_page) page_cpupid_xchg_last(old_page, (1 << LAST_CPUPID_SHIFT) - 1); |
9637a5efd [PATCH] add page_... |
2058 2059 2060 |
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... |
2061 |
if (ptep_set_access_flags(vma, address, page_table, entry,1)) |
4b3073e1c MM: Pass a PTE po... |
2062 |
update_mmu_cache(vma, address, page_table); |
72ddc8f72 do_wp_page: clari... |
2063 |
pte_unmap_unlock(page_table, ptl); |
9637a5efd [PATCH] add page_... |
2064 |
ret |= VM_FAULT_WRITE; |
72ddc8f72 do_wp_page: clari... |
2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 |
if (!dirty_page) return ret; /* * Yes, Virginia, this is actually required to prevent a race * with clear_page_dirty_for_io() from clearing the page dirty * bit after it clear all dirty ptes, but before a racing * do_wp_page installs a dirty pte. * |
f0c6d4d29 mm: introduce do_... |
2075 |
* do_shared_fault is protected similarly. |
72ddc8f72 do_wp_page: clari... |
2076 2077 2078 |
*/ if (!page_mkwrite) { wait_on_page_locked(dirty_page); |
ed6d7c8e5 mm: remove unused... |
2079 |
set_page_dirty_balance(dirty_page); |
41c4d25f7 mm: Update file t... |
2080 2081 2082 |
/* file_update_time outside page_lock */ if (vma->vm_file) file_update_time(vma->vm_file); |
72ddc8f72 do_wp_page: clari... |
2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 |
} put_page(dirty_page); if (page_mkwrite) { struct address_space *mapping = dirty_page->mapping; set_page_dirty(dirty_page); unlock_page(dirty_page); page_cache_release(dirty_page); if (mapping) { /* * Some device drivers do not set page.mapping * but still dirty their pages */ balance_dirty_pages_ratelimited(mapping); } } |
72ddc8f72 do_wp_page: clari... |
2099 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2100 |
} |
1da177e4c Linux-2.6.12-rc2 |
2101 2102 2103 2104 |
/* * Ok, we need to copy. Oh, well.. */ |
b5810039a [PATCH] core remo... |
2105 |
page_cache_get(old_page); |
920fc356f [PATCH] unpaged: ... |
2106 |
gotten: |
8f4e2101f [PATCH] mm: page ... |
2107 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2108 2109 |
if (unlikely(anon_vma_prepare(vma))) |
65500d234 [PATCH] mm: page ... |
2110 |
goto oom; |
a13ea5b75 mm: reinstate ZER... |
2111 |
|
62eede62d mm: ZERO_PAGE wit... |
2112 |
if (is_zero_pfn(pte_pfn(orig_pte))) { |
a13ea5b75 mm: reinstate ZER... |
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 |
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); |
00501b531 mm: memcontrol: r... |
2123 |
if (mem_cgroup_try_charge(new_page, mm, GFP_KERNEL, &memcg)) |
8a9f3ccd2 Memory controller... |
2124 |
goto oom_free_new; |
6bdb913f0 mm: wrap calls to... |
2125 |
mmun_start = address & PAGE_MASK; |
1756954c6 mm: fix build war... |
2126 |
mmun_end = mmun_start + PAGE_SIZE; |
6bdb913f0 mm: wrap calls to... |
2127 |
mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end); |
1da177e4c Linux-2.6.12-rc2 |
2128 2129 2130 |
/* * Re-check the pte - we dropped the lock */ |
8f4e2101f [PATCH] mm: page ... |
2131 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
65500d234 [PATCH] mm: page ... |
2132 |
if (likely(pte_same(*page_table, orig_pte))) { |
920fc356f [PATCH] unpaged: ... |
2133 |
if (old_page) { |
920fc356f [PATCH] unpaged: ... |
2134 |
if (!PageAnon(old_page)) { |
34e55232e mm: avoid false s... |
2135 2136 |
dec_mm_counter_fast(mm, MM_FILEPAGES); inc_mm_counter_fast(mm, MM_ANONPAGES); |
920fc356f [PATCH] unpaged: ... |
2137 2138 |
} } else |
34e55232e mm: avoid false s... |
2139 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
eca351336 [PATCH] Fix missi... |
2140 |
flush_cache_page(vma, address, pte_pfn(orig_pte)); |
65500d234 [PATCH] mm: page ... |
2141 2142 |
entry = mk_pte(new_page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); |
4ce072f1f [PATCH] mm: fix a... |
2143 2144 2145 2146 2147 2148 |
/* * 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. */ |
34ee645e8 mmu_notifier: cal... |
2149 |
ptep_clear_flush_notify(vma, address, page_table); |
9617d95e6 [PATCH] mm: rmap ... |
2150 |
page_add_new_anon_rmap(new_page, vma, address); |
00501b531 mm: memcontrol: r... |
2151 2152 |
mem_cgroup_commit_charge(new_page, memcg, false); lru_cache_add_active_or_unevictable(new_page, vma); |
828502d30 ksm: add mmu_noti... |
2153 2154 2155 2156 2157 2158 |
/* * 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... |
2159 |
update_mmu_cache(vma, address, page_table); |
945754a17 mm: fix race in C... |
2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 |
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... |
2183 |
page_remove_rmap(old_page); |
945754a17 mm: fix race in C... |
2184 |
} |
1da177e4c Linux-2.6.12-rc2 |
2185 2186 |
/* Free the old page.. */ new_page = old_page; |
f33ea7f40 [PATCH] fix get_u... |
2187 |
ret |= VM_FAULT_WRITE; |
8a9f3ccd2 Memory controller... |
2188 |
} else |
00501b531 mm: memcontrol: r... |
2189 |
mem_cgroup_cancel_charge(new_page, memcg); |
8a9f3ccd2 Memory controller... |
2190 |
|
6bdb913f0 mm: wrap calls to... |
2191 2192 |
if (new_page) page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2193 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2194 |
pte_unmap_unlock(page_table, ptl); |
1756954c6 mm: fix build war... |
2195 |
if (mmun_end > mmun_start) |
6bdb913f0 mm: wrap calls to... |
2196 |
mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end); |
e15f8c01a mlock: fix race w... |
2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 |
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... |
2209 |
return ret; |
8a9f3ccd2 Memory controller... |
2210 |
oom_free_new: |
6dbf6d3bb memcg: page_cache... |
2211 |
page_cache_release(new_page); |
65500d234 [PATCH] mm: page ... |
2212 |
oom: |
66521d5aa mm/memory.c: remo... |
2213 |
if (old_page) |
920fc356f [PATCH] unpaged: ... |
2214 |
page_cache_release(old_page); |
1da177e4c Linux-2.6.12-rc2 |
2215 2216 |
return VM_FAULT_OOM; } |
97a894136 mm: Remove i_mmap... |
2217 |
static void unmap_mapping_range_vma(struct vm_area_struct *vma, |
1da177e4c Linux-2.6.12-rc2 |
2218 2219 2220 |
unsigned long start_addr, unsigned long end_addr, struct zap_details *details) { |
f5cc4eef9 VM: make zap_page... |
2221 |
zap_page_range_single(vma, start_addr, end_addr - start_addr, details); |
1da177e4c Linux-2.6.12-rc2 |
2222 |
} |
6b2dbba8b mm: replace vma p... |
2223 |
static inline void unmap_mapping_range_tree(struct rb_root *root, |
1da177e4c Linux-2.6.12-rc2 |
2224 2225 2226 |
struct zap_details *details) { struct vm_area_struct *vma; |
1da177e4c Linux-2.6.12-rc2 |
2227 |
pgoff_t vba, vea, zba, zea; |
6b2dbba8b mm: replace vma p... |
2228 |
vma_interval_tree_foreach(vma, root, |
1da177e4c Linux-2.6.12-rc2 |
2229 |
details->first_index, details->last_index) { |
1da177e4c Linux-2.6.12-rc2 |
2230 2231 |
vba = vma->vm_pgoff; |
d6e932177 mm: use vma_pages... |
2232 |
vea = vba + vma_pages(vma) - 1; |
1da177e4c Linux-2.6.12-rc2 |
2233 2234 2235 2236 2237 2238 2239 |
/* 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... |
2240 |
unmap_mapping_range_vma(vma, |
1da177e4c Linux-2.6.12-rc2 |
2241 2242 |
((zba - vba) << PAGE_SHIFT) + vma->vm_start, ((zea - vba + 1) << PAGE_SHIFT) + vma->vm_start, |
97a894136 mm: Remove i_mmap... |
2243 |
details); |
1da177e4c Linux-2.6.12-rc2 |
2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 |
} } static inline void unmap_mapping_range_list(struct list_head *head, struct zap_details *details) { struct vm_area_struct *vma; /* * In nonlinear VMAs there is no correspondence between virtual address * offset and file offset. So we must perform an exhaustive search * across *all* the pages in each nonlinear VMA, not just the pages * whose virtual address lies outside the file truncation point. */ |
6b2dbba8b mm: replace vma p... |
2258 |
list_for_each_entry(vma, head, shared.nonlinear) { |
1da177e4c Linux-2.6.12-rc2 |
2259 |
details->nonlinear_vma = vma; |
97a894136 mm: Remove i_mmap... |
2260 |
unmap_mapping_range_vma(vma, vma->vm_start, vma->vm_end, details); |
1da177e4c Linux-2.6.12-rc2 |
2261 2262 2263 2264 |
} } /** |
72fd4a35a [PATCH] Numerous ... |
2265 |
* 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: ... |
2266 |
* @mapping: the address space containing mmaps to be unmapped. |
1da177e4c Linux-2.6.12-rc2 |
2267 2268 |
* @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... |
2269 |
* boundary. Note that this is different from truncate_pagecache(), which |
1da177e4c Linux-2.6.12-rc2 |
2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 |
* 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 |
2299 |
|
1da177e4c Linux-2.6.12-rc2 |
2300 |
|
48ec833b7 Revert "mm/memory... |
2301 |
i_mmap_lock_write(mapping); |
6b2dbba8b mm: replace vma p... |
2302 |
if (unlikely(!RB_EMPTY_ROOT(&mapping->i_mmap))) |
1da177e4c Linux-2.6.12-rc2 |
2303 2304 2305 |
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); |
48ec833b7 Revert "mm/memory... |
2306 |
i_mmap_unlock_write(mapping); |
1da177e4c Linux-2.6.12-rc2 |
2307 2308 |
} EXPORT_SYMBOL(unmap_mapping_range); |
1da177e4c Linux-2.6.12-rc2 |
2309 |
/* |
8f4e2101f [PATCH] mm: page ... |
2310 2311 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. |
9a95f3cf7 mm: describe mmap... |
2312 2313 2314 2315 |
* We return with pte unmapped and unlocked. * * We return with the mmap_sem locked or unlocked in the same cases * as does filemap_fault(). |
1da177e4c Linux-2.6.12-rc2 |
2316 |
*/ |
65500d234 [PATCH] mm: page ... |
2317 2318 |
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... |
2319 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
2320 |
{ |
8f4e2101f [PATCH] mm: page ... |
2321 |
spinlock_t *ptl; |
56f31801c mm: cleanup "swap... |
2322 |
struct page *page, *swapcache; |
00501b531 mm: memcontrol: r... |
2323 |
struct mem_cgroup *memcg; |
65500d234 [PATCH] mm: page ... |
2324 |
swp_entry_t entry; |
1da177e4c Linux-2.6.12-rc2 |
2325 |
pte_t pte; |
d065bd810 mm: retry page fa... |
2326 |
int locked; |
ad8c2ee80 rmap: add exclusi... |
2327 |
int exclusive = 0; |
83c54070e mm: fault feedbac... |
2328 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2329 |
|
4c21e2f24 [PATCH] mm: split... |
2330 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
8f4e2101f [PATCH] mm: page ... |
2331 |
goto out; |
65500d234 [PATCH] mm: page ... |
2332 2333 |
entry = pte_to_swp_entry(orig_pte); |
d1737fdbe HWPOISON: Add bas... |
2334 2335 2336 2337 2338 2339 2340 |
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... |
2341 |
ret = VM_FAULT_SIGBUS; |
d1737fdbe HWPOISON: Add bas... |
2342 |
} |
0697212a4 [PATCH] Swapless ... |
2343 2344 |
goto out; } |
0ff922452 [PATCH] per-task-... |
2345 |
delayacct_set_flag(DELAYACCT_PF_SWAPIN); |
1da177e4c Linux-2.6.12-rc2 |
2346 2347 |
page = lookup_swap_cache(entry); if (!page) { |
02098feaa swapin needs gfp_... |
2348 2349 |
page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, vma, address); |
1da177e4c Linux-2.6.12-rc2 |
2350 2351 |
if (!page) { /* |
8f4e2101f [PATCH] mm: page ... |
2352 2353 |
* Back out if somebody else faulted in this pte * while we released the pte lock. |
1da177e4c Linux-2.6.12-rc2 |
2354 |
*/ |
8f4e2101f [PATCH] mm: page ... |
2355 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1da177e4c Linux-2.6.12-rc2 |
2356 2357 |
if (likely(pte_same(*page_table, orig_pte))) ret = VM_FAULT_OOM; |
0ff922452 [PATCH] per-task-... |
2358 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
65500d234 [PATCH] mm: page ... |
2359 |
goto unlock; |
1da177e4c Linux-2.6.12-rc2 |
2360 2361 2362 2363 |
} /* Had to read the page from swap area: Major fault */ ret = VM_FAULT_MAJOR; |
f8891e5e1 [PATCH] Light wei... |
2364 |
count_vm_event(PGMAJFAULT); |
456f998ec memcg: add the pa... |
2365 |
mem_cgroup_count_vm_event(mm, PGMAJFAULT); |
d1737fdbe HWPOISON: Add bas... |
2366 |
} else if (PageHWPoison(page)) { |
71f72525d HWPOISON: comment... |
2367 2368 2369 2370 |
/* * hwpoisoned dirty swapcache pages are kept for killing * owner processes (which may be unknown at hwpoison time) */ |
d1737fdbe HWPOISON: Add bas... |
2371 2372 |
ret = VM_FAULT_HWPOISON; delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
56f31801c mm: cleanup "swap... |
2373 |
swapcache = page; |
4779cb31c HWPOISON: Fix pag... |
2374 |
goto out_release; |
1da177e4c Linux-2.6.12-rc2 |
2375 |
} |
56f31801c mm: cleanup "swap... |
2376 |
swapcache = page; |
d065bd810 mm: retry page fa... |
2377 |
locked = lock_page_or_retry(page, mm, flags); |
e709ffd61 mm: remove swap t... |
2378 |
|
073e587ec memcg: move charg... |
2379 |
delayacct_clear_flag(DELAYACCT_PF_SWAPIN); |
d065bd810 mm: retry page fa... |
2380 2381 2382 2383 |
if (!locked) { ret |= VM_FAULT_RETRY; goto out_release; } |
073e587ec memcg: move charg... |
2384 |
|
4969c1192 mm: fix swapin ra... |
2385 |
/* |
31c4a3d3a mm: further fix s... |
2386 2387 2388 2389 |
* 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... |
2390 |
*/ |
31c4a3d3a mm: further fix s... |
2391 |
if (unlikely(!PageSwapCache(page) || page_private(page) != entry.val)) |
4969c1192 mm: fix swapin ra... |
2392 |
goto out_page; |
cbf86cfe0 ksm: remove old s... |
2393 2394 2395 2396 |
page = ksm_might_need_to_copy(page, vma, address); if (unlikely(!page)) { ret = VM_FAULT_OOM; page = swapcache; |
cbf86cfe0 ksm: remove old s... |
2397 |
goto out_page; |
5ad646880 ksm: let shared p... |
2398 |
} |
00501b531 mm: memcontrol: r... |
2399 |
if (mem_cgroup_try_charge(page, mm, GFP_KERNEL, &memcg)) { |
8a9f3ccd2 Memory controller... |
2400 |
ret = VM_FAULT_OOM; |
bc43f75cd mm: fix pageref l... |
2401 |
goto out_page; |
8a9f3ccd2 Memory controller... |
2402 |
} |
1da177e4c Linux-2.6.12-rc2 |
2403 |
/* |
8f4e2101f [PATCH] mm: page ... |
2404 |
* Back out if somebody else already faulted in this pte. |
1da177e4c Linux-2.6.12-rc2 |
2405 |
*/ |
8f4e2101f [PATCH] mm: page ... |
2406 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
9e9bef07c [PATCH] mm: do_sw... |
2407 |
if (unlikely(!pte_same(*page_table, orig_pte))) |
b81074800 [PATCH] do_swap_p... |
2408 |
goto out_nomap; |
b81074800 [PATCH] do_swap_p... |
2409 2410 2411 2412 |
if (unlikely(!PageUptodate(page))) { ret = VM_FAULT_SIGBUS; goto out_nomap; |
1da177e4c Linux-2.6.12-rc2 |
2413 |
} |
8c7c6e34a memcg: mem+swap c... |
2414 2415 2416 2417 2418 2419 2420 2421 |
/* * The page isn't present yet, go ahead with the fault. * * Be careful about the sequence of operations here. * To get its accounting right, reuse_swap_page() must be called * while the page is counted on swap but not yet in mapcount i.e. * before page_add_anon_rmap() and swap_free(); try_to_free_swap() * must be called after the swap_free(), or it will never succeed. |
8c7c6e34a memcg: mem+swap c... |
2422 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2423 |
|
34e55232e mm: avoid false s... |
2424 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
b084d4353 mm: count swap usage |
2425 |
dec_mm_counter_fast(mm, MM_SWAPENTS); |
1da177e4c Linux-2.6.12-rc2 |
2426 |
pte = mk_pte(page, vma->vm_page_prot); |
30c9f3a9f Remove internal u... |
2427 |
if ((flags & FAULT_FLAG_WRITE) && reuse_swap_page(page)) { |
1da177e4c Linux-2.6.12-rc2 |
2428 |
pte = maybe_mkwrite(pte_mkdirty(pte), vma); |
30c9f3a9f Remove internal u... |
2429 |
flags &= ~FAULT_FLAG_WRITE; |
9a5b489b8 mm: set VM_FAULT_... |
2430 |
ret |= VM_FAULT_WRITE; |
ad8c2ee80 rmap: add exclusi... |
2431 |
exclusive = 1; |
1da177e4c Linux-2.6.12-rc2 |
2432 |
} |
1da177e4c Linux-2.6.12-rc2 |
2433 |
flush_icache_page(vma, page); |
179ef71cb mm: save soft-dir... |
2434 2435 |
if (pte_swp_soft_dirty(orig_pte)) pte = pte_mksoft_dirty(pte); |
1da177e4c Linux-2.6.12-rc2 |
2436 |
set_pte_at(mm, address, page_table, pte); |
00501b531 mm: memcontrol: r... |
2437 |
if (page == swapcache) { |
af34770e5 mm: reduce rmap o... |
2438 |
do_page_add_anon_rmap(page, vma, address, exclusive); |
00501b531 mm: memcontrol: r... |
2439 2440 |
mem_cgroup_commit_charge(page, memcg, true); } else { /* ksm created a completely new copy */ |
56f31801c mm: cleanup "swap... |
2441 |
page_add_new_anon_rmap(page, vma, address); |
00501b531 mm: memcontrol: r... |
2442 2443 2444 |
mem_cgroup_commit_charge(page, memcg, false); lru_cache_add_active_or_unevictable(page, vma); } |
1da177e4c Linux-2.6.12-rc2 |
2445 |
|
c475a8ab6 [PATCH] can_share... |
2446 |
swap_free(entry); |
b291f0003 mlock: mlocked pa... |
2447 |
if (vm_swap_full() || (vma->vm_flags & VM_LOCKED) || PageMlocked(page)) |
a2c43eed8 mm: try_to_free_s... |
2448 |
try_to_free_swap(page); |
c475a8ab6 [PATCH] can_share... |
2449 |
unlock_page(page); |
56f31801c mm: cleanup "swap... |
2450 |
if (page != swapcache) { |
4969c1192 mm: fix swapin ra... |
2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 |
/* * 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... |
2462 |
|
30c9f3a9f Remove internal u... |
2463 |
if (flags & FAULT_FLAG_WRITE) { |
61469f1d5 memcg: when do_sw... |
2464 2465 2466 |
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 |
2467 2468 2469 2470 |
goto out; } /* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
2471 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
2472 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2473 |
pte_unmap_unlock(page_table, ptl); |
1da177e4c Linux-2.6.12-rc2 |
2474 2475 |
out: return ret; |
b81074800 [PATCH] do_swap_p... |
2476 |
out_nomap: |
00501b531 mm: memcontrol: r... |
2477 |
mem_cgroup_cancel_charge(page, memcg); |
8f4e2101f [PATCH] mm: page ... |
2478 |
pte_unmap_unlock(page_table, ptl); |
bc43f75cd mm: fix pageref l... |
2479 |
out_page: |
b81074800 [PATCH] do_swap_p... |
2480 |
unlock_page(page); |
4779cb31c HWPOISON: Fix pag... |
2481 |
out_release: |
b81074800 [PATCH] do_swap_p... |
2482 |
page_cache_release(page); |
56f31801c mm: cleanup "swap... |
2483 |
if (page != swapcache) { |
4969c1192 mm: fix swapin ra... |
2484 2485 2486 |
unlock_page(swapcache); page_cache_release(swapcache); } |
65500d234 [PATCH] mm: page ... |
2487 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2488 2489 2490 |
} /* |
8ca3eb080 guard page for st... |
2491 2492 |
* 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 ... |
2493 |
* doesn't hit another vma. |
320b2b8de mm: keep a guard ... |
2494 2495 2496 2497 2498 |
*/ 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... |
2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 |
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 ... |
2509 |
|
d05f3169c mm: make expand_d... |
2510 |
expand_downwards(vma, address - PAGE_SIZE); |
320b2b8de mm: keep a guard ... |
2511 |
} |
8ca3eb080 guard page for st... |
2512 2513 2514 2515 2516 2517 2518 2519 2520 |
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 ... |
2521 2522 2523 2524 |
return 0; } /* |
8f4e2101f [PATCH] mm: page ... |
2525 2526 2527 |
* 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 |
2528 |
*/ |
65500d234 [PATCH] mm: page ... |
2529 2530 |
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... |
2531 |
unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
2532 |
{ |
00501b531 mm: memcontrol: r... |
2533 |
struct mem_cgroup *memcg; |
8f4e2101f [PATCH] mm: page ... |
2534 2535 |
struct page *page; spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
2536 |
pte_t entry; |
1da177e4c Linux-2.6.12-rc2 |
2537 |
|
11ac55247 mm: fix page tabl... |
2538 2539 2540 2541 |
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 ... |
2542 |
return VM_FAULT_SIGBUS; |
11ac55247 mm: fix page tabl... |
2543 |
/* Use the zero-page for reads */ |
593befa6a mm: introduce mm_... |
2544 |
if (!(flags & FAULT_FLAG_WRITE) && !mm_forbids_zeropage(mm)) { |
62eede62d mm: ZERO_PAGE wit... |
2545 2546 |
entry = pte_mkspecial(pfn_pte(my_zero_pfn(address), vma->vm_page_prot)); |
11ac55247 mm: fix page tabl... |
2547 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
a13ea5b75 mm: reinstate ZER... |
2548 2549 2550 2551 |
if (!pte_none(*page_table)) goto unlock; goto setpte; } |
557ed1fa2 remove ZERO_PAGE |
2552 |
/* Allocate our own private page. */ |
557ed1fa2 remove ZERO_PAGE |
2553 2554 2555 2556 2557 |
if (unlikely(anon_vma_prepare(vma))) goto oom; page = alloc_zeroed_user_highpage_movable(vma, address); if (!page) goto oom; |
52f37629f THP: fix comment ... |
2558 2559 2560 2561 2562 |
/* * The memory barrier inside __SetPageUptodate makes sure that * preceeding stores to the page contents become visible before * the set_pte_at() write. */ |
0ed361dec mm: fix PageUptod... |
2563 |
__SetPageUptodate(page); |
8f4e2101f [PATCH] mm: page ... |
2564 |
|
00501b531 mm: memcontrol: r... |
2565 |
if (mem_cgroup_try_charge(page, mm, GFP_KERNEL, &memcg)) |
8a9f3ccd2 Memory controller... |
2566 |
goto oom_free_page; |
557ed1fa2 remove ZERO_PAGE |
2567 |
entry = mk_pte(page, vma->vm_page_prot); |
1ac0cb5d0 mm: fix anonymous... |
2568 2569 |
if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); |
1da177e4c Linux-2.6.12-rc2 |
2570 |
|
557ed1fa2 remove ZERO_PAGE |
2571 |
page_table = pte_offset_map_lock(mm, pmd, address, &ptl); |
1c2fb7a4c ksm: fix deadlock... |
2572 |
if (!pte_none(*page_table)) |
557ed1fa2 remove ZERO_PAGE |
2573 |
goto release; |
9ba692948 ksm: fix oom dead... |
2574 |
|
34e55232e mm: avoid false s... |
2575 |
inc_mm_counter_fast(mm, MM_ANONPAGES); |
557ed1fa2 remove ZERO_PAGE |
2576 |
page_add_new_anon_rmap(page, vma, address); |
00501b531 mm: memcontrol: r... |
2577 2578 |
mem_cgroup_commit_charge(page, memcg, false); lru_cache_add_active_or_unevictable(page, vma); |
a13ea5b75 mm: reinstate ZER... |
2579 |
setpte: |
65500d234 [PATCH] mm: page ... |
2580 |
set_pte_at(mm, address, page_table, entry); |
1da177e4c Linux-2.6.12-rc2 |
2581 2582 |
/* No need to invalidate - it was non-present before */ |
4b3073e1c MM: Pass a PTE po... |
2583 |
update_mmu_cache(vma, address, page_table); |
65500d234 [PATCH] mm: page ... |
2584 |
unlock: |
8f4e2101f [PATCH] mm: page ... |
2585 |
pte_unmap_unlock(page_table, ptl); |
83c54070e mm: fault feedbac... |
2586 |
return 0; |
8f4e2101f [PATCH] mm: page ... |
2587 |
release: |
00501b531 mm: memcontrol: r... |
2588 |
mem_cgroup_cancel_charge(page, memcg); |
8f4e2101f [PATCH] mm: page ... |
2589 2590 |
page_cache_release(page); goto unlock; |
8a9f3ccd2 Memory controller... |
2591 |
oom_free_page: |
6dbf6d3bb memcg: page_cache... |
2592 |
page_cache_release(page); |
65500d234 [PATCH] mm: page ... |
2593 |
oom: |
1da177e4c Linux-2.6.12-rc2 |
2594 2595 |
return VM_FAULT_OOM; } |
9a95f3cf7 mm: describe mmap... |
2596 2597 2598 2599 2600 |
/* * The mmap_sem must have been held on entry, and may have been * released depending on flags and vma->vm_ops->fault() return value. * See filemap_fault() and __lock_page_retry(). */ |
7eae74af3 mm: do_fault(): e... |
2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 |
static int __do_fault(struct vm_area_struct *vma, unsigned long address, pgoff_t pgoff, unsigned int flags, struct page **page) { struct vm_fault vmf; int ret; vmf.virtual_address = (void __user *)(address & PAGE_MASK); vmf.pgoff = pgoff; vmf.flags = flags; vmf.page = NULL; ret = vma->vm_ops->fault(vma, &vmf); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) return ret; if (unlikely(PageHWPoison(vmf.page))) { if (ret & VM_FAULT_LOCKED) unlock_page(vmf.page); page_cache_release(vmf.page); return VM_FAULT_HWPOISON; } if (unlikely(!(ret & VM_FAULT_LOCKED))) lock_page(vmf.page); else VM_BUG_ON_PAGE(!PageLocked(vmf.page), vmf.page); *page = vmf.page; return ret; } |
8c6e50b02 mm: introduce vm_... |
2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 |
/** * do_set_pte - setup new PTE entry for given page and add reverse page mapping. * * @vma: virtual memory area * @address: user virtual address * @page: page to map * @pte: pointer to target page table entry * @write: true, if new entry is writable * @anon: true, if it's anonymous page * * Caller must hold page table lock relevant for @pte. * * Target users are page handler itself and implementations of * vm_ops->map_pages. */ void do_set_pte(struct vm_area_struct *vma, unsigned long address, |
3bb977946 mm: consolidate c... |
2647 2648 2649 2650 2651 2652 2653 2654 2655 |
struct page *page, pte_t *pte, bool write, bool anon) { pte_t entry; flush_icache_page(vma, page); entry = mk_pte(page, vma->vm_page_prot); if (write) entry = maybe_mkwrite(pte_mkdirty(entry), vma); else if (pte_file(*pte) && pte_file_soft_dirty(*pte)) |
9aed8614a mm/memory.c: don'... |
2656 |
entry = pte_mksoft_dirty(entry); |
3bb977946 mm: consolidate c... |
2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 |
if (anon) { inc_mm_counter_fast(vma->vm_mm, MM_ANONPAGES); page_add_new_anon_rmap(page, vma, address); } else { inc_mm_counter_fast(vma->vm_mm, MM_FILEPAGES); page_add_file_rmap(page); } set_pte_at(vma->vm_mm, address, pte, entry); /* no need to invalidate: a not-present page won't be cached */ update_mmu_cache(vma, address, pte); } |
3a91053ae mm: mark fault_ar... |
2669 2670 |
static unsigned long fault_around_bytes __read_mostly = rounddown_pow_of_two(65536); |
a9b0f8618 mm: nominate faul... |
2671 |
|
a9b0f8618 mm: nominate faul... |
2672 2673 |
#ifdef CONFIG_DEBUG_FS static int fault_around_bytes_get(void *data, u64 *val) |
1592eef01 mm: add debugfs t... |
2674 |
{ |
a9b0f8618 mm: nominate faul... |
2675 |
*val = fault_around_bytes; |
1592eef01 mm: add debugfs t... |
2676 2677 |
return 0; } |
b4903d6e8 mm: debugfs: move... |
2678 2679 2680 2681 2682 |
/* * fault_around_pages() and fault_around_mask() expects fault_around_bytes * rounded down to nearest page order. It's what do_fault_around() expects to * see. */ |
a9b0f8618 mm: nominate faul... |
2683 |
static int fault_around_bytes_set(void *data, u64 val) |
1592eef01 mm: add debugfs t... |
2684 |
{ |
a9b0f8618 mm: nominate faul... |
2685 |
if (val / PAGE_SIZE > PTRS_PER_PTE) |
1592eef01 mm: add debugfs t... |
2686 |
return -EINVAL; |
b4903d6e8 mm: debugfs: move... |
2687 2688 2689 2690 |
if (val > PAGE_SIZE) fault_around_bytes = rounddown_pow_of_two(val); else fault_around_bytes = PAGE_SIZE; /* rounddown_pow_of_two(0) is undefined */ |
1592eef01 mm: add debugfs t... |
2691 2692 |
return 0; } |
a9b0f8618 mm: nominate faul... |
2693 2694 2695 |
DEFINE_SIMPLE_ATTRIBUTE(fault_around_bytes_fops, fault_around_bytes_get, fault_around_bytes_set, "%llu "); |
1592eef01 mm: add debugfs t... |
2696 2697 2698 2699 |
static int __init fault_around_debugfs(void) { void *ret; |
a9b0f8618 mm: nominate faul... |
2700 2701 |
ret = debugfs_create_file("fault_around_bytes", 0644, NULL, NULL, &fault_around_bytes_fops); |
1592eef01 mm: add debugfs t... |
2702 |
if (!ret) |
a9b0f8618 mm: nominate faul... |
2703 |
pr_warn("Failed to create fault_around_bytes in debugfs"); |
1592eef01 mm: add debugfs t... |
2704 2705 2706 |
return 0; } late_initcall(fault_around_debugfs); |
1592eef01 mm: add debugfs t... |
2707 |
#endif |
8c6e50b02 mm: introduce vm_... |
2708 |
|
1fdb412bd mm: document do_f... |
2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 |
/* * do_fault_around() tries to map few pages around the fault address. The hope * is that the pages will be needed soon and this will lower the number of * faults to handle. * * It uses vm_ops->map_pages() to map the pages, which skips the page if it's * not ready to be mapped: not up-to-date, locked, etc. * * This function is called with the page table lock taken. In the split ptlock * case the page table lock only protects only those entries which belong to * the page table corresponding to the fault address. * * This function doesn't cross the VMA boundaries, in order to call map_pages() * only once. * * fault_around_pages() defines how many pages we'll try to map. * do_fault_around() expects it to return a power of two less than or equal to * PTRS_PER_PTE. * * The virtual address of the area that we map is naturally aligned to the * fault_around_pages() value (and therefore to page order). This way it's * easier to guarantee that we don't cross page table boundaries. */ |
8c6e50b02 mm: introduce vm_... |
2732 2733 2734 |
static void do_fault_around(struct vm_area_struct *vma, unsigned long address, pte_t *pte, pgoff_t pgoff, unsigned int flags) { |
aecd6f442 mm: close race be... |
2735 |
unsigned long start_addr, nr_pages, mask; |
8c6e50b02 mm: introduce vm_... |
2736 2737 2738 |
pgoff_t max_pgoff; struct vm_fault vmf; int off; |
aecd6f442 mm: close race be... |
2739 2740 2741 2742 |
nr_pages = ACCESS_ONCE(fault_around_bytes) >> PAGE_SHIFT; mask = ~(nr_pages * PAGE_SIZE - 1) & PAGE_MASK; start_addr = max(address & mask, vma->vm_start); |
8c6e50b02 mm: introduce vm_... |
2743 2744 2745 2746 2747 2748 |
off = ((address - start_addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1); pte -= off; pgoff -= off; /* * max_pgoff is either end of page table or end of vma |
850e9c69c mm: fix typo in c... |
2749 |
* or fault_around_pages() from pgoff, depending what is nearest. |
8c6e50b02 mm: introduce vm_... |
2750 2751 2752 2753 |
*/ max_pgoff = pgoff - ((start_addr >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) + PTRS_PER_PTE - 1; max_pgoff = min3(max_pgoff, vma_pages(vma) + vma->vm_pgoff - 1, |
aecd6f442 mm: close race be... |
2754 |
pgoff + nr_pages - 1); |
8c6e50b02 mm: introduce vm_... |
2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 |
/* Check if it makes any sense to call ->map_pages */ while (!pte_none(*pte)) { if (++pgoff > max_pgoff) return; start_addr += PAGE_SIZE; if (start_addr >= vma->vm_end) return; pte++; } vmf.virtual_address = (void __user *) start_addr; vmf.pte = pte; vmf.pgoff = pgoff; vmf.max_pgoff = max_pgoff; vmf.flags = flags; vma->vm_ops->map_pages(vma, &vmf); } |
e655fb290 mm: introduce do_... |
2773 2774 2775 2776 2777 2778 |
static int do_read_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, pgoff_t pgoff, unsigned int flags, pte_t orig_pte) { struct page *fault_page; spinlock_t *ptl; |
3bb977946 mm: consolidate c... |
2779 |
pte_t *pte; |
8c6e50b02 mm: introduce vm_... |
2780 2781 2782 2783 2784 2785 2786 |
int ret = 0; /* * Let's call ->map_pages() first and use ->fault() as fallback * if page by the offset is not ready to be mapped (cold cache or * something). */ |
c118678bc mm: do not call d... |
2787 |
if (vma->vm_ops->map_pages && !(flags & FAULT_FLAG_NONLINEAR) && |
aecd6f442 mm: close race be... |
2788 |
fault_around_bytes >> PAGE_SHIFT > 1) { |
8c6e50b02 mm: introduce vm_... |
2789 2790 2791 2792 2793 2794 |
pte = pte_offset_map_lock(mm, pmd, address, &ptl); do_fault_around(vma, address, pte, pgoff, flags); if (!pte_same(*pte, orig_pte)) goto unlock_out; pte_unmap_unlock(pte, ptl); } |
e655fb290 mm: introduce do_... |
2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 |
ret = __do_fault(vma, address, pgoff, flags, &fault_page); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) return ret; pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); unlock_page(fault_page); page_cache_release(fault_page); return ret; } |
3bb977946 mm: consolidate c... |
2807 |
do_set_pte(vma, address, fault_page, pte, false, false); |
e655fb290 mm: introduce do_... |
2808 |
unlock_page(fault_page); |
8c6e50b02 mm: introduce vm_... |
2809 2810 |
unlock_out: pte_unmap_unlock(pte, ptl); |
e655fb290 mm: introduce do_... |
2811 2812 |
return ret; } |
ec47c3b95 mm: introduce do_... |
2813 2814 2815 2816 2817 |
static int do_cow_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pmd_t *pmd, pgoff_t pgoff, unsigned int flags, pte_t orig_pte) { struct page *fault_page, *new_page; |
00501b531 mm: memcontrol: r... |
2818 |
struct mem_cgroup *memcg; |
ec47c3b95 mm: introduce do_... |
2819 |
spinlock_t *ptl; |
3bb977946 mm: consolidate c... |
2820 |
pte_t *pte; |
ec47c3b95 mm: introduce do_... |
2821 2822 2823 2824 2825 2826 2827 2828 |
int ret; if (unlikely(anon_vma_prepare(vma))) return VM_FAULT_OOM; new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); if (!new_page) return VM_FAULT_OOM; |
00501b531 mm: memcontrol: r... |
2829 |
if (mem_cgroup_try_charge(new_page, mm, GFP_KERNEL, &memcg)) { |
ec47c3b95 mm: introduce do_... |
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 |
page_cache_release(new_page); return VM_FAULT_OOM; } ret = __do_fault(vma, address, pgoff, flags, &fault_page); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) goto uncharge_out; copy_user_highpage(new_page, fault_page, address, vma); __SetPageUptodate(new_page); pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); unlock_page(fault_page); page_cache_release(fault_page); goto uncharge_out; } |
3bb977946 mm: consolidate c... |
2848 |
do_set_pte(vma, address, new_page, pte, true, true); |
00501b531 mm: memcontrol: r... |
2849 2850 |
mem_cgroup_commit_charge(new_page, memcg, false); lru_cache_add_active_or_unevictable(new_page, vma); |
ec47c3b95 mm: introduce do_... |
2851 2852 2853 2854 2855 |
pte_unmap_unlock(pte, ptl); unlock_page(fault_page); page_cache_release(fault_page); return ret; uncharge_out: |
00501b531 mm: memcontrol: r... |
2856 |
mem_cgroup_cancel_charge(new_page, memcg); |
ec47c3b95 mm: introduce do_... |
2857 2858 2859 |
page_cache_release(new_page); return ret; } |
f0c6d4d29 mm: introduce do_... |
2860 |
static int do_shared_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
16abfa086 Fix sys_remap_fil... |
2861 |
unsigned long address, pmd_t *pmd, |
54cb8821d mm: merge populat... |
2862 |
pgoff_t pgoff, unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
2863 |
{ |
f0c6d4d29 mm: introduce do_... |
2864 2865 |
struct page *fault_page; struct address_space *mapping; |
8f4e2101f [PATCH] mm: page ... |
2866 |
spinlock_t *ptl; |
3bb977946 mm: consolidate c... |
2867 |
pte_t *pte; |
f0c6d4d29 mm: introduce do_... |
2868 |
int dirtied = 0; |
f0c6d4d29 mm: introduce do_... |
2869 |
int ret, tmp; |
1d65f86db mm: preallocate p... |
2870 |
|
7eae74af3 mm: do_fault(): e... |
2871 2872 |
ret = __do_fault(vma, address, pgoff, flags, &fault_page); if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY))) |
f0c6d4d29 mm: introduce do_... |
2873 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2874 2875 |
/* |
f0c6d4d29 mm: introduce do_... |
2876 2877 |
* Check if the backing address space wants to know that the page is * about to become writable |
1da177e4c Linux-2.6.12-rc2 |
2878 |
*/ |
fb09a4642 mm: consolidate c... |
2879 2880 2881 2882 2883 |
if (vma->vm_ops->page_mkwrite) { unlock_page(fault_page); tmp = do_page_mkwrite(vma, fault_page, address); if (unlikely(!tmp || (tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) { |
f0c6d4d29 mm: introduce do_... |
2884 |
page_cache_release(fault_page); |
fb09a4642 mm: consolidate c... |
2885 |
return tmp; |
4294621f4 [PATCH] mm: rss =... |
2886 |
} |
fb09a4642 mm: consolidate c... |
2887 |
} |
f0c6d4d29 mm: introduce do_... |
2888 2889 2890 2891 2892 2893 |
pte = pte_offset_map_lock(mm, pmd, address, &ptl); if (unlikely(!pte_same(*pte, orig_pte))) { pte_unmap_unlock(pte, ptl); unlock_page(fault_page); page_cache_release(fault_page); return ret; |
1da177e4c Linux-2.6.12-rc2 |
2894 |
} |
3bb977946 mm: consolidate c... |
2895 |
do_set_pte(vma, address, fault_page, pte, true, false); |
f0c6d4d29 mm: introduce do_... |
2896 |
pte_unmap_unlock(pte, ptl); |
b827e496c mm: close page_mk... |
2897 |
|
f0c6d4d29 mm: introduce do_... |
2898 2899 |
if (set_page_dirty(fault_page)) dirtied = 1; |
d82fa87d2 mm/memory.c:do_sh... |
2900 2901 2902 2903 2904 2905 |
/* * Take a local copy of the address_space - page.mapping may be zeroed * by truncate after unlock_page(). The address_space itself remains * pinned by vma->vm_file's reference. We rely on unlock_page()'s * release semantics to prevent the compiler from undoing this copying. */ |
f0c6d4d29 mm: introduce do_... |
2906 2907 2908 2909 2910 2911 2912 2913 |
mapping = fault_page->mapping; unlock_page(fault_page); if ((dirtied || vma->vm_ops->page_mkwrite) && mapping) { /* * Some device drivers do not set page.mapping but still * dirty their pages */ balance_dirty_pages_ratelimited(mapping); |
d08b3851d [PATCH] mm: track... |
2914 |
} |
d00806b18 mm: fix fault vs ... |
2915 |
|
f0c6d4d29 mm: introduce do_... |
2916 2917 2918 |
/* file_update_time outside page_lock */ if (vma->vm_file && !vma->vm_ops->page_mkwrite) file_update_time(vma->vm_file); |
b827e496c mm: close page_mk... |
2919 |
|
1d65f86db mm: preallocate p... |
2920 |
return ret; |
54cb8821d mm: merge populat... |
2921 |
} |
d00806b18 mm: fix fault vs ... |
2922 |
|
9a95f3cf7 mm: describe mmap... |
2923 2924 2925 2926 2927 2928 |
/* * We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults). * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). */ |
54cb8821d mm: merge populat... |
2929 2930 |
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... |
2931 |
unsigned int flags, pte_t orig_pte) |
54cb8821d mm: merge populat... |
2932 2933 |
{ pgoff_t pgoff = (((address & PAGE_MASK) |
0da7e01f5 calculation of pg... |
2934 |
- vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; |
54cb8821d mm: merge populat... |
2935 |
|
16abfa086 Fix sys_remap_fil... |
2936 |
pte_unmap(page_table); |
e655fb290 mm: introduce do_... |
2937 2938 2939 |
if (!(flags & FAULT_FLAG_WRITE)) return do_read_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
ec47c3b95 mm: introduce do_... |
2940 2941 2942 |
if (!(vma->vm_flags & VM_SHARED)) return do_cow_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
f0c6d4d29 mm: introduce do_... |
2943 |
return do_shared_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
54cb8821d mm: merge populat... |
2944 |
} |
f4b81804a [PATCH] do_no_pfn() |
2945 |
/* |
1da177e4c Linux-2.6.12-rc2 |
2946 2947 2948 |
* 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 ... |
2949 2950 2951 |
* * We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. |
9a95f3cf7 mm: describe mmap... |
2952 2953 2954 |
* We return with pte unmapped and unlocked. * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). |
1da177e4c Linux-2.6.12-rc2 |
2955 |
*/ |
d0217ac04 mm: fault feedbac... |
2956 |
static int do_nonlinear_fault(struct mm_struct *mm, struct vm_area_struct *vma, |
65500d234 [PATCH] mm: page ... |
2957 |
unsigned long address, pte_t *page_table, pmd_t *pmd, |
30c9f3a9f Remove internal u... |
2958 |
unsigned int flags, pte_t orig_pte) |
1da177e4c Linux-2.6.12-rc2 |
2959 |
{ |
65500d234 [PATCH] mm: page ... |
2960 |
pgoff_t pgoff; |
1da177e4c Linux-2.6.12-rc2 |
2961 |
|
30c9f3a9f Remove internal u... |
2962 |
flags |= FAULT_FLAG_NONLINEAR; |
4c21e2f24 [PATCH] mm: split... |
2963 |
if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) |
83c54070e mm: fault feedbac... |
2964 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
2965 |
|
2509ef26d badpage: zap prin... |
2966 |
if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) { |
65500d234 [PATCH] mm: page ... |
2967 2968 2969 |
/* * Page table corrupted: show pte and kill process. */ |
3dc147414 badpage: replace ... |
2970 |
print_bad_pte(vma, address, orig_pte, NULL); |
d99be1a8e mm: sigbus instea... |
2971 |
return VM_FAULT_SIGBUS; |
65500d234 [PATCH] mm: page ... |
2972 |
} |
65500d234 [PATCH] mm: page ... |
2973 2974 |
pgoff = pte_to_pgoff(orig_pte); |
e655fb290 mm: introduce do_... |
2975 2976 2977 |
if (!(flags & FAULT_FLAG_WRITE)) return do_read_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
ec47c3b95 mm: introduce do_... |
2978 2979 2980 |
if (!(vma->vm_flags & VM_SHARED)) return do_cow_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
f0c6d4d29 mm: introduce do_... |
2981 |
return do_shared_fault(mm, vma, address, pmd, pgoff, flags, orig_pte); |
1da177e4c Linux-2.6.12-rc2 |
2982 |
} |
b19a99392 mm/memory.c: mark... |
2983 |
static int numa_migrate_prep(struct page *page, struct vm_area_struct *vma, |
04bb2f947 sched/numa: Adjus... |
2984 2985 |
unsigned long addr, int page_nid, int *flags) |
9532fec11 mm: numa: Migrate... |
2986 2987 2988 2989 |
{ get_page(page); count_vm_numa_event(NUMA_HINT_FAULTS); |
04bb2f947 sched/numa: Adjus... |
2990 |
if (page_nid == numa_node_id()) { |
9532fec11 mm: numa: Migrate... |
2991 |
count_vm_numa_event(NUMA_HINT_FAULTS_LOCAL); |
04bb2f947 sched/numa: Adjus... |
2992 2993 |
*flags |= TNF_FAULT_LOCAL; } |
9532fec11 mm: numa: Migrate... |
2994 2995 2996 |
return mpol_misplaced(page, vma, addr); } |
b19a99392 mm/memory.c: mark... |
2997 |
static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, |
d10e63f29 mm: numa: Create ... |
2998 2999 |
unsigned long addr, pte_t pte, pte_t *ptep, pmd_t *pmd) { |
4daae3b4b mm: mempolicy: Us... |
3000 |
struct page *page = NULL; |
d10e63f29 mm: numa: Create ... |
3001 |
spinlock_t *ptl; |
8191acbd3 mm: numa: Sanitiz... |
3002 |
int page_nid = -1; |
90572890d mm: numa: Change ... |
3003 |
int last_cpupid; |
cbee9f88e mm: numa: Add fau... |
3004 |
int target_nid; |
b8593bfda mm: sched: Adapt ... |
3005 |
bool migrated = false; |
6688cc054 mm: numa: Do not ... |
3006 |
int flags = 0; |
d10e63f29 mm: numa: Create ... |
3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 |
/* * The "pte" at this point cannot be used safely without * validation through pte_unmap_same(). It's of NUMA type but * the pfn may be screwed if the read is non atomic. * * ptep_modify_prot_start is not called as this is clearing * the _PAGE_NUMA bit and it is not really expected that there * would be concurrent hardware modifications to the PTE. */ ptl = pte_lockptr(mm, pmd); spin_lock(ptl); |
4daae3b4b mm: mempolicy: Us... |
3019 3020 3021 3022 |
if (unlikely(!pte_same(*ptep, pte))) { pte_unmap_unlock(ptep, ptl); goto out; } |
d10e63f29 mm: numa: Create ... |
3023 3024 3025 3026 3027 3028 3029 3030 3031 |
pte = pte_mknonnuma(pte); set_pte_at(mm, addr, ptep, pte); update_mmu_cache(vma, addr, ptep); page = vm_normal_page(vma, addr, pte); if (!page) { pte_unmap_unlock(ptep, ptl); return 0; } |
a1a46184e mm: numa: Do not ... |
3032 |
BUG_ON(is_zero_pfn(page_to_pfn(page))); |
d10e63f29 mm: numa: Create ... |
3033 |
|
6688cc054 mm: numa: Do not ... |
3034 3035 3036 3037 3038 3039 3040 |
/* * Avoid grouping on DSO/COW pages in specific and RO pages * in general, RO pages shouldn't hurt as much anyway since * they can be in shared cache state. */ if (!pte_write(pte)) flags |= TNF_NO_GROUP; |
dabe1d992 sched/numa: Be mo... |
3041 3042 3043 3044 3045 3046 |
/* * Flag if the page is shared between multiple address spaces. This * is later used when determining whether to group tasks together */ if (page_mapcount(page) > 1 && (vma->vm_flags & VM_SHARED)) flags |= TNF_SHARED; |
90572890d mm: numa: Change ... |
3047 |
last_cpupid = page_cpupid_last(page); |
8191acbd3 mm: numa: Sanitiz... |
3048 |
page_nid = page_to_nid(page); |
04bb2f947 sched/numa: Adjus... |
3049 |
target_nid = numa_migrate_prep(page, vma, addr, page_nid, &flags); |
d10e63f29 mm: numa: Create ... |
3050 |
pte_unmap_unlock(ptep, ptl); |
4daae3b4b mm: mempolicy: Us... |
3051 |
if (target_nid == -1) { |
4daae3b4b mm: mempolicy: Us... |
3052 3053 3054 3055 3056 |
put_page(page); goto out; } /* Migrate to the requested node */ |
1bc115d87 mm: numa: Scan pa... |
3057 |
migrated = migrate_misplaced_page(page, vma, target_nid); |
6688cc054 mm: numa: Do not ... |
3058 |
if (migrated) { |
8191acbd3 mm: numa: Sanitiz... |
3059 |
page_nid = target_nid; |
6688cc054 mm: numa: Do not ... |
3060 3061 |
flags |= TNF_MIGRATED; } |
4daae3b4b mm: mempolicy: Us... |
3062 3063 |
out: |
8191acbd3 mm: numa: Sanitiz... |
3064 |
if (page_nid != -1) |
6688cc054 mm: numa: Do not ... |
3065 |
task_numa_fault(last_cpupid, page_nid, 1, flags); |
d10e63f29 mm: numa: Create ... |
3066 3067 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
3068 3069 3070 3071 3072 3073 3074 3075 3076 |
/* * 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... |
3077 3078 |
* We enter with non-exclusive mmap_sem (to exclude vma changes, * but allow concurrent faults), and pte mapped but not yet locked. |
9a95f3cf7 mm: describe mmap... |
3079 3080 3081 3082 |
* We return with pte unmapped and unlocked. * * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). |
1da177e4c Linux-2.6.12-rc2 |
3083 |
*/ |
c02925540 thp: consolidate ... |
3084 |
static int handle_pte_fault(struct mm_struct *mm, |
71e3aac07 thp: transparent ... |
3085 3086 |
struct vm_area_struct *vma, unsigned long address, pte_t *pte, pmd_t *pmd, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3087 3088 |
{ pte_t entry; |
8f4e2101f [PATCH] mm: page ... |
3089 |
spinlock_t *ptl; |
1da177e4c Linux-2.6.12-rc2 |
3090 |
|
e37c69827 mm: replace ACCES... |
3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 |
/* * some architectures can have larger ptes than wordsize, * e.g.ppc44x-defconfig has CONFIG_PTE_64BIT=y and CONFIG_32BIT=y, * so READ_ONCE or ACCESS_ONCE cannot guarantee atomic accesses. * The code below just needs a consistent view for the ifs and * we later double check anyway with the ptl lock held. So here * a barrier will do. */ entry = *pte; barrier(); |
1da177e4c Linux-2.6.12-rc2 |
3101 |
if (!pte_present(entry)) { |
65500d234 [PATCH] mm: page ... |
3102 |
if (pte_none(entry)) { |
f4b81804a [PATCH] do_no_pfn() |
3103 |
if (vma->vm_ops) { |
3c18ddd16 mm: remove nopage |
3104 |
if (likely(vma->vm_ops->fault)) |
54cb8821d mm: merge populat... |
3105 |
return do_linear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3106 |
pte, pmd, flags, entry); |
f4b81804a [PATCH] do_no_pfn() |
3107 3108 |
} return do_anonymous_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3109 |
pte, pmd, flags); |
65500d234 [PATCH] mm: page ... |
3110 |
} |
1da177e4c Linux-2.6.12-rc2 |
3111 |
if (pte_file(entry)) |
d0217ac04 mm: fault feedbac... |
3112 |
return do_nonlinear_fault(mm, vma, address, |
30c9f3a9f Remove internal u... |
3113 |
pte, pmd, flags, entry); |
65500d234 [PATCH] mm: page ... |
3114 |
return do_swap_page(mm, vma, address, |
30c9f3a9f Remove internal u... |
3115 |
pte, pmd, flags, entry); |
1da177e4c Linux-2.6.12-rc2 |
3116 |
} |
d10e63f29 mm: numa: Create ... |
3117 3118 |
if (pte_numa(entry)) return do_numa_page(mm, vma, address, entry, pte, pmd); |
4c21e2f24 [PATCH] mm: split... |
3119 |
ptl = pte_lockptr(mm, pmd); |
8f4e2101f [PATCH] mm: page ... |
3120 3121 3122 |
spin_lock(ptl); if (unlikely(!pte_same(*pte, entry))) goto unlock; |
30c9f3a9f Remove internal u... |
3123 |
if (flags & FAULT_FLAG_WRITE) { |
1da177e4c Linux-2.6.12-rc2 |
3124 |
if (!pte_write(entry)) |
8f4e2101f [PATCH] mm: page ... |
3125 3126 |
return do_wp_page(mm, vma, address, pte, pmd, ptl, entry); |
1da177e4c Linux-2.6.12-rc2 |
3127 3128 3129 |
entry = pte_mkdirty(entry); } entry = pte_mkyoung(entry); |
30c9f3a9f Remove internal u... |
3130 |
if (ptep_set_access_flags(vma, address, pte, entry, flags & FAULT_FLAG_WRITE)) { |
4b3073e1c MM: Pass a PTE po... |
3131 |
update_mmu_cache(vma, address, pte); |
1a44e1490 [PATCH] .text pag... |
3132 3133 3134 3135 3136 3137 3138 |
} 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... |
3139 |
if (flags & FAULT_FLAG_WRITE) |
61c77326d x86, mm: Avoid un... |
3140 |
flush_tlb_fix_spurious_fault(vma, address); |
1a44e1490 [PATCH] .text pag... |
3141 |
} |
8f4e2101f [PATCH] mm: page ... |
3142 3143 |
unlock: pte_unmap_unlock(pte, ptl); |
83c54070e mm: fault feedbac... |
3144 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3145 3146 3147 3148 |
} /* * By the time we get here, we already hold the mm semaphore |
9a95f3cf7 mm: describe mmap... |
3149 3150 3151 |
* * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). |
1da177e4c Linux-2.6.12-rc2 |
3152 |
*/ |
519e52473 mm: memcg: enable... |
3153 3154 |
static int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, unsigned int flags) |
1da177e4c Linux-2.6.12-rc2 |
3155 3156 3157 3158 3159 |
{ pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; |
ac9b9c667 [PATCH] Fix handl... |
3160 |
if (unlikely(is_vm_hugetlb_page(vma))) |
30c9f3a9f Remove internal u... |
3161 |
return hugetlb_fault(mm, vma, address, flags); |
1da177e4c Linux-2.6.12-rc2 |
3162 |
|
1da177e4c Linux-2.6.12-rc2 |
3163 |
pgd = pgd_offset(mm, address); |
1da177e4c Linux-2.6.12-rc2 |
3164 3165 |
pud = pud_alloc(mm, pgd, address); if (!pud) |
c74df32c7 [PATCH] mm: ptd_a... |
3166 |
return VM_FAULT_OOM; |
1da177e4c Linux-2.6.12-rc2 |
3167 3168 |
pmd = pmd_alloc(mm, pud, address); if (!pmd) |
c74df32c7 [PATCH] mm: ptd_a... |
3169 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3170 |
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma)) { |
c02925540 thp: consolidate ... |
3171 |
int ret = VM_FAULT_FALLBACK; |
71e3aac07 thp: transparent ... |
3172 |
if (!vma->vm_ops) |
c02925540 thp: consolidate ... |
3173 3174 3175 3176 |
ret = do_huge_pmd_anonymous_page(mm, vma, address, pmd, flags); if (!(ret & VM_FAULT_FALLBACK)) return ret; |
71e3aac07 thp: transparent ... |
3177 3178 |
} else { pmd_t orig_pmd = *pmd; |
1f1d06c34 thp, memcg: split... |
3179 |
int ret; |
71e3aac07 thp: transparent ... |
3180 3181 |
barrier(); if (pmd_trans_huge(orig_pmd)) { |
a1dd450bc mm: thp: set the ... |
3182 |
unsigned int dirty = flags & FAULT_FLAG_WRITE; |
e53289c0c mm: reinstante dr... |
3183 3184 3185 3186 3187 3188 3189 |
/* * If the pmd is splitting, return and retry the * the fault. Alternative: wait until the split * is done, and goto retry. */ if (pmd_trans_splitting(orig_pmd)) return 0; |
3d59eebc5 Merge tag 'balanc... |
3190 |
if (pmd_numa(orig_pmd)) |
4daae3b4b mm: mempolicy: Us... |
3191 |
return do_huge_pmd_numa_page(mm, vma, address, |
d10e63f29 mm: numa: Create ... |
3192 |
orig_pmd, pmd); |
3d59eebc5 Merge tag 'balanc... |
3193 |
if (dirty && !pmd_write(orig_pmd)) { |
1f1d06c34 thp, memcg: split... |
3194 3195 |
ret = do_huge_pmd_wp_page(mm, vma, address, pmd, orig_pmd); |
9845cbbd1 mm, thp: fix infi... |
3196 3197 |
if (!(ret & VM_FAULT_FALLBACK)) return ret; |
a1dd450bc mm: thp: set the ... |
3198 3199 3200 |
} else { huge_pmd_set_accessed(mm, vma, address, pmd, orig_pmd, dirty); |
9845cbbd1 mm, thp: fix infi... |
3201 |
return 0; |
1f1d06c34 thp, memcg: split... |
3202 |
} |
71e3aac07 thp: transparent ... |
3203 3204 3205 3206 3207 3208 3209 3210 |
} } /* * Use __pte_alloc instead of pte_alloc_map, because we can't * run pte_offset_map on the pmd, if an huge pmd could * materialize from under us from a different thread. */ |
4fd017708 mm: Check if PTE ... |
3211 3212 |
if (unlikely(pmd_none(*pmd)) && unlikely(__pte_alloc(mm, vma, pmd, address))) |
c74df32c7 [PATCH] mm: ptd_a... |
3213 |
return VM_FAULT_OOM; |
71e3aac07 thp: transparent ... |
3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 |
/* 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 |
3224 |
|
30c9f3a9f Remove internal u... |
3225 |
return handle_pte_fault(mm, vma, address, pte, pmd, flags); |
1da177e4c Linux-2.6.12-rc2 |
3226 |
} |
9a95f3cf7 mm: describe mmap... |
3227 3228 3229 3230 3231 3232 |
/* * By the time we get here, we already hold the mm semaphore * * The mmap_sem may have been released depending on flags and our * return value. See filemap_fault() and __lock_page_or_retry(). */ |
519e52473 mm: memcg: enable... |
3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 |
int handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, unsigned int flags) { int ret; __set_current_state(TASK_RUNNING); count_vm_event(PGFAULT); mem_cgroup_count_vm_event(mm, PGFAULT); /* do counter updates before entering really critical section. */ check_sync_rss_stat(current); /* * Enable the memcg OOM handling for faults triggered in user * space. Kernel faults are handled more gracefully. */ if (flags & FAULT_FLAG_USER) |
494264208 mm: memcg: handle... |
3251 |
mem_cgroup_oom_enable(); |
519e52473 mm: memcg: enable... |
3252 3253 |
ret = __handle_mm_fault(mm, vma, address, flags); |
494264208 mm: memcg: handle... |
3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 |
if (flags & FAULT_FLAG_USER) { mem_cgroup_oom_disable(); /* * The task may have entered a memcg OOM situation but * if the allocation error was handled gracefully (no * VM_FAULT_OOM), there is no need to kill anything. * Just clean up the OOM state peacefully. */ if (task_in_memcg_oom(current) && !(ret & VM_FAULT_OOM)) mem_cgroup_oom_synchronize(false); } |
3812c8c8f mm: memcg: do not... |
3265 |
|
519e52473 mm: memcg: enable... |
3266 3267 |
return ret; } |
e1d6d01ab mm: export find_e... |
3268 |
EXPORT_SYMBOL_GPL(handle_mm_fault); |
519e52473 mm: memcg: enable... |
3269 |
|
1da177e4c Linux-2.6.12-rc2 |
3270 3271 3272 |
#ifndef __PAGETABLE_PUD_FOLDED /* * Allocate page upper directory. |
872fec16d [PATCH] mm: init_... |
3273 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3274 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3275 |
int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3276 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3277 3278 |
pud_t *new = pud_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3279 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3280 |
|
362a61ad6 fix SMP data race... |
3281 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3282 |
spin_lock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3283 |
if (pgd_present(*pgd)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3284 |
pud_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3285 3286 |
else pgd_populate(mm, pgd, new); |
c74df32c7 [PATCH] mm: ptd_a... |
3287 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3288 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
3289 3290 3291 3292 3293 3294 |
} #endif /* __PAGETABLE_PUD_FOLDED */ #ifndef __PAGETABLE_PMD_FOLDED /* * Allocate page middle directory. |
872fec16d [PATCH] mm: init_... |
3295 |
* We've already handled the fast-path in-line. |
1da177e4c Linux-2.6.12-rc2 |
3296 |
*/ |
1bb3630e8 [PATCH] mm: ptd_a... |
3297 |
int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) |
1da177e4c Linux-2.6.12-rc2 |
3298 |
{ |
c74df32c7 [PATCH] mm: ptd_a... |
3299 3300 |
pmd_t *new = pmd_alloc_one(mm, address); if (!new) |
1bb3630e8 [PATCH] mm: ptd_a... |
3301 |
return -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
3302 |
|
362a61ad6 fix SMP data race... |
3303 |
smp_wmb(); /* See comment in __pte_alloc */ |
872fec16d [PATCH] mm: init_... |
3304 |
spin_lock(&mm->page_table_lock); |
1da177e4c Linux-2.6.12-rc2 |
3305 |
#ifndef __ARCH_HAS_4LEVEL_HACK |
1bb3630e8 [PATCH] mm: ptd_a... |
3306 |
if (pud_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3307 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3308 3309 |
else pud_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3310 |
#else |
1bb3630e8 [PATCH] mm: ptd_a... |
3311 |
if (pgd_present(*pud)) /* Another has populated it */ |
5e5419734 add mm argument t... |
3312 |
pmd_free(mm, new); |
1bb3630e8 [PATCH] mm: ptd_a... |
3313 3314 |
else pgd_populate(mm, pud, new); |
1da177e4c Linux-2.6.12-rc2 |
3315 |
#endif /* __ARCH_HAS_4LEVEL_HACK */ |
c74df32c7 [PATCH] mm: ptd_a... |
3316 |
spin_unlock(&mm->page_table_lock); |
1bb3630e8 [PATCH] mm: ptd_a... |
3317 |
return 0; |
e0f39591c [PATCH] Workaroun... |
3318 |
} |
1da177e4c Linux-2.6.12-rc2 |
3319 |
#endif /* __PAGETABLE_PMD_FOLDED */ |
1b36ba815 mm: wrap follow_p... |
3320 |
static int __follow_pte(struct mm_struct *mm, unsigned long address, |
f8ad0f499 mm: introduce fol... |
3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 |
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... |
3337 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
f8ad0f499 mm: introduce fol... |
3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 |
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... |
3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 |
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... |
3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 |
/** * 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... |
3395 |
#ifdef CONFIG_HAVE_IOREMAP_PROT |
d87fe6607 x86: PAT: modify ... |
3396 3397 3398 |
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... |
3399 |
{ |
03668a4de mm: use generic f... |
3400 |
int ret = -EINVAL; |
28b2ee20c access_process_vm... |
3401 3402 |
pte_t *ptep, pte; spinlock_t *ptl; |
28b2ee20c access_process_vm... |
3403 |
|
d87fe6607 x86: PAT: modify ... |
3404 3405 |
if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) goto out; |
28b2ee20c access_process_vm... |
3406 |
|
03668a4de mm: use generic f... |
3407 |
if (follow_pte(vma->vm_mm, address, &ptep, &ptl)) |
d87fe6607 x86: PAT: modify ... |
3408 |
goto out; |
28b2ee20c access_process_vm... |
3409 |
pte = *ptep; |
03668a4de mm: use generic f... |
3410 |
|
28b2ee20c access_process_vm... |
3411 3412 |
if ((flags & FOLL_WRITE) && !pte_write(pte)) goto unlock; |
28b2ee20c access_process_vm... |
3413 3414 |
*prot = pgprot_val(pte_pgprot(pte)); |
03668a4de mm: use generic f... |
3415 |
*phys = (resource_size_t)pte_pfn(pte) << PAGE_SHIFT; |
28b2ee20c access_process_vm... |
3416 |
|
03668a4de mm: use generic f... |
3417 |
ret = 0; |
28b2ee20c access_process_vm... |
3418 3419 3420 |
unlock: pte_unmap_unlock(ptep, ptl); out: |
d87fe6607 x86: PAT: modify ... |
3421 |
return ret; |
28b2ee20c access_process_vm... |
3422 3423 3424 3425 3426 3427 3428 |
} 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 __... |
3429 |
void __iomem *maddr; |
28b2ee20c access_process_vm... |
3430 |
int offset = addr & (PAGE_SIZE-1); |
d87fe6607 x86: PAT: modify ... |
3431 |
if (follow_phys(vma, addr, write, &prot, &phys_addr)) |
28b2ee20c access_process_vm... |
3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 |
return -EINVAL; maddr = ioremap_prot(phys_addr, PAGE_SIZE, prot); if (write) memcpy_toio(maddr + offset, buf, len); else memcpy_fromio(buf, maddr + offset, len); iounmap(maddr); return len; } |
5a73633ef mm: make generic_... |
3443 |
EXPORT_SYMBOL_GPL(generic_access_phys); |
28b2ee20c access_process_vm... |
3444 |
#endif |
0ec76a110 [PATCH] NOMMU: Ch... |
3445 |
/* |
206cb6365 mm: factor out ma... |
3446 3447 |
* Access another process' address space as given in mm. If non-NULL, use the * given task for page fault accounting. |
0ec76a110 [PATCH] NOMMU: Ch... |
3448 |
*/ |
206cb6365 mm: factor out ma... |
3449 3450 |
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... |
3451 |
{ |
0ec76a110 [PATCH] NOMMU: Ch... |
3452 |
struct vm_area_struct *vma; |
0ec76a110 [PATCH] NOMMU: Ch... |
3453 |
void *old_buf = buf; |
0ec76a110 [PATCH] NOMMU: Ch... |
3454 |
down_read(&mm->mmap_sem); |
183ff22bb spelling fixes: mm/ |
3455 |
/* ignore errors, just check how much was successfully transferred */ |
0ec76a110 [PATCH] NOMMU: Ch... |
3456 3457 3458 |
while (len) { int bytes, ret, offset; void *maddr; |
28b2ee20c access_process_vm... |
3459 |
struct page *page = NULL; |
0ec76a110 [PATCH] NOMMU: Ch... |
3460 3461 3462 |
ret = get_user_pages(tsk, mm, addr, 1, write, 1, &page, &vma); |
28b2ee20c access_process_vm... |
3463 |
if (ret <= 0) { |
dbffcd03d mm: change confus... |
3464 3465 3466 |
#ifndef CONFIG_HAVE_IOREMAP_PROT break; #else |
28b2ee20c access_process_vm... |
3467 3468 3469 3470 |
/* * Check if this is a VM_IO | VM_PFNMAP VMA, which * we can access using slightly different code. */ |
28b2ee20c access_process_vm... |
3471 |
vma = find_vma(mm, addr); |
fe936dfc2 mm: check that we... |
3472 |
if (!vma || vma->vm_start > addr) |
28b2ee20c access_process_vm... |
3473 3474 3475 3476 3477 |
break; if (vma->vm_ops && vma->vm_ops->access) ret = vma->vm_ops->access(vma, addr, buf, len, write); if (ret <= 0) |
28b2ee20c access_process_vm... |
3478 3479 |
break; bytes = ret; |
dbffcd03d mm: change confus... |
3480 |
#endif |
0ec76a110 [PATCH] NOMMU: Ch... |
3481 |
} else { |
28b2ee20c access_process_vm... |
3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 |
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... |
3498 |
} |
0ec76a110 [PATCH] NOMMU: Ch... |
3499 3500 3501 3502 3503 |
len -= bytes; buf += bytes; addr += bytes; } up_read(&mm->mmap_sem); |
0ec76a110 [PATCH] NOMMU: Ch... |
3504 3505 3506 |
return buf - old_buf; } |
03252919b x86: print which ... |
3507 |
|
5ddd36b9c mm: implement acc... |
3508 |
/** |
ae91dbfc9 mm: fix memory.c ... |
3509 |
* access_remote_vm - access another process' address space |
5ddd36b9c mm: implement acc... |
3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 |
* @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 ... |
3523 |
/* |
206cb6365 mm: factor out ma... |
3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 |
* 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 ... |
3543 3544 3545 3546 3547 3548 3549 |
/* * 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... |
3550 3551 3552 3553 3554 3555 |
/* * Do not print if we are in atomic * contexts (in exception stacks, etc.): */ if (preempt_count()) return; |
03252919b x86: print which ... |
3556 3557 3558 3559 3560 3561 |
down_read(&mm->mmap_sem); vma = find_vma(mm, ip); if (vma && vma->vm_file) { struct file *f = vma->vm_file; char *buf = (char *)__get_free_page(GFP_KERNEL); if (buf) { |
2fbc57c53 mm: use kbasename() |
3562 |
char *p; |
03252919b x86: print which ... |
3563 |
|
cf28b4863 d_path: Make d_pa... |
3564 |
p = d_path(&f->f_path, buf, PAGE_SIZE); |
03252919b x86: print which ... |
3565 3566 |
if (IS_ERR(p)) p = "?"; |
2fbc57c53 mm: use kbasename() |
3567 |
printk("%s%s[%lx+%lx]", prefix, kbasename(p), |
03252919b x86: print which ... |
3568 3569 3570 3571 3572 |
vma->vm_start, vma->vm_end - vma->vm_start); free_page((unsigned long)buf); } } |
51a07e50b mm/memory.c:print... |
3573 |
up_read(&mm->mmap_sem); |
03252919b x86: print which ... |
3574 |
} |
3ee1afa30 x86: some lock an... |
3575 |
|
662bbcb27 mm, sched: Allow ... |
3576 |
#if defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_DEBUG_ATOMIC_SLEEP) |
3ee1afa30 x86: some lock an... |
3577 3578 |
void might_fault(void) { |
95156f005 lockdep, mm: fix ... |
3579 3580 3581 3582 3583 3584 3585 3586 |
/* * 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... |
3587 3588 3589 3590 3591 |
/* * it would be nicer only to annotate paths which are not under * pagefault_disable, however that requires a larger audit and * providing helpers like get_user_atomic. */ |
662bbcb27 mm, sched: Allow ... |
3592 3593 3594 3595 3596 3597 |
if (in_atomic()) return; __might_sleep(__FILE__, __LINE__, 0); if (current->mm) |
3ee1afa30 x86: some lock an... |
3598 3599 3600 3601 |
might_lock_read(¤t->mm->mmap_sem); } EXPORT_SYMBOL(might_fault); #endif |
47ad8475c thp: clear_copy_h... |
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 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 |
#if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_HUGETLBFS) static void clear_gigantic_page(struct page *page, unsigned long addr, unsigned int pages_per_huge_page) { int i; struct page *p = page; might_sleep(); for (i = 0; i < pages_per_huge_page; i++, p = mem_map_next(p, page, i)) { cond_resched(); clear_user_highpage(p, addr + i * PAGE_SIZE); } } void clear_huge_page(struct page *page, unsigned long addr, unsigned int pages_per_huge_page) { int i; if (unlikely(pages_per_huge_page > MAX_ORDER_NR_PAGES)) { clear_gigantic_page(page, addr, pages_per_huge_page); return; } might_sleep(); for (i = 0; i < pages_per_huge_page; i++) { cond_resched(); clear_user_highpage(page + i, addr + i * PAGE_SIZE); } } static void copy_user_gigantic_page(struct page *dst, struct page *src, unsigned long addr, struct vm_area_struct *vma, unsigned int pages_per_huge_page) { int i; struct page *dst_base = dst; struct page *src_base = src; for (i = 0; i < pages_per_huge_page; ) { cond_resched(); copy_user_highpage(dst, src, addr + i*PAGE_SIZE, vma); i++; dst = mem_map_next(dst, dst_base, i); src = mem_map_next(src, src_base, i); } } void copy_user_huge_page(struct page *dst, struct page *src, unsigned long addr, struct vm_area_struct *vma, unsigned int pages_per_huge_page) { int i; if (unlikely(pages_per_huge_page > MAX_ORDER_NR_PAGES)) { copy_user_gigantic_page(dst, src, addr, vma, pages_per_huge_page); return; } might_sleep(); for (i = 0; i < pages_per_huge_page; i++) { cond_resched(); copy_user_highpage(dst + i, src + i, addr + i*PAGE_SIZE, vma); } } #endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */ |
49076ec2c mm: dynamically a... |
3673 |
|
40b64acd1 mm: fix build of ... |
3674 |
#if USE_SPLIT_PTE_PTLOCKS && ALLOC_SPLIT_PTLOCKS |
b35f1819a mm: create a sepa... |
3675 3676 3677 3678 3679 3680 3681 3682 |
static struct kmem_cache *page_ptl_cachep; void __init ptlock_cache_init(void) { page_ptl_cachep = kmem_cache_create("page->ptl", sizeof(spinlock_t), 0, SLAB_PANIC, NULL); } |
539edb584 mm: properly sepa... |
3683 |
bool ptlock_alloc(struct page *page) |
49076ec2c mm: dynamically a... |
3684 3685 |
{ spinlock_t *ptl; |
b35f1819a mm: create a sepa... |
3686 |
ptl = kmem_cache_alloc(page_ptl_cachep, GFP_KERNEL); |
49076ec2c mm: dynamically a... |
3687 3688 |
if (!ptl) return false; |
539edb584 mm: properly sepa... |
3689 |
page->ptl = ptl; |
49076ec2c mm: dynamically a... |
3690 3691 |
return true; } |
539edb584 mm: properly sepa... |
3692 |
void ptlock_free(struct page *page) |
49076ec2c mm: dynamically a... |
3693 |
{ |
b35f1819a mm: create a sepa... |
3694 |
kmem_cache_free(page_ptl_cachep, page->ptl); |
49076ec2c mm: dynamically a... |
3695 3696 |
} #endif |